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 |
|---|---|---|---|---|---|
Computational procedures for predicting metabolic interventions leading to the overproduction of biochemicals in microbial strains are widely in use . However , these methods rely on surrogate biological objectives ( e . g . , maximize growth rate or minimize metabolic adjustments ) and do not make use of flux measurements often available for the wild-type strain . In this work , we introduce the OptForce procedure that identifies all possible engineering interventions by classifying reactions in the metabolic model depending upon whether their flux values must increase , decrease or become equal to zero to meet a pre-specified overproduction target . We hierarchically apply this classification rule for pairs , triples , quadruples , etc . of reactions . This leads to the identification of a sufficient and non-redundant set of fluxes that must change ( i . e . , MUST set ) to meet a pre-specified overproduction target . Starting with this set we subsequently extract a minimal set of fluxes that must actively be forced through genetic manipulations ( i . e . , FORCE set ) to ensure that all fluxes in the network are consistent with the overproduction objective . We demonstrate our OptForce framework for succinate production in Escherichia coli using the most recent in silico E . coli model , iAF1260 . The method not only recapitulates existing engineering strategies but also reveals non-intuitive ones that boost succinate production by performing coordinated changes on pathways distant from the last steps of succinate synthesis .
An overarching challenge for metabolic engineers is to optimize the conversion of biomass and other renewable resources into useful metabolic products through fermentation and other biological conversions [1] , [2] . Metabolic reaction fluxes are a fundamental determinant of the cell physiology , primarily because they provide a degree of engagement of various pathways in metabolic processes [3] . Earlier efforts addressed parts of metabolism with an emphasis on dynamics using kinetic approximations of reaction rates [4]–[7] . These approximations included the popular the S-system representation [4] , [8]–[12] and Michaelis-Menten based descriptions [13]–[15] . Despite many success stories , it is increasingly becoming accepted that strain optimization requires taking account of the totality of biotransformations present in a production strain . This global view of metabolism is needed to enable the complete elucidation of all carbon fluxes diverted away from the desired product , diagnose unbalanced cofactor requirements limiting the extent of reactions as well as remedy deficiencies in the production of all biomass components leading to growth arrest . Flux balance analysis ( FBA ) has emerged as an important framework [16]–[19] to assess the metabolic potential of a microbial production system . By taking a complete inventory of all ( known ) metabolic capabilities of an organism , FBA can assess the maximum possible yield of a desired product for different substrates and growth levels [20] . Given the lack of a truly predictive nature , FBA results must be carefully interpreted as performance limits and supplemented with MFA data whenever possible . Shortly after the introduction of FBA , a number of computational tools emerged that identified strain engineering modifications leading to targeted overproductions . One of the earliest efforts was the OptKnock [21] procedure that suggested gene knockouts leading to targeted overproductions . A bilevel optimization framework was postulated that computationally coupled the desired overproduction target to growth with unforeseen , at the time , implications for strain stability . Later , OptReg [22] extended OptKnock to consider not only knockouts but also overexpressions and down regulations of various reactions in the network . In addition , OptStrain [23] allowed for knock-ins of non-native functionalities from a comprehensive universal database of reactions to enable production of desired biochemicals . Evolutionary search procedures for solving the resulting combinatorial optimization problems were explored in OptGene [24] and applied for the production of succinic acid , glycerol and vanillin in yeast . The Ensemble Modeling approach [25] circumvented the kinetic modeling approach by incorporating flux measurements from knockout and enzyme overexpression experiments . Recently , the GDLS algorithm [26] was used for reduced metabolic models employing GPR associations to predict gene knockouts for succinate and acetate production in E . coli . So far , computational strain design procedures have been applied for a variety of metabolic engineering projects including the overproduction of lactic acid [21] , [27] , succinate [24] , [28]–[31] , 1 , 3-propanediol [21] , hydrogen [23] , amino acids [32] , L-lysine [33] , L-valine [34] , threonine [35] , lycopene [36] , [37] , ethanol in E . coli [22] , [38] , [39] and Saccharomyces cerevisiae [40] and bioelectricity in Geobacter sulfurreducens [41] . The use of computational tools operating on metabolic reconstructions to identify strain modifications is becoming commonplace . Nevertheless , a number of shortcomings plague all existing approaches . All are sequential in nature generating a single engineering strategy per run thus requiring multiple restarts to generate a set of candidate list of alternatives ( i . e . , typically less than ten ) that is dwarfed by the myriads of engineering possibilities afforded by genome-scale models spanning thousands of reactions . Furthermore , in the absence of kinetic descriptions OptKnock and other methods rely on the maximization of surrogate biological fitness functions ( e . g . maximization of biomass yield [21] or minimization of metabolic adjustments MOMA [42] ) to estimate flux redirection upon strain engineering . These estimates may or may not be an accurate representation of how metabolism responds to genetic or environmental perturbations with significant consequences in the quality of the suggested re-designs . Existing methods do not pro-actively make use of flux measurements for the wild-type and/or an engineered strain to identify which fluxes need to be actively engineered in response to a production target . To remedy these limitations , we introduce a new computational framework termed OptForce that identifies all possible engineering interventions for a wild-type strain characterized by specific metabolic flux data consistent with an imposed production target ( s ) .
The key concept of OptForce is to maximally resolve which fluxes ( or combinations thereof ) must depart away from the range of values allowed to span in the wild-type strain in response to an overproduction target . This maximal range of flux variability for the wild-type strain can be elucidated by iteratively maximizing and minimizing each flux [20] , [43] subject to the stoichiometric constraints , uptake conditions and MFA flux data ( either exact values or ranges ) whenever available for the wild-type strain . This yields a set of lower and upper bounds for every flux in the metabolic network . Narrow ranges for the bounds are indicative of fluxes whose value is well bracketed given the information available for the wild-type strain whereas wide ranges indicate fluxes that are not significantly limited by the imposed ( stoichiometric , MFA , etc . ) constraints . Flux ranges can be used not only for characterizing the metabolic flux limits of the wild-type strain but also for identifying all flux combinations consistent with a single ( i . e . , v>vtarget ) or multiple desired overproduction targets ( see Appendix A of Text S1 for optimization formulations ) . The flux ranges consistent with the overproduction target ( s ) can be derived as before by iteratively maximizing and minimizing every flux in the metabolic network subject to stoichiometric constraints , uptake conditions and overproduction targets . Contrasting the flux ranges for the ( wild-type ) metabolic network against the ones consistent with the overproduction target ( s ) provides the cornerstone of OptForce . Figure 1 pictorially illustrates the proposed concept . By superimposing the flux ranges for a given reaction in the wild-type vs . the overproducing network a number of possible outcomes are revealed . If there is any degree of overlap between the two reaction flux ranges ( Figure 1a ) then it may be possible to achieve the overproduction target without changing the value of the corresponding reaction flux in the wild-type strain . In contrast , if the flux ranges for a reaction in the wild-type metabolic network are completely to the left ( Figure 1b ) or to the right ( Figure 1c ) of the corresponding ranges for the overproducing metabolic network then the overproduction target cannot be achieved unless the reaction flux is directly or indirectly changed . The case depicted in Figure 1b calls for an increase whereas the one shown in Figure 1c requires a decrease in the reaction flux value . Note that if the reaction flux range collapses to zero then the corresponding reaction needs to be eliminated ( e . g . , through a gene knock-out ) . The gap between the two flux ranges quantifies the degree of required reaction flux modification . This reaction flux modification does not necessarily have to be realized by actively engineering the gene that codes for the enzyme catalyzing the reaction ( e . g . , through changed promoter , codon usage , or gene disruption/knock-out ) . It may come about indirectly by propagating through stoichiometry the effect of modifications occurring in other parts of metabolism ( e . g . , coupled reactions in series , cofactor coupling , etc . ) . We refer to reaction fluxes that must increase ( see Figure 1b ) in the face of the imposed overproduction requirements as MUSTU whereas the ones that must decrease ( see Figure 1c ) as MUSTL . Fluxes of reactions with overlapping ranges ( see Figure 1a ) between the wild-type and overproducing network do not provide any imperatives on network modifications when considered one at a time . Therefore , we further scrutinize them by considering sums of two reaction fluxes at a time and subsequently calculating their ranges in the wild-type and overproducing metabolic networks . This concept is similar to the use of residue doubles in the dead-end elimination algorithm for protein design [44] . As was the case of single reaction fluxes , three outcomes are possible ( see Figure 1d–f ) . Non-overlapping ranges imply that in the overproducing network either one or the other reaction flux ( but not necessarily both ) must increase ( Figure 1d ) or decrease ( Figure 1e ) in value . These pairs of reactions form sets MUSTUU and MUSTLL respectively . One can extend this concept further by analyzing the range of not just the sum of two fluxes but also their difference for the wild-type and overproducing networks ( see Figure 1f ) . As before , non-overlapping ranges imply that either the first reaction flux must increase or the second reaction flux must decrease . By extension , these pairs of reactions form the equivalent sets MUSTUL and MUSTLU , respectively . One can systematically extend this analysis by considering sums and/or differences of three , four , etc . reactions at a time . Collectively , the derived sets ( e . g . , MUSTL , MUSTU , MUSTUU , MUSTLLL , MUSTUULL , etc . ) encompass all the necessary reaction flux changes that MUST take place in the wild-type metabolic network for the desired overproduction . Appendix B in Text S2 introduces a bilevel formulation for identifying all MUST sets without relying on exhaustive enumerations inspired by a similar representation introduced earlier [45] for identifying synthetic lethal deletions . The next step of OptForce is to identify how the collective set of changes ( encoded within the MUST sets ) can be imparted on the wild-type metabolic network with the minimal number of direct interventions ( i . e . , knock-up/down/outs ) . The identified MUST sets encode Boolean choices regarding which fluxes ( or combinations thereof ) must change in value . Upon the incorporation of these constraints , an optimization formulation is proposed ( see Appendix C in Text S3 ) that finds the minimum number of imparted changes ( through gene knock-outs/up/downs ) so as the overproducing metabolic network involves no feasible metabolic phenotypes that fail to meet the imposed production target . The collective set of minimal network modifications that yield the desired overproduction target is referred to as the FORCE set and is typically represented as a Boolean diagram globally depicting all minimal alterative choices for engineering the wild-type network . Many of the reactions in the FORCE set are also members of various MUST sets . The optimization formulations for computing the allowable flux values for all reactions in the wild-type metabolic network are provided in Appendix A ( see Text S1 ) . The derivation and solution procedure of bilevel optimization formulations for exhaustively elucidating the membership in the MUST sets are provided in Appendix B ( see Text S2 ) . The bilevel optimization formulation for identifying the FORCE set of engineering interventions is given in Appendix C ( see Text S3 ) . All optimization problems were solved using the GAMS/CPLEX ( version 9 . 1 ) solver on a 2 . 6 GHz AMD Opteron Processor with 32 GB of ECC RAM .
Figure 2 lists the identified MUSTU and MUSTL sets of reactions whose fluxes must depart the original ranges . Note that because all members of set MUSTL involve fluxes set to zero we re-designate them as MUSTX to signify that they all correspond to reaction eliminations . Not surprisingly , the transport reaction directing succinate out of the cytosol ( SUCCt3rpp ) was classified into the MUSTU whereas transport reactions for competing by-products such as ethanol ( ETOHt2rpp , ETOHtex ) , acetate ( ACtex ) , formate ( FORtex ) and acetaldehyde ( ACALtpp , ACALDtex ) were completely blocked ( i . e . , members of the MUSTX set ) . In addition , a number of reactions from hisitidine ( ATPPRT , HISTD , HISTP , HSTPT , IG3PS , IGPDH , PRAMPC , PRATPP and PRPPS ) and methionine metabolism ( AHCYSNS , DHPTDCs , HCYSMT and RHCCE ) were also set to zero . Note that these reactions are essential for amino acid biosynthesis and are fully coupled to growth . Therefore , the drain of carbon flux from the pentose phosphate pathway towards histidine and methionine synthesis is prevented thus halting the production of biomass . While results for MUSTU and MUSTL involve primarily intuitive negations of by-products formation , sets MUSTUU , MUSTUL and MUSTLL allude to more complex flux re-allocations ( see Figure 3 ) . For example , in the MUSTUU set the increase in the flux for reaction phosphoenolpyruvate carboxylase ( PPC ) can only be compensated by the simultaneous increase in the flux of five TCA cycle reactions ( i . e . , MALS , CS , ACONTa , ACONTb and ICL ) . This implies that at least one of two possible avenues for succinate production must be increased under anaerobic conditions ( see Figure 3a ) . Specifically , either the flux along the traditional succinate synthesis route through the reductive pathway that converts oxaloacetate ( oaa ) to malate and fumarate or the flux through the glyoxylate shunt needs to increase . Interestingly , the higher succinate yield of the latter mechanism due to NADH availability has been implemented in E . coli by deactivating the iclR repressor ( to activate the glyoxylate bypass ) under anaerobic conditions by [59] . Figure 3a reveals that a number of flux up-regulations ( e . g . , PPC , PGM , CS , ICL , ACONTa/b , PGM , ATPS4rpp , ALDD2x , ACALD ) and down-regulations ( e . g . , PFL , TPI , RPI , ASPTA , PGK ) appear frequently as choices in multiple pairs . These mutually compensatory flux changes can be more clearly discerned by fusing all interacting components from MUSTUU , MUSTUL and MUSTLL into a single graph ( see Figure 3b ) where fluxes that increase are shown in green and those that decrease are shown in red . The importance of PPC up-regulation is manifested by the fact that as many as ten separate reaction flux modifications would be needed to replace it . Similarly , the decrease in flux through PFL can only be compensated by up-regulating the flux of four reactions along the glyoxylate shunt while the down-regulation of the flux through ENO can only be replaced by the up-regulation of four reactions supplying flux to the TCA cycle . The compensatory interconnections in Figure 3b suggest that not all depicted flux modifications are simultaneously needed to reach the desired phenotype ( i . e . , 100% yield of succinate ) . Instead , all flux modifications implied by sets MUSTLL , MUSTUU and MUSTUL can be satisfied by up- or down-regulating a minimal set of reactions . We identified all such minimal reaction flux modification sets and depicted them in the form of a Boolean diagram in Figure 3c . As expected , up-regulation of the flux through PPC is a consensus choice while the up-regulation of only one out of ACONTa , ACONTb , CS and ATPS4rpp is needed . Interestingly , the down-regulation of PFL which diverts flux towards organic acids such as formate , lactate , acetate , ethanol , etc . emerged as a required change despite its relatively low connectivity in the diagram of Figure 3b . Figure 4 depicts the reaction flux modifications needed when considering three reaction fluxes at a time ( one out of three ) . The reactions are denoted as ovals where green nodes represent the flux of the reaction that increases and red nodes indicate those that decrease . They span up-regulations ( MUSTUUU ) , down-regulations ( MUSTLLL ) or combinations thereof ( MUSTUUL and MUSTULL ) . Figure 4a re-affirms the key role of up-regulating PPC but also reveals the importance of redirecting the flux of reactions from pyruvate metabolism ( i . e . PFL , ACS , ACALD , ACKr , PTAr ) towards acetyl-CoA . Furthermore , Figure 4a reveals that the decrease in the value of the flux for phosphotransacetylase ( PTAr ) and acetate kinase ( ACKr ) reduces the export of acetate and increases the amount of acetyl-CoA available for the glyoxylate pathway . These results are in agreement with the knockouts for ackA and pta in strain SBS990MG constructed for succinate synthesis [59] . The reaction modifications implied in MUSTLLL , MUSTUUU , MUSTUUL and MUSTULL can also be distilled into a minimal set of modifications ( see Figure 4b ) . Many of these modifications were present in Figure 3c , however , a number of new imperatives such as reducing the flux of FUM emerge . One can methodically , continue to identify additional constraints that need to be satisfied to achieve the desired phenotype by looking into higher-order combinations of fluxes . The results for reactions quadruples are provided as supplementary material ( see supporting information - Text S4 and Figure S1 ) . We next used the bilevel optimization formulation ( refer Appendix C in Text S3 ) to identify the minimal set of reaction modifications ( i . e . , FORCE set ) that guarantee the imposed yield ( 100% succinate yield ) . Note that the identified MUST reaction flux modifications were added as constraints in the FORCE set formulation . However , we found that the flux restraints ( single , double and triple reaction combinations ) in the MUST sets were insufficient to guarantee the target yield for succinate ( i . e . , min Vsuccinate = 64% of theoretical ) . This suggested that additional reactions that participate in higher-order ( unexplored ) MUST sets were required to guarantee the target yield for succinate . Upon allowing reactions absent from the MUST sets to become members of the FORCE set the imposed target for succinate production was met . The identified minimal set of forced modifications ( see Figure 5a ) is comprised of ten different interventions . The up-regulation of PPC and CS ensures that the pool of oxaloacetate is diverted towards the TCA cycle . The up-regulation for PGK and TPI increases the glycolytic activity providing precursor metabolites such as phosphoenol pyruvate , oxaloacetate etc . to succinate synthesis . The down-regulation of PFL , GLUDy and ASPTA prevents the formation of by-products such as formate , lactate , ethanol , glutamate , aspartate and 2-ketoglutarate . The up-regulation for ACALD converts any residual acetate back into acetyl-CoA , which in turn is converted to succinate . Notably , for two such interventions there exist two equivalent alternatives . The first one involves the up-regulation of either of ACONTa/b isozymes to ensure conversion of citrate into glyoxylate and succinate . The second one requires either the down-regulation of malate dehydrogenase ( MDH ) that converts malate into oxaloacetate or the down-regulation of ICDHy that diverts flux away from the glyoxylate shunt . Interestingly , none of the transport reaction regulations identified in the MUSTU and MUSTX sets are present in the FORCE sets . The optimization formulation for the FORCE set identified more economical upstream flux modifications that negated the formation of multiple by-products . A consequence of imposing 100% yield to succinate is that biomass formation is halted as histidine and methionine formation is seized . In the next section , we examine how the identified engineering interventions change when a 1% biomass requirement is imposed simultaneously with a 98% yield requirement for succinate . In addition , we contrast the magnitude of the imposed flux changes for the two different scenarios . Figure S2 ( see supplementary information ) lists all MUST sets involving single , double and triple reaction combinations . As expected , we find that by dialing back the requirement for succinate production the number of flux modifications that must happen in the network to meet the new requirement is reduced . Lowering the yield of succinate from 100 to 98% eliminates all reaction deletions ( i . e . , members of the MUSTX set ) belonging to competing pathways . The ethanol transport reactions ( ALCD2x and ETOHt2rpp ) do not have to be completely eliminated but rather lowered in value to 3 mmol/gDW . hr from a wild-type flux value of 19 mmol/gDW . hr . Despite the differences in the MUST sets between cases 1 and 2 the corresponding FORCE sets of reactions were identical . Up-regulations for PPC , CS , MALS , ICL and ACONTa and down regulations for reactions along the pathways leading to competing by-products were required for the 98% succinate yield case . Even though the membership of the FORCE set is the same the corresponding required levels of up or down-regulation are slightly different . Figure 6 depicts the original wild-type flux ranges and the new values that the reaction fluxes must reach to guarantee the imposed succinate production targets under cases 1 and 2 respectively . The largest difference between the two arises for the down-regulation of ACALD where a value of 7 . 5 mmol/gDW . hr suffices for case 2 while a value of 1 . 4 mmol/gDW . hr is needed for case 1 . Note that a number of glycolytic fluxes are set at their stoichiometric upper bounds ( i . e . , PPC , PGK and TPI ) implied by the uptake of 100 moles of glucose . Next , we explore how the addition of a single heterologous reaction ( i . e . , pyruvate carboxylase ) radically changes the way that the network needs to be re-engineered . Pyruvate carboxylase ( PYC ) has been overexpressed in E . coli from Lactococcus lactis [58] , [59] and Rhizobium etli [49] . The addition of the new reaction to the metabolic network boosts the succinate yield by 15 . 3% above the original theoretical maximum ( 1 . 72 moles/mole of glucose ) . PYC using ATP directly converts pyruvate into oxaloacetate which serves as a precursor for the glyoxylate and the fermentative pathway . In this study , we allowed the production of biomass at 1% of theoretical yield and identified the flux changes when succinate was produced at 98% of theoretical maximum ( 1 . 7 moles/mole of glucose ) . Figure 7 shows the results for the MUST set of reactions . As expected , the transport reaction for succinate and ATP are both members of the MUSTU set whereas the transport reaction for acetaldehyde is classified as MUSTL . The required increase in the flux for ATP is due to the ATP consuming pyruvate carboxylase . Unlike cases 1 and 2 , the synthesis route for by-products ( formate and acetyl-CoA ) consuming pyruvate through the pyruvate formate lyase ( PFL ) , alcohol dehydrogenase ( ALCD2x ) and formate dehydrogenase ( FDH5pp ) reactions are completely shut off to afford a complete conversion of pyruvate to OAA . This suggests that the presence of PYC provides an alternative route to PPC whereby OAA can be replenished either by increasing the flux through PPC or PYC . This is in agreement with the experimental findings by Ka-Yiu San and coworkers [59] that a drop in the activity of one the two enzymes can be compensated by the other . The FORCE set of engineering interventions for this scenario is contrasted against cases 1 and 2 and is shown in Figure 5b . The addition of the PYC reaction significantly reduces the number of engineering interventions required to guarantee the target yield for succinate . The interventions required to reduce the drain of carbon away from the pyruvate metabolism are absent indicating that the pyruvate carboxylase enzyme can safeguard against the consumption of pyruvate towards side-products . However , the down regulation for ASPTA is again needed to reduce the secretion of aspartate and glutamate . Importantly , the up-regulation for PYC could be substituted by up-regulating PPC which suggest that the OAA pool can be replenished by either of these two reactions . The increase in activity for some reactions in the glycolytic pathways ( TPI , PGK ) and the TCA cycle ( ACONTa , ACONTb and MDH ) is required as before . In contrast with the previous case-study , the complete elimination of PFL and isocitrate dehydrogenase ( ICDHy ) , rather than just their down-regulation is needed . The elimination of PFL is imposed to completely prevent the conversion of pyruvate into by-products . The elimination of ICDHy blocks the flow of carbon flux through the TCA cycle into the glutamate pathway thus ensuring the complete conversion of isocitrate into glyoxylate and succinate .
In this paper , an optimization-based methodology called OptForce was introduced for predicting all possible metabolic modifications that could guarantee , subject to the model stoichiometry and conditions , a pre-specified overproduction level of a desired biochemical . The results for succinate overproduction in E . coli reveal that the needed interventions results remain the same upon requiring the production of a small amount of biomass but change significantly upon the addition of a key reaction to the model . Many of the suggested interventions recapitulate existing strain redesign strategies for succinate synthesis . For example , experimental evidence suggests that the overexpression of PPC from Sorghum vulgare and Actinobacillus succinogenes in E . coli not only increases the yield of succinate but also reduces the secretion of acetate [31] , [58] , [59] , [63]–[65] . In addition , succinate production has been enhanced by the increased carboxylation of PEP and pyruvate ( to increase the pool of OAA for TCA cycle ) in the E . coli mutant NZN111 by decreasing the activity for pyruvate formate lyase ( PFL ) and lactate dehydrogenase [47] , [53] . Furthermore , Vemuri et al . [62] , [66] made use of the glyoxylate pathway for succinate synthesis thus overcoming the limitation of NADH availability for the fermentation pathway . The up-regulations for the isozymes ACONTa/b and the down regulations for ICDHy , ASPTA and GLUDy predicted by OptForce allude to the same strategy of glyoxylate shunt utilization for succinate synthesis . Finally , multiple studies [31] , [59]–[61] have shown that the deletion of adhE and ackA-pta coding for acetaldehyde dehydrogenase ( ACALD ) reduces the formation of by-products ethanol , acetate and acetaldehyde as suggested by OptForce . The up-regulation of citrate synthase ( CS ) , aconitase ( ACONTa/b ) and reactions from the glycolytic pathway ( PGK and TPI ) are engineering strategies suggested by OptForce that to the best of our knowledge have not yet been implemented for succinate production . Heterologous overexpression of the citZ gene from Bacillus subtillis that encodes citrate synthase increased the activity through the TCA cycle towards isocitrate and 2-ketoglutarate [67] . However , when this gene was overexpressed in E . coli strain SBS550MG , an increase in the yield of succinate was not observed [59] . The reason for this could be the absence of the down regulations for ICDHy and GLUDy that lead to the production of glutamate and other amino acids required for growth . The results predicted by OptForce suggest that by collectively incorporating the flux modulations for citrate synthase , isocitrate dehydrogenase and glutamate dehydrogenase along with the existing strategies , the yield of succinate can be further enhanced from the current experimental yield ( 1 . 7 moles/mole of glucose ) as observed for strains SBS550MG and SBS990MG [59] . The genetic interventions predicted by OptForce underscore the importance of up-regulating key fluxes along the succinate pathway in addition to the knockouts for by-products . Existing strain optimization procedures ( e . g . OptKnock [21] and OptReg [22] ) that couple the maximization of growth rate and secretion of the product tend to prevent the yield of succinate from reaching the theoretical maximum . Table 1 contrasts the yields predicted for succinate overproduction by OptKnock [21] , OptReg [22] and OptForce . OptKnock and OptReg rely on biomass maximization to perform flux allocation in the metabolic network whereas OptForce reports the most conservative value for succinate production allowed by the stoichiometry and conditions . It is noteworthy that for more than two interventions even the worst-case succinate yield predictions by OptForce are far more superior to strategies predicted by OptKnock and OptReg . Notably , OptForce suggested the down regulation but not the knockout of PFL and GLUDy [59] along with a number of additional interventions missed by both OptKnock and OptReg due to their inconsistency with biomass maximization . The OptForce procedure allows for the complete enumeration of engineering modifications consistent with an overproduction target ( s ) . The incorporation of metabolic flux information about the wild-type network allows for a sharper elucidation of engineering interventions . The engineering interventions predicted by OptForce depend on the available flux measurements for the initial strain . OptForce can be modified to predict globally valid metabolic interventions by utilizing biological objectives ( i . e . maximization of biomass ) when sufficient metabolic flux data are not available . Furthermore , the procedure can hierarchically be applied at intermediate stages of a metabolic engineering project by re-calculating the set of engineering interventions as new flux data for ( multiple ) mutant strains become available . The restriction of minimality in the calculated FORCE set can be relaxed allowing for the exploration of less parsimonious engineering interventions . For example , we studied the case for identifying additional interventions after retaining the best eight out of the ten interventions originally identified by the OptForce method ( for cases 1 and 2 ) . However , we found that even after allowing seven additional interventions ( i . e . K = 15 ) , the resulting FORCE set was not sufficient to increase the yield to more than 80% of the theoretical maximum . In addition , reactions that cannot ( e . g . , diffusion limited transport , non-gene associated reactions , etc . ) be directly manipulated can be excluded from consideration during the derivation of the FORCE set . It is to be noted that the OptForce procedure provides targets for genetic manipulations at the metabolic flux level . The lack of a completely quantitative mapping between gene expression and flux levels implies that multiple rounds of experimental strain modifications may be needed to translate the FORCE set of reaction fluxes to the required gene expression levels . An algorithmic implementation of the procedure is available as supplementary material ( see supporting information - Text S5 ) . | Over the past few years , there has been an unprecedented increase in the use of microorganisms for the production of biofuels , industrial chemicals and pharmaceutical precursors . In this regard , biotechnologists are confronted with the challenge to efficiently convert biomass and other renewable resources into useful biochemicals . With the advent of organism-specific mathematical models of metabolism , scientists have used computations to identify genetic modifications that maximize the yield of a desired product . In this paper , we introduce OptForce , an algorithm that identifies all possible metabolic interventions that lead to the overproduction of a biochemical of interest . Unlike existing techniques , OptForce does not rely on the maximization of a fitness function to predict metabolic fluxes . Instead , OptForce contrasts the metabolic flux patterns observed in an initial strain and a strain overproducing the chemical at the target yield . The essence of this procedure is the identification of all coordinated reaction modifications that force the network towards the overproduction target . We used OptForce to predict metabolic interventions for succinate overproduction in Escherichia coli . The results described in this paper not only uncover existing strain designs for succinate production but also elucidate new ones that can be experimentally explored . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"computational",
"biology/systems",
"biology",
"biochemistry/bioinformatics",
"computational",
"biology/metabolic",
"networks",
"biotechnology/bioengineering"
] | 2010 | OptForce: An Optimization Procedure for Identifying All Genetic Manipulations Leading to Targeted Overproductions |
Rotavirus ( RV ) is the major cause of severe gastroenteritis in young children . A virus-encoded enterotoxin , NSP4 is proposed to play a major role in causing RV diarrhoea but how RV can induce emesis , a hallmark of the illness , remains unresolved . In this study we have addressed the hypothesis that RV-induced secretion of serotonin ( 5-hydroxytryptamine , 5-HT ) by enterochromaffin ( EC ) cells plays a key role in the emetic reflex during RV infection resulting in activation of vagal afferent nerves connected to nucleus of the solitary tract ( NTS ) and area postrema in the brain stem , structures associated with nausea and vomiting . Our experiments revealed that RV can infect and replicate in human EC tumor cells ex vivo and in vitro and are localized to both EC cells and infected enterocytes in the close vicinity of EC cells in the jejunum of infected mice . Purified NSP4 , but not purified virus particles , evoked release of 5-HT within 60 minutes and increased the intracellular Ca2+ concentration in a human midgut carcinoid EC cell line ( GOT1 ) and ex vivo in human primary carcinoid EC cells concomitant with the release of 5-HT . Furthermore , NSP4 stimulated a modest production of inositol 1 , 4 , 5-triphosphate ( IP3 ) , but not of cAMP . RV infection in mice induced Fos expression in the NTS , as seen in animals which vomit after administration of chemotherapeutic drugs . The demonstration that RV can stimulate EC cells leads us to propose that RV disease includes participation of 5-HT , EC cells , the enteric nervous system and activation of vagal afferent nerves to brain structures associated with nausea and vomiting . This hypothesis is supported by treating vomiting in children with acute gastroenteritis with 5-HT3 receptor antagonists .
Rotavirus ( RV ) is the major cause of infantile gastroenteritis worldwide and the infection is associated with approximately 600 , 000 deaths every year , predominantly in developing countries [1] . Most of the deaths result from excessive loss of fluids and electrolytes through vomiting and diarrhoea . Despite its significant clinical importance and the research conducted over several decades , knowledge of the pathophysiological mechanisms that underpin this life-threatening disease remains limited . Several mechanisms have been proposed to account for the watery diarrhoea associated with RV infection . These include osmotic diarrhoea following a virus-induced loss of epithelial absorptive function , the effect of NSP4 , a virus-encoded enterotoxin , and an active role of the enteric nervous system ( ENS ) and serotonin ( 5-hydroxytryptamine , 5-HT ) [2]–[5] . However , the pathophysiological basis of virus-induced emesis , a hallmark of illnesses caused by RV and norovirus , is poorly understood . The human ENS contains about 100 million neurones which are sensory- , inter- and motor neurons [6] . The luminal enterochromaffin ( EC ) cells “taste” and “sense” the luminal contents and can release mediators such as 5-HT to activate ENS as well as extrinsic vagal afferents to the brain . 5-HT is located in the secretory granules of the EC cells , which are most abundant in the duodenum and comprise the single largest enteroendocrine cell population . They are strategically positioned in the intestinal mucosa to release mediators of endocrine signalling from the basolateral surface to activate afferent neuron endings within the lamina propria [7] , [8] . Following stimulation by several agents ( e . g . hyperosmolarity , carbohydrates , mechanical distortion of the mucosa , cytostatic drugs ) including the cholera toxin [8] , [9] , EC cells mobilize intracellular Ca2+ and release 5-HT [10] . 5-HT is involved in the regulation of gut motility , intestinal secretion , blood flow and several gastrointestinal disorders [11]–[15] including RV diarrhoea [3] , chemotherapy-induced nausea and vomiting [16] , [17] and Staphylococcal enterotoxin-induced vomiting [18] . We have previously shown that RV infection results in stimulation of the ENS and that RV diarrhoea in mice can be attenuated with 5-HT3 receptor antagonists , such as granisetron [3] . This drug is frequently used to reduce vomiting in humans [19] , including children with acute gastroenteritis [20]–[23] , suggesting that EC cells could be an important mediator of RV-induced diarrhoea and vomiting , symptoms regarded as naturally inherited host responses [24] . The proposed anti-emetic mechanism of 5-HT3 receptor blockade involves an action on receptors located on vagal afferents communicating with the medullary vomiting centre [25] , [26] which is supported by observations that the stimulation of vagal 5-HT3 receptors in ferrets , shrews and dogs triggers an emetic reflex , evoking reverse peristalsis in response to chemotherapeutic agents and radiation treatment [19] , [27]–[29] . The present study was designed to test the novel hypothesis that RV can stimulate human EC cells in the gut , causing release of 5-HT , which activates vagal afferents and the brain stem vomiting centre , a reaction cascade associated with vomiting .
To determine whether RV can infect EC cells , primary EC tumor cells ( t . c . ) were harvested from patients with liver metastasis undergoing surgery for the midgut carcinoid syndrome . In addition , a human midgut carcinoid t . c . line ( GOT1 ) , was examined for susceptibility to RV infection . The EC cell phenotype was confirmed by staining for 5-HT and the neuroendocrine granule marker chromogranin A . Immunofluorescence microscopy revealed that 100% of the GOT1 cells and 95% of the primary EC t . c . were positive for chromogranin A , and 68% of GOT1 cells and 40% of the primary EC t . c . were positive for 5-HT ( Figure . S1 ) ; this confirmed that the vast majority of the cells were EC specific . To determine whether RV replicates in EC cells , we infected confluent GOT1 cells ( 200 , 000 cells/well ) with trypsin-activated Rhesus rotavirus ( RRV ) ( MOI 0 , 1 ) , collected cells and media at 3 , 24 and 48 h post infection ( p . i . ) and determined the progeny viral titre . The EC t . c . did indeed support RRV replication ( Figure . S2 ) and the viral titres increased >100 fold from ≤1×101 plaque forming units ( pfu ) /ml in samples collected 3 h p . i . to 2×103 pfu/ml at 48 h p . i . RV infection has previously been shown to increases the intracellular Ca2+ concentration in human intestinal cells [30]–[34] . To investigate whether RRV has a similar effect on GOT1 and primary t . c . , the cells were loaded with the ratiometric fluorescent Ca2+ indicator Fura-2 prior to infection with trypsin-activated RRV at an MOI of 10 and 3 respectively ( Protocol . S1 ) . The Ca2+-images were captured at 1 , 5 and 23 h p . i . These experiments showed that RV infection caused an elevation in intracellular Ca2+ in EC t . c . within 60 min . ( Protocol . S2 , Figure . S3 ) . GOT1 cells had a maximum release of Ca2+ at 60 min p . i . and primary cells had a maximum release at 5 h p . i . This observation is important since the release of 5-HT is known to be a Ca2+-dependent process in EC cells [35] , including human enterochromaffin-like cells [36] . Using the cholera toxin ( 2 nM , 20 nM , 200 nM ) as an agonist we observed a dose-dependent release of 5-HT from primary EC t . c . within 24 h ( Figure . S4 ) . RV was shown to stimulate the release of 5-HT in a dose and time-dependent manner with an increase of 5-HT release after 6 h p . i . in primary EC t . c . ( Figure . 1A ) . Furthermore , when primary EC t . c . and GOT1 cells were incubated with supernatant from RRV-infected MA104 cells , a release of 5-HT was observed within 60 min ( Figures . 1B , 1C ) . Such an early 5-HT response indicated that viral replication was unlikely to be required for 5-HT release . More likely , a release of viral protein ( s ) from infected cells could explain this effect . Purified double- and triple-shelled RV [37] ( 1 and 2 µg/ml , respectively ) and virus particle-free supernatant ( 130000× g; 4 h , SW40 ) were individually added to GOT1 ( 500 , 000 cells/well ) and primary EC t . c . ( 200 , 000 cells/well ) for 60 min . While the ultracentrifuged supernatant stimulated 5-HT release within 60 min ( Figures . 1B , 1C ) , no such effect was observed using the purified particles ( data not shown ) . Therefore , it was concluded that double- and triple-shelled RV were not responsible for this early effect on 5-HT secretion . The enterotoxic glycoprotein NSP4 is secreted from human intestinal cells following RV infection [38] , [39] and has been reported to mobilize intracellular Ca2+ from human HT-29 cells [40] . Therefore , we examined whether NSP4 secreted from polarized human intestinal Caco-2 cells infected with rotavirus [39] or recombinant NSP4 [41] could stimulate the release of 5-HT from GOT1 and primary EC t . c . Indeed NSP4 at concentrations ranging from 0 . 25–2 . 5 µM resulted in the secretion of 5-HT from primary EC t . c ( Figure . 2 ) . In a subsequent experiment , the addition of 2 µM NSP4 to GOT1 cells for as short time as 60 min revealed an increase of 204% of 5-HT in the media from cells stimulated with recombinant NSP4 ( data not shown ) , and an increase of 84% of 5-HT from the secretory NSP4 ( data not shown ) . Thus , both recombinant NSP4 and NSP4 secreted from rotavirus-infected Caco-2 cells were capable of stimulating the release of 5-HT from human EC t . c . Both GOT1 and primary EC t . c . were loaded with the ratiometric fluorescent Ca2+ indicator Fura-2 and stimulated with NSP4 . Following the establishment of a stable Fura-2 fluorescence baseline , recombinant NSP4 ( 2 µM ) was added to cells and the Ca2+ signal measured continuously for 40–50 min . The intracellular Ca2+ concentration increased 1 . 5–2-fold in EC t . c . ( GOT1 n = 8 cells , primary EC t . c . n = 10 cells ) within 30 min of NSP4 stimulation ( Figure . 3 ) , which was consistent with observations in Caco-2 cells [34] . Ionomycin ( 1 µM ) was added at the end of the experiment as a control to compare the magnitude of the increase , but it had little further effect , suggesting that NSP4 is a strong mobilizer of intracellular free Ca2+ ( Figure . 3 ) . Exogenous NSP4 can stimulate IP3 production in human intestinal HT-29 cells [40] , and a human pancreatic carcinoid cell line ( BON ) was shown to produce IP3 after mechanical stimulation and activation of purinergic receptors [42] . We therefore determined weather NSP4 caused IP3 production in EC cells . The production of IP3 was indeed stimulated in GOT1 cells ( 400 , 000 cells/well ) when incubated for 60 min with recombinant or secretory NSP4 ( 2 µM ) . Caracole ( 100 µM ) , a muscarinic receptor agonist known to activate the PLC pathway , was also used as an agonist . Both forms of NSP4 activated the PLC pathway , albeit to a less degree than carbachol ( Figure . 4A ) . Since activation of adenylyl cyclase ( AC ) and cyclic adenosine monophosphate ( cAMP ) signalling has been previously associated with Ca2+ increase and 5-HT release in GOT1 [43] , KRJ-I [44] and BON cells [36] , we next determined whether NSP4 could stimulate this pathway . GOT1 cells ( 650 , 000 cells/well ) were thus stimulated with 2 µM NSP4 and the AC agonist isoproterenol ( 10 µm ) for 5 and 30 min , followed by collection of the supernatant and quantification of cAMP as an indicator of AC activation . While isoproterenol stimulated cAMP in GOT1 cells , no effect of NSP4 was observed at 5 min ( Figure . 4B ) or 30 min ( data not shown ) , suggesting that the cAMP pathway in GOT1 cells was not activated by NSP4 . The distribution and occurrence of neuroendocrine cells in the small intestine of the mouse duodenum , jejunum and ileum was assessed by immunohistochemistry for the neuroendocrine secretory granular marker chromogranin A . The duodenum , jejunum and ileum all contained neuroendocrine cells ( Figures . 5A–D ) . They were observed in the crypts as well as among mature villous enterocytes , most abundantly in the duodenum . To investigate whether RV could infect enterocytes in the close vicinity of EC cells or infect chromogranin/5-HT-containing EC cells , infant mice were infected with murine rotavirus ( strain EDIM ) as described [5] and processed for histopathology and immunohistochemistry at different time points p . i . Figure 6A illustrates the typical vacuolization of infected mature enterocytes 48 h p . i . The vacuoles were occasionally found in the close vicinity of chromogranin-containing cells ( Figure . 6B ) . No infected crypt cells were seen . Moreover , confocal microscopy revealed a co-localization between RV proteins and 5-HT-containing EC cells in jejunum of infected mice ( Figures . 6C , 6D ) . Evidently , RV can infect enterocytes in the close vicinity of EC cells in the small intestine , and occasionally EC cells as well . To further support our previous observations [3] , that 5-HT participates in RV-induced illness , 50 µL 5-HT at 5 mg/kg ( Serotonin creatinine sulfate monohydrate , Sigma Aldrich , Code: 85030 ) were administered intraperitoneally to five to seven days old Balb/C pups . Our objective was to investigate whether diarrhoea would occur . Following 30 min post administration , 5 of the 7 pups responded with diarrhoea , after 45 min 6 out of 7 and after 60 min all 7 responded with diarrhoea indistinguishable from RV–induced diarrhoea [3] . To determine whether RV infection leads to activation of brain structures involved in sickness symptoms such as nausea and vomiting , immunohistochemical detection of Fos , a commonly used marker of neuronal activity was carried out on brain sections of RV-infected and uninfected mice [45] . Fos is an immediate early gene that is expressed upon repeated depolarization of neurons . In brains harvested 48 h . p . i . , there was a robust activation of nucleus of the solitary tract , the main target structure for incoming fibers from the vagal nerve . This was seen in 3/5 of the infected animals . Thus , RV infection activated brain areas considered to be the vomiting centre . No activity was seen in uninfected animals ( 0/6 ) ( Figure . 7 ) . Our observation shows for the first time that RV activates brain structures associated with vomiting , presumably through activation of vagal afferents . No clear infection-induced Fos expression was seen in other parts of the brainstem .
Despite the clinical importance of vomiting induced by RV infection and its contribution to severe dehydration , no mechanism has yet been proposed for emesis . Progress in elucidating the RV-induced emetic mechanism has been hindered by lack of appropriate small animal models , because most commonly used animal models that are susceptible to RV , i . e . mice and rats , do not exhibit an emetic response [24] . Progress has been further restricted by the limited availability of human primary and established EC cell lines . Primary human carcinoid EC cells as well as the carcinoid EC cell line ( GOT1 ) , enable investigatigation of the role of EC cells in RV pathogenesis . The EC cells have been identified from the mid- to the top of villi in the duodenum , jejunum and ileum , which are the segments associated with RV replication and histopathological lesions [46] . A cross-talk between EC cells and infected enterocytes is supported by immunohistochemistry and confocal microscopy experiments . The neuroendocrine cells were found in the close vicinity of cells with vacuoles , a characteristic feature of RV-infected enterocytes [46]–[48] . Moreover , RV sometimes co-localized with EC cells , suggesting a paracrine signalling of NSP4 and a cross-talk between EC cells and NSP4 in vivo . Consistent with the proposal of EC cells being key cells in RV pathophysiology , we demonstrated that: ( i ) RV cause release of 5-HT from primary EC t . c . , ( ii ) crude and virus particle-free supernatants from RV-infected MA104 cells stimulated 5-HT release within 60 min in primary and GOT1 EC t . c . , ( iii ) recombinant and secretory NSP4 caused release of 5-HT from primary and GOT1 EC t . c . The latter finding is particularly interesting , and is to the best of our knowledge , the first observation of a virus/viral protein-mediated effect on human EC t . c . , stimulating 5-HT release . This release exhibited similar time kinetics as the Ca2+ mobilization , which is consistent with previous findings in EC cells and BON cells [35] , [36] . RV also stimulated 5-HT release from EC t . c . in a time and dose-dependent manner beginning about 6 h p . i . It is reasonable to believe that the 5-HT release and Ca2+ mobilization were associated with expression of NSP4 . Exogenous NSP4 can also cause the mobilization of intracellular Ca2+ [30] , [31] via PLC-mediated IP3 production in human intestinal cells [40] , [49] . Moreover , NSP4 is secreted from human intestinal cells [38] in a polarized fashion [39] . Therefore , we explored the novel hypothesis that RV , and particularly NSP4 , stimulate EC t . c . via mobilization of intracellular Ca2+ and release of 5-HT . Since the elevation of Ca2+ after addition of NSP4 occurred after approx 20 min , similar to RV infections in Caco-2 cells [34] , we speculate that the effect was not mediated by the opening of ion channels in the plasma membrane . It is more likely that it originated from internal ER stores as previously shown [31] , although we cannot exclude that other possible mechanisms were involved . The kinetics of the Ca2+ response were similar to those of non-infected human intestinal Caco-2 cells inoculated with supernatant from RRV-infected Caco-2 cells at 18 h p . i . [34] . As the presence of proteins in the supernatant at this time point could not be explained by cell lysis , it was suggested that the Ca2+ mobilization is stimulated by viral proteins or peptides secreted from RV-infected cells [34] . Indeed , release of soluble NSP4 from RV-infected human intestinal cells has been demonstrated [38] , [39] , and we can now report that both recombinant NSP4 and NSP4 secreted from virus infected intestinal cells mobilize Ca2+ in EC t . c . and stimulate 5-HT release . The EC cells express an ensemble of ligand-gated ion channels , chemo- and mechanosensitive-ion channels and G-protein-coupled receptors on their surface [50] , [51] . G-protein-coupled AC and PLC are key enzymes involved in 5-HT release in EC cells [36] , [44] , [50] , [52] , [53] . Kolby and co-workers previously showed that GOT1 cells weakly responded to carbachol , a muscarinic receptor agonist and activator of PLC , suggesting that these cells lacked functional muscarinic receptors or had an impaired PLC-dependent formation of IP3 [43] . Consistent with this observation , we found only a modest response to carbachol and NSP4 , indicating that NSP4 increases Ca2+ in an PLC-independent way [54] . In order to explain the mechanism , by which NSP4 alter the ER , it has been proposed that NSP4 stimulates Ca2+ signal transduction mechanisms by binding to specific surface membrane receptors , activates PLC [31] , [40] and thus creating IP3 . Suggested membrane receptors for NSP4 , increasing intracellular Ca2+ are muscarinic receptors [31] , [40] but also α1β1 and α2β1 integrins [55] . The mechanism by which NSP4 may alter ER is unknown . It has been hypothesized that NSP4 acts as a viroporin [54] , forming a cation channel in the ER membrane or having a direct action on IP3 receptors in the ER membrane , with action or no action on the membrane itself [34] . Brunet and co-workers [34] reported that at a late stage of infection in Caco-2 cells , Ca2+ is partially increased by a PLC-dependent Ca2+ release from the ER through the opening of IP3-sensitive channels . However , they did not exclude an efflux of Ca2+ due to a direct alteration of the ER membrane . GOT1 cells are believed to respond with an increase of intracellular Ca2+ concentration upon stimulation with the AC activator isoprotenerol [43] . We confirmed that isoprotenerol did indeed stimulate the formation of cAMP , but no such effect was observed with NSP4 , suggesting that NSP4 does not induce the release of 5-HT through the AC pathway in these cells . Activation of the vagal afferent fibres by toxins in the gut appears to operate via detection of toxins by EC cells , which release 5-HT to activate 5-HT3 receptors on vagal afferent fibres [56] . Induction of Fos has previously been observed in vomiting animals [57] , [58] at the nucleus of the solitary tract ( NTS ) of CNS [58] . Our finding that RV-infected mice responded with a strong neural activation of the primary target site of vagal afferents , i . e . the NTS , is in line with the hypothesis that RV-induced activation of vagal afferents triggers vomiting . The fact that not all infected mice showed such activation ( 3/5 ) may be due to the time-kinetic variation of Fos activation between different animals . The peak period of Fos expression has shown to be 60–120 min after stimulation [59] , and prolonged for up to 6 hours . Another study showed a long-term increase in Fos expression in the area postrema , the NTS and the nucleus amygdalae in conjunction with vomiting after cisplatin treatment in a animal species with an emetic response ( the house musk shrew , Suncus murinus ) [60] . They also showed that Fos activation in these brain stem areas could be suppressed by palonosetron , a 5-HT3 receptor antagonist , which is used as an anti-emetic drug . The time kinetics of RV-induced Fos expression needs to be further investigated , since the onset of clinical symptom ( diarrhoea ) in mice started around 24 h p . i and persisted , at least up to 96 h . A limitation of using mice in these studies is the absence of a functional emetic reflex , but there are reports of “retching” but not vomiting , which may suggest that they have a degenerate reflex rather than none at all [56] , [61] . Further support that 5-HT is associated with RV disease was provided by the observation that 5-HT induced diarrhoea in 7/7 animals within 60 min which is consistent with our previous observation that 5-HT3 receptor antagonists attenuate RV-induced diarrhoea [3] . The participation of EC cells in diarrhoea , as revealed by studies of 5-HT release and/or use of pharmacological blocking agents , has been demonstrated in such diverse fluid secretory states as those caused by cholera toxin [62] , [63] , the enterotoxins produced by E coli [64] , bile salts [65] and Salmonella typhimurium [66] . Furthermore , an ENS involvement has been associated with cholera toxin [67] , E coli heat stable toxin [68] , certain bile salts [69] , and gut inflammation [70] . These observations indicate that the interaction between EC cells and ENS is a pathophysiological mechanism common to many intestinal secretory states . A Staphylococcal enterotoxin was recently reported to induce emesis by releasing 5-HT into the intestine , an effect inhibited by a 5-HT3 receptor antagonist [18] . Similarly , our data show that the NSP4 enterotoxin stimulated release of 5-HT and , sometimes , RV was localized to EC cells in the small intestine . Support for our hypothesis that RV-induced vomiting includes stimulation of EC cells , 5-HT and activation of vagal afferents is derived from several clinical studies with 5-HT3 receptor antagonists . For example , ondansetron , a 5-HT3 receptor antagonists , has successfully been used to attenuate vomiting in paediatric gastroenteritis [20] , [21] , [23] , [71] , [72] . Furthermore , it has also been reported that ondansetron-treatment of vomiting in American children has become quite common [21] . While it is established that vomiting during acute gastroenteritis in young children can be treated with 5-HT3 receptor antagonists , it remains to be shown in clinical studies that children with RV-induced vomiting can be successfully treated . Moreover , the emetic responses to the Staphylococcus toxin in the musk shrew animal model [18] , suggests that this animal model might be explored to study RV–induced vomiting . Our present and previous studies of the pathophysiology of RV infections [2] , [3] suggest a common triggering mechanism for the fluid loss and the emesis as schematically illustrated on Figure 8 . The results of the present study strongly suggest that RV per se and/or NSP4 released from adjacent virus-infected enterocytes increase intracellular Ca2+concentration in the EC cells , which , in turn , stimulates the release of 5-HT from EC cells . We propose that the released 5-HT activates both intrinsic and extrinsic afferent nerve fibres located in close vicinity to the EC cells . Hence , EC cells function as sensory transducers of different luminal stimuli . Such a mechanism has been demonstrated in several experimental models [8] , [9] , [42] . As pointed out above intrinsic afferent nerves stimulated by the released 5-HT are part of intramural nervous reflex ( es ) , which in the case of RV infection increase fluid secretion from intestinal crypts via the release of VIP ( vasoactive intestinal peptide ) at the crypt epithelium [3] , [4] . The released 5-HT also activates vagal afferents that project to the medullary vomiting centre of the central nervous system , triggering the emetic reflex [25] , [26] , [42] , [50] . It is apparent that EC cells and nerves play an important role for RV-induced diarrhoea and vomiting and the present findings may be of more general importance for our understanding of pathophysiological mechanisms of many different types of infection-induced diarrhoea and vomiting . Our observations may be of clinical importance , since the possibility to reduce vomiting by 5-HT3 receptor antagonists in acute viral gastroenteritis will both favour oral rehydration therapy by preventing vomiting and attenuate fluid loss , thus reducing hospitalisation of children [21] .
Primary tumor EC cells [43] and a human midgut carcinoid tumor cell line ( GOT1 ) , previously characterized for specific EC cell markers [43] , were cultured in RPMI 1640 medium supplemented with 10% FCS , 0 . 73 mg/ml L-glutamine , 5 µg/ml apo-transferrin , 5 µg/ml insulin , and PEST ( 100 U penicillin , 100 µg/ml Streptomycin ) . Primary EC cells were obtained from patients with liver metastasis and midgut carcinoid syndrome [43] . RRV was cultivated , quantified and purified as described [37] , [73] . The GOT1 cells ( 200 , 000/well ) were infected with trypsin-activated RRV at an MOI of 0 . 1 as previously described [73] . Briefly , after infection the cells were washed twice and then incubated with Minimal Essential Medium ( MEM ) containing trypsin ( T8353 , bovine pancreas , type III , Sigma Aldrich ) , 1 µg/ml medium , for 3 , 24 and 48 h . At each time point cells and supernatants were collected and frozen at −80°C . Cell suspensions were freeze-thawed twice , centrifuged to remove cell debris , followed by determination of the progeny virus , as previously described [74] . The GOT1 cells ( 450 , 000 cells/plate ) and primary EC t . c . ( 200 , 000 cells/plate ) were seeded onto coverslip-bottomed plastic Petri dishes used for fluorescent microscopy ( MatTek Corporation ) . The cells were washed twice with MEM without FCS and then loaded with the fluorescent Ca2+ indicator Fura-2-AM ( Molecular Probes ) , 10 µM in the presence of Pluronic-F 127 ( Sigma Aldrich ) , 20% w/v in DMSO , 10 µl/plate in a total volume of 1 ml MEM without FCS for 45 min at 37°C . The cells were washed twice with MEM without FCS and incubated in fresh medium . Ratiometric imaging of Fura-2-loaded cells was performed using a Photon Technology International ( Monmouth Junction , NJ ) system and a Zeiss Axiovert 100 M ( Jena , Germany ) microscope equipped with a ×100 glycerol-immersion Fluar objective ( 1 . 3 numerical aperture ) and a PTI IC-200 camera for fluorescence imaging . Before adding NSP4 to the cell cultures , initial 10 min fluorescence ( F340/F380 ) was captured to obtain a Ca2+-baseline . Bright-field images were taken simultaneously using a PTI IC-100 camera by passing the transmission light through a 700 nm band-pass filter in front of the halogen lamp to avoid stray light in the fluorescence channel . NSP4 were added to a final concentration of 125 nM and 2 µM , respectively , and continuously measured in real time for further 30 min . For measuring of intracellular Ca2+ using RRV see supporting information ( Protocol . S1 and S2 ) . The GOT1 and primary EC-cells were stained for 5-HT and chromogranin A . Cells were fixed with 4% paraformaldehyde/PBS on microscope slides ( Histolab , Göteborg , Sweden ) overnight at 4°C in a humidity chamber and then processed for immunofluorescence ( Protocol . S3 ) . Histidine-tagged NSP4 from a simian rotavirus strain ( SA11 ) was produced using the baculovirus expression vector system and Spodoptera frugiperda ( Sf9 ) insect cells . The NSP4 was purified by column chromatography , as previously described [41] . Secreted NSP4 was purified from the media of polarized epithelial Caco-2 cells infected with bovine rotavirus ( UK strain ) [39] . Briefly , the medium was ultracentrifuged to remove virions and NSP4 purified by sequential concanavalin A affinity chromatography and monoS cation exchange chromatography . The protein was judged as >99% pure by SDS PAGE and silver staining . The cholera toxin was purchased from List Biological Laboratories , Campbell , California; Code 101A . The 5-HT in cell culture medium from primary EC t . c . and GOT1 cells was quantified using a commercial serotonin ELISA kit ( IBL International , Hamburg , Germany; Code: RE59121 ) according to the manufacturer's instructions , or by HPLC [43] . GOT1 cells ( 650 , 000/well ) were pre-incubated for 30 min in MEM without FCS with ( 100 µM ) or without the AC inhibitor 2′ , 5′-Dideoxyadenosine , ( Sigma Aldrich; Code: D7408 ) , at 37°C with 5% CO2 . Recombinant or secretory NSP4 ( 2 µM ) , were added for 5 and 30 min . Cells were lysed and analysed for intracellular cAMP with a commercial cAMP EIA kit as described by the manufacture ( R&D systems , United Kingdom; Cat . No KGE002B ) . MEM without FCS was used as a negative control and Isoprotenerol ( 10 µM; Sigma Aldrich; Code:I6504 ) as a positive control . IP3 production was measured indirectly through the accumulation and analysis of the metabolite IP-one . GOT1 cells ( 650 , 000/well ) were stimulated with recombinant NSP4 or secretory NSP4 , 2 µM , for 60 min at 37°C with 5% CO2 in a buffer containing LiCl ( 50 mM ) to prevent degradation of IP-one . Carbachol ( Sigma Aldrich , USA ) , 100 µM , were used as positive control . Cells were lysed and analysed using a commercial IP-one ELISA kit ( Cisbio Bioassay , France; Code: 72IP1PEA ) , according to the manufacturer's instructions . RV naïve , five to seven days old BALB/c mice ( B&K Laboratories , Sollentuna , Sweden ) were orally infected with 10 µL/animal ( 100DD50 diarrhoea doses ) of murine rotavirus strain EDIM [3] , [46] . For mice receiving 5-HT ( Sigma Aldrich ) , 50 uL at 5 mg/kg were administered intraperionally and observed for signs of diarrhoea at 30 , 45 and 60 minutes after administration . The small intestines were removed and processed as previously described [3] , [46] . For immunohistochemistry of mice intestinal tissue , paraffin-embedded specimens were cut into thin sections as previously described [46] . Intestinal sections were hydrated and processed for immunohistochemistry ( Protocol . S4 ) . For immunoflourescence staining , the paraffin-embedded intestinal specimens were cut into thin sections , hydrated and processed for immunohistochemistry ( Protocol . S5 ) . Brains were cut at a freezing microtome at 30 micrometer and further processed for free floating immunohistochemistry . Sections were incubated with a primary antibody directed against Fos ( Santa Cruz Biotechnology; sc-52; 1∶1000 ) and the antibody was visualized using avidin-biotin complex amplification and DAB as chromogen according to previously published protocols [75] . The results are expressed as mean ± standard errors of the mean ( SEM ) , unless indicated . Statistical analysis of all data was performed using the repeated measures analysis . The Mann Whitney test was used unless stated . P-value of ≤0 . 05 was considered significant . All animal experiments in this study were carried out in strict accordance with the recommendations in the guide for the care and use of laboratory animals conformed to Swedish animal protection laws and applicable guidelines ( djurskyddslagen 1988:534; djurskyddsförordningen 1988:539; djurskyddsmyndigheten DFS 2004:4 ) . Animal experiments were approved by the local Ethical Committee ( Stockholm Norra Djurförsöksetiska nämnd , Stockholm , Sweden; Approval No: N289/09 ) . All efforts were made to minimize suffering and surgical operations were performed afterwards the animals were euthanized by overdose of Isoflurane . | Rotavirus ( RV ) can cause severe dehydration and is a leading cause of childhood deaths worldwide . While most deaths occur due to excessive loss of fluids and electrolytes through vomiting and diarrhoea , the pathophysiological mechanisms that underlie this life-threatening disease remain to be clarified . Our previous studies revealed that drugs that inhibit the function of the enteric nervous system can reduce symptoms of RV disease in mice . In this study we have addressed the hypothesis that RV infection triggers the release of serotonin ( 5-hydroxytryptamine , 5-HT ) from enterochromaffin ( EC ) cells in the intestine leading to activation of vagal afferent nerves connected to brain stem structures associated with vomiting . RV activated Fos expression in the nucleus of the solitary tract of CNS , the main target for incoming fibers from the vagal nerve . Both secreted and recombinant forms of the viral enterotoxin ( NSP4 ) , increased intracellular Ca2+ concentration and released 5-HT from EC cells . 5-HT induced diarrhoea in mice within 60 min , thereby supporting the role of 5-HT in RV disease . Our study provides novel insight into the complex interaction between RV , EC cells , 5-HT and nerves . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"virology",
"biology",
"microbiology",
"viral",
"diseases"
] | 2011 | Rotavirus Stimulates Release of Serotonin (5-HT) from Human Enterochromaffin Cells and Activates Brain Structures Involved in Nausea and Vomiting |
PKR is well characterized for its function in antiviral immunity . Using Toxoplasma gondii , we examined if PKR promotes resistance to disease caused by a non-viral pathogen . PKR−/− mice infected with T . gondii exhibited higher parasite load and worsened histopathology in the eye and brain compared to wild-type controls . Susceptibility to toxoplasmosis was not due to defective expression of IFN-γ , TNF-α , NOS2 or IL-6 in the retina and brain , differences in IL-10 expression in these organs or to impaired induction of T . gondii-reactive T cells . While macrophages/microglia with defective PKR signaling exhibited unimpaired anti-T . gondii activity in response to IFN-γ/TNF-α , these cells were unable to kill the parasite in response to CD40 stimulation . The TRAF6 binding site of CD40 , but not the TRAF2 , 3 binding sites , was required for PKR phosphorylation in response to CD40 ligation in macrophages . TRAF6 co-immunoprecipitated with PKR upon CD40 ligation . TRAF6-PKR interaction appeared to be indirect , since TRAF6 co-immunoprecipitated with TRAF2 and TRAF2 co-immunoprecipitated with PKR , and deficiency of TRAF2 inhibited TRAF6-PKR co-immunoprecipitation as well as PKR phosphorylation induced by CD40 ligation . PKR was required for stimulation of autophagy , accumulation the autophagy molecule LC3 around the parasite , vacuole-lysosomal fusion and killing of T . gondii in CD40-activated macrophages and microglia . Thus , our findings identified PKR as a mediator of anti-microbial activity and promoter of protection against disease caused by a non-viral pathogen , revealed that PKR is activated by CD40 via TRAF6 and TRAF2 , and positioned PKR as a link between CD40-TRAF signaling and stimulation of the autophagy pathway .
PKR , also known and eukaryotic translation initiation factor 2-alpha kinase 2 ( IEF2AK2 ) is a ubiquitously expressed serine-threonine kinase , constitutively present at low levels as inactive monomers in the cytoplasm of mammalian cells [1]–[3] . This kinase was discovered as a component of the interferon-inducible cellular antiviral defenses . PKR consists of a kinase domain ( KD ) and two tandem dsRNA binding domains ( dsRBD ) that regulate the kinase activity [2] , [3] . Under resting conditions , dsRBD interact with the KD maintaining the molecule in a closed , inactive form [4] , [5] . Binding of dsRNA to the dsRBD results in a conformational change that is believed to relieve the KD from the autoinhibitory effect of the dsRBD , allowing PKR to dimerize and autophosphorylate , thus becoming active [4] , [5] . Activated PKR can then catalyze the phosphorylation of its best-characterized substrate , the subunit of eukaryotic initiation factor 2 ( eIF-2α ) leading to inhibition of protein synthesis [1]–[3] . The antiviral role of PKR has been well characterized . dsRNA produced during infection with RNA and DNA viruses causes PKR activation with resulting eIF2α phosphorylation and inhibition in viral protein translation . Moreover , in vivo studies revealed that PKR restricts replication of viruses such as vesicular stomatitis virus [6] , [7] , lymphocytic choriomeningitis virus [8] and Herpes simplex virus 2 ( HSV-2 ) [9] . The importance of PKR in anti-viral immunity is emphasized by the fact that most animal viruses utilize various strategies to impair the action of PKR [3] . In addition to its role as a translational regulator , PKR is involved in signal transduction . PKR can signal to NF-κB , the signal transducer and activator of transcription ( STAT ) -1 and -3 , IFN regulatory factor ( IRF ) -1 , activating transcription factor ( ATF ) -3 and -4 , p53 AP-1 , Jun N-terminal protein kinase ( JNK ) and p38 [2] , [3] . In addition , dsRNA , cytokines ( IFN-γ , TNF-α , IL-1 ) [3] , [10] , [11] , LPS [12] , [13] and platelet-derived growth factor ( PDGF ) [14] can activate PKR . Moreover , the intracellular protein PKR-associated protein PACT ( also called RAX in mice ) can activate PKR in the absence of dsRNA [5] , [15] , [16] . While the role of PKR in antiviral immunity is well characterized , there is limited evidence for the involvement of this kinase during infections with non-viral pathogens . Induction of IL-6 and IL-12 p40 was defective in PKR−/− fibroblasts exposed to LPS [13] . In addition , serum levels of these two cytokines were reduced in PKR−/− mice challenged with LPS [13] . PKR promoted IL-6 and TNF-α production by mouse alveolar macrophages stimulated with TLR2 and TLR4 ligands [17] . Bacillus Calmette-Guerin ( BCG ) induced PKR-dependent IL-6 , IL-10 and TNF-α production by human monocytes [18] . More recently , PKR activation was shown to enhance in vitro replication of the protozoan Leishmania amazonensis in human and mouse macrophages , an effect that appears to be mediated by PKR-dependent IL-10 production [19] . In addition , PKR−/− mice exhibit improved in vivo control of Mycobacterium tuberculosis that is accompanied by increased apoptosis of infected macrophages and reduced macrophage production of IL-10 [20] . However , to our knowledge , it has not been reported whether PKR stimulates anti-microbial activity against a non-viral pathogen and enhances resistance against disease caused by such a pathogen . Toxoplasma gondii is an obligate intracellular protozoan parasite that infects an estimated 30% of the human population worldwide . Tachyzoites , the invasive form of the parasite , penetrate host cells and reside within parasitophorous vacuoles that resist lysosomal fusion thereby avoiding eradication [21] . Tissue cysts are formed primarily in the brain and skeletal muscle during the chronic phase of infection and persist in the host for life . Infection with T . gondii can cause severe illness in children with congenital infection and in immunocompromised adults . Toxoplasmic encephalitis and ocular toxoplasmosis are two important manifestations of toxoplasmosis . Host protection against T . gondii infection is mediated primarily by T cell-mediated immunity [22]–[24] . IFN-γ , TNF-α and NOS2 are major mediators of resistance to both toxoplasmic encephalitis and ocular toxoplasmosis [25]–[31] . Using models of T . gondii infection we report that PKR contributes to protection against ocular and cerebral toxoplasmosis , triggers anti-microbial activity against this pathogen in macrophages and microglia and we identified molecular events involved in induction of this activity .
To begin to explore the relevance of PKR during T . gondii infection , wild-type ( B6 ) and PKR−/− mice were infected with tissue cysts of the ME49 strain of T . gondii . Parasite load as assessed by qPCR for the T . gondii B1 gene was examined at different times throughout the first month post-infection . B6 and PKR−/− mice had similar T . gondii parasite loads in the spleen , liver and lung at days 3 , 7 , 14 and 28 post-infection and both strains of mice were able to restrict the parasite load in these organs ( Table 1 ) . In contrast to peripheral organs , the parasite loads in the brain and eye were significantly higher in PKR−/− compared to B6 mice , a difference that became more pronounced at day 28 post-infection ( Table 1 ) . At approximately this time , PKR−/− mice exhibited piloerection and hunched posture . Brain homogenates of PKR−/− mice collected at this time contained significantly greater numbers of tissue cysts than control B6 mice ( Figure 1A ) . Histopathological examination at 4 weeks post-infection revealed that while there was minimal inflammation in brain sections from B6 mice ( Figure 1B ) , brains of PKR−/− mice exhibited a significant increase in perivascular inflammation , microglial nodules and presence of numerous tissue cysts ( p<0 . 01 ) ( Figure 1C ) . Areas of acute focal inflammation were noted ( Figure 1D ) in which tachyzoites and parasite antigens were detected ( Figure 1E ) . Also , while B6 mice revealed minimal histopathological changes in the retina ( Figure 1F ) , eyes from PKR−/− mice had remarkable histopathological changes characterized by presence of inflammatory cells in the vitreous and retina including perivascular inflammation , distortion of the retinal architecture , hypertrophy of the retinal pigment epithelial cells ( RPE ) and invasion of the retina and vitreous by RPE ( p<0 . 01 ) ( Figure 1G , H ) . In addition , infected PKR−/− mice exhibited increased mortality ( Figure 1I ) . Thus , PKR promotes resistance against ocular and cerebral toxoplasmosis . PKR can enhance cytokine and NOS2 expression [13] , [17] , [18] , [32]–[34] . IFN-γ is critical for control of T . gondii , and the production of this cytokine is dependent on IL-12 [25] , [35]–[37] . TNF-α is also important for protection [30] , [31] . mRNA levels of these cytokines were similar in infected B6 and PKR−/− mice , with the exception of IFN-γ mRNA levels that were higher in the brain and eye of infected PKR−/− mice ( Figure 2A ) . Moreover , serum levels of IFN-γ , IL-12 and TNF-α were similar in infected B6 and PKR−/− mice ( Figure 2B ) . Splenocytes from PKR−/− mice produced higher amounts of IFN-γ and IL-12 in response to T . gondii lysate antigens ( Figure 2C ) . We also examined Immunity-related GTPases ( IRG ) expression and nitric oxide production , key effector molecules downstream of IFN-γ . Expression of IRGM3 was similar in the spleens and lungs from infected B6 and PKR−/− mice ( Figure 3A ) . Nitric oxide production by splenocytes incubated with T . gondii lysate antigens was higher in PKR−/− mice ( Figure 3B ) . Data shown on serum cytokine levels , cytokine and nitric oxide production by splenocytes and expression of IRGM3 are from samples obtained on day 7 post-infection . Samples collected on day 14 post-infection also revealed that the expression of these molecules was similar in B6 and PKR−/− mice ( not shown ) . In addition to IFN-γ , TNF-α , cerebral and ocular mRNA levels of IL-6 , NOS2 and IL-10 are increased in the brain and eye of mice infected with T . gondii , and these molecules promote protection against toxoplasmosis while in the case of IL-10 , this cytokine modulates susceptibility to disease [25]– . We conducted a separate set of experiments to determine whether less effective parasite control in PKR−/− mice might be due to defective expression of mediators of protection in the brain and eye or changes in IL-10 expression . As shown before , brains of infected PKR−/− mice exhibited significantly higher mRNA levels of IFN-γ at 4 weeks post-infection ( Figure 4A ) . The levels of TNF-α , NOS2 and IL-6 in PKR−/− mice were comparable or even higher than those from B6 mice ( Figure 4A ) . Assessment of expression of these molecules in the eye revealed similar results ( Figure 4B ) . The mRNA levels of IL-10 in the brain and eye of PKR−/− mice did not differ from B6 mice ( Figure 4A–B ) . Taken together , PKR−/− mice are more susceptible to T . gondii despite unimpaired expression of IFN-γ , TNF-α , NOS2 and IL-6 and lack of changes in IL-10 expression . In addition , it is unlikely that PKR deficiency promotes toxoplasmosis by impairing type I IFN signaling since we could not detect defective parasite control in the brains and eyes of IFN-α/βR−/− mice ( data not shown ) . PKR can control cellular proliferation , differentiation and apoptosis [41] , [42] . Accordingly , we examined whether lack of PKR could perturb the cellular composition of a lymphoid organ . The frequencies of CD4+ T cells , CD8+ T cells , NK cells , monocytes and B lymphocytes were similar in the spleens of PKR−/− and B6 mice confirming previous findings [43] ( Figure 5A; p>0 . 05 ) . Next , we examined the expansion of IFN-γ-producing T cells . Intracellular IFN-γ expression was examined after splenocytes were incubated with anti-CD3 mAb [44] . In contrast to splenocytes from uninfected animals , splenocytes from infected mice showed an expansion in the percentages of CD4+ T cells and CD8+ T cells that expressed IFN-γ as well as in the absolute numbers of IFN-γ-producing T cells ( Figure 5B ) . The percentages and numbers of IFN-γ+ T cells were similar in B6 and PKR−/− mice ( Figure 5B ) . Similarly , the percentages of IFN-γ+ CD4+ and CD8+ T cells as well as the absolute numbers of these cells were comparable in brain mononuclear cells from infected B6 and PKR−/− mice ( Figure 5B ) ( p>0 . 2 ) . These data indicate that lack of PKR does not affect the phenotypic composition of a lymphoid organ and does not impair the expansion of IFN-γ-producing T cells . PKR can modulate antibody production [45] and B cells play a protective role against T . gondii [46] . However , the levels of anti-T . gondii IgG antibodies as assessed by ELISA were similar in infected B6 and PKR−/− mice . The anti-T . gondii IgG antibodies titers were 1∶25 , 600 in both groups of mice and the O . D . values ( 450 nm ) at a serum dilution of 1∶200 were comparable ( Figure 5C ) . Macrophages and microglia are key effector cells that mediate resistance to T . gondii [26] , [47] , [48] . Thus , studies were performed to determine whether PKR is required for killing of T . gondii in these cells . Interestingly , treatment with IFN-γ plus TNF-α induced killing of T . gondii as efficiently in macrophages ( Figure 6A–B ) and microglia ( Figure 6C ) derived from PKR−/− mice as in cells from B6 mice . These findings suggested a role for PKR in regulating another aspect of immune response to T . gondii . CD40 and its ligand CD154 are central for protection against ocular and cerebral toxoplasmosis [49] , [50] and CD40 ligation activates macrophages and microglia to acquire anti-T . gondii activity [49] , [51]–[54] . Accordingly , we assessed the role of PKR in the CD40-induced anti-T . gondii activity . Regardless of whether macrophages were infected with a type I ( RH ) or type II ( ME49 ) strain of T . gondii , CD40 ligation caused a marked decrease in the number of tachyzoites in macrophages ( Figure 6A–B ) . Anti-T . gondii activity was the same regardless of whether CD40 ligation took place before or after infection ( not shown ) [53] . CD40 ligation also caused anti-T . gondii activity in primary brain microglia ( Figure 6C ) from B6 mice . On the contrary , the parasite load did not decrease in CD40 stimulated macrophages and microglia from PKR−/− mice . The lack of anti-T . gondii activity by PKR−/− macrophages/microglia was not due to defective expression of CD40 by these cells as assessed by flow cytometric analysis ( data not shown ) . To further characterize the role of PKR in CD40-induced killing of T . gondii we utilized RAW 264 . 7 cells that express a chimera that consists of the extracellular domain of human CD40 and the intracytoplasmic domain of mouse CD40 ( hmCD40 ) [53] . HmCD40-RAW 264 . 7 cells transiently transfected with plasmids encoding WT-PKR , DN-PKR ( K296R ) or empty plasmid were incubated with or without CD154 followed by challenge with T . gondii . While CD40 stimulation induced killing of T . gondii in cells transfected with either empty plasmid or wild-type PKR , anti-T . gondii activity was impaired in cells expressing DN-PKR ( Figure 6D ) . Next , we determined whether PKR was relevant for controlling T . gondii in human macrophages . Prior to stimulation with CD154 , human monocyte-derived macrophages ( MDM ) were treated with vehicle or 2-amino purine ( 2-AP ) , a pharmacological inhibitor of PKR kinase activity . Cells were then challenged with T . gondii . Stimulation with CD154 induced anti-T . gondii activity in MDM treated with vehicle alone . In contrast , 2-AP ablated anti-T . gondii activity in response to CD154 stimulation ( Figure 6E ) . Taken together , these findings indicate that PKR is required for CD40-induced anti-T . gondii activity in macrophages and microglia but is dispensable for the IFN-γ/TNF-α arm of resistance to T . gondii in these cells . We determined whether CD40 ligation causes activation of PKR signaling in macrophages . Bone marrow derived macrophages from B6 mice incubated with a stimulatory anti-CD40 mAb exhibited PKR phosphorylation as assessed by immunoblot ( Figure 7A ) . Similar results were obtained with monocyte-derived macrophages from humans ( data not shown ) . In addition , CD40 stimulation caused phosphorylation of eIF2α , a signaling molecule classically activated by PKR ( Figure 7A ) . PACT , TNF-α , IL-1 and IFNs can activate PKR [3] , [5] , [10] , [11] , [15] , [16] . However , bone marrow-derived macrophages from PACT−/− , TNF-α−/− , IL-1R1−/− , IFN-α/βR−/− and IFN-γ−/− mice were not defective in phosphorylation of PKR in response to CD40 stimulation ( Figure 7B ) . Thus , CD40 ligation induced phosphorylation of PKR that was independent of PACT , TNF-α , IL-1 , IFN-α/β and IFN-γ . The cytoplasmic tail of CD40 lacks intrinsic catalytic activity and signals through its ability to recruit TNF receptor-associated factors ( TRAFs ) [55] , [56] . Membrane-distal domains of CD40 directly bind TRAF2 and TRAF3 ( TRAF3 inhibits CD40 signaling ) whereas TRAF6 binds to a different membrane-proximal domain [55] , [56] . CD40-induced toxoplasmacidal activity in macrophages is dependent exclusively on the TRAF6 binding site of CD40 [52] , [54] . In order to examine the role of TRAF binding sites on PKR phosphorylation , RAW 264 . 7 cells that express WT hmCD40 or hmCD40 with mutations at the TRAF2 , 3 binding sites ( ΔT2 , 3 ) , TRAF6 binding site ( ΔT6 ) or TRAF2 , 3 plus TRAF6 binding sites ( ΔT2 , 3 , 6 ) [54] were stimulated with CD154 . RAW 264 . 7 cells that express WT hmCD40 or hmCD40 with a mutation that disrupts binding to TRAF2 , 3 ( ΔT2 , 3 ) exhibited unimpaired phosphorylation of PKR ( Figure 8A ) . In contrast , phosphorylation of PKR was impaired in RAW 264 . 7 cells that express hmCD40 with a mutation that disrupts binding to TRAF6 ( ΔT6 ) or mutations that disrupt binding to TRAF6 as well as TRAF2 , 3 ( ΔT2 , 3 , 6 ) ( Figure 8A ) . These results could not be explained by differences in the levels of CD40 expression ( Figure 8B; p>0 . 1 ) . Thus , PKR phosphorylation in response to CD40 ligation was dependent on the TRAF6 binding site . PKR can associate with TRAF proteins including TRAF2 , TRAF5 and TRAF6 [57] . We therefore hypothesized that TRAF6 links CD40 to PKR signaling . To test this hypothesis , FLAG-tagged PKR was transiently expressed in WT hmCD40-RAW 264 . 7 cells . Following stimulation with CD154 , PKR immunoprecipitated with TRAF6 ( Figure 8C ) . Interestingly , while PKR has been reported to contain binding motifs for TRAF2 , 3 [57] , no TRAF6 binding motifs are identifiable in mouse PKR . However , TRAFs can form heterocomplexes [58] , [59] . Thus , we examined whether CD40-induced association of TRAF6 and PKR is dependent on TRAF2 . Following stimulation of WT hmCD40-RAW 264 . 7 cells with CD154 , endogenous TRAF2 immunoprecipitated with endogenous TRAF6 ( Figure 8D ) . In addition , TRAF2 immunoprecipitated with FLAG-tagged PKR ( Figure 8E ) . Next , we examined the effects of TRAF2 knockdown to further determine the role of TRAF2 in the CD40-induced association between TRAF6 and PKR . Transfection of hmCD40-RAW 264 . 7 cells with TRAF2 siRNA effectively diminished TRAF2 protein levels ( Figure 8F ) . TRAF2 knockdown diminished immunoprecipitation of TRAF6 and FLAG-tagged PKR in response to CD40 stimulation ( Figure 8F ) . To further explore the role of TRAF2 in CD40-induced PKR phosphorylation wt MEF and TRAF2−/− MEF that stably express hmCD40 were incubated with CD154 . Whereas CD154 induced PKR phosphorylation in wt MEF , this effect was not observed in TRAF2−/− MEF ( Figure 8G ) . Taken together , while the TRAF2 , 3 binding sites of CD40 do not play an appreciable role in CD40-induced activation of PKR , the association of TRAF6 with PKR and the activation of PKR were dependent on TRAF2 . CD40 stimulation of T . gondii-infected macrophages and microglia leads to fusion of the parasitophorous vacuole with late endosomes/lysosomes , leading to lysosomal degradation and killing of the parasite [50] , [53] , [54] . This process is dependent on the autophagy machinery [50] , [53] , [54] . Autophagy is a conserved cellular homeostatic process whereby a double membrane autophagosome sequesters portions of the cytoplasm and damaged organelles and fuses with lysosomes culminating in the formation of an autolysosome and enzymatic degradation of its cargo [60] . The autophagy pathway is important for the control of T . gondii not only in vitro but also in vivo [50] . Accordingly , we examined whether PKR is required for stimulation of autophagy induced by CD40 stimulation of macrophages . We utilized a plasmid that encodes tandem fluorescent LC3 ( tfLC3; RFP-GFP-tagged LC3 ) that enables to monitor both the presence of autophagosomes and the flux to autophagosome fusion with lysosomes ( autolysosomes ) [61] . CD40 stimulation of hmCD40 RAW 264 . 7 cells that expressed WT-PKR and were transfected with tfLC3 caused an increase in the percentages of cells with LC3+ autophagosomes and autolysosomes indicative of enhanced autophagy flux ( Figure 9A ) . However , autophagy flux was markedly impaired in cells expressing DN-PKR ( Figure 9A ) . Inhibition of PKR signaling did not affect the enhanced autophagy triggered by rapamycin ( not shown ) . Next , we investigated whether PKR is required for recruitment of LC3 around the parasite . HmCD40-RAW 264 . 7 cells expressing LC3-EGFP plus either WT-PKR or DN-PKR were incubated with or without CD154 followed by challenge with transgenic T . gondii tachyzoites that express cytoplasmic RFP . CD40 stimulation of cells expressing WT-PKR resulted in accumulation of LC3 around the parasite ( Figure 9B ) . In contrast , accumulation of LC3 was abrogated in cells expressing DN-PKR , pointing to the relevance of PKR in the CD40-induced recruitment of the autophagy protein LC3 around the parasite ( Figure 9B ) . We previously showed that CD40 ligation results in fusion of late endosomes/lysosomes with the parasitophorous vacuole and killing of T . gondii [53] . Indeed , T . gondii infected bone marrow-derived macrophages or primary brain microglia from B6 mice exhibited accumulation of the late endosomal/lysosomal marker LAMP-1 around the parasite after CD40 ligation ( Figure 10 ) . In contrast , LAMP-1 accumulation was markedly impaired in cells derived from PKR−/− mice ( Figure 10 ) . Taken together , PKR links CD40 to stimulation of the autophagy pathway , recruitment of the autophagy protein LC3 around T . gondii and vacuole-lysosomal fusion , the mechanism by which CD40 has been reported to mediate killing of T . gondii in macrophages and microglia [50] , [53] .
While the role of PKR in antiviral immunity has been extensively characterized , the involvement of PKR in mechanisms of protection against non-viral pathogens remains underexplored . We report herein that PKR−/− mice exhibited increased parasite load and worsened histopathology in the eye and brain after infection with T . gondii . This was accompanied by impaired ability of macrophages and microglia to control the parasite in response to CD40-CD154 stimulation , molecules important for protection against the parasite in the eye and brain . Furthermore , we identified TRAF6 and TRAF2 as molecular links between CD40 and PKR activation . These findings indicate that PKR plays an important role in activation of mechanisms of protection against a non-viral pathogen . In addition to induction of anti-viral activity in host cells via transcriptional inhibition , PKR has been reported to promote cytokine production , enhance nitric oxide production and appears to promote viral clearance mediated by CD8+ T cells [8] , [13] , [17] , [18] , [32]–[34] . We found no evidence of impaired expression of IFN-γ in PKR−/− infected with T . gondii . Indeed , brain/eye IFN-γ mRNA levels were higher in these animals compared to controls . This phenomenon could be explained by the higher parasite loads detected in infected PKR−/− mice and/or by a potential role of PKR in regulating transcription of IFN-γ mRNA [62] . Similar to studies of IFN-γ mRNA levels , the expression of IFN-γ+ CD4+ and CD8+ T cells was not impaired in PKR−/− mice . This is relevant because CD4+ and CD8+ T cells are considered to mediate resistance against the parasite by producing IFN-γ [63]–[65] . Pertinent to our findings , the induction of LCMV-reactive CD8+ T cells is not defective in PKR−/− mice [8] . In addition to IFN-γ production , CD8+ T cells exhibit cytotoxic activity against T . gondii-infected cells [66] , [67] . However , it would appear unlikely that impaired induction of CD8+ T cell cytotoxic activity represents the major mechanism by which PKR promotes resistance to toxoplasmosis . PKR−/− mice have increased tissue cysts in the brain and more severe encephalitis . In contrast , while CD8+ T cells that express perforin diminish the numbers of tissue cysts in the brain [68] and perforin−/− mice infected with T . gondii exhibit higher tissue cyst numbers [65] , these animals do not exhibit worse histopathology [65] . TNF-α , NOS2 and IL-6 are important mediators of protection against T . gondii in the brain and eye [26]–[31] , [39] . However , we did not detect a defect in expression of these molecules in T . gondii-infected PKR−/− mice . In addition to promoting cytokine production , PKR has been reported to mediate activation of NF-κB , MAPK and Akt in cells treated with IFN-γ , TNF-α and/or the combination of these cytokines [11] , [69] , [70] . Nevertheless , the induction of anti-T . gondii activity in response to IFN-γ/TNF-α was unimpaired in macrophages/microglia from PKR−/− mice , a mouse macrophage line that expresses DN PKR and in human macrophages treated with 2-AP . Relevant to our results is the evidence that PKR plays a selective role in cytokine signaling since there are responses triggered by IFN-γ and TNF-α can occur independently of PKR [70]–[72] . In addition to being an IFN stimulated gene , PKR can promote type I IFN production [73] . However , our studies with IFN-α/βR−/− mice argue against defective type I IFN signaling as explaining susceptibility to cerebral and ocular toxoplasmosis in PKR−/− mice . The CD40 - CD154 pathway activates macrophages/microglia to acquire anti-T . gondii activity [49]–[53] . Our studies revealed that in contrast to IFN-γ/TNF-α , CD40 stimulation requires PKR for induction of anti-T . gondii activity in macrophages/microglia . Relevant to this differential role of PKR is the evidence that CD40 does not require IFN-γ to activate macrophages/microglia to kill the parasite [50] , [52] . However , despite the latter findings , the CD40-CD154 pathway functions in synergy with IFN-γ to enhance resistance to T . gondii in vitro and also likely in vivo [50] , [51] , [74] . The role of PKR in mediating anti-T . gondii activity induced by CD40 ligation may provide an explanation for susceptibility to ocular and cerebral toxoplasmosis in PKR−/− mice since macrophages/microglia are considered to be key effectors of protection against the parasite in neural tissue and CD40−/− and CD154−/− mice are susceptible to these forms of the disease [26] , [47]–[50] . Ligands for CD40 in the eye and brain of T . gondii infected mice may include not only infiltrating T cells , but potentially non-T cells that acquire CD154 expression as a result of cytokine stimulation as well as HSP70 , a molecule upregulated in T . gondii-infected cells [75]–[77] . CD40 ligation induces phosphorylation of PKR in macrophages . PKR mediates signaling induced by TNF-α , IFN-γ , IFN-α/β and IL-1β [3] , [10] , [11] , [69]–[71] . Moreover , CD40 enhances production of TNF-α , IL-1β and likely IFN-α [78] , [79] . However , studies using macrophages from TNF-α−/− , IL-1R1−/− , IFN-γ−/− and IFN-α/βR−/− mice indicate that PKR activation induced by CD40 is unlikely to be mediated by autocrine secretion of these cytokines . PACT/RAX is considered to be the intracellular mediator that links a wide variety of cellular stresses to PKR activation [15] , [16] . Our studies also indicate that PKR phosphorylation induced by CD40 was independent of PACT/RAX . Moreover , CD40-induced killing of T . gondii was unimpaired in bone marrow derived macrophages from PACT−/− mice ( Portillo et al; unpublished observations ) . These findings indicate that there is a mechanism distinct from cytokine secretion and PACT activation by which CD40 induces PKR phosphorylation . The cytoplasmic tail of CD40 lacks intrinsic kinase activity and therefore signals through recruitment of adaptor proteins . TRAFs are central mediators of CD40 signaling . Our data herein show that the TRAF6 binding site in the intra-cytoplasmic tail of CD40 is required for phosphorylation of PKR . These results are consistent with the pivotal role of the TRAF6 binding site in the induction of anti-T . gondii activity in CD40-activated macrophages [52] . TRAF recruitment to cytoplasmic domains of receptor molecules can lead to assembly of larger signaling complexes . Indeed , in our immunoprecipitation experiments , TRAF6 was observed to interact with PKR in response to CD40 ligation . While TRAFs are proposed to act downstream of human PKR and mediate NF-κB activation [57] , we are unaware of studies on TRAFs as upstream regulators of PKR . Human PKR can interact with TRAFs in HeLa and 293T cells infected with vaccinia virus expressing PKR , transfected with PKR- or TRAF-encoding plasmids or treated with IFN-α/β [57] . Two putative TRAF-binding sites exist in human PKR: one in the dsRBD II subdomain ( TKQE ) and another in the KD ( PEQIS ) [57] . Both sites have been reported to interact with TRAF2 [57] . However , whereas the TRAF interacting site in dsRBD II is preserved in mouse PKR , the KD site in mice exhibits an altered motif ( PEQLF ) . The presence of phenylalanine in the C-terminus of this motif is predicted to ablate recruitment of TRAF2 [80] . Thus , TRAF-PKR interaction in mice would most likely occur at the level of the dsRBD II subdomain since no other putative TRAF binding domain is apparent . Binding of TRAF proteins to dsRBD II subdomain could potentially result in an open ( active ) PKR conformation given that the dsRBD – KD interaction keeps PKR in a closed ( inactive ) form . In this regard , PACT is believed to activate PKR by binding to the site in KD that interacts with dsRBD II subdomain resulting in allosteric changes in PKR and an open conformation [5] . Of importance to our studies , no apparent TRAF6 binding site is detected in mouse PKR [81] . This suggested the possibility that the interaction between TRAF6 and PKR is indirect via TRAF2 . Indeed , our studies indicate that upon CD40 ligation there is TRAF6-TRAF2 and TRAF2-PKR interaction . Moreover , TRAF2 deficiency impairs CD40-induced TRAF6-PKR association and PKR phosphorylation even though the TRAF2 , 3 binding site of CD40 plays no appreciable role in PKR activation . Of relevance to our studies , TRAFs can form heterocomplexes through their TRAF domains . TRAF3 can interact with TRAF5 [58] , while TRAF2 may recruit TRAF6 to the cytoplasmic tail of CD40 in non-hematopoietic cells [59] . It has recently been reported that CD40 appears to induce PKR phosphorylation in B cells although the molecular mechanisms responsible for this effects were not elucidated [45] . It remains to be determined if CD40-induced TRAF6-TRAF2 signaling may be responsible for PKR activation in these cells . Autophagy can act as an anti-microbial mechanism against several pathogens including T . gondii [82] . CD40 stimulation of macrophages and microglia results in killing of T . gondii dependent on the autophagy proteins Beclin 1 and Atg 7 [50] , [53] , [54] . PKR signaling has been implicated in regulation of the autophagy pathway . PKR controls autophagy triggered by starvation [83] . In addition , PKR can modulate autophagy in response to virus infection . Using long-lived protein degradation as an indicator of autophagy activity , PKR was shown to promote autophagy in response to HSV-1 infection , a process that is inhibited by the HSV-1 neurovirulence factor ICP34 . 5 , a protein that antagonizes autophagy by binding to Beclin 1 [84] . While wild-type mice infected with a mutant HSV-1 that lacks the neurovirulence factor ICP34 . 5 do not develop encephalitis , PKR−/− mice are susceptible to this disease suggesting a role of autophagy in protection against HSV-1 encephalitis [85] . In this study , we identified PKR as a molecular link between CD40 and the autophagy pathway since PKR is required for the CD40-induced autophagy flux , the accumulation of the autophagy protein LC3 around T . gondii , vacuole-lysosomal fusion and killing of the parasite . PKR appears to promote autophagy induced by selective stimuli since we have not observed a significant role for PKR in autophagy enhanced by rapamycin . While CD40 likely functions in synergy with IFN-γ to enhance resistance to T . gondii in vivo , the role of PKR as a link between CD40 and killing of T . gondii via the autophagy pathway may contribute to increase the resistance to ocular and cerebral toxoplasmosis . T . gondii can impair the effects of IFN-γ by manipulating cell signaling in host cells and by inducing in vivo production of cytokines that can have antagonistic activity against IFN-γ [86] . The presence of a process such as CD40-induced autophagy that can occur in the absence of IFN-γ and that is likely under different regulation than IFN-γ-induced anti-T . gondii activity may explain a role in enhancing protection in the setting of parasite-induced impairment of optimal IFN-γ signaling . CD40-TRAF6 signaling triggers macrophage anti-T . gondii activity that is dependent on autocrine production of TNF-α [51] . However , TNF-α alone is not sufficient to induce anti-T . gondii activity in macrophages [52] indicating that induction of this activity requires synergy between CD40-TRAF6-induced TNF-α production and additional signals downstream of TRAF6 [53] , [54] . Our studies indicate that PKR is unlikely to be activated by autocrine TNF-α production . However , we cannot rule out a potential role of PKR in mediating signals downstream of TNF-α . While PKR can regulate various signaling pathways and modulates many cellular responses , one of the best-characterized roles of PKR is that of restricting viral replication . This study uncovered a role for PKR in inducing anti-microbial activity against a non-viral pathogen and enhancing protection against disease caused by this organism . In addition , this study reports a molecular link between CD40 ( an important mediator of resistance to pathogens ) , TRAF2 , TRAF6 , PKR and activation of anti-microbial activity in macrophages/microglia .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee of Case Western Reserve University School of Medicine ( Protocol Number 2009-0095 ) . C57BL/6 ( B6 ) mice were purchased from Jackson Laboratories . B6 and PKR−/− mice ( B6 background; lacking the majority of the amino-terminal dsRBD [87] ) were maintained at the Animal Resource Center ( Case Western Reserve University ) and the Biological Resources Unit ( Lerner Research Institute , Cleveland Clinic ) . Female mice were 8–12 weeks old when used for the studies ( 4–8 mice per group ) . Mice were infected intraperitoneally ( i . p . ) with 10 cysts of the ME49 strain of T . gondii ( gift from Dr . George Yap , University of Medicine and Dentistry of New Jersey ) . In addition to B6 and PKR−/− mice , PACT−/− ( Lerner Research Institute , Cleveland Clinic ) , IFN-γ−/− , TNF-α−/− , IL-1R1−/− , BALB/c ( all from Jackson Laboratories ) , IFN-α/βR−/− ( gift from Dr . Clifford Harding , Case Western Reserve University ) and 129SvEv mice ( Taconic Farms ) were used to obtain bone marrow-derived macrophages . Tachyzoites of the RH and PTG-ME49 strain were maintained in human foreskin fibroblasts . Transgenic parasites expressing cytoplasmic yellow fluorescent protein ( YFP ) or cytoplasmic DsRed ( RFP ) have been described [88] , [89] . Animals were anesthetized , perfused with PBS and euthanized . Four 5 µM sections of different areas of the brains and eyes were stained with periodic acid Schiff hematoxylin ( PASH ) or hematoxylin and eosin stain respectively . Histopathologic changes were scored following previously described criteria [26] , [27] . In addition , immunohistochemistry to detect T . gondii parasites and antigens was performed as described [31] , [38] . Total RNA was isolated from brains and eyes using RNeasy kit ( QIAGEN ) according to the manufacturer's protocol . RNA ( 0 . 5 µg ) was treated with DNase ( Ambion ) and reverse transcribed to cDNA with Super-Script III reverse transcriptase ( Invitrogen ) and oligo ( dT ) 12–18 primers ( Invitrogen ) . cDNA ( 2 . 5 µl ) was used as template for quantitative RT-PCR using SYBR GREEN PCR mix ( Applied Biosysytems ) and 20 pM of primer in 50 µl . Primer sequences for IFN-γ [90] , TNF-α [90] , NOS2 [91] , IL-6 [92] , IL-10 [90] , IL-12 p40 [93] and 18S rRNA [94] were previously described . Gene expression was assessed using 7300 Real Time PCR System ( Applied Biosystems ) . Each sample was run in duplicate and normalized to the content of 18S rRNA [50] . Genomic DNA was isolated from organs using DNeasy kit ( QIAGEN ) and subjected to quantitative RT-PCR using SYBR GREEN PCR mix . A standard curve of DNA from 1 to 105 ME49 tachyzoites per reaction was used to quantitate parasite load . Each sample was run in triplicate [50] . Splenocytes were stained with anti-CD3 , anti-CD4 , anti-CD8 , anti-CD40 , anti-CD45R , anti-CD49d ( DX5 ) , anti-CD11b or isotype control mAb ( all from eBiosciences ) . Cells were fixed with 1% paraformaldehyde and analyzed by use of an LSR II flow cytometer ( BD Biosciences ) . Expression of intracellular cytokines was assessed in splenocytes and brain mononuclear cells . The latter cells were isolated as previously described [50] . Splenocytes obtained at 7 and 14 d post-infection as well as brain mononuclear cells obtained at 28 d post-infection were incubated with or without anti-CD3 plus Brefeldin A ( 10 µg/ml; eBiosciences ) as described [44] . Cells were first stained with anti-CD3 , anti-CD4 or anti-CD8 . Cells were permeabilized using IntraPrep permeabilization reagent ( Counter-Immunotech ) , following the manufacturer's protocol . Cells were then stained with anti-IFN-γ or anti-IL-4 mAb ( eBiosciences ) . After fixation with 1% paraformaldehyde , cells were analyzed by use of an LSR II flow cytometer ( BD Biosciences ) . Primary bone marrow-derived macrophages and brain microglia were obtained from control or PKR−/− mice as described [50] , [95] . Prior to infection , these cells were incubated overnight with isotype control or stimulatory anti-CD40 mAb ( 1C10; 10 µg/ml ) or with IFN-γ ( 100 U/ml; PeproTech ) plus TNF-α ( 250 pg/ml; PeproTech ) . RAW 264 . 7 cells stably transfected with linearized pRSV . 5 plasmid encoding a chimera of the extracellular domain of human CD40 and the intracytoplasmic domain of mouse CD40 ( hmCD40-RAW 264 . 7 ) were previously described [53] . HmCD40-RAW 264 . 7 cells were treated with or without CD154 ( 3 µg/ml; gift from W . Fanslow , Amgen , Thousand Oaks , California , USA ) prior to infection with T . gondii . Human monocyte-derived macrophages were obtained as described [51] and were incubated with the PKR inhibitor , 2-aminopurine ( 2-AP; 2 mM; Sigma ) or vehicle for 30 minutes followed by incubation with or without CD154 . Tachyzoites of the RH or PTG-ME49 strains of T . gondii were used to infect monolayers as described [95] . Monolayers were fixed with Diff-Quick ( Dade Diagnostics ) and the number of tachyzoites per 100 cells was determined by light microscopy by counting at least 200 cells per monolayer as previously described [95] . HmCD40-RAW 264 . 7 cells were transiently transfected with a plasmid that encodes either FLAG-tagged wild-type ( WT ) -PKR , dominant negative ( DN ) -PKR ( K296R ) , empty plasmid ( gifts from Bill Sudgen , University of Wisconsin ) , TRAF2 siRNA [96] or control siRNA ( Dharmacon ) using an Amaxa Nucleofector ( Amaxa ) according to the manufacturer's protocol . For assessment of autophagy , hmCD40-RAW 264 . 7 cells were transfected with LC3-EGFP or a plasmid encoding tandem monomeric RFP-GFP-tagged LC3 ( tfLC3 ) [61] ( gifts from T . Yoshimori , National Institute for Basic Biology , Okazaki , Japan ) . Parent RAW 264 . 7 cells ( >96% CD40− ) were transduced with previously described EGFP-encoding MIEG3 retroviral vectors that encode WT hmCD40 or hmCD40 with mutations at the TRAF2 , 3 binding site ( ΔT2 , 3 ) , TRAF6 binding site ( ΔT6 ) or TRAF2 , 3 plus TRAF6 binding sites ( ΔT2 , 3 , 6 ) [54] . Briefly , parent RAW 264 . 7 cells were incubated with retroviral supernatants for 8 h in the presence of polybrene ( 8 µg/ml; Sigma Chemical ) . EGFP+ cells were sorted by FACS after 4 days . Mouse embryonal fibroblasts from wt and TRAF2−/− mice ( gifts from Hiroyasu Nakano ) were also transduced with the retroviral vector that encodes wt hmCD40 . To assess autophagy flux , hmCD40-RAW 264 . 7 cells expressing tfLC3 plus either WT-PKR or DN-PKR were cultured with or without CD154 for 4 hr and fixed with 4% paraformaldehyde . Slides were mounted with Flouromount G ( Southern Biotech ) and analyzed by fluorescent microscopy for distinct LC3 positive structures that measure at least 1 µm in diameter [53] . For assessment of accumulation of LC3 around the parasite , hmCD40-RAW 264 . 7 cells expressing LC3-EGFP plus either WT-PKR or DN-PKR were cultured with or without CD154 overnight prior to challenge . Monolayers were infected with RH T . gondii that express cytoplasmic RFP . Five hr post challenge , monolayers were fixed with paraformaldehyde and assessed for LC3-EGFP accumulation around T . gondii as described [53] . For assessment of LAMP-1 accumulation around the parasite , macrophages or microglia were cultured with isotype control or stimulatory anti-CD40 mAb overnight prior to challenge with transgenic RH T . gondii that express cytoplasmic YFP . Monolayers were incubated with LAMP-1 antibodies ( gift from Dr . Clifford Harding , Case Western Reserve University ) followed by incubation with Alexa flour 568 conjugated secondary antibody ( Jackson ImmunoResearch ) and accumulation of LAMP-1 around the parasite was assessed 8 hr post infection as described [53] . Cells and organs ( spleen , lung ) were lysed in buffer supplemented with protease and phosphatase inhibitors ( Cell Signaling ) . Equal amounts of protein were subjected to SDS-PAGE and transferred to a PVDF membrane . Membranes were probed with antibodies to total PKR or phospho PKR ( Thr 451 ) ( Santa Cruz Biotechnology ) , IRGM3 ( Abcam , Cambridge , MA ) or actin ( Santa Cruz Biotechnology ) followed by incubation with corresponding secondary Ab conjugated to horseradish peroxidase ( Santa Cruz Biotechnologies ) . Bands were visualized by using a chemilluminescent kit ( Pierce Bioscience ) . For immunoprecipitation , hmCD40-RAW 264 . 7 cells were transfected with a plasmid encoding FLAG-tagged WT-PKR or remained untransfected , and after 48 h , cells were incubated with or without CD154 for 30 min . In certain experiments , hmCD40-RAW 264 . 7 cells were transfected with either control or TRAF2 siRNA followed by transfection with FLAG-tagged WT-PKR after 24 h . Lysates were immunoprecipitated by incubation with anti-FLAG antibody ( Sigma ) , anti-TRAF2 C20 antibody ( Santa Cruz Biotechnology ) or anti-TRAF6 D-10 antibody ( Santa Cruz Biotechnology ) overnight at 4°C . Protein complexes were then captured by incubation with 50 µl of protein G beads ( Sigma ) for 2 hr at 4°C and then washed with wash buffer supplemented with protease and phosphatase inhibitors . The beads were resuspended in 35 µl of sample buffer and boiled . Lysate from the immunoprecipitation was immunoblotted for TRAF2 , TRAF6 or FLAG . Splenocytes ( 2×106/ml ) were incubated with or without TLA ( 10 µg/ml ) . Supernatants were collected at 24 h and used to determine concentrations of IL-12p 40 and TNF-α while supernatants collected at 72 h were used to measure IFN-γ ( eBiosciences , San Diego , CA ) . Concentrations of nitric oxide were assessed in 72 h supernatants using Griess reaction ( Promega Corporation , Madison , WI ) . Data are expressed as µM of nitrite . Statistical significance was assessed by 2-tailed student's t test and Analysis of Variance . Histopathologic changes were analyzed using Mann-Whitney U test . Differences were considered statistically significant when p was <0 . 05 . | PKR was identified more than 30 years ago as an inhibitor of viral replication . It is unknown if PKR promotes protection against disease caused by non-viral pathogens . We addressed this question using Toxoplasma gondii , a major parasitic pathogen . T . gondii can cause cerebral and/or eye disease primarily in immunosuppressed patients and newborns . After infection with T . gondii , PKR-deficient mice exhibited high parasite loads in the eye and brain and were more susceptible to ocular and cerebral toxoplasmosis . Macrophages and microglia are important effectors of protection against T . gondii . These cells required PKR signaling to kill the parasite in response to stimulation via CD40 , a molecule that promotes protection against ocular and cerebral toxoplasmosis . CD40 functioned only through its TRAF6 binding site to activate PKR , but this process was also dependent on TRAF2 where this molecule likely acted as an intermediary that promoted TRAF6-PKR association and PKR activation . PKR linked CD40-TRAF signaling to stimulation of the autophagy pathway and T . gondii killing . Our studies identified a previously unappreciated role of PKR as mediator of anti-microbial activity and promoter of resistance against disease caused by a non-viral pathogen , as well as provided new insight on the molecular link between CD40 and PKR . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"parastic",
"protozoans",
"protein",
"kinase",
"signaling",
"cascade",
"signal",
"transduction",
"toxoplasma",
"gondii",
"protozoology",
"immunology",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"immune",
"response",
"signaling",
"cascades"
] | 2013 | The Protein Kinase Double-Stranded RNA-Dependent (PKR) Enhances Protection against Disease Cause by a Non-Viral Pathogen |
Histidine kinase QseE and response regulator QseF compose a two-component system in Enterobacteriaceae . In Escherichia coli K-12 QseF activates transcription of glmY and of rpoE from Sigma 54-dependent promoters by binding to upstream activating sequences . Small RNA GlmY and RpoE ( Sigma 24 ) are important regulators of cell envelope homeostasis . In pathogenic Enterobacteriaceae QseE/QseF are required for virulence . In enterohemorrhagic E . coli QseE was reported to sense the host hormone epinephrine and to regulate virulence genes post-transcriptionally through employment of GlmY . The qseEGF operon contains a third gene , qseG , which encodes a lipoprotein attached to the inner leaflet of the outer membrane . Here , we show that QseG is essential and limiting for activity of QseE/QseF in E . coli K-12 . Metabolic 32P-labelling followed by pull-down demonstrates that phosphorylation of the receiver domain of QseF in vivo requires QseE as well as QseG . Accordingly , QseG acts upstream and through QseE/QseF by stimulating activity of kinase QseE . 32P-labelling also reveals an additional phosphorylation in the QseF C-terminus of unknown origin , presumably at threonine/serine residue ( s ) . Pulldown and two-hybrid assays demonstrate interaction of QseG with the periplasmic loop of QseE . A mutational screen identifies the Ser58Asn exchange in the periplasmic loop of QseE , which decreases interaction with QseG and concomitantly lowers QseE/QseF activity , indicating that QseG activates QseE by interaction . Finally , epinephrine is shown to have a moderate impact on QseE activity in E . coli K-12 . Epinephrine slightly stimulates QseF phosphorylation and thereby glmY transcription , but exclusively during stationary growth and this requires both , QseE and QseG . Our data reveal a three-component signaling system , in which the phosphorylation state of QseE/QseF is governed by interaction with lipoprotein QseG in response to a signal likely derived from the cell envelope .
Two-component systems ( TCSs ) allow bacteria to perceive information from the environment and to adapt gene expression and behavior in a meaningful way . Typically , a membrane-bound histidine kinase senses a stimulus via its N-terminal input domain leading to auto-phosphorylation at a histidine residue in the C-terminal transmitter domain [1] . Subsequently , the phosphoryl-group is transferred to an aspartate residue in the receiver domain of the cognate response regulator , thereby activating the associated output domain , which is often a transcription factor . While the downstream functions of many TCSs are well characterized , the stimuli sensed by the kinases and the underlying mechanisms often remain elusive . Histidine kinases may perceive their cognate stimuli directly or through employment of accessory proteins [2–5] . The model organism E . coli K-12 encodes 29 TCSs , each dedicated to a specific function [6 , 7] . Albeit intensively investigated , the roles of some TCSs still remain weakly defined including the TCS QseE/QseF ( a . k . a . GlrK/GlrR , a . k . a . YfhK/YfhA ) , which is conserved in Enterobacteriaceae [8] . Response regulator QseF comprises an N-terminal receiver domain , a σ54 interaction domain and a C-terminal DNA-binding helix-turn-helix ( H-T-H ) motif . In E . coli K-12 , QseF activates σ54-dependent promoters located upstream of genes glmY and rpoE , respectively [9 , 10] . GlmY is a small RNA ( sRNA ) controlling cell envelope biosynthesis ( see below ) and rpoE encodes σ24 , a master regulator of the cell envelope stress response [11 , 12] . The glmY gene is located immediately upstream of the qseEGF operon encoding the QseE/QseF TCS and this synteny is conserved [8 , 13] . Assisted by the integration host factor , QseF binds three upstream activating sequences ( UAS ) with the consensus TGTN12ACA thereby triggering transcription of glmY from its σ54-promoter [8 , 9] . An overlapping weak σ70-promoter contributes to low basal expression levels ( Fig 1C; [8] ) . UAS similar to those present upstream of glmY are also observed upstream of the recently identified σ54-dependent rpoE P2 promoter shown to be activated by QseF [10] . QseF requires phosphorylation by kinase QseE for activity . Phosphorylation of QseF increases its DNA-binding affinity and activity of the glmY σ54-promoter is abolished in mutants lacking QseE [8 , 9] . QseE/QseF is one of few TCSs residing in the “on” state , at least partially , which is in contrast to many other TCSs , which require a specific signal for activation that is usually absent from standard laboratory conditions . GlmY has a crucial role for the bacterial cell: Together with the homologous sRNA GlmZ and the RNA-binding adaptor protein RapZ , it controls the levels of enzyme GlmS , which synthesizes glucosamine-6-phosphate , an essential precursor for peptidoglycan and the outer membrane [11] . GlmZ is an Hfq-dependent sRNA and base-pairs with the glmS transcript , thereby stimulating translation and stabilizing the mRNA [14 , 15] . When not required , GlmZ is recruited by RapZ to degradation by RNase E [16–18] . The latter process is counteracted by sRNA GlmY , which accumulates when the intracellular glucosamine-6-phosphate concentration decreases [16 , 19] . GlmY is not an Hfq-binding sRNA [17] . It acts as decoy RNA and serves to sequester RapZ , thereby inhibiting decay of GlmZ , which then stimulates GlmS production to replenish glucosamine-6-phosphate [16] . This feedback mechanism also operates in Salmonella and perhaps in all Enterobacteriaceae ensuring homeostasis of cell envelope precursors [20] . Accumulation of GlmY in response to glucosamine-6-phosphate depletion occurs post-transcriptionally indicating that QseE/QseF are not sensing this metabolite [9] . Thus , the stimulus for the QseE/QseF TCS in E . coli K-12 remains unknown so far . In addition to its crucial function in governing expression of cell envelope regulators , the QseE/QseF TCS is required for virulence of pathogenic Enterobacteriaceae . As a common theme , deletion mutants of the qseEGF operon are attenuated in virulence , as demonstrated for Citrobacter rodentium , the fish pathogen Edwardsiella tarda , enterohemorrhagic Escherichia coli ( EHEC ) , Salmonella enterica and Yersinia pseudotuberculosis [21–25] . This phenomenon has been studied most thoroughly in EHEC , in which QseE/QseF together with the QseB/QseC TCS complexly regulate virulence genes encoded within and outside of the locus of enterocyte effacement ( LEE ) , a pathogenicity island organized in 5 operons ( for recent reviews , see: [26–28] ) . Regulation by QseE/QseF is indirect and occurs through GlmY/GlmZ , which promote translation of virulence gene espFU and selectively destabilize transcripts of the LEE4 and 5 operons [29] . Kinase QseC is a sensor of epinephrine ( Epi ) and norepinephrine in EHEC [30] and the QseE/QseF TCS has been described to participate in Epi sensing and signal transduction involving extensive cross-talk between both TCSs [29 , 31] . EHEC and also other bacteria including C . rodentium and Salmonella sense these host hormones to activate virulence gene expression and colonize the host [26] . Epi was shown to stimulate autophosphorylation of QseC as well as QseE in vitro [30 , 31] . In addition , QseF can also be cross-phosphorylated by the non-cognate kinase QseC in vitro [32] . By integration of these phosphorylation signals QseF is proposed to modulate expression of virulence genes in response to Epi [33] . Whether Epi also plays a role for activity of QseE/QseF in commensal bacteria is unknown . The operon encoding the QseE/QseF TCS contains a third gene , qseG . QseG carries an N-terminal signal sequence recognized by the general Sec secretory pathway or the Tat twin arginine translocation system . Consistently , in EHEC QseG was shown to reside in the outer membrane facing the periplasmic leaflet [24 , 31] . In pathogenic bacteria including EHEC , C . rodentium and Salmonella , qseG is required for virulence and host colonization , but the underlying mechanisms remain unclear [24] . In the current work , we investigated the role of QseG in E . coli K-12 . We show that QseG is essential for activity of the QseE/QseF TCS and thereby for glmY transcription . In vivo phosphorylation assays demonstrate that QseG is mandatory for kinase QseE activity and thereby for QseF phosphorylation . We show that QseG interacts with the periplasmic domain of kinase QseE and mutational analysis indicates that this interaction is required for QseE activity . Finally , we show that Epi slightly increases QseF phosphorylation and thereby glmY expression in a QseE- and QseG-dependent manner in the stationary growth phase . Taken together , our data show that QseG operates together with QseE/QseF constituting a three-component system . QseG is likely involved in sensing of the cognate stimulus .
First , we confirmed that QseG is present in the periplasmic space in E . coli K-12 . To this end , we isolated the E . coli cell envelope containing periplasmic and outer membrane proteins using an extraction method , which was shown to produce clean envelope extracts [34] . To allow for its detection , QseG was provided with a C-terminal Strep-tag epitope . E . coli cells carrying a plasmid encoding qseG-strep or the empty expression vector were grown to the exponential as well as to the stationary growth phase . Western blot analysis of total protein extracts confirmed proper synthesis of QseG-Strep ( S1 Fig , top panel , lanes 1–4 ) . Envelope extracts were prepared and analyzed by SDS-PAGE and Western blotting . Comparison of the protein bands revealed a distinctive pattern of the periplasmic extracts as compared to the total extracts , indicating successful fractionation ( S1 Fig , bottom panel ) . Indeed , the periplasmic maltose binding protein ( MBP; MW = 43 . 39 kDa ) could be readily detected in the envelope extracts , whereas the cytoplasmic ribosomal protein S1 ( MW = 61 . 16 KDa ) was absent , confirming successful isolation of cell envelope proteins ( S1 Fig , panels 2 and 3 ) . Western analysis detected QseG-Strep in the envelope extracts and this localization was unaffected by the growth stage ( S1 Fig , top panel , lanes 7 and 8 ) . In conclusion , QseG is present in the cell envelope of E . coli K-12 , in agreement with previous results in EHEC [31] . To address the role of QseG for activity of the QseE/QseF TCS , we first studied the impact of a qseG deletion on GlmY steady state levels . Total RNA was extracted from bacteria harvested at various times during growth and analyzed by Northern blotting . In the wild-type strain GlmY accumulated over time showing highest levels during transition to the stationary growth phase , recapitulating previous observations ( Fig 1A; [9 , 16] ) . In the ΔqseG strain , GlmY levels were drastically decreased albeit weak hybridization signals remained detectable . GlmY levels were perfectly restored upon introduction of a plasmid expressing qseG from a heterologous promoter , excluding negative interference of the qseG deletion with synthesis of the downstream encoded response regulator QseF ( Fig 1A ) . To determine whether QseG affects GlmY levels at the transcriptional or post-transcriptional level , we measured expression of an ectopic glmY’-lacZ reporter fusion integrated into the chromosomes of the respective strains ( Fig 1B ) . Expression of the glmY’-lacZ fusion increased over time in the wild-type strain , whereas only low activities were detectable in the ΔqseG mutant . Complementation of the ΔqseG mutant with a multi-copy plasmid expressing qseG from a heterologous promoter restored glmY’-lacZ expression to levels that exceeded the activities measured in the wild-type ( Fig 1B ) . The requirement of qseG for glmY transcription was not only detectable in strain CSH50 derivatives , which were used here , but also in MG1655 , indicating that this is a general phenomenon affecting E . coli K-12 strains ( S3 Fig ) . In conclusion , QseG is required for efficient transcription of glmY . Next , we dissected whether QseG controls the σ54- or the σ70- or both promoters upstream of glmY . To this end , we used mutated reporter gene fusions carrying nucleotide exchanges in the glmY transcriptional control region , which abolish activity of one promoter while leaving the respective second promoter unaffected ( Fig 1C top ) . We determined the activities of these reporter constructs in exponentially growing wild-type and ΔqseG strains ( Fig 1C bottom ) . Expression of the fusion solely driven by the σ54-promoter was abolished in the ΔqseG mutant and perfectly restored upon complementation with a plasmid carrying qseG . Complementation was observed regardless whether qseG was expressed from an IPTG-inducible or an arabinose-inducible expression vector ( Fig 1C and S4 Fig ) . In contrast , activity of the σ70-promoter was unaffected by qseG deletion or overexpression ( Note that basal expression levels are elevated in this case , because the σ70-promoter is usually repressed by binding of σ54 to the overlapping σ54-promoter; [9] ) . Hence , qseG is essential for activity of the σ54-promoter of glmY , but has no role for the σ70-promoter , explaining the low residual expression of glmY that remained detectable in the qseG deletion mutant ( Fig 1A–1C ) . These results strongly resemble previous data obtained in a mutant lacking kinase QseE [9] , i . e . the ΔqseG allele phenocopies a ΔqseE mutation . As the σ54-promoter of glmY is controlled by QseE/QseF , one likely explanation for these results is that QseG has a role for activity of this TCS . To determine whether QseG acts up- or downstream of QseE/QseF on glmY , we performed epistasis experiments . To this end , we tested the effects of plasmid-driven qseF , qseG and qseE overexpression in qse deletion mutants . In absence of qseF , transcription of glmY’-lacZ is solely driven by the σ70-promoter and therefore significantly decreased as compared to the wild-type strain ( Fig 2A , compare columns 1 and 2; [9] ) . Similar low expression levels were observed in the ΔqseG and ΔqseE mutants confirming that QseE and QseG are required for activity the σ54 glmY promoter ( Fig 2A , columns 6 and 10 ) . Complementation of the various deletion mutants with corresponding genes on plasmids restored high glmY’-lacZ expression levels , ruling out impaired expression of the remaining qse genes in the individual deletion mutants ( Fig 2A , columns 3 , 8 , 13 ) . Notably , plasmid-driven overexpression of qseG in the ΔqseF and ΔqseE mutants had no effect on the weak glmY expression level ( Fig 2A , columns 4 and 12 ) . Likewise , overexpression of qseE was without any effect when tested in the ΔqseF and ΔqseG mutants ( Fig 2A , columns 5 and 9 ) . These data show that QseG requires both QseF as well as QseE to stimulate glmY expression . Moreover , QseE is apparently unable to stimulate QseF activity when QseG is absent . Interestingly , when the qseG expression plasmid was used to complement the ΔqseG mutant strain , a very high glmY expression level was observed , suggesting that QseG is limiting for QseF activity in the wild-type strain ( Fig 2A , columns 1 and 8 ) . This conclusion is further supported by an experiment in which qseG was transcribed from the arabinose-inducible PAra promoter on a low copy plasmid and expression was gradually increased using incremental arabinose concentrations . A concomitant increase of glmY expression was observable indicating that glmY promoter activity directly correlates with the QseG level ( S5 Fig ) . To learn whether QseG has a role for QseF phosphorylation , we studied the effect of QseG on QseF variants carrying mutations in the D56 phosphorylation site , i . e . QseF-D56A and QseF-D56E variants mimicking non-phosphorylated and phosphorylated QseF , respectively [8] . Plasmids encoding wild-type QseF and the mutant QseF variants were used to complement strains deleted for chromosomal qseF , qseG or both genes , respectively . Introduction of the plasmid encoding wild-type QseF into the ΔqseF mutant restored glmY’-lacZ levels above the levels observed in the wild-type strain ( Fig 2B , compare columns 1–3 ) . Complementation of the ΔqseF mutant with the plasmid encoding the phospho-ablative QseF-D56A variant resulted in activities , which were 2-fold lower but clearly above background levels ( Fig 2B , compare columns 2–4 ) . It is well-known that upon overproduction even non-phosphorylated response regulators are able to activate their target genes to some extent [4 , 35 , 36] and this also applies to QseF [8] . In contrast , introduction of the plasmid coding for the phospho-mimetic QseF-D56E variant generated a much higher glmY expression level ( Fig 2B , column 5 ) , confirming that phosphorylated QseF is active and strongly stimulates glmY expression [8] . We observed very similar glmY expression patterns when the various qseF expression plasmids were tested in ΔqseG , ΔqseGF and ΔqseE mutant strains , but there was one striking exception: In the latter mutants , comparable activities were produced by wild-type QseF and the non-phosphorylatable QseF-D56A variant , respectively . In contrast , when tested in the ΔqseF mutant two-fold higher activities were generated by plasmid-borne wild-type QseF as compared to QseF-D56A ( Fig 2B , compare columns 3–4 with 7–8 , 11–12 and 15–16 ) . These observations suggest that QseG , just as QseE , can stimulate the activity of wild-type QseF , but not of QseF variants bearing exchanges in the D56 phosphorylation site . To investigate the role of QseG for QseF activity in more detail , we compared the activities of the various plasmid-encoded QseF variants in isogenic ΔqseF and ΔqseGF strains during growth . In this case , we used a glmY’-lacZ fusion solely driven from the σ54-promoter and monitored β-galactosidase activities at regular time intervals ( Fig 2C ) . Expression of the phospho-mimetic qseF-D56E variant resulted in very high glmY expression levels , whereas much lower activities were measured when the phospho-ablative qseF-D56A mutant was expressed ( Fig 2C , compare blue and green columns ) . Of note , presence or absence of qseG had no role for the activities generated by these qseF alleles . In contrast , glmY expression levels triggered by wild-type QseF always decreased in the absence of qseG to the levels observed for the phospho-ablative QseF-D56A variant ( Fig 2C , compare dark red with light red and blue columns ) . Taken together , these data suggest that QseG stimulates activity of response regulator QseF in a dosage-dependent manner , most likely by triggering its phosphorylation . Several two-component systems are known , which are subject to autoregulation at the transcriptional level [5] . A putative autoregulation could potentially interfere with our genetic analysis addressing the role of QseG for QseE/QseF activity . To investigate a possible feedback regulation , we measured expression of ectopic transcriptional lacZ fusions to the qseEGF promoter . The qseEGF operon is transcribed from a σ70-promoter , which is located immediately downstream of the glmY gene , and starts transcription 25 bp upstream of the qseE start codon ( Fig 3A top; [9] ) . A fusion of lacZ to a DNA fragment comprising this promoter ( position -70 to +107 relative to the qseE start ) , generated only low β-galactosidase activities that were not affected by qseF and qseG mutations ( Fig 3A , fusion I ) . To account for potential glmY-qseE read-through transcripts , a fusion of lacZ to a fragment comprising positions -480 to +107 relative to qseE was additionally tested . The latter fusion carried the complete glmY locus including its transcriptional control region upstream of the qseE promoter and qseE’-lacZ ( Fig 3A , fusion II ) . However , the activities generated by this construct were virtually indistinguishable from the activities observed for the shorter fusion ( Fig 3A , compare fusions I and II ) . To account for a possible intrinsic instability of the qseE’ ( +107 ) -lacZ fusion mRNAs , we additionally tested isogenic constructs , in which the lacZ gene was fused further downstream at position +266 to qseE ( Fig 3A , fusions III and IV ) . Indeed , these fusions generated somewhat higher activities as compared to fusions I and II , but once again activities were not affected by deletion of qseF or qseG or presence of the glmY locus upstream of qseE’ ( +266 ) -lacZ . Similar expression patterns were observed in the exponential and stationary growth phases ( compare Fig 3A and S6 Fig ) . These results argue against an autoregulation of qseEGF expression . Moreover , the data are in agreement with previous Northern results indicating that qseE is only weakly expressed , and with previous semi-quantitative RT-PCR data suggesting that read-through from the upstream located glmY promoter into qseE does virtually not occur [9] . Low expression of qseE is also reflected by weak signals obtained for FLAG-tagged QseE in Western blot analyses of total protein extracts as shown later in this study . In agreement , a global proteomics study measured 11 molecules QseG and 36 molecules QseF per E . coli-K12 cell , whereas QseE could not be detected at all [37] . To corroborate these data and to account for a hypothetical internal promoter that could be present in the qseE-qseG intergenic region , we additionally monitored the levels of endogenously encoded QseG protein . To allow for detection , the 3×FLAG epitope sequence was fused to the 3’ end of the chromosomal qseG gene . Reporter gene measurements confirmed that the QseG-3×FLAG protein retained full functionality in respect to activation of glmY transcription ( S7 Fig ) . To test for autoregulation , we refrained from analysis of deletions within the qseEGF operon as this procedure would generate shorter qse transcripts with likely altered stabilities , thereby leading to ambiguous results . In lieu thereof , we tested the effects of plasmid-driven over-expression of qseF-D56E , qseG and qseE on synthesis of the chromosomally encoded QseG-3×FLAG protein , respectively . The same plasmids trigger a strong expression of the glmY’-lacZ reporter fusion ( Fig 2 ) , reflecting the properties of a fully activated QseE/QseF TCS . Since downstream effectors are sometimes involved in feedback regulation of two-component systems [5] , a plasmid overexpressing glmY was included in this analysis . However , none of the tested plasmids had any significant effect on the QseG-3×FLAG level , neither during exponential growth nor in the stationary growth phase ( Fig 3B ) . In conclusion , QseG controls activity of QseE/QseF rather than expression of corresponding genes . The genetic data ( Fig 2 ) point to a mechanism in which QseG stimulates phosphorylation of response regulator QseF through modulation of activity of kinase QseE . To address this possibility , we studied phosphorylation of QseF in vivo by metabolic labeling of cells using [32P] phosphorus . To this end , qseF was expressed under control of the IPTG-inducible Ptac promoter from a plasmid in wild-type as well as ΔqseE cells . The QseF variant carrying the D56A exchange of the phosphorylation site in the receiver domain served as negative control . The bacteria were grown in absence and presence of IPTG and subsequently labeled with [32P] phosphoric acid . Total protein extracts were separated by SDS-PAGE and analyzed by autoradiography ( Fig 4A ) . Among various bands representing abundant phosphoproteins , a single phosphorylation signal became visible exclusively in the presence of IPTG and its position on the gel roughly matched the molecular weight of QseF ( MW = 49 . 15 kDa ) . To our surprise , this phosphorylation signal was also detectable in the ΔqseE strain and when the QseF-D56A variant was employed . To confirm that the IPTG-inducible phosphorylation signal indeed corresponds to QseF , we used QseF variants carrying Strep-tags at their C-termini allowing for their pull-down following [32P] labeling . In this case , the QseF variants were produced from plasmids in ΔqseF ( qseG+ ) as well as in ΔqseFG cells ( Fig 4B , top panel ) , metabolically labeled and subsequently isolated by pull-down using StrepTactin coated magnetic beads . The obtained fractions were separated by SDS-PAGE and analyzed by Western blotting using anti-Strep antiserum as well as by autoradiography ( Fig 4B , middle and bottom panels ) . The Western blot confirmed successful isolation of the QseF proteins from the cultures induced by IPTG ( Fig 4B middle panel , lanes 5–8 ) , whereas QseF could not be recovered from non-induced cells ( Fig 4B middle panel , lanes 1–4 ) . Autoradiography once again detected phosphorylation of QseF under all conditions , regardless of the D56A substitution and also not affected by QseG ( Fig 4B bottom panel , lanes 5–8 ) . To explain the surprising results of the in vivo phosphorylation assays , we reasoned that QseF is phosphorylated at a second site , masking its phosphorylation at Asp56 . We speculated that the additional phosphorylation ( s ) may take place in the QseF C-terminus comprising the σ54 interaction domain and the DNA-binding domain ( subsequently designated as QseF-CTD ) . Phosphorylation of response regulators outside their receiver domains has been observed in several cases [38] . Therefore , we split the protein and expressed the QseF N-terminus comprising the receiver domain ( subsequently designated as QseF-NTD ) and the QseF-CTD separately , both provided with C-terminal Strep-tags for subsequent pull-down and detection ( Fig 5A ) . In addition , a QseF-NTD variant was generated carrying the D56A exchange of the phosphorylation site . Bacteria producing the various QseF-Strep variants from plasmids ( Fig 5B , left panel ) were labeled with [32P] followed by pull-down of the QseF variants , which was confirmed by Western blotting ( Fig 5B , middle panel ) . Indeed , autoradiography detected phosphorylation signals for both , the QseF-NTD and the QseF-CTD ( Fig 5B , right panel ) . Importantly , no phosphorylation of the QseF-NTD carrying the D56A substitution was observable ( Fig 5B , right panel ) . These data show that D56 is the single site phosphorylated in the QseF receiver domain , whereas the additional phosphorylation signal localizes in the QseF-CTD . Western blot analysis of purified proteins using an antiserum specific for phospho-tyrosine residues generated no signals . However , a phospho-threonine specific antiserum detected full-length QseF and the QseF-CTD ( Fig 5C , top and middle panel ) . In contrast , the QseF-NTD and the response regulator PhoB , which was included as a control , were not detectable ( Fig 5C , middle panel , lanes 1 and 4 ) . Treatment of the PVDF membrane with alkaline phosphatase prior to application of the antiserum erased the signals for full-length QseF and the QseF-CTD ( Fig 5C , bottom panel ) . Thus , QseF is phosphorylated at D56 in the receiver domain and presumably at unknown threonine or serine residue ( s ) in the CTD . Next , we used the QseF-NTD construct to clarify the question whether QseE and QseG are required for phosphorylation of QseF at the D56 residue in the receiver domain . To this end , the plasmid encoding the C-terminally Strep-tagged QseF-NTD was introduced in isogenic wild-type , ΔqseG and ΔqseE strains . Once again , the bacteria were grown in LB supplemented with IPTG to induce synthesis of recombinant proteins ( Fig 6A , left panel ) and subsequently labelled with [32P] phosphoric acid . Finally , the QseF-NTD was isolated by pull-down and eluates were analyzed by Western blotting using anti-Strep antiserum and by autoradiography . Western blotting proved successful isolation of the QseF-NTD from all three strains ( Fig 6A , middle panel ) . The autoradiograph revealed a strong phosphorylation signal for the QseF-NTD isolated from the wild-type strain . In contrast , 7- and 8-fold reduced phosphorylation signal intensities were obtained , when the QseF-NTD was isolated from the ΔqseG and ΔqseE mutants ( Fig 6A , right panel ) . These data show that both QseE and QseG are required for efficient phosphorylation of the QseF receiver domain . As QseG is unable to increase phosphorylation of QseF in the qseE mutant ( Fig 6A ) , it can be concluded that QseG acts through QseE to stimulate phosphorylation of QseF and thereby glmY expression . Some histidine kinases are bi-functional and exhibit in addition to phosphotransferase also phosphatase activity towards the cognate response regulator [1] . Kinase activity of QseE towards QseF was demonstrated previously [39] , but whether QseE has also phosphatase activity is unknown . In principle , QseG could increase phosphorylation of QseF either by stimulating phosphotransferase activity or by inhibiting phosphatase activity of QseE . To gain initial insight into how QseG governs phosphorylation of the QseF receiver domain , we carried out [32P] pulse-chase experiments to follow the fate of the QseF-D56 phosphorylation signal in a time course . Therefore , we once again labelled the wild-type , ΔqseG and ΔqseE strains producing the Strep-tagged QseF-NTD ( pulse ) , but subsequently stopped further incorporation of the [32P] label by addition of “cold” phosphorus ( chase ) . Samples were harvested at 0 , 5 and 15 min following chase and the QseF-NTD was subsequently isolated by StrepTactin pull-down and analyzed as before ( Fig 6B ) . The phosphorylation signal for the QseF-NTD rapidly diminished within 15 min in the wild-type strain , whereas such a decrease was not observable in the ΔqseE mutant ( Fig 6B , compare lanes 1–3 with 7–9 ) . This result indicates that QseE possesses phosphatase activity and is responsible for dephosphorylation of the QseF receiver domain in the wild-type strain . In case QseG would act by inhibition of QseE phosphatase activity , an accelerated dephosphorylation of the QseF-NTD is expected in the ΔqseG mutant as compared to the wild-type . However , this was not the case: The QseF-NTD phosphorylation signal also decreased over time in the ΔqseG mutant , but not faster than in the wild-type strain ( Fig 6B , lanes 4–6 ) . In conclusion , QseG appears not to act by inhibition of phosphatase activity , suggesting that it increases phosphorylation of the QseF receiver domain through stimulation of QseE phosphotransferase activity . QseG faces the periplasmic leaflet of the outer membrane [24 , 31] . On the other hand , kinase QseE contains a helical periplasmic domain of 140 amino acids between its two transmembrane domains ( TMs; [40] ) . This topological arrangement makes a physical interaction of QseG with the periplasmic domain of QseE feasible . Physical interaction of outer membrane lipoproteins with the periplasmic domains of cytoplasmic membrane proteins has been demonstrated in several cases [41–43] . To investigate whether QseG and QseE interact , we used a ligand fishing approach based on StrepTactin affinity chromatography , which allows for pull-down of membrane proteins by cytoplasmic or periplasmic interaction partners as demonstrated previously [4 , 44] . For detection of the prey protein QseE , the sequence encoding the 3×FLAG epitope was fused in frame to the 3’ end of qseE encoded at its natural chromosomal locus . An isogenic strain carrying the 3×FLAG epitope sequence fused to the 3’ end of phoQ served as negative control . Similar to QseE , PhoQ is a histidine kinase that possesses two N-terminal TMs encompassing a large domain extruding into the periplasm . QseG carrying a C-terminal Strep-tag was used as bait and produced from a plasmid in the latter two strains . A complementation assay confirmed functionality of the QseG-Strep protein ( S8 Fig ) . The same strains , but producing solely the Strep-peptide rather than QseG-Strep served as negative controls . Analysis of total cell extracts by Western blotting revealed a clear signal for the PhoQ-3×FLAG protein ( MW = 58 . 12 kDa ) in addition to several non-specific bands , whereas only a faint band for QseE-3×FLAG ( MW = 56 . 15 kDa ) was detectable ( Fig 7A , “input” ) , reflecting the notoriously weak expression level of qseE ( see above and [9] ) . The various strains were subjected to the StrepTactin affinity purification protocol and inspection of the eluates proved successful purification of QseG-Strep ( Fig 7A , “output” bottom panel ) . Western blotting analysis of the eluates revealed a strong enrichment of QseE-3×FLAG when QseG-Strep was used as bait , whereas no signals were obtained when the Strep-peptide was produced or when PhoQ-3×FLAG was assessed as potential prey , providing proof of specificity of the QseE-QseG interaction detected by this approach ( Fig 7A ) . To further characterize interaction of QseG and QseE , we used the bacterial adenylate-cyclase based two-hybrid system ( BACTH ) , which relies on interaction-mediated reconstitution of adenylate cyclase activity in E . coli [45] . In BACTH , the complementary T18- and T25-fragments of Bordetella pertussis adenylate cyclase are assembled to a functional enzyme through interaction of candidate proteins that are fused to these fragments . The classical BACTH is restricted to interactions within the cytoplasm or the cytoplasmic membrane , but more recently modified BACTH vectors have been developed , which allow to assess extra-cytoplasmic protein interactions [46] . In this case , a membrane domain of the E . coli OppB protein is inserted in the fusion protein between the CyaA-fragment and the candidate protein , resulting in extrusion of the latter into the periplasm , while the N-terminal CyaA fragment stays in the cytoplasm . Therefore , we fused the sequence encoding QseG ( but lacking the first 25 codons encoding the N-terminal export signal ) to the 3’ end of the T18-TMoppB fusion gene ( Fig 7B ) . Of note , deletion of the export signal in the context of the wild-type QseG protein rendered the protein inactive , supporting the idea that QseG must leave the cytoplasm to stimulate glmY expression ( S9 Fig , columns 1–4 ) . The T18-TMoppB-QseG BACTH fusion construct was then tested for interaction with QseE , which was fused to the C-terminus of the T25 fragment ( Fig 7B ) . Indeed , β-galactosidase assays reflecting cAMP synthesis detected activity fairly above the level of the negative control , in which the unfused CyaA fragments were addressed ( Fig 7C , columns 2 and 8 ) . Activity even exceeded the positive control , which detects homodimerization of the leucine zipper of the yeast transcription factor Gcn4 in the periplasm ( Fig 7C , compare columns 2 and 7 ) . No interaction was detectable when QseG [Δ aa 1–25] was directly fused to the T18 fragment omitting the TMoppB domain in the fusion protein ( Fig 7C , column 1 ) , confirming that QseG must leave the cytoplasm to interact with QseE . To provide further proof of specificity of the detected QseG-QseE interaction , we also tested interaction of the T18-TMoppB-QseG fusion with membrane-bound histidine kinases CpxA and PhoQ , which exhibit similar membrane-topologies as QseE , i . e . they possess large periplasmic domains encompassed by two N-terminally located TMs . However , only back-ground activities could be measured in these cases ( Fig 7C , columns 5 and 6 ) . BACTH assays addressing homodimerization of the kinases proved functionality of the fusion proteins ( Fig 7D , columns 1 , 4 , 5 ) . Next , we wanted to confirm that QseG interacts with the N-terminus of QseE comprising the periplasmic loop . Therefore , we tested interaction of the QseG fusion protein with the N-terminus of QseE ( residues 1–250; subsequently designated QseENTD ) lacking the C-terminal transmitter domain ( Fig 8A ) . However , in this case only background activities were detectable ( Fig 7C , columns 3 and 8 ) . Dimerization of histidine kinases is usually mediated through the transmitter domains [1] and accordingly homodimerization of the QseENTD was greatly impaired ( Fig 7D , columns 1 and 2 ) . To test , whether the loss of interaction with QseG resulted from the inability of QseENTD to form dimers , we fused the leucine zipper homodimerization domain of Gcn4 to the C-terminus of QseENTD . Indeed , this procedure rescued dimerization ( Fig 7D , column 3 ) and also partially restored interaction with QseG ( Fig 7C , column 4 ) . Taken together , the data indicate that QseG binds the N-terminus of QseE in the periplasm and that dimerization of QseE is a prerequisite for this interaction . Our data suggested that activation of QseE by QseG may require their physical interaction in the periplasm . To obtain insight , we searched for mutations in the QseE N-terminus decreasing interaction with QseG . To this end , we randomly mutagenized the sequence encoding the QseE N-terminus ( aa 1–258 ) within the full-length T25-qseE construct by error prone PCR and screened the resulting library of QseE mutants in context of BACTH for variants showing decreased interaction with QseG ( Fig 8A ) . In addition to mutants carrying stop- or frameshift mutations , which were not further analyzed , two mutants carrying exclusively amino acid exchanges were isolated . One mutant ( subsequently designated “QseG-M5 ) carried five exchanges ( i . e . F19L , L21H , I22R , L23P , L24P ) in TM1 , while the other mutant received a single amino acid exchange ( S58N ) in the periplasmic loop ( Fig 8A and S10 Fig ) . Quantitative assays revealed that interaction of QseE with QseG is abrogated by the M5 mutation and significantly decreased when the S58N exchange was present ( Fig 7B , columns 1–3 ) . A pull-down assay using QseG-Strep as bait confirmed the decreased interaction potential of the QseE-S58N variant ( S11 Fig ) . In this case , presence of the S58N mutation reduced the amount of co-purifying QseE-3×FLAG protein ~3-fold ( S11 Fig , compare lanes 7 and 9 ) . Thus , interaction of QseE with QseG is impaired by the S58N mutation but not completely abolished . To test for kinase activity , the various QseE variants were placed on plasmids under Ptac-promoter control and used to complement a ΔqseE mutant strain carrying the glmY’-lacZ reporter fusion on the chromosome ( Fig 8C ) . Indeed , the M5 mutation abolished QseE activity as judged from comparison with the empty vector control and an inactive QseE-H259A mutant carrying a substitution in the QseE autophosphorylation site ( Fig 8C , compare columns 1–6 ) . However , as indicated by BACTH , the QseE-M5 mutant was also strongly impaired in homodimerization ( Fig 8B , columns 4 and 5 ) . Therefore , the mutations in TM1 might interfere with proper membrane insertion of QseE rather than to specifically abrogate interaction with QseG . In contrast , the QseE-S58N mutant was not significantly impaired in homodimerization ( Fig 8B , columns 4 and 6 ) . Strikingly , the QseE-S58N mutant showed a 3-fold decreased potential to activate transcription of glmY as compared to wild-type QseE ( Fig 8C , compare columns 4 and 7 ) . The residual activation potential of QseE-S58N is still dependent on QseG ( S12 Fig ) . Thus , the S58N exchange diminishes interaction with QseG and concomitantly lowers activity of QseE . This result supports a model in which QseG activates QseE kinase activity through interaction . A previous study reported that QseE of EHEC responds to epinephrine in vitro by increased autophosphorylation [31] . To learn whether epinephrine has also a role for QseE activity in E . coli K-12 , we studied the impact of epinephrine on glmY transcription during growth . Epinephrine did not change glmY transcription during the exponential and early stationary growth phase ( Fig 9A ) . However , after 10 h growth in presence of epinephrine a somewhat higher glmY transcription level became evident in the epinephrine treated culture ( Fig 9A ) . To confirm this result , we determined glmY transcription levels in overnight cultures incubated for ~16h in epinephrine containing LB medium . Once again , higher glmY’-lacZ levels were observable in the wild-type strain in presence of epinephrine and Northern blot analysis confirmed that GlmY accumulated to higher amounts in this case ( Fig 9B ) . In contrast , the ΔqseG and ΔqseE mutant strains showed only low glmY transcription levels and failed to respond to epinephrine ( Fig 9B ) . These results suggested that epinephrine might stimulate QseE autophosphorylation in a QseG-dependent manner . To address this possibility , we studied phosphorylation of the QseF receiver domain by metabolic [32P] labelling followed by pull-down in vivo . The Strep-tagged QseF-NTD was overproduced from a plasmid in wild-type , ΔqseG and ΔqseE strains ( Fig 9C , left ) and subsequently cells grown to stationary phase were labelled in the absence and presence of epinephrine and the QseF-NTD was isolated by pull-down using StrepTactin coated magnetic beads ( Fig 9C , middle panel ) . Indeed , epinephrine moderately stimulated phosphorylation of the QseF-NTD in the wild-type strain ( Fig 9C , right panel and diagram ) . In the ΔqseG and ΔqseE mutants , however , phosphorylation of the QseF-NTD was strongly decreased as observed before ( Fig 6 ) and epinephrine had no effect on the remaining phosphorylation signal ( Fig 9C , right panel and diagram ) . Thus , epinephrine is capable to stimulate QseE phosphorylation even in E . coli K-12 , but this effect requires QseG and solely occurs in the stationary growth phase .
In this work , we show that the outer membrane lipoprotein QseG is an indispensable component of the QseE/QseF TCS , reflecting the conserved co-localization of the qseEGF genes in one operon . Genetic and in vivo phosphorylation studies indicate that QseG triggers phosphorylation of QseE/QseF in vivo , constituting a “three-component system” ( Figs 1 , 2 and 6 ) . QseG binds the large periplasmic domain of kinase QseE ( Fig 7 ) and this interaction likely triggers QseE autophosphorylation activity or stimulates QseE/QseF phosphoryl-group transfer ( Fig 10 ) . Such a model is supported by identification of the S58N exchange located in a conserved region in the periplasmic domain of QseE ( S10 Fig ) , which impairs interaction with QseG and concomitantly decreases activity of QseE/QseF ( Fig 8; S11 Fig ) . The data are consistent with a model , in which the outer membrane protein QseG activates kinase QseE by interaction thereby increasing the level of phosphorylated QseF , which in turn activates the σ54-dependent promoters upstream of glmY and rpoE , both encoding central regulators of cell envelope homoeostasis ( Fig 10 ) . In agreement with our results , QseG was shown also to be required for activation of the σ54-dependent rpoE promoter by response regulator QseF [10] . In this case , QseG was identified in a screen as a multi-copy activator of the rpoE σ54-promoter . QseG carries a 25 aa long export sequence at the N-terminus , including a so-called “lipobox” , and does not contain a Lol avoidance motif ( S13 Fig ) . Therefore , it is exported to the periplasm ( S1 Fig ) and predicted to attach to the inner leaflet of the outer membrane via lipidation of residue Cys26 , which should become the new N-terminal amino acid following cleavage of the signal peptide [47] . Consistently , EHEC QseG , which is identical with QseG from E . coli K-12 ( S13 Fig ) , was shown to localize to the outer membrane , but being inaccessible to proteinase K digestion from the exterior [24 , 31] . This extracytoplasmic localization is in perfect agreement with our observation that QseG must leave the cytoplasm in order to interact with QseE and to activate glmY transcription ( Fig 7C and S9 Fig ) . However , it should be stressed that the exact localization of QseG within the periplasmic compartment is apparently not crucial for its activity . Mutation of the presumably lipidated Cys26 residue has only a moderate effect on QseG activity and QseG even retains significant activity when carrying a V27D Lol avoidance motif ( S9 Fig ) leading to its retention in the cytoplasmic membrane [47] . Obviously , QseG can reach and bind QseE regardless of its specific localization within the periplasmic space . In any case , the 237 amino acids long QseG protein is sufficiently large to form a trans-envelope complex with QseE [42] . The architecture of the QseE/QseG/QseF three-component system is remarkably reminiscent of the Cpx and Rcs envelope stress response systems , which are also built around two-component systems that employ outer membrane lipoproteins for signal perception and activation of the phosphorylation cascade [48] . Under normal conditions , the lipoprotein RcsF is threaded into β-barrel Omp proteins and thereby sequestered at the outer membrane [41 , 49] . Stress prevents incorporation of RcsF into these complexes leading to accumulation of RcsF remaining exposed in the periplasm . This enables the outer-membrane attached RcsF to interact with the inner membrane protein IgaA , thereby releasing the Rcs phospho-relay system from IgaA-mediated inhibition [41 , 42] . In the Cpx system , the outer membrane lipoprotein NlpE activates the CpxA/CpxR TCS , presumably through direct interaction with kinase CpxA or its periplasmic inhibitor CpxP [50] . As a common principle , phosphorylation of the Rcs and Cpx systems is triggered by availability of the cognate lipoproteins for interaction in the periplasm . Accordingly , both systems can be activated by artificially increasing the levels of these lipoproteins [51 , 52] . A similar scenario is observed here , as activity of the QseE/QseF TCS directly correlates with qseG expression levels ( Fig 2 and S5 Fig ) . Apparently , QseE/QseF phosphorylation activity is limited by availability of QseG in the periplasm . It is tempting to speculate that cells control the levels of “free” QseG available for interaction with QseE , to adjust QseE/QseF activity accordingly . Interestingly , in EHEC QseG was recently found to interact with the LEE-encoded protein SepL , which serves as gate-protein for the type III secretion system used to translocate effector proteins into host cells [24] . Albeit the role of this interaction remained unclear , it could serve to sequester QseG making it unavailable for interaction with QseE . Such a mechanism could fine-tune synthesis of type III secretion system components , as their expression is controlled by QseE/QseF through GlmY/GlmZ [29] . However , SepL is absent in E . coli K-12 indicating that interaction of QseG with QseE must be differently controlled , which will be the subject of future studies . Moreover , we show that QseF also responds to Epi in E . coli K-12 , but moderately and exclusively in the stationary growth phase . Under these conditions Epi increases QseF phosphorylation and concomitantly glmY transcription 1 . 5-fold and this effect requires both , kinase QseE and QseG ( Fig 9 ) . In respect to its limited impact , it appears that Epi is not a major stimulus for the QseE/QseF TCS in E . coli K-12 , which apparently is , at least partially , already in the “on-state” under standard laboratory conditions ( Figs 1 , 2 and 6 ) . As QseE requires QseG to respond to Epi ( Fig 9 ) , it appears debatable whether QseE is able to sense Epi on its own [31] . Interaction of Epi with the QseE periplasmic domain could also not be observed by NMR [40] . Therefore , interaction of QseG with epinephrine appears to be a possible mechanism . However , we cannot exclude that the weak stimulatory effect of Epi on QseE/QseF phosphorylation is indirect , and may perhaps involve a putative interaction partner of QseG or even epinephrine degradation products . EHEC QseE was also reported to sense phosphate and sulfate ions as it responds with increased autophosphorylation to these signals in vitro [31] . However , when tested in a minimal medium , phosphate and sulfate had no impact on glmY transcription in E . coli K-12 ( S14 Fig ) . Apparently , through employment of QseG , QseE senses different cues in E . coli K-12 , most likely signal ( s ) derived from the cell envelope . In agreement , activity of QseF was observed to increase in a waaC mutant [10] . Gene waaC encodes LPS heptosyltransferase I and its absence causes defects in LPS biosynthesis . In addition , there is also no Epi-dependent cross-phosphorylation of QseF by histidine kinase QseC in E . coli K-12 , as Epi is unable to increase phosphorylation of QseF in the absence of QseE ( Fig 9C ) . This is in contrast to EHEC , in which QseC was reported to contribute to QseF phosphorylation as it is able to cross-phosphorylate QseF in vitro [32] . In EHEC , the QseB/QseC TCS was also described to cross-talk with the QseE/QseF TCS at the level of glmY transcription: In addition to QseF , also response regulator QseB was shown to bind to the EHEC glmY promoter region , thereby stimulating glmY expression two-fold [29] . However , in E . coli K-12 deletion of qseB or qseC has no effect on glmY transcription ( S15 and S16 Figs ) , which might be explained by differences in the sequences of the predicted QseB binding sites [29] . Therefore , the activities of the QseB/QseC and QseE/QseF TCSs appear to be well separated in E . coli K-12 . It even appears unlikely that QseF can receive phosphoryl-groups from any other histidine kinase than QseE in vivo , as glmY transcription from the σ54-promoter is abolished in a mutant lacking kinase QseE ( Figs 2 and 8C; [9] ) . The weak phosphorylation of the QseF receiver domain remaining detectable in the absence of QseE or QseG ( Figs 6 and 9C ) might result from non-physiological cross-talk as we had to overproduce the QseF-NTD in these experiments , potentially providing a sink for non-cognate phosphorylations . Alternatively , these cross-phosphorylations may not be robust as they could be removed through QseE phosphatase activity in wild-type cells ( Fig 6B ) . Phosphatase activities of histidine kinases were shown to be crucial to prevent aberrant phosphorylations of response regulators by non-cognate kinases [53] . In this work , we studied phosphorylation of QseF in vivo using metabolic [32P] labelling , which is a method usually not considered in TCS research [54] , albeit it allows to snapshot protein phosphorylation states in the living cell [55] . Using this approach , we also detected an additional phosphorylation signal for the QseF-CTD , suggesting that QseF is at least doubly phosphorylated ( Figs 4 and 5 ) . The C-terminal phosphorylation was also detectable by an antiserum recognizing phosphorylated Thr- and Ser-residues ( Fig 5C ) . Meanwhile several response regulators are known to become additionally phosphorylated by Ser/Thr kinases interfering with their function [38 , 56] . In E . coli , two serine/threonine kinases , SrkA ( a . k . a YihE ) and YeaG , have been characterized [57 , 58] , but they are not required for phosphorylation of the QseF-CTD ( S17 Fig ) . The source and role of this additional phosphorylation signal must be addressed in future research .
E . coli strains were routinely grown in Lysogeny broth ( LB medium ) at 37°C or in case of bacterial two hybrid assays at 28°C . When required , antibiotics were added to the following concentrations: ampicillin ( 100 μg/ml ) , kanamycin ( 30 μg/ml ) , spectinomycin ( 75 μg/ml ) and chloramphenicol ( 15 μg/ml ) . E . coli strains and plasmids used in this study are described in S1 and S2 Tables and oligonucleotides are listed in S3 Table under “Supporting information” . Details on plasmid constructions are described in S1 Text under “Supporting information” . Deletions in the chromosomal qseEGF operon were constructed by λ red recombination using plasmid pKD3 as template as described and the oligonucleotides specified in S1 and S3 Tables [59] . FLAG-tagging of chromosomal genes was performed as described previously [60] using oligonucleotides BG1305/BG1306 for qseG , BG902/BG903 for phoQ , BG968/BG969 for qseE and plasmid pSUB11 as template . Ectopic integration of glmY’-lacZ reporter gene fusions into the λattB site on the E . coli chromosome was achieved as described before [8 , 61] . Established alleles were moved between strains by general transduction using E . coli phage T4GT7 [62] . Strains were cured from resistance gene cassettes using FLP recombinase encoded on plasmid pCP20 as described [59] . Cell envelope fractions containing soluble periplasmic and outer membrane proteins were isolated as described [34] . Briefly , E . coli strain Z197 harboring either plasmid pYG191 coding for QseG-Strep or the isogenic plasmid pBGG237 encoding only the Strep-peptide was grown in 100 ml M9 minimal medium supplemented with 1% maltose , 0 . 1% casamino acids , thiamine ( 1 μg/ml ) and L-proline ( 40 μg/ml ) . One half of each culture was harvested in the exponential growth phase ( OD600 ~0 . 5 ) , whereas the remaining half was harvested in the stationary growth phase . Cells were pelleted by centrifugation , gently re-suspended in 200 μl TSE buffer ( 200 mM Tris-HCl pH 8 . 0 , 500 mM sucrose , 1 mM EDTA ) and incubated on ice for 60 min . The TSE-soluble proteins were separated from the insoluble fractions by centrifugation ( 16000 g , 4°C , 45 min ) and 6 . 25 μg of the supernatants containing the periplasmic extracts were analyzed by SDS PAGE and Western blotting , respectively . RNA extraction and Northern blotting was performed as described before [19] . Bacteria were grown in LB for the indicated times and cells were harvested by centrifugation ( 2 min , 4°C , 11000 rpm ) and frozen in liquid nitrogen . RNA was extracted using the RNeasy mini kit ( Qiagen ) according to the manufacturer’s instructions . Digoxigenin-labeled RNA probes against GlmY and 5S RNA were obtained by in vitro transcription using the DIG-Labelling kit ( Roche Diagnostics ) and specific PCR fragments as templates . Primer pairs used for PCR were BG260/BG261 for glmY and BG287/BG288 for rrfD ( 5S ) . T7 RNA polymerase promoter sequences were introduced during PCR by incorporation of the reverse primer . 2 . 5 μg of total RNA/lane were separated on a 7 M urea/TBE/8% polyacrylamide gel and subsequently transferred to a positively charged nylon membrane ( GE Healthcare ) by electroblotting in 0 . 5×TBE at 15 V for 1 h . Probe hybridization and detection were performed according to the supplier’s instruction ( DIG RNA Labelling kit , Roche Diagnostics ) . β-Galactosidase activity assays were performed as described previously [63] . Activities were determined from exponentially growing cells ( OD600 = 0 . 5–0 . 8 ) if not otherwise indicated . Reported values are the average of at least three measurements using independent cultures . Metabolic labeling of phosphorylated proteins using H3[32P]PO4 was carried out as described previously with slight modifications [64 , 65] . Briefly , bacteria were grown in LB medium to late exponential phase ( OD600 ~ 0 . 5–0 . 8 ) and subsequently expression of plasmid-encoded proteins was induced using 1 mM IPTG . Following an additional incubation for 30 min , cells were washed and further incubated for 1 h in phosphate-depleted TG-medium containing 1 mM IPTG if required . Cells were collected by centrifugation and re-suspended to an OD600 of 0 . 5 in the same medium . Of these suspensions 50 μl were labeled with H3[32P]PO4 as described before [65] . Phosphorylated proteins were separated by 13% SDS-PAGE and finally analyzed by phospho-imaging ( Typhoon FLA-9500; GE Healthcare ) . Cells were grown as described in the section above . Following incubation in TG-medium for 1 h , cultures had a cell density corresponding to OD600 = 2–3 . Cells equivalent to 5 OD600 units were collected by centrifugation and re-suspended in 1 . 8 ml TG-medium ( i . e . OD600 = 2 . 8 ) . Aliquots of the cultures were subjected to SDS-PAGE and Coomassie blue staining or Western blotting to assess proper synthesis of IPTG-inducible proteins . For labeling , 150 μCi H3[32P]PO4 ( Hartmann Analytic ) were added to the cells and incubation was continued for 45 min at 37°C . If required , 150 μM L-epinephrine ( Sigma-Aldrich ) was added 5 min prior addition of H3[32P]PO4 . Following labeling , cells were pelleted and lysed in 350 μl lysis buffer ( 100 mM Tris/HCl pH 7 . 5 , 200 mM KCl , 20% sucrose , 1 mM EDTA , 1 mg/ml Lysozyme , 20 μg/ml RNase A ) by at least 5 freeze thawing cycles . In case of pulse-chase experiments , the assay was scaled up accordingly , i . e . 20 OD600 units cells were collected and re-suspended in 7 . 2 ml TG-medium containing IPTG and 400 μCi H3[32P]PO4 were added for labeling . Labeling was stopped after 45 min by addition of 40 mM Na2HPO4 and 20 mM KH2PO4 . Subsequently , 1 . 8 ml samples were removed at indicated times and subjected to lysis and protein pull-down . For pull-down , crude extracts were cleared by centrifugation ( 15000 rpm , 1 h , 4°C ) and subsequently 500 μl buffer W ( 100 mM Tris/HCl pH 8 . 0 , 150 mM NaCl , 1 mM EDTA ) and 10 μl MagStrepXT magnetic beads ( IBA , Germany ) were added to the cleared lysates and further incubated for 30 min on ice . The magnetic beads were washed 2× using 500 μl buffer W and finally dissolved in 50 μl Laemmli buffer ( 62 . 5 mM Tris/HCl pH 6 . 8 , 5% ( v/v ) 2-mercaptoethanol , 2% ( w/v ) SDS , 10% ( v/v ) glycerol , 0 . 05% ( w/v ) bromophenolblue ) . Dissolved beads ( 5–10 μl ) were loaded on SDS-PAA gels ( 12 . 5–15% ) and analyzed by phospho-imaging or Western blotting using anti-Strep antiserum ( 1:20000 , Promokine ) . Loading volumes were adjusted according to protein amounts detected in pilot Western blots . Signal intensities were quantified using software ImageQuant TL 8 . 1 ( GE Healthcare ) . Strep-tagged proteins were purified as described previously [66] . Recombinant proteins were overproduced in strain Z196 using the following plasmids encoding the proteins in parentheses: pDL35 ( Strep-PhoB ) , pYG278 ( QseF-Strep ) , pYG278-D56A ( QseFD56A-Strep ) , pYG279 ( QseF-NTD-Strep ) , pYG279-D56A ( QseF-NTDD56A-Strep ) and pYG280 ( QseF-CTD-Strep ) . Bacteria were grown in 100 ml LB to an OD600 of ~0 . 8 and synthesis of proteins was induced by addition of 1 mM IPTG for 1 h . Cells were harvested by centrifugation ( 20’ , 4000 rpm , 4°C ) , washed in buffer W and disrupted by passage through a French pressure cell . Lysates were cleared by centrifugation ( 14000 rpm , 1 h , 4°C ) and loaded on pre-equilibrated columns containing 100 μl StrepTactin resin ( IBA , Germany ) . Samples were 4× washed using 2 ml buffer W prior to elution with 150 μl buffer E ( 100 mM Tris/HCl pH 8 . 0 , 150 mM NaCl , 1 mM EDTA , 2 . 5 mM desthiobiotin ) . Following addition of 20% glycerol , protein fractions were stored at -20°C until further use . Protein samples were dissolved in Laemmli buffer and heated for 5 min at 65°C ( for samples containing magnetic beads heat denaturation was omitted ) . Proteins were separated on 12 . 5–15% SDS PAA gels and blotted onto a polyvinylidene difluoride ( PVDF ) membrane ( GE Healthcare ) by semi-dry blotting for 60–120 min at 2 . 0 mA/cm2 . Rabbit polyclonal antisera directed against the 3×FLAG-Tag ( Lactan ) and the Strep-epitope ( Promokine ) were used in a dilution of 1:5000 and 1:20000 , respectively , containing 3% BSA . The phospho-threonine specific antibody ( Cell Signaling Technology ) was used in a dilution of 1:2000 containing 5% BSA . Primary S1 antiserum was used in a 1:20000 dilution containing 3% BSA . Secondary goat anti-rabbit IgG antibodies conjugated to alkaline phosphatase ( 1:100 000; Promega ) were used together with the CDP* detection system ( Roche Diagnostics ) to detect the primary antibodies . The maltose binding protein MalE was detected using recombinant monoclonal mouse anti-MBP antibody ( Sigma Aldrich ) in a dilution of 1:10000 . The primary antibody was detected using a secondary HPR coupled anti-mouse antibody . MalE protein was visualized using the Westar sun ECL system ( WESTAR ) and a chemiluminescence detector ( ChemiDoc , BioRad ) . Ligand fishing experiments were carried out as described previously [4 , 66] . Bait plasmids for expression of the Strep-tag only ( pBGG237 , negative control ) or QseG-Strep ( pYG191 ) were introduced into strains Z952 and Z986 carrying qseE-3xFLAG or phoQ-3xFLAG preys on the chromosome , respectively . Cells were grown in LB to late exponential phase and expression of bait proteins was induced with 1 mM IPTG for one additional hour . Cells were harvested , lysed and proteins were purified by StrepTactin affinity chromatography as described before [66] . Eluates were diluted in Laemmli buffer and separated by SDS-PAGE followed by Western blotting analysis using an anti-FLAG antiserum . For SDS-PAGE 5 μg of the cleared lysates ( total extracts ) were loaded onto the gel . Output samples were normalized to the eluted QseG-Strep bait protein amount ( i . e . 0 . 5 μg QseG-Strep ) in case QseG-Strep was the bait ( Fig 7A , lanes 7 and 9 ) . Corresponding volumes of the eluates obtained from the Strep-tag only co-purifications ( Fig 7A , lanes 6 and 8 ) were loaded . For monitoring of protein-protein interactions in vivo , the BACTH system was used [45 , 46] . BACTH relies on reconstitution of activity of the split Bordetella pertussis adenylate cyclase toxin . Reconstitution and thus cAMP production occurs through interaction of candidate proteins fused to the separately encoded T18- and T25-fragments of the B . pertussis toxin . The plasmid-encoded fusion genes are tested in E . coli strain BTH101 , which lacks endogenous adenylate cyclase activity . Interaction can be monitored quantitatively by measuring activity of β-galactosidase , whose synthesis depends on cAMP-CRP . Plasmid pKT25 and pUT18C were used for construction of in-frame fusions of the candidate genes to the 3′ ends of the sequences encoding T25 and T18 , respectively . Plasmid pUTM18C is a derivative of pUT18C that allows translocation of the C-terminally fused candidate protein into the periplasm , while the N-terminal T18-fragment remains in the cytoplasm [46] . BTH101 was co-transformed with the plasmids carrying the desired T18 and T25 fusion genes using selection for kanamycin and ampicillin . The resulting transformants were grown at 28°C in selective LB medium containing 1 mM IPTG for inactivation of the Lac repressor and β-galactosidase activities were determined from cells grown to the stationary phase . The qseE gene was amplified by error prone PCR [67] using primers BG646/BG647 and plasmid pYG199 as template . Three independent reactions were performed and PCR products were digested with PstI and BspHI . The 789 bp DNA fragment carrying the qseE-5’ end was isolated and used to replace the corresponding wild-type sequence in the BACTH plasmid pYG199 . The ligation reactions were introduced into strain BTH101 carrying plasmid pYG242 coding for T18-TMoppB-QseG and recombinants were selected at 28°C on LB agar plates containing the required antibiotics , 40 μg/ml X-Gal and 1 mM IPTG . Plasmids were isolated from colonies exhibiting colorless or pale blue phenotypes indicating impaired QseE/QseG interaction and re-introduced into BTH101/pYG242 to confirm persistence and uniformity of the phenotype . Plasmids passing this test were isolated once more and sequenced . Plasmids carrying qseE alleles with stop- or frameshift mutations were not further analyzed . Finally , two plasmids ( named pYG199_1 . 6 and pYG199-TM1; S2 Table ) were obtained encoding QseE variants with amino acid exchanges . | Bacteria use two-component systems , composed of a membrane-bound histidine kinase and a cytoplasmic response regulator , to sense environmental cues and adapt gene expression accordingly . The enterobacterial two-component system QseE/QseF controls functions related to the cell envelope . In pathogens , QseE/QseF were suggested to regulate virulence in response to the host hormone epinephrine , a process known as interkingdom signaling . Here , we analyzed the role of qseG , which co-localizes with qseE/qseF and encodes a periplasmic lipoprotein . We show that QseG is a prerequisite for QseE/QseF activity in E . coli K-12 . Without QseG , kinase QseE is unable to phosphorylate and activate response regulator QseF . Furthermore , QseG and QseE interact in the periplasm and a mutation in QseE impairing interaction concomitantly decreases activity . Our data reveal a regulatory cascade likely conserved in other Enterobacteriaceae , in which membrane-bound QseE is stimulated through interaction with periplasmic QseG to phosphorylate cytoplasmic QseF . Finally , we show that epinephrine is a minor stimulus for QseE/QseF in E . coli K-12 and that its sensing strictly depends on QseG . Therefore , QseG is not only required for activity of QseE/QseF , but also for signal perception . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"cell",
"physiology",
"blood",
"serum",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"body",
"fluids",
"chemical",
"compounds",
"neuroscience",
"organic",
"compounds",
"plasmid",
"construction",
"hormones",
"membrane",
"proteins",
"regulat... | 2018 | Interaction of lipoprotein QseG with sensor kinase QseE in the periplasm controls the phosphorylation state of the two-component system QseE/QseF in Escherichia coli |
For a rapid induction and efficient resolution of the inflammatory response , gene expression in cells of the immune system is tightly regulated at the transcriptional and post-transcriptional level . The control of mRNA translation has emerged as an important determinant of protein levels , yet its role in macrophage activation is not well understood . We systematically analyzed the contribution of translational regulation to the early phase of the macrophage response by polysome fractionation from mouse macrophages stimulated with lipopolysaccharide ( LPS ) . Individual mRNAs whose translation is specifically regulated during macrophage activation were identified by microarray analysis . Stimulation with LPS for 1 h caused translational activation of many feedback inhibitors of the inflammatory response including NF-κB inhibitors ( Nfkbid , Nfkbiz , Nr4a1 , Ier3 ) , a p38 MAPK antagonist ( Dusp1 ) and post-transcriptional suppressors of cytokine expression ( Zfp36 and Zc3h12a ) . Our analysis showed that their translation is repressed in resting and de-repressed in activated macrophages . Quantification of mRNA levels at a high temporal resolution by RNASeq allowed us to define groups with different expression patterns . Thereby , we were able to distinguish mRNAs whose translation is actively regulated from mRNAs whose polysomal shifts are due to changes in mRNA levels . Active up-regulation of translation was associated with a higher content in AU-rich elements ( AREs ) . For one example , Ier3 mRNA , we show that repression in resting cells as well as de-repression after stimulation depends on the ARE . Bone-marrow derived macrophages from Ier3 knockout mice showed reduced survival upon activation , indicating that IER3 induction protects macrophages from LPS-induced cell death . Taken together , our analysis reveals that translational control during macrophage activation is important for cellular survival as well as the expression of anti-inflammatory feedback inhibitors that promote the resolution of inflammation .
In their function as innate immune cells , macrophages are highly sensitive to endogenous and exogenous danger signals . They sense pathogen-associated molecular patterns through Toll-like receptors ( TLRs ) and mount a tightly controlled immune response . The secretion of cytokines and chemokines by macrophages recruits , activates and polarizes other immune cells , while reactive oxygen species and phagocytosis directly kill microorganisms . Lipopolysaccharide ( LPS ) , a cell wall component of gram-negative bacteria , potently activates macrophages via TLR4 . Upon receptor ligation , the NF-κB pathway together with the p38 MAPK , ERK1/2 and JNK pathways causes a highly orchestrated , transient induction of numerous inflammatory genes . Such dynamic gene expression patterns are achieved by regulation at multiple levels , as exemplified by the pro-inflammatory cytokine TNF . The promoter of Tnf contains a cAMP responsive element and binding sites for NFAT , ETS1/ELK1 , SP1 , EGR proteins and NF-κB [1] . LPS also acts at the post-transcriptional level and controls the splicing , nuclear export , stability and translation of Tnf mRNA [2] . In their 3′ untranslated region ( UTR ) , many cytokine mRNAs contain an AU-rich element ( ARE ) , which recruits specific RNA-binding proteins [3] . In resting cells , TIA1 , FXR1 and ZFP36 ( also known as TTP ) recognize the ARE and repress Tnf mRNA translation [4]–[6] , and ZFP36 additionally causes degradation of Tnf mRNA [7] . Activation of the p38 MAPK pathway leads to the phosphorylation of ZFP36 , whereby Tnf mRNA becomes partially stabilized and its translation activated [6] , [8] . MicroRNAs [3] and a recently discovered stem-loop motif that acts as a constitutive RNA decay element ( CDE ) [9] further suppress the expression of Tnf and other immune-related mRNAs at the post-transcriptional level . Not only rapid induction , but also the timely shut down of inflammatory responses is essential for immune homeostasis . The acute , excessive and systemic release of TNF , for example , can lead to septic shock , while the chronic production of pro-inflammatory cytokines sustains auto-immune diseases such as rheumatoid arthritis and Crohn's disease . In contrast , physiological immune responses induce negative feedback loops that resolve inflammation . TLR4 signaling , for example , limits itself by the induction of inhibitors that interfere with signaling complexes downstream of TLR4 . Activation of the NF-κB pathway occurs via the proteasomal degradation of the NF-κB inhibitor NFKBIA ( IκBα ) , which retains NF-κB dimers in the cytoplasm . Once in the nucleus , NF-κB dimers activate the transcription of target genes , which comprise not only cytokines but also inhibitors of NF-κB that re-export nuclear NF-κB to the cytoplasm , degrade it in the nucleus or prevent it from binding to target promoters [10] , [11] . The cytoplasmic NFKBIA pool is re-filled by NF-κB-induced transcription of Nfkbia , whereby NF-κB activation terminates itself . In addition , rapid shut down of gene expression requires clearance of the previously synthesized mRNAs . Activation of ZFP36 , for example , causes degradation of 8% of all LPS-induced mRNAs [12] , and more generally the regulation of mRNA half-lives was found to strongly shape the temporal expression pattern of inflammatory genes in macrophages and dendritic cells [13] , [14] . While translation of Tnf mRNA has been studied extensively as an individual example , the general role of translational regulation during macrophage activation remains unclear . Parallel measurements show a poor correlation between mRNA and protein abundance in many systems [15] , in line with the notion that translation efficiency is a major determinant of steady-state protein levels in mouse fibroblasts [16] . So far , three studies addressed the role of translational regulation at a transcriptome-wide scale during activation of innate immune cells: In LPS-stimulated dendritic cells , mRNAs of ribosomal proteins were found to be translationally repressed , which correlated with a global drop in translation in the late phase of activation [17] . In monocytes stimulated with interferon gamma , the so-called GAIT ( gamma interferon-activated inhibitor of translation ) element was found to cause translational inhibition of several chemokine and chemokine receptor mRNAs [18] . Both studies focused on late stages of the response when inflammatory gene expression is shut off by negative feedback loops . A third study described translational repression of mRNAs encoding components of the mitochondrial respiratory chain early after LPS-stimulation of J774 . 1 cells [19] . Our goal was to investigate translational regulation early during immune activation in a more comprehensive manner by including an assessment of changes in mRNA levels . In the present study , we conducted a systematic genome-wide analysis to explore the contribution of changes in mRNA translation to the early phase of macrophage activation . To identify individual mRNAs whose translation is regulated after stimulation with LPS , we measured polysome association of mRNAs by sucrose density gradient fractionation and microarrays ( Figure 1 , left part of scheme ) . Because changes in mRNA levels are especially strong early after cell activation and can affect ribosome density without active regulation of translation efficiency , we additionally quantified mRNA levels by RNASeq at a high temporal resolution ( right part of scheme ) . By combining both data sets , we were able to distinguish mRNAs whose translation is actively regulated from mRNAs whose translational changes are more likely to be passive . Our analysis revealed that the most frequent mode of active regulation is translation de-repression , which is prominent among inhibitors of NF-κB signaling and a factor that supports macrophage survival .
mRNAs can be separated according to the number of bound ribosomes by sucrose density gradient centrifugation , a method widely applied to monitor changes in translation initiation [20] . Polysome profiles revealed a significant increase in the percentage of polysomal ribosomes ( from 74 to 82% , n = 3 ) after stimulation of RAW264 . 7 macrophages with LPS for 1 h ( Figure 2A ) , indicating that a higher proportion of ribosomes is engaged in translation . Such a change can be due to increased initiation rates , decreased elongation rates or a reduced stoichiometric ratio of ribosomal subunits to mRNA molecules in the cell . Since we did not observe a significant change of protein synthesis by measuring [35S]-methionine/cysteine incorporation ( Figure S1 ) , the relative increase in polysomes is most likely not due to a general increase in the rate of translation initiation . In order to assess translation of individual mRNAs , RNA was extracted after sucrose gradient fractionation and pooled as follows: free RNA ( F ) , 40S-associated RNA ( S ) , RNA associated with 1–3 ribosomes ( L , light polysomes ) and RNA associated with >3 ribosomes ( H , heavy polysomes ) ( Figure 2B ) . As a control for equal purification efficiency , we added a rabbit β-globin ( HBB2 ) in vitro transcript ( Figure 2B ) . The RNA pools from three biological replicates were separately hybridized to microarrays , and after annotating the probes to all mouse RNAs in the RefSeq database , the distribution of every RNA across the four pools was calculated ( GEO accession number GSE52451 ) . As expected , protein-coding RNAs ( mRNAs ) showed preferential association with heavy polysomes ( H ) , whereas non-translated RNAs such as small nucleolar RNAs ( snoRNAs ) showed a different distribution ( Figure 2C ) . The global up-regulation of polysome-association in LPS-activated macrophages ( Figure 2A ) was reflected by a shift within the polysome fractions: The median proportion of mRNAs in H increased from 69% to 73% , the median proportion in L decreased from 25% to 21% ( Figure 2C ) . In contrast to ribosome profiling , which uses sequencing of ribosome-protected fragments to obtain the density of ribosomes as a proxy for translation efficiency [21] , association with different polysome fractions ( F , S , L and H in our analysis ) reflects the absolute number rather than the average density of ribosomes on an mRNA . A disadvantage of our method is that association with polysome fractions might not be directly proportional to the number of associated ribosomes , and that our method does not provide information on the position of ribosomes on the mRNA . On the other hand , our method preserves information on the distribution of mRNAs along the polysome profile , which is lost in ribosome profiling . Since mRNAs in general showed only minimal association with F and S ( Figure 2C ) , we concluded that most mRNA molecules in RAW264 . 7 macrophages are associated with at least one ribosome . Therefore , we restricted our analysis to L and H for identification of mRNAs regulated individually at the level of translation . As a measure for the translation efficiency , we calculated for every mRNA the ratio of its proportion in H to its proportion in L ( H/L ) , which represents a measure of ribosome load . Ncl mRNA , for example , has a very high ribosome load ( Figure 3A , left panel and Figure 2B ) , and does not change its position after LPS stimulation . Some mRNAs , such as Nfkbiz ( Figure 3A , middle panel ) , show a strong shift from L to H , while others , such as Cpd ( right panel ) , shift in the opposite direction against the general trend of increased polysome association . By this approach , H/L was determined for 14 , 320 mRNAs in macrophages stimulated for 1 h with LPS and plotted against H/L in unstimulated macrophages ( Figure 3B ) . The regression line in this plot was shifted “upwards” ( y-intercept at 0 . 3 ) , which reflects the general increase in polysome association after LPS stimulation . The extent to which the ribosome load of an individual mRNA deviates from the general trend corresponds to its orthogonal distance from the regression line . We chose 2 SD from the regression line as our cut-off , and asked that the mean value and at least two out of three biological replicates were outside of this cut-off . By applying these criteria , we identified 90 mRNAs that increase their ribosome load after LPS- stimulation ( Figure 3B , orange dots , and Table S1 ) , and 129 mRNAs that decrease their ribosome load after LPS stimulation ( Figure 3B , blue dots , and Table S2 ) . For 20 mRNAs selected across the entire range of observed shifts , we confirmed the change in H/L by qPCR and obtained an excellent correlation between the qPCR and the microarray data ( Pearson's correlation coefficient RP = 0 . 96 , p = 2 . 4×10−11 , Figures 3C and S2 ) . Importantly , 15 out of 19 mRNAs showed a similar shift in ribosome load in bone-marrow derived macrophages ( BMDM , polysome fractionation shown in Figure S3 ) , demonstrating that the observed regulation also occurs in primary cells ( RP = 0 . 86 , p = 1 . 8×10−6 , Figure 3D and S2 ) . To our surprise , only four cytokines ( out of 68 detectable in both conditions , see Table S3 ) were translationally up-regulated ( Tnf , Cxcl2 , Il23a and Tnfsf9 ) , whereas 8 feedback inhibitors of TLR4 signaling ( out of 51 , see Table S4 ) were among the most highly up-regulated at the level of translation ( Figure 3B , hypergeometric p-value for enrichment of feedback inhibitors = 9×10−10 ) . NFKBID ( IκBδ ) , NFKBIZ ( IκBζ ) and IER3 are reported to be direct inhibitors of RELA ( p65 subunit of NF-κB ) transactivation , while NR4A1 is a transcription factor for NFKBIA . DUSP1 inhibits the p38 MAPK pathway . ZFP36 , ZFP36L2 ( BRF1 ) and ZC3H12A are mRNA-binding proteins that inhibit the expression of cytokines at the post-transcriptional level . Among the translationally down-regulated mRNAs , we found one feedback inhibitor ( Tnip3 ) and three cytokines ( Csf3 , Il1b and Lif ) . As suggested previously [20] , [22] , changes in mRNA levels can affect the average ribosome load of a transcript . When newly transcribed mRNA exits the nucleus , there is a delay until it is fully loaded with ribosomes , which leads to a transient increase of its proportion in the free or light fractions . Likewise , degradation of mRNAs is thought to occur preferentially on non-translated transcripts , which will reduce its proportion in the free or light fractions and cause a relative increase in heavy fractions [20] . Such passive shifts in ribosome load would be most prominent in the early phase of cell activation , when mRNA levels change strongly . To determine whether changes in mRNA levels are responsible for some of the shifts we observe in our ribosome load data set , we isolated total RNA at a high temporal resolution during the first 2 h of macrophage activation , and measured transcriptome-wide mRNA expression patterns by RNASeq ( n = 1 , GEO accession number GSE52451 ) . Indeed , we found that the change of translation and the change of mRNA abundance show a significant negative correlation ( Figure 4A , left panel , Spearman's rank correlation coefficient RS = −0 . 23 ) . Thus , changes in mRNA levels have to be taken into account when interpreting polysome association . To do so , all mRNAs with a significant change in expression were divided into four groups that reflect the behavior of mRNA abundance before and immediately after 1 h of stimulation , the time point when translation was measured ( Figure 4B ) . Within these groups , mRNA expression patterns correlate well with each other ( Figure 4C ) . All mRNAs without a significant change in expression are in group ( g ) 0 ( Figure 4D ) . g1 contains 1123 mRNAs whose levels go up at the 1 h time point and reach their first significant maximum at or after 1 h of stimulation . As predicted , we found that the ribosome load of g1 mRNAs is significantly decreased by LPS stimulation compared to g0 mRNAs ( p<0 . 001 , two-sided Wilcoxon rank sum test , Figure 4E ) . g1 mRNAs comprise 57 . 4% of all mRNAs that were identified as translationally down-regulated ( Table S2 ) , but only 7 . 8% of all mRNAs considered in our microarray analysis . g2 contains 1117 mRNAs whose levels go down at the 1 h time point and reach their first significant minimum at or after 1 h of LPS stimulation . Similar to g1 , g2 represents 34 . 4% of the translationally up-regulated mRNAs ( Table S1 ) , but only 7 . 8% of all mRNAs in our analysis . Their ribosome load is significantly increased compared to g0 mRNAs ( p<0 . 001 , Figure 4E ) . Hence , increasing mRNA levels ( group g1 ) are often associated with decreasing ribosome load , while decreasing mRNA levels ( group g2 ) are often associated with increasing ribosome load , as the shape of the bulk distribution in Figure 4A indicates . Therefore , mRNAs in g1 and g2 undergo passive shifts in ribosome load ( Figure 4A , right panel ) . g3 mRNAs ( n = 111 ) show an early and transient induction of expression before 1 h of macrophage activation . As a group these mRNAs do not show a significant difference in their change of ribosome load compared to g0 ( Figure 4E ) , yet many of the mRNAs with the strongest increase in ribosome load are part of g3 ( Table S1 ) . We concluded that the change of ribosome load of g3 mRNAs results from active translational regulation and not from passive shifts due to changes in mRNA levels . This is supported by the fact that we also observed many g3 mRNAs which do not show an increase of their ribosome load although their expression patterns are very similar to those mRNAs whose translation is strongly up-regulated ( Figure S4 ) . Similar to g3 , g4 mRNAs ( n = 45 ) , which show an early and transient decrease before 1 h of LPS stimulation , do not significantly differ from g0 mRNAs in their change of ribosome load ( Figure 4E ) . Because the ARE is often found in the 3′UTR of inflammation-related genes [3] and was shown to regulate translation of Tnf mRNA in LPS-stimulated macrophages [6] , we looked at the frequency of AREs in the different mRNA groups described above . For this purpose , we used the AREScore algorithm , which assigns a numeric value to the putative strength on an ARE [23] . As shown in Figure 5A , mRNAs whose levels increase significantly and therefore belong to group g1 or g3 ( Figure 4 ) have significantly higher AREScores than mRNAs in g0 ( p<0 . 001 , two-sided Wilcoxon rank sum test ) . For translational regulation , we applied the categories as shown on the right side of Figure 4A: Translation was considered to be actively up-regulated unless the mRNA levels were decreasing ( g2 mRNAs ) and actively down-regulated unless the mRNA levels were increasing ( g1 mRNAs ) . By this analysis , a significant increase in AREScores is only observed in the group of mRNAs whose translation is actively up-regulated ( p<0 . 001 , two-sided Wilcoxon rank sum test , Figure 5B ) . We then chose a few examples to further characterize the relationship between mRNA levels , translation and protein production . Because the ribosome load strongly correlates with the length of the open reading frame ( ORF ) ( Figure S5 ) , we defined for each of the selected mRNAs a control group of mRNAs with similar ORF length ( ±25 nt ) . Il1a mRNA , for example , has a much lower ribosome load in activated macrophages than its control group ( Figure 6A , middle panel ) , and although the mRNA is induced >3000-fold ( left panel ) , IL1A protein cannot be detected in the supernatant ( right panel; for positive control and sensitivity of the assay see Figure S6 ) . A similar example is Il1b: Its ribosome load is below that of its control group , and despite a >2500-fold induction of the mRNA during the first 2 h of LPS- stimulation , the protein is not detectably secreted into the supernatant ( Figure 6B ) . Both IL1A and IL1B are translated as precursors that undergo proteolytic cleavage before they are secreted . LPS alone was described to prime macrophages for IL1A and IL1B production , but a second stimulus is required for cleavage of the precursors and efficient secretion [24] . In addition , it has been shown that translational repression of Il1b mRNA is mediated by the Janus kinase TYK2 and strongly contributes to the lack of IL1B secretion by LPS-stimulated macrophages [25] . Our data suggest that Il1a mRNA is subject to similar translational repression , and that the release of these cytokines in macrophages stimulated with LPS alone is prevented through a combination of translation repression and secretion . In contrast to Il1a and Il1b , Ccl4 mRNA is expressed at much higher levels ( 25 rpkm versus 0 . 02 rpkm for Il1a and 0 . 27 rpkm for Il1b ) , yet its translation is repressed compared to the control group ( Figure 6C ) . After LPS stimulation , Ccl4 mRNA is induced 80-fold and its translation is de-repressed , which is reflected by efficient secretion of CCL4 ( Figure 6C ) . Similarly , Tnf mRNA levels are high ( 50 rpkm ) in resting macrophages , and its translation is strongly suppressed ( Figure 6D ) . Upon activation with LPS , Tnf mRNA shows an oscillatory induction and translation is de-repressed . TNF secretion is low but detectable in unstimulated macrophages , and induced efficiently by LPS ( Figure 6D ) . While Tnf is an exception among the cytokines , five translationally regulated feedback inhibitors belong to g3 with peak expression before 1 h of LPS stimulation ( Table S4 ) . Ier3 and Zfp36 mRNAs , for example , show a behavior very similar to Tnf: Their mRNAs are well expressed in resting macrophages ( 8 and 17 rpkm , respectively ) , whereas translation is strongly repressed and the protein is barely detectable ( Figure 6E and 6F ) . Upon LPS stimulation , mRNA levels oscillate , translation is de-repressed , and protein production is induced ( Figure 6E and 6F ) . These examples illustrate that de-repression of translation is a frequent mode of regulation , which we observed for two cytokines and seven negative feedback inhibitors ( see Figure 7 ) . A systematic comparison of the 72 cytokines ( including chemokines , Table S3 ) and 51 feedback inhibitors ( Table S4 ) that we could track in RAW264 . 7 macrophages shows a distinct pattern of regulation for these two groups of genes: The majority of cytokine mRNAs is expressed at very low levels in resting macrophages ( median rpkm = 4 . 75×10−2 ) , and is strongly induced after 1 h of LPS stimulation ( median rpkm = 1 . 89×10−1 , 3 . 97-fold induction ) ( Figure 7A ) . Feedback inhibitors start with higher mRNA levels ( median rpkm = 8 . 94 ) , yet their increase in mRNA levels after 1 h of LPS stimulation is much weaker ( median rpkm = 12 . 67 , 1 . 42-fold induction ) . The effect of LPS on the translation of cytokines is heterogeneous: Most cytokine mRNAs follow the general trend ( Figure 7B ) , only a few show enhanced translation ( Tnf , Il23a , Cxcl2 , Tnfsf9 ) , while others are translated less efficiently ( Il1b , Lif , Csf3 ) ( Figure 3B ) . In contrast , feedback inhibitors more often show a low ribosome load in resting cells and are de-repressed upon LPS stimulation ( e . g . Ier3 , Nfkbid , Nfkbiz , Nr4a1 , Dusp1 , Zfp36 and Zc3h12a; Figure 7B ) . TNF is the most highly induced cytokine at the level of translation ( Figure 3B ) , and its mRNA levels oscillate with a first peak during the first hour of the LPS response ( Figure 6D ) . Several feedback inhibitors of NF-κB signaling have a very similar behavior . Out of all g3 mRNAs , Nfkbia and Ier3 show the highest correlation with the expression profile of Tnf mRNA ( RP = 0 . 96 and 0 . 85 , respectively ) . While translation of Nfkbia is not regulated ( Table S4 ) , Ier3 , like Tnf , is translationally de-repressed after LPS stimulation ( Figure 6E ) . Moreover , the two genes show a striking similarity of regulatory elements . The promoters of both Tnf and Ier3 contain binding sites for the transcription factors NF-κB , ETS1 and SP1 ( Figure 8A ) . In their 3′UTRs , they share three post-transcriptional regulatory elements: Both Tnf and Ier3 mRNA contain a highly conserved CDE stem loop [9] and in mouse , both harbor a miR-125b binding site , which was shown to regulate the expression of mouse Tnf mRNA [26]; both mRNAs also contain an ARE and are validated targets of ZFP36 [7] , [27] . By Luciferase assays , we confirmed that the 3′UTR of Ier3 mediates translational regulation upon LPS stimulation . Due to the low transfection efficiency in RAW264 . 7 cells , we used HEK293 cells that stably express the TLR4 receptor and are therefore responsive to LPS . Because the reporter mRNAs were expressed from a heterologous MMLV promoter in HEK293 cells , their levels did not change strongly upon LPS stimulation ( Figure 8B ) . In unstimulated cells , translation of the reporter mRNA containing the complete Ier3 3′UTR was suppressed more than 2-fold compared to the control reporter that contains the rabbit β-globin ( HBB2 ) 3′UTR alone ( Figure 8C ) . After stimulation with LPS , translation of the Ier3 3′UTR reporter was significantly increased 2 . 9-fold and was even more efficient than translation of the control reporter . When the ARE was deleted from the Ier3 3′UTR , translation was neither repressed in resting cells nor enhanced after stimulation with LPS ( Figure 8C ) , which indicates that the ARE is involved in translational regulation of Ier3 mRNA . Since the transcriptional and post-transcriptional regulation of Ier3 parallels that of Tnf , we speculated that IER3 may play an important role during early macrophage activation . IER3 was reported to inhibit NF-κB and limit induction of CCL2 , IL6 , CXCL1 and IL1B upon TLR2 stimulation [28] . Hence , we first tested whether IER3 might also affect TNF expression after TLR4 ligation . BMDM were derived from wt and Ier3−/− mice [29] , and confirmed to be equally differentiated by the time of the experiment ( Figure S7A ) . After LPS stimulation , both wt and Ier3−/− BMDM secreted similar amounts of TNF ( Figure 9A ) , suggesting that IER3 does not affect TNF production . IER3 was also reported to have , depending on the system , pro- or anti-apoptotic effects [30] . We therefore compared cell death in wt and Ier3−/− BMDM . Prior to LPS stimulation , there was no significant difference in the proportion of dead cells . After LPS stimulation , Ier3−/− BMDM showed twice as many dead cells as wt controls ( Figure 9B ) . Dead BMDM were permeable for both Annexin V and propidium iodide ( PI ) , whereas the early apoptotic population ( Annexin V positive , PI negative ) was barely affected by LPS stimulation ( Figure S7B ) . We concluded that the induction of IER3 has a protective effect and contributes to the survival of macrophages during the early inflammatory response .
When macrophages initiate an inflammatory response , numerous secreted and intracellular proteins have to be synthesized . Despite efficient translation initiation in LPS-stimulated macrophages , the transcription of new mRNAs does not lead to an immediate increase in protein production . Rather , newly transcribed mRNAs have to compete for components of the translation machinery , and some mRNAs are loaded more efficiently with ribosomes than others . Among the 90 mRNAs that we identified as translationally up-regulated early after macrophage stimulation , 20 show a significant increase in mRNA levels ( group 1 or group 3 in Table S1 ) . For these 20 , the increase in transcription and/or mRNA stability acts in synergy with enhanced translation , allowing for efficient induction of protein synthesis ( Figure 4 ) . In the subset of genes with concomitant induction of mRNA levels and translation , we find four cytokines , including Tnf , and two genes required for the efficient induction of Tnf: Map3k8 [31] and Dusp2 [32] . Surprisingly , the largest group of genes with concomitant induction encode negative feedback regulators of the inflammatory response ( Table S4 ) : Three direct inhibitors of RELA transactivation ( Nfkbid , Nfkbiz , Ier3 ) , one transcription factor for Nfkbia ( Nr4a1 ) , one phosphatase that inactivates p38 MAPK ( Dusp1 ) and two RNA-binding proteins that inhibit the expression of cytokines at the post-transcriptional level ( Zfp36 and Zc3h12a ) . Four out of the seven genes ( Nfkbid , Dusp1 , Zfp36 and Zc3h12a ) are antagonists of septic shock in LPS-injected mice [33]–[36] . Translational regulation of Nfkbid , Nfkbiz and Zc3h12a was also described in IL1A-stimulated HeLa cells [37] , suggesting that translational control observed in macrophages also operates in other cell types . For many of the translationally up-regulated mRNAs we observe that translation is repressed in resting cells in comparison to a control group of mRNAs with similar ORF length ( Figure 6 ) . For Tnf , the importance of suppression at the post-transcriptional level was demonstrated by deletion of the ARE in the 3′UTR: Mice lacking the ARE in one Tnf allele spontaneously develop chronic inflammatory arthritis and inflammatory bowel disease [38] . In contrast to most other cytokines , Tnf shows comparatively high mRNA levels in resting cells ( Figure 6D ) , which explains why suppression at the translational level is crucial . The advantage of such an expensive system is that high levels of a labile and translationally repressed mRNA prior to stimulation allow for the immediate induction of protein synthesis . Several feedback inhibitors show a pattern similar to Tnf: The mRNA is already transcribed but translationally repressed in resting cells ( Figure 7 ) . Stimulation with LPS relieves translational repression and induces a transient or oscillatory induction of mRNA expression ( Figure 6 and S4 ) . Several of the translationally up-regulated cytokines and feedback inhibitors also share regulatory elements in their mRNAs: Tnf , Cxcl2 , Il23a , Dusp1 , Ier3 and Zfp36 contain known AREs [3] , [39] . Indeed , by the analysis of reporter mRNA translation we could show that the ARE of Ier3 is required for both repression of translation in unstimulated cells and de-repression upon stimulation with LPS ( Figure 8 ) . Moreover , the entire group of mRNAs that we identified as actively up-regulated at the level of translation ( Figure 4A ) has a significantly higher content in predicted AREs than mRNAs whose translation is not regulated ( Figure 5B ) . This group of 59 mRNAs encodes proteins of various functions , which suggests that translational regulation mediated by AREs is involved in multiple processes besides cytokine expression and negative feedback . Notably , ZFP36 , an ARE-binding protein that was shown to repress translation of Tnf mRNA [6] , is part of this group and is therefore induced along with its targets ( Figure 6F ) . During the first hours of the response to LPS , however , ZFP36 is strongly phosphorylated ( Figure 6F , right panel ) and therefore not active as a translation suppressor [6] . Nevertheless , not all mRNAs with an ARE are translationally up-regulated ( Figure 5B ) . In fact , some cytokines that are well known to bear a strong ARE , such as Il1b and Csf3 , show a decrease in ribosome load after LPS stimulation ( Figure 3B and 6B ) . This might be due to the strong increase in mRNA levels , which is typical for cytokines ( Figure 7A ) , or due to other regulatory mechanisms . In addition to the ARE , CDE stem-loop motifs accelerate decay of Tnf mRNA in both resting and LPS-stimulated macrophages , and are also active in Nfkbid , Nfkbiz and Ier3 mRNA [9] . Therefore , it appears that similar post-transcriptional mechanisms mediate the switch from suppression to rapid production of Tnf and feedback inhibitors . Interestingly , our data suggest that translational repression of feedback inhibitors in resting macrophages is just as important as inhibition of pro-inflammatory effectors like Tnf . Presumably , suppression of feedback inhibitors renders cells susceptible to stimulation . Immediately after stimulation , negative feedback loops are induced through translational de-repression of abundant , pre-existing mRNAs encoding feedback inhibitors , which are important for turning off the inflammatory response . In contrast to these feedback inhibitors , the induction of most cytokines strongly relies on a rapid increase in mRNA levels ( Figure 7 ) . The highly similar expression patterns of Tnf and Ier3 led us to investigate the role of Ier3 during early macrophage activation . Besides multiple promoter elements , Tnf and Ier3 share several post-transcriptional regulatory elements in their 3′UTRs ( Figure 8A ) . Our analysis revealed that both mRNAs oscillate and are translationally de-repressed in LPS-stimulated macrophages ( Figure 6 ) . Taking both mRNA abundance and translation into account , Ier3 shows the highest correlation with Tnf among all genes in our data set . IER3 was reported to inhibit the production of the pro-inflammatory cytokines CCL2 , IL6 , CXCL1 and IL1B in macrophages upon TLR2 stimulation [28] . In a mouse model of inflammatory bowel disease , Ier3−/− mice showed an aggravated phenotype with a stronger activation of the NF-κB pathway and increased cytokine production [28] . LPS-stimulated Ier3−/− BMDM , however , did not produce more TNF than wt macrophages ( Figure 9A ) . Inhibitors of NF-κB including NFKBIZ , NR4A1 and IER3 regulate not only cytokine expression but also the susceptibility to apoptosis [30] , [40] , [41] . Indeed we found that twice as many macrophages die after 2 and 4 h of LPS treatment when Ier3 is deleted ( Figure 9B ) , demonstrating that Ier3 protects macrophages from LPS-induced cell death . In our experiments , BMDM became Annexin V-positive and permeable for PI ( Figure S7B ) , which is indicative of necrosis rather than apoptosis . Macrophage activation involves a complex network of pro- and anti-inflammatory signals , which ensures that inflammation is initiated , but also limited and resolved in due time . On this tightrope walk of immune homeostasis , macrophages not only have to find the right balance between activators of inflammation and negative feedback regulators , but also protect themselves from damage . While post-transcriptional regulation of cytokines has been studied extensively , our work reveals that translational regulation primarily controls feedback inhibitors .
RAW264 . 7 cells were cultured in Dulbecco's modified Eagle's medium ( DMEM , Gibco ) supplemented with 10% fetal bovine serum ( FBS , Biochrome ) , 2 mM L-Glutamine , 100 U/ml penicillin and 0 . 1 mg/ml streptomycin ( all PAN Biotech ) at 37°C in a humidified atmosphere with 5% CO2 . HEK293 cells stably expressing the TLR4 receptor ( HEK-Blue mTLR4 ) were purchased from Invivogen and cultured in DMEM containing 10% FBS , 2 mM L-Glutamine , 100 U/ml penicillin , 0 . 1 mg/ml streptomycin , 0 . 1 mg/ml Normocin and 1× HEK-Blue Selection ( Invivogen ) at 37°C in a humidified atmosphere with 5% CO2 . To obtain BMDM , tibia and femur of wild type or Ier3−/− mice [29] were flushed with PBS . Bone-marrow cells were frozen at −80°C in FBS with 10% DMSO at a density of 1 . 6×107 cells/ml and stored in liquid nitrogen . For differentiation of BMDM , bone-marrow cells were cultured in DMEM with 10% FBS , 2 mM L-Glutamine , 100 U/ml penicillin , 0 . 1 mg/ml streptomycin , 55 µM β-mercaptoethanol , 18 mM HEPES and 30% conditioned medium of L929 cells . BMDM were stained with APC anti-mouse/human ITGAM ( CD11b; BLD-101211 ) and Alexa Fluor ( R ) 488 anti-mouse EMR1 ( F4/80; BLD-123119 ) to assess differentiation by flow cytometry . RAW264 . 7 macrophages or BMDM were stimulated with 100 ng/ml LPS ( E . coli O111:B4 , Sigma L2630 ) for 1 h . Ribosomes were stalled by addition of 100 µg/ml cycloheximide ( CHX ) for 5 min , and cells were lysed in Polysome lysis buffer ( 15 mM Tris-HCl pH 7 . 4 , 15 mM MgCl2 , 300 mM NaCl , 1% Triton-X-100 , 0 . 1% β-mercaptoethanol , 200 U/ml RNAsin ( Promega ) , 1 complete Mini Protease Inhibitor Tablet ( Roche ) per 10 ml ) . Nuclei were removed by centrifugation ( 9300× g , 4°C , 10 min ) and the cytoplasmic lysate was loaded onto a sucrose density gradient ( 17 . 5–50% in 15 mM Tris-HCl pH 7 . 4 , 15 mM MgCl2 , 300 mM NaCl and , for fractionation from BMDM , 200 U/ml Recombinant RNAsin Ribonuclease Inhibitor , Promega ) . After ultracentrifugation ( 2 . 5 h , 35 000 rpm at 4°C in a SW60Ti rotor ) , gradients were eluted with a Teledyne Isco Foxy Jr . system into 16 fractions of similar volume . A rabbit HBB2 in vitro transcript was added to each fraction as a spike-in control and RNA was purified by phenol chloroform extraction . To assess RNA quality and equal purification efficiency across all fractions , the HBB2 in vitro transcript and endogenous Ncl mRNA were detected by Northern blotting . 1 . 5×106 RAW cells per 6 well plate were seeded 8–12 h before experiments in DMEM supplemented with 10% dialyzed FBS ( PAA Laboratories ) , 2 mM L-Glutamine , 100 U/ml penicillin , and 0 . 1 mg/ml streptomycin ( all PAN Biotech ) . Before treatment , cells were cultured in methionine- and cysteine-free medium for 1 h . Cells were treated with 100 ng/ml LPS for 1 h , and for the last 30 minutes of the treatment , 11 µCi of [35S]-labeled methionine and cysteine ( EasyTag; PerkinElmer ) was added to each well . Cells were then washed with 1× PBS , collected , and solubilized in 150 µl of lysis buffer containing 15 mM Tris , pH 7 . 4 , 15 mM MgCl2 , 300 mM NaCl , and 1% Triton-X-100 . After centrifugation at 7800× g for 3 min , proteins were precipitated out of the supernatant by spotting 20 µl of each lysate onto Whatman paper and soaking in 5% trichloroacetic acid followed by acetone . Incorporation of 35S was measured in 4 ml of Ultima Gold F ( PerkinElmer ) using a scintillation counter ( LS 6000IC; Beckman Coulter , Brea , CA ) . For normalization , the total protein concentration of each sample was determined using the bicinchoninic acid protein assay reagent kit ( Sigma-Aldrich ) . Cytoplasmic RNA and RNA from polysome fractions was quantified with GeneChip Mouse Gene 1 . 0 ST Arrays . Labeling , hybridization and scanning were performed by the GeneCore Genomics Core Facility at EMBL , Heidelberg . Random primers were used for cDNA synthesis with the Ambion WT Expression Kit to avoid any bias due to poly ( A ) tail length . Labeling was performed with the Affymetrix GeneChip WT Terminal Labeling Kit . Probe sequences of all perfect match probes were retrieved using the Bioconductor [42] package oligo ( version 1 . 22 . 0 ) [43] . Probes were mapped to the mouse RefSeq transcriptome as downloaded from the UCSC Genome Browser mm10 refGene table on February 5 , 2013 . Probes with perfect complementarity to transcripts of more than one gene ( as defined by a common gene symbol ) were excluded . For mapping and further processing of probe information , the R packages seqinr [44] and Biostrings [45] were used together with in-house developed Perl scripts . Expression values were quantile normalized and summarized at the gene level with the basicRMA ( ) function of the Bioconductor package oligo and the target gene symbols as probe set names . The different pools ( cytoplasmic , free , 40S-associated , light and heavy ) were pre-processed as separate groups ( 6 samples per group ) , because their signal distributions might differ due to biological and not technical reasons and therefore should not be quantile normalized together . To obtain the proportion of each mRNA in a specific pool , we had to take into account how much of each pool was used for quantification . After pre-processing , the signals were corrected for the different average proportions of each pool that were used for cDNA synthesis . For example , on average 14 . 2% ( volume ) of the free RNA pool ( F , control condition ) was used for cDNA synthesis , but only 0 . 4% of the heavy polysome pool ( H , control condition ) . The corrected signal of an individual mRNA in a specific pool was then divided by the sum of its signal in all four pools . Only protein-coding genes with at least four specific probes and well detectable expression values in the cytoplasmic samples of treated and untreated cells were included into our analysis . Pre-processed expression values and the distribution over the four pools are represented in Dataset S1 . Post-transcriptional inhibitors of cytokine expression and genes involved in negative feedback loops of the TLR4 response and NF-κB signaling in general were collected based on recent reviews and a PubMed search with the following terms: “TLR4” AND “negative feedback” , “LPS” AND “negative feedback” , “NF-kappaB” AND “negative feedback” and “p38” AND “negative feedback” . PubMed IDs of all sources are listed for each gene in Supplemental Table S4 . For RT-qPCR , mRNA was reverse transcribed with random hexamer primers . Primer efficiencies were obtained from dilution curves , and a HBB2 ( rabbit β-globin ) in vitro transcript was used for normalization in each pool of polysome fractions . The following primers ( forward/reverse ) were used: HBB2 ( gaaggctcatggcaagaagg/atgatgagacagcacaataaccag ) , C3ar1 ( tctcagtgtgcttgactgagccat/agaccaagaatgaccatggaggca ) , Cpd ( tgacgtggaaggtggtatgcaaga/tcttgtcgaagctgagaagcaggt ) , Csf2 ( gcatgtagaggccatcaaaga/cgggtctgcacacatgtta ) , Cxcl2 ( aaagtttgccttgaccctgaagcc/tctttggttcttccgttgagggac ) , Icosl ( tgaacttacagaccacgcctgaca/tccatcacagcccataagcagaca ) , Ier3 ( gagcgggccgtggtgtc/cttggcaatgttgggttcctc ) , Il15ra ( agctggaacatccaccctgattga/tgtcactactgttggcactggact ) , Map3k8 ( aagaatggcgtgcaaactgatccc/aggacggcaccatataactcagca ) , Ncl ( agggggcagaaattgatggacgat/tgggttctggggcactttg ) , Nfkb1 ( atgatccctacggaactgggcaaa/tgggccatctgttgacagtggtat ) , Nfkbid ( atattcgtgaacataaaggcaaga/tcagtggcgttaggctctg ) , Nfkbiz ( caggtgaacaccacggatt/ctcacagctcccttctggat ) , Nr4a1 ( tgcacagcttgggtgttgatgttc/agcaatgcgattctgcagctcttc ) , Pif1 ( tgactcccgagtgctgcatttcta/aggtcagaggtttgggtccatgtt ) , Plk3 ( ggctggcagctcgattag/gttgggagtgccacagatg ) , Tk1 ( tctccacacatgatcggaacacca/cagcgctgccacaattactgtctt ) , Tnf ( tgcctatgtctcagcctcttc/gaggccatttgggaacttct ) , Zc3h12a ( tgtgcctatcacagaccagcacat/tgaagcggtcatcatagcacacca ) , Zfand2a ( tcaccctgggaggaacagaaacaa/ctgtgctgaatgcagaagttgcca ) and Zfp36 ( tctcttcaccaaggccattc/atcgactggaggctctcg ) . For quantification of RNA by RNASeq , RNA was purified with the EURx GeneMATRIX universal RNA purification kit , including a DNase on-column digestion . RNA libraries were prepared for sequencing using the NEBNext Ultra Directional RNA Library Prep Kit after ribosomal RNA was removed with the Ribo-Zero Magnetic Kit ( Epicentre ) . Library preparation and sequencing was performed by the CellNetworks Deep Sequencing Core Facility at the University of Heidelberg . As spike-in controls , in vitro transcripts ( rabbit HBB2 and firefly luciferase ) were added at a concentration of 0 . 4 fmol per 1 µg RNA . Reads were mapped to the mouse RefSeq transcriptome as downloaded from the UCSC Genome Browser mm10 refGene table on February 5 , 2013 . The sequences of the in vitro transcripts were included in the index . For mapping , Bowtie [46] was used allowing a maximum of two mismatches and reporting all alignments in the best stratum ( settings: -a –best –stratum –v 2 ) . With an in-house developed Perl script , read counts were summarized at the gene level discarding all reads that map to transcript isoforms of more than one gene ( as defined by a common gene symbol ) . To calculate fold changes relative to the control condition , library size factors were estimated with the DESeq package [47] . Expression patterns were obtained as follows: A maximum was defined as a time point with a significant increase compared to the last significant minimum ( or the control ) . Unless the maximum is the last time point , it has to be followed by time points with a smaller or not significant fold change compared to the last significant minimum ( or the control ) , until the end of the time course or the next significant minimum is reached . A minimum was defined in an analogous way . Significance was defined as a log2-transformed fold change of >0 . 5 for maxima or <0 . 5 for minima and a p-value of <0 . 05 ( see Statistical Procedures ) . The group g0 contains all mRNAs without any significant changes compared to the control . G1 mRNAs have the first maximum at or after 1 h of stimulation , and no minimum before the first maximum . G2 mRNAs have the first minimum at or after 1 h of stimulation , and no maximum before the first minimum . G3 mRNAs have the first maximum before 1 h of stimulation and the first minimum at or after 1 h of stimulation . G4 mRNAs have their first minimum before 1 h of stimulation and the first maximum at or after 1 h of stimulation . Rpkm values were calculated with the following equation: The number of 58-mers ( the read length ) that are unique to the transcript isoforms of one gene was obtained with an in-house developed Perl script . Read counts and the number of unique 58-mers are represented in Dataset S2 . For plasmid MXp-GFβ ( p3113 ) , the MMLV promoter was amplified by PCR from MXh-GFP-control [48] with primers G2670/G2671 and ligated as a BamHI/KpnI fragment into the BglII/KpnI sites of pCI-puro [49] . In a second step , a GFP/β-globin fusion reporter gene was amplified with primers G2677/G2678 from pcDNA3-GFβ ( p2732 ) [9] and ligated into the MluI/NotI sites of the first cloning product , thereby introducing an XhoI site between the MMLV promoter and GFP . For MXp-FLB ( p3249 , construct a in Figure 8B ) , GFP/β-globin was replaced with a Firefly Luciferase/β-globin fusion reporter gene , which was PCR amplified from pFLB ( p2524 ) [50] using primers G3093/G3094 and cloned into the XhoI/EcoRI sites of MXp-GFβ ( p3113 ) . MXp-FLB-Ier3-3′UTR ( p3324 , construct b in Figure 8B ) contains ClaI/SpeI sites that had been introduced into the BglII site at the beginning of the β-globin 3′UTR by annealing oligos G2432/G2433 . These ClaI/SpeI sites were used to insert the murine Ier3 3′UTR after amplification by PCR with primers G2632/G2633 , placing it between the β-globin ORF and the β-globin 3′UTR . To obtain MXp-FLB-Ier3-3′UTR-ΔARE ( p3325 ) lacking the ARE ( construct c in Figure 8B ) , the regions upstream and downstream of the ARE were amplified with primers G2632/G2680 and G2633/G2679 , respectively . These two PCR products were annealed , amplified with primers G2632/G2633 and ligated into the ClaI/SpeI sites of MXp-FLB-Ier3-3′UTR ( p3324 ) . The following primers were used: G2670 ( agtggatcccatatgggcccttccgtttc ) , G2671 ( actcatcgattaatgcgaagagccgacgcagtctatc ) , G2677 ( atagcggccgcgccgccatgctcgaggtgagcaagggcga ) , G2678 ( aatacgcgttccttccgagtgagag ) , G3093 ( attactcgaggaagacgccaaaaacataaagaaag ) , G3094 ( ctgaggagtgaattctttgcca ) , G2432 ( gatcctgactcatgctcagtgacgtatcgatgactactagtgtca ) , G2433 ( gatctgacactagtagtcatcgatacgtcactgagcatgagtcag ) , G2632 ( tgatcgataacgcgatgggtca ) , G2633 ( gacactagtgacaggcaaatcaa ) , G2680 ( accgaccgacacggagaaagtct ) , G2679 ( tccgtgtcgggtcggtaagacag ) . HEK-Blue mTLR4 cells were transiently co-transfected with the MXp-FLB Firefly Luciferase Reporters and a Renilla Luciferase expressing plasmid ( pCIneo-RL , p2443 , [51] ) , and split into two wells 24 h after transfection . The cells in one well were treated with LPS ( 100 ng/ml , 1 h ) 48 h after transfection . Cells were lysed in 300 µl of passive lysis buffer ( dual-luciferase reporter assay system; Promega ) per well of a 6-well dish and incubated at room temperature for 20 min . Nuclei were removed by centrifugation for 1 min at 17 000× g . 20 µl of the supernatant was mixed with 50 µl of substrate from the dual-luciferase reporter assay system , diluted 1∶3 . Firefly and Renilla luciferase activities were measured on a Fluostar Optima ( BMG Labtech ) plate reader . In parallel , RNA was extracted and subjected to Northern blot analysis to determine the FLB reporter mRNA levels . Translation efficiency was calculated by first dividing the Firefly by the Renilla luciferase activity . This value was then normalized to the relative FLB reporter mRNA level . Supernatant of stimulated BMDM was collected and stored at −80°C . Cytokine concentrations were measured with the Basic Kit mouse FlowCytomix ( eBioscience ) in conjunction with the mouse IL1A ( BMS8611FF ) , IL1B ( BMS8602FF ) , CCL4 ( BMS86014FF ) and the TNF ( BMS860712FF ) FlowCytomix Kits . For detection of proteins by Western blotting , cells were lysed in sample buffer ( 50 mM HEPES pH 7 . 4 , 2% SDS , 10% Glycerol , 100 mM DTT ) . After separation on a SDS/5–20% polyacrylamide gradient gel and transfer to a 0 . 2 µm pore size nitrocellulose membrane ( Peqlab ) , membranes were blocked in PBS containing 0 . 1% sodium azide and 5% bovine serum albumin ( for IER3 ) or milk powder . Proteins were detected with the following antibodies diluted in PBS: ZFP36 ( Carp3 , abcam , ab36558-200 ) , IER3 ( Santa Cruz , sc-8454 ) and EIF3B ( Santa Cruz , sc-16377 ) , as well as HRP-coupled anti-goat ( Santa Cruz , sc-2020 ) or anti-rabbit ( Jackson ImmunoResearch , 711-036-152 ) antibody . Between the antibody incubation steps , membranes were washed in 150 mM NaCl , 50 mM Tris-HCl ( pH 7 . 5 at 25°C ) , 1% Tween-20 . As a luminol reagent , Western Lightning Plus-ECL Enhanced Luminol Reagent ( Perkin Elmer ) was used . RNA was resolved by 1 . 2% agarose-2% formaldehyde-MOPS ( morpholinepropanesulfonic acid ) gel electrophoresis and blotted overnight with 8× SSC ( 1× SSC is 0 . 15 M NaCl , 0 . 015 M sodium citrate ) onto Hybond-N+ Nylon membranes ( Amersham , GE ) . Membranes were hybridized overnight with digoxigenin-labeled RNA probes at 55°C and washed twice with 2× SSC/0 . 1% SDS for 5 min and twice with 0 . 5× SSC/0 . 1% SDS for 20 min at 65°C . Alkaline phosphatase-labeled anti-DIG Fab fragments and CDP-Star substrate ( both Roche ) were used for detection according to the manufacturer's instructions . Templates for in vitro transcription of RNA probes with SP6 polymerase were obtained by PCR from cDNA of cells expressing rabbit HBB2 ( β-globin ) , human Ncl and human Rps7 with the following primer pairs: HBB2 ( gtgcatctgtccagtg/gccgatttaggtgacactatagaataccctgaagttctc ) , Ncl ( ttacaaagtcactcaggatg/gccgatttaggtgacactatagaatacttagcgtcttcg ) , Rps7 ( ggtggtcggaaagctatc/gccgatttaggtgacactatagaatactatagacaccag ) . Following 10 days of differentiation , BMDM were detached with non-enzymatic cell dissociation solution ( SIGMA ) and seeded at a density of 1 . 6×105 cells per 12-well . After stimulation with 100 ng/ml LPS ( E . coli O111:B4 , Sigma L2630 ) , cells were again detached with non-enzymatic cell dissociation solution and pelleted by centrifugation ( 150× g , 3 min ) . Cells were washed once in cold flow cytometry buffer ( PBS with 0 . 2% FBS and 0 . 5 mM EDTA ) and stained in 100 µl Annexin V Binding Buffer with 5 µl Alexa Fluor 647 Annexin V ( BioLegend ) and 5 µg/ml propdidium iodide for 15 min at room temperature . Flow cytometry measurements were performed with a BD FACSCanto II flow cytometer of the Flow Cytometry Core Facility at ZMBH , Heidelberg . For most statistical methods , R was used . Pearson's product moment correlation coefficients and Spearman's rank correlation coefficients were calculated with the R function cor . test ( ) . Wilcoxon rank sum tests were performed with the function wilcox . test ( ) . The hypergeometric p-value for enrichment was determined with phyper ( ) . To test for differences in cell death or normalized Luciferase activity , two-sided unpaired t-tests were performed assuming equal variance . Differences in read counts of RNASeq samples were tested with the Bioconductor package DESeq [47] . DESeq estimates the dispersion that exists between biological replicates in addition to the sampling error and uses the negative binomial distribution to account for the additional variance . With the DESeq function estimateDispersions ( X , method = “blind” , sharingMode = “fit-only” ) , we estimated the dispersion by treating the eight different samples of the LPS time course like biological replicates , assuming that the majority of genes does not change in expression . Hence , variance is rather over- than underestimated . | When macrophages encounter pathogens , they initiate inflammation by secreting pro-inflammatory factors such as the cytokine TNF . Because a prolonged or overshooting release of these factors is harmful for the organism , their production needs to be tightly controlled and shut off in due time . To ensure a rapid but transient inflammatory response , gene expression is regulated at multiple levels , including transcription , stability and translation of mRNAs . While control of transcription and mRNA stability has been studied extensively , little is known about translational regulation in macrophages . In this study , we measured the translation of all mRNAs expressed in mouse macrophages . Upon activation of macrophages with the bacterial cell wall component lipopolysaccharide , we found that many feedback inhibitors , which are important for dampening the inflammatory response , are translationally up-regulated . Translation of these mRNAs is repressed in resting cells and de-repressed after stimulation . In contrast to feedback inhibitors , most cytokines are primarily regulated by changes in mRNA abundance . Furthermore , we could show that one of the feedback inhibitors , IER3 , protects macrophages from cell death during activation . Therefore , regulation at the level of translation is important for the induction of negative feedback loops and cellular survival . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology",
"protein",
"translation",
"gene",
"expression",
"genetics",
"biology",
"and",
"life",
"sciences",
"immunology",
"molecular",
"cell",
"biology",
"immune",
"response"
] | 2014 | Translational Regulation of Specific mRNAs Controls Feedback Inhibition and Survival during Macrophage Activation |
Chagas disease is a trypanosomiasis whose agent is the protozoan parasite Trypanosoma cruzi , which is transmitted to humans by hematophagous bugs known as triatomines . Even though insecticide treatments allow effective control of these bugs in most Latin American countries where Chagas disease is endemic , the disease still affects a large proportion of the population of South America . The features of the disease in humans have been extensively studied , and the genome of the parasite has been sequenced , but no effective drug is yet available to treat Chagas disease . The digestive tract of the insect vectors in which T . cruzi develops has been much less well investigated than blood from its human hosts and constitutes a dynamic environment with very different conditions . Thus , we investigated the composition of the predominant bacterial species of the microbiota in insect vectors from Rhodnius , Triatoma , Panstrongylus and Dipetalogaster genera . Microbiota of triatomine guts were investigated using cultivation-independent methods , i . e . , phylogenetic analysis of 16s rDNA using denaturing gradient gel electrophoresis ( DGGE ) and cloned-based sequencing . The Chao index showed that the diversity of bacterial species in triatomine guts is low , comprising fewer than 20 predominant species , and that these species vary between insect species . The analyses showed that Serratia predominates in Rhodnius , Arsenophonus predominates in Triatoma and Panstrongylus , while Candidatus Rohrkolberia predominates in Dipetalogaster . The microbiota of triatomine guts represents one of the factors that may interfere with T . cruzi transmission and virulence in humans . The knowledge of its composition according to insect species is important for designing measures of biological control for T . cruzi . We found that the predominant species of the bacterial microbiota in triatomines form a group of low complexity whose structure differs according to the vector genus .
Chagas disease [1] , which was first described by Chagas [2] as the American human trypanosomiasis , is a potentially life-threatening illness caused by the protozoan parasite Trypanosoma cruzi , which is transmitted via obligate hematophagous insect vectors classified within the family Reduviidae , subfamily Triatominae , commonly known as kissing bugs . Chagas disease is a tropical endemic disease found over large areas of South and Central America and has been ranked as one of the most important diseases in Latin America in terms of social and economic impacts [3] , [4] . According to the WHO statistics for 2010 , an estimated 10 million people are infected with T . cruzi worldwide , mostly in Latin America , with 30% of chronically infected individuals exhibiting cardiac alterations and 10% showing digestive , neurological or mixed alterations . More than 25 million people are at risk of the disease . It is estimated that in 2008 , Chagas disease killed more than 10 , 000 people ( http://www . who . int/mediacentre/factsheets/fs340/en/index . html ) . In Colombia , the annual treatment costs for chronic Chagas disease patients vary from $46 for a patient with cardiomyopathy without congestive heart failure treated in a basic care facility to approximately $7 , 900 for a patient with congestive heart failure requiring a specialized level of care . In Mexico , the cost per admitted patient varies from $4 , 463 to $11 , 839 , while it has been reported to be $3 , 864 on the average in Brazil . Healthcare costs must be considered together with prevention costs , i . e . , a minimum of $30/house . These costs multiplied by the infected population provide an idea of the total costs of Chagas disease for Latin American economies [5] . Despite the continuous pressure for vaccine or drug development , no suitable solution for addressing this situation has been developed thus far . The best strategy for combating Chagas disease is still controlling the triatomine population via insecticide application . According to their association with humans , it is common to characterize triatomines as domestic , peridomestic or sylvatic . Domestic and peridomestic triatomines create the greatest public concern due to their impact on human populations . The species comprising these groups depend on latitude [6] . Similarly , the type of trypanosome transmitted varies according to geographic localities and the species of insect vectors . Because of co-evolutionary processes , it is often observed that local vector species show a higher rate of infection with local T . cruzi strains [7] . Many factors can affect T . cruzi gut colonization , including the intestinal microbiota [8] , [9] . Thus , the vector may act as a biological filter for the parasites [9] , [10] . Based on GMO technology and the coprophagous habits of triatomines , it has been proposed that triatomine resistance of T . cruzi gut infection could be increased by natural autoinoculation with paratransgenic symbionts [11] . Different bacterial groups have been found in triatomine guts ( see [10] ) . However , the spectrum of bacterial species identified in laboratory cultures does not necessarily reflect the relative frefquency of these species under natural conditions [10] . Identification of the main bacterial groups in triatomine guts is important , as it may influence the selective pressures on T . cruzi [12] . Cultivation-independent methods , such as PCR-DGGE , full-length 16S rDNA sequencing and other molecular approaches , have been widely applied for describing insect gut microbiota and have revealed substantial bacterial diversity as well as groups of uncultivable microbes [13] . These methods offer the advantage of being performed independent of culture medium and providing a quantitative picture of the dominant microbiota present . In this report , we applied PCR-DGGE and library sequencing approaches to assess the diversity of bacterial communities in the guts of triatomine specimens from insectary colonies and from the field based on 16S rDNA analysis . We show that the microbiota of the triatomine guts are predominately composed of a few bacterial species that tend to be specific to the vector species; i . e . , Arsenophonus was preferentially associated with vectors of the Panstrongylus and Triatoma genera , while Serratia and Candidatus Rohrkolberia were typical of Rhodnius and Dipetalogaster .
A total of 54 triatomines in the 5th instar larval stage including both males and females belonging to different genera ( Dipetalogaster maximus , Panstrongylus megistus , Triatoma infestans , Triatoma vitticeps , Rhodnius prolixus and Rhodnius neglectus ) were obtained from insectary colonies maintained over approximately 20 generations at the Laboratório de Doenças Parasitárias ( Fiocruz , IOC ) using chicken as a blood source . Additionally , nine individuals of Rhodnius prolixus in the 5th instar larval stage fed with rabbit blood were obtained from another insect collection maintained as described by Garcia and Azambuja [14] , and seven sylvatic 5th instar larvae or adults of Rhodnius sp . were directly captured from palm trees ( Attalea maripa ) at Oriximiná , PA , Brazil ( Amazon region ) as described by Abad-Franch et al . [15] and kept isolated from the other specimens without receiving a blood meal until dissection . Insects that were separated from colonies were dissected 7–10 days after feeding . Dissection of the insects was performed using two fine forceps to open the dorsal side of specimens from the posterior end of the abdomen toward the last thoracic segment . Meticulous dissection of the whole gut was performed using a sterile ultrafine insulin syringe needle . Feces were obtained by abdominal compression or spontaneous dejections immediately after feeding . Guts and feces were collected in sterile Eppendorf tubes and maintained at −20°C until use . All steps were performed under aseptic conditions . DNA was extracted from feces and gut samples using the Fast DNA Spin Kit for soil ( Qbiogene , BIO 101 Systems , CA , USA ) according to the manufacturer's protocol . DNA concentrations were determined using a NanoDrop spectrophotometer ( Thermo Fisher Scientific Inc . ) . The DNA extracts were visualized on 0 . 8% ( w/v ) agarose gels to assess their integrity and purity . Fragments of 16S rDNA ( corresponding to the V6–V8 region of the E . coli 16S rDNA gene ) were amplified via PCR using the primers 968F-GC ( 5′-CGC CCG CCG CGC CCC GCG CCC GTC CCG CCG CCC CCG CCC G AAC GCG AAG AAC CTT AC-3′ ) and 1401R ( 5′-CGG TGT GTA CAA GAC CC-3′ ) as described by Nübel et al . [16] . The 50 µl reaction mix contained 1 µl of template DNA ( corresponding to approximately 50 ng ) , 10 mM Tris-HCl ( pH 8 . 3 ) , 10 mM KCl , 2 . 5 mM MgCl2 , 0 . 2 µM of each dNTPs , 1 . 25 U of Taq DNA polymerase ( Promega , Madison , WI , U . S . A . ) , and 0 . 2 µM of each primer . The amplification conditions were 1×2 min at 94°C followed by 35×1 min at 94°C , 1 . 5 min at 48°C , and 1 . 5 min , 72°C , and a final 10 min extension at 72°C . Negative controls ( without DNA ) were included in all amplifications . The PCR products were analyzed by agarose gel electrophoresis ( 1 . 4% gel ) and ethidium bromide staining [17] . Amplicons were stored at −20°C until DGGE analysis . DGGE was carried out as described by da Mota et al . [18] using a Bio-Rad DCode Universal Mutation Detection System ( Bio-Rad Laboratories , Munich , Germany ) . PCR products ( 15 µl ) were applied onto 6% ( w/v ) polyacrylamide gels in 1× TAE buffer ( 40 mM Tris-acetate [pH 7 . 4] , 20 mM sodium acetate , 1 mM disodium EDTA ) containing a denaturing gradient of urea and formamide varying from 45% to 65% . The gels were run for 15 h at 60°C at 100 V . After electrophoresis , the gels were stained for 30 min with SYBR Green I ( Invitrogen - Molecular Probes , SP , Brazil ) and photographed under UV light by using a Typhoon Trio apparatus ( Amersham Pharmacia Biotech ) . SYBR Green-stained bands were retrieved by excision from the DGGE gels under UV illumination and eluted in water for sequencing . The procedure described by Massol-Deya et al . [19] was used to amplify ∼1 . 5 Kb fragments of the 16S rDNA gene of each insect specimen via PCR with the universal primer pair pAf ( 5′-AGA GTT TGA TCC TGG CTC AG-3′ ) and pHr ( 5′-AAG GAG GTG ATC CAG CCG CA-3′ ) . The amplification conditions were as follows: 35 cycles of 92°C for 1 min 10 s , 48°C for 30 s and 72°C for 2 min 10 s . A hot start ( 2 min 10 s at 92°C ) was applied to avoid initial mispriming and enhance the specificity of the amplifications . A final extension step was run for 6 min 10 s at 72°C , and the reaction tubes were then cooled to 4°C . A negative control ( without DNA ) was included in all amplifications . DNA preparation and PCR products were visualized after electrophoresis as described above . Following PCR amplification or band purification with a QIAquick Gel Extraction Kit ( Qiagen Inc . ) , DNA fragments were cloned into the pJET2 . 1/blunt vector using the CloneJET PCR Cloning Kit according to the instructions of the manufacturer ( Fermentas ) . After transformation of competent E . coli DH5α cells , clones were picked , and the presence of inserts of the correct size was assessed via PCR using forward ( 5′-CGACTCACTATAGGGAGAGCGGC-3′ ) and reverse ( 5′-AAGAACATCGATTTTCCATGGCAG-3′ ) pJET1 . 2 primers . The clones were sequenced using the same primers in an ABI Prism 3730 automatic sequencer ( Applied Biosystems , Foster City , CA , USA ) . The 101 sequences obtained were deposited in GenBank under the accession numbers JQ410794–JQ410894 . Base calling and low quality sequence trimming from electropherogram files were performed with Phred [20] . The taxonomic position of bacterial genera corresponding to 35 DGGE bands was assessed by comparing their sequences to sequences in GenBank ( Rel . 184 , June 2011 ) according to the best BLASTn hits ( http://blast . ncbi . nlm . nih . gov/ , accessed by 2011-06-08 ) . We first removed plasmid vector sequences with cross-matching and then discarded putative chimeras using the Mallard program [21] . Full-length 16S rDNA sequences were obtained by assembling valid insert sequences with CAP3 [22] . The full-length 16S rDNA sequences and their homologous pairs from GenBank corresponding to the best BLASTn hits were aligned using MUSCLE v3 . 7 [23] . We constructed maximum likelihood phylogenetic trees with 1 , 000 bootstrap replicates using the generalized time-reversible ( GTR ) model [24] . The GTR model is the most general , as it is a neutral , independent , finite-site and time-reversible model . This process allowed accurate inference of the phylogenetic relationships of the full 16S rDNA sequences with their closest relatives of known taxonomic positions . Sequences aligned with MUSCLE v3 . 7 were formatted according to PHYLIP and used to construct distance matrices for each library with DNADIST ( provided in the PHYLIP 3 . 6 package , [25] ) using the default parameters and Jukes-Cantor as the substitution model . The distance matrices were used as input files for MOTHUR v1 . 14 [26] to define operational taxonomic units ( OTU ) on the basis of a similarity distance cutoff of 0 . 03 ( OTU0 . 03 ) . Sequences belonging to the same cluster based on reference to OTU0 . 03 were circumscribed with ellipses in the phylogenic trees that we identified using Greek symbols for the purpose of clarity . Although this cutoff distance can be seen as arbitrary , it is often helpful to think of OTUs defined by distances of 0 . 03 as corresponding to a species [27] . Then , we calculated the Chao1 index [28] , which measures the absolute value of species richness . To estimate the relationship between the expected OTU richness and the sampling depth , we used rarefaction curve methodology [29] , [30] . Good's coverage estimator [31] was used to calculate the sample representativeness with the formula C = 1− ( ni/N ) ×100 [28] , where N is the total number of clones analyzed , and ni is the number of clones that occurred only once among the total number of clones analyzed using OTU0 . 03 [32]
DGGE is a simple method that is well suited to characterize the global complexity of bacterial communities , such as found in triatomine guts from insectary colonies or field specimens . When analyzing the electrophoresis migration patterns of 16S amplicons from guts of Dipetalogaster maximus ( Fig . 1 , lanes A–C ) , Panstrongylus megistus ( Fig . 1 , lanes D–F ) , Triatoma infestans ( Fig . 1 , lanes G–I ) , Triatoma vitticeps ( Fig . 1 , lanes J–L ) and Rhodnius neglectus ( Fig . 1 , lanes M–O ) using this system , we observed band patterns that are characteristic of these species and were conserved among triplicate specimens of the same vector species . Further characterization of these bands by DNA sequencing revealed that their corresponding bacterial species were essentially members of the Enterobacteriaceae , particularly of the three genera Candidatus Rohrkolberia ( bands 1–3 ) , Arsenophonus ( bands 4–22 ) and Serratia ( bands 23–29 ) . Some bands ( 23–25 ) related to Serratia were shared by different insect genera , i . e . , Triatoma and Rhodnius . Additionally , Arsenophonus sequences were found in Panstrongylus ( bands 4–18 ) and Triatoma ( bands 19–22 ) , but not in Rhodnius or Dipetalogaster . The fingerprints of Triatoma infestans and Triatoma vitticeps were similar , suggesting that the main species in the bacterial communities in triatomine guts are specific to the genus of their hosts . When the DGGE fingerprints of wild specimens of Rhodnius sp . collected from the Amazon ( Fig . 2 , lanes D–L ) were compared to those of specimens of Rhodnius prolixus ( Fig . 2 , lanes A–B ) and Rhodnius neglectus ( Fig . 2 , lane C ) from insect collections grown under captivity and fed with rabbit or chicken blood , we actually observed similar profiles ( Fig . 1 , bands 28 and 29; Fig . 2 , bands 1 and 2 ) , although some specific bands ( Fig . 2 , bands 3 and 4 ) could be only observed in the sylvatic specimens . The profiles were similar whether guts ( Fig . 2 , lanes D–J ) or only feces ( Fig . 2 , lanes L–M ) were analyzed , showing that these bacteria ( Fig . 2 , bands 1 , 2 , 5 and 6 ) are most probably free in the Rhodnius gut lumen . The profiles were also similar in insects at the 5th instar larval ( Fig . 2 , lanes H–J ) and adult stages ( Fig . 2 , lanes D–G ) . The sequences of the specific bands could be associated to Candidatus Rohrkolberia cinguli ( GenBank: FR729479 . 1 ) or Erwinia chrysanthemi strain NZEC151 ( GenBank: EF530551 . 1 ) ( Fig . 2 , band 3 ) and Wolbachia ( Fig . 2 , band 4 ) , which is an endosymbiont that infects a wide range of insect hosts ( 20 to 75% of insect species , see refs in [33] ) , such as Microcerotermes sp . ( GenBank: AJ292347 . 1 ) , Pseudolynchia canariensis ( GenBank: DQ115537 . 1 ) and Supella longipalpa ( GenBank: FJ152101 . 1 ) . The full-length 16S rDNA sequences ( 1 . 5 kb ) are larger than the partial sequences of 16S bands separated by DGGE ( approximately 430 pb ) . Therefore , the full-length sequences allow classification of bacteria at the species level rather than just the genus level . Thus , full-length 16S rDNA sequences obtained from libraries of 16S rDNA clones allowed the description of the bacterial community structure in the investigated specimens in more details . The simplest microbiota structure was observed in D . maximus ( Fig . 3 ) , which presented only one OTU0 . 03 cluster ( D1–25 ) , designated α , including all sequences obtained from its rDNA library . Sequences belonging to this α OTU0 . 03 cluster were also observed in the R . prolixus microbiota ( Fig . 4 ) . The sequences from the α cluster were grouped along the same phylogenetic branch and were closely related to Candidatus Rohrkolberia cinguli ( DQ418491 . 1 ) . One of the sylvatic specimens of Rhodnius sp . also showed a DGGE band ( Fig . 2 , lane D , band 3 ) related to Candidatus Rohrkolberia cinguli . Candidatus Rohrkolberia cinguli is a new genus and species name recently proposed for the newly characterized clade of obligate intracellular symbiotic bacteria found in the midgut epithelium of the bulrush bug Chilacis typhae [34] . In R . prolixus , we found two additional OTU0 . 03 clusters representing approximately 85% of the bacterial microbiota related to Serratia marcescens strains ( Fig . 4 ) . Cluster δ was the major cluster and represented approximately 73% of the sequences found in the R . prolixus library . The minor cluster ( R2 , R9 , R17 ) is most likely related to Serratia sp . Sequences similar to the δ OTU0 . 03 cluster were also observed in the T . infestans microbiota ( Fig . 5 ) . Among the triatomine vectors that we investigated here , T . infestans presented the most complex microbiota structure , with six OTUs0 . 03 clusters . In addition to the δ cluster , we found ( i ) an OTU0 . 03 represented by only one sequence ( T14 ) closely related to the δ cluster , ( ii ) an additional uncharacterized singleton OTU0 . 03 represented by T15 , ( iii ) an OTU0 . 03 including two sequences ( T17 , T19 ) related to Serratia rubidaea and ( iv ) two clusters ( β and γ ) associated with the Arsenophonus endosymbionts , representing 62% of the whole set of sequences . Sequences similar to clusters β ( P1–8 , P10–18 , P20–25 ) and γ ( P9 , P19 ) from T . infestans were also observed in P . megistus ( Fig . 6 ) , where they represented 92% of the whole sequence set . To better characterize the relative bacterial species richness among the triatomines , we used the Chao index . The Ti library exhibited the highest species richness , with 6 . 5 expected OTUs0 . 03 ( confidence interval: 6 . 03–14 . 30 ) . The R . prolixus library showed 3 OTUs0 . 03 , while the P . megistus library showed 2 OTUs0 . 03 , and D . maximus showed 1 OTU0 . 03 according to the Chao index . There was no confidence interval associated with R . prolixus , P . megistus and D . maximus because there are no unseen species expected in these triatomine species given the rarefaction curve saturation ( Fig . 7 ) . The trend of the rarefaction curves suggests that the bacterial composition spectrum observed in the T . infestans library could indeed be slightly larger than reported here ( Fig . 7 ) . The Good's coverage estimator calculated from the 16S rDNA sequences of R . prolixus , P . megistus and D . maximus libraries was 100% , demonstrating the representativeness of our results compared to real conditions . In comparison , the same index showed that our sequencing coverage of the T . infestans library was 91 . 6% .
A low level of microbiota complexity seems to be frequent in insect guts [34]–[38] , except in particular cases such as termites [39] . This study shows that the triatomine gut microbiota , as revealed by DGGE , are composed of a few predominant bacterial species that ultimately differ according to the insect vector . Interestingly , the banding patterns were conserved among specimens of the same species , suggesting that the bacterial communities in triatomines are well adapted to their hosts . In addition , we found the same spectrum of bacterial species when using different primers pairs for DGGE and full-length 16S rDNA sequencing , which reduces the probability of a bias at the level of PCR amplification . Blood is of course sterile , and the sucking mouth parts of triatomines are adapted to blood consumption directly from vertebrate host capillaries , minimizing contamination by skin penetration . However , triatomines have multiple opportunities to inoculate themselves with bacteria , as they are coprophagous [40] , [41] . The habit of feces consumption by triatomines may explain the similarity between the DGGE fingerprints of the gut microbiota of individual insects within the same genus . It also suggests that the bacterial communities found within these vector species are rather stable and well adapted to their environment . Given the potentially large spectrum of bacterial species in triatomine feces due to environmental contamination , it must be concluded that the low number of prevalent bacterial populations in triatomine guts despite their coprophagous behavior is due to regulation by the host vector [42] . Several humoral [38] , [42] , [43] and cellular mechanisms involved in vector defenses have in fact been described . These mechanisms mainly include ( i ) antimicrobial peptides that could restrict the bacterial diversity in the gut , such as prolixicin produced by Rhodnius prolixus [44] , ( ii ) lysozyme activity , ( iii ) prophenoloxidase activation , ( iv ) phagocytosis and hemocyte microaggregation , ( v ) nitric oxide and superoxide production and ( vi ) trypanolytic proteins ( see refs in [10] ) . Moreover , the microbiota also participates in other basic functions , such as digestion and vitamin production [45] . It is the specific biochemical balance between host and microbiota factors that determines the features of the environment where a parasite has to adapt , and these features are expected to interact with its virulence [10] . The detection limit of bacterial DNA among eukaryotic DNA via PCR under normal conditions has been reported to be between approximately 4×102 and 4×103 fragments [37] . As a consequence , we believe that our PCR results simply reflect the most frequent bacterial species , which does not exclude potential presence of other species in minute amounts . In addition , some weak bands from the DGGE gels were not characterized , and the possibility cannot be excluded that some taxa escaped our analyses . However , we also showed that more than 90% of the information related to the gut microbiota of the triatomine specimens in our investigation has been considered here via analysis of rarefaction curves obtained for 16S rDNA libraries . One advantage of DGGE analysis is the possibility of extracting bands from the gels followed by performing sequence analysis of the purified amplicons and identifying community members belonging to different phyla , such as Actinobacteria , Proteobacteria , Firmicutes , and Deferribacteres , by comparison to reference databases [18] , [37] , [46] . Investigation of the gut microbiota of different triatomine species using bacterial culture methods revealed a bacterial community of limited diversity characterized by several species of Enterobacteriaceae [10] , [40] , with some of them being eventually pathogenic to humans . In our analyses , we mainly detected members of the Enterobacteriaceae ( Serratia , Candidatus Rohrkolberia/Pectobacterium and Arsenophonus ) , which suggests that this group is predominant in the guts of triatomines . Enterobacteriaceae appear to be frequent in insects , particularly in insect vectors whose diets are limited to a few food sources [35] , [45] , [47] . In their review of microbiota complexity , Vallejo et al . [10] reported 6 , 25 and 26 bacterial species in T . infestans , R . prolixus and P . megistus , respectively . However , these numbers were obtained by different authors across very different conditions and under in vitro culture . It is clear that this list does not show the relative abundance of these species in the triatomine gut . In contrast , the present study provides an indication of the predominant species of the bacterial microbiota across triatomine species in insectaries and field conditions , including obligatory intracellular symbionts that cannot be detected using traditional culture methods [34] . Many complex molecules produced by hosts or microbiota are present in the insect gut lumen , such as hydrolytic enzymes , peptides , vitamins , cofactors and antimicrobial factors . These molecules can stimulate some bacteria , but they can inhibit the growth of many other competing members in this environment . For these reasons and many others , it has been recognized that the in vitro conditions of bacteria culture on artificial media do not mimic those of the natural conditions in insect guts [13] . Diagnosis of the number of predominant bacterial species is a quantitative concept when it involves 16S rDNA , which deserves some comment . Because variation is continuous among sequences , criteria based on genetic distance must be applied to discriminate among species , which is the reason that we applied the cutoff of 0 . 03 to discriminate among OTUs . Although this cutoff distance can be seen as arbitrary , it is often helpful to think of OTUs that are defined by a distance of 0 . 03 as corresponding to a species , of 0 . 05 as corresponding to a genus , of 0 . 15 as corresponding to a class , and of 0 . 20 to 0 . 30 as corresponding to a phylum [27] . The concept of species among bacteria is a difficult issue . At first glance , the large sequence number per cluster may surprise , but deeper observation shows that the genetic diversity between these sequences is equivalent to that among strains of the same bacterial species , as found in GenBank . This is particularly obvious for Serratia in cluster δ . We found relatively good agreement between the description of microbiota complexity according to ( i ) OTU0 . 03 criteria [27] , ( ii ) branch numbers in phylogenetic trees , and ( iii ) the species richness index of Chao . The number of dominant bacterial OTUs0 . 03 clusters in T . infestans was ∼6 according to the Chao index , with an upper limit of approximately 14 according to the inference of unseen species . This index has the advantage of presenting clear and rigorous non-parametric statistics [28] . Based on this observation , we can conclude that the microbiota diversity according to the prevalent bacterial OTUs0 . 03 is at least two times greater in T . infestans compared to the other vector species , suggesting a different relationship between microbiota and vector factors with possible consequences for protozoan parasitism . A methodology similar to that used in this study , including DNA extraction , universal primers and PCR-DGGE , allowed successful recovery of 16S rDNA sequences related to Rhodococcus sp . ( GU585554 ) , Gordonia sp . ( GU585556 , GU585557 ) and other Actinomycetales from different environments [18] , [48] . Although sequences belonging to Rhodococcus were not recovered in the present study , we cannot exclude the association of weak DGGE bands with Rhodococcus [40] . In Rhodnius specimens , S . marcescens was the predominant species observed; this species is a free-living bacterium that produces a red pigment known as prodigiosine , which has recently received renewed attention due to its reported antibacterial , antifungal , antiprotozoan [49] , immunosuppressive and anticancer properties [50] . Moreover , Serratia marcescens has been reported to utilize the type VI secretion system ( T6SS ) to target bacterial competitors [51] . It was demonstrated that T6SS exhibits dramatic antibacterial killing activity against several other bacterial species [52] , [53] , favoring S . marcescens strains in Drosophila [51] . The S . marcescens found in Triatoma sp . and Rhodnius sp . guts could reduce the number and diversity of other extracellular bacteria in the gut lumen via the same process , as the antibacterial activity of S . marcescens T6SS appears to act through direct bacterium-to-bacterium contact [51] . In addition , S . marcescens can also lyse T . cruzi through the action of D-mannose fimbriae , which adheres to the parasite surface [54] . The genus Arsenophonus represents a group of endosymbiotic , mainly insect-associated bacteria with a broad spectrum of insect and even plant hosts [55] . Arsenophonus includes lineages with a rapidly increasing number of closely related symbionts reported from phylogenetically distant hosts [55] , [56] . A member of this genus , Candidatus Arsenophonus triatominarum , has been isolated from T . infestans [56] . This bacterium is an intracellular endosymbiont found in hemolymph , heart tissue , salivary glands , neural ganglia , visceral muscles , nephrocytes , ovaries , testes , and dorsal vessels that lives in the cytoplasm of host cells and displays pleiomorphy , with forms ranging from spherical to highly filamentous . In contrast to Arsenophonus nasoniae ( a symbiont of parasitoid wasps of the genus Nasonia ) , A . triatominarum does not grow on artificial culture media but does grow well on Aedes albopictus cell lines , which demonstrates that it must be considered to be a P-symbiont [55] , and the two species form a distinct lineage of bacteria within the family Enterobacteriaceae [56] . The draft sequence of the complete genome of A . nasoniae [57] , [58] shows that it carries putative hemolysins , alkaline metalloproteases , serralysin , ( an insecticidal toxin of Serratia ) and several other pseudogenized toxin genes that most likely indicate past parasitic activity typical of Enterobacteriaceae [58] . In addition , it also carries 8–10 copies of the rDNA operon [57] . Therefore , although we detected several Arsenophonus bands by DGGE in P . megistus , we attributed this to paralogy rather than orthology . Analysis of these sequences actually revealed a high level of similarity among them , and the Chao1 species richness estimator indicated the presence of only two OTUs in the library , which confirms paralogy , as discussed by Nováková et al . [55] . Moreover , the Arsenophonus OTUs from the P . megistus and T . infestans libraries showed a high level of similarity to the Arsenophonus accession in GenBank previously isolated from Triatoma melanosoma [59] . When Candidatus Rohrkolberia cinguli was reported for the first time in Chilacis typhae , a 1 . 5 kb segment of the eubacterial 16S rRNA gene was amplified by PCR from DNA samples from the midgut epithelium of the hemipteran species Chilacis typhae ( 52 individuals were used for PCR ) , then cloned and typed by RFLP . All RFLP types of 40 clones were identical . Furthermore , when a 1 . 65 kb segment of the gammaproteobacteria groEL gene was amplified , cloned and sequenced , the RFLP types and sequences of the clones were all the same [34] , indicating that the Rohrkolberia population is prevalent and well adapted to the host . Some sequences belonging to the same Rohrkolberia OTU cluster in D . maxima were also found in R . prolixus , demonstrating that Candidatus Rohrkolberia is not restricted to one insect genus . In fact , to our knowledge , this is the first time that Candidatus Rohrkolberia has been reported in triatomine guts , and this species , together with Arsenophonus and Wolbachia , the other symbionts found in this study , deserve more attention with respect to paratransgenic strategies . Interestingly , Wolbachia , which is a symbiont that can be transmitted together with Arsenophonus [57] by parasitoid wasps of the genus Nasonia [56] , has been found in a Rhodnius specimen from the Amazon . Wolbachia has been described in several organs and feces of Rhodnius pallescens by Espino et al . [33] . These symbionts belong to α-Proteobacteria and can cause postzygotic reproductive incompatibilities in insects , as they display a tropism for the reproductive tissues of their hosts and are transmitted vertically from insect to insect through ovules or horizontally through parasitoids . Despite the fact that infected insects do not show pathological signs , the presence of Wolbachia can result in diverse reproductive alterations in their hosts , including parthenogenesis , feminization , male killing and unidirectional or bidirectional cytoplasmic incompatibility . The relationship between Wolbachia and their arthropod hosts ranges from mutualistic to parasitic depending on the Wolbachia strain and arthropod species [33] . Introduction of a Wolbachia wMel strain isolated from Drosophila melanogaster to an adult vector of Aedes aegypti allowed successful suppression of dengue transmission in two natural populations of A . aegypti within only a few months [60]–[62] . The reduced number of sylvatic samples analyzed in the present study does not permit us to reach a definitive conclusion regarding variations in microbiota between insectary and sylvatic individuals of Rhodnius . However , some differences were observed between the prevalent populations of bacteria , including Wolbachia , although only seven individuals of sylvatic Rhodnius sp . were analyzed . If DGGE profiles show that the complexity of predominant bacterial species of the Rhodnius gut microbiota is apparently low , exhaustive cloning and sequencing will be necessary to reveal bacterial species with a relative minor number in the microbiota analyzed . The present pilot investigation should be used as a basis for future analyses with other specific focuses that would require a more refined approach , such as high-throughput sequencing . Rhodnius has long served as an important physiological laboratory model . Since Wigglesworth's pioneering work [63] on molting and reproduction , a large body of knowledge has been accumulated worldwide . Ultimately , it has been proposed that protozoan parasites inside the guts of vectors should be killed by infecting triatomines under natural conditions with paratransgenic bacteria that are able to produce antimicrobial factors [11] , [64] , [65] . However , this strategy , known as paratransgenesis [64] , should take into consideration potential interactions with the intestinal microbiota under natural conditions . As shown above , the composition of the gut microbiota varies according to the species of insect vector and includes Serratia , Arsenophonus , Rohrkolberia or Wolbachia populations , which could affect the success of paratransgenic approaches over time and , thus , long-term host protection . Possible degradation or bioaccumulation of paratransgenic factors by natural microbiota is another matter of concern . Finally , the long-term effect of paratransgenic factors on the mechanisms through which symbiotic microbes can influence the ability of their host to transmit pathogens may be questioned [38] , [42] , [43] . | Chagas disease is one of the most important endemic diseases of South and Central America . Its causative agent is the protozoan Trypanosoma cruzi , which is transmitted to humans by blood-feeding insects known as triatomine bugs . These vectors mainly belong to Rhodnius , Triatoma and Panstrongylus genera of Reduviidae . The bacterial communities in the guts of these vectors may have important effects on the biology of T . cruzi . For this reason , we analyzed the bacterial diversity hosted in the gut of different species of triatomines using cultivation-independent methods . Among Rhodnius sp . , we observed similar bacterial communities from specimens obtained from insectaries or sylvatic conditions . Endosymbionts of the Arsenophonus genus were preferentially associated with insects of the Panstrongylus and Triatoma genera , whereas the bacterial genus Serratia and Candidatus Rohrkolberia were typical of Rhodnius and Dipetalogaster , respectively . The diversity of the microbiota tended to be the largest in the Triatoma genus , with species of both Arsenophonus and Serratia being detected in T . infestans . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biology",
"microbiology"
] | 2012 | Cultivation-Independent Methods Reveal Differences among Bacterial Gut Microbiota in Triatomine Vectors of Chagas Disease |
Open source drug discovery offers potential for developing new and inexpensive drugs to combat diseases that disproportionally affect the poor . The concept borrows two principle aspects from open source computing ( i . e . , collaboration and open access ) and applies them to pharmaceutical innovation . By opening a project to external contributors , its research capacity may increase significantly . To date there are only a handful of open source R&D projects focusing on neglected diseases . We wanted to learn from these first movers , their successes and failures , in order to generate a better understanding of how a much-discussed theoretical concept works in practice and may be implemented . A descriptive case study was performed , evaluating two specific R&D projects focused on neglected diseases . CSIR Team India Consortium's Open Source Drug Discovery project ( CSIR OSDD ) and The Synaptic Leap's Schistosomiasis project ( TSLS ) . Data were gathered from four sources: interviews of participating members ( n = 14 ) , a survey of potential members ( n = 61 ) , an analysis of the websites and a literature review . Both cases have made significant achievements; however , they have done so in very different ways . CSIR OSDD encourages international collaboration , but its process facilitates contributions from mostly Indian researchers and students . Its processes are formal with each task being reviewed by a mentor ( almost always offline ) before a result is made public . TSLS , on the other hand , has attracted contributors internationally , albeit significantly fewer than CSIR OSDD . Both have obtained funding used to pay for access to facilities , physical resources and , at times , labor costs . TSLS releases its results into the public domain , whereas CSIR OSDD asserts ownership over its results . Technically TSLS is an open source project , whereas CSIR OSDD is a crowdsourced project . However , both have enabled high quality research at low cost . The critical success factors appear to be clearly defined entry points , transparency and funding to cover core material costs .
The vast majority of drug research and development ( R&D ) performed globally is directed towards the needs of high-income countries [1] . The former Global Forum for Health Research and the work that led to its establishment asserted that 90% of all health R&D investment is spent on areas that concern only 10% of the world's population [2]–[4] . High-income countries have the resources to pay , either publicly or privately , a price which gives the innovator a profitable return on investment . The problem , of course , is that the medical needs of high-income countries are not the same as low-income countries . There are a host of diseases that are primarily endemic to low-income countries , diseases like dengue fever , malaria and schistosomiasis . Incentivizing R&D investments by standard incentives like patents simply does not produce the greatly needed , new medicines or diagnostics for these diseases ( which are often labeled “neglected” ) . These are neglected because the market does not offer sufficient purchasing power . This market failure is an internationally recognized problem and has been a major focus of the World Health Organization ( WHO ) . In 2003 a Commission on Intellectual Property Rights , Innovation and Public Health was established under the auspices of WHO in order to apprise appropriate funding and incentive mechanisms for these neglected diseases . A number of initiatives have resulted from the Commission's recommendations including the formation of an expert working group to suggest and evaluate options to incentivize R&D for these diseases [5] . A large variety of financing and coordinating mechanisms have been proposed . One that has received some support is open source drug discovery . Open source drug discovery is a model based upon the open source movement within the computer software industry . Basically it takes two primary attributes , namely the collaboration of volunteers and free access to the results , and applies them to drug discovery . This should ultimately translate into new drugs entering the market at prices determined by generic competition . The concept has been discussed within the academic literature for almost a decade . One of the first proposals by Maurer , Rai and Sali [6] laid out the concept and applied it particularly to tropical diseases . Subsequently , there have been several high-level descriptions of example projects [7] , [8] and more recently , empirical examples [9] of models , methods , processes and tools . However , the literature has not united behind a single , comprehensive definition of the concept . This combined with the multitude of titles given to the concept ( e . g . precompetitive collaboration , data sharing , open access R&D , etc . ) makes a common understanding difficult . Luckily , the non-profit research organization , Results for Development Institute ( “R4D” ) , has recently undertaken a high-level review of open source drug discovery projects aimed at neglected diseases [10] . One of the results of this review is a straightforward definition . R4D defines a set of attributes that a project must comply with in order to be considered open source: If a project adheres to all three requirements , the resulting advantages should be: verified content , collaborative projects , the creation of a commons of knowledge and reduced costs for the project ( resulting in lower prices for the end product ) . In an open source project data is made publicly-available for anyone and everyone to verify . In drug discovery this means that all virtual and laboratory results are published with as much of the raw data available as possible . This should include enough data for someone knowledgeable in the topic to review and critique the data . Collaboration across organizational and geographical boundaries offers several benefits . If enough researchers can be incentivized to collaborate , even small contributions by many researchers can significantly progress a project . It also opens a project to new external ideas and approaches . It is anticipated that the majority of the researchers will contribute on a volunteer basis , thereby reducing the cost of the project . A commons of knowledge is knowledge that is owned by the public , meaning that there is no individual owner . All sciences contain vast commons of knowledge . For example , in mathematics , algebra , geometry and calculus are all a part of the commons of knowledge . No one owns them; they are public knowledge . These knowledge commons grow when researchers place their data in the public domain . This is most commonly done by publishing the data without first patenting it . Knowledge residing in the public domain may not be patented since novelty is required to patent . This means that anyone can use , distribute and further develop the research without paying a royalty to , or even notifying , the innovator . If all the data necessary to manufacture a new drug are placed in the public domain , anyone may undertake the necessary regulatory steps for approval and begin to manufacture the drug . In open source computing it is more common to utilize specialized licenses rather than the public domain since software code is most commonly protected by copyright which is awarded automatically . These licenses allow the innovator to maintain some level of control over the innovation , generally ensuring that attribution is given and that the code is freely accessible for anyone to redistribute and modify . Any license in compliance with the Open Source Definition [11] is considered open source . These same aims can also be achieved by pairing a patent with a standard license allowing free use of the patent so long as the use adheres to a set of conditions . Examples include instances where innovators allow patented medicines to be manufactured by producers in low-income countries for local use only ( i . e . equitable licenses ) . Project costs of open source projects are significantly reduced based upon the percentage of work performed by volunteers as well as the absence of the administrative costs that accompany contract creation and royalty payment . Since the research is placed in the public domain , the price of the manufactured product is essentially de-linked from the cost of the R&D . Manufacturers set a price point based solely upon their own costs and expectations of the market's willingness to pay . Ideally generic competition is introduced immediately . Three similar concepts ( open access , open innovation and crowdsourcing ) are often confused with open source . Open access means that anyone can view , copy or distribute some form of content ( e . g . an article , book , etc . ) free-of-charge; it does not permit changing the content [12] . Open innovation is simply the use of external sources of R&D [13] . This may include paying royalties to the innovator and does not necessitate any type of transparency or commons formation and is therefore not related to the general “open definition” . For example , AstraZeneca recently agreed that a certain set of external scientists could access all of the data related to approximately 20 experimental drugs that they have stopped researching . This data is not open to the public . These drugs are under patent , and AstraZeneca will commercially benefit if the scientists manage to determine a profitable use of these molecules [14] . Open innovation offers the potential benefits of collaborative projects and reduced costs of both the project and the end product , but does not offer verified content or the creation of a commons of knowledge . “Crowdsourcing” is the use of volunteers to perform a specified task , generally through an open call [15] . For example , the FoldIt game has players fold proteins into their most chemically stable configuration , introducing new possibilities to the scientists who gather and research the players' efforts [16] . The contributors do not own their output , and crowdsourced outputs may or may not be protected by intellectual property rights . Crowdsourcing offers the same benefits of open innovation - collaborative projects and reduced costs of both the project and the end result , but does not necessarily offer verified content or the creation of a commons of knowledge . Open source is an important model for neglected diseases R&D because it offers the opportunity to accelerate the discovery progress while keeping expenditures to a minimum . Patents in these instances are neither desired nor justifiable since the cost of patenting will likely exceed any potential profits . A current gap within the academic literature is detailed profiles and evaluations of ongoing open source initiatives for neglected-disease research . This is the objective of our case study – to learn from the first movers of open source drug discovery , their successes and failures , in order to generate a better understanding of how a much discussed theoretical concept actually works in practice . After a search for relevant cases , we have studied two cases in detail: The Council for Scientific and Industrial Research Team India Consortium's Open Source Drug Discovery project ( CSIR OSDD ) and The Synaptic Leap's Schistosomiasis project ( TSLS ) . The objective of the case study is to answer the following research questions: Our results demonstrate that open source drug discovery initiatives can make significant achievements . However , there is no one formula for success . Critical success factors are clearly defined entry points , transparency and funding to cover all material costs .
We chose open source drug discovery projects targeted towards neglected diseases that have had at least one year of continuous data from multiple individuals . We identified twelve potential cases of an open source approach to drug discovery , mainly through our ongoing research of the topic but also through other articles reviewing the topic [8]–[10] . The potential cases identified along with their conformance to the selection criteria are given in Table 1 . Two cases fit our selection requirements: The Council for Scientific and Industrial Research Team India Consortium's Open Source Drug Discovery project ( CSIR OSDD ) and The Synaptic Leap's Schistosomiasis Project ( TSLS ) . The other potential cases were excluded either because the project's collaboration efforts were not viewable ( meaning that data was shared but the process of producing the data was not shared or collaboratively performed ) or the project was inactive ( meaning that a small number of individuals would occasionally make a posting which was most often an interesting article about the topic ) . Data were gathered from four sources: an analysis of the cases' websites , interviews of participating members , a survey of potential members of CSIR OSDD and a literature review . Additionally the project managers of both cases were sent our findings , and their comments have been incorporated into this paper . All websites of the two projects have been reviewed focusing on aspects of collaboration and progress . The licenses have also been reviewed to understand how intellectual property is managed . Telephone and written interviews were performed from November 2010 to April 2011 . Interview content focused on collaboration , intellectual property and progress . An interview template was devised and reviewed by two external researchers familiar with open source drug discovery ( Annex S1 ) . We posted introductions to our case study on both the CSIR OSDD and TSLS websites , asking interested individuals to e-mail us if interested in participating . We also directly e-mailed participants where we could find contact information ( n = 99 ) . Fourteen ( 14 ) individuals responded , representing both project leaders and active members . Among the 14 , only ten completed all interview topics and this was disproportionately members of TSLS project ( n = 9 ) . The individual completing the interview from the CSIR OSDD project had observed the project but not contributed . However , four CSIR OSDD project members partially completed the interview . A survey ( Annex S2 ) of potential members of the CSIR OSDD project was performed in February and March 2011 . The CSIR OSDD project was selected because they are performing general tuberculosis drug discovery activities where as the TSLS project is performing a very specific development task in regards to making a new synthesis of a known molecule , making it more difficult to identify researchers with similar research interests . PubMed was searched on January 31 , 2011 for articles published within the last year containing the phrase “Mycobacterium tuberculosis genome” . A second search was performed on February 10 , 2011 for articles published within the last year containing the phrase “Tuberculosis drug discovery” . The searches resulted in 221 and 112 articles respectively . The corresponding author's e-mail address was retrieved from each of these articles and then duplicates were removed . Sixty-one individuals completed the survey ( n = 46 from the genome group and n = 15 from the drug discovery group ) . A literature review was performed to identify any academic articles relevant to our research questions . This was done by searching Google Scholar on December 6 , 2011 with the following strings , achieving the following results: These articles were read . We sought approval for our research portfolio ( including interviews and surveys ) from the Norwegian Committees for Medical and Health Research . The Committee decided that our research did not require their ethical approval since we are studying collaboration amongst scientists and not patients . With that said , all interview participants were informed orally that their interview responses would be treated confidentially and that their participation was completely voluntary . Written consent was deemed unnecessary since interview participants responded individually to a call for interviews from a website posting . The survey data were analyzed anonymously . The interview data were analyzed in combination with the scientists' postings on publicly-available websites .
The Council for Scientific and Industrial Research Team India Consortium's Open Source Drug Discovery project ( CSIR OSDD ) started in 2008 with an initial grant from the Government of India of approximately US $35 million ( of which US $12 million has been released to date ) . Their vision is “to provide affordable healthcare to the developing world by providing a global platform where the best minds can collaborate & collectively endeavor to solve the complex problems associated with discovering novel therapies for neglected tropical diseases like Malaria , Tuberculosis , Leshmaniasis , etc . ” Initially they have targeted tuberculosis as their primary research area ( see Table 2 ) . CSIR OSDD aims to discover novel therapies for tuberculosis . Its activities are spread throughout every stage of the discovery process ( from drug target identification to lead optimization ) . It has 54 molecules in process and has initiated discussions with pharmaceutical companies regarding pre-clinical and clinical trials . Its main achievements to date are: the re-annotation of the Mycobacterium tuberculosis genome and the generation of 11 models for prediction of anti-tuberculosis activity [18] . The genome of the Mycobacterium tuberculosis strain H37Rv was first published in 1998 [19] . Since publishing , new research has been performed in such areas as gene functionality , associated proteins , interactions and potential drug targets . Most of this research is available electronically but on many different websites . Data curation involves establishing and developing long-term repositories of reference data [20] . The CSIR OSDD project created a data repository for genome-level information regarding the strain H37Rv , by recruiting volunteers to gather relevant research articles , extract the data and transcribe it into a standardized format . The aggregation of this process is TBrowse , a publicly-available integrative genomics map , http://tbrowse . CSIR OSDD . net/ [21] . The formation of TBrowse demonstrated that students could successfully contribute to open source drug discovery . With this proof of concept performed , CSIR OSDD moved onto a more complex task called Connect to Decode , annotating the tuberculosis genome . Again students collated the data contained in published articles regarding the approximate 4 , 000 genes contained in the tuberculosis genome . For those genes whose function was unknown , participants computationally extrapolated the possible function ( s ) . This work created a metabolome ( a complete set of small molecules involved in growth , development and reproduction ) and protein-protein functional network for Mycobacterium tuberculosis that is being used to identify potential drug targets . This data is contained on website called Sysborg . Eleven groups have worked independently to develop models for prediction of anti-tuberculosis activity . Two of these models have been published [22] and the other nine are in the process of being written up . CSIR OSDD has purchased the virtual screening data of 20 , 000 molecules , where 140 of these molecules have shown promising anti-tubercular properties . CSIR OSDD has built a new repository [23] ( the OSDD Chemical Database ) to gather data on these and other promising molecules . As of February 22 , 2012 , 304 molecules reside in the virtual repository , submitted by 17 individuals . Four molecules have been screened against tuberculosis , 14 against malaria . To perform these accomplishments , CSIR OSDD has created a significant amount of infrastructure . They utilize several websites including: According to a description of the project [24] , the workflow follows a standard process comprised of the following steps: No content may be viewed on Sysborg without first logging on . When registering , the user must accept the terms and conditions of the CSIR OSDD license , a non-standard license written specifically for the project [44] . The license affirms that CSIR OSDD owns all content posted to Sysborg ( §3 . 1 ) . Therefore , content is not a part of the public domain . All improvements based upon data within Sysborg must be contributed back to CSIR OSDD under a worldwide royalty-free non-exclusive license ( §3 . 5–6 ) . There is no stipulation in the license that CSIR OSDD must adopt non-exclusive licensing of the resulting products or any stipulations regarding the final price of these products . However , the mission states clearly that they aim “to make available affordable medicines to every single person of the developing world . ” CSIR OSDD has mapped out a process for discovering and developing new tuberculosis medicines . They have 54 molecules in the pipeline , including two candidates in the hit to lead phase which are being optimized in collaboration with private partners ( which seem to follow the same overall process ) [24] . They have instigated talks with pharmaceutical industry to perform the preclinical and clinical trials . Their approach to clinical trials is to build facilities specifically for clinical trials within publicly-funded hospitals . These trials would be conducted by CSIR OSDD in combination with the hospital personnel and experts from private pharmaceutical companies . All data will be made available ( presumably anonymized ) [24] . We found no evidence of clinical trials on Sysborg so we presume that these are planning activities in anticipation of forthcoming trials . The government of India has committed to grant CSIR OSDD INR 1 . 5 billion ( or about US $35 million ) of which US $12 million has already been paid out [33] . These funds pay the administrative costs of the project including equipment and material costs at the partner institutions and the salaries of a few contributors . Most work is done by unpaid volunteers . However , the project does hire individuals at times to perform specific tasks . For example , 20 female scientists are planned ( or have been ) hired to work from their homes for four hours a day [30] . Expert mentors are paid to attend meetings [30] . Vacancies are regularly posted on the website for paid positions such as project assistants [45] . The Synaptic Leap website was launched in 2006 with an aim “to provide a network of online research communities that connect and enable open source biomedical research” [46] . It was launched with four pilot disease research areas: malaria , schistosomiasis , toxoplasma and tuberculosis . Each area had a project leader with the responsibility of gathering and motivating international researchers to contribute to the Synaptic Leap community by sharing results , giving feedback and possibly undertaking new research tasks . Since launch , the malaria , toxoplasma and tuberculosis communities have been relatively silent . However , the schistosomiasis community has consistently utilized the website to share findings , discuss research results and identify new , necessary research tasks ( see Table 3 ) . The aim of the TSLS project was a well-defined drug development task – to generate the off-patent schistosomiasis drug , praziquantel , as a single enantiomer . This would remove the bitter taste of the original drug making it more palatable for children as well as remove some of its side effects . This has been needed for years but companies would not invest , likely because the innovation was not suitably lucrative since an inexpensive drug already existed . Additionally the patent on praziquantel expired in the 1990s [47] , and the needed change was likely not sufficiently novel to warrant a new patent . The optimization of praziquantel had long been a high priority of WHO which was affirmed in TDR's ( Special Programme for Research & Training in Tropical Diseases ) Scientific Working Group on Schistosomiasis in 2005 and repeated in its Business Plan of 2008–2013 . [48] This led to the funding of the TSLS project in 2008 by both WHO and the Australian government . The TSLS project completed this task in 2011 . To perform these accomplishments , TSLS has made use of web tools that were already available such as The Synaptic Leap website and an open source online laboratory notebook [49] . The laboratory notebook was chosen because it allowed contributors to enter scientific data more easily than The Synaptic Leap website . Dr . Matthew Todd became the leader of the schistosomiasis project in 2006 . He was already working on the problem of the production of praziquantel as a single enantiomer but wanted the project to go faster than typical academic speed . He thought that open source might be a solution to attract industry participation . The project was first discussed on the TSLS website in January 2006 [50] . However , even though Todd regularly updated the website , there was little external interest shown in the project . From 2006 to 2008 there were 35 postings initiated on the website , with only four of these coming from individuals other than Todd . In 2008 the project received their funding ( although contracting delays resulted in the laboratory work actually not starting until January 2010 ) . This allowed the project to hire a full-time postdoctoral researcher and cover laboratory expenses for Ph . D . students , mentored by Todd at the University of Sydney . This gave the project some needed momentum . From project initiation in 2006 until project funding in the beginning of 2010 , 10% of new postings were initiated from external contributors ( those not a part of Todd's team at the University of Sydney ) . After the funding was received 30% of postings were made by external contributors . However , comments posted by external contributors did not vary significantly ( increasing only from 50% to 53% ) . At the time of funding , significant external marketing efforts were also undertaken ( see below ) . The data from the on-going experiments were regularly posted in the online publicly-available laboratory notebook [49] and summarized on The Synaptic Leap , without peer review . Todd did not want to slow the speed of sharing the data by implementing an offline peer review process . He expected project contributors to give the researchers feedback , and this turned out to be the case . Key findings have received as many as 14 comments; entries average 1 . 5 comments each , with 50% of all new postings receiving comments . This process has been an adjustment for some of the contributors . There were concerns that mistakes would be published with name attribution . One researcher stated that he used more time to check his results before publishing them online . Ultimately , Todd expected peer review to be done through publishing , and two articles summarizing the results of this project have been published in September and October 2011 ( one with a focus on the project results and one focused on the process ) [51] , [52] . In order to make contributions as easy as possible , Todd regularly posted an update on TSLS detailing progress and descriptions of the next tasks needed [53] , [54] . This minimized the time that potential contributors needed to sift through backdated postings to come up to speed . It also avoided duplication of efforts . The project did not have an official project plan or deadlines , but it was time-constrained by funding parameters ( three years ) . Even after the postdoctoral researcher was hired to contribute , there was a hope that greater external interest could be raised for the project . Todd began giving speeches including a Google TechTalk in April 2010 [55] . After each article , blog and presentation , the project experienced significant increases in website traffic [51] . It was also decided to reach out to a closed chemistry networking forum on LinkedIn . This positively resulted in 20 comments from 11 different scientists , new to the project , and four private e-mails [51] . One of the respondents was a Dutch contract research organization interested in participating in the project [51] . This was an important milestone for the project because the CRO had the equipment and expertise to perform some of the necessary tasks very quickly ( they completed tasks in weeks as opposed to the months it would probably have otherwise taken ) . This industry-academic support enabled the project to complete the project before the funding ran out . Ninety-seven ( 97 ) individuals have registered on the Synaptic Leap indicating that they are actively participating or are interested in participating in research for schistosomiasis . Thirty-seven ( 37 ) contributed to the TSLS project . The contributors include six members of Todd's team , four industry representatives , 15 academics/researchers , one retiree , two informatics professionals , and 9 of unknown affiliation . Contributors were based in Africa , Europe , Oceania and North America . Only one postdoctoral researcher from the University of Sydney was paid specifically to work on the project . Motivations for participation included accelerating own research , intellectual stimulation , signaling abilities and a belief in the benefits of open collaboration . Their contributions ranged from one-off comments regarding the project to substantial postings regarding laboratory results . TSLS places all scientific discoveries in the public domain , therefore , obviating the ability to patent them . All of the website content is copyright protected according to the Creative Commons Attribution 2 . 5 License unless otherwise stipulated [56] . All content may be viewed without a username and password . If an individual wants to make a posting on the Synaptic Leap website , he/she can either leave a comment as a guest or as a registered user . A guest must supply a valid e-mail address which is not viewable with the comment . Registering requires a username and e-mail address . An automated system sends a log-on password . There is no requirement to accept a license at time of registration . Intellectual property does not play a major role in this project since a version of praziquantel has been in the public domain for almost two decades . The scope of this project was limited to a specific problem . Once they managed to generate a single enantiomer of praziquantel , the expectation was that the project would be complete ( although the project continues looking at more elegant solutions to the problem ) . The next steps of scaling up the modified drug to commercial quantities and any regulatory approvals needed would be performed externally by a pharmaceutical manufacturer in partnership with WHO . Funding was important to the project because it allowed for the recruitment of a full-time postdoctoral researcher whose postings provided fresh , regular content giving the project momentum . The grant money paid the salary of the postdoctoral student , all administrative supplies and covered the cost of shipping the samples to any interested laboratory . Contributing organizations did not receive any monies from the project . Firstly , we would like to acknowledge that both cases have made great accomplishments in meeting their aims . CSIR OSDD has persuaded a large number of volunteers to contribute and published four articles in 2011 , a significant accomplishment for a group of volunteers . TSLS has gathered contributors from around the globe , both from academia and the private sector and has managed to fulfill its goal . The two cases operate very differently and differ greatly in magnitude . CSIR OSDD is a vast project , encouraging international collaboration on its website , but in actuality , geared principally towards Indian researchers and students . The funding from the Indian government applies only to activities within India [24] . There are many workshops and face-to-face meetings in India as well as private e-mail correspondence between teacher and pupil . This , in essence , translates into an Indian-centric project . TSLS , on the other hand , has attracted contributors internationally , albeit substantially fewer than CSIR OSDD , with a variety of motivations . Both have obtained funding used to pay for access to facilities , physical resources and , at times , labor costs . TSLS releases its results into the public domain , where as CSIR OSDD asserts ownership over its results . If we return to R4D's definition of open source – the application of open access , open collaboration and open rules – it is useful to analyze each case's adherence to the definition in order to understand the impact of this adherence ( see Table 4 ) . CSIR OSDD's scientific research results are placed on Sysborg which requires a user to log on before any content may be viewed . The content is not searchable through general search engines like Google . Technically , the content is open access because a username and password are eventually granted to users allowing them to view the data free-of-charge . However , we believe that this tight control of the data is actually a barrier to entry . Most potential contributors will want to browse the website before contributing , and they may lose interest in the two days or more that it takes to receive access to the full content . Indeed , a few TSLS contributors reported through the interviews that they had tried to access CSIR OSDD and had given up in frustration . CSIR OSDD' process limits contributors to only those who have a strong motivation to contribute . CSIR OSDD has assigned certain tasks to partner institutions . This is likely a practical solution to achieving progress . These institutions receive funding and have commitments back to CSIR OSDD . They must follow an agreed structure and process . Other institutions or individuals can no doubt assist in any activity . However , since much of the process is opaque ( through face-to-face meetings , Skype or private e-mail ) and not reported back through Sysborg , open collaboration is difficult . This opaqueness does not promote cross-organizational or geographical linkages . Until the processes and decision-making are made more transparent and easier to follow on Sysborg , CSIR OSDD does not fit the definition for open collaboration . CSIR OSDD's license awards the project ownership over all data . Data may not be used by other entities without entering into a contract with CSIR OSDD . The license may also be considered viral since all improvements based upon CSIR OSDD data are to be granted back to CSIR OSDD ( i . e . future generations of improvements are subject to the CSIR OSDD license if any of the original CSIR OSDD data was used ) . This may make industry shy away from participating in the project . CSIR OSDD has taken a very protective approach of its data likely so that it is not expropriated and exploited by a third party . This is understandable considering the potential commercial value of new tuberculosis medicines . However , CSIR OSDD's license does therefore not mandate “openness” . CSIR OSDD states that the project will shepherd its new products up through regulatory approval and then make them available to the generic drug industry without any exclusivity [24] . It is unclear whether they intend to patent the drugs and offer a non-exclusive license to generic manufacturers , utilize the public domain or an alternative intellectual property strategy . Perhaps they have not yet decided themselves . The license language , however , does not mandate openness . We believe that rather than a strictly defined open source project , CSIR OSDD is actually a highly successful crowdsourcing project , using volunteers to perform specified and structured tasks . They have achieved most of the advantages of open source identified by R4D . The data results are verified ( although offline ) , but the project's impressive publishing demonstrates that its work has passed peer review muster . The contributions of 400+ volunteers result in a significant cost savings . Undoubtedly , any medicines that they develop will enter the market at a low price point . They have not , however , succeeded in creating open collaboration or a public commons of knowledge . They have created a proprietary knowledge repository . TSLS is largely in adherence to the open source drug discovery definition . All of TSLS' data are publicly-available without a password . Searches within Google for related TSLS content return all of TSLS' related websites . This makes it easy for potential contributors to firstly find the project and then browse the content to get a feeling for the project . However , TSLS' website could also be improved . Postings are not necessarily in chronological order and there is no easy method to see all postings related to one disease area . Thanks to TSLS' project manager's continuous efforts to summarize the current state of play , these inconveniences are minimized . TSLS' websites allow for open collaboration across organizational and geographical boundaries . It is stressed that e-mails should be avoided . Raw data is placed directly on the website awaiting virtual peer review . Observers can easily follow the threads of the process . TSLS uses well known legal concepts with the public domain and a creative commons license . Both mandate “openness” . Results may be utilized by third parties without contracts or royalties . TSLS has achieved all of the open source advantages . Its content is transparently verified on the website with the additional peer review of publishing in top-ranked journals . The data forms a knowledge commons . Global collaboration was achieved between representatives from both academia and industry . The grant funding and volunteer contributions of industry significantly sped up the progress of the project , achieving cost savings .
These two cases demonstrate that drug innovation can be performed using an open source approach , albeit in very different ways and not necessarily in strict adherence to the definition of open source drug discovery . Adherence to the definition is not necessarily that important . As a crowdsourced project , CSIR OSDD has still achieved great success by persuading volunteers to perform high quality research at low cost , which , of course , is the goal of open source collaboration . The definition is still useful though , to evaluate how different projects approach transparency , collaboration and access to results , but not necessary to spur on high quality , low cost drug discovery . The cases do point to three common critical success factors: clearly defined entry points , transparency and funding . Both projects attracted volunteers by publicizing the respective projects through descriptive articles in academic journals and utilizing social media and networks . CSIR OSDD has also effectively paired up with Indian universities and colleges , incentivizing students to volunteer as parts of classroom assignments or positioning participation as valuable hands-on experience . They have also built in an element of patriotism , linking finding cures for tuberculosis as an Indian responsibility due to the high prevalence of tuberculosis in India . This patriotic effect is reinforced through project marketing efforts , like the project's music video [57] . The entry point into CSIR OSDD is through the classroom which is likely to limit international participation in the project . Rather TSLS' entry point is through the website , using frequent status updates to pinpoint exactly the tasks currently needed . The two cases approach collaboration and progress in different ways . TSLS takes a very transparent approach , posting raw data , containing the discussion to publicly-available websites and placing results in the public domain . CSIR OSDD takes a more cautious approach with a significant amount of work being performed through face-to-face or Skype meetings , greater use of private e-mail exchanges and a license that emphasizes mostly trust in the project's mission rather than legally-binding clauses stipulating open access to the data . Funding was an absolute necessity for both projects . Without it , they would not have been able to access the laboratories and physical supplies needed for drug innovation , hire the minimum number of employees needed to give the projects their initial momentum , or perform routine administrative functions ( such as website hosting ) . How much savings each project has achieved through the use of volunteers is uncertain . The Global Alliance for TB Drug Development calculated in 2001 that the estimated costs of discovering and developing a new anti-tuberculosis drug ( including the costs of failure ) where between US$115 million and US$240 million [58] . CSIR OSDD has about US$35 million at its disposal but they are still in early days , having yet to embark upon the most expensive part of the process , clinical trials . Maurer [59] in 2005 estimated that lead compound optimization costs between millions to tens of millions of US dollars . The chemists from TSLS achieved their result with about US$330 , 000 . However , they were working with a known , effective lead compound with a specific problem . The results of this case study cannot be generalized to all open source drug discovery projects since we only examined two , separate efforts . However , we believe that our findings are relevant to other projects interested in the open source model . Firstly , the cases give an indication of the number of participants necessary to achieve different drug discovery tasks . TSLS managed to complete its task with a relatively modest 37 individuals , with only a few of these dedicating large amounts of time to the project . On the other hand CSIR OSDD will require hundreds of contributors to discover and develop a new tuberculosis medicine . The market realities of the potential drugs may also play a role in a project's adherence to the strict definition of open source . CSIR OSDD has reasonable grounds for protecting their data through a gated community and a protective license , namely that new tuberculosis products offer private companies with a profit potential in both developed and developing countries , especially lucrative if they have not had to invest in R&D . The public domain is not actually an intellectual property right but the absence of one . If a patent were to be granted to others on the knowledge developed by CSIR OSDD , the only way to defend against that claim would be a costly court trial . One can therefore argue that CSIR OSDD has utilized a protective license as a negative measure to safeguard others trying to patent the knowledge . In contrast the risk that TSLS' version of praziquantel will be expropriated and patented is next to null since schistosomiasis is only endemic to developing countries and the generic form of praziquantel is already available cheaply . Unfortunately our case study is weakened by a rather low interview response rate from the CSIR OSDD project . We surmise that our timing was unlucky as an article critical of the project appeared just before we started recruitment [60] . This paper criticized the project for not publishing its first results in a peer-reviewed journal . A few potential interviewees expressed skepticism that we did not harbor an ulterior , negative motive . We debated the benefits of offering a cash prize to gather additional respondents but decided that this may only fuel the skepticism surrounding our study . We attempted to compensate by closely examining the content on the websites including interactions and self-reported data . We also submitted our results to the project manager of each of the two cases and incorporated their feedback into the final paper . Are CSIR OSDD and TSLS model cases for open source drug discovery ? It is too early to tell . Since there are so few instances of open source drug discovery , the model is still being developed , most recently with an interesting new joint project between TSL and CSIR OSDD with a focus on malaria initiated in 2011 [61] . More modeling is still needed , especially in evaluating the potential of hybrid models that combine open source with standard intellectual property mechanisms like data exclusivity and secrecy . Interesting examples ( like the public-private partnership , the Archipelago to Proof of Clinical Mechanism [62] ) are combining these approaches in the areas of neurology and oncology . The recently released report of WHO's Consultative Expert Working Group on R&D Financing and Coordination [5] has called for greater use of “open knowledge innovation” . This concept is more general than open source and groups open source drug discovery with equitable licensing , patent pools and prizes ( in other words , creating a grouping of the drug discovery and access business models with a primary focus of sharing of open knowledge , particularly to meet the needs of low-income countries ) . As organizations consider acting on the expert group's recommendations , and possibly funding organizations begin requiring a certain level of adherence to the open source model , the model will become more mainstream , giving a new level of transparency and access to the data needed to more efficiently finding cures for neglected diseases . | Open source drug discovery can be an influential model for discovering and developing new medicines and diagnostics for neglected diseases . It offers the opportunity to accelerate the discovery progress while keeping expenditures to a minimum by encouraging incremental contributions from volunteer scientists . Publishing raw data and results in the public domain is positive within the context of neglected diseases since it facilitates open collaboration while obviating the ability to patent any results . In this way it effectively de-links the research and development costs from the sales price of the end product , the new medicine or diagnostic . This case study demonstrates that implementations of the open source model can differ while still achieving the ultimate goal of obtaining high quality research at reduced costs . However , the importance of clearly defined entry points , transparency and funding are shared success factors . These findings present the practical challenges of implementing a theoretical concept and hopefully will assist other scientists in organizing future open source drug discovery projects . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"medical",
"devices",
"drug",
"research",
"and",
"development",
"drugs",
"and",
"devices",
"global",
"health",
"pharmacology",
"drug",
"discovery"
] | 2012 | Open Source Drug Discovery in Practice: A Case Study |
Inverted repeats ( IRs ) can facilitate structural variation as crucibles of genomic rearrangement . Complex duplication—inverted triplication—duplication ( DUP-TRP/INV-DUP ) rearrangements that contain breakpoint junctions within IRs have been recently associated with both MECP2 duplication syndrome ( MIM#300260 ) and Pelizaeus-Merzbacher disease ( PMD , MIM#312080 ) . We investigated 17 unrelated PMD subjects with copy number gains at the PLP1 locus including triplication and quadruplication of specific genomic intervals—16/17 were found to have a DUP-TRP/INV-DUP rearrangement product . An IR distal to PLP1 facilitates DUP-TRP/INV-DUP formation as well as an inversion structural variation found frequently amongst normal individuals . We show that a homology—or homeology—driven replicative mechanism of DNA repair can apparently mediate template switches within stretches of microhomology . Moreover , we provide evidence that quadruplication and potentially higher order amplification of a genomic interval can occur in a manner consistent with rolling circle amplification as predicted by the microhomology-mediated break induced replication ( MMBIR ) model .
Inverted repeats ( IRs ) are a common architectural feature within the human genome and can predispose loci to rearrangement [1–3] . An IR-mediated inversion that disrupts the Factor VIII gene causes ~45% of severe hemophilia A cases [4] . The importance of IRs to human genomic rearrangements and resultant genomic disorders and the expanded scope by which IRs can facilitate genomic change are now apparent [2 , 3 , 5–7] . The abundance of inverted low copy repeats ( LCRs ) or segmental duplications genome-wide suggests that ~12% of the genome may be susceptible to inversion mediated by IRs [2] . Fosmid paired-end sequencing of 8 human genomes from diverse populations shows that ~50–100 large genomic inversions not represented in the human genome reference sequence are present in the personal genome of each individual . In total , 224 non-redundant inversions were identified in 8 genomes; these events are primarily mediated by larger blocks of homology [8] . Earlier work provided experimental evidence for genome-wide inversions and suggested these can occur somatically and with aging [9] . Moreover , inverted repetitive regions that are smaller than conventional LCRs , designated self-chains , are also associated with genomic instability furthering the impact of IRs on both structural human differences and phenotypes [3] . Recently , IRs were shown to mediate complex duplication—inverted triplication—duplication ( DUP-TRP/INV-DUP ) rearrangements , leading to MECP2 duplication syndrome ( MIM#300260 ) , Duchenne Muscular Dystrophy ( MIM#310200 ) , VIPR2 triplication , CHRNA7 triplication , and Pelizaeus-Merzbacher disease ( PMD , MIM#312080 ) [1 , 10–13] . The mechanisms for such complex genomic rearrangements ( CGRs ) have only begun to be elucidated . Genomic rearrangements leading to the duplication of the X-linked proteolipid protein 1 ( PLP1 ) gene are the major mutational cause for PMD and explain ~80% of patients; point mutations in PLP1 occur less frequently , and higher copy number gains ( e . g . triplications ) and deletions are rare [14–17] . CGR can cause PMD by duplicating PLP1 via a mechanism that results in a DUP-TRP/INV-DUP structure [1] . Consistent with a gene dosage hypothesis , and as established for both homozygous duplication [18] and heterozygous triplication [19] at the CMT1A locus , triplication of PLP1 can lead to a more severe form of PMD than duplication [13 , 14] . Using high-density array comparative genomic hybridization ( aCGH ) , DUP-TRP/INV-DUP rearrangements primarily contain one variable breakpoint at the proximal ( centromeric ) end [1]; however , distal breakpoints for the triplication to duplication and duplication to normal copy number transitions cluster at inverted LCRs distal to MECP2 [1] . The proposed mechanism for these CGR involved a two-step process: i ) break-induced replication ( BIR ) within homologous regions of the inverted LCRs forming a breakpoint junction ( Jct1 ) and ii ) microhomology-mediated BIR ( MMBIR ) or non-homologous end-joining forming a second junction ( Jct2 ) . Mutational signatures observed at the latter junction include microhomology , templated insertions , and increased point mutation frequency [1 , 20] . However , in both MECP2 and PLP1 DUP-TRP/INV-DUP rearrangements , delineation of unique breakpoint junctions within the IR has been hampered by the complexity of large blocks of homologous sequences creating challenges to mapping Jct1 at base pair resolution . To further investigate mechanisms for CGR formation we analyzed a cohort of 17 unrelated PMD patients with copy number gains at the PLP1 locus , including duplications , triplications and quadruplication . Analysis of phenotypically normal individuals elucidated a common inversion polymorphism associated with the IRs distal to PLP1 . Southern blotting experiments established an estimated frequency for the inversion . We postulated and confirmed that the LCR substrates responsible for the inversion are also responsible for one breakpoint junction ( Jct 1 ) in each PMD associated CGR . Additionally , we document a DUP-TRP/INV-DUP rearrangement product structure at the PLP1 locus in the personal genomes of 16 subjects with PMD and provide evidence that such CGR can occur by replicative mechanisms [21] . Finally , we investigated the quadruplication of a genomic segment proximal to PLP1 and found the potential mechanism of formation to be consistent with rolling-circle replication leading to amplification—a mechanism predicted by the MMBIR model [22] .
The 186 kb genomic interval ( ChrX: 103 , 172 , 000–103 , 358 , 000 in hg19 ) located ~150 kb distal to PLP1 contains a complex genomic architecture in the haploid reference genome . This region consists of an array of IRs , with the ~40 kb outer C and D repeats having ~93% identity , the middle A1a and A1b repeats ~20 kb in size and ~99% identical , and the innermost ~10 kb A2 and A3 repeats showing ~87% identity both with each other and with A1a and A1b ( Fig . 1A ) [16 , 23 , 24] . The IR architecture predicts the potential for inversion mediated by non-allelic homologous recombination ( NAHR ) , resulting in at least two structural haplotypes , analogous to the H1 and H2 structural variant ( SV ) alleles at the MECP2 locus [1] . Indeed , in silico analysis of the human genome SV track from the UCSC Genome Browser ( www . genome . ucsc . edu ) suggests the existence of such an SV allele [25] . The browser track indicates fosmids consistent with inversions spanning both of the A1a and A1b LCRs in 5 of 9 individuals ( S1 Fig ) [8 , 26] . These data indicate that there was an inversion between A1a and A1b LCRs and that the inversion haplotype exists at a relatively high allele frequency as a non-pathogenic rearrangement in HapMap individuals ( Fig . 1B ) . Further investigation mapped the apparent ectopic crossover for the NAHR-mediated inversion in a fosmid from the G248 library to nucleotide-level resolution ( S2 Fig ) . To directly examine the inversion SV polymorphism between A1a and A1b , we designed a Southern blotting assay and genotyped multiple individuals from different populations of origin for reference ( arbitrarily designated H1 ) or inversion ( H2 ) structural haplotypes . The scheme of the assay is depicted in Fig . 1C , wherein Southern analysis leads to predicted visible fragments of 25 kb for H1 and/or 29 kb for H2 . As the rearrangement is on the X chromosome , males should have only one allele , and females , two . Genotyping 17 individuals ( including 3 males ) with this assay discerned 31 haplotypes of the X chromosome ( Figs . 1C , S3 , and S1 Table ) . The frequencies of structural haplotypes were 13/31 H2 ( ~42% ) and 18/31 H1 ( ~58% ) , with 4 individuals hemizygous or homozygous for H2 and 7 for H1 . The remaining 6 females were heterozygous for both H1 and H2 . The 17 individuals were of Japanese , CEPH Northern European , Han Chinese , Yoruban and unknown populations of origin , and all populations contained both H1 and H2 structural haplotypes ( S2 Table ) . We hypothesized that the similarity between LCRs A1a and A1b and their relatively large length and proximity ( ~20 kb repeats of ~99% identity separated by ~50 kb ) could predispose to recurrent events [27 , 28] . We analyzed the genomic region encompassing the two LCRs and identified multiple adjacent single nucleotide polymorphisms ( SNPs ) spanning the region in linkage disequilibrium and delineating a haplotype block extending for ~0 . 5 Mb with a recombination rate of 0 . 3 centimorgans per Mb . The two SNP haplotype blocks were evenly distributed between the 14 different populations from the 1 , 000 genomes project [29] . Superimposing Southern blotting results for individuals homozygous or hemizygous for SV haplotypes on top of the SNP haplotypes enabled phasing; 6/7 inversion ( H2 ) alleles were on one SNP haplotype and 1/7 was on the other ( belonging to individual NA18947 , S4 Fig ) , whereas homozygous H1 alleles occurred on either SNP haplotype . Heterozygous calls are uninformative , as the structural haplotype information cannot be phased to the SNP data . These data suggest that the inversion is likely recurrent in the population and makes population estimation of the structural variant using SNP genotyping unlikely to reflect the true population frequency . Sixteen patients with PMD and one diagnosed with spastic paraplegia type two ( SPG2; MIM#312920 ) were examined by aCGH for copy number variation ( CNV ) in PLP1 and the surrounding genomic region . A schematic of CNV observed in the personal genomes from 17 patients is depicted in Fig . 2A . PLP1 duplications were detected in 10 patients ( BAB1290 , BAB2389 , P250 , P298 , P500 , P558 , P842 , P1389 , P1407 and P113 ) , whereas triplications were detected in 6 patients ( BAB3698 , P518 , P642 , P674 , P820 , and P1150 ) . The one SPG2 patient , BAB1612/P374 has been described previously , and the phenotype of this individual is ascribed to a potential position effect [30] . The distal breakpoints in all subjects appear to cluster in approximately the same genomic location; however , there are few probes on the arrays that can specify unique loci within the C/D , A1a/A1b , and A2/A3 LCRs due to the repeat nature of the region . Thus , determining the precise LCRs involved in the breakpoints required alternate mapping approaches . Array and semi-quantitative PCR data , summarized in Fig . 2A ( see also S5 Fig and Table 1 ) , indicate that the region of rearrangement spans from 145 kb ( BAB1612/P374 ) to ~4 , 000 kb ( BAB2389 ) . Triplicated genomic segments range in size from 254 bp ( P298/P255 ) to 575 kb ( P642 ) . Proximal triplication and duplication copy number transitions differed in each individual and were not located within LCRs . The distal copy number transitions group within a 100 kb region of uncertainty as described above . The triplication present in P298/P255 was too small to be detected by aCGH; however , a 254 bp triplicated genomic segment was detected both by amplification and sequence analysis with unique flanking primers and by quantitative PCR ( S6 Fig ) . FISH was performed on nuclei prepared from peripheral blood lymphocytes from P642 , P1150 , and P113 . This independently corroborated interpretation of array and semi-quantitative PCR data using an orthogonal experimental approach . Moreover , FISH determined whether extra copies of the genomic segments were located in or near the PLP1 locus as opposed to elsewhere in the genome; arrays determine neither the orientation nor the position of a copy number segment , but only specify the genomic segment that underwent a gain in copy number ( Fig . 2B-D ) . Interphase nuclei of patients P642 and P1150 showed , as expected , one green control probe signal and revealed three closely-spaced red PLP1 probe signals indicating triplication at the PLP1 locus; P113 had two red PLP1 signals indicating duplication at that locus , but four proximal probe signals confirmed the additional quadruplication ( Fig . 2D ) . Metaphase spreads of all 3 patients gave one green control probe signal , one red presumably merged PLP1 probe signal on the X chromosome , and no signals on other chromosomes , also indicating that the triplications and duplications were at the PLP1 locus rather than being located elsewhere in the genome and that the triplications were too small to resolve on metaphase chromosomes . We investigated haplotypes using genetic markers , 2 short tandem repeats ( STRs ) and 9 SNPs , mapping over a 258 kb region of the duplications with 4 markers mapping within PLP1 and the remainder distal to it . We observed that 12 of 13 patients tested ( P250 , P255/298 , BAB1612/P374 , P500 , P518 , P558 , P642 , P674 , P820 , P842 , P1150 , P1389 , and P1407 ) , were monomorphic displaying only one form for each marker genotype ( S3 Table ) . The DNA from BAB1612/P374 was only interrogated at the 7 sites distal to PLP1 , since this is where his triplicated/duplicated region lies . In this subject , only one form was detected for all markers except the STR furthest distal to PLP1 where two were detected . The finding of an absence of bi-allelic loci in these multi-copy regions of X is most parsimoniously explained by the occurrence of intra-chromosomal rearrangement events , as has also been observed for DUP-TRP/INV-DUP rearrangements at the MECP2 locus [20] . In P1150 , we obtained a breakpoint junction between the proximal ( centromeric ) endpoint of the rearrangement and the proximal end of the triplicated region via inverse PCR . We then hypothesized that our other patients with duplication-triplication-duplication copy number changes could potentially have the same CGR product structure and explored this hypothesis by long-range PCR on each individual personal genome . We were able to amplify and sequence across the proximal breakpoint junction in all 16 patients ( Table 1 ) . The 16 junctions each indicate that the triplicated region is inverted with respect to the proximal duplication region ( S6 Fig ) ; in 6 cases the triplication encompasses PLP1 . This is a potentially analogous rearrangement structure to that previously described for MECP2 CGRs; therefore , we denote the non-recurrent junctions in Table 1 as Jct2 [1] . We had previously mapped and sequenced across the duplication breakpoint junction of patient P255 , who had a 254 bp inverted duplication [24] . We no longer had DNA from P255 to interrogate the copy number of the region; therefore , we tested an affected family member , P298 , by qPCR and found , as anticipated , that the region is triplicated . The Jct2 sequences in the 16 patients are shown in Table 1 . Fourteen patients contain one or more breakpoint junctions displaying microhomology . Patients P558 and P842 have blunt junctions . In 13 of the patients , endpoints at Jct2 are in repetitive element sequences , and in P1389 , one end was in an LCR ( Table 1 ) . Patient BAB1612/P374 contained a LINE2-mediated event ( L2/L2 , both within the same LCR ) that did not result in a chimeric element . Patients P518 and BAB3698 contain chimeric AluS elements formed in the generation of this junction . In BAB3698 , there are 47 bp of identity between the two AluSx elements at the transition from triplication to duplication . In P518 , the rearrangement occurs through the formation of two AluS chimeric junctions , the first ( from proximal to middle segment ) in the same orientation in 14 bp of identity , and the second ( from middle segment to triplication ) in 34 bp of identity ( see S6 Fig ) . This complex breakpoint junction contains a segment of 488 bp that consists of an AluSx3 , an L1 sequence ( L1ME ) , and an AluSq2 . Interestingly , the distal ( triplication ) to middle junction occurs between these two Alu elements that are only separated by 310 bp , suggesting a potential U-turn caused by inverted Alus within close proximity [31 , 32] , similar to the situation in Jct1 but mediated by short Alu sequences instead of LCRs ( S6 Fig ) . The breakpoint junction mutational signatures are consistent with replicative mechanisms such as MMBIR or a homeologous ( near homologous ) recombination event between similar Alu elements at each instance of Jct2 [22 , 33 , 34] . Complexities that included an additional template switch were observed in Jct2 from individuals P500 , P518 , BAB1290 and BAB2389 ( Table 1 and S6 Fig ) . Such events have been postulated to reflect reduced processivity of the replisome mediating MMBIR during initial template switching [22] . We also amplified across a breakpoint junction present in P1150 , indicating a 27 kb deletion on one of the duplicated copies ( Figs . 2 and S5 ) . At that junction , there is a bp of microhomology ( S6 Fig ) . The overall findings for Jct2 are consistent with both long distance template switching and a microhomology-mediated mechanistic process such as FoSTeS/MMBIR [21 , 22] . After Jct2 was determined for the 16 patients , we hypothesized that a likely genomic arrangement consistent with this junction was one in which one copy of the triplicated region was situated in an inverse orientation between the two copies of the duplicated region and that the other two copies of the triplication were embedded within the duplicated regions , i . e . a DUP-TRP/INV-DUP structure [1] . Patients with presumed DUP-TRP/INV-DUP rearrangements with sufficient DNA available were subjected to Southern blotting ( 10/16 total ) to examine whether the same repeats involved in the common inversion polymorphism are also involved in the CGR and to investigate on which structural haplotype the rearrangement occurred . The Southern scheme in Fig . 1C was used to analyze patient DNAs; however , in a male with PMD caused by DUP-TRP/INV-DUP involving the A1a and A1b repeats , the Southern blot does not reflect the normal copy number of one allele of the X chromosome ( either H1 or H2 ) ( Fig . 3A , S4 Table ) . Instead , the rearrangement gives rise to two copies of the original haplotype plus an additional “flipped” haplotype in an affected individual with DUP-TRP/INV-DUP leading to PMD , similar to the observation described for the MECP2 locus [1] . This assay can presumably distinguish the SV haplotype on which the genomic rearrangement occurred . A representative gel and labeled blot are shown in Fig . 3B , with the dosage of the bands indicating that subjects BAB1290 and BAB1612/P374 both carried the inversion H2 structural haplotype prior to the rearrangement . Interestingly , the 10 individuals examined by this assay appeared to use the A1a and A1b LCRs as the substrates for their rearrangements , in spite of two other IRs being located in close proximity ( BIR between C/D would lead to duplication of H1 or H2 and A2/A3 would lead to triplication ) ( Figs . 1A , 3C and D ) . Individuals BAB1290 , BAB1612/P374 , BAB2389 , BAB3698 , P500 , P518 , and P642 all contained a rearrangement that had occurred on the inverted H2 allele , while P250 , P298 , and P558 had a Southern blot result indicating the rearrangement occurred on an H1 haplotype ( S4 Table ) . A three-generation family was studied in which the two maternal grandparents were unaffected , and subsequent Southern blotting and aCGH data indicated that the grandmother ( BAB4179 ) was not a carrier and that she had two copies of the inverted H2 locus ( Figs . 3C and S5 ) . The grandfather was unavailable , but did not have PMD; therefore , the de novo rearrangement can be inferred to have occurred in between the grandparent and the maternal generation . The mother ( BAB3700 ) was a carrier of the rearrangement and had equal dosage of H1 and H2 on a Southern Blot . The affected son ( BAB3698 ) had Southern results consistent with rearrangement on H2 , and his carrier sister ( BAB3699 ) had similar results to the mother . These findings are consistent with the de novo DUP-TRP/INV-DUP occurring in association with “flipping” the H2 haplotype to an H1 haplotype , a mechanism similar to that observed for CGRs at the MECP2 locus ( Fig . 3A ) [1] . The assay results in this family are most parsimonious with the rearrangement occurring on one of the grandmother’s inversion-containing alleles ( H2 ) , and having balanced copy number in BAB3700 and BAB3699 due to the additional allele being a reference ( H1 ) 25 kb band . This would result in a 2:2 dosage of 29 kb:25 kb bands on the Southern blot , which we observe in both BAB3700 and 3699 ( Fig . 3C and S4 Table ) . As Jct1 occurs within the LCR region distal to PLP1 , the junctional products are not readily amplified and sequenced by long PCR with primers anchored to unique flanking sequence . We adopted an alternative strategy to complement the Southern blotting assay above . Using a semi-quantitative PCR approach , we first confirmed that each of the patients has duplication of A1a and A1b LCRs ( black primer pair in Fig . 3E ) and triplication of a region proximal to A1a ( red primer pair and black/red primer pair in Figs . 3E , S7 ) . This PCR approach independently verified the Southern Blot results and suggested a crossover breakpoint within the A1a or A1a/A1b chimera present on H2 ( Fig . 3E ) . We attempted to more narrowly define the crossover region in our patients by using sequence differences between the LCRs ( paralogous sequence variants or PSVs ) , but patients appeared to lack apparent PSVs between A1a and A1b that were at the corresponding genomic locations in the hg19 reference sequence [35] . To determine sequences across Jct1 , we designed a PCR-cloning assay that allowed us to amplify large ( ~12–16 kb ) , overlapping portions of both A1a and A1b LCRs that are implicated in the rearrangements [35] ( Figs . 3F , S8 ) . Three individuals were subjected to this analysis ( BAB1612/P374 , BAB2389 , and BAB1290 ) , however BAB2389 and BAB1290 appear to have Jct1 within a large region of identity ( >8 kb ) in the center of the LCR that lacks PSVs between cloned segments; therefore , further refinement of the breakpoint junction was intractable using this method . Additionally , in P255/298 , a PCR approach using one primer at the proximal duplication junction and one within the LCR corroborated that the breakpoint indeed occurred within this >8kb stretch of identity . In contrast to the three other individuals for whom we sought to find Jct1 at base pair resolution , in BAB1612/P374 we were able to detect an LCR-mediated breakpoint within 24 bp of microhomology flanked by A1a and A1b sequences ( Fig . 3F ) . The point of crossover within this sequence was confirmed by direct PCR amplification and sequence analysis from genomic DNA followed by comparison to the PSVs present on cloned A1a and A1b sequences from the same individual; its identification elucidates Jct1 within an LCR , a heretofore un-investigated junction at the nucleotide level of resolution . The DUP-TRP/INV-DUP structure hypothesized for these 16 individuals postulates that although there are 4 copy number transitions in these patients , there are only two breakpoint junctions ( Fig . 4A ) . We have sequenced Jct2 in all 16 patients; Southern blotting and quantitative PCR were used to determine Jct1 , and direct junction sequencing was successful for BAB1612/P374 ( Figs . 3 , 4 and S6 ) . Additionally , due to the small size ( ~ 254 bp ) of the triplication in P255/298 , a PCR approach using one primer at the proximal duplication junction and one within the LCR validated the overall structure of this rearrangement as DUP-TRP/INV-DUP . We have discerned two junctions from patient P113 with proximal quadruplication and duplication of PLP1 using long-range PCR ( Figs . 5A , S6 ) . Junction 1 consists of one fork stalling and template switching ( FoSTeS ) event—FoSTeS 1 ( Fig . 5A ) . The second junction , between the proximal end of the triplication and the distal end of the quadruplication , consists of FoSTeS events 2 and 3 ( S6 Fig for sequences of all junctions ) . We determined that the rearrangement was on the inverted H2 allele using PCR genotyping of the haplotype present in P113 ( S9 Fig ) . Additionally , digital PCR ( dPCR ) data indicate that the FoSTeS 1 occurs in one copy , and FoSTeS 2/3 occurs in 2 copies ( S5 Table ) . This quadruplication rearrangement is also associated with a de novo point mutation ( G insertion ) ~50 bp away from the junction that appeared to occur concurrent with the rearrangement , as observed for other CGR mediated by a replicative process ( S6 Fig ) [20] . The mechanism by which copy number increased from 3 to 4 copies and generated the quadruplication is suggestive of a rolling circle amplification , wherein one breakpoint is repeated twice in the process of replicating ~280 kb ( S5 Table ) [22 , 36–38] . The FISH data for this individual shows the rearrangement to be contained on the X chromosome , and family data including the proband P113 , his mother P154 , his affected uncle P117 , and grandmother P088 suggest that the structure is stable in 4 individuals from 3 generations , diminishing the likelihood of recombination-based amplification ( Figs . 2D and S5 ) . If the amplification were mediated by NAHR ( see S10 Fig ) , this rearrangement would contain two templates for subsequent rounds of amplification . Therefore , the rearrangement in P113 should be twice as likely to undergo expansion as the proposed intermediate . Additionally , the presence of monomorphic SNPs throughout the region of copy number gains on SNP arrays indicates that the rearrangement was intra-chromosomal , as was found for the 13 patients with DUP-TRP/INV-DUP rearrangements ( S11 Fig ) . The quadruplication-containing CGR was observed to transmit stably and co-segregate with disease through three generation ( S5 and S11 Figs ) . A proposed mechanism for the rearrangement occurring in one complex quadruplication event is shown in Fig . 5 and consists of a rolling-circle amplification of the triplicated and quadruplicated segments .
PLP1 is surrounded by LCRs of variable size and sequence similarity; previous studies have shown that such genomic architecture renders this region unstable and susceptible to rearrangements , leading to PMD [16 , 23 , 24] . We show that DUP-TRP/INV-DUP rearrangements are a frequent CGR product at the PLP1 locus and that they are facilitated by a complex IR but specifically mediated via the ~20 kb A1a and A1b 99% identical repeats . These particular IRs are not only driving CGR observed in patients but additionally mediate a common SV polymorphism—copy-number neutral inversions at Xq22 . 2 . The latter can complicate interpretation of CGRs in the region; a proposed breakpoint can also appear in an unaffected individual in the guise of an inverted allele [13] . Additionally , the recurrence of this inversion might confound the correlation of diagnostic SNPs with structural information , leading to the underestimation of the frequency in a population [39] . These data suggest that IRs with a high degree of identity that are involved in non-pathogenic inversions can also drive seemingly recurrent breakpoints in non-recurrent rearrangements associated with disease and that this occurs at multiple genomic loci [1 , 12 , 13] . Indeed , a previous determination of genes potentially subject to CNV via DUP-TRP/INV-DUP due to proximity of homologous IRs predicted the PLP1 gene might be affected [2] . The proximal junctions , or Jct2 , in the DUP-TRP/INV-DUP rearrangements at the PLP1 locus are depicted in S6 Fig . Jct2 is non-recurrent in the 16 individuals , with different genomic coordinates for each breakpoint . Interestingly , investigation revealed that 14 of 16 Jct2 sequences contained microhomology of 1–4 bp at one or more of the FoSTeS events in the junction . Two of these Jct2 sequences involved larger stretches of microhomology; one contained an Alu-Alu chimeric event with 47 bp of perfect identity at the junction and the second CGR contained two Alu-Alu chimeras , one containing 14 bp and the other with 34 bp of perfect identity at the junction ( Table 1 ) . These data suggest that a replicative mechanism is involved in the formation of Jct2 in a majority of cases . Previously , we proposed that MMBIR or NHEJ could be responsible for Jct2 [1] . Here , we expand this ‘”two-step hypothesis” to include homeologous recombination within divergent repeats or similar sequences [33 , 40] . This is especially relevant to Alu-Alu mediated junctions , where the region of perfect identity may not be extensive enough to employ homology-driven repair , but extensive base-pairing outside the region of identity could aid in driving a recombination coupled replication driven rearrangement process at these loci [34] . In the 16 patients with DUP-TRP/INV-DUP rearrangements presented , 2 contain a Jct2 breakpoint resulting in the formation of a chimeric Alu element . We hypothesize that the PMD-associated CGR are caused by BIR or MMBIR; these replicative processes have been shown to be error-prone , perhaps because they utilize a polymerase/replisome with reduced fidelity ( induced point mutations ) as well as reduced processivity ( template switching ) relative to intergenerational DNA polymerases [20 , 41] . Evidence now indicates that BIR/MMBIR-associated mutation results from conservative replication coupled with a migrating bubble [42 , 43] . Thus , DUP-TRP/INV-DUP CGRs involving PLP1 have the potential to additionally impact patient health through point mutations on the X chromosome . These hypotheses need further investigation through large-scale genomic sequencing . Nevertheless , although few in number , de novo point mutations apparently acquired concomitantly with the DUP-TRP/INV-DUP rearrangement in P250 ( insertion of an A ) and the quadruplication rearrangement in P113 ( insertion of a G ) were not seen in the corresponding , contiguous ( non-breakpoint containing ) section of the X chromosome for these intrachromosomal events , a finding consistent with observations made at the MECP2 locus and de novo mutation with CGR formation [20] . Given that on average , ~600 bp were sequenced at each junction , this suggests a rate of 2 mutations in ~15 kb of sequencing , consistent with the elevated point mutation rate observed in association with replication-based mechanisms of repair [20 , 41] . Junction 1 is present at seemingly identical loci , occurring within a complex inverted repeat structure in the 16 DUP-TRP/INV-DUP rearrangements studied . Further analysis has shown that at least one of these breakpoint junctions is in a region of 24 bp of microhomology and three occur within a >8 kb region of identity within A1a and A1b . The proposed mechanism for Jct1 is BIR within a region of ectopic , inverted homology [1] . Our data reveal that the template switch can occur within smaller regions of identity within A1a and A1b , suggesting that either MMBIR or homeologous recombination , rather than an homologous recombination within IRs may be an alternative mechanism for the formation of these seemingly recurrent junctions [22 , 31] . Previously , a study of 36 PMD patients identified 3 cases with duplicated copies of PLP1 inserted outside of Xq22 [24] . Conversely , in this study all 16 subjects with junctions in IRs contain the extra copy or copies of the gene on Xq22 , therefore suggesting that the mechanism of CNV results in a contiguous rearrangement ( triplicated or duplicated regions in tandem , Fig . 2 , Table 1 ) . Additionally , all 16 individuals queried by Southern blotting and/or qPCR methodologies indicate that the A1a and A1b inverted LCRs mediate PLP1 DUP-TRP/INV-DUP rearrangements . Although two other IRs in the region , albeit with less sequence identity ( the 93% identical outer C/D and 87% identical innermost A2/A3 repeats ) , could presumably mediate the junction between distal duplication and distal triplication breakpoints , these 16 cases use the A1a and A1b specific repeats . A1a and A1b are ~20 kb in length ( versus ~30 kb for C/D and ~10 kb for A2/A3 ) and are separated by ~50 kb ( versus ~140 kb for C/D and ~30 kb for A2/A3 ) . Therefore , the higher level of sequence identity between the A1a and A1b repeats ( ~99% ) , added to the shorter inter-repeat distance and the length of the LCR may both increase the likelihood of NAHR leading to the inversion [28 , 44] and potentiate these repeats as substrates for replication pausing , fork invasion , and reversal through BIR [31] . This is the second locus for which DUP-TRP/INV-DUP cases with recurrent Jct1 mediated by IRs has been described . In MECP2 DUP-TRP/INV-DUP , the K1 and K2 LCRs participate in both non-pathogenic inversions and the rearrangements present in patients [1 , 27] . Such empirical studies may enable refinement of current predictions for IRs that can predispose regions of the genome to DUP-TRP/INV-DUP [2] . Our data further implicate a “two-step process” of BIR paired with MMBIR to generate CGRs resulting in duplication of copy number sensitive genes proximal to IRs [1] . The rearrangements in the 16 patients with DUP-TRP/INV-DUP contain just two junctions that result in four copy number transition states . This complex pattern on array CGH is due to just two template switches , Jct1 occurring within the LCRs A1a and A1b distal to the PLP1 gene and resulting in an inversion , and Jct2 occurring at varying locations proximal to junction 1 and resuming the pattern of normal replication , resulting in a rescue from the potential formation of a dicentric chromosome ( Fig . 4 ) [45] . The observations at the quadruplication-containing CGR in P113 are consistent with rolling-circle amplification ( Fig . 5 ) . The rarity of quadruplication at PLP1 could be due to selective pressures from the increased severity of PMD with additional copies of PLP1 ( 4 versus 3 ) ; it is notable that the quadruplication observed herein does not include the dosage sensitive PLP1 gene [14] . One junction in this CGR ( between IRs A1a and A1b ) occurs at a similar location as in the DUP-TRP/INV-DUP structures , and PCR genotyping suggests that the interpretation of the rearrangement is complicated by the inversion structural variation , resulting in H2 ( S9 Fig ) . At the proximal junction , the fork template switches twice , invading upstream and leading to a rolling-circle [22 , 36–38] . After almost two complete copies of the circle ( 35 kb short of the overall 280 kb ) , the next junction is a template switch from the proximal end of the quadruplicated region to the distal end of the duplicated region within the LCR region . Our observations are most parsimoniously explained by a rolling-circle amplification event , as predicted for higher-order genomic segment amplification in the MMBIR model [22] . Due to the observations of: i ) triplicated and quadruplicated segments , ii ) the accompanying point mutation associated with CGR formation , and iii ) the prevalence of intrachromosomal rearrangements at this locus , a replicative model for CGR formation is likely [20 , 42 , 43] . The quadruplication-containing CGR provides evidence for an important next step in the MMBIR model , allowing for higher-order amplification to occur , as is often observed in cancer [22 , 46 , 47] . In summary , our studies confirmed a unique rearrangement product consisting of a DUP-TRP/INV-DUP structure in 16 individuals , with 6 containing triplication of PLP1 [1] . We also elucidated a common , recurrent inversion polymorphism between two IRs distal to this gene . Jct1 occurs between the same repeats that mediate the non-pathogenic inversion , and sequencing of a DUP-TRP/INV-DUP breakpoint within the LCRs showed that these junctions can occur within short stretches of identity within a larger repeat of ~20 kb . This study of breakpoint junctions involved in both DUP-TRP/INV-DUP and higher-order amplification leading to quadruplication implicate replicative mechanisms in the generation of these CGRs . Additionally , we provide experimental evidence supporting the contentions that: i ) IRs contribute to genome instability , ii ) LCRs can mediate replication-based mechanisms , and iii ) short repetitive sequences , such as Alu , can provide microhomology to facilitate template switching . The prevalence of DUP-TRP/INV-DUP events involving PLP1 brings attention to the importance of this mechanism and the potentially broader impact of this rearrangement structure in gene and genome evolution .
To determine whether there is a polymorphic inversion in the LCRs distal to PLP1 , we examined the genomic information for 9 individuals contained in the human genome structural variation ( HGSV ) track of the UCSC Genome Browser [8 , 25 , 26] . The HGSV track ( hg18 ) contains data on discordant fosmid end sequences from libraries of 9 individuals from diverse geographical regions . Discordant end sequence orientations of fosmids spanning LCRs A1a or A1b [23] and having both ends present in unique sequence ( not LCRs ) indicate potential inversions [26] . Individuals with at least one clone independently spanning each of the LCRs suggest that there is an inversion between the two repeats ( S1 Fig ) . Phased data from the 1000 genomes project [29] was used to create plots of two haplotypes in the region spanning from LCRs A1a to A1b ( Hg19 coordinates , ChrX:103223669–103324337 ) [48] . One thousand genomes data was cross-correlated with homozygous genotypes determined from Southern Blots to elucidate phased haplotypes that contain inversion alleles . Results were plotted using custom ( in-house ) scripts implemented in the R programming language ( S4 Fig ) . Families with PMD or rearrangements of Xq22 . 2 including PLP1 were obtained by physician referral or self-referral . Patients were enrolled through informed consent in research protocols approved by the Institutional Review Boards at Baylor College of Medicine ( BCM ) and the Nemours Alfred I . duPont Hospital for Children . The rearrangements present in patients BAB1290 , BAB1612/P374 , BAB2389 , P250 , P255 , P500 , P518 , and P558 were published previously [1 , 24 , 30] . Two of the patients with PLP1 triplication ( P518 and P674 ) were described as having more severe disease than patients with duplication [14] . Control DNAs from HapMap individuals [48] were obtained from the Coriell Institute for Medical Research cell repositories . Approximately 10 μg of genomic DNA from each patient was digested using BssSI . The DNA was diluted to 60 μl and digested for 4 hours at 37°C with 16U , heat inactivated at 80°C for 20 minutes , and the digest was then repeated with 12U for 3 hours and subsequent heat inactivation ( leading to a 10-fold overdigestion ) . The digested DNAs were then precipitated and concentrated using standard sodium acetate precipitation , and were reconstituted in 25 μl of water with gentle mixing overnight . Concentrations were determined using a NanoDrop spectrophotometer , and samples were then loaded along with an 8–48 kb ladder on a 0 . 6% Tris- Boric Acid-EDTA ( TBE ) gel and run in 1X TBE buffer for ~3 days at 50–60 volts . DNA restriction digestion products , i . e . bands on gels , were then visualized with ethidium bromide staining . Probe DNA was prepared using primers A1a proximal probe For- 5′-AATGCAGCTCAAAGGAAAGC-3′ and A1a proximal probe Rev- 5′-AGCCACTGACCAGTGATTTTC-3′ and amplifying a 514 bp fragment from BAC clone RP11–462K21 ( https://bacpac . chori . org ) DNA prepared using a QIAprep spin miniprep kit . The resultant PCR bands were resolved on 1% agarose and Tris-Acetate-EDTA gels and purified using a Zymoclean Gel DNA Recovery Kit ( Zymo Research ) . Probe DNA was validated by Sanger sequencing , using both forward and reverse primers and was frozen at-20°C in 90 ng aliquots . Southern Blotting was carried out as previously described [23] . Briefly , DNAs were subjected to electrophoresis for sufficient duration to distinguish 25 and 29 kb fragment sizes and were then transferred to a Sure Blot positively charged nylon membrane by standard ‘sandwich’ methodology for 2–3 days . Approximately 80 ng of DNA was labeled with 32P-dCTP by random priming for 2–4 hours at 37°C using the Random Primed DNA labeling kit ( Roche ) . Membranes were pre-hybridized for 4 hours in 10% dextran sulfate/1M NaCl/1%SDS ( hybridization solution ) with 4mg of sheared salmon sperm DNA at 65°C . Probe was pre-associated in hybridization solution with ~1mg sheared placental DNA at 65°C for ~2 hours , then added to the pre-hybridized membrane . Hybridization was carried out at 65°C overnight ( ~18 hours ) . The following day , the blot was washed and analyzed using autoradiography for bands corresponding to PLP1 A1a structural haplotype information ( ~25 and 29 kb ) . To determine the size , genomic content , and extent of PLP1 rearrangements , a high-density oligonucleotide array from Agilent was custom-designed to examine PMD patients . The 4 x 44 K microarray was designed using the Agilent eArray website ( https://earray . chem . agilent . com/earray/ ) and was used to visualize the rearrangements of three patients in this study ( BAB1290 , BAB1612/P374 , and BAB2389 ) , the family containing individuals BAB3698 , BAB3699 , BAB3700 , and BAB4179 , and to complement existing array data for P500 , P1407 , and P113 . The family of P113 was explored using Agilent arrays , including patients P113 and P117 , as well as the mother of P113 , P154 , and grandmother , P088 , who are both carriers . Probe labeling and hybridization were conducted as previously described , with NA15510 and NA10851 used as reference DNAs for female and male individuals , respectively ( Accession GSE63594 ) [1] . Purified DNA samples from P113 , P250 , P500 , P518 , P558 , P642 , P674 , P820 , P842 and P1150 were submitted to NimbleGen for array service with normal male control NM002 as a reference . The NimbleGen X chromosome CGH fine-tiling array with oligonucleotide probes of 45 to 85 bases in length with median spacing of 106 bp throughout the whole X chromosome was used . Patient DNA samples P255/P298 , BAB1612/P374 , P1389 and P1407 were submitted to the Biomolecular Core Lab at duPont Hospital for Children for hybridization to Affymetrix Cytogenetics 2 . 7M Array . DNA sample P1407 was submitted to Coriell’s Genotyping and Microarray Center for hybridization on Affymetrix Genome Wide Human SNP Array 6 . 0 . All data from Affymetrix arrays were analyzed with GeneChip Command Console Software AGCC . NimbleGen and Affymetrix Cytogenetics array data were aligned with qPCR data and plotted using the R programming language ( Affymetrix and NimbleGen data are under Accession GSE64122 ) ( Figs . 2B , C , D and S5 ) . Semi-quantitative multiplex PCR was performed using a QIAGEN Multiplex PCR kit according to the manufacturer’s protocol to analyze regions on the X chromosome in and surrounding PLP1 to determine copy number . Primer pairs were selected using the NCBI primer design tool ( primers available upon request ) . In each experiment , five control DNAs were used , two known to have duplications in the region of interest without CGRs and three normal controls known to be single copy in the region of interest . A primer pair that amplifies a region of the human dystrophin ( DMD ) gene on the p-arm of the X chromosome was included in each multiplex reaction for amplification of a single-copy region . Products were separated by electrophoresis on a 4% NuSieve 3:1 agarose gel ( Lonza , Walkersville MD ) and stained with ethidium bromide . Net intensity of each band was determined using a Molecular Imaging system with Kodak Gel Logic Imaging software or AlphaImager HP . Copy number was determined by calculating the ratio of the net intensities of bands in the test region to dystrophin single-copy region for each DNA sample and then normalized by dividing by the average of the ratios of test region to dystrophin of the three normal controls . Theoretical ratios were: one , single-copy; two , duplication; three , triplication . Alternatively , quantitative PCR was performed as above except that one primer of each pair was labeled with 6-FAM and samples were submitted to the Biomolecular Core Lab at duPont Hospital for Children for capillary electrophoresis on an ABI PRISM 3130XL DNA Analyzer . Analysis of copy number was determined as above , by using area under the peak as determined by Peak Scanner software rather than net intensity . Triplicated , quadruplicated and duplicated regions were mapped to within several kb of their endpoints using these methods . Interphase nuclei and metaphase chromosomes were prepared from 700 μl of whole blood stored in sodium heparin Vacutainer tubes as follows . Blood was placed in α-MEM , 20%FBS , 1% L-glutamine , 50 μg/ml gentamycin and treated with 150 μl Phytohemagluttenin ( Invitrogen , Carlsbad CA ) . The cultures were incubated at 37°C for 72 hours in upright position after which they were treated with 100 μl colcemid by trituration followed by incubation at 37°C for 30 min . Cultures were then subjected to centrifugation at 350 x g for 6 minutes . The supernatant liquid was discarded and the pellet was suspended in 10 ml 75mM KCl pre-warmed to 37°C and incubated at 37°C for 15 minutes . Then 1 ml of fixative ( 3:1 mixture of methanol:acetic acid ) was added slowly . The preparation was washed 3 times in 10 ml of fixative with pelleting by centrifugation at 350xg for 6 minutes after washes . The resulting pellet of interphase and metaphase chromosomes was then stored in fixative at -20°C . Chromosomes and nuclei were dropped onto pre-cleaned Fisherbrand slides in a CDS-5 Glovebox environmental Chamber , ( Thermotron , Holland Michigan ) set at 25°C and 50% humidity . Slides were stored at -20°C in a vacuum under dessication until use . FISH was performed using cosmid clone U125A1 and BAC clone RP13–188A5 obtained from the BACPAC resource center . Cosmid and BAC DNAs were isolated using the QIAGEN Plasmid purification and QIAGEN Large-Construct kits , respectively . One μg of U125A1 DNA was labeled with Biotin-16-dUTP and 1μg RP13–188A5 DNA was labeled with digoxigenin using the DIG-Nick Translation Mix . Labeled probes were purified using Nuctrap probe purification columns according to the manufacturer’s protocol . After hybridization to chromosomes and nuclei according to standard protocol , biotinylated U125A1 was bound to Cy3-labeled streptavidin , further amplified with biotinylated antiavidin ( Vector Laboratories , Burlingame CA ) and detected with a second layer of Cy3-labeled streptavidin . Simultaneously , RP13–188A5 labeled with digoxigenin was coupled with mouse antidigoxigenin , detected with rabbit anti-mouse FITC ( Jackson ImmunoResearch Laboratories ) and further amplified with goat anti-rabbit FITC antibody . Nuclei and chromosome spreads were counterstained with 4’ , 6-diamidino-2-phenylindole ( DAPI ) and cover slips were mounted using Vectashield antifade solution . Images were captured using a Leica DM RXA2 fluorescence microscope and Openlab imaging software ( Perkin Elmer , Waltham MA ) . To examine whether an intra- or inter-chromosomal origin occurred for the extra genomic segments in each patient’s genome , we analyzed 2 STRs and 9 single SNPs within the duplicated/triplicated region common to most patients . Sites were chosen based on marker genotypes displaying a high degree of heterozygosity in HapMap samples . S6 Table depicts the dbSNP identifiers , locations with respect to Chromosome X sequence NT_011651 . 17 , and primers used to amplify the SNP or STR . Regions of interest were amplified from DNA with HotStar Taq DNA polymerase ( Qiagen ) for products <1kb or Expand High Fidelity PCR system ( Roche ) for products >1kb . Patients P250 , P255/298 , BAB1612/P374 , P500 , P518 , P558 , P642 , P674 , P820 , P842 , P1150 , P1389 , and P1407 were analyzed . Products containing the STR were amplified with one primer within the pair fluorescently-labeled so product sizes could be evaluated by capillary electrophoresis on ABI’s PRISM 3130 XL DNA Analyzer . Products containing the SNP were purified using QIAquick PCR or Gel purification kits , then sequenced with the Big Dye Terminator kit v . 3 . 1 ( Life Technologies ) , according to the manufacturer’s instructions . Patients P113 , P117 and carriers P154 and P088 were subjected to genotyping using the Illumina OmniExpress SNP array analyses at the human genome sequencing center of BCM . Data from the analyses was visualized by plotting the B allele frequencies versus the X chromosome coordinates encompassing the quadruplication genomic rearrangement , as well as the Log ratio of SNP intensity ( S11 Fig ) . Inverse PCR was used to obtain the first junction of patient P1150 . Briefly , DNA was digested with NheI and ligated to form circles . PCR primers were designed to amplify in opposite directions around the circle by long-range PCR using the Expand High Fidelity PCR system ( S6 Table ) . When the PCR products were analyzed on an agarose gel , a product was found that was unique to the patient . The product was subjected to DNA sequencing according to the manufacturer’s instructions , then purified with the Filtration Cartridge ( Edge Biosystems , Inc . , Gaithersburg MD ) and separated using an ABI PRISM 3130 xl Genetic Analyzer . DNA sequence was analyzed using Vector NTI sequence analysis software . Proximal junctions were obtained for the personal genomes from the remaining triplication patients by long-range PCR using appropriately positioned primers at the endpoints of copy number changes ( Table 1 and S6 Table ) in 25 μl reactions with 50–100 ng of patient DNA using TaKaRa LA Taq or using the Expand High Fidelity PCR dNTPack kit according to the manufacturers’ instructions . PCR products were prepared for sequencing by using the standard ExoSAP-IT protocol ( Affymetrix , Santa Clara CA ) or by using the Qiagen PCR purification kit and DNA sequencing reactions were performed as indicated above using primers used in amplification or internal primers as indicated in S6 Table . Sequences were aligned to the human genome reference sequence , and breakpoints are depicted in S6 Fig . We had previously reported the sequence across the junction in P255 [24] . PCR was conducted across Jct1 from DNAs prepared from patients with DUP-TRP/INV-DUP CGRs using a QIAGEN Multiplex PCR . Two control DNAs duplicated through this region and three control DNAs with a single copy at this locus were amplified in parallel . Along with a dystrophin primer pair ( Hdys 23F-6FAM and Hdys 23 R ) for a single copy region of the human dystrophin gene , we used primers pairs V362H12-F19–6FAM and V362H12-R19 ( red arrows ) , and V362H12-F24–6FAM and V362H12-R24 ( black arrows ) , and V362H12-F19–6FAM and V362H12-R24 ( one red , one black arrow ) ( Figs . 3E , S8 , S6 Table ) . Fluorescently labeled PCR products were diluted 1:100 in sterile HPLC water and subjected to capillary electrophoresis using an ABI PRISM 3130 XL DNA Analyzer . Copy number analysis was performed as previously described using the Peak Scanner software [24] . To subclone breakpoints in PMD DUP-TRP/INV-DUP patients , we amplified patient DNAs containing rearrangements ( from BAB1612/P374 , BAB2389 , and BAB1290 ) with PCR primers that anneal within the A1a and A1b LCRs and uniquely flanking primers . This yielded four overlapping segments of the two LCRs ( S6 Fig ) . These PCR products were then subjected to electrophoresis in crystal violet 0 . 8% agarose gels , purified using the SNAP purification kit from Invitrogen , and cloned into TOPO XL cloning vectors . Resultant clones for each of the four segments were screened by digestion and sequenced in their entirety . At least two clones for each region , obtained from independent PCR reactions , were screened for the breakpoint and the corresponding A1a or A1b region . Sequence analysis was conducted using the Lasergene 9 DNA analysis software suite . Copy number of junctions in the quadruplication patients and a carrier were determined by dPCR using QuantStudioTM 3D Digital PCR System ( Life Technologies ) , according to the manufacturer’s instructions . Concentration of DNA was determined by QubitR dsDNA BR assay ( Life Technologies ) using the Qubit 2 . 0 fluorometer ( Life Technologies ) . Sample DNA was digested with SphI ( NEBiolabs ) to separate multiple copies of interest that may be located on the same molecule without disrupting the region of amplification . Digests were performed using 400 ng of DNA in a 10 μl reaction containing 10U of SphI and incubating at 37°C for 1 . 5 hr , followed by heat-inactivation of the enzyme at 65°C for 20 min . The digest was diluted to 40 μl with RNase-free water to yield a concentration of 10 ng/μl DNA . Primers and probes used in the dPCR assays are in S6 Table . Reactions for dPCR included 1x QuantStudioTM 3D Digital PCR Master Mix , 1x TaqMan Copy Number Reference Assay for human RNaseP ( Life Technologies , Cat . # 4403328 , VIC label ) , 1–1 . 5x PrimeTime qPCR 5’ nuclease assay ( IDT , FAM label ) for jct1 or jct2/3 and 40–60ng DNA in a 16 μl volume . Fifteen μl of this mix was used to load the Digital PCR 20K Chip ( Life Technologies ) ; chips were processed according to the manufacturer’s instructions . | Genomic architecture , such as direct or inverted repeats , can facilitate structural variation ( SV ) of the human genome . SV can consist of deletion , duplication , or inversion of a genomic segment , or combinations thereof , the latter referred to as complex genomic rearrangements ( CGR ) . CGR are defined as requiring two or more novel DNA breakpoint junctions . We described a CGR product at the MECP2 locus with an unusual pattern consisting of an inverted triplicated segment flanked by duplicated segments of the genome . This complex CGR is facilitated by inverted repeats in a process that mechanistically could occur by two template switches mediated by replicative DNA repair . We now investigate the PLP1 locus and demonstrate that 16/17 CGR independent events present with duplication—inverted triplication—duplication pattern facilitated by two inverted repeats , similar to events involving MECP2 . We show that the same inverted repeats facilitating CGR formation are also responsible for an inversion polymorphism observed frequently in the normal population . Intriguingly , one CGR was found to have a quadruplication resulting in the presence of four copies of a genomic segment . Breakpoint studies suggest this quadruplication occurred in a manner consistent with rolling circle amplification as predicted by previously postulated models . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Complex Genomic Rearrangements at the PLP1 Locus Include Triplication and Quadruplication |
A critical feature of Mycobacterium tuberculosis , the causative agent of human tuberculosis ( TB ) , is its ability to survive and multiply within macrophages , making these host cells an ideal niche for persisting microbes . Killing the intracellular tubercle bacilli is a key requirement for efficient tuberculosis treatment , yet identifying potent inhibitors has been hampered by labor-intensive techniques and lack of validated targets . Here , we present the development of a phenotypic cell-based assay that uses automated confocal fluorescence microscopy for high throughput screening of chemicals that interfere with the replication of M . tuberculosis within macrophages . Screening a library of 57 , 000 small molecules led to the identification of 135 active compounds with potent intracellular anti-mycobacterial efficacy and no host cell toxicity . Among these , the dinitrobenzamide derivatives ( DNB ) showed high activity against M . tuberculosis , including extensively drug resistant ( XDR ) strains . More importantly , we demonstrate that incubation of M . tuberculosis with DNB inhibited the formation of both lipoarabinomannan and arabinogalactan , attributable to the inhibition of decaprenyl-phospho-arabinose synthesis catalyzed by the decaprenyl-phosphoribose 2′ epimerase DprE1/DprE2 . Inhibition of this new target will likely contribute to new therapeutic solutions against emerging XDR-TB . Beyond validating the high throughput/content screening approach , our results open new avenues for finding the next generation of antimicrobials .
About one third of the world's population is estimated to be infected with Mycobacterium tuberculosis . In nine out of ten cases , M . tuberculosis persists in a latent state throughout an individual's lifetime [1] . The bacillus is found in a variety of host cells such as alveolar macrophages , dendritic cells and type II alveolar pneumocytes in infected lungs [2] , [3] , [4] , as well as in adipocytes [5] . Whereas dendritic cells and adipocytes are not permissive for in vitro growth , M . tuberculosis replicates actively in macrophages and type II alveolar pneumocytes [2] , [3] , [5] , [6] . The ability of M . tuberculosis to survive and multiply within host cells certainly contributes to the pathogenesis of tuberculosis ( TB ) . Though the exact means of ensuring intracellular survival is still a matter of debate [7] , [8] , [9] , it is clear that potential new anti-tuberculosis drugs have to be active against M . tuberculosis inside host cells [10] . As this feature is not normally taken into account in traditional drug-screening procedures at an early stage , we developed a target-free cell-based assay suitable for high throughput screening that enables an unbiased search for compounds that kill intracellular M . tuberculosis without affecting the viability of the host macrophage . Such molecules would then serve as tools to identify novel druggable mycobacterial targets . Target-based screens for antimicrobial agents have been disappointing to date [11] , [12] whereas whole cell-based approaches with M . tuberculosis are fraught with logistic difficulties and hampered by long incubation periods . In this study , we developed a rapid phenotypic assay based on the use of automated confocal fluorescent microscopy to monitor intracellular growth of GFP-expressing M . tuberculosis H37Rv in Raw264 . 7 macrophages . The assay was set-up for the high throughput screening ( HTS ) of large chemical libraries in 384-well format and its robustness was validated with known antibiotics . By screening several thousand small molecules , new series of compounds were identified as well as some sharing structural similarities with known TB drugs . Among these , the benzamide series was then used as a bait to identify a new putative target . Using a combination of biochemical assays and genetic approaches , we showed that nitrobenzamide derivatives inhibited arabinan synthesis , which has not been observed for any of TB drugs so far . Altogether , these results demonstrate the feasibility of large scale screen for intracellular M . tuberculosis growth and open new avenues for enriching the TB drug pipeline as well as for finding new druggable targets .
To set up the optimal conditions of M . tuberculosis infection , Raw264 . 7 macrophages were first infected with mycobacteria that constitutively express green fluorescent protein ( GFP ) using different multiplicities of infection followed by kinetic analysis of intracellular bacterial growth . Confocal images of live samples were acquired using an automated confocal microscope ( Opera™ ) over 7 days ( Figure 1A ) . During the first twenty-four hours , a few discrete weakly fluorescent bacteria localized within the cells . At day 2 , the average number of cells had increased and mycobacteria had started to spread into neighboring cells leading to zones of strongly fluorescent bacteria . At day 3 , the number of cells had significantly diminished and the bacteria formed large , highly fluorescent aggregates , which covered almost the entire image from day 5 onwards . As a control , non-infected cells grew to confluence at day 2 and remained alive until day 5 . Customized image analysis was developed to automatically quantify several different parameters such as the number of host macrophages , the percentage of infected cells and the average surface area of bacterial aggregates [13] , [14] . Representative results of the cell segmentation method are displayed on Figure 1B . After two hours of infection , between 2 and 10% of Raw264 . 7 cells were found to harbor intracellular bacilli ( Figure S1A ) . The percentage of infected cells steadily increased reaching 50% by day five with a MOI of 1 . This augmentation correlated with substantial macrophage mortality due to the known cytopathogenic effects of M . tuberculosis [2] ( Figure S1B ) . From day 5 to day 7 , the percentage of infected cells continued to increase slowly up to 70%; however , the cell number dramatically decreased . Therefore to ensure that a sufficient number of cells was recorded in each field were recorded , the incubation time with M . tuberculosis was set at 5 days for the next series of experiments . To validate the system , we first tested the effect of the standard anti-tuberculosis drugs such as isoniazid ( INH ) and rifampin ( RIF ) in our model ( Figure 1C ) . As expected , these drugs demonstrated a dose-dependent decrease in both bacterial load and percentage of infected macrophages . Interestingly , an increase in host cell number was seen at effective concentrations , clearly demonstrating the ability of these drugs to prevent M . tuberculosis induced cytopathogenicity ( Figure 1D , E ) . Taking into account both the amount of green fluorescent bacteria and the host cell number , this assay enables a dual and independent determination of intracellular anti-mycobacterial drug efficacy . In addition , the reference drugs were also applied against M . tuberculosis H37Rv-GFP grown in liquid broth without host cells . No major differences in minimal inhibitory concentrations ( MIC ) between the two assays were noticed for INH ( Figure 1F ) , ethambutol , ethionamide and PA-824 [15] ( Table 1 ) , whereas RIF was 100-fold less efficient in the cell-based assay ( Figure 1G ) , confirming the previously reported reduced activity of RIF against intracellular bacteria [16] . This finding clearly demonstrates that our dual read-out , cell-based drug screening system indeed measured the intracellular antibacterial activity of drugs , which allowed us to further adapt the system for High Throughput/Content Screening . A diverse library of 56 , 984 synthetic compounds was first screened at a single concentration . A normal distribution of the compounds was obtained using PCA-1x analysis ( Figure 1H ) . 486 fully active hits were then confirmed by means of serial dilution experiments . The MIC of each hit was then determined using both the percentage of infected cells and the total cell number by taking advantage of the dual visual effect described above as an independent confirmation of compound activity . More than one-quarter of the hits ( 135 hits ) had an MIC less than 5 µM , and 8% had a MIC below 1 µM , which is equivalent to that of INH ( Figure 1I ) . A few compounds , such as compound CPD1 , showed cytotoxicity at high concentrations as seen by a significant decrease in the cell number above 5 µM ( Figure S2C ) . Chemo-informatic cluster analysis of the 135 hits was performed and hits fell into 9 clusters plus 13 singletons ( Table S1 ) . The largest cluster had 69 members with an isonicotinohydrazide moiety similar to that of INH , used as a positive reference in our assay , which validated our approach . The second largest cluster of 24 derivatives shares a common benzamide scaffold . As no antimycobacterial effect had been previously reported for this particular chemical structure , a series of related derivatives was synthesized for further studies . To identify the chemical substituents necessary for benzamide antibacterial activity , over 155 additional derivatives were synthesized and their structure-activity relationship was analyzed using both our intracellular assay and the in vitro growth assay . The most potent compounds exhibited substitutions of the benzene moiety with a nitro group at positions 3 and 5 ( Figure 2 and Figure S3 ) . The reduction of one nitro- to hydroxylamine and amino groups led to totally inactive compounds . In contrast , derivatives with an N-substitution by benzyloxy-ethyl or by phenoxy-ethyl showed enhanced activity with an MIC below 0 . 2 µM . More importantly , cyclic-benzamides had an MIC below 80 nM in the in vitro assay . However , these compounds turned out to be much less potent in the intracellular assay . Furthermore , substitution of the benzyloxy moiety by a chlorine- or fluorine atoms at position 3 led to increased potency in both assays in contrast to carboxyl substitutions . In parallel , we selected two compounds , N- ( 2- ( 4-methoxyphenoxy ) ethyl ) -3 , 5-dinitrobenzamide ( DNB1 ) and N- ( 2- ( benzyloxy ) ethyl ) -3 , 5-dinitrobenzamide ( DNB2 ) , for further mechanistic studies and target identification ( Figure 2 ) . DNB1 and DNB2 were pursued further since their activities on intra-cellular and extra-cellular M . tuberculosis were particularly favorable ( Figure 3A–C and Figure S3 ) . Their effects on primary macrophages were further determined . Host cells that had been pre-incubated with DNB1 harbored fewer bacteria compared to the DMSO control , and were more abundant at day 7 of infection as shown in Figure 3D . Conventional CFU determination was then performed after seven days of infection to quantify the remaining bacterial load . More than a ten-fold decrease in the number of CFUs was observed with both human and mouse primary cells at a DNB1 concentration above 5 µM ( Figure 3E ) . Similar data were obtained for DNB2 ( data not shown ) . This confirms the potency of this series of compounds . In parallel , no cell toxicity was noted for these compounds using conventional cytotoxicity assays of uninfected cells , indicating that our high content assay can reliably predict cytotoxicity ( Table S2 ) . Analysis of the broad antimicrobial spectrum was undertaken and revealed that the effect of these dinitrobenzamide derivatives was mainly restricted to actinomycetes with the most potent activity observed against Mycobacterium with an MIC of 75 ng/mL ( 0 . 2 µM ) ( Table S2 ) . Of particular importance , DNB1 and DNB2 were also highly active against multidrug-resistant ( MDR ) and extensively drug-resistant ( XDR ) clinical isolates . Moreover , these two compounds were also associated with low levels of spontaneous resistance . Resistant mutants arose at frequencies between 1 . 2×10−6 and 1×10−8 on agar containing 2–16× the MICs of DNB1 or DNB2 , a frequency similar to that with INH ( Table S3 ) . The potential for the development of resistance to dinitrobenzamides , in vitro , is therefore analogous to major anti-TB drugs . Interestingly , the bactericidal effect on M . tuberculosis of DNB1 and DNB2 was found to be time-dependent ( Figure S4A ) and to require several days to reach bacterial clearance , implying that they could interfere with de novo mycobacterial component biosynthesis . This is further corroborated by the fact that the DNB compounds lost their activity in a non-replicating M . tuberculosis system [17] . Altogether these results suggested that the DNB compounds might act on different targets than current antituberculosis compounds . To gain insight into the possible targets of dinitrobenzamides , we investigated the effect of DNB1 and DNB2 on the lipid composition of the cell envelope of M . tuberculosis; no effects on the biosynthesis of fatty acids , mycolic acids and/or other lipids were noted ( data not shown ) . By contrast , DNB1 and DNB2 showed a clear-cut effect on the synthesis of the arabinan domains of arabinogalactan ( AG ) and lipoarabinomannan ( LAM ) ( Figure S4B , C ) . Decaprenyl-phospho-arabinose ( DPA ) is the only known arabinofuranose ( Araf ) donor in the biogenesis of AG and LAM in mycobacteria and is thus an essential precursor [18] , [19] . To determine whether the effects of DNB were attributable to the inhibition of the synthesis of DPA or to that of DPA-dependent arabinosyltransferases involved in the elongation of both heteropolysaccharides , we set out to monitored DPA formation in treated and untreated extracts of M . smegmatis mc2155 . Analyses revealed complete inhibition of DPA formation in the DNB-treated extracts concurrent with the accumulation of decaprenyl-phospho-ribose ( DPR ) ( Figure 3F ) , indicating that the target of both DNB inhibitors is probably the heteromeric decaprenyl-phospho-ribose 2′ epimerase encoded by the rv3790c ( dprE1 ) /rv3791c ( dprE2 ) genes in M . tuberculosis H37Rv [20] . DprE1 has been recently described as the target of benzothiazinones ( BTZ ) , a new class of antitubercular unrelated nitro-compounds [21] . BTZ- resistant mutants of M . smegmatis and M . bovis BCG were isolated and characterized as having a mutation in dprE1 , in which Cysteine 387 had been replaced by a Glycine residue . This led us to test these dprE1 mutants for their sensitivity to DNB1 and DNB2 ( Tables S4 , S5 ) . They all displayed resistance to the DNB compounds corroborating our biochemical data ( Figure 3F ) . These findings demonstrate the remarkable intracellular vulnerability of DprE1 and highlight the importance of pursuing the route of DPA production as a drug target .
Although the location and state of latent bacteria remains a matter of debate [22] , one commonly shared hypothesis for mycobacterial persistence is that M . tuberculosis bacilli are able to survive in macrophages for prolonged periods of time and , unlike other bacteria , are able to actively replicate . It has clearly been established that the tubercle bacillus adopts a different phenotype in the host macrophage's phagosome compared to growth in extracellular conditions [7] , [8] . The intraphagosomal transcription profile of M . tuberculosis is complex; a large variety of genes are over-expressed and temporally regulated in response to environmental cues . Altogether , this makes the identification of one specific factor in the tubercle bacillus that could be selected as the ideal drug target difficult . Consequently , non-target cell-based assays have emerged as a critical tool in the search for intracellular M . tuberculosis inhibitors . Identification of antimycobacterial inhibitors active within host cells has long been limited due to cumbersome CFU plating , slow bacillary growth , safety requirements and difficulties in setting-up appropriate infection conditions . As a consequence , this approach was always used as a secondary assay after the initial selection of compounds that are active on broth grown bacteria . With the advent of automated confocal microscopy , the above-mentioned limitations could be circumvented and here we demonstrate the feasibility of large scale compound screening . To minimize the steps and to cope with HTS requirements , we performed suspension macrophage batch infection . To this end , careful attention was paid to remove the extracellular non-phagocytosed mycobacteria through the use of judicious centrifugation conditions and amikacin . Mycobacteria are able to grow independently of host cells and consequently any remaining extracellular bacilli would greatly compromise the validity of our model . Consequently , an extra amikacin treatment step was added to the protocol to further eliminate any remaining mycobacteria . Additional washing steps after amikacin treatment removed the antibiotic thereby minimizing the introduction of any bias towards compounds that could act synergistically with amikacin during screening . Thus with the optimized protocol , there are almost no non-phagocytosed mycobacteria left by the time compounds are added . Our results demonstrate that our assay specifically measured the effect of compounds on intracellular mycobacteria . Indeed , we observed weak inhibition with rifampin , an antibiotic that is known to be poorly active on intracellular mycobacteria . The reproducible 100-fold decrease in MIC for rifampin in the intracellular assay compared to the in vitro growth assay proved that the targeted bacteria are not extracellular . Otherwise no difference would have been seen in MIC between the two assays . As is well established and as we confirmed , macrophages are able to support high bacterial loads , which occupy a large part of the cell cytoplasm , eventually leading to macrophage cell death . Taking this into account , it was decided to set the data acquisition at day 5 post-infection when the cell number in the DMSO-control samples had significantly decreased relative to the antibiotic protected controls . Thus , monitoring cell number was an additional parameter enabling us to confirm the compound's antibacterial activity . This confocal imaging-based assay could likely be adapted to other type of cells such as non-phagocytes in which M . tuberculosis is known to reside [3] , [6] . Firstly , one could envision searching for drugs that will be active in different host settings . Secondly , our system could be adapted to the screening of compounds that target mycobacterial granulomas as these multi-cellular structures can be generated in vitro [23] and have been shown to promote infection [24] . One of the current challenges for TB drug discovery is the identification of compounds that are active against MDR and XDR bacteria . Compound-based approaches have lately proven to be effective for the development of new antitubercular drugs and have identified compounds with new mechanisms of actions such as TMC207 [25] , [26] . The library of compounds that we screened contained more than 1 , 500 different heterocycles and was initially designed to be unbiased . This led to the identification of 23 clusters of molecules , among which the only known anti-tubercular compounds were INH derivatives . However , screening of another library led to a list of another set of hits including analogs of the nitroimidazopyran PA-824 [15] ( data not shown ) . Also , our set of compounds does not include the typical chemical structures of some common antibacterials such as rifampin and streptomycin , which do not meet the Lipinski criteria on size and lipophilicity used for our library selection [27] . Taken together , this showed that the 57 , 000 member library does not contain the full repertoire of active small molecules . The step-like shape of the dose-response curves ( DRC ) resulting from this cell-based phenotypic assay is unusual compared to the classical sigmoid profile of DRC for in vitro enzymatic or ligand-receptor type based assays . However , we can clearly rule out possible artifacts such as precipitation for several reasons . Firstly , our assay was calibrated with known TB inhibitors such as INH and ethambutol , which are water soluble and the curves obtained with these compounds displayed step-like shape . Secondly , the classical sigmoid with Hill coefficient value around 1 correspond to a fitting equation whose parameters are based on a model relying on the interaction of one unique substrate/ligand with one enzyme/receptor . In contrast , in our phenotypic assay , a large number of proteins are likely to be involved in the inhibition process , which may require compound intracellular uptake , pro-drug activation and target inhibition . In this system , the classical sigmoid model may not be the best fitting model . Thirdly , a similar step-like shape is observed for classical microbiological assays on whole mycobacterium such as for the rezasurin reduction assay . Thus , determination of MIC values as used for conventional TB drug susceptibility testing turned out to be more appropriate than half maximal inhibitory concentration IC50 measurement . Structure-activity relationship studies were thus undertaken on the benzamide scaffold , which initially contained the largest number of molecules identified from the screen after the INH-like molecules . Cyclic-benzamides showed a 200-fold diminished intracellular growth inhibitory effect relative to its in vitro antibacterial effect , thereby demonstrating that compounds have to be efficiently taken up by cells to be effective against the intracellular bacillus . This intracellular assay may thus prove to be suitable for counter selection of compounds that have impaired membrane uptake . Further development of the DNB series into lead compounds active in vivo requires improvement of their pharmacokinetics properties . Indeed , we observed that the nitro groups that are necessary for DNB antimycobacterial activity were very rapidly reduced into an amino group by mammalian liver enzymes . The mean half-life of the most active DNB compounds was about 8 minutes in a mouse microsomes assay and could already be significantly increased using encapsulation within nanoparticles . In a preliminary experiment using the acute mouse model of M . tuberculosis , a one log reduction of the CFU in the lungs of DNB treated animals compared to non-treated controls was observed after a three week daily treatment with 30 mg/kg/day following an intranasal infection ( data not shown ) . Additional optimization of ADME properties and in vivo delivery of the DNB compounds is currently in progress . Strikingly , chemical genomics identified decaprenyl phospho-ribose 2′-epimerase as the main target of dinitrobenzamides . This epimerase is encoded by dprE1 and dprE2 genes that are adjacent to the embA-C gene cluster whose products are also involved in the biosynthesis of LAM and AG and are targets of the first-line drug , ethambutol [28] . Consistent with their involvement in the synthesis of DPA , Rv3790c ( DprE1 ) and Rv3791c ( DprE2 ) have been suggested to be essential for the in vitro growth of M . tuberculosis as determined by transposon site hybridisation ( TraSH ) [29] . Interestingly , the fact that potent inhibitors of DprE1 could directly be isolated from the primary screening may indicate that this target is not only essential for bacterial growth inside the macrophages but also is easily accessible to small molecules . Though it is evident that the presence of the DNB scaffold in our library largely contributed to the identification of DrpE1 as a very druggable target , screening of another non-biased library could have resulted in similar findings . This is supported by the fact that as part of an independent study , DprE1 was recently identified as the target of benzothiazinones ( BTZ ) , a different class of compounds that show potent antimycobacterial activity [21] . Moreover , BTZ-resistant mutants all displayed cross-resistance indicating that two chemically un-related nitro-compounds probably inhibit DPA production by the same mechanism . Further biochemical analyses will definitely contribute to a better understanding of the pharmacology of this new druggable mycobacterial target . The mechanisms of action of the other scaffolds found in this study remains to be characterized and will likely contribute to the discovery of new bacterial as well as cellular targets . For example , derivatives from Scaffold IX ( Table S1 ) are effective against XDR isolates and have no effect against DprE1 activity and ATP synthesis , which suggests that they may act on an unknown target . In addition , molecules sharing Scaffold III displayed selective inhibition of intracellular growth within macrophages , raising the possibility that a host cellular target could be involved in the antibacterial effect . Alternatively , using other libraries could lead to the identification of scaffolds with different chemical structures . For instance , screening another set of 120 , 000 molecules in our cell-based assay revealed analogs of the nitroimidazopyran PA-824 [15] , which was shown to induce bacterial killing by nitric oxide release [30] . Taken together this clearly shows that both the repertoire of druggable targets and potential antitubercular compounds has not yet been fully uncovered . Finally , we would like to point out that high throughput/content screening is a powerful generic approach that can be used to discover inhibitors for other intracellular pathogens that are genetically tractable .
The 56 , 984-compound library was purchased from Timtec ( 26 , 500 molecules ) , Cerep ( 10 , 484 ) and ChemBridge™ ( 20 , 000 ) and each sub-library consisted of a selection of molecules based on their chemical diversity and drug-like properties . An in-house evaluation showed that ≥80% of the compounds met the criteria of the ‘rule of 5’ of Lipinski [31] . Small molecules from the screening libraries , CPD1 , N- ( 2- ( 4-methoxyphenoxy ) ethyl ) -3 , 5-dinitrobenzamide ( DNB1 ) and N- ( 2- ( benzyloxy ) ethyl ) -3 , 5-dinitrobenzamide ( DNB2 ) were dissolved in pure DMSO ( Sigma , D5879 ) and added to the assay plates using an EVObird liquid handler ( PerkinElmer ) to reach a final concentration of 20 µM . The description of all the mycobacterial strains used in this study is given in Table S6 . Mycobacterium tuberculosis H37Rv , H37Ra and BCG Pasteur were used as reference strains . The recombinant strain of M . tuberculosis H37Rv expressing the green fluorescent protein ( H37Rv-GFP ) bears an integrative plasmid ( based on Ms6 ) carrying a gfp gene constitutively expressed from the promoter pBlaF [32] . All strains were precultured at 37°C in Middlebrook 7H9 broth ( Difco ) supplemented with 0 . 05% Tween 80 ( Sigma , P8074 ) and oleic acid-albumin-dextrose-catalase ( OADC ) for 14 days . 384-well plates ( Greiner , #781091 ) were first preplated with 0 . 5 µl of compound dispensed by EVOBird ( Evotec ) in 10 µl of Middlebrook 7H9-OADC medium supplemented with 0 . 05% Tween 80 . Forty microliters of H37Rv-GFP bacterial suspension diluted to 2×106 CFU/mL ( based on GFP fluorescence assessment and a reference curve ) was then added to the diluted compound resulting in a final volume of 50 µl containing 1% DMSO . Plates were incubated at 37°C , 5% CO2 for 7 days . Mycobacterial growth was determined by measuring GFP-fluorescence using a Victor 3 reader ( Perkin-Elmer Life Sciences ) . The resazurin reduction method was used for reference strains , MDR , XDR and clinical isolates [33] . Isoniazid at 0 . 05 µg/mL and 1 µg/mL ( Sigma , I3377 ) , Rifampin at 1 µg/mL ( Euromedex ) and DMSO were used as controls . Drug susceptibility testing on benzothiazinone-resistant mycobacteria with various mutations in the rv3790 gene was performed as recently reported [21] . 384-well Evotec plates ( #781058 ) were first preplated with 0 . 5 µl of compound dispensed by EVOBird ( Evotec ) in 10 µl of RPMI 1640 ( Gibco ) supplemented with 10% heat-inactivated fetal calf serum ( FCS , Gibco ) . Raw 264 . 7 ( ATCC # TIB-71 ) ( 1 . 5×108 cells ) were infected with H37Rv-GFP [32] in suspension at a MOI of 1∶1 in RPMI 1640 supplemented with 10% heat-inactivated FCS for 2 hours at 37°C with shaking . After two washes by centrifugation , the remaining extracellular bacilli in the infected cell suspension were killed by a 1 hour Amikacin ( 20 µM , Sigma , A2324 ) treatment . After a final two-wash centrifugation , 10 000 infected cells were dispensed into each plate well pre-plated with compounds and controls . Infected cells were then incubated for 5 days at 37°C , 5% CO2 . After five days , macrophages were stained with SYTO 60 , 5 µM ( Invitrogen , S11342 ) for 1 hour at 37°C and image acquisition was performed on an EVOscreen-MarkIII fully automated platform ( PerkinElmer ) integrated with an Opera™ ( 20X-water objective , NA 0 . 70 ) and located in a BSL-3 safety laboratory . Mycobacteria-GFP were detected using a 488-nm laser coupled with a 535/50 nm detection filter and SYTO 60 labelled cells with a 635-nm laser coupled with a 690/40 nm detection filter . Four fields were recorded for each plate well and each image was then processed using dedicated in-house image analysis software ( IM ) described elsewhere [14] . Briefly , the algorithm first segments the cells on the red channel using a sequence of processing steps [13] . Firstly the contour of each macrophage is delineated using an algorithm based on the intensity signal given by the red channel ( Figure 1B ) . The number of red delineated surfaces corresponds to the number of macrophages . The host cell is then considered to be infected by M . tuberculosis if there is an overlap of at least 3 pixels in the green channel above a given intensity threshold within the cell surface . The ratio of infected cells to the total number of cells determines the percentage of infected cells . Another parameter deduced from the images is the bacterial load that refers to the total surface area of all the green objects that partly cover the delineated macrophages . Eight parameters that include cell number , cell surface , infected cell number , number of green objects , green object intensity , green object surface , green surface in infected cells and infection ratio were then processed plate by plate in a PCA protocol developed using PipelinePilot™ ( Accelrys ) . Briefly , the values from both positive and negative controls of each plate were first used to create a PCA model in 1 dimension for the plate ( Minimum Variance explained = 0 . 75 , Center-and-Scale data pre-transformation ) . The model was then applied to calculate the new coordinates of compounds and controls for that plate . Similar analysis was then repeated for each new screened plate . A Z′ score was then calculated , and controls and plates were accepted with a Z′ score above 0 . For each plate , the model described more than 99% of the variance . PCA-1x analysis improved active compound separation compared to the analysis based on the infected cell percentage parameter as demonstrated by achievement of better Z′ values ( Figure S2B ) . Hits were selected with a PCA-1x value below 0 . 5 , corresponding to the separation value between the DMSO and the INH 1 µg/mL populations . The compound library was screened at a single concentration of 20 µM . Hits were then cherry-picked and tested in ten- 2-fold serial dilutions ( from 20 µM to 0 . 5 nM ) in duplicate . 486 hits ( 0 . 85% ) were then confirmed . Data obtained from either the intracellular assay image analysis or from the conventional antibacterial assay were then processed using ActivityBase ( IDBS ) to calculate statistical data ( % of inhibition , Z score for each compound , Z′ , coefficient of variation ( CV ) etc . for the control plates ) . Results visualization was performed with Spotfire ( Tibco ) . If not specified in the figure legend , data are expressed as mean+/−SD from 2 independent experiments . Mouse bone-marrow-derived macrophages were obtained by seeding 107 bone marrow cells from C57BL/6 mice in 75 cm2 dishes in RPMI 1640 ( Gibco™ ) supplemented with 10% heat-inactivated FCS and 10% L-cell conditioned medium ( L-929 ) . Peripheral Blood Mononuclear Cells ( PBMC ) were isolated from buffy coat from healthy volunteers . 15 ml of Ficoll-Paque Plus ( Amersham Biosciences , Sweden ) were added to PBS diluted buffy coat diluted and centrifuged at 2500×g for 20 min . PBMC were obtained by CD14+ beads separation ( Miltenyi Biotec , Germany ) , washed 3-times with PBS containing 1% FCS and transferred to 75 cm2 culture flask containing RPMI 1640 media , 10% FCS and 50 ng/ml of recombinant-human macrophage colony stimulating factor ( rh-MCSF , R & D systems , Minneapolis ) . After 6 days , murine or human macrophages were harvested with Versene ( Gibco™ ) and seeded at a density of 1 . 5×104 cells per well in 384-well Evotec plates in 50 µl RPMI 1640 supplemented with 10% heat-inactivated FCS and 10% L-929 or 50 ng/ml of rh-MCSF respectively . Adherent cells were then infected with bacterial suspensions at a MOI of 2 . 5 to 1 bacteria per cell and incubated for 2 h . Cells were then washed three times with PBS supplemented with 1% FCS and further incubated with different concentration of DNB compounds for 7 days . Cells were then lysed with 0 . 1% Triton X-100 ( Sigma ) in H2O and serial dilutions were performed to quantify CFUs as previously reported [34] . The frequency of spontaneous mutation was determined on 7H10-OADC plates containing increasing concentrations of DNB1 and DNB2 at 0 . 4 , 0 . 8 , 1 . 6 and 3 . 2 µM . 105 , 106 , 107 and 108 CFU containing bacterial suspensions were spread on dinitrobenzamides containing agar plates . After 5–6 weeks at 37°C , colonies were counted and frequency of mutation was evaluated as the ratio of colonies grown relative to the original inoculum . DMSO and INH were used as negative and positive controls respectively . Reaction mixtures contained 1 mg of M . smegmatis membrane and cell wall ( P60 ) proteins [20] , 80 µM ATP , 120 , 000 dpm p ( 14C ) Rpp [35] , 50 mM MOPS pH 7 . 9 , 5 mM 2-mercaptoethanol and 10 mM MgCl2 . Reactions were stopped by the addition of CHCl3∶CH3OH ( 2∶1 ) and the organic phase backwashed with CHCl3∶CH3OH∶H2O ( 3∶47∶48 ) . After drying under N2 , the radiolabeled material was dissolved in CHCl3∶CH3OH∶H2O∶NH4OH ( 65∶25∶3 . 6∶0 . 5 ) for TLC analysis . For assessing the effects of DNB1 and DNB2 on whole M . tuberculosis H37Ra , 0 . 6 to 80 µM ( 0 . 2 to 28 µg/mL respectively ) of the compounds were added to bacterial cultures grown to mid-log phase in glycerol-alanine-salts medium and incubated for 16 hrs at 37°C with shaking , after which 1 µCi/mL [U-14C]glucose ( specific activity , 317 Ci mol−1 , MP Biomedicals Inc . ) or 0 . 5 µCi/mL [1 , 2-14C]acetic acid ( specific activity , 60 Ci mol−1 , NEN Radiochemicals ) were added and the cultures were incubated for another 24 hrs at 37°C . Untreated and inhibitor-treated bacteria were collected by centrifugation , washed and their lipids , lipoglycans ( LM and LAM ) and mycolyl-arabinogalactan-peptidoglycan ( mAGP ) complex , were extracted essentially as described [36] . 14C-glucose- and 14C-acetate-derived lipids and fatty acids ( including mycolates ) were analyzed by TLC on silica gel 60 aluminum-backed plates ( Merck , Darmstadt , Germany ) in a variety of solvent systems [37] . 14C-glucose-derived lipoglycans were separated on Tricine gels , transferred to nitrocellulose membranes and revealed by autoradiography . The amount of radioactivity incorporated into the individual sugars of the mAGP complex was determined by hydrolysis of the 14C-glucose-derived material with 2M CF3COOH for 3 hrs at 120°C and separation of the individual monosaccharides ( upon removal of fatty acids ) on aluminum-backed TLC plates developed twice in pyridine∶ethyl acetate∶acetic acid∶water ( 5∶5∶1∶3 ) . Autoradiograms were produced by exposure of the TLCs and nitrocellulose membranes to KODAK-Biomax MR films at −70°C . | Tuberculosis is still a major threat to global health . The disease in humans is caused by a bacterium , Mycobacterium tuberculosis , and treatment of an infected individual requires more than six months of chemotherapy . Because such a long course of treatment is required , compliance is low , which can result in the development of multidrug resistant strains ( MDR-TB ) and even extremely resistant strains ( XDR-TB ) . Identifying new drug targets and potential lead therapeutic compounds are needed to combat MDR-XDR-TB . We developed a new type of assay based on the visualization of mycobacterium replication within host cells and applied it for the search of compounds that are able to chase the pathogen from its hideout . As a result , we found 20 new series of drug candidates that are effective against the bacilli in its hiding place , potentially addressing a crucial aspect in the resilience of the disease . We also showed that one series of compounds acts by inhibiting a key enzyme required for the synthesis of an essential component from the mycobacterial cell wall that is not targeted by any of the commercially available antituberculosis drugs . Altogether , our results pave the way for development of the next generation of antibacterial agents . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"chemical",
"biology/small",
"molecule",
"chemistry",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance"
] | 2009 | High Content Screening Identifies Decaprenyl-Phosphoribose 2′ Epimerase as a Target for Intracellular Antimycobacterial Inhibitors |
Exome sequencing coupled with homozygosity mapping was used to identify a transition mutation ( c . 794T>C; p . Leu265Ser ) in ELMOD3 at the DFNB88 locus that is associated with nonsyndromic deafness in a large Pakistani family , PKDF468 . The affected individuals of this family exhibited pre-lingual , severe-to-profound degrees of mixed hearing loss . ELMOD3 belongs to the engulfment and cell motility ( ELMO ) family , which consists of six paralogs in mammals . Several members of the ELMO family have been shown to regulate a subset of GTPases within the Ras superfamily . However , ELMOD3 is a largely uncharacterized protein that has no previously known biochemical activities . We found that in rodents , within the sensory epithelia of the inner ear , ELMOD3 appears most pronounced in the stereocilia of cochlear hair cells . Fluorescently tagged ELMOD3 co-localized with the actin cytoskeleton in MDCK cells and actin-based microvilli of LLC-PK1-CL4 epithelial cells . The p . Leu265Ser mutation in the ELMO domain impaired each of these activities . Super-resolution imaging revealed instances of close association of ELMOD3 with actin at the plasma membrane of MDCK cells . Furthermore , recombinant human GST-ELMOD3 exhibited GTPase activating protein ( GAP ) activity against the Arl2 GTPase , which was completely abolished by the p . Leu265Ser mutation . Collectively , our data provide the first insights into the expression and biochemical properties of ELMOD3 and highlight its functional links to sound perception and actin cytoskeleton .
Many molecular components that are necessary for the development and maintenance of hearing have been discovered by identifying the genes that underlie hearing impairment in humans and mice [1]–[4] . Hearing requires the precise and efficient functioning of intricately structured mechanosensory hair cells and supporting cells in the inner ear [3] . One of the key structures in the mechanotransduction process is the hair cell stereocilium . Protruding from the apical surface of the hair cells , stereocilia are organized in three rows of decreasing height in a staircase pattern . Each stereocilium is composed of an actin core that contains cross-linked and bundled γ- and β-actin microfilaments that are uniformly polarized , with the barbed ( positive ) ends localized at the tip . At the tapered end of the stereocilium , the actin filaments form a rootlet that has been proposed to anchor the structure in the actin-rich meshwork of the cuticular plate [5] . Interestingly , among the identified hearing loss-associated genes , nineteen encode proteins that interact with actin [6] , [7] . Numerous studies have demonstrated that actin cytoskeleton-associated proteins are involved in the development , maintenance and stabilization of the stereocilia ( for review , see [6] ) . Continuous depolymerization of actin filaments at the base and polymerization at the barbed end , termed treadmilling , is thought to be critical to the maintenance of the length of stereocilia [8] , [9] . However , a recent study demonstrated a rapid turnover of the actin filaments only at the tip of the stereocilia , without a treadmilling process [10] , emphasizing the specific role of proteins at the stereocilia tip in the regulation of actin filaments . Regardless of the precise site , it is quite clear that the proper regulation of actin dynamics is critical to the generation and maintenance of stereocilia as sensory structures . The Rho/Rac/Cdc42 family of GTPases is well known as a regulator of actin . Rho and Rac in the inner ear are involved in the morphogenesis and growth of the otocyst [11] , [12] . The depletion of Rac1 or both Rac1 and Rac3 in the murine inner ear leads to a shorter cochlear duct with an abnormal sensory epithelium . Rac may participate in cell adhesion , proliferation , and movements during otic development [11] , [12] . Several studies have suggested that the activation/inhibition of Rho pathways control the actin depolymerization rate in the outer hair cells [13] , [14] . Although best known for their roles in the regulation of membrane traffic , there is growing evidence that GTPases in the Arf family can also act via changes in actin . [15] . Here , we report the identification of a new deafness gene , which encodes an ELMO/CED 12 domain containing protein , ELMOD3 . Our biochemical studies demonstrated that ELMOD3 possesses GAP activity against a small GTPase in the Arf family , Arl2 , providing a functional link between Arf family signaling pathways and stereocilia actin-based cytoskeletal architecture . GAPs are regulators and effectors of the Ras superfamily of GTPases , which are increasingly recognized as providing specificity as well as temporal and spatial regulation to GTPase signaling [16] . Thus , we believe that the identification of ELMOD3 role in the inner ear provides new insights into signaling processes that are important to hearing in humans .
Family PKDF468 ( Figure 1A ) was recruited after obtaining Institutional Review Board approval and written informed consent . The family history revealed that the onset of hearing loss was pre-lingual , with no clear vestibular impairment among the deaf individuals . Pure-tone bone and air-conduction audiometry revealed severe-to-profound mixed ( conductive and sensorineural ) hearing loss in the affected individuals of family PKDF468 ( Figure 1B ) . Individual V:2 exhibited severe-to-profound mixed hearing loss , with bone conduction thresholds for the right ear displaying a mild downward slope to the severe hearing loss range . The left ear displayed slightly better bone conduction thresholds with normal values for lower frequencies and a downward slope to severe hearing loss at higher frequencies ( Figure 1B ) . The audiograms of individual V:5 revealed bilateral severe-to-profound mixed hearing loss , with a large conductive component in both ears . The bone conduction thresholds exhibited a mild downward slope to moderately severe hearing loss for the right ear and were slightly better on the left , for which the thresholds ranged from borderline normal to moderate hearing loss ranges ( Figure 1B ) . The clinical evaluation revealed no clear signs of skin , renal , or retinal abnormalities . To determine the temporal bone malformation , we performed computed tomography ( CT ) scans of two affected ( V:2 and V:11 ) along with a normal hearing sibling ( V:7 ) . CT scan of individual V:2 revealed all three semicircular and internal auditory canals were intact on both sides . The middle ear and mastoid appeared well-aerated bilaterally . Imaging of individual V:11 demonstrated a slightly narrow appearing internal auditory canal on the right side only . The mastoid air cells and middle ear cleft were well-aerated bilaterally . The external auditory canal appeared normal as well for both affected individuals . We initially observed that deafness in family PKDF468 did not co-segregate with short tandem repeat ( STR ) markers for 74 of the reported recessive nonsyndromic deafness loci ( data not shown ) . We therefore performed a genome-wide linkage analysis and observed that the deafness phenotype of family PKDF468 exhibited significant evidence of linkage to STR markers on chromosome 2p12-p11 . 2 ( Figure 1A ) . Additional STRs on 2p were genotyped , and haplotype analysis revealed a 0 . 91 Mb linkage interval that was delimited by the markers D2S1387 and D2S2232 ( Figure 1A ) . Under a recessive model of inheritance , with a disease allele frequency of 0 . 001 and full penetrance , a maximum two-point LOD [17] score of 4 . 74 ( θ = 0 ) was obtained for the marker D2S2333 . These results define and delimit DFNB88 [Human Genome Nomenclature Committee ( HGNC ) approved locus symbol] , a novel recessive deafness locus on chromosome 2p11 . 2 . The DFNB88 locus partially overlaps with the dominant deafness locus DFNA43 ( Figure 1C ) [18] . Four known candidate genes were identified within the DFNB88/DFNA43 overlapping linkage region ( Figure 1C ) . However , Sanger sequencing of these genes did not reveal any pathogenic variants . Approximately 85% of the disease-causing mutations in Mendelian disorders reside in coding regions or in exon-intron canonical splice junctions [19] . We therefore performed exome sequencing of an affected individual from family PKDF468 . The sample was enriched using the NimbleGen SeqCap EZ Exome Library v2 . 0 ( Roche Diagnostics; San Francisco , CA ) , and 100 bp , paired-end sequencing was performed on the Illumina HiSeq 2000 platform ( Illumina ) . An average of 78 . 94% of bases were sequenced with 20× coverage within the targeted regions . This yielded a total of 64 , 863 single-nucleotide variants , of which 1 , 928 were not found in the dbSNP133 database ( Table S1 ) . Based on the recessive mode of inheritance evident in the pedigree , we analyzed genes with homozygous changes and potential compound heterozygous changes . Additionally , we removed all of the variants that were present in six ethnically matched control samples ( Table S1 ) . No mutation segregating with hearing loss in family PKDF468 was identified in any of the known deafness-causing genes ( Table S1 ) . We identified one homozygous transition mutation , c . 794T>C ( p . Leu265Ser ) , in ELMOD3 ( Figure S1 ) on chromosome 2p11 . 3 ( Figure 1C ) that segregated with DFNB88-linked deafness ( Tables S1 and S2 ) . The c . 794T>C change was not present near the canonical splice junctions and was not predicted to create any aberrant splice site . However , to confirm that c . 794T>C did not affect splicing of ELMOD3 transcripts , we generated cDNA libraries using the total RNA extracted from the white blood cells of two affected and one normal hearing individual . Sanger sequencing of sub-cloned PCR products , amplified using primers in either exons 9 and 11 or in exons 9 and 12 ( Figure S2 ) , did not reveal any aberrant splicing product in affected individuals . Thus , the likely pathogenic affect of the c . 794T>C change is substitution of a highly conserved leucine residue at amino acid position 265 of the human ELMOD3 protein with serine ( Figure 2C ) . No carrier of c . 794T>C was identified among 524 ethnically matched control chromosomes , in the 1000 Genome database or in the 6500 individuals who are listed in the NHLBI-ESP variant database ( http://evs . gs . washington . edu/EVS/ ) . Moreover , Polyphen-2 [20] , SNPs3D [21] , MutationTaster [22] , PMut [23] , and SIFT [24] predicted that the ELMOD3 mutation would be deleterious ( Table S3 ) . To further confirm that the p . Leu265Ser allele of ELMOD3 is the only mutation that was associated with hearing loss at the DFNB88 locus , we sequenced the coding , non-coding , and approximately 75 bp flanking sequences of the exon-intron boundaries of all the known candidate genes present within the linkage region in two affected individuals of family PKDF468 ( Figure 1C ) . No other potentially pathogenic mutation was identified in the affected individuals of family PKDF468 . Although , ELMOD3 is located outside the reported linkage interval of DFNA43 ( Figure 1C ) [18] , nevertheless we sequenced DNA samples of two affected individuals from the original DFNA43 family and no mutation was found . We next examined the gene structure and expression of ELMOD3 . Seven alternatively spliced isoforms of human ELMOD3 were identified ( Figure 2A ) . Isoform A ( reference sequence NM_ 032213 . 4 ) has a translation initiation codon ( AUG ) in exon 2 , ten coding exons that encode a polypeptide of 391 residues ( Figure S1B ) . Exons 7 to 11 encode the engulfment and cell motility ( ELMO or CED12 ) domain , which consists of 164 amino acid residues ( Figure S1B; blue box ) . ELMOD3 isoforms B to D include alternatively spliced exons in the 5′ untranslated region ( UTR ) but harbor the same coding exons and encode identical 381 residue polypeptides that differ from isoform A only at their carboxy termini ( Figures 2A and S1B ) . The human ELMOD3 isoforms , A and B , share 87% identity , with all the differences clustered near the C-terminus . Isoforms E , F , and G do not encode the full-length ELMO domain due to alternate splicing of exons in the carboxy terminus ( Figure 2A ) . The c . 794T>C transition mutation is predicted to result in the substitution of serine for a highly conserved leucine in all of the ELMO domain-containing isoforms of ELMOD3 ( Figures 2A and 2C ) . In comparison to the human sequence , mouse Elmod3 includes only three known alternatively spliced transcripts ( Figure 2A ) . RT-PCR and real-time quantitative PCR analysis of multiple human and mouse tissue cDNAs ( Tables S4 and S5 ) revealed the ubiquitous expression of isoforms A/a and B–D/b–c ( Figures S3 and 2B ) . We also assayed the relative mRNA expression of murine Elmod3 isoforms a and b–c with real-time quantitative RT-PCR of RNA that was extracted from cochlear and vestibular inner ear tissues from postnatal day 0 ( P0 ) , P10 , and P30 C57BL/6J mice ( Figure 2D ) . The expression of Elmod3 isoform b–c was several-fold higher than isoform a , in both cochlear and vestibular tissues at all of the time points examined ( Figure 2D ) . Therefore , we focused on the ELMOD3 isoform B for the subsequent biochemical and cellular studies . The mouse ELMOD3 protein is 70% and 80% identical to the A and B isoforms of human ELMOD3 , respectively , and again the differences are greatest at the C-terminus , although single amino acid changes are scattered throughout the alignments . To characterize the cellular localization of ELMOD3 , we produced a rabbit polyclonal antiserum against synthetic peptide immunogens from mouse ELMOD3 isoform b . The sensitivity and specificity of the ELMOD3 antibody was validated in immunoblot and immunofluorescence analyses , in transfected cells and mouse tissues ( Figures S4 and S7 ) . Our antibodies specifically recognized ELMOD3 isoform b but not murine ELMOD1 , ELMOD2 , or ELMOD3 isoform a ( Figure S4 ) . We next performed immunolocalization of ELMOD3 in the rat and mouse organ of Corti ( Figures 3 and 4 ) . In rat cochlea , ELMOD3 immunoreactivity was observed in the stereocilia , kinocilia and cuticular plate of developing hair cells ( Figures 3 and S5 ) . Before P07 , ELMOD3 staining was very weak in the inner hair cells stereocilia . By P07 , in auditory hair cells , patchy labeling of ELMOD3 immunostaining was detected along the length of stereocilia ( Figure 3 ) . In contrast to actin staining , ELMOD3 immunoreactivity was not uniform along the length of each stereocilium and the protein seemed to be excluded from a region near the tip ( Figure 3E ) . ELMOD3 immunoreactivity was also found in the supporting cells , including pillar and Dieters' cells ( Figure 3 ) . Similar to that seen in the rat ( Figure 3D ) , the stereocilia of inner hair cells in the mouse organ of Corti were more intensely labeled than those of outer hair cells ( Figure 4A ) . In contrast to the cochlear hair cells , ELMOD3 antibody labeling was observed within the hair cell bodies in the vestibular end organs of both rat and mouse inner ear , but no prominent immunoreactivity was observed in the hair bundles ( Figures 4B–4C and S6 ) . These observations suggest a unique role for ELMOD3 in cochlear sensory cells and may reflect the functional or structural differences between cochlear and vestibular hair bundles . No specific immunoreactivity was observed when the primary antibody was omitted ( data not shown ) or when the antibody was pre-incubated with the ELMOD3 peptide antigen ( Figure S7 ) . We examined LLC-PK1-CL4 epithelial ( CL4 ) cells to understand the mechanism and effect of the hearing loss-associated allele of ELMOD3 . CL4 cells contain actin-rich microvilli and have been used as in vitro models of stereocilia to examine F-actin and protein dynamics [25] . We transiently co-transfected GFP-ELMOD3 constructs with Espn constructs , where the latter was used to over-elongate the microvilli at the CL4 cell surface [26] ( Figure 5A–5B ) . We observed a significant expression of GFP-ELMOD3 in the apical ( microvillar ) plasma membrane twenty-four hours post-transfection ( Figure 5A ) . We also observed expression of GFP-ELMOD3 in the cytosol of transfected cells ( Figure 5A ) . In contrast to the wild type protein , the p . Leu265Ser mutation in the ELMO domain yielded a protein that displayed either weak or no labeling in the microvilli of the transfected CL4 cells . Additionally , the protein appeared to be diffusely located throughout the cytoplasm , with a nuclear concentration in approximately half of the transfected cells ( Figure 5B ) . Identical results were observed with tdTomato-tagged wild-type and mutant ELMOD3 constructs ( data not shown ) . To determine the effect of p . Leu265Ser mutation on the localization of ELMOD3 in the mouse inner ear , we performed gene gun-mediated transfection of wild-type and p . Leu265Ser mutant GFP-tagged ELMOD3 cDNA constructs in organotypic cultures of inner ear sensory epithelia of P2 C57BL/6J mice ( Figure 5C–5D ) . Over-expressed wild-type GFP-ELMOD3 localized along the length of the stereocilia of cochlear hair cells ( Figure 5C ) . We also observed homogeneous distribution throughout the hair cell bodies ( Figure 5C ) . Similar to the results that were observed in CL4 cells , GFP-ELMOD3 harboring the p . Leu265Ser mutation failed to target to the stereocilia , and the protein was apparently distributed throughout the cochlear hair cell bodies ( Figure 5D ) . Taken together , these results support our conclusion that ELMOD3 localizes to actin-based microvilli and stereocilia ( Figure 5 ) but that a point mutation in the ELMO domain can prevent its normal localization and potentially affect its function in the stereocilia . To further investigate the ELMOD3-actin association , we transfected GFP-tagged ELMOD3 into MDCK cells , which is a highly polarized cell model system ( Figure S8 ) . Forty-eight hours post-transfection , GFP-ELMOD3 accumulation was apparent at the periphery of the transfected cells near the plasma membrane ( Figure S8A ) . The expression of GFP-ELMOD3 harboring the p . Leu265Ser mutation in MDCK cells resulted in a protein that failed to target or accumulate at the plasma membrane and instead , appeared to concentrate in the nuclei ( Figure S8B ) . To determine whether GFP-ELMOD3 associates with the actin cytoskeleton at the plasma membrane ( Figure 6A ) , we treated the cells with cytochalasin D ( cyto-D ) , which is a potent inhibitor of actin polymerization , to disrupt the actin cytoskeleton [27] , [28] . We hypothesized that if GFP-ELMOD3 associated with the actin cytoskeleton at the cell membrane , then treatment of the cells with cyto-D would also affect ELMOD3 localization . Indeed , we observed a significant decrease in the GFP-ELMOD3 signal at the cell membrane following disruption of the actin cytoskeleton ( Figure 6B , 6D ) . Four hours following cyto-D treatment ( i . e . , the recovery period for actin re-polymerization ) [27] , we observed that ELMOD3 re-accumulated at the cell membrane ( Figure 6C , 6D ) . These results suggest that the localization of ELMOD3 is dependent on the actin cytoskeleton and/or may contribute to a mechanism that supports its maintenance . To decipher the link between F-actin and ELMOD3 , we performed a two-color stochastic optical reconstruction microscopy ( STORM ) imaging of EGFP-ELMOD3 transfected MDCK cells . While conventional confocal acquisitions revealed co-localization of ELMOD3 and the actin-cytoskeleton , this high resolution imaging technique allowed us to determine more precisely the relative positions of ELMOD3 and actin ( Figure 7 ) . ELMOD3 and actin were each found in close apposition to the plasma membrane and in irregularly shaped puncta ( Figure 7C ) . Many regions of extensive overlap in staining between ELMOD3 and actin , suggest the possibility that a subset of the actin-based structures may contain ELMOD3 but that each protein is also found localized independently of the other at the plasma membrane ( Figure 7C ) . To test the possibility that ELMOD3 binds to actin-based structures , we performed a high-speed co-sedimentation assay that pellets actin along with its associated proteins . To obtain a source of purified ELMOD3 , His6-Trigger factor-ELMOD3 ( TF-ELMOD3 ) fusion protein was expressed in bacteria , and the recombinant protein was purified by Ni-NTA chromatography . The TF-ELMOD3 or control proteins were incubated with polymerized F-actin and subjected to high-speed centrifugation at 150 , 000× g for 1 . 5 hrs ( Figure S9 ) . TF-ELMOD3 co-sedimented , albeit weakly or incompletely , with F-actin in this assay ( Figure S9 ) . Under these conditions , the p . Leu265Ser mutation did not significantly impact the level of TF-ELMOD3 that co-sedimented with F-actin ( Figure S9 ) . Human ELMOD1 and ELMOD2 each possess Arl2 GAP activity [29] . We therefore investigated whether ELMOD3 also possesses GAP activity against Arl2 . Previous tests of bacterially expressed human ELMOD3 as either maltose binding proteins or trigger factor fusion proteins were negative , but the homologous preparations of ELMOD1 and ELMOD2 were found to possess very low specific activities as Arl2 GAPs , compared to the preparation purified from bovine tissues . To obtain a potentially more active preparation of human ELMOD3 , we expressed ELMOD3 and the mutant p . Leu265Ser in HEK293T cells as N-terminal GST-fusion proteins to facilitate protein purification . Protein expression and purification from ∼108 HEK293T cells that expressed GST-ELMOD3 or the mutant each yielded ∼0 . 6 mg protein . These preparations were stable at 4°C and against freeze-thaw cycles , as judged by either GAP activity or lack of precipitation . We expressed and purified GST alone and used it as a negative control in all of our assays . We also evaluated the effect of cleavage of the GST fusion tag by TEV protease on ELMOD3 activity; no changes in activity in the Arl2 GAP assay were observed compared to un-cleaved proteins ( data not shown ) . Thus , we believe that the presence of the GST moiety at the N-terminus does not interfere with access to the substrate or with enzymatic activity in our assay . When we varied the amount of GST-ELMOD3 protein in the Arl2 GAP assay , we observed a dose-dependent response and evidence of saturation at higher protein concentrations ( data not shown ) . Using the lower concentrations of GST-ELMOD3 to estimate the initial rates of GAP-dependent activity and estimating the purity of the preparation at 50% ( based on visual inspection of Coomassie blue-stained gels ) , we obtained a specific activity of 24 pmol of GTP hydrolyzed/min/mg ( Figure 8 ) . This specific activity of GST-ELMOD3 as an Arl2 GAP is approximately 32-fold lower than that determined for GST-ELMOD1 and nearly 1000-fold lower than that of GST-ELMOD2 or bovine testes ELMOD2 , which is the most active reported preparation of any Arl2 GAP [29] . Thus , in contrast to our earlier report that it is inactive , GST-ELMOD3 does exhibit Arl2 GAP activity , and we believe that its lower specific activity when expressed in bacteria likely contributed to the earlier negative findings [29] . The differences in specific activity among the three human GST-ELMOD preparations from HEK293T cells are predicted to result from differences in substrate specificity , sensitivities to co-activators ( as known for Arf GAPs ) , or both . Thus , more studies are required to determine whether the biologically relevant substrate of ELMOD3's GAP activity in the inner ear is Arl2 or a related GTPase . We next assessed the effects of the Leu265Ser point mutation on the Arl2 GAP activity of GST-ELMOD3 . The mutant protein was expressed at the same levels in HEK293T cells and was purified in the same way , resulting in equivalent amounts of protein , indicating that the protein is equally stable in mammalian cells and in solution . However , when assayed for Arl2 GAP activity , the mutant was inactive ( Figure 8 ) . Although we observed small amounts of activity over our no protein control , this level of activity seen for the mutant was not different from that observed with GST alone ( Figure 8 ) . These activities are so low as to be at or near the lower limits of our assay . Thus , we can safely conclude that the point mutant has at least a 10-fold lower specific activity than the wild-type protein , but it might be completely inactive as an Arl2 GAP .
Our study revealed that ELMOD3 is important for hearing in humans as a missense mutation in the gene leads to profound hearing impairment . ELMOD3 belongs to the engulfment and cell motility ( ELMO ) protein family , which includes six known members in mammals ( ELMO1-3 and ELMOD1-3 ) . Our ex vivo studies reveal that fluorescently tagged ELMOD3 localized with the actin-based microvilli of LLC-PK1-CL4 epithelial cells , in the stereocilia of sensory hair cells of mouse organ of Corti explants , and to a lesser extent to the actin cytoskeleton of MDCK cells , whereas the deafness-associated allele ( p . Leu265Ser ) was deficient in each case . Similarly , we show that human ELMOD3 possesses Arl2 GAP activity but the mutant has at least a 10-fold loss in activity . While ELMOD3 antibody reactivity was detected in outer hair cells stereocilia at P02 , more pronounced accumulation of ELMOD3 immunoreactivity was detected in rat cochlear inner hair cell stereocilia only by P12 , which is when hair bundles are in the late phase of maturation . During this period , the inner hair cell stereocilia undergo a rapid elongation [30] . The observed staining suggests that ELMOD3 might be necessary for the initial development of the outer hair cell stereocilia or the organization of the bundle in a staircase pattern but may play a different role in the stereocilia of inner hair cells . Nevertheless , it is tempting to speculate that ELMOD3 may play a role in the maturation or maintenance of the cochlear stereociliary bundle . Recently , two spontaneous mutations ( rda and rda2J ) in mouse Elmod1 were shown to result in profound deafness and vestibular dysfunction [7] , demonstrating that the function of ELMOD1 is essential for regulating the shape and maintenance of inner ear hair cell stereocilia in mice [7] . Elmod1 has been shown to be part of a large cluster of genes expressed in the developing inner ear , while Elmod3 level was below the detectable range [31] . These observations are consistent with findings from other studies , like the SHIELD database , which demonstrated that the level of Elmod3 mRNA is ∼100-fold lower than that of Elmod1 in the developing inner ear ( P0–P1 ) . Besides stereocilia bundles , immunoreactivity for ELMOD3 was also detected along the kinocilium in developing cochlear hair cells . Recently , it has been shown that ELMO1 can act at the interface between the actin-cytoskeleton and microtubule network by interacting with ACF7 ( Actin crosslinking family 7 ) [32] . Moreover , microtubule polymerization depends on Arl2 activity [33] , and we have shown that ELMOD3 exhibits a GAP activity against Arl2 . Therefore , ELMOD3 expression in the kinocilium might have a role in the assembly of the kinocilium architecture and in pathways regulating planar cell polarity . The ELMO family proteins are functionally poorly characterized , and more information is currently available for the ELMOs than for the ELMODs , with no structural information available for any of them . So far , only one activity has been ascribed to ELMOD proteins: we previously reported that recombinant human ELMOD1 and ELMOD2 display in vitro Arl2 GAP activity , whereas ELMOD3 and bacterially expressed ELMO1-3 did not [29] . More recently , we performed additional phylogenetic and functional analyses of the ELMO domain that led us to re-examine whether ELMOD3 shares the Arl2 GAP activity of ELMOD1 and ELMOD2 . Our data contrast with the earlier-published claim that ELMOD3 lacks Arl2 GAP activity; we determined that it does indeed possess this activity , albeit at a substantially lower specific activity than that of its two closest human paralogs . The large differences in specific activities observed in the Arl2 GAP assay may be due to differences in the specificities of ELMODs as GAPs for different GTPases , including the lack of one or more binding partners ( e . g . , one that is perhaps analogous to Dock180 binding to ELMO1 or a co-activator for activity such as has been proposed for COP-I and ArfGAP1 [34] ) , and/or the lack of post-translational modification . Our in vitro experiment revealed that ELMOD3 harboring the p . Leu265Ser mutation , unlike ELMOD3 , has no or few GAP activity against Arl2 . The lack of Arl2 GAP activity of the mutant may suggest a reduced affinity for the Arl2 GTPase , which may play an important role in ELMOD3 localization . Elmod3 and Arl2 are expressed in developing mouse cochlear tissues and weakly in vestibular tissues ( Figure S10 ) . Even though ELMOD3 is active and has been defined as “Arl2 GAPs” , we expect it to be active against other GTPases in the Arf family as well . We therefore speculate that ELMOD3 functions as a GAP for Arl2 and perhaps other GTPases that participate in actin organization , polymerization or depolymerization in the cochlear hair bundles . If ELMOD3 is an active GAP for other GTPases , these GTPases are likely to be part of the Arf family given that GAPs are not known to cross family boundaries within the Ras superfamily . However , it is plausible that ELMOD3 functions in a signaling pathway that includes Arl2 ( or an Arf family GTPase ) , Rac , Rho , and , ultimately , affects the actin cytoskeleton . Future studies will address the specific roles of ELMOD3 in the development of the inner ear sensory epithelium , cytoskeletal organization , and ELMOD3-mediated signaling pathways . Revealing the interacting partners , substrate specificities for its GAP activities , as well as the means of regulation of ELMOD3 and other ELMO family proteins , will shed light on the overlapping functions of the Ras superfamily in the inner ear . These fundamental functions of this unique protein family are likely to be important in all eukaryotic cells .
Family PKDF468 was enrolled in the present study from the Punjab province of Pakistan , and written informed consent was obtained from all participating family members . The Institutional Review Boards at the Center for Excellence in Molecular Biology ( Pakistan ) , at the National Institute on Deafness and Other Communication Disorders , and at Cincinnati Children's Hospital ( USA ) approved the present study . Hearing loss in the affected family members was evaluated using pure-tone audiometry , which tested frequencies that ranged from 125 Hz to 8 kHz . The family medical history stated that the onset of hearing loss was pre-lingual , and we observed no evidence of vestibular dysfunction or other balance issues using the Romberg and tandem gait tests . There were no other significant findings from the clinical exam , and the affected members had basic metabolic panel results within the normal range , indicating that they had nonsyndromic hearing loss . We conducted a genome-wide scan on family PKDF468 using 388 STR markers and performed linkage analysis using GeneMapper software ( Applied Biosystems; Carlsbad , CA ) . The LOD score was calculated using a recessive model of inheritance assuming a fully penetrant disorder and a disease allele frequency of 0 . 001 . The primers were designed with Primer3 to sequence all of the coding exons and 75 bp of the exon-intron boundaries of all of the known genes within the DFNB88 locus ( Table S2 ) . The products were amplified using either Taq polymerase ( Genscript; Piscataway , NJ ) or Amplitaq Gold 360 ( Applied Biosystems ) for the GC-rich regions . The chromatograms were read using SeqMan software ( DNAStar; Madison , WI ) . Exome sequencing was conducted on one affected individual from family PKDF468 and was enriched using the Nimblegen SeqCap EZ Exome v2 . 0 Library ( Roche Diagnostics; San Francisco , CA ) . One hundred base pair paired-end sequencing was performed on an Illumina Hi-Seq 2000 system . The sequencing data were analyzed following the guidelines that are outlined in the Broad Institute's Genome Analysis Toolkit [35] , [36] . The row data were mapped using the Burrows Wheeler Aligner [36] , the variants were called using the Unified Genotyper , and the data underwent further processing and quality control [35] , [36] . Low-quality reads ( less than 10× coverage ) were removed , and the remaining variants were filtered against the dbSNP133 database and all of the known variants in the NHLBI 6500 Exome Variant database that had a minor allele frequency ( MAF ) of greater than 0 . 05% . We also filtered out additional variants that were observed in six ethnically matched control exomes . Primers were designed , using Primer3 , to screen the remaining candidate gene variants , and we performed segregation analysis by performing Sanger sequencing of the variants of all of the participating family members . Human and mouse ELMOD3/Elmod3 isoform-specific primers and TaqMan probes were designed , using Primer3 web-based program , and the transcripts were amplified from human and mouse cDNA libraries ( Clontech Laboratories; Mountain View , CA ) . Mouse inner ear tissues were harvested from 3 or more C57BL/6J mice at P0 , P10 , and P30 . The cochlea and vestibular system were separated from the inner ear , and the total RNA was extracted from each tissue using TRIreagent ( Life Technologies , Grand Island , NY ) . The RNA was reverse-transcribed into cDNA using the SMARTscribe Kit ( Clontech ) . Real-time PCR was performed in triplicate on a StepOne Plus instrument ( Applied Biosystems ) . The data were analyzed using the comparative Ct method , with Gapdh as the endogenously expressed reference gene . RT-PCR was performed by using LA Taq ( Clontech ) . The products were run on a 2% agarose gel that was stained with ethidium bromide and each isoform was verified by sequencing . To determine the in vivo effect of c . 794T>C allele , if any , on the splicing of ELMOD3 transcripts , total RNA was isolated from fresh blood samples of two affected individuals ( V:2 and V:11 ) and one normal hearing individual ( V:3 ) by use of TRIzol reagent ( Life Technologies ) . Oligo dT and randomly primed first strand cDNA libraries were generated using SMART 1st strand cDNA synthesis kit ( Clontech ) . Touchdown PCR was performed with GenScript Taq ( GenScript ) and 1 . 5 mM MgCl2 at an annealing temperature of 63°C for 30 cycles using ELMOD3-specific primer pairs with a common forward primer in exon 9 and reverse primer either in exons 11 ( hELMOD3_ex9-11; Table S5 ) or in exon 12 ( hELMOD3_ex9-12; Table S5 ) . GAPDH was used as a control and amplified under the same conditions . PCR fragments were subcloned into pCR-TOPO cloning vector ( Life Technologies ) , and the sequences were verified . Human ELMOD3 B isoform , murine Elmod1 , Elmod2 and Elmod3 isoforms a and b open reading frame have been amplified from commercially available human and mouse cDNA libraries ( Clontech ) and inserted in pEGFP-C2 vector ( Clontech ) to generate proteins with GFP fused to their N-termini . The construct encoding p . Leu265Ser ELMOD3 was prepared through site-directed mutagenesis ( Agilent Technologies , Santa Clara , CA ) using the wild-type ELMOD3 isoform B as a template . The full-length open reading frame of human ELMOD3 ( isoform B ) was cloned into the pCOLD-TF ( Takara Bio , Inc . ; Otsu Shiga , Japan ) vector at the BamHI and SalI sites by PCR amplification of the cDNA using primers that inserted the appropriate restriction sites . This generated a fusion protein with a His6 tag at the N-terminus . This tag was followed by trigger factor ( ∼48 kDa ) , a thrombin cleavage site , and the ELMOD3 open reading frame . To insert the Leu265Ser mutation , we performed site-directed mutagenesis on the construct using the QuikChange Lightening Kit ( Agilent Technologies ) . Full-length open reading frames of human ELMOD3 ( Isoform B ) and the Leu265Ser mutant were cloned into the pLEXm-GST vector [37] using KpnI and SphI sites that were inserted into the PCR primers , with subsequent confirmation of the correct DNA sequence . The parent vector was used to express GST alone . Inner ear explants were harvested from C57BL/6J mice at P2 . The explants were cultured in a glass-bottom Petri dish ( MatTek , Ashland , MA ) that was coated with Matrigel ( BD Biosciences , San Jose , CA ) and were maintained in DMEM that was supplemented with 7% fetal bovine serum ( FBS ) ( Life Technologies ) for 24 hrs at 37°C with 5% CO2 . The cultures were transfected using a Helios gene gun ( Bio-Rad , Hercules , CA ) , as described elsewhere [38] . HEK293T , CL4 and MDCK cells were grown in DMEM that was supplemented with 10% FBS , 2 mM L-glutamine , and penicillin/streptomycin ( 50 U/ml ) ( Life Technologies ) and were maintained at 37°C in 5% CO2 . The cells were transfected using Fugene HD Transfection Reagent ( Promega; Sunnyvale , CA ) , according to the manufacturer's instructions . The cells were then cultured for an additional 48 hrs prior to immunostaining . Forty-eight hrs following transfection , we added 2 . 5 µM Cytochalasin D ( EMD Millipore , Billerica , MA ) in fresh DMSO medium to the cells for two hrs to disrupt the actin cytoskeleton . The Cytochalasin D was then washed out , and the cells were grown for four additional hrs in complete medium . ELMOD3 antiserum was raised in rabbits against two synthetic mouse ELMOD3-specific peptides ( corresponding to residues 143–156 and 346–361 of the mouse sequence [GenBank accession number GI:358679299] ) . The immunizations and sera collections were performed by Covance ( Princeton , New Jersey ) . The antiserum was affinity purified ( AminoLink Plus Immobilization Kit; Thermo Scientific , Rockford , IL ) either using both synthetic peptides in combination or individually . Antibody specificity was assessed by transfections ( Fugene HD Transfection Reagent; Promega ) of GFP-tagged mouse ELMOD1 , ELMOD2 and ELMOD3 into HEK293T cells followed by western blot analysis , as described elsewhere [39] . Inner ear and olfactory bulbs tissues were harvested from C57BL/6J mice at P30 and followed by western blot analysis , as described elsewhere [39] . Antigen competition was performed by incubating the primary antibody for 30 min at room temperature with the two immunizing peptides prior to use in Western Blot or immunofluorescence analyses . C57BL/6J mice were obtained from Jackson Laboratories ( Bar Harbor , ME ) and bred in house . Sprague–Dawley rats were purchased from Charles River Breeding Laboratories ( Raleigh , North Carolina ) and bred in house . All experiments and procedures were approved by the Institutional Animal Care and Use Committee of the Cincinnati Children's Hospital Medical Center . The inner ears from rats and mice were fixed with 4% PFA at 4°C overnight . P12 and P14 rat cochlea were incubated for one day , at 4°C , in 0 . 25M EDTA . The sensory epithelia were dissected in PBS . Following permeabilization with 0 . 25% Triton X-100 for 45 min , the samples were incubated in blocking solution ( 5% normal goat serum in PBS ) . The samples were then incubated overnight at 4°C with primary antibody ( anti-ELMOD3; anti acetylated Tubulin ( Sigma-Aldrich , St Louis , MO ) ) in 3% NGS/PBS . This step was followed by three washes in PBS and consecutive incubation with Alexa-488 conjugated secondary antibody and Alexa-647 conjugated secondary antibody ( Life Technologies ) at 1∶500 dilution and with rhodamine phalloidin ( Life Technologies ) at 1∶200 dilution in 3% NGS/PBS for one hour . After three washes with PBS , the samples were mounted using ProLong Gold Antifade Reagent ( Life Technologies ) . Transfected cells and inner ear explants were washed with PBS and fixed for 20 min in 4% paraformaldehyde . Filamentous actin was detected with rhodamine phalloidin ( Life Technologies ) in PBS/0 . 1%Triton-X100 for one hour . Following subsequent washes , the coverslips were mounted using FluoroGel medium ( Electron Microscopy Sciences , Hatfield , PA ) for the cell monolayers or with ProLong Gold Antifade Reagent ( Life Technologies ) for the explant cultures . MDCK cells were washed with PBS , fixed for 20 min in 4% paraformaldehyde and blocked with 10%NGS/PBS/0 . 1% Triton-X100 . The cells were incubated overnight at 4°C with a primary antibody ( ZO1 , Life Technologies; anti-ELMOD3 ) in 3% NGS/PBS/0 . 1%Triton-X100 . The cells were then washed in PBS/0 . 1%Triton-X100 and incubated with the Alexa-fluor 546 conjugated secondary antibody ( Life Technologies ) in 3% NGS/PBS/0 . 1% Triton-X100 for one hour at room temperature . Filamentous actin was detected with Alexa647-phalloidin ( Life Technologies ) in PBS/0 . 1%Triton-X100 for one hr . Following subsequent washes , the coverslips were placed using FluoroGel mounting medium ( Electron Microscopy Sciences ) . All images were acquired using a Zeiss LSM 700 scanning confocal microscope that was equipped with 63× and 100× objectives , and the analyses were performed using ImageJ software . The pixel intensity analyses were performed using ImageJ software on images that were acquired with the same microscope settings . The statistical analyses were performed using GraphPad Prism software and the ANOVA test function . Transfected MDCK cells were washed with PBS and fixed for 15 min in 4% paraformaldehyde . Following a one hour incubation in blocking solution ( 10% normal goat serum in PBS with 0 . 1% Triton X-100 ) , the cells were then incubated one hour with primary antibody: polyclonal chicken anti-GFP ( Aves Labs Inc . , Tigard , OR ) custom conjugated with APEX Alexa Fluor 568 Antibody ( Life Technelogies ) , in PBS/0 . 1%Triton-X100/3%NGS . Filamentous actin was labeled concomitantly with Alexa647-phalloidin ( Life Technologies ) . After extensive washes in PBS/0 . 1%TritonX-100/3%NGS , the cells were fixed 15 min in 4% paraformaldehyde and stored at 4°C in PBS . For imaging , PBS was replaced by the following imaging medium: 2-mercaptoethanol , buffer B ( 10% glucose , 50 mM Tris-HCl pH = 8 , 10 mM NaCl ) and the GLOX system ( 14 mg Glucose Oxidase , 50 ul Catalase , 200 ul Tris buffer ) in a 1∶100∶1 volume ratio . N-STORM imaging was performed with a Nikon N-STORM super-resolution microscope system ( Nikon Instruments Ltd , Melville , NY ) based on an inverted microscope Nikon A1Rsi equipped with a perfect-focusing system and a 100× TIRF APO NA 1 . 49 oil objective . A 561 nm wavelength laser was applied for bleaching and excitation of Alexa Fluor 568 while a 647 nm wavelength laser was applied for Alexa Fluor 647 . The images were acquired with an Andor Xion 897 EMCCD camera using 16 ms exposition with one frame of imaging for 5000 cycles . The analysis was performed with Nikon Elements Storm software . BL21 ( DE3 ) Gold E . coli were transfected with the pCOLD-TF-ELMOD3 and pCOLD-TF-ELMOD3 Leu265Ser plasmids , and single colonies were used to inoculate cultures in LB medium with 100 µg/ml ampicillin , which were grown at 37°C with shaking . When OD600 = 0 . 5 , the cells were moved to 15°C for 30 min without shaking . IPTG was then added to 1 mM , and the culture was grown overnight at 15°C with shaking . The cells were collected and lysed by passage through a French press in 20 mL of 20 mM HEPES ( pH 7 . 5 ) , 150 mM NaCl , and 5 mM imidazole with a protease inhibitor cocktail ( Sigma-Aldrich ) . The lysates were clarified by centrifugation at 100 , 000× g for one hour at 4°C , and the supernatants were loaded onto a Ni-NTA column ( GE Healthcare ) that was pre-equilibrated in the same buffer . The column was washed with 20 mL of 20 mM HEPES ( pH 7 . 5 ) , 150 mM NaCl , and 55 mM imidazole prior to elution in 15 mL buffer that contained 250 mM imidazole . The protein was further purified and buffer-exchanged by gel filtration chromatography using a Superdex S200 column ( 24 mL ) that was run in 20 mM HEPES ( pH 7 . 5 ) , 150 mM NaCl , and 1 mM dithiothreitol . Typical yields from this protocol were ∼10 mg/L TF-ELMOD3 that was >90% pure , as estimated by visual inspection of a Coomassie blue-stained gel . The actin binding experiment was performed with purified TF-ELMOD3 and TF-ELMOD3 Leu265Ser using an Actin Binding Protein Biochem Kit ( Cytoskeleton Inc , Denver , CO ) , following the manufacturer's protocol and later repeated with the GST-ELMOD3 proteins purified from HEK cells . GST-ELMOD3 , the point mutant , or GST alone was expressed in HEK293T cells using the pLEXm-GST vector ( a kind gift from Dr . James Hurley ( NIDDK ) ) and a modification of the method that was described in Aricescu et al [37] . Briefly , HEK293T cells ( 10×10 cm plates ) were transfected at 90% confluency with 1 µg/mL DNA after mixing with polyethyleneimine ( PEI-MAX; Polysciences , Inc . ; Warrington , PA ) at a 1∶3 ratio of DNA∶PEI in Opti-MEM medium ( Life Technologies ) . The mixture was then added to cells in DMEM medium that contained 2% FBS and grown for two days . The cells were collected by centrifugation and lysed by resuspension in 1 . 5 ml 25 mM HEPES ( pH 7 . 4 ) , 100 mM NaCl , and 1% CHAPS . The solution was clarified by centrifugation for 30 min in a microfuge at a maximum speed ( ∼14 , 000× g ) at 4°C . Glutathione Sepharose 4B ( GE Healthcare ) beads were added and incubated at 4°C for 3 hrs with mixing . The beads were then pelleted , washed twice in 25 mM HEPES ( pH 7 . 4 ) and 100 mM NaCl and eluted ( 2×0 . 5 mL ) in the same buffer containing 20 mM glutathione . The eluted protein was concentrated to 0 . 25 mL in a spin concentrator ( Amicon Ultra-4; EMD Millipore ) . The protein concentration was determined using a Bradford assay . Typical yields of preparations from 10×10 cm plates were ∼600 µg of GST-ELMOD3 or mutant and ∼6 mg of GST . The purified proteins were quick frozen and stored at −80°C . The Arl2 GAP assay was performed as described previously by Bowzard et al [29] . Briefly , 2 µM purified recombinant Arl2 , prepared as described by Clark et al [40] , was pre-loaded with [γ-32P]GTP in 25 mM HEPES ( pH 7 . 4 ) , 2 . 5 mM MgCl2 , 100 mM NaCl , 1 mM EDTA , 25 mM KCl , and 0 . 5 mM ATP in a total volume of 100 µL . The incubation was performed at 30°C for 30 min . The GAP reaction was performed in a buffer that contained 25 mM HEPES ( pH 7 . 4 ) , 2 . 5 mM MgCl2 , 100 mM NaCl , 1 mM dithiothreitol , 2 mM ATP , 1 mM GTP and the pre-loaded Arl2 [γ-32P]GTP . The total 50 µL reaction was initiated by the addition of the sample that contained the Arl2 GAP and stopped after 4 min at 30°C by the addition of 750 µL of ice-cold activated charcoal ( 5% activated charcoal ( Sigma-Aldrich , St . Louis , MO ) in 50 mM Na2HPO4 ( pH 7 . 4 ) . The samples were clarified by centrifugation , and 400 µL was taken for counting in a liquid scintillation counter . Each GAP sample was also assayed in parallel in a tube that contained all of the above reagents except Arl2 ( i . e . , the same amount of [γ-32P]GTP ) . The resulting “blank” values were subtracted from the results that were obtained in the presence of Arl2 to determine the amount of hydrolyzed 32Pi that was dependent on Arl2 GAP activity . More detail and descriptions regarding how the specific activities were calculated from this assay can be found in the report by Bowzard et al [29] . | Autosomal recessive nonsyndromic hearing loss is a genetically heterogeneous disorder . Here , we report a severe-to-profound mixed hearing loss locus , DFNB88 on chromosome 2p12-p11 . 2 . Exome enrichment followed by massive parallel sequencing revealed a c . 794T>C transition mutation in ELMOD3 that segregated with DFNB88-associated hearing loss in a large Pakistani family . This transition mutation is predicted to substitute a highly invariant leucine residue with serine ( p . Leu265Ser ) in the engulfment and cell motility ( ELMO ) domain of the protein . No biological activity has been described previously for the ELMOD3 protein . We investigated the biochemical properties and ELMOD3 expression to gain mechanistic insights into the function of ELMOD3 in the inner ear . In rodent inner ears , ELMOD3 immunoreactivity was observed in the cochlear and vestibular hair cells and supporting cells . However , ELMOD3 appears most pronounced in the stereocilia of cochlear hair cells . Ex vivo , ELMOD3 is associated with actin-based structures , and this link is impaired by the DFNB88 mutation . ELMOD3 exhibited GAP activity against Arl2 , a small GTPase , providing a potential functional link between Arf family signaling and stereocilia actin-based cytoskeletal architecture . Our study provides new insights into the molecules that are necessary for the development and/or function of inner ear sensory cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | An Alteration in ELMOD3, an Arl2 GTPase-Activating Protein, Is Associated with Hearing Impairment in Humans |
During embryogenesis , the neural stem cells ( NSC ) of the developing cerebral cortex are located in the ventricular zone ( VZ ) lining the cerebral ventricles . They exhibit apical and basal processes that contact the ventricular surface and the pial basement membrane , respectively . This unique architecture is important for VZ physical integrity and fate determination of NSC daughter cells . In addition , the shorter apical process is critical for interkinetic nuclear migration ( INM ) , which enables VZ cell mitoses at the ventricular surface . Despite their importance , the mechanisms required for NSC adhesion to the ventricle are poorly understood . We have shown previously that one class of candidate adhesion molecules , laminins , are present in the ventricular region and that their integrin receptors are expressed by NSC . However , prior studies only demonstrate a role for their interaction in the attachment of the basal process to the overlying pial basement membrane . Here we use antibody-blocking and genetic experiments to reveal an additional and novel requirement for laminin/integrin interactions in apical process adhesion and NSC regulation . Transient abrogation of integrin binding and signalling using blocking antibodies to specifically target the ventricular region in utero results in abnormal INM and alterations in the orientation of NSC divisions . We found that these defects were also observed in laminin α2 deficient mice . More detailed analyses using a multidisciplinary approach to analyse stem cell behaviour by expression of fluorescent transgenes and multiphoton time-lapse imaging revealed that the transient embryonic disruption of laminin/integrin signalling at the VZ surface resulted in apical process detachment from the ventricular surface , dystrophic radial glia fibers , and substantial layering defects in the postnatal neocortex . Collectively , these data reveal novel roles for the laminin/integrin interaction in anchoring embryonic NSCs to the ventricular surface and maintaining the physical integrity of the neocortical niche , with even transient perturbations resulting in long-lasting cortical defects .
The cues responsible for maintaining the physical and molecular architecture of the stem cell niche of the developing mammalian brain are not well known . In the mammalian neocortex , the radial glia neural stem cells ( NSC ) that generate neurons are bipolar and have a radial morphology that spans the developing neocortical wall [1] , [2] . These NSC have their soma located within the ventricular zone ( VZ ) adjacent to the ventricle , and their apical and basal processes make contact with the ventricular surface and the pial basement membrane , respectively [3] , [4] . The basal ( pial ) process is important for informing the fate of the NSC daughter cells and then acting as a guidepost for their migration [5]–[7] . In contrast , the apical process contains cilia that extend into the ventricle and are thought to be important for morphogen signalling within the VZ microenvironment or niche [8] , [9] . In addition , the apical process is necessary for the interkinetic nuclear migration ( INM ) , which takes place during VZ cell proliferation [4] , [10] , and its transection results in translocation of the NSC soma away from the ventricular surface [7] . The apical processes of adjacent radial glia cells are attached to one another via cadherin-based adherens junctions that are Numb and Numbl-dependant [11] , [12] . Recent studies have also highlighted the importance of Cdc42 [13] , αE-catenin [14] , β-catenin [15] , [16] , and the adenomatous polyposis coli protein ( APC ) [17] in the maintenance of this morphology . Deletion of Cdc42 as well as Numb/Numbl in NSCs disrupts apical adherens junctions resulting in defects in cell proliferation and disorganized cortical lamination [12] , [13] . Likewise , αE-catenin and β-catenin , components of adherens junctions , regulate NSC cell cycle progression and thereby cerebral cortical size [14]–[16] , [18] . Recently , APC has been shown to regulate the development and maintenance of the radial glial scaffold during corticogenesis [17] . However the specific adhesive molecules required for anchorage of these interconnected apical processes within the ventricular niche and their impact on neocortical development have not yet been determined . The integrin α6β1 heterodimer is expressed at high levels in the apical regions of NSC [19] . Laminins , which serve as ligands for integrins in the extracellular matrix ( ECM ) , are also present in the VZ niche [19] , suggesting a possible role of laminins and integrins in providing these adhesive signals for NSC within the VZ . However , previous studies examining conditional deletion of β1 integrin in NSC [20] or α6−/− mice [21] did not report abnormalities of NSC behaviour in the VZ . We reasoned that this might reflect compensation for the long-term loss of one integrin by other heterodimer combinations , as has been described for β subunit integrin mutants in Drosophila midgut development [22] . To test the potential role of laminin/integrin binding in VZ maintenance and proliferation , we circumvented this possible compensation by transiently disrupting β1 integrin/laminin binding specifically in the VZ using blocking antibodies injected into the ventricle of the embryonic mouse brain . We also developed a novel ex vivo multiphoton time lapse imaging method that enables the effect of targeting of the blocking antibody to the cortical niche to be seen in real time . Furthermore , we analyzed VZ cell morphology and proliferation in laminin α2 deficient embryos . Together , our data demonstrate a novel role for laminin/integrin binding in the regulation of NSC proliferation and adhesion within the embryonic VZ , as well as its requirement to maintain the architecture of the neocortical niche .
While β1 integrin ( accession number Swiss Prot P09055 , http://www . ebi . ac . uk/swissprot ) has previously been shown to be present in the VZ of the developing cortex [19] , [20] , [23] , we confirmed the expression levels in the neocortical wall on the embryonic days at which we performed the perturbation studies . At E13 . 5 , there is a high level of β1 integrin in the VZ , as shown by double labelling with a mitotic marker of M-phase , phospho histone 3 ( PH3 , Figure 1A and 1B ) . The high level of β1 integrin continues into the cortical subventricular zone ( SVZ ) as marked by the second layer of PH3+ cells , and β1 integrin is also highly expressed at the pial surface and in blood vessels ( Figure 1A and 1B ) . Importantly , there are particularly high levels of β1 integrin on the apical surface of the VZ and on radial glia apical fibers ( as assessed by double labelling with RC2 , Figure 1E–1J ) . Analysis of the subcellular localization of β1 integrin within the ventricular processes reveals that this receptor is mainly located immediately basal to the adherens junctions ( Figure S1 ) . At E16 , as large numbers of neurons begin to differentiate in the cortex , the level of β1 integrin remains high in the VZ/SVZ but decreases in the neuronal layers ( Figure 1C and 1D ) . Because β1 integrin is also expressed at the pial surface , where it is involved in the organization of the cortical marginal zone [20] , one major challenge was to preferentially inactivate β1 signalling only at the apical surface to determine the particular contribution of β1 integrin in the cellular dynamics that take place in the VZ during formation of the neocortical wall . To accomplish this , we delivered a blocking antibody ( Ha2/5 ) [24] into the cerebral ventricle in utero to specifically block β1 integrin function in cells bordering the ventricle . To determine the efficacy of this approach , we first assessed the in vivo dynamics of the antibody by injecting fluorescently conjugated Ha2/5 into the ventricles of E14 mice in utero . We observed widespread localization of the antibody within the VZ and SVZ after 6 h ( Figure S2B ) and 24 h ( unpublished data ) . The antibody penetration was confined to apical regions and did not reach the pial surface . To confirm in vivo that the antibody inhibited integrin signalling , we evaluated phospho-Akt 1 ( p-Akt 1 ) levels in the cortex 30 min after injection . p-Akt-1 is known to be highly expressed in NSC and is a well-recognized downstream signalling molecule in the integrin pathway [25] , [26] . Brain lysates were prepared from E12 and E15 embryos injected either with the Ha2/5 or with an isotype control ( ITC ) antibody . Western blotting revealed a reduced level of p-Akt 1 expression 30 min after injection ( Figure S2C ) . By 2 h after antibody injection , the differences in p-Akt 1 levels were absent ( unpublished data ) demonstrating that the perturbation of β1 integrin signalling is transient . We took advantage of the movements of the NSC soma , which take place during cell cycle progression , to determine whether β1 integrin signalling affects the positioning of the NSC . Mitosis in the VZ normally occurs on the ventricular surface , after which the NSC soma transitions to the abventricular side of the VZ before entering S-phase ( which can be identified by the incorporation of 5-bromo-2-deoxyuridine [BrdU] into the newly synthesized DNA ) . This cell cycle-dependent nuclear movement is known as INM [27] . Injection of 10 ng of the β1 integrin blocking antibody into the lateral ventricles of E12 . 5 and E15 . 5 embryos disrupted this pattern 18 h after injection , with mitotic PH3+ cells now scattered throughout the VZ ( Figure 2A and 2B ) . Injection of a higher concentration of Ha2/5 ( 100 ng ) produced identical results ( unpublished data ) . Quantitative analysis ( Figure 2C and 2D ) revealed a significant increase in the number of PH3+ cells away from the ventricular surface ( nonventricular surface or nVS ) in both E12 . 5- ( Figure 2C; p<0 . 01 , unpaired two-tailed t-test ) and E15 . 5-injected brains ( Figure 2D; p<0 . 001 , unpaired two-tailed t-test ) without any changes seen in the number of PH3+ cells at the ventricular surface . Due to INM , a short 1 h pulse of BrdU normally labels cells clustered in S-phase at the abventricular boundary of the VZ ( Figure 2E ) . Indeed , a maximum labelling index ( calculated by the percentage of BrdU+ cells ) was observed 60–70 µm away from the ventricular surface in the embryos injected with ITC antibodies ( Figure 2G ) . In contrast , perturbation of β1 integrin signalling shifted the maximum labelling index 80–110 µm away from the ventricular surface ( Figure 2F and 2G; p<0 . 05 , two-way ANOVA ) , and also resulted in an overall increase in the number of BrdU+ cells . Thus in addition to the PH3+ mitotic cell ectopias , supernumerary S-phase cells were found in abnormal positions following β1 integrin blockade . To investigate somal translocation towards the VZ surface ( i . e . , M-phase reentry ) , E15 . 5-injected embryos were pulsed with BrdU 6 h prior to sacrifice , allowing the majority of proliferative cells to transit through S-phase and be on the ventricular surface either in or approaching mitosis at the time of analysis . In the embryos injected with the ITC antibody , this nuclear movement was indeed observed with the highest labelling index occurring in bin 1 nearest the ventricle ( Figure 2H ) . While bin 1 also contained the highest labelling index in Ha2/5-injected brains , there was a significant increase in the labelling index of abventricular bins ( bins 7–20 ) compared to controls . Thus , although continued ventricular divisions were apparent following blockade of β1 signalling , the abventricular dividing progenitor population was significantly increased . To determine whether changes in the SVZ may be related to the mode of division in the VZ , we analyzed the distribution of cleavage plane angles . It has previously been shown that the mitotic spindle undergoes significant rotation during metaphase [28]–[30] , leading to changes in the cleavage orientation of mitotic figures , which may be an indication of cell fate [29]–[33] . In addition , β1 integrin signalling has previously been shown to affect mitotic spindle formation in Chinese Hamster Ovary ( CHO ) cells in vitro [34] . We therefore assessed the orientation of cell divisions in the VZ 18 h after antibody injection and found that the β1 integrin blocking antibody caused a significant change in the pattern of cell divisions ( Figure 3 ) . In the developing telencephalon , the majority of mitoses occur vertically with cleavage angles greater than 60 degrees relative to the ventricular surface [30] , [31] , although a small percentage ( 15%–20% ) can be seen with lower degrees of cell division ( i . e . , horizontal divisions ) . Indeed , this is what was observed with the ITC-injected embryos 18 h after antibody injection on E12 . 5 , E13 . 5 , E14 . 5 , and E15 . 5 at rostral , medial , and caudal levels of the telencephalon ( Figure 3A , 3C–3E ) . However , there was a reduction in the amount of ( horizontal ) cell divisions with cleavage angles below 60 degrees in the β1 integrin antibody injected embryos in the medial and caudal regions of the dorsal telencephalon ( Figure 3C–3E ) . Using a statistical model to analyze the distribution of the VZ cleavage angles throughout neurogenesis ( from E13 to E16 ) , we found that the proportion of horizontally dividing VZ cells ( 0–30 degrees ) is significantly lower at the medial and caudal levels of the forebrain after disruption of β1 signalling ( Figure S3A ) . Interestingly , the effect of blockade of β1 integrin signalling was not seen in neural precursors located rostrally ( Figures 3C , 3E , and S3 ) . Together , these data indicate that the ventricular divisions that remain following β1 integrin blockade exhibit altered cleavage parameters coincident with the increased number of abventricularly proliferating cells . Our BrdU and cell division studies clearly demonstrate that disruption of β1 integrin signalling leads to the presence of ectopic mitotic cells . We considered the possibility that these positioning defects lead to precocious differentiation . First , we determined the effects of β1 integrin blockade on the number of intermediate progenitor cells ( IPC ) , which express the transcription factor T-brain 2 ( Tbr2 ) [35] . IPC are neuronal progenitors that are generated from the NSC in the VZ , and which undergo further rounds of division just outside of the VZ in the SVZ [2] , [35] . We found no difference in the number of Tbr2+ cells between ITC and Ha2/5-injected brains at E13 ( Figure S4A–S4D ) . In addition , no premature neuronal ( β3 tubulin , Figure S4E and S4F ) or glial ( NG2 , unpublished data ) differentiation was detected in the neocortical wall . Thus , β1 integrin blockade did not lead to abnormalities in cell differentiation within 18 h , and notably , although the number of proliferating cells in abventricular positions was increased , we found no increase in the number of Tbr2+ IPCs . To determine whether the cell positioning defects following β1 integrin blockade are due to disruption of NSC morphology , we simultaneously performed electroporation of an RFP-expressing plasmid ( CAG-RFP ) with the Ha2/5 antibody injection to fluorescently label a population of VZ cells at the time of β1 integrin inhibition ( Figure 4 ) . We used electroporation parameters previously shown to transfect only VZ cells [36] and determined whether cells had detached from the ventricle surface within 18 h of co-electroporation/injection ( Figure 4A and 4B ) . To do this , sections were stained with phalloidin to label the actin ring at the border of the NSC apical membranes so that the apical processes attached at the ventricular surface could be unambiguously identified ( Figure 4C–4F; further 3-D examples can be seen in Figure S5 and Video S1 ) . Volumetric reconstructed slices were created by image analysis and both the numbers of cell soma and apical processes were counted to generate a soma∶process ( S∶P ) ratio ( Figure 4C–4H ) , and the percentage of apical processes still attached at the ventricular surface was also determined ( Figure 4I and 4J ) . There was a significant difference between the two groups ( * , p<0 . 05 , unpaired two-tailed t-test ) , with ITC-injected embryos having a lower S∶P ratio and a higher percentage of apical processes in contact with the ventricular surface compared to the brains injected with the β1 integrin blocking antibody at both E13 . 5 ( Figure 4G and 4I ) and E15 . 5 ( Figure 4H and 4J ) . This 3-D analysis therefore identified a morphometric change in neocortical VZ cells following Ha2/5 injection , with β1 integrin blockade resulting in detachment of apical processes from the ventricular surface as shown in Figure 4A . Furthermore , we also identified many dystrophic ascending basal processes emanating from the VZ cells ( Figure 4A and 4B ) indicating that apical detachment has widespread morphological effects on VZ cells and may therefore adversely affect neuronal migration . To visualize the impact of β1 integrin signalling blockade on VZ cell morphology in real time , we performed time lapse imaging of neocortical VZ cell dynamics in living slices following electroporation with farnesylated enhanced green fluorescent protein ( eGFP-F ) ( Figure 5 ) . To specifically block β1 integrin at the apical surface without disrupting its function at the pial surface , we applied a drop of growth factor-reduced matrigel containing the antibody ( β1 integrin blocking or ITC control ) inside the lateral ventricle of the living slices prepared from E14 . 5 embryos electroporated 24 h earlier ( Figure 5A ) . Analysis of the diffusion of β1 integrin blocking antibody-FITC from the drop of matrigel into the neocortical wall via pixel intensity profiles revealed that β1 integrin blocking antibody is mainly present in the 1/5 of the neocortical wall next to the ventricle in living slices; this corresponds to the VZ/SVZ compartment and indicates that , as with the in utero experiments , the blocking antibody does not reach the pial surface ( Figure S6 ) . This enabled us to monitor the effect of localized antibody blockade on eGFP-F+ cell morphology in the slices . After 10 h of contact with the antibody , we noted the progressive bending of both basal and apical processes ( red arrows , Figure 5C and Video S3 ) as well as the detachment of apical end-feet from the ventricular surface ( red arrowheads , Figure 5C and Video S3 ) . In contrast , in the control experiment , NSC apical and basal processes are unaffected ( Figure 5B and Video S2 ) . Collectively , these data demonstrate that β1 integrin signalling disruption at the ventricular surface results in a progressive destabilization of the VZ architecture due to the simultaneous loss of both NSC bipolar morphology and apical end-feet at the ventricular surface . The laminin α2 chain ( accession number Swiss Prot Q60675 ) mediates cell adhesion through β1 integrins [37] and is expressed at the embryonic ventricular surface [19] . Thus , laminin α2 chain-β1 integrin interactions may be involved in the NSC adhesion at the ventricular surface during corticogenesis . To test this possibility , we analyzed cell proliferation and mitotic cleavage parameters in the VZ of laminin α2 deficient mice ( Lnα2−/− mice ) [38] . As with β1 integrin blockade , more proliferating cells were present outside the VZ/SVZ after a 1 h BrdU pulse in Lnα2−/− embryos ( Figure 6A–6C ) . Furthermore , the angle of VZ cell division was also altered in Lnα2−/− embryos with the proportion of horizontal divisions ( 0–30 degrees ) significantly lower in the medial region of the telencephalon ( Figures 6D and S3B ) . Using the same experimental paradigms as in the β1 integrin blockade experiments , we performed in utero electroporation of the CAG-RFP plasmid in E15 . 5 Lnα2−/− mutant embryos ( Figure 6F ) and control wild-type littermates ( Figure 6E ) . We then quantified the numbers of cell soma and apical processes and determined both the S∶P ratio ( Figure 6G ) and the percentage of apical processes ( Figure 6H ) still in contact with the ventricular surface . There was a significant difference between the two groups ( * , p<0 . 05 , unpaired two-tailed t-test ) with a higher S∶P ratio in the Lnα2−/− mutants compared to the controls , consistent with an apical detachment of electroporated NSC . These results show that disruptions of either β1 integrin or of a ligand expressed in the VZ lead to identical alterations in cell position , NSC proliferation , orientation of cell division , and apical process detachment . These results with the Lnα2−/− mice therefore identify laminin α2 as a key ligand for the integrins expressed in the VZ and thus provide a genetic corroboration of our antibody perturbation studies . To investigate the long term consequences of β1 integrin blockade and detachment of VZ cells on neocortical morphogenesis and layer formation , we utilized the co-electroporation/antibody injection strategy to mark cells at E15 . 5 and then allowed cortical development to proceed until postnatal day ( P ) 4 . We reasoned that VZ cell detachment may lead to disruption of cortical layering since the detached NSCs with dystrophic radial fibers that we observed in the short-term experiments would not generate the proper amount of committed neurons and would alter the migration route to the cortical plate . Indeed , we found a reduction in the width of cortical layers I-V ( Figure 7A–7C ) , as well as in the radial distribution of RFP+ cells in the somato-sensory cortex following β1 integrin antibody injection ( Figure 7D–7F ) . Interestingly , in keeping with the rostro-caudal differences in β1 integrin blockade described in Figure 3 spatial discrepancies were also found in the postnatal cortex of animals injected with β1 integrin blocking antibody at E15 . 5; cortical layer thickness was reduced in somato-sensory but not in the primary motor cortex , although these results did not reach statistical significance . These results therefore provide evidence that proper maintenance of apical process attachment during embryogenesis is critical not only for INM and NSC proliferation , but also for neuronal migration and cortical cell layer formation , as a result of which transient disruption of β1 integrin signalling can have long lasting effects .
Using a multidisciplinary approach that includes cellular/molecular analysis and multiphoton time lapse imaging , we have revealed a hitherto unsuspected role for β1 integrin during neocortical development . Previously , β1 integrin has been suggested to be important for neocortical formation through its regulation of the radial glial contacts on the pial basement membrane [20] , [39] . In our study , we combined in utero electroporation and injection of a specific blocking antibody to specifically inactivate the β1 integrin receptor by preventing binding to its ligand ( laminin ) at the ventricular surface . Compared to transgenesis or siRNA knock down , which cause widespread effects throughout both cell and tissue , this novel approach resulted in a focused and transient disruption at the subcellular level that resulted in detachment of the apical processes of many NSC from the ventricular surface and led to increased numbers of ectopic proliferating cells as well as perturbations to INM . Confirming that integrins act at least in part through interactions with laminins in the neocortical VZ , we found similar abnormalities in the laminin α2-deficient mouse . Together , our data clearly demonstrate for the first time in vivo , to our knowledge , that integrin/laminin interactions at the apical VZ surface play a critical role in the adhesion that maintains the stem cells within their niche and preserves the architecture of the VZ . The adhesion of stem cells to their niche is critical for the molecular programmes that promote maintenance . For example , altered expression of adhesion-related genes is known to cause depletion of haematopoietic and epidermal stem cell niches [40] , [41] . A recent report has also shown that niche-supporting gonad cells in Drosophila also require integrin signalling to ensure niche integrity [42] . Although the VZ lacks a basal lamina , which is well recognized as a principal site of cell/extracellular matrix interactions , we have shown previously that both laminins ( α2 , α4 , and α2 chains ) and integrins are expressed at the apical surface of the neocortical wall in the embryonic mouse VZ [19] . Our present observations suggest that integrin/laminin interactions are necessary to enable the retention of apical processes seen for at least 5 h after mitosis , and which may be critical for key cell-cell interactions that instruct behaviour [6] . So , while no perturbation of cell differentiation following β1 integrin blockade has been detected in our study , premature loss of these interactions resulting from apical process detachment has profound consequences on other aspects of NSC behaviour , including dysregulated proliferation of the NSC and altered allocation to the developing cortical plate . Interestingly a recent investigation of the role of α6β1 integrin in the adhesion of adult SVZ progenitor cells to endothelial cells using the same in vivo blocking antibody paradigm demonstrated two similar phenotypes to those we observed in the embryo—separation of SVZ progenitor cells from their normal location ( adjacent to blood vessels ) and enhanced proliferation [43] . Integrin/laminin interactions may therefore play similar roles in the regulation of neural stem and progenitor behaviour in embryonic and adult central nervous system . Whether blood vessels in the embryonic NSC niche provide some of these laminins as they do in the adult remains unknown , but recent studies do suggest a critical role of the developing cortical vasculature in regulating cortical neurogenesis [44] , [45] and laminin α2 is expressed in blood vessels in the embryonic VZ [19] . Further work to test the hypothesis that laminin/integrin interactions in the vicinity of blood vessels contribute to the embryonic niche as they do in the adult is therefore required . The two other phenotypes we observed after disruption of integrin signalling in the VZ , the loss of the subset of VZ cells that divide with horizontal cleavage planes and abnormal cortical layer formation may not simply be explained by an effect solely on VZ adhesion . Under normal circumstances , horizontal cleavages are the minority , and daughter cell fate cannot be predicted solely by the cell division orientation of its parent cell [31] , [46] , [47] , because cells undergoing vertical cleavage during mitosis can give rise to either identical ( via symmetric division ) or different ( via asymmetric division ) daughter cells [48] . However , several recent studies have suggested an important link between the precise regulation of mitotic spindle orientation and the fate of neocortical neural progenitors . In particular , disruptions of centrosomal proteins such as Aspm [33] , Nde1 [49] , doublecortin-like kinase [50] , and Cep120 [51] , all of which play crucial roles in mitotic spindle function , affect the neural progenitor pool size and lead both to alterations in INM [51] and reductions in cerebral cortical size [49] . Although the link between mitotic spindle orientation and daughter cell fate is still debated , several recent studies demonstrate the close relationship between VZ cell cleavage angle and the location of the resulting daughter cells ( i . e . , ventricular surface versus abventricular location ) [46] , [48; present study] . While the detailed molecular mechanisms by which the integrin/laminin interaction influences the NSC cleavage orientation are still not known , compelling data linking integrin signalling to spindle assembly have already been reported for Chinese Hamster Ovary cytokinesis in vitro [34] . Thus our present results extend the importance of this role by demonstrating that β1 integrin signalling is required for the regulation of NSC mitotic spindle dynamics for the cells that normally undergo oblique cleavages during neocortical neurogenesis in vivo ( Figure 8 ) . Furthermore , our data suggest regional differences in that medial and caudal telencephalic progenitors are most sensitive to β1 integrin signalling . In keeping with this , it is interesting to note that human congenital muscular dystrophy caused by deficiency of the laminin α2 chain has been associated with significant abnormalities of cortical development in the occipital but not frontal regions of the telencephalon [52] . The thinning of the cortical layers that we observed in the postnatal mouse brain following transient blockade of integrin signalling in the embryo might reflect the alterations in the plane of cell division and subsequent effects on neurogenesis , but a key observation argues against this . We found that the NSC proliferation and morphological defects occurring after β1 integrin blockade had long-term consequences on the migration of both the newly formed neurons as well as those previously generated before the antibody injection . For example , the deep cortical layers ( IV and V ) , which contain neurons born before the perturbation on E15 . 5 , were also thinner than in controls . This deep layer defect is not likely to be caused solely by a disruption in neurogenesis at midgestation , since the earlier born neurons would be expected to establish proper laminar positions . Rather , it points to a phenotype resulting from the dystrophic radial glia processes we observed in the antibody-injected tissue , with cells born several days prior to the injection and still en route to the cortical plate affected by the morphological changes to the radial glia in the VZ . The most parsimonious explanation is that the loss of apical adhesion leads to NSC detachment and shortening and dystrophy of the basal process , and this in turn perturbs the migration of the cells on these processes . Supporting this proposed mechanism , several reports have demonstrated that mechanical forces play an important role in shaping the developing brain [7] , [53] , [54] . The present data suggest that anchorage of the apical endfeet provides the physical tension required for maintenance of position and morphology of radial glia cells during corticogenesis . Thus , our data using short-term blocking approaches reveal functions not shown by knock-out experiments and clearly define the novel contribution of integrins to neocortical development by elucidating a number of key roles in the regulation of NSC behaviour in the mammalian VZ .
Control ICR mice were produced in the Children's National Medical Center ( CNMC ) animal facility . ICR ( CNMC ) , C57/BL6 ( National Institute on Aging , NIA ) , and laminin α2 deficient mice ( Oregon Health and Science University , OHSU and also SUNY Stony Brook , NY ) were housed under standard conditions with access to water and food ad libitum on a normal 12 h light/dark cycle . Genotyping for the laminin α2 deficient mice was performed as previously described [38] . 6- or 20-µm thick coronal sections from frozen Tissue-Tek embedded embryonic brains were harvested from three different levels along the rostro-caudal axis depending on the experiment . Most of the studies focused on the medial region of the forebrain , corresponding to E13/E14 plate 4 ( for E13 to E14 embryos ) or E15/E16 plate 5 ( for E15 to E16 embryos ) in the atlas by Jacobowitz and Abbott ( 1998 ) . The analysis of the cleavage plane angle was also performed on the rostral ( E13/E14 plate 3; E15/E16 plate 3 ) and caudal ( E13/E14 plate 5; E15/E16 plate 8 ) regions of the embryonic forebrain [55] . Intraventricular injections were done with approval from the NIA and CNMC Institutional Animal Care and Use Committees using methods described previously [36] . Briefly , timed-pregnant mice ( from E12 to E15 ) were anesthetized with ketamine/xylazine and a midline laparotomy was performed exposing uterine horns . The lateral ventricle in the brain of each embryo was visualized with transillumination and the injections were performed with a glass capillary pipette ( 75–125 µm outer diameter with bevelled tip ) driven either by a Sutter micromanipulator ( Sutter Instrument Company ) equipped with 20-µl Hamilton gas-tight syringe or a nitrogen-fed Microinjector ( Harvard Apparatus ) . For integrin blocking studies , approximately 1 µl of either β1 integrin blocking antibody ( 10 ng or 100 ng; Ha2/5 , with or without FITC conjugation , BD Pharmigen ) or an ITC antibody ( anti-hamster , BD Pharmigen ) solution ( combined 3∶1 with sterile fast green dye to enable monitoring of the injection into the cerebral ventricles , Sigma ) was injected alone or mixed with DNA . Two different plasmid vectors were used: a plasmid encoding red fluorescent protein under the control of the chicken β actin promoter ( CAG-RFP ) and a plasmid expressing eGFP-F ( Clontech ) . For the in utero electroporation procedure , the anode of a Tweezertrodes ( Genetronics ) was placed above the dorsal telencephalon and four 40-V pulses of 50 ms duration were conducted across the uterine sac . Following intrauterine surgery , the incision site was closed with sutures ( 4-0 , Ethicon , ) and the mouse was allowed to recover in a clean cage . Mice were humanely killed 8–24 h after the injection unless indicated otherwise and embryonic brains were harvested . Organotypic slices were prepared 24 h after in utero electroporation performed on E14 . 5 brains with an eGFP-F plasmid as described previously [30] , [56] . Briefly , 300-µm-thick slices containing EGFP-F+ cells were collected in ice-cold Complete Hank's Balanced Salt Solution using a vibrating microtome ( Leica VT1000S ) and transferred into serum-free medium ( SFM; neurobasal medium supplemented with B27 , N2 and glutamax; Invitrogen ) . After 1 h of recovery , the slices were placed in a 35-mm glass bottom culture dishes . ITC control or β1 integrin blocking antibodies were diluted ( 1∶100 ) in growth factor-reduced matrigel ( BD Biosciences ) and a drop ( 0 . 5 µl ) of this solution carefully introduced in the ventricular space of the embryonic brain slices . A slice holder immobilized the slices and 3 ml of SFM were added . The 35-mm glass bottom culture dishes containing the slices with matrigel were positioned in a heated micro-incubation chamber ( DH-40i; Warner Instruments ) . Preheated SFM was pumped over the slices for the length of the imaging experiment ( usually 10 h ) , the slice temperature was maintained at 37°C and the imaging preparation was maintained in 5% CO2/95% air for the entire period . All multiphoton imaging was performed on a Zeiss LSM 510 Meta NLO system equipped with an Axiovert 200M microscope ( Zeiss ) direct coupled to a Mira 900F laser pumped by an 8-W Verdi laser ( Coherent Laser Group ) . EGFP was excited at 850 nm and time-series experiments were conducted under oil-immersion with 25× objective . Time-series images consisted of 40-µm-thick z-stacks and were collected at multiple locations at 2 min intervals to repetitively record both β1 integrin blockade and control slices . The experiments were analyzed with LSM 510 software . For the presentation of videos , each z-stack was projected onto one optical slice per time period and the resulting frames were assembled and compressed using Volocity software ( Improvision ) . For analysis of the diffusion of the blocking antibody in the experiments using matrigel to deliver antibodies within organotypic slices , these slices were prepared as above . After 10 h of incubation with the ITC control or FITC-labelled β1 integrin blocking antibodies in a drop of growth factor-reduced matrigel placed in the ventricular cavity , slices were fixed in 4% PFA and nuclei stained with dapi . 20-µm-thick z-stacks were collected and were analyzed with LSM 510 software . Each neocortical length was divided in five bins , each representing 20% of the total cortical thickness . The pixel intensity calculated by the LSM software was summed for each bin and then averaged and plotted in a graph ( Figure S6 ) . 20-µm-thick coronal sections were imaged using a Zeiss LSM 510 NLO system direct coupled to an inverted Axiovert 200M microscope ( Zeiss ) . 25× ( DIC , 0 . 8 na; Zeiss ) image stacks ( 1-µm intervals ) containing the region/cells of interest were collected with conventional detectors and then analyzed post hoc . For the orientation of cell division , studies were conducted with a 40× oil-immersion lens ( DIC , Plan Neofluar , 1 . 3 na; Zeiss ) . Each frame of the series , consisting of a z-stack of images , was reconstructed in 3-D using Zeiss LSM software and was then rotated around the y-axis to bring the edge of the mitotic figures at the VZ surface into view so that the mitotic spindle plane was parallel to the computer screen . The angle of the mitotic spindle was then measured by projecting a line through the spindle to a reference line parallel with the ventricular surface . This procedure was repeated for each mitotic figure in each frame from the beginning of metaphase ( a discrete organized metaphase plate ) until the beginning of chromatid separation in anaphase . The spindle angles were then documented manually and graphed using SigmaPlot software . The 3-D reconstruction of CAG-RFP cell attachment to the ventricular surface labelled with phalloidin was performed using Volocity software ( Improvision ) . For integrin signalling validation , half the litter was injected with the blocking antibody and the other half with the control antibody for 30 min . Telencephali were isolated by rapid dissection 30 min after injection and then flash frozen . Total brain lysates were prepared by resuspending the tissue in cell lysis buffer . Tissue protein was extracted using T-PER tissue protein extraction buffer with protease inhibitor cocktail ( Sigma ) and protein concentration was determined by the BCA protein assay kit ( Pierce ) . 50 µg of protein was separated by SDS-PAGE ( 8%–12% ) and transferred to nitrocellulose membranes . The membranes were blocked in 5% nonfat milk for 1 h at room temperature , followed by an overnight incubation at 4°C with antibodies raised against p-Akt1 ( BD Pharmigen ) , total Akt ( T-AKT , BD Pharmigen ) , or β-actin ( Sigma ) . Membranes were then washed and incubated with secondary antibodies for 1 h at room temperature . Protein bands were visualized using a chemiluminescence detection kit ( Amersham Biosciences ) . Embryonic brains were fixed in 4% paraformaldehyde ( PFA ) in PBS overnight at 4°C before being transferred to sequential 20% and 30% solutions of sucrose ( w/v ) and left at 4°C overnight or until the brains equilibrated . The brains were then embedded in TissueTek ( Sakura ) prior to cryostat sectioning ( Leica CM3050S ) . For immunofluorescence , sections were blocked for a minimum of 30 min in PBS containing 0 . 1% Triton X-100 and 10% normal goat serum ( Sigma ) . Sections were incubated overnight with primary antibodies at 4°C . After incubation with the appropriate secondary antibodies and counter-staining with 4′ , 6-diamidino-2-phenylindole dihydrochloride ( Dapi , Sigma ) to visualize the DNA , images were acquired using an Olympus IX50 fluorescence microscope . Images were processed using MagnaFire and Photoshop 6 . 0 ( Adobe ) and adjusted such that the entire signal was in the dynamic range . The following antibodies were used for immunofluorescence: anti-β1 integrin ( used 1∶100 in blocking buffer ) , anti-Tbr2 ( used 1∶200 in blocking buffer following a 5 min boil in 10 mM sodium citrate , Millipore ) , anti-RC2 ( used 1∶5 in blocking buffer , Developmental Studies Hybridoma Bank ) , anti-phospho histone H3 ( used 1∶500 in blocking buffer , Millipore ) , anti-BrdU ( used 1∶5 in BrdU blocking buffer , Accurate Chemicals ) , anti-β3 tubulin ( used 1∶500 in blocking buffer , Sigma ) . For BrdU staining , the blocking buffer consisted of DMEM ( Sigma ) supplemented with 1% tween-20 ( Sigma ) and 7 mg of DNAse ( Sigma ) per 1 ml of blocking solution . The angle of the cleavage plane was determined in cells in anaphase identified by propidium iodide staining performed on 20-µm-thick sections from three different levels ( rostral , medial , and caudal ) of the forebrain from E13 to E16 embryos . F-actin filaments were visualized at the ventricular surface using Alexa Fluor 488 phalloidin ( 165 nM final concentration , Molecular Probes ) by incubation for 1 h after a prior 5 min incubation in 0 . 1% Triton X-100 in PBS and 30 min in 10% normal goat serum in PBS for blocking . Nuclear counterstaining was performed by 10 min incubation at room temperature in TOPRO-3 iodide ( 1∶100 , Molecular Probes ) . For the subcellular localization analysis of β1 integrin , first a postfixation with methanol at −20°C for 10 min was performed on the cryosections before a blocking step in a solution containing bovine serum albumin 3% and Tween 0 . 05% for 1 h at room temperature ( RT ) . Then , anti-β1-integrin antibody ( clone MB1 . 2 from Chemicon Int . , 1/100 dilution ) and Alexa Fluor 546 coupled phalloidin ( Molecular Probes , 1/200 dilution ) were incubated overnight in the blocking buffer at RT . Alexa Fluor 488-conjugated donkey anti-rat antibody ( Molecular probes , 1/250 dilution ) was incubated for 1 h at RT along with Hoechst for nuclei staining . Images were captured with a Zeiss confocal using an oil immersion 63× objective with a zoom of 2 . The profile function of the Zeiss acquisition software was used to determine the fluorescence intensity of each marker at a defined xy position . For PH3 analysis performed 18 h following antibody ( ITC or β1 blocking ) injection at E12 . 5 , positive cells were calculated from the average of three sections from five separate embryos from two litters . For E15 . 5 injected embryos , positive cells were calculated from the average of three sections from 11 ( ITC ) or 17 ( β1 integrin blocking antibody ) separate embryos from nine litters . For BrdU analysis , labelling index was calculated on the basis of three sections from three embryos from 1 litter for both E12 . 5 and E15 . 5 injected embryos . For Tbr2 expression analysis , a 200 µm×100 µm ( width×height ) region adjacent to the ventricular surface was analyzed and an average was calculated on the basis of three sections from three embryos from one litter . All statistical analysis was performed using GraphPad Prism version 4 . 00 for Windows , GraphPad Software ( www . graphpad . com ) . The specific statistical test is indicated in both the text and figure legends . Apical process quantification was performed on 150-µm vibrating microtome-cut coronal sections of co-antibody injected/CAG-RFP electroporated E13 . 5 and E15 . 5 brains stained with phalloidin to label the ventricle surface . 80-µm z-stacks were collected in 2-µm steps with a 1 , 024×1 , 024 pixel frame size and each z-stack was analyzed with LSM examiner ( Zeiss ) and Volocity software ( Improvision ) to determine the number of RFP+ cell bodies within 200 µm from the ventricle surface and to determine the number of apical processes attached to the ventricle ( colocalized with the phalloidin staining ) . Both of these counts were calculated from the average of three embryos for each condition . The ratio soma/apical processes ( S∶P ) representing the total number of cell bodies divided by the number of apical processes was determined and the results were analyzed by unpaired two-tailed t-test . The postnatal phenotype of embryos co-injected/electroporated at E15 . 5 was assessed at P4 by determining the thickness of the individual cortical layers ( I , II/III/IV , V , VI as shown in Figure 7A ) and by the radial distribution ( i . e . , layer specification ) of RFP+ cells in multiple areas at both the primary motor and somato-sensory cortical levels . The data from two ITC and three β1 integrin blocking antibody injected animals were analysed by unpaired two-tailed t-test . In preparation of performing statistical analysis we checked assumptions of normality and homogeneity of variance , and found that the data for the orientation of cell division were not normally distributed and could not be transformed to achieve acceptable levels of normality to permit linear regression analysis . Thus , an ordinal logistic regression model was developed to estimate the tendency toward having greater angles of cleavage in one group ( β1 integrin blocking antibody injected brains or Lnα2−/− brains ) compared to another ( controls: ITC injected brains or wild type littermates from Lnα2−/− embryos ) . To perform these analyses , angles of cleavage were stratified as , <30 degrees , 30 to <60 degrees , and 60–90 degrees . The model , which included covariates to account for brain region and study group , enabled the study to estimate and compare differences in the frequency of angle of cleavage strata in one study group compared to another . The model adjusted variance estimates to account for the correlation between repeated measurements on the same embryo . | The developing cerebral cortex contains bipolar neural stem cells that span the cortical layers between the inner ventricular surface and the outer pial surface of the embryonic brain . The nuclei of these cells remain near the ventricular cavity , a microenvironment or niche thought to provide vital signals . It is not known how this inner end of the bipolar stem cell is held in place , or what would happen if its attachment to the inner surface were lost . Genetic manipulation can be used to disrupt candidate molecules involved in this adhesion , but this will affect all adhesion points and complicate the results . We have therefore developed an approach to target the stem cell attachments specifically at the ventricular surface by placing blocking antibodies directly into the ventricles of mouse embryos and then expressing fluorescent markers in the stem cells to see the effects of losing this attachment in living tissue . We examined laminins and integrins , whose expression and properties make them excellent candidates . Blocking integrin signalling by antibody application caused the inner end of the stem cells to rapidly detach and then undergo aberrant cell division . We also showed that a transient block of integrins ( for <2 hours ) resulted in permanent malformations of the cortical layers and disrupted neuronal migration . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/neuronal",
"and",
"glial",
"cell",
"biology",
"developmental",
"biology/stem",
"cells",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"neuroscience/neurodevelopment",
"developmental",
"biology/neurodevelopment",
"cell",
"biology/extra-cellular",
"m... | 2009 | β1 Integrin Maintains Integrity of the Embryonic Neocortical Stem Cell Niche |
The worldwide neglect of immunotherapeutic products for the treatment of snakebite has resulted in a critical paucity of effective , safe and affordable therapy in many Third World countries , particularly in Africa . Snakebite ranks high among the most neglected global health problems , with thousands of untreated victims dying or becoming permanently maimed in developing countries each year because of a lack of antivenom—a treatment that is widely available in most developed countries . This paper analyses the current status of antivenom production for sub-Saharan African countries and provides a snapshot of the global situation . A global survey of snake antivenom products was undertaken in 2007 , involving 46 current and former antivenom manufacturers . Companies producing antivenom for use in sub-Saharan Africa were re-surveyed in 2010 and 2011 . The amount of antivenom manufactured for sub-Saharan Africa increased between 2007 and 2010/11 , however output and procurement remained far below that required to treat the estimated 300 , 000–500 , 000 snakebite victims each year . Variable potency and inappropriate marketing of some antivenoms mean that the number of effective treatments available may be as low as 2 . 5% of projected needs . Five companies currently market antivenom for sale in Africa; three others have products in the final stages of development; and since 2007 one has ceased production indefinitely . Most current antivenom producers possess a willingness and capacity to raise output . However inconsistent market demand , unpredictable financial investment and inadequate quality control discourage further production and threaten the viability of the antivenom industry . Financial stimulus is urgently needed to identify and develop dependable sources of high-grade antivenoms , support current and emerging manufacturers , and capitalise on existing unutilised production capacity . Investing to ensure a consistent and sustainable marketplace for efficacious antivenom products will drive improvements in quality , output and availability , and save thousands of lives each year .
Since Edward Jenner's controversial inoculation of James Phipps with cowpox in 1796 , immunotherapy has developed into a diverse industry [4] . Calmette's groundbreaking work with equine antiserum resulted in the first , unrefined antivenom in 1894 . Pope's improvements to antivenom refinement in the 1930s were another major step forward in safety and potency of antivenom . Unfortunately , further advances since then have been limited . Despite snakebite being over-represented in morbidity and mortality tables [5] , investment in this type of immunotherapy has not been characterised by the same level of publicity or resolve that has characterised vaccine production or monoclonal antibody research . This under-recognition of bites and stings as major medical and social problems , and snakebite's association with poverty , have contributed to the current antivenom crisis [6] . The introduction of antivenom to Africa in the 1950s heralded a decline in morbidity and mortality from snakebites that led to its widespread use and production . Sadly , over the last 30 years , production of this life-saving medication has been neglected by governments and non-government organisations , and abandoned by some manufacturers [7] . The 1970s and 1980s were characterised by a decline in the sale of antivenom in Africa due to growing neglect and prohibitive costs [8] . By 1998 , it was estimated that fewer than 100 , 000 vials of antivenom were available across Africa , constituting less than 25% of the amount needed [9] . A number of recent publications state the availability of antivenom in Africa has reduced to <1% of what is needed , or “<20 , 000 vials reduced from ∼250 , 000 doses/year 25 years ago” [10] . The WHO has estimated that antivenom supply failure in Africa is imminent [11] , which is further compounded by the presence of non-specific or fake products , inappropriate clinical use and poor community awareness of the benefits of antivenom [12]–[14] .
Data for this paper was collected from primary and secondary sources , including interviews , surveys , product inserts and literature searches . Market research surveys were sent to representatives from 46 known antivenom manufacturers in 2007 . Previous , current and future producers for sub-Saharan African markets were again contacted in 2010 and 2011 . One current and one future company did not respond in 2010/2011 . Companies responded with information regarding the following: Calculations regarding the number of vials that constitute an “effective treatment” are based on company information and product inserts for an average , or moderately severe , envenomation . Independent testing of potency and proteomic analysis to validate the species of origin was outside the scope of this study , although verification was sought through literature reviews .
The global incidence of clinically significant snakebite has been calculated to be between 421 , 000 and 2 . 5 million annually [15] , [16] , with up to 500 , 000 occurring in Africa each year [10] , [17] , [18] . Inadequate record keeping and limited primary epidemiological studies makes accurate assessment difficult , and most authors concur that estimates of snakebite incidence under-represent the problem . Up to 20–70% of victims in some regions do not present to hospital because they are either unaware treatment is available , cannot afford it , or instead utilise ineffective traditional healing methods [19]–[21] . However a recent metaanalytical study of reported data concluded that probably 314 , 000 snakebites occur in Africa annually [22] . The rate of snakebite in sub-Saharan Africa varies between 150–250/100 , 000 population [23]–[25] , with a peak incidence in some regions in Nigeria of 497/100 , 000 [26] . At least 20 , 000 deaths each year are attributed to snakebite in Africa [17] , although this is also considered conservative . The recorded annual mortality in Nigeria , Senegal and Kenya is between 2–16/100 , 000 population , and across Africa the case fatality rates from untreated snakebite ranges from 4% to 24% [27]–[30] . The WHO estimates that 10% of envenomings results in serious , non-fatal sequelae , while other reports have stated that 12 , 000–14 , 000 amputations and other sequelae result from snakebites in Africa annually [19] , [31] , [32] . Other debilitating morbidities result from the neurotoxic , coagulopathic or necrotic components of different venoms , with clinical effects ranging from chronic ulceration , osteomyelitis , chronic renal failure , endocrine disorders , paralysis , stroke and blindness . All companies currently producing antivenom for sub-Saharn Africa indicated a willingness to increase output should market demand improve . Manufacturers identified factors that prevented them from raising production , despite a willingness to do so . Whilst not all manufacturers listed the same reasons , there was some concordance and the responses below have been listed in descending order of frequency:
This survey of antivenom manufacturers highlights the paucity of antivenom products for sub-Saharan Africa and the unhelpful variability that exists within the current industry . It also illustrates that despite the exodus of manufacturers in the 1970s and 1980s , willing producers do exist and they possess substantial unutilised production capacity . Unfortunately , inadequate government and non-government funding for procurement and regulatory oversight restrains production of commercial antivenom . This lack of investment is not only the reason for the current crisis in antivenom availability , but also represents the greatest challenge to future improvements in quantity and quality . Although inexpensive and efficacious antivenoms do exist , and compelling moral and legal arguments advocate increased purchase and distribution [39] , [40] , a lack of funding for antivenom acquisition and regulation of quality standards has catalysed the vicious cycle responsible for the decline in production and use over the last 30 years ( figure 2 ) . This cycle has also contributed to conditions that have allowed lesser quality products and inappropriate marketing to emerge . The arrival of new manufacturers and the presence of spare capacity within some current facilities provide hope , but uncertain market conditions and inadequate financial support will continue to restrict growth of trustworthy antivenoms . This cycle is a variation on that proposed by Stock et al in 2007 [9] , and demonstrates the importance of future financial stimulus in reinvigorating competition and viability of the antivenom market . Inadequate financing within the antivenom industry is the major factor underpinning its decline over the last 40 years , and strategies to solve this crisis must recognise and unwind the economic and commercial drivers on both sides of the supply and demand equation . It is unrealistic to expect that pharmaceutical companies will commit to long-term production of antivenom for an inconsistent and unreliable market that is starved of investment . Even if greater volumes of appropriate antivenom could be produced , without adequate subsidisation it will be priced out of range for most snakebite victims living in underprivileged rural and remote areas . Similarly , corporate executives and regulatory bodies must also accept that there exists a moral imperative for them to contribute their expertise and capabilities , and that existing business models and production frameworks may be inappropriate for the supply of humanitarian products to developing countries . Encouragingly , there has been a small increase in financial support for the development and procurement of new African antivenoms between 2007 and 2010 . Whilst the >$60 million in global antivenom revenue and $10 . 3 million from African antivenom sales are small by pharmaceutical standards , this represents valuable investment and an encouraging base from which the industry can grow . Better utilisation of spare production capacity and improved economies of scale will produce greater yields , reduce costs , increase revenues and further enhance the commercial viability of antivenoms . The second major problem eroding the antivenom market is the lack of accountability in quality standards . Possessing the capacity to produce vast amounts of antivenom for sub-Saharan African communities is meaningless if the products are poorly made and ineffective against the snakes in those regions . A current lack of interest , insufficient investment and poor competition are allowing unscrupulous behaviours within the marketplace to go unchecked . Given the ongoing severe shortage of antivenom and the continuing high incidence of envenoming , it is not surprising that opportunistic manufacturers seek to fill the void . The advent of seemingly inexpensive , but low quality or inappropriate antivenoms with poor neutralising ability , not only compromises the reputation of antivenoms in general but also drains important financial resources away from proven snakebite treatment programs and products . Some manufacturers have cited this uneven playing field as a key impediment to future innovation and productivity . Nevertheless , the very high volume output by some manufacturers of alleged inappropriate products still make them key players in the antivenom industry , and potentially integral to future strategies for increasing output of higher quality products . Improving standards and maximising efficiencies ought to be the common goal for all manufacturers . The three groups with emerging new African antivenoms provide hope for the future [41]–[44] , however ensuring that these products , as well as existing antivenoms , are of sufficient quality to be incorporated into a properly funded and sustainable market is paramount [8] . The final quality control checkpoint for all antivenoms entering a country should be the national regulatory authorities . It is essential that NRAs are adequately resourced and transparent to ensure the integrity and robustness of their mechanisms are above reproach . Linking funds for antivenom procurement to improved quality control and assurance measures would enhance the crucial role of local regulatory bodies and incentivise the maintenance of minimum standards . Antivenom's usually rapid and curative effects make it a highly cost-effective intervention [40] , and together with snakebite's surpassing morbidity and mortality [6] , ought to attract attention from global health funding bodies . If improved efficiencies , technical support and collaboration within the antivenom industry were achieved , the cost of an effective antivenom treatment would fall below the current average of $124 , and may ultimately be significantly less than $100 . Supplying sufficient quantities of antivenom to the whole of Africa at that price would require an annual input of less than $30–$50 million , which is considerably lower than the budgets for many other global health programs . Leadership and support from groups such as the Global Snakebite Initiative and the World Health Organisation may help to secure essential funds from donors and provide important coordination , transparency and accountability . It will also help to recruit and reform manufacturers capable of contributing a greater supply of effective and appropriate antivenoms . The declining availability of high quality antivenom in sub-Saharan Africa is a real and unnecessary tragedy , and constitutes a major neglected global health concern . The amount of suitable antivenom marketed in these countries has fallen to crisis levels , representing only a fraction of the amount required . Although recent output of antivenom for Africa has increased , and the number of manufacturers able to boost production is growing , inadequate financial support and market uncertainty continue to suppress growth and compromise quality standards . The provision of sufficient funds to identify satisfactory antivenoms , maintain quality control , maximise efficiencies and increase procurement is desperately needed to break the vicious cycle that currently constrains the antivenom industry . The mechanisms to achieve this are realistic and available; science , business and government must collaborate to secure a brighter future for snakebite victims in developing countries . Only then will the goal of providing effective , safe and affordable antivenoms to all who need them , be realised . | Antivenom is the only specific treatment for systemic envenoming from snakebite , but remains unavailable to thousands of snakebite victims around the world . A cycle of inconsistent and low market demand , sub-optimal utilisation , rising costs and reduced output of antivenoms have resulted from long term under-investment in procurement and quality regulatory programs . This study provides a contemporary overview of the African antivenom market within the context of the global market . Globally , 35 companies sold at least 4 million vials of antivenom in 2007 . Five companies had established African antivenom markets in 2010/11; three other institutions have antivenoms for Africa in development; and another ceased production indefinitely . Between 2007 and 2011 , production of sub-Saharan African antivenoms rose from 227 , 400 to at least 377 , 500 vials , constituting ∼83 , 000 effective treatments for moderate envenomings . However , recent reports have identified that some products , which comprise up to 90% of the total antivenom supply in sub-Saharan Africa , may lack efficacy or specificity against relevant snake species . Despite this , revenues from antivenom marketed in sub-Saharan Africa increased from $6 . 6 million in 2007 to $10 . 3 million in 2010/11 . The average cost of a stated effective treatment in 2010/11 was $124 , and the price of antivenom is inversely proportional to the amount produced . Combined unutilised production capacity far exceeds the total projected antivenom needs for Africa . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"drugs",
"and",
"devices",
"toxicology",
"occupational",
"and",
"industrial",
"health",
"global",
"health",
"neglected",
"tropical",
"diseases",
"pharmacology",
"infectious",
"diseases",
"toxin",
"binding",
"public"... | 2012 | Consequences of Neglect: Analysis of the Sub-Saharan African Snake Antivenom Market and the Global Context |
Certain pathogens deliver effectors into plant cells to modify host protein targets and thereby suppress immunity . These target modifications can be detected by intracellular immune receptors , or Resistance ( R ) proteins , that trigger strong immune responses including localized host cell death . The accelerated cell death 11 ( acd11 ) “lesion mimic” mutant of Arabidopsis thaliana exhibits autoimmune phenotypes such as constitutive defense responses and cell death without pathogen perception . ACD11 encodes a putative sphingosine transfer protein , but its precise role during these processes is unknown . In a screen for lazarus ( laz ) mutants that suppress acd11 death we identified two genes , LAZ2 and LAZ5 . LAZ2 encodes the histone lysine methyltransferase SDG8 , previously shown to epigenetically regulate flowering time via modification of histone 3 ( H3 ) . LAZ5 encodes an RPS4-like R-protein , defined by several dominant negative alleles . Microarray and chromatin immunoprecipitation analyses showed that LAZ2/SDG8 is required for LAZ5 expression and H3 lysine 36 trimethylation at LAZ5 chromatin to maintain a transcriptionally active state . We hypothesize that LAZ5 triggers cell death in the absence of ACD11 , and that cell death in other lesion mimic mutants may also be caused by inappropriate activation of R genes . Moreover , SDG8 is required for basal and R protein-mediated pathogen resistance in Arabidopsis , revealing the importance of chromatin remodeling as a key process in plant innate immunity .
Unlike vertebrates , plants lack a somatic , adaptive immune system and immunological memory [1] . Therefore , plants rely on a large repertoire of pre-existing immune receptors , encoded by hypervariable Resistance ( R ) genes , which recognize specific pathogens and activate strong defense responses . These responses include the programmed cell death ( PCD ) of host cells at infection sites to restrict pathogen access in a process called the hypersensitive response ( HR ) . R proteins are triggered by pathogen-specific effector proteins that have evolved to perturb or disrupt host processes to facilitate infection . While some pathogen effectors are recognized extracellularly , the majority are targeted to various intracellular compartments of the plant host and identified there . In most cases , R proteins are activated by detecting modifications to host proteins targeted by pathogen effectors . This model , known as the “guard hypothesis” [2] , [3] , has been supported in numerous instances . For example RIN4 , a host protein with key roles in basal defense , is under surveillance by multiple R proteins , and at the same time is the target of multiple pathogen effectors [4] . Most R proteins have been classified as NB-LRRs , named after their central nucleotide-binding ( NB ) and C-terminal leucine-rich repeat ( LRR ) domains , although various exceptions exist [5] . The N-terminal domains of NB-LRR R proteins fall into two broad categories: those with homology to Drosophila Toll and mammalian Interleukin-1 Receptor ( TIR ) , and those with predicted coiled-coil ( CC ) regions [6] . Members of the animal NOD-like receptor ( NLR ) family exhibit similar domain architecture to plant NB-LRRs , and NLRs are likewise involved in immunity [7] , [8] . Like NB-LRR proteins , NLRs have several types of amino-termini including protein–protein interaction domains associated with proteins involved in programmed cell death and inflammation . Several autoimmune diseases in humans have been associated with mutations in NLRs [9] . In plants , there are numerous examples of mutants with autoimmunity-related phenotypes . These so-called “lesion-mimics” are , in many cases , caused by mutations in genes hypothesized to be negative regulators of the HR [10] . Other examples include point mutations in NB-LRR R proteins [11] , [12] . Since R proteins have the potential to trigger host PCD , their activity is tightly regulated . R genes are typically constitutively expressed at low levels and some are up-regulated in response to pathogen-derived peptides or to the accumulation of the phytohormone salicylic acid ( SA ) [13] , [14] . Little is known about the transcriptional control of R genes . Intriguingly , members of a cluster of related Arabidopsis R genes are endogenously suppressed at the post-transcriptional level by RNA silencing , suggesting that pathogens that interfere with the silencing machinery unwittingly up-regulate steady-state R protein levels [15] . At the protein level , inappropriate activation is likely prevented by autoinhibition , high rates of turnover , and alternatively spliced products . Recently , it has become clear that hybrid necrosis , a deleterious genetic incompatibility observed in many intra- and interspecific plant hybrids , is associated with autoimmunity [16] . One example of this type of autoimmune response in Arabidopsis was shown to be dependent on an NB-LRR R protein , suggesting that these immune receptors have a broad mandate over PCD that extends beyond pathogen defense [17] . The lethal , recessive accelerated cell death 11 ( acd11 ) mutant of Arabidopsis is characterized by constitutive activation of immune responses and PCD in the absence of pathogen attack [18] . ACD11 encodes a putative sphingosine transfer protein with homology to HET-C2 of the fungus Podospora anserina . Allelic variants of het-c determine compatibility during fusion of hyphae from different strains , causing PCD in combination with specific alleles at other het loci [19] . acd11 mutants develop normally until the 2–4 leaf stage , and PCD involves the phytohormone SA such that expression of a bacterial SA hydroxylase ( NahG ) strongly suppresses cell death . Application of SA agonists , such as benzothiadiazol-S-methyl ester ( BTH ) , restores autoimmunity in acd11 . Interestingly , the genetic requirements for acd11 cell death are similar to those for the HR triggered by TIR-NB-LRR immune receptors [18] , [20] . We report here that cell death in acd11 is suppressed by mutations in genes encoding a histone methyltransferase and a TIR-NB-LRR R protein . In addition , the expression of the R gene is dependent on the activity of the histone modifying enzyme . We propose that the TIR-NB-LRR is triggered by the absence of ACD11 , implying that ACD11 ( or a complex containing ACD11 ) may be a guarded pathogen effector target . Alternatively , since ACD11 may be involved in production of a lipid signal , the absence of this signal may induce LAZ5 expression in an SA-dependent manner . Our study provides strong evidence that a specific type of histone modification is directly involved in chromatin remodeling and transcriptional control of a subset of R genes including LAZ5 .
To isolate genes required for cell death in acd11 , Landsberg erecta ( Ler ) ecotype acd11-1 plants harboring the NahG transgene were mutagenized with ethyl-methanesulfonate ( EMS ) , diepoxybutane ( DEB ) or γ-irradiation . ∼200 suppressors of acd11 were subsequently identified as plants that survived following BTH treatment . Genetic analyses of 43 such suppressors grouped them into 12 recessive and 2 dominant loci referred to as lazarus ( laz ) mutants , after the biblical resurrection . One of the laz mutants found in the suppressor screen , laz2 , abolished cell death in response to BTH in the acd11 NahG background , and exhibited similar levels of cellular ion leakage as wild type ( Fig . 1 , A and B ) . laz2-1 acd11-1 NahG plants also exhibited abnormal development ( e . g . early flowering , increased shoot branching ) that , along with acd11 suppression , was inherited recessively ( data not shown ) . Two other laz2 alleles with similar morphology , laz2-2 and laz2-3 , were confirmed by complementation tests ( Fig . S1A ) . Global transcript profiles of laz2-1 acd11-1 NahG , Ler wild-type , NahG , and acd11-1 NahG plants were acquired by hybridizing total mRNA , isolated before and 72 h after BTH treatment , to Affymetrix ATH1 GeneChip arrays . laz2-1 exhibited dramatic suppression of the top 500 most significantly regulated genes in acd11-1 after 72 h BTH ( Fig . S2A ) In addition , a strong negative Pearson correlation of −0 . 87 was obtained for global expression fold change between laz2-1 acd11-1 and acd11-1 , indicating that gene expression in acd11-1 was strongly affected by the laz2-1 mutation ( Fig . S2B ) . The LAZ2 locus was identified using a map-based approach . Briefly , Ler laz2-1 acd11 NahG was crossed to Columbia ecotype ( Col-0 ) acd11 NahG to generate a segregating F2 mapping population after BTH treatment . Ecotype-specific linkage markers were used to map laz2-1 to a ∼150 kb region at the bottom of chromosome 1 ( Fig . S3 ) . Candidate genes were selected and sequenced based on annotated mutant phenotypes at The Arabidopsis Information Resource ( TAIR; http://www . arabidopsis . org ) , revealing an irradiation-induced 28-bp deletion in the third exon of the gene At1g77300 ( Fig . 2A ) . This locus was also sequenced in laz2-2 acd11-1 NahG , revealing an EMS-induced G to A transition converting tryptophan 1536 to a premature stop . Sequence analysis revealed that LAZ2 encodes the histone lysine methyltransferase ( HKMT ) SET ( Su ( var ) 3-9 , E ( z ) and Trithorax-conserved ) DOMAIN GROUP 8 ( SDG8 ) , otherwise known as EARLY FLOWERING IN SHORT DAYS ( EFS ) and CAROTENOID CHLOROPLAST REGULATORY 1 ( CCR1 ) [21] , [22] . The mutation in laz2-1 causes a frame-shift just upstream of the sequence encoding the conserved SET associated cysteine-rich domains , while that in laz2-2 introduces a stop codon upstream of a motif conserved within the RPB1 subunits of RNA polymerase II [23] . SDG8 is homologous to yeast SET2 , which is associated with methylations at histone 3 lysine 36 ( H3K36 ) . Another yeast HKMT , SET1 , modifies H3K4 . Both H3K4 and H3K36 methylation marks are typically associated with active transcription [24] . While Arabidopsis has 43 annotated SDG proteins , SDG8 groups with H3K36-specific HKMTs in fungi and animals along with 4 other Arabidopsis proteins [25] . During transcription in yeast , SET1 and SET2 are recruited to active chromatin by the RNA polymerase II-associated PAF1 complex , where they promote gene expression by facilitating chromatin opening , thus enhancing transcription initiation and elongation , respectively [26] . A similar mechanism seems to be conserved in Arabidopsis based on studies of sdg mutants . SDG8 was first identified as a gene that controlled flowering time via its activity on the transcription of the key floral repressor FLOWERING LOCUS C ( FLC ) , an epigenetically regulated MADS box transcription factor ( TF ) [27] , [28] . Expression of the FLC paralog MADS AFFECTING FLOWERING 1 ( MAF1 ) is also dependent on SDG8 , which is required for di- and trimethylation of H3K36 [25] . In addition to flowering time , SDG8 regulates carotenoid composition and shoot branching via modification of chromatin at specific loci [22] , [29] . Our microarray expression analysis revealed that MAF1 and CRTISO , both recently confirmed as direct targets of SDG8 [22] , [25] , exhibited very low expression levels in the absence of LAZ2 ( Fig . S4 ) . Deficient expression of these and similar genes likely contributes to the developmental phenotypes observed in laz2 . Furthermore , the loss-of-function mutant sdg8-2 ( SALK_026642 ) shared laz2 morphology ( Fig . S1B ) and suppressed acd11-2 , an ACD11 knockout in the Col-0 ecotype ( Fig . 2B ) . Transcriptome analysis of genes normally induced in acd11-1 NahG after BTH treatment showed that one of the most affected genes in laz2-1 was At5g44870 , annotated as an NB-LRR R gene ( Fig . 3A ) . This agrees with data from a previous study showing that At5g44870 is severely down-regulated in ccr1-1 ( sdg8 ) leaf tissues [22] . A number of acd11 suppressors found in the same screen as laz2 were dominant . One of these , laz5 Dominant 1 ( laz5-D1 ) , was mapped to a region close to this R gene ( Fig . S5 ) . Sequencing of At5g44870 in laz5-D1 revealed a G to A transition at the splice donor site ( +1 position ) of intron 4 likely resulting in deletion of exon 5 ( Fig . 3B ) . To confirm that this mutation resulted in suppression of acd11 , two allelic dominant suppressors , laz5-D2 and laz5-D3 , were sequenced: both had lesions in At5g44870 ( below ) , hereafter referred to as LAZ5 . LAZ5 encodes a TIR-class NB-LRR of unknown pathogen specificity with sequence similarity to RPS4 ( Fig . S6 ) , an R protein conferring resistance to Pseudomonas syringae expressing the effector AvrRPS4 [30] . The DEB-induced laz5-D2 mutation is a T to A transversion changing isoleucine 287 to asparagine ( I287N ) . This mutation is within the P-loop motif of the NB domain essential for coordination of bound nucleoside triphosphates [5] . The EMS-induced point mutation in laz5-D3 ( G811E ) lies in the LRR domain , which provides pathogen recognition specificity and has been implicated in R protein activation [31] . Accelerated cell death in acd11-1 was suppressed by laz5-D1 and laz5-D2 ( Fig . 3C ) , and laz5-D alleles suppressed acd11 cell death irrespective of BTH induction or the presence of NahG ( Fig . 3D ) . Furthermore , over-expression of laz5-D2 or laz5-D3 ( 35S:laz5-D2 or 3 ) suppressed acd11 death after induction , confirming that dominant negative mutations in LAZ5 are responsible for suppression of the acd11-dependent autoimmune response ( Fig . S7 ) . Transgenic plants over-expressing R-genes can exhibit spontaneous cell death and/or constitutive defense responses [32] . In agreement with these observations and the phenotype associated with deletion of ACD11 , over-expression of wild-type LAZ5 ( 35S:LAZ5 ) in the Col-0 background resulted in 30 out of 38 transgenic plants exhibiting acd11-like cell death which did not survive to set seed ( Fig . S8 ) . Since LAZ5 transcription is likely dependent on SDG8 HKMT activity , and the suppression of acd11 by laz2/sdg8 is recessive , we predicted that a loss-of-function mutation in LAZ5 would suppress acd11 in a recessive manner . As expected , a null T-DNA insertion mutant of At5g44870 ( SALK_087262; here termed laz5-1 ) suppressed acd11-2 cell death recessively in plants without NahG ( Fig . 4A ) . A second T-DNA insertion mutant allele of LAZ5 ( SAIL_874-D10 ) also suppressed cell death in acd11-2 ( data not shown ) . Expression of LAZ5 was assayed by real-time PCR in wild-type , laz5-1 , and sdg8-2 plants 24 hours after syringe inoculation with the virulent bacterial pathogen Pseudomonas syringae tomato ( P . s . t . ) DC3000 or with 10 mM MgCl2 ( mock control ) . While pathogen treatment induced LAZ5 expression in wild type , transcript levels in sdg8-2 were comparable to that in the laz5-1 null mutant ( Fig . S9A ) . This confirms the microarray expression data shown in Fig . 3A . The apparent lack of LAZ5 expression in sdg8-2 was seen in several independent experiments with plants at different stages and/or treated with other pathogen strains ( data not shown ) . Moreover , ACD11 expression was unaffected in laz5-1 and sdg8-2 ( Fig . S9B ) , and transcript accumulation of several TIR-NB-LRR-encoding genes homologous to LAZ5 was seemingly unaffected in 3-week old sdg8-2 plants compared to wild-type control with the possible exception of At5g45230 ( Fig . S10 ) . An important question is whether LAZ5 is the relevant target of SDG8 required for acd11 cell death . To help answer this question , we transformed laz2-1 acd11-1 NahG plants with a genomic construct of LAZ5 under control of a constitutive promoter and monitored cell death by ion leakage after BTH treatment compared to relevant controls ( Fig . S11 ) . LAZ5 over-expression restored cell death in leaf discs between 3 and 8 days after induction , indicating that lack of LAZ5 expression in sdg8 is a major cause of the suppression of acd11 cell death . However , it cannot be excluded that other targets of SDG8 histone methyltransferase activity also contribute to BTH-induced cell death in acd11 . To test whether laz2 directly affects histone methylation at the LAZ5 locus , chromatin immunoprecipiation ( ChIP ) was conducted using antibodies against specifically modified histones . In laz2-1 acd11-1 NahG , trimethylated ( me3 ) H3K36 levels were reduced in chromatin associated with the 5′ coding regions of MAF1 ( control ) and LAZ5 , when compared to the acd11-1 NahG control ( Fig . 4B ) . Enrichment of H3K36me3 in LAZ5 chromatin was not influenced by BTH treatment or acd11 homozygosity ( Fig . S12A ) . This suggests that activation of cell death in acd11 does not result in hyper-trimethylation at H3K36 , but rather that this histone modification is required for proper LAZ5 expression . There was no effect of genotype on levels of total H3 ( Fig . 4C ) . H3K36me3 is not a general mark for genes up-regulated in acd11 , such as FMO1 [18] , since we found no enrichment at FMO1 chromatin 72 h after BTH induction ( Fig . S12B , C ) . Moreover , absence of LAZ2/SDG8 had no effect on H3K36me3 levels at the constitutively expressed ACTIN locus ( Fig . 4C ) or the MAP KINASE KINASE 4 ( MKK4 ) locus ( Fig . S12D ) . To elucidate H3K36 methylation status irrespective of acd11 and NahG , we also conducted ChIP assays on sdg8-2 single mutant and Col-0 wild-type seedlings . It was previously shown that loss of SDG8 resulted in both a decrease in global H3K36me3 levels and a coincident increase in global monomethylated ( me1 ) H3K36 , a mark associated with transcriptional repression in Arabidopsis [25] . In wild-type plants , MAF1 and LAZ5 chromatin was enriched for H3K36me3 , whereas the level of H3K36me3 was diminished in sdg8-2 ( Fig . 4D ) . Conversely , H3K36me1 levels at these loci were higher in sdg8-2 and reduced in wild type . Treatment of seedlings for 3 hours with an HR-inducing bacterial pathogen had no effect on the methylation status of H3K36 ( data not shown ) . Also , H3 trimethylation of LAZ5 chromatin at other lysine residues ( K4 , K9 , K27 ) , was not affected by loss of SDG8 ( Fig . S12E ) . To determine whether SDG8 and/or LAZ5 are required for basal resistance to virulent pathogens , leaves of 4-week old sdg8-2 , laz5-1 , wild-type and an allele of enhanced disease susceptibility 1 ( eds1-2 introgressed into Col-0 ) mutants were syringe-inoculated with P . s . t . DC3000 and growth was assayed after 4 days . Bacteria grew to ∼9-fold higher titers in sdg8-2 than in wild-type or laz5-1 , while titers in eds1 were yet another order of magnitude higher ( Fig . 5A ) . Growth of another strain of bacterial pathogen , Pseudomonas syringae maculicola ( P . s . m . ) ES4326 , was tested on sdg8-2 , laz5-1 , wild-type and eds1 with similar results ( Fig . 5B ) . We did not observe elevated bacterial growth in sdg8-2 when we used P . s . t . DC3000 HrcC- ( Fig . S13A ) , a non-pathogenic mutant defective in delivery of effectors to host cells [33] . These data indicate that SDG8 , but not LAZ5 , is required for full resistance to virulent pathogens . Furthermore , we found that SDG8 is involved in resistance to avirulent pathogens mediated by other R proteins , for example RPM1 . Plants were syringe-inoculated with P . s . t . DC3000 expressing HR-inducing AvrRpm1 , AvrRpt2 , AvrRps4 or AvrPphB and growth was assayed after 3 or 4 days . Bacterial titers were ∼15-fold higher in sdg8-2 than in wild-type or laz5-1 for P . s . t . expressing AvrRpm1 ( Fig . 5C ) . This suggested that RPM1-mediated resistance is defective in sdg8-2 . To confirm this , growth of P . s . m . ES4326 expressing AvrB was assessed after 3 days: AvrB is also recognized by RPM1 , and resistance to this avirulent pathogen was affected in sdg8-2 to a similar level as P . s . t . with AvrRpm1 ( Fig . 5D ) . In both cases , bacterial titers were comparable to the rpm1-3 null mutant [34] . Defects in SDG8 had a consistent , yet statistically insignificant effect on growth of P . s . t . DC3000 expressing AvrPphB , ( Fig . S13B ) resistance to which is dependent on the R gene RPS5 [35] . In addition , sdg8-2 did not affect RPS2- or RPS4-mediated resistance to AvrRpt2 [36] , [37] ( Fig . 5E ) and AvrRps4 [30] ( Fig . 5F ) . Corroborating the pathogen growth assay , transcript levels of RPM1 and RPS5 were low or absent in 4-week old sdg8-2 compared to wild-type , whereas expression of RPS2 and RPS4 in sdg8-2 was similar to that in wild-type ( Fig . 5G and S13C ) . Defects in LAZ5 did not have a detectable effect on transcript accumulation of RPM1 , RPS5 , RPS2 or RPS4 ( data not shown ) . As with LAZ5 , we conducted ChIP assays at the RPM1 locus in untreated seedling tissue from laz2-1 acd11-1 NahG versus acd11-1 NahG ( in Ler ) and sdg8-2 versus wild-type ( in Col-0 ) . We observed lower H3K36me3 and higher H3K36me1 levels at RPM1 chromatin in the absence of functional LAZ2/SDG8 , indicating that RPM1 is an example of another R gene that is regulated by histone methylation ( Fig . S14 ) . These results indicate that SDG8 targets a subset of R genes and other genes involved in more general aspects of basal defense .
Chromatin remodeling has emerged as a complex regulator of transcription and an epigenetic mechanism to maintain lasting changes in gene activity states . Dynamic post-translational modifications of various residues of histones tails , including methylation , phosphorylation , acetylation , and ubiquitination , play important roles in both promoting and repressing gene expression by recruiting histone binding proteins and chromatin remodeling enzymes [38] . The combinatorial nature of histone modifications results in a complex “histone code” that adds an important level of control to fine-tune gene-specific responses to broader transcriptional inputs [39] . Changes in chromatin state may therefore modulate gene expression in a context-dependent manner to maintain a flexible response to pathogen attack . In plants , this process has been proposed as a mechanism for priming SA-responsive loci during systemic acquired resistance to pathogens [40] . So far , relatively few studies directly associate epigenetic processes related to chromatin modification to plant innate immunity and/or PCD . Defects in HISTONE DEACETYLASE 19 ( HDAC19 ) and HISTONE MONOUBIQUITINATION 1 ( HUB1 ) increase susceptibility to necrotrophic fungal pathogens in Arabidopsis [41] , [42] . Furthermore , defects in genes involved in histone variant replacement , and the variant H2A . Z itself , result in increased resistance to virulent bacterial pathogens , some spontaneous cell death , and up-regulation of defense genes [43] . More commonly , the “memory” of chromatin remodeling activity is observed as increased levels of open chromatin marks ( H3Ac , H3K4me2 , etc ) at the promoters of many SA-responsive genes , such as PATHOGENESIS-RELATED 1 ( PR-1 ) and WRKY TFs [40] , [44] , [45] . The clearest example of immune response at the level of chromatin comes from Alvarez-Venegas and colleagues , who showed that the HKMT ARABIDOPSIS TRITHORAX 1 ( ATX1 , also known as SDG27 ) controls expression of WRKY70 , a TF involved in pathogen response [46] . ATX1-dependent H3K4me3 signatures at the promoter of WRKY70 correlated with WRKY70 transcriptional up-regulation . Intriguingly , although ATX1 regulates expression of a large set of genes , a high proportion of immunity-related genes exhibited reduced expression in the knockout mutant , including various TIR-NB-LRR R genes [47] . Numerous examples exist of microbes and viruses manipulating host chromatin remodeling machinery or histones directly in animals [48] , [49] . Strikingly , toxins from unrelated bacterial pathogens of animals have evolved to modify host histones , reducing transcriptional activity of key immunity genes [50] . The only clear instance of related phenomena identified among plant pathogens is the case of the Crown Gall disease-causing bacterium Agrobacterium tumefaciens which selectively modulates the expression of host variant histone genes to allow genomic integration of its T-DNA [51] , [52] . There is conflicting data on whether loss of sdg8 influences H3K4 methylation , H3K36 methylation , or both [22] , [23] , [25] , [28] . We detected a dramatic effect of laz2/sdg8 on H3K36 methylation status of chromatin at various loci and no difference in H3K4me3 levels at LAZ5 , although the H3K4 methylation status of chromatin at other loci in laz2 backgrounds remains to be investigated . In addition , our data suggest that monomethylation of H3K36 at MAF1 and LAZ5 chromatin relies on HKMTs other than SDG8 . One of these , SDG26 , was previously shown to act antagonistically to SDG8 by repressing FLC expression , although global H3K36me1 levels were unaffected in the sdg26 mutant [25] . The significance of H3K36me1 enrichment in sdg8-2 remains unknown . One hypothesis is that H3K36 methylation proceeds in a stepwise fashion , with the accumulation of H3K36me1 ( due to activity of an unknown HKMT ) being a consequence of a block in further di- and trimethylation at this residue normally mediated by SDG8 . Alternatively , monomethylation of H3K36 may represent a transcriptionally repressive mark that accumulates only in the absence of di- and trimethylation due to disruption of the balance between antagonistic chromatin modifiers . For example , the SET-domain containing Arabidopsis proteins TRITHORAX-RELATED PROTEIN 5 ( ATXR5 , also known as SDG15 ) and ATXR6/SDG34 are H3K27-specific monomethyltransferases essential for transcriptional repression in heterochromatin [53] . Further studies should examine if other predicted H3K36-specific HKMTs , namely SDG4 , SDG7 , SDG24 and SDG26 , have any role in H3K36 monomethylation , trimethylation and/or antagonistic control of expression of LAZ5 and other genes with roles in immunity or are required for cell death in acd11 . Moreover , further work is required to determine the mechanisms by which SDG8-dependent changes in H3 methylation regulates the expression of specific genes . A clue to the function of LAZ5 activation comes from the isolation in our screen of dominant alleles . This indicates that the mutant form ( laz5-D ) of the R protein likely interferes with activity of the wild-type copy since plants heterozygous for the laz5 null mutation do not suppress acd11 , indicating haplosufficiency of LAZ5 . Dominant negative activity has been described for mutations in the R gene N from tobacco , and indeed for a point mutation ( G216E ) in the P-loop motif of N [54] . N was later found to oligomerize in the presence of a Tobacco mosaic virus elicitor , likely through interaction of TIR domains [55] . This oligomerization was an early event in pathogen perception and was independent of mutations that have an effect on HR induction . Therefore , it is possible that laz5-D mutants form inactive oligomers with wild-type LAZ5 and/or accessory proteins . An example of this scenario from animal innate immunity comes from NOD2 , an NLR involved in recognition of bacterial cell wall components: an endogenously truncated form , NOD2-S , interacts with full-length NOD2 to potentiate signaling [56] . In plants , there are examples of truncated R proteins , generated by alternative splicing , playing a key role in signaling [57] , [58] . At present , it is an open question whether LAZ5 oligomerizes and how this relates to cell death activation . It should be noted that , while all the laz5 alleles isolated thus far in the acd11 suppressor screen were dominant negative , only 43 of the ∼200 unknown recessive mutants were placed into complementation groups , and even fewer were mapped . Therefore , a recessive laz5 knockout allele may exist among our unmapped suppressors . In this study we have identified the chromatin modifying enzyme SDG8 , and its specific target LAZ5 , as regulators of autoimmune cell death in acd11 . Furthermore , sdg8 mutants exhibit enhanced susceptibility to virulent and avirulent pathogens , whereas laz5 mutants do not , suggesting that other targets of SDG8 are important for general resistance . We also show that transcription of a subset of R genes , including LAZ5 and RPM1 , is likely to be directly or indirectly dependent on LAZ2 activity . One scenario that may account for the enhanced susceptibility of sdg8 mutants to virulent pathogens could be the consequence of SDG8 action on multiple NB-LRR loci . If the suite of effectors delivered by Pseudomonas triggers a weak R gene response , in sdg8 a subset of these do not accumulate and thus are no longer available to signal for defense against the invading pathogen . Intriguingly , SDG8 is not expressed until 8 days after germination [28] , a stage preceding the initiation of cell death in acd11 . SDG8 may therefore developmentally regulate targets such as LAZ5 , and may exemplify a key difference in the programmed defenses required during seed maturation and the inducible defenses used during plant growth . Lesion mimic mutants such as acd11 are useful tools in the genetic dissection of innate immunity in plants [10] . Whereas several of these mutants have putative roles in ceramide signaling or synthesis [59] , [60] or auto-activate R proteins [11] , the majority of lesion mimic mutants represent proteins with no straightforward connection to PCD . Milder autoimmunity , associated with constitutive activation of defense responses and dwarf morphology without coincident HR , can similarly be the result of point mutations in immune receptors ( Zhang et al . , 2003 ) , or deletion of signaling intermediates such as MAP kinases [61] . Knockout mutants that eliminate host guardees mimic the effects of pathogen effectors , and have been found to exhibit R-gene-dependent lethality [62] . Therefore , it is possible that many lesion mimic/autoimmune mutants may correspond to gene functions that are guarded by NB-LRRs . If so , the diverse functions of these genes may be “red herrings” not directly related to PCD but only implicated in this process due to their targeting by pathogen effectors . Such may be the case for acd11 , although we have been unable to detect any interaction between full-length or truncated LAZ5 and ACD11 in yeast or in planta ( data not shown ) . Previously , we reported the identification of ACD11-interacting proteins [63] , which we are testing for interaction with LAZ5 . Two predictions about wild-type products of autoimmune mutants emerge from this model . First , suppressor screens should identify R genes . Second , pathogen effectors should target them either directly or indirectly via interacting partners or products of their activities . We currently have no evidence that ACD11 is targeted by pathogen effectors , or that ACD11 contributes to disease resistance in the absence of LAZ5 . While future work may strengthen this hypothesis , an alternative model is that ACD11 is involved in negatively regulating SA-dependent expression of LAZ5 ( or a subset of R genes ) perhaps via some lipid signal .
Arabidopsis plants were grown on soil or MS-agar plates at 21°C with an 8 h or 12 h photoperiod . sdg8-2 ( SALK_026642 ) and laz5-1 ( SALK_087262 ) T-DNA insertion lines , both previously described as null mutants [23] , [64] , were generated by SIGnAL [65] and obtained from the Nottingham Arabidopsis Stock Centre ( NASC; Nottingham , UK ) . Homozygous genotyping primers were 5′-TAAAGAGGGTCTGCATCATGG-3′ with 5′-CACTGTCCAGTAAAAGCTGGC-3′ for sdg8-2 and 5′-TATGTTTTTCCCAGATGCCAG-3′ with 5′-ATCATGCATCTCAACTCGACC-3′ for laz5-1 . Sequences of primers used to detect acd11-1 , acd11-2 , and NahG are available upon request . Three lots of 920–950 mg Ler acd11-1 NahG seeds were incubated for 4 hr in either 0 . 74% ( w/v ) EMS ( Sigma-Aldrich , St Louis , MO , USA ) prepared in 0 . 1M sodium phosphate buffer , pH 5 , with 5% DMSO , or 10 mM DEB ( Sigma-Aldrich ) in water , followed by rinsing . γ-irradiation of 300 mg acd11-1 NahG seeds was performed at the Risø Reference Laboratory ( Denmark ) with 500 Gy from a Cobalt-80 source . M1 plants were grown in families of 125 individuals , 3500 M2 plants per family were screened for BTH-resistant suppressors . ∼3 million M2 plants from 845 M1 pools or ∼100 . 000 M1 plants were scored . Putative mutants were genotyped to be homozygous for acd11-1 by PCR . Conductivity assays were conducted essentially as previously described [66] . Total RNA was isolated from three independent biological replicates of relevant genotypes at 0 and 72 hr after BTH treatment . RNA was labeled and amplified according to the MessageAmp Biotin-enhanced kit ( Ambion ) protocol and hybridized to 51 ATH1 GeneChips after Affymetrix protocols . ChIP antibodies purchased from Abcam ( Cambridge , UK ) included anti-H3 ( ab1791 ) , anti-H3K36me1 ( ab9048 ) , anti-H3K36me3 ( ab9050 ) and anti-H3K27me3 ( ab6002 ) . ChIP antibodies against H3K4me3 ( pAb-056-050 ) and H3K9me3 ( pAb-003-050 ) were purchased from Diagenode ( Liège , Belgium ) . Quantitative PCR primers for ChIP analysis were LAZ5: 5′-GAGTCGTGGCAAGTGTTCATC-3′ with 5′- GAAGATGGACAGTGCGATTTC-3′; FMO1: 5′-CTCAGATGGCTTCTAACTATG-3′ with 5′-CTATTATTGGGCCATGGAAAG-3′; MAF1: 5′-CCCTTATCGGAGATTTGAAGC-3′ with 5′-GGAGGATTCACAGAGAATCG-3′; ACTIN: 5′-GGAAACATCGTTCTCAGTGG-3′ with 5′-ACCAGATAAGACAAGACACAC-3′ . ChIP was performed essentially as described [67] , using 1µg of each antibody . Real-time PCR to quantify the immunoprecipitated DNA was performed using Brilliant II SYBR Green qPCR kit ( Stratagene ) , and reactions were run on an iCycler IQ ( Bio-Rad , Hercules , CA , USA ) . In all cases , ChIP values were calculated using the Delta-Delta-Ct ( ddCt ) algorithm to determine relative gene expression utilizing the ‘percent input method’ . Briefly , signals obtained from the ChIP were divided by signals obtained from an input sample representing the amount of chromatin used in the ChIP . The ‘% input’ value shows what proportion of this starting material is found in the eluate after IP with appropriate Ab . For expression analyses , RNA was extracted from relevant genotypes using the Qiagen RNeasy RNA extraction kit followed by DNase treatment as per the manufacturer's instructions . Equal amounts of RNA were subjected to one-step real-time PCR using the same kit as described for ChIP except with reverse transcriptase included . For all sample/primer combinations , a control without reverse transcriptase was included to exclude genomic DNA contamination . 3 . 9-kb fragments of laz5-D alleles were amplified from genomic DNA ( laz5-D1 acd11-2 NahG , laz5-D2 acd11-2 NahG , laz5-D3 acd11-2 NahG ) and cloned into modified pCAMBIA-3300 as described [68] , using a uracil-excision based cloning technique ( USER , New England Biolabs ) . Cloning primers were 5′-ggcttaaUATGGCAGCATCTTCCGAAATAC-3′ and 5′-ggtttaaUTTACAATAAACCCAAGTATAATTTAG-3′ . A 3 . 9-kb fragment of LAZ5 was amplified from genomic DNA ( wild type Ler ) , cloned into pENTR/D-TOPO ( Invitrogen ) and transferred to Gateway-compatible constitutive expression vectors pGWB502Ω or pGWB521 [69] by LR recombination reaction ( Invitrogen ) . Cloning primers used were 5′-CACCATGGCAGCATCTTCCGAAATAC-3′ and 5′-TTACAATAAACCCAAGTATAATTTAG-3′ . The final constructs were verified by sequencing , electroporated into Agrobacterium tumefaciens strain GV3101 and used to transform acd11-1 NahG or wild type plants by floral dip method [70] . Transgenic plants were selected on soil with glufosinate ( 35S:laz5-D alleles ) or on MS-agar plate with ( 20mg/L ) hygromycin B followed by transplanting to soil ( 35S:LAZ5 ) . At2g34690 ( ACD11 ) : NP_181016 . At1g77300 ( LAZ2/SDG8 ) : NP_177854 . At5g44870 ( LAZ5 ) : NP_199300 . At1g77080 ( MAF1 ) : NM_180648 . At5g10140 ( FLC ) : NM_121052 . At1g19250 ( FMO1 ) NP_173359 . At5g09810 ( ACTIN ) : NP_196543 . At1g06820 ( CRTISO ) : NP_172167 . At3g48090 ( EDS1 ) NM_114678 . At3g20600 ( NDR1 ) : NP_188696 . At3g07040 ( RPM1 ) : NP_187360 . At4g26090 ( RPS2 ) : NP_194339 . At5g45250 ( RPS4 ) : NP_199338 . At1g12220 ( RPS5 ) : NP_172686 . At5g17880 ( CSA1 ) : NP_197290 . At4g36150: NP_195338 . At5g45200: NP_199333 . At5g45230: NP_199336 . | Plants defend themselves against pathogens via immune receptors that trigger responses including the suicide of infected cells to limit pathogen growth . The accelerated cell death 11 ( acd11 ) knockout mutant of the model plant Arabidopsis thaliana kills itself in the absence of invading pathogens . By screening for secondary mutations that resurrect acd11 , we discovered two LAZARUS ( LAZ ) genes required for death . The first , LAZ2 , encodes an enzyme that methylates histones , the major protein component of chromatin . This particular histone modification is generally involved in epigenetic remodeling of chromatin to a more permissive state for transcription of associated DNA . We show that expression of the second gene , LAZ5 , is dependent on LAZ2 activity , suggesting that LAZ5 is a direct target of LAZ2 . LAZ5 is a member of an immune receptor class involved in detection of specific pathogens and subsequent cell death . We propose that acd11 , and other suicidal mutants , result from autoimmunity triggered by immune receptors controlled by chromosomal modifications . Interestingly , we found that defects in LAZ2 result in enhanced susceptibility to bacterial pathogens , suggesting that it controls other genes involved in innate immunity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant",
"genetics",
"and",
"gene",
"expression",
"plant",
"biology/plant-biotic",
"interactions",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"immunology/inna... | 2010 | Autoimmunity in Arabidopsis acd11 Is Mediated by Epigenetic Regulation of an Immune Receptor |
Hepatitis E Virus ( HEV ) is the leading cause of acute viral hepatitis globally . Symptomatic infection is associated with case fatality rates of ~20% in pregnant women and it is estimated to account for ~10 , 000 annual pregnancy-related deaths in southern Asia alone . Recently , large and well-documented outbreaks with high mortality have occurred in displaced population camps in Sudan , Uganda and South Sudan . However , the epidemiology of HEV is poorly defined , and the value of different immunisation strategies in outbreak settings uncertain . We aimed to estimate the critical epidemiological parameters for HEV and to evaluate the potential impact of both reactive vaccination ( initiated in response to an epidemic ) and pre-emptive vaccination . We analysed data from one of the world's largest recorded HEV epidemics , which occurred in internally-displaced persons camps in Uganda ( 2007–2009 ) , using transmission dynamic models to estimate epidemiological parameters and assess the potential impact of reactive and pre-emptive vaccination strategies . Under baseline assumptions we estimated the basic reproduction number of HEV in three separate camps to range from 3 . 7 ( 95% Credible Interval [CrI] 2 . 8 , 5 . 1 ) to 8 . 5 ( 5 . 3 , 11 . 4 ) . Mean latent and infectious periods were estimated to be 34 ( 95% CrI 28 , 39 ) and 40 ( 95% CrI 23 , 71 ) days respectively . Assuming 90% vaccine coverage , reactive two-dose vaccination of those aged 16–65 years excluding pregnant women ( for whom vaccine is not licensed ) , if initiated after 50 reported cases , led to mean camp-specific reductions in mortality of 10 to 29% . Pre-emptive vaccination with two doses reduced mortality by 35 to 65% . Both strategies were more effective if coverage was extended to groups for whom the vaccine is not currently licensed . For example , two dose pre-emptive vaccination , if extended to include pregnant women , led to mean reductions in mortality of 66 to 82% . HEV has a high transmission potential in displaced population settings . Substantial reductions in mortality through vaccination are expected , even if used reactively . There is potential for greater impact if vaccine safety and effectiveness can be established in pregnant women .
Knowledge of the course of HEV infection is based mainly on two volunteer studies [13 , 14] and three patient studies [15–17] . The two single volunteer studies ( with known dates of exposure ) gave incubation periods of 38 and 39 days respectively [13 , 14] . Clinical illness typically lasts 1–4 weeks [17] . Viral shedding in the stool begins about four to five weeks after infection and continues for up to four to five weeks [13 , 14 , 18–20] . The start of the infectious period appears to approximately coincide with the onset of the prodromal phase of disease ( i . e . early non-specific symptoms ) [14] . Transmission of HEV is widely thought to occur predominantly via the faecal-oral route , usually through contact with contaminated water; it might therefore be assumed that household person-to-person transmission is rare . However , a case-control study of 112 symptomatic cases and 145 controls in Paloga , Uganda found only two behavioural risks associated with symptomatic HEV infection ( with adjusted odds ratios of 3 and 2 respectively ) : use of wide-mouthed water storage vessels and communal hand washing [21] . Drinking water from the river and having a borehole as a primary source of drinking water were not associated with HEV risk . Based on these findings , and the presence of HEV RNA in hand-rinse samples ( but not in any of the 15 drinking water samples collected ) , the authors concluded that water storage practices could have played an important role and that transmission was likely to have included household-level or person-to-person spread . A study during a large HEV outbreak in Madi Opei , Uganda reached a similar conclusion , and reported multiple lines of evidence to suggest that person-to-person household transmission of HEV contributed to the epidemic , and was unable to detect HEV in drinking water or zoonotic sources [22] . There is a recombinant vaccine , HEV 239 , which is approved for use in China in those aged 16–65 years who are not pregnant [23] . It is produced with a genotype 1 isolate and efficacy against both genotypes 1 and 4 has been established in non-human primates . The vaccine has been demonstrated to have >90% efficacy with a three dose schedule ( 0 , 1 and 6 months ) based on a clinical trial involving 109 , 959 people at risk of HEV infection in an endemic setting , primarily with genotype 4 [23] . Preliminary observations suggest the vaccine is also safe and effective in pregnant women [24] . Clinical trial data are lacking for those aged <16 or >65 years , in areas where genotypes 1 and 2 dominate , and in outbreak settings . We aimed to quantify key epidemiological parameters for HEV in IDP camp settings and evaluate the potential benefits of vaccination . We consider both pre-emptive ( prior to HEV cases occurring ) and reactive vaccination ( once HEV outbreaks are already underway ) , and evaluate the potential impact of selecting different target groups to receive the vaccine . To do this we fitted dynamic transmission models to data from three large HEV outbreaks in IDP camps . We used a Bayesian framework to combine data from previous studies with observed epidemic data to obtain an improved understanding of the natural history of HEV infection , quantify the transmission potential , and evaluate the potential for vaccination to reduce the number of clinical cases and associated mortality .
Data came from three outbreaks in 2007–2009 from IDP camps in the district of Kitgum , Uganda: Agoro , Madi Opei , and Paloga ( estimated populations 16 , 689 , 10 , 442 , and 10 , 555 respectively ) . These outbreaks were part of a larger epidemic in Kitgum district . Conditions in the camps were crowded and in all camps access to water and sanitation was initially poor [21] . Though the camps were in close proximity , they were not within easy walking distance of each other and there was little population movement between camps . Jaundice cases were recorded in facility-based passive surveillance systems , with all suspected cases referred to the MSF clinic at Madi Opei . Evidence from serology and reverse transcription–PCR confirmed HEV genotype 1 to be the outbreak cause; other causes of viral hepatitis were rare [10] . All patient data used in this study were anonymized . We fit a series of deterministic transmission models to the data . We assumed latent and infectious periods and the probability of infections being reported were common to all camps , as the demographics and provision of healthcare at the three camps was similar . We allowed transmissibility to vary by camp , as this might be expected to depend on local camp conditions ( Fig 1 ) . In our baseline model ( Model 1 ) individuals were assumed to be in one of four possible states: susceptible to infection ( S ) ; latently-infected but not yet infectious ( E ) ; infectious ( I ) ; and recovered and immune ( R ) . The rate at which susceptibles became infected was assumed to scale linearly with the number currently infectious . Information from previous studies was used to construct informative prior distributions ( priors ) for natural history parameters . These priors represent knowledge about disease progression parameters before fitting the model to the outbreak data . When combined with analysis of the data they give rise to posterior distributions , which represent what we know about the parameters after analysing the new data . A human challenge study had previously reported that viable virus could be found in the faeces four days before onset of the icteric phase of disease ( i . e . jaundice ) [14] . We therefore assumed that patients become infectious one week before the icteric phase of disease . A prior for the mean infectious period was derived from a study of faecal shedding in 11 patients with sporadic acute HEV infection acquired in Bangladesh , Vietnam , Nepal , and Japan ( Table 1 ) [20]; this study found that viable HEV could be recovered from faecal samples up until 2–5 weeks after hepatitis onset . We also performed sensitivity analyses where we based this prior on duration of faecal shedding from the single patient with HEV genotype 1 in this study . The prior distribution for the latent period was based on a single observation from the same study where there was a delay of 34 days between inoculation and viable HEV in faeces . Faecal shedding of HEV in asymptomatically infected people is known to occur [25]; we assumed no difference in faecal viral shedding between symptomatic and asymptomatic individuals . We used seroprevalence data to derive an informative prior for the proportion of infections that are reported ( Table 1 ) . By default we assumed no immunity to HEV in the IDP camp populations prior to first reported case , consistent with the absence of reports of previous HEV epidemics in Kitgum district and serological data [10] . We performed multiple sensitivity analyses , considering models with: i ) a different prior for the infectious period; ii ) camp-specific transmission rates affected by a water and sanitation intervention ( Models 2a and 2b ) ; iii ) different assumptions about the distribution of the latent and infectious periods ( Models 3–5 ) ; and iv ) allowing for between 10% and 30% of the population to be initially immune to HEV infection , prior to the start of the outbreak . All the above models assumed nothing about the relative importance of different modes of transmission and are consistent with both transmission mediated by a contaminated environment ( though without long-term virus persistence in this environment ) and direct person-to-person spread . We also compared our findings with those from a model ( Model 6 ) explicitly accounting for a second mode of transmission , an unobserved environmental reservoir of HEV where virus may persist for longer time periods . We performed an extensive sensitivity analysis by running this model under 25 different prior assumptions representing combinations of five different assumptions about the relative importance of this environmental reservoir in the early stages of an epidemic and five different assumptions about persistence of viable virus in the environment . Model fitting was performed within a Bayesian framework using a Markov chain Monte Carlo ( MCMC ) algorithm to derive the posterior distributions for unknown parameters . For each model we used at least 4 million Markov chain iterations and assessed convergence by visual inspection of the trace plots . If φj represents the set of unknown parameters for model j and if p ( φj |D ) is the posterior distribution of these parameters given model j , data D and priors p ( φj ) then p ( φj|D ) ∝p ( D|φj ) p ( φj ) For given parameter values , φj , a system of differential equations was used to determine the expected number of new infectious people in each seven-day period , Zi ( S1 Appendix ) . For the likelihood term , p ( D|φj ) , we assumed that the observed number of cases in each camp in week i followed a negative binomial distribution , with a mean given by the product of Zi and π , the proportion of infections which are reported . In all cases parameters to be estimated included this proportion ( π ) , the dispersion parameter of the negative binomial distribution , the time of the first case in each camp , the rates of leaving latently infected and infectious compartments , and a transmission parameter , β . Compared to the baseline model ( Model 1 ) , Models 2a and 2b estimated three additional parameter corresponding to the camp-specific estimates of the water and sanitation intervention effect on rates of transmission assuming stepwise effects associated with the intervention , and effects that scaled in proportion to the water sources per person respectively . This analysis was intended to evaluate the evidence that this intervention contributed to epidemic control . Model 5 estimated one additional parameter ( the proportion of transmission events that resulted from contacts with individuals in the first infectious period ) . Model 6 estimated two additional parameters: λ , the rate of increase in the contamination of the saturating environment per infected host and υ , the rate of loss of contamination from this environmental reservoir . To evaluate the potential impact of vaccination we used estimates of vaccine effectiveness after two and three doses derived from data in Zhu et al [23] ( Table 2 ) . This gave posterior means ( and central 95% credible intervals ) of 80 . 2% ( 16 . 4% , 99 . 6% ) for two doses and 93 . 3% ( 74 . 3% , 99 . 8% ) for three ( S2 Appendix ) and assumed 90% coverage for the first two doses in target groups . The intervals between the first and second and the second and third doses were one and five months respectively . There was no evidence of any effect of a single dose of vaccine in the clinical trial , so this was excluded from the analysis . In the absence of evidence to the contrary , we assumed vaccine effectiveness did not vary by age . We assumed no loss of vaccine or infection derived immunity over the timescales considered , in accordance with findings of long-term follow-up studies which found consistent vaccine-induced protection over 4 . 5 years and a slow rate of decline of immunity derived from infections [28 , 29] . Case fatality ratios amongst those pregnant and those not pregnant were derived from the meta-analysis of Rein et al . [6] . We assumed a threshold of 50 or 100 reported cases as the starting point for reactive vaccination . Other assumptions are given in Table 2 . Regardless of the threshold number of cases for initiating vaccination , a three dose vaccination strategy was considered incompatible with reactive vaccination because the third dose would have to be given close to the end of the epidemic ( Fig 2 ) . We therefore only consider two dose reactive vaccination scenarios . A web application that uses the baseline model and allows for the simulation of the effects of different vaccination strategies ( including three dose reactive vaccination ) under user-specified assumptions is available at https://moru . shinyapps . io/HEVmodel/ . In all simulations we accounted for uncertainty in model parameters by drawing these from the posterior distributions obtained by fitting the model to data from the three camps . Analysis was performed in R [30] . Model code is available at https://github . com/BenSCooper/HEVmodel .
The baseline model ( Model 1a ) gave good fits to the three epidemic curves , with the observed number of weekly cases usually within the 95% prediction intervals for observed cases ( grey shaded region ) ( Fig 2 ) . Under the baseline model the mean latent period was estimated to be 34 days , 95% CrI ( 29 , 39 ) ( Fig 3 and Table 3; example MCMC output is shown in S3 Appendix ) . Similar estimates were obtained in sensitivity analyses , though when the prior for the infectious period was based on a single patient with HEV genotype 1 ( with 22 days post-onset shedding ) the posteriors for the infectious period and camp-specific reproduction numbers gave more support to lower values than under baseline assumptions ( Model 1b , Table 3 ) . The effect of assuming prior immunity in the population was to slightly increase estimates of the proportion of infections which are reported ( from 13% with no immunity to 18% if 30% of the population were initially immune ) and to lead to higher estimates of basic reproduction numbers . These were inflated by about 50% if 30% of the population were assumed initially immune ( S1 Table ) . A shorter mean latent period ( 19 days ( 10 , 34 ) ) was inferred if alternative distributional assumptions were made about the latent period ( Model 3 , Table 3 ) . The mean infectious period was estimated to be 36 days , 95% CrI ( 21 , 64 ) , in the baseline model , though this reduced to 27 days ( 21 , 37 ) under different distributional assumptions ( Model 4 , Table 3 ) . The estimated proportion of infections reported was 12 . 5% ( 11 . 4% , 13 . 6% ) in the baseline model and similar in all sensitivity analyses . Under baseline assumptions the basic reproduction numbers were estimated to be similar in two of the three camps ( Agoro 6 . 5 ( 4 . 5 , 9 . 9 ) ; Paloga 8 . 5 ( 5 . 3 , 11 . 4 ) ) , but smaller in Madi Opei ( 3 . 7 ( 2 . 8 , 5 . 1 ) ) . Central estimates for these reproduction numbers were similar in most sensitivity analyses ( Table 3 ) , though higher in models that explicitly accounted for two modes of transmission ( Table 4 ) . Analysis of the data using models explicitly accounting for the water and sanitation intervention ( Models 2a and 2b , Table 3 ) did not provide evidence that these interventions were effective in reducing transmission . However , results were unable to rule out both substantial beneficial and harmful effects ( WATSAN coefficients less than one and greater than one respectively ) , indicating that the data contained little information about the effects of the water and sanitation responses on transmission . This reflects the fact that in all three camps the interventions were not fully in place until the HEV epidemics were almost over ( S1 Fig ) . Fitting the models accounting for a second mode of transmission corresponding to an unobserved environmental reservoir showed that the data were only consistent with the majority of transmission occurring via this route if mean persistence of viable virus in this environmental reservoir was two weeks or more ( S2 Fig ) . Considering the potential effects of different vaccine usage scenarios , under the baseline model , reactive vaccination ( assuming two doses after the first 100 or 50 cases ) was capable of producing important though relatively modest reductions in cases and deaths ( Fig 4 , top row ) . These reductions were sensitive to the threshold number of cases before reactive vaccination was initiated: when the threshold was 100 , reductions in mortality compared to no vaccination scenarios were unlikely to exceed 25% even assuming the vaccine was given without age or pregnancy restrictions; for a threshold of 50 , mortality reductions of about 40% were plausible . In contrast , with no age-restriction on vaccine recipients and pre-emptive vaccination ( Fig 4 , bottom row ) , reductions in mortality of 100% were possible indicating that the herd immunity threshold had been reached . This was true whether two or three doses were administered , though in the former case uncertainty was far larger reflecting the lower precision in the estimated vaccine effectiveness of two doses ( S2 Appendix ) . The impact of excluding pregnant women from the population to vaccinate varied with scenario . In cases where vaccination had a high chance of achieving herd immunity ( i . e . when vaccine was used pre-emptively without age restriction ) excluding pregnant women had only a moderate negative impact on reductions in mortality . In contrast , when herd immunity was less likely to be obtained through vaccination ( i . e . when the vaccine was used reactively or with age restrictions ) excluding pregnant women led to substantially smaller reductions in total mortality ( and much smaller reductions in mortality in pregnant women ) . Comparing vaccination policies with and without age restrictions , restricting receipt of the vaccine to those between the ages of 16 and 65 years had a small impact on reductions in mortality in the reactive vaccination strategy , but a far larger negative impact on the pre-emptive vaccination strategy , a consequence of reducing the chance of achieving herd immunity in these latter strategies . These broad conclusions about the impact of different vaccination strategies were robust to precise details of model specification . In particular , under all 25 scenarios in the model with two modes of transmission ( Model 6 ) , similar patterns were seen ( S3 Fig ) , though the smallest reductions in mortality were seen when priors expressed the belief that most transmission was via this environmental route and viral persistence in this environment was long .
Our results indicate that in displaced population settings HEV can be highly transmissible . In one camp ( Paloga ) the estimated mean number of secondary cases per primary case at the start of the epidemic ( the basic reproduction number , R0 , ) exceeded 6 . 5 under all model assumptions . This number has important implications for vaccination policies . To achieve herd immunity requires successful immunization of a percentage of individuals given by 100-100/ R0 [32] . Thus , even taking the optimistic R0 value of 6 . 5 , we would require 85% of the population to be effectively immunized to achieve herd immunity . Herd immunity is desirable as it means a major epidemic will not be possible even though many in the population remain susceptible to infection . It is of particular relevance here because HEV-infected pregnant women face a greatly increased risk of mortality but the safety and efficacy of the vaccine in pregnant women remains to be established . Even when restricting vaccine use to non-pregnant 16–65 year olds ( for which the vaccine is licenced in China ) , the benefits could be substantial , with reductions in mortality of over 40% likely under baseline assumptions given pre-emptive vaccination with three doses . A robust finding was that pre-emptive vaccination was much more effective at preventing HEV cases and deaths than reactive vaccination . Estimates of reductions in mortality were based on the case fatality estimates taken from the meta-analysis of HEV genotypes 1 and 2 outbreaks by Rein et al [6] , as we lacked reliable local mortality data . In practice mortality rates might be expected to vary considerably between locations , and substantially higher mortality rates have been reported by some studies [7 , 33] . Evaluation of the potential benefits of HEV vaccination in an emergency setting should take into account the best estimates of local mortality rates . Our work sheds light on other important aspects of HEV epidemiology: we found consistent evidence that the mean latent period is between about 20 and 40 days and that a little over 10% of individuals infected with HEV are identified as cases . The data were less informative about the mean infectious period , though were consistent with the range 20–70 days suggested by previous data . Previous epidemiological investigations had suggested that household-level person-to-person spread may have been important in this epidemic [21 , 22] . Our results show that the epidemic curves in all three camps can be reproduced without positing the existence of a saturating environmental reservoir , and models with such a reservoir did not improve fits to data . In fact , models 1 and 4 had substantially lower DICs than more complex models ( including those explicitly accounting for an environmental reservoir ) suggesting that these simpler models should be preferred in the absence of strong evidence that a saturating environmental reservoir played a role in this epidemic . We found no evidence that the water and sanitation interventions reduced transmission , though these interventions were introduced late and the data provide little evidence for or against their effectiveness . Previous modelling has shown that such interventions have the potential to be highly effective if assumed to be capable of interrupting the dominant mode of transmission [34] . More recently , a matched case control study of HEV infection in a 2014 outbreak in a predominantly rural-nomadic population in Napak District , Uganda found that eating roadside food , drinking untreated water , and not always cleaning utensils were strongly associated with risk of HEV infection [33] . These findings are consistent with fecal oral transmission being the dominant mode of spread , and suggest that both contaminated drinking water and household transmission might play important roles . While our analysis does not allow us to quantify the relative importance of different transmission routes , the results do suggest that if an environmental reservoir was important for transmission in this epidemic the current number of infectious people was a good proxy for the infection pressure from the environment . Given the uncertainties about transmission routes and fundamental methodological challenges in making inferences about an unobserved environmental reservoir [35] , we performed extensive sensitivity analyses . While this did not make it possible to quantify the relative importance of different transmission routes it did shed light on the circumstances where a saturating reservoir would be consistent with the observed epidemic data . A key finding is that if the mean persistence of viable virus in the environment was seven days or fewer , the saturating environmental reservoir was estimated to play only a minor role in the epidemic . If mean persistence was two weeks or more , however , the data were compatible with such an environmental reservoir representing the dominant mode of transmission . Strengths of our work include the use of high quality epidemic data , extensive sensitivity analysis , and an analysis that allows us to incorporate information from previous HEV investigations . This work also has important limitations . First , estimates of HEV vaccine effectiveness from a trial in China may not generalise to a typical emergency setting or to age/ethnic groups not included in the trial . Moreover , the model assumes that when an individual is vaccinated they have an all-or-nothing response: either full protection against disease and ability to transmit infection or no protection . In practice , components of vaccine protection may be more complex [36] , and it is not known whether vaccinated individuals developing subclinical infections have reduced transmissibility [28] . Second , our analysis neglects age , spatial and household structuring , behavioural , biological and temporal heterogeneities that might affect HEV transmission . These may all play important roles in HEV epidemiology in emergency settings but we lacked data of sufficient resolution to meaningfully incorporate them . Delineating such factors is an important area for future work . There is some evidence that disease severity is affected by age [10] . By analogy with other viral infections , we might anticipate that transmissibility , infectious period and vaccine effectiveness also vary be age . Such differences could have an important influence on the relative effectiveness of different vaccination policies . Understanding such age effects , and incorporating them into models should be considered a priority . We assumed no prior immunity though lacked pre-epidemic sera to enable us to assess this assumption . If some people were immune prior to the epidemic , we are likely to have underestimated the basic reproduction numbers . Third , lacking any data on the efficacy of a single vaccine dose , we assumed it conferred no protective effect . This is a conservative assumption and may mean that we have underestimated the potential vaccine benefits . Given the logistical challenges of delivering three doses over a six month period , future research to better quantify the value of one and two doses of vaccine and perhaps a reduced interval between doses 2 and 3 would be valuable . Finally , it is not clear how relevant the modelling framework used here will be for community outbreaks where HEV is already endemic , such as the recent urban outbreak in Chad [37]; further work is needed to evaluate the potential impact of hepatitis E vaccination in different contexts . In conclusion , this work has shown a high transmission potential for HEV in displaced population settings , shed light on important natural history parameters , and has shown that mass vaccination campaigns in such high risk populations have the potential to lead to substantial reductions in mortality . In 2015 the Strategic Advisory Group of Experts on immunization could not recommend the routine use of the vaccine for population sub-groups including children aged less than 16 years and pregnant women but emphasized that the use of the vaccine during outbreaks of hepatitis E should be considered [38] . Our findings show that such an intervention , while not as effective as pre-emptive vaccination , could nonetheless have a major impact , particularly if vaccination can be safely extended to high risk groups excluded from vaccine trials . In particular , these results underline the need to prioritise evaluations of the vaccine in pregnant women [39] . | Hepatitis E virus is a leading cause of acute viral hepatitis in developing countries . About 20% of those infected develop clinical symptoms; of those , about 2% of non-pregnant cases and 20% of pregnant cases die . There is a safe and effective HEV vaccine that is licensed in China for those aged 16–65 years who are not pregnant . The potential for using this vaccine in outbreak settings has not previously been examined . We analysed data from one of the world’s largest recorded HEV epidemics . We estimated that one case infects , on average , between 4 and 9 others at the start of an epidemic . We found that vaccination restricted to those aged 16–65 who are not pregnant could reduce mortality in outbreak settings by between about 10 and 30% if used reactively ( initiating vaccination after the start of an epidemic ) ; pre-emptive vaccination of the same group could reduce mortality by 35–65% . Substantially higher reductions in mortality are likely if vaccination can be safely extended to pregnant women and other age groups without loss of effectiveness . However , even if this is possible , reactive vaccination is unlikely to reduce mortality by more than 50% while pre-emptive vaccination can reduce mortality by 80 to 100% . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"death",
"rates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"liver",
"diseases",
"health",
"care",
"viruses",
"preventive",
"medicine",
"vaccines",
"gastroenterology",
"and",
"hepato... | 2018 | Reactive and pre-emptive vaccination strategies to control hepatitis E infection in emergency and refugee settings: A modelling study |
More than 5% of alternatively spliced internal exons in the human genome are derived from Alu elements in a process termed exonization . Alus are comprised of two homologous arms separated by an internal polypyrimidine tract ( PPT ) . In most exonizations , splice sites are selected from within the same arm . We hypothesized that the internal PPT may prevent selection of a splice site further downstream . Here , we demonstrate that this PPT enhanced the selection of an upstream 5′ splice site ( 5′ss ) , even in the presence of a stronger 5′ss downstream . Deletion of this PPT shifted selection to the stronger downstream 5′ss . This enhancing effect depended on the strength of the downstream 5′ss , on the efficiency of base-pairing to U1 snRNA , and on the length of the PPT . This effect of the PPT was mediated by the binding of TIA proteins and was dependent on the distance between the PPT and the upstream 5′ss . A wide-scale evolutionary analysis of introns across 22 eukaryotes revealed an enrichment in PPTs within ∼20 nt downstream of the 5′ss . For most metazoans , the strength of the 5′ss inversely correlated with the presence of a downstream PPT , indicative of the functional role of the PPT . Finally , we found that the proteins that mediate this effect , TIA and U1C , and in particular their functional domains , are highly conserved across evolution . Overall , these findings expand our understanding of the role of TIA1/TIAR proteins in enhancing recognition of exons , in general , and Alu exons , in particular .
Alternative splicing of mRNA precursors allows the synthesis of multiple mRNA isoforms from a single primary transcript [1]–[4] . Recent analyses indicate that the majority of human genes are alternatively spliced , thus contributing significantly to human transcriptome diversity [5] , [6] . Accurate removal of introns occurs by a two step reaction , conserved from yeast to mammals , that takes place in a large macromolecular complex termed the spliceosome . The spliceosome consists of five small nuclear RNAs ( snRNAs; U1 , U2 , U4 , U5 and U6 ) and over 200 associated proteins . Four degenerate sequences are recognized by the spliceosome: the 5′ and 3′ splice sites ( 5′ss and 3′ss ) , located at the 5′ and the 3′ end of each intron , the polypyrimidine tract ( PPT ) and the branch point sequence ( BPS ) both located upstream of the 3′ss [7] . The 5′ss consensus sequence in higher eukaryotes is comprised of nine bases that bridge the exon-intron boundary; this region is bound by a complementary region along the RNA component of the U1 snRNP . In most pre-mRNAs the base pairing of U1 snRNP and 5′ss is not perfect . Increased complementarity between U1 snRNP and the 5′ss strongly contributes to 5′ss selection [8] , [9] and can shift the splicing pattern from alternative to constitutive [10] , [11] . In metazoans , the four main splice signals are insufficient to allow accurate splicing . It has been estimated that these splicing signals provide , at most , half of the information required for recognition by the splicing machinery [12] . Studies of the molecular basis of splicing revealed the existence of exonic and intronic cis-acting regulatory sequences ( ESRs and ISRs , respectively ) , which bind trans-acting factors and regulate the precise excision of introns from within eukaryotic pre-mRNA . These cis-acting elements are classified as exonic or intronic splicing enhancers and silencers , which promote or inhibit splicing , respectively . These sequences have been identified using a wide array of experimental and computational methodologies [13]–[21] and interact in a complex manner to allow precise splicing [22] . Aberrant regulation of splicing is linked with a wide array of disease states , including cancer [23]–[25] . The ESRs and ISRs are recognized by trans-splicing factors , which usually contain one or more RNA binding domains as well as additional domains that are essential for recruitment of the splicing apparatus and for splice site pairing . The TIA1 ( T-cell intracellular antigen 1 ) and TIAR ( TIA1 related protein or TIAL ) proteins are examples for two such splicing factors . These proteins contain an RNA-recognition motif ( RRM ) known as RRM2 that specifically binds U-rich RNA sequences within introns [26] . The proteins are also characterized by two additional RRMs and a glutamine rich carboxyl terminal region [26]–[28] . Binding of TIA1 protein to uridine-rich sequences downstream of weak 5′splice sites helps to recruit U1 snRNP to the 5′ss through protein-protein interactions involving the glutamine rich domain of TIA1 and the U1-specific protein U1C [29]–[32] . TIAR can also recruit U6 snRNP to a pseudo-5′ss that is followed by a U-rich sequence located within a 200-bp element regulating alternative splicing of the calcitonin/CGRP gene [33] . Because of their affinity for U-rich sequences , TIA proteins are often antagonized by the pyrimidine tract binding protein ( PTB ) , a general repressor of exon inclusion [34]–[37] . The functions of TIA1/TIAR proteins and homologues have been demonstrated in several model organisms . In yeast , NAM8 , PUB1 and NGR1 are related to the TIA proteins and have similar domain organizations . NAM8 stabilizes commitment complexes and facilitates weak 5′ss recognition by interacting with non-conserved sequences downstream of the 5′ss [38] , [39] . The mouse homologs of the TIA proteins were shown to be functional as well [29] , [40] . In Drosophila , Rox8 was shown to be the functional homolog , based on RNA interference experiments [41] , and in plants , the related proteins UBP1 and RBP45 were shown to interact with U-rich elements and enhance splicing in [42]–[44] . Much less is known about the TIA homologs among other eukaryotes . More than 5% of alternatively spliced internal exons in the human genome are derived from Alu elements . Throughout the course of evolution , some intronic Alus have accumulated mutations that led the splicing machinery to select them as internal exons , a process called “exonization” [10] , [45]–[47] . The majority of Alu-derived exons are alternatively spliced [46] , [48] allowing the enrichment of the human transcriptome with new isoforms without compromising its original repertoire [49] . Alus originated from the 7SL RNA gene [50] . They belong to the short interspersed elements ( SINE ) family of repetitive elements and are unique to primates [51] , [52] . More than one million copies are dispersed throughout the human genome with a majority located in introns [46] . A typical Alu element is ∼300 nucleotides long , consisting of two arms ( left and right ) joined by an A-rich linker and followed by a poly ( A ) tail . The right and left arms are highly similar , sharing ∼80% of their sequence . Both arms contain potential splicing signals and both can undergo exonization , although exonizations tend to occur from the right arm [53]–[55] . When Alus insert into introns in the antisense orientation ( relative to the coding sequence ) , the poly ( A ) tail becomes a poly ( U ) in the mRNA precursor and thus can serve as a PPT . This PPT presumably leads the splicing machinery to select a downstream AG as the 3′ss and a further downstream GT or GC sequence as the 5′ss [55] . Exonizations can occur either from the right Alu arm or from the left arm . In the first case , both the 3′ss and the 5′ss are selected from the right arm , whereas in the latter both signals are selected within the left arm . Only few cases were known to us in which the 3′ss occurs in one arm , and the 5′ss in the other , although there are many cases in which potential splicing signals are present [46] , [56] . We thus hypothesized that the second PPT sequence , located within the Alu element and separating the two Alu arms from each other , limits splice site selection and causes both splicing signals to be selected from within the same arm . To evaluate this hypothesis , we created an Alu-based model system of two competing 5′ss separated by a PPT . The PPT in this system is not the classical PPT located upstream of the 3′ss , but rather is a pyrimidine-rich stretch located downstream of the 5′ss . We showed that the presence of the PPT sequence led to selection of the upstream 5′ss even in the presence of a stronger 5′ss downstream . Deletion of the PPT sequence shifted selection to the stronger 5′ss . We show that this enhancing effect depended on the strength of the downstream 5′ss and the efficiency of base pairing to U1 snRNA . PPTs of 3-to-9 nucleotides modulated different levels of 5′ss usage . We also show that this enhancing effect is mediated by the binding of TIA proteins to the Alu PPT and that the function of these proteins is distance-dependent . To obtain a wide-scale overview on the evolution of the TIA proteins and their binding sites , we analyzed over 1 million introns from 22 eukaryotes and found that throughout eukaryotic evolution there has been an increased tendency for PPTs to occur within ∼20 nt downstream of the 5′ss . Among most metazoans , the strength of the 5′ss inversely correlates with the presence of a downstream PPT , indicating the functional importance of this signal . Finally , we searched for TIA homologs across evolution and found that functional regions of these proteins are highly conserved . Taken together , these findings indicate that throughout eukaryotic evolution , the TIA proteins have served as key players that have helped shape introns and that these proteins also mediate the formation of new exons , as in the context of Alu exonizations .
The ADAR2 minigene , containing the human genomic sequence of exons 7 , 8 and 9 ( 2 . 2kb ) , was previously cloned [45] . The PCR products were restriction digested and inserted between the KpnI/BglII sites in the pEGFP-C3 plasmid ( Clontech ) , which contains the coding sequence for Green Fluorescent Protein ( GFP ) . The 350-nt intronic sequence originating from intron 11 of the IMP gene was amplified by PCR using 5′ phosphorylated primers and inserted downstream of the PPT sequence of the intronic left arm of the Alu element . For RNA pull-down assays , three fragments containing the 5′ss of the Alu exon and the PPT downstream of it were amplified by PCR from WT , ΔPPT and rep_PPT minigenes and cloned into the BamHI/EcoRI sites of pBluescript KS+ . The TIA1b and TIARb cDNAs ( kind gifts from Juan Valcárcel ) were cloned into the pEGFP-C1 vector and the U1 gene was cloned into the pCR vector . For the sequences of the ADAR minigene insert and pBluescript KS+ inserts see Text S1 . Site-directed mutagenesis was carried out to introduce mutations into the ADAR2 and U1 minigenes by PCR using oligonucleotide primers containing the desired mutations . Mutations creating deletions in wild-type minigenes were performed by PCR using 5′ phosphorylated primers flanking the sequence to be deleted ( see Supplementary Table 1 in Text S1 for list of primers ) . PCR was performed using PfuTurbo DNA polymerase ( Stratagene ) with an elongation time corresponding to 2 min for each kb . The PCR products were treated with DpnI ( 20 U , New England BioLabs ) at 37°C for 1 h . Plasmid mutants were ligated using T4 DNA Ligase ( New England BioLabs ) at 37°C for 2 h . The mutant DNA was transformed into E . coli XL1-competent cells . DNA was extracted from selected colonies by mini-prep extraction ( Promega ) . All plasmid sequences were confirmed by sequencing . 293T cells were cultured in Dulbecco's Modification of Eagle medium , supplemented with 4 . 5 g/mL glucose ( Biological Industries , Inc . ) , 10% fetal calf serum ( FCS ) , 100 U/mL penicillin , 0 . 1 mg/mL streptomycin and 1 U/mL nystatin ( Biological Industries , Inc . ) . Cells were cultured in 6-well plates under standard conditions at 37°C in 5% CO2 . Cells were grown to 50% confluence and transfection was performed using 3 µL TransIT LT1 ( Mirus ) with 1 µg of plasmid DNA . RNA was isolated and harvested after 48 h . Total RNA was extracted using Trizol Reagent ( Sigma ) , followed by treatment with 1 U RNase-free DNase ( Ambion ) . Reverse transcription ( RT ) was preformed for 1 h at 42°C using an oligo dT reverse primer and 2 U reverse transcriptase of avian myeloblastosisvirus ( AMV , Roche ) . The spliced cDNA products derived from the expressed minigenes were detected by PCR using an ADAR2 exon 7 forward primer ( 5′CCCAAGCTTTTGTATGTGGTCTTTCTGTTCTGAAG3′ ) and a pEGFP-specific reverse primer ( 5′CGCTTCTAACATTCCTATCCAAGCGT3′ ) . Amplification was performed for 28 cycles to maintain a linear relationship between the input RNA and signal [18] . Each cycle consisted of 30 sec at 94°C , 45 sec at 61°C and 1 . 5 min at 72°C . The RT-PCR products were separated on a 2% agarose gel and confirmed by sequencing . The relative ratios of RNA products using 5′ssA or 5′ssB were measured using ImageJ software ( http://rsb . info . nih . gov/ij/index . html ) , as we previously established that ImageJ quantification for ADAR2 RT-PCR products correlates with real-time RT-PCR quantification produced by the Roche LightCycler PCR and detection system [45] . Semi-quantitative RT-PCR of three independent biological replicates of three ADAR minigene mutants revealed standard deviations of 0 . 6% to 5 . 3% of the relative ratios of RNA products . Linearized pBluescript KS+ plasmids were used as templates for the synthesis of biotinylated RNAs by using T7 RNA polymerase ( Promega ) and biotinylated-16-UTP ( Roche ) following manufacture recommendations . Total cell extract from 1 mg of HeLa cells was incubated with 1 µg of biotin-labeled RNA and rotated for 4 h at 4°C in binding buffer containing 10 mM HEPES , pH 7 . 5 , 40 mM KCl , 3 mM MgCl2 , 5% glycerol , supplemented with 40 units of RNasin ( Promega ) and 5 mg/ml heparin ( Sigma ) . The biotin-labeled RNA was isolated using streptavidin-conjugated beads ( Fluka ) and was washed with binding buffer for four times . The presence of TIA1/TIAR in the pull-down pellet was verified by western blot analysis as described below . Lysis buffer ( 50 mM Tris at pH 7 . 5 , 1% NP40 , 150 mM NaCl , 0 . 1% SDS , 0 . 5% deoxycholic acid , protease inhibitor cocktail and phosphatase inhibitor cocktails I and II; Sigma ) was used for protein extraction . Lysates were centrifuged for 30 min at 14 , 000 rpm at 4°C . Total protein concentrations were measured using BioRad Protein Assay ( Bio-Rad ) . Proteins were separated in 12% SDS-polyacrylamide gel electrophoresis ( SDS-PAGE ) and then electroblotted onto a Protran membrane ( Schleicher and Schuell ) . The membranes were probed with anti-TIA1 ( C-20 , Santa Cruz Biotechnology ) , anti-TIAR ( C-18 , Santa Cruz Biotechnology ) , anti-HSC70 ( B6 , Santa Cruz Biotechnology ) , anti-GFP ( MBL ) or anti-α-tubulin ( Sigma ) , followed by the appropriate secondary antibody . Immunoblots were visualized by enhanced chemiluminescence ( Lumi-Light Western Blotting Substrate; Roche ) and exposure to X-ray film . To examine the prevalence of Alu exons with a 3′ss selected within the right arm and a 5′ss within the left arm , we began by querying the TranspoGene webserver [57] for cases of exons overlapping Alu elements in the antisense orientation that were supported by at least one EST . This query yielded 744 such exons . Since we were interested only in cases in which both the 5′ss and the 3′ss occurred within the Alu sequence , we next filtered out all cases in which either of these signals occurred outside of the Alu sequence; this yielded 548 sequences . To map the 3′ss and the 5′ss of each exonization event to either the left or the right arm performed pairwise alignments between each Alu and the Alu-Jo consensus sequence based on the Needleman-Wunsch algorithm for global alignment [58] . To identify PPTs , we used the algorithm we previously developed for identifying polypyrimidine tracts that is described in detail in [54] . We set a minimum score threshold of 6 , which dictates that a PPT sequence must consist of at least six consecutive pyrimidines . Notably , the identified stretch may also be longer and may contain non-pyrimidines as long as the overall enrichment score is ≥6 . For each intron of each organism , we first masked the 30 terminal nucleotides and then searched for pyrimidine-rich stretches within the 300 first nucleotides of the intron or within the entire remaining stretch of the intron in cases of introns shorter than 330 nucleotides . The 30 terminal nucleotides were masked in order to avoid contamination by PPT at the 3′ end of the intron . To derive the plots indicating the presence of PPTs for each organism , we summarized for each of the first 100 intronic positions the number of PPTs covering that position and divided this number by the number of introns reaching that position . The 5′ss of all introns were scored based on their adherence to a position-specific scoring matrix ( PSSM ) for the 5′ss consensus for each organism . The 5′ss was defined as 12 positions as in [54] , including four exonic and eight intronic positions . The 5′ss score was calculated as:where A is the sequence to be scored and fi , Ai is the PSSM frequency at position i of the ith nucleotide in sequence A . A dataset of 3634 alternative 5′ss events , based on the AltSplice track in University of California Santa Cruz ( UCSC ) Genome Browser ( http://genome . ucsc . edu/ ) was compiled . We discarded all events in which the distance between the two alternative 5′ss was less than 12 nt , in order to allow the presence of a PPT . This left 2 , 296 alternative 5′ss events . PPTs within the first 25 nt ( or less ) downstream of each of the two competing 5′ss were found as described above . We compiled a set of 684 known splicing factors with RNA binding domains from multiple species . We then grouped these proteins into 38 families ( see Supplementary Table 3 in Text S1 ) . We grouped known TIA1/TIAR and polyuridylate binding proteins ( PUB1 ) proteins into the same family and known NAM8 and NGR1 proteins into a different family . For each family we built a hidden Markov model ( HMM ) for each of the RNA binding domains ( RRM or KH-type ) using Hmmer [59] ( http://hmmer . janelia . org ) . We considered as candidate homologs those proteins that had collinear hits for a multidomain protein in the right order or a single hit for a single domain protein . For each of the sets of homologous RRMs we built a maximum parsimony tree using the close-neighbour-interchange algorithm with search level 3 . The initial trees were obtained with random addition of sequences using 10 replicates . A candidate protein was labeled as an ortholog of a known protein if its RRMs grouped consistently in the trees with most of the known RRMs ( see Dataset S1 ) . Multiple alignments were built using t-coffee [60] and phylogenetic analyses were performed with MEGA4 [61] . In order to establish the conservation between proteins or between domains we used two measures: the average pairwise identity and the multiple alignment conservation score . For the average pairwise identity we calculated , for each pair , the proportion of identical amino acids over the gapless positions and averaged over all pairs in the multiple sequence alignment ( MSA ) . To calculate the conservation score ( MSA score ) , we first calculated the score for each gapless column of the MSA by determining the proportion of amino acid pairs M in the column that were identical:where δ ( Ai , Aj ) is 1 if Ai = Aj and 0 otherwise . The MSA score was then computed as the average of the column scores over all the gapless columns N:
The left and right arms of Alu elements are highly similar and both contain potential splice sites [53] . However , in most cases exonization occurs almost exclusively within either the right arm or the left , but not both . We were interested in determining how often exonizations overlapped both arms . We compiled a dataset of Alus in the antisense orientation involved in exonization events , based on a TranspoGene query [57] . We then used pairwise alignments against an Alu consensus sequence to map each signal to the right or the left arm . Of 548 cases of exonization events within Alu elements , 405 ( 74% ) occurred from within the right arm only , 114 ( 21% ) occurred from within the left arm only , and in only 29 ( 5% ) was the 3′ss selected from within the right arm and the 5′ss from within the left . In light of our finding that Alu exonization events do not tend to cross the border between the two arms , we hypothesized that the PPT sequence separating the two Alu arms prevents exonization into downstream sequences . To examine this hypothesis , we used a modified version of the ADAR2 minigene as a model system . The original ADAR2 minigene contains exons 7 to 9 of the human ADAR2 gene , along with the introns between them . Exon 8 is an Alu exon that originated from the right arm of the Alu element and is alternatively spliced . In order to investigate the effect of the PPT in isolation of the pseudo-exon effect of the left arm [53] , we began by inserting a 350-nt sequence between the PPT and the potential 3′ss of the left arm ( Figure 1A ) . By separating the right arm from the left by 350nt , we eliminated the effect of the intronic arm on the Alu exon , thus shifting splicing of the Alu exon from alternative to constitutive splicing [53] . We next generated a 5′ss 68 nucleotides downstream of the PPT of the intronic left arm ( 5′ssB in Figure 1A ) . This 5′ss is stronger in terms of Senapathy score ( http://ast . bioinfo . tau . ac . il/SpliceSiteFrame . htm ) than the 5′ss of the Alu exon ( 5′ssA in Figure 1A ) . Thus , this system contains two potential 5′ss separated by a PPT sequence ( Figure 1A ) ; we will henceforth refer to this minigene as ADAR WT , and to the PPT following the Alu exon ( originating from the left arm ) as PPT . The minigene was transfected into 293T cells , total cytoplasmic RNA was extracted after 48 hours and 5′ss selection was examined by RT-PCR analysis using primers specific to the minigene mRNA . Although 5′ssA is weaker than site B , it was almost exclusively selected ( Figure 1B , lane 1 ) . However , when the PPT sequence was deleted or replaced by a sequence that did not contain any splicing regulatory elements ( see sequence in Supplementary Methods in Text S1 ) there was a shift in 5′ss selection from site A to site B ( Figure 1B , lanes 2 and 9 , respectively ) . These results indicated that the PPT enhances selection of a weaker upstream 5′ss in the presence of a stronger 5′ss downstream . To determine whether this enhancing effect of the PPT was dependent on the strengths of the 5′ss , we made mutations in 5′ssB to strengthen it over a Senapathy score range of 79 . 87 to 100 . Specifically , we inserted different combinations of T→A mutations in positions 3 and 4 , and a T→C mutation in position −3 . As site B was strengthened , there was a gradual shift towards selection of this site despite the presence of the PPT sequence ( Figure 1B , compare lane 1 to lanes 3–7 ) . Strengthening of 5′ssA in combination with a deletion of the PPT sequence resulted in its constitutive selection ( Figure 1B , lane 8 ) . These results imply that there is a delicate interplay between the PPT and the strengths of the splice sites flanking it . The presence of a PPT sequence enables selection of a weak 5′ss upstream , but only if the downstream 5′ss is weaker than a certain level . Once the competing 5′ss is strong enough , it is selected despite the presence of the PPT . To determine the length of the PPT required for efficient selection of site A , we shortened the 14-nt PPT sequence separating site A from site B to nine , six and three consecutive uridines . Shortening the PPT resulted in a shift from site A to B ( compare Figure 1C , lane 1 to lanes 3–5 ) . When the two adenosine bases within the PPT sequence in the WT minigene ( see minigene sequence in Supplementary Methods in Text S1 ) were replaced with uridines to obtain a PPT of 14 consecutive uridines , there was little change in 5′ss selection compared to the WT ( Figure 1C , lane 2 ) . These results indicated that a PPT with at least nine consecutive uridines results in maximal selection of 5′ssA . We then set out to examine whether the competition between the two putative 5′ss is mediated through the binding to U1 snRNA . 293T cells were co-transfected with the ADAR WT minigene and with a U1 gene containing mutations to enhance complementarity to site B ( Figure 2A ) . A schematic illustration of the base pairing between site B and U1 is presented in Figure 2B . Mutations were made at positions 5 , 6 and 11 of U1 snRNA to improve its base pairing to 5′ssB ( these U1 snRNA mutations are complementary to positions 4 , 3 and −3 in 5′ssB , respectively ) . Improving the binding of U1 snRNA to 5′ssB by insertion of all three mutations enhanced its selection ( Figure 2A , compare lane 1 to lane 6 ) , indicating that complementarity to U1 snRNA is critical to 5′ss selection in this competitive situation . Notably , an individual mutation at position 5 of U1 snRNA or the combination of mutations in positions 5 and 11 did not improve base pairing of U1 snRNA to 5′ssB . This is presumably explained by the fact that the mutation at position 5 enhances the ability of U1 snRNA to base pair not only with 5′ssB but also with 5′ssA ( Figure 2B ) . The reciprocal experiment , in which a U1 snRNA was designed with complementarity to 5′ssA , caused activation of a cryptic intronic site that resembles 5′ssA ( data not shown ) . It has been previously shown that TIA proteins ( TIA1 and TIAR ) activate weak 5′ss that are located upstream of U-rich sequences [29]–[33] . To test whether the enhancing effect of the PPT sequence on the selection of the weak 5′ssA is mediated by the binding of TIA1/TIAR , we transfected 293T cells with three mutant minigenes that contained 5′ssB of different strengths and thus exhibited different levels of site B selection . For variants B ( 3A4A ) , B ( 3A ) and B ( 4A ) , 5′ssB was selected in 100% , 82% and 44% of the transcripts , respectively ( Figure 1B ) . We also co-transfected the cells with vectors containing TIA1 and TIAR . In addition , we co-transfected the cells with a vector containing the PTB cDNA , which is also known to bind pyrimidine rich sequences [35] . As shown in Figure 3 , co-transfection of the indicated mutants with TIA1 and TIAR cDNA induced a shift of splicing towards use of 5′ssA . Western blot analysis revealed that both proteins were expressed at the same level ( see Supplementary Figure 1 in Text S1 ) . However , co-transfection of the same mutants with PTB did not affect the splicing pattern of any of these ADAR mutants . We subsequently depleted levels of the TIA proteins via siRNA experiments . In these experiments we did not observe a shift in the 5′ss selection , which may be explained either by functionality of the residual levels following depletion or by involvement of additional factors ( data not shown ) . We then determined whether TIA1 and TIAR could bind to the PPT sequence downstream of 5′ssA . Three fragments containing the 5′ss of the Alu exon and the downstream PPT were amplified by PCR from the WT ADAR minigene and from the mutant minigenes in which the PPT sequence was deleted or replaced ( ΔPPT and rep_PPT minigenes , respectively , see Figure 1B ) . The fragments were cloned into pBluescript KS+ plasmids ( see insert sequences in Supplementary Methods in Text S1 ) and in vitro transcription using T7 RNA polymerase and biotinylated-16-UTP was performed . Biotinylated transcripts were incubated with HeLa extracts , isolated by streptavidin-conjugated beads and TIA1 and TIAR was detected using western blot analysis . Our results indicate that TIA proteins strongly interact with the RNA transcript corresponding to 5′ssA and the PPT sequence downstream of it: The anti-TIAR and anti-TIA1 antibodies detected double bands at 40 and 44 kD , corresponding to two different isoforms of TIAR and TIA1 , respectively [62] , when the WT biotinylated RNA was used ( Figure 3D , lane 1 ) . The TIA1 and TIAR bands were completely absent when the PPT sequence was deleted or replaced ( Figure 3D , lanes 2 and 3 , respectively ) . Previous studies have experimentally demonstrated that the splicing-enhancing function of U-rich sequences is observed when they are located immediately downstream from the activated 5′ss [29] , [30] . In our model system the PPT is located 18 nt from 5′ssA yet still enhances selection of 5′ssA . To examine whether positioning of the PPT sequence in closer proximity to 5′ssA would enhance its selection further , we deleted five nucleotides from the 18-nt sequence separating the PPT from 5′ssA ( indicated as −5nt_PPT in Figure 4 ) , using the B ( 3A ) , B ( 4A ) and B ( 3A4A ) mutants . Deletion of five nucleotides from the 18-nt sequence separating the PPT from 5′ssA also shortened the distance between 5′ssA and B . Interestingly , deletion of five nucleotides resulted in a shift of splicing from 5′ssB to 5′ssA ( Figure 4 , compare lanes 1 and 2 in each panel ) . Deletion of five or ten nucleotides of the sequence lying between the PPT sequence and 5′ssA had the same effect on 5′ss usage in mutants B ( 3A ) , B ( 4A ) and B ( 3A4A ) ADAR mutants ( Supplementary Figure 2A in Text S1 ) and a negligible effect on the splicing pattern of the ADAR WT minigene ( Supplementary Figure 2B in Text S1 ) . Furthermore , deletion of five nucleotides from the sequence separating the PPT sequence and 5′ssA in the presence of TIAR resulted in predominant selection of 5′ssA ( Figure 4 , compare lanes 1 and 4 in each panel ) . Taken together , these results demonstrate that the enhancing effect of TIAR on the selection of a weak 5′ss decreases with distance . Our analyses thus far indicated that in our Alu model , the PPT between the two Alu arms was bound by TIA proteins and enhanced selection of the weaker , upstream 5′ss . We were thus interested in understanding the impact of TIA proteins across evolution . Specifically , we focused on three components: the TIA binding sites on pre-mRNA , the TIA proteins , and the protein U1C , which serves as a link between the TIA proteins and the 5′ss [31] . It has been previously shown that the 5′ end of human introns are enriched in U-rich tracts [63] , but other organisms have not been analyzed for this phenomenon . To determine how wide-spread this enrichment is , we determined the prevalence of PPTs downstream of the 5′ss in a dataset of over 1 million introns from 22 organisms spanning all four major eukaryotic kingdoms: plants , protozoans , fungi and metazoans ( Figure 5A ) . Strikingly , we found an enrichment of PPTs downstream of 5′ss in almost all organisms in the dataset ( Figure 5C and Supplementary Figure 3A in Text S1 ) . PPTs were found in ∼20 to 40% of the introns and , in most cases , the center of the PPT was located between positions 15 and 25 downstream of the 5′ss ( see Supplementary Table 2 in Text S1 ) . The mean lengths of the PPTs ranged from 10 to 14 nucleotides depending on the organism ( Supplementary Table 2 in Text S1 ) . Notably , among several fungi , including S . pombe , U . maydis , Y . lipolytica and E . gossypi , as well as in the protozoan C . parvum , the pyrimidine-rich peaks were less pronounced . This may be indicative either of functional aspects , or may result from the fact that these organisms have fewer introns , making our measurements in these organisms less reliable . We hypothesized that if the PPTs downstream of the 5′ss are of functional importance in the context of splicing , the presence of these sequences would anti-correlate with the strength of the 5′ss , as they are expected to compensate for weak 5′ss . To assess whether such an anti-correlation exists , we divided all introns into four equally-sized bins of increasing 5′ss strengths . For each bin , we calculated the prevalence of a PPT beginning within the first 20 nt of the intron . Our results demonstrate a clear inverse correlation between 5′ss strength and the presence of a pyrimidine-rich stretch downstream of the 5′ss among all metazoans , excluding C . elegans ( Figure 6A ) . Such an anti-correlation was observed in the plant A . thaliana as well . These correlations were all highly statistically significant ( Supplementary Table 2 in Text S1 ) . However , these anti-correlations were not observed among most fungi and protozoans ( Supplementary Figure 3B in Text S1 ) . Thus , these results suggest that among most metazoans and in the plant A . thaliana , a pyrimidine-rich stretch downstream of the 5′ss compensates for the presence of a weak 5′ss . This is in agreement with our results pertaining to the Alu sequence and with previous molecular studies that found that pyrimidine-rich stretches support the inclusion of weakly defined exons [29] , [30] , [63] . Given our observation that PPTs downstream of the 5′ss are prevalent throughout evolution , we were next interested in obtaining an evolutionary perspective regarding the TIA proteins , which potentially bind this signal . TIA1 and TIAR proteins are quite similar ( 81% identity ) , each contains three RNA-recognition motifs ( RRMs ) and a glutamine ( Q ) rich C-terminus [27] , [28] and were shown to have redundant activities in splicing [34] , [37] , [64] . Additionally , we considered two proteins in S . cerevisiae that have high similarity to TIA1/TIAR , namely PUB1 and NAM8 . Both bind RNA [38] , [65] , [66] and also have three RRMs . NAM8 , which is a constitutive component of the U1 snRNP , binds in a non-specific manner downstream of the 5′ss and affects 5′ss selection [38] and has no counterpart in the mammalian U1 snRNP . As negative controls we included proteins that share high sequence similarity and have similar domain configurations , like the Negative Growth Regulatory protein ( NGR1 ) from S . cerevisiae and additional protein families with RNA binding domains ( Supplementary Table 3 in Text S1 ) . Using a combination of hidden Markov models ( HMMs ) and construction of phylogenetic trees for the candidates ( Supplementary Figure 4 in Text S1 ) , we found homologs for TIA1/TIAR in all analyzed metazoans ( Figure 5B ) . In addition , we found that A . thaliana and all fungi , except for S . pombe , have homologs of PUB1 ( Figure 5B ) . We also found that all fungi , except for C . neoformans and U . maydis , have homologs of NAM8 , whereas its close relative , NGR1 , is only present in the group of the Saccharomycetaceae ( D . hansenii , A . gossypii , K . lactis , C . glabrata and S . cerevisiae ) . Finally , we could not detect any clear homologs of TIA1/TIAR , NAM8 or PUB1 in the protozoa D . discoideum or C . parvum . These results highlight several points . First , among all analyzed organisms excluding protozoa , at least one TIA1/TIAR or PUB1 homolog was found . Second , most organisms for which we demonstrated an anti-correlation between PPT prevalence and 5′ss strength have either TIA1 or TIAR . One exception to this is C . elegans , in which there is a TIA1/TIAR homolog , but not a PPT/5′ss anti-correlation , and another is A . thaliana , in which an anticorrelation was observed but we found no TIA1/TIAR homologs ( see Discussion ) . Finally , S . pombe is an exception among fungi since it lacks any TIA1/TIAR or PUB1 homologs; it also lacks a clear pyrimidine-rich peak downstream of the 5′ss . The N-terminal RRM domain in TIA1/TIAR ( RRM1 ) is important for TIA1 activity and enhances the interaction of the Q-rich C-terminal domain with the U1 snRNP [31] . The other two RRMs , RRM2 and RRM3 , contact the pre-mRNA , although only RRM2 binds specifically to uridine-rich motifs [31] . RRM2 is the most conserved domain across all homologous proteins ( TIA1/TIAR , NAM8 and PUB1 ) , with multiple alignment conservation score of 0 . 65 , as opposed to 0 . 37 and 0 . 4 for RRM1 and RRM3 , respectively ( Figure 5B ) , and 47% average pairwise identity , as opposed to 37% and 41% for RRM1 and RRM3 , respectively . A multiple alignment depicting the conservation of RRM2 across TIA homologs is presented in Figure 6C and alignments for RRM1 and RRM3 are presented in Supplementary Figures 5 and 6 in Text S1 , respectively . This conservation underscores the evolutionary importance of the TIA proteins and implies that the mechanism by which TIA homologs bind to RNA has remained conserved throughout evolution . The recruitment of the U1 snRNP by TIA1 takes place through the interaction of the glutamine-rich ( Q-rich ) C-terminus of TIA1 with N-terminus of U1C , a protein component of U1 snRNP [31] . We therefore examined U1C conservation . We found U1C homologs in all species analyzed and observed a high degree of conservation among N-terminal regions ( Figure 6D ) with an average of 69% pairwise similarity in the first 20 positions and much lower conservation levels in downstream residues . In parallel , we examined the extent of conservation of the Q-rich C terminus of the TIA proteins . Although the precise order of amino acids at the C terminus varies , a distinct and statistically significant enrichment was observed in the Q-rich region among the vast majority of TIA1/TIAR/PUB homologs with respect to all other proteins of similar size . Furthermore , no enrichment in Qs was found among relevant controls with high sequence similarity to TIA proteins in other regions ( see Supplementary Results in Text S1 for a detailed analysis ) . Thus , the machinery involved in TIA regulation of splicing is conserved throughout evolution , from the sequences of functional regions of the involved proteins to the binding sites in the pre-mRNA .
This study was motivated by our finding that Alu exonization events involving both Alu arms occur in only ∼5% of Alu exons . Several factors probably limit exonization events across the arms of Alu elements . For example , the lengths of exons are known to be constrained with internal exons averaging 145 nucleotides in length . Alu exons within right arms average 110 nucleotides in length [67] , whereas exons that encompass sequence from both arms tend to be between 200 and 250 nucleotides long . Thus , exonizations occurring from a single arm yield exons that are more optimal in length . However , approximately 20% of human exons are longer than 200 nt [68] , strongly contrasting with only 5% of Alu exons that contain sequences from both arms . We hypothesized , and subsequently demonstrated , that the PPT sequence separating the two arms may be involved in limiting exonization across arms . The presence of a PPT enhanced the selection of the 5′ss of the right arm Alu exon even in the presence of a stronger splice site downstream . Conversely , in the absence of a PPT sequence between the two splice sites , the stronger downstream site was selected , indicating that in the absence of the PPT , the rules of simple competition apply . In subsequent analyses we were able to determine that the effect of the PPT on the Alu 5′ss selection is mediated by TIA1/TIAR proteins . This led us to conduct a bioinformatic analysis in which we examined the machinery involved in TIA regulation across evolution . This machinery , from the binding signal on the pre-mRNA to the sequences of the TIA and U1C proteins , is conserved and , for most metazoans , the presence of a polypyrimidine stretch anti-correlates with 5′ss strength . Interestingly , our findings may also explain why most exonizations tend to occur predominantly from the right arm of Alu elements and not from the left [53] . A previous study showed that exons from within left arms tend to be shorter , depleted in exonic splicing enhancers ( ESEs ) and enriched in exonic splicing silencers with respect to those from right arms [67] . Here we showed that the presence of a PPT downstream of the right arm Alu 5′ss , which is intrinsically embedded in the structure of a typical Alu element , enhances the selection of right arm Alu exons . Such an effect is not possible in the left arm and this might reduce the potential for Alu exonizations from the left arm . Our study using the Alu model system highlights a novel aspect of TIA1/TIAR proteins: These proteins activate a splice site at some distance from their binding site . Previous studies in human systems demonstrated that TIA1 only activates 5′ splice sites immediately followed by U-rich sequences [30] , [32] , although one study suggested , but did not conclusively prove , that TIA1 may be active from greater distances [62] . In our model system , the PPT was located 18 nt from the 5′ss of the Alu exon and the TIA1/TIAR proteins activated its selection . This is similar to the activity of the yeast TIA homolog NAM8 which can activate a 5′ss 46 nt downstream of its binding site [39] . In this respect , our results concur with recent findings , based on depletion of TIA proteins , that demonstrated a correlation between the magnitude of the change in exon skipping and the distance between U-rich motifs and the 5′ss [63] . The function of TIA proteins from a distance may be mediated by other splicing factors or by a formation of pre-mRNA secondary structures that bring together the U-rich sequence and the 5′ss to be activated . In our experimental system , we focused on the regulative role of the TIA proteins . The reason we focused on these proteins are ( 1 ) that the regulation was mediated through the binding to a pyrimidine-rich stretch downstream of the 5′ss , which is a classical mode of regulation of the TIA proteins , and ( 2 ) we ruled out PTB , which could potentially also have played a role in this context . However , other splicing factors can bind pyrimidine-rich stretches on the one hand , and play a role in splicing , on the other . Two such proteins are U2AF65 and PUF60: U2AF65 facilitates 3′ splice-site recognition at the early stages of spliceosome assembly , and PUF60 was found to functionally substitute for U2AF65 [69] , [70] . Despite the fact that classically these two proteins are mostly known for their involvement in the context 3′ss selection , two considerations could suggest that they might potentially play a role in our system as well: First , the fact that we observed an effect of the PPT when it was distanced up to 18 nucleotides from the 5′ss may suggest that in fact this regulation did not act on the 5′ss but on the 3′ss , since human introns can be as short as 25 nt . In such a scenario , PUF60 and U2AF65 could be involved as factors regulating 3′ss selection . However , we consider this scenario unlikely since the effect we observed increased once the PPT was brought into closer proximity with the 5′ss . Second , it was previously demonstrated that U2AF65 also plays an enhancing regulatory role when binding downstream of the 5′ss [71] . An additional recently discovered protein which might potentially play a role is nSR100 , which was shown to bind pyrimidine-rich sequences within alternative exons and in the intronic regions flanking them , and to enhance their recognition [72] . Thus , we cannot rule out that in addition to the TIA proteins , additional factors such as U2AF65 and/or additional factors play a role in Alu exonization . Our bioinformatic analysis provided evidence that our experimental conclusions are applicable to a wide variety of organisms . This analysis showed that the PPT region tends to be located within 20 nt of the beginning of an intron; this is the situation in Alu elements . Moreover , this analysis revealed the presence of an inverse correlation between 5′ss strength and prevalence of PPT tracks within metazoan introns . This anticorrelation may be indicative of the functional role of the interaction between these two signals , consistent with previous findings showing that PPTs downstream of the 5′ss support the inclusion of weakly defined exons [29] , [30] , [32] . It is noteworthy , however , that while our observations establish a correlative relationship between the two signals , it will require experimental analysis in different organisms to establish a cause-effect relationship between the 5′ss and the PPT downstream of it . Our analysis further demonstrated the high extent of conservation of the TIA proteins and their binding sites on pre-mRNA . For most organisms there was a clear PPT peak downstream of splice sites . In all analyzed eukaryotes , excluding the two protozoa , we found at least one TIA homolog . Moreover , the RRM2 domain , which is responsible for binding U-rich sequences , was particularly conserved and most homologs have retained a glutamine-rich C-terminal region . Finally , the N-terminal domain of U1C , which mediates the recruitment of U1 snRNP by the TIA proteins , was highly conserved among eukaryotes . Our analysis did , however , show that this machinery may have undergone modifications over the course of evolution . In S . pombe , for example , there is no clear PPT downstream of splice sites and we found no PUB1 or TIA1 homolog . This might be related to the extremely short intron length in S . pombe , which allows this organism to maintain intron selection without the need for TIA1 or PUB1 proteins . Two additional organisms in which modifications may have occurred are A . thaliana , for which no TIA1/TIAR homolog was found , and C . elegans , in which no anticorrelation with 5′ss strength existed . One possibility is that in these organisms additional factors compensate for the loss of the factor , or of the signal . Indeed , in plants two related proteins UBP1 and RBP45 , can interact with intronic U-rich elements and enhance the recognition of suboptimal splice sites [42]–[44] . This could explain why we still observe a PPT/5′ss anti-correlation in A . thaliana . Alternatively , the role of the PPT downstream of the 5′ss , and perhaps also the relationship between the PPT and the 5′ss , may have changed over time . In C . elegans , for example , the binding of TIA1/TIAR proteins to the PPT downstream of the 5′ss may occur regardless of the strength of the latter . Alternatively , the PPT downstream of the 5′ss may be an evolutionary ‘fossil’ which has lost its function in C . elegans . As in S . pombe , such loss of function may be a function of intron length as C . elegans introns are considerably shorter than those in other analyzed metazoans [54] . Such loss of function may also be linked with dramatic differences in C . elegans splicing compared to other organisms tested in this study , as attested , for example , by the high prevalence of trans-splicing in this organism [73] . A further intriguing result is the balance of power we observe between different splicing signals . Despite the presence of an intervening PPT , a weaker , upstream splice site is only selected as long as the stronger , competing splice site is weaker than a set threshold . Once this threshold is exceeded , the stronger splice site is selected even in the presence of a PPT . The strength of the PPT is yet another factor as also shown by [32] . In this context , we found that in a dataset of 2 , 296 alternative 5′ss events , in 25 . 4% and 33 . 9% of the cases there is a PPT within 25 nt downstream of the proximal and distal 5′ss , respectively . These cases are potential candidates for TIA regulation . Taken together , our findings demonstrate the role of TIA proteins in the specific context of Alu exonizations and also in the much wider context of exon selection in organisms from throughout the evolutionary tree . | Human genes are composed of functional regions , termed exons , separated by non-functional regions , termed introns . Intronic sequences may gradually accumulate mutations and subsequently become recognized by the splicing machinery as exons , a process termed exonization . Alu elements are prone to undergo exonization: more than 5% of alternatively spliced internal exons in the human genome originate from Alu elements . A typical Alu element is ∼300 nucleotides long , consisting of two arms separated by a polypyrimdine tract ( PPT ) . Interestingly , in most cases , exonization occurs almost exclusively within either the right arm or the left , not both . Here we found that the PPT between the two arms serves as a binding site for TIA proteins and prevents the exon selection process from expanding into downstream regions . To obtain a wider overview of TIA function , we performed a cross-evolutionary analysis within 22 eukaryotes of this protein and of U1C , a protein known to interact with it , and found that functional regions of both these proteins were highly conserved . These findings highlight the pivotal role of TIA proteins in 5′ splice-site selection of Alu exons and exon recognition in general . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"computational",
"biology/alternative",
"splicing",
"molecular",
"biology/rna",
"splicing",
"computational",
"biology/genomics",
"computational",
"biology/protein",
"homology",
"detection"
] | 2009 | The Pivotal Roles of TIA Proteins in 5′ Splice-Site Selection of Alu Exons and Across Evolution |
Members of the highly conserved class of BEACH domain containing proteins ( BDCPs ) have been established as broad facilitators of protein–protein interactions and membrane dynamics in the context of human diseases like albinism , bleeding diathesis , impaired cellular immunity , cancer predisposition , and neurological dysfunctions . Also , the Arabidopsis thaliana BDCP SPIRRIG ( SPI ) is important for membrane integrity , as spi mutants exhibit split vacuoles . In this work , we report a novel molecular function of the BDCP SPI in ribonucleoprotein particle formation . We show that SPI interacts with the P-body core component DECAPPING PROTEIN 1 ( DCP1 ) , associates to mRNA processing bodies ( P-bodies ) , and regulates their assembly upon salt stress . The finding that spi mutants exhibit salt hypersensitivity suggests that the local function of SPI at P-bodies is of biological relevance . Transcriptome-wide analysis revealed qualitative differences in the salt stress-regulated transcriptional response of Col-0 and spi . We show that SPI regulates the salt stress-dependent post-transcriptional stabilization , cytoplasmic agglomeration , and localization to P-bodies of a subset of salt stress-regulated mRNAs . Finally , we show that the PH-BEACH domains of SPI and its human homolog FAN ( Factor Associated with Neutral sphingomyelinase activation ) interact with DCP1 isoforms from plants , mammals , and yeast , suggesting the evolutionary conservation of an association of BDCPs and P-bodies .
BEACH ( beige and Chediak Higashi ) domain containing proteins ( BDCPs ) represent a highly conserved protein family in eukaryotes [1 , 2] . Initially , the BEACH domain was described as a protein motif in the human lysosomal trafficking regulator protein ( LYST ) . Mutations in LYST cause the autosomal recessive human Chediak Higashi Syndrome ( CHS ) [3] . Its corresponding mouse model known as beige was characterized in parallel [4 , 5] . Individuals concerned suffer from severe morphological symptoms , like decreased pigmentation , bleeding diathesis , impaired cellular immunity [6] , cancer growth [7] , and neurological dysfunctions [8] . To date , all experimental data point to a role of BDCPs in the regulation of membrane dynamics . Mutations in BDCPs have been shown to impair diverse cellular mechanisms , including vesicle transport , membrane fission and fusion events , receptor signaling , autophagy , and apoptosis [9] . In plants , only the BDCP encoding gene SPIRRIG ( SPI ) has been characterized in more detail in the plant model Arabidopsis thaliana [2] . Like most BDCPs , SPIRRIG encodes a large protein of 3571aa . Its C-terminally located BEACH domain is preceded by a Pleckstrin-Homology ( PH ) domain , which is followed by five WD40 repeats . The structural organization of these three domains is highly conserved across all known BDCPs and might function as an independent cassette within these proteins [1 , 2 , 10] . Therefore , this region is referred to as the PH-BEACH-WD40 ( PBW ) module [11] . In the N-terminus , SPI contains multiple ARMADILLO repeats and a Concanavalin A ( ConA ) -like lectin domain , similarly as the human LYST protein [12] . As mutations in SPI cause defects in epidermal cell expansion and vacuolar integrity , plant BDCPs are also involved in membrane-dependent cellular processes [2] . Our molecular analysis of SPI sheds light on an unexpected , membrane-independent function for a BDCP . We identified the DECAPPING PROTEIN1 ( DCP1 ) as a direct interactor of SPI . DCP1 is known to activate the pyrophosphatase activity of the DECAPPING PROTEIN2 ( DCP2 ) , by forming a complex together with VARICOSE ( VCS ) and the DECAPPING PROTEIN5 ( DCP5 ) [13 , 14] . Their combined action leads to a 5´monophosphorylated mRNA body , which in turn is accessible for its final decay by the EXORIBONUCLEASE4 ( XRN4 ) [15–21] . Frequently , several of those decapping complexes accumulate with each other . These membrane-unbound ribonucleoprotein ( RNP ) particles are present in the cytoplasm of all eukaryotic organisms and harbor not only the 5’ to 3’ mRNA decay machinery but also translationally repressed mRNAs , as well as components responsible for mRNA quality control and miRNA-dependent gene silencing . Following the principles of phase transitions , the individual protein constituents and mRNA substrates locally concentrate and form a specific type of microscopically visible RNA granule , termed mRNA processing body ( P-body ) [22–24] . A general prediction is that P-body formation is directly proportional to the pool of cytoplasmic , translationally repressed mRNAs [25] . However , the molecular mechanisms required for transition of mRNAs from sites of active translation to P-bodies is currently poorly understood . In this study , we report that the BDCP SPI is a regulator of RNP particle formation in the context of post-transcriptional salt stress response . We show that SPI interacts with DCP1 , and that it localizes to P-bodies under salt stress conditions . In addition , salt stress-dependent P-body assembly is impaired in spi mutants . Genome-wide transcript analysis by RNA-seq indicated that RNA abundance is pleiotropically altered in spi mutants under control and salt stress conditions . Under salt stress , one-third of the transcriptional response is altered in the mutant as compared to wild type . We tested whether the loss of P-body formation in spi mutants is indicative for the post-transcriptional regulation of salt stress-regulated transcripts . We found that SPI is required for the stabilization of a subset of salt stress-regulated mRNAs and their recruitment to P-bodies . As spi mutants display salt hypersensitivity , it is conceivable that the salt stress-dependent regulative function of SPI at P-bodies is biologically relevant . Finally , we show that an interaction between BDCPs and DCP1 is evolutionarily conserved , suggesting that the role in RNP particle formation is a general feature of eukaryotic BDCPs .
Consistent with a role for BDCPs in membrane trafficking and dynamics , several studies identified membrane-associated proteins as binding partners of BDCPs [26–30] . To identify interactors of plant BDCPs , we performed yeast two-hybrid cDNA library screens using the C-terminal fragment of SPI containing its PBW domain module ( referred to as SPI-PBW hereafter , Fig 1A ) as bait . Surprisingly , we identified the evolutionarily conserved P-body core component DCP1 as an interaction partner ( Fig 1B ) . All other tested decapping complex proteins including DCP2 , DCP5 , or VCS did not show interactions with SPI in yeast two-hybrid assays . The interaction of SPI-PBW and DCP1 was confirmed in pull-down experiments with bacterially expressed proteins . Gluthatione S-Transferase ( GST ) /His6-fusions of SPI-PBW were efficiently bound to resins labeled with Maltose Binding Protein ( MBP ) -tagged DCP1 , while no binding was detected with the negative control MBP alone ( Fig 1C ) . To analyze the interaction between full-length SPI and DCP1 in Arabidopsis leaf epidermis cells , we performed Förster-Resonance Energy Transfer ( FRET ) -Acceptor Photobleaching ( AP ) experiments ( Fig 1D and 1E ) . We expressed YFP-tagged full-length genomic SPI ( acceptor ) and DCP1-CFP ( cyan fluorescent protein; donor ) under the 35S promoter and measured their FRET efficiencies ( FRETE ) . We measured FRETE of about 27% in whole cells , indicating that the interaction between DCP1 and SPI occurs in vivo . To test whether the interaction takes place at P-bodies , we analyzed the fraction of immobile P-bodies [31] . Here , the FRETE was 23% ( Fig 1E ) , indicating that SPI and DCP1 interact at P-bodies . No significant FRET was detected between DCP1-CFP and free YFP ( yellow fluorescent protein ) as a negative control . Donor emissions of cells transfected with DCP1-CFP alone were used as a photobleaching corrective ( Fig 1D ) . The intracellular localization of the SPI-PBW/DCP1 interaction was independently analyzed by Bimolecular Fluorescence Complementation ( BiFC ) assays in transiently transformed Nicotiana benthamiana leaf epidermis cells . Like shown for full-length SPI and DCP1 in FRET-AP assays , we found SPI-PBW and DCP1 interacting at distinct cytoplasmic dot-like structures . These completely colocalized with DCP2-mCHERRY ( mCHERRY is a monomeric mutant of DsRED ) ( Fig 2A ) . To exclude that the presence of another P-body component influences the interaction behavior of SPI-PBW and DCP1 , we confirmed our observations in BiFC assays coexpressing free RFP ( red fluorescent protein ) instead of DCP2-mCHERRY ( Fig 2B ) . As negative controls , we coexpressed YFPC ( C-terminal half of YFP ) -DCP1 and YFPN ( N-terminal half of YFP ) -SPI-PBW with YFPN C-terminally fused to VPS20 . 2 ( Vacuolar Sorting Protein 20 . 2 ) and YFPC N-terminally fused to AtMYC1 ( MYC related protein 1 ) , respectively ( Fig 2C–2F; S1A and S1B Fig ) . We did not observe any YFP fluorescence , confirming the specificity of our BiFC analysis ( S1 and S2 Tables ) . The integrity of VPS20 . 2-YFPN was confirmed in cells cotransfected with its known interactor VPS25 ( Vacuolar Sorting Protein 25 ) , C-terminally fused to YFPC ( S1C Fig ) [32 , 33] . The integrity of AtMYC-YFPC was confirmed by showing BiFC interaction with GL1 ( GLABRA1 ) , C-terminally fused to YFPN ( S1D Fig ) [34] . Taken together , these data show that Arabidopsis SPI interacts with DCP1 and that this interaction occurs at P-bodies . The finding that SPI-PBW interacts with DCP1 at P-bodies raised the question of where the SPI protein is localized . In contrast to our expectations from FRET and BiFC assays , we found 35S promoter-driven N-terminal YFP fusions with full-size genomic SPI ( 35Spro:gSPI ) as well as SPI-PBW evenly distributed in the cytoplasm ( Fig 3A ) . In less than 10% of cells analyzed , YFP-gSPI accumulated in cytoplasmic dot-like structures ( S1E Fig ) . We did not notice a correlation between the expression strength of YFP-gSPI and its localization behavior . In cells cotransfected with fluorescently tagged DCP1 and YFP-gSPI , YFP-gSPI was efficiently relocalized to P-bodies ( 84 . 8% , S1E and S1F Fig ) . This relocalization behavior suggests that DCP1 can recruit SPI to P-bodies . However , the functional role of DCP1 in the recruitment of SPI to P-bodies could not be analyzed , as dcp1 mutants are embryonic lethal [35] . As not only the accumulation frequency of the decapping complex but also the localization of various mRNA and protein constituents to P-bodies is highly stress-regulated [36–39] , we reasoned that the association of SPI with P-bodies might be triggered by stress . Different abiotic stress conditions , including salt stress or hypoxia , are known to induce repression of de novo protein synthesis in Arabidopsis [40 , 41] . In turn , polysomes disassemble and release their transcripts , which are trapped by RNA binding proteins that frequently aggregate into cellular RNP particles such as P-bodies [42] . Consequently , P-bodies increase in number and size [24 , 38 , 43–45] . As P-body number increases under salt stress conditions in A . thaliana ( Table 1 ) , we tested whether salt treatments can induce the recruitment of SPI to P-bodies . In transiently transfected leaves , incubated on ½ Murashige and Skoog ( MS ) control medium or ½MS medium supplemented with different NaCl concentrations ( S1G Fig ) , we observed a salt-dependent accumulation of YFP-gSPI in cytoplasmic dots that colocalized with the P-body marker DCP2-mCHERRY ( Fig 3B ) . In contrast , Mannitol treatments had no effect on the localization of YFP-gSPI indicating that the relocalization of SPI is not triggered by osmotic stress in general ( S1H Fig ) . The salt stress-dependent localization of SPI to P-bodies is likely mediated by its PBW module , as the corresponding fragment alone was sufficient for the localization to P-bodies ( Fig 3A and 3B , second rows ) . Free YFP was never observed to accumulate in dots ( S1I Fig ) . In summary , these data show that the localization of SPI to P-bodies is triggered by salt stress but not by osmotic stress in general . P-body formation is determined by the relative entry and exit rates of mRNAs and mRNA binding proteins . The efficiency of material uptake and release from P-bodies can be monitored after blocking translation elongation with Cycloheximide ( CHX ) , which causes the trapping of mRNAs in polysomes . As a consequence , the pool of translationally repressed mRNAs shrinks and the influx into P-bodies decreases , resulting in a reduced number of P-bodies [14 , 24 , 44–46] . In stably transformed 35S:DCP1-YFP Col-0 plants , the number of P-bodies was reduced by 50% , 45 min after CHX treatment ( Fig 4A and 4B ) . The reduction of P-body numbers was not significantly different in two different spi mutant alleles ( S2A and S2B Fig ) , indicating that uptake and release from P-bodies are not generally affected in spi mutants under nonstress conditions . Mock treated cells did not show any significant changes in P-body numbers ( S2C Fig ) . We confirmed the response of P-bodies to CHX treatments in leaf epidermis cells of Col-0 , spi-2 , and spi-4 transiently transfected with another P-body marker , 35Spro:DCP2-mCHERRY ( S2D Fig ) . To test whether the DCP1-dependent recruitment of SPI requires the bulk RNA flow from polysomes to P-bodies , we exposed transiently transfected cells to 0 . 5 mM CHX for 150 min . Under these conditions , YFP-gSPI still accumulated in DCP1-mCHERRY labeled P-bodies ( S1G and S2E Figs ) . In a second step , we examined P-body assembly under salt stress conditions in two spi alleles . In contrast to wild-type , both spi mutants showed no significant changes in P-body number ( Table 1 ) . In samples treated simultaneously with CHX ( 0 . 5 mM ) and NaCl ( 140 mM ) for 90 min , the number of P-bodies decreased significantly by 20% in both Col-0 and spi mutants ( S2F Fig ) . This indicates that in Col-0 , the salt stress-induced increase of P-body number depends on the RNA flow from polysomes to P-bodies . In addition , the data show that in spi mutants P-body maintenance depends on polysome-P-body RNA shuttling also under salt stress conditions . Taken together , our data suggest that SPI is functionally important for the accumulation of mRNA-protein complexes in P-bodies under salt stress conditions . The salt stress-dependent function of SPI at P-bodies suggested that SPI might be relevant for the salt stress tolerance of Arabidopsis . We first compared root growth efficiencies between Col-0 and three spi mutant alleles at different NaCl concentrations under nontranspiring conditions . The relative growth of primary roots did not differ between wild-type and spi under nonstress conditions . With increasing NaCl concentrations , we observed a stronger inhibition of primary root growth in spi than in Col-0 ( Fig 5A ) . The stress hypersensitive phenotype of spi mutants is salt specific , as the relative root growth of spi and Col-0 did not significantly differ after Mannitol treatments ( Fig 5B ) . The notion that spi mutants are salt hypersensitive was supported by cotyledon greening assays . While cotyledon greening of spi mutants was undistinguishable from Col-0 plants under nonstress conditions , a clear whitening of more than 50% of spi seedlings was observed on MS medium supplemented with 150 mM NaCl after an incubation time of 14 d . More than 90% of Col-0 plants remained unaffected under these conditions ( Fig 5C ) . Next , we assessed the salt sensitivity of more adult plants under transpiring conditions in NaCl irrigation experiments ( Fig 5D and 5E ) . When treated with 50 mM or 100 mM NaCl , growth of spi mutants was much more restricted than growth of wild type plants . In summary , these data show that SPI is required for Arabidopsis salt stress but not for osmotic stress tolerance in general . To test whether the hypersensitive response of spi mutants is accompanied by changes in the transcript levels , we performed a genome-wide analysis of the transcriptome by RNAseq of wild type and spi mutants under control and salt stress conditions . 10-d-old seedlings were incubated in liquid ½MS control medium or ½MS medium supplemented with 200 mM NaCl for 4 h . Ten libraries were Illumina-sequenced , mapped , and the results were visualized ( for additional information see S1 Text , S3–S8 Figs , and S3 Table; complete raw data are available under http://www . ncbi . nlm . nih . gov/bioproject/278120 ) . Between spi mutant and wild type , 483 ( control condition ) and 474 transcripts ( salt-treated condition ) were significantly different between genotypes ( q < 0 . 01 , Benjamini-Hochberg ( BH ) corrected ) . The salt treatment significantly altered the abundance of 8 , 469 transcripts in Col-0 and 8 , 482 transcripts in the spi mutant ( Table 2 ) . In Col-0 , the salt treatment altered transcriptional abundance in stress-related gene categories ( Fig 6A and 6B ) . At the same time , transcript abundance in categories related to growth was changed ( S5 and S6 Figs ) . This transcript abundance pattern reflected the cross activation of different stress pathways , the preparation for reduced nutrient uptake with regard to nitrate and iron , and , likely as a consequence , a down-regulation of photosynthesis ( Fig 6A ) . In spi , the salt treatment altered transcriptional abundance in similar gene categories , including the main abiotic responses , biotic responses , and metabolic Gene Ontology ( GO ) term categories ( Fig 6C and 6D , S7–S9 Figs ) . The gene-by-gene comparison of the transcriptional abundance changes in Col-0 , and spi showed that one-fourth of the responses are specific to each genotype and three-fourths are shared ( Fig 6E ) . In the shared response , the strength of regulation was similar between Col-0 and spi , indicating that the response to salt stress is not attenuated but qualitatively changed ( S7–S9 Figs ) . As the biological relevance for the salt stress response has not been validated for most of the differentially regulated genes yet , we directly compared the expression levels of a subset of genes , functionally and/or genetically shown to regulate Arabidopsis salt stress response [47–53] . While most of the candidate genes tested were significantly up-regulated under salt stress conditions in Col-0 as well as in spi mutants , a subset of genes was either not up-regulated or significantly less up-regulated in spi mutants ( S4 Table ) . Our RNA-Seq analysis demonstrated that SPI affects transcripts abundance pleiotropically in a global manner rather than specifically salt stress-regulated mRNAs . Among the latter , only a subset of transcripts was identified to be differentially regulated by SPI . Since the steady-state transcript levels provide only a snapshot of the balance between transcription and mRNA decay , we determined the stabilities of SPI-dependent ( TANDEM ZINC FINGER PROTEIN 3 ( TZF3 ) and ABA INSENSITIVE 1 ( ABI1 ) ) and SPI-independent regulated mRNAs ( RESPONSIVE TO DESSICATION 29B ( RD28B ) and CBL-INTERACTING PROTEIN KINASE 9 ( CIPK9 ) ) ( S4 Table ) . Towards this end , we measured transcript levels by qPCR under nonstress and salt stress conditions 3 h and 6 h after inhibition of transcription with Actinomycin D ( ActD ) [54 , 55] . Under nonstress conditions , the mRNA decay rates of RD29B and TZF3 were similar in wild type and spi mutants ( Fig 7A and 7B ) , while ABI1 and CIPK9 transcripts were significantly stabilized in all three spi alleles ( Fig 7C and 7D ) . The incubation of seedlings in ½MS liquid medium supplemented with 200 mM NaCl for 4 h resulted in a stabilization of RD29B in wild type . In contrast , RD29B mRNA decay was enhanced in spi mutants under these conditions ( compare Figs 7A and 8A ) . Similarly , TZF3 was slightly stabilized under salt stress conditions in wild type ( compare Figs 7B and 8B ) , whereas it was significantly destabilized in the absence of SPI ( Fig 8B ) . CIPK9 and ABI1 mRNA stability was not significantly different in wild type and spi mutants ( Fig 8C and 8D ) . Given that in spi mutants some mRNAs are stabilized under control conditions ( CIPK9 , ABI1 ) whereas others are destabilized under salt stress conditions ( RD29B , TZF3 ) , it is conceivable that SPI does not regulate RNA stability directly but rather the uptake into P-bodies where each mRNA experiences its specific fate . The salt stress-dependent recruitment of SPI to P-bodies suggested to us that SPI might regulate the localization of mRNAs . As salt-regulated transcripts were selectively destabilized under salt stress conditions in spi mutants , we tested whether their spatial and temporal distributions were also impaired . We monitored the localization of two different mRNA targets of SPI in vivo using the LambdaN22 reporter system [56] . We placed 16 BoxB repeats N-terminally to the 5’ UTR of full-size genomic TZF3 ( gTZF3 ) and RD29B ( gRD29B ) , and coexpressed them with the LambdaN22 protein fused to mVENUS in Arabidopsis leaf epidermis cells . Binding of LambdaN22-mVENUS to the BoxB repeats enables the indirect visualization of the RNAs of interest . Three different negative control experiments were performed: first , the LambdaN22-mVENUS reporter was coexpressed with the BoxB repeats without the mRNA target ( S10A Fig ) . Second , the full-size genomic target constructs without the stem loops but with an N-terminally fused mCHERRY were coexpressed with the reporter construct ( S10B and S10C Fig ) . Third , the LambdaN22-mVENUS reporter was coexpressed with full-size genomic ABF3 ( gABF3 ) , a transcript unaffected in its stability in spi mutants under all conditions investigated ( S10D–S10F Fig ) . Under all three control conditions , the LambdaN22-mVENUS reporter was detected almost exclusively in the nucleus or in both the nucleus and the cytoplasm . In cells coexpressing 16BoxB-gTZF3 or 16BoxB-gRD29 , we observed accumulations of the reporter constructs in cytoplasmic dot-like structures , indicating that these accumulations are caused by the presence of the 16BoxB-fused target mRNAs ( S11A and S11B Fig ) . Colocalization studies with DCP1 , C-terminally fused to mCHERRY , revealed that several mRNA positive accumulations overlap with P-bodies ( Fig 9 ) . As P-body and mRNA granule formation follow the principles of classical liquid–liquid phase separations , comprising multivalent and low affinity interactions between proteins constituents and mRNA substrates involved [57–59] , their assembly is highly dynamic . As a consequence , the number of mRNA granules and their size is highly variable from cell to cell [39 , 46 , 60 , 61] . This was also observed in our study; however , a quantitative analysis revealed clear differences between nontreated and salt stress-treated samples . In transfected Col-0 cells , 33% of RD29B and 52% of TZF3 positive dots colocalized with P-bodies under nonstress conditions . Their amount in cells treated with 140mM NaCl increased to 57% and 81% , respectively ( Table 3 ) . This salt stress-dependent recruitment of RD29B and TZF3 to P-bodies was not observed after CHX treatments suggesting that these two mRNAs are delivered from the polysomes to the P-bodies ( S11C and S11D Fig ) . Additionally , the total amount of RD29B positive mRNA dots increased drastically in comparison to untreated cells under salt stress conditions ( Table 4 ) . The average number of TZF3 mRNA dots per cell was unchanged under salt stress conditions in comparison to nonstress conditions ( Table 4 ) . In two spi alleles , we observed an impaired granule formation behavior for the transcripts of TZF3 and RD29B in two respects . First , no salt stress-dependent relocalization of TZF3 and RD29B mRNA dots to P-bodies took place ( Table 3 ) . Second , the total number of RD29B mRNA dots was not increased like in Col-0 cells ( Table 4 ) . Our RNA localization data show that SPI functions as a positive regulator of post-transcriptional RNP particle formation under salt stress conditions . BDCPs and the Decapping machinery are evolutionarily highly conserved and are present in a wide range of eukaryotic organisms . In inter- and intraspecific interaction studies , we also investigated whether an association between BDCPs and P-bodies is evolutionarily conserved . In yeast two-hybrid assays and coprecipitation experiments , we observed the interspecific interactions between the Arabidopsis SPI-PBW and both mammalian DCP1 isoforms ( DCP1a and DCP1b ) as well as its yeast counterpart ( DCP1p ) ( Fig 10A and 10B ) . The interaction to the yeast DCP1p was especially surprising , as its protein structure differs strongly from those of mammalian and plant homologs ( S12 Fig ) . In contrast to human and Arabidopsis DCP1 proteins , yeast DCP1p lacks an extended C-terminal domain . In the remaining N-terminal region , we found only one domain , the EVH1 ( Enabled/VASP Homology 1 Domain ) domain that is conserved in all four DCP1 proteins suggesting that this domain mediates the DCP1 BDCP interactions . To substantiate our hypothesis of an evolutionarily conserved association of BDCPs and the Decapping machinery , we studied the interactions of the PH-BEACH domain containing fragment of the human FAN ( Factor Associated with Neutral sphingomyelinase activation ) protein with the corresponding human DCP1 isoforms and the Arabidopsis DCP1 homolog . Interactions were found in yeast two-hybrid assays and coprecipitation experiments ( Fig 11A and 11B ) . As we used a C-terminal truncation of FAN lacking its WD40 repeats , we exclude that this protein region is required for an interaction to DCP1 . Furthermore , sequence alignments revealed that the eponymous BEACH domain exhibits the highest level of sequence conservation in the C-terminal fragments of SPI and FAN ( S13 Fig ) . All together our data on intra- and interspecific interactions between Arabidopsis SPI , human FAN , and the DCP1 homologs from Arabidopsis , human , and yeast demonstrate that the association of BDCPs and P-bodies is not plant specific but rather evolutionarily conserved .
Col-0 and spi react to salt stress with transcript abundance changes related to specific and general stress responses . In comparison to Col-0 , spi mutants display changes in pathways , including carbohydrate- and polysaccharide-dependent biosynthetic processes , transportations of and response to anorganic and organic substances , immune responses and photosynthesis , pointing to a pleiotropic function of SPI ( S3B and S4 Figs ) . However , we observed similar transcript abundance changes in pathways regulating the responses to wounding , hormones , pathogens and chitin , callose deposition , circadian regulation , metabolic processes , and polysaccharide localization , as well as salinity and water deprivation ( Fig 3 and S5–S8 Figs ) . The comparison of transcriptional changes in a selection of functionally and/or genetically characterized regulators of Arabidopsis salt stress response revealed that most are not affected in spi mutants and that few show a lower expression ( e . g . ABI1 ) or a reduced salt stress response ( TZF3 ) ( S4 Table ) . Thus , SPI regulates salt stress-dependent transcriptional regulation in a differential manner . The salt stress-dependent relocalization of SPI to P-bodies and its requirement for their assembly suggest a functional role of SPI in post-transcriptional mRNA regulation . The assembly of microscopically visible RNP particles depends on three processes . First , the amount of cytoplasmic-available mRNA is tightly regulated . Here , translational repression is a prerequisite for all the following steps . This is exemplified by yeast Dhh1p and Pat1p that stimulate P-body formation by translational inhibition [77] . Second , the assembly of RNP particles takes place . This is best shown for the yeast Edc3p and Lsm4p [78 , 79] . Edc3p and Lsm4p are important scaffold proteins for the assembly of RNP particles and the corresponding mutants display reduced amount of P-bodies . Third , the activity of the mRNA Decapping machinery controls P-body number . Mutations in the decapping enzyme DCP2 or decapping enhancers lead to hyperassembly and enhanced formation of P-bodies [45 , 80–90] . At what step does SPI contribute to the regulation of P-body formation ? It is unlikely that SPI contributes to P-body formation at the level of translational regulation , as P-body disassembly is similar in wild-type plants and spi mutants after CHX treatment ( Fig 4A and 4B ) . In contrast , a role for SPI in the second step , the assembly of P-bodies , is supported by several observations: ( 1 ) SPI is recruited to P-bodies in response to salt treatments ( Fig 3A and 3B; S1G Fig ) ; ( 2 ) P-body formation was greatly impaired after salt treatments in spi mutants but increased in wild-type plants ( Table 1 ) ; ( 3 ) the absence of relocalization of TZF3 and RD29B mRNAs to P-bodies under salt stress conditions in spi mutants suggests that SPI is important for their recruitment into P-bodies ( Table 3 ) . Furthermore , these observations shed new light on the function of P-bodies , which have been mainly described as cytoplasmic spots responsible for transcript decay so far . However , our data suggest that mRNAs can also be stabilized in P-bodies and thereby support recent studies performed in yeast [91 , 92]; and ( 4 ) in contrast to wild-type plants , the number of RD29B mRNA positive granules was not increased upon salt treatments in spi mutants , indicating that SPI is required for the assembly of RD29B mRNA containing RNP particles ( Table 4 ) . In interspecific interaction assays , we observed Arabidopsis SPI to physically interact with the core P-body component DCP1 from human and yeast via its structurally conserved PBW module ( Figs 1 , 10A and 10B; S12 Fig ) . Therefore , the functions of BDCPs in RNP particle formation and post-transcriptional gene regulation seem to be evolutionarily conserved . This conclusion is supported by our finding that the PH-BEACH domain of human FAN interacts with both isoforms of human DCP1 as well as with the Arabidopsis homolog ( Fig 11A and 11B ) . This conclusion creates a new perspective in understanding the molecular reasons for altered gene expression patterns observed in BDCP mutant mice . For heterozygous NBEA ( Neurobeachin ) mutant mice , the misregulation of specific hypothalamic genes was reported in response to calorie deprivation [93] . In TNF ( Tumor Necrosis Factor ) -stimulated FAN-deficient mice , expression of inflammatory genes was selectively impaired [94] . However , the molecular mechanisms causing altered gene abundance remain unclear . An evolutionarily conserved function of BDCPs in stress-dependent RNP particle formation presents an elegant explanation for these findings that are difficult to explain with the current concept of BDCP functioning only in membrane trafficking .
Arabidopsis thaliana spi-2 ( GK_205G08 ) , spi-3 ( SALK_065311 ) , and spi-4 ( GK_420D09 ) mutants ( Columbia ecotype ) were obtained from the National Arabidopsis Stock Center . Positions of all T-DNA insertions ( S2A Fig ) were confirmed by genotyping and sequencing the flanking genomic regions . Loss of expression of full-length transcripts in the T-DNA insertion mutants was confirmed by qualitative RT-PCR spanning the insertion site ( S2B Fig ) . 35S:DCP1-YFP was stably transformed in Col-0 , using the Agrobacterium tumefaciens strain GV3101 pMP90RK as described previously [95] . Homozygous DCP1-YFP lines in spi-2 and spi-4 backgrounds were obtained by crosses . Seeds were surface-sterilized and grown on ½MS medium or on soil under long day conditions at 22°C and 110 μmol m-2 s-1 light intensity . For salt treatments under transpiring conditions , 18-day-old seedlings ( grown on ½MS medium supplemented with 1% sucrose ) were transferred to pots containing a sand–soil mixture ( 9:1; v/v ) . For the first week plants were watered with ½MS medium only . 24-d-old plants were irrigated daily with ½MS medium ( as a control ) or ½MS supplemented with NaCl on every second day . The initial NaCl concentration of 50 mM NaCl was either kept or stepped up to 100 mM NaCl after four days of watering , as indicated . Diameter of rosette leaves were taken as indicators for salt tolerance [96] and measured using ImageJ . For root growth inhibition assays under nontranspiring conditions , seeds were transferred to ½MS medium supplemented with different concentrations of NaCl and grown vertically orientated [97] . Root growth rates of salt-treated seedlings were measured 10 d post-transfer and normalized with those of untreated plants . For salt treatments of single transfected cells , rosette leafs of 14- to 16-d-old plants were transferred to culture medium containing different NaCl concentrations for the indicated durations . For blocking translation elongation , rosette leaves were incubated in culture medium containing 0 . 5 mM CHX ( 100 mM Stock in 100% DMSO ) for the indicated durations [13] . For transcript measurements , 8-d-old vertically grown seedlings were transferred from solid to liquid ½MS ( nonstress ) medium as well as liquid ½MS medium supplemented with 200 mM NaCl for 4 h . Afterwards , Actinomycin D ( ActD ) was added to a final concentration of 200 μmol [54] . Plants were kept in the light during incubation . Samples were taken prior to ( time point 0 ) as well as 3 h and 6 h after transcriptional block . Coding sequences of SPI-PBW ( AT1G03060 ) , DCP1 ( AT1G08370 ) , DCP2 ( AT5G13570 ) , DCP5 ( AT1G26110 ) , and VCS ( AT3G13300 ) were amplified from Col-0 cDNA; full-length genomic TZF3 ( gTZF3;AT4G29190 ) , RD29B ( gRD29B; AT5G52300 ) and ABF3 ( gABF3; AT4G34000 ) were amplified from Col-0 genomic DNA; full-length genomic SPI ( gSPI ) was cloned by homologous recombination in E . coli SW102 [98] . In an overlapping PCR , 0 . 6 kbp of the 5´ and 3´ end of SPI ORF were amplified from genomic Arabidopsis Col-0 DNA-adding attB sites and cloned into pDONR207 . The resulting plasmid was linearised with PdmI and XbaI and transformed in E . coli SW102 containing BAC F10O3 that harbors the complete SPI genomic region . Homologous recombination was induced and gentamycin-positive colonies were analyzed in detail . A clone with the expected restriction pattern was verified by sequencing and used in this study . DCP1p ( Q12517 ) , DCP1a ( Q9NPI6 ) , and DCP1b ( Q8IZD4 ) were described previously [31] . The PH-BEACH domain-comprising fragment of human FAN ( FAN-PB ) was amplified from pEGFP-C3:FAN-ΔWD [99] . Primer sequences are provided in S5 Table . Expression vectors containing AtMYC1 and GL1 , VPS20 . 2 and VPS25 , were published previously [33 , 34] . All constructs used in this study were confirmed by sequencing . The following GATEWAY vectors were used for protein expression driven by the CaMV 35S promoter in planta: pENSG or pEXSG-YFP [100] pAMARENA or pAUBERGINE for N- and C-terminal fusions with mCHERRY ( M . Jakoby , GenBank ID: FR695418 ) , pCL112 or pCL113 ( donated by J . F . Uhrig , S14 Fig , S2 and S3 Texts ) for N-terminal fusion with YFPN or YFPC , pSCJ232 for N-terminal fusions of 16 BoxB repeats [56]; for protein expression in yeast: pAS or pACT ( Clontech ) ; for protein expression in E . coli: pGEX-2TM-GW ( kindly received from Imre Sommsich and Bekir Ülker ) for creating fusion proteins with an N-terminal GST- and a C-terminal His6-tag , pDEST17 ( Invitrogen ) , pETG-40A ( EMBL , Heidelberg , Germany ) . Yeast two-hybrid assays were done as described previously [101] . Interactions were analyzed by selection on synthetic dropout interaction media lacking leucine , tryptophan and histidine , supplemented with 3-Amino-1 , 2 , 3-Triazole ( 3AT ) . For coprecipitation studies , MBP tagged fusion proteins and MBP as negative control , as well as GST or GST/His6-tagged proteins were expressed in E . coli ( BL21 ( DE3 ) RIL ) . Bacteria were grown in TB medium ( 37°C , OD600 = 1 ) , induced by isopropyl b-D-1-thiogalactopyranosie ( IPTG ) and incubated ( 20°C , 6 h ) . Cells were harvested and resuspended in TRIS purification buffer ( 100 mM TRIS , 150 mM NaCl , 1% Triton X-100 , pH 8 . 0 ) . After addition of Lysozyme ( 100 μg/ml ) and incubation on ice ( 20 min ) , samples were sonicated . Lysates of MBP fusion proteins were cleared by centrifugation ( 4 , 000 g , 15 min , 4°C ) . Amylose resin was labeled according to the manufactures’ instructions ( New England Biolabs ) . Prior to centrifugation of lysates of SPI-PBW- and FAN-PB-fusion proteins ( 10 , 000 g , 5 min , 4°C ) , N-Lauroylsarcosine sodium salt ( Sigma-Aldrich ) was added ( 1% final concentration ) . Triton X-100 was added to the cleared lysates of SPI-PBW- and FAN-PB-fusion proteins ( 1% final concentration ) . Labeled amylose resins were incubated in cleared lysates of SPI-PBW-or FAN-PB-fusion proteins ( 1 h , 4°C ) under constant shaking . MBP-fusion proteins were purified according to the manufacturer’s instructions ( New England Biolabs ) using TRIS purification buffer for all wash steps . Purifications and coprecipitations were analyzed by immunoblotting as described previously [31] . Nicotiana benthamiana leaves were transiently transformed by infiltration with Agrobacterium tumefaciens ( GV3101 pMP90RK ) . Coinfiltrated cultures were mixed in equal proportions ( 1:1:1 ) and incubated for 4 h at RT [102] . Transfection of Arabidopsis leaves was performed by biolistic transformation [103] and analyzed after 12 to 16 h by Confocal Laser Scanning Microscopy ( CLSM ) . CLSM was done as described previously [31] . Colocalization of dot-like structures was analyzed manually . For BiFC analysis , cells were considered to be transfected when a cotransfected marker protein was expressed . As negative controls , noninteracting proteins that are found in the cytoplasm were included [104] . In three biological replicates ( n = 30 cells ) , no YFP reconstruction was observed . To ensure comparability between transfected cells , laser intensities and exposure times were fixed and no automated corrections used . Transfected leaf epidermis cells were analyzed for FRET between 35Spro:DCP1-CFP ( donor molecule ) and 35Spro:YFP-gSPI ( acceptor molecule ) using a Leica SP8 confocal microscope , equipped with a 20 x 0 . 75 objective ( HC PL AP IMM CORR C52 ) at a 512 x 512 resolution format . Pre and post-AP of the emission spectra of the donor and acceptor were recorded by sequential scanning at 475 nm +/- 10 nm upon excitation at 458 nm and at 540 nm +/- 15 nm upon excitation at 514 nm in z-stacks of whole transfected cells , respectively . The detection of donor and acceptor emissions occurred via high-efficiency hybrid detectors . Laser intensities of 1 . 3% ( DCP1-CFP ) , 10 . 7% ( for YFP-gSPI ) , or 0 . 5% ( for free YFP ) were fixed . Targeted AP was done at 514 nm with 60% laser intensity of a 30% activated Argon laser on a defined region of interest ( ROI ) covering the whole cell and parts of the background by scanning through the complete z-axis of the selected part in 1 μm steps . Photodestruction of the acceptor was 40% on average ( +/- 6% ) . For each cell , maximum projections of z-stacks pre- and post-AP were created and fluorescence intensities of whole cell areas and stationary PBs were measured separately by setting defined ROIs manually using the quantification tool of the LAS AF ( Leica Application Suite Advanced Fluorescence 2 . 4 . 1 ) software . The intensity of an untransfected leaf area ( pre- and post-AP ) was measured and subtracted from the donor intensities as background corrective . Cells transfected with 35Spro:DCP1-CFP only were treated similarly to double transfected cells to determine the photobleaching corrective of the donor molecule in the whole cell ( -6 . 38%; n = 14 ) and in stationary PBs ( -8 . 4%; n = 50 ) . Donor emissions after AP were collectively photobleaching corrected ( Intensity Donor emission post-AP ( corr ) ) [105] . The FRETE was determined according to the calculation: E= ( 1− ( IntensityDonoremissionpriortoAP−BackgroundIntensity ) ( IntensityDonoremissionpostAP ( corr ) −BackgroundIntensity ) ) . Total RNA was isolated ( Qiagen , RNeasy Mini kit ) , treated with DNaseI ( Thermoscientific ) according to the manufacturer´s recommendations , and quantified spectrophotometrically . 1 μg of total RNA was reverse-transcribed ( SuperScriptIII , Invitrogen ) . qPCRs were performed on an Applied Biosystems 7 , 300 real-time PCR system ( http://www . appliedbiosystems . com ) using POWER SYBR Green PCR-Master Mix ( Applied Biosystems ) . Transcript levels were normalized to the 18S rRNA [106] . All qPCR data represent the average of three biological and two technical replicates . For determination of mRNA decay , Ct values after transcription inhibition were normalized to those of 18S rRNA and compared further to their normalized values prior to Actinomycin D treatment . Primer sequences are provided in S6 Table . Isolated RNA was checked for RNA integrity ( RIN > 7 , Agilent 2010 Bioanalyzer [Agilent] ) and prepared for Illumina sequencing using the TruSeq RNA sample prep kit versus Illumina . The resulting libraries were sequenced by Illumina HiSeq according to the manufacturers protocol using the Beckmann Coulter service as 12 samples multiplexed on one lane ( please find the complete data set under http://www . ncbi . nlm . nih . gov/bioproject/278120 ) . The resulting reads were mapped against the Arabidopsis genome with the TAIR10 gff files for annotation using CLC genomics workbench ( Quiagen ) , and total reads per gene were extracted as the measure for gene expression ( S7 Table ) . Differential expression was called using the Bioconductor package edgeR [107] . Read counts were normalized to reads per mapped million , averages were calculated , and log-fold changes were calculated adding one value to avoid division by zero . Gene IDs were functionally annotated using the descriptions from TAIR10 ( www . arabidopsis . org ) and the MapMan annotation ( S8 Table ) [108] . Functional enrichments were calculated based on GO terms using GOrilla [109] using the significantly regulated genes as the target group and all genes tested as the background . p-values are calculated using the hypergeometric distribution and corrected with Benjamini Hochberg for multiple hypotheses testing . Additional functional enrichments were calculated for the metabolism-centric MapMan categories using the embedded Wilcoxon sum rank algorithm with Benjamini Hochberg correction [110] . Using both categorizations for enrichments exploits the strength of the GO annotation ( regulation ) and the strength of the MapMan categorization ( metabolism ) . | BEACH ( beige and Chediak Higashi ) domain containing proteins ( BDCPs ) are a highly conserved protein family in eukaryotes . BDCPs are known to be important for membrane dynamics such as vesicle transport , membrane fission and fusion events , and autophagy . Here we describe a new , membrane-independent molecular function of the Arabidopsis thaliana BDCP SPIRRIG ( SPI ) in the regulation of mRNA stability in P-bodies . We report that SPI associates with P-bodies—cytoplasmic mRNA or protein aggregations responsible for the posttranscriptional regulation of RNA storage and decay—in a salt stress-dependent manner . We show that spi mutants display salt hypersensitivity suggesting that SPI regulates the uptake of salt stress-regulated mRNAs to P-bodies concomitant with their stability . Finally , we found a direct interaction between the P-body core component DECAPPING PROTEIN1 ( DCP1 ) and SPI and show that this interaction is evolutionarily conserved . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The BEACH Domain Protein SPIRRIG Is Essential for Arabidopsis Salt Stress Tolerance and Functions as a Regulator of Transcript Stabilization and Localization |
Disruption of the blood-brain barrier ( BBB ) is a hallmark event in the pathophysiology of bacterial meningitis . Several inflammatory mediators , such as tumor necrosis factor alpha ( TNF-α ) , nitric oxide and matrix metalloproteinases ( MMPs ) , contribute to this disruption . Here we show that infection of human brain microvascular endothelial cells ( HBMEC ) with Neisseria meningitidis induced an increase of permeability at prolonged time of infection . This was paralleled by an increase in MMP-8 activity in supernatants collected from infected cells . A detailed analysis revealed that MMP-8 was involved in the proteolytic cleavage of the tight junction protein occludin , resulting in its disappearance from the cell periphery and cleavage to a lower-sized 50-kDa protein in infected HBMEC . Abrogation of MMP-8 activity by specific inhibitors as well as transfection with MMP-8 siRNA abolished production of the cleavage fragment and occludin remained attached to the cell periphery . In addition , MMP-8 affected cell adherence to the underlying matrix . A similar temporal relationship was observed for MMP activity and cell detachment . Injury of the HBMEC monolayer suggested the requirement of direct cell contact because no detachment was observed when bacteria were placed above a transwell membrane or when bacterial supernatant was directly added to cells . Inhibition of MMP-8 partially prevented detachment of infected HBMEC and restored BBB permeability . Together , we established that MMP-8 activity plays a crucial role in disassembly of cell junction components and cell adhesion during meningococcal infection .
Despite improved antimicrobial therapy , bacterial meningitis is still a cause of high mortality and severe neurological morbidity in childhood [1] . Neisseria meningitidis is one of the most common causes of bacterial meningitis in Northern Europe and the United States [2] , [3] . During meningitis , the host inflammatory response encompasses a variety of detrimental pathophysiological changes , involving increased blood–brain barrier ( BBB ) permeability , increased CSF outflow resistance , brain edema , elevated intracranial pressure , and alterations in cerebral blood flow [4] . These pathophysiological changes lead to long-term neurological deficits in approximately one-third of the patients [5]–[7] . Several mediators have been shown to affect the BBB permeability . These include reactive oxygen species , nitric oxide , peroxynitrite , matrix metalloproteinases ( MMPs ) , tumour necrosis factor-α ( TNFα ) -converting enzyme ( TACE ) , transforming growth factor-β1 ( TGFβ1 ) , arachidonic acid metabolites , proinflammatory neuropeptides and caspases [8]–[12] . Moreover , experimental and clinical studies suggested that cytokines and chemokines also play an important role in the pathophysiology of BBB disruption during bacterial meningitis . However , the mechanism by which the BBB is damaged during bacterial meningitis is still a matter of debate . A role of MMPs in BBB damage in bacterial meningitis has been implicated in several studies [13] , [14] . In particular MMP-8 and MMP-9 are upregulated in CSF of children with bacterial meningitis , levels being 10 to 1000-fold higher than in viral meningitis [15] . The increase of MMP-8 is a specific feature of bacterial meningitis [14] . MMPs are a family of zinc-dependent endopeptidases that catalyze the proteolysis of a broad spectrum of extracellular matrix ( ECM ) and basement membrane proteins [16] . MMPs also cleave a range of other molecules , including cytokines , chemokines and growth factors . Neutrophils , glial cells , vascular smooth muscle cells and endothelial cells can produce MMPs upon stimulation . The ability to disrupt the subendothelial basement membrane in cerebral capillary endothelial cells make MMPs likely candidates as effector molecules of BBB breakdown . Intriguingly , MMPs are also implicated in the regulation of cell survival and death [17] . The adherence of cells to the ECM provides survival signals through mechanisms that include the activation of integrin receptors that have engaged particular ECM proteins . When such anchored cells are detached from the substratum , the loss of integrin signaling can result in apoptosis , a phenomenon named anoikis [18] , [19] . Alterations of the cerebral microvascular endothelium during bacterial meningitis have been intensively studied . Former experimental studies with Escherichia coli , Streptococcus pneumoniae , Haemophilus influenzae revealed that these bacteria induced functional and morphologic alterations of the BBB [20] , which were characterized by an early increase in pinocytotic vesicle formation and a preceding disruption of intercellular tight junctions during time course . On the other hand in vitro BBB models have turned out that a number of meningitis-causing pathogens transmigrate across the BBB without barrier disruption [21]–[27] . N . meningitidis has been found to transmigrate across T84 epithelial cell monolayers without disruption of tight junctions and sustained electrical resistance [28] , [29] . However , only recently type IV pili of N . meningitidis have been shown to induce the formation of ectopic early junction-like domains causing the weakening of endothelial cell-cell tight junctions [30] . In this study , we demonstrate that N . meningitidis infection results in physiological and morphological alterations of brain endothelial cells after prolonged time of infection . Bacteria induced progressive cell detachment from the matrix and disruption of the tight junction protein occludin . We evaluated MMP-8 activity and the influence of MMP-8 activity on both cell junction components and cell adhesion interaction . We also addressed the question of whether the major virulence factor of meningococci , the polysaccharide capsule , had any influence on cell detachment and/or tight junction proteins .
To analyze the physiological properties of human brain microvascular endothelial cells ( HBMEC ) after exposure to N . meningitidis , we first evaluated the integrity of polarized endothelial cells by measuring the transendothelial flux of fluorescently labeled dextrans ( FITC-dextran ) . HBMEC were grown onto matrigel-coated Transwell filters for 5 days to obtain confluence . HBMEC monolayers were treated with a well-charaterized encapsulated serogroup B strain ( MC58 ) [31] , [32] for indicated times and diffusion of FITC-dextrans with molecular masses of 4 and 40 kDa from the apical to the basal compartment was estimated . About 38% of FITC-dextran was transported in a 30 min period from the apical to the basal compartment when matrigel-coated filters without cells were analyzed , indicating that matrigel coating allowed free diffusion of the tracer ( data not shown ) . Changes in the permeability of HBMEC after exposure to N . meningitidis strain MC58 ( 107 bacteria ml−1 ) were seen . As shown in Figure 1 , infection of HBMEC with N . meningitidis increased the transport of FITC-dextrans from the apical chamber to the basolateral chamber in a time-dependent manner . After 24 h of infection , the percentage of dextran flux in uninfected cells was 1 . 73±0 . 53 and 1 . 25±0 . 24 for 4–kDA and 40–kDA FITC-dextran , respectively , whereas N . meningitidis significantly increased the dextran flux to 6 . 62±0 . 25 and 2 . 8±0 . 18 for 4–kDA and 40–kDA FITC-dextran , respectively , at 24 h post-infection ( p . i . ) ( Figure 1 ) . A further increase in FITC-dextran transport across the HBMEC monolayer was seen when 108–1010 bacteria per ml were added ( data not shown ) . Multiple mechanisms may be involved in increased endothelial permeability , such as cell contraction and retraction , enhanced transcellular vesicle transport , disruption of intercellular junctions and cell death . Since the maintenance of the impermeabilty of polarized endothelial cells requires formation of specialized complexes consisting of tight junctions at the apicolateral cell surface [33] , we first visualized the localization and expression of the tight junction proteins occludin , claudin-1 and ZO-1 in HBMEC during N . meningitidis infection . Non-infected HBMEC showed uniform staining of occludin , claudin-1 and the tight junction-associated cytoplasmatic protein ZO-1 ( Figure 2A and B ) . No differences of the localization of occludin were detected for up to 4 h following bacterial exposure . At 8 h p . i . , the distribution of occludin was no longer uniform , and a dotted appearance was observed . When HBMEC were infected with N . meningitidis MC58 for a 24-hour period , a complete dissociation of occludin from the membrane was seen . Cells were more irregular in shape and appeared more rounded in interference contrast images ( data not shown ) . We furthermore observed that numerous cells were detached from glass slides or filters at 24 h p . i . during the experimental process of immunfluorescence microscopy . In contrast to the disruption of occludin localization , ZO-1 and claudin-1 staining was unchanged even in the residual cells at 24 h p . i . ( Figure 2A and B ) . HBMEC cells were furthermore characterized for tight junction expression by immunoblotting . Cell lysates were generated at different time points after infection with N . meningitidis MC58 and analyzed for occludin expression . To further evaluate the role of the major virulence factor of meningococci , the polysaccharide capsule , an isogenic unencapsulated mutant MC58 siaD [34] were included in the experiments . While both isolates efficiently adhere to HBMEC , loss of capsule formation resulted in significant increase of meningococcal uptake , which is due to unmasking of outer membrane proteins such as the Opc protein ( Figure S1 ) . The Opc protein has recently been shown to be mainly involved in internalization into HBMEC [34] . Infection with both meningococcal strains resulted in generating a weak lower-sized 50-kDa occludin fragment in infected HBMEC at 24 h p . i . ( Figure 3A ) . Whole cell lysates were furthermore subjected to immunoprecipitation: after immunoprecipitation with an occludin-specific monoclonal antibody ( clone OC-3F10 ) recognizing the C-terminal part of occludin , the samples were again analyzed by immunoblotting . As shown in Figures 3B and C the 50-kDa cleavage fragment was clearly visible , while the intensity of the 62-kDa band representing the full-length protein decreased . We next assessed the amounts and presence of claudin-1 and ZO-1 in HBMEC infected with N . meningitidis by immunoblot analysis . Immunoblots revealed no changes in the levels of these proteins in infected HBMEC analyzed for a 24-hours time period ( Figure 3A ) . Taken together , these findings indicate that N . meningitidis infection induced selective degradation of occludin , which resulted in its dissociation from the tight junctions . To verify that occludin cleavage was a specific result in N . meningitidis-infected cells , HBMEC were treated with Cytochalasin D , which also promotes a round cell shape in endothelial cells , and analyzed for occludin expression by immunoblot analysis . As shown in Figure 3D no cleavage product was observed when HBMEC were treated with Cytochalasin D even at prolonged time of >8 h ( data not shown ) . As occludin contains a putative extracellular matrix metalloproteinase ( MMP ) cleavage site within the first extracellular loop [35] , we hypothesized that the cleavage product was the result of MMP-dependent cleavage . The 50-kDa weight cleavage fragment shown in Figure 3 B and C can be attributed to a cleavage within the first extracellular loop because cleavage at any other site would result in smaller or bigger sized fragments . To verify this hypothesis , HBMEC were infected with N . meningitidis strain MC58 in the presence of the broad-spectrum MMP inhibitor GM6001 and cell lysates were again analyzed by immunoblotting . Actually , formation of the 50-kDa fragment was blocked in the presence of the GM6001 ( Figure 4A ) . The inactive form of GM6001 did not prevent occludin proteolysis ( Figure 4A ) . Since GM6001 is a general MMP inhibitor with low-nanomolar inhibition of MMP-1 , -2 , -3 , -8 , and -9 , we next assessed the involvement of specific MMPs by using the specific MMP-3 , MMP-8 and MMP-2/9 inhibitors , MMP-3 inhibitor II , MMP-8 inhibitor I and MMP-2/9 inhibitor I . In the presence of specific MMP-8 I inhibitor formation of the occludin cleavage product was again prevented , whereas the inactive form of MMP-8 did not prevent occludin cleavage ( Figure 4B ) . A caspase cleavage site in the C-terminal cytoplasmic domain has recently been described [35] . To further exclude the possibility that the 50–kDa occludin fragment was due to apoptotic effects in HBMEC after N . meningitidis infection , cells were also pre-incubated with a pan-caspase inhibitor . In the presence of the membrane-permeable caspase inhibitor Z-DEVD-fmk , generation of the 50-kDA fragment was not impaired ( Figure 4A ) . To support the involvement of MMP-8 in N . meningitidis-induced cleavage of occludin , we used RNA-mediated interference to knock down the expression for MMP-8 . HBMEC were transfected with 150 nM MMP-8 siRNA or with control siRNA . 72 h post transfection , cells were infected with N . meningitidis for 24 h , lysed , immunoprecipitated and analyzed for occludin expression . HBMEC transfected with MMP-8 siRNA demonstrated the fully preserved 62-kDa band representing the full-length protein in infected HBMEC ( Figure 4C ) , while transfection with control siRNA did not affect occludin cleavage ( Figure S2 ) . Inhibition of MMP-8 expression by siRNA was monitored from cell lysates from transfected cells and examined by Western blot ( Figure 4D ) . β–actin detection was used to control for integrity of samples and equal protein loading . To prove whether inhibition of MMP activity could preserve occludin localization at the cell periphery , occludin distribution was visualized in the presence of GM6001 and the specific MMP-8 inhibitor by immunofluorescence microscopy . As shown in Figure 4E , inhibition of MMP-activity partly prevented relocation of occludin in MC58-infected HBMEC . Finally , analyses of N . meningitidis-induced occludin degradation and cell-detachment by means of permeability studies were also suggestive to an involvement of MMPs in this process: Inhibition of MMP activity by GM6001 and the specific MMP-8 inhibitor abolished N . meningitidis-induced paracellular permeability of 40 kDa FITC-dextran measured at 24 h p . i . ( P<0 . 05 ) ( Figure 4F ) . Since MMP-8 resulted in cleavage of occludin we next analyzed whether infection of HBMEC triggers release of active MMPs . We determined a time-course of MMP-8 activity in the supernatants of infected HBMEC using a 5-FAM/QXL 520 FRET peptide as a substrate . HBMEC were infected with both strains MC58 and MC58 siaD using different bacterial concentrations and MMP-8 activity was assayed at 2 , 4 , 8 , and 24 h p . i . As shown in Figure 5 , infection of HBMEC with both strains induced a time and dose-dependent secretion of active MMP-8 in the supernatant . Time-course of active MMP secretion paralleled occludin cleavage . Increasing the MOI from 30 to 500 significantly enhanced the release of active MMP-8 . Capsule expression did not contribute to active MMP-8 release as no difference between both strains was observed . Matrixmetalloproteinase activity can result in the processing of the extracellular matrix ( ECM ) , integrins and other proteins . Since numerous cells were detached from glass slides and/or filters during immunfluorescence analysis we decided to analyze HBMEC adherence to the matrix during N . meningitidis infection in detail . HBMEC were infected as described above and cell detachment was either visualized or estimated by crystal violet staining . A progressive detachment of HBMEC during infection was observed . The extent of HBMEC detachment varied according to the time of infection: Detachment from the culture support was visible as soon as 8 h p . i . and clearly visible at 24 h p . i . in response to N . meningitidis ( Figure 6A ) . Determination of specific cell detachment revealed that approximately 60% of cells detached at 24 h after infection with the wild-type strain N . meningitidis MC58 and approximately 40% of cells after infection with the unencapsulated mutant strain MC58 siaD , respectively ( Figure 6B ) , indicating that HBMEC detachment occurred independent from capsule expression . To determine whether soluble factors released from bacteria could promote HBMEC detachment , bacteria were placed above a transwell membrane and detachment of endothelial cells placed in the lower chamber was assessed . As shown in Figure 6B , HBMEC detachment was not induced unless direct bacteria-cell contact occurred , suggesting that injury of HBMEC monolayers required adhesion of the bacterium to the cell . Moreover , no detachment was observed when bacterial supernatant was collected and directly added to the cell monolayer ( Figure 6B ) . To investigate the cellular response in more detail , scanning EM was performed on monolayers of HBMEC that were infected with meningococci: In contrast to uninfected cells , cells infected with MC58 or MC58 siaD released cell-cell contact and acquired a rounded morphology that was consistent with bacteria-induced cell detachment ( Figure 6C ) . We next considered the possibility that endothelial cell detachment after interaction with N . meningitidis might involve matrix-degrading MMPs . HBMEC were infected with N . meningitidis strains MC58 and MC58 siaD in the presence of the broad-spectrum MMP inhibitor GM6001 and the specific MMP-8 inhibitor and analyzed for cell detachment . As shown in Figures 6D and E , HBMEC detachment could efficiently be inhibited by GM6001 and the specific MMP-8 inhibitor assessed by crystal violet staining and determination of specific cell detachment . To further prove data observed by crystal violet staining , we established an impedance-based real-time cell electronic sensing system ( xCELLigence ) ( Protocol S1 ) [36]–[38] . First , we determined the optimal concentration for cell proliferation and adhesion and estimated that seeding of 25 , 000 cells per well was best suited for further experiments ( data not shown ) . Next , 72 h after seeding HBMEC were infected with N . meningitidis MC58 and MC58 siaD at an MOI of 30 or left uninfected in the presence or absence of MMP inhibitors . Cell index values , corresponding to the intensity of cell adhesion and detachment , were monitored every 15 min using the xCELLigence system for additional 40 h . The increase of the number of attached cells on the electrodes leads to higher Cell index ( CI ) values , in which the CI represents a dimensionless unit due to the relative change in electrical impedance , while cell detachment will lead to a decreased CI values . As shown in Figure S3 , CI values drastically decreased in N . meningitidis-infected HBMEC at 24 h p . i . : The CI decreased from 6 . 24±0 . 93 and 6 . 19±0 . 81 at the beginning of the infection assay ( t = 0 h ) to 0 . 72±0 . 34 and 0 . 54±0 . 26 at 24 h p . i . in MC58 and MC58 siaD-infected HBMEC , respectively . Addition of MMP-inhibitors significantly decelerated the decrease of CIs in N . meningitidis-infected HBMEC corroborating our findings observed by crystal violet staining ( 1 . 29±0 . 47 and 1 . 57±0 . 49 in MC58 and MC58 siaD-infected HBMEC in the presence of GM6001 and 1 . 57±0 . 49 and 1 . 49±0 . 32 for MC58 and MC58 siaD in the presence of the specific MMP-8 inhibitor ( Figures S3 B and C ) ) . Detachment from the matrix is also a late feature of apoptosis in endothelial cells [39] . Pathogenic Neisseriae have been shown to induce the expression of apoptosis-related genes and to trigger apoptosis in different cell types [40]–[42] . Yet , caspase-independent detachment of infected cells has recently been shown after infection with the closely related species N . gonorrhoeae [43] . To analyze whether caspase activity is required for HBMEC detachment in response to N . meningitidis , cells were infected in the presence of the pan-caspase-inhibitor z-VAD-fmk and detachment was measured over a 24-hours time period . Interestingly , N . meningitidis-infected HBMEC still continued to detach when Z-VAD-fmk was present in the medium at a concentration of 25 µM ( Figure 7 ) . To determine a mutual dependence from MMP activity of apoptosis signaling , infection assays were carried out in the presence of the pan-caspase inhibitor and MMP-8 activity was estimated in the supernatant . Intriguingly , when infection of HBMEC with N . meningitidis MC58 was carried out in the presence of the pan-caspase inhibitor Z-VAD-fmk at a concentration of 25 µM , we still detected an increase in MMP-8 activity during time-period ( Figure 7B ) , indicating that release of active MMP-8 does not necessitate caspase activity . Furthermore , in order to test , whether an apoptotic stimulus was able to induce release of active MMP-8 , Staurosporin ( STS ) , a potent apoptosis inductor , was used as control . HBMEC were therefore treated with 1 µM and 5 µM STS for 2 , 4 and 8 h and analyzed for active MMP-8 secretion in the supernatant . Indeed , induction of apoptosis by STS resulted in significant release of active MMP-8 already after 2 h post treatment ( Figure 7C ) . Finally , we explored apoptosis rate in adherent and detached ( ‘floating’ ) HBMEC after infection with N . meningitidis strain MC58 . Apoptosis was scored using Annexin-V staining followed by flow cytometry analysis and TdT-mediated dUTP-biotin nick end labeling ( TUNEL ) staining . Rates of cells in apoptosis are given in Table S1 and Figure S4 . While adherent cells displayed low signs of apoptosis , we observed significant apoptosis in cells collected from the supernatant ( detached cells , ‘floating’ cells ) . About 10±2 . 3% and 54 . 4±8 . 6% of detached cells after infection with MC58 and MC58 siaD , respectively , at MOI 30 were found to be apoptotic ( Table S1 ) .
The present study analyzes the physiological and morphological alterations of endothelial cells comprising the BBB after exposure to Neisseria meningitidis . This endothelium differs functionally and morphologically from the endothelial cells of the leaky peripheral vasculature owing to the presence of tight junctions [44] . These cells form the basis of the BBB , which is the primary route penetrated by N . meningitidis during meningococcal meningitis [45] . In this study we observed an increase of permeability of brain endothelial cells after prolonged time of infection with N . meningitidis . By investigating the mechanism of permeability changes we discovered an unexpected connection between pathogen-induced matrix metalloproteinase ( MMP ) secretion and disruption of cell-cell connections . Infection with bacteria induced the secretion of active MMPs in the supernatants of infected HBMEC . In particular MMP-8 then participated in the cleavage of the tight junction protein occludin , which resulted in its diffuse accumulation in the cell . Moreover we observed that bacterial adhesion to HBMEC in cell culture resulted in a progressive detachment of the endothelial cell from the underlying matrix . Likewise as observed for occludin disruption MMP activity accounted for cell detachment . MMPs have already been shown to increase capillary permeability and act as effectors of BBB opening [14] . They target the subendothelial basement membrane ECM as well as cytokines and their receptors [46] . It has only recently been recognized that MMPs can cleave other host proteins , such as cell adhesion molecules – for example , CD44 and αv integrin – as well as some cytokines and tumor necrosis factor ( TNF- ) α [47] and that they may also play a crucial role in regulation of tight junction dynamics [48] , [49] . MMP-7 for example can shed VE-cadherin , a major component of endothelial adherens junctions [50] . MMPs do not have specific cleavage sequences on their target molecules , and cleavage sites are based on the tertiary structure of the protein and not on the primary amino acid sequence . Therefore , it is difficult to predict a priori which proteins may be cleaved by each MMP . Occludin contains a putative MMP-cleavage site in the first extracellular loop [35] and cleavage of this loop would result in fragments of about 50-kDa weight as observed in N . meningitidis-infected HBMEC . Recent studies have clearly demonstrated that occludin serves as a substrate for MMP-3 and MMP-9 [51] , [52] . Furthermore , MMP-7 is involved in disruption of occludin shown in vaginal epithelial cells [53] . However , we are unaware of examples of occludin cleavage involving MMP-8 activity . In this study , we provide several lines of evidence that MMP-8 is involved in occludin disruption . Inhibition of MMP-8 activity using the broad-spectrum inhibitor GM6001 as well as a specific MMP-8 inhibitor reduced N . meningitidis-induced occludin cleavage . Furthermore blocking MMP-8 activity by silencing of MMP-8 resulted in inhibition of occludin degradation . In addition , our data showed the involvement of MMP-8 in morphological ( occludin relocation ) and functional ( permeability increase ) alterations , and blocking MMP-8 activity preserved occludin attached at the cell membrane under infection and reduced increase of permeability . The cleavage of occludin mediated by MMP-8 in the first extracellular loop suggested that the C-terminal part of occludin is not affected during infection and therefore remains associated with the membrane and still interacts with ZO-1 . This would explain the appearance of ZO-1 in our immunfluorescence microscopy analysis , where ZO-1 localization was not affected during infection and still remained localized to the cell periphery . Otherwise , it has been shown that the function of ZO-1 is not exclusively linked to that of tight junctions [54] , but that ZO-1 also interacts with components of adherence junctions such as cadherins . The appearance of ZO-1 at the cell periphery may therefore also be due to binding to adherence junction components . It is therefore an open question , whether the remaining part of occludin stays associated with the apical tight junction or becomes distributed . Occludin cleavage was observed in response to both meningococcal strains . Western blot analyses suggested that the occludin band in the lane with MC58 siaD was lower in intensity compared with MC58 . However , densitometric analyses of cleaved occludin and full-length occludin levels taken from three replicates did not show statistically significant differences between both isolates ( data not shown ) . This is in accordance with the observed MMP secretion , since both strains triggered the release of equal amounts of active MMPs . An interesting study recently published by Coureuil et al . [30] gave new insights into the mechanism of BBB transcytosis by N . meningitidis . In this work type IV pili of N . meningitidis are shown to induce the recruitment of the polarity complex Par3/Par6/PKCζ leading to formation of ectopic junction-like domains at the site of bacterial microcolonies . The formation of the novel junctions weakens the endothelial cell-cell tight junction with opening of the intercellular junctions allowing meningococcal crossing of the BBB by a paracellular route . This cell-cell junction leakage is triggered by type IV pili and occurs at 2 h p . i . The temporal sequence of events suggests that cell-cell junction disruption by rerouting the intercellular junction molecules precedes MMP-8-mediated occludin disruption . However , this initial weakening of tight junctions might favor further protein degradation processes as described for occludin in this study . In addition to the effect of MMP-8 on tight junction components , adherence of the cell to the basement membrane was also affected in our infection model system . We observed that the contact between the bacteria and HBMEC in culture resulted in the detachment of the endothelial cell from the underlying matrix . We have recently shown that the Opc protein plays a pivotal role in the interaction with brain endothelial cells [34] . Several groups have shown that expression of the polysaccharide capsule , which confers strong protection against the host's immune defense [55] , blocks the interaction of meningococci with epithelial and endothelial cells , and only unencapsulated bacteria were able to enter these cell types . It was therefore concluded that the capsule functionally masks membrane proteins , like the Opa and Opc proteins , that mediate a close contact to and the internalisation into host cells . To study the effect of direct contact of N . meningitidis via the Opc protein we therefore incorporated an unencapsulated isogenic mutant in our experiments . Our preliminary data on the initiation of cell detachment triggered by N . meningitidis revealed that both the encapsulated and unencapsulated mutant strains were capable to promote HBMEC detachment , indicating that the endothelial cell detachment is either not dependent on capsule polysaccharide expression nor on the invasive capacity . We enhanced our end-point assays using an instrument with impedance technology , which allows dynamic monitoring of cell adhesion and detachment during N . meningitidis infection . Initial data revealed that 25 , 000 cells per well were the optimum concentrations for a dynamic monitoring of cell adhesion and detachment . Using this technology real-time analysis of cell detachment nicely correlated with our data observed with the end-point crystal violet staining assays . Detachment paralleled MMP secretion and was inhibited by the general MMP inhibitor GM6001 and a specific MMP-8 inhibitor . HBMEC detachment might occur due to cleavage of protein components of the ECM by MMPs , resulting in inappropriate ECM interaction . Alternatively , dissociation of transmembrane matrix receptors from the cytoskeleton anchoring proteins might contribute to the phenomenon . The adherence of cells to the ECM provides survival signals through mechanisms that include the activation of integrin receptors that have engaged particular ECM proteins . When such anchored cells are detached from the substratum the loss of integrin signaling can result in apoptosis , a phenomenon named anoikis [18] , [19] . Indeed , detached HBMEC displayed significant signs of apoptosis . On the other hand , endothelial cell detachment has been demonstrated as a late feature of endothelial cells undergoing apoptosis [56] and pathogenic Neisseriae have been shown to induce the expression of apoptosis-related genes and to trigger apoptosis in different cell types [40]–[42] , [57] . Caspase 3 has been shown to act as an effector for the cytoskeletal remodelling involved in cell “rounding” that occurs before the apoptotic cell detaches [39] . However , we could show that HBMEC still continued to detach in the presence of the pan-caspase inhibitor Z-VAD-fmk , indicating caspase-independend cell detachment . These findings are in line with a recently published study by Kepp and co-workers , who demonstrated that the closely related species N . gonorrhoeae triggers a caspase-independent detachment of infected epithelial cells [43] . Moreover , we observed a low rate of apoptosis in remaining adherent cell after infection with N . meningitidis strain MC58 . The inability of strain MC58 to induce apoptosis has recently been reported by Deghmane et al . [42] for HecIB cells . MMPs have been described to exert cytotoxic effects [17] , [58] . Apoptosis can be affected by direct proteolysis of death-inducing signaling components: MMP-7 for example has been described to mediate apoptosis through the generation of a soluble form of Fas ligand that initiates Fas-dependent apoptosis [59] . However , whether both enzyme families are separately activated and impact on cell detachment or cause themselves mutually remains to be elucidated . Taken together our data support the central role of MMP-8 in the disassembly of both host cell junction components and cell adhesion to the ECM as a consequence of meningococcal infections . The mechanisms by which MMP-8 is activated during meningococcal infection will be one focus of further investigations , as delineating this process will be fundamental to increasing our understanding of the interaction between this major meningitis causing pathogen and brain endothelial cells . However , so far our data suggest that adhesion contributes to detachment of infected cells . This could occur through increasing local delivery of bacterial factors , such as the lipooligosaccharid ( LOS ) . The importance of endotoxin of Gram negative bacteria in mediating cell damage has been documented [60] , [61] . Analyzing the specific effect of meningococcal LOS on HBMEC detachment as well as exploring the influence of N . meningitidis on further cell junction components will be will be an ambitious task for further investigations .
Goat anti-actin ( I19 ) antibody and rabbit anti-ZO-1 ( clone I19 ) were purchased from Santa Cruz Biotechnology ( Santa Cruz , CA , USA , 1∶100 ) . Mouse anti-occludin ( clone OC-3F10; against C-terminus ) was from Zymed ( Invitrogen , Ca , USA ) . Secondary antibodies goat anti-mouse IgG alkaline phosphatase-conjugated antibody and goat anti-rabbit IgG alkaline phosphatase-conjugated antibody were obtained from Invitrogen ( Invitrogen , Ca , USA ) . Alexa Fluor546-conjungated goat anti-rabbit , Alexa Fluor488-conjugated goat anti-mouse antibodies and 4′ , 6-diamidino-2-phenylindole , dihydrochloride ( DAPI ) were obtained from Invitrogen . The broad spectrum matrix-metalloproteinase ( MMP ) inhibitors GM6001 ( and its inactive form ) , the specific MMP inhibitors ( MMP-2/MMP-9 Inhibitor I , MMP-8 Inhibitor I , MMP-3 Inhibitor II ) , as well as the caspase inhibitor Z-D ( OMe ) E ( OMe ) VD ( OMe ) -FMK ( Z-DEVD-fmk ) and the pan-caspase inhibitor Z-VAD fmk were obtained from Calbiochem ( Schwalbach , Germany ) . Staurosporine ( STS ) were purchased from Sigma Aldrich ( Taufkirchen , Germany ) . FITC-Annexin V kit was from BioCat ( Heidelberg , Germany ) . Neisseria meningitidis strain MC58 , a serogroup B isolate ( United Kingdom , 1983 ) of the ST-32 complex characterized as serotype B:15:P1 . 7 , 16 was kindly provided by E . R . Moxon . Non-encapsulated mutant MC58 siaD , used in this study was previously described in detail [34] . The simian virus 40 large T antigen-transformed human brain microvascular endothelial cells ( HBMEC ) were kindly provided by K . S . Kim [22] and were cultured as previously described [34] . Cells between the 10th and 25th passages were used for infection assays . HBMEC were cultured in T25 flasks ( Corning Costar Corporation , Cambridge , MA , USA ) to a confluent monolayer . At 48 hours prior to infection , HBMEC were split and seeded on matrigel ( BD Matrigel Matrix , Heidelberg , Germany ) –coated 24–well tissue culture plates ( Sarstedt; Germany ) or on matrigel–coated Transwell cell culture chambers ( polycarbonate filters , 3 . 0-µm pore size; Corning Costar Corporation , Cambride , MA , USA ) at a density of 5×104 cells per well . Cells were grown to approximately 1×105 cells prior to infection . Monolayers of HBMEC were infected with bacteria at an MOI of 30 unless indicated otherwise for a 24-hour time period . Infections were carried out in the presence of 10% human serum ( HS ) supplemented RPMI medium . HS were derived from a serum pool ( voluntary staff ) and heat-inactivated for 30 min at 56°C . Use of HS–supplemented RPMI medium is based on the observation that meningococcal entry is supported by binding of the Opc protein via fibronectin to integrins on HBMEC [34] . Adhesion assays were performed as described previously [34] . We repeatedly tested the wild-type strain and the isogenic capsule deficient mutant for pili , Opa and Opc expression before application to infection assays and after re-isolation from the cell culture using Western blot analysis to exclude variation in the expression level of these meningococcal components . To compare the effects of soluble factors , cells were cultured in transwells with bacteria placed in the upper chamber , separated by a permeable filter ( 0 . 4 µm pore size ) . Paracellular permeability was studied by measuring the apical-to-basolateral flux of Fluorescein isothiocyanate ( FITC ) -dextran ( Sigma , St Louis , MO , USA ) through confluent HBMEC monolayers . HBMEC were seeded onto matrigel-coated Transwell filters at 5×104 cells/filter in 200 µl of HS–supplemented RPMI medium . The lower compartment was filled with 800 µl of the same medium . HBMEC were grown for 5 days to obtain confluence . About 1 . 7% of 4–kDa FITC-dextran and 1 . 1% of 40–kDA FITC-dextran was found in the lower chamber when flux was analyzed in a 30 min period . Matrigel was used in a concentration of 10 µg ml−1 which allowed free diffusion of tracers and bacteria ( data not shown ) . To measure paracellular flux , 4–kDa or 40–kDa FITC-dextrans were dissolved in P buffer [10 mM 4- ( 2-hydroxyethyl ) -1-piperazine-ethanesulphonic acid ( HEPES ) , pH 7 . 4; 1 mM sodium pyruvate; 10 mM glucose; 3 mM CaCl2 , 145 mM NaCl] . Bacteria were grown as described above , resuspended in HS–supplemented RPMI medium and inoculated on the apical surface of the cell layer at a MOI of 30 unless indicated otherwise . After indicated time points of infection , transwell inserts were replaced to measure paracellular flux . Cells were allowed to equilibrate in P buffer for 20 min and FITC-dextrans were added to the apical compartment to give a final concentration of 1 mg ml−1 . After 30 min , the basolateral medium was collected and the concentrations of FITC-dextrans were measured with a fluorometer in the presence or absence of inhibitors ( excitation 485 nm , emission 535 nm ) . Non-infected cells served as a negative control while non-infected cells incubated in phenylarsenoxid ( PAO/DMSO ) buffer served as a positive control for tight junction disruption as described previously [48] . For scanning electron microscopy , cells were fixed in situ with 2% glutaraldehyde/3% formaldehyde buffered in 0 . 1 M cacodylate , 0 . 09 M sucrose , 0 . 01 M CaCl2 , and 0 . 01 M MgCl2 ( pH 6 , 9 ) for at least 1 h at 4 C . Samples were dehydrated in a graded series of acetone on ice . After critical point drying from liquid CO2 , samples were sputter coated with gold/palladium and examined at 15 kV of accelerating voltage in a field emission scanning electron microscope ( model DSM962 , Zeiss , Oberkochen , Germany ) . Images were digitally recorded and processed in Adobe Photoshop CS . HBMEC were infected with bacteria ( MOI 30 ) as described . After infection , cells were gently washed with 1xPBS while shaking for 10 min to remove loosely attached cells . Adherent cells were fixed , dyed with crystal violet ( 0 , 75% crystal violet , 50% ethanol , 0 , 25% NaCl , 1 , 75% formaldehyde ) and incubated at room temperature for 30 min . Cells were washed twice with PBS and air dried . Cells were examined with a Zeiss Axiovert 40 CFL inverted light microscope using high NA 10X objective and microscopic photos were taken using the Zeiss AxioCam ICc1 digital camera system . After microscopy , cells were lysed in elution solution ( 1% SDS in 1xPBS ) overnight . The staining intensity was measured in a microplate reader ( Tecan ) at 620 nm . Bacteria were furthermore separated from the cell monolayer by a transwell filter system ( Transwell cell culture chambers , polycarbonate filters , 0 . 4-µm pore size; Corning Costar Corporation , Cambride , MA , USA ) or bacteria were grown to midlog phase , filtered and supernatants were added directly to the cell monolayer . The percentage of detached cells is expressed as a percentage ( mean ± SD of two wells from five independent assays ) relative to detachment observed in uninfected control cells . HBMEC were grown to confluence on matrigel–coated glass coverslips or on matrigel–coated Transwell filters for localization studies . Monolayers were infected with N . meningitidis strains MC58 and MC58 siaD ( MOI of 30 ) for 2 , 4 , 6 , 8 and 24 h in HS–supplemented RPMI in the presence or absence of inhibitors . Cells were rinsed gently three times with PBS to remove extracellular non–adherent bacteria and fixed in PBS with 3 . 7% paraformaldehyde for 20 min at RT . Cells were then rinsed , permeabilized with 0 . 1% Trition-X-100 in PBS for 15 min and blocked in 1% BSA in PBS for further 45 min . Monolayers were incubated with primary antibodies overnight at 4°C . Primary antibodies were used at the following dilutions: rabbit anti-ZO-1 ( 1∶50 ) , mouse anti-occludin ( clone 19 ) ( 1∶500 ) , and mouse anti-occludin ( clone OC-3F10 ) ( 1∶150 ) . Following incubation with the first antibodies monolayers were washed three times with PBS and than incubated with the appropriate secondary antibodies Alexa 546-conjugated goat anti-rabbit ( 1∶200 ) and Alexa 488-conjugated goat anti-mouse ( 1∶200 ) . Monolayers were viewed on a Zeiss Axio Imager . Z1 microscope equipped with ApoTome . Images were photographed using AxioCam digital Camera and AxioVision software and imported into Adobe Photoshop CS for manuscript preparation . HBMEC monolayers were incubated with N . meningitidis strains MC58 and MC58 siaD for indicated time points and then washed three times in ice-cold PBS . Proteins were extracted using ice-cold buffer containing 1% Triton-X-100 , 20 mM Tris-HCl , 150 mM NaCl , 0 . 1% SDS , 1% deoxycholic acid , 5 mM EDTA , 1 mM Na3VO4 , 2 mM phenylmethylsulphoyl fluoride ( PMSF ) , 50 µg ml−1 pepstatin , 50 µg ml−1 chymostatin and 50 µg ml−1 antipain . Extracts were spun at 14 . 000×g for 10 min and the supernatant was removed . β–actin detection was used to control for integrity of samples and equal protein loading . MMP inhibitor GM6001 and the inactive form of GM6001 ( 20 µM ) , as well as the specific MMP-8 inhibitor ( 20 µM ) and the inactive form of the MMP-8 inhibitor ( 20 µM ) were added 60 min prior to infection , Z-DEVD-fmk ( 50 µM ) was added 1 h prior to infection . All inhibitors were again added at 6 h p . i . Protein ( 30–50 µg ) from total cell lysates was used for immunoblot analysis . SDS-PAGE ( 6% acrylamide for ZO-1 and 10–12% acrylamide for occludin and claudin-1 ) was performed as described previously [62] using Mini-Protean electrophoresis System ( Bio-Rad , München , Deutschland ) . Proteins were transferred electrophoretically on nitrocellulose membranes ( Millipore , Bedford , MA ) using 30 mA for 1 h . The membranes were incubated overnight at 4°C in blocking buffer ( Tris-buffered saline , 0 . 1% Tween 20 ( TBST ) , 5% skim milk ) and than incubated with primary antibodies diluted in blocking buffer for 1 h at RT ( rabbit anti-ZO-1 ( 1∶80 ) , mouse anti-occludin ( clone 19 ) ( 1∶50 ) , mouse anti-occludin ( clone OC-3F10 ) ( 1∶2000 ) ) . After washing in TBST , the membranes were incubated with the appropriate secondary antibody ( goat anti-mouse IgG alkaline phosphatase-conjugated antibody: 1∶10000; goat anti-rabbit IgG alkaline phosphatase-conjugated antibody: 1∶5000 ) diluted in blocking buffer for 1 h at RT . After washing in TBST , the bands were detected using an enhanced chemiluminescence kit ( Amersham ) , according to the manufactures' instructions . All Western blots are representative of at least three experiments carried out . 3 µg monoclonal anti-occludin antibody/sample were added to cleared lysates containing equivalent amounts of protein and incubated for 4 h at 4°C . After addition of protein A/G plus agarose ( Santa Cruz Biotechnology , Santa Cruz , CA ) and 1 h incubation at 4°C , samples were washed twice with RIPA buffer and twice with Triton buffer ( 25 mM Hepes ( pH 7 . 4 ) , 1% Triton X-100 , 150 mM NaCl , 20 mM MgCl2 , 10% glycerol , 10 mM sodium pyrophosphate , 100 mM NaF , 1 mM Na3VO4 , and 10 µg ml−1 of each aprotinin , leupeptin , and pepstatin ) . For Western blot analysis , the precipitates were taken up in reducing 2x SDS sample buffer and analyzed as described above . siRNA ( MMP-8 and Control ) was synthesized by Santa Cruz Biotechnology . MMP-8 siRNA ( sc-35949 ) is a pool of three target-specific 20–25 base sequences siRNAs designed to knock down expression . Control siRNA ( sc-37007 ) is a non-targeting 20–25 nt siRNA , which consists of a scrambled sequence that will not lead to the specific degradation of any known cellular mRNA . At 75% confluency , HBMEC were mock transfected on 24-well plates or transfected with either 150 nM MMP-8 siRNA or 150 nM control siRNA using 3 µl of HiPerfect Transfection reagent ( Qiagen ) according to the manufacture's guidelines . MMP-8 expression was monitored 72 h post transfection by Western blot analysis with an anti-MMP-8 antibody ( Acris , Herford , Germany ) . 72 h after transfection cells were infected with bacteria for 24 h before cell lysate collection and immunoprecipitation . Blots were reprobed with a goat anti-β-actin antibody to control for protein loading . MMP activity was determined using the Sensolyte 520 MMP-8 Assay Kit ( AnaSpec , San Jose , CA , USA ) following the manufacturer's instructions . Briefly , media from MC58 and MC58 siaD-infected HBMEC and non-infected HBMEC were collected 2 , 4 , 8 , and 24 h p . i . and 50 µl of these culture supernatant were incubated with the FAM/QXL 520 fluorescence resonance energy transfer substrate for 1 h in a black 96-well plate at room temperature in the dark . Measurements were made using a Tecan microplate reader ( excitation at 360 nm , emission at 465 nm ) . Statistical analyses were performed using Student's t test , and data were considered significant if P was <0 . 05 . | A crucial step in the pathogenesis of bacterial meningitis is the disturbance of cerebral microvascular endothelial function , resulting in blood-brain barrier ( BBB ) breakdown . Matrix metalloproteinases ( MMPs ) have been implicated in BBB damage in bacterial meningitis in several studies . MMPs are a family of zinc-dependent endopeptidases that catalyze the proteolysis of extracellular matrix proteins , but can also cleave a range of other molecules , including cell adhesion molecules . In this study we showed that brain endothelial cells produced MMPs—in particular MMP-8—upon infection with Neisseria meningitidis , a bacterium that causes meningitis and septic shock . We found that MMP-8 was then involved in disruption of the tight junction protein occludin . In addition to the effect of MMP-8 on the tight junction component , MMP-8 activity also accounted for brain endothelial cell detachment that occurred during prolonged time of infection with N . meningitidis . When we inhibited MMP-8 activity , occludin disruption was completely abolished and cell detachment could be partially prevented , which resulted in restored BBB permeability . Our data reveal a molecular mechanism of cellular dysfunction during meningococcal meningitis that enhances our understanding how MMPs affect cerebral endothelial function and that can aid in our understanding and prevention of this disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis",
"microbiology"
] | 2010 | Neisseria meningitidis Induces Brain Microvascular Endothelial Cell Detachment from the Matrix and Cleavage of Occludin: A Role for MMP-8 |
Our earlier genome-wide expression study revealed up-regulation of a tryptophan-catabolizing enzyme , indoleamine 2 , 3-dioxygenase ( IDO1 ) , in patients with scrub typhus . This gene has been previously reported to have anti-microbial activity in a variety of infectious diseases; therefore , we aimed to prove whether it is also involved in host defense against Orientia tsutsugamushi ( OT ) infection . Using LC-MS , we observed an increased ratio of serum L-kynurenine to serum L-tryptophan in patients with scrub typhus , which suggests an active catalytic function of this enzyme upon the illness . To evaluate the effect of IDO1 activation on OT infection , a human macrophage-like cell line THP-1 was used as a study model . Although transcription of IDO1 was induced by OT infection , its functional activity was not significantly enhanced unless the cells were pretreated with IFN-γ , a potent inducer of IDO1 . When the degree of infection was evaluated by quantitative real-time PCR , the relative number of OT 47 kDa gene per host genes , or infection index , was markedly reduced by IFN-γ treatment as compared to the untreated cultures at five days post-infection . Inhibition of IDO1 activity in IFN-γ treated cultures by 1-methyl-L-tryptophan , a competitive inhibitor of IDO1 , resulted in partial restoration of infection index; while excessive supplementation of L-tryptophan in IFN-γ treated cultures raised the index to an even higher level than that of the untreated ones . Altogether , these data implied that IDO1 was partly involved in restriction of OT growth caused by IFN-γ through deprivation of tryptophan . Activation of IDO1 appeared to be a defensive mechanism downstream of IFN-γ that limited intracellular expansion of OT via tryptophan depletion . Our work provided not only the first link of in vivo activation of IDO1 and IFN-γ-mediated protection against OT infection but also highlighted the promise of this multifaceted gene in scrub typhus research .
Scrub typhus is a potentially life-threatening infectious disease caused by Orientia tsutsugamushi ( OT ) , an obligate intracellular gram-negative bacterium transmitted to human through the bite of a larval trombiculid mite . The disease is one of major public health problems in Asia-Pacific region , where approximately one billion people are at risk , and one million cases are affected each year [1] . Exposure to the field where its host vectors inhabit is an important risk factor for the infection especially among people who live in rural areas or work in a farm . So far , no effective strategy has ever succeeded in providing long lasting immunity to this particular infection [2] . Earlier studies have shown that IFN-γ is essential in protection against OT infection in animal and cell-based models [3] , [4] , [5] . Apart from its well-known roles in macrophage activation and the development of type I immune response , it was found to directly inhibit intracellular growth of OT in non-immune host cells [6] . Nevertheless , the underlying molecular mechanisms remain unclear . In natural infection of scrub typhus , a marked elevation of IFN-γ was consistently observed in acute serum of the infected patients [7] , [8] , [9] . Accordingly , our recently published genome-wide expression data in patients with scrub typhus have revealed that IFN-γ and a number of IFN-related genes are up-regulated during acute phase of the illness [10] . Among these , indoleamine 2 , 3-dioxygenase-1 ( IDO1 ) is an interesting gene with a growing number of studies for its multifaceted roles in the immune system . It encodes an intracellular heme-containing dioxygenase [11] that is an initial and rate-limiting enzyme of kynurenine pathway , a major catabolic pathway of tryptophan . Since tryptophan is the rarest essential intracellular amino acid; activation of IDO1 , which leads to deprivation of tryptophan , becomes a potential strategy of the host to control the population of an invading microorganism . This notion was first proven in various cell models of toxoplasma infection , including human glioblastoma cell line [12] , primary endothelial cells [13] , uroepithelial cells [14] , as well as monocyte-derived macrophages [15] . Later on , IDO1-mediated microbial-stasis was also reported in a number of intracellular and extracellular infections; including those caused by Chlamydia spp . [16] , Mycobacterium avium [17] , herpes simplex virus [18] , [19] , measles virus [20] , cytomegalovirus [21] , dengue virus [22] , group B streptococcus [23] , enterococci [24] , and Staphylococcus aureus [25] . In addition , IDO1 also exerts an indirect antimicrobial activity via production of some certain downstream catabolites of kynurenine pathway [26] , [27] , [28] . For example , 3-hydroxy-kynurenine , a subsequent product of IDO1 activation , was found to limit replication of Trypanosoma cruzi in vivo; treatment of the infected mice with the metabolite can improve resistance to the infection as well as survival of the host [26] . Other downstream catabolites , including picolinic acid , 3-hydroxyanthranilic acid , and quinolinic acid , can also inhibit the growth of methicillin-resistant S . aureus , S . epidermidis , Escherichia coli , and multidrug-resistant Pseudomonas aeruginosa in vascular allograft [27] . Despite its protective role in a variety of infections , IDO1 activation paradoxically appears to be involved in suppression of the immune responses . This was first revealed by a key study showing that IDO1-mediated tryptophan degradation prevents allogeneic fetal rejection in mice [29] . It was later demonstrated in vitro that the enzyme exerts anti-proliferative effects on T cells , NK cells , as well as tumor cells via degradation of tryptophan and production of downstream metabolites , resembling its impact on microorganisms [30] , [31] , [32] , [33] . Indeed , IDO1 expression in certain dendritic cell subsets was found to induce tolerogenic responses to antigenic stimuli through a variety of mechanisms; including induction of T cell anergy [34] , apoptosis [35] and regulatory T cell differentiation [36] . When such concept is applied to a scenario of infection , it seems that IDO1 activation in some certain settings may contribute to ineffective development of adaptive immunity and allow a microorganism to persist . For example , HIV-induced IDO1 activation in peripheral blood of HIV-infected patients , which mainly contributed by plasmacytoid dendritic cell subpopulation , appeared to be responsible for unresponsiveness of CD4-positive T cell to stimulation of T cell receptor in vitro [37] . Moreover , levels of IDO1 expression in peripheral blood mononuclear cells from these patients also correlates with their viral loads; which further supports a link between IDO1 , T cell dysfunction , and ineffective viral control [37] . In rickettsial infection , it was reported that IFN-γ-mediated IDO1 activation inhibited the growth of Rickettsia conorii in vitro , and such restriction could be relieved by tryptophan supplementation [38] . In contrast , proliferation of Rickettsia prowazekii was insensitive to IDO1-mediated tryptophan depletion [39] . In vivo up-regulation of IDO1 , which positively correlated with IFN-γ and TNF-α expression , was observed at the skin lesion of patients with Mediterranean spotted fever , an illness caused by R . conorii infection [40] . However , the role of IDO1 in scrub typhus has never been directly investigated . We hypothesized that up-regulation of IDO1 upon infection with scrub typhus could lead to rescriction of OT growth in host cells according to the fact that OT lacks an enzyme to generate tryptophan , which is an essential material for its expansion [41] . Since monocytes are the main source of active IDO1 in human peripheral blood [42] , [43] and also seem to be a potential target of infection to carry the intracellular organisms to remote organs , a human macrophage cell line THP-1 was used as a model in subsequent in vitro infection experiments . In the present study , we reported that IDO1 acitivity was increased in patients with scrub typhus . Activation of IDO1 by IFN- γ resulted in a lower number of OT load in THP-1 macrophages , and the degree of infection could be partly restored by an inhibitor of IDO1 enzyme . Supplementation with high-dose tryptophan did not only reverse the suppression of OT growth but markedly increased the number of OT in IDO1 active cultures . Our work not only provided the first link of in vivo activation of IDO1 and IFN-γ-mediated protection against OT infection but also highlighted the promise of this multifaceted gene in scrub typhus research .
The study was conducted after obtaining the approval of the Ethics Committee of Faculty of Medicine Siriraj Hospital , Mahidol University , Bangkok , Thailand . Informed and written consent was derived from all patients and control subjects before their blood samples were collected . Serum samples were derived from patients with acute undifferentiated fever at the first visit to Siriraj Hospital and were stored at −80°C until being used . As a control group , serum samples from ten healthy blood donors were derived and handled in a similar manner . As detected by an indirect immunofluorescent assay , definite diagnosis of scrub typhus was made according to the following criteria: 1 ) presence of OT-specific antibody at a titer of ≥1∶400 in a single acute sera; or 2 ) ≥four-fold rising of OT-specific antibody in paired sera collected two weeks apart [44] . Only confirmed cases of scrub typhus were selected to undergo LC-MS analysis , and their clinical and laboratory data were retrospectively reviewed . Analysis of L-Trp and L-Kyn was performed using a validated high performance liquid chromatography with tandem mass spectrometry ( HPLC-MS/MS ) method in accordance with the USFDA guidelines [U . S . Department of Health and Human Services Food and Drug Administration Center for Drug Evaluation and Research 2001] . Briefly , 3-nitro-L-tyrosine was added to each sample as an internal standard , and protein precipitation was performed using 10% trichloroacetic acid . Chromatographic separation of subsequent organic layer was carried out on LC-MS/MS with C18 , 2 . 5 µm ( 50×3 . 00 mm i . d . ) . A mobile phase consisting of acetonitrile and 0 . 1% formic acid ( Gradient condition ) was delivered at a flow rate of 0 . 2 ml/min . Mass spectra were obtained using a Quattro Premier XE mass spectrometer ( Micromass . UK ) , operated in multiple reaction monitoring mode . Sample introduction and ionization was performed by electrospray ionization in the positive ion . The mass transition ion-pair for L-Trp [M+H]+ and L-Kyn L-Trp [M+H]+ ions was selected as m/z 205 . 08>188 . 00 and 205 . 08>146 . 08 respectively . The mass transition ion-pair for 3-nitro-L-tyrosine [M+H]+ ions was selected as m/z 227 . 02>181 . 03 . The data acquisition was ascertained by Masslynx 4 . 1 software . The detectable ranges of L-Trp and L-Kyn in both human serum and culture media were 0 . 05–50 µg/ml and 0 . 15–10 µg/ml respectively . The best linear fit was achieved with a 1/x weighting factor , showing a mean correlation coefficient ( r2 ) ≥0 . 998 . OT Standard Kato strain ( CSUR R163 ) was propagated using mouse fibroblast cell line L929 as a host . When more than 90% of the host cells were infected , as determined by Giemsa staining , the media was replaced with fresh media . The infected cells were then dislodged from the flask and broken down using 1 . 0-mm-diameter glass beads with vigorous vortexing . The disrupted cell suspension ensuing from multiple parallel cultures were pooled together and stored as multiple small aliquots in liquid nitrogen . Before being used , a few aliquots of frozen OT inoculum were thawed in water bath at 37°C . Afterwards , the cell suspension was thoroughly mixed and broken down once again with glass beads and vortex before being centrifuged at 500×g for five minutes to sediment host cell lysate . Subsequent OT-containing supernatant was immediately used to infect THP-1 as well as to determine the infectivity of the inoculum by infected cell counting method with some modifications [45] , [46] . Briefly , two-fold serial dilution of OT-containing inoculum was inoculated onto L929 monolayers on the glass coverslips . After 24 hours of incubation , the degree of infection was determined using an indirect immunofluorescence assay . Infected-cell-counting unit ( ICU ) was calculated according to the following formula: ICU = ( total number of cells used in infection ) × ( ratio of infected cells to counted cells ) × ( dilution fold of OT inoculum ) [45] . Twenty-four hours prior to an in vitro infection experiment , 2×105 cells of THP-1 cell suspension were transferred into each well of 24-well plate and treated with 100 nM pharbol 12-myristate 13-acetate ( PMA ) ( Sigma-Aldrich ) to induce cell adherence . When IDO1 expression was required , the cells were be treated with 20 ng/ml IFN-γ ( R&D Systems , Minneapolis , MN ) , along with 1 mM 1-methy-L-tryptophan ( 1-MT ) ( Sigma-Aldrich ) , L-Trp ( Sigma-Aldrich ) supplement , or neither . On the day of the experiments , the culture media was aspirated; and OT-containing inoculum of 5 . 9×105 ICU , prepared as described in the previous section , was inoculated onto the THP-1 monolayer . Infection process was facilitated by centrifugation at 1 , 450×g for five minutes and further incubation in a humidified 5% CO2 atmosphere at 37°C for one hour . The inoculum was then replaced with fresh PMA-containing media , together with additional treatments corresponding to each pre-infectious condition; this time point was designated as zero hour post-infection ( p . i . ) . At specified time points , culture media was collected for determination of IDO1 activity or just discarded . The cell layer was then rinsed three times with phosphate buffer saline to wash out extracellular organisms . Finally , the infected cells were collected and further processed for RNA or DNA study . Total RNA was extracted from each cell culture that had been lyzed in Trizol reagent ( Invitrogen , Calsbad , CA ) in accordance to the manufacturer's instruction . cDNA was synthesized from subsequent RNA extract using SuperScript III First-Strand Synthesis System ( Invitrogen , Carldbad , CA ) . Specific amplification of TATA binding protein ( TBP ) and IDO1 gene transcripts was performed in duplicate using LightCycler FastStart DNA Master SYBR Green I reagents ( Roche Applied Science ) , two µL of an appropriate dilution of a cDNA sample , and a specific primer pair for TBP or IDO1 . The primer sequences were given in Table S1 . Relative quantification analysis of qPCR data was then performed by LightCycler 480 software ( Roche Applied Science ) based on comparative ΔΔCt method . IDO1 and TBP were set as the target gene and the reference gene respectively . Finally , a mean expression ratio of IDO1 and TBP in each culture was normalized to that of mock-infected culture harvested at 0 hour p . i . . DNA was isolated from infected cell cultures using standard phenol/chloroform method [47] . Specific amplification of human methylenetetrahydrofolate reductase ( MTHFR ) gene and OT 47 kDa gene was performed in triplicate using LightCycler FastStart DNA Master SYBR Green I reagents ( Roche Applied Science ) , two µL of an extracted DNA sample , and a specific primer pair for MTHFR or OT 47 kDa . The primer sequences were given in Table S1 . Relative quantification analysis of qPCR data , based on comparative ΔΔCt method , was adopted to indirectly assess the degree of OT infection in each culture . OT 47 kDA and MTHFR were set as the target gene and the reference gene respectively . Finally , the relative number of OT 47 kDa per MTHFR in each culture at different time points was normalized to the mean OT 47 kDa/MTHFR ratio of infected cultures that did not expose to IFN-γ harvested at 6 hours p . i . . The normalized OT 47 kDa/MTHFR value was designated as “infection index” , which reflects the extent of intracellular harbour of OT in each culture condition . To reflect IDO1 activity , the proportion of L-Kyn to L-Trp was calculated for each sample . Differences in the levels of serum L-Trp , L-Kyn or L-Kyn/L-Trp ratio between any two groups of the subjects were evaluated using Mann-Whitney test . For continuous clinical and laboratory variables , their association with the levels of serum L-Trp , L-Kyn or L-Kyn/L-Trp ratio was assessed using Spearman correlation . For experimental data , differences of L-Trp level , L-Kyn level , L-Kyn/L-Trp ratio , IDO1 expression , and infection indexes between any two culture conditions were evaluated using unpaired t-test with or without Welch's correction as appropriate .
Twenty patients with confirmed scrub typhus were enrolled into the study . Clinical and laboratory data of seventeen patients were available and were summarized in Table 1 . To check whether functional activity of IDO1 enzyme actually increased in concordance to our earlier transcriptional study [10] , serum levels of L-Trp and L-Kyn of patients with scrub typhus were determined ( n = 20 ) . As shown in Figure 1A and 1B , serum level of L-Trp was significantly lower in scrub typhus infected patients as compared to healthy individuals ( P = 0 . 0146 ) , whereas a reverse trend was observed for serum L- Kyn level ( P = 0 . 0002 ) . To see how IDO1 activity differs between the two groups of subjects , the ratio of L- Kyn to L-Trp was evaluated . As shown in Figure 1C , the enzyme activity in patients with scrub typhus was about nine times higher than that of healthy controls ( P<0 . 0001 ) . Correlation analysis between levels of serum L-Kyn , serum L-Trp , or IDO1 activity and characteristics of the patients revealed no association for most of the clinical and laboratory parameters except for serum AST , whose level significantly correlates with serum L-Kyn ( ρ = 0 . 6412 , P = 0 . 0074 ) ( Figure S1 and Table S2 ) . When the patients were classified by their clinical features , no significant difference in levels of serum L-Kyn , serum L-Trp , or IDO1 activity was observed between any two subgroups of the patients . ( Table S3 ) To prove our hypothesis on the role of IDO1 in OT infection , we first evaluated temporal profiles of IDO1 gene transcription in an experimental model . As shown in Figure 2A , IDO1 expression was evidently induced in OT-infected THP-1 at 24 hours p . i . ( P<0 . 0001 ) and increased up to 18 times greater than mock-infected cultures at 120 hour p . i . ( P = 0 . 0001 ) . IFN-γ treatment dramatically enhanced IDO1 transcription to even more than 2000 folds since 6 hours p . i . ( P<0 . 0007 ) , but the expression became comparable to that of unstimulated infected cultures at 72 and 120 hour p . i . ( Figure 2B ) . The degree of IDO1 induction in cultures co-treated with IFN-γ and 1-MT was lower than those treated with IFN-γ alone at 6 and 24 hours p . i . ( P = 0 . 077 and 0 . 001 respectively ) , yet overall kinetics of IDO1 expression in both conditions were quite similar ( Figure 2B ) . To assess functional activity of IDO1 in a cell model , levels of L-Trp and L-kynurenine in culture supernatant of THP1 cells was determined at corresponding time points . As shown in Figure 3A , the level of L-Trp in OT-infected cultures was relatively lower than that of mock infection at 6 hours p . i . ( P = 0 . 001 ) yet still significantly greater than that of IFN-γ treated ones ( P = 0 . 0011 ) . Co-treatment of IFN-γ and 1-MT retarded the rate of L-Trp consumption as compared with IFN-γ treatment alone at the same time point ( P = 0 . 0334 ) . However , the amino acid levels in culture supernatant fell below the limit of detection in all conditions by 24 hours p . i . ( Figure 3A ) . Despite the different rate of L-Trp consumption , the dynamic of L-Kyn level in OT-infected cultures was similar to that in mock-infected counterparts ( Figure 3B and 3C ) . IFN-γ treatment significantly doubled L-Kyn level at the first 6 and 24 hours p . i . ( P = 0 . 0027 and 0 . 0009 respectively ) , but the level became comparable to that of mock infection at later stages ( Figure 3B ) . Unexpectedly , a continuous rise of L-Kyn level was observed in 1-MT treated cultures ( Figure 3C ) . Since the difference of L-Trp levels among the four culture conditions was relatively unremarkable in comparison with that of L-Kyn levels , the profiles of L-Kyn/L-Trp ratio resembled that of L-Kyn as shown in Figure 3D and 3E . To investigate the effect of IDO1 activation on OT growth; an infection index , which is a relative copy number of OT 47 kDa gene per a human gene MTHFR , was determined in OT-infected cultures with or without IFN-γ-induced IDO1 activation . As shown in Figure 4A , the infection index was significantly lowered by IFN-γ treatment at 5 days p . i . ( P = 0 . 0273 ) . When IDO1 activity in IFN-γ treated cultures was inhibited by 1-MT , the infection index was partially restored . Replenishment of L-Trp in the culture media at a concentration of 400 µg/ml could not reverse the suppressive effect of IFN-γ-induced IDO1 induction on OT growth ( Figure 4A ) . However , supplementation of the amino acid at 1 mg/µl markedly increased the infection index of IFN-γ-treated cultures to even higher than that the untreated counterparts as early as 3 days p . i . ( P = 0 . 0326 ) ( Figure 4B ) .
Activation of IDO1 is a host defensive mechanism downstream to IFN-γ that has been proven to limit the growth of various infectious pathogens in both immune and non-immune cells in vitro [12] , [13] , [17] , [18] , [19] , [20] . It also appears to be induced in human subjects infected with some particular pathogens , such as Leishmania guyanensis [40] , HIV-1 [37] and dengue virus [22] . According to the fact that OT lacks tryptophan-synthesizing enzyme [41] , IDO1-mediated deprivation of tryptophan is a potential protective mechanism to limit the activity of this particular organism . From our earlier analysis of genome-wide expression , we have observed an up-regulation of IDO1 in peripheral blood leukocytes of patients with acute scrub typhus [10] . Thus , activation of IDO1 was first confirmed at functional level in patients with acute scrub typhus in the present study . Considering that monocytes appear to be the main source of active IDO1 in human peripheral blood [42] , [43] , THP-1 cells were used as an experimental model of infection in further testing of our hypothesis . Indeed , up-regulation of IDO-1 was previously demonstrated in OT-infected monocyte-derived macrophages [10] . Besides , we believe that macrophages could be a secondary target of infection to harbour the intracellular organisms to remote organs; which was supported by the detection of OT in circulating monocyte-like blood cells of patients with acute scrub typhus [48] and in macrophages located in the liver , the spleen , and the lymph nodes of fatal cases of scrub typhus [49] , [50] . When THP-1 macrophages were infected with OT , transcription of IDO1 was directly induced by OT infection early after infection . However , a decreased level of L-Trp without a concomitant elevation of L-Kyn in OT-infected cultures as compared with mock infection may imply that higher rate of tryptophan consumption was contributed by increased utilization of the amino acid by the organism for its intracellular activity rather than enhancement of tryptophan catabolism by IDO1 enzyme . We postulated that the increase of IDO1 activity in vivo was a secondary event following the release of IFN-γ rather than a direct induction by the infection itself . For this reason , THP-1 was treated with IFN-γ prior to the in vitro infection to imitate natural infection in human body , in which other IFN-γ-producing cells such as NK cell , γδT cell , Th-1 cells and cytotoxic T cells , are also present . Such postulate seems sensible considering that the rise of IFN-γ during acute scrub typhus has been consistently observed by a number of earlier studies [7] , [8] , [9] , [10] . After pre-treatment with exogenous IFN-γ , a marked increase in transcription of IDO1 along with its functional activity was detected as early as 6 hours p . i . . Concurrently , the intracellular number of OT was significantly depressed by IFN-γ treatment at 5 days p . i . . Even though IFN-γ is known to cause multiple changes in macrophage biology that could influence the outcome of an intracellular infection , our data suggest that IFN-γ-mediated IDO1 activation is responsible for growth restriction of OT to some extent since the number of OT per host cell was partially restored by a competitive inhibitor of the enzyme . However , it must be noted that induction of IDO1 only limited the rate of OT proliferation , but neither froze the growth nor reduced the number of the intracellular organism . For this reason , development of adaptive immune mechanisms seems to be required for eradication of the infection . From earlier studies , anti-microbial activity of IDO1 has been explained by two non-mutually exclusive mechanisms: deprivation of tryptophan and formation of its downstream metabolites , collectively known as kynurenines . The first mechanism is based on the fact that tryptophan is the rarest essential amino acid in a human cell; therefore , IDO1-mediated tryptophan degradation , which limits the availability of the amino acid to be exploited by the organism , could lead to restraint of reproduction of tryptophan-sensitive microorganisms . In this study , we demonstrated that L-Trp supplementation at 400 µg/ml caused no change in the infection index of IFN-γ treated culture , whereas replenishment of the amino acid up to 1 mg/ml did not just restore but accelerated the growth of OT since the third day post infection . For this reason , it is not surprising why an earlier study that supplemented the culture media with 100 µg/ml of trytophan failed to rescue the growth of OT in IFN-γ-treated murine embryonic cell lines and refuted the role of tryptophan deprivation in control of OT infection [6] . Indeed , the number of the organisms in the culture with no IFN-γ treatment in the same experiment was demonstrated to be significantly increased by such small dose of tryptophan supplementation [6] , which supported the dependence of OT proliferation on the availability of L-Trp similar to ours . Altogether , these data implies that IDO1-mediated tryptophan deprivation can indeed restrict OT growth; but restoration or enhancement of its reproduction requires adequate replenishment of tryptophan that allows an excess of the amino acid to be available in IDO1-active environment long enough for exploitation by this particularly slow-growing organism . However , it should be aware that comparing the findings derived from experiments in murine and human cell models might not be straightforward since defensive mechanisms to an infection may vary from species to species . For example , it was found that IDO1 , but not iNOS , is essential in control of various infectious organisms in human mesenchymal stromal cells , while a reversed scenario is revealed for a similar cell type in mice [51] . In much the same way , it is possible that IDO1 is critical in control of OT growth in human macrophages but not in murine embryonic cells . For the second mechanism involving IDO1-mediated kynurenine formation , it has been proved that some tryptophan metabolites , particularly 3-hydroxy-DL-kynurenine and alpha-picolinic acid but not L-Kyn , can exert anti-microbial activity against several extracellular bacteria in a dose-dependent manner [27] . In our experiments , the role of these metabolites in OT control was not directly investigated . However , supplementation of L-Trp into the culture , which was assumed to result in a greater amount of downstream metabolites produced in the system , did not increase the extent of OT growth restriction . Thus , we concluded that that formation of kynurenines was unlikely to play a major role in suppression of OT growth . Despite the key findings that have already been discussed , we are also aware of some factors that could influence the interpretation of our experimental data . Firstly , the infection index , used as an indicator of OT growth in this study , was derived by a relative quantification method . It is , therefore , not only sensitive to a change in the number of OT but also to that of host cells at the time of the assessment . In the cultures supplemented with 1-mg L-Trp , we observed a reduction in the number of host cells by time , which might be partly responsible for a marked increase in the infection index of the condition compared with the others . Such cell loss could be a consequence of overwhelming expansion of OT in either tryptophan-rich environment or excessive accumulation of toxic metabolites in the culture system , or both . Secondly , the use of L-Kyn to L-Trp ratio as an indicator of IDO1 activity in THP-1 cells was unfortunately limited by the drop of L-Trp below the detectable level since the first day post infection . Moreover , some ambiguity in the data interpretation also resulted from the surprising rise of L-Kyn level in the culture supernatant of cells treated with 1-MT . Unlike other cultures whose L-Kyn level similarly reached its equilibrium at 72 hours p . i . , the concentration of L-Kyn in 1-MT treated cultures continuously rose beyond 120 hour p . i . . This might reflect a higher rate of L-Kyn formation exceeding the capacity of downstream enzymes to catabolize the molecule into downstream metabolites even in the presence of the IDO1-specific inhibitor . However , it is unlikely that the other two tryptophan-degrading enzymes , IDO-2 and tryptophan 2 , 3-dioxygenase ( TDO ) , were responsible for such phenomenon because the former appears to be expressed as an inactive form in human dendritic cells [52] and tumor cells [53] , while the latter is exclusively expressed by hepatocytes [54] . Other possible explanations include an incidental blockage of a downstream kynurenine-degrading enzyme by 1-MT treatment , which results in accumulation of L-Kyn . However , further investigation is needed to determine the exact underlying mechanisms for such finding . Apart from its anti-microbial activity , IDO1 also appears to be involved in suppression of immune responses as well as development of immunological tolerance . This concept originated from a key finding that demonstrated a role of IDO1-mediated tryptophan degradation in prevention of allogeneic fetal rejection [29] . It was later revealed that the enzyme exerts anti-proliferative effects on T cells , NK cells and tumor cells via degradation of tryptophan as well as production of its downstream metabolites , resembling its impact on microorganisms [30] , [31] , [32] , [33] . Then , the issue has been actively concentrated in a variety of research areas , including infection , transplantation , autoimmunity , and cancer , with the hope to develop effective therapeutic strategies for these conditions . In certain dendritic cell subsets , IDO1 expression also appears to induce tolerogenic response to antigenic stimuli through other varieties of mechanisms , including induction of T cell anergy [34] , apoptosis [35] and differentiation of regulatory T cell [36] . Based on our earlier study in patients with scrub typhus , such anti-proliferation of immune cells seemed unlikely to happen in vivo since we still observed marked leukocytosis , as well as up-regulation of genes in cell division process and leukocyte activation among the infected patients [10] . Nevertheless , further investigation is warranted to see whether IDO1 activation upon the infection also leads to tolerogenic responses or other modulation of the immune system . Recently , a high level of serum L-Kyn/L-Trp ratio was reported to be associated with the development of severe complications , like septic shock and multiple organ failure , in patients with major trauma [55] , sepsis [42] , and bacteremia [56] . IDO1 activity during the course of the illness also appears to predict the severity and fatality for these patients [42] , [55] , [56] . According to its regulatory roles in the immune system as mentioned earlier , hyperactivity of IDO1 has been suspected to contribute to immune dysregulation that might underlie the development of such complications [55] , [57] . Seeing that similar serious conditions are also common in severe cases of scrub typhus , activation of IDO1 might be involved in the development of such complications; its high functional activity could be a poor prognostic marker for this life-threatening disease as well . However , if we consider that activation of IDO1 would restrain the expansion of OT , an expected outcome would be in reverse . Considering that a positive correlation between the bacterial load in peripheral blood of scrub typhus patients and severity of the disease was recently reported [58]; it remains compelling to investigate a relationship between IDO1 activity and OT load , as well as , to reevaluate its association with clinical features in a larger set of patients with variable severity . In conclusion , this is the first report for IDO1 activation in patients with scrub typhus , which has brought this multifaceted gene with promising therapeutic potential into focus in the field of scrub typhus research . We demonstrated here that IDO1-mediated tryptophan deprivation was a downstream mechanism of IFN-γ that helps restrain intracellular expansion of OT . However , further studies are deserved to investigate other potential effects of IDO1 activation on the outcome of OT infection in a more complex experimental model as well as in more patients with scrub typhus . | Scrub typhus is a potentially life-threatening infectious disease that is a major cause of acute undifferentiated fever in Asia-Pacific region . It is caused by Orientia tsutsugamushi ( OT ) , an obligatory intracellular gram-negative bacterium in family rickettsiaceae . Earlier studies have shown that IFN-γ is essential in protection against OT infection in animals and cell-based models , but molecular mechanisms underlying such phenomenon remain largely unclear . In the present study , we are the first to demonstrate that activation of IDO1 , a key enzyme in tryptophan-degrading pathway , is a mechanism downstream to IFN-γ in control of OT infection . IDO1 was found active in patients with acute scrub typhus . Subsequent in vitro experiments suggested that IDO1 restrains OT growth via deprivation of tryptophan , an important material for proliferation of OT , rather than via production of its downstream anti-microbial metabolites . Since IDO1 is a multifaceted gene that also plays a part in several processes of the immune system , including induction of tolerogenic responses; other possible consequences of IDO1 activation deserve further investigation in a more complex experimental model as well as in naturally infected human before subsequent clinical implication of this basic knowledge can be ultimately applied . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"zoonoses",
"clinical",
"immunology",
"immunity",
"immunity",
"to",
"infections",
"immunology",
"biology",
"scrub",
"typhus"
] | 2012 | Activation of Indoleamine 2,3-Dioxygenase in Patients with Scrub Typhus and Its Role in Growth Restriction of Orientia tsutsugamushi |
Myosin VI , found in organisms from Caenorhabditis elegans to humans , is essential for auditory and vestibular function in mammals , since genetic mutations lead to hearing impairment and vestibular dysfunction in both humans and mice . Here , we show that a missense mutation in this molecular motor in an ENU-generated mouse model , Tailchaser , disrupts myosin VI function . Structural changes in the Tailchaser hair bundles include mislocalization of the kinocilia and branching of stereocilia . Transfection of GFP-labeled myosin VI into epithelial cells and delivery of endocytic vesicles to the early endosome revealed that the mutant phenotype displays disrupted motor function . The actin-activated ATPase rates measured for the D179Y mutation are decreased , and indicate loss of coordination of the myosin VI heads or ‘gating’ in the dimer form . Proper coordination is required for walking processively along , or anchoring to , actin filaments , and is apparently destroyed by the proximity of the mutation to the nucleotide-binding pocket . This loss of myosin VI function may not allow myosin VI to transport its cargoes appropriately at the base and within the stereocilia , or to anchor the membrane of stereocilia to actin filaments via its cargos , both of which lead to structural changes in the stereocilia of myosin VI–impaired hair cells , and ultimately leading to deafness .
The molecular motor myosin VI is known to function as either an actin-based anchor or as a transporter , based on biochemical , biophysical and cell biological studies ( reviewed in [1] ) . This protein moves along actin toward the minus end , in the opposite direction to all other characterized myosins to date [2] . Myosin VI achieves ‘gating’ or coordination of movement along actin by first having the rear head of the dimer strongly bound to actin , while blocking the lead head from binding ATP and thus continuing through its ATPase cycle until the rear head is released [3] . This is achieved by a unique insert near the myosin transducer region that facilitates communication between the actin interface , myosin lever arm , and nucleotide-binding elements of the motor domain . Myosin VI plays an essential role in Drosophila , where it was originally discovered [4] , [5] , as well as in C . elegans [6] , mice [7] , and zebrafish [8] . Mutations in these organisms have confirmed myosin VI function in sperm individualization during spermatogenesis by remodeling of the plasma membrane , unequal portioning of organelle and cytoskeletal components , basal protein targeting and spindle orientation in mitotic neuroblasts , and regulation of actin-based interactions with the plasma membrane . In mammalian cells , myosin VI appears to be involved in clathrin-mediated endocytosis and maintenance of Golgi morphology and protein secretion , as well as movement and clustering of receptors and vesicle scission of endocytic vesicles [9]–[11] . Most compelling , mutations in the myosin VI gene are associated with a dominantly inherited form of human hearing loss , DFNA22 [12] , a recessively inherited form of human deafness , DFNB37 , and hypertrophic cardiomyopathy [13] , [14] . Interactions with proteins have been defined for myosin VI with GIPC1 ( GAIP-interacting protein , C-terminus ) that recruits myosin VI to uncoated vesicles [15] , SAP97 [16] and Dab2 [17] that recruit myosin VI to clathrin-coated pits and vesicles , and optineurin , required for myosin VI localization at the Golgi complex [18] . Most recently , TRAF6-binding protein ( T6BP ) and nuclear dot protein 52 ( NDP52 ) interactions with myosin VI have implicated a role for this protein in membrane trafficking pathways with cell adhesion and cytokine-dependent cell signaling [19] . Myosin VI and vinculin were recently shown to interact , suggesting that myosin VI acts as an actin-based anchor to facilitate vinculin's link with cadherin complexes on actin filaments in epithelial cells [20] . Though by convention a cytoplasmic protein , myosin VI has been identified in the nucleus of mammalian cells , where it modulates the RNA polymerase II-dependent transcription of active genes [21] . In the inner ear , though myosin VI is clearly essential , its exact role remains a mystery . Myosin VI is one of the earliest hair cell-specific proteins , expressed in the mouse cochlea as early as embryonic day ( E ) 13 [22] . In the inner and outer hair cells of the organ of Corti , myosin VI was suggested to serve as an anchor , maintaining the structure of the stereocilia [23] . Most of our information regarding protein function in the inner ear has come from the study of mouse mutants with mutations in genes encoding these proteins . Over 190 deaf and circling mutants exist in which the mutant gene is known and in 46 of these cases , they serve as corresponding models for human hereditary hearing loss ( The Jackson Laboratory's Hereditary Hearing Impairment in Mice database , http://hearingimpairment . jax . org/master_table . html; [24] ) . The mutants include those that arose spontaneously , radiation-induced mutants , mutants produced by gene-targeted mutagenesis and N-ethyl-N-nitrosourea ( ENU ) -generated mutants ( reviewed in [25] ) . The mouse Snell's waltzer mutant , a spontaneous mutation that arose in a colony at the Jackson Laboratory in 1966 , was found to contain mutations in myosin VI ( Myo6 ) approximately 30 years later [7] . Instrumental for this search was a radiation-induced mutant , sesv . Research using mouse models for deafness have emphasized the need for multiple alleles , since each mutation may lead to new phenotypic descriptions , allowing for elucidation of new functions for a given protein . Tailchaser , an ENU-generated dominant mouse mutant that arose from a large-scale mutagenesis screen [26] , [27] , was identified as a deaf and circling mouse mutant and shown to display a gradual deterioration of both hearing and balance function , similar to forms of dominant nonsyndromic deafness in humans . A scanning electron microscopy ( SEM ) study revealed that Tlc stereocilia bundles fail to form the characteristic V-shape pattern at birth , and by adulthood , hair bundles are severely disorganized and eventually degenerate . We have now identified a missense mutation in myosin VI in Tlc and revealed new insights into myosin VI function with this new Myo6 allele .
Linkage for the Tlc mutation was found to chromosome 9 between markers D9Mit104 and D9Mit182 in a genome scan of 21 randomly selected N2 Tlc/+ mice from a [C3HeB/FeJ-Tlc/+×C57BL/6J]×C3HeB/FeJ backcross . There was no evidence of linkage to any of the other autosomal chromosomes . Linkage to sex chromosomes was excluded by analyzing the mating data . Extending the analysis to a total of 84 N2 mutant mice and using an additional marker within the identified linkage interval confirmed linkage to mouse chromosome 9 to a region of 29 Mbp between D9Mit75 and D9Mit182 . Genotyping of additional polymorphic markers on all 183 available N2 mutant mice revealed 22 mice with recombinations in the region of the mutation , narrowing the region to 6 Mbp between D9Mit74 and D9Mit133 ( Figure 1A ) . This linkage interval contained 28 known RefSeq genes ( http://genome . ucsc . edu ) . Previously , we reported that the Tlc mutation was most likely localized to chromosome 2 , although there were some inconsistent genotypes [27] . Analysis to further define the region demonstrated that C57 backcrossed N2 mutants ( from a [C3HeB/FeJ-Tlc/+×C57BL/6J]F1×C57BL/6J backcross ) presented a milder mutant phenotype when compared to the N2 progeny of the C3H backcross mutants , suggesting reduced penetrance . In the original analysis , only C57 backcrossed N2 mutants were analyzed , that may have led to erroneous phenotyping and subsequently incorrect matings and therefore mistaken localization . The unconventional myosin VI ( gene symbol , Myo6 ) is localized in the center of the chromosome 9 non-recombinant interval . Myosin VI is an actin-based molecular motor that has been previously shown to underlie hereditary hearing loss in both human and mice ( see Introduction ) . The Myo6 gene encodes a 1265 amino acid protein ( 140 kD ) that consists of an N-terminal motor domain , a calmodulin interacting neck domain and a C-terminal tail domain that is necessary for binding cargo and forming myosin VI homodimers ( reviewed in [1] ) . Direct sequencing of Myo6 cDNA extracted and amplified from Tlc/+ brains revealed a c . G694T transversion in a region that corresponds to exon 6 of the Myo6 gene ( NM_008662 ) ( Figure 1B ) . This mutation was confirmed by reverse sequencing of new mutant and wild type cDNA and direct sequencing of Tlc/+ , Tlc/Tlc , wild type , sv/sv , C3H , C57 and Spretus genomic DNA ( data not shown ) . The mutation was identified only in Tlc mutant mice and in none of the other DNA samples tested . A restriction digestion assay of PCR-amplified Myo6 genomic DNA was tailored to identify the mutation as an alternative method of genotyping to screen all mice ( see Materials and Methods ) . The G694T mutation is predicted to result in an aspartic acid to tyrosine amino acid substitution at position 179 of the myosin VI protein ( p . D179Y ) . The conservation pattern of myosin VI was analyzed using ConSurf , which uses the Rate4Site algorithm for estimating the evolutionary rates at each amino acid site ( Figure 1C ) [28] . Aspartic acid in position 179 is highly conserved . Noteworthy is the fact that in the positions aligned with the D179 residue in the multiple sequence alignment ( MSA ) , none of the 77 sequences , found in a hidden Markov model ( HMM ) search based on similarity to known myosin VI sequences , harbor an aromatic amino acid . The expression pattern of myosin VI in Tlc mutant inner ears was evaluated by immunohistochemistry with an antibody that detects the tail of myosin VI , and phalloidin , which labels filamentous actin . The specificity of the antibody used was validated by Hasson and colleagues [29] and was confirmed using auditory sensory epithelia of Snell's waltzer mice ( Figure S2 ) . Previously , myosin VI staining was observed in the cell body , the cuticular plate and pericuticular necklace of hair cells [29] and between the actin core and plasma membrane of stereocilia [30] . We confirmed this general expression pattern in wild type mice , in which specific immunostaining was evenly distributed along the length of control stereocilia ( Figure 2A–D , I ) , with an increase at the base . Myosin VI expression was also observed in hair cells from Tlc/Tlc mutants , but with a significant difference in distribution in adults ( P70 ) , with stereocilia often showing enhanced myosin VI staining in the upper third portion ( Figure 2E–H , J ) . The same staining pattern of myosin VI was observed in all turns of the cochlea despite much more pronounced hair cell degeneration and hair cell death in the basal turn . Only fused stereocilia forming giant protrusions showed a more diffused staining pattern . A comparison of green pixel intensity profiles revealed that the intensity of myosin VI specific staining was higher in Tlc/Tlc stereocilia ( 53 . 81±30 . 25 , n = 95; t-test: 4 . 7×10−16 ) than in control stereocilia ( 15 . 81±7 . 93 , n = 53 ) . In wild type mice , at E18 . 5 , all hair cells are already present and their hair bundles are aligned with a slight degree of hair bundle disorientation , as demonstrated by SEM . Both inner and outer hair cell stereocilary bundles in all turns of the cochlea show a staircase-like organization with rows of stereocilia of graded height ( data not shown ) . By P1 , the stereocilia of the second row have a larger diameter than those from other rows on each cell and their tips are pointed in middle and basal turns [31] . In wild type mice , supernumerary stereocilia can still be found at the front of all hair bundles . At E18 . 5 , mutant hair bundles were indistinguishable from those of controls and all of them showed a staircase-like organization of the stereocilia within middle and basal turns of the cochlea . Some slight degree of misalignment of hair bundle orientation was observed in all three genotypes at E18 . 5 but this is normal and there is no indication of a planar cell polarity defect at this age . Surprisingly , by P1 , hair bundles of wild type littermate controls ( Figure 3A–D ) were easily distinguishable from those of Tailchaser heterozygotes ( Figure 3 E–H ) and homozygotes ( Figure 3I–L ) . Both homozygotes and heterozygotes for this mutation show disorganization of hair bundles , reflected in a loss of the characteristic ‘V’ hair bundle shape and an overall flattening or even a concave-like shape of the bundle compared with the characteristic convex ‘V’ shape . In addition , when we measured the width of the P1 stereocilia from the tallest row , we noticed that the wild type stereocilia have a classic Gaussian normal distribution with 0 . 3 micron average ( n = 80 ) , whereas the width of the tallest Tlc/Tlc stereocilia distributes in a wide width range with 0 . 38 micron average ( n = 80 ) ( Figure S1 ) . Finally , the hair bundle disorganization is accompanied by a variable position of the kinocilium . To further evaluate the position of the kinocilium we labeled cochlea from P1 wild type , Tlc/+ and Tlc/Tlc mice with an antibody for acetylated tubulin , which stains the kinocilium , and with phalloidin to stain filamentous actin ( Figure 3D , H , L ) . While the kinocilia of the P1 wild type hair cells were uniformly localized to the center of the lateral ( strial ) side of the hair cell , the kinocilia of the Tlc/+ mice were dispersed at varying positions between the lateral side of the hair cells and the center of the hair cell . In order to quantify this phenomenon , a region spanning 17–20 hair cells was randomly selected from the basal turn of wild type , Tlc/+ and Tlc/Tlc mice . The location of the kinocilia in reference to the apical surface of the cell was then plotted and quantified as percentage distance from the center of the apical surface of the cell both in the X ( apical to basal ) and Y ( modiolar to striolar ) dimensions , with 0% representing the center of the cell and 100% representing the cell surface perimeter ( Figure 3M ) . While in wild type mice the kinocilia were localized to 65% of the distance from the center of the cell to its perimeter in the Y axis ( SD 13 . 22 ) and 8 . 6% ( SD 14 . 7 ) towards the base in the X axis , kinocilia localization in the Tlc/+ mice was 28% ( SD20 . 48 ) and −5 . 25% ( 28% ) in the Y and X axis respectively and −7 . 6% ( SD18 . 38 ) and 5 . 9% ( SD 14 . 6 ) in the Y and X axis of the Tlc/Tlc mice . A t-test analysis shows that while the differences in the localization of the kinocilia along the X-axis of the hair cells is not significantly different between the different genotypes ( p values >0 . 05 ) , the localization of the kinocilia in the Y axis is significantly different between the wild type and Tlc/+ mice ( p value 1 . 17E-07 ) , wild type and Tlc/Tlc mice ( p-value 5 . 09E-15 ) and between the Tlc/+ and Tlc/Tlc mice ( p value 3 . 24E-06 ) . Interestingly , the average localization of kinocilia of the Tlc/Tlc mice was centralized and even closer to the modiolar side of the hair cell apical surface than to the lateral side of the hair cells' apical surface . At postnatal day 21 ( P21 ) morphologically mature hair bundles on the apical surface of outer and inner hair cells of wild type mice show a characteristic staircase-like organization of stereocilia graded in height ( Figure 4A , D ) . At this stage of hair bundle development both tip links ( Figure 4G ) ( connecting the pointed tips of shorter stereocilia with the side of adjacent longer stereocilia ) and horizontal top connectors ( Figure 4H ) ( forming links along the stereocilia length ) were clearly visible on both types of hair cells . As previously described in Tailchaser heterozygotes [27] , the stereocilia on outer hair cells form staircase-like bundles but of highly variable shape ( Figure 4B ) . In contrast , in Tailchaser homozygotes , the staircase-like arrangement of stereocilia of graded height was very unclear and hair bundles were disorganized ( Figure 4C ) . Despite severe changes in bundle shape , both tip links and horizontal top connectors were visible in Tailchaser homozygotes and heterozygotes ( Figure 4I , J ) . The morphological changes observed were less pronounced in inner hair cell bundles that still showed a staircase-like stereocilia organization in both hetero- ( Figure 4E ) and homozygotes ( Figure 4F ) for the Tailchaser allele . To check if the observed disorganization of the stereocilia in mature hair bundles was caused by an earlier effect of the Tailchaser mutation on the presence or structure of transient interstereocilial links , we analyzed mouse cochleae at P7 , expecting both lateral and ankle links to be fully developed by then [32] . At P7 , in hair bundles of wild type littermates , lateral links form irregular rays of fine fibers between the upper two thirds of the length of adjacent stereocilia within and between rows and are clearly visible in all bundles analyzed ( Figure 4K ) , while ankle links connecting the basal third of the stereocilia length were seen only sporadically , probably due to the tissue processing method used ( Figure 4L ) . Interestingly , both lateral and ankle links were also present in hetero- ( Figure 4M ) and homozygous ( Figure 4N ) Tailchaser mutants and they appeared normal in structure . The original null mutation of the Myo6 gene causes extensive and early stereocilia fusion in Snell's waltzer mice mutants [23] . Tailchaser homozygotes also exhibit stereocilia fusion , which can be observed first at P1 in the apical turn ( Figure 5A ) . At P21 stereocilia fusion was spread along the entire length of the cochlea affecting both outer and inner hair cell bundles in the apical turn ( Figure 5B ) and mostly outer hair cells in middle and basal turns of the cochlear duct ( Figure 5C ) . In addition , high-resolution analyses of hair bundles from the apical turn of the cochlea of Tailchaser homozygotes at P1 revealed stereocilia branching ( Figure 5D , E ) , a feature also found in Snell's waltzer mutants [23] . In this study we examined Snell's waltzer mice at P6 , when degeneration is quite advanced in the middle and basal turns but mild in the apical region of the cochlea . We confirmed the presence of stereocilia branching in many outer and inner hair cell bundles along the entire length of the cochlea ( Figure 5F ) . Myosin VI functions in epithelial cells to transport nascent uncoated endocytic vesicles through actin-dense regions [33]–[35] . Disruption of myosin VI activity blocks this process , providing us with an assay to ask whether the D179Y mutation is sufficient to disrupt myosin VI function . We introduced the D179Y point mutation into wild type human myosin VI tagged with Green Fluorescent protein ( GFP ) [35] . The incorporation of this mutation into the construct , HGFP-M6 ( D179Y ) , was confirmed by DNA sequencing and the proper expression was confirmed by Western blot ( data not shown ) . Since the tail of myosin VI is sufficient to target to nascent uncoated vesicles ( UCV ) , we hypothesized that HGFP-M6 ( D179Y ) would also target properly to these vesicles . ARPE-19 cells were transfected with HGFP-M6 ( D179Y ) and then fixed and stained with antibodies specific for GIPC , an adapter protein that collocates with myosin-VI on the UCV surface [33] , [35] ( Figure 6A ) . HGFP-M6 ( D179Y ) targeted to peripherally located vesicles with 73% of GIPC-associated UCV colocalizing with the myosin VI protein ( Figure 6A , B ) . Control experiments ( not shown ) established no overlap between HGFP-M6 ( D179Y ) and clathrin , consistent with the identity of the GIPC and myosin-VI-labeled structures as UCV [33] , [34] . Mutations that disrupt myosin-VI motor activity block delivery of UCV vesicles to the early endosomes [33] , [34] . This block can be easily visualized by following steady-state uptake of rhodamine-conjugated transferrin ( R-Tsfn ) [33] , [34] . We therefore predicted that overexpression of HGFP-M6 ( D179Y ) would alter R-Tsfn uptake . ARPE-19 cells were transfected with HGFP-M6 ( D179Y ) , HGFP-M6 ( which has no effect on trafficking [33] ) , or PGFP-M6tail ( which lacks a myosin-VI motor domain and disrupts trafficking [33] ) , then incubated with R-Tsfn for 15 minutes . Following fixation , the transfected cells were scored for accumulation of transferrin in the perinuclear recycling and early endosome compartment ( Figure 6C , D ) . After 15 minutes of incubation , 52 . 7+/−2 . 5% of HGFP-M6 cells transfected exhibited a prominent perinuclear accumulation of R-Tsfn ( Figure 6C , D ) . In contrast , overexpression of GFP-M6 ( D179Y ) caused a drastic decrease in delivery of R-Tsfn to the early endosome with only 20 . 5+/−0 . 7% of cells exhibiting a perinuclear accumulation ( Figure 6C , D ) . These results are equivalent to those seen for cells expressing GFP-M6tail ( 19 . 7+−2 . 0% of cells exhibit endosomal delivery ) , confirming that introduction of the D179Y mutation is sufficient to disrupt myosin VI's function as an endocytic motor . The D179Y mutation is in a helix that follows a loop ( Loop 1 ) that is involved in altering the steady state ATPase rate and ADP release rate of myosins [36] , and precedes the nucleotide-binding element known as switch I [37] ( Figure 7A ) . It is thus in a position to alter communication in the region that has been termed the transducer in myosin motors , which rearranges as myosin goes through its force producing cycle on actin . To evaluate if the D179Y mutation may disrupt the transducer region , we assessed the maximal steady state actin-activated ATPase rates , which for wild type dimers shows half the rate per head than does a monomer , since the lead head cannot cycle until the rear head detaches . As shown in Table 1 , the steady state ATPase rates ( actin-activated ) for a myosin VI monomer ( S1 ) incorporating the D179Y mutation is decreased compared to wild type . However note that in the mutant dimer , the rate per head is the same as the monomer . Thus gating , or communication between the heads of the dimer , has been destroyed and the mutant myosin is incapable of performing its cell biological functions .
We have identified a Myo6 mutation affecting the motor domain that permits normal protein expression levels and sub-cellular localization , but disrupts its function as an endocytic motor and prevents appropriate coordination of the myosin heads , known as ‘gating’ , to move along actin . Eventually , this leads to disorganized hair cell bundles and branching of stereocilia in mammalian inner ear hair cells . The identified mutation is a G694T transversion leading to an aspartic acid to tyrosine amino acid substitution at position 179 , adding to the catalogue of Myo6 mutations that cause deafness and balance defects in mice and humans . There are several lines of evidence that implicate the D179Y mutation as being responsible for the Tailchaser phenotype . First , genetic mapping placed the Tlc mutation in the same chromosomal interval as Myo6 and all Tlc mutant mice identified by their abnormal behavior carry the G694T mutation , showing that the mutation is present in mice with the mutant phenotype . Furthermore , D179 is evolutionarily conserved . Second , the stereocilia fusion seen in Tlc mutant hair cells is very similar to that described in Snell's waltzer myosin VI-null mice . The branching seen for the first time in Tlc mutants was also found to be present in Snell's waltzer hair cells . Third , a functional assay revealed that both wild type and a D179Y mutant myosin VI was able to target to nascent uncoated vesicles , but the mutant D179Y myosin VI blocked delivery of UCV vesicles to the early endosomes , similar to the behavior of a motorless myosin VI . Fourth , the location of the D179Y mutation suggests that it would affect the transducer region of the myosin and indeed , measurements of actin-activated ATPase rates demonstrated that the disruption of the transducer region by D179Y leads to loss of ‘gating’ , or coordination of myosin VI in dimer form to move along actin . Thus , the D179Y mutation appears to impair the motor function of myosin VI . Two new features were found in the Tailchaser mutants that confer additional functions for myosin VI in the inner ear . First , myosin VI is expressed specifically in the inner ear hair cells early in development , beginning at E13 . 5 [22] , yet in mutant form , now identified in more than one mouse mutant , hair cells remain indistinguishable from wild type hair cells up to at least E18 . 5 . Detailed SEM analyses of auditory sensory epithelia of homozygous mice revealed that the Tailchaser mutation affects the overall organization of the stereociliary bundle . However , the mutation does not influence the formation of interstereocilial links or planar cell polarity of embryonic hair cells . The stereociliary bundles of outer hair cells in Tailchaser homozygotes progressively lose their staircase-like arrangement but despite these severe morphological rearrangements the formation and maintenance of interstereocilial links is not affected . This indicates that myosin VI is not required for the initial polarization of the hair cells or the proper positioning of the cilia of the bundle or the kinocilium . However , it is necessary for maintenance of the bundle orientation and overall morphology once the bundle has formed and is maturing , possibly via interaction with the cuticular plate . The first steps in hair bundle-formation are marked by the appearance of a central kinocilium ( a microtubule-based true cilium ) around mouse E13 on the microvilli covered hair cell . Many microvilli eventually develop into stereocilia . By birth , the kinocilium relocates to the middle of the lateral side on the apex of the hair cell , and the ‘V’ shaped hair bundles are all uniformly oriented . The kinocilia of the P1 Tlc/+ mice are clearly mislocalized and more centrally and variably distributed than the wild type kinocilia , indicating an early defect in their migration towards the lateral pole of the cell . Interestingly , the center of the hair cell bundle seems to follow the kinocilium and the overall organization of the stereociliary bundles in the P1 Tlc mutant mice is more flat or concave . Second , the Tlc mutation in myosin VI causes stereocilium branching . The presence of the branching was confirmed in Snell's waltzer , another Myo6 mutant . This phenomenon occurs at the same time as changes of stereocilia dimensions and suggests myosin VI involvement in not only dynamics of stereocilia actin but also its role in maintenance of parallel actin bundles . These findings are consistent with localization of myosin VI between the actin paracrystal and plasma membrane of stereocilia [30] . The processing myosin VI dimer could transiently interact with both the plasma membrane and actin paracrystal on its way towards the minus ends of the actin filaments localized at the stereocilia bases , and thus create a force that would push the actin paracrystal towards the stereocilia tip . In the absence of functional myosin VI , the forces upon the actin paracrystal and the putative weaker linkage of the plasma membrane to the actin could lead to dysregulation of actin treadmilling in stereocilia , affecting in turn the staircase-like organization of the hair bundle . It remains unclear , however , how presence or absence of myosin VI would influence the stereocilia actin core , leading to formation of stereocilia branches . Myosin VI could create tension between the plasma membrane and actin paracrystal and in this way mechanically inhibit branching . Myosin VI could also interact with the ARP2/3 complex , a complex of seven proteins that binds actin filaments and nucleates new actin filament assembly [38] , inhibiting actin filament nucleation and preventing formation of unwanted branches in normal stereocilia . Indeed , in Drosophila , myosin VI and the Arp2/3 complex colocalize , suggesting that myosin VI is concentrated in a region of active actin assembly [39] . In this work , we show myosin VI-specific staining along the length of control stereocilia at the level of immunofluorescence . Our current results are consistent with myosin VI immunogold localization [30] and immunofluorescence [40] . Others previously described myosin VI-specific immunofluorescence only in the cell cytoplasm , cuticular plate and stereocilia base [29] . Data obtained by Belyansteva and colleagues using gene gun transfections with Myo6-GFP [41] are consistent with the lack of stereocilia staining as well . The discrepancy between the negative results obtained in 1997 and the current positive stereocilia staining may be explained by differences in sensitivity of the detection systems used . There may be a number of reasons for false negative results of immunohistochemistry and expression of a fluorescently tagged protein ( e . g . tag effect on protein function , cell response to mechanical damage ) . Despite the inherent uncertainty in interpreting negative staining and expression data we do need to consider the possibility that the discrepancies in stereocilia localization that we describe here and previously [30] , [41] are false positive results for immunofluorescence and immunogold staining . However , the fundamental novel finding described in this manuscript is that the lack of functional myosin VI predominantly affects the shape and integrity of stereocilia bundles ( Figures 3 and 4 ) , indicating that myosin VI plays a role in stereocilia maintenance . The phenotypes that we observe are therefore consistent with stereocilial expression of myosin VI . For myosin VI to perform its anchoring and processive trafficking functions , it must be able to gate its heads [3] . In the case of myosin VI , the mechanism of gating , or communication between the heads of the dimer , during processive movement involves the lead head being unable to bind ATP until the rear head has released from actin [3] . During anchoring , the lead head would be unable to bind ATP [3] , and ADP would tend to out-compete ATP for binding to the rear head [42] , preventing either head from releasing from actin . Thus the simplest prediction of why the D179Y mutation , which is in the region of a helix that follows Loop 1 ( Figure 7A ) , causes deafness is that disruption of the transducer region of the myosin destroys gating , and thus allows ATP to bind and dissociate both the lead and rear heads simultaneously . This is easily assessed by the maximal steady state actin-activated ATPase rates , which for wild type dimers shows half the rate per head than does a monomer , since the lead head cannot cycle until the rear head detaches . In the mutant , the actin-activated rate per head is the same in the monomer and dimer , indicative of a loss of gating . Most compelling , the consequence of this immobility renders myosin VI to remain ‘stuck’ at the tips of the stereocilia , rather than expressed along the length , as demonstrated by immunofluoresence studies in Tlc/Tlc mutant inner ears . The presence of myosin VI along the length of the stereocilia in wild type mice , while it is known to move along actin towards the minus end of actin filaments ( towards the base of stereocilia ) , suggests that myosin VI molecules must be transported to the tip by another mechanism not involving the myosin VI motor . Myosin VI could then proceed down the actin filaments using its own motor in wild types , but in the Tlc mutants the defective motor would not permit this movement . This hypothesis was supported by our observation of an accumulation of myosin VI labeling at the tips of Tlc stereocilia . How does the loss of myosin VI as a processive motor , either due to its inability to move along actin properly , in the case of Tailchaser , or its absence in the case of Snell's waltzer , translate into the pathology seen in myosin VI-mutated hair cells ? The myosin VI motor spends a significant portion of its catalytic cycle bound to actin with a high duty ratio and requires mechanical coordination between the dimer heads while ‘walking’ along actin in relatively large steps [43] . This coordination and communication , or gating , between the two heads is essential . The kinetics of ATP hydrolysis is linked to the length of time each dimer head is attached to actin , determining whether myosin VI will act as a transporter or anchor . When myosin VI cargo is anchored in the membrane , then this motor would serve to apply force on the actin filament to remain close to the membrane , as may be the case in the cuticular plate at the base of the stereocilia ( Figure 7B ) . Lack of myosin VI , or mutant myosin VI that no longer can serve as this anchor , would permit the membrane to detach from the actin filament and the consequence would be fusion of two stereocilia at the base ( Figure 7B ) . A similar mechanism could be taking place that would allow branching of stereocilia to be formed in the absence ( Snell's waltzer ) or mutant form of myosin VI ( Tailchaser ) ( Figure 7B ) . This anchoring in the normal state may be achieved , for example , in part by the Arp2/3 complex ( described above ) . Cell adhesion , or lack of it , might be responsible for branching and/or stereocilia fusion due to altered coordination between myosin VI and vinculin and cadherin complexes [20] . While a multitude of cargoes have been identified for myosin VI that bind to its tail region , including GIPC , Ddab2 , Sap97 , and opineurin [15]–[18] , their role in clathrin-coated pits and vesicles in receptor-mediated endocytosis and Golgi secretion has been studied only in epithelial cells . Their task , however , in the inner ear , remains to be elucidated . The identification of the cargos in the stereocilia and cuticular plate may hold the key to understanding how mutant myosin VI causes hair bundle structural changes and ultimately , loss of auditory and vestibular function .
The founder mouse carrying the Tlc mutation was generated in a large-scale ENU mutagenesis program [26] . All procedures involving animals met the guidelines described in the National Institutes of Health Guide for the Care and Use of Laboratory Animals , were approved by the Animal Care and Use Committee of Tel Aviv University ( M-00-65 ) and were in compliance with UK Home Office regulations . The colony was maintained on the original C3HeB/FeJ genetic background . Tlc/+ ( C3H ) males were outcrossed to wild type C57BL/6 ( C57 ) females in order to generate F1 mutants . F1 mice were phenotyped and identified as mutant if they displayed a strong mutant phenotype that consisted of an impaired reaching response , head bobbing , hyperactivity and severely compromised performance in a swimming test [27] . Phenotyping was performed by two lab colleagues independently . A total of 183 N2 mutants were produced from the two backcross mating protocols , consisting of 46 mutants from the C57 backcross , and 137 mutants from the C3H backcross . The N2 mutant mice from the C57 backcross presented a milder mutant phenotype when compared to the N2 mice of the C3H backcross , resulting in fewer identified mutant mice than would be predicted by Mendelian inheritance . Genomic DNA was analyzed with MapPairs Mouse microsatellite markers ( Invitrogen , http://www . invitrogen . com/ ) , developed at the Whitehead Institute/MIT Genome Center . Fifty-nine markers , polymorphic between C57 and C3H and evenly distributed over the entire mouse genome , were used for a low resolution genome scan using DNA from 21 randomly selected N2 ( C3H ) mice ( Table S1 ) . PCR products were electrophoresed on a 4% MetaPhor ( FMC BioProducts , http://www . fmc . com/ ) gel . After linkage to chromosome 9 was discovered , linkage was further confirmed by typing all N2 mutant mice . Finally , additional chromosome 9 markers were selected for high resolution mapping using N2 mutant mice with informative recombination breakpoints in chromosome 9 ( Table S1 ) . High and low resolution genome scan results were analyzed using ‘Pattern , ’ a PERL-based computer software developed for this purpose , designed to facilitate analyzing mouse genome scan results by generating a graphic representation of haplotypes and predicting potential linkage . Pattern uses PCR genome scan results as an input and generates graphic haplotypes and predicts potential linkage to specific genomic intervals . Potential linkage is assigned to a chromosomal marker ( available at http://www . tau . ac . il/karena/pattern . html ) . For each adjacent pair of markers ( vertically coupled table analysis ) , if the number of linked pairs ( at least one linked marker out of the pair ) through all the patterns is at least 5 times higher than the number of unlinked pairs ( both markers unlinked ) , potential linkage is indicated . For Tlc mapping , genome scan haplotypes were then carefully read and the chromosome of linkage , chromosome 9 , was identified . Brain RNA was isolated using RNeasy Mini kit ( Qiagen , http://www1 . qiagen . com/ ) and samples were treated with DNAase for removal of genomic DNA , according to the manufacturer's protocol . Mutation analysis was carried out by sequencing of PCR products of Myo6 cDNA or genomic DNA . Primers available as Supplementary Material ( Table S2 ) . A genotyping assay was developed specifically to identify the Tlc mutation . The G694 nucleotide resides in a sequence that mimics the BclI restriction enzyme recognition site with only one mismatch . Genomic DNA was amplified with a reverse primer that contains a 3′ mismatch to artificially create a BclI restriction enzyme recognition site in the wild type DNA . The Tlc G694T mutation eliminates this newly formed restriction enzyme recognition site . Digesting the PCR product with BclI generates a single band of 118 bp in Tlc/Tlc mice , two bands of 32 bp and 86 bp in wild type mice and three bands of 118 , 86 and 32 bp in Tlc/+ mice . Forward primer 5′-CAATATTATTGTTATTCAAGGATTTTTTTTG-3′ , reverse primer 5′ AAATTAACAATACCTTCAACAATTCTATG3′ . In order to obtain an extensive multiple sequence alignment , known myosin VI proteins were used to build a hidden Markov model ( HMM ) profile ( using the HMMER package , version 2 . 0; http://hmmer . janelia . org/; [44] , [45] , which was then used to search for similar sequences in the SWISSPROT database ( Release 48; http://www . expasy . ch/cgi-bin/sprot-search-ful; [46] ) . Thereafter , redundant sequences were removed using a 90% identity threshold limit . The resulting 77 sequences were multiply-aligned using MUSCLE [47] ( http://phylogenomics . berkeley . edu/cgi-bin/muscle/input_muscle . py ) . The multiple sequence alignment was then used as input to the ConSeq and ConSurf web-servers [48]–[50] ( http://conseq . tau . ac . il/ and http://consurf . tau . ac . il/ ) . Since the mouse myosin VI structure has not been solved yet , the NEST program [51] was used to create a homology model . The structure used as template was that of the head domain of porcine myosin VI ( PDB ID: 2bkh ) [52] . Since the porcine myosin VI protein shares high sequence similarity ( 89% identity ) with the mouse myosin VI head region , it is likely to be structurally similar and serves as a potentially good template . Immunofluorescence of whole mount cochleae harvested from 3 wild type mice at P6 , 3 wild type mice at P40 and 8 P70 Tailchaser mice ( 4 homozygotes and 4 littermate heterozygotes ) was performed as described [53] . For protein detection , samples were incubated with a myosin VI antibody at a dilution of 1∶400 ( Proteus Biosciences , http://www . proteus-biosciences . com/ ) , and a monoclonal anti α-tubulin antibody ( mouse ) at a dilution of 1∶200 ( Sigma-Aldrich , http://www . sigmaaldrich . com/ ) . DAPI ( 4′ , 6-Diamidine-2′-phenylindole dihydrochloride , Roche Applied Science , http://www . roche-applied-science . com/ ) was used to stain the nuclei in a working solution of 2mg/ml dissolved in water . The secondary antibody ( Alexa 488 , 1∶1000 ) and rhodamine conjugated phalloidin were from Invitrogen-Molecular Probes . Images were acquired with a Zeiss LSM510 META confocal microscope , equipped with 63x 1 . 4NA objective , and processed with LSM Image Browser Rel . 4 . 2 and Adobe Photoshop CS2 . Pixel intensity analyses were performed using Image J software on images acquired at the same settings of the microscope , with background subtracted . Statistical analyses were performed with Xcel software , using the two-tailed test . ARPE-19 cells [54] were grown at 37°C with 5% CO2 in DMEM-F12 with 10% FBS , fungizone and glutamine and transfected with GFP-tagged myosin- VI constructs as described [33] . HGFP-M6 ( D179Y ) was created using GFP-M6+LI ( full length human myosin-VI containing both the small and large tail domain splice insertions fused to GFP [35] using the Quick Change XL Site-Directed Mutagenesis Kit ( Stratagene , http://www . stratagene . com/ ) . PGFP-M6tail is a dominant negative construct containing the tail domain of porcine myosin-VI fused to GFP [33] . The primers used were as follows: Forward: 5′-GGAACAGGTCAAGATATTTATGACAGAATTGTTGAAGC-3′ and Reverse: 3′-GCTTCAACAATTCTGTCATAAATATCTTGACCTGTTCC-5′ . The mutation in the primer set is underlined . Isolated clones were sequenced to verify that the point mutation was incorporated and that no other mutations were introduced by PCR . Coverslip grown cells were processed for immunofluorescence in six-well plates [55] . Affinity-purified rabbit anti-GIPC domain antibodies were used as described [35] . All fixed samples were observed with a Leica DMR upright light microscope fitted with a Hamamatsu ORCA 10bit CCD Digital Camera [33] . Plots and statistical analyses of vesicle properties were generated with Microsoft Excel 2000 . Pulse-chase and steady-state uptakes of rhodamine-conjugated transferrin were undertaken and quantified as described [33] . Quantification of steady-state R-Tsfn uptake to the pericentriolar endosome and quantification of percent overlap between GFP-tagged constructs , R-Tsfn and endocytic markers was carried out as described [33] . Error bars represent the standard deviation from three experiments . A total of 78 Tailchaser mice including 14 E18 . 5 ( 5 littermate controls , 4 homozygotes and 9 heterozygotes ) , 32 P1 ( 14 littermate controls , 11 homozygotes and 7 heterozygotes ) , 8 P7 ( 2 littermate controls , 1 homozygote and 5 heterozygotes ) , 24 P21 ( 8 littermate controls , 8 homozygotes and 8 heterozygotes ) and a total of 12 P6 Snell's waltzer mice ( 6 homozygotes and 6 heterozygotes ) were investigated by SEM . Freshly isolated cochleae were locally perfused through oval and round windows with 4% glutaraldehyde in 0 . 07 M ( or 0 . 1 M for adult cochleae ) sodium cacodylate buffer pH 7 . 4 with 3 mM CaCl2 and then fixed for 3 h at RT in the same fixative . Samples were then carefully washed in PBS and processed with the OTOTO method ( osmium tetroxide/thiocarbohydrazide ) adapted from Hunter-Duvar [56] , dehydrated in ethanol , critical point dried ( CPD 20 , Bal-Ted , http://www . bal-tec . com/ ) , mounted on stubs with conductive paint and viewed with a Hitachi FE S-4800 Scanning Electron Microscope operated at 3–5 kV . Samples were viewed without coating or coated with 2 nm of gold ( Sputter coater SCD 050 , Bal-Tec ) . Postacquisition image analyses were performed using Adobe Photoshop CS2 and NIH Image softwares ( http://rsb . info . nih . gov/nih-image/ ) . Myosin VI “zippered” dimer constructs ( with and without the D179Y mutation ) were created by truncation at Arg-994 ( NP_999186 myosin VI [Sus scrofa] ) , followed by a leucine zipper ( GCN4 [57] ) to ensure dimerization . The myosin VI-S1 ( monomer ) constructs ( with and without the D179Y mutation ) were created by truncation at amino acid 839 . In all cases , a Flag tag was appended to the C-terminus to facilitate purification [36] . These constructs were used to create a baculovirus for expression in SF9 cells [36] . ATPase assays were performed as previously described [58] . | Human deafness is extremely heterogeneous , with mutations in over 50 genes known to be associated with this common form of sensory loss . Among them , mutations in five myosins are associated with human hereditary hearing impairment , demonstrating that this family of proteins is essential for the proper function of the inner ear . Myosins , motor proteins found in eukaryotic cells , are responsible for actin-based motility . Composed of a motor domain and a tail , the former binds filamentous actin and uses ATP hydrolysis to generate force and move along the filaments , while the latter binds to cargos in the cell . Myosin VI is unique among myosins due to its movement along actin towards the minus or pointed end , rather than the positive or barbed end . Mutations in this myosin are associated with human deafness . Much of our information regarding myosin VI comes from studies in cell culture or mouse mutants with mutations leading to deafness . Here , we describe a deaf mouse mutant , Tailchaser , with a mutation in myosin VI . Our data describe new functions for myosin VI in the hair cells of the inner ear , showing how alterations in this motor can lead to a human sensory disorder . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/gene",
"discovery",
"neuroscience/sensory",
"systems",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"computational",
"biology/evolutionary",
"modeling",
"genetics",
"and",
"genomics/d... | 2008 | A Myo6 Mutation Destroys Coordination between the Myosin Heads, Revealing New Functions of Myosin VI in the Stereocilia of Mammalian Inner Ear Hair Cells |
Regulatory proteins can bind to different sets of genomic targets in various cell types or conditions . To reliably characterize such condition-specific regulatory binding we introduce MultiGPS , an integrated machine learning approach for the analysis of multiple related ChIP-seq experiments . MultiGPS is based on a generalized Expectation Maximization framework that shares information across multiple experiments for binding event discovery . We demonstrate that our framework enables the simultaneous modeling of sparse condition-specific binding changes , sequence dependence , and replicate-specific noise sources . MultiGPS encourages consistency in reported binding event locations across multiple-condition ChIP-seq datasets and provides accurate estimation of ChIP enrichment levels at each event . MultiGPS's multi-experiment modeling approach thus provides a reliable platform for detecting differential binding enrichment across experimental conditions . We demonstrate the advantages of MultiGPS with an analysis of Cdx2 binding in three distinct developmental contexts . By accurately characterizing condition-specific Cdx2 binding , MultiGPS enables novel insight into the mechanistic basis of Cdx2 site selectivity . Specifically , the condition-specific Cdx2 sites characterized by MultiGPS are highly associated with pre-existing genomic context , suggesting that such sites are pre-determined by cell-specific regulatory architecture . However , MultiGPS-defined condition-independent sites are not predicted by pre-existing regulatory signals , suggesting that Cdx2 can bind to a subset of locations regardless of genomic environment . A summary of this paper appears in the proceedings of the RECOMB 2014 conference , April 2–5 .
Profiling the activity of regulatory proteins in multiple cell types is important for understanding cellular function , as a single regulator can bind to distinct sets of genomic targets depending on the cellular context in which it is expressed . Characterizing the determinants of such binding specificity is key to understanding how a single regulator can play multiple roles during development and other dynamic cellular processes . For example , pre-existing genomic context such as chromatin accessibility or the binding of other regulators may determine the binding of some developmental transcription factors ( TFs ) [1]–[3] , while other ‘pioneer’ TFs may find their binding targets independently of the established chromatin state [4] , [5] . Here we introduce MultiGPS , an integrated machine learning approach for the analysis of condition-specific binding events from multi-condition ChIP-seq data . MultiGPS performs binding event analysis across multiple conditions , sharing information across conditions to produce accurate joint binding estimates while simultaneously allowing for condition-specific binding events . MultiGPS employs a flexible framework for incorporating prior information into binding event discovery , allowing models of joint binding and sequence dependence to be used . The novel multi-experiment modeling approach of MultiGPS identifies the read enrichment associated with binding events that are bound in specific conditions , enabling principled methods of discovering differential binding [6]–[9] . Most current strategies for defining consistent ChIP-seq binding event locations across multiple experiments either analyze each experiment independently or pool reads for analysis . For example , the ENCODE2 project used standard ChIP-seq event finders on each experiment independently , and then merged event locations across experiments using a fixed-sized window to define event identity [10] , [11] . Related methods specifically developed for multi-condition ChIP-seq analysis require that binding events be called in each condition individually as a preprocessing step , then apply statistical models to matched regions to detect differential effects [9] , [12] . Other multi-condition approaches focus on ChIP-seq signals arising from broad regions of enrichment , such as histone modifications . These methods instead search for larger genomic regions where coverage patterns differ across experiments [8] , [13]–[15] . In contrast , MultiGPS uses a joint multi-experiment model that considers the read data from all experiments to produce accurate location estimates of punctate binding events . Approaches that first identify binding events and then attempt to merge locations across conditions may inappropriately combine distinct binding events that happen to be located within the same window . In genomic regions with a high density of binding events , the problem of matching sites across conditions is difficult and may lead to erroneous comparisons between binding strengths . Furthermore , the experiment-by-experiment event calling approach fails to use the full power of the experimental data when a large fraction of binding events are shared across conditions . An alternative method is to pool ChIP-seq reads from all experiments and then use a single event finding run to yield a consistent set of binding event locations that can be subsequently quantified in each individual experiment . However , this pooling approach may not discover weak condition-specific binding locations that are swamped by noise from other experiments in the pooled set of reads . Additionally , applying a single detection threshold in the pooled read set may bias the binding event calls to experiments that had higher sequencing coverage , better antibody batches , or fewer technical sources of error . Similarly , varying experimental parameters such as the fragmentation distribution could render the pooled read dataset harder to analyze by algorithms that assume a single , consistent set of experimental properties . MultiGPS combines the theoretical benefits of pooling and separate ChIP-seq experimental analysis by using a Bayesian prior to couple the analysis of independent experiments together . This multi-experiment model is one aspect of a novel modeling approach that enables external sources of information to be included as priors in binding event identification ( see Methods ) . In this work , we use the following priors , while recognizing that other directions are also possible: MultiGPS detects binding events independently in each experiment in each step of its iterative optimization , allowing it to model experiment-specific parameters such as the distribution of reads around binding events and the properties of background noise . The iterative optimization procedure analyzes each experimental condition in turn , using binding event locations from other experiments to form an inter-experiment prior term for a single experiment optimization . MultiGPS therefore encourages the base locations of binding events to align across experiments when appropriate , and automatically produces coherent sets of binding events that are linked across experiments without any potentially noisy windowed analysis . To our knowledge , MultiGPS is the first ChIP-seq analysis approach that uses read data from multiple experiments in a joint and fully integrated method for identifying consistent and accurate binding event locations . As a case study of our framework's sensitive and accurate multi-condition analysis , we applied MultiGPS to Cdx2 binding data in three developmentally relevant cellular contexts and found that condition-specific Cdx2 binding events are predicted by preexisting chromatin state . Surprisingly , condition-independent Cdx2 binding events that are bound in multiple contexts do not appear to be predetermined by accessibility or other chromatin signatures , and instead may be predicted on the basis of cognate motif occurrence . Our results suggest that Cdx2 can act as a pioneer factor at a subset of sites , while also being influenced by preexisting genomic context at other sites . Therefore , our results have consequences for understanding where TFs will bind when introduced into an established regulatory state during development , or when induced artificially during cellular programming techniques .
We find that MultiGPS's inter-experiment and motif priors encourage binding location consistency on CTCF biological replicate experiments . The binding events that are called in both CTCF replicates should by definition be located at the same base location . As we can see in Figure 1a , when MultiGPS is run without either prior , predicted binding events do not typically align to each other or to cognate motif instances . Each prior alone makes a significant , though incomplete , improvement in binding event accuracy ( Figure 1b–c ) . The inter-experiment prior is able to significantly improve the distance to the nearest motif when compared to sites identified without any positional priors ( p<5×10−5 , Mann-Whitney U test comparing binned distance to nearest motif match ) . The motif prior significantly improves the distance to the nearest site in another experiment ( p<1×10−12 , Mann-Whitney U test comparing binned distance to the nearest event in another experiment ) . In these two comparisons , we used information sources not considered by the prior as validation ( motif distance for the inter-experiment prior and inter-experiment distance for the motif prior ) . The use of both priors together fully utilizes available sequence and multi-experiment information and allows almost all binding events in this example to be aligned to consistent ( typically motif-associated ) locations ( Figure 1d ) . These comparisons are not meant as absolute performance assessments for the MultiGPS modeling approach , but instead as relative measurements of the benefit of using additional types of prior information within a single modeling framework . MultiGPS facilitates the detection of differential binding events by accurately quantifying read count levels associated with each binding event in each analyzed experiment . Since at present no ChIP-seq datasets exist for which absolute binding levels are known across multiple conditions , we generated simulated ChIP-seq datasets to test the relative performance of MultiGPS in defining differential binding events . In our simulated data , the distribution of reads at binding events mirrors the properties of real ChIP-seq datasets ( see Methods ) . A subset of binding events is chosen to be differentially enriched across conditions , and while we chose to set the absolute level of differential enrichment to be constant at all differential events ( 4-fold in Figure 2 , 8-fold in Figure S1 ) , simulated sampling noise leads to a wide array of apparent fold differences ( Figure 2a , blue dots ) . Using the simulated data , we compared MultiGPS with other approaches for determining differential binding events . We used MultiGPS ( without the motif prior since no sequence information was used to simulate the data ) , MultiGPS in single-condition mode ( i . e . without using either inter-experiment or motif priors ) , and the single-condition event finders MACS [18] and SISSRs [19] to predict binding events in each simulated condition . All methods made comparable numbers of binding event predictions in each dataset ( Figure S2 ) . For the methods other than MultiGPS , differential binding events were defined using: a ) binding event list comparison , where differential binding events are those that are detected in one condition and no binding event is detected within 200 bp in the other condition; b ) using the software DBChIP [9]; or c ) by counting reads that occur within the enriched regions and inputting the resulting tables into edgeR [6] ( using the same parameters as used by edgeR within MultiGPS ) . The results illustrate the problems with defining differentially bound events using binding event list comparison . Regardless of which event finding method was used to provide input binding events , list comparisons have poor sensitivity when predicting differentially bound events with higher mean read counts ( Figure 2b , dashed lines ) . Such events are more likely to be detected in both conditions and hence would be treated as non-differential binding events regardless of quantitative differences in ChIP enrichment levels . Conversely , binding event list comparisons have low specificity when predicting differentially bound events with lower mean read counts ( Figure 2c , dashed lines ) . Low enrichment binding events may have read counts that are just above a binding event detection threshold in one condition , and just below in another , even if there is no significant quantitative difference in the underlying ChIP enrichment levels . Such events would appear as false positive differential binding event predictions according to the binding event list comparison approach . In contrast , approaches that test differential binding using statistical analyses of read count tables have uniformly high specificity across our test datasets ( Figure 2c , solid lines ) . These methods also have higher sensitivity when predicting differential binding events with higher mean read counts ( Figure 2b , solid lines ) or involving greater absolute differences in binding levels ( Figure S1b , solid lines ) . EdgeR attains the highest overall sensitivity using the read count tables generated by MultiGPS , thus illustrating the advantages of MultiGPS' probabilistic approach to quantifying read enrichment at binding sites across conditions . MultiGPS models experiment-specific parameters such as the distribution of reads around binding events and the properties of background noise . To investigate whether these parameters yield improved quantification of binding event ChIP enrichment , we ran the complete MultiGPS model on 14 ChIP-seq experiment sets in which the ENCODE2 project has performed replicated ChIP-seq of a given protein in all three human Tier 1 cell lines . While no gold standard exists for measuring the accuracy of ChIP-enrichment quantification , we reasoned that accurate quantification estimates should be correlated across biological replicate experiments . For each of the 14 experiment sets , MultiGPS yields per-replicate estimates of binding enrichment for binding events discovered in any cell line . We compared these values to those produced by the widely used approaches of counting read occurrences in a window around the binding event locations ( here we use a 400 bp window centered on the MultiGPS-defined binding event locations ) , or by using the peak heights defined by MACS [18] analyses of the same data . Quantified read counts were compared across biological replicate pairs using Spearman's rank correlation , a nonparametric assessment of statistical dependence that makes no distributional assumptions that could artificially favor one model over another . Note that MACS does not produce per-replicate read counts or peak heights at each event , and so to compare MultiGPS with MACS we ran MACS on each replicate separately and compared read counts and heights at only those binding events detected in both replicates by MACS and MultiGPS . Read counts at these reproducibly detected binding events may be more highly correlated than read counts associated with the wider sets of binding events tested in the comparison between MultiGPS estimates and windowed read counts . As shown in Figure 3 , MultiGPS improves the cross-replicate correlation of binding event quantification estimates in most tested datasets , implying that MultiGPS has reduced the effects of inter-replicate noise in comparison to the window counting approaches . We expect that reducing the degree of over-dispersion between replicates will yield greater sensitivity in detecting significant differences between conditions . Indeed , in all 14 tested datasets we find substantially greater numbers of statistically significant differentially enriched binding events between cell lines when we run edgeR [6] on the MultiGPS quantification table as opposed to the table of read counts produced by the window approach ( Table S1 ) . Therefore , MultiGPS improves the quantification of binding event ChIP-enrichment and the detection of condition-specific binding events . To demonstrate the ability of MultiGPS to analyze biologically relevant condition-specific binding events , we examined if MultiGPS improves upon the independent analysis of experiments when identifying Cdx2 events in multiple conditions . Cdx2 is a mammalian caudal-type homeobox protein that plays a key role in regulating the development of diverse tissue types . For example , Cdx2 is a master regulator of the intestinal lineage when expressed in endoderm [20] , and also plays a key role in defining caudal motor neuron fate when expressed in motor neuron progenitors ( pMNs ) [21] . In addition , over-expression of Cdx2 in embryonic stem ( ES ) cells forces cells to differentiate into the trophectoderm lineage [22] , [23] . We thus wanted to elucidate how Cdx2 performs its different regulatory functions in these three developmental contexts . Does it bind to the same genomic targets in all cell types , or does it bind distinct targets in each context ? If the latter , how is such specificity achieved ? To determine the context-dependent binding activity of Cdx2 , we performed ChIP-seq analysis of Cdx2 after it was over-expressed in ES cells , endoderm , and pMNs . We call these cell types after Doxycycline-dependent Cdx2 induction ES+Dox Cdx2 , endoderm+Dox Cdx2 , and pMN+Dox Cdx2 , respectively . Since Cdx2 is not natively expressed in any of these three cell types , our experiments provide a useful model of how a transcription factor responds to a new cellular environment . We found that MultiGPS outperformed an independent binding event analysis ( i . e . using independent runs of MultiGPS without the use of priors ) on the three Cdx2 conditions using a binding event list comparison approach to determine differentially bound sites . While this is a common approach in the literature , it leads to highly misleading results . As can be seen in Figure 4 , the binding event list comparison suggests that 95% of pMN+Dox sites are not bound in ES+Dox cells . However , the apparent degree of differential binding is largely caused by the disparity in the total numbers of binding events predicted in each condition ( 3 , 704 in ES+Dox and 36 , 651 in pMN+Dox ) . The difference in the total number of events is in turn caused by differences in read coverage between the conditions and the thresholds employed to determine bound events . In addition , the binding event list comparison approach may miss differences at events when the level of ChIP enrichment varies significantly between conditions . To perform a more principled analysis of Cdx2 differential binding , we analyzed the ChIP-seq data collection using MultiGPS ( Table 1 ) . With the coupled MultiGPS method only 24% of all pMN+Dox Cdx2 binding events are significantly differentially enriched in pMN+Dox cells compared with ES+Dox cells ( p<10−3 ) , while 37% of all ES+Dox Cdx2 binding events are significantly differentially enriched in ES+Dox cells compared with pMN+Dox ( p<10−3 ) . Since MultiGPS identifies a large proportion of condition-specific Cdx2 binding events without finding any evidence for a corresponding change in Cdx2's DNA-binding preference , we asked whether ES cell genomic context could predict the observed condition-specific binding of Cdx2 after induction . To answer this question , we examined the ES genomic patterns at the locations of Cdx2 sites that are significantly enriched in ES+Dox cells according to MultiGPS . Interestingly , we found that ES+Dox-specific Cdx2 sites are enriched for ES signatures of chromatin accessibility ( DNaseI hypersensitivity ) , enhancers ( H3K4me1 and H3K27ac ChIP-seq ) , and TF binding ( Oct4 , Sox2 , and Nanog ChIP-seq ) , but not active transcription ( H3K4me3 ChIP-seq ) ( Figure 5 ) . Conversely , pMN+Dox-specific Cdx2 sites and endoderm+Dox-specific Cdx2 sites show no enrichment for these ES cell chromatin signatures ( Figure 5 & Figure S3 ) . To more rigorously test the capacity of ES cell genomic context to predict ES+Dox-specific Cdx2 binding events , we trained support vector machines ( SVMs ) to classify Cdx2 binding events vs . unbound Cdx2 motif instances using the read count information from a collection of 55 ES experiments ( 2 DNaseI-seq , 13 histone modification ChIP-seq , 35 TF , co-activator and chromatin modifier ChIP-seq , and 5 Pol2 ChIP-seq experiments ) . Cross-validation was used to generate disjoint training and test sets ( see Methods ) . Our SVMs discriminate ES+Dox-specific Cdx2 sites from unbound sites with an area under true-positive vs . false-positive curve ( AUC ) of 0 . 95–0 . 96 , suggesting that the pre-existing genomic context in ES cells is highly predictive of future Cdx2 binding . Conversely , our SVMs are unable to discriminate pMN+Dox-specific Cdx2 sites from unbound Cdx2 motif instances using ES genomic context ( AUC = 0 . 63 , Figure 6 ) . Our results therefore suggest that condition-specific Cdx2 binding events are more likely to be located in genomic regions that already displayed regulatory activity or accessibility before Cdx2 expression was induced . Since condition-specific Cdx2 binding events appear highly correlated with immediately pre-existing genomic context , we reasoned that the condition-independent Cdx2 sites that are bound in multiple conditions might also display the same associations . For example , Cdx2 sites that are bound in two conditions may represent locations that happened to have pre-existing regulatory activity or accessibility in both conditions . Surprisingly , the Cdx2 sites bound in both ES+Dox and pMN+Dox conditions are not enriched for accessibility , enhancer chromatin marks , or TF binding in ES cells ( Figure 5 ) . Furthermore , SVMs trained as before are unable to discriminate between these shared Cdx2 sites and unbound motif instances using ES genomic context information ( AUC = 0 . 61 , Figure 6 ) . These results suggest that the condition-independent Cdx2 sites are not determined by pre-existing genomic context , in contrast with the condition-specific sites . Given that the condition-independent Cdx2 sites do not seem to have any distinguishing chromatin features before Cdx2 induction , we asked how Cdx2 recognizes these sites regardless of genomic context . We hypothesized that such sites may have sequence features that enable condition-independent binding . To test this hypothesis , we trained SVMs to discriminate condition-independent Cdx2 sites from condition-specific Cdx2 sites using only 4-mer word frequencies in 200 bp windows around the sites . Surprisingly , even these crude sequence features were sufficient to discriminate between the two types of sites ( AUC = 0 . 89–0 . 92 , Figure 7a ) , suggesting that some sites contain sequence information that enables condition-independent Cdx2 binding . We next used the discriminative motif finders DEME and DECOD [24] , [25] to determine which sequence motifs discriminate between Cdx2 site types . Interestingly , both tools returned the primary Cdx2 motif as being the most discriminative , even though most condition-specific and condition-independent sites contain instances of the same primary motif . This apparent contradiction is resolved by considering features of the motif instances in each set of Cdx2 sites . SVMs trained with just three simple primary Cdx2 motif-related metrics – the maximum motif score in the 200 bp window around sites , the number of motif instances above a threshold , and a score that integrates motif scores across the entire 200 bp window [26] – were able to discriminate between condition-independent and condition-specific sites with reasonable accuracy ( AUC = 0 . 81 , Figure 7b ) . In other words , the strength and multiplicity of motif instances are somewhat predictive of condition-independent Cdx2 binding . Taken together , our results suggest that sequence information allows Cdx2 to act as a pioneer TF at some sites , overriding the lack of pre-existing accessibility or chromatin markers .
MultiGPS provides a principled platform for the analysis of differential protein-DNA binding across multiple experimental conditions by preferring consistent binding locations across related experiments while also modeling condition-specific experimental parameters . Rather than treating reads from all experiments as equivalent , MultiGPS models experiment-specific read distributions around binding events . MultiGPS can thus correctly analyze collections of related ChIP experiments that were performed according to different protocols such as mixtures of related ChIP-seq and ChIP-exo [27] experiments . As demonstrated above , MultiGPS improves the quantification of ChIP enrichment at binding events in comparison with the typically used window-counting approaches , thus enabling more sensitive analyses of differential binding enrichment between conditions . Since MultiGPS prefers but does not force binding events to align across experiments , it may also be used to study possible forms of differential binding activity that we did not illustrate . For instance , it may be of interest to examine locations where the underlying read evidence overrides the MultiGPS inter-experiment prior , resulting in differing reported binding locations across experiments . Such locations may represent shifts in binding location between conditions , which may be useful for studies of nucleosome positioning or regulators that might bind alternate nearby locations in different conditions . We demonstrated that MultiGPS can characterize condition-specific binding and then used MultiGPS to characterize the nature of both condition-specific and condition-independent binding of Cdx2 . Our results suggest that many condition-specific Cdx2 binding events are located in regions that had pre-existing regulatory activity , thus agreeing with hypotheses proposed to explain the observed binding of other developmental TFs [1]–[3] . However , Cdx2 also appears to act as a ‘pioneer’ at a subset of sites that are bound condition-independently . Our analysis suggests that such sites on average contain stronger and more frequent Cdx2 motif instances than condition-specific sites , thus suggesting a possible mechanism by which condition-independent sites can be bound regardless of preexisting genomic context . These findings also accord with our recent demonstration that TF combinations can override pre-existing cellular state to synergistically bind composite motifs during motor neuron programming [28] , perhaps pointing to a deeper relationship between sequence information and ‘pioneer’ binding activity .
In our previously described GPS [16] and GEM [17] approaches to binding event detection , ChIP sequencing data are modeled as being generated by a mixture of binding events along the genome , and an Expectation Maximization ( EM ) learning scheme is used to probabilistically assign sequencing reads to binding event locations . The assignment of reads is achieved via an empirically estimated multinomial distribution , Pr ( rn|x ) , which gives the probability of observing read rn from a binding event located at genomic coordinate x . Conceptually , every base position is treated as a potential binding event , although the use of a sparse prior [29] has the effect of allowing only a small subset of these potential binding events to take responsibility for observed reads and survive the EM training process . In MultiGPS , we decouple the relationship between a binding event's index and its spatial ( genomic ) location . Specifically , we introduce a vector of component locations μ where μj is the genomic location of event j . We initialize a large number of potential events , M , such that the events are evenly spaced in 30 bp intervals along the genome . Note , however , that the use of a sparse prior will again result in only a subset of events remaining active in the model after training ( i . e . components having mixing probability πj>0; see MAP estimation of π below ) . In the new mixture model , the likelihood of observing the N total ChIP read locations r is given by:where Pr ( rn|μj ) is the distribution over ChIP-seq read positions conditioned on membership in a binding event at location μj . This distribution is initialized to a strand-specific shape typical of many ChIP-seq datasets ( see Figure S5 ) , and is iteratively re-estimated during EM training using the distribution of reads observed around high-confidence binding site locations . The above expression calculates the observed data likelihood of a mixture model by taking the product over all reads , where each read averages over each possible binding event that may have caused it . This extension of the model allows us to apply prior knowledge directly to the positions of the binding events ( μ ) , without affecting the binding event strength estimation or the sparsity-promoting prior , which continues to act on raw expected read counts . We introduce a Bernoulli prior over each genomic location where each element ki of the parameter k corresponds to the probability that location i is a binding event ( that is , i μ ) . This prior assumes that there can be only one or zero binding events at a single position and that binding positions are selected independently along the genome according to this weighting . The prior assigns likelihoods to a set of binding events on a genome of size L as follows: As in the original framework , the latent assignments of reads to binding events are represented by the vector z . The complete-data log posterior can thus be derived as follows:Here , C is a normalization constant that does not involve any of the terms to be optimized . It can be seen that the overall binding event sparsity-inducing negative Dirichlet prior α acts only on the mixing probabilities π , which controls the total fraction of reads assigned to each binding event , and the positional prior k acts only on the binding event locations μ . Therefore , the E-step that calculates the relative responsibility of each binding event in generating each read is unchanged from our original framework , following standard mixture model approaches:Furthermore , the maximum a posteriori probability ( MAP ) estimation of π is also unchanged:where Nj is the effective number of reads assigned to binding event j . The α parameter can thus be interpreted as the minimum number of ChIP-seq reads required to support a binding event remaining active in the mixture model . We set the value of α per experiment to be the maximum number of reads that would be expected to occur ( p>10−7 ) in a window equal to the effective range of the binding distribution should the experiment's reads be distributed uniformly along the mappable portion of the genome . We can estimate μ component-wise since it only participates in sums in the log likelihood . However , no closed form solution exists since the prior k has no parametric form . We can determine the MAP ( integer ) value of μj by simply enumerating over all possible values of μj . Specifically , the MAP value of μj is: . If the maximization step results in two components sharing the same location , they are combined in the next iteration of the algorithm . One practical use for the positional prior k is to bias the estimated binding locations towards biologically appropriate base positions . For example , a TF's position weight matrix scores along the genome can be directly encapsulated in k in the above framework . As described previously for our GEM approach [17] , we can estimate binding motifs from current estimates of binding locations , and reciprocally use those motifs as prior information to re-estimate binding event locations . Note that motif priors are incorporated quite differently in GEM and MultiGPS . In practice , MultiGPS uses MEME [30] to discover a set of over-represented motifs in the top 500 most enriched binding events ( 80 bp windows ) , chooses the motif with the highest true-positive vs . false-positive AUC for discriminating bound regions from random sequences ( if any motif AUC≥0 . 7 ) , and incorporates the genomic log-odds scores for that motif in the positional prior . Unlike our previously described approaches , MultiGPS incorporates an additional mixture component that explicitly models noise ( i . e . reads arising from nonspecific binding and independent of any binding event ) . Whereas binding component read distributions have approximately finite support ( and therefore only allow binding events to take responsibility for reads in their local vicinity ) , the noise component is defined as having a global distribution . The form of the noise distribution can be defined as uniform or can be parameterized using the read density observed in a control experiment . In the latter case , the shape of the noise distribution is defined by smoothing the control experiment's read counts using a 50 bp sliding window ( adding fractional pseudocounts to 50 bp windows that contain no control reads ) . For a more efficient and stable training process , some parameters in MultiGPS are re-estimated only periodically , including the form of the binding event read distribution , the noise component mixing probability ( πM+1 ) , and the binding motif position weight matrix . We can therefore think of MultiGPS as an instance of a generalized EM algorithm . Generalized EM algorithms increase the expected log likelihood in each M step without necessarily achieving a maximum in each iteration ( as in the original EM algorithm ) [31] . Convergence to a local optimum is guaranteed with generalized EM algorithms , as it is with the EM algorithm [31] . As with GPS and GEM , MultiGPS filters predicted binding events to require that their associated read counts are significantly enriched ( p<10−3 , Benjamini-Hochberg corrected Binomial test ) over the corresponding read count from an appropriately normalized control experiment , such as a mock-IP experiment . The control experiment normalization factors are estimated via regression on the read count ratios in 10 Kbp windows . Control read counts are associated with individual binding events via maximum likelihood assignments using the trained model ( i . e . assigning control reads to binding events without changing the π and μ parameters learned from the ChIP data ) . MultiGPS can be run in a multi-condition analysis mode by providing multiple input datasets and structured annotation as to how these datasets are related ( i . e . which datasets represent technical or biological replicates of others , which collections of datasets represent distinct experimental conditions , and which datasets serve as controls for others ) . MultiGPS then runs semi-independent mixture model training across all provided data . Since reads from distinct conditions are not pooled , MultiGPS can maintain condition-specific and replicate-specific parameters , including distinct binding event read distributions per replicate , distinct noise component read distributions and mixing probabilities per replicate , and distinct binding motifs per condition . However , the goal is to report binding event locations that are consistent across conditions . This is achieved using another form of prior information during the maximization of binding event locations μ . We motivate our approach by imagining a TF that binds to N locations in cellular condition A and N locations in cellular condition B . In typical analysis scenarios , the number of bound locations will be much fewer than the number of bases on the genome ( i . e . ) , and a non-zero set of S locations will be bound in both A and B conditions . We present the model for two conditions with a symmetric number of binding sites here for notational simplicity , but note that the same process can be applied to any number of conditions with more complex binding site sharing patterns . A schematic example ( not to scale ) of bound and unbound bases in two conditions as a fraction of the genome is shown in Figure 8 . Now , the distribution that generates binding positions is extended from the single-condition case of a Bernoulli distribution to a multivariate Bernoulli distribution . As suggested by the schematic in Figure 8 , this distribution generates a sample from { ( 0 , 0 ) , ( 0 , 1 ) , ( 1 , 0 ) , ( 1 , 1 ) } at each base in the genome , where each element in a sample corresponds to whether a binding site is present at that position in that condition . This generative model induces the following distribution over genome positions i with respect to binding site positions in conditions A and B:We parameterize the above distribution during each iteration of the MultiGPS algorithm by choosing appropriate values for N and S ( L is fixed , being the length of the genome ) . While N can be taken from MultiGPS' current estimate of the number of binding events in each condition , we do not typically know S . We therefore define S by setting the ratio S/N as described below . We need to know the contribution of the location prior in the optimization step for the binding site locations μ . For the multi-condition analysis , we jointly optimize two binding sites when they fall within 100 bp of each other ( range chosen empirically as the maximum range for which the inter-experiment prior will have an effect at most binding events , see Figure S7 ) . The model optimization step determines whether the two binding positions in question are separate ( and therefore two site-specific positions contribute to ) or shared ( and therefore one shared site contributes to ) . All other bases will be the same during this optimization since all other binding sites are fixed , and can be ignored in this step . Using the distribution above gives the following contribution to the prior :where in most experimental studies . If the binding events share a location across conditions , we choose the optimal shared position w by maximizing the expected complete-data log posterior ( with terms not affecting the minimization omitted ) as follows:Alternatively , the two binding component locations are independent , in which case the two positions are optimized independently: The decision to use the coupled or uncoupled estimate is based on which scenario yields the higher expected complete-data log posterior probability . Higher values of the ratio S/N encourage the coupling of nearby binding event locations across conditions by increasing with respect to ( see Figures S6 & S7 ) . In MultiGPS , we set the ratio S/N to be equal to 0 . 9 , although in practice we observe few differences in the proportion of aligned binding components when varying the ratio in the range 0 . 5<S/N<0 . 99 . This is because a number of nearby genomic locations give similar probabilities when maximizing μj ( Figure S7 ) , allowing the penalties associated with moving the components away from the optimal positions in each condition to be overridden by the positive-valued prior over a range of S/N ratios . Note , however , that MultiGPS will still prefer the uncoupled scenario in situations where the read evidence supports distinct binding locations across conditions . This behavior represents a data-driven joint analysis mode that weighs the statistical confidence given by the reads against prior knowledge of the experimental setup in a probabilistically optimal way . We also note that positional prior terms encapsulating per-condition TF position weight matrix scores can be accounted for in the μ maximization terms above in a manner analogous to that described in the previous section . MultiGPS can therefore account for both motif positional priors and the inter-experiment prior . Assessing all possible scenarios of coupled and uncoupled binding events during the update of each μj becomes prohibitive when analyzing more than two conditions . Therefore MultiGPS assesses a limited number of scenarios when updating μj in such cases: 1 ) event j is uncoupled across all conditions; 2 ) event j is coupled with a corresponding event in one other condition; or 3 ) event j is coupled with corresponding events in all other conditions . The scenario that yields the best overall likelihood is chosen . A table containing the replicate-specific read counts associated with each binding event is generated from the MAP-estimated responsibilities γ . MultiGPS uses the edgeR Bioconductor package [6] to detect differential ChIP enrichment between conditions from the read count table . We use edgeR's TMM method to calculate normalization factors , and the glmLRT method to calculate likelihood ratios . In the Cdx2 example described here , we used a fixed overdispersion parameter of 0 . 15 across all experiments , which results in a stricter definition of significant differential enrichment than the overdispersion parameters estimated by edgeR . To computationally simulate multi-condition ChIP-seq data , we defined a hypothetical system in which a protein has 20 , 000 binding events in the mouse genome ( version mm9 ) . The relative strengths of each of these binding events was drawn randomly from a distribution of relative read counts observed for Cdx2 binding events in our pMN ChIP-seq experiments . For two hypothetical experimental conditions , A & B , we randomly chose 20% of the binding events to be differentially enriched in condition A with respect to B , and we modify the relative binding event strengths of these sites such that they are 4-fold ( or 8-fold in separate simulations ) greater in condition A versus B . We similarly chose a non-overlapping 20% of binding events to be differentially enriched in condition B with respect to A . The binding events were placed along the genome in 10 Kbp intervals . We then generated 20 million read positions for each of two replicates in each of the two conditions . To reflect the typical signal-to-noise ratio observed in real ChIP-seq experiments , 95% of the read positions are spread randomly across the entire genome . The remaining reads ( averaging 1 million per replicate ) are distributed amongst the binding events according to the relative strength of the event in each relevant condition , and accounting for read sampling noise using a negative binomial distribution with an over-dispersion parameter of 0 . 1 . The MA plot in Figure 2a shows the log2 mean read count and log2 fold difference for each binding event in the simulated experiments . The position of generated reads with respect to the defined binding event location is drawn from a bimodal distribution typical of ChIP-seq binding sites ( Figure S5 ) . We ran the following binding event analysis methods on the simulated data: a ) MultiGPS on the entire dataset , using default parameters with the exception of turning off the use of sequence information and the motif prior ( since motif information was not used in generating the simulated data ) ; b ) MultiGPS without the inter-experiment prior or the motif prior on the entire dataset , which has the effect of calling binding events in each condition independently; c ) MACS [18] using default parameters on each condition independently , merging reads across replicates; and d ) SISSRs [19] using default parameters ( with the exception of using a p-value cutoff of 0 . 05 ) on each condition independently , merging reads across replicates . For binding event list comparison approaches , per-condition events were compared with each other using a 200 bp window . In other words , if an event prediction in one condition is located within 200 bp of an event prediction in the other condition , it is treated as being in the intersection of the binding event list comparison , and thus not differentially bound . EdgeR [6] was run either internally in MultiGPS ( as described above ) or , using the same parameters , on read count tables built by counting reads that overlap the peak regions found by MACS or SISSRs . We also ran DBChIP [9] using default parameters with the exception of an FDR threshold <0 . 01 and using the MACS peaks as inputs . Sensitivity and specificity in Figures 2 and S1 are defined by comparing predicted binding events to the positions of the simulated differential binding events using a 100 bp window . Support vector machines were trained using the libSVM [32] interface in Bioconductor ( e1071 ) . In all cases , classification accuracy was determined using a randomly selected held-out test set of 100 datapoints , and training of each SVM application was repeated 20 times ( using different held-out test sets each time ) to calculate average true-positive vs . false-positive AUC values . To train SVMs using ES chromatin state data , we first gathered 55 mouse ES ChIP-seq and DNaseI-seq experiments from a variety of sources [33]–[42] . We defined positive training sets from the top-most Cdx2 binding events for each condition-specific and condition-independent permutation ( up to a maximum of 4 , 000 binding events ) , and we also defined a negative training set of 10 , 000 matches to the Cdx2 cognate binding motif ( as defined by UniProbe [43] ) that were not bound by Cdx2 in any experiment . Reads were counted in 1 , 000 bp windows around each of the positive and negative locations for each of the 55 mouse ES experiments , and SVMs were trained on the resulting 55-dimensional vectors without any normalization . SVMs were trained on k-mer frequencies by enumerating the occurrences of each 4-mer ( accounting for reverse-complement redundancies ) in 200 bp windows around each of the top-most Cdx2 binding events for each condition-specific and condition-independent permutation ( up to a maximum of 4 , 000 binding events ) . Similarly , SVMs were also trained on three pieces of information from the same 200 bp windows: the maximum log-likelihood ratio score for the Cdx2 motif in the window; the number of matches to the motif in the window that score more than a 5% FDR threshold; and the probability of binding occupancy in the window [26] . An ES cell line harboring Dox-inducible Flag-tagged Cdx2 was generated as previously described [44] . Anti-Flag ChIP-seq experiments were performed as previously described [44] after 24 hours of Dox-induced expression of Cdx2 in the ES cells or in motor neuron progenitors or endoderm cells that were differentiated from the same ES cell line . Differentiation of the ES cells to pMN and endoderm lineages was also described previously [20] , [21] . Mock IP control experiments were performed using the same system . Sequenced ChIP-seq reads were mapped to the mm9 reference genome using Bowtie [45] . ChIP-seq data generated during this study were deposited in GEO under accession numbers GSE39433 and GSE39435 . MultiGPS is available as an open-source Java package , released under the MIT license , from: http://mahonylab . org/software and https://github . com/shaunmahony/seqcode . Simulated multiple condition ChIP-seq datasets are also available from the same webpage . A summary of this paper appears in the proceedings of the RECOMB 2014 conference , April 2–5 [46] . | Many proteins that regulate the activity of other genes do so by attaching to the genome at specific binding sites . The locations that a given regulatory protein will bind , and the strength or frequency of such binding at an individual location , can vary depending on the cell type . We can profile the locations that a protein binds in a particular cell type using an experimental method called ChIP-seq , followed by computational interpretation of the data . However , since the experimental data are typically noisy , it is often difficult to compare the computational analyses of ChIP-seq data across multiple experiments in order to understand any differences in binding that may occur in different cell types . In this paper , we present a new computational method named MultiGPS for simultaneously analyzing multiple related ChIP-seq experiments in an integrated manner . By analyzing all the data together in an appropriate way , we can gain a more accurate picture of where the profiled protein is binding to the genome , and we can more easily and reliably detect differences in protein binding across cell types . We demonstrate the MultiGPS software using a new analysis of the regulatory protein Cdx2 in three different developmental cell types . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"gene",
"regulation",
"molecular",
"genetics",
"biology",
"computational",
"biology"
] | 2014 | An Integrated Model of Multiple-Condition ChIP-Seq Data Reveals Predeterminants of Cdx2 Binding |
The concept of "protein-based inheritance" defines prions as epigenetic determinants that cause several heritable traits in eukaryotic microorganisms , such as Saccharomyces cerevisiae and Podospora anserina . Previously , we discovered a non-chromosomal factor , [NSI+] , which possesses the main features of yeast prions , including cytoplasmic infectivity , reversible curability , dominance , and non-Mendelian inheritance in meiosis . This factor causes omnipotent suppression of nonsense mutations in strains of S . cerevisiae bearing a deleted or modified Sup35 N-terminal domain . In this work , we identified protein determinants of [NSI+] using an original method of proteomic screening for prions . The suppression of nonsense mutations in [NSI+] strains is determined by the interaction between [SWI+] and [PIN+] prions . Using genetic and biochemical methods , we showed that [SWI+] is the key determinant of this nonsense suppression , whereas [PIN+] does not cause nonsense suppression by itself but strongly enhances the effect of [SWI+] . We demonstrated that interaction of [SWI+] and [PIN+] causes inactivation of SUP45 gene that leads to nonsense suppression . Our data show that prion interactions may cause heritable traits in Saccharomyces cerevisiae .
Prions are proteins that convert between structurally distinct states , of which one or more is transmissible [1] . Prion formation leads to DNA-independent changes in heritable traits in microorganisms such as Saccharomyces cerevisiae and Podospora anserina [2 , 3] . For example , [PSI+] and [ISP+] prions , whose structural proteins are Sup35 and Sfp1 , respectively , modulate nonsense suppression [2 , 4]; Swi1 in its prion state , [SWI+] , causes a partial loss of function in utilizing non-glucose sugars and completely abolish yeast multicellularity [5 , 6] , while Ure2 in its [URE3] form changes nitrogen catabolism [2] . The HET-s protein of P . anserina in prion state determine heterokaryon incompatibility corresponding to a cell death reaction , an event which occurs upon fusion of genetically distinct strains [7] . Thus , prion formation changes heritable information encoded at the protein level . There are a number of works dedicated to the study of prion interactions , but currently we only know that pre-existing prions may facilitate the induction or elimination of other prions . It has been shown that pre-existing prions , such as [PIN+] or [SWI+] , are required for induction but not for maintenance of [PSI+] [5 , 8–11] . Excluding Mod5 [12] , all yeast prion proteins that form amyloid-like aggregates contain similar prion-forming regions rich in glutamine ( Q ) and/or asparagine ( N ) residues [13 , 14] . According to the cross-seeding model , pre-existing aggregates of one prion serve as the conformational template for newly forming prions [15] . Aggregates of Rnq1 or Swi1 in [PIN+] or [SWI+] strains , respectively , colocalize with overexpressing Sup35 only at the early stage of the initiation of [PSI+] prion formation , but at the latter stages the aggregates of these proteins do not colocalize [16] . Coexisting conformers of different prions do not physically interact probably because their conformations are spatially distinct . Some prions exhibit antagonistic relationships . For instance , the presence of [URE3] leads to the elimination of [PSI+] [17] . In this paper , we showed that coexisting prions may genetically interact , and this interaction causes heritable traits in Saccharomyces cerevisiae . Previously , we described [NSI+] ( Nonsense Suppression Inducer ) prion factor [18] . [NSI+] was shown to suppress the ade1-14UGA and trp1-289UAG nonsense alleles in background of modified Sup35 variants with decreased functional activity [18 , 19] . We demonstrated that the nonsense suppressor phenotype of [NSI+] cells , i . e . , their growth on–Ade or–Trp synthetic media , is caused by defects in translation termination [19 , 20] . This factor does not depend on [PSI+] prion , because [NSI+] phenotype is manifested in the strains containing deletion of N-terminal prion-forming domain of Sup35 . ” Additionally , [NSI+] cells exhibit growth defects on media containing galactose or glycerol as the sole carbon source [19] . Like known yeast prions , [NSI+] shows reversible curability , non-Mendelian inheritance , and cytoplasmic infectivity . Deletion of the chaperone Hsp104 or its inactivation by Guanidine-Hydrochloride ( GuHCl ) causes elimination of [NSI+] [18] . We previously used large-scale overexpression screens to reveal the genes affecting [NSI+] manifestation , but we did not identify the structural gene of this factor [19 , 21] . Recently we developed the proteomic method for identification of yeast prion proteins that form amyloid-like aggregates resistant to treatment with ionic detergents [22] . Here , using this method we demonstrated that the nonsense suppression in the [NSI+] strain is a result of interaction between [PIN+] and [SWI+] prions . We showed that prion inactivation of Swi1 protein decreases the expression of the translation termination factor eRF1 , and [PIN+] enhances this effect .
To identify the proteins whose prion conversion determines the [NSI+] phenotype , we used a method for proteomic screening and identification of amyloid proteins ( PSIA ) which we had previously developed and successfully applied for identification of yeast prions [22] . This method is based on the resistance of prion aggregates to treatment with ionic detergents such as sodium dodecyl sulfate ( SDS ) . To reveal prion proteins that determine the manifestation of the [NSI+] factor , we carried out a comparative analysis of proteins forming SDS-resistant aggregates in the 1-1-D931 [NSI+] and 1-1-1-D931 [nsi-] isogenic strains . Proteins forming aggregates resistant to 1% SDS were solubilized , labeled with Cy5 ( [NSI+] ) and Cy3 ( [nsi-] ) fluorescent dyes and analyzed by two-dimensional gel electrophoresis ( 2D-DIGE ) ( Fig 1 ) . Proteins present only in the test sample ( [NSI+] ) are pseudocolored in red , proteins present only in the control sample ( [nsi-] ) are pseudocolored in green and yellow spots correspond to proteins that were present in both samples . Such yellow spots were identified as the aminopeptidases Ape1 and Ape4 ( Fig 1 , S1 and S2 Figs ) that we had detected in our previous study in different yeast strains [22] . Red spots specific to the [NSI+] sample ( Fig 1 ) were identified as Rnq1 ( S3 Fig ) , which is the structural protein of [PIN+] prion [10 , 23] . To confirm the presence of [PIN+] in the [NSI+] strain , we transformed 1-1-D931 [NSI+] and 1-1-1-D931 [nsi-] cells with pCUP1-RNQ1-CFP ( LEU2 ) plasmid and grew the cells for 48 h at 30°C in liquid–Leu selective medium containing 150 μM CuSO4 . Next , we analyzed Rnq1-CFP aggregation in these strains using semi-denaturing detergent agarose gel electrophoresis ( SDD-AGE ) assay ( Materials and Methods ) . Rnq1-CFP formed SDS-resistant oligomers in the [NSI+] strain but not in [nsi-] ( Fig 2A ) . Thus , we confirmed that the [NSI+] strain bears the [PIN+] prion . Next , we tested whether [PIN+] affects the phenotypic manifestation or maintenance of the [NSI+] factor . We obtained three independent clones containing PCR-generated deletion of RNQ1 in the 1-1-D931 [NSI+] strain ( Materials and Methods ) and analyzed the phenotypic manifestation of the resulting strain 5-1-1-D931 [NSI+] rnq1Δ . All clones containing deletion of RNQ1 were characterized by a decreased level of nonsense suppression ( Fig 2B ) which was intermediate between those of [NSI+] and [nsi-] . However , deletion of RNQ1 did not affect cell growth on media containing galactose as the sole carbon source ( Fig 2B ) . The weak suppressor phenotype of 5-1-1-D931 [NSI+] rnqΔ was stably inherited in mitotic progeny , and reintroduction of the RNQ1 gene on the plasmid did not restore strong nonsense suppression ( Fig 2B ) . Moreover , the weak suppressor phenotype of 5-1-1-D931 [NSI+] rnqΔ strain was completely eliminated by curing on YPD medium containing 5 mM Guanidine Hydrochloride ( GuHCl ) ( Fig 2B ) . These data suggest that the strong nonsense suppressor phenotype of the [NSI+] strains is a result of interaction between [PIN+] , which acts as the enhancer of nonsense suppression , and a second , unknown prion . Since we proposed that weak nonsense suppression in the [NSI+] rnqΔ strain was caused by an unknown prion that was undetectable by the standard PSIA approach , we modified the PSIA protocol to improve its sensitivity . Although 2D-DIGE is a useful comparative method , it has strong limitations . For example , minor proteins cannot be detected on the gel , and proteins with extreme pI typically do not enter the gel . In addition , not all amyloids are soluble in UTC ( 8 M urea , 2 M thiourea , 4% CHAPS , and 30 mM TrisHCl pH 8 . 5 ) buffer [22] . In this study , we used a novel variant of PSIA called PSIA-LC-MALDI . This method consists of ( i ) the previously described procedure of isolation of detergent-resistant protein fractions [22] followed by ( ii ) solubilization of proteins with formic acid and by boiling in SDS-PAGE loading buffer , ( iii ) purification of proteins from detergent , trypsinization and ( iv ) separation of tryptic peptides by high-performance liquid chromatography coupled with matrix-assisted laser desorption/ionization mass spectrometry ( LC-MALDI ) . For a detailed description of PSIA-LC-MALDI , see Materials and Methods . We applied this method to identify detergent-resistant proteins from the 4-1-1-D931 [NSI+] and 1-4-1-1-D931 [nsi-] strains . Yeast core ribosomal proteins that were presented in SDS-resistant fraction were excluded from this table because they form SDS-resistant non-amyloid complexes [24] . From this analysis we identified 46 proteins with Mascot score >60 at a significance level of p<0 . 05; 42 of them were presented in SDS-resistant fraction of both , [NSI+] and [nsi-] strains , while 4 were identified in the [NSI+] strain only: Rnq1 , Swi1 , Mit1 , and Sis1 . ( S1 Table; MS/MS spectra of Rnq1 , Swi1 , Sis1 , and Mit1 proteins are presented in S4–S7 Figs , respectively ) . The presence of Rnq1 in the SDS-resistant fraction of the [NSI+] strain supports our conclusion that the [NSI+] strain contains the [PIN+] prion . Swi1 is the structural protein of the [SWI+] prion [5] , whereas Mit1 is an evolutionally conserved transcriptional regulator of pseudohyphal growth [25] whose amino acid sequence contains an extremely asparagine-rich region . The presence of Sis1 chaperone in the SDS-resistant fraction of the [NSI+] strain is not surprising , because Sis1 binds [PIN+] aggregates [26] and is important for its propagation [27] . To analyze whether the [NSI+] strain contains Swi1 in the [SWI+] state , the [NSI+] and [nsi-] strains were transformed with pCUP1-SWI1 ( 1–297 ) -YFP ( URA3 ) plasmid , cell lysates were separated by centrifugation onto the soluble ( S ) and insoluble ( I ) fractions and Swi1 ( 1–297 ) -YFP protein was detected by Western blotting . The Swi1 ( 1–297 ) -YFP protein in the [NSI+] strain was detected in the insoluble fraction only , whereas in the [nsi-] strain it was presented mostly in the soluble fraction ( Fig 3A ) . Next , the insoluble fractions from the [NSI+] and [nsi-] strains were analyzed by semi-denaturing detergent agarose gel electrophoresis ( SDD-AGE ) . We showed that Swi1 ( 1–297 ) -YFP forms SDS-resistant polymers in the [NSI+] strain but not in [nsi-] ( Fig 3A ) . These data support the results of PSIA-LC-MALDI and confirm that the [NSI+] strain contains Swi1 in the [SWI+] state . We demonstrated that the PCR-generated deletion of SWI1 leads to very strong nonsense suppression and an almost complete absence of growth on medium containing galactose as the sole carbon source ( Fig 3B ) . To analyze the effect of [SWI+] elimination , we re-introduced SWI1 by transformation of the 11-1-1-D931 swi1Δ strain with the YSC4613 genomic library ( Open Biosystems , USA ) plasmid containing the SWI1 gene . Next , we analyzed nonsense suppression and growth of the transformants on media without adenine or with galactose as the sole carbon source . Cells that lost the [SWI+] prion had a phenotypic manifestation identical to [nsi-] ( Fig 3B ) . Probably , [SWI+] is the key determinant of the [NSI+] factor that regulates both nonsense suppression and growth on the medium containing galactose as the sole carbon source . However , considering that the strong suppressor phenotype manifested only in strains containing both [SWI+] and [PIN+] prions , we can assume that the manifestation of the [NSI+] factor results from interaction between [SWI+] and [PIN+] prions . The third asparagine-glutamine-rich protein identified only in SDS-resistant aggregate fraction of the [NSI+] strain was Mit1 . We analyzed whether Mit1 is present in the aggregated state in the [NSI+] strain by SDD-AGE . The 4-1-1-D931 [NSI+] and 1-4-1-1-D931 [nsi-] strains were transformed with a pMIT1-MIT1-GFP ( URA3 ) centromeric plasmid that expresses the Mit1 protein fused with GFP under the control of its endogenous promoter . Interestingly , the results of SDD-AGE demonstrated that a small portion of Mit1-GFP protein formed detergent-resistant aggregates in both [NSI+] and [nsi-] strains ( Fig 4A ) . Considering that the level of Mit1-GFP expression in our experiment was close to physiological , we propose that some portion of this protein permanently forms amyloid-like aggregates in yeast cells . Further , we analyzed the influence of MIT1 deletion on the phenotypic manifestation of the [NSI+] strain . The levels of nonsense suppression and growth on galactose-containing medium were identical between the 2–936 [NSI+] mit1Δ strain and 1-1-D931 [NSI+] strain ( Fig 4B ) . Additionally , expression of MIT1 on the plasmid in the 2–936 [NSI+] mit1Δ strain did not affect the [NSI+] phenotype ( Fig 4B ) . Together , these results indicate that Mit1 forms SDS-resistant aggregates independently of the [NSI] status of the cell and does not affect the maintenance or phenotypic manifestation of [NSI+] . Once we demonstrated that the [NSI+] phenotype is determined by two prions , [SWI+] and [PIN+] , we decided to analyze their interactions . First , we compared the growth of strains containing different combinations of [SWI+] and [PIN+] prions as well as deletions of the SWI1 or RNQ1 genes . The phenotypes of all these strains were analyzed on medium without adenine and on medium containing galactose as the sole carbon source ( Fig 5A ) . As shown in the figure , [swi-][pin-] and [swi-][PIN-+] strains did not grow on–Ade medium . [SWI+] caused a weak nonsense suppression in the absence of [PIN-+] . Comparative analysis of the growth of the [SWI+][PIN+] and [SWI+][pin-] strains on–Ade medium showed that [PIN-+] enhanced [SWI+]-dependent nonsense suppression . The strongest suppressor phenotype was observed on the background of SWI1 gene deletion in three independently obtained clones , and in this case suppression did not depend on [PIN+] ( Fig 5A ) . These data suggest that suppression of nonsense mutation depends on Swi1 inactivation . Prion inactivation of Swi1 protein causes weak nonsense suppression , [PIN+] enhances this effect , whereas deletion of SWI1 leads to strongest nonsense suppressor phenotype . To reproduce these results , the strains with different combination of prions ( 26-1-4-1-1-D931 [swi-][PIN+] , 12-1-4-1-1-D931 [SWI+][pin-] , and 16-1-4-1-1-D931 [SWI+][PIN+] ) were obtained by transformation of [swi-][pin-] cells with protein lysates ( see “Materials and Methods” ) . The recipient 1-4-1-1-D931 [swi-][pin-] strain was transformed with protein extract from the 1-1-D931 [SWI+][PIN+] strain and with plasmid pRNQ1-GFP ( URA3 ) . The clones that acquired strong and weak suppressor phenotypes were selected . All these clones lost nonsense suppressor phenotype after GuHCl treatment . The [swi-][PIN+] protein transformants which had the same phenotype as the recipient [swi-][pin-] cells , but contained Rnq1-GFP aggregates , were selected by fluorescent microscopy ( S8 Fig ) . The [SWI+] status of cells that acquired strong or weak suppressor phenotype was confirmed by analysis of Swi1 ( 1–297 ) -YFP fluorescent aggregates ( S9 Fig ) . Thus , we obtained the [SWI+][PIN+] , [SWI+][pin-] , and [swi-][PIN+] strains by two different methods and showed that suppression of ade1-14UGA nonsense mutation depends on interaction of [SWI+] and [PIN+] prions ( Fig 5A ) . Such interaction of two prions is very similar to the classical complementary interaction of two genes , where dominant allele of one gene enhances the manifestation of dominant allele of other gene . The second analyzed phenotype , i . e . , growth on medium containing galactose as the sole carbon source , was determined by [SWI+] only , whereas the presence or absence of [PIN+] did not affect this phenotypic trait ( Fig 5A ) . Since nonsense suppression in [SWI+] strains depends on [PIN] status , we proposed that [PIN+] might affect aggregation of Swi1 protein in prion form . To test this hypothesis , we compared the levels of Swi1 aggregates in the 12-1-4-1-1-D931 [SWI+][pin-] and 16-1-4-1-1-D931 [SWI+][PIN+] strains by centrifugation analysis . The data presented in Fig 5B demonstrate that Swi1 ( 1–297 ) -YFP protein is detected in the [SWI+][PIN+] cells in the insoluble fraction only , whereas in the [SWI+][pin-] cells Swi1 ( 1–297 ) -YFP is presented in both , soluble and insoluble fractions . These data show that [PIN+] directly or indirectly enhances Swi1 aggregation in the [SWI+] strains . Recently , it was shown that coexisting aggregates of Swi1-YFP and Rnq1-CFP do not colocalize in [PIN+][SWI+] cells [16] . We also demonstrated that Swi1-YFP and Rnq1-CFP proteins do not colocalize in the D938 [SWI+][PIN+] strain ( Fig 5C ) . Thus , the presence of [PIN+] enhances formation of Swi1-YFP aggregates in [SWI+] strains , but this effect is not mediated by a physical interaction between Rnq1 and Swi1 proteins . In a previous work , we demonstrated that levels of SUP45 mRNA expression and Sup45 production were 2–3 times higher in the [nsi-] strain than in [NSI+] [28] . Moreover , nonsense suppression does not manifest , when the [NSI+] strain is transformed by the centromeric plasmid containing SUP45 under the control of endogenous promoter [29–31] . Thus , even two-fold increase in the level of SUP45 expression completely prevents the appearance of nonsense suppression in the [NSI+] strains . To determine whether decreased expression of SUP45 and nonsense suppression in the [NSI+] strain depend on the [SWI+] and [PIN+] prions , we compared SUP45 mRNA levels in the [SWI+][PIN+] , [SWI+][pin-] , and [swi-][pin-] strains . RNA extraction , cDNA synthesis , and real-time PCR were performed as described in “Materials and Methods” . The data obtained show that SUP45 mRNA levels in the [swi-][pin-] strain were approximately 1 . 5 and 2 . 5 times higher ( p<0 . 01 ) than in the [SWI+][pin-] and [SWI+][PIN+] strains , respectively ( Fig 6A ) . We may conclude that Swi1 is a positive regulator of SUP45 , whereas the prion inactivation of Swi1 decreases the SUP45 expression level . Importantly , SUP45 expression was higher ( p<0 . 01 ) in the [SWI+][pin-] strain than in [SWI+][PIN+] ( Fig 6A ) . These data show that [PIN+] not only influences prion aggregation of Swi1 ( Fig 5B ) , but also [PIN+] enhances the effect of [SWI+] on SUP45 expression ( Fig 6A ) . These effects correlate with the differences in growth on–Ade medium between the [swi-][pin-] , [SWI+][pin-] , and [SWI+][PIN+] strains ( Fig 6B ) . Thus , interaction between [SWI+] and [PIN+] causes decreased SUP45 expression and leads to nonsense suppression .
Previously , we performed a set of unsuccessful attempts to identify the [NSI+] factor using genetic approaches [19 , 21] . Genetic screening for the detection of prion structural genes may be useless because the overexpression of prion-forming proteins does not always lead to prion induction [16] , and deletion screens are always incomplete , as they cannot be applied to essential genes . Here , for the first time , we identified protein determinants of an unknown prion factor using a proteomic approach . Two similar methods , TAPI and PSIA , based on the universal feature of amyloids to form detergent-resistant aggregates , were recently developed [22 , 24] . Both methods have some limitations and include the step of protein extraction from an agarose or polyacrylamide gel . Using a new gel-free modification of this proteomic approach named PSIA-LC-MALDI , we have shown that [NSI+] cells , in contrast to [nsi-] , contain SDS-resistant aggregates of Swi1 , Rnq1 and Mit1 proteins ( S1 Table , S4–S6 Figs ) . Note , using the original PSIA method , we detected only Rnq1 protein in the SDS-resistant fraction of the [NSI+] strain ( Fig 1 , S3 Fig ) and did not identify Swi1 or Mit1 . These data suggest that PSIA-LC-MALDI is a more powerful method for the identification of proteins forming amyloid-like aggregates . Deletion of MIT1 does not affect the manifestation or maintenance of [NSI+] ( Fig 4B ) ; thus , we can conclude that MIT1 is not a structural gene for [NSI+] . At the same time , our SDD-AGE experiment demonstrated that a small portion of Mit1-YFP protein forms SDS-resistant aggregates in both [NSI+] and [nsi-] cells . It is probable that Mit1 was not identified in proteomic screening in SDS-resistant fraction of the [nsi-] strain , because this protein is expressed in yeast at an extremely low level [32 , 33] , and only a small portion of Mit1 forms SDS-resistant aggregates that may be detected at the lower limit of sensitivity of our PSIA-LC-MALDI approach . Mit1 is a transcriptional regulator of pseudohyphal growth [25] , whose sequence contains a region extremely rich in asparagine ( http://www . yeastgenome . org/locus/S000000733/protein ) . Based on the data obtained , we propose that Mit1 forms amyloid-like aggregates at physiological conditions , though the possible functional roles of these aggregates are unclear and may represent a subject for further study . We have shown that the chimeric proteins Rnq1-CFP and Swi1-YFP form SDS-resistant aggregates in the [NSI+] strain , but not in [nsi-] ( Figs 2A and 3A ) . Thus , the [NSI+] cells , in contrast to [nsi-] , contain Rnq1 and Swi1 proteins in their prion forms . Surprisingly , the elimination of the [PIN+] prion in [NSI+] cells causes a strong decrease in the nonsense suppression level ( Fig 2B ) . The weak suppressor phenotype was stably inherited in the strain that lost the [PIN+] factor and efficiently eliminated by GuHCl . These data suggest that the strong suppressor phenotype in [NSI+] strains is determined not only by [PIN+] but also by another prion , [SWI+] , which was identified by PSIA-LC-MALDI . Elimination of [SWI+] causes the complete loss of all manifestations of the [NSI+] phenotype ( Fig 3B ) . Taking into consideration that the strong suppressor phenotype manifests only in [PIN+][SWI+] strains , we conclude that the heritable trait detected in our strains is the result of the interaction of [PIN+] and [SWI+] prions ( Fig 5A ) . Our data showing that [PIN+] enhances Swi1-YFP aggregation ( Fig 5B ) strongly support this conclusion ( Fig 5B ) . At first glance , the appearance of a heritable trait related to a prion-prion interaction is surprising , because previous studies have shown that coexisting prion polymers typically do not physically interact [16 , 34] . Moreover , it was recently shown that aggregates of Swi1-YFP and Rnq1-CFP do not colocalize in [PIN+][SWI+] cells [16] . Nevertheless , although [PIN+] and [SWI+] prions show no colocalization ( Fig 5C ) , they exhibit a functional interaction that is mediated by other components of the proteomic network and can be monitored by the level of nonsense suppression . Prion conversion may lead not only to protein inactivation but also to the acquisition of novel functions . For example , Rnq1 protein only in its prion state causes hyperphosphorylation of Pin4 [35] . The authors of this work suggest that [PIN+] prion could serve as an epigenetic switch to promote the post-translational modification of yeast proteins . We have shown that Rnq1 in the [PIN+] state increases [SWI+]-dependent nonsense suppression ( Fig 5A ) . One cannot exclude the possibility that Rnq1 in its prion form causes posttranslational modification of Swi1 prion aggregates . On the other hand , it can be assumed that [PIN+] polymers may affect chaperone machinery which interacts with prion [SWI+] and can influence its properties . Our data according to that Sis1 chaperone is presented in fraction of SDS-resistant aggregates only in the [PIN+] [SWI+] strain ( S1 Table ) support this hypothesis . Swi1 is a global transcriptional regulator that affects the transcription of a number of yeast genes [36 , 37] . We showed that Swi1 positively regulates the transcription of the SUP45 gene ( Fig 6A ) that encodes the translation termination factor eRF1 [29] . Prion inactivation of Swi1 causes a decrease in SUP45 expression that leads to the weak suppressor phenotype ( Fig 6 ) . [PIN+] increases [SWI+]-dependent inactivation of Sup45 and enhances nonsense suppression ( Fig 5A and Fig 6 ) . A scheme illustrating the effect of interaction between [SWI+] and [PIN+] on nonsense suppression is presented in Fig 7 . Overall , we can conclude that [SWI+] and [PIN+] , like classical genes , demonstrate complementary interactions , and this prion-prion interaction causes heritable traits in Saccharomyces cerevisiae . [SWI+] is the key determinant of nonsense suppression , while [PIN+] does not cause nonsense suppression by itself , but strongly enhances the effect of [SWI+] . Thus , by analogy with monogenic and polygenic inheritance , in the framework of the prion concept we can distinguish “monoprionic” and “polyprionic” types of inheritance . We assume that new examples of polyprionic inheritance will be revealed using the proteomic methods for identification of prions .
Standard yeast genetic techniques , media , and cultivation conditions were used [38] . Yeast cultures were grown at 30°C . 150 μM copper sulfate ( CuSO4 ) was added to synthetic and YPD media to induce the expression of genes under the PCUP1 promoter . To eliminate prions , yeast cultures were grown for three consecutive passages on the solid YPD medium in the presence of 5 mM Guanidine Hydrochloride ( GuHCl ) . Growth on the selective medium containing 20 g/l galactose ( Gal ) as the sole carbon source was analyzed as described below . Yeast were grown for one day on the solid YPD medium and then passaged threefold , for one day per passage , on solid medium containing 20 g/l galactose as the sole carbon source at 30°C . 1-1-D931 [NSI+] and 1-1-1-D931 [nsi-] strains were described previously [18] . The genotype of these strains is MATa sup35Δ::HIS3 ade1-14 his3 leu2 lys2 ura3 trp1-289 [pU-Aβ-Sup35MC] . The 4-1-1-D931 [NSI+] and 1-4-1-1-D931 [nsi-] strains have the same genotype but contain the pL-Aβ-Sup35MC plasmid [18] . The pU-Aβ-Sup35MC and pL-Aβ-Sup35MC plasmids contain hybrid Aβ-SUP35MC gene under the control of the PCUP1 promoter , which compensates for SUP35 deletion in the chromosome [18] . Presence of the [NSI+] factor was detected by suppression of the ade1-14UGA mutation resulting in the growth of [NSI+] strains on the synthetic medium without adenine ( –Ade ) . This medium was supplemented with 150 μM CuSO4 to overexpress Aβ-SUP35MC that is essential for the detection of differences in growth of [NSI+] and [nsi-] cells on–Ade medium . The 5-1-1-D931 [NSI+] rnqΔ strain ( MATa sup35Δ::HIS3 ade1-14 his3 leu2 lys2 ura3 trp1-289 rnq1Δ::KanMX [pU-Aβ-Sup35MC] ) contains a deletion of RNQ1 chromosomal copy substituted with a KanMX cassette that provides resistance to the antibiotic geneticin ( G418 ) . This strain was obtained by the PCR-generated gene deletion technique [39] , during which KanMX cassette flanked by 5’ and 3’ regions of RNQ1 was PCR-amplified from the plasmid pFA6-kanMX4 with the primers FRNQ1deltaKanMX4 and RRNQ1deltaKanMX4 ( first PCR ) , FRNQ1delta and RRNQ1delta ( second PCR ) ( S2 Table ) and transformed into the strain 1-1-D931 [NSI+] . Transformants were selected on YPD plates containing 200 mg/L G418 ( Promega , USA ) . The deletion of RNQ1 was PCR-verified with the primers FRNQ1deltach and RRNQ1deltach ( S2 Table ) in three independently obtained transformants . Analogously , the 11-1-1-D931 [NSI+] swi1Δ strain ( MATa sup35Δ::HIS3 ade1-14 his3 leu2 lys2 ura3 trp1-289 swi1Δ::KanMX [pU-Aβ-Sup35MC] ) was obtained from 1-1-D931 [NSI+] strain . KanMX cassette flanking by sequences from promoter and middle region of SWI1 was PCR-amplified from the plasmid pFA6-kanMX4 with the primers FSWI1deltaKanMX4 and RSWI1deltaKanMX4 ( first PCR ) ; and FSWI1delta and RSWI1delta ( second PCR ) ( S2 Table ) . Deletion was verified by PCR with FSWI1deltach and RSWI1deltach ( S2 Table ) in three independently obtained transformants . The 2–936 [NSI+] mit1Δ strain ( MATa sup35Δ::HIS3 ade1-14 his3 leu2 lys2 ura3 trp1-289 mit1Δ::KanMX [pU-Aβ-Sup35MC] ) was obtained by mating the mit1Δ strain ( MATα his3Δ1 leu2Δ lys2Δ ura3 ) from the BY4742 deletion collection ( Invitrogen , USA ) to the 1-1-D931 [NSI+] strain followed with sporulation and dissection of the resulting diploids . MIT1 deletion was PCR-verified with the primers FMIT1deltach and RMIT1deltach ( S2 Table ) . The [nsi-] derivatives of the [NSI+] strains bearing deletions of the MIT1 , RNQ1 or SWI1 genes were obtained by GuHCl treatment . The D938 [SWI+][PIN+] diploid strain ( SUP35/sup35Δ::HIS3 ADE1/ade1-14 his3/his3 leu2/leu2 lys2/lys2 ura3/ura3 TRP1/trp1-289 ) was obtained by mating the 1-1-D931 strain to BY4742 ( MATα his3Δ1 leu2Δ lys2Δ ura3 ) ( Invitrogen , USA ) followed by elimination of the pU-Aβ-Sup35MC plasmid . The 26-1-4-1-1-D931 [swi-][PIN+] , 12-1-4-1-1-D931 [SWI+][pin-] , and 16-1-4-1-1-D931 [SWI+][PIN+] strains were obtained by transformation of the 1-4-1-1-D931 [swi-][pin-] recipient yeast cells with 1-1-D931 [SWI+][PIN+] protein lysates ( see “Transformation of yeast cells with protein lysates” ) . The pRNQ1-GFP ( URA3 ) plasmid that contains chimeric RNQ1-GFP gene under the control of the copper-inducible PCUP1 promoter was described early [10] . The pCUP1-RNQ1-CFP ( LEU2 ) plasmid contains RNQ1 fused with the sequence encoding cyan fluorescent protein ( CFP ) under the control of the PCUP1 . To obtain this plasmid , the fragment encoding CFP was PCR-amplified from the plasmid pDH5 ( Yeast Resource Center , University of Washington , USA , http://depts . washington . edu/~yeastrc ) with the primers FCFPSacII and RCFPSacI ( S2 Table ) . Next , the sequence encoding green fluorescent protein ( GFP ) in the plasmid pRNQ1-GFP ( URA3 ) [10] was substituted with PCR-amplified CFP digested with SacI and SacII . Finally , the XhoI-SacI restriction fragment of the pRNQ1-CFP ( URA3 ) plasmid containing PCUP1-RNQ1-CFP was inserted into the pRS415 multicopy vector [40] . pMIT1-MIT1-GFP ( URA3 ) plasmid was constructed as follows . The Mit1-encoding sequence was PCR-amplified using the primers MIT1F and MIT1R ( S2 Table ) using YGPM21o12 plasmid of the YSC4613 yeast genomic library ( Open Biosystems , USA ) as a template . Next , PCR-amplified MIT1 was inserted into the pRNQ1-GFP ( URA3 ) plasmid by the BamHI and SacII digestion sites . As a result , pCUP-MIT1-GFP ( URA3 ) was obtained . Further , MIT1 promoter was PCR-amplified using MIT1_prom_F and MIT1_prom_R primers ( S2 Table ) and the YGPM21o12 plasmid as a template . The resulting PMIT1 sequence was inserted into the pCUP-MIT1-GFP ( URA3 ) with digestion sites ClaI and BamHI . The pCUP1-SWI1 ( 1–297 ) -YFP ( URA3 ) plasmid contains a Swi1 ( aa 1–297 ) -encoding sequence fused in-frame with the sequence encoding yellow fluorescent protein ( YFP ) under the control of PCUP1 promoter . To construct this plasmid , SWI1 fragment was PCR-amplified using FSWI1 ( 1 ) HindIII and RSWI1 ( 889 ) BamHI primers ( S2 Table ) and the YGPM19p21 plasmid of YSC4613 ( Open Biosystems , USA ) as a template . Next , the PCR-amplified SWI1 fragment was cloned into the pU-CUP1-YFP plasmid [41] with the digestion sites HindIII and BamHI . The p426GPD–SWI1YFP plasmid contains the sequence encoding full-length Swi1 fused in-frame with YFP under the control of a strong constitutive PGPD promoter [16] was kindly provided by L . N . Mironova ( St . Petersburg State University ) . Transformation of yeast cells with total protein lyzate was performed as described previously [18] . To introduce [PIN+] or [SWI+] prions , the 1-4-1-1-D931 [swi-][pin-] spheroplasts were co-transformed with 1-1-D931 [SWI+][PIN+] protein lysate and pRNQ1-GFP ( URA3 ) plasmid . The transformants were selected on–Leu–Ura medium with 1M sorbitol , tested of mating type and analyzed for presence of [PIN+] or [SWI+] prions as follows . To analyze the presence of [PIN+] , the aggregation of Rnq1-GFP was analyzed by the fluorescence microscopy ( see “Fluorescence microscopy” ) . The [SWI+] status of transformants was phenotypically detected by the growth on–Ade medium with 150 ΔM CuSO4 . To confirm [SWI+] status of protein transformants the pRNQ1-GFP ( URA3 ) plasmid was replaced with the pCUP1-SWI1 ( 1–297 ) -YFP ( URA3 ) plasmid , and fluorescence analysis of the Swi1 ( 1–297 ) -YFP aggregation was performed . The PSIA ( Proteomic Screening and Identification of Proteins ) approach was described previously ( for details , see [22] ) . In general , PSIA consists of three steps: ( i ) isolation of detergent-resistant protein aggregate fractions , ( ii ) separation of proteins from aggregates by two-dimensional difference gel electrophoresis ( 2D-DIGE ) , and ( iii ) identification of separated proteins . The isolation of proteins forming detergent-resistant aggregates is comprised of a series of ultracentrifugations of protein lysates at 151000 x g coupled to treatment with ionic detergents [22] . In this study , samples were treated with 1% sodium dodecyl sulfate ( SDS ) . Additionally , 0 . 1% SDS was added to the sucrose cushion for ultracentrifugation followed the detergent treatment . The proteins from the [NSI+] and [nsi-] strains were labeled at lysine residues with Cy5 and Cy3 fluorescent dyes , correspondingly , according to recommendations of the manufacturer . The proteins were dissolved in UTC buffer ( 8 M urea , 2 M thiourea , 4% CHAPS , and 30 mM TrisHCl pH 8 . 5 ) and separated by 2D-DIGE . Gel slices were washed twice with deionized water and washed once with 40% acetonitrile in 50 mM ammoniumbicarbonate solution . Next , dehydration was performed in 100% acetonitrile followed by removing of liquid and air-drying of gel slices . The dried samples were incubated for 4 h with 5 ml of sequencing grade trypsin ( Promega ) 5 mg/ml solution , 100 mM ammonium bicarbonate ( pH 7 . 0 ) at 37°C . Peptides were extracted with 0 , 5 ml of 0 . 1% trifluoroacetic acid in water . Mass spectrometric peptide analysis was performed using an Ultraflextreme MALDI-TOF/TOF mass spectrometer ( Bruker Daltonics , DE ) equipped with an Nd laser ( 354 nm ) in reflecto-mode ( the mass range 700–4500 m/z ) . The matrix was α-cyano-4-hydroxycinnamic acid . Peak lists were generated by the flexAnalysis 3 . 2 software ( Bruker Daltonics ) . Proteins were identified by Mascot software release version 2 . 4 . 2 ( Matrix Science , http://www . matrixscience . com ) in the database of National Center for Biotechnology Information ( NCBI ) [22] . This variant of PSIA uses high-performance liquid chromatography coupled with mass-spectrometry . The first step , isolation of detergent-resistant protein aggregate fractions , was performed as described previously [22] . Isolated proteins were lyophilized using the vacuum concentrator Labconco CentriVap ( Labconco , USA ) . Next , lyophilized samples were treated with formic acid ( 90% ) , dried in the vacuum concentrator Labconco CentriVap ( Labconco , USA ) , solubilized in Tris-buffered saline ( TBS ) , and boiled in SDS-PAGE loading buffer . Then , detergents and salts were removed from the samples using HiPPR Detergent Removal columns ( Thermo Scientific , USA ) and Zeba Desalting columns ( Thermo Scientific , USA ) , respectively , according to the manufacturers’ protocols . Final samples ( volume 50 μl , total protein concentration 0 . 2–0 . 4 mg/ml ) were supplemented with 1 μl of freshly prepared 50 mM DTT in 50 mM ammonium bicarbonate , incubated for 15 min at 50°C , supplemented with 1 μl 100 mM iodoacetamide in 50 mM ammonium bicarbonate and incubated for 15 min at 20°C in the dark . Then the samples were supplemented with 1 μl DTT to inactivate iodoacetamide and 5 μl trypsin ( 10 ng/μl; Sigma ) and incubated overnight at 37°C . The trypsin was inactivated by adding 0 . 5 μl 10% TFA followed by centrifuging for 30 min ( 20 , 000g , 4°C ) . The final peptide mixtures were loaded ( 1 μl ) onto an Acclaim PepMap 300 HPLC reverse-phase column ( 150 mm , 75 μm , particle size 5 μm; Thermo Scientific , USA ) and separated in an acetonitrile gradient ( 2–90% ) during 45 min using an UltiMate 3000 UHPLC RSLC nano high-performance nanoflow liquid chromatograph ( Dionex , USA ) . Peptide fractions were collected every 10 s and loaded onto a 384-sample MTP AnchorChip 800/384 microtiter plate ( Bruker Daltonics ) using spotter Proteineer fc II ( Bruker Daltonics ) [42] . Peptides were identified using the Ultraflextreme MALDI-TOF/TOF mass spectrometer ( Bruker Daltonics , DE ) . MS-spectra for each peptide fraction were determined and analyzed using WARP-LC software . An array of unique peptides characterized by specific retention time , charge , and molecular weight was determined . MS/MS-analysis was performed for these peptides in fractions ( spots ) with maximal concentration ( peak intensity ) of these peptides . Match between the experimental spectra and corresponding proteins was analyzed using Mascot version 2 . 4 . 2 software ( Matrix Science; http://www . matrixscience . com ) in the UniProt database ( http://www . uniprot . org/ ) restricted to Saccharomyces cerevisiae . As matrix , α-cyano-4-hydroxycinnamic acid was used . During analysis , preset parameters of “Mass tolerance” were used ( precursor mass tolerance 100 ppm , fragment mass tolerance 0 . 9 Da ) . As a standard sample , Peptide Calibration Standard II 8222570 ( Bruker Daltonics ) was applied . Carboxymethylation of cysteine , partial oxidation of methionine , and one skipped trypsinolysis site were considered as permissible modifications [42] . The BioTools software ( Bruker , Bremen , Germany ) was used for manual validation of protein identification . Preparation and fractionation of protein lysates by centrifugation were performed as described previously [43] , with modifications . Total lysate was fractionated by centrifugation at 100 000 g for 20 min , 4°C . The supernatant was placed into a fresh tube , and the insoluble fraction was resuspended in an equal amount of lysis buffer . SDS , glycerol , β-mercaptoethanol , and Tris-HCl ( pH 6 . 8 ) were added to each sample up to final concentrations of 3% , 10% , 3% , and 0 . 15 M , respectively . Resulting samples were heated at 95°C for 10 min and run on the standard SDS-polyacrylamide gel . Next , proteins were transferred onto Immobilon-P PVDF membrane ( GE Healthcare , USA ) , reacted to antibodies against GFP [E385] ( ab32146 ) ( Abcam , Great Britain ) , and detected by Amersham ECL Prime Western Blotting Detection Reagent kit ( GE Healthcare , USA ) . Semi-Denaturing Detergent Agarose Gel Electrophoresis ( SDD-AGE ) [44 , 45] was performed with 1% agarose gel . Before separation , proteins were treated for 10 min with 1% SDS at room temperature . The separated proteins were transferred onto Immobilon-P PVDF membrane ( GE Healthcare , USA ) . Proteins fused with CFP , GFP , and YFP were detected using monoclonal rabbit primary antibodies against GFP [E385] ( ab32146 ) ( Abcam , Great Britain ) and the Amersham ECL Prime Western Blotting Detection Reagent kit ( GE Healthcare , USA ) . RNA extraction was performed with “Trizol” reagent ( “Invitrogene” , USA ) . Reverse transcription was carried out using SuperScript III cDNA synthesis kit ( “Invitrogene” , USA ) . Real-time PCR was performed with primers ( ACT1F , ACT1R , SUP45F , and SUP45R ) and “TaqMan” probes ( ACT1probe , SUP45probe ) listed in S2 Table . The probes were conjugated with FAM ( for ACT1 ) or R6G ( for SUP45 ) fluorophores as well as with BHQ quencher ( “Beagle” , Russian Federation ) . Actin-encoding gene ACT1 was used as the reference gene . Results of real-time PCR were normalized with the 2-ΔΔC ( t ) method [46] . This method uses ΔCt parameter indicating the difference in the intensities of signals between gene of interest ( SUP45 ) and reference ( ACT1 ) . Next , ΔΔCt value , which is the difference between ΔCt parameters in experiment and control , is calculated . Finally , 2-ΔΔC ( t ) is calculated . This value demonstrates the relative amounts of mRNAs of interest in comparison between the experiment and control samples . The results are presented as the means ± the standard deviations . Fluorescence microscopy assays of GFP , YFP and CFP-fused proteins were performed with a Leica DM6000B microscope using GFP , YFP and CFP cubes and the Leica QWin Standart 3 . 2 . 0 software ( Leica Microsystems GmBH , Germany ) . | The data presented in the paper deepens and enriches the concept of protein-based inheritance . According to this concept , prion conformational switches change protein functional activity , and such changes are inherited . Here , for the first time , we demonstrate that heritable traits may appear not only due to a conformational switch of one protein but also can be caused by interactions between different prions . To identify the novel epigenetic factor that causes suppression of nonsense mutations in yeast , we applied our original method of proteomic screening of prions . We have shown that two yeast proteins , which normally do not interact , in prion form demonstrate genetic interaction: one is the key determinant of the suppression of nonsense mutation , while the second enhances this effect . Thus , by analogy with monogenic and polygenic inheritance , in the framework of the prion concept , we can distinguish “monoprionic” and “polyprionic” inheritance . We assume that new examples of polyprionic inheritance will be revealed using modern proteomic methods for identification of prions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"chemical",
"compounds",
"prions",
"light",
"microscopy",
"carbohydrates",
"galactose",
"organic",
"compounds",
"plasmid",
"construction",
"fungi",
"model",
"organisms",
"microscopy",
"materials",
"science",
"experimental",
"organism",
"systems",
"dna",
"construction",
"m... | 2016 | Interaction of Prions Causes Heritable Traits in Saccharomyces cerevisiae |
Schistosomiasis , caused by parasitic flatworms of the genus Schistosoma , is a neglected tropical disease affecting hundreds of millions globally . Praziquantel ( PZQ ) , the only drug currently available for treatment and control , is largely ineffective against juvenile worms , and reports of PZQ resistance lend added urgency to the need for development of new therapeutics . Ion channels , which underlie electrical excitability in cells , are validated targets for many current anthelmintics . Transient receptor potential ( TRP ) channels are a large family of non-selective cation channels . TRP channels play key roles in sensory transduction and other critical functions , yet the properties of these channels have remained essentially unexplored in parasitic helminths . TRP channels fall into several ( 7–8 ) subfamilies , including TRPA and TRPV . Though schistosomes contain genes predicted to encode representatives of most of the TRP channel subfamilies , they do not appear to have genes for any TRPV channels . Nonetheless , we find that the TRPV1-selective activators capsaicin and resiniferatoxin ( RTX ) induce dramatic hyperactivity in adult worms; capsaicin also increases motility in schistosomula . SB 366719 , a highly-selective TRPV1 antagonist , blocks the capsaicin-induced hyperactivity in adults . Mammalian TRPA1 is not activated by capsaicin , yet knockdown of the single predicted TRPA1-like gene ( SmTRPA ) in S . mansoni effectively abolishes capsaicin-induced responses in adult worms , suggesting that SmTRPA is required for capsaicin sensitivity in these parasites . Based on these results , we hypothesize that some schistosome TRP channels have novel pharmacological sensitivities that can be targeted to disrupt normal parasite neuromuscular function . These results also have implications for understanding the phylogeny of metazoan TRP channels and may help identify novel targets for new or repurposed therapeutics .
Trematode flatworms of the genus Schistosoma cause schistosomiasis , a tropical parasitic disease that affects hundreds of millions globally [1 , 2] , causing severe morbidity , compromised childhood development , and an estimated 280 , 000 deaths annually [3–5] . There is no vaccine , and treatment and control depend almost entirely on a single drug , praziquantel ( PZQ ) [6–8] , which , though indispensable , has significant limitations , including reduced effectiveness against immature worms [9 , 10] . Field isolates exhibiting reduced PZQ susceptibility have been reported , and PZQ resistance can be experimentally induced [10–12] , suggesting that reliance on a single treatment for a disease of this scope may be particularly dangerous . Ion channels are membrane proteins that form a gated pore , allowing ions to flow by diffusion down the electrochemical gradient established across cell membranes . They are vital for normal functioning of the neuromusculature , as well as for other cell types , and are validated and outstanding therapeutic targets [13 , 14] . Indeed , the majority of current anthelmintic drugs target ion channels of the parasite's neuromuscular system [15–17] , and there is increasing evidence that PZQ itself acts on voltage-gated Ca2+ channels [18 , 19] . To date , however , only a small subset of parasite ion channels has been investigated for potential targeting by new anthelmintics . One almost entirely unexplored group of schistosome ( and other parasite ) ion channels is the transient receptor potential ( TRP ) channel family [20] . TRP channels are non-selective cation channels that display an extraordinary diversity of functions and activation mechanisms [21 , 22] . Indeed , a single TRP channel can be activated through different , seemingly unrelated , mechanisms . TRP channels were initially discovered and characterized in Drosophila , with later identification of ~30 vertebrate isoforms . Though the full array of physiological functions of these channels is only gradually becoming clear , one unifying theme appears to be their key role in responding to all major classes of external stimuli ( eg , light , sound , chemicals , temperature , and touch ) . The ability of TRP channels to transduce these signals depends largely on their role in modulating intracellular Ca2+ concentrations [23] . The huge potential of TRP channels as therapeutic targets has recently been extensively reviewed [24] . In addition to diverse activation mechanisms , TRP channels also show differences in ion selectivity , structure , and tissue distribution [25] . Based on sequence similarity , however , TRP channels fall into several subfamilies [21] . Mammalian types include TRPC , TRPM , TRPA , TRPV , TRPML , and TRPP . These classes , as well as others ( TRPN , TRPVL ) , are found throughout the animal kingdom . A small subset of TRP channels are also expressed in protists , including protozoan parasites [26 , 27] . Representatives of five metazoan subfamilies , including TRPV , appear in the choanoflagellates [27] , the unicellular common ancestors to metazoans , indicating that most classes of TRP channels emerged prior to the appearance of multicellular animals [28 , 29] . Schistosomes contain a wide diversity of TRP channels , but were reported to lack any predicted TRPV homologs [20 , 26] . Searches of current schistosome genome databases [30–32] confirm this finding . As noted , TRP channels are often polymodal , responding to multiple stimuli . For example , TRPA1 and TRPV channels ( as well as others ) can be thermosensitive [33] , but are also activated by chemical and mechanical stimuli . Thus , TRPV1 , the mammalian vanilloid receptor , is sensitive to heat ( >43°C ) , pH , and inflammatory factors; it is also activated with high potency by capsaicin [34–36] , an active ingredient in chili peppers . Capsaicin and related compounds are selective for TRPV1; other members of the TRPV channel subfamily do not appear to respond to capsaicin [37 , 38] , and differences in capsaicin sensitivity between different vertebrate species have been localized to a few amino acid residues in the S3 and S4 domains of the TRPV1 sequence [39–43] . Many invertebrates have genes for single or multiple TRPV-like channels , although the mammalian TRPV subtypes , including the capsaicin-sensitive TRPV1 , are found only in the vertebrates [21] . Nonetheless , a few invertebrates have been reported to exhibit some sensitivity to capsaicin . Lophotrochozoans such as molluscs [44–46] and leeches [47] show cellular activation and avoidance behaviors in response to relatively high concentrations ( >100 μM ) of capsaicin , and capsaicin-like compounds inhibit zebra mussel ( Dreissena polymorpha ) macrofouling at micromolar concentrations [48] . In contrast , ecdysozoans such as Drosophila [49] and Caenorhabditis [50] do not exhibit acute capsaicin avoidance behaviors , though capsaicin appears to enhance thermal avoidance behavior in C . elegans [51] , and Drosophila has been reported to exhibit a preference for the compound [49] . Interestingly , the planarian Dugesia dorotocephala , which , like S . mansoni , is also a platyhelminth , shows no detectable response to 10 μM capsaicin , though it does respond with increased locomotor activity to the TRPM8 agonist icilin [52] . On the other hand , oil extracts of the leaves and fruit of a Brazilian species of pepper ( Capsicum annuum ) , which likely contain capsaicin , appear to have potent effects against S . mansoni cercariae , killing 90–96% within 15 min [53] . However , since the S . mansoni genome predicts no representatives of the TRPV channel subfamily [20 , 26] it would be reasonable to expect that schistosomes would be unresponsive to capsaicin and other TRPV1 channel modulators . TRPA is another TRP channel subfamily , with one member , TRPA1 , in mammals . TRPA1 channels are non-selective cation channels characterized by a large group of ankyrin repeats in the N- terminal domain . TRPA channels , like TRPV channels , transduce noxious thermal , mechanical , and nociceptive signals , and also mediate chronic itch [reviewed in 54] . Members of the TRPA1 clade of TRPA channels act as chemosensors for a wide variety of pungent irritants , most notably electrophilic compounds such as allyl isothiocyanate ( AITC ) , found in mustard oil [54 , 55]; neither human [56] nor mouse [57] TRPA1 is activated by capsaicin . TRPA1 channels are found in a variety of organisms , and the structure of the channel has recently been reported [58] . The sub-family is represented by a single gene in S . mansoni [20 , 26] , which we have named SmTRPA , coding for a protein with eight predicted N-terminal ankyrin repeats . Here , we show that , though they lack a TRPV homolog , S . mansoni are sensitive to capsaicin , which elicits dramatic hyperactivity at 10−5 M concentrations . A TRPV1 competitive antagonist blocks these effects . Surprisingly , essentially the entire response to capsaicin is eliminated by using RNA interference to suppress SmTRPA expression in adult worms . The effect of SmTRPA knockdown on capsaicin sensitivity appears to be specific; knockdown of SmTRPA has no impact on 5-hydroxytryptamine ( 5-HT; serotonin ) -elicited hyperactivity . Our results suggest that in schistosomes and perhaps other platyhelminths , TRPA1 channels may exhibit novel pharmacological sensitivities , opening the possibility for selective targeting and perhaps providing clues to the phylogenetic relationship of these TRP channel sub-families .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the U . S . National Institutes of Health . Animal handling and experimental procedures were undertaken in compliance with the University of Pennsylvania's Institutional Animal Care and Use Committee ( IACUC ) guidelines ( Animal Welfare Assurance Number: A3079-01 ) . The IACUC approved these studies under protocol number 804217 . Capsaicin and SB 366791 were from Cayman Chemical ( Ann Arbor , MI ) , resiniferatoxin ( RTX ) was from LC Laboratories ( Woburn , MA ) , and allyl isothiocyanate ( AITC ) and 5-hydroxytryptamine ( serotonin ) were from Sigma-Aldrich ( St . Louis , MO ) . Reagents were dissolved in dimethyl sulfoxide ( DMSO ) for stock solutions and then diluted to an appropriate concentration in culture media . All oligonucleotides and siRNAs were from Integrated DNA Technologies ( IDT , Coralville , IA ) . Female Swiss Webster mice infected with S . mansoni ( NMRI strain ) were provided by the Schistosomiasis Resource Center for distribution by BEI Resources , NIAID , NIH ( S . mansoni , Strain NMRI—exposed Swiss Webster mice , NR-21963 ) . Adult worms were perfused at 6–7 weeks post infection as described [59] . Perfused worms were maintained in RPMI ( Life Technologies , Inc . , Grand Island , NY ) plus 10% FBS ( Sigma-Aldrich ) and 1% penicillin/streptomycin ( Sigma-Aldrich ) at 37°C and 5% CO2 . Schistosomula were obtained by in vitro transformation of cercariae [59] and maintained in the same culture conditions as adults . There is a single gene in the S . mansoni genome ( Smp_125690 ) predicted to code for a TRPA channel , which we have dubbed SmTRPA . However , the coding region found in the database for SmTRPA appears to be incomplete . Though the predicted SmTRPA protein contains the series of ankyrin repeats that define TRPA channels , a large portion of the channel domain itself is missing . We used 5' and 3' RACE on S . mansoni RNA to obtain a complete coding region . The total RNA we used as template was from adult male worms , and was provided by the Biomedical Research Institute ( distributed by BEI Resources , Manassas , VA ) . 500 ng of this RNA was used in the SMARTer RACE 5'/3' Kit ( Clontech , Mountain View , CA ) according to the manufacturer's instructions , with the following gene-specific primers: for 3' RACE , 5' TCAAGGTCCAGGAATCAGAACAGTCCTA 3' and the nested primer 5' CGTGGGGCTTCTGCATTAGAACGTGAT 3'; for 5' RACE , 5' CGTTCTAATGCAGAAGCCCCACGTA 3' and the nested 5' TAGGACTGTTCTGATTCCTGGACCTTGA 3' ( both with the sequence 5' GATTACGCCCAAGCTT 3' at their 5' ends to facilitate In-Fusion cloning to the Clontech pRACE vector , as per the manufacturer's instructions ) , to obtain the full-length coding sequence that now contains the full channel domain . We also used spliced leader primers [60] to verify the 5' end . The 3' end of the cDNA coding region is extended 993 bp beyond base 1936 of the predicted 3' end of Smp_125690 . This additional sequence includes the remaining transmembrane regions of the channel domain . In contrast , the 5' start of the coding region that was predicted in the genome database is the same as that found by us in the cDNA following RACE . However , in addition to the 3' extension , the cDNA also contains three insertions not found in the predicted amino acid sequence in the genome database . These include a ~250 bp ( 82 aa ) insertion following base 319 ( of the genomic predicted sequence ) , a ~200 bp ( 67 aa ) insertion following base 712 , and a 108 bp ( 36 aa ) insertion following base 1764 . The full-length coding sequence was amplified by PCR using high-fidelity Q5 DNA polymerase ( NEB ) , using terminal primers with 5' overlapping vector sequence , cloned into the EcoRV site of the pcDNA3 . 1-zeo vector ( Life Technologies , Inc . ) by In-Fusion cloning ( Clontech ) , and re-sequenced . Adult worms were first placed individually in standard schistosome medium in single wells of a 24-well plate for 15 min . During this period , control motility readings were taken ( see below ) . TRP channel modulators ( or DMSO carrier ) were then added to the medium to appropriate final concentrations , and motility measured again over the course of another 15 min . Each worm thus served as its own control . Serotonin ( 5-HT ) at 40 μM was used as a control for increased motility and 500 nM PZQ served as a control for reduced motility ( i . e . , paralysis ) . Vehicle controls included 0 . 1% DMSO . In vitro-transformed schistosomula were exposed in a similar manner , but contained several worms per well . For inhibition studies with SB 366791 , adult males or female worms were first imaged for their control motility , then pre-incubated for 30 min in antagonist ( with a 15-min measurement of motility ) , followed by addition of capsaicin and imaging for measurement of motility again , as described above . For measurement of adult worm motility , we adapted an imaging system and software used for monitoring of C . elegans locomotor activity [61] . The imaging system consists of a USB monochrome camera ( 2592 x 1944 pixel resolution , DMK 72BUC02 , The Imaging Source , Charlotte , NC ) , a 12 . 5 mm , f/1 . 4 fixed-focus lens ( HF12 . 5SA-1 , Fujinon , Fujifilm , Valhalla , NY ) , a red flexible LED strip for illumination , and other mechanical components , as described [61] . Images were acquired for the entire 24-well plate , each well of which contained a single worm , over the course of 15 min at 15 frames per second , and saved to the hard drive in BMP format using Image Capture software . After completion of image acquisition we used custom MATLAB software to calculate motility for each worm by measuring absolute differences in gray scale values between consecutive images within each region of interest ( eg , each well ) , a metric that is sensitive to any type of movement [61] . For each pair of successive frames , we summed all pixels in which a change in intensity greater than a threshold occurred , yielding a measurement of the amount of motion within the region of interest between the two frames . Using these values , we calculated an average change in pixel values per frame across the 15-minute window for each worm , and normalized that value to 100 for the control worms . For schistosomula , we used a Leica stereomicroscope with Qicam Fast 1394 camera ( Qimaging ) and Q-Image capture software to create videos at 2 frames per second over a 2 . 5-minute time span ( 300 frames ) . These video recordings were then used to analyze motility of individual worms using MaxTraq-Lite+ motion analysis software ( Innovision Systems , Columbiaville , MO ) , as described [62] . We measured the distance between the ends of individual worms as an estimate of body length every 3 frames , and then calculated the change in distance between each measurement ( as the worm moves more rapidly , body length will change with increased frequency ) . Knockdown of RNAs encoding the single S . mansoni homologs of TRPA1 ( SmTRPA1 , Smp_125690 ) and TRPM7 ( SmTRPM7 , Smp_035140 ) , and one of the three S . mansoni TRPM3-like sequences ( SmTRPM3a , Smp_165170 ) in adult worms was as described [63 , 64] . The luciferase siRNA used as a control ( Silencer Firefly Luciferase , GL2+GL3 , Life Technologies , Inc . ) shows no significant similarity to any sequences from the S . mansoni gene database . siRNAs against the S . mansoni TRP channel were designed using the SciTools software suite from IDT . The target sequences were: SmTRPA siRNA , 5'-GAGTTGAAACGTGAAGAGTTATTAATT-3' , which maps to bp 1561–1587 of the predicted coding region ( in the S . mansoni database ) of SmTRPA ( Smp_125690 ) ; SmTRPM7 siRNA , 5'-ACCTGATGAAGAGAATAGTAATTTGAA-3' , corresponding to bp 2792–2818 of SmTRPM7 ( Smp_035140 ) ; and 5'-GGAGTGCATACCAATGCATTTGT-3' , corresponding to bp 2128–2152 of SmTRPM3a ( Smp_165170 ) . Following overnight incubation in schistosome medium , adult worms ( 5 males and 5 females ) were placed in a 0 . 4 cm electroporation cuvette ( USA Scientific , Ocala , FL ) containing 50 μl media plus 5 μg of SmTRP or control luciferase siRNAs ( IDT ) . Worms were electroporated in this solution with a 20 ms square-wave pulse of 125 volts ( Pulser XCell , BioRad , Hercules , CA ) . Following electroporation , worms were incubated in schistosome medium for 2 days , then sorted into a single male or female per well of a 24-well plate These worms were tested for sensitivity to capsaicin and other compounds as described above . Real-time RT-PCR was used to measure levels of knockdown by RNAi . Total S . mansoni RNA was extracted from adult worms using Direct-Zol RNA Mini Prep ( ZYMO Research , Irvine , CA ) according to the manufacturer's instructions . qRT-PCR was performed using the Brilliant II SYBR green qRT-PCR Master kit ( Agilent Technologies , Santa Clara , CA ) on an Applied Biosystems 7500 according to the manufacturer's recommendations and as described [62] . Primers used for the amplification of 18S ribosomal RNA have been described [62] . Primers used for amplification of SmTRPA1 were: TRPA FWDSET1 ( 5′-TCGTCTTGGAGCAAATCCTAAT-3′ ) and TRPA REVSET1 ( 5′-CTGGTAGGACTGTTCTGATTCC-3′ ) . Primers used for amplification of SmTRPM7 were: TRPM7 FWDSET1 ( 5′-GAGAACCCAGTCCAGGTTTAAG-3′ ) and TRPM7 REVSET1 ( 5′-GCTAACATCGGTCGTATCCATT-3′ ) . Data were analyzed using the 2−ΔΔCt method [65] to determine the relative expression ratio between targets ( TRP channels ) and reference genes ( 18S RNA ) . Data were analyzed with GraphPad Prism or Microsoft Excel , expressed as arithmetic means ± SEM , and tested for statistical significance using statistical tests noted in the figure legends . In the drug response studies , each worm served as its own control , and we therefore compared means using paired t-tests ( on the raw data , prior to normalization ) . The time course of motility ( Fig 1C ) was analyzed and plotted using R v . 3 . 13 , ggplot2 package . For the knockdown experiments , we compared motility between knockdown and control worms in each of the drug concentrations; in this case , differences between means were therefore tested using unpaired t-tests .
Capsaicin , a vanilloid compound that is an active ingredient in hot peppers , is a potent and selective activator/enhancer of TRPV1 [37] . Exposure of S . mansoni adult males or females to capsaicin produces a marked increase in worm motor activity that persists for at least 15 min . Fig 1A shows responses of male schistosomes to different concentrations of capsaicin , ranging from 10 μM to 200 μM , using an imaging system that assesses the number of pixels where differences in gray scale values occur between consecutive video frames as a measure of motility [61] . Adult males in capsaicin show a dose-dependent , saturable enhancement of motility , with a ~10-fold increase at 10 μM capsaicin and a 20-30-fold increase in concentrations of capsaicin equal to or higher than 60 μM . Controls for both hyperactivity ( 40 μM 5-HT ) and paralysis ( 500 nM PZQ ) indicate that this system for measuring motility is robust ( Fig 1A ) . Adult female worms also show a dose-dependent enhancement of motility in 10–200 μM capsaicin ( Fig 1B ) . As with males , the response appears to saturate at capsaicin concentrations of 60 μM and higher . Attempts to measure effects of capsaicin on paired male and female worms ( the more biologically relevant condition in which schistosomes reside within the host ) were not possible , as worm pairs rapidly and invariably separate when exposed to capsaicin ( S1 Video ) , perhaps indicative of a nociceptive response . The effects of 10 μM capsaicin on worm motility are sustained through a 15-minute time period , as shown in Fig 1C . Capsaicin ( 20 μM ) also elicits increased activity in in vitro-transformed schistosomula , indicating that different schistosome stages are susceptible to vanilloid compounds that target the TRPV1 channel in mammals . ( Fig 1D ) . Although the change in activity in schistosomula does not appear to be as striking as in adult worms , the two sets of data were obtained using different assays of motility , and direct comparisons are not applicable . Interestingly , consistent with previous findings showing cercaricidal effects of an oil extract from the pepper plant C . annuum [53] , we find that either 60 μM or 100 μM capsaicin effectively abolishes cercarial swimming . The cercariae retain their tails , but instead of swimming , remain largely in one place , moving in a fashion reminiscent of schistosomula ( see S2 Video and S3 Video ) . Resiniferatoxin ( RTX ) , a toxin found in the dried latex of Euphorbia resinifera , acts as an analog of capsaicin , potently activating the TRPV1 channel [37] and binding to the TRPV1 vanilloid binding site [41] . Like capsaicin , RTX ( 1–10 μM ) increases motility in both male and female adult schistosomes , with significant effects seen at concentrations as low as 3 μM ( Fig 2 ) . We tested the potent , highly-selective [66] competitive TRPV1 antagonist SB 366791 to determine whether it blocks worm responses to capsaicin . SB 366791 has been shown to have no effect on AITC-evoked ( ie , TRPA1-mediated ) responses in mice [67] , although a recent report [68] indicates that 10 μM SB 366791 can antagonize responses to AITC in particular leech neurons . As shown in Fig 3 , the significant increases in motility induced by capsaicin do not occur when worms are pre-incubated in concentrations of SB 366791 ranging from 400 nM to 100 μM . Despite indications in Fig 3 that SB 366791 may have motility effects on its own , particularly in females , direct tests ( S1 Fig ) show that the compound exhibits no significant effect on worm motility . These results indicate that in schistosomes , which appear to lack TRPV-like channels , the capsaicin receptor has pharmacological sensitivities that are similar to those of mammalian TRPV1 channels . The S . mansoni genome predicts no TRPV homologs . We therefore initiated experiments to determine if other S . mansoni TRP channels might be mediating the parasite's response to capsaicin . We first examined SmTRPA , the only TRPA1-like gene predicted in the S . mansoni genome . TRPA channels appear to be most closely related phylogenetically to TRPV channels , are frequently co-expressed with TRPV channels , and often serve analogous functions and transduce signals ( nociceptive , thermal , inflammatory , etc . ) similar to those transduced by TRPV channels [28 , 69–71] . Mammalian TRPA1 does not respond to concentrations of capsaicin that activate TRPV1 channels [56 , 57] . Electroporation of adult worms with SmTRPA siRNA reduced SmTRPA expression ~90% in males and ~60% in females ( S2A Fig ) . SmTRPA knockdown significantly attenuated the response to capsaicin in both adult male and female worms . Thus , following knockdown of SmTRPA in males ( Fig 4A ) , the response to capsaicin drops from a 14-fold increase in motility at 20 μM capsaicin to a 3 . 4-fold increase in worms exposed to SmTRPA siRNA . At higher capsaicin concentrations , the effects of knockdown are even more dramatic , with essentially the entire capsaicin-dependent increase in motility eliminated . In females ( Fig 4B ) , knockdown of SmTRPA abolishes effectively the entire response to capsaicin at all concentrations tested ( electroporation with a control luciferase siRNA has no significant effect on capsaicin-dependent hyperactivity in males or females ) . In contrast ( Fig 4C ) , knockdown of SmTRPA expression does not affect stimulation of motility by serotonin [72] , indicating that suppression of SmTRPA expression does not compromise the parasite neuromuscular system nonspecifically; these worms can still respond to agents that evoke hyperactivity through other pathways . To further assess whether the effects of SmTRPA knockdown on capsaicin sensitivity are specific , we used RNAi to knock down -other two S . mansoni TRP channel genes: SmTRPM7 ( Smp_035140 ) , which encodes a TRPM7/M6-like protein; and one of the TRPM3-like genes ( SmTRPM3a; Smp_165170 ) found in the S . mansoni genome . In adult male worms , knockdown of SmTRPM7 ( S2B Fig ) produces no significant effect on capsaicin sensitivity compared to control worms ( S3 Fig ) . Thus , male worms with suppressed SmTRPM7 expression continue to exhibit large ( >15-fold ) increases in motility in response to capsaicin . On the other hand , there is some evidence for an effect of SmTRPM7 knockdown on responses of females to capsaicin , with a significant decrease from controls at 100 μM capsaicin . Nonetheless , these SmTRPM7-knockdown worms continue to exhibit a ~20-fold increase in activity when exposed to 100 μM capsaicin . A double knockdown of SmTRPA and SmTRPM7 ( S2C Fig ) decreases responsiveness to 20 μM capsaicin in males approximately 4-fold ( P<0 . 001 ) and to 60 μM capsaicin approximately 2-fold in females ( P<0 . 05 ) compared to the single SmTRPA knockdowns ( S4 Fig ) , again indicating the possibility of some contribution of other schistosome TRP channels to capsaicin-induced hyperactivity . Interestingly , knockdown of SmTRPM3a does produce a significant effect on capsaicin responsiveness ( S5 Fig ) . The worms still react to the compound , but far less dramatically ( ~4-5-fold vs . 15-25-fold ) . However , in contrast to schistosomes with SmTRPA knocked down , which continue to respond to serotonin ( Fig 4 ) , worms with SmTRPM3a knocked down exhibit a significantly diminished response to serotonin ( S5 Fig ) . These results suggest that SmTRPM3a may play a role in the ability of the schistosome neuromuscular system to respond to a variety of agents that stimulate activity . Knockdown of SmTRPM3a apparently compromises this capability , but is not specific to any particular stimulant . These worms appear to have normal basal levels of motility however . AITC is an organosulphur compound found in mustard and wasabi . It is one of several pungent electrophilic compounds that activate TRPA1 channels through covalent modification of cysteines [73–75] . AITC also activates/sensitizes TRPV1 , though at relatively high concentrations compared with TRPA1 , via a mechanism independent of cysteine modification [76–78] . Adult male and female schistosomes exposed to AITC display a significant increase in activity ( Fig 5 ) . In contrast with capsaicin , to which females exhibit more pronounced responses than males , AITC elicits substantially higher levels of hyperactivity in male worms that in females . When exposed to 10 μM AITC , males show a 4-5-fold increase in motility ( Fig 5 ) , while females show no significant change . At 20 μM AITC , motility increases approximately 10-fold in males compared to control worms , but only slightly ( ~1 . 5-fold ) , though significantly , in females . Interestingly , though 40 μM AITC continues to evoke increasing hyperactivity in females , producing a 2 . 5-fold rise , in males , 40 μM and 60 μM AITC produce a significant ( P<0 . 0001 and P<0 . 001 , respectively ) decrease in levels of hyperactivity compared to the 20 μM concentration . Nonetheless , in males as in females , 40 μM and 60 μM AITC still elicit 2 . 5- and 3 . 5-fold increases in motility over controls , respectively . Interestingly , at 100 μM AITC , the increase in motility in males returns to the level at 20 μM AITC . It is not clear why there is a diminished response at the 40 μM and 60 μM concentrations , but it perhaps reflects a balance of increasing activation and non-selective toxicity . We also tested the effects of suppressing SmTRPA expression on AITC sensitivity in adult worms . As with capsaicin , knockdown of SmTRPA essentially eliminates schistosome sensitivity to AITC ( Fig 5 ) , indicating that SmTRPA expression is required for parasite responsiveness to AITC . The amino acid sequence of SmTRPA ( Smp_125690 ) predicted in the S . mansoni genome database is missing a large portion of the channel domain found in all TRP channels , and thus appears to be incomplete . We therefore cloned the full-length coding region of SmTRPA , using RACE and other protocols ( see Methods ) , extending the predicted open reading frame 993 bp at the 3' end . The cDNA also contains inserted stretches not found in the open reading frame predicted in the genome database . This sequence has been deposited to GenBank , with accession #KT266713 . Analysis of the SmTRPA sequence reveals that it contains many , but not all , of the residues thought to be required for AITC activity . As noted above , a wide range of electrophilic compounds such as AITC react with mammalian TRPA1 via covalent modification of cysteine residues [69 , 79] . The three cysteine residues in mouse TRPA1 that appear to be required for this activity are C415 , C422 , and C622 [75] . In human TRPA1 , C619 ( the equivalent of mouse C622 ) is also crucial , as are C639 and C663 , as well as a lysine residue , K708 [74] . As shown in Fig 6A , SmTRPA contains the equivalent of the C619/622 , C639 , and K708 residues found in mouse and human TRPA1 proteins; C415 , C422 , and C663 are not conserved .
Ion channels are validated targets for anthelmintics , yet only a small subset of this superfamily has been characterized in parasitic helminths . Notably , TRP channels , which are critical to sensory transduction , ion homeostasis , and other cellular functions , are almost entirely unexplored in these organisms . The S . mansoni genome contains 15–20 genes predicted to code for TRP channels that represent nearly all of the major metazoan sub-families . Notably , however , there appear to be no representatives of the TRPV channel sub-family [20 , 26] . Indeed , we have verified the absence of TRPV-like schistosome sequences by querying schistodb . net [30] , which contains current databases for the three major human schistosome species , for TRPV-like sequences , using BLAST ( we have also searched genedb . org and NCBI ) . Consistent with previous reports [20 , 26] , we have never found any schistosome sequences with significant similarity to a TRPV channel , other than some predicted non-TRPV proteins containing homologous ankyrin domains . Though a lack of TRPV channels is unusual , it does appear to occur in some other metazoans as well [28] . In this communication , we show that the S . mansoni TRPA1 channel has seemingly taken on at least some of the pharmacological characteristics of the mammalian TRPV1 channel . Schistosomes show sensitivity to capsaicin and RTX , selective activators of TRPV1 in mammals [37 , 38 , 56 , 57] , as well as the highly-selective TRPV1 antagonist SB366791 , but lose capsaicin sensitivity when SmTRPA expression is suppressed . Functional characterization of SmTRPA , either in a heterologous system , or in dissociated schistosome tissues [80] could confirm this novel pharmacology as well as provide opportunities to define other pharmacological and biophysical properties of this channel . It would also be interesting to determine if TRPA1 channels in other metazoans that lack TRPV channels show TRPA-dependent sensitivity to TRPV1 ligands as well . The question of which structural features of SmTRPA might be responsible for capsaicin sensitivity is intriguing . Particular amino acid residues appear to be required for binding of vanilloid ligands to TRPV1 channels . These include Y511 and S512 in transmembrane segment 3 , and M/L547 and T550 in transmembrane segment 4 [39–41] . Examination of SmTRPA aligned with rat and human TRPV1 and other TRPA1 sequences ( Fig 6B ) reveals that , with one exception , it does not appear to share those residues . The exception is M/L547 , for which the equivalent residue in SmTRPA ( and Drosophila TRPA1 ) is a leucine ( L824 ) , as it is in human ( and rabbit ) TRPV1 . A switch from leucine to methionine at this position in human TRPV1 is associated with a 20-30-fold increase in affinity for RTX [39–43] . Our results raise interesting questions regarding the relationships between different TRP channels , including how these channels have evolved and the level of fluidity between the different TRP channel types . They also could have important implications for development of antischistosomal therapeutics . The parasite neuromuscular system is the site of action for many anthelmintics . More specifically , ion channels , required for normal neuromuscular function , have proven to be outstanding anthelmintic targets . Nonetheless , only a few ion channel families have been investigated in detail in parasitic helminths; TRP channels notably represent one of these ion channel families that has remained largely unexplored . Our results show that expression of schistosome SmTRPA is required for vanilloid-dependent ( and AITC-dependent ) dysfunction ( hyperactivation ) of the parasite neuromuscular system ( though not serotonin-dependent activation ) , perhaps providing an entrée for a novel set of targets for new or repurposed antischistosomals . SmTRPM3a , which appears to play a broader role in stimulation of schistosome motility , may represent another , more general point of attack for compromising normal parasite neuromuscular activity . | Schistosomes , infect hundreds of millions of people worldwide , causing schistosomiasis , a disease with devastating effects on human health and economic development . Despite the prevalence of this disease , there is only a single drug , praziquantel , available for treatment and control . Praziquantel is effective , but has limitations , and reports of drug resistance lend increased urgency to development of new treatments . Ion channels are critical components of animal neuromuscular systems , and have proven to be excellent targets for drugs against parasitic worm infections . TRP ion channels , which play important roles in sensory functions , have received little attention in parasitic helminths . One class of TRP channels , TRPV , is activated by capsaicin . However , schistosomes do not appear to contain any TRPV channels . Nonetheless , we find that they are highly sensitive to capsaicin and similar compounds , responding by dramatically increasing their motor activity . Unexpectedly , suppressing expression of a different schistosome TRP channel , TRPA1 , which in mammals is not sensitive to capsaicin , almost completely eliminates this response . Thus , it appears that the pharmacology of schistosome TRP channels differs from that of host mammalian channels , a characteristic that might be exploitable for development of new antischistosomal drugs . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Evidence for Novel Pharmacological Sensitivities of Transient Receptor Potential (TRP) Channels in Schistosoma mansoni |
Uncoupling protein 2 ( UCP2 ) is a mitochondrial transporter that has been shown to lower the production of reactive oxygen species ( ROS ) . Intracellular pathogens such as Leishmania upregulate UCP2 and thereby suppress ROS production in infected host tissues , allowing the multiplication of parasites within murine phagocytes . This makes host UCP2 and ROS production potential targets in the development of antileishmanial therapies . Here we explore how UCP2 affects the outcome of cutaneous leishmaniosis ( CL ) and visceral leishmaniosis ( VL ) in wild-type ( WT ) C57BL/6 mice and in C57BL/6 mice lacking the UCP2 gene ( UCP2KO ) . To investigate the effects of host UCP2 deficiency on Leishmania infection , we evaluated parasite loads and cytokine production in target organs . Parasite loads were significantly lower in infected UCP2KO mice than in infected WT mice . We also found that UCP2KO mice produced significantly more interferon-γ ( IFN-γ ) , IL-17 and IL-13 than WT mice ( P<0 . 05 ) , suggesting that UCP2KO mice are resistant to Leishmania infection . In this way , UCP2KO mice were better able than their WT counterparts to overcome L . major and L . infantum infections . These findings suggest that upregulating host ROS levels , perhaps by inhibiting UPC2 , may be an effective approach to preventing leishmaniosis .
UCP2 is a mitochondrial carrier expressed in a wide variety of tissues , including white adipose tissue , skeletal muscle and the immune system [1] . UCP2 activity presumably lowers the efficiency of oxidative phosphorylation by increasing the membrane proton conductance . This effect would result in an increased rate of respiration leading to a downregulation of the mitochondrial production of ROS . This uncoupling activity is widely considered as an additional element of the cellular antioxidant defence [2]–[4] . In fact , UCP2 is upregulated in many pathological processes in which ROS play an important role in the development of the disease ( atherosclerosis , cancer , chronic inflammation , etc . ) [2] . In addition , UCP2 has been postulated to help regulation of insulin secretion by pancreatic β-cells [5] . Indeed , loss-of-function mutations of UCP2 have been identified in patients suffering from congenital hyperinsulinism . It has been shown that UCP2 modulates glucose-stimulated insulin secretion by reducing mitochondrial ROS [6] . By decreasing ROS production , UCP2 may weaken the immune system when faced with attack by bacterial and parasitic pathogens . Mice lacking UCP2 showed higher survival rates than WT mice following experimental infection with Listeria monocytogenes [7] or Toxoplasma gondii [8] . A recent in vitro study demonstrated that Leishmania donovani upregulates UCP2 and thereby suppresses mitochondrial ROS production , leading to increased production of anti-inflammatory cytokines and parasite survival inside murine macrophages [9] . The same authors found that silencing UCP2 in vivo in splenic tissue of BALB/c mice 3 days prior to L . donovani infection stimulates ROS production , shifts the balance of pro- vs . anti-inflammatory cytokines towards the pro-inflammatory phenotype and reduces splenic parasite burden after 4–6 weeks of infection . These findings raise the possibility of targeting host UCP2 activity and ROS production as a strategy to develop effective therapies against leishmaniosis . To make this work possible , the role that host UCP2 plays in the host immune response to Leishmania infection should first be investigated in vivo . The leishmaniases comprise a group of diseases caused by infection by several species of intracellular protozoan parasites of the genus Leishmania , which are transmitted by the bite of an infected female phlebotomine sandfly [10] . Infections with L . major and L . infantum are distinguished by clinical manifestations ranging from localized skin ulcers at the site of the sandfly bite ( CL ) , to potentially lethal visceral disease ( VL ) , respectively . The leishmaniases represent a global public health problem , affecting an estimated 12 million people around the world . Indeed , 1 . 5 million new cases of CL and 0 . 5 million new cases of VL are reported in humans each year [11] . Current research has focused on the development of diagnostic methods , drugs and vaccines . Unfortunately , despite considerable progress , treatments are toxic and expensive [12] and no vaccines are available for any form of human leishmaniosis [13] , [14] . The leishmaniases are classified as neglected tropical diseases ( NTDs ) , which still require improved treatment strategies and new prophylactic vaccines [15] . To investigate the role of UCP2 in host immune defense against intracellular pathogens , we explored in detail the outcomes of CL and VL due to L . major and L . infantum infection of WT C57BL/6 mice and UCP2 knockout ( UCP2KO ) C57BL/6 mice . Infection was analyzed in vivo by measurement of footpad swelling , quantification of parasite load and assays for the production of cytokines and Leishmania-specific antibodies . Better understanding of the role of host UCP2 in the immune response to Leishmania parasites should help develop more effective strategies to control disease .
The animal research described in this manuscript complied with Spanish ( Ley 32/2007 ) and European Union legislation ( 2010/63/UE ) . The protocols used were approved by the Animal Care Committee of Complutense University of Madrid . C57BL/6 mice were purchased from Harlan Interfauna Ibérica ( Barcelona , Spain ) . UCP2KO mice with a C57BL/6 background were obtained from JAX mice ( The Jackson Laboratory , Bar Harbor , USA ) . In infection experiments , mice were matched for age and sex ( females , 8–9 weeks old ) . L . major parasites ( clone V1: MHOM/IL/80/Friedlin ) and L . infantum parasites ( MCAN/ES/96/BCN/150 , MON-1 ) were cultured at 26°C in Schneider's medium ( Sigma-Aldrich , Madrid , Spain ) containing 20% heat-inactivated fetal calf serum ( FCS , Sigma-Aldrich ) , streptomycin and penicillin . Soluble Leishmania antigen ( SLA ) was prepared from stationary cultures of L . major promastigotes as previously described [16] . Leishmania pifanoi axenic amastigote strain MHOM/VE/60/Ltrod ( kindly provided by A . A . Pan , The Johns Hopkins School of Hygiene and Public Health , Baltimore , Maryland 21205 , USA ) was cultured at 32°C in M199 medium supplemented with 20% heat-inactivated FCS , 2 . 5% ( w/v ) trypticase peptone , 13 . 8 mM d-glucose , 76 µM hemin , and 48 µg/mL gentamicin sulfate at pH 7 . 2 [17] . Three days after intraperitoneal injection of 1 ml of 4% thioglycollate ( Difco Laboratories , Detroit , MI ) , cells were obtained by peritoneal lavage with cold phosphate-buffered saline ( PBS ) . Then cells were washed and plated as needed for each experiment in RPMI medium containing 10% heat-inactivated FCS . L . pifanoi amastigotes were labeled using the fluorescent dye CFSE ( Invitrogen , San Diego , CA ) . Briefly , parasites were incubated in PBS at 4×106 parasites/ml CFSE ( 0 . 4 µg/ml ) for 30 min at 32°C . This process does not affect parasite multiplication [18] . To evaluate intracellular killing of L . pifanoi amastigotes we seeded peritoneal macrophages onto LabTek culture chamber incubation slides ( Thermo Scientific , Madrid , Spain ) at 5×104 per chamber . On the following day , adherent macrophages were incubated for 4 h with L . pifanoi amastigotes in the ratio 4∶1 . Then cells were washed until all free parasites were removed . After 24 h , confocal images were acquired using a confocal laser scanning microscope ( Leica TCS – SP2 ABOS ) , and the infection rate was assessed by counting both the number of infected macrophages and the number of parasites per infected macrophage . ROS production in peritoneal macrophages isolated from UCP2KO and WT mice was estimated from the rate of fluorescence increase of the cell-permeant redox-sensitive fluorescent indicators dihydroethidium ( DHE ) and 2′ , 7′-dichlorodihydrofluorescein diacetate ( DCFDA ) . Cells ( 25×103 per well ) were seeded in 96-well plates in RPMI containing 10% FCS . When required , 4 h before the determination , macrophages were infected with L . pifanoi at a ratio 3∶1 and incubated at 32°C in RPMI without FCS and phenol red . After the incubation , plates were transferred to a Varioskan Flash microplate reader and the fluorescent probe of interest added to each well . Fluorescence intensity was recorded for 50 min , with measurements every 2 min , and the maximum rate of fluorescence increase calculated after subtraction of the rate obtained in the absence of cells . The rate of hydrogen peroxide formation was estimated with the probe DCFDA ( 10 µM , excitation 485 nm , emission 535 nm ) and the rate of superoxide formation with the probe DHE ( 50 µM , excitation 535 nm , emission 610 nm ) . To measure NO release , peritoneal macrophages were plated at 5×104 per well in 96-well plates in the presence of 10 ng/ml recombinant mouse IFN-γ ( Sigma-Aldrich ) . After 24 h , parasites were added and NO release was measured 24 h later using Griess reagent as described below . To assay interleukin-12 ( IL-12 ) , peritoneal macrophages were cultured at 2×105 cells per well in 24-well plates and incubated in the presence of 10 ng/ml recombinant mouse IFN-γ for 24 h , as previously described [19] . Subsequently macrophages were stimulated with lipopolysaccharide ( LPS , 20 ng/ml ) and infected with Leishmania parasites ( L . pifanoi , L . major and L . infantum , parasite/cell ratio = 5∶1 ) . After 24 h , culture supernatants were harvested and IL-12 p40 secretion was assayed by ELISA ( BD Biosciences , Madrid , Spain ) , according to the manufacturer's instructions . We obtained bone marrow stem cell progenitors from the femurs and tibiae of naïve C57BL/6 mice and cultured the cells in the presence of 20 ng/ml murine granulocyte macrophage colony-stimulating factor ( GM-CSF; PeproTech , London , UK ) , as previously described [20] . We added fresh medium containing GM-CSF to the cell cultures every 3 days , generating many CD11c+ DCs largely free of granulocyte and monocyte contamination . On day 7 , DCs were plated at 1×106 cells/ml in 6-well plates and primed in the presence or absence of SLA ( 50 µg/ml ) . DCs were collected at 24 h after pulsing with SLA and used for in vitro stimulation of T cells as described below . Experimental CL was initiated by subcutaneous injection of 5×103 metacyclic L . major promastigotes in 30 µl PBS into the left footpad ( at least 6 mice per group ) . Metacyclic forms had previously been isolated from stationary cultures by negative selection using peanut agglutinin ( Vector Laboratories , Barcelona , Spain ) [21] . Footpad thickness was measured weekly with a metric caliper . Draining lymph nodes ( DLNs ) and spleens from euthanized mice were removed after 4 and 12 weeks of infection and subjected to a limiting dilution assay . Experimental VL was initiated by intravenous injection of 5×105 L . infantum promastigotes in 100 µl PBS ( at least 6 mice per group ) . Spleens and livers from euthanized mice were removed after 8 weeks of infection and subjected to a limiting dilution assay . In addition , NO release , arginase activity , cytokine profile and humoral response were determined in both CL and VL experimental models , as described below . Parasite burdens in DLNs , spleens and livers were determined by limiting dilution culture [22] . Briefly , organs were harvested aseptically and homogenized in 1 ml of Schneider's medium ( Sigma-Aldrich ) containing 20% FCS , streptomycin and penicillin . Four-fold serial dilutions were carried out in 96-well culture plates and incubated at 26°C . After culturing for 10 days , the highest dilution yielding growth of viable Leishmania promastigotes was determined using a phase contrast microscope . Mice were infected as described above . At 4 weeks after L . major infection , or 8 weeks after L . infantum infection , mice were euthanized and single-cell suspensions of DLNs or spleens were prepared , respectively . Previously , bone marrow-derived DCs from naïve C57BL/6 mice had been generated and pulsed with SLA ( 50 µg/ml ) , as described above . T cells were washed , resuspended at a final concentration of 2×106 per ml in complete Dulbecco's modified Eagle's medium ( DMEM supplemented with 10% heat-inactivated FCS , 2 mM l-glutamine , 100 U/ml penicillin , and 100 µg/ml streptomycin ) , and plated at 1 ml/well in 24-well plates . T cells and DCs were mixed at a ratio of 5∶1 and cultured for 72 h . Culture supernatants were harvested and ELISA was used according to the manufacturer's instructions to assay secretion of the following mouse cytokines: IFN-γ ( Diaclone , Madrid , Spain ) , IL-17 ( R&D Systems , Madrid , Spain ) , IL-4 ( eBioscience , Barcelona , Spain ) , IL-10 ( Diaclone ) and IL-13 ( R&D Systems ) . Concentration of nitrite , which is a byproduct of nitric oxide ( NO ) production , was measured in the culture supernatant after 72 h using the Griess assay as described [23] . Briefly , 100 µl of culture supernatant was mixed with an equal volume of Griess reagent ( Sigma-Aldrich ) and incubated at room temperature for 10 min . Absorbance was then measured at 540 nm and nitrite concentration calculated by comparison with a standard curve of serial dilutions of sodium nitrite . After removing supernatants to measure NO release at 72 h , cells were incubated for 30 min in lysis buffer ( 0 . 1 M Tris-HCl , pH 7 . 5 , 300 µM NaCl , 1 µM PMSF , 1% Triton X-100 ) . Lysates were then assayed for intracellular arginase activity as previously described [24] , [25] . One unit of enzyme activity was defined as the amount of enzyme that catalyzes the formation of 1 mmol of urea/min . Leishmania-specific antibodies were quantified by ELISA . Standard plates were coated overnight at 4°C with 100 µL of SLA ( 4 µg/mL ) diluted in PBS . Then wells were washed with PBS supplemented with 0 . 05% ( v/v ) Tween-20 and blocked with 2% ( w/v ) bovine serum albumin ( BSA ) in PBS . Sera were serially diluted in order to determine the titer , which was defined as the inverse of the highest serum dilution factor giving an absorbance >0 . 2 . Peroxidase-conjugated goat anti-mouse IgG1 and IgG2 were used as secondary antibodies ( 1∶2500 and 1∶500 , respectively; SouthernBiotech , Madrid , Spain ) . After washing and adding peroxidase substrate ( Ultra TMB-ELISA , ThermoScientific , Madrid , Spain ) , the reactions were stopped with 2 M sulfuric acid . Finally , sample absorbance was measured at 450 nm . Since measurements showed a standard normal distribution , Student's t test was used to evaluate the significance of differences between means of the experimental groups . Differences were considered significant when P<0 . 05 . Statistical analyses were performed using SigmaPlot ( version 12 . 2 , Systat Software ) .
In this study , in vitro results will be reported primarily for the axenic Leishmania pifanoi amastigote line , whose similarity to the amastigotes isolated from infected macrophages has been exhaustively demonstrated [26] . To investigate the ability of macrophages from UCP2KO and WT mice to eliminate L . pifanoi amastigotes in vitro , we determined infection rates in macrophages . As demonstrated by confocal imaging ( Fig . 1A ) , the percentage of infected macrophages and the number of parasites per macrophage after 48 h were found to be lower in UCP2KO macrophages than in WT ones ( Fig . 1B–C ) . Our data showed that the rate of superoxide formation in macrophages from UCP2KO mice was higher than in those from WT mice . In contrast , hydrogen peroxide levels seemed to be identical and did not vary upon infection with L . pifanoi ( Fig . 2 ) . Intracellular killing of Leishmania amastigotes by macrophages requires production of both IL-12 and NO [27] . Therefore we compared IL-12 and NO production in macrophages from UCP2KO and WT mice following in vitro infection with L . pifanoi , L . major or L . infantum . Peritoneal macrophages were primed with IFN-γ before the addition of LPS and parasites , as this is known to be required for considerable production of IL-12 by macrophages [28] , [29] . We observed that in vitro infection with any of the Leishmania species did not suppress IL-12 p40 production . In addition , IL-12 release revealed no significant difference between peritoneal macrophages from UCP2KO and WT mice ( Table 1 ) . In contrast , in vitro infection with any of the three Leishmania species increased nitrite production in macrophages from UCP2KO mice but not in macrophages from WT mice ( Table 2 ) . Disease progression in mice infected with L . major was evaluated by weekly monitoring of the appearance and extent of footpad swelling . Infection led to significantly smaller footpad swelling ( after 6 weeks of infection ) and lower parasite burden in DLNs in UCP2KO mice than in WT mice ( Fig . 3A , 3B ) . Lesions ultimately resolved at the same time in both groups . We also analyzed parasite loads in spleen in order to investigate systemic infection . We did not detect L . major parasites in spleens of UCP2KO mice after 12 weeks of infection ( Fig . 3C ) , while 50% of WT mice had not yet eliminated the parasites by this time . It is important to point out that C57BL/6 mice are resistant to cutaneous L . major infection and develop small , self-resolving lesions [10] . To investigate whether the increased resistance of UCP2KO mice to L . major infection correlates with arginase and iNOS activity , we measured arginase enzymatic activity and levels of nitrite in the popliteal lymph nodes that were draining the infection site after 4 weeks of infection . Arginase activity in the draining popliteal lymph node was significantly lower in UCP2KO mice than in WT mice ( Fig . 3D ) . As expected , arginase activity and NO production were undetectable in uninfected UCP2KO and WT mice ( data not shown ) . We also measured iNOS activity indirectly by assaying nitrite production after stimulating DLN cells with SLA-pulsed DCs . Nitrite levels were significantly higher in DLN cells from UCP2KO mice than in DLN cells from WT mice ( Fig . 3E ) . The Th1/Th2 paradigm of resistance/susceptibility to intracellular infection is based largely on studies using L . major . Th1 cells secrete cytokines that promote cell-mediated immunity , while Th2 cells secrete cytokines like IL-4 that induce antibody production [30] . Leishmania is an intracellular pathogen that avoids the humoral defenses of the host immune system . Thus , anti-Leishmania IgG1 antibody production fails to protect against this parasite and contributes to disease progression [31] . To characterize the humoral immune response to L . major infection , we measured levels of anti-SLA IgG1 and IgG2a antibodies in sera from both groups of mice ( Fig . 3F ) . Leishmania-specific IgG1 antibody levels were significantly lower in UCP2KO mice than in WT mice after 4 and 12 weeks of infection . Nevertheless , the mean Th1/Th2 humoral ratio after infection was similar between WT and UCP2KO mice . In order to determine what type of immune response is elicited after L . major infection , we assayed IFN-γ , IL-17 and IL-4 production in popliteal DLNs after in vitro stimulation with SLA-pulsed DCs . After 4 weeks of infection , DLN cells of L . major-infected UCP2KO mice produced significantly lower levels of SLA-specific IL-4 than those of WT mice , but they produced significantly larger amounts of SLA-specific IFN-γ and IL-17 ( Table 3 ) . To investigate the effects of host UCP2 deficiency on L . infantum systemic infection , we evaluated parasite loads in the spleen and liver by limiting dilution after 8 weeks of infection . Parasite loads in both organs were significantly smaller in infected UCP2KO mice than in infected WT mice ( Fig . 4A ) . Anti-SLA IgG1 antibody levels were significantly lower in UCP2KO mice than in WT mice after 4 and 12 weeks of infection . However , IgG2a antibody levels were similar between both groups of mice ( Fig . 4B ) . To examine whether host UCP2 deficiency might play another role in the VL model , we evaluated the pattern of cytokine production in the spleen in both groups of mice . Spleens of UCP2KO mice showed significantly higher levels of IFN-γ and IL-13 than the spleens of WT mice ( Table 4 ) . However , IL-10 levels were similar between the two groups of mice .
Since mitochondrial UCP2 has been proposed to negatively regulate ROS levels and thereby help protect against oxidative damage , we have decided to investigate the ability to mount responses against Leishmania parasites in macrophages from UCP2KO and WT mice . Protozoan parasites of the genus Leishmania reside and replicate predominantly within macrophages , cells involved in the innate immune response . It has recently been suggested that Leishmania upregulates UCP2 after infection in order to decrease host ROS levels and thereby suppress macrophage defense machinery [9] . In agreement with previous studies using L . pifanoi [32] , a member of the Leishmania mexicana complex causing CL in the New World [26] , we consider that infection of macrophages with axenically cultured L . pifanoi amastigotes represent a suitable in vitro model to study mechanisms of phagocytic uptake , killing activity and ROS production in mammalian cells . In laboratory experiments ( unpublished data ) , we found no significant differences in the early phagocytosis , 2 h and 4 h after infection with L . pifanoi , between UCP2KO and WT peritoneal macrophages . Nevertheless , we demonstrate here an enhanced resistance ( Killing activity at 24 h ) of UCP2-deficient macrophages to infection with the intracellular protozoan L . pifanoi . In addition to intracellular killing of parasites , we assessed ROS production by peritoneal macrophages infected with L . pifanoi amastigotes , the intracellular form of the mammalian stage of the parasite life . These results are of great mechanistic importance since the activity of UCP2 should lower superoxide formation at the mitochondrial respiratory chain [4] . Therefore , it suggests that mitochondrial ROS formation is an important contributor to resistance of UCP2KO mice to L . pifanoi infection and it would be in line with the reported resistance to infection with Toxoplasma gondii [8] . We hypothesize that our results reflect that natural synergy between increased ROS production and increased nitrite production was greatly enhanced in macrophages from UCP2KO mice . Several studies have addressed the role of ROS and NO in killing of Leishmania parasites [33]–[35] . Our present work showed that thioglycollate-elicited macrophages population from peritoneum of both WT and UCP2KO mice , stimulated by IFN-gamma and LPS , produced IL-12 p40 independently of Leishmania infection . Furthermore , none of the three Leishmania species analyzed in our study ( L . pifanoi , L . major or L . infantum ) was able to inhibit IL-12 p40 production by peritoneal macrophages during the early stage of infection ( 24 h ) , as characterized in previous studies [36] . The early induction and maintenance of IL-12-dependent IFN-gamma production is critical to the host resistance to Leishmania infection [36] , [37] . The IFN-γ induced production of NO is not only crucial for the direct leishmanicidal activity of macrophages but also triggers the IL-12 signaling cascade that leads to further IFN-γ production by natural killer cells [38] . Regarding IL-12 production , it has been suggested that Leishmania parasites of different species are potent inhibitors of macrophage IL-12 production [29] , [39] , [40] . However , the ability of macrophages to produce IL-12 seems to be related to their maturation stage [36] , [41] . Also , this capacity varies depending on the stage of Leishmania infection [36] . We found NO , a principal effector molecule that mediates the intracellular killing of Leishmania parasites [42] , to be present at higher levels in UCP2KO peritoneal macrophages than in WT ones after infection with any of three Leishmania species . Therefore , our study provides in vitro evidence that macrophages from UCP2KO mice have developed several mechanisms to combat intracellular Leishmania pathogens more efficiently . These include not only maintenance of IL-12 p40 production throughout infection that is required to prevent the loss of Th1 cells [37] , but also significant high amounts of ROS and NO after Leishmania infection . Subsequently we found that UCP2KO C57BL/6 mice mounted a stronger host defense against L . major and L . infantum infections in vivo than did the corresponding WT mice . In contrast to BALB/c mice , C57BL/6 mice after L . major infection develop self-healing cutaneous lesions and control parasite multiplication [43] . It has been reported that CL in C57BL/6 mice occurs in two phases following infection in the ear: a remarkably silent phase lasting about 4 weeks favors parasite amplification in the dermis without the formation of a lesion , followed by the development of a cutaneous lesion that coincides with a reduction in parasite load at the site of infection [44] . In our study mice were inoculated in the footpad and the evolution of infection occurred in the two mentioned phases above . UCP2KO C57BL/6 mice were significantly better than their WT counterparts at controlling L . major: after 6 weeks of infection , the UCP2KO mice showed significantly smaller footpad swelling and lower parasite burden . Nevertheless , the lesions in the UCP2KO mice resolved at the same time as those in WT mice , suggesting that lack of UCP2 accelerated the development of an adaptive immune response at the level of parasite load but not at the level of lesion resolution . This may reflect the intrinsic features of the resistant C57BL/6 strain , which , like humans , develops self-healing skin lesions following L . major infection [45] , [46] . Taken together , our data showed that UCP2 deficiency was dispensable for the resolution of the lesions at the site of infection , but was advantageous to achieve an early reduction of parasite load in the DLN and was essential for efficient parasite clearance in the spleen after 12 weeks p . i . These results are consistent with previous research showing that the capacity to kill intracellular L . major parasites was dramatically reduced in mice deficient in the production of reactive oxygen intermediates ( ROI ) [34] . In addition , it is well established that NO acts as a leishmanicidal molecule [42] , interfering with arginase activity in infected host cells and subsequently with cellular metabolism of Leishmania . Inhibition of arginase activity during infection has a clear therapeutic effect , as evidenced by markedly reduced pathology and efficient control of parasite replication [47] . In the present study , arginase activity was lower and leishmanicidal nitrite levels were higher in phagocytic cells from UCP2KO mice than in such cells from WT mice . This reflects greater intracellular killing of Leishmania parasites in UCP2KO mice . Because susceptibility to Leishmania infection correlates with the development of CD4+ Th2 cells , whereas resistance correlates with the development of CD4+ Th1 cells [43] , we compared the production of the cytokines IFN-γ , IL-17 and IL-4 in DLNs from L . major-infected UCP2KO and WT mice after in vitro stimulation with SLA-pulsed DCs . We found that infected UCP2KO mice showed a higher Th1 and Th17 anti-Leishmania immune response and a weaker Th2 response than did WT mice . Consistent with our findings , UCP2KO mice infected with another intracellular pathogen , Listeria monocytogenes , showed high levels of ROS in the spleen , and this was associated with a switch away from anti-inflammatory cytokine production toward proinflammatory cytokine production , as well as with a significant increase in recruited splenic phagocytes [7] . We further found UCP2KO mice to have elevated IL-17 levels , which is consistent with previous reports that acquisition of a resistant phenotype in both VL and CL murine models is related to Th1- and Th17-mediated , parasite-specific cellular responses [48] , [49] . In addition , we assessed the relative production of IgG2a and IgG1 isotypes , since these have been widely used as markers , respectively , of the induction of Th1- and Th2-type immune responses [50] . Indeed , anti-Leishmania IgG1 antibody production fails to protect against this intracellular pathogen and contributes to disease progression [31] . In accordance with the notion that UCP2KO mice are highly resistant to Leishmania infection , we found not only higher levels of Leishmania-specific IgG1 antibodies in WT mice than UCP2-deficient ones , but also higher IL-4 levels in DLNs from WT mice . Th2 IL-4 cytokine has been reported to induce antibody production [30] . Taken together , these data support our hypothesis that UCP2KO mice have developed increased resistance to L . major that led to a faster resolution of the infection . The visceral form of the disease ( VL ) was also evaluated in order to investigate the role of UCP2 in host immune defense . C57BL/6 and BALB/c strains of mice susceptible to L . infantum infection possess a non-functional Slc11a1 gene , preventing them from controlling early parasite growth in the liver [21] , [51] , [52] . However , during later stages of L . infantum infection , these strains develop mature granulomas that can eliminate the parasites in the liver [53] , [54] . It has been reported that while ROS are not crucial for the activation of leishmanicidal abilities of hepatic mature granulomas during later stages of VL , NO and ROS do contribute to parasite control [33] . A recent study of murine VL reported that granuloma maturation is delayed in mice deficient in IL-13 due to defective IFN-γ production and elevated IL-4 and IL-10 levels [55] . That study suggested that IL-13 plays a crucial role in ensuring efficient hepatic granuloma maturation to control parasite load during VL [55] . Therefore we evaluated the pattern of cytokine production by the spleen in UCP2KO and WT mice in the present study . Our finding that IFN-γ and IL-13 levels in the spleen were significantly higher in UCP2KO mice supports the notion that they are better able to control L . infantum infection . This is consistent with a recent study in which siRNA-mediated silencing of the UCP2 gene in a mouse model of VL stimulated mitochondrial ROS production , which induced a proinflammatory cytokine response and subsequently reduced parasite burden [9] . Also , anti-Leishmania IgG1 antibody production , associated with increased susceptibility to the disease as mentioned above , was significant higher in WT mice as compared to UCP2-deficient ones . In conclusion , our study for the first time provided in vivo evidence that UCP2 deficiency was beneficial to achieve an early control of parasite growth in the DLN and was critical to avoid visceralization of the parasite during CL . Also , it does contribute to parasite control in murine model of VL . The role of UCP2 as part of the cellular anti-oxidant defence is well established . Since alterations in UCP2 levels are involved in a number of pathologies , this protein is currently considered as a drug target [2] . A good example comes from the cancer field and , thus , it has been shown that UCP2 levels are elevated in drug-resistant tumour cells and that UCP2 inhibitors sensitize cancer cells to chemotherapeutic agents [3] . In our work , we demonstrate that knock-out of the UCP2 gene allows mice to respond better to CL and VL . Therefore , it can be envisaged that UCP2 inhibitors may improve efficacy of combined therapies against leishmaniosis . | The leishmaniases comprise a group of diseases caused by infection by several species of intracellular protozoan parasites of the genus Leishmania , which are transmitted by the bite of an infected sandfly . The leishmaniases represent a global public health problem , affecting an estimated 12 million people around the world and ranging from self-healing skin lesions to potentially fatal systemic infections . Here , we use mouse models of CL and VL to investigate the effect of a host gene called UCP2 . Uncoupling protein 2 ( UCP2 ) is a mitochondrial carrier expressed in a wide variety of tissues , including white adipose tissue , skeletal muscle and the immune system . Intracellular pathogens such as Leishmania upregulate UCP2 and may weaken the immune system . Consequently , parasites survive and multiply within mouse macrophages . We demonstrate here that mice lacking the UCP2 gene are better able than WT mice to control both CL and VL . Infection was analyzed in vivo by measurement of footpad swelling , quantification of parasite load and assays for the production of cytokines and Leishmania-specific antibodies . These findings could have important implications in designing an effective approach to preventing leishmaniosis . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"immunology",
"microbiology",
"host-pathogen",
"interaction",
"parasitic",
"diseases",
"parasitology",
"animal",
"models",
"model",
"organisms",
"neglected",
"tropical",
"diseases",
"veterinary",
"science",
"infectious",
"diseases",
"zoonoses",
"microbial",
"pat... | 2013 | UCP2 Deficiency Helps to Restrict the Pathogenesis of Experimental Cutaneous and Visceral Leishmaniosis in Mice |
Dengue is a reportable disease in Brazil; however , pregnancy has been included in the application form of the Brazilian notification information system only after 2006 . To estimate the severity of maternal dengue infection , the available data that were compiled from January 2007 to December 2008 by the official surveillance information system of the city of Rio de Janeiro were reviewed . During the study period , 151 , 604 cases of suspected dengue infection were reported . Five hundred sixty-one women in their reproductive age ( 15–49 years ) presented with dengue infection; 99 ( 18 . 1% ) pregnant and 447 ( 81 . 9% ) non-pregnant women were analyzed . Dengue cases were categorized using the 1997 WHO classification system , and DHF/DSS were considered severe disease . The Mann-Whitney test was used to compare maternal age , according to gestational period , and severity of disease . A chi-square test was utilized to evaluate the differences in the proportion of dengue severity between pregnant and non-pregnant women . Univariate analysis was performed to compare outcome variables ( severe dengue and non-severe dengue ) and explanatory variables ( pregnancy , gestational age and trimester ) using the Wald test . A multivariate analysis was performed to assess the independence of statistically significant variables in the univariate analysis . A p-value<0 . 05 was considered statistically significant . A higher percentage of severe dengue infection among pregnant women was found , p = 0 . 0001 . Final analysis demonstrated that pregnant women are 3 . 4 times more prone to developing severe dengue ( OR: 3 . 38; CI: 2 . 10–5 . 42 ) . Mortality among pregnant women was superior to non-pregnant women . Pregnant women have an increased risk of developing severe dengue infection and dying of dengue .
Since the reintroduction of DENV-1 in 1986 in RJ , dengue has become a major public health problem in Brazil [1] . The occurrence of dengue fever ( DF ) and dengue hemorrhagic fever ( DHF ) has increased over the past several years in Brazil , in part due to the rapid spread and simultaneous circulation of the DENV-1 , DENV-2 , DENV-3 [1] . In 2008 , over 600 , 000 cases of DF and 4 , 455 cases of DHF were reported in Brazil , with 40% and 42% , respectively , occurring in the state of RJ [2] , [3] . A surveillance information system of reportable diseases , SINAN , was implemented in Brazil in the early 1980s [4] , and since then , dengue has been a compulsory reportable disease . However , pregnancy was a reportable item on the form only after 2006 . Globally , there are increasing reports of dengue during adulthood , increasing the risk for dengue during pregnancy . In the literature only approximately 400 cases of dengue during pregnancy have been reported , primarily describing the maternal and fetal outcomes [5] , [6] . If diseases such as malaria and cholera are more severe during pregnancy , would dengue also be more severe ? During the 2007/2008 epidemic in the city of RJ , the highest rate of laboratory-positive dengue samples was among those in the age group under 15 years , followed by those 15–29 years; 99% of all births during this period occurred in mothers aged 15–49 years [7] . To estimate the severity of maternal dengue , the available data provided by SINAN related to the epidemic period of January 1 , 2007 , through December 31 , 2008 , in the city of RJ , were reviewed . Laboratory-confirmed dengue cases in reproductive-age women ( 15–49 years ) were included . Mortality and severity of the disease were compared between pregnant and non-pregnant women .
A suspected dengue case is routinely reported to SINAN within 24 hours of attendance in a healthcare unit , using a standardized form [8] . When the laboratory results are available , the form is completed by a health staff member who reviews the chart information and adds the final dengue classification , usually after a period of no more than 3 months . Suspected cases are reported from all healthcare facilities in RJ . The SINAN form includes information on basic demography , laboratory data , hospitalization and outcomes ( death or cure ) . Dengue cases are classified according to the WHO 1997 [9] , adapted by the Brazilian Ministry of Health to include the category of dengue with complications [10] for the cases that do not fulfill all three criteria for DHF . Laboratory confirmed cases were considered when either virus isolation , PCR testing , paired IgM or IgG testing or single IgM test was positive . Pregnancy is categorized in the SINAN form according to trimester: 1st trimester ( up to 14 weeks of gestation ) , 2nd trimester ( 14–28 weeks ) , 3rd trimester ( after 28 weeks ) or unidentified gestational age . Patients were categorized according to the WHO 1997 classification system as DF , DHF or DSS [10] . Dengue classification of patients ( n = 117 ) categorized in the SINAN form as ‘dengue with complications’ were reviewed . If patients had evidence of plasma leakage they were categorized as having DHF/DSS and thus considered as severe cases . Otherwise , patients were categorized as having DF . The Mann-Whitney U test was used to test the difference between the mean age of pregnant and non-pregnant women and the difference between the mean age of pregnant women by dengue classification ( DF and DHF/DSS ) . A chi-square test was used to evaluate the differences in the proportion of dengue severity between pregnant and non-pregnant women . A p-value of <0 . 05 was considered significant in all statistical tests . A univariate analysis was performed using DHF/DSS ( dependent variable ) and pregnancy , maternal age ( as a continuous variable ) and trimester ( independent variables ) using the Wald test . Multiple logistic regression analysis was used to determine whether statistically significant variables were independently associated with dengue severity . Variables with a p-value<0 . 05 in the univariate analysis were included in the multivariate analysis . Finally , the residuals of the fitted model were analyzed . With this modeling , the odds ratio and their respective confidence intervals ( 95% ) were obtained . All statistical analyses of data were performed using R software , version 2 . 11 . 1 . Our study was reviewed and approved by the Ethical Committee of the Municipal Secretary of the City of Rio de Janeiro: Comitê de Ética em Pesquisa da Secretaria Municipal de Saúde e Defesa Civil . Protocolo de pesquisa: 51/08 . CAAE: 0122 . 1 . 314 . 000-08 e 0130 . 1 . 314 . 000-08 . Inform consent was not obtained because the data were analyzed anonymously .
The incidence of laboratory confirmed dengue among women in reproductive age was 234/100 , 000 inhabitants/2y , with similar rates between pregnant ( 238/100 , 000 ) and non-pregnant women ( 233/100 , 000 ) . Mortality of dengue was 3 , 6/100 , 000 inhabitants/2y among pregnant women and 1 , 7/100 , 000 inhabitants/2y among non-pregnant women . Case fatality rate was 7 , 4 and 1 , 5% respectively . Data on 546 eligible reproductive-age women who had confirmed cases of dengue were analyzed: 99 ( 18 . 1% ) were pregnant and 447 ( 81 . 9% ) were not ( table 1 ) . The mean ( ± standard deviation ) maternal age was significantly different: 26 . 3±8 . 5 years in pregnant women compared with 31 . 5±10 . 7 years in non-pregnant women ( p<0 . 05 ) . No significant difference was observed in the mean age between pregnant women with DHF/DSS ( 25 . 5±8 . 2 ) and DF ( 26 . 9±8 . 5 ) . Most cases were classified as DF ( n = 417 , 76 . 4% ) , 123 as DHF ( 22 . 5% ) and 6 as DSS ( 1 . 1% ) . A higher proportion of pregnant women than non-pregnant women had DHF/DSS ( table 1 ) . Hospitalization information available for 186 ( 34 . 1% ) patients occurred in 61 ( 34 . 1% ) pregnant women , and in 118 ( 65 . 9% ) non- pregnant women . The proportion of severe dengue among hospitalized women was similar: 73 . 8% and 66 . 9% for pregnant and non-pregnant women , respectively . Information on death was available for 395 ( 72 . 3% ) of the eligible cases: three pregnant and five non-pregnant women died ( table 1 ) . Shock syndrome ( n = 3 ) and cavity effusion ( n = 2 ) were associated with deaths . The cause of death was unknown in three patients . A higher prevalence of DHF/DSS that increased with gestation age was observed ( table 2 ) . Pregnant women were 3 . 4 times more likely to have DHF/DSS , primarily in the last trimester; OR 3 . 38; CI 2 . 1–5 . 42 ( table 3 ) .
This study suggests that dengue during pregnancy can increase maternal mortality , as previously reported [11] . It also suggests that pregnancy is associated with DHF/DSS and that the susceptibility to severe disease increases with pregnancy age . Severe dengue has been associated with maternal deaths , with fatality rates ranging from 2 . 9%–22% [5]–[6] , [11]–[13] . The maternal dengue fatality in this study was 7 . 4% . The differences in dengue fatality in pregnant women likely result from differences in the designs and in the heterogeneity of the studies sample sizes . Additionally , it may represent different regional management of dengue in pregnant women . More than half of pregnant women were hospitalized and it was twice the rate of hospitalization for non-pregnant women , since it was a recommendation of Rio de Janeiro's healthcare authorities to prevent dengue complications in this group . Moreover , the proportion of DHF could still be underestimated as the identification of plasma leakage syndrome through the hemoconcentration or hypoproteinemia may be compromised from the seventh to the 32rd week of gestation , by the physiological increase of intravascular volume of this period [14] . The reasons for the association of DHF/DSS with pregnancy were not assessed in this study . The amount of vascular leakage during early versus late pregnancy may have different effects on the clinical presentation and on the perceived severity level . The higher risk for developing severe disease in the 2nd and 3rd trimesters should be confirmed by prospective studies as the selection bias related to admission because of risk of preterm delivery cannot be excluded . The non-laboratory confirmed dengue cases were not analyzed to avoid a detection bias , and the confusion of dengue with pregnancy complications , such as HELLP syndrome . The findings of the study are based on a retrospective review of routinely collected data , with laboratory confirmed dengue , which introduces some limitations such as bias resulted from incomplete data and possible misclassification . Although pregnant women were more likely to be hospitalized for fever and illness in general compared to their non-pregnant counterparts , it would be expected a lower frequency of severity among this group as pregnant women had a preventive hospitalization . As all the uncompleted data about death were attributed to non-pregnant women , the mortality rate among pregnant women might still be underestimated . SINAN has also been used in Brazil to conduct studies on dengue [15] . Although citywide surveillance system of information has no specific clinical plasma leakage signs data and may be incomplete , it is a population-base registry from which maternal dengue severity could be inferred by the access to dengue classification . Further longitudinal studies are needed to confirm these findings and to determine on how these two subgroups presents clinically and how their presentations differ . | Dengue represents a major worldwide public health problem . According to the WHO , up to 50 million dengue infections occur each year . The occurrence of dengue fever and dengue hemorrhagic fever has increased in Brazil , in part due to the simultaneous circulation of DENV-1 , DENV-2 and DENV-3 . Although a primary infection with one serotype confers a partial or transient immunity against other serotypes , any subsequent infections harbor the risk of increased morbidity/mortality . Several case reports have been published regarding maternal and fetal outcomes from dengue infection , but it is still inconclusive if pregnancy is associated with severity . To estimate the severity of maternal dengue infection , available data that were compiled from 2007 to 2008 by the official surveillance information system of the city of Rio de Janeiro were reviewed . The cases of dengue were analyzed using the 1997 WHO classification . Pregnant women were 3 . 4 times more prone to developing severe dengue than non-pregnant women . Mortality among pregnant women was superior to non-pregnant women . The increased risk of severe outcomes in pregnant women merits further attention to effective public health and medical interventions that will aid in avoiding morbidity/fatalities within this population . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine"
] | 2013 | Is Pregnancy Associated with Severe Dengue? A Review of Data from the Rio de Janeiro Surveillance Information System |
Bcl-2 family proteins control a decisive apoptotic event: mitochondrial outer membrane permeabilization ( MOMP ) . To discover MOMP-regulating proteins , we expressed a library of intracellular single-chain variable fragments ( scFvs ) ( “intrabodies” ) and selected for those rescuing cells from apoptosis induced by BimS ( the short isoform of Bim ) . One anti-apoptotic intrabody , intrabody 5 ( IB5 ) , recognized pyruvate kinase M2 ( PKM2 ) , which is expressed in cancer cells . PKM2 deletion ablated this clonogenic rescue; thus , IB5 activated a latent cytoprotective function of PKM2 . This resulted not from pyruvate kinase activity per se but rather from the formation of an active tetrameric conformation of PKM2 . A stably tetrameric PKM2 mutant , K422R , promoted cell survival even in the absence of IB5 , and IB5 further increased survival . Mitochondria isolated from IB5-expressing cells were relatively resistant to MOMP in vitro . In cells , IB5 expression up-regulated Mitofusin-1 ( Mfn1 ) and increased mitochondrial length . Importantly , Mfn1 deficiency abrogated IB5’s cytoprotective effect . PKM2’s anti-apoptotic function could help explain its preferential expression in human cancer .
Apoptosis is a cellular suicide process that is important for certain aspects of normal animal development [1] and is dysregulated in various diseases , especially cancer [e . g . , 2 , 3] . Members of the Bcl-2 protein family act at the mitochondrial outer membrane ( MOM ) to regulate the central events in apoptotic cell death [4–13] . Venetoclax , a drug targeting Bcl-2 , is currently approved for the treatment of a refractory form of chronic lymphocytic leukemia [14 , 15] , and other drugs that directly target Bcl-2-family proteins are now in cancer clinical trials [16–18] . Bcl-2 family proteins function in a complex network of heterodimeric interactions that collectively decide between cell survival and death [12] . Several Bcl-2 subfamilies carry out different functions [19] . In particular , the proteins Bax and Bak comprise the effector subfamily responsible for the critical mitochondrial events in cell death . Genetic and in vitro studies [6 , 20–23] have shown that Bax/Bak can be activated by transient interactions with other Bcl-2 family proteins belonging to the “Bcl-2 homology domain 3 ( BH3 ) -only” category ( including Bim , Bid , Puma , and others . ) Once activated , Bax/Bak undergo conformational changes to become fully integrated in the MOM . As a result , these proteins produce large lipidic membrane pores [24 , 25] , in an event known as mitochondrial outer membrane permeabilization ( MOMP ) [13 , 19 , 26] . MOMP allows soluble mitochondrial proteins ( e . g . , cytochrome c , Smac , and Omi ) to escape into the cytoplasm , where they trigger the activation of caspase proteases that carry out the cell death program . MOMP and cell death are decisively regulated by Bcl-2 family interactions [6 , 21 , 27 , 28] , and this underlies the importance of targeting these proteins for cancer therapy . In this regard , Letai and colleagues have shown that the in vitro response of mitochondria from patient tumor samples to BH3 domain peptides can often predict the effect of therapy [29–31] . Bcl-2 family members can also be regulated by proteins outside the Bcl-2 family . For example , p53 can act at mitochondria both to activate Bax directly and to sequester Bcl-xL [32] . Similarly , the Retinoblastoma protein pRB is reported to translocate to mitochondria to promote Bax activation in a nontranscriptional manner [33] , and oncogenes such as Myc and Ras also modulate the expression of key Bcl-2-family proteins [34] . The ability of proto-oncoproteins to inhibit or activate apoptosis is an important facet of their homeostatic function , inasmuch as cell death serves as a critical counterbalance to cell proliferation . To discover molecules regulating the core mechanism of mitochondria-dependent cell death , we developed an unbiased functional selection approach that used libraries of “intrabodies”: intracellularly expressed single-chain antibodies ( scFv ) . We found that some of the selected intrabodies specifically recognized a key metabolic regulatory protein , pyruvate kinase M2 ( PKM2 ) . This suggests that PKM2 , aside from its well-documented role in glycolytic metabolism , could also have an expressly anti-apoptotic function . PKM2 is an important regulator of tissue homeostasis as well as tumor growth and metabolism [e . g . , 35] and is currently a subject of intense research [reviewed in 36 , 37–39] . PKM2 is a glycolytic enzyme that promotes the “Warburg effect , ” also termed aerobic glycolysis , in which cells exhibit increased glucose to lactate conversion even in the presence of oxygen [40] . In cancer cells , PKM2 is typically expressed preferentially over its related isoform PKM1 , even when the tissue of origin does not express PKM2 . Hypothetically , cancers gain some selective advantage from the highly regulated functions of PKM2 . The adaptive metabolic functions of PKM2 also come into play in some cell types that quickly transition to a proliferative state , such as lipopolysaccharide ( LPS ) -activated macrophages [41] . PKM1 and PKM2 are generated from transcripts of the PKM gene by alternative mRNA splicing . Both isoforms can catalyze the last step in glycolysis , in which phosphoenolpyruvate ( PEP ) and ADP are converted to pyruvate and ATP . Isoforms M1 and M2 are identical except for the region encoded by the one alternatively spliced exon ( exon 9 for PKM1 and 10 for PKM2 ) , yielding a difference in only 22 amino acids . PKM1 exists as a constitutively active tetramer , whereas PKM2 is subject to many forms of regulation . Various metabolites , including fructose-1 , 6-bisphosphate ( FBP ) , serine , phenylalanine , and triiodo-L-thyronine ( T3 ) , can allosterically regulate PKM2’s glycolytic activity [40 , 42] . In vitro biochemical studies have shown that PKM2 exists in equilibrium between a glycolytically active tetramer form and less active dimer or monomer forms [43 , 44] . Based on crystallographic data , it has also been proposed that PKM2 tetramers can transition between inactive T-state and active R-state conformations [45] . Paradoxically , it is the ability of PKM2’s glycolytic activity to be reduced that favors rapid cell proliferation . Reduced PK activity correlates with increased biosynthesis of metabolites important for cell proliferation , potentially explaining why tumor cells prefer the M2 isoform [40 , 46] . Consistent with this idea , treatment of cells with small-molecule activators of PKM2 [47 , 48] or the replacement of PKM2 with the constitutively active isoform PKM1 [35] can reduce cell proliferation in some situations . In primary Mouse Embryonic Fibroblasts ( MEFs ) , deletion of PKM2 results in increased PKM1 expression , and this in turn impairs nucleotide availability for DNA synthesis , thereby inhibiting cell cycle progression [49] . PKM2 is reported also to have nonglycolytic functions . Many PKM2 interaction partners have been described , including multiple transcription factors [50] . For example , PKM2 is reported to cooperate with Hif-1α to regulate the transcription of multiple glycolysis-related proteins , which contribute to metabolic remodeling and the Warburg effect [41 , 51–53] . These transcriptional functions require the nuclear import of PKM2 [52–55] . PKM2’s nuclear translocation can be promoted by epidermal growth factor receptor ( EGFR ) activation [56] and regulated by Erk1/2 and JMJD5 [57 , 58] . In the nucleus , PKM2 can promote β-catenin transactivation , leading to the expression of cyclin D1 and tumorigenesis [56] . A PKM2-activating compound , TEPP-46 , which causes PKM2 tetramerization , inhibits Hif-1α–dependent transcriptional effects [41] , supporting the idea that the dimeric form of PKM2 is responsible for transcriptional functions . Dimeric PKM2 is also reported to possess protein kinase activity , targeting multiple oncogenic factors [54 , 59–61] . However , PKM2 protein kinase activity is controversial , as Vander Heiden and colleagues found no evidence of protein kinase activity for PKM2 in cell lysates [62] . In some cases , PKM2 ablation can produce or enhance cell death [63–69] . Precisely how PKM2 affects apoptosis is unclear . PKM2 silencing has been reported to stabilize proapoptotic Bim [70] or down-regulate the expression of the anti-apoptotic proteins Bcl-xL or Mcl-1 [71 , 72] . However , PKM2 knockdown produces an artificial situation . PKM2 has multiple functions that may be regulated independently , and experiments in which this protein is ablated would involve a simultaneous loss of all of these activities , along with a compensatory up-regulation of PKM1 , making interpretation difficult . In contrast to the studies just mentioned , Sabatini and colleagues showed that the inhibition of PKM2 activity under ischemic conditions had the effect of promoting cell survival rather than cell death [73] . The cells bordering necrotic foci in gliomas expressed higher levels of the enzyme SHMT2 , leading to an allosteric inhibition of PKM2’s glycolytic activity . This provided a significant protection from ischemic cell death . In another ischemia model , these authors found that overexpression of PKM2 or treatment with the PKM2-activating compound TEPP-46 eliminated the increased cell viability produced by SHMT2 . It is unclear whether this connection between reduced PKM2 activity and survival is a general phenomenon or only applies to certain cancer cell subsets or environments . In contrast to studies emphasizing PKM2 loss of function , our results now show that PKM2 possesses a positive cytoprotective function that can be activated by a PKM2-specific intrabody We show that this latent function of PKM2 counteracts the central Bax/Bak-dependent mitochondrial apoptotic mechanism . Moreover , the stably tetrameric mutant PKM2 ( K422R ) supported intrabody 5 ( IB5 ) ’s cytoprotective effect , arguing that the anti-apoptotic function involves the cytoplasmic tetramer form rather than the nuclear dimer form of PKM2 . The K422R mutant also produced BimS resistance in MEFs after expression for several passages , even in the absence of IB5 . This mutant’s ability to counteract the central apoptotic pathway could provide a selective advantage for these cells , and indeed , this mutation was found to be spontaneously selected in Bloom syndrome patient tumor cells . The IB5/PKM2-induced cytoprotective function depended in part on up-regulation of the mitochondrial fusion–related protein Mitofusin-1 ( Mfn1 ) . Therefore , we propose that PKM2 can activate an Mfn1-dependent general anti-apoptotic pathway , which could help explain why human cancer cells often preferentially express the M2 isoform of PK .
Because glucose metabolism can influence cell survival [52 , 80–82] , we asked whether IB5 could rescue cells simply through stimulating PKM2’s glycolytic activity . First , to analyze the antibody’s interaction with PKM2 in vitro , we produced both PKM2 and a monovalent scFv corresponding to IB5 ( scFv 5 ) as recombinant proteins in E . coli . We then mixed these proteins at various ratios and analyzed them by blue native gel electrophoresis ( Fig 4A ) . We found that monovalent scFv 5 strongly increased the tetrameric PKM2 species and shifted up the tetramer band to a degree dependent on the molar input ratio of scFv . Thus , the antibody bound directly to PKM2 , promoting its stable tetramerization . Moreover , increasing the input ratio of scFv:PKM2 gradually altered PKM2’s electrophoretic mobility , apparently reflecting the binding stoichiometry . This altered mobility in native gels might indicate either an increased mass or an altered tertiary conformation of the PKM2 tetramers . We then found that purified scFv 5 stimulated PKM2’s glycolytic activity in a concentration-dependent manner ( Fig 4B ) , confirming indirectly that the scFv interacts with PKM2 . The activity declined at higher concentrations of scFv 5 . This might result from PKM2 aggregation . These data suggest that scFv 5 activates PKM2 allosterically . ( Note also that scFv 5 altered neither the electrophoretic mobility nor the PK activity of PKM1 , consistent with IB5’s specificity for PKM2 . ) Based on these results , we considered the possibility that the anti-apoptotic effect of IB5 purely reflected an increased glycolytic activity of PKM2 . However , we found that treating cells with the PKM2-activating compound TEPP-46 ( Fig 2A and 2C ) [47] alone did not protect cells from BimS-induced death . Similarly , reconstituting PKM2-null MEFs with the constitutively active PKM1 failed to rescue cells ( Fig 2A ) . Thus , high PK activity by itself was insufficient to produce an anti-apoptotic effect . On the other hand , culturing 293T cells in the presence of TEPP-46 enhanced the prosurvival effect of IB5 expression to a modest but statistically significant extent ( Fig 2A and 2C ) . This could mean that increased glycolytic activity does contribute somewhat to cell survival . Alternatively , the effect of IB5 , alone or in combination with TEPP-46 , might result from stabilizing a tetrameric conformation of PKM2 , whose glycolytic activity may be irrelevant to the survival function . To help define the aspects of PKM2 function required for the IB5-induced anti-apoptotic effect , we reconstituted PKM2-null MEFs with WT or mutant forms of PKM2 . As expected , WT PKM2 increased cell viability when coexpressed with IB5 ( Fig 2A ) . We then analyzed the K270M mutation , reported to be catalytically dead [52 , 83 , 84] . This mutant indeed lacked basal glycolytic activity in vitro , but the addition of high concentrations of scFv 5 stimulated its PK activity somewhat ( Fig 4B ) . PKM2 ( K270M ) failed to support the cytoprotective effect of IB5 ( Fig 3A ) . If the K270M mutation merely inactivated the catalytic site , this would suggest that PKM2’s glycolytic activity is required for the cell survival effect . However , using blue native gel electrophoresis , we found that the K270M mutation also prevented the protein from forming stable tetramers in vitro , when incubated with scFv 5 and/or FBP ( S5A Fig ) . These results have at least three possible explanations ( which are not mutually exclusive ) : 1 ) PKM2’s glycolytic activity is required for the anti-apoptotic function , 2 ) a specific tetrameric conformation of PKM2 that produces a nonglycolytic activity is required , or 3 ) IB5 binding is reduced by the K270M mutation . We note that IB5 seems to have some affinity for PKM2 ( K270M ) because higher concentrations of scFv 5 stimulated this mutant’s glycolytic activity ( Fig 4B ) . We analyzed another mutant , PKM2 ( K367M ) , which also was reported to be inactive for glycolysis [56] . We confirmed that recombinant PKM2 ( K367M ) indeed had little PK activity , even when incubated with scFv 5 or FBP ( S5B Fig ) . Furthermore , like K270M , K367M lacked anti-apoptotic activity when coexpressed with IB5 ( S5C Fig ) . However , blue native gel electrophoresis showed that recombinant K367M mutant protein did not form stable tetramers when mixed with scFv 5 ( S5A Fig ) . Furthermore , when both scFv 5 and FBP were added , this mutant primarily formed an aberrantly migrating species , possibly a malformed tetramer . ( We did also observe a minor species migrating with authentic tetramers . ) Thus , because the K270M and K367M mutations impaired both PKM2’s glycolytic activity and tetramerization in vitro , the data did not resolve whether PKM2’s glycolytic activity is required for the IB5-induced cytoprotective effect . In any case , taking together our observations that treating cells with the PKM2-activating compound TEPP-46 or replacing PKM2 with the constitutively active PKM1 failed to rescue cells from BimS-induced death , we conclude that high PK activity alone is insufficient for cell rescue by IB5 . In addition , we found that the addition of 2-deoxyglucose ( 20 mM ) to the cultures did not prevent cell rescue by IB5 ( S6 Fig ) , suggesting that glycolysis in general was unnecessary for the cytoprotective effect of IB5 . We next considered the possibility that IB5-induced PKM2 tetramerization per se is important for the cell survival activity . To test this , we reconstituted PKM2-deficient MEFs with the stably tetrameric PKM2 ( K422R ) mutant . As we confirmed in Fig 4C , this mutant is glycolytically inactive , unless an allosteric activator such as FBP is added , causing a quaternary conformational change from T-state to R-state [45] . Blue native gel electrophoresis confirmed the spontaneous formation of K422R tetramers in the absence of FBP ( Fig 4A ) . It is unclear why IB5 did not shift up the PKM2 ( K422R ) tetramer band , whereas it did shift up the WT tetramer . We speculate that the tertiary structure of this mutant tetramer is more rigid than that of WT PKM2 . Nevertheless , it does appear that IB5 interacts with PKM2 ( K422R ) , as the recombinant scFv stimulated this mutant’s glycolytic activity ( Fig 4B ) , especially upon the addition of FBP or TEPP-46 ( Fig 4C ) . It should be noted that within cells , the allosteric activator FBP is likely to be present , at concentrations dependent on the metabolic state . In MEFs reconstituted with PKM2 variants for a short time ( 3–4 passages ) , the K422R mutant slightly increased the numbers of viable cells compared with WT PKM2 , and IB5 expression further increased viability ( Fig 5 ) . At this early time , IB5 expression enhanced clonogenic survival with PKM2 ( K422R ) in a manner similar to WT PKM2 . ( For unknown reasons , survival in the absence of IB5 was somewhat reduced with this mutant , compared with WT PKM2 . ) However , in MEFs that had expressed the K422R mutant for a longer time ( passage 7 ) , clonogenic survival was increased even in the absence of IB5 and was further enhanced by IB5 expression ( Fig 5 ) . This argues that stably tetrameric PKM2 promoted a cell survival function that developed over time ( see below ) . The nuclear form of PKM2 is thought to be dimeric , whereas tetramers are restricted to the cytoplasm [54] . If so , our data imply that the anti-apoptotic effect of IB5 involves cytoplasmic PKM2 molecules and does not require the transcriptional activities ascribed to dimeric PKM2 in the nucleus . Consistent with this , we found that the K270M mutant , which formed dimers but not stable tetramers , as seen in blue native gels , even in the presence of FBP ( S5A Fig ) but is reportedly competent in nuclear transactivational activity [52] , failed to support IB5-induced cell rescue ( Fig 2A ) . Our results showed that IB5 and PKM2 inhibited apoptosis triggered by transient expression of BimS or tBid , which directly activate the “intrinsic” pathway involving Bax/Bak-dependent permeabilization of mitochondrial outer membranes . This raised the possibility that PKM2 can directly inhibit the activity or function of Bax/Bak at mitochondria . However , we observed no effect of adding recombinant scFv 5 and PKM2 to our well validated in vitro systems based on isolated mitochondria or liposomes mixed with Bax and cleaved Bid , which recapitulate the basic aspects of Bcl-2 family protein function in membranes [6 , 24 , 25 , 85] . Thus , we saw no evidence that PKM2 acts directly on the process of Bax/Bak-mediated MOMP . On the other hand , mitochondria isolated from 293T cells expressing IB5 were reproducibly more resistant than control mitochondria to MOMP induced by treatment with cleaved Bid protein ( Fig 6A ) . This suggests that PKM2 could produce mitochondrial changes that could explain the cellular rescue we observed ( Fig 2 ) . As Bcl-2 family proteins are the most prominent regulators of apoptotic death at mitochondria , we first considered whether altered levels of these proteins could be responsible for MOMP resistance . In this regard , one study reported that , under conditions of oxidative stress , PKM2 can interact with and stabilize the Bcl-2 protein [86] . However , we found that IB5 expression and TEPP-46 treatment ( alone or in combination ) failed to change the cellular levels of Bcl-2 and other major family members Bax , Bak , Bid , Bim , Puma , Bcl-xL , and Mcl-1 ( Fig 6B ) . We next used microscopy to analyze the effect of IB5 and PKM2 on mitochondrial morphology . We found that , in PKM2-null MEFs reconstituted with WT PKM2 , IB5 expression increased the average mitochondrial length ( in cells examined at passage 3–4 after transduction with IB5; Fig 7A and 7B ) . Furthermore , MEFs reconstituted with PKM2 ( K422R ) displayed a similar mitochondrial lengthening , even without IB5 expression . These results raised the possibility that PKM2-dependent mitochondrial lengthening and apoptosis resistance could involve alterations of proteins that regulate mitochondrial dynamics . In this regard , a recent study reported that PKM2 overexpression promoted mitochondrial fusion by binding to p53 and MDM2 , promoting p53 ubiquitination and degradation , and thereby inhibiting expression of Drp1 , a protein required for mitochondrial fission [87] . However , Fig 7C shows that the cytoprotective effect produced by IB5 was not accompanied by changes in the levels of Drp1 or p53 , in MEFs reconstituted with PKM2 WT or PKM2 ( K422R ) . We did find that IB5 expression substantially increased the levels of Mfn1 , a protein involved in mitochondrial fusion ( Fig 7C ) , but left Mfn2 levels unchanged . Importantly , reconstituting MEFs with PKM2 ( K422R ) alone increased Mfn1 levels , and IB5 expression elevated Mfn1 even further ( Fig 7C ) . To determine whether Mfn1 up-regulation occurred at the transcriptional level , we used qPCR to measure MFN1 mRNA ( S6B Fig ) . The results show that IB5 expression did not elevate MFN1 mRNA levels but rather tended to decrease them , and IB5 + PKM2 ( K422R ) produced the greatest decrease . We conclude that Mfn1 protein up-regulation is post-transcriptional and can even outweigh a concomitant decrease in mRNA levels . The reason for decreased MFN1 mRNA levels is unknown , but we hypothesize that it may involve either a loss of PKM2’s nuclear functions or a gain of cytoplasmic functions , as PKM2 ( K422R ) is mostly tetrameric and thus expected to be excluded from the nucleus . To determine whether Mfn1 is required for the cytoprotective effect of IB5 and PKM2 , we measured BimS-resistant clonogenic survival in WT and Mfn1- or Mfn2-deficient MEFs . The Mfn1 and Mfn2 deletions were confirmed by western blot ( S7 Fig ) . Deletion of Mfn1 or Mfn2 did not affect PKM2 expression ( S7 Fig ) . Importantly , IB5 failed to rescue Mfn1-deficient MEFs from BimS-induced clonogenic death ( Fig 7D ) . Even when IB5 was not expressed , Mfn1-null MEFs showed greater sensitivity to BimS-induced apoptosis than WT MEFs ( Fig 7D ) . In contrast , Mfn2-null MEFs responded similarly to WT . These results appear consistent with a previous report that Mfn1 directly inhibits mitochondria-mediated apoptosis at the step of Bax activation , downstream of Bax mitochondrial translocation [88] . This effect did not trivially arise from an unhealthiness of Mfn1-null MEFs , as they grew at approximately the same rate as WT cells . ( Note: we did not show a well untreated with BimS in Fig 7D because , in the absence of BimS transfection , both the WT and mutant cells overgrew the cultures , causing many of the cells to die and detach from the substrate . ) As mentioned above , we found that MEFs expressing PKM2 ( K422R ) developed a significant resistance to BimS-induced death after extended culture ( passage 7 ) , even in the absence of IB5 ( Fig 5B ) . We hypothesize that Mfn1 up-regulation gradually increases cellular resistance to apoptosis to some extent by enhancing mitochondrial fusion . Increased fusion could be expected to gradually improve the overall health of the mitochondrial network and may , for example , limit the production of reactive oxygen species . We previously observed a similarly delayed but detrimental effect on mitochondrial function in MEFs haploinsufficient for the optic atrophy 1 ( Opa1 ) protein [89] . In that case , inefficient mitochondrial fusion caused the late-passage cells to accumulate dysfunctional mitochondria that were deficient in Complex IV subunits . Even in cells expressing PKM2 ( K422R ) , which had up-regulated Mfn1 to some degree , IB5 expression further elevated Mfn1 levels and enhanced cell survival ( Fig 7C and Fig 5B ) . Thus , the extent of survival was correlated with Mfn1 levels . Moreover , our results suggest that Mfn1 promotes survival by two different mechanisms , which could be differentially engaged depending on Mfn1 levels: 1 ) a slowly developing process involving enhanced mitochondrial fusion , leading to healthier mitochondria and 2 ) an event in which Mfn1 immediately acts to inhibit Bax/Bak-mediated MOMP [88] . How IB5 cooperates with PKM2 to up-regulate Mfn1 is unknown . One possibility is that IB5 , by driving PKM2 molecules into the tetramer form , could reduce the amount of dimeric nuclear PKM2 and thereby abrogate transcriptional functions of PKM2 that could down-regulate Mfn1 . Alternatively , PKM2 tetramers could act in the cytoplasm to regulate the postsynthetic degradation of Mfn1 . Mfn1 turnover is reported to be controlled by E3 ligases such as MARCH5 , thereby regulating apoptosis [90] . However , to our knowledge , specific proteasomal degradation of Mfn1 but not Mfn2 has not been reported . In summary , our data show that high PK activity by itself was insufficient to produce an anti-apoptotic effect , as expression of the constitutively glycolytic PKM1 or treatment of WT cells with the PKM2-stimulator TEPP-46 did not rescue cells from BimS-induced death . This argues that the anti-apoptotic effect induced by IB5 involves a nonglycolytic activity of PKM2 . On the other hand , TEPP-46 significantly enhanced the cytoprotective activity of IB5 , and a stably tetrameric mutant of PKM2 , K422R , enhanced the effects of IB5 and TEPP-46 . Taken together , these results argue that IB5’s anti-apoptotic activity involves a tetrameric conformation of PKM2 . In the absence of IB5 , cells expressing PKM2 ( K422R ) for multiple passages displayed a degree of apoptosis resistance , and IB5 expression further enhanced this resistance . Such a cytoprotective effect of K422R may help explain why this mutation promoted oncogenesis in mice and occurred spontaneously in Bloom syndrome patient cells [91] . Bloom syndrome involves a mutation-prone mechanism and can therefore be considered an in vivo phenotypic selection process , in effect similar to our intrabody selection approach . Finally , the anti-apoptotic activity induced by IB5 was not accompanied by changes in the levels of major Bcl-2 family proteins . In contrast , IB5 did up-regulate Mfn1 , and apoptosis resistance was ablated by Mfn1 deletion . This is consistent with reports that Mfn1 protein can oppose Bax-dependent MOMP [88] . Our observation that mitochondria isolated from IB5-expressing cells were more resistant to Bax-mediated apoptosis ( Fig 6 ) may reflect increased levels of Mfn1 in mitochondria . PKM2-deficient cells can form tumors in mice . Often the rapidly proliferating subset of tumor cells remodel glucose utilization by expressing low PKM1 levels , whereas nonproliferating tumor cells are more likely to express higher levels of PKM1 [92] . These observations reinforce the idea that reduced PK activity , and not necessarily PKM2 expression per se , is important for rapid cell proliferation . However , they also pose a question: if PKM2 is not strictly required for tumor formation , why is PKM2 expression overwhelmingly favored in human cancers ? Although some human cancers harbor PKM2 loss-of-function mutations , these mutations are typically heterozygous . Thus , cancer cells presumably benefit from retaining at least one WT allele of the M2 isoform , which , unlike M1 , provides adaptive glycolytic regulation and nonglycolytic functions [38] . Our results suggest another potential benefit for cells expressing PKM2: resistance to apoptosis . Because PKM2 inhibits the central mechanism of apoptosis involving mitochondria , PKM2 could promote cell survival despite circumstances that would otherwise be cytotoxic . We can conjecture that particular subsets of neoplastic or preneoplastic cells could engage this mechanism to survive under adverse conditions , favoring oncogenesis . Because glycolytically active PKM2 typically corresponds with lower rates of proliferation , we could hypothesize that this cell survival function of active PKM2 tetramers might be seen primarily in slowly proliferating tumor cell subsets . IB5 most likely promotes cell survival by altering the interaction of PKM2 with one or more protein partners . It is tempting to hypothesize that IB5 mimics a natural PKM2-interacting protein . However , the identity of such a putative ligand is still unknown , as are the circumstances under which it is potentially engaged . Perhaps this cell survival function of PKM2 occurs only under specific conditions ( e . g . , allosteric activation of PKM2 combined with another regulatory event ) , which may explain why it has not been identified through conventional approaches . An anti-apoptotic function of PKM2 could be important both in cancer cells and in normal cell populations that preferentially express PKM2 , such as macrophages [41 , 93–96] and podocytes in the kidney [97 , 98] .
Primary and immortalized WT and PKM2-deficient MEFs ( PKM2Δ/Δ ) were maintained in MEMα medium supplemented with 10% FBS , penicillin and streptomycin ( Gibco-Invitrogen ) , and 0 . 1 mM of 2-mercaptoethanol [49] . 293T cells were maintained in DMEM containing 10% ( vol/vol ) FBS and antibiotics . pLHCX-Flag-mPKM2 ( Plasmid #42512 ) was obtained from Addgene . The intrabody scFv library was prepared using a naïve human combinatorial scFv phage library [76] . The scFv phagemid library was digested with SfiI , and the approximately 800-bp insert scFv-coding sequence was ligated into the SfiI-digested lentiviral vector , driven by an EF1α promoter ( without a secretion leader sequence ) , followed by a FLAG tag . Lentiviral particles were produced in 5 x 107 293T cells using pCMVD8 . 9 and pVSVg viral packaging vectors at a ratio of 1:1:1 . For the first round of selection , culture supernatants containing lentiviral particles were collected , filtered , and used for infection of 1 × 107 293T cells per 10-mm plate . For the recloning step after rounds 2 and 3 , 5 x 106 cells were used . Forty-eight h post infection , the culture medium was replaced with fresh MEMα medium supplemented with 10% FBS and penicillin/streptomycin ( Gibco-Invitrogen ) . Immunofluorescence microscopy using anti-FLAG antibody showed that over 90% of cells expressed IB5 . Human BimS cDNA was subcloned into pShooter mammalian expression vector ( pCMV/myc/cyto; Invitrogen ) to allow the expression of BimS driven by a CMV promoter . Approximately 5 x 107 293T cells were then infected with the intrabody library and then transfected with 4 μg/ml BimS plasmid using 10 μl of Lipofectamine 2000 transfection reagent ( Thermo Fisher ) . After 24 h post infection , the culture medium was replaced with fresh MEMα medium supplemented with 10% FBS and penicillin/streptomycin ( Gibco-Invitrogen ) . The integrated intrabody-coding sequences from the surviving cells were recovered after 48 h incubation and used to construct a secondary lentiviral library , as follows . Genomic DNA from the surviving 293T cells was recovered using a DNeasy Blood & Tissue kit ( Qiagen ) . A sample of 100 ng of the genomic DNA was used as a PCR template . A pair of primers matching the regions flanking the scFv fragment was used to amplify the integrated antibody fragment from the genomic DNA . The PCR product was digested with SfiI and inserted back into the lentiviral vector for a subsequent round of BimS selection , as described above . In total , over 300 clones with distinct DNA sequences were harvested and tested individually for the ability to confer BimS resistance . Sequences were analyzed with Vbase2 . scFv-coding sequences subcloned into pET28a plasmid were introduced into Rosetta ( DE3 ) pLys cells ( Novagen ) . Single colonies were picked and grown in 2 l of LB medium containing 50 μg/ml of kanamycin at 30°C for 8 h , then incubated for 12 h at 4°C with 0 . 2 mM IPTG under vigorous shaking . Cells were pelleted by centrifugation , frozen/thawed , resuspended in 50 ml of lysis buffer ( Tris 25 mM pH 8 . 0 , NaCl 300 mM ) , incubated 1 h on ice , and then lysed by sonication . The scFv was recovered from the soluble fraction by passage over a Ni++-NTA affinity column ( GE Healthcare ) . FLAG-tagged intrabody was introduced along with a tandem Strep-tag by PCR into the same lentiviral vector used for selection . 293T cells infected with the intrabody lentivirus were incubated at 30°C for 72 h , as described above . After washing with cold PBS , 5 × 108 cells were lysed for 15 minutes on ice in lysis buffer ( 50 mM Tris HCl , pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 1% Triton X-100 ) . Cell lysates were clarified by centrifugation for 15 min at 4°C at 16 , 000 x g . The total protein content of the soluble fraction was quantified using the BCA assay . For pull-down experiments , 10 mg of protein lysate was incubated with 200 μl of EZview Red anti-FLAG M2 Affinity Gel ( Sigma-Aldrich ) for 2 h at 4°C . Beads were washed three times in wash buffer ( 50 mM Tris HCl , pH 7 . 4 , 150 mM NaCl , 1 mM EDTA , 0 . 1% Triton X-100 ) . Elution was performed under native conditions by competition with 3X FLAG peptide following the manufacturer’s protocol . Eluates were used for the second-step purification using Strep-Tactin Superflow Plus ( Qiagen ) , following the manufacturer’s instructions . The final two-step purified protein was used for SDS-PAGE analysis . Bands of interest were cut out from the gel and subjected to in-gel digestion with trypsin ( PR omega , Fitchburg , WI , USA ) , followed by MALDI TOF/TOF mass spectrometry analysis ( Biomolecular and Proteomics Mass Spectrometry Facility , University of California at San Diego ) . Cell viability was measured with a Countess Automated Cell Counter ( Invitrogen ) using trypan blue . For the clonogenic survival assay , 293T or MEFs cells were seeded in 6-well plates at 1 x 105 cells/well in a 2-ml volume , transfected with BimS cDNA , and incubated for 24 h . Afterwards , the medium was replaced and the cells cultured for 3–4 d . Thereafter , the plates were rinsed with PBS and fixed and stained with a solution containing crystal violet ( 0 . 5% w/v ) and glutaraldehyde ( 6% v/v ) , as described [99] . The results were quantified using ImageJ software ( either total cell area or number of colonies , as indicated in the figure legends ) . pET28a-His-hPKM2 plasmid was obtained from Addgene . The PKM2 mutants pET28a-His-hPKM2 ( C358S ) and pET28a-His-hPKM2 ( K270M ) were generated by Quick-Change mutagenesis ( Stratagene ) . All plasmids were verified by DNA sequencing and transformed into E . coli strain BL21 ( DE3 ) . WT and mutant PKM2 proteins were overexpressed in LB medium at 30°C with 200 mM IPTG for 3 h . Cells were harvested and lysed in buffer containing 25 mM Tris ( pH 8 . 0 ) , 300mM NaCl . The supernatants were loaded on a Ni++-NTA affinity column ( GE Healthcare ) for protein purification . PK activity was measured by using Kinase-Glo Plus Luminescent Kinase Assay kit ( Promega Corporation , Madison , WI , USA ) . Purified WT or mutant PKM2 ( 50–100 nM ) was added in 100 μl assay buffer containing 50 mM Tris pH 7 . 5 , 100 mM KCl , 10 mM MgCl2 , 200 μM PEP , 200 μM ADP , and 3% DMSO . After a 15-min incubation , Kinase-Glo Plus reagent was added , according to the manufacturer’s instructions . In some cases , 0–40 μM FBP and 0–150 nM scFv 5 were added . Mitochondria were isolated from 5 x 108 cells , as described [100] . The freshly isolated mitochondria ( 100 mg protein/ml ) were then incubated with recombinant cleaved Bid protein [6] at the indicated concentrations in the presence or absence of purified scFv 5 . After incubation for 30 min at 37 °C , mitochondria were collected by centrifugation at 10 , 000 × g and analyzed by immunoblotting , as described [100] . 293T cells were washed with PBS and lysed with PBS containing 1% NP-40 . Protein concentration was determined using Pierce BCA reagent ( ThermoFisher; 23221 , 23224 ) . 35 μg protein was loaded in each lane of 12% SDS-PAGE gels . Proteins were transferred to nitrocellulose membrane and immunoblotted with the following primary antibodies: anti-Bax antibody ( Santa Cruz N20 ) , anti-Bak antibody ( Cell Signaling 3814 ) , anti-Bid antibody ( R&D Systems AF860 ) , anti-Bim antibody ( Sigma B7929 ) , anti-Puma antibody ( Cell Signaling 4976 ) , anti-Bcl-2 antibody ( Abcam 32124 ) , anti-Bcl-xL antibody ( Cell Signaling 2764 ) , and anti-Mcl-1 ( Santa Cruz S19 ) at 1:1 , 000 dilution . The secondary anti-rabbit and mouse antibodies , conjugated with HRP , were obtained from Santa Cruz and were used at 1:2 , 000 dilution . The luminescence signal was detected using ECL reagent ( ThermoFisher 32106 ) . Untreated or IB5-infected 293T cells ( 5 × 105 per well ) were seeded into 6-well plates ( Falcon ) and transfected with 30 nM siRNA . Lipofectamine 2000 ( Invitrogen , Carlsbad , CA , USA ) was used for transient transfection , according to the manufacturer’s protocol . After a 30-h incubation , fresh medium containing 30 nM siRNA and 4 μg/ml BimS expression plasmid was added . Cell viability assay was assayed after a further 36-h incubation . The PKM2-siRNA and control siRNA were purchased from Dharmacon ( SiGENOME SMART pool hPKM2 , Si156 , and ON-TARGET plus nontargeting siRNA 2 ) [101] . Isolation of total RNA was performed using the RNeasy mini-kit ( Qiagen , CA ) . After removal of contaminating DNA using DNase I ( Invitrogen , CA ) , cDNA was synthesized using SuperScript III first-strand synthesis system for qPCR ( Invitrogen ) . MFN1 mRNA levels were normalized against mouse GAPDH mRNA . | Proteins belonging to the Bcl-2 family regulate a common form of cell death known as apoptosis . Typically , these proteins function in apoptosis by controlling the formation of large pores in the mitochondrial outer membrane ( MOM ) . While many proteins that regulate apoptosis have been identified over the years , some may still be unknown . Here , we used an unbiased approach in which we first expressed in cultured tumor cells a library of intracellular single-chain antibodies termed “intrabodies . ” We then selected for intrabodies that allowed cells to evade apoptosis . We identified pyruvate kinase isoform M2 ( PKM2 ) , a major glycolytic enzyme that has been linked to cancer development , as the specific target of one such anti-apoptotic intrabody . We showed that the PKM2-specific intrabody promoted cell survival not by neutralizing its target but rather by activating an anti-apoptotic function of PKM2 . While this cell survival function of PKM2 was not related to changes in the levels of Bcl-2 family proteins or to effects on the enzymatic activity of PKM2 , we found that cell survival requires the increased expression of a MOM protein , Mitofusin-1 ( Mfn1 ) , known to regulate mitochondrial fusion . We conclude that this cell survival function of PKM2 could contribute to a role in cancer progression for this protein . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"death",
"chemical",
"compounds",
"293t",
"cells",
"gene",
"regulation",
"ketones",
"cell",
"processes",
"biological",
"cultures",
"pyruvate",
"plasmid",
"construction",
"membrane",
"proteins",
"outer",
"membrane",
"proteins",
"mitochondria",
"molecular",
"biolo... | 2019 | Phenotypic selection with an intrabody library reveals an anti-apoptotic function of PKM2 requiring Mitofusin-1 |
To eliminate Lymphatic filariasis ( LF ) as a public health problem , the World Health Organization ( WHO ) recommends that any area with infection prevalence greater than or equal to 1% ( denoted by presence of microfilaremia or antigenemia ) should receive mass drug administration ( MDA ) of antifilarial drugs for at least five consecutive rounds . Areas of low-antigen prevalence ( <1% ) are thought to pose little risk for continued transmission of LF . Five low-antigen prevalence communes in Haiti , characterized as part of a national survey , were further assessed for transmission in this study . An initial evaluation of schoolchildren was performed in each commune to identify antigen-positive children who served as index cases for subsequent community surveys conducted among households neighboring the index cases . Global positioning system ( GPS ) coordinates and immunochromatographic tests ( ICT ) for filarial antigenemia were collected on approximately 1 , 600 persons of all ages in the five communes . The relationship between antigen-positive cases in the community and distance from index cases was evaluated using multivariate regression techniques and analyses of spatial clustering . Community surveys demonstrated higher antigen prevalence in three of the five communes than was observed in the original mapping survey; autochthonous cases were found in the same three communes . Regression techniques identified a significantly increased likelihood of being antigen-positive when living within 20 meters of index cases when controlling for age , gender , and commune . Spatial clustering of antigen-positive cases was observed in some , but not all communes . Our results suggest that localized transmission was present even in low-prevalence settings and suggest that better surveillance methods may be needed to detect microfoci of LF transmission .
Lymphatic filariasis ( LF ) is one of 13 neglected tropical diseases ( NTDs ) known to chronically infect some of the worlds' poorest individuals [1] . While LF has been shown to be endemic in over 80 countries world-wide , it is one of six diseases that were deemed to be eradicable in 1993 by the International Task Force for Disease Eradication [2] . Since LF was made a priority by the World Health Organization ( WHO ) in 1997 , there has been much progress in the control and elimination of LF across the globe [3] . In 2000 , the WHO developed the Global Programme for the Elimination of Lymphatic Filariasis ( GPELF ) and established a goal to eliminate LF by 2020 . A “two-pillar” approach has been implemented for the control and elimination of LF that focuses on the interruption of transmission through Mass Drug Administration ( MDA ) and limiting the disability caused by infection through morbidity management programs . The original definition of elimination used by the WHO was based on the demonstration that the microfilaria ( Mf ) or antigen prevalence at the community level was <1% and that the cumulative incidence in children born after the start of a MDA was less than 1 per 1000 children . National mapping of lymphatic filariasis identifies areas requiring MDA by utilization of tests for microfilaremia or antigenemia . The immunochromatographic test ( ICT ) , a rapid antigen test , is considered to be one of the most practical tools for the rapid mapping of endemic areas [4] . Mapping methods are generally based on convenience sampling to identify implementation units ( administrative units , identified by the Ministries of Health , as the unit for implementing MDA ) in need of MDA [5] , [6] . These approaches to mapping facilitate programmatic decision-making , but because of the known heterogeneity of LF , microfoci of LF transmission may be missed when overall infection prevalence is low . Strategies for defining residual foci of transmission in low-prevalence settings are relevant to the global elimination effort , both from the perspective of targeting communities for MDA and for understanding surveillance requirements following cessation of MDA . Haiti is one of only four LF-endemic countries in the Americas and bears 90% of the LF disease burden in the region [7] . The LF-causing filaria Wuchereria bancrofti has been historically documented in Haiti as far back as the 1700s , primarily transmitted by the Culex quinquefasciatus species [7] . Based on nation-wide mapping carried out in 2001 , antigen prevalence ranged from zero to 45% among 6 to 11 year olds and 88% of the 120 communes that had been defined at the time had a prevalence greater than 1% [8] . In the current study , we conducted follow-up surveys to analyze potential transmission of LF in five communes that did not exceed this 1% threshold . Antigen surveys were performed in schools and selected antigen-positive children were defined as index cases for subsequent community surveys . Households near the residence of index cases were mapped and persons from a random sample of these households were tested for antigenemia using the ICT rapiddiagnostic . The analysis was designed to determine if active transmission of LF occurred in these settings and if infection prevalence exceeded the 1% threshold for MDA in some communities .
Surveys were conducted according to study protocols approved by the Centers for Disease Control and Prevention ( CDC ) and University of Notre Dame Institutional Review Boards ( IRBs ) and the Ethics Committee of Ste . Croix Hospital . During previous research in the area , researchers observed low rates of literacy in the population; therefore , all information regarding the study was translated into Haitian Creole , and verbal consent and assent were requested from parents and participants in both the school and community surveys . The consent form was read to potential study participants and/or parents . The reader of the consent form and a witness were then asked to sign the form to indicate the subject's agreement , in accordance with the IRB-approved protocol . Before surveys began in the schools , approval was obtained from the Haitian Ministry of Health and Population ( MSPP ) , the Ministry of Education , department directors , and schoolmasters . Subsequently , meetings were held with parents to provide an opportunity to explain the survey and the potential risks and benefits to their child . Prior to the community survey , community leaders were informed of the survey's procedures and provided approval for the study to commence . At the time of the survey , household members were informed of survey objectives and procedures , at which time oral consent or assent was obtained for all participants . In 2001 , national mapping for LF was performed , as previously described , using antigen testing of 100–250 schoolchildren aged 6–11 years of age per commune ( a district-sized political entity in Haiti ) across all Haitian communes [8] . Since financial resources were limited , communes of highest antigen prevalence ( ≥10% prevalence ) were prioritized for MDA . Our evaluation focused on five , predominantly rural , communes in which antigen prevalence in the survey was ≤1%: Grand Goâve ( 0 . 8% ) , Hinche ( 1 . 0% ) , Thomazeau ( 0 . 6% ) , Moron ( 0 . 8% ) , and St . Louis du Sud ( 0 . 4% ) . Within each of the communes , five to seven schools were selected for antigen testing . These public and private schools were in urban and rural areas . Following the acquisition of informed consent and assent , as detailed above , blood samples were collected from students and tested as described below . ICTs were performed on all consenting children on site and additional blood was taken for ELISA testing upon return to laboratory facilities . Questionnaires were administered to a parent or guardian of ICT-positive children . The questionnaires were designed to identify autochthonous cases—defined as those children thought to have acquired the infection in the community where testing was conducted based on responses to questions about the absence of travel and residence in the same community for the past five years . Five to eight ICT-positive children were chosen from each commune as index cases for that area . Index cases were children identified as antigen-positive by ICT in the school survey , with recoded GPS coordinates who responded to the questionnaire , and , when possible , were chosen based on those with confirmatory ELISA results . The index cases became central points in the subsequent community surveys . For the community surveys , households of index cases defined the center for each testing radius . Circles of 50–75 meters were used in more densely populated urban or peri-urban areas , and circles of 100–250 meters were used in sparsely-populated rural settings . After index cases were identified , all consenting members of index households , and a systematic random sample of the neighboring households were selected for testing by ICT . All neighboring houses within the test radius were mapped using global positioning system ( GPS ) TerraSync ( Sunnyvale , CA ) . In an effort to test 100 persons per community , approximately 20 households were chosen , based on an estimate of five persons per household ( unpublished data ) . To select these 20 houses , the total number of houses in the zone was divided by 20 to determine the sampling interval . Houses were selected from a numbered list using a randomly selected starting point and this sampling interval . The questionnaire and methods used for blood collection and testing were the same as those used for the school survey . A total of 1 , 633 persons of all ages were evaluated in the community survey . For our study , subjects were included in the analysis if they had not been previously defined as an index case in the school study , received an ICT test result , and GPS coordinates were available for their households ( n = 1290 ) . A blood specimen was collected from each person tested in the school or community survey . Filarial antigen-status was determined by ICT ( Binax , Portland , ME ) by trained laboratory personnel at the time of blood collection . Finger prick blood ( 100 µl ) was collected and results were read at 10 minutes following application to the antigen test card . Technicians were supervised , and all efforts were made to uphold the quality of the test in regards to environment and timing . An additional 200 µl of blood was collected for confirmatory antigen testing . Collected blood was stored overnight at 4°C . The tubes were centrifuged the following day and the collected sera were stored in the field at 4°C for several days until return to Hôpital Ste . Croix where they were stored at −20°C . Sera were used for subsequent serologic assays , including confirmation of antigen status for persons with positive or questionable ICT results by Og4C3 antigen enzyme-linked immunosorbent-assay ( ELISA , TropBio , Queensland , Australia ) . For the school survey and the subsequent community survey , ICT results were used as the indicator of antigen status for all persons; furthermore , autochthonous index cases were confirmed as antigen-positives by Og4C3 ELISA , the current gold standard for identifying circulating filarial antigens . Antigen-positive persons were treated with a single dose of diethylcarbamazine ( DEC , 6 mg/kg ) . Due to the low percentage of ICT positive persons in each commune and the logistic difficulty of night blood surveys , microfilaria levels were not assessed . Data were analyzed using SAS 9 . 3 ( Cary , NC , USA ) , Epi Info 6 ( CDC , Atlanta , USA ) and ArcGIS ( v . 9 . 3 . 1 , Environmental Systems Research , Inc . , Redlands , CA , USA ) . Univariate , Mantel Haenszel chi-square and logistic regression techniques were utilized . The outcome of interest for this analysis was antigen positivity as denoted by the ICT results performed in the field . Two separate case definitions were employed as indicators of possible exposure in this analysis . The definition of index case only required a positive ICT among children tested in the school survey and the availability of GPS data . Index cases would therefore serve as potential , but unconfirmed , reservoirs of infection . A second , more stringent case definition was applied for children with positive Og4C3 results and who were defined as autochthonous cases based on their answers to the survey . These individuals were referred to as autochthonous index ( AI ) cases . The exposure of interest was the distance from each person tested to the nearest index or autochthonous index case . In order to determine the ordinal categories which best represent the distance to cases , a sensitivity analysis was performed for dichotomized distances of 10 , 20 , 40 , 80 and 160 meters . Analysis of distance when using the AI case definition revealed no antigen-positive persons in the 59–99 m group , so the categories of 59–99 and 100+ meters were combined into a 60+ meter group , which was then used as the referent for the crude and multivariate regression analyses . Potential confounders and effect modifiers , including age , gender and commune were also considered based on previous literature and anticipated heterogeneity among the communes . For the purpose of modeling , age was dichotomized into <15 years and ≥15 years , with the age of 15 or younger to denote school-aged children . A spatial cluster analysis was performed on mapped households in the four communes recording antigen positivity . The analysis tested the spatial clustering of antigen-positive persons ( excluding index cases ) through the use of a Bernoulli model in SatScan ( version 9 . 1 . 1 , Boston , MA ) . A separate cluster analysis was performed for each of the four communes that included confirmed antigen-positives to better elucidate micro-clusters . Both general and isotonic simulations were performed on the commune-specific data , the latter of which accounts for the inverse relationship between risk and distance from the center of the cluster [9] . This type of simulation holds biological plausibility in representing the transmission patterns of vector-borne diseases .
Crude odds ratios were calculated to evaluate the likelihood of being antigen-positive compared across the individual covariates: distance from index case , age , gender , locale ( meaning urban or rural habitation ) and commune . A distance of less than 20 m from an index case produced a crude prevalence odds ratio of 4 . 99 [95% CI 1 . 60 , 15 . 51] when compared with distances of 100 m or more from an index case . All other individual covariates were evaluated for significance , but none besides distance of less than 20 m was statistically significant . Multivariate logistic regression techniques were applied and evaluated for collinearity , interaction and confounding , and the final model is presented in Table 3 where the exposure of interest is distance from an index case . The odds of positive antigen status among persons living within 20 meters of an index case is 5 . 41 [95% CI 1 . 64 , 17 . 83] times the odds of positive antigen status among persons living 100 meters or more from an index case , when controlling for age , gender and commune . The communes of Grand Goâve and Hinche showed significantly higher odds ratios for antigen prevalence ( 5 . 72 [95% CI 1 . 26 , 25 . 90] , and 7 . 17 [95% CI 1 . 53 , 33 . 50] respectively ) compared to Thomazeau . The parallel analysis using the AI case definition determined that there were no AI cases in the initial school survey in Moron and St . Louis du Sud , thus both were excluded from further analysis . Similar results were obtained between the index and the AI case definitions in both the crude and multivariate analyses ( Table 4 ) . A proximity of less than 20 m to an AI case had a statistically significant increased odds of being antigen positive compared to distances of 60 meters or more from the AI case ( cPOR 6 . 76 [95% CI 2 . 31 , 19 . 78] ) in the crude analysis , while no other covariates yielded statistically significant results . Multivariate logistic regression techniques identified an even larger increased odds of antigen positivity with close proximity to AI cases ( 6 . 70[95% CI 2 . 02 , 22 . 21] ) when controlling for age , gender and commune; and the communes of Grand Goâve and Hinche showed slightly higher odds of being ICT-positive when compared to Thomazeau; all of which were statistically significant . Spatial analyses were carried out in each commune , looking at the clustering of antigen-positive persons compared to the total number of persons tested . The Bernoulli model analyzed spatial clustering of cases and non-cases from a total of 319 households , each with an average of four people tested per household . Results shown in Table 5 demonstrate statistically significant clustering of cases in Hinche and Thomazeau , when evaluated at the 5% significance level in both the general and isotonic Bernoulli analyses . Examples of clustering can be seen in Figure 4 .
The original mapping for Haiti , carried out in 2001 , identified the communes of Grand Goâve , Hinche , Moron , St . Louis du Sud and Thomazeau as areas of low antigen prevalence ( ≤1% ) . As transmission of lymphatic filariasis was presumed to not be occurring , the original conclusion was that MDA was not required in these areas . A subsequent school survey was conducted to determine if there was evidence of ongoing transmission in such areas . The results from our survey showed both higher than expected ( >1% ) antigen prevalence in three of the five communes and unexpected presence of autochthonous cases ( Figure 1 ) . These observations provide evidence that transmission of LF is occurring in settings previously identified as below the prevalence threshold that would trigger a MDA . We do not know whether the results we have observed in low-prevalence areas reflect historic foci not detected through previous surveys , or recently established transmission as a consequence of population migration , or expansion of vector populations [10] . These results were shared with the MSPP and led to the decision to carry out MDA across all Haitian communes , independent of the initial mapping results . The statistically significant clustering of antigen-positive cases and increased odds of antigen positivity that were observed in the crude and multivariate analyses , as a function of distance to index and AI cases , all suggest that transmission might be occurring in microfoci , that is , among people living in very close proximity to one another , posing challenges for current mapping strategies . The model demonstrated a statistically significant increased likelihood of having a positive ICT result when residing within 20 meters of an index or AI case controlling for age , gender and commune , suggesting that antigen-positive children can serve as indicators of microfoci of transmission , and that proximity to these microfoci may be associated with the risk of acquiring LF . Risk associated with proximity to infected persons becomes of particular interest as communities see fewer and fewer instances of new infections . Different studies have reached independent conclusions regarding the possible risk . One study from Brazil found that one antigen-positive individual did not seem to pose a significant risk for transmission , as no one in the vicinity had become infected in the 10 years he lived in this non-endemic community [11] . However , this study was only observational and based on a single individual and therefore may not be generalizable to the overall population . Washington et al . addressed the probability of acquiring LF as a function of the distance from antigen-positive cases through an analysis of changes in antifilarial antibodies [12] . This study determined that for every 10 meter increase in distance from an antigen-positive case , there was a 5 . 6% decrease in IgG1 antibody levels , when controlling for age , gender and treatment status ( p = 0 . 04 ) [12] . These observations coupled with our present study indicate substantial risk with spatial proximity to an antigen-positive person in both exposure as well as acquisition of LF arguing that clustering may play a substantial role in transmission dynamics . The latest published research pertaining to the clustering and identification of “hotspots” of LF was performed by Joseph et al [13] , in which researchers examined the spatial clustering of LF in Samoa . This study looked at the clustering not only antigen positive , but also microfilaremic and antibody positive persons . Their results revealed statistically significant clustering of antigen positive individuals in three of the tested communities with radii ranging from 0 to 1160 meters . Our study complements this work by documenting the presence of clustering prior to the implementation of MDA in low prevalence settings . The final multivariate logistic regression model included commune as a significant variable , consistent with the conclusion that differences in transmission exist between communes ( Table 3 ) . Differences in the transmission of LF are likely due to different physical environments or population densities , either of vectors or humans , which may be more compatible with transmission of LF and could warrant further exploration . We did not collect data on mosquito densities for this study; however , it is likely that clustering and dispersal of infections are influenced by the behavior and flight range of the vector mosquito as well . Future studies also should examine the micro-environment surrounding the households of study participants to better address possible heterogeneity . This study is a preliminary analysis of the factors associated with antigen prevalence in five communes in Haiti in 2003 , and the findings may not be generalizable to all endemic settings . Infection risk could not be established definitively in this cross-sectional study , and we suggest a cohort study be conducted in order to confirm these results . The cross-sectional study design also does not allow for chronology to be established so there is no way to determine if cases identified in the school survey were infected before or after their ICT-positive neighbors . Thus , although we can argue that antigen-positive children were indicators of community infection , we could not determine the actual reservoir or source of this transmission . Due to the low percentage of ICT positive persons in each commune and the logistic difficulty of night blood surveys , microfilaria levels were not assessed . While generally highly specific , the ICT is not considered the most accurate test for LF infection because of problems with test interpretation in the field [14] . Testing using Og4C3 ELISA provides a quantitative measure of circulating antigens and is generally accepted as being a more sensitive test of antigenemia; however , due to financial and logistical considerations , the ICT was used for all participants in the study . Presence of microfilaria as well as biting rates from the vector would allow for the calculation of rate of transmission , and transmission risk; however , such tests were not performed for this study . We used antigen positivity as an indicator of transmission in lieu of the acquisition of such entomologic data which can be challenging and expensive to collect , especially in the context of short term surveys . Since this study was carried out in low-prevalence settings , there were few persons found to be antigen-positive . This is a challenge of sampling in a low-prevalence setting and it is accompanied by decreased statistical power . Lastly , in order to evaluate the exposure of interest , we required that GPS coordinates be available , in addition to ICT results . Since this information was not available for all study participants , our sample size was reduced . We have demonstrated that transmission , using antigen prevalence as a proxy , is occurring in areas that had previously been categorized as areas with low risk of LF transmission , suggesting that areas of low-prevalence may not be without transmission risk . The nation-wide mapping techniques in 2001 revealed a prevalence of ≤1% antigenemia for all five communes in our study , but we observed prevalence values ranging from zero to 5 . 6% in the school survey and 0 to 4 . 35% in the community survey within those same 5 communes . The identification of autochthonous index cases indicates that transmission is occurring at the level of microfoci . Since our analysis revealed that living within 20 meters of an index case significantly increased the likelihood of being antigen-positive , such microfoci may represent a particular challenge in terms of surveillance following MDA . MDA is expected to reduce infection prevalence to the point that only small isolated foci of transmission are expected to remain . The concern that these foci might represent groups of persons who are systematically noncompliant has been hypothesized on several occasions [15] , [16] , [17] . How long such microfoci can persist is unknown . In any case , the presence of such foci should be addressed and the school survey strategy we have described may represent one approach to detecting such foci through active surveillance , as proposed by Huppatz et al . [18] . Such efforts will be aided by the development of new and more sensitive diagnostic tools , based on the detection of parasite-specific antibody [18] , [19] . Ramaiah et al . reported residual microfilaria prevalence ranging from 0 . 03 to 0 . 43% in the population when tested annually over a period of 20 years post MDA [20] . It is impractical to require a surveillance period of 20 years post MDAs; however , increasing the period of surveillance from five years to ten years , prior to certification , might be necessary to ensure that transmission has indeed stopped or at least slowed to a point which cannot sustain the filarial lifecycle . Additional research is needed to address this issue . With a comprehensive program and stringent monitoring and evaluation of remaining infection we can make greater strides towards the elimination of lymphatic filariasis . | Lymphatic filariasis ( LF ) is among the leading causes of disability among tropical diseases and is caused by a mosquito-transmitted parasite but can be prevented using mass drug therapy and vector-control . In recent years , an international effort has been mounted to eliminate LF . In order to focus limited resources on areas with the highest disease burden , the World Health Organization ( WHO ) has suggested that mass drug treatment programs be focused in areas with >1% prevalence of the infection , working under the assumption that areas with <1% prevalence are equivalent to areas of limited or no transmission . We carried out an additional assessment in low-prevalence areas and observed evidence of active transmission and clustering of antigen-positive persons . Our results imply that a 1% infection threshold may not be sufficient to capture all remaining reservoirs of transmission . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"global",
"health",
"neglected",
"tropical",
"diseases",
"spatial",
"epidemiology",
"parasitic",
"diseases",
"lymp... | 2012 | Secondary Mapping of Lymphatic Filariasis in Haiti-Definition of Transmission Foci in Low-Prevalence Settings |
Progress in systems medicine brings promise to addressing patient heterogeneity and individualized therapies . Recently , genome-scale models of metabolism have been shown to provide insight into the mechanistic link between drug therapies and systems-level off-target effects while being expanded to explicitly include the three-dimensional structure of proteins . The integration of these molecular-level details , such as the physical , structural , and dynamical properties of proteins , notably expands the computational description of biochemical network-level properties and the possibility of understanding and predicting whole cell phenotypes . In this study , we present a multi-scale modeling framework that describes biological processes which range in scale from atomistic details to an entire metabolic network . Using this approach , we can understand how genetic variation , which impacts the structure and reactivity of a protein , influences both native and drug-induced metabolic states . As a proof-of-concept , we study three enzymes ( catechol-O-methyltransferase , glucose-6-phosphate dehydrogenase , and glyceraldehyde-3-phosphate dehydrogenase ) and their respective genetic variants which have clinically relevant associations . Using all-atom molecular dynamic simulations enables the sampling of long timescale conformational dynamics of the proteins ( and their mutant variants ) in complex with their respective native metabolites or drug molecules . We find that changes in a protein’s structure due to a mutation influences protein binding affinity to metabolites and/or drug molecules , and inflicts large-scale changes in metabolism .
Synergistic advances in pharmacogenomics , genome-wide association studies ( GWAS ) and next-generation sequencing bring promise to future applications of personalized medicine . Exploring the mechanistic link between human sequence variation and responses to drug therapy is likely to shed light on why certain drugs show a reduced or even harmful effect on specific individuals . For example , if an individual has a specific polymorphism or rare variant , the consequences of administering a given drug are potentially immense if a life-threatening gene-drug association has not yet been identified [1] . While numerous harmful gene-drug associations have been identified from GWAS ( and those with significant side effects now have warnings on pharmaceutical labels [2] ) , screening genome-wide associations across the broad scope of available pharmaceutical compounds is currently limited by both the cost of carrying out such studies [3] as well as a lack of statistical power due to the rarity of deleterious mutations . To address these limitations , a number of recent studies have developed mechanistic , computational analyses and the construction of omics-based workflows that identify , for example , the mode of action of common drug side effects [4] . Genome-scale modeling enables the analysis of disease-causing mutations in mechanistic detail . Genome-scale models of metabolism ( GEMs ) encompass the known interactions of diverse biological components , or the reactome of a target organism , into a unified , functional framework . This framework contains all known metabolic reactions , the genes that encode each enzyme , and all metabolites in a given organism and therefore provides a direct mapping from genes , to gene products , to the phenotypic responses of cellular activity . Mapping sequence variations in a gene to changes in the biological states of an entire metabolic network enables characterizing the effects of sequence variation in simplified cellular systems , such as the human erythrocyte [5 , 6] . Furthermore , a recently updated version of the erythrocyte metabolic model ( iAB-RBC-283 ) , based on the global reconstruction of the human metabolic network ( Recon 2 ) [7] has been used to study the response of the cell to deleterious single nucleotide polymorphisms ( SNPs ) as well as drugs with known targets [5 , 8 , 9] . Predicting the wide range of possible effects that SNPs and single nucleotide variations ( SNVs ) can have on structure-function relationships in proteins requires extending a systems-level description to include details from physics-based approaches , such as molecular dynamics simulations . To this end , three-dimensional structures of proteins provide complementary data for further elucidating changes in drug-protein interaction networks . Much attention has been placed on developing bioinformatics tools for the statistical analysis of large-scale data sets , ( which contain information on non-synonymous , exonic mutations on individual proteins ) , and generating hypotheses that explain how mutations affect stability , protein-protein interactions , ligand binding , or catalytic function [10] . Atomistic simulations have been used as a complement to experimental methods to assess changes in relative binding affinities of potential lead compounds to key enzymatic targets [11] . While these approaches are rich in molecular-level details , they are limited in their ability to address how significant the observed changes are in the context of an entire biochemical pathway or , ultimately , a whole cell . This limitation thus motivates the need to develop novel workflows that integrate systems-level and molecular-level details to characterize biological processes at graded levels of chemical detail [12–14] . The growing field of structural systems biology brings promise to the integration of systems and molecular sciences , enabling applications in personalized medicine [13 , 15–17] , drug discovery [18–20] , understanding off target binding [21–23] or mechanisms of action , [24–26] and also to enhance pharmacokinetic/pharmacodynamic models [27] . Here , we build upon previous studies which integrate protein structural information into GEMs [22 , 23 , 28] , by developing a multi-scale framework to analyze the effects of sequence variation on drug responses in human erythrocyte metabolism ( Fig 1 ) . Using genome-scale modeling approaches , we identify key proteins in erythrocyte metabolism that are perturbed in the presence of ( i ) pharmaceutical drugs and ( ii ) sequence variants . Using atomistic simulations , we characterize changes in structure and function relationships for different metabolic proteins in the form of drug or metabolite binding differences resulting from reported sequence variants . Finally , we integrate the knowledge gained from these simulations into a detailed genome-scale model of the erythrocyte , allowing for both constraint-based and kinetic methods of analysis to understand the systems-wide effect of these variants .
We were interested in quantifying the number of proteins in the human erythrocyte metabolism that ( i ) are known pharmaceutical targets and ( ii ) have been documented with both disease and non-disease causing mutations ( Fig 2 ( A ) ) . The erythrocyte presents a valuable and tractable model system for studying the effects of human genetic variation on drug metabolism . First , it is widely appreciated that the erythrocyte possesses drug metabolizing capabilities such that extracts of erythrocyte enzymes are commonly used as a general measure of enzyme activity [31 , 32] . Second , genetic changes that occur in cells other than the erythrocyte are often manifested in the erythrocyte , assuming correct isoforms and similar genetic control [33–36] . The ease of collection of human erythrocyte samples and subsequent purification of enzymes of interest motivates the study of the erythrocyte as an in silico model that can be tested against . Lastly , the erythrocyte outnumbers any other cell type in the human body ( 85% of the total cell count ) [37] . Starting from the set of metabolic genes in the genome-scale model , iAB-RBC-283 [8] , we mapped gene identifiers to cross-referenced information from dbSNP [38] , OMIM [39] , and UniProt [40] . We find that for 6800 exon coding SNPs in genes which are expressed in the erythrocyte , the majority ( >90% ) are missense SNPs as opposed to frameshift or insertion/deletion variations . These SNPs map to 247 of the 281 genes ( 88% ) in the erythrocyte model . The majority of these annotated as “disease-causing” map to enzymes within the heme biosynthesis , glycolysis , and galactose metabolism pathways , which is consistent with hemolytic dysfunction . Other non-disease causing SNPs , ( or SNPs with unknown associations ) , occur in nucleotide metabolism . Harmful mutations also tend to alter the type of amino acid much more than non-disease causing SNPs . For instance , mutations from a hydrophobic residue to another hydrophobic residue are quite common , but disease causing SNPs greatly increase this type of amino acid change to a polar , non-polar , or positive amino acid ( Fig D in S1 Text ) . Our pipeline also identifies variants that potentially influence drug-binding capabilities of respective proteins . Of the metabolic proteins in the erythrocyte , 143 are found to be potential targets for pharmaceutical action . We find 343 drugs ( approved , experimental , withdrawn drugs , or drug metabolites ) that bind to different proteins in the model [41 , 42] . In addition , mapping to the PharmGKB database , we find 274 deleterious SNP-drug associations , or documented adverse reactions ( i . e . , pharmaceutical complications ) in patients ( referred to herein as SNP-drug association ) . To summarize , our systems pharmacological database provides details on all documented missense SNPs in erythrocyte metabolism , whether they are causal for disease or cause pharmaceutical complications in a significant percentage of the human population with a sequence variation [29] . In addition , our dataset contains information on drug-binding capabilities of all proteins in the model . This combined source of information for genetic and pharmacological information within the erythrocyte allows for the selection of interesting targets to further analyze with both molecular and systems simulations . To address the structural implications of changes to sequence or drug-binding capacity , we were interested in mapping all protein-encoding genes within the metabolic network of the erythrocyte to their three-dimensional ( 3D ) macromolecular structures . Integration of protein structural data and GEMs has previously been described through the construction of GEnome-scale models of Metabolism with PROtein structures ( GEM-PRO ) . The established pipelines for constructing a GEM-PRO have been recently updated [28] . Applying this procedure for the human erythrocyte metabolic model , we start from the existing GEM , iAB-RBC-283 [8] , and the final outcome is a mapping of all protein-encoding genes to the 3D structures of their catalyzing enzymes . The selected protein structures have been quality-controlled and ranked to ensure the highest quality structures are retained . The new GEM-PRO model , iNM-RBC-283-GP , initially contained structural coverage for 181 of the 346 proteins in the metabolic network ( Fig 2 ( A ) ) , and includes a total of 1766 unique PDB entries ( the original GEM is comprised of 281 genes which encode 346 unique proteins ) . In addition , 312 homology models were obtained for proteins from existing homology model databases [43] , using the I-TASSER suite of programs [44] . Our QC/QA pipeline identifies experimental structures and homology models that can be used with high confidence in molecular modeling simulations [28] . Several quality metrics are used to rank-order structures , including: ( i ) coverage of the wild-type amino acid sequence ( with a wild-type being defined as the canonical UniProt sequence ) ; ( ii ) X-ray structure resolution; ( iii ) number of missing or unresolved parts of the structure . The final QC/QA statistics indicate that 36% of proteins in the GEM model ( 125/346 ) have high quality structural information , whereas the remaining 64% ( 221/346 proteins ) can be represented by template-based and ab initio generated homology models ( see Fig C in S1 Text for detailed statistics on subsystem coverage ) . Interestingly , when we combine the structural data and the pharmacogenomic data , we are able to assess SNP data in the context of protein structural information and derive new association . For example , we find that , on average , disease causing SNPs are 4 Å closer to annotated enzyme active sites than non-disease causing SNPs . All structural annotations , mapped database information , and quality statistics are included as a supplementary database ( S1 Database ) . One of the main advantages of assembling a structural systems pharmacological dataset for the erythrocyte is that it can be used to address questions requiring multi-scale perspectives , such as “Can mutating a single amino acid in a protein influence network-level perturbations , and , ultimately lead to disease phenotypes ? ” Considering the availability of information ( pharmacogenomic and structural ) that emerged from our mapping efforts , we were interested in focusing on several specific cases that could be studied in greater molecular detail , using a combined systems and molecular modeling approach . To this end , we assessed the available experimental , pharmacogenomic , protein structural and metabolic information available for all proteins in the erythrocyte model . Given the data collected from publically available datasets ( described above ) , we classified proteins based on: ( i ) availability of experimental protein structure , drug or metabolite binding information , ( ii ) known harmful gene-drug associations and ( iii ) if the knockout of this gene within the context of erythrocyte caused significant changes in metabolite import and export ( see Methods ) , resulting in four different classes of proteins based on these criteria ( Fig 2 ( B ) ) . This categorization mainly aids in the next steps of our contributed workflow , in studying the effects of SNVs on metabolite and drug binding using all-atom molecular simulations . As shown in Fig 2 ( B ) , Class I targets have the most information available , including 3D protein structures ( some in complex with a metabolite , drug or analogue ) , known drug-protein interactions , gene-drug associations , and clinically relevant phenotypic responses to a drug therapy . This group of proteins includes six proteins: catechol-O-methyltransferase ( COMT ) , aldehyde dehydrogenase ( ALDH3A1 ) , adenosine deaminase ( ADA ) , glucose-6-phosphate dehydrogenase ( G6PD ) , glutathione peroxidase 1 ( GPX1 ) , and uridine 5'-monophosphate synthase ( UMPS ) . Class II targets provide case-studies amenable to experimental testing SNV or drug-induced effects . Class III & IV targets are proteins found to be important in the genome-scale model , but do not have other sources ( structural or pharmacogenomic ) of information available , and therefore constitute examples of where our molecular modeling framework is useful for filling in missing information ( Table B in S1 Database ) . Here , we focus the rest of this study on three distinctive proteins in erythrocyte metabolism ( Fig 3 ) : ( i ) catechol-O-methyltransferase ( COMT ) , a class I protein ( according to our above classification scheme ) ; ( ii ) glucose-6-phosphate dehydrogenase ( G6PD ) , a class I protein; ( iii ) glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) , a class II protein . For the purpose of validation , we study the class I proteins , which have ample experimental , structural and pharmacological data associated with their roles in metabolism . To assess the predictive value of this workflow , we study the class II protein , a rare variant where population data was not available to understand the impact of documented sequence variants . Such an example serves as a demonstration for how this structural systems biology framework can be used in the absence of experimental and pharmacological data . The targets chosen for this study and their pharmacogenomic importance are outlined in Table 1 . The next stage of our proposed workflow builds on previous methods [22 , 23 , 45 , 46] and leverages systems modeling with molecular dynamics ( MD ) simulations . How SNPs/SNVs affect structure/function relationship is a question that requires analysis beyond a comparison of crystal structures . Here , we take advantage of using an ensemble of protein conformations , generated from explicit solvent MD simulations , to study the effects of clinically relevant SNVs/SNPs on drug and/or native metabolite binding ( Fig 4 ( A ) ) . While understanding protein-drug interactions provides information on how sequence variation changes protein structure and reactivity , evaluating the downstream effects of these changes requires a systems-level perspective ( Fig 5 ( A ) ) . Changes in metabolic networks can be assessed using a variety of systems methods including constraint-based and kinetic modeling techniques [5 , 77–79] . To test the susceptibility of the metabolic network of the human erythrocyte to the harmful variants detailed above , we utilized both constraint-based modeling of the iAB-RBC-283 model [8] and a recently developed in silico kinetic rate law model derived from the Mass Action Stoichiometric Simulation ( MASS ) approach [80 , 81] . For a number of proteins , disease causing mutations can cause systemic changes within the metabolic network or in the transport of certain metabolites [8 , 82] . With regards to the erythrocyte , understanding these differences in metabolite transport can be correlated with changes in metabolite concentrations within biofluids , which potentially expands the use of this model as a diagnostic tool for human disease . Similar perturbations can also be linked to the specific phenotypic responses of the erythrocyte , such as to drug treatments , or the ability to respond to changes in oxidative ( rate of NADPH use in order to combat oxidants ) or energy ( rate of ATP use ) load [5] . Here , we propose a framework for mapping protein structural information to genome-scale models of human erythrocyte metabolism for the characterization SNP-drug associations . Three case studies presented in this contribution point to the complexity of pharmacogenomic associations and being able to conduct integrated in silico simulations that extend from the molecular scale to the systems level . Using parameters from molecular simulations to guide genome-scale modeling , we are able to study how changes in protein structure and binding affinity influence the phenotypic states of an entire metabolic network . We find that the union of genome-scale modeling and molecular , physics-based methods , presents , to the best of our knowledge , the first workflow capable of systematically integrating data from pharmacogenomics research , in conjunction with 3D high resolution protein structural information , to model changes on both the pathway ( i . e . metabolic network ) and molecular ( i . e . protein ) scales . The information gained through molecular modeling simulations can be utilized to supply parameters to both kinetic models and constraint-based modeling approaches and has been found to be amenable to the study of other enzymopathies [5 , 91] . Our findings indicate that there is consistency between experimental and computational trends in substrate and drug compound binding in wild-type versus mutant proteins . Currently , most systems biology approaches lack the ability to utilize insights from structure-based analyses related to metabolite and/or drug binding . Fortunately , atomistic molecular simulations have evolved to become powerful tools for the characterization of binding mechanisms and as such constitute valuable assets for systems modeling . Extending analysis beyond crystallographic structures through the use of ensemble confirmations substantially enhances the predictive scope of docking methods by identifying alternative binding modes for a drug molecule [56–60] . Ensembles of the thermodynamically accessible states of a protein , generated from molecular dynamics , allows for the mechanistic characterization of how sequence and structural variation may influence metabolite or drug binding [92] . The scalability of this workflow is mainly limited ( i ) to the documentation and experimental analysis of exonic SNVs/SNPs , and ( ii ) by the execution of molecular dynamics simulations , which takes a significant manual effort and requires high performance computing resources . For the second point , certain efforts have already shown that high-throughput simulations using classical MD can be performed on large numbers of proteins [93 , 94] . However , performing high accuracy computations on a systems scale is currently intractable , due to the intense computational and time requirements of quantum-based simulations or free energy calculations . Therefore , a trade-off between accuracy and cost must be considered ( see Fig B in S1 Text and recent reviews on the subject [95–97] ) . In light of these limitations , we find that the additional information gained from protein structure greatly contribute to our understanding of causal mutations and can assist in selecting protein targets for more detailed molecular studies . Thus , when combined with other developing frameworks [4] and experiments [98] , the contributed workflow provides a first step in the translation of Big Data in the pharmaceutical industry to practical therapeutic applications and is expected to have a positive transformative impact on the fields of systems medicine , population studies and drug discovery efforts .
The techniques used here are a consolidation of 4 previous methods to add protein structural information to genome-scale models [22 , 23 , 99 , 100] , and described in detail in [28] . To do so , the SBML model of the erythrocyte genome-scale model was first obtained from the BiGG Models website ( http://bigg . ucsd . edu/models/iAB_RBC_283 ) [101] , and all gene IDs were mapped to their corresponding amino acid sequences ( UniProt and RefSeq entries ) . This model differed from the construction of previous GEM-PROs due to the appearance of protein isoforms , and required additional manual mapping to ensure correctness . Gene isoforms led to inconsistencies between database entries and additional difficulty linking to available homology models ( discussed in the section “Homology Modeling” ) . Additional QC/QA steps were taken in order to ensure the correct sequence was being retrieved , as described below . Previous work was done to map data from the Online Mendelian Inheritance in Man ( OMIM ) database in order to find disease causing mutations that could map to erythrocyte proteins [8] . We also collected all known SNPs from dbSNP , and filtered them down to variations in exons that could be studied utilizing protein structure information . Information was additionally cross-referenced with UniProt variant annotations [109] . There are a number of drug target databases that were queried for this study . DrugBank was used in a previous study to gather drug targets based on sequence [8] . In order to be as comprehensive as possible , we also obtained data from ChEMBL [110] and MATADOR [42] , with MATADOR providing annotations for indirect interactions . With this , we were able to verify targets that appeared in all 3 databases . Drug adverse effects due to variation were mainly gathered from the PharmGKB , a pharmacogenomics database with information from clinical studies , research articles , and individual cases [111] . The PharmGKB further annotates for the significance of an association , as well as details of the clinical trial or GWAS study carried out . Finally , the DrugBank contains a simple list of SNP-drug associations in their SNP-ADR and SNP-FX sub-databases [41] , which was cross-referenced with all information found in the PharmGKB . As a final source of parameters for validation of our model , experimentally determined kinetic values for binding of a drug or inhibitor to a target ( wild-type as well as mutant ) were obtained from BRENDA and the BindingDB [112 , 113] . As expected , information for this step was much sparser than the previous information , which indicates the need for experimental assays if we are to validate the predictions made from this model . For the targets in this study , we also manually searched for additional information from published biochemical studies . Finally , for the selection of interesting targets to study with molecular and systems modeling techniques , we also wanted to understand the essentiality of each gene within the erythrocyte model . Gene knockouts were performed for each gene contained within iAB-RBC-283 , as per [8] . A gene was marked as interesting to study within the context of the erythrocyte if there were significant changes within the reaction fluxes of metabolite import and export through the membrane using flux variability analysis ( FVA ) simulations [114] . In order to detect these significant differences , all reaction fluxes were compared to the normal “wild-type” state of the cell . Specifically , similar procedures to Shlomi et al . and Bordbar et al . were followed [8 , 82] . Changes in exchange fluxes were categorized into i ) activation/inactivation , ii ) shift to a fixed direction , iii ) a change in magnitude of flux , or iv ) no change ( refer to [8] , Fig 5 ) . For changes in magnitude of flux , if the new flux span ( defined as maximum flux—minimum flux ) was less than 40% of the original flux span , it was considered to be a significant change . Experimental PDB structures or homology models representing the genes of interest in this study were taken from the GEM-PRO data frame following ranking and QC/QA . Mutant forms of the enzymes were either taken directly from the PDB , if available , or modeled by point mutations of the structure . Next , the general approach for each target was to first understand the binding position and energetics of either the native metabolite or a drug of interest to a wild-type protein structure and its corresponding mutant . Flexible docking simulations using DOCK6 were carried out with default parameters and binding sites defined when known [115] . Furthermore , simulations were conducted with and without cofactors , to account for competitive binding drugs or cases where the order of substrate binding was not known . To compare flexible docking results to ensemble docking , simulations were repeated under different random seeds for a total of 500 docking runs . Molecular dynamics simulations were run utilizing the PMEMD module of the AMBER14 toolkit [116] . Initial parameterization of ligands and cofactors were carried out utilizing the Gaussian 09 software [117] or obtained from previously published data sets ( see S1 Text for protein-specific methods and S2 Database for parameter sets ) . For generating topologies as input to AMBER , 99SB force field charges and atom types were then used and then solvated in a periodically repeated TIP3P 12 Å water box with counterions being added as needed ( Na+ or Cl- ) . Minimization was carried out under constant volume conditions at while being heated to 300 K . Structures were then equilibrated under constant temperature and pressure conditions with restraints being released . Finally , the structures were run in production phase of 75 ns or more under a Langevin thermostat and Particle Mesh Ewald ( PME ) cutoff of 12 Å . At least 4 separate MD simulations ( representing WT and SNP structures in cofactor unbound and bound states , more for additional cofactor bound states ) were carried out on each enzyme ( see Tables D-F in S1 Text for all simulation information ) . Every 100 frames from these trajectories were utilized as input for ensemble docking of the substrate of interest . All docked positions were clustered into 5 representative poses based on the distances from known binding residues . Specifically , distances from 3 known binding or interacting residues to the atoms of the drug or metabolite were calculated for each extracted frame , and k-means clustering of the Euclidean distance separated these frames into 5 distinct binding modes for use in further simulation . These docked positions were subject to additional MD production runs of 10 ns each , in order to examine the stability of the bound position and if they would converge into one distinct pose . We conducted free energy calculations for each of the ligands in the cofactor bound state of the WT and SNP enzymes . MM-PBSA calculations were carried out to predict the difference in free energies of binding ( ΔΔG ) . The binding energies of all 5 representative conformations were averaged per ligand , and the resulting value indicates if the ligand is more favorable to bind to WT ( negative ΔΔG ) or SNP ( positive ΔΔG ) structures . MM-GBSA/MM-PBSA calculations utilizing the MMPBSA . py script available in the AMBER14 toolkit were carried out on the 10 ns simulated receptor-ligand complexes [61] . The first nanosecond of simulations was discarded before running calculations to account for initial stabilization of the docked ligand . Thermodynamic integration ( TI ) calculations were calculated utilizing the Simulated Annealing with NMR-derived Energy Restraints ( SANDER ) module within AMBER14 [118] . The dual topology paradigm was utilized with a three step alchemical transformation , with state 0 representing a wild-type enzyme and state 1 the mutant form . Step 1 carried out the decharging of the WT utilizing 10 λ points and simulations of 1 ns each . Step 2 transformed the residue atoms of the WT to the SNP again utilizing 10 λ points and simulations of 1 ns each . Step 3 carried out the recharging of the mutant residue atoms with the same number of λ points and simulation time . This was run for both ligand bound and unbound states . Finally , the change in potential energy of the system with ligand bound was calculated by integration over the λ points and subtracted from the ligand unbound state . For full information on docking , MD , MM-PBSA , and TI parameters , please refer to the section entitled “Molecular modeling simulations” in S1 Text . The constraint-based modeling approach was carried out for all enzymes in this study by simulating a normal ( wild-type ) and perturbed ( mutant ) erythrocyte condition utilizing FVA followed by a Markov chain Monte Carlo ( MCMC ) based sampling approach [83 , 91 , 119] . Previous simulations for identifying biomarkers have simulated perturbed states by setting the upper and lower bounds of flux through affected enzymes of the cell to 0 , effectively mirroring a full gene inhibition , and then analyzing the exchange conditions [8 , 82] . For the purposes of this study , we are now able to understand the relative differences in native metabolite catalysis utilizing the ratio of differences in the binding affinity between wild-type and mutant forms of the enzymes . This ratio was then converted into a ratio of flux in wild-type to mutant enzymes , assuming equal concentration of substrate and enzyme ( see Equation S3 ) . From this , the determined normal wild-type minimum and maximum fluxes through the corresponding reaction were adjusted to a perturbed mutant state , and both FVA and MCMC simulations were then run with the goal of analyzing 1 ) the flux differences through the exchange reactions ( import/export of metabolites ) of the erythrocyte ( as described above in the section “Genetic variation , drug-target interactions , and essential genes” ) and 2 ) significant flux shifts within the internal network . In this way , hypotheses for the altered phenotypic state of the erythrocyte and its impact on the body could be deduced based on the differences of uptake or secretion of metabolites or large-scale internal network changes . For MCMC simulations , significant shifts in the distribution of fluxes were considered ( p-value < 0 . 05 ) . Additional information on MCMC sampling is included in the section entitled “Systems modeling” in S1 Text . With the kinetic rate law model , we are able to directly integrate the predicted Km and experimental Kcat values as well as simulate the cell under oxidative or energy load conditions . This detailed model was utilized for the simulations of normal and perturbed G6PD and GAPDH enzymes . Simulation of COMT within the kinetic model was not available due to the current model being limited to core metabolic enzymes . We utilize the model to also understand the erythrocyte’s capability to withstand oxidative stress or increased energy needs and compare wild-type to mutant states . Oxidative stress is simulated as an increase in the rate of NADPH usage , to mirror the fact that a cell under stress requires NADPH to neutralize reactive oxygen species . Energy load is simulated as an increase in the rate of ATP usage . The normal , wild-type cell was first simulated and the maximum oxidative and energy loads were determined for comparison to the mutant state . Integration of the predicted Km without any change in Kcat was then simulated for the mutant state , to understand if only changes in binding affinity led to a change in maximum tolerable oxidative or energetic load . Finally , changes from predicted Km , experimental Km , and experimental Kcat were fully integrated to investigate the model’s accuracy to the known phenotype . | Structural systems pharmacology is an emerging field of computational biology research that aims to merge network and molecular views of biology . Genome-scale models are in silico , network models of metabolism , and by integrating the detailed knowledge we can gain from molecular simulations with these models , we can begin to understand whole cell phenotypes at a more complete scale . In this study , we use and integrate a variety of simulation tools at both the network and molecular levels to allow us to understand how a mutation can change an enzyme’s ability to bind to drugs or metabolites . We look at three different enzymes within red blood cell metabolism , and find that these computational tools reflect what we know about them relatively well , and also potentially serve as a workflow for understanding other traits in the overall theme of personalized medicine . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"protein",
"metabolism",
"enzymology",
"red",
"blood",
"cells",
"enzyme",
"metabolism",
"protein",
"structure",
"protein",
"structure",
"databases",
"pharmacology",
"drug",
"metabolism",
"enzyme",
"chemistry",
... | 2016 | A Multi-scale Computational Platform to Mechanistically Assess the Effect of Genetic Variation on Drug Responses in Human Erythrocyte Metabolism |
The characterization of mutational spectra is usually carried out in one of three ways–by direct observation through mutation accumulation ( MA ) experiments , through parent-offspring sequencing , or by indirect inference from sequence data . Direct observations of spontaneous mutations with MA experiments are limited , given ( i ) the rarity of spontaneous mutations , ( ii ) applicability only to laboratory model species with short generation times , and ( iii ) the possibility that mutational spectra under lab conditions might be different from those observed in nature . Trio sequencing is an elegant solution , but it is not applicable in all organisms . Indirect inference , usually from divergence data , faces no such technical limitations , but rely upon critical assumptions regarding the strength of natural selection that are likely to be violated . Ideally , new mutational events would be directly observed before the biased filter of selection , and without the technical limitations common to lab experiments . One approach is to identify very young mutations from population sequencing data . Here we do so by leveraging two characteristics common to all new mutations—new mutations are necessarily rare in the population , and absent in the genomes of immediate relatives . From 132 clinical yeast strains , we were able to identify 1 , 425 putatively new mutations and show that they exhibit extremely low signatures of selection , as well as display a mutational spectrum that is similar to that identified by a large scale MA experiment . We verify that population sequencing data are a potential wealth of information for inferring mutational spectra , and should be considered for analysis where MA experiments are infeasible or especially tedious .
Knowledge of the mutational spectrum is central to the study of molecular evolution . However , mutational spectra are difficult to characterize because spontaneous mutations are scarce and thus rarely observed in large enough numbers for precise measurements . In addition , mutational spectra vary across species , between individuals , and across genomic segments , placing a demand for methods that can identify a large set of mutational events genome-wide , while remaining applicable to a wide range of species . One direct approach to the study of spontaneous mutations on a genome-wide scale is through mutation accumulation ( MA ) experiments . MA experiments allow the accumulation of mutations under minimal selection conditions in a controlled lab environment , usually over many generations [1–4] . If following individual clonal lineages is not feasible , minimal selection conditions are usually achieved in unicellular cultures through repeated extreme bottlenecks , sometimes down to a single individual , such as in Saccharomyces cerevisiae [5–13] , Dictyostelium discoideum [14] , Arabidopsis thaliana [1] , and Chlamydomonas reinhardtii [15 , 16] . It can also be achieved through generations of inbreeding in species such as Drosophila melanogaster [17–19] , or rhabditid nematods [20] . The final progeny are then sequenced and compared to the starting ancestor to identify de novo mutations that occurred within the span of the experiment . The throughput of this process has been greatly aided by recent advances in next generation sequencing , and MA experiments have thus provided significant insights into overall mutation rates , relative frequencies of mutation classes , mutational biases , and repair pathways . While powerful , MA experiments face certain limitations that cannot be easily rectified . One limitation is technical . Many species cannot be considered for lab studies due to space , life span , ecological , or ethical limitations , if they can be maintained under lab conditions at all . The other limitation is theoretical . Genome stability can be dependent upon environmental factors and life cycle stages [21–23] . For many organisms , including the majority of microbes , such parameters are difficult to characterize . The complex habitats of ‘wild’ populations are thus important but unknown , and therefore cannot be replicated in the lab . In addition , a complex network of genes and pathways regulate DNA repair . Differences in genes involved in DNA fidelity-associated pathways may result in the mutation spectrum varying across sub-populations or even individual strains . As MA experiments usually involve less than a handful of genomic backgrounds that are extremely well adapted to a lab environment , it is possible that they are not representative of the mutational patterns in the species as a whole . In addition , most MA experiments utilize a relatively small number of lines that are allowed to accumulate relatively large number of mutations for a fairly long period of time . While it is possible to shorten MA experiments , this is often accomplished through the use of mismatch-repair ( MMR ) impaired strains that accumulate mutations at an artificially fast rate . Such experiments are used to survey large numbers of mutations in a short period of time in a fashion that is specific to the MMR pathway affected . For example , recent work on conditional or complete MMR defect [10 , 24–26] , nucleotide pool imbalance [27] , and replicative polymerase variants [9 , 13] has made use of such systems . These experiments are powerful but extremely specific means of probing the DNA replication and repair system , and all mentions of MA experiments in the rest of this paper do not specifically refer to MMR based studies . In regular MA experiments , where the aim is to study ‘natural’ mutations spectrum , only ‘wild-type’ strains are used . For such studies , the MA approach is certainly economical , in that the sequence of a single genome can reveal the presence of a large number of mutations . But the savings come with the cost of two possible sources of bias . First , the MA lines lose fitness as they accumulate mutations and less fit lines might have a very different mutational bias compared to the more fit , naturally occurring lines [28 , 29] . Second , some MA lines might go extinct–indeed , in most MA experiments they invariably do [7] . The extinct lines are likely to contain some of the most deleterious mutations that will be missed in the final sample of mutations; thus the sequencing of the surviving lines necessarily does not provide a fully unbiased sample of mutations . An alternative approach to MA experiments relies on the identification of mutations from sequencing of genomes of natural strains . Unlike controlled laboratory experiments , such sequencing can be carried out with most species . Sampling from natural populations further removes many potential biases introduced by lab conditions and experimental set up . Methods that infer mutational spectra from sequence data usually rely upon the assumption that mutations at certain genomic locations are strictly neutral , such as pseudogenes or dead transposable elements [30] that are presumably under no selection pressure , or mutations that lead to a synonymous change in a protein-coding sequence . If this assumption holds , it can be shown that the rate of substitution between species at these sites would directly reflect variation in mutation rates [31–33] . However , it is increasingly apparent that almost no mutations are truly neutral , and even very mild selection or selection-like forces such as biased gene conversion can significantly influence patterns of substitution [34–38] . The overwhelming majority of substitutions observed from sequence data would therefore be survivors of selection and selection like forces , albeit to varying degrees . While extremely informative in their own right , these are necessarily highly biased subsets of the true spectrum of spontaneous mutations . While divergence data are almost certainly biased by selection , existing polymorphisms within a population need not all be . Segregating alleles can be effectively neutral if they are observed while still under the selection-drift barrier . Because spontaneous mutations necessarily enter the population at a frequency of 1/N , where N is the number of the chromosomes in the population , identifying a cohort of extremely rare polymorphisms will enrich for very young mutations [39] . Mutational spectra from rare variants through deep population sequencing has already been employed in viral systems such as HIV [40] , where the main challenge lies in accurately calling extremely rare variants from a heterogeneous viral population [41–43] . Rare variants have also been applied to characterizing context dependent mutational patterns in 202 human genes [44] , although in species where single individual sequencing is accessible and populations are not homogeneous , population structure must be accounted for [45] . One elegant solution would be limiting analysis to de novo variants in parent offspring genome comparisons , such as the comparison of family trios in drosophila , butterfly , and humans [46–49] . In many other species , it is not always possible to identify relatedness between individuals ahead of time and selectively sequence parent-offspring genomes . In such instances the relatedness of sampled genomes or genomic regions must be estimated post hoc . For a hypothetical organism that reproduces asexually and does not undergo recombination , relatedness between individuals simply involves genomic sequence identity . If two genomes are nearly identical , any variant between them is likely a relatively young mutation that occurred after their last common ancestor . In actual datasets , recombination and/or sexual reproduction result in genomes with mosaic evolutionary history across genomic segments . To obtain recent mutations from such sequences , regions of identity by descent ( IBD ) would be more appropriate . However , proper IBD analysis requires haplotype information , which may not always be available , or might be difficult to impute in species such as yeast where ploidy can vary between 1n and 4n in natural isolates [50] . In the absence of IBD information , on the basis that rare polymorphisms are younger on average , the density of unique SNPs serves as a proxy for IBD information . Genomes with close relatives in the dataset share most of their polymorphisms with at least one other strain and carry few unique mutations , most of which will be young , while genomes with no close relatives share fewer polymorphisms and appear to carry an excessively large number of unique mutations ( singletons ) , most of which will be old . The density of singletons in a genome or genomic region [51] , as defined by all polymorphisms present in a sampled population , can serve as a measure of the age of rare variants on that genome . To test the practicality and accuracy of this technique , we sequenced 141 individual strains of Saccharomyces cerevisiae to high genomic coverage and analyzed the mutational spectrum that could be obtained from identified young mutations . By comparing how closely our results matched both theoretical expectations and the mutational spectrum derived from a large-scale MA experiment in yeast , we determined that we could recapitulate the mutation spectrum of a species through broad population sequencing , that is , the sequencing of a large number of individuals .
To sample a set of non-experimental individuals from a relatively diverse population , we sequenced 141 S . cerevisiae strains in their natural ploidy states [52] . The majority of these strains were clinical isolates , with around a dozen well-studied commercial and lab strains . Because yeasts are known opportunistic pathogens , this set of strains likely represents the diversity in human-associated yeast populations . SNPs were only called in comparison to the reference sequence of S288C in non-repeat regions after meeting filter requirements ( S1 Fig ) . Excluding one strain where sequencing failed due to contamination , a final set of 423 , 387 SNPs passed these quality filters ( Methods ) . The site frequency spectrum of the observed population of polymorphisms shows the expected gamma shape of population sequencing datasets , with a small bump around freq = 1 ( S2 Fig ) . New spontaneous mutations , as a group , should show none of the classical signatures of selection . Three criteria were employed as indicators of our ability to identify very young SNPs: 1 ) the percentage of nonsynonymous polymorphism ( %Pn ) , 2 ) the transition transversion ( Ts/Tv ) ratio , and 3 ) the GC equilibrium percentage ( GCeqm ) . In divergence data , the ratio of nonsynonymous changes tends to be much lower than the ratio of 0 . 75 expected in the absence of selection , Ts/Tv values are usually > 2 . 5 , and the GCeqm ( roughly ) matches the genomic GC content ( which is 38% in yeast ) . The mutations from a previous large-scale genome-wide MA experiment in yeast yield a %Pn value close to the neutral expectation of 0 . 75 , a Ts/Tv value of 1 , and a GCeqm of 32% [12] . We therefore explored our ability to obtain similar values from our polymorphism data . We first segregated SNPs by their frequencies in the population and summarized all three values for each frequency class . We expected that with decreasing frequency of polymorphisms , the proportion of young SNPs should increase , and the three values should approach those observed in MA experiment ( Fig 1 green dotted lines ) . While the %Pn and Ts/Tv ratios did shift towards MA values , especially in the lowest SNP frequencies , the changes did not reach expected MA values . However , a similar trend was not seen for the value of GCeqm ( Fig 1 ) . Indeed , even at the frequency of 1/141 , none came close to matching MA values . Because there is substantial population structure in the sampled strains [52] we tested whether controlling for relatedness between strains could further refine our analysis , this time focusing on just the singletons . We used the density of singletons/kb as a measure of singleton age . For example , if a chromosome carried n singletons , each of the n singletons is given the ‘age’ of n/length of the chromosome in kb , approximating the time unit it takes for a mutation to occur once per 1 kb since its last common ancestor with the closest sampled relative . Often , chromosomes will carry multiple singletons , and though the singleton mutations must have occurred at different times , it was impossible to accurately identify the order in which these mutations happened . We chose to be conservative in our age categorization and assign the same age to all singleton mutations on a given chromosome . We binned SNPs by age into groups of roughly the same sample size , with higher resolution at the youngest ages , ranging from 0 . 001/kb through 2 . 25/kb . We then tested whether patterns derived from the younger age groups came closer to the MA experimental values . Plots of the %Pn , Ts/Tv , and GC equilibrium values for each age group showed a clear trend in which the 5 youngest categories ( ages <0 . 005/kb ) matched MA values for both Pn/Ps and Ts/Tv ratios ( Fig 2 ) . Surprisingly , for GCeqm , the youngest singleton classes suggested an average value of around 25% , below the 32% derived from the MA experiments ( Fig 2 ) . While mutations are indeed AT biased , this value is more extreme than previously reported . To ensure that the youngest singletons as a group were not dominated by low quality SNPs , we noted that coverage depth , genotype qualities , and mapping qualities were not significantly different between young singletons with density <0 . 005/kb as compared to older singletons , and SNP quality was capped at a minimum of 20 ( S3 Fig ) . There were 829 singletons of ages <0 . 005/kb that matched the Pn/Ps and Ts/Tv values from the MA experiment . Coincidentally , this sample size is similar to the 864 SNPs from the MA experiment . Because MA results were based on a single homozygous diploid strain that was exposed to a constant , stable environment , the mutation spectra of a population that is far less homogenous may be different . To determine how the mutation spectra presented by the young singletons differ from old singletons , or from MA data , we calculated the relative mutation rates for all six possible nucleotide changes ( Fig 3 ) . Young singleton rates for each nucleotide change were compared to corresponding old singletons and MA rates ( Z-test , Bonferroni corrected ) . There were significant differences in rates between young singletons and old singletons , but also between young singletons and the MA mutations . We further pursued the context dependent difference in mutation rates previously found in MA data , and divided singletons into groups based on their neighboring bases . A previous MA experiment showed a potentially elevated mutation rate at the middle nucleotide C in CCG and TCG environments , suggestive of low but detectable levels of methylation [12] . However , this particular bias was not clearly observed in the young singletons . Indeed , the highest rate was observed at ACG sites in the young singletons . Intriguingly , all four *CG sites had higher mutation rates in the old singletons ( Fig 4 ) . The biological significance of these results remains to be determined as there is more recent evidence that there is in fact no methylation in S . cerevisiae [53] . Our results do suggest , however , that there might be subtle differences between MA estimates and mutational biases in nature . Additional data should be able to resolve this question . We tested if this classification system could potentially be employed for another mutation class–indels . Indels have been more difficult to study and analyze than SNPs due to their exceedingly rare nature ( observed at least an order of magnitude less often than SNPs ) and their strong fitness effects ( that do not usually allow them to persist in natural populations ) . In most MA experiments , indels are observed in very low numbers in unique sequences , particularly in coding regions . Broad population sequencing allows larger numbers of such events to be observed , but mapping errors can increase false discovery rate ( FDR ) around repetitive regions . We filtered and aged indels following the same protocol as SNPs , and utilized the percentage of indels seen within coding regions ( which span ~70% of the analyzed portion of the yeast genome ) as the main signature for the action of selection . We confirmed that GC content of genomic sequences ±10bp of 3 , 389 high quality singleton indels were not significantly biased , but the incidence of simple tandem repeats ( STRs ) were more common than expected by chance ±10bp of indels , particularly for A/T monomers ( Fig 5 ) . This is in spite of the prior masking of 600Mb of known repetitive sequences . The indel singletons also did not occur randomly within the genome , with only 20% found in coding sequences , although this may be partly due to context dependent variation in error rates [13 , 54–55] . However , the youngest indels of age <0 . 002/kb were clearly less constrained by selection than older indels ( Fig 6 ) .
In yeast , we used singleton density per Kb of genomic sequence as the arbitrary genomic unit for singleton counts . Any genomic unit or segment can conceivably be used , as long as they are long enough such that mutations within that region are rare , but not so long that the sample does not contain individuals closely related enough as to be nearly identical across all of it . It is also important to note that yeast has an atypical life cycle that is neither obligate asexual or sexual . It is thought to reproduce predominantly through clonal means with occasional sexual reproduction ( reviewed in [56] ) . Yeast also has a marked tolerance for large-scale copy number changes ( e . g . [57] ) . It may even be highly tolerant of hybridization with closely related species ( e . g . [58–60] ) , and is known to carry introgression from sister species ( e . g . [57] ) as well as more distant relatives ( e . g . [61] ) . The impact of such irregular life cycles ( found in many fungi/moss species ) on segregating sites within a population sample is not clear . A similar study in more species may help to resolve this question . In obligate sexual reproducers , there may still be large variations in mutation rate , recombination rate , or population diversity that can make sampling closely related genomes difficult without prior knowledge . For such species , more care must be taken during sample collection . Another point to keep in mind is that this method can only identify what selection doesn’t immediately remove . There is a practical limit to how closely related individuals from a random population sampling can be , unless the population is extremely inbred , or there is genealogy information . The youngest mutation we can identify is consequently lower bound by how recently the two closest related individuals diverged . If selection is so strong that many mutations have already been removed within that short divergence time , we would be limited to only describing the trends that follow . An extreme example of such scenarios would be lethal mutations , although this was also seen to a certain degree in indels , which are removed by selection at a rate that is 10 times as fast as nonsynonymous mutations [19] , and unsurprisingly never reached neutral expectations in our analysis . The power of this method lies in numbers . Sequencing of just 141 strains was able to give us 829 putative young SNPs , a number nearly matching that from a large diploid yeast MA experiment involving nearly ~311 , 000 generations under controlled lab conditions [7] . A subset of these young mutations may have accumulated during lab propagation for DNA extraction , but the large numbers suggest that the majority were in fact ‘natural’ mutations . In addition , we identified 168 singleton indels with an age of <0 . 002/Kb , a class of mutation only very rarely seen in experimental settings . We were able to show that even here , where selection acts strongly and quickly to change the overall signature , some trends can still be observed with a decrease in indel age . There are multiple benefits unique to this analysis . First , instead of correcting for population structure , an issue common to most population samples , it takes advantage of the varying degrees of relatedness in a sample set to classify singletons into age groups . These continuous age groups , in addition to investigating whether young mutations match neutral expectations , also allow observation of trends across time . Second , unlike methods dealing with divergence data , there are no phasing or haplotype issues . Young singletons are necessarily the derived allele , and they are so rare the effect of linkage is negligible . The yeast strains used were in fact in various states of natural ploidy [52] , as can be expected of a natural population sample . However , note that <1% of sites were suspected of carrying more than 2 alleles , and were not considered in this analysis . One major limitation of this method is that it doesn’t provide the ability to accurately estimate mutation rate , which is something that naturally follows from an MA experiment . The means of accurately estimating generation time separating such closely related individuals is beyond the scope of this manuscript . A second issue is that the number and identity of singletons will heavily depend upon which and how many strains are sequenced . As more strains are sequenced , some singletons will be lost , while others will be identified . A logical further extension of this approach would be to try to age not just singletons , but doubletons , tripletons etc . , based on population frequency and shared haplotype lengths , though it is unclear how much this would modify the overall conclusions . As broad population sequencing becomes increasingly accessible , the amount of information we can extract from resulting sequence data becomes the limiting factor to their scientific value . Well-described mutational spectra form one area of molecular evolution for which extensive work has been difficult to amass , and which can benefit from this new application .
DNA library construction , read mapping , and variant calling protocol was detailed in earlier publication [52] . Briefly , DNA was extracted from liquid cultures using a modified glass bead lysis protocol . 500bp paired-end Illumina sequencing libraries were prepared at The Genome Institute , Washington University School of Medicine , and run on an Illumina HiSeq to an average of 100-fold coverage . Resulting fastq files were mapped to the reference genome with bwa v0 . 5 . 9 [62] , sorted and indexed with samtools v0 . 1 . 18 [63] , and assigned strain IDs with picard tools v1 . 55 . Duplicated read pairs were removed and remaining reads locally realigned with GATK v2 . 1–8 [64] . The UnifiedGenotyper was used to call candidate variants across each sample independently . The resulting VCF files were filtered for variants with MQ>40 , GQ>20 , Qual>20 , coverage depth >8X , >2 reads and >15% of reads supporting alternative variant . Around 600kb of the genome–annotated in the SGD database as simple repeats , centromeric regions , telomeric regions , or LTRs were excluded from analysis due to their susceptibility to mismapping and associated miscalls . Custom scripts were written to parse , identify , count , and summarize variants for every frequency , age , mutation type , and neighborhood category . Error bars were calculated as sampling errors where possible , or else estimated with 500 bootstraps . | The mutational spectrum is central to our understanding of molecular evolution . However , mutational spectra are difficult to study because spontaneous mutations are rare , difficult to observe , and a large number of events is required to detect subtle differences between mutational bias , selection and selection like forces . The possibility of estimating mutational spectra from population polymorphism data , with neither the need for tedious experiments nor the restrictions and biases of lab conditions , is a crucial step in overcoming such difficulties . We show that with sufficiently broad population sequencing and proper identification of young polymorphisms , it is possible to recapitulate the experimental yeast mutation spectrum . This holds implications for future applications to all species where population sequencing is possible . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biotechnology",
"fungal",
"genetics",
"invertebrate",
"genomics",
"mutation",
"fungi",
"plant",
"science",
"model",
"organisms",
"substitution",
"mutation",
"experimental",
"organism",
"systems",
"plant",
"genomics",
"saccharomyces",
"research",
"and",
"analysis",
"metho... | 2017 | Extremely Rare Polymorphisms in Saccharomyces cerevisiae Allow Inference of the Mutational Spectrum |
Innate immunity is regulated by cholinergic signalling through nicotinic acetylcholine receptors . We show here that signalling through the M3 muscarinic acetylcholine receptor ( M3R ) plays an important role in adaptive immunity to both Nippostrongylus brasiliensis and Salmonella enterica serovar Typhimurium , as M3R-/- mice were impaired in their ability to resolve infection with either pathogen . CD4 T cell activation and cytokine production were reduced in M3R-/- mice . Immunity to secondary infection with N . brasiliensis was severely impaired , with reduced cytokine responses in M3R-/- mice accompanied by lower numbers of mucus-producing goblet cells and alternatively activated macrophages in the lungs . Ex vivo lymphocyte stimulation of cells from intact BALB/c mice infected with N . brasiliensis and S . typhimurium with muscarinic agonists resulted in enhanced production of IL-13 and IFN-γ respectively , which was blocked by an M3R-selective antagonist . Our data therefore indicate that cholinergic signalling via the M3R is essential for optimal Th1 and Th2 adaptive immunity to infection .
The role of acetylcholine ( ACh ) as a neurotransmitter is well established , both in the central nervous system and the periphery , where it regulates smooth muscle contraction and many other functions of the autonomic nervous system . Cholinergic signalling also influences the immune system , most notably in the cholinergic anti-inflammatory pathway , which results in the α7 nicotinic receptor subunit-dependent inhibition of macrophage TNF-α , IL-1β and IL-6 production [1 , 2] . The influence of cholinergic signalling on adaptive immunity however is largely unexplored , although there is evidence that nicotinic receptors influence B lymphocyte development and activation [3] . Expression of both nicotinic receptors ( nAChRs ) and muscarinic receptors ( mAChRs ) is affected by CD4 T cell activation in vitro [4] , and mAChRs influence differentiation of CD8 T cells in vivo [5] . To our knowledge , nothing is known about the role of mAChRs in the adaptive response to infection . Nippostrongylus brasiliensis is a common laboratory pathogen used to study T helper 2 immune response-mediated disease resolution , and biologically closely resembles the important human hookworms Ancylostoma duodenale and Necator americanus [6] . The Th2 response drives resolution of infection , and IL-13 signalling through IL-4Rα is an important component of the protective response [7] . This signalling pathway also enhances smooth muscle contractility , which is thought to contribute to parasite expulsion [8 , 9] . Previous studies in our laboratory showed delayed parasite expulsion in mice with smooth muscle cells deficient in IL-4Rα . Associated with this defect was reduced Th2 cytokine production , delayed goblet cell hyperplasia and lower expression of the M3 muscarinic receptor ( M3R ) in the intestine [10] . The mAChR family consists of five subtypes ( M1-M5 ) of G protein-coupled receptors [11] , which regulate a range of physiological activities including heart rate , smooth muscle contractility , and endocrine and exocrine gland secretion [12–14] . The M3R is the major mAChR expressed on smooth muscle , and drives contractile responses in the ileum [15] . Our previous investigation determined that upregulation of M3R expression induced by N . brasiliensis infection is related to IL-4Rα , sensitive to host immunity , and may therefore also contribute to the immune response [10] . In this study , we investigated the contribution of signalling through the M3R to protective immunity against N . brasiliensis , using both infection of M3R gene deficient mice ( M3R−/− mice ) and ex vivo CD4 T cell assays . M3R deficiency significantly abrogated the ability of BALB/c mice to launch an effective adaptive immune response to primary and secondary infection , and underlying this defect were reduced CD4 T cell-associated protective cytokine responses . Stimulation of CD4 T cells from N . brasiliensis-infected wild-type ( WT ) control mice with ACh and muscarinic agonists enhanced their secretion of Th2 cytokines , and this effect could be blocked by muscarinic and M3R-selective antagonists . A similar impairment in immunity was found in M3R−/− mice following systemic infection with Salmonella typhimurium . Immunity to S . typhimurium is dependent on a robust Th1 immune response [16] , with production of IFN-γ by CD4 T cells critical for host protection and bacterial clearance [17] . In the absence of M3R expression , strikingly higher bacterial loads were observed , which again correlated with impaired CD4 T cell cytokine responses . Ex vivo stimulation of lymphocytes from S . typhimurium-infected WT mice with ACh and muscarinic agonists enhanced IFN-γ secretion , and this effect was blocked with an M3R-selective antagonist . Collectively , these results show for the first time that the magnitude and efficacy of the adaptive response to infection depends on cholinergic signalling via the M3R .
To determine if expression of the M3R inherently influenced steady-state immune cell populations , we analysed the composition of leukocytes in secondary lymphoid tissues . There was no significant difference in the relative proportions of B cells , CD4 and CD8 T cells , dendritic cells or macrophages in spleens from naïve WT and M3R−/− mice ( Fig . 1A ) , and equivalent proportions of B cells , CD4 and CD8 T cells were also recorded in the mesenteric lymph nodes ( MLN ) ( Fig . 1A ) . M3R expression also did not appear to affect T cell activation status: WT and M3R−/− mice had equal proportions of naïve ( CD3+CD4+CD44loCD62Lhi ) and activated ( CD3+CD4+CD44hiCD62Llo ) CD4 T cells in spleen and MLN ( Fig . 1B ) . Furthermore , in vitro differentiation of CD4 T cells into Th1 populations by addition of IFN-γ and anti-IL-4 antibody , or Th2 populations via IL-4 and anti-IFN-γ antibody , resulted in equivalent cytokine responses in WT and M3R−/− mice ( Fig . 1C ) . However , anti-CD3 or PMA/ionomycin stimulation of either whole MLN cell suspensions or sorted CD4 T cells for 24 hours demonstrated an impaired ability of M3R−/− CD4+ T cells to express markers of activation when compared to WT CD4+ cells ( Fig . 1D ) . This finding shows that an absence of M3R expression in CD4 T cells results in an impaired ability to respond to a non-specific activating stimulus . We determined whether the M3R contributed to CD4+ T cell immune responses in an N . brasiliensis infection . Primary infection with N . brasiliensis was self-resolving in WT BALB/c control mice , with parasite clearance from the small intestine by day 9 post-infection ( p . i . ) . M3R−/− mice exhibited delayed expulsion of worms , with significantly higher recovery of parasites from the intestine at days 7 and 9 p . i . ( Fig . 2A ) . M3R−/− mice have previously been shown to be refractory to ACh-induced smooth muscle contraction [15] , and intestinal smooth muscle hypercontraction is hypothesised to be a potential effector mechanism for expulsion of nematode parasites [9] . We observed increased ACh-driven smooth muscle contraction compared to naïve animals in infected WT mice but not infected M3R−/− mice ( Fig . 2B ) , suggesting that M3R-mediated smooth muscle hypercontractility may indeed contribute to parasite expulsion . Surprisingly , in addition to this altered physiological response , levels of mRNA for IL-13 were reduced 4-fold in mesenteric lymph node ( MLN ) cells of M3R−/− mice compared to those of WT controls at day 7 p . i . ( Fig . 2C ) . This was associated with reduced numbers of IL-13+ CD4 T cells ( Fig . 2C ) . Additionally , M3R−/− mice had reduced CD3+CD4+CD44hiCD62Llo effector memory ( activated ) T cells in the MLN when compared to WT mice ( Fig . 2D ) . Elevation of intracellular calcium ( Ca2+ ) i is an essential event which is necessary for CD4 T cell activation and cytokine production [18] . We tested whether CD4 T cells from N . brasiliensis-infected M3R−/− mice had a reduced ability to mobilise Ca2+i . This was the case; elevation of Ca2+i by the calcium ionophore ionomycin was significantly impaired in M3R−/− CD4 T cells when compared to WT CD4 T cells . No difference was observed between M3R−/− and WT CD4 T cells when extracellular calcium was chelated by EGTA , although the response was much lower , suggesting that lack of the M3R impaired uptake of extracellular calcium across the plasma membrane . These data indicate that M3R expression is required for optimal T cell activation and protective immunity to primary infection by N . brasiliensis . CD4 T cells isolated from the MLN of WT BALB/c mice 7 days p . i . with N . brasiliensis expressed mRNA for the M1 , M3 , M4 and M5 receptor subtypes ( Fig . 3A ) . Cholinergic stimulation influenced several parameters of CD4 T cell function which are important in the immune response to infection . ACh enhanced expression of Ox40 on effector memory CD4 T cells either directly , or in cells stimulated with sub-maximal concentrations of anti-CD3 ( Fig . 3B ) . When cells were treated with anti-CD3 in the presence or absence of ACh or the muscarinic agonists muscarine and oxotremorine-M ( oxo-M ) all agonists enhanced production of IL-4 and IL-13 approximately 2-fold , and potentiation of cytokine production by muscarine was demonstrated to be dose-dependent ( Fig . 3C ) . Cholinergic enhancement of IL-13 secretion was blocked by the pan-specific muscarinic receptor antagonist atropine , and no enhancement was observed in cells from M3R−/− mice ( Fig . 3D ) . Further confirmation that the cholinergic co-stimulatory signal acts through the M3R was provided by use of the M3R-selective antagonist J104129 , which blocked potentiation of cytokine secretion by ACh ( Fig . 3D ) . To test if cholinergic regulation of CD4 T cell cytokine responses influenced adaptive immunity , we next examined how this affected the outcome of secondary infection , as this requires a CD4 T cell-driven Th2 immune response in the lung [19 , 20] . Immunity to secondary infection was strikingly abrogated in M3R−/− mice , with higher numbers of larvae recovered from the lungs in comparison to WT controls ( Fig . 4A ) . Levels of IL-13 and IL-4 in the lungs of M3R−/− mice were significantly decreased ( Fig . 4B ) , and this was reflected by diminished numbers of IL-13+ and IL-4+ CD4 T cells ( Fig . 4C ) , accompanied by reduced goblet cell hyperplasia and associated mucus production ( Fig . 4D ) . Immunity to N . brasiliensis can be accelerated by adoptive transfer of CD4 T cells sensitised by primary infection into naïve mice [20] . We used this approach to test whether optimal adaptive immunity is dependent on signalling through the M3R on CD4 T cells . CD4 T cells from previously infected WT and M3R−/− mice were adoptively transferred i/v into naïve BALB/c mice and animals subsequently infected with N . brasiliensis . Recipients of WT CD4 T cells showed a significant reduction in the numbers of adult worms recovered from the small intestine at day 5 post infection , whereas recipients of M3R−/− CD4 T cells showed no reduction in worm burden ( Fig . 4E ) . This defective ability to reduce worm burden in recipients of M3R−/− CD4 T cells was associated with an impaired induction of pulmonary alternatively activated macrophages ( AAM ) as demonstrated by reduced expression of RELMα and YM1 on alveolar macrophages ( Fig . 4E ) . These data demonstrate that CD4 T cell responsiveness to cholinergic signalling via the M3R is essential for effective adaptive immunity to N . brasiliensis . Having established that cholinergic signalling via the M3R contributed directly to CD4 Th2-driven immunity to nematode infection , we next tested whether this influenced immunity in a Th1 setting . We therefore infected WT and M3R−/− mice with Salmonella typhimurium , control of which is dependent upon CD4 Th1 immune responses [16 , 17] . No difference in recovery of bacteria was observed early in infection ( day 7 ) , but at later time points M3R−/− mice harboured significantly higher bacterial loads in the spleen ( Fig . 5A ) and displayed significant weight loss compared to WT mice ( Fig . 5A ) . This increased susceptibility to infection correlated with reduced IFN-γ secretion by splenocytes at days 18 and 27 p . i . ( Fig . 5B ) . Underlying this reduction in the protective cytokine response was reduced antigen specific IFN-γ production by CD4 T cells as demonstrated by both real-time PCR and flow cytometry analysis ( Fig . 5C ) . Additionally , fewer activated ( CD4+CD44hiCD62Llo ) T cells were found in M3R−/− mice in comparison to WT controls infected with S . typhimurium at day 18 p . i . ( Fig . 5D ) . To confirm that cholinergic signalling influenced T cell cytokine responses during S . typhimurium infection , we carried out ex vivo restimulation experiments with splenocytes removed at day 26 p . i . Both ACh and muscarinic agonists enhanced production of IFN-γ in T cells stimulated sub-maximally with anti-CD3 . Moreover , as with ex vivo stimulation of T cells from nematode-infected mice , the stimulatory effects were blocked by addition of atropine or the M3R-selective inhibitor J104127 ( Fig . 5E ) . Collectively , these data indicate that cholinergic signalling via the M3R plays a common role in enhancing CD4 T cell activation and cytokine responses in both Th1 and Th2 settings .
Understanding of the influence of cholinergic signalling on immune system function has thus far been largely restricted to innate immunity , principally focussed on dissection of the cholinergic anti-inflammatory pathway . The ultimate source of ACh in this reflex response is not neuronal , but a population of splenic noradrenaline-responsive CD4 T cells [21] . B cells also have the capacity of synthesizing and secreting ACh and , like T cells , this can regulate innate immunity , as B cell-derived ACh has been shown to inhibit neutrophil recruitment during sterile endotoxemia [22] . In contrast , little is known about cholinergic regulation of adaptive immunity , and the role of muscarinic receptors , despite numerous studies showing their expression on cells of the immune system [23] . A number of studies have observed that administration of muscarinic antagonists attenuates inflammation in vivo , suggesting a pro-inflammatory role for muscarinic receptors [24 , 25] , although this conflicts with others which suggest that signalling through muscarinic receptors on vascular endothelium suppresses expression of cell adhesion molecules [22] . We demonstrate here that activation of the M3R is essential for optimal immune responses to two different pathogens; one which triggers Th2 cytokine production and another which triggers Th1 cytokines . In the case of primary infection with N . brasiliensis , absence of the M3R was associated with inhibition of smooth muscle contraction , reduced activation of CD4 T cells , lower IL-13 production and delayed expulsion of parasites from the small intestine . The impact of a lack of cholinergic signalling through the M3R was more evident during secondary infection , in which immunological memory in the lungs is particularly important [20] [26] . M3R deficient mice are highly susceptible to secondary infection , with a 4-fold higher recovery of parasites on day 2 p . i . , lower levels of IL-4 and IL-13 in the lungs and reduced goblet cell hyperplasia in the airways . Mucus secretion by goblet cells is an important mediator of immunity to N . brasiliensis [27] , and enhanced mucus production in the airways is correlated with reduced recovery of larvae following secondary infection [28] . Transfer of intact CD4 T cells from a prior infection confers protective pulmonary immunity to N . brasiliensis in naïve mice [20] . Using this approach we demonstrated that WT CD4 T cells from N . brasiliensis-infected mice were able to confer significant protection which was accompanied by expansion of AAMs in the lungs ( which have recently been shown to be protective against lung stage N . brasiliensis infection [26] ) of naive recipients , an effect which was not observed with CD4 T cells from M3R deficient mice . The requirement of cholinergic signalling through the M3R for full CD4 T cell activation and cytokine production is not restricted to Th2 responses , but also extends to synthesis of IFN-γ and protection against bacterial infection . Our demonstration of an impaired response to the calcium ionophore ionomycin in M3R−/− mice ( Fig . 2 ) may well underlie this common M3R-dependent effect on Th1 and Th2 driven immunity , as sustained elevation of Ca2+ is necessary for the action of NFAT which in turn is required for full cytokine transcription in T cells [18] . Muscarinic receptor signalling has been shown to cause the rapid release of Ca2+ from intracellular stores and a sustained influx of external Ca2+ in neuronal cells [29] as well as T and B cell lines [23] . Moreover , the use of antagonists suggested that this was most likely acting through the M3R in lymphocytes [30] . Our demonstration that M3R−/− CD4 T cells from N . brasiliensis-infected mice have an impaired ability to mobilize Ca2+ in response to ionomycin is indicative of a profound defect in cellular function . No difference in recovery of S . typhimurium from the spleens of WT and M3R−/− mice was observed early in infection ( day 7 ) , when control is largely affected through innate myeloid cells [31] . However , eventual resolution of infection is dependent on CD4 T cell-coordinated responses [32] , and higher bacterial loads in M3R−/− mice were associated with reduced IFN-γ from day 18 onwards ( Fig . 5 ) . Others have shown indirectly that cholinergic signalling can protect against bacterial infection . Thus , administration of paraoxon , a specific inhibitor of acetylcholinesterase , rendered mice more resistant to infection with S . typhimurium and enhanced production of IL-12 . These protective effects were abrogated by co-administration of an oxime which reactivated acetylcholinesterase , implicating ACh as the agent driving potentiation of this response [33] . We show conclusively here that cholinergic signalling through the M3R is a major contributor to immunity to parasitic and bacterial infection , and that in its absence CD4 T cell responses are significantly impaired . Moreover , muscarinic agonists enhanced cytokine responses to both pathogens . The immunological defect in both scenarios may also be influenced by the lack of M3R expression on other cell types . However , the impaired immunity to S . typhimurium infection in M3R−/− mice was not related to altered macrophage function and was also independent of M3R influence on B cells [34] . Our data therefore suggest that M3R-dependent protection against infection may primarily be a function of altered CD4 T cell responsiveness . CD4 T cells also express the muscarinic receptors M1 , M3 and M5 . However , our experiments with M3R-deficient mice and use of specific antagonists strongly support the conclusion that cholinergic enhancement of cytokine responses by CD4 T cells is driven by the M3R . This is reinforced by pharmacological experiments which indicate that elevation of intracellular calcium in T cell lines stimulated with muscarinic agonists is most likely effected through the M3R [30] . The current study represents an important development in our understanding of cholinergic signalling in immunity , and may have implications for the use of functionally selective M3R antagonists such as tiotropium , widely used clinically as bronchodilators to treat chronic obstructive pulmonary disease [35] and more recently asthma [36] . Several studies have shown that tiotropium can inhibit allergen-induced eosinophilia and cytokine production [25] , while others conclude that the M3R promotes allergen-induced tissue remodelling and goblet cell hyperplasia without directly affecting inflammation [37] , perhaps indicative of varying effects in different animal models . As expression of the M3R has been demonstrated on other immune cell populations including macrophages and dendritic cells ( DCs ) [23] , these cells may also respond directly to ACh . Studies of human DCs have shown that cholinergic stimulation in vitro enhanced expression of receptors involved in antigen presentation ( CD86 and HLA-DR ) and influenced mixed lymphocyte reactions to these agonists . Moreover , these effects could be blocked by atropine but not mecamylamine , suggesting that signalling was effected via muscarinic rather than nicotinic receptors . However , cholinergic stimulation of LPS-treated DCs led to contrasting effects; namely a reduction in antigen presentation . These findings indicate that complex muscarinic-dependent effects on myeloid cells exist which may differ depending on , for example , the maturation status of cells [38] . Furthermore , a study on human DC from nasal mucosa showed that addition of metacholine to DCs upregulated Ox40L , which can induce and interact with Ox40 on activated T cells [39] . Both these data and our demonstration that ACh upregulates expression of Ox40 on T cells provide a mechanism for cholinergic enhancement of T cell activation and lymphoid-myeloid cell interactions . In summary , our data indicate that signalling through the M3R promotes CD4 T cell activation and cytokine production in two disparate murine models of infection , suggesting that there may be potential for the development of drugs aimed at enhancing immune function for immunoprophylaxis or other applications .
All animal work was approved by the UCT Health Sciences Animal Ethics Committee ( Project licence 012/054 ) to be in accordance with guidelines laid down by the South African Bureau of Standards . Research at Imperial College was approved and in accordance with regulations of the Home Office ( PPL70/6957 ) . M3R−/− mice were backcrossed to BALB/c background for 10 generations for this study . Mice were bred under specific pathogen-free conditions and used aged 6–8 weeks . Protocols for all experiments were reviewed and approved by the UCT and ICL Animal Ethics committees . For primary parasite infection , mice were infected subcutaneously with 500 N . brasiliensis infective larvae . To enumerate adult worms , mice were killed at various times post-infection ( p . i . ) , intestines opened longitudinally , incubated in 10 ml saline for 3 hrs at 37°C and parasites counted under a dissecting microscope . For secondary infections , mice were treated at day 9 p . i . with ivermectin via drinking water to eliminate parasites , rested for 28 days and re-infected with 500 L3 . Larvae were recovered from lungs by finely slicing the tissues , placing them in 5 ml saline for 3 hrs , and parasites subsequently enumerated . Salmonella enterica serovar Typhimurium aroA− ( SL3261 ) is an attenuated strain of STm SL1344 ( 44 ) and was maintained and used to infect mice intraperitoneally with an infectious dose of 5 × 105 CFU as described previously [40] . Tissue bacterial burdens were determined by direct culturing . CD4 T cells were isolated from mesenteric lymph nodes of naïve mice using flow cytometry ( >99% purity ) . Sorted CD4 T cells were plated at 1 × 105 cells/well on plates coated with 10 μg ml−1 anti-CD3 ( BD Bioscience ) and 5 μg ml−1 anti-CD28 ( BD Bioscience ) . Th1 polarization conditions: 5 ng ml−1 rIL-12 ( BD Bioscience ) and 50 μg ml−1 anti-IL-4 ( homemade Clone: 11B11 ) ; Th2 polarization: 50 ng ml−1 rIL-4 ( BD Bioscience ) and 50 μg ml−1 anti-IFN-γ . Cells were cultured in a final volume of 100 μl for 72 hrs , then transferred to fresh round-bottom 96 well plates and resuspended in appropriate antibody cocktails with the addition of 20 U ml−1 IL-2 ( BD Bioscience ) and cultured for another 48 hrs . Finally , cells were plated at 2 × 105 cells/well and incubated in 96 well plates coated with 20 μg ml−1 anti-CD3 ( BD Bioscience ) . Supernatants were harvested after 48 hrs restimulation and used for ELISA . Cytokine ELISAs were performed as previously described [41] using coating and biotinylated detection antibodies from R&D , with the exception of homemade coating antibodies for IL-4 ( 11B11 ) and IFN-γ ( ANK18KL6 ) . Streptavidin-conjugated HRP was used for detection with a commercially available substrate solution . MLN and lung cells were plated at 1 × 106 cells per well in 48 well plates pre-coated with 20 ug ml−1 anti-CD3 and restimulated for 72 hrs . Homogenates of lung and intestinal sections were prepared using a Polytron homogenizer and all samples standardized to 5 mg ml−1 protein prior to ELISA . Changes in the cytosolic concentration of free Ca2+ were measured using the calcium indicator Fura-4-AM . Cellular suspensions of MLN were stained with anti-CD3 and anti-CD4 antibodies and then washed and resuspended to a concentration of 0 . 5 × 107 cells/ml in Ca2+ flux buffer ( Hank’s balanced salt solution ( HBSS ) containing 1 mM CaCl2 , 1 mM MgCl2 , and 0 . 1% BSA ) , and labelled with 5 μM Fura-4-AM for 30 min at 37°C in the dark . Labelled cells were washed in Ca2+ flux buffer . Changes in Ca2+ in CD3+CD4+ cells following stimulation with 10 μM ionomycin in the presence or absence of 10 μM EGTA were monitored by flow cytometry . Cells were stimulated with 0 . 1 μg ml−1 ( sub-optimal ) anti-CD3 , 10 μM ACh + 10 μM BW284C51 , a specific inhibitor of acetylcholinesterases used to increase the half-life of acetylcholine ( ACh ) , 10 μM Oxotremorine M ( Oxo M ) , 10 μM muscarine or buffer controls for 48 hrs . The muscarinic receptor antagonist atropine ( AT ) was used at a concentration of 100 μM , and the M3R-selective antagonist J104129 [42] at a concentration of 40 nM . Supernatants were analysed for cytokines as described . Jejunum sections ( 1 cm ) were obtained and hooked onto a force transducer , placed in PBS maintained at 37°C in an organ bath , and stimulated with ACh from 10−9 M to 10−3 M . In between stimulations , the intestinal segment was allowed to return to baseline contraction ( at least 5 min ) . All measurements were recorded using the Powerlab acquisition unit and analysed using the Chart5 program . The amplitude was measured as the difference between the peak and trough of the contraction and reported in millinewtons ( mN ) . Single cell suspensions were prepared and 1 × 106 cells incubated in PBS + 0 . 1% BSA , 1% normal rat serum and appropriate antibody cocktails . Cell populations were determined and acquired on a BD FACS Fortessa ( Becton Dickinson ) . Cell populations were identified by the following antibody staining strategies: CD4 T cells: CD3+CD4+; CD8 T cells: CD3+CD8+; B cells: CD19+B220+; Macrophages: CD11b+F4/80+; Dendritic cells: CD11c+ . CD4 T cells populations were additionally stratified into naïve ( CD44loCD62Lhi ) and activated ( CD44hiCD62Llo ) T cell populations and stained for Ox40 ( CD134 ) . Alternatively activated macrophages were characterised by staining for YM1 and RELMα . Intra-cellular cytokine staining was carried on MLN , spleen or lung cells . Cells were re-suspended in complete media ( IMDM ( GIBCO/Invitrogen; Carlsbad , CA ) , 10% FCS , P/S ) at 2 . 5×107/ml and stimulated with either10 μg/ml PMA/ionomycin or antigen and GolgiStop ( as per manufacturer’s protocol; BD Pharmingen ) at 37°C for 4 hours . After re-stimulation , cells were surface stained for CD3 , CD4 then fixed and permeabilized with Cytofix/ Cytoperm Plus ( as per manufacturer’s instructions; BD Pharmingen ) . Intracellular staining was performed by staining cells with either IL-13 , IFN-γ or appropriately labeled isotype control . All analyses were performed with FlowJo software . CD4 T cells were purified from MLNs by positive selection using CD4 MACS beads ( L3T4 , MACS Miltenyi ) according to the manufacturer’s protocol . Cells were further purified by flow cytometry to obtain purities above 95% , and 5 × 105 purified CD4 T cells from infected or naive animals transferred into naive WT or M3R−/− mice intravenously . Recipient mice were infected 24 hrs later with 500 L3 and killed 5 days post infection . RNA was extracted using the Qiagen RNeasy Mini kit as per manufacturer’s protocols . RNA was converted to cDNA using random primers and Superscript II . The following primer pairs were used M1R: 5’-GGACAACAACACCAGAGGAGA-3’; 5’-GAGGTCACTTTAGGGTAGGG-3’; M2R: 5’-TGAAAACACGGTTTCCACTTC-3’ , 5’-GATGGAGGAGGCTTCTTTTTG-3’; M3R: 5’-TTTACATGCCTGTCACCATCA-3’ , 5’-ACAGCCACCATACTTCCTCCT-3’; M4R: 5’-TGCCTCTGTCATGAACCTTCT-3’ , 5’-TGGTTATCAGGCACTGTCCTC-3’; M5R: 5’-CTCTGCTGGCAGTACTTGGTC-3’ , 5’-GTGAGCCGGTTTTCTCTTCTT-3’; β-actin: 5’-TGGAATCCTGTGGCATCCATGAAAC-3’ , 5’-TAAAACGCAGCTCAGTAACAGTCCG-3’; IL-13: 5’-CTCCCTCTGACCCTTAAGGAG-3’; 5’-GAAGGGGCCGTGGCGAAACAG-3’ Lung and intestinal sections were fixed with 4% formalin in PBS solution , embedded in wax and cut into sections , then stained with Periodic Acid Schiff ( PAS ) stain to distinguish mucus-producing goblet cells . The histological mucus index ( HMI ) was used to quantify pulmonary goblet cell hyperplasia in individual mice , as described before ( 49 ) . Values are expressed as mean ± standard deviation and significant differences were determined using either Mann-Whitney U test or ANOVA ( GraphPad Prism4 ) . | Recent data indicate that acetylcholine ( ACh ) , a neurotransmitter which regulates a variety of physiological functions , also influences the immune system , and that lymphocytes have the capacity to synthesise and release ACh , controlling local innate immune responses and suppressing inflammation . Thus far however there has been little evidence to suggest that ACh influences adaptive immunity , characterised by activation and effector functions of lymphocytes . We show here that during the immune response to two different pathogens , ACh signals through muscarinic receptors , and the M3 receptor subtype specifically , resulting in enhanced activation and cytokine production by ‘helper’ T lymphocytes which protect the host against infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The M3 Muscarinic Receptor Is Required for Optimal Adaptive Immunity to Helminth and Bacterial Infection |
Transcriptional activation in response to hypoxia in plants is orchestrated by ethylene-responsive factor group VII ( ERF-VII ) transcription factors , which are stable during hypoxia but destabilized during normoxia through their targeting to the N-end rule pathway of selective proteolysis . Whereas the conditionally expressed ERF-VII genes enable effective flooding survival strategies in rice , the constitutive accumulation of N-end-rule–insensitive versions of the Arabidopsis thaliana ERF-VII factor RAP2 . 12 is maladaptive . This suggests that transcriptional activation under hypoxia that leads to anaerobic metabolism may need to be fine-tuned . However , it is presently unknown whether a counterbalance of RAP2 . 12 exists . Genome-wide transcriptome analyses identified an uncharacterized trihelix transcription factor gene , which we named HYPOXIA RESPONSE ATTENUATOR1 ( HRA1 ) , as highly up-regulated by hypoxia . HRA1 counteracts the induction of core low oxygen-responsive genes and transcriptional activation of hypoxia-responsive promoters by RAP2 . 12 . By yeast-two-hybrid assays and chromatin immunoprecipitation we demonstrated that HRA1 interacts with the RAP2 . 12 protein but with only a few genomic DNA regions from hypoxia-regulated genes , indicating that HRA1 modulates RAP2 . 12 through protein–protein interaction . Comparison of the low oxygen response of tissues characterized by different levels of metabolic hypoxia ( i . e . , the shoot apical zone versus mature rosette leaves ) revealed that the antagonistic interplay between RAP2 . 12 and HRA1 enables a flexible response to fluctuating hypoxia and is of importance to stress survival . In Arabidopsis , an effective low oxygen-sensing response requires RAP2 . 12 stabilization followed by HRA1 induction to modulate the extent of the anaerobic response by negative feedback regulation of RAP2 . 12 . This mechanism is crucial for plant survival under suboptimal oxygenation conditions . The discovery of the feedback loop regulating the oxygen-sensing mechanism in plants opens new perspectives for breeding flood-resistant crops .
In higher plants , respiratory metabolism requires molecular oxygen as the terminal electron acceptor to generate ATP . Limited oxygen availability ( hypoxia ) can occur in plant cells due to floods , frosts , and excessive respiration , requiring physiological acclimation to constraints in ATP availability for growth and development [1] , [2] . The switch from aerobic respiration to anaerobic ethanolic fermentation , as a means to maintain substrate-level ATP production from available carbohydrates , is essential for plant survival in conditions of oxygen deprivation [3] , [4] . For instance , mutants lacking pyruvate decarboxylase ( PDC ) or alcohol dehydrogenase ( ADH ) , key enzymes in ethanolic fermentation , are less tolerant to hypoxia and soil waterlogging [4]–[6] . On the other hand , uncontrolled or constitutive fermentation is also detrimental to plant survival , due to rapid depletion of carbohydrate resources needed for basic cellular homeostasis [7] . The repression of catabolic metabolism is the basis of the quiescence survival strategy of the flash-flood–tolerant varieties of rice that have been recently adopted by many farmers in South and Southeastern Asia [8] . Molecular responses must be , thus , accurately balanced to meet plant requirements for survival under fluctuating oxygen conditions [9] . Low oxygen responses are coordinately regulated in plants by ethylene-responsive factor group VII ( ERF-VII ) transcription factors ( TFs ) , primary activators of anaerobic gene expression [1] . In Arabidopsis thaliana , the presence of the ERF-VII factor RAP2 . 12 in the nucleus is inversely correlated to cellular oxygen levels , due to an oxygen-dependent branch of the N-end rule pathway of targeted protein degradation [7] , [10]–[12] . It has been observed that constitutive accumulation of versions of RAP2 . 12 that are insensitive to the N-end rule degradation leads to a decreased submergence or hypoxia stress tolerance , whereas overexpression of the native RAP2 . 12 factor improves survival [7] . Therefore , excessive up-regulation of the stress-responsive genes appears to be detrimental , leading to the hypothesis that fine-tuning of transcription is a prerequisite for cellular homeostasis under hypoxia . Among the genes that are induced by oxygen deficiency , those encoding known or putative TFs deserve special attention as candidate modulators of transcription under hypoxia . When hypoxia-responsive genes are compared across different plant species , a few TFs in addition to ERF-VIIs are consistently up-regulated by oxygen deprivation [13] . These include zinc finger , MADS , LOB domain proteins , and trihelix TF gene family members . The trihelix family , in particular , encompasses plant-specific TFs that have been so far linked to embryo development , trichome formation , seed shattering , and tolerance to biotic and abiotic stresses [14] . This study was aimed at the molecular and physiological characterization of a hypoxia-inducible trihelix TF gene ( At3g10040 ) , which we named HRA1 ( HYPOXIA RESPONSE ATTENUATOR 1 ) . Here , we present evidence that HRA1 encodes a transcriptional repressor that attenuates the anaerobic response induced by ERF-VIIs in a tissue-specific manner . We show that HRA1 imposes an additional level of negative regulation on RAP2 . 12 , besides the ERF-VII's oxygen-dependent instability . RAP2 . 12 transcriptionally activates HRA1 , which in turn binds RAP2 . 12 and restrains its function . Additionally , HRA1 interacts with its own promoter , limiting its activation by RAP2 . 12 through a negative feedback mechanism . Thus , transcriptional activation by RAP2 . 12 is controlled under normoxia by its N-end rule susceptibility and under oxygen deficiency by HRA1 . The spatial and temporal regulation of both factors appears to be a key to modulation of transcriptional activity and survival of transient hypoxia .
The A . thaliana Columbia-0 genome encodes 30 genes belonging to the plant-specific family of trihelix TFs [14] . A survey of public transcriptomic data showed that the gene At3g10040 ( HRA1 ) is the only trihelix family member up-regulated by oxygen deprivation ( Figure S1 ) . Trihelix TFs are induced by low oxygen in different species ( Table S1 and Figure S2 ) and therefore appear to be a component of the conserved stress response strategy in land plants [13] . Moreover , the HRA1 transcript , detected at medium-to-low levels throughout the plant life cycle ( Figure S3A ) , was most strongly induced by short-term oxygen deficiency in plants subjected to a range of abiotic stress conditions ( Figure 1A ) . Hypoxia enhanced the activity of the HRA1 promoter , as visualized by means of a promHRA1:GUS transgenic line ( Figure 1B ) , and led to over 15-fold elevation of HRA1 mRNA in both the leaves and roots of seedlings ( Figure S3B ) . A survey of our previously generated data of polysome-associated transcripts under the same hypoxia system [15] showed that this was accompanied by active loading of HRA1 mRNA onto polysomes ( Figure S3C ) , indicating that the synthesis of HRA1 protein occurs during the stress . When HRA1 expression was monitored over time in hypoxia-treated seedlings , HRA1 mRNA accumulation was induced rapidly but transiently , whereas that of hypoxia marker ADH1 increased slowly and steadily during the stress ( Figure 1C ) . This peculiar dynamics of gene expression hinted at a possible role for HRA1 in the early phase of the low oxygen response . A green fluorescent protein ( GFP ) reporter fusion demonstrated that the HRA1 protein localized to the cell nucleus ( Figure 2A ) , consistent with its prediction as a TF . To investigate the role of HRA1 in transcription , we performed a microarray analysis and compared the hypoxic reconfiguration of the transcriptome between Cauliflower Mosaic Virus 35S:HRA1:FLAG transgenics ( OE-HRA1 ) and wild type Arabidopsis seedlings ( Figure 2B and Table S2 ) . Overexpression of HRA1 significantly reduced the up-regulation of 30 out of the 49 ( 61% ) core hypoxia-responsive genes [13] induced in the wild type after short-term ( 2 h ) hypoxia ( |SLR|≥1 , FDR<0 . 01 ) ( Figure 2C ) , revealing the ability of HRA1 to broadly affect the hypoxic transcriptome . This included genes encoding key enzymes for anaerobic metabolism , such as the marker genes ADH1 and PDC1 . The inhibition of hypoxic transcript accumulation by HRA1 overexpression contrasted to the constitutive up-regulation of the hypoxia-responsive genes observed in 35S:HA:RAP2 . 12 transgenic seedlings , which constitutively accumulate RAP2 . 12 due to masking of its N-terminus from the N-end rule machinery [7] . The hypoxic induction of 43% ( 7/16 ) of the RAP2 . 12 up-regulated genes was dampened by ectopic HRA1 expression ( Figure 2C ) , suggesting there is antagonism between HRA1 and RAP2 . 12 . Consistent with the dramatic reduction in ADH1 mRNA up-regulation , we found out that ADH activity was significantly repressed in hypoxic OE-HRA1 seedlings ( Figure 2D ) . It was this squelching of low oxygen induction of many hypoxia-responsive genes that led us to name At3g10040 HYPOXIA RESPONSE ATTENUATOR 1 . The impact of HRA1 on hypoxia-responsive gene expression prompted us to assess how altered levels of HRA1 expression affect plant performance under submergence-induced hypoxia . We compared two independent 35S:HRA1:FLAG transgenic genotypes ( OE-HRA1#1 and #2 ) ( Figure S4A ) and two independent T-DNA insertion homozygous mutants ( hra1-1 and hra1-2 ) ( Figures S4B–C and S5 ) with the wild type . This revealed that both overexpression and failure to produce a full-length HRA1 transcript reduced the ability of plants to withstand the stress . When tested for tolerance to complete submergence in the dark with two distinct experimental setups , wild type Arabidopsis plants endured the stress significantly longer than OE-HRA1 plants at the 10-leaf rosette stage ( Figure S6A ) and , in older rosettes prior to bolting , recovered better than either OE-HRA1 or hra1-1 plants ( Figure 3A and B ) . Underwater petiole elongation , a trait recognized as a part of the escape strategy from flooding in semiaquatic species ( i . e . , deepwater rice and the wetland species Rumex palustris ) [8] , was unaffected by HRA1 ( Figure S6B ) , consistent with previous reports of a limited overall correlation between the trait and flooding survival of mutants in genes up-regulated by hypoxia and Arabidopsis accessions [16] , [17] . Moreover , the analysis of the total soluble carbohydrate content in plants prior to submergence allowed us to rule out that a significant difference in the available reserves accounts for the poorer performance of the noticeably smaller OE-HRA1#1 plants ( Figure S7 ) . This was again in line with previous reports of a lack of correlation between carbohydrate content before submergence and stress survival in Arabidopsis [17] . The observation that altered HRA1 levels modified performance under submergence , at two stages of rosette development and in distinct growth environments , supports the hypothesis of a distinct role for the factor during the stress . A closer examination of the phenotype of the plants at the end of the recovery period revealed that susceptibility to submergence-induced hypoxia differed in young and older rosette leaves . As compared to the wild type , young leaves emerging from the shoot apex and the shoot apical meristem region were more sensitive in the hra1-1 mutant , and generally unable to recover during postsubmergence . Contrastingly , the fully expanded and mature leaves of OE-HRA1#1 plants were more sensitive to dark submergence than the wild type , but the shoot apical meristem performance was less damaged ( Figure 3A , see magnified insert ) . This suggested that HRA1 is imperative for an effective anaerobic response in the meristematic zone and young leaves . Following preliminary evaluation of temporal regulation of HRA1 expression in rosette leaves during submergence , which demonstrated an early peak of gene expression after 2 h of stress ( Figure S8 ) , we selected 4 h of submergence as a suitable time to study distinctions in gene transcript and protein accumulation in young and fully expanded leaves of wild type and HRA1 mutant genotypes . Firstly , we then found out that in young leaves , but not in older ones , the expression of the hypoxia marker genes ADH1 and PDC1 were differentially regulated by manipulation of HRA1 . In control and submergence treated plants , ADH1 and PDC1 expression was enhanced in hra1-1 and dampened in young OE-HRA1 rosette leaves as compared to the wild type ( Figure 3C ) . In all three genotypes , hypoxic gene expression was promptly reversed to presubmergence levels upon reaeration of the plants ( Figure 3C , “Reoxygenation” ) , in agreement with earlier studies [7] , [18] . The enhancement in anaerobic gene expression observed in young leaves of hra1-1 was also seen in the independent hra1-2 mutant ( Figure S9 ) , reinforcing the hypothesis that mutation of HRA1 leads to altered regulation of the hypoxic response . Secondly , Western blot analyses performed to detect the products of ADH1 and all five PDCs [19] indicated that elevated levels of these transcripts in young leaves of hra1-1 plants was accompanied by higher hypoxic production of the encoded proteins ( Figures 3D and S10 ) . Although in older leaves some enhancement of PDC accumulation in submergence was visible in the mutant as compared to the wild type , ADH was always below the limit of detection . In younger leaves , contrastingly , ADH protein levels were enhanced in hra1-1 already under normoxic conditions . These results suggest that HRA1 plays a key role in negatively regulating the induction of ADH1 and PDC1 in younger tissues of rosette-stage plants . We also considered that HRA1 expression could impact steady-state levels of its upstream regulator RAP2 . 12 . To do so , we evaluated the effect of HRA1 on RAP2 . 12 protein stability , using mesophyll protoplasts that were transiently transfected with a 35S:RAP2 . 12:RrLuc plasmid construct . The stability of RAP2 . 12 was inferred from the activity of a C-terminal translational fusion of RAP2 . 12 to the Renilla reniformis luciferase ( RrLuc ) reporter . As RrLuc activity was unaffected by concurrent transfection of the 35S:HRA1 effector , we conclude that HRA1 expression does not affect RAP2 . 12 stability , at least in isolated leaf protoplasts ( Figure 3E ) . Altogether , these data , presented in Figure 3 , support the conclusion that HRA1 acts to limit accumulation of transcripts and the encoded proteins associated with anaerobic metabolism , even in air , particularly in younger rosette tissue . The submergence survival studies suggested that HRA1 expression provides vital control of the anaerobic response in the meristematic region and young leaves . To further investigate the spatial and temporal regulation of HRA1 , we monitored transgenics expressing promHRA1:GUS . The beta-glucuronidase ( GUS ) reporter confirmed that basal HRA1 promoter activity , detectable under normal growth conditions , was restricted to the shoot apical region and leaf vasculature in aboveground tissues ( Figure S11 , “Control” ) , and was pronounced in roots as well ( Figure 1B ) . This pattern of expression is consistent with the hypothesis that HRA1 is active in cells experiencing physiological hypoxia , due to higher oxygen demand , lower permeability to oxygen , or a hypoxic environment [20] , [21] . The elevated levels of hypoxia marker gene transcripts in the shoot apical area under normoxia suggest that this region is physiologically hypoxic ( Figure 3C ) , as reported previously [22] . By use of the promHRA1:GUS transgenics , we also determined that submergence primarily enhanced GUS activity in the younger rosette tissues and to a much lesser extent in adult leaves , except within the vasculature ( Figure S11 , “Submergence” ) . This pattern of GUS staining correlated well with the tissue-specific effect exerted by HRA1 on submergence tolerance , as described above ( Figure 3A ) . In genotypes with altered HRA1 expression , the absence of a fully functional HRA1 protein in hra1 shoot meristem tissue led to its higher susceptibility to submergence . On the other hand , ectopic expression of HRA1 in older rosette leaves of OE-HRA1 plants reduced their survival of submergence and prolonged darkness , possibly due to accelerated senescence of older leaves ( Figure S12 ) . Our data showed that HRA1 balances low oxygen acclimation responses , but also suggested that proper spatial expression of HRA1 is required for normal vegetative development . Overexpression of either the native HRA1 protein in OE-HRA1#3 plants or a FLAG-tagged protein in OE-HRA1#1 and #2 plants caused a pleiotropic phenotype that included reduced rosette size , due to shortened petiole length and altered leaf index , slower rosette growth ( Figure S13A and B ) , increased leaf anthocyanin content ( Figure S13B ) , partial loss of apical dominance , delayed flowering , and reduced seed production ( Figure S13C ) . The conservation of these phenotypes across three independent transgenic lines allowed us to recognize their cause in the ectopic expression of high HRA1 levels in the whole plant , rather than ascribe it to random integration of the transgenes in unrelated genomic loci . We speculate that , although sustained HRA1 expression is beneficial in rapidly dividing and expanding leaf primordia under normal growth conditions , abnormal HRA1 accumulation in mature leaves has a negative impact on plant development . Moreover , because hra1-1 and hra1-2 showed no differences from the wild type under normal growth conditions ( Figure S13 ) , we can conclude that mutations in the 3′ region of the HRA1 transcript ( Figure S4B and S5 ) particularly affect the hypoxic pathway ( Figure 3A–D ) but do not relate to the developmental phenotypes shown here . To gain insight into the role of HRA1 dampening core hypoxia gene transcription during hypoxia , we performed chromatin immunoprecipitation followed by deep sequencing ( ChIP-Seq analysis ) using seedlings deprived of oxygen for 2 h , in the same manner as the transcriptome analysis . To identify chromatin bound by HRA1-FLAG , OE-HRA1#1 and Col-0 seedling tissue was cross-linked , nuclei were isolated , and immunopurification performed with a FLAG antibody . The Col-0 sample was used as a control to monitor nonspecific immunopurification . Deep sequencing of ∼100 bp fragments yielded 146 peak-to-gene associations ( Table S3 ) , corresponding to putative HRA1 target genes , 42% of which ( 62 elements ) fell 5′ of the predicted transcription start sites ( Figure S14A ) . We then focused on the 1 , 295 differentially expressed genes ( DEGs ) in the microarray dataset and found out that seven of the HRA1 binding sites resided on genes significantly regulated by hypoxia and/or HRA1 overexpression ( HRA1; RAV1 , At1g13260; HUP7 , At1g43800; a hydroxyproline-rich glycoprotein family protein coding gene , At1g31310; WRKY7 , At4g24240; CYP78A6 , At2g46660; DCL1 , At1g01040 ) ( Figure S14B ) . The small number of stress-responsive genes identified by ChIP of HRA1 suggests that its effect on the hypoxia-responsive gene transcription may be mediated by other DNA binding factors , rather than HRA1's ability to recognize DNA . Finally , of the candidate targets of HRA1 , only HUP7 and HRA1 itself ( Figure 4A and Table S4 ) were constitutively up-regulated by HA-RAP2 . 12 and less hypoxia-induced in OE-HRA1 plants ( Figure S14C ) . This led us to consider that repression of hypoxia-responsive genes by HRA1 was largely independent of its direct association to chromatin of the genes it regulates . We hypothesized that HRA1 could attenuate hypoxia-responsive gene expression by directly inhibiting RAP2 . 12 activity . To address this , we first examined whether protein–protein interaction occurs between HRA1 and RAP2 . 12 . Previously , interaction between rice GTγ-clade trihelix protein LOC_Os11g06410 ( SAB18 ) ( Figure S2 ) and the ERF-VII TFs SUBMERGENCE1A ( SUB1A ) and the related SUB1C ( LOC_Os09g11460 ) was reported in a nondirected yeast-two-hybrid screen [23] . We confirmed that HRA1 and RAP2 . 12 interact in the heterologous yeast-two-hybrid system ( Figures 4B and S15A–C ) . By systematically testing different combinations of full-length and truncated versions of HRA1 and RAP2 . 12 ( Figure S15A ) , we determined that , firstly , the conserved C-terminal region rather than the trihelix domain of HRA1 was required for RAP2 . 12 association ( Figures 4B and S15B ) and , secondly , the N-terminal portion of the ERF-VII ( RAP2 . 121–123 ) was sufficient for interaction ( Figure S15C ) , while its DNA binding domain might enhance the association ( Figure S15B ) . The interaction between HRA1 and RAP2 . 12 was subsequently validated in planta by means of bimolecular fluorescence complementation using Arabidopsis protoplasts ( Figure 4C ) . We also tested HRA1 interaction with the other four Arabidopsis ERF-VII factors in the yeast-two-hybrid system ( Figure 4B ) . Interestingly , only RAP2 . 12 interacted with the HRA1 protein , consistent with the fact that the RAP2 . 12 1–123 region , which was sufficient for binding , is poorly conserved among Arabidopsis ERF-VII sequences ( with the exception of the N-terminal region , critical to N-end rule regulation , which was anyway not required for HRA1 interaction in planta; Figure 4C ) . It remains an open question why no interaction was observed with RAP2 . 2 , which has the highest level of sequence similarity with the RAP2 . 12 protein in the interaction region . With the knowledge that HRA1 and RAP2 . 12 interact , we investigated if the interaction could account for HRA1-mediated attenuation of RAP2 . 12-driven transcriptional activation . Towards this goal , we used a firefly luciferase ( PpLuc ) reporter fusion to measure the activity of the RAP2 . 12-responsive PDC1 promoter ( −911 to −1 relative to the start codon ) in transiently transfected Arabidopsis mesophyll protoplasts . We confirmed that 35S:RAP2 . 1214–358 effector plasmid DNA enhanced the luciferase activity of promPDC1:PpLuc ( Figure 4D ) . When increasing amounts of 35S:HRA1 were co-transfected , the luciferase activity gradually fell towards basal levels , indicating that RAP2 . 12 potential in promPDC1:PpLuc transactivation was negatively affected by HRA1 , presumably due to the interaction between factors . Inhibition of promPDC1:PpLuc expression was similarly achieved by concurrent expression of 35S:HRA1194–431 ( Figure 4D ) , which lacked the trihelix DNA binding domain ( Figure S15A ) but retained the C-terminal region that allowed interaction with RAP2 . 12 in the yeast-two-hybrid assay ( Figures 4B and S15B ) . This demonstrates that the repression of RAP2 . 12 activation of PDC1 transcription by HRA1 was independent of the trihelix domain and , therefore , most likely the TF's binding of DNA . This is consistent with the absence of PDC1 and many other HRA1-regulated genes in the immunoprecipitated chromatin . To further validate the hypothesis that HRA1 inhibits RAP2 . 12 through direct interaction rather than DNA binding , we took advantage of an artificial UAS promoter , made up of four repetitions of the yeast GAL4 upstream activating sequence [7] , which cannot be recognized by endogenous plant factors . By this approach we confirmed that the activation of the UAS:PpLuc construct by a chimeric RAP2 . 12-GAL4DBD factor was inhibited by coexpression of 35S:HRA1 in protoplasts , in spite of the inability by HRA1 to recognize the UAS promoter ( Figure 4E ) . Altogether , these observations demonstrated that HRA1 inhibits RAP2 . 12 function by direct protein–protein interaction , rather than by competition for DNA binding . Additional evidence of the impact of RAP2 . 12 inhibition by HRA1 was obtained in protoplasts , whose survival of hypoxia was enhanced by transfection with 35S:RAP2 . 12 only if 35S:HRA1 was not concurrently transfected ( Figure S16A ) , and in planta , where overexpression of a stabilized RAP2 . 1214–358 protein in the OE-HRA1#1 background was sufficient to suppress the alteration in OE-HRA1 rosette morphology and return the overall phenotype to that of the wild type ( Figure S16B ) . Because the molecular response to hypoxia might entail a balance between RAP2 . 12 stabilization and attenuation under low oxygen stress , tight regulation of HRA1 was anticipated . As for the promoter of PDC1 , we found that the HRA1 promoter ( −849 to −1 relative to the start codon ) was transactivated in a dosage-dependent manner by RAP2 . 12 and repressed by HRA1 itself in mesophyll protoplasts ( Figure 5A ) . An additional mechanism contributing to HRA1 regulation involves binding of HRA1 to its own promoter , as revealed by ChIP-Seq and confirmed by ChIP-qPCR ( Figure 4A ) . We hypothesize that this binding dampens RAP2 . 12 activation of this promoter . In support of this , when specific RT-qPCR was performed to detect the expression of the endogenous HRA1 gene ( Figure S4B ) , strong down-regulation was observed in OE-HRA1 plants ( Figure 5B ) . These results demonstrate that HRA1 transcription is activated upon hypoxia following the nuclear accumulation of RAP2 . 12 but is subsequently subjected to negative self-regulation . This can be considered a “double check” mechanism that takes advantage of HRA1's ability to both repress RAP2 . 12 activity and directly bind its own promoter , possibly competing with RAP2 . 12 binding . The double regulation of HRA1 transcription is most likely responsible for the transient dynamics of HRA1 transcript accumulation during hypoxia ( Figure 1C ) and allows the plant to limit hypoxic gene expression over time , as detected in the wild type and to a lesser extent in the hra1-1 mutant ( Figure 5C ) .
Gene expression is tightly regulated in response to low oxygen stress . In order to maximize the efficiency of ATP utilization , the transcription of many genes , whose function is not essential for survival , is repressed under low oxygen stress , whereas polyribosomes dissociate from their mRNA to limit translation [18] , [24] . At the same time , the metabolism of plants is adapted to hypoxia through a reconfiguration of the energetic pathways that enables fermentation to maintain substrate-level ATP production through glycolysis after replacement of the oxidative phosphorylation [1] , [8] . This requires transcriptional activation of genes such as PDC1 and ADH1 , encoding essential enzymes for ethanolic fermentation . Although the transcriptional rearrangement following exposure to hypoxia is not limited to the expression of fermentation-related genes , this pathway contributes largely to survival in low oxygen conditions , as mutants lacking PDC and ADH genes are hypersensitive to hypoxia and conditions with a hypoxic component [4]–[6] . Moreover , the transcriptional induction of ADH and PDC genes is a conserved feature in the anaerobic response of all higher plants studied so far [13] . Transcriptional activation of fermentative genes is downstream of the oxygen-sensing machinery , which relies on the N-end-rule–dependent stabilization of the ERF-VII TFs , such as RAP2 . 12 [1] . However , this mechanism may not be an on–off process but rather modulated in intensity by additional hypoxic players , as both environmental fluctuations in oxygen availability [25] , as well as local hypoxic microenvironments in developing tissues and organs [20] , [26] , may necessitate temporal and spatial flexibility in the hypoxic response . This is because stabilization of RAP2 . 12 would trigger activation of the core hypoxia genes , with their down-regulation reliant upon reoxygenation and the destabilization of RAP2 . 12 . Such inflexibility could expose cells to unregulated fermentative metabolism that may rapidly exhaust the limited respiratory substrates [27] , [28] , preventing endurance of prolonged stress and limiting recovery upon reoxygenation . Here , we show that the strategy adopted by cells to respond to decreased oxygenation entails the induction of a repressor of hypoxic gene expression , the nuclear-localized trihelix protein HRA1 , and confirm this protein acts as a direct attenuator of the low oxygen stabilized transcriptional activator RAP2 . 12 in Arabidopsis . It cannot be excluded that HRA1 may mediate additional mechanisms of repression , starting from the cascade activation of hypoxia-specific transcriptional repressor ( s ) , either at the transcriptional or post-translational level of regulation ( i . e . , activation of a repressor via protein-protein interaction ) . The absence of candidate transcriptional repressors among HRA1 targets ( according to our microarray and ChIP-seq analyses ) , along with HRA1's ability to restrain anaerobic promoter activation even after ablation of its DNA binding domain ( Figure 4D ) , supports the conclusion that attenuation of RAP2 . 12 by HRA1 is not accomplished through DNA binding . Contrastingly , HRA1's ability to bind its own promoter appears to provide a second tier of activity , namely inhibition of its transcription . The present study expands the knowledge of the hypoxia-response transcription network mediated by the low oxygen stabilized ERF-VIIs . The fast induction of HRA1 , notably directed by RAP2 . 12 at the onset of hypoxia , confers the ability to prevent excessive expression of anaerobic genes , particularly in younger tissues exposed to submergence ( Figure 3C ) . The up-regulation of the attenuator HRA1 serves to limit the activity of stabilized RAP2 . 12 . This may enable the cells expressing HRA1 to limit carbon catabolism through fermentation , conserving energy reserves required at the restoration of normoxia . Interestingly , in SUB1A-containing varieties of rice the ability to resume meristem development upon desubmergence is linked to an energy-saving quiescence strategy associated with submergence tolerance [28] . The recognition of a trihelix protein that interacts with SUB1A in rice [23] leads to the question whether the regulation of plant ERF-VIIs may broadly rely on trihelix-dependent attenuation mechanisms similar to the one we described in Arabidopsis . We show that HRA1 acts through a sophisticated mechanism that involves physical interaction with RAP2 . 12 to down-regulate its transactivation capacity and generates a feedback loop of negative self-regulation ( Figure 6 ) . This latter mechanism may make it possible for the cell to start a new pulse of gene expression if hypoxia is prolonged . It is important to highlight that HRA1 interacts with RAP2 . 12 , but apparently not with HRE1 and HRE2 . This is suggestive of a hierarchy in the involvement of ERF-VIIs in the anaerobic response , with the initial burst resulting from the action of RAP2 . 12 and HREs taking over during prolonged hypoxia , in line with the hypoxia susceptibility of hre1hre2 double mutants corresponding to their inability to sustain the expression of the hypoxia-responsive genes [29] . The relative contribution of the different ERF-VIIs requires further exploration and will likely reveal additional layers of complexity of the anaerobic transcriptional response network . The tissue-specific expression of HRA1 unveils the importance of differential modulation of the anaerobic response in rosette leaves of distinct developmental age . HRA1 is predominantly expressed in tissues prone to physiological hypoxia and mutants lacking HRA1 display organ-specific susceptibility to hypoxia ( Figure 3A ) . This implies that the need for a finely-tuned hypoxic response is varied across organs and during development . Fast growing tissues , such as young expanding leaves , required HRA1-dependent dampening of the anaerobic response for survival , and this is likely related to the need to preserve resources for resumption of growth following reoxygenation . Mature leaves , instead , can devote available carbon to fuel fermentation to preserve leaf tissue homeostasis , with less requirement for biosynthetic processes . Carbon will be more rapidly available at reoxygenation by resuming of the photosynthetic activity in source leaves , as compared to younger , sink leaves . Although plants possess a vascular system for transporting nutrients , its ability to transport oxygen relies on anatomical features , such as aerenchyma , that are absent in many instances . Physiological and molecular acclimation to rapidly changing oxygen availability , due to environmental perturbations such as flooding or on a daily basis as a consequence of the light/dark cycle , requires a sophisticated mechanism to fine-tune the anaerobic response . We can conclude that two components of this system in Arabidopsis are the N-end-rule–regulated ERF-VIIs and the trihelix HRA1 . It is well established that genetic variation of ERF-VIIs in rice confer distinct survival strategies and manipulation of these proteins in Arabidopsis can be used to bolster low oxygen and submergence survival . Our evidence of a mechanism regulating the efficacy of the RAP2 . 12-dependent transcriptional regulation provides experimental support for the existence of an elaborated system allowing the plants to respond dynamically to hypoxia . This mechanism is based on the equilibrium between the induction of the anaerobic response by group VII ERFs and repression by HRA1 . Alteration of this equilibrium by misexpression of HRA1 results in lower tolerance to submergence , suggesting that crops with higher tolerance to flooding conditions might be bred through fine-tuning of the relative contribution of ERFs and HRA1 to the overall response to hypoxia . The presence of HRA1 orthologues in crops provides additional opportunity for engineering or breeding varieties with enhanced tolerance to flooding .
A . thaliana Columbia-0 ( Col-0 ) was used as the wild type ecotype . hra1-1 ( N541486; SALK_041486 ) and hra1-2 ( N560275; SALK_060275 ) T-DNA mutants were obtained from the European Arabidopsis Stock Center ( uNASC ) and the Arabidopsis Biological Resource Center , respectively . Mutants were genotyped using standard nonquantitative PCR on genomic DNA , using primers listed in Table S7 . See also Figure S4B for a graphical representation of primer binding sites . Seeds were sown in a moist mixture of soil∶perlite∶sand mixture 3∶1∶1 , stratified at 4°C in the dark for 48 h and germinated at 23°C day/18°C night under a neutral day cycle ( 12 h light/12 h darkness , ∼80 µmol photons m−2 s−1 light intensity ) . Experiments in sterile conditions were performed with 4-d-old seedlings grown in liquid MS medium [0 . 43% ( w/v ) Murashige–Skoog ( MS ) salts ( Sigma-Aldrich ) , 1% ( w/v ) sucrose , pH 5 . 7] under continuous shaking conditions , or with 2- and 3-wk-old plants grown vertically on solid MS medium [liquid MS medium , 0 . 4% ( w/v ) Phytagel ( Sigma–Aldrich ) ] . For the DNA microarray , chromatin immunopurification , and ADH assay experiments , sterilized seeds were grown for 7 d on solid MS medium in vertical orientation in a growth chamber ( Model # CU36L5 , Percival Scientific , Perry , IA ) under a long day cycle ( 16 h light/8 h darkness , ∼80 µmol photons m−2 s−1 ) , at 23°C , before treatments [18] . Hypoxic treatments were performed as described previously [18] . For submergence treatments , 4-wk-old plants ( stage 3 . 50 [30] ) grown in soil as described above were used . Treatments started at ZT ( Zeitgeber Time ) 2 . Plants were submerged with deionized water in glass tanks , until the water surface reached 20 cm above the rosettes , and kept in the dark for the duration of the treatment . Submergence was for 72 h , after which plants were transferred to normal photoperiodic conditions ( 12 h light/12 h darkness ) and allowed to recover for 1 wk before the phenotypic evaluation . Plants that were able to progress in vegetative development were scored as survivors ( Figure 3A ) . The dry weight of whole rosettes was measured before submergence and at the end of the recovery phase . Five separate tanks were used in every submergence experiment , each containing five plants per genotype , and the experiment was repeated three times . Samples for gene transcript abundance and Western blot analyses were , instead , collected after 4 h of submergence . Each sample was composed of young leaves ( youngest three emerging leaves and shoot meristem ) or old leaves ( 10th to 12th leaves ) from five plants . Three biological replicates were used , and the experiment was repeated two times with comparable results . Gene expression data are mean ± s . d . In an independent submergence survival assessment system [16] , [17] , seedlings at the 10-leaf rosette stage ( stage 1 . 10 [30] ) were submerged in complete darkness or held in complete darkness in air for 3 , 5 , 7 , or 10 d ( Figure S6 ) . After desubmergence or re-illumination , the number of plants with alive apical meristem ( green , nonwater-soaked ) was recorded each day for 12 d . The median lethal time ( LT50 ) , standard error , and the 95% confidence interval were determined using the 9-d recovery time point exactly as described previously [17] . Additional abiotic stress treatments were carried out on liquid-grown 4-d-old seedlings . The following conditions were used: 2 h at 4°C ( cold stress ) ; pinching of the seedlings with 10 consecutive pin pricks , 1 h before sampling ( mechanical wounding ) ; 3 h in the presence of 150 mM sodium chloride ( salt stress ) ; 3 h in the presence of 5 mM hydrogen peroxide ( oxidative stress ) ; 3 h in the presence of 100 mM mannitol ( osmotic stress ) ; 3-h-long desiccation under laminar air flux ( dehydration stress ) ; 90 min at 38°C ( heat stress ) . Control plants were maintained at 23°C with continuous shaking . Coding and upstream regulatory sequences were amplified from appropriate Arabidopsis cDNA or genomic DNA templates using Phusion High Fidelity DNA-polymerase ( New England Biolabs ) . Fusion sequences were generated by overlapping PCR . Whenever the GATEWAY cloning system ( Life Technologies ) was exploited , sequences were cloned into pENTR/D-TOPO and the resulting entry vectors were recombined into destination vectors using the LR reaction mix II ( Life Technologies ) . A list of plasmid constructs generated in this study and primers used for cloning can be found in Tables S6 and S7 , respectively . A construct for overexpression of HRA1 in the OE-HRA1#1 and OE-HRA1#2 transgenics , named 35S-HRA1-FLAG , was prepared by cloning the full-length HRA1 cDNA with gwHRA1_5′UTR_Fw and gwHRA1_3′UTR_Rv primers and recombination into the p35S:GATA-HF vector [15] , in which a CaMV 35S promoter drives the expression of HRA1 cDNA linked to a C-terminal FLAG tag [NH2-Gly7-FLAG ( AspTyrLysAsp4Lys ) Gly3-His6-COOH] . A further 35S-HRA1 overexpression construct , used to obtain a third transgenic lacking any C-terminal epitope tag ( OE-HRA1#3 ) , was produced by amplification of HRA1 coding sequence with gwHRA1_Fw and gwHRA1_Rv primers and subsequent cloning in the pK7WG2 vector [31] . The 35S:RAP2 . 12:RrLuc construct exploited for Figure 3E was produced by GATEWAY cloning of a RAP2 . 12:RrLuc DNA sequence into p2GW7; this sequence , in turn , was produced by overlapping PCR after separate amplification of the RAP2 . 12 full CDS , from a cDNA template , and Renilla reniformis luciferase CDS , from the 35S:RrLuc plasmid ( see Table S6 ) . Moreover , the normalization vector 35S:PpLuc was generated by amplification of the firefly luciferase gene from pBGWL7 [31] and GATEWAY cloning into p2GW7 . Finally , the pGWL7 GATEWAY destination vector used for transactivation experiments in plant protoplasts was obtained by cutting an ApaI/SpeI fragment from pBGWL7 and ligating it into the p2GW7 backbone [31] . Stable transgenic Arabidopsis plants were generated by Agrobacterium-mediated transformation following the floral dip method [32] . T0 seeds were screened on the appropriate selection plates , and single-insertion homozygous lines were identified . T3 or later generations of single insertion homozygotes were evaluated . Arabidopsis mesophyll protoplasts were obtained from rosette leaves and transfected according to [33] . In planta protein–protein interactions were investigated via bimolecular fluorescence complementation ( BiFC ) [34] , using the C-terminal split-YFP constructs 35S:HRA1:YFPn and 35S:RAP2 . 1214–358:YFPc [35] . As the negative control for nonspecific YFP complementation , empty 35S:YFPn and 35S:YFPc vectors were co-transfected into protoplasts . For each construct , 10 µg plasmid DNA was used . Fluorescence was observed with a Nikon ViCo microscope using filters for YFP ( excitation wavelengths , 495–510 nm; barrier , 520–550 nm ) , TRITC ( excitation wavelengths , 540–565 nm ) , and DAPI ( excitation wavelengths , 385–400 nm ) . Micrographs are representative of three independent experiments . In promoter transactivation assays performed with protoplasts , 3 µg transformation−1 35S:RrLuc plasmid DNA [36] was used for normalization of the PpLuc activity . Test constructs ( test promoter:PpLuc , 3 µg transformation−1 ) harboring the Photinus pyralis luciferase gene were co-transfected into protoplasts , along with the specified effector plasmid ( s ) encoding TFs ( 35S:effector , up to 6 µg transformation−1 ) . pAVA 393 [37] was used as the 35S:GFP construct , when needed . Samples were subsequently processed with the Dual-Luciferase Reporter Assay System ( Promega ) , and luciferase activity was quantified with a Lumat LB 9507 luminometer ( Bechtold Technologies ) . Each experiment was repeated three times and a representative replicate was shown . Relative luciferase intensity values ( PpLuc/RrLuc ) are presented as mean ± s . d . of four independent transfections . RAP2 . 12:RrLuc protein stability from Figure 3E was assessed likewise through the Dual-Luciferase system . For each individual transfection , 5 µg 35S:RAP2 . 12:RrLuc plasmid DNA was supplemented with increasing amounts of the 35S:HRA1 effector construct and transfected into a mesophyll protoplast suspension . In this case , 35S:PpLuc was used for normalization . Relative Renilla luciferase intensity values ( RrLuc/PpLuc ) are presented as mean ± s . d . of four independent transfections . For HRA1 localization in plant tissue , a 35S:HRA1:GFP-His6-FLAG translational fusion construct ( named 35S-HRA1-GFP ) , was generated by subcloning of a HRA1:GFP-His6 construct , obtained by recombining the full-length HRA1 cDNA in the pEarleyGate103 vector [38] , into the p35S:GATA-HF plasmid [15] . The construct was transformed into Arabidopsis Col-0 to produce transgenic plants accumulating the HRA1:GFP protein . Three-day-old seedlings were vacuum infiltrated with 5 µg ml−1 4′ , 6-diamidino-2-phenylindole ( DAPI ) for 10 min , washed in water for 10 min under vacuum , and observed with a Leica SP2 ( Bannockburn , IL ) confocal microscope at the Microscopy Core Facility , Institute for Integrative Genome Biology , University of California , Riverside . GFP was viewed by excitation at 488 nm and emission at 500–600 nm . DAPI stained nuclei were visualized with a UV laser by excitation at 350 nm and emission at 399–600 nm . Histochemical GUS staining was carried out according to [39] . Briefly , plant material was fixed immediately after sampling in ice-cold 90% acetone for 1 h , rinsed several times in 100 mM phosphate buffer ( pH 7 . 2 ) , and then stained in a freshly prepared reaction solution [0 . 2% Triton X-100 , 2 mM potassium ferrocyanide , 2 mM potassium ferricyanide , and 2 mM X-Gluc ( 5-bromo-4-chloro-3-indolyl ß-D-glucuronide , sodium salt dissolved in DMSO ) in 100 mM phosphate buffer pH 7 . 2] . Plants were stained for 2–4 h ( seedlings ) or overnight ( adult plants ) . Chlorophyll was eliminated from green tissues by washing them with absolute ethanol . RNA extraction , removal of genomic DNA , cDNA synthesis , and RT-qPCR analyses were performed as described previously [36] . Steady-state mRNA levels were normalized using Ubiquitin10 ( At4g05320 ) or β-TUB2 ( At5g62690 ) as reference genes and relative expression values were calculated using the comparative Ct method [40] . The complete list of qPCR primers employed is reported in Table S8 . Multiple qPCR primer couples were designed on HRA1-derived transcripts: Those named “sgHRA1_Endo” and “sgHRA1_Tot” were exploited to measure HRA1 mRNA levels in Figure 5B , while elsewhere “sgHRA1” primers were used . Data ( mean ± s . d . ) are representative of at least two independent experiments , each one carried out with three biological replicates , unless differently stated . Total RNA was extracted from frozen tissue using the RNeasy Plant Kit ( Qiagen , Chatsworth , CA ) and quantified with a NanoDrop 1000 spectrophotometer ( ThermoScientific , Wilmington , DE ) . RNA quality was checked using the Agilent 2100 Bioanalyzer ( Santa Clara , CA ) , and biotin-labeled cRNA was prepared with the GeneChip IVT Labeling Kit ( Affymetrix , Santa Clara , CA ) . Hybridizations against the Arabidopsis ATH1 Genome Array were performed by the Institute for Integrative Genome Biology ( IIGB ) Genomics Core Facility , University of California , Riverside . Transcriptomes of each of the three genotypes—namely , Col-0 , OE-HRA1#1 , and OE-HRA1#2—were profiled under the two conditions . CEL files of two OE lines were processed as biological replicates along with two Col-0 replicates , using R and Bioconductor package . The NCBI Gene Expression Omnibus ( http://www . ncbi . nlm . nih . gov/geo/ ) accession number for the generated dataset is GSE50679 . The microarray dataset generated by [7] was obtained from NCBI GEO ( http://www . ncbi . nlm . nih . gov/geo/ ) accession no . GSE29187 . CEL files of aerobic-treated 5-wk-old rosette tissues of Col-0 and a transgenic overexpressing N-terminally HA-tagged RAP2 . 12 ( RAP ) line were processed together with the HRA1 microarray dataset ( Col-0 , OE-HRA1#1 , and OE-HRA1#2 ) . After computing the absent and present calls using the Affymetrix MAS 5 . 0 algorithm [41] , datasets were normalized using the robust multichip average ( RMA ) method [42] . Mitochondrial and plastid gene probe pair sets were removed , and probe pair sets with present calls in greater than 50% of the samples were used in further analyses . DEGs were identified by comparisons using linear models for microarray data ( LIMMA ) available in the Bioconductor package [43] . A total of 1 , 295 DEGs were selected that satisfied the two following criteria , |SLR|>1 and adj . p<0 . 01 ( SLR , signal log2 ratio; adj . p , false discovery rate adjusted p value ) , in at least one comparison for each Affymetrix probe set . The DEGs were further analyzed using fuzzy k-means clustering with FANNY function . Clustering results were visualized using the Multi Expression Viewer ( MEV ) software ( http://www . tm4 . org/mev/ ) [44] . Gene ontology ( GO ) enrichment was evaluated for each cluster with the GO annotation file of A . thaliana from http://geneontology . org ( downloaded 17 Jan 2012 ) . ChIP-seq libraries were prepared using the protocol of [45] with modifications . Arabidopsis seedlings were grown for 7 days on solid MS medium and hypoxia stressed for 2 hours as described above . Immediately at the termination of treatment , 1 g of plant material was transferred to a 50 ml Falcon tube and fixed in 25 ml MC buffer ( 10 mM sodium phosphate , pH 7 , 50 mM NaCl , 0 . 1 M sucrose ) containing 1% ( w/v ) formaldehyde by incubation on ice for 20 min . The fixation was stopped with the addition of 2 . 5 mL of 1 . 25 M glycine . After three washes with 25 mL MC buffer , the seedlings were frozen , ground and hydrated in 25 mL M1 buffer [10 mM sodium phosphate , pH 7 , 0 . 1 M NaCl , 1 M 2-methyl 2 , 4-pentanediol , 10 mM β-mercaptoethanol and 0 . 5 tablet Complete Protease Inhibitor Cocktail ( Roche Molecular Diagnostics , Pleasanton , CA ) per 25 ml] . The slurry was filtered and centrifuged at 1000 g for 20 min at 4°C to obtain a nuclear pellet that was washed five times with 5 ml of M2 buffer ( 10 mM sodium phosphate , pH 7 , 0 . 1 M NaCl , 10 mM MgCl2 , 1 M 2-methyl 2 , 4-pentanediol , 10 mM β-mercaptoethanol and 0 . 5% ( v/v ) Triton X-100 , 0 . 5 tablet Complete Protease Inhibitor Cocktail per 25 ml ) . The final wash was performed with 5 ml M3 buffer ( 10 mM sodium phosphate , pH 7 , 0 . 1 M NaCl , 10 mM β-mercaptoethanol and 0 . 5 tablet of Complete Protease Inhibitor Cocktail per 25 ml ) . The nuclear pellet was resuspended in 1 ml sonication buffer ( 10 mM sodium phosphate , pH 7 , 0 . 1 M NaCl , 0 . 5% Sarkosyl , 10 mM EDTA ) and sonicated using a Bioruptor UCD-200 ( Denville , NJ ) on ice , following the manufacturer's instruction . The sample was centrifuged twice at 15600 g for 10 min at 4°C and the supernatant was used for chromatin immunoprecipitation ( ChIP ) with 20 µl of EZview Red ANTI-FLAG® M2 Affinity Gel ( Sigma-Aldrich , St . Louis , MO ) following the manufacturer's instruction with IP buffer ( 50 mM HEPES , pH 7 . 5 , 150 mM NaCl , 5 mM MgCl2 , 10 µM ZnSO4 , 1% ( v/v ) Triton X-100 , 0 . 05% ( w/v ) SDS ) . After the samples were reverse-crosslinked , , the DNA was purified using the Qiagen PCR purification kit . Library construction involving end repair , A-tailing , and ligation to an adapter was conducted using the End-It DNA End-Repair Kit ( Epicentre , Madison , WI ) , Klenow fragment ( New England Biolabs , Ipswich , MA ) and Fast-Link DNA Ligation Kit ( Epicentre ) . Two custom barcodes of four nucleotides were used for multiplexing ( Col-0 [5′-GTAT-3′] and OE-HRA1#1 [5′-ACGT-3′] ) . Samples were submitted to the IIGB Genomics Core Facility , University of California , Riverside , for single-end sequencing with 100 cycles using the Illumina Hiseq2000 platform . Raw data in fastq file format were imported into R using the ShortRead package [46] , the barcode sequence was removed from the 100 bp read sequences and reads were aligned to the A . thaliana genome TAIR10 version ( http://www . arabidopsis . org ) using Bowtie ( ver . 0 . 12 . 7 ) [47] , allowing two nucleotide mismatches . Peaks were generated with the Model-based Analysis for ChIP-Seq ( MACS ) software [48] using Col-0 mock ChIP data as the control with default settings . ChIPpeakAnno was used to acquire gene annotation , determine location of peak regions from the nearest genes , and obtain DNA sequences of peak regions [49] . Peaks of ChIP-seq data were visualized using the Integrative Genomics Viewer software ( 2 . 1 ) [50] . Soluble protein samples from total tissue extracts were separated by SDS-PAGE on 10% polyacrylamide Bis-Tris NuPAGE midigels ( Life Technologies ) and then transferred onto a polyvinylidene difluoride membrane by means of the Trans-Blot Turbo System ( Bio-Rad ) . Detection of the HRP-conjugated secondary antibody ( goat anti-rabbit IgG , Agrisera , product code AS09 602 ) was performed with the LiteAblot Turbo Extra-Sensitive Chemiluminescent Substrate ( EuroClone ) . Antibodies against PDC ( product code AS10 691 ) and ADH ( product code AS10 685 ) were purchased from Agrisera and antisera against FLAG ( A8592 ) from Sigma-Aldrich . Equal loading of total protein samples was checked by amido black staining , as described in [19] . ADH-specific activity was measured as described previously [51] with minor modifications , using Arabidopsis 7-d-old seedlings . Soluble carbohydrates analyzed in Figure S7 were extracted from whole rosettes using perchloric acid and analyzed enzymatically in the neutralized supernatant , as described previously by [52] . The ProQuest Two-Hybrid System ( Life Technologies ) was used . Saccharomyces cerevisiae strain MaV203 was transformed with the different combinations of bait ( obtained after recombination of the inserts into pDEST32 ) , prey ( obtained after recombination of the inserts into pDEST22 ) , and control vectors . Empty pDEST32 and pDEST22 were used as negative controls . Yeast transformation was performed according to the LiAc/SS carrier DNA/PEG method [53] . After transformation , yeast containing both vectors was grown for 3 d at 28°C on minimal selective dropout medium lacking Leu and Trp ( SC-LW medium ) to select colonies containing two vectors . Plating was then replicated on selective dropout medium ( SC-LWH+3AT medium ) lacking Leu , Trp , and His , supplemented with 10 mM 3-aminotriazole ( 3AT ) , in order to select colonies containing interacting partners . The strength of the interaction was further verified by β-galactosidase staining ( LacZ ) following the manufacturer's instructions . Significant variations between genotypes or treatments were evaluated statistically by Student's t test or one-way ANOVA , coupled with Tukey's posttest , for general comparisons , or Dunnet's posttest , for multiple comparisons with a reference sample , where appropriate . Mean values that were significantly different ( p<0 . 05 ) from each other are marked with lower case letters or asterisks inside the figures . The statistical evaluation of the submergence survival , DNA microarray , and ChIP-seq experiments is described under the respective subsections . Full protein coding sequences of plant trihelix proteins were obtained from GenBank and aligned using ClustalW [54] . The maximum likelihood algorithm in the MEGA5 . 0 framework [55] was used with 500 bootstrap replicates to evaluate evolutionary relatedness . To identify trihelix genes positively regulated by low oxygen conditions across plant species ( Table S1 ) , existing transcriptomic data were surveyed . Genes belonging to the trihelix family in each of the organisms taken into consideration were extracted from the PlantTFDB [56] and used to query the selected public microarray datasets . | Respiratory metabolism in land plants requires oxygen availability to be able to generate ATP , which is essential for biosynthetic processes . Cellular hypoxia can be triggered as a consequence of environmental events ( mainly floods ) , anatomical constraints ( low tissue permeability to gases ) , or elevated cellular respiration , and it is unfavorable to growth due to the resultant decline in ATP . The adaptation of plants to fluctuating oxygen levels inside tissues requires the dynamic regulation of mechanisms that ensure cell viability and ultimately organism survival , but only a few molecular components of this homeostatic network are known . Direct hypoxia-sensing entails the posttranslational stabilization of a subgroup of plant ethylene-responsive factor ( ERF ) transcription factors , which coordinate the expression of hypoxia-inducible genes . Turnover of these ERFs is determined by an oxygen-dependent pathway of proteasomal degradation . Here , we demonstrate that the hypoxia-inducible transcription factor gene HRA1 is transcriptionally activated upon ERF-VII RAP2 . 12 stabilization and encodes a trihelix DNA binding protein that functionally interacts with RAP2 . 12 to curtail its activity . In addition to its negative regulation of RAP2 . 12 , HRA1 negatively regulates activation of its own promoter . This RAP2 . 12-HRA1 control unit allows plants to modulate the extent of the response to hypoxia , including anaerobic enzyme production , to levels that improve endurance of the stress . Our results emphasize the importance of a strategy that can counterbalance energy-inefficient survival responses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biotechnology",
"cellular",
"stress",
"responses",
"plant",
"growth",
"and",
"development",
"cell",
"processes",
"brassica",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"genetically",
"modified",
"plants",
"plants",
"arabidopsis",
"thaliana",
... | 2014 | A Trihelix DNA Binding Protein Counterbalances Hypoxia-Responsive Transcriptional Activation in Arabidopsis |
Hox genes that determine anteroposterior body axis formation in all bilaterians are often found to have partially overlapping expression pattern . Since posterior genes dominate over anterior Hox genes in the region of co-expression , the anterior Hox genes are thought to have no function in such regions . In this study we show that two Hox genes have distinct and essential functions in the same cell . In Drosophila , the three Hox genes of the bithorax complex , Ubx , abd-A and Abd-B , show coexpression during embryonic development . Here , we show that in early pupal abdominal epithelia , Ubx does not coexpress with abd-A and Abd-B , while abd-A and Abd-B continue to coexpress in the same nuclei . The abd-A and Abd-B are expressed in both histoblast nest cells and larval epithelial cells of early pupal abdominal epithelia . Further functional studies demonstrate that abd-A is required in histoblast nest cells for their proliferation and suppression of Ubx to prevent first abdominal segment like features in posterior segments while in larval epithelial cells it is required for their elimination . We also observed that these functions of abd-A are required in its exclusive as well as the coexpression domain with that of Abd-B . The expression of Abd-B is required in histoblast nest cells for their identity while it is dispensable in the larval epithelial cells . The higher level of Abd-B in the seventh abdominal segment , that down-regulates abd-A expression , leads this segment to be absent in males or of smaller size in females . We also show that abd-A in histoblast nest cells positively regulates expression of wingless for the formation of the abdominal epithelia . Our study reveals an exception to the rule of posterior prevalence and shows that two different Hox genes have distinct functions in the same cell , which is essential for the development of abdominal epithelia .
Anteroposterior ( AP ) body axis in all the bilaterians is determined by a set of homeobox ( Hox ) containing genes , the Hox genes [1] . The eight Hox genes in Drosophila are arranged in two clusters , the Antennapedia complex ( ANT-C ) and the bithorax complex ( BX-C ) [2]–[4] . BX-C has three Hox genes Ultrabithorax ( Ubx ) , abdominal-A ( abd-A ) and Abdominal-B ( Abd-B ) [3] , [5] . During development , Hox genes express in a collinear manner where the order of genes in the genomic locus is similar to their spatial expression pattern along the AP axis of the embryo [6]–[9] . This expression pattern of Hox genes determines the identity of the body segments along the AP body axis [10] , [11] . In several instances , expression of Hox genes is found to be overlapping [9] . In such cases , posteriorly expressed Hox genes are known to suppress function of anterior genes at transcriptional or post-translational level of gene regulation , a phenomenon known as posterior dominance [10] , [12]–[15] . Therefore , the function of the anterior Hox gene in such regions of overlapping expression with the posterior Hox gene is thought to be irrelevant . This is supported by the observation that mutants for anterior Hox genes do not show distinct phenotypes in the region of overlapping expression up to early larval stage and die later during development . Owing to this , the role of anterior Hox genes in the region where they co-express with posterior ones has not been investigated extensively during larval and pupal stages . A few studies , however , suggest that anterior Hox genes can have non-homeotic functions in the region of co-expression with posterior Hox genes [16] , [17] . In order to investigate the role of three Hox genes Ubx , abd-A and Abd-B in their non-overlapping and overlapping domains of expression , we analyzed their expression pattern in the early pupal abdominal epithelial cells . Each abdominal segment at larval stage has two cell types the polytenized larval epithelial cells ( LECs ) and diploid histoblast nest cells ( HNCs ) . The HNCs are maintained in a quiescent state during the larval development and proliferate during the early pupal development to differentiate into adult abdominal epithelial cells [18] , [19] . During this process , HNCs induce apoptosis in the LECs and replace them with a layer of abdominal epithelium . For consistency we have used early pupal abdominal epithelia of 0 to 32 h after puparium formation ( APF ) having both LECs and HNCs . The abdominal epithelia formed after 32 h APF by proliferation and differentiation of HNCs and removal of LECs are termed as pupal abdominal epithelia . These pupal abdominal epithelial cells further develop into adult abdominal epithelia with features like bristles and pigmentation . We found that Ubx is expressed only in HNCs and LECs of first segment of early pupal abdominal epithelia and does not overlap with abd-A or Abd-B . On the contrary , abd-A co-expresses with Abd-B in HNCs and LECs . Here we show that in the coexpressing HNCs , abd-A is required for formation of adult abdominal epithelia while Abd-B is required for its identity . We also observed that the higher expression of Abd-B in abdominal segment 7 suppresses expression of abd-A that leads to smaller segment in females and complete elimination in males . These findings , for the first time , show that abd-A is required for abdominal epithelia formation not only in its exclusive expression domain but also in the region where it overlaps with Abd-B .
Three Hox genes of the BX-C in Drosophila melanogaster determine the identity of third thoracic and all abdominal segments [9] , [20] . The identity of third thoracic ( T3 ) and first abdominal segment ( A1 ) is determined by Ubx , A2 to A4 by abd-A and A5 to A9 by Abd-B [9] . Although these Hox genes regulate the identity of specific segments in adults , their expression is not restricted only to the corresponding parasegments ( PSs ) in embryos [7] , [21] , [22] . The Ubx gene is expressed from PS5 to PS12 , abd-A expresses from PS7 to PS12 while Abd-B from PS10 to PS14 in embryos [6] , [7] , [23] . This results in the overlap of Ubx with abd-A in PS7 to PS9 , and all three Hox genes in PS10 to PS12 ( Figure 1A , Figure S1A and B ) [8] , [21] . This overlapping expression is seen not only in the same parasegment but also in the same cells ( Figure S1A and B ) , raising the question of why an anterior Hox gene should express in the domain of a posterior Hox gene if it has no apparent functions there . We further explored if the overlapping expression pattern seen during embryonic development also persists in LECs and HNCs of early pupal abdominal epithelia . In immunostaining experiments , we observed the expression of Ubx in T3 and A1 segment , Abd-A from A2–A7 and Abd-B from A5–A7 ( Figure 1B , C and D respectively ) , which is similar to earlier observations [24] , [25] . The expression of all the three genes is seen in bigger polytenised LECs ( yellow arrowheads ) and smaller diploid HNCs ( white arrowheads ) . In contrast to the embryonic expression pattern , Ubx is observed only in third thoracic ( T3 ) and first abdominal segment ( A1 ) and not in any of the posterior segments ( Figure 1B and C ) . This suggests that Ubx expression does not overlap with the other two genes while overlapping expression of Abd-A and Abd-B is seen from A5 to A7 ( Figure 1D ) . A closer look reveals that the expression of abd-A is not uniform and shows very weak expression in A2 and A7 in comparison to other segments ( Figure 1D ) . The expression of Abd-B continues to show a lower to higher gradient from A5 to A7 as seen in embryonic CNS ( Figure 1A and D ) . Furthermore , we observed that from A5 to A7 the Abd-A and Abd-B expression coexists not only in same segment but also in the same nuclei ( Figure 1E ) . These observations establish that , unlike the embryonic expression pattern , at pupal stage the expression of Ubx does not overlap with abd-A or Abd-B while the overlap between abd-A and Abd-B persists in LECs and HNCs of early pupal abdominal epithelia ( Figure 1F ) . To assess the role of abd-A and Abd-B in both overlapping and non-overlapping expression domains , we chose to knock down abd-A and Abd-B in HNCs and LECs using UAS-RNAi and Gal4 approach . We used esg-Gal4 and Eip-71CD-Gal4 for exclusive knockdown in HNCs and LECs , respectively and 71B-Gal4 and Pnr-Gal4 for knockdown in the both cell types [26] , [27] ( Figure 2 A1-4 ) . The esg-Gal4 is known to express in all histoblast nest cells in early pupa ( before 24 h ) , which fades away later in development ( Figure 2 A1 ) [19] . We found that the expression of 71B-Gal4 is seen mainly in larval epithelial cells till 18h APF but later it also starts expressing in HNCs at low levels and increases with time . Interestingly , its expression in HNCs is stochastic and not uniform as seen in esg-Gal4 . ( Figure 2 A3 and Figure S2 ) . The expression of Pnr-Gal4 is not seen in early pupa and starts expressing in LECs after 14 h APF , which later increases and is seen in all LECs and in few rows of HNCs that are leading towards dorsal midline ( Figure 2 A4 and Figure S3 ) . Knockdown of abd-A in HNCs using esg-Gal4 driver shows lethality at larval stages and only ∼20% larvae pupated although none of them hatch . These pupae showed complete loss of abdominal epithelia in the expression domain of abd-A ( A2–A7 ) but not in A1 ( Figure 2 B2 ) suggesting this to be an abd-A specific phenotype . The loss of epithelia was observed from both dorsal and ventral sides of the segment and no tergite or sternite was seen in these segments ( Figure 2 B2 and Figure S4 ) . This brings out the critical role of abd-A in HNCs for development of adult abdominal epithelia at pupal stage of development . The knockdown of abd-A in LECs using Eip-71CD-Gal4 shows complete lethality at very early stage ( before 24 h ) of pupal development . These pupae did not even grow enough to show any epithelia formation ( Figure 2 B3 ) . The simultaneous knocking down of abd-A in both LECs and HNCs using 71B-Gal4 also shows developmental arrest at pupal stage leading to lethality . Manually eclosed pharates show loss of abdominal epithelia similar to what we see with esg-Gal4 ( Figure 2 B4 ) . Similarly , the abd-A knockdown using Pnr-Gal4 also showed lethality at pupal stage but manually eclosed pupae showed dorsal closure defect of abdominal epithelia ( DDA ) ( Figure 2 B5: marked by dotted line ) . In this case we observed loss of epithelia only close to dorsal mid line , which corresponds to the expression pattern of this Gal4 driver in LECs and HNCs . We also generated milder abd-A RNAi lines ( see materials and methods ) and used them to knockdown abd-A in HNCs and LECs . As expected , these lines gave less severity and penetrance of the phenotypes . The phenotype of one of the lines is shown in figure 2 B2′-5′ . Knockdown of abd-A in HNCs using this line with esg-Gal4 shows less lethality at pupal stage and adult flies show partial loss of abdominal epithelia in adults , which is always restricted to A2–6 ( Figure 2 B2′ ) . The adult abdominal epithelium of the knockdown flies also show lesser and smaller bristles in comparison to the wild type flies ( Figure 2 B2′ and 2B1′ respectively ) . This indicates the homeotic transformation of the epithelia of posterior segments into A1 segment that has smaller bristles . This surprised us because in the abd-A knockdown the features of posterior segments transform into features similar to Ubx expression domain A1 , although we did not observe Ubx expression in HNCs of these segments . We reasoned that the knockdown of abd-A might be leading to the derepression of Ubx in posterior segments thus showing this phenotype . To test this , we did immunostaining of Ubx in abd-A RNAi background . We observed the expression of Ubx only in HNCs of posterior segments ( Figure S5 ) , which was not seen in wild type epithelia ( Figure 1B ) , suggesting that in wild type epithelia abd-A suppresses expression of Ubx in its expression domain . Furthermore , knockdown of abd-A using Eip-71CD-Gal4 and 71B-Gal4 with milder abd-A RNAi line also showed viability and DDA phenotypes in adults ( Figure 2 B3′–4′ ) . The DDA phenotype is always restricted to A2 to A6 segment . Knockdown of abd-A by Pnr-Gal4 shows loss of pigmentation across the dorsal midline , suggesting that the expression of abd-A is also required for pigmentation of adult cuticle ( Figure 2B5′ ) . The phenotype of these abd-A RNAi experiments shows that the knockdown of abd-A in HNCs leads to a loss of epithelial cells while that in LECs causes DDA phenotype . This indicates that the level of abd-A is critical in HNCs and LECs during development of adult abdominal epithelia at pupal stage . Interestingly this role of abd-A in epithelia formation is not limited only in its exclusive expression domain but also in the domain where it overlaps with Abd-B ( A5 and A6 segments ) . We further assessed the role of abd-A in HNCs using live cell imaging . The HNCs are quiescent during larval stage and get signal to proliferate as soon as pupation starts . The proliferation of HNCs is biphasic , where in the first phase ( 4–12 hours APF ) , cells divide without growing much in size while in the second phase ( 15–36 hours APF ) cells divide normally and make complete epithelia [19] . In the abd-A knockdown using esg-Gal4 the live cell imaging of HNCs showed almost 50% reduction in the proliferation of HNCs during first six hours of the first phase of proliferation ( Figure 3A ) . There are 16 HNCs at 0 h APF , after 6 h of proliferation their number become 60±5 HNCs in the case of esg-Gal4 while it is only 31±5 in the case of abd-A RNAi . We also observed that about 30% of the analyzed HNCs show complete arrest at this stage and do not proliferate . Further analysis of LECs and HNCs in these pupae also showed that non-proliferating HNCs are unable to eliminate LECs , which results in the arrest in development of abdominal epithelia ( Figure 3B and C ) . This clearly shows that abd-A is required in HNCs for their proliferation during early pupal development . To further analyze the role of abd-A in the HNCs during the development of adult epithelia , we made mitotic clones of abd-A using its loss of function allele , abd-AD24 [28] . Mitotic clones were observed in all segments from A2 to A6 implying a common role of abd-A in all these segments ( Figure 3D ) . Mitotic clones are seen as either small white patches on the tergites without bristles ( Figure 3E , red dotted line ) or relatively larger light pigmented regions with small bristles ( Figure 3E , white dotted line ) . One of such white clones shows bristles only at the places where pigmentation is low while white areas are devoid of bristles ( Figure 3F ) . In the middle of the clone we also see a pigmented patch ( marked with green dotted line ) with stubble bristle representing an abd-A positive region . The small bristles in the mitotic clones suggest that these cells are transformed into A1 like cells as seen on knocking down of abd-A in HNCs ( Figure 2B2′ ) . We further extended our analysis and gave heat shock at 0 h , 12 h , 24 h and 36 h APF to evaluate the role of abd-A in HNCs during pupal development . We observed that mitotic clones show phenotypes only in the flies that were given heat shock at 0 h , 12 h APF and not later stages . All pupae ( 40 out of 40 ) that were given heat shock at 0 h APF show phenotype in mitotic clones of adult flies while only 15% of ( 6 out of 40 ) pupae that were given heat shock at 12 h APF show phenotype in mitotic clones . These results confirmed the abd-A RNAi result that abd-A is required in HNCs for identity and proliferation during the first phase of HNC proliferation . These observations also suggest that expression of abd-A is required for normal size bristles not only in its exclusive expression domain but also in the expression domains ( A5 and A6 ) where it overlaps with Abd-B . We investigated the role of abd-A in LECs by knocking down abd-A using Eip-71CD- Gal4 . Since the original strong abd-A RNAi line gave early pupal lethality , we used a milder RNAi line . In this case also , we observed LEC specific reduction in abd-A expression ( Figure 4A ) . These LEC nuclei do not show the distinct apoptotic feature of irregular nuclear morphology , unlike what is seen in wild type nuclei ( Figure 4A ) . The development of these pupae was arrested during early pupal stage and we see that HNCs do not grow and LECs are not removed as compared to wild type ( Figure 4A and B ) . This suggests that the loss of abd-A prevents cell death in LECs and , thereby , leaves no space for HNCs to proliferate and make complete abdominal epithelia . To confirm if the persistence of LECs during early pupal development leads to DDA phenotype as seen in abd-A knockdown we over-expressed anti-apoptotic factor P35 in LECs using Eip-71CD Gal4 . These flies show DDA phenotype in all the abdominal segments ( Figure 4C ) similar to what is seen in abd-A RNAi using Eip-71CD- Gal4 ( Figure 2B3′ ) . This demonstrates that inefficient clearing of LECs prevents HNCs from growing and leads to fusion of abdominal epithelia at dorsal midline resulting in DDA phenotype . Finally , to validate the abd-A RNAi results we also carried out loss of abd-A function by other independent ways . As miR-iab-8-5p is known to knock down Antp , Ubx and abd-A [29] , [30] , we overexpressed this miRNA using Eip-71CD-Gal4 . Here we observed pupal lethality with few flies emerging with DDA phenotype , similar to what is seen in corresponding abd-A RNAi flies ( Figure 4D and 2 B3′ , respectively ) . In the control experiment with miR-iab-4-5p over expression , which knocks down only Antp and Ubx but not abd-A , this phenotype is not observed indicating that DDA phenotype seen in miR-iab-8-5p overexpression is specifically due to abd-A knockdown ( Figure 4E ) . Similarly , we also observed the DDA phenotype in a cis-regulatory mutant , McpH27 Fab71 , with 100% penetrance ( Figure 4F ) . This mutant has deletion of boundary and PRE combination from Mcp and Fab7 regions , which regulates Abd-B gene in segment specific manner [31] , [32] . We hypothesize that deletion of these cis-regulatory elements leads to ectopic expression of Abd-B in LECs of anterior segment causing suppression of the anterior gene abd-A that results into DDA phenotype . Immunostaining for Abd-B protein in McpH27 Fab-71 mutant indeed shows ectopic and enhanced expression of Abd-B in LECs of A4 and A5 segments ( Figure 4I ) . Further , we knocked down Abd-B using Eip-71CD-Gal4 and observed complete rescue of the DDA phenotype in almost 55% of the flies , while rest of them showed partial rescue confirming that the ectopic expression of Abd-B causes DDA phenotype ( Figure 4G ) . To further establish Abd-B dependent loss of abdominal epithelia , we over expressed Abd-B in LECs using Eip-71CD-Gal4 and observed 100% lethality at pupal stage . Most of the pupae died at early stages but few ( 15% ) of them survived till later stages and showed the anticipated loss of abdominal epithelia ( Figure 4H ) . This partial to complete loss of abdominal epithelia by Abd-B over expression is similar to what we observed in the case of abd-A RNAi ( Figure 2 B2 and 4 ) . These results confirm our abd-A knockdown results and establish that abd-A is required in LECs for their removal during abdominal epithelia development . Taken together , these observations suggest that abd-A plays dual role during development of abdominal epithelia , on the one hand it is required for proliferation of HNCs and on the other hand it is required for removal of LECs . We also did cell type specific knockdown of Abd-B to understand its role in abdominal epithelia formation . To analyze the cell type specific knockdown of Abd-B in HNCs and LECs we did immunostaining against Abd-B protein in early pupal abdominal epithelia in wild type and knockdown background . We observed HNC specific loss of Abd-B in the case of knockdown using esg-Gal4 and LEC specific loss of Abd-B with Eip-71CD-Gal4 ( Figure 5A ) . Knockdown of Abd-B in HNCs of all the abdominal segments using esg-Gal4 driver leads to anteriorization of A5-A7 segments ( Figure 5B ) while adult epithelia formation is unaffected . The loss of pigmentation in A5 indicates its transformation to A4 while appearance of bristles in sternites of A6 is indicative of A6 to A5 transformation . We also see the homeotic transformation of A7 into A5 that is evident from its appearance as a discrete segment , which is otherwise absent in males [33] , [34] , and bristles in the sternite ( Figure 5B ) . The transformation of these posterior segments into anterior is a typical Abd-B loss of function phenotype . We also observed defective genital and anal organs in these knockdown flies , which is in line with the known role of Abd-B in genital and anal development [35] . On the other hand , knocking down of Abd-B in LECs using Eip-71CD-Gal4 did not show any detectable phenotype ( Figure 5B ) . Knock down of Abd-B in both HNCs and LECs driven by 71B-Gal4 and Pnr-Gal4 , leads to anteriorization phenotypes in A5 , A6 and A7 segments ( Figure S6B and C ) . The Abd-B knockdown using 71B-Gal4 shows anteriorization phenotypes similar to esg-Gal4 , however , it is milder than the esg-Gal4 ( Figure S6B ) . This can be attributed to the lower , stochastic and late expression of 71B-Gal4 as compared to esg-Gal4 ( Figure 2A1 and 3 and Figure S2 ) . The Abd-B RNAi using Pnr-Gal4 shows complete loss of pigmentation on both sides of the dorsal midline in A5 to A6 segments indicating transformation of these cells into A4 like identity ( Figure S6C ) . This phenotype is seen only in cells close to dorsal mid line , which very well corresponds to expression pattern of Pnr-Gal4 in the HNCs ( Figure S3 ) . These homeotic phenotypes seen in the case of Abd-B knockdown by esg , 71B and Pnr Gal4 drivers establish that expression of Abd-B in HNCs determines the identity of abdominal epithelia . The knockdown of Abd-B in LECs by using Eip-71CD-Gal4 does not give any phenotype ( Figure 5B ) indicating that expression of Abd-B in these cells is dispensable . From the results of abd-A and Abd-B RNAi , we conclude that the expression of abd-A is critical in both HNCs and LECs for development of the complete epithelia while expression of Abd-B is required in HNCs for their identity but dispensable in LECs . Recent studies show that higher level of Abd-B expression in A7 segment leads to loss of HNCs and LECs that results to elimination or smaller size of A7 segment in males and females , respectively [33] , [34] . While analyzing expression of abd-A and Abd-B in abdominal epithelia , we observed that in the A7 segment expression of Abd-B is maximum and abd-A expression is very weak as compared to anterior segments . ( Figure 1D ) . Thus we hypothesized that Abd-B being a posterior gene may be suppressing abd-A in A7 leading to the removal of A7 segment from the abdomen of males . To understand the Abd-B mediated suppression of abd-A , we analyzed the expression pattern of Abd-A in Fab-71 mutant and Abd-B RNAi driven by esg-Gal4 . The Fab-71 mutant has a deletion of boundary element which leads to higher expression of Abd-B in A6 segment resulting in A6 to A7 transformation and thus loss of both the segments [36] . In wild type fly abd-A expresses weakly in A7 as compared to A5 and A6 ( Figure 6A top panel ) , while in Fab-71 mutant we observed very weak expression of abd-A in both A6 and A7 , suggesting that higher levels of Abd-B in A6 suppresses expression of abd-A in this segment ( Figure 6A middle panel ) . To further confirm this observation , we knocked down Abd-B in HNCs of A7 using esg-Gal4 and observed increased expression of abd-A in this segment ( Figure 6A lower most panel ) . The gain of abd-A expression in A7 is specific to HNCs and it was not seen in LECs , confirming Abd-B dependent suppression of abd-A . This establishes that expression of Abd-B similar to that in A7 segment , suppresses abd-A expression while lower expression in A5 and A6 does not effect abd-A expression . These experiments also clearly show that loss of abd-A expression correlates with the loss of abdominal segment and gain of abd-A correlates with the gain of abdominal segment in adults ( Figure 6B ) . To test if abd-A is required and is enough for abdominal segment formation , we overexpressed abd-A in A7 segment using Abd-B-Gal4 and observed formation of an extra segment in males ( Figure 7A lower panel ) and a bigger A7 segment in females with 100% penetrance ( Figure S7 ) . This establishes that abd-A expression is required and is enough for adult abdominal epithelia formation during pupal development . And in A7 , higher expression of Abd-B suppresses abd-A that leads to the loss of this segment in males and smaller segment in females . We further extended our study to understand how abd-A expression helps development of abdominal epithelia . Earlier studies have shown the role of Wingless ( Wg ) morphogen in regulating development of abdominal epithelia [37]–[39] . In A7 of male the expression of wg is known to be suppressed by higher Abd-B expression [34] . In this study we also observed that higher Abd-B levels suppress abd-A expression in HNCs of male A7 to promote elimination of this segment . This suggests that higher expression of Abd-B suppress both abd-A and wg expression in A7 for its elimination from male abdomen . We further wanted to understand if suppression of wg expression by Abd-B in A7 is through abd-A or independent of it . To study this , we analyzed the expression of wg in male A7 in the background of abd-A overexpression using Abd-B Gal4 . Consistent with earlier expression we do not detect Wg morphogen in HNCs of male A7 ( Figure 7B upper panel ) . The ectopic expression of abd-A leads to ectopic expression of wg in male A7 HNCs , clearly indicating that wg expression is positively regulated by abd-A ( Figure 7B lower panel ) . This shows that wg expression in abdominal epithelial cells is activated by abd-A expression . To further prove that wg pathway is required for the formation of extra A7 segment we also did over expression of wg in A7 segment using Abd-B Gal4 . We observed an extra A7 ( Figure 7C ) in all flies expressing wg under Abd-B-Gal4 similar to what we saw in over-expression of abd-A . This establishes that Abd-A protein activates wg expression for formation of abdominal epithelia . The higher expression of Abd-B in A7 of male induces elimination this segment by suppressing wg expression through suppression of abd-A expression ( Figure 7D ) .
The expression pattern of anterior Hox genes is often found to be overlapping with the posterior genes although the functional importance of such an expression pattern is unknown . In this study we show that the overlapping expression of Hox genes Ubx and abd-A during embryonic development of Drosophila becomes spatially non-overlapping in abdominal epithelia of pupa . The expression of Ubx is seen in T3 and A1 and not in posterior segments while the expression of abd-A is seen from A2 to A7 , which overlaps with that of Abd-B from A5 to A7 . Using UAS/GAL4 based RNAi and FRT/FLP based genetic mosaic techniques we show that abd-A is required in HNCs of A2 to A6 segment for their proliferation and suppression of Ubx expression to provide bigger size of the bristles . This suggests that abd-A is required in HNCs of A2 to A6 segments for development of adult epithelia with bigger bristle size . In contrast to this , in LECs abd-A expression is required for apoptosis , which allow proliferation of HNCs . Here , the interesting point is that these roles of abd-A in abdominal epithelia development are not limited only in the segments of exclusive expression ( A2 to A4 ) but also in the segments where its expression overlaps with that of Abd-B ( A5 to A6 ) . We further show that Abd-B expression is seen in both HNCs and LECs of A5 to A7 and functional studies suggested that it determines the identity of HNCs but it seems to be dispensable in the LECs . Knockdown of Abd-B in the HNCs of A5 to A7 segment of male leads to loss of pigmentation in A5 segment , bristles in sternite of A6 segment and formation of an extra A7 with bristles . This means that loss of Abd-B in HNCs of A5 leads to its transformation into A4 like features , transformation of A6 and A7 into A5 . Loss of function results of abd-A and Abd-B RNAi bring out the function of both the genes in the HNCs of A5 and A6 during the development of adult abdominal epithelia . The two genes not only function together in same segment but also in same nuclei for their distinct roles . This is in contrast to the posterior dominance rule where an anterior gene does not function in the presence of a posterior Hox gene . Interestingly , the situation in A7 is just the opposite , where a higher level of Abd-B suppresses abd-A expression . This suggests that abd-A can coexpress with Abd-B in the regions where Abd-B expression is less than that of A7 . The over expression of abd-A in A7 shows an additional abdominal segment in males and bigger A7 segment in female , thus finally proving that Abd-A protein is required and sufficient for adult abdominal epithelia formation . This also implies that the down-regulation of abd-A in A7 by Abd-B protein is required only for the reduction in the size of abdominal segment . Earlier studies have shown additional male specific roles of Abd-B protein and sex-determination regulator Doublesex that regulate the elimination of A7 in males [33] , [34] . The suppression of abd-A in A7 is seen as loss of Abd-A protein suggesting that posterior prevalence may be operating at transcriptional or post transcriptional level and not at post-translational level where concentration dependent interaction with common cofactor ( s ) was implicated in deciding the dominant function of posterior Hox gene [15] . Our study , with other published data , suggests that the expression of an anterior Hox gene in the domain of posterior Hox gene is not futile and that it has adopted new roles to play in such regions [16] , [17] . This study and earlier work , taken together , explain how Hox genes of the bithorax complex control development of A1 to A7 segments of adult abdominal epithelia . The expression of Ubx , abd-A and Abd-B is required for the identity of A1 , A2–A4 and A5–A7 , respectively ( Figure 7D ) . We show that the expression of abd-A is also required in A2 to A6 for the proliferation of HNCs and elimination of LECs . In A7 , however , higher levels of Abd-B suppress epithelia formation by down regulating expression of abd-A . Interestingly , smaller size of A2 compared to A3 or A4 also correlates with the relatively lower level of expression of abd-A in A2 as compared to that in A3 or A4 . This raises the possibility of the level of Abd-A determining the sizes of segments in adult fly . While the role of Hox genes in determining the identity of body segments is well established , our findings bring into light the collective role of Hox genes in determining the size , shape and identity of body segments . Our observations are in agreement with a recent study which shows that in embryonic CNS of Drosophila the expression of abd-A is not suppressed by Abd-B and that the two genes coexpress in same nuclei [40] . These observations suggest that posterior dominance rule operating between abd-A and Abd-B is tissues specific and not a universal phenomenon . Further studies will be required to understand the likely functional significance of such coexpression patterns of Hox genes in other tissues and animals as well . Much of what we understand about function of Hox genes is in the context of positional identity along the AP body axis . Our study suggests the need to explore the Hox code to understand development of organs where rules like posterior prevalence may not hold .
Flies were grown in standard cornmeal yeast extract medium at 25°C unless otherwise specified . We used following stocks during this study: CantonS ( CS ) as wild type strain , esg-Gal4 UAS-GFP ( Nobert Perrimon ) , Abd-B-RNAi ( Vienna Drosophila RNAi Centre ) , abd-A-RNAi , UAS-abd-A , UAS-Abd-B ( Yacine Graba ) , McpH27 Fab-71 , Fab-71 , abd-AD24 , Pnr-Gal4 ( Francois Karch ) , UAS-miR-iab8-5p ( Eric Lai ) and Abd-B-Gal4LDN , Abd-B-Gal4199 ( Ernesto Sánchez-Herrero ) , UAS-wg and UAS-Abd-B ( LS Sashidhara ) and w hsflp122;FRT82 abd-AD24/TM6 and w y hsflp122;FRT82 GFP/TM3 , neoFRT82B Sb/TM6 , 71B Gal4 , Eip-71CD-Gal4 were procured from the Bloomington Drosophila Stock Centre . During this study we observed that abd-A RNAi line has multiple insertion of the P-element based RNAi vector in same chromosome . This gave us a clue that separating these P-elements would give us lines with less number of insertions and thus weaker abd-A RNAi effect . We recombined the abd-A RNAi carrying chromosome with the wild type chromosome and recovered lighter eye color line . In the case of Abd-B RNAi in McpH27 Fab-71 mutant background we recombined McpH27 Fab-71 mutant chromosome with third chromosome Abd-B RNAi to see the rescue in homozygous condition . In all RNAi and over expression experiments , we always used females from Gal4 lines while male from RNAi or over expression transgenic lines to employ maternal effect . All the phenotypic quantifications in the knockdown or over expression experiments is done using at least 40 larvae , pharates or adult flies of correct genotype . The pupae and larvae of correct genotypes were identified either by loss of Tubby marker present on balancer chromosome or Gal4 driven GFP expression . 10–14 hours old embryos of desired genotypes were used for immunostaining as per the published protocol [7] . Monoclonal antibodies; Anti Ubx ( 1∶50 dilution ) and anti Abd-B ( 1∶10 dilution ) were procured from Developmental Studies Hybridoma Bank , while polyclonal goat anti Abd-A ( 1∶200 ) antibody from Santa Cruz ( 27063 ) . The secondary antibody tagged with HRP or fluorophore was used at 1∶200 dilutions . In the case of HRP chromogenic reaction was performed to stain the embryo and then CNS of the correct stage embryos were dissected out using fine needles and images were taken by using Zeiss AxioScop 2 . For imaging of embryos stained with fluorophore conjugated secondary antibody we used confocal microscope . For immunostaining of abdominal epithelial cells during early pupal development we followed the protocol described by Wang et al . [41] . Briefly , newly pupated pupae were collected on the basis of light color and staged as per the requirement . They were cut longitudinally into two halves from dorsal side with a sharp razor . Internal organs of cut pupae were removed by flushing 1XPBS using 20 µl pipette . Cleaned abdominal epithelia were fixed in 1XPBS +4% paraformaldehyde +0 . 2% dioxycholic acid for one hour . Tissue was further blocked in PBSB ( 1XPBS +1 . 0% BSA ) for 3 hours and then incubated overnight with the antibody in same solution . The primary antibody anti-Ubx was used at 1∶20 dilution , anti-Abd-A at 1∶100 and anti-Abd-B at 1∶10 dilution . Further tissue was washed with PBSB and subjected to secondary antibody ( 1∶200 ) for 2 to 3 hours . After this the tissue was washed and mounted in anti-fade medium with DAPI ( Vectashield ) . Images were taken by using confocal microscope . Pupae of correct stage and genotype were collected in microfuge tubes . They were washed with PBS to remove any adherent dirt and then fixed by boiling for 5 minutes in 1XPBS . The pupal case was removed gently with fine needles while in PBS to hatch the pharates [42] . The imaging of abdominal region of the fly was done by collecting the flies of interest and taking the pictures immediately or storing them at −30°C for later use . Wings and legs were cut if required for proper positioning and visualization of the abdomen . During acquisition of images light was given only from one side to avoid glittering of cuticle . For live cell Imaging of HNCs we followed the protocol described by Nivon et al . [19] . Newly pupated milky white color pupae of correct genotype were collected and placed in moist chamber to avoid drying . Then pupae were aligned in correct orientation so that HNCs face towards the bottom of chamber with the help of halocarbon oil . Imaging of HNCs was done in gap of two hours using Zeiss multiphoton confocal . For calculation of the change in proliferation rate we did live cell imaging of 15 abd-A RNAi HNCs and 10 wild type HNCs . In this analysis we did not include HNCs that showed complete arrest in the proliferation . For making mitotic clones of abd-AD24 mutant , the mutant was recombined with FRT82 chromosome and recombinants were selected on G418 media . The hsflp; FRT82 abdAD24/TM6 females were crossed with FRT82/TM3 and FRT82 Sb/TM6 males separately . Progenies were given heat shock ( 37°C for 2 hours ) at different stages of development for activation of the flipase . We at least screened 30 flies of correct genotype of each stage to score the percent of flies showing mitotic clones . For making flat preparation of adult abdomen cuticle , flies of interest were collected and stored in 3∶1; ethanol:glycerol solution until use . Flies were heated at 90°C in 10% KOH for 10 minutes to dissolve all the tissue except cuticle . The cuticle was washed with PBS and stored in 50% glycerol . For mounting , abdomen of the fly was separated from the whole body by cutting between thorax and abdominal segment 1 . Then the abdomen was cut from dorsal mid line to open the cuticle and remove all other tissues . Finally , cuticles were mounted with dorsal side facing up in 50% glycerol for imaging . | The spatially non-overlapping function of Hox genes is known to determine Antero-posterior body axis in all the bilaterians . The expression of Hox genes is found to be overlapping in several cases . According to the posterior prevalence rule , posterior Hox genes suppress the function of anterior Hox genes in the overlapping expression domains . Our findings show an exception to the rule of posterior prevalence . We show that in the overlapping expression domains of abd-A and Abd-B in early pupal abdominal epithelia , both the genes have essential roles . While abd-A is required for cell proliferation , Abd-B determines the segmental identity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"organism",
"development",
"biology",
"and",
"life",
"sciences",
"organogenesis",
"morphogenesis",
"pattern",
"formation"
] | 2014 | Role of abd-A and Abd-B in Development of Abdominal Epithelia Breaks Posterior Prevalence Rule |
An unrealized potential to understand the genetic basis of aging in humans , is to consider the immense survival advantage of the rare individuals who live 100 years or more . The Longevity Gene Study was initiated in 1998 at the Albert Einstein College of Medicine to investigate longevity genes in a selected population: the “oldest old” Ashkenazi Jews , 95 years of age and older , and their children . The study proved the principle that some of these subjects are endowed with longevity-promoting genotypes . Here we reason that some of the favorable genotypes act as mechanisms that buffer the deleterious effect of age-related disease genes . As a result , the frequency of deleterious genotypes may increase among individuals with extreme lifespan because their protective genotype allows disease-related genes to accumulate . Thus , studies of genotypic frequencies among different age groups can elucidate the genetic determinants and pathways responsible for longevity . Borrowing from evolutionary theory , we present arguments regarding the differential survival via buffering mechanisms and their target age-related disease genes in searching for aging and longevity genes . Using more than 1 , 200 subjects between the sixth and eleventh decades of life ( at least 140 subjects in each group ) , we corroborate our hypotheses experimentally . We study 66 common allelic site polymorphism in 36 candidate genes on the basis of their phenotype . Among them we have identified a candidate-buffering mechanism and its candidate age-related disease gene target . Previously , the beneficial effect of an advantageous cholesteryl ester transfer protein ( CETP-VV ) genotype on lipoprotein particle size in association with decreased metabolic and cardiovascular diseases , as well as with better cognitive function , have been demonstrated . We report an additional advantageous effect of the CETP-VV ( favorable ) genotype in neutralizing the deleterious effects of the lipoprotein ( a ) ( LPA ) gene . Finally , using literature-based interaction discovery methods , we use the set of longevity genes , buffering genes , and their age-related target disease genes to construct the underlying subnetwork of interacting genes that is expected to be responsible for longevity . Genome wide , high-throughput hypothesis-free analyses are currently being utilized to elucidate unknown genetic pathways in many model organisms , linking observed phenotypes to their underlying genetic mechanisms . The longevity phenotype and its genetic mechanisms , such as our buffering hypothesis , are similar; thus , the experimental corroboration of our hypothesis provides a proof of concept for the utility of high-throughput methods for elucidating such mechanisms . It also provides a framework for developing strategies to prevent some age-related diseases by intervention at the appropriate level .
Aging is associated with a decline in the frequency of survivors attaining older ages; i . e . , the frequency of centenarians in human populations is only ∼1/10 , 000 persons . Given the evidence of a genetic basis for longevity [14–16] , we would expect the prevalence of favorable genotypes in genes contributing to prolonged lifespan—i . e . , longevity genes—to be significantly higher among centenarians relative to their prevalence in a younger control population , as can be observed for the favorable genotypes of CETP-VV and APOC-3 in Figure 1 ( see the discussion below ) . Furthermore , although at birth the probability of living more than 100 years is ∼1/10 , 000 , this probability increases to ∼1/250 when an individual reaches his or her life expectancy ( ∼80 years old ) . We would therefore expect the frequencies of longevity genotype to increase monotonically in progressively older age groups . Figure 2 shows the expected monotonic increase for the same genotypes , CETP-VV and APOC-3 ( see also Figure S1 ) . In contrast , deleterious genotypes associated with “aging phenotype” ( e . g . , age-related disease genes ) would be expected to decrease monotonically as mortality selects out individuals with these deleterious genotypes . These considerations suggest that changes in genotypic frequencies affecting lifespan in different age groups can be detected and used to determine the relevant genes associated with the aging process . However , the two components , longevity and an aging phenotype , cannot be disentangled , since an increase of a favorable genotype in a longevity gene necessarily results in a decrease of the frequency of its complementary genotypes . Similarly , a decline in prevalence of a deleterious genotype at an age-related disease gene necessarily implies an increase in its non-deleterious alternatives . Thus , a simple analysis of monotonic increase or decrease in genotypic prevalence across age groups is not sufficient to discriminate between longevity genes and their counterpart age-related disease genes . In this study , we introduce a possible solution to overcome this limitation . As a first step , we must consider the protective effect that longevity genes might confer and what effect this might have on lifespan . It has been shown that offspring of centenarians are healthier than their appropriate age-matched controls , supporting the notion of inheritance from centenarians to their offspring [10 , 12 , 16–20] . This observation lends support to the hypothesis that longevity genes might buffer the deleterious effect of certain genetically determined age-related diseases . Genes associated with the latter set of diseases are here termed buffered disease genes . Figure 3 shows the frequency trend of the deleterious genotypes of two age-related diseases genes , KLOTHO and LPA , and indeed , centenarians are endowed with significantly higher deleterious genotype than the elder control group ( 80 years old ) , and in the case of LPA , even than that of the younger group . The presence of a molecular buffering mechanism has already been discussed and observed in model organisms . Studies of buffering mechanisms such as the chaperone HSP90 have demonstrated that in the wild-type , under normal conditions , the hidden accumulated , mostly deleterious , genetic variation is buffered and not expressed phenotypically; however , when HSP90′s functionality is compromised , that same genetic variation is translated into , mostly deleterious , phenotypic variation [21 , 22] . Further theoretical analyses on the capacity to harbor and express such phenotypic variation have also been reported [23 , 24] . These theoretical findings have been corroborated by experimental data from model organisms [23 , 24] without any a priori assumptions regarding the potential genetic pathways . It is therefore reasonable to extend this approach to genetic hierarchies and pathways responsible for the process of aging in humans . We recognize that not all longevity genes act as buffering mechanisms . Our study therefore focuses on the discovery of those genes that have a buffering effect , and their targets , age-related disease genes . This equivalence suggests that the buffering effect longevity genes are hypothesized to possess , will allow the accumulation of deleterious allelic variants in buffered disease genes . In turn , we expect the prevalence of a deleterious age-related genotype among centenarians to be maintained at a level similar to that prior to the onset of the age-related disease within the ( younger ) control population . Centenarians , however , are rare in human populations; thus , an initial decline in the prevalence of the deleterious genotype in buffered disease genes is expected . As the population ages , the proportion of individuals endowed with favorable genotypes at longevity gene loci increases , as should the proportion of individuals carrying ( the now buffered ) deleterious age-related disease genotype . Thus , such deleterious genotypes should exhibit a U-shaped trend with progressive aging . The following more formal argument describes this more clearly . We first divide the population ( at birth ) into those protected by the longevity allele ( group P ) and those lacking it . We then further subdivide the latter group into those with deleterious allele in age-related disease gene ( unprotected group U ) and those with wild-type allele ( group N ) . Since they are unprotected from the deleterious genes they are carrying , group U is expected to have the shortest mean lifespan , dU , of the three groups ( P , U , and N ) . Assuming that the favorable genotype group P has the longest mean lifespan , dP , ( dP > dN > dU ) , in part because the longevity allele confers a buffering effect against the deleterious alleles they may have , then we will observe a U-shaped curve , which is a consequence of the final population being dominated by group P ( see Text S1 and Figure S3 for simulation results ) . Figure 3 shows such a trend for the deleterious variant of two age-related disease genes , KLOTHO and LPA ( see also Figures S1–S3 ) . An important final step in our analysis is required—that is , to associate longevity genes with their potential targets—buffered disease genes . In the absence of favorable genotypes in longevity genes , the protective effect , and with it the accumulation of deleterious genotype in buffered disease genes , will not occur . Therefore , in a subpopulation lacking longevity-favorable genotypes , a monotonic decline in the prevalence of a deleterious genotype in a buffered disease gene is expected . In contrast , in a subpopulation possessing the favorable genotype , no change , or an increase , in the prevalence of deleterious buffered disease genes will be observed . Figure 4 shows such an interaction between the deleterious variant of LPA and the two variants of CETP ( for further details see also Figure S2 ) . A limiting factor in this final step is that , since centenarians are rare , we expect the favorable genotype in longevity buffering genes also to be rare among younger populations . To overcome this limitation , we make novel use of the centenarian offspring data . Inheritance assures that the genotypes among offspring will be enriched with favorable longevity genotypes . Thus , by admixing the offspring with a control population , we artificially enrich the prevalence of rare longevity genotypes . Because we only make use of the offspring population in this final step , the identification of potential longevity genes and buffered disease genes ( and the analysis of their interactions ) is not affected by the introduction of this artificial enrichment .
We present the analysis of the 66 common allelic site polymorphisms in 36 candidate genes for a lipoprotein phenotype we have studied ( see Methods for the population and statistical considerations ) . Figure 1 visually represents the frequencies of their genotypes in ∼70 and ∼100-year-old subjects ( see Table S1 and Text S1 for a complete list ) . Only the frequencies of homozygosity for the codon 405 valine ( V ) allele of CETP ( VV genotype ) and the homozygote CC genotype in the APOC-3 promoter region APOC-3 C ( -641 ) A , have been determined as having a significantly greater prevalence among centenarians ( both p-values from chi-square tests are less than 0 . 0001 , and show statistical significance even after Bonferroni correction ) . Both genotypes have been previously associated with increased lipoprotein particle size , and thus are associated with a reduced risk for CVD . For that reason , we considered them favorable candidates for longevity genotypes [11 , 25] . The observed greater prevalence of these genotypes in centenarians compared with the control group as a whole does not , however , completely satisfy our first hypothesis . That is , for CETP-VV and APOC-3 CC to be considered favorable longevity genotypes , a monotonic increase with age is expected . Indeed , this turns out to be the case . Figure 2 shows a highly significant ( p < 0 . 0006 ) monotonically increasing frequency trend for APOC-3 CC . For CETP , the monotonic increase trend is also found to be statistically significant ( p < 0 . 047 ) , making both APOC-3 CC and CETP longevity gene candidates . The statistical significance for a monotonic increase of favorable genotypes with age was tested using logistic regression ( see Text S1 for details ) . Most of the genes we have studied , however , have not been reported to differ significantly in frequency between control subjects and centenarians—e . g . , lipoprotein lipase ( LPL ) , lymphotoxin alpha ( LTA ) , low-density lipoprotein receptor ( LDLR ) , and others ( see Figure 1 ) . Lack of statistically significant differences , however , do not signify the irrelevance of these genes to the aging process , since the frequency analysis of the intermediate age groups may well reveal a more complex pattern , as will be shown below . To identify buffered disease genes further , a more detailed examination is required . A deleterious genotype at an age-related buffered disease gene is predicted to decline initially as the population ages . As the population approaches extreme longevity , the initial decline should reverse , and the prevalence of the deleterious genotype should increase . Indeed , studies in French and Italian centenarian subjects [26–28] reported the paradox of an unfavorable genotype and phenotype that are more common in centenarians . Among the set of genes tested , we have identified two such genes: one is LPA ( see Figure 3 ) , which is associated with increased risk for vascular diseases in the elderly [29] . The other gene shown in Figure 3 is the age-related disease gene KLOTHO , an aging gene that is associated with low HDL and reduced CVD risk [30 , 31] . The reported frequency trends for both LPA and KLOTHO follow the expected U-shape trends based on individuals from the control group only , as described above . Using statistical assumptions described above ( see Methods and Text S1 ) , we arrived at a statistically significant quadratic component for the genotypic frequency of LPA ( p < 0 . 035 ) . A similar statistically significant result was shown for KLOTHO ( p < 0 . 035 ) . Interestingly , LPA plasma levels , known to be a factor for coronary artery disease [32] , have also been measured and found to reflect precisely the trends of the genotype: 15 . 3 ± 1 . 8 mg/dl in subjects 60–70 years old , declined to 10 . 8 ± 2 . 1 mg/dl in subjects aged 71–80 years old , with an increment to 15 . 7 ± 2 . 3 mg/dl when subjects achieve an age of 100 years old . To complete our analysis , the buffering effect of longevity genes on their target age-related disease genes needs to be determined . As explained above , to ensure the presence of favorable longevity genes in a younger population , we supplemented the control population with the offspring of centenarians . We then divided the pooled , control offspring population into two subpopulations—those endowed with , and those not endowed with , a favorable genotype in a longevity gene , i . e . , CETP-VV versus CETP-IV / CETP-II . When a chi-square test to reveal the interaction between LPA and CETP genotypes is performed between the 92-year-old age group ( advanced in age , though not yet centenarians ) and the younger 80-year-old group , we find a significant interaction with respect to CETP-VV and CETP-IV / CETP-II ( p < 0 . 026 , chi-square ) . However , the hypothesis described above suggests a stronger result: in a subpopulation endowed with the favorable CETP genotype ( CETP-VV ) , the frequency trend of the deleterious genotype of LPA will exhibit no change , or may even increase , with age . In contrast , a significant decline should be observed in the subpopulation lacking the favorable longevity CETP genotype . Indeed , such a difference in frequency trends is observed ( see Figure 4 ) . The reported differences between the observed trends is statistically significant , p < 0 . 037 . Although we expect to have multiple targets for each of the longevity genes , not all age-related disease genes may be targeted by all longevity genes . In case such protection is not provided , we would not therefore expect to see a significant change in frequency trends between the two subpopulations . For example , when the same pooled control offspring population was subdivided on the basis of the favorable genotype at the APOC-3 locus to test its effect on the deleterious LPA genotype , no significant changes in trend behavior were observed . In addition , as a test of our hypothesis , a similar interaction analysis was performed on the KLOTHO genotype . Since no data from centenarian's offspring were available for the gene–gene interaction analysis , we included individuals from the centenarian population . When an interaction between CETP and KLOTHO exists , one would expect the admixed , control-centenarians population to exhibit a more significant interaction term than that obtained by an admixed control-offspring population . However , no significant interaction was found when associating CETP or APOC-3 genotypes individually with the unfavorable KLOTHO genotype ( P = 0 . 38 and P = 0 . 85 , respectively ) . Figure 5 shows that the KLOTHO's frequencies for the two subpopulations , with and without favorable CETP genotype , follow a similar U-shape trend . The lack of interaction suggests that , though CETP and KLOTHO may both have an influence on lipoprotein size , no buffering mechanism can be inferred . Finally , to corroborate our findings with existing knowledge about interactions among the set of genes we have identified , we applied a two-stage network analysis . First , using GRID ( General Repository of Interaction Database ) , we identified all the possible targets of CETP , LPA , and APOC-3 ( primary interactions ) , and then extended the scope to include secondary and tertiary interactions which resulted in a network of 248 nodes and 450 edges . From this extended network , we then selected those genes that lie on the minimum pathways ( shortest path length linking two nodes ) among CETP , LPA , and APOC-3 . This resulted in the following additional genes: APOE , LPAL2 , PLTP , and APOA1 . In the second stage , we used the PathwayArchitect software application ( Stratagene , http://www . strategene . com ) to further corroborate the resulting interaction network . Results of this analysis yielded a similar outcome , that is , the same genes resulted in having the highest confidence index of interactions in the pathway layout graph . Figure 6 delineates a subnetwork of known interactions among proteins relevant to our hypothesis . As can be observed , CETP indeed interacts with LPA . This interaction is two-fold , via LPAL2 , a LPA-like 2 lipoprotein , and again , through LPAL2 via APOA1 apolipoprotein A-1 [33 , 34] . These findings call for further analysis of the role of the LPAL2 and APOA1 proteins . Interestingly , although the pathway analysis revealed direct contact between APOC-3 and LPA [35–38] , the hypothesis test revealed that buffering does not occur between APOC-3 and LPA , since there exists no significant interaction ( p = 0 . 076 ) . This observation suggests that given the similar biological effect APOC-3 and CETP has on lipoprotein size , buffering is most likely mediated by other , yet to be discovered , biological means . Finally , when the KLOTHO protein was introduced into the pathway analysis , no additional links were found , nor did we find any link between KLOTHO and any of the subset of proteins tested . This finding is in accord with our hypothesis , due to lack of interaction between CETP or APOC-3 and KLOTHO . Corroborating our hypothesis with known protein interactions indicates that our analysis can predict and be used to further suggest interactions not yet known .
Recent progress in the search for candidate aging genes in centenarian studies has been significantly helped by the availability of more sensitive statistical techniques [39 , 40] . For example , an allele-specific association with longevity was found for SOD2 [41] due to the increased sensitivity of the relative risk method [42 , 43] over the gene frequency method [41] . Given the important polygenic aspects of the aging phenotype , an approach that searches for candidates within their genetic context is urgently needed . Previous work has already demonstrated the importance of genetic background for longevity , suggesting a role for epistasis [44 , 45] . The study we present here moves further in this direction by not only giving the genetic background a central role , but furthermore suggesting a novel mechanistic explanation based upon previous theoretical results on gene–gene interactions and buffering . A previously unfavorable LPA genotype has been reported to be paradoxically increased in French and Italian studies of centenarians [26–28] , representing an increased frequency of disease genotype in centenarians . Our results suggest that a harmful genotype probably does not turn into a protective one , but rather indicates a protection by other favorable longevity genotypes . Indeed we show that increased CETP-VV genotype , CETP levels , and favorable lipoprotein sizes may buffer the deleterious effects of LPA genotypes , thus allowing the accumulation of unfavorable genotypes among centenarians . Similar U-shaped patterns of genotypic frequency with respect to age have also been reported by Tan et al . [42] for the C282Y allele and by Cavallone et al . [46] for the PON1 gene , making them candidates for further analysis as buffered age-related disease genes . To identify the frequency pattern of genotypes associated with aging and longevity , any analysis must include representation of all age groups . Because the nadirs of frequencies for LPA and KLOTHO genotypes were at age ∼80 , previous studies that identified putative longevity genotypes ( due to rising life expectancy from ages 80 to 100 ) , may have identified protected aging genes but not true longevity genes . Comparing populations at ∼50 and ∼100 years old would potentially exclude individuals possessing aging genes that may be buffered at a later age . It is also important to realize that , because we have obtained a large number of rare individuals with exceptional longevity , classical genetic rules , such as the Hardy-Weinberg equilibrium , do not apply in this population . This should be retrospectively considered in groups of centenarians whose results were dropped because they violated the Hardy-Weinberg equilibrium . In other words , because the genotypes of survivors are “selected , ” the greater the attribution of a genotype to longevity , the greater is the divergence from Hardy-Weinberg equilibrium among the elderly . The focus of the current study has been on the longevity genes of centenarians . However , given the complexity of the trait , which may derive from multiple , redundant pathways , pleiotropic interactions , and other environmental factors , it is highly likely that different individuals achieve longevity by different means . Parallel studies in other isolated populations , such as the Icelandic , will provide the means to address new pathways for longevity . Confirming these findings in the general population will also facilitate identification of longevity genes at the individual level . It is most likely that longevity involves a far more complex relationship among longevity and disease genes than the pairwise interactions we have introduced here . Yet , our results suggest that this approach can contribute to our understanding of processes as complex as aging . We used two examples of genotypes that seemed to protect from several age-related conditions . Thus , we expect each of these genotypes to provide protection against several aging genotypes . Indeed , we conclude that investigating the genetic pathways for “aging phenotypes , ” such as age-related disease and the pathways that buffer these effects , combined with analyses of quantitative traits , may suggest strategies to modulate the disease phenotypes of aging . For example , if any single longevity gene buffers against several aging genes , agents could be developed to exploit such a drug with widespread protective effects .
Our population , Ashkenazi Jews , is an ideal study group because social , political , and religious pressures limited this population to a relatively few founders [47] . This genetic homogeneity is paralleled by a relatively homogeneous socioeconomic and educational status . Inbreeding in this population has allowed successful genetic research in Ashkenazi Jews , including the characterization of multiple rare autosomal recessive disorders such as Tay-Sachs disease [48] , factor XI deficiency [49] , and hyperinsulinemic hypoglycemia of infancy ( i . e . , sulfonylurea receptor mutations ) [50] , as well as common diseases such as breast and ovarian cancer ( i . e . , BRCA1 and BRCA2 gene mutations ) [8] . A limitation of this model is the lack of inclusion of a complex gene-environment interaction . However , by selecting an environmentally homogeneous population , we minimize the effects of such interaction . Three hundred and five probands with exceptional longevity ( 228 females and 77 males , age 98 . 2 ( 0 . 36 ) years [mean ( SE ) ] , range 95–109 years; 48% over the age of 100 years ) were recruited to participate in the study . Birth certificates or dates of birth as stated on passports defined the participants' ages . Probands were required to have been living independently at 95 years of age as a reflection of good health , although at the time of recruitment they could be at any level of dependency . In addition , for inclusion probands were required to have a child who was willing to participate in the study . The offspring group consisted of 227 females and 203 males ( age 68 . 3 [0 . 45] years , range 54–89 years ) . Finally , the control group consisted of 265 females and 203 males ( age 69 . 5 [0 . 52] years , range 54–90 years ) , matched in age to the offspring . Details regarding the recruitment and demographics of this group can be found in [11 , 16 , 17] ( see Table 1 for age distribution of the subjects recruited ) . To test our hypothesis , we determined the prevalence of 66 common polymorphic sites in 36 genes that are known to be risk factors for cardiovascular disease ( CVD ) using a multilocus PCR-based genotyping assay . Briefly , DNA was amplified using multiplex reaction containing biotinylated primer pairs . Amplified fragments within each PCR product pool were then detected colorimetrically with sequence-specific oligonucleotide probes immobilized in a linear array on nylon membrane strips . Probe specificities had previously been confirmed by sequencing and by use of DNA genotyped independently through other methods such as restriction length polymorphism analysis [51] . The following are the statistical considerations in identifying potential buffered disease genes contributing to the aging phenotype . Among the SNPs which demonstrate a significant decline in frequency with age in the control group , we examined those in which the prevalence is significantly higher in the centenarian population relative to the control 80–90 age group . The identified SNPs are those associated with candidate-buffered disease genes . For those SNPs which show an initial significant decline with age followed by a significant increase , that is , a U-shape trend of age-related target genes , we confirm the pattern by fitting a generalized linear model with data from the combined control and centenarian groups . We use a binomial model with an identity link function and both linear and quadratic terms for age , and test for the significant quadratic component . More specifically , the binary response ( Y ) of having ( Y = 1 ) or not having ( Y = 0 ) the deleterious genotype of an age-related disease gene is modeled as P ( Y = 1 ) = b0 + b1age + b2age2 . Note that the standard logistic regression does not apply to this case since it models probability as monotonic to covariates . Maximum likelihood estimates of the coefficients b0 , b1 , b2 are obtained by the Fisher Scoring method . The statistical significance of the quadratic term is then determined by the likelihood ratio test that compares the likelihood of the model of b2 ≠ 0 with the model of b2 = 0 . If the quadratic term is significant and the minimum of the quadratic function −b1/ ( 2b2 ) falls within age range of our subjects , the relevant gene is further considered . To formally test the statistical significance of the interaction term between frequency trends of the two subpopulations ( those endowed with and without favorable genotypes at the longevity gene ) , it is equivalent to test the statistical significance of interaction effects between the factor “age” and the “longevity genotype” factor , using logistic regression . The binary response in the logistic regression is whether or not a subject has the deleterious genotype of the buffered disease gene . To test the significance of age–gene interaction , the model with main effect only and the model with the interaction effects are compared using log-likelihood ratio test ( for additional information see Text S1 ) . | Previous research showed that the frequency of deleterious genotype of some age-related disease decreases its prevalence as the population ages , as expected , since subjects with deleterious genotype are weeded out due to mortality . There exists , however , a set of age-related genes whose deleterious genotype indeed decreases up to ages 80–85 , but subsequently increases monotonically , until by age 100 its prevalence is similar to that at age ∼60 . Why is a known harmful genotype so prevalent among centenarians ? Most likely because this genotype is protected by longevity genes . We corroborated this hypothesis by studying gene–gene interactions between age-related disease genotypes and longevity genotypes . Our findings suggest that individuals with the favorable longevity genotype can have just as many deleterious aging genotypes as the rest of the population because their longevity genotype protects them from the harmful effects of the other . We identify genes contributing to extreme lifespan as well as their counterpart , age-related disease genes . Our findings provide a proof of concept for the utility of high-throughput methods , and for elucidating mechanisms by which longevity genes buffer the effects of disease genes . Our approach gives hope for developing new medications that will protect against several age-related diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"buffering",
"systems",
"biology",
"longevity",
"computational",
"biology",
"evolutionary",
"biology",
"gene",
"networks",
"genetics",
"and",
"genomics",
"evolution"
] | 2007 | Buffering Mechanisms in Aging: A Systems Approach Toward Uncovering the Genetic Component of Aging |
Astroglia from the postnatal cerebral cortex can be reprogrammed in vitro to generate neurons following forced expression of neurogenic transcription factors , thus opening new avenues towards a potential use of endogenous astroglia for brain repair . However , in previous attempts astroglia-derived neurons failed to establish functional synapses , a severe limitation towards functional neurogenesis . It remained therefore also unknown whether neurons derived from reprogrammed astroglia could be directed towards distinct neuronal subtype identities by selective expression of distinct neurogenic fate determinants . Here we show that strong and persistent expression of neurogenic fate determinants driven by silencing-resistant retroviral vectors instructs astroglia from the postnatal cortex in vitro to mature into fully functional , synapse-forming neurons . Importantly , the neurotransmitter fate choice of astroglia-derived neurons can be controlled by selective expression of distinct neurogenic transcription factors: forced expression of the dorsal telencephalic fate determinant neurogenin-2 ( Neurog2 ) directs cortical astroglia to generate synapse-forming glutamatergic neurons; in contrast , the ventral telencephalic fate determinant Dlx2 induces a GABAergic identity , although the overall efficiency of Dlx2-mediated neuronal reprogramming is much lower compared to Neurog2 , suggesting that cortical astroglia possess a higher competence to respond to the dorsal telencephalic fate determinant . Interestingly , however , reprogramming of astroglia towards the generation of GABAergic neurons was greatly facilitated when the astroglial cells were first expanded as neurosphere cells prior to transduction with Dlx2 . Importantly , this approach of expansion under neurosphere conditions and subsequent reprogramming with distinct neurogenic transcription factors can also be extended to reactive astroglia isolated from the adult injured cerebral cortex , allowing for the selective generation of glutamatergic or GABAergic neurons . These data provide evidence that cortical astroglia can undergo a conversion across cell lineages by forced expression of a single neurogenic transcription factor , stably generating fully differentiated neurons . Moreover , neuronal reprogramming of astroglia is not restricted to postnatal stages but can also be achieved from terminally differentiated astroglia of the adult cerebral cortex following injury-induced reactivation .
While exerting diverse functions within the brain parenchyma [1] , astroglia are remarkable in that they also function as neural stem or progenitor cells in specific regions of the postnatal and adult brain [2] , such as the ventricular subependymal zone [3] and the subgranular zone of the hippocampus [4] , [5] . This raises the possibility that even astroglia from non-neurogenic regions such as the cerebral cortex may be reprogrammed towards neurogenesis when provided with the appropriate transcriptional cues . Indeed , we could previously show that astroglia from the early postnatal cerebral cortex can be reprogrammed in vitro towards the generation of neurons capable of action potential ( AP ) firing by a single transcription factor , such as Pax6 or its target , the pro-neural transcription factor neurogenin-2 ( Neurog2 ) [6] , [7] . These findings may open interesting avenues towards the potential activation of endogenous astroglia for neuronal repair of injured brain tissue . However , several major obstacles remained to be overcome to fully exploit the potential of reprogrammed astroglia as an endogenous cellular source for neuronal repair . Firstly , reprogramming of astroglia towards neurons remained incomplete as the astroglia-derived neurons failed to establish a functional presynaptic output [7] , an obvious hurdle towards functional repair that requires participation in a neural network . Secondly , given the lack of functional presynaptic output , we could not determine the neuronal subtype generated by the reprogrammed astroglial cells [7] . This raises the conceptual concern of whether neurons derived from astroglial cells in a given brain region may be restricted towards the generation of a respective neuronal subtype . During development of the forebrain in rodents , stem/progenitor cells in the dorsal telencephalon generate exclusively excitatory glutamatergic neurons , directed by Pax6 and Neurog1/2 [8]–[10] , while stem/progenitor cells in the ventral telencephalon give rise primarily to inhibitory GABAergic neurons , governed by the fate determinants mammalian achaete-schute homolog 1 ( Mash1 ) [11] , [12] and Dlx1/2 [13] . Region-specific fate restriction also seems to apply for adult neural stem cells that are intrinsically specified towards the generation of distinct neuronal subtypes [14] . This implies that despite their multipotent nature in regard to generating different glial cell types and neurons , the subtype identity of the neurons generated from these stem cells is predetermined ( see also [15] ) and is not altered following transplantation [14] . This raises the important question of whether neuronal reprogramming of astroglia derived from the cerebral cortex , a region derived from the dorsal telencephalon , may be restricted towards the generation of glutamatergic neurons , or whether this region-specific bias could be overcome by forced expression of the appropriate neurogenic fate determinants . Such an ability to generate both glutamatergic and GABAergic neurons from astroglia may be a crucial step towards restoring a damaged or imbalanced neuronal network . Towards this end , we first aimed at a more potent neuronal reprogramming by inducing higher and more persistent expression of neurogenic fate determinants in astroglial cells . This allowed us not only to obtain fully functional neurons that also establish synapses from astroglial cells in vitro but also to demonstrate that distinct neurogenic transcription factors , such as on the one hand Neurog2 and on the other Dlx2 alone or in combination with Mash1 , can indeed instruct the selective generation of different neuronal subtypes , such as glutamatergic and GABAergic neurons , respectively . Moreover , we found that the reprogramming efficiency of postnatal cortical astroglia towards GABAergic neurons by Dlx2 could be enhanced by first expanding the astroglial cells under neurosphere conditions prior to forced expression of Dlx2 . Given that following brain injury reactive astroglia from the adult cerebral cortex de-differentiate , resume proliferation , and can give rise to self-renewing neurospheres in vitro [16] , we finally show that neuronal reprogramming and subtype specification are not restricted to postnatal stages but can also be achieved from adult cortical astroglia responding to injury .
Failure to establish a functional presynaptic compartment by astroglia-derived neurons may be due to an incomplete reprogramming [7] . Here , we hypothesized that stronger and more persistent expression of neurogenic fate determinants may be a prerequisite for a more complete reprogramming of astroglia towards synapse-forming neurons . We have previously shown that expression levels from LTR ( long terminal repeat ) -driven MMLV ( Moloney Murine Leukemia Virus ) -derived retroviral constructs , which we employed in previous studies , are only about 2–3-fold of the endogenous expression [6] , [17] . Moreover , these viral vectors are prone to silencing [18] and we observed a severe decrease in Neurog2 or green fluorescent protein ( GFP ) reporter expression already 7–14 d after transduction [7] , [19] . Thus , in order to overcome the limitations in synaptogenesis of neurons derived from reprogrammed astroglia , we examined the effect of stronger and more persistent expression of Neurog2 on neuronal reprogramming of astroglia from the cerebral cortex . We therefore subcloned Neurog2 into a self-inactivating retroviral vector driving gene expression under the control of a chicken beta-actin promoter ( pCAG ) optimized for long-term expression over months in the adult mouse brain [20] . Astroglia cultures were prepared from postnatal day 5–7 ( P5–P7 ) cerebral cortex as described previously [7] and 1 wk later cells were passaged and subsequently transduced with a retroviral vector encoding Neurog2 and DsRed ( pCAG-Neurog2-IRES-DsRed ) or with a control virus encoding DsRed only ( pCAG-IRES-DsRed ) . Consistent with a stronger and more persistent expression driven by the pCAG promoter , high levels of Neurog2 and DsRed protein were still detected at 5–6 wk following retroviral transduction of cortical astroglia ( unpublished data ) . In agreement with our previous observation on the high efficiency of neurogenesis from astroglia following forced Neurog2 expression , the vast majority of Neurog2-transduced astroglia had differentiated into βIII tubulin-positive , GFAP-negative neurons after 10 d in culture ( Figure S1A and S1A'; 70 . 2%±6 . 3% at 9 . 8±3 . 1 days post-infection ( DPI ) , 5 independent experiments , n = 1 , 022 DsRed-positive cells counted ) , in contrast to control retrovirus transduced cells ( 1 . 8%±1 . 8% of βIII tubulin-positive cells at 7 . 3±1 . 0 DPI , 3 independent experiments , n = 3 , 235 DsRed-positive cells counted ) . Time-lapse video microscopy revealed that the initial conversion of astroglia into neurons requires approximately 4 d , confirming previous results [7] , and can occur at high efficiency ( Video S1 ) . By 2–3 wk post-transduction , neurons derived from Neurog2-transduced astroglia had acquired MAP2 immunoreactivity , indicative for dendritic maturation ( Figure 1A and 1B ) . Most strikingly , immunostaining for the vesicular glutamate transporter 1 ( vGluT1 ) , present in synaptic vesicles within presynaptic terminals of glutamatergic neurons , revealed that the vast majority of astroglia-derived neurons exhibited a dense labelling with vGluT1-positive puncta outlining their soma and their MAP2-positive processes 4 wk post-infection with Neurog2 ( Figure 1A and 1B , 85 . 4%±5 . 0% of DsRed-positive neurons at 26 . 3±2 . 2 DPI , n = 3 independent experiments , n = 170 DsRed-positive neurons counted ) . This was in pronounced contrast to the virtual absence of such staining upon transduction with the LTR-driven construct ( pCLIG-Neurog2 ) as described previously [7] and also no vGluT1 immunoreactivity could be detected in astroglial cultures transduced with the control vector ( unpublished data ) . Thus , these data suggest that astroglia reprogrammed with the pCAG-Neurog2-containing retroviral vector acquire a glutamatergic phenotype forming presynaptic specializations . As vGluT1 immunoreactivity does however not allow to ascertain the neurotransmitter identity of an individual labelled neuron , as the vGluT1-positive puncta may arise from other neurons in the next set of experiments we assessed with single and pair electrophysiological recordings whether astroglial cells reprogrammed by Neurog2 indeed give rise to functional glutamatergic autapses or synapses after a period of 14–32 DPI . As shown in Figure 1D , suprathreshold step-depolarisation of a DsRed-positive neuron ( i . e . presynaptic neuron , black asterisk , Figure 1C ) resulted in an autaptic response in the stimulated neuron and an inward current in a nearby DsRed-positive neuron with a short delay typical of a monosynaptic connection ( i . e . postsynaptic neuron , red asterisk , Figure 1C ) . In addition , the AMPA/kainate glutamate receptor antagonist CNQX completely abolished both the autaptic and the synaptic current , demonstrating the glutamatergic nature of the presynaptic neuron ( Figure 1D ) . Among all the Neurog2-transduced astroglia-derived neurons recorded ( n = 36 , average age of cells: 24 . 6±0 . 9 DPI ) , 58 . 3% exhibited either glutamatergic autaptic connections onto themselves or glutamatergic synapses onto nearby neurons ( Figure S2A ) . In none of the recordings from neurons derived from Neurog2-transduced astroglia was a GABAergic connection observed ( Figure S2A ) . In accordance , cultures transduced with Neurog2 encoding retrovirus were devoid of any vesicular GABA transporter ( vGaT ) immunoreactivity ( unpublished data ) . Thus , these data provide evidence that Neurog2 does not only induce a generic neuronal fate in postnatal astroglia but selectively elicits differentiation along the glutamatergic lineage , in exclusion of GABAergic neurogenesis . Consistent with the specification of postnatal astroglia towards a glutamatergic identity , forced expression of Neurog2 resulted in the induction of the T-box transcription factors Tbr2 ( Figure S1B and S1B' ) in 20 . 7%±1 . 9% of the DsRed-positive cells at 4 DPI ( n = 4 coverslips , n = 634 DsRed-positive cells counted ) and Tbr1 ( 48 . 2% of DsRed-positive neurons at 7 DPI , n = 112 DsRed/βIII tubulin-double positive cells counted; Figure S1C and S1C' ) as shown previously [7] , hence of two well characterised hallmarks of glutamatergic neurogenesis [21] . Moreover , by 4 wk of forced Neurog2 expression , astroglia-derived neurons expressed high levels of the forebrain glutamatergic neuron specific Ca2+/Calmodulin dependent kinase subunit IIα [22] , accumulating at dendritic spine-like structures which were typically in opposition of vGluT1-positive presynaptic terminals ( Figure 1E–1F' ) . Consistent with the development of excitatory networks in Neurog2-reprogrammed astroglia cultures , we also observed the emergence of self-driven synaptic activity , resulting eventually in the occurrence of barrages of synaptic currents ( Figure 2A ) . To monitor such self-driven activity , we performed calcium imaging experiments of neurons derived from Neurog2-reprogrammed astroglia . Figure 2B illustrates two neurons that exhibited spontaneous , recurrent , and synchronous Ca2+ transients ( Figure 2B–2B” ) . These Ca2+ transients were completely abolished in the presence of CNQX ( Figure 2B” ) . The majority of the DsRed-positive neurons that we analysed ( 63 . 8% , n = 47 imaged neurons , 3 independent experiments ) exhibited Ca2+ transients at 14–43 d after transduction with Neurog2 , thus indicating the high degree of incorporation of Neurog2-transduced astroglia into excitatory neuronal networks . These data clearly demonstrate that forced expression of Neurog2 driven by the pCAG retroviral vector is sufficient to instruct postnatal cortical astroglia to generate fully functional synapse-forming glutamatergic neurons . In order to ascertain the astroglial nature of the cells that gave rise to functional glutamatergic synapses following reprogramming by Neurog2 , we took advantage of a transgenic mouse line in which GFP expression can be induced in astroglia and is maintained in their progeny . Heterozygous mice in which the expression of a tamoxifen-inducible Cre recombinase is driven by the astroglia specific L-glutamate/L-aspartate transporter promoter ( GLAST::CreERT2 ) [23] were crossed to a reporter mouse line ( Z/EG ) [24] to generate double heterozygous mutants ( GLAST::CreERT2/Z/EG ) that were used in the present study . Cre-mediated recombination of the reporter locus was induced via tamoxifen administration from postnatal day 2 ( P2 ) until sacrifice ( P5–P7 ) . Astroglia cultures were prepared as described above and , 1 wk later , cells were passaged onto glass coverslips . The vast majority of GFP reporter-positive cells were immunoreactive for GFAP ( 98 . 7%±0 . 7% ) at 1 d after plating , with few cells being positive for the oligodendroglial markers NG2/O4 ( 1 . 2%±0 . 7% ) and none ( 0 . 1%±0 . 1% ) for the neuronal marker βIII tubulin ( Figure 3A; n = 3 independent experiments , n = 1 , 560 GFP-positive cells counted ) . These data indicate that , under our culture conditions , most reporter-positive cells at the time of transduction possess an astroglial identity . These cells largely remain within their astroglial lineage ( 86 . 9%±12 . 7% of GFP-positive cells expressing GFAP ) when analysed at later stages ( Figure 3A'; n = 4 independent experiments , n = 1 , 363 GFP-positive cells counted; 9–21 d following plating ) . We noted , however , a slight increase in the number of NG2/O4-positive cells ( 13 . 0%±12 . 8% ) , likely due to the expansion of few reporter-positive clones of oligodendrocyte precursors . Also at later stages reporter-positive cells did not give rise to βIII tubulin-positive neurons ( 0 . 1%±0 . 1%; Figure 3A' ) . To determine the identity of fate-mapped astroglial cells following retroviral transduction , we performed immunostaining for GFP ( identifying cells of astroglial origin ) , DsRed ( identifying transduced cells ) , and either βIII tubulin , MAP2 , or GFAP ( identifying neuronal and astroglial cells , respectively ) . Notably , the stochastic infection of the subset of genetically recombined cells results in a limited number of double-targeted cells . When cultures of adherent astroglia were transduced with the control retrovirus encoding DsRed only , fate-mapped astroglial cells co-expressing GFP and DsRed remained in the glial lineage , as revealed by their astroglial morphology and GFAP expression 1 mo after transduction ( Figure 3B–3B” ) . In sharp contrast , when cultures of tamoxifen-induced astroglia were transduced with the new retrovirus encoding Neurog2 and DsRed , most GFP/DsRed-double-positive fate-mapped astroglia were reprogrammed into neurons expressing the neuronal markers βIII tubulin and MAP2 ( 67 . 3%±12 . 7% among GFP/DsRed-double positive cells at 8 . 0±1 . 0 DPI , n = 3 independent experiments , n = 217 double-positive cells counted; Figure 3C–3C” ) . Single cell tracking of GFP-reporter positive cells following Neurog2-transduction allowed the direct visualisation of the glia-to-neuron conversion of fate-mapped cells over the time course of 5 d ( Figure 3D and 3D'; Video S1 and Video S2 ) . Perforated patch clamp recordings of these fate-mapped astroglia-derived cells reprogrammed by Neurog2 revealed their functional neuronal identity as these cells fired APs following step-current injection in current clamp ( n = 8; Figure 4A–4C ) . In the next set of experiments , we assessed whether neurons derived from fate-mapped astroglia could give rise to functional glutamatergic autapses ( Figure 4D–4I ) . Step-depolarisation of GFP/DsRed-double-positive neurons at 0 . 05 Hz evoked a sequence of both autaptic and polysynaptic components ( 2 out of 8 cells recorded ) consistent with the excitatory nature of the recorded neurons ( average age of the cells: 18 . 1±2 . 2 DPI; Figure 4D–4I , insets ) , while at higher stimulation frequency ( 1 Hz ) the autaptic component with a short decay time typical of glutamatergic synaptic transmission [25] could be observed in isolation ( Figure 4F and 4I ) . Consistent with their glutamatergic nature , fate-mapped astroglia reprogrammed by forced expression of Neurog2 also exhibited a dense labelling of vGluT1-positive puncta ( Figure 4J and 4K ) . These data clearly demonstrate that Neurog2 instructs fate-mapped astroglia from the postnatal cerebral cortex to acquire a glutamatergic identity . Given that our reprogramming strategy is based on retrovirally mediated expression of neurogenic fate determinants , only cells undergoing cell division will be targeted . In order to examine whether cell division is required for fate conversion to occur , we assessed whether neuronal reprogramming can be also achieved when the Neurog2 and DsRed encoding plasmid is delivered to the postnatal astroglia by transfection , i . e . a gene transfer strategy which does not select for dividing cells , and tracked single transfected cells by time-lapse video microscopy . Transfection with the Neurog2 encoding plasmid resulted in a similar degree of reprogramming after 4 d ( 14 cells out of 17 , Figure 5 ) as obtained after retroviral transduction . Of note , in four cases neurons were generated directly from single astrocytes without a prior cell division ( Figure 5 and Video S3 ) . Thus direct lineage conversion can occur in the absence of cell division . Based on our finding that forced expression of Neurog2 can selectively drive cortical astroglia towards the generation of functional and synaptically integrated glutamatergic neurons , we next asked whether cortical astroglia may also be directed towards distinct neuronal subtypes . In particular , we asked whether neuronal fate determinants known to instruct the genesis of GABAergic neurons during embryonic development may be sufficient to exert a similar effect on postnatal astroglia . As the homeobox transcription factor Dlx2 is one of the key factors involved in GABAergic neuron specification in the developing ventral telencephalon [13] and in adult neurogenesis [26] , we examined whether forced expression of Dlx2 is also sufficient to induce a neuronal and possibly a GABAergic fate in cortical astroglia . To test this , astroglia cultures from P5–P7 cortex of C57BL/6J or GLAST::CreERT2/Z/EG mice were transduced with the same high-expressing retrovirus encoding in this case Dlx2 and DsRed ( pCAG-Dlx2-IRES-DsRed ) , and cells were immunostained for GFP ( to identify cells of astroglial origin ) , DsRed ( to identify Dlx2-transduced cells ) , and the neuronal markers βIII tubulin or MAP2 after various differentiation time periods in culture . Upon forced expression of Dlx2 , a substantial number of postnatal cortical astroglia were redirected towards a neuronal identity as revealed by βIII tubulin or MAP2 expression ( 35 . 9%±13 . 0% at 10 . 7±2 . 0 DPI , n = 3 independent experiments , n = 392 DsRed-positive cells counted ) . Notably , however , the efficacy of neurogenesis elicited by Dlx2 was significantly lower than the one elicited by Neurog2 ( see above ) . Fate-mapping analysis confirmed the astroglial nature of the cells reprogrammed by forced expression of Dlx2 as observed 22 DPI ( Figure 6A–6A” ) . Next , to confirm the neuronal identity of the astroglia-derived cells , we performed patch clamp recordings . All the cells expressing Dlx2 and exhibiting a neuronal morphology , that we recorded , were capable of AP firing in response to step current injection ( n = 33 ) . In particular , this also held true for GFP-positive neurons originating from fate-mapped astroglia that had been reprogrammed by Dlx2 ( n = 9; Figure 7A–7C” ) . Notably , neurons derived from Dlx2-transduced astroglia exhibited distinct firing patterns , with most of them revealing immature characteristics ( single to few spikes , 22 out of 30 cells recorded ) ( Figure 7A” and Figure 8 ) . The eight remainder cells exhibited firing patterns which could be classified into three categories [27] , [28] , namely regular , stuttering , and low-threshold burst spiking ( Figure 7B”–7C” and Figure 8 ) , suggestive of the maturation into distinct types of non-fast spiking interneurons [29] . Similarly , the majority of fate-mapped astroglia reprogrammed by Dlx2 ( 7 out of 9 cells recorded ) exhibited immature firing patterns ( Figure 7A–7A” ) , while 2 out of 9 fate-mapped cells developed more mature interneuron-like firing ( Figure 7B–7C” ) . Consistent with the generation of regular- and burst-spiking interneurons [29] , we observed calretinin immunoreactivity in a small subset of the Dlx2-expressing cells ( Figure 6C ) , while no parvalbumin immunoreactivity could be detected . The predominant appearance of immature firing patterns , however , suggests an overall hampered maturation of Dlx2-reprogrammed astroglia . Accordingly , astroglia-derived neurons reprogrammed by Dlx2 exhibited much higher input resistance values than Neurog2-derived neurons after the same time in culture ( Figure S2B and S2C; 2 , 319 . 2±187 . 9 MΩ at 26 . 0±1 . 4 DPI ( n = 26 ) versus 1 , 111 . 8±211 . 1 MΩ at 23 . 7±1 . 5 DPI ( n = 20 ) , Dlx2 versus Neurog2 , respectively ) . Surprisingly , the high input resistance of Dlx2-expressing neurons did not decrease but even slightly increased with time in culture ( Figure S2B; 2 , 786 . 4±440 . 3 MΩ at 35 . 9±1 . 2 DPI ( n = 7 ) ) , while in the case of Neurog2-reprogrammed astroglia input resistance decreased over time ( 608 . 6±125 . 0 MΩ at 26 . 8±1 . 2 DPI; n = 14; Figure S2C ) . Taken together , these data show that some postnatal cortical astroglia can be redirected by forced expression of Dlx2 towards a neuronal identity; however , in sharp contrast to the progressive maturation of Neurog2-transduced cells , most of the astroglia-derived neurons reprogrammed by Dlx2 remain in a rather immature state , suggesting a comparatively less efficient reprogramming by Dlx2 . Next we assessed whether some of these relatively immature neurons derived from Dlx2-reprogrammed astroglia may nevertheless establish functional autaptic or synaptic connections . We first performed immunocytochemistry for vGluT1 and for vGaT , the latter known to be expressed in synaptic vesicles located in presynaptic terminals of GABAergic neurons . In sharp contrast to reprogramming by Neurog2 , astroglia-derived neurons reprogrammed by Dlx2 were devoid of vGluT1 immunoreactivity ( unpublished data ) , but some of them ( 33 . 7%±3 . 6% at 22 . 0±0 . 6 DPI , n = 339 DsRed-positive neurons counted; n = 3 independent experiments ) were found to exhibit labelling of vGaT-positive puncta outlining both their soma and their processes ( Figure 6D ) . In addition , a small subset of DsRed-positive neurons exhibited GAD67 immunoreactivity ( Figure 6B and 6B' ) . These findings therefore suggest that Dlx2 induces a GABAergic identity in the reprogrammed astroglia . Consistent with an interneuron phenotype , we could also record in 9 out of 33 neurons spontaneous synaptic currents exhibiting a slow decay time , characteristic of GABAergic synaptic events ( Figure 7D and 7E ) . Finally , in few cases , step-depolarisation in voltage clamp evoked an autaptic response of the stimulated neuron ( 6 . 1% of the DsRed-positive neurons recorded , n = 33 , age of the cells: 26 . 9±1 . 4 DPI; Figures 7F and S2A ) . In accordance with the above data , these autaptic responses exhibited slow decay time kinetics and were abolished by the GABAA receptor antagonist bicuculline ( Figure 7F ) , thus demonstrating the GABAergic nature of these autapses . Taken together , these data strongly indicate that forced expression of Dlx2 , in sharp contrast to Neurog2 , can induce the reprogramming of astroglia from the postnatal cortex towards a GABAergic phenotype . However , whereas Neurog2 redirected the majority of astroglia towards functional glutamatergic neurons , only few astroglial cells reprogrammed by Dlx2 differentiated into fully functional , GABAergic neurons ( 58% versus 6% , respectively ) , thus indicating that Dlx2-induced reprogramming remains partial in most of the cells . Since we have previously shown that Mash1 , a transcription factor located up-stream of Dlx2 in the interneuron fate specification [30] , the direct targets of which overlap only partially with that of Mash1 [31] , can also reprogram postnatal astroglia towards neurogenesis [7] , we tested whether co-expression of these two transcription factors further promote neurogenesis and subsequent interneuron differentiation of reprogrammed astroglia [32] . Consistent with previous data [7] , 33 . 5%±17 . 8% of astroglia expressing Mash1 alone developed into βIII tubulin-positive neurons ( Figure 9B , n = 3 independent experiments , n = 226 DsRed-positive cells counted at 10 . 7±2 . 0 DPI ) . In contrast , co-expression of Mash1 and Dlx2 significantly augmented neurogenesis from postnatal astroglia ( 93 . 0%±3 . 1% of βIII tubulin-positive neurons amongst DsRed-positive cells , n = 3 independent experiments , n = 548 DsRed-positive cells counted at 10 . 7±2 . 0 DPI; Figure 9B ) , indicating that these two factors indeed act synergistically . Moreover , compared to cells expressing Dlx2 alone , Mash1/Dlx2 co-expressing neurons exhibited lower input resistance values ( 1 , 237 . 5±278 . 8 MΩ at 18 . 4±1 . 0 DPI; n = 15; Figure 9C ) . Consistent with a more mature status , a higher proportion of Mash1/Dlx2 co-expressing neurons exhibited specific interneuronal firing patterns ( 6 out of 15 cells recorded , Figure 8 and Figure 9A–9A” ) compared to Dlx2 ( 8 out of 30 cells recorded , Figure 8 ) . Despite this enhanced degree of differentiation , none of the recorded cells co-expressing Mash1 and Dlx2 showed an autaptic response ( Figure S2A , n = 15 ) . Taken together , our data provide evidence that postnatal astroglia from the cerebral cortex can be driven towards the generation of interneurons with distinct functional properties by forced expression of Dlx2 or Dlx2 in combination with Mash1 . Next , we examined whether a complete and more efficient reprogramming of astroglia towards synapse-forming functional GABAergic neurons may be achieved by first expanding astroglial cells as neurospheres in presence of mitogens that on the one hand promote de-differentiation of astroglia and on the other hand up-regulate fate determinants normally involved in the generation of GABAergic neurons in the telencephalon , such as Mash1 [33]–[35] . We therefore cultured postnatal astroglial cells as neurospheres before transducing the astroglia-derived neurosphere cells with Dlx2 . It has been previously shown that early postnatal cortical astroglial cells can give rise to neurospheres until P11 [36] . Cortical tissue from P5–P7 C57BL/6J , GLAST::CreERT2/Z/EG , or hGFAP-GFP mice was cultured as neurospheres under non-adherent conditions in serum-free medium and in the presence of EGF/FGF2 , and 1 wk later , astroglia-derived neurosphere cells were passaged and subsequently transduced with retrovirus encoding either Dlx2-DsRed or Neurog2-DsRed , and were allowed for differentiation . The astroglial origin of the neurosphere founder cells was confirmed by culturing single GFP-positive cells derived from the postnatal cortex of hGFAP-GFP mice which gave rise to neurospheres ( Figure S3A–S3D ) . Moreover , quantitative RT-PCR demonstrated that during expansion in EGF/FGF2 neurosphere cells expressed mRNAs for different astroglial markers , such as the specific pan-astrocyte marker Aldh1L1 [37] , at similarly high levels as adherent astroglia after 1 wk in culture , and the mRNA encoding GFAP at even higher levels ( Figure 10A–10E ) . In contrast , no βIII tubulin mRNA could be detected ( Figure 10F ) . Likewise , similar to adherent astroglia neurosphere cells did not express detectable levels of endogenous Neurog2 mRNA ( Figure 10G ) . These data support the notion that during the expansion in EGF/FGF2 , neurosphere cells have an astroglial character . In contrast to adherent astroglia cultures , 94 . 7%±0 . 3% ( n = 3 independent experiments , n = 644 DsRed-positive cells counted ) of the astroglia-derived neurosphere cells transduced with Dlx2 differentiated into MAP2-positive neurons ( Figure S4A and S4C ) . Strikingly , even at younger stages in culture ( 20 . 1±1 . 9 DPI ) , Dlx2-expressing neurons derived from neurosphere cells exhibited substantially lower input resistances ( 1 , 266 . 4±294 . 3 MΩ , n = 7 ) compared to Dlx2-reprogrammed adherent astroglia recorded at 4 wk in culture ( 2 , 319 . 2±187 . 9 MΩ at 26 . 0±1 . 4 DPI , n = 26; Figure S2B ) , indicative of a more advanced neuronal maturation . Immunostaining for vGaT revealed a dense labelling of vGaT-positive puncta ( Figure 11A and 11B ) , thus suggesting that astroglia-derived neurosphere cells transduced with Dlx2 had acquired a GABAergic identity . Finally , electrophysiological recordings demonstrated the GABAergic phenotype of the fate-mapped astroglia expanded as neurospheres and reprogrammed by Dlx2 ( n = 9; Figure 11C–11E ) . Consistent with the widespread vGaT expression , Dlx2-expressing neurons received spontaneous synaptic activity displaying slow decay time kinetics characteristic of GABAergic currents ( 9 out of 10 cells recorded , unpublished data ) . As shown in Figure 11E , step-depolarisation of a GFP-positive , Dlx2-expressing neuron ( Figure 11C and 11D , black asterisk ) evoked an autaptic response of the stimulated neuron that was blocked by bicuculline . Four out of 10 Dlx2-transduced neurosphere-derived neurons recorded were found to form GABAergic autapses , despite being analysed at a younger age compared to the adherent astroglia ( n = 10 , age of the cells: 20 . 1±1 . 6 DPI; Figure S2A ) . Consistent with the development of GABAergic networks in Dlx2-transduced cultures , calcium imaging experiments performed at 27 DPI did not reveal any spontaneous Ca2+ transients in the analysed DsRed-positive neurons ( n = 70 neurons recorded , n = 2 independent experiments; Figure S5 ) . Thus , culturing astroglia under neurosphere conditions clearly eases the reprogramming towards functional GABAergic neurons by Dlx2 transduction compared to the effects obtained in adherent astroglia ( 40% versus 6% , neurosphere cells versus adherent astroglia , respectively ) . Interestingly , astroglia cultured as neurosphere cells could still be reprogrammed by Neurog2 towards a glutamatergic neuronal phenotype . Virtually all astroglia-derived neurosphere cells transduced with Neurog2 differentiated into MAP2-positive neurons ( 91 . 4%±2 . 2% , n = 2 , 272 DsRed-positive cells counted , n = 3 independent experiments , in agreement with [7] ) that exhibited a large soma size and extended several MAP2-positive processes ( Figure S4B and S4D ) . Again , fate-mapping analysis corroborated the astroglial origin of the reprogrammed cells ( Figure 12D–12F ) . These neurons derived from astroglia-derived neurosphere cells exhibited quite low input resistances similarly to the adherent astroglial cells reprogrammed by Neurog2 , therefore suggesting that they had reached a mature neuronal state ( Figure S2C; 751 . 9±118 . 8 MΩ at 15 . 0±1 . 4 DPI ( n = 10 ) versus 608 . 6±125 . 0 MΩ at 26 . 8±1 . 2 DPI ( n = 14 ) , Neurog2-reprogrammed neurosphere cells versus Neurog2-reprogrammed adherent astroglia , respectively ) . Immunostaining for vGluT1 revealed a massive labelling of vGluT1-positive puncta indicating that the Neurog2-transduced cells had acquired a glutamatergic identity ( Figure 12A–12C ) . Indeed , electrophysiological pair recordings unambiguously demonstrated the glutamatergic phenotype of the fate-mapped cortical astroglia expanded as neurospheres and reprogrammed by Neurog2 ( n = 5 ) ( Figure 12E–12G ) . In addition , step-depolarization of Neurog2-expressing neurons also evoked in several cases a sequence of polysynaptic components consistent with the development of excitatory networks in these cultures ( Figure S3G and S3H ) . Nine out of 21 DsRed-positive neurons recorded exhibited glutamatergic autaptic or synaptic connections ( age of the cells: 14 . 2±0 . 7 DPI; Figure S2A ) . In addition , calcium imaging experiments performed 3–4 wk post-infection in these Neurog2-reprogrammed cultures revealed a high degree of self-driven , synchronous excitatory network activity , which was blocked by CNQX and AP5 treatment ( 97 out of 98 DsRed-positive cells imaged; n = 3 independent experiments; Figure 13 ) . As can be appreciated from Figure 13 many of the neurons recruited to the self-driven networks were derived from fate-mapped GFP-positive neurosphere cells indicating their astroglial origin ( 28 out of 29 fate-mapped GFP/DsRed-double-positive cells imaged ) . These data suggest that astroglia initially expanded as neurosphere cells are more plastic in regard to their differentiation into various neuronal subtypes . What may be the molecular changes underlying the increased plasticity of neurosphere cells compared to adherent astroglia given their common astroglial origin ? The striking increase in the efficiency of reprogramming by Dlx2 could be accounted for by a loss in the expression of molecular cues associated with a glutamatergic bias . However , quantitative RT-PCR showed that there was no difference in the expression of Emx1 or Emx2 mRNAs [38] between astroglial cells cultured adherently or under neurosphere conditions ( Figure 10H and 10I ) . In contrast , astroglial cells cultured under neurosphere conditions expressed drastically higher levels of Sox2 mRNA compared to adherent astroglia ( Figure 10J ) . Given the high level of expression of Sox2 in neural stem cells [39] , these data are in agreement with the observation that culturing postnatal astroglia from the cerebral cortex under neurosphere conditions leads to their de-differentiation towards a more stem cell-like state [33] , [34] . The above finding that Neurog2 or Dlx2 overexpression can induce with high efficiency the generation of functional neurons from postnatal astroglia-derived neurosphere cells prompted us to examine whether reactive astroglia from the adult cortex following injury can also be reprogrammed to generate functional neurons after prior expansion as neurospheres . Indeed , previous work of our laboratory has shown that following a local injury such as a stab wound lesion , reactive astroglia isolated from the adult cerebral cortex de-differentiate , resume proliferation , and have the capacity to give rise to self-renewing neurospheres in vitro , in contrast to the intact contralateral cortex [16] . To examine the reprogramming potential of reactive astroglia isolated from the adult injured cerebral cortex , we performed local stab wound lesions in the right cortical hemisphere of adult C57BL/6J mice and dissociated both the control and injured cortical hemispheres 3 d later for subsequent neurosphere cultures . While control cortical tissue did not generate neurospheres , the injured hemisphere gave rise to neurospheres as reported [16] . Confirming previous results [16] , neurospheres could also be obtained from single GFP reporter-positive cells derived from either GLAST::CreERT2/Z/EG mice ( unpublished data ) or hGFAP-GFP mice ( Figure 14A–14A” ) . After 1–2 wk , single neurospheres were plated , subsequently transduced with retrovirus encoding Neurog2 and DsRed or Dlx2 and DsRed , and then allowed for differentiation . When astroglia-derived neurosphere cells obtained from lesioned cortex of wild type mice were transduced with Neurog2 , virtually all the DsRed-positive cells had developed into MAP2-positive neurons at 15 DPI ( >50 DsRed-positive cells per sphere; 25 spheres analysed , Figure 14C and 14D ) . Importantly , when GFP-positive neurospheres originating from the injured cortex of hGFAP-GFP mice ( Figure 14A–14B ) were transduced with Neurog2 , we observed the generation of numerous DsRed-positive neurons ( total number of neurons >50 , 10 spheres analysed ) . In contrast to untransduced lesion cortex neurosphere cells , GFP expression was lost following neuronal differentiation ( Figure 14B ) , consistent with the astroglia-to-neuron fate change . Following step-current injection , Neurog2-expressing neurosphere-derived cells responded with a train of repetitive APs demonstrating their neuronal nature ( n = 30; Figure 15A and 15A' ) . In addition , these neurons derived from adult glia exhibited relatively low input resistance values similar to Neurog2-transduced neurons derived from the postnatal cortical neurosphere cells ( 1 , 009 . 2±177 . 4 MΩ at 21 . 9±1 . 2 DPI ( n = 19 ) versus 751 . 9±118 . 8 MΩ at 15 . 0±1 . 4 DPI ( n = 10 ) , respectively ) . We next assessed whether these neurons could establish functional synaptic connections . Single adult cortex-derived neurospheres showed a massive immunostaining of vGluT1-positive puncta as shown 28 DPI , that outlined the dense network of intermingled MAP2-positive processes ( Figure 15B ) and the soma of Neurog2-transduced cells ( Figure 15C ) . Consistent with vGluT1 expression , immunostaining for the presynaptic protein synapsin also revealed a dense labelling of synapsin-positive puncta , thus suggesting the development of synaptic contacts between adult lesioned cortex-derived neurosphere cells ( Figure 15D ) . Furthermore electrophysiological recordings revealed the emergence of CNQX-sensitive spontaneous synaptic currents in these neurons in accordance with vGluT1 and synapsin expression ( 8 out of 30 cells recorded at 22 . 5±0 . 9 DPI; Figure 15E–15F” ) . These data indicate that adult astroglia-derived neurosphere cells transduced with Neurog2 mature into functional glutamatergic neurons . To examine the extent of plasticity of glial cells derived from the adult lesioned cortex we also tested as a proof-of-principle experiment whether adult lesioned cortex-derived neurosphere cells could also be directed by forced expression of Dlx2 towards MAP2-expressing neurons ( Figure S6A and S6A' ) . However , Dlx2 reprogrammed neurons were rather few and fragile due to their rather small soma size , thus hampering extensive electrophysiological analysis . Nevertheless we could record from one cell shown in Figure S6 , where step-depolarisation evoked an autaptic response that was blocked by bicuculline , indicating the development of functional GABAergic connections ( Figure S6B and S6B' ) . These data show that even adult cells isolated from the injured cortex and expanded as neurospheres can be instructed by forced expression of Neurog2 or Dlx2 to generate mature neurons able to establish functional glutamatergic or GABAergic connections , respectively .
We have previously shown that postnatal astroglia can be reprogrammed into neurons [6] , [7] . However , as the astroglial origin of the reprogrammed cells is a very important issue , here we sought to provide new experimental evidence via genetic fate-mapping of astroglia by using the GLAST::CreERT2/Z/EG mouse line developed in our laboratory [23] . To ensure the specificity of this mouse model we showed that virtually all fate-mapped , i . e . GFP-reporter positive , cells remain in the glial lineage under our culture conditions , with the vast majority being identified as astrocytes based on GFAP expression as well as GLAST ( unpublished data ) and a minor population as oligodendroglial cells , while none of the fate-mapped cells spontaneously gave rise to neurons ( Figure 3A and 3A' ) . These data support the notion that the cells cultured under these conditions do not possess an intrinsic neurogenic potential . This is consistent with the finding of virtually absent endogenous Neurog2 expression compared to cortical precursors isolated at the embryonic stage ( Figure 10G ) and in agreement with the epigenetic silencing of the neurogenin-1 and -2 loci at the transition between neurogenic and astrogliogenic precursors [40] . These data do not rule out the possibility that in vivo a small subset of astroglial cells can still give rise to neurons as suggested by hGFAP::CreERT2-mediated fate mapping showing that neurons can be generated at early postnatal stages from genetically marked cells [41] . However , so far it could not be experimentally distinguished whether the postnatal generated neurons indeed had been derived from astroglia local to the cerebral cortex or would be derived from astroglial stem cells in the subependymal zone that had subsequently immigrated into the cortex [42] . In any case our cultures do not sustain conditions for the genesis of neurons from astroglia ( even in the presence of EGF/FGF2 ) without forced expression of neurogenic fate determinants . Yet despite generating only glia when adherently grown in serum containing medium , postnatal astroglia exhibit a remarkable degree of plasticity as indicated by the fact that at least some can give rise to self-renewing , multipotent neurospheres ( [36] and present study ) . The latter fact could be taken as evidence that early postnatal astroglia possess stem cell character , particularly in the light of the notion that cells from the postnatal cerebral cortex may contribute to the pool of radial/astroglial stem cells within the dorsal adult subependymal zone [14] ( for review see [43] ) . However , several lines of evidence argue against a stem cell character of the astroglial cells studied here . Firstly , neither wild-type nor genetically fate-mapped astroglia spontaneously give rise to neurons , which is inconsistent with the stem cell defining hallmark of multipotency ( Figure 3A–3B” ) . Moreover , the large number of neurons generated following forced expression of Neurog2 argues against the possibility that the successfully reprogrammed cells derive from rare stem cells within this culture ( Video S1 ) . Secondly , while Dlx2 very efficiently directs adult neural stem cells in vitro towards neurogenesis [26] , the responsiveness of adherent astroglia is much more limited , suggesting a reduced susceptibility to Dlx2 transcriptional activity . The third line of evidence is based on the striking difference in Sox2 mRNA levels following expansion of the postnatal astroglia as neurosphere cells . Sox2 is a transcription factor well known to play a key role in neural stem cell self-renewal [39]; thus the massive up-regulation of Sox2 following exposure of astroglia to neurosphere culture conditions suggests that these cells undergo de-differentiation eventually acquiring indeed stem cell properties , while the comparatively lower levels of Sox2 in the adherent cultures would be in agreement with their non-stem cell character . Consistent with the higher degree of plasticity characterizing neural stem cells , the efficiency of reprogramming of astroglia-derived neurosphere cells by Dlx2 was found to reach levels comparable to bona fide neural stem cells ( Figure S4; [26] ) . Intriguingly , exposure of non-stem cell astroglia to EGF/FGF2 in the absence of serum factors may thus mimic similar extrinsic signals encountered by astroglial cells during postnatal development that later give rise to the stem cell compartment in the adult subependymal zone [44] . In fact the conversion of quiescent astrocytes from the adult cerebral cortex into stem cell-like cells following injury ( [16] and present study ) clearly demonstrates that cells apparently devoid of stem cell properties can acquire stem cell hallmarks such as self-renewal and multipotency when exposed to the appropriate environment . While our previous findings of eliciting neurogenesis by a single transcription factor from postnatal astroglial cells demonstrated the potency of these neurogenic fate determinants [6] , [7] , a major obstacle towards reprogramming into fully functional neurons was encountered in the failure of the astroglia-derived neurons to provide functional presynaptic output to other neurons [7] . In our previous study neuronal reprogramming was achieved by transcription factor-encoding retroviral vectors that exhibit relatively low levels of overexpression and are subject to substantial silencing in neurons [18] . Of note , long-term expression of an exogenous transcription factor appears to be required for maintaining a new phenotype following cellular reprogramming [45] . For instance , it has been shown that reprogramming of fibroblasts into macrophage-like cells remains unstable , resulting in the loss of macrophage markers , following silencing of retrovirally expressed PU . 1 and C/EBPα [46] . Here we demonstrated that stronger and more prolonged expression of neurogenic fate determinants from retroviral constructs more resistant to silencing [20] indeed permits more complete reprogramming of postnatal cortical astroglia towards synapse-forming neurons . In support of our hypothesis , we found that Dlx2 expression driven from a weaker and silencing-prone retroviral vector ( pMXIG ) [26] resulted in nearly negligible neurogenesis in postnatal astroglia cultures ( unpublished data ) , while the same fate determinant encoded by the pCAG vector induced substantial neurogenesis . Consistent with a more efficient reprogramming via a strong and silencing-resistant retroviral expression system , we found that forced expression of Neurog2 or Dlx2 endowed astroglia-derived neurons not only with the ability to receive synaptic input but also to form functional presynaptic output onto other astroglia-derived neurons to such degree as generating networks of spontaneously active neurons . Thus , one of the major obstacles that may impair incorporation of astroglia-derived neurons into a neuronal network , namely the inability to give rise to functional presynaptic output , can be overcome by appropriate expression of neurogenic fate determinants . Interestingly , neuronal reprogramming of astroglia by Neurog2 towards mature neurons appears to involve similar developmental steps as in newborn neurons during embryonic and adult neurogenesis . For instance , while GABA acts as an inhibitory neurotransmitter onto mature astroglia-derived neurons reprogrammed by Neurog2 , at a more immature stage these neurons respond with a rise in free intracellular calcium ( Blum et al . , submitted ) , very similar to immature embryonic- or adult-generated neurons [47] , [48] and such a differential response may be required for proper maturation to proceed [49] . Of note , despite its continued expression , Neurog2 , a transcription factor normally expressed in progenitors and only transiently maintained in postmitotic neurons [50] , does not seem to interfere with a surprisingly normal maturation of the reprogrammed cells . Firstly , input resistances of Neurog2-expressing neurons reach levels that correspond well to values observed for neurons recorded in slices from postnatal mice of roughly matching age [51] . Secondly , as a sign of proper morphological maturation dendrites of Neurog2-expressing neurons were covered with spines after 1 mo in culture , in agreement with progressive formation of excitatory synapses . Finally , astroglia-derived neurons reprogrammed by Neurog2 exhibit prominent expression of the Ca2+/Calmodulin-dependent kinase IIα subunit , which in vivo is exclusively expressed in excitatory neurons starting from the first postnatal week [22] . Moreover , expression of Ca2+/Calmodulin-dependent kinase IIα is a pre-requisite for the occurrence of changes in synaptic efficacy [52] , suggesting that Neurog2-reprogrammed cells potentially also acquire the ability to undergo synaptic modification . The fact that cells committed to the astroglial lineage can be reprogrammed into fully functional synapse-forming neurons also sheds some light on the more general question of whether the conversion across cell lineages induced by a single transcription factor can generate fully differentiated and stable cell fates that closely mirror cell types found in vivo [45] . Our study provides definitive positive evidence that such a cell lineage conversion is indeed possible and does not require passing through a pluripotent ground state . Interestingly , a recent study has shown that even mouse embryonic or perinatal fibroblasts , i . e . cells of the mesodermal lineage , can be converted by combined forced expression of three defined factors ( Mash1 , Brn2 , Myt1l ) into functional neurons [53] . Unexpectedly this combination seems to favour the generation of glutamatergic neurons , while the possibility to generate neurons of other phenotypes remains to be explored . Our study indeed reveals for the first time the feasibility of direct conversion of a somatic cell type into distinct neuronal subtypes by selective expression of transcription factors . Moreover , the efficiency of fibroblast conversion into neurons remains markedly lower ( ∼20% ) despite the joint use of three transcription factors . Conversely , we demonstrate here that cells of closer lineage-relationship to ectoderm-derived neurons , namely the astroglia , require only a single transcription factor ( Neurog2 ) to be converted towards fully functional neurons with 60% efficiency . This does not only confirm the notion that lineage reprogramming is achieved with best results by using the closest related cells but is also of profound relevance in regard to the eventual translational potential towards regenerative medicine . Activation of endogenous brain cells towards neuronal repair may be feasible if only one factor needs to be activated , e . g . by transvascular delivery of small molecules and RNAs [54] . In addition , the regional specification of astroglial cells characterized by distinct transcription factor profiles [55] , [56] may create a specific bias towards the generation of the type of neurons normally residing within the respective brain region and hence favour the fate conversion into the appropriate neuronal subtypes . Our results therefore further pave the way towards neuronal reprogramming from endogenous cells residing within the brain , circumventing the complications associated with transplantation . Intriguingly , although the majority of astroglial cells undergoing reprogramming by Neurog2 also undergo cell cycle division , single cell tracking demonstrated that astroglia can give rise to neurons without dividing . These data show that cell division is not a sine qua non condition for successful reprogramming , providing additional evidence for a direct fate conversion by-passing a proliferative state . Future studies will have to reveal whether even adult quiescent astrocytes could be reprogrammed into neurons without the requirement of entering the cell division cycle . Notably , given that expression of neurogenic fate determinants allows for the generation of synapse-forming neurons , we could assess whether distinct transcription factors promote neuronal subtype specification of the reprogrammed cells . Indeed , our immunocytochemical and electrophysiological analysis demonstrated that astroglia derived from the cerebral cortex can be reprogrammed not only towards the generation of glutamatergic neurons by Neurog2 , a fate determinant that regulates glutamatergic neuron generation in the developing dorsal telencephalon [8] , but also towards the generation of synapse-forming GABAergic neurons by Dlx2 . Of note , glutamatergic neurogenesis from cortical astroglia following reprogramming by Neurog2 was accompanied by the ( transient ) up-regulation of Tbr2 and Tbr1 , T-box transcription factors that characterize the genesis of glutamatergic neurons throughout the forebrain [21] . Moreover , Neurog2 induced the expression of the Ca2+/Calmodulin-dependent kinase IIα , which is selectively expressed in glutamatergic neurons of the forebrain [22] . The fact that Neurog2-reprogrammed astroglia derived from the cerebral cortex generate glutamatergic neurons rather than other neuronal subtypes which also require Neurog2 expression for their specification , such as spinal cord cholinergic motoneurons , is consistent with the notion that astroglial cells retain region-specific identity characterized by distinct transcription factor profiles [55] , [56] , in this case dorsal telencephalic cues [57] . In contrast to Neurog2 , Dlx2 is normally expressed in progenitor cells derived from the ventral telencephalon and has been shown to play a crucial role in the genesis of GABAergic neurons during development [13] and in adulthood [26] in this region . Yet the fact that astroglial cells of dorsal telencephalic origin can be forced to adopt a “ventral” fate is consistent with previous findings that ventral transcription factors can instruct a GABAergic fate in dorsally derived progenitors including induction of endogenous Dlx gene expression [12] , [58] . Of note , however , this cellular competence to respond appropriately to “wrong” regional transcriptional cues is not restricted to relatively unspecified precursors but can be even observed in cells committed to the astroglial lineage . However , the ability of Dlx2 to induce neuronal reprogramming from cortical astroglia is limited . In contrast to reprogramming induced by Neurog2 that occurred with very high efficiency , only a third of all Dlx2-transduced astroglia differentiated into neurons and most of these exhibited high input resistances typical of immature neurons even after prolonged periods of culturing . Moreover , the apparent impairment of maturation resulted in an even lower number of Dlx2-expressing neurons forming functional GABAergic synapses . Also only a minority of the neurons obtained from Dlx2-reprogrammed astroglia displayed more mature interneuronal firing patterns , with the majority of neurons often responding with one or few spikes to prolonged current injection . Intriguingly , however , among those neurons acquiring mature firing properties we could clearly discern distinct types of patterns which have been classified as regular spiking , stuttering , and low-threshold burst spiking . These data indicate that Dlx2-reprogrammed astroglia can eventually mature into specific interneuron subtypes . Consistent with this we found that a small subset of the reprogrammed cells expressed the calcium binding protein calretinin , which is normally expressed within the cortex in subpopulations of regular or low-threshold burst spiking interneurons [29] . The apparent limitations of Dlx2-mediated astroglia-to-neuron fate conversion could be due to inaccessibility of some downstream targets of Dlx2 in cortex-derived astroglia or to an overall lower potency of Dlx2 compared to Neurog2 in reprogramming and/or the need for additional co-factors missing in cortical astroglia . Along these lines we examined whether co-expression of Mash1 with Dlx2 may further promote the maturation and specification of astroglia into synapse-forming interneurons . Mash1 is a ventral telencephalic transcription factor upstream of Dlx2 in the cascade of interneuron specification during development [11] , [12] . Importantly , while being upstream of Dlx2 , which is a direct transcriptional target of Mash1 [30] , the latter factor is also known to activate targets that are not shared with Dlx2 , suggesting that complete interneuron specification may require the activity of both factors [31] , [32] . In agreement with this , we found that co-expression of Mash1 and Dlx2 promoted neurogenesis from astroglia in a synergistic manner to levels similar to Neurog2 ( Figure 9 ) . Moreover , input resistances were lower compared to those of Dlx2 only expressing neurons . Finally , the relative proportion of neurons exhibiting more mature interneuronal firing patterns was increased ( Figure 8 ) . However , there was no apparent enhancement of synapse formation following co-expression . This may point to the possibility that full interneuronal maturation requires extrinsic signals provided by their target cells , i . e . excitatory neurons , which are absent in cultures from Dlx2-reprogrammed astroglia . Alternatively , additional transcription factors such as Dlx5/Dlx6 may be required for complete maturation to occur . Given that different subtypes of GABAergic neurons are generated in the telencephalon ranging from medium spiny projection neurons of the striatum to various types of aspiny interneurons throughout the telencephalon [29] , it will be of great interest to develop strategies to further refine the subtype specification of GABAergic neurons generated upon forced Dlx2 or Mash1/Dlx2 expression , by using additional transcriptional cues in order to generate the full spectrum of GABAergic interneurons . Such directing of astroglia towards specific interneuron phenotypes may allow for the development of alternative approaches for the treatment of pharmaco-resistant epileptic disorders , particularly during early childhood [59] , [60] . Interestingly , culturing astroglial cells prior to transduction under neurosphere conditions improved the efficiency of reprogramming towards GABAergic neurons quite dramatically . Indeed , the synapse-forming capacity of Dlx2-expressing astroglia-derived neurosphere cells exceeded that of adherent astroglia-derived neurons expressing Dlx2 by a factor of seven . In addition , neurons derived from Dlx2-expressing neurosphere cells exhibited significantly lower input resistances consistent with a more mature state . These findings are not only of great importance in regard to eliciting the generation of neurons of different subtypes but may also support the concept that culturing cells under neurosphere conditions ( in serum-free medium containing high levels of EGF and FGF2 ) induces the erasure of region-specific transcription [33] , [34] . Indeed culturing neural cells of different origins under neurosphere conditions has been shown to induce a partial loss of region-specific transcription factor expression while resulting in the up-regulation of Olig2 and Mash1 [34] , providing a more permissive transcriptional environment that favours reprogramming towards distinct neuronal subtypes . However , we found no striking differences in the expression of the mRNAs encoding the cortical patterning factors Emx1 and Emx2 between cortical astroglia cultured adherently or expanded under neurosphere condition ( Figure 10H and 10I ) . In contrast , the two populations differed drastically in their expression levels of Sox2 mRNA ( Figure 10J ) , which may indicate a more stem cell-like status of the astroglia under neurosphere conditions and which may be the molecular correlate for their higher degree of plasticity . The successful reprogramming of astroglia following expansion under neurosphere conditions also encouraged us to investigate whether neurosphere cells derived from the adult cerebral cortex after injury can be similarly directed towards neurogenesis . Thus , we assessed here whether adult cortex-derived neurosphere cells also can be driven towards fully functional neurons following forced expression of Neurog2 or Dlx2 . Indeed , virtually all the cells expressing Neurog2 acquired a neuronal identity as revealed by MAP2 staining and their ability to fire APs . Moreover , we also provide evidence by vGluT1 immunoreactivity that the neurons derived from these lesion cortex-derived neurosphere cells undergo a subtype specification similar to reprogrammed postnatal astroglia and differentiate into glutamatergic neurons . Consistent with the development of functional synapses by Neurog2-reprogrammed adult cortex-derived neurosphere cells , we could record spontaneous glutamatergic events in these cultures . Conversely , following forced expression of Dlx2 , we could observe the generation of functional GABAergic neurons from adult cortex-derived neurosphere cells , indicating that the same dichotomy of subtype specification observed in postnatal astroglia also holds true for adult cortex-derived neurosphere cells following injury . While this first demonstration that functional neurons of different subtypes even undergoing synaptic connectivity can be derived in vitro from adult glial cells isolated from the injured cortex is an exciting step forward to utilize endogenous glial cells for repair of neurons [61]–[63] , it will still be a major challenge to translate these in vitro findings into the context of the injured brain .
Experiments were conducted either on C57BL/6J mice or transgenic GLAST::CreERT2/Z/EG double heterozygous mice . Briefly , heterozygous GLAST::CreERT2 mice [23] were crossed with the Z/EG reporter mouse line [24] to generate double heterozygous mutants . The activation of the tamoxifen-inducible form of the Cre recombinase was done as follows: From postnatal day 2 , i . e . at the peak of cortical astrogliogenesis [64] , until sacrifice at postnatal day 7 , tamoxifen ( 20 mg/mL , dissolved in corn oil , Sigma-Aldrich , Munich , Germany ) was administered as described by Mori et al . [23] to mothers and mice pups thus received tamoxifen via the milk from their mother during lactation . Using the same mouse line as in this study , Mori et al . found that by recombination already at E18 virtually all fate-mapped cells give rise to glia [23] , indicating that at that stage GLAST expressing cells have lost their intrinsic neurogenic potential . In some experiments , hGFAP-GFP transgenic mice were also used [65] . All animal procedures were carried out in accordance with the policies of the use of Animals and Humans in Neuroscience Research , revised and approved by the Society of Neuroscience and the state of Bavaria under licence number 55 . 2-1-54-2531-144/07 . All efforts were made to minimize animal suffering and to reduce the number of animals used . Retroviral transduction of astroglia cultured as adherent astroglia or expanded as neurosphere cells was performed 2–3 h after plating on coverslips , using VSV-G ( vesicular stomatitis virus glycoprotein ) -pseudotyped retroviruses encoding neurogenic fate determinants . Neurog2 or Dlx2 were expressed under control of an internal chicken β-actin promoter with cytomegalovirus enhancer ( pCAG ) together with DsRed located behind an internal ribosomal entry site ( IRES ) [20] . The Neurog2 coding cDNA was subcloned from the pCLIG-Neurog2 construct [67] into the EcoRI site of the pSKSP shuttle vector , from where it was then subcloned between the 5′SfiI and 3′PmeI restriction sites of the pCAG retroviral vector to generate pCAG-Neurog2-IRES-DsRed . The Dlx2 coding cDNA was subcloned from the pMXIG-Dlx2 construct [26] and inserted into the pCAG retroviral vector following the same cloning strategy to generate pCAG-Dlx2-IRES-DsRed . For control , cultures were transduced with a virus encoding only DsRed behind an IRES under control of the chicken β-actin promoter ( pCAG-IRES-DsRed ) . Viral particles were produced using gpg helperfree packaging cells to generate VSV-G ( vesicular stomatitis virus glycoprotein ) -pseudotyped viral particles [68] . Viral particles were titered by clonal analysis after transduction of E14 cortical cultures . Twenty-four hours after transduction , the medium of astroglia cultured as adherent astroglia was completely replaced by a differentiation medium consisting of DMEM/F12 , 3 . 5 mM glucose , penicillin/streptomycin , and B27 supplement , and the cells were allowed for differentiation for different time periods . Similarly , the medium of astroglia-derived neurosphere cells was replaced by a differentiation medium consisting of DMEM/F12 , 3 . 5 mM glucose , penicillin/streptomycin , supplemented with B27 , and buffered with HEPES . As it has been found that Brain-Derived Neurotrophic Factor ( BDNF ) is required for robust synapse formation of neurons derived from neural stem cells [69] , 20 ng/mL of BDNF ( Calbiochem ) were added to the cultures every fourth day during the differentiation period . Cells were cultured at a CO2 concentration of 9% , resulting in a pH of the differentiation medium of ∼7 . 2 . After 7–41 d following transduction , cells were used either for immunocytochemistry , electrophysiology , or calcium imaging experiments . The number of days after retroviral transduction is indicated as DPI . Transfection via DNA-liposome complexes was performed 2 h after plating of passaged cortical astroglia on poly-D-lysine coated 24-well tissue plates . DNA-liposome complexes were prepared in Optimem medium ( Invitrogen ) using the retroviral plasmid pCAG-Neurog2-IRES-DsRed or the control plasmid pCAG-IRES-DsRed and Lipofectamine 2000 ( Invitrogen ) as cationic liposome formulation . Astrocyte cultures were exposed to DNA-liposome complexes at a concentration of 0 . 5 µg DNA per 400 µL of Optimem medium for 4 h . Subsequently the medium was replaced by differentiation medium consisting of DMEM/F12 , 3 . 5 mM glucose , penicillin/streptomycin , and B27 supplement , and the 24-well tissue plates were placed into the time-lapse incubating chamber . Time-lapse video microscopy [70] , [71] of P7 cortical astrocyte cultures was performed with a cell observer ( Zeiss ) at a constant temperature of 37°C and 8% CO2 . Phase contrast images were acquired every 4 min and fluorescence images every 6–12 h for 6–8 d using a 20× phase contrast objective ( Zeiss ) and an AxioCamHRm camera with a self-written VBA module remote controlling Zeiss AxioVision 4 . 7 software [72] . Single-cell tracking was performed using a self-written computer program ( TTT ) [72] . Videos were assembled using Image J 1 . 42q ( National Institute of Health , USA ) software and are played at speed of 4 frames per second . Adult C57BL/6J mice of 8–10 wk of age ( 20–25 g ) were injured in the neocortex as described previously [61] . Briefly , mice received Rimadyl ( 4 mg/kg , s . c . , Carprofen ) as analgesic treatment and were anesthetized with ketamine ( 100 mg/kg , i . p . , Ketavet , GE Healthcare , Germany ) and xylazine ( 5 mg/kg , i . p . , Rompun , Bayer , Germany ) and placed in a stereotaxic frame in a flat skull position . After trepanation , a stab wound was made in the right cerebral sensorimotor cortex by using a sharp and thin scalpel ( Ophthalmic Corneal V-lance knife , Alcon , Germany ) at the following coordinates: anteroposterior ( AP ) = from −1 . 6 to −2 . 4 , mediolateral ( ML ) = −1 . 5 , dorsoventral ( DV ) = −0 . 6 mm with Bregma as reference . After surgery , mice were housed in individual Plexiglas cages with food and water ad libitum and kept in a 12 h light-dark cycle ( room temperature = 22±1°C ) . Three days after stab wound lesion , mice were killed by cervical dislocation following euthanasia in rising CO2 concentrations . After removal of the meninges grey matter tissue surrounding the stab wound injury of the neocortex and a similar piece at the same rostrocaudal level in the contralateral hemisphere were dissected and neurospheres were generated as described above . After 7–14 d , neurospheres generated from the adult injured cerebral cortex were collected and plated as single neurosphere on poly-D-lysine ( Sigma-Aldrich ) coated coverslips in 24-well plates ( BD Biosciences ) in a medium consisting of DMEM/F12 supplemented with B27 , EGF , FGF2 , penicillin/streptomycin , and buffered with HEPES . Two to 3 h after plating on coverslips , neurosphere cells were transduced as described above with retroviral vectors encoding Neurog2 ( pCAG-Neurog2-IRES-DsRed ) , Dlx2 ( pCAG-Dlx2-IRES-DsRed ) , or the control retrovirus ( pCAG-IRES-DsRed ) and were then allowed for differentiation . For immunocytochemistry , cultures were fixed in 4% paraformaldehyde ( PFA ) in phosphate buffered saline ( PBS ) for 15 min at room temperature . Cells were first pretreated in 0 . 5% Triton X-100 in PBS for 30 min , followed by incubation in 2% BSA and 0 . 5% Triton X-100 in PBS for 30 min . Primary antibodies were incubated on specimen overnight at 4°C in 2% BSA , 0 . 5% Triton X-100 in PBS . The following primary antibodies were used: anti-GFP ( GFP , chicken , 1∶1000 , Aves Labs , GFP-1020 ) , polyclonal anti-Glial Fibrillary Acidic Protein ( GFAP , rabbit , 1∶4000 , DakoCytomation , Z0334 ) , polyclonal anti-Red Fluorescent Protein ( RFP , rabbit , 1∶500 , Chemicon , AB3216 ) , polyclonal anti-RFP ( rabbit , Rockland , 1∶2000 , 600-401-379 ) , polyclonal anti-vesicular glutamate transporter 1 ( vGluT11 , rabbit , 1∶1000 , Synaptic Systems , 135302 ) , monoclonal anti-Microtubule Associated Protein 2 ( MAP2 , mouse IgG1 , 1∶200 , Sigma-Aldrich , M4403 ) , monoclonal anti-synapsin 1 ( mouse IgG2 , 1∶2000 , Synaptic Systems , 106001 ) , polyclonal anti-vGaT ( guinea pig , 1∶200 , Synaptic Systems , 131004 ) , polyclonal anti-Tbr1 ( rabbit , 1∶1000 , Millipore , AB9616 ) , polyclonal anti-Tbr2 ( rabbit , 1∶500 , Millipore , AB9618 ) , monoclonal anti-βIII tubulin ( mouse IgG2b , 1∶500 , Sigma , T8660 ) , polyclonal anti-GAD1 ( GAD67 ) ( rabbit , 1∶500 , Synaptic Systems , 198013 ) , monoclonal anti-calretinin ( mouse IgG1 , 1∶200 , Millipore , MAB1568 ) , and monoclonal anti-CaM kinase IIα ( mouse IgG1 , 1∶200 , Abcam , ab2725 ) . After extensive washing in PBS , cells were incubated with appropriate species- or subclass-specific secondary antibodies conjugated to Cy™2 , Cy™3 , Cy™5 ( 1∶500 , Jackson ImmunoResearch ) , Alexa Fluor 488 ( 1∶500 , Invitrogen ) , FITC ( fluorescein isothiocyanate , 1∶500 , Jackson ImmunoResearch ) , TRITC ( tetramethyl rhodamine isothiocyanate , 1∶500 , Jackson ImmunoResearch ) , or biotin ( 1∶500 , Jackson ImmunoResearch or Vector Laboratories ) for 2 h in the dark at room temperature , followed by extensive washing in PBS . Following treatment with secondary antibodies conjugated to biotin , cells were subsequently incubated for 2 h at room temperature with AMCA streptavidin ( 1∶200 , Vector Laboratories ) or Alexa Fluor 647 streptavidin ( 1∶500 , Invitrogen ) . Coverslips were finally mounted onto a glass slide with an anti-fading mounting medium ( Aqua Poly/Mount; Polysciences , Warrington , PA ) . Stainings were first examined with an epifluorescence microscope ( BX61 , Olympus , Hamburg , Germany ) equipped with the appropriate filter sets . Stainings were further analyzed with laser-scanning confocal microscopes ( SP5 , Leica , Wetzlar , Germany or LSM710 , Carl Zeiss , Göttingen , Germany ) . Z-stacks of digital images were captured using the LAS AF ( Leica ) or ZEN software ( Carl Zeiss ) . Single confocal images were then extracted from the Z-stacks . Alternatively , the Z-stacks were collapsed in one resulting picture using the maximum intensity projection function provided by the above mentioned softwares . Cell counts were performed by taking pictures of several randomly selected views per coverslip analysed by means of a Zeiss LSM 710 confocal microscope using a 25× objective . Subsequently , pictures were analysed for cell quantification using Image J 1 . 42q ( National Institute of Health , USA ) software . For each quantification , values are given as mean ± SEM . Cell counting data from reprogramming induced by Dlx2 , Mash2 , and Mash1 in combination with Dlx2 were subjected to a two-tailed Student's t test for statistical significance . Differences were considered statistically significant when the probability value was <0 . 05 . Electrophysiological properties of neurons derived from reprogrammed astroglial cells were analyzed 11–41 d following retroviral transduction . Single or dual perforated patch-clamp recordings [73] , [74] were performed at room temperature with amphotericin-B ( Calbiochem ) for perforation . Micropipettes were made from borosilicate glass capillaries ( Garner , Claremont , CA , USA ) . Pipettes were tip-filled with internal solution and back-filled with internal solution containing 200 µg/mL amphotericin-B . The electrodes had resistances of 2–2 . 5 MΩ . The internal solution contained 136 . 5 mM K-gluconate , 17 . 5 mM KCl , 9 mM NaCl , 1 mM MgCl2 , 10 mM HEPES , and 0 . 2 mM EGTA ( pH 7 . 4 ) at an osmolarity of 300 mOsm . The external solution contained 150 mM NaCl , 3 mM KCl , 3 mM CaCl2 , 2 mM MgCl2 , 10 mM HEPES , and 5 mM glucose ( pH 7 . 4 ) at an osmolarity of 310 mOsm . The recording chamber was continuously perfused with external solution at a rate of 0 . 5 mL/min . Cells were visualized with an epifluorescence microscope ( Axioskop2 , Carl Zeiss ) equipped with the appropriate filter sets . For patch clamp recordings , virally transduced cells were selected on the basis of their DsRed immunoreactivity . In addition , to ascertain the astroglial origin of the recorded neurons , DsRed- and GFP-expressing cells from GLAST::CreERT2/Z/EG animals were also selected for patch clamp recordings . Digital pictures of the recorded cells were acquired using a digital camera ( AxioCam , Carl Zeiss ) . Signals were sampled at 10 kHz with Axopatch 200B patch-clamp amplifiers ( Axon Instruments , Foster City , CA , USA ) , filtered at 5 kHz and analyzed with Clampfit 9 . 2 software ( Axon Instruments ) . For assessing a cell's ability to fire APs , cells received depolarizing step-current injections . AP amplitudes were measured by subtracting the threshold voltage of the AP from the AP maximum amplitude . For determining input resistance , hyperpolarizing currents of small amplitudes were injected into the cells under current clamp condition at a holding potential of −70 mV and input resistances were calculated from the corresponding voltage deviation . To examine spontaneous synaptic input into a given neuron , cells were kept in voltage clamp at a holding potential of −70 mV and synaptic events were recorded throughout a period of 1 to 5 min . In order to assess autaptic connections , single cells were step-depolarized in voltage clamp for 1 ms from −70 to +30 mV at a frequency of 0 . 05 Hz and responses were recorded in the same cell . Responses were considered to be autaptic when they occurred within 3 ms after the step-depolarization [75] . Synaptic connectivity was investigated by means of pair recordings in voltage clamp mode . One neuron was stimulated at low frequency ( 0 . 05–0 . 1 Hz ) by a 1 ms step-depolarization from −70 to +30 mV and the response was recorded from the other neuron , and vice versa . To determine the nature of the autaptic or synaptic responses , neurons were step-depolarized as described above and we assessed whether responses could be abolished in the presence of either the GABAA receptor antagonist bicuculline ( 10 µM ) or the AMPA/kainate receptor antagonist CNQX ( 5 µM ) . Finally , the recovery of the autaptic or synaptic response was assessed following washout of the pharmacological drugs . Neurons derived from reprogrammed astroglial cells were further analyzed 22 , 27 , and 29 d after retroviral transduction with calcium imaging experiments . A 5 mM stock solution of Oregon-Green BAPTA1 , AM ( KD: 170 nM , Invitrogen , O6807 ) was prepared in 8 . 9 µL 20% Pluronic F-127 ( Invitrogen ) in dimethylsulfoxide ( DMSO ) by means of a sonifier bath ( Bandelin , Berlin , Germany ) for 3 min . Reprogrammed astroglial cells were incubated with 5 µM Oregon-BAPTA1 , AM in artificial cerebrospinal fluid solution ( ACSF: in mM: 127 NaCl , 4 . 5 KCl , 2 . 5 NaH2PO4 , 2 CaCl2 , 2 MgCl2 , 23 NaHCO3 , and 25 D-glucose . , bubbled with 95% O2/5% CO2 . ) . The incubation of neurons with the dye-ACSF was performed in a cell culture incubator ( 37°C , 9% CO2 ) for a loading time of 10–15 min . Cells were washed with ACSF in a perfusion chamber ( Volume: 150–200 µL ) for at least 10–20 min with a flow rate of approximately 2–3 mL/min . Cells were imaged under continuous perfusion with ACSF solution at 26–30°C . Ligands ( Ascent scientific ) were bath-applied and used at the following concentrations: CNQX ( 10 µM ) , D-AP5 ( 10 µM ) , and tetrodotoxin ( TTX , 500 nM ) . For confocal Ca2+ imaging ( 256×256 pixels , 2 . 16 Hz ) , an inverted confocal microscope ( Olympus IX70 , equipped with a Fluoview 300 laser scanning system ) was used in combination with an Olympus , UPlanApo 20×/0 . 7 objective . Transduced cells were identified by means of DsRed expression . Oregon Green-derived fluorescence was excited with a 488 nm laser line ( emission filter: band pass 510/540 nm ) . DsRed was excited at 543 nm ( emission filter: band pass: 580 nm±40 nm ) . In some experiments , astroglia from GLAST::CreERT2/Z/EG mice , reprogrammed by forced expression of neurogenic fate determinants , were used for calcium imaging experiments . To ascertain the astroglial origin of the DsRed-positive reprogrammed cells that will be analyzed , GFP expression was first pictured at the confocal microscope and was carefully bleached by using the 488-laser afterwards . The DsRed-positive reprogrammed astroglial cells were subsequently incubated with Oregon Green and calcium imaging experiments were processed as described above . Images were analyzed using IMAGEJ software ( WS Rasband , IMAGEJ , US National Institutes of Health , Bethesda , Maryland , USA , http://rsb . info . nih . gov/ij/ , 1997–2006 ) . XY-time Calcium imaging results were analyzed by a region-of-interest analysis ( pixel intensity ) in the extended TIFF format as described before [76] . Total RNA was extracted with RNeasy Plus MicroKit ( Qiagen ) , according to the manufacturer's instructions . 1–1 . 5 µg of total RNA was retro-transcribed using Super-ScriptIII Reverse Transcriptase ( Invitrogen ) and random primers . Each cDNA was diluted one to ten , and 2 µl was used for each real-time reaction . mRNA quantitation was performed on a DNA Engine Opticon 2 System ( Bio-Rad ) following the manufacturer's protocol using the IQ SYBR Green SuperMix ( Bio-Rad ) . The following oligonucleotide primers were used for the qPCR: Gapdh , Emx1 , and Emx2 [34]; Sox2 [77] , Ngn2 [78] , βIIItub [79] , Glt1 [80] , Glu1 ( Glutamine Synthetase ) ( CCTGGACCCCAAGGCCCGTA; TGGCAGCCTGCACCATTCCAG ) , Aldh1l1 ( TGTTTGGCCAGGAGGTTTAC; AGGTCACCAGTGTCCAGACC ) , S100β ( GATGTCTTCCACCAGTACTCC; CTCATGTTCAAAGAACTCAT ) , and Gfap [81] . The amount of each gene was analyzed in triplicate , and the analysis was repeated on at least three independent samples ( n = 3 for NPC and postnatal neurospheres , n = 5 for postnatal adherent astrocytes ) . Data analysis was performed with the ΔΔCt method [82] . | The brain consists of two major cell types: neurons , which transmit information , and glial cells , which support and protect neurons . Interestingly , evidence suggests that some glial cells , including astroglia , can be directly converted into neurons by specific proteins , a transformation that may aid in the functional repair of damaged brain tissue . However , in order for the repaired brain areas to function properly , it is important that astroglia be directed into appropriate neuronal subclasses . In this study , we show that non-neurogenic astroglia from the cerebral cortex can be reprogrammed in vitro using just a single transcription factor to yield fully functional excitatory or inhibitory neurons . We achieved this result through forced expression of the same transcription factors that instruct the genesis of these distinct neuronal subtypes during embryonic forebrain development . Moreover we demonstrate that reactive astroglia isolated from the adult cortex after local injury can be reprogrammed into synapse-forming excitatory or inhibitory neurons following a similar strategy . Our findings provide evidence that endogenous glial cells may prove a promising strategy for replacing neurons that have degenerated due to trauma or disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/cell",
"differentiation",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration",
"neuroscience/neuronal",
"and",
"glial",
"cell",
"biology",
"developmental",
"biology/stem",
"cells"
] | 2010 | Directing Astroglia from the Cerebral Cortex into Subtype Specific Functional Neurons |
We previously reported interferon gamma secretion by human CD4+ and CD8+ T cells in response to recombinant E . coli-expressed Rv1860 protein of Mycobacterium tuberculosis ( MTB ) as well as protection of guinea pigs against a challenge with virulent MTB following prime-boost immunization with DNA vaccine and poxvirus expressing Rv1860 . In contrast , a Statens Serum Institute Mycobacterium bovis BCG ( BCG-SSI ) recombinant expressing MTB Rv1860 ( BCG-TB1860 ) showed loss of protective ability compared to the parent BCG strain expressing the control GFP protein ( BCG-GFP ) . Since Rv1860 is a secreted mannosylated protein of MTB and BCG , we investigated the effect of BCG-TB1860 on innate immunity . Relative to BCG-GFP , BCG-TB1860 effected a significant near total reduction both in secretion of cytokines IL-2 , IL-12p40 , IL-12p70 , TNF-α , IL-6 and IL-10 , and up regulation of co-stimulatory molecules MHC-II , CD40 , CD54 , CD80 and CD86 by infected bone marrow derived dendritic cells ( BMDC ) , while leaving secreted levels of TGF-β unchanged . These effects were mimicked by BCG-TB1860His which carried a 6-Histidine tag at the C-terminus of Rv1860 , killed sonicated preparations of BCG-TB1860 and purified H37Rv-derived Rv1860 glycoprotein added to BCG-GFP , but not by E . coli-expressed recombinant Rv1860 . Most importantly , BMDC exposed to BCG-TB1860 failed to polarize allogeneic as well as syngeneic T cells to secrete IFN-γ and IL-17 relative to BCG-GFP . Splenocytes from mice infected with BCG-SSI showed significantly less proliferation and secretion of IL-2 , IFN-γ and IL-17 , but secreted higher levels of IL-10 in response to in vitro restimulation with BCG-TB1860 compared to BCG-GFP . Spleens from mice infected with BCG-TB1860 also harboured significantly fewer DC expressing MHC-II , IL-12 , IL-2 and TNF-α compared to mice infected with BCG-GFP . Glycoproteins of MTB , through their deleterious effects on DC may thus contribute to suppress the generation of a TH1- and TH17-dominated adaptive immune response that is vital for protection against tuberculosis .
The scourge of tuberculosis which claimed close to a million non-HIV infected victims in 2011 worldwide [1] aided by multiple ( MDR ) and extremely drug resistant ( XDR ) strains [2] of the causative organism Mycobacterium tuberculosis ( MTB ) , has entrenched itself in the human population in its latent form and is undisputedly one of the most dreaded human bacterial diseases . MTB employs multiple mechanisms to interfere with both the innate and adaptive arms of the vertebrate immune system . These include inhibition of ( i ) phagolysozome fusion within antigen presenting cells [3] , ( ii ) maturation of human monocytes into DC [4] , ( iii ) dendritic cell migration to secondary lymphoid organs [5] as well as antigen processing and presentation to T cells [6] , [7] . In addition , MTB-infected macrophages , but not DC , prevented the development of a TH1-polarized T cell response [8] . The ability of the infected host to control infection by MTB depends on the capacity of the innate immune cells , primarily professional antigen-presenting cells such as DC and macrophages to prime an early and effective adaptive T cell response [9] , [10] . The presence of numerous pattern recognition receptors ( PRR ) on DC that are linked to intracellular signaling pathways allows these specialized cells to readily perceive invading pathogens and upregulate surface co-stimulatory molecules as well as secrete inflammatory and regulatory cytokines [11] , both of which have a crucial bearing on the subsequent development of T cell responses . It is therefore to be expected that a successful pathogen such as MTB would target this subset of cells to subvert the generation of effective host-protective immune responses . While the presence of complex lipid and carbohydrate moieties such as lipoarabinomannan , mycolic acids , phenolic glycolipids , peptidoglycan , phosphatidyl inositol mannosides etc . on the mycobacterial cell surface has been recognized for a very long time , awareness of the existence of glycosylated proteins in prokaryotic organisms has only come about over the last couple of decades . The pathogenic nature of several bacteria that possess glycosylated proteins , such as Mycobacterium and Clostridium species , Campylobacter jejuni , Treponema pallidum , Pseudomonas aeruginosa , Helicobacter pylori , Neisseria meningitis , N . gonorrhoea , and Streptococcus parasanguis ( reviewed in [12] ) suggests a role for these glycoproteins in mediating virulence and/or pathogenicity of these organisms . M . tuberculosis codes for at least forty one glycoproteins based on mass spectrometric characterization of concanavalin-A ( Con-A ) binding proteins [13] , [14] . The two secreted glycoproteins that have been well characterized , namely Rv1860 of MTB [15] , M . bovis and BCG [16] and MPB83 of M . bovis [17] carry one to three mannose residues linked to each other by α1→2 and 1→3 glycosidic bonds on threonine residues , respectively . The varied range of carbohydrate structures present in the MTB cell is dominated by the hexose mannose . Lipoarabinomannan ( LAM ) from MTB was originally credited with much of the subversion of host immunity effected by this pathogen , by binding to the carbohydrate receptors mannose receptor ( MR ) on human macrophages [18] and dendritic cell-specific intercellular adhesion molecule-3 grabbing non-integrin ( DC-SIGN ) on human dendritic cells ( DC ) [19] . Subsequent investigations pointed to significant contribution from other ligands carrying glycosylated structures from MTB [20] , [21] such as glycoproteins . Early reports relied primarily on Con-A-binding as proof of protein glycosylation [22] , [23] in mycobacteria . One of these glycoproteins , Rv1860 , also referred to as 45/47 kDa culture filtrate protein or APA owing to the presence of repeating units of alanine-proline-alanine motifs , was first identified as a proline-rich culture filtrate protein capable of eliciting both a delayed type hypersensitive ( DTH ) [24] and an antibody response [25] in guinea pigs immunized with live , but not killed Mycobacterium bovis BCG . The MTB homolog coding for a 50–55 kDa , 325 amino acid long Rv1860 protein [26] , was subsequently cloned and expressed both in M . smegmatis and E . coli [27] . Elegant analysis of the glycosylation moieties by proteolytic digestion of the purified 45 kDa culture filtrate-derived Rv1860 protein of MTB followed by mass spectrometry revealed the amino acid residues glycosylated to be threonines 10 , 18 , 27 and 277 and the attached carbohydrate to be single mannose , mannobiose , or mannotriose units strung together by α1→2 linkages [15] . Alteration of glycosylation of Rv1860 by expression in M . smegmatis as well as loss of glycosylation by enzymatic digestion or expression in E . coli resulted in reduced ability to elicit a DTH reaction in guinea pigs [16] , [28] . Both 45 and 47 kDa species had lost their 39 amino acid long N-terminal signal sequence; while the 45 kDa species carried predominantly a single mannose per molecule , the 47 kDa protein was dominated by 6 to 9 mannose residues per molecule [16] . We had initially identified Rv1860 as a target of antibody responses in sputum positive pulmonary TB patients [29] and demonstrated the ability of live , but not killed bacilli to elicit an antibody response in guinea pigs against several MTB proteins including Rv1860 [30] . We subsequently reported that recombinant E . coli-expressed Rv1860 was a robust in vitro stimulator of both CD4+ and CD8+ T cells from healthy PPD-positive volunteers [31] and that Rv1860 expressed in eukaryotic systems including a DNA vaccine vector and poxvirus protected guinea pigs against a challenge dose of virulent MTB . The contrasting loss of BCG's protective efficacy upon expression of Rv1860 from a genomically integrated copy of the MTB-derived gene in BCG-TB1860 suggested subversion of innate immune cells by native glycosylated Rv1860 . Since DCs possess multiple receptors to sense glycosyl moieties unique to pathogens , including mannose receptor , dectins and several C-type lectins including DC-SIGN , SIGNR1 to R4 and pulmonary surfactant proteins [32]–[35] , we investigated the effect of MTB Rv1860 on functions of primary murine bone marrow-derived dendritic cells ( BMDC ) .
All experiments with animals were carried out in strict accordance with the Institutional Animal Ethics Committee-approved protocols of the Indian Institute of Science for mouse experiments ( CAF/Ethics/220-2011 dated February 10 , 2011 ) and National Tuberculosis Institute for guinea pig experiments ( NTI-IAEC-2608-2610 dated 25 January , 2005 ) . All protocols adhered to the guidelines provided by the “Committee for the Purpose of Control and Supervision of Experiments on Animals” ( CPCSEA ) , a statutory body of the Government of India , Ministry of Social Justice and Empowerment . We constructed the plasmids pDK-Hyg-Rv1860 , pDK-Hyg-GFP and pDK-Hyg-Rv1860-6His ( Figure S1 in Text S1 ) expressing the Rv1860 gene of H37Rv , the control gene GFP and Rv1860 with a C-terminal 6× Histidine tag , respectively as described in Supplementary Methods in Text S1 . BCG-SSI-1331 from Statens Serum Institute was grown in Middlebrook 7H9 liquid medium ( Difco , Detroit , USA ) supplemented with 10% albumin/dextrose/catalase ( Difco ) , 0 . 2% glycerol , and 0 . 05% Tween 80 . Aliquots of mid log phase cultures were frozen at −85°C and viable bacteria were enumerated by plating serial dilutions on 7H10 agar plates supplemented with 10% oleic acid albumin dextrose catalase ( Difco ) and 0 . 2% glycerol ( 1 A600 nm = 109 colony forming units [cfu]/ml ) . To construct BCG strains expressing GFP ( BCG-GFP ) , and M . tuberculosis Rv1860 without and with the C-terminal 6× Histidine tag , ( BCG-TB1860 and BCG-TB1860His; Table 1 ) , M . bovis BCG-SSI was transformed with pDK-Hyg-GFP , pDK-Hyg-Rv1860 and pDK-Hyg-Rv1860-6His , respectively by electroporation at 2 . 5 kV , 200 µF and 1000 Ohms in a BIORAD Gene Pulser Xcell electroporator and selected on 7H10 plates supplemented with 75 µg/ml hygromycin . Single colonies were grown in 7H9 broth supplemented with 75 µg/ml hygromycin and the presence of the endogenous Rv1860 homolog as well as the integrated Rv1860 and GFP were verified by PCR amplification from genomic DNA as described in legend to Figure S2 in Text S1 . Expression from endogenous and inserted copies of Rv1860 protein in the BCG strains was quantitated by Western blotting lysates of the BCG strains with mouse anti-Rv1860 serum ( Supplementary Methods in Text S1 ) or mouse anti-6His antibody ( BD Biosciences ) , followed by chemiluminescence detection , with ribosome release factor serving as loading control . Western blotting of bacterial lysates obtained by sonication of cell pellets as described below , utilized 50 µg protein per lane of the polyacrylamide gel . For immunoprecipitation , lysates of BCG strains ( 400 µg protein ) and culture filtrates ( 10 ml from cultures with A600 nm = 0 . 7 ) were incubated with 10 µl of rabbit serum raised to recombinant Rv1860 protein ( described in Supplementary methods in Text S1 ) or 2 . 5 µg mouse monoclonal antibody specific to 6×Histidine ( Cell Biolabs ) coated on Protein A-Sepharose beads in binding buffer ( 50 mM Tris , pH 8 . 0 , 0 . 5 M NaCl , 0 . 1% deoxycholate , 0 . 1% sodium lauryl sulphate , 0 . 2% Nonidet P40 , and 0 . 05% Tween-20 ) overnight at 4°C . Beads were washed thrice with binding buffer and eluted by boiling with Laemmli sample buffer for 3 min . Binding and washing of BCG lysates and culture filtrates to Con-A-sepharose beads ( GE Healthcare ) was carried out in 20 mM Tris-HCl pH 7 . 4 , 0 . 5 M NaCl , 1 mM MnCl2 , 1 mM CaCl2 . Beads were eluted with 0 . 1 M borate buffer , pH 6 . 5 . Samples were electrophoresed on SDS-10% PAGE , transferred to nitrocellulose membranes and probed with appropriate mouse anti-Rv1860 serum , mouse monoclonal antibody specific to 6× Histidine ( BD Pharmingen clone F24-796 ) or Con-A-HRP as described in Figure legends followed by chemiluminescence detection . 8 to 10 week old female BALB/c mice were euthanized and femurs and tibia recovered by dissection . Following removal of muscle and connective tissues , bones were carefully cut at the ends to expose the marrow which was flushed out using plain Iscove's Modification of Dulbecco's Minimum Essential Medium ( IMDM ) . Bone marrow cells were cultured in IMDM-10% FBS containing 2 mM β-mercapto-ethanol essentially according to standard published protocols [8] with 50 U/ml murine rGM-CSF ( R&D Systems , Minneapolis , MN ) in the absence of antibiotics . 50% of the culture medium was removed and replenished every two days . Non-adherent cells were harvested on day 7 and analysed by flow cytomtery using antibodies against MHC-II , CD80 , CD86 , CD54 and CD40 to confirm that >90% of the cells represented immature dendritic cells which were then used for all experiments . Peritoneal macrophages were obtained from female BALB/c mice by washing the peritoneal cavity with IMDM . 2×105 cells were seeded per well of a 96 well plate . Adherent cells obtained following washing to remove the non-adherent population were infected at 1 multiplicity of infection ( MOI ) . DCs were infected with BCG strains mentioned above at an MOI of 5 for all experiments unless otherwise mentioned . MOI of one was used to assess survival of BCG strains within DC by colony enumeration of infected DC lysates . Mycobacterial cultures were concentrated to give 100 A600 nm units ( 1011 cfu/ml ) and sonicated for 10 min . using 30 sec pulses in a Branson Model 450 cup horn sonifier . Sonicates were plated on 7H10 agar plates to confirm absence of live cells and sonicate comparable to 5 MOI was added to BMDC cultures . Purified glycosylated native Rv1860 protein ( g1860 ) from M . tuberculosis ( NR-14862 ) obtained from Biodefense and Emerging Infections Resources , established by the National Institute of Allergy and Infectious Diseases ( NIAID ) and purified recombinant E . coli-expressed MTB Rv1860 ( Table 1 ) were each used at 10 ng per well of a 96 well plate , a concentration that was determined to be comparable to that delivered by infection at 5 MOI based on Western blotting serial dilutions of the purified proteins and BCG-TB1860 lysates with mouse anti-Rv1860 followed by quantitation of chemiluminescence signals . LPS from E . coli ( Sigma ) was used at 0 . 1 µg/ml . For in vitro recall responses of splenocytes , female BALB/c mice were infected intraperitoneally with 5×108 cfu of BCG-SSI . Three weeks later , spleens were harvested and used for in vitro stimulation assays described below . To study the effect of MTB-Rv1860 on spleen DC in vivo , we infected mice with 107 cfu of the two BCG strains BCG-GFP or BCG-TB1860 . Production of cytokines 1L12p40 , IL-2 and TNF-α as well as up regulation of MHC class II on spleen DC ex vivo was measured 12 hrs later by flow cytometry using the gating strategy shown in Figure S4 . For MLR studies , BMDC from eight weeks old female BALB/c mice were infected at an MOI of 5 with BCG strains for 24 hours , washed , gamma irradiated ( 15 Gray ) and cultured in triplicate wells with non-adherent splenocytes from female C57/Black6 mice for 72 hours at DC: splenocyte ratios of 1∶10 , 1∶20 and 1∶100 in 96 well culture dishes . For syngeneic polarization studies , the above mentioned infected BMDC were co-cultured with 5×105 splenocytes per well from BALB/c mice infected 3 weeks previously with 5×108 BCG-GFP or BCG-TB1860 , at DC: splenocyte ratio of 1∶20 for 72 hrs . Supernatants were collected for cytokine measurements and cells were pulsed with 0 . 5 µCi 3H-thymidine for 16 hrs . Cells were harvested in a Perkin Elmer Filtermate harvester and counted in a LKB Rack Beta liquid scintillation counter ( Model 1209 , LKB-Wallac , Turku , Finland ) . Con-A at 5 µg/ml was included as positive control . Irradiated DC alone incorporated between 40 and 90 cpm while T cells alone incorporated between 280 to 970 cpm . Con-A stimulation resulted in incorporation of 24 , 000 to 43 , 000 cpm in multiple experiments . To assess the effect of BCG strains on in vitro recall responses of immunized mice , splenocytes from mice immunized with 5×108 cfu of BCG-SSI three weeks before , were cultured at 5×105 cells per well in triplicate wells of a 96 well plate for 72 hrs in the presence of 20 MOI BCG-GFP or BCG-TB1860 . Supernatants were collected for cytokine measurements . Con-A at 5 µg/ml was included as positive control . Staining for surface markers on BMDC was done by resuspending up to 1×106 cells in 100 µl FACS buffer ( PBS supplemented with 1% heat-inactivated FBS , 0 . 1% NaN3 , and 1 mM EDTA ) containing combination of anti-CD11c FITC ( cloneN418 ) with one of the following antibodies , all from eBioscience: anti-CD40 PE ( 1C10 ) , anti-CD86 PE ( PO3 . 1 ) anti-CD80 PE ( 16-10A1 ) , anti-MHC class II PE ( M5/114 . 15 . 2 ) , anti-CCR7 PE ( 4B12 ) for 30 min at 4°C ( or at 37°C for CCR7 ) . Splenocytes from mice immunized twice 3 weeks apart with 5×108 BCG-SSI were obtained one week after the second immunization . 6×106 million splenocytes seeded in 3 ml per 35 mm dish were infected at an MOI of 20 for 36 hrs with BCG-GFP or BCG-TB1860 . 10 µg/ml brefeldin and 0 . 75 µM monensin were added for the final 26 hrs of culture . Harvested splenocytes were washed once with PBS-0 . 05% azide , permeabilized in 500 µl of PBS-0 . 05% N3-1% BSA with 0 . 1% saponin for 30 min . and intracellular IL-17 was detected using an antibody cocktail made up of CD3-FITC ( clone 145-2C-11 ) , CD4-PE ( clone GK1 . 5 ) , CD25-PE-Cy7 ( clone PC 61 . 5 ) from BD Biosciences and IL-17-APC ( clone eBio17B7 ) for 30 min on ice . IL-17 producing cells within the CD3-high , CD4-high and CD25-low population were enumerated . Detection of cytokines produced by mouse spleen DC ex vivo following 12 hr infection with 107 cfu of the BCG strains was carried out using 2×106 fresh splenocytes in the absence of in vitro re-stimulation as above using an antibody cocktail consisting of CD11c-fluorescein isothiocyanate ( FITC ) , MHC class II-PE/IL-12-phycoerythrin ( PE ) ( clone C17 . 8 ) , IL-2-PECy7 ( clone JES6-5H4 ) , anti-mouse TNF-α-phycoerythrin-cyanine-7 ( PECy7 ) ( clone MP6-XT22 ) F4/80-Alexa 647 ( clone BM8 ) , anti-mouse CD19-Alexa647 ( clone 1D3 ) , anti-mouse CD3-allophycocyanain ( APC ) ( clone 145-2C11 ) , anti-mouse Gr-1-APC ( clone RB6-8C5 ) obtained from BD Biosciences or eBiosciences . Cells were washed twice and fixed in 2% paraformaldehyde for 15 min . at 4°C . Data were acquired using a BD FACS Canto flow cytometer . The APC/Alexa647-high T and B cells , macrophages and neutrophils were gated out and cytokine production by CD11c-high DC within the APC-low population was enumerated . Data were analysed using FlowJo software ( Treestar ) . Outbred albino guinea pigs weighing 300 to 350 g bred at the National Tuberculosis Institute since 1979 and referred to as NTI-bred strain were used . Groups of eight animals were immunized intradermally with one of the following: ( i ) BCG , 106 CFU one time only ( ii ) BCG-Rv1860 , 106 CFU one time only . Animals were challenged via the intramuscular route in the thigh muscle 4 weeks after immunization with 2×105 viable MTB strain NTI64719 [29] . This challenge dose was sufficient to cause weight loss followed by death during weeks 7 to 8 post challenge in all control animals [36] . Animals were allowed free access to standard laboratory food and water for 6 weeks after which spleens were removed aseptically , homogenized and cultured for CFU of MTB as described [37] , [38] . Humane endpoints for euthanasia established by the institutional animal ethics committee ( NTI ) , which included being moribund or exceeding acceptable weight loss and/or being affected in their respiratory rate , were strictly followed . 7-day GM-CSF-differentiated DCs or mouse peritoneal macrophages were infected with BCG-GFP and BCG-TB1860 at an MOI of 1 for four hours . Following 3 washes to remove excess bacteria , remaining extracellular bacteria were killed by treatment with 50 µg/ml gentamicin for 1 hr . Infected DCs were lysed immediately or 72 h later , serially diluted , and colonies appearing on 7H10 agar plates were counted after 20 days . IFN-γ , IL-2 , IL-4 , IL-10 , IL-12p40 and p70 , TNF-α and TGF-β were quantitated by ELISA using commercially available antibody pairs ( Duo set , R&D Systems ) according to manufacturer's instructions . The lower limit of detection for all cytokines was 15 pg/ml . Comparisons between groups of guinea pigs were performed using t tests on log-transformed data . Statistical comparisons of cytokines , stimulation indices , surface marker expression on BMDC and cytokine-producing T cells were carried out using non-parametric Student's t test in GraphPad Prism for Windows ( GraphPad Software , San Diego California USA ) . P values less than 0 . 05 were considered statistically significant .
BCG-SSI ( 1331 ) obtained from the Statens Serum Institute , Copenhagen , Denmark was electroporated with plasmids pDK-Hyg-TB1860 , pDK-Hyg-Rv1860-6His and pDKHyg-GFP ( Figure S1 in Text S1 ) and selected on hygromycin agar plates to obtain BCG-TB1860 , BCG-TB1860His and BCG-GFP , respectively ( Table 1 ) . Genomic DNA from the parent and recombinant strains were analyzed by polymerase chain reaction using primers shown in Table 2 to amplify the endogenous Rv1860 gene as well as the integrated exogenous copies of Rv1860 and GFP genes as described in the Methods section . As expected , the parent BCG-SSI strain carried only the endogenous Rv1860 gene while BCG-TB1860 , BCG-TB1860His and BCG-GFP carried in addition , the exogenous Rv1860 , Rv1860-6×His and GFP genes , respectively ( Figure S2 in Text S1 ) . Western blotting with mouse serum specific to Rv1860 revealed a 1 . 84 fold and 2 . 65 fold increase in expression of Rv1860 in the BCG-TB1860 and BCG-TB1860His strains , respectively ( Figure 1 , A and B ) . To determine whether the recombinant Rv1860 protein expressed by BCG-TB1860 was mannosylated and secreted similar to the endogenous BCG homologue , we utilized BCG-TB1860His , carrying the C-terminally 6× Histidine-tagged Rv1860 gene . Western blotting with a mouse anti-6×Histidine antibody detected the 47 kDa 6×Histidine tagged protein in lysates from BCG-TB1860His but not in lysates from BCG-GFP ( Figure 1A ) . Immunoprecipitates of cell lysates and culture filtrate proteins from the BCG-GFP and BCG- TB1860His strains with rabbit antiserum to Rv1860 protein or with Con-A-sepharose , when analyzed by Western blotting , contained the 47 kDa protein detected by anti-6×Histidine antibody only in BCG-TB1860His ( Figure 1 , C and D , respectively ) . The 45 kDa form of Rv1860 , previously demonstrated to represent the mono-mannosylated species [16] , was consistently not detected by the anti-6×His antibody , confirming earlier reports that it represents a C-terminally processed Rv1860 protein which thereby lost the 6×Histidine tag [16] , [27] . Western blotting of rabbit anti-Rv1860 immunoprecipitates with Con-A-HRP ( Figure 1C , lower panel ) as well as inefficient precipitation of the 45 kDa species from lysates and culture filtrates with Con-A-Sepharose beads ( Figure 1D , upper panel ) , both revealed the paucity of mannosylation on the 45 kDa species as shown earlier by mass spectrometry [16] , despite its greater abundance in the lysates of both BCG-TB1860 and BCG-TB1860His as seen in Figure 1A . Specific immunoprecipitation of the 47 kDa mannosylated form of Rv1860 from lysates and culture filtrates of BCG-TB1860His by mouse anti-6× Histidine antibody ( Figure 1E , upper panel ) , which was demonstrated to be glycosylated based on Con-A-HRP reactivity ( Figure 1E , lower panel ) , corroborated these findings . These results showed that the recombinant 6×Histidine tagged protein was glycosylated and secreted similar to the endogenous protein . Additionally , comparison of upper and lower panels in Figure 1 C suggested that glycosylation is significantly higher in the secreted 47 kDa form compared to its counterpart in the lysate both in BCG-GFP and in BCG-TB1860His , in keeping with the translocation-mediated mycobacterial protein glycosylation model proposed by VanderVen et . al . [39] . In order to query the surface localization of Rv1860 protein , we carried out staining of the BCG strains with antibodies specific to Rv1860 and the 6×Histidine tag followed by flow cytometry . While the anti-Rv1860 serum detected a 2 . 2 fold increase in percentage of bacteria with positive surface staining for the protein in BCG-TB1860His relative to BCG-GFP ( 2 . 600±0 . 2082 N = 3% versus 1 . 200±0 . 1155%; Figure S3 in Text S1 ) the mouse Anti-6×Histidine-specific monoclonal antibody did not detect surface expression of the Rv1860-6His protein by flow cytometry ( data not shown ) . We were therefore unable to confirm the surface localization of Rv1860-6His . We also observed that both BCG-GFP and BCG-TB1860 multiplied at comparable levels when grown in liquid 7H9 broth ( Figure 1F ) . It was the impressive protective efficacy of a DNA vaccine and a poxvirus recombinant expressing Rv1860 of MTB [31] , that encouraged us to construct the above BCG recombinant expressing MTB Rv1860 from a copy of the gene inserted into the mycobacteriophage L5 integration site of BCG . As shown above , BCG carries in its genome the homologue of Rv1860 with a total length identical to that in MTB , of 325 amino acids but with a single amino acid difference at residue 140 which is phenylalanine in MTB and leucine in BCG . To our surprise , guinea pigs immunized with the recombinant BCG-TB1860 failed to control bacterial burden from a challenge dose of virulent MTB , with bacterial counts six weeks post challenge reaching those of unvaccinated controls ( Figure 2 ) . Thus , BCG expressing MTB Rv1860 sufferred loss of its protective efficacy . We then queried the mechanism by which the Rv1860 protein of MTB abrogates protective immunity elicited by BCG . Owing to the glycosylated nature of native Rv1860 , we surmised that this glycoprotein perhaps interfered with cells of the innate immune system . Indeed , compared to BMDC infected with BCG-GFP , levels of pro-inflammatory and regulatory cytokines including IL-2 , IL-6 , IL-12p40 , IL-12p70 , IL-10 , and TNF-α secreted by BMDC infected with BCG-TB1860 at 5 MOI , were significantly reduced ( Figure 3A ) , with reproducible reduction of 80 to 95% in secreted cytokine levels , while that of TGF-β , a well-established immune regulatory cytokine [40] , [41] was unaltered . BCG-TB1860His infection also resulted in similar reduction of IL-2 , IL-6 , IL-12p40 and TNF-α ( Figure 3A ) . The loss of IL-2 , IL-6 and TNF-α was nearly complete in several experiments despite variation among mice observed as reported by other workers also [42] . IL-12p40 showed the least inhibition , with BCG-TB1860-infected DC still secreting as much as 20 to 30% of the levels produced by DC infected with BCG-GFP ( Figure 3A ) . Interestingly , we observed reduction also in the low levels of IL-10 secreted by BMDC infected with BCG-TB1860 ( Fig . 3A ) compared to that induced by BCG-GFP , although this cytokine has been reported to have an antagonistic effect on IL-12 secretion by DC infected with MTB [8] , [43] . Sonicated preparations of the BCG strains added to BMDC at levels comparable to live infection at 5 MOI resulted in secretion of identical levels of cytokines ( Figure 3B ) , suggesting that MTB Rv1860 mediated inhibition of cytokine secretion by binding to a cell surface receptor . As expected therefore , addition of purified glycosylated Rv1860 ( g1860; Table 1 ) of MTB to BMDC cultures treated with live BCG-GFP or sonicates thereof also resulted in inhibition of cytokine secretion to levels observed following infection with live BCG-TB1860 ( Figure 3 , A and B ) , while g1860 added alone had no effect on BMDC , reminiscent of the lack of effect of LAM alone on human blood monocyte derived DC [19] . Non-glycosylated E . coli-expressed recombinant Rv1860 added alone or with live/sonicated BCG-GFP had no effect on BMDC ( data not shown ) . The ability of dendritic cells to prime an adaptive T cell immune response is critically dependent on the cognate interaction between the co-stimulatory molecules such as MHC-II , CD40 , CD80 and CD86 on the APC and the corresponding ligand on the T cell [44]–[46] . We therefore asked if expression of Rv1860 would adversely affect the capacity of BCG-infected DC to up regulate surface expression of co-stimulatory molecules . While BCG-GFP brought about a dramatic increase in the surface expression of CD40 , CD54 , CD80 , CD86 and MHC-II ( Figure 4A , blue line ) , infection of BMDC with BCG-TB1860 ( Fig . 4A , green line ) drastically reduced the expression of all five co-stimulatory molecules studied , to levels comparable to that found on uninfected DC ( Fig . 4A , red line ) . The reduction was found to be significant for all the markers analyzed ( Figure 4B ) . We did not observe significant variation in CCR7 expression on DC infected with these BCG strains ( data not shown ) . Robust response of BMDC to LPS was observed as expected ( Fig . 4A , brown dotted lines ) . Dendritic cells possess the unique ability to prime naïve T cells and activate them to mature into effector T cells [11] , mediated by secretion of inflammatory cytokines in combination with the cognate interaction with T cells through the upregulated surface molecules . Having observed the reduction in cytokine secretion as well as surface co-stimulatory molecule expression by DC infected with BCG-TB1860 , it was logical to ask if such DC showed impaired ability to polarize T cells towards a TH1 cytokine profile , a vital prerequisite for establishing protective immunity against MTB . In allogeneic MLR assays , BALB/c BMDC infected with BCG-GFP elicited significant proliferation of and IFN-γ secretion from C57Black6 splenocytes ( Figure 5 , A and B ) whereas similarly cultured DC infected with BCG-TB1860 failed to induce either proliferation ( Figure 5A ) or IFN-γ secretion ( Figure 5B ) from C57BL/6 splenocytes . Interestingly , we did not detect increase in IL-4 secretion from splenocytes cultured with BCG-TB1860 infected DC ( Figure 5C ) . This , coupled with the reduction in IL-10 secretion by DC infected with BCG-TB1860 , suggested that this glycoprotein functioned primarily to prevent the initiation of a TH1-polarized adaptive immune response , without causing TH2-polarization . In syngeneic polarization assays using splenocytes from mice immunized either with BCG-GFP or BCG-TB1860 also , BMDC infected with BCG-TB1860 inhibited the proliferation ( Figure 6A ) as well as secretion of IFN-γ ( Figure 6B ) and IL-17 ( Figure 6C ) induced by BCG-GFP-infected BMDC . The response of splenocytes from BCG-GFP-immunized mice was significantly higher than splenocytes from BCG-TB1860-immunized mice , to antigen presentation by DC infected in vitro with either BCG-GFP or BCG-TB1860 ( Figure 6 , A to C ) . We did not detect any changes in levels of IL-4 secreted by splenocytes exposed to the different DC populations ( data not shown ) . In addition , relative to BCG-GFP , BCG-TB1860 inhibited in vitro TH1 and TH17 recall responses of splenocytes from BCG-immunized mice . Secretion of IFN-γ , IL-2 and IL-17 by splenocytes from mice immunized with BCG-SSI and infected in vitro with BCG-TB1860 were significantly reduced relative to those infected with BCG-GFP ( Figure 7 , A , B and C ) while levels of IL-10 were significantly increased ( Figure 7D ) . Intracellular staining for IL-17 of splenocytes from BCG-immunized mice also revealed a significant reduction in IL-17 secreting CD3+CD4+CD25− T cells following in vitro stimulation with BCG-TB1860 relative to BCG-GFP ( Figure 7E ) . Although DC , unlike macrophages do not possess the ability to kill phagocytized bacteria , we asked if the ability to prevent BMDC activation was accompanied by a survival advantage within these two cell types in vitro for BCG-TB1860 . The bacterial load within BMDC and mouse peritoneal macrophages 3 days post infection with BCG-GFP or BCG-TB1860 were similar; both strains suffered a similar 4 to 5 fold reduction in viable bacteria 72 hrs post infection compared to that at 5 hrs post infection in BMDC ( Figure S4A in Text S1 ) . Mouse peritoneal macrophages infected at an MOI of 1 also supported similar 3 fold increase of the two strains over a period of 72 hrs . ( Figure S4B in Text S1 ) . We next asked if the in vitro effects of BCG-TB1860 infection on BMDC would be corroborated by similarly reduced DC functions in vivo in mice infected with BCG-TB1860 relative to BCG-GFP . A spurt of IL-12p40 production was reported by spleen dendritic cells as early as 5 hours following infection of mice with BCG [47] . We analyzed spleen dendritic cells ex vivo from mice infected for 12 hrs , for surface expression of MHC class II as well as synthesis of cytokines IL-12 , and IL-2/TNF-α by intracellular cytokine staining followed by flow cytometry using the gating strategy shown in Figure S5A in Text S1 . We detected significant reduction both in the numbers of DC up-regulating surface MHC-II as well as producing the above mentioned cytokines in splenic DC from mice infected 12 hrs previously with BCG-TB1860 compared to mice infected with BCG-GFP ( Figure 8 , A , B and C and Figure S5B in Text S1 ) . Significant reduction in DCs that simultaneously expressed a combination of either IL-2/TNF-α and IL-12 or IL-2/TNF-α and MHC-II ( Figure 8 , D and E and Figure S4C in Text S1 ) was also observed . We next wished to determine the identity of the pattern recognition receptor ( PRR ) on DC that bound Rv1860 as ligand . Blocking antibodies to the mouse mannose receptor , Dectin-1 and SIGNR1 added to BMDC prior to infection brought about no alleviation of the inhibition of BCG-GFP-induced cytokine secretion that we observed with BCG-TB1860 ( Figure S6 in Text S1 ) . In the same experiment , a TLR-4 blocking antibody resulted in approximately 50% reduction of cytokine secretion elicited by LPS ( data not shown ) , similar to the inhibition reported earlier [48] for this antibody . Control isotype antibody and pre-immune rabbit serum samples did not cause perturbation of the BCG-TB1860-induced inhibition of cytokine production , as expected . The non-availability of commercial blocking antibodies for murine DC-SIGN , the human homologue of which the Rv1860 protein has previously been reported to bind [21] , precluded addressing the role of this PRR in mediating the Rv1860 effect on mouse BMDC .
Immune protection against MTB is an extremely complex phenomenon where the multiple cells of the immune system along with their mediators work both in concert and at cross purposes , often succeeding and occasionally ( in 5 to 10% of individuals ) failing to achieve just the right balance required to prevent disease {reviewed in O'Garra , 2013 #183} . Subversion of cells comprising the innate immune system by MTB has been extensively documented . Macrophages , dendritic cells and neutrophils are all infected by MTB in the mouse model of aerosol infection [5] with resultant subversion of their functions . In this model , DC were demonstrated to be one of the earliest infected population , second only to neutrophils . Following intranasal inoculation of mice with as little as one million BCG , 5% of lung cells were infected by 24 hrs , of which 15% were DC [49] , corroborated by another study that demonstrated phagocytosis of BCG by almost 40% of lung DC by 24 hrs post intranasal infection [50] . This latter study also showed the superior production of IL-12 and TH1 polarization of CD4+ T cells by recruited DC over alveolar macrophages . Both these studies reproducibly implicated both alveolar macrophages and DC as the predominantly infected cells early in lung infection . Both in infected mouse lung and spleens , the subset of DC infected early were lymphoid [47] , [50] , replaced by myeloid DC post 2 weeks following infection [5] , [47] . DC , despite constituting a mere 6 . 8% of total lung tissue , were found to represent 50% of BCG-infected cells in lungs by 2 to 3 weeks post low dose aerosol infection [5] . All these studies point to the high propensity of tissue DC to be infected regardless of route of infection by the slow growing mycobacteria . Keeping in mind that immature DC are highly phagocytic , it is not surprising that both in vivo and in vitro , DC are efficiently infected , not to mention the numerous pathogen recognition receptors that they boast of , many of which such as DC-SIGN and mannose receptor , have already been shown to play a role in internalization of mycobacteria [5] , [51] , [52] . Mouse BMDC and BMDM cultures infected in vitro by MTB have also been extensively used to demonstrate loss or reduction of their normal functions by this pathogen [8] , [53] . Sections of infected human lung tissue revealed infection of DC as well as alveolar macrophages by MTB [52] , [54] and MTB infection resulted in a differential cytokine response from human monocyte-derived DC and macrophages [55] . The major receptor on human DC used by MTB to gain entry into these cells was demonstrated to be DC-SIGN [19] , [52] which recognized the abundant mannose capped lipoarabinomannan ( Man-LAM ) from the cell wall of MTB as the dominant ligand [19] . This binding resulted in secretion of IL-10 and inhibition of LPS- and M . bovis BCG-mediated maturation of DC . Later studies however revealed that human DC-SIGN binds to multiple ligands derived from MTB in addition to Man-LAM , including lipomannan and the mannose-containing 19 and 45 kDa glycosylated proteins [21] . Subsequent investigations using live mycobacteria in the place of purified LAM clearly revealed that LAM makes at best a minor contribution to DC-SIGN binding during infection of human DC by mycobacterial pathogens [20] . One report concluded , based on inhibition by mannan , that Man-LAM binds to the mannose receptor ( MR ) on human DC [56] , resulting in inhibition of IL-12 secretion . However , a later study that used blocking antibodies to MR and DC-SIGN revealed that ManLAM binds to DC-SIGN and not to MR on human DC [57] . Our results demonstrated sizeable loss of multiple BMDC functions both in vitro and in vivo , including cytokine secretion , up regulation of surface markers and downstream effects on TH1 and TH17 cytokine secretion by in vitro-polarized T cells along with significantly reduced cytokine secretion by mouse splenic dendritic cells recruited early , within 12 hrs of BCG infection , suggesting that Rv1860 ( along with other glycoproteins of MTB ) makes a measurable contribution towards suppressing DC functions very early following infection . That these early events ultimately lead to severely compromised protective immunity , was suggested by the dramatic increase in challenge MTB burden in the guinea pig vaccination model . We observed a dramatic reduction in the secretion of several cytokines including IL-2 , IL-10 , IL-12 p40 and p70 , and TNF-α by BCG-TB1860-infected BMDC . Early IL-2 secretion by DC was shown to be key for efficient T cell stimulation [58] , [59] and IL-2–deficient DCs when activated by bacteria were severely impaired in the ability to induce allogeneic CD8+ and CD4+ T-cell proliferation compared with wild-type ( WT ) DCs [58] . BCG-TB1860 infection extinguished IL-2 secretion by BMDC almost completely . The ability of DC to migrate to the draining lymph nodes following encounter with pathogens and subsequent initiation of the T cell response is governed by the cytokines they secrete soon after encounter with pathogens . IL-12 p40 levels have been shown to be critical for DC migration and TH1 T cell priming in MTB infection [55] , [60] , [61] . The importance of IL-12p70 in resistance to tuberculosis has been convincingly demonstrated [62] , [63] . Thus , the significant reduction in BMDC secretion of both IL-12p40 and IL-12p70 caused by Rv1860 poses a threat to effective initiation of a protective TH1-biased T cell immune response against TB . BMDC infected with BCG-GFP secreted low levels of IL-10 ( 50 to 100 pg/ml , similar to that reported earlier [53] following infection of 5 day cultured C57BL/6 with the Erdman strain of MTB ) , which was completely extinguished by BCG-TB1860 . IL-10 secreted at high concentrations by DC is also known to induce development of tolerogenic and regulatory T cells [64] . The failure of MTB-infected macrophages to secrete IL-12 has in turn been demonstrated to be mediated through IL-10 production , with both DCs and macrophages from IL-10 knockout mice secreting multiple fold more IL-12 p70 compared to wild type mice [8] . The loss of IL-10 from BMDC exposed to MTB Rv1860 suggests that modulation of this cytokine is not one of the mechanisms deployed by Rv1860 to prevent TH1 polarization of T cells . Expression of co-stimulatory molecules by DC is critical for their ability to deliver the second signal to T cells and activate both CD4+ and CD8+ T cells . CD80 ( B7-1 ) , CD54 ( ICAM-1 ) and LFA-3 were shown to function synergistically to boost proliferation , cytokine secretion and effector functions of T cells [65] , [66] while the regulated recruitment of CD86 to lipid rafts on DC was shown to be essential to achieve optimal T cell activation [67] . Ligation of CD40 on human DC was also shown to result in the dramatic up regulation of the surface co-stimulatory molecules CD80 and CD86 [68] and to serve as efficient stimulus for IL-12 production and subsequent TH1 polarization of T cells [69] . Adoptively transferred naïve antigen-specific CD4+ T cells lacking CD40-ligand failed to divide when the recipient mice were challenged with antigen [44] which could be attributed to their inability to engage CD40 on antigen-presenting DC and induce them to secrete IL-12 . CD40-CD40 ligand interaction is also vital for generating TH17 T cells [70] . It is therefore striking that Rv1860 of MTB was capable of suppressing a wide range of co-stimulatory molecule expression on DC to levels found on uninfected cells . These in vitro effects on BMDC and splenocytes were replicated during infection of mice by these BCG strains , as evidenced by the depressed activation of spleen DC at early time points after infection of mice . Surface expression of MHCII as well as secretion of IL-2 , IL-12 and TNF-α were all compromised when spleen DC were analyzed ex vivo . The results we report here have significant implications beyond innate immune responses , for the generation of a TH1- and TH17-biased adaptive cell mediated immune response that is vital for protection against TB [71] , [72] . In keeping with the significantly reduced secretion of inflammatory and regulatory cytokines and loss of surface co-stimulatory molecules by BCG-TB1860-infected BMDC , splenocytes stimulated with these DC suffered almost complete loss of ability to secrete IFN-γ , a cytokine that again has been shown to be vital for anti-TB immunity [73] , [74] as well as IL-2 and IL-17 . Rv1860 did not affect the levels of TGF-β secreted by BMDC while significantly reducing the IL-6 levels produced . The combination of these two cytokines is required to generate TH17 cells while TGF-β alone causes polarization of the common precursors to regulatory T cells ( Treg ) [75] , [76] . Selective loss of IL-6 without decrease in TGF-β as seen following BCG-TB1860 infection of BMDC would therefore polarize precursors to differentiate into Treg at the expense of TH17 . In fact , we did observe significant decrease in the percentage of IL17-secreting CD3+CD4+CD25− T cells and a significant increase in splenocyte secretion of IL-10 , a cytokine secreted by Treg cells , following infection with BCG-TB1860 , compared to BCG-GFP . IL-10 is elevated in human TB disease [77] and has been shown to prevent fibrotic granuloma formation and initiation of CD4+ T cells with a TH1 profile in the mouse model of TB {Cyktor , 2013 #161} . Splenocytes from BCG-TB1860-infected mice could not regain their TH1 and TH17 cytokine secretion to levels seen in BCG-GFP-immunized mice despite in vitro exposure to high MOI of BCG-GFP or to BCG-GFP-infected BMDC , revealing a long lasting deleterious effect of Rv1860 on the ability of dendritic cells to initiate a TH1 and TH17-dominated T cell response , reflecting the loss of TH1 and TH17 cells observed in human TB disease [72] . The guinea pig animal model provided convincing evidence that this subversion of DC function and subsequent TH1 T cell activation by Rv1860 had far reaching deleterious effects on protection afforded by BCG . BCG-immunized mice splenocytes stimulated in vitro with BCG-TB1860 would model the effect of MTB infection in BCG-vaccinated individuals and perhaps explain the recorded failure of BCG vaccination against adult pulmonary TB [78] in India and Africa . That MTB actively delays the initiation of adaptive immunity is well documented [79] , shown to be mediated through delayed migration of infected DC from the sites of infection to the draining lymph nodes [80] . Our results suggest that the impairment of DC functions reported in these earlier studies may have a contribution from glycoproteins such as Rv1860 through binding of DC receptors . The mycobacteriophage L5 integration locus within the glycyl tRNA ( glyV ) locus of BCG and MTB is the most widely used site for integrating foreign genes into these slow growing mycobacteria [81] . The Rv1860 protein expressed from the integrated copy of the MTB gene was expressed at similar levels to and was secreted and glycosylated , also similar to the native protein . Similar to previous recombinant strains of BCG carrying integrations , we too did not observe any alteration in growth of BCG-TB1860 relative to BCG-SSI or BCG-GFP . BCG-TB1860 brought about its suppressive effects on DC without exhibiting growth advantage , either in axenic cultures or within in vitro-infected BMDC and peritoneal macrophages . This was surprising but leaves open the possibility that BCG-TB1860 indeed survives better than BCG in infected animals in vivo , thanks to the presence of the Rv1860 protein . This crucial issue that we were unable to address in this study will remain the focus of our ongoing investigations to understand the mechanisms by which Rv1860 co-opts/subverts host innate immune cells for the benefit of the pathogen . The receptor on innate immune cells to which Rv1860 binds , to bring about the inhibition of DC functions that we report here awaits identification . Owing to non-availability of commercial blocking antibodies for murine DC-SIGN , we were unable to address this question in its entirety . Indeed , the 45 kDa glycoprotein Rv1860 has been reported to bind several C-type lectins on human innate immune cells , including DC-SIGN [21] and the pulmonary surfactant protein ( PSP ) -A [33] . Recent studies [13] , [14] have identified a vast array of mannosylated proteins in MTB . Whether alpha 1→2 linkages , which are unique to Rv1860 and are required for binding to DC-SIGN [82] , are shared by any of these other mycobacterial glycoproteins , and if so , their contribution to pathogenesis of this organism is worth investigating . M . bovis-BCG also carries the gene for and expresses the homologue of Rv1860 which differs from the MTB protein by a single amino acid at residue 136 which is phenylalanine in MTB and leucine in BCG . Despite conservation of the threonines that are glycosylated , the MTB protein is extremely potent at inhibiting DC functions , in contrast to the BCG homologue . The structural motifs that functionally distinguish the MTB Rv1860 from the BCG homologue are currently not known although they do not appear to reside within the glycosylation pattern , since the MTB Rv1860 in BCG-TB1860 is glycosylated by the BCG enzymatic machinery . Thus , the possible role of F136 in the MTB Rv1860 protein in affecting the folding of Rv1860 and its consequent binding to its cognate receptor on BMDC deserves investigation . The existence of numerous individuals living in TB-hyper endemic regions with no reactivity in the purified protein derivative ( PPD ) skin test despite continual exposure to MTB [83] attests to the remarkable ability of the innate immune response to control infection by MTB to a degree that prevented T cell activation . In light of the recent demonstration that BCG immunization results in long lasting beneficial epigenetic alterations in monocytes [84] , it would be worthwhile to investigate the effect of Rv1860 on chromatin remodeling in cells of the innate immune system . The suppressive effect of Rv1860 on BMDC would point to the exciting possibility of developing efficacious vaccines against TB , based on removal of genes encoding such immune suppressors from the genome of BCG/MTB . | Tuberculosis ( TB ) , although recognized as an infectious disease for centuries , is still the leading cause of human deaths , claiming a million lives annually . Successful control of TB , either through drugs or effective preventive vaccines has not been achieved despite decades of research . We have studied the role for mannosylated protein Rv1860 of MTB in interfering with the early response of dendritic cells , which belong to the host's innate immune arsenal , to this mycobacterium . We were able to show that incorporating the gene coding for Rv1860 of MTB into the safe vaccine strain BCG resulted in loss of BCG's protective ability in the guinea pig animal model . Using primary mouse bone marrow derived dendritic cells in vitro as well as spleen dendritic cells from infected mice , we show in this study that exposure to mannosylated Rv1860 leads to loss of dendritic cell functions such as cytokine secretion and T cell activation . This leads to defective downstream T cell responses to the mycobacteria . We suggest that altering or extinguishing the expression of such glycoproteins by mycobacteria may be a strategy for developing better vaccines against TB . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"host-pathogen",
"interactions",
"medical",
"microbiology",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"immunology",
"microbiology",
"pathogenesis",
"bacterial",
"p... | 2014 | The Glycosylated Rv1860 Protein of Mycobacterium tuberculosis Inhibits Dendritic Cell Mediated TH1 and TH17 Polarization of T Cells and Abrogates Protective Immunity Conferred by BCG |
Negative frequency-dependent selection ( NFDS ) is an evolutionary mechanism suggested to govern host-parasite coevolution and the maintenance of genetic diversity at host resistance loci , such as the vertebrate MHC and R-genes in plants . Matching-allele interactions of hosts and parasites that prevent the emergence of host and parasite genotypes that are universally resistant and infective are a genetic mechanism predicted to underpin NFDS . The underlying genetics of matching-allele interactions are unknown even in host-parasite systems with empirical support for coevolution by NFDS , as is the case for the planktonic crustacean Daphnia magna and the bacterial pathogen Pasteuria ramosa . We fine-map one locus associated with D . magna resistance to P . ramosa and genetically characterize two haplotypes of the Pasteuria resistance ( PR- ) locus using de novo genome and transcriptome sequencing . Sequence comparison of PR-locus haplotypes finds dramatic structural polymorphisms between PR-locus haplotypes including a large portion of each haplotype being composed of non-homologous sequences resulting in haplotypes differing in size by 66 kb . The high divergence of PR-locus haplotypes suggest a history of multiple , diverse and repeated instances of structural mutation events and restricted recombination . Annotation of the haplotypes reveals striking differences in gene content . In particular , a group of glycosyltransferase genes that is present in the susceptible but absent in the resistant haplotype . Moreover , in natural populations , we find that the PR-locus polymorphism is associated with variation in resistance to different P . ramosa genotypes , pointing to the PR-locus polymorphism as being responsible for the matching-allele interactions that have been previously described for this system . Our results conclusively identify a genetic basis for the matching-allele interaction observed in a coevolving host-parasite system and provide a first insight into its molecular basis .
Host-parasite interactions are ubiquitous among all living organisms and are thought to represent one of the strongest contributing factors to shaping the evolution of biological organisms [1] . The antagonistic nature of host-parasite interactions leads to reciprocal selection of the antagonists on each other that can drive rapid coevolutionary change [1–3] . Hosts are expected to evolve mechanisms to reduce the likelihood of infection and to minimize the fitness costs associated with infections , while parasites are expected to evolve mechanisms to evade the hosts’ defense mechanisms . Host-parasite interactions are thought to contribute to diversification , speciation , maintenance of sexual reproduction , and maintenance of genetic diversity in natural populations [1 , 4–6] . Multiple evolutionary mechanisms have been proposed to underlie host-parasite evolutionary dynamics . These include heterozygote advantage , selective sweeps , and negative frequency-dependent selection ( NFDS ) [2 , 7–9] . NFDS , whereby common host genotypes have a selective disadvantage , can result in balancing selection and is therefore proposed to contribute to the maintenance of genetic diversity in natural populations . The selective disadvantage for common host genotypes comes about because parasites are expected to adapt to these common genotypes [10 , 11] . Signatures of balancing selection have been found in gene families associated with disease resistance in vertebrates ( the Major Histocompatibility Complex , MHC ) and plants ( R-gene ) [12 , 13] . An assumption underlying this form of coevolution is that no parasite can infect all host types and no host can resist all parasite types . The matching-allele-model is one of the genetic mechanisms suggested to prevent the rise of such super-genotypes and thus contributing to the maintenance of genetic diversity [10 , 11 , 14] . However , despite of in-depth knowledge of the molecular structure of immune-related loci , the genetics underlying the interactions between hosts and parasites have not yet been resolved [15–17] . The Daphnia–Pasteuria system is a model for studies in host-parasite coevolution . Pasteuria ramosa is an obligate bacterial pathogen of the crustacean Daphnia magna that causes strong disease phenotypes with major fitness consequences for the host [8] . In short , feeding hosts pick up dormant P . ramosa spores . Contact with the host results in the activation of spores , which then attach to the hosts’ foregut . If attachment is successful , the spores penetrate into the D . magna body cavity initiating infection and disease . P . ramosa eventually kills the host and its spores are then released into the environment [18] . Importantly , spore attachment is genetically determined and fully consistent with infection success , i . e . resistant host genotypes prevent spore attachment whereas attachment is successful in susceptible host genotypes [19–22] . Here we use the terms resistance and susceptibility to refer to both spore attachment and overall infection . In this host-parasite system fluctuating selection in natural populations have been observed [23] and the D . magna—P . ramosa interactions follow a matching-allele model with no universally resistant host genotype being found [20–22 , 24] . Thus , the Daphnia-Pasteuria host-parasite system fulfils the core assumptions of models for coevolution by NFDS [10 , 11 , 14] , making it a promising model to explore the underlying genetic mechanisms of host-parasite interactions . We aimed to investigate the molecular genetic basis of this host-pathogen system and to gain insight into the genetic basis of coevolution by NFDS . Using a Quantitative Trait Locus ( QTL ) approach on a D . magna F2 recombinant panel , one large effect QTL associated with resistance to infection by the P . ramosa C19 genotype was detected [25] . The F2 recombinant panel showed Mendelian segregation of approximately 75% resistant and 25% susceptible genotypes . We build upon this work to explore and characterize the Pasteuria Resistance locus ( PR-locus ) in D . magna . We show that the PR-locus is highly polymorphic with striking structural genetic polymorphisms and , additionally , gene content and gene expression divergence in the PR-locus between resistant and susceptible haplotypes . The most striking aspect of these differences in gene content is related to a cluster of glycosyltransferase genes located within the PR-locus . Finally , we show that genetic variation at the PR-locus explains variation in resistance to spore attachment observed in natural D . magna populations following the predictions of a matching-allele model .
Routtu and Ebert ( 2015 ) detected one major effect QTL underlying D . magna resistance to infection by the P . ramosa C19 genotype located within a scaffold of approx . 2 . 3 Mb of the D . magna draft genome 2 . 4 ( Fig 1A ) [25] . We reduced the interval of the D . magna resistance locus and fine-mapped the QTL interval using microsatellites and SNP markers to find recombination breakpoints within the QTL interval ( S1 File ) . Microsatellite marker P34 and SNP g311b ( S1 Table ) defined the closest recombination breakpoints at positions 1369860 and 1506194 of scaffold00944 in the D . magna genome draft 2 . 4 , leaving a mapping interval of approximately 130 kb that we call here the PR-locus ( Fig 1B ) . Within this region no further recombination event was detectable among 360 F2 clones . Interestingly , we detected a genomic region of approximately 50 kb within the interval map where none of the designed genetic markers ( g294 and g350 ) could be amplified in the resistant parental D . magna clone Iinb1 , while the genetic markers placed outside this region ( g292 and g351 ) did amplify in both parent clones ( Fig 1C ) . As genetic markers were designed to match the D . magna Xinb3 based draft genome ( D . magna 2 . 4 ) , this result could be explained by structural polymorphism—a single indel polymorphism where the entire 50 kb region is absent in D . magna Iinb1 genotype or by a genomic region of such high sequence divergence between haplotypes that all the primer pairs based on D . magna Xinb3 clone would not produce an amplicon with D . magna Iinb1 DNA . In order to understand the polymorphism between the parental genotypes we applied high-throughput sequencing and long-read PacBio sequencing of both parental clones with the goal to improve the existing assembly of PR-locus in the D . magna Xinb3 clone and to obtain an independent de novo assembly of the same region in the Iinb1 clone . We obtained two complete haplotypes from the D . magna clones Xinb3 and Iinb1 for the PR-locus that correspond to the interval between positions 1366653 and 1520041 of the scaffold00944 in draft genome 2 . 4 and call them the xPR-locus and iPR-locus , respectively . The most striking feature found was that each haplotype contains a large genomic region where little homology was found corresponding to the region where we had previously found amplicon presence/absence polymorphism ( Fig 1C ) . We call this the Non-Homologous Region ( NHR ) , and the haplotypes we obtained from clones Xinb3 and Iinb1 are called xNHR and iNHR , respectively ( Fig 2 ) . xPR-locus and iPR-locus differ in their nucleotide lengths: xPR-locus is 159 kb long while iPR-locus is 215 kb long . In addition , considering the entire PR-locus haplotypes 34% of xPR-locus and 46% of iPR-locus have no homology to each other ( Fig 2 ) ( S2 Table ) . However , these differences in length and lack of homology are unevenly distributed across PR-locus . It is the NHR that differs substantially in length: iNHR ( from the Iinb1 clone ) was 121 kb in length , in contrast to xNHR ( Xinb3 clone ) with only 55 kb ( Figs 2 and 3 ) . The two NHR haplotypes contain only few fragments with homologous sequences: in iNHR a total of 25 kb had a significant alignment in xNHR , representing only 20% of the total sequence; in xNHR only 13 . 7 kb could be homologized to iNHR ( Figs 2 and 3 ) ( S2 Table ) . This region of non-homology at the NHR contrasts to high homology ( >90% ) at the flanking regions of the NHR , i . e . in the remainder of the PR-locus ( Figs 2 and 3 ) ( S2 Table ) . A large proportion of both PR-locus haplotypes was composed of repeated sequences . We divide the repeated sequences in two groups according to the location of their copies: sequences that are repeated in the host genome but outside PR-locus–extra-locus repeats; and sequences that were repeated within PR-locus–intra-locus repeats . A large proportion of both PR-locus haplotypes sequences were made of extra-locus repeats . In spite of the differences observed in length between xPR-locus and iPR-locus haplotypes , both had approx . 25% of their total sequence composed of these extra-locus repeats representing 54 . 7 kb and 39 . 9 kb , respectively ( Fig 2 ) ( S3 Table ) . Looking into the distribution of extra-locus repeats we observed that they were unevenly distributed as the NHR contains by far the largest proportion of these extra-locus repeat elements , representing 33% of iNHR and 38% of xNHR ( Fig 2 ) ( S3 Table ) . In addition , the remaining extra-locus repeats found outside the NHR were concentrated in a 20 kb region immediately upstream of NHR ( Fig 2 ) ( S3 Table ) . Interestingly , extra-locus repeats accounted for a significant proportion of sequences non-homologous between PR-locus haplotypes . Specifically , 53% of the non-homologous iPR-locus sequences and 51% of the non-homologous xPR-locus are extra-locus repeats . Second , iPR-locus and xPR-locus diverged in number and nature of intra-locus repeats . In xPR-locus , we detected 14 intra-locus repeats , covering 17 . 3 kb or 11% of the sequence total ( Fig 2 ) ( S4 Table ) . In contrast , in iPR-locus haplotype we detected 30 intra-locus repeats , representing 68 kb and nearly 32% of the total sequence ( Fig 2 ) ( S4 Table ) . Most of these intra-locus repeats were located within the NHR , specifically 97% and 67% of the intra-locus repeat sequence in iNHR and xNHR , respectively ( Fig 2 ) ( S4 Table ) . In summary , PR-locus is characterized by dramatic structural polymorphism that in its overwhelming majority is contained within a defined genomic region , the NHR . In particular a large proportion of PR-locus sequences here investigated are non-homologous between the resistant and susceptible haplotypes; a large proportion of both PR-locus haplotypes was composed of repeat elements; the repeat sequences could be repeated extra-locus , intra-locus or both; a large part of the sequence that was non-homologous between the PR-locus haplotypes was composed of extra-locus and/or intra-locus repeats; PR-locus haplotypes diverged in their sequence nucleotide length and in the number and nature of both extra and intra-locus repeats ( Figs 2 and 3 ) . The NHR , where most of the variation described here is found , is therefore a strong candidate to harbor variation underlying D . magna resistance to P . ramosa . We annotated the expressed genes in each PR-locus haplotype . Orsini et al . ( 2016 ) produced an RNAseq database for D . magna Xinb3 and Iinb1 clones investigated in this article , as well as for D . magna F1 lineage resulting from a cross between D . magna Xinb3 and Iinb1 clones [26] . This D . magna ( Xinb3 x Iinb1 ) F1 clone was in turn used to generate the F2 recombinant panel genotypes used for QTL mapping [27] . In addition to control conditions , the Orsini et al . ( 2016 ) study also investigated gene expression in the same genotypes when exposed to multiple environmental stress factors , including exposure to spores of P . ramosa [26] . Using this resource we produced a de novo transcriptome and carried out reciprocal blasts between this transcript database and the PR-locus haplotype sequences that we generated from D . magna Xinb3 and Iinb1 genotypes in order to find which expressed transcripts map to each of the PR-locus haplotypes . We annotated a total of 83 expressed genes that map to the PR-locus haplotypes . Of these , 20 mapped exclusively to the iPR-locus and 18 exclusively to the xPR-locus , whereas 45 annotated expressed transcripts mapped to both haplotypes ( S5 Table ) . The 20 annotated genes that mapped only to the iPR-locus represented one sulfoquinovosyltransferase , and 19 uncharacterized proteins ( UP ) ( S5 Table ) . The 18 annotated genes that mapped only to the xPR-locus represented five fucosyltransferases , one alpha 1 , 4-glycosyltransferase , one PC-Esterase and 11 UPs ( S5 Table ) . These observations revealed that the differences in gene content between PR-locus haplotypes resulted for the most part from differences in the representation of fucosyltransferases and UPs . Importantly , all the genes that were exclusive of one or another haplotype , mapped entirely to the NHR region at the center of the PR-locus with the exception of one fucosyltransferase mapping to xPR-locus . This result is consistent with the lack of homology between haplotypes at the NHR . Finally , the 45 expressed transcripts that were shared between the PR-locus haplotypes represented four PC-Esterases , two fucosyltransferases , one methyltransferase , one alpha 1 , 4-glycosyltransferase , one galactosyltransferase , one sestrin , one DNA mismatch-repair protein , one zinc-finger binding domain , one glutamate synthase , one calcipressin , one spermidine synthase , one acyl-CoA Thioesterase and 29 UPs ( S5 Table ) . We investigated differences in expression of genes shared between clones Xinb3 ( susceptible to P . ramosa C19 ) and Iinb1 ( resistant to P . ramosa C19 ) . Among the 45 transcripts resulting in annotated genes that mapped to both PR-locus haplotypes , 20 were differentially expressed between Xinb3 and Iinb1 clones ( S6 Table ) . Using the Xinb3 clone ( the chosen clone for the 2 . 4 D . magna draft genome ) as the focal genotype we identified 11 upregulated and nine downregulated expressed transcripts ( S6 Table ) . The 11 transcripts upregulated in the Xinb3 clone represented one methyltransferase , one fucosyltransferase , one DNA mismatch-repair protein , one PC-esterase and seven UPs ( S6 Table ) . The nine transcripts downregulated in the Xinb3 clone represented one calcipressin , one DNA mismatch-repair protein , one fucosyltransferase , one sestrin and five UPs ( S6 Table ) . In order to narrow down the number of candidate genes in the PR-locus haplotypes , we compared expression of transcripts mapping to the PR-locus haplotypes between the Xinb3 and Iinb1 clones and the hybrid F1 ( Xinb3 x Iinb1 ) clone . The hybrid F1 clone was resistant to the P . ramosa C19 genotype just as the Iinb1 clone and in contrast to the Xinb3 clone . Thus , we searched for those transcripts that were consistently down- or upregulated in the Xinb3 clone in comparison to both of the Iinb1 and F1 clones , as those represented the best candidates to underlay the variation in resistance to P . ramosa observed in the previous QTL study [25] . Only one transcript of calcipressin was downregulated in the Xinb3 clone when compared to both of the Iinb1 and F1 clones . In contrast , seven transcripts were upregulated in the Xinb3 clone , including one methyltransferase , one DNA mismatch-repair protein , and five UPs ( S6 Table ) . In Orsini et al ( 2016 ) , a number of transcripts were differentially expressed between P . ramosa infected and non-infected individuals of the same genotype ( same D . magna clone ) [26] . We investigated these transcripts to find if any of them would map to our interval . Importantly , we found no significant differences in gene expression between controls and P . ramosa treatments for transcripts mapping to PR-locus ( data not shown ) ( but see McTaggart et al . 2015 ) [28] . Rather , the significant differences in expression were identified when comparing the control treatments of the Xinb3 and Iinb1 clones . This is not surprising given that we are here investigating the host’s first line of defense , while genes expected to be expressed differently are genes whose expression is induced once the parasite succeeds in infecting its host—the second line of defense [18] . One model was suggested , whereby three D . magna resistance loci govern the Daphnia-Pasteuria host-pathogen system , regarding the two P . ramosa genotypes , C1 and C19 [22] . In this model , variation in locus C determines resistance to both P . ramosa genotypes whereas variation in loci A and B determines D . magna resistance to P . ramosa genotypes C1 and C19 , respectively . Epistasis between loci can be described as follows: the presence of the resistant allele in C masks the genotypes at loci A and B , and the presence of the resistant allele in A masks the genotype at locus B ( Fig 4 ) . A hierarchy of dominance between D . magna resistance phenotypes is observed: RR ( C1 , C19 double resistant ) > RS ( C1 resistant , C19 susceptible ) > SR ( C1 susceptible , C19 resistant ) > SS ( double susceptible ) [20 , 22] . Our analysis so far allows us to conclude that the predicted locus C ( Fig 4 ) is located within PR-locus . However , it does not resolve if different locus C alleles result from structural variation at the NHR or from variation in the flanking region . In addition , since all F2 recombinant clones were either RR ( double resistant ) or RS ( C1 resistant/C19 susceptible ) resistance phenotypes , we cannot withdraw any conclusions on whether loci A and B are located within PR-locus even though the three loci are expected to be closely linked [22] ( Fig 4 ) . Therefore , we undertook an association study , testing for a link between structural variation at the PR-locus and variation in resistance to P . ramosa spore attachment in D . magna clones collected from a metapopulation in the Tvärminne archipelago in Finland . We tested 447 Tvärminne clones from 27 different populations ( rock pools ) ( on average 16 . 5 clones per population ) for resistance to P . ramosa genotypes C1 and C19 using the attachment test and observed high resistance phenotype diversity between and within the rock pool populations ( S7 Table ) . We then tested two genetic markers ( g294 and g350 ) designed within xNHR unique coding sequences based upon the current draft genome ( ver 2 . 4 ) for the susceptible D . magna clone Xinb3 for presence/absence patterns . We had two predictions: i ) that these markers ( S1 Table ) would produce an amplicon when the xNHR haplotype was present either in a homozygote or heterozygote form , but not when it was absent from the tested genotype; and ii ) that since the RS phenotype ( observed in Xinb3 clone ) is dependent on the dominant allele of locus A , these amplicons would be produced irregularly in RR clones , always in RS clones and never in SR ( C1 susceptible/C19 resistant ) and SS ( double susceptible ) clones . Our analysis revealed two groups of host genotypes . There were genotypes where the xNHR diagnosis markers amplified together ( as does the Xinb3 clone ) and other genotypes where none of the markers could be amplified ( as is the case for the Iinb1 clone ) ( Fig 1C ) . As expected , this amplification pattern was strongly associated to resistance to P . ramosa C1 genotype . Specifically , clones susceptible to C1 almost never showed xNHR diagnostic marker amplification ( resistance phenotypes SR and SS ) . Clones that are at the same time resistant to C1 genotype and susceptible to C19 genotype ( RS ) always show amplification ( this is also the case for the Xinb3 genotype ) , whereas double resistant clones ( RR ) could show amplification or not ( Fig 5 ) . The double resistant Iinb1 clone does not show amplification of any of these xNHR diagnostic markers . We tested whether these results would be confirmed within a single D . magna population . We chose a rock pool population ( K-8 ) with only RS and SR resistance phenotypes being present and predicted that this polymorphism is associated with presence and absence of the xNHR . In our K-8 population sample we found that 56 out of 60 RS clones showed xNHR marker amplification , whereas only one out of 36 SR clones showed such amplification ( Table 1 ) . Thus , we find a strong association between presence of xNHR haplotype and RS resistance phenotype , and xNHR absence and C1 susceptibility both within and between populations .
The fine mapping and sequence analysis of the Daphnia magna PR-locus revealed an unusual pattern of structural polymorphism between haplotypes . Remarkably , we find lack of homology between PR-locus haplotypes in restricted regions of 55 kb and 121 kb , the xNHR and iNHR , respectively ( Figs 2 and 3 ) ( S2 Table ) . In the PR-locus haplotypes , and particularly within the NHR sequences we found a complex pattern of repeated sequences , which likely represent a history of evolutionary events with multiple classes of structural mutations playing a role . The existence of large-scale repetition of sequences found elsewhere in the D . magna genome , the extra-locus repeats ( Fig 2 ) ( S3 Table ) , argues against horizontal gene transfer in creating the NHR , while suggesting that gene conversion might be a recurrent phenomenon influencing its evolution . The difference in length between the two haplotypes is explained by a far higher prevalence of intra-locus repeats in the iNHR in comparison to the xNHR that suggests a higher number of segment duplication events in iNHR ( Fig 2 ) ( S4 Table ) . Finally , the lack of homology between the two NHR haplotypes together with the observation that this region seem to segregate as one unit in natural populations , suggests the absence of , or very low rates of local recombination . Taken together , our results indicate that the NHR represents a defined and highly divergent genomic region whose structural genetic variation underlies the natural variation in D . magna resistance to P . ramosa . The characteristics that we find in the NHR of the D . magna PR-locus largely overlap with what is known of the genetics , origin , structure and evolution of supergenes . Supergenes are clusters of multiple loci , each affecting different traits that together control complex phenotypes within a species and segregate as a block that is characterized by restricted or suppressed recombination [29] . Supergenes can emerge due to new mutations leading to beneficial interactions with closely linked loci , or to structural large-scale mutations such as gene duplication and translocation [29] . Large-scale structural polymorphisms are one of the main reasons for recombination suppression in supergenes and there are examples of supergenes being located in genomic fragments that are absent in alternative haplotypes [29 , 30] . Finally , NFDS seems to be the main evolutionary mechanism to maintain supergene polymorphism [29] . Thus , it is tempting to suggest that the NHR of D . magna PR-locus may represent an immunity supergene . We collected more than 400 clones from a well-studied D . magna metapopulation located in the Tvärminne archipelago in South-Western Finland and made an association study between their resistance phenotypes for P . ramosa genotypes C1 and C19 and the presence of diagnostic markers of the xNHR . We find that the presence of the xNHR haplotype is tightly associated to the RS phenotype ( C1 resistance and C19 susceptibility ) , while xNHR markers are absent in D . magna clones with SR and SS phenotypes ( Fig 5 ) . On the other hand , the presence of xNHR markers shows no association with RR phenotypes ( Fig 5 ) . We verified the association between xNHR and the RS phenotype in a single population ( rock pool K-8 ) , which was polymorphic only for RS and SR phenotypes . In this population the matching-allele matrix–already described for this host-parasite system–is clearly seen [21 , 22] . D . magna clones showing RS phenotype are homozygote or heterozygote for the dominant xNHR , while this haplotype is absent in SR clones ( Fig 4 ) ( Table 2 ) . Gene conversion , rare events of homologous recombination at NHR , or errors while determining the resistance phenotypes or the marker could explain the few instances where xNHR diagnostic markers are absent in RS clones or present in SR clones ( Table 1 ) ( S7 Table ) . Our results are consistent with previous work showing a dominance hierarchy between D . magna resistance phenotypes and epistasis between resistance loci [20 , 22] . The NHR corresponds to the A-locus in these earlier studies . The xNHR contains the dominant allele of the A-locus whereas the iNHR contains the recessive allele . The phenotype associated to xNHR is hidden in RR clones , as its effect is suppressed by the dominant C allele at the C-locus ( Fig 4 ) . Conversely , the presence of the xNHR is strongly associated with the RS phenotype and completely absent in SR and SS clones . The presence of the xNHR masks the effect of the B-locus , which defines the SR and SS resistance phenotype polymorphism ( Fig 4 ) . In population K-8 the C-locus is apparently fixed for the recessive c-allele , while the B-locus is fixed for the dominant B-allele ( Table 2 ) . On the other hand , the results of the QTL mapping leading to PR-locus , is based on a polymorphism at the C-locus ( parents are CC—Iinb1 , and cc—Xinb3 , while the F1 is Cc ) , because the parental genotypes used , Iinb1 and Xinb3 clones , have RR and RS phenotypes and no other phenotype was found in over 400 tested F2 recombinants [22] . Thus , the C-locus is also located within the PR-locus ( Fig 4 ) . Finally , a report of recombination between the three linked resistance loci concluded that the B-locus is located between loci A and C [22] , suggesting loci A , B and C loci would all sit within the PR-locus ( Fig 4 ) . Until now few empirical examples of matching-allele interactions have been described in host-parasite systems [31] , which can result from this type of genetic interactions being rare . However , in the D . magna-P . ramosa system the ease of collecting , large samples are easily available for collection , genotyping and phenotyping . Furthermore , the clonal system of reproduction of D . magna permits the maintenance of stable genotypes without the need to produce inbred lines [8 , 21 , 22 , 24] . Together , these traits increase the probability of finding existing matching-allele interactions . In addition , many studies of host-parasite systems rely on the overall infection results whereas the infection process requires a series of steps , each with its own genetic basis [18] . In the D . magna-P . ramosa system the spore attachment step is the only infection step that fulfils the requirements of a matching-allele model: binary response; lack of environmental variability and; host-parasite genotype-to-genotype interactions . It is possible that by focusing on infection steps that show the same characteristics and using large numbers of host and parasite genotypes , future studies reveal more examples of matching-allele interactions . In parallel to large structural polymorphisms found in the NHR region of D . magna PR-locus we found differences in the gene content between the i- and the x- haplotypes at the PR-locus . Most differences in gene content are associated with genes that map to the NHR region ( S4 Table and S5 Table ) . Gene annotation reveals that genes of the glycosyltransferase family are over-represented within xPR-locus including seven fucosyltransferases , two alpha 1 , 4-glycosyltransferase and one galactosyltransferase transcripts ( S5 Table ) . In contrast , iPR-locus has only two fucosyltransferase transcripts , one alpha 1 , 4-glycosyltransferases and one galactosyltransferase ( S5 Table ) . Glycosyltransferases are known to play fundamental roles in innate and acquired immunity-related traits in multiple organisms [32–34] . Thus , the differences in the presence and activity of fucosyltransferases and alpha 1 , 4-glycosyltransferases indicate that these are good candidates genes that may determine variation in D . magna resistance to P . ramosa . D . magna—P . ramosa is a host-pathogen system where growing evidence suggests NFDS as the primary responsible of the coevolutionary process [20–23] . Here we describe the first steps into the molecular basis of evolution by NFDS and find evidence that suggest a role for glycosyltransferase genes in our study system . Next , it is important to identify which particular genes are responsible for the observed polymorphism . That requires to fine-map the A , B and C loci ( Fig 4 ) and to then carry out functional tests on the remaining candidate genes ( e . g . gene knock-outs ) to verify their role . Furthermore , it is important to describe more D . magna PR-locus haplotypes associated with different resistance phenotypes to better understand the extent of the genetic variation associated to D . magna resistance to P . ramosa and the relative roles that gene conversion and homologous recombination have in shaping it .
The D . magna ( Xinb3 x Iinb1 ) F2 recombinant panel is a resource available at the Ebert laboratory in Basel , Switzerland , that was generated from a single cross between the Xinb3 mother clone and the Iinb1 father clone [27] . A QTL analysis based on this resource revealed one major effect QTL for resistance against P . ramosa genotype C19 [25] . In the region of the major QTL for resistance to P . ramosa , single nucleotide polymorphism ( SNP ) and microsatellite markers were designed based on the D . magna 2 . 4-genome draft ( S1 Table ) . We amplified each marker via standard PCR and Sanger sequenced them in all F2 clones with a recombination event in the region around the resistance QTL . We then searched for the recombination breakpoints in each F2 recombinant clone . Since the region around the QTL was poorly assembled in version 2 . 4 of the D . magna draft genome ( http://wfleabase . org/ ) , we undertook a number of additional sequencing and assembly methods in order to better resolve the focal region . For Xinb3 we generated high coverage ( ~60X ) PacBio sequencing in order to perform de novo genome assembly . For Iinb1 we took a hybrid Illumina short-read/PacBio long-read approach , generating ~80X 125bp PE Illumina coverage and ~ 15X PacBio long-read coverage ( see S1 Methods ) . We used the D . magna Xinb3 and Iinb1 haplotype sequences obtained to search for homologies within and between haplotypes and other genomic regions ( see S1 Methods ) . In order to understand how expression of individual genes localized to the focal genome regions and to other parts of the genome differed between the Xinb3 and Iinb1 clones , we conducted a de novo transcriptome assembly of the data set described in Orsini et al . ( 2016 ) ( see S1 Methods ) [26] . Finally , we constructed a de novo annotation of each of the transcripts mapping to PR-locus by performing blastx ( nucleotide to protein ) searches in the NCBI database ( see S1 Methods ) . The aim of this assessment was to link the structural polymorphism observed in the QTL panel with genetic variation for resistance in natural populations . D . magna females were collected from fresh water rock pools in the long term study area of the Tvärminne archipelago , South-Western Finland . The Tvärminne archipelago is composed of many skerry islands of varying sizes , each with multiple rock pools that freeze in winter , forcing the Daphnia to survive as sexually produced resting stages called ephippia . It is the location where the ancestor of the D . magna Xinb3 genotype ( our three times selfed reference genome clone ) was first collected . Each rock pool represented one population , but together these populations form a metapopulation with frequent migration . Females were freshly hatched from sexually produced resting stages ( ephippia ) in the wild right after the winter season and thus each of them represented a unique genotype ( clone ) . In the laboratory , we separated females into individual jars initiating a clonal line . Clones were kept in ADaM media at 20°C , fed with Scenedesmus sp . three times a week and moved to fresh media once a week [20 , 35] . Resistance phenotypes were determined using the attachment protocol described in Duneau et al . ( 2011 ) [19] . Two cloned P . ramosa genotypes , C1 and C19 , were used in this study [24] . In short , three replicates of each D . magna clone were placed individually into 96-well plates and exposed for one hour to spores of P . ramosa C1 or C19 genotypes marked with fluorescein5 ( 6 ) isothiocyanite [19] , after which the attachment of spores to an individual was assessed under fluorescent microscope . Attachment of spores to the esophagus of the host indicated that this host genotype was susceptible to the pathogen genotype tested whereas absence of spore attachment implied host resistance [19] . Primers for genetic structural markers were designed based on the available Xinb3 D . magna genome draft ( version 2 . 4 ) at the time . Each primer pair was selected so that it amplified one coding sequence predicted to be present in the annotated genome ( S1 Table ) . Absence or presence of visible amplicons on an agarose gel ( 1 . 5% w/v ) was used as indicator of PR-locus genotypes ( absence indicating homozygotes for absence , while presence indicates homozygotes for presence or heterozygotes ) . Statistical analysis was based on contingency tables of expected vs . observed values to which a Chi-square test was applied to test statistical significance to both the full dataset and to pairwise comparisons between resistance phenotypes . | Negative frequency-dependent selection , whereby common genotypes are disfavored , resulting in cyclic change of gene frequencies and maintenance of genetic diversity in host and parasite populations , is one the mechanisms predicted to drive host-parasite coevolution . Specific matching-allele interactions between hosts and parasites are a mechanism predicted to underpin this mode of selection . In spite of in depth research , little is known about the genetic basis of such matching-allele interactions and few empirical examples have been described . Recent research has suggested that the Daphnia-Pasteuria host-parasite system follows a model of negative frequency-dependent selection . We map a Daphnia magna locus of resistance to Pasteuria ramosa . We use next-generation genome and transcriptome sequencing to characterize resistant and susceptible haplotypes of the resistance locus . We find large-scale structural polymorphism between resistance locus haplotypes and we find evidence that gene conversion , segment duplication and restricted homologous recombination contribute to produce the observed polymorphisms . We analyse natural populations and find that the resistance locus structural polymorphisms reproduce the matching-allele interactions predicted for the Daphnia-Pasteuria system . This work presents rare and conclusive evidence of the genetic basis of matching-allele interactions in host-parasite systems while opening research avenues to find the underlying molecular mechanisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"quantitative",
"trait",
"loci",
"cloning",
"parasitic",
"diseases",
"alleles",
"animals",
"genetic",
"mapping",
"molecular",
"biology",
"techniques",
"research",
"an... | 2017 | The genetic basis of resistance and matching-allele interactions of a host-parasite system: The Daphnia magna-Pasteuria ramosa model |
Electronic patient records remain a rather unexplored , but potentially rich data source for discovering correlations between diseases . We describe a general approach for gathering phenotypic descriptions of patients from medical records in a systematic and non-cohort dependent manner . By extracting phenotype information from the free-text in such records we demonstrate that we can extend the information contained in the structured record data , and use it for producing fine-grained patient stratification and disease co-occurrence statistics . The approach uses a dictionary based on the International Classification of Disease ontology and is therefore in principle language independent . As a use case we show how records from a Danish psychiatric hospital lead to the identification of disease correlations , which subsequently can be mapped to systems biology frameworks .
With the consolidation of EPR systems in modern healthcare , massive amounts of clinical data and phenotype data are gradually becoming available for researchers [1] , [2] , [3] , [4] , [5] , [6] . Alone , or integrated with existing biomedical resources , these EPR systems constitute a rich resource for many types of data driven knowledge discovery as we demonstrate in this paper . In the coming years , as these data are also coupled to the expected explosion in personal genomic data , the translational meeting of ‘bench and bedside’ is expected to push scientific advancements in personalized medicine [4] , [7] , [8] , [9] , [10] . EPR systems document patient morbidity , treatment and care over time . They comprise different types of structured and unstructured data , ranging from coded diagnoses , ordinary physiological measures , biobank data , laboratory test results over medication prescriptions , and treatment plans , to free text notes such as admission notes , discharge notes and nursing notes [11] , [12] . We focus here on the assigned structured diagnosis codes and the free text notes . In our Danish setting , assigned codes are coded in the EPR according to the International Classification of Disease version 10 ( ICD10 ) , and are ultimately reported to the discharge registries for reimbursement . This process has known ( but poorly quantified ) biases since codes result in different reimbursement sums [13] , [14] . Assigned codes will also typically pertain strictly to the current hospitalization and the morbidity deemed strictly relevant to it . These bias and completeness issues are also documented in insurance claims data with ICD9 [15] . In contrast free text notes should not have this bias , and contain much additional information , but in an inherently unstructured form ( refs ) . In this paper we demonstrate how text- and data mining techniques can be used to extract clinical information hidden in text to augment coded data . The result is a much more complete phenotypic description of patients , than what could be obtained from just structured data and registries . There is an increasing focus on the research potential of both structured and textual data collected in EPR systems and registries . Examples of this work is classical database knowledge discovery and association mining [16] , [17] , [18] , identifying and classifying specific medical cases or conditions in an EPR [19] , [20] , [21] , [22] , patient safety and automated surveillance of adverse events , contraindications and epidemics [23] , [24] , [25] , comorbidity and disease networks [26] , [27] , [28] , autocoding of clinical text [29] , [30] , [31] , [32] , medication information extraction [33] , [34] and identifying suitable individuals for clinical trials [35] , [36] . Also see review by Meystre et , al [37] . Some of this work deals strictly with structured data , while some use text mining techniques to extract information from text . Much of the latter work builds on existing Natural Language Processing ( NLP ) text mining tools designed for recognizing clinical terms and findings and mapping them to controlled vocabularies such as the United Medical Language System ( UMLS ) . Some of these tools are MedLee , MetaMap , cTakes and HITEx ( [29] , [38] , [39] , [40] ) . For Danish text , unfortunately no such EPR Information Extraction tools exist . To extract data from the text for our analysis , we therefore constructed our own text mining module compatible with Danish classification resources and easily adapted to any language with a translation of ICD10 . Our comparatively simple approach significantly enriches structured EPR data , and allows a higher resolution analysis than otherwise possible . Independently of the research assisted by the information presented in the patient records , several approaches have been developed to discover novel disease associations , either based on shared disease causing genes or on overlapping pathways [26] , [41] , [42] . Known disorder–gene associations from available resources like OMIM have been used to establish links between diseases , thus creating a network of disorders [26] . Common to many of these approaches is the extensive use of protein-protein interactions from large-scale proteomic studies . Linking disease-gene information with the growing data present in EPR systems will allow for a better understanding of disease etiology and phenotype-genotype associations . The PheWAS work at Vanderbuilt University . [43] , [44] is a recent illustration of this . Here we describe a strategy for exploring EPR data from a patient cohort in the context of subsequent systems biology analysis . By mining the free-text parts of the EPR from a psychiatric hospital we are able to augment the disease information assigned in structured formats as ICD10 codes , and thus obtain a much richer phenotype profile of each patient . Treating these profiles as phenotype vectors [41] in the controlled vocabulary space of the ICD10 disease classification , we demonstrate how they can be used to investigate disease comorbidity and patient stratification , paving the way for discovery of the underlying molecular level disease etiology in the form of overlapping genes and pathways . A longer-term perspective is to also include genetic profiles of the individuals in these data integration schemes , but this is not explored in the present paper .
We based our study on a corpus of 5 , 543 patient records from the Sct . Hans Hospital ( the largest Danish psychiatric hospital ) collected in the period 1998–2008 . A manually curated subset of the records was used to assess the precision of the text mining approach . From structured fields in the EPR , we extracted 31 , 662 ICD10 codes , representing 351 different level 3 codes and corresponding to 2 . 7 unique codes associated to each patient on average . In the selected text found in the EPR our text mining approach matched 218 , 963 text strings to strings in a compiled dictionary of ICD10 terms and generated term variants ( see Materials and Methods and Text S1 for additional detail ) . A further 22 , 956 matches were disqualified by a negation module whenever a negating word or mention of another subject ( e . g . mother , sister or friend ) was found in the preceding part of the sentence . The corresponding codes of these terms covered 554 different level 3 ICD10 codes , on average 9 . 5 unique codes per patient . Combining mined and assigned codes results in 674 different ICD10 codes with 12 . 3 average codes per patient The combined data was gathered in a Patient-ICD10 association matrix , by assigning each Patient–ICD10 combination both a binary and a TF-IDF ( [45] ) weighted value indicating whether or not a given code was associated with a given patient and how strongly . Rows thus represent the morbidity of a patient as a vector in ICD10 space , and columns represent the prevalence of a ICD10 as a vector in patient space . The precision of our text mining was quantitatively assessed by manually checking all 2 , 724 mining hits for 48 patients ( Table 1 ) . The validation set covered 214 full level ICD10 codes , corresponding to 151 level 3 codes . A hit was considered correctly assigned when it was possible to infer a direct clinical link between the term and the patient from the record context . We defined precision in two ways: Incidence precision of all curated hits , and association precision , where an ICD10 code is considered correctly associated with a patient if it h77as at least one correct incidence . In both cases we considered how the precision was distributed among the different chapters . We found a total incidence precision of 87 . 78% and an association precision of 84 . 03% . False text mining hits fall in the categories: Negations , 3 . 9%; false subject , 0 . 6%; Delusion , 0 . 3%; Putative , 1 . 5%; Polysemic , 0 . 3%; Information to patient , 3 . 3%; Other , 2 . 2% ( see Text S1 ) . For the same 48 patients we also manually curated the 411 hits ( 373 negations and 38 subject ) disqualified by the negation module . 330 of these were correctly disqualified giving an 80% precision of the negation module . 122 text mining hits out of 2 , 724 are due to hits categorized as negations or false subject that were not caught by the negation module . Combining the numbers the negation module identifies 73% of all relevant negations ( 330/ ( 330+122 ) ) . The negation module is similar to the approach of the NegEX method [46] , [47] . A further breakdown of the validation is available in Text S1 . ICD10 is organized into 22 chapters according to disease areas ( see Materials and Methods ) . To discover the degree of comorbidity between chapters , we constructed an ICD10 chapter network ( Figure 1 ) . Based on which diseases belonging to a specific chapter each patient has in the corpus , we calculated a similarity score between the different chapters , ranging between 0 ( for the lowest comorbidity ) , to 1 ( highest comorbidity ) , see Materials and Methods . Codes for chapter V ‘Mental and behavioral disorders’ account for over 80% of the assigned codes given by physicians at Sct . Hans Hospital , while codes for chapter XXI ‘Factors influencing health status and contact with health services’ have a frequency of around 7% . These are also the two most correlated chapters . The strong correlation between mental disorders of chapter V and the observational Z-diagnoses of chapter XXI is most likely explained by a large ward in the hospital for forensic psychiatry , where patients are frequently admitted for mental observation following a criminal offence . When including both the assigned and the mined codes from the textual records we capture many symptomatic descriptions for diseases . As seen on Figure 1b , more than 35% of all codes are pertaining to chapter XVIII ‘Symptoms , signs and abnormal clinical and laboratory findings , not elsewhere classified’ , e . g . general medical complaints , edema , back pain , and elevated blood glucose . Chapter XIX ‘Injury , poisoning and certain other consequences of external causes’ , as well as chapter XVIII , exhibit a high correlation with chapter V . Assigned codes are often restricted to the principal psychiatric illness and important for billing and social purposes , not necessarily reflecting the actual psychiatric treatment and care , nor the somatic disorders affecting the patient . For this reason , introducing the mined codes in the analysis allows capturing correlations that were previously impossible to find . In our attempt to identify pairs of interesting unexpected co-morbidities , as well as general trends of correlation , we investigated pairs of ICD10 code vectors in patient space ( columns in the patient-ICD10 association matrix ) . We used two measures to rank the 226 , 801 possible pairs of the 674 ICD10 codes , according to their co-association , compared to what would be randomly expected . Pairs were sorted based on p-values and a cut-off was imposed based on a comorbidity score and a false discovery rate of 1% ( see Materials and Methods ) . The result is a list of 802 candidate ICD10 diagnostic pairs that occur more than twice as often as expected by random , and that are statistically significant at a false discovery rate of 1% ( Data S1 ) . Using the comorbidity score as a similarity measure we clustered all 674 ICD10 codes and created a corresponding heatmap of the comorbidity scores for the ICD10 pairs . Figure 2 shows a truncated version of the entire heatmap , containing the scores of all the interactions for the top ranking 100 ICD10 codes ( i . e . , the top 100 codes found when sorting the list of 802 candidate pairs by their comorbidity score ) . The full heatmap for all 674 ICD10 codes extracted from the corpus can be inspected in Figure S1 . Figure 2 illustrates the general ability of our approach to capture correlations between different disorders . Several clusters of ICD10 codes relating to the same anatomical area or type of disorder can be identified along the diagonal of the heatmap . They range from trivial correlations ( e . g . , different arthritis disorders ) , to correlations of cause and effect codes ( e . g . , stroke and mental/behavioral disorders ) , to social and habitual correlations like drug abuse with liver diseases and HIV . Another interesting observation on the composition of the corpus is the lower than expected co-occurrence between the codes of the ‘mental and behavioral disorders’ cluster and the ‘drug abuse , liver disease , HIV’ cluster , as indicated by the blue areas in the upper and lower corners . These are very different groups of disorders that strongly stratify the patient corpus , and inspection of the specific diagnoses indicates that the correlation reflects two of the primary causes for admittance to the Sct . Hans Hospital ( i . e . , two distinct clinical departments ) : psychiatric disorders caused by stroke or brain injury , and mental illness accompanied by drug abuse . Our approach will , and should , for the most part return trivial or already known co-morbidities . This is a result of the non-independence of ICD10 codes . These will to a certain extent be expected to correlate according to anatomical and functional similarity , which again is what the taxonomy of the ICD10 classification attempts to capture . This is also reflected in figure 1 where e . g . chapter V , Mental and behavioral disorders and chapter VI , Diseases of the nervous system exhibit correlation . One could attempt to reduce this type of dependency , by imposing filters for intra chapter pairs , or in other ways use the taxonomy as a filter or weighing scheme [48] . However since the candidate list resulting from the described pipeline was manageable for manual curation , we choose to not impose further filtering with the risk of losing interesting comorbidities . Trivial pairs occur for example between two codes for essentially the same disease ( e . g . , E11 ‘Non-insulin-dependent diabetes mellitus’ and R73 ‘Elevated blood glucose level’ ) , between trivial disease-symptom pairs ( e . g . , N30 ‘Cystitis’ and R30 ‘Pain associated with micturition’ ) , or between pairs of well-established correlations ( e . g . , E51 ‘Thiamine deficiency’ and H55 ‘Nystagmus and other irregular eye movements’ ) . To discriminate potentially interesting , novel candidate co-morbidities from the many trivial ones , an experienced medical doctor manually inspected the candidate list of 802 pairs and flagged 93 surprising co-morbidities A list of all code pairs as well as flagged pairs can be seen in Supplementary Data S1 . Disease correlations may or may not have genetic causes . To identify a possible molecular basis for the flagged pairs , we extracted genes implicated in those particular diseases when a good mapping from ICD10 to OMIM was possible ( see Materials and Methods ) . We then created a protein-protein interaction network by determining the first order interactions of those genes in refined experimental proteomics data ( see Materials and Methods ) . For each disease pair , we searched for shared first order interactions connecting the two networks . Despite the difficulty of mapping the different terminologies and genes with this approach [27] , the analysis revealed several connected proteins which are novel in relation to the diseases used to generate the networks . For example , we narrowed down an interesting case story between Alopecia ( i . e . , hair loss , ICD10 L65 ) and Migraine ( ICD10 G43 ) . We found that THRA , thyroid hormone receptor , not previously associated with any of the two diseases , is a shared interaction partner of Protein Hairless ( HR , a putative single zinc finger transcription factor protein ) involved in alopecia [49] , and the Estrogen Receptor 1 ( ESR1 ) associated with migraine [50] , with a p-value of 1 . 17×10−3 ( Materials and Methods ) . This may suggest that these two diseases share a similar molecular mechanism of action . A network view of these proteins and their interaction partners can be seen on Supplementary Figure S2 . Migraine and alopecia were associated to 210 and 38 patients respectively , with 12 cooccurences ( comorbidity score of 1 . 92 , p-value of 2 . 07×10−6 ) . To confirm these associations , which primarily came from text mining , we checked the surrounding textual contexts of all the mining associations to check their validity . For the 12 overlaps a medical doctors looked for confirmation in the full EPR record . In the case of migraine , in some cases a more correct clinical description would have been ‘headache’ , and for alopecia some cases covered fear of or delusion of hair loss . The corrected contingency numbers were 168 ( migraine ) , 26 ( alopecia ) , 9 ( both ) , and results in a comorbidity score of 0 . 4 and a p-value of 2 . 81×10−6 . Of the remaining 9 patients with migraine and alopecia , six are women aged 21–63 and three are men aged between 47 and 54 . The observed comorbidity may reflect different side effects from medication [51] , [52] , [53]; most prominently seen with SSRIs ( Selective Serotonin Reuptake Inhibitors for treatment of depression ) that have been associated with cutaneous reactions , including alopecia , and migraine [54] . Also , frequently prescribed oral contraceptives are associated with migraines [55] . In fact , inspection of the nine comorbidity cases revealed that three patients were being treated with SSRIs ( with a possible link to hair loss mentioned in the medical notes ) , two patients were administered oral contraceptives and one patient was treated with calcium antagonists and antiepileptic drugs . Removing 3 of the co-morbid cases corresponding to the SSRI treated patients results in a recalculated p-value of 2 . 9×10−4 . The comorbidity may also have an etiological cause that relates to schizophrenia , the primary disease of the patients . It has previously been shown that schizophrenia is associated with celiac disease , i . e . the highly under-diagnosed condition of gluten allergy [56] , which in turn has been linked to both alopecia , and migraine; in fact the two latter conditions are now indications for diagnostic work-up for celiac disease according to the recent guidelines from the American Gastroenterological Association [57] , [58] . In a specific hospital corpus the most important level of stratification is generally based on the primary diagnosis , or inclusion , which dictates treatment and care . The stratification can be very specific and based on lab results and tests for molecular markers , such as in the case of hormone receptor variants in breast cancer [59] . We were interested in determining if the combined mined and structured data could lead to a richer structure in the patient population , spanning a wider range of phenotypes , not typically considered when stratifying a specific corpus by assigned codes . In the patient-ICD10 association matrix each patient is represented as a vector of associated ICD10 codes in the space of all the 674 ICD10 codes . We calculated cosine similarity [41] between the ICD10 vectors of all possible pairs of patients , and used this as the basis for a hierarchical clustering of patients . We used TF-IDF [45] weighted values in the association of each ICD10 code to the vector of a patient . ( see Materials and Methods ) . Figure 3 shows those 26 clusters with at least 25 members resulting from the clustering . They are laid out according to the patient-patient similarity and colored by group membership . The ICD10 characteristics of each group are seen in figure 3b ( see Materials and Methods ) . In all but one cluster , 54 , a single ICD10 code stands out as the most discriminating code . The TF-IDF value for this code constitutes up to 18–40% of the sum of all TF-IDF values in the vector . Furthermore , no two clusters share the same main code . The ICD10 characteristics of each cluster are shown in Figure 3b . From this figure , we see that Schizophrenia has a strong component in several clusters , primarily located in the top left of the network . As pictured , many of these clusters are also characterized by various codes for alcohol/drug use , indicating the type of abuse as a good sub-stratification of schizophrenia . Similarly , alcohol seems to be a common denominator for clusters 48–54 , which are primarily characterized by depressive disorders , anxiety disorders , and other personality disorders . What is also interesting is that many patients fall into clusters characterized by somatic codes like diabetes and psoriasis , which have certainly not been the initial reason for admittance to the hospital . This is largely attributable to data coming from text mining ( see Supplementary Data S1 ) .
As EPR systems become the norm in modern health care , focus is naturally turned to exploring this treasure trove of data for improving health care and research [60] . Extracting the data is a first step , and as EPR systems in many countries maintain the use of free text to complement structured data , text-mining approaches are necessary for extracting data usable in further analyses . The enrichment of existing structured patient data by text mining significantly expands phenotype profiles , both within the specific pathology of the corpus , but especially into other disease areas . We present one example of comorbidity between two diseases that are very often not coded in the record by the physician , but show up in the patient record text and are later picked up by mining . The enrichment from mining is also visible in our attempts to stratify patients , where potential is shown for uncovering additional layers of the population structure . More detailed stratification of patient cohorts could help improve population homogeneity and signal strength in genome wide association studies , and lead to increased power in case-control studies [35] , [36] . The procedure described here represents , in our opinion , a practical non-hypothesis driven approach for extracting valuable information from patient records for any patient corpus where manual inspection and ICD10 association would turn into an otherwise impossible task . Furthermore , we show how this information can be used in researching disease comorbidity and patient stratification and how it can be mapped to the underlying systems biology revealing possible causes for the observed correlations . The results obtained from a data driven approach like this one will obviously depend on the composition and domain of the patient corpus and on the amount and quality of the available data . In that sense , some of the found correlations and results will be domain or cohort specific , and do not necessarily translate to general population wide conclusions . In the case of patient stratification , this is inherently true . Even in these cases however , novel correlations can still be highly valuable and suggest hypothesis for causality within the cohort in terms of treatments , procedures , responses , and co-morbidities that are not necessarily genetically founded .
Patient data was analyzed anonymously and the project was ethically approved by the Danish National Board of Health ( No . J . nr . 7-604-04-2/33/EHE ) . The patient population data was collected from the Sct . Hans Mental Health Centre , in Roskilde , Denmark . All analyses were performed on an anonymous data set . A total of 5 , 543 patients were followed from 1998–2008 , and their records stored in an EPR database . 70% of the patients ( 4 , 822 ) are from the Copenhagen area , 61% of these are males . The average age is 30 years . The records are a mixture of structured diagnose assignments of ICD10 codes , ATC codes ( http://www . whocc . no/atc ) for medication usage , patient care notes from nurses and doctors , admission and personal information , etc . A corpus was created from the relevant tables of the Sct . Hans EPR , containing all unique text entries for each patient that were verified and signed by a physician . To each entry we assign an entry date , the note type , and the text . The note type identifies the type of text entry , such as the epicrisis , discharge note , treatment note , nursing note etc . A few non-medical notetypes such as ‘Social worker’ notes were excluded . In total , the corpus contains text for 4 , 765 patients with an average of 25 , 000 words per patient . In addition , we extracted all ICD10 codes assigned to patients that were stored in a structured format . The dictionary used in our text mining approach is based on the Danish translation of the Danish translation of the WHO International Classification of Diseases ( ICD10 ) , downloaded from the Danish National Board of Health the 2nd Nov 2009 . The ICD10 classification is a hierarchical classification of diseases and symptoms , divided into 22 anatomical/functional chapters with increased specification of terms in each lower level . The Danish translation of ICD10 consists of 22 , 261 terms , each uniquely matched to a code of between 3–5 characters . To increase the scope of the dictionary , we augmented existing terms with variants created by simple rules reflecting common semantic structures ( [61] , [62] ) in the Danish ICD10 terms . E . g . adding truncated versions of terms containing specifiers like ‘ . . forårsaget af . . ’ ( caused by ) , keeping just the preceding part . Terms containing commas and parenthesis are treated similarly . These variant terms are mapped to the same code as their parent . Since truncation throws away the detailed information in the case of low-level code-term pairs we ensure the code-term information content by rounding all codes to level 3 . In this way all terms are essentially treated as synonyms of the more generic level 3 meaning . With variants the final dictionary consisted of 53 , 452 terms . Generated term variants were responsible for 24% of the total number of hits . More detail about the ICD10 dictionary is available in 1 . For relevant reviews on methods in text mining see e . g . ( [37] , [63] , [64] , [65] ) . The compiled text for each patient was normalized for orthographic variation like the dictionary , and a simple sentence splitter was used to split the text into smaller units . For each unit , a stepping algorithm created all possible strings of 1–10 words and looked them up in the dictionary . Exact matches were required . The longest possible match was always chosen . Candidates matching a blacklist of polysemic or otherwise mis-informative terms were disqualified . Negations and false subject-term associations were handled by disqualifying matches when the preceding sentence contained tokens from a list of negations ( ‘never’ , ‘no’ , etc ) and subjects ( ‘mother’ , ‘friend’ , etc ) . Validated performance characteristics were covered in the results section . Further details about the text mining approach and its validation is contained in Text S1 . For each disease we created a vector mapping its presence or absence from a patient record . This resulted in 22 vectors for each disease chapter . The pair-wise overlap between vectors was quantified by calculating the cosine of the angle between normalized vector pairs [41] . The result is a score between 0 and 1 , mapping the comorbidity value of each of the chapter pairs . We also calculated the frequency of each chapter in relation to the total number of chapter assignments . In Figure 1 , the roman numerals represent the different ICD10 chapter numbers: I , Certain infectious and parasitic diseases; II , Neoplasms; III , Diseases of the blood and blood-forming organs and certain disorders involving the immune mechanism; IV , Endocrine , nutritional and metabolic diseases; V , Mental and behavioral disorders; VI , Diseases of the nervous system; VII , Diseases of the eye and adnexa; VIII , Diseases of the ear and mastoid process; IX , Diseases of the circulatory system; X , Diseases of the respiratory system; XI , Diseases of the digestive system; XII , Diseases of the skin and subcutaneous tissue; XIII , Diseases of the musculoskeletal system and connective tissue; XIV , Diseases of the genitourinary system; XV , Pregnancy , childbirth and the puerperium; XVI , Certain conditions originating in the perinatal period; XVII , Congenital malformations , deformations and chromosomal abnormalities; XVIII , Symptoms , signs and abnormal clinical and laboratory findings , not elsewhere classified; XIX , Injury , poisoning and certain other consequences of external causes; XX , External causes of morbidity and mortality; XXI , Factors influencing health status and contact with health services; XXII , Codes for special purposes . For the purpose of exploring comorbidity between ICD10 codes we used two measures to rank the 226 , 801 possible ( ( 674*674-674 ) /2 ) pairs of different codes , according to how often they come together in patients , compared to what would be randomly expected assuming no a-priori correlations . The two measures represent our desire to ensure statistical significance , while focusing on pairs with a noticeably increased co-association . First , for each pair of ICD10 codes A and B , the patient corpus is divided and counted in the four categories: A & B , A NOT B , B NOT A and NOT A NOT B , according to their association to A and B . Using this , p-values are calculated using Fishers exact test , and the pairs are sorted accordingly . We then filtered this list by imposing a cut-off value of 1 . 0 of a comorbidity score between diseases A and B defined as:Where Obs is the observed number of ICD10 co-associations , and Expt is the expected number . Expected overlaps are calculated based on the prevalence of each disease in the actual corpus ( nA and nB ) . To make the tendency to favor pairs of low prevalence ICD10 codes less pronounced , a pseudo-count of 1 is added to nominator and denominator . Since we take log2 of this ratio , a cut-off value of 1 . 0 means we restrict our focus to pairs with a higher than two fold ( approximately ) over co-association . This comorbidity measure is very similar to the one used by Hidalgo et al . [66] . Finally we used a Benjamini-Hockberg false discovery rate method [67] on the ranked list to correct for multiple testing . The p-values for all pairs are multiplied by the total number of pairs ( 226 , 801 ) and divided by the rank of the pair in the sorted list . A cut-off is then imposed where the corrected p-value drops below 0 . 01 . The result is a selection of 802 potentially interesting candidate pairs , with a false discovery rate of 1 percent , from the total of 226 , 801 pairs . There is no direct mapping between ICD10 codes and the OMIM [68] record entries . Furthermore , the disease names used by ICD10 and OMIM are not identical , so there was a need to map OMIM disease names into ICD10 codes . Work has been done mapping the online database and ICD9 codes , a previous version of the ICD [27] . We used the ICD10 to ICD9 General Equivalence Mapping available online from CMS ( http://www . cms . gov/ICD10/ ) to map the ICD codes to their previous version . With the mappings in place , OMIM was parsed for phenotypic descriptions of defects in genes , as described in Lage et al . , 2007 [41] . From the OMIM records , the clinical synopsis field was extracted for retrieving phenotypic descriptions regarding a certain disease . Additional information was retrieved from the morbid map tables , a map of disorders included in OMIM that have the syndrome name , chromosomal localization , and name of the disease causing gene . A manual curation step by a medical doctor ensured that each ICD10 code to be included in the analysis was assigned the correct OMIM entries . For each disease , a network was generated by taking the disease causing genes extracted from OMIM and determining their first order interactions in a human protein interaction network of refined experimental proteomics data . This procedure is described in detail elsewhere [41] , [69] , [70] . For determining genetic overlaps between two ICD10 diseases , we take their networks and identify those genes which are shared and have first order interactions with the seed genes . After a round of automatic overlap detection , we manually curated the results of the different steps in the pipeline , in order to detect erroneous assignments of disease names or genes , and reran the overlap detection in those cases . For those pairs where overlapping protein-protein interaction networks indicate underlying biological evidence , a final round of validation was done by manually checking if the binary associations from text mining of patients to the ICD10 codes were correct . Based on the corrected data , new p-values were calculated by Fishers exact test , and it was controlled that the p-value remained lower than the lowest p-value of the list of 802 candidates . The candidate genes found to overlap in the two disease networks were scored using the enrichment of OMIM seed genes in their first order interaction network , in a similar procedure as the one used by Lage et al . , 2010 [69] . The score assigned to a candidate was the hyper geometric p value of observing the amount of interactions to the OMIM set out of all the interaction partners of the candidate . Our example of THRA has a total of seventeen interaction partners in the network , and two are with the input genes ( HR and ESR1 ) , having a p-value of 1 . 17×10−3 . By looking at the Patient-ICD10 matrix by rows , or patient vectors in ICD10 space , we can stratify patients based on the similarity of their ICD10 associations . Instead of a binary association of a given code to a given patient , we weighted the significance of ICD10 occurrences using the term frequency – inverse document frequency measure ( TF-IDF ) [45] . TF-IDF rewards high code frequency in the individual record , and penalizes high prevalence across the corpus . As a patient-patient stratification measure , we used the cosine similarity CS [41] to calculate the cosine of the angle between all pairs of vectors . We included only patients with at least three associated codes , and exclude a number of trivial/symptom codes ( e . g . , pain , coughing , itching ) . A total of 2 , 584 patients were found to have at least three associated codes . We used 1-CS as a distance measure and calculated average linkage clustering to divide patients into clusters . Manual inspection of the clustering dendrogram led us to cut the tree at a CS value of 0 . 6 , which created a total of 307 clusters . 26 clusters contained 25 or more members , accounting for a total of 1 , 800 patients . Taking all edges with CS greater than 0 . 6 between these patients , the network in Figure 3a of 1 , 497 patients was created . The network layout is based purely on an edge weighted layout algorithm . In order to investigate the clinical characteristics of each cluster , we concatenated the assigned and mined data for all members of a cluster , and calculated a new TF-IDF code vector for the entire cluster in ICD10 space . Figure 3b illustrates these characteristics . | Text mining and information extraction can be seen as the challenge of converting information hidden in text into manageable data . We have used text mining to automatically extract clinically relevant terms from 5543 psychiatric patient records and map these to disease codes in the International Classification of Disease ontology ( ICD10 ) . Mined codes were supplemented by existing coded data . For each patient we constructed a phenotypic profile of associated ICD10 codes . This allowed us to cluster patients together based on the similarity of their profiles . The result is a patient stratification based on more complete profiles than the primary diagnosis , which is typically used . Similarly we investigated comorbidities by looking for pairs of disease codes cooccuring in patients more often than expected . Our high ranking pairs were manually curated by a medical doctor who flagged 93 candidates as interesting . For a number of these we were able to find genes/proteins known to be associated with the diseases using the OMIM database . The disease-associated proteins allowed us to construct protein networks suspected to be involved in each of the phenotypes . Shared proteins between two associated diseases might provide insight to the disease comorbidity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"medicine",
"text",
"mining",
"computer",
"science",
"diagnostic",
"medicine",
"mental",
"health",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2011 | Using Electronic Patient Records to Discover Disease Correlations and Stratify Patient Cohorts |
The human whipworm Trichuris trichiura is a parasite that infects around 500 million people globally , with consequences including damage to physical growth and educational performance . Current drugs such as mebendazole have a notable lack of efficacy against whipworm , compared to other soil-transmitted helminths . Mass drug administration programs are therefore unlikely to achieve eradication and new treatments for trichuriasis are desperately needed . All current drug control strategies focus on post-infection eradication , targeting the parasite in vivo . Here we propose developing novel anthelmintics which target the egg stage of the parasite in the soil as an adjunct environmental strategy . As evidence in support of such an approach we describe the actions of a new class of anthelmintic compounds , the 2 , 4-diaminothieno[3 , 2-d]pyrimidines ( DATPs ) . This compound class has found broad utility in medicinal chemistry , but has not previously been described as having anthelmintic activity . Importantly , these compounds show efficacy against not only the adult parasite , but also both the embryonated and unembryonated egg stages and thereby may enable a break in the parasite lifecycle .
The benzimidazole anthelmintics albendazole and mebendazole are typically used to treat human whipworm infection but are compromised by lack of single-dose efficacy and the risk of resistance . Thus , existing drugs lack sufficient efficacy in mass drug administration ( MDA ) programs to adequately control or potentially eradicate whipworm . This is a major stumbling block in the WHO target to eliminate morbidity from soil transmitted helminthiases in children by 2020 . The current approach for controlling soil-transmitted helminths such as Trichuris is mass drug administration of a single-dose of albendazole or mebendazole , typically repeated annually [1] . However for infection with T . trichiura , single doses of benzimidazoles lead to low cure rates , only 28% and 36% for albendazole and mebendazole respectively [2] . These cure rates are much lower than those of other major human soil-transmitted helminths , Ascaris lumbricoides and hookworm , demonstrating the need for improvements to therapy specifically targeting Trichuris . Indeed modelling studies have demonstrated that , due to these low cure rates , MDA with benzimidazoles does not interrupt whipworm transmission and thus cannot achieve eradication in many settings [3] . Furthermore , the experience from studies on veterinary parasites is that widespread usage of anthelmintics can lead to rapid development of resistance . The discovery of isolates of two species of gastrointestinal nematodes resistant to monepantel only four years after its introduction [4] underlies the real threat to control programmes imposed by emerging drug resistance . Indeed , the combination of MDA programs and low single-dose cure rates may facilitate the development of drug resistance in populations of human parasites . For example , resistance to benzimidazole drugs is caused by point mutations in β-tubulin . Such resistance mutations have been found in T . trichiura after mass drug administration [5] , and have been found to increase in frequency after MDA . High frequency of resistance mutations in a population may be associated with lower egg-reduction rates after MDA [6] . Whilst there is no clear evidence yet of widespread anthelmintic resistance in human populations , identification of new drugs with novel mechanisms of actions is warranted to slow the development of drug resistance . A T . trichiura infection becomes patent when adult female worms , embedded in the gut of the host , start to lay eggs . A single female worm can lay up to 20 , 000 eggs per day and these unembryonated eggs pass out with the faeces and embryonate in the soil . Development only proceeds further if the embryonated eggs are accidentally consumed via contact of the next host with contaminated food , water or soil . Once ingested , signals for hatching are received when the eggs reach the large intestine [7 , 8] , the newly emerged first stage larvae invade the mucosal epithelium and development to the adult stage of the parasite occurs through a succession of larval moults . Importantly , even when active infections are successfully treated , hosts are constantly re-infected due to high levels of infective eggs present within the water and soil , which can remain viable for years . Current anthelmintic programmes , including those targeting Trichuris , focus on post-infection eradication of existing infections . However , lifecycle stages outside of the host are also potential viable targets for small molecule drugs . Thus , both preventing egg embryonation and reducing the infectivity of embryonated eggs prior to ingestion offer targets that would break the parasite lifecycle . The mouse whipworm , T . muris , is a convenient model of the human whipworm as it can be grown routinely in the laboratory via infection of severe combined immune deficiency ( SCID ) mice . Screening ex vivo adult T . muris has been used to test the anthelmintic activity of a variety of compounds , including approved drugs with the potential for repurposing , and also plant extracts [9–11] . We recently reported a small molecule screen utilising an automated assay for assessment of the motility of ex vivo T . muris adults . This screen led to the identification of a class of molecules termed dihydrobenzoxazepinone ( DHB ) which demonstrated encouraging activity in this assay , as well as the ability to reduce in vivo infectivity of treated eggs [12] . Most of the active molecules identified from that screen belonged to the dihydrobenz[e][1 , 4]oxazepin-2 ( 3H ) -one chemotype , but interestingly one additional active was from a completely different structural class . Here we report the identification , synthesis and characterisation of a series of compounds belonging to this second chemotype , which has not previously been described as having anthelmintic activity , the 2 , 4-diamino thieno[3 , 2-d]pyrimidines ( henceforth called diaminothienopyrimidines or DATPs ) .
All animal experiments were approved by the University of Manchester Animal Welfare and Ethical Review Board and performed under the regulation of the Home Office Scientific Procedures Act ( 1986 ) and the Home Office project licence 70/8127 . T . muris worms were cultured using severe combined immune deficiency ( SCID ) mice , at the Biological Services Facility at the University of Manchester . Male and female mice were infected with 200 infective embryonated T . muris eggs via oral gavage . Thirty-five days later , the mice were sacrificed . Adult T . muris were obtained from the intestine as previously described [12] . Worms were maintained in Roswell Park Memorial Institute ( RPMI ) 1640 media supplemented with penicillin ( 500 U/mL ) and streptomycin ( 500 μg/mL ) at approximately 37°C and studied on the same day . Individual adult worms were added to wells containing 75 μL of RPMI-1640 medium , penicillin ( 500 U/mL ) , streptomycin ( 500 μg/mL ) plus 1% v/v final concentration of dimethylsulfoxide ( DMSO ) or compound dissolved in DMSO . Plates were incubated at 37°C , 5% CO2 . Motility was determined after 24 hours . An automated system was used to quantify worm movement . An earlier version of this system has been previously described [13 , 14] . Two hundred frame movies of the whole plate were recorded at 10 frames per second and then motility determined by an algorithm based on thresholding pixel variance over time [15] . For the hit confirmation and expansion assays , library material was used at a final concentration of 100μM . Dose-response curves were calculated with the four factor log-logistic model using the R package drc [16] or using GraphPad Prism . Thin layer chromatography ( TLC ) was performed on aluminium sheets coated with 60 F254 silica . All solvents are used anhydrous unless stated otherwise . NMR spectra were recorded on Bruker AV400 ( 400 MHz ) , Bruker AVII 500 ( 500 MHz ) or AVIIIHD 600 ( 600 MHz ) instruments in the deuterated solvent stated . All chemical shifts ( δ ) are quoted in ppm and coupling constants ( J ) , which are not averaged , in Hz . Residual signals from the solvents were used as an internal reference using the stated deuterated solvent . Infrared spectra were recorded on a Perkin-Elmer 1750 IR Fourier Transform spectrophotometer using thin films on a diamond ATR surface ( thin film ) . Only the characteristic peaks are quoted . Melting points were determined using a Stanford Research Systems EZ-Melt . Low resolution mass spectra ( m/z ) were recorded on an Agilent 6120 spectrometer and high resolution mass spectra ( HRMS m/z ) on a Bruker microTOF mass analyzer using electrospray ionization ( ESI ) . Compounds were synthesised from commercially available starting materials , and fully characterised by Infrared ( IR ) Spectroscopy , Mass Spectrometry ( ESI-MS , HRMS-ESI ) and Nuclear Magnetic Resonance ( 1H and 13C NMR ) . Spectra supporting the synthesis of these compounds are provided in the S1 File . A mixed-stage C . elegans N2 population was obtained by liquid culture ( 20°C ) according to standard methods [17] . It was then bleached to obtain an egg population with 1 . 5 mL 4M NaOH , 2 . 4 mL NaOCl , 2 . 1 mL water , washed three times , and allowed to hatch in 50 mL S-basal buffer at 20°C overnight to obtain a synchronised L1 population . For the growth assay , 49 μL of S-complete buffer and 1 μL of DMSO or DMSO plus compound were added to each well of 96-well plates . 50 μL of a worm suspension ( approximately 20 synchronised L1 worms , 1% w/v E . coli HB101 in S-complete buffer ) were then added to each well . Plates were incubated at 20°C before imaging 5 days later . Worm movement was stimulated by inserting and removing a 96-well PCR plate into/from the wells of the assay plate , and then whole plate 200 frame movies were recorded at 30 frames per second . Growth was quantified as a correlate of movement using the same automated system described earlier , which estimates movement for each well by categorising pixels as imaging movement if their variance is greater the mean plus one standard deviation of the variances of all the pixels on the plate [15] . The mouse rectal epithelial cell line CMT-93 ( LGC Promochem , Teddington , United Kingdom ) was used for these studies . The WST-8 and neutral red cytotoxicity assays were performed as described [12] . Briefly , cells were cultured with test compounds , chlorpromazine positive control or DMSO alone ( final compound concentrations of 0 to 100 μM ) for 72 hours . The WST-8 assay was then carried out using the Cell Counting Kit– 8 ( Sigma Aldrich # 96992 ) with an incubation time of 2 hours . This time was chosen according to the manufacturer’s instructions and was such that the absorbance of the WST-8 formazan dye was within the linear range of the microplate reader . Following this assay , the medium was exchanged , and the ability of the cells to take up the dye neutral red ( concentration 33 μg/mL , incubation time 2 hours ) was determined using a microplate reader ( absorbance at 540 nm ) . Results were analysed using GraphPad Prism and fitted using a log-logistic model . 100 infective embryonated eggs were incubated in deionised water with 1% v/v DMSO or test compounds at a final concentration of 100 μM in 1% v/v DMSO for 14 days at room temperature in the dark . Eggs were then washed and resuspended in deionised water . For in vitro hatching assays 100 eggs were added to 1 mL of E . coli bacterial culture grown in LB broth overnight at 37°C shaking at 200 rpm . Egg-bacterial cultures were incubated for 24 hours at 37°C , 5% CO2 and hatching determined following blinding by visual examination under a dissecting microscope . For in vivo hatching assays , 40 eggs were counted under a dissecting microscope and given to a SCID mouse in 200 μL water . At day 15 post-infection mice were culled and the number of L2 larvae present in the caecae and colon enumerated in a blinded manner under a dissecting microscope . The experiment was conducted in two ‘experimental batches’ . For batch one there were 5 mice in each of the DMSO and OX02926 groups . For batch two there were 9 mice in each of the DMSO and OX02926 groups . The raw data ( number of worms that established infection in each mouse ) are shown separated by batch and treatment in the S1 Fig . To analyse the data we used a two-way ANOVA ( worm number ~ treatment * batch ) . This showed a significant effect of treatment [F ( 1 , 24 ) = 8 . 520 , P = 0 . 00752] . It also showed a significant effect of batch [F ( 1 , 24 ) = 10 . 956 , P = 0 . 00294] . There was no significant interaction between treatment and batch [F ( 1 , 24 ) = 0 . 296 , P = 0 . 59153] . The significant effect of batch reflected that in both DMSO- and OX02926-treated groups , the number of worms that established infection was generally lower in mouse batch 1 than in batch 2 ( S1 Fig ) . Variation in control worm establishment , which is commonplace in Trichuris infections due to natural variation in egg infectivity from a standardised egg number , was within expected ranges . We therefore took the approach of normalising each data point by dividing by the mean of the DMSO-treated group for that batch . This yielded the % batch normalised infection establishment . We used a two-way ANOVA ( % batch normalised infection establishment ~ treatment * batch ) to analyse the data . There was a significant effect of treatment [F ( 1 , 24 ) = 9 . 569 , P = 0 . 00497] but no effect of batch [F ( 1 , 24 ) = 0 . 083 , P = 0 . 77618] or interaction [F ( 1 , 24 ) = 0 . 083 0 . 77618] . We therefore conducted a post-hoc Tukey HSD test which showed that infection establishment in the OX02926-treated group was significantly different from the DMSO-treated control group ( P = 0 . 0050 ) . One hundred unembryonated eggs were treated with water , 1% v/v DMSO in water or test compounds at a final concentration of 100 μM ( unless stated ) with 1% v/v DMSO , in the dark at 26°C , either for 56 days or for shorter periods as described . Images were collected on an Olympus BX63 upright microscope using a 60x / 1 . 42 PlanApo N ( Oil ) objective and captured and white-balanced using an DP80 camera ( Olympus ) in monochrome mode through CellSens Dimension v1 . 16 ( Olympus ) . Images were then processed and analysed using the image analysis platform Fiji [18] .
We have recently described a small molecule screen for new anthelmintics , which used reduction or loss of motility of adult ex vivo T . muris as an endpoint for screening [12] . This screen was designed to identify compounds active on Trichuris as existing drugs are notably less efficacious against this nematode , and it is comparatively evolutionarily distant to nematodes typically screened in anthelmintic-discovery efforts , such as H . contortus , M . incognita and C . elegans . From this primary screen , we found 13 members of the dihydrobenzoxazepinone chemotype , which had not previously been shown to have anthelmintic activity . In this report we describe the identification of a second new anthelmintic chemotype from this screen . A single 2 , 4-diaminothieno[3 , 2-d]pyrimidine ( DATP ) compound was found in the primary screen . This has been given the identifier OX02926 ( Fig 1A ) . We confirmed this activity in a secondary screen using the same source sample ( DMSO solution containing 10 mM compound ) , and also tested a number of structurally-related compounds from our small molecule collection using the same assay ( Fig 1B ) . The rationale for this was to gain greater confidence in the screening hit and also to explore the activity of “near-neighbour” molecules with the same core 2 , 4-diaminothieno[3 , 2-d]pyrimidine structure , which could support the early development of the series . The hit expansion process led to the identification of three further active molecules in this series OX02925 , OX03143 and OX03147 ( Fig 1C ) . Two structurally-related compounds were however not active in this assay ( Fig 1D ) . Having identified promising active DATPs from testing of DMSO solution samples of compounds , these were then resynthesised to obtain authentic , unambiguously characterised samples from which confirmatory screening could take place . Compound resynthesis is important since DMSO solution samples can degrade over time , and this often leads to so-called ‘false positive’ hits [19] . These compounds could be readily prepared in two steps from commercially available 2 , 4-dichlorothieno[3 , 2-d]pyrimidine 1 , via two sequential nucleophilic aromatic substitution reactions . Treatment of 1 with 2- ( 2-chlorophenoxy ) ethylamine or 2-phenoxyethylamine gave exclusively monosubstitution affording 2a and 2b as a single regioisomer in 64% and 80% yield respectively . Subsequent displacement reaction at C4 gave authentic samples of OX02925 , OX02926 , OX03143 and OX03147 in 57–91% yield ( Fig 2 ) . The resynthesized hits were then tested in this screen and a concentration-response curve constructed , thereby confirming the anthelmintic activity of several examples of this structural class ( Fig 3 , Fig 4 ) . This class has ‘lead-like’ or ‘drug-like’ chemical properties [22] , although it is important to note that in the contemporary medicinal chemistry literature this term is usually applied in the context of imparting oral bioavailability characteristics ( Fig 4 ) . For agents targeting the gastrointestinal located Trichuris , minimal systemic exposure of the host is desirable and therefore it is critical to differentiate between the conventionally used terminology and parameters for ‘drug-like’ molecules , which affect solubility and permeability , compared to properties that would be relevant to agents targeting other body compartments . Recent literature has described this important caveat for non-peripheral CNS drugs [23] , and indeed for anti-parasitic drug development [24] . Importantly , there is considerable scope for generating the large number of structural variants of the DATPs needed for the iterative improvement of compound properties during the downstream lead optimisation process . Although we are focused on developing an anthelmintic with improved efficacy over existing drugs against Trichuris , activity across the nematode phylum is valuable , particularly as efficacy against economically significant agricultural animal parasites would make further development more economically viable . We therefore wanted to test the activity of the DATP chemotype against the clade V nematode Caenorhabditis elegans . Using a quantitative development assay to measure the growth of synchronised L1 stage worms , we tested varying concentrations of the compounds to determine the concentration-response effects . As shown in Fig 5 , all four DATP compounds were active in this assay with EC50 values from 7–87 μM . Interestingly , the DATPs display differing trends in activity between the Trichuris and C . elegans assays . At this stage we do not know whether this reflects different potency at the target or different patterns of drug access between the species , but the findings highlight the importance of screening against Trichuris in the search for novel anthelmintic agents targeting whipworm . The data from each of these assays as well as structural descriptors and Lipinski rule assessment for the four DATP compounds and other anthelmintics are summarised in Fig 4 . The leading member of the dihydrobenzoxazepinone class OX02983 is shown in Fig 4 for comparison . EC50 values for the two series are currently in a similar range . It was critical to ensure that this series of compounds showed minimal cytotoxicity towards mammalian cells , and showed selective activity against the parasite . For example , gut cytotoxicity may result in the compounds having too narrow a therapeutic window . Selected examples of the DATPs were assessed for cytotoxicity using the mouse gut epithelial cell line CMT-93 ( Table 1 ) . Although , the DATPs exhibited increased in vitro cytotoxicity in these assays compared to the previously reported DHB series [12] , an encouraging overall profile was exhibited for these early stage molecules . Furthermore , the nematode cuticle often limits drug access which reduces target engagement by small drug-like molecules [25 , 26] . This means that compound optimisation to improve uptake through the cuticle may be a fruitful route to improved anti-nematode selectivity , as well as improving the cytotoxicity profile . It is interesting to note that the activity against Trichuris did not correlate with cytotoxicity , with the most cytotoxic compound ( OX03143 ) showing the lowest activity in the T . muris adult paralysis assay , with an EC50 > 80μM . This suggests that either anti-Trichuris activity is distinct from cytotoxic action , or that differential drug access can be exploited to achieve differential host-parasite activity . Either possibility is encouraging and suggests that continued exploration and iterative improvement of the DATP structure might be anticipated to deliver a more potent anthelmintic with acceptable host toxicity . Developing novel anthelmintics to disrupt the T . trichiura life cycle at the egg stage represents an exciting and complementary strategy to an oral therapy and is particularly attractive as T . trichiura eggs are highly resistant to extreme temperature changes and ultraviolet radiation , thereby remaining viable in the environment for many years [27] . We assessed whether the DATP derivatives were capable of affecting either infection establishment or embryonation of eggs . We first explored whether the compounds could alter the establishment of infection by soaking embryonated T . muris eggs in the test compounds for 14 days , washing the eggs and then determining infectivity both in vitro and in vivo ( Fig 6A ) . To determine effects on in vitro hatching , a protocol modified from that previously described [8] was established whereby eggs were induced to hatch when incubated in a culture of Escherichia coli at 37°C . The results are summarised in Fig 6B . Strikingly all DATPs were capable of significantly reducing in vitro hatching compared to the DMSO control . To extend this finding , we selected OX02926 to test in an in vivo hatching and infection establishment assay , as this compound showed both a significant decrease in in vitro hatching and a small standard deviation between samples . The eggs were soaked as for the in vitro experiment and SCID mice were infected with 40 treated eggs ( OX02926 or DMSO ) by oral gavage . Egg infectivity was quantified at day 15 post infection by culling the mice and counting the number of established L2 larvae in the gut . All L2 larvae counted had a normal morphology as viewed under a dissecting microscope . This experiment was carried out in two batches and the raw data are shown in the S1 Fig . Because variation in control worm establishment is commonplace in Trichuris infections due to natural variation in egg infectivity from a standardised egg number , we took the approach of normalising data for each batch relative to the mean of the DMSO-only control group for that batch . This allowed us to determine the effects of OX02926 treatment ( a full statistical description is given in the Methods section ) . The results are shown in Fig 6C . We used a two-way ANOVA ( % batch normalised infection establishment ~ treatment * batch ) to analyse the data . There was a significant effect of treatment [F ( 1 , 24 ) = 9 . 569 , P = 0 . 00497] but no effect of batch [F ( 1 , 24 ) = 0 . 083 , P = 0 . 77618] or interaction [F ( 1 , 24 ) = 0 . 083 0 . 77618] . We therefore conducted a post-hoc Tukey HSD test which showed that infection establishment in the OX02926-treated group was significantly different from the DMSO-treated control group ( P = 0 . 0050 ) . Treatment of eggs with OX02926 was able to significantly reduce the burden of worms in vivo by an estimated 40% . This is likely to reflect reduced infectivity of DATP-treated eggs . The ability of the DATPs to alter the embryonation of T . muris eggs was investigated by soaking unembryonated T . muris eggs collected overnight from live adult T . muris in the test compounds at 26°C for the duration of the embryonation process ( 56–60 days ) . During embryonation the first larval stage of the parasite develops within the egg shell ( Fig 7A ) from a ball of cells ( Fig 7B ) . Treatment with the DATPs OX02925 and OX03147 resulted in a significant increase in the percentage of unembryonated eggs present compared to the DMSO control ( Fig 7C ) . Importantly , although the other DATPs did not alter the percentage of eggs unable to undergo the embryonation process , the larvae that developed were atypical ( Fig 7D–7I ) . These atypical larvae were morphologically altered with the granules present within the larvae appearing less distinct . As OX03147 had the clearest phenotype with a significant increase in the number of unembryonated eggs , a concentration response study was performed to determine if an effect could be seen at lower treatment doses . Additionally , we repeated the experiment at room temperature to allow for more physiological conditions rather than the constant 26°C utilised in the initial study to standardise conditions across experiments . Although the increased number of unembryonated eggs was only detected at the highest drug dose tested ( 100 μM ) at both 26°C and room temperature ( Fig 7C and 7J ) striking effects on egg morphology was detectable at concentrations as low as 1 μM with significant larval stunting observed ( Fig 7K ) . To determine if an effect on embryonation could be observed following a shortened drug exposure we soaked unembryonated eggs in 100 μM OX03147 at 26°C for weeks 0–2 , 0–3 , 2–4 or 4–6 of embryonation . Although there was no increase in the proportion of unembryonated eggs observed in any treatment group ( S2F Fig ) there were clear morphological alterations in the L1 larvae within the egg following exposure to OX03147 during weeks 0–3 , 2–4 or 4–6 of the embryonation process ( S3A–S3E Fig ) . The most striking observation was the clear larval stunting observed following drug soaking from weeks 0–3 ( S2C and S2G Fig ) . A one-way ANOVA test showed a significant effect of treatment on larval length , F ( 4 , 21 ) = 3 . 984 , P = 0 . 0147 . A post-hoc Dunnett’s test showed a significant difference in the Weeks 0–3 treatment group compared to the DMSO-only control group ( P = 0 . 0076 ) . This appeared to phenocopy the effect OX03147 had when treated for the duration of the embryonation process at 1 μM ( Fig 7K ) . Additionally , in the 2–4 week and 4–6 week groups , although larval length was not affected , there was evidence of structural alterations in the L1 larvae with a less distinct structure present and altered granulation within the larvae ( S2D and S2E Fig ) . To the known range of applications of DATPs in medicinal chemistry we can now add anthelmintic activity . This study suggests they have significant potential for further development into dual-acting therapeutic agents for both the reduction of Trichuris egg infectivity , and embryonation in the environment . Thus , their actions on both the embryonated and unembryonated egg stages may enable a break in the parasite lifecycle .
We recently reported a small molecule screen for new anthelmintics targeting the gastro-intestinal ( GI ) nematode parasite Trichuris muris that identified the dihydrobenzoxazepinone ( DHB ) chemotype . The DHBs had not previously been ascribed anthelmintic activity [10] . Here , we describe a second class of novel anthelmintic , the diaminothienopyrimidines ( DATPs ) . The potential for this early stage series is significant; their chemical synthesis is facile and lends itself to iterative optimisation , which will facilitate structural modifications aiming , for example , to increase local epithelial penetrance and hence improve efficacy during future development . Furthermore , their straightforward production imparts a favourable cost benefit aspect to the series . Thienopyrimidines have received much interest in medicinal chemistry as they are bioisosteres for purines , such as the nucleic acid components adenine and guanine . They are also related to quinazolines , an important class of kinase inhibitors , including gefitinib and erlotinib , which act by recognizing the ATP-binding site of the enzyme [30] . Thieno[2 , 3-d]pyrimidines are a particularly important scaffold , with many reported examples of protein kinase inhibitors , as well as inhibitors of dihydrofolate reductase , kainate receptor agonists , and α1-adrenoreceptor antagonists [31] . The thieno[3 , 2-d]pyrimidine scaffold found in the compounds reported in this study , has also been investigated . A series of 2-aryl 4-morpholino derivatives have been identified as phosphatidylinositol-3-kinase inhibitors [32] , leading to the discovery of the PI3K inhibitor GDC-0941 ( pictilisib ) [33] and the dual PI3K/mTOR inhibitor DGC-0980 ( apitolisib ) [34] . The structures of these compounds are shown in the S3 Fig , in comparison with the 2 , 4-diaminothieno[3 , 2-d]pyrimidine OX02926 . Pictilisib and apitolisib are under development as anti-cancer agents , have been tolerated in Phase I trials for solid tumors , and Phase II trials have commenced [35 , 36] . A series of 2 , 4-diaminothieno[3 , 2-d]pyrimidines have been described as orally active antimalarial agents [37] , with activity in the low nanomolar range against Plasmodium falciparum . The structures of these compounds are shown in the S3 Fig in comparison with OX02926 . This anti-malarial series was later improved by systematic modification giving improved antimalarial activity , but unfortunately continued hERG inhibition [38] . Whilst our DATP compounds have the same core scaffold as the anti-malarial series , they have different substituents , and in particular lack the 6-aryl substituent that is critical for anti-malarial activity and found in all compounds tested for hERG activity . However , the authors were able to demonstrate that hERG activity could be removed through modification of the C1 substitutuent , suggesting that this potential liability is not instrinsic to the 2 , 4-diaminothieno[3 , 2-d]pyrimidine core . We have not yet performed hERG assessment of our compounds , but this will form an important part of the future development of this series . A series of 2 , 4-diaminothieno[3 , 2-d]pyrimidines has also recently been reported as active against the endosymbiotic bacterium Wolbachia , with potential use against filarial nematodes [39] . In neither the anti-malarial or anti-Wolbachia case is the molecular target of the compounds known . The major goal of our research is to develop a new oral therapy for trichuriasis , which could be widely used in mass drug administration programs leading to the eradication of human whipworm . Such an agent should have a substantially higher single-dose cure rate than the current drugs used in mass drug administration , albendazole and mebendazole . Drug development is long process , and recent work has defined a set of criteria , tailored to neglected infectious diseases , for progression in the hit to lead and lead optimisation stages [40 , 41] . Our DATP series members are early-stage compounds in the development process . The compounds meet almost all of the criteria for hit selection in neglected diseases , including confirmed activity with resynthesized material , dose-dependent in vitro activity , a tractable chemotype that passes drug-likeness filters such as the Lipinksi rule of five , and an established synthetic route of only two steps [40] . The most pressing weakness of the series is the small selectivity window for their activity against the parasite compared to cytotoxicity in a mammalian cell line . Improving this property for these early stage compounds must be a priority for future development . The DATP compounds also meet some of the milestones in the hit to lead process , particular in terms of drug-likeness and the exploitability of the structure , giving the ability to generate variants and establish the structure-activity relationship and hence improve potency and selectivity [41] . The in vitro activity of OX02926 in the adult whipworm motility assay ( EC50 = 27μM , equivalent to 10 . 2μg/ml ) also reaches the activity threshold for lead compounds that has been determined for drug development against the microfilarial nematode Brugia malayi [41] . In summary the DATP series are promising early-stage compounds with a number of lead-like features . Improvement of potency , together with an understanding of parasite/host selectivity and pharmacokinetic properties will be the focus of the next steps of development . In addition to activity against the adult stage of whipworm , the DATPs were also able to significantly reduce egg hatching , both in vitro and in vivo . These data are in keeping with members of the DHB series , which also were able to inhibit parasite egg hatching . However , unlike the DHB series , we identified members of the DATPs that also significantly reduced the percentage of eggs embryonating ex vivo , with other members of the DATP series appearing to disrupt the embryonation process , resulting in defects in embryonic elongation and abnormal egg shape . Trichuris egg embryonation occurs gradually and the mechanism by which it occurs is currently a poorly understood process . A detailed characterisation of the morphological changes which occur with the Trichuris suis egg during embryonation has been described and other Trichuris species appear to undergo the same process . Once the unembryonated , unsegmented eggs are deposited , the two clear , nuclei-like areas move together and fuse . Cellular division then begins , initially occurring asymmetrically with two blastomeres of unequal size . The larger blastomere then divides again and then subsequently each blastomere divides in two until a blastula formed of many small blastomeres develops . The initial larval differentiation then occurs with the appearance of a motile cylindrical embryo , which gradually turns into an infective larva with its characteristic oral spear . The fully developed larva is no longer motile and is thought to be an L1 larva as no moult is observed within the egg [42] . The embryonation process is temperature sensitive . The effect of temperature on egg embryonation has been characterised in detail in recent years for T . suis eggs with the embryonation process accelerated at 30–32°C compared to 18°C , with degeneration of the eggs rather than embryonation observed at higher temperatures ( 40°C ) . At low temperatures ( 5–10°C ) no embryonation occurs , however once these eggs are then transferred to optimal embryonation temperatures normal embryonation proceeds [43] . Similar temperature sensitivity has been described for other Trichuris species including Trichuris trichiura with different species embryonating with different kinetics [44 , 45] . More research is required to understand the mechanisms behind this embryonation process , which may then allow an even more targeted approach to breaking the life cycle . Humans become infected with Trichuris via a faecal oral route . Adult parasites in the intestine shed unembryonated eggs , which pass out with the faeces and embryonate in the external environment over a period of five weeks . Eggs can remain viable in the environment for many months [46] . Parasite eggs are only infective if fully embryonated upon ingestion . Thus , the ability of the DATPs to disrupt both the infectivity of embryonated eggs and the embryonation process itself suggests a potential environmental control to decrease Trichuris infection rates in the field without the need to develop and administer a new oral anthelmintic to the infected population . In particular , it has been noted that the environmental pool of infectious eggs makes those individuals successfully treated , typically once or twice per year , in mass drug administration programs at risk of reinfection [47] . It has therefore been proposed that improvements in sanitation are required in addition to anthelmintic MDA . We suggest that an environmentally-acting , egg-targeting agent , potentially developed from our DATP series compounds , could play a complementary role to help break transmission in parallel with MDA and santitation improvements . Clearly it is not possible to widely treat large areas of endemic regions with such an environmental control . Instead , we envisage the targeted use of DATPs in the environment at sites of high parasite egg density; these might include for example focusing treatment around pit latrines , as it is known that pit latrines may be a focal point of infection with a high concentration of eggs of soil-transmitted helminths [48] . In a study in Ethiopia , Trichuris trichiura prevalence was higher in communities with greater latrine usage ( compared to field or yard defecation ) , suggesting that basic pit latrines may in some circumstances be ineffective at reducing infection [49] . However improved sanitation facilities generally , including pit latrines , ventilated improved pit latrines , and flush toilets , do reduce STH infection rates [47 , 50] . Such an egg-targeted agent should have a limited negative effect on the environment , have a suitable formulation for practical delivery , and be able to block egg viability at low concentration in the environment . The DATP series , which damage egg development and infectivity when applied at fairly high concentrations ( 1 to 100μM ) for quite long periods of time ( from 2 to 3 weeks to 60 days ) show potential for developing such an agent . However these properties need to be improved during future development , while achieving an appropriate safety and environmental profile .
In summary we report the discovery of a new class of anthelmintic , the DATPs , which possesses activity directed against adult stage T . muris parasites and the egg stage . Importantly , as a chemical series the DATPS are notable , since they are relatively facile to produce synthetically thereby presenting considerable scope for structural modifications to improve efficacy and deliver an optimised agent . | The human whipworm , Trichuris trichiura , infects around 500 million people globally , impacting on their physical growth and educational performance . There are currently huge mass drug administration ( MDA ) programs aiming to control whipworm , along with the other major soil transmitted helminths , Ascaris and hookworm . However single doses of albendazole and mebendazole , which are used in MDA , have particularly poor effectiveness against whipworm , with cure rates less than 40% . This means that MDA may not be able to control and eliminate whipworm infection , and risks the spread of resistance to albendazole and mebendazole in the parasite population . We are attempting to develop new treatments for parasitic worm infection , particularly focused on whipworm . We report the identification of a class of compounds , diaminothienopyrimidines ( DATPs ) , which have not previously been described as anthelmintics . These compounds are effective against adult stages of whipworm , and also block the development of the model nematode C . elegans . Our DATP compounds reduce the ability of treated eggs to successfully establish infection in a mouse model of human whipworm . These results support a potential environmental spray to control whipworm by targeting the infectious egg stage in environmental hotspots . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion",
"Conclusions"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"caenorhabditis",
"tropical",
"diseases",
"parasitic",
"diseases",
"animals",
"toxicology",
"animal",
"models",
"developmental",
"biology",
"caenorhabditis",
"elegans",
... | 2018 | 2,4-Diaminothieno[3,2-d]pyrimidines, a new class of anthelmintic with activity against adult and egg stages of whipworm |
The assembly of ribosomal subunits in eukaryotes is a complex , multistep process so far mostly studied in yeast . In S . cerevisiae , more than 200 factors including ribosomal proteins and trans-acting factors are required for the ordered assembly of 40S and 60S ribosomal subunits . To date , only few human homologs of these yeast ribosome synthesis factors have been characterized . Here , we used a systematic RNA interference ( RNAi ) approach to analyze the contribution of 464 candidate factors to ribosomal subunit biogenesis in human cells . The screen was based on visual readouts , using inducible , fluorescent ribosomal proteins as reporters . By performing computer-based image analysis utilizing supervised machine-learning techniques , we obtained evidence for a functional link of 153 human proteins to ribosome synthesis . Our data show that core features of ribosome assembly are conserved from yeast to human , but differences exist for instance with respect to 60S subunit export . Unexpectedly , our RNAi screen uncovered a requirement for the export receptor Exportin 5 ( Exp5 ) in nuclear export of 60S subunits in human cells . We show that Exp5 , like the known 60S exportin Crm1 , binds to pre-60S particles in a RanGTP-dependent manner . Interference with either Exp5 or Crm1 function blocks 60S export in both human cells and frog oocytes , whereas 40S export is compromised only upon inhibition of Crm1 . Thus , 60S subunit export is dependent on at least two RanGTP-binding exportins in vertebrate cells .
The synthesis of ribosomal subunits is a major cellular task in all living organisms . Ribosomal subunit assembly requires deposition of numerous ribosomal proteins on ribosomal RNA ( rRNA ) to yield two ribonucleoprotein particles , a small and a large subunit , that together form a ribosome competent for protein translation . Our knowledge about ribosomal subunit biogenesis is most advanced for unicellular model organisms such as bacteria ( Escherichia coli ) and budding yeast ( S . cerevisiae ) . Pioneering early experiments demonstrated that simple subunits such as E . coli 30S [1] and 50S [2] can be reconstituted from rRNA and ribosomal proteins in vitro . Still , in vivo various non-ribosomal factors assist subunit assembly by promoting rRNA folding , modification , and protein deposition [3] , [4] . With the increase in ribosome complexity and the establishment of cellular compartments in eukaryotes , ribosome biogenesis became dependent on a greater variety and number of trans-acting factors [5]–[7] . In yeast , more than 200 proteins are necessary for efficient ribosomal subunit assembly , which starts in the nucleolus with RNA polymerase I-mediated transcription of a large rRNA precursor molecule containing rRNA pieces of both subunits . Co-transcriptionally , the first trans-acting factors and ribosomal proteins assemble with the nascent pre-rRNA , generating an 80–90S precursor particle [8] , [9] . During nuclear pre-rRNA maturation , numerous small nucleolar RNPs ( snoRNPs ) are involved in guiding and promoting various rRNA modification and processing reactions . The U3 snoRNP as part of the small subunit ( SSU ) processome is required for early endonucleolytic cleavage of the pre-rRNA at sites A0 , A1 , and A2 [6] . Cleavage at site A2 results in the conversion of the 90S pre-ribosome into separate pre-40S and pre-60S particles [10] . Subsequently , the two subunit precursors mature independently of each other through a series of rRNA processing and RNP remodeling steps in the nucleolus and nucleoplasm . After export from the nucleus , both subunits are subjected to final maturation in the cytoplasm , including the release of trans-acting factors and the completion of rRNA processing [11] , [12] . Systematic approaches such as proteomic analysis of preribosomal particles and genetic screens have greatly helped the rapid advancement of our understanding of ribosome biogenesis in yeast and have yielded a comprehensive picture of the biogenesis pathway [7] , [13]–[20] . In higher eukaryotes , this global view is still lacking , but particular aspects of ribosome biogenesis such as rRNA transcription and processing , snoRNP biogenesis and function , and nuclear export have been studied in different vertebrate model systems such as cultured cells and frog oocytes [6] , [21]–[23] . These studies revealed that many aspects of ribosome synthesis are well conserved from yeast to human and many yeast trans-acting factors possess functional homologs in vertebrate species [12] , [24]–[38] . However , for most human homologs of yeast trans-acting factors , a function in ribosome biogenesis has not yet been established . Moreover , vertebrate ribosome synthesis is dissimilar in several aspects , including the genomic organization of rDNA [39] , differences in rRNA processing [6] , [31] , [40]–[43] , and links to stress response pathways unique for higher eukaryotes [44] . Here we present a systematic approach to directly assess the functional requirement of 464 selected proteins in ribosome maturation in human cells . We depleted these factors by RNAi and scored for ribosome maturation defects using microscopic readouts . Our analysis provides a list of 153 proteins required for human ribosome synthesis . For 40S biogenesis , we can attribute their requirement to nucleolar , nucleoplasmic , or cytoplasmic maturation steps . Collectively , our data show that core features of 40S biogenesis are indeed conserved from yeast to humans . However , we also uncover unexpected differences such as the requirement of the pre-miRNA export receptor Exp5 for 60S subunit export to the cytoplasm .
To identify proteins required for ribosome biogenesis by RNAi in HeLa cells , we established visual readouts suitable to detect defects in ribosomal subunit maturation in a high-throughput manner using microscopic endpoint assays . For the 40S subunit , we applied two different readouts [45] , namely an inducible , fluorescent ribosomal reporter protein , Rps2-YFP , and immunofluorescence ( IF ) analysis of the trans-acting factor Enp1 ( BYSL ) [24] . Tetracycline-induced expression of Rps2-YFP allows for selective visualization of newly synthesized 40S subunits [45] . After 14 h of induction , Rps2-YFP-containing 40S have largely completed maturation , giving rise to a prominent cytoplasmic signal of the reporter ( Figure 1A , control RNAi ) . RNAi against Crm1 , the RanGTP-dependent nuclear export receptor required for both pre-40S and pre-60S export [34] , [35] , resulted in a nucleoplasmic accumulation of Rps2-YFP , reflecting the dependence of pre-40S export on Crm1 . In contrast , inhibition of nucleolar steps of ribosome biogenesis , as exemplified by depletion of the box C/D snoRNPs-associated methyltransferase fibrillarin ( Fbl ) [10] , led to accumulation of Rps2-YFP in nucleoli ( Figure 1A ) . Thus , two phenotypes can be distinguished using the Rps2-YFP reporter , reflecting early ( nucleolar ) and late ( nucleoplasmic ) defects in nuclear pre-40S maturation . The IF analysis of Enp1 complements the Rps2-YFP readout , as it allows for detection of both nuclear and cytoplasmic defects in 40S biogenesis . Enp1 is nucleolar at steady state but accompanies pre-40S to the cytoplasm from where it is recycled during cytoplasmic subunit maturation [45] . Inhibition of 40S export ( Crm1 RNAi ) resulted in a nucleoplasmic accumulation of Enp1 , whereas interference with cytoplasmic 40S maturation led to cytoplasmic Enp1 localization , as observed upon downregulation of the 40S trans-acting factor Ltv1 ( Figure 1A ) . Altogether , these readouts enable us to monitor the entire 40S maturation pathway , from the nucleoli through the nucleoplasm to the cytoplasm , and to identify factors required for subunit assembly and nuclear export as well as recycling of trans-acting factors in our RNAi screening approach . To monitor 60S biogenesis , we established a HeLa cell line carrying an inducible copy of the ribosomal reporter protein Rpl29-GFP . Rpl29-GFP is efficiently incorporated into 60S subunits ( Figure S1 ) , and the GFP-readout faithfully illustrates the Crm1 dependency of 60S export ( Figure 1A ) . Thus , this cell line is suitable to study nuclear 60S biogenesis , but currently we do not have the means to detect defects in cytoplasmic 60S maturation . Based on curation of the literature , we compiled a list of proteins with a known or potential function in human ribosome biogenesis ( Figure 1B and Table S1 ) , including ribosomal proteins , nuclear pore complex ( NPC ) components , proteins of the nucleo-cytoplasmic trafficking machinery , and potential or established human homologues of yeast trans-acting factors . Additionally , we included factors that function in various other cellular pathways that may or may not impact on ribosome biogenesis , like mRNA or tRNA metabolism , protein modification , and degradation ( miscellaneous category ) . The RNAi screen was performed in a 96-well format using three different siRNAs per target at an siRNA concentration of 10 nM . On each plate , three negative ( Allstars siRNA , Qiagen ) and three positive ( Crm1 siRNA ) controls were present . Image analysis was performed in an automated fashion , using image segmentation based on the identification of cell nuclei by Hoechst fluorescence as a first step . Then , 30 different features including the mean nuclear and cytoplasmic fluorescence intensities , morphological descriptors , and texture marks were extracted from the images using a customized version of Cell Profiler [46] . For example , nucleolar and nucleoplasmic accumulation of Rps2-YFP were distinguished by textural features . Phenotypic analysis of the dataset was performed applying a newly developed supervised machine learning software , the Advanced Cell Classifier ( http://acc . ethz . ch ) . Based on this analysis , we obtained for each siRNA and readout a numerical score , termed hit rate , that was defined by the ratio of cells displaying phenotypes indicative of defects in ribosome synthesis to all reporter-positive interphase cells ( Figure S2 and Text S1 ) . SiRNAs causing a strong defect on cell growth were excluded from subsequent analysis . Based on the hit rates , we further selected all targets , which were represented by at least two siRNAs resulting in more than twice the hit rate of the negative controls . Subsequently , these targets were ranked relative to both the negative and positive controls of the respective 96-well plate ( for details , see Text S1 ) . Next , we defined a cutoff ( see Text S1 ) , generating a high confidence hit list of targets that are involved in ribosome biogenesis based on the phenotypic classification of the different readouts ( Figure 2 ) . Lists of all numerical data are given in Tables S2 , S3 , and S4 . All primary images are stored in a newly developed database , accessible via a web browser interface at http://hcpb . ethz . ch . We note that this database has all the required features to provide a framework for integrating future high content RNAi screening results . Altogether , we identified 153 targets that displayed a phenotype in one or several of our readouts ( Figure 2 ) . A Venn diagram of these high confidence hits ( Figure 3A ) shows that there is a large coherence for the 40S readouts , with 84% of nuclear Enp1 IF hits also scoring as Rps2-YFP hits . Interestingly , there is also a significant fraction of targets common to nuclear steps in both 40S and 60S biogenesis ( see below ) . Hits shared between all readouts are highly enriched for nucleoporins , the constituents of NPCs . Analysis of the hit distribution with respect to the functional categories of targets revealed that , as expected , ribosomal proteins were retrieved to a large extent , whereas only very few hits were found within the miscellaneous category , further illustrating the specificity of the used readouts ( Figure 3A , B ) . To test the reproducibility of the screening results , we performed two tests . First , the screen was repeated for a random subset of plates ( ∼50% ) . This analysis verified the high quality of the hit annotation , as 88% of the included high confidence hits for Rps2-YFP , 86% for Enp1 , and 88% for Rpl29-GFP were confirmed ( unpublished data ) . Notably , original hits that were not reproduced mostly turned up either just below the cutoff or were confirmed by only one siRNA during repetition . Second , to address whether the siRNA concentration was limiting , we repeated the whole analysis for the Rps2-YFP readout at an siRNA concentration of 25 nM ( Figure S3 ) . Importantly , this analysis confirmed 93% of our hits obtained at 10 nM . Although we obtained more hits at 25 nM , the higher siRNA concentration caused stronger effects on cell growth and likely increased off-target effects . Exportin 5 ( Exp5 ) was detected in the screen as a 60S-specific hit . Because only one siRNA gave rise to a robust phenotype in the screen , we verified the result using two previously validated siRNAs against Exp5 ( [56] , unpublished data ) . Depletion of Exp5 indeed resulted in nuclear accumulation of Rpl29-GFP ( Figure 5 ) , comparable to depletion of the positive controls Crm1 and Nmd3 , which serves as an adaptor protein for Crm1 in 60S export [34] , [35] , [57] , [58] . Northern blot analysis using an ITS2-specific probe that detects precursors of 28S rRNA did not reveal any rRNA processing defects induced by depletion of Exp5 ( Figure S5 ) . Exp5 is an RNA-binding exportin that functions in nuclear export of miRNA precursors [56] , [59] , [60] . Recently , miRNA-10a was shown to positively regulate the expression of certain mRNAs , including mRNAs coding for ribosomal proteins of both the small and the large subunit [61] . These data showed that both 40S and 60S biogenesis were slightly compromised upon miRNA-10a inhibition . We therefore tested whether components of the miRNA biogenesis pathway would be required for 40S and 60S synthesis . In these experiments we found no evidence for a phenotype in nuclear 60S biogenesis similar to that caused by Exp5 depletion ( Figures S6 , S7 ) . Depletion of Exp5 also gave rise to detectable defects in 40S biogenesis , as based on nuclear accumulation of the Rps2-YFP reporter ( Figure S6 , Table S2 , and unpublished data ) . However , defects in 60S biogenesis were more prominent . Thus , there appears to be a stronger dependence of 60S than 40S biogenesis on the presence of Exp5 . Taken together , an impairment of miRNA synthesis or function seemed insufficient to explain the strong 60S-specific phenotype for Exp5 depletion in our assay . Therefore , we next addressed whether Exp5 could have a direct function in 60S export . An exportin/cargo relationship between Exp5 and pre-60S particles would predict a physical , RanGTP-regulated interaction between Exp5 and pre-60S particles [62] . To test this , we first purified a pre-60S particle from HEK293 cells by tandem affinity purification ( TAP ) using the TAP-tagged trans-acting factor MRTO4 ( homologous to yeast Mrt4 ) as bait ( Figure 6 and Figure S8 ) . Protein composition analysis of this particle showed the presence of 60S trans-acting factors like the export adaptor Nmd3 , C15orf15 ( Rlp24 , ribosomal-like protein 24 ) , and ribosomal proteins of the large subunit ( Figure 6B and Figure S8A ) . These particles were then incubated with HeLa cell extract in the absence or presence of RanQ69L-GTP , a Ran mutant locked in the GTP-bound state , to analyze RanGTP-dependent binding of nuclear export receptors to pre-60S . Note that the HeLa cell extract had been pre-depleted for ribosomes and tRNAs . The depletion of ribosomes ensured that the TAP-purified particles were the sole ribosomal particles present . The presence of tRNA might hamper the detection of a potential Exp5/60S interaction because tRNA binds efficiently to Exp5 [63] , [64] . Reflecting their requirement for pre-60S export , both Exp5 and Crm1 bound to this pre-60S particle in a RanGTP-dependent manner , whereas exportin-t ( Exp-t ) , another exportin that was not detected as a hit in the screen , did not bind ( Figure 6B ) . We confirmed these results using a pre-60S particle purified with ZPR9 ( REI1 ) , another TAP-tagged 60S trans-acting factor ( Figure S8B ) . The addition of competitor pre-miRNA to the binding reaction inhibited Exp5 binding but left Crm1 binding unaffected ( Figure S8C ) , showing that Crm1 binds independently of Exp5 and indicating that Exp5 uses its RNA-binding interface for pre-60S particle association . Overall , the observed RanGTP-dependent binding of Crm1 and Exp5 to pre-60S TAP particles reflects a bona fide exportin/cargo interaction . Since the depletion of Exp5 had also affected 40S biogenesis , albeit to a lesser extent than 60S synthesis , we analyzed whether Exp5 binding is 60S-specific or can also be observed for pre-40S particles . Pre-40S particles were similarly purified by TAP using the 40S trans-acting factor PNO1 ( Dim2 ) as bait ( Figure 6C ) . These pre-40S particles contain other expected 40S trans-acting factors such as Rio2 and Nob1 , as well as 40S ribosomal proteins ( Figure 6C and Figure S8A ) . Crm1 bound to this particle in a RanGTP-dependent manner and the particle can therefore be considered an export-competent pre-40S . In contrast , Exp5 did not bind to this pre-40S particle ( or another purified pre-40S particle , unpublished data ) , further illustrating the specificity of 60S binding . These data substantiate a direct role for Exp5 in 60S export . It should be pointed out that RanGTP-dependent binding of Crm1 to pre-40S and pre-60S particles has not been demonstrated previously . Thus , this analysis of exportin binding to pre-ribosomal subunits also provides the first biochemical evidence for the prevailing model of a direct role of Crm1 in the export of both ribosomal subunits . To confirm the function of Exp5 in 60S export in another vertebrate species and by a different experimental approach , we analyzed the exportin dependence of rRNA export in Xenopus oocytes . Neutralizing antibodies specific for Crm1 or Exp5 were injected into oocyte nuclei , and subsequently the maturation and export of newly made , radio-labeled rRNAs were assessed after dissection of the oocytes into nuclear ( N ) and cytoplasmic ( C ) fractions ( Figure 7 ) . Under control conditions , various pre-rRNA species are detected in the nuclear fraction , whereas mature rRNAs of the 60S subunit ( 28S rRNA , 5 . 8S rRNA ) and the 40S subunit ( 18S rRNA ) are found in the cytoplasmic fraction . Injection of Crm1 antibodies resulted in nuclear accumulation of both 28S and 18S rRNAs , consistent with previous experiments [34] , [35] . Upon injection of the Exp5 antibody , 28S rRNA and 6S rRNA , the precursor to 5 . 8S rRNA , accumulated in the nucleus , but 18S rRNA export was unaffected . Thus , Exp5 is the second exportin besides Crm1 required for 60S export in higher eukaryotes . In yeast , deletion of Msn5 , the homolog of Exp5 , does not impair ribosomal subunit export [18] , [19] , indicating that pre-60S export is different between yeast and man . Moreover , recent studies have identified two additional export factors besides Crm1 that contribute to 60S export in yeast: the general mRNA export receptor Mex67/Mtr2 and the trans-acting factor Arx1 [65]–[67] . The domains in Mex67/Mtr2 and Arx1 required for pre-60S binding and NPC interaction , respectively , are not conserved from yeast to humans; hence , their human homologs are not expected to contribute to 60S export [65] , [67] . Notably , we did not detect the human homolog of Mex67 ( Tap , NXF1 ) on our purified pre-60S particles ( Figure S8B ) and also found no evidence for the involvement of Tap or EBP1 ( Arx1 ) in 60S biogenesis in our screen . However , we note that in yeast the depletion of Arx1 alone causes no 60S export defect [65] . Thus , at least two export receptors are required for 60S export in both humans and yeast . Here , we demonstrate for the first time the binding of two exportins to a pre-60S particle . Importantly , neither receptor on its own is sufficient to promote 60S export in vivo ( Figure 7 ) . Together , these data suggest that a particle as big as a 60S pre-ribosome requires more than one transport receptor for NPC passage . The usage of the pre-miRNA export receptor Exp5 for 60S subunit export opens the possibility for crosstalk between miRNA biogenesis and 60S export , but it remains to be seen whether cells make use of this . In yeast , a crosstalk between mRNA and 60S export has been proposed based on the shared usage of the export receptor Mex67/Mtr2 [68] . We present a first , partial protein inventory for human ribosome biogenesis . The scope of this RNAi screen was to evaluate candidate genes for human ribosome synthesis , building largely on the knowledge about eukaryotic ribosome biogenesis obtained in yeast . We provide a list of 153 proteins ( including 91 non-ribosomal proteins ) that are required for ribosome synthesis in human cells . Moreover , both primary and analyzed data of all investigated targets can be accessed through an online database ( http://hcpb . ethz . ch ) . Additionally , we demonstrate a direct requirement for the RanGTP-binding transport receptor Exp5 in the biogenesis of 60S subunits in vertebrates . The presented RNAi screen can be extended in future to perform unbiased genome-wide searches for additional factors that participate in ribosome biogenesis . We anticipate that many features of the process in vertebrates will be found to be analogous to those in yeast , but differences have already been identified . These include proteins that participate in ribosome biosynthesis in yeast or humans and are not conserved between yeast and vertebrates , for example vertebrate nucleophosmin [69]–[71] , as well as the many uncharacterized proteins present in human nucleoli [72] , [73] . Moreover , regulation of ribosome synthesis is likely to be more complex in vertebrates than in unicellular organisms . Systematic RNAi together with complementing approaches might ultimately lead to a comprehensive inventory of proteins involved in this fundamental process in mammalian cells and may pave the way towards a deeper mechanistic knowledge . Given the emerging evidence for numerous links between human diseases and ribosome biogenesis/function [74]–[76] , such progress is eagerly awaited .
HeLa Rps2-YFP has been described [45] . The HeLa Rpl29-GFP cell line was generated by integrating Rpl29-GFP ( cloned into the KpnI and NotI sites of pcDNA/FRT/TO; Invitrogen ) into HeLaK FRT TetR cells [45] . siRNAs ( 10 µl of a 100 nM stock in OptiMEM; Invitrogen ) were added to the transfection reagent ( 0 . 0625 µl Oligofectamin ( Invitrogen ) in 20 µl OptiMEM ) in wells of 96-well plates and incubated at RT for 30 min . Then , 70 µl of cells were added to each well ( 1 , 750 cells for HeLa Rps2-YFP and HeLa Rpl29-GFP cell lines , 1 , 250 cells of HeLa cells for Enp1 IF ) . 58 h after transfection , Rps2-YFP cells were induced with tetracycline ( final concentration of 125 ng/ml ) for 14 h . Rpl29-GFP cells were treated similarly , except induction with tetracycline was followed by incubation in tetracycline-free medium for 6 h . HeLa cells ( used for Enp1 IF analysis ) were incubated for 72 h after siRNA transfection . Cells were fixed with 4% PFA and DNA stained with Hoechst . Enp1 IF was performed as described [45] . Cells were automatically imaged using a 20× objective on a BD pathway 855 microscope . The Allstars Negative Control ( Qiagen ) was included three times on each plate . All siRNAs used for screening were designed by and purchased from Qiagen . All images and numerical data are accessible at http://hcpb . ethz . ch . Generation of HEK293 cell lines expressing TAP-tagged proteins and subsequent purification has been described [77] . HeLa extracts depleted of ribosomes by ultracentrifugation [78] and of tRNA [79] were added to purified particles without or with the addition of 10 µM RanQ69L-GTP . Binding reactions were performed in 50 mM Tris/HCl pH 7 . 5 , 150 mM KOAc , 2 mM MgCl2 , and 0 . 001% Triton X-100 . Antibodies against the following human proteins have been described: Enp1 , Nmd3 , Nob1 , Rio2 and Rps3 [45] , Crm1 , hExp5 [56] , and Exp-t [80] . The α-Rlp24 antibody was generated in rabbits using 6His-tagged Rlp24 as antigen . Anti-β-Actin ( Sigma-Aldrich ) and α-HA IgG beads ( Sigma ) are commercially available . Anti-Xenopus Exp5 antibodies were a kind gift from D . Görlich [59] . A custom assembled siRNA library was purchased from Qiagen . For individual analysis , RNAi was performed as described above , but in 6 wells . SiRNAs ( sense ) : Crm1 ( 5′-UGUGGUGAAUUGCUUAUAC ) , Exp5-1 ( 5′-AGAUGCUCUGUCUCGAAUU ) , Exp5-2 ( 5′-UGUGAGGAGGCAUGCUUGU ) [59] , and Nmd3 ( 5′-GAAUGGUGCUAUCCUUCAA ) . 32P-labeled RNAs encoding U3 snoRNA and pre-miR-31 were transcribed in vitro from PCR templates using α-[32P]-GTP as previously described [81] . The microinjection and dissection of Xenopus oocytes and 32P-labeling and analysis of rRNAs were performed according to [35] . | Ribosomes are molecular machines that synthesize proteins and are found in every cell . Each ribosome has a small and a large subunit , each of which is in turn composed of ribosomal RNA and many ribosomal proteins . In eukaryotic cells , the generation of these ribosomal subunits is a complex process requiring the participation of hundreds of factors . Originating from the nucleolus ( a nuclear region specialized in ribosome synthesis ) immature ribosomal subunits pass through the nuclear interior and are exported to the cytoplasm , where their assembly is finalized . In recent years , it has become apparent that defects in ribosome production are associated with human diseases , but our knowledge about this fundamental process is largely based on knowledge derived from yeast , a unicellular eukaryote . We set out to systematically identify factors involved in making ribosomes in human cells by individually depleting around 500 different cellular proteins . Using microscopic analysis , we identified approximately 150 human ribosome synthesis factors , thereby significantly extending our knowledge about human ribosome biogenesis . Our dataset not only revealed many evolutionarily conserved aspects of this essential cellular process but also led us to characterize an export route for the large ribosomal subunit that is specific for higher eukaryotic cells . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Material",
"and",
"Methods"
] | [
"biochemistry",
"cell",
"biology/nuclear",
"structure",
"and",
"function",
"cell",
"biology",
"biochemistry/macromolecular",
"assemblies",
"and",
"machines",
"computational",
"biology"
] | 2010 | A Protein Inventory of Human Ribosome Biogenesis Reveals an Essential Function of Exportin 5 in 60S Subunit Export |
Intercellular signaling plays an important role in controlling cellular behavior in apical meristems and developing organs in plants . One prominent example in Arabidopsis is the regulation of floral organ shape , ovule integument morphogenesis , the cell division plane , and root hair patterning by the leucine-rich repeat receptor-like kinase STRUBBELIG ( SUB ) . Interestingly , kinase activity of SUB is not essential for its in vivo function , indicating that SUB may be an atypical or inactive receptor-like kinase . Since little is known about signaling by atypical receptor-like kinases , we used forward genetics to identify genes that potentially function in SUB-dependent processes and found recessive mutations in three genes that result in a sub-like phenotype . Plants with a defect in DETORQEO ( DOQ ) , QUIRKY ( QKY ) , and ZERZAUST ( ZET ) show corresponding defects in outer integument development , floral organ shape , and stem twisting . The mutants also show sub-like cellular defects in the floral meristem and in root hair patterning . Thus , SUB , DOQ , QKY , and ZET define the STRUBBELIG-LIKE MUTANT ( SLM ) class of genes . Molecular cloning of QKY identified a putative transmembrane protein carrying four C2 domains , suggesting that QKY may function in membrane trafficking in a Ca2+-dependent fashion . Morphological analysis of single and all pair-wise double-mutant combinations indicated that SLM genes have overlapping , but also distinct , functions in plant organogenesis . This notion was supported by a systematic comparison of whole-genome transcript profiles during floral development , which molecularly defined common and distinct sets of affected processes in slm mutants . Further analysis indicated that many SLM-responsive genes have functions in cell wall biology , hormone signaling , and various stress responses . Taken together , our data suggest that DOQ , QKY , and ZET contribute to SUB-dependent organogenesis and shed light on the mechanisms , which are dependent on signaling through the atypical receptor-like kinase SUB .
How intercellular communication mechanisms coordinate the activities of cells during organogenesis is an important topic in biology . In higher plants shoot apical meristems and floral meristems are the ultimate source of above-ground lateral organs , such as leaves , flowers , and floral organs [1] . Meristems are organised into three distinct meristematic or histogenic layers , called L1 , L2 , and L3 [2] , and cells of all histogenic layers contribute to organogenesis [3] , [4] . The L1 layer gives rise to the epidermis while the L2 and L3 layers contribute to internal tissues . In Arabidopsis ovules , for example , the integuments that eventually develop into the seed coat are entirely made up of L1-derived cells , while L2 cells generate the inner tissue [5] . Classic studies have demonstrated that meristematic layers communicate [6] , [7] , but it is only recently that the biological relevance and the molecular mechanisms are being elucidated [8]–[11] . For example , work on the receptor-like kinase ( RLK ) BRASSINOSTEROID INSENSITIVE 1 ( BRI1 ) has provided evidence that the epidermis both promotes and restricts organ growth [12] . Furthermore , microsurgical experiments indicated that the epidermis also maintains cell division patterns in subtending layers [13] . These are but two examples that highlight the importance of the epidermis and inwards-oriented signaling in this inter-cell-layer cross-talk required for correct organ size and shape . At the same time , radial outward-oriented signaling also takes place during organogenesis . Known scenarios include transcription factors or small proteins that are synthesized in inner layers and move outwards into overlaying cell layers in a controlled fashion [14]–[17] . The so far best-characterised case of such a movement underlies radial patterning of the root [18] , [19] . In addition , the epidermally-expressed RLKs CRINKLY4 ( CR4 ) from corn or its Arabidopsis homolog ACR4 are necessary for epidermis development and may receive signals from underlying cell layers [20]–[24] . Inter-cellular communication during floral morphogenesis in Arabidopsis also depends on signaling mediated by the leucine-rich repeat transmembrane receptor-like kinase ( LRR-RLK ) STRUBBELIG ( SUB ) [25] . Analysis of sub mutants indicated that SUB is required for proper shaping of floral organs such as carpels , petals and ovules . At the cellular level SUB participates in the control of cell shape and/or the orientation of the cell division plane in floral meristems and ovules . In addition , SUB , also known as SCRAMBLED ( SCM ) , affects specification of hair cells in the root epidermis [26] , [27] . Recent evidence suggests that the SUB protein is confined to interior tissues in floral meristems , developing ovules and young roots although SUB mRNA is monitored throughout those organs [28] . In particular functional SUB:EGFP fusion protein is absent from cells that show a mutant phenotype in sub mutants , but can either be found in adjacent cells , as in floral meristems and ovules , or in cells that are separated from mutant cells by two cell diameters , like in the root . The non-cell-autonomous effects of SUB were corroborated by an analysis of sub-1 plants expressing a functional SUB:EGFP transgene under the control of different tissue-specific promoters . Thus the data indicate that SUB undergoes posttranscriptional regulation , acts in a non-cell-autonomous fashion and mediates cell morphogenesis and cell fate across clonally distinct cell layers in an inside-out fashion [28] . The SUB protein is a member of the LRRV/STRUBBELIG-RECEPTOR FAMILY ( SRF ) family of receptor-like kinases [29] , [30] . It is predicted to carry an extracellular domain with six leucine-rich repeats , a transmembrane domain , and a cytoplasmic intracellular domain with the juxtramembrane and kinase domains . Interestingly , phosphotransfer activity of the SUB kinase domain is not essential for its function in vivo [25] and thus SUB seems to belong to the family of atypical or “dead” receptor kinases [31] , [32] . Very little is known regarding signaling through atypical receptor-like kinases in plants [31] . In addition , it remains to be understood how cellular morphogenesis is coordinated across cell layers [8]–[11] . It is therefore of great interest to investigate the molecular basis of SUB signaling and function . Here we present the identification and analysis of three genetic factors that may relate to SUB signaling . Our results show that mutations in QUIRKY ( QKY ) , ZERZAUST ( ZET ) , and DETORQUEO ( DOQ ) result in a sub-like phenotype . Molecular cloning of QKY revealed that the predicted QKY protein is likely a transmembrane protein with four C2 domains indicating a role for QKY in Ca2+-dependent signaling . Global gene expression profiling of the mutants corroborates the morphological analysis but also suggests additional and distinct roles for each gene . Furthermore , the data indicate that SUB signaling plays previously unknown roles in cell wall and stress biology .
We applied a forward genetic approach to isolate additional factors of the SUB signaling pathway , based on the hypothesis that mutations in some of the genes that are part of the SUB pathway should result in sub-like ( slm ) mutant phenotypes . We thus screened M2 families of an ethylmethane sulfonate-mutagenized Ler population for slm mutants ( see Materials and Methods ) . In this experiment we identified two new sub alleles [25] as well as several novel mutants with sub-like phenotypes ( Figures 1–3 ) . These fell into three different complementation groups , which map to distinct positions on chromosome 1 ( Table 1 ) . We termed two of the genes DETORQUEO ( DOQ ) and ZERZAUST ( ZET ) , respectively . DETORQUEO refers to a latin term that means “to twist out of shape” . ZERZAUST is a German term for “disheveled” . We also isolated three mutant alleles of QUIRKY ( QKY ) , which plays a role in fruit dehiscence ( L . F . , unpublished results; S . J . and Martin F . Yanofsky , unpublished observations ) and the numbering of qky alleles was coordinated . Thus , our genetic approach resulted in the identification of three loci , DOQ , QKY , and ZET , mutations in which result in a sub-like phenotype and that , together with SUB , define the STRUBBELIG-LIKE MUTANT ( SLM ) class of genes . We identified one mutant allele of DOQ and two and three independent alleles of ZET and QKY , respectively ( Table 1 ) . All sub , doq , zet and qky alleles were recessive and behaved in a Mendelian fashion ( not shown ) . The various zet and qky mutants did not noticeably differ in their respective phenotypes and the three qky alleles are likely to be nulls ( see below ) . Thus , zet-2 and qky-8 or qky-9 were used as reference alleles for further analysis . In addition , we used the well-characterised sub-1 mutant for comparison [25] . This mutation likely represents a null-allele since it results in a stop codon and a predicted shorter SUB protein lacking the transmembrane and intracellular kinase domains . Thus , it is expected that SUB-dependent signaling across the plasma membrane is blocked in sub-1 mutants . At the macroscopic level sub-1 mutants are known to be affected in several above-ground organs [25] ( Figures 1–4 ) ( Tables 2 , 3 ) . Inflorescences are characterised by reduced height , an irregularly twisted stem , and an aberrant phyllotaxis of flowers . Flowers open prematurely and show a large percentage of twisted and often notched petals . Furthermore , all flowers exhibit twisted carpels and about 70 percent of sub-1 ovules showed aberrant initiation of the outer integument ( Figure 2 ) ( Table 2 ) . This results in outer integuments with gaps that often resemble “multifingered clamps” or “scoops” . Also ovules with a fully developed outer integument show defects . In particular the distal or micropylar cells of the outer integument can show aberrant size and shape . In addition , about 40 percent of sub-1 plants show at least one leaf per rosette with twisted petioles . At the cellular level sub-1 exhibits aberrant cell shape and robustly scorable numbers of periclinal , rather than anticlinal cell division planes in cells of the L2 layer of stage 3 floral meristems [25] ( Figure 3 ) ( Table 3 ) . Interestingly , in this analysis we could also observe a previously unnoticed low number of periclinal divisions in the L1 of sub-1 floral meristems . Furthermore , sub/scm mutants develop root hairs at discordant positions in the epidermis [26] , [27] . This defect in root hair patterning can be followed by expressing the bacterial β-glucuronidase gene under the control of the Arabidopsis GLABRA2 ( GL2 ) promoter ( GL2::GUS ) [27] , [33] ( Figure 3 ) . This reporter conveniently labels the regular files of non-hair cells in the epidermis of wild-type roots and exhibits an irregular expression pattern in sub/scm mutants . Upon examination , many aspects of the phenotypes of sub-1 and the other slm mutants were comparable . The doq-1 , qky-8 and zet-2 mutants showed aberrant floral phyllotaxis and flowers with twisted petals ( Figure 1 ) . In addition , we observed premature floral bud opening in doq-1 and zet-2 . Petals of zet-2 mutants showed notches similar to sub-1 petals but zet-2 floral organs were generally more misshapen . In rosettes , nearly all plants showed examples of leaf petiole twisting . In particular , qky-8 mutants showed leaf twisting comparable to sub-1 , whereas doq-1 rosette leaves showed in addition elongated petioles and narrow blades . In contrast , zet-2 rosette leaves did not show major defects at the gross morphology level . Irregular twisting of siliques and stems was apparent in doq-1 , qky-8 and zet-2 mutants , although the twisting in doq-1 siliques and stems was more subtle . Plant height was most affected in qky-8 while doq-1 and zet-2 showed only a slight reduction . Mature ovules of doq-1 mutants were mildly but consistently affected ( Figure 2 ) ( Table 2 ) . The outer integument did not fully extend to the funiculus ( reduced campylotropy ) . In addition , its distal cells were shorter and of irregular shape . Mature ovules of qky-8 and zet-2 closely resembled ovules of sub-1 mutants with respect to gaps and altered cell shape in the outer integument . Interestingly , zet-2 showed a slightly higher percentage of malformed outer integuments . As was the case for sub [25] the inner integument in all other slm single mutants appeared to be unaffected . Further , doq-1 , qky-9 and zet-2 mutants were investigated for cellular defects in the root epidermis and in L1/L2 cells of stage 2 to 4 floral meristems ( Figure 3 ) ( Table 3 ) . In 4 day old main roots , all three mutants showed misregulation of GL2::GUS reporter expression comparable to sub-1 , indicating that root hair patterning was similarly affected . The three mutants also exhibited cell shape and periclinal cell division plane defects in the L1 and L2 cells of floral meristems . Floral meristems of doq-1 showed a higher percentage of those defects while in zet-2 , cell separation and disintegration could also be observed . The latter finding indicates that ZET is also required for cell viability . Taken together the analysis of the slm single mutants revealed that there is a large functional overlap of the corresponding genes and that individual SLM genes participate in subsets of SUB-dependent processes , such as ovule development , cellular behavior of L2 cells of stage 3 floral meristems , and root hair patterning . The results also suggest , however , that SLM genes have additional functions unrelated to each other . For example , ZET has a particular function in cell survival in floral meristems as has DOQ in the regulation of leaf shape . To investigate further the genetic relationship between SLM genes we generated all possible double-mutant combinations and analysed the respective mutant phenotypes . The results are summarized in Figures 4 and 5 and Table 2 . Petals of sub-1 doq-1 double mutants mostly resembled doq-1 petals ( Figure 4A ) , however , stem twisting was more similar to sub-1 . Other aspects of the sub-1 doq-1 phenotype , such as silique twisting , plant dwarfism and rosette leaf petiole twisting were more exaggerated when compared to either single mutant . Ovules of sub-1 doq-1 plants showed an increase in outer integument defects , although not as strong as in some other double mutant combinations . In addition , about 11% of ovules of sub-1 doq-1 plants showed defects in inner integument morphology ( Figure 5C ) ( Table 2 ) , with gaps of variable sizes and finger-like protrusions . The fertility of sub-1 doq-1 plants was severely reduced ( Figure 5R ) . Petal and carpel twisting in sub-l qky-8 double mutants was similar to that shown by each single mutant ( Figure 4G ) . In contrast , twisting of siliques , stems , and leaf petioles was more pronounced as was the reduction in plant height ( Figure 4I–L ) . In addition , ovule development was more heavily affected compared to the single mutants ( Figure 5D–F ) ( Table 2 ) , with 13% of ovules showing inner integument defects and a corresponding reduction in fertility ( Figure 5R ) . About 5% of ovules were hardly recognizable as such , but rather resembled a mass of cells with integument-like outgrowths ( Figure 5E ) . Overall , the sub-1 zet-2 double mutants showed the most disturbed morphology ( Figure 4M–R ) . Perianth organs were twisted , narrower and notched , while carpels were heavily twisted . As a result the overall structure of the flower was irregular . The stem phenotype was knotted rather than twisted . Leaf morphology was misshapen with more pronounced leaf petiole twisting and plants showed prominent dwarfism . Consistent with this exaggerated phenotype ovules of sub-1 zet-2 plants showed severe defects with 21% of ovules resembling a mass of cells with integument-like outgrowths ( Figure 5H ) ( Table 2 ) . About 18% of sub-1 zet-2 ovules exhibited short gapped outer integuments and similarly malformed inner integuments ( Figure 5I ) . Again , fertility was reduced in sub-1 zet-2 plants ( Figure 5R ) . Perianth organs and carpels in doq-1 qky-8 mutants were more twisted as compared to the parental lines ( Figure 4S ) . Twisting of leaf petioles , stems , and siliques was also more pronounced as was dwarfism . Ovules of doq-1 qky-8 mutants showed a less exaggerated phenotype compared to some of the other double mutant combinations . Still , a large proportion of doq-1 qky-8 ovules showed gaps in the outer integument ( Figure 5J–L ) ( Table 2 ) and aberrant inner integument morphology was seen in 9% of ovules ( Figure 5L ) . Fertility was strongly reduced in doq-1 qky-8 double mutants ( Figure 5R ) . Perianth morphology in doq-1 zet-2 mutants was about equivalent to that observed in sub-1 zet-2 and doq-1 qky-8 flowers ( Figure 4Y ) . Twisting of stems appeared slightly more pronounced compared to each single mutant while siliques were drastically more twisted . Plant height was comparable to zet-2 single mutants . Overall a higher percentage of doq-1 zet-2 ovules showed gaps in the outer and inner integuments ( Figure 5M–O ) , with 4% of ovules exhibiting severe malformations comparable to sub-1 zet-2 mutants ( Table 2 ) . The reduction in fertility was comparable to that of doq-1 qky-8 double mutants ( Figure 5R ) . The qky-8 zet-2 double mutants largely phenocopied zet-2 single mutants in above-ground morphology ( Figure 4Ae–Aj ) including ovules ( Figure 5P , Q ) ( Table 2 ) . One exception to this was leaf petiole twisting as this aspect was most similar to the phenotype in qky-8 single mutants . In summary , the pleiotropic phenotypes of slm single and double mutants complicated the double mutant analysis . Although an exaggerated phenotype was often observed in a double mutant combination it was usually difficult to decide whether a double mutant displayed an additive or synergistic phenotype , or whether a particular mutation was epistatic to another . Overall , the double mutant analysis reinforced the notion that SLM genes likely do not act in a single linear pathway but have both overlapping and separate functions . To further explore the genetic relationship between SLM genes we first tested whether SUB expression was responsive to other genes of this group . We investigated SUB expression in inflorescence apices and stage 10–12 flowers ( see below ) from several slm mutants by quantitative real time PCR ( qRT-PCR ) . As can be seen in Figure 6A only very moderate changes in SUB expression were detected with a slight but consistent reduction of SUB expression in flowers of doq-1 , qky-8 and zet-2 mutants . In contrast , SUB transcript levels were unaltered in those tissues and mutants when assessed by a transcriptome analysis ( see below ) . Given that SUB is expressed at very low levels to begin with [34] we reasoned that the observed mild effects may not be the result of direct effects but rather be a consequence of indirect influences due to tissue sampling or the altered morphology of the mutants . To test the relevance of the effects we asked whether rendering SUB independent of its normal transcriptional regulation , by ectopically expressing SUB using the cauliflower mosaic virus 35S promoter , could result in phenotypic rescue of doq , qky and zet mutants . We analysed the phenotypes of 35S::SUB doq-1 , 35S::SUB qky-8 and 35S::SUB zet-2 plants ( Figure 6B ) . In particular we scored stem twisting , flower morphology and silique twisting of individual transgenic T1 lines . Transgene functionality was demonstrated by the wild-type appearance of 35S::SUB sub-1 plants . Despite demonstrating comparable transgene expression ( Figure 6C ) , the other tested combinations did not show phenotypic rescue ( 35S::SUB doq-1: 42 T1 lines scored , 35S::SUB qky-8: >100 , and 35S::SUB zet-2: >100 ) , indicating that SUB does not act as a downstream target gene of DOQ , QKY or ZET . Taken together our data suggest that SUB is not directly regulated at the transcriptional level by the other SLM genes . To shed light on the molecular nature of SUB-dependent processes and to systematically investigate the similarities between the SLM genes at the mechanistic level , we applied a transcriptome analysis . Since mutants with defects in genes of overlapping functions should also exhibit overlapping alterations in transcript profiles we compared wild type ( Ler ) with plants of two sub alleles , as well as doq-1 , zet-2 and qky-8 mutants using the Affymetrix ATH1 GeneChip platform ( see Materials and Methods ) . The analysis of two independent sub alleles was aimed at obtaining a robust set of SUB-responsive genes for the different pairwise comparisons and the list of SUB-responsive genes represents the overlap of misexpressed genes in sub-1 and sub-3 mutants [25] . Since we had observed tissue-specific differences in the phenotypes of slm mutants , which suggested tissue-specific sub-functions for the corresponding genes , we tried to capture these differences also in the sampling for transcriptome analysis . To this end , we sampled inflorescence apices plus flowers up to stage 9 ( apex data set ) , representing mostly proliferative tissues , while stage 10–12 flowers were collected to represent maturation stages ( flower data set ) . The samples thus covered SUB-related aspects such as the control of cell shape/division plane in L2 cells of floral stage 3 meristems , and the regulation of petal , carpel and ovule development . Since traditional analysis tools to identify differential gene expression in whole genome transcriptome data are not well suited for comparisons of multiple samples we developed a meta analysis tool based on Z-score statistics ( see Materials and Methods ) . It offers a variety of benefits for the simultaneous analysis of multiple pairwise comparisons , such as normalization for the overall biological effect observed in the individual experiments , increased sensitivity , and the possibility of querying data using Bolean logic . Using this tool we identified 89 and 193 significantly misexpressed genes , respectively , in the apex and flower samples of sub plants ( Table 4 ) . The other mutants were characterised by higher numbers of misexpressed genes ( Table 4 ) indicating that DOQ , QKY , and ZET affect more processes than SUB . More importantly , we systematically analyzed the overlap of misexpressed genes from individual slm mutants compared to wild type . The results of all pairwise comparisons are given in Table 5 . For example , a pairwise comparison between sub and doq-1 in the apex sample revealed a 42 gene overlap . This corresponded to 47% of genes misexpressed in sub and about 17% of genes aberrantly expressed in doq-1 . Comparable values were observed for sub versus qky-8 and sub vs zet-2 comparisons . The overlap with SUB-dependent genes was even more pronounced in the flower data set and in the sub-doq-1 comparison we found 118 common genes , representing 61% and 17% of the genes misexpressed in sub and doq-1 , respectively . The sub qky-8 and sub zet-2 comparisons revealed higher values with 76% and 81% of SUB-responsive genes being found to overlap . Thus , a significant number of SUB-responsive genes are also sensitive to DOQ , QKY and ZET function . Similar pairwise comparisons were made between doq-1 , qky-8 and zet-2 ( Table 6 ) . Again there was considerable overlap of misexpressed genes between different mutants . Most strikingly , in stage 10–12 flowers 67% and 71% of the QKY-responsive genes are also misregulated in doq-1 and zet-2 , respectively . This finding implies that the activity of more than two-thirds of QKY-responsive genes also depends on DOQ and ZET function in stage 10–12 flowers . Taken together the microarray data corroborate the morphological analysis of the single mutants and are compatible with the notion that the different SLM genes share overlapping functions . To further study the complexity of the functional relationship among SLM genes we carried out additional pairwise comparisons , examined the identity of genes present in the overlap lists , and identified transcripts which are misregulated in all four mutants ( i . e . , a sub , doq-1 , qky-8 and zet-2 “quadruple” overlay ) ( Figures 7 , 8; Tables 5 , 6; Datasets S1 , S2 ) . While doing so we uncovered striking trends in the transcriptome data . For example , we found that more than 30% of the misregulated genes found in any sub/slm overlap identified in the apex samples are indeed shared by all mutants ( Figure 7A ) . More significantly , 13 of the 14 transcripts identified as common SLM responsive in the apex were consistently elevated in expression in all mutants , suggesting that a core function of the SLM genes is to repress this set of genes in proliferating tissue ( Figure 7B ) . Strikingly , this trend was reversed in differentiating tissue represented by the floral samples . Here the misregulated genes shared by all mutants made up more than 65% of transcripts identified in pairwise comparisons ( Figure 7A ) and 88 of the 93 genes were reduced in expression ( Figure 7B ) . In addition to supporting the notion of functional overlap between the SLM genes , this analysis also demonstrated that diversification of SLM function observed in different tissues is based on a dramatic switch in the underlying molecular mechanism . While signaling through SLM proteins in proliferating tissue mainly causes repression of target gene transcription , SLM mediated signals lead to a transcriptional activation of downstream genes in differentiating cells , suggesting that SUB and other SLM proteins interact with a radically different regulatory environment . This notion is supported by another striking observation: while we found strong similarities between different genotypes within one tissue we observed very little overlap in misregulated genes between the same genotype in the two tissues analysed ( Figure 8 ) . For example , only four misregulated transcripts were shared in apex and flower samples of sub mutants . Similar observations were made for the other single and pairwise comparisons . In addition , the core target gene sets of apex and flower , derived from quadruple comparisons , did not share any transcripts . To investigate the nature of SLM-dependent processes we assessed the known or predicted functions of the apex and flower core sets of SLM-responsive genes by searching The Arabidopsis Information Resource ( TAIR ) , the literature and making use of the AtGenExpress expression atlas [34] with the help of the AtGenExpress Visualisation Tool ( AVT , http://jsp . weigelworld . org/expviz/ ) ( Tables 7 , 8 ) . In both core sets we observed a strong representation of genes that are involved in processes , such as cell wall biosynthesis and function , hormone signaling and abiotic and biotic stress responses . For example , two genes in the apex core set encode the inositol oxygenase family enzymes MIOX2 and MIOX4 required for biosynthesis of uridine-diphospho-glucuronic acid ( UDP-GlcA ) , a precursor for various cell-wall matrix polysaccharides [35] . In addition , GERMIN-LIKE PROTEIN 6 ( GLP6 ) is a member of a gene family encoding putative extracellular GERMIN-LIKE proteins [36] some of which play a role in biotic stress responses and cell wall biology . The flower core set of SLM-responsive genes also included several genes coding for cell wall proteins , such as ECS1 [37] , At2g05540 ( Glycine-rich protein ) , At5g03350 and At5g18470 ( lectin family proteins ) , and At2g43570 and At4g01700 ( chitinases ) . In another example that relates to hormone signaling , five of the 13 upregulated apex core set genes were inducible by jasmonates ( JA ) while the flower core set was characterized by a large group of genes that relate to different aspects of signaling mediated by salicylic acid ( SA ) ( see Discussion ) . SLM-responsive genes whose expression is sensitive to JA encoded for example JAZ1 , a central negative regulator of JA signaling [38] , [39] , ERD5 a mitochondrial proline dehydrogenase ( ProDH ) [40] , AtLEA5/SAG21 [41] , [42] , and AtTTPG , a class III trehalose-6-phosphate phosphatase involved in the biosynthesis of trehalose [43] , [44] . All of these genes are believed to contribute to the plant's response to dehydration , cold and oxidative stress [40] , [41] , [44]–[50] . We also observed SLM-dependency of genes involved in the homeostasis of the auxin indole-3-acetic acid ( IAA ) and the production of the Arabidopsis phytoalexin camalexin and indole glucosinolates . Glucosinolate metabolites play important roles in plant nutrition , growth regulation and the interactions of plants with pathogens and insect herbivores [51] , [52] . In the apex core set we identified IAA-LEUCINE RESISTANT-LIKE GENE 6 ( ILL6 ) , which belongs to a gene family encoding IAA amino acid conjugate hydrolases involved in the release of the active auxin from corresponding inactive amino acid conjugates [53] , [54] . CYP83B1 and CYP79B2 encode cytochrome P450 monooxygenases that regulate the pool of indole-3-acetaldoxime ( IAOx ) [55]–[58] a key branching point of primary and secondary metabolic pathways and a precursor of indole glucosinolates , camalexin [59] , and IAA ( reviewed in [51] , [52] ) . In addition , the flower core set contained two downregulated cytochrome P450 genes , CYP71A13 and CYP71B15 , which promote distinct steps in camalexin biosynthesis [60]–[63] . As a first step to molecularly define the SLM pathway , the QKY gene was identified by map-based cloning ( Figure 9 and Materials and Methods ) . In all three EMS-induced qky alleles we could identify single point mutations in At1g74720 ( Figure 9A ) . Further , two distinct T-DNA insertions in this locus result in qky mutants phenocopies ( not shown ) . Combined this demonstrates that At1g74720 is QKY . Sequence analysis predicted that QKY contains no introns and encodes a transmembrane protein of 1081 amino acids with a calculated molecular mass of 121 . 4 kDa harboring four C2 domains ( Figure 9B ) . In addition , two transmembrane domains are embedded in a plant-specific phosphoribosyltransferase C-terminal region ( PRT_C , PFAM identifier PF08372 ) , which according to the PFAM database occurs characteristically at the carboxy-terminus of phosphoribosyltransferases and phosphoribosyltransferase-like proteins [64] . Interestingly , this domain often appears together with C2-domains , as in QKY . A related domain topology , albeit with only three C2 domains , is found in several proteins present in humans , Drosophila melanogaster and Caenorhabditis elegans ( C . elegans ) , collectively referred to as multiple C2 domain and transmembrane region proteins ( MCTPs ) [65] . The role of animal MCTPs is not well defined despite the fact that in C . elegans a single MCTP gene is present and essential for embryo viability [66] . QKY represents the first described plant MCTP-class gene . In accordance with proposed transmembrane topography for MCTPs the QKY C2 domains most likely have an intracellular localisation as QKY lacks a distinguishable N-terminal signal peptide and all known C2 domains are cytoplasmic . C2 domains are autonomously folding modules and form Ca2+-dependent phospholipid complexes , although some exceptions are known that do not bind Ca2+ or phospholipids or both [67]–[69] . Other animal proteins with multiple C2 domains , but only one transmembrane region , include the synaptotagmins [70] , the extended synaptotagmins [71] , and the ferlins [72] . Synaptotagmins and ferlins are known to play a role in membrane trafficking . Although there is general resemblance in domain topology , very little primary sequence conservation is observed between QKY and the animal MCTPs , synaptotagmins , and ferlins . The qky-7 , qky-8 , and qky-9 alleles likely cause a complete loss of QKY function . All three mutations are predicted to introduce stop codons leading to shorter proteins with variable numbers of C2 domains but always lacking the two transmembrane domains ( Figure 9 ) . Thus , the three mutants likely contain truncated QKY variants that are not properly located to the membrane . Membrane localisation , however , seems essential for QKY function as all three mutants show identical phenotypes , regardless of the number of C2 domains still present in the predicted truncated proteins . In summary , the results suggest that QKY is a membrane-bound Ca2+-signaling factor .
In this work we set out to genetically identify additional components of the SUB-dependent mechanism regulating ovule development and floral morphogenesis . To this end we identified second-site mutations that resulted in a sub-like phenotype , reasoning that such mutations should reside in genes that either act directly in the SUB signaling pathway , or , that SUB and these genes affect parallel pathways that converge on a common molecular or developmental target . A defined set of criteria was applied in the identification of candidate mutants . At the developmental level , candidates were chosen that showed a sub-like ovule phenotype and twisting of petals , carpels , siliques and the stem . In addition , they were analysed for two types of sub-like cellular defects: an increased frequency of periclinal cell divisions in the L1/L2 layers of stage 3 floral meristems and a defect in root hair patterning . Mutations in three genes , DOQ , QKY , and ZET , satisfied all criteria and were considered as sub-like mutants . Although slm mutants were phenotypically very similar there were some noteworthy distinctions . For example , doq-1 was generally characterised by a weak ovule phenotype and mild silique/stem twisting . At the same time , doq-1 showed the highest frequency of periclinal cell divisions in the L1/L2 layers of stage 3 floral meristems . Furthermore , rosette leaves of doq-1 plants had elongated petioles and narrow leaf blades , features not observed in other slm mutants . These findings indicate that DOQ is less important for ovule , silique and stem development compared to other SLM genes , but plays a more prominent role in the regulation of cell shape/division plane in the young floral meristem and leaf development . In contrast ZET does not seem to be involved in leaf development as zet mutants exhibited apparently normal leaves . This clearly distinguishes ZET from the other SLM genes , which all seem to be important for leaf development . Plants with a defect in ZET activity also showed more defects in perianth development when compared to other slm mutants . Cell separation and disintegration observed in stage 3 floral meristems of zet mutants indicated that ZET may function in cell adhesion and viability , roles that seem to be unique among SLM genes . Since the phenotypes of sub and the other slm mutants overlap we assessed the genetic relationship among SLM genes by testing whether the expression of SUB is sensitive to SLM activity . This was apparently not the case , at least not in a relevant manner , since SUB expression was unaltered in the respective other slm mutants in our transcriptome analysis and showed only minor downregulation in slm flowers when assayed by qRT-PCR . In addition , a functional 35S::SUB transgene was unable to rescue the phenotype of doq , qky and zet mutants , including the floral defects . To further investigate the genetic interactions between SLM genes we performed a comprehensive double mutant analysis . The pleiotropic and variable phenotypes of slm single and double mutants complicated this task . Although an exaggerated phenotype was often observed in double mutants it was sometimes difficult to decide whether this represented an additive or synergistic phenotype . In addition , the various effects were often tissue dependent and thus precluded conclusions at the whole-plant level . One exception was qky-8 zet-2 where zet-2 appears to be largely epistatic to qky-8 . This result indicates that ZET either acts upstream of QKY in a genetic pathway , or that ZET generally acts prior to QKY . Focusing on ovule development simplified our analysis , since a stringent set of morphological criteria could be applied . In sub-1 zet-2 , sub-1 qky-8 , and doq-1 zet-2 double mutants a mass of cells with integument-like structures often developed in place of true ovules . Furthermore , all double mutants , with the exception of qky-8 zet-2 , often showed aberrant inner integuments , in contrast to slm single mutants that only displayed defects in the outer integument . Taken together , these synergistic effects indicate that the SLM genes contribute to similar aspects of ovule development and show that QKY and ZET function is still present in ovules of sub plants . In addition , the analysis supports the notion that DOQ may play a more peripheral role in ovule development in comparison to SUB , QKY , and ZET . The exact genetic relationship between SLM genes during this process remains to be elucidated . The phenotypic analysis of slm single mutants suggested that SLM genes variably contribute to several common functions but it also showed that they have distinct roles , both of which depend on tissue context . To characterise the molecular mechanisms that are under control of SLM genes and to narrow down the overlap of SLM gene function , we used transcript profiling of mRNA isolated from two different tissues . Our Z-score based meta-analysis revealed 89 and 193 largely non-overlapping SUB-responsive genes in the apex and flower data sets , respectively ( Table 4 ) . That we consistently observed larger numbers of affected transcripts in the flower data set might reflect the relatively higher importance of SLM genes for later stages of flower development , but it could also be explained by differences in the nature of the samples . While the flower sample only contained tissue from two well-defined floral stages , the apex sample contained the shoot apical meristem as well as flowers from stage 1 to stage 9 and thus was much more diverse . Consequently , the sensitivity of the array to detect changes that are limited to only a few developmental stages was reduced . When comparing the number of misregulated transcripts across the various genotypes , we found that fewer genes were affected in sub mutants than in the other slm mutants . This observation suggests that DOQ , QKY and ZET play broader roles than SUB . The result might have been expected for ZET , as zet mutants show the most severely affected flowers of all slm mutants , while sub , qky and doq mutant flowers , however , are more alike and thus their morphology is less congruent with a more dramatic change at the molecular level . Meta-analysis revealed a consistent overlap in misregulated genes between sub and the other slm mutants within a given tissue and even allowed us to identify a set of core targets shared by all four SLM genes . Interestingly , these transcripts were consistently up- or downregulated in all genotypes , while the direction of the change was dependent on tissue context . These results demonstrate not only that SLM genes affect common processes with a high tissue specificity , but also that the molecular mechanisms employed by SLM genes in the different tissues are distinct . In addition , the identification of common targets that show consistent behaviour across all genotypes underlines the sensitivity and specificity of our meta-analysis . Sequence analysis of QKY suggests that QKY is the first described plant representative of the previously described MCTP family [65] and that it might act as a membrane-bound protein involved in a process regulated by Ca2+ and phospholipids . While the function of animal MCTPs is unknown , human MCTP2 is a membrane protein located to intracellular vesicular structures [65] . Interestingly , the C2 domains of human MCTPs were found to bind Ca2+ with high affinity but lacked any phospholipid binding capacity [65] . Other membrane-bound proteins with multiple C2 domains are the synaptotagmins and ferlins [70] , [72] . Synaptotagmins contain two C2 domains and an amino-terminal transmembrane domain while most ferlins carry between four and six C2 domains and a carboxy-terminal transmembrane domain . Members of these two protein families function during regulated exocytosis , a process in which specific vesicles are signaled to fuse with the plasma membrane . Processes that rely on regulated exocytosis include neurotransmitter release at synapses and plasma membrane repair , a basic cellular process that mends physical injuries inflicted upon the plasma membrane [73]–[75] . During the repair process lysosomes [76] or specialised vesicles , such as the enlargosomes [77] , [78] , fuse with the plasma membrane providing new membrane material and facilitating resealing . The synaptotagmins Syts 1 and 2 are required for Ca2+-regulated synaptic vesicle exocytosis during neurotransmitter release into the synaptic cleft [79] , [80] , [70] while Syt VII promotes lysosomal exocytosis during plasma membrane repair in fibroblasts [76] . The C . elegans ferlin FER-1 is localized to the membranes of membranous organelles ( MOs ) and promotes the fusion of MOs with the plasma membrane during the development of crawling spermatozoa [81] , [82] . Mutations in dysferlin result in progressive muscular dystrophies [83]–[85] and dysferlin appears to be required for Ca2+-dependent sarcolemma resealing during membrane repair in skeletal muscle fibres [84] , [86] . Given the predicted domain topology of QKY one could speculate that QKY , and possibly other SLM genes , could participate in the control of vesicle trafficking . In principle , the predicted membrane localisation of QKY also allows for the possibility that the SUB and QKY proteins interact directly . A role of QKY in vesicle trafficking would help to explain the non-cell-autonomy of SUB function in inter-cell-layer signaling [28] . Thus , in one possible scenario SUB could directly or indirectly regulate QKY which in turn might affect vesicular transport of factors mediating the non-cell-autonomy of SUB-dependent signaling . Further work needs to address this exciting hypothesis . To obtain additional insights concerning the molecular processes influenced by SLM function we investigated the known and predicted functions of SLM-responsive genes . The core sets of direct or indirect SLM targets included many genes encoding proteins relating to cell wall biology and a notable enrichment of genes that are inducible by hormones , such as JA and SA , and presumed to play a role in the plant's response to various stresses . For example , a group of genes misregulated in slm mutants encode proteins with a known or predicted role in the biosynthesis and/or function of the cell wall , such as MIOX2/4 , GLP6 , and putative glycine-rich proteins , lectin family proteins , and chitinases . Defects in SLM-activity also resulted in a deregulation of the basal expression of JA-inducible genes with a presumed role in abiotic stress responses to cold or dehydration . The altered expression of genes responsible for camalexin and glucosinolate metabolism , secondary metabolites involved in the defense against pathogens , indicates that SLM-activity appears to play a role in biotic stress responses as well . The notion that SLM activity somehow relates to stress is reinforced by the interesting finding that a major group of SLM-responsive genes in flowers relate to various aspects of salicylic acid ( SA ) -dependent signaling . SA is synthesized upon exposure of plants to abiotic stresses , such as ozone or UV-C light , and plays an important role in pathogen defense ( for reviews see for example [87]–[91] ) . Three of the SLM-responsive genes are known to be required for the pathogen-induced production of SA: ICS1/SID2/EDS16 encodes an isochorismate synthase that synthesizes SA [92] , [93] , EDS5/SID1 encodes a MATE family transporter potentially involved in the transport of SA intermediates [94] , and PBS3/GDG1 encodes an adenylate-forming enzyme required for SA accumulation [95] , [96] . Other SLM-dependent genes are involved in further aspects of SA signaling . EDS1 encodes a protein with a lipase-like domain [97] , contributes to rapid SA accumulation in response to many SA-inducing stimuli , and is part of a central regulatory node of SA signaling . NPR1/NIM1 constitutes another , though SLM-independent , central regulator of SA signaling . At low SA levels NPR1 is present in the cytoplasm as a homo-multimeric complex . Increased SA levels result in the dissociation of NPR1 oligomers into monomers , which enter the nucleus where they interact with specific TGA transcription factors . These NPR1-TGA complexes regulate expression of defense genes , such as PATHOGENESIS-RELATED ( PR ) genes and WRKY-type transcription factor genes . Interestingly , we found NIMIN1 and GRX480 , both of which act in the NPR1 pathway , to be downregulated in slm mutants . NIMIN1 encodes an interactor of NPR1 that counteracts NPR1 activity [98] . GRX480 codes for a glutaredoxin that interacts with TGA2 [99] and is possibly involved in cross-talk between SA and JA signaling . Another important class of regulators implied in plant immunity , the regulation of PR genes and SA-mediated signaling are WRKY transcription factors . Interestingly , we found five WRKY genes to be downregulated in slm mutants and three of them are known direct targets of NPR1: WRKY38 , WRKY53 , and WRKY70 [100] . WRKY53 is a positive modulator of systemic acquired resistance ( SAR ) [100] while WRKY70 is involved in the cross-talk between SA and JA responses by acting as a negative regulator of JA-inducible genes [101] , [102] . In addition , WRKY70 is required for resistance against several bacteria , fungi and an oomycete [100] , [101] , [103] , [104] . Another SLM-responsive WRKY gene , WRKY33 , is a positive regulator of resistance against the necrotrophic fungi Botrytis cinerea and Alternaria brassicicola [105] . Fitting in the overall picture was the observation that expression of five PR genes was reduced in slm mutants , including PR1 , PR2 , and PR5 whose activities often undergo coordinated changes in response to various stimuli including SA [106] . In addition , PR1 is likely a direct target of a NPR1/TGA complex [107] , [108] . Many of the SLM-responsive genes involved in SA-signaling and pathogen defense are normally transcribed at basal levels while their expression is induced several fold in the case of pathogen attack and increased levels of SA . Our transcriptome analysis of slm mutants suggests that SLM function is required for basal steady-state expression of SA-responsive genes . Thus , SLM function could be involved in priming components of the cellular defense machinery , thereby enabling floral cells to respond faster to invading pathogens . In this scenario SLM genes would contribute to basal resistance . Another scenario is based on the hypothesis that QKY may play a role in membrane traffic . Since membrane repair or cell wall synthesis require extensive vesicle trafficking [74] , [109] SLM-dependent processes could be involved in the regulation of the composition of the plasma membrane or the cell wall and alterations in the two cellular compartments of slm mutants could be interpreted as wounding stress . For example , it was suggested that the SLM- and SA-responsive WALL-ASSOCIATED RECEPTOR KINASE1 ( WAK1 ) gene transmits changes in the cell wall to the interior of the cell . In addition , WAK1 protects plant cells against negative aspects of the pathogen response [110] . The RLK THESEUS1 ( THE1 ) has been proposed to act as a sensor of cell wall integrity [111] . It was shown that in the absence of proper cellulose synthesis THE1 modulates the activity of genes affecting pathogen defense , cell wall cross-linking and stress responses . Although THE1 and the SLM genes do not share common transcriptional targets , they affect a functionally related spectrum of genes . Thus , it is conceivable that one aspect of SLM function relates to the surveillance of cell wall integrity or the repair of plasma membranes . The two proposed models are not mutually exclusive and it will be an exciting challenge to further dissect the biological function of SLM genes in development , cell biology , and stress response .
Arabidopsis thaliana ( L . ) Heynh . var . Columbia ( Col-0 ) and var . Landsberg ( erecta mutant ) ( Ler ) were used as wild-type strains . The sub-1 mutant was described previously [25] . Plants were grown in a greenhouse under Philips SON-T Plus 400 Watt fluorescent bulbs on a long day cycle ( 16 hrs light ) . Dry seeds were sown on soil ( Patzer Einheitserde , extra-gesiebt , Typ T , Patzer GmbH & Co . KG , Sinntal-Jossa , Germany ) overlying perlite , stratified for 4 days at 4°C and then placed in the greenhouse . Plant trays were covered for 7–8 days to increase humidity and support equal germination . Ler seeds mutagenized with ethylmethane sulfonate ( EMS ) were obtained from Lehle Seeds ( Round Rock , TX , USA ) . 60'000 M2 plants , corresponding to about 7'500 M1 plants , were screened for plants exhibiting a sub-like phenotype . All sub-like mutants described in this paper were outcrossed three times to Ler prior to further analysis . Two qky T-DNA insertion mutants ( line SALK_140123 and SALK_043901 ) [112] were obtained from the ABRC ( http://www . arabidopsis . org ) . The GL2::GUS line [33] in Ler was crossed into slm mutants for analysis of root hair specification . For DNA and RNA work standard molecular biology techniques were used [113] . PCR-fragments used for cloning were obtained using either PfuUltra high-fidelity DNA polymerase ( Stratagene ) or TaKaRa PrimeSTAR HS DNA polymerase ( Lonza , Basel , Switzerland ) . All PCR-based constructs were sequenced . Information regarding all primers used in this study is given in supporting Table S1 . In order to generate a translational fusion between SUB and the small c-myc tag SUB cDNA was amplified by PCR from plasmid H2H6T7 [25] using primers SUB-CmycF and SUB-Cmyc-R . A 3xmyc-tag was amplified from a pFastBac-HT A plasmid ( Invitrogen ) containing a 2xmyc tag using primers cmyc-F and cmyc-R . Both PCR fragments were gel purified and an overlap PCR was set up using primers SUB-Cmyc-F and cmyc-R . The overlap PCR product was cloned into vector PCRII TOPO ( Invitrogen ) and was designated PCRII TOPO SUB:3xmyc . SUB:3xmyc was then cloned into pCAMBIA2300 ( www . cambia . org ) . To this end pCAMBIA2300 was digested with BamHI and PmlI to remove the 35S::GUS cassette . PCRII TOPO SUB:3xmyc was digested with BamHI and BsrBI and the resulting SUB:3xmyc fragment was cloned into BamHI/PmlI-digested pCAMBIA2300 resulting in plasmid SUB:3xmyc pCAMBIA2300 . To obtain the 35S promoter [114] the vector pART-7 [115] was digested with NotI , blunt-ended by T4 DNA polymerase treatment and the 35S:NOS cassette was isolated by gel purification and digested with XbaI , resulting in a 35S promoter fragment with a 5′ blunt and 3′ sticky end . SUB:3xmyc pCAMBIA2300 was blunt-ended after SpeI digestion at the 5′ end and redigested with BlnI to create an XbaI-compatible 3′ end . Ligation resulted in the final 35S::SUB:myc construct ( abbreviated 35S::SUB ) . The plasmid was verified by sequencing . Transformation of sub-1 , doq-1 , qky-8 and zet-2 mutants with the 35S::SUB construct was performed using the floral dip method [116] and Agrobacterium tumefaciens strain GV3101 [117] . Transgenic T1 plants were selected on 50 µg/ml Kanamycin plates and subsequently transferred to soil for phenotypic inspection . About 41 percent of the independent 35S::SUB sub-1 T1 plants scored showed a wild-type phenotype ( 51/126 scored ) . This indicates that the c-myc tag at the carboxy-terminus of SUB may result in a reduction of SUB functionality . To assay transgene expression RNA was isolated from inflorescences using the NucleoSpin RNA II kit ( Macherey-Nagel , Düren , Germany ) . First-strand cDNA was synthesized from 2 µg of total RNA using Moloney Murine Leukemia Virus ( M-MuLV ) reverse transcriptase ( New England Biolabs , Frankfurt , Germany ) . Semi-quantitative PCR was performed using Taq DNA polymerase ( New England Biolabs , Frankfurt , Germany ) and a transgene-specific primer pair ( sRT-SUBcmyc_F , sRT-SUBcmyc_R ) . Between 24 and 32 thermal cycles were tested . The GAPC gene was used as positive control [118] . To map the QKY locus at high resolution , an F2-mapping population was generated . F1 plants from qky/qky ( Ler ) and QKY/QKY ( Col ) crosses were allowed to self-pollinate , and the F2 progeny were screened for qky individuals based on twisted inflorescence morphology . DNA was isolated and used for PCR-based amplification of molecular markers . Indel polymorphism data was derived from the Monsanto Ler sequence database at TAIR [119] . Primer sequences for indel and CAPS markers are shown in supporting Table S1 . Marker amplifications from 598 mutant individuals restricted the qky map interval to 103 kb , as defined by markers F25A4 ( BglII ) and 27 . 99 ( RsaI ) . Candidate genes were analysed via T-DNA insertion mutant analysis and/or sequence determination revealing that At1g74720 carried mutations in various qky alleles . To confirm qky allelic mutations , nucleotide sequences were obtained from both strands of PCR-amplified fragments . In all three EMS-induced qky alleles transitions result in stop codons . The qky-7 allele carries a G to A transition at position 2229 ( genomic DNA , relative to the start ATG triplet ( +1 ) ) resulting in a shorter predicted protein ( W743* ) . The qky-8 allele carries another G to A transition at position 2706 ( W902* ) , and in qky-9 a C to T transition was found at position 649 ( Q217* ) . We also determined the genomic integration sites of two T-DNA insertions in At1g74720: SALK_140123 is located at position 2576 and SALK_043901 at position 3056 . Both lines exhibit a qky phenotype . SALK_140123 is predicted to carry a shorter QKY protein of 878 residues with the 29 last residues ( RSHKGSHVMTPADDAGQAVLRLELTEPQR* ) being encoded by the T-DNA . SALK_043901 results in a predicted shorter protein of 1022 residues with residues 1019–1022 ( LFVV* ) being encoded by the T-DNA . Near full-length QKY cDNA sequence was assembled from sequences derived from four publicly available RIKEN RAFL cDNA clones [120] . The cDNA clones partially overlapped ( RAFL16-35-G10 , RAFL22-02-B15 , RAFL22-66-H20 , RAFL22-96-F19 ) with one clone ( RAFL22-96-F19 ) containing the 3′ poly ( A ) stretch . Additional 5′ RACE experiments [121] did not result in more extended 5′ cDNA sequences and comparisons of the available QKY genomic and cDNA sequences did not reveal introns . Protein domain searches were conducted using the PFAM database [64] . Transmembrane topology was predicted using the TMHMM webserver [122] . Tissue was harvested as described for the whole-genome transcriptome analysis . RNA was extracted as outlined above . First-strand cDNA was synthesized from 1 . 5 µg of total RNA via reverse transcription , using the Transcriptor High Fidelity cDNA Synthesis Kit ( Roche Diagnostics GmbH , Mannheim , Germany ) . Quantitative real-time PCR was performed on a Roche LightCycler using the SYBR Green I detection kit according to the manufacturer's recommendations ( Roche ) . Amplification of UBC21/At5g25760 served as a normalization control [123] . Primer sequences are summarized in supporting Table S1 . Using the comparative Ct method , all gene expression levels were calculated relative to UBC21 . Preparation and analysis of propidium iodide-stained samples for confocal laser scanning microscopy , scanning electron microscopy , and histochemical localisation of ß-glucuronidase ( GUS ) activity in whole-mount tissue was done essentially as described [124] , [33] , [125]–[128] . Confocal laser scanning microscopy was performed with an Olympus FV1000 setup using an inverted IX81 stand and FluoView software ( FV10-ASW version 01 . 04 . 00 . 09 ) ( Olympus Europa GmbH , Hamburg , Germany ) . After excitation at 488 nm with a multi-line argon laser , propidium iodide fluorescence ( 580–630 nm slit width ) and autofluorescence ( 500–530 nm slit width ) was detected . One-way scan images ( scan rate 12 . 5 s/pixel , 512×512 pixels , Kahlman frame , average of four scans ) were obtained using an Olympus 40× objective ( UApo/340 40×/1 . 35 Oil Iris ) . Confocal Z-stack imaging of floral meristems was performed using 1 . 5 m sections . Plants or various plant organs were analysed under an Olympus SZX12 stereo microscope . Images were obtained using a ColorView III digital camera ( Olympus ) and Cell∧P software ( version 2 . 4 build 1131 , Olympus ) , saved as TIFF files , and adjusted for color and contrast using Adobe Photoshop CS3 ( Adobe , San Jose , CA , USA ) software . Tissue was harvested from 21-day plants grown in continuous light at 23°C . Wild type and mutants were in the Ler background . The inflorescence apex plus flowers up to stage 9 were kept separate from later stage 10–12 flowers ( the oldest 3–4 closed flower buds ) . The tissue was immediately frozen in liquid nitrogen and stored at −80°C . RNA was extracted with the Plant RNeasy Mini kit ( Qiagen ) . 1 µg of total RNA was used for probe synthesis with the MessageAmp II-Biotin Enhanced kit ( Ambion ) according to the manufacturer's instructions . Affymetrix ATH1 GeneChips were hybridised , washed , and stained as described [129] . The gcRMA implementation in the Bioconductor and R software packages was used for background correction , quantile normalization and expression estimate computation [130] , [131] ( www . bioconductor . org ) [132] . Expression estimates were transformed to linear scale and the fold change ( FC ) was calculated for each gene in each mutant-wildtype comparison by dividing the expression estimate of the mutant condition by the value of an according control . Comparisons were done on a single replicate basis for all possible pair wise comparisons . A Z-score for each fold change was calculated by dividing the difference of log2 transformed FC and the mean of the FC population by the standard deviation of the FC population ( ) . Genes were considered as differentially expressed if they displayed a Z-score>1 in a single comparison and an average Z-score>2 over all comparisons or , if they displayed a Z-score<−1 in a single comparison and an average Z-score over all comparisons <−2 . The QKY cDNA sequence was deposited at GenBank under the accession number FJ209045 . The expression profile data have been deposited to the EMBL-EBI ArrayExpress repository under the accession E-MEXP-1592 . | Plant organs , such as flowers or leaves , are made up of distinct cell layers . Although communication across these cell layers is essential for organ development , we have only recently gained some insight into the underlying mechanisms . Receptor-like kinases are cell-surface receptors that perceive and relay intercellular information . In Arabidopsis , the receptor-like kinase STRUBBELIG is required for inter–cell-layer communication during floral development , amongst other functions; little is known , however , concerning its exact signaling mechanism . Here , we identified three new genes called DETORQUEO , QUIRKY , and ZERZAUST . Plants defective in any of these genes strongly resemble the strubbelig mutant , both at the whole-organ and cellular levels . Thus , all four genes may share or contribute to a common signaling pathway essential for plant morphogenesis . Analyses revealed complex interactions between the genes , indicating that each has additional and distinct activities . We provide the molecular nature of QUIRKY; the encoded protein is likely membrane-localised and predicted to require Ca2+ for activity . In light of analogous animal models , we speculate that QUIRKY facilitates transport of molecules to the cell boundary and may support a STRUBBELIG-related extracellular signal . These results open new inroads into a molecular understanding of inter-cellular communication during flower development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant-biotic",
"interactions",
"cell",
"biology/cell",
"signaling",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"cell",
"biology/cell",
"growth",
"and",
"division",
"plant",
"biology/plant",
"cell",
"biology",
"cell",
"biology/developmental"... | 2009 | DETORQUEO, QUIRKY, and ZERZAUST Represent Novel Components Involved in Organ Development Mediated by the Receptor-Like Kinase STRUBBELIG in Arabidopsis thaliana |
We report the identification and characterization of a previously unknown suppressor of myopathy caused by expansion of CUG repeats , the mutation that triggers Myotonic Dystrophy Type 1 ( DM1 ) . We screened a collection of genes encoding RNA–binding proteins as candidates to modify DM1 pathogenesis using a well established Drosophila model of the disease . The screen revealed smaug as a powerful modulator of CUG-induced toxicity . Increasing smaug levels prevents muscle wasting and restores muscle function , while reducing its function exacerbates CUG-induced phenotypes . Using human myoblasts , we show physical interactions between human Smaug ( SMAUG1/SMAD4A ) and CUGBP1 . Increased levels of SMAUG1 correct the abnormally high nuclear accumulation of CUGBP1 in myoblasts from DM1 patients . In addition , augmenting SMAUG1 levels leads to a reduction of inactive CUGBP1-eIF2α translational complexes and to a correction of translation of MRG15 , a downstream target of CUGBP1 . Therefore , Smaug suppresses CUG-mediated muscle wasting at least in part via restoration of translational activity of CUGBP1 .
Myotonic Dystrophy type 1 ( DM1 ) is a multisystemic neuromuscular disorder that has become a paradigm of a class of diseases caused by RNA toxicity . DM1 arises from expansion of a CTG triplet repeat in the 3′ untranslated region of the DMPK gene , and it accounts for the majority of adult cases of muscular dystrophy [1]–[5] . In DM1 the CUG-expanded mRNA is trapped in the nuclei where it forms nuclear foci and sequesters MBNL1 protein leading to loss of its activity [6] , [7] . In addition , the mutant mRNA leads to increased steady-state levels of CUGBP1 ( a . k . a CELF1 ) [8] , [9] through its stabilization as a result of PKC phosphorylation [10] . Both MBNL1 and CUGBP1 are RNA-binding proteins involved in regulation of splicing [11]–[14] , and aberrant splicing of the insulin receptor [12] , muscle-specific chloride channel [13] , [15] and many other genes [16] , [17] occur in DM1 . The critical significance of MBNL1 sequestration for DM1 pathogenesis is eloquently demonstrated in loss of function and overexpression experiments . MBNL1 mutant mice show cataracts , myotonia , and other muscle abnormalities [7] that closely resemble a number of DM1 pathological features , and they also share many of the splicing aberrations observed in transgenic mice expressing CUG repeats [16] , [17] . Importantly , MBNL1 overexpression ameliorates , muscle wasting in a Drosophila DM1 model [18] , and myotonia and splicing aberrations in mouse models [19] . Evidence of the relevance of increased steady-state levels of CUGBP1 in DM1 pathogenesis comes from overexpression experiments . Transgenic mice expressing CUGBP1 show delays in muscle development and differentiation [20] , muscle wasting [21] , splicing misregulation [22] and DM1-like cardiac abnormalities [23] . Besides its nuclear role in splicing , CUGBP1 also has other functions in the cytoplasm including regulation of mRNA translation and stability [24]–[26] . Alterations of protein [25] and mRNA [16] levels occur in DM1 consistent with the idea that perturbation of CUGBP1 cytoplasmic functions contribute to DM1 pathogenesis . CUGBP1 cellular localization depends on its phosphorylation status [25] . Several kinases phosphorylate CUGBP1 at different residues and affect its localization within the cell . Activation of the Akt pathway increases CUGBP1 phosphorylation at Ser-28 altering the transition from proliferating myoblasts to differentiated myotubes in DM1 [27] . On the other hand , DM1 cells show decreased activity of cyclin D3-cdk4 , another kinase that phosphorylates CUGBP1 . This renders higher levels of unphosphorylated CUGBP1 , which forms inactive complexes with eIF2α ( CUGBP1-eIF2α ) affecting translation of mRNAs required for myoblast differentiation . These inactive complexes containing CUGBP1 accumulate in the cytoplasm of DM1 cells in stress granules ( SG ) [25] . The richness of evidence implicating CUGBP1 in DM1 pathogenesis suggests the possibility that correcting the abnormal levels and activity of CUGBP1 may be a therapeutic approach to ameliorate DM1 pathogenesis . In support of this idea , Wang and colleagues used a pharmacological approach to inhibit PKC in mice expressing ( CUG ) 960 in the heart; this treatment ameliorates the mortality rates and cardiac conduction as well as contractile abnormalities in this heart-specific DM1 mouse model [28] . Additional evidence comes from the observation that overexpression of a nuclear dominant negative CUGBP1 protein reverses dysregulation of a splicing minigene reporter in cultured cells , and of the CUGBP1 target Nrap exon 12 in DM1 mice [29] . Here we report that smaug , a gene not previously known to be implicated in DM1 , is a powerful suppressor of CUG-induced myopathy when overexpressed in Drosophila . We show that human SMAUG1 protein ( a . k . a SAMD4A ) interacts with CUGBP1 and decreases its abnormally high steady-state levels in DM1 nuclei . Furthermore , increasing the levels of SMAUG1 in myoblasts of DM1 patients decreases the amount of inactive CUGBP1-eIF2α translational complexes . This suggests that SMAUG1 improves the activity of the CUGBP1-containing translational complexes that are dysfunctional in DM1 , a hypothesis that is supported by data showing SMAUG1-mediated increased translation of the CUGBP1 translational target MRG15 in DM1 myoblasts .
To identify previously unknown genes implicated in DM1 pathogenesis , we used a well characterized Drosophila DM1 model [18] . Since DM1 is caused by expansion of an untranslated transcript , and MBNL1 and CUGBP1 are themselves RNA-binding proteins , we hypothesized that DM1 modifier genes may be enriched among genes encoding RNA binding proteins ( RNA-BPs ) . Thus , we screened a collection of 93 loss of function and 17 overexpression alleles in 73 RNA-BP genes for their ability to modulate pathogenesis caused by expanded CUG repeats . First , we used an external eye phenotype induced by expression of ( CUG ) 480 as a primary screen to identify genes able to ameliorate or enhance CUG-induced toxicity . To validate the identified modifiers we tested the ability of the primary screen hits to modify CUG-induced muscle wasting . Among the RNA-BPs tested , we uncovered the Drosophila gene smaug as a strong modifier of both the eye and muscle degeneration . As shown in Figure 1 increased levels of Smaug rescue the eye disorganization and loss of bristle phenotypes induced by ( CUG ) 480 expression ( compare Figure 1C with 1B ) . Consistent with this result , partial loss of function of smaug caused by a heterozygous mutation enhances the ( CUG ) 480-induced eye phenotype ( compare Figure 1D with 1B ) . As shown in Figure S1 these overexpression and loss-of-function alleles do not induce any abnormal phenotypes in control animals that do not express expanded CUG repeats . The DM1 Drosophila model shows progressive muscle wasting , which is easily studied in the indirect flight muscles of the thorax . While 1-day-old ( CUG ) 480 flies have muscles that appear wild type , animals that are 20 days old show muscle disorganization , vacuolization and loss of muscle fibers [18] . We investigated the effect of increasing the levels of smaug on ( CUG ) 480-induced muscle wasting . As shown in Figure 1E overexpression of smaug dramatically suppresses CUG-induced myopathy . Next we investigated whether increased smaug levels could restore muscle function in addition to muscle integrity . Animals expressing ( CUG ) 480 show a severe impairment in flying ability prior to showing any signs of muscle wasting by histological analysis ( see green bars in Figure 1F ) . Increased levels of smaug sharply improve flying ability in animals expressing ( CUG ) 480 . ( compare orange and green bars in Figure 1F ) . These muscle histology and behavioral data further support the idea that smaug is a suppressor of expanded-CUG toxicity in a variety of cellular contexts . In addition , we investigated whether Drosophila Smaug protein and the expanded-CUG RNA co-localize in nuclear foci . Previous studies have shown that Smaug accumulates in cytoplasmic foci similar to stress granules , and that it can shuttle between the nucleus and the cytoplasm [30] . To determine whether Smaug protein localization is altered due to expression of ( CUG ) 480 , we performed in situ and immunofluorescense analysis of larval muscles of animals expressing ( CUG ) 480 . As shown in Figure 1G Smaug accumulates mainly in the cytoplasm in the form of granules ( Figure 1G , green ) , and it does not co-localize with the nuclear CUG-containing foci ( Figure 1G , red , NF ) . This observation suggests that the mechanism by which Smaug modulates expanded-CUG toxicity does not involve direct interaction with the nuclear foci . The data described above and shown in Figure 1G does not suggest sequestration of Smaug in nuclear foci as a mechanism for Smaug modification of expanded-CUG toxicity . Consequently , we investigated possible interactions between smaug and the known key players in DM1 pathogenesis: MBNL1 and CUGBP1 . Overexpression of human MBNL1 or CUGBP1 in the Drosophila eye leads to a mild disorganization phenotype [18] , and Figure 2A and 2D . We used these phenotypes as assays to test potential genetic interactions with smaug . We found that smaug overexpression suppresses the phenotype induced by CUGBP1 expression ( compare Figure 2B with Figure 2A ) . In addition , smaug partial loss of function enhances this phenotype ( compare Figure 2C with Figure 2A ) . In contrast , altering smaug levels does not have an effect on the MBNL1-induced eye phenotype ( Figure 2D–2F ) . In summary , we find that CUGBP1 and Smaug interact genetically in Drosophila . To further investigate the interaction between SMAUG1 and CUGBP1 , we performed immunofluorescense on COSM6 cells transfected with SMAUG1 and ( CUG ) 960 . We found that CUGBP1 localizes predominantly in the nucleus in cells transfected only with ( CUG ) 960 ( see arrowhead in Figure 3A ) , an observation that is consistent with previous reports [8] , [9] , [31] , [32] . We found , however , that nuclear CUGBP1 steady-state levels are significantly decreased in cells transfected with both ( CUG ) 960 and SMAUG1 ( Figure 3B , arrowhead ) . CUGBP1 can be seen in these cells both diffuse in the cytoplasm as well as co-localizing with SMAUG1 in cytoplasmic granules ( Figure 3B , arrow ) . As control we transfected with GFP and we could not observe differences in CUGBP1 signal between GFP-transfected ( Figure 3C , arrow ) and GFP-untransfected ( Figure 3C , arrowhead ) cells . A similar experiment was performed with MBNL1 and SMAUG1 , but we found no evidence of changes in the accumulation of MBNL1 in nuclear foci following expression of SMAUG1 ( Figure S2 ) . To validate these data on a cell type more relevant to DM1 , we investigated whether CUGBP1 distribution is altered by SMAUG1 expression in myoblasts from DM1 patients . DM1 myoblasts transfected with GFP show predominantly nuclear CUGBP1 signal ( Figure 4A ) . In contrast , DM1 myoblasts transfected with SMAUG1 show significantly decreased levels of nuclear CUGBP1 ( Figure 4B , compare intensity of CUGBP1 staining in SMAUG1-transfected ( arrowhead ) vs . untransfected myoblasts ) . We quantified the intensity of the signal of nuclear CUGBP1 staining in DM1 myoblasts transfected with SMAUG1 versus controls transfected with GFP , and we found that SMAUG1-transfected DM1 myoblasts show a significant decrease in the nuclear signal intensity compared to controls transfected with GFP ( Figure 4D , p<0 . 0001 ) . In addition we find that in SMAUG1-transfected DM1 myoblasts CUGBP1 and SMAUG1 co-localize in cytoplasmic granules ( Figure 4C , arrows , and Figure S3 ) . Cytoplasmic co-localization of both proteins was also observed in normal myoblast ( Figure 5B ) . In spite of the observation that SMAUG1-expressing DM1 myoblasts show reduced nuclear CUGBP1 , we did not detect an increase on cytoplasmic CUGBP1 in DM1 myoblasts transfected with SMAUG1 when compared to GFP-transfected controls ( Figure S3 , see also western blot of cytoplasmic fraction in Figure 6A ) . In control non-DM1 myoblasts transfected with SMAUG1 nuclear CUGBP1 signal remains the same ( Figure 5 ) . Prompted by the genetic interaction between Drosophila smaug and human CUGBP1 and the co-localization of SMAUG1 and CUGBP1 in cells , we also investigated whether human SMAUG1/SAMD4A and CUGBP1 proteins physically interact . To do so , we performed co-immunoprecipitation experiments with cellular extracts from human myoblasts expressing SMAUG1 . As shown in Figure 4E , SMAUG1 signal is detected after pull-down with anti-CUGBP1 antibody both in normal and DM1 myoblasts . The intriguing finding that increased levels of SMAUG1 leads to decreased nuclear accumulation of CUGBP1 suggests that restoration of normal alternative splicing patterns may explain SMAUG1-mediated suppression of CUG-induced myopathy . To test this potential mechanism of SMAUG1 suppression , we examined the alternative splicing changes induced by either expanded CUG repeats or CUGBP1 overexpression . Using a cTNT minigene reporter we found no evidence that overexpression of SMAUG1 restores normal alternative splicing changes caused by either expanded CUG repeats or CUGBP1 overexpression ( Figure S4 ) . Since we did not find evidence that increased SMAUG1 restore alternative splicing patterns , we investigated whether they restore CUGBP1 normal function in the cytoplasm . CUGPB1 regulates the translation and stability of mRNAs , and these activities are impaired in DM1 [24]–[26]; thus , we asked if the translational activity of CUGPB1 is influenced by SMAUG1 . In the cytoplasm , CUGBP1 interacts with eukaryotic translation initiation factor eIF2α ( eIF2α ) , and its translational activity is mediated by CUGBP1-eIF2α complexes [33] . CUGBP1 phosphorylated at S302 binds to unphosphorylated eIF2α ( non-pS51-eIF2α ) making active CUGBP1-eIF2α translational complexes , whereas CUGBP1 unphosphorylated at S302 binds with higher affinity to inactive pS51-eIF2α forming CUGBP1-eIF2α inactive translational complexes [25] . In DM1 cells , the levels of inactive eIF2α ( pS51-eIF2α ) are increased , and formation of inactive CUGBP1-eIF2α complexes inhibits translation of certain mRNAs in DM1 myoblasts [25] . Therefore , we examined the effects of SMAUG1 on the abundance of inactive CUGBP1-eIF2α complexes . Western blot analysis of cytoplasmic extracts from transfected normal and DM1 myoblasts and fibroblasts show that the total cytoplasmic levels of CUGBP1 are increased in DM1 myoblasts and fibroblasts , and are not affected significantly by SMAUG1 ( Figure 6A–6B and Figure S5 ) . Additionally , we investigated if levels of total eIF2α are altered by SMAUG1 expression in cytoplasm . As shown in Figure 6A and 6B , total levels of eIF2α remain unchanged upon SMAUG1 transfection in both normal and DM1 myoblasts and fibroblasts . We then performed co-IP experiments on cytoplasmic protein extracts from normal and DM1 myoblasts and fibroblasts to test whether SMAUG1 expression alters the levels of inactive CUGBP1-eIF2α complexes . We found that in control GFP-transfected DM1 myoblasts/fibroblasts pS51-eIF2α-CUGBP1 inactive complexes are abundant . This is in striking contrast to SMAUG1-transfected DM1 myoblasts ( Figure 6A , see CUGBP1-IP ) and Figure S6 ) and fibroblasts ( Figure 6B , see CUGBP1-IP ) and Figure S6 ) where these complexes are undetectable . Thus , increasing SMAUG1 levels decreases the steady-state levels of CUGBP1-eIF2α inactive translational complexes . Previous reports have shown that MRG15 mRNA translation is controlled by CUGBP1-eIF2α complexes . Particularly , inactive CUGBP1-eIF2α complexes trap MRG15 mRNA in stress granules and reduces protein levels of MRG15 in DM1 compared to normal myoblasts [25] . Since we found that expression of SMAUG1 reduces amounts of inactive CUGBP-eIF2α complexes , we investigated if this reduction corrects translation of MRG15 , a target of the CUGBP-eIF2α complex . Western blot analysis of nuclear protein extracts shows that DM1 cells contain reduced amounts of MRG15; however , SMAUG1 restores translation of MRG15 in DM1 cells to near normal levels in both myoblasts and fibroblasts ( Figure 6C , Figure S7 ) . In summary , these data indicate that expression of SMAUG1 significantly reduces the amounts of inactive CUGBP1-eIF2α complexes and enhances translation of MRG15 .
Here we show that increased expression levels of smaug , a conserved gene involved in translational regulation , suppresses CUG-induced muscle wasting and , notably , it also restores normal muscle function in a Drosophila model of DM1 . Experiments in DM1 myoblasts indicate that the human homolog SMAUG1/SAMD4A suppresses the toxic effects of expanded CUG repeats at least in part by restoring impaired CUGBP1 translational functions . Early during DM1 pathogenesis CUGBP1 steady-state levels increase as a consequence of PKC-mediated phosphorylation [10] . This leads to disrupted regulation of alternative splicing , as well as impairments in mRNA stability and mRNA translation , all of which contribute to the multiple features of the disease ( reviewed in [34] , [35] ) . In addition , CUGBP1 overexpression in wild-type mice mimics some of the functional , histopathological and molecular features of DM1 [22] , [36] , [37] , while CUGBP1 overexpression in Drosophila enhances expanded-CUG induced pathology [18] . Together these observations suggest that restoring normal CUGBP1 levels and activities may reverse DM1 pathology . This approach however may prove difficult to execute . First , there are several CUGBP1-like proteins in mammals and in Drosophila making proof-of-principle experiments using loss-of-function alleles complicated . To circumvent the problem of functional redundancy , a dominant-negative CUGBP1 construct was expressed in culture cells and mice , and this resulted in the reversion of abnormal alternative splicing [29] . Expression of dominant-negative CUGBP1 , however , also leads to cardiac and skeletal muscle pathology [38] , [39] . A second and perhaps more important caveat is that CUG expansion leads to increased nuclear levels of CUGBP1 [8] , [40] ( i . e . , a gain of function ) , while in the cytoplasm the same mutation leads to the inactivation of CUGBP1 translational complexes [25] ( i . e . , loss of function ) . Hence , restoring normal CUGBP1 activities in both nucleus and cytoplasm by modulating CUGBP1 itself seems challenging . An alternative approach is to target other factors modulating CUGBP1 function . One such factor is PKC , a kinase that phosphorylates and stabilizes CUGBP1 [10] . Indeed , PKC pharmacological inhibition ameliorates the cardiac phenotypes in a heart-specific DM1 mouse model [28] . The data presented here reveals that SMAUG1/SAMD4 is able to restore CUGBP1 normal levels and activities . We found that increasing the levels of SMAUG1 leads to decreased levels of nuclear CUGBP1 in DM1 myoblasts . This intriguing observation suggested that rescue of alternative splicing alterations may be a possible mechanism to explain the observed suppression of CUG-induced myopathy . This is an open possibility because even though we did not find evidence of SMAUG1 modulating splicing on a cTNT minigene , we cannot rule out its effects on other unknown splicing targets . We showed that SMAUG1 can re-activate impaired CUGBP1 translational activities in the cytoplasm . smaug was first discovered in Drosophila as a translational regulator of nanos mRNA in the posterior pole of the embryo [41] . In this context it functions as a translational repressor by capturing transcripts containing Smaug recognition elements , forming stable ribonucleoprotein particles , and displacing the eIF4G initiation factor [42] . Smaug also promotes the destabilization and degradation of nanos mRNA by recruiting a deadenylation factor [43]–[45] . There are two smaug homologous genes in mammals [30] . One of them , SMAUG1/SAMD4A , forms mRNA-silencing foci at postsynapses of hippocampal neurons that respond to NMDA and modulate synapse formation [46] . We find that SMAUG1 has a positive function in the context of CUGBP1-dependent translation in myoblasts suggesting that SMAUG1 is not a dedicated repressor of translation , but rather a translational regulator whose function is context dependent . In DM1 , high levels of CUGBP1 unphosphorylated at S302 form inactive translational complexes with pS51-eIF2α . We found that increased levels of SMAUG1 lead to a dramatic reduction of CUGBP1-eIF2α inactive complexes . It is unlikely that this is a result of nuclear CUGBP1 being exported to the cytoplasm because we did not detect an increase of CUGBP1 in western blots of cytoplasmic extracts from SMAUG1-transfected DM1 myoblasts or fibroblasts . This observation was confirmed by immnofluorescence experiments showing similar levels of cytoplasmic CUGBP1 between SMAUG1-transfected and GFP-transfected DM1 myoblasts . An attractive possibility is that the interaction between SMAUG1 and CUGBP1 promotes repair of defective initiation complexes . In support of this hypothesis we observe an increase in translation of MRG15 . Translation of MEF2A , C/EBPbeta , p21 , and other translational CUGBP1 targets such as cyclin D1 and HDAC1 are promoted by active CUGBP1/elF2 complexes ( i . e . , formed by p-S302-CUGBP1 and elF2 not phosphorylated at S51 ) . However , we only know of one target , MRG15 , whose translation is inhibited by inactive CUGBP1/elF2 complexes ( i . e . , formed by CUGBP1 not phosphorylated at S302 , and p-S51-elF2 ) [25] . Thus , we expect that other mRNA targets of CUGBP1 whose translation is impaired in DM1 may be corrected as well; however , these other mRNAs have not been identified yet . The only therapy available for DM1 patients is used to treat the symptoms rather than the cause of the disease . Efforts to develop therapeutic avenues for DM1 pathogenesis include: 1 ) to revert the instability of the expansion , 2 ) to target the toxic RNA with ribozymes or antisense oligonucleotides [47]–[51] , 3 ) to target the CUG RNA hairpins with siRNA [52] . Potential alternatives are to develop therapeutic approaches to restore CUGBP1 and MBNL1 protein levels and activities [18] , [19] , [28] ( reviewed in [53] ) . The data reported here suggests that therapeutics designed to increase SMAUG1 protein levels could be useful to ameliorate the toxicity of the mutant RNA in DM1 .
The transgenic lines UAS- ( CTG ) 480 , UAS-MBNL1 , and UAS-CUGBP1 have been previously described [18] . Mhc-GAL4 was obtained from G . Davis ( UCSF ) . gmr-GAL4 , smg1 and smgEP3556 were obtained from Bloomington Stock Center ( Indiana ) and Szeged Stock Center ( Hungary ) . Processing of flies for SEM and image acquisition were performed following previously published procedures [54] . For paraffin sections , adult thoraxes were dissected out , fixed overnight in 4% formaldehyde in PBS , washed in PBS and dehydrated in increasing concentrations of ethanol . Thoraxes were embedded in paraffin . Serial sections of 10 µm were obtained and rehydrated to water . Sections were stained with eosin ( Sigma ) and the fluorescent images were captured using an AxioCam MRc camera ( Zeiss ) attached to a Microphot-FXA microscope ( Nikon ) . Individual adult flies were dropped one at a time from the top of a 12-inch cylinder and the landing position in the cylinder was recorded . One hundred flies per genotype were scored and each fly was tested three times . The protocols were previously described in [18] . Anti-Smaug antibody ( provided by C . A . Smibert ) was used at a concentration of 1∶50 . Constructs used for transfection were ( CUG ) 960 ( T . A . Cooper ) and SMAUG1-ECFP ( G . L . Boccaccio ) . COSM6 cells were transfected with ( CUG ) 960 alone or together with SMAUG1-ECFP using Amaxa Cell Line Nucleofector Line R ( Lonza ) . Two days after transfection cells were fixed in 4% formaldehyde for one hour , washed and hybridized with a Cy3-labelled 5′-CAG-3′ LNA probe for one hour , followed by incubation with mouse anti-CUGBP1 3B1 antibody ( 1∶120 ) overnight at 4C . Secondary anti-mouse antibody labelled with Cy5 was used to visualize CUGBP1 . Human primary myoblasts derived from control individuals or from DM1 patients with 300 CTG repeats were grown for no more than 12 passages and transfected with SMAUG1-ECFP or control pmaxGFP using Amaxa Cell Line Nucleofector Line NHDF ( Lonza ) . Two days after transfection in situ and immunofluorescense was performed as described in the above paragraph . For quantification of CUGBP1 nuclear signal , pictures taken at the confocal microscope under the same conditions were analyzed using ImageJ software . Pictures of at least 50 different cells were taken . Data sets were compared using ANOVA followed by Student's t analysis . Transfection of myoblasts and fibroblasts with SMAUG1-V5 ( G . L . Boccaccio ) was performed using Amaxa Nucleofector Line NHDF ( Lonza ) . For co-immunoprecipitation of myoblasts in Figure 4E , anti-CUGBP1 3B1 antibody ( Novus Biologicals ) was coupled to Dynabeads M-270 Epoxy ( Invitrogen ) , and co-IP was performed with Dynabeads Co-Immunoprecipitation kit ( Invitrogen ) using anti-V5 ( Invitrogen ) antibodies . For western blot analysis , control and DM1 myoblasts ( 300 CTG repeats ) and fibroblasts ( 2000 and 1600 CTGs ) were transfected as described above . Two days after transfection nuclear and cytoplasmic protein fractions were extracted [31] . Twenty five µg of cytoplasmic proteins were separated by gel electrophoresis , transferred on membrane and incubated with anti-CUGBP 3B1 and anti-eIF2α ( FL-315 , Santa Cruz , CA , USA ) . Co-IP with 100 µg of cytoplasmic protein was performed using the protocol associated with Trueblot Antibodies from eBioscience . Antibody for pS51-eIF2α was S51-sc12412-R from Santa Cruz . For detection of MRG15 , nuclear protein fractions of cells transfected with SMAUG1 or GFP were separated by gel electrophoresis , transferred on membrane and incubated with anti-MRG15 ( F-19 ) and anti-β-actin ( AC-15 ) from Santa Cruz . | Myotonic dystrophy type 1 ( DM1 ) is the most common among the muscular dystrophies causing muscle weakness and wasting in adults , and it is triggered by expansion of an untranslated CUG repeat . To identify potential therapeutic approaches , we used a Drosophila DM1 model to screen for genes capable of suppressing CUG-induced toxicity . Here we report that increased levels of the smaug gene prevent muscle wasting and , perhaps more impressively , also prevent muscle dysfunction caused by the DM1 mutation . Smaug interacts genetically and physically with CUGBP1 , an RNA–binding protein previously implicated in DM1 . We used myoblasts from DM1 patients and control individuals to investigate how Smaug suppresses CUG-induced myopathy . We found that increased human SMAUG1 ( a . k . a . SMAD4A ) levels revert the abnormal accumulation of CUGBP1 in myoblasts nuclei and restore normal translation of at least one mRNA regulated by CUGBP1 in the cytoplasm . These findings demonstrate that manipulating Smaug activity protects against the effects of the DM1 mutation , and they also support the idea that restoring normal CUGBP1 function is a potential therapeutic approach . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"genetic",
"screens",
"genetics",
"biology",
"human",
"genetics",
"molecular",
"cell",
"biology",
"autosomal",
"dominant"
] | 2013 | Smaug/SAMD4A Restores Translational Activity of CUGBP1 and Suppresses CUG-Induced Myopathy |
Attachment proteins from the surface of eukaryotic cells , bacteria and viruses are critical receptors in cell adhesion or signaling and are primary targets for the development of vaccines and therapeutic antibodies . It is proposed that the ligand-binding pocket in receptor proteins can shift between inactive and active conformations with weak and strong ligand-binding capability , respectively . Here , using monoclonal antibodies against a vaccine target protein - fimbrial adhesin FimH of uropathogenic Escherichia coli , we demonstrate that unusually strong receptor inhibition can be achieved by antibody that binds within the binding pocket and displaces the ligand in a non-competitive way . The non-competitive antibody binds to a loop that interacts with the ligand in the active conformation of the pocket but is shifted away from ligand in the inactive conformation . We refer to this as a parasteric inhibition , where the inhibitor binds adjacent to the ligand in the binding pocket . We showed that the receptor-blocking mechanism of parasteric antibody differs from that of orthosteric inhibition , where the inhibitor replaces the ligand or allosteric inhibition where the inhibitor binds at a site distant from the ligand , and is very potent in blocking bacterial adhesion , dissolving surface-adherent biofilms and protecting mice from urinary bladder infection .
Receptor-ligand interactions are among the most basic biological phenomena involved in cell signaling , adhesion and pathogen attachment . Antibody- or small molecule-based inhibitors of these interactions are of great importance for various preventive and therapeutic implications , including development of protective vaccines . Two general types of inhibitory mechanisms have been described to date . Orthosteric inhibitors directly compete with ligands for the binding pocket and , thus , their receptor-inhibitory activity is of a competitive nature [1] . In contrast , allosteric inhibitors exert their effects via interaction with a site that is separate from the ligand-binding pocket and accomplish the inhibition in a non-competitive manner [2] . Non-competitive inhibition is less sensitive to endogenous ligand and thus is generally more effective pharmacologically [3] . In the current study , we describe a type of inhibitory monoclonal antibody against the mannose-binding adhesin of E . coli , FimH , that does not fall into either of the expected two categories of inhibitors . Like an allosteric inhibitor , this antibody exerts non-competitive inhibition , but like an orthosteric inhibitor , it binds within the ligand-binding pocket . Unlike the latter , however , it forces the conversion of the binding pocket to an open , inactive conformation , even when the pocket is occupied by the ligand mannose . FimH is a 30 kDa lectin-like protein that is incorporated into the tip of surface hair-like structures of E . coli and other enterobacteria called type 1 fimbriae [4] . It exhibits specificity to glycoproteins carrying terminally exposed mannose and is critical for the virulence of uropathogenic strains of E . coli [5 , 6 , 7 , 8 , 9] . FimH has two domains: the C-terminal pilin domain that anchors the adhesin to the fimbrial rod and the N-terminal lectin domain that is responsible for mannose binding [10] . The binding pocket in the lectin domain shifts between open and tightened conformations with low ( KD = 298 μM ) - and high ( KD = 1 . 2 μM ) - affinity for mannose , respectively [11 , 12 , 13] . The low-affinity ( inactive ) state of the lectin domain is allosterically stabilized by its interaction with the pilin domain that sustains a finger-trap-like twist in the β-sheets of the binding domain [11] . The high-affinity ( active ) state is induced by ligand binding and/or separation of the domains , with the latter facilitated by force during bacterial adhesion under flow conditions . FimH-like force-activated adhesion has been described in several other adhesive systems of different bacterial species as well as eukaryotic cells . For example , proteins like integrins [14] or P/L-selectins [15] exhibit a shift between inactive and active conformations under shear force . The existence of two alternative conformations of the mannose-binding pocket of FimH reflects a broad phenomenon in the biology of receptor-ligand interactions , including enzyme binding to substrates and products . In fact , the century-old static ‘lock-and-key’ model of the interaction mechanism is considered now to be too rigid for many if not the majority of receptor proteins and enzymes . It has been shown that ligand-binding pockets are typically composed of residues on flexible loops and dynamically shift between active and inactive conformations , with relatively high and low ( often unmeasurable ) affinity for the ligand , respectively [16 , 17 , 18 , 19 , 20] . Generally , the ligand-bound active pocket assumes a more contracted shape than the ligand-free inactive pocket , so the corresponding receptor conformers are commonly referred to as open vs closed ( or tightened ) states [20 , 21 , 22 , 23] . Some well-studied examples of receptors with such pocket dynamics include allosteric proteins such as maltose-binding protein [24 , 25 , 26] , and G-protein-coupled receptors ( GPCRs ) [21 , 23 , 27] . Two general models have been proposed to describe the effect of ligand on the conformation of receptor binding pockets . In the ‘induced fit’ model , the active state of the pocket is assumed only after ligand binds to the inactive state , while in the ‘conformational selection’ model , the inactive and active states coexist in the absence of ligand , but the active state is stabilized by ligand binding [28 , 29 , 30] . More complex models of ligand-receptor recognition that combine the two models are also considered [31] . All models allow for initial weak interaction of the ligand with the inactive state of binding pocket , and this weak interaction has been repeatedly shown to involve only a subset of the receptor residues that interact with the ligand in the strongly-binding active state [16 , 20 , 22 , 23 , 31] . Partial interaction of the ligand with the binding pocket leaves the remaining residues , which only interact with the ligand when the pocket is in the active state , free in theory to bind to an additional compound . Such a compound could potentially act as an inhibitor by interfering with the switch of the pocket into the active state . Such an inhibitor would not fit the accepted definition of either orthosteric inhibitors that cannot bind simultaneously with ligand , or allosteric inhibitors that should bind away from the binding pocket . Instead , because such inhibitors would bind next to the ligand , they could be classified as parasteric inhibitors . Interestingly , the parasteric binding of effector molecules and ligand was predicted previously for an enzymatic protein , low-molecular-weight acid phosphatase 1 ( ACP1 ) , though the structural basis or mechanism was not investigated [32] . In the current study we compared the inhibitory mechanism of different anti-FimH antibodies and describe an antibody that blocks the adhesive function in a distinct manner consistent with the parasteric model of inhibition . Compared to an orthosteric antibody , the parasteric antibody was a more potent inhibitor against bacterial adhesion , surface-bound biofilms and in vivo colonization , demonstrating that design of parasteric inhibitors potentially represents a very powerful approach toward the development of anti-adhesive preventive and therapeutic strategies .
In a previous study , we characterized anti-FimH inhibitory antibodies raised against the lectin domain of FimH in the inactive conformation and described in detail the mAb475 antibody that directly competes with mannose binding [33] . Using IMGT/V-Quest software [34 , 35] we now have compared the germline origins of mAb475 and several other FimH-inhibiting monoclonal antibodies . One of the antibodies , mAb926 , was of a different germline origin from mAb475 , with the amino acid sequence homology of the mAb V-regions being only 55% and all three complementarity-determining regions being highly diverse ( S1 Fig ) . We therefore compared the abilities of mAb926 and mAb475 to inhibit FimH in greater detail . Type 1 fimbriated bacterial cells expressing the wild-type FimH adhesin variant of the uropathogenic E . coli strain J96 ( FimHwt ) were pre-incubated with different concentrations of the antibodies and then allowed to bind to mannosylated surfaces ( Fig 1 ) . mAb926 inhibited bacterial adhesion with half-maximal inhibitory concentration ( IC50 ) of 0 . 4 ± 0 . 1 nM and mAb475 with IC50 of 14 ± 1 nM indicating much higher inhibitory potency of mAb926 ( Fig 1 ) . To determine if the 32-fold difference ( P≤0 . 005 ) in inhibitory potency of the antibodies reflected a difference in binding affinity , we characterized binding of the antibodies to the purified fimbriae carrying FimHwt with surface plasmon resonance . As shown in S2 Fig , the KD of mAb926 was 7-fold lower than KD of mAb475 ( 0 . 58 vs . 4 . 15 nM , respectively ) . Thus , the difference in IC50 between the two antibodies could not be explained by a difference in affinity . Indeed , mAb926 demonstrates >50% inhibition at its KD concentration , while mAb475 demonstrates no measurable inhibition at its KD concentration ( Fig 1 ) . Moreover , because mAb926 had a 17 . 6-fold higher association rate relative to mAb475 ( 49 . 7 vs . 2 . 8 x104 M-1s-1 , respectively ) , and a 2 . 5-fold higher dissociation rate ( 2 . 89 vs . 1 . 17 x10-4s-1 , respectively ) , the difference in affinity was due to the faster association rate of mAb926 . The SPR experiments were performed in parallel at the Analytical Biopharmacy Core , at the University of Washington ( with separate preparations of the antigenic substrate ) , with results that are completely consistent with those reported here , adding confidence to our analyses ( S1 Table ) . Since in the inhibition assay the antibodies were pre-incubated with the bacteria for one hour , the saturation of binding has likely been reached for both antibodies . Thus , at that point the difference in the dissociation rate of the inhibitory antibodies would be more important than their association rates , so the increased effectiveness of mAb926 is even more remarkable considering its slightly higher dissociation rate . Thus , taken together , these results demonstrate the significantly higher inhibitory potential of mAb926 than of mAb475 antibodies cannot be explained by differences in binding kinetics or affinity of the two antibodies . Instead , the large difference in inhibitory potency must reflect some unknown difference in the mechanism of inhibition . We next compared the ability of mAb475 and mAb926 to bind FimH in the presence of soluble mannose . As shown in Fig 2A , mannose strongly inhibited mAb475 binding , causing a significant shift of its binding curve towards higher concentrations of the antibody . The mAb475 half-maximal effective concentration ( EC50 ) increased 179-fold in the presence of soluble mannose . In contrast , binding of mAb926 was affected by mannose to a much lesser extent resulting in a relatively small rightward shift of the binding curve with a 6 . 2-fold increase in the mAb926 EC50 ( Fig 2B ) . To distinguish between competitive versus non-competitive interactions of mannose and the antibodies , we compared the observed EC50 ratio values for antibody binding with an EC50 ratio for a model of two ligands binding to a receptor [36] . Based on the model , mannose and antibody can compete for binding to the same site on FimH according to their relative concentrations and affinities , or bind to the receptor simultaneously ( S3 Fig ) with the affinities altered by a cooperativity factor α [37] . The calculated EC50 ratio for competitive binding ( see Material and Methods ) was 175 ± 30 , which is consistent with the EC50 ratio experimentally determined for mAb475 ( 179 ) confirming that the antibody is a direct competitor . Thus , binding of mAb475 and mannose to FimH is not simultaneous but mutually exclusive , implying binding to a structurally identical site , consistent with our previously reported determination of the mAb475 epitope [33] . However , the 6 . 2-fold alteration of mAb926 EC50 by mannose cannot be explained by the competitive inhibition model but instead indicates that mannose inhibits mAb926 binding in a non-competitive manner consistent with simultaneous binding of the antibodies and mannose to FimH . The non-competitive relationship of mAb926 with the mannose ligand indicates that , unlike with mAb475 , inhibition of FimH activity by mAb926 is not via a direct orthosteric mechanism . It rather resembles more the mechanism exerted by allosteric inhibitors that , however , would have to involve structurally distant sites for antibody and ligand binding . We next determined the mAb926 binding epitope in FimH by site-directed mutagenesis ( S2 Table ) and compared it with the locations of the mAb475 epitope and mannose-interacting residues of the adhesin defined according to lectin domain crystalized in the high-affinity ( active ) , mannose-bound conformation ( Fig 3A ) . Alteration of positions 52 , 135–138 and 140 in FimH abrogated mAb926 binding ( Table 1 and S2 Table ) . These positions form a compact epitope located on the top of the lectin domain ( Fig 3A , blue spheres ) , i . e . on the side of the beta-barrel where the mannose-binding pocket is positioned . This epitope location is opposite from the domain-domain interaction interface that comprises the natural allosteric site of the lectin domain ( Fig 3A , grey spheres ) [11] . Three out of 6 of the residues in the mAb926 epitope , I52 , N135 and N136 , are also part of the mAb475 epitope ( which include positions 1 , 46 , 52 , 54 , 133 , 135 , and 136 ) ( Table 1 and Fig 3B ) [33] . The predicted structural overlap of mAb926 and mAb475 epitopes is also supported experimentally by the fact that mAb926 and mAb475 strongly cross-interfere with each other’s binding to FimH ( S4 Fig ) . Epitopes of both antibodies overlap with the mannose-binding pocket ( Table 1 and Fig 3B ) , in particular with the network of 7 side chain residues that form 11 hydrogen bonds with the ligand [38] . However , the mAb475 epitope is positioned on at least three different areas of the pocket ( 46–54 and 133–142 loops and N-terminal end ) and includes a total of 5 of these mannose-interacting residues that form a total of 9 hydrogen bonds with the ligand [33 , 38] . In contrast , almost the entire mAb926 epitope is limited to just one side of the pocket formed by loop 133–142 , with only two residues—135 and 140—forming a total of 3 hydrogen bonds with mannose . Consistent with structural data , mutation of residues that are directly ( N135 and D140 ) - or indirectly ( N138 ) involved in hydrogen bonds with ligand substantially decreased ligand binding [33 , 38 , 39] , while mutation of remaining mAb926 epitope residues had no or only minor effect on the interactions with mannose [33] . Thus , the epitope of non-competitively inhibiting mAb926 overlaps with the mannose-binding pocket of the active FimH but , in contrast to the mAb475 epitope , is mostly limited to just one loop of the pocket . Still , because two residues of the mAb926 epitope contribute to the network of hydrogen bonds with the mannose ligand , it is plausible to expect some inhibitory effect of mAb926 against mannose binding , consistent with results reported above . While the overlap of mAb926 epitope with the mannose-binding residues in the active FimH explains the inhibitory potential of the antibody , it does not explain the non-competitive nature of the mAb926 inhibition . Thus , we turned to the alternative FimH structure ( 3JWN ) , where the lectin domain assumed a more twisted conformation and interacts with the pilin domain ( Fig 3D ) . This structure was obtained in the absence of mannose ligand and its binding pocket is in an open , low-affinity ( i . e . inactive ) conformation . As the inactive FimH structure was obtained in the absence of ligand , we first determined the potential position for mannose in the open configuration of binding pocket . Mannose was docked into the pocket of the 3JWN crystal structure using coordinates present in the active 1UWF structure followed by energy minimization using the CHARMM and the PARAM22 force field . As shown in Table 1 and Fig 3E mannose is predicted to take a position in the open pocket that retains 8 out of 11 hydrogen interactions of the active FimH with side chains of 5 out of 7 mannose-interacting residues—Phe1 , Asn46 , Asp47 , Asp54 and Gln133 [38] . Interestingly , these five residues in the inactive binding pocket retain essentially the same position relative to each other as in the active FimH ( Fig 3C and 3F ) , with the distance shift being 0 . 1 to 1 Å ( 0 . 44 Å ± 0 . 3 on average ) ( S5 Fig ) . This position of mannose is also supported by previous studies that employed molecular dynamics simulations of the active pocket or resolved crystal structure of mutationally inactivated FimH [38 , 40] ( see Discussion ) . According to the predicted position of mannose , two of the residues that interact with mannose in the active structure—Asn135 and Asp140—would lose their contacts with mannose in the open binding pocket ( Fig 3E and Table 1 ) . In the alternative FimH structures , these two residues also shifted significantly relative to one another and the other five mannose-interacting residues ( Fig 3F ) , with the shift being 1 . 2 to 7 . 7 Å ( 3 . 9 ± 2 . 2 Å on average ) ( S5 Fig ) . Thus , based on the projected mannose position in the inactive FimH , the mAb926 epitope residues that form hydrogen bonds with mannose in the active conformation are shifted relatively further away from the ligand in the inactive conformation . Thus , while in the active FimH pocket a portion of the mAb926 epitope is occupied by mannose , in the inactive FimH the entire mAb926 epitope is potentially accessible to the antibody . Because the mannose binding pocket of FimH is allosterically coupled with the rest of the lectin domain , we compared the effects of mAb926 and mAb475 antibodies on the conformation of FimH . For this , we used the mAb21 antibody that recognizes only the active conformation of FimH . The mAb21 epitope ( Fig 3A , magenta spheres ) is located distal to the mannose-binding pocket and close to the interdomain interface , presumed natural allosteric site ( Fig 3A , grey spheres ) [11] . As shown in Fig 4A , binding of mAb475 and mannose to fimbrial FimHwt converts it from the inactive to the active conformation as determined by binding of active conformation-specific mAb21 . In contrast , mAb926 failed to induce such a conversion . We then performed the same experiment but using a FimH mutant variant that has the A188D mutation in the pilin domain that interferes with its interaction with the lectin domain and , in contrast to FimHwt , sustains FimH in active state even in the absence of mannose [13] . Pre-treatment of FimHA188D fimbriae with mAb926 entirely abolished its recognition by mAb21 ( Fig 4B ) , again in sharp contrast with the pretreatment with mAb475 which enhanced subsequent mAb21 binding . Moreover , when FimHA188D was first pre-treated with mAb21 followed by incubation with the inhibitory mAbs , mAb475 stabilized mAb21 binding , while mAb926 resulted in almost complete displacement of the active state-specific antibody from the adhesin ( Fig 4C ) . Thus , not only is mAb926 binding to the FimH pocket unable to induce the shift from the inactive to the active conformation of the lectin domain , but it does the opposite—induces a shift away from the active conformation . We assessed whether soluble mannose and mAb926 could interact with the binding pocket simultaneously as predicted by the non-competitive inhibition model . We hypothesized that if mannose and mAb926 do bind together to FimH , then mannose should be able to bind to , and destabilize , a pre-formed complex of FimH with the antibody , resulting in a higher off-rate of mAb926 . In contrast , this should not occur for mAb475 bound to FimH as the competitive antibody would fully prevent access of mannose to the pocket and the effect of mannose on the mAb475-FimH complex would be insignificant . In other words , the mannose effect after antibody binding will be opposite from the effect before/during the binding ( Fig 2 above ) . Thus , we measured the effect of mannose ligand on the antibody-FimH complexes using surface plasmon resonance . Surface coated with FimHwt fimbriae were first allowed to bind mAb926 or mAb475 and then antibody-FimH complexes were exposed to running buffer with and without a high concentration ( 1% ) of soluble mannose . As shown in Fig 5 , the dissociation rate of mAb926 from FimH was dramatically increased in the presence of mannose . At the same time , mannose had no significant effect on the stability of the complexes between mAb475 and FimH over the observed time period . These results demonstrate that addition of soluble mannose affects the stability of the FimH-mAb926 complex and , thus , the antibody and the ligand must be able to bind to FimH simultaneously , consistent with the non-competitive nature of their interaction . In contrast , there is no such evidence for simultaneous interaction of mannose and mAb475 with FimH consistent with the direct orthosteric inhibitory mechanism of that antibody . Thus , the effect of mannose on the pre-formed antibody-FimH complexes was opposite from the antibodies effect on the complexes formation . Considering that mAb926 was found to stabilize the low-affinity conformation of FimH , we determined whether any antibodies from our original panel [33] have analogous activity . We found that indeed one of the antibodies , mAb824 , can also prevent binding of active state-specific mAb21 to FimHwt in the presence of soluble mannose ( Fig 6A ) , i . e . mAb824 stabilizes the low-affinity state of the adhesin similar to mAb926 . Unlike the latter , however , mAb824 recognized an epitope located away from the mannose-binding pocket ( residues G79 , S80 , Y82 , and P91; S3 Table ) suggesting that the stabilization of the low-affinity conformation of FimH occurs via an allosteric , i . e . away-from-ligand , mechanism . While mannose could not displace mAb824 from FimHwt in SPR experiments ( S6A Fig ) , SPR studies with mAb824 could not be reliably performed due to difficulties in re-generating the antigen surface upon the mAb824 antibody binding . Thus , the stability of mAb926- and mAb824-FimHwt complexes in the presence of soluble mannose was measured by ELISA . Unlike mAb926 , mAb824 is not displaced from FimHwt even at high ( 8% ) concentration of ligand ( Fig 6B ) . Notably , in the absence of mannose , both antibodies were binding to FimH at the same level upon 4 h-long incubation with PBS ( S6B Fig ) . These results suggested that while stabilization of the low-affinity conformation of FimH by antibodies could be achieved via an allosteric mechanism , the parasteric mechanism may provide unique properties of interference between the mannose ligand and mAb926 binding , not provided by an allosteric mechanism . We evaluated the effect of inhibitory antibodies against an E . coli biofilm formed on a mannose-coated surface by the model uropathogenic strain UTI89 expressing type 1 fimbriae . Both mAb475 and mAb926 ( as well as soluble mannose ) prevented formation of E . coli biofilm on a mannan-coated surface when they were added to the bacteria prior to growth over the surface ( P<0 . 005 ) , indicating that biofilm formation is dependent on mannose-specific bacterial adhesion ( Fig 7A ) . However , when we first allowed the surface biofilms to form on a mannan-coated surface overnight and then added the antibodies ( or mannose ) , neither mAb475 nor soluble mannose caused substantial detachment of the surface-attached biofilm ( Fig 7B ) . In contrast , mAb926 resulted in effective dissolution of the biofilm ( 93% biofilm reduction vs 12% for mAb475 , P<0 . 0005 ) . We then compared the two antibodies for their ability to block E . coli infection in vivo . As shown in Fig 7C , incubation of E . coli UTI89 with mAb926 prior to inoculation of mice via bladder catheter blocked bladder colonization more effectively than mAb475 . There was an 83% reduction in bacteria recovered from bladder of mice infected with E . coli UTI89 bacteria that were pre-treated with mAb926 ( Fig 7C ) , while inhibition with mAb475 was 52% ( P = 0 . 0087 ) . Thus , the non-competitively-inhibiting mAb926 is more effective than is the competitively-inhibiting mAb475 in assays that are most physiologically relevant , such as detachment of biofilms and prevention of bladder infections by uropathogenic E . coli .
Creating antibodies targeting ligand-binding-site epitopes of receptor proteins is a primary goal in the development of protective or therapeutic antibodies . These antibodies are expected to block receptor binding functions by directly competing with the ligand . By definition , for competitive inhibition to occur , the binding pocket cannot be occupied by the ligand at the moment of inhibitor binding ( Fig 8A ) . Thus , such orthosteric inhibitors cannot reverse ligand binding by triggering detachment of ligand from the pocket , and are ineffective in the presence of high concentrations of endogenous ligand , which limits their utility . The only inhibitors able to detach bound ligand are thought to be those of an allosteric nature that induce a weakly-binding inactive receptor state by signaling a conformational change from the site that is positioned distal to the ligand-binding pocket ( Fig 8B ) . However , design of allosteric inhibitors is problematic for proteins where conformational regulation is unknown , not existing or complex . We demonstrated here that the ligand-binding site of a receptor protein provides epitopes for powerful inhibitory antibodies that interfere with ligand binding within the pocket ( like orthosteric inhibitors ) but in a non-competitive manner ( like allosteric inhibitors ) , via a mechanism that we refer to as parasteric ( next-to-ligand ) inhibition . Allosteric inhibitors have been described that bind near the pocket [41 , 42 , 43] , however , unlike the parasteric inhibitor , they did not bind the ligand-interacting pocket residues themselves . The term ‘parasteric inhibition’ was suggested previously to highlight at least a theoretical possibility that inhibitor and ligand could bind in close proximity to each other rather than to fully overlapping or distant sites as expected for orthosteric and allosteric inhibitors , respectively [32] . That study was focused on modulation of enzymatic activity of ACP1 by purine modulators , but structural or mechanistic details of the inhibition were not examined . We show here that one of the striking apparent properties of the parasteric antibody is to bind to the binding pocket simultaneously with the ligand and prevent its conversion into the active state ( Fig 8C ) . In this way , the parasteric concept is also distinct from the concept of inverse agonists , like those shown for the human G-protein coupled receptors , which can stabilize the inactive state of the pocket , but do not bind simultaneously with ligand [23 , 44] . Parasteric inhibition is potentially applicable to a wide range of receptors . Conformational dynamics of the binding pocket is considered to be an essential property of all ligand-binding proteins [45 , 46 , 47 , 48] . At least two different conformational states of the binding pocket are proposed to exist for receptor proteins—the active state that binds the ligand strongly and the inactive state that binds the ligand relatively weakly . In many or even the majority of cases , the active state pocket tightens around the ligand relative to the inactive state . For example , the two domains of the maltose binding protein hinge close to increase affinity [25 , 26] , a ‘lid’ over the binding site of adenylate kinase closes to prevent substrate exit [49] , a ‘gate’ of the plant hormone abscisic acid receptor closes around the hormone ligand [20] , and the binding pocket of the beta-adrenergic receptor contracts around catecholamines [23 , 50] . Similarly , the ligand-binding loops of various receptor proteins have been shown to be highly flexible by NMR and FRET analysis [51 , 52] . This suggests that parasteric inhibitors could potentially bind simultaneously with ligand to the more loosely binding open pocket , preventing it from tightening in many different receptor-ligand systems . Notably , many of these receptors undergo only localized conformational dynamics that are not allosteric in nature , so the parasteric mechanism should be applicable to a wider range of receptors than the allosteric mechanism . The wealth of accumulated structural and functional data on different FimH states and the availability of various conformation-specific monoclonal antibodies provided an opportunity to use FimH as a prototype molecule to study dynamics of conformational shifts between active and inactive states and test various types of inhibitors and conformational modulators . To gain insight into the mAb926 inhibitory mechanism , we turned to the crystal structure of inactive FimH with an open mannose-binding pocket , which was obtained without mannose [11] . Our previous studies on locking the inactive conformation have suggested that the open pocket of FimH retains some ability to interact with mannose [11] . Also , studies of others have shown that mutation of one of the mannose-binding residues , Gln133 to Asn , virtually eliminated the binding function of FimH , but mannose still could be co-crystallized with the mutant [38] . In the latter structure , mannose retained the same interactions as in the native active pocket with all binding residues except for the N135 and D140 residues that shifted away from the ligand , supporting our conclusion that the latter residues are on a flexible pocket loop . Further support for our model is provided by previous molecular dynamics simulations that predicted that residues F1 and D54 ( i . e . those not on the flexible loop ) form the strongest interaction with the ligand [40] . Thus , our model pointing to the dynamic nature of the 133–140 loop is consistent with previous studies . It remains unclear whether the binding pocket must undergo a shift from the active to the inactive conformation in order for mAb926 ( or any parasteric inhibitor in general ) to gain access to its epitope when ligand is already bound . In the case of FimH , the 132–140 loop of the pocket that carries mAb926 epitope may transiently shift away from the attached mannose . Indeed , the intrinsic opening rate of maltose-binding protein has been recently shown to determine ligand dissociation [46] . Another possible scenario for mAb926 action could involve its intermediate binding to the active conformation of the mannose-occupied pocket of FimH via a portion of the epitope that does not involve the mannose-binding residues N135 and D140 . In turn , the possible intermediate step of binding could facilitate the full opening of the flexible loop , thus de-activating the active state . Indeed , mutation of mAb926 epitope residues that are not involved in hydrogen bonding with ligand and facing outside the pocket cleft ( I52 , N136 and Y137 ) did not decrease or had only minimal effect on mannose ligand binding [33] suggesting that they could be accessible for interactions with the antibody even if the pocket is occupied by the ligand . Further studies are needed to elucidate the structural details of the inhibitory action of mAb926 and that of other potential parasteric inhibitors . Although interaction of the lectin domain with the pilin domain in FimH allosterically facilitates the inactive conformation of the binding pocket , soluble pilin domain failed to stabilize the low affinity state of the adhesin [12] . In this study , however , we found an antibody , mAb824 that appears to prevent shifting of FimHwt from the low- to high-affinity state in an apparently allosteric manner by binding an epitope located away from the binding pocket . This epitope is positioned on a side of the split sheet of the lectin domain of FimH that may be a critical region in the conformational pathway of the switch , but details of mAb824 action require further investigation . Importantly , we observed that , unlike mAb926 , mAb824 is not displaced from FimH by soluble mannose suggesting that the ligand-induced conformational shift in FimH cannot overcome the conformation-stabilizing effect of mAb824 . It is plausible to expect that at least in some receptor proteins , allosteric antibodies that are bound to the inactive conformation might be displaced by the ligand-induced activation of the protein and conversely that ligand bound to the active conformation might be displaced by allosteric inhibitory antibodies . However , allosteric antibodies are fundamentally different from parasteric antibodies in a way that is likely to affect the issue of induced dissociation . Allosteric antibodies bind to sites that are distinct from , and relatively weakly coupled to , the binding site [2] . This means that the antibody epitope can be in the inactive conformation while the binding site is in the active conformation . In contrast , a parasteric antibody by our definition has an epitope that to some extent overlaps with the binding site , so that binding kinetics must be changed if the antibody binds , and vice versa . We hypostatize that this difference explains the lack of mAb824 displacement by mannose from FimH but more extensive studies of mAb824 are needed to address this question . Thus , although mAb926 binding to FimH has an allosteric effect on the lectin domain , its inhibition of mannose-binding is not via the allosteric mechanism per se and is fundamentally distinct from the action exerted by mAb824 or other classic allosteric inhibitors described for other receptor-ligand interaction systems [42 , 43 , 53 , 54] . The model of non-competitive inhibition by mAb926 implies that the antibody and the ligand can bind simultaneously to FimH , i . e . form at least a transient FimH-mannose-mAb926 complex . This is experimentally supported by the observation that mannose accesses the FimH pocket occupied by mA926 but not by mAb475 , resulting in unbinding of the former . The exact mechanism of how the three-way complex forms and why it is unstable is unclear , but is likely that antibody is eluted from FimH due to a structural distortion or steric hindrance caused by ligand binding . Reciprocally then , the co-binding property of mAb926 would lead to ligand unbinding when the receptor is already occupied by the ligand . Indeed , we observed that bacterial biofilm formed on a mannose-coated surface can be effectively detached only by mAb926 compared with the relatively minor effect of mAb475 antibody or even high concentrations of soluble mannose . To our knowledge , mAb926 is the only antibody shown to dissolve a bacterial biofilm . The phenomenon of apparent simultaneous binding of the inhibitory mAb926 and the mannose ligand in spite of overlapping binding sites , makes it distinct from both the orthosteric and allosteric inhibitors , providing the rationale for defining a novel , parasteric mechanism of ligand binding inhibition . The advantage of parasteric inhibitors vs orthosteric inhibitors is that the former would be more potent in unbinding the ligand from the binding pocket and more effective in the presence of high concentrations of endogenous ligands . Either effect can explain the significantly stronger inhibition by mAb926 relative to the competitive inhibitor mAb475 , and the superior ability of mAb926 to block FimH-mediated mouse bladder colonization . Although we demonstrated that the ligand can enter the binding pocket with bound mAb926 and in fact displace the antibody , this only occurred at very high , non-physiologic concentrations , and would not compromise the effectiveness of the antibody as a ligand-binding inhibitor . Many bacterial adhesins mediate shear-enhanced adhesion similar to FimH [55 , 56 , 57 , 58] , suggesting that they may also undergo conformational changes and be candidates for parasteric inhibition . The advantage of parasteric inhibitors vs allosteric inhibitors is that the effectiveness of the parasteric inhibition is not limited by weak coupling of the allosteric site to the ligand binding site , and that parasteric inhibition does not require the receptor to be allosteric . In comparison to allosteric inhibitors which present difficulties for rational design due to lack of knowledge of the location of allosteric sites , development of parasteric inhibitiors may be more universally applicable to any protein with defined ligand binding sites , whether or not the protein is allosteric . Taken together , our findings suggest that antibodies binding to just one loop of the ligand-binding site have the potential to be very effective for both inhibition and reversal of bacterial adhesion via the novel parasteric mechanism . The binding pocket side loops of bacterial adhesins appear to be good targets for generation of antibodies with such potential . The observation that the mAb926 epitope is largely formed by a single loop makes this loop a good candidate immunogen for induction of parasteric antibodies using synthetic cyclic peptides . Loop-shaped epitopes have been shown to be extremely potent as synthetic peptide—based vaccines due to their structural properties when synthesized in circular form that not only well mimic the native epitopes but also exhibit extremely high stability as immunogenic agents [59 , 60] .
All mouse work in this study was carried out in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol no . 120304–03 was reviewed and approved by the New York University Institutional Animal Care and Use Committee . The Escherichia coli clinical isolate UTI89 and recombinant strains of E . coli K12 expressing type 1 fimbriae with different structural FimH variants were previously described [12 , 13 , 61] . Briefly , the recombinant strain of E . coli K12 ( AAEC191A ) carries pPKL114 plasmid containing the entire fim gene cluster from the E . coli strain K12 , but with the inactivated fimH gene . For type 1 fimbriae expression , pPKL114 plasmid harboring bacteria were transformed with isogenic pGB2-24-based plasmids carrying different alleles of the fimH gene . Hybridoma cell cultures producing mice anti-FimH monoclonal antibodies were described earlier [13 , 33] . Antibodies , mAb475 , mAb926 , mAb824 and mAb21 were purified from hybridoma culture supernatants using protein G-agarose ( Millipore ) according to manufacture recommendations . Germline origin of anti-FimH antibodies was determined based on the V-region sequence of their light chains [33] using IMGT/V-Quest software ( http://www . imgt . org/IMGT_vquest/vquest ? livret=0&Option=mouseIg ) [34 , 35] . Mapping of mAb926 and mAb824 epitopes was performed as described previously [33] . Briefly , the antibodies were tested for the ability to recognize purified isogenic fimbriae carrying different mutations in LD of FimH . Parental ( not mutated ) fimbriae were used as a reference against which binding of the antibodies to all other mutant fimbriae was normalized . Epitopes of mAb926 and mAb824 were mapped using the high affinity FimH variant ( FimHwt: ( 186–201 ) FocH , [12] and the low affinity FimH variant ( FimHwt ) , respectively . Microtiter plate wells were coated with purified fimbriae [33] at concentration 0 . 1 mg/ml in 0 . 02 M NaHCO3 buffer for 1 h at 37°C . The wells were washed with PBS and quenched for 20 min with 0 . 2% BSA in PBS . To test the effect of mannose on monoclonal antibody binding , immobilized fimbriae were incubated with serial dilutions of pure mAbs in the absence or presence of 52 mM α-methyl-D-mannopyranoside ( αmm , hereinafter also termed ‘mannose’ ) . Bound antibodies were detected with a 1:5 , 000 diluted HRP-conjugated goat anti-mouse antibody ( Bio-Rad ) . The reaction was developed using 3 , 3′ , 5 , 5′-tetramethylbenzidine ( TMB , KPL ) , and absorbance was read at 650 nm . EC50 values were determined by non-linear regression curve fitting using Prism 6 . 0 software ( GraphPad , La Jolla , CA ) for each antibody independently . To test the effect of antibodies on the adhesin conformation , immobilized fimbriae were incubated with 50 μg/ml pure antibodies , or 52 mM mannose for 1 h and then 0 . 5 μg/ml biotinylated mAb21 was added to wells . After washing , binding of biotinylated mAbs was detected using a 1:5 , 000 diluted HRP-conjugated streptavidin ( Sigma-Aldrich ) . In some experiments , surface-immobilized fimbriae were first incubated with 0 . 5 μg/ml biotinylated mAbs ( in the absence or presence of 52 mM mannose ) followed by incubation with purified mAbs 50 μg/ml for 1 h . To test the effect of ligand on the stability of FimH-antibody complexes , antibodies at concentration 0 . 4 μg/ml were first bound to surface-immobilized fimbriae , followed by incubation with 8% mannose , or PBS for 1–4 h time . SPR analyses of mAb926 , mAb475 and mAb824 binding to FimHwt followed by the absence or presence of 1% ( w/v ) mannose were conducted at 25°C in a running buffer ( RB ) of HBS-EP+ ( 0 . 01 M Hepes pH 7 . 4 , 0 . 15 M NaCl , 3 mM EDTA , 0 . 05% ( v/v ) Surfactant P20 ) with 0 . 1 mg/mL BSA on a Biacore T100 system ( GE Healthcare ) . Using standard amine coupling chemistry , ~2000 RUs of FimHwt fimbriae were amine-coupled at 20 μg/mL in 10 mM glycine , pH 2 . 5 to 2 flow cells of a Series S CM5 chip ( GE Healthcare ) . Two reference surfaces were prepared by activating and deactivating flow cells without the addition of protein . Duplicate ( single for mAb824 ) samples at a single concentration were injected at a flow rate of 10 μL/min using a “dual” injection command in the T100 control software ( v2 . 0 . 4 ) with injection 1 at 5 mins , injection 2 at 10 mins and a final dissociation time of 1 min . MAb alone curves were generated by injecting mAb followed by an injection of RB and double referenced [62] by subtracting a dual injection of RB followed by RB . mAb + mannose curves were generated by injecting mAb ( in RB without mannose ) followed by an injection of RB with 1% mannose and double referenced by subtracting a dual injection of RB followed by RB with 1% mannose . Optimal regeneration was achieved by injection of either one ( for mAb926 ) or 2 ( for mAb475 ) 30 second pulses of 10 mM glycine , pH 1 . 5 at a flow rate of 50 μL/min followed by a 2 min buffer stabilization phase . Optimal regeneration conditions for mAb824 sample binding were not found , and so required the generation of two FimHwt fimbriae surfaces with only a single mAb824 injection on each . MAb926 and mAb475 injections with and without mannose were run on each FimHwt fimbriae surface prior to the mAb824 injection in order to match mAb824 binding surfaces as closely as possible , as well as to provide a control for comparison . Sensorgrams were double-referenced in Scrubber 2 . 0b software ( BioLogic Software ) , saved as text files , and re-plotted in Prism GraphPad 6 software . To determine apparent kinetic rate and equilibrium binding constants , FimHwt fimbriae were amine-coupled as noted above to a density of 1300 RUs , with an activated/deactivated surface used as reference . Serial 2-fold dilutions of analyte starting at 12 . 5 nM ( mAb926 ) or 200 nM ( mAb475 ) , and buffer blanks were injected in random order and run in duplicate in HBS-EP+ with 0 . 1 mg/mL BSA at a flow rate of 30 μL/minute with 700 s of association and 1200 s of dissociation . Surfaces were regenerated with either one 30 s injection ( mAb926 ) or two 30s injections ( mAb475 ) of 10 mM glycine , pH 1 . 5 at 50 μL/minute followed by 2 mins of buffer stabilization . Double-referenced data were fit with a 1:1 binding model with BIAevaluation 2 . 0 . 4 software ( GE Healthcare ) . Microtiter 96-well plates were coated with 20 μg/ml of yeast mannan ( Sigma-Aldrich ) in 0 . 02 M NaHCO3 buffer at pH 9 . 6 . The wells were quenched with 0 . 2% bovine serum albumin ( BSA , Sigma-Aldrich ) in PBS for 20 min . Bacteria expressing FimHwt ( OD = 1 ) were first preincubated with different concentrations of mAbs for 1 h at 37°C and then allowed to adhere to mannan-coated surface for another 1 h . After an extensive washing with PBS , plates were dried and bound bacteria were stained with 0 . 1% crystal violet ( Becton Dickinson ) for 20 min at room temperature ( RT ) . The wells were washed with water and 50% ethanol was added to the wells . The absorbance was measured at 600 nm . Microtiter 96-well plates were coated with 20 μg/mL of yeast mannan in 0 . 02 M NaHCO3 buffer at pH 9 . 6 . Bacterial strains grown overnight in 3 ml LB media were spun and washed 1x with minimal essential media ( MEM , Difco ) . Bacterial suspensions in MEM , at final concentration OD = 0 . 2 , were added to mannan-coated wells in the absence and presence of 52 mM mannose or 50 μg/ml mAbs and incubated 16 h at RT without shaking . After washing with PBS , formed biofilms were stained with 0 . 1% ( v/v ) crystal violet ( Becton Dickinson ) as described above for bacterial adhesion assay . For biofilm detachment , 16 h-old biofilm produced by E . coli UTI89 on mannan-coated microtiter plates was washed 3 times with PBS and incubated in the absence or presence of 52 mM mannose or 50 μg/ml mAbs at RT with mild shaking . The wells were washed 3 times with PBS and stain for biofilm detection as described above . Infection of 10- to 11-week-old C57BL/6 female mice was performed as described elsewhere [33] . Briefly , bacteria were grown in LB medium without shaking for 48 h , harvested , washed twice in PBS and resuspended in PBS at a final concentration of 108 CFU per ml . Bacteria were pretreated with 500 μg/ml mAbs for 1 h at 37°C prior to inoculation . Mice were anesthetized with ketamine/xylazine and twenty-five microliters of mAb-pretreated bacteria in PBS were inoculated transurethrally into mouse bladders via catheter . After 24 h , mice were sacrificed and bladders were aseptically removed and homogenized in 1 mL PBS . Serial dilutions were plated and total bacterial load per bladder was calculated . Statistical significance was determined using two-tailed Mann-Whitney test ( GraphPad Prism 6 . 0 software , La Jolla , Ca ) . To dock α-D-mannose to the binding site of the inactive conformation of FimH , the crystal structure of the lectin domain ( residues 1–158 , with PDB code 3JWN ) was aligned onto the high affinity structure of FimH lectin domain ( PDB code 1UWF ) followed by minimization of the RMSD of residues 1 to 6 and 44 to 48 . The coordinates of mannose which are present in 1UWF structure were then used to create a complex between LD in low affinity conformation and the ligand . The entire system was then subjected to 100 steps of steepest descent minimization in vacuo and 500 steps of conjugate gradient minimization in a dielectric continuum using the program CHARMM . 13 [63] and PARAM22 force field [64] . Based on published structural data , [38 , 65 , 66] , the mannose-ring retains the same position and makes the same network of hydrogen bonds in the pocket , regardless the nature of mannosylated ligand ( i . e α-D mannose , alkyl-derivatives of the mannose or oligomannose substrate ) . Thus , for simplicity , only the mannose-ring of α-D mannose was modeled and is presented in FimH structures . In the 1UWF structure , α-D mannose was modeled in by alignment with the sugar ring of the mannose residue of the original crystal structure . The spatial distribution of amino acid residues involved in mAb epitopes and distances between atoms forming hydrogen bonds and mannose ligand in 3JWN and 1UWF structures were measured using the molecular visualization software program PyMOL ( DeLano Scientific LLC ) . All values , unless otherwise indicated , are expressed as mean and SEM . Statistical significance was determined by two-tailed student test using Prism 6 . 0 software ( GraphPad , La Jolla , Ca ) . The receptor occupancy by antibody in the presence of mannose was calculated from the equation: [37] EC50ratio=1+[M]KD1+α[M]KD where , α is a cooperativity factor for mannose and antibody , [M] is mannose concentration and KD is its equilibrium constant . For the strongest possible negative cooperativity ( if mannose and antibody are direct competitors ) α = 0 and EC50 ratio=1+[M]KD . As the dissociation constant for mannose and fimbrial FimHwt is 298 ± 50 μM [13] , and antibody binding was tested at a 52 mM concentration of mannose , the expected EC50 ratio for competitive binding is 175 ± 30 . | A common approach in the development of selective inhibitors for ligand-receptor interactions is targeting the receptor binding site with the expectation that inhibitors will sterically interfere with ligand binding and thus block receptor function via a competitive ( orthosteric ) mechanism . However , using monoclonal antibodies specific for the mannose-binding Escherichia coli adhesin , FimH , we demonstrate that the binding site epitopes allow for non-competitive inhibition that is more effective than orthosteric blocking . FimH , similar to other binding proteins , exhibits conformational flexibility of the ligand-binding pocket shifting between open ( inactive ) and tight ( active ) conformations , with relatively low- and high- affinity towards mannose . We show that an antibody that binds just one of the mannose-binding pocket loops prevents the shift from the inactive to the active conformation and hence blocks formation of high-affinity ligand-receptor complexes . This antibody type was more effective in inhibition of bacterial adhesion than anti-FimH antibodies competitively blocking mannose binding , and unlike the latter or a soluble ligand , showed the ability to detach an established bacterial biofilm from a ligand-coated surface . As the newly described antibody can bind the FimH pocket simultaneously with ligand , we refer to it as a parasteric ( next-to-ligand ) inhibitor that exhibits non-competitive inhibition from within the binding-pocket of the receptor . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Inhibition and Reversal of Microbial Attachment by an Antibody with Parasteric Activity against the FimH Adhesin of Uropathogenic E. coli |
Large experimental efforts are characterizing the regulatory genome , yet we are still missing a systematic definition of functional and silent genetic variants in non-coding regions . Here , we integrated DNaseI footprinting data with sequence-based transcription factor ( TF ) motif models to predict the impact of a genetic variant on TF binding across 153 tissues and 1 , 372 TF motifs . Each annotation we derived is specific for a cell-type condition or assay and is locally motif-driven . We found 5 . 8 million genetic variants in footprints , 66% of which are predicted by our model to affect TF binding . Comprehensive examination using allele-specific hypersensitivity ( ASH ) reveals that only the latter group consistently shows evidence for ASH ( 3 , 217 SNPs at 20% FDR ) , suggesting that most ( 97% ) genetic variants in footprinted regulatory regions are indeed silent . Combining this information with GWAS data reveals that our annotation helps in computationally fine-mapping 86 SNPs in GWAS hit regions with at least a 2-fold increase in the posterior odds of picking the causal SNP . The rich meta information provided by the tissue-specificity and the identity of the putative TF binding site being affected also helps in identifying the underlying mechanism supporting the association . As an example , the enrichment for LDL level-associated SNPs is 9 . 1-fold higher among SNPs predicted to affect HNF4 binding sites than in a background model already including tissue-specific annotation .
Despite large ongoing efforts to characterize regulatory regions in the human genome ( e . g . , ENCODE [1] , Roadmap Epigenomics [2] ) , the lack of a regulatory genetic code to discriminate functional from silent non-coding variants in regulatory sequences poses severe limitations in interpreting the results of many human and population genetic analyses . For example , large numbers of genetic variants associated with disease and normal trait variation have been identified through genome-wide association studies ( GWAS ) [3]; yet a formidable challenge remains in determining the specific molecular mechanisms underlying association signals in non-coding regions . Similar challenges also arise when exploring the evolutionary functional significance of non-coding variants , for example through analysis of differences in genotype distribution across populations [4 , 5] . This is also complicated by the fact that GWAS hits and signals of selection are usually found in large regions of association and do not directly pinpoint the true causative variants . In general , we do not know in which cell types/tissues these variants may have a functional impact . Computationally and experimentally derived annotations for regulatory regions have been used to functionally characterize GWAS hits [1 , 6–12] . However , a simple positional overlap between a genetic variant and regulatory regions is a necessary but not a sufficient condition to demonstrate an impact on TF binding . Many experimentally derived annotations are very useful to identify broad genomic regions across many cell-types , but lack the resolution necessary to pinpoint the regulatory sequences . High resolution functional assays like DNase-seq and ATAC-seq combined with computational methods that integrate sequence motif models [8 , 9 , 13 , 14] can effectively dissect the regulatory elements; yet the motif models for transcription factor ( TF ) binding are generally not sufficiently well calibrated to predict the binding impact of a sequence change . Alternative ChIP based approaches ( such as ChIP-seq and ChIP-exo ) , may provide increased TF and regulatory element specificity , but rely upon the availability of antibodies to target specific TFs or tagged TFs [15 , 16] . The consequence is that we cannot provide a satisfactory answer to the following questions: Which genetic variants are more likely to impact binding of specific TFs ? What is the fraction of genetic variants in regulatory regions that are not neutral ? If we can adequately answer these questions , we may further ask: Did polygenic adaptation occur at binding sites for the same TF ? Do variants in certain types of TF footprints and tissues contribute to variation in specific complex traits ? To help answer these questions , we have extended the CENTIPEDE approach to generate a catalog of regulatory sites and binding variants encompassing more than 600 experimental samples from the ENCODE and Roadmap Epigenomics projects with DNase-seq data , and recalibrated sequence motif models for more than 800 TFs . We then incorporated ASH information to provide additional empirical evidence , to validate the accuracy of the computational predictions and to estimate the fraction of genetic variants in regulatory regions that are not neutral . Importantly , our annotation is specific at the motif level ( i . e . , TF-specific ) and at the sample level ( i . e . , tissue-specific ) . We then compare our results with the only alternative TF-centric annotation that has been recently published [17] , but we also compare with non TF-centric SVM derived annotations [18] . Using our new catalog , we then examined genomic properties of the annotations , identifying characteristics that predict variants that disrupt binding , and demonstrated the action of natural selection on TF binding sites . Finally , we annotated and interpreted variants associated with complex traits , and we validated their allele-specific enhancer activity by reporter gene assays .
The CENTIPEDE approach allows to predict TF activity by integrating sequence motif models together with functional genomics data , and gains the most information from high-resolution data such as DNase-seq or ATAC-seq [19] . The spatial pattern in which reads are distributed , or footprint , is specific for each TF and can be very useful for discriminating between classes of TFs with distinct profiles [13] . In the original CENTIPEDE approach , the sequence models are pre-determined; e . g , k-mers or previously defined position weight matrix ( PWM ) models . However , many sequence models in existing databases were created with very few instances of known TF binding sites and do not represent the full spectrum of sequence variation that can be tolerated without affecting binding . Here , we have extended CENTIPEDE to readjust the sequence models for TF binding ( Fig 1 and S1 Fig ) using DNase-seq data and sequence orthologs ( Methods ) . Compared to the original motif models the consensus sequence is largely maintained in the recalibrated motifs ( S6 Fig ) . However , when we consider ChIP-seq peaks as validation we obtain superior precision recall characteristics ( S7 Fig , Section 6 . 1 in S1 Text ) and a much higher correlation with the prior probability of binding calculated by CENTIPEDE ( S8 Fig , Section 6 . 2 in S1 Text ) . Across all 653 DNase-seq samples , we identified a total of 6 , 993 , 953 non-overlapping footprints corresponding to 1 , 372 motifs active in at least one tissue and collectively spanning 4 . 15% of the genome . Each individual sample contained , on average , 280 , 000 non-overlapping footprints for 600 motifs and spanning 0 . 162% of the genome , indicating that footprints are highly tissue specific . Considering all SNPs from 1000 Genomes Project ( 1KG ) at any allele frequency ( even singletons ) , we found 5 , 810 , 227 ( 0 . 19% of the genome ) unique genetic variants in active footprints ( footprint-SNPs ) , 3 , 831 , 862 ( 66% ) of which are predicted to alter the prior odds of binding ≥20-fold ( effect-SNPs ) based on the logistic sequence model hyperprior in the CENTIPEDE model ( Fig 1C and 1D , Equation 2 in S1 Text ) . Effect-SNPs are further classified as switch-SNPs ( 264 , 965 ) if the allele flips the prior odds of binding . Importantly , in any of these categories we retain for each prediction the motif identity ( TF-specific ) and the underlying sample ( cell-type specific ) information . These functional categories we computationally defined provide an answer to the question of which genetic variants in DNaseI sensitive regions are more likely to affect binding . To experimentally assess the accuracy of our answer , we used Quantitative Allele-Specific Analysis of Reads ( QuASAR ) [20] to perform joint genotyping and ASH analysis within DNase I hypersensitivity ( DHS ) regions ( S2 Fig ) . While the initial quality filtering is the same as for the CENTIPEDE analysis , the parameters of the QuASAR model also allowed us to detect tissues with chromosomal abnormalities or samples from pooled individuals ( Section 4 . 2 in S1 Text ) . These DNase-I samples were therefore excluded from ASH analysis ( S9 and S10 Figs , S6 Table ) . Across the remaining 316 samples suitable for ASH analysis , we identified 204 , 757 heterozygous SNPs ( hSNPs ) in DHS sites ( DHS-hSNPs ) with coverage > 10x and with MAF > 0 . 05 . Overlapping our predictions with the DHS-hSNPs , 55 , 044 are footprint-hSNPs , 26 , 773 of these are effect-hSNPs , and 5 , 991 of these are switch-hSNPs . Overall , our computational predictions are highly concordant with the direction of ASH; 75% of the sequence models show positive correlation between the predicted and observed ASH ( S11 Fig , S7 Table , Section 5 . 4 in S1 Text ) . Each of the nested SNP functional categories have marked differences in p-value distribution ( Fig 2A ) for the QuASAR test of ASH . Compared to what would be expected from the null uniform distribution , effect-hSNPs and switch-hSNPs have 8x and 14x times more SNPs with p < 0 . 001 respectively , showing that our functional annotations can predict ASH . Furthermore , these enrichments for lower p-values are much higher than those of DHS-hSNPs ( 4x ) and footprint-hSNPs ( 6x ) , indicating that identifying SNPs in DHS regions and/or footprints alone is not enough to predict functional effects on binding . A similar observation can be made using the observed allelic ratios across CENTIPEDE annotations ( S12 Fig ) . The result that SNPs that are just located in footprints or DHS regions tend to be silent is also true for other existing annotations ( S13 Fig ) or if we change the threshold for discriminating between footprints-SNPs and effect-SNP ( S14 Fig ) . We also see that conservation score alone is not accurate enough to predict which SNPs have a functional impact on binding ( S16 Fig ) . To quantify the fraction of genetic variants that in each annotation will truly affect TF binding , we used ASH p-values as input evidence and followed the strategy of Benjamini et al . [21] to perform multiple testing correction in each category separately using Storey’s q-value procedure [22] . At an FDR threshold of 20% , we detected 3 , 217 unique hSNPs displaying significant ASH ( Table 1 ) , hereafter referred to as ASH-hSNPs . Taking into account LD ( R2 < 0 . 8 ) these ASH-hSNPs constitute at least 3 , 158 independent loci . Several of the ASH-hSNPs were significant in more than one cell-type , giving a total of 4 , 940 observations of ASH-hSNPs across all samples . The 20% FDR threshold was chosen because this data was not originally sequenced to the depth that is generally required to call ASH at a single site with high confidence . In this reanalysis , we instead focus on the aggregate distribution of p-values to estimate the proportion of true null hypotheses ( Storey’s procedure π ^ 0 estimate ) . We estimate that 56% of the effect-SNPs show evidence of ASH . While this conservative estimate can be considered a lower bound , it is still much higher than the estimates for DHS-SNPs ( 2 . 1% ) and footprint-SNPs ( 3 . 1% ) , indicating that most SNPs in DHS regions and even in the putative binding sites do not affect binding . In addition to the DNase-seq ASH validation , we compared our annotations to the results of QTL analyses targeting DNase-seq sensitivity sites ( dsQTLs , [23] ) , and CTCF binding sites from ChIP-seq [24] . For dsQTLs , using the same PROC analysis ( see Fig 2B ) as in [18] demonstrates that effect-SNPs have a good performance compared to SNPs identified using a SVM approach or CATO [17] . Note that we have not repeated the PROC analysis for the methods studied by [18] , but we used directly the results provided by them , as PROC analysis could be sensitive to a redefinition of the underlying true labels of the set used to evaluate performance ( see discussion in Section 7 in S1 Text ) . If we constrain the gk-SVM model to those predictions that overlap with our CENTIPEDE footprints , the precision ( at 10% recall ) improves to 80% . This indicates that SVMs are better sequence models than PWMs , but are not as specific without footprint information . To further investigate the TF-specificity accuracy of our annotations we used CTCF QTLs . CTCF is a very special type of TF with insulation [25] , DNA loop organization [26] , and barrier functions [27] . Compared to training an SVM on the DNase-seq data-set ( non TF-centric ) , models that are TF-centric such as CATO and our effect-SNPs ( integrating the footprint and sequence preferences ) demonstrate a superior accuracy in discriminating dsQTLs that are also CTCF QTLs from those that may affect other factors ( see Fig 2C ) . Among all CTCF footprint-SNP instances , all those that are also effect-SNPs are enriched for low CTCF QTL p-values and we predicted the correct direction ( the allele with higher binding ) in 100% of the cases ( Fig 2D , Section 3 . 3 in S1 Text ) . Some of the alternative methods include information such as conservation , distance to the TSS and allele frequency , however we have not included them in our annotation as we wanted to use those measures for analyzing the potential impact on organismal function and study differences among distinct TF motifs . Regions of the genome with demonstrated molecular function ( e . g . genic regions ) generally show reduced diversity [28] and a site frequency spectrum skewed towards rare variants . This is due to negative ( purifying ) selection , which prevents alleles from reaching high frequencies in the population if the molecular trait translates to a negative impact on organismal function . We investigated whether a similar skew in the site frequency spectrum exists at functional non-coding variants ( effect-SNPs ) . We observed that effect-SNPs display an enrichment for rare variants ( <0 . 5% ) comparable to what it is observed in coding regions ( Fig 3A ) , where rare variants are 1 to 2 times more likely to be non-synonymous changes than synonymous [29] . eQTL studies have found that variants associated with gene expression tend to occur close to the transcription start site ( TSS ) [30–33] . We detect a similar trend among our annotations , with 83% of footprint-SNPs occurring within 100kb of the TSS . However , we find a 1 . 12-fold depletion of effect-SNPs within 300 bases of a TSS ( Fig 3B ) , which represents the core promoter region [34] . Effect-SNPs in this region are also enriched among rare variants ( MAF < 0 . 001 , 1 . 15-fold enrichment , Fisher’s test p-value = 6 . 027 × 10−13 ) . This is likely because effect-SNPs in these regions have a major impact on regulatory processes that are shared across tissues . Accordingly , we also discovered a 1 . 18-fold enrichment for effect-SNPs in footprints active in 5 or fewer samples and a 1 . 38-fold depletion for effect-SNPs in footprints active in 50 or more samples ( Fig 3C ) . Since allele frequency can be correlated with distance to the TSS or sequence conservation , and shared footprints may also be more common at the promoter region , we tested several features ( individually explored in Fig 3 ) in a joint model ( Methods ) . All tested factors are significant predictors when considered together in a multiple regression logistic model , and the direction of the effect is the same as when they are considered separately ( S8 Table ) . These results support the hypothesis that factors binding closer to the TSS and/or active in many tissues are housekeeping factors and those that recruit the transcriptional machinery and as a consequence are less likely to harbor common regulatory variants . To examine the distribution of ASH-hSNPs across the different regulatory factors , we calculated the ASH enrichment ratio for each TF defined as the fraction of ASH-hSNPs over hSNPs relative to the average fraction across all TF ( S17 Fig , Section 8 . 3 in S1 Text ) . At a nominal p-value cutoff of p < 0 . 01 ( Binomial test ) , we detected 32 motifs enriched for ASH and 56 depleted for ASH ( Fig 4A; S9 Table ) . In cases where multiple motifs correspond to the same factor , we observe similar enrichment for ASH-hSNPs ( S10 Table ) , most notably for the factor AP-1 , showing a >2 . 5-fold enrichment for ASH SNPs in all but one of the seven motif models . We see the same pattern for motifs significantly depleted of ASH-hSNPs , such as CTCF ( 1 . 5-fold median depletion ) and E2F ( 1 . 8-fold median depletion ) . ASH enrichment ratios are also consistent across factors with similar functions . For example , three factors in addition to AP1 with roles in the immune response , CREB [35] , c/EBP [36] , and NF-κB [37] are over 2-fold enriched for ASH-hSNPs within their binding sites ( S11 Table ) . We then examined the genomic characteristics at TF binding sites to identify features that distinguish motifs enriched for ASH versus those that are not . We found that motifs enriched for ASH are significantly farther from the TSS , having an average median distance to the TSS of 23kb compared to 17kb for those depleted ( Mann-Whitney p = 3 . 2 × 10−8; Fig 4B ) . Furthermore , motifs enriched for ASH are active in significantly fewer samples , on average active in 20% vs 40% for those depleted ( Mann-Whitney p = 1 . 9 × 10−7; Fig 4C ) , indicating that TFs with a high degree of ASH across their binding sites tend to be active in fewer tissues . This further confirms that changes in footprints active in a large number of tissues ( constitutionally active ) are more likely to have pleiotropic effects and therefore impact negatively the fitness of the organism and suggests polygenic mechanisms of evolution on motifs categories ( i . e . groups of binding sites for a given TF or for TFs regulating genes with similar functions ) . An important question in evolutionary biology is the extent to which selection has acted on cis-regulatory elements in humans [38–41] . While methods are being developed to address this question [42 , 43] , such methods have only been applied to a narrow subset of TFs , and , in the case of [43] , rely on RNA expression data to classify mutations as up- or downregulating transcription relative to the reference enhancer sequence . Given our categorization of footprint-SNPs relative to their effect on factor binding , we performed an initial survey of selection across TF binding sites using a test similar to the McDonald-Kreitman ( MK ) test [44] ( S3 Fig , Section 8 . 4 in S1 Text ) . Applying our modified motif-wise MK test , we obtained a selection score for TF motifs with a sufficient number of binding sites ( Fig 5A , S12 Table ) . At an FDR of 1% , we observe 84 factors whose binding sites are enriched for fixed functional differences ( higher selection scores ) , suggestive of positive selection acting on those sites . Among the top scoring motifs are several factors that regulate neural and neuro-developmental processes , including POU1F1 , PHOX2B , DBX2 , UNCX , and YY1 which were not previously seen [42] . Among the factors with the lowest selection scores , we find ARNT , RBPJ , CREB1 , POU2F2 , and MYC which match with what has previously been observed [42] . While the interpretation of a positive selection score is generally that of positive selection , interpreting negative scores is more challenging . Generally , deleterious alleles are much less likely to reach fixation in populations than neutral alleles , however a negative selection score could also be explained by relaxation of selection or balancing selection . To identify the most likely evolutionary scenario for variation in binding motifs with negative selection scores , we calculated the derived allele frequency ( DAF ) for SNPs in binding sites . We observed an excess of rare alleles for SNPs in binding sites with a negative selection score ( Fig 5B , S19 Fig , Section 8 . 5 in S1 Text ) , suggestive of weak purifying selection , rather than relaxation of selection ( similar DAF spectrum across categories ) or balancing selection ( excess of intermediate frequency alleles ) . We next asked whether the excess of functional polymorphism relative to functional divergence were influenced by background selection from nearby genes ( S18 Fig ) , as functional regulatory variants may occur closer to the TSS , compared to silent variants . We find a mild but significant positive correlation between selection score and median TSS distance ( Spearman ρ = 0 . 16 , p = 5 . 6 × 10−9 ) . Additionally , there is a negative correlation between tissue specificity and selection score ( Spearman ρ = −0 . 20 , p = 1 . 2 × 10−13 ) . While some of the selection signal may come from nearby genes , there does appear to be a pattern of selective constraint on broadly active factors binding in promoter regions . Given that our annotations comprise predicted functional effects across multiple cell-types/tissues and are anchored at footprints for known TF motifs , we asked if they could help interpret genomic hits reported in the GWAS catalog . We first considered a gross overlapping approach that considers each variant in a GWAS hit region equally likely to be causal ( using an r2 cutoff of 0 . 8 from 1KG Project data , as in Ward et al . [10] ) . In GWAS hit regions , we compared the proportion of effect-SNPs over footprint-SNPs and found a moderate 1 . 11-fold enrichment for effect-SNPs ( p < 2 . 2 × 10−16 , 95% CI: 1 . 10—1 . 14 ) . These moderate but statistically significant enrichments are typical of other annotations as well and are likely due to the fact that: i ) we only consider the strongest GWAS hits ( missing variants with moderate and small effects ) , ii ) not all the factors and tissues may have the same enrichment , and iii ) lack of resolution , as expanding the GWAS hit region makes the enrichment effects more moderate . Nevertheless , if we add our annotation to category 2 SNPs from the RegulomeDB [8] ( SNPs with multiple regulatory annotations , but not yet shown to be functional ) , we detect a 1 . 6-fold enrichment for GWAS hits compared to category 2 SNPs alone ( p = 6 . 11 × 10−5 , 95% CI: 1 . 27—1 . 99 ) . This result demonstrates that our annotation adds relevant information as it filters genetic variants not likely to be functional , but the overlap approach employed cannot take full advantage of the resolution and contextual information provided by our CENTIPEDE predictions . To better test if the annotated effect-SNPs can help fine-mapping and give a mechanistic support for variants associated with complex traits , we integrated them into GWAS meta analyses for 18 traits ( see S13 Table ) using the recently developed hierarchical model fgwas [45] . Importantly , in this analysis we used as input the association p-values measured or imputed to all known common variants in the genome . Furthermore , for each trait we compare to a baseline model [45] that considers previously defined annotations [11 , 46] and confounders ( e . g . , distance to TSS , coding region , and others ) . For each trait , we identified factors whose binding sites were enriched for associated SNPs ( Fig 6A and 6B , S20 Fig and S14 Table ) over the baseline model ( the enrichments reported by fgwas are log-odds ratios from the model parameters ) . Overall , we observed high enrichments for biologically relevant factors . For example , the enrichment for effect-SNPs in OCT-4 ( POU5F1 , a TF with a key role in embryonic development and stem cell pluripotency [47] ) regulatory sequences when considering genetic variants associated with human height is 6 . 6-fold higher ( 95%CI: 3 . 7-8 . 2 ) than in the baseline model . This is consistent with previous observations of genetic variants associated with height being enriched in embryonic stem cell DHS sites [48] . We also observed an enrichment for the developmental regulators TBX15 ( 3 . 9x ) , FOXD3 ( 3 . 9x ) , and NKX2-5 ( 4 . 7x ) for genetic variants associated with height . From a study of low-density lipoprotein ( LDL ) levels in the blood , enriched factors include the liver-specific factor HNF4A ( 9 . 1x ) , as well as several regulators of immune function , including CREB1 ( 3 . 7x ) , IRF1 ( 6 . 2x ) , and IRF2 ( 7 . 1x ) . Our high resolution annotations allowed us to dissect the most likely functional variant ( posterior probability of association , PPA > 0 . 2 ) in 88 previously identified GWAS regions ( S15 Table , S23 Fig ) . For all 88 but 2 of these SNPs we have at least a 2-fold increase on the posterior odds of picking the potentially causal genetic variant according to fgwas ( 8 . 5x median fold increase ) when compared to the comprehensive baseline annotation used by [45] . We then performed reporter gene assays for 21 SNPs to validate the predicted allelic effect on gene expression and the underlying molecular mechanism ( Fig 7A and 7B , S16 Table , Methods ) . Among the regions tested we validated that 11 have enhancer/repressor activity and 10 have variants with allele-specific activity ( p < 0 . 05 , BH-FDR = 10% ) . This corresponds to 48% validation rate which is much greater than the 5% that would be expected by chance ( Binomial test p = 2 . 01 × 10−8 ) . Overall the predicted effect on binding and the change in gene expression are well correlated ( Spearman ρ = 0 . 612 , p-value = 0 . 0032 ) , and the three SNPs with opposite effects may represent binding sites for repressors . Spearman correlation is robust to outliers , removing potential outlier rs540909 results in ρ = 0 . 657 ( p-value = 0 . 002 ) . We also achieve a similar correlation when we use our predictions to evaluate mutations in enhancers from a previously published reporter assay [49] that match our CENTIPEDE footprints ( Spearman ρ = 0 . 76 , p-value = 4 . 37 × 10−5 , S22 Fig , Section 9 . 4 in S1 Text ) . As an example , rs4519508 , associated with a 2 . 1cm decrease in height [50] , is in a binding site for the cell-cycle regulator family E2F ( Fig 6D ) . Our annotation increased the PPA from a baseline of 10 . 5% to 44 . 4% , and it is the highest associated SNP in the association block ( S21A Fig ) . This E2F footprint is active in >300 tissues ( most of them fetal ) and we detected ASH at this SNP in lung fibroblasts , validating that the reference allele at rs4519508 confers stronger binding than the alternate . Interestingly , in the reporter assay we observed 1 . 5-fold increased expression in the presence of the alternate allele , suggesting that at this location , E2F is acting as a repressor . Finally , this SNP is located within the promoter of PPP3R1 , a regulatory subunit of calcineurin important for cardiac and skeletal muscle phenotypes; and a SNP in the same region has been shown to be associated with endurance [51] in humans . The p-value of association for this GWAS locus ( p = 8 . 1 × 10−6 ) does not reach genome-wide significance in the height meta-analysis data we used [50]; however , in a recent more extensive meta-analysis for height [52] this locus achieves genome-wide significance p = 8 . 4 × 10−10 , demonstrating that our annotation can be useful to rescue relevant loci . Finally , a SNP associated with LDL levels , rs532436 , is within a footprint for USF , an E-box motif ( Fig 6C ) . Adding our annotation increased the PPA of the SNP from 39 . 7% to 94 . 7% ( S21B Fig ) . We found that the alternate allele , associated with a 0 . 0785 mg/dL increase of LDL in the blood , is predicted to have a lower binding probability and results in 1 . 8-fold lower expression , compared to the reference allele . This SNP is identified by GTEx [53] as an eQTL for two proximal genes in whole blood: ABO ( p = 5 × 10−5 ) and SLC2A6 ( GLUT6 , a class III glucose transport protein; p = 8 × 10−5 ) . The SNP has an opposite effect on expression of the two genes , with the alternate allele showing lower expression for ABO and higher expression for SLC2A6 . These results show that our integrated analysis provides support for likely mechanisms linking regulatory sequence changes to complex organismal phenotypes . Furthermore , these mechanisms can be directly investigated through molecular studies , providing additional support that these sequence changes are truly functional .
We have developed an approach for assessing functional significance of non-coding genetic variants in DNase-seq footprints . Our strategy integrates sequence information with functional genomics data to predict the impact of single nucleotide changes on tissue-specific TF binding . This is achieved while integrating footprint information that preserves the identity of the underlying factor with high specificity . By borrowing data from ENCODE and Roadmap Epigenomics , we generated one of the most comprehensive catalogs available to date annotating regulatory regions and functional genetic variants across the genome . Thus far , most common approaches for identifying regulatory variants from functional genomics data assume that each SNP in a regulatory region is equally likely to be functional . A key finding in this study is that genetic variants in active regulatory sequences , as defined by DNase I sensitivity and footprinting , are mostly silent; only 2 . 1% of SNPs in DHS regions and 3 . 1% of SNPs in CENTIPEDE footprints are estimated to have ASH . This is analogous to SNPs in coding regions , where most genetic changes are synonymous and do not result in an amino acid change [29] . The sequence model developed in this study provides a very useful filter for non-coding genetic variants that are not functional , resulting in a tissue-specific and motif-specific annotation of effect-SNPs ( 56 . 5% of which are estimated to have an impact on ASH ) . This is crucial information to take into account when we attempt to understand the molecular mechanism behind GWAS hits and evolutionary signals of selection . As additional functional genomics studies are performed , across larger sample sizes , tissue types and cellular conditions , it will be important to further determine the functional subset of regulatory variants within binding sites to achieve greater power in functionally annotating genetic variants associated with complex traits . We find that genetic variants that are predicted to impact TF binding are depleted in the core promoter regions , exhibit higher sequence conservation in closely related species , tend to have low allele frequency and are enriched in tissue-specific footprints . These properties largely reflect the family-wise characteristics of motifs , which are further reflected in signals of selection . Future studies could incorporate tissue breath , conservation and distance to TSS as features to further filter effect-SNPs that may not show ASH . It should also be noted that our definition of functional regulatory variants is connected to the predicted effect on binding in the specific subset of cell-types/conditions that were available . Analyzing the allelic effects of non-coding variants in the context of other tissue types , conditions and functional genomic assays may potentially identify a functional role for some of the sites here defined as silent . In this study , we treated each TF separately , but future work should further explore the combinatorial grammar that different groups of motifs may define by cooperative binding to determine tissue specific binding sites . This will probably require more complex sequence models ( e . g . , SVMs [18 , 54] or deep neural networks [55 , 56] ) than the PWMs used here . Here we show that the footprint information helps in predicting functional variants by further identifying the underlying TF compared to a sequence-fits-all model . More sophisticated footprint models [57] may also offer additional improvements to dissect the complexity of the regulatory grammar . As not all genetic variants that have an impact on binding may lead to changes in gene expression and ultimately an organismal phenotype , combining these predictions with eQTL data across several tissues or environmental conditions would be important to further refine this annotation . As an example , Wen et al . [33] , using an early release of this annotation in lymphoblastoid cell-lines demonstrates that effect-SNPs are 1 . 49 fold ( with 95%CI[1 . 38 , 1 . 63] ) more likely than baseline SNPs ( SNPs that are not located in a footprint ) to be eQTLs ( p = 4 . 93 × 10−22 ) ; in contrast , silent footprint-SNPs are 1 . 15 fold ( with 95%CI[1 . 04 , 1 . 27] ) enriched in eQTLs , comparing to baseline SNPs ( p = 0 . 0035 ) . A key feature of our annotation is that it spans a large collection of tissues and transcription factor motifs . This allowed us to trace some of the evolutionary history of TF binding and identify evolutionary constraints on specific molecular functions , which may reflect selective pressures during human history . For example , we observed that immune TFs are enriched for ASH sites , which supports the hypothesis that this may be a consequence of human adaptations to pathogen exposures [58] . On the other hand , we identified neural development TFs that may have undergone positive selection in humans . The large number of regulatory variants predicted in our study , together with previously reported eQTL signals [59–61] , and the overall relevance that they have in explaining complex traits provide further support for polygenic models of complex traits in humans . By taking advantage of the factor-specific annotations in our study , we identified motifs that are enriched for regulatory variants associated with relevant GWAS traits and we provide examples of molecular mechanisms behind the association signals; e . g . , immune TFs in the lipids study , and developmental TFs for height . Finally , we show how regulatory annotations improve the identification of potential causal SNPs in GWAS . Overall , the GWAS meta-analysis and selection signals in our study support the concept that polygenic variation in binding sites has been a major target of evolutionary forces and a key contributor to disease risk and complex phenotypes in human populations .
We used 1 , 949 PWM sequence models ( motifs ) from the TRANSFAC [62] and JASPAR [63] databases to scan the genome for a set of representative motif matches ( Section 3 . 1 in S1 Text ) . For each motif , we used the matching sequences to calculate a new PWM model which we then used to scan the genome and identify all genome-wide motif matches using a two step approach: Step 1: Initial CENTIPEDE scan and motif recalibration . For each motif , we extracted DNase-seq data at sequence matches across 653 samples ( corresponding to 153 unique tissues ) publicly available from the ENCODE and Roadmap Epigenomics projects ( Sections 1 and 2 . 1 in S1 Text ) . The motifs and samples used are summarized in S1 and S2 Tables . For each motif and only for this initial step , we used a reduced subset of motif matches that include the top 5 , 000 best sequence matches , and up to 10 , 000 additional low-scoring sequences ( Section 3 . 1 in S1 Text , note that for Step 2 we will use all motif matches in the genome ) . To avoid overfitting and to heuristically reduce the search space , these low scoring motif instances are human sequences that have orthologous very high scoring motif instances in the chimp or rhesus genome . We then applied the CENTIPEDE model to survey TF activity for each 1 , 272 , 697 tissue-TF pair . For each pair we then determined that the TF is active if the sequence matches that exhibit a CENTIPEDE footprint can be predicted from the PWM score ( Z-score > 5 , S4 and S5 Figs ) . Using this criterion , we determined that 1 , 891 TF motifs are active in at least one tissue . The full list of motifs active in each tissue can be found in S3 Table . We then recalibrated the PWM model for each active motif using the sequences of all motif matches that have a DNase-seq footprint ( CENTIPEDE posterior > 0 . 99 ) . Step 2: Full genome CENTIPEDE scan and genetic variant analysis . Using the recalibrated sequence models we scanned the human genome again for all possible sequence matches . We used the CENTIPEDE algorithm to assess the probability that each motif instance is bound by a TF , both to the reference and to alternate alleles when the match contained a genetic variant catalogued in the 1KG Project [29] . In this second step , we included all high and low scoring PWM matches down to the threshold corresponding to a CENTIPEDE prior probability of binding of 10% ( Equation 2 and Section 3 . 2 in S1 Text ) . To evaluate whether the updated sequence models derived from DNase-seq data are better at predicting TF binding than the original seed motifs , we compared to ChIP-seq data available for a small set of TFs from the ENCODE project ( as these data are generated in independent experimental assays that should be highly TF-specific ) . Using precision recall operating characteristic ( P-ROC ) curve analysis ( see Section 6 . 1 in S1 Text ) , we determined that for a given precision ( precision = 1—FDR , false discovery rate ) , the updated sequence models have higher recall ( sensitivity ) than the original PWM in detecting ChIP-seq peaks ( S7 Fig ) . Additionally , we compared the correlation between the prior probability of binding ( calculated by CENTIPEDE based on the PWMs ) and the number of ChIP-seq reads overlapping motif matches ( S8 Fig , Section 6 . 2 in S1 Text ) . We classified a SNP in a CENTIPEDE footprint ( footprint-SNP ) as having a predicted effect on binding ( effect-SNP ) if the difference in the prior log odds ratio ( from the logistic sequence model in CENTIPEDE , Equation 2 in S1 Text ) between the two alleles was ≥3 , indicating a ≥20-fold change in the prior odds of TF binding . We further classified an effect-SNP as switching the likelihood of binding ( switch-SNP ) if the prior log odds ratio flips; i . e , if it is ≥0 for one allele and ≤0 for the other . To generate a final set of annotated SNPs , we aggregated the data from each sample and motif into one table . For cases where a SNP is within multiple predicted binding sites , we selected the factor whose CENTIPEDE likelihood ratio was the greatest , i . e . , the factor most likely to be binding at that location . Starting from raw sequencing reads , we used a custom mapper [23] to align the reads to the hg19 reference genome . As allele-specific analysis is extremely sensitive to mapping errors and PCR duplicates , we employed several methods to reduce these sources of potential bias ( Sections 2 . 2—2 . 4 in S1 Text ) . To detect allele-specific hypersensitivity , we applied QuASAR [20] to the processed read data to infer genotypes for all 1KG SNPs and determine the likelihood of allelic imbalance at heterozygous sites . Note that we only test a SNP with QuASAR if it is covered by ≥10 reads . To adjust for multiple testing , we used the q-value method [22] on the p-values produced by QuASAR . We overlapped heterozygous SNPs ( DHS-hSNPs ) identified by QuASAR with CENTIPEDE footprints-SNPs and effect-SNPs catalogued for each sample . SNPs were then partitioned based on their predicted effect on binding into three non-overlapping categories: 1 ) hSNPs in predicted footprints whose binding effect is in the direction predicted , 2 ) all other hSNPs in footprints , 3 ) all other DHS-hSNPs . Because each annotation has a different prior expectation of being functional , we re-adjusted for multiple testing within each annotation separately using the q-value method [22] on p-values produced by the QuASAR model . We denote as ASH-hSNPs those hSNPs with a q-value < 20% in any of the partitions . To determine which features of a SNP are predictors of functional effect , we performed multiple regression analysis using a logistic model considering the dependent binary variable El , indicating whether the footprint-SNP , l , is also an effect-SNP . logit ( E l ) ∼ C l + F l + T l + N l + P l We considered the following variables related to the probability of a footprint-SNP being an effect-SNP: the footprint likelihood ratio ( without the sequence model ) ( Cl ) ; the minor allele frequency ( Fl ) ; the absolute distance to the nearest transcription start site ( Tl ) ; the number of tissues for which the motif containing the footprint-SNP was predicted to be bound ( Nl ) ; the phyloP conservation scores calculated from primates ( Pl ) . This model does not evaluate the sequence , rather it combines the results shown separately in Fig 2 into a single model to characterize the predictions made by CENTIPDE . The model was fit using the GLM function in R . The result of this regression analysis can be seen in S8 Table . To identify divergent TF binding sites , we used the UCSC liftOver tool on binding sites without a known polymorphism to obtain orthologous regions in the chimpanzee genome . Using the PWM model , we calculated PWM scores and CENTIPEDE prior probabilities of binding on the chimpanzee sequences . Sites with a sequence change in the motif instance ( prior probability of binding differs from the humans sites ) were classified as divergent , and were further categorized by the difference in binding affinity: “functional” for sites that change ≥20-fold between species ( analogous to effect-SNPs ) , and “silent” for those that do not . For the binding sites containing a polymorphism , we used the definition of effect-SNPs to identify functional for silent sites and footprint -SNPs for silent sites . For each factor motif , we then calculated the number of binding sites belonging to each of the four categories ( divergent functional , divergent silent , polymorphic functional , and polymorphic silent ) and calculated a selection score similar to the McDonald-Kreitman test ( Section 8 . 4 in S1 Text ) . To integrate functional annotations and GWAS results , we used the fgwas command line tool [45] . fgwas computes association statistics genome wide using all common SNPs from European populations in the 1KG Project , splitting the genome into blocks larger than LD . Summary statistics were imputed with ImpG using Z-scores from meta-analysis data . Using an empirical Bayesian framework implemented in the fgwas software , GWAS data were then combined with functional annotations . We then compared the informativeness of these annotations from each of the 1891 motifs with CENTIPEDE predicted regulatory sites to a baseline model ( see Section 9 . 2 in S1 Text ) consisting of previously used genomic annotations identified as relevant [45] . For each locus that contains at least one SNP with a PPA > 0 . 2 , we only consider the SNP with the highest p-value or PPA from fgwas . Rather than look at a credible set , we pick a single SNP most likely to be causal and see if that SNP has a higher PPA with the annotation than without it . While reduction in size of the credible set is very important for assessing fine-mapping methodologies , here our focus is on combining annotations to identify the single most likely causal SNP per GWAS locus . GWAS-relevant effect-SNPs located in active footprints in LCLs ( the cell line used for transfection ) were ranked on the Spearman correlation coefficient in S7 Table . We initially selected the top 25 SNPs with a positive correlation , but the assays for 4 of them failed for several technical reasons ( e . g . , cloning step failed ) . To test allele-specific effects on expression for the remaining 21 SNPs , we first constructed inserts containing the reference or alternate allele for each SNP of interest ( see Section 9 . 3 in S1 Text ) . Cloning of these inserts in the pGL4 . 23 vector was performed using the Infusion Cloning HD kit ( Clontech ) and DNA was extracted using the PureYield kit ( Promega ) . Transfections were performed into GM18507 using the standard protocol for the Nucleofector electroporation ( Lonza ) . Luciferase activity was measured for up to 20 replicate experiments using the Dual-Glo Luciferase Assay Kit ( Promega ) . We contrasted the activity of each construct to the pGL4 . 23 vector , to assess enhancer/repressor activity of each region . To evaluate allele-specific effects , we contrasted the activity of the reference allele to the alternate allele for each region and we used a t-test to assess significance at a p < 0 . 05 threshold . We used the Benjamini-Hochberg [64] procedure to assess FDR across all 21 SNPs tested . Unless otherwise noted , tests for enrichment on two-way categorical variables are based on Fisher’s exact test . Tests involving multiple categorical , discrete or continuous variables use a logistic regression model and Wald’s test on each enrichment parameter , and are identified as such . The generated annotation files are available as supplementary tables and at http://genome . grid . wayne . edu/centisnps/ . All other relevant data are available in the manuscript and its Supporting Information files . | A large fraction of genetic variants that have been associated with complex traits are found outside of protein coding genes and likely affect gene regulation . Many experimental efforts have been dedicated to mapping regulatory regions in the genome but there are not many systematic methods that integrate functional data and regulatory sequences to predict the potential effect of any genetic variant on any given tissue and motif . Here we present a tissue and factor specific annotation that provides a predicted functional effect for both common and rare genetic variants . These predictions , certain of which are validated experimentally , show that the majority of genetic variants in gene regulatory regions are actually silent . Annotating those that are not silent allows us to investigate the molecular basis for the genetic architecture of many common traits and also to study the evolutionary properties that different types of regulatory sequences have across tissues or transcription factors . Overall , our study supports the concept that polygenic variation in binding sites for distinct classes of transcription factors has been a major target of evolutionary forces contributing to disease risk and complex trait variation in humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"sequencing",
"techniques",
"genome-wide",
"association",
"studies",
"functional",
"genomics",
"gene",
"regulation",
"population",
"genetics",
"dna",
"footprinting",
"genome",
"analysis",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"population",
... | 2016 | Which Genetics Variants in DNase-Seq Footprints Are More Likely to Alter Binding? |
Of the Orthomyxoviridae family of viruses , only influenza A viruses are thought to exist as multiple subtypes and has non-human maintenance hosts . In April 2011 , nasal swabs were collected for virus isolation from pigs exhibiting influenza-like illness . Subsequent electron microscopic , biochemical , and genetic studies identified an orthomyxovirus with seven RNA segments exhibiting approximately 50% overall amino acid identity to human influenza C virus . Based on its genetic organizational similarities to influenza C viruses this virus has been provisionally designated C/Oklahoma/1334/2011 ( C/OK ) . Phylogenetic analysis of the predicted viral proteins found that the divergence between C/OK and human influenza C viruses was similar to that observed between influenza A and B viruses . No cross reactivity was observed between C/OK and human influenza C viruses using hemagglutination inhibition ( HI ) assays . Additionally , screening of pig and human serum samples found that 9 . 5% and 1 . 3% , respectively , of individuals had measurable HI antibody titers to C/OK virus . C/OK virus was able to infect both ferrets and pigs and transmit to naive animals by direct contact . Cell culture studies showed that C/OK virus displayed a broader cellular tropism than a human influenza C virus . The observed difference in cellular tropism was further supported by structural analysis showing that hemagglutinin esterase ( HE ) proteins between two viruses have conserved enzymatic but divergent receptor-binding sites . These results suggest that C/OK virus represents a new subtype of influenza C viruses that currently circulates in pigs that has not been recognized previously . The presence of multiple subtypes of co-circulating influenza C viruses raises the possibility of reassortment and antigenic shift as mechanisms of influenza C virus evolution .
Influenza A , B and C viruses are members of the Orthomyxoviridae family that can cause influenza in humans [1] . Influenza A viruses exist in humans , various other mammal species , and birds; migratory or domestic waterfowl are their largest reservoir . Humans are thought to be the primary hosts and reservoir of influenza B and C viruses , although both have been identified in other hosts after reverse zoonotic transmission from humans . While influenza B virus is a common seasonal human pathogen similar to influenza A virus in its clinical presentation , influenza C virus causes primarily upper respiratory tract infections in children [2] . Clinical manifestations ( cough , fever , and malaise ) are typically mild , but infants are susceptible to serious lower respiratory tract infections [3] . Influenza C viruses co-circulate with influenza A and B viruses and causes local epidemics [4] , [5] . Six genetic and antigenic lineages of influenza C viruses have been described , and as in influenza B viruses , are considered monsubtypic [6] , [7] . Co-circulation of multiple subtypes of influenza allows for rapid viral evolution through the process of antigenic shift , a property previously only shown for influenza A viruses . Thus , both influenza B and C viruses do not have pandemic potential . In contrast , the Influenza A genus includes 17 hemagglutinin and 9 neuraminidase subtypes , and reassortment among different subtypes has repeatedly generated pandemic viruses to which the human population is naïve [8]–[10] . It is the animal reservoirs of diverse influenza A viruses that give them the unique property within orthomyxoviruses of causing human pandemics . Aside from humans , influenza C virus has been isolated only from swine in China ( in 1981 ) [11] . Genetic analysis showed a close relation between Japanese human and Chinese swine influenza C isolates [12] , [13] . Serological surveys in Japan and the United Kingdom found 9 . 9% and 19% of swine , respectively , to have positive HI antibody titers to human influenza C viruses , suggesting that the virus is not uncommon in swine [14] , [15] . Swine inoculated with influenza C virus had mild respiratory disease and transmitted the virus to naive swine by direct contact [11] . Here we characterize an orthomyxovirus isolated from a clinically ill pig and show that the virus is distantly related to human influenza C virus and readily infects and is transmissible in both ferrets and pigs . Genetic and antigenic analysis suggest that this virus represents a new subtype of influenza C virus , raising the possibility of reassortment and antigenic shift as mechanisms for influenza C virus evolution which could pose a potential threat to human health .
In April 2011 , nasal swabs from 15-week old swine exhibiting influenza-like illness were submitted to Newport Laboratories , Worthington , Minnesota , for virus isolation . Real-time reverse transcription PCR ( rt-RT-PCR ) was negative for influenza A virus [16] . In swine testicle ( ST ) cells , the viruses caused influenza-like cytopathic effects ( CPE ) by day 3 . The cell culture harvests were again negative for influenza A virus by rt-RT-PCR . Electron microscopic ( EM ) studies of the cell cultures demonstrated features characteristic of an Orthomyxovirus ( Fig . 1 ) . Negative-staining EM showed enveloped spherical to pleomorphic viral particles approximately 100–120 nm in diameter ( Fig . 1A ) . The virion surface contained dense projections 10–13 nm in length and 4–6 nm in diameter . Thin-section EM studies of infected cells revealed filamentous budding of virions from the plasma membrane ( Fig . 1B ) . These data strongly suggested the virus to be a member of the family Orthomyxoviridae . Enzymatic assays revealed that the virus had negligible neuraminidase but detectable O-acetylesterase activity using 4-nitrophenyl acetate , suggesting it to be a member of the influenza C genus . However , further RT-PCR analysis was negative for influenza B and C viruses [17] . RT-PCR or PCR assays to detect porcine reproductive and respiratory syndrome virus , porcine coronavirus , and porcine circovirus were also negative ( data not shown ) . The virus was purified by ultracentrifugation and sequenced on an Ion Torrent Personal Genome Machine . De novo genome assembly found that most of the sequence reads mapped to seven contigs of approximately 1000–2400 bp . Open reading frame ( ORF ) analysis of the contigs found a single ORF for all segments , with the exception of two ORFs for the smallest contig . BlastP searches of the putative proteins identified modest homology to human influenza C virus , suggesting that this virus was distantly related to human influenza C virus ( Table 1 ) . Consequently , the virus was provisionally designated C/swine/Oklahoma/1334/2011 ( C/OK ) . The genomic coding sequences of all segments were determined and used for subsequent genetic and phylogenetic analyses . Because PB1 is reported to be the most conserved influenza virus protein , it is frequently used to evaluate the evolutionary relationship among influenza viruses [18] . We firstly performed ClustalW alignment of predicted polymerase basic 1 ( PB1 ) amino acid sequences of influenza A , B , and C viruses . C/OK shared approximately 69%–72% mean pairwise identity to influenza C viruses and 39%–41% identity to influenza A and B viruses . Homology between the influenza A and B PB1 proteins was approximately 61% and intrasubtype PB1 proteins of influenza A viruses are extremely conserved reaching up to 90% homology . PB1 protein alignments indicated that C/OK was more closely related to influenza C viruses than to influenza A and B viruses but more distant from individual members of human influenza C type . Pairwise identity between C/OK and influenza C viruses was considerably lower for polymerase basic 2 ( PB2 ) and polymerase 3 ( P3 ) ( 53% and 50% , respectively ) . In influenza A and B viruses segment 3 is referred to as polymerase acidic ( PA ) protein because of its pKa of approximately 5 . 2 . Conversely , segment 3 of influenza C viruses encodes a polymerase with a neutral pH ( pKa ∼7 . 2 ) and is referred to as P3 [18] . Interestingly , the predicted pKa of C/OK P3 is 6 . 2 , which is between those of the influenza A/B and influenza C viruses . In influenza C virus , a hemagglutinin esterase ( HE ) protein is responsible for receptor binding , receptor destroying ( acetylesterase ) , and membrane fusion activities , whereas in influenza A and B viruses , separate hemagglutinin ( HA ) and neuraminidase ( NA ) proteins perform these functions in a cooperative fashion . The pairwise sequence identity of the C/OK and human influenza C virus HE proteins was 53% , similar to the 49% observed across the influenza A HA subtypes [10] but clearly higher than the HA homology ( approximately 25–30% ) between influenza A and B viruses . NS1 of C/OK virus had the lowest homology to its counterpart in human influenza C viruses ( 29%–33% identity ) , similar to the less conserved influenza A and B NS1 proteins ( 22% identity ) . Like PB1 , the nucleoprotein ( NP ) and matrix ( M ) proteins are highly conserved among members of each genus of influenza viruses . Despite the high intragenic homology ( >85% ) , NP and M1 are highly variable among the three influenza virus genera and their intergenic homologies are only about 20–30% , which serve as genus-specific antigens that distinguish between the influenza A , B , and C viruses [19] , [20] . The amino acid sequence of the C/OK NP had 38%–41% identity to influenza C viruses . Unspliced mRNA from the C/OK M segment 6 encodes the polyprotein P42 , which is cleaved by a signal peptidase to yield M1′ and CM2 [21] . P42 was analyzed due to unknown mRNA splice and protein cleavage sites used by C/OK virus to generate M1 and CM2 , respectively . The C/OK P42 had 38% identity to influenza C viruses . Relative low homologies of NP and M proteins between C/OK and human influenza C viruses are interesting but seem to be consistent with pairwise protein homology analysis for polymerase and non-structural proteins . In addition to the coding region , each RNA segment of influenza viruses also contains noncoding ( NC ) regions at its 5′ and 3′ ends . These NC regions are highly conserved , particularly those at the terminal ends , among the genome segments of each species . These regions form panhandle structures by partial inverted complementarity between the 5′ and 3′NC regions and play a critical role in genome replication and packaging [22]–[25] . Using 5′ and 3′ RACE coupled with direct PCR sequencing by the Sanger method , we determined the complete sequences of the 3′ and 5′ NC regions of the seven segments of the C/OK virus ( Table S1 ) . The 3′ and 5′ NC region sequences of C/OK genome segment were similar to those of human influenza C viruses with the exception of one nucleotide ( position 5 from the 3′-terminus ) and polymorphism at position 1 of the 3′terminus . Viral RNA packaging sequences are composed of the 5′ and 3′NC regions and the terminal coding sequences of each segment . Incompatibility between homologous segment packaging sequences has been shown to prevent segment reassortment [26] . Nearly identical NC sequences at the proximal ends of seven RNA segments observed between C/OK and human influenza C viruses suggest a potential for viral segment reassortment in nature . Significant variability was observed in the NC regions immediately adjacent to each coding region for C/OK as compared to human influenza C; however previous work demonstrated that the highly conserved NC region at the proximal ends of the segment plays a key role in transcription and replication [27] . Our phylogenetic analysis used representative influenza A , B , and C viruses ( Fig . 2 ) . The segments encoding the C/OK virus PB2 , PB1 , P3 , NP , M and NS clustered most closely with influenza C viruses , suggesting that these C/OK genes diverged from known human influenza C viruses after they diverged from influenza A and B viruses but before they diverged from previously sequenced influenza C viruses . As HE does not occur in influenza A and B viruses , only influenza C viruses were included in that analysis . Previous studies have found that multiple genetically and antigenically distinct but related lineages of influenza C virus co-circulate and frequently reassort [6] , [28]–[30] . Given this evidence , it is puzzling that the seven segments of C/OK are only slightly to moderately homologous to characterized influenza C viruses . An HI assay was performed to determine the antigenic cross-reactivity and seroprevalence of C/OK virus in humans and swine . The assay included reference strains of influenza A , B , and C genera and their matched antisera ( Table S2 ) . No cross-reactivity was observed between C/OK virus and heterologous antisera . For the human cohort , we used a set of 316 serum samples . These sera originated from patients recruited in the Greater Vancouver area of British Columbia , Canada , or in the vicinity of the Greater Hartford area of Connecticut during the 2007–2008 and 2008–2009 influenza seasons as described in Marcelin et al . [31] . All but four of these sera had undetectable C/OK HI titers ( ≤10 ) . Three had HI titers of 20 , but each of these also had high titers ( 160 , 320 , and 1280 ) to the human influenza C isolate C/Yamagata/10/1981 . The remaining positive sample had a HI titer to C/OK of 40 with no corresponding titer to the human influenza C isolates tested . The low titers and number of positive samples ( 1 . 3% ) obtained are inconclusive in determining circulation of C/OK in the human population , especially as thirty-four percent of the serum samples had HI titers ≥20 to C/Yamagata/10/1981 which is consistent with previous studies showing approximately 60% of elderly humans retain influenza C virus antibody titers [32] . Swine serum samples ( n = 220 ) submitted to Newport Laboratories for unrelated diagnostic testing by commercial swine production facilities nationwide were similarly analyzed . Sera were collected from pigs aged 3–20 weeks from March through September 2011 . HI titers ( range , 10–80 ) were detected in 9 . 5% of samples , with a GMT of 20 . 7 . To assess the specificity of the HI titers to C/OK in swine sera , we performed HI assays using the human influenza C virus C/Taylor/1233/47 ( C/Taylor ) . Only 2 . 8% of the swine sera had measurable titers ( range , 10–20 ) . Taken together , these results suggest that C/OK virus circulates in swine populations but is not widespread in humans . Further serologic studies focusing on individuals occupationally exposed to swine are required . To better understand the pathogenesis and epidemiology of C/OK , we performed infection studies with ferrets and swine . We first addressed the zoonotic potential of C/OK virus by conducting a pathogenesis and transmission study in the ferret model . After intranasal inoculation of ferrets , C/OK virus was first detected in nasal washes on day 3 ( mean titer , 3 . 3 log10 TCID50/mL ) ( Fig . 3 ) . C/OK virus was first detected in ferrets exposed by direct contact to inoculated ferrets on day 7 , reaching a mean titer of 4 . 3 log10 TCID50/mL by day 10 . Virus was not detected in ferrets exposed to respiratory droplets . No clinical signs of disease were observed . In the tissues of ferrets on day 5 post-inoculation ( p . i . ) , a mean titer of 3 . 9 log10 TCID50/mL was observed in the nasal turbinates , but no virus was detected in the upper and lower trachea , lung , small intestine , liver , or spleen . Histopathological examination of lung tissues showed no typical influenza lesions . These results are consistent with a previous study that investigated human influenza C replication in ferret alveolar macrophage cells where viral replication with titers >104 egg infectious dose 50 from days 4 to 9 were measured with no cytopathic effects [33] . All ferrets that were inoculated or exposed by direct contact and 1/3 of the ferrets exposed to respiratory droplets seroconverted 3 weeks after exposure as measured by HI assay ( GMT = 780 ) . To assess the pathogenicity and transmissibility of the virus in swine , we similarly challenged swine intranasally with C/OK ( Fig . S1 ) . Virus was first detected in nasal swabs on day 3 p . i . by using an rt-RT-PCR method specifically developed for C/OK virus . Virus shedding peaked at day 8 p . i . and remained detectable on day 10 . Virus was detected in swine exposed by direct contact on days 7 and 9 after exposure . No clinical signs of illness were observed . Lung samples collected from inoculated swine on day 7 p . i . showed no evidence of the virus by rt-RT-PCR . Histopathological examination of lung tissues showed no typical influenza lesions . Sera collected on day 14 p . i . from donor pigs were positive for antibodies to C/OK virus in an HI assay ( GMT = 30 . 3 ) . All 5 pigs were positive for antibodies to C/OK . Additionally , 2 of the 5 direct contact pigs seroconverted by day 13 post exposure . These data suggest that in animals the replication kinetics is slower for C/OK virus than for influenza A viruses and infection in both swine and ferrets was limited to the upper respiratory tract . The ability of C/OK to readily transmit to contact ferrets suggests that a level of transmission potential to humans is possible . Zoonotic H5N1 and H9N2 influenza A viruses are typically unable to transmit in ferrets . We compared in vitro cellular tropism between C/OK and human influenza C viruses by rt-RT-PCR in several cell lines , including ST , adenocarcinomic human alveolar basal epithelial ( A549 ) , Madin-Darby canine kidney ( MDCK ) , Green African monkey kidney ( Marc-145 ) , human rectal tumor ( HRT-18G ) , baby hamster kidney ( BHK-21 ) and porcine kidney ( PK-15 ) cells . C/OK replicated , in order of highest replication , in ST , MDCK , Marc-145 , HRT-18G and A549 cells ( Fig . S2A ) . Minimal replication was observed in BHK-21 and PK-15 cells . In marked contrast , the human influenza C virus C/Taylor showed poor growth and replicated only in ST and HRT-18G and not in other cells tested ( Fig . S2B ) . It should be noted however that cultivation of influenza C/Taylor virus was performed at 33°C , the optimal temperature that is typically used to propagate human influenza C virus [34] . This virus failed to replicate at 37°C in our hands which differs from the C/OK virus that can replicate efficiently at this temperature ( data not shown ) . These results suggest that C/OK has a broader cellular tropism than C/Taylor and is also not restricted at elevated temperatures for replication . The influenza C virus utilizes the 9-O-acetyl-N-acetylneuraminic acid ( Neu5 , 9Ac2 ) as the primary receptor for attachment to the cell surface to initiate infection [35] . The receptor binding specificity and affinity are mainly determined by the 9-O-acetyl group of the Neu5 , 9Ac2 [36] . As a result , the virus encodes a sialate-O-acetylesterase , not neuraminidase , in order to release virions from infected cells by cleavage of the 9-O-acetyl group [37] . To provide structural insights to the observed different tropism , we conducted structural modeling of the C/OK HE protein in complex with the receptor based on the solved X-ray crystallographic structure of a human influenza C virus ( C/Johannesburg/1/66 ) HE protein [36] . The overall 53% sequence identity between the two HE proteins allows us to predict important structural features such as the receptor-binding pocket and the enzymatic active site of the C/OK HE protein . Based on the assumption that the HE protein uses similar sites for function , our structural modeling analysis identified a conserved enzymatic active site but revealed a variable receptor-binding pocket between two HE proteins ( Figs . 4 and S3 ) . These results are consistent with the observed difference in cellular tropism . For example , both HE proteins possess an identical catalytic triad: S71/H369/D365 for C/Johannesburg and S73/H375/D372 for C/OK ( Fig . 4A ) . The other two substrate-interacting residues for optimal enzymatic function are also completely conserved in the two HE proteins ( G99/N131 for C/Johannesburg and G101/N133 for C/OK ) . In addition , both HE proteins utilize two conserved arginine residues for substrate binding ( R72/R332 for C/Johannesburg and R74/R342 for C/OK ) . The conserved enzymatic site between C/Johannesburg and C/OK suggests that C/OK utilizes 9-O-acetyl sialic acid as the cellular receptor for infection . Analysis of the receptor-binding pocket formed by a cluster of noncontiguous amino acid residues revealed some interesting similarities and differences between C/Johannesburg and C/OK . The influenza C HE protein uses two binding pockets for recognizing the receptor: one binds to the 9-O-acetyl group while the other engages the 5-N-acetyl group [36] . As shown in Fig . 4 , the 5-N-acetyl binding pocket of C/OK HE becomes smaller as compared to that of C/Johannesburg due to L→W substitution , i . e . L198 in C/Johannesburg is replaced by W201 at C/OK HE position 201; the large rigid aromatic side-chain of W201 extends into the binding pocket ( Fig . 4B , 4C , and 4D ) . The binding pocket for the 9-O-acetyl group is nearly identical in C/Johannesburg and C/OK , implying the utility of the 9-O-acetyl-neuraminic acid for C/OK virus infection . For interacting with the 9-O-acetyl group , human influenza C viruses utilize a cluster of amino acid residues Y141 , F239 , Y241 , R250 , and R302 , which is also used by the C/OK virus except for two phenylalanine residues replacing tyrosine ( Y141 ) and arginine ( R250 ) , respectively . We hypothesize that these amino acid differences may alter the binding specificity and affinity of the HE protein to the receptor that in turn result in the observed difference in cellular tropism between two viruses .
Previous studies have indicated the presence of multiple lineages and antigenic groups in influenza virus type C virus [28]–[30] . Despite such variation , it has been long viewed that the influenza C viruses consist of a single subtype [1] . In contrast to this conventional wisdom , here we describe the characterization of a novel influenza C virus from swine with influenza-like illness . The phylogenetic analyses , together with the observations of genomic structure , indicated that this novel virus is more closely related to influenza C than to other members of the Orthomyxoviridae family including influenza A and B viruses , and suggested that the virus could be considered a new subtype of influenza C virus , despite its divergence from human influenza C viruses , which is similar to the divergence between influenza A and B viruses . Identification of this virus was not an isolated case , as we have identified four additional swab samples from swine showing influenza-like symptoms that were positive for this virus in a RT-PCR assay ( data not shown ) . These samples were collected from different pig farms across the U . S . between 2010 and 2012 . The finding that influenza C virus , like influenza A , harbors multiple subtypes is significant . It suggests the possibility of reassortment between subtypes , which could potentially generate viruses with phenotypes that may pose a threat to public health . Nine and a half percent of surveyed swine possessed antibodies specific to C/OK indicating that this previously unidentified virus circulates in U . S . swine . A causal relationship is evidently supported by the experimental infection of pigs . As such , reassortment may occur between C/OK and human influenza C viruses in pigs because both viruses can infect and transmit among pigs and because swine has been documented in serving as a mixing vessel for reassortment of influenza A viruses [38] , [39] . Despite a lack of compelling evidence of C/OK virus infection in humans , we suspect that the virus may infect and replicate in the human population because of the following reasons: First of all , the virus infects and transmits in ferrets , a surrogate for human influenza pathogenesis studies . The ability of C/OK virus to readily transmit to contact ferrets suggests that a level of transmission potential to humans is possible . Second , C/OK virus displays a broader cellular tropism compared to human influenza C virus , and this virus seems quite plastic in terms of propagation because high temperatures such as 37°C do not restrict its replication at least in cell culture . Human influenza C virus causes a mild respiratory disease in humans and the infection is normally confined to the upper respiratory tract , although occasionally it can also cause lower respiratory infection [5] . Considering that C/OK virus varies significantly from currently circulating human influenza C viruses in the amino acid sequences of predicted proteins , assessment of clinical disease in humans , particularly in children , caused potentially by C/OK is justified . Knowledge of its presence in human clinical settings is important to any future attempt to manage and control the disease outbreak . By using the nucleotide sequence of C/OK virus reported in this work , sensitive and specific diagnostic methods can be developed to investigate the pathogenesis and epidemiology of this novel virus in humans . Influenza C virus is not readily isolated and cultured , and primary isolation can be challenged . The obstacle is largely due to a lack of suitable cell lines for influenza C virus isolation as suggested previously [5] , [34] . In contrast , identification and cultivation of the C/OK in the ST cell line is relatively straightforward . The emerging but puzzling question then is why this virus has not been identified until now . We suspect that several factors including use of less susceptible cell lines and complicated co-infection often involving influenza A viruses and other viruses having the capacity to agglutinate red blood cells may account for the previous failures in identifying this virus . Alternatively , the C/OK virus may have spread to swine in recent years from an unknown animal reservoir . On-going retrospective seroepidemiological analyses will help to address this question . The difference in cellular tropism between C/OK and human influenza C may be a result of differences in the receptor recognition of the HE protein . To explore the possible structural origins of this difference , we created a homology model for C/OK using the crystal structure of the influenza C HE protein ( Fig . S3 ) . The high sequence identity ( 53% , Fig . S3 ) between C/OK HE and influenza C suggest that the quality of this model will be high , given previous work in homology modeling and structure validation for influenza A and B HA proteins in which similar sequence similarities gave RMSD errors of ∼1 Å [40] , [41] . The predicted receptor binding for C/OK HE appears on the top face of the receptor binding domain similar to human influenza C HE protein and remote homolog HE proteins of other viruses such as coronavirus and torovirus ( ∼30% sequence similarity to influenza C HE ) [42] . Here , four out of nine residues of the receptor binding site and residues around the binding pocket of HE of human influenza C are retained in that of C/OK ( Fig . S4 ) . Previous work has shown that receptor binding specificity and affinity are sensitive to substitutions in the receptor binding site of HE and HA [43] , [44] . Examining the predicted receptor-binding site of C/OK HE revealed changes in the receptor binding site relative to human influenza C HE . The most notable difference is a reduction in the size of the 5-N-acetyl binding pocket of C/OK HE relative to C HE due to the L198 in C HE being replaced by W in C/OK HE ( Fig . 4 ) . Alternatively , the binding pocket for the 9-O-acetyl group is similar between the two HE proteins ( Fig . 4 ) . We speculate that these differences may indicate that C/OK HE utilizes a different substrate than influenza C HE . Further experimental verifications of are required to test this prediction . It has been suggested that influenza A , B , C viruses have a common precursor , and of the three virus types , influenza A and B viruses are much more similar to each other in genome organization and protein homology than to C viruses , which suggests that influenza C virus diverged well before the split between A and B viruses [9] . Numerous studies have shown that influenza C viruses have the slowest evolutionary rate among influenza viruses [13] , [19] , [45]–[47] . One theory is that influenza C viruses , like influenza B viruses , are close or at an evolutionary equilibrium in humans , whereas influenza A viruses have not yet reached an equilibrium [48] . Consistent with this hypothesis is that only a single subtype is thought to exist for influenza B and C viruses and humans , not other mammals , are the primary hosts of influenza B and C viruses . The discovery of C/OK in pigs , being distantly related to human C viruses , seems to challenge these accepted views and warrant future studies of influenza C virus evolution . Virus nomenclature is the subject of discussion and there is still a possibility that C/OK virus can be assigned as the prototype of a new genus of the Orthomyxoviridae family . Most compelling evidence in support of the tentative designation of the C/OK virus are ( i ) seven genomic segments , ( ii ) 3′ and 5's NC regions similar to those of human influenza C viruses , and ( iii ) HE protein sharing approximately 53% homology with that of influenza C viruses . The last parameter is the primary determinant to classify subtypes of influenza A virus . However , the overall divergence between C/OK and human influenza C viruses is similar to that observed between influenza A and B viruses and argues for classification of C/OK into a potential new virus genus . Influenza A , B , and C viruses are classified on the basis of antigenic differences between their nucleoprotein ( NP ) and matrix ( M ) proteins [1] . Intriguingly , only modest homologies of these structural proteins between C/OK and human influenza C further confound the provisional classification for this newly discovered virus . Recently the family Orthomyxoviridae was expanded by including two novel genera , Thogotovirus , consisting of three viruses that infect birds and ticks , as well as the genus Isavirus , consisting of infectious salmon anemia virus [49] , [50] . For the novel swine influenza virus reported here , perhaps not until attaining more detailed serological , virological , and molecular data , a final classification of this virus can be made . Of particular importance are reassortment experiments between C/OK and human influenza C viruses . While similar , two discrepancies were found in the NCR's of C/OK as compared to human influenza C viruses . It is not known whether these mutations prevent reassortment between C/OK and human influenza C viruses . Preliminary reassortment experiments between C/OK and C/Taylor have been performed and have failed to identify reassortant viruses . More detailed studies are underway . In summary , we identified a novel influenza C virus that infects and spread among pigs or ferrets by direct contact . The ability of this novel pathogen to infect ferrets; a surrogate for human influenza infection suggests that such viruses may become a potential threat to human health . Our finding reported in the present study raises several interesting questions . Does this influenza C-like virus have the capability of generating a viable reassortant with currently circulating human C viruses ? If so , could such a reassortment allow influenza C virus to diverge and to have greater pathogenicity ? When and where did this novel virus emerge ? What is its animal reservoir in nature ? Future elucidation of these questions will provide insights into the ecology , virology , and pathobiology of influenza C virus .
Ferret experiments were conducted in an Animal Biosafety Level 2+ ( level 2 with enhanced biocontainment for pandemic H1N1 influenza A virus ) facility at St . Jude Children's Research Hospital , in compliance with the policies of the National Institutes of Health and the Animal Welfare Act and with the approval of the St . Jude Children's Research Hospital Animal Care and Use Committee ( IACUC No . 428 ) . Pig experiments were performed at Newport Laboratories under biosafety level 2 conditions in accordance with the Guide for the Care and Use of Agricultural Animals in Research and Teaching and were approved by the Institutional Animal Care and Use Committees at Newport Laboratories ( IACUC No . 02-2012 ) . Nasal swabs were collected from 15-week-old pigs exhibiting influenza-like illness at a commercial swine production facility in Oklahoma , USA , in April , 2011 . Viral isolation was performed on swine testicle cells and cytopathic effects were evident by day 3 post inoculation . Detailed information on cell culture conditions is available in SI Appendix , Supplementary Materials and Methods . Hemagglutination assays were performed using chicken red blood cells . C/OK was concentrated from ST cell supernatants by ultracentrifugation and subsequently purified through a 20% sucrose cushion by ultracentrifugation . Viral RNA was isolated using the Qiagen Viral RNA Isolation Kit and converted to cDNA using random primers included in the GoScript Reverse Transcription Kit ( Promega ) . The cDNA was made double stranded with DNA polymerase and used to construct a library for Ion Torrent Sequencing . Detailed sequencing methodology is available in SI Appendix , Supplementary Materials and Methods . Contigs were assembled de novo by using SeqMan NGen software ( DNAStar ) . Contigs encoding proteins with homology to influenza C proteins were identified by BlastP analysis . The genome sequence of C/OK was submitted to Genbank under accession no . JQ922305-JQ922311 , relating to segments 1–7 , respectively . Phylogenetic analyses were performed by using Mega 5 software [51] . Evolutionary analyses were conducted by using the Maximum Likelihood algorithm , and the tree topology was verified by performing 1000 bootstrap replicates . The PB1 sequence was used to design primers and a Taqman probe for detection of C/OK ( position of primers and probe in PB1 gene: forward , nucleotides 1420–1439; reverse , nucleotides 1555–1535; probe , 1482–1460 ) . Viral RNA was extracted by using the MagMAX-96 viral RNA isolation kit ( Life Technologies ) according to the manufacturer's instructions . rt-RT-PCR was performed by using QIAGEN Quantitect RT-PCR with the C/OK primers and probe . Method specificity was assessed by using influenza A , B , and C reference viruses , and no cross-reaction was observed . Swine testicle ( ST ) cells were grown in DMEM containing 5% fetal bovine serum . Influenza C/Taylor/1233/47 virus was provided by BEI Resources ( NIAID ) . For infection , cell medium was replaced with DMEM and viral inoculum was added at a multiplicity of infection of 0 . 001 . Viral growth studies were performed on a monolayer of ST , A549 , Marc145 , HRT-18G , BHK-1 , or PK-15 cells using an inoculum of 1 . 0–3 . 0 TCID50/ml in duplicate ( multiplicity of infection 1×10−5–1×10−3 ) . C/OK virus replication was performed at 37°C while C/Taylor replication was done at 33°C . Samples were removed at 0 , 24 , 48 , and 72 hours p . i . and virus was titrated by rt-RT-PCR . Experiments were performed three times in duplicate . Human sera were treated with receptor-destroying enzyme ( Denka Seiken Co . , Tokyo , Japan ) overnight at 37°C , heat-inactivated at 56°C for 30 min , diluted 1∶10 with PBS , and tested by hemagglutination inhibition ( HI ) assay with 0 . 5% packed chicken red blood cells ( cRBCs ) as described in the WHO Manual on Animal Influenza Diagnosis and Surveillance [52] . The pathogenicity and transmission of the virus was tested in 3- to 4 month-old male ferrets . Detailed pathogenicity and transmission methodology is available in SI Appendix , Supplementary Materials and Methods . Three donors were inoculated intranasally under light isoflurane anesthesia with 106 TCID50 of swine/Oklahoma/1334/2011 virus in 1 ml of sterile PBS . Two additional ferrets were similarly inoculated and were housed separately for virus titration and histopathology in organs . At 23 h p . i . , each of the three remaining donor ferrets was housed in a cage with one naïve direct-contact ferret ( n = 3 ) . An additional ferret ( n = 3 ) was placed in an adjacent cage separated from the donor's cage by a two layers of wire mesh ( ∼5 cm apart ) that prevented physical contact but allowed the passage of respiratory droplets . Clinical signs of infection , relative inactivity index [53] , weight , and temperature were recorded on days 0 , 3 , 5 , 7 , and 10 p . i . . Nasal washes were collected from ferrets 3 , 5 , 7 , and 10 days p . i . Two donor animals were euthanized 5 dpi , and tissue samples were collected . Samples were homogenized and virus was titrated ( log10 TCID50 per gram of tissue ) in ST cells . Tissues were also subjected to histopathologic analysis . Swine challenge studies were performed at Newport Laboratories under biosafety level 2 conditions . Twenty-eight swine approximately 10 weeks of age were obtained from a commercial high-health herd . Eleven swine were placed in a single room and inoculated intranasally with 6 . 0 log10 TCID50 of C/OK . On day 1 p . i . , 11 naïve direct-contact swine were introduced into the room . Temperatures were recorded and nasal swabs were collected on days 0 , 2 , 3 , 6 , 8 , and 10 p . i . Six inoculated swine and three mock-inoculated swine were euthanized on day 7 p . i . and lung specimens were fixed in 10% neutral buffered formalin and submitted for histopathological analysis . The remaining swine were euthanized on day 14 p . i . Nasal swabs and lung tissue were analyzed by rt-RT-PCR as described above . The structure of C/OK HE ( aa 17–620 ) was modeled using Modeller 9 . 10 [54] . The structure of C/Johannesburg/1/66 HE was used as template ( PDB id: 1FLC ) [36] because the two HE proteins show both high sequence and secondary structure similarity . The quality of the modeled structure was estimated by Verify_3D and 95 . 47% of the modeled residues are compatible with the structure ( averaged 3D-1D score >0 . 2 ) [55] , suggesting that the quality of the modeling is good . The distance between the modeled structure and template is 0 . 49 angstrom . Electrostatic surface maps were generated using APBS [56] . | Influenza C viruses infect most humans during childhood . Unlike influenza A viruses , influenza C viruses exhibit little genetic variability and evolve at a comparably slower rate . Influenza A viruses exist as multiple subtypes and cause disease in numerous mammals . In contrast , influenza C viruses are comprised of a single subtype in its primary human host . Here we characterize a novel swine influenza virus , C/swine/Oklahoma/1334/2011 ( C/OK ) , having only modest genetic similarity to human influenza C viruses . No cross-reaction was observed between C/OK and human influenza C viruses . Antibodies that cross react with C/OK were identified in a significant number of swine but not human sera samples , suggesting that C/OK circulates in pigs . Additionally , we show that C/OK is capable of infecting and transmitting by direct contact in both pigs and ferrets . These results suggest that C/OK represents a new subtype of influenza C viruses . This is significant , as co-circulation of multiple subtypes of influenza allows for rapid viral evolution through antigenic shift , a property previously only shown for influenza A viruses . The ability of C/OK to infect ferrets along with the absence of antibodies to C/OK in humans , suggests that such viruses may become a potential threat to human health . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"emerging",
"infectious",
"diseases",
"viral",
"transmission",
"and",
"infection",
"virology",
"viral",
"classification",
"biology",
"microbiology",
"viral",
"evolution"
] | 2013 | Isolation of a Novel Swine Influenza Virus from Oklahoma in 2011 Which Is Distantly Related to Human Influenza C Viruses |
Swarming , a collective motion of many thousands of cells , produces colonies that rapidly spread over surfaces . In this paper , we introduce a cell-based model to study how interactions between neighboring cells facilitate swarming . We chose to study Myxococcus xanthus , a species of myxobacteria , because it swarms rapidly and has well-defined cell–cell interactions mediated by type IV pili and by slime trails . The aim of this paper is to test whether the cell contact interactions , which are inherent in pili-based S motility and slime-based A motility , are sufficient to explain the observed expansion of wild-type swarms . The simulations yield a constant rate of swarm expansion , which has been observed experimentally . Also , the model is able to quantify the contributions of S motility and A motility to swarming . Some pathogenic bacteria spread over infected tissue by swarming . The model described here may shed some light on their colonization process .
Bacterial swarming , a coordinated motion of many bacterial cells , facilitates their spread on the surface of a solid medium , like agar [1] . Swarming may have evolved to permit the bacteria in a colony to expand their access to nutrients from the subsurface and to oxygen from above . When the surface is a tissue in a live host , pathogenic bacteria swarm to create a biofilm and to spread the infection . Swarming is observed in cells that are propelled by rotating flagella [2] , by the secretion of slime [3] , and by retracting type IV pili [4 , 5] . Bacterial swarming has been studied quantitatively in the modeling context of self-propelled particle systems [6–8] . Most models , such as those for Bacillus subtilis and Escherichia coli ( see [8] for a review ) , are based on long-range cellular interactions facilitated by chemical gradient or nutrient level ( chemotaxis ) . However , myxobacteria show no evidence of long-range communicating systems to guide their collective motion; they have only local contact signaling and use social interactions between neighboring cells for swarming [9] . How interactions between cells facilitate swarming is still an open question . Understanding this question might shed light on the self-organizing process in bacteria , such as the spreading of a biofilm in an infected tissue and the development of multicellular fruiting bodies [4 , 5] . In this paper , we describe a new cell-based model and study the effects of social interactions between cells , including the interaction mediated by slime trails and by type IV pili , on swarming . Type IV pili are found at one pole of a wide range of bacteria , including many pathogens that cause plant and animal disease . We chose to examine Myxococcus xanthus because it swarms rapidly , has typical type IV pilus engines at the front end of cells , has slime secretion engines at the rear , and coordinates the two engines with each other . M . xanthus has been studied for more than a century; numerous swarming mutants have been identified and characterized . Myxobacteria are commonly found in cultivated soils , where they feed on other bacteria . On the surface of nutrient agar , they swarm away from a point inoculum , spreading outward at a constant rate for 2 wk . Although the bacteria are growing ( and in fact they must grow to swarm ) , 90% of the swarm expansion rate is due to motility and to interactions associated with motility , as shown by the low spreading rate of nonmotile mutants [10 , 11] . Individual M . xanthus cells are rod shaped , roughly 5 μm in length and 0 . 5 μm in width . They have two types of molecular motors that provide the thrust necessary for their gliding movement over a surface [9] . At the leading end of the cell are retractile type IV pili , long and thin hairs responsible for S motility . When a cell is close to a group of other cells , the cell's type IV pili can attach to the fibrils , which cover the surface of the neighboring group of cells like a fisherman's net . After attachment , the pilus retracts , and the retraction force pulls the piliated cell forward , while the group hardly moves . This pilus-mediated interaction produces many asymmetric cell clusters that often have tips that are pointed at one end ( arrowhead-shaped ) and is characteristic of S motility . Arrowheads can be seen in the young swarms of A−S+ mutants [11] . S motility is found among pathogenic Neisseria and Pseudomonas , where it is called twitching motility [4 , 5] . At the trailing end of myxobacterial cells are several hundred pores , from which slime is secreted . There are roughly 150 pores scattered over the sides of the cell , also secreting slime that becomes a thin layer , protecting the cell from lysis by cell-wall digestive enzymes being secreted by all the cells [3] . Both the lateral and the polar slimes are thought to be the same polysaccharide that is part of A motility ( hereafter simply referred to as slime ) . Importantly , slime is completely distinct from the fibril polysaccharide that serves S motility [11] . Slime secretion from the rear pushes the cell forward , leaving a trail of slime behind the cell [3 , 12] and generating movements called A motility . When a moving cell encounters a slime trail , it tends to turn through the acute angle to follow the slime trail . When an A motile cell collides with the side of another cell , the pushing of the slime engines at the rear causes the cell , which is flexible , to bend . The colliding cell thus reorients parallel to the other cell , producing a side-by-side cluster of cells . Such clusters are transient because the two cells do not adhere and often slide past one another . The A and S motility engines , which are located at opposite poles of the rod-shaped cells , have engine-specific social interactions . During movement , a cell's polarity reverses regularly every 10 min or so [13 , 14] , and reversal is required for swarming [15 , 16] . A wild-type cell ( A+S+ ) expresses both A and S motilities . A+S− mutants express only A motility , while those with S motility but no A motility are called A−S+ mutants [9] . Because wild-type and A+S− mutants are self-propelled by A motility engines , a comparison can expose the social interactions specific to the type IV pili . In both cases , individual cells are observed to move , stop , and move again , sometimes slightly changing direction and regularly reversing [3] . To investigate the coordinated motion within M . xanthus swarms , culture droplets of each mutant were placed on agar plates , and the swarm expansion rates were measured [10] . Figure 1 shows the edge of a typical swarm of wild-type ( A+S+ ) cells . It is observed that swarm expansion rates remain constant until the swarm covers the entire surface available [10] . The expansion rates for various initial cell densities in K-S units were measured and plotted against the cell densities . ( K-S is Klett-Summerson unit; a measurement of cell density in suspensions [10] . A sample of cell suspension with 100 K-S units has approximately 4 ×108 cells/ml . Using the experimental data in [10] , we find that 100 K-S units correspond to a close-packing arrangement of cells in a 2-D area . ) The fitted functions of expansion rate data for the three cell types are shown as solid lines in Figure 2 . To a first approximation , the velocity of individual cells , when they are moving , is the same for S− mutants ( A+S− ) and wild-type ( A+S+ ) cells , about 4 μm/min , but their swarm expansion rates are different [10] . The A+S− and A−S+ mutants swarm with a maximum rate of 0 . 67 μm/min and 0 . 46 μm/min , respectively . Surprisingly , when S motility cooperates with A motility in wild-type M . xanthus ( A+S+ ) , the maximum swarming rate is 1 . 55 μm/min , about 2 . 3-fold larger than that of A+S− ( [10] , as shown in Figure 2 ) . Previously , we used a lattice-based model to study myxobacterial fruiting body development after starvation [17 , 18] . Swarming with sufficient nutrient supply has been studied using a continuous model in the form of partial differential equations ( PDEs ) [19] . The effects of engine mechanics and cell shape have yet to be taken into account . Recently , we introduced a simplified off-lattice stochastic description of swarming [20] , and herein add our current understanding of engine mechanics to investigate swarming and the role of social interactions . This paper is organized as follows . We start by describing the model of cell behavior and social interactions . Then , we present the simulation results and compare them with the experimental observations . We demonstrate a constant rate of swarm expansion and show that the model accounts for the significant difference in swarming rates between wild-type and A+S− myxobacteria arising from the loss of S motility . We also study in detail the order of collective motion in myxobacterial swarms . A detailed description of the computational model is given in the Methods section .
In this paper , we focus on the collective motion of a large number of cells in a swarm of high cell density , taking only the local , contact-mediated interactions between cells into account . We represent each cell as a string of N nodes in a 2-D space , following our earlier work [20] ( Figure 3 ) . The vector pointing in the direction from the tail node to the head node represents the orientation of a cell . We define an energy function ( Hamiltonian ) for the node configuration of a cell body and use it to constrain the cell length and the cell shape to a certain range . The active motion of an individual cell is then modeled as follows . After the head moves in a particular direction , a Monte Carlo approach [21] is used to reconfigure positions of other nodes in an attempt to minimize the Hamiltonian ( see Methods ) . This allows the cell body to bend and to change its length by random fluctuations , which reflects the experimental observations [22] . As mentioned in the introduction , the measured velocities of individual cells vary over a wide range , but the average velocities of A+S− and A+S+ cell types are similar . To a first approximation , we take the cell velocity to be constant and the same for wild-type A+S+ and A+S− cells , with a magnitude of 4 μm/min [10] . The direction of cell movement is determined dynamically by the model , which takes the interactions between neighboring cells into account . Frequently cells reverse their motion by 180° . Reversals are regulated by an internal biochemical clock that is not affected by collisions or other interactions between cells [15 , 16] . We model regular reversals of cell motility engines by switching roles of head and tail nodes in accordance with an internal clock ( see Methods ) . The swarming efficiency ( the ratio of the swarm expansion rate to the speed of individual cells ) of myxobacteria primarily depends on social interactions between neighboring cells . The expansion rate of a swarm without social interactions would be zero , since the cells would move back and forth equally without any net displacement in the long run . Social interactions help a swarm of reversing cells to spread . Ideally , interactions between all the more than 107 cells in a swarm would be considered , but that is not possible in practise . Instead , we try to identify for each cell a neighborhood within which a majority of its interactions are expected to be found . Social interactions arise in S motility when the type IV pili of one cell attach to the fibrils that surround other cells . Social interactions arise in A motility from the tendency of a cell to follow the trail of slime left by another cell , and from collisions between cells that cause a moving cell to stop and its engines to stall , or those cause a cell to change its direction . Using the experimental data of Figure 2 , an area of interaction for each cell type , A+S− , A−S+ , or A+S+ , was defined as the statistically averaged area around a cell within which most of its social interactions occur . The interaction areas were taken to be proportional to the inverse of parameters in the exponential term of the formula obtained when an exponential curve was fitted to the experimental data in Figure 2 . Fitting functions are specified in the legend to Figure 2 . Each curve represents the observed swarm expansion rate as a function of the initial cell density of the culture . The interaction area for wild-type cells was found to be smaller than the sum of the interaction areas of the A+S− and A−S+ mutants . We suggest that this unanticipated finding results when both engines are working because the two engines on a wild-type cell are not statistically independent but are constrained by the structure of a cell to propel it in the same direction . Pilus-mediated interactions depend on the dynamics of pilus retraction [23] and on the spatial distribution of the fibrils to which the pilus tips have attached [24 , 25] . Although these factors are mechanically complex and not yet understood in detail , the interaction has straightforward effects . Pilus retraction provides a driving force for cell movement that happens to be large , several times larger than the force developed by muscle acto-myosin . And , because the force is almost never directed along the cell's long axis , the force tends to reorient the direction of gliding . Because we are confined by the approximation that isolated cells move with constant speed , we need only consider the reorienting effect of pilus retraction . No effect on cell speed is considered , except that it drops to zero when one cell collides with another . Inasmuch as the fibrils tend to bundle groups of cells , as will be described below , the large size of the cell cluster prevents a significant reorientation of the bundle; only the cell whose pili have attached is reoriented . We model the reorientation effect of pilus-mediated interactions as driving the local alignment of cells ( see area I of Figure 4 and Methods ) . Although we represent the interaction area by a rectangle , a circle or some irregular domain could have been used . The important quality of an interaction domain is its area . That area is proportional to the probability that a cell has an interaction . Swarms of wild-type cells cover a larger area than those of A+S− or A−S+ mutants [11] . Moreover , the peninsulas are denser with cells that are well-aligned side by side [10] . Both effects illustrate reorientation due to pilus retraction . Cell clusters tend to be narrow in the case of an A+S− mutant and wide in the case of wild-type bacteria . A motility engines at the rear of the cell push it forward in the direction of their long axis . A motility also produces slime trails , and cells tend to follow them due to the adhesion of newly secreted slime to the older slime in the trail . The resulting alignment of the slime polysaccharide chains also reorients the direction of gliding . Slime trails are represented in the model by the paths that were taken by the last cells to have passed through area II ( Figure 4 ) . Further details are given in the slime orientation field described in Methods . Rod-shaped A motile cells , which are pushing at their tail ends , tend to form parallel arrays if they collide or come into close contact with each other . These effects are illustrated in area II of Figure 4 and are elaborated in Methods . Alignment results from inelastic collisions between cells that change their orientations . More generally , alignment in regions of high cell density arises spontaneously from the physical clustering of self-propelled rods [26] . For wild-type cells , we first model A and S motilities , individually . Then , we combine them under the approximation that isolated cells move at a constant rate , as described above . The persistent active motion is taken to be led by the head of the cell , no matter which engines are functioning . Finally , we model the reorientation due to pilus retraction and to the alignment of A motile cells with their neighbors , or with the slime field . To test the consistency of the model , we simulated the motion of cells near the edge of the swarm , and studied the expansion of the swarm . Although M . xanthus swarms consist of many millions of cells , the radial symmetry of a swarm makes it possible to consider a small rectangular sector of the swarm ( Figure 1 ) . A rectangular area of 200 μm by 200 μm ( Figure 5 ) was convenient . To compare the simulation with experimental measurements ( shown in Figure 2 ) , we considered that growth in the center of the swarm was driving a net radial outflow of cells from their center [15] , and that the swarm was expanding at a constant rate . A constant cell density near the swarm edge was observed experimentally as the edge moved out [10] . Skipping the early transient phases , we start the simulation after the steady state has been reached . Although cells in the initial area are oriented in all directions , the orientations are radially symmetric . Both conditions apply to the “Initial Area of Cells” in Figure 5 . Denoting cell density as p ( r , θ ) and the radial density as P ( r ) , due to the symmetry and the steady state , we have: This relation shows that the cell number flux across the lower boundary of the initial area ( or the increase rate of total cell number in the whole simulation domain ) is linearly correlated with the colony expansion rate in Figure 2 . We calculate the cell number flux rather than expansion rate directly . Therefore , we do not have to increase the simulation domain or the total number of simulated cells . Further details of the simulation setup , implementation of the algorithm , and the choices of parameters are described in Methods and Table 1 . Simulations show formation of long clusters ( peninsulas ) in both A+S+ and A+S− cases ( see Figure 6A and 6B ) , which was observed experimentally [10] . Simulations were performed for cell densities ranging from 2 to 200 K-S units , and linear increase of cell number was observed in all cases ( for example , see Figure 6C ) . This implies that the cell number flux is almost constant during the whole swarming process for a given initial cell density , in full agreement with experiments [10] . Figure 6 corresponds to an initial density of 50 K-S units , which is a near saturation density for the rate curves of Figure 2 . We have calculated linear fits for the cell number increase data at various cell densities , and taken the slopes to be the average cell number fluxes , as shown in Figure 6C . The results for both A+S+ and A+S− cells are plotted against cell density in Figure 7A . We found that the cell number flux of the wild-type cells ( A+S+ ) is greater than that of the A+S− mutant at all cell densities . At densities higher than 50 K-S units , the cell number flux for A+S+ is 2-fold larger than the A+S− . To see this effect more clearly , we fit the average cell number flux data into the first order exponential decay function , which is similar to the function used in Figure 2 from [10] . The fitting functions for wild-type A+S+ cells and A+S− mutants are found to be f ( x ) = 5 . 6 − 6 . 6 × exp ( −x / 32 . 1 ) and g ( x ) = 2 . 8 − 3 . 1 × exp ( −x / 29 . 4 ) , respectively . The ratio of these two fitting functions is plotted against cell densities in Figure 7B . It is equivalent to the ratio of colony expansion rates since the cell number flux is linearly correlated with the expansion rate . The ratio first increases and then saturates around 2 for cell densities higher than 50 K-S units . Experimental data shows that the A+S+ rates are 2-fold to 2 . 5-fold larger at cell densities higher than 50 K-S units ( Figure 2 ) . Therefore , our result shows a significant difference in swarming rates between wild-type and A+S− , arising from the contribution of S motility that agrees with the experiment . Collision occurs in the model whenever the head of one cell overlaps the area occupied by another cell . At this point , the moving cell stops; it is not permitted to glide on top of the other cell . As a consequence , at high cell densities the movement of individual cells is reduced . In reality , cells do glide over each other . Reduction becomes significant above 100 K-S units , because at 100 K-S units the average area occupied by an individual cell is close to the area of a cell body ( i . e . , the area is closely packed with cells ) . In practice , due to the tendency of cells to cluster , cell movement is reduced beginning at concentrations of 60 K-S units . This effect explains the decrease in cell flux observed at higher cell densities ( >60 K-S units ) for the wild-type cells in Figure 7A . The decrease results in a smaller value of maximum ratio ( about 2-fold; see Figure 7B ) than experimental data ( about 2-fold to 2 . 5-fold ) . Comparing the three curves of Figure 2 shows clearly that S motility contributes to the swarming of wild-type ( A+S+ ) cells . Figure 2 also shows that A−S+ swarms expand without help from A motility , although the rate of expansion is less than one-third that of the wild-type cells at every cell density . With these data in mind , a puzzle takes shape: how are pili able to support expansion of an A−S+ swarm when there should be no fibrils to which the type IV pili might attach beyond the edge of the swarm ? The surface ahead of the swarm edge never had cells upon it . Must the belief that pili attach to fibrils before they can retract be abandoned ? This section describes an attempt to solve the puzzle by examining the evidence that pili bind fibrils specifically , by offering a mechanism whereby specific binding and retraction can bring about the expansion of an A−S+ swarm , and by testing the mechanism proposed . Evidence for specific binding includes the observation that A−S+ cells move only when they are within a pilus length of another cell [10 , 27] . Fibrils are present in profusion , and they envelope clusters of adjacent cells ( see Figure 2 from [28] ) . Although only half of the fibril mass is polysaccharide ( the other half is protein [24] ) , several experiments have revealed that removing the protein has no effect on pilus binding [24 , 25 , 29–31] . Evidently , M . xanthus pili bind fibril polysaccharide . Therefore , side-by-side clusters of M . xanthus cells , like the peninsulas in Figure 8 , are viewed as a bundle that is enveloped by an elastic fisherman's net formed by association of polysaccharide fibrils that the cells have secreted . Bundling of cells by fibrils offers an explanation for the pointed shape , which A−S+ peninsulas tend to have . The points aim away from the swarm center ( Figure 8 and [10 , 32] ) and in the general direction of swarm expansion . The shape and orientation of the peninsula tips suggest that cells at the tip of the peninsula have been pushed into their position at the tip . Consider a cell within the body of the peninsula that happens to be moving toward the tip of the peninsula . This cell will have projected its pilus forward and attached it to the fibril network on cells ahead of it and closer to the tip . Retraction of that pilus could pull the cell forward and upward to add a new layer of cells to the peninsula . Indeed , most peninsulas have a second ( or third ) layer near their tips , which are evident in Figure 8 . On other occasions , retraction would pull the piliated cell right up to the end of a cell in the bottom layer that lies just ahead of our piliated cell . Recalling the description of A motility in the Introduction , each cell is also covered by the slime polysaccharide , which protects them from autolysis . Since the network of fibrils that envelops cells of a peninsula bundles them , both the elasticity of the fibrils and the cohesion between the slime on adjacent cells would tend to prevent their separation , by wedging action of the rounded end of the pushing cell , from cells to their left and right in the tip of the peninsula . Consequently , complete retraction of the pilus would cause the moving cell to push the cell in the peninsula that is immediately ahead of it . The pushed cell might slide forward while adhering through its slime covering to the cells on either side . Localized sliding would be reflected in a sharpening of the tip contour to a point , as observed ( Figure 8 ) . The hypothesis of pushing by S motile cells was tested by analysis of seven time-lapse movies of the advancing edges of A−S+ swarms , each movie of 1 h to 3 h in duration . Figure 8 is a single frame from one of the movies . In that frame , numerous single cells and many peninsulas of various sizes are evident . Several observations relevant to A−S+ swarming could be made from the movies ( Key and Kaiser , unpublished data ) . First , almost all of the many thousands of cell movements were found within clusters of ten or more cells . No isolated cell moved significantly , unless the cell was within pilus-striking distance of another cell . This shows that the cells are moving with S motility alone . Second , although the peninsulas either elongated or moved forward , the translocation rate was much less than the rate of individual cell movement in the same field . A lower rate correlates the peninsula's advance to its being pushed from behind , because the hypothesis has the pushed cell sliding past its neighbors in the peninsula . The sliding friction would decrease the rate of advance . Finally , the movies show many examples of individual cells , which appear to be moving more or less randomly , behind an arrowhead or a peninsula . Individual cells advancing toward the rear edge of the peninsula could have pushed it . A quantitative analysis of cell movement in the movies will be published separately , but this qualitative analysis supports pushing . In previous sections , we have shown that our model for social interactions is consistent with experimental results at the level of individual cells . In this section , we investigate how microscopic social interactions facilitate swarming at the population level . We demonstrate that social interactions lead to an increase in the order of collective motion , which is strongly correlated with swarming efficiency . We start by introducing an order parameter to characterize collective motion of bacteria in swarms with complex clustering patterns . After analyzing experimental data and taking into account regular reversals of myxobacteria cells , we define the most ordered state as follows: all cells move side by side in close contact with each other in the same or opposite direction . The collective motion is considered purely nonordered when either one of the following criteria is satisfied: ( 1 ) the orientations of neighboring cells of any given cell are random ( or uniformly distributed ) ; and ( 2 ) any pair of cells is well-separated so that cells are not in direct contact . Vicsek et al . [33] used the average velocity as a global order parameter for analyzing the motion of self-propelled particles . However , myxobacteria cells reverse regularly , and two opposite directions should be considered as being equivalent to each other . There are always cells moving in the direction opposite to the net motion of the whole cluster in most cell clusters in experimental movies . Also , as shown in the inset of Figure 1 , the swarming pattern often exhibits localized clusters of aligned cells with different orientations of motion , and one would need to take local order into account when measuring global order of motion . Therefore , the average velocity is not the best way of measuring the nematic order in myxobacteria swarms . We first define two local measuring components to describe the local orientational order and positional order of a given cell , denoted as Ψ and P , respectively . For a given cell k ( k = 1 , 2…M , M is the total cell number ) , we choose the rectangular domain ( of area s0 ) illustrated in Figure 9 as the local measuring domain ( one cell length by two cell lengths ) , centered at the center of mass of a cell . We then measure the total area S occupied by neighboring cells within the local measuring domain and define the local positional order as the following: We record the orientations θj , with j = 1 , 2 , …n of the neighboring cells with either head node or tail node inside the local measuring domain , in a way used in Equation 11 in Methods . Then the angles between these orientations and the x-axis in Figure 9 , , with j = 1 , 2 , …n , and , are calculated . If cell k has no neighbors ( j = 0 ) , we define Ψ as 0 . Otherwise , the local orientation order function is defined as follows: with and Φk determines how ordered the distribution of is . The most ordered state corresponds to the case when all are equal and Φk has a maximum value of 1 . Φk is rescaled to the range between 0 and 1 . Therefore , in the case of uniform ( random ) distribution of , the local orientational order function Ψk is equal to 0 . In the most ordered state , all are equal and Ψk is equal to 1 . Finally , we combine both the local orientational order and positional order components from Equations 3–5 to define the global order parameter for the collective motion of myxobacteria: where M is the total number of cells . The order parameter Ω has been specifically designed for myxobacterial swarming . Figure 10 shows values of Ω for the simulation of swarming near colony edge with initial cell density of 50 K-S units . We find that the order of collective motion in both A+S+ and A+S− swarms steadily increase , and that A+S+ cells achieve a much higher ( about 2-fold ) order than A+S− cells . Further , we look at the order of cellular motion in the inner area of myxobacteria colony . In Figure 11A , cells are randomly distributed in a square area of size 167 μm × 167 μm with a density of 50 K-S units . All boundary conditions are periodic . This is different from the previous simulations for cells near the colony edge , because we do not assume a preorganized orientation distribution of cells . Figure 11B and 11C are the simulation pictures after 3 h for A+S− mutant and wild-type cell ( A+S+ ) swarms , respectively . We see that the pattern of A+S− mutant exhibits lower order , while wild-type ( A+S+ ) cells form large clusters oriented in various directions . Plots of the parameter Ω are presented in Figure 11D . Again , we see that the order of motion in both A+S+ and A+S− cases increase with time , while A+S+ cells achieve a much higher order than A+S− cells . Therefore , we demonstrate that social interactions lead to an increase in the order of collective motion . Type IV pilus-mediated interactions increase the order much greater than social interactions associated with A motility . This is consistent with the experimental findings by Pelling et al . [32] , who observed higher-order patterns within wild-type ( A+S+ ) swarms in comparison with motility mutant swarms . Comparison of Figure 10B with the ratio of cell number fluxes ( Figure 7B ) indicates that the order of collective motion strongly correlates with the swarming efficiency . We suggest that higher order of motion results in greater swarming rates as observed in wild-type myxobacteria experiments . It explains the origin of the significant difference in swarming rates between wild-type and A+S− myxobacteria arising from the coupling of S and A motilities .
We have developed an off-lattice cell-based computational model to study the role of social interactions in bacterial swarming . The model is stochastic and is based on detailed description of the bacterial motility engines and their regulation . The model demonstrates how social interactions facilitate bacterial swarming , and provides an explanation to the significant difference in swarming rates between wild-type and A+S− mutants arising from the effects of S motility . Our simulations indicate that the order of collective motion strongly correlates with the swarming efficiency , which provides a connection between microscopic social interactions and population-level swarming behavior . The model is two-dimensional and provides a very good approximation for the bacterial behavior near the edge of the swarming population . However , in experiments at higher densities , cells were observed to glide on top of each other , resulting in multiple cell layers just behind the edge of the swarm . As discussed in Results , the 2-D nature of our model causes a slight decrease in cell number flux at higher cell densities ( >60 K-S units ) for wild-type myxobacteria ( Figure 7A ) , and results in a smaller value of maximum ratio ( about 2-fold; see Figure 7B ) than experimental data ( about 2-fold to 2 . 5-fold ) . A 3-D extension of the model will avoid such affects , and allows us to study cell clustering inside of a swarm as well as during fruiting body development under starvation [17 , 18] . We did not quantitatively study the motion of mutants with impaired A motility ( A−S+ mutants ) in this paper . As discussed in Results , A−S+ cells only have persistent active motion when they are within a pilus length of other cells so that the type IV pili can attach to the fibril materials on the surfaces of other cells [34] . Wild-type and A+S− mutants both have A motility that can produce persistent active motion . The only difference between wild-type cells and the A+S− mutant is the effects of S motility , so it is more convenient to take wild-type cells and the A+S− mutant as the modeling systems . By comparing their movements , we could investigate the role of pilus–cell interactions during swarming , which was one of our aims . In Results , we have presented a qualitative analysis of A−S+ swarming , which demonstrated that the pushing of cells near the swarming edge can explain the expansion of A−S+ swarms . Preliminary simulations with the pushing mechanism show qualitative agreement with the experiment in terms of peninsula shape and cell ordering ( unpublished data ) . Quantitative modeling of A−S+ swarming dynamics will require more knowledge of the distribution and mechanical properties of the fibril . When studying the effects of social interactions , we have related the swarming efficiency with the order of collective motion . This order parameter may provide a novel perspective on quantifying the condition of bacterial swarming . Further experimental investigation of this concept will rely on advances in microscope and image processing of microphotographs . Such experiments require very-high-resolution imaging that can cover large areas of a live bacterial colony [35] . We defined an appropriate order parameter , which characterizes the combined local orientational and positional order . Not limited to the case of myxobacteria swarming , the order parameter provides a quantitative measurement of collective motion in nematic biological systems where local interactions play a dominant role . We have shown that social interactions mediated by type IV pili , when coupled with active motion , have an alignment effect on neighboring cells and significantly facilitate swarming . Many pathogenic bacteria swarm within infected tissues and have type IV pili as virulence factors . It is likely that the ascent of Proteus mirabilis up the urinary tract is a result of growth and swarming with flagella [36] . Similarly , the spreading of Neisseria in infected tissue is related to swarming with type IV pili , since those pili are necessary for virulence [37] . The bacterial swarming model described in the paper may therefore shed light on the colonization and infection process of pathogens .
In the model , each cell is represented by a flexible string of N nodes ( Figure 3 ) consisting of ( N − 1 ) segments , each of length r . There are ( N − 2 ) angles θi between neighboring segments . For each cell , we define the following energy function ( Hamiltonian ) : The first term in Equation 7 is the stretching energy determined by the cells' length . The second term is a bending energy . Kb and Kθ are stretching and bending dimensionless coefficients , analogous to the spring constants in Hooke's Law . They determine the extent to which the segment length and angles can change in the presence of fluctuations , respectively . They are the same for all segments and angles . r0 is the target length of a segment . In our simulations , we choose the number of nodes N = 3 ( Figure 3 ) , so that r0 is 2 . 5 μm ( half-cell length ) . Kb and Kθ are set at 5 and 2 , respectively , based on experimental observation that cells do not change their length a lot , but can bend rather easily . Let's denote the dark cell in the center of Figure 4 as cell k . In the absence of cell–cell collisions , the velocity direction of cell k is determined by three contributions: a motility direction , orientation from slime trail , and orientation from type IV pili . A motility direction . The cells secrete slime ( polysaccaride ) from their tail end , which expands as it leaves the cell body and pushes the cell directly forward [12] . We model this motility by trying to orient the cell along its long axis , which is the tail-to-head direction . The corresponding term in formulae below is denoted as . Small deviations from the direction of long axis are observed [10] . This is modeled using a Monte Carlo reconfiguration algorithm . Orientation from slime trail . When a moving cell encounters a slime trail , it tends to turn through an acute angle to follow the trail . We define a 2-D slime-orientation vector field that records the slime trail orientation as a vector assigned to each position . This vector coincides with orientation of a cell that passed through most recently . We make a simplifying assumption of all orientation vectors having unit length . Once a slime trail is laid down at position , it will be cleared after the slime aging time Ts . Orientation from type IV pili . As discussed in an earlier section ( Model of cell behavior and social interactions ) , type IV pilus-mediated interactions are assumed to align neighboring cells . For a particular cell k , we average the orientations of its neighboring cells within the pilus-cell interacting area ( Figure 4 , area I ) , and define this averaged direction as the contribution of pilus-mediated interactions to the head velocity direction of cell k . This term is denoted as . Cell velocity direction . When there are no collisions between cell k and its neighbors , the direction of its head velocity ( denoted as ) is determined by the sum of A motility direction , orientation from slime trail , and orientation from type IV pili: a motility direction: orientation from type IV pili: In Equation 8 , C is a constant cell speed ( 4 μm/min ) ; α , β , and γ are parameters representing the relative strength of each motility term . Denominators are used in all equations for normalization . The experiments suggest that the forces generated from A and S motilities are nearly the same , approximately 150 pN [9 , 12] . For an A+S+ cell , we choose α = γ = 1 . 0 , and for an A+S− mutant , we choose α = 1 . 0 and γ = 0 . The strength of slime orientation effect is set as β = 0 . 5 . Note that the slime-orientation vector field is recorded in a discrete 2-D lattice , with each lattice site having a slime-orientation vector . Slime trails interact with the newly secreted slime , not with the head of a cell . We analyze slime-orientation vectors at the lattice sites covered by the front half of a cell body , and take the direction , which occurs most frequently , as slime-orientation direction to be followed by the cell . It is denoted as slime in Equation 8 . Equation 9 determines cell orientation , which is the direction from the tail node ( ) to the head node ( ) and which is considered the A motility direction . In Equation 10 , n denotes the total number of neighboring cells of the cell k . We multiply the expression by the factor of n because we think that type IV pili have a stronger effect on the direction of motion of the head node ( pili are located at the head of a cell ) , and that this effect depends on the number of neighboring cells . The terms cosθj and sinθj are the x and y components of the orientation vector of the j-th neighboring cell . These vector components are then averaged ( <cosθj> and <sinθj> ) and are taken as the x and y components of the average direction . and denote unit vectors along the x and y axes . We model the alignment in such a way that cells orient with their neighbors to the acute angle . That is , if the dot product of the tail-to-head directions of cell k and its j-th neighbor cell is negative , we choose the opposite direction to as its orientation . Therefore , we have: This approach is different from that taken by Vicsek et al . [33] . The alignment is determined through acute angles because we use cell orientations instead of velocity directions . Collision-resolving algorithm . When the head node of cell k collides with the body of cell j , this collision is resolved as follows: Calculate distances between the head node of cell k and two end nodes of cell j; If one of these distances is less then a cell width , choose at random a new direction such that the dot product of new direction of cell k and orientation of cell j is positive , and move; Otherwise , take the average direction of both cells k and j as the new direction and stall until next time step . ( The same method is used in the above alignment algorithm of type IV pilus-mediated interactions . ) Reversal of gliding direction . Each myxobacterial cell reverses its gliding direction every 10 min or so . ( Reversal periods of myxobacteria follow a distribution with an average of about 10 min . ) For simplicity , we choose the reversal periods in accordance with binomial distribution from 5 min to 15 min [38] . Each cell is assigned an inner reversal clock . The initial values of the clock are assigned at random . At each time step of a simulation , the clock value increases by a unit of time . Cell reverses when the clock reaches the value of the reversal period and the clock is reset to zero . Simulation setup . The simulation domain is chosen in the form of a rectangle 200 μm by 200 μm ( Figure 5 ) . In simulations , a unit length is equal to 0 . 166 μm and one time step is equal to 0 . 2 min so that the initial cell length ( 5 μm ) is equal to 30 units of length , and cell width ( 0 . 5 μm ) is 3 units . As mentioned in the text , we approximate that myxobacteria move at the constant speed of 4 μm/min so that in the simulations , a cell moves a distance of 5 units in each time step . Initially , cells are distributed within the “Initial Area of Cells” ( see Figure 5 ) . Cell centers are distributed at random , but cell orientations are distributed around the radial direction in accordance with the normalized distribution function f ( x ) with a peak at ( π / 2 ) : From experimental observations , it follows that a steady rate of swarm expansion is reached only when most cells behind the swarm edge orient themselves outward along the radial direction . Ideally , one would need to choose the initial orientation distribution f ( x ) according to the experimental data measured at the beginning of the steady swarming . However , due to the lack of such data , we select the initial orientation distribution function f ( x ) in such a form that most cells initially point outward from the swarming edge . Cell growth and division are included in our model as maintaining the average density in the simulation domain near the edge . Algorithm implementation . At each time step , we implement the following sequence of operations for each cell . First , check the inner reversal clock and decide whether to reverse polarity of the cell or not . Then , calculate the velocity direction of the head node according to the model for motility systems . If no collision occurs , move the head node at a distance of five units; otherwise , use the collision-resolving algorithm to resolve the collision . Then , apply Monte Carlo algorithm to reconfigure the positions of other nodes of the cell . Use the procedure suggested in [20] . After moving the head node to a new position , repeat the following operations for ( integer part of 2 . 5N ) number of steps ( N is the number of nodes per cell ) : ( i ) choose node i at random and move it in the direction from node i to node ( i − 1 ) at a distance of 5 unit lengths; ( ii ) calculate the energy change ΔE due to the relative position change of the nodes . Use the Metropolis algorithm [21] to determine the acceptance probability for the positional change of a node: Then , record slime-orientation vectors in the end of individual cell movement at all positions passed through by the cell . After all cells move , calculate the cell number flux through the boundary into the free space and add the same number of cells into the initial area to keep the cell number in the “Initial Area of Cells” constant . Table 1 provides values of modeling parameters . Our model depends on two parameters characterizing properties of the slime trail: the slime aging time Ts and the relative strength of slime guidance . In this section , we describe simulation results for different ranges of these parameters to test the robustness of the model . The slime aging time ( Ts ) is defined as the lifetime of a slime trail during which it has the ability to guide the motion of a bacteria . We used a value of 20 min in our simulations . In Figure 12 , we simulate the swarming of wild-type cells and A+S− mutant at the density of 50 K-S units ( the same simulation setup as in Figure 6 ) , and varied Ts from 10 min to 200 min ( the whole time span of the swarming simulations ) . We make linear fits for the data points and find that the value of Ts has little effect on simulation results ( the cell number flux ) . This is because slime guidance is primarily a local effect , and slime trails will be washed out by other cells' slimes at short times when the cell density is high . Therefore , the parameter Ts is quite robust for the results in Figure 7B , which is the main validation of our model . The relative strength of slime guidance is modeled by the parameter β in Equation 8 . We used a value of 0 . 5 in the simulations ( see Methods ) . Here , we varied β from 0 to 1 . 5 , and calculate the cell number flux in swarms of wild-type cell and A+S− mutant type at the density of 50 K-S units ( the same simulation setup as in Figure 6 ) . The simulation data are plotted in Figure 13 along with the linear fits . We find that as the slime guidance effect gets stronger , the cell number flux increases . It increases slightly faster in the case of A+S− mutants than in the case of wild-type cells , with the slopes being 0 . 33 and 0 . 19 for A+S− mutant and wild-type cells , respectively . This result suggests that as the effect of slime guidance gets stronger , the local alignment of cells and the order of collective motion are both increased . However , this does not affect the results in Figure 7B much , since the increase of cell number flux in the case of A+S− mutants is only slightly faster than that in the case of wild-type cells . The ratio of two fitting functions remains greater than 2-fold until β = 18 . 5 . This demonstrates robustness of our model with respect to the relative slime strength ( see Figure 7B ) . | Many bacteria are able to spread rapidly over the surface using a strategy called swarming . When the cells cover a surface at high density and compete with each other for nutrients , swarming permits them to maintain rapid growth at the swarm edge . Swarming with flagella has been investigated for many years , and much has been learned about its regulation . Nevertheless , its choreography , which is somewhat related to the counterflow of pedestrians on a city sidewalk , has remained elusive . It is the bacterial equivalent of dancing toward the exit in a crowd of moving bodies that usually are in close contact . Myxococcus xanthus expands its swarms at 1 . 6 μm/min , about a third the speed of individual cells gliding over the same surface . Each cell has pilus engines at its front end and slime secretion engines at its rear . Using the known mechanics of these engines and the ways they are coordinated , we have developed a cell-based model to study the role of social interactions in bacterial swarming . The model is able to quantify the contributions of individual motility engines to swarming . It also shows that microscopic social interactions help to form the ordered collective motion observed in swarms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"infectious",
"diseases",
"developmental",
"biology",
"eubacteria",
"computational",
"biology"
] | 2007 | Social Interactions in Myxobacterial Swarming |
Complex phenotypes such as the transformation of a normal population of cells into cancerous tissue result from a series of molecular triggers gone awry . We describe a method that searches for a genetic network consistent with expression changes observed under the knock-down of a set of genes that share a common role in the cell , such as a disease phenotype . The method extends the Nested Effects Model of Markowetz et al . ( 2005 ) by using a probabilistic factor graph to search for a network representing interactions among these silenced genes . The method also expands the network by attaching new genes at specific downstream points , providing candidates for subsequent perturbations to further characterize the pathway . We investigated an extension provided by the factor graph approach in which the model distinguishes between inhibitory and stimulatory interactions . We found that the extension yielded significant improvements in recovering the structure of simulated and Saccharomyces cerevisae networks . We applied the approach to discover a signaling network among genes involved in a human colon cancer cell invasiveness pathway . The method predicts several genes with new roles in the invasiveness process . We knocked down two genes identified by our approach and found that both knock-downs produce loss of invasive potential in a colon cancer cell line . Nested effects models may be a powerful tool for inferring regulatory connections and genes that operate in normal and disease-related processes .
Carcinogenesis involves a host of cell-cell communication breakdowns that include the loss of contact inhibition , an increased potential to proliferate , and the ability to invade and spread into foreign tissue [1] . The molecular events involved in this transformation are still poorly understood . New systematic methods are needed to infer the key events responsible for these disease processes . The ability to measure gene expression changes for the entire genome in the presence of molecular perturbations , such as specific gene knock-downs , provides a new opportunity to infer gene networks in a data-driven manner . Our goal is to identify the genetic mechanisms underlying a phenotype , such as cancer cell deregulation . We take a network-based approach to the problem , starting with a set of signaling genes or S-genes , known to act in a common pathway . The input to the method is a matrix in which gene expression has been measured under the knock-down of each of the S-genes . Genes exhibiting differential expression across the knock-downs , here referred to as effect genes or E-genes , are used to predict a set of interactions among the S-genes , and expand the pathway by identifying newly implicated frontier genes based on their expression changes . We hypothesize that using a structured model of the interactions among the S-genes will improve the identification of frontier genes for inclusion in the network for subsequent rounds of investigation . Previous approaches for pathway expansion have used methods based on expression correlations to a phenotype of interest . These methods search for genes with expression profiles that are highly correlated with a particular phenotype or disease state and have led to promising results [2]–[5] . Methods using Analysis of Variance [6] , false-discovery [7] , and non-parametric methods [8] also have been proposed . For example , one method is to measure the correlation of gene expression levels with an idealized vector representing the phenotype ( e . g . indicator variables with zeroes for disease and ones for lack of disease ) [9] . One disadvantage of these methods is that they make no explicit use of the known members of a pathway or how these members interact with each other . More recently , several approaches have demonstrated learning a structured model from perturbation experiments [10]–[13] . Approaches based on Bayesian Networks have also been proposed [11] , [12] . However , these approaches attempt to identify networks over the E-genes rather than the S-genes and therefore require many replicated microarray experiments to distinguish signal from noise . Instead , perturbing genes of interest and constructing networks from observations of downstream changes allows powerful interventional reasoning , as well as reconstruction of interactions not directly reflected in expression levels , such as phosphorylation . In one approach , Carter et al . ( 2007 ) [14] decompose the matrix of expression changes under single- and double-gene deletions to infer a transcriptional regulation network from which phenotypes and gene expression responses following knock-downs can be predicted . An alternative approach is the Nested Effects Model ( NEM ) of Markowetz et al . ( 2005 , 2007 ) [10] , [15] , which has been used to predict interactions , including non-transcriptional interactions . Rather than searching for genetic networks that explain observational data , as several Bayesian Network approaches have done [11] , [16] , NEMs are useful in situations in which perturbations have been carried out on a focused set of genes . In this case , NEMs assume the interest is in a finer description of the interactions among the silenced genes rather than identifying a network of unrestricted connections between potentially additional genes . The NEM approach takes as input a matrix of expression changes , X . A column of X corresponds to a single gene knock-down ( or knock-out ) of one S-gene; a row corresponds to the response of an E-gene to all of the knock-downs . The method searches for approximate subset relations among the expression changes of the E-genes to organize the S-genes into a network . To do this it assumes , for example , that S-gene A is above S-gene B if the set of E-genes that change under gene A's knock-down are an approximate superset of the effected genes under B's knock-down . The current NEM approach uses binary set membership relations to identify a network and thus the exact nature of interaction between S-genes ( e . g . activation or inhibition ) is not determined . However , an appreciable extent of inhibition occurs in real genetic networks . To estimate the amount of inhibition present in living cells , we estimated the proportion of genes up-regulated in deletion mutants relative to wild-type from a yeast knock-out compendium [17] . Over half of the genes had increasing expression changes across the deletion strains , consistent with a high degree of inhibitory interactions in the yeast genetic network ( see Figure S1 ) . Thus , the inability to distinguish between stimulatory and inhibitory interactions may be a critical shortcoming of current NEM approaches . To address this limitation , we developed a generalization of the NEM approach using a probabilistic graphical model called a factor graph that allows a broader set of S-gene interactions to be recovered from the secondary effects of E-gene expression . This paper offers three methodological contributions . First , we present a factor graph formulation called FG-NEM that allows for an efficient search over all possible NEM structures for a high-scoring model . Second , we show how FG-NEMs extend the NEM approach for expanding the network beyond the current set of S-genes . Third , we show that FG-NEMs can model a more general class of S-gene interactions than NEMs , which increases the accuracy of network identification over an approach that considers a more restricted set of interactions . We demonstrate the usefulness of FG-NEMs on both simulated and biologically relevant signaling networks that contain both inhibition and activation . We apply FG-NEMs to identify novel genes not previously implicated in colon cancer cell invasiveness . Finally , we experimentally test FG-NEM predictions and report that knock-downs of the top-scoring genes lead to a loss-of-invasion phenotype , validating the approach . Source code is available as an R library from our website: http://sysbio . soe . ucsc . edu/projects/fgnem .
Our goal is to automatically identify genetic interactions among a set of signaling genes from gene expression changes observed under their knock-down . The signaling genes represent a set of genes that prior experimental evidence suggests participate in a common pathway . To infer a network , we use an extension of the Nested Effect Model ( NEM ) introduced by Markowetz et al . ( 2005 ) [10] . The set of silenced genes are denoted as the set S ( or S-genes ) . An NEM is a probabilistic formulation that measures how well a directed graph of the S-genes is consistent with expression changes collected under the separate silencing of each S-gene ( i . e . only single knock-downs are considered in NEM ) . While the method can make use of either complete deletion mutants or genes that may be partially silenced , here we use the term knock-down to refer to either case . We denote the knock-down of S-gene A as ΔA . We also refer to a set of effect genes as the set E ( or E-genes ) , for which gene expression data is available . The expression of an E-gene e is assumed to be influenced by at most one S-gene . The key assumption of NEMs is the expression changes observed under ΔA are an approximate superset of the changes observed under ΔB if gene A acts upstream of gene B in a pathway . We use the shorthand A>B to represent this generic directed interaction . In addition to identifying A>B , the E-gene expression changes on the microarray can be used to infer the “sign” of the interaction , either activating or inhibiting . In our framework , we extend the interactions so that an upstream gene can have either an inhibitory or stimulatory effect on downstream genes . Figure 1A presents an example , similar to Fröhlich et al . ( 2008 ) [18] that motivates the use of signed interactions . E-genes E1 through E13 are listed from top to bottom according to where they are attached to the network . Depending on the connections of the S-genes to one another and to the E-genes , a disruption in an S-gene will cause E-genes to either increase or decrease in expression relative to wild-type . For example , E-gene E7 decreases under ΔB relative to wild-type because the wild-type activation by B is absent in the deletion . On the other hand , the expression of E10 also decreases under ΔB relative to wild-type but as a result of a different mechanism . In wild-type , E10 is expressed at a baseline level because its repressor , the product of gene D , is inhibited by B's product . However , in the B deletion , D is derepressed , leading to inhibition of E10 . This toy example illustrates that the disambiguation of inhibition and activation , both for S-gene interactions and E-gene attachments , make it possible to account for an expanded set of mechanisms leading to the observed expression changes . The E-gene expression changes are available in a data matrix X where each column gives the difference in expression of each E-gene under the deletion of a single S-gene relative to wild-type . X may also contain replicates in the form of repeated S-gene knock-downs . The entry XeAr represents e's expression change under the rth replicate of ΔA . Furthermore , we assume that an unknown expression “state” for each E-gene under each knock-down , determines its set of expression changes observed across the {XeAr} replicates in the microarray data . The matrix , Y , records a hidden state for each E-gene under each knock-down , where entry YeA is the state of E-gene e under ΔA . We allow the states to be ternary-valued {+1 , −1 , 0} representing whether e is up-regulated , down-regulated , or unchanged under ΔA relative to wild-type respectively . Nested effects models include two sets of parameters . The parameter set Φ records all pair-wise interactions among the S-genes and the parameter set Θ describes how each E-gene is attached to the network of S-genes . In the original NEM formulations [10] , [15] , [18] Φ is a binary matrix with entry φAB set to one if S-gene A acts above S-gene B and zero otherwise . If φAB = φBA = 1 then the S-genes are assumed to operate at an equivalent position in the pathway . Note that indirect interactions are also represented in Φ so that if φAB = 1 and φBC = 1 it implies φAC = 1 . A parsimonious network among the S-genes is solved for by computing the transitive reduction of Φ . To allow for both stimulatory and inhibitory interactions in our formulation , φAB can assume six possible values for each unique unordered S-gene pair {A , B} . We refer to these values as interaction modes . The possible values are: ( i ) A activates B , A→B; ( ii ) A inhibits B , A⊣B; ( iii ) A is equivalent to B , A = B; ( iv ) A does not interact with B , A≠B; ( v ) B activates A , B→A; and ( vi ) B inhibits A , B⊣A . Plotting the response of E-genes under ΔA and ΔB yields a scatter-plot that may provide a signature for the type of interaction between A and B . For example , Figure 1B shows a scatter-plot of gene expression changes from the Hughes et al . ( 2000 ) yeast knock-out compendium [17] for a pair of knock-outs of the well-known pheromone-response genes: ΔSTE12 and the ΔDIG1/DIG2 double knock-out . Comparing the scatter-plot for these pheromone-response genes to the patterns in Figure 1C , it can be seen to match the inhibitory interaction mode more closely than the other modes , which is consistent with DIG1/DIG2's known inhibition of STE12 . Figure 1C shows an example of the first four modes . Shaded regions denote consistent E-gene responses for each mode . An interaction mode determines a constraint on the observed E-gene expression changes . For example , plotting the expression changes of E-genes that act downstream of either A or B for the generic A>B interaction mode produces points in one of the seven shaded regions shown in Figure 1Cv . Figure 1Cii shows an example where the inhibitory interaction mode is the best match to the data because a higher number of E-gene changes fall within consistent regions ( filled circles in the figure ) . In this manner , genomewide expression changes detected on the microarrays can be used as quantitative phenotypes to identify a variety of interactions between pairs of S-genes . Note that two genes are equivalent if their knock-downs lead to significantly similar expression changes , which may predict , for example , that they form a complex . Figure 1C also illustrates the generic interaction mode A>B used in an unsigned version of our method . We compare FG-NEM results to two unsigned variants to estimate the change in predictive power as a function of the introduction of sign . In effect , both variants consider four interaction modes: ( i ) A>B; ( ii ) B>A; ( iii ) A≠B; and A = B . For comparison purposes , a predicted unsigned interaction was treated as activation . In the FG-NEM AVT variant , FG-NEM is run on the absolute value of the data . In the uFG-NEM method , we remove the component of FG-NEM which models induced expression , resulting in interaction modes where the top and right five regions are disallowed in all interaction modes . Our goal is to find a structure among the S-genes that provides a compact description of X . To find a network that best “fits” the data , we take a maximum a posteriori approach as in [15] , [18] jointly identify Φ and Θ that maximize the posterior: ( 1 ) ( 2 ) where we introduce the hidden E-gene states by summing over all possible configurations of the Y matrix . Applying Bayes' Rule and dropping P ( X ) , which is constant with respect to the maximization , we obtain: ( 3 ) ( 4 ) The approximation in the last step uses the assumption that any E-gene attachments are equally likely given a network structure; i . e . P ( Θ|Φ ) is assumed to be uniformly distributed and is ignored in our approach . P ( Φ ) represents a prior over S-gene networks . As in previous NEM formulations , we assume that each E-gene is attached to a single S-gene and that each E-gene observation vector across the knock-downs is independent of other E-gene observations . The maximization function can then be written: ( 5 ) ( 6 ) ( 7 ) where Xe and Ye are the row vectors of data and hidden states for E-gene e respectively , and θe records the attachment point information for E-gene e . After rearranging the products and sums , we introduce the shorthand Le to represent the likelihood of the data restricted only to E-gene e . Previous approaches decompose Le over the knock-downs , which assume the S-gene observations are independent given the network and attachments ( see [18] for an example of such a derivation ) . To facilitate scoring the expanded set of interaction modes mentioned earlier , we replace Le with a function proportional to Le , Le′ . Le′ is defined as a product of pair-wise S-gene terms: ( 8 ) where θeAB represents the attachment of E-gene e relative to the pair of S-genes A and B . Note that both θeAB and φAB are indexed by the unordered pair , {A , B} , so that φAB and φBA are references for the same variable . We refer to θeAB as e's local attachment which can take on five possible values from the set {A , −A , B , −B , 0} representing that e is either up- or down-regulated by A , attached and either up- or down-regulated by B , or not affected by either S-gene . φAB defines the mode of interaction between S-genes A and B . Assuming the replicates are independent given the E-gene states , P ( XeA | YeA ) can be written as a product over replicate terms: , where P ( XeAr | YeA ) is modeled with a Gaussian distribution having mean and standard deviation σ estimated from the data ( see Text S1 ) . Substituting Le′ for Le into Eq . ( 7 ) and distributing the maximization over attachment points , we obtain the maximizing function used in our approach: ( 9 ) The interaction factors P ( YeA , YeB | φAB , θeAB ) have a value of one if the E-gene e is attached to either A or B and e's state is consistent with the interaction mode between A and B . If e's state is inconsistent with the interaction and attachment , then the factor has value zero . While we used hard constraints to model consistent and inconsistent expression changes ( corresponding to the rigid boundaries of the regions drawn in Figure 1C ) , such constraints could be softened to use factors with belief potentials between zero and one . Note that , to simplify the example , the interaction modes in Figure 1C show defined regions . However , P ( XeA | YeA ) is modeled as a Gaussian distribution and therefore assigns non-zero probabilities over all possible expression values rather than classifying some as allowed and others disallowed ( i . e . probability zero ) . The prior over interactions , P ( Φ ) , can represent preferences over specific interactions in the S-gene graph , allowing the incorporation of biologically-motivated constraints to guide network search . For example , the interaction priors for genes in a common pathway or genes whose products have been detected to interact in protein-protein interaction screens could be set higher than the priors for arbitrary pairs of S-genes . In this study , we chose to test the approach both with and without external biological information . Without external biological information , the prior encodes a basic property of the S-gene graph: that it should exhibit transitivity to force pair-wise interaction modes to be consistent among all triples . Using transitivity , all paths between any two genes , A and B , are guaranteed to have the same overall effect; i . e . the product of the signs of individual links along different paths between A and B are equal . In order to preserve the transitivity of identified interaction modes , the prior is decomposed over interaction configurations into transitivity constraints on all triples of S-genes; i . e . : ( 10 ) where τ is zero if the triple of interactions are intransitive , and one if the interactions are transitive ( see Text S1 for full definition ) . Using transitivity constraints forces the search to find consistent models that best explain the observed changes . The transitivity constraint includes both the direction of interactions and the sign of interactions . As S-gene interactions are signed , the transitivity constraint forces the sign of the product of two edges to equal the sign of the third; e . g . if A⊣B and B⊣C , then A→C . A result of modeling transitivity is that a directed cycle of stimulatory interactions will also imply activation between any pair of S-genes in the cycle , in both directions . Therefore , the method clusters such S-genes into equivalence interactions . The product over ρ factors in Eq . ( 10 ) encode evidence from high-throughput assays , such as protein-protein binding and protein-DNA binding interactions ( see “Physical Structure Priors” in Text S1 ) . While network structures are constrained to reflect more intuitive models , the decomposition introduces interdependencies among the interactions , adding complexity to the search for high-scoring networks . Importantly , max-sum message passing in a factor graph [19] provides an efficient means for estimating highly probable S-gene configurations . We next describe how the problem is recoded into message-passing on a factor graph . The formulation above provides a definition of the objective function to be maximized but says nothing about how to search for a good network . The search space of networks is very large making exhaustive search [10] intractable for networks larger than five S-genes . To apply the method to larger networks , we require a fast , heuristic approach . Markowetz et al . ( 2007 ) introduced a bottom-up technique to infer an S-gene graph . They identify sub-graphs of S-genes ( pairs and triples ) and then merge the sub-graphs together into a final parsimonious graph . Fröhlich et al . ( 2008 ) [18] use hierarchical clustering to first identify modules , subsets of S-genes with correlated expression changes . Networks among the modules are exhaustively searched and a final network is identified by greedily introducing interactions across modules that increase the likelihood . Here , we introduce the use of a graphical model called a factor graph to represent all possible NEM structures simultaneously . The parameters that determine the S-gene interactions , Φ , are explicitly represented as variables in the factor graph . Identifying a high-scoring S-gene network is therefore converted to the task of identifying likely assignments of the Φ variables in the factor graph . A factor graph is a probabilistic graphical model whose likelihood function can be factorized into smaller terms ( factors ) representing local constraints or valuations on a set of random variables . Other graphical models , such as Bayesian networks and Markov random fields , have straightforward factor graph analogs . A factor graph can be represented as an undirected , bi-partite graph with two types of nodes: variables and factors . A variable is adjacent to a factor if the variable appears as an argument of the factor . Factor graphs generalize probability mass functions as the joint likelihood function requires no normalization and the factors need not be conditional probabilities . Each factor encodes a local constraint pertaining to a few variables . Figure 2 shows the factor graph representing the NEM for the example S-gene network from Figure 1A . Each random variable is represented by a circle and each conditional probability term in Eqs . ( 9–10 ) is represented by a square . The factor graph contains three types of variables . First , every unique unordered pair of S-genes {A , B} has a corresponding variable , φAB , that takes on values equal to one of the previously mentioned interaction modes ( Figure 2 , “S-Gene Interactions” level ) . Second , every E-gene-S-gene pair is associated with a variable , YeA for the hidden expression state of effect gene e under knock-down A , ( Figure 2 , “E-gene Expression State” level ) . Third , every observed expression value is associated with a continuous variable , XeAr , where r indexes over replications of ΔA ( Figure 2 , “E-gene Expression Observation” level ) . Figure 2 also shows the expression factors , interaction factors , and transitivity factors of Eqs . ( 9–10 ) . To validate the involvement of predicted invasiveness frontier genes , HT29 colon cancer cells were resuspended in DMEM medium containing 0 . 1% FBS and seeded into the top wells ( 2×105 per well ) containing individual Matrigel inserts ( BD Biosciences , San Jose , CA ) according to manufacturer's protocol . The lower wells were filled with 800 µl medium with 10% fetal bovine serum as chemoattractant . Six to ten hours following seeding , the cells in the upper wells were transfected with the appropriate shRNA-expressing pSuper constructs [25] using Lipofectamine 2000 ( Invitrogen , Carlsbad , CA ) . Final concentration of pSuper constructs was 1 . 6 µg/ml . The transfected cells were incubated at 37°C for 48 hours before assaying for invasion . Media was aspirated from the top wells and non-invading cells were scraped from the upper side of the inserts with a cotton swab and invading cells on the lower side were fixed and stained using DiffQuick ( IMEB , Inc . San Marcos , CA ) . Total number of invading cells was counted for each insert using a light microscope . Invasion was assessed in quadruplicate and independently repeated at least five times . The shRNA-expressing portion of the construct was designed using the siRNA Selection Program of the Whitehead Institute for Biomedical Research ( http://jura . wi . mit . edu/bioc/siRNAext/ ) , synthesized by Invitrogen and subcloned into the XhoI and BamHI sites of pSuper plasmid . Sequences for shRNA constructs are available in the Text S1 . shRNA construct MYO1G targets the myosin 1G mRNA ( GenBank accession number NM _033054 ) . shRNA construct BMPR1A targets the bone morphogenetic protein receptor , type IA mRNA ( NM_004329 ) . shRNA construct COLEC12 targets the collectin sub-family member 12 mRNA ( NM_130386 ) . shRNA construct AA099748 targets an expressed sequence tag mRNA ( AA099748 ) . shRNA construct CAPN12 targets the calpain 12 mRNA ( NM_144691 ) . shRNA construct scrambled serves as a nonsense sequence negative control .
We hypothesized that an estimate of genetic pathway structure based on modeling observed expression changes could facilitate the identification of new pathway members . To test this , we evaluated the ability of FG-NEMs , uFG-NEMs , and TM to identify genes involved in a diverse set of pathways in S . cerevisae using the well-studied gene expression dataset from the Hughes et al . ( 2000 ) knock-out compendium elucidated by Rosetta [17] . This compendium contains whole-genome expression profiles of 276 yeast gene-deletion mutants and P values for differential gene expression . We applied the FG-NEM approach to a human colon cancer invasiveness network elucidated by Irby et al . ( 2005 ) [26] . In this work , the authors identified several “tiers” of genes implicated in the invasion process under the control of SRC kinase . Genes were included in a tier if their knock-downs were found to produce a significant drop in the invasive potential of HT29 colon cancer cells as defined by invasion through Matrigel . To identify additional genes involved in the invasion process , the authors measured gene expression under an RNA interference knock-down of each gene in the tier . Genes whose expression was lower in the knock-downs producing loss-of-invasiveness , and higher in knock-downs that did not produce loss-of-invasiveness , were considered candidates for inclusion in the next tier . In this fashion , each tier was formed by knocking-down each candidate gene and assaying for loss-of-invasion in Matrigel .
We applied FG-NEMs to discover a human signaling network among genes involved in colon cancer cell invasiveness . The method formalizes and extends analysis of genetic interactions using high-dimensional quantitative phenotype data in the form of gene expression changes observed under specific perturbations . It makes explicit use of the knock-downs of known members of a pathway to identify how the members interact with one another and for identifying new members . The method predicts several genes with new roles in the cancer invasiveness process , two of which were verified to act in the pathway based on an ex vivo invasion assay . Thus , the FG-NEM approach may be a powerful tool for inferring regulatory connections and for identifying new partners of genes known to operate in a process of interest . The application of structured causal models for pathway identification and expansion promises to greatly accelerate the discovery of genetic pathways from genetic knock-downs and other intervention-based experiments . | Biological processes are the result of the actions and interactions of many genes and the proteins that they encode . Our knowledge of interactions for many biological processes is limited , especially for cancer where genomic alterations may create entirely novel pathways not present in normal tissue . Perturbing gene expression ( for example , by deleting a gene ) has long been used as a tool in molecular biology to elucidate interactions but is very expensive and labor intensive . The search for new genes that may participate can be a daunting “fishing expedition . ” We have devised a tool that automatically infers interactions using high-throughput gene expression data . When a gene is silenced , it causes other genes to be switched on or off , which provide clues about the pathway ( s ) in which the gene acts . Our method uses the genomewide on/off states as a fingerprint to detect interactions among a set of silenced genes . We were able to elucidate a network of interactions for several genes implicated in metastatic colon cancer . Genes newly connected to the network were found to operate in cancer cell invasion in human cells , validating the approach . Thus , the method enables an efficient discovery of the networks that underlie biological processes such as carcinogenesis . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"computational",
"biology/systems",
"biology",
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/functional",
"genomics",
"genetics",
"and",
"genomics/cancer",
"genetics",
"genetics",
"and",
"genomics/bioinformatics"
] | 2009 | A Factor Graph Nested Effects Model To Identify Networks from Genetic Perturbations |
Parasitic infections are still of considerable public health relevance , notably among children in low- and middle-income countries . Measures to assess the magnitude of ill-health in infected individuals , however , are debated and patient-based proxies through generic health-related quality of life ( HrQoL ) instruments are among the proposed strategies . Disability estimates based on HrQoL are still scarce and conflicting , and hence , there is a need to strengthen the current evidence-base . Between November 2011 and February 2012 , a national school-based cross-sectional survey was conducted in Côte d'Ivoire . Children underwent parasitological and clinical examination to assess infection status with Plasmodium and helminth species and clinical parameters , and responded to a questionnaire interview incorporating sociodemographic characteristics , self-reported morbidity , and HrQoL . Validity analysis of the HrQoL instrument was performed , assessing floor and ceiling effects , internal consistency , and correlation with morbidity scores . Multivariate regression models were applied to identify significant associations between HrQoL and children's parasitic infection and clinical status . Parasitological examination of 4 , 848 children aged 5–16 years revealed Plasmodium spp . , hookworm , Schistosoma haematobium , Schistosoma mansoni , Ascaris lumbricoides , and Trichuris trichiura prevalences of 75 . 0% , 17 . 2% , 5 . 7% , 3 . 7% , 1 . 8% , and 1 . 3% , respectively . Anemic children showed a significant 1-point reduction in self-rated HrQoL on a scale from 0 to 100 , whereas no significant negative association between HrQoL and parasite infection was observed . The 12-item HrQoL questionnaire proofed useful , as floor and ceiling effects were negligible , internally consistent ( Cronbach's alpha = 0 . 71 ) , and valid , as revealed by significant negative correlations and associations with children's self-reported and clinically assessed morbidity . Our results suggest that HrQoL tools are not sufficiently sensitive to assess subtle morbidities due to parasitic infection in Ivorian school-aged children . However , more advanced morbid sequelae ( e . g . , anemia ) , were measurable by the instrument's health construct . Further investigations on health impacts of parasitic infection among school-aged children and refinement of generic HrQoL questionnaires are warranted .
Malaria and the neglected tropical diseases ( NTDs ) are still of considerable public health relevance in the tropics and subtropics and their successful control is a key issue toward progress of the millennium development goals ( MDGs ) and the post-2015 agenda of sustainable development [1]–[4] . Preschool-aged children are considered at highest risk of malaria , whereas school-aged children are the most affected by parasitic worm infections ( helminthiases ) [5]–[7] . The assessment of the precise burden attributable to parasitic infections , however , is a difficult issue and there is ongoing discussion and debate [8] , [9] . Over the past 20 years , the magnitude of health loss due to diseases , injuries , and risk factors has been increasingly expressed in disability-adjusted life years ( DALYs ) . This metric is a combined measure of premature death and years of life lived with disability . For measuring the burden of helminthiases and other NTDs , specific disability weights ( DWs ) of morbid sequelae are considered and , by convention , scaled on an axis from 0 ( no health loss ) to 1 ( health loss equivalent to death ) [10] . Former estimates were often criticized for underestimating the true burden of infectious diseases , due to separating out morbidity ( e . g . , anemia ) , although such morbidity is partially associated with infection ( e . g . , hookworm and Plasmodium ) . Additionally , cultural and socioeconomic contexts are insufficiently taken into account , and DWs were usually based on expert opinion; thus , ignoring community- or patient-based appraisal [11]–[13] . Meanwhile , the Global Burden of Disease ( GBD ) consortium presented estimates for the year 2010 by incorporating different sequelae to capture direct consequences of infections and judgments on health losses from the general public in culturally and socioeconomic diverse settings [14] , [15] . Nonetheless , the use of generic health status measurement instruments to expand the GBD approach has been discussed by the lead authors of the GBD 2010 study [15] , thus partially addressing concerns that have been articulated a decade ago [16] , [17] . The discussed generic health status measurement instruments evaluate health burden in a comprehensive way based on health-related quality of life ( HrQoL ) and typically include domains on physical , mental , and social wellbeing , and a visual analogue scale ( VAS ) for subjective health rating [18]–[20] . Thus far only few studies have assessed HrQoL and derived DWs in individuals with parasitic diseases , indicating the early stage of this approach in the field of parasitology . This issue is further underscored by conflicting results; while negative associations between HrQoL measures and Trichuris trichiura , Schistosoma mansoni , Schistosoma haematobium , and advanced Schistosoma japonicum infections were observed [21]–[23] , other studies failed to show significant differences in HrQoL and DWs between infected children and their non-infected counterparts [24]–[26] . A weaker explanatory power in previous studies may partly be explained by a lack of cross-cultural validity of the questionnaires . HrQoL instruments have been developed and broadly validated in Europe and the United States of America and were originally designed for adult respondents . Child-friendly versions meanwhile exist [27] , [28] , but application in different cultural settings imply careful adaptations in language and scoring , thorough pre-testing , and validity analysis . Considering the scarcity of empirical data on HrQoL assessments in school-aged children with single and multiple species infections , the aim of the present study is to strengthen the current evidence-base of disability due to parasitic diseases among pupils in Côte d'Ivoire . Hence , a cross-sectional school-based survey was carried out using standardized , quality-controlled parasitological and questionnaire tools . Furthermore , we discuss the utility and validity of a HrQoL questionnaire tailored to a given setting with basic elements from readily available tools .
The study protocol was approved by the institutional research commissions of the Swiss Tropical and Public Health Institute ( Basel , Switzerland ) and the Centre Suisse de Recherches Scientifiques en Côte d'Ivoire ( Abidjan , Côte d'Ivoire ) . Ethical approval was obtained from the ethics committees in Basel ( EKBB; reference no . 30/11 ) and Côte d'Ivoire ( CNER; reference no . 09-2011/MSHP/CNER-P ) . Additionally , permission to carry out the study was sought from the Ministry of National Education in Côte d'Ivoire . Directors and teachers of the selected schools , district and local health and education authorities were informed about the purpose and procedures of the study . Written informed consent was obtained from parents and legal guardians of children , whilst children assented orally . Participation was voluntary , and hence , children could withdraw from the study at any time without further obligations . All collected data were coded and kept confidential . Participating children benefited from free of charge deworming with albendazole ( single oral dose of 400 mg ) . Children identified to harbor Schistosoma spp . were given praziquantel ( single oral dose of 40 mg/kg ) . In schools where the prevalence of Schistosoma infection was above 25% , the entire study sample was treated with praziquantel . Symptomatic malaria cases , defined as having a positive rapid diagnostic test ( RDT ) and fever , were offered artemisinin-based combination therapy ( ACT; using artesunate-amodiaquine ) and paracetamol against fever . Between November 2011 and February 2012 ( i . e . , dry season ) we conducted a national cross-sectional , school-based study , including parasitological and clinical examinations , and administered a questionnaire . Our aim was to select approximately 100 schools across Côte d'Ivoire , which we considered as a maximum number of locations that we would be able to visit within a 3-month period and our financial and human resources would allow to cover . A lattice plus close pairs design [29] , [30] was applied for the sampling of the schools . In brief , a grid indicating latitude and longitude at a unit of 0 . 5° was overlaid on a map of Côte d'Ivoire that divides the country into two major ecological zones [31] . The southern ecozone is characterized by abundant rainfall ( >1 , 000 mm per annum ) and dense forest vegetation cover , whereas the northern ecozone corresponds to a savannah-type profile with markedly less precipitation . In order to achieve a representative sample of the country , 58 and 42 possible survey locations were retained after randomly drawing from each or every second grid cell of ecozones 1 and 2 , respectively , taking into account population density from the last available census in 1998 . About 27% of the population was estimated to live in the major urban centers in 2007 [32] . We aimed at including at least one fifth of all schools from urban areas . In total , 94 schools were selected and we double-checked that the schools comprised a minimum of 60 children attending grades 3 and 4 , using a recent school inventory from a national UNICEF education program ( UNICEF 2010; personal communication ) . Children attending grades 3 and 4 were considered as capable to express themselves and give reliable answers to questionnaire items on household assets , experienced symptoms and diseases , and HrQoL and may be retrievable in case of followed-up studies . The sample size per school was delimited to 60 children due to financial and operational constraints , considering the high number of schools to be surveyed and the maximum number of children that a survey team could sample in a single day , including questionnaire interviews and detailed laboratory work-up of blood , stool , and urine specimens . This sample size exceeds the minimum of 50 children to be surveyed in a school , as recommended by the World Health Organization ( WHO ) for collection of baseline information on helminth prevalence and intensity in the school-aged population within large-scale surveys [7] . Two schools were omitted in the final analysis . One school refused to participate , while another school was subjected to recent deworming . The latter would have biased the results , since signs and symptoms due to chronic helminth infections and HrQoL are likely to change after anthelmintic treatment . The remaining 92 schools are mapped by ecozone , and stratified by rural and urban setting characteristics ( Figure 1 ) . In advance of the study conduct , directors and teachers of the selected schools were contacted and they were invited to inform parents or legal guardians of 60 children attending grades 3 and 4 . Whenever necessary , children from grade 5 were invited to complement sampling to reach the targeted number of 60 children . Children whose parents/guardians had provided written informed consent were invited for participation . The objectives and procedures of the study were explained on the day of the visit . Children were then asked to provide fresh urine and stool samples in plastic containers distributed upon arrival at school . Additionally , a finger-prick blood sample was taken for preparation of an RDT of malaria ( ICT ML01 malaria Pf kit; ICT Diagnostics , Cape Town , South Africa ) and thick and thin blood films on microscope slides for subsequent analysis of Plasmodium infection . All biological samples were transferred to nearby laboratories and processed the same day . In brief , urine reagent strips ( Hemastix; Siemens Healthcare Diagnostics GmbH , Eschborn , Germany ) were used to assess microhematuria in urine samples , as a proxy for S . haematobium infection [33] . Of note , reagent strips show a high specificity for indirect diagnosis of S . haematobium among school-aged children in endemic areas [34] . Duplicate Kato-Katz thick smears [35] , using 41 . 7 mg templates , were prepared from each stool sample . Kato-Katz thick smears were allowed to clear for 30–45 min prior to microscopic examination by experienced laboratory technicians . The number of helminth eggs was counted and recorded for each species separately ( i . e . , S . mansoni , A . lumbricoides , T . trichiura , hookworm , and other helminths ) . Blood films were stained with a 10% Giemsa solution and examined under a microscope for Plasmodium species identification and quantification of parasitemia ( parasites/µl of blood ) [36] . For quality control , 10% of the Kato-Katz thick smears and stained blood film slides were re-examined by a senior microscopist . In case of discrepancies ( e . g . , positive versus negative results or counts of parasitic elements differing by more than 10% ) , slides were read by a third technician and findings discussed until agreement was achieved . All participating children underwent a clinical examination , conducted by experienced medical staff , which included hemoglobin ( Hb ) measurement using a HemoCue analyser ( Hemocue Hb 301 system; Angelholm , Sweden ) to assess anemia , palpation for liver and spleen enlargement , and measurement of body temperature using an ear thermometer ( Braun ThermoScan IRT 4520; Kronberg , Germany ) for identification of fever cases ( ≥38 . 0°C ) . Two anthropometric measurements were taken ( i . e . , height in cm and body weight in kg , precision 0 . 5 kg ) for subsequent calculation of children's nutritional status . A questionnaire assessing the socioeconomic status , self-reported symptoms and diseases , and HrQoL was administered to all children . Questions on household asset ownership , diseases , and disease-related symptoms were adapted from an instrument previously used in school-based surveys conducted in Côte d'Ivoire [37] . Children were asked for 11 different symptoms ( i . e . , abdominal pain , blood in stool , blood in urine , diarrhea , dysentery , fatigue , fever , headache , loss of appetite , respiratory problems , and vomiting/nausea ) and eight diseases ( i . e . , cold , cough , eye disease , malaria , malnutrition , schistosomiasis , skin disease , and worms ) using a recall period of 2 weeks . To evaluate self-rated HrQoL , the French version of the WHOQOL-BREF tool [18] served as template . Specific questions were dropped and some questions were slightly rephrased to be more specific for the current context , interviewing school-aged children in Côte d'Ivoire . In addition to specific questions focusing on HrQoL , children were asked to rate their general health status using an adapted VAS [38] . This single-item measure basically consists of a thermometer-like scale , in which the anchors are ‘best imaginable health’ and ‘worst imaginable health’ , in our case defined as a maximum and minimum value of 10 and 0 , respectively . The complete questionnaire instrument was further refined in several rounds of pre-testing in a primary school that was not otherwise involved in the current study . In this pre-testing , children attending grades 2–5 with different cultural backgrounds were included . We determined interview duration using a stopwatch and comprehensibility and appropriateness of the HrQoL part , which was not yet validated from earlier studies , with the goal to achieve a compact , understandable , and locally valid instrument . Questionnaire interviews in the field were conducted by members of the study team and teachers from the selected schools , who were trained beforehand . Data were double-entered and cross-checked using EpiInfo version 3 . 5 . 3 ( Centers for Disease Control and Prevention; Atlanta , United States of America ) and analyzed in Stata version 10 . 1 ( Stata Corp . ; College Station , United States of America ) . Only data from children with written informed consent , completed questionnaire , valid parasitological results , and clinical assessments were considered for further analysis . Socioeconomic data were utilized to calculate a wealth index following an asset-based approach as adopted and explained elsewhere [37] , [39] . According to their index score , children were stratified into five economic groups according to wealth quintiles ( i . e . , most poor , very poor , poor , less poor , and least poor ) . Data on helminth infections were classified into light , moderate , and heavy , following WHO guidelines [7] . Anemia was defined as having a Hb level below 115 g/l in children aged 5–11 years and 120 g/l in children aged12–15 years [40] . The presence of organ enlargement was defined as having a palpable liver or spleen; the latter of grade 1 or higher using a Hackett's scale [41] . Indicators for malnutrition were calculated according to WHO child growth standards for children aged 5–19 years [42] . They included stunting ( height-for-age ) , wasting ( body mass index ( BMI ) -for-age ) , and underweight ( weight-for-age ) . The latter is considered a valid measure for nutritional status in children up to 10 years only and was incorporated in a summary measure for malnutrition , defined as Z-score <−2 for any of the three nutritional indicators . HrQoL questionnaire answers were coded as 1 , 2 , or 3 ( in question 1 up to five codes; Appendix S1 ) with higher scores indicating fewer problems for a certain issue or activity . HrQoL questionnaire scores were summarized into three main domains on ( i ) physical , ( ii ) psychosocial , and ( iii ) environmental wellbeing . The first comprised the sum of scores from questions 2–6 , the second from questions 7–9 , and the third from questions 10–12 . Each child's overall score on HrQoL was built by summing up individual scores from questions 1–12 . Domain and overall raw scores were further converted to a 100-point scale ( formula: [ ( raw score−lowest possible score ) /raw score range]×100 ) [43] . Cronbach's alpha coefficient was used to assess for internal consistency of the HrQoL scores . Overall HrQoL , domain , and VAS scores were subjected to analysis on floor and ceiling effects . Floor or ceiling effects ( >15% of respondents achieved lowest or highest possible score ) can indicate limited content validity and reduced reliability , whilst responsiveness may be limited since changes in respondents with lowest or highest possible scores cannot be measured [44] . The validity of the HrQoL instrument was further evaluated by assessing relationships of domain , overall HrQoL and VAS scores with symptoms reporting and clinical signs using Spearman rank correlation and linear regression analysis , as appropriate . In order to relate the questionnaire measures with self-reported and clinically assessed morbidity , additional summary variables providing the total number of self-reported symptoms and diseases ( n = 19 ) and clinical signs ( n = 7 ) for each child was generated , with possible ranges of 0 to 19 and 0 to 7 , respectively . Chi square ( χ2 ) , Fisher's exact , Student-t , Kruskal-Wallis , and Wilcoxon rank sum tests were applied , as appropriate , to investigate significant univariate differences between groups for sociodemographic , parasitological , clinical , and HrQoL indicators . Associations between the HrQoL outcome and parasitic infection , infection intensity , and clinical status were assessed using multivariate linear regression analysis with random effects to account for clustering within schools . In case of censored data , we additionally applied tobit regression models . Particular emphasis was placed on total HrQoL and physical wellbeing domain scores as outcome in order to make explicit the physical and non-physical impacts of the health conditions assessed . Explanatories of regression models included sociodemographic , parasitological , and clinical variables . The final models were built , following a stepwise backward elimination approach . Covariates were excluded from the model at a significance level of 0 . 20 or higher . Relationships between the outcome and remaining explanatory variables were expressed as adjusted mean differences with corresponding 95% confidence intervals ( CIs ) .
A total of 94 schools across Côte d'Ivoire were visited during the study and 5 , 491 children invited to participate . Figure 2 depicts the study compliance and participation in the various assessments undertaken . The final sample used for in-depth analysis consisted of 4 , 848 children from 92 schools with a mean age of 9 . 8 years ( range: 5 to 16 years ) . These children had complete questionnaire , parasitological , and clinical data and had not received deworming drugs within the past 4 weeks prior to the survey . There were slightly more boys than girls ( 2 , 579 versus 2 , 269 ) . 72 schools were considered rural , whilst the remaining 20 ( 21 . 7% ) were based in urban settings . 4 , 101 ( 84 . 6% ) of the children belonged to the two targeted school grades , 3 and 4 . The data set is provided as supplementary information ( Data set S1 ) . Table 1 summarizes overall prevalence and intensity of parasitic infections , clinical signs , and self-reported symptoms and diseases . Overall 3 , 635 of the 4 , 848 children ( 75 . 0% ) harbored any malaria parasite . P . falciparum was the predominant species ( 74 . 1% ) , followed by P . malariae ( 3 . 9% ) and P . ovale ( 0 . 3% ) . The latter two Plasmodium species occurred mainly as co-infections with P . falciparum . Helminth infections; namely , hookworm , S . mansoni , A . lumbricoides , and T . trichiura were observed in 17 . 2% , 3 . 7% , 1 . 8% , and 1 . 3% of the children , respectively . Microhematuria was found in 5 . 7% of the children . The majority ( 95 . 6% ) of soil-transmitted helminth infections were of light intensity , whereas about half of the S . mansoni-infected children had moderate- to heavy-intensity infections ( ≥100 eggs per gram of stool ) . More than a fourth of all children were found to be anemic ( 28 . 7% ) or malnourished ( 28 . 4% ) and a mean number of 6 . 1 experienced symptoms or diseases were reported . Detailed information on parasitic infections and clinically assessed and self-reported morbidity stratified by sex , age group , residential area , and ecozone are provided in Supporting Information Tables S2 and S3 . Boys showed significantly higher infection rates for P . falciparum , hookworm , and S . mansoni ( Table S1 ) . Prevalence rates differed between age groups; while P . malariae was more often found in younger children , infections with Schistosoma and soil-transmitted helminths were more prevalent in children aged 11–16 years than in their younger counterparts . Plasmodium spp . and soil-transmitted helminth infections were most prevalent among the poorest and rural households ( all p<0 . 001 ) . Plasmodium spp . was more common in children living in the northern ecozone . Clinical morbidity , such as anemia and indicators for malnutrition , was more pronounced in boys than girls and in older children compared to their younger counterparts ( Table S2 ) . Splenomegaly was found to be more common in the younger age group ( p = 0 . 014 ) and in children from rural and northern settings compared to children living in urban and southern environments ( both p<0 . 001 ) . Anemia ( p = 0 . 049 ) , splenomegaly ( p<0 . 001 ) and stunting ( p<0 . 001 ) were significantly lower in children from wealthier households . Furthermore , helminth ( OR = 1 . 69 , p<0 . 001 ) and Plasmodium ( OR = 1 . 44 , p<0 . 05 ) mono-infected as well as co-infected ( OR = 2 . 0 , p<0 . 001 ) children showed significantly higher odds ratios ( ORs ) for anemia than their non-infected peers in multivariable logistic regression analysis . Symptom and disease reporting was higher in girls compared to boys , in older compared to younger individuals , in children from northern regions compared to their counterparts living in the southern ecozone ( all p<0 . 001 ) , and in children from poorer households ( p = 0 . 025 ) . Table 2 shows the results from the utility and validity analysis of the HrQoL measures . For the summary scores , floor and ceiling effects were negligible . In contrast , relevant ceiling effects were observed for single HrQoL domains and the VAS scores . Internal consistency of the 12-item HrQoL questionnaire was above the recommended threshold of 0 . 7 for Cronbach's alpha needed for comparison between groups . The item-rest correlations were all above 0 . 25 , indicating that single items measured the same construct as the remaining ones and removal of a specific item would not have increased Cronbach's alpha . Self-reported symptoms and diseases were reflected in the HrQoL . All HrQoL measures showed significant negative correlations and associations with increasing number of self-reported morbidities . For an incremental increase of 1 self-reported morbidity , the overall HrQoL decreased by 1 . 4 points ( p<0 . 001 ) . Clinical signs were mainly captured by the physical domain of the HrQoL tool , showing a decreased domain score of 1 . 2 points ( p = 0 . 001 ) by each supplemental clinical morbidity observed . VAS scores showed a statistically significant correlation and association with self-reported symptoms and diseases ( Table 2 ) and also a statistically significant correlation with overall HrQoL ( all p<0 . 001 ) , but the correlations were only weak ( ρ = −0 . 22 and ρ = 0 . 30 , respectively ) . The VAS results were not considered for in-depth analysis and calculation of DWs due to deviance between actual data collected and the original concept of the scale . Univariate analysis showed several differences in overall HrQoL among groups with different sociodemographic factors and observed clinical signs ( Table 3 ) . Boys reported higher overall HrQoL scores , which were mainly driven by higher self-rated environmental wellbeing . Children from the most poor wealth quintile showed significantly lower scores for all three HrQoL domains . Lower scores for psychosocial and environmental wellbeing , and thus lower overall HrQoL scores , were observed in older children and children living in urban areas . Children from the northern regions reported higher physical but lower environmental wellbeing than their peers from the southern zone . Children's HrQoL with regard to parasitic infections mainly showed differences for the physical domain . Microhematuria negatively affected physical wellbeing , while light-intensity soil-transmitted helminth infections and low Plasmodium parasitemia were associated with fewer problems in this domain compared to non-infected counterparts . Comparison for Plasmodium-helminth co-infection categories and the number of concurrent parasitic infections ( including malaria parasites ) showed that children harboring two or more concurrent infections reported the highest physical wellbeing scores compared to their mono- or non-infected counterparts . Anemic children's HrQoL was considerably compromised compared to non-anemic children . A similar but less pronounced decrease in HrQoL was found in children with splenomegaly . Other observed clinical signs showed no significant effects on children's overall HrQoL , but wasted children reported a significantly increased psychosocial wellbeing , while generally malnourished children reported not only higher psychosocial but also higher environmental wellbeing . Table 4 provides an overview on significant associations between sociodemographic , parasitological , and clinical variables on one hand and self-reported HrQoL on the other hand , placing emphasis on summary and physical wellbeing scores , derived from multivariate linear regression with a stepwise backward elimination procedure . Sex , socioeconomic status , anemia , Plasmodium spp . infection , Plasmodium-helminth co-infection , and number of concurrent parasitic infections remained significant predictors for overall HrQoL . If only physical wellbeing was considered , negative associations of clinical manifestations such as anemia and malnutrition were more pronounced . Interestingly , several single species parasitic infections ( i . e . , Plasmodium spp . , and soil-transmitted helminths ) and multiple species parasitic infections ( i . e . , Plasmodium-helminth , and number of concurrent infections ≥2 ) showed a significant positive association with self-reported physical wellbeing .
We present HrQoL measures among 4 , 848 school-aged children surveyed during a 3-month cross-sectional survey in the dry season in Côte d'Ivoire , and explore associations with parasitic infections and clinical and sociodemographic measures . Parasitological examination revealed a very high prevalence of Plasmodium spp . infection ( 75 . 0% ) . Helminth infections were considerably lower; 17 . 2% , 10 . 6% , 3 . 7% , 1 . 8% , and 1 . 3% for hookworm , S . haematobium ( microhematuria ) , S . mansoni , A . lumbricoides , and T . trichiura , respectively . More than a quarter of the surveyed children showed clinical signs of anemia and malnutrition . Findings from multivariate linear regression analysis revealed that the children's self-rated overall HrQoL and physical wellbeing is lower among those affected by anemia and malnutrition compared to their counterparts without anemia and malnutrition . Surprisingly , associations between HrQoL and parasitic infection status were of positive rather than negative direction . Sociodemographic variables such as sex , age group , socioeconomic status , and setting characteristics had considerable influences on children's perceived HrQoL . The locally adapted HrQoL instrument employed showed acceptable utility considering minimal floor and ceiling effects and a robust internal consistency ( Cronbach's α≥0 . 7 ) . Significant correlations and associations between HrQoL scales and self-reported and clinically assessed morbidity were found and even though the effect sizes were weak , they may further support the concept of health measured by the HrQoL tool . Interestingly , we could not identify significantly lower HrQoL scores in Plasmodium- and helminth-infected children compared to their non-infected peers . Possible explanations for this finding are offered for consideration . First , in Côte d'Ivoire 100% of the population is at risk of Plasmodium infection [3] and previous research concluded that malaria transmission is perennial [45] , [46] . Constant exposure from early childhood onwards leads to naturally acquired immunity to malaria at an early age [47] . Thus , most of the Plasmodium infections we identified in the school-aged population surveyed were asymptomatic ( >98% ) . Levels of transmission and endemicity of parasitic infections has been shown to influence children's HrQoL . For example , Kenyan school-aged children infected with S . haematobium from a high endemicity setting reported similar HrQoL measures than their non-infected counterparts , whilst infected children in a low prevalence village reported significantly lower HrQoL compared to non-infected children [23] . Second , our study focused on children who were present at school the day of the survey . Children experiencing a clinical disease episode , perhaps related to a parasite infection , and who might have expressed lowered HrQoL , were more likely to be absent from school than their healthier peers . Helminth infections , additionally , might still be at a less advanced stage with regard to disability in children compared to adolescents or adults . Smaller studies conducted in the People's Republic of China and Kenya also found no evidence of significant differences in self-rated HrQoL between helminth-infected and non-infected school children [24] , [26] . Those studies that reported negative associations between HrQoL and helminth infections either focused on adult populations [22] or investigated chronic and advanced clinical stages of an infection [21] , [48] . Third , polyparasitism , particularly Plasmodium-helminth co-infections , is common in Côte d'Ivoire [49]–[51] . It follows that interactions between multiple species parasitic infections and their influence on ill-health must be considered . Indeed , potentially beneficial effects from light-intensity helminth infections on clinical outcomes ( i . e . , anemia ) and subtle morbidity ( i . e . , physical fitness ) in school-aged populations from malaria co-endemic settings in Côte d'Ivoire have been indicated [46] , [50] , [51] . Underlying mechanisms might be seen in the immunomodulatory features of helminth infections that down regulate the pro-inflammatory immune response needed to combat intracellular parasites like Plasmodium . Consequently , this may negatively affect resistance but simultaneously promote tolerance to malaria-related pathology by controlling harmful associated inflammation [52] . The current study confirms these prior observations , as we observed a positive association between soil-transmitted helminth infections , Plasmodium-helminth co-infections , and two or more concurrent infections including malaria parasites , and reported physical wellbeing . We found significantly lower HrQoL among anemic children compared to non-anemic children . Parasitic infections , most notably Plasmodium and hookworm contribute to the development of anemia [53] . Plasmodium and helminth mono- or co-infected children in our sample showed significantly higher odds ratios for anemia than their non-infected counterparts ( all ORs>1 . 4 ) . Consequently , we suggest the attribution of direct disease consequences ( sequelae ) – such as anemia due to specific parasitic infections – to the etiological cause in future burden estimates [14] . We found a 1-point lower HrQoL score overall and a 2-point lower physical wellbeing score on a 100-point scale . If divided by 100 , these findings might translate to DWs of 0 . 01 and 0 . 02 on the DW scale that ranges from 0 to 1 . Such DWs are within the range of recent DW estimates of the GBD 2010 Study , which were set at 0 . 005 , 0 . 058 , and 0 . 164 for mild , moderate , and severe anemia [15] , [54] . The HrQoL concept attempts to evaluate the impact of diseases and injuries from a comprehensive point of view , incorporating psychological , social , and environmental wellbeing on top of physical health [12] , [15] . We found that particularly psychosocial and environmental measures of wellbeing were significantly associated with sociodemographic variables like sex , age , socioeconomic status , and residential area . Associations of the physical component of HrQoL with parasite infections and clinical signs were observed to be more pronounced and indicated that the perceived health status varies between and depends importantly on different sociocultural settings . Our results are in line with previous observations [22] , [23] , [25] and highlight the importance of inclusion of social determinants for more integrative burden of disease assessments . Our data stem from a large-scale nation-wide survey , which subjected almost 5 , 000 children to detailed clinical and parasitological examinations , coupled with a questionnaire . A major weakness of previous studies was their small sample sizes [22] , [24] , [25] . Further , we consider the setting-tailored , applied HrQoL tool as a useful and valid instrument . Its internal consistency was good ( Cronbach's α>0 . 7 ) and floor and ceiling effects for the overall HrQoL were minimal , despite its shortness , including only 12 items compared to 26 questions in the WHOQOL-Bref , which was used as a template to develop our tool [18] . Particularly the ceiling effects were more pronounced when looking at single domain scores , which is , however , not surprising , considering the lower number of items in each domain . The ceiling effects found for the physical domain , were addressed by utilizing tobit regression analysis , which have been shown to provide more reliable estimates in censored data [55] , in parallel to linear regression models . The negative associations and correlations between HrQoL and symptom and disease reporting followed the logic of lower self-rated HrQoL in simultaneously higher experienced morbidity and supported the construct that our instrument measures . Data collection on a national scale entails several limitations . To respect the tight time schedule and in view of limited financial and human resources , all parasitological , clinical , and questionnaire information had to be collected within a single day at each location by dedicated field teams . Consequently , teachers of the selected schools were trained to administer our questionnaire and they assisted in the conduct of the interview . Given our time constraints and restricted resources , we were not able to assess inter-observer agreement and cannot exclude measurement errors due to variation between interviewers . Another limitation regarding the questionnaire was the difficult implementation of the VAS , as already observed elsewhere [24] . The concept of this scale , the range of 0 to 100 , and the fact that children had to point out their respective health status on a sheet was poorly understood . As an adaptation , children were asked to rate and orally express their health status according to a scale they were more familiar with , a scale of school marks ( ranging from 0 to 10 ) . However , this procedure resulted in a categorical rather than an interval distribution of scores . Unfortunately , this limitations hindered us to fully exploit these data and derive DWs , which could have been compared with previous research conducted elsewhere focusing on chronic S . japonicum infection [48] . Hence , there is a pressing need for a culturally accepted alternative to the VAS . We conclude that the assessment of HrQoL in school-aged children in areas where parasitic infections are still widespread tends to be difficult and may not be sensitive enough to capture subtle morbidities . Important factors blurring the picture might be the often asymptomatic course due to acquired immunity in malaria and more subtle morbidities at this age for helminth infections , which therefore may not be perceived as disabling by infected children . School absenteeism adds bias , as non-inclusion of children who might experience more measurable disability will not be part of the analysis . Importantly though , the applied instrument showed acceptable utility and validity and was able to identify significant disability of more chronic sequelae such as anemia . Further refinement and more rigorous reliability measurements of the tool are needed . Surveys in settings targeting specific parasite endemicity levels and efforts to include non-enrolled and otherwise absent school-aged children might resolve some of the limitations highlighted here . The aim of developing , validating , and applying setting-specific HrQoL tools that will allow comparison between areas and measuring changes over time remains – particularly as large-scale control efforts targeting malaria and the NTDs are underway . | Infectious diseases like malaria and parasitic worms affect hundreds of millions of people , and impact physical and cognitive development of children in Africa , Asia , and the Americas . Over the past 20 years , it was debated how the magnitude of ill-health due to these conditions should be assessed . One proposed strategy was to include patient-based ratings of wellbeing by administration of health-related quality of life ( HrQoL ) questionnaires . In order to provide new evidence on disability from parasitic infections , we conducted HrQoL interviews with children aged 5–16 years from 92 schools across Côte d'Ivoire . Children were examined for parasitic infections and clinical signs like anemia , malnutrition , and organ enlargement . We compared the self-rated HrQoL of infected and non-infected children and also considered their sociodemographic background . We could not identify lowered HrQoL in infected children , but we found that children with anemia reported a 1-point lower score on a 100-point HrQoL scale in comparison with their non-anemic counterparts . We consider our HrQoL questionnaire as useful and valid , but would recommend its further testing and development in few purposefully selected settings . Further investigation of disability induced by malaria and parasitic worm infections is warranted . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"helminth",
"infections",
"medicine",
"and",
"health",
"sciences",
"plant",
"science",
"schistosomiasis",
"parasitic",
"intestinal",
"diseases",
"infectious",
"disease",
"epidemiology",
"epidemiology",
"plant",
"pathology",
"biology",
"and",
"lif... | 2014 | Health-Related Quality of Life among School Children with Parasitic Infections: Findings from a National Cross-Sectional Survey in Côte d'Ivoire |
Dietary restriction ( DR ) is a dietary regimen that extends lifespan in many organisms . One mechanism contributing to the conserved effect of DR on longevity is the cellular recycling process autophagy , which is induced in response to nutrient scarcity and increases sequestration of cytosolic material into double-membrane autophagosomes for degradation in the lysosome . Although autophagy plays a direct role in DR-mediated lifespan extension in the nematode Caenorhabditis elegans , the contribution of autophagy in individual tissues remains unclear . In this study , we show a critical role for autophagy in the intestine , a major metabolic tissue , to ensure lifespan extension of dietary-restricted eat-2 mutants . The intestine of eat-2 mutants has an enlarged lysosomal compartment and flux assays indicate increased turnover of autophagosomes , consistent with an induction of autophagy in this tissue . This increase in intestinal autophagy may underlie the improved intestinal integrity we observe in eat-2 mutants , since whole-body and intestinal-specific inhibition of autophagy in eat-2 mutants greatly impairs the intestinal barrier function . Interestingly , intestinal-specific inhibition of autophagy in eat-2 mutants leads to a decrease in motility with age , alluding to a potential cell non-autonomous role for autophagy in the intestine . Collectively , these results highlight important functions for autophagy in the intestine of dietary-restricted C . elegans .
Dietary restriction ( DR ) , defined as limited food intake without malnutrition , is the most robust and conserved intervention currently known to delay aging . Since many physiological effects of DR are evolutionarily conserved , genetically tractable model organisms such as the nematode C . elegans can be exploited as tools to identify the molecular events underlying DR-mediated lifespan extension [1] . Autophagy is an evolutionarily conserved process induced in response to nutrient deprivation via important metabolic regulators such as the nutrient-sensing kinase TOR . During autophagy , cytoplasmic components are first encapsulated within a double-membrane structure called an autophagosome . This structure subsequently fuses with an acidic lysosome to form an autolysosome , in which the sequestered cargo is degraded by hydrolases [2] . The autophagy process is orchestrated by >30 conserved autophagy proteins ( encoded by ATG genes ) , many of which function in complexes at different steps in the process . Of particular note is ATG8 , a family of lipid-binding proteins , which play essential roles in autophagosome formation , cargo recruitment , and autophagosome–lysosome fusion [2] . During autophagy , ATG8 proteins get post-translationally modified and inserted into both the inner and outer autophagosomal membranes and GFP-tagged ATG8 proteins are commonly used as markers to assess steady-state levels of autophagosomes in many species [2] , including C . elegans , which contain two ATG8 homologs , LGG-1 and LGG-2 [3] . Autophagy is increasingly recognized to play a critical role in lifespan extension promoted by multiple conserved longevity paradigms [4] . In particular , C . elegans subjected to DR require autophagy genes for lifespan extension and have increased levels of the ATG8 autophagy marker LGG-1 in their hypodermis [5–8] . These observations suggest a model in which the induction of autophagy by DR , plays a beneficial role in contributing to lifespan extension [4] . However , it is currently unclear if the effects of DR on organismal aging originate from individual tissues . While autophagy is modulated in multiple tissues of dietary-restricted mammals , such as the liver , skeletal muscle , and cardiac muscle of mice [9–13] and rats [14] , it remains to be directly tested if tissue-specific changes in autophagy are sufficient to contribute to the improved healthspan and longevity of dietary-restricted animals . Previous experiments in C . elegans and Drosophila have highlighted an important role for the intestine in several longevity paradigms . Specifically , modulation of the expression of certain genetic determinants specifically in the intestine can influence lifespan [15–18 , 40] . It however remains unclear how such genetic interventions may affect physiological functions of the intestine . A key role of the intestinal epithelial layer in all organisms is to form a selective permeability barrier that permits absorption of water and critical solutes while maintaining a defense against potentially toxic substances . Studies in Drosophila have shown that the intestinal epithelium in flies , as in mammals , possesses a barrier function . This “intestinal barrier function” declines with age but the rate of decline can be slowed by DR [18–20] . However , it is unknown how DR affects this physiological function of the intestine , and if additional longevity paradigms besides DR can similarly prevent its age-dependent decline . To study the functional role of the intestine in DR-mediated longevity and learn more about how autophagy contributes to the fitness and aging of animals subjected to DR , we examined the autophagy process in the intestine of dietary-restricted C . elegans . We used eat-2 ( ad1116 ) mutants , which carry a mutation that compromises food uptake , thereby providing a genetic model of DR . Notably , specific inhibition of autophagy genes in the intestine was sufficient to prevent the long lifespan of eat-2 ( ad1116 ) mutants , indicating an important role for autophagy in this organ in the DR longevity paradigm . Consistent with this , the intestine of eat-2 ( ad1116 ) mutants displayed an enlarged lysosomal compartment and increased turnover of an autophagosomal marker compared to wild-type ( WT ) animals . Using a novel assay in C . elegans to measure intestinal barrier function , we found that DR slowed the rate at which intestinal integrity decreased with age , and this protective effect was dependent on intestinal expression of autophagy genes . Similarly , whole-body or intestine-specific autophagy gene knockdown alone caused a reduction in motility , an age-dependent trait , irrespective of the genetic background . In contrast , knockdown of autophagy in the muscle did not cause significant changes in motility , but prevented full lifespan extension . Collectively , our results indicate that inhibition of autophagy in the intestine of dietary-restricted animals prevented lifespan extension and resulted in phenotypes reminiscent of aging , thus highlighting important tissue-specific functions for autophagy that contribute to healthspan and longevity .
In C . elegans , the eat-2 gene is essential for pharyngeal pumping , and animals carrying mutations in this gene have a reduced rate of food intake and are longer lived [21] . Such mutants therefore provide a genetic model of DR . Autophagy is required for DR to extend lifespan in C . elegans , since RNAi of the autophagy-related genes unc-51/ATG1/Ulk1 , bec-1/ATG6/Beclin1 , vps-34 , and atg-7 shortens the long lifespan of animals with eat-2 mutations [5 , 6 , 8] , while having a relatively small or non-significant effect on the lifespan of wild-type ( WT , N2 ) animals . However , it is not yet known how autophagy regulation in select tissues contributes to DR-induced longevity . Since the intestine is a central metabolic tissue key to nutrient uptake and for healthy aging [17] , we addressed if intestinal autophagy plays a role in DR-mediated longevity . To inhibit autophagy in the intestine , we utilized tissue-specific RNA interference ( RNAi ) . To this end , we crossed eat-2 ( ad1116 ) mutants to rde-1 ( ne219 ) loss-of-function mutants with reconstituted rde-1 expression from the intestinal nhx-2 promoter [22] . rde-1 ( ne219 ) mutants carry a loss-of-function mutation in the piwi/argonaute family member essential for the initiation of RNAi [23] . We confirmed the intestinal RNAi specificity of this strain by using RNAi clones targeting different genes expressed inside and outside of the intestine ( S1A Fig ) . Lifespan analyses revealed that eat-2 ( ad1116 ) ; rde-1 ( ne219 ) double mutants , and eat-2 ( ad1116 ) ; rde-1 ( ne219 ) mutants expressing the nhx-2::rde-1 transgene had lifespans largely comparable to that of eat-2 ( ad1116 ) single mutants ( Fig 1A and 1B and S1 Table ) . These data indicate that neither systemic rde-1 deficiency nor tissue-specific re-expression of rde-1 in the intestine significantly affected the lifespan of the eat-2 ( ad1116 ) mutants . To determine the effect of intestinal autophagy inhibition on lifespan , we fed the eat-2 ( ad1116 ) ; rde-1 ( ne219 ) ; nhx-2p::rde-1 strain with bacteria carrying empty vector or expressing dsRNA targeting atg-18/Wipi , or the two ATG8 homologs lgg-1 and lgg-2 from Day 1 of adulthood . These genes were selected because whole-body inhibition of their expression had potent effects on the lifespan of eat-2 ( ad1116 ) animals ( S1 Table ) . Autophagy gene RNAi caused a significant ~10–22% reduction in the mean lifespan of eat-2 ( ad1116 ) ; rde-1 ( ne219 ) animals expressing rde-1 from the nhx-2 intestinal promoter ( Fig 1A and 1B and S1 Table ) . These results were supported by an eat-2 ( ad1116 ) ; sid-1 ( qt9 ) double mutant expressing sid-1 , a transmembrane protein that acts as a channel for dsRNA entry [24] , from the gly-19 promoter ( S1 Table , see also Methods ) . In parallel to these experiments , we assayed the lifespan of eat-2 ( ad1116 ) mutants subjected to whole-body RNAi , which generally resulted in stronger lifespan-shortening effects ( Fig 1 and S1 Table ) . These observations are consistent with autophagy contributing to DR-mediated lifespan in the intestine and we conclude that autophagy plays an important function in longevity within the intestine of dietary-restricted animals . During autophagy , LGG-1/ATG8 is sequestered at the membrane of nascent autophagosomes and can be visualized as characteristic punctate structures in animals expressing a GFP-tagged form of LGG-1 [25 , 26] . We and others have previously reported that DR ( via the eat-2 mutation or by dilution of bacterial food source ) increases the abundance of GFP::LGG-1–positive punctae in the hypodermal seam cells of L3 larvae and 1-day-old adult animals [5 , 7 , 8] . To investigate the role of autophagy in the intestine of DR animals , we monitored autophagy by quantifying autophagosomes in the intestine of WT and eat-2 animals [27] . Unexpectedly , we found that the number of GFP::LGG-1 punctae was significantly lower in the intestine of eat-2 ( ad1116 ) mutants compared to WT animals on Day 4 of adulthood ( Fig 2A ) . This decrease in GFP::LGG-1 punctae was independent of the type of bacteria the animals were fed ( S2A Fig ) , and was also observed by confocal microscopy in older Day 7 animals ( S3A Fig ) , collectively suggesting that the steady-state levels of autophagosomes in the intestine of dietary-restricted eat-2 mutants are lower than in WT animals . Since the hypodermis [5 , 7] and the intestine ( Figs 2A , S2A and S3A ) of eat-2 mutants showed a divergent number of autophagosomes , we additionally counted GFP-positive LGG-1 punctae by confocal microscopy in two other tissues , i . e . , body-wall muscle and neurons of 3-day-old eat-2 mutants ( S2B and S2C Fig ) . Similar to the hypodermis , the neurons displayed an increased number of autophagosomes while the muscle showed no change compared to WT , highlighting the intestine as the only tissue in which we observed decreased numbers of GFP::LGG-1 punctae . Notably , all C . elegans longevity models tested to date require autophagy genes for lifespan extension and have increased numbers of LGG-1 punctae in hypodermal cells at the late larval stages compared to wild-type animals [28] . Thus , our intestinal-specific lifespan experiments ( Fig 1 ) and GFP::LGG-1 analyses ( Figs 2A , S2 and S3A ) are the first to show a lack of correlation between a requirement for autophagy genes and an increase in LGG-1-positive punctae in long-lived animals . The reduced number of GFP::LGG-1 positive punctae we observed in eat-2 mutant intestines may be due to a decrease in the formation of autophagosomes , or an increase in the conversion or turnover of autophagosomes . Consistent with the latter hypothesis and in support of increased autophagic activity we observed that eat-2 ( ad1116 ) mutants had increased mRNA levels of autophagy-related genes ( lgg-1/ATG8 , sqst-1/SQSTM1 ) as well as lysosomal genes ( vps-11 , vha-15/V-ATPase , vha-16/V-ATPase ) compared with the levels in WT animals ( Fig 2B ) . These results corroborate our previous analysis of eat-2 ( ad1116 ) mutants [29] , which showed increased expression of additional autophagy-related genes important for autophagosome formation ( atg-18/Wipi and atg-9 ) as well as lysosomal genes ( lmp-1/Lamp1 , vps-18 , and cathepsins ( cpr-1 , asp-1 ) ) , and collectively support that eat-2 mutants have increased autophagy . Based on these observations , we hypothesized that the intestine of eat-2 mutants might possess an enlarged lysosomal compartment , which could facilitate autophagosome–lysosome fusion and subsequent autophagosome conversion irrespective of a possible concomitant increase in autophagosome formation . To test this , we used transmission electron microscopy to image the intestines of WT and eat-2 ( ad1116 ) animals to quantify the number of autolysosomal-like structures . Autolysosomes were identified as vacuoles containing contents undergoing degradation [27] . We found that eat-2 ( ad1116 ) mutants had a significant increase in the number of autolysosomes at Day 3 of adulthood compared to WT ( Fig 2C and 2D ) . Similar observations were made in animals at Day 7 of adulthood ( S3B Fig ) . Additionally , we found that eat-2 ( ad1116 ) mutants had increased intestinal staining of acidic organelles by the pH-sensitive fluorophore LysoTracker Red DND-99 , a red-fluorescent dye supplemented to media from Day 1 of adulthood and whose fluorescence intensity was quantified on Day 5 ( Fig 2E and 2F ) . Collectively , these observations are consistent with an expanded lysosomal compartment in the intestine of C . elegans during DR . This increased steady-state level of lysosomes could facilitate an increased conversion rate of autophagosomes , which might be reflected in lower steady-state levels of GFP::LGG-1 positive punctae , as we observed in the intestine of eat-2 mutants ( Figs 2A , S2A and S3 ) . In such a scenario , autophagosomal contents would be turned over more frequently . To test this , we blocked autophagy by subjecting WT and eat-2 ( ad1116 ) animals to RNAi targeting vha-15 , which encodes a subunit of an essential lysosomal enzyme , V-ATPase [30] . RNAi from the L4 larval stage to Day 4 of adulthood caused a ~4-fold accumulation of GFP::LGG-1 punctae in the intestine of eat-2 ( ad1116 ) mutants whereas a small , but insignificant ~5% increase was observed in WT animals ( Fig 1A ) . Taken together , our results are consistent with the notion that autophagosome conversion and by extension autophagic activity is increased in the intestine of eat-2 mutants , at least in part facilitated by a larger lysosomal compartment . The data additionally emphasize the importance of employing multiple autophagy assays , including ‘flux’ assays to infer autophagy activity , to complement GFP::LGG-1 reporter assays and thus provide more accurate assessments of the status of autophagy within cells . To further investigate roles of autophagy in the intestine , we next sought to establish an assay that could measure functions of this organ in C . elegans . To this end , we developed a non-invasive assay to measure C . elegans intestinal barrier function , as previously done in Drosophila [18 , 19] . In our protocol , animals were fed for 3 h by submersion in a liquid bacterial culture mixed with a non-absorbable blue food dye , and then visualized microscopically . After incubation , the dye was clearly visible in the intestine of both young and older WT and eat-2 ( ad1116 ) mutants ( Fig 3A and 3B ) . While the dye was contained within the intestine in younger animals , older animals also displayed dye in their body cavity ( Fig 3A and 3B; as in the Drosophila studies , we refer to this as a 'Smurf' phenotype [18 , 19] ) . Specifically , dye leakage was observed around mid-adulthood ( Day 7; Fig 3C ) and increased in frequency with age ( ~60 percent on Day 15; Fig 3C ) , indicating an age-dependent decline in intestinal integrity similar to observations in Drosophila [18 , 19] . Notably , the intestinal integrity of eat-2 ( ad1116 ) mutants was remarkably improved as far fewer eat-2 ( ad1116 ) than WT animals displayed the Smurf phenotype ( ~20 percent on Day 15; Fig 3B and 3C ) . These data suggest that the age-related decrease in intestinal integrity in C . elegans is improved by DR , as observed in Drosophila [19 , 20] . We also tested daf-2 ( e1370 ) insulin/IGF-1 receptor mutants , another long-lived mutant [31] in the Smurf assay . Interestingly , these mutants maintained their intestinal integrity remarkably well over time ( only ~5 percent on Day 16; S4A Fig ) , suggesting a broader correlation between longer lifespan and improved intestinal integrity . Based on our observations that eat-2 mutants display improved intestinal barrier function and enhanced autophagosomal turnover in the intestine compared with WT animals , we next asked whether autophagy was required to maintain the intestinal integrity of these mutants . WT and eat-2 mutants were subjected to whole-body atg-18/Wipi RNAi during adulthood , and the dye-leakage assay was performed on Days 7 , 11 , and 15 of adulthood . More eat-2 ( ad1116 ) mutants with reduced atg-18/Wipi levels displayed the Smurf phenotype by Day 11 and furthermore at Day 15 ( Fig 4A and 4B ) than control animals , largely coinciding with the time at which eat-2 ( ad1116 ) animals subjected to atg-18/Wipi RNAi would start to be at high risk of dying ( Fig 1 and S1 Table ) . In contrast , atg-18/Wipi RNAi had no significant effect on dye leakage in WT animals at any time point measured ( S4B Fig ) . Similar trends were observed in eat-2 ( ad1116 ) mutants subjected to whole-body vha-15/V-ATPase RNAi ( S4C Fig ) or lgg-1/lgg-2/ATG8 RNAi ( S4D Fig ) at Day 15 of adulthood . These data indicate that whole-body RNAi of autophagy genes reduced the protective effect of the eat-2 mutation on intestinal barrier function just like inhibition of autophagy prevents the long lifespan of eat-2 mutants , yet cause relatively small or non-significant effects on lifespan or intestinal barrier function in WT animals [5 , 6] ( S1 Table ) . To determine whether autophagy specifically within the intestine contributes to the barrier function , we subjected eat-2; rde-1 mutants expressing rde-1 from the nhx-2 promoter to atg-18/Wipi RNAi during adulthood and performed the dye-leakage assay on Day 15 . We found that intestine-specific atg-18/Wipi RNAi was sufficient to significantly increase intestinal leakage in eat-2 ( ad1116 ) mutants compared to control-treated animals ( Fig 4C ) . These findings indicate that autophagy can act cell autonomously in the intestine of eat-2 mutants to maintain barrier function , a conserved healthspan parameter that can be experimentally measured in C . elegans . Additional healthspan assays can be used to assess physiological functions in C . elegans [32] . To more broadly explore roles for autophagy in C . elegans , we investigated the effects of autophagy on motility . This activity engages muscular and neuromuscular functions of the animal and is a common measure of C . elegans healthspan [33–35] . To quantify motility , we counted body bends in WT and eat-2 animals swimming freely in liquid media . While Day 3 eat-2 ( ad1116 ) animals were slightly more active than WT animals , the motility of older eat-2 ( ad1116 ) mutants did not significantly differ from that of WT animals ( S5 Fig ) , as previously reported for eat-2 ( ad1113 ) mutants [34] . Interestingly , whole-body RNAi of atg-18/Wipi caused a dramatic decline in motility that reached statistical significance on Day 11 in both WT ( Fig 5A ) and eat-2 ( ad1116 ) animals ( Fig 5B ) . Motility defects generally appeared prior to body-cavity leakage , which first became highly significant on Day 15 in eat-2 ( ad1116 ) mutants subjected to whole-body autophagy gene RNAi ( Fig 4B ) . Thus , motility decline induced by whole-body autophagy gene knockdown in eat-2 mutants may therefore not be linked to lifespan or to intestinal leakage . Consistent with this notion , WT animals subjected to whole-body atg-18/Wipi RNAi displayed decreased motility ( Fig 5A ) , but had the same degree of intestinal leakage at least up to Day 15 ( S4B Fig ) as control animals . The decline in motility caused by whole-body autophagy gene knockdown might reflect changes in neuromuscular functions . To specifically investigate the contribution of muscle , we introduced a myosin MYO-3::GFP reporter highlighting muscle fibers into eat-2 ( ad1116 ) mutants and subjected these transgenic animals as well as WT animals expressing MYO-3::GFP to whole-body atg-18/Wpi and control RNAi . While Day 15 WT animals experienced a significant sarcomere deterioration in response to atg-18/Wipi RNAi compared to control treatments , we observed no obvious differences in the appearance of body-wall muscle fibers in Day 15 eat-2 ( ad1116 ) mutants subjected to control or whole-body atg-18/Wipi RNAi ( S6A and S6B Fig ) despite atg-18/Wipi RNAi-treated eat-2 animals being clearly movement impaired ( Unc ) . These data indicate that different mechanisms may underlie the motility decline observed in WT animals compared to dietary-restricted eat-2 animals following autophagy inhibition . To further address the role of autophagy in the muscle in eat-2 mutants , we created an eat-2 ( ad1116 ) ; sid-1 ( qt9 ) double mutant expressing sid-1 from the myo-3 promoter to be able to inhibit autophagy specifically in the muscle ( S1A Fig ) . Muscle-specific knockdown of atg-18/Wipi in adult eat-2 ( ad1116 ) animals however caused only minor effects on motility ( Fig 5C ) . In contrast , muscle-specific atg-18/Wipi RNAi significantly shortened the long lifespan of these animals ( S7 Fig and S1 Table ) . These observations indicate that the body-wall muscle in eat-2; sid-1; myo-3p::sid-1 transgenic animals indeed possesses RNAi processing ability ( S1A Fig ) , and that autophagy in the body-wall muscle of eat-2 mutants is critical for longevity , as is the case in WT mice [36] . We similarly tried to address functions for autophagy in the neurons of eat-2 mutants by making an eat-2 ( ad1116 ) ; sid-1 ( qt9 ) double mutant expressing sid-1 from the neuronal rab-3 promoter . While we observed that atg-18/Wipi RNAi significantly shortened the long lifespan of this strain ( S1 Table ) , the rab-3 promoter appeared leaky to the intestine starting around Day 5 of adulthood ( S1A and S1B Fig , see also Methods ) . Therefore we were not able to conclusively determine the neuronal effects of autophagy at this point . While autophagy in the body-wall muscle of eat-2 mutants may contribute to their longevity , the motility defects observed after whole-body , but not muscle-specific , disruption of autophagy in eat-2 mutants may originate as a consequence of autophagy disruption in tissues other than the body-wall muscle . Interestingly , we found that eat-2 ( ad1116 ) ; rde-1 ( ne219 ) mutants expressing nhx-2p::rde-1 subjected to intestinal-specific atg-18/WWipi RNAi displayed significantly reduced motility compared with animals subjected to control RNAi , even relatively early in adulthood ( Fig 5D ) . These observations indicate that autophagy in the intestine of dietary-restricted animals has physiological effects on distal tissue function . As such , our data indicate that autophagy inhibition in the intestine during DR not only accelerates a cell-autonomous function ( i . e . , decline in intestinal barrier function ) but also has cell non-autonomous consequences in reducing motility .
Here we show a critical role for autophagy in the intestine to ensure lifespan extension of dietary-restricted eat-2 mutants . Moreover , autophagy inhibition in this major tissue accelerated an age-dependent decline in the intestinal barrier function and in motility of dietary-restricted animals , thus highlighting both cell autonomous and non-autonomous consequences of inhibiting autophagy in a single tissue . Since we found that dietary-restricted animals display an enlarged lysosomal compartment and possibly increased autophagosome conversion in the intestine , we propose that increased autophagic activity in this tissue extends the lifespan of dietary-restricted animals by improving multiple aspects of organismal fitness . We used several assays to monitor autophagy in dietary-restricted eat-2 animals , including a widely used GFP::LGG-1 reporter that reflects steady-state levels of autophagosomes [25 , 26] . Its subcellular localization into GFP-positive punctae is increased in the hypodermis of larvae of all long-lived C . elegans mutants investigated so far , including eat-2 ( ad1116 ) mutants [5 , 7 , 8] . We observed GFP::LGG-1 punctae in adult WT and eat-2 animals in all tissues inspected , i . e . , intestine , muscle and neurons , consistent with a recent report using an enzymatic assay to show that these major tissues can modulate autophagy [37] . Notably , of the three tissues analyzed here the intestine stood out as the only tissue with a decrease in GFP::LGG-1 counts in eat-2 animals compared to WT . We used additional steady-state analyses as well as a ‘flux’ assay that collectively indicated that the intestine of eat-2 mutants possesses increased autophagosome turnover . Together , these intestinal experiments represent the most comprehensive analysis to date of the various steps in the autophagy process in a tissue of C . elegans , and highlight the importance of using multiple assays to monitor the autophagy process [3] . While neurons showed an increase and body-wall muscle showed no difference in the number of GFP::LGG-1 punctae in 3-day-old adult eat-2 mutants compared to WT animals , additional assays are needed to conclusively evaluate the status of autophagy in these tissues . One interpretation of our intestinal data is that autophagosome conversion is more rapid in the intestine of dietary-restricted eat-2 mutants than in WT animals , possibly facilitated by the enlarged lysosomal compartment observed in dietary-restricted mutants . Additional biochemical experiments will be needed to measure the kinetics of the overall autophagy process and its individual steps in WT and dietary-restricted animals to directly address this possibility . We found that inhibition of several autophagy genes ( i . e . , atg-18/Wipi and lgg-1/lgg-2/ATG8 ) in the intestine of eat-2 animals was sufficient to significantly shorten the lifespan , as previously observed following whole-body knockdown of these ( this study ) and additional autophagy genes ( i . e . , unc-51/ATG1/Ulk1 , bec-1/ATG6/Beclin1 , vps-34 and atg-7 ) [5 , 6 , 8] . These observations suggest that intestinal autophagy contributes to DR-mediated lifespan in C . elegans . To investigate the possible functions of intestinal autophagy , we developed a novel dye-leakage assay ( termed the ‘Smurf’ assay ) that allowed us to quantify the intestinal barrier function in C . elegans , as has previously been done in Drosophila [18 , 19] . ( We note that , while this manuscript was under revision , Michael Rera and his collaborators independently reported a dye-leakage assay in C . elegans , as well as in zebrafish [38] ) . Our results indicate that eat-2 ( ad1116 ) animals are more resistant than WT animals to the age-associated increase in dye leakage , similar to observations in flies [19 , 20] , a scenario that appears to extend to additional long-lived mutants , since daf-2 insulin/IGF-1 receptor mutants were remarkably resistant to dye leakage . While we cannot rule out a potential contribution from hypodermal leakage in our assay ( since the animals are soaked in dye ) , our findings are consistent with a previous study of WT C . elegans that used several cytological methods to demonstrate an age-dependent decline in intestinal integrity , associated with irregular lumenal shape , changes in microvilli , and loss of nuclei [39] . Interestingly , whole-body autophagy gene inhibition during adulthood accelerated the breakdown of intestinal integrity in dietary-restricted eat-2 animals but had no significant effect in WT animals; this is similar to the lifespan-extending effects of autophagy inhibition in eat-2 mutants , but generally with small or no significant effect in WT animals [5 , 6] . This finding implies that increased levels of autophagy in the intestine is required for the improved intestinal barrier function in dietary-restricted animals , and could underlie the longevity of these animals . Indeed , the decline in intestinal integrity of eat-2 animals generally coincided with the decrease in survival observed at Day 15 in these animals . However , this observation does not exclude the possibility that basal autophagy could be important for the intestinal barrier function in WT animals . Our finding that autophagy inhibition within the intestine of eat-2 mutants is sufficient to accelerate the decline in intestinal integrity suggests that autophagy normally functions cell autonomously to support intestinal barrier function in dietary-restricted animals . This is consistent with observations in Drosophila , where increased expression of Atg1/Ulk1 induces autophagy and improves the intestinal barrier function [40] . The barrier function itself is maintained via protein-protein networks that form adhesive complexes of desmosomes , adherens junctions , and tight junctions [41] . Notably , a recent study suggested a role for autophagy in turnover of the pore-forming tight junction protein Claudin-2 in nutrient-starved human intestinal epithelial cells , leading to less permeable junctions [42] . Since our findings suggest a link between beneficial autophagic turnover and improved intestinal barrier function in eat-2 mutants , it would be interesting to investigate autophagy-dependent maintenance of tight junctions in dietary-restricted animals . We also assessed the relationship between autophagy and motility , an important healthspan parameter that is known to decline with age , and found that autophagy knockdown significantly reduced the motility of both WT and eat-2 animals . This observation is consistent with reduced climbing activity in adult Atg7 mutant flies [43] . Since autophagy gene RNAi significantly shortens the lifespan of eat-2 mutants but generally has relatively small or non-significant effects in WT animals , these findings uncouple the motility decline from lifespan shortening and suggest that animals with reduced autophagy levels may not die simply as a consequence of decreased motility . Autophagy could influence C . elegans motility in a number of ways , including via autonomous effects in muscle cells . In this regard , inhibition of autophagy in skeletal muscle cells of WT mice reduces lifespan , exacerbates mitochondrial dysfunction , and significantly impairs both muscle strength and neuromuscular synaptic function [9 , 10 , 36] . We observed similar effects on sarcomere integrity in WT C . elegans subjected to autophagy gene inhibition , whereas the muscle fibers in eat-2 mutants with impaired autophagy appeared similar to the control . Moreover , we found that muscle-specific autophagy inhibition had no significant effect on the motility of eat-2 animals , while the same treatment was sufficient to prevent lifespan extension in eat-2 mutants . These observations suggest a longevity role for autophagy not only in the intestine but also in body-wall muscle , as suggested in WT mice [36] , and indicate differences between the role of basal autophagy versus autophagy induced by DR in maintaining muscle integrity that are still to be elucidated . Interestingly , we observed that intestinal autophagy was required for the motility of eat-2 mutants , highlighting a cell non-autonomous function of autophagy in the intestine . One explanation for this observation is that intestinal inhibition of autophagy may cause deterioration of this organ , as implied by our Smurf assays . In turn , this could have deleterious effects on distal tissues such as muscle or motor neurons by reducing nutrient availability or other signals from the intestine . However , eat-2 animals subjected to whole-body autophagy RNAi showed significant motility defects on Day 11 , before significant intestinal barrier dysfunction was detected on Day 15 . These observations therefore indicate that intestinal dysfunction ( as measured by the Smurf assay ) may not be causally related to the motility decline observed upon whole-body autophagy gene knockdown . Rather other potential functions of autophagy in the intestine ( e . g . , turnover of intestinal nuclei [39] , or of mitochondria [44] ) may be important for the integrity of this tissue . As noted above , inhibition of intestinal autophagy in eat-2 mutants could affect motility because metabolic and/or endocrine functions of the intestine could be impaired . Such metabolic/endocrine functions of the intestine may be particularly important in dietary-restricted animals in which nutrient availability is already limited . Alternatively , intestinal inhibition of autophagy in eat-2 mutants could impair the detection/receipt of signals from distal , non-intestinal tissues that are important for maintaining intestinal function . To this point , we created eat-2; sid-1 animals expressing sid-1 from the neuronal rab-3 promoter , and we observed that autophagy knockdown in this strain significantly shortened the long lifespan of these animals ( S1 Table and S1A Fig ) . While this observation indicates a possible longevity role for autophagy in neurons , as in the intestine and body-wall muscle , we found the rab-3 promoter in addition to neurons also drove expression in the intestine in older animals , thus compromising the interpretation of our lifespan data . Further studies will be needed to elucidate the interaction between autophagy in the intestine and other tissues , and to determine how they affect intestinal integrity , motility , and longevity . In conclusion , our study provides insights into the contribution of intestinal autophagy and the importance of inter-tissue communication in the healthspan of dietary-restricted C . elegans . Further work to identify the cellular and molecular pathways that sense DR , regulate specific steps of the autophagy process , and mediate inter-tissue signaling will be important to gain a more complete understanding of the homeostatic role of autophagy in maintaining the health and longevity of an organism .
Strains were maintained and cultured under standard conditions at 20°C using E . coli OP50 as a food source [45] . For RNAi experiments , animals were grown on HT115 bacteria transformed with empty vector or plasmid encoding the appropriate gene-specific dsRNA ( see below ) . See S2 Table for strains created for and used in this study . RNAi efficiency was assessed in all eat-2 tissue-specific RNAi strains with RNAi clones encoding genes expressed in specific tissues ( S1A Fig ) , as we have done previously [46 , 47] . As expected , neither eat-2 ( ad1116 ) ; rde-1 ( ne219 ) nor eat-2 ( ad1116 ) ; sid-1 ( qt9 ) double mutants showed any phenotypes in these tests . We also tested RNAi efficiency in eat-2 ( ad1116 ) ; sid-1 ( qt9 ) strains expressing sid-1 from different promoters with a tdTomato RNAi clone against the fluorescent co-injection marker ( S1A Fig ) . In these analyses , the newly made tissue-specific eat-2; sid-1 transgenic strains all showed the expected processing ( S1A Fig ) . However , the eat-2; sid-1; rab-3p::sid-1 neuronal RNAi strain also showed signs of intestinal RNAi-processing capabilities , i . e . , on elt-2 RNAi . Consistent with this observation , we found that transgenic animals expressing a rab-3p::mCherry transcriptional reporter started showing expression in the intestine around Day 5 ( S1B Fig ) . Collectively , these observations argue that the rab-3 promoter is leaking to the intestine . In our RNAi analyses , we also observed that bli-4 and bli-3 RNAi produced phenotypes in eat-2; sid-1; gly-19p::sid-1 mutants ( S1A Fig ) . While this is consistent with the gly-19 promoter possibly expressing in the hypodermis , in addition to the intestine , we did not observe visible ectopic expression of the gly-19p::tdTomato co-injection marker in eat-2; sid-1 animals up to 10 days of age . However , we have occasionally observed old WT animals carrying the same array display ectopic expression to head neurons . In a different line of experiments , we observed GFP knockdown in the intestine of eat-2; sid-1 double mutants crossed to a translational lgg-1p::gfp::lgg-1 reporter when subjected to gfp RNAi . Although this observation suggests that some degree of RNAi processing may take place in the intestine of all strains , this did not appear to contribute to the changes in lifespan , body-cavity leakage or motility we observed in eat-2; sid-1 transgenic strains subjected to tissue-specific autophagy gene RNAi , since eat-2; rde-1 or eat-2; sid-1 double mutants did not show significant phenotypes in these assays when subjected to autophagy RNAi . For the rgef-1:gfp::lgg-1 construct , PCR products were generated from the C . elegans rgef-1/F25B3 . 3 cDNA using primers: Fwd 5′ GGG GAC AAG TTT GTA CAA AAA AGC AGG CTG GGC ATG CTA AGT GAT CTG ACC TCG CGC CCC 3′ and Rvs 5′ GGG GAC CAC TTT GTA CAA GAA AGC TGG GTG GGT ACC GTC GAT GCC GTC TTC 3′ ( 1 . 9 kb ) . The rgef-1 promoter-containing plasmid was a gift from Dr . Andrew Dillin . The PCR products were sequenced and introduced into the original lgg-1p::gfp::lgg-1 plasmid [25] to generate pMH882 , which was injected at 20 ng/μl , with 34 ng/μl of unc-122p::rfp as a co-injection marker [48] , into wild-type animals and integrated by γ-irradiation followed by outcrossing to WT ( N2 ) animals . To construct plasmids expressing sid-1 cDNA driven by various promoters , full-length sid-1 cDNA ( 2330 bp ) was cloned from first-strand worm cDNA by PCR amplification and inserted in the C . elegans expression vector pPD95 . 77 using Xma I and Age I restriction enzymes . The unc-54 3’ UTR was PCR amplified with a 5’ Age I site and a 3’ BsiW I site and cloned into the sid-1/pPD95 . 77 vector in the place of the gfp::unc-54 3’ UTR fragment . To create the intestinal expression construct , the gly-19 promoter was cloned in front of the sid-1 cDNA using Sph I and Xma I enzymes . As co-injection marker for the intestinal expression vector , tdTomato was amplified from the original tdTomato sequence ( in pCMV/tdTomato , a gift from Dr . Roger Tsien ) and cloned into the C . elegans expression vector pPD95 . 77 using Age I and Bsm I ( s260 ) , followed by cloning in the gly-19p promoter sequence using Sph I and Xma I . The sid-1-containing vector , and the tdTomato-containing co-injection marker were microinjected into the gonads of adult eat-2 ( ad1116 ) ; sid-1 ( qt9 ) hermaphrodite animals at a concentration of 10 ng/μl . The DNA mix for injection was brought to a final total concentration of 100 ng/μl using pPD61 . 125 as “Filler” DNA . Integration was performed by γ-irradiation followed by outcrossing to eat-2 ( ad1116 ) ; sid-1 ( qt9 ) animals . Plasmids expressing mCherry cDNA driven by the rab-3 promoter were purchased from Addgene and injected ( 15 ng/μg ) into WT ( N2 ) animals along with 100 μg/μg of pRF4/rol-6 marker . HT115 bacteria carrying empty vector ( plasmid L4440 ) were used as RNAi controls . The following RNAi clones were obtained from the Ahringer library ( JA ) or the Vidal RNAi library ( MV ) : atg-18/F41E6 . 13 ( MV ) , lgg-1/C32D5 . 9 ( JA ) , lgg-2/ZK593 . 6 ( JA ) . Construction of tdTomato RNAi clone: tdTomato was amplified from vector s260 and cloned into plasmid L4440 . All RNAi clones were verified by sequencing . We note an oversight in our earlier report that unc-51/Y60A3A . 1 , atg-7/M7 . 5 , and atg-18 RNAi failed to significantly reduce lifespan in eat-2 mutants [5]: atg-18 RNAi was erroneously included in this list . For RNAi experiments , HT115 bacteria were grown in liquid LB medium containing 0 . 1 mg/ml carbenicillin ( BioPioneer ) , and then 80 μl samples were spotted onto 6 cm NGM plates supplemented with carbenicillin and allowed to grow for 1–2 days at room temperature . dsRNA expression was induced by addition of 80 μl 0 . 1 M IPTG ( Promega ) to the bacterial lawn , and eggs or adult worms were then transferred to the plates . For whole-life RNAi , animals were synchronized by hypochlorous acid treatment and the eggs were placed on NGM plates seeded with dsRNA-expressing bacteria . For adult-only RNAi , animals were synchronized by hypochlorous acid treatment , eggs were placed on NGM plates seeded with OP50 bacteria and allowed to hatch , and worms were then transferred to NGM plates seeded with dsRNA-expressing bacteria on Day 1 of adulthood . Lifespan was measured at 20°C as previously described [49] . In brief , ~100 synchronized animals were examined every second day of adulthood and were scored as dead if they failed to respond to gentle prodding with a platinum wire pick . Integrated , outcrossed strains were used for tissue-specific RNAi experiments . Animals were censored from the analysis if they became desiccated on the edge of the plate , escaped , ruptured , or suffered from internal hatching . Statistical analysis using the log-rank ( Mantel-Cox ) method was performed with Stata software ( StataCorp ) . We used multiple versions of eat-2 ( ad1116 ) mutant strains ( i . e . , CF1908 , MAH95 , and MAH458; S2 Table ) , which were regularly outcrossed due to occasional increases in male progeny or sterility in the population . We note that multiple methods of DR have been shown to increase lifespan in C . elegans [50] . While we have shown that several of these paradigms cause changes in the GFP::LGG-1 reporter [5] , it remains possible that the tissue-specific role and regulation of autophagy in animals subjected to DR in different ways may differ . For tissue-specific analyses , WT or mutant animals expressing GFP::LGG-1 in the muscle and intestine ( from the lgg-1 promoter , DA2123 [51] ) or in neurons ( from the rgef-1 promoter , see above ) were raised at 20°C , and GFP::LGG-1 punctae were visualized and counted on Day 3 of adulthood . Day 3 was chosen for our autophagy analyses since this was the first time point in motility tests . Animals were mounted on a 2% agarose pad in M9 medium containing 0 . 1% NaN3 , and Z-stack images were taken at 0 . 6 μm intervals using an LSM Zeiss 710 scanning confocal microscope at 630× magnification . GFP::LGG-1 punctae were quantified as follows: for the muscle , punctae in one 1000 μm2 area per 0 . 6 μm slice per animal; for the neurons , total punctae between the pharyngeal bulbs in one 0 . 6 μm slice per animal; for the intestine , punctae in one proximal cell ( with visible nucleus ) per 0 . 6 μm slice per animal . For vha15/V-ATPase RNAi experiments , animals expressing GFP::LGG-1 from the endogenous reporter were raised from L4 larval stage to Day 4 of adulthood on bacteria containing empty vector or expressing vha-15/T14F9 . 1 dsRNA ( JA ) . Animals were mounted as described above and GFP::LGG-1 punctae were quantified using Zeiss Imager Z1 fluorescence microscope , as previously described [27] . GFP::LGG-1 punctae were scored in the proximal region of the intestine , encompassing the first three pairs of intestinal cells starting beyond the pharyngeal grinder . Animals were raised at 20°C , collected on Day 3 of adulthood , and prepared for analysis as described [27] . Grids were viewed using a Philips CM10 electron microscope ( FEI ) equipped with a Morada digital camera ( Olympus ) and iTEM software ( Olympus SIS ) . Autolysosomes were identified as vesicles containing material undergoing degradation in 5000-fold magnified micrographs of a cross-section of the intestine . Statistical analysis was performed using GraphPad Prism . LysoTracker Red DND-99 ( Invitrogen ) was added at 2 μM to NGM medium before plates were poured . Animals were raised at 20°C , placed on LysoTracker-supplemented plates on Day 1 of adulthood , and examined on Day 5 . Animals were mounted on a 2% agarose pad in M9 medium containing 0 . 1% NaN3 and analyzed using a Zeiss Imager Z1 fluorescence microscope . Image analysis was performed by selecting a 51 . 3 × 51 . 3 pixel area of the anterior intestine and measuring the integrated intensity of LysoTracker fluorescence in images taken after a 100 ms exposure ( ImageJ software; National Institutes of Health ) . Statistical analysis was performed using GraphPad Prism on data acquired from ~10 animals per condition . Total RNA was isolated from an age-synchronized population of ~2000 animals flash frozen in liquid nitrogen on Days 1 or 5 of adulthood . RNA was extracted with TRIzol ( Life Technologies ) and purified using a Qiagen RNeasy kit with an additional DNA digestion step performed with a Qiagen DNase I kit . M-MuLV reverse transcriptase and random 9-mer primers ( New England Biolabs ) were used for reverse transcription of 1 μg of RNA per sample [52] . Quantitative PCR was performed using SYBR Green Master Mix and a Roche LC480 LightCycler . A standard curve was included for each primer using serial dilutions of a mixture of cDNAs , and the observed CT values were converted to relative values according to the standard curve obtained with the relevant primers . Target gene mRNA levels were normalized against the geometric mean mRNA levels of the housekeeping genes ama-1 ( large subunit of RNA polymerase II ) and nhr-23 ( nuclear hormone receptor ) [29 , 53] . Primer sequences can be found in S3 Table . Each biological sample was analyzed with three technical replicates . The mean ± SEM for each mRNA was calculated and the data were analyzed by one-way ANOVA using GraphPad Prism . Animals were raised as described above for lifespan assays . On specific days , ~10 animals were removed from the NGM plates and suspended for 3 h in liquid cultures of standard OP50 bacteria ( grown overnight ) mixed with blue food dye ( Spectrum FD&C Blue #1 PD110 , 5 . 0% wt/vol in water ) . We detected no reduction in the rate of dye uptake in eat-2 mutants , despite their reduced pumping rate [21] . Animals were then transferred to NGM plates seeded with OP50 bacteria and analyzed for the presence or absence of blue food dye in the body cavity using a Leica DFC310 FX microscope under 40x magnification . For each time point , three or more independent experiments were carried out , each with 8–10 animals per strain and/or treatment . Data were analyzed using GraphPad Prism . We often observed an age-related increase in the frequency of WT and eat-2 mutant worms with blue food dye in the germline , increasing from ~10% of worms on Day 7 to ~50% on Day 15 . It is unclear whether the dye entered the germline through the body cavity or the vulva; however , many animals had dye in the germline but not the body cavity , suggesting entry through the vulva . For this reason , we included in the analysis animals with dye present in both the germline and body cavity , and excluded animals with dye present only in the germline . Multiple versions of outcrossed eat-2 ( ad1116 ) mutants were used for this assay ( see Lifespan analysis section ) . Animals were raised as in lifespan assays . On specific days , animals were removed from NGM plates , suspended individually in M9 medium , and the number of body bends in a 30 s period was counted . One body bend was counted every time the part of the worm just behind the pharynx reached a maximum bend in the opposite direction from the bend last counted . Data were analyzed using GraphPad Prism . Multiple versions of outcrossed eat-2 ( ad1116 ) mutants were used for this assay ( see Lifespan analysis section ) . Animals were raised at 20°C , moved to RNAi plates on Day 1 and analyzed on Day 15 . Animals were mounted on a 2% agarose pad in M9 medium containing 0 . 1% NaN3 and analyzed using a Zeiss Imager Z1 fluorescence microscope . Image analysis was done manually similar to a previous study [54] , and we scored for fragmentation , irregular orientation or broken MYO-3::GFP-positive structures . The degree of these phenotypes was scored on a scale between 1–3 ( 1 least severe , 3 most severe ) , and the sum of all scored events were averaged . The analysis was repeated by another person and found to give similar results . Statistical analysis was performed using GraphPad Prism on data acquired from 6–10 animals per condition . | Dietary restriction ( DR ) without inducing malnutrition has robust beneficial effects on lifespan in many species , including humans . The cellular recycling process of autophagy contributes to DR-mediated longevity . Autophagy is triggered by nutrient scarcity and increases the degradation of cytosolic molecules and organelles in the lysosomes . Using the nematode Caenorhabditis elegans as a model organism , we previously showed that genes involved in autophagy are required for lifespan extension through DR; however , it is not clear whether autophagy in individual tissues plays critical roles in DR-mediated longevity . Here , we investigated the contribution of autophagy in genetically dietary-restricted eat-2 mutants . Our major findings include: ( i ) Inhibition of autophagy in the intestine prevents the long lifespan observed in eat-2 mutants; ( ii ) the intestine of eat-2 mutants contains an expanded lysosomal compartment and flux assays indicate increased autophagosome turnover , consistent with elevated autophagy in this tissue; ( iii ) intestinal autophagy is required for the improved intestinal integrity observed in eat-2 mutants; ( iv ) autophagy inhibition impairs motility in older animals; and ( v ) inhibition of autophagy in the intestine accelerates the motility decline in eat-2 mutants . Collectively , these studies suggest a critical role for intestinal autophagy in dietary-restricted animals , and highlight the importance of this process in maintaining fitness and longevity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"death",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"autophagic",
"cell",
"death",
"lysosomes",
"rna",
"interference",
"pathology",
"and",
"laboratory",
"medicine",
"caenorhabditis",
"pathogens",
"cell",
"processes",
"cloning",
"animals",
"animal"... | 2016 | Intestinal Autophagy Improves Healthspan and Longevity in C. elegans during Dietary Restriction |
Single-cell variability in gene expression is important for generating distinct cell types , but it is unclear how cells use the same set of regulatory molecules to specifically control similarly regulated genes . While combinatorial binding of transcription factors at promoters has been proposed as a solution for cell-type specific gene expression , we found that such models resulted in substantial information bottlenecks . We sought to understand the consequences of adopting sequential logic wherein the time-ordering of factors informs the final outcome . We showed that with noncommutative control , it is possible to independently control targets that would otherwise be activated simultaneously using combinatorial logic . Consequently , sequential logic overcomes the information bottleneck inherent in complex networks . We derived scaling laws for two noncommutative models of regulation , motivated by phosphorylation/neural networks and chromosome folding , respectively , and showed that they scale super-exponentially in the number of regulators . We also showed that specificity in control is robust to the loss of a regulator . Lastly , we connected these theoretical results to real biological networks that demonstrate specificity in the context of promiscuity . These results show that achieving a desired outcome often necessitates roundabout steps .
A fundamental question in systems biology is how a small number of signaling inputs specifies a large number of cell fates through the coordinated expression of thousands of genes . This problem is especially challenging given that gene regulatory and other types of networks in biology tend to be highly interconnected and their regulators promiscuous , with regulators affecting multiple targets and targets being affected by multiple regulators . Examples of this architecture include: transcription factor binding networks in bacteria [1] , yeast [2 , 3] , plants [4] , and animals [5 , 6]; cellular signalling pathways involved in growth and differentiation [7–9]; the interactome of protein kinases and phosphatases [10 , 11]; and synaptic connections between different layers of the brain [12] . Furthermore , because the targets and regulators are often well-mixed and mutually accessible in the cell , most actions are likely to have nonspecific and undesired effects . At the same time , regulatory molecules drive networks to a large number of highly specific outcomes or cell fates . Although there are approximately four hundred canonical cell types in the adult human [13] , recent single-cell RNA expression profiling experiments in the developing embryo [14 , 15] , brain [16] , hematopoietic system [17 , 18] , and other organs [19 , 20] , have indicated that there may be thousands more . Given there are only a few signaling pathways used in metazoan development [21 , 22] , understanding how cells reach their final outcomes when there are fewer regulators than fates and/or targets is an unsolved problem . One extensively studied solution for the control of promiscuous gene networks is combinatorial binding of DNA-binding transcription factors ( TFs ) at the promoter [23–31] . At the level of individual promoters , combinatorial binding ensures that individual genes are ON only when specific combinations of TFs are present ( Fig 1A ) . However , on the genome level , combinatorial regulation restricts which sets of genes may be ON at the same time . For example , using AND logic , gene H in Fig 1A is only ON in the case that the three TFs K1 , K2 , and K3 are concurrent at the H promoter; but these stringent requirements mean that H can never be transcribed independently of the less highly-regulated genes A-G . ( A similar conclusion holds for OR logic . ) In fact , using combinatorial control , there is a one-to-one correspondence between configurations of the targets and configurations of the regulators . As shown in Fig 1A , the ON/OFF states of 3 TFs uniquely define the binding combinations at 23 = 8 promoters . A similar conclusion holds when the regulators are expressed in a graded fashion . This one-to-one correspondence is the fundamental limitation of combinatorial regulation: it requires an equal number of regulators and independently controlled targets and/or cell fates . Applied to embryonic development , combinatorial control requires that hundreds or thousands of cell-type specific TF combinations be generated in a spatially precise manner at the start . However , the combinatorial scheme does not explain how the TF states are regulated in the first place , and thus it offers no new insight into how cell fate is specified . The limitations of combinatorial logic can also be understood from an information theoretic point of view . In particular , it is impossible to specify arbitrary cell fates if the regulatory layer bottlenecks the capacity of the targets to receive messages from extracellular signals . It is known that some ten to twenty types of signals [21 , 22] converge onto membrane-bound regulators in many different combinations , permitting messages to be passed to the downstream targets . Much of this information stands to be lost , however , if the network relies on combinatorial logic alone: the regulatory layer simply cannot transmit messages in their entirety if there are more signals than regulators . Thus , combinatorial logic strongly circumscribes what fates are ultimately reachable . Cell fate information is lost not only if the signals are more numerous than the regulators , but also if the connections between signals and regulators are promiscuous ( Fig 1B ) . When different signals activate the same regulators ( Fig 1Bi ) , certain signaling inputs become redundant . On the other hand , when same signal activates different regulators ( Fig 1Bii ) , some of the regulators become redundant . One may determine by direct enumeration exactly how redundancy decreases the number of configurations available to the targets ( Materials and Methods Sections 1 and 2 ) . These preliminary conclusions are at odds with the observation that signaling molecules are deployed over time in a complex code [32] . How do these messages in the signal space reach the targets if the regulatory layer imposes a bottleneck on information flow ? In addition , feedback regulation—a common feature of regulatory networks—exacerbates information bottlenecks when coupled with combinatorial logic . Stated another way , feedback merely widens the basin of attraction of certain promoter configurations at the expense of the number of distinct configurations . In Fig 1C , constitutive expression of K1 by C means that C is never ON independently of the targets regulated by K1 . Thus , the number of accessible configurations decreases from 8 to 6 without allowing new target configurations to be explored . We need an alternative to combinatorial logic in cell fate specification that overcomes information bottlenecks . Here , we considered time-ordered control schemes , which we refer to as sequential logic . In this scheme , regulators can be applied in a stepwise manner; the entire sequence matters , so the final configurations can differ if the same regulators are permuted in time . In order for different temporal sequences to carry distinct information , the actions of the regulators must be noncommutative . This is the case , for example , when a regulator protects its targets from the action of another regulator , as when loci recruited to repressive chromatin compartments are protected from further modification [33 , 34] . While it is not surprising that noncommutative sequences like this result in different outcomes at the single promoter level , these simple mechanisms may have nontrivial implications for regulation at the genome level . In particular , noncommutativity permits the same regulators to be used at different times with distinct effects . This is seen in development when ubiquitous signaling molecules like FGF family members exert different effects depending on the time and context of their expression [35–38] . Reuse of factors could greatly expand the information capacity of the major signaling pathways . A number of examples show that noncommutativity may be a general strategy in other areas of biology . In hematopoietic stem cells , activation of GATA2 and C/EBPα in different orders results in different cell fates [39] . In neurobiology , different temporal orderings of the same inputs lead to distinct firing patterns [40–42] . In the field of synthetic biology , a DNA switch was developed that could detect the order in which invertase enzymes were applied [43] . And in evolutionary biology , the order in which mutations arise was recently implicated in determining a genotype’s fitness [44–47] . There is also accumulating evidence for sequential logic in transcriptional control: signaling molecules and TFs in mammalian cells , including ERK [48] , NF-κB [49 , 50] , p53 [51] , as well as in yeast [52–54] have been observed to pulse , suggesting that TF timing may be used to control the transcriptional state of the cell . By applying sequential logic , we show that , even in complex and promiscuously regulated networks , specific target configurations can be reached using a temporal sequence of regulators . In particular , we consider two models inspired by ( i ) kinase/neural networks and ( ii ) chromosome folding and show analytically that both scale super-exponentially . We further show that noncommutative networks are robust to the loss of regulators , suggesting a mechanism for regulator evolution . We also show that regulators induce different orbits in expression space , which is related to the number of networks that can be controlled in parallel . We conclude by discussing how these models apply to interconnected networks in and outside biology and by providing possible experimental tests of the theoretical concepts . Theorems and proofs are given in the Materials and Methods .
To consider how time-ordered sequences of regulators can specifically control groups of targets , we begin by analyzing a generic two-layer network that is an extension of combinatorial logic ( Fig 2 , Materials and Methods Sections 1 and 2 ) . In this model , each regulator controls multiple targets , and each target is accessible to any of its regulators . The model is meant to be analogous to the cellular environment wherein regulators and targets are well-mixed . For example , targets could be substrate proteins capable of multi-site phosphorylation [55 , 56] , and regulators the kinases and phosphatases . Targets could also be neurons and regulators their upstream excitatory and inhibitory inputs [12] . We denote by K the set of activators ( i . e . kinases ) and P the set of deactivators ( i . e . phosphatases ) . Each target has a ladder of ( integer-valued ) states , and together the states of the targets are a configuration of the network . ( This distinction is in contrast to the common usage of “state” as a gene expression vector . ) An additional parameter , the threshold T , determines the number of rungs on the ladder . Regulators ratchet the targets through their states , and only targets that have reached threshold will be ON at the end of a sequence of regulators . If each target in the group can be controlled by a unique combination of K’s and P’s , what ON/OFF configurations are possible ? In this model , termed the ratchet network ( Fig 2A ) , each of n K’s and m P’s control N = ( n l n ) ( m l m ) unique targets , with the connectivity parameters ln and lm specifying the number of regulators to which each target connects . Consider the sequence K1 K2 P1 acting on the targets A , B , C , and D ( Fig 2B ) . In the final configuration , B and D are ON together even though no single K connects to both , and A and C are OFF , even though both share and activator with B and D . Therefore , this simple model illustrates the important point that similarly regulated targets can be in independently controlled using sequential logic . With threshold T = 1 , not all configurations are reachable . Observe that there is no way to specifically activate A and D while leaving B and C OFF . This result is surprising given that A and D share no regulators: specificity depends on the network as a whole , not just individual targets . By going to T = 2 , the forbidden configuration becomes accessible ( Fig 2C ) , along with all ON/OFF states ( below ) . The model described above can be formalized as a combinatorial object that we refer to as the connectivity matrix A . This formulation is useful because it is amenable to studying scaling , and it permits a direct comparison between noncommutative ratchet networks and standard combinatorial logic . For the interested reader , the models considered in this paper have a universal formulation as noncommutative matrix operators on the vector space of configurations ( Materials and Methods Section 9 ) . Typically , the state of N targets is represented as an N-dimensional vector . If each target is controlled by a unique ( Ki , Pj ) pair ( i . e . ln = lm = 1 ) , the N = nm-dimensional vector can be re-formulated as an n × m matrix A = P 1 ⋯ P m K 1 ⋮ K n ( A 1 , 1 ⋯ A 1 , m ⋮ ⋱ ⋮ A n , 1 ⋯ A n , m ) ( 1 ) where each entry Ai , j ∈ {0 , 1 , … , T} is the state of the target regulated by Ki and Pj . For example , the connectivity matrix for the network in Fig 2 is A = P 1 P 2 K 1 K 2 ( A B C D ) . ( 2 ) In general , a regulator may connect to multiple targets ( i . e . ln , lm > 1 , see below ) , so that each entry of A may be thought of as an M-dimensional vector ( M determined in Materials and Methods Section 1 ) . It turns out that this is an unnecessary complication; we instead let each Ai , j = 1 if at least one of the M targets regulated by Ki and Pj is ON , and Ai , j = 0 only if all M targets are OFF . In this formulation Ki and Pj are raising and lowering operators that map n × m matrices to n × m matrices via the rules K i A i , j = A i , j + 1 if A i , j < T A i , j if A i , j = T P i A i , j = A i , j - 1 if A i , j > 0 0 if A i , j = 0 . ( 3 ) From Eq ( 3 ) , any sequence Ki1 Ki2⋯Kik of all K’s is commutative , because any target controlled by t ≤ k of the K’s will be in state t ≤ T at the end of the sequence , regardless of the order . A similar argument holds for the P’s . However , sequences consisting of both K’s and P’s are in general noncommutative . This is due to edge effects when Ai , j = 0 or T . If Ai , j = T , for example , then Ki Pj results in Ai , j = T − 1 , whereas Pj Ki gives Ai , j = T . Therefore , A gives insight into both the configuration of the targets and the noncommutativity of the regulators . The problem of determining the number of accessible configurations in a network is reduced to finding the number of matrices satisfying certain patterns . For example , combinatorial logic with T = 1 corresponds to the special case in which the only sequences are the 2n combinations of the n K’s . In an n × 1 connectivity matrix , activating Ki corresponds to turning all 0’s in row i into 1’s . There are 2n matrices generated by this procedure . More complicated cases of combinatorial logic can be studied this way ( Materials and Methods Section 2 ) , but it turns out that the total number of network configurations is always less than 2n + m , with n + m the total number of regulators . This is important because noncommutative models can bypass the exponential limit . We used the connectivity matrix representation of the ratchet network to determine the scaling as function of the number of regulators n and m , with each target connected to a unique ( K , P ) pair ( i . e . ln = lm = 1 ) and the threshold T = 1 . Ki turns 0’s to 1’s in row i and Pj turns 1’s to 0’s in column j . The rules are consistent with the one-pot reaction model in which all substrates receptive to Ki are promoted when Ki is active . For example , the sequence K1 K2 P1 in Fig 2B can be recast as 0 0 0 0 → K 1 1 1 0 0 → K 2 1 1 1 1 → P 1 0 1 0 1 . ( 4 ) The main result is that A must avoid the patterns ( 1 0 0 1 ) and ( 0 1 1 0 ) in any 2 × 2 sub-block ( Materials and Methods Section 3 ) . Brewbaker [57] enumerated the n × m binary matrices avoiding these patterns and showed that they scale as the poly-Bernoulli numbers [58] B m − n = B n − m = ∑ j = 0 m ( − 1 ) ( n + j ) j ! ( j + 1 ) n { n j } = ∑ j = 0 min ( n , m ) ( j ! ) 2 { m + 1 j + 1 } { n + 1 j + 1 } , ( 5 ) where { nj} is a Stirling number of the second kind , defined combinatorially as the number of ways to put j labelled balls into n unlabelled boxes such that no box is empty [59] . These numbers scale not quite as fast as 2N = 2nm , but much faster than 2n + m , the maximum number of states in the equivalent combinatorial network ( Fig 2D ) . Thus , a simple time-sequence model is able to generate super-exponential scaling . Are more configurations accessible if multiple activation events are needed before reaching threshold ? For example , neurons require the summation of multiple excitatory inputs to reach action potential , and proteins need to be phosphorylated at multiple sites before they are activated [55 , 56] . We found that by increasing the threshold to T = 2 , all 2N ON/OFF configurations of the N targets become reachable . In the connectivity matrix formulation , ( 1 0 0 1 ) and ( 0 1 1 0 ) are no longer forbidden , which we show with an inductive proof ( Materials and Methods Section 4 ) . This scaling law ( Fig 2D ) , achieves the maximum of reachability and specificity; it far exceeds the scaling 2n + m of the combinatorial model . Being able to reach the entire ON/OFF space of N targets is overkill for most biological networks , which only display a relatively small number of stable configurations . The major implication of this result is that multiple levels of activity permit more targets to be controlled independently . As sequential logic allows a large number of configurations to be reached in a complex network , we asked whether increasing the connectivity of the network ( ln and lm ) can maintain the specificity of the network while making it robust to the loss of a regulator . This is potentially relevant to evolution of biological networks , because redundant connections allow the network to repurpose regulators for new functions without severely impairing existing ones [60] . In the ratchet model , an increase in the connectivity parameters to ln = 2 K’s and lm = 2 P’s permits multiple targets to share a common ( K , P ) pair ( Fig 3A ) . The connectivity matrix incorporating the extra links in the network in Fig 3A is P1 P2 P3A=K1K2K3 ( ABDEACDFCDEFABGHACGIBCHIDEFGDFGIEFHI ) . ( 6 ) Now that each entry of A is a group of M > 1 targets , it makes sense to track the state of the group as a whole with a single number Ai , j . Even though a target appears in multiple entries of A , the rules prevent a regulator from altering the state of groups at remote locations ( e . g . K1 cannot change the state of the group at A2 , 2 ) . We prove in the Materials and Methods that all sequences using at least n − ln + 1 K’s and m − lm P’s are redundant with shorter sequences ( Fig 3B and 3C , Materials and Methods Section 5 ) . For example , the sequences K1 K2 K3 is required to turn ON all targets in the case ln = lm = 1 , but if ln = lm = 2 , the shorter sequences K1 K2 , K1 K3 , and K2 K3 have the same effect . We derived a recursive formula that eliminates the redundant sequences in each ( n , m , ln , lm ) instance to derive the number of sequences in ( n , m , ln + 1 , lm ) and ( n , m , ln , lm + 1 ) ( Fig 3D and S2 Fig ) . The formula agreed exactly with an algorithm designed to find all minimal length sequences ( Materials and Methods 5 ) . Notably , increasing ln , lm reduced the number of configurations . We observed a similar effect in combinatorial logic ( S1 Fig ) . To investigate the robustness of sequential logic networks , we studied the effect of deleting regulators in increasingly connected networks on the number of reachable configurations ( Fig 4A ) . We hypothesized that sequences that activate similar subsets of targets should be able to recoup permanently lost configurations . To test this , we computed the normalized correlation coefficient between configurations in the network using all K’s ( the full network ) and configurations in the network without K1 ( the impaired network ) , subject to the constraint that those configurations were reached using longer sequences ( Fig 4B ) . To focus on the recoverable fraction , we deleted all configurations that had an exact match . Highly similar configurations ( yellow ) clustered to the right of the plot , indicating that longer sequences can be used to recover lost configurations . How similar are the recouped configurations ? As connectivity increased , the maximum similarity became increasingly concentrated above 0 . 8 ( Fig 4C ) . There is generally a tradeoff between reachability and the size of the fraction above 0 . 8 ( Fig 4D ) . The tradeoff is nonlinear , however: using ln = 2 gave the greatest increase recoverability for the smallest loss of configurations , showing that an intermediate level of redundancy can buffer the network to loss of regulators . The above analyses demonstrate that specificity of control is not compromised when regulators are lost or repurposed in heavily interconnected networks . In the ratchet model , all targets are accessible to their regulators at all times . However , in some cases targets may be shielded from regulators: for example , genes can be silenced by sequestration in various nuclear compartments [61 , 62] . This was seen in a landmark study by Filion et al [63] , who used a DNAse accessibility assay to show that genes associate with different regulators depending on their chromatin “color” or accessibility status . To study the effect of accessibility and silencing on activating specific subsets of genes , we constructed the following sequestration model . In addition to the OFF state 0 and the ON state 1 , each target/gene is endowed with additional orthogonal states 2 to n ( allowing for a total of 2n − 1 − 1 genes ) . If RNA polymerase ( RNAP ) is associated with K1 , what genes can be independently activated ? In this model ( Fig 5 ) a regulator Ki promotes targets in the 0 state to state i , and Pi returns targets in state i to 0 . Any target in state i is protected from regulators other than Pi . As an example of gene regulation on a three-dimensional chromosome ( Fig 5A ) , the sequence K3 K4 K1 P3 P4 first clusters all genes having a 3 in a repressive compartment , and then the remaining genes having a 4 in another repressive compartment . The net effect is that RNAP can only act on the gene represented by {1 , 2} . We represent this abstractly as a configuration vector of k-armed targets ( Fig 5B ) , where each entry corresponds to the state {0 , 1 , … , n} of a gene able to access k ≤ n of the states ( see below for a mathematical description of the model ) . Therefore , protected states in the sequestration model allow genes to be transcribed specifically in a well-mixed environment . We derived ( see below ) that the number of reachable configurations scales with the number of regulator pairs n as f n = 2 2 n - 1 - 1 - ∑ m = 2 n - 1 n - 1 m 2 ∑ k = 3 m m k - 1 - 1 2 ∑ k = 3 m n - 1 k - 1 - m k - 1 . ( 7 ) For n = 1 , 2 , 3 , 4 , 5 , 6 , this formula gives f ( n ) = 1 , 2 , 7 , 89 , 16897 , 780304385 ( Fig 6 ) . We also relaxed the constraint that all genes have a 1 state ( allowing for a total of 2n − 1 genes ) and found that the number of configurations scales as cn = 2 , 7 , 94 , 37701 with n = 1 , 2 , 3 , 4 . The full model does not have an analytical solution , but it does have upper and lower bounds related to Eq ( 7 ) ( Materials and Methods Section 7 , S3 Fig ) . Combinatorial scaling laws of this sort are not uncommon [44 , 64 , 65] . Edwards and Glass [64] saw an explosion in the number of states when studying trajectories on n-cubes , and Green and Rees [65] saw a super-exponential jump when enumerating certain types of nonrepeating sequences on n letters . Furthermore , a similar small number ( four ) of factors are necessary and sufficient to reprogram fibroblasts to stem cells [66] . Together , these results indicate that sequences can far exceed the 2n limit set by combinatorial regulation , and that only a few regulators are necessary to make large changes in the configuration of a cell . The sequestration network with n regulator pairs ( referred to as the n-network ) is described using the 1 × 2n − 1 configuration vector x . This is a simpler description than the connectivity matrix because a target affected by Ki is necessarily affected by Pi . The entries of x are the states of each target g able to be controlled by k ≤ n of the regulator pairs . Each target g is is a list {0 , i1 , … , ik} of the k regulators to which it responds . Because of their radial appearance , such targets are said to have k arms ( see Fig 5B ) . The regulators act on x according to the rules K i x g = x g + i if i ∈ g and x g = 0 x g else P i x g = 0 if i ∈ g and x g = i x g else . ( 8 ) Eq ( 8 ) guarantees that the regulators are orthogonal in the sense that a target in state j is protected from Ki and Pi if i ≠ j; and also idempotent in that K i 2 = K i . Furthermore , sequences of regulators are noncommutative unless the only actions are P’s . This is a consequence of the fact that P’s put all affected targets into the 0 state . Although these rules are different from the ratchet model , a formulation exists that generalizes the K’s and P’s to matrix operators consistent with both models ( Materials and Methods Section 9 ) . If x is restricted to the 2n − 1 − 1 targets all able to be regulated by K1 and at least one other K , the network is said to be reduced; otherwise we say x is full . This distinction was used in Fig 5 . A one-coloring is a configuration of x that uses only one of the states and 0 . For example , the configuration x = ( 1 , 0 , 0 , 1 , 1 , 0 , 0 ) in the full n = 3-network is a one-coloring of 1; so is the reduced network formed by ( x4 , x5 , x7 ) = ( 1 , 1 , 0 ) . This concept is easily extended to k > 1-colorings . One-colorings are particularly important because they resemble the ON/OFF configurations of genes in an RNA-seq experiment , and we would like to know how many such configurations can be reached . As in the ratchet model , finding the accessible states of the sequestration network amounts finding restricted patterns in x . We determined that the restricted one-colorings are those that violate a property referred to as connectivity ( Materials and Methods Section 7 ) . A configuration of x is said to be connected if all k > 3-arm targets g i ( k ) = { 0 , i 1 , … , i k } match the state of at least one of k of the 2-arm targets {0 , i1 , i2} , … , {0 , ik − 1 , ik} sharing the indices i . If the network is reduced , no k-arm target may be in the 1 state when all of 2-arm targets with which it overlaps ( i . e . shares an index other than 1 ) are in the 0 state . This restricts the one-colorings and suggests a method to determine the scaling law for the model in Fig 5 . As an example , in the n = 4 network on the reduced set of 23 − 1 targets illustrated in Fig 5 , {0 , 1 , 3} and {0 , 1 , 4} both being 0 constrains {0 , 1 , 3 , 4} to be 0 as well . Furthermore , even though {0 , 1 , 2} is in the 1 state , {0 , 1 , 2 , 4} and {0 , 1 , 2 , 3 , 4} may be 0 . It is only the two-arm targets that constrain the possible configurations: for example , the longer sequence K2 K4 P2 K3 K2 P4 K1 P3 K4 P1 K3 P4 K1 P2 P3 obtains the state x = ( 0 , 0 , 1 , 0 , 0 , 0 , 1 ) in which only the targets {1 , 4} and {0 , 1 , 2 , 3 , 4} are ON , showing that {0 , 1 , 2 , 3 , 4} need not be in the same state as {0 , 1 , 2 , 3} , {0 , 1 , 2 , 4} , or {0 , 1 , 3 , 4} . In Fig 6A and 6B we show the allowed states and the sequences that generate them for n = 4; there are 90 out of a possible 224 − 1 − 1 = 128 configurations . There are 22n − 1 − 1 one-colorings on 2n − 1 − 1 targets . How many of these violate the connectivity rule ? Suppose there are m 0’s among the 2-arm targets . If m = 1 , then ( m k - 1 ) = ( 1 k - 1 ) = 0 of the k ≥ 3-arm targets are constrained to be 0 , as there is always another 2-arm target ( in the 1 state ) that each k-arm target can match . If m > 1 and m − 1 < k , however , then ( m k - 1 ) > 0 , so ( m k - 1 ) k-arm targets whose states {i1 , … , ik − 1} are completely contained within the set of 2-arm targets {0 , 1 , j1} , … , {0 , 1 , jm} must be 0 . Hence in any violation of the connectivity rules at least one of ∑ k = 3 m ( m k - 1 ) k-arm targets will be in the 1 state and the remaining ∑ k = 3 m ( n - 1 k - 1 ) - ( m k - 1 ) k-arm targets will be 0 or 1 . Furthermore , there are ( n - 1 m ) ways of specifying m 0’s , so the total number of violations is ∑ m = 2 n - 1 n - 1 m 2 ∑ k = 3 m m k - 1 - 1 2 ∑ k = 3 m n - 1 k - 1 - m k - 1 . ( 9 ) Subtraction from 2n − 1 − 1 gives Eq ( 7 ) . Until now we have considered the reachable space of a single group of targets each starting in 0 . An ensemble of networks could each start with their targets in some arbitrary state , and when a sequence is applied to the ensemble the different networks will in general span different configurations . Determining the number of orbits ( defined precisely in Materials and Methods Section 8 ) within the set of possible configurations tells us how many networks can be controlled in parallel . Enumerating the reachable space for both the ratchet and sequestration networks involved finding configurations that violated at least one rule . If two configurations have distinct violations , then there is no way they can communicate using the regulators . Therefore , the different orbits are the groups of configurations having the same forbidden patterns . It is possible that a violation could be alleviated by an action that changes the state of an offending target , so we require that each orbit be immune to a subset of the regulators . This could be achieved in biological networks by locking targets in protective chromatin states or by shutting down certain cellular receptors . We determined a recursive formula for the number of orbits in the ratchet network for an arbitrary n , m ( Materials and Methods Section 8 ) . In Fig 7A we plot the orbits for the n = 4 , m = 2 case . There is one large component of size B n - m and several smaller orbits of size B i - j with i ≤ n and j ≤ m . There are only a handful of singleton orbits in Fig 7A , but the number of isolated states dominates the space as n , m increase . We were unable to find a similar solution for the sequestration network because we lack a general solution for the number of states in the main orbit . However , Fig 7B shows the computationally discovered orbits for the full network on 2n − 1 targets . A nontrivial feature is that there are orbits which use all pairs of regulators , but which do not communicate with the main orbit . For example , the sequence K2 K3 from x = ( 1 , 0 , 0 , 0 , 0 , 0 , 1 ) reaches the same configuration as the sequence K1 starting from x = ( 0 , 2 , 3 , 2 , 2 , 3 , 0 ) ; these configurations are part of the same orbit because both violate the connectivity rule between x7 = {0 , 1 , 2 , 3} and the 2-arm targets x4 , x5 , and x6 . Another observation is that some pathways cannot be reversed by a legal action in the ratchet network orbits ( indicated by a directed arrow in Fig 7 ) , whereas there always exists a reversible path between configurations in the sequestration network orbits ( no arrowheads ) . It can be proved that this is true in general for the sequestration network ( Materials and Methods Section 8 ) . This feature permits orbits to be found computationally by looking for reversible one-step paths in the entire configuration space . The orbits are one explanation for the phenomenon the same signal can cause cells to behave differently [38] . More generally , the orbits demonstrate an intriguing symmetry between the targets responding to a restricted subset of the regulators on one hand , and the orbits restricted to the same subset on the other .
In this paper we first show how noncommutative , sequential logic can relieve information bottlenecks in multilayer networks . Bottlenecks in combinatorial logic may occur whenever a downstream layer has fewer elements than the layer upstream , which poses the problem of how networks process complex signals without loss of information . Noncommutative solutions such as the ratchet and sequestration models , in which the number of configurations scales super-exponentially in the number of regulators ( Eqs ( 5 ) and ( 7 ) ) , permit longer , more complex messages to reach the targets via information “pulses . ” These pulses encode a large diversity of signals into configurations of the targets that would otherwise be lost using combinatorial logic . Noncommutativity has long been recognized as a central concept in control theory , because it allows systems with few controllers to explore a broader configuration space . For example , one generates z rotations in 3D by R − x Ry Rx , so control over z is generated by a pulse sequence of rotations in x and y , as in airplane control where roll and pitch generate yaw [67] . Infinitesimal motions in the form of generating matrices are translated into flows in a vector space by exponentiation . Because matrix multiplication is noncommutative , composition of flows is not simply the addition of generators , but rather a higher order polynomial of commutators of the generators given by the Baker-Campbell-Hausdorff formula [68] . Noncommutativity also appears in experimental physical chemistry where pulse sequences can prepare spin systems in nontrivial population configurations [69] . A formal description of these phenomena is based on the Heisenberg picture of quantum mechanics , wherein evolution of a system of many variables is given by a differential equation involving the commutator of a Hamiltonian operator . The significance of noncommutative control for systems biology is that it becomes possible to independently control targets that would otherwise be activated by the same promiscuous regulator . In this paper , we argue that noncommutative sequences permit control over new directions in gene expression space , allowing more specific sets of targets to be controlled . Several studies have shown that TFs that can bind genes in one tissue type are in fact precluded from binding the same genes in another [70 , 71] . The C . elegans TF LIN35 fails to bind targets in the germline that it binds in the intestine [71] , and the SMARCA4 complex in mouse binds enhancer elements in heart , limb , and brain tissue in a tissue-specific manner [70] . One hypothetical explanation for these observations , based on the sequestration model , is that cell-type specific gene expression is the result of noncommutative sequences like K1 K2 and K2 K1 that silence certain promoters . The three-dimensional structure of the genome is a likely setting for this type of regulation . Gene regulation is known to take place in three-dimensions , as observations of DNA looping [72] , nonrandom chromosome packing [73] , and clustered transcription factories [74] have shown . However , the factors that affect chromosome structure are non-specific . One such factor is the ubiquitous zinc finger protein CCCTC binding factor ( CTCF ) [75] , which functions as both an activator of transcription by bringing enhancers and promoters together [76 , 77] and as a repressor by insulating genes [78 , 79] . Epigenetic modifications , such as histone methylation and acetylation [80–82] , also affect three-dimensional structure . In addition , DNA looping was observed in the context of allelic exclusion during B- and T-cell lineage specification where individual alleles were recruited to heterochromatic regions while the other underwent recombination [33 , 34] . Consequently , the sequestration model predicts that temporal permutations of a small set of chromatin modifying factors could specify a large number of potential chromosomal conformations and lead to different expression states and corresponding cell fate decisions . New technologies such RNA-seq and ChIP-seq can be used to test the predictions of the noncommutativity hypothesis at the genome level . Epigenetic drugs such as azacytidine and trichostatin A inhibit DNA methylation [83] and histone deacetylation [84] , respectively , and have been shown to cause global changes in gene expression alone and in combination [83 , 85] . The sequestration hypothesis predicts that perturbations to the three-dimensional structure of the chromosome are noncommutative , so distinct gene expression states may be reached by permuting the order in which epigenetic drugs are applied . While the sequestration model may underlie chromosome folding , the ratchet model could form the basis of phosphorylation networks . For example , mass spectrometry studies have revealed complex phosphorylation patterns [86 , 87] , though the number of kinases and phosphatases is comparatively small and the networks are highly interconnected [10 , 11] . As phosphoproteins are the mediator of extracellular signals , ordered disruption of signaling pathways could also lead to distinct gene expression configurations . Analogously , the ratchet model may aid in the specification of distinct neural activity patterns , owing to the fact that connections between the different hippocampal layers overlap [12 , 88] . While superficial neurons can be activated in response to spatial cues , deeper layers can be selectively activated by time sequences of inputs [40 , 41 , 89] . These results suggest the hypothesis that neural networks may be noncommutative . In particular , experimental support exists for the role of the dentate gyrus in pattern separation and orthogonalization by way of ensuring that even quite similar memory representations use distinct subsets of neurons [90 , 91] . The ratchet model , by ordering inputs in time , is one way of reaching these specific subsets if the number of input neurons is smaller than the number of targets neurons . Memories share many common elements , including shape , color , smell , and sound , which poses problems for recall . We hypothesize that older , “fuzzier” memories could be those relegated to very long ratchet sequences . According to this hypothesis , memories are not forgotten , but are instead increasingly difficult to access , and memories that are not consolidated are those that never formed a unique ratchet sequence . Beyond resolving bottlenecks and generating specificity , noncommutative actions offer a new interpretation of how cell fate decisions and other stepwise processes occur on abstract regulatory landscapes . The classical Waddington landscape view of development holds that cells decay to attractor configurations representing terminal outcomes [92]; this is consistent with a boolean network with many variables X converging to a fixed point [93] . In a static landscape , the final outcome is determined a priori by the nearest energy minimum . What then determines the initial configuration ? In organisms such as Drosophila , maternal patterning of the embryo may account for this initial bias [94]; but in other organisms that employ mechanisms like multilineage priming [82 , 95] , it is not clear that every cell fate decision is made at the beginning . Sequential logic allows cells to reach their final fate on a dynamic landscape . In the system of Fig 8A ( top ) , for example , it is not possible for cells in the blue configuration to transition to the red fate by increasing X2 , because this involves an uphill climb . However , the regulators of genetic networks may also affect the landscape directly . This is seen in Fig 8A ( bottom ) where the sequence K1 K2 P1 changes the landscape in such a way that the overall cost of reaching the same endpoint is much lower than the direct path ( Fig 8A , top ) . This can be understood as the effect of regulators acting on additional variables V , which modulates the landscape in X space . For example , TFs can recruit chromatin regulators that modify global three-dimensional chromosome structure and future TF accessibility [74 , 76 , 96 , 97] , or kinases can sequester substrates in the nucleus to prevent their subsequent activation [53 , 54] . Because sequential logic acts on the V’s as well as the X’s , changes that appear to be small in one dimension ( Fig 8B , left ) actually involve large excursions in the full space ( Fig 8B , right ) . As a consequence , in noncommutative regulation , the landscape changes and cells can take on fates that were not accessible at the beginning . Previous theoretical models have explored dynamic regulatory landscapes in the form of bifurcations [98 , 99] . In these models , a set of kinetic parameters determines the positions of minima and maxima in the landscape . However , the noncommutative model advanced here is fundamentally different , in that using the regulators to move through X changes the landscape directly . This could happen , for example , if acting on X1 with K1 hides it from the effect of K2 . Uncoupling of targets in this way may underlie the distinct effects of signals like FGF at different stages of development [35–38] . It will be interesting to explore time series data for hints that some genes pulse ON and OFF in order to protect their promoters from the actions of promiscuous regulators . Multistep processes other than development can benefit from the type of noncommutative regulation highlighted in Fig 8 . What seems like an intractable problem at the start becomes much more feasible if one realizes that the effects of actions change with time and context . This intuition is why thinking in terms of commutators [A , B] = AB − BA can make complex problems more soluble: the desired effect is often what is leftover after performing and undoing a sequence of actions . Several examples illustrate this concept . With its increased capacity for generating diversity , sequential logic is likely to be used in evolution . A recent theoretical example in social bacteria demonstrated that in evolving a new quorum sensing receptor-ligand pair , adding new receptors prior to ligands is preferred over the opposite path [45] . An analysis of the stability and catalytic activity of a family of bacterial β-lactamase mutants showed that the ability to evolve new substrate specificity is contingent on mutations that first stabilize the protein active site [46 , 100] . Finally , biological networks evolve the same functions in different orders , but the order in which these functions arise dictates which other genotypes can be reached by neutral mutations [44] . These results suggest that permuted sequences of mutation events may have different fitness costs . With extensive artificial evolution experiments underway in protein engineering [100] and bacterial mutation accumulation [47] , coupled with progress in sequencing technologies , it will be possible to test this hypothesis by permuting the conditions that promote mutation . Sequential logic can also be applied in synthetic biology to build circuits with memory [43 , 101–103] . In general , the toolkit that permits up- and downregulation of genes is small , with a few staples like Lac , Tet , and Ara [104] . Significant effort has been put into generating logic gate ( AND/OR ) promoters [30] . To further expand the toolkit , it has been proposed that more orthogonal regulators be developed [105] . We suggest that sequential logic may be a more promising strategy to scale up the number of targets that can be independently controlled by permuting in time a small number of controllers . More broadly , sequential logic can be used to accomplish experimental goals not possible in single-step approaches . For example , in multiplexing mRNA detection in single cells , we previously used a sequential hybridization scheme that permits the number of barcodes to exponentially [106] , whereas combinatorial schemes can only specify approximately 30 barcodes . We expect many single-cell experiments to benefit from a sequential strategy in which detours facilitate achievement of the main goal with high efficiency . Finally , our results connect outside of biology to strategic planning in social , political , and economic arenas . Anyone familiar with negotiating knows about the limitations inherent in trying to make interconnected groups of people move in specific directions , especially when the actions affect all participants at once . Multiparty negotiations and tournaments may benefit from time-ordered strategies in which enemies temporarily team up , or fringe interest groups are transiently pacified . Indeed , a conclusion from the sequestration model is that the most highly regulated targets need to be protected prior to satisfying the ones with fewer connections . Determining whether this prediction is borne out in congressional and international negotiations , for example , is an interesting question for political science . Evidence for noncommutative effects in games exists in that the initial seeding in a tournament can bias its outcome [107] , and that long-term goals change players’ strategies in in the repeated prisoner’s dilemma [108] . In conclusion , the direct path to an outcome in a networks with many interacting parts may have many unintended and prohibitively expensive consequences . A multi-step strategy may achieve the same outcome with minimal cost and side effects .
In this section we determine how many targets are controlled by the same regulators in the connectivity matrix A . Then we extend A to more than 2 dimensions . If ln = lm = 1 it is clear that each Ai , j corresponds to a single target and that each target appears only once . In general , however , a target can appear in multiple entries of A ( cf . Eq ( 6 ) ) . To see this , consider the bipartite graph formed by all the targets and all the K’s , but none of the P’s . The handshaking lemma from graph theory [59] says that the total number of edges is one half the sum of the degrees of each vertex , which is either ln for a target or some number pn for a K regulator . There are Nln total edges , so we find 1 2 ( N l n + n p n ) = N l n or p n = N n l n for the number of links coming from each K . Similarly , the number of links emanating from each P is p m = N m l m . In terms of the connectivity matrix , pn and pm correspond to the number of unique targets in each row and column , respectively . Because K1 connects to a fraction p n N of the targets , it follows that K1 and P1 together connect to a fraction p n p m N 2 of the targets . Therefore , the total number of targets connecting to K1 and P1 is M = N ( p n p m N 2 ) = p n p m N . Another way to see this is to consider one target in the intersection of K1 and P1 . This one target uses up one of each of the regulators and one unit of connectivity , leaving a total of M = ( n - 1 l n - 1 ) ( m - 1 l m - 1 ) ways to connect other targets to the same pair of regulators . It is easily verified that these two formulations for the number of targets per matrix entry M are equivalent . This illustrates that there is not simply a one-to-one correspondence between the entries of A and the targets . There was nothing special about the labels K and P in the above paragraphs . Thus , the connectivity matrix can easily be extended to a u-dimensional connectivity tensor where u is the number of pools of regulators . Each pool has ni regulators connecting to lni targets , and each target connects to p n i = N n i l n i regulators of pool i , ∀i ∈ {1 , … , u} . The total number of targets and the total number of targets per entry are extensions of the u = 2 case , giving N = ∏ i = 1 u n i l n i ( 10 ) distinct targets and M = ∏ i = 1 u p n i N u - 1 = ∏ i = 1 u n i - 1 l n i - 1 ( 11 ) targets controlled by one factor from each of the u pools . S1A Fig shows an example network with u = 3 pools . The number of configurations in combinatorial logic is the number of ways that N targets can each be bound by exactly u regulators , where each regulator comes from a different pool . In the main text we analyzed the case u = 1 and ln = 1 . Here we extend those results to arbitrary u and ln . First consider the case u = 2 , corresponding to a pool of K’s and a pool of P’s . Whereas in the ratchet model , Ki and Pj acted separately on the entries of A , in combinatorial logic the pair ( Ki , Pj ) is needed to switch Ai , j from 0 to 1 . Many such pairs may be active at any one time . We write this formally as K , P A i , j = 1 if K i ∈ K and P j ∈ P 0 else , ( 12 ) where {K} denotes a subset of the K’s . The notation ( ⋅ , ⋅ ) means that a combination of factors acts on the target , instead of just a single factor . If ln = lm = 1 there are ( 2n − 1 ) ( 2m − 1 ) + 1 ways to pick at least one of n K’s and one of m P’s , plus one way to pick nothing . If lm = 1 and ln > 1 , then for a certain number α ≤ n of the K’s , any subset containing α or more K’s has the same effect as activating all n K’s at once . For example , in Eq ( 6 ) , the action of ( {K1 , K2} , {P1 , P2} ) is sufficient to activate all targets in the n = m = 3 , ln = lm = 2 network . To determine α , recall that there are M targets in each entry of the connectivity matrix A . Choosing i K’s means that the total number of targets is M × i , but a single column of A only contains pm unique targets . Each target is connected to ln K’s , so for a target in the intersection of i K’s and a single P , there are ln − i spots left over to choose n − i K’s and lm − 1 spots left over to choose m − 1 P’s , or ( n - i l n - i ) ( m - 1 l m - 1 ) ways total . Using the principle of inclusion-exclusion [59] this means that α is the smallest i such that M × i - ∑ i ′ = 2 min i , l n - 1 i ′ i i ′ n - i ′ l n - i ′ m - 1 l m - 1 ≥ p m . ( 13 ) By choosing α K’s , the number of unique targets in a column of A that can be turned ON is exactly the number represented in that column . Because all subsets with α , α + 1 , … , n − 1 K’s are redundant , here are only ( 2 n - 1 ) - ∑ i = α n - 1 ( n i ) subsets of K’s that contribute to unique configurations , leaving a total of [ ( 2 n - 1 ) - ∑ i = α n - 1 ( n i ) ] ( 2 m - 1 ) + 1 unique configurations . If the P’s also have redundant connections , the result generalizes to Theorem 1 The number of configurations in combinatorial logic with parameters n , m , ln , lm , and u = 2 is 2 n - 1 2 m - 1 + 1 - ∑ i = α n - 1 n i 2 m - 1 - 2 n - 1 ∑ i = β m - 1 m i + ∑ i = α n - 1 n i ∑ i = β m - 1 m i , ( 14 ) where α ( resp . β ) is the smallest number of K’s ( resp . P’s ) having the same effect as all K’s ( resp . P’s ) at once . This result is obtained by counting all pairings of K’s and P’s , then subtracting those pairings that have a redundant effect . For example , any combination using K3 is redundant in the connectivity matrix of Eq ( 6 ) . Finally , those pairings that were excluded twice are added back in . This result generalizes to all u with slight modifications . Because one factor from each of u pools is now required , the combinatorial equation determining state of a target is K 1 , K 2 , … , K u A i , j , … , k = 1 if K 1 i ∈ K 1 , K 2 j ∈ K 2 , … , K u k ∈ K u 0 else . ( 15 ) Here the double subscript Kik indicates the kth factor in the ith pool . Determining αi for each pool i of regulators requires finding the pool j ≠ i which maximizes the number Ni of targets controlled in two dimensions . If we choose αi or more regulators in the ith pool , then there is a redundancy in the jth dimension , whereas any choice of fewer than αi regulators activates fewer than Ni targets . Write N i = max j ≠ i { ( n i l n i ) ( n j l n j ) } the total number of targets and p n j = N i n j l n j the number of targets in any column of the the equivalent ni × nj connectivity matrix regulated by pools i and j . It is easy to see that these parameters reduce to their previous definitions for u = 2 . Now define M i = ( n i - 1 l n i - 1 ) ( n j - 1 l n j - 1 ) as the number of targets in each entry of the equivalent ni × nj connectivity matrix . As above , αi is now the smallest r such that M i × r - ∑ r ′ = 2 min r , l n i - 1 r ′ n i - r ′ l n i - r ′ n j - 1 l n j - 1 ≥ p n j . ( 16 ) Once αi is determined for each pool i , the inclusion-exclusion sum can be extended using standard arguments [59] . Define by S k = ∑ σ ∈ u k ∏ i ∈ σ ∑ j = α i n i - 1 n i j ∏ i ∉ σ 2 n i - 1 , ( 17 ) where σ denotes all k-subsets of {1 , … , u} . Then we have the final result Theorem 2 The total number of configurations in combinatorial logic with u pools and parameters ni , lni , i ∈ {1 , … , u} is S = 1 + ∑ k = 0 u - 1 k S k . ( 18 ) This result reduces to Theorem 1 when there are only u = 2 pools . At most there are ∏ i = 1 u ( 2 n i - 1 ) ways to specify at least one target , corresponding to the 0th-order term in Eq ( 18 ) . Increasing the connectivity through the lni can only reduce the number of configurations . This behavior is shown in S1B Fig for the symmetric case that all the ni and lni are equal . As u is increased the number of configurations increases dramatically , but the scaling is actually subexponential , i . e . less than 2N . Increasing connectivity through lni shifts the curves to the right . To establish the correspondence between the reachable configurations of ratchet network ( ln = lm = 1 , T = 1 ) and the lonesum matrices , we must show ( i ) that A avoids the patterns ( 1 0 0 1 ) and ( 0 1 1 0 ) in any 2 × 2 sub-block , and ( ii ) that any lonesum matrix can be constructed from K and P actions . First observe that the value 1 in Ai , j indicates the last K affecting that index must have followed a P , whereas 0 indicates the last P must have followed a K . For the first restriction we have ( 1 0 0 1 ) implies ( P 1 … K 1 K 1 … P 2 K 2 … P 1 P 2 … K 2 ) . This means P2 follows K1 follows P1 follows K2 follows P2 , which is a contradiction , showing that this 2 × 2 block is unreachable . The other five unique 2 × 2 blocks are all reachable with elementary sequences . This establishes point ( i ) that the reachable configurations are a subset of the lonesum matrices . To establish point ( ii ) that the lonesum matrices are a subset of the reachable configurations , we use an equivalent formulation of the lonesum matrices as staircase matrices composed of the rows aj = ( 1 , … , 1 , 0 , … , 0 ) with the last 1 appearing at position ij subject to the constraint that ij ≤ ij − 1 for all ∀j ∈ {2 , … , n} [109] . It is easy to see that the pattern of ones resembles an inverted staircase . We show via induction that any staircase matrix can be constructed from K and P actions . The nth row is obtained by the sequence Kn Pin + 1⋯Pm which leaves 1’s at the first in indices and 0’s at the remainder . Now assume that the kth row is obtained by the sequence Kk Pik + 1⋯Pm without affecting any of the rows n , n − 1 , … , k + 1 . Then the sequence Kk − 1 Pik − 1 + 1⋯Pm puts 1’s at the first ik − 1 indices of row k − 1 . Because ik − 1 ≥ ik ≥ ⋯ ≥ in , none of the Pik − 1 + 1 , … , Pm turn a 1 to a 0 in rows n , n − 1 , … , k + 1 , k . This proves the induction hypothesis and shows that the staircases matrices are a subset of the reachable configurations . Together with the fact that the reachable configurations are a subset of the staircase matrices , this implies that the reachable configurations and the lonesum matrices are in fact the same set , and we have Theorem 3 The number of reachable configurations in the ( n , m ) ratchet network with ln = lm = 1 and threshold 1 scales as the poly-Bernoulli numbers B m - n = B n - m . With T = 2 , only targets in state 2 are ON . Once a 0-1 configuration of A is obtained , however , it is a simple matter to convert it into an ON/OFF configuration by applying all the K’s . Here we use the fact that 1’s can be reached from above and below to prove the Theorem 4 In the ratchet network represented by the matrix A with ln = lm = 1 and threshold T = 2 , all binary 0-1 configurations are reachable . Proof . We use an induction argument analogous to the proof of Theorem 3 . Suppose that in row n a set of r ≤ m indices {nj} = {nj1 , … , njr} should be ON . First prepare every target in row n in the 1 state using Kn , then use the sequence Kn Pjr + 1⋯Pjm to obtain 2’s at {nj1 , … , njr} and 1’s at {njr + 1 , … , njm} . Now assume that we can prepare rows n , n − 1 , … , k + 1 in a similar 1-2 configuration with the rest of the matrix 0 . We want to show that we can add row k to this set without affecting any of the previous rows . Assuming that a set of s ≤ m indices {kj1 , … , kjs} should be ON , apply the sequence P j 1 … P j s K k 2 P j s + 1 … P j m to obtain 2’s at {kj1 , … , kjs} and 1’s at {kjs + 1 , … , kjm} . Now , because {Pj1 , … , Pjs}∪{Pjs + 1 , … , Pjm} = {P1 , … , Pm} , all 2’s and 1’s in rows n , n − 1 , … , k + 1 are now 1’s and 0’s , respectively . Applying the sequence Kn Kn − 1⋯Kk + 1 reestablishes the 1-2 configuration we had prior to fixing row k and leaves 0’s at rows 1 , … , k − 1 . Now that row k is also in the proper 1-2 configuration , we have proved the induction hypothesis . Once all rows in the proper 1-2 configuration , the sequence P1⋯Pm obtains the matrix in the 0-1 configuration . Since this procedure can be repeated for any collection of indices {{1j} , … , {nj}} , it follows that all binary 0-1 matrices are reachable . When the connectivity parameters ln and lm exceed 1 , certain sequences in the threshold 1 ratchet network become redundant . Our goals in this section are to ( i ) to characterize the redundant sequences by the number of K’s and P’s , and ( ii ) count the non-redundant sequences . This will obtain an upper bound on the number of configurations . We want the shortest sequences that can activate or ( deactivate ) all targets; any sequences longer than this are redundant . To see why this is so , we need the concept of a cycle . We say that a target has gone through a cycle if has traversed the states 0 , 1 , 0 at some subsequent time points . We have the following lemma . Lemma 5 Any sequence that takes all targets through a cycle is redundant . Proof . The final configuration of any sequence is represented by the positions of the 1’s and 0’s of the connectivity matrix . Recall that Ai , j = 0 if an only if all targets represented by Ai , j are OFF in the final configuration . Permute the rows and columns of A until it is in staircase form with r ≤ min ( n , m ) steps , where a step is a group of adjacent rows or columns having the same number of 1’s and 0’s . The steps partition the rows and columns of A into subsets of indices {i1 , i2 , … , ir} and {j1 , j2 , … , jr} where the kth step is defined by 1’s at rows ik to ik + 1 − 1 and 0’s at columns jk to jk + 1 − 1 . Then the sequence ∏ k = 1 r K i k … K i k - 1 P j k … P j k - 1 obtains the desired configuration of 1’s and 0’s . Being able to write a staircase matrix for the final configuration means that every target ON in the final configuration occurs only where there are 1’s in the matrix . These targets are never affected by a P in this procedure; they do not go through a cycle . Because any allowed configuration can be reached from this procedure , it follows that any sequence that uses a cycle is redundant . Knowing that the non-redundant sequences must avoid cycles , it suffices to find the longest sequences that can be written before cycles appear . Lemma 6 For each value of ln ( lm ) , the maximum number of K’s ( P’s ) that can be used before all targets are activated ( deactivated ) is n − ln + 1 ( m − lm ) . Proof . A sequence that activates all targets has no intervening P’s . Recall that a single K activates at most N n l n targets . Then , prior to the last K being used , the number of activated targets is N - N n l n = N n ( n - l n ) ≤ N n l n ( n - l n ) . This means there are at most n − ln groups of targets controlled by different K’s . Thus , at most n − ln K’s are used before the last K is used , and n − ln + 1 K’s must be sufficient to activate the complete set . The maximum number of P’s that can be used is only m − lm because we can think of every sequence starting in the zero configuration as having been preceded by a single P; this modification puts the P’s on equal footing with the K’s . With this characterization of the non-redundant sequences our goal is to recursively eliminate sequences that use n − ln + 1 K’s and m − lm P’s . We first find the number of sequences that use up to m − lm P’s , which forms the top row in each ( n , m ) block in S2 Fig . Then we use these values to recursively find the number of sequences using up to n − ln + 1 K’s . The strategy is to subtract from the total number of sequences at a given ( ln , lm ) all those sequences using the forbidden number of regulators in order to get the new total . Denote by a n m the number of sequences using m P’s when the total number of K’s is n . If m = 1 , then all B 1 - n = 2 n sequences ( except for the empty sequence ) use a K and none use a P . If m = 2 , the maximum number of P’s that can be used is m − lm = 1 . Discarding the 2n sequences with no P , the number of sequences using a single P is a n 1 = B 2 - n - 2 n 2 . ( 19 ) Division by m = 2 is required to account for the fact that there are ( m 1 ) = m different ways of starting each sequence with a P , and we consider both of these equivalent . Having determined a n m , it is straightforward to determine a n m + 1 . Because there are m + 1 P’s to choose from , there are ( m + 1 m ) a n m ways to write sequences with m P’s , ( m + 1 m - 1 ) a n m - 1 ways to write sequences with m − 1 P’s , … , ( m + 1 0 ) 1 ways to write sequences with 0 P’s , the only remaining sequences are those with m + 1 P’s . Knowing that the total number of sequences is B m - n , this leaves a n m + 1 = B m - n - 2 n - ∑ j = 0 m m + 1 j a n j m + 1 ( 20 ) total sequences using m + 1 P’s when the total number of K’s is n . Having determined this number , we can sum up all the sequences using m − lm P’s to get the first row of the ( n , m ) block in S2 Fig . Denote by c n m ( l n , l m ) the lmth column and lnth row of the ( n , m ) block . The column headers c n m ( 1 , l m ) are given by c n m 1 , l m = 2 n + ∑ j = 0 m - l m m j a n j . ( 21 ) We can determine the row entries for ln > 1 in the same way that we determined the column headers , the only difference being that the total number of sequences is c n m ( 1 , l m ) , not B m - n unless lm = 1 . Denote by b m n ( l m ) the number of sequences using n K’s when the total number of P’s is m and the P connectivity is lm . For fixed m , lm and n = 1 , there are b m 1 l m = 2 m - ∑ j = 0 l m - 1 m j , ( 22 ) sequences , as all but the empty sequence use a single K . In complete analogy to Eq ( 20 ) we find there are b m n + 1 l m = c n m 1 , l m - ∑ j = 0 n - l n + 1 n + 1 j b m j l m ( 23 ) sequences using n + 1 K’s when the total number of P’s is m . Unlike in the equation for a n m , there is no division by n + 1 because all sequences starting with a different K are different . Finally , we can sum up all the sequences using n − ln + 1 K’s to get the Theorem 7 The number of minimal length sequences in the ( n , m , ln , lm ) ratchet network with threshold T = 1 using no more than n − ln + 1 K’s and m − lm P’s is c n m l n , l m = ∑ j = 0 n - l n + 1 n j b m j l m . ( 24 ) We used this formula to compute each entry in S2 Fig . Because of the complexity of this procedure , we checked it against a computer algorithm operating with the following steps . In step 1 find all B m - n sequences in the ln = lm = 1 case . In step 2 increase the connectivity ( ln or lm ) and find all sequences of a given length; group them by the configuration they generate . Some of these sequences will not appear in the list generated by step 1: for example , both K1 K2 and K2 K1 will be found in step 2 . We are interested in index permutation e . g . 1 → 3 , not letter permutation , so in step 3 delete all sequences in each length group not appearing in step 1 . Repeat steps 1–3 with this new list of sequences until ln = n − 1 . This code , implemented in Matlab Version 2015b , gave exact agreement with Theorem 7 . We now show that rules restrict the reachable configurations of the sequestration model in the main text to the connected one-colorings of the reduced n-network . Theorem 8 There is a one-to-one correspondence between the reachable configurations of the reduced n-network and the connected one-colorings . Proof . The converse direction , reachable implies connected , is easier to prove and will be discussed first . Assume that all configurations in the reduced n-network so far reached are connected . The next configuration will be reached by turning all 0’s to i’s or all j’s to 0’s by application of Ki or Pj , respectively . The k-arm targets sharing state i with the 2-arm target {0 , 1 , i} are either in the same state as some other 2-arm target {0 , 1 , i′} or are in the 0 state . So application of Ki cannot change the connectivity of the configuration . Furthermore , a k-arm target can be in the j state only if the target {0 , 1 , j} is in the j state , so these targets will still be matched after application of Pj . Thus , any configurations reachable from a reachable configuration must be connected . The forward direction , connected implies reachable , is less trivial . In order to prove that all connected one-colorings in the n-network are reachable , we will use the strong form of mathematical induction . Assume the theorem holds for all networks up to n − 1 . Embedded within the full n-network of 2n − 1 targets is the reduced n-network on 2n − 1 targets . Within the reduced n-network is a set of 2n − 2 targets able to access {0 , 1 , 2} and all subsets ( including Ø ) of the integers {3 , … , n} . Thus , we can substitute 2 → 1 as the ON state in this embedded network and all connected one-colorings ( of 2 ) will be reachable . The same holds in general for all 2n − k targets able to access {0 , 1 , k} and all subsets of the integers {k + 1 , … , n} . In each of these embedded networks the substitution k → 1 as the ON state will enable us create any connected one-coloring . Pick any connected one-coloring ( of 1 ) in the n-network . Its opposite configuration is formed by the transformation at each target g of 1 → 0 and 0 → kmin , where kmin = min{k∈g|xpos ( {0 , j , k} ) = 0} is the smallest index that g shares with a corresponding 2-arm target at position pos ( {0 , j , k} ) of x ( possibly in the full network ) currently in the 0 state . The opposite of a connected one-coloring is clearly connected , because all the connected 1’s are now 0 , and all the 0’s are in the same state as the 2-arm target {0 , j , kmin} . If it is possible to reach the opposite configuration , then application of the sequence K1 P2…Pn yields the desired one-coloring of the n-network . To show that the opposite configuration of the chosen one-coloring is indeed reachable , isolate the embedded networks one-by-one by application of the sequence Kk K1 Pk for k = 2 , … , n , so that the targets in the n − k + 1-network are the only targets in the 0 state . By hypothesis , the connected one-colorings are reachable in all embedded networks which have at most n − k states besides 0 , 1 , and k . The opposite configuration in the n-network is composed of connected one-colorings ( of k ) in each embedded network; these are are reachable . Therefore , the one-coloring of the n-network is reachable via K1 P2…Pn . This procedure holds for any one-coloring . How many configurations are reachable in the full n-network ? Let this number be cn . The following theorems derive lower and upper bounds for cn in terms of the number of one-colorings . Theorem 9 The formula f ( n + 1 ) for the number of connected one-colorings in the reduced n + 1-network is a lower bound for cn . Proof . The full n + 1-network can be partitioned into a set of 2n targets having a 1 and all subsets of {2 , … , n + 1} , and 2n − 1 targets that lack 1 but have all nonempty subsets of {2 , … , n + 1} . The latter set of targets is an embedded full n-network , while the former is the reduced n + 1-network . All 2 ( n + 1 ) letters are needed to form the one-colorings in the reduced n + 1-network . Every one-coloring is finally obtained by applying some permutation of K1 , P2 , … , Pn + 1 to a configuration that uses ( at most ) the states 2 , … , n + 1 and 0 , i . e . the full n-network . Because K1 and P1 do not affect the targets of the the embedded full n-network , there must be ( at least ) one sequence using only {K2 , … , Kn + 1} and {P2 , … , Pn + 1} that prepares the embedded full n-network in the aforementioned configuration , which means we may associate a one-coloring with ( at least ) one of the cn sequences in the embedded full n-network . Therefore , multiple configurations in the full n-network may map to the same one-coloring in the reduced n + 1-network . Conversely , if two one-colorings are different , they are distinguishable by their configurations immediately preceding the final K1 , P2 , … , Pn + 1 sequence , and must therefore map to different configurations in the full n-network . Together , these statements imply that the map from configurations in the full n-network to one-colorings in the reduced n + 1-network is many-to-one , but the map from one-colorings to configurations in the full n-network is one-to-one . Therefore , f ( n + 1 ) ≤ cn . Theorem 10 An upper bound on cn is n f n + n n - 1 f n - 1 f n - 1 + ⋯ + n ! f n - 1 ⋯ f 2 - 1 f 1 + 1 = ∑ k = 1 n n k ∏ j = n - k + 2 n f j - 1 f n - k + 1 + 1 . ( 25 ) where ( n ) k = n ( n − 1 ) ⋯ ( n − k + 1 ) is the falling factorial . Proof . There are nf ( n ) one-colorings in the full n-network , plus one origin . Each one of the one-colorings can be thought of as the origin of an n − 1-network , which in turn generate ( n − 1 ) f ( n − 1 ) one-colorings in an embedded n − 1-network , for a total of n f n n - 1 f n - 1 configurations using 1 , 2 , and perhaps 0 , hence termed two-colorings . However , one of the f ( n ) one-colorings is the 0 state of the n-network , so it does not generate any two-colorings . Thus , there are at most 1 + nf ( n ) + n ( n − 1 ) ( f ( n ) − 1 ) f ( n − 1 ) zero- , one- , and two-colorings . Now assume that the number of k-colorings is n n - 1 ⋯ n - k + 1 f n - 1 f n - 1 - 1 ⋯ f n - k + 2 - 1 f n - k + 1 . Of these , n n - 1 ⋯ n - k + 1 f n - 1 f n - 1 - 1 ⋯ f n - k + 2 - 1 are origins of an n − k-network , meaning they are actually k − 1-colorings; they cannot generate any k + 1-colorings . The remaining n n - 1 ⋯ n - k + 1 f n - 1 f n - 1 - 1 ⋯ f n - k + 2 - 1 f n - k + 1 - 1 are genuine k-colorings which can generate f ( n − k ) one-colorings in the n − k-network , or equivalently , k + 1-colorings . Thus , the total number of zero- , one- , two- , … , k + 1-colorings is no more than n n - 1 ⋯ n - k + 1 f n - 1 f n - 1 - 1 ⋯ f n - k + 2 - 1 f n - k + 1 - 1 . This induction argument proves the statement . First we define what it means to be an origin and an orbit in the threshold-1 ratchet network and determine the number of orbits as a function of n and m . Then we prove that the configurations in the sequestration network are defined by reversible paths . A forbidden configuration in the ratchet network contains some row or column permutation of the pattern ( 1 0 0 1 ) on any 2 × 2 sub-block of the connectivity matrix A . This is the minimum violation , but larger blocks may violate this pattern as well , for example ( 1 0 0 1 1 0 ) has 2 violations . Furthermore , application of any of the K’s or P’s in this sub-block will relieve at least one of these violations . Therefore , we define an i , j-orbit in the ratchet network as the locus of configurations having a forbidden configuration on an i × j sub-block that does not use the corresponding set of i K’s and j P’s . The origin of any i , j-orbit is the configuration having all remaining nm − ij entries of A equal to 0 ( or all 1 to make the case of having only P actions symmetric with having only K’s ) . A matrix X having the same forbidden i × j sub-block as an origin Y is not considered to be in the orbit of Y if ( i ) there is no sequence of actions that transforms Y to X , or ( ii ) if the sequence involves one of the forbidden K’s or P’s . With these restrictions , the number of origins is equal to the number of orbits . Denote by c i j the number of orbits in a ratchet network of size n × m with violations involving i ≤ n K’s and j ≤ m P’s . If i = j = 2 there are 2 i j - B 2 - 2 = 2 forbidden configurations that turn into origins for the remaining n − i K’s and n − j P’s . There are more orbits in these smaller networks . For every i′ , j′ ≥ 2 there are ( i i ′ ) ( j j ′ ) c i ′ j ′ B i - i ′ - ( j - j ′ ) configurations reached by orbits using i′ K’s and j′ P’s . Only configurations not reached by these orbits are available as new origins when the number of K’s and P’s not to be used is i and j , respectively . Finally , there are ( n i ) ( m j ) ways to specify i ≤ n K’s and j ≤ m P’s . Then we have the Theorem 11 For a given set of i ≤ n K’s and j ≤ m P’s , the number of i , j-orbits is cij= ( 2ij−Bi−j ) −∑i′ , j′≥2i′+j′≤i+j−1i , j ( ii′ ) ( jj′ ) ci′j′B′i−i′− ( j−j′ ) , ( 26 ) and the the total number of i , j-orbits in the n × m ratchet network is C i j n , m = n i m j c i j , ( 27 ) where B ′ i - i ′ - j - j ′ = B i - i ′ - j - j ′ i f i - i ′ > 0 and j - j ′ > 0 2 i - i ′ i f j - j ′ = 0 2 j - j ′ i f i - i ′ = 0 . ( 28 ) The modification B′ ensures that an orbit lacking allowable P’s ( K’s ) can still use K’s ( P’s ) . A table of values of Eq ( 27 ) is given in S4 Fig . We noted in the main text that configuration in the sequestration network can be joined by reversible paths . A path Ki Pj or Pj Ki is reversible if a configuration reached by the sequence of actions w is also reached by the either the sequence wKi Pj or wPj Ki , but not wKi or wPj , respectively . Thus we can also prove the Theorem 12 There always exists a reversible path between any two configurations in an orbit of the sequestration network . Proof . Let x be a configuration in an orbit using m ≤ n of the actions , and let P denote the locus of configurations reached from x . We now need to show that P must be reversibly reached from the origin . Denote by P ¯ the complement of P , so that any y ∈ P ¯ is reversibly reached from the origin . In order for there to be no reversible path between x ∈ P and y ∈ P ¯ , there must always be a state i such that Ki increases the number of targets {⋅ , i} in the i state and Pi increases the number of targets {⋅ , i} in the zero state . Now assume there is a configuration z ∈ P using all m allowed states . z must have at least one target in the 0 state , but this is un-allowed , because then z would violate the connection rule . Therefore , there is a maximum number m′ < m of states used by any x ∈ P . Now assume there is a configuration z′ ∈ P using all m′ allowed states . But this implies that there is a single-arm target {0 , j} that must be in the zero state . Then the action Kj takes z′ to a configuration y ∈ P ¯ and Pj takes y to z . This path must be reversible , and z′ is reached reversibly from the origin . By induction we conclude that m′ = 0 and that P = Ø . Finally , because any two configurations are reached reversibly from the origin , there is a reversible path between them . Theorem 12 defines the orbits of the sequestration network as those configurations connected by reversible paths . In this section we show how to write the K and P regulators as matrix operators in a manner consistent with both models considered in the paper . First we define the vector space V of configurations of the N targets , then we derive the operators that transform V . Let x ∈ V . For a network with N targets we require that ∑i xi = N . This means that x has at least N entries , and in general dim x ≥ N . Therefore , we cannot use the standard state space of N-dimensional vectors , because the operators will not conserve the number of targets . Each target has a 0 state . The number D of independent directions accessible from 0 is called the dimension of the network , and the number T of steps one can move along each dimension is called the threshold . In the ratchet model , each target has a single ladder of states with variable threshold , so D = 1 and T is allowed to vary; in the sequestration model D = n but the threshold is T = 1 . Denote by Adi the fraction of the targets of type A in state i ∈ {0 , 1 , … , T} along dimension d . For a subset of the targets a K-type action causes population transfer between states ( d , j ) and ( d , i ) with i = j + 1 , and a P-type action the reverse . If a K regulator acts for a short time we can write the “reaction rate” equation as x ˙ A j = - g A x A j x ˙ A i = + g A x A j ( 29 ) where gA > 0 is a proportionality constant . This defines a matrix differential equation x ˙ = G dj · x ( 30 ) with x ∈ R N ( D T + 1 ) × 1 the vector of populations of the DT + 1 states of the N targets and G dj ∈ R N ( D T + 1 ) × N ( D T + 1 ) the block diagonal matrix of rate constants between the j and j + 1 population states along dimension d . Eq ( 29 ) can be rewritten x ˙ A j x ˙ A i = - g A 0 g A 0 · x A j x A i . ( 31 ) Because Gdj is block diagonal , Eq ( 30 ) can be solved by exponentiation on each block: x A j t x A i t = exp - g A 0 g A 0 t · x A j 0 x A i 0 = e - g A t 0 1 - e - g A t 1 · x A j 0 x A i 0 . ( 32 ) The restriction of the model from a continuous range of population states xAi ∈ [0 , 1] to the boolean values {0 , 1} formally emerges by considering the “reaction” K catalyzes on its targets to have gone to completion . We do this by taking the the limit t → ∞ in Eq ( 32 ) to get x A j t x A i t = 0 0 1 1 · x A j 0 x A i 0 , ( 33 ) so that the matrix Kdj defined by K dj = lim t → ∞ exp G dj t ( 34 ) is the block diagonal matrix having 1’s at ( row , column ) positions ( 1 + ( d − 1 ) T + i , 1 + ( d − 1 ) T + j ) of each block that responds to K in dimension d and admits population transfer between from j to i . Because K acts on all targets at once , it is insensitive to the initial state j . Thus the matrix corresponding to the action of K is K d = ∏ j K dj , ( 35 ) which is the block diagonal matrix having 1’s at ( row , column ) positions 1 + d - 1 T + 1 , 1 + d - 1 T + 0 , … , 1 + d - 1 T + T , 1 + d - 1 T + T - 1 and 1 + d - 1 T + T , 1 + d - 1 T + T of each block that responds to K in dimension d . This derivation can be repeated in the case that population goes in the opposite direction from at state j to a state i < j using a different set of rate matrices Hdj corresponding to the reverse of Eq ( 31 ) . We obtain the block diagonal matrix Pd corresponding to the action of P in dimension d having 1’s at ( row , column ) positions 1 + d - 1 T + 0 , 1 + d - 1 T + 1 , … , 1 + d - 1 T + T - 1 , 1 + d - 1 T + T and 1 + d - 1 T + 0 , 1 + d - 1 T + 0 of each block that responds to P in dimension d . Whereas Kd is sub-diagonal , Pd is super-diagonal . The Baker-Campbell-Hausdorf expansion shows that Kd in Eq ( 35 ) and in general any product of matrices Kd and Pd are generated by matrix exponentiation of commutators of the generators Gdj , Hdj . This is the origin of noncommutativity in both the ratchet and sequestration models . An example in the sequestration network illustrates population transfer between states . In the n = 2 network on the targets A , B , and C the initial configuration of the network is represented by ( A 0 A 11 A 21 B 0 B 11 B 21 C 0 C 11 C 21 ) T = ( 1 0 0 1 0 0 1 0 0 ) T . Only targets A and C can access dimension 1 , and only targets B and C can access dimension 2 . Therefore the t → ∞ action of K1 on the network is given by e - g A t 0 0 0 0 0 0 0 0 1 - e - g A t 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 e - g C t 0 0 0 0 0 0 0 0 1 - e - g C t 1 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 1 0 0 → t → ∞ 0 1 0 1 0 0 0 1 0 . ( 36 ) Only A and C advance to state 1 and the number of targets ( 3 ) is conserved . | DNA is the blueprint of life . Yet the order in which a cell follows these instructions makes it capable of generating thousands of different fates . How this information is extracted from underlying gene regulatory networks is unclear , especially given that biological networks are highly interconnected , and that the number of signaling pathways is relatively small ( approximately 5–10 ) . The conventional approach for increasing the information capacity of a limited set of regulators is to use them in combination . Surprisingly , combinatorial logic does not increase the diversity of target configurations or cell fates , but instead causes information bottlenecks . A different approach , called sequential logic , uses noncommutative sequences of a small set of regulators to drive networks to a large number of novel configurations . If certain targets are first protected , then even promiscuous regulators can activate specific subsets of lineage-specific targets . In this paper we show how sequential logic outperforms combinatorial logic , and argue that noncommutative sequences underlie a number of cases of biological regulation , e . g . how a small number of signaling pathways generates a large diversity of cell types in development . In addition to explaining biological networks , sequential logic may be a general experimental design strategy in synthetic and single-cell biology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"phosphorylation",
"sequencing",
"techniques",
"gene",
"regulation",
"neuroscience",
"regulator",
"genes",
"gene",
"types",
"epigenetics",
"molecular",
"biology",
"techniques",
"chromatin",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"chromosome",
"b... | 2016 | Noncommutative Biology: Sequential Regulation of Complex Networks |
This paper introduces a time- and state-dependent measure of integrated information , φ , which captures the repertoire of causal states available to a system as a whole . Specifically , φ quantifies how much information is generated ( uncertainty is reduced ) when a system enters a particular state through causal interactions among its elements , above and beyond the information generated independently by its parts . Such mathematical characterization is motivated by the observation that integrated information captures two key phenomenological properties of consciousness: ( i ) there is a large repertoire of conscious experiences so that , when one particular experience occurs , it generates a large amount of information by ruling out all the others; and ( ii ) this information is integrated , in that each experience appears as a whole that cannot be decomposed into independent parts . This paper extends previous work on stationary systems and applies integrated information to discrete networks as a function of their dynamics and causal architecture . An analysis of basic examples indicates the following: ( i ) φ varies depending on the state entered by a network , being higher if active and inactive elements are balanced and lower if the network is inactive or hyperactive . ( ii ) φ varies for systems with identical or similar surface dynamics depending on the underlying causal architecture , being low for systems that merely copy or replay activity states . ( iii ) φ varies as a function of network architecture . High φ values can be obtained by architectures that conjoin functional specialization with functional integration . Strictly modular and homogeneous systems cannot generate high φ because the former lack integration , whereas the latter lack information . Feedforward and lattice architectures are capable of generating high φ but are inefficient . ( iv ) In Hopfield networks , φ is low for attractor states and neutral states , but increases if the networks are optimized to achieve tension between local and global interactions . These basic examples appear to match well against neurobiological evidence concerning the neural substrates of consciousness . More generally , φ appears to be a useful metric to characterize the capacity of any physical system to integrate information .
First , we need to evaluate how much information is generated by a system when it enters a particular state , x1 , out of its repertoire ( a repertoire is a probability distribution on the set of output states of a system ) . The information generated should be a function of how large the repertoire of possible states is , and how much uncertainty about the repertoire is reduced by entering state x1 . Also , the reduction of uncertainty must be produced by interactions among the elements of the system acting through their causal mechanisms , which is why we call it effective information . Let us first consider an isolated system , as in Figure 1 . The system consists of three AND-gates and transitions from state x0 = 110 at time zero to state x1 = 001 at time one . How much effective information does the system generate ? To answer the question we need to precisely describe: i ) the alternative states available to the system ( the a priori repertoire ) ; ii ) those states that the architecture of the system specifies as causes of x1 ( the a posteriori repertoire ) . Effective information captures the information generated by the system by measuring the difference between these two repertoires . Effective information is defined as the entropy of the a posteriori repertoire relative to the a priori repertoire , which we write as: ( 1A ) The a priori repertoire is the probability distribution on the set of possible outputs of the elements considered independently , with each output equally likely . This repertoire includes all possible states of the system prior to considering the effects of its causal architecture and the fact that it entered state x1 . This distribution is imposed onto the system , i . e . we perform a perturbation in the sense of [6] . The a priori repertoire coincides with the maximum entropy ( maxent ) distribution on the states of the system; we denote it by pmax ( X0 ) . No perturbation can be ruled out a priori , since it is only by passing a state through the mechanism that the system generates information . The maximum entropy distribution formalizes the notion of complete ignorance [7] . In Figure 1 the a priori repertoire distribution assigns equal probability to each of the 23 = 8 possible outputs of the system . The a posteriori repertoire p ( X0 → x1 ) is the repertoire of states that could have led to x1 through causal interactions . We determine the a posteriori repertoire by forcibly intervening in the system and imposing each state in the a priori repertoire , thus we implement a perturbational approach [1] , [2] , [4] , [6]; see also [8] , [9] which apply perturbations to measure the average interaction between subsets for general distributions . Considering each a priori perturbation in turn we find that some perturbations could have caused ( led to ) x1 and others not ( either deterministically or with a certain probability ) . The a posteriori repertoire is formally captured by Bayes' rule , which keeps track of which perturbations cause ( lead to ) the given effect ( see Text S1 , section 3 ) . In Figure 1 x0 is the unique perturbation that causes x1 , so the a posteriori repertoire assigns weight 1 to x0 and weight 0 to all other perturbations . Relative entropy ( also known as Kullback-Leibler divergence , see Text S1 , section 1 ) is the uncertainty reduction provided by an a posteriori repertoire with respect to an a priori repertoire . It is always non-negative , and is zero if and only if the repertoires are identical . In our case the information is generated by the system when , through causal interactions among its elements , it enters state x1 and thereby specifies an a posteriori distribution with respect to an a priori distribution . By comparing the a priori and a posteriori repertoires effective information measures those “differences that make a difference” [10] . Given that the second term is a maximum entropy distribution , Equation 1A can be more simply written as a difference of entropies , so that ( 1B ) Here H ( p ( − ) ) is the entropy of probability distribution p . Entropy of the a priori repertoire n bits in a system of n binary elements . The second term is the entropy of the a posteriori repertoire , and lies between 0 and n bits depending on the state x1 and the architecture of the system . It follows that a system of n binary elements generates at most n bits of information . In Figure 1 the entropy of the a priori repertoire is 3 bits and that of the a posteriori is 0 bits , so 3 bits of effective information are generated by the system when it enters x1: one out of eight perturbations is specified by the system as a cause of its current state , and the other 7 perturbations are ruled out , thus reducing uncertainty ( generating information ) . In Figure 2 , we show that effective information depends both on the size of the repertoire and on how much uncertainty is reduced by the mechanisms of the system . Figure 2A depicts a system of two elements . The a priori repertoire is smaller than in Figure 1 , and effective information is reduced to 2 bits . Figure 2B shows the AND-gate system entering state x1 = 000 . In this case the a posteriori repertoire specified by the system contains four perturbations that cannot be distinguished by its causal architecture , since each of the four perturbations leads to 000 . Fewer alternatives from the a priori repertoire are ruled out , so effective information is 1 bit . Finally , Figure 2C and 2D illustrate two systems that generate no effective information . In Figure 2C the elements fire no matter how the system is perturbed , so the system always enters state x1 = 111 . The process of entering x1 does not rule out any alternative states , so the a posteriori repertoire coincides with the a priori repertoire and effective information is zero . In Figure 2D the elements fire or not with 50% probability no matter how the system is perturbed . In other words , the behavior of the system is completely dominated by noise . Again , the process of entering x1 does not rule out any alternative states , so the a posteriori repertoire coincides with the a priori repertoire and effective information is zero . Next , we must evaluate how much information is generated by a system above and beyond what can be accounted for by its parts acting independently . Consider Figure 3A . Effective information ei ( X0 → x1 ) generated by the system , considered as a single entity , is 4 bits . In this case , however , it is clear that the two couples do not constitute a single entity at all: since there are no causal interactions between them , each of the disjoint couples generates 2 bits of information independently ( Figure 3B ) . Effective information tells us how much information is generated without taking into account the extent to which the information is integrated . What we need to know , instead , is how much information is generated by the system as a whole , over and above the information generated independently by its parts , that is , we need to measure integrated information . Integrated information φ ( I for information and O for integration ) is defined as the entropy of the a posteriori repertoire of the system relative to the combined a posteriori repertoires of the parts: ( 2A ) where M and μ stand for parts , and PMIP is the minimum information partition , which represents the natural decomposition of the system into parts . The a posteriori repertoires of the parts are found by considering each part as a system in its own right ( averaging over inputs from other parts and extrinsic to the system , Figure 4 ) . Each part has an a priori repertoire , given by the maximum entropy distribution . The product of the a priori repertoires of the parts is the same as the a priori repertoire of the system , since the elements are treated independently in both cases . The a posteriori repertoire of each part Mk is specified ( as for the whole , X , in the previous section ) by its causal architecture and current state , after averaging over external inputs . Thus the rest of the system is treated as a source of extrinsic noise by each part . The effective information generated independently by the parts , shown in red in Figure 4 , is the sum of the entropies of their a priori repertoires relative to their a posteriori repertoires . Integrated information , shown in dark blue , measures the information generated by the system through causal interactions among its elements ( its a posteriori repertoire ) with respect to ( over and above ) the information generated independently by its parts ( their combined a posteriori repertoires ) . In particular , integrated information is zero if and only if the system can be decomposed into a collection of independent parts . Thus , φ ( x1 ) of a system captures how much “the whole is more than the sum ( or rather the product ) of its parts . ” To exemplify , consider again the system of Figure 3 , where the natural decomposition into parts is given by the subsets M1 and M2 , as shown in Figure 3B . The a posteriori repertoire specifies perturbation 10 . Similarly the a posteriori repertoire of M2 specifies perturbation 01 . The combined a posteriori repertoire of the parts specifies perturbation 1001 ( red notch ) , coinciding with the a posteriori perturbation specified by the entire system . No alternatives are ruled out by the system as a whole , so integrated information is The system generates no information as a whole , over and above that generated by its parts . Of note , a related measure is stochastic interaction [11] , which quantifies the average interactions between subsets of a system . Briefly , our approach is distinguished by comparing the whole to the parts , rather than the parts to one another; see Text S1 , section 8 , for detailed discussion and technical motivation . For any given system X , we are now in a position to identify those subsets that are capable of integrating information , the complexes . A subset S of X forms a complex when it enters state s1 if φ ( s1 ) >0 and S is not contained in some larger set with strictly higher φ . A complex whose subsets have strictly lower φ is called a main complex . For instance , the complex in a given system with the maximum value of φ necessarily forms a main complex . ( 3A ) In addition , ( 3B ) At each instant in time any system of elements can be decomposed into its constituent complexes , which form its fundamental units . Indeed , only a complex can be properly considered to form a single entity . For a complex , and only for a complex , it is meaningful to say that , when it enters a particular state out of its repertoire , it generates an amount of integrated information corresponding to its φ value .
Figure 9 shows four discrete systems . Elements fire if they receive two or more spikes . We refer to the number of elements firing as the firing rate of the system . Graphed alongside each system is integrated information , computed across bipartitions , as a function of the firing rate . The graph shows average integrated information , averaged over all output states ( that can arise from the dynamics of the system ) with the given firing rate . Small synchronously updated Hopfield networks [27] , [28] provide a class of examples that are computationally tractable and have interesting dynamics . Hopfield networks are probabilistic systems constructed so that for any initial condition the network tends to one of a few stable firing patterns called attractors . The integrated information generated by a firing pattern depends , in an interesting way , on the relationship between the firing pattern and the attractors embedded in the network . A Hopfield network consists of N elements with all-to-all connectivity . The probability of the ith element firing at time t is given bywhere nj ( t−1 ) is 0 or 1 according to whether the jth element fired at time t−1; and β = 1/T The temperature T is a measure of the amount of indeterminacy in the system: higher temperatures correspond to more noise . The connection matrix Cij is constructed so that the network contains certain attractors . For each attractor stored deliberately there will be additional “spurious” attractors: for example a network designed to store the firing pattern {0…01…1} will also contain its mirror image {1…10…0} . This is a quirk of the Hopfield network design . The construction is as follows . Suppose we wish to store attractor states ξ1 , …ξP . Set With this connection matrix and the probabilistic firing rule above the network will typically – depending largely on the temperature – settle into one of the attractor states ( including the spurious states ) given any initial condition . The construction crucially depends on the near orthogonality of the attractors considered as vectors in the N dimensional space determined by the network . Choosing N to be large – hundreds of elements – and picking the attractors randomly , most easily arranges this near orthogonality . Since Hopfield networks possess all-to-all connectivity and identical elements they are similar to homogeneous systems . The crucial difference is that the weights on the arrows afferent to each element vary . Figure 17 depicts a Hopfield network consisting of 8 elements with 6 embedded attractors . Since we work with a small network randomly chosen attractors will not be orthogonal; instead we carefully choose the attractors so the patterns do no interfere with one another . The attractors are 00001111 , 00110011 , 01010101 , and their mirror images . A sample run is shown at temperature T = . 45 and initial state 11111111 . The network quickly relaxes into an attractor . The graphs show integrated information as a function of temperature , which ranges between . 05 and 2 . We analyze integrated information generated by the system in detail for different states to better understand how integrated information and the repertoire reflect the dynamics . The specific choice of attractors has an interesting consequence . Notice that the pairs of elements n1 and n8 have opposite outputs in all 6 attractors . Similarly for the pairs {n2 , n7} , {n3 , n6} , and {n4 , n5} . It turns out that the connection matrix for the Hopfield network ( embedding the 6 attractors above ) has stronger connections within each couple than between them . The couples are the dominant small-scale feature of the network , a structural feature that will be reflected in the integrated information generated by the system .
In the paper we have extended the notions of effective information and integrated information to discrete non-stationary systems . Integrated information , φ , measures the information generated by a system above and beyond what can be accounted for by its parts acting independently . The subsets of a system capable of generating integrated information form complexes , which constitute individual entities and are the fundamental units into which a system decomposes . Finding the integrated information generated by a physical system requires analyzing it from the ground up , without preconception regarding the nature of its elementary units . In the applications we analyzed a variety of systems to uncover how φ reflects network dynamics and architecture . A few broad lessons can be extracted . First , the integrated information generated by a system depends on the current state of the system . In general , integrated information is higher when there is a balance between the number of active and inactive elements of a system . By contrast , when a system is completely inactive or hyperactive , φ values are low . Second , integrated information can differ substantially for systems with identical or similar surface dynamics , because the latter does not necessarily reflect the causal architecture of a system . For instance , a system composed of causally interacting elements can generate large amounts of integrated information , while a mere copy or “replay” of its surface dynamics generates none . More generally , integrated information appears to be a function of the complexity of the interactions leading to the observed dynamics . Third , we observed that certain classes of network architectures have low φ . Modular and homogeneous systems are unable to generate high φ because the former lack integration whereas the latter lack information . Feedforward and lattice architectures are capable of generating high φ , but they are extremely inefficient . Everything else being equal , it appears that high values of integrated information can be obtained by architectures that conjoin functional specialization with functional integration . Finally , from the probabilistic ( Hopfield-style network ) examples we conclude that high φ can be produced by tension between local and global interactions . Conventional Hopfield networks relax into attractor states and so cannot sustain high φ . However , random Hopfield networks can be optimized to maintain higher values of φ over the course of their dynamics . The notion of integrated information is motivated by the need for a measure that captures the two basic phenomenological properties of consciousness: i ) each conscious experience generates a huge amount of information by virtue of being one of a vast repertoire of alternatives; and ii ) each conscious state is integrated , meaning that it is experienced as a whole and does not decompose into independent parts . We have shown that the way φ behaves in simple simulated networks differing in causal architecture and dynamics fits available neurobiological evidence concerning the neural substrates of consciousness . For example , φ is low for simple network analogues of inactive ( “comatose” ) and hyperactive ( “epileptic” ) states , in line with the loss of consciousness when the brain enters such states . Conversely , high φ requires balanced states similar to those observed when the brain is spontaneously active during waking consciousness . We also saw that a simplified model of bistable dynamics , loosely resembling slow-wave sleep early in the night , when consciousness fades , is not able to sustain high values of integrated information . We provided evidence that , everything else being equal , causal architectures characterized by a coexistence of functional specialization and integration are best suited to generating high values of φ , whereas strongly modular systems fare much less well . Neurobiological evidence suggests that human consciousness is generated by the thalamocortical system [3] , the paradigmatic example of a functionally specialized and functionally integrated network . The cerebellum , which is instead organized into strong local modules with little communication among them , does not seem to contribute to consciousness , though it is as rich in neurons and connections as the cerebral cortex . Finally , the analysis of Hopfield networks shows that tension between the local and global connectivity of a system results in high φ . This suggests that metastable systems , which arise when a collection of neuronal groups are loosely coupled , may be highly integrated . Intriguingly , some initial evidence obtained with multiunit recordings suggests that in awake , behaving animals populations of neurons may undergo a similar metastable dynamics [30] , [31] . A few general observations about the present measure of integrated information are also in order . First , φ measures a process: integrated information is generated by a system transitioning from one state to the next – it does not make sense to ask about the information value of the state of a system per se . Second , φ is a causal measure: integrated information is generated only to the extent that the system transitions into a given state due to causal interactions among its elements . Thus , a system that enters a particular state due to extrinsic noise generates no integrated information , as in Figure 2D . The same is true for a system whose elements update their state without interacting , as in Figure 11 . Importantly , causal interactions can only be made explicit by perturbing the system in all possible ways . Third , φ captures an intrinsic property of a system: integrated information is a function of the possible causal interactions within a system , independent of external observers . In this sense , integrated information is closer to other intrinsic properties of physical systems , such as charge or spin , than to observer-dependent properties that vary with the frame of reference , such as position or velocity . Specifically , integrated information is associated with and indeed identifies complexes – sets of elements that cannot be meaningfully decomposed into independent parts – independently of external observers . For example , elements forming two independent complexes may be lumped together into an externally defined “system” by an observer , as in Figure 7 , but such arbitrary entities generate no integrated information – from an “intrinsic” perspective , they do not really exist . The intrinsic nature of integrated information , which only exists to the extent that it makes a difference from the perspective of the complex itself , is usefully contrasted with the traditional , observer-dependent definition of information , in which a set of signals are transmitted from a source to a receiver across a channel ( or stored in a medium ) , and their “integration” is left to an external human interpreter . Finally , we should mention some of the many limitations of the present work . Foremost among them is that our examples are restricted to small-scale models , so it is unclear to what extent the principles suggested by our partial explorations would scale with larger networks . The impossibility of measuring integrated information for larger systems is due to the combinatorial explosion in the partitions of a system as the number of elements is increased . An inevitable consequence is that computing φ for parts of the human brain , even if the connectivity and causal architecture of the neurons were known , is not a feasible undertaking , though heuristics , estimates , and relative comparisons remain possible . Applying the measure to biological system also introduces the practical issue of correctly identifying the causal architecture and a priori repertoire , a difficult empirical problem ( for example , when dealing with neural networks , should all possible firing patterns be considered , including those differing by just a millisecond , or should they be lumped together ? ) From a theoretical perspective this problem should be addressed according to what may be called a principle of “causal ontology”: only those differences that make a difference within a system matter; differences between a priori perturbations that cannot be detected by the system can be considered as if not existing . There are a number of further issues that will be addressed in future work . A limitation of the present work is the exclusive focus on the amount of information integrated by a given network , with no consideration given to the kind of informational relationships among its elements . To address this we will move beyond quantifying integrated information as a single number and investigate the informational relationships between interacting parts by exposing the geometry of the causal interactions . Another shortcoming is that it focuses exclusively on memoryless systems , in which integrated information can only be generated over one time step . In a forthcoming paper we will coarse-grain discrete systems and develop techniques to find the natural spatiotemporal scale at which a system generates integrated information . This will allow us to deal with systems with memory , as well as to make a first step towards analyzing large-scale , hierarchically organized systems . Finally , the networks considered here were analyzed as isolated entities , without consideration for their environment ( or rather by averaging over possible extrinsic inputs ) . In future work we will discuss how discrete systems interact with and incorporate information from the environment , as well as the relationship between integrated information and learning . | We have suggested that consciousness has to do with a system's capacity to generate integrated information . This suggestion stems from considering two basic properties of consciousness: ( i ) each conscious experience generates a large amount of information , by ruling out alternative experiences; and ( ii ) the information is integrated , meaning that it cannot be decomposed into independent parts . We introduce a measure that quantifies how much integrated information is generated by a discrete dynamical system in the process of transitioning from one state to the next . The measure captures the information generated by the causal interactions among the elements of the system , above and beyond the information generated independently by its parts . We present numerical analyses of basic examples , which match well against neurobiological evidence concerning the neural substrates of consciousness . The framework establishes an observer-independent view of information by taking an intrinsic perspective on interactions . | [
"Abstract",
"Models",
"Results",
"Discussion"
] | [
"neuroscience/theoretical",
"neuroscience",
"computational",
"biology/computational",
"neuroscience",
"computational",
"biology"
] | 2008 | Integrated Information in Discrete Dynamical Systems: Motivation and Theoretical Framework |
Closed-loop neurotechnologies often need to adaptively learn an encoding model that relates the neural activity to the brain state , and is used for brain state decoding . The speed and accuracy of adaptive learning algorithms are critically affected by the learning rate , which dictates how fast model parameters are updated based on new observations . Despite the importance of the learning rate , currently an analytical approach for its selection is largely lacking and existing signal processing methods vastly tune it empirically or heuristically . Here , we develop a novel analytical calibration algorithm for optimal selection of the learning rate in adaptive Bayesian filters . We formulate the problem through a fundamental trade-off that learning rate introduces between the steady-state error and the convergence time of the estimated model parameters . We derive explicit functions that predict the effect of learning rate on error and convergence time . Using these functions , our calibration algorithm can keep the steady-state parameter error covariance smaller than a desired upper-bound while minimizing the convergence time , or keep the convergence time faster than a desired value while minimizing the error . We derive the algorithm both for discrete-valued spikes modeled as point processes nonlinearly dependent on the brain state , and for continuous-valued neural recordings modeled as Gaussian processes linearly dependent on the brain state . Using extensive closed-loop simulations , we show that the analytical solution of the calibration algorithm accurately predicts the effect of learning rate on parameter error and convergence time . Moreover , the calibration algorithm allows for fast and accurate learning of the encoding model and for fast convergence of decoding to accurate performance . Finally , larger learning rates result in inaccurate encoding models and decoders , and smaller learning rates delay their convergence . The calibration algorithm provides a novel analytical approach to predictably achieve a desired level of error and convergence time in adaptive learning , with application to closed-loop neurotechnologies and other signal processing domains .
Recent technological advances have enabled the real-time recording and processing of different invasive neural signal modalities , including the electrocorticogram ( ECoG ) , local field potentials ( LFP ) , and spiking activity [1] . This real-time recording capability has allowed for the development of various neurotechnologies to treat neurological disorders . For example , motor brain-machine interfaces ( BMI ) have the potential to restore movement to disabled patients by recording the neural activity—such as ECoG , LFP , or spikes—in real time , decoding from this activity the motor intent of the subject , and using the decoded intent to actuate and control an external device [2–12] . Closed-loop deep brain stimulation ( DBS ) systems , e . g . , for treatment of Parkinson’s disease , use recordings such as ECoG or LFP to decode the underlying diseased state of the brain and adjust the electrical stimulation pattern to an appropriate brain region , e . g . , the subthalamic nucleus ( STN ) [13–16] . These neurotechnologies are examples of closed-loop neural systems . Closed-loop neural systems need to learn an encoding model that relates the neural signal ( e . g . , spikes ) to the underlying brain state ( e . g . , motor intent ) for each subject . The encoding model is often taken as a parametric function and is used to derive mathematical algorithms , termed decoders , that estimate the subject’s brain state from their neural activity . These closed-loop neural systems run in real time and often require the encoding model parameters to be learned in closed loop , online and adaptively ( Fig 1 ) . For example , in motor BMIs , neural encoding can differ for movement of the BMI compared to that of the native arm or to imagined movements [17–20] . Hence encoding model parameters are better learned adaptively in closed-loop BMI operation [17 , 21–30] . Another reason for real-time adaptive learning could be the non-stationary nature of neural activity patterns over time , for example due to learning in motor BMIs [17–19] , due to new experience in the hippocampus [31 , 32] , or due to stimulation-induced plasticity in DBS systems [14 , 33 , 34] . Adaptive learning algorithms in closed-loop neural systems , such as adaptive Kalman filters ( KF ) , are typically batch-based . They collect batches of neural activity , fit a new set of parameters in each batch using maximum-likelihood techniques , and update the model parameters [22 , 23 , 27] . In addition to these methods , adaptive point process filters ( PPF ) have also been developed for tracking plasticity in offline datasets [31 , 32 , 35 , 36] . Recently , control-based state-space algorithms have been designed for adaptive learning of point process spike models during closed-loop BMI operation , and have improved the speed of real-time parameter convergence compared with batch-based methods [28 , 29] . A critical design parameter in any adaptive algorithm is the learning rate , which dictates how fast model parameters are updated based on a new observation of neural activity ( Fig 1 ) . The learning rate introduces a trade-off between the convergence time and the steady-state error of the estimated model parameters [37] . Increasing the learning rate decreases the convergence time , allowing for parameter estimates to reach their final values faster . However , this faster convergence comes at the price of a larger steady-state parameter estimation error . Similarly , a smaller learning rate will decrease the steady-state error , but lower the speed of convergence . Hence principled calibration of the learning rate is critical for fast and accurate learning of the encoding model , and consequently for both the transient and the steady-state performance of the decoder . To date , however , adaptive algorithms have chosen the learning rate empirically . For example , in batch-based methods , once a new batch estimate is obtained , the parameter estimates from previous batches are either replaced with these new estimates [22] or are smoothly changed by weighted-averaging based on a desired half-life [23 , 27] . In adaptive state-space algorithms , such as adaptive PPF , learning rate is dictated by the choice of the noise covariance in the prior model of the parameter decoder , which is again chosen empirically [28 , 36 , 38] . Given the significant impact of the learning rate on both the transient and the steady-state performance of closed-loop neurotechnologies , it is important to develop a principled learning rate calibration algorithm that can meet a desired error or convergence time performance for any neural recording modality ( such as spikes , ECoG , and LFP ) and across applications . In addition to neurotechnologies , designing such a calibration algorithm is also of great importance in general signal processing applications . Prior adaptive signal processing methods have largely focused on non-Bayesian gradient decent algorithms . These algorithms , however , do not predict the effect of the learning rate on error or convergence time ( except for a limited case of scalar linear models; see Discussions ) and hence can only provide heuristics for tuning the learning rate [39 , 40] . A calibration algorithm that can write an explicit function for the effect of the learning rate on error and/or convergence time for both linear and nonlinear observation models would also provide a novel approach for learning rate selection in other signal processing domains [41–47] . For example , in image processing [43] , in electrocardiography [41] , in anesthesia control [44] , in automated heart beat detection [46 , 47] , and in unscented Kalman filters [42] , adaptive filters with learning rates are used in decoding system states or in learning system parameters in real time ( see Discussions ) . Here , we develop a mathematical framework to optimally calibrate the learning rate for Bayesian adaptive learning of neural encoding models . We derive the calibration algorithm both for learning a nonlinear point process model for discrete-valued spiking activity—which we term point process encoding model— , and for learning a linear model with Gaussian noise for continuous-valued neural activities ( e . g . , LFP or ECoG ) —which we term Gaussian encoding model . Our framework derives an explicit analytical function for the effect of learning rate on parameter estimation error and/or convergence time . Minimizing the convergence time and the steady-state error covariance are competing requirements . We thus formulate the calibration problem through the fundamental trade-off that the learning rate introduces between the convergence time and the steady-state error , and derive the optimal calibration algorithm for two alternative objectives: satisfying a user-specified upper-bound on the steady-state parameter error covariance while minimizing the convergence time , and vice versa . For both objectives , we derive analytical solutions for the learning rate . The calibration algorithm can pre-compute the learning rate prior to start of real-time adaptation . We show that the calibration algorithm can analytically solve for the optimal learning rate for both point process and Gaussian encoding models . We use extensive Monte-Carlo simulations of adaptive Bayesian filters operating on both discrete-valued spikes and continuous-valued neural observations to validate the analytical predictions of the calibration algorithm . With these simulations , we demonstrate that the learning rate selected analytically by the calibration algorithm minimizes the convergence time while satisfying an upper-bound on the steady-state error covariance or vice versa . Thus the algorithm results in fast and accurate learning of the encoding model . In addition to the encoding model , we also examine the influence of the calibration algorithm on decoding by taking a motor BMI system , which uses discrete-valued spikes or continuous-valued neural activity ( e . g . , ECoG or LFP ) , as an example . We perform extensive closed-loop BMI simulations [38 , 48] that closely conform to our non-human primate BMI experiments [28 , 29 , 49–51] ( see Discussions ) . Using these simulations , we show that analytically selecting the optimal learning rate can improve both the transient operation of the BMI by allowing its decoding performance to converge faster , and the steady-state performance of the BMI by allowing it to learn a more accurate decoder . We also demonstrate that large learning rates lead to inaccurate encoding models and decoders , and small learning rates delay the convergence of encoding models and decoder performance . By providing a novel analytical approach for learning rate optimization , this calibration algorithm has significant implications for closed-loop neurotechnologies and for other signal processing applications ( see Discussions ) .
In this section , we derive the calibration algorithm for continuous signals such as LFP and ECoG . We first present the observation model and the adaptive KF for these signals . We then find the steady-state error covariance and the convergence time as functions of the learning rate . Finally , we derive the inverse functions to select the optimal learning rate . We now follow the same formulation used for continuous-valued signals , such as LFP or ECoG , to derive the calibration algorithm for the discrete-valued spiking activity . The derivation follows similar steps but , due to the nonlinearity in the observation model , has some differences that we point out . Given the nonlinearities , in this case , the calibration algorithm can be derived for the main first objective , i . e . , to keep the steady-state error covariance below a desired upper-bound while minimizing convergence time ( Fig 2; see Discussions ) . For both discrete and continuous signals , we considered a periodic behavioral state ( e . g . , intended velocity ) in the training data for the derivations to satisfy the mild conditions in Appendix C in S1 Text . However , the derivation of ( 7 ) , ( 8 ) and ( 19 ) are based on Have and Mave for the continuous and discrete signals , respectively , which are simply the average values of functions of the state {vt} . So the core information needed in the calibration algorithm is not the state periodicity , but its expected value , which we can compute empirically for any state evolution . As detailed in Appendix E in S1 Text , the periodicity of vt is simply required to ensure that the mean of the prediction covariance St+ 1|t is well-defined at steady state . If we ignore some mathematical rigorousness and instead assume that St+ 1|t has bounded steady-state moments ( which is a relatively mild requirement ) , then this calibration algorithm can be generalized to the case with non-periodic vt directly . That is precisely why , as we show using simulations in the Results section , the calibration algorithm works even in the case of random evolution for the states {vt} in the training experiment . Periodicity is simply required to guarantee the existence of the mean of St+ 1|t at steady state ( instead of assuming this existence ) in the derivations , as detailed in Appendix E in S1 Text . To validate the calibration algorithm , we run extensive closed-loop numerical simulations . We show that the calibration algorithm allows for fast and precise learning of encoding model parameters , and subsequently for a desired transient and steady-state behavior of the decoders ( Fig 1 ) . While the calibration algorithm can be applied to learn encoding models and decoders for any brain state , as a concrete example , we use a motor BMI to validate the algorithm . In motor BMIs , the relevant brain state is the intended movement . The BMI needs to learn an encoding model that relates the neural activity to the subject’s intended movement . We simulate a closed-loop BMI within a center-out-and-back reaching task with 8 targets . In this task , the subject needs to take a cursor on a computer screen to one of 8 peripheral targets , and then return it to the center to initiate another trial [29 , 56] . To simulate how subjects generate a pattern of neural activity to control the cursor , we use an optimal feedback-control ( OFC ) model of the BMI that has been devised and validated in prior experiments [28 , 29 , 48 , 49] and is inspired by the OFC models of the natural sensorimotor system [78–80] . We then simulate the spiking activity as a point process using the nonlinear encoding model in ( 12 ) and simulate the ECoG/LFP log-powers as a Gaussian process linearly dependent on the brain state [55] using the linear encoding model in ( 1 ) . We test the calibration algorithm for adaptive learning of the ECoG/LFP and the spike model parameters . We assess the ability of the calibration algorithm to enable fast and accurate learning of the encoding models , and to lead to a desired transient and steady-state performance of the decoder . To simulate the intended movement , we use the OFC model . We assume that movement evolves according to a linear dynamical model [28 , 29 , 48 , 49] x t + 1 = A x t + B u t + w t , ( 20 ) where x t = [ d t ′ , v t ′ ] ′ is the kinematic state at time t , with dt and vt being the position and velocity vectors in the two-dimensional space , respectively . Here ut is the control signal that the brain decides on to move the cursor and wt is white Gaussian noise with covariance matrix W . Also , A and B are coefficient matrices that are often fitted to subjects’ manual movements [22 , 23 , 28 , 29 , 56 , 80] . Similar to prior work [28 , 29 , 48 , 49] , we write ( 20 ) as [ d 1 ( t + 1 ) d 2 ( t + 1 ) v 1 ( t + 1 ) v 2 ( t + 1 ) ] = [ 1 0 Δ 0 0 1 0 Δ 0 0 α 0 0 0 0 α ] [ d 1 ( t ) d 2 ( t ) v 1 ( t ) v 2 ( t ) ] + [ 0 0 0 0 1 0 0 1 ] [ u 1 ( t ) u 2 ( t ) ] + [ 0 0 w 1 ( t ) w 2 ( t ) ] , ( 21 ) where Δ is the time-step and α is selected according to our prior non-human primate data [28 , 29] . The OFC model assumes that the brain quantifies the task goal within a cost function and decides on its control commands by minimizing this cost . For the center-out movement task , the cost function can be quantified as [28 , 29 , 48 , 49 , 78 , 80] J = ∑ t = 1 ∞ ∥ d t - d * ∥ 2 + w v ∥ v t ∥ 2 + w r ∥ u t ∥ 2 , ( 22 ) where d* is the target position , and wv and wr are weights selected to fit the profile of manual movements . For the linear dynamics in ( 20 ) and the quadratic cost in ( 22 ) , the optimal control command is given by the well-known infinite horizon linear quadratic Gaussian ( LQG ) solution as [28 , 29 , 48 , 49 , 81] u t = - L ( x t - x * ) , ( 23 ) where x* = [d*′ , 0′]′ is the target for position and velocity ( as the subject needs to reach the target position and stop there ) . Here L is the gain matrix , which can be found recursively and offline by solving the discrete-time Riccati equation [81] . By substituting ( 23 ) in ( 20 ) , we can compute the intended kinematics of the subject in response to visual feedback of the current decoded cursor kinematics xt in our simulations [28] . Details are provided in our prior work [28 , 38 , 48] . Note that we use a single OFC model to simulate the brain strategy throughout all closed-loop numerical simulations—i . e . , both during training experiments in which parameters are being learned in parallel to the kinematics being decoded ( Fig 1 ) , or after training is complete and during pure decoding experiments when the learned parameters are fixed and the learned decoder is used to move the cursor . Indeed prior work have suggested that the brain strategy in closed-loop control largely remains consistent , e . g . , regardless of whether parameters are being adapted or not ( e . g . , [22 , 23 , 26 , 28 , 29 , 49 , 82 , 83] ) . The subject’s intended velocity vt is in turn encoded in neural activity . We first test the performance of the calibration algorithm for continuous ECoG/LFP recordings . We then test this performance for discrete spike recordings . For the continuous signals , we simulate 30 LFP/ECoG features whose baseline powers and preferred directions in ( 1 ) are randomly selected in [1 , 6] dB and [0 , 2π] , respectively . The modulation depth , ‖η‖ , in each channel is randomly-selected in [7 , 10] and the noise variances are randomly-selected in [320 , 380] . The initial value , ψ0|0 , and the true value , ψ* , of each channel are selected randomly and independently . The eight targets are around a circle with radius 0 . 3 . Each trial including the forward and the back movement for a selected target in the center-out-and-back task takes 2 secs . During the training experiment , the subject reaches the targets in the counter-clockwise order repeatedly . To assess whether the calibration algorithm can analytically compute the steady-state error covariance and convergence time for a given learning rate accurately , we simulate 3000 trials under each learning rate considered . For spikes , we simulate 30 neurons . Here since the state vt is the intended velocity , we can also interpret ( 13 ) as a modified cosine-tuning model [75 , 84] by writing it as λ c ( v t ) = exp ( β c + ∥ α c ∥ ∥ v t ∥ cos ( θ t - θ c ) ) , ( 24 ) where θt is the direction of vt , θc is the preferred direction of the neuron ( or direction of αc = ‖αc‖[cos θc , sin θc]′ ) , and finally ‖αc‖ is the modulation depth . For each neuron , we select the baseline firing rate randomly between [4 , 10] Hz and the maximum firing rate randomly between [40 , 80] Hz . We select each neuron’s preferred direction in ( 24 ) randomly between [0 , 2π] . The task setup is equivalent to the continuous signal case . We simulate 1000 trials for each learning rate considered . We also examine the effect of the calibration algorithm on kinematic decoding . For continuous signals , we use a KF kinematic decoder as in prior work ( e . g . , [22 , 23 , 55] ) . For the discrete spike signals , we use a PPF kinematic decoder as in prior work ( e . g . , in real-time BMIs [28 , 29] ) . Kinematic decoder details have also been provided in Appendix B in S1 Text for convenience .
We first assess the accuracy of the analytically-computed error covariance and convergence time by the calibration algorithm . As described in detail in Numerical Simulation section , we run a closed-loop BMI simulation in which the subject performs a center-out-and-back task to eight targets in counter-clockwise order . We simulate 30 LFP/ECoG features . We define the convergence time as the time when the estimated parameters reach within 5% of their true values , i . e . , ‖ψt|t − ψ*‖≤0 . 05 × ‖ψ0|0 − ψ*‖ ( so Erest = 0 . 05; as defined before ψt|t , ψ* , and ψ0|0 are the current parameter estimate , the true parameter value , and the initial parameter estimate , respectively . ) Fig 3A shows the true and the analytically-computed error covariance and convergence time as a function of the learning rate , across a wide range of learning rates . The analytically-computed values are close to the true values . From Fig 3A , the average normalized root-mean-squared errors ( RMSE ) between the true and the analytically-computed values for the convergence time and the steady-state error covariance are 3 . 6% and 1 . 6% , respectively ( where normalization is done by dividing by the range of possible convergence time and covariance values ) . Fig 3A shows that as the learning rate s increases , the error covariance increases and the convergence time decreases . Also , the error covariance is inversely related to the convergence time . These trends also demonstrate the fundamental trade-off between steady-state error covariance and convergence time . In the above analysis , we considered estimating the encoding model parameters ψt|t in ( 6 ) . As derived in ( 11 ) , when the noise variance Z in ( 1 ) is unknown , we can also estimate this variance in real time and simultaneously with the parameters . We thus repeated our closed-loop BMI simulations , this time simultaneously estimating the noise variance Zt|t to show that it converges to the true value regardless of the learning rate s . Fig 3B shows that Zt|t converges to the true value with all tested learning rates , which cover a large range ( 5 × 10−7 to 5 × 10−3 ) . Moreover , even when estimating both ψt|t and the noise variance Zt|t jointly , the analytically-computed error covariance is still close to the true one ( normalized RMSE is 4 . 5% ) . Overall , the analytically-computed error covariance is robust to the uncertainty in Zt|t because Zt|t converges to the true value at steady state regardless of the learning rate ( Fig 3B ) . Here we show how the inverse functions in Theorem 2 can be used to select the learning rate . In our example , we require the 95% confidence bound of the estimated encoding model parameters ( i . e . , ±2 standard deviations of error ) to be within 10% of their average value . Thus this constraint provides the desired upper-bound on the steady-state error covariance Vbd . In general , Vbd can be selected in any manner desired by the user . Once Vbd is specified , we use ( 9 ) and find the optimal value of the learning rate as s1 = 5 . 6 × 10−5 . Hence the calibration algorithm dictates that the learning rate should be smaller than s1 to satisfy the desired error covariance upper-bound . Let’s now suppose that we want to ensure that the convergence time is within a given range . In our example , we require the estimation error to converge within 7 minutes , where convergence is defined as reaching within 5% of the true value ( Erest = 0 . 05 ) . This constraint sets the upper-bound on the convergence time to be Cbd = 7min = 420 sec . The calibration algorithm using ( 10 ) dictates that the learning rate needs to be larger than 4 . 75 × 10−5 . Taken together , for the above constraints for error covariance and convergence time , any learning rate 4 . 75 × 10−5 < s < 5 . 6 × 10−5 is admissible . For conciseness and as an illustrative example , we select the learning rate s = 5 × 10−5 , which satisfies both criteria above . In the next section , we examine the effect of this learning rate on the estimated model parameters over time , i . e . , on the adaptation profiles ( Fig 4 ) . We also examined the evolution of the estimated encoding model parameters ψt|t in time , which we refer to as the parameter adaptation profiles . Plotting the adaptation profile provides a direct way of investigating the influence of the learning rate on the estimated encoding model . We plot the adaptation profiles for the optimal learning rate in our example above , i . e . , s = 5 × 10−5 . We also show these profiles for a smaller and a larger learning rate ( Fig 4 ) . We used these adaptation profiles to further assess the accuracy of the calibration algorithm . The adaptation profiles confirm the accuracy of the calibration algorithm as expected from Fig 3A . We used ( 7 ) to find the steady-state error covariance for each learning rate in Fig 4 and consequently to compute the 95% confidence bounds for the parameter estimates ( which are equal to ±2 square-root of the analytically-computed error covariance ) . We then empirically found the percentage of time during which the steady-state parameter estimates were within this 95% bound . If the covariance matrix is accurately computed by the calibration algorithm , then this percentage should be close to 95% . We found that about 96% of the time , the steady-state estimated parameters lie within the 95% confidence bound calculated by the calibration algorithm for all learning rates . Finally , we also simulated the case where the parameters may shift from day to day ( see Discussions ) to see the application of the calibration algorithm in this case . We confirmed , as shown in S1 Fig , that the same KF with a learning rate calculated from the calibration algorithm ( Fig 4B ) can track the parameters and satisfy the criteria on steady-state error and convergence time on both days . In the algorithm derivation and for rigorousness to ensure the existence of the mean of the prediction covariance St+ 1|t at steady state ( instead of simply assuming this existence; Appendix E in S1 Text ) , we assume that the evolution of behavioral state {vt} , e . g . , the trajectory , is periodic in the training data . However , in computing the error covariance and the convergence time , the only aspect of vt needed by the calibration algorithm is not periodicity , but an average of a function of vt over time , which is Have . Indeed , if we assume St+ 1|t has bounded steady-state moments , then our derivation directly applies to the general non-periodic case ( Appendix E in S1 Text , S2 Fig ) . To show that the calibration algorithm also extends to the case of non-periodic state evolutions , we run a closed-loop BMI simulation with a non-periodic trajectory . In this simulation , in each trial , one of eight targets is instructed randomly according to a uniform distribution over the targets . So the trajectory is no longer periodic ( in contrast to when the targets are instructed one by one and in counter-clockwise order ) . The comparison between the true error covariance and convergence time and their values computed analytically by the calibration algorithm are shown in Fig 5A , across a wide range of learning rates . The analytically-computed values are still close to the true values , with an average normalized RMSE of 2 . 1% and 7 . 4% for the steady-state error covariance and the convergence time , respectively . Similarly , when the noise variance Z needs to be estimated , its estimate Zt|t from ( 11 ) still converges to the true value for all learning rates ( Fig 5B ) . Even when estimating Zt|t simultaneously with parameters , the calibration algorithm can approximate the error covariance well ( normalized RMSE is 2 . 6% ) . Taken together , these results demonstrate that the calibration algorithm can generalize to a wide range of problems since the training state-evolution when adapting the encoding models could have a general form . We also validate the calibration algorithm for discrete-valued spiking observations . We run multiple closed-loop BMI simulations with either a periodic or a non-periodic trajectory . The simulation setting is the same as that for continuous signals and given in Numerical Simulation section . Fig 6 shows that the analytically-computed error covariance is close to its true value across a wide range of learning rates with any type of trajectory ( i . e . , periodic or not ) . The average normalized RMSE between the true and the analytically-computed error covariance is around 5% with either periodic or non-periodic trajectory . This result shows that the calibration algorithm can also accurately compute the learning rate effect for a nonlinear point process model of spiking activity . The result also verifies the generality of the calibration algorithm to different state evolution profiles during adaptation , as was the case for continuous signals . In the case of spikes , the inverse function can again be used to select the learning rate for a given upper-bound on the steady-state error covariance . For example , we can require the error covariance to be within 7% of the average values for all parameters , which provides the value of Vbd . Again , Vbd can be selected as desired by the user . Once Vbd is specified , we use the inverse function using Theorem 3 and Eq ( 9 ) and find that the corresponding optimal learning rate r is 10−7 . We also confirm the accuracy of the calibration algorithm using the parameter adaptation profiles . We plot three realizations of the estimated point process parameters , ϕt|t , under different learning rates r to examine whether the 95% confidence bounds computed by the calibration algorithm are accurate ( Fig 7; similar analysis to the case of continuous signals ) . Note that the confidence bounds are given by twice the square-root of the analytically-computed covariance matrix . We use the optimal learning rate computed for our example above , i . e . , r = 10−7 , and a smaller and a larger learning rate in Fig 7 . We find that at steady state , the estimated parameters are within the 95% confidence bound about 96% of time . This shows the accuracy of the analytically-computed confidence bound ( if this bound is correct , about 95% of the time the estimates should be within confidence bounds ) . This result is consistent with the good match between the true and analytically-computed covariances in Fig 6 . Finally , even though the convergence time cannot be analytically obtained in the case of spike observations , it is still significantly affected by the learning rate r . For a small learning rate ( r = 10−9 ) , the parameter estimate ϕt|t does not converge to its true value even in 2000 sec . In comparison , this convergence time is only about 200 sec for an intermediate learning rate ( r = 10−7 ) . Hence to allow for fast convergence , it is critical to select the maximum possible learning rate that satisfies a desired upper-bound constraint on error covariance . This was the basis for the calibration algorithm . The selection of the optimal learning rate is critical not only for fast and accurate estimation of the encoding model , but also for accurate decoding of the brain state . Here we show that the selection of the appropriate learning rate by the calibration algorithm can improve both the transient and the steady-state operation of decoders . We simulate closed-loop BMI decoding under various learning rates . Since the optimal trajectory for reaching a target in a center-out task should be close to a straight line connecting the center to the target , as the measure of decoding accuracy we use the RMSE between the decoded trajectory and these straight lines [22 , 23 , 28 , 29 , 56] ( the error is the perpendicular distance of the decoded position to the straight line at each time ) . To study the effect of the learning rate on steady-state BMI decoding , we adaptively estimate the encoding model parameters under different learning rates . We fix the estimated parameters after varying amounts of adaptation time . We then use the obtained fixed models to run the closed-loop BMI simulations without adaptation . We run these simulations both for continuous LFP/ECoG observations decoded with a KF kinematic decoder , and for discrete spike observations decoded with a PPF kinematic decoder ( Figs 8 and 9 , respectively ) . By comparing the small and medium learning rates , we find that a small learning rate results in a slow rate of convergence for the decoder performance , without improving the steady-state performance ( two-sided t-test P > 0 . 36; Figs 8 and 9 ) . Moreover , large learning rates result in poor and unstable steady-state decoding due to inaccurate estimation of the model parameters . This is evident by observing that for large learning rates , BMI decoding RMSE widely oscillates as a function of time at which adaptation stops for both continuous ECoG/LFP observations and discrete spike observations ( Figs 8B and 9B , respectively ) . This result shows that due to the large steady-state error , steady-state parameter estimates change widely depending on exactly when we stop the adaptation . Thus the decoder does not converge to a stable performance . Taken together , optimally selecting the learning rate to achieve a desired level of steady-state parameter error covariance is also important for fast convergence and accuracy of decoding . It is interesting to note that due to feedback-correction in closed-loop BMI , the decoder can tolerate a larger steady-state parameter error than we would typically allow if our only goal is to track the encoding model parameters . This is evident by noting , for example , that using a learning rate of s = 5 × 10−3 for continuous signals results in a relatively large steady-state parameter error as shown in Fig 4 ( The 95% confidence bound is about ±30% of the modulation depth ) . However , for the purpose of BMI decoding , this learning rate results in no loss of performance at steady state compared to smaller learning rates , and allows for a faster convergence time ( Fig 8 ) . Hence the user-defined upper-bound on the steady-state error covariance is dependent on the application and the goal of adaptation . For closed-loop decoding , a larger error covariance could be tolerated , and as a result , a faster convergence time can be achieved . In contrast , if the goal is to accurately track the evolution of encoding models over time , for example to study learning and plasticity , a lower steady-state error covariance should be targeted . Regardless of the desired upper-bound on the error covariance , the calibration algorithm can closely approximate the corresponding learning rate that satisfies this upper-bound while allowing for the fastest possible convergence .
Developing invasive closed-loop neurotechnologies to treat various neurological disorders requires adaptively learning accurate encoding models that relate the recorded activity—whether in the form of spikes , LFP , or ECoG—to the underlying brain state . Fast and accurate adaptive learning of encoding models is critically affected by the choice of the learning rate [37] , which introduces a fundamental trade-off between the steady-state error and the convergence time of the estimated model parameters . Despite the importance of the learning rate , currently a principled approach for its calibration is lacking . Here , we developed a principled analytical calibration algorithm for optimal selection of the learning rate in adaptive methods . We designed the calibration algorithm for two possible user-specified adaptation objectives , either to keep the parameter estimation error covariance smaller than a desired value while minimizing convergence time , or to keep the parameter convergence time faster than a given value while minimizing error . We also derived the calibration algorithm both for discrete-valued spikes modeled as point processes nonlinearly dependent on the brain state , and for continuous-valued neural recordings , such as LFP and ECoG , modeled as Gaussian processes linearly dependent on the brain state . We showed that the calibration algorithm allows for fast and accurate learning of encoding model parameters ( Figs 4 and 7 ) , and enables fast convergence of decoding performance and accurate steady-state decoding ( Figs 8 and 9 ) . We also demonstrated that larger learning rates make the encoding model and the decoding performance inaccurate , and smaller learning rates delay their convergence . The calibration algorithm provides an analytical approach to predict the effect of the learning rate in advance , and thus to select its optimal value prior to real-time adaptation in closed-loop neurotechnologies . To derive the calibration algorithm , we introduced a formulation based on the fundamental trade-off that the learning rate dictates between the steady-state error and the convergence time of the estimated parameters . Calibrating the learning rate analytically requires deriving two functions that describe how the learning rate affects the convergence time and the steady-state error covariance , respectively . However , currently no explicit functions exist for these two relationships for Bayesian filters , such as the Kalman filter or the point process filter . We showed that the two functions can be analytically derived ( Eqs ( 7 ) , ( 8 ) and ( 19 ) ) and can accurately predict the effect of the learning rate ( Figs 3 and 6 ) . We obtained the calibration algorithm by deriving two inverse functions that solve for the learning rate based on a given upper-bound of the error covariance ( Eq ( 9 ) ) or the convergence time ( Eq ( 10 ) ) , respectively . To allow for rigorous derivations in finding tractable analytical solutions for the learning rate , we performed the derivations for the case in which the behavioral state in the training experiment evolved periodically over time . This is the case in many applications; for example , in motor BMIs , models are often learned during a training session in which subjects perform a periodic center-out-and-back movement . However , we found that the calibration algorithm only depended on an average value of the behavioral state rather than on its periodic characteristics . Indeed , we showed that with a simplifying assumption , the derivation extends to the general non-periodic case ( Appendix E in S1 Text , S2 Fig ) ; moreover , using extensive numerical simulations , we demonstrated that the calibration algorithm can accurately predict the effect of the learning rate on parameter error and convergence time for a general behavioral state evolution in the training experiments ( Figs 5 and 6B ) . The match between the analytical prediction of the calibration algorithm and the simulation results suggest the generalizability of the calibration algorithm across various behavioral state evolutions . We derived the calibration algorithm for Bayesian adaptive filters , i . e . , KF for continuous-valued activity and PPF for discrete-valued spikes . Here the KF and PPF were used to adaptively learn the neural encoding model parameters , which were assumed to be unknown but essentially fixed within the time-scales of parameter learning . This scenario is largely the case that arises in neurotechnologies for learning encoding models/decoders for two reasons . First , in neurotechnologies , such as BMIs , the parameters of the encoding models are initially unknown because they need to be learned in real time during closed-loop operation ( cannot be learned offline and a-priori before actually using the BMI ) . Second , even though these parameters are unknown , they are largely fixed at least within relevant time-scales of parameter learning ( e . g . , minutes ) in BMIs ( and even typically within time-scales of BMI operation in a day , e . g . , hours; see for example [17–19 , 21–24 , 26 , 28 , 29 , 49–51 , 57–64] ) . Even in scenarios where these parameters may change over time for example due to plasticity or task learning , the time-scale of parameter variation will be substantially slower than the time-scale of parameter estimation/learning in the KF or PPF . For example , as we show here and as observed in prior experiments through trial and error , with a well-calibrated adaptive algorithm the parameters can typically be learned within several minutes ( e . g . , [22–29] ) . In contrast , the time-scale of changes in encoding model parameters is typically on the order of days [18 , 19 , 56] . So even in the case that parameters may be changing , for the purpose of selecting the learning rate in the adaptive algorithm , they can be considered as essentially constant . We also showed that the calibration algorithm combined with the Bayesian adaptive filter can be used on an as-needed basis to re-learn parameters in case they shift over these relevant longer time-scales , e . g . , from day to day . Finally , while Bayesian adaptive filters such as the KF and PPF can be used to track time-varying parameters , they can also be used to estimate fixed but unknown parameters as shown both in neurotechnologies and in other applications such as climate modeling , control of fluid dynamics , and robotics [28 , 29 , 65–69] , and confirmed in our derivations and simulations here . In deriving the calibration algorithm , we assumed that recorded signals ( whether continuous or discrete ) are conditionally independent over channels and in time , similar to prior work [17 , 22 , 23 , 26–29 , 49 , 54–58 , 61 , 72] . This assumption enables the derivation of tractable real-time decoders ( i . e . , KF and PPF ) , adaptive algorithms , and in our case the analytical calibration algorithm , for both linear and nonlinear observation models ( Eqs ( 1 ) and ( 12 ) ) for continuous neural signals and binary spike events , respectively . While conditional dependencies could exist in general , prior experiments have shown that algorithms derived with these conditional independence assumptions work well for neural data analysis [17 , 22 , 23 , 26–29 , 49 , 54–58 , 61 , 72] . Finally , given the high dimensionality of neural recordings obtained in current neurotechnologies , modeling correlations between channels would introduce a large number of unknown neural parameters that need to be learned in real time . This real-time learning becomes computationally quite expensive , and would require more data ( and thus longer time in real-time applications ) for parameters to be learned without overfitting . Thus the conditional independence assumption makes the parameter learning algorithms and setups amenable for real-time applications by reducing the number of model parameters and complexity . The selected learning rate in the calibration algorithm depends on the user-specified upper-bound on the error covariance or convergence time . The values of these upper-bounds could be chosen by the user based on the goal of adaptation . If the adaptation goal is to accurately estimate the encoding model parameters ( e . g . , to study learning ) , then the acceptable error upper-bound may be selected to be small . In such a case , the calibration algorithm would select a small learning rate . However , we showed that if the goal of calibration is to enable accurate decoding in a closed-loop BMI , then larger errors in the estimated parameters may be tolerated . This is due to feedback-correction in BMIs , which can compensate for the parameter estimation error ( Figs 8 and 9 ) . The calibration algorithm would then select larger learning rates to improve how fast decoding performance converges to high values . However , even in this case , there is a limit to how large the learning rate can be . A learning rate that is too large will result in unstable and inaccurate performance of the decoder ( Figs 8 and 9 ) . This result shows the importance of the calibration algorithm regardless of the goal of adaptation . The calibration algorithm may also serve as a tool to help examine the interaction between model adaptation and neural adaptation . In closed-loop neurotechnologies , neural representations can change over time resulting in neural adaptation , e . g . , due to learning over multiple days . For example , in motor BMIs , the brain can change its encoding of movement ( e . g . , the directional tuning of neurons ) to improve neuroprosthetic control [17–19 , 56 , 85] . Neural and model adaptation result in a “two-learner system” and can interact [56] . It is important to study whether model adaptation interferes with neural adaptation in these closed-loop systems , and if so whether this interference depends on how fast models are adapted . By accurately adjusting the convergence time and hence the speed of model adaptation , the calibration algorithm may provide a useful tool in studying such interference in careful experiments . Moreover , if neural adaptation is significantly affected by the speed of model adaptation , the calibration algorithm could help carefully adjust this speed for a desired neural adaptation outcome . It is also important to examine this interference problem theoretically [86] . To validate the calibration algorithm , we used a motor BMI as an example . The calibration algorithm , however , can be applied to other closed-loop neurotechnologies that need to decode various brain states , for example , interest score in closed-loop cortically-coupled computer vision for image search [87] or mood in closed-loop DBS systems [88] . Also , while our main goal was to derive the calibration algorithm for closed-loop neurotechnologies , this algorithm can be used in other domains of signal processing . We derived the calibration algorithm to select the learning rate and predict its effect on error and convergence time in Bayesian adaptive filters . Prior work in other signal processing applications have focused vastly on the non-Bayesian LMS or steepest-decent adaptive filters [37 , 40] . However , LMS is only applicable to linear observation models [37] . Moreover , steepest-decent filters that use non-linear cost functions to specify the goal of adaptation cannot predict the effect of learning rate on error or convergence time and thus only provide heuristics for learning rate selection [37] . Finally , LMS or steepest-decent filters are not Bayesian filters , unlike the KF or the PPF ( Eqs ( 3 ) – ( 6 ) and ( 15 ) – ( 18 ) ) . Using a Bayesian filter for parameter adaptation has the advantage that it can extend to nonlinear stochastic observation models ( such as the point process model of spikes ) [28 , 29 , 36] . Here , we derived a learning rate calibration algorithm for Bayesian filters both with continuous linear observation models ( KF ) and with discrete nonlinear observations models ( PPF ) . Importantly , we derived explicit analytical functions ( 9 ) and ( 10 ) to predict the effect of the learning rate on steady-state error and convergence time for a Bayesian filter . This allowed us to analytically compute an optimal value for the learning rate to achieve a desired user-specified performance metric . Our main contribution is the derivation of a novel analytical calibration algorithm for both nonlinear point process and linear Gaussian encoding models ( Eqs ( 1 ) and ( 12 ) ) ; this calibration algorithm optimally selects the learning rate based on the trade-off between convergence time and steady-state error covariance . In deriving closed-form expressions for the calibration algorithm , we needed to analytically compute the steady-state error covariance in both the PPF and the KF . Note that , even in the case of the KF , this analytical computation cannot be achieved through the general steady-state analysis of the KF . First , the steady-state analysis of the KF does not formulate a tradeoff between the steady-state error covariance and convergence time , and thus does not provide a calibration algorithm . Second , in order to derive the calibration algorithm , we need to derive novel analytical closed-form expressions for the steady-state error covariance and convergence time in the KF ( so that we can find the inverse function to compute the optimal learning rate for a given covariance or convergence time ) . To obtain these expressions , we need to find an analytical solution for a special form of the discrete Riccati equation ( DRE ) [89] . While the DRE is solved numerically and recursively in the general steady-state analysis of a KF , there exists no analytical solution with a closed-form expression for a DRE in general . Obtaining such an analytical solution is critical for calculating the optimal learning rate in ( 9 ) and ( 10 ) . Therefore , unlike the steady-state analysis of KF , we additionally had to derive the analytic solution of a special form of DRE first ( Appendix J in S1 Text ) . Third , we also needed analytical expressions for the convergence time of the KF during the transient phase , which again the steady-state analysis of the KF does not provide . Finally , note that we also provide the calibration algorithm for the point process model of the binary spike time-series and thus for the nonlinear PPF in addition to the linear KF . Here our focus was on deriving an analytical calibration algorithm for both nonlinear point process and linear Gaussian encoding models for spikes and continuous neural recordings , respectively . Thus to validate our analytical approach , we used extensive closed-loop Monte-Carlo simulations . These simulations allowed us to examine the generalizability of the calibration algorithm across different neural signal modalities . The closed-loop simulations closely conformed to our prior non-human primate experiments [28 , 29] . Prior studies have shown that these closed-loop simulations can mimic the observed experimental effects and thus provide a useful validation testbed for algorithms [28 , 38 , 48 , 90] . Moreover , the calibration algorithm adjusted the learning rate of adaptive PPF and adaptive KF decoders , which have been shown to be successful for real-time BMI training and control using spikes or LFP in non-human primate and human experiments both in our work and other studies [17 , 21–30 , 55] . However , prior experiments , including ours , selected the learning rates empirically in these decoders . Given that the calibration algorithm is run prior to experiments , and based on the success of adaptive PPF and KF in prior animal and human experiments , we expect our calibration algorithm to be seamlessly incorporated in BMIs regardless of the neural signal modality . The calibration algorithm allows the optimal learning rate to be computed prior to running the adaptation experiments to achieve a predictable speed and accuracy in adaptive learning . Implementing the calibration algorithm in animal models of adaptive BMIs using both spikes and LFP is the topic of our future investigation . Finally , the calibration algorithm has the potential to be generalized to Bayesian filters beyond the KF and PPF , e . g . , the unscented Kalman filter [42] , an adaptive filter with a binomial distribution as the observation model [44] , or hybrid spike-LFP filters [91] . The derivations of Eqs ( 7 ) and ( 8 ) in theorems 1 and 3 are based on the recursive equation for estimation error dynamics , which is derived from the desired Bayesian filter . This implies that for other observation models different from a linear model with Gaussian noise in KF or a nonlinear point process model in PPF , once we write down their corresponding Bayesian adaptive filters [92] , we can derive the calibration algorithms by writing the corresponding recursive error equations . Thus this calibration algorithm has the potential to be generalized and applied to other types of signals with various stochastic models . This will be a topic of our future investigation . | Closed-loop neurotechnologies for treatment of neurological disorders often require adaptively learning an encoding model to relate the neural activity to the brain state and decode this state . Fast and accurate adaptive learning is critically affected by the learning rate , a key variable in any adaptive algorithm . However , existing signal processing algorithms select the learning rate empirically or heuristically due to the lack of a principled approach for learning rate calibration . Here , we develop a novel analytical calibration algorithm to optimally select the learning rate . The learning rate introduces a trade-off between the steady-state error and the convergence time of the estimated model parameters . Our calibration algorithm can keep the steady-state parameter error smaller than a desired value while minimizing the convergence time , or keep the convergence time faster than a desired value while minimizing the error . Using extensive closed-loop simulations , we show that the calibration algorithm allows for fast learning of accurate encoding models , and consequently for fast convergence of decoder performance to high values for both discrete-valued spike recordings and continuous-valued recordings such as local field potentials . The calibration algorithm can achieve a predictable level of speed and accuracy in adaptive learning , with significant implications for neurotechnologies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"learning",
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"classical",
"mechanics",
"engineering",
"and",
"technology",
"electronics",
"applied",
"mathematics",
"membrane",
"potential",
"social",
"sciences",
"random",
"variables",
"electrophysiology",
"... | 2018 | Optimizing the learning rate for adaptive estimation of neural encoding models |
The ability of human immunodeficiency virus type 1 ( HIV-1 ) to develop high levels of genetic diversity , and thereby acquire mutations to escape immune pressures , contributes to the difficulties in producing a vaccine . Possibly no single HIV-1 sequence can induce sufficiently broad immunity to protect against a wide variety of infectious strains , or block mutational escape pathways available to the virus after infection . The authors describe the generation of HIV-1 immunogens that minimizes the phylogenetic distance of viral strains throughout the known viral population ( the center of tree [COT] ) and then extend the COT immunogen by addition of a composite sequence that includes high-frequency variable sites preserved in their native contexts . The resulting COT+ antigens compress the variation found in many independent HIV-1 isolates into lengths suitable for vaccine immunogens . It is possible to capture 62% of the variation found in the Nef protein and 82% of the variation in the Gag protein into immunogens of three gene lengths . The authors put forward immunogen designs that maximize representation of the diverse antigenic features present in a spectrum of HIV-1 strains . These immunogens should elicit immune responses against high-frequency viral strains as well as against most mutant forms of the virus .
The failure of AIDS vaccine efforts in the past 20-plus years is thought to be due , in part , to the enormous viral antigenic diversity found within and among patients with human immunodeficiency virus type 1 ( HIV-1 ) infection . However , until recently , relatively little effort had been devoted to choosing particular viral variant sequences or designing sequences to include within vaccines [1 , 2] . There were early attempts to design vaccines by concatenating commonly recognized T cell and antibody epitopes [3] , but these did not produce a viable vaccine candidate . New methods of combining epitopes are being explored in vaccine design , including production of pseudoprotein strings of T cell epitopes [4] , and the synthetic scrambled antigen vaccine ( SAVINE ) [5] , which employs consensus overlapping peptide sets from HIV-1 proteins scrambled together . Focusing on the use of whole viral protein sequences , natural strains ( NSs ) as well as consensus ( CON ) sequences are being used as a means to minimize the abrogating effect of antigenic diversity in vaccine antigens [2 , 6 , 7] , as are the inferred most recent common ancestors ( MRCA , or ANC ) [6 , 8–10] of targeted virus populations defined as sequences that reside at the basal node of the set of in-group sequences in a phylogenetic tree reconstruction [11] . HIV-1 env sequences representing both the CON and ANC have been prepared and studied , but neither has generated exceptionally broad humoral immune reactivity in initial small animal studies [7 , 12] . In an effort to develop antigens that capture both the summary of circulating variation found in CON estimates , and the coupling of mutations generated with inferred ANC sequences , we have developed an alternative computational method that reconstructs the ancestral state sequence at the center of tree ( COT ) ( [13] and Rolland M , Jensen MA , Nickle DC , Learn GH , Heath L , et al . , unpublished data ) . The COT sequence explicitly minimizes genetic distance , as does the CON , and because it is derived from a phylogenetic tree , it embodies the most likely mutational coupling relationships found in the ANC . Despite these efforts , it may be that no single unit-length antigen , including any NS , CON , ANC , or COT , will encompass sufficient antigenicity to elicit protective immune responses against a broad array of viruses [7 , 12] , as will be required of an AIDS vaccine . This led us to hypothesize that we would need more than one antigenic sequence , or greater than one gene length of the antigen , to elicit protection against the broad antigenic diversity encountered in natural infections . However , cocktails of large numbers of native , full-length NS antigens would quickly become unmanageably complex for practical use as vaccines . Here , we propose a means to cope with HIV-1 diversity by engineering vaccine antigen constructs to include short protein sequences present at high frequencies in natural viral populations . Currently , this method is explicitly directed toward developing CD8+ cytotoxic T lymphocyte ( CTL ) responses , which are critical to controlling viremia during infection [14–17] . Because the cumulative strength of the CTL-mediated immune response depends on the presence of recognizable epitopes ( often approximately nine amino acids in length ) in the target proteins , it is logical to seek to maximize epitope coverage within a vaccine antigen . However , although substantial , our current catalog of known CTL epitopes appears to be woefully incomplete [18] , hence our strategy relies on the universe of HIV sequences and not solely on known epitope content . Thus , here we will define coverage as the sum of the frequencies of all nine amino acid segments ( 9mers ) where the frequency is derived from random independent HIV-1 subtype B isolates found in the vaccine construct . As our epitope catalog increases and our knowledge of protein degradation , CTL epitope binding , and HLA presentation is expanded , this epitope-specific data can be integrated into the measure of coverage ( e . g . , by weighting epitope frequencies in accordance to their relative “importance” when computing coverage ) . In this study , we applied our method to Nef because it is highly variable and is potentially very difficult to design an immunogen against , and to Gag because it is immunologically important yet more conserved . We considered subtype B sequences because more immunological information is available about this subtype than any other . This clearly makes the vaccine construct described here as region-specific because of the biogeographic nature of the distribution of viral subtypes across the globe [19] . However , our purpose is to illustrate and demonstrate that this method has promise at producing a vaccine against highly variable infectious agents such as HIV .
( 1 ) A COT sequence is calculated as described ( [13 and Rolland M , Jensen MA , Nickle DC , Learn GH , Heath L , et al . , unpublished data ) from a phylogenetic tree that captures the relationships among genes in the sample using maximum likelihood methods [23] . Briefly , from aligned sequences we estimate a maximum likelihood tree under a HKY + Γ + I model of evolution in PAUP*v4beta10 [24] . The resulting tree is re-rooted at the point that describes the least-squares distance to all the tips on the phylogeny ( the COT node ) . We then infer the maximum likelihood state using the same model of evolution as above . ( 2 ) A table of unique 9mer peptides [20 , 21] with their corresponding frequencies ( the 9mer distribution ) is constructed from translated protein sequences . To illustrate this , note that if our sample contained N identical sequences of length L each , but every 9mer in the COT peptide library was unique , then each peptide would be at equal frequency . On the other hand , if every sequence were different from all others , to the extent that no 9mer was represented twice , the frequency for each peptide would equal . Actual samples will yield an intermediate distribution that can be exploited for vaccine design ( see Figure 1 ) . We used this distribution to compute “coverage”; that is , as we select candidate fragments to be included in the potential vaccine , we will select only those fragments that are highly represented under the 9mer curve . ( 3 ) Unique or rare 9mers , which by definition are unlikely to be common in circulating viral strains , are likely to derive from low-fitness variants [25 , 26] and , because of their low frequency , have low probability of being incorporated in our vaccine constructs . Specifically , we calculate the frequency of all observed mutations at each site , and revert any mutation with a frequency below a fixed “smoothing” threshold , M , to the corresponding character in the COT sequence . All 9mers present in the COT sequence are then removed from the 9mer distributions before proceeding to the next step . ( 4 ) Given a fixed window size F ( ranging from 9 to L , where L = the length of the protein sequence [we start with 9 because that is the size of the peptide that is most often found to encode epitope sequences] and a stride parameter S [ranging from 1 to L , the length of the protein] ) , we generate all sequence fragments from the sampled sequence by iteratively shifting the frame S residues at a time . We then compute the coverage for each sequence fragment not already present in the COT sequence , and append the sequence fragment to the COT string , compressing with possible overlap to yield a COT+ molecule with the highest ratio of coverage per length . Specifically , fragments are chosen by their level of coverage and whether or not they have differences with respect to the COT sequences . The highest coverage fragments are chosen first , with subsequent fragments with lower coverage being chosen subsequently . This process is repeated until the sequence of desired length is derived . The length of the COT+ sequence is arbitrarily chosen by taking into account plasmid size limitations for producing and delivering an antigen construct and the amount of variability that can be efficiently incorporated as the length is extended , which in turn depends on the variability found in circulating strains that have been sampled for a particular gene . We note that it is possible to arrange the order in which sequence fragments are added to COT+ to maximize the overlap of consecutive fragments , thereby further compressing the antigen . ( 5 ) The values of window size F , stride step S , and smoothing threshold M are varied to achieve maximum coverage ( Figure 2A and 2B ) . We compared our constructs of various lengths to randomly drawn sequences from the curated dataset of 169 sequences using the optimal values for F and S . We generated COT+ for both Gag and Nef at ever-increasing unit protein lengths until we reached 100% coverage . For comparison , we concatenated randomly sampled protein sequences 100 times at ever-increasing unit protein lengths from both Gag and Nef and measured 9mer coverage across the same gene lengths ( Figure 3A and 3B ) . We chose protein unit length for our comparison , but COT+ can be derived for any partial unit protein length desired . To ensure that we were not overestimating the coverage of our constructs due to the finite size of our dataset , we repeated our approach using 10-fold cross-validation . We partitioned the data into ten sets , and for each we estimated COT+ from the remaining 90% of the data and then measured its coverage of the sequences in the chosen set . Thus , given that our assessment of coverage is on a set of sequences not seen in training , we yield an estimated lower bound on the coverage we would obtain for a larger population . We report this lower bound as a percentage of similarity to the estimated upper-bound COT+ , derived from training and testing on all 169 sequences for both Gag and Nef . This study is geared to understand the effect of sample size on the on the COT+ estimation and to show that we are not overfitting the estimations .
Although the list of known HIV-specific CD8 T-cell epitopes is far from complete [18] , we sought to determine how well our 9mer coverage-based constructs identified known epitopes . To this end , we obtained all available HIV CTL epitopes from the Los Alamos National Laboratory ( LANL ) HIV immunology database [27] and counted the perfect matches to our constructs . Because many true epitopes are listed multiple times and larger peptides are reported frequently where the true epitope is embedded , we curated the database to remove any larger epitope that had a smaller embedded known epitope with the same supertype HLA response pattern , and removed any duplicates .
We inferred COT sequences from databases of Gag and Nef protein sequences from HIV-1 subtype B from 169 independently infected individuals , and then added frequently observed variant 9mer peptides to create COT+ sequences . The frequencies of unique 9-mer peptides are shown in Figure 1 . We find that maximal coverage occurs when the window size , F , is 17 , the stride length , S , is 1 , and when smoothing M is 0 ( Figure 2A and 2B ) . One possible reason for why an S value equal to 1 leads to the highest coverage is that it gives every amino acid in the sequences a chance to be in every possible position in a high-scoring peptide . Counterintuitive to this is the observation that S values greater than 1 do not get penalized with big drops in 9mer coverage . We think the explanation for this observation has to do with the fact that even with S larger than 1 , every amino acid in the sequences is still considered when building a construct . This is exemplified by the fact that the biggest drops in 9mer coverage come when S is larger than F , because it is in this parameter space that some amino acids have the probability of not being considered at all in the resulting construct . Adding peptides to generate a three-gene-length COT+ construct achieved 82% 9mer coverage for Gag and 62% for Nef , whereas an antigen constructed from several random concatenated database sequences [22] needed to achieve the same level of coverage required ten gene lengths for Gag and approximately 11 for Nef ( Figure 3A and 3B ) . When COT+ is compared with 100 constructs of the same length obtained by concatenating randomly selected sequences from the Los Alamos National Laboratory database [22] , the COT+ estimate had a higher level of coverage in every case ( randomization test , p < . 01 ) for both Gag and Nef . The flattening of the curves in Figure 3A and 3B suggests that after the COT+ construct has grown past a few gene lengths , the benefit of adding more length is dramatically reduced . For example , the extension of the COT+ construct from one to three gene lengths results in a 16% increase in coverage for Gag and a 13% increase in coverage for Nef . However , extending COT+ from three to five gene lengths yields only 5% additional coverage for both Gag and Nef . The COT+ sequence reaches 100% coverage at 33 gene lengths for Gag and 67 gene lengths for Nef , while the randomly sampled sets reach 100% coverage only after all 169 sequences are included . The latter observation is due to the fact that many of the mutations found in HIV are private ( i . e . , found only within the lineage infecting a particular person ) . When applied to small datasets , our algorithm generates COT+ constructs with high coverage . An extreme example is making a three-gene construct from just three genes in the training set . In this scenario , we can trivially achieve 100% coverage . The larger the training set , the lower the coverage in a three-gene-length vaccine construct . A 10-fold cross-validation study was therefore designed to determine the effects of sample size on our COT+ constructs . Specifically , at three protein lengths , the cross-validated coverage of Gag and Nef are 96% and 93% , respectively . This suggests that for both proteins these inferences are generalizable across HIV-1 subtype B and that adding more sequence data into the training dataset would add very little to these estimations . That is to say , 10% of the original 169 sequences produce estimations of the COT+ that are highly consistent with the estimations from the entire dataset , supporting the notion that there is a saturation effect and that adding sequences beyond the 169 will not give rise to better estimations . Assessing the inclusion of functional CTL epitopes in our constructs is problematic . The majority of the known CTL epitopes were mapped using peptides derived from a limited number of HIV strains ( e . g . , laboratory-adapted strains and consensus sequences ) . The CTL database is also incomplete ( e . g . , a recent study that used a subset of autologous peptides from a single patient enabled recognition of 28% more epitopes in the virus than were previously reported [18] ) , and it is unclear whether characterized epitopes form an unbiased sample of naturally occurring antigenic peptides . It is also necessary that the epitope be presented in the proper context of adjacent amino acids for efficient immunoproteasome cleavage . We therefore assessed the overall size of the peptides needed to obtain maximal coverage of included 9mers . As shown in Figure 2A and 2B , maximal coverage of both the Gag and Nef datasets was obtained with a window size of 17 amino acids and a stride of one amino acid and no smoothing required ( see Methods ) . Hence , we are able to construct immunogens that preserve much of the extended local amino acid environment of the epitope without sacrificing coverage . This enhances the likelihood that the desired peptide epitope will be properly cleaved by cellular proteases and presented efficiently on HLA molecules . Next , we assessed the inclusion of known CTL epitopes in our constructs by comparing the number of known HIV-1 Nef and Gag epitopes [27] contained in the three-gene-length COT+ constructs to that of 1 , 000 combinations of three randomly selected database sequences ( Figure 4 ) . Sequences from the viral strains used to map these epitopes were excluded from the randomization study . Although our algorithm does not attempt to explicitly enrich for known CTL epitopes , the number of known epitopes in COT+ is significantly higher than in a random three-gene construct ( p < 0 . 001 ) for both Gag and Nef . This suggests that COT+ provides a substantial boost in the number of epitopes shared between the immunogen and a random circulating database variant , and thus may have enhanced potential as an immunogen .
COT+ constructs provide a means to extensively compress epitope variation into an immunogen of minimal size . Much of the known variation of both the relatively conserved HIV-1 Gag gene and the quite variable Nef gene can be successfully compressed into COT+ constructs of a few gene lengths . Little increases in variation coverage are noted , however , beyond three to four gene lengths . Coverage grows with length approximately in a y = mlog ( x ) + b form where y is coverage and x is length of the construct . The difference between COT+ construct of Gag and Nef can be broken down into these terms . The coverage intercept parameter b is higher for Gag constructs than for Nef simply because Gag is a more conserved protein than Nef . However , the parameter m is larger for Nef than it is for Gag because the benefits of 9mer compression on coverage are higher with constructs made from variable proteins . Our COT+ generation algorithm is a rapid , computationally efficient heuristic approximation , though it is not guaranteed to find the antigen that achieves maximal epitope coverage for a fixed length . More computationally intensive approaches , such as genetic algorithm searches or approximate solutions to the classic Traveling Salesman problem ( see http://mathworld . wolfram . com/TravelingSalesmanProblem . html ) , could also be brought to bear on the problem of antigen design . Surprisingly , selecting the high-frequency 9mers alone and appending them to the COT sequence does poorly in terms of total coverage ( unpublished data ) . This observation is due to the fact that many of the 9mers do not overlap , and therefore the fragments cannot be efficiently joined . By going back and selecting high-coverage peptide windows from the original data , we obtain better compression in the vaccine construct leading to higher coverage constructs for the same length . It is a reasonable assumption that the retention of native protein structures might be advantageous in generating CTL epitopes , since epitopic peptides are generated in vivo by protein degradation within infected cells . Nef and Gag COT clearly adopt a native structure , as they retain biological activity ( Rolland M , Jensen MA , Nickle DC , Learn GH , Heath L , et al . , unpublished data ) . However , the extended COT+ component of antigens generated in the manner proposed here does not preserve a sequence that is necessarily collinear with the native gene over the second and third gene lengths ( Figure 5A ) . Hence , we have also considered additional means of optimizing immunogen structures that also preserve native structure . First , we can assemble high-frequency variable elements in a pattern collinear with the native gene , with some segments redundant with COT to retain collinearity ( Figure 5B ) . We can also use NS sequences in combination with the COT sequence to optimize coverage ( Figure 5C ) . We can also do very well in generating coverage by exclusive use of NS sequences that maximize 9mer coverage ( Figure 5D ) . Although it is not guaranteed , these additional constructs ( Figure 5B–5D ) should have biologically acceptable tertiary structures . The COT+ approach captures more of the 9mer distribution and more of the known CTL epitopes than any of the potential constructs presented here . Applying high-frequency peptides onto COT to create a collinear pattern provides the second highest level of diversity and epitope enrichment , but the use of COT plus two NSs is not beneficial relative to judicious choice of three NSs . Last , it should be noted that all of these methods substantially exceed the coverage afforded by the use of a single strain as a vaccine . Immunodominance gives rise to a rank order of immune responses to specific epitopes [28] , and the underlying biological mechanisms giving rise to these rank orders are poorly understood . The antigen designs we report here do not take immunodominance into account . One can argue that the combination of epitopes we have derived could elicit an immunodominant response that does not reflect what is found in circulating HIV strains and hence could be a poor choice for vaccine design . However , since the strings of peptides in our immunogen design are captured by their frequency in the circulating viral population , we surmise that these antigens have epitopes that are shared across many potential challenge strains and could thus lead to potentially broad immune response . However , immunodominance rank order patterns can be partially illuminated by expressing epitopes from different vaccine vectors [29–31] . By vaccinating with different combinations of vectors encoding a single or more antigens , they found that using separate vectors elicited broader CD8+ T cell responses . Because COT+ is directed towards capturing high-frequency fragments from a variable protein , it is well-suited to being expressed as segments on separate vectors . The COT+ algorithm can be generalized to produce sets of immunogens that can take advantage of this phenomenon . COT+ constructs are able to capture significantly more known epitopes and potential antigen variability than much longer constructs composed by combining circulating strains . Considering the substantial expense and difficulty involved in production and testing of candidate vaccines , careful crafting of potential antigens using computational methods , including that shown here , may be beneficial . Furthermore , this approach is applicable not only to HIV vaccine design , but to the design of vaccines targeting any pathogen capable of rapid escape from immune recognition . | The ability of human immunodeficiency virus type 1 ( HIV-1 ) to acquire mutations that preserve virus viability yet evade immune responses contributes to the current failure in producing a vaccine . We describe the generation of candidate HIV-1 immunogens that include multiple forms of variable elements of the virus including some that retain colinearity with the virus and thus are expected to retain protein function . These antigens compress the variation found in many viral strains into lengths suitable for vaccine immunogens . For example , we can capture 62% of the variation found in the Nef protein and 82% of the variation in the Gag protein into immunogens of three gene lengths . We put forward immunogen designs that maximize representation of the diverse antigenic features present in a spectrum of HIV-1 strains . These immunogens should elicit immune responses against high frequency viral strains as well as against most mutant forms of the virus . | [
"Abstract",
"Introduction",
"Methods",
"Known",
"Epitope",
"Coverage",
"Results",
"Discussion"
] | [
"viruses",
"infectious",
"diseases",
"virology",
"immunology",
"microbiology"
] | 2007 | Coping with Viral Diversity in HIV Vaccine Design |
Syncytins are envelope genes of retroviral origin that have been co-opted for a role in placentation . They promote cell–cell fusion and are involved in the formation of a syncytium layer—the syncytiotrophoblast—at the materno-fetal interface . They were captured independently in eutherian mammals , and knockout mice demonstrated that they are absolutely required for placenta formation and embryo survival . Here we provide evidence that these “necessary” genes acquired “by chance” have a definite lifetime with diverse fates depending on the animal lineage , being both gained and lost in the course of evolution . Analysis of a retroviral envelope gene , the envV gene , present in primate genomes and belonging to the endogenous retrovirus type V ( ERV-V ) provirus , shows that this captured gene , which entered the primate lineage >45 million years ago , behaves as a syncytin in Old World monkeys , but lost its canonical fusogenic activity in other primate lineages , including humans . In the Old World monkeys , we show—by in situ analyses and ex vivo assays—that envV is both specifically expressed at the level of the placental syncytiotrophoblast and fusogenic , and that it further displays signs of purifying selection based on analysis of non-synonymous to synonymous substitution rates . We further show that purifying selection still operates in the primate lineages where the gene is no longer fusogenic , indicating that degeneracy of this ancestral syncytin is a slow , lineage-dependent , and multi-step process , in which the fusogenic activity would be the first canonical property of this retroviral envelope gene to be lost .
Syncytins are genes of retroviral origin that have been co-opted by their host for a function related to placentation . They correspond to the envelope ( env ) gene of ancestral retroviruses that entered the germline of evolutionarily distant animals and were endogenized ( reviewed in [1] , [2] ) . Two such genes have already been identified in simians , namely syncytin-1 [3]–[7] and -2 [8] , [9] , as well as two distinct , unrelated ones in muroids , syncytin-A and –B [10] , one in leporids , syncytin-Ory1 [11] , one in carnivores , syncytin-Car1 [12] , and more recently one in ruminants , syncytin-Rum1 [13] . Their canonical characteristic features -allowing them to be named “syncytins”- comprise i ) placenta-specific expression , ii ) cell-cell fusion activity , and iii ) conservation in evolution of mammalian species for extended periods of time [e . g . >10 million years ( My ) ] . Syncytin proteins are expected to participate in the formation of the placental syncytiotrophoblast at the maternal-fetal interface , via fusion of the mononucleated cytotrophoblasts [3] , [14]–[18] . Some of them additionally possess an immunosuppressive activity , as classically observed for infectious retroviral envelope glycoproteins , which may be involved in maternal-fetal tolerance [19] . The direct involvement of syncytins in placentation has been recently demonstrated unambiguously through the generation of knockout mice for syncytin-A and –B [20] , [21] , whose embryonic placenta displayed defects in cell-cell fusion , resulting in decreased maternal-fetal exchange and impaired embryo survival . A remarkable feature of syncytins is that these genes , which have been acquired “by chance” , repeatedly and independently in the course of evolution , are “necessary” for a basic function in placental mammals ( reviewed in [2] ) . It has therefore been proposed that syncytins might be present in all placental mammals , and that the capture of a founding syncytin by an oviparous ancestor has been pivotal for the emergence of placentation -approximately 150 My ago . This founding syncytin would then have been subsequently replaced in the diverse emerging mammalian lineages , upon successive and independent germline infections by new retroviruses and co-optation of their env gene , each new gene providing its host with a positive selective advantage . This would account for the diversity in the nature and age of the captured syncytins that can be presently identified , concomitant with the diversity of placental architectures [2] . A consequence of this evolutive scenario is that evidence should exist for “decaying syncytins” , which could possibly be unraveled in eutherian mammals . In fact , screening of the human genome for envelope protein-coding sequences [22]–[24] led to the identification , in addition to syncytin-1 and syncytin-2 , of three env genes that share some but not all of the characteristic features of syncytins: the envR , envV and envPb genes . The envR gene is strongly expressed in the placenta [22] , [25] but lacks fusion activity due to a stop codon before the membrane-anchoring domain of the protein , that most probably arose very early in primate evolution being already present in Old World monkeys [26] . The envV gene is specifically expressed in the placenta , but its fusogenicity could not be demonstrated , either due to an intrinsic defect , or to the lack of its cognate receptor on the panel of cells used for the ex vivo cell-cell fusion assay [24] . Yet , envV can be found in all simians , with the orthologous copy displaying a complete open reading frame ( ORF ) suggesting that it has been subject to purifying selection ( although its function was not investigated ) [27] . Finally , envPb in humans was demonstrated to be fusogenic , and orthologous copies can be found in most simians , but it is only poorly expressed and in a non-specific manner in all the human tissues tested , including the placenta [24] , [28] . Altogether , close examination of the status of these env genes shows that , presently , they cannot be formally considered as syncytins , but still possess some characteristic features suggesting that they could be the remnants of ancestrally co-opted syncytins that are progressively losing their function in some primates . This could be a consequence of the incorporation into the genome of the latter of “new” syncytin genes –such as syncytin-1 and syncytin-2- which might have functionally replaced them in the course of primate evolution , for syncytin-2 about 60 My ago ( Mya ) , after the divergence between prosimians and simians , and for syncytin-1 about 45 Mya , after the divergence between New and Old World monkeys [29] . In fact , the natural history and the time course of these processes might be different depending on the env gene . For instance , the intrinsic lack of fusogenicity of envR due to a stop codon upstream of the membrane-anchoring domain sequence in all simians , the complete loss of the gene in the gorilla [26] , and finally the occurrence of a high polymorphism in this gene among the human population , with 1% of the population being homozygous for a premature second stop codon subsequently acquired in the N-terminal domain of the protein [30] , strongly suggest an early loss of function in placentation . In humans , envV has no demonstrable fusogenic activity , but this property has not been assayed for the orthologous genes found in other simians , where it could be detected . Here , we cloned the envV genes from a series of primates where they have been shown to be present over a 45 My period of time , from New World monkeys to humans , and assayed the cloned genes for their fusogenic activity . A scheme for the “life cycle” of syncytin genes in evolution , with recurrent new incorporation into and progressive decay within the genomes of their hosts can be proposed to account for the observed status of these genes in placental mammals .
Previous studies have shown that ERV-V is a very ancient endogenous retrovirus [24] , which entered the primate genome after the simian and prosimian divergence but before the separation of New World and Old World monkeys , more than 45 Mya ( Figure 1 ) [27] . It was also shown that the unique human envV gene -at chromosomal position 19q13 . 41- is in fact part of a post-integrative provirus duplication , with the envV1 ( C-terminally truncated ) and envV2 ( full-length ) gene sequences −20 kb apart- displaying >94% nucleotide identity ( Figure 2A , 2B ) and frequent events of gene conversion [27] . Using the human env genes as a query , we screened orthologous env genes in the available primate genome libraries ( http://genome . ucsc . edu and http://www . ensembl . org ) , in order to design specific PCR primers flanking either the envV1 or the envV2 gene . Using appropriate primer pairs , both genes were PCR-amplified and cloned into a CMV-driven expression vector to assay their functional activity . Amplified envV1 and envV2 sequences could be recovered from a wide range of simian species , including hominoids ( human , chimpanzee , gorilla , orangutan and gibbon ) , Old World monkeys ( macaque , baboon , African green monkey and langur ) and New World monkeys ( marmoset , cotton top tamarin and white-faced saki ) , consistent with their previously documented occurrence in the simian lineage ( they were not found in prosimians ) [27] . For each amplified fragment , several cloned genes were sequenced to recover only those devoid of PCR-induced mutations . The reference sequences ( sequences deposited in GenBank with accession numbers KC010496-KC010519 ) were determined by sequencing the whole PCR product before cloning , and were confirmed -when available- by the corresponding sequences found in genome databases . As illustrated in Figure 2C , all primate envV2 genes among the 12 species tested display a full-length ORF ( with high sequence conservation , see below ) , whereas the envV1 genes were severely altered , with evidence for frameshift mutations ( human , langur and common marmoset ) , stop codons ( gibbon and orangutan ) and deletion ( orangutan ) ( Figure 2C ) . Of note , the other primate envV1 genes , namely from chimpanzee , gorilla , macaque , baboon , African green monkey , cotton top tamarin and white-faced saki , have a full-length ORF ( still with significant sequence conservation , i . e . 84 to 99% nucleotide identity , most probably due to previously described gene conversion events with envV2 [27] ) , and all were therefore also cloned for further studies . The main expected function for captured retroviral env genes with specific expression in the placenta and involvement in syncytiotrophoblast formation is cell-cell fusion . This was assayed as previously described [8] by transient transfection of cell lines in culture with the above env-expression vectors , and follow-up of syncytium formation 1–2 days post-transfection ( Figure 3A , 3B ) . As illustrated in Figure 3B , 3C and in agreement with previously published results [24] , the human envV1 and envV2 genes were found to have very limited –if any- fusogenic activity . However , and rather surprisingly , Figure 3C also shows that all four tested envV2 genes from Old World monkeys are highly fusogenic , as well as that from marmoset –but not that from the two other New World monkeys , i . e . tamarin and saki . It can also be observed that the envV2 genes from gibbon , which is part of the hominoids , display a weak , intermediate fusogenicity , and that envV1 from all simians is fusion-negative ( Figure 3C ) . Of note , the data in Figure 3B , 3C were obtained with human 293T cells , i . e . with cells where the receptor for EnvV –still to be identified- should in principle be best fitted to the human EnvV protein . To eliminate any possible artifact due to unexpected species-dependent receptor properties that would discriminate between the Old World monkey –and the New World monkey marmoset- and the other primate EnvV2 proteins , we carried out the same assay with the evolutionarily “remote” G355 . 5 cat cells ( from the superorder Laurasiatheria , which diverged from the superorder Euarchontoglires to which belong primates more than 100 Mya ) : as illustrated in Figure 3D , the same fusion profile was obtained ( and similarly with other cell lines from human , simian and rodent ) , thus strongly suggesting that the observed differences are intrinsic differences in fusogenic activity of the EnvV2 proteins . To analyze further the molecular basis of the observed differences , we carried out a series of control experiments . First , we checked that representative cloned envV2 genes that are fusion-negative indeed directed the expression of a membrane-associated protein , similarly to fusion-positive genes used as controls . To do so , we tagged the Env proteins from the 3 New World monkey genes as well as from the human and macaque genes , with the hemagglutinin ( HA ) epitope . We placed it at the C-terminus of the protein , where the HA-tag is less likely to alter protein folding and function . Indeed , as illustrated in Figure 4A , the HA-tag had only a minimal effect , if any , with the macaque and marmoset EnvV2s being still fusogenic , and the human as well as the two tamarin and saki New World monkey EnvV2s still fusion-negative . Expression of the proteins at the level of the plasma membrane was then investigated by Western blotting , after biotinylation and subsequent streptavidin isolation of the biotinylated cell surface proteins . As illustrated in Figure 4B , one specific band with an apparent molecular mass of ∼80 kDa , most probably corresponding to the full-length Env precursor , was observed . All envelope proteins were found to be equally expressed and at the expected location on the plasma membrane . It can therefore be concluded that the lack of fusogenicity of the human , tamarin and saki EnvV2s is not simply related to a lack of expression or a mislocalization of the EnvV2 proteins . A second series of experiments was further devised to determine whether the differences in fusogenic activity among the EnvV2 proteins could be due to a shift in or loss of their capacity to recognize their cognate receptor . Accordingly a series of chimaeras were constructed and assayed for fusion activity . As illustrated in Figure 5 for the human/macaque chimaera , the loss of function of human EnvV2 is not simply due to a defect in the SU moiety –which carries the receptor binding domain- since the human SU is fully active when associated with the macaque TM subunit . Yet the reverse chimaera retains some fusion activity , suggesting that the human TM cannot be considered as defective per se , but rather that fusogenicity is the result of complex structural and dynamic interactions between both subunits , as examplified for several retroviral envelope proteins by specific mutations/truncations ( reviewed in [31] ) . This conclusion is further supported by the two reverse chimaeras involving only the exchange of the cytoplasmic tails between the human and macaque proteins , which both retain some fusion activity . In conclusion , the envV2 gene has clearly conserved , in all the tested Old World monkeys , the fusogenic activity that was most probably associated with the ERV that primitively entered the primate lineage , but this property has been lost on several occasions , including in the human lineage , in some New World monkeys , and to some extent in non-human hominoids . It can be also concluded that this loss of fusion activity is not associated –at least for the human gene- with an inadequacy between the receptor for this protein and the EnvV2 SU subunit . This raises two questions: first , can envV2 still be a “syncytin” in Old World monkeys , and second what can be the function –if any- , of this conserved gene in the other primates . The above fusion assays together with the conservation of the gene for million years of primate evolution strongly suggest that envV2 could be a syncytin gene , still active in the Old World monkey lineage . To test this hypothesis , the third canonical characteristic feature of syncytins was investigated , namely the restriction of its expression to the placenta , at the level of the specific cytotrophoblast/syncytiotrophoblast cells which contribute to the formation of the materno-fetal interface ( reviewed in [1] , [2] ) . First , as illustrated in Figure 6 , a quantitative RT-PCR for envV2 expression in a large panel of macaque tissues –including the placenta- clearly demonstrated placenta-specific expression , which was found to be at least 10-fold higher than in any other tissue tested . The pattern of expression of envV2 in the macaque is closely related to that of the orthologous gene in humans ( Figure 6 ) . The placenta of simians is of the hemochorial type and is characterized by the presence of fetal villi in direct contact with maternal blood [32]–[35] . The villi have an inner mononucleated cytotrophoblast layer underlying the surface syncytial layer , the syncytiotrophoblast ( Figure 7A ) . During gestation , the cytotrophoblast layer becomes discontinuous , whereas the syncytiotrophoblast remains continuous although it can develop regions of unequal thickness , with accumulation of nuclei in some areas . The macaque and human placentae are closely related , notably at the level of the structure of the placental villi [36]–[39] . To localize precisely envV2 expression , in situ hybridization on macaque placental serial sections was performed with specific digoxigenin-labeled antisense probes as well as with the corresponding sense probes as controls . As illustrated in Figure 7B , specific labeling was obtained with the antisense probe , and not with the control probes . In the macaque placenta , envV2 expression is localized at the level of the syncytiotrophoblast , at the materno-fetal interface , as expected for a gene involved in syncytiotrophoblast morphogenesis . Of note , envV2 expression is not detected at the level of the cytotrophoblasts which can be distinguished by their mitotic activity using an anti-Ki67 antibody [40] ( Figure 7B , right panels ) . Finally , the figure also shows a similar localization of envV2 transcripts in the human placenta . In conclusion , envV2 in the macaque Old World monkey possesses the expected features for a syncytin , and as such could now be named “mac-syncytin-3” to refer to the species and to indicate that it belongs to a family clearly distinct from the syncytin-1 and syncytin-2 families . As illustrated in Figure 8A , 8B , the envV2 gene is conserved among simians: comparison of the aa sequences from the 12 genes cloned and assayed for the fusogenicity of the corresponding proteins displays high sequence conservation ranging from 80 to 99% , as well as evidence for purifying selection for the entire env sequences , with low values for the non-synonymous to synonymous substitution ratios ( dN/dS ) between all pairs of species ( methods in [41] ) , ranging from 0 . 04 to 0 . 49 ( Figure 8B ) . This pattern of dN/dS ratio is classically observed for cellular genes with a physiological function , in which non-synonymous mutations are strongly selected against . We obtained a similar pattern for syncytin-2 whose entry into the primate lineage coincides with that of envV2 and that we analyzed for the same extended set of species , with dN/dS ratios ranging from 0 . 08 to 0 . 47 ( not shown ) . Of note , a similar analysis carried out on envV1 ( after elimination of the stop/frameshift mutations for some of the sequences ) provides related low dN/dS values , a paradoxical result for a pseudogene-like sequence that can be accounted for by the frequent conversion events taking place with envV2 , as shown in [27] . To characterize further the conservation and evolution of the envV2 gene and detect possible differences of selective pressure on distinct sites of the protein and/or between different branches among the gene phylogeny , we performed a more refined analysis using algorithms implemented in the PAML and HyPhy programs [42] . The site model analysis , allowing for the dN/dS ratio to vary between codons , provided support for a model ( Model M7 ) in which most of the codons are under purifying selection ( 0<dN/dS<0 . 84 , 80% of the sequence ) , while a few codons are evolving nearly neutrally ( 0 . 97<dN/dS<1 , 20% of the sequence ) . The alternative model allowing for some codons to be under positive selection ( dN/dS>1 , Model M8 ) did not fit better to the data ( Model M8 versus M7 , degree of freedom = 2 , chi2 = 1 . 5 , p-value = 0 . 47 ) . The distribution of dN/dS values is homogenous among codons , with no particular domain emerging . Analysis using the HyPhy package [43] , with slightly different site-specific models ( i . e . Random Effect Likelihood or Fixed Effect Likelihood ) , leads to similar conclusions . Finally , a branch model analysis using the HyPhy package ( GA-Branch analysis ) , allowing for the dN/dS ratio to vary between branches , does not provide support for some phylogenetic branches to be under positive selection nor for significant differences in strength of selection between hominoids , Old World monkeys and New World monkeys ( Figure 9 ) . A similar analysis using the PAML package provides support for a model in which all branches are under strong purifying selection , with dN/dS = 0 . 36 . In conclusion , no specificity for the evolution of the Old World monkey lineage can be unraveled from the sequence analyses , with the envV2 genes from all species being under strong purifying selection .
The envV2 gene entered the simian lineage >45 Mya together with syncytin-2 , and was previously demonstrated to be conserved as a full-length envelope protein encoding gene and to be placenta-specific but not fusogenic in humans [8] , [24] , [27] . We show here that envV2 has retained fusogenic activity in Old World as well as in some New World monkey species . Furthermore , in the macaque , expression of the fusogenic envV2 gene is placenta-specific , and in situ hybridization demonstrates that it is expressed at the materno-fetal interface , as expected for a syncytin gene . Accordingly , it can be proposed that envV2 was primitively captured and domesticated as a syncytin gene , prior to the simian radiation , and that its fusogenic activity was lost on several occasions , firstly before emergence of the hominoid branch , approx . 30 Mya , and then in several New World monkey species , except Callithrix jacchus . It shows that the fusogenic function of envV2 is not uniformly subjected to selection , which could indicate either i ) progressive degradation of this primitive syncytin gene , or ii ) selection for another function of the encoded protein . The progressive decay of the gene could be accounted for by the presence of other syncytin genes captured in the course of primate evolution: syncytin-2 is present in all simians where it has retained fusogenic activity [8] , and its functions might therefore have substituted for part or all of those of envV2 in the course of evolution; syncytin-1 entered the primate lineage later but retained fusogenic activity in all the hominoids , whereas it was lost in Old World monkeys [5]–[7] , i . e . in a mirror-like manner as compared to envV2 evolution –thus consistent with a possible functional complementation between the two env genes . Another possibility , which would be consistent with the conservation of the full-length envV2 ORF with low dN/dS ratios , the demonstrated branch-independent purifying selection , and the sustained placenta-specific expression , could be that fusogenic activity is not the property for which this gene has been selected for , and that there is another EnvV2 function that is being selected . We propose that this function is linked to the presence of an immunosuppressive domain ( ISD ) on the EnvV2 protein , a domain that we have previously unambiguously demonstrated to be functional [19] . As illustrated in Figure 8A , this sequence , which corresponds to a highly structured polypeptide domain , is the longest unaltered sequence among the 12 EnvV2 aligned sequences , with a stretch of more than 30 amino acids without a single mutation . This domain , also present in the envelope protein of infectious retroviruses and most endogenous retroviruses ( with the significant exception of Syncytin-1 in primates and Syncytin-A in muroids ) , carries an immunosuppressive function that can inhibit both the humoral and cellular immune response , and which has been proposed to participate –among other factors- in the establishment of materno-fetal tolerance . Along this line , conservation of the canonical ISD sequence among the EnvV2 orthologs would be the driving force for the observed conservation of the EnvV2 sequence in evolution , despite the presently demonstrated variable selective pressure on its fusogenic activity . Another possible role that could have been selected for is protection against infection by the so-called process of receptor interference , with the well-known examples of the murine Fv4 env-like gene , and of the enJSRV ovine genes ( reviewed in [44] ) . Yet , we do not favour such a role for envV2 whose expression is severely restricted to the placenta , at variance with the Fv4 and enJSRV genes which disclose a broader expression pattern as would be expected for an efficient protective effect against different routes of infection . In addition , there is no known infectious retrovirus that presently possesses an env gene closely related to envV2 . Whatever the interpretation , it clearly emerges that captured syncytins are prone to disruption or complete deletion and , in this respect , it is interesting to note that EnvR , also an ancestrally captured retroviral env gene , which lost its fusogenic activity very early in the course of primate evolution ( with a stop codon just upstream of the transmembrane domain of the TM subunit conserved among all simians ) , retained immunosuppressive activity and high-level expression in the placenta , but became dispensable , for instance in the gorilla where the entire gene is deleted , and in 1% of the human population where the gene is interrupted in a homozygous state by a premature stop codon [26] , [30] . A model in which syncytin genes can be acquired and lost in the course of evolution is therefore very likely , and would be further consistent with the rather paradoxical situation whereby the acquisition of all the presently discovered syncytins has been dated between 10–80 My depending on the species , which is compatible with a fundamental role of these genes in placental mammal evolution only if one assumes that primitive syncytins were present earlier and have been progressively “replaced” by the stochastically acquired new syncytins that are presently found to be active .
This study was carried out in strict accordance with the French and European laws and regulations regarding Animal Experimentation ( Directive 86/609/EEC regarding the protection of animals used for experimental and other scientific purposes ) . The macaque placenta tissues used were obtained in accordance with the animal protocol approved by the Institutional Animal Care and Use Committee at the Commissariat à l'Energie Atomique ( CEA ) . First trimester human placenta tissues were obtained from legal induced abortions and term placenta tissues after cesarean sections , with parent's written informed consent . The source of human genomic DNA is given in [45] . Chimpanzee ( Pan troglodytes ) , gorilla ( Gorilla gorilla ) , orangutan ( Pongo abelii ) , gibbon ( Hylobates ) , and cotton-top tamarin ( Saguinus oedipus ) genomic DNAs were obtained from ECACC and rhesus macaque ( Macaca mulatta ) from Zyagen . The baboon ( Papio papio ) DNA was a gift from Guy Dubreuil ( Station de Primatologie , Rousset , France ) , the African green monkey ( Chlorocebus aethiops ) and marmoset ( Callithrix jacchus ) DNAs from Helene Gachot ( Institut Pluridisciplinaire Hubert Curien , Strasbourg , France ) and the Hanuman gray langur ( Semnopithecus entellus ) and white-faced saki ( Pithecia pithecia ) DNAs were extracted from blood samples provided by Baptiste Mulot ( Zooparc de Beauval , Saint-Aignan , France ) . Human first trimester placentas were obtained from legal induced abortions ( 8–12 weeks of gestation ) and term placentas after cesarean section from healthy mothers near term with uncomplicated pregnancies from the Department of Obstetrics and Gynecology at the Saint-Vincent-de-Paul and Cochin Hospitals , Paris , France . Macaque term placentas were obtained from Guy Germain ( CEA , Fontenay-aux-roses , France ) after cesarean section on pregnant females ( Macaca cynomolgus ) at day 152 of gestation . The orthologs of the env genes from the human endogenous retroviruses ERV-V1 and -V2 were PCR amplified from primate genomic DNAs . PCRs were carried out for 25 cycles ( 30 s at 94°C , 30 s at 56°C , and 3 min 30 s at 68°C ) , with sets of appropriate primers ( see Table S1 ) in 50 µl containing 300 ng of genomic DNA , 1× AccuPrime PCR buffer II and 2 U of AccuPrime Taq DNA polymerase ( Life technologies ) . XhoI ( or BamHI ) -containing primers were used as forward primers and MluI-containing primers as reverse primers ( Table S1 ) . Each PCR product was then XhoI ( or BamHI ) -MluI restricted and cloned into the phCMV-G vector opened with the same enzymes . Sequencing of each of the primate env genes was performed directly on the PCR products ( Applied Biosystem sequencing kit ) . For C-terminal tagging of the envelope proteins with the hemagglutinin ( HA ) epitope , primate envV2-amplified fragments were generated with appropriate primers ( Table S1 ) from the phCMV-envV2 vectors described above . Each PCR product was restricted with AgeI-XbaI and inserted in frame into a HA-containing pCMV4 plasmid ( Gift of M . Malim , King's College , London , UK ) opened with the BspEI-XbaI enzymes . The EnvV2-HA fragments were then restricted ( SnaBI-EcoRV for human and macaque , SnaBI-EcoNI for marmoset and cotton-top tamarin and SnaBI-BglII for saki ) and re-inserted into the phCMV vector . The 293T human embryonic kidney cells ( ATCC CRL11268 ) , the A23 hamster fibroblasts [46] and the G355 . 5 feline astrocyte cells ( ATCC CRL2033 ) were grown in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% fetal calf serum ( Life Technologies ) , 100 µg/ml streptomycin and 100 U/ml penicillin . For cell-cell fusion assays , 5×104 to 1×105 cells seeded in 24-well plates were transfected by using lipofectamine LTX ( Life technologies ) with 250 ng of env expression plasmid . Fusion activity of each envelope protein was visualized 24 to 48 h after transfection with the corresponding vectors by May-Grünwald and Giemsa staining , according to the manufacturer's instructions ( Sigma ) . The fusion index , which represents the percentage of fusion events in a cell population , is defined as [ ( N – S ) /T ]×100 , where N is the number of nuclei in the syncytia , S is the number of syncytia , and T is the total number of nuclei counted . Cell surface biotinylation assays were performed 48 h post-transfection . Cells were chilled on ice , washed twice with phosphate-buffered saline ( PBS ) supplemented with 0 . 7 mM CaCl2 and 0 . 25 mM MgSO4 ( PBS++ ) for 30 min and then incubated with 0 . 75 mg/ml of sulfo-N- hydroxysuccinimide-biotin ( Thermo Scientific ) for 30 min on ice . Biotinylation was stopped by incubating the cells with 1 M glycine in PBS++ for 5 min at 4°C . The cells were washed three times with PBS/100 mM glycine , pH 7 . 4 and then lysed for 30 min on ice with PBS/100 mM glycine , supplemented with 1% Triton X-100 and a protease inhibitor cocktail ( cOmplete ULTRA tablets , Mini Easypack , Roche ) . Cell lysates were incubated with streptavidin-coated magnetic beads ( Dynabeads Streptavidin M-280 , Life Technologies ) , previously washed three times with PBS/100 mM glycine , for 30 min at 4°C with gentle agitation . Biotinylated cell surface proteins immobilized on streptavidin beads were pelleted by magnetic separation and intracellular unmodified ( non-biotinylated ) proteins were collected from the supernatant . The streptavidin beads were then washed four times with PBS/100 mM glycine and the bead-associated proteins together with the intracellular proteins present in the supernatant were examined by Western blot analysis . For Western blot analyses , biotinylated cell surface protein fractions were separated by SDS/PAGE on gradient precast gels ( NuPAGE Novex 4–12% Bis-Tris gels , Life Technologies ) and transferred onto a nitrocellulose membrane using a semi-dry transfer system . After blocking in PBS containing 0 . 2% Tween-20 and 5% nonfat milk , membranes were incubated 1 h at room temperature ( RT ) with primary antibodies , washed and then incubated with species-appropriate horseradish peroxidase ( HRP ) -conjugated secondary antibodies for 45 min at RT . The Super Signal West Pico Chemiluminescent substrate ( Thermo Scientifics ) was used to reveal proteins . HA-tagged EnvV2 proteins were detected using a rat anti-HA monoclonal antibody ( 3F10 , Roche ) and a goat HRP-conjugated anti-rat IgG secondary antibody ( AbD Serotec ) . The antibody used to detect the cellular glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) was a goat polyclonal HRP-conjugated anti-GAPDH ( Antibodies-online ) . Human and macaque total RNAs were extracted from frozen organs with the RNeasy RNA isolation kit ( Qiagen ) or obtained from Zyagen . Reverse transcription was performed with 1 µg DNase-treated RNA as in [22] . Real-time quantitative PCR was carried out with 5 µL diluted cDNA ( 1/25 ) in a final volume of 25 µL using the SYBR Green PCR Master Mix and the ABI PRISM 7000 sequence detection system ( Applied Biosystems ) . To normalize the transcript levels , PCR amplifications using primers for the peptidylpropyl isomerase A ( PPIA ) gene mRNA encoding cyclophilin A were performed as an internal control . Freshly collected placentas were fixed in 4% paraformaldehyde at 4°C and embedded in paraffin . Serial sections ( 7 µm ) were either stained with the anti-Ki67 antibody by immunohistochemistry or used for in situ hybridization . For human and macaque envV2 mRNA detection , three non-overlapping PCR-amplified envV2 fragments of: 1 ) 472 bp ( human ) or 478 bp ( macaque ) , 2 ) 324 bp and 3 ) 309 bp were cloned in both orientations into the pGEM-Teasy vector ( Promega ) to be used as templates for in vitro synthesis of the antisense and control sense riboprobes in the presence of T7 or SP6 RNA polymerase and digoxygenin-11-UTP ( Roche Applied Science ) , following the manufacturer's instructions . Paraffin-embedded placenta tissue sections were processed , hybridized at 42°C overnight with each set of pooled riboprobes ( human or macaque , antisense or sense ) and incubated further at room temperature for 2 h with alkaline phosphatase-conjugated anti-digoxigenin antibody Fab fragments ( Roche Applied Science ) . Staining was carried out with nitroblue tetrazolium ( NBT ) and 5-bromo-4-chloro-3-indolyl phosphate ( BCIP ) alkaline phosphate substrates , as recommended by the manufacturer ( Roche Applied Science ) . For Ki67 immunohistochemical staining , paraffin sections were processed for heat-induced antigen retrieval , incubated with a rabbit monoclonal anti-Ki67 antibody ( Neomarkers , LabVision ) , stained by using the peroxidase/diaminobenzidine Rabbit PowerVision kit ( ImmunoVision ) and counterstained with Mayer's hemalum . Multiple alignments of sequences were carried out by using the Seaview program [47] under ClustalW protocol . Maximum likelihood trees were constructed with RAxML 7 . 3 . 2 [48] with bootstrap percentage ( BP ) computed after 1000 replicates , using the global time reversible ( GTR ) model with gamma distribution ( GTR + GAMMA ) and the rapid bootstrapping algorithm . The phylogenetic analysis by maximum likelihood ( PAML4 ) package [42] was used to run site-specific or branch-specific selection tests , and to obtain the non-synonymous vs synonymous ( dN/dS ) substitution rate ratios . We used likelihood ratio tests to compare the improvement in likelihood between the different models . The site-specific and branch-specific models analyzed assumed no molecular clock ( clock = 0 ) . Site-specific models take into account a single dN/dS for all tree branches ( model = 0 ) and a beta distribution of codons among site classes ( models M7 and M8; NS site = 7 8 ) . Branch-specific models take into account a single dN/dS for all codons within a phylogenetic group ( NS site = 0 ) , and different dN/dS values for pre-defined phylogenetic groups ( model = 2 ) . Each analysis ran until convergence ( Small_Diff = 0 . 5e-6 ) , and the control file is available on request . The HyPhy software package [43] was implemented on the datamonkey webserver ( www . datamonkey . org ) for selection analysis using two different selection tests: i ) the site-specific Random Effect Likelihood ( REL ) and Fixed Effect Likelihood ( FEL ) tests , and ii ) branch-specific genetic algorithm ( GA ) -branch analyses . The best branch-specific model was selected using the Akaike Information Criterion . | Syncytins are “new” genes encoding the envelope protein of captured endogenous retroviral elements . Their unambiguous status of “cellular gene” was recently demonstrated by knocking them out in genetically modified mice , showing their absolute requirement for placenta formation and embryo survival , via formation by cell–cell fusion of the feto-maternal syncytium interface . These genes are remarkable , as they are “necessary” for a basic function in placental mammals and yet they were acquired “by chance” on multiple occasions and independently in diverse mammalian species . We proposed that syncytins have been pivotal for the emergence of animals with a placenta from those laying eggs via the capture of a founding retroviral env gene , then subsequently replaced in the diverse mammalian lineages upon successive and independent germline infections by new retroviruses and co-optation of their env gene , each new gene providing its host with a positive selective advantage . This hypothesis would account for the diversity of the captured syncytins that can be currently found , concomitant with the diversity of placental architectures . A consequence of this paradigm is that evidence for “decaying syncytins” in eutherian mammals should exist , and this is precisely what we sought—and found—in this study . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"molecular",
"cell",
"biology",
"virology",
"genetics",
"biology",
"genomics",
"evolutionary",
"biology",
"microbiology",
"genetics",
"and",
"genomics"
] | 2013 | Differential Evolutionary Fate of an Ancestral Primate Endogenous Retrovirus Envelope Gene, the EnvV Syncytin, Captured for a Function in Placentation |
The accurate segregation of chromosomes during cell division is achieved by attachment of chromosomes to the mitotic spindle via the kinetochore , a large multi-protein complex that assembles on centromeres . The budding yeast kinetochore comprises more than 60 different proteins . Although the structure and function of many of these proteins has been investigated , we have little understanding of the steady state regulation of kinetochores . The primary model of kinetochore homeostasis suggests that kinetochores assemble hierarchically from the centromeric DNA via the inclusion of a centromere-specific histone into chromatin . We tested this model by trying to perturb kinetochore protein levels by overexpressing an outer kinetochore gene , MTW1 . This increase in protein failed to change protein recruitment , consistent with the hierarchical assembly model . However , we find that deletion of Psh1 , a key ubiquitin ligase that is known to restrict inner kinetochore protein loading , does not increase levels of outer kinetochore proteins , thus breaking the normal kinetochore stoichiometry . This perturbation leads to chromosome segregation defects , which can be partially suppressed by mutation of Ubr2 , a second ubiquitin ligase that normally restricts protein levels at the outer kinetochore . Together these data show that Psh1 and Ubr2 synergistically control the amount of proteins at the kinetochore .
Accurate chromosome segregation is necessary for the equal distribution of genetic material between daughter cells during cell division and is achieved by kinetochores which link chromosomes to spindle microtubules [1] . Perturbations of kinetochore function result in aneuploidy , i . e . changes in chromosome number , and genome instability [2 , 3] . Thus kinetochore regulation is of critical importance in replicating cells . A number of different cancers overexpress kinetochore genes [4 , 5] leading to the notion that disrupting kinetochore stoichiometry and regulation may be a driver of aneuploidy and genomic instability . Budding yeast is a key model to study kinetochore composition and assembly because of its comparatively simple structure; there is only one microtubule attachment per chromosome and per kinetochore [6 , 7] . Kinetochores are composed of more than 60 proteins organized into various sub-complexes that are thought to assemble hierarchically initiating at the centromeres [1] . The inner part of the kinetochore mediates centromere binding whereas the outer part mediates microtubule binding . Kinetochore structure and composition is remarkably well conserved from yeast to humans [8] . In budding yeast the position of the centromeres is sequence specific . Cbf1 and the CBF3 complex associate to centromere DNA elements ( CDE ) , CDEI and CDEIII , respectively [9–13] . The CDEII region wraps around the centromeric nucleosome that contains the centromeric histone H3 variant CENP-A ( Cse4 in budding yeast ) [14–17] . Mif2 ( CENP-C ) and the COMA complex mediate the association between centromere and outer kinetochore . Mif2 binds to both the Cse4 nucleosome and the outer kinetochore MIND complex [18–20] . The COMA complex proteins Okp1 and Ame1 form a dimer that binds directly to DNA and the MIND complex [20 , 21] . The outer kinetochore mediates interactions with microtubules emanating from opposite spindle pole bodies . The yeast homologues of the KNL1/ MIS12/ NDC80 network ( KNM ) are the essential complexes SPC105 , MIND and NDC80 , respectively [1] . The MIND complex is composed of two heterodimers: Mtw1-Nnf1 , which associates with both Mif2 and the COMA complex , and Dsn1-Nsl1 , which associates with the NDC80 complex [21 , 22] . Both the NDC80 complex and the yeast-specific DAM-DASH complex , which may play an orthologous function to the human SKA proteins [23] , bind to microtubules in a cooperative process [24 , 25] . Although the centromeric DNA sequence ( CEN ) is essential to assemble kinetochores , protein degradation has been shown to be important to control cellular levels of various kinetochore proteins . The E3 ubiquitin ligase Psh1 restricts the localization of Cse4 to centromeres [26] . Psh1 localizes to centromeres throughout the cell cycle , and its destabilizing role is opposed by the Cse4 chaperone Scm3 [27 , 28] . Levels of Cse4 are increased in psh1Δ cells [26] and these cells have a chromosomal instability phenotype [29] . More recently , the E3 ubiquitin ligase Ubr2 has been shown to control levels of the MIND complex protein Dsn1 [30] . Thus kinetochore assembly may be regulated differently from steady state homeostasis . Surprisingly , yeast kinetochores can assemble in reverse from the microtubule interface back to the inner kinetochore as shown via artificial recruitment of proteins to DNA [31] . In this situation , the conserved yeast centromere is not necessary , although inner kinetochore proteins are required [32] . These data point to a kinetochore with more flexibility in its assembly and stoichiometry than was previously assumed . Numerous studies in budding yeast have revealed the stoichiometry of the various protein sub-complexes forming the kinetochore [20 , 21 , 33–37] . It is thought that the kinetochore assembles hierarchically from the centromere [37] . However , little is known about how these sub-complexes assemble to form the kinetochore in vivo and how much flexibility exists in kinetochore composition . To investigate this , we tested how increased levels of kinetochore proteins affect kinetochore composition . We used fluorescence microscopy to quantify the levels of proteins at kinetochore foci . We found that Mtw1 levels at the kinetochore correlate with chromosome number and they are not transcriptionally controlled . Moreover , we found that psh1Δ mutants , in addition to the elevated Cse4 protein , have increased levels of inner kinetochore proteins but not outer kinetochore proteins . However , the levels of outer kinetochore proteins are increased in the psh1Δ ubr2Δ double mutant , in which both Cse4 and Dsn1 are unconstrained . Finally , we found that ubr2Δ suppresses psh1Δ mitotic and meiotic defects . These findings are consistent with multiple regulatory pathways acting independently on the different kinetochore complexes .
To investigate whether we could perturb kinetochore homeostasis by overexpression of kinetochore genes , we chose to study MTW1 . Mtw1 forms part of the essential MIND complex [21 , 38] and the levels of one of these proteins , Dsn1 , is controlled via phosphorylation status and subsequent ubiquitylation by the E3 ligase , Ubr2 [30] . We used an ectopically-expressed plasmid-encoded version of Mtw1 to elevate the levels of Mtw1 within the cell and assessed the recruitment of Mtw1 to kinetochores by fluorescence imaging . The plasmid is a single copy CEN plasmid and its MTW1 gene is driven by a constitutively-active copper promoter ( CUP1 ) [39] . We used differential fluorescence tagging of endogenously-encoded and plasmid-encoded Mtw1 to differentiate between and quantitate the proteins loaded into kinetochores ( Fig 1A , 1B and 1C ) . The MTW1 plasmid produced significant ectopic expression as judged by loading of plasmid-encoded Mtw1 at the kinetochore ( Fig 1A ) . We quantified the levels of fluorescence at kinetochores using Volocity image analysis software . In brief , the mean fluorescence within a 3-dimensional spherical region around each kinetochore was assessed and a background region around each kinetochore was also measured by dilating each kinetochore selection ( Fig 1E ) . Each background measurement was subtracted from each kinetochore measurement to produce a relative value representing the levels of fluorescence signal from the kinetochore . When we expressed an ectopic MTW1-CFP gene in cells containing MTW1-YFP at the endogenous locus , we found that the resulting fluorescence at kinetochores was approximately 50% of the haploid CFP signal and 50% of the haploid YFP signal ( Fig 1B ) . This is consistent with an approximately equal contribution of the two proteins to the kinetochore , but not consistent with an elevation of Mtw1 loading at the kinetochore . To determine whether one fluorescent tag is preferred over the other , we then performed the same analysis but with the tags reversed i . e . ectopic MTW1-YFP and endogenous MTW1-CFP . In this case the levels of the plasmid encoded Mtw1-YFP at the kinetochore are somewhat higher than the CFP signal , although both still contribute to the kinetochore signal ( Fig 1B ) . Again , no increase in total kinetochore fluorescence was measured . We also examined the effect of deleting the endogenous MTW1 gene in cells containing an MTW1-YFP plasmid . The level of YFP fluorescence in this stain is the same as an endogenously-encoded MTW1-YFP strain , ( Fig 1B ) . Finally , we transformed the MTW1-YFP plasmid into an untagged strain . We find that the Mtw1-YFP level of fluorescence is equivalent to the strain with both endogenously and ectopically-encoded Mtw1 , approximately 50% ( Fig 1B ) . We also assessed whether changes in the background levels of fluorescence in the cells over-expressing kinetochore proteins were increased , resulting in an artificially low kinetochore signal . However , we find that changes to background fluorescence do not mask an effect of MTW1 expression on kinetochore protein levels ( S1A and S1B Fig ) . Thus , these quantitative data support the notion that the fluorescently tagged proteins compete for inclusion into the kinetochore and that the total levels of kinetochore Mtw1 remain constant . There are two likely reasons for this homeostasis of Mtw1 at the kinetochore . First , an uncharacterised negative feedback mechanism could limit transcription , translation or protein stability of the endogenous Mtw1 , thus maintaining a steady state level of Mtw1 protein within the cell . Second , the loading of Mtw1 onto the kinetochores is limiting , such that there is a strong affinity to load Mtw1 as part of the MIND complex but once the protein reaches a threshold level ( perhaps through stoichiometric interaction with other kinetochore components ) , no more Mtw1 is loaded . To discriminate between these two ideas we used western blotting to assess the total cellular levels of Mtw1 . We find that the ectopic expression of MTW1 causes an increase in the levels of Mtw1 protein in the cell ( Fig 1F ) . Thus , we exclude the possibility that total Mtw1 protein levels are tightly regulated by translation or protein stability . Our results are also consistent with the notion of hierarchical assembly of the kinetochore building up from inner kinetochore components such as Cse4 . To test this notion we compared the loading of Mtw1 in diploid strains with MTW1-YFP at either one or two of the endogenous MTW1 alleles . We find that diploid kinetochore Mtw1 levels are approximately double that of haploids and heterozygous mtw1Δ/MTW1-YFP strains compensate by loading equivalent Mtw1 as diploid strains ( Fig 1D ) . We note here that these heterozygous mtw1Δ/MTW1-YFP strains are haplo-sufficient in that they do not show sensitivity to microtubule poison drug benomyl ( S2B Fig ) . We also confirmed that overexpression of MTW1 does not render cells sensitive to benomyl ( S2C Fig ) , nor does it affect cell cycle progression ( S3A Fig ) , plasmid loss ( S3B Fig ) , or chromosome segregation ( S3C and S3D Fig ) . We also checked whether MTW1 overexpression resulted in changes to the levels of other kinetochore proteins and consistent with the levels of Mtw1 , we find no change in Dsn1 or Ndc80 ( S3E and S3F Fig ) . In order to test more generally the effects of high levels of kinetochore proteins , we expressed various inner and outer kinetochore proteins from a CEN plasmid under the control of a CUP1 promoter . Only NDC10 overexpression showed a reduced growth in the presence of benomyl ( S4 Fig ) We then tested whether Mtw1 kinetochore levels were affected by the deletion of genes encoding several inner kinetochore components: the DNA-binding protein Cbf1 , the Monopolin complex components Mam1 and Csm1 , and the COMA complex component Ctf19 . We found no change in Mtw1 levels in any of these mutants ( S5A and S5B Fig ) , consistent with Mtw1 loading hierarchically based upon the number of centromeres present in the cell . The hierarchical loading model is consistent with the hypothesis that the loading of inner kinetochore proteins is critical for determining kinetochore stoichiometry as a whole . To test this idea we decided to attempt to manipulate the levels of an inner kinetochore protein to test whether the MIND complex is regulated in parallel . The levels of the inner kinetochore protein Cse4 are controlled in part by degradation via an ubiquitylation-dependent degradation pathway . Psh1 was identified as the E3 ubiquitin ligase responsible for restricting Cse4 levels at the kinetochore [26 , 27] . In a psh1Δ strain Cse4 levels are elevated and furthermore overexpression of the CSE4 is lethal in psh1Δ cells , consistent with a failure to constrain Cse4 loading [26 , 27] . We used the same fluorescence quantitation method described above to compare endogenous kinetochore protein levels of wild-type cells with those of psh1Δ cells . Consistent with previous studies we find that psh1Δ cells have elevated levels of Cse4 at kinetochore foci , although with considerable heterogeneity between cells ( Fig 2A ) . We found no change in the protein levels of the inner kinetochore protein Ndc10 ( Fig 2B ) . In addition , we find that Mif2 , the ortholog of human CENP-C , ( Fig 2C ) and members of the Ctf19/COMA complex are also elevated in the psh1Δ ( Fig 2D , 2E and 2F ) . However , contrary to our expectation Mtw1 kinetochore levels are unchanged in a psh1Δ strain compared with wild type ( Fig 2G ) . We therefore examined whether other outer-kinetochore complexes are affected by deletion of PSH1 . Like Mtw1 , the kinetochore levels of Ndc80 and Ask1 ( a member of the decameric DAM1/DASH complex ) are both unaffected in psh1Δ cells ( Fig 2H and 2I ) . These data show that although Cse4 levels may influence the inner kinetochore , the protein levels of the entire kinetochore are not affected . This result shows that for the fluorescence focus that is widely considered to represent the structural kinetochore the stoichiometry is not fixed . One possible reason for the non-stoichiometric increase in kinetochore protein levels in psh1Δ cells is that the increased Cse4 , Ctf19 etc . are not part of the canonical kinetochore structure , but rather represent a pericentromeric ‘cloud’ of protein . There is precedent for this from fluorescence studies of Cse4 [40 , 41] . We therefore re-analysed our images to evaluate the size each of the fluorescence foci . The rationale is that pericentric protein recruitment will result in a larger area of fluorescence , which can be measured by fitting a Gaussian distribution to the kinetochore foci ( Fig 3A ) . We find that psh1Δ Cse4 foci are considerably larger than WT , consistent with the notion of a cloud of pericentric Cse4 and this is rescued by overexpressing PSH1 ( Fig 3B and 3C ) . However , the other kinetochore proteins had psh1Δ foci comparable in size to WT cells ( Fig 3C–3K ) . We cannot say for sure that protein that is located in a comparably-sized focus is part of a structural complex , it is possible that for certain proteins the kinetochore can accommodate additional proteins within the confines of the WT diffraction limited region . We next asked whether the effect of Psh1 upon kinetochore protein levels would function in synergy with the Mub1/Ubr2 ubiquitylation pathway . The MIND complex member Dsn1 is ubiquitylated by the E3 ubiquitin ligase Ubr2 [30] . Dsn1 contains two AuroraB ( Ipl1 ) phosphorylation sites ( serines 240 and 250 ) and versions of Dsn1 that cannot be phosphorylated at these residues are ubiquitylated and degraded [30 , 42] . Such a mechanism may restrict the levels of MIND proteins even in the presence of excess inner kinetochore proteins . Since psh1Δ , ubr2Δ and the double mutant cells are all viable we were able to assess their relative contribution to the kinetochore focus fluorescence levels . We find that UBR2 deletion has no effect upon inner kinetochore protein levels of Cse4 or Ndc10 . Cse4 levels are elevated by PSH1 deletion , but not further affected by the additional deletion of UBR2 ( Fig 4A ) . Also addition of ubr2Δ mutation did not further increase the size of Cse4-GFP foci ( S6A Fig ) . Ndc10 is unaffected by either of these mutants ( Fig 4B ) . Mif2 is elevated in a psh1Δ mutant , but unaffected by further deletion of UBR2 ( Fig 4C ) . The MIND complex shows little change in either of the single mutants but both Mtw1 and Dsn1 are modestly elevated in the double psh1Δ ubr2Δ strain ( Fig 4D and 4E ) . The size of Mif2 and Dsn1 foci was unaffected in the ubr2Δ and in the double psh1Δ ubr2Δ cells ( S6B and S6C Fig ) . Another MIND complex protein Nnf1 is also elevated in psh1Δ ubr2Δ cells ( Fig 4F ) . Other outer kinetochore proteins Spc105 , Spc24 , from NDC80 complex , and Ask1 were unaffected by either of these mutants ( Fig 4G , 4H and 4I ) . The degradation of Dsn1 is controlled by phosphorylation/ dephosphorylation of serines 240 and 250 . The double dsn1-S240A , S250A mutant is inviable , but can be rescued by either its overexpression or by deleting UBR2 [30] . We reasoned that if increased Dsn1 was responsible for the MIND phenotype , this should be epistatic with a dsn1-S240D , S250D mutant , which would be hyper-stable . However , we find that the elevated levels of Mtw1 in a psh1Δ ubr2Δ mutant are increased further when the two Dsn1 serines are changed to aspartic acid ( Fig 5A and 5B ) . Furthermore , we examined cellular levels of both Mtw1 and Dsn1 in psh1Δ , ubr2Δ and the psh1Δ ubr2Δ mutants and find that these are comparable with wild-type cells ( S6D and S6E Fig ) These data suggest that Ubr2 plays additional , potentially indirect , roles in regulating the levels of kinetochore components in addition to its function on dephosphorylated Dsn1 or that there are other mechanisms to remove dephosphorylated Dsn1 from kinetochores . These data also strengthen our observation that the stoichiometry of the various kinetochore sub-complexes is not fixed in these mutants . Although these ubiquitin ligase mutants affect kinetochore protein levels , they are all viable and the cells appear to grow normally [26 , 30] . Since there is considerable interest in the possibility that altered kinetochore protein levels would lead to kinetochore dysfunction and the resulting aneuploidy [4 , 5 , 43] , we asked whether the psh1Δ and ubr2Δ mutants affected the mitotic or meiotic phenotype of yeast . We did not find strong defects in cell cycle progression , although S-phase was slightly faster in ubr2Δ and psh1Δ ubr2Δ mutants ( S7 Fig ) . It has previously been reported that ubr2Δ mutants have an enhanced sporulation phenotype [44] . Consistent with this we found that the sporulation of homozygous ubr2Δ mutants is enhanced compared with wild-type diploids ( Fig 6A ) . Addition of the psh1Δ mutant did not modify this phenotype . In all cases spore viability was similar ( Fig 6B ) . We tested whether the increase in Mtw1 kinetochore levels in psh1Δ ubr2Δ mitotic cells ( Fig 3D ) was recapitulated in meiosis . Diploid cells were induced to sporulate and arrested in pachytene , prior to the two meiotic divisions by depletion of the Ndt80 transcription factor . Then , meiosis I was triggered by induction of NDT80 expression from the GAL1-10 promoter [45] ( see Materials and Methods for details ) . We found elevated Mtw1 kinetochore levels in psh1Δ ubr2Δ in meiosis I , and to a lesser extent in meiosis II ( Fig 6C and 6D ) . As Psh1 is known to have a role in maintaining chromosome stability [29] , we used an assay for homozygosity of chromosome III [2 , 3 , 29] to analyse the rate of chromosomal instability ( CIN ) in diploids cells , and we also tested the rate of loss of a CEN plasmid . Consistent with previous reports , we find that psh1Δ cells show elevated rates of both chromosome III loss ( Fig 7A ) and CEN plasmid loss ( Fig 7B ) , whereas ubr2Δ cells are unaffected . Surprisingly , we found that the addition of ubr2Δ to a psh1Δ mutant leads to a reduction of these CIN phenotypes ( Fig 7A and 7B ) . To investigate the effect of the ubiquitin ligases Psh1 and Ubr2 on checkpoint function , we assessed the synthetic effects of combining mutations in these genes with those of checkpoint genes . We deleted the MAD1 gene , which encodes a protein required for the activation of Mad2 [46] and also MAD3 , which encodes a key member of the mitotic checkpoint complex [47] . These mutants were combined with psh1Δ , ubr2Δ or the double mutant . The resulting strains were all viable ( Fig 8 ) , so to test their checkpoint proficiency we grew them in the microtubule poison benomyl . We found that deletion of psh1Δ decreases the ability of both mad1Δ and mad3Δ to grow in the presences of benomyl ( Fig 8 ) . Moreover , deletion of ubr2Δ partially rescued the ability of mad1Δ and mad3Δ to grow on benomyl . Finally , we also found that ubr2Δ partially rescues the benomyl sensitivity of mad1Δ psh1Δ and mad3Δ psh1Δ double mutants ( Fig 8 ) . We then tested if increased Dsn1 levels could explain the rescue of ubr2Δ . However , we found that DSN1 over-expression from a CUP1 promoter did not rescue benomyl sensitivity ( S8 Fig )
A number of studies have shown correlation between the overexpression of kinetochore genes and tumorigenic status [4 , 5 , 43] . These observations raise the possibility that increased levels of kinetochore proteins result in aberrant kinetochore function , which then leads to chromosomal instability . We wished to test the idea that overexpression of kinetochore genes would affect kinetochore protein loading . We overexpressed the kinetochore gene , MTW1 that encodes a core member of the outer kinetochore MIND complex . The MIND complex plays an essential role in linking the inner kinetochore and the outer kinetochore [48 , 49] . Using quantitative fluorescence imaging we find that although overexpression of MTW1 does lead to increased Mtw1 protein in the cell , the loading of Mtw1 onto the kinetochores is unaffected ( Fig 1 ) . Our data supports the idea that kinetochores are assembled hierarchically from the inner kinetochore , likely directed by Cse4 inclusion into centromeric nucleosomes [37] . Similarly , Aravamudhan and colleagues found that the levels of Cse4 at the kinetochore did not change after increasing total Cse4 cellular levels in budding yeast [50] . The effects of kinetochore gene overexpression may be subtle and/or different in mammalian cells , however , our data do not support the idea that kinetochore gene overexpression would , a priori , lead to a kinetochore defect ( Figs 1 , S2–S4 ) . On the contrary , our data also support the idea that the kinetochore focus represents the structural assembly of kinetochore proteins loaded onto centromeres [37 , 51] and that kinetochore protein levels scale with centromere number ( Fig 1 ) [52] . However , recent work using synthetic kinetochores has demonstrated that a functional kinetochore can assemble backwards from the microtubule associated DAM1/DASH complex [31 , 32] . Recruitment of outer kinetochore proteins to a non-centromere sequence is sufficient to generate an artificial kinetochore that no longer requires a specific CEN sequence but does require inner kinetochore proteins . These observations challenge the hierarchical assembly model , albeit in an artificially tethered system and suggest that the kinetochore structure may be more adaptable than previously imagined . In an effort to perturb the kinetochore structure we examined kinetochores in mutants of two ubiquitin ligases that are known to affect the degradation of kinetochore proteins , Psh1 and Ubr2 . The Psh1 ubiquitin ligase regulates the levels of Cse4 protein at the kinetochore focus [26 , 27] . We confirmed that the levels of Cse4 are increased in psh1Δ cells , and additionally found that the levels of inner kinetochore proteins Mif2 , Okp1 , Ame1 and Ctf19 also increase ( Fig 2 ) . The increase in kinetochore-loaded Cse4 was higher than the other inner kinetochore proteins , suggesting that some of the excess Cse4 is not able to recruit these additional proteins and maybe part of a local ‘cloud’ of Cse4 adjacent to the kinetochore [40] or that it is in a form that is unable to recruit the other components . Consistent with the former notion , we find that the increased Cse4 in a psh1Δ mutant is spread over a larger area , although this is not true for all kinetochore proteins that are elevated in psh1Δ cells ( Fig 3 ) . This may explain why a large increase in Cse4 levels results in only a modest increase in , for example , members of the COMA complex . Surprisingly , we found that outer kinetochore protein levels are unaffected in psh1Δ cells ( Fig 2 ) . These data support the idea that in these mutants the stoichiometry of the kinetochore is flexible . We found that mutating both PSH1 and UBR2 is sufficient to modestly increase the levels of members of the MIND complex ( Fig 4 ) . In budding yeast , if we assume two Cse4 molecules per centromere , there are about 6–7 MIND complexes per kinetochore in anaphase [7 , 53] . In the psh1Δ ubr2Δ double mutants , the ~ 30% increase of Mtw1 and Dsn1 would correspond to ~2 additional MIND complexes per kinetochore . It is unlikely that the chromosome instability phenotype found in psh1Δ and psh1Δ ubr2Δ ( Fig 7 ) accounts for the difference in kinetochore protein levels ( Fig 2 and Fig 4 ) . If these mutant cells would have a higher number of chromosomes ( due to their CIN phenotype ) , we would expect all kinetochore components to be similarly increased . Instead , we find no change in Ndc10 protein levels in the absence of Psh1 , Ubr2 or both ( Fig 2 and Fig 4 ) , and we also did not find an increase in the outer kinetochore proteins in psh1Δ cells . It is possible that the additional proteins at the kinetochore focus in psh1Δ and psh1Δ ubr2Δ are not part of the structural kinetochore assembly . However , the magnitude of the increase of Mtw1 and Dsn1 in the psh1Δ ubr2Δ double mutant ( Fig 4 ) is similar to the increase in Mif2 and COMA complex proteins in the psh1Δ mutant ( Fig 2 ) . This suggests that the amount of MIND complex binding to the kinetochore is still limited by the amount of inner kinetochore components , consistent with a hierarchical kinetochore assembly . The double psh1Δ ubr2Δ mutant does suppress some characteristics of the psh1Δ phenotype; including meiotic sporulation defects ( Fig 6 ) and mitotic genome instability ( Fig 7 ) . It is possible that partially restoring the stoichiometry between inner and outer kinetochore proteins contributes to this phenotypic suppression . However , it is important to note that there is no evidence that the increased Cse4 levels at the kinetochore in psh1Δ cells cause their CIN phenotype . Collectively our data show that inclusion of kinetochore proteins into the kinetochore focus is flexible in mutant backgrounds . Furthermore , that the genomic instability of psh1Δ cells , which may result from increased Cse4 loading , is suppressed by second mutation , ubr2Δ , that also increases the levels of MIND complex members . In psh1Δ cells , Cse4 is increased at kinetochore foci ( Fig 2 ) and also deposited ectopically in non-centromeric regions [26 , 27] . Both kinetochore and non-kinetochore ectopic pools of Cse4 could contribute to psh1Δ chromosomal instability phenotype [29] ( Fig 6 ) . The negative interaction of psh1Δ with spindle assembly checkpoint components mad1Δ and mad3Δ in the presence of microtubule poison ( Fig 8 ) suggests a decreased kinetochore function in psh1Δ . Surprisingly , ubr2Δ partially rescued benomyl sensitivity of both mad1Δ and mad3Δ also in combination with psh1Δ ( Fig 8 ) . This ubr2Δ suppressor effect was not recapitulated by DNS1 overexpression ( S8 Fig ) , suggesting an additional role of Ubr2 . It is possible that the upregulation of other Ubr2/Mub1 complex targets , such as Rpn4 [54] and Sml1 [55] , contribute to the suppression of mitotic and meiotic phenotypes of ubr2Δ . Ubr2 has been previously shown to reduce Dsn1 protein stability by ubiquitylation [30] , but the impact of Ubr2 in kinetochore composition was not known . Ipl1 phosphorylation on Dsn1 promotes the interactions of the MIND complex with the inner kinetochore proteins [42] . However , the presence of dsn1-S240D/S250D did not increase Mtw1 kinetochore levels in wild type or psh1Δ cells , but only in psh1Δ ubr2Δ double mutant and slightly in ubr2Δ ( Fig 5 ) . Our data suggest an important role of Ubr2 on limiting outer kinetochore loading by restricting MIND complex availability ( Figs 4 and 5 ) . From our data , we cannot be sure whether the changes in kinetochore protein levels are a direct result of changes in ubiquitylation status of kinetochore proteins , the effects may be indirect . We note that the artificial recruitment of Ubr2 and Mub1 to kinetochores does not cause a growth defect [56] . Our data also show that Ubr2 is upstream of Ipl1 in the regulation of outer kinetochore assembly ( Fig 5 ) . Regardless of the mechanism of action of Psh1 and Ubr2 , the flexibility of kinetochore stoichiometry may have some functional significance . Kinetochore components are remarkably well conserved from S . cerevisiae to H . sapiens although the centromeres to which they bind are highly divergent both in length and sequence . It is hard to imagine that an inflexible kinetochore structure would be sufficient to support the rapid evolution that is typically seen for centromere sequences [57 , 58] . Our data in yeast show that overexpression of the kinetochore gene MTW1 is not sufficient to disrupt kinetochore function , however this may not be true for all kinetochore genes or in nascent tumor cells . This is further supported by the observation that overexpression of CSE4 is not lethal without further perturbations to the kinetochore [26 , 27 , 59] .
Yeast strains used in this study are either W303 or S288C background , as indicated in S1 Table . For plasmid construction ( see S2 Table ) , the SPC42-RFP sequence containing 200 bp of the SPC42 promoter was cloned into pX29 plasmid ( CEN6 , LEU2 , CUP1 promoter ) . Then , YFP ( pHT5 ) , CFP ( pHT222 ) , MTW1-YFP ( pHT15 ) or MTW1-CFP ( pHT223 ) were cloned downstream of the CUP1 promoter by gap repair . A sequence encoding four alanine residues was used as a linker between MTW1 and the fluorescent tags , and between SPC42 and RFP . Plasmids were transformed into appropriate strains by lithium acetate transformation and continuously selected in synthetic media lacking leucine . MTW1 , PSH1 and UBR2 genes were disrupted by transforming with PCR products containing either MX6-KAN or MX6-NAT selection cassettes flanked with ~250 bp of sequences upstream and downstream the corresponding coding regions . Gene deletions were confirmed by PCR . Since MTW1 is an essential gene , it was disrupted in a haploid strain harbouring CUP1-pMTW1-YFP::LEU2 plasmid ( pHT15 ) . Transformants were selected in synthetic media lacking leucine and containing G418 and confirmed by PCR . Diploid strain MTW1-YFP/MTW1-CFP ( PT11 ) was transformed using mtw1Δ::KANMX PCR to obtain heterozygous diploids MTW1-YFP/mtw1Δ::KANMX ( PT69 and PT70 ) . Loss of CFP or YFP kinetochore foci was tested by fluorescence microscopy and insertion of the KANMX cassette at one of the MTW1 locus was confirmed by PCR . For microscopy and western blot analysis cells were grown in synthetic complete ( SC ) or lacking leucine SC–LEU media supplemented with 100mg/ml of adenine ( +ADE , 100 mg/mL ) . Yeast strains were grown overnight at 23°C . Cultures were diluted in fresh media to ≈ OD600 0 . 3 and grown for 3 hours before imaging or protein extraction . Cells from log-phase cultures were mounted on microscope slides with 0 . 7% LMP agarose in SC +ADE or SC-LEU +ADE , and covered with 0 . 17 mm glass coverslips . Our microscope system uses a Zeiss AxioImager Z2 microscope , 63X Plan Apo , 1 . 4NA , oil immersion objective and a Hamamatsu CCD ORCAII camera ( 2X2 binning and maximum analog gain ) . The resulting pixel size was 0 . 205 μm . Excitation light was provided by LED Colibri system ( excitation band-pass filter ) : CFP 445 nm ( 445/25 ) , YFP 505 nm ( 510/15 ) , GFP 470 nm ( 474/28 ) and RFP 590 nm ( 585/35 ) . Emission band-pass filters were as follows: CFP 47HE ( 480/40 ) , YFP 46HE ( 535/30 ) , GFP 38HE ( 525/50 ) , and RFP 63HE ( 629/62 ) . Exposure times were optimized for each fluorescent protein and ranged from 100 to 250ms . Z stacks consisted of 17 vertically separated slices with 0 . 4 μm spacing . The theoretical dynamic range of our system is ~3000 levels of brightness , however , in practice this will be somewhat lower . A custom-made protocol in Volocity software was used to quantify fluorescence intensity at kinetochore foci . The protocol finds the brightest spots in the image . Spots within 3 pixels from x , y , z edges of the image were removed from the analysis . A 3D box was drawn concentric to the brightest pixels ( 1 . 36 μm3 ) . The background region was 2 pixels separated from the kinetochore box ( 23 . 51 μm3 ) . Average intensity of the background was subtracted from average kinetochore intensity to obtain the final fluorescence value . Finally , fluorescence values were normalized to the average of wild type or control populations . For quantitation , only post-anaphase kinetochores of dividing cells were selected . To measure the size of individual kinetochore foci we fit two Gaussian distributions to each kinetochore . A five pixel square box was selected for each kinetochore and a local background subtracted . The pixel values in each column and each row were summed and for both the rows and columns and then we used ImageJ’s fitDoFit function to fit a Gaussian curve to the values , separately both the rows and columns ( Fig 3A ) . The two values for the full width at half maximum ( FWHM ) , vertical and horizontal Gaussian fits , were averaged to give a mean FWHM measurement for each focus . The mean FWHM measurements for each experiment were normalized relative to the level in WT cells . Cell were harvested by centrifugation and resuspended in 1 . 5X Laemmli buffer with protease inhibitors ( Roche ) and transferred to a fresh tube containing 0 . 5 mm glass beads . Cells were disrupted with a cell homogenizer . Cells extracts were harvested into a fresh tube and boiled for 5 minutes . Cells debris was pelleted and 20 μL of the protein extracts were loaded in a 12% acrylamide gel ( Biorad ) . Proteins were transferred into a PVDF blotting membrane ( GE Healthcare Amersham ) . The western blot was performed with monoclonal anti-GFP antibody ( Roche ) , anti-PGK1 ( Invitrogen ) , goat anti-mouse HRP antibody ( Abcam ) , and ECL kit ( GE Healthcare Amersham ) . Yeast strains were grown o/n at 30°C in YPD or selective media . Cultures were adjusted to OD600 = 1 , serially diluted and spotted into YPD or selective media plates with 0 . 2% DMSO and 10–15 μg/ml benomyl . For testing effects of overexpression increasing concentrations of CuS04 were added to the media as indicated . Plates were incubated for 2 days at 30°C before images were captured . Diploid strains were grown in YPD at 23°C for 24 hours . Then , cultures were diluted 100X in YEPA media and grown at 23°C until OD600 reached 0 . 6 ( 2X107 cells/ml ) . Cultures were washed once with water , resuspended in SPO media and incubated at 23°C for 3 days . Four independent cultures were tested for each genotype . To test spore viability , 22 tetrads per genotype were dissected in YPD and grown for 2 days at 30°C . Diploid strains were grown in YPD for 24 hours at 30°C . Cultures were diluted to OD600 0 . 3 in YPA ( 1% yeast extract , 2% Bacto-peptone , 1% potassium acetate ) and grown for 12–15 h at 30°C . Cells were then resuspended in sporulation media ( 1% potassium acetate pH7 ) at 23°C for 12 hours . Finally , 1μM β-estradiol ( Sigma ) was added to induce NDT80 expression . Cells were imaged every hour to follow meiotic divisions . MATa strains lacking the Bar1 protein were used to facilitate α-factor G1 synchronization . Strains were grown overnight at 30°C , diluted to OD600 = 0 . 3 and grown for 1 hour . The asynchronous sample was collected at this time , then α-factor was added and cells were incubated for additional for 2 . 5 hours . G1 arrest was confirmed by the presence of the characteristic ‘shmoo’ morphology . Cells were washed twice with water and resuspended in YPD with Pronase E . Samples were taken every 30 minutes until 180 minutes . Cells were prepared for flow cytometry as in [60] . Briefly , cells were fixed overnight in 70% ethanol at 4°C , washed once with water , resuspended in RNAase solution and incubated at 37°C for 2 hours . Cells were then washed once with water and resuspended in protease solution for 30 minutes . For FACS analysis , cells were resuspended in 1μM SYTOX solution ( Invitrogen ) . Cell cycle profiles were generated in a BD Canto Flow cytometer using the GFP filter . G1 , S and G2/M populations were calculated using FCS Express ( De Novo Software ) . For S3A Fig , cell cycle progression was scored by fluorescence microscopy . Cells containing a single Mtw1-YFP ( kinetochore ) and Spc42-RFP ( spindle pole body , SPB ) foci and without bud were scored as G1 cells . Budding cells with a single kinetochore and SPB were scored as S phase . Cells with one kinetochore and two SPB or two kinetochores and two SPBs were scored as G2/M ( Metaphase to Telophase ) . Diploid his3-/HIS1 strains were streaked on fresh YPD plates and grown for 2 days at 30°C . Five colonies of each strain were resuspended in YPD . 3x106 cells were mixed with 3x107 cells of log-phase cultures of haploid mating tester strains ( HIS3/his1- ) . Cells were concentrated by gentle centrifugation and incubated overnight at 23°C . The next day these cells were plated on synthetic dropout plates and incubated for 3 days at 30°C to select for HIS+ mating products . For each colony , mating products originating from both mating type MATa and MATα tester strains were summed . For each strain , the median number of colonies from the 5 colonies was calculated . Strains with a tetracycline operator array , inserted at the URA3 locus of chromosome V and a tetracycline repressor linked to mRFP , were grown overnight in synthetic media at 23°C . The day after the culture was diluted and further grown until log phase . Cells were imaged as explain above . In each image , cells showing aberrant chromosome segregation were identified as containing two TetR-mFRP foci in G1 or S-M Strains were transformed with a CEN plasmid with a selectable marker and grown for two days . 9 colonies were grown overnight in YPD and then plated in either YPD or selective media . The percentage of plasmid loss was calculated by subtracting the amount of cells growing in the selective media to the number of cells growing in YPD . The data is presented as the median of percentage plasmid loss of 9 colonies . | As cells divide , their replicated chromosomes must be correctly allocated to the two nascent daughter cells . This is achieved by the kinetochore , which provides a physical link between the chromosomes and the microtubules that drive their movement . If chromosome separation fails , the resulting cells have an abnormal number of chromosomes . This state is called aneuploidy and is a hallmark of cancer cells . The regulation of the kinetochore is therefore of critical importance in maintaining genome integrity . Since a number of cancer cells have over-active kinetochore genes , it has been proposed that an excess of kinetochore proteins can disrupt the normal assembly or maintenance of kinetochores . We tested this idea in yeast by increasing the amount of a specific kinetochore protein , but found no effect upon the normal loading of kinetochore proteins . Instead , we find that two ubiquitin ligases play a role in maintaining the normal balance of the different kinetochore proteins and that this correlates with correct segregation of the chromosomes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microtubules",
"chromosome",
"structure",
"and",
"function",
"centromeres",
"enzymes",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"enzymology",
"ubiquitin",
"ligases",
"molecular",
"biology",
"techniques",
"cellular",
"structures",
"and",
"organel... | 2016 | Synergistic Control of Kinetochore Protein Levels by Psh1 and Ubr2 |
Dominant mutations in CACNA1A , encoding the α-1A subunit of the neuronal P/Q type voltage-dependent Ca2+ channel , can cause diverse neurological phenotypes . Rare cases of markedly severe early onset developmental delay and congenital ataxia can be due to de novo CACNA1A missense alleles , with variants affecting the S4 transmembrane segments of the channel , some of which are reported to be loss-of-function . Exome sequencing in five individuals with severe early onset ataxia identified one novel variant ( p . R1673P ) , in a girl with global developmental delay and progressive cerebellar atrophy , and a recurrent , de novo p . R1664Q variant , in four individuals with global developmental delay , hypotonia , and ophthalmologic abnormalities . Given the severity of these phenotypes we explored their functional impact in Drosophila . We previously generated null and partial loss-of-function alleles of cac , the homolog of CACNA1A in Drosophila . Here , we created transgenic wild type and mutant genomic rescue constructs with the two noted conserved point mutations . The p . R1673P mutant failed to rescue cac lethality , displayed a gain-of-function phenotype in electroretinograms ( ERG ) recorded from mutant clones , and evolved a neurodegenerative phenotype in aging flies , based on ERGs and transmission electron microscopy . In contrast , the p . R1664Q variant exhibited loss of function and failed to develop a neurodegenerative phenotype . Hence , the novel R1673P allele produces neurodegenerative phenotypes in flies and human , likely due to a toxic gain of function .
Voltage-gated calcium channels ( VGCCs ) are a family of calcium ion selective proteins that both mediate calcium entry into neurons at synapses upon depolarization and are required for calcium-dependent functions in the cell [1 , 2] . VGCCs are composed of multiple subunits , including the α1 subunit which is encoded by the CACNA1A gene and is responsible for Ca2+ entry in neurons . Across species the α1 subunit is required in the nervous system . In mice the tottering ( tg ) mutants have mutations in Cacna1a and exhibit ataxia , motor seizures and cerebellar degeneration [3] . In Drosophila the channel is required for synaptic transmission and cacophony ( cac ) null alleles are lethal [4 , 5] . In mosaic clones mutations in cacophony produce neurodegeneration with defective lysosomal fusion with autophagosomes [6] . Given the essential role of VGCCs in the nervous system it is not surprising that mutations in CACNA1A cause a spectrum of neurological disorders in humans[7] . Spinocerebellar ataxia type 6 ( SCA6 , OMIM #183086 ) is characterized by ataxia , dysarthria , and dysphagia with onset typically in middle adulthood but ranging from 20 to 60 years of age[8] . SCA6 is due to heterozygous expansions of CAG repeats within CACNA1A resulting in large polyglutamine tracts within the protein which cause cytoplasmic aggregations and Purkinje cell degeneration [9 , 10] . Episodic ataxia type 2 ( EA2 , OMIM #108500 ) is characterized by episodes of ataxia with onset in childhood or early adulthood , and affected individuals are often responsive to carbonic anhydrase inhibitors such as acetazolamide [11–13] . EA2 is due to heterozygous deletions , stop-gains , frameshifts , or missense mutations [11 , 14 , 15] . Functional analysis based on patch-clamping suggests that they correspond to loss of channel function and hence reveal haploinsufficiency of the locus [16 , 17] . While EA2 is typically episodic , many patients also have progressive ataxia due to ongoing neurodegeneration which is generally slow [15] . Familial hemiplegic migraine ( FHM , OMIM # 141500 ) is a form of migraine with aura and transient hemiplegia and an age of onset between 5 years and early adulthood [18] . CACNA1A variants associated with FHM are typically missense alleles [19] , and in contrast to EA2-associated missense alleles , electrophysiological evidence suggests that they are gain-of-function mutations [20] . FHM mutations appear to lead to hyperactivity of the channel [21] both by altering the biophysical properties as well as decreasing the inhibitory G-protein association with the channel [22] . While the disease phenotypes for EA2 and FHM appear to relate to different underlying genetic mechanisms , considerable phenotypic overlap between FHM and EA2 means that some individuals exhibit features of both disorders [23 , 24] . Indeed even within the same family , the same CACNA1A variants can produce phenotypes more similar to FHM than EA2 [23 , 25] . Several de novo missense alleles in CACNA1A have been reported in children with congenital ataxia and intellectual disability [26] , as well as with non-progressive congenital ataxia with seizures[27] . However , channel function and its relation to phenotype are not well studied in these severe ataxias [26–28] . Most previous studies used in vitro electrophysiological analyses to assess channel function [20 , 21 , 29] , leaving a need for functional annotation of CACNA1A and its variants in a model organism .
We ascertained 5 individuals who underwent exome sequencing for global developmental delay and congenital ataxia , in whom de novo missense variants in CACNA1A were discovered . Patient 1 was enrolled in the Undiagnosed Diseases Network ( UDN ) at Baylor College of Medicine and concurrently in the Baylor-Hopkins Center for Mendelian Genomics ( BHCMG ) [30] as part of a large-scale research re-analysis of clinical exomes [31] . Patients 2–5 were identified at Baylor Genetics Laboratories ( BGL ) [32 , 33] . All families gave written consent for exome sequencing . The clinical findings of these individuals are summarized in Table 1 ( S1 Case Histories ) . All individuals ( 5/5 ) exhibited global developmental delay , expressive language delay and dysarthric ( 4/5 ) or no expressive speech ( 1/5; Patient 1 ) . Interestingly all subjects had ataxia ( 5/5 ) but to varying degrees , with independent ambulation and unsteady gait in some ( 3/5 ) and more severely impaired ambulation requiring use of walker in others ( 2/5 ) . Other neurological features such as behavioral problems , sensory processing disorders , aggressive behavior and attention deficit were also noted , although these features were not consistent among the patients . Neuroimaging for the patients differed; one of the five subjects had evidence of progressive cerebellar atrophy . In Patient 1 , the initial MRI at 10 months showed a normal cerebellum ( Fig 1A ) , while imaging at 22 months revealed mild cerebellar atrophy ( Fig 1B ) , which progressed at 3 . 5 years ( Fig 1C ) and 8 years ( Fig 1D ) . Patient 1 is the only subject with progressive cerebellar degeneration . In Patient 3 , the cerebellum appeared normal in size at 2 years , although the corpus callosum was thin posteriorly ( Fig 1E ) . In Patient 5 the MRI is normal ( Fig 1F ) . In Patient 4 there is some cerebellar hypoplasia involving the vermis ( S1A Fig ) but not the lobes of the cerebellum ( S1B Fig ) Trio-based exome sequencing for Patient 1 through BHCMG [31] revealed a de novo missense variant ( NM_001127221:c . 5018G>C: p . R1673P; chr19:13346480C>G [hg19] ) in CACNA1A ( S1C Fig ) . Patients 2–5 had clinical proband exome sequencing at BGL [32 , 33] ( see Materials and Methods ) and were all found to have a recurrent missense de novo CACNA1A variant ( NM_001127221:c . 4991G>A: p . R1664Q; chr19:13346507C>T [hg19] ) ( S1C Fig ) . Notably , the four individuals with p . R1664Q and the single individual with p . R1673P all exhibit some similarities to a child previously reported with a de novo p . R1664Q allele with early onset ataxia without seizures or migraine[34] . For each case the ratio of variant reads to total reads and the Sanger confirmation suggested approximately 50% variant alleles meaning the patients are heterozygous . Both de novo changes occur within CpG dinucleotides , hypermutable sites prone to methylation and deamination leading to de novo events [35–37] . These have been noted to affect arginine residues in a number of disease contexts [38–40] . The de novo missense changes in these patients affect conserved arginine residues at the S4 transmembrane segment of domain IV of the protein ( Fig 2A and 2A’ ) . Several pathogenic alleles associated with a range of phenotypes are reported to affect this transmembrane segment of domain IV in ClinVar [41] ( Fig 2A’ ) . For example , p . R1661H is associated with EA2 , while p . R1667W is associated with FHM . The R1664Q allele seen in our patients 2–5 is associated with ataxia and global developmental delay . We observed that the intervals of the transmembrane domains ( I-IV ) all appear intolerant to missense variation ( Fig 2B ) . Notably these S4 arginine variants in domain IV display diverse phenotypes , and other severe cases reported also carry missense variants within the S4 segments in domains III [26 , 27 , 42] and IV [34] of CACNA1A . However , these mutations have not been modeled in vivo , and the functions of these missense variants are therefore not defined ( i . e . , haploinsufficient loss or gain of function ) . As drugs are available to either boost or inhibit calcium channel function , knowledge of mutation mechanism may have important implications for patients . In Drosophila cacophony ( cac ) , the homolog of CACNA1A , is required for synaptic transmission and lysosomal fusion , and cac null alleles are embryonic lethal [4 , 5] . Previously , we isolated numerous alleles of cac in a forward genetic screen for essential genes that affect the function of photoreceptors based on defective ERGs in homozygous mutant eye clones [20]: cacJ is an early nonsense mutation which is an embryonic lethal , and cacF is a missense mutation affecting a key glutamate residue in the calcium ion selectivity filter loop and is larval lethal ( Fig 3A and 3B ) . In mosaic eye clones , both mutations lead to expanded nerve terminals , synaptic vesicle accumulations , and aberrant lysosome-autophagosome fusion defects [6] . To test the functional consequences of the variants identified in the five subjects , we designed a rescue-based strategy . To rescue the phenotypes associated with the cac alleles , we first selected a 77 kb P[acman] transgenic construct [6] that contains the entire 53 kb genomic region of cac , including endogenous enhancers to drive proper expression of the transgene . We introduced the two variants found in the subjects by recombineering and we inserted these genomic rescue ( GR ) constructs into the identical VK37 docking site [6] in the fly genome by phiC31-mediated recombination [6] to avoid position effects ( Fig 3A ) . We labeled these constructs GR-WT ( fly wild type ) , GR-R1673P ( Patient 1 ) and GR-R1664Q ( Patients 2–5 ) corresponding to the human proteins/variants . These mutations are within the S4 transmembrane segment of domain IV that is nearly identical between flies and humans ( Fig 3C ) . The wild type P[acman] transgene rescues the lethality associated with Drosophila cacJ and cacF ( Fig 3D ) . However , GR-R1673P ( Patient 1 ) mutation failed to rescue lethality , whereas GR-R1664Q ( Patients 2–5 ) was able to rescue lethality partially ( 41% of expected viable progeny ) . This data suggest that the two mutations have functional consequences in vivo , and that the R1673P mutation seems to behave as a more severely impaired allele . Next we performed ERGs [6] in cac mutant clones rescued with either a wild type or a mutant P[acman] GR construct . ERG recordings reveal two key features: the ‘on’ and ‘off’ transients ( red dotted circles , Fig 3E ) typically reflect synaptic transmission between pre- and post-synaptic cells , whereas the amplitude of the depolarization ( red bracket in Fig 3E ) is a measure of the phototransduction activity . cac mutations typically affect the ‘on’ and ‘off’ transients but have little or no effect on the amplitude , consistent with the role of cac in synaptic vesicle release [6] . To assess the function of the variants from the subjects , we generated homozygous mutant eye clones of cacJ and cacF in young animals ( 3 days old ) and compared the ERGs in flies that carry the wild type and mutant P[acman] GR constructs . Both cac alleles exhibit loss of synaptic transmission as evidenced by loss of the ‘on’ and ‘off’ transients ( Fig 3E , S2A Fig ) . Mutant flies carrying the wild type P[acman] cac transgene exhibited normal ‘on’ and ‘off’ transients ( Fig 3E , S2A Fig ) . Interestingly , GR-R1673P ( Patient 1 ) dramatically increased the amplitude of the on and off transients , whereas GR-R1664Q ( Patients 2–5 ) failed to rescue the synaptic transmission defect caused by the cacJ and cacF mutations . These data suggest that the R1673P allele seen in Patient 1 is a gain-of-function mutation , and R1664Q found in Patients 2–5 is a loss-of-function allele based on the ERG defects observed in the photoreceptors in Drosophila . We also examined the ‘on’ and ‘off’ transients of flies with the P[acman] transgenes in a wild-type y w ( cac +/cac+ ) background as well as heterozygous backgrounds ( cacJ /+ and cacF /+ ) . Although there was a slight nominal increase in the ‘on’ transient in the R1673P animals , it was not statistically significant ( S2B Fig ) . As increased Ca2+ influx typically causes excitotoxicity which may lead to the demise of neurons [6] , we recorded the ERGs of 8 genotypes ( cacF or cacJ mutations rescued by GR-WT , GR-1673P , GR-1664Q , or no rescue ) in 30-day-old flies . We observed a loss of depolarization amplitude in the R1673P ( Patient 1 ) flies in the cacF ( missense ) mutant background ( Fig 4A ) . In contrast , GR-R1664Q ( Patients 2–5 ) did not lead to a significant decrease in amplitude . Interestingly , when the transgenes were tested in the cacJ ( nonsense ) mutant background the reduction in amplitude for R1673P was less severe although still statistically significant ( S2C Fig ) . Similar to those tested in cacF background , the R1664Q variant did not lead to a significant decrease in amplitude ( S2C Fig ) . These data suggest that the increase in activity observed in young flies that carry the R1673P variant leads to age-dependent deterioration of the phototransduction pathway , a phenotype that has not been previously observed in cac loss-of-function mutations [6] . To assess if the R1673P allele may also affect the ultrastructure of the photoreceptors , we performed transmission electron microscopy ( TEM ) . 30-day-old cacF mutant flies show a slight change in retinal morphology , but the photoreceptor cell bodies retain all seven rhabdomeres and normal overall structure compared to cacF rescued by the wild type cac containing P[acman] clone ( Fig 4C versus 4B ) . Moreover , the cacF mutants rescued with GR-R1664Q ( Patients 2–5 ) did not exhibit obvious morphological defects of photoreceptors at 30 days ( Fig 4D ) . In contrast , the cacF mutant photoreceptors rescued by GR-R1673P ( Patient 1 ) show obvious features of photoreceptor neurodegeneration ( Fig 4E ) . The rhabdomeres of the photoreceptors are severely disrupted ( arrows ) , and the cytoplasms of these cells are filled with autophagic vesicles implying neurodegeneration . The severe neurodegenerative phenotype observed here has never been seen in any cac alleles , including the null alleles , thus suggesting that R1673P may act via a ( toxic ) gain-of-function mechanism . We also performed TEM at the level of the lamina , where the presynaptic photoreceptors make synaptic connections with post-synaptic neurons . These data show even more dramatic differences in phenotype between the cacF mutants rescued by wild type and mutant P[acman] transgenes ( Fig 5 ) . cacF rescued by the wild type cac P[acman] clone displayed normal morphology of six photoreceptor terminals ( green areas in Fig 5A ) . Consistent with our previous findings [6] , the cacF mutant exhibited aberrantly expanded terminals , accumulation of autophagic vacuoles ( AVs ) and some signs of synaptic degeneration ( red area in Fig 5B ) . The P[acman] clone containing the R1664Q ( Patients 2–5 ) variant partially rescued this terminal expansion phenotype ( Fig 5C ) , whereas the cacF mutant rescued by GR-R1673P ( Patient 1 ) shows smaller size of their terminals ( Fig 5D , Quantification in Fig 5E ) , a phenotype which is likely due to the depletion of more synaptic vesicles with gain of channel function . In addition , we saw evidence of a dramatic synaptic degeneration in the GR-R1673P rescued flies ( red outlined areas in Fig 5D ) compared to the other genotypes . These data indicate that the R1673P mutation causes severe neurodegeneration in both photoreceptor cell bodies and terminals , likely due to gain of channel function that is toxic to neurons during aging . Interestingly , we did not observe any severe neurodegeneration in cacJ ( nonsense ) mutants rescued by GR-R1673P in both retinae and laminae ( S3 and S4 Figs ) . Instead , we observed partial rescue of the terminal expansion phenotype , and notably autophagic vacuoles did not accumulate ( S4D Fig ) . The latter suggests that lysosome-autophagosome fusion function is still present in cac with the R1673P mutation . The differences in the phenotypes observed in cacF and cacJ mutants rescued by GR-R1673P further support that R1673P is a gain-of-function variant . In the cacF partial loss-of-function background , the R1673P variant can generate toxic gain-of-function effects that lead to severe neurodegeneration . In contrast , the cacJ null allele expresses no endogenous cac and therefore alleviates the toxicity that arises from the R1673P variant . In summary , our experiments in Drosophila strongly suggest that the R1673P and R1664Q mutations are likely to be functional in human and likely to exert their effects through distinct mechanisms .
We report five individuals with similar clinical presentations of ataxia , expressive speech delay , motor incoordination , and age of onset . The severe early-onset ataxias seen in these patients are similar to reports of severe early-onset ataxia associated with CACNA1A missense variants observed in the S4 transmembrane segment of domain III ( e . g . p . I1342T , p . V1396M , and p . R1352Q ) [27] or domain IV ( R1664Q ) [34] . Given these unique clinical features , it had been proposed that these represented loss-of-function mutations of the calcium channel [34] . These severe ataxia phenotypes were thought to represent the most severe end of the spectrum of EA2 rather than the gain-of-function mechanisms seen in hemiplegic migraines [29 , 34] . However , in our series , Patient 1 , a girl with a de novo R1673P variant , also exhibited a progressive cerebellar neurodegenerative process of the most severe end of the CACNA1A clinical spectrum . Our Drosophila studies indicate distinct functional consequences when comparing R1673P and R1664Q alleles . Initially we observed the R1673P allele failed to rescue lethality whereas R1664Q partially rescued , suggesting R1673P is a more severe allele . Importantly , the R1673P allele causes a neurodegenerative phenotype based on functional and morphological criteria which was not seen in either R1664Q or in loss-of-function alleles of cac . Despite the overall clinical similarity between the patients , the two alleles exhibit dramatic functional differences in Drosophila , a functional spectrum not observed previously in severe CACNA1A variants . In retrospect , we note that Patient 1 , the only subject with the R1673P variant , had progressive neurodegeneration of the cerebellum , which was a distinguishing feature between her and the other four subjects . The molecular mechanism for the special arginine at position 1673 in regulating the Ca2+ channel function remains unclear . Interestingly , the recently solved crystal structure of CaV1 . 1 suggests that both R1664 and R1673 are positively charged residues within the voltage sensor[43] . Since many disease-associated variants are found in positively charged residues in the S4 segment of domain IV of CACNA1A ( Fig 2A’ ) , to explore their functional differences will be an exciting future topic to eventually establish a mechanistic model for these key residues in Ca2+ channel function . Responsiveness to treatment and medication differs between reported cases of loss-of-function and gain-of-function CACNA1A alleles [11 , 18] . Patients with EA2 and loss-of-function alleles are often responsive to acetazolamide , while patients with gain-of-function alleles and FHM may respond but generally tend to be less responsive . One would predict that calcium channel blockers might be more effective for gain-of-function alleles . Whether patients with the more severe ataxias also differ clinically in their response to treatment remains to be tested . We note that , while Patient 5 had a strong positive response to acetazolamide , Patient 1 did not respond , consistent with this observation . Indeed as a result of our study Patient 1 was started on a calcium channel blocker . In conclusion , deciphering the functional impact of these severe CACNA1A alleles may provide insight into the pathogenic mechanisms and help direct therapeutic interventions .
All human subjects research was approved by the Institutional Review Board at Baylor College of Medicine ( Studies- Undiagnosed Diseases Network protocol , 15-HG-0130 , approved by the National Human Genome Research Institute IRB , and in the Baylor-Hopkins Center for Mendelian Genomics protocol , H-29697 , and "Evaluation of Sequence Variants" H-22769 approved by the IRB for Baylor College of Medicine ) . All families gave written informed consent for whole exome sequencing and publication . All sequencing studies were performed on genomic DNA from blood samples . Patient 1 obtained trio-based exome sequencing through BHCMG [31] . In brief , for patient 1 DNA samples were obtained and prepared into Illumina paired-end libraries and whole-exome capture with BCM-HGSC core design ( 52 Mb , Roche NimbleGen ) , and then sequencing on the Illumina HiSeq 2000 platform ( Illumina ) . The produced data were aligned and mapped to the human reference genome ( hg19 ) through the Mercury pipeline [44] . Single-nucleotide variants ( SNVs ) were called with the ATLAS ( an integrative variant analysis pipeline optimized for variant discovery ) variant calling method and annotated by the in-house Cassandra annotation pipeline that adapts ANNOVAR ( Annotation of Genetic Variants ) and additional in-house tools [45–47] . De novo variants were calculated by an in-house developed pipeline ( DNM-Finder ) [31] for in silico subtraction of parental variants from the proband’s variants in vcf files while accounting for the read number information extracted from BAM files . Bioinformatic tools predicted conservation and pathogenicity of candidate variants , and variants were compared against both an internal database and public databases such as the Exome Aggregation Consortium ( ExAC ) database . Patients 2–5 had clinical proband exome sequencing at BGL [32 , 33] cacF and cacJ mutants were isolated from a chemical mutagenesis screen as described previously [6 , 48] . The mapping and sequencing of the mutants was performed as described [49] . The P[acman] BAC construct that contains the full length cac genomic region was selected from a large P[acman] library that we previously described [49] . A Transgenic line from this BAC ( CH321-60D21 ) was generated previously and named Dp ( 1;3 ) DC131 [49] which we refer to as GR-WT in this paper . Two point mutations ( R1664Q or R1673P ) were introduced into the CH321-60D21 BAC by recombineering using the modified DH10B strain SW102 and a galK positive/counter selection cassette [50] . The reagents for recombineering were obtained from Biological Resources Branch at National Cancer Institute ( NCI ) -Frederick . The microinjections to generate transgenic flies that contain wild type or mutant 77kb P[acman] clones were performed by GenetiVision , Houston , TX . Lethality rescue experiments in Drosophila were done by crossing y w , cacF FRT19A/FM7c or y w , cacJ FRT19A/FM7c with y w/Y; Dp ( 1;3 ) DC131 ( or -R1673P or -R1664Q ) flies . The males that have no FM7c marker in the next generation were considered rescued flies . The numbers of rescued flies were also compared with FM7c males in the same progeny . In order to circumvent the lethal phenotype , we generated mosaic clones in the Drosophila eyes for ERG and EM experiments . Virgin females of y w , cacF ( or cacJ ) FRT19A / FM7c; Dp ( 1;3 ) DC131-R1673P ( or + ) / CyO were crossed with cl ( 1 ) * FRT19A/ Y; ey-FLP males and we examined y w , cacF ( or cacJ ) FRT19A / cl ( 1 ) * FRT19A; ey-FLP/ Dp ( 1;3 ) DC131-R1673P ( or + ) progeny for ERG and EM defects . Homozygous mutant cells were marked by w- and heterozygous cells were marked by w+ . Homozygous wild-type cells were eliminated by the recessive cell lethal mutation ( cl ( 1 ) * ) to give the mutant clones a growth advantage . ERG recordings were performed as previously described [51] . Briefly , adult flies were glued to a glass slide , a recording electrode was placed on the surface of the eye while a reference probe was inserted in the thorax . A fly eye was exposed to a flash of white light for 1 second . Responses were recorded and analyzed with AXON TM-pCLAMP8 software . Data were analyzed by two-tailed unpaired Student’s t test . A p-value of <0 . 05 was considered statistically significant . Laminae in adult flies were processed for TEM imaging as described [52] . Samples were processed using a Ted Pella Bio-Wave microwave oven with vacuum attachment . Adult fly heads were dissected at 25°C in 4% paraformaldehyde , 2% glutaraldehyde , and 0 . 1 M sodium cacodylate ( pH 7 . 2 ) . Samples were subsequently fixed at 4°C for 48 hours . 1% osmium tetroxide was used for secondary fixation and subsequently dehydrated in ethanol and propylene oxide , and then embedded in Embed-812 resin ( Electron Microscopy Science , Hatfield , PA ) . 50 nm ultra-thin sections were obtained with a Leica UC7 microtome and collected on Formvar-coated copper grids ( Electron Microscopy Science , Hatfield , PA ) . Specimens were stained with 1% uranyl acetate and 2 . 5% lead citrate and imaged using a JEOL JEM 1010 transmission electron microscope with an AMT XR-16 mid-mount 16 megapixel CCD camera . The genotypes of the fly strains used are as follows: | Calcium channels control the levels of calcium within cells and are important in human health . Indeed , groups of patients with disorders of balance known as ataxia have been found to have mutations in a calcium channel gene in the human genome called CACNA1A . CACNA1A mutations have also been observed in patients with particular forms of migraine leading to temporary paralysis on one side of the body ( hemiplegia ) . Mutations in CACNA1A are increasingly found in even more severe brain phenotypes in childhood . This research focused on a group of 5 patients with that particularly severe CACNA1A-related disease . One of the patients had a particular genetic misspelling in CACNA1A while the other four had nearby misspellings . We used the fruitfly , Drosophila melanogaster , to generate flies with these same misspellings in a genetic background that lacked the fly version of the calcium channel . Interestingly , by studying these flies we saw differences between the mutation in Patient 1 and the other four patients . These differences suggest one of the mutations produces more neurodegeneration , and indeed we see more degeneration in that patient . The fly studies allowed us to understand the function of the mutations in these patients , and were helpful in guiding treatment decisions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"neurodegenerative",
"diseases",
"social",
"sciences",
"cloning",
"neuroscience",
"animals",
"invertebrate",
"genomics",
"alleles",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"microscopy"... | 2017 | Clinically severe CACNA1A alleles affect synaptic function and neurodegeneration differentially |
Horizontal gene transfer ( HGT ) can promote evolutionary adaptation by transforming a species’ relationship to the environment . In most well-understood cases of HGT , acquired and donor functions appear to remain closely related . Thus , the degree to which HGT can lead to evolutionary novelties remains unclear . Mucorales fungi sense gravity through the sedimentation of vacuolar protein crystals . Here , we identify the octahedral crystal matrix protein ( OCTIN ) . Phylogenetic analysis strongly supports acquisition of octin by HGT from bacteria . A bacterial OCTIN forms high-order periplasmic oligomers , and inter-molecular disulphide bonds are formed by both fungal and bacterial OCTINs , suggesting that they share elements of a conserved assembly mechanism . However , estimated sedimentation velocities preclude a gravity-sensing function for the bacterial structures . Together , our data suggest that HGT from bacteria into the Mucorales allowed a dramatic increase in assembly scale and emergence of the gravity-sensing function . We conclude that HGT can lead to evolutionary novelties that emerge depending on the physiological and cellular context of protein assembly .
The acquisition of new protein functions through horizontal gene transfer ( HGT ) is known to confer selective advantages and enable the occupation of new ecological niches [1–6] . Examples include the acquisition of antibiotic resistance [7] , virulence-promoting factors [8] , expanded enzymatic capability [9–17] , and tolerance of environmental extremes [18 , 19] . In well-understood cases of HGT , the transferred genes generally encode enzymes whose functions appear to be retained in the recipients . The ability to sense gravity allows plants and fungi to orient the growth of shoots and roots , and fruiting bodies , respectively . This response , known as gravitropism , depends on sedimentation of dense cytoplasmic bodies [20–22] , which generate cell elongation-promoting signals at the cell cortex . Plant gravity sensing is mediated by starch bodies that form within specialized plastids [20] . In the fungi , gravitropism has been demonstrated in the multicellular Basidiomycota [21] and the Mucorales [22] . However , gravity-sensing organelles have only been examined in the Mucoralean Phycomyces blakesleeanus [23] where giant single-celled sporangiophores exhibit gravitropism through a combination of buoyant lipid globules and sedimenting protein crystals that form within vacuoles [24] . A crystal-less mutant grows normally , but displays defective gravitropism , indicating that the crystals indeed serve as gravity sensors [24–26] . Similar structures have been observed in other members of the Mucorales [22] , suggesting that this function arose early in this lineage . However , its basis and evolutionary origin remain unknown . Here , we identify the octahedral crystal matrix protein ( OCTIN ) . Phylogenetic analyses indicate that octin was acquired from a gram-negative bacterium . Both Phycomyces crystals and bacterial OCTIN form disulfide-bonded high-order oligomers , suggesting that they share elements of a conserved assembly mechanism . Given the size of bacterial cells , thermal fluctuations are expected to dominate the movement of OCTIN oligomers . This precludes any speculated role in bacterial gravity sensing . We conclude that HGT of a bacterial octin into the common ancestor of the Mucorales is likely to have relieved constraints on OCTIN oligomer size , allowing evolution of the gravity-sensing function . The data exemplify a general mechanism for the evolution of adaptations based on HGT and high-order protein assembly .
To determine the molecular basis of gravity sensing , we purified vacuolar crystals from P . blakesleeanus sporangiophores using the method of Ootaki and Wolken ( Fig 1 ) [27] . As previously observed , a highly purified crystal fraction contains two major proteins , p55 and p14 ( Fig 1C ) [28] . Mass spectrometry indicates that peptides from these bands are derived from the N- ( p14 ) and C-terminus ( p55 ) of a single predicted protein , which we named OCTIN ( Fig 1D ) . Edman degradation defines the N’-termini of p14 and p55 and full-length octin transcript is detected exclusively in sporangiophores ( Figs 1D and S1A ) . These data indicate that p14 and p55 are derived through proteolytic processing of an OCTIN precursor . Furthermore , sequencing the octin gene from the crystal-less mutant reveals a stop codon at W326 ( Figs 1D and S1B ) . Together , these observations identify two OCTIN-derived proteins as primary components of Phycomyces gravity-sensing crystals . Full-length OCTIN is sporadically present in eukaryotes and bacteria ( Figs 1E and S2 ) . In the fungi , OCTIN is found exclusively in members of the Mucoromycotina , suggesting that it was acquired early on in this lineage . Homologs are also found in the protozoan Stramenopiles , including all sequenced Oomycetes , the Pelagophyceae diatom Aureococcus anophagefferens and both sequenced Haptophytes ( the brown alga Emiliania huxleyi and the phytoplankton Chrysochromulina ) . OCTIN also occurs sporadically in diverse bacterial clades , where it is found in Proteobacteria , Acidobacteria , Actinobacteria , and Bacteroidetes ( Fig 1E ) . Mucorales octin sequences do not encode a predicted signal sequence , suggesting localization through the cytoplasm-to-vacuole targeting pathway , which has been associated with the import of oligomeric vacuolar resident proteins [29] . Predicted signal sequences are found in OCTIN homologs from gram-negative bacteria and the Oomycetes , suggesting that these proteins are directed to the periplasm and secretory pathway , respectively . The sporadic distribution of OCTIN in eukaryotes ( Fig 1E ) could be explained by an early origin followed by extensive gene loss . However , both maximum likelihood ( ML ) and Bayesian analyses provide strong support for independent acquisition of OCTIN by the Mucoromycotina and Oomycetes through HGT from bacteria . In the ML tree , the Mucorales and Oomycetes each have a distinct sister bacterial group ( Figs 2 and S3 and S4 ) , while in the Bayesian tree , the Mucorales are nested within a clade of acido- and proteobacteria ( S5 Fig ) . Enforcing eukaryote monophyly on the ML OCTIN phylogeny results in a topology significantly less likely than the unconstrained phylogeny as judged by the Shimodaira’s Approximately Unbiased ( AU ) test ( p-value = 0 . 021 , S1 Table ) . The trees further suggest HGT among bacteria: acidobacteria and proteobacteria , as well as actinobacteria and proteobacteria , are interspersed to form distinct well-supported monophyletic groups ( Figs 2 and S3 , S4 and S5 ) , while the constrained topology consistent with vertical transmission is significantly less likely ( AU test p-value = 0 . 009 , S2 Table ) . OCTIN is found in a large number of species in deep branching clades in the proteobacteria and acidobacteria ( S6 and S7 Figs ) , suggesting an ancient origin in bacteria . Together , the phylogenetic analyses support an origin for the gravity-sensing protein crystal through HGT from a gram-negative bacterium . The OCTIN C-terminus contains a full-length formylglycine-generating enzyme ( FGE ) domain ( Fig 3A ) . In metazoans , FGE catalyzes the oxidation of cysteine to Cα-formylglycine to activate sulfatase enzymes in the endoplasmic reticulum ( ER ) . In humans , its loss-of-function causes the fatal genetic disorder multiple sulfatase deficiency ( MSD ) [30] . Alignment between OCTIN from diverse species and human FGE reveals high overall sequence conservation , with many residues mutated in MSD being conserved in the OCTIN FGE domain . However , key FGE catalytic cysteines are absent in OCTIN sequences , suggesting that OCTIN does not function in sulfatase activation ( S8 Fig ) . Interestingly , many other bacterial FGE domain-containing proteins lack FGE catalytic residues , and like OCTIN , have N-terminal sequence extensions ( S9 Fig ) . In some cases , these extensions show similarity to known domains , which include Kinase , Caspase , DinB , NATCH , and PEGA domains . DinB-FGE has been shown to function as a sulfoxide synthase . This activity depends on DinB catalytic residues that form contacts with the FGE domain [31 , 32] . Together , these data identify a bacterial superfamily of OCTIN-related proteins . The extent to which these function through structural or enzymatic mechanisms remains to be determined . The position and number of OCTIN cysteine residues show significant variation between the diverse OCTIN-containing clades . However , within clades , cysteine residues can be well conserved ( Fig 3A ) , suggesting that they tailor OCTIN to its taxa-specific functions . When Phycomyces crystals are analyzed by SDS-PAGE under non-reducing conditions , p55 shifts to a high-molecular–weight species that migrates as smear around 250 kDa . By contrast , the migration of p14 is unchanged . These data indicate that p55 forms a disulphide-bonded sub-assembly ( Fig 3B ) . Rapid swelling and disintegration of crystals upon treatment with DTT ( dithiothreitol ) reveal the importance of disulphide bonds for crystal lattice stability . ( Fig 3C and S1 Movie ) . Centrifugation confirms this effect—p55 and p14 are pelleted by centrifugation at 100 , 000 x g , whereas DTT treatment shifts both into the supernatant fraction . Together , these data further show that p14 associates with p55 through non-covalent interactions . Crystals also swell and dissolve upon addition of the protein denaturant sodium dodecyle sulfate ( SDS ) ( S2 Movie ) . Neither DTT nor SDS fully solubilizes p55 . However , when combined , they synergize to promote disassembly ( Fig 3D ) . Together , these data show that disulphide-bonded p55 sub-assemblies form a crystal lattice through additional non-covalent interactions . p14 is physically associated with the p55 lattice . However , its role in stabilizing this structure is unclear . The origin of a gravity-sensing crystal through HGT from a gram-negative bacterium raises the important question of how bacterial OCTIN might be predisposed to this function . Bacteria descended from the likely octin donor are not currently genetically manipulable . To investigate this question , we expressed OCTIN from the gram-negative acidobacterium Terriglobus saanensis ( OCTINT ) in Escherichia coli . OCTINT encodes a predicted signal sequence ( SST ) and a SST-mCherry fusion protein is targeted to the periplasm as indicated by a fluorescent ring around the cell periphery . By contrast , a full-length OCTINT-mCherry fusion protein produces punctate fluorescence at the cell periphery ( Fig 4A ) . Both proteins are released upon lysis of the outer membrane , indicating that they are indeed periplasmic ( Fig 4B ) . However , only OCTINT can be pelleted by centrifugation , suggesting that patches seen by fluorescence represent stable high-order oligomers ( Fig 4C ) . Non-reducing SDS-PAGE shows that like Phycomyces OCTIN , OCTINT forms intermolecular disulphide bonds ( Fig 4D ) . Furthermore , as with Phycomyces OCTIN , SDS and DTT synergize to promote OCTINT oligomer disassembly ( Fig 4E ) . Compared with Phycomyces OCTIN , DTT alone has little effect , suggesting that these assemblies rely more on non-covalent interactions . Nevertheless , these data support a related underlying mechanism of self-assembly for Phycomyces and bacterial OCTIN . Phycomyces sporangiophores are approximately 100 μm in diameter [33] and OCTIN crystals have an average edge length of 5 μm [27] . By contrast , octin-containing bacteria whose sizes are known have diameters ranging from 0 . 3 to 0 . 8 μm [34–39] . To the best of our knowledge , bacterial gravitropism has not been observed . Moreover , assuming an OCTIN assembly size of 1 μm or less , and taking into account cytoplasmic viscosity [40] , the density of OCTIN crystals [41] , and the bacterial cytoplasm [41] , an estimation of sedimentation velocity based on Stokes’ law indicates that bacterial OCTIN oligomers would be too small to function as gravity sensors . The low ratio of particle movement by gravitational force relative to Brownian motion ( Péclet number , [42] ) for oligomers in this size range further demonstrates that their movements would be dominated by thermal fluctuations ( S10 Fig and S1 Text ) [22] . While the function of OCTIN in bacteria remains unknown , its ability to form high-order oligomers is likely to have predisposed neo-functionalization towards a role in gravity sensing in the Mucorales . This is likely to have required the accumulation of mutations relating to crystal lattice assembly , vacuole targeting , and proteolytic processing . If primitive assemblies were too small to function as gravity sensors ( S10 Fig ) , what factors could account for the retention of octin ? Phycomyces OCTIN crystals are found in clusters ( Fig 1A ) , which increases their effective size and sedimentation velocity [24] . Similarly , early OCTIN oligomers could have acted as sensors by clustering . Other scenarios involving neutral selection or another function could also have played a role in the evolutionary transition . In the latter scenario , we note that presently available information does not preclude an enzymatic activity for OCTIN . The periplasm of gram-negative bacteria and the eukaryotic secretory pathway are both oxidizing environments that share a related machinery for translocation of proteins from the cytoplasm [43] . Indeed , OCTINT-mCherry is targeted to the ER when expressed in mammalian tissue culture cells ( Fig 4F ) . To determine whether Phycomyces OCTIN ( OCTINP ) can self-assemble upon heterologous expression , we expressed an ER-targeted version in mammalian cells . This version of OCTIN co-localizes with an ER lumenal marker , but does not display a punctate signal , suggesting an absence of self-assembly . Western blotting further shows an absence of proteolytic processing ( Fig 4G ) . This indicates that OCTIN crystal assembly is likely to require taxa-specific processing activities . Many vacuolar hydrolases are synthesized as auto-inhibited precursors , which are activated upon delivery to the vacuole through processing by resident proteases [44] . We speculate that the region between p14 and p55 functions to inhibit crystal lattice formation through an analogous mechanism ( see S11 Fig for a model of OCTIN assembly ) . Phycomyces has yet to be transformed [33] , and this limits its use as a model system . Thus , understanding the control of crystal assembly will require the identification of OCTIN processing factors and reconstitution in a genetically amenable model system . Phylogenetic analyses strongly support the acquisition of bacterial OCTIN by the Mucorales ancestor through HGT ( Figs 2 and S3 , S4 and S5 ) . Through its signal sequence , this protein would have been targeted to the endomembrane system ( Fig 4F ) . In this context , the size constraint on OCTIN oligomers was relieved , allowing eventual increase in assembly scale and emergence of the gravity-sensing novelty . The case of OCTIN exemplifies how HGT of a protein undergoing high-order assembly can lead to a novel function that emerges depending on a combination of cellular potentialities and physiological imperatives .
P . blakesleeanus wild-type strain NRRL155 [25] and crystal-less mutant strain C2 [24] were grown as previously described [41] . Octahedral crystals were purified as previously described [27] . Bands corresponding to p14 and p55 were analyzed by mass spectrometry and Edman degradation ( Alphalyse A/S , Odense , Denmark ) . Peptides p14 , p46 , and p55 identified the same P . blakesleeanus protein ( National Center for Biotechnology Information [NCBI] accession: XP_018295118 . 1 ) . The search for OCTIN homologs was performed with BLASTP [45] against the NCBI nonredundant database [46] using the OCTIN-specific N-terminal domain ( amino acids 1–500 ) as the query . HMMER3 [47] was used to confirm the presence of the FGE domain ( PF03781 ) [48] in BLAST hits . The accessions of these hits are reported in S3 Table . The extended bacterial species trees ( S2 and S9 Figs ) were constructed based on a previously reported microbial phylogeny [49] . The original tree , which contains multiple strains from the same species , was pruned to retain 1 strain per species whose annotated genome is available in the NCBI Reference Sequence Database ( RefSeq , ftp . ncbi . nlm . nih . gov/refseq/ ) . PhyloPhlAn [49] was used to insert additional octin-possessing species that are not present in the original tree ( S4 Table ) . All other species trees ( S4 , S6 and S7 Figs ) were constructed from 400 conserved protein sequences by PhyloPhlAn using RefSeq bacterial proteomes . The presence of signal sequence was predicted using Phobius [50] . Phylogenetic trees were visualized with ETE3 [51] . To construct OCTIN protein trees ( Figs 2 and S3 and S5 ) , OCTIN sequences from the NCBI reference protein database were used . MAFFT [52] with the option E-INS-i was used to obtain sequence alignments , which were trimmed using Trimal [53] at a gap threshold of 70% . ML bootstrap analysis was performed with RAxML [54] using the automatic bootstrapping option [55] ( 300 replicates ) and the PROTGAMMAILGX model as suggested by ProtTest [56] . The human FGE sequence , which serves as the outgroup ( Fig 2 ) , was placed on the ML tree a posteriori using the RAxML option -f v [57] . Bayesian trees were constructed using MrBayes [58] , run with 12 chains , temperature 0 . 05 , sampling every 500th generation for 300 , 000 generations . Convergence was assessed using RWTY [59] . The ML and Bayesian phylogenies , as well as the matrix used to derive them are accessible under the identifier S22330 at TreeBASE ( https://treebase . org/ ) . To compare the ML trees with and without the monophyly constraint , the best-scoring tree with monophyly constraint was constructed using RAxML with the same parameters specified above for the construction of unconstrained trees . Phylogenetic hypothesis testing using the resampling estimated log-likelihood ( RELL ) test , Shimodaira–Hasegawa ( SH ) test , Kishino–Hasegawa ( KH ) test , and AU test was then performed with the PAML package ‘codeml’ [60] and CONSEL [61] . The search for FGE domain-containing proteins was performed with HMMER3 [47] , using the FGE alignment ( PF03781 ) downloaded from http://pfam . xfam . org . The search was performed on RefSeq proteomes of species present in the bacterial phylogeny shown in S2 Fig . Sequences containing at least 100 amino acids upstream of the FGE domains were selected . Annotated domains within these sequences were identified using the hmmscan function of HMMER3 [47] . Homologs of the gliding motility protein GldK , whose N-terminal domain is not annotated , were manually added based on similiarity to the known GldK sequence from Flavobacterium johnsoniae ( NCBI accession: AAW78679 . 1 ) . T . saanensis octin was codon-optimized for expression in E . coli and the synthetic sequence was obtained from Genscript . Full-length octinP was amplified by reverse transcription polymerase chain reaction ( RT-PCR ) from Phycomyces sporangiophore total RNA . Octin sequences and mCherry fusions were integrated into the pETDuet-1 vector ( Novagen , cat #71146 ) for transformation in E . coli strain HMS174 ( Novagen , cat #69453 ) . Primers used in generating the expression plasmids are listed in S5 Table . E . coli periplasmic extract was obtained following a previously described protocol [62] with modifications . The induced culture was centrifuged at 2 , 500 x g and 4 oC for 10 minutes . The pellet was then gently resuspended in ice-cold PE buffer ( 20% sucrose , 1 mM EDTA , 50 mM Tris pH 7 . 4 ) and placed on a nutating mixer at 4 oC for 15 minutes . This was followed by centrifugation at 2 , 500 x g and 4°C for 10 minutes . The supernatant was transferred to a clean tube and supplemented with Halt protease and phosphatase inhibitor cocktail ( ThermoFisher 78440 ) . This extract was aliquoted and flash-frozen for disassembly assays and western blot . OCTINT and mCherry variants were detected by western blotting using horseradish-peroxidase–conjugated rat anti-HA antibodies ( ROCHE , cat# 12013819001 ) or mouse anti-mCherry ( SAB2702286 SIGMA ) and secondary goat anti-mouse IgG ( SAB4600004 SIGMA ) . Blot images were acquired using the ChemiDoc Touch Imaging System ( Bio-Rad ) . Crystals suspended in Tris-buffered saline buffer ( TBS; 10 mM Tris pH 7 . 2 , 150 mM NaCl ) were mounted on a microscope slide . DTT or SDS was added to one side of the coverslip to a final concentration of 50 mM or 0 . 1% , respectively . Crystal disassembly was recorded using an epifluorescence microscope ( BX51; Olympus ) and a digital camera ( Coolsnap HQ; Photometrics ) controlled by Metamorph . Synergistic disassembly of Phycomyces crystals ( Fig 3D ) and bacterial OCTIN oligomers ( Fig 4E ) by SDS and DTT was performed by incubating the crystals or periplasmic extract with the indicated combinations of SDS and DTT for 30 minutes at 25 oC . This was followed by centrifugation at 100 , 000 x g for 30 minutes at 25 oC . The total sample and the resulting supernatant and pellet fractions were analyzed by SDS-PAGE . Overnight cultures of transformed HMS174 cells were diluted into fresh media and allowed to grow to OD600 of 0 . 7 before induction with 1 mM IPTG . After 4 hours , 5 μl of the suspension was diluted into 1 ml of fresh LB media and 300 μl was placed on a 35-mm microscopy dish ( Matek P35G-1 . 5-10-C ) that had been pre-treated with 50 μg/ml poly-D-lysine ( Sigma P7886 ) . After 1 hour the media was removed and replaced with 2 ml of fresh media . Imaging was carried out with a Leica SP8 inverted laser-scanning confocal microscope fitted with a white-light laser and 100x lens of numerical aperture ( NA ) 1 . 4 . Each image is composed of 4 averaged frames taken at 1% laser power at 587-nm excitation with a scan speed of 400 MHz . HeLa cells cultured in 6-well dishes or 8-well chamber slides were transiently transfected with the indicated plasmids using lipofectamine 3000 ( ThermoFisher ) and cultured for 48 hours before fixing for microscopy or harvesting for western blot analysis . Cells were fixed with 4% Paraformaldehyde ( EMS #15700 ) in phosphate buffered saline ( PBS ) and then kept in 90% glycerol PBS for imaging . Imaging was carried out using a Leica SP8 fitted with a 63x objective NA of 1 . 4 . The white-light laser was set to 488 nm and 587 nm for GFP and mCherry , respectively . Images are a single z plane taken with 8 line averages at 5% laser power , with a scan speed of 200 MHz , 50% gain and a pixel size of 70 nm . To extract protein for western blotting , HeLa cells were lysed in RIPA buffer ( 50 mM Tris-HCl pH7 . 4 , 150 mM NaCl , 1% Triton-X100 , 0 . 1% Sodium Deoxycholate , 1% SDS ) supplemented with Halt protease and phosphatase inhibitor cocktail . Insoluble material was pelleted at 10 , 000 x g and the supernatant fraction boiled in SDS-PAGE loading dye . 10 μg of total cell extract was run per lane . Western blotting was carried out as stated above . | A central question in evolutionary biology is how novel traits arise . Gravitropism—the orientation of growth based on gravity—evolved independently in different groups of plants and fungi . Dense bodies formed within distinct organelles play a key role by sedimenting under the influence of gravity to activate signaling pathways that orient growth . Here we identify the octahedral crystal matrix protein ( OCTIN ) as the structural protein that makes up sedimenting vacuolar crystals in the fungal order Mucorales ( pin molds ) . OCTIN appears to have originated by horizontal gene transfer ( HGT ) from bacteria , but the dominance of Brownian motion at bacterial size scales makes an ancestral role in gravity sensing unlikely , indicating that OCTIN evolved its new function after HGT . However , bacterial OCTIN self-assembles in a manner similar to fungal OCTIN . Our data suggest that fungal OCTIN evolved by retaining elements of the original bacterial assembly mechanism , while acquiring new mutations that increase assembly size . Most genes taken up by HGT encode enzymes whose original and acquired function remains closely related; this study by contrast identifies an assembly-based mechanism for the emergence of novelties through HGT . We speculate that these new functions can emerge depending on changes not only in scale , but also in assembly number , shape , and dynamics . | [
"Abstract",
"Introduction",
"Results",
"and",
"discussion",
"Methods"
] | [
"taxonomy",
"horizontal",
"gene",
"transfer",
"crystal",
"structure",
"chemical",
"compounds",
"plant",
"growth",
"and",
"development",
"condensed",
"matter",
"physics",
"crystal",
"lattices",
"plant",
"physiology",
"organic",
"compounds",
"gene",
"transfer",
"developme... | 2018 | Evolutionary novelty in gravity sensing through horizontal gene transfer and high-order protein assembly |
Triosephosphate isomerase ( TPI ) deficiency is a poorly understood disease characterized by hemolytic anemia , cardiomyopathy , neurologic dysfunction , and early death . TPI deficiency is one of a group of diseases known as glycolytic enzymopathies , but is unique for its severe patient neuropathology and early mortality . The disease is caused by missense mutations and dysfunction in the glycolytic enzyme , TPI . Previous studies have detailed structural and catalytic changes elicited by disease-associated TPI substitutions , and samples of patient erythrocytes have yielded insight into patient hemolytic anemia; however , the neuropathophysiology of this disease remains a mystery . This study combines structural , biochemical , and genetic approaches to demonstrate that perturbations of the TPI dimer interface are sufficient to elicit TPI deficiency neuropathogenesis . The present study demonstrates that neurologic dysfunction resulting from TPI deficiency is characterized by synaptic vesicle dysfunction , and can be attenuated with catalytically inactive TPI . Collectively , our findings are the first to identify , to our knowledge , a functional synaptic defect in TPI deficiency derived from molecular changes in the TPI dimer interface .
Triosephosphate isomerase ( TPI ) is a glycolytic enzyme that converts dihydroxyacetone phosphate ( DHAP ) into glyceraldehyde-3 phosphate ( GAP ) . TPI is a non-linear member of the glycolytic pathway , enhancing the efficiency of the catabolic process , and several missense mutations within TPI lead to a disease known as TPI deficiency [1] . TPI deficiency is one member of a group of disorders caused by mutations in glycolytic enzymes , collectively called glycolytic enzymopathies . Glycolytic enzymopathies are largely characterized as blood disorders , with all patients experiencing hemolytic anemia [2 , 3] . TPI deficiency is one of the few glycolytic diseases associated with patient neurologic dysfunction , and by far the most severe [1 , 4] . Clinical examinations of TPI deficiency patients have established that this disease is often characterized by episodic seizures , periodic dystonia , and progressive weakness and flaccidity in extremities [5–11] . Cellular studies have centered on erythrocytes and lymphocytes , leaving it unclear how TPI molecular dysfunction influences the nervous system . Further , the absence of neurologic dysfunction in many other glycolytic enzymopathies has made it unclear whether these symptoms are related to glycolysis or an as-yet-unidentified function of TPI . Several reports have suggested that TPI deficiency is a disease caused by changes in protein conformation rather than metabolic defects [4 , 12–14] . A human protein structure and yeast genetic studies have asserted that defects in TPI dimerization are the primary determinants of pathology [15 , 16] , revealing no catalytic defects in vitro or from cell lysate . However , not all TPI deficiency mutations lack catalytic defects . An erythrocyte study examining two Hungarian brothers with identical TPI alleles revealed equivalent reductions in TPI activity in both individuals [17] , yet one exhibited severe neurologic dysfunction and the other was asymptomatic . Further , a recent structural study demonstrated that the hTPII170V substitution significantly altered enzyme kinetics and protein stability through a molecular alteration near the catalytic pocket [18] . Collectively , each of these studies failed to establish a causal relationship between TPI activity and disease . Previously , a pathogenic substitution in Drosophila TPI ( dTPIM80T ) was identified , eliciting mechanical- and thermal-stress dependent paralysis [19 , 20] . These behavioral phenotypes have been independently established as hallmarks of neurologic dysfunction and each has been used in forward genetic screens to identify novel components of neuronal transmission [21–23] . The dTPIM80T allele was identified in such a screen [20 , 24] , and to-date D . melanogaster is the only model organism to exhibit neurologic dysfunction caused by TPI deficiency . The dTPIM80T protein was found to be prematurely degraded with reduced catalytic activity [25 , 26] . This reduction in catalytic activity was shown to inhibit glycolytic flux as well as induce metabolic stress [19 , 25] , yet did not change ATP levels in vivo [20] . Subsequent studies demonstrated that the dTPIM80T point mutation could be complemented by the addition of a catalytically inactive TPI without increasing lysate isomerase activity or alleviating metabolic stress [25] , suggesting that dTPIM80T may elicit pathology through a change in protein conformation . Thus , we initially examined the molecular source of dTPIM80T pathology . To determine whether Drosophila TPI deficiency was caused by changes in protein conformation , we purified and assessed the physical characteristics of hTPIM82T , the human equivalent of dTPIM80T , and revealed impaired TPI dimerization . These results were further supported when independent alleles bearing mutations at the TPI dimer interface phenocopied the Drosophila behavioral dysfunction seen in the dTPIM80T allele . These experiments provided novel insight into the pathogenesis of TPI deficiency leading to the conclusion that alterations of TPI dimerization are sufficient to elicit neuropathology . Defining the molecular source of TPI neurologic dysfunction led to the generation of new alleles containing dimer-interface mutations . These novel TPI alleles were characterized by extreme behavioral defects , directing new investigations into the neuropathogenic mechanism of TPI deficiency . An examination of vesicle dynamics at the larval neuromuscular junction ( NMJ ) revealed a severe impairment that appears to be related to vesicle recycling . Further , complementation with a TPI allele encoding catalytically inactive TPI rescued both synaptic dysfunction and behavior , thereby characterizing a cellular mechanism of TPI deficiency neuropathology . Collectively , the results of this study support the conclusion that an improperly formed dimer interface is sufficient to elicit TPI deficiency neuropathology . Further , our experiments establish that a functional synaptic defect occurs in our Drosophila model of TPI deficiency .
TPI is a homodimeric enzyme with catalytic sites in the C-terminal of a triose isomerase ( TIM ) barrel tertiary structural motif [27] . Each catalytic site is rigidified through dimerization to increase catalytic turnover , yet each active site works independently [28–30] . A dTPIM80T substitution was previously isolated and demonstrated to elicit pathology in a Drosophila model of TPI deficiency [19 , 20] . The dTPIM80T substitution is physically located in a solvent-exposed region of the protein near the dimer interface [25] . Numerous misfolding events could be hypothesized to occur as a function of the TPIM80T substitution , among them alterations of dimerization [15 , 16] and aggregation [31] . To examine the structural change elicited by M80T in vitro we purified Drosophila dTPIM80T . Previous purification experiments had yielded Drosophila TPI enzyme , but these samples proved aggregation prone at high concentrations . Conversely , purified human TPI ( hTPI ) was well-behaved; therefore , in order to physically characterize TPI we studied the human protein in vitro . To validate the use of human protein in vivo we generated human WT ( hTPIWT ) and human M80T ( hTPIM82T ) alleles in the Drosophila TPI gene locus using an established genomic engineering ( GE ) system [25]; the hTPIM82T substitution is equivalent to dTPIM80T ( Fig 1A ) . We found that hTPIM82T was able to recapitulate the disease phenotypes observed in dTPIM80T ( S1 Fig ) . The phenotypes of hTPIM82T were remarkably similar but less severe than dTPIM80T , possibly due to subtle organism-specific changes in the dimer interface [32] . Confirmation that hTPIM82T pathologically phenocopied dTPIM80T indicated that any conformational change elicited by dTPIM80T was likely retained in the human protein . We utilized dynamic light scattering ( DLS ) to examine potential conformational differences between hTPIWT and hTPIM82T . Analyses of 15 μM solutions of hTPIWT revealed a hydrodynamic radius of 4 . 3±0 . 08 nm , while hTPIM82T exhibited a significant reduction to 3 . 3±0 . 06 nm ( Fig 1B ) ; these results were consistent across two additional protein concentrations , 3 . 75 μM and 30 μM ( Fig 1B ) . The linear slope generated by plotting the intensity correlation data suggested the hTPIWT sample was largely monodispersed , much like that of the 15 μM sample of bovine serum albumin ( Fig 1C ) . Conversely , hTPIM82T samples exhibited a non-linear slope ( Fig 1C ) , suggesting the possibility of a polydisperse protein population . Polydisperse protein populations indicated the sample was a mixed population in solution , and the observed reduction in TPI mean hydrodynamic radius suggested the sample could be a mixture of monomer and dimer TPI species . We examined enzyme dimerization by assessing protein size via gel filtration chromatography . A standard curve was used to establish column resolution , 15 μM samples of hTPIWT and hTPIM82T were injected onto the gel filtration column , and their migration monitored by UV light at 280 nm . hTPIWT samples separated into two distinct peaks–one at ~24 min . and another at ~27 min . corresponding to ~50 kDa and ~28 kDa , respectively ( Fig 1D ) . Dimeric and monomeric hTPI are 54 and 27 kDa , respectively . Integrating the peak areas revealed an ~80:20 split in dimer:monomer ratio of hTPIWT ( Fig 1D inset ) . In contrast , the majority of the hTPIM82T sample eluted at 27 min . , resulting in a ~5:95 dimer:monomer ratio ( Fig 1D inset ) . These data led us to conclude that the hTPIM82T substitution elicited a dramatic conformational change in TPI resulting in a disruption of dimerization . Interestingly , the gel filtration results did not precisely reflect the monodisperse vs . polydisperse observations of the DLS experiments; we believe this could be due to dilution effects as the proteins migrated over the large gel filtration column . Having established that the hTPIM82T mutation alters enzyme dimerization in vitro , we sought to assess whether other substitutions at the TPI dimer interface were sufficient to elicit neuropathology . Two novel TPI alleles ( dTPIT73R and dTPIG74E ) were generated using GE . Our hTPI dimer analyses ( S2 Fig ) and data from previous TPI studies [28] indicated these substitutions would result in dimer defective TPI . dTPIT73R and dTPIG74E dimer interface mutants elicited a more severe pathology than dTPIM80T , and stocks required maintenance over balancer chromosomes due to their poor viability . Test crosses of balanced stocks yielded significantly fewer homozygous animals than the Mendelian predicted 33% , and homozygous animals were extremely short-lived ( Fig 2A ) , with median lifespans of 2 and 5 days for dTPIT73R and dTPIG74E , respectively . Mechanical- and thermal stress-dependent behavioral defects were assessed at Day 1 and Day 2 , respectively , as these phenotypes have been demonstrated to be hallmarks of Drosophila TPI deficiency [19 , 20 , 25 , 26 , 33 , 34] . dTPIM80T was previously described to exhibit a modest phenotype at early time points [19] , and these data corroborate our analyses of the GE dTPIM80T allele ( Fig 2B and 2C ) . Comparatively , the dimer interface mutants displayed a more severe degree of behavioral dysfunction than that seen in dTPIM80T ( Fig 2B and 2C ) . These data support the hypothesis that mutations at the dimer interface are sufficient to induce neurologic dysfunction . Lysate isomerase activity was compared between samples taken from animals homozygous for the dimer interface mutants . First , it was noted that all dimer interface mutants exhibited reductions in TPI activity ( Fig 2D ) . However , a comparison of the dTPIM80T , dTPIT73R , and dTPIG74E lysates revealed a striking observation–the least phenotypically severe mutation ( dTPIM80T ) was characterized by the lowest isomerase activity ( Fig 2D ) . These data support previous observations that TPI activity does not predict the presence or severity of TPI deficiency [25] . Many conformational diseases are elicited through changes in protein structure and stability leading to misfolding , then either sequestration and degradation , or aggregation [35] . First , we examined whether these new dimer interface alleles produced robust levels of TPI protein . We determined TPI levels in our dimer interface mutants as previously [34] , and found that both dTPIT73R and dTPIG74E homozygotes exhibited reduced protein levels ( Fig 3A and 3B ) . It has previously been shown that TPI has the capacity to aggregate and thereby seed the aggregation of other proteins such as tau [31] . When measuring protein levels via SDS-PAGE , it is important to note that not all aggregate species are SDS soluble and a reduction in protein levels can indicate that the aggregates are not passing through the gel matrix . To determine whether the dTPIT73R and dTPIG74E proteins aggregate we used a dot-blot filter trap assay to assess retention differences between TPI mutant isoforms , as performed previously [36] . Lysates were collected from homozygous animals , and PC12 cell lysates expressing EGFP-huntingtin-Q97 ( GFP htt-Q97 ) were used as a positive aggregation control . The results indicate that little TPI was trapped on the 200 nm filter , yet each sample showed a concentration-dependent increase in signal ( Fig 3C ) . Importantly , no differences were observed in TPI signal between the WT and mutant alleles ( Fig 3C ) . These data support similar findings established by sedimentation assays performed on dTPIM80T [34] , and led us to conclude that although these dimer interface mutants display reduced protein levels via SDS-PAGE , this is not due to the insolubility of large aggregates . To date , all but one study examining TPI deficiency in Drosophila have highlighted a reduction in TPI protein levels in disease-associated alleles [18–20 , 25 , 26] . To independently examine the importance of TPI protein levels in vivo , we employed the GAL4-UAS expression system to knock down wild type ( WT ) TPI using a UAS-RNAi line directed toward dTPI messenger RNA ( mRNA ) [37] . These lines were driven with actin-GAL4 + UAS-GAL4 ( actin/UAS-GAL4 ) to obtain a dramatic reduction of TPI in all tissues . Using UAS-RNAi in conjunction with actin/UAS-GAL4 , we found that w;actin-GAL4 , UAS-GAL4/+;UAS-RNAiTPI/+ animals exhibited a dramatic reduction in TPI protein levels similar to that seen in head and thorax tissue from w;;dTPIM80T homozygotes ( S3A and S3B Fig ) . Head and body tissues were assessed separately to ensure equivalent knockdown in both tissues . Next , we examined animal behavior in these knockdown populations to determine whether depletion of cellular TPI was sufficient to elicit TPI deficiency behavioral abnormalities . Mechanical stress responses were used to quantify behavioral dysfunction . None of the knockdown genotypes exhibited abnormal mechanical-stress dependent responses ( S3C Fig ) . No paralysis or seizure-like activity was observed in the knockdown genotypes at elevated temperatures , though hypoactivity was noted , with the knockdown animals consistently dwelling near the bottom of the vial relative to their TPI+ and UAS only controls . These observations suggested that a general depletion of TPI is not sufficient to elicit paralysis or seizure-like locomotor dysfunction , yet do not exclude the possibility that changes in protein conformation , localized subcellular depletions , or changes in protein stability may play a greater role in animal pathology . Previous work demonstrated that a catalytically inactive allele of TPI ( dTPIΔcat ) complemented the behavior and longevity defects of the dTPIM80T allele [25] , a mutation now established to disrupt enzyme dimerization . This previous study suggested that TPI deficiency is a loss-of-function disease caused by either i ) the depletion of cellular TPI , or ii ) a conformational change that could be rescued through the addition of a properly folded yet catalytically open/inactive isoform [25] . Having utilized knockdown strategies to examine the necessity of total TPI levels , we sought to confirm the capacity of dTPIΔcat ( Lys-to-Met , position 11 , Fig 1A ) to complement additional dimer-interface mutations . To evaluate whether dTPIΔcat was sufficient to support normal behavior and longevity , dTPI+/dTPI+ , dTPI+/dTPIT73R , dTPIT73R/dTPIT73R , dTPIT73R/dTPIΔcat , dTPI+/dTPIG74E , dTPIG74E/dTPIG74E , and dTPIG74E/dTPIΔcat animals were collected and tested as outlined above . These experiments demonstrated that the dTPIT73R allele was nearly fully complemented by dTPIΔcat ( Fig 4A–4C ) , similar to the results found with dTPIM80T [25] . It should be noted that this complementation was not fully penetrant; 5 out of the 30 dTPIT73R/dTPIΔcat animals did paralyze after an extended thermal stress period ( Fig 4B ) . The penetrance of the thermal stress complementation is reflected in an increased time to paralysis relative to the homozygous mutant animals ( Fig 4B and 4E ) . dTPIG74E was also complemented by dTPIΔcat , although more modestly than was observed for dTPIT73R . Mechanical stress responses were unchanged in dTPIG74E/dTPIΔcat relative to dTPIG74E homozygotes , though the penetrance of thermal stress sensitivity was decreased to 20 out of 30 animals , and the median lifespan of the dTPIG74E mutants was extended from 5 to 21 days ( Fig 4D–4F ) . Importantly , neither of the dimer interface mutants elicited dominant negative effects within the dTPI+ heterozygotes; to the contrary , dTPIT73R and dTPIG74E promoted a significant increase in animal health , extending the median 48 day dTPI+/dTPI+ lifespans to 77 and 71 days , respectively ( Fig 4C and 4F ) . In toto , dTPIΔcat partially but significantly complemented each of the new TPI dimer alleles . Our experiments with TPI dimer mutations demonstrated that alterations of the dimer interface were sufficient to elicit TPI neurologic dysfunction , and that these phenotypes were able to be complemented with dTPIΔcat . To address how the TPIΔcat substitution may influence its structure , we purified , crystallized , and determined the structure of hTPIΔcat at 1 . 7Å resolution , refining against native data to Rwork and Rfree values of 15 . 8% , and 19 . 6% , respectively ( Table 1 ) . These crystals grew in conditions that were nearly identical to conditions in which we have previously determined the structure of wild-type human TPI [18] , minimizing the effects that changes in the crystallization condition or crystal packing might have on the resulting structure . While the overall fold of hTPIΔcat is highly similar to wild-type ( r . m . s . d of 0 . 35 Å over all atoms ) there are a number of important differences within the catalytic pocket and neighboring regions . First , the active site pocket of our previous hTPIWT structure contained a highly ordered phosphate and bromide ion located where the phosphate and triose groups of the natural substrate , DHAP , would be located [18] . In contrast , the active site pocket of hTPIΔcat was filled with solvent . At the site of the hTPIK13M substitution ( hTPIΔcat ) , the M13 side chain adopts a different conformation than its lysine counterpart , shifting 4 Å away from the catalytic site and interacting with N11 , G233 and L236 at the back of the pocket ( Fig 5A ) . The sidechain positions of important active site residues S96 and E165 are also altered in hTPIΔcat , breaking critical solvent networks and shifting E165 2 . 7 Å away from the position it adopts in wild-type TPI [18] and substrate analog bound structures [38–40] ( Fig 5A ) . Lastly , the lid moves as much as 7 Å away from the active site pocket , adopting an open conformation [40 , 41] ( Fig 5A ) and corroborating established kinetic data demonstrating that this enzyme is catalytically inactive [42] . These data are in agreement with a structure of yeast TPIK12M , G15A containing two mutations within the active site [43] , but were an important control to isolate the structural impact of hTPIK13M . Importantly , examinations of the dimer interface of hTPIΔcat revealed that it is unchanged relative to hTPIWT ( Fig 5B ) . The peptide backbone and side chains of Loop3 form the majority of the TPI dimer interface , and as shown previously , perturbations of this loop disrupt TPI dimer stability [28–30 , 44] . The new crystal structure revealed that the backbone of Loop3 along with important side chains M14 , T75 , G76 , M82 , and E104 , are unaltered in hTPIΔcat ( Fig 5B ) . These structural data indicated that TPIΔcat homodimers are catalytically inactive , with no observable alterations of the overall folding of the monomers or their dimeric assembly ( Fig 5B ) . Complementation of the TPI dimer mutant alleles with dTPIΔcat suggested a physical and/or functional interaction between the two enzymes . Given the dimeric nature of TPI , we sought to first examine physical interactions between TPI species . All of the dimer-interface substitutions used in this study disrupt homodimerization ( Figs 1 and S2 ) , though no experiments had yet addressed how these alterations may change heterodimerization with dTPIΔcat . These putative heterodimers could support or inhibit a critical function of TPI . To examine heterodimer formation in vivo , we measured the capacity of the dimer-mutant TPI isoforms to co-precipitate using a C-terminal Cerulean cyan fluorescent protein ( CFP ) tagged variant of dTPIΔcat-CFP; this allele was previously confirmed to complement dTPIM80T [25] . anti-GFP was covalently conjugated to the AminoLink resin , and the CFP tag was immunoprecipitated ( IPed ) in dTPIWT/dTPIΔcat-CFP , dTPIM80T/dTPIΔcat-CFP , dTPIT73R/dTPIΔcat-CFP , and dTPIG74E/dTPIΔcat-CFP animal lysates and probed . Unconjugated resin was incubated with dTPIWT/dTPIΔcat-CFP lysate and used as a negative control ( - ) ( Fig 6A , IP ) . Upon elution and SDS-PAGE separation , protein size was used to discriminate between the tagged and untagged TPI isoforms; the CFP tag roughly doubled the molecular weight of dTPI-CFP monomer ( ~50 kDa ) relative to dTPI monomer ( ~25kD ) ( Fig 6A , Input ) . Robust amounts of dTPIWT precipitated with dTPIΔcat-CFP , establishing substantial heterodimerization between the two species ( Fig 6A and 6B ) in agreement with the similarities between their respective dimer interfaces ( Fig 5B ) . Conversely , dTPIM80T and dTPIT73R displayed markedly reduced associations with dTPIΔcat-CFP , corroborating their previously established dimerization deficiencies and reflecting their overall prevalence in the lysate ( Fig 6A ) . Finally , it was surprising to see that dTPIG74E produced heterodimerization similar to that seen in dTPIWT ( Fig 6A and 6B ) ; it was predicted that the rotational flexibility of G74 was necessary for the appropriate positioning of loop 3 and establishment/rigidification of the dimer interface [28] . The coIP experiments suggested that the TPI species responsible for dTPIT73R phenotype suppression in the animals was not a dTPIΔcat heterodimer; the heterodimer was a very small fraction of the total TPI enzyme in lysate ( Fig 6 ) . Conversely , the dTPIΔcat-CFP::dTPIG74E heterodimer existed as a substantial fraction of the total TPI ( Fig 6 ) , yet exhibited modest complementation of the abnormal behavioral phenotypes ( Fig 3 ) . The substantial and unanticipated presence of the dTPIΔcat-CFP::dTPIG74E heterodimer could indicate an allele-specific dominant interaction . To examine whether we could enhance the capacity of dTPIΔcat to suppress dTPIG74E , we designed a double-mutant aiming to revert heterodimer formation . An allele was generated bearing both substitutions , dTPIT73R , G74E , and animals homozygous for this allele displayed aggressive behavioral phenotypes and shortened lifespans ( S4A , S4B and S4D Fig ) . Immunoprecipitation experiments found that the double substitution reduced dTPIT73R , G74E heterodimerization with dTPIΔcat-CFP relative to dTPIG74E ( Fig 6 ) . Finally , when paired with the dTPIΔcat allele , the addition of the T73R substitution to dTPIG74E enhanced the capacity for dTPIΔcat-behavioral complementation ( S4A and S4B Fig ) . The mean time to recovery after mechanical stress was reduced from 204 sec . in dTPIT73R , G74E homozygotes to 50 sec . in dTPIT73R , G74E/dTPIΔcat animals with approximately 60% of the animals no longer responding to the stressor ( defined as a recovery time ≤ 5 sec . ) ( S4A Fig ) ; similar complementation was observed in the thermal stress assay ( S4B Fig ) . Curiously , the longevity of the dTPIT73R , G74E/dTPIΔcat animals was unchanged relative to dTPIT73R , G74E homozygotes ( S4D Fig ) ; this is the second time that TPI deficiency behavioral abnormalities and longevity have not paralleled each other [18] , suggesting the possibility of independent pathogenic mechanisms ( see Discussion ) . The inverse correlation between dTPIΔcat heterodimerization and behavioral complementation suggested that dTPIΔcat did not complement TPI deficiency behavioral phenotypes via heterodimer formation . Further , disease severity did not correlate with isomerase activity ( S5B Fig ) ; complementation of the dimer-mutant alleles with dTPIΔcat failed to increase isomerase activity and in all but one case significantly decreased activity ( S4C and S5 Figs ) . These data led us to conclude that dTPIΔcat does not “suppress” TPI deficiency behavioral phenotypes through a general influence on TPI catalytic activity . TPI deficiency complementation was not corroborated by an enhancement of TPI catalysis; however , the mean temperature-dependent time to paralysis of the dTPIT73R allele ( 27 sec ) was a striking result ( Fig 2B ) and suggested a previously unknown role of TPI . Rapid ( <60 sec . ) temperature-dependent paralysis had only been identified in a handful of mutants in Drosophila and typically results from neural conductance or synaptic vesicle recycling defects [45] . To determine whether TPI was influencing vesicle dynamics , we first examined vesicle endocytosis at the synapse using the lipophilic dye , FM1-43 . FM1-43 is a water soluble membrane dye that increases its fluorescence when bound to cellular membranes . During endocytosis , the dye will bind to the outer leaflet of the plasma membrane and become internalized within the synapse providing an optical measurement of endocytosis . Measuring vesicle dynamics in this context allowed us to assess two possibilities; i ) a primary recycling defect due to impaired endocytosis , or ii ) a secondary recycling defect due to aberrant exocytosis . We dissected larvae homozygous for dTPIWT , dTPIT73R , and Shits1 as previously detailed [46] . The NMJ preparations were heated to 38°C over 3 min . and a loading curve was generated from a series of three different high [K+] + FM1-43 loading times– 15 sec . , 30 sec . , and 60 sec . as previously detailed [47] . dTPIWT displays a progressive increase in dye loading from 15 sec . to 60 sec . ( Fig 7A ) , while the temperature sensitive dynamin mutant control Shits1 showed no signs of vesicle recycling at any heated time points ( Fig 7D , data not quantified ) . Conversely , although dTPIT73R displayed similar loading to dTPIWT at 15 and 30 sec . , dTPIT73R exhibited a striking 50% decrease in loading at 60 sec . ( Fig 7A , 7B and 7D ) . This progressive decrease in endocytosis was stimulation and temperature dependent; loading experiments performed at room temperature did not exhibit an endocytic defect ( Fig 7C ) . As previous experiments had highlighted the capacity of dTPIΔcat to complement the adult behavioral defects of dTPIT73R , we examined dTPIT73R/dTPIΔcat larvae to assess the relationship between vesicle endocytosis and animal behavior . The dTPIT73R/dTPIΔcat animals displayed a significant increase in vesicle endocytosis relative to dTPIT73R ( Fig 7B and 7D ) . These results demonstrate that dTPIΔcat complements adult behavior and vesicle endocytosis defects . The utilization of chemical stimulation in these preparations demonstrates a synaptic defect arising from the severe dTPIT73R dimer mutation as this methodology bypasses conductance requirements . A reduction in vesicle dye uptake could be derived from defects in endocytosis or exocytosis , and indeed , these activities are intimately linked [48] . To examine temperature-dependent changes in vesicle fusion , dTPIWT and dTPIT73R animals were i ) loaded with dye at RT for 3 min . , ii ) washed with 0 mM Ca2+ HL-3 , iii ) imaged , iv ) heated to 38°C , v ) vesicle fusion initiated with 30 sec . of high [K+] HL-3 , and vi ) reimaged . Care was taken to ensure the same synapses were imaged at loading and unloading timepoints . Preliminary experiments demonstrated that 60 sec . of high [K+] stimulation completely unloaded the synapses in each genotype; therefore 30 sec . was analyzed to achieve a measurable dynamic range . Unloading experiments at elevated temperatures demonstrated no change in vesicle exocytosis between dTPIWT and dTPIT73R at 38°C ( Fig 7E and 7F ) . Finally , functional changes at the synapse can be the result of acute impairments in recycling machinery or more chronic developmental defects . Mutations in the E3- ubiquitin ligase Highwire or alterations in the trans-synaptic signaling proteins wingless and Glass-bottom boat have been shown to alter synaptic function through primary changes in development [49–51] . These changes in synaptic physiology are accompanied by dramatic alterations in synaptic morphology , a hallmark of neurodevelopmental defects . To examine whether aberrant neurodevelopment may contribute to this recycling deficit , we morphologically characterized the Drosophila NMJ from segment A2 , muscle 6/7; this particular NMJ is highly elaborate and therefore sensitive to developmental perturbations . The dTPIM80T , dTPIT73R , and dTPIG74E alleles all exhibited early lethality if maintained at 25°C , therefore development was scored at RT . Third instar larva were dissected , and an assessment of bouton number and branches revealed no significant developmental differences in the thermal-stress sensitive mutants relative to dTPIWT ( Fig 8 ) . These results suggest that the synaptic defect is an acute disruption of function , and not likely a secondary defect caused by altered development . Collectively , these data demonstrate that TPI deficiency thermal-stress sensitivity is characterized by acute perturbation of synaptic vesicle dynamics .
The pathogenic hTPIM80T substitution impairs TPI dimerization . These results were obtained from purified proteins and do not corroborate those from non-denaturing gel filtration experiments performed on animal lysates [26] . However , several in vitro studies have found that mutations that impair TPI dimerization severely destabilize the protein [28 , 30 , 44 , 52 , 53] . In vivo , unstable proteins are bound by chaperones and either refolded , targeted to the proteasome , or aggregate [54] . The results presented here suggest that dTPIM80T does not cause TPI to aggregate ( Fig 3 ) , while previous work extensively details the recruitment of Hsp70 and Hsp90 to dTPIM80T and its degradation through the proteasome [34] . Therefore , we hypothesize that TPI monomer may not have been detected previously in animal lysates due to its rapid sequestration and degradation . We believe these data , along with the previous inability to identify monomer in vivo , collectively suggest that TPI does not stably exist in vivo as a soluble monomer . We utilized our GE system to generate two additional TPI alleles with point mutations at the dimer interface that have previously been shown to impair homodimerization [28] . These substitutions were located at the tip of the 3rd loop of TPI that extends into its dimer partner and stabilizes/rigidifies a network of hydrophobic interactions and hydrogen bonds which form the dimer interface [28 , 29 , 55] . The substitution of these dimer interface residues resulted in severely pathogenic TPI alleles , eliciting greater behavioral dysfunction and shorter lifespans than dTPIM80T ( Fig 2 ) . A universal molecular mechanism of TPI deficiency pathogenesis is currently unclear . To date , two crystal structures of disease-associated TPI mutations have been reported [15 , 18] . These mutations were found in two distinct structural regions of the TPI homodimer . The first structure to be solved was from the most commonly diagnosed TPI deficiency substitution , hTPIE104D [15] . The hTPIE104D substitution is a conservative alteration of a charged residue at the dimer interface that results in reduced dimer stability , but unchanged catalytic activity [15 , 16] . The second disease-associated structure was an hTPII170V substitution , a conservative substitution found on the catalytic lid of the enzyme that enhances thermal stability and reduces catalytic activity [16 , 18] . The dimer substitutions used in this study share several molecular characteristics with the TPIE104D human mutation , and given the conservation of the TPI enzyme , we propose that these patients likely share similar molecular and cellular dysfunction as identified in this study . Conversely , the dimer substitutions are not predicted to share many molecular similarities with the hTPII170V mutation indicating that dimerization defects are sufficient but not necessary to elicit TPI deficiency . It is interesting to note that although dTPIT73R and hTPII170V both exhibit mechanical and thermal stress sensitivities in Drosophila , the behavioral dysfunction caused by the dimer substitutions is far more severe , and hTPII170V does not influence animal longevity [18] . Further , the capacity to attenuate behavioral dysfunction but not longevity in dTPIT73R , G74E/dTPIΔcat suggests an independent pathogenic mechanism ( S4 Fig ) that may not be determined by TPI dimerization . TPI dimerization and protein stability in vivo will ultimately influence catalytic capacity , and many TPI activity measurements from animal models and patient tissue samples have identified a reduction in isomerase activity [7 , 14 , 17 , 18 , 25 , 56–58] . However , our measurements of isomerase activity from healthy and affected animal lysates argue that TPI dimer integrity is a stronger determinant of behavioral dysfunction . Still , this could indicate that reduced TPI activity is corollary to the disease or a contributing factor to an alternative pathogenic mechanism . Previous studies have demonstrated that the redox state is altered in TPI deficient cells and organisms [33 , 59 , 60] . The redox status in TPI deficiency is proposed to be altered by abnormal flux through the pentose-phosphate pathway as well as potential accumulation of advanced glycation end-products ( AGEs ) [16 , 20 , 33 , 59–61] , and the accumulation of redox damage in the nervous system is strongly linked with several neurodegenerative diseases [62–64] . Interestingly , a study in yeast demonstrated differences in redox responses between the E104D and I170V mutations [16]; these results could imply different modes of pathogenesis that are dependent on the conformational and catalytic states of TPI . One unresolved aspect of this study was the inability of dTPIΔcat to fully complement the behavioral defects of dTPIG74E . Co-IP experiments suggested that complementation correlated with an inability of dimer-interface mutants to form heterodimers with dTPIΔcat ( Fig 6 ) . These data would imply that dTPIG74E may be exhibiting a dominant negative effect as a heterodimer , but this conclusion was inconsistent with our genetic analyses of dTPIWT/dTPIG74E animals ( Fig 4 ) . To investigate whether TPIG74E may be interacting differently with TPIWT than with TPIΔcat , we purified , crystallized , and determined the structure of hTPIΔcat at 1 . 7Å resolution . While the hTPIK13M substitution ( hTPIΔcat ) resulted in multiple rearrangements that mimic the open or non-catalytic TPI conformation , the dimeric interface remained essentially unchanged , including the peptide backbone of Loop3 and sidechain positions hT75 , hG76 , and hM82 ( Fig 5 ) . To address how hTPIT75R and hTPIG76E substitutions may influence the dimer interface , we generated models of heterodimers in which one subunit contained either hT75R , hG76E , or both hT75R and hG76E substitutions , while the other monomer remained unaltered . Models were made using either wild-type TPI hTPIWT ( PDB: 4POC ) or the hTPIΔcat ( PDB: 4ZVJ ) as the structural template , and subjected to analysis by RosettaBackrub [65] . Briefly , Rosetta scores are predictions of the most energetically stable conformations with higher scores indicating less favorable positioning of the model . The algorithm was run 50 times for each mutation to be modeled . Of these 50 simulations , the lowest scores of hTPIΔcat::hTPIT75R and hTPIΔcat::hTPIG76E were selected and shown ( Fig 9A and 9B ) , while the modeled structures whose Rosetta scores fell within the best 10% of its respective ensemble were collected for analysis ( Fig 9C ) . Modeling the hTPIT75R and hTPIG76E substitutions as homodimers or heterodimers with the hTPIWT structure produced high Rosetta scores , predicting poor energetic favorability ( Fig 9C ) in agreement with our gel filtration experiments ( S2 Fig ) . To examine TPIΔcat heterodimers we used the new hTPIΔcat structure to model hTPIΔcat::hTPIT75R and hTPIΔcat::hTPIG76E . These experiments predicted a high Rosetta score for hTPIΔcat::hTPIT75R and a very low one for hTPIΔcat::hTPIG76E , corroborating the results of our animal lysate coIP experiments and suggesting the simulated heterodimers may accurately represent the conformations of these molecules . The hTPIΔcat::hTPIT75R heterodimer with the lowest Rosetta score suggests the hR75 residue may orient toward the catalytic pocket of hTPIΔcat , lining the floor of the substrate-binding pocket through the displacement of hE165 and hK13M ( Fig 9A ) . This orientation of hR75 into the catalytic pocket is similar to that previously described by Wierenga and colleagues [28] . The hTPIΔcat::hTPIG76E heterodimer with the lowest Rosetta score suggests that hE76 finds a stable position in the dimer interface through the displacement of hE104 and hR98 , possibly via coordination of the terminal amide of hN65 ( Fig 9B ) . Interestingly , perturbation of hE104 has been shown to significantly alter the TPI dimer interface and elicits TPI deficiency in humans through a conservative hTPIE104D substitution [15] . These modeling predictions suggest that the character of the hTPIΔcat::hTPIG76E dimer interface is drastically altered relative to hTPIΔcat homodimers . Drosophila TPI deficiency neurologic dysfunction is characterized by impaired vesicle dynamics at the neuronal synapse , a defect we believe is likely conserved in human patients . The key to deciphering this pathogenic mechanism was the behavioral severity of the newly generated TPI dimer interface mutants . The dTPIT73R allele exhibited temperature-dependent paralysis at a mean time of approximately 27 sec . , an acute behavior that is rare and highly enriched for synaptic or conductance defective mutants . Only a handful of Drosophila mutant alleles have been identified with rapid temperature-dependent paralysis , including those of voltage-gated Na+ , K+ , and Ca2+ channels ( para , sei , cac ) [66–68] , the sodium-potassium exchanging ATPase ( ATPα ) [69] , and components of vesicle fusion and recycling ( N-ethylmaleimide sensitive factor–dNSF1 , dynamin–Shi ) [70 , 71]; and after noting these phenotypic similarities we broadly examined synaptic function . Stimulation was conducted using high [K+] bath applications , thereby bypassing the participation of Na+ and K+ channels . The dTPIT73R mutants were characterized by normal endocytosis during acute stimulations ( 15 and 30 sec . ) but exhibited a dramatic reduction after 60 sec . of stimulation ( Fig 7A ) , suggesting a time/excitation dependent phenotype . Further , complementation with the dTPIΔcat allele significantly increased the temperature-dependent FM1-43 loading at these terminals ( Fig 7B and 7D ) , similar to its complementation of adult thermal stress-induced paralysis . Finally , measurements of vesicle exocytosis at these elevated temperatures did not indicate an exocytic defect ( Fig 7E and 7F ) . We were unable to detect terminal signal above background after 60 sec . of unloading , so we cannot unequivocally define the nature of the vesicular dysfunction , but the data suggest it is likely due to impaired endocytosis . The observed synaptic defect provides insight into why the majority of human patients present with TPI mutations affecting the dimer interface . First , all substitutions that disrupt the dimer interface have been shown to destabilize the enzyme in vitro [28 , 29 , 55] . This destabilization is likely responsible for the reduced cellular TPI found in patient samples , and our work with the dTPIM80T , dTPIT73R , and dTPIG74E mutants provide additional evidence that dimer interface substitutions reduce TPI levels in vivo ( Fig 3A and 3B ) . Secondly , the cellular depletion of dTPIM80T has been shown to be mediated by heat shock protein sequestration and proteasomal degradation [34] . If chaperones sequestered and degraded these misfolded or unstable proteins , this would likely prevent the distribution and maintenance of TPI at specific subcellular locales . Recent work has shown that the anterograde transport of globular/soluble proteins to the terminals is a slow process , moving at a rate of approximately 0 . 008–0 . 01μm/sec [72] . To put this in the context of the Drosophila nervous system , the length of the relatively short larval motor axon innervating muscle 4 of segment A3 has been measured to be ~220μm [73] . Based on these approximations , one could estimate that it would take ~6 hrs for TPI translated in the soma to be transported to the axonal terminal . In this way , substitutions that affect protein stability would likely result in improper localization or sequestration of TPI during distal transport , ultimately depleting TPI at the synapse . This proposal is also consistent with the inability of RNAi knockdown to recapitulate Drosophila TPI deficiency behavioral phenotypes , as RNAi alters mRNA transcript levels rather than enzyme conformation or stability . RNAi knockdown of TPIWT would reduce , but still allow the transport and stable accumulation of TPIWT at the synaptic terminal . Two other Drosophila glycolytic mutants , aldolase and phosphoglycerate kinase , have been shown to exhibit temperature-dependent paralysis though with longer onsets [74 , 75] . The phosphoglycerate kinase mutant was shown to exhibit synaptic dysfunction , and the authors asserted that an inhibition of vesicle recycling was likely the cause of the functional defect [75] . In both cases , the animals were found to have depletions in lysate ATP [74 , 75] . It is attractive to speculate that all three of these glycolytic mutants may suggest a pivotal role for glycolysis within synaptic vesicular dynamics , and indeed , recent measurements of ATP consumption in the synapse suggest that glycolytic ATP is the primary substrate used to support synaptic function [76 , 77] . However , the role of glycolytic proteins and their putative energetic importance at the synapse is controversial . Many research groups assert the preeminent utilization of mitochondrial ATP at these sites [78–81] , while the lactate shuttle hypothesis largely circumvents a role for neuronal glycolysis [82 , 83] . Further , the absence of a correlation between TPI catalytic activity and behavioral phenotypes suggests that the enzyme may be complexing with another molecule to facilitate synaptic vesicle cycling . How the dimerization or integrity of the TPI dimer interface impacts synaptic vesicle dynamics remains a mystery , though one candidate for a molecular complex is the actin-regulatory protein cofilin . In Drosophila , cofilin ( twinstar ) and twinfilin mutants have been demonstrated to elicit functional and developmental neurologic defects [84–86] , and actin-regulatory proteins such as cofilin , actin-depolymerizing protein ( ADP ) , and twinfilin are known to influence synaptic vesicle dynamics [85 , 87 , 88] . Recently , cofilin was found to bind to TPI in both its inactive and active forms [89] . The precise binding site between cofilin and TPI is unknown , though with its mixture of charged and hydrophobic pockets , the TPI dimer interface may provide a suitable site for this interaction . Additional studies will be needed to specifically delineate the role of TPI in the synapse . In conclusion , to our knowledge this work is the first to highlight a critical role for TPI in the cycling of vesicles at the synapse , with behavioral correlates similar to the inactivation of vesicle fusion/recycling proteins . These observations help clarify the neurologic symptoms seen in patients and will direct future therapeutic strategies . The findings of this study will guide future investigations regarding the contribution of TPI localization and function to synaptic vesicle dynamics , and ultimately how these properties are perturbed in TPI deficiency .
The Vienna Drosophila RNAi Center ( VDRC ) line used for knockdown experiments was stock #25644 [37]; experiments were also conducted with #25643 with similar results . The w;actin-GAL4 , UAS-GAL4; animals were generated by recombining the second chromosomes of the Drosophila Genetic Resource Center ( DGRC ) stock y1 w1118; P ( w+mC = UAS-Gal4 . H ) 12B and Bloomington Stock Center stock y1w*; P ( Act5C-GAL4 ) 25FO1/CyO , y+; recombinants were screened molecularly and balanced in a w1118 background . The following TPI alleles in this study were generated using the GE system: dTPIWT , hTPIWT , dTPIWT-CFP , dTPIM80T , hTPIM82T , dTPIT73R , dTPIT73R-CFP , dTPIG74E , dTPIG74E-CFP , dTPIT73R , G74E , dTPIΔcat , and dTPIΔcat-CFP . The development of the GE system and the production of the dTPIWT , dTPIM80T , dTPIΔcat , and dTPIΔcat-CFP alleles were initially described elsewhere [25] . Briefly , GE involves the replacement of the TPI gene locus with a phiC31 integration site through homologous recombination . The phiC31 integration system allows nearly seamless integration of complementary vector constructs directly into the TPI gene locus to maintain endogenous spatial and temporal regulation . The TPInull allele used in S1 Fig was generated previously , formerly known as TPIJS10 [19] . RNAi experiments used the TPIM80T allele formerly known as TPIsgk [19] , due to similarities with the UAS-RNAi and GAL4 genetic backgrounds . This study uses the established nomenclature for TPI , assuming the start methionine is removed following translation [13]; all residue numbering in this study uses the same convention . An alignment is included for clarification ( Fig 1A ) . All animal populations assessed were approximately equivalent mixtures of males and females . Site directed mutagenesis was performed using the QuikChange Lightening Site-Directed Mutagenesis Kit ( Agilent Technologies ) . Mutagenesis primers were generated ( Integrated DNA Technologies ) to introduce a Thr-to-Arg codon change at position 73 , and a Gly-to-Glu change at position 74 –both separately and together for the purpose of creating the double-mutant . Mutagenesis was performed on the previously published pGE-attBTPI+ plasmid [25] and confirmed by sequencing . Once the constructs were generated , TPI GE was performed using previously published methods [25 , 90 , 91] . Briefly , the PGX-TPI founder animals were mated to vasa-phiC31ZH-2A animals expressing the integrase on the X chromosome and their progeny injected with pGE-attBTPI constructs . Integration events were identified via the w+ phenotype and verified molecularly . Human TPI enzyme was purified as outlined previously [18] . DLS measurements were taken using a DynaPro Plate reader ( Wyatt Technology ) equipped with a temperature control unit . Purified BSA ( Sigma Aldrich ) , hTPI+ , and hTPIM80T were diluted to concentrations of 3 . 75 μM , 15 μM and 30 μM in 100 mM triethanolamine ( TEA ) ; pH 7 . 6 . Three 75 μl aliquots were loaded onto a 384-well microplate and read at 37°C . Ten measurements were taken per sample and Dynamics V6 software ( Wyatt Technology ) was used to process the scattering data , generating autocorrelation functions . Autocorrelation functions were then analyzed to obtain the hydrodynamic radii . Student’s T test was used to compare samples . DLS experiments were performed three times . Gel filtration was performed as outlined previously [26] . Purified TPI samples were diluted to 15 μM in mobile phase , 100 μl were injected in triplicate , and their elution monitored at 280 nm . Experiments were performed three times . Chromatography traces were collected and analyzed using EZStart 7 . 3 ( Shimadzu ) to quantify the relative monomer and dimer populations . Curve integration data were compared using Student’s T test . Isomerase activity was determined using an NADH-linked assay as previously detailed [25 , 92] . Lysates were diluted to 0 . 1 μg/μl in 100 mM TEA pH 7 . 6 + inhibitors and enzyme activity was assessed . Reaction assays were performed in triplicate using 80 μl mixtures composed of 0 . 5 mM NADH , 0 . 752 mM GAP , 1 unit glycerol-3-phosphate dehydrogenase and 1 μg of lysate protein in 100 mM TEA; pH 7 . 6 . Consumption of NADH was monitored at 340 nm and 25°C using a SpectraMax Plus 384 microplate reader ( Molecular Devices ) . All reactions were performed at least three times . Reaction components were purchased from Sigma-Aldrich . Enzyme activity curves were normalized to reactions performed without GAP . A one-way ANOVA was performed to assess variance , and data sets were compared using Tukey’s post-hoc analysis . Mechanical stress sensitivity was examined on Day 1 by vortexing the animals in a standard media vial for 20 seconds and measuring time to recovery , similar to [93] . Briefly , recovery is defined as two purposeful WT movements including righting , grooming , climbing , or walking . Thermal stress sensitivity was assessed on Day 2 by acutely shifting animals to 38°C and measuring time to paralysis , as previously described [24 , 69] . In these assays , the animals typically seized and either flipped over onto their backs or fell sideways with no successive coordinated movements , i . e . righting , climbing , walking , grooming . Behavioral responses were capped at 360 and 600 seconds where indicated and reported as 360 and 600 sec . Animal lifespan determinations were performed at 25°C as previously described [69] . All assays used approximately equivalent numbers of males and females . One-way ANOVAs were performed with Tukey's post-hoc analysis to compare behavior , and lifespans were assessed with Log-rank ( Mantel–Cox ) survival tests . Animals were collected and aged 1–2 days at room temperature . Ten fly heads were obtained in triplicate from each genotype and processed as described previously [34] . Blots were incubated with anti-TPI ( 1:5000; rabbit polyclonal FL-249; Santa Cruz Biotechnology ) , anti-ATPalpha ( 1:10 , 000; mouse monoclonal alpha5; Developmental Studies Hybridoma Bank ) , or anti-β tubulin ( 1: 6 , 000; rabbit polyclonal H-235; Santa Cruz Biotechnology ) . Densitometric analyses of the scanned films were performed on unsaturated exposures using ImageJ software available from the National Institutes of Health . A one-way ANOVA was performed to assess variance of TPI levels and data sets were compared using Tukey's post-hoc analysis . The filter-trap dot blot was modified from methods published previously [36] . Animals were aged 1–2 days , collected and homogenized in 1X PBS ( 2 . 7 mM KCl , 137 mM NaCl , 2 mM NaH2PO4 , 10 mM Na2HPO4; pH 7 . 4 ) supplemented with cOmplete mini Protease Inhibitors ( Roche Diagnostics ) , and diluted to 1 μg/μl; wells were loaded as indicated . Samples were diluted 1:2 in 1% SDS , 1X PBS , boiled for 5 min . , and filtered through a cellulose acetate membrane ( Whatman , 0 . 2μm pore ) using a 96-well vacuum dot blot apparatus . Positive controls were collected from PC12 cells transfected with a construct expressing huntingtin “exon1” bearing a stretch of 97 glutamines and C-terminally tagged with EGFP . The membrane was washed four times with 1% SDS-PBS , blocked with Odyssey Blocking Buffer ( LiCor ) , and primary antibodies applied in Odyssey Blocking Buffer . Blots were incubated with anti-TPI ( 1: 5000 ) and anti-GFP ( 1: 5000; rabbit polyclonal; FL; Santa Cruz Biotechnology ) . The membranes were then washed and incubated with the secondary antibody IRDye 800-conjugated goat anti-rabbit ( LiCor ) at 1: 20 , 000 in Odyssey Blocking Buffer . Direct-to-scanner detection and blot visualization were performed using a LiCor Odyssey scanner . Filter-trap experiments were performed twice . Coimmunoprecipitation was performed using the Pierce Co-Immunoprecipitation Kit ( Thermo Scientific ) as per manufacturer’s instructions . Lysates were generated by mechanically homogenizing 50 animals in 0 . 5 ml of IP Lysis buffer ( 25 mM Tris , 150 mM NaCl , 1 mM EDTA , 1% NP-40 , 5% glycerol; pH 7 . 4 ) supplemented with cOmplete mini Protease Inhibitors . After homogenization , lysates were frozen in liquid nitrogen , thawed , then centrifuged twice at 5 , 000 g to pellet exoskeletal debris . Supernatants were collected and diluted to 1 μg/μl , and 400 μg were loaded onto 25 μl of gel pre-coupled with 10 μg of anti-GFP . A negative control was performed using uncoupled gel and dTPI+/dTPIΔcat lysate . Samples were incubated overnight at 4°C and washed ten times with IP Lysis buffer at 4°C . Beads were eluted by boiling with 70 μl of 2X SDS–PAGE sample buffer , separated via SDS-PAGE , immunoblotted , and analyzed as outlined above . Coimmunoprecipitations were performed three times . Images were taken with an Olympus BX51WI fluorescence microscope with Till Photonics Polychrome V monochromator excitation , and Hamamatsu C4742-95 digital camera . Heterozygous TPIT73R larvae were maintained over TM6B , and Tb+ 3rd-instar larvae selected for analysis . Dissection and preparation of larval NMJs were performed as described previously [46] . FM1-43FX dye [Molecular Probes , Invitrogen] loading was performed as previously detailed [47] . Briefly , animals were dissected in ice cold 0 mM Ca2+ HL-3 with 0 . 5 mM EGTA , motor neurons severed , and the preps heated to room temperature or 38°C over the course of 3 min . Bath temperature was monitored throughout the experiments with a microthermal probe to ensure consistency [Fisher Scientific] . Loading experiments were performed with room temperature or 38°C preheated 90 mM KCl , 1 . 5 mM CaCl2 HL-3 supplemented with 4 μM FM1-43FX , and preparations were washed quickly and thoroughly during the experiments to avoid Ca2+ chelation . After loading , preparations were washed with 15 ml of 0mM Ca2+ HL-3 with 0 . 5 mM EGTA at room temperature for 10 min . Unloading experiments were performed as follows: preparations were loaded for 3 min at room temperature; washed with 15 ml of 0mM Ca2+ HL-3 with 0 . 5 mM EGTA at room temperature; imaged; washed with 38°C 0 mM Ca2+ HL-3; heated to 38°C over 3 min . ; unloaded using 38°C 90 mM KCl , 1 . 5 mM CaCl2 HL-3 for 30 sec . ; and the same synapse was imaged again . Preparations were imaged with a water immersion 60X objective , using 450 nm excitation and a 500 nm longpass filter [Chroma Technology] . Simple PCI imaging software was used for acquisition and ImageJ for analysis . Two NMJs from muscles 6/7 were assessed per animal , one from segment A2 and A3 . Six animals were assessed per genotype per time point for a total of 12 NMJs per experimental condition . After acquisition , images were relabeled by an independent researcher and blinded analysis was performed on the raw images . Boutons were outlined and intensity measured , with background subtracted from adjacent tissue . Pair-wise analyses were performed using a two-tailed Student’s t test , while comparisons among multiple experimental conditions were performed using a one-way Analysis of Variance ( ANOVA ) with Tukey’s post-hoc analysis . For NMJ morphological analyses , 3rd-instar larvae were collected and dissected as detailed above without transection of the descending motor neurons . Preparations were fixed in 3 . 5% paraformaldehyde HL-3 , permeabilized with 0 . 1% Triton X-100 in 1X PBS ( PBST ) , and blocked with 0 . 2% BSA in PBST ( PBSTB ) for 2 hrs at room temperature . Preps were washed and incubated with goat anti-HRP [Jackson Laboratories] at 1: 200 in PBSTB for 2 hrs at room temperature . Primary antibodies were removed , washed in PBSTB , and incubated with Cy3-labeled donkey anti-goat in PBSTB at 1: 400 for 1 . 5 hrs at room temperature . Preps were washed , mounted in VectaShield [Vector Laboratories] , and imaged within three days . Images were acquired with an Olympus confocal FV1000 microscope , using a 559 nm excitation laser . Z stacks of segment A2 of muscle 6/7 were taken using 1 μm steps . The Z stacks were merged using Olympus FV1000 Fluoview Viewer , and morphology determined . Ten animals were assessed per genotype , one NMJ per animal , for a total of ten NMJs per experimental condition . Boutons were defined as varicosities at least 2 μm in diameter , and branches defined as extensions containing at least 2 boutons . Images were relabeled by an independent researcher for blinded analysis . Variance within the data set was examined using a one-way ANOVA , with comparisons made using Tukey’s post hoc test . All image quantification was performed on raw image files acquired below saturation . Representative images were selected on the basis of raw image measurements , and post-acquisition processing was performed uniformly with grouped images in parallel using ImageJ; in agreement with published guidelines [94] . Recombinant hTPIΔcat containing the K13M substitution was expressed and purified as previously described [18] using affinity , anion exchange , and size exclusion chromatography . Purified protein was dialyzed into a buffer containing [20 mM Tris pH 8 . 8 , 25 mM NaCl , 2 . 0% glycerol and 1 mM β–mercaptoethanol] , and concentrated to 6 mg/ml prior to crystallization . Crystals of TPIΔcat were obtained using the vapor diffusion method with sitting drops containing 1 μl of protein and 2 μl of well solution [28–34% PEG2000 MME , 50 mM KBr] . Initial crystals grew within 3 days and were improved by successive rounds of microseeding . Crystals were cryoprotected in 40% PEG 2000MME , 20% glycerol , 50 mM KBr , prior to flash freezing in liquid nitrogen . Data collection was performed at the National Synchrotron Light Source at beamline X25 and using a Pilutas 6M detector . Diffraction data was integrated , scaled , and merged using HKL2000 [95] . hTPIΔcat crystals belong to space group P212121 and contain a dimer in the asymmetric unit . Initial phases were estimated for hTPIΔcat via molecular replacement using a previously determined structure of wild-type as our search model [18] . Model bias was reduced through simulated annealing and the model was further improved by manual model building combined with positional and anisotropic B-factor refinement within Phenix [96] . Model quality was validated using MolProbity [97] . Figs were generated using PyMOL ( PyMOL Molecular Graphics System , Schrödinger , LLC ) . Coordinates and structure factors for hTPIΔcat have been deposited within the Protein Databank under accession code 4ZVJ . The effect of hT75R and hG76E substitutions were modeled onto hTPIΔcat or hTPIWT structural templates using the RosettaBackrub analysis as implemented within the RosettaBackrub server [98] . For all predictions , Rosetta version 3 . 1 was used as the algorithm with a backrub radius of 15 Å to ensure that perturbations extending away from the site of the substitution could be sampled . An ensemble of 50 structures was predicted for each TPI model , and their reported Rosetta scores normalized to the starting template . Structures whose Rosetta scores were in the most favorable 10% of the ensemble were used in Fig 9D . Triosephosphate isomerase research has been split between those studying the enzymatic and structural properties of the protein , and those studying its role in disease . Researchers focusing on the pathology of TPI deficiency typically use the abbreviation “TPI” , whereas enzymologists and structural biologists used the abbreviation “TIM” . The aim of this study was to determine the molecular mechanism of a disease mutation , and as such we have used the abbreviation “TPI” . | Glycolysis is the metabolic pathway that cells use to break down the sugar glucose , and mutations in the genes that control the glycolytic pathway elicit a collection of diseases known as glycolytic enzymopathies . Glycolytic enzymopathies are rare genetic diseases that lead to the degeneration of patient red blood cells . Triosephosphate isomerase is a gene that encodes a part of this glycolytic process , and patients with mutations in this gene experience the typical blood disorder , as well as severe neurologic dysfunction and often infant death . Until now , a molecular source of neurologic dysfunction in triosephosphate isomerase mutants was unknown . We have discovered that mutations that disrupt the self-association of the triosephosphate isomerase enzyme lead to neurologic dysfunction in fruit flies . This neurologic dysfunction is characterized by the abnormal cycling of neuronal vesicles that contain neurotransmitters . Given the evolutionary conservation of triosephosphate isomerase and neuronal synaptic function , we believe that these observations represent the dysfunction seen in human patients . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"dimers",
"(chemical",
"physics)",
"invertebrates",
"vesicles",
"classical",
"mechanics",
"crystal",
"structure",
"enzymes",
"isomerases",
"condensed",
"matter",
"physics",
"enzymology",
"mechanical",
"stress",
"animals",
"animal",
"models",
"drosophila",
"melanogaster",
... | 2016 | Structural and Genetic Studies Demonstrate Neurologic Dysfunction in Triosephosphate Isomerase Deficiency Is Associated with Impaired Synaptic Vesicle Dynamics |
Hedgehog ( Hh ) signaling regulates multiple aspects of metazoan development and tissue homeostasis , and is constitutively active in numerous cancers . We identified Ubr3 , an E3 ubiquitin ligase , as a novel , positive regulator of Hh signaling in Drosophila and vertebrates . Hh signaling regulates the Ubr3-mediated poly-ubiquitination and degradation of Cos2 , a central component of Hh signaling . In developing Drosophila eye discs , loss of ubr3 leads to a delayed differentiation of photoreceptors and a reduction in Hh signaling . In zebrafish , loss of Ubr3 causes a decrease in Shh signaling in the developing eyes , somites , and sensory neurons . However , not all tissues that require Hh signaling are affected in zebrafish . Mouse UBR3 poly-ubiquitinates Kif7 , the mammalian homologue of Cos2 . Finally , loss of UBR3 up-regulates Kif7 protein levels and decreases Hh signaling in cultured cells . In summary , our work identifies Ubr3 as a novel , evolutionarily conserved modulator of Hh signaling that boosts Hh in some tissues .
Hedgehog ( Hh ) signaling regulates numerous developmental processes and is implicated in multiple cancers , wound healing and pain sensation in adults [1–3] . The Hh ligand acts as a morphogen to induce differential cell responses based on distinct activity thresholds of its signaling transduction cascade [4–6] . Mis-regulation of Hh signaling affects cell specification and proliferation during development and causes several types of cancer such as glioblastoma or basal cell carcinoma [7 , 8] . In the absence of Hh , the receptor Patched ( Ptc ) inhibits the G-protein coupled receptor Smoothened ( Smo ) [9] . Inhibition of Smo promotes the assembly of an antagonistic molecular complex composed of Costal 2 ( Cos2 ) , a kinesin-related motor protein , Cubitus interruptus ( Ci ) , the key transcriptional effector of Hh [10 , 11] , and several protein kinases [12] . This complex phosphorylates the full length , transcriptionally active form of Ci , Ci155 . Phosphorylated Ci155 is ubiquitinated by a SCF ( Skp1-Cullin1 ( Cul1 ) -F-box ) E3 ligase complex [13] and partially cleaved to generate a transcriptional repressor form , Ci75 , which leads to the transcriptional silencing of Hh target genes [14 , 15] . The Hh signaling cascade is activated by the binding of Hh to Ptc and Ihog ( Interference hedgehog ) [16] , resulting in the release of Smo inhibition . Activated Smo can interact physically with Cos2 [17–20] . This interaction prevents the formation of the Hh signaling antagonistic complex and cleavage of Ci155 . As a result , levels of Ci155 increase in the cytoplasm , promoting its translocation to the nucleus and the transcription of downstream target genes such as decapentaplegic ( dpp ) or ptc ( Fig 1A ) . Previous studies have shown that Cos2 is a key modulator of Hh signaling , and that it facilitates kinase-mediated phosphorylation of Ci and promotes partial degradation of Ci [21] . Loss of Cos2 leads to ectopic activation of Hh signaling and pattern duplications in the Drosophila wing [11] , whereas over-expression of Cos2 inhibits Hh signaling [22] , suggesting that Cos2 is both necessary and sufficient for Hh signaling . In vertebrates , the core components of Hh signaling are conserved , including Cos2 . Cos2 has two vertebrate orthologs , Kif7 and Kif27 [23 , 24] . Kif7 has been proposed to function similarly to Cos2 , because Kif7 knockout mice and zebrafish mutants show an up-regulation of Sonic Hedgehog ( Shh ) signaling [25–27] . In addition , Kif7 can interact physically and modulate the activity of the GLI transcription factors , the mammalian homologs of Ci [27 , 28] . Moreover , Cos2 can functionally replace Kif7 [27] , demonstrating a molecular conservation between vertebrate and invertebrate homologues . In humans , patients carrying KIF7 allelic variants display a spectrum of phenotypic severity ranging from hydrolethalus or Acrocallosal syndromes to Meckel and Joubert syndromes [28 , 29] . Hence , proper function of Kif7 activity is essential for correct Hh signal transduction and is likely to be regulated tightly . Previous studies have shown that Cos2 ( Kif7 ) is phosphorylated by a kinase , Fused , which mediates the strength of differential Hedgehog signaling [30 , 31] . To date , however , no data support a role for ubiquitination in the regulation of Cos2 . Ubiquitination plays an important role in several steps of Hh signaling [32–34] . Ubiquitination is catalyzed by a cascade of enzymes consisting of ubiquitin-activating ( E1 ) , -conjugating ( E2 ) , and –ligating ( E3 ) enzymes [35] . E3 enzymes bind , transfer and ligate ubiquitin to particular substrates . The two major types of E3 ligase are the Really Interesting New Gene ( RING ) domain E3s and the Homologous to E6AP Carboxyl Terminus ( HECT ) domain E3s [36] . We describe the identification and characterization of Ubr3 , a novel regulator of Hh signaling . Ubr3 belongs to the UBR protein superfamily , characterized by a 70-residue zinc finger domain UBR box [37] . Recent studies showed that Ubr3 can polyubiquitinate target proteins [38] involved in multiple biological processes , including olfactory organ function in mice [39] , denticle patterning in Drosophila [40] , DNA damage repair in yeast [38] , apoptosis in flies [41] , homoeostasis in the heart [42] , and breast cancer [43] . Here we show that Ubr3 promotes Hh signaling by mediating the ubiquitination and degradation of Cos2/Kif7 . Loss of Ubr3 elevates the levels of Cos2 , resulting in a decrease in Ci155 and transcriptional silencing of Hh target genes . Loss of ubr3 in flies and zebrafish affects eye development , as well as neuronal specification and somite development in zebrafish . Ubr3 regulates the ubiquitination and degradation of Kif7 in mammalian cells , and transcription of the Shh target ptch2 is strongly decreased in the retina of ubr3 mutant zebrafish . Taken together , our data suggest that Ubr3 is an evolutionarily conserved , positive regulator of Hh signaling that regulates Cos2/Kif7 ubiquitination and degradation .
To identify novel components in developmental signaling pathways , we isolated mutations that affect eye and/or wing morphogenesis in a mosaic forward genetic screen of approximately 6000 X-linked lethal mutations in Drosophila [44–48] . We identified an essential complementation group ubr3 , consisting of two alleles ( ubr3A and ubr3B ) . Both ubr3A and ubr3B hemizygous mutants die as 1st instar larvae . Homozygous mutant clones of both alleles cause delayed differentiation of photoreceptors in the morphogenetic furrow of eye imaginal discs ( Fig 1B and S1A Fig ) . This is revealed by the delayed expression of Senseless , an R8 photoreceptor marker [49 , 50] and Elav ( Embryonic lethal abnormal vision ) , a marker for photoreceptors [51] . Since delayed differentiation of photoreceptors is observed when Hh signaling is lost [52] , we hypothesized that ubr3 mutations may impair Hh signaling . To assess the activation of Hh signaling in ubr3 mutant clones , we examined expression of a Hh reporter , dpp-lacZ [53] and the active form of Ci , CiA . Both dpp-lacZ and Ci155 are lost in ubr3 mutant clones in the morphogenetic furrow ( Fig 1C–1D’ and S1B Fig ) . We and others also noticed an increase in apoptosis in ubr3 mutant cells [41] . To exclude the possibility that the Hh signaling defect in ubr3 mutant cells is due to apoptosis , we over-expressed the anti-apoptotic gene p35 in ubr3 clones . As shown in S1C Fig , the delayed differentiation of photoreceptors is not rescued although apoptosis is suppressed ( S1D and S1E Fig ) . Hence , Hh signaling is impaired in ubr3 mutant cells . Alleles of ubr3 map to a small deficiency that uncovers ~11 genes including ubr3 ( CG42593 ) and l ( 1 ) G0193 ( S1F Fig ) . Both alleles ( ubr3A and ubr3B ) fail to complement the lethality associated with a P-element insertion in ubr3 ( S1F and S1H Fig ) [54] . ubr3B carries a Leu788>STOP and ubr3A carries a Phe949>Leu in ubr3 ( Fig 1E ) . No mutations were found in l ( 1 ) G0193 . A genomic rescue construct rescued the lethality of both ubr3 alleles ( S1F and S1H Fig ) , and over-expression of the ubr3 cDNA in ubr3B mutant clones rescued the loss of Ci155 expression in the morphogenetic furrow ( S1G Fig ) . Together , these data show that ubr3 is required for Hh signaling . ubr3 encodes a 2219 amino acid protein , the Drosophila homolog of the mammalian RING-type E3 ubiquitin ligase n-recognin 3 ( UBR3 ) gene ( Fig 1E ) . Most UBR superfamily member proteins , including UBR1 , UBR2 , UBR4 and UBR5 , function in the N-end rule pathway , a ubiquitin-dependent system where E3 ligases recognize N terminal residues of their targets and degrade them [37] . However , UBR3 does not bind to known N-end rule substrates , suggesting a different molecular function of Ubr3 from N-end rule E3 ligases [55] . Ubr3 contains a UBR moiety , a RING domain and a C-terminal auto-inhibitory ( AI ) domain ( Fig 1E ) [38 , 39] . All three domains are highly conserved among fly , mouse and human ( S1I Fig ) , suggesting that the molecular function of Ubr3 may be conserved . To determine the expression pattern and protein localization of Ubr3 , we raised a polyclonal antibody against a region between UBR domain and RING domain of Ubr3 ( see Materials and Methods ) . The Ubr3 antibody specifically recognized a single 250 kDa band on Western blots of protein extracts from larval eye-brain complexes ( Fig 2A ) . This band became more intense when a Ubr3 transgene was expressed ( Fig 2A ) . Furthermore , immunofluorescent labeling of eye imaginal discs with our Ubr3 antibody revealed that the signal was severely diminished or lost within ubr3B mutant clones ( Fig 2B ) . Ubr3 is cytosolic and broadly expressed ( Fig 2C ) and is enriched in the morphogenetic furrow of developing eye discs ( Fig 2B ) , where Ci155 and dpp-lacZ expression is elevated . The Ubr3 proteins in the cytosol are present in puncta that do not show obvious co-localization with a markers for different organelles ( S2A–S2G Fig ) . These data suggest that elevated levels of Ubr3 positively correlate with the activation of Hh signaling . To assess whether the enriched Ubr3 protein in the morphogenetic furrow ( Fig 2B ) results from increased transcription of ubr3 , we performed in situ hybridization experiments . As shown in Fig 2D , ubr3 was transcribed most abundantly in the morphogenetic furrow , in agreement with the protein enrichment shown in Fig 2B . Over-expression of ubr3 with an eyegone-Gal4 driver ( eyg-Gal4; Fig 2E ) expanded ubr3 expression domain in eye discs ( Fig 2F ) , whereas ubr3 RNAi knockdown decreased expression of ubr3 in the center of the eye disc ( Fig 2G ) , showing the specificity of the RNA probes . We activated Hh signaling in the eyg positive cells by expressing a dominant-negative Ptc ( ptcDN ) [56] or by down-regulating the expression of negative Hh regulators Cos2 or Cul1 by RNAi . In all cases , activation of Hh signaling elevated ubr3 mRNA levels in eye discs ( Fig 2H–2J ) . In contrast , down-regulation of Ci by expressing CiRNAi in the equator region of the morphogenetic furrow through eyg-Gal4 ( arrow in S2H Fig ) resulted in moderate loss of ubr3 transcription ( white arrow in S2I Fig ) . Hence , Hh signaling positively regulates ubr3 expression at both the mRNA and protein levels . To assess whether different levels of Ubr3 proteins contribute in a dosage-dependent manner to Hh signaling , we manipulated the expression levels of Ubr3 in ubr3B/B mutant cells by expressing a ubr3 cDNA at low or high levels . The actin-Gal4 driver used to express the ubr3 cDNA is temperature sensitive and leads to low expression at 18°C and medium to high expression at 25°C [57] . We then assessed Ci155 expression in the mutant clones expressing discrete levels of Ubr3 . Interestingly , when Ubr3 was expressed at low level at 18°C , Ci155 expression was only partially restored ( arrows in Fig 2K ) . However , high level of Ubr3 expression in ubr3 mutant cells fully rescued Ci155 expression . In some cells , Ubr3 over-expression induced ectopic expression of Ci155 posterior to the morphogenetic furrow ( arrowheads in Fig 2L–2L” ) . In summary , these data suggest that Hh activation up-regulates transcription of ubr3 , which in turn promotes Hh signaling . To determine how Ubr3 promotes Hh signaling , we assessed the protein expression of key components of the Hh pathway in ubr3 mutant clones . Expression of Ptc and Fused ( Fu ) , a kinase interacting with Cos2 , was not obviously affected ( S3A–S3B’ Fig ) , but Cos2 ( Fig 3A and 3A’ ) and Cul1 ( S3C and S3C’ Fig ) were up-regulated in ubr3 mutant eye clones . Cos2 up-regulation is obvious in the morphogenetic furrow ( arrows in Fig 3A’ ) , suggesting that Hh regulates the Ubr3-mediated down-regulation of Cos2 . Cos2 and Cul1 are both negative regulators of Hh signaling and loss of function of either gene causes ectopic activation of Hh signaling in eye discs [11 , 13 , 58] . Because both genes are up-regulated in cells lacking Ubr3 , we tested whether over-expression of either gene is sufficient to phenocopy the ubr3 mutation . Over-expression of Cos2 , but not Cul1 , results in loss of Ci155 in the morphogenetic furrow , similar to ubr3 mutants ( Fig 3B–3C’ ) . Labeling with a Cos2 antibody showed that a subtle increase of Cos2 is sufficient to inhibit Ci155 expression ( S3D and S3D’ Fig ) , implicating that Cos2 up-regulation in ubr3 mutant cells is relevant . Hence , up-regulation of Cos2 , but not Cul1 , is likely to be responsible for the Hh signaling defects observed in ubr3 mutants . This hypothesis is supported by the observation that reducing Cos2 protein levels in ubr3 mutant clones through cos2RNAi restored Ci155 levels and suppressed the morphogenetic furrow defects ( arrows in Fig 3D and 3D’ and S3E and S3E’ Fig ) . In contrast , over-expression of cul1RNAi in ubr3 mutant clones did not restore Ci155 expression in the morphogenetic furrow ( arrowheads in Fig 3E and 3E’ ) , suggesting that Cul1 up-regulation was not the cause of Ci155 loss . One likely reason why Ci155 expression is not restored by Cul1 RNAi in ubr3 mutant clones in the morphogenetic furrow is that Cul1 RNAi does not completely remove Cul1 in ubr3 clones and the residual Cul1-Slimb E3 ligase activity may suffice to mediate processing of Ci155 . Moreover , expression of ptcDN in ubr3 mutant clones did not rescue Ci155 loss ( arrowhead in Fig 3F and 3F’ ) . These data show that loss of ubr3 causes a decrease in Hh signaling and a reduction in Ci155 that can be restored by Cos2 down-regulation . Hence , ubr3 acts to attenuate the levels of Cos2 , which enhances the activity of Hh signaling in the morphogenetic furrow . The RING domain of Ubr3 is not a canonical RING domain ( Fig 4A and 4A’ ) [59] . To assess whether Ubr3 has E3 ligase activity , we performed an in vitro ubiquitination assay . Immunoprecipitation-purified Ubr3::GFP fusion proteins were incubated with E1 and E2 enzymes ( HR6A ) [39] and Flag-tagged Ubiquitin ( Flag::Ub ) peptides . Interestingly , Ubr3 poly-ubiquitinated itself , as shown in Fig 4B . Moreover , the UBR domain fragment may form a dimer when over-expressed , because a band of twice the molecular weight of GFP::UBR ( ~80 kDa ) is detected ( Fig 4B ) . Co-immunoprecipitation assays with the over-expressed UBR domain indicated that it interacts with the Ubr3 full-length protein present in whole cell extracts of S2 cells ( Fig 4C ) . This suggests that Ubr3 interacts with the UBR domain of another Ubr3 molecule and that Ubr3 proteins poly-ubiquitinate each other . To test whether the up-regulation of Cos2 in ubr3 mutant cells is due to defective degradation by the proteasome , we performed a degradation assay of Cos2 in Drosophila S2 cells . We found degradation of Cos2 proteins begins 6 hours after treatment with a translational inhibitor cycloheximide ( CHX ) and that the level of Cos2 decreased to 10% after 10 hours of treatment ( Fig 5A ) . Addition of the proteasomal inhibitor MG132 suppressed the degradation of Cos2 ( Fig 5A ) , suggesting that Cos2 proteins are degraded via the proteasome . The degradation of Cos2 is partially suppressed by down-regulation of Ubr3 by Ubr3RNAi and promoted by over-expression of Ubr3 ( Fig 5B ) , suggesting that Ubr3 mediates the degradation of Cos2 . Because ubiquitination is known to regulate protein abundance through proteasome-mediated degradation , Cos2 levels may be regulated via Ubr3-mediated ubiquitination . To determine whether Ubr3 interacts physically with Cos2 and to map which domains are required for this interaction , we performed co-immunoprecipitation assays . As shown in Fig 5C , both the UBR domain fragment and the full length Ubr3 protein interact with Cos2 ( lane 2 and lane 3 ) . To exclude the possibility that Cos2 binds to Ubr3 indirectly via microtubules , we treated S2 cells with the microtubule-destabilizing agent Colchicine . The Cos2-Ubr3 interaction is not affected by Colchicine treatment ( lane 4 in Fig 5C ) , suggesting that Cos2 does not bind Ubr3 via microtubules . To identify which domain of Cos2 is critical for the interaction with Ubr3 , we tested a series of deletion constructs of Cos2 ( Fig 5D ) in co-IP assays with the UBR domain . We found that only the fragments bearing the N-terminal motor domain ( MD ) of Cos2 ( Cos2ΔC1 , ΔC2 , and ΔC3 ) interacted with the UBR domain ( Fig 5E ) . Hence , Ubr3 binds to the N-terminal MD of Cos2 with its UBR domain . To detect the ubiquitination of Cos2 , we performed immunoprecipitation assays and examined the ubiquitination of Cos2 in S2 cells that express ubr3 ( Fig 4C ) . As shown in Fig 5F , the ubiquitinated Myc-tagged Cos2 ( Myc::Cos2 ) was detected by an anti-hemagglutinin ( HA ) antibody in S2 cells co-transfected with an HA-tagged ubiquitin construct ( HA::Ub; Fig 5F , lane 1 , top panel ) . In addition , the HA signal exhibited a lower mobility shift compared to the major band detected by anti-Myc antibody ( Fig 5F ) , indicating that these bands correspond to the ubiquitinated forms of Cos2 . Over-expression of an E3 ligase dead form of Ubr3 , in which the residues required for RING domain activity ( Fig 4A ) were mutated to alanines , did not cause an increase in ubiquitination of Cos2 ( Fig 5G , lane 1–3 ) , suggesting that the E3 ligase activity of Ubr3 mediates the ubiquitination of Cos2 . In addition , removing the Ubr3 binding domain of Cos2 , Cos2ΔN1 , abolished most of the ubiquitination of full length Cos2 ( Fig 5G , lane 4–6 ) . The residual ubiquitination of Cos2ΔN1 may result from endogenous full length Cos2 that co-precipitates with Cos2ΔN1 through dimerization [10 , 11] . To determine whether Ubr3 regulates Cos2 ubiquitination , we examined the levels of Cos2 ubiquitination when Ubr3 was either over-expressed or knocked down by RNAi . As shown in Fig 5F , the co-expression of Ubr3 with Cos2 increased Cos2 ubiquitination , whereas inactivation of Ubr3 by RNAi decreased ubiquitination ( lane 2 and lane 3 , top panel ) . A control GFP RNAi ( negative control ) did not significantly change the level of Cos2 ubiquitination ( lane 4 , top panel ) . These results suggest that Ubr3 interacts with and ubiquitinates Cos2 . We next tested whether Hh signaling regulates the ubiquitination of Cos2 . Interestingly , we found that the ubiquitination of Cos2 was strongly enhanced by Hh treatment ( Fig 6A , lane 2 ) . This increased ubiquitination was abolished by down-regulation of Ubr3 ( Fig 6A , lane 4 ) , suggesting that Ubr3 mediates Hh induced ubiquitination of Cos2 . This implied that Ubr3-mediated ubiquitination of Cos2 was tightly controlled by Hh signaling . Because Ci is not expressed in S2 cells , Hh-induced ubiquitination of Cos2 cannot be mediated by a positive , transcriptional feedback loop that depends on Ci . We therefore tested whether Hh may promote binding of Ubr3 to Cos2 . We performed co-IP assays between the Ubr3 and Cos2 in the presence or absence of Hh . As shown in Fig 6B , the interactions between Ubr3 full-length protein and Cos2 ( lanes 1 and 2 in Fig 6B ) were strongly increased by Hh . These data show that Hh induces the ubiquitination of Cos2 by promoting the association of Ubr3 with Cos2 . Consistent with Hh-induced poly-ubiquitination of Cos2 , we also observed a faster degradation of Cos2 upon Hh treatment ( Fig 6C and 6C’ ) . The ladder pattern of the HA signal in Fig 5E suggests that Cos2 is poly-ubiquitinated . We further determined the ubiquitination chain pattern by using a panel of ubiquitin mutant constructs [34] . Compared to wild-type ubiquitin ( Fig 6D , lane 1 , top panel ) , a mutated lysine 48 in ubiquitin ( HA::UbK48R ) abolished the formation of the ubiquitin chain ( lane 4 , top panel ) , whereas altered lysine 11 ( K11R ) , lysine 29 ( K29R ) , or lysine 63 ( K63R ) did not affect chain formation . In addition , mutating all of the lysine residues except lysine 48 ( HA::UbK48 only ) leads to longer ubiquitination chains ( Fig 6D , lane 6 , top panel ) . The single sharp band of Cos2 ubiquitination by K48R indicates a mono-ubiquitinated Cos2 that cannot be further elongated due to the lack of K48 . Together , these data indicate that Cos2 undergoes K48-linked poly-ubiquitination . To determine whether Ubr3 plays a conserved function in vertebrates , we created two independent zebrafish ubr3 mutant alleles using CRISPR/Cas9 . The ubr3 gene is predicted to encode a protein of 1808 amino acids , and the ubr3b1250 allele lacks 28 nucleotides ( Del 378–405 ) downstream of the predicted ATG ( S4A Fig ) leading to a frameshift and early stop codon . The mutant protein should encode only 129 amino acids ( S4C Fig ) , lacking the UBR and RING domains . The second allele , ubr3b1251 carries a 4 nt insertion at position 220 ( S4B Fig ) , also causing a frameshift and early stop codon . ubr3b1251 is predicted to encode a 78 aa protein lacking all functional domains ( S4C Fig ) . Using an anti-Ubr3 antibody , we detected expression of Ubr3 in the developing retina , central nervous system and trunk , which are lost in ubr3b1250/b1251 mutant zebrafish ( Fig 7A and 7B ) . Three independent crosses between single carriers heterozygous for the b1250 and b1251 alleles resulted in progeny with a distinguishable and reproducible retinal phenotype in a Mendelian frequency ( f = 0 . 22 , f = 0 . 27 , f = 0 . 23 , n = 270 ) . At the 5-6-somites stage , phenotypically wild-type siblings display optic vesicles characterized by a compacted and stratified epithelium ( Fig 7C , 7E and 7G ) . The optic vesicles of the ubr3 trans-heterozygous mutants failed to form a cohesive and stratified epithelium ( Fig 7D , 7F and 7H ) . Because appropriate levels of Sonic Hedgehog ( Shh ) signaling are essential for eye morphogenesis [60 , 61] , we examined the transcriptional levels of ptch2 . In zebrafish , ptch2 is a direct target of Shh signaling [62 , 63] . In wild-type embryos , a gradient of ptch2 expression was observed within the optic vesicle ( dotted area in Fig 7G and 7I ) . This gradient was characterized by high levels of ptch2-expressing cells localized in the ventral border of the vesicle , and low level expressing cells localized in the dorsal border region and vesicle core ( Fig 7I ) . In ubr3 mutants , ptch2 expression was strongly decreased ( Fig 7H and 7J ) . Consistent with decreased Shh signaling , ubr3b1250/b1251 trans-heterozygous mutants show a 30% increase in the angle of the somite in comparison with phenotypically wild-type siblings ( S5A , S5C and S5E Fig ) . The opening of the somite angle is a common morphological phenotype of mutants with reduced Hedgehog signaling [64–67] . In addition , we observed a gain of Rohon-Beard sensory neurons at the level of the posterior central nervous system ( CNS ) in ubr3 mutants ( S5B and S5D Fig ) , assessed by expression of a Rohon-Beard sensory neuron marker islet2 [68] . Because Hedgehog restricts CNS dorsal fate acquisition [69] , this result supports the interpretation that Hedgehog signaling is decreased in ubr3 mutants . This finding is also consistent with our observation of decreased retinal ptch2 expression in the absence of ubr3 ( Fig 7H and 7J ) . Because Kif7-depleted zebrafish embryos do not show de-repression of Hh target genes in the CNS [27] , our findings further suggest that , at least in zebrafish , Ubr3 may regulate not only Kif7 but also other intracellular negative regulators of Hedgehog signaling in the CNS . Different zebrafish Hh signaling mutants show distinct degrees of severity , highlighting the tissue-specific requirements of Hh levels during development [60 , 70–74] . Similarly , loss of ubr3 does not result in cyclopia or inner ear defects , showing that these mutants have a less severe phenotype when compared to smoothened mutant animals . Hence , ubr3 zebrafish mutants retain some residual Hh signaling . Thus , our data show that Ubr3 positively regulates Hedgehog signaling in tissues sensitive to high levels of Hh like the mesoderm and neuroectoderm . In addition , the transcription of ubr3 is strongly reduced in smohi1640-/- mutant animals [70] , which lose Shh activity ( S5F and S5H Fig ) . In contrast , ectopic activation of the Shh pathway by injection of the mRNA encoding a dominant negative form of PKA ( dnPKA ) [75] expands the expression domain of Ubr3 ( S5G and S5I Fig ) . These data suggest that Shh signaling promotes the transcription of ubr3 in zebrafish , similar to what we observed in Drosophila . In summary , Ubr3 is required for the transduction of Hh signaling and proper eye morphogenesis in zebrafish . To test whether UBR3 also plays a role in Shh signaling in mammals , we used C3H10T1/2 mouse mesenchymal cells . These cells respond to Shh and activate Shh target genes [76] . We first confirmed that Ubr3 is expressed in C3H10T1/2 cells by RT-PCR ( see Fig 8B’ ) . We then infected these cells with a lentivirus bearing 7 tandem binding sites for Gli ( the vertebrate homologue of Ci ) that control the expression of a GFP reporter . Addition of either the Shh ligand or purmorphamine , an agonist of Smo [77] , to C3H10T1/2 cells induced GFP expression in about 25% of the cells ( Fig 8A and 8A’ ) . To determine whether knockdown of UBR3 impairs Shh signaling , we measured the proportion of GFP-expressing C3H10T1/2 cells transfected with one of four different siRNAs against UBR3 or a scrambled siRNA control , followed by purmorphamine treatment . Induction of the Gli::GFP reporter by purmorphamine was suppressed when siRNA reduced the UBR3 levels ( Fig 8B ) , as judged by real time PCR ( Fig 8B’ ) . In addition , down-regulation of UBR3 resulted in up-regulation of Kif7 ( Fig 8C ) , the mammalian homolog of Cos2 . To assess poly-ubiquitination of Kif7 , we purified Kif7 through immunoprecipitation and loaded the Western blot lanes with equal amounts of protein ( unlike in Fig 8C where we loaded equal amounts of cells ) . We observed decreased poly-ubiquitination of Kif7 upon knockdown of UBR3 ( Fig 8D ) . These data indicate that UBR3 regulates Shh signaling through poly-ubiquitination of Kif7 in vertebrate cells , a process that seems to be evolutionarily conserved .
Numerous studies have shown that Cos2 plays a central role in Hh signaling [10 , 11 , 22 , 25 , 26 , 78 , 79] . Cos2 is both necessary and sufficient to regulate Ci [11 , 22] and the level of Cos2 protein is critical for activating Hh signaling [80 , 81] . Here , we identified Ubr3 as a novel regulator of Cos2 in a forward genetic screen in Drosophila and showed that this gene is conserved in vertebrates and affects Hh signaling . We present evidence that the level of Cos2 protein is tightly controlled through a Ubr3-mediated poly-ubiquitination pathway ( Fig 8E ) . Although most of the core components of Hh signaling are evolutionarily conserved , there are differences in Hh signaling between vertebrates and invertebrates [82] . For example , Cos2 can be phosphorylated by the kinase Fused [30 , 31] , but the kinase that phosphorylates Kif7 remains to be identified , because mice lacking Fused have no apparent defects in Hh signaling [83 , 84] . Given that a Kif7 phosphatase affects Hh signaling in vertebrates [85] it is likely that phosphorylation of Kif7 is important even if these sites are different than those observed in Cos2 . Here , we present the first evidence that the levels of Cos2 and Kif7 proteins are also controlled by poly-ubiquitination via a conserved Ubr3 E3 ligase . The conservation of this mechanism is supported by the finding that fly Cos2 rescues the Kif7 mutant phenotypes in zebrafish [27] . Although we have shown that the degradation of Cos2 protein is regulated by Ubr3 mediated ubiquitination , the increased Cos2 proteins in ubr3 mutant cells may also result from up-regulated transcription of Cos2 . Although Hh promotes poly-ubiquitination and degradation of Cos2 ( Fig 6A and 6C ) , we did not observe a decrease of Cos2 proteins at the morphogenetic furrow , where Hh signaling is activated . Instead , the level of the Cos2 protein is modestly elevated when compared to surrounding tissues/cells ( S6A and S6A’ Fig ) , consistent with a previous finding [11] . We also observe that activation of Hh in S2 cells up-regulates Cos2 ( Fig 6C ) . The observation that activation of Hh signaling promotes the degradation of Cos2 and that Cos2 protein level is increased , but not decreased by Hh activation , suggests that some mechanism other than ubiquitination up-regulates the level of Cos2 protein ( Fig 8F ) . Ubr3-mediated degradation of Cos2 may function as a mechanism to prevent aberrantly high levels of Cos2 , thereby toning down Hh signaling . This may also underlie the observation that not all cells that require Hh signaling are affected in flies . This hypothesis is also supported by the finding that loss of ubr3 in zebrafish affects developmental processes that rely on high levels of Shh signaling but does not affect those that respond to low Shh signaling ( Fig 7 and S4 Fig ) . Cul1 functions downstream of Cos2 to process Ci155 , one would anticipate that Cul1 is epistatic to Cos2 . This is inconsistent with the observation that down-regulation of Cul1 in ubr3 clones in the morphogenetic furrow of Drosophila eye discs fails to restore Ci155 expression ( Fig 3E ) , whereas down-regulation of Cos2 restores Ci155 levels ( Fig 3D ) . This may be because the RNAi expression does not deplete the protein sufficiently , or because , Cos2 may regulate Ci155 through a mechanism independent of Cul1 . Although our data clearly show that Ubr3 plays a role in Hh signaling at the morphogenetic furrow , we do not observe a loss of Ci155 in ubr3 mutant clones in wing discs ( S6B and S6B’ Fig ) . However , we observed a similar up-regulation of Cos2 in ubr3 mutant clones in wing discs ( S6C and S6C’ Fig ) , implying that Ubr3 mediated poly-ubiquitination of Cos2 may be present in wing discs . The lack of a Hh phenotype in posterior compartment cells of wing discs may be due to another E3 ligase that is functionally redundant and downregulates Cos2 . Alternatively residual Ubr3 in ubr3 mutant cells due to perdurance of Ubr3 products may partially downregulate Cos2 , allowing activation of Hh signaling . When we sensitized the background by over-expressing ptcDN to ectopically activate Hh signaling , we find that loss of ubr3 strongly suppresses the activation of Hh signaling in clones , gauged by the reduced clone sizes and Ci155 levels ( S6D–S6E’ Fig ) . This may also be the reason why not all tissues display the typical Shh phenotype in zebrafish . In addition , ectopic activation of Hh signaling leads to up-regulated transcription of ubr3 ( S6F–S6J Fig ) , suggesting that the positive feedback of Ubr3 is present in the wing . Hh signaling shares many similarities with Wnt signaling [86] . Both pathways regulate many developmental processes and induce human cancers when the pathways are aberrantly activated . Moreover , the principal signaling mechanisms are based on similar features . Each pathway is activated through ligand binding of a G-protein coupled receptor , leading to the downstream activation of a transcription factor through phosphorylation-dependent proteolysis . Axin is the scaffold protein that recruits an activation complex in Wnt signaling , which mediates phosphorylation of β-catenin [87] . This function is similar to that of Cos2 in Hh signaling . Interestingly , previous studies have shown that the levels of Axin protein are also regulated by an E3 ligase , RNF146 , through poly-ubiquitination [88–90] . Upon activation of Wnt signaling , Axin undergoes tankyrase-dependent poly ADP-ribosylation , which promotes RNF146-Axin interaction [89] . Ubr3 seems to regulate the poly-ubiquitination of Cos2 in a similar manner , given that Hh activation promotes the Ubr3-Cos2 interaction and the ubiquitination of Cos2 . Hence , our data suggest further similarities between the Hh and Wnt signaling pathways .
ubr3A and ubr3B mutants were isolated in a forward genetic screen as previously described [45 , 48] . y w ubr3A FRT19A/FM7c Kr-Gal4 , UAS-GFP and y w ubr3B FRT19A/FM7c Kr-Gal4 , UAS-GFP flies were crossed to , y w tub-Gal80 , eyFLP , FRT19A; actin-Gal4 , UAS-CD8::GFP/CyO and y w UbxFLP , tub-Gal80 FRT19A; UAS-CD8::GFP , actin-Gal4 to generate GFP-labeled ubr3 homozygous mutant clones using the MARCM technique [91] . The ubr3 genomic rescue transgenic fly strain was generated using the P[acman] system , BAC recombineering and transgenic platform developed in our laboratory [92] . ubr3 cDNA transgenic flies were generated through φC31-mediated transgenesis [92] . Additional strains used in the study are as follows: dpp-lacZ [93] , Df ( 1 ) BSC622 [[94] , Bloomington Drosophila Stock Center] , P[lacW]CG42593G0307a [[54] , Bloomington Drosophila Stock Center] cul1EX , FRT42D/CyO [58] , FRT42D/CyO [95] , UAS-p35 ( a kind gift from Andreas Bergmann ) , eyg-Gal4 [96] , UAS-cos2/CyO [81] , UAS-cul1/CyO [58] , UAS-ubr3RNAi [P{GD12698} , [97] , Vienna Drosophila Resource Center] UAS-cos2RNAi [[97]; Vienna Drosophila Resource center] , UAS-cul1RNAi [TRiP . HM05197 , [98]; Bloomington Drosophila Stock Center]; UAS-CiRNAi [TRiP . JF01272 , [98]; Bloomington Drosophila Stock Center]UAS-ptcDN ( a kind gift from Michael Galko ) . All flies were maintained on standard food at 25°C . Zebrafish strains were AB wild-type , ubr3b1250 , ubr3b1251 and smohi1640 . The ubr3 mutations are recessive alleles . Phenotypically wild-type siblings were used as controls and labeled as wild type in Fig 7 . Animals were raised in a 10 hour dark and 14 hour light cycle and maintained as previously described [99] . Embryos were staged according to the standard series [100] . All animal use protocols were IACUC-approved . CRISPR mutagenesis was carried as previously described with minor modifications [101 , 102] . The zebrafish ubr3 reference sequence used in this study was XM_009304449 . 1 . Identification of target sequences was done using Zifit software [103 , 104] . Candidate sequences were then blasted against the zebrafish genome ( Zv9 ) and those with unique hits were selected . The following target sequences were selected b1250: 5’- GGGGCCTGTGACTGCGGGGA-3’ , located in the sense strand , and b1251: 5’-GGCGTTATCGTAGGATCGGA3’ , located in the antisense strand ( Fig 7 , S1 Fig ) . A guide RNA ( gRNA ) template was created by PCR . A T7 promoter site was incorporated in the gene specific oligonucleotide , followed by the target sequence and the start of the guide RNA sequence ( 5’-gttttagagctagaaatagc-3’ ) . The complementary guide RNA scaffold oligonucleotide sequence used was 5’-gatccgcaccgactcggtgccactttttcaagttgataacggactagccttattttaacttgctatttctagctctaaaac-3’ . PCR was performed using Phusion polymerase ( NEB ) following the manufacturer’s recommendations . 10μM of each primer was used for the reaction . The first denaturation step was carried out at 98°C for 30 sec , followed by 40 cycles of denaturation at 98°C for 10 seconds , annealing at 60°C for 10 seconds , and extension at 72°C for 15 seconds . A final extension step was introduced at 72°C for 10 minutes . PCR products were purified using a PCR purification kit ( QIAGEN ) . RNA was transcribed using a MEGAscript T7 kit following the manufacturer recommendations . A volume of 2nl of Cas9 RNA and gRNA were co-injected at a concentration of 100ng/μl each . Screening of F0 founders and genotyping of F1 carriers were done by PCR and sequencing using the following primers: Primer b1250F ( position 101–119 ) : 5’-CTGCAGGAACTGCTGGATAG-3’; Primer b1250R ( position 415–433 ) : 5’-ACCCGCTCTCTCTCATCAC-3’ . Primer b1251F ( position75-94 ) : 5’-TGACAACAGTTCAGGCTTGC-3’; Primer b1251R ( position326-345 ) : 5’-GTGGCGTTATCGTAGGATCG-3’ . 250 pg of RNA encoding for a dominant negative regulatory subunit of the Protein Kinase A ( dnPKA ) [75] was injected into 1 cell stage embryos . dnPKA construct was linearized with NotI and transcribed with SP6 using a mMessage mMachine kit following the manufacturer´s recommendations . Embryos were fixed at 27hpf and processed for in situ hybridization against ubr3 . Fly tissues were dissected in phosphate-buffered saline ( PBS ) at room temperature and fixed with 3 . 7% formaldehyde in PBS for 20 minutes , followed by permeabilization with 0 . 2% Triton-X100 in PBS . The primary antibodies and secondary fluorescently-labeled antibodies used were: chicken anti-GFP ( 1:1000 , Abcam ) , rat-Elav [1:1000 , 7E8A10 , DSHB , [51] , guinea pig anti-Sens [1:1000 , [50]] , rat anti-Ci [1:50 , 2A1 , DSHB [105]] , rabbit anti- β-galactosidase ( lacZ; 1:1000 , Abcam ) , guinea pig anti-Ubr3 ( 1:1000 , this study , see below ) , mouse anti-Cos2 [1:50 , 17E11 , DSHB [18]] , mouse anti-Ptc [1:100 , DSHB [106]] , mouse anti-Fu [1:100 , DSHB [18]] , rabbit anti-Cul1 [1:250 , [107]] , rabbit anti-GM130 ( 1:500 , Abcam ) , rabbit anti-Rab5 ( 1:500 , Abcam ) , rabbit anti-Rab7 ( 1:500 ) [108] , mouse anti-Rab11 ( 1:100 , BD Biosciences ) [109] , mouse anti-Complex V ( 1:500 ) [110] , ER-GFP ( 1:1000 incubate with cells over night , CellLight ER-GFP , BacMam 2 . 0 . Thermo Fisher Scientific ) , PNA-biotin ( Vector Laboratories ) . Alexa488- , Cy3- and Cy5- or DyLight649 conjugated affinity purified donkey secondary antibodies ( 1: 500 , Jackson ImmunoResearch Laboratories ) and DAPI ( 0 . 5 μg/ml , Life Technologies ) . Zebrafish immunolabeling was performed as previously described [111] with the following minor modifications . 18-somites stage embryos were fixed in BT-fix overnight at room temperature . Embryos were permeabilized in PBS+1% Tween20 for 5 hours at room temperature . Anti-UBR3 antibody ( Sigma Prestige , catalogue #HPA035390 ) was diluted in 1/500 . Biotinylated anti-rabbit was used at 1/500 . To detect signal , ABC kit ( Vectorlabs ) was used . A and B reagents were mixed together at an 1/100 dilution in PBS- Block and pre-incubated for 20 minutes at room temperature , then added to the samples for 25 minutes . Tyramide from the TSA kit ( Perkin-Elmer ) was diluted 1:50 in pre-warmed buffer reagent , and added to samples for 20 minutes following the manufacturer’s recommendations . The detection reaction was stopped by adding cold PBS+0 . 1%Tween20 , followed by 4 washes in PBS+0 . 1%Tween20 . Images were acquired using LSM510 and LSM710 confocal microscopes ( Zeiss ) and examined and processed using LSM viewer ( Zeiss ) , ZEN ( Zeiss ) and Photoshop ( Adobe ) software . Immunostained zebrafish embryos were immersed in Vectashield , mounted laterally in a slide chamber and imaged with a Zeiss LSM5 confocal microscope . Live embryos were mounted laterally in 3% methylcellulsose and imaged on a compound microscope using DIC . ISH treated embryos were dissected in 90% glycerol , flat-mounted in 100% glycerol in a slide chamber and imaged on a compound microscope using DIC . A ubr3 genomic rescue construct was constructed by cloning a 18 . 3 kb fragment of genomic DNA that contains the ubr3 gene ( X: 7 , 935 , 666… 7 , 953 , 967 ) [Release 6 Drosophila reference genome , [112]] into P[acman] [92 , 113] . ubr3 cDNA was constructed from exon sequences and cloned into pUASTattB using a GENEART Seamless Cloning and Assembly Kit ( Life Technologies ) . GFP was tagged to the carboxyl terminus of the full length ubr3 sequence or a partial sequence encoding only the UBR domain ( aa 222–292 ) . The flag sequence was conjugated to the carboxyl terminus of the full length ubr3 sequence in the primer . Ubr3::flag was then amplified through PCR and cloned into pUASTattB through XhoI and XbaI . To generate E3 dead form of flag tagged Ubr3 expression construct , mutations results in all residues shown in red box in Fig 4A changed to alanines were introduced through synthesized DNA which spans 500 bp downstream from RsrII . This synthesized DNA fragment was then cloned together with PCR amplified flag tagged carboxyl fragment of Ubr3 into pUASTattB-Ubr3-flag through RsrII , BbsI and XbaI . The HA::Cos2N1 to 3 constructs were cloned into pUASTattB through EcoRI and XbaI . HA::Cos2ΔN1-2 , HA::Cos2ΔC1-3 have been described [114] . Myc::Cos2 was constructed by fusion of 5xMyc tags to the N-terminus of the Cos2 coding sequence . The HA::Ub transgene has been described previously [34] . Kif7::GFP construct is a gift from Dr . Chi-Chung Hui [78] . The 7Gli:GFP reporter of Hedgehog signaling activity contains 7 repeats of the Gli binding site ( 5’-TCGACAAGCAGGGAACACCCAAGTAGAAGCTC ) followed by GFP [115] . The primer pair ( forward 5’-TGAAGCTTGCATGCCCTGCAGGACAAGCAGGGAACGCCCAAGTAG and reverse 5’ CTCGAGTACCGGATCCATTATATACCCTCTGCAGACTTGGGTGTTCCCTGCTTGTCG ) was used to amplify the Gli binding sequences from the 8Gli-Luc plasmid by PCR . The reverse primer also contains a TATA sequence , which was used to rebuild the TATA box after the Gli binding sites . The destination plasmid pRRL . sin-18 . ppt . TCF/LEF:GFP . pre [116] was linearized by PstI and BamHI digestion to cut out the TCF/LEF sequence and the TATA box . The resulting products were recombined into a linearized destination plasmid by infusion cloning according to the manufacturer’s protocol ( Clontech ) . For Ubr3 antibody production , the sequence encoding aa 751–1500 of Ubr3 was cloned into pET21 expression construct and expressed in E . coli . Purified inclusion bodies were used to immunize guinea pigs . ubr3 in situ hybridization probes I and II anti-sense sequences contain 2558 to 3576 nt and 3775 to 4770 nt of ubr3 cDNA , respectively . Anti-sense sequences were cloned into pGEM-T vector ( Promega ) . Before transcription , the construct was linearized by SalI . ubr3 RNA in situ probes were transcribed and labeled with a digoxigenin [113] RNA labeling kit ( Roche ) . In situ hybridization to whole-mount discs was performed as previously described [117] . Probe I was used in images shown in Fig 2F , 2G , 2H and 2I and S6J Fig . Probe II was used in images shown in Fig 2E and 2J and S2I , S6F , S6H and S6I Figs . Zebrafish whole-mount in situ hybridization was carried out as described with minor modifications [118] . Digoxigenin-labeled probes were prepared according to manufacturer´s instructions ( Roche ) . Probe signal was detected using NBT/BCIP mix ( Roche ) . The ptch2 probe was kindly shared by Stone Elworthy ( University of Sheffield , UK ) . S2 cells were cultured at 25°C in Schneider’s medium ( Life Technologies ) plus 10% heat-inactivated fetal bovine serum ( Sigma ) , 100 U/mL penicillin ( Life Technologies ) , and 100 μg/mL streptomycin ( Life Technologies ) . Cells were split every 3 days and plated at a density of 106 cells/well in 12-well cell culture plates for experiments . Transfections were carried out using Effectene transfection reagent ( Qiagen ) . Ubr3 dsRNA was synthesized against nt 652–1 , 191 . dsRNAs transfection and dsRNA against GFP have been described [114] . CHX ( 100 μM , Sigma ) and MG132 ( 50 μM , Sigma ) in dimethyl sulfoxide ( DMSO ) were added to S2 cells 48 hours after transfection and incubated for indicated time . An equal amount of DMSO was added as a negative control ( - ) . 1ug/ml Colchicine ( sigma ) was incubated with cells for 5 hours before harvest . C3H10T1/2 cells were cultured at 37°C in 5% CO2 in air in Eagle's Basal medium with Earle's BSS , 2 mM L-glutamine , 1 . 5 g/L sodium bicarbonate and 10% fetal bovine serum , as described by the ATCC ( http://www . atcc . org/ ) . Cells were split when reaching 80–90% confluence . Cells were transduced with lentivirus containing Gli::GFP reporter construct for 16h and then plated on 12-well cell culture plates . On the second day , siRNAs were incubated with Lipofectamine RNAiMAX reagent ( Invitrogen ) overnight . The sequences of siRNAs against ubr3 are as follows: siRNA 1: GTTATAGCTTTGAATCAGT; siRNA 2: CAGAGTTTGCCTCACGACA; siRNA 3: CAAGATTGGTTTGATGCTA; siRNA 4: CAGAAATTGCTCGCAGAGT . Stealth RNAi™ siRNA Negative Control Med GC Duplex ( Invitrogen ) was used as control siRNA . 3μg/ml Shh ( R&D systems ) or 10 μM purmorphamine ( Calbiochem ) in culture medium , or the same amount of vehicle in culture medium for uninduced controls , was added to cells on the third day . Cells were photographed for GFP fluorescence and harvested 48 h after purmorphamine induction . The number of GFP-positive cells was manually counted and statistical testing was performed with a one way ANOVA followed by Dunnett’s test using uninduced cells as a control . For RNA extraction , total RNA from C3H10T1/2 cells was isolated by using Absolutely RNA miniprep Kit ( Agilent Technologies ) . cDNA was synthesized using Superscript III First Strand Synthesis System for RT-PCR ( Invitrogen ) . Quantitative real-time PCR ( qPCR ) was conducted with a Master SYBR Green kit ( Applied Biosystems ) and gene-specific primer sets on a Step One Plus real-time PCR system ( Applied Biosystems ) . Each experiment was performed with three biological sample repeats and each PCR was performed in triplicate . L19 was used as an endogenous reference . The gene-specific primer sets used were as follows: L19 ( RpL19 ) : 5’-GGTCTGGTTGGATCCCAATG-3’ and 5’-CCCGGGAATGGACAGTCA-3’; UBR3: 5’-CTGATTCATAGAGGAGGCAG-3’ and 5’-ATGGAACAGCTGATTCAGAC-3’ . S2 cells were lysed 48 h after transfection with plasmids in lysis buffer ( Tris-HCl 25mM , pH 7 . 5 , NaCl 150 mM , EDTA 1mM , NP-40 1% , Glycerol 5% , DTT 1mM ) plus Complete proteinase inhibitor ( Roche ) for 30 minutes on ice , followed by centrifugation . In these experiments to detect the ubiquitination of Cos2 ( Figs 5F , 5G and 6A ) , we treated S2 cells with 50 uM of MG132 24 h before harvesting the cells . The supernatant was then immunoprecipitated with agarose beads conjugated to antibodies recognizing different epitope tags , which had been previously equilibrated with lysis buffer , overnight at 4°C . The beads were then washed 3 times in washing buffer ( Tris-HCl 10mM , pH 7 . 5 , NaCl 150mM , EDTA 0 . 5 mM ) before boiling in loading buffer . Western blotting was then performed with each sample . The following beads were used for immunoprecipitation: Chromotek-GFP-Trap Agarose Beads ( Allele Biotechnology ) , Monoclonal Anti-HA−Agarose antibody ( Sigma ) . Protein A resin and anti-Myc ( 9E10 , Santa Cruz ) were used for Myc immunoprecipitation . To examine the levels of Cos2 ubiquitination , a denaturing method was used as previously described [34] . Briefly , S2 cells were transfected with Myc::Cos2 and then lysed with denaturing buffer ( 1% SDS , 50mM Tris , pH 7 . 5 , 0 . 5 mM EDTA , and 1 mM DTT ) and incubated at 100°C for 5 min . The lysates were then diluted 10-fold with regular lysis buffer containing 1 . 5 mM MgCl2 and subjected to immunoprecipitation with the anti-Myc antibody . The proteins were then resolved on an 8% SDS-PAGE , and an immunoblot was performed using an anti-HA antibody to detect the HA::Ub or HA::Ub mutants . The antibodies used in Western blot analysis are as follows: anti-GFP ( 1:1000 Zymed or 1:1000 , Millipore ) , anti-Myc ( 1:5000 , 9E10 , Santa Cruz ) , anti-HA ( 1:5000 , Santa Cruz , F7 or 1:1000 , 16B12 , Covance ) , anti-Ubr3 ( 1:5000 ) , anti-actin ( 1:5000 , C4 , MP Biomedicals ) , anti-α-tub ( 1:1000 , Cell Signaling ) . The intensities of the bands in Fig 5B were quantified using image J software . In vitro auto-ubiquitination assays were performed as described previously [119] with modifications . In brief , S2 cells were first transiently transfected with GFP , UBR-GFP or Ubr3-GFP . S2 cell cultures were collected 48 hours post-transfection and lysed on ice for 45 minutes under stringent conditions to minimize interactions with other proteins , using 100 μl RIPA buffer ( 150 mM NaCl , 1 . 0% NP-40 , 0 . 5% sodium deoxycholate , 0 . 1% SDS , 50 mM Tris , pH 8 . 0 ) containing 1x Complete protease inhibitors cocktail ( Roche ) for every 106 cells seeded . The lysates were then added to 30 μl bed volume of Chromotek GFP Trap beads ( Allele Biotechnology ) , previously equilibrated with RIPA buffer , and incubated by rocking at 4°C for 3 hours . After washing with RIPA buffer , 20% of the beads were retained for assessing the expression of GFP protein by Western blot analysis using anti-GFP ( 1:1000 , Zymed ) . The remainder of the lysates were equilibrated by rinsing twice in 1x Ubiquitination Reaction buffer ( 50 mM Tris-HCl , pH 7 . 4 , 5 mM MgCl2 , 50 mM NaCl , 1 mM dithiothreitol DTT , 1x protease inhibitors cocktail ) . The ubiquitination reaction was assembled by adding rabbit UBE1 E1 ( Boston Biochem Cat . #302 ) and human recombinant His6-hHR6A E2 ( Boston Biochem Cat . E2-612 ) conjugating enzymes and FLAG-Ubiquitin ( Sigma ) on ice and incubated at 30°C for 30 minutes . The reactions were stopped by adding 1x Laemmli buffer , after which the samples were boiled for 10 minutes and analyzed by SDS-PAGE and Western blot using anti-FLAG M2 monoclonal antibody ( 1:1000 , Sigma ) . For Cos2 ubiquitination assays , S2 cell culture and RNAi were performed as described previously [114] . Transfections were carried out using Effectene transfection reagent ( Qiagen ) . The immunoprecipitation and immunoblot analysis were performed using standard protocols . Myc-Cos2 was constructed by fusion of 5xMyc tag to the N-terminus of Cos2 coding sequence . HA-Ub and Ub mutants have been described [34] . The HA::UbK48 only has mutations at all of the lysine residues with the exception of K48 . GFPRNAi has been described . Ubr3 dsRNA was synthesized against nucleotides 652–1191 . The following antibodies were used: mouse anti-Myc ( 1:5000 , 9E10 , Santa Cruz ) , anti-GFP ( 1:1000 , Millipore ) , anti-HA ( 1:5000 , F7 , Santa Cruz ) , and anti-β-tubulin ( 1:2000 , E7 , DSHB ) . | Hedgehog signaling regulates many important biological processes and has been linked to developmental disorders , wound healing , and cancer . Although the major components in the pathway have been well studied in Drosophila and vertebrates , how the signaling is regulated by different modulators is not well understood . Here , we take advantage of a fly forward genetic screen to isolate Ubr3 . We show that it is a novel modulator that regulates Hh signaling . Loss of ubr3 leads to Hh signaling defects in developing eyes of Drosophila , and affects eye , and somite and sensory neuron development in zebrafish . However , Hh signaling is not affected in all cells known to be dependent on Hh signaling as loss of ubr3 in the fly wing and zebrafish inner ear are not affected . This suggests that Ubr3 is a modulator that is only required in some Hh dependent organs/cells . We have shown that Ubr3 down-regulates the levels of Cos2 and its mammalian homolog Kif7 , key negative regulators of Hh signaling , through poly-ubiquitination . The poly-ubiquitination of Cos2 by Ubr3 is enhanced by Hh activation , suggesting that it functions in a positive feedback that modulates Hh activation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"gene",
"regulation",
"vertebrates",
"cloning",
"animals",
"animal",
"models",
"osteichthyes",
"drosophila",
"melanogaster",
"model",
"organisms",
"immunoprecipitation",
"molecular",
"biology",
"techniques",
"eyes",
... | 2016 | Ubr3, a Novel Modulator of Hh Signaling Affects the Degradation of Costal-2 and Kif7 through Poly-ubiquitination |
Human rhinoviruses ( HRV ) cause the majority of common colds and acute exacerbations of asthma and chronic obstructive pulmonary disease ( COPD ) . Effective therapies are urgently needed , but no licensed treatments or vaccines currently exist . Of the 100 identified serotypes , ∼90% bind domain 1 of human intercellular adhesion molecule-1 ( ICAM-1 ) as their cellular receptor , making this an attractive target for development of therapies; however , ICAM-1 domain 1 is also required for host defence and regulation of cell trafficking , principally via its major ligand LFA-1 . Using a mouse anti-human ICAM-1 antibody ( 14C11 ) that specifically binds domain 1 of human ICAM-1 , we show that 14C11 administered topically or systemically prevented entry of two major groups of rhinoviruses , HRV16 and HRV14 , and reduced cellular inflammation , pro-inflammatory cytokine induction and virus load in vivo . 14C11 also reduced cellular inflammation and Th2 cytokine/chemokine production in a model of major group HRV-induced asthma exacerbation . Interestingly , 14C11 did not prevent cell adhesion via human ICAM-1/LFA-1 interactions in vitro , suggesting the epitope targeted by 14C11 was specific for viral entry . Thus a human ICAM-1 domain-1-specific antibody can prevent major group HRV entry and induction of airway inflammation in vivo .
Human rhinoviruses ( HRVs ) infect the upper respiratory tracts of healthy subjects and cause around three quarters of common colds [1] , [2] . Recently , it has become clear that HRVs are also major causes of severe , life-threatening illnesses in susceptible populations: 50–85% of acute asthma exacerbations are now known to be caused by HRV infections across all age groups [3] , [4] , [5] , [6] . Further , wheezing illnesses in infancy caused by HRVs are strongly associated with a very high risk of later development of childhood asthma [7] , [8] , [9] . HRV infections are also associated with the majority of acute exacerbations of chronic obstructive pulmonary disease ( COPD ) [10] , [11] and cystic fibrosis [12] and cause life-threatening illnesses in susceptible populations such as infants [13] , the elderly [14] and immuno-compromised individuals [15] . Against this background , there is an urgent need to develop effective medications to combat HRV induced illnesses . To date there are no licensed treatments or vaccinations . Rhinoviruses are positive-stranded RNA viruses of the Picornaviridae family and more than 100 antigenically distinct serotypes have been identified [1] , [16] , [17] . Serotyped strains of HRVs are classified by use of their entry receptor into minor group viruses , which comprise ∼10% of identified serotypes and use low-density-lipoprotein receptor and related molecules [17] and major group viruses , comprising ∼90% of identified serotypes , which use intercellular adhesion molecule 1 ( ICAM-1 , CD54 ) as their receptor [2] , [16] . Thus ICAM-1 is an attractive target for development of new therapies to combat major group HRV infections . However , apart from its role as HRV receptor , ICAM-1 also has important host functions and in particular it is critical in host defence against numerous pathogens via its role in leukocyte recruitment and activation [18] , [19] , [20] . Thus inhibiting the function of ICAM-1 has potential to interfere not only with HRV cell binding , but also to impair host defence by inhibiting its interaction with its natural host ligands . ICAM-1 is a member of the immunoglobulin ( Ig ) superfamily and the extracellular portion of the molecule has five Ig-like domains termed D1 as the most distal , to D5 as the most proximal to the cell membrane . HRV serotypes binding ICAM-1 have been reported to bind only to D1 , while LFA-1 and Mac-1 bind to domains D1 and D3 , respectively [21] , [22] , [23] . Thus antibodies specific to D1 of ICAM-1 would be needed to interfere with HRV infection; however they could potentially interfere with ICAM-1/LFA-1 interactions . In the present report we characterise an anti-human ICAM-1-specific antibody ( 14C11 ) and show its binding is specific to human ICAM-1 D1 , that it is inhibitory against multiple major group HRV serotypes in vitro , and prevents major group HRV infection and HRV induced exacerbation of allergic airway inflammation in vivo . We also show that this antibody does not inhibit human ICAM-1/LFA-1 interactions and therefore should not interfere with cellular trafficking .
ICAM-1 plays an important role in trans-endothelial migration of leukocytes and activation of T cells via LFA-1 . In addition it is also the entry receptor for major group HRVs . We were interested in identifying antibodies suitable to inhibit HRV-ICAM-1 interactions for the potential treatment of HRV-induced exacerbations in airway disease without compromising the host leukocyte trafficking by interference of ICAM-1 interactions with the integrin LFA-1 . We thus initially investigated the binding properties of the commercially available antibodies to human ICAM-1 - 14C11 and 84H10 - to determine whether they bind to domain 1 of human ICAM-1 ( Hu ICAM-1 ) the putative binding site for viral entry [2] , [16] . Biotinylated Hu ICAM-1 , a chimeric protein consisting of domain 1 of human ICAM-1 and domains 2–5 of mouse ICAM-1 ( Hu1 Mu2-5 ) , murine ICAM-1 ( Mu ICAM-1 ) and insulin as a control protein were incubated with 14C11 or 84H10 to test their binding specificity . We found binding of 14C11 and 84H10 to both Hu ICAM-1 and Hu1 Mu2-5 confirming that 14C11 binds specifically to domain 1 of human ICAM-1 , but not to murine ICAM-1 ( Figure 1A and 1B ) . Additionally , cells were stimulated with PMA in a human T cell ( Jurkat cell ) adhesion assay , to determine whether 14C11 or 84H10 could inhibit human ICAM-1/LFA-1 interactions . Neither 14C11 nor the isotype control was able to prevent cell adhesion via LFA-1 whereas the anti-ICAM-1 antibody 84H10 blocked the ICAM-1/LFA-1 interaction and failed to retain the labelled Jurkat cells ( Figure 1C ) . These results emphasise that the domain specificity of 14C11 suggests it will not block human ICAM-1/LFA-1 interaction . Regarding binding to other members of the immumoglobulin ( Ig ) superfamily structurally related to huICAM-1 the specification provided by the manufacturer stated that 14C11 showed no cross-reactivity with rabbit ICAM-1 , human ICAM-2 , human ICAM-3 , human VCAM-1 , and ‘mouse DCC’ . We also demonstrated the binding specificity of 14C11 observing binding to human ICAM-1 and a chimeric mouse ICAM-1 with domain 1 substituted for human ICAM-1 Ig domain 1 . In the same assay no binding to mouse ICAM-1 , human ICAM-2 , human ICAM-3 , human ICAM-5 or human VCAM-1 was observed . ( Fig . S1 ) . Next we tested the effect of 14C11 in an in vitro HRV infection assay . Using a cytopathic effect assay ( CPE ) we showed the antiviral effect of 14C11 in inhibiting the replication of HRV16 ( Figure 2A ) and HRV14 ( Figure 2B ) with increasing concentration of antibody . 14C11 did not inhibit CPE caused by infection with minor group serotype HRV25 ( Figure 2C ) consistent with ICAM-1 blockade-specific mechanism of action . Isotype control antibody did not prevent replication of HRV16 or HRV14 in HeLa Ohio cells . Additionally , the median tissue culture inhibitory concentration ( IC50 ) was assessed for a wide range of major group HRVs as indicated in Figure 2D , suggesting a consistent prevention of replication by 14C11 for many major type HRVs , the great majority with sub-nanoMolar IC50s . These results emphasise that the domain specificity of 14C11 indicates it inhibits human ICAM-1-HRV binding but does not block human ICAM-1/LFA-1 interaction . To test the effect of the anti-ICAM-1 antibody 14C11 in an in vivo major group HRV infection model , we challenged transgenic Balb/c mice over-expressing extracellular domain 1 and 2 of human ICAM-1 ( tg+ ) and non-transgenic littermates ( tg− ) with major group HRV16 as described previously [24] . Transgenic positive ( tg+ ) mice were dosed 2 hours prior to viral challenge with 3 doses of 14C11 intranasally; 1 µg , 10 µg and 100 µg/mouse 14C11 or 100 µg/mouse isotype control and were subsequently infected intranasally with major group HRV16 . Transgenic negative ( tg− ) mice were challenged with the same dose of HRV16 as the tg+ littermates . BAL was performed 2 days after infection . Tg+ mice showed a robust cell infiltration and cytokine and chemokine production in the lung compared to tg− mice for all parameters measured . Isotype control treatment did not alter HRV induced airway inflammation for any endpoint ( Figure 3 ) . Antibody treatment in the absence of infection did not cause any measurable immune response in the lung ( data not shown ) . 14C11 significantly reduced total BAL cell , macrophage , neutrophil and lymphocyte numbers in a dose dependent manner ( Figure 3A ) . HRV induction of the pro-inflammatory cytokines IL-1β and IL-6 , the chemokines CXCL1 , CXCL11 and CXCL10 , as well as type III interferon IFNλ2/3 were all significantly reduced in BAL by 14C11 treatment ( Figure 3B ) . Induction of IL-1β , IL-6 and CXCL1 were also significantly reduced in lung homogenate ( Figure 3C ) . The kinetics of HRV16 viral load was assessed by real-time PCR . Tg+ mice and tg+ mice pre-treated with isotype control showed increased HRV16 vRNA levels at 6 to 9 hours after challenge compared to tg− mice , while the HRV16 vRNA levels were completely abolished by 14C11 treatment ( Figure 3D ) . H&E staining of lung sections from tg+ mice infected with HRV or isotype pre-treated HRV infected mice showed increased inflammatory cells in the lung . However , in 14C11 treated animals as well as control tg− animals challenged with HRV only marginal inflammation was observed ( Figure 3E ) . The human ICAM-1 specific antibody 14C11 could therefore prevent HRV16 entry and replication as well as induction of airway inflammation in vivo . To investigate further how 14C11 influences airway inflammation induced by another major group virus , we analysed BAL from HRV14 infected mice . We chose this serotype as HRV14 is in the HRV-B type group and is genotypically distinct from HRV16 ( which belongs to the HRV genotype group A ) . BAL cell analysis revealed increases in total BAL cell , neutrophil and lymphocyte numbers in HRV14 infected tg+ mice and HRV14 infected tg+ mice pre-treated with isotype control compared to tg− controls . Virus induction of total BAL cell , neutrophil and lymphocyte numbers were significantly reduced in tg+ mice pre-treated with 14C11 ( Figure S1A ) . The induction of proinflammatory cytokine IL-6 , the chemokines CXCL1 , CXCL11 and CXCL10 as well as type III interferon IFNλ2/3 were not different in tg+ mice infected with HRV14 and isotype pre-treated tg+ infected mice . However , induction of cytokines , chemokines and type III interferon IFNλ2/3 in both BAL ( Figure S1B ) and lung homogenate ( Figure S1C ) were reduced by 14C11 treatment in HRV infected tg+ mice , to levels similar to those in HRV14 challenged tg− control mice . Thus 14C11 is able to inhibit group B HRV-as well as group A HRV-induced inflammation . As therapeutic antibodies for the treatment of respiratory diseases are typically dosed parenterally [25] we next evaluated the pharmacological activity of 14C11 via the intravenous route of administration . Mice were pre-treated 24 hours prior to viral challenge and then cell influx and pro-inflammatory cytokine release were assessed . Control groups of HRV16 infected tg+ and isotype treated control mice showed elevated levels of total BAL cells , lymphocytes and neutrophils on day 2 after infection compared to HRV16 challenged tg− mice ( Figure 4A ) . ) . Administration of intravenous 14C11 24 h prior to HRV infection dose dependently reduced airways inflammation in terms of total numbers of BAL cells , lymphocytes and neutrophils . ( Figure 4A ) . 14C11 significantly reduced in a dose-dependent manner the induction of CXCL1 , CXCL11 and CXCL10 in the BAL of tg+ HRV16 infected mice compared to untreated or isotype pre-treated HRV16 infected tg+ mice ( Figure 4B ) . Induction of IL-1β and IL-6 as well as IFNλ2/3 was also reduced by 14C11 treatment in tg+ HRV16 infected mice ( Figure S2A ) . Similar results were found in lung homogenate for protein levels of IL-1β , IL-6 and CXCL1 ( Figure S2B ) . Systemic administration of 14C11 is therefore protective against HRV16 infection in terms of induction of airway inflammation in vivo . Antibodies are large complex proteins that not only bind antigen via the variable domain but also bind Fc receptors on antigen presenting cells by binding via the constant domain . Moreover 14C11 is a mouse anti-human IgG1 type antibody and could potentially bind to mouse Fcγ receptors on innate immune cells , such as FcγRIIB , providing a potential inhibitory signal . To investigate whether the 14C11 antibody non-specifically inhibited inflammation by cross-linking human ICAM-1 and binding Fcγ receptor cells we therefore investigated whether 14C11 could inhibit cellular inflammation evoked in tg+ mice , but by a non human ICAM-1 mediated infection using minor group HRV1B infection . Transgenic negative and transgenic positive littermates were dosed intravenously 24 hours prior to viral infection with or without 20 mg/kg isotype control or 14C11 antibody and consequently infected with UV-inactivated HRV1B or with HRV1B . Significant induction of total BAL cells and neutrophils at day 1 post infection and lymphocyte numbers at day 4 could be observed compared to mice infected with UV-inactivated HRV1B in both types of mice ( Figure 5A ) . Induction by HRV1B of each cell type was virtually identical in tg+ and tg− mice , suggesting that the transgenic over-expression of the chimeric human/mouse ICAM-1 molecule did not influence cell trafficking via potential interactions with mouse LFA-1 . Pre-treatment with isotype control or 14C11 antibody in transgene positive mice infected with HRV1B did not alter the induction of cellular inflammation , while infiltration of inflammatory cells in the lung of HRV16 infected transgene positive mice was significantly reduced by 14C11 , showing that 14C11 was able to block HRV entry and replication in the same experiment ( Figure 5A and S3A ) . Similarly , 14C11 administration did not reduce HRV1B-induction of CXCL1 , CXCL10 or CXCL11 , nor the cytokines IL-1β , IL-6 or IFNλ2/3 in BAL or lung homogenate ( Figure S3B and C ) . We also explored whether 14C11 would non-specifically inhibit inflammatory responses using an LPS challenge model in the presence or absence of 14C11 antibody . Although we would not expect 14C11 to interfere with mouse LFA-1 - mouse ICAM-1 interaction it was still necessary to demonstrate the specificity of 14C11 by performing studies in a model of airway inflammation that did not require human ICAM-1 . Mice challenged with LPS , LPS and systemic administered isotype control or 14C11 showed significant increases in total BAL cells on day 1 and neutrophil and lymphocyte numbers on day 4 after infection ( Figure 5B ) , which were not altered by pre-treatment with 14C11 . In addition , pro-inflammatory cytokines were also unaffected by 14C11 treatment ( Figure S3D and E ) . Taken together , this indicates that the human ICAM-1 specific antibody 14C11 could prevent airway inflammation in vivo for two major group HRVs , HRV16 and HRV14 , but had no non-specific effects on inflammation induced by a stimuli involving human ICAM-1 independent mechanisms . We then asked whether the progression of inflammatory responses to viral infection at later time points is altered by 14C11 to shed light on whether 14C11 simply retards , or completely arrests HRV16 infection . Therefore , we dosed tg+ mice intranasally 2 hours prior to viral infection either with or without 100 µg/mouse 14C11 antibody or 100 µg/mouse isotype control and screened for airway inflammation at day 2 , 4 and 7 post infection . Tg− mice received the same dose of HRV16 as all other groups ( Figure 6 ) . In HRV16 infected tg+ mice as well as in infected tg+ mice pre-treated with isotype control total BAL cell numbers peaked at day 2 after infection and stayed increased till day 7 . Counts of total BAL cells in tg+ mice pre-treated with 14C11 revealed the same total BAL cell numbers as those observed in tg− mice challenged with HRV16 up to day 7 . Macrophages and lymphocytes peaked at day 4 after infection for the tg+ positive HRV16 group and mice pre-treated with isotype control; however macrophage numbers showed a prolonged increase while lymphocyte numbers decreased after day 4 and returned close to baseline . 14C11 significantly reduced macrophage and lymphocyte numbers to levels found in tg− littermates challenged with HRV16 . Neutrophils were also recruited and highest levels of neutrophils could be observed at day 2 post-infection in HRV16 infected tg+ mice or isotype-control pre-treated and HRV16 infected tg+ mice . A single dose of 14C11 significantly reduced the number of neutrophils to control levels found in challenged tg− mice , over a sustained period of 7 days ( Figure 6A ) . IL-1β , IL-6 and the chemokines CXCL1 , CXCL11 and CXCL10 as well as IFNλ2/3 in the BAL peaked at day 2 after infection in HRV16 infected tg+ mice and those pre-treated with isotype control , whereas tg− littermates and 14C11 treated HRV16 infected tg+ mice failed to induce cytokine and chemokine production . All cytokines and chemokines returned to baseline levels on day 4 of infection , except IFNλ2/3 which could still be detected on day 4 in tg+ infected with HRV16 and isotype control treated animals , but not in tg− mice or tg+ mice pre-treated with 14C11 ( Figure 6B and S4A ) . Similar results were found in lung homogenate for levels of IL-1β , IL-6 and CXCL1 protein ( Figure S4B ) . We next checked the influence of 14C11 on the adaptive immune response by measuring HRV16-specific antibodies . HRV16 infected tg+ and isotype control pre-treated mice showed increased levels of HRV16-specific IgG1 and IgG2a in serum on day 7 post-infection . Levels of HRV16 specific IgG1 and IgG2a were reduced in infected and 14C11 treated and in tg− HRV16 challenged mice ( Figure 6C ) . A single dose of 14C11 was therefore able to decrease HRV16-induced cellular infiltration in the lung , induction of cytokines and chemokines , and the induction of HRV16-specific antibodies over a sustained period of 7 days . Allergic asthma is exacerbated by viral infections , especially by HRV infections . To model virus-induced exacerbations of allergic airway inflammation , transgenic positive mice were sensitised and challenged with OVA , and infected with HRV16 . We asked whether 14C11 could prevent HRV16-induced exacerbation of allergic airway inflammation . The asthma-like phenotype with eosinophilic airway inflammation was confirmed in all 3 groups challenged with OVA on day 6 after infection ( Figure 7B ) . Mice not challenged with OVA , but HRV16 infected ( RV-PBS ) showed increased neutrophil numbers on day 2 compared to UV-HRV16 infected mice ( UV-PBS ) . Compared to mice challenged with UV-inactivated HRV16 and OVA ( UV-OVA ) , HRV16 and OVA challenge in isotype pre-treated tg+ mice ( RV-OVA iso ) significantly increased total BAL cell and neutrophil numbers on day 2 after infection , lymphocyte numbers on day 6 and increased airway hyper-responsiveness as measured by PenH at 24 hrs post-challenge . 14C11 pre-treated , HRV16 and OVA challenged mice ( RV-OVA 14C11 ) showed reduced numbers of total BAL cells , neutrophils and lymphocytes ( Figure 7A and B ) . RV-OVA-14C11 treatment also reduced the exacerbation of airway hyper-responsiveness to UV-OVA control levels on day 1 after infection ( Figure 7C ) . To further investigate the impact of 14C11 pre-treatment on HRV16-induced allergic airway inflammation we analysed Th2-type cytokine concentrations as well as the neutrophil and eosinophil chemokine attractants CXCL1 in BAL and eotaxin-1 levels in lung homogenate at day 2 after infection . Higher levels of IL-4 , IL-5 and IL-6 were found in RV-OVA iso mice compared to mice receiving RV-OVA 14C11 treatment ( Figure 7D ) . The level of IL-13 in OVA challenged groups was not increased by virus and therefore no effect of 14C11 treatment was observed . The neutrophil attractant CXCL1 was induced in all groups infected with HRV16 but not modulated by OVA . However , 14C11 administration reduced the level of CXCL1 in RV-OVA challenged animals . RV-OVA 14C11 treated mice exhibited a small but significant reduction in eotaxin-1 levels compared to RV-OVA iso treated controls ( Figure 7D ) . Total IgE levels measured in serum were not significantly increased by virus infection and not altered by 14C11 pre-treatment . Because in asthma exacerbations , increased inflammation and mucus production have been observed we analysed the impact of 14C11 on mucus production in HRV16-exacerbated allergic airway inflammation . RV-OVA iso challenged mice showed a small increase in the production of mucins MUC5AC and MUC5B compared to mice challenged with UV-OVA challenge . 14C11 pre-treatment of RV-OVA challenged mice significantly reduced the induction of both MUC5AC and MUC5B mucus proteins 6 days after infection ( Figure 7E ) . Our data showed that neutralisation of the HRV16 entry receptor ICAM-1 by 14C11 reduced HRV16-induced allergic airway inflammation in vivo . These data indicate the possibility to design a domain-specific anti-ICAM-1 antibody to specifically hinder major group HRV entry and replication without blocking the crucial LFA-1 function for unaltered cell recruitment to the site of infection . We have shown for the first time a major group allergen-induced asthma exacerbation model and the possible use of a domain-specific anti-ICAM-1 antibody as possible treatment for rhinoviral triggered exacerbations of asthma .
Human rhinovirus infection is the most common infection afflicting mankind and is responsible for enormous morbidity and societal costs as the major cause of the common cold . It also causes severe morbidity , mortality and health care costs as a key trigger of acute exacerbations of lung diseases such as asthma , COPD and cystic fibrosis . In this study , we demonstrate that an anti-ICAM-1 domain-specific antibody can effectively block entry and replication in an in vivo model of human rhinovirus infection and reduce rhinovirus-induced exacerbation of allergic airway inflammation , without preventing ICAM-1 interaction with its cellular ligand LFA-1 , which is required for cell recruitment during respiratory infections . Thus 14C11 and other antibodies with similar specificities could be useful agents to help reduce rhinovirus induced respiratory exacerbations . . The majority of HRVs use a single cellular receptor , ICAM-1 , for attachment to cells and therefore ICAM-1 is an interesting target to block virus-receptor binding to prevent rhinovirus infection . Several alternative approaches have been investigated , but have not been taken further in clinical trials or had to be abandoned because of side effects . For example , soluble ICAM-1 ( tremacamra , BIRR4 ) reduced the severity of experimental colds [26] , but was not developed further potentially due to the expense of manufacturing recombinant proteins as drugs , as well as the frequency of dosing required . Interferon-alpha used intranasally was also shown to be effective against HRV infection , but had side effect including leukopenia and nasal bleeding [27] . The HRC 3C protease inhibitor ( Ruprintrivir ) was reported to reduce the proportion of subjects with positive viral cultures and viral titres but did not decrease the frequency of colds [28] , [29] . Pleconaril has also been shown to be clinically effective in reducing the duration of colds , but did not get approval due to its side effects profile [30] , [31] . Lack of cross-protection between different serotypes of HRV has also meant that no successful vaccine has been developed and thus no specific therapy against HRVs exists , even though there is a very high medical need . Antibody therapy is an attractive alternative therapeutic approach , as a single injection can provide protection of long duration . In this present study we have identified a human ICAM-1 specific antibody , 14C11 , which specifically binds to domain 1 of human ICAM-1 ( Figure 1 ) , the rhinovirus binding domain [32] . 14C11 was able to block infection in vitro with HRV16 and HRV14 , members of the HRV A-type and B-type group , respectively . Moreover , 14C11 was able to block a wide range of other major group rhinovirus strains , suggesting that blocking of rhinovirus entry and replication with 14C11 could be seen as broad spectrum blocking agent for major group rhinoviruses ( Figure 2 ) . ICAM-1 , through its interaction with LFA-1 is also important in cell trafficking , and ICAM-1/LFA-1 interactions are reported to involved domains 1 of the ICAM-1 molecule [32] , thus antibodies to ICAM-1 have theoretical potential to inhibit cell trafficking through blocking ICAM-1/LFA-1 interaction . Binding assays using the Jurkat human T cell line have shown that 14C11 , in contrast to another anti-ICAM-1 antibody 84H10 , did not interfere with LFA-1 interaction and therefore that this human ICAM-1 domain 1-specific antibody potentially only hinders HRV entry and replication ( Figure 1 ) . To test the 14C11 antibody in vivo we infected transgenic mice over-expressing a chimeric ICAM-1 molecule , consisting of mouse ICAM-1 in which domains 1 and 2 were replaced with human ICAM-1 domains 1 and 2 [24] , with the major group rhinovirus HRV16 . We observed a cellular infiltration in the lung in HRV16 infected transgenic positive mice , comprising an increase in total BAL cells , lymphocytes and neutrophils , which was decreased in a dose dependent manner by topical administration of 14C11 for all parameters . Chemokines and cytokines in BAL and lung homogenate were also significantly reduced by 14C11 treatment . In addition , rhinovirus replication in transgene positive mice was completely blunted with 14C11 treatment , suggesting that blocking of all ICAM-1 binding sites in the lung can prevent HRV entry and replication in vivo . Histological analysis confirmed that treatment with 14C11 reduced inflammatory cell infiltration of the lung and was comparable to the lack of infiltration obtained by challenge of non-transgenic littermates with HRV16 ( Figure 3 ) . Genome sequence places HRV16 in group A which also contains the majority of ICAM-utilising major group viruses ( 63 out of a currently identified 88 serotypes ) . HRV14 is amongst a significant minority of genetically distinct major group viruses ( 25 serotypes ) that belong to group B viruses [17] . We could additionally show that infection with HRV14 could be prevented by 14C11 in vivo . Cellular infiltration in the lung , as well as cytokine and chemokine levels in BAL and lung homogenate were all reduced by topical administration of 14C11 ( Figure S2 ) . Thus ICAM-1 blockade has shown efficacy in vivo against representative serotypes of both species of major group rhinoviruses . Of the 99 serotyped rhinovirus strains , major group viruses which use ICAM-1 and which can potentially be specifically blocked by 14C11 constitute 88 serotypes , while minor group viruses , which do not use ICAM-1 and which therefore would not be treated with 14C11 , constitute only 11 serotypes , thus 14C11 has potential to treat 89% of serotyped rhinovirus strains . A third group of rhinoviruses named group C viruses has been recently identified based on RNA genome sequence analysis [33] . The size of this group of rhinoviruses is estimated on sequence analysis to be around 60 virus strains . Viruses in this group have not been serotyped as they have not yet been successfully cultured in a cell line [16] , however based on sequence analysis , it is assumed that this group of viruses will not use ICAM-1 as their cellular receptor [16] . If this is confirmed by experimental data antibodies targeting the same epitope to 14C11 and thus only inhibiting major group virus entry would be anticipated to block∼55% of rhinovirus strains known to date . Therapeutic use of antibodies in man normally involves systemic rather than inhaled dosing , for this reason we wished to determine if systemic dosing with 14C11 would successfully inhibit rhinovirus induced airway inflammation . Systemic dosing of 14C11 did prevent HRV16 induced inflammation and proinflammatory cytokine release , thus showing pharmacological activity of 14C11 via the intravenous route and suggesting the suitability of parenteral dosing for route of administration for an anti-ICAM1 based therapy ( Figure 4 and Figure S3 ) . To investigate the human ICAM-1-specificity of the action of 14C11 in vivo and to exclude possible interactions of the Fc part of the 14C11 antibodies we tested for possible inhibition by 14C11 in minor group HRV infection , which does not use human ICAM-1 as virus entry receptor and in an LPS challenge model where induction of airway inflammation is independent of human ICAM-1 ( Figure 5 and Figure S3 ) . In both models of airway inflammation , we found that 14C11 , which does not bind mouse ICAM-1 in vitro ( Figure 1 ) , also had no effect on human ICAM-1 independent airway inflammation in vivo . These data confirmed that the inhibition of HRV infection and induction of airway inflammation in vivo was specifically a consequence of blockage of rhinovirus entry and replication , and not due to non-specific effects . We next investigated whether a single administration of 14C11 prior to HRV infection was effective in significantly inhibiting later outcomes of HRV infection . Complete inhibition of HRV induced total cell counts , macrophages , lymphocytes and neutrophils as well as virus-specific antibody induction were achieved up to 7 days after administration of 14C11 ( Figure 6 ) , thus strengthening our hypothesis that domain specific blocking of HRV entry and replication could be a potential anti-viral approach in man . These data also confirmed that 14C11 treatment completely arrested , rather than simply retarded HRV infection in vivo . Having shown that this antibody could inhibit rhinovirus infection and induction of inflammation in vivo , we next wished to test the antibody in a model of disease where the costs in man would justify the development of a relatively expensive treatment such as a therapeutic antibody . Rhinovirus infections are responsible for the majority of acute exacerbations of asthma and exacerbations are responsible for around 50% of asthma related health care costs [34] . We therefore tested the antibody in a mouse model of rhinovirus exacerbation of allergic airway inflammation [24] . We show that 14C11 was effective in suppressing each of the outcomes that were exacerbated by rhinovirus infection superimposed on a model of ovalbumin induced allergic airway inflammation , including HRV-induced exacerbation of total cellular , neutrophilic and lymphocytic airway inflammation , as well as Th2 ( IL-4 and IL-5 ) and pro-inflammatory ( IL-6 ) cytokine and neutrophil ( CXCL1 ) and eosinophil ( eotaxin ) recruiting chemokine release . Additionally , 14C11 treatment significantly suppressed HRV exacerbation of the two major respiratory mucins MUC5AC and MUC5B , and importantly , airway hyper-responsiveness . In conclusion we have shown that the human ICAM-1 domain 1-specific antibody 14C11 inhibits major group HRV infection in vitro and in vivo , as well as inhibiting major group HRV-induced exacerbation of allergic airway inflammation in vivo . This antibody however did not interfere with human ICAM-1 binding to its ligand on human T cells and further it had no non-specific effects on cell recruitment in models of airway inflammation induced by mechanisms independent of human ICAM-1 . This antibody and others with similar properties would therefore be good candidates for development of novel treatments for diseases induced by major group rhinoviruses – principally acute exacerbations of airway diseases such as asthma , COPD and cystic fibrosis .
All animal work was completed in accordance with UK Home Office guidelines following approval via the ethical approval process ( UK project licence PPL 70/7234 valid 03/03/2011 to 03/03/2016 ) . ICAM-1 antigens were generated using mammalian expression systems . Human ICAM-1 domain 1/mouse ICAM-1 domains 2–5 chimera , and mouse ICAM-1 domains 1–5 were generated in-house . Human ICAM-1 domains 1–5 was obtained commercially ( R&D Systems 720-IC ) . All ICAM-1 antigens were biotinylated via free amines using EZ link Sulfo-NHS-LC-Biotin ( Thermo Scientific ) . Biotinylated bovine insulin ( Sigma I2258 ) was used as control for non-specific binding . Streptavidin plates ( Thermo Scientific ) were coated with the biotinylated antigens at 1 µg/ml in PBS and incubated overnight at 4°C . Plates were washed , blocked ( PBS+0 . 1% BSA ) for 1 hour and serial dilutions of anti-ICAM-1 antibody 14C11 ( MAB720 , R&D Systems ) were added in blocking buffer for 1 hour as indicated in the figures . ELISA to test 14C11 binding specificity used 14C11 and control antibodies from R&D Systems: anti-mouse ICAM-1 ( MAB796 ) anti-human ICAM-2 ( MAB244 ) , anti-human ICAM-3 ( MAB813 ) , anti-human ICAM-5 ( MAB1173 ) and anti-human VCAM-1 ( MAB809 ) . Plates were washed and developed with anti-human IgG light chain HRP antibody ( Sigma ) , TMB and H2SO4 . Plates were read on an EnVision™ plate reader at 450 nm . Human ICAM-1-Fc ( R&D Systems ) was coated onto 96-well black walled plates at 10 µg/ml in PBS overnight at 4°C . Plates were then washed and blocked with assay buffer ( phenol-red free HBSS plus 1% BSA ) for 1 h . Antibodies ( 14C11 , 84H10 ( Serotec ) and muIgG1 isotype control were serially diluted in assay buffer ( 4× final concentration ) and 50 µl/well transferred to ICAM-1-Fc coated plates . After a 15 min incubation period at room temperature , a further 50 µl/well of PMA ( 50 ng/ml ) in assay buffer was then added . During this time , Jurkat E6 . 1 cells were harvested and labelled with Calcein-AM dye ( 5 µM in phenol-red free RPMI 1640 ) for 30 minutes at 37°C . Following washing , 105 Jurkat E6 . 1 cells were added to each well in 100 µl assay buffer . Plates were then incubated for a further 3 h at 37°C/5% CO2 . Adhesion of the cells to the plate was assessed following aspiration of the plate and washing with 250 µl wash buffer ( 150 mM HEPES , 0 . 1% glucose , 2 mM MgCl2 pH 7 . 2 ) . A final 250 µl of wash buffer was added and plates were read for fluorescein detection . Data are expressed without PMA ( 0% signal ) and with PMA and no antibody ( 100% signal ) . HeLa Ohio cells ( ECACC ) were seeded at 3×104/well into 96 well plates in MEM ( minimal essential medium ) supplemented with 1× non-essential amino acids , penicillin ( 100 U/ml ) , streptomycin ( 100 µg/ml ) ( each from Invitrogen , Paisley , UK ) and heat inactivated bovine calf serum ( 10% ) , ( SAFC Biosciences , Andover , UK ) . After an overnight incubation ( 37°C/5% CO2 ) the media was removed and replaced with serially diluted anti-ICAM-1 antibody 14C11 , mIgG1 isotype control ( R&D Systems , Abingdon ) or medium alone for untreated controls and incubated at 37°C for 30 minutes . Following this , 30 µl of HRV16 or HRV14 culture supernatant ( multiplicity of infection predetermined to induce 80–90% CPE ) was then added to the wells ( or medium alone for un-infected controls ) and virus attachment was allowed to proceed for 2 hours at 33°C/5% CO2 . Assay supernatants were then removed and 100 µl of fresh medium was added to each well and incubated for a further 23 hours at 33°C/5% CO2 . Remaining viable cells were fixed in 4% formaldehyde ( Sigma-Aldrich , Poole , UK ) , stained with 0 . 5% w/v crystal violet ( Sigma-Aldrich , Poole , UK ) for 5 minutes , washed 3 times in water to remove excess stain and allowed to dry . The extent of CPE was quantified by measuring absorbance at 600 nm and calculated according to the equation % CPE = 100×[ ( test sample−virus only ) / ( uninfected control−virus only ) ] . Human-mouse ICAM-1 transgenic mice were generated as described previously [24] . Human-mouse ICAM-1 transgenics were bred in-house and transmission of the chimeric transgene was confirmed by PCR . Mice bearing the human-mouse ICAM-1 transgene are referred in the paper as tg+ and mice not expressing the transgene are referred as tg− . HRV were grown in HeLa Ohio cells . Infected cells were harvested after 24 hours and HRV was concentrated and purified as described previously [24] . Viral titers were assessed by TCID50 assay ( 50% tissue culture infectivity dose ) . The identity of HRV1B , HRV14 and HRV16 working stocks were confirmed by neutralisation with serotype specific antibodies ( ATCC ) . The HRV infection and HRV-induced asthma exacerbation models have been described previously [24] . For infection studies mice were infected with 5×106 TCID50/ml of the indicated HRV . For the mouse asthma exacerbation model mice were infected with 2 . 5×106 TCID50/ml . To demonstrate HRV replication-specific responses HRV1B and HRV16 were UV-irradiated using a UV Stratalinker 2400 ( Stratagene ) . Mouse anti-human ICAM-1 antibody 14C11 ( MAB720 , R&D Systems ) and mouse IgG1 isotype control ( MAB002 , R&D Systems ) was used as a 0 . 2 µm filtered solution dissolved in PBS . 5 to 7 week old transgenic positive ( tg+ ) or transgenic negative ( tg− ) females were lightly anaesthetised and dosed intranasally with mouse anti-human ICAM-1 antibody 14C11 or isotype control 2 hours prior to challenge with HRV16 or HRV14 . For intravenous treatment , 14C11 or isotype control was administered in the tail vein 24 hours before infection with HRV16 , HRV1B or UV-inactivated HRV1B ( UV-HRV1B ) . Bronchoalveolar lavage ( BAL ) was performed using 1 . 5 ml BAL buffer ( EBSS , 55 mM EDTA , 12 mM lidocaine ) 1 , 2 , 4 or 7 days after infection , as indicated . BAL supernatant was aliquoted and stored for cytokine and chemokine analysis . BAL cells were counted and processed onto slides by cytospin , differentially stained ( Quick Diff , Reagena , Finland ) and counted blind to experimental conditions . 300 cells were counted per slide . The small upper lung lobe was stored in RNAlater ( Qiagen ) for mRNA analysis . The remaining lung tissue was homogenised in 2 ml of PBS , clarified by centrifugation and stored for cytokine and chemokine analysis . 5 to 7 week old transgenic positive females were injected intraperitoneally ( i . p . ) with 50 µg OVA ( Albumin from chicken egg , Calbiochem ) and 2 mg aluminium hydroxide ( Sigma ) in a volume of 200 µl on day −13 . Lightly anesthetised mice were challenged with 40 µg OVA/mouse on day −2 and day −1 to induce allergic airway inflammation . On day 0 anesthetised mice were challenged either with 20 µg OVA or PBS and infected with HRV16 or dosed with UV-inactivated HRV16 ( RV-OVA , UV-OVA or RV-PBS , UV-PBS ) . Airway hyper-reactivity ( AHR ) was assessed by whole body plethysmography on day 1 after infection in response to increasing doses of nebulised methacholine ( MCh ) . BAL supernatant and cells were collected at day 2 or day 6 after infection and processed as described above . Cytokines and chemokines in BAL or lung homogenate were measured by ELISA using commercial kits from R&D Systems or Meso Scale Discovery . IL-4 , IL-5 and IL-13 were measured with R&D Quantikine ELISA kits; eotaxin , IP-10 , ITAC and IFNλ2/3 were measured with R&D Duosets; IL-1β , IL-6 and CXCL1 were measured with MSD multi-array according to the manufacturer's recommendations . Total IgE was measured in serum with BD OptEIA ( BD Bioscience ) according to the manufacturer's instructions . Mucins MUC5AC and MUC5B were determined by semiquantiative ELISA: BAL fluid was diluted in PBS and dried over night in Maxisorp plates ( Nunc ) . MUC5AC was detected using a biotinylated anti-MUC5AC antibody ( Neomarkers ) and detected with streptavidin-HRP ( Invitrogen ) and TMB ( Sigma ) . MUC5B was detected using an anti-MUC5B antibody ( Santa Cruz Biotechnology ) and developed using a biotinylated anti-mouse IgG antibody ( Sigma ) , streptavin-HRP and TMB . Lung tissues were freshly fixed in 10% buffered formaldehyde , embedded in paraffin wax , sectioned 5 µm thick and stained with haematoxylin-eosin ( H&E ) . The morphology of H&E stained sections were visualized and images acquired using an Axioskop 40 light microscope with AxiCamMRc 5 digit camera ( Zeiss , West Germany ) . Nunc Maxisorp Immuno plates were coated with purified HRV stock in PBS overnight . After blocking , diluted serum samples were incubated for 1 hour , followed by detection with biotinylated rat-anti-mouse IgG1 or IgG2a antibodies . The assay was developed with streptavidin-HRP and TMB as substrate . Transgenic positive mice were dosed intravenously with 14C11 24 hours prior to intranasal challenge with 1 µg LPS/mouse . Lipopolysaccharide ( LPS ) from E . coli 055:B5 was purchased from Sigma . BAL was collected and processed as described above . The left upper lung lobe was stored in RNAlater ( Qiagen ) until RNA was purified using the RNeasy Mini kit ( Qiagen ) and the RNase-free DNase set ( Qiagen ) . Isolated RNA was reverse-transcribed using random hexamer primers ( OmniscriptRT kit , Qiagen ) . Real-time PCR was performed with the Quantitect Probe PCR master mix ( Qiagen ) . Primers and probes for 18S rRNA ( forward: 5′cgc cgc tag aggtgaaattct; revers: 5′cat tcttggcaaatgctttcg and probe: FAM5′acc ggcgcaagacggaccaga ) and HRV specific primers ( forward: 5′ gtgaagagccscrtgtgc t; reverse: gctscagggttaaggttagcc and probe FAM5′tga gtcctccgg ccc ctgaat g ) were used for quantification of HRV16 vRNA . All data were distributed normally and data are expressed as means ± SEM . Data were analysed with Prism3 ( Graph Pad ) or SAS Version 8 . | Viruses exploit receptors on the host cell to cause infection . Therapies aimed at blocking virus-receptor interactions have the potential to prevent viral disease . Cellular receptors are also important for normal host cell function . Therefore , new therapies targeting these receptors to block viral infection may also inadvertently alter the physiology of the host cell . Viral pathogens , such as the cold virus ( rhinovirus ) , are believed to be the major cause of asthma attacks and exacerbations in chronic obstructive pulmonary disease ( COPD ) . In this study , we show that it is possible to identify novel therapeutic antibodies that block infection with rhinovirus without impairing the receptors' main function of cell adhesion . We then use animal models that show that an antibody can inhibit virus-induced lung inflammation and disease . Moreover , we show that this antibody can also inhibit a virally induced asthma exacerbation . This work is relevant in that it shows that antibodies can be tailored to distinct regions of viral receptors to block infection without inhibiting the receptors' normal cellular function . This is important for the development of new treatments that will prevent diseases caused by infection with rhinovirus , such as exacerbations of asthma and COPD . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immune",
"cells",
"cytokines",
"viral",
"transmission",
"and",
"infection",
"immune",
"activation",
"immunity",
"to",
"infections",
"immunology",
"host-pathogen",
"interaction",
"microbiology",
"animal",
"models",
"viral",
"attachment",
"model",
"organisms",
"immune",
... | 2013 | An Anti-Human ICAM-1 Antibody Inhibits Rhinovirus-Induced Exacerbations of Lung Inflammation |
Multiple recent outbreaks of Rift Valley Fever ( RVF ) in Africa , Madagascar , and the Arabian Peninsula have resulted in significant morbidity , mortality , and financial loss due to related livestock epizootics . Presentation of human RVF varies from mild febrile illness to meningoencephalitis , hemorrhagic diathesis , and/or ophthalmitis with residual retinal scarring , but the determinants for severe disease are not understood . The aim of the present study was to identify human genes associated with RVF clinical disease in a high-risk population in Northeastern Province , Kenya . We conducted a cross-sectional survey among residents ( N = 1 , 080; 1–85 yrs ) in 6 villages in the Sangailu Division of Ijara District . Participants completed questionnaires on past symptoms and exposures , physical exam , vision testing , and blood collection . Single nucleotide polymorphism ( SNP ) genotyping was performed on a subset of individuals who reported past clinical symptoms consistent with RVF and unrelated subjects . Four symptom clusters were defined: meningoencephalitis , hemorrhagic fever , eye disease , and RVF-not otherwise specified . SNPs in 46 viral sensing and response genes were investigated . Association was analyzed between SNP genotype , serology and RVF symptom clusters . The meningoencephalitis symptom phenotype cluster among seropositive patients was associated with polymorphisms in DDX58/RIG-I and TLR8 . Having three or more RVF-related symptoms was significantly associated with polymorphisms in TICAM1/TRIF , MAVS , IFNAR1 and DDX58/RIG-I . SNPs significantly associated with eye disease included three different polymorphisms TLR8 and hemorrhagic fever symptoms associated with TLR3 , TLR7 , TLR8 and MyD88 . Of the 46 SNPs tested , TLR3 , TLR7 , TLR8 , MyD88 , TRIF , MAVS , and RIG-I were repeatedly associated with severe symptomatology , suggesting that these genes may have a robust association with RVFV-associated clinical outcomes . Studies of these and related genetic polymorphisms are warranted to advance understanding of RVF pathogenesis .
Rift Valley fever virus ( RVFV ) is a negative strand RNA virus of the family Bunyaviridae . Episodic epidemics of Rift Valley Fever ( RVF ) present a significant natural threat to human health in many countries of Africa and the Middle East , causing retinitis , encephalitis and hemorrhagic fever [1 , 2] . Epizootics of RVFV also seriously affect livestock , including sheep , cattle , goats , buffalo , and camels , creating serious economic disruption and risk of famine [3] . Two of the largest RVF outbreaks have occurred in Kenya over the last decade , the first in 1997–98 [4] , and another more recently in 2006–2007 [5] . Both epidemic human disease , ( including hemorrhagic fever ) , and enzootic livestock disease , ( including excess mortality and miscarriage ) , are most prevalent in semi-arid areas that experienced prolonged excess rainfall during El Nino-Southern Oscillation ( ENSO ) weather anomalies [6] . Given the recent US experience with West Nile Virus , we could expect that , after either accidental or intentional introduction , RVFV will have the potential to become a widespread multi-state or multinational problem in North America . Our ongoing field studies aim to better define the epidemiology of RVF viral transmission at the local community level [7–11] . However , little is known about the pathogenesis of the variable disease progression observed between different RVFV-infected human subjects . In communities where 20–30% of persons are exposed , only about 1% of infections progress to severe liver dysfunction and hemorrhagic disease , and late onset encephalitis is rare , although 10–30% develop some form of anterior or retinal eye disease [12 , 13] . Currently there is no specific treatment for RVF . It remains unclear why the majority of infected humans recover from RVFV infection after only a brief febrile illness . Evidence from experimental animal models suggests that early activation of innate immunity provides the greatest protection against lethal RVFV infection [14 , 15] , and that differences in interferon-mediated response pathways [16–18] could be responsible for resistance to lethal infection [18–20] . Later-onset , adaptive immunity ( with the production of neutralizing antibody responses ) likely also plays a role in modulation of RVFV-infection associated disease . However , given the rapid time course of lethal disease progression , with most major symptoms developing within the first week of illness [21] , a study of the variability in innate immune responses appears to be the logical first step to elucidate inter-subject variation in disease progression . Several classes of innate receptors are important in host anti-viral defense , including membrane bound Toll-like Receptors ( TLRs ) [22] , cytoplasmic DExD/H box RNA helicases such as retinoic acid-inducible gene-I ( RIG-I ) and melanoma differentiation-associated gene 5 ( MDA5 ) [23–26] and NOD-like receptors ( NLRs ) including inflammasomes [27–32] . TLRs are innate receptors that recognize specific structures expressed by microorganisms . Surface TLR2 and −4 recognize viral proteins including hemagglutinin of measles virus [33] and components of HSV [34] and CMV [35] and RSV [36] . Endosomal TLRs act as receptors for nucleic acids , including TLR3 ( dsRNA [37] ) ; TLR7 and TLR8 , ( single-stranded RNA [38 , 39] ) ; and TLR9 , ( unmethylated CpG DNA motifs [40] ) . There is strong evidence in support of a role for endosomal TLRs in the detection of viruses including Influenza virus and Vesicular Stomatitis virus ( VSV ) ( TLR7 ) and HSV ( TLR9 in certain cell types ) [38 , 41–43] . Signaling by all TLRs originates from a conserved intracellular domain ( Toll–IL-1–resistance; TIR ) , which mediates recruitment of members of a family of adaptor molecules . Recruitment of the common adaptor , myeloid differentiation factor 88 ( MyD88 ) [44 , 45] leads to the interaction and activation of the IRAK family members [46] and the subsequent activation of TRAF6 [47] resulting in NF-κB activation . Activation of the Interferon Regulatory Factors ( IRFs ) , important mediators of IFN gene transcription also occurs downstream of the TLRs . Thus , these pathways could potentially play an important role in modulating the severity of RVFV-associated disease in humans . A role for TLR3 was demonstrated in a murine model of another phlebovirus , Punta Toro virus [48] . Other reports have shown a protective role for poly I-C ( ligand for TLR3 and MDA5 ) when used as a pretreatment prior to RVFV infection with virulent ZH501 strain [14 , 49] . In our previous study , TLRs did not play a predominant role in IFN production [50]; however , their role in human innate responses and RVF disease pathogenesis remains unclear . Viral nucleic acid recognition can also occur via the RNA helicases RIG-I and MDA5 [23 , 24] . Both proteins are expressed in the cytoplasm and contain caspase recruitment domains ( CARDs ) as well as a C-terminal region harboring ATP-dependent RNA helicase activity [24] . RIG-I and MDA5 activate downstream signaling via the adaptor MAVS ( mitochondrial anti-viral signaling [51] , also called IPS1 [52 , 53] , CARDif [54] or VISA [55] ) , which relays signals to downstream kinases to trigger IFN gene transcription . RIG-I is required for triggering anti-viral responses against several classes of RNA viruses including ( Flaviviridae , Paramyxoviridae , Orthomyxoviridae and Rhabdoviridae ) [52] , whereas MDA5 is required for the response against picornaviruses like encephalomyocarditis virus ( EMCV ) [56 , 57] . RIG-I senses viral and synthetic RNA containing 5’-triphosphate caps whereas MDA5 detects synthetic poly ( I-C ) in vivo , although the nature of the viral ligand for MDA5 remains unclear [53 , 56 , 58] . Our previous work emphasized the importance of the RNA helicase adaptor MAVS for RVFV induced type I IFN production in primary immune cells as well as for protection against mortality and morbidity during mucosal challenge in mice [50] . We showed that type I IFN responses were mediated through RIG-I in mice and in vitro human cell systems , although MDA5 also played a role at the earliest time points of viral entry . In the present study , we hypothesized that distinct RVFV-associated clinical syndromes are related to differences in early innate host responses to viral infection , and that variation in these host responses may be associated to differences in the makeup of innate immune response pathways . In this first genetic epidemiologic study of human RVF , we sought to examine the association between genes in innate immunity pathways and clinical phenotypes linked to acute RVFV infection . This manuscript describes the genetic associations we discovered among variants within host immunologic pathways that may influence susceptibility to RVFV-associated disease .
The study protocol was approved by the University Hospitals Case Medical Center Institutional Review Board ( IRB ) , Cleveland , Ohio ( No . 11-09-01 ) and the Ethical Review Committee of the Kenya Medical Research Institute , Nairobi , Kenya ( Non-SSC Protocol No . 195 ) . Before participation , written informed consent was obtained from adult study subjects , and parents provided written informed consent for participating children; children over 7 years of age also provided individual assent . Study participants were recruited from six villages located in the Sangailu area of Ijara District , located in Northeastern Province , Kenya ( Fig . 1 ) . All local residents were eligible for participation , with the exception of those living in the area for less than 2 years , and children < 1 year of age , who were excluded . After an initial demographic census was performed , consented subjects were enrolled , surveyed via structured interview for potential RVFV exposure history and past symptoms suggestive of RVF , and examined by a nursing officer with particular attention to current visual acuity and eye disease , as previously described [10 , 59] . Whole blood was collected by phlebotomy ( ∼ 5 mL venous blood samples from persons > 5 years of age and 1 mL from children under the age of 5 ) . Individual sera and associated blood clots were separated and stored frozen at −80°C . The study sample was representative of the local mix of 99% Somali ethnicity , and < 1% Bantu , Indian , or other Asian . Out of 1134 household residents identified , 1128 completed survey questionnaires ( parents served as proxies for young children ) , and 1110 completed a basic physical examination and vision check . A total of 1114 provided blood for anti-RVFV antibody screening , and 1082 provided full data from survey , exam and serology testing . This paper's genetic analysis focuses on 363 individuals from an RVF-endemic study area , Ijara Constituency of Garissa County . The study group included 117 individuals who were antibody-seropositive for RVFV and had DNA available for analysis , and 246 unrelated local control subjects from nearby households or villages . The study subjects ( seropositive and seronegative ) were sampled from 251 households in Sangailu Location ( centered around coordinates 1 deg . 19 min S , 40 deg . 44 min E ) . In this area , residents per household ranged from 1 to 9 ( median = 3 ) , and for our study 1 to 4 persons were tested per household ( average = 1 . 5 per household ) . To test for evidence of past RVFV infection , serum specimens were screened for the presence of anti-RVF IgG via ELISA [4 , 7] . Briefly , high protein binding plates ( Corning ) were coated with RVFV variant rMP-12 viral antigens prepared in Vero cell lysates , and blocked in 5% non-fat dairy milk . Serum samples were diluted 1:100 in PBS/5% milk solution and allowed to absorb for 1 h at 37°C . After washing , a HRP-conjugated secondary anti-human IgG antibody was applied , again in the milk solution at a 1:2000 dilution . Plates were incubated at 37°C for 1 h then developed using a TMB substrate ( Thermo ) and absorbance was read at 405nm . Each sample was run in duplicate , and OD values were normalized to background values from wells coated with uninfected Vero cell lysate and averaged . Samples were considered positive with OD values greater than the mean + 2 SD for pooled control sera obtained from unexposed North Americans . Genomic DNA was isolated from frozen blood clots using a 96 well DNeasy Blood and Tissue kit ( Qiagen ) with some slight modifications . Approximately 500 μL of thawed blood clot material was homogenized in ALT lysis buffer using a mixer mill ( Retsch ) . Proteinase K ( Qiagen ) was added to each tube and incubated at 56°C for 60 min with occasional agitation followed by pulse centrifuged at low speed to pellet debris . Supernatant was removed and DNA was eluted using spin columns according to the manufacturer’s recommendations . Total DNA was quantified using the Quant-iT PicoGreen dsDNA Assay Kit following the manufacturer’s recommendations ( Life Technologies ) . The results were verified on random samples by spectrophotometry ( NanoDrop 1000 , Thermo Scientific ) . Low yielding samples were concentrated by re-precipitating the DNA in ethanol in the presence of 20mg/ml glycogen ( Thermo Scientific ) and re-suspended in TE buffer ( Qiagen ) . This study focused on genes encoding molecules likely to be involved in early innate immune responses to RVFV including: IL6 , IL6R , TLR3 , TLR7 , TLR8 , TRIF ( TICAM1 ) , MyD88 , RIG-I ( DDX58 ) , LGP-2 ( DHX58 ) , MAVS , IFNAR1 , IFNB1 , Mda5 ( IFIH1 ) , CCR5 , DC-SIGN ( CD209 ) , and CFH . Single nucleotide polymorphisms ( SNPs ) within these genes were chosen if they were in the promoter region , the coding region of the gene , or in the untranslated regions ( 3’UTR or 5’UTR ) of the gene , and having at least a 10% minor allele frequency in the Maasai ( MKK ) or Luhya ( LWK ) Kenyan HapMap populations [60–62] . The exception to this was CD209 , where an intergenic SNP was chosen , as no putatively functional SNPs met the allele frequency criteria in HapMap . A total of 48 SNPs were genotyped using the Illumina VeraCode platform . Two SNPs failed quality control because of poor intensity; the remaining SNPs were all in Hardy-Weinberg equilibrium . Allelic frequencies are included in S1 Table . RVFV-related disease phenotypes were defined on the basis of subjects’ self-reported symptoms on the study intake questionnaires . The time period of interest for occurrence was any time during or after the 2006–2007 RVF epizootic in Northeastern Province . The nineteen questions included in the symptom review covered known RVF complications including isolated eye symptoms ( eye pain , scleral injection , poor vision or blurry vision ) , central nervous system symptoms ( photophobia , meningismus , vertigo/dizziness , reduced consciousness , confusion , coma , or seizures ) , symptoms of a hemorrhagic diathesis ( bleeding gums , non-traumatic bruising , hematemesis , hematochezia ) , or non-focal systemic symptoms ( fever , malaise , back ache , nausea , anorexia ) . Presence vs . absence of the clinical phenotype clusters plus positive RVFV serology were the phenotypes of interest . SNP genotypes were analyzed both according to a dominant model with respect to the minor allele , and according to an additive model with increasing counts of the minor allele ( e . g . , AA = 0 , AB = 1 , BB = 2 ) ; these analyses were conducted using PLINK [64] ( http://pngu . mgh . harvard . edu/∼purcell/plink/ ) . If there were fewer than 10 rare homozygotes plus heterozygotes , the Fisher’s exact test was used to define related p values for significance testing . If there were fewer than 10 homozygotes for the rare allele in either the affected or unaffected group for a given trait , the results from the additive model were not considered . In addition , the categorical cluster score trait was analyzed using Kendall’s tau to model increasing severity of the phenotype in association with increasing counts of the minor allele , using dominant and additive models as before; the analyses were conducted using SPSS version 20 . Haplotype association analyses were conducted using the proxy association method in PLINK . The most likely haplotypes were evaluated using the EM algorithm . Two statistical tests are reported: an omnibus test , which compares the distribution of most likely haplotypes in cases versus controls , and haplotype-specific tests , comparing the presence versus absence of that specific haplotype in cases versus controls . Finally , gene-level and pathway-level analyses were conducted using PLINK . Because this was an exploratory pilot study , results significant at α = 0 . 10 are presented . Of note , a large number of associations ( 46 SNPs ) were tested in our exploratory analysis , and none of the results were significant after Bonferroni correction for multiple test comparisons .
The analysis included 363 individuals , of whom 219 ( 67 . 1% ) were female ( Table 1 ) . A total of 117 ( 32 . 2% ) were found to be seropositive for RVFV by ELISA . Subject- reported RVFV-associated symptoms were clustered to system-related and severity-related groupings to facilitate association analysis with genetic polymorphisms ( Table 1 ) . Many individuals ( 72% ) reported at least one non-specific RVF-associated symptom , such as past fever ( 66% ) or malaise ( 51% ) . Meningoencephalitis symptoms ( photophobia , meningismus , vertigo/dizziness , reduced consciousness , confusion , coma , or seizures ) were common and at least one symptom was reported by 44 . 6% of individuals . Individuals reporting at least 3 symptoms of meningoencephalitis , a more stringent classification , were much less common ( 8 . 3% ) . Any hemorrhagic ( HE ) fever-associated symptom was reported by 9 . 4%; three or more HE symptoms were reported by only 4 subjects ( 1 . 1% ) . By contrast , eye disease symptoms were commonly reported ( 38 . 6% for any single symptom; 29 . 5% for 2 or more symptoms ) . Our computer-assisted severity group clustering , based on overall number and type of symptoms per individual , classified 28/177 ( 24% ) seropositive subjects as having been more mildly symptomatic , 44/117 ( 38% ) as moderately symptomatic , and 45 ( 38% ) as more severely symptomatic ( Table 1 ) . To analyze associations of clinical disease with individual SNPs , a single major locus model was first applied , and significance considered at α = 0 . 10 . Using a dominant model , a number of SNPs showed association with clinical phenotypes with p < 0 . 10 ( Table 2 ) . We looked at several polymorphisms in complement factor H ( CFH ) , a gene previously found to be associated with eye disease in published genome-wide association studies [65 , 66] as well as host susceptibility to meningococcal disease [67] . A single polymorphism in CFH ( rs1061147 ) was associated with the presence of any eye symptom ( p = 0 . 059 ) . However , two other SNPs in this gene ( rs1065489; rs3753396 ) were not significantly associated with individual symptoms or with clusters of clinical symptoms ( S3 Table ) . A polymorphism in the gene of the pro-inflammatory cytokine interleukin-6 ( IL-6 ) ( rs2069849 ) was associated with presence of 3 or more non-specific symptoms ( p = 0 . 025 , Table 2 ) . Two SNPs in the 3’ UTR region of the IL-6 receptor were associated with meningoencephalitis or hemorrhagic symptoms with a p < 0 . 10 ( rs4072391; rs7514452 ) . Also detailed in Table 2 , several single polymorphisms in the RNA helicase pathway showed associations with clinical symptoms . A polymorphism in DDX58 ( RIG-I ) ( rs2274863 ) was associated with the subject report of 3 or more ME symptoms ( ME3 , p = 0 . 026 ) , and with the past experience of any ME symptom ( p = 0 . 03 ) . A SNP in the 3’ UTR of the common adaptor MAVS ( rs3746660 ) was significantly associated with the experience of any eye symptom ( p = 0 . 041 ) , any ME symptom ( p = 0 . 059 ) and also with a history of having had two or more eye symptoms ( eye2 , p = 0 . 0598 ) and two different SNPs were associated with positive serology ( rs7262903; rs7269320 ) . In the TLR pathway , TLR7 SNP rs864058 was associated with positive RVFV serology ( p = 0 . 032 ) as well as a history of any hemorrhagic symptom ( HE_any , p = 0 . 02803 ) . TLR8 SNPs rs3747414 and rs5744088 were associated with having three or more meningoencephalitic symptoms ( ME3 ) and positive serology , respectively . The SNP rs6853 , in the adaptor molecule MyD88 which mediates signaling by both TLR7 and TLR8 , showed association with the presence of at least one HE symptom ( p = 0 . 01776 ) . The adaptor TRIF ( TICAM1; rs229151 ) was associated with ME3 ( Fisher’s exact p = 0 . 002 ) , as well as eye2 , any eye , ME_any , and the presence of non-specific symptoms . An additive analysis was next performed to examine the impact of multiple copies of the polymorphisms of interest . Because of the rarity of some of these phenotypes and SNP minor alleles , the dominant model ( as shown in Table 2 ) was more significant , with a few SNPs showing robust associations in the additive model ( Table 3 ) . CCR5 , RIG-I , LGP2 and IFNAR1 all had SNPs associated with clinical symptom traits at the p < 0 . 1 level ( Table 3 ) . Next , we examined the association between the SNPs we selected for study and disease severity as determined by cluster analysis , again considering significance at α = 0 . 10 level . As shown in Table 4 , rank correlation of SNP genotype with severity cluster scores revealed significant associations with exon SNPs in the LGP2 helicase ( p = 0 . 08 ) , the MAVS adaptor molecule ( p = 0 . 047 ) , and the interferon-receptor IFNAR1 ( p = 0 . 013 ) . To better understand the effect of polymorphisms in the overall genes of interest ( versus specific SNP associations ) , we conducted gene- and pathway-level tests of association . In this analysis , all of the single SNP associations within a gene or pathway were included and significance considered at the α = 0 . 10 level . As shown in Table 5 , TLR7 variation was associated with the presence of at least one HE symptom ( p = 0 . 062 ) and TLR3 gene was associated with presence of 2 or more HE symptoms ( p = 0 . 022 ) . In the pathway analysis , the TLR3-TRIF pathway was associated with HE2 ( p = 0 . 035 ) and ME3 ( p = 0 . 0176 ) , the TLR7-MyD88 pathway and TLR8-MyD88 pathways were both associated with HE_any ( p = 0 . 07; p = 0 . 039 ) and the combined TLR7-TLR8-MyD88 pathway was associated with presence of at least one HE symptom ( p = 0 . 0458 ) . The IL6-IL6R pathway was associated with presence of 3 or more non-specific symptoms ( p = 0 . 065 ) . Finally , we conducted a haplotype analysis based on the most likely haplotype phases using the EM algorithm as implemented by the proxy association method in PLINK , considering significance at α = 0 . 10 . First , we found that haplotypes in TLR3 had a different distribution in individuals with and without HE2 ( Table 6 ) ; the overall distribution was significantly different ( p = 0 . 0102 ) . The TG haplotype was associated with a 5-fold increased risk of HE2 ( p = 0 . 00775 ) , and the AG haplotype had a significant protective effect ( p = 0 . 0103 ) . These results confirm the gene-level association between HE2 and TLR3 in Table 5 . Second , we found haplotypes in MAVS were distributed differently in individuals with and without Any3 ( omnibus p = 0 . 0514 ) , with the GCT haplotype resulting in a 1 . 88 increased risk of Any3 ( p = 0 . 00295 ) . Though single SNP analyses of MAVS revealed associations with other clinical phenotypes , there were no associations observed with Any3 , suggesting that an untyped polymorphism on the GCT haplotype may be associated with Any3 .
In this study we examined polymorphisms in human genes of the innate immune system using diverse approaches and demonstrated association with a variety of clinical phenotypes in an ethnically Somali population of long-term residents in a RVFV endemic area . Our analysis included documentation of symptom recall using a structured interview administered by trained study personnel and we acknowledge that there may be inaccuracies with self-reported symptoms including memory lapses , selective recall of more severe symptomatology and other potential biases which are difficult to control in a retrospective study of this nature . Additional epidemiological factors associated with seropositivity and with severity of disease in this population are described elsewhere [11] . We analyzed a total of 46 SNPs in 16 genes ( CFH , IL6 , IL6R , IFIH1 , DDX58 , DHX58 , MAVS , CCR5 , TLR3 , TLR7 , TLR8 , MYD88 , TICAM1 , IFNB , IFNAR1 CD209 ) , and have identified innate immunity pathways that may play an important role in the pathogenesis of clinical RVF associated symptoms . These genes included those for the RNA helicases RIG-I ( DDX58 ) , LGP2 ( DHX58 ) and their common adaptor MAVS ( also called IPS-1 ) as well as endosomal Toll-like receptors TLR3 , TLR7 and TLR8 and their signaling adaptors MyD88 and TRIF . A strong association was observed in our analysis of the inflammatory cytokine IL-6 and its receptor , IL6R . We found association of the IL6 SNP and 3 of the 4 IL6R SNPs in our single gene analysis ( Table 2 ) . In an analysis of all SNPs in the IL6 and IL6R pathway , a significant association was found with non—specific symptoms including fever , anorexia , and backache . Therefore , although our data does not show a strong association between IL-6 and severe RVF symptoms , there is likely a role for IL-6 response , along with those for other inflammatory cytokines , in the pathogenesis of severe RVF . We have previously shown that IL-6 is one of several inflammatory responses to RVFV infection in a murine model of mucosal RVFV infection [50] . It is possible that robust IL-6 responses may lead to a cytokine “storm” via IL-6 receptor signaling , resulting in more severe clinical pathology such as hepatic inflammation , encephalitis , and risk for death . One gene that was of interest , based on previously published GWAS studies was serum complement factor H . Several studies have shown an association of CFH mutations with age related macular degeneration [68–70] as well as with other eye diseases including uveitis [71] . Retinitis is a serious long-term complication of human RVF and we have previously observed prevalence as high as 21% in our study population [10 , 59]; therefore , we hypothesized that CFH may contribute to the pathogenesis of retinal disease . Surprisingly , we did not see a strong association of any of the 6 SNPs in the CFH gene with RVF specific eye disease symptoms , although one SNP ( rs1061147 ) showed weak association with a cluster of general eye symptoms ( Table 2 ) . As other viruses , including a related member of the Phlebovirus group , the ssRNA virus Punta Toro virus ( PTV ) , have been associated with TLR activation [48 , 72] , we decided to look for associations of clinical RVF symptoms with common polymorphisms in TLRs and signaling adaptor molecule genes . Although we did not find an association with individual SNPs in TLR3 , we did find a TLR3 gene-level as well as haplotype association with having had two or more symptoms of hemorrhagic fever . Also , there was a pathway association between TLR3-TRIF SNPs and multiple symptoms of hemorrhagic fever , as well as 3 or more symptoms of meningoencephalitis , which suggests that this innate pathway is important in the pathogenesis of RVFV-associated severe disease . Although these association results were not all significant at α = 0 . 05 , the consistency of association across phenotypes suggests the associations are robust and indicate a role in RVFV pathology . This was not surprising as there has been a clear association between TLR3 mediated innate responses and poor outcomes in a murine model of PTV [48]; however , in a murine model of mucosal RVFV infection we did not see an impact of the TLR3/TRIF pathway on severe disease or type I IFN responses [50] . A recent paper found Toll-7 dependence of RVFV induced autophagy in Drosophila and MyD88 dependence in a human osteosarcoma cell line [73] , although the role of these pathways in human primary immune cell autophagy or immune responses remains unclear . In previous studies we also did not see TLR7 or TLR8 dependence for IFN and other cytokine responses to RVFV , yet in this genetic analysis we do see associations between polymorphisms in human TLR7 and TLR8 with clinical symptoms at the level of individual SNPs , as well as single gene and pathway analysis [50] . We conclude that the impact of the endosomal TLRs , including TLR3 , TLR7 and TLR8 in the innate responses to RVFV and the pathogenesis of severe RVF is unclear . There may be differences between the utilization of endosomal TLRs for viral sensing at the cellular level and the impact of these important innate receptors at the whole organism level . Also , it is increasingly being recognized that there are important differences between mouse and human innate receptor activity in health and disease which may be contributing to the differences that we observe between laboratory studies and analysis of human field collected samples . We have previously shown the importance of the RNA helicases RIG-I and MDA5 as well as the common signaling adaptor MAVS ( also known as IPS-1 ) in RVFV induced IFN responses [50] . In a single gene analysis , we observed a trend towards a significant impact of one individual RIG-I ( DDX58 ) SNP located in the exon of the gene . In the additive model , significant correlations were found between the helicase family member LGP2 ( DHX58 ) and serology and between a 3’UTR SNP in RIG-I and any symptom . Interestingly , in a symptom cluster analysis both LGP2 ( DHX58 ) and MAVS showed association and MAVS also showed association using a haplotype an analysis . These findings point out the challenge of single allelic association testing; whereas using more complex haplotype and cluster analysis demonstrated association of this well established viral innate sensing pathway with clinical phenotypes in RVF . The type I interferon receptor , IFNAR , is formed by class II helical cytokine receptor family members IFNAR1 and IFNAR2 [74–76] . Although the role of type I IFN in host defense to multiple viruses is well established , and the role of IFNAR in modulating susceptibility and severity of disease has been established in multiple models of viral infection , including RVFV [77] , no previous human studies have shown a correlation of genetic polymorphisms in the IFNAR genes with disease phenotype in RVF . Other groups have shown association with polymorphisms in IFNAR1 and Hepatitis B and C infection and disease [78–80] . Our current studies found significant associations between two polymorphisms ( rs2257167 , rs17875834 ) and disease phenotypes using an additive model and phenotype cluster / severity analysis . Our findings point to important innate immune pathways in the pathogenesis of RVF associated symptoms . Polymorphisms in TLR3 , TLR7 , TLR8 , MyD88 , TRIF , MAVS , and RIG-I were repeatedly associated with severe symptomatology , suggesting that these genes may have a robust association with RVFV-associated clinical outcomes . Future studies to further explore the importance of these pathways in RVFV associated disease in different populations as well as correlation with in vivo and in vitro models of RVF are warranted . | The underlying risk factors that lead to severe human Rift Valley Fever disease are unknown , but are likely multi-factorial . Host factors , such as innate immune genetic makeup , are likely important determinants of disease phenotype . This study investigated the association of 46 single nucleotide polymorphisms ( SNPs ) in genes encoding innate immune receptors , signaling pathways or mediators with RVF disease phenotype . Of the 46 SNPs tested , TLR3 , TLR7 , TLR8 , MyD88 , TRIF , MAVS , and RIG-I were repeatedly associated with severe RVF symptomatology , suggesting that these genes may have a robust association with RVFV-associated clinical outcomes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Association of Symptoms and Severity of Rift Valley Fever with Genetic Polymorphisms in Human Innate Immune Pathways |
Dengue is a disease of great complexity , due to interactions between humans , mosquitoes and various virus serotypes as well as efficient vector survival strategies . Thus , understanding the factors influencing the persistence of the disease has been a challenge for scientists and policy makers . The aim of this study is to investigate the influence of various factors related to humans and vectors in the maintenance of viral transmission during extended periods . We developed a stochastic cellular automata model to simulate the spread of dengue fever in a dense community . Each cell can correspond to a built area , and human and mosquito populations are individually monitored during the simulations . Human mobility and renewal , as well as vector infestation , are taken into consideration . To investigate the factors influencing the maintenance of viral circulation , two sets of simulations were performed: ( 1st ) varying human renewal rates and human population sizes and ( 2nd ) varying the house index ( fraction of infested buildings ) and vector per human ratio . We found that viral transmission is inhibited with the combination of small human populations with low renewal rates . It is also shown that maintenance of viral circulation for extended periods is possible at low values of house index . Based on the results of the model and on a study conducted in the city of Recife , Brazil , which associates vector infestation with Aedes aegytpi egg counts , we question the current methodology used in calculating the house index , based on larval survey . This study contributed to a better understanding of the dynamics of dengue subsistence . Using basic concepts of metapopulations , we concluded that low infestation rates in a few neighborhoods ensure the persistence of dengue in large cities and suggested that better strategies should be implemented to obtain measures of house index values , in order to improve the dengue monitoring and control system .
Dengue is currently the most important arthropod-borne disease , affecting around 50 million people worldwide every year , mostly in urban and semi-urban areas [1] . During the last decades , the disease has spread to most tropical countries and has become an important cause of death and hospitalizations by dengue hemorrhagic fever and dengue shock syndrome [2] . South-east Asia is one of the most affected regions , where dengue hemorrhagic fever is a leading cause of morbidity and death among children [1] . In the Americas , a significant increase in dengue incidence has been observed in the last two decades [3] . Dengue can be caused by four distinct but antigenically related serotypes which are mainly transmitted by Aedes aegypti mosquitoes . The wide clinical spectrum ranges from asymptomatic infections or mild illness , to the more severe forms of infection such as dengue hemorrhagic fever and dengue shock syndrome . Infection by one serotype produces long-life immunity to that serotype but does not protect against infection by others [4] . A wide variety of factors influence the spatial and temporal dynamics of mosquito populations and , therefore , dengue transmission patterns in human populations [5] . Temperature , rainfall and humidity interfere in all stages of vector development from the emergence and viability of eggs , to the size and longevity of adult mosquitoes , as well as their dispersal in the environment [6]–[13] . Additionally , factors such as unplanned urbanization , high human population density [14] , the precariousness of garbage collection systems and water supply [15] , [16] - frequent problems in developing countries - favor the proliferation of breeding sites and infection spread . While the development of dengue vaccines is still underway [17] , [18] and assuming that mosquito eradication is a remote possibility , the only alternative of controlling dengue transmission remains in keeping the vector population at the lowest possible levels [19] , [2] . However , the threshold has not been established yet [20] . For dengue control programs to be effective , information on the occurrence of infection and disease in the population are essential . However , as most dengue infections are asymptomatic or unapparent , presenting themselves as non-differential fevers of unknown etiology , surveillance systems based on the monitoring and notification of symptomatic cases have low sensitivity and are not capable of detecting low or sporadic transmission [2] , [21] . Mathematical and statistical models have been developed in order to provide a better understanding of the nature and dynamics of the transmission of dengue infection , as well as predict outbreaks and simulate the impact of control strategies in disease transmission [22] , [16] . Most of these approaches are based on ordinary differential equations or statistical models without exploring the spatial pattern of disease transmission; e . g . [23]–[26] . A summary of the approaches used up to 2006 was reviewed by Nishiura [27] . More recently , models have been developed which incorporate the spatial structure of dengue spread [28]–[32] , as well as models that use complex networks [33] , [34] . Another class of models used to investigate the disease transmission process is that of cellular automata ( CA ) [35]–[39] which are self-reproductive dynamic systems , where time and space are discretized [40] . They are composed of a finite regular lattice of cells , called cellular space , each one with an identical pattern of local connections to other cells , and subjected to given boundary conditions [39] , [41] . Each cell can assume a state , among a finite set of states , which can change at every time-step according to local transition rules ( deterministic or stochastic ) based on the states of the cell and of its neighbors . Models based on cellular automata have the advantage of being spatially explicit in the sense that their elements can be individually tracked in space through which the simulations are carried out . They constitute a class of spatio-temporal dynamics models that allow the development of a virtual environment that creates and explores different scenarios of the dynamics of disease . CA-based models have been used to study the dynamics of dengue fever [42] , [43] , [34] . Santos et al . [42] considered the immature forms of Aedes aegipty in their model to study the patterns of dengue in Salvador city , in the Northeastern coast of Brazil . Ramchurn et al . [43] and Silva et al . [34] used a combination of cellular automata and scale free network ideas to map the evolution of dengue fever . We propose a stochastic cellular automata model that simulates dengue transmission in a hypothetical population , aiming to perform a qualitative analysis of factors that influence disease transmission . Unlike the mathematical models based on differential equations , the proposed CA-based model of diffusion of dengue fever uses heterogeneous rules for human mobility . The role of human mobility in the transmission of infectious diseases has been previously investigated [44]–[46] , including in dengue epidemics [32] . This article investigates the influence of factors related to both humans ( renewal rate and population size ) and vectors ( house infestation index , vector density per human and biting frequency ) in the maintenance of viral circulation for extended periods . The approach was based on the urban shape of the populous Brasilia Teimosa neighborhood within the city of Recife , Brazil . Previous surveys have found high Aedes aegypti infestation rates [19] and prevalence of dengue seropositivity higher than 90% in this area [47] .
The CA-based model consists of two bidimensional square lattices , H and M , both of same size and spatial location , representing the spaces occupied by humans and mosquitoes , respectively . Each cell of H and M corresponds to a lot that can be occupied by a building or be empty . The probability of a lot being occupied by humans is ρh . Each cell with position ( i , j ) that contain humans is represented by a matrix , named H ( i , j ) , where information related to the humans living in the existing building ( intrinsic incubation period ( τi ) , period of infectivity ( τvir ) , status of the individual in relation to disease and infection time ) are stored . Figure 2 illustrates the information stored in a non-empty cell H ( i , j ) of the H lattice . Assuming that Aedes aegypti are usually located in the places where humans reside , the model states that a percentage ρv of non empty cells in the H lattice is infested by mosquitoes . This percentage - called house index ( HI ) – represents the proportion of mosquito-infested buildings . The model considers only Aedes aegypti females . The population of female vectors in each cell is a function of the number of humans and the vector/human ratio within the corresponding cell in H lattice . The vector/human ratio varies from building to building , following a uniform distribution in the interval [0 , maxv] , where maxv is the maximum number of vectors for each human assumed in the model . In the M lattice each cell of position ( i , j ) which contain mosquitoes is represented by a matrix M ( i , j ) that contains information on the existing vector population in the corresponding building . The matrix M ( i , j ) contains the following information pertaining to each mosquito: the extrinsic incubation period ( τe ) , the age of the vector , the state of the mosquito in relation to the disease and the time of infection . Figure 3 illustrates the information in a non empty unit in the M lattice . At the beginning of each simulation , the model generates an initial configuration for the H and M lattices , assuming that the entire population ( humans and vectors ) are susceptible , except for a single randomly chosen infected human . For this initial configuration , the following parameters in each cell are pre-determined: ( 1 ) the human population ( Nh ( i , j ) ) ; ( 2 ) the vector population ( Nv ( i , j ) ) ; ( 3 ) the intrinsic incubation periods and infectivity periods for each human and ( 4 ) the extrinsic incubation periods for each vector . These parameters are summarized in Table 1 . The values assigned to individual parameters τi , τvir and τe are in agreement with literature [20] , [4] , [48] , [55] , [60] . The dynamics of human-mosquito interactions is based on the following rule: every day each mosquito randomly selects one or a few humans to bite , according to a daily frequency of bites bfv ( number of blood meals per day ) . Contact between humans and mosquitoes can occur two ways: local and global contact . The local contact is determined by the search strategy of mosquitoes for human targets which reside nearby . The global contact is determined by the movement of humans , which may come from elsewhere and visit buildings where mosquitoes are found . During the process of interaction between humans and mosquitoes , each human can assume one of four states with respect to each serotype: susceptible ( S ) , exposed ( E ) , infectious ( I ) or recovered ( R ) and each vector can assume one of three states with respect to each serotype: susceptible ( S ) , exposed ( E ) and infectious ( I ) . The duration of the exposed state ( infected but not infectious ) corresponds to the incubation period . If there is a contact between a susceptible human and an infectious vector , the human may become exposed with probability βvh . On the other hand , if an infectious human has contact with a susceptible mosquito , the latter becomes exposed at a probability βhv . The human population was modeled considering a single annual renewal rate ( ρnh ) , as a combination of births , deaths , immigration and emigration . All newcomers are assumed to be susceptible to the dengue virus . The total amount of humans and mosquitoes was kept constant during all simulations . Mosquito survival rate is assumed to satisfy a Poisson distribution . The boundary conditions are periodic , which means that opposite borders of the lattice are connected to each other to form a toric topology [61] . Each time step corresponds to one day . The constant parameters of the model are: Human population size ( Nh ) , percentage of human occupation ( ρh ) , house index ( ρv ) , maximum ratio of vectors per human ( maxv ) , mosquito daily biting frequency ( bfv ) , vector daily survival probability ( ps ) , transmission probability from human to vector ( βhv ) , transmission probability from vector to human ( βvh ) , annual human renewal rate ( ρnh ) and mobility parameters . The parameters that vary by cell , individual or per unit of time are described in Table 1 , wherein function int ( · ) means the integer part of .
Figure 5 shows the evolution of the SEIR framework for humans and the SEI pattern for mosquitoes . Here , as well as in subsequent figures , we considered for simplicity the infected state as the sum of the individuals of exposed or infectious state at time step t . Also , we considered a neighborhood with approximately 10 , 000 inhabitants and house infestation index of 90% . This agrees qualitatively with the patterns of compartmental epidemiological models [39] , . To study the effects of human movement , we conducted simple experiments with different mobility configurations . The stochastic parameters used are those shown in Table 1 and the fixed parameters are given in Table 2 . Other constants were: population size of 10 , 000 , annual human renewal rate of 0% , house index of 70% , mosquito daily bite rate of 1 and maximum ratio of vectors per human of 2 . Figures 6 to 8 illustrate the spatial spread of dengue fever in humans and mosquitoes through time . Each cell of the lattices in these figures corresponded to a building or a empty lot and the colors represent cell states , whose meanings are described in Table 5 . The wave front is clear when human mobility is not considered ( Figure 6 ) . In the case of concentrated mobility in public locations ( Figure 7 ) , small and clear foci of disease emerge over time . As human mobility becomes more homogenous , transmission foci become less clear . The different propagation speeds of the disease can be observed in Figure 9 . The human movement rates and patterns influence the shape of epidemic curves: the higher and more homogeneous the mobility , the higher the amplitude of the epidemic curve and more rapid its duration . Indeed , it was found that a human mobility rate ρmob of 10% would reduce the duration of the epidemic to almost half . Figure 10 shows the epidemic wave-front pattern for different annual human renewal rates ρnh in a population with approximately 10 , 000 inhabitants . While for ρnh = 0% the epidemics ended after 18 months , the viral transmission was kept active for non zero renewal rates . In fact , the amplitude of the viral transmission after the epidemic outbreak ( in the second phase ) was related to ρnh . However , we found that the renewal rates had no effect in the duration and amplitude of the initial outbreak ( not shown ) . The periodicity of the epidemics is shown in Figure 11 . Fixing the annual human renewal rate of 3 . 2% in an area with 10 , 000 inhabitants and considering a house index of 90% , we can note the periodic behavior of the epidemics and the endemic state . After the first major epidemic , small outbreaks occur at intervals of about four years . This pattern of periodicity is consistent with patterns observed in countries of Southeast Asia and in America [1] , [66] , [67] . For the first set of simulations , using the range of parameters described in Table 3 , the percentage of replications in which an epidemic outbreak occurred and viral transmission in the first six months was over 70% in all sets of 50 replications . Figures 12 and 13 illustrate the proportion of cases , among those which the virus was transmitted in the first six months , for which transmission was maintained for a long period after the appearance of the serotype , for both biting frequencies bfv = 1 and bfv = 1 . 5 , respectively . The results showed that for both frequencies of bites and for all population sizes , the human renewal rate of 1% was not sufficient to maintain viral transmission for more than three years , while for 2% of human renewal , in very few cases , viral circulation was maintained for many years . The viral transmission was not sustained with the combination of small human population with low human renewal . In order to maintain viral transmission for a long period it was necessary that at least one of these parameters were not low . In the case of 8 , 000 inhabitants and 5% of annual human renewal rate , the chance of sustained viral circulation was higher than 50% ( for both biting frequencies ) . Therefore , we chose these values for the second set of simulations . Figure 14 represents the percentage of cases that presented viral transmission in six months for each set of 200 replications with parameters of Table 4 . The percentage of cases of viral transmission in six months decreased with decreasing house index . Nevertheless , we considered the cases with small values of house index ( less than or equal to 10% ) and found that viral transmission was not sustained for more than one year ( not shown ) . Figures 15 and 16 show the percentages of cases among those with initial viral transmission , for which transmission was maintained for extended periods . The results showed that the combination of bfv = 1 . 5 with maxv = 2 and high house index was sufficient to maintain a high probability of transmission for 7 years . However , for both frequency of bites and maxv = 1 , in very few cases it was possible that house index between 20% and 30% maintained viral transmission active for at least five years . The results also show that the vector/human ratio influences the maintenance of viral transmission: the lower this value , the lower the viral transmission persistence .
Noting the limitations inherent to any mathematical modeling , we discuss the problem of viral transmission maintenance between successive epidemic periods . This question was motivated by the high incidence rates of dengue in densely populated areas of Recife [19] in 2004 and 2005 . For this , we created a stochastic cellular automata model to represent the dynamics of dengue transmission in a community in which important characteristics were considered: human mobility and human renewal . Human movement transcends the spatial and temporal scales , with different influences on disease dynamics , because it influences the exposure to other individuals and thus the transmission of pathogens [44] . The simplest and traditional mathematical models for the spread of infectious diseases assume homogeneous mixing among individuals and although such models are robust , they do not reflect reality . Here we presented a non homogeneous mobility in the sense that every day most people visit public locations containing mosquitoes . Although the model considers mosquitoes in households , public locations are the main source of disease spread . We showed that human movements , concentrated or not in public locations , are responsible for the rapid development of the epidemic , reaching a very large amount of people . The other feature considered , human renewal , is responsible for the continuous increase of susceptible humans , and therefore for maintenance of viral transmission and the recurrence of outbreaks . The simulations qualitatively repeated the cyclical pattern of dengue epidemics [1] , [66] , [67] . With respect to the investigation of the maintenance of viral transmission for extended periods , the question to be answered was: Since the number of susceptible individuals in a naive population is virtually exhausted after an epidemic outbreak , how can the virus remain active between outbreaks ? This issue was exhaustively addressed in different scenarios , where we analyzed the influence of some human and vector factors in the maintenance of viral circulation during seven years , a sufficient period for equilibrium of viral transmission [48] . The results of numerical experiments showed that with high house index values combined with high/moderate vector/human ratio , viral transmission was maintained for long periods , whereas it was not when considering the combination of small human population and low human renewal rates . The latter combination led to disease extinction in the model . Therefore , for the maintenance of viral transmission it was necessary that at least one of these parameters were not low . The extinction situation also happened when we considered house index values below 10% , for human populations with approximately 8 , 000 inhabitants in all cases of vector/human ratio . However , the SET model also showed that viral transmission is possible for several years ( with low probability ) considering low house index ( between 20% and 30% ) , moderate ratio of vector per human ( 0–1 vector per person ) and small human populations ( approximately 4 , 000 people ) . For these cases , we believe that the random combination of factors in the initial configuration of the CA-based model allowed the virus to circulate for many years . The results of the SET model are consistent with findings from the model of Newton and Reiter [58] , who concluded that viral transmission can be maintained with low house index . As the neighborhoods of large cities generally have populations of at least 8 , 000 inhabitants , the model suggests that it is possible that in these cities a small percentage of its neighborhoods have the potential to sustain the virus for extended periods . For example , considering a hypothetical metropolis of 6 million inhabitants with house index of 30% and 750 neighborhoods of approximately 8 , 000 inhabitants , the SET model showed that about 1 . 5% of the city's neighborhoods sustain viral circulation for 5 years ( or roughly 11 neighborhoods ) . The persistence of viral circulation is in agreement with the classic notion of extinction risk and persistence in metapopulations [68] . If ρe is the probability that one of N independent and identical occupied patches becomes extinct in a certain period of time , the probability that all of them become extinct is ( ρe ) N , thus the probability of persistence of at least one patch is 1- ( ρe ) N . For the hypothetical metropolis considered , the estimated probability for the persistence of viral transmission in 5 years was 0 . 015 , that means ρe = 1-0 . 015 = 0 . 985 . As we have N = 750 neighborhoods , a number sufficiently large so that 1- ( ρe ) N is nearly 1 , the persistence of viral transmission in at least one neighborhood is guaranteed . To illustrate , Figure 17 shows the relation between the probability of persistence of at least one patch ( disease persistence ) in five years and the number of patches N , for three values of ρe . In the case of ρe = 0 . 985 , for N below 15 neighborhoods ( corresponding to cities with less than 120 , 000 inhabitants ) , the probability of disease extinction is high . If we had N = 305 neighborhoods ( which corresponds to a city with 2 , 440 , 000 inhabitants ) , it was sufficient to guarantee 99% of chance of persistence of at least one patch . In fact , a simple analysis of the expression 1- ( ρe ) N says that the greater the value of ρe , the greater the value of N to ensure a high probability that at least one patch persists . This can be examined from the viewpoint of the relationship between the number of neighborhoods N and the combination of house index and vector/human ratio: the lower the vector infestation , the greater the value of ρe , so the greater the value of N to ensure the disease persistence . The same rule applies to the converse: the greater the vector infestation , the lower the value of N to guarantee disease persistence . Moreover , interactions between individuals from different neighborhoods ensure a possibility of disease transmission to other districts [45] . Thus , there is a likelihood of rotation of the neighborhoods with viral circulation . This theory explains the maintenance of viral transmission in large cities . However , in real situations , the vector population fluctuates according to a combination of meteorological factors [6]–[8] , [69] , [70] , which modulates the number of vectors in some seasons or years , although the house index virtually does not change; Figure 1 and [19] . On the other hand , in big cities where dengue is endemic , while some districts have low infestation by vectors , others have greater abundance ( thus increasing the likelihood of maintaining viral transmission for extended periods ) . The latter will ensure sustaining the population of mosquitoes even at low levels , despite the occurrence of seasonal variations in vector population . This occurred in some neighborhoods of the city of Recife in 2004 and 2005 , where evidence showed that the vector population was not eliminated entirely by natural factors [19] . In practice , house index values should be zero or very close to zero in order to eliminate viral transmission [56] , [71] . The SET model also recommends the implementation of control measures to drastically reduce the vector infestation , mainly for large cities . Moreover , the model suggests that measured house index values from field data are incorrect , since the circulation of the virus has been found even in situations with measured house index below 3% [19] , [72] , [73] . In a survey in a district of the city of Recife in the years 2004 and 2005 [19] , a high density of Aedes aegypti eggs was found in the region ( site 1 in part B of Figure 1 ) , while the house index measured by health workers based on larval survey in the same neighborhood and at the same time was 0% . This apparent contradiction can be explained when considering the method of calculating the house index . The big problem with regard to the values of this index obtained from field data , is that the methodology used in most programs for controlling Aedes aegypti , based on larval survey , is not suitable for measuring the abundance of mosquitoes [74] , disguising the true value of the house index . Thus , in agreement with Regis et al . [74] , the SET model suggests that better strategies should be implemented to obtain the house index , in order to ensure better efficiency in the control programs of Aedes aegypti . | Dengue is the most rapidly spreading mosquito-borne viral disease in the world and approximately 2 . 5 billion people live in dengue endemic countries . In Brazil it is mainly transmitted by Aedes aegypti mosquitoes . The wide clinical spectrum ranges from asymptomatic infections or mild illness , to the more severe forms of infection such as dengue hemorrhagic fever or dengue shock syndrome . The spread and dramatic increase in the occurrence of dengue cases in tropical and subtropical countries has been blamed on uncontrolled urbanization , population growth and international traveling . Vaccines are under development and the only current disease control strategy is trying to keep the vector quantity at the lowest possible levels . Mathematical models have been developed to help understand the disease's epidemiology . These models aim not only to predict epidemics but also to expand the capacity of phenomena explanation . We developed a spatially explicit model to simulate the dengue transmission in a densely populated area . The model involves the dynamic interactions between humans and mosquitoes and takes into account human mobility as an important factor of disease spread . We investigated the importance of human population size , human renewal rate , household infestation and ratio of vectors per person in the maintenance of sustained viral circulation . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology/epidemiology",
"computer",
"science/applications",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2011 | Modeling the Dynamic Transmission of Dengue Fever: Investigating Disease Persistence |
Integrating data from multiple regulatory layers across cancer types could elucidate additional mechanisms of oncogenesis . Using antibody-based protein profiling of 736 cancer cell lines , along with matching transcriptomic data , we show that pan-cancer bimodality in the amounts of mRNA , protein , and protein phosphorylation reveals mechanisms related to the epithelial-mesenchymal transition ( EMT ) . Based on the bimodal expression of E-cadherin , we define an EMT signature consisting of 239 genes , many of which were not previously associated with EMT . By querying gene expression signatures collected from cancer cell lines after small-molecule perturbations , we identify enrichment for histone deacetylase ( HDAC ) inhibitors as inducers of EMT , and kinase inhibitors as mesenchymal-to-epithelial transition ( MET ) promoters . Causal modeling of protein-based signaling identifies putative drivers of EMT . In conclusion , integrative analysis of pan-cancer proteomic and transcriptomic data reveals key regulatory mechanisms of oncogenic transformation .
Central to the understanding of cancer cells are their epithelial or mesenchymal traits , which are governed by epithelial-mesenchymal transition ( EMT ) . Cells that have undergone EMT display increased invasiveness and metastatic potential [1] . The transition is reversible , in that cells can also undergo mesenchymal-to-epithelial transition ( MET ) [2] . This plasticity plays a role in cancer progression and metastasis by increasing the capacity of cancer cells to invade and colonize at remote tissue [3] . EMT is thought to be governed by a few master regulators that induce epigenetic and transcriptional reprogramming , affecting the expression of multiple downstream genes [4] . The transition is characterized by the down-regulation of E-cadherin , which has been the gene most extensively studied , resulting in disruption of adherens junctions [5] . The inhibition of E-cadherin expression is known to be mediated by the transcription factor Snail [6] . At the CDH1 loci , Snail recruits protein complexes containing histone deacetylases ( HDACs ) that deacetylate H3 and H4 histones , silencing the transcription of E-cadherin [7] . Other key transcription factors implicated in EMT are ZEB1/2 and TWIST [8] . The regulation of EMT-TFs by miR200 and miR34 constitutes a double-negative feedback mechanism [2] , predicting a bistable system with binary transition between cellular states . Essentially , EMT is controlled by multiple interconnected regulatory networks , which include transcriptional and post-transcriptional mechanisms . Due to high regulatory complexity , proteomic and transcriptomic technologies provide an opportunity to obtain a more global understanding of EMT and MET , while possibly discovering additional molecular mechanisms with implications for targeted cancer therapeutics . The reverse phase protein array ( RPPA ) is a high-throughput proteomics method that utilizes antibody binding to quantify protein expression and post-translational modifications including phosphorylation , acetylation , and protein cleavage . Compared to mass spectrometry proteomics , RPPA has higher sensitivity for low-abundance proteins and is characterized by increased throughput; however , it relies on high-quality antibodies , so it cannot identify proteins or post-translational modifications de novo [9] . Using RPPA , 736 cancer cell lines have been assayed for 450 proteins and phosphoproteins covering well-established cancer-related signaling pathways [10] . This data complements prior efforts to characterize basal mRNA expression across many of the same cancer cell lines for different cancer types [11] . In addition , tumor samples have been characterized by similar RPPA experiments for samples from the Cancer Genome Atlas ( TCGA ) [12] , which are publicly available through the Cancer Proteomics Atlas ( TCPA ) [13] . Most genome-wide studies of EMT in cancer cell lines and tumors have focused on particular cancer types . Combining EMT signatures based on cell lines and tumors of multiple cancer types can identify general transcriptomic features of EMT in cancer cells , which are clinically relevant for multiple types of cancer . More recently , transcriptomic data from TCGA and Cancer Cell Line Encyclopedia ( CCLE ) have been used to define a pan-cancer EMT signature based on the expression of E-cadherin and Vimentin alone [14] . In this study , we integrate transcriptomics and RPPA data from multiple cancer cell lines to study pan-cancer cellular states associated with EMT .
The Cancer Cell Line Encyclopedia ( CCLE ) contains 1037 cancer cell lines with profiled transcriptomes [11] , and the MD Andersen Cell Line Project ( MCLP ) contains 736 cancer cell lines profiled by RPPA [10] . Out of these cancer cell lines , 381 have both available RPPA and microarray data . RPPA measurements are available for 450 proteins and phospho-proteins , of which 311 genes can be matched to mRNAs measured in CCLE . In the RPPA data , 79 proteins are measured both at the basal expression and phosphorylation levels ( Fig 1A ) . To our knowledge , this data set , although far from genome-wide at the protein level , represents the largest collection of cancer cell line data measured at the transcriptional , translational , and post-translational levels . Transcriptional profiling of human tumor samples accurately predicts the tissue of origin for common cancer types [15] . This suggests that despite oncogenic transformation , cancer cells retain cellular identity and molecular features of their ancestral cell lineage , which is a key confounding factor in pan-cancer analyses . To assess how cancer cell lines relate based on transcript and protein expression , we visualized distances between cell lines using the t-Distributed Stochastic Neighbor Embedding ( t-SNE ) method [16] . As expected , the cell lines are clustered predominantly by their tissue of origin for both RPPA and transcriptomic data ( Fig 1B , top ) . Cancer types with ill-defined or multiple clusters included breast and ovary as well as cell lines from the most common targets of metastasis: liver , lung , and bone . Nonetheless , most cell lines were correctly classified by nearest neighbor classification ( S1 Fig ) , even when the t-SNE perplexity parameter was varied widely . In addition , independently from the t-SNE analysis , Gap statistics [17] from the average linkage hierarchical clustering at different tree cuts resulting in clusters of varying cardinalities displays similar grouping of cell lines . Furthermore , the inflection points in the number of clusters formed at different thresholds demonstrate that the cell lines are organized into distinct clusters based on expression vector similarity ( S3 Fig ) . Next , we asked: to what extent the RPPA and transcriptomic data is concurrent . Although the cell line distances for protein and mRNA data were correlated ( r = 0 . 58 ) , they were surprisingly different for particular pairs of cell lines ( Fig 1C ) . To quantify these differences and rank cell lines with the most characteristic protein or transcript signatures , we calculated the residuals of a linear regression between the protein and mRNA cell line distances . According to this model , we found that the 49 breast cancer cell lines had the largest distances at the protein vs . mRNA levels compared with other sets of cells from other tissues of origin , suggesting that the RPPA measurements better distinguish breast cancer subtypes . Interestingly , hierarchical clustering based on the RPPA data supports three luminal breast cancer subtypes compared with two subtypes identified by transcriptomic data ( Fig 1D ) here and elsewhere [12] . More broadly , combining data from microarray and RPPA data strengthened the cancer type clustering of cell lines , further suggesting that these measurements of global cellular states are complementary . Overall , clustering of cell lines by transcriptomic and RPPA data is consistent with some cancer types being well-defined and others spanning a wide spectrum of molecular states , while retaining few but important distinguishing differences at both the cancer type and subtype level . To test the hypothesis that EMT governs the molecular states of cell lines across cancer types , we colored the z-scores of E-cadherin expression on the points on the t-SNE maps . For both transcripts and proteins , the cancer cell lines were globally organized by a gradient of E-cadherin expression ( Fig 1B , middle ) . This organization indicates a central role for EMT in characterizing the molecular states of cancer cell lines . Most cancer types associated with common carcinomas had cell lines that spanned this E-cadherin gradient , with lung and breast cancer displaying the largest span . In contrast , cell lines from skin , bone , blood , and kidney were exclusively found in regions with low E-cadherin expression; whereas cell lines from pancreatic and large intestine cancers were found mostly in regions with high E-cadherin expression with only few cell lines expressing E-cadherin at low levels . To ensure the robustness of these findings , we ran independent t-SNE analyses by varying the perplexity parameter , which recapitulated both the E-cadherin gradient and the cancer type-specific clusters ( S2 Fig ) . In comparison , principal component analysis ( PCA ) yielded less separation of cancer types and a less prominent gradient of E-cadherin expression ( S4 Fig ) . Using cell line annotations from the Catalogue of Somatic Mutations in Cancer ( COSMIC ) , we found no obvious association to whether the cell lines were derived from primary or metastatic tumors ( Fig 1B , bottom ) . This suggests that the arrangement of cell-lines on the t-SNE plots , and thus global expression at the mRNA and protein levels , is dominated by tissue of origin much more than metastatic status . Nonetheless , we propose that collections of pan-cancer cell lines can be used to study aspects of EMT related to E-cadherin expression , which is also clearly bimodal ( Fig 2A ) . Oncogenesis is a multi-step process by which cells acquire cancerous traits , which often mimic physiological cellular processes such as embryogenesis [14] . Such processes are governed by molecular switches that turn on or off coordinated cellular programs . Hence , analyzing the bimodality of protein expression can potentially illuminate cellular states of oncogenesis . To evaluate this idea , we fit a univariate two-component Gaussian mixture model to each RPPA measurement using the expectation-maximization ( EM ) algorithm . We evaluated bimodality against unimodal distributions using the Bayesian Information Criterion ( BIC ) . Out of the 450 antibody-based RPPA measurements , 260 were bimodal across 736 pan-cancer cell lines ( Fig 2B , S1 Table ) . Among the most bimodal proteins were E-cadherin , Claudin7 , and Rab25 , all of which have been previously associated with EMT or MET [18] . However , because of the preponderance of tissue-specific signatures among pan-cancer cell lines , bimodal protein expression could more simply be explained by cell type-specific expression . For example , LCK was highly expressed only in a subset of blood cancer cell lines ( Fig 2C ) concordant with its specific roles in T cell development [19] . To account for residual effects of ancestral cell types , we quantified the tissue diversity of the cell lines assigned to the low- and high expression states by the Shannon entropy of the tissue distributions . We then excluded the lower tertile of the minimum tissue entropy of the low- and high expression states . This approach yielded 172 bimodal proteins and phosphosites ( Fig 2B , S1 Table ) . Out of these , 90 had balanced bimodal distributions including E-cadherin , Claudin7 , and Rab25 , indicating common pan-cancer oncogenic switches , while 82 were classified as rare transitions ( Fig 2D ) . This filtering and classification is likely prone to false positives due to other confounding factors such as different stages of the circadian clock at time of measurement . Compared to non-bimodal proteins , the proteins associated with the common switches were uniquely enriched for 107 Gene Ontology terms ( p < 0 . 05 , after Benjamini-Hochberg correction ) many of which can be linked to metastasis and invasion ( Fig 2E ) . To identify whether the observed bimodal protein expression across cancer cell lines correlate with transcriptional regulation , we evaluated the bimodality of matching transcripts from CCLE ( Fig 3A ) . We then defined bimodal coupling coefficients between mRNA and protein measurements as the Spearman’s correlation between the posterior probabilities of the mixture model . Overall , 14 . 0% of proteins measured in MCLP had highly coupled ( rb > 0 . 5 ) bimodal expression of mRNA and protein . Slightly fewer proteins ( 10 . 8% ) were uniquely bimodal only at the protein level , including important cancer-related proteins such as MEK1 , mTOR , E2F1 , TTF1 , EIF4G , and JAB1 . Hence , these proteins are bimodally expressed due to post-transcriptional regulatory mechanisms such as protein translation and degradation . In addition , we compared the bimodality of proteins and their phosphosites as measured by antibody binding in the RPPA data ( Fig 3B ) . Here , we found weaker bimodal coupling , indicating that phosphosignaling leading to bimodal phosphorylation is mostly independent from basal protein expression . Interestingly , bimodal HER2 phosphorylation at Y1248 was moderately coupled to HER2 protein expression ( rb = 0 . 46 ) , most likely due to autophosphorylation on increased dimerization at higher expression [20] . The bimodal EMT proteins E-cadherin , Claudin7 , and Rab25 all had high bimodal mRNA-protein coupling ( Fig 3C ) , confirming that these EMT switches are mostly determined by transcriptional regulation . Nonetheless , 30 cancer cell lines had high expression of the E-cadherin transcript but low protein expression ( Fig 3D and 3E ) , suggesting that E-cadherin could be translationally or post-translationally controlled in some cellular contexts . Among these cell lines , 3 out of 4 CDH1 genotyped cell lines in COSMIC had either nonsense ( MDA-MB-453 and HT115 ) or frameshift ( MDA-MB-134-VI ) mutations in CDH1 , which validate our ability to identify effects on E-cadherin translation . Inactivating mutations in CDH1 are frequently observed in breast and gastric cancers with cancer type-specific mutational patterns and are associated with loss of cell-cell adhesion and increased cell motility [21] . The nature of the low E-cadherin protein expression in the other 26 cell lines remains unknown , but likely includes inactivating mutations and possibly translational or post-translational regulation . Transcriptional mechanisms that determine molecular switches are regulated by upstream signaling , such as phosphorylation cascades , which leads to coordinated expression of multiple genes . To detect candidates for such signaling and further characterize the EMT-related states in cancer cell lines , we analyzed the network of bimodal coupling coefficients among bimodal protein and phosphosites associated with high tissue diversity . We first trimmed the protein network by including only significant bimodal coupling coefficients ( FDR < 5% , Bonferroni ) with |rb| > 0 . 3 . This yielded a network of 172 protein nodes connected by 507 edges , from which network communities were defined based on the leading non-negative eigenvector [22] . In total , we detected 8 protein communities that likely reflects shared underlying signaling or cellular events ( Fig 4A ) . One community ( EMT1 ) was clearly linked to EMT , containing E-cadherin , β-catenin , Fibronectin and Twist among several other EMT-related proteins ( Fig 4A ) . E-cadherin was connected to EPPK1 , INPP4B , Stathmin , Jagged1 , UGT1A , and PDCD1L1 , indicating that these might be involved in EMT . The strong bimodal coupling to EPPK1 could help explain why loss of E-cadherin is associated with migratory phenotypes and not just loss of cell-cell adherence; in mice keratinocytes , EPPK1 knockout cells exhibit faster migration and increased wound healing [23] . E-cadherin was also positively coupled to the phosphosite EGFR pY1068 , which was in turn positively coupled to SRC pY416 and STAT3 pY705 , suggesting a role for phosphorylation of these sites in EMT . Strikingly , all detected communities contained multiple proteins with known mechanisms linking them to EMT but also identified potentially undiscovered components ( Fig 4A ) . The dispersion of these EMT-related proteins among the identified protein communities suggests that they are either part of separate biological processes , or that their involvement in EMT depends on cancer subtypes . Another intriguing possibility is that the multiple protein communities associated with EMT reflect partial cellular states in-between epithelial and mesenchymal phenotypes ( Fig 4B ) . In support of this idea , P-cadherin has previously been suggested as a marker of metastable EMT states [24] . Here we find P-cadherin in the EMT2 community ( Fig 4A ) . In addition , Claudin7 is highly coupled to E-cadherin ( rb = 0 . 70 ) , but found in a separate community ( EMT3 ) , along with Rab25 and N-cadherin . Looking closer at this correlation , cell lines had high E-cadherin and low Claudin7 expression but not conversely ( Fig 4C ) . Two other communities are identified , EMT4 and EMT5 . The EMT4 cluster contains key cell-cycle transcription-factors such as FoxM1 , Cyclin-B1 , and Elk1 , together with protein kinases that are known to positively regulate their activity , including PLK1 , MAPK , and MEK1 . Consequently , this cluster indicates changes in cell proliferation regulation . The EMT5 cluster contains a clique made of 3 protein kinases known to regulate the protein translation machinery: RICTOR , P70S6K , and PDK1; and S6 a key protein in the 40S ribosomal subunit . Hence , this cluster likely represents changes in protein translation activity related to overall cell growth . It should also be noted that highly studied proteins and phosphoproteins such pAKT , Cyclin D1 , PTEN , and PKC are known to be central to many other pathways , not just to EMT . Hence , labeling all identified clusters as EMT clusters needs to be considered with such general functions in mind . Altogether , it is possible that the protein communities EMT1 and EMT3 may reflect a two-step transition ( Fig 4B ) . In summary , the quintessential EMT marker E-cadherin was found centrally in a large protein and phosphosite network community with clear associations to known EMT markers . For these reasons , we focused subsequent analyses around the expression of E-cadherin , arguing that this approach reflects core aspects of EMT that are invariant across cancer types . Because E-cadherin is primarily transcriptionally controlled , we next sought to characterize the coordinated transcriptional program associated with E-cadherin down-regulation . First , we defined an EMT signature based on the bimodal coupling ( |rb| > 0 . 5 ) between E-cadherin protein expression and transcriptomic measurements from CCLE , resulting in 239 transcripts—215 positively and 24 negatively coupled ( Fig 5A , S2 Table ) . To our knowledge , these 239 genes include many novel epithelial and mesenchymal markers , while recovering many known EMT markers previously described ( Fig 5C ) , for example , Axl which was reported for non-small cell lung carcinoma [25] , or KPNA2 for ovarian carcinoma [26] . The preponderance of positively coupled transcripts suggests that the EMT signature is predominantly characterized by down-regulation of genes governing epithelial traits rather than by gain of mesenchymal traits . Nonetheless , the bimodal coupling coefficients were shifted towards negative values ( Fig 5B ) and we did find negatively coupled mesenchymal markers such as ZEB1/2 [4] . To further characterize the EMT signature , we performed enrichment analysis on the epithelial markers ( Fig 5B ) . Enrichment analysis [27 , 28] for Gene Ontology ( GO ) cellular components and biological processes clearly demonstrated epithelial phenotypes ( Fig 5D ) . We also found enrichment for localization to the perinuclear region , which is a cytosolic region next to the nuclear envelope with largely unknown composition and biological function . This suggests that the epithelial markers can be used to prioritize spatial cellular regions not widely considered to be affected by EMT such as the perinuclear region , where components of endocytosis aggeregate , although it is well established that endocytosis is central to cell migration . Enrichment for TF binding , using aggregated results from ChIP-seq studies [29] , identified SOX2 , SMAD2-4 , TP63 , GATA3-4 , and GATA6 , which likely act to down-regulate epithelial genes during EMT ( Fig 5E ) . The identified enriched TFs OCT4 , SOX2 , NANOG , KLF4 , and ESRRB are all known to be essential for maintaining pluripotency of human and mouse embryonic stem cells [30] . These TFs bind to super-enhancer regions and through the Mediator complex [31] . Therefore , large parts of the observed transcriptional bimodality could be explained by TF co-operation at super-enhancers resulting in switch-like regulation at numerous genomic loci . At the CDH1 loci , ENCODE ChIP-seq data supports the involvement of super-enhancers since the loci is marked by high H3K27ac correlated with E-cadherin expression ( S5 Fig ) . Previously , super-enhancers have been proposed to control partial EMT through the putative master regulator TFs ETS2 , HNF4A , and JUNB [32] , the first two of which we also identified through the TF enrichment analyses . Taken together , pan-cancer bimodality uncovers oncogenic states and regulatory mechanisms of EMT and MET . Gene expression-based , high-throughput screening is a promising approach to identifying small-molecule candidates that can reverse or mimic changes in expression observed in transition to a disease state [33] . To detect small molecules that would maximally push cells toward the EMT or MET expression state , we queried the EMT signature against signatures from ~20 , 000 small-molecule perturbations of ~50 human cell lines generated by the library of network-based cellular signatures ( LINCS ) project L1000 dataset [34] . We searched for small molecules that down-regulate epithelial genes and up-regulate mesenchymal genes , resulting in candidate EMT inducers ( Fig 5E ) . Small molecules with the opposite effects were interpreted as MET inducers . Strikingly , most small molecules predicted to induce EMT were HDAC inhibitors , whereas most small molecules predicted to induce MET were kinase inhibitors . The identified HDAC inhibitor Trichostatin A has been shown to induce EMT in prostate cancer cells through modification of H3 near promoters of EMT-related genes [35] . Of the candidate MET inducers , Selumetinib , Trametinib , and PD-0325901 are thought to inhibit MEK , while Saracatinib and Dasatinib to inhibit SRC among other kinome targets . In agreement with these findings , a prior high-content chemical screen aimed at identifying inhibitors of EMT has predominantly identified other similar kinase inhibitors based on cell growth and migration assays [36] . Hence , in summary , caution should be placed in utilizing HDAC inhibitors as therapeutics due to their putative potential to enhance EMT as predicted by chemogenomics screening . The bimodal coupling model we implemented to analyze EMT is essentially correlative and hence not causal . However , establishing causal interactions based on RPPA data is challenging without time-series or direct perturbation data such as gene knock-downs or knock-outs [37] . With sufficient sample size and coverage of diverse cell lines , it is in principle possible to identify causal , regulatory interactions between measured signaling components . Despite not satisfying the observation that cell signaling regulatory networks contain feedback loops [38] , learning Bayesian network learning algorithms can be applied to construct causal models of cellular regulatory networks , including cell signaling networks , from observational data [39–41] . To infer causal relationships among proteins and phosphosites measured by RPPA , we used a Fast Greedy Search algorithm to estimate a Bayesian network over all 450 RPPA measurements ( Fig 6 ) . Based on the resulting directed acyclic graph , we calculated betweenness centrality , subgraph centrality , in-degree , and out-degree for each analyte . Bimodal phosphosites overall had higher betweenness centrality ( p = 0 . 036 , t-test ) . By considering measures of network influence , several proteins and phosphosites were identified as promising candidate drivers ( Fig 6A and 6B ) . In particular , SRC pY416 had the highest out-degree . This phosphosite is known to be highly predictive of patient survival [42] . Furthermore , we analyzed the network neighborhood of E-cadherin , Claudin7 , and Rab25 , which spanned the proposed two-step transition: EMT3 followed by EMT1 ( Fig 6C ) . Using a hierarchical network layout algorithm , E-cadherin was found downstream of Claudin7 and Rab25 concurrent with EMT3 preceding EMT1 in cancer cell lines . In addition , several proteins and phosphosites upstream of the EMT markers were plausible oncogenic drivers for specific cancer types , supported by prior reports . For example , MACC1 is associated with pancreatic EMT and metastasis [43] , which the analysis found for both pancreatic and pleural cancer cell lines , while suggesting the opposite effect in lung , endometrium , and upper digestive tract cancers . Lastly , we also found LKB1 , CHK1 , and Stathmin upstream of EMT markers . The inactivation of LKB1 , which is frequently mutated in lung adenocarcinomas , induces EMT in lung cancer cells through activation of ZEB1 [44] , whereas CHK1 mediates DNA damage response as part of EMT by stabilizing ZEB1 [45] . Inhibiting the microtubule destabilizer Stathmin impedes EMT by increased microtubule formation [46] . In conclusion , inferred Bayesian protein networks based on pan-cancer cell lines can potentially identify key drivers of EMT . To validate the causal models , we carried out bootstrapping for 200 iterations and considered the average network for both cell line and tumor data ( Supporting Information , S3 Table , Fig 6E ) . Although the reproducibility of particular edges in the ensemble of Bayesian networks was relatively low for the cell line data , the connection from Claudin7 to Rab25 was present in 48% of the networks , and in all the networks inferred from tumor samples . In contrast , the statistical reproducibility of the tumor networks was higher , most likely due to the larger sample size ( n = 3 , 161 ) . Overall , the average Bayesian networks were significantly correlated between cell line and tumor data ( Fig 6D ) . The bimodal coupling coefficients , however , were lower in the tumor data , indicating that bimodal expression is less pervasive in tumors compared to cell lines . This result might indicate that tumors , more-so than cell lines , contain mixtures of cell types that are in multiple cellular states . Overall , the learning Bayesian Network strategy employed here is exploratory and needs further evaluation , parameter tuning and validation .
Cellular transitions from epithelial to mesenchymal phenotypes share common characteristics such as down-regulation of E-cadherin in a variety of tissues and cancer contexts [47] . In this study , we demonstrate that in pan-cancer cell lines , bimodal coupling of transcript , protein , and phosphosite expression reveal epithelial and mesenchymal states . However , many known EMT markers are dispersed across bimodally coupled network modules , suggesting that they are involved in distinct regulatory programs . Different modules likely correspond to intermediate cellular states of an EMT spectrum , whereby transcriptional down-regulation of E-cadherin , and other genes , represents a decisive loss of epithelial traits . In agreement , we find that E-cadherin expression is primarily transcriptionally controlled , possibly with context-dependent control at the translational or post-translational level . By anchoring the investigation around the transcriptional program associated with E-cadherin down-regulation , we identified 239 bimodal EMT markers , many of which have not previously been associated with EMT . The observation that EMT markers are particularly bimodal suggests that cell lines are unequivocally either epithelial or mesenchymal in cell culture . It follows that the EMT decision for cells is determined by the growth medium and the genetics of the cancer cells , rather than by stochastic processes leading to heterogeneous mixtures of cells . Therefore , the identified bimodal switches likely reflect deterministic rather than stochastic architecture . However , in bulk experiments of cell lines , rare but important populations of cells such as mesenchymal stem cells could be neglected . The lack of obvious association between metastasis and E-cadherin expression raises some questions . Possibly , the in vitro conditions , lack of cues from tumor microenvironments , and cell culture passages might mask the original metastatic events from which the cell line is derived . More broadly , cell culture conditions may fail to model crucial aspects of how EMT occurs in complex tissue environments . Yet , the identified deterministic mechanisms may be valid only under the right conditions . We find that cancer cell lines down-regulate epithelial and up-regulate mesenchymal genes when treated with HDAC inhibitors . This observation warrants caution for the use of HDAC inhibitors as cancer and other therapies . If HDAC inhibitors induce EMT in cancer cells , this could explain the disappointing outcomes in clinical trials of HDAC monotherapies for solid tumors [48] . Several protein kinase inhibitors were predicted to revert cancer cell lines to a more epithelial state , but most of these kinase inhibitors are not currently in clinical use . Therefore , these kinase inhibitors may be effective as metastatic repressors and could be under-investigated due to the contemporary focus on targeted drug treatments rather than broad functional effects . Furthermore , the mechanisms of action for the small-molecules may inform us about EMT or MET drivers . For example , the identification that the two SRC inhibitors Dasatinib and Saracatinib are potentially MET inducers , the co-clustering of SRC pY416 with Claudin7 , and its large out-degree in the Bayesian network , all corroborate evidence to the importance of SRC activity for regulating EMT during cancer progression . In addition , the identification of kinase inhibitors rather than other classes of small-molecules suggests that phospho-signaling in general is particularly important for driving MET . Lastly , we show that causal models of protein expression and phosphorylation in cancer cell lines identify known and putative drivers of EMT . Due to the promising preliminary results from the causal models that we constructed , identifying molecular drivers of EMT , despite the lack of statistical power to robustly detect individual causal interactions , it is clear that measuring more cell lines , under more conditions , would substantially increase the sensitivity and in turn quality of such models . Also , if a sufficient number of pan-cancer cell lines could be profiled by mass spectrometry proteomics , the developed bimodal methodology could be reapplied to confirm and discover novel associations between proteins and post-translational modifications that drive oncogenic state transitions .
The RPPA data for 736 cancer cell lines were generated at the MD Anderson Cancer Center . The selection criteria of the 474 measured proteins were based on the aim to cover known cancer-related signaling pathways . We excluded antibodies with missing values across cell lines by requiring that each RPPA measurement is present in at least 40 cell lines . This resulted in a dataset with 450 antibody-based measurements . The CCLE mRNA data and cell line annotations of 1 , 037 cancer cell lines were retrieved from the CCLE portal at: https://portals . broadinstitute . org/ccle . We used the gene-centric RMA-normalized data . For all methods relying on geometric distances , Euclidean distances were computed considering only pairwise complete features . Sparse RPPA measurements were excluded , requiring that each protein is measured in at least 100 cell lines , which resulted in the inclusion of 263 protein measurements . To reduce the dimensionality of the RPPA and mRNA data , we used t-SNE implemented in the R package ‘tsne’ with perplexity value of 30 and at 5 , 000 iterations , and all other arguments at their default values [16] . Only cell lines with available RPPA and mRNA data were included . For the combined RPPA and CCLE mRNA embedding , the distance matrices for each data set were weighted by the sum of all distances . In this way , each data type contributed equally to the combined analysis . To more rigorously assess the number of clusters supported by the RPPA and CCLE data sets , we calculated the Gap statistic [17] from the average linkage hierarchical clustering at tree cuts resulting in clusters of varying cardinalities . PCA was performed using the R package ‘pcaMethods’ from data that were centered and scaled to unit variance , while imputing missing values with the ‘svdImpute’ method . Furthermore , we analyzed patterns in the classification and misclassification of the tissue of origin for the RPPA and mRNA data using 3-nearest-neighbor classification according to a leave-one-out cross-validation scheme . Linear regression was carried out between the RPPA and CCLE mRNA distance matrices with the RPPA distances considered the target variable . The residuals of the regression thus quantify the deviation from expected distance for RPPA data for each pairwise cell line distance . To compare the clustering of breast cancer cell lines , we computed tanglegrams using the dendextend R package . The tanglegram method uses a random search to rotate tree nodes minimizing the overlap of lines drawn between leaves of two trees . We fit univariate two-component Gaussian distributions using the expectation-maximization ( EM ) algorithm implemented in the ‘mixtools’ R package with default parameters . To compare the fitted distribution to unimodal Gaussian distributions , we calculated the difference between the Bayesian Information Criterion ( BIC ) . The data were determined to be bimodal if the BIC difference was larger than 2 . Based on the fitted Gaussian mixture model , we calculated , using Bayes’ theorem , the posterior probabilities of measurements being generated from the high expression component . Note that the probability of belonging to the low component is 1-p . To estimate the tissue diversity of each bimodal fit , we first calculated the frequencies of tissues assigned to the low ( p < 0 . 5 ) and the high ( p > = 0 . 5 ) component . We then calculated the Shannon entropy of the tissue distributions associated with the low- and the high components . The bimodal RPPA measurements were classified into groups of low , medium , and high tissue diversity by the tertiles of the minimum tissue entropy associated with low- and high expression . The bimodal expression was considered common if the fitted mixture coefficients were above 1/4 and rare if below . Based on the posterior probabilities of the bimodal fits associated with high tissue diversity , we calculated a network of bimodal coupling coefficients defined as Spearman’s correlations between the posterior probabilities . To detect robust communities in this network , we set a cutoff of |rb| > 0 . 3 and calculated the leading non-negative eigenvectors using the igraph R package . The network was visualized in Cytoscape with node size proportional to the mixing parameter of the two-component Gaussian fit and with edge coloring based on the coupling coefficients . The coupling coefficients between the E-cadherin RPPA measurements and matched CCLE transcript data were used to define an EMT signature ( rb > 0 . 5 ) . Enrichment analysis was performed with Enrichr [28] . L1000CDS2 was used to query small molecules as potential inducers or reversers of EMT [34] . We summarized the EMT and MET small-molecule predictions by reporting the top-50 small molecules identified using boxplots to aggregate small molecules with multiple experimental conditions such as cell lines , dosage , or timing . The causal models of the RPPA measurements across cancer cell lines were inferred using the Fast Greedy Search algorithm [49] implemented by the BD2K Center for Causal Discovery [50] . We used the rcausal R package version 0 . 99 . 5 to run the Java implementation with penalty discount 4 and depth 3 . To visualize causal neighborhoods , we computed graph cuts and rendered the subnetwork in R using a Sugiyama layout of the directed acyclic graph . The tissue-specific correlations were layered on top of the edges as histograms . To estimate the robustness of the resulting causal network , we ran the algorithm several times in a bootstrap scheme ( M = 200 ) by sampling with replacement . | Profiling molecular and phenotypic characteristics of large collections of cancer cell lines can be used to identify distinct and common oncogenic pathways across cancer types . So far , most large-scale data obtained from cancer cell lines have been at the genomic , transcriptomic , and phenotypic levels . Recently , high-quality data at the level of cell signaling through protein abundances and phosphorylation sites has become available . By integrating this newly generated protein data with prior transcriptomic data , and by visualizing all cancer cell lines using dimensionality reduction techniques , pan-cancer cell lines are strikingly shown to organize into a gradient of epithelial to mesenchymal types . Interestingly , many of the measured proteins and transcripts display bimodality; the expression of genes , proteins , and protein phosphorylations is either high or low , strongly suggesting that they act as molecular switches . Focusing on further characterizing molecular switches of epithelial-mesenchymal transitions , we identify candidate regulators and small molecules that can induce or reverse such transition , as well as potential causal relationships between proteins . Since the mesenchymal state of tumors is known to be associated with metastasis and later-stage cancer development , better understanding the regulatory mechanisms of epithelial-to-mesenchymal transition can lead to improved targeted therapeutics . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biotechnology",
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"small",
"molecules",
"gene",
"regulation",
"messenger",
"rna",
"organic",
"compounds",
"oncology",
"protein",
"expression",
"genome",
"analysis",
"molecular",
"biology",... | 2018 | Integration of pan-cancer transcriptomics with RPPA proteomics reveals mechanisms of epithelial-mesenchymal transition |
Relay cells are prevalent throughout sensory systems and receive two types of inputs: driving and modulating . The driving input contains receptive field properties that must be transmitted while the modulating input alters the specifics of transmission . For example , the visual thalamus contains relay neurons that receive driving inputs from the retina that encode a visual image , and modulating inputs from reticular activating system and layer 6 of visual cortex that control what aspects of the image will be relayed back to visual cortex for perception . What gets relayed depends on several factors such as attentional demands and a subject's goals . In this paper , we analyze a biophysical based model of a relay cell and use systems theoretic tools to construct analytic bounds on how well the cell transmits a driving input as a function of the neuron's electrophysiological properties , the modulating input , and the driving signal parameters . We assume that the modulating input belongs to a class of sinusoidal signals and that the driving input is an irregular train of pulses with inter-pulse intervals obeying an exponential distribution . Our analysis applies to any order model as long as the neuron does not spike without a driving input pulse and exhibits a refractory period . Our bounds on relay reliability contain performance obtained through simulation of a second and third order model , and suggest , for instance , that if the frequency of the modulating input increases or the DC offset decreases , then relay increases . Our analysis also shows , for the first time , how the biophysical properties of the neuron ( e . g . ion channel dynamics ) define the oscillatory patterns needed in the modulating input for appropriately timed relay of sensory information . In our discussion , we describe how our bounds predict experimentally observed neural activity in the basal ganglia in ( i ) health , ( ii ) in Parkinson's disease ( PD ) , and ( iii ) in PD during therapeutic deep brain stimulation . Our bounds also predict different rhythms that emerge in the lateral geniculate nucleus in the thalamus during different attentional states .
Relay neurons are found in various brain nuclei including the thalamus [1]–[3] . Experiments have suggested that the inputs to a thalamic relay neuron can be divided into two categories: driving and modulating . The driving input typically contains sensory information ( e . g visual , motor ) and the modulating input controls relay of this sensory information back to cortex [4] . The driving input is made up of a few synapses on the proximal dendrites whereas the modulating input comprises all other synapses [5] , [6] ( see Figure 1 A ) . For example , the lateral geniculate nucleus ( LGN ) in the thalamus receives the driving input from the retina and projects to the primary visual cortex . The modulating input comprises descending inputs from layer 6 of the visual cortex and ascending inputs from the brain stem . The function of the LGN is to selectively relay sensory information from the retina subject to attentional needs [4] , [7] . It has been observed that during different attentional needs ( which translate into different relay demands ) , local field potentials ( LFPs ) in the LGN have a concentration of power in different frequency bands ( ) [8] , [9] . LFPs may be reflected in the modulating input because they are believed to arise from ensemble synaptic activity [10] . This would then suggest that one mechanism that controls relay in the LGN cell is the frequency of the modulating input . Similarly , relay neurons in the motor thalamus receive driving inputs from sensorimotor cortex , and modulating inputs from the basal ganglia ( BG ) , specifically the Globus Pallidus internal segment ( GPi ) [4] , [11] . The driving input contains information about the actual movement via proprioception , and the modulating input facilitates/impedes relay of this information to motor cortex [12]–[16] . It has been observed that neural activity in the GPi changes its oscillatory patterns when a subject's cognitive state moves from being idle to planning a movement [17] . In particular , GPi activity has prominent beta band oscillations when the subject is idle , which then get suppressed when the subject plans to move . This suppression coincides with an emergence of gamma band oscillations . This would suggest , again , that one mechanism that controls relay in the motor thalamic cell is the frequency of the modulating input . In this study , we set out to quantify when and how these thalamic cells relay driving inputs . Previous attempts to study relay neurons are made in [15] , [16] , [18]–[20] . Specifically , in [18] , [19] in-vitro experiments are used to understand how background synaptic input modulates relay reliability of a thalamic neuron . These studies suggest that the neuron's reliability of relaying an incoming spike is governed by the background synaptic input ( the modulating input ) combined with intrinsic properties of the neuron . In particular [19] , showed that if the variance of the background synaptic input increases , the transmission reliability goes down , and [18] showed that the feedback inhibition from the nucleus reticularis modulates the excitability of the thalamic cell membrane and hence gates transmission of spikes from the retina . An attempt to analytically study relay neurons is made in [15] , where in they studied the effects of BG inhibition on the thalamic relay reliability . They used a order non-bursting model and phase-plane analysis to study relay neuron properties . However , they only considered a constant and a low frequency periodic modulating input . Additionally , only one deterministic periodic waveform was considered for the driving input . A follow up study with a similar objective is presented in [20] , wherein the authors analyzed a relay neuron driven only by a driving input ( no modulating input ) . Using Markov models , they studied how different distributions of driving pulse arrival times affect relay reliability . However , they did not present an explicit expression for the dependence of reliability upon input distributions and relay neuron properties . The work presented here is different from the above computational studies in that we include classes of modulating and driving inputs in our analysis , and we employ systems theoretic tools to obtain explicit analytical bounds on reliability as a function of the neuron's electrophysiological properties ( i . e . , model parameters ) , the modulating input signal , and the driving signal parameters . Our analysis is applicable to any order model as long as the neuron does not spike without a pulse in the driving input and exhibits a refractory period . Consequently , our analysis is relevant for relay cells whose electrophysiological dynamics , including bursting , may be governed by several different ion channels and is more rigorous than previous works . Our lower and upper bounds contained reliability computed through simulation of both a second- and third-order model , and suggest , for example , that if the frequency of the modulating input increases and/or its DC offset decreases , then relay reliability increases . The methods used here are generally applicable to understanding cell behavior under various conditions . In the discussion section , we show how our analysis shed new insights into motor signal processing in health and in Parkinson's disease with and without therapeutic deep brain stimulation . We also discuss how our bounds predict neural activity generated in the LGN during visual tasks with different attentional needs as well as during sleep . In particular , we show how our bounds predict the following observations in the LGN: ( i ) prominent and rhythms ( ) in the LGN LFPs during high attentional tasks [9]; ( ii ) phase locking between rhythm ( ) in LFPs and spiking activity in the LGN in awake behaving cats [21]; ( iii ) rhythms ( ) in drowsy cats; and , ( iv ) even slower rhythms in sleeping cats [8] .
A relay neuron receives two kinds of inputs: a driving input , and a modulating input , and generates one output , , as shown in Figure 1 B . The function of this type of neuron is to generate an output that relays the driving input at appropriate times . The modulating input does as its name implies i . e . it modulates the neuron's ability to relay the driving input [4] . This relay neuron model structure has been widely used to model thalamic relay neurons [15] , [16] , [22]–[26] . We would like to understand exactly how the modulating input affects relay reliability of the neuron . To do so , we use a biophysical-based model to describe the electro-physiological dynamics of the relay neuron . We first begin with a second order model to highlight structure in the model dynamics , and then we present an order generalization . Recall that the output of the cell , , is the membrane voltage of the neuron . Then for time , ( 1a ) ( 1b ) ( 1c ) ( 1d ) ( 1e ) In ( 1 ) , are the membrane capacitance , ionic current , external current and synaptic reversal potential , respectively . is composed of currents , which is a low threshold calcium ion current , and which is the neuron's membrane leakage current . is a constant external current , and , is an internal state of the system representing the probability that a calcium channel inactivation gate is open at a time . are temperature correction factor , maximum calcium current and leakage current conductance , respectively . The details of , and and numerical values used in our simulations are given in Tables 1 and 2 . This is a simplified model of a thalamic neuron that is driven only by calcium ion and leak currents . We begin with this model because it is simple and still contains low threshold calcium currents which are shown to govern input selectivity of relay neurons , in a computational study [23] . This model has also been used to model neurons in the inferior olive for the purpose of studying sub-threshold oscillations [27] . Before we define relay reliability , we first define a relayed pulse . A relayed pulse is a successful response , , that occurs within after a pulse in the driving input , . See Figure S2 ( Supplementary Material ) . Let , ( 11a ) ( 11b ) then the empirical reliability is defined as: ( 12 ) This definition of reliability is similar to the one defined in [15] and is not meaningful if spikes without a pulse in . But since our neuron is a stable neuron , this will never happen . In the limit that we observe the neuron for an infinite amount of time , the empirical reliability converges to ( 13 ) Let us define events ( 14a ) ( 14b ) We then see that ( 15 ) Here we have used the total probability law and the definition of conditional probability [35] to go from ( 13 ) to ( 15 ) . Because we cannot generate a spike in the refractory zone , , we get that ( 16 ) For most neurons , the dynamics of the first component of the state , , are faster than the other states in the region , see Figure 2 C . Therefore , when , it returns to only if it is close to , otherwise it returns to . The return process to is much faster as compared to the return process to , due to slower dynamics arising near . Therefore , when , it spends most of its time close to , and hence we assume that the . Furthermore , since the , this assumption does not affect our results much . We will convince the reader that these assumptions are mild in the results section . Essentially , we will show that our reliability expressions under these assumptions match well to numerically computed curves for different relay neurons . Finally , since and are disjoint sets , we get: ( 17 ) Although not explicitly in ( 17 ) , relay reliability is a function of the driving input parameters , and , the modulating input parameters , and and the neuron's dynamics ( i . e . model parameters ) denoted by . In the next sections , we compute closed-form approximations of lower and upper bounds of reliability as a function of and , by computing and bounds on . To compute we first find a solution for the orbit tube and then find a solution for the response to a driving pulse given the state starts in . This solution shows us when the neuron generates a successful response . We later use this information to compute . In this section , we compute in ( 17 ) to ultimately obtain an expression for . Since a driving pulse that arrives at time can only result in either a successful response or an unsuccessful response , we can equivalently write the definition of as: ( 38a ) ( 38b ) ( 38c ) Here , we have used the law of total probability and the definition of conditional probability [35] to arrive at ( 38c ) . We know that after a successful response at , the system state , only for . Therefore ( 39 ) Similarly , if denotes time spent in refractory zone after unsuccessful response , then we get: ( 40 ) Now by combining ( 13 ) , ( 38c ) , ( 39 ) and ( 40 ) we get: ( 41 ) Since has a complicated dependence on the input and model parameters , it is difficult to calculate . However , it is certain that . This implies that , by properties of cumulative distributive functions [35] . Therefore , we get the following bounds: ( 42 ) Putting ( 41 ) and ( 42 ) together , we get: ( 43a ) ( 43b ) ( 43c ) Now , we calculate . Recall that the inter pulse intervals of , , here is generated from an exponential distribution and is the refractory period . Therefore: ( 44a ) ( 44b ) ( 44c ) It can be easily shown that: ( 45 ) is the average inter pulse interval , . Finally , by combining ( 43c ) and ( 44 ) we get: ( 46a ) Now we compute bounds on relay reliability i . e . Recall that: ( 47a ) ( 47b ) ( 47c ) Similarly , we can write lower bound on reliability as: ( 48 ) Combining ( 47 ) and ( 48 ) we get: ( 49 ) From ( 49 ) and ( 44 ) , one can see that if , which makes . This result is intuitive because if pulses in occur at a slow rate , then the solution of ( 4 ) has enough time to return to the orbit tube after each pulse . Therefore , and . Another interesting case emerges if . In this case and . This case has two interesting extremes: 1 . , making , 2 . , and both and approach . In case 1 , an average a number of pulses occur in the time interval after a successful response . All of these pulses generate unsuccessful responses because the system state is inside during this interval . Therefore , for each successful response , we get unsuccessful responses making . However , in the second case , exactly one pulse occurs during the period after a successful response . Therefore , for every successful response we get at least unsuccessful response . Now , if , we get exactly one unsuccessful response for each successful response making .
In Figure 5 , we plot and vs for given by ( 3 ) with , and superimpose it with a numerically obtained curve through simulation of the original model ( 1 ) . is estimated by doing repeated simulations on ( 4 ) with given by ( 3 ) , and . We see that empirical reliability plus and minus its standard deviation are essentially within bounds and . From Figure 5 B , we see that increases with the frequency of the modulating input , . In Figure 6 A , we plot and vs for , along with empirical reliability computed numerically . We see that reliability decreases as ( i . e . the mean value of modulating input ) increases . In Figure 6 B , we plot vs for , . Reliability again decreases as increases . The dependence of reliability on the cell's input parameters is explicit in our bounds . However , dependence of reliability on the model parameters is captured implicitly by the gain , and . The refractory period , , is well studied in literature and depends on inactivation gate time constants [38] . Therefore , in this section we discuss how the gain and depends on the properties of a relay neuron membrane dynamics . In Figure 7 A , we plot vs conductances and . We see that first decreases with increasing and then increases forming a parabola . Furthermore , with increasing , decreases . In Figure 7 B , we plot the dependence of the gain on and . is essentially a low pass filter whose amplitude decreases as frequency increases . Consequently , reliability increases with frequency ( see ( 49 ) ) . From the Figure , we can see that the gain , , in the high frequency range ( ) increases with and decreases with . For lower frequencies , , has a complex dependence on & . This is an important result as we can increase/decrease reliability of the relay neurons by increasing/decreasing T-type or leak channel conductances which can be further used to treat diseases such as Parkinson's disease ( see discussion ) . In this section , we will apply ( 49 ) to a third order model of a thalamic relay neuron . In this case , the parametrs in the equation are computed from the third order model . We chose the 3rd order thalamic model used in [15] , [16] , [22] , which is a simplification of model used in [39] , [40] . This model exhibits bursting activity in the hyperpolarized state and non bursty firing in the depolarized state . The two responses of the model for an oscillating modulating input and a Poisson driving input ( inter-pulse interval is given by ( 7 ) ) are shown in Figure 8 A and 9 A . The equations and parameters of the model are the same as those used in [15] , [22]: ( 50a ) ( 50b ) ( 50c ) In the ( 50 ) , , , are the leak current , sodium and potassium current , respectively . and are the low threshold potassium current and external current respectively . are the temperature correction factors . All the parameters used are given in Table 3 . A thalamic neuron generates a single spike when depolarized in the relay mode [15] , [41] . However , it generates a burst of spikes when it receives a depolarizing input when it is in a hyperpolarized state [42] . We used , to model the hyperpolarized or bursty state . Whereas , models a single spike state of thalamic neuron . We can rewrite the ( 50 ) in the form of ( 4 ) by defining the state vector with: ( 51 ) In Figure 8 A , we plot the time profile of the voltage for a bursty neuron along with a zoomed in view of the burst in Figure 8 B . Figure 8 C plots our reliability bounds ( 49 ) along with empirical reliability computed numerically through simulation of the 3rd order model . We see that our bounds predict reliability well even for a bursty neuron . Note that we consider a burst response to a pulse as a successful response . In Figure 9 A , we plot the time profile of voltage for a non bursty neuron along with a zoomed in view of a successful spike in Figure 9 B . Figure 9 C plots our reliability bounds ( 49 ) along with empirical reliability computed numerically through simulation of the 3rd order model . Note that here therefore . We see that our bounds predict reliability well in this case also . In general , our analytical bounds are applicable as long as the model 1 . does not generate a spike if there is no pulse in , and 2 . has a threshold behaviour as defined in Materials and Methods section , and 3 . shows a refractory period . The second condition is true for most neurons that satisfy the first condition . Our analysis may also be extended to include neurons that spike without any driving input ( see Discussion ) , but in this manuscript we neglect such dynamics .
Our reliability bounds were calculated assuming that the relay neuron does not fire spontaneously . However , many relay neurons show spontaneous firing in the absence of any input . This spontaneous firing is usually periodic ( period ) because it arises from the emergence of a limit cycle [43] and can be thought of as responses to background noise . Our analysis can therefore be extended to capture this by adding a periodic noise pulse train in the reference input , therefore the new reference input becomes: ( 52 ) Since a successful response to a pulse in is undesirable , we must modify our definition of reliability . To do this , we assume that the arrival of a pulse in cannot coincide with an arrival of a pulse in and thus successful responses to pulses in each signal are disjoint events . This leads us to define reliability as ( 53a ) With this approach , our analysis can be extended to spontaneously firing neurons . We believe that the reliability will approximately be bounded as: ( 54 ) The above expression is reduced to ( 49 ) in the case i . e the noise period is much larger than the period of the driving input . In the case when the reliability becomes negative because noise pulses occur very frequently as compared to desirable driving input pulses . This generates undesirable successful responses making reliability negative . Note that ( 54 ) is only an approximate solution for the reliability of spontaneously firing relay neurons and we leave the exact solution to this problem for the future work . In the motor circuit , thalamocortical neurons receive a driving input from the motor cortex and a modulating input from the GPi segment in the basal ganglia ( BG ) . See Figure 10 A . The function of the GPi input is hypothesized to enable/disable thalamic cells to relay cortical stimuli related to movement when movement is intended/not intended [14] . This is consistent with evidence that the BG both inhibits unwanted movements and enables intended movements in a timely manner [12] , [13] . This GPi modulated thalamic relay ultimately enables reliable transfer of information from higher cortical layers to lower layers which then command the musculoskeletal system to generate planned movements [44] . The thalamic relay hypothesis is supported by previous studies [4] , [16] , [22] . In [16] , [22] , it is shown that relay reliability computed from a data-driven computational model of a thalamic neuron is low in Parkinson's disease ( PD ) , and high in both healthy and when therapeutic DBS is applied to the BG in PD . Previous works emphasize the inhibitory projections from GPi to motor thalamus [45]–[48] . They argue that when movements are intended/not intended , appropriate task-related GPi neurons decrease/increase their firing rates . This in turn disinhibits/inhibits thalamus and consequently enables/disables thalamic relay , respectively . Our analysis as well as recent experimental observations show that the story is a bit more complicated . GPi firing rates alone may not be the mechanism for thalamic relay , rather , the dynamics of the GPi activity control thalamic relay . In particular , it appears that the oscillatory dynamics of GPi activity control relay . Our relay bounds predict that if one intends to move , then the GPi neurons that project to motor thalamus should initially generate LFP activity that has prominent low frequency oscillations which allows the subject to remain idle , and then generate activity that has prominent high frequency oscillations which allows the subject to plan an intended movement and then move . We first discuss how our analysis concurs with observations obtained from a computational model of the motor circuit that characterizes neural activity dynamics in the BG and motor thalamus in health and in PD with and without therapeutic DBS . The computational model simulates neural activity when movements are planned and hence when motor thalamus should relay information from the cortex at all simulated times . We then discuss how our relay bounds accurately predict how GPi activity recorded from two healthy primates modulates during a structured behavioral task that forces an idle phase , and a planning phase during each task trial . As mentioned in the Introduction , neurons in the LGN receive driving input synapses from the retina and modulating input synapses from layer 6 of the visual cortex and the brain stem . The LGN then relays the driving input to visual cortex for perception . The LGN functions as a “gatekeeper” and allows only the relevant information to go through depending on attentional demands [7] , [64] . In the LGN , the spatial map of the visual field is conserved [64] , [65] . Here , we hypothesize that the LGN functions as a filter of the spatial map which shows a high relay reliability in spatial areas requiring high attention and lower reliability otherwise . Our analysis suggests then that LGN neurons relaying attended areas of the visual field receive higher frequency modulating inputs as compared to LGN neurons relaying areas which are ignored . Note that the modulating input represents the synaptic background activity , which is a major contributor to LFPs and EEG recordings [10] . Therefore , the frequency content of LFPs and EEG reflect the frequency of the modulating input . This hypothesis is supported by [8] , where it was shown that the frequency of the LFPs in LGN depends on the arousal state of the cat . Specifically , they showed a prominent rhythm ( ) in awake and naturally behaving cats , a rhythm ( ) in drowsy cats and a slow rhythm ( ) during sleep . Additionally [21] , showed that , in wakeful naturally behaving cats , the spiking activity of relay-mode ( non-bursty ) neurons in the LGN is correlated with the phase of the alpha rhythm of the LFPs . Specifically , some neurons spike more at the peaks of the alpha wave while other neurons spike more at the valleys of the alpha rhythm . Using ( 36 ) , we may be able explain why such phase locking occurs . In words , this equation says that relay neuron reliably relays the driving input only during a fixed phase interval of modulating input , and this phase interval depends on neuron membrane properties [21] . Finally , during deep sleep slow delta rhythms are observed in the EEG which are believed to be of thalamic origin [66] . This may cause even lower reliability in LGN and filter out all the visual information , resulting in deep sleep . On the other hand , high frequency & rhythms are observed during visual attentional tasks in the LFPs of cat LGN [9] . Our analysis shows that reliability increases with modulating input frequency , therefore we propose that the reliability during these tasks is greater than during natural wakeful behaviour for most LGN neurons . This results in larger relay of information which increases general productivity . In addition to the observed relationship between the LGN LFP oscillations and attention , it has been observed that during sleep , LGN neurons become hyperpolarized [42] , [67] . In our model , this means that the DC offset of the modulating input , , is large which decreases reliability according to our analysis . The LGN neurons relay poorly and also exhibit a bursty behaviour ( see Figure 6 A and 8 ) . The lower reliability may result in less information relay from the LGN to the visual cortex , inducing sleep whereas the bursty behaviour may only be a by-product of hyperpolarization and may have nothing to do with information suppression . This agrees with [68] where in it is shown experimentally that although all bursts combine carry lesser information than all single spikes , individual burst is more informative than a single spike in the LGN output . The information carried in the bursty mode may be critical for waking up [42] . | In cellular biology , it is important to characterize the electrophysiological dynamics of a cell as a function of the cell type and its inputs . Typically , these dynamics are modeled as a set of parametric nonlinear ordinary differential equations which are not always easy to analyze . Previous studies performed phase-plane analysis and/or simulations to understand how constant inputs impact a cell's output for a given cell type . In this paper , we use systems theoretic tools to compute analytic bounds of how well a single neuron's output relays a driving input signal as a function of the neuron type , modulating input signal , and driving signal parameters . The methods used here are generally applicable to understanding cell behavior under various conditions and enables rigorous analysis of electrophysiological changes that occur in health and in disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"bioengineering",
"biomedical",
"engineering",
"biotechnology",
"biological",
"systems",
"engineering",
"control",
"engineering",
"control",
"systems",
"computational",
"neuroscience",
"single",
"neuron",
"function",
"biology",
"computational",
"biology",
"engineering"
] | 2012 | Performance Limitations of Relay Neurons |
Dengue is a potentially fatal acute febrile illness ( AFI ) caused by four mosquito-transmitted dengue viruses ( DENV-1–4 ) that are endemic in Puerto Rico . In January 2010 , the number of suspected dengue cases reported to the passive dengue surveillance system exceeded the epidemic threshold and an epidemic was declared soon after . To characterize the epidemic , surveillance and laboratory diagnostic data were compiled . A suspected case was a dengue-like AFI in a person reported by a health care provider with or without a specimen submitted for diagnostic testing . Laboratory-positive cases had: ( i ) DENV nucleic acid detected by reverse transcriptase-polymerase chain reaction ( RT-PCR ) in an acute serum specimen; ( ii ) anti-DENV IgM antibody detected by ELISA in any serum specimen; or ( iii ) DENV antigen or nucleic acid detected in an autopsy-tissue specimen . In 2010 , a total of 26 , 766 suspected dengue cases ( 7 . 2 per 1 , 000 residents ) were identified , of which 46 . 6% were laboratory-positive . Of 7 , 426 RT-PCR-positive specimens , DENV-1 ( 69 . 0% ) and DENV-4 ( 23 . 6% ) were detected more frequently than DENV-2 ( 7 . 3% ) and DENV-3 ( <0 . 1% ) . Nearly half ( 47 . 1% ) of all laboratory-positive cases were adults , 49 . 7% had dengue with warning signs , 11 . 1% had severe dengue , and 40 died . Approximately 21% of cases were primary DENV infections , and 1–4 year olds were the only age group for which primary infection was more common than secondary . Individuals infected with DENV-1 were 4 . 2 ( 95% confidence interval [CI]: 1 . 7–9 . 8 ) and 4 . 0 ( 95% CI: 2 . 4–6 . 5 ) times more likely to have primary infection than those infected with DENV-2 or -4 , respectively . This epidemic was long in duration and yielded the highest incidence of reported dengue cases and deaths since surveillance began in Puerto Rico in the late 1960's . This epidemic re-emphasizes the need for more effective primary prevention interventions to reduce the morbidity and mortality of dengue .
Dengue virus ( DENV ) transmission is endemic throughout most of the tropics and sub-tropics and is estimated to result in ∼50 million symptomatic infections and ∼20 , 000 deaths each year [1] , [2] . Infection with any DENV can result in dengue , an illness characterized by fever , headache , retro-orbital eye pain , myalgia and rash [2] . In some cases , dengue can progress to severe dengue [2] , which includes dengue hemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) [3] and is characterized by thrombocytopenia , increased vascular permeability with plasma leakage , severe organ involvement , and/or clinically significant bleeding [2] . Supportive care with appropriate intravascular volume repletion has been shown to lower mortality associated with severe dengue [2] . The four related but serotypically distinct DENV-types , DENV-1 , -2 , -3 and -4 , are transmitted by Aedes aegypti or Ae . albopictus mosquitoes [4] , [5] . Following infection , individuals develop short-lived , heterotypic immunity and long-lived , type-specific immunity [6] . Primary infection is an individual's first DENV infection , and secondary infection is any subsequent infection with a DENV-type different from the first . Severe dengue is more common upon secondary infection [2] , [7] and may be affected by the order in which an individual is infected with the respective DENV-types [2] , [8] . Thus , increases in the force of DENV infection can result in a decrease in the age of primary and secondary infection [2] . Both local patterns of circulation of the four DENV-types and force of infection can influence the age groups most affected by dengue and severe dengue . The unincorporated United States territory of Puerto Rico is composed of 78 municipalities , an area of 3 , 515 square miles , and a population of 3 , 725 , 789 [9] . The demographics of Puerto Rico are similar to the United States as median age is 36 years and 78 . 6% are white , although 99% are self-described Hispanic [9] . Since the mid-1990's the health care system in Puerto Rico has included both public and private health care services , and dengue has been a reportable condition for several decades . Ae . aegypti is the predominant DENV vector on the island . Dengue was first described in Puerto Rico in 1915 [10] and outbreaks have been recognized since 1963 [11] , [12] . DHF was first reported in 1975 [13] , [14] , all four DENV-types have been identified on the island since 1982 [15] , [16] , and the first confirmed dengue-related death was reported in 1986 [17] . Recent epidemics were detected in 1994–1995 , 1998 and 2007 , with 24 , 700 [18] , 17 , 000 [19] and 10 , 508 [20] reported suspect cases , respectively ( Table S1 ) . During both epidemic and non-epidemic periods , 10–19 year olds have been the most affected age group for several decades . In the present investigation , we describe a dengue epidemic that occurred in 2010 , including differences in the epidemiology of cases infected with different DENV-types with respect to primary versus secondary infection .
A retrospective analysis of suspected dengue cases reported to surveillance systems was performed to: 1 ) describe the epidemiology of the 2010 dengue epidemic , including disease severity; 2 ) determine the proportion of primary and secondary DENV infections , and the molecular epidemiology of the DENVs responsible for the epidemic; and 3 ) describe relationships between demographic variables ( e . g . age , sex , municipality of residence ) and characteristics of illness ( e . g . infecting DENV-type , severity of illness ) . This investigation underwent institutional review at CDC and was determined to be public health practice and not research , including the post-hoc determinations of DENV molecular epidemiology and primary/secondary infection rates in reported cases; as such , Institutional Review Board approval was not required . Surveillance data from five sources were used to identify cases . First , Centers for Disease Control and Prevention Dengue Branch ( CDC-DB ) and Puerto Rico Department of Health ( PRDH ) jointly operate the island-wide Passive Dengue Surveillance System ( PDSS ) that requires an acute serum specimen and completion of a Dengue Case Investigation Form ( DCIF ) ( cdc . gov/dengue/resources/dengueCaseReports/DCIF_English . pdf ) for case reporting and diagnostic testing . Second , the Enhanced Dengue Surveillance System ( EDSS ) operates solely in the municipalities of Patillas and Guayama and utilizes an on-site nurse epidemiologist to encourage case reporting and patient follow-up to obtain a convalescent serum specimen [21] . Third , identification of fatal dengue cases is conducted via PDSS and EDSS [22] , and enhanced fatal case surveillance was initiated in January 2010 in collaboration with the Instituto de Ciencias Forenses de Puerto Rico , which obtains blood and tissue specimens at autopsy from suspected dengue-related deaths . Fourth , PRDH operates the Notifiable Diseases Surveillance System ( NDSS ) wherein suspected dengue cases are reported without diagnostic testing at CDC-DB . Last , in addition to dengue diagnostic testing performed at CDC-DB for PDSS and EDSS , testing is performed by two private diagnostic laboratories outside of Puerto Rico according to their internal protocols [23] . Diagnostic test results from these laboratories and patient data , including sex , age , and date of illness onset ( if unavailable , specimen collection date was used instead ) , were entered into an independent database . Deduplication of individuals reported to more than one data source was achieved by matching records on name and date of birth and consolidation into a single case if two or more reports from any data source had symptom onset dates within 14 days of each other . As case-patients' travel history is not well captured via the surveillance systems used in this investigation , reported cases may represent both locally-acquired as well as travel-associated cases . All diagnostic testing was performed at CDC-DB for serum specimens received through PDSS or EDSS using the following algorithm: acute specimens ( collected ≤5 days after symptom onset ) were tested by DENV-type-specific real-time reverse-transcriptase-polymerase chain reaction ( RT-PCR ) [24] adapted for high throughput using MDX-10 Universal and M48 systems ( Qiagen , Valencia , CA ) ; acute specimens collected 5 days after symptom onset and all convalescent specimens ( collected ≥6 days after symptom onset ) were tested for the presence of anti-DENV immunoglobulin M ( IgM ) antibody with an antibody-capture enzyme-linked immunosorbent assay ( MAC ELISA ) and a cut-off value of the OD450 of the specimen versus that of the negative control ( ie . P/N ratio ) ≥2 . 0 [25] , [26] . All serum specimens from fatal cases were tested by both RT-PCR and MAC ELISA . Tissue specimens were tested at CDC Infectious Diseases Pathology Branch in Atlanta , GA by immunohistochemistry ( IHC ) [27] and flavivirus-specific RT-PCR [28] followed by sequencing . A suspected dengue case was a dengue-like illness in a person in Puerto Rico whose health care provider: 1 ) submitted a DCIF and serum or tissue specimen to CDC-DB for dengue diagnostic testing; 2 ) submitted a serum specimen to a private laboratory for dengue diagnostic testing; or 3 ) reported the case via NDSS . A laboratory-positive case was a suspected dengue case that met the following criteria for current ( criteria 1 and 2 ) or recent ( criterion 3 ) DENV infection: 1 ) detection of DENV nucleic acid in a serum or tissue specimen; 2 ) detection of DENV antigen in a tissue specimen; or 3 ) detection of anti-DENV IgM antibody in a serum specimen . A laboratory-negative case was a suspected dengue case with: 1 ) no anti-DENV IgM antibody detected in a convalescent specimen; or 2 ) no DENV nucleic acid or antigen detected in a fatal case with only a tissue specimen submitted . A laboratory-indeterminate case was a suspected dengue case with no DENV nucleic acid or anti-DENV IgM antibody detected in an acute specimen with no convalescent specimen available for testing . Dengue with warning signs and severe dengue were defined according to 2009 WHO clinical guidelines [2]; dengue , DHF and DSS were defined according to 1997 WHO clinical guidelines [3] . A representative sample of all RT-PCR-positive cases reported to PDSS or EDSS with illness onset between January 1 and December 31 , 2010 was selected to determine the rates of primary and secondary DENV infection . Cases were stratified by age group with optimal allocation to allow for comparison between age groups , and further allocated to reflect the proportion of DENV-types that occurred during 2010 to allow for comparison between DENV-types and age groups . Sample size was calculated using an estimate of the proportion of secondary infections by age group based on data from the 2007 dengue epidemic [20] , an error of 20% , 95% significance , and an expected 20% of specimens having insufficient specimen volume remaining for testing to be completed . Of the 1 , 000 selected cases , 818 had sufficient specimen volume and were tested at a dilution of 1∶100 for the presence of anti-DENV IgG antibody by ELISA using DENV-1–4 antigen and a cut-off value of OD450≥0 . 15 [29] , [30] . A secondary DENV infection was defined by detection of anti-DENV IgG antibody in an acute specimen , and a primary DENV infection by lack of anti-DENV IgG antibody detection in an acute specimen . Serum specimens with DENV-1 ( n = 7 ) , DENV-2 ( n = 2 ) or DENV-4 ( n = 4 ) detected by RT-PCR were randomly selected from municipalities with the highest incidence of the respective DENV-type and inoculated into cultured C6/36 cells; the presence of virus was confirmed by RT-PCR and indirect immunofluorescence [31] . Isolates were further propagated and viral RNA was extracted from culture supernatants using the M48 BioRobot System ( Qiagen; Valencia , CA ) . The envelope glycoprotein ( E ) gene was amplified and sequenced; sequence data were restricted to the E gene open reading frame ( 1 , 485 basepairs ) . Multiple sequence alignment was performed using MUSCLE available in MEGA 5 ( megasoftware . net ) and GTR+Γ+I4 was selected as the best nucleotide substitution model as determined by MODELTEST v3 . 7 . Genetic relatedness was inferred and represented with phylogenetic trees using the maximum likelihood method in MEGA 5 . MCMC was run in BEAST v1 . 6 . 1 ( beast . bio . ed . ac . uk ) under Bayesian skyline prior , constructed in TreeAnnotator found in the same BEAST package , and visualized in FigTree v1 . 3 . Both trees rendered almost identical tree topologies , therefore confirming genetic relatedness . Evolutionary distances were corroborated by pairwise alignment in BioEdit v7 . 1 . 3 and E gene sequences from GenBank were included in the phylogenetic tree to support tree topology by currently circulating genotype . Tree topology was supported by bootstrapping with 1 , 000 replicates . Genotypes were referred to by previously described nomenclature [32] , [33] . The frequencies of clinical , demographic and laboratory data were calculated by performing descriptive analyses of all suspected dengue cases identified in 2010 . Rates of suspected dengue and laboratory-positive cases were calculated using population denominators obtained from the 2010 United States Census [9] . Statistical differences in proportions were tested by applying the Chi-squared test and Fisher's exact test when applicable . Unless otherwise noted , relative risk ratios were used to calculate all differences between effect sizes . All data analyses were conducted using SAS 9 . 2 ( SAS Institute Inc . , Cary , NC ) , graphs were produced in Microsoft Excel ( Microsoft Corp . , Redmond , WA ) , and maps were created using ArcView ( ESRI , Redlands , CA ) .
We identified 26 , 766 suspected dengue cases with illness onset between January 1 and December 31 , 2010 ( 7 . 2 suspected dengue cases per 1 , 000 residents ) . Of these , 22 , 496 ( 84 . 0% ) were reported to PDSS , 1 , 846 ( 6 . 9% ) were identified though diagnostic testing at a private laboratory , 1 , 304 ( 4 . 9% ) were reported to NDSS , and 1 , 120 ( 4 . 2% ) were reported to EDSS ( Fig . S1 ) . Suspected dengue cases exceeded the PDSS epidemic threshold in the first week of 2010 , increased steeply in week 20 ( May 14–20 ) , and peaked at 1 , 157 in week 32 ( August 6–12 ) ( Fig . 1 ) . Suspected dengue cases slowly declined thereafter and returned to below the historic average in mid-December . Of all suspected dengue cases , 25 , 852 ( 96 . 6% ) had a specimen tested for evidence of DENV infection , of which 25 , 246 ( 97 . 7% ) were tested by CDC-DB and the remainder by a private laboratory; paired specimens were available for 1 , 996 ( 7 . 5% ) cases . Of all cases with a specimen tested , 3 , 664 ( 14 . 2% ) were laboratory-negative , 10 , 140 ( 39 . 2% ) were laboratory-indeterminate , and 12 , 048 ( 46 . 6% ) were laboratory-positive ( 3 . 2 laboratory-positive cases per 1 , 000 residents ) . The median weekly proportion of cases that tested laboratory-positive was 48 . 3% , and was highest ( 64 . 5% ) in week 24 ( June 11–17 ) and lowest ( 11 . 1% ) in week 53 ( December 31 ) . Laboratory-positive case-patients resided in all 78 municipalities of Puerto Rico ( Fig . 2A ) , and the median rate of laboratory-positive cases by municipality was 2 . 68 per 1 , 000 residents . Rates were the highest in the municipality of Patillas ( 16 . 34 cases per 1 , 000 residents ) , the southeastern municipality where the EDSS site is located [21] , and lowest in Aibonito ( 0 . 12 cases per 1 , 000 residents ) in the mountainous center of Puerto Rico . Of 7 , 426 RT-PCR-positive cases , DENV-1 was detected in 5 , 126 ( 69 . 0% ) and incidence was highest in the southeast ( Fig . 2B ) . DENV-2 was detected in 545 ( 7 . 3% ) cases primarily in the west ( Fig . 2C ) , whereas DENV-4 was detected in 1 , 757 ( 23 . 7% ) cases and incidence was highest in south-central and northwestern Puerto Rico ( Fig . 2D ) . DENV-3 was detected in just two ( <0 . 1% ) cases in early 2010 . The age distribution of laboratory-positive cases was significantly different from suspected dengue cases only for case-patients between 30 and 69 years of age ( Fisher's exact , p≤0 . 04 ) . The median age of laboratory-positive case-patients was 18 years ( Table 1 ) . The most affected age group was 10–14 year olds ( 7 . 8 cases per 1 , 000 individuals ) , followed by 15–19 year olds ( 7 . 4 cases per 1 , 000 individuals ) ( Fig . 3A ) . Five-to-nine year olds were the next most affected age group followed by individuals <1 year of age ( 4 . 6 and 4 . 1 cases per 1 , 000 individuals , respectively ) . Individuals 50–59 years of age were the least affected age group ( 1 . 7 cases per 1 , 000 individuals ) . The distribution of RT-PCR-positives cases among age groups was not significantly different from that of laboratory-positive cases ( Fisher's exact , p>0 . 05 ) except for the 50–59 year-old age group , for which serum specimens were collected later ( median: 6 days post-illness onset [DPO] ) than all other age groups ( median: 4 DPO ) ( Fisher's exact , p = 0 . 04 ) and thus tested less frequently by RT-PCR . Despite this , the distribution of DENV-types was not consistent among age groups ( Fig . 3B ) . The strong majority ( 89 . 3% ) of RT-PCR-positive cases in individuals 1–4 years of age were due to infection with DENV-1 , whereas 8 . 1% and 2 . 6% were due to infection with DENV-4 and -2 , respectively . The percent of infections due to DENV-1 decreased and those due to DENV-4 increased with age until a plateau of approximately 60% DENV-1 , 30% DENV-4 and 10% DENV-2 was reached in the 20–29 year old age group . From the sample of 818 RT-PCR-positive specimens tested for primary versus secondary DENV infection , 169 ( 20 . 7% ) were primary and 649 ( 79 . 3% ) were secondary . The median age of individuals experiencing primary infection was 14 years , compared to 23 years for individuals experiencing secondary infection . Eighty-one percent of individuals 1–4 years of age had primary infection and were the only age group for which primary infection was significantly more common than secondary ( p = 0 . 003 ) ( Figure 3C ) . More than 89% of infections in all adult age groups ( i . e . age ≥20 years ) were secondary . The frequency with which anti-DENV IgG antibody was detected in specimens taken from infants was likely due to the presence of maternal antibody [2] . Whereas 28 . 5% of all DENV-1 infections were primary , significantly fewer DENV-2 ( 6 . 8% ) and DENV-4 ( 7 . 1% ) cases were primary infections ( p<0 . 0001 ) ( Table 2 ) . Calculation of relative risk ratios ( RR ) indicated that individuals infected with DENV-1 were 4 . 2 and 4 . 0 times more likely to be experiencing primary infection than were individuals infected with DENV-2 or -4 , respectively ( Table 2 ) . Sequencing and phylogenetic analyses of randomly selected DENV isolates showed that DENV-1 belonged to the American-African genotype ( genotype V [34] ) , but to a clade distinct from virus isolated during the 1998 Puerto Rico epidemic ( Fig . 4A ) . Available sequence data suggest that close ascendants of the 2010 DENV-1 clade had been circulating in Puerto Rico and the Caribbean since at least 2006 ( Fig . 4A ) . DENV-2 sequencing indicated that the virus belongs to clade 1B of the American-Asian genotype ( genotype IIIb [35] ) ( Fig . 4B ) , which is composed of DENV strains endemic to Puerto Rico [36] . DENV-4 belonged to the Indonesian genotype ( genotype II [37] ) , but was distinct from virus isolated in 1998 ( Fig . 4C ) . Viruses closely-related to the DENV-4 isolated in 2010 were first detected in Puerto Rico in 2004 ( Fig . 4C ) . Of 12 , 048 laboratory-positive cases , 31 . 5% had at least one hemorrhagic manifestation and sufficient clinical data was provided to classify 74 . 0% as dengue and 2 . 4% as DHF ( Table 1 ) . Nearly half ( 49 . 7% ) of all laboratory-positive cases had dengue with at least one warning sign , and 11 . 1% had severe dengue . Of 128 suspected dengue deaths , 40 ( 31 . 3% ) were laboratory-positive cases . While adults represented nearly half of laboratory-positive cases with dengue ( 47 . 1% ) , dengue with warning signs ( 44 . 6% ) , and severe dengue ( 49 . 7% ) , they accounted for nearly all ( 92 . 5% ) fatal dengue cases . Laboratory-positive severe and fatal dengue occurred at a rate of 0 . 36 and 0 . 01 cases per 1 , 000 residents , respectively; laboratory-positive fatal dengue cases occurred at a rate of 30 . 0 per 1 , 000 severe dengue cases . From the sample of cases for which primary and secondary DENV infection status was determined , secondary infection was identified in 102 ( 87 . 9% ) case-patients with severe dengue and 547 ( 77 . 9% ) case-patients without severe dengue ( RR = 1 . 2; 95% CI = 1 . 1–1 . 2 ) . Case-patients with DHF were more likely to have been infected with DENV-4 than DENV-1 , and those with severe dengue were more likely to have been infected with DENV-4 than DENV-1 or -2 ( Table 2 ) . There was no significant difference between infection with DENV-1 , -2 or -4 and likelihood of being a fatal case .
In 2010 , Puerto Rico experienced the largest and longest dengue epidemic ever documented on the island . In total , more than 12 , 000 individuals had laboratory-confirmed dengue , of which more than 1 , 300 experienced severe dengue and 40 died . The most common DENV identified was DENV-1 , and 1–4 years old were the only age group more frequently experiencing a primary versus secondary DENV infection . Individuals infected with DENV-1 were four times more likely to have a primary infection than were those infected with DENV-2 or -4 . A strength of this investigation was utilization of multiple surveillance systems to identify all reported suspect dengue cases . However , a minor weakness was that data obtained from each system may not be directly comparable due to different diagnostic algorithms used by CDC-DB and private laboratories , and we were not able to determine status of primary versus secondary infection or perform sequencing on specimens from private laboratories . Because private laboratories contributed <5% of all laboratory-positive dengue cases , this likely did not affect the conclusions of this investigation . The 2010 dengue epidemic was similar in several respects to the 1998 epidemic: both began in January during El Niño events accompanied by above average temperatures , which while not a determinant of epidemics in Puerto Rico [38] may contribute to increased DENV transmission [39]; and both epidemics peaked in week 32 of the calendar year and were predominated by transmission of DENV-1 and -4 [19] . A notable difference was that DENV-3 was essentially absent in 2010 , whereas it accounted for ∼6% of cases during the 1998 epidemic [19] . DENV-3 was re-introduced into Puerto Rico in 1998 following a 20-year absence and was the predominant virus-type in the 2007 dengue epidemic [20] . Thus , susceptibility to DENV-3 infection was likely high in 1998 and low in 2010 , which likely explains these observations . The American-African and Indonesian genotypes of DENV-1 and -4 have been circulating in Puerto Rico since introduced in 1978 and 1981 , respectively [16] , [40] . However , the DENV-1 isolated in 2010 was distinct from the DENV-1 isolated during the 1998 epidemic ( Fig . 4A and [41] ) and was more closely related to the DENV-1 isolated during the 2007 epidemic ( Fig . 4A ) . Similarly , the DENV-4 isolated during the 2010 epidemic was distinct from the DENV-4 isolated in 1998 and was more closely related to viruses circulating since 2004 ( Fig . 4B ) . These findings suggest that DENV-1 and -4 may have both experienced clade replacements at some point after 1998 but prior to 2007 . After the re-introduction of DENV-3 into Puerto Rico in 1998 , DENV-1 was not detected between 2001 and 2006 and DENV-4 was not detected between 2000 and 2005 [42] . Nonetheless , apparent re-introductions of DENV-1 in 2007 and DENV-4 in 2006 were soon followed by the disappearance of DENV-3 in 2010 ( this paper and [42] ) . In place of the convenience sample used in this investigation to describe the DENVs responsible for the epidemic , sequencing of a representative sample of specimens and longitudinal sequence analysis will be necessary to both confirm apparent clade replacements and determine if other DENV clades contributed to the 2010 epidemic . Similar to previous epidemics in Puerto Rico ( Table S1 ) , 10–19 year olds were most affected during the 2010 epidemic; however , unlike previous epidemics , 5–9 year olds were the next most affected age group . The median age of individuals experiencing secondary DENV infection declined from 27 years in 2007 [20] to 23 years in 2010 , likely due to the relative proximity of the periods of high infection pressure . Taken together , these observations indicate an increase in incidence of dengue and a decrease in the age of secondary infection , suggesting that the overall force of DENV transmission may have been higher in 2010 than in previous epidemic years . The observation that DENV-2 and -4 cause relatively infrequent clinical apparent illness upon primary DENV infection is consistent with previous studies [43]–[48] . Similarly , our finding that DENV-1 was a more frequent cause of clinically apparent illness upon primary infection has also been previously reported [43] , [49] , including the observation of increased disease severity during primary infection with DENV-1 compared to other DENV-types [44] , [50] , [51] . Nonetheless , of 545 DENV-2 and 1 , 755 DENV-4 infections , roughly 7% were primary , indicating that primary infection with these DENVs can cause clinically apparent illness , contrary to previous assertions [46] , [47] . The relative abundance of DENV-1 compared to DENV-2 and -4 is unlikely to be responsible for the observed differences in likelihood of causing clinically apparent illness upon primary infection , as relative risk ratios compare the proportion of exposed individuals experiencing the outcome of interest . This is supported by the findings in the 1–4 year-old age group , of which ∼80% experienced a primary infection with DENV-1 . Alternative explanations for these observations include potential variations in the sensitivity of detection of DENV-type-specific anti-DENV IgG antibody and differences in force of infection between the DENV-types circulating in 2010 . We also saw that DENV-1 and -2 were less frequently a cause of severe dengue than DENV-4 . This is in contrast to previous studies where DENV-1 was a more frequent cause of DHF than DENV-4 [52] , and a study where DENV-2 was twice as likely to result in DHF as DENV-4 [43] . Possible explanations for these differences include: the comparatively small number of DENV-4 infections observed in previous studies; differences in clade and/or viral fitness leading to differential pathogenicity [33] , [53] , [54]; and/or the DENV-type ( s ) and sequence to which individuals were previously exposed , which may affect the likelihood of developing severe dengue [44] , [55] , [56] . This investigation had several limitations . First , because individuals experiencing secondary infection may have a diminished anti-DENV IgM antibody response [57] , suspected dengue cases tested solely for anti-DENV IgM antibody may have been misclassified . Second , although DENV is the sole flavivirus known to cause clinically apparent illness in humans in Puerto Rico ( CDC , unpublished data ) , some proportion of anti-DENV IgM or IgG positive results could have been due to infection with or vaccination against another flavivirus [58] , resulting in misclassification . Third , because clinical data was provided for >90% of case-patients on only one occasion and some data variables were incompletely reported ( e . g . only 56% of suspected cases had a reported status of hospitalization ) , severity of disease and the rates of dengue with warning signs and severe dengue reported here were likely underestimated . Finally , the description of the epidemiology and molecular characteristics of dengue reported here is only representative of reported , clinically apparent DENV infections and may not be reflective of asymptomatic and sub-clinical DENV infections . The 2010 dengue epidemic in Puerto Rico demonstrated that dengue continues to be a public health concern for Puerto Rico residents and visitors , and surveillance systems and control initiatives should continue to be supported and strengthened . This epidemic also highlights the need for effective primary prevention tools such as a dengue vaccine to reduce disease morbidity and mortality . | Dengue is a potentially fatal acute febrile illness that is endemic throughout the tropics and sub-tropics . Dengue has been endemic in Puerto Rico for several decades and recent epidemics occurred in 1994–5 , 1998 and 2007 . In January 2010 , dengue surveillance indicated that an epidemic had begun . The epidemic peaked in early August and ended in December with a total of 26 , 766 suspected dengue cases identified , of which 128 were fatal . The 2010 epidemic was one of the longest in Puerto Rico history and resulted in the greatest number of cases and deaths ever detected . We analyzed the epidemiologic and immunologic characteristics of laboratory-confirmed dengue cases and age group-specific attack rates , and determined the frequency of first DENV infection and DENV-types among persons experiencing their first infection . This analysis indicated that 10–19 year-olds were most affected during the epidemic , and that DENV-1 was roughly four times more likely to be associated with clinically apparent illness upon first DENV infection than were DENV-2 or -4 . The 2010 dengue epidemic demonstrated the heavy burden of illness due to dengue in Puerto Rico , re-emphasizing the critical need for effective primary prevention tools to reduce the morbidity and mortality due to dengue worldwide . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"public",
"health",
"and",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"emerging",
"viral",
"diseases",
"dengue",
"microbiology",
"neglected",
"tropical",
"diseases",
"infectious",
"diseases",
"molecular",
"epidemiology",
"dengue",
"fever",
"epide... | 2013 | Virus-Specific Differences in Rates of Disease during the 2010 Dengue Epidemic in Puerto Rico |
The response of a neuron to repeated somatic fluctuating current injections in vitro can elicit a reliable and precisely timed sequence of action potentials . The set of responses obtained across trials can also be interpreted as the response of an ensemble of similar neurons receiving the same input , with the precise spike times representing synchronous volleys that would be effective in driving postsynaptic neurons . To study the reproducibility of the output spike times for different conditions that might occur in vivo , we somatically injected aperiodic current waveforms into cortical neurons in vitro and systematically varied the amplitude and DC offset of the fluctuations . As the amplitude of the fluctuations was increased , reliability increased and the spike times remained stable over a wide range of values . However , at specific values called bifurcation points , large shifts in the spike times were obtained in response to small changes in the stimulus , resulting in multiple spike patterns that were revealed using an unsupervised classification method . Increasing the DC offset , which mimicked an overall increase in network background activity , also revealed bifurcation points and increased the reliability . Furthermore , the spike times shifted earlier with increasing offset . Although the reliability was reduced at bifurcation points , a theoretical analysis showed that the information about the stimulus time course was increased because each of the spike time patterns contained different information about the input .
Neural recordings in vivo are often analyzed with the peristimulus time histogram , which measures increases or decreases in firing rate in response to stimulus onset [1] . Recordings at the sensory periphery , such as the retina and lateral geniculate nucleus ( LGN ) , indicate that spiking responses can be tightly locked to stimulus features with a temporal resolution as high as 1 ms [2] , [3] , [4] , [5] . Ensemble recordings in cortex and hippocampus have shown that populations of cells could dynamically reactivate during sleep and quiet awake periods with high precision [6] , [7] . These precisely timed spikes drive target neurons [8] , [9] , [10] , but only a few studies have reported stimulus-locked responses in cortex [11] , [12] , [13] . The question of how cortical neurons use the information encoded in spike times is fundamental in systems neuroscience [14] , [15] . Temporally coherent synaptic inputs due to background cortical activity , uncorrelated with stimulus onset , could reduce the precision of spikes relative to stimulus onset [14] . Thus , whether cortical neurons in vivo respond as precisely as those measured in vitro depends on the impact of the external stimulus in the context of the background cortical state [14] . We hypothesize that many cortical neurons receive common and/or synchronized inputs [16] , because there are 100 times more neurons in the primary visual cortex than there are in the retina or LGN [17] , and each spiny stellate cell receives inputs from tens of LGN cells [18] , [19] , suggesting that the same LGN cell projects onto many cortical cells . We refer to a group of neurons with common input as a neural assembly or ensemble [20] , which will be further explained in the Discussion . We performed in vitro experiments to determine how the time-course of neural spike trains is modulated by the strength of an aperiodic current injected into the soma . Although these experiments were conducted in vitro , they may shed light on the role of spike timing in vivo because in vitro , background synaptic activity can be tightly controlled [14] , [15] , [21] , [22] , [23] , [24] , [25] . The synaptic inputs to a neuron in vivo can be simulated in vitro by injecting aperiodic fluctuating current at the soma . In vivo , neurons also receive local cortical recurrent inputs that are modulated by other top-down inputs [26] such as those responsible for covert attention [27] , [28] . We approximated these effects in vitro by changing the amplitude and offset of the stimulus waveform ( Figure 1A and B ) . The response of one neuron obtained across multiple trials for different amplitudes or offsets can be interpreted as the response of a neural ensemble under the assumption that the neurons in the ensemble do not strongly interact with each other ( Figure 1C and D ) . According to this interpretation , precise spike times of a single neuron measured across trials can also be understood as synchronous volleys of a neural ensemble [22] , [29] , which are effective in driving postsynaptic cells both in vivo [30] and in vitro [31] . The strength of a volley depends both on the number of cells in the ensemble that produce a spike , which is related to the reliability across trials , and on their degree of synchrony , which is related to the precision . In a real neural ensemble , the interactions between the neurons will affect both the number of cells emitting a spike and their amount of synchrony , but the independence assumption [32] provides a starting point for thinking about how a stimulus can be represented across a population of similar neurons . Reliability is in principle different from precision [14] . We developed an event-based analysis , for which the sets of spike trains are represented as a set of spike time events . This analysis can be performed for spike trains recorded from the same neuron across trials as well as for spike trains of different neurons recorded on the same trial , as in ensemble reactivation in cortex or hippocampus . For ease of presentation , we will describe the analysis in terms of trials . Spike time occurrences , called events , are temporally localized concentrations of spike time density across trials , which can be characterized in terms of their occurrence time , precision and reliability . We investigated how these quantities vary with stimulus amplitude and offset . In the following we will use precision and jitter interchangeably to refer to the temporal resolution of spike times . We report three key results: First , we found that spike trains change with stimulus amplitude in such a way that the information about the stimulus time course was preserved . Some information about the amplitude is only reflected in the trial-to-trial variability and thus needs to be reconstructed based either on multiple trials or multiple units operating in parallel . Second , the general behavior as a function of drive parameter ( amplitude or offset current ) could be characterized in terms of spike patterns and bifurcation points in the dynamics of the membrane potential . Spike patterns are within-trial spike correlations , which may be due to afterhyperpolarization currents [33] and other slower currents activated by action potentials . At bifurcation points the spike times changed rapidly in response to a small change in parameter value; this change resulted in multiple spike patterns . The number of different spike patterns provided important information about the stimulus time course . We found that this number was highest for the intermediate amplitudes used in our experiments , as a consequence of the presence of bifurcation points . Third , we used an event-based analysis , which made it possible to semi-automatically analyze spike train data [34] . The entire procedure is characterized by four parameters , for which heuristics are given [34] . The main advantage of the event-based analysis is that it does not rely on fitting a specific parametric model for the neural dynamics based on the stimulus [35] , [36] , [37] , [38]; rather it models the data directly . Preliminary reports have appeared in abstracts [39] , [40] , [41] , [42] .
We define the stimulus amplitude β and DC offset α as follows: Given a fixed , fluctuating aperiodic current waveform , h ( t ) with a duration of 1000 ms , constructed to have an average value of zero and a given variance , we inject a current stimulus of the form I ( t ) = α+β h ( t ) ( Figure 1A and B ) . The mean DC current injected is α and the root mean square ( RMS ) size of the fluctuation is β times the RMS size of h . On any given set of trials we predetermined a set of relative amplitudes b and offsets a , which we scale with a multiplicative factor ν . This factor is determined on a cell-by-cell basis in order to adjust each suite of stimuli to the cell's intrinsic firing properties ( as described below ) . The resulting stimulus is I ( t ) = ν ( a+b h ( t ) ) , so the stimulus offset is given by α = ν a , and the stimulus amplitude is given by β = ν b . As the quantities a and b are common across experiments , the results will be presented in terms of a and b ( the latter as a percentage ) rather than α and β . The current so constructed was injected into the soma of layer 5 pyramidal cells in a slice of rat prefrontal/infralimbic cortex . The fluctuating drive was the same on each trial , but for the first experiment we used eleven different amplitudes , expressed as percentages . The first step was to determine the scaling factor ν , which varied from neuron to neuron according to its intrinsic properties . We used between 18 and 51 trials per amplitude and performed experiments on 10 different cells . For 8/10 cells , b ranged from 0% to 100% in steps of 10% , whereas for 2/10 cells b was 20% to 100% in steps of 8% . Because the injected waveform was prepared off-line and stored in a file , at the time of recording we could only adapt the overall gain ν to the properties of each neuron . The overall gain ( ν , taking values between 0 . 4 and 5 ) was chosen such that the neuron produced at least one spike for the lowest amplitude ( b = 0 or 20% ) , which was achieved for 8/10 cells . Rastergrams for one representative cell are shown in Figure 2A , in response to an injected current waveform . A spike-time histogram for all values of the current step and scaling factor are shown in the bottom panel . The rastergram consists of blocks of constant amplitude , with the highest amplitude on top . Within each block the trials are in the order they were collected , with the earliest at the bottom . In the rastergram ( Figure 2Aa ) , events are visible as spike alignments that appear for low to intermediate amplitudes and that become sharper as the amplitude is increased . Overall this graph suggests that both the precision , the jitter in the spike times belonging to an event , and the reliability , the fraction of trials on which a spike is present in an event , improve with amplitude . The trial-to-trial variability was characterized using the R-reliability ( see Methods ) with a sigma value of 3 ms [43]; that is , spikes in two different trials with a time difference of less than 3 ms are considered effectively coincident . This measure does not distinguish between changes in reliability and changes in precision ( as demonstrated previously [14] , [34] ) , which would require an event-based analysis . Overall , R increased with amplitude ( Figure 2Ab ) , but there were dips , indicated by the arrow . As will be shown below , the presence of a dip is characteristic of a bifurcation point , where a small change in the stimulus can lead to a large shift in the pattern of spikes . The spike times for an event shifted only slightly as a function of stimulus amplitude , as demonstrated by the relatively sharp peaks in the spike time histogram across all amplitudes ( Figure 2Ac ) . The distribution of spikes was broad , reflecting the responses for the lowest amplitudes , which were not strongly stimulus-locked , but had a sharp , approximately symmetric peak , corresponding to the highest amplitudes . To interpret these results , consider a neural ensemble comprised of different neurons receiving feedforward inputs with strengths that are different because the synaptic inputs could have different decay time constants or short-term plasticity [44] and the neurons could have different input resistances and resting membrane potentials [45] . The in vitro experiments show that despite this diversity the ensemble would produce volleys that are effective in driving postsynaptic neurons and transmitting information about the time course of the input [21] . This effect was also observed in vivo comparing the responses of different cat LGN neurons to the same flickering light stimulus [4] . Despite a wide range of firing rates , some of the events in these recordings were consistent across neurons of the same type even in different cats . For the second set of experiments , the current offset , a , was varied between 0 . 05 and 0 . 3 nA in 10 equal steps , while the amplitude of the stimulus waveform was held constant at b = 0 . 05 nA ( for the example in Figure 2B , in which amplitude is expressed as a percentage ) . We presented 7 such stimulus sets ( amplitudes b = 0 . 02 to 0 . 06 nA ) to 5 different cells and recorded the responses on 17 to 36 trials , with overall gain factors ν between 2 . 7 and 3 . The overall behavior was similar: Precise spiking was obtained at all current offsets , with some common event times ( Figure 2Ba and c ) , and the reliability measure R increased with current offset ( Figure 2Bb ) and also displayed a dip ( arrow in Figure 2Bb ) . The differences between the effects of varying offset and varying amplitude were , first , the firing rate increased when current offset was varied ( from 1 . 0±1 . 4 to 15 . 5±0 . 8 Hz for Figure 2B , mean ± standard deviation across trials , corresponding to an increase from the lowest to the highest offset , relative to the maximum rate , of 0 . 92; across the population , N = 9 , this was: mean 0 . 88 , range 0 . 32 to 1 . 0 ) more than for the case where the amplitude was varied ( from 10 . 5±1 . 1 to 14 . 5±0 . 9 Hz for Figure 2A , corresponding to 0 . 28 , across the population , N = 6 , mean 0 . 75 , range 0 . 28–1 . 0 ) . This is because the overall level of depolarization increased , whereas for increasing amplitude not only did the peaks increase , but the troughs also got deeper , which meant that some spikes would appear and other spikes would disappear . Because we could not perfectly adapt the relative level of amplitude and offset , such that for zero amplitude there already was some spiking , the across-the-population difference in firing rate between the two cases was smaller . Second , for a nonzero firing rate , the neurons immediately phase locked to the waveform in the current offset case , which resulted in higher reliability ( we excluded the first current offset , for which the neurons only spiked a few times ) , even at a rate of a few Hz , compared to the amplitude case . Third , the distribution of spike times in an event was asymmetric , with the peak skewed to earlier times . This was because the increasing depolarization made the spikes reach threshold earlier . Note , however , that for the low current offset trials the first spike times appeared to drift with current offset , but they actually shifted to earlier events on some trials . Identification of events in a peristimulus time histogram is essential for the analysis of stimulus encoding via spike time patterns . Because the events visible in the multi-amplitude rastergrams ( Figure 2 ) persist across amplitudes , the precision , reliability and mean spike time of events can be compared across amplitudes ( see Methods ) . One strategy for event-based analysis would be to find events for each amplitude and merge events common across amplitudes . In Figure 3 we show the results of an alternative strategy , in which spike train ensembles generated by the five highest amplitudes are analyzed at the same time . With this approach the underlying pattern-finding procedure ( see Methods ) is more robust because there are more trials in each pattern [34] . Our analysis revealed the presence of four patterns ( Figure 3B ) , which led to 8 events , some of which were common to multiple patterns . We recall that events are temporally localized concentrations of spike time density across trials and patterns are transient multiplicity of spike response . In panel B the patterns are divided by gray horizontal lines , whereas all the spikes belonging to an event are enclosed in a gray vertical box . Because there were more amplitudes ( five ) in the data set than there were patterns ( four ) , a given pattern had to persist across multiple amplitudes . We plotted the rastergram in blocks of constant amplitude , and sorted the trials on the basis of the pattern they expressed ( Figure 3C ) . The fraction of patterns that were present varied across amplitude and was quantified in Figure 3D . As the fraction of trials with the second pattern increased ( Figure 3D , line 2 ) , the fraction of trials on which the first pattern was present decreased ( Figure 3D , line 1 ) . Hence , the reliability of events in the second pattern increased ( black ellipses in Figure 3C ) , whereas the reliability of events in the first pattern decreased ( gray ellipses in Figure 3C ) . As patterns fade in and fade out during variation of amplitude , the mix of patterns present across trials at a given amplitude varies with amplitude . Thus the non-monotonic change in reliability with amplitude indicated by the arrows in Figure 2Bb reflects the changes in the mix of patterns . The diversity of patterns present for a given amplitude is quantified using the entropy of the pattern distribution ( Figure 3E ) . For these data , the entropy was maximal at a specific amplitude ( see Methods , arrow in Figure 3E ) . In other data sets and segments , the entropy decreased monotonically with increasing amplitude because for higher amplitudes only one pattern survived . These results are relevant to the amount of information that can be extracted about the temporal dynamics of the injected current from the spike time patterns and is discussed below . An important question is what aspects of the membrane potential are reflected in , or can be reconstructed from , the measured spike times , because the membrane potential itself is not accessible from in vivo extracellular recordings . Without the membrane potential little information can be obtained about excitatory and inhibitory inputs to the neuron [46] , needed to test hypotheses about computational mechanisms . To address this in vitro , we obtained multiple trials and analyzed recordings where the same waveform with the same amplitude was injected on each trial . For these experiments , the fluctuating waveform was extended to 1700 ms and was preceded by a constant current offset of 200 ms ( in addition to 50 ms zero current at the start of all current injections ) . In these experiments , the initial current offset took eleven different values , the influence of which are discussed below . We injected this drive in nine experiments using eight cells , with between 10 and 35 trials each ( 110 to 385 trials overall , ignoring the initial current offset ) and an overall gain with values between ν = 0 . 9 and 4 . 7 . The main difference between this data set and the ones used in Figure 2 and 3 is that there are more trials available for statistical analysis . Spike patterns correspond to within-trial correlation between spike times . We determined how long these correlations persisted by applying the spike pattern analysis to approximately 500 ms long segments from seven of the nine available data sets from which we show one ( Figure 4A ) . The segment length was chosen such that at least two events , and no more than six , were present . Within each segment the trials were ordered according to the pattern they expressed in that segment . This shuffles the trials differently in each of the segments and the spikes on a single row of the rastergram most likely correspond to a sequence of segments from different trials . We then determined how well the pattern expressed on a trial during one time segment predicted which pattern was expressed in a preceding or following segment , that is , the between-segment correlation of the patterns that neurons express . Strong correlation means that if in one segment a group of trials express the same pattern , they will also do so in the other segment . In Figure 4B , the trials were reordered in each segment based on the order in the last ( fourth ) segment ( indicated by the asterisk ) . For this case , each row is one and the same trial on each segment in contrast to the display in panel A . This panel shows that even though a group of trials expressed the same pattern during segment 4 , that same group expressed a mixture of patterns during segment 3 – indicative of a low between-segment correlation . This association between different segments is best expressed as the normalized mutual information between the pattern classification of a trial in two segments ( IN , see Methods , which is primarily used here to summarize a two-dimensional array of transfer probabilities rather than to make statements about information content ) . The maximum value was normalized to one , which occurs if the classifications are identical . The IN ( bias; std ) between the classification in segment 4 and that in segments 3 , 2 or 1 was 0 . 20 ( 0 . 01; 0 . 04 ) , 0 . 003 ( 0 . 009; 0 . 012 ) , or 0 . 025 ( 0 . 007; 0 . 020 ) , respectively . We further analyzed the three patterns uncovered in segment 4 between 1500 and 1900 ms . For each pattern , the voltage traces were averaged across all corresponding trials and the standard deviation was used as an estimate for the trial-to-trial variability . The mean voltage traces differed not only because the neurons spiked at different times , but also because the conductances associated with the afterhyperpolarization following the spike had a long-lasting influence on the response to the current injection ( Figure 4E , top; averages from patterns 1 to 3 ) . The spikes reflected periods where the injected current had a large positive slope , but each pattern was triggered by a different subset of these upswings ( Figure 4E , bottom ) . Once a spike was produced , the neuron did not spike during an otherwise viable upswing shortly thereafter , even though it had produced a spike there on other trials during which it expressed an alternative pattern . For instance ( Figure 4E , bottom ) , on trials labeled 1 , the neuron did not spike in response to the upswing that caused the neuron to spike on trials labeled 2 because the membrane was hyperpolarized . To uncover correlations that persisted across segments , for each pattern expressed on segment 4 , we averaged the voltage traces across all trials belonging to that pattern for the entire duration of the trial . For clarity , we only show three time intervals ( Figure 4C to E ) . In the first interval between 100 and 250 ms there was a small difference in the mean membrane potential ( Figure 4C ) , which had disappeared by t = 1000 ms ( Figure 4D ) , but reappeared after a deep hyperpolarization ( Figure 4D , arrow ) . This difference then led to three clearly distinct voltage patterns in the last interval ( Figure 4E ) . Thus , whether or not a neuron spiked at a given transient depolarization determined the subsequent firing pattern for hundreds of milliseconds , a time scale comparable to the time needed to process a visual image . The neuron's internal state determined whether or not a neuron fired a spike at a given time , which depended on its previous history as well as the current membrane potential . This could include the height of the depolarizing step that preceded the fluctuating current , or whether the trial was at the beginning or end of the experiment . The normalized mutual information between the trial number and the pattern on segment 4 ( the degree of non-stationarity ) was IN = 0 . 059 ( bias: 0 . 001; std: 0 . 011 ) , and between the height of the offset and the pattern it was IN = 0 . 050 ( bias: 0 . 006; std: 0 . 013 ) . This analysis shows that there was an influence but only a small fraction of the variability can be explained by these two factors . Among the other eight data sets , two had too low firing rates , which meant that there was less than one spike during segment 4 , these were not further analyzed . For the remaining six data sets ( see Figure S1 ) , we found that in five of them non-stationarity caused high normalized mutual information between the trial number and the resulting pattern on segment 4 , which also led to large values for the IN between patterns on different segments . The one remaining data set was stationary ( 052701; Figure S1C ) , but , as assessed by the normalized mutual information , there was little correlation between patterns on different segments . Our procedure was robust against non-stationarity in the sense that for small shifts in the spike times during the course of the experiment , these spikes would still be assigned to the correct event , whereas for large shifts , the later trials would be classified as a different pattern . Nevertheless , the typical procedure was to remove the last few trials from a non-stationary data set to obtain an approximately stationary one to analyze . Apart from suggesting a mechanism for how intrinsic neural properties generated correlations between patterns in different segments , we also showed that the analysis procedure was robust against the effects of non-stationarity Without knowing the internal state of a neuron it is difficult to draw any conclusions about dynamical mechanisms that might be responsible for the observed diversity of spike patterns . The same current inputs were used to study the Wang-Buzsaki ( WB ) model neuron [47] modified by including a small additive noise current as a source of independent trial-to-trial variability ( see Methods ) . We simulated 50 trials for each of 101 different amplitudes b between 0 and 100% ( the stimulus h ( t ) was normalized to have zero mean and unit standard deviation ) . Figure 5A shows the reliability curve , which was smoothed by a three-point running average , and Figure 5B shows the corresponding rastergram with matching amplitudes . For clarity we only displayed half of the trials and half of the amplitude values in the rastergram . The R-reliability generally increased with amplitude but showed multiple local minima . Two such minima are highlighted by the arrows and correspond to the spike train features inside the circle of matching gray-scale in Figure 5B . The bottom half of the gray circle intersects two reliable spike times and at higher amplitudes the top half of the circle intersects three spike times . The spike train ensemble exhibited a bifurcation within this region because one pattern of events branched into another . For continuous models ( Hodgkin-Huxley , as opposed to the leaky integrate-and-fire ) the spike time “bends” , that is , shifts to earlier times , after which it stops shifting because it arrived at a previously-subthreshold peak , and at some point the original spike time ( the one that was shifted ) emerges again . For the leaky integrate-and-fire model , new spike times get inserted de novo . Moreover , dynamics changed rapidly for small changes in the parameter value . Inside the circle there were fuzzy clouds of spikes corresponding to multiple competing patterns for a given amplitude . As the amplitude increased , the fraction of patterns with two spikes decreased , whereas the fraction of those with three spikes increased . A similar transformation took place within the black circle , where a three spike-time pattern occurred on the bottom half of the circle and as b increased a four spike-time pattern appeared on the top half of the circle . The peaks in the reliability ( gray curve ) coincided with plateaus where the spike count was constant as a function of stimulus amplitude b ( black curve ) ( Figure 5C ) , and the variability of the spike count was minimal ( Figure 5D ) . A similar behavior was obtained when the current offset a was varied , holding the amplitude b fixed . In both the model and in the in vitro experiments , the firing rate increased more rapidly with increasing normalized offset ( 0≤a≤100% ) than with normalized amplitude ( 0≤b≤100% ) ; and the reliability was high even at low firing rates . In contrast , for small amplitudes , a low R-reliability similar to that in response to a current step was obtained , because the stimulus lacked temporal structure . The spike train ensemble represents specific features of the stimulus with the timing , reliability and precision of spike event patterns in the ensemble . Away from bifurcation points , there is only one spike pattern and the neuron spikes with a high precision and reliability at a subset of stimulus upswings . When there are more upswings than there are spikes , information about the time course of the stimulus is lost . By contrast , near a bifurcation point the dynamics is more sensitive to noise , and multiple spike patterns are obtained with non-overlapping event times . Each pattern provides information about a subset of stimulus features . Given a natural ( fluctuating ) input , the precisely timed spikes provide evidence for an upswing in the stimulus . Absence of a spike could reflect a downswing of the stimulus or just the relative refractoriness of the cell following a spike . The coexistence of multiple patterns means the spike train ensemble provides a richer description of the stimulus and is therefore in principle more informative , in the sense that an ideal observer can extract more information about the stimulus [48] . We investigated the ability of an ensemble to reconstruct an input waveform in different noise regimes . In order to focus on stimulus reconstruction via ensemble event detection , we drove the Wang-Buzsaki model neuron with a frequency-modulated ( FM ) waveform . The FM waveform is simpler than a general Gaussian process , in that comprises a series of distinct upswings of equal amplitude , but at variable intervals and slopes . We compared the response at a bifurcation point under two circumstances: low noise and medium noise . In this way we could compare responses at the same amplitude with a similar , but not identical , spike rate . For the low-noise case in Figure 6A , there were six events during the time interval displayed ( bottom , the ticks representing the spikes coalesced into gray vertical lines ) , but for the medium-noise case there were additional events with a reduced reliability and precision ( top ) . We applied the event-based analysis to the entire simulation time interval ( 1100 ms ) and determined for each event the reliability and precision ( Figure 6B ) . We also determined the spike-triggered average ( STA ) of the stimulus for both cases ( Figure 6C ) . Using all detected events with a reliability exceeding 5% and the measured STA , we reconstructed the stimulus waveform ( Figure 6D ) . The ensemble of medium-noise spike trains , with their multiple spike patterns , yielded a better reconstruction ( middle ) than the low-noise case ( top ) from a single spike pattern . To interpret these results , consider the response of a neuron to a volley with an average number n of spikes ( n = R×N , where R is the event reliability and N is the number of neurons in the ensemble ) each with some jitter . Each event contributes to a reconstruction of the input according to its reliability and precision . The gray shading in Figure 6B contains those events that might contribute to the reconstruction because they have the most reliable spikes with the lowest jitter . The shape of the shaded area represents the intuitive idea that when the precision is less , you need more input spikes to still produce an output spike , i . e . a higher reliability . The shown curve is hypothetical , but it is based on the simulation results reported in [49] , [50] , [51] . Experimental support derives from studies using two-photon uncaging of glutamate to produce synchronous ensembles of inputs that trigger dendritic action potentials and thereby reliably and precisely produce somatic action potentials [31] . The beneficial effects of the medium noise regime demonstrated for a frequency-modulated drive hold for more general drives , including the aperiodic drive used in the experimental studies . The basic requirement is that there are three noise regimes , which is the case for fluctuating drives , be it periodic or aperiodic , but their ranges in terms of the range of noise standard deviation vary with the specific details of the drive . For weak noise , the neuron spikes at perfect reliability and the jitter is proportional to the square root of the noise strength . The spikes sample only some of the upswings . For medium noise , there are still precise events , but the reliability is reduced , because there are multiple patterns . Due to the multiplicity of patterns more upswings are sampled . For strong noise , there are no longer discrete spiking events any more , rather the spike time density follows the temporal dynamics of the driving current . The beneficial effect occurs for medium noise strength , when more upswings are sampled , but the events are still precise enough to have a strong postsynaptic effect .
Our overall goal was to use the in vitro experiments to gain insights into the dynamics of neural ensembles in vivo . This raises two issues: First , what is the nature of neural ensembles , and second , how do they represent information ? There are two extreme hypotheses about how information is represented in an ensemble . The first holds that slow modulation ( 50–500 ms ) of the firing rate encodes stimulus properties . This firing rate can be estimated by averaging across the relevant neurons in the ensemble , from which the time course of the stimulus can be reconstructed . The second hypothesis is that ensembles of neurons produce precisely timed spikes , which lead to synchronous volleys or events that are effective in driving postsynaptic neurons . Experimental results , reviewed in [14] , suggest that a combination of these two strategies is at work at the level of the cortex . First , volleys represent the fast fluctuations in the inputs to the cortex [19] or internally generated events [6] , [7] . Second , slow modulations change the number of volleys and the number of spikes per volley . The two information channels can work together on two different time scales and even interact . More spikes are fired in neurons whose tuning properties better match the sensory input , which raises the level of activity in the ensemble and engages new dynamical mechanisms , such as gamma oscillations , that in turn generate synchronous events within the cortex and increase downstream firing rates [66] , [67] . Spatially clustered , temporally precise synaptic inputs to pyramidal neurons in vitro are effective at eliciting reliable and precise spikes through dendritic action potentials [31] . This decoding mechanism is decoupled from the spikes generated by slow input modulations and has an all-or-none character . If , for instance , 50 spikes in a 2 ms long interval are enough to elicit an output spike , then increasing the number of spikes or their precision will generally not increase the number of spikes produced in response to this volley . Any volley meeting the minimum requirement will elicit a spike and will be able to transmit information about the upswing that generated it . Hence , when there are multiple precise spike patterns , the neural ensemble provides more information than a single pattern , despite the reduced reliability at the single neuron level . This suggests that it is beneficial for the nervous system to keep ensembles close to bifurcation points so that they are most informative about the dynamics of their inputs . In general , a neural assembly refers to a group of cells sharing dynamically in functionally related activity . An assembly may or may not correspond to an anatomically distinguished set of cells such as a cortical column . In our simulations , the network is essentially feedforward; experimentally , synaptic transmission was blocked . In the systems we considered , therefore , correlations resulted from common inputs . However , our analyses did not assume any specific underlying architecture , and so would apply generally to any ensemble of cells . In a separate recent work , we have explored the reliability of a similar network with added feedforward inhibition , and established that the influence of inhibitory inputs was negligible when compared to the influence of feedforward excitatory drive and synchrony [19] . There is ample experimental evidence for the existence of functional neural ensembles . For example , fewer than 5% of the synaptic inputs to spiny stellate cells in layer 4 of the cat visual cortex are from thalamic relay neurons [18] , [68] . Dual recordings from thalamic neurons and cortical neurons reveal a high degree of convergence [10] , and analysis of these data show that as few as 5 thalamic cells with 5 synaptic contacts are sufficient to reliably elicit a spike [19] . Thus synchronous volleys of spikes from thalamocortical neurons insure that visual information enters the cortex despite background activity in both the thalamus and the cortex . As another example , recordings in hippocampus and cortex show coordinated reactivation of small assemblies during sleep and quiet awakeness [6] , [7] . Similarly , recent in vivo recordings in rodent barrel cortex [69] show that under certain circumstances nearby neurons have correlated membrane potentials , indicating that they receive similar inputs . Furthermore , their spikes are preceded by large , sharp deflections of the membrane potential , indicating the presence of synchronized volleys . Nevertheless , even neurons in the same cortical column have diverse morphologies , different input conductance and other physiological differences . Our analysis shows that despite this heterogeneity , temporal information is robust across a range of parameter values and can be combined into precise volleys even across a moderately heterogeneous ensemble . Bifurcation points allow noise to enrich the representation of stimulus features . This effect superficially resembles subthreshold stochastic resonance ( SR ) , but is distinct from it . In traditional SR [70] , [71] , [72] noise can enhance signal detection by allowing a spiking neuron to reach threshold even when the applied stimulus remains below threshold . SR is effective when a subthreshold stimulus is close enough to threshold to allow noise facilitated spiking . Such a stimulus is a fortiori near a bifurcation point , because increasing either the current offset or the amplitude of the fluctuating input will bring it above threshold . In this sense SR may be viewed as a special case of a bifurcation point phenomenon analogous to that described here . Both SR and spike-time bifurcation phenomena involve a time varying deterministic input signal and a detection process . In classic SR both the mean input signal and the entire signal remain below threshold , while in our experiments the mean input signal was suprathreshold . This distinction may be readily appreciated in a simple threshold-crossing model such as a leaky integrate-and-fire ( LIF ) neuron driven by additive noise currents . For simplicity , suppose a LIF model cell with unit membrane time constant is driven by a combination of fixed ( deterministic or “frozen noise” ) and noisy input currents:with the voltage reset to V0 = 0 upon reaching the threshold Vth = 1 . Here dWi represents white noise forcing ( increments of a Wiener process ) , which is taken to be independent on each trial ( or for each cell , in the simultaneous ensemble interpretation ) . The expression represents the injected current drive , with mean , and h ( t ) a zero-mean , function with , and a fluctuating component amplitude set by . The function h ( t ) can either be a deterministic function , such as a sinusoid or combination of sinusoids , or a predetermined “frozen noise” stimulus . The detailed analysis of this system is a source of open mathematical problems [73] . The analysis of the system is qualitatively different in the suprathreshold case ( ) , the subthreshold case ( ) and the mixed or perithreshold case ( ) [74] , [75] . If one were to assume the existence of a current threshold for spiking as in the LIF model , then in a similar way the applied stimulus ( whether deterministic , such as a sinusoid , or stochastic , as in the frozen noise paradigm ) may either be subthreshold ( spikes cannot occur without noise ) , suprathreshold ( spikes can occur at any time; spike occurrence is not noise dependent; noise only weakly modulates the times of individual spikes ) or perithreshold ( the stimulus repeatedly crosses the current threshold; spike occurrence and spike timing depend both on the noise and the stimulus; the hyperpolarizing downswings of the fluctuating component become large enough to exclude spiking over certain time intervals; this regime corresponds to the “forbidden zones” mechanism explored in [76] ) . In our preparation , the DC component of the input current ( , above ) was always set to be large enough to guarantee that the cell would fire a train of action potentials . The mechanism of spike time precision and reliability in both the suprathreshold and perithreshold regimes is distinct from the mechanism underlying standard stochastic resonance in the subthreshold regime . The experiments reported here fall in the suprathreshold case when the stimulus amplitude is low , and transition to the perithreshold case as the stimulus amplitude increases . The regime in which stochastic resonance may occur corresponds to the subthreshold case , which was not explored here . Hunter and Milton demonstrated , in both in vitro and computer experiments , that spike timing reliability is sensitive to the frequency content of an aperiodic injected current in a way that depends on the relative amplitude of the fluctuating drive component [77] . For low amplitude inputs they found that noise dominated the spike times and that the reliability was low regardless of drive frequency content . For high amplitude inputs the spike times were determined by large current upswings , leading to high reliability regardless of frequency content . For intermediate amplitude inputs , however , the reliability was strongly influenced by input frequency content . Here we surveyed the entire range from weak amplitudes to strong . We found that bifurcation points enhanced the representation of inputs by neural ensembles primarily for intermediate amplitude values . This observation suggests that the presence of bifurcation points , and the frequency dependence of spike-time reliability [78] , may occur in overlapping intermediate amplitude regimes . The operating point of the cortex depends on many factors , including the level of arousal and the behavioral state of the animal . The presence of spontaneous background activity even during relaxed behavioral states suggests that the intracellular membrane potential is balanced just below threshold [79] . This allows neurons to react rapidly to sudden changes in synaptic inputs and also insures a high sensitivity to synchronous events in neural ensembles . Thus , the spontaneous activity should not be considered noise but a dynamical variable that can be adjusted in magnitude and variance to enhance the function of the cortical area . In particular adjusting the background activity could be one of the targets of some neuromodulators , such as acetylcholine . The experimental results presented here apply to uncoupled neural ensembles receiving feedforward inputs . This clearly is an approximation to in vivo dynamics because: ( 1 ) There are recurrent synaptic connections that generate coherent activity in various frequency ranges and across different spatial scales [80]; ( 2 ) There is feedforward inhibition that follows feedforward excitatory volleys at a small delay [81] and ( 3 ) There are recurrent loops between the different cortical layers [82] , [83] , [84] . Nevertheless , at the soma/spike generating zone the sum of these inputs leads to a fluctuating drive . This fluctuating drive consists of fast fluctuations riding on top of slower modulations . We have shown that spike timing is generated by these fast fluctuations and modulated by slow fluctuations . The relative phase between gamma oscillations in two brain areas can modulate the effectiveness of communication [85] . Hence , an important issue for further study is how fast feedforward volleys interact with internally generated oscillations in the gamma frequency range [67] . In the in vitro experiments , current is injected at the soma . But in vivo the excitatory and inhibitory synaptic inputs are spatially distributed across the soma and dendritic tree . In particular , somatic current injection bypasses the nonlinear integration that takes places in the dendrites [86] , [87] . In addition , the changes in conductance due to opening of synaptic channels are absent from the current injection . We have partially accounted for these dendritic effects by varying the amplitude and offset of the current injection motivated by analyses of the effects on the firing rate versus current ( f-I ) characteristic . Adding a constant conductance approximately shifts the f-I curve of a leaky integrate-and-fire neuron to the right , similar to the response of a hyperpolarizing current offset [88] . Synaptic inputs add a fluctuating conductance , which in models can change the gain of the f-I multiplicatively and increase the impact of other fluctuations . In vitro experiments show random background activity produced by a network can cause changes in the gain or sensitivity to other inputs [89] , [90] . This is not the same as changing the amplitude , but the effects on the neuron are similar . The offset and amplitude approximate a number of effects . First , because neurons in the ensemble are not identical , there is diversity of offset and amplitude values across the ensemble . Our results show that despite this diversity , the ensemble can produce synchronous volleys . Second , when the network state changes , as might occur in response to top-down modulation , the overall offset and amplitude changes . In vivo experiments have documented corresponding changes in gain and sensitivity in response to top-down activation of cortical networks [28] , [91] . Interestingly , the in vitro experiment across trials may be a better approximation of a neural ensemble on one trial than the response of one in vivo neuron across multiple trials because the drive component that varies from trial to trial has a large shared component among different neurons in the ensemble [60] . Precision and reliability are distinct properties of neural dynamics , although they may be modulated in a correlated fashion . There is a need both for an easy way to characterize the overall variability as well as for parsing out the reliability and precision separately . A reliability measure such as R-reliability [43] is appropriate for the former purpose , because it quantifies the overlap between pairs of spike trains , at a given temporal scale . But it does not directly indicate whether this overlap is due to a reduced reliability or precision , nor whether there are multiple patterns . Because reliability and precision are event properties , the event structure needs to be extracted in order to quantify them . Distinct events may be overlapping , however , making it difficult to separate them . As an example , consider sampling from two Gaussian densities with means of −0 . 5 and 0 . 5 , respectively , both with a standard deviation of 1 . It is not possible to say to which of the two distributions a sample point at 0 belongs to . However , by exploiting the context provided by the history of the spike train , the overlapping events can be separated . Hence the ensemble of spike patterns plays a crucial role in disentangling the reliability and precision of the spike train ensemble . Because the event-based analysis procedure only depends on a few well-defined parameters for which heuristics are available , it is reproducible from lab to lab [34] . Briefly , the temporal resolution at which the spike patterns can be optimally distinguished is first determined; then the number of patterns is found; after which the events for each pattern are determined and finally , events common to multiple patterns are merged . Mathematically , bifurcations of spike time patterns due to addition or removal of individual spikes resemble so-called grazing bifurcations present in non-smooth dynamical systems such as impact oscillators [92] , [93] , [94] , [95] . Recent advances in the understanding of stochastic suprathreshold leaky integrate and fire model neurons may help clarify the structure of spike time bifurcations in mathematical detail , although many details remain for future study [73] . Optogenetic technologies utilizing light-activated channels and pumps together with the read-out of neural activity via two-photon microscopy makes it possible to manipulate and record from neural ensembles in vivo [96] , [97] , [98] , [99] , [100] . This should make it possible to move neural ensembles away from or towards bifurcation points . Furthermore , in vitro , at the single neuron level , rapid spatially-distributed glutamate uncaging [31] , [101] can be used to determine how neurons respond to the feedforward inputs generated by a neural ensemble and how these inputs interact with pharmacologically generated fast oscillations . Taken together , these technologies offer the opportunity to test in vivo predictions about the functional role of neural ensembles positioned at bifurcation points .
The voltage response of cortical neurons was measured in a rat slice preparation as described previously [102] . The Salk Institute Animal Care and Use Committee approved protocols for these experiments; the procedures conform to USDA regulations and NIH guidelines for humane care and use of laboratory animals . Briefly , coronal slices of rat pre-limbic and infra limbic areas of prefrontal cortex were obtained from 2 to 4 weeks old Sprague-Dawley rats . Rats were anesthetized with isoflurane and decapitated . Their brains were removed and cut into 350 µm thick slices on a Vibratome 1000 ( EB Sciences , Agawam , Mass . ) . Slices were then transferred to a submerged chamber containing standard artificial cerebrospinal fluid ( ACSF , mM: NaCl , 125; NaH2CO3 , 25; D-glucose , 10; KCl , 2 . 5; CaCl2 , 2; MgCl2 , 1 . 3; NaH2PO4 , 1 . 25 ) saturated with 95% O2/5% CO2 , at room temperature . Whole cell patch clamp recordings were achieved using glass electrodes containing ( 4–10 MΩ; in mM: KmeSO4 , 140; Hepes , 10; NaCl , 4; EGTA , 0 . 1; Mg-ATP , 4; Mg-GTP , 0 . 3; Phosphocreatine , 14 ) . Patch-clamp was performed under visual control at 30–32 °C . In most experiments Lucifer Yellow ( RBI , 0 . 4% ) or Biocytin ( Sigma , 0 . 5% ) was added to the internal solution for morphological identification . In all experiments , synaptic transmission was blocked by D-2-amino-5-phosphonovaleric acid ( D-APV; 50 µM ) , 6 , 7-dinitroquinoxaline-2 , 3 , dione ( DNQX;10 µM ) , and biccuculine methiodide ( Bicc; 20 µM ) . All drugs were obtained from RBI or Sigma , freshly prepared in ACSF and bath applied . Data were acquired with Labview 5 . 0 and a PCI-16-E1 data acquisition board ( National Instrument , Austin Tex . ) at 10 kHz , and analyzed with MATLAB ( The Mathworks ) . We applied the event finding method to data collected to study the effect of varying the amplitude and offset of a repeated “frozen noise” stimulus . For all experiments the same frozen noise waveform h ( t ) was used . A white noise waveform ( sampling rate 10 kHz , with samples uniformly distributed on the unit interval ) was generated using the MATLAB function rand with the state of the random number generator set to zero . It was twice filtered using the MATLAB routine filter ( afil , bfil ) . First , we applied a low-pass filter with a corner frequency of 100 Hz , obtained by setting afil = [1 , −0 . 99] and bfil = 1 . Second , we performed a 50-sample ( 5 ms ) running average ( afil = 1 , bfil has fifty elements equal to 1/50 ) . The first 500 samples ( i . e . 50 ms ) of the transient were discarded and the remaining waveform was centered on zero by subtracting the mean and normalized to have unit variance by dividing by the standard deviation . Depending on the cell and the quality of the seal , the waveform was multiplied by a factor ν representing the maximum amplitude . Waveforms with fractional amplitudes ( b ) ranging from zero to one were presented to the cell , as listed in the main text . In other experiments the waveform amplitude was held constant but its offset ( a , i . e . mean ) was varied instead . We also conducted experiments to determine the effect of the initial state of the neuron , which for our purposes was the membrane potential at the start of the actual stimulation , on the response to a fluctuating stimulus waveform . In these experiments , the level of depolarization of the initial constant-current step was varied , but the amplitude and offset of the fluctuating current was held constant . The fluctuating stimulus was increased in length to 1650 ms without changing the initial 1050 ms and a longer constant step was put in front of this stimulus . Trials were separated by at least 15 sec of zero current injection , to let the membrane return to its resting state . Throughout the experiment , a few hyperpolarizing pulses were injected to monitor the access resistance of the preparation . These pulses were clearly separated from other stimuli . Spike times were detected from recorded voltage traces as the time the membrane potential crossed 0 mV from below . The firing rate was the number of spikes recorded during a trial , averaged across all similar trials and normalized by the duration of the trial in seconds . Here “similar” means having the same amplitude and offset . In the rastergram , each row represented a spike train from a different trial . Each spike is represented as a tick or a dot , with the x-ordinate being the spike time and the y-ordinate being the trial number . Often we group trials together based on the stimulus amplitude or re-order trials based on which pattern they belong too . This is indicated in the corresponding figure caption . The spike time histogram is an estimate for the time-varying firing rate . It was obtained by dividing the time range of a trial into bins ( typically 0 . 5 ms wide ) and counting the number of spikes that fell in each bin across all trials . The bin count was normalized by the number of trials and by the bin width in seconds . The latter was to assure that a bin entry had the dimensions of a firing rate , Hz . The histogram was subsequently smoothed by a Gaussian filter with a standard deviation equal to 4 bins . The spike-triggered average ( STA ) was obtained for each spike by selecting the 25 ms stimulus segment prior to the spike and averaging across all spikes . Events were detected using the procedure summarized below . At the end of this procedure , all spikes were either assigned to an event or were classified as noise . The event-reliability is the fraction of trials on which a spike was observed during that event , and the event-jitter is the standard deviation of the spike times belonging to the event . The event-precision is the inverse of the event-jitter . For a given condition ( amplitude , offset or initial current step ) the reliability , precision and jitter are defined as the event-reliability , event-jitter and event-precision averaged across all events . The R-reliability is calculated based on all spike times without detecting events . The spike trains are first transformed into a continuous waveform , where each spike is convolved with a Gaussian distribution with a standard deviation sigma [43] , [103] . This procedure eliminates quantization noise that would otherwise result from a priori binning of the spike time data [104] . The cosine of the angle between the two waveforms , when considered as vectors , is computed as the inner product between the waveforms of trial i and j , normalized by the product of the square roots of the inner product of each trial with itself . This quantity is a number between 0 and 1 ( the waveforms are positive valued ) and is called the similarity Sij . The reliability estimate R is the mean of Sij across all distinct pairs <ij> . Intuitively , the inner product measures the degree of overlap between spike times: the closer two spike times are , the larger the overlap and thus their contribution to the inner product . Sigma sets the time scale of the reliability measure and determines which spike times between the pairs are considered overlapping . For sigma approaching zero , all spike times are considered different ( except when the spike trains are identical to machine precision ) , hence R = 0 . For sigma much larger than the trial length , all spikes overlap and R = 1 . In the first case , the emphasis is on precise spike times; in the second case , the emphasis is on the global amount of spikes ( spike rate ) . We used a more efficient method for calculating R by summing , for each pair of trials <ij> , the following expression across all spike pairs <kl> that are separated by less than six sigma's , ( here is the kth spike on the ith trial and for simplicity the normalization was omitted , see [104] for details . Briefly , the Victor-Purpura ( VP ) metric [105] calculates the distance between two spike trains A and B by calculating the cost of transforming A into B ( or B into A - the measure is symmetric ) . This distance is obtained as the minimum cost of transformation under the following rules: adding or removing a spike from A costs +1 point , while sliding spikes forward or backward in time by an interval dt costs q times . The variable q ( units 1/ms ) represents the sensitivity of the metric to the timing of spikes . For large q values it is frequently cheaper to add and remove spikes than to move them . Hence , for very large q , the metric is simply the number of spikes with different times between the two trains . For small q values , spike moving transformations are cheap , leaving the majority of the metric's value to the difference in the number of spikes which must be added or removed to produce train B; in the limit , the metric becomes the difference in the number of spikes in each spike train . In a recent study we uncovered multiple spike patterns in trials obtained in response to repeated presentation of the same stimulus both in vitro and in vivo [4] , [54] . Spike patterns are present when trials , or at least short segments thereof , can be separated in two ( or more ) distinct groups of spike sequences . As an example consider the case where on some trials the neuron spikes at 10 and 35 ms ( relative to stimulus onset ) , whereas on other trials it spikes at 15 and 30 ms . These trials would consist of two distinct patterns as long as there are no trials with spikes at 15 and 35 ms or at 10 and 30 ms . Hence , spike patterns correspond to a within-trial correlation , because , in the example , a spike at 10 ms implies that a spike will be found at 35 ms with a high probability . The problem of finding spike patterns is made harder by the presence of spike time jitter and trial-to-trial unreliability . We designed a method to uncover patterns independently of the event structure and then used the patterns to construct the event structure . The full method is described elsewhere [34] , here we only summarize the basic steps of the method . The method itself is unsupervised , but four parameters need to be provided . The four parameters are: the threshold for finding events ( parameter tISI ) ; the temporal resolution for which two spike times are considered similar ( parameter: q ) ; and the number of patterns the clustering algorithm looks for ( parameter Nc , which stands for the number of clusters ) ; and a threshold tROC which determines which events need to be merged because they are common to multiple spike patterns . For the parameter settings used here , each cluster corresponds to a spike pattern , hence we will use these designations interchangeably . The method has been tested on short temporal segments with on average approximately 2 or 3 spikes per trial . These segments can be found by cutting spike trains at times with a low or zero spike rate in the spike time histogram . In the following text , we assume that the data has been divided in segments and discuss the analysis on one segment . The basic premise is that two trials on which the same pattern was produced are more similar to each other than two trials on which different patterns were produced . We used the VP distance [105] to quantify the similarity . The distance between spike trains i and j is represented as a matrix dij . The distance matrix appeared typically unstructured when the trials were arranged in the order in which they were recorded . We seek to re-order the matrix so that it becomes block-diagonal . The block diagonals correspond to trials that have a mutually small distance between all pairs in the block and are more distant to trials outside the block . That is , blocks on the diagonal correspond to spike patterns . This goal is achieved using the fuzzy c-means ( FCM ) method [106] applied to the columns of the distance matrix . FCM finds Nc clusters and assigns to each trial a probability of belonging to a cluster . If the clustering is ‘good’ each trial belongs to only one cluster with a high probability , it is ‘poor’ if a trial has similar probabilities of belonging to two or more different clusters . We have developed a heuristic to find appropriate values for Nc based on the gap-statistic [107] , [108] . Once patterns have been uncovered , a preliminary event structure is determined using the interval method outlined in [64] on each pattern . The interval method operates on the aggregate spike train , comprised of the time-ordered set of all spikes across all trials . The interval method is based on the principle that in the aggregate spike train the distance between spikes within an event is less than the distance between spikes in different events . The common events that occur in multiple spike patterns were found and merged based on a receiver operating characteristic ( ROC ) analysis [109] , which quantified how distinguishable the spikes in the two events were . After this analysis a cluster-assisted event structure is available . The result is illustrated in Figure 3B: each trial is assigned to a pattern on each segment , and each spike is assigned to an event . See [34] for additional examples . The outcome of the pattern-finding ( clustering ) procedure is that each spike train is assigned to a cluster . Formally , a set of trials has a classification ci , where i is the trial index between 1 and Ntrial; and the classification c is a number between 1 and the number of clusters Nc . The class distribution pj is the fraction of trials that were classified as class j , ( 1 ) where denotes the Kronecker delta . The diversity of the classification was characterized by the entropy ( 2 ) where the sum was over all nonzero pj values because 0 log 0 was defined to be zero [110] . The entropy S was zero ( minimal diversity ) when all trials were assigned to the same class and was maximal at S = log2Nc when all classes had the same probability of occurring ( maximal diversity ) . It is well known that the estimation of entropy from histogram data is biased [111] , [112] . We obtained approximate estimates for the bias and variance of the entropy estimates using a resampling procedure . Using the probabilities estimated from a finite number of trials as the exact probabilities we generated a thousand data sets from this probability distribution with the same number of trials and determined the entropy for each data set . The bias was the difference between the mean across the resampled entropy values and the original entropy , the variance was the standard deviation across the resampled entropy values . The mutual information was used to measure the similarity between two classifications . The joint distribution between two classifications ci and dj with Nc and Nd classes , respectively , was computed as: ( 3 ) The mutual information was then expressed as ( 4 ) ( 5 ) In the above formulas , the class distributions for c and d have a different subscript in order to distinguish them , hence i also is a class index rather than a trial index . To obtain a measure between 0 and 1 we normalized the mutual information by the maximum entropy , yielding the normalized mutual information: ( 6 ) The neuron was modeled as a single compartment with Hodgkin-Huxley-type voltage-gated sodium and potassium currents and a passive leak current [47] , [113] . The equation for the membrane potential of the model neuron is: ( 7 ) where is the leak current , is the sodium current , is the potassium current , Iinj is the injected fluctuating current , which is the same on each trial and is a noise current that is different on each trial . The values for the maximum conductance and reversal potential are listed in Table 1 . The gating variables are m , n , and h and they satisfy the equation ( 8 ) Here the label x stands for the kinetic variable , and ζ = 5 is a dimensionless time scale that can be used to tune the temperature dependent speed with which the channels open or close . The rate constants are: ( 9 ) and the asymptotic values of the gating variables are: ( 10 ) where x stands for m , n , and h . We made the approximation that m follows the asymptotic value m∞ ( V ) instantaneously [47] . The noise ξi in the current of neuron i is chosen such that 〈ξi ( t ) 〉 = 0 and 〈ξi ( t ) ξj ( t′ ) 〉 = 2λδ ( t−t′ ) δij . On each integration time step , the noise was drawn independently from a uniform distribution between −12λ/dt and 12λ/dt , where dt was the time step . The random noise value was treated as a constant for the duration of the time step . For Figure 5 we used λ = 0 . 00025 mV2/ms , whereas in Figure 6 we used λ = 0 . 0001 mV2/ms ( low noise ) and λ = 0 . 025 mV2/ms ( medium noise ) . For Iinj we either used the same 1050 ms long fluctuating drive as in experiment ( Figure 5 , amplitude between 0 and 1 , offset 0 . 2 ) , but without the constant depolarizing current pulses preceding the fluctuating drive in vitro; or we used a sinusoidal drive with time-varying frequency ( illustrated in Figure 6 ) . For Figures S2 and S3 we performed simulations with an additional potassium channel added to the WB model . The gating variable was denoted by z and satisfied the following equation with and when and for higher membrane potential values . This resulted in an additional current on the right-hand side of Eq . ( 7 ) , with gslow = 0 . 5 mS/cm2 . Additional parameters , Figure S2: offset 0 . 6 , amplitude 0 . 4 , λ = 0 mV2/ms; Figure S3A: as 2 , initial z-value as shown on the y-axis; Figure S3B: as 2 , offset as shown on the y-axis; Figure S3C offset as shown on the y-axis and λ = 0 . 01 mV2/ms . The initial values of the membrane potential at the beginning of the simulation were set to a fixed value , usually −70 mV . The gating variables were set to their asymptotic stationary values , , corresponding to the starting value , V , of the membrane potential . The differential equations were integrated using a second-order Runge-Kutta method with a time step of dt = 0 . 05 ms [114] , [115] . | Neurons respond with precise spike times to fluctuating current injections , leading to peaks in ensemble firing rate . The structure of these peaks , or spike events , provides a compact description of the neural response . We explore the consequences of precise spike times for neural coding in vivo , by investigating the spike event structure of virtual cell assemblies constructed in vitro . We incorporate diversity of electrophysiological response properties by varying the amplitude and offset of a common stimulus waveform injected in vitro . Across multiple trials , spike trains produce precise events in response to upswings in the stimulus , suggesting that such upswings in in vivo assemblies may effectively drive postsynaptic targets and transmit information about the stimulus . In simulations and in vitro , we identified bifurcations in the event structure as the amplitude or the offset was varied . Near bifurcation points , the neural response showed heightened sensitivity to intrinsic neural noise , leading to multiple competing response patterns , and enriching the representation of stimulus features by the ensemble output . The presence of bifurcations could therefore allow an ideal observer to extract more information about the stimulus . Our results suggest that event structure bifurcations may provide a mechanism for boosting information transmission in vivo . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"information",
"science",
"social",
"and",
"behavioral",
"sciences",
"neural",
"networks",
"neuroscience",
"mathematics",
"computational",
"neuroscience",
"biophysics",
"simulations",
"biophysics",
"theory",
"biology",
"nonlinear",
"dynamics",
"biophysics",
"physics",
"cent... | 2012 | Multiple Spike Time Patterns Occur at Bifurcation Points of Membrane Potential Dynamics |
In Neurospora crassa , the transcription factor COL-26 functions as a regulator of glucose signaling and metabolism . Its loss leads to resistance to carbon catabolite repression . Here , we report that COL-26 is necessary for the expression of amylolytic genes in N . crassa and is required for the utilization of maltose and starch . Additionally , the Δcol-26 mutant shows growth defects on preferred carbon sources , such as glucose , an effect that was alleviated if glutamine replaced ammonium as the primary nitrogen source . This rescue did not occur when maltose was used as a sole carbon source . Transcriptome and metabolic analyses of the Δcol-26 mutant relative to its wild type parental strain revealed that amino acid and nitrogen metabolism , the TCA cycle and GABA shunt were adversely affected . Phylogenetic analysis showed a single col-26 homolog in Sordariales , Ophilostomatales , and the Magnaporthales , but an expanded number of col-26 homologs in other filamentous fungal species . Deletion of the closest homolog of col-26 in Trichoderma reesei , bglR , resulted in a mutant with similar preferred carbon source growth deficiency , and which was alleviated if glutamine was the sole nitrogen source , suggesting conservation of COL-26 and BglR function . Our finding provides novel insight into the role of COL-26 for utilization of starch and in integrating carbon and nitrogen metabolism for balanced metabolic activities for optimal carbon and nitrogen distribution .
Filamentous fungi are one of the primary degraders of plant biomass because of their ability to produce enzymes that break down complex polysaccharides , including cellulose , hemicellulose , and pectin in the plant cell wall and starch , which is the major storage component in plants . Starch consists of two types of polysaccharides , amylose and amylopectin . Amylose is composed of linear chains of α-1 , 4-linked glucose units , while amylopectin is composed of α-1 , 4-linked glucose polymers , with branched α-1 , 4-glucan connected through α-1 , 6 glucosidic bonds at branch points . Our understanding of starch degradation by filamentous fungi mainly comes from work in Aspergillus spp . ( reviewed in [1] ) , which are industrially important producers of starch-degrading enzymes . Three types of enzymes , α-amylases , glucoamylases , and α-glucosidases , hydrolyze starch synergistically to produce glucose . α-Amylases hydrolyze α-1 , 4-glucan chains endolytically to produce maltose , while α -glucosidases and glucoamylases hydrolyze maltose and α -1 , 4-linkage exolytically from non-reducing ends to form glucose . Glucoamylases also hydrolyze α -1 , 6 linkages at branch connections . Recently , a new family of lytic polysaccharide monooxygenases ( LPMO ) active on starch was identified in Neurospora crassa [2] . The starch-active LPMOs together with a biological redox partner oxidatively cleave amylose , amylopectin , and starch . The expression of genes encoding amylolytic enzymes can be induced by starch and its degradation products , maltose and glucose [3–5] . In Aspergillus spp . , expression of genes encoding amylolytic enzymes requires the transcriptional activator AmyR , a zinc binuclear cluster ( Zn ( II ) 2Cys6 ) DNA-binding protein belonging to the Gal4p family of transcription factors [6] . Disruption of amyR in A . oryzae and A . nidulans leads to significantly decreased amylolytic enzyme activities and restricted growth on starch medium [7 , 8] . A similar role in starch hydrolysis was demonstrated for amyR homologs in Penicillium decumbens [9] , Fusarium verticillioides and F . graminearum [10] . Genome sequencing of two Trichoderma reesei mutant strains , RUT C30 and PC-3-7 , with enhanced cellulase production and resistance to carbon catabolite repression ( CCR ) identified SNPs in the bglR gene , a homolog of amyR [11 , 12] . Although a T . reesei strain bearing a deletion of bglR was reported having reduced growth on maltose and glucose , further investigation on the phenotype of the ΔbglR mutant was not reported . Instead , Nitta et al . ( 2012 ) suggested that BglR regulates genes encoding β-glucosidases and belongs to a new functional transcription factor group distinguishable from AmyR based on two observations [11] . First , when induced by cellobiose , expression of some β-glucosidase genes was lower in the ΔbglR mutant as compared to the parental PC-3-7 strain . Second , AmyR and BglR form two separate clusters in phylogenetic analyses . However , the AmyR homologs in F . graminearum and F . verticillioides ( FgART and FvART , respectively ) are in the same cluster as BglR and are essential for starch utilization [10] . COL-26 is the N . crassa ortholog of BglR and was named colonial-26 ( col-26 ) for its colonial phenotype on medium containing sucrose , glucose or fructose as a sole carbon source [13 , 14] , suggesting COL-26 plays a role in regulating glucose metabolism . In N . crassa , COL-26 was shown to function synergistically with CRE-1 , a transcription factor important for CCR and in regulating cellulase gene expression and enzyme production [14] . The Δcol-26 mutant is also resistant to 2-deoxyglucose , suggesting it has impaired CCR . In this study , we tested growth phenotypes of the Δcol-26 mutant on a variety of carbon sources and determined that COL-26 is essential for maltose and starch utilization . We determined that the absence of col-26 led to a decrease in expression of amylolytic genes . Metabolic analyses of the Δcol-26 mutant in comparison to WT cells indicated that mis-regulation of the TCA cycle , GABA shunt , and amino acid biosynthesis occurs in the Δcol-26 mutant . Replacing ammonium as a nitrogen source on preferred carbon sources with glutamine alleviated the growth defects of Δcol-26 on glucose , but not on maltose medium . Our study indicates that COL-26 has an important and conserved role in the regulation of starch degradation as coordinating primary carbon and nitrogen metabolism in filamentous fungi , and provides insight for the rational design of strains for the food and biofuel industries .
The Δcol-26 mutant poorly utilizes simple sugars , including glucose , fructose , and sucrose , but grows well on complex polysaccharides such as cellulose [14] . To test whether COL-26 is important for the utilization of other carbon sources , we tested the growth of the Δcol-26 mutant on different mono- , di- or polysaccharides as a sole carbon source . As observed previously , the Δcol-26 mutant showed reduced growth in glucose , fructose and sucrose [14] , but also showed reduced growth on xylose and cellobiose and essentially no growth on maltose or trehalose ( Fig 1A ) . On complex polysaccharides , such as xylodextrins and albumin , the Δcol-26 mutant grew similarly to the WT parental strain . However , the Δcol-26 mutant showed a severe growth defect on amylopectin ( Fig 1A ) . To verify that col-26 is causative for these growth phenotypes , we introduced a copy of the col-26 gene under regulation of the A . nidulans gpd promoter at the csr-1 locus in the Δcol-26 mutant ( see Materials and methods ) . This Pgpd-col-26; Δcol-26 strain showed a similar growth phenotype as the WT strain on these different carbon sources ( Fig 1A ) . Consistent with the hypothesis that COL-26 plays a role in regulating genes encoding enzymes required for utilization of starch , trehalose and maltose , the expression level of col-26 was induced 4 to 8 -fold by a 4-hr exposure to trehalose , maltose , amylopectin and amylose ( Fig 1C ) . A genetic interaction between cre-1 and col-26 was revealed in the regulation of cellulase production; increased expression levels of cre-1 was observed in the Δcol-26 mutant [14] . This observation suggested that mis-regulation of cre-1 ( and thus inappropriate triggering of CCR ) may play a role in the growth phenotype of the Δcol-26 mutant . To test this hypothesis , we examined the growth phenotype of the Δcre-1; Δcol-26 double mutant as compared to the WT strain and the Δcol-26 and Δcre-1 single mutants on a variety of carbon sources . The Δcre-1; Δcol-26 mutant grew similarly to the Δcol-26 mutant when glucose , xylose , sucrose , cellobiose , maltose , trehalose , or amylopectin was used as the sole carbon source ( Fig 1B ) , indicating that the mis-regulation of cre-1 expression was not causative for the poor growth phenotype observed in the Δcol-26 mutant . Neurospora has long been known to be a starch utilizer ever since its discovery over 170 years ago on contaminated bread in a French Bakery [15] . Although mutants deficient for the utilization of starch ( sor-4 , gla-1 and gla-2 ) have been identified [16] , how N . crassa transcriptionally responds to starch in its environment has not been previously investigated . To provide systematic data on expression changes in response to defined polysaccharide constituents of starch , we performed transcriptional profiling of WT cells exposed to Vogel’s minimal medium ( VMM ) [17] containing amylose or amylopectin as the sole carbon source ( 1% w/v ) and WT cells exposed to VMM containing maltose as the sole carbon source ( 2% w/v ) . RNA-seq data from N . crassa cultures exposed to VMM with no carbon ( NC ) or VMM with 2% ( w/v ) sucrose were included as controls . The fifteen sets of RNA-seq data were first evaluated using principle component analysis ( PCA ) . Biological replicate samples from the same carbon condition clustered tightly ( Fig 2A ) . Expression patterns from cultures exposed to amylose and amylopectin also clustered closer to each other than to the NC , maltose and sucrose samples , suggesting a common transcriptional response in N . crassa upon exposure to polysaccharides of starch . Additionally , expression patterns from cultures exposed to maltose were distant from those exposed to sucrose in the PCA plot , suggesting substantial transcriptional changes specifically induced by maltose . Pairwise comparison between the transcriptome of WT cells exposed to amylose or to amylopectin compared to that of VMM-NC revealed genes with differential expression levels ( fold change greater than 2 , and false discovery rate ( FDR ) -corrected p value below 0 . 01 ) . After subtracting genes that were also differentially induced or repressed in sucrose as compared to VMM-NC , we identified 322 genes that increased in expression level in WT cells upon exposure to amylose/amylopectin and 108 genes that showed reduced expression levels in amylose/amylopectin ( Fig 2B; S1 Table , Sheet 1 ) . We name this 322-gene set the “starch regulon” . Indeed , the only overrepresented KEGG pathway in this set of 322 genes was starch and sucrose metabolism ( adjusted p-value: 4 . 8e-3 ) . No KEGG pathway was overrepresented in the 108 reduced expression gene set . Analyses of RNA-seq data from WT on maltose revealed a set of 1871 genes that increased in expression level and 1881 genes that decreased in expression level compared to data from WT on NC . After subtracting genes that were similarly regulated by sucrose , we identified 736 genes with increased expression level and 696 genes with decreased expression level in WT cells on maltose medium ( Fig 2B; S1 Table , Sheet 2 ) . The maltose-inducible gene set was enriched in genes from functional categories of biogenesis of cell wall , perception of nutrient and nutritional adaptation , and electron transport and membrane-associated energy conservation . Additionally , the maltose-inducible gene set overlapped the starch regulon by 111 genes ( S1 Table , Sheet 3 ) . A search in the Carbohydrate Active Enzymes ( CAZyme ) database ( http://www . cazy . org/ ) [18] revealed that 7 of the 111 genes were predicted to act on carbohydrates . Three of them , NCU04674 ( gh31-3 ) , NCU01517 ( gla-1 ) , and NCU08746 have annotated functions in degrading starch . gh31-3 encodes a α-glucosidase , gla-1 encodes a glucoamylase and NCU08746 encodes a lytic polysaccharide monoxygenase that acts on starch [2] ( Fig 2C ) . A BLASTP search of the transporter classification databases ( TCDB ) ( http://www . tcdb . org/ ) with cut-off value less than 1e-20 identified 14 genes likely encoding transporters . Four of them , hgt-1 ( NCU10021 ) , NCU05627 , NCU04963 , and NCU04537 are annotated as sugar transporter genes ( Fig 2C ) . hgt-1 shows high affinity glucose transport activity [19] , while NCU05627 ( xyt-1 ) has xylose transporting activity [20] . The transport substrates of NCU04963 and NCU04537 remain to be determined . There are also 5 TF genes induced by all three starch components , tah-3 , vad-2 , kal-1 , hac-1 , and NCU03975 ( Fig 2C ) . tah-3 was found to be required for tolerance to a harsh plasma environment [21] . For VAD-2 and KAL-1 , a role in nutrient metabolism or sensing has been proposed [13] . HAC-1 is involved in the unfolded protein response and is necessary for growth on cellulose , but not hemicellulose in N . crassa [22] . Genes in the starch-regulon , but that were not in the maltose-inducible gene set , included gh13-6 , tre-1 , and clr-2 ( Fig 2D ) . gh13-6 encodes an α-amylase , tre-1 encodes a trehalase , and CLR-2 is the major transcriptional regulator of cellulase genes in N . crassa and is essential for the utilization of cellulose [23–25] . Among genes significantly induced by maltose , but not by starch polysaccharides were NCU00801 ( cdt-1 ) and NCU12154 ( Fig 2D ) . CDT-1 is a cellodextrin transporter and NCU12154 was annotated as maltose permease . The latter shows low homology to the yeast maltose permease ( P53048; TCDB database ) . Interestingly , the α-amylase gene ( gh13-6 ) in the starch regulon was not induced by maltose ( Fig 2D ) . Instead , two other α-amylase genes ( NCU09805 gh13-1 and NCU08131 gh13-2 ) were significantly induced ( Fig 2D ) . The Δcol-26 mutant failed to grow on maltose and amylopectin ( Fig 1A ) . To investigate the functions of COL-26 required for utilization of these substrates , we evaluated transcriptional changes in the Δcol-26 mutant when switched to medium containing amylose or maltose under identical conditions as with the WT parental strain ( see above ) . RNA-seq data from the WT and Δcol-26 biological replicates were subjected to PCA analysis and data from the same strain grown under the same growth conditions clustered together ( Fig 3A ) . On the PCA plot , data from the Δcol-26 mutant exposed to amylose and data from the Δcol-26 mutant exposed to maltose did not cluster . Under amylose conditions , the expression level of 1242 genes was significantly lower in the Δcol-26 mutant , while the expression level of 1124 genes increased ( Fig 3B , S2 Table ) . Strikingly , 252 genes out of the 322-gene starch regulon gene set ( 78% ) were down regulated in the Δcol-26 mutant ( Fig 3B ) , including three of the four amylolytic genes , gla-1 , the starch-active LPMO ( NCU08746 ) , gh13-6 and 19 transporter genes including hgt-1 , xyt-1 , NCU04963 , and NCU04537 ( Fig 3C ) . Seventeen TF genes in the starch regulon were also down regulated in the Δcol-26 mutant , including tah-1 , tah-3 , vad-2 , ada-5 , and kal-1 ( S2 Table , sheet 1 ) . The majority of the remaining 70 starch-regulon genes ( 41 genes ) whose expression levels were not affected by the col-26 deletion were annotated as hypothetical . These data indicate that COL-26 is a major regulator of the starch regulon of N . crassa . The down regulation of expression of the starch regulon genes by deletion of col-26 is consistent with the growth defect observed in the Δcol-26 mutant on starch polysaccharides ( Fig 1A ) . Under maltose conditions , the expression levels of 1110 genes were significantly increased in Δcol-26 mutant , while the expression levels of 988 genes decreased ( Fig 3B and S2 Table , sheet 2 ) . The three amylolytic genes , i . e . , gla-1 , gh13-2 , and the starch LPMO ( NCU08746 ) and the four sugar transporter genes , hgt-1 , xyt-1 , NCU04963 , and NCU04537 were members of the down-regulated gene set ( Fig 3C ) . This down-regulated gene set also included additional 100 transporter genes , many from the Mitochondrial Protein Translocase ( MPT ) family , the Nuclear mRNA Exporter ( mRNA-E ) family , the Mitochondrial Carrier ( MC ) family , and the Major Facilitator Superfamily ( MFS ) ( S2 Table ) . The down-regulated 988-gene set in the Δcol-26 mutant only overlapped the 736 maltose-inducible set in WT cells by 63 genes ( Fig 3B ) . Directly comparing the amylose and maltose RNA-seq data between the WT and the Δcol-26 mutant showed that 363 genes were down regulated and 421 genes were up regulated in absence of col-26 . We named the 363 genes as the “COL-26-dependent gene set” and the 421 genes as the “COL-26-reduced expression gene set” ( S3 Table , Fig 3D ) . Only 33 genes in the COL-26-dependent gene set ( less than 10% ) were induced in WT cells by exposure to maltose , amylose , or amylopectin . For the COL-26-dependent gene set , a functional enrichment analysis using FunCat [26] showed that transcription and protein synthesis were overrepresented , including rRNA processing , where 79 of 198 genes in this category were identified as being COL-26 dependent ( p = 9e-52 ) . Genes from categories such as RNA binding functions , ribosome biogenesis , rRNA modification , mRNA synthesis and mitochondrial transport were also enriched ( p = 3e-17 , 2e-13 , 3e-9 , 2e-7 , and 1e-2 respectively ) . The COL-26-dependent gene set contained 6 TF genes besides col-26 . Three were annotated to be hypothetical , and the other three were vib-1 ( NCU03725 ) , nit-2 ( NCU09068 ) , and cpc-1 ( NCU04050 ) ( Fig 3E ) . VIB-1 ( vegetative incompatibility block-1 ) is required for extracellular protease secretion in response to both carbon and nitrogen starvation [27] and for the utilization of cellulose [14] . The cpc-1 gene ( cross-pathway control-1 ) is the ortholog of S . cerevisiae GCN4 , and is required in N . crassa for the expression of many amino acid biosynthetic genes in response to amino acid starvation [28–30] . Ten genes in CPC-1 regulon [30] were also found in the COL-26-dependent gene set ( S3 Table ) . The nit-2 gene ( nitrate nonutilizer-2 ) is the major regulatory transcription factor in N . crassa regulating nitrogen catabolism and is critical for utilization of nitrate , nitrite , purines , and most amino acids as a nitrogen source ( reviewed in [31] ) . Also in this set were genes encoding catabolic enzymes in nitrogen metabolism and amino acid synthesis such as am , which encodes the NADP-glutamate dehydrogenase ( NADP-GDH ) , gln-1 ( NCU06724 ) and gln-2 ( NCU04856 ) , both of which encode glutamine synthases . Several transporters in this set are also predicted to be involved in nitrogen and amino acid assimilation , including uc-5 ( NCU07334 ) , mtr ( NCU06619 ) , pmb ( NCU05168 ) and tam-1 ( NCU03257 ) . uc-5 encodes a uracil permease [32] . The mtr mutant is defective in transport of neutral aliphatic and aromatic amino acids . The pmb mutant is defective in basic L-amino acid transport and has reduced uptake of L-arginine , L-lysine , and L-histidine [16] and tam-1 encodes a predicted ammonium transporter . We also compared the COL-26-dependent gene set to the set of genes that showed reduced expression levels in WT cells in carbon-free medium as compared to WT on maltose or on amylose to reflect the effects of carbon starvation under these two conditions . This comparison revealed that 291 of the 363 COL-26-dependent genes also showed reduced expression in WT cells when no carbon source was available ( S3 Table ) , including vib-1 , nit-2 , uc-5 , mtr , pmb , am , and 5 of the 10 CPC-1 regulon genes . However , cpc-1 , gln-1 and gln-2 were not among these 291 genes . The COL-26-reduced expression gene set was enriched with genes in the functional categories of non-vesicular cellular import ( p = 7e-8 ) , secondary metabolism ( p = 4e-6 ) , degradation or biosynthesis of phenylalanine ( p = 2e-5 ) , allantoin and allantoate transport ( p = 2e-7 ) , polysaccharide metabolism ( p = 2e-6 ) , and C-compound and carbohydrate transport ( p = 9e-6 ) . This gene set also included 39 transporter genes , 7 TF genes , and 26 CAZyme genes ( S3 Table ) . All predicted TF genes in this set have no assigned function . The majority of the transporter genes ( 25 of 39 ) belong to the MFS family , but only two , cdt-2 and cbt-1 ( NCU08114 and NCU05853 ) have been characterized ( Fig 3E ) . CDT-2 transports cellodextrins and xylodextrins [33–35] , while CBT-1 has transporting activity for cellobionic acid [36 , 37] . The 26 CAZymes are from 21 CAZyme families , and two of them , a cellulose LPMO gene pmo-3 ( NCU07898 ) and a cellobiose dehydrogenase gene cdh-2 ( NCU05923 ) have been characterized in N . crassa [38 , 39] . Of these 421 genes , 195 were also de-repressed in WT cells under carbon-free conditions relative to WT on maltose or amylose . COL-26 was essential for the utilization of starch components and was essential for expression of a large fraction of genes associated with utilization of starch in WT cells ( Figs 1 and 3B ) . However , its growth defect on preferred carbon sources was unique , as other mutants , such as Δclr-1 and Δclr-2 are unable to grow on cellulose , but have WT growth rates on preferred carbon sources [23] . Our observation of the down regulation of tam-1 , am , gln-1 and gln-2 genes ( Fig 3E ) in the Δcol-26 mutant and the fact that loss-of-function of glutamine synthase renders N . crassa dependent upon glutamine for normal growth [40] prompted us to test whether the reduced growth of the Δcol-26 mutant in VMM-glucose was due to impaired nitrogen metabolism . To test this hypothesis , we examined growth of the Δcol-26 mutant in VMM where ammonium nitrate was replaced by glutamine , VMM ( Gln ) . Bird’s minimal medium ( BMM ) [41] was also used for assessing the effect of glutamine substitution , BMM ( Gln ) ; BMM ( NH4Cl ) has ammonium chloride as the sole nitrogen source . The carbon source for both VMM ( NH4NO3 ) and BMM ( NH4Cl ) was glucose; the Δcol-26 mutant showed decreased growth on VMM ( NH4NO3 ) -glucose ( Fig 1A ) . Mycelial biomass from 24-hr cultures of the WT and Δcol-26 strains , as well as a glutamine synthetase mutant ( Δgln-1 ) [16] were compared ( Fig 4A ) . In VMM ( NH4NO3 ) with 2% ( w/v ) glucose as the carbon source , both Δcol-26 and Δgln-1 grew poorly , reaching only 11~ 12% of WT biomass . Both mutants grew much better in VMM ( Gln ) , with Δcol-26 and Δgln-1 reaching 47% and 72% of WT biomass , respectively ( Fig 4A ) . Similar rescue effects of glutamine were observed in the Δcol-26 and Δgln-1 mutants in BMM ( Gln ) as compared to that in BMM ( NH4Cl ) ( Fig 4A ) . Substitution of glutamine for ammonia also partially rescued the growth of the Δgln-1 mutant on maltose or amylopectin , but failed to rescue the growth of the Δcol-26 mutant under the same conditions ( Fig 4B ) . The distinctive rescue effect of glutamine in the Δcol-26 mutant on medium with glucose versus medium with maltose argues for glutamine being used as a nitrogen source rather than a carbon source . To test this hypothesis , WT , and the Δcol-26 , Pgpd-col-26 and Δgln-1 strains were grown on VMM with either glutamine , glutamate , arginine or proline as both the carbon and nitrogen source or as the carbon source with ammonium nitrate as the nitrogen source ( S1 Fig ) ; none of the strains efficiently utilized these amino acids as a carbon source , with only minimal mycelial biomass observed after 9 days of growth ( S1 Fig ) . The Δcol-26 and Δcre-1 mutants are resistant to 2-deoxyglucose ( 2-DG ) , a glucose analog that cannot be metabolized but is able to trigger CCR [14] . It is often used to select for , or evaluate , impairment of CCR or glucose repression in filamentous fungi [42 , 43] . As Δcol-26 grows extremely slowly in VMM ( NH4NO3 ) with glucose , we hypothesized that its insensitivity to 2-DG may be a result of its mis-regulation of carbon/nitrogen metabolism . Since the growth of the Δcol-26 mutant in VMM ( Gln ) was enhanced , we evaluated whether this restoration in growth rescued the sensitivity to 2-DG in the Δcol-26 mutant . We used cellobiose as the sole carbon source , as the Δcol-26 mutant grows better under these conditions ( Fig 1A ) . The WT , Δcre-1 , Δcol-26 , and Δgln-1 strains were grown in VMM-2% ( w/v ) cellobiose with or without 0 . 2% ( w/v ) 2-DG , with either NH4NO3 or glutamine as the nitrogen source . The Δcol-26 mutant showed a clear resistance to 2-DG , independently of whether NH4NO3 or glutamine was used as the nitrogen source ( Fig 4C ) . These data indicate that Δcol-26 resistance to 2-DG inhibition ( and thus impaired COL-26-mediated CCR ) remains independent of the nitrogen source . To further understand the changes in primary carbon , nitrogen , and amino acid metabolism in the Δcol-26 mutant relative to the WT strain , we profiled 45 intracellular metabolites from WT , Δcol-26 and Δgln-1 strains grown on VMM ( NH4NO3 ) or VMM ( Gln ) with glucose as the carbon source ( S4 Table ) . Strains were first grown in VMM ( NH4NO3 ) -cellobiose ( 2% w/v ) to accumulate fungal biomass , then switched to VMM ( NH4NO3 ) -NC for 18 hrs and subsequently grown in VMM ( NH4NO3 ) or VMM ( Gln ) with glucose ( 2% w/v ) for an additional 5 . 5 hrs . Intracellular metabolites were extracted and subjected to analyses using gas chromatography coupled to mass spectrometry ( GC-MS ) augmented with liquid chromatography coupled to tandem mass spectrometry ( LC-MS/MS ) ; normalized abundances of metabolites were compared between WT and the mutants . Relative quantitative analysis showed that 17 metabolites were significantly different between WT and the mutants ( p < 0 . 05 ) ( Fig 5A ) . However , surprisingly little similarity in metabolite profile was observed when the Δgln-1 and Δcol-26 mutants were compared . In the Δgln-1 mutant , glutamine levels were significantly lower than WT when grown in VMM ( NH4NO3 ) , but the intracellular glutamine deficiency was rescued by growth on VMM ( Gln ) media ( Fig 5A ) . However , although the growth defect of the Δcol-26 mutant on glucose was partially alleviated when grown on VMM ( Gln ) media , the intracellular glutamine levels in the mutant were similar to that of WT grown on either VMM ( NH4NO3 ) or VMM ( Gln ) ( Fig 5A ) . Instead , the Δcol-26 mutant accumulated high levels of four metabolites under both conditions: 4-aminobutanoic acid ( GABA ) , phenylalanine , cysteine and succinate and was deficient in three metabolites: valine , threonine and succinic semialdehyde . Only the high level of GABA and the low level of lysine were shared phenotypes between the Δcol-26 and the Δgln-1 mutant ( Fig 5A ) . GABA and succinic semialdehyde are two intermediate metabolites in the GABA shunt , a metabolic pathway that bypasses two enzymatic steps of the TCA cycle to produce succinate from α-ketoglutarate via glutamate ( Fig 5B ) . As the GABA shunt links primary nitrogen and carbon metabolism , the abnormal level of these intermediates suggests a mis-regulation of primary carbon and nitrogen metabolism occurs in the Δcol-26 mutant . The levels of several amino acids showed a difference between the Δcol-26 mutant and WT grown on VMM ( NH4NO3 ) , including homoserine , valine , lysine and threonine . In filamentous ascomycete fungi , COL-26 , ART , and AmyR are conserved in their functions in regulating starch degradation [7 , 8 , 10] ( this study ) . We further demonstrated critical functions of COL-26 in integrating nitrogen and carbon metabolism , a role not previously reported for AmyR/BglR/ART orthologs in other fungi . Although phylogenetic analyses have been performed to infer functional conservation of these homologs [10 , 11] , either a single homolog per fungal genome was chosen or homologs from very few model organisms were included in the analyses . Our search for col-26 homologs in 44 fungal species within the Ascomycota using BLASTP with cut-off E value of e-20 revealed that many fungi have more than one predicted col-26 homolog and that the number of col-26 homologs varies within each species ( S5 Table ) . For example , some Fusarium species and Trichoderma species have 5 or 6 col-26 homologs , while other species such as Metarhizium spp . , Verticillium spp . , Myceliophthora thermophile , Thielavia terrestris , Chaetomium globosum , Cordyceps militaris , and Beauveria bassiana , each have only one homolog of col-26 . Three Aspergillus spp . have 3 col-26 homologs , including amyR , but amyR from both A . oryzae and A . nidulans was not the best hit by col-26 . In order to gain a broader view regarding functional conservation of the col-26 homologs , we constructed a phylogenetic tree of the 86 COL-26 protein sequences using a Maximum Likelihood algorithm . CLR-2 ( NCU08114; also identified as Neucr2 6271 in Mycocosm ) was used as outgroup to root the tree ( Fig 6 ) . Although two COL-26 homologs exist in T . reesei ( BglR/Trire2 52368 and Trire2 55109 ) , only BglR was within the same clade as COL-26 . Similarly , although F . graminearum and F . verticillioides possess 5 and 6 homologs of COL-26 respectively , only FgART and FvART were in the same clade as COL-26 and BglR . The genome of Magnaporthe oryzae ( also called Magnaporthe grisea ) has a single COL-26 homolog , named MoCOD1 [44] . Interestingly , the ΔMocod1 mutant showed significant growth reduction on glucose and maltose-containing medium but not on starch-containing medium , while the ΔFgART1 mutant displayed a severe growth defect on glucose and starch-containing medium , but not on maltose-containing medium [10 , 44] . AmyR from A . oryzae , A . niger , and A . nidulans together with two COL-26 homologs from A . flavus and A . terreus , respectively , form a clade distant from the COL-26 clade , while a MalR ( AO90038000235 ) from A . oryzae and two homologs from A . flavus and A . nidulans , respectively , are in a clade more closely aligned to the COL-26 clade . Although AmyR is reported to be required for growth on both starch and maltose in A . nidulans [8] , A . oryzae largely relies on MalR for growth on maltose [45] . Our phylogenetic tree indicated that BglR is the closest T . reesei homolog of COL-26 . Reduced growth of the ΔbglR mutant on maltose has been reported [11] . To test if BglR functions similarly to COL-26 , we replaced the endogenous bglR coding sequence in T . reesei with the pyr4 gene in a Δpyr4 auxotrophic mutant [46] . All PCR verified transformants grew slowly on MM with 2% glucose agar plates . Three independent ΔbglR mutants were selected for further assessment . We subsequently tested growth of the ΔbglR mutant in minimal medium with ammonium sulfate , MM ( ( NH4 ) 2SO4 ) as the sole nitrogen source with glucose , maltose , trehalose , amylose or amylopectin as the sole carbon source . In contrast to parental WT strain QM6a , almost no growth of the ΔbglR mutants in MM-glucose , MM-amylopectin or MM-trehalose was observed ( Fig 7A ) . Surprisingly , neither the parental QM6a strain nor the ΔbglR mutant grew in MM when maltose or amylose was used as the sole carbon source . To assess the influence of glutamine on glucose utilization in the ΔbglR mutant , we measured changes in glucose concentration in liquid cultures of QM6a and the ΔbglR mutant in MM ( ( NH4 ) 2SO4 ) or MM ( Gln ) with 2% glucose as the sole carbon source . This approach was chosen because T . reesei utilized glutamine as a carbon source more efficiently than N . crassa , which prevented an unambiguous conclusion about glucose consumption rate based on growth phenotypes ( S2 Fig ) . In MM ( ( NH4 ) 2SO4 ) , only 11% of the glucose in the medium was used after 2 days by the ΔbglR mutant versus a 90% reduction in glucose levels in the parental QM6a strain ( Fig 7B ) . In MM ( Gln ) , an increase of glucose consumption to 60% by the ΔbglR mutant was detected when glutamine was used as the nitrogen source , while QM6a showed similar glucose consumption on MM ( Gln ) as MM ( ( NH4 ) 2SO4 ) ( Fig 7B ) . These data suggest that , like COL-26 in N . crassa , BglR also plays a critical role in regulating starch degradation and primary carbon and nitrogen metabolism in T . reesei . Based on the phylogenetic analyses , such a multi-regulatory role may also be conserved in col-26 orthologs in many other filamentous fungal species .
In nature , filamentous fungi must integrate data from available carbon sources to coordinate with nitrogen , phosphorus and sulfur assimilation for optimal growth . How this coordination is achieved in these organisms is currently not clear , as most studies evaluate physiological/transcriptional differences based on comparison between single carbon or nitrogen sources . In this study , we identified a conserved regulator , COL-26 , that plays a role in coordinating the utilization of starch components with nitrogen regulation . By comparing the amylose and maltose RNA-seq data between the WT and the Δcol-26 mutant , we identified a 363-COL-26-dependent gene set . This gene set contained many genes with functions in primary nitrogen and amino acid metabolism , including the transcription factors vib-1 , nit-2 , and cpc-1 . A large percentage of these genes were also induced in WT when cells were exposed to maltose or amylose , indicating coordinate regulation of nitrogen metabolism with carbon metabolism . The down regulation of genes such as gln-1 and gln-2 in the Δcol-26 mutant and its unique phenotype of poor growth on preferred carbon sources led us to speculate that an inability to coordinate carbon and nitrogen metabolism may occur in the Δcol-26 mutant . This hypothesis was supported by the partial rescue of growth defects in the Δcol-26 mutant by the use of glutamine as the sole nitrogen source with glucose as the carbon source . As in N . crassa , growth defects on glucose medium were noted for a F . graminearum ΔFgART mutant [10] , a M . oryzae ΔMoCOD1 mutant [44] and a T . reesei ΔbglR mutant [11] . Here we provided experimental evidence that in T . reesei , BglR was essential for amylopectin and trehalose utilization and for ammonium assimilation on preferred carbon sources . Our data also showed that T . reesei cannot grow on amylose or maltose . These data indicate that , unlike N . crassa , T . reesei may rely on α-1 , 6 linkages for signaling for starch degradation . Whether these functions are all conserved by the COL-26 orthologs in other filamentous fungi awaits further verification . However , as many filamentous fungi are plant pathogens of starch crops , such as F . graminearum and M . oryzae , and deletion of the col-26 orthologs reduced pathogenicity in both these fungi , understanding the regulatory mechanisms by the COL-26 orthologs could shed light on the development of future anti-fungal strategies . The data from metabolic analyses showed that several metabolites in the TCA cycle and GABA shunt pathway were either at a higher or lower level in the Δcol-26 mutant as compared to WT cells . In particular , high levels of succinate and GABA persisted and succinic semialdehyde remained below detectable levels in the Δcol-26 mutant even when glutamine was provided as the nitrogen source and growth was partially restored . The GABA shunt pathway is a metabolic route conserved among bacteria , fungi , plant and vertebrates . The role of the GABA shunt has been extensively investigated in animals and plants due to GABA being a key neurotransmitter in the central and peripheral nervous system of vertebrates and a signal molecule in response to many biotic and abiotic stresses in plants [47] . The GABA shunt in fungi has received less attention , but has been associated with nitrogen metabolism , spore germination , asexual sporulation , redox homeostasis , acidogenic growth , response to hypoxia , oxidative stress and virulence [48–53] . Besides the GABA shunt , an alternative pathway exists for GABA catabolism in many eukaryotes including S . cerevisiae , through which the intermediate metabolite , succinic semialdehyde ( SSA ) , is reduced to γ-hydroxybutyric acid ( GHB ) [54] . In the Δcol-26 mutant , it is possible that mis-regulation of enzyme activities at either the transcriptional and ( or ) post-transcriptional level and/or defects in the transport of glutamate or GABA between cytoplasm and the mitochondria may occur . Whether the reduction of succinic semialdehyde or the other metabolites in the mutant is caused by increased activity of enzymes in one pathway versus a re-wiring the metabolite to other pathways warrants further investigation . Our metabolite data is consistent with a regulatory role of COL-26 in the GABA shunt and in coordinating primary carbon and nitrogen metabolism for optimal fungal growth . In addition to a role in the coordination of primary carbon and nitrogen metabolism , COL-26 is essential for the utilization of starch . In A . niger , transcriptional analyses via microarrays of carbon-limited chemostat or batch cultures growing on maltose versus xylose revealed that only three amylolytic genes aamA ( acid α-amylase ) , glaA ( glucoamylase ) , agdA ( α-glucosidase ) were induced by maltose [55 , 56] . In A . oryzae , ten genes annotated to encode glucoamylase , maltose permease , maltase , sugar transporters and maltose O-acetyltransferase were up regulated by maltose [57] . In this study , we performed systematic transcriptional profiling of N . crassa on different components of starch , including maltose , amylose , and amylopectin . From these analyses , we identified a starch regulon consisting of 322 genes; COL-26 is required for WT expression patterns of 252 of these 322 genes . Surprisingly , our data showed that expression changes in N . crassa in response to polysaccharides of starch differed substantially from those induced by maltose ( Fig 2A ) , where only ~1/3 of the starch regulon genes were induced ( Fig 2B ) . Such transcriptional differences may reflect changes in signaling or utilization strategies by Neurospora for optimal uptake of nutrients of different forms ( disaccharides versus polysaccharides , for example ) . The function of a COL-26 homolog in Aspergilli , AmyR , the transcriptional regulator associated with maltose and starch utilization in Aspergillus spp . , shows some divergence in function even among Aspergillus species . In A . oryzae , where maltose-utilizing ( MAL ) clusters are found , AmyR is important for starch degradation , but MalR is required for maltose utilization and AmyR activation [45] . In A . nidulans and A . niger , which lack MAL clusters , AmyR is critical for both maltose and starch utilization [5 , 8] . N . crassa does not have MAL clusters and no protein exhibits higher homology to MalR than COL-26 . Here , we demonstrated that COL-26 is essential for the utilization of maltose , amylopectin and amylose , all components of starch . Consistent with this essential role , the expression of col-26 increased in presence of amylose , amylopectin , and under a low concentration of maltose ( 2 mM ) , while deletion of col-26 led to decrease in expression level of 78% of the starch-regulon genes . Genes related to cellulose degradation were among the genes that increased in expression level in the Δcol-26 mutant . These included cellodextrin and cellobionic acid transporter genes , cdt-2 and cbt-1 , respectively and cellulase genes pmo-3 and cdh-2 . Substrates of CDT-2 and CBT-1 are in fact products from PMO-3 and CDH [38 , 39] . A screen for N . crassa hypersecretors of cellulases also identified a modest increase of cellulase production in the Δcol-26 mutant [46] . These data support the hypothesis that cellulose degradation by N . crassa is negatively regulated by a COL-26-mediated glucose repression , consistent with the robust 2-DG resistance in Δcol-26 mutant . In support of a conserved function of COL-26 , a Penicillin oxalicum ΔamyR mutant also showed decreased amylase activity and increased cellulase expression on cellulose [58] . These observations suggest an antagonizing effect between activation of amylolytic genes versus cellulase genes in filamentous fungi , which is mediated by COL-26/AmyR . In this study , although we focused on elucidating the essential roles of COL-26 in regulating starch degradation and primary carbon and nitrogen metabolism , we also demonstrated that COL-26 and BglR were essential for trehalose utilization . Trehalose is the major internal carbohydrate reserve in N . crassa and other fungi and trehalose mobilization occurs during germination of fungal spores , a process that can be enhanced by glucose combined with a nitrogen source [59] . The tre-1 gene , encoding trehalase , was within the starch regulon , but was not differentially expressed in the Δcol-26 mutant on starch components . Whether the inability to utilize trehalose is a consequence of the inability of the Δcol-26 mutant to efficiently utilize glucose ( cleavage of trehalose yields two glucose molecules ) is unclear . In the insect pathogen Metarhizium acridum , enhancing fungal utilization of trehalose , the main carbon source in insect hemolymph , has been shown to improve virulence [60] . Single col-26 orthologs occur in the genomes of the insect-pathogenic fungi Metarhizium acridum and Metarhizium robertsii . Further study of functions of the COL-26 orthologs in trehalose utilization in these fungi may aid in developing more potent strains for insect biocontrol . Finally , we identified a number of predicted transporter genes within the starch regulon , including hgt-1 , xyt-1 , NCU04963 , and NCU04537 , while cdt-1 and NCU12154 were significantly induced by maltose . NCU12154 has been annotated as maltose permease based on bioinformatics analyses , although biochemical evidence is lacking . It is possible that one of these uncharacterized transporters encode a maltooligosaccharide transporter that accompanies activity of intracellular α-amylase , which are part of the starch regulon . Testing transporting activity of the predicted transporters will aid in our understanding of diverse nutrient assimilation pathways by filamentous fungi .
N . crassa Δcol-26 ( FGSC 11031 ) and Δgln-1 ( FGSC 19959 ) were obtained from the Fungal Genetics Stock Center ( http://www . fgsc . net/ ) . The Pgpd-col-26; Δcol-26 strain was constructed by transforming the Δcol-26 mutant with a DNA fragment containing the A . nidulans gpdA promoter , the open reading frame and 3’ untranslated region ( UTR ) of col-26 , and flanking regions homologous to the upstream and downstream genomic sequence of the csr-1 gene . Transformants were selected for resistance for cyclosporin [61] and tested for genotypes by diagnostic PCR . The transformants with positive results were backcrossed to FGSC 2489 to obtain a csr-1::PgpdA-col-26; Δcol-26 homokaryotic strain . The T . reesei ΔbglR mutants were created by transforming protoplasts of an uridine auxotrophic strain made from QM6a ( Δpyr-4 ) [46] with two split-marker DNA fragments using method described in [62] . One of the split-marker fragment contains a ~1 kbp sequence homologous to upstream genomic sequence of the bglR gene followed by the promoter and the first half of the pyr-4 coding sequence and the other contained the second half the pyr-4 coding sequence with ~400 bp of overlap sequence with the first half of the pyr-4 coding sequence and a ~1 kb sequence homologous to the downstream genomic sequence of the bglR gene . Transformants were first grown on the plates with minimal media and subsequently transferred to PDA plates for conidiation . Conidia were tested for correct integration of the pyr-4 gene at the bglR locus using diagnostic PCR . The strains with the bglR gene disrupted were subjected to single colony purification . Three verified ΔbglR homokaryotic strains were used for downstream analysis . N . crassa cultures were grown on slants , each with 3 mL of Vogel’s minimal medium ( VMM ) with 2% sucrose [17] and 2% agar , at 30°C in dark for 24 hours , followed by 4–10 days in constant light at 25°C to stimulate conidia production . For growth phenotype testing in 24-well plates , conidia were inoculated at 106/ml into 3 mL of VMM with selected carbon and nitrogen sources in 24-well plates covered with breathable rayon film seal , and the culture were grown at 25°C in constant light with shaking at 200 rpm . The film was taken off before imaging . At least two replicates were included in each experiment and the same experiments were done at least twice . For mycelial biomass measurement , conidia were inoculated at 106/ml into 100 mL of VMM with selected carbon and nitrogen sources and grown at 25°C in constant light with shaking at 200 rpm . For crosses , one parental strain was grown on plates with synthetic crossing medium [63] for 2 weeks at room temperature for protoperithecial development . Conidia of the other parental strain were added to the plates for fertilization . Plates were kept for 3 weeks at room temperature . Ascospores were collected and activated as described [64] , plated on VMM with 1% sucrose , and incubated at room temperature for 18 hrs . Germinated ascospores were transferred to VMM slants supplemented with cyclosporin or hygromycin B and screened for desired genotypes by diagnostic PCR . T . reesei cultures were grown in either minimal media [65] for selecting transformants or with PDA for conidiation . For growth phenotype testing , conidia were inoculated at 106/ml into 3 mL minimal media with a selected carbon source in 24-well plates , and the culture were grown at 28°C in dark with shaking at 200 rpm . For RNA-seq experiments on VMM with 2% ( w/v ) maltose and metabolite analyses , conidia were inoculated at 106 conidia/mL into 3 mL VMM with 2% cellobiose and grown at 25°C in constant light and shaking at 200 rpm for 28 hrs . The mycelial biomass was washed twice with VMM-NC , followed by 18 hrs of incubation in VMM-NC . Mycelia were then transferred to VMM with maltose and grown 5 . 5 hrs for RNA-seq experiments , or transferred to VMM or VMM ( Glu ) with 2% ( w/v ) glucose and grown 5 . 5 hrs for metabolite profiling experiments . For RNA-seq experiments on VMM with other carbon sources , conidia were inoculated at 106 conidia/mL into 3 mL VMM with 2% sucrose and grown at 25°C in constant light and shaking at 200 rpm for 16 hrs . The mycelial biomass was washed twice with VMM-NC and then transferred to VMM with the selected carbon sources for 4 hrs prior to RNA extraction . Concentrations of carbon sources were glycerol ( 2 mM ) , fructose ( 2 mM ) , mannose ( 2 mM ) , trehalose ( 2 mM ) , sorbose ( 2 mM ) , xylose ( 2 mM ) , sucrose ( 2% w/v ) , cellobiose ( 2 mM ) , maltose ( 2 mM ) , avicel ( 1% w/v ) , amylose ( 1% w/v ) , amylopectin ( 1% w/v ) , xyloglucan ( 1% w/v ) . Mycelia of cultures were harvested by filtration and flash frozen in liquid nitrogen . RNA was extracted using the Trizol method ( Invitrogen ) and further purified using RNeasy kits ( QIAGEN ) . RNA-seq libraries of WT and Δcol-26 from 2% ( w/v ) maltose were prepared at the Functional Genomics Lab , a QB3-Berkeley Core Research Facility at UC Berkeley and sequenced on an Illumina HiSeq2000 at the Vincent J . Coates Genomics Sequencing Lab . Other libraries were prepared and sequenced at JGI as part of the Neurospora ENCODE CSP project . Total RNA starting material was 1 μg per sample and 10 cycles of PCR was used for library amplification . The prepared libraries were then quantified using KAPA Biosystem’s next-generation sequencing library qPCR kit and run on a Roche LightCycler 480 real-time PCR instrument . The quantified libraries were then multiplexed into pools of 9 libraries , and the pool was then prepared for sequencing on the Illumina HiSeq sequencing platform utilizing a TruSeq paired-end cluster kit , v3 , and Illumina’s cBot instrument to generate a clustered flowcell for sequencing . Sequencing of the flowcell was performed on the Illumina HiSeq2000 sequencer using a TruSeq SBS sequencing kit , v3 , following a 1x100 indexed run recipe . The sequencing reads that passed filtering from the CASAVA 1 . 8 FASTQ files were subjected to quality score checking using the FASTX-Toolkit ( http://hannonlab . cshl . edu/fastx_toolkit/ ) . Only reads with all bases scoring greater than 22 were used to map against predicted transcripts from the N . crassa OR74A genome v12 ( Neurospora crassa Sequencing Project , Broad Institute of Harvard and MIT http://www . broadinstitute . org/ ) with Tophat v2 . 0 . 4 [66] . The output bam files were sorted and indexed using the SAMtools package [67] and the indexed files were visualized in Integrative Genomics Viewer [68] . Transcript abundance reflected in FPKM was estimated with Cufflinks v2 . 0 . 2 [66] mapping against reference isoforms . Profiling data are available at the GEO ( http://www . ncbi . nlm . nih . gov/geo/; Series Record GSE GSE92848 and GSE95350 ) . For differential gene expression analysis , the bam files were first processed using the HTSeq package v0 . 6 . 0 [69] to generate raw counts , and the raw counts are subjected to differential analysis using the DESeq2 package version 1 . 10 . 1 [70] . The FungiFun2 online resource tool was used in functional enrichment analysis ( https://elbe . hki-jena . de/fungifun/fungifun . php ) [71] . The gene to category associations was tested for over-representation using hypergeometric distribution and the probability for false discovery rate was controlled by the Benjamini-Hochberg procedure . Mycelia from 3 mL cultures in 24-well plates were harvested by filtration followed by a quick wash in distilled water . Half of biological replicates were used for metabolite extraction and the other half were dried for biomass measurement . Washed mycelia for metabolite extraction were quickly put into a tube containing 200 μL zirconia beads ( 0 . 5 mm ) and 500 μL extraction buffer ( 80% acetonitrile , 20% water , 0 . 1 M formic acid ) and snap frozen in liquid nitrogen . Samples were stored at -80°C until extraction before mass spectrometry ( MS ) analysis . For metabolite extraction , the frozen samples were immediately put in a bead-beater ( BioSpec ) and homogenized for 1 min , and cooled on ice . The homogenate were centrifuged at 4°C at 14 000 rpm for 5 minutes , and the supernatants were subjected to either GC-MS or LC-MS analysis . For GC-MS analysis , 20 μL of the supernatant was collected and transferred to 1 . 5 mL micro-tubes containing 50 μL internal standard solution ( d27-Myristic acid in methanol , 250 μM ) . Samples were dried under reduced pressure using a speedvac ( Savant ) . Samples were derivatized for GC-MS analysis according to the method of Kind et al [72] . Briefly , 10 μL of methoxyamine hydrochloride dissolved in pyridine ( 40 mg/mL ) was added to each dried sample , and shaken at 30°C at maximum speed for 90 min using a thermomixer ( Eppendorf ) . A mixture of retention time marker standards were prepared by dissolving fatty acid methyl esters ( FAMEs ) of different linear chain lengths in chloroform ( C8 , C9 , C10 , C12 , C14 , C16 FAMES at 0 . 8 mg/ml , and C18 , C20 , C22 , C24 , C26 , C28 , C30 at 0 . 4 mg/ml ) . The FAME mixture ( 20 μL ) was added to 1 mL of N-methyl-N-trimethylsilytrifluoroacetamide ( MSTFA ) containing 1% trimethylchlorosilane ( TMCS ) , and 90 μL of the FAMEs/MSTFA solution was added to each sample . Samples were shaken at 37°C at maximum speed in a thermomixer for 30 min , and then transferred to and sealed in amber GC-MS sample vials containing glass inserts ( Agilent ) . Extraction blanks were prepared following the above procedure but starting with empty Eppendorf tubes . For LC-MS analysis , supernatant ( 350 μL ) was collected and filtered through 0 . 2 μm spin filters ( Pall ) by centrifugation for 1 min at 14000 rpm . Fifty μL of the filtrate was transferred to HPLC vials containing 50 μL of an internal standard mixture solution . Samples were kept at 4°C in the LC-MS autosampler chamber . Extraction blanks were prepared in triplicate by following the above sample preparation procedure with empty microtubes . For GC-MS analysis , samples were analyzed using an Agilent 7890 gas chromatograph ( Agilent Technologies , Santa Clara , CA ) connected to an Agilent 5977 single quadrupole mass spectrometer , all controlled by Agilent GC-MS MassHunter Acquisition software . Samples were injected using a Gerstel automatic liner exchange MPS system ( Gerstel , Muehlheim , Germany ) controlled by Maestro software . Sample injection volume was 2 μL , and the injector was operated in splitless mode . Samples were injected into the 50°C injector port which was ramped to 270°C in a 12°C/s thermal gradient and held for 3 min . The gas chromatograph was fitted with a 30m long , 0 . 25mm ID Rtx5Sil-MS column ( Restek , Bellefonte , PA ) , 0 . 25 mm 5% diphenyl film with a 10 m integrated guard column . Initial oven temperature was set at 50°C , and the over program was as follows: ramp at 5°C/min to 65°C , held for 0 . 2 min; ramp at 15°C/ min to 80°C , held for 0 . 2 min; ramp at 15°C/min to 310°C , hold for 12 min . The mass spectrometer transfer line and ion source temperature was 250°C and 230°C , respectively . Electron ionization was at 70 eV and mass spectra were acquired from 50 to 700 m/z at 8 spectra per second . Raw data was visually inspected using Agilent MassHunter Qualitative Analysis software ( Agilent Technologies , Santa Clara , CA ) . Agilent MassHunter Unknowns Analysis software v . B . 07 . 00 ( Agilent Technologies , Santa Clara , CA ) was used to perform peak deconvolution and library matching . A library match score was calculated for using FAME markers for retention time calibration , and matching mass fragmentation spectra to those in the Fiehn GC-MS Metabolomics RTL Library [72] . Metabolites of interest were only included in further analysis if their library match scores were greater than 75% . The identities of some metabolites with scores lower than 90% were confirmed by comparing mass spectra and retention times with that of authentic reference standards ( S6 Table ) . Mass spectral and retention time data from identified target metabolites were used to make an analysis method in MassHunter Quantitative Analysis Software for GCMS ( v . B . 07 . 00 ) . For each metabolite , a quantifier ion and two qualifier ions were defined to produce an extracted ion chromatogram in a specified retention time window . Integration of the extracted ion chromatogram peaks yielded peak areas that were further normalized by the mean of dry fungal biomass from biological replicates . The normalized peak areas were used for comparing the relative abundance of metabolites across samples . Targeted LC-MS analysis was performed for select metabolites not detected by GC-MS ( S7 Table ) . Samples were analyzed on an Agilent 6550 ESI-QTOF LCMS fitted with a Merck SeQuant Zic-HILIC column ( 150 x 1 mm , 3 . 5 mm , 100 Å ) with a guard column . Mobile phase consisted of 5% ammonium acetate in water ( solvent A ) , and 5% ammonium acetate in water-acetonitrile ( 10:90 ) ( solvent B ) . The following LC solvent time-table was used: 0 min , 100% B; 1 . 5 min , 100% B; 25 min , 50% B; 26 min , 35% B; 32 min , 35% B; 33 min , 100% B; 40 min , 100% B . Flow rate: 0 . 25 ml/min; injection volume: 2 μL . Each sample was analyzed in positive and negative ionization mode . Raw data was analyzed using Agilent MassHunter Qualitative analysis . Extracted ion chromatograms were produced from raw scan data using calculated m/z values for target metabolites , corresponding to their molecular ion and potential adducts: ( M+H ) + , ( M+Na ) + , ( M+K ) + for Positive mode; ( M-H ) - , ( M+COOH ) - , ( M+CH3COOH ) - for Negative mode . The identity of detected ions were confirmed by comparing retention time with reference standards , or checked by performing MS/MS analysis of the target ion , and comparing ion fragments with those in the METLIN online database . Integration of the extracted ion chromatogram peaks yielded peak areas that were further normalized by the mean of dry fungal biomass from biological replicates . Normalized peak areas were used for comparing the relative abundance of target metabolites across samples . For differential metabolite analysis , the normalized peak areas were log transformed and then used in the independent t-test of hypothesis that there is no difference between WT and the mutant . P values of less than 0 . 05 were considered significantly different and values between 0 . 05 and 0 . 1 were interpreted as indicating a trend toward statistical significance . Four biological replicates measured by GC-MS and two biological replicates measured by LC-MS were used in differential analysis . All metabolites that were found to be either significantly different or with a trend toward statistical significance were subjected to hierarchical clustering analysis . Hierarchical clustering analysis is performed with Cluster 3 . 0 [73] using log transformed mean of normalized peak areas from biological replicates . The values were centered to the mean across different growth conditions and normalized on a per metabolite basis . Average linkage clustering was performed with Euclidean distance as the similarity metric . Protein sequences of selected ascomycetes were downloaded from JGI Mycosm [74] and used to construct a local protein database using the NCBI BLAST+ application ( version 2 . 2 . 31 ) ( S5 Table ) . The putative COL26 orthologs were searched in the database using BLASTP with a cut-off E value less than e-20 . All hits were tested by reciprocally BLASTP against the N . crassa database and only ones that resulted in COL26 as the best hit were retained for protein sequence alignment . Protein sequences of the putative COL26 orthologs from selected species were aligned using three different programs: Clustal Omega [75] , MAFFT [76] , and MUSCLE [77] , and the best alignment was chosen and further trimmed using trimAl [78] . The trimmed alignment file was used for phylogenetic tree construction by the RAxML program with 200 bootstraps [79] . The result were visualized and edited in iTOL ( http://itol . embl . des/ ) [80] . | In nature , filamentous fungi sense nutrient availability in the surrounding environment and adjust their metabolism for optimal utilization , growth and reproduction . Carbon and nitrogen are two of major elements required for life . Within cells , signals from carbon and nitrogen catabolism are integrated , resulting in balanced metabolic activities for optimal carbon and nitrogen distribution . However , coordination of carbon and nitrogen metabolism is often missed in studies that are based on comparisons between single carbon or nitrogen sources . In this study , we performed systematic transcriptional profiling of Neurospora crassa on different components of starch and identified the transcription factor COL-26 to be an essential regulator for starch utilization and needed for coordinating carbon and nitrogen regulation and metabolism . Proteins with sequence similar to COL-26 widely exist among ascomycete fungi . Here we provide experimental evidence for shared function of a col-26 ortholog in Trichoderma reesei . Our finding provides novel insight into how the regulation of carbon and nitrogen metabolism can be integrated in filamentous fungi by the function of COL-26 and which may aid in the rational design of fungal strains for industrial purposes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"carbohydrate",
"metabolism",
"chemical",
"compounds",
"gene",
"regulation",
"carbohydrates",
"glucose",
"metabolism",
"organic",
"compounds",
"glucose",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"acidic",
"amino",
"acids",
"amino",
"acids",
... | 2017 | A fungal transcription factor essential for starch degradation affects integration of carbon and nitrogen metabolism |
We identified the dsRNA binding protein RbdB as an essential component in miRNA processing in Dictyostelium discoideum . RbdB is a nuclear protein that accumulates , together with Dicer B , in nucleolar foci reminiscent of plant dicing bodies . Disruption of rbdB results in loss of miRNAs and accumulation of primary miRNAs . The phenotype can be rescued by ectopic expression of RbdB thus allowing for a detailed analysis of domain function . The lack of cytoplasmic dsRBD proteins involved in miRNA processing , suggests that both processing steps take place in the nucleus thus resembling the plant pathway . However , we also find features e . g . in the domain structure of Dicer which suggest similarities to animals . Reduction of miRNAs in the rbdB- strain and their increase in the Argonaute A knock out allowed the definition of new miRNAs one of which appears to belong to a new non-canonical class .
MicroRNAs ( miRNAs ) are an abundant class of regulatory RNAs that are encoded in the genome of eukaryotes . They are involved in many biological processes , e . g . in development , differentiation and metabolism [summarized in 1 , 2] . Key proteins of the RNAi-machinery are required for miRNA processing as well as for binding to cognate sequences in mRNA targets and subsequent gene silencing . At least one Argonaute-like protein , one Piwi-like protein , one Dicer and one RNA dependent RNA polymerase ( RdRP ) was suggested to have constituted the basic RNAi machinery in the last common ancestor of eukaryotes [3] . Research has focused on plants and animals which belong to the supergroups of archaeplastida and opisthokonta respectively [4] . Of special evolutionary interest is the supergroup of amoebozoa with the model organism Dictyostelium discoideum which branched off the tree of life after the plants but before the animal/fungi split [5] . The D . discoideum genome encodes five Argonaute proteins of the Piwi clade ( AgnA to AgnE ) , two Dicer homologues ( DrnA and DrnB ) and three RdRPs ( RrpA to RrpC ) . Previous experiments have shown that AgnA [6] and RrpC [7] are required for the production of siRNAs while DrnB is essential for miRNA processing [8 , 9] . DrnA is probably the nuclease that generates siRNAs but apparently , a knockout is lethal . The structures of the D . discoideum RNase III proteins differ from those in animals or plants . Based on sequence similarities of the RNase III domains , DrnA and DrnB were classified as Dicer homologues , even though they lack a PAZ domain , the DUF283 domain and a helicase domain [3] . However , the latter is encoded in D . discoideum RdRPs instead [10] . Like animal Dicers , DrnA and DrnB encode one dsRBD instead of two as in plant Dicer homologues [3] . Surprisingly , the dsRBD is present at the N-terminus of DrnA and DrnB . Drosha like RNase III enzymes that are only present in animals are distinct in sequence from Dicer and consist of tandem RNase III domains and a C-terminal dsRBD [3] . Genes encoding proteins with these features are not present in the genome of D . discoideum . However , the domain architecture of DrnA and DrnB is more similar to that of Drosha-like enzymes than to Dicer homologues of animals and plants ( Fig 1 ) MiRNA transcripts ( pri-miRNAs ) are cleaved in two steps by RNase III nucleases: first , a hairpin-like precursor miRNA ( pre-miRNA ) is generated and then processed to a 21 nt long miRNA duplex . DsRNA binding domain proteins ( dsRBPs ) contribute miRNA processing [summarized in 11 , 12] . One strand of the miRNA duplex is stably incorporated into an Argonaute protein . Upon binding to perfect or partially complementary sequences in mRNA targets , the latter are endo- and exonucleolytically degraded or translationally repressed [summarized in 13] . Despite the similarities in miRNA processing of animals and plants , there are significant differences in the participating proteins and in compartmentalization . In animals , pri-miRNAs are processed by the RNase III enzyme Drosha [14] in the nucleus which is assisted by the DiGeorge syndrome chromosomal region 8 ( DGCR8 ) in humans [15–17] and Pasha in D . melanogaster and C . elegans [17 , 18] . Both are dsRNA binding proteins with two dsRBDs in a complex known as the microprocessor . Drosha and DGCR8 are nuclear proteins and partially co-localize at the nucleoli [19 , 20] . Pre-miRNAs are exported into the cytoplasm by Exportin-5 [21 , 22] , where they are further processed to mature miRNAs by Dicer [23–25] in concert with dsRBPs: In D . melanogaster Dcr-1 interacts with Loquacious ( Loqs ) , a protein with three dsRBDs that facilitates pre-miRNA formation [26 , 27] . Human Dicer is associated with the dsRNA binding protein TRBP or PACT and with Ago2 [28–30] . In plants , miRNAs processing depends on a nuclear complex consisting of the Dicer homologue DCL1 [31] , the dsRBP HYL1 [32–34] and the zinc-finger protein Serrate ( SE ) [35 , 36] . DCL1 and HYL1 co-localize in discrete subnuclear structures which are known as dicing bodies ( D-bodies ) while Serrate only partially resides in these structures [37 , 38] . D-Bodies share different components with Cajal bodies [39] . HYL1 and DCL1 contain two dsRBDs with coordinated functions in miRNA processing: they mediate dsRNA binding , protein-protein interactions and help to target the proteins to D-bodies [32 , 37 , 40 , 41] . While the miRNA biogenesis pathway in animals and plants is quite well investigated [summarized in 42] , much less is known in the Amoebozoa . With their unique position between animals and plants , the Amoebozoa and their siRNA machinery which cannot be unambiguously grouped to one or the other , provide an excellent model to understand RNAi evolution and the modular flexibility of this system . Here we investigated the role of dsRBPs in D . discoideum miRNA generation in comparison to plants and animals .
Since RNase III family members interact with dsRBD proteins to process miRNAs in plants and animals , we screened the genome of D . discoideum for proteins with homologous function . Using the InterPro algorithm , we were able to identify 10 annotated proteins with dsRBDs [43] ( Table 1 ) . We eliminated proteins which are involved in translation or ribosomal proteins and also excluded the RNase III proteins DrnA and DrnB . Dhx9 and HelF were interesting candidates , since they contain an RNA helicase domain which is absent in the D . discoideum dicer proteins . The knockout of helF had no effect on the miRNA processing in the amoeba [44] and the deletion of dhx9 is apparently lethal since no gene disruption could be obtained in several independent attempts and could not be analyzed . Two further uncharacterized proteins ( DDB_G0275735 and DDB_G0269426 ) were identified and annotated rbdA and rbdB respectively . They contain a single dsRBD and no additional defined protein domains ( see Fig 2 ) . We designed knockout strains where the promoter region and most of the dsRBD of rbdA and rbdB were deleted ( S2 Table ) . Schematic presentation of the AX2 wild type ( wt ) and the mutant allele is shown in S1A Fig . At least two rbdA- and rbdB- strains each were obtained by independent transformations . Gene deletions were confirmed by PCR analysis and the BsR-cassette from the rbdB- strains was removed ( S1B Fig ) . Deletion strains are referred to as rbdA- [2] , rbdA- [3] and rbdB- [1] and rbdB- [2] , respectively . RbdA- and rbdB- strains were analyzed for molecular phenotypes in miRNA processing by Northern Blot using two known miRNAs . As a control , we analyzed RNA from the agnA- and the drnB- strains . The latter was known to lack endogenous miRNAs [9] ( Fig 3 ) . While miRNA expression was similar to the wild type in the rbdA- strains , we could not detect any mature miRNAs in the rbdB- strains ( Fig 3 ) . They thus showed a similar molecular phenotype as the drnB- strain [9] . The same results were obtained with the original parent strains rbdB- [3] and rbdB- [5] . RbdB is thus necessary for proper miRNA processing . To our surprise , miRNA levels were strongly enriched in agnA- strains . Quantification revealed a 4-6-fold increase in the agnA mutant strain ( Fig 3 ) . The same molecular phenotype was recently observed upon deletion of the RdRP RrpC [8] . To analyze if RbdB was involved in pri-miRNA or pre-miRNA processing , we performed RT-PCR analysis with primers outside the predicted pre-miRNA sequence . We show that primary miRNAs accumulated in rbdB- and in drnB- strains indicating that both proteins were involved in processing of pri-ddi-miR-1176 and pri-ddi-miR-1177 ( Fig 4 ) . Since pri-miRNA processing occurs in the nuclei in plants as well as in animals , we further analyzed the subcellular localization of tagged RbdB . We expressed RbdB GFP fusions proteins in the wild type background and monitored fixed and living cells by fluorescence microscopy . RbdB GFP was found in the nuclei and was concentrated in discrete foci often associated with nucleoli ( Fig 5 ) . These structures were similar to plant D-bodies , in which the proteins DCL1 , HYL1 and SE co-localize and interact [37] . The subnuclear distribution patterns of RbdB GFP were similar to those observed for DrnB fusion proteins [46] and in addition , both proteins were shown to co-localize ( Fig 5C ) . We further showed that subcellular localization of DrnB and RbdB fusion proteins were not affected in rbdB- and drnB- strains , respectively ( S2 Fig ) . To prove the interaction of RbdB and DrnB we co-expressed both proteins in the AX2 wt as GFP and 3xHA fusion proteins , respectively . For technical reasons , we used a truncated version of RbdB GFP , ( RbdB Δ 504–612 ) that was shown to fully complement the mutant phenotype ( see below ) . We used this protein as bait and could precipitate full length DrnB-3xHA ( Fig 6 ) . Serrate is involved in plant miRNA processing and is a component of D-bodies [35–37] . An ortholog ( DDB_G0277375 ) in D . discoideum was predicted by InPranoid7 [47] and denominated srtA ( Fig 7A ) . SrtA was cloned and expressed as a GFP fusion in AX2 wt cells . We observed a diffuse localization in the nuclei and no nucleoli associated accumulation ( Fig 7B ) . We then analyzed srtA mutant strains for miRNA processing . Since a knockout strain could not be generated , we constructed a srtA [RNAi] knockdown strain [48] and examined miRNA levels by Northern Blot . In contrast to expectations , ddi-miR-1176 was enriched ( Fig 7C and 7D ) . A similar enrichment was observed for miRNA ddi-miR-7097 ( S4A Fig ) and for ddi-miR-1177 . Analyses of a control knockdown strain verified that the effect was specific for srtA ( S4B Fig ) . We confirmed the presence of srtA specific siRNAs by Northern Blot analysis ( S4C Fig ) and quantified knockdown efficiency by qPCR ( S4C Fig ) to an approx . 40% reduction in mRNA levels compared to AX2 . We then tested if the rbdB- phenotype in miRNA processing could be complemented by ectopic expression of RbdB GFP fusion proteins . MiRNA levels ( ddi-miR-1177 and ddi-miR-1176 ) were very similar to those in the AX2 wt , no matter if the protein was expressed from a high copy or from a low copy vector ( Fig 8 ) . Expression of the high copy and the low copy transgenes [49 , 50] , as determined by qPCR for one biological replicate each , differed at least by a factor of 8 ( S5A Fig ) and GFP fusion protein expression was not even detectable in the low copy variant ( S5B Fig ) . Overexpression of RbdB variants did not cause any mis-expression of miRNAs processing but always approached wild type expression levels . RbdB contains a C-terminal region rich in Prolin and Threonin residues , reminiscent to the P-rich site in human Drosha . We generated two mutants: one where 230 C-terminal amino acids were deleted ( RbdB Δ504–733 ) and one where only the P-rich site was deleted ( RbdB Δ504–612 ) ( Fig 9A ) . Both were introduced by extrachromosomal plasmids into rbdB- strains . Fluorescence microscopy of RbdB ( Δ504–612 ) GFP fusion proteins showed the same localization as RbdB GFP . In contrast , RbdB ( Δ504–733 ) GFP was not detectable by fluorescence microscopy ( Fig 9B ) but could be seen in Western Blots ( Fig 9B ) . This was either due to misfolding of the GFP domain or to diffuse localization throughout the cell . Northern Blot analysis and quantification revealed that both deletion constructs complemented the mutant phenotype ( Fig 9C ) . Using cNLS Mapper [51] two overlapping bipartite nuclear localization signals ( NLS1 aa 643–660 , NLS2 aa 643–666 ) were predicted in the C-terminal region of RbdB . This region was still present in the truncated variant RbdB ( Δ504–612 ) but not in RbdB ( Δ504–733 ) . In addition , we detected a putative nucleolar localization sequence ( NoLS residues 693–713 ) using NoD [52] . We fused the coding regions for amino acids 643–713 ( containing both signal sequences ) and the NLS2 sequence alone to GFP for expression in the AX2 wt . The NLS2 sequence was sufficient to bring the reporter into the nucleus but no nucleoli associated foci could be observed in the presence of the NoLS signal sequence ( S6 Fig ) . Since a strong decrease of known miRNAs has been observed in the rbdB- strain and an increase in the agnA- strain , we sequenced small RNAs from both strains and the wild type [6] . To identify putative miRNAs , we applied the following criteria . ( 1 ) Small RNAs should have a length between 20–24 nt . ( 2 ) The relative expression of putative miRNAs was at least 3-fold higher in agnA- strains compared to AX2 wt cells . ( 3 ) The relative expression of putative miRNAs was at least 3-fold lower in rbdB- strains than in AX2 wt cells . ( 4 ) The putative miRNA resides in a hairpin-like structure . ( 5 ) There is a putative corresponding miRNA-5p or -3p sequence . ( 6 ) miRNA-5p or miRNA-3p are detectable by Northern Blot . We considered a miRNA as validated when at least four of these criteria were met . In addition to some of the known miRNAs , including those of a recent study [8] , we detected 4 new species ( Table 2 ) , three of which fulfilled the criteria for canonical miRNAs . Since the read number per strain was relatively low , this analysis is by far not complete but only demonstrates a proof of principle . S3 Table shows absolute read counts of miRNAs in the different strains . S7 Fig shows relative expression of the miRNA candidates . Notably , the miRNA-like non-canonical small RNA ( miRNA-like_D4 ) matched four positions ( S4 Table ) with two very close to the telomeres . The 22 nt long RNA showed an elevated expression level in the agnA- strain based on the deep sequencing studies but no significant differences between the AX2 wt and the rbdB- strain . Additionally , the flanking sequences did not fold into a canonical hairpin-like structure and did thus not fulfill the official criteria of a canonical miRNA [53] . In Northern Blots the small RNA behaved like a canonical miRNA in terms of expression patterns in the RNAi-mutant strains . It may constitute a new class of miRNAs which is generated by cleavage in the adjacent stem-loop structures . This is somewhat supported by the observation that no corresponding miRNA-5p or -3p was found . Another non-canonical miRNA ( miRNA-like _D3 ) was located in the intron of the DNA transposon thug-S [54] and is thus encoded several times in the genome ( S4 Table ) . Thug-S derived miRNAs were already published by Avesson et al . 2012 . However , the great majority of these RNAs were found in developed cells ( 16 hours and 24 hours RNA libraries ) [8] . We were able to detect thug-S derived miRNAs by Northern Blots in vegetative cells although very weakly ( S8 Fig ) . Northern Blot analysis and folding analysis for the remaining small RNAs are also shown in S8 Fig . The canonical and miRNA-like miRNAs are listed in Table 2 .
We have shown that the dsRNA binding protein RbdB is a necessary component for miRNA processing in D . discoideum . It interacts with the nuclear protein DrnB , which has previously been shown to be required for miRNA accumulation [8 , 9] . Disruption of the rbdB gene did not result in an obvious mutant phenotype in growth or development under laboratory conditions but in a molecular phenotype in that previously identified miRNAs were almost entirely lost . In contrast , the disruption of the closely related rbdA gene had no such effect . In agreement with the subcellular localization of RbdB , pri-miRNAs were found to be enriched in rbdB- strains and the same was true for drnB- cells . Both proteins are thus at least required to convert pri-miRNAs to pre-miRNAs . Since rbdA- and helF- strains show normal miRNA accumulation , no other cytoplasmic dsRBP appears to be involved in the accumulation of mature miRNAs . We can , however , not rule out the unusual case that some other non-dsRBD proteins adopted this function . The knockout of DrnA is apparently lethal since no clones could be obtained in multiple independent attempts . In Arabidopsis , protein interactions in D-bodies are mediated by the second dsRBD of DCL1 and probably by the second dsRBD of HYL1 [32 , 41] . Drosha requires DGCR8 and its two encoded dsRBDs for accurate and efficient pri-miRNA processing [55] , [16] . The zinc-finger protein Serrate is another component of the microprocessor-like complex in plants and contributes to miRNA processing . A knockdown of the D . discoideum homologue SrtA resulted in an unexpected enrichment of miRNAs . This suggests that SrtA plays a direct or indirect role in miRNA processing which is different from that in plants . However , due to the relatively low knockdown efficiency of 40% mRNA reduction , we can only speculate that Serrate in D . discoideum has adopted the role of a negative regulator of miRNA processing and may compete with the microprocessor complex of DrnB and RbdB . DrnB localizes in distinct subnuclear compartments which can often be found in close association with nucleoli [46] . RbdB GFP fusions had a similar distribution and both proteins co-localize in these structures . This is reminiscent of plant Dicing Bodies , where HYL1 and DCL1 co-localize . The microprocessor in D . discoideum appears to be strictly confined to the periphery of the nucleoli . A canonical NLS was found in RbdB and we show that it was necessary and sufficient for nuclear import but not for localization in D-bodies . Interestingly , this precise localization seems not to be necessary for function since several mutant constructs diffusely localize in the nuclei or even in the cytoplasm but still rescue the phenotype in rbdB- strains . We assume that small but sufficient amounts of the protein , undetected by fluorescence microscopy , are transported to the localized microprocessor . This occurs apparently not by co-import of DrnB and RbdB since both proteins localize correctly even if the other one is knocked out . Neither in plants nor in D . discoideum it is , however , known if D-bodies are the location of miRNA processing , they may also be storage particles for the microprocessor and be dispensable for miRNA generation . We have previously shown that the argonaute protein AgnA is required for siRNA production [6] . Here we demonstrate a direct or indirect negative effect of AgnA on miRNA accumulation since all examined miRNAs display significantly higher levels in agnA- cells . By the criteria of overexpression in agnA- cells and underexpression in rbdB- cells we have identified , as a proof of principle , 4 new miRNAs and verified their existence by Northern blot . Surprisingly , one of them could not be folded into a hairpin by Mfold . We termed this a miRNA-like small RNA . Consistently , we find no evidence for a corresponding miRNA5p or -3p at this locus . A second miRNA with essentially the same features was detected by deep sequencing but was barely detectable in Northern Blots . Such miRNA-like small RNAs would probably have escaped most prediction tools . It will require further investigations to determine if these molecules are derived from a new class of precursors that may be processed by alternative pathways . One may speculate that the predicted hairpin structures adjacent to the unpaired miRNA region may serve as unconventional processing sites for DrnB . Taken together , our data suggest a hybrid mechanism from plants and animals in D . discoideum: the RNase III enzymes are Dicer-like in sequence but more similar to animal Droshas in domain structure . There is a nuclear Serrate homolog in D . discoideum that is directly or indirectly involved in miRNA processing but it appears to have adopted a different function from that in plants . A similar functional role was found for AgnA since deletion of either gene resulted in an accumulation of miRNAs . Probably , both miRNA processing steps are carried out in the nucleus as in plants . This is suggested by the lack of suitable cytoplasmic dsRBD candidates that are usually required for generating mature miRNAs and by the presence of D-bodies in the nucleolar periphery . Furthermore , our data emphasize the modular character and the flexibility of the RNAi machinery: functional domains have been exchanged between proteins in the microprocessor during evolution and can still be shuttled by molecular methods without affecting the functionality of the complex .
All D . discoideum strains were grown axenically in HL5+ medium ( Formedia ) supplemented with Blasticidine S and/or Geneticin at concentrations of 10 μg/mL when required . Transformation into the axenic strain AX2 or derivates was done by electroporation as described previously [56] . When integrating plasmids or knockout constructs were transformed , cells were subcloned in order to isolate single colonies . Clones were considered independent when they were derived from different transformations . After transformation with extrachromosomal vectors , cell populations were used for further analysis . DNA oligonucleotides ( Invitrogen ) used in this study are listed in S1 Table . Isolation of total RNA from D . discoideum and Northern Blot analysis of small RNAs were performed as described previously [6] . Blots were probed with 5’ 32P labeled DNA oligonucleotides that are listed in S1 Table . qRT-PCR analysis were performed as described elsewhere [6] . Illumina sequencing of small RNAs ( <400 nt ) from the AX2 wt and from the agnA- strains was described previously [6] . The small RNA fraction from the rbdB- strain was prepared and sequenced in the same way . Around 5 , 3 M , 4 , 9 M and 6 , 6 M reads with sufficient quality for the AX2 wild type , the angA- strains and the rbdB- strain respectively were obtained . The processed and trimmed reads were mapped by the short read mapper segmehl [61] against the D . discoideum chromosomes ( DDB0169550 , DDB0215151 , DDB0232428 , DDB0232430 , DDB0232432 , DDB0237465 , DDB0215018 , DDB0220052 , DDB0232429 , DDB0232431 , DDB0232433 ) . The results were visualized and analyzed by the Integrative Genomics Viewer [62] or by the Integrated Genome Browser [63] . The data have been deposited in NCBI’s Gene Expression Omnibus ( GEO ) [64] and are available through GEO Series accession number GSE56111 . Western Blot analysis to verify gene expression of fusion proteins or to analyze Co-IP experiments were performed a described elsewhere [6] . Co-IP experiments were performed as described elsewhere [65] , except for the following modifications . 5 x 108 cells were resuspended in 5 mL lysis buffer ( 10 mM Tris-Cl [pH 7 . 5] , 150 mM NaCl , 0 . 5 mM EDTA , 0 . 5% NP-40 , 25 mM MgCl2 , 1 tablet of Roche proteinase inhibitor cocktail mini ) . After binding , the GFP-Trap beads ( ChromoTek ) were washed four times with a buffer containing 10 mM Tris-Cl ( pH 7 . 5 ) , 150 mM NaCl , 0 . 5 mM EDTA , 25 mM MgCl2 . Bound protein was boiled off the beads in 80 μl Laemmli buffer . Aliquots were taken from the intermediate steps ( Input , Preclear ) and compared to bound protein by SDS-PAGE and subsequent Western blotting . Images were acquired on a Leica DMIRB inverted microscope with a DC350 camera and IM50 Acquisition software ( Leica Microsystems , Wetzlar , Germany ) or on a Leica DM 5500 with a DFC365 camera and MMAF acquisition software . Around 2 x 105 cells were plated on a coverslip to settle down for 20 minutes . Cells were washed with phosphate buffer , fixed in 4% Formalin for 5 min at 22°C and then permeabilized for 5 min in ice-cold methanol . Alternatively , cells were treated 7 minutes with methanol , only . Afterwards , cells were stained with DAPI for 3 minutes ( DAPI stock solution ( 1 mg/mL ) was diluted 1:15 . 0000 in 1 x PBS ) and washed two times with 1 x PBS . The coverslips were mounted on a slide with a drop of mounting medium ( 90% ( vol/vol ) glycerol , 20 mM Tris-HCl , and 1 g ml1 1 , 4-diazabicyclo[2 . 2 . 2]octane ( pH 8 . 3 ) ) . Living cells were incubated in Low Fluorescence Axenic Medium ( Formedium , Hunstanton , UK ) and analyzed on the Leica DMIRB . | miRNAs are essential regulators in eukaryotic cells and serve to control translation and stability of mRNAs . Processing of primary miRNA transcripts is carried out in two steps by evolutionary conserved machineries consisting mainly of double-strand specific RNases of the Dicer family and accessory double-strand RNA binding proteins ( dsRBPs ) . Regulation occurs by effector proteins of the Argonaute family . While processing in plants is confined to the nucleus , the mechanisms is split into a nuclear and a cytoplasmic step in animals . By knock-out and complementation experiments , we identify RbdB in the amoebozoa Dictyostelium as the accessory dsRBP processing component for both steps . Fluorescence microscopy shows that RbdB co-localizes with the RNaseIII Dicer B in nucleolar foci suggesting mechanistic similarities to plants . Functional domain analysis of RbdB and the structure of Dicers , however , indicate similarities to animals . This places Dictyostelium at an evolutionary branch point between plants and animals . Deep sequencing reveals that the rbdB knock-out strain shows reduced accumulation of microRNAs . Comparison with the wild type and the miRNA overexpressing agnA knock-out strain , allowed for the identification of new miRNAs in Dictyostelium which may have escaped detection by other methods . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"rna",
"interference",
"molecular",
"probe",
"techniques",
"gene",
"regulation",
"dictyosteliomycota",
"cloning",
"dictyostelium",
"plasmid",
"construction",
"northern",
"blot",
"micrornas",
"model",
"organisms",
"dna",
"construction",
"molecular",
"biology",
"techniques",
... | 2016 | Analysis of the Microprocessor in Dictyostelium: The Role of RbdB, a dsRNA Binding Protein |
Although homologous recombination is an important pathway for the repair of double-stranded DNA breaks in mitotically dividing eukaryotic cells , these events can also have negative consequences , such as loss of heterozygosity ( LOH ) of deleterious mutations . We mapped about 140 spontaneous reciprocal crossovers on the right arm of the yeast chromosome IV using single-nucleotide-polymorphism ( SNP ) microarrays . Our mapping and subsequent experiments demonstrate that inverted repeats of Ty retrotransposable elements are mitotic recombination hotspots . We found that the mitotic recombination maps on the two homologs were substantially different and were unrelated to meiotic recombination maps . Additionally , about 70% of the DNA lesions that result in LOH are likely generated during G1 of the cell cycle and repaired during S or G2 . We also show that different genetic elements are associated with reciprocal crossover conversion tracts depending on the cell cycle timing of the initiating DSB .
A double-stranded break ( DSB ) is a potentially lethal DNA lesion that can lead to genomic instability and chromosome rearrangements if not promptly repaired . DSBs and other forms of DNA damage ( for example , single-stranded nicks or base damage ) can result from exogenous ( for example , gamma- or UV-radiation ) or endogenous sources [1] . Inverted repeats have been called “at-risk motifs” because of their ability to form recombinogenic secondary structures such as hairpins and cruciforms when intrastrand pairing takes place [2] . Hairpin structures can cause replication fork pausing and lead to DSBs [3] . Cruciform structures are thought to promote genome instability because they resemble Holliday junctions that can be cleaved by resolvases . The studies of the effects of inverted repeats on genomic stability have usually been done using repeats introduced into the genome by transformation [4] , [5] or in strains with defective DNA replication [6] , [7] . Below , we will show that naturally-occurring inverted repeats act as recombination hotspots in strains with unperturbed DNA replication . In S . cerevisiae , DSBs can be repaired by non-homologous end joining ( NHEJ ) or by homologous repair ( HR ) [8] . In HR events , an intact donor DNA molecule ( a sister chromatid or a homolog ) is used to repair the broken molecule . HR can result in conversions ( the non-reciprocal transfer of DNA sequence ) and crossovers . Conversion events can be associated or unassociated with reciprocal crossovers ( RCOs ) . Conversion events result from either correction of mismatches in heteroduplex DNA by the mismatch repair system [9] , or by repair of a double-stranded DNA gap located near the site of the DSB [10] . Mitotic crossovers often result in LOH distal to the crossover ( Figure 1 ) . Previously , we showed several different types of gene conversion events that likely reflect differences in the timing of the recombination-initiating DSBs in the cell cycle [11] , [12] . Figure 1 shows a comparison of conversion tracts resulting from repair of a DSB formed during S/G2 ( Figure 1A ) or G1 ( Figure 1B and 1C ) ; the red and black lines represent homologs with heterozygous SNPs . Figure 1A shows a DSB on a black chromatid during G2 of the cell cycle that is repaired by HR , resulting in a crossover and associated 3∶1 conversion tract ( three chromatids with red SNPs and one with black SNPs ) . Figure 1B shows a DSB on the black homolog during G1 of the cell cycle that is replicated to generate two broken chromatids . Since both chromatids are broken in the same location , only the homolog is available as a template for HR . If one chromatid is repaired resulting in a crossover and associated conversion event , and the second chromatid is repaired as a conversion event unassociated with crossovers , there will be a 4∶0 conversion tract . Figure 1C also shows a DSB on the black homolog during G1 of the cell cycle . Following replication , repair occurs as in Figure 1B , but one broken chromatid has been processed to a greater extent than the other , resulting in unequal-sized conversion tracts . The resulting conversion tract is a 4∶0/3∶1 hybrid tract . In this study , we map spontaneous crossovers and associated gene conversion events on the right arm of chromosome IV , an interval of about 1 . 1 Mb , using SNP microarrays to map events at approximately 500 bp resolution . We find that events occur throughout the chromosome arm , although their distribution is non-random . In particular , inverted pairs of transposons at two different locations result in homolog-specific hotspots for crossovers initiated by DSBs during G1 . We also demonstrate that LOH events are more frequently a consequence of G1-initiated DSBs than G2-initiated DSBs , and that conversion tracts associated with the G1 DSBs are longer than those associated with G2-initiated DSBs . Finally , we show that different genetic elements are associated with crossovers initiated during G1 and G2 . For example , regions of converging replication forks ( TER sites; [13] ) are associated with G2 , but not G1 , conversion events .
We used a colony-color screening assay to identify reciprocal crossovers on chromosome IV , similar to the previously described system [11] , [14] ( Figure 1A ) . The diploid JSC25 was generated by mating two sequenced haploid strains , W303a and YJM789 , that have 55 , 000 SNPs relative to each other [15] . We inserted the KANMX gene near the telomere on the right arm of the W303a chromosome IV homolog and SUP4-o ( encoding an ochre-suppressing tRNA ) allelically on the YJM789-derived homolog . The KANMX gene renders the cell resistant to geneticin ( GenR ) , a drug related to the bacteria-specific antibiotic kanamycin . JSC25 is also homozygous for the ochre-suppressible ade2-1 mutation . Diploid strains of this genotype that lack SUP4-o form red colonies , strains with one copy of SUP4-o form pink colonies , and those with two copies form white colonies . If there is a crossover between the centromere and the SUP4-o locus , there is a 50% chance , following chromosome segregation that the two daughter cells will be homozygous for the region centromere-distal to the crossover . If the crossover occurs at the time that the diploid JSC25 is plated , a red/white sectored colony will be formed ( Figure 1A ) , in which the white side of the sector is geneticin-sensitive due to loss of the KANMX gene , and the red side of the sector is Ade− due to loss of SUP4-O . The frequency of red/white sectored colonies ( 3 . 1×10−5 ) in JSC25 is reflective of the rate of crossovers between CEN4 and the KANMX/SUP4-o markers . Since only one half of reciprocal crossovers result in a sectored colony , the rate of crossovers on the right arm of chromosome IV is 6 . 2×10−5/division . We mapped crossovers in about 140 sectored colonies of JSC25 . We monitored about 2300 SNPs between CEN4 and the KANMX/SUP4-o loci within the 1 . 1 Mb interval . As in our previous study [16] , for each SNP , we designed four 25-base oligonucleotides: one for each Watson and Crick strand of each homolog . The normalized ratio of experimental to reference hybridization of these samples to each oligonucleotide allowed us to determine whether the experimental strains were heterozygous or homozygous for a specific SNP . Figure 2 shows an example of the analysis of genomic DNA from a sectored colony ( JSC25 SP 129 ) . The hybridization of genomic DNA to the W303a- and YJM789-derived SNPs is shown by red and blue lines , respectively . In the low-resolution analysis ( Figure 2A ) , genomic DNA hybridizes equally efficiently to both W303a- and YJM789-derived SNPs ( normalized ratio of 1 ) from SGD coordinates 445 kb ( the location of CEN4 ) to 770 kb . Near coordinate 770 kb in Figure 2A , there is a transition from heterozygosity to homozygosity for the W303a and YJM789 forms of the SNPs in the red ( top ) and white ( bottom ) sectors , respectively . Based on the higher resolution depiction ( Figure 2B ) , it is clear that the transition points between heterozygosity and homozygosity are different between the two sides of the sector , occurring near 762 kb and 772 kb in the red and white sectors , respectively . The boxed region in Figure 2B represents a 3∶1 gene conversion tract ( Figure 1A ) . This pattern indicates that the gene conversion tract was initiated by a DNA lesion on the YJM789-derived chromosome , since the chromosome with the lesion acts as a recipient during the conversion event [17] . Microarray data illustrating a hybrid conversion tract are shown in Figure S1 . Of 139 crossovers analyzed , 121 ( 87% ) were associated with a contiguous gene conversion event . Of the 121 conversion tracts , there were 29 simple 3∶1 conversions , 7 simple 4∶0 conversions , and 46 simple 4∶0/3∶1 or 3∶1/4∶0/3∶1 hybrid conversions; 39 tracts had more complicated patterns of conversion ( Text S1 ) . Summaries of various features of all conversion tracts observed in our study are presented in Tables S1 , S2 , S3 . No aneuploidy of chromosome IV was observed for any of the sectored colonies . In addition , only one large deletion was detected . In the red sector of colony 42R/W ( Class E34 in Table S1 ) , we detected a deletion of about 100 kb located between pairs of inverted Ty elements at SGD coordinates 872 kb and 981 kb . We do not know whether this deletion occurred prior to the crossover or was associated with the crossover . We collected and analyzed 138 sectors and 139 reciprocal crossovers in the strain JSC25; one sectored strain had a double crossover . The locations of crossovers and associated conversion tracts are shown in Figure 3 . The crossovers associated with conversion tracts were separated based on which homolog had the initiating lesion . Events initiated in W303a- or YJM789-derived homologs are presented in the top or bottom of Figure 3 , respectively . The conversion tracts are depicted as pairs of lines ( discussed in the figure legend ) with the lengths of the conversion tracts indicated on the Y-axis . Crossovers unassociated with conversions are not shown in Figure 3 , because it is impossible to determine which homolog had the originating DSB . It is apparent that the distribution of conversion events along the chromosome arm is not even . In addition , in certain regions of the chromosome ( between SGD coordinates 900 kb and 1000 kb ) , the conversion tracts appear longer than most of those in other regions of the chromosome . Figure 4 summarizes the location of crossover-associated gene conversion events in a different way . In this figure , the X-axis shows the SGD coordinates and the Y-axis shows the number of conversion events that include the SNP located at that coordinate summed over all mapped events . The probability of a SNP being included within a conversion tract is a function of both the frequency of nearby DNA lesions and of the lengths of conversion tracts emanating from the initiating DNA lesions . In Figure 4 , we label regions where multiple contiguous SNPs were involved multiple times in a gene conversion event as potential hotspots ( HS1-HS7 ) . In Figure 4B , we show separately those events initiated on the W303a- and YJM789-derived homologs . These distributions are strikingly different . It is also clear that the distributions of G1-initiated conversions ( tracts with 4∶0 segments or other features that indicate a G1-associated DSB ) and G2-associated conversions ( 3∶1 conversions without a 4∶0 segment ) are different ( Figure 4C ) . HS3 and HS4 are both G1-specific and specific for the W303a-derived homolog . The mechanism responsible for this specificity will be discussed below . We used several different statistical methods to determine if the distribution of crossovers along the chromosome arm is non-random . In our first statistical analysis , we divided the right arm of chromosome IV into five bins of 200 kb ( starting at CEN4 ) and a sixth bin of about 80 kb . We counted the number of crossovers that occurred within each region ( additional details discussed in Text S1 ) . If all events are examined , the number of crossovers per bin is approximately proportional to the size of each bin ( p = 0 . 96 by chi-square Goodness-of-Fit ) . We also analyzed the events separated by the homolog and timing of the initiating lesion ( depicted in Figure S2 ) . Both the homolog-specific distributions of crossovers and cell cycle-specific distributions of events were significantly different with p values of 0 . 01 and 0 . 007 , respectively . The bin from 845–1045 kb ( containing HS3 and HS4 ) had a very significant enrichment of crossovers initiated on the W303a-derived homolog during G1 of the cell cycle compared to a random distribution of this class of event among all of the bins ( p = 2 . 3×10−5 ) . In Figure 5 , we show a comparison between the physical map of the right arm of chromosome IV with the genetic maps based on recombination data from JSC25 . In addition , we show separately the data from the W303a- and YJM789-derived homologs . This figure emphasizes the differences in recombination activity between the two homologs . For example , the region between 927 and 980 kb in the map of W303a is larger than expected from the physical map , but there are no events in this segment in the YJM789 map . Conversely , the segment between 768 and 821 kb has no events in the W303a map , but a greater-than-average number of events in the YJM789 map . Such differences are expected because of the differences in the distributions of hotspots on the two homologs . For example , the hotspot activity of HS4 is specific to the W303a chromosome . Thus , we expect an expansion of the genetic map on the W303a-derived homolog near HS4 , but no expansion of the map on the YJM789-derived homolog . We compared the DNA sequences of W303a and YJM789 to determine if any features of the sequence would suggest possible mechanisms for the observed differences in hotspot activity on the two homologs . Although there are many SNPs that distinguish the two homologs , insertions/deletions ( in/dels ) greater than 50 bp are relatively rare [15] , and most of these involve Ty elements or delta elements . These alterations and their approximate position by SGD coordinates are listed below; in this list , boldface indicates that the element is present in W303a and not in YJM789 , and the regular typeface indicates that the element is present in YJM789 and not in W303a: 1 ) delta element at 489 kb , 2 ) delta element at 513 kb , 3 ) delta element at 520 kb , 4 ) Ty element at 645 kb , 5 ) sigma element at 668 kb , 6 ) partial Ty1 element at 805 kb , 7 ) inverted pair of Ty2/Ty1 elements at 872 kb , 8 ) sigma element at 946 kb , 9 ) inverted pair of Ty2/Ty1 elements at 981 kb , and 10 ) unannotated partial delta element at 1151 kb . In addition to retrotransposon-related sequences , there were a small number of other in/dels . The W303a-derived homology has three tandem ENA genes located near SGD coordinate 534 kb , whereas the YJM789-derived homolog has only two . The YJM789-derived homolog has a 153 bp insertion in YDR077W relative to the same gene on the W303a-derived homolog . The YJM789-derived homolog has a partial deletion of ARS420 relative to the W303a-derived homolog . Finally , the YJM789 allele of YDR150W has more copies of an internal 192 bp repeat than observed in the W303a allele . Except for these in/dels and numerous SNPs , the genomic sequences of the two homologs are well conserved . For example , no changes were observed for the positions of tRNA genes or G-quadruplex structures . The significance of these genomic alterations on recombination hotspot activity is unclear with two exceptions . As discussed below , the inverted pairs of Ty elements located at positions 872 kb and 981 kb are likely to be important for the hotspot activities of HS3 and HS4 in the W303a-derived homolog . Other differences in hotspot activity between the two homologs could reflect multiple small changes or homolog-specific alterations in chromatin modifications . In addition , although it is difficult to determine what chromosome elements are associated with most of the hotspots , by analyzing the sequences within all of the conversion tracts , we were able to detect significant associations with certain chromosome elements; this analysis is described below . Differences in hotspot activities on the two homologs demonstrate the difficulty and , perhaps , the futility of generating a universal mitotic recombination map for S . cerevisiae . As shown below , however , differences in the recombination activities of the two homologs can lead to mechanistic insights into mitotic recombination . The HS4 hotspot , located between coordinates 970 and 1 , 000 kb , has the highest level of recombination as measured by the number of conversion events ( Figure 4A ) . This hotspot is specific to the W303a-derived homolog ( Figure 4B ) and is associated with G1-initiated gene conversion events ( Figure 4C ) . Analysis of the DNA sequences in the hotspot region shows that the peak of recombination activity of HS4 overlaps with a closely-spaced inverted pair of Ty elements ( YDRWTy2-3 and YDRCTy1-3 ) located between coordinates 981 and 992 kb . Ty elements are 6 kb retrotransposons present in about 40 copies per haploid genome [18] . To determine whether the hotspot activity of HS4 was related to the inverted pair of Ty elements , we first examined whether both of the haploid parental strains contained the inverted pair of elements . PCR analysis demonstrated that the inverted repeat is in the W303a homolog but not the YJM789 homolog ( Figure S3 ) . The observation that the HS4 hotspot is specific to the W303a-derived homolog that contains the inverted pair of Ty elements implicates this structure in the hotspot activity . The spacer between the Ty elements is about 25–66 bp , depending on the degree of mismatches allowed between the repeats ( Figure S4 ) . To examine some of the structural features of the HS4 hotspot , we measured hotspot activity in four different isogenic strains ( Figure 6 ) with the wild-type HS4 hotspot ( JSC71-1 ) , with a deletion of the HS4-associated Ty2 element ( JSC73-2 ) , with an expansion of the spacer between Ty2 and Ty1 ( JSC74-1 ) , and with a deletion of one of the delta elements associated with Ty2 . For all four strains , we inserted URA3 about 23 kb centromere-distal to HS4 ( 1013 kb ) and the hygromycin-resistance gene HYG about 20 kb centromere-proximal to HS4 ( 957 kb ) . These strains are depicted in Figure 6A and 6B . A crossover initiated at HS4 will result in a daughter cell that is HygR and 5-FOAR half of the time . A crossover initiated between CEN4 and HYG will result in a daughter cell that is HygS and 5-FOAR half of the time . By monitoring the ratio of HygR 5-FOAR recombinants to all 5-FOAR recombinants , we measured the recombination rate between the HYG and URA3 markers . The distance between CEN4 and URA3 is about 568 kb and the distance between HYG and URA3 is about 56 kb . Thus , if the numbers of crossovers are proportional to the size of the physical intervals , 10% of the 5-FOAR strains should be HygR . The observed frequency of derivatives of this class in the strain with the wild-type hotspot ( JSC71-1 ) was 0 . 17 ( Figure 6C ) . The difference in the number of observed and expected crossovers is statistically significant ( p<0 . 001 ) , as expected since this region contains the HS4 hotspot . In a strain with the deletion of the Ty2 element ( JSC73-2 ) or a strain with an expanded spacer ( JSC74-1 ) , the hotspot activity is significantly reduced ( Figure 6C ) . In contrast , in a strain in which the centromere-proximal δ element of Ty2 is deleted ( JSC77-1 ) , no significant loss of hotspot activity is observed . Deletion of the Ty2 element eliminates the possibility of hairpin formation at HS4 and an increase in the spacer between the repeated elements would be expected to substantially reduce the probability of hairpin formation [4] , [7] . The 330 bp δ elements flank Ty elements [18] and the 5′ δ element acts as a transcriptional promoter [19] . In the strain containing the deletion of the 5′ δ , Ty2 should no longer be transcriptionally active . The results summarized in Figure 6C , therefore , indicate that the formation of a secondary structure between the two Ty elements is necessary , but the transcriptional activities of both of the Ty elements are not required for hotspot activity at HS4 . The HS3 hotspot ( Figure 4 ) also maps at the positions of an inverted pair of Ty elements ( YDRWTy2-2 and YDRCTy1-2 ) separated by a spacer of about 100 bp . As with HS4 , this hotspot is specific to the W303a-derived homolog and G1-specific . Since the YJM789-derived homolog does not have this inverted pair of Ty elements [15] , we assume that the hotspot activity of HS3 is a consequence of formation of a secondary structure . Most ( 121/139; 87% ) of the crossovers examined in our study were associated with gene conversion . Of these conversion tracts , about two-thirds ( 82/121 ) are simple 3∶1 conversions , simple 4∶0 conversions , or hybrid 3∶1/4∶0 or 3∶1/4/0/3∶1 tracts as diagrammed in Figure 1 . As discussed previously , the simple 3∶1 tracts are most simply interpreted as events initiated in S or G2 , and the other two classes are likely to reflect events initiated in G1 . Thus , if we count only the simple conversions as illustrated in Figure 1 , two-thirds of the spontaneous crossovers have gene conversions indicative of a G1-initiated event , supporting previous conclusions based on a more limited dataset [11] . The numbers of conversion events of all types , as well as their schematic depictions , are given Table S1 . About 1/3 of the conversion tracts were more complicated than those predicted by Figure 1 . These complex tracts are discussed further in Text S1 . We also examined the conversion tract lengths for the G2- and G1-associated conversions . The median tract lengths of G2- and G1-associated crossovers ( Table S1 and S2; Text S1 ) were 4 . 7 kb ( 95% confidence limits of 2 . 6–9 . 5 kb ) and 14 . 8 kb ( 11 . 7–17 . 5 kb ) , respectively . By the Mann-Whitney test , the G1 and G2 conversion tract lengths are significantly different ( p<0 . 0001 ) . The median tract length for all conversion events was 10 . 6 kb ( 8 . 2–13 . 6 kb ) . From many genetic studies in yeast , it has been shown that the site at which recombination initiates is within or adjacent to the conversion tract [17] . There are a number of elements of chromosome structure in yeast that have been identified as potential chromosome fragile sites or potential hotspots of recombination . We compared our mapping of gene conversion events with the 18 possible recombination-inducing elements listed in Table S4 . The locations of some of these elements ( G-quadruplexes , replication termination regions , Ty retrotransposons , and long terminal repeats ) are depicted in relation to the crossover map in Figure S5 . The methods used for examining potential element enrichments are explained in detail in Text S1 . In brief , we determined whether there was a statistically significant excess of elements within the gene conversion tracts relative to that expected based on the known number of elements on the right arm of chromosome IV . In addition to examining all events collectively , we performed data analysis on more exclusive categories: conversion events initiated on the W303a-derived chromosome , conversion events initiated on the YJM789-derived chromosome , G1 conversion events , and G2 conversion events . The statistically significant associations from these analyses are in Table S5 . In the analysis in which all conversions are included , the only significant enrichments were with Ty elements , long terminal repeats ( δ elements ) that flank Ty elements , and tRNA genes . These same associations were also significant for G1 events and events initiated on the W303a-derived homolog . These associations are expected since the HS3 and HS4 hotspots are G1- and W303a-specific ( Figure 4B and 4C ) . No significant associations were observed between conversion tracts and δ elements unassociated with Ty elements . The non-random linkage between Ty elements and tRNA genes in the genome [18] may explain the tRNA enrichment . Replication forks do not proceed at a constant rate in the genome and replication-pause sites have been associated with a number of chromosome elements including the ribosomal RNA fork barrier , centromeres , tRNA genes , convergent replication forks , G-quadruplex motifs , and highly-transcribed RNA polymerase II genes ( reviewed in [20] ) . We found several similar motifs associated with conversion events . G2 conversion events were significantly associated with replication-termination regions . We also found a significant association between G-quadruplex motifs [21] and G1 conversion events , as well as an association between G-quadruplex motifs and conversion events initiated on the YJM789-derived chromosome; G-quadruplex structures slow replication fork progression [22] . Finally , the conversion events that are initiated on the YJM789-derived chromosome have an over-representation of sites of stalling of the Rrm3p helicase; this helicase is thought to aid replication through genomic regions at which DNA polymerase is paused [20] . We also found some elements of chromosome structure that were negatively associated with the conversion tracts ( Table S6 ) . In addition to looking for associations of various elements within the conversion tracts , we looked for associations of these elements with the termini of the tracts . The details of this analysis are described in Text S1 . In brief , we examined windows defined by transitions between heterozygous and homozygous markers at the ends of the conversion tracts for over- or under-representation of the chromosome elements described in Table S4 . One significant positive association was observed . We found that the LTR elements associated with Ty elements were significantly over-represented at the termini of conversion events . Although it is possible that these regions are preferred sites for the termination of conversion events , there are several alternative interpretations . First , since the YJM789-derived chromosome IV is missing five of the eight Ty elements that are present on the right arm of the W303a-derived chromosome IV [15] , several of these LTRs border large regions of heterology . It is possible , therefore , that the conversion tracts are terminated as a consequence of this heterology rather than as a consequence of the sequence/structure of the LTR . Second , HS3 and HS4 , the two strongest hotspots , are flanked by Ty-associated LTR events . Thus , these preferred sites for initiating DSBs may lead to preferred sites of tract termination indirectly . A choice among these alteratives will require additional experiments . Finally , it should be emphasized that , despite an enrichment for Ty-associated LTRs at the ends of conversion tracts , the termini of most tracts are not associated with these elements .
The right arm of chromosome IV has about 6 . 2×10−5 crossovers/division or about 6 . 2×10−8 crossovers per kb . As a convenient unit of mitotic crossovers , we suggest that 10−6 crossovers/division be defined as one micro Stern ( µS ) , named after Curt Stern , the discoverer of mitotic recombination [23] . We emphasize that this unit is restricted to crossovers that occur between homologs; sister-chromatid exchanges are undetectable by our methods since they do not result in loss of heterozygosity . We calculate that the right arm of chromosome IV has a genetic length of about 62 µS . Extrapolating the data from chromosome IV , we calculate that the yeast genome has a genomic crossover rate of about 6 . 2×10−4/cell division and a genetic map length of about 620 µS . This value is only a rough estimate since the frequency of mitotic recombination hotspots may vary from strain-to-strain , chromosome-to-chromosome , and homolog-to-homolog . However , since mitotic recombination hotspots appear weaker than meiotic recombination hotspots ( as described below ) , some of these variables may have only small effects . For example , although the W303a-derived chromosome IV has more hotspots than the YJM789-derived chromosome IV , the numbers of events initiated on each homolog ( determined by which homolog is the donor in the conversion event ) are almost identical: 61 events initiated on the W303a-derived homolog and 59 events initiated on the YJM789-derived homolog . Most previous studies of the rates of spontaneous crossovers utilize a diploid strain that is heterozygous for a can1 mutation on chromosome V [24] . The heterozygous diploid is sensitive to canavanine , and mitotic crossovers resulting in LOH for the can1 locus produce cananvanine-resistant derivatives . In several representative studies performed in different genetic backgrounds , the rates of crossovers per division in the 120 kb interval between CEN5 and CAN1 were measured as 6×10−6 [24] , 2×10−6 [25] , and 2×10−6 [11] . These rate estimates are in reasonable agreement with our rate measurement of 6 . 2×10−5 on chromosome IV , since the interval measured in our study is about nine-fold greater than the interval on chromosome V . Measurements of the rate of mitotic gene conversion events are usually done using diploids that have different non-complementing mutant alleles ( heteroalleles ) of genes affecting the biosynthesis of an amino acid or nucleotide . The rates of gene conversion , therefore , are estimated by the rate that prototrophic derivatives are produced from the auxotrophic diploid . The rates of heteroallelic gene conversion vary considerably in different studies . Representative rates per division ( the heteroallelic gene indicated in parentheses ) are: 2 . 5×10−7 ( arg4 ) [26] , 2 . 6×10−6 ( trp5 ) [27] and 3×10−7 ( leu2 ) [28] . Since gene conversion events are initiated by DSBs , the frequency of these events will be affected by the hotspot distribution . In addition , the distance between the mutant substitutions would be expected to influence the rate of gene conversion with this assay . It is important to note that gene conversion requires a recombinogenic DSB near the heteroalleles , whereas the crossover assays detect DSBs distributed throughout the region between the reporter gene and the centromere . In the present study , our analysis of conversion events was restricted to those associated with crossovers on chromosome IV . In a previous study , using microarrays that covered the entire yeast genome , no unselected spontaneous gene conversion events were detected in thirteen strains examined [16] . The yeast genome has a genetic length of about 4100 cM ( SGD ) , representing about 81 meiotic crossovers/division . Thus , the rate of crossovers in meiosis is about 105-fold higher than the mitotic rate . Although the rate of mitotic crossovers/division is low relative to the meiotic rate , since the number of mitotic divisions is likely to greatly exceed the number of meiotic divisions , mitotic crossovers are likely to be a potent mechanism for generating novel combinations of alleles . The distributions of mitotic crossovers and their associated gene conversion tracts are shown in Figure 3 and Figure 4 . Two points concerning the distribution should be emphasized . First , by our statistical analysis , not all of the peaks labeled as hotspots have significantly elevated levels of recombination . Second , for conversion-associated crossovers , the position of the initiating DNA lesion can be anywhere within the conversion tract . Previously , we showed that 230 repeats of the GAA trinucleotide was a hotspot for DSB formation [29] . In individual conversion events associated with this hotspot , the conversion tracts were propagated either toward the centromere , away from the centromere or bidirectionally from the tract . Our mapping of DNA lesions and the resolution of the mitotic recombination map , therefore , are limited by the size of the conversion tracts rather than the distribution of SNPs . An important conclusion from our analysis is that two homologs can differ significantly in their distribution of recombination events ( Figure 4B and Figure 5 ) . From the data shown in Figure 4B , it appears that HS1 , HS3 , and HS4 are hotspots on the W303a-derived chromosome , HS2 and HS6 are hotspots on the YJM789-derived chromosomes , and HS5 and HS7 are hotspots on both homologs . Thus , construction of a universal genetic map of mitotic events is difficult . Meiotic recombination events have been mapped throughout the genome using various types of microarrays [30]–[35] . The patterns of meiotic and mitotic recombination events on the right arm of chromosome IV show no evident similarity ( Figure S6 ) . Since meiotic recombination are initiated by meiosis-specific cleavage of the genome by Spo11p and the DNA lesions that initiate mitotic recombination are likely to have a variety of sources ( discussed further below ) , this difference in meiotic and mitotic patterns of recombination is expected . The levels of meiosis-specific DSBs at different places in the genome vary at least 400-fold [34] . We can estimate the relative “heat” of a chromosome region by determining the number of times that SNPs within that region are included within conversion tracts . For example , the peak SNP in HS4 ( the strongest hotspot ) was included in eleven recombination events ( Figure 4 ) . Since the average number of events for all SNPs throughout the right arm of chromosome IV is two , HS4 is only about five times more active than an average region . Thus , our data suggest that hotspots and coldspots are more pronounced for meiotic than for mitotic recombination . Both HS3 and HS4 co-localize with inverted Ty elements . Inverted repeats have been previously shown to be hotspots for certain types of mitotic recombination [5] , [36] . Most of these studies have involved non-yeast DNA sequences inserted into the genome by transformation . In our previous study [7] , a hotspot for chromosome rearrangements on chromosome III in cells that have low levels of DNA polymerase alpha was an inverted pair of Ty elements separated by about 280 bp . In contrast , HS3 and HS4 function as hotspots in cells with wild-type levels of DNA polymerases . The conversion tracts associated with inverted repeats at HS3 and HS4 are primarily 4∶0 or hybrid events , arguing that the recombinogenic lesion is formed prior to replication . Two types of secondary structures have been associated with inverted repeats: cruciforms and hairpins . Evidence for cruciform processing ( presumably by a Holliday junction-like resolvase ) resulting in two hairpin-capped ends has been obtained [5] . However , spacers greater than 20 bases substantially inhibit cruciform formation in E . coli [37]–[39] , and reduce the recombinogenic properties of palindromes in yeast [40] . An alternative model is that a single-stranded nick near the junction of the inverted repeats is processed into a single-stranded DNA gap , allowing formation of a hairpin on the intact DNA strand . The loop formed by the spacer in the hairpin could then be nicked by a nuclease such as the Mre11p complex [5] . We do not know the source of the nick that initiates hairpin formation . It is possible that DNA supercoils produced by Ty transcription are nicked by topoisomerases [41] . Our deletion analysis , however , demonstrates that convergent transcription of both Ty elements is not necessary for HS4 activity . It is also possible that the nick that initiates hairpin formation could represent a lesion generated by mismatch repair . Since the hotspot activity of HS3 and HS4 appears G1-specific , it is likely that the initiating nick occurs in G1 . In addition to inverted repeats , several other factors have been associated with mitotic recombination or DSB formation . Using several different assays , researchers have found that elevated levels of transcription stimulate recombination ( reviewed in [1] ) . Since mutations in many enzymes involved in DNA replication result in a hyper-Rec phenotype , it is likely that some spontaneous recombination events result from DNA lesions generated during DNA replication [42] . It has been shown that some DNA sequences or motifs are associated with slow or stalled replication forks [42] . Consequently , we analyzed our conversion tracts to determine if there was an over-representation of these structure/sequence motifs . We found that G4 quadruplex motifs were significantly associated with G1-initiated DSBs ( Table S5 ) . In yeast cells , G4 motifs have been shown previously to be enriched in areas of the genome that have high γ-H2AX binding , a signal associated with DNA damage [21] . In our study , the events initiated during S or G2 of the cell cycle were significantly associated with replication-termination ( TER ) regions . This association is especially intriguing because sister chromatids are intercalated following replication , and need to be resolved by Top2p [43] , [44] . Since Top2p induces DSBs to allow decatenation of sister-chromatids , it is possible that some of these Top2p induced cleavages are misrepaired , leading to homologous recombination . The median conversion tract length for all analyzed crossovers on the right arm of chromosome IV was 10 . 6 kb ( 95% confidence limits: 8 . 2–13 . 6 kb ) . As discussed in Results , we found that conversion tract lengths associated with G1-initiated breaks were significantly longer than G2-initiated breaks . Since the crossovers of both G1- and G2-initiated events occur in G2 , the broken ends generated by a G1-initiated lesion have a longer time for 5′-to-3′ resection prior to repair of the DSBs . In addition , the G1-initiated events involve the repair of two broken chromatids . If the conversion tracts associated with these two repair events are propagated in different directions , we would expect that the G1-initiated conversions would be longer than those initiated in G2 . As noted previously [11] , the mitotic conversion tracts characterized in our system are considerably longer than the average meiotic conversion tract of about 2 kb [45] . One of the unique characteristics of recombination events associated with HS4 is that their conversion tracts are substantially longer than those that occur at other locations along chromosome IV . The median conversion tract length associated with HS4 is 48 . 4 kb ( 95% confidence limits: 17 . 1–118 . 3 kb ) , much longer than the median tract length for all conversions ( 10 . 6 kb ) and longer than the median length of G1-induced conversions ( 14 kb ) . One factor that would be expected to affect tract length is that HS3 has two intact Ty elements on the W303a-derived homolog whereas the YJM789-derived homolog has only a 2 kb fragment of a Ty ( Figure 6 , Figure S3 ) . Thus , a DSB formed in this hemizygous insertion on the W303a-derived chromosome would need to be processed about 10 kb before exposing homology on the YJM789-derived chromosome . In addition , since the homology would be located internally on the resected strand , a recombination intermediate with single-stranded branches would be formed [17] . Removal of these single-stranded branches likely requires the nucleotide excision repair proteins Rad1p and Rad10p [46]–[47] , as well as the mismatch repair proteins Msh2p and Msh3p [48]–[51] . Thus , the hotspot activity of HS4 would also be dependent on these proteins . It is possible that the more extensive processing of the DSBs occurring in HS4 delays completion of the crossover , resulting in a longer crossover-associated gene conversion tract . As described above , about two-thirds of the observed conversion events involve the repair of two sister chromatids broken at approximately the same position . We infer these events are a consequence of a G1-induced DSB on one of the homologs , followed by replication of the broken chromosome . The repair of two broken chromatids subsequently occurs in G2 ( Figure 1B and 1C ) . This model is supported by the observation that γ-irradiation of G1-synchronized yeast cells results in 4∶0 and 4∶0/3∶1 hybrid tracts , whereas irradiation of G2-synchronized cells results in primarily 3∶1 tracts [12] . Although we prefer the interpretation that these events are initiated by a G1 DSB , we cannot rule out the possibility that a DNA lesion produces two broken chromatids during DNA synthesis . Conversion tracts more complicated than those depicted in Figure 1 were observed . These are discussed in Text S1 , and a mechanism to explain a class E2 ( Table S1 ) conversion tract is presented in Figure S7 . Although about two-thirds of the LOH-inducing events were caused by G1-DSBs in our study , it is likely that more spontaneous DSBs occur in S than in G1 . Rad52 foci , indicative of HR , are more frequent in the S- and G2-phases than in G1 [52] . A related observation is that the extent of resection of broken ends is substantially greater in G2 than in G1 [53] , [54] . One simple model that reconciles these observations is shown in Figure 7 . We suggest that most spontaneous DSBs are generated in S as a consequence of broken replication forks . These DSBs are preferentially repaired by sister chromatid recombination with only a small fraction involving an interaction with the homolog; this preference is likely to be ensured by sister-chromatid cohesion [55] . In yeast , DSBs generated by ionizing radiation during G2 of the cell cycle are usually repaired using the sister chromatid as the template [56]; such events are undetectable by our analysis . In contrast , the DSBs generated in G1 , even though less frequent than DSBs initiated in S , are preferentially repaired by recombination between the homologs , because both sister chromatids have a break at the same location . Thus , the model accounts for the preferential the use of the homolog in a G1-initiated event . In summary , our analysis shows that mitotic recombination events are distributed broadly along the chromosome . One well-defined class of hotspots is a consequence of secondary structure formation between inverted repeats . We suggest that most of the recombination events leading to loss of heterozygosity are initiated by a DSB in G1 since DSBs that occur in S or G2 are usually repaired by sister-chromatid exchanges .
A list of all strains used in this study is in Table S7 , and a list of all primers used is in Table S8 , unless otherwise specified . Mapping of crossovers was performed with the strain JSC25 ( MATa/MATα::HYG leu2-3 , 112/LEU2 his3-11 , 15/HIS3 ura3-1/ura3 GAL2/gal2 ade2-1/ade2-1 trp1-1/TRP can1-100::NAT/CAN1::NAT RAD5/RAD5 IV1510386::KANMX-can1-100/IVI1510386::SUP4-o ) , a diploid generated by mating the haploids W303a and YJM789 . Details of strain constructions are in Text S1 . All media were prepared using standard recipes [57] , except that the SD-Arg plates contained only 10 ìg/mL of adenine , as previously described [14] . To detect crossovers , we struck JSC25 to generate single colonies on rich solid medium ( YPD plates ) . The cells were grown for three days at 30° . A single colony was then resuspended in water and plated onto SD-Arg plates at a density of about 3000 cells per plate . Plates were incubated at 25° for three days and then at 4° for an additional day to let the red pigmentation develop . Red/white sectors were detected using a dissecting microscope . In colonies in which the red sector was at least one-eighth of the colony , we purified colonies from both the red and white sides of the sector . We confirmed that the red colonies were Ade− ( due to loss of SUP4-o ) and that the white colonies were GenS ( due to loss of the KANMX gene ) . Single red and single white colonies from each sectored colony were subsequently analyzed . Microarray samples were prepared , hybridized , and analyzed as previously described [16] . Briefly , agarose plugs containing genomic DNA were prepared from strains derived from the sectored colonies and from the control diploid reference strain . DNA from both the experimental and reference strains was sonicated to yield DNA fragments of about 200–400 bp; these samples were labeled with Cy5-dUTP and Cy3-dUTP , respectively . DNA samples from the experimental and control strains were mixed and competitively hybridized to SNP microarrays . Following washing and scanning of the arrays , probe signals were analyzed as ratios of hybridization of the experimental and control samples for each SNP-specific oligonucleotide represented on the arrays . The chromosome IV SNP microarray was designed and optimized following the principles described previously [16] . For the majority of the probes , the SNP is located in the center of the oligonucleotide , and the melting temperature for the oligonucleotide/genomic DNA hybrid is between 55 and 59°C . Minor deviations from these principles are described in Text S1 . The location and sequence of each SNP on the microarray are listed in Table S9 . Recombination between the URA3 and HYG markers was assayed by monitoring the HYG gene in 5-FOAR strains . Each strain was struck for single colonies on YPD plates . After incubating the cultures at 30° for 30 hours , we made patches of individual colonies on plates containing medium with 5-FOA . The plates were then incubated for two days at 30° . A single 5-FOAR colony from each patch was then transferred as a patch to YPD plates . After one day of growth at 30° , the samples were replica-plated to YPD plates containing hygromycin . About 400 5-FOAR colonies were analyzed for each strain . We used chi-square goodness-of-fit tests to compare expected values with observed values . VassarStats ( http://vassarstats . net/ ) and Microsoft Excel were used for this statistical test . We used Table B11 of [58] to calculate 95% confidence intervals on median estimates of conversion tract length . The Mann-Whitney test from the VassarStats website was used to compare conversion tract lengths . Data were corrected for multiple comparisons in the element enrichment analysis [59] . | Double-strand breaks ( DSBs ) are DNA lesions that can be fatal to a cell if left unrepaired . They can be caused by exogenous sources , such as gamma radiation , or endogenous stresses , such as high levels of transcription . Yeast cells primarily repair DSBs that are initiated outside of meiosis by mitotic recombination , which can result in physical exchanges between chromosomes , known as crossovers . We created a mitotic recombination map of one chromosome arm , representing 10% of the genome . This recombination map allows us to determine which regions of the chromosome arm are more susceptible to DNA damage than other regions . We were able to determine that most DSBs that result in detectable genomic changes were initiated prior to DNA replication and that some secondary DNA structures can be recombination hotspots . Recombination can also occur during meiosis , as a method of ensuring proper chromosome segregation . However , previously reported meiotic recombination maps have no correlation with our mitotic recombination map . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"nucleic",
"acids",
"genetics",
"dna",
"biology",
"dna",
"recombination"
] | 2013 | High-Resolution Mapping of Spontaneous Mitotic Recombination Hotspots on the 1.1 Mb Arm of Yeast Chromosome IV |
Eukaryotic release factors 1 and 3 , encoded by SUP45 and SUP35 , respectively , in Saccharomyces cerevisiae , are required for translation termination . Recent studies have shown that , besides these two key factors , several genetic and epigenetic mechanisms modulate the efficiency of translation termination . These mechanisms , through modifying translation termination fidelity , were shown to affect various cellular processes , such as mRNA degradation , and in some cases could confer a beneficial phenotype to the cell . The most studied example of such a mechanism is [PSI+] , the prion conformation of Sup35p , which can have pleiotropic effects on growth that vary among different yeast strains . However , genetic loci underlying such readthrough-dependent , background-specific phenotypes have yet to be identified . Here , we used sup35C653R , a partial loss-of-function allele of the SUP35 previously shown to increase readthrough of stop codons and recapitulate some [PSI+]-dependent phenotypes , to study the genetic basis of phenotypes revealed by increased translational readthrough in two divergent yeast strains: BY4724 ( a laboratory strain ) and RM11_1a ( a wine strain ) . We first identified growth conditions in which increased readthrough of stop codons by sup35C653R resulted in different growth responses between these two strains . We then used a recently developed linkage mapping technique , extreme QTL mapping ( X-QTL ) , to identify readthrough-dependent loci for the observed growth differences . We further showed that variation in SKY1 , an SR protein kinase , underlies a readthrough-dependent locus observed for growth on diamide and hydrogen peroxide . We found that the allelic state of SKY1 interacts with readthrough level and the genetic background to determine growth rate in these two conditions .
High fidelity in translation , one of the key steps in the expression of genetic information , is essential for functional integrity of the cell . Efficient termination is an important aspect of translational fidelity , and a multitude of factors participate in this process [1] , [2] . The efficiency of translation termination depends on the competition between stop codon recognition by release factors and decoding by near-cognate tRNAs ( tRNAs that can pair with two of the three bases of the stop codon ) [3] . Recent studies of translation termination in Saccharomyces cerevisiae have revealed genetic and epigenetic regulatory mechanisms that modify translation termination efficiency , which can affect cellular processes such as mRNA degradation and , in some cases , can confer a beneficial phenotype to the cell [4] . The most studied example of such mechanisms is the yeast prion [PSI+] , which is formed by a conformational change in Sup35p , a subunit of the translation termination complex [5] . [PSI+] is an epigenetic modifier of translation termination efficiency in S . cerevisiae [6] . Sup35p carries an intrinsically disordered prion-determining region at its amino terminus . When this domain switches to the aggregating amyloid conformation ( the prion conformation ) , much of the protein becomes unavailable for translation terminations , which in turn increases readthrough of stop codons [7] , [8] . [PSI+] was reported to generate different phenotypes in different genetic backgrounds , and most of these phenotypic effects were shown to be recapitulated by a partial loss-of-function allele of SUP35 , sup35C653R [9] . Previous studies have shown that some of the observed [PSI+]-dependent phenotypic effects are due to ribosomal frame-shifting [10] . It has also been proposed that some of the observed phenotypic variation in different yeast strains can be due to [PSI+]-dependent increase in readthrough , which results in ribosomes bypassing stop codons and reading into regions such as sequences at the 3′ untranslated regions or pseudogenes [11] . These regions are thought to be under less selective pressure than coding sequences , and therefore may be more divergent among different yeast strains . However , specific loci underlying phenotypic differences due to increased readthrough of stop codons have yet to be identified . Here , we used sup35C653R to examine the phenotypic effects of decreasing translation termination efficiency in various growth conditions in two divergent yeast strains , BY4724 ( a laboratory strain hereafter referred to as BY ) and RM11_1a ( a wine strain hereafter referred to as RM ) . Using a quantitative dual luciferase assay [12] , we showed that this partial loss-of-function allele of SUP35 increased readthrough of stop codons , as previously reported [13] . We identified nine growth conditions ( about one quarter of the growth conditions tested ) in which increased readthrough of stop codons resulted in different growth responses between BY and RM . Then , we used a recently developed linkage mapping technique , extreme QTL mapping ( X-QTL ) [14] , to find the genetic basis for the observed readthrough-dependent growth differences . We found one to six readthrough-dependent loci for the growth conditions examined , suggesting that phenotypes revealed by increased translational readthrough are often genetically complex . We further showed that variation in SKY1 underlies a readthrough-dependent locus observed for growth in diamide and hydrogen peroxide . We found that a complex interplay between sup35-mediated increase in readthrough , the allelic state of SKY1 , and genetic background determines growth in these two conditions . Our results provide new insights into the genetic basis of phenotypes revealed by decreased translation termination efficiency in yeast .
To compare the growth effects of increased readthrough in BY and RM , we replaced the wildtype allele of SUP35 in each strain with a partial loss-of-function allele , sup35C653R . Sequence comparison showed no differences between the BY and RM amino acid sequences of SUP35 in either the N-terminal ( prion forming ) domain or the C-terminal ( translation termination ) domain . We previously showed that the baseline readthrough level is different between BY and RM [15] . We used a quantitative dual luciferase assay [12] to show that sup35C653R increases readthrough in both strains by approximately four folds ( Figure S1 ) . We then measured growth rates of BY and RM carrying the wildtype alleles of SUP35 ( hereafter “wildtype” ) and sup35C653R ( hereafter “sup35” ) . We tested 33 different growth conditions including alternative carbon sources , different temperatures , and growth in the presence of small molecules that perturb varied cellular processes ( Table S1 ) . These growth conditions have previously been shown to induce different growth phenotypes in isogenic [PSI+] and [psi−] strains [16] and/or different growth phenotypes in BY and RM [17] . For each genetic background , we then calculated the ratio of the sup35 strain growth rate and the wildtype strain growth rate ( hereafter , “growth rate ratio” ) . We found that in control medium ( YPD ) , the growth rate ratios for both strains were not significantly different from one ( Figure 1 ) . This showed that in both strains , sup35-mediated increase in readthrough had no significant effect on growth rates in rich medium . For 24 out of 33 growth conditions tested , we found that the sup35-mediated increase in readthrough had the same effect on growth rate ratio in BY and RM; it decreased or did not alter either strain's growth rate ratio ( data not shown ) . However , we found nine growth conditions in which the growth rate ratio was significantly different between BY and RM ( uncorrected p<0 . 05; False Discovery Rate ( FDR ) ∼10%; Figure 1 ) . In five growth conditions ( chlorpromazine , cobalt chloride , cycloheximide , ethanol and hydrogen peroxide ) increase in readthrough did not alter growth rate in one strain ( growth rate ratio not significantly different from one ) while it decreased growth rate in the other strain ( growth rate ratio significantly less than one ) . In the presence of tunicamycin , increase in readthrough did not change growth rate in BY ( growth rate ratio not significantly different from one ) while it increased growth rate in RM ( growth rate ratio significantly greater than one ) . In the remaining three growth conditions ( diamide , E6-berbamine , and neomycin ) , increase in readthrough increased growth rate in one genetic background while decreasing growth rate in the other ( Figure 1 ) . We used X-QTL [14] to examine the genetic basis of the observed readthrough-dependent differences in growth rate ratio between BY and RM . For each growth condition , we performed X-QTL on two segregant pools in parallel: a wildtype pool from a cross between wildtype BY and RM , and a sup35 pool from a cross between BY and RM both carrying sup35C653R ( Materials and Methods ) . We grew these pools on selection plates ( rich medium plus the chemical agent of interest ) and control plates ( Materials and Methods ) , and compared the allele frequencies between the selected pools and control pools by microarray-based single nucleotide polymorphism ( SNP ) genotyping as previously described [14] . A locus that affects growth rate in a given condition independent of sup35 is expected to be detected as an allele frequency skew of similar direction and magnitude in both the wildtype and sup35 selected pools . In contrast , a locus whose effects depend on sup35 is expected to show a difference in the allele frequency skew between the two pools . The number of loci detected for growth rates in the nine conditions at an FDR of 5% ranged from one to 20 in both wildtype and sup35 crosses ( Figure S2A–S2I ) . The results in the wildtype cross were similar to those previously described for these growth conditions [14] , which showed the reproducibility of X-QTL . Most loci showed similar allele frequency skews in the wildtype and sup35 pools; however , 18 loci showed significant differences between these pools ( FDR = 5% , Materials and Methods ) ( Figure S2A–S2I ) . We refer to these loci as “readthrough-dependent” . One to six readthrough-dependent loci were detected in the growth conditions tested ( Figure 2A ) . These results showed that sup35-mediated effects on growth in certain conditions are genetically complex , as was previously suggested for some [PSI+]-dependent growth phenotypes [9] . Each readthrough-dependent locus had an effect in one to five conditions , with a total of ten distinct loci detected ( Figure 2B ) . In order to gain more insight into readthrough-dependent effects on growth rate , we focused on the locus on the right arm of chromosome XIII , which affected growth in five conditions ( Figure 2B ) . We chose one of the growth conditions in which this locus had the strongest effect for further investigation: diamide , a sulfhydryl-oxidizing agent [18] ( Figure S2D ) . At this locus , we detected a frequency skew in favor of the RM allele in the sup35 pool but not in the wildtype pool , which suggested that in the presence of increased readthrough , strains carrying an RM allele at this locus grow better on diamide than strains carrying the BY allele . Based on sequence comparison between BY and RM for the genes in this region ( Figure S3 ) , we selected SKY1 and MRE11 for further investigation , as both contained nonsynonymous changes in their open reading frames and in the downstream regions that might be translated due to increased readthrough . Comparison of the coding sequence of SKY1 between BY and RM showed 13 single nucleotide polymorphisms ( SNPs ) between the two strains , including seven nonsynonymous substitutions . The downstream sequence contained a two-nucleotide deletion in BY at the 129th nucleotide after stop codon , which results in the addition of three amino acids in BY before the next stop codon is reached . Comparison of the coding sequence of MRE11 between BY and RM showed eight SNPs between the two strains , including five nonsynonymous substitutions . The downstream sequence contained five SNPs , including four nonsynonymous substitutions . To test the causality of SKY1 and MRE11 polymorphisms for the effects of this locus , we replaced the SKY1 and MRE11 genes in both wildtype and sup35 RM with the BY versions . For both genes , we replaced the downstream sequence along with the coding sequence . We previously showed that the expression level of SKY1 is lower in RM than in BY , and that this difference maps to the location of the SKY1 gene , suggesting the presence of a cis-regulatory polymorphism [19] . Therefore , we included the upstream regulatory sequence in the SKY1 allele replacement along with the coding and downstream sequence . We then repeated the X-QTL experiments for growth on diamide with both wildtype and sup35 segregant pools from crosses using the RM parent strains with SKY1 and MRE11 allele replacements ( that is , both parents carried the BY allele of SKY1 or MRE11 , respectively ) . In the crosses with the MRE11 allele replacement , the results were unchanged; that is , we still saw a skew in the direction of the RM allele at this locus in the sup35 pool despite the fact that MRE11 was no longer polymorphic , ruling it out as the causal gene for this locus ( Figure 3A ) . In contrast , this allele frequency skew disappeared in the sup35 pool from the cross with the SKY1 allele replacement , and there was no longer any difference in allele frequency at this locus between the wildtype and sup35 pools ( Figure 3B ) . These results demonstrate that polymorphisms in SKY1 are causal for the effects of this locus , and that the difference in growth between the RM and BY alleles of SKY1 is revealed when readthrough is increased from the wildtype level by sup35C653R . To better understand the effect of SKY1 on the readthrough-dependent difference in growth rate between BY and RM on diamide , we measured growth rates of SKY1-swapped wildtype and sup35 BY and RM and compared them to the original strains . For this experiment , we constructed wildtype and sup35 BY strains carrying the RM allele of SKY1 . Similar to the previous replacements strains , we replaced the upstream regulatory region along with the coding and downstream sequences of SKY1 . We also measured growth rates of wildtype and sup35 BY and RM strains with SKY1 knocked out ( sky1Δ strains ) . In the presence of sky1Δ , there was no significant difference in growth between wildtype and sup35 strains in either genetic background ( Figure 4A ) implying that readthrough-dependent differences in growth are mediated by SKY1 . As expected from the X-QTL results , replacing SKY1 in the BY background with the RM allele increased growth rate in diamide in the presence of sup35 , and replacing SKY1 in the RM background with the BY allele decreased growth rate in the presence of sup35 ( Figure 4B ) . These results confirmed the growth effects of SKY1 polymorphisms . A readthrough-dependent locus for growth on hydrogen peroxide was also observed at this genomic location ( Figure 2B , Figure S2G ) . Therefore , we tested whether SKY1 polymorphisms underlie the effects of this locus on hydrogen peroxide as well . X-QTL experiments with allele replacement strains showed that polymorphisms in SKY1 are indeed causal ( Figure S4 ) . Growth rate measurements in wildtype and sup35 BY and RM strains and the corresponding SKY1 swapped and sky1Δ strains confirmed the effects of SKY1 polymorphisms on growth in the presence of hydrogen peroxide ( Figure S5A–S5B ) . When we first measured the growth rates of wildtype and sup35 BY and RM in the presence of diamide , we found that BY-sup35 grew significantly slower than wildtype BY , while there was little difference in growth rate between RM-sup35 and wildtype RM ( Figure 4C ) . When the BY allele of SKY1 was replaced with the RM allele , the difference in growth between the sup35 and wildtype strains was reduced , although it remained significant ( Figure 4C ) . When the RM allele of SKY1 was replaced with the BY allele , the sup35 strain grew somewhat slower than the wildtype strain ( Figure 4C ) . We found similar results for growth of these strains in presence of hydrogen peroxide ( Figure S5C ) . The observation that SKY1 replacement strains recapitulated the direction but not the magnitude of the effects of readthrough on growth in these conditions seen in the parent strains suggested the presence of interactions between SKY1 polymorphisms , readthrough , and genetic background . We used Analysis of Variance ( ANOVA ) ( Materials and Methods ) to formally test the effects of these factors and the interactions among them ( Table 1 , Table S2 ) . The model showed a major effect of the genetic background , with RM growing better in presence of diamide and hydrogen peroxide than BY . SKY1 allelic status also had a significant effect , with the RM allele increasing growth . Readthrough level did not show a significant effect on its own but did show significant interaction effects with both genetic background and SKY1 allelic status . These results suggest a complex interplay between the effects of sup35-mediated increase in readthrough , the allelic state of SKY1 , and genetic background in determining growth on diamide and hydrogen peroxide . At wildtype readthrough levels , RM grows better than BY , and swapping SKY1 in either strain with the version from the other strain has little effect ( Figure 4D , Figure S5D ) . When readthrough is increased by introduction of sup35C653R , growth rate decreases dramatically in BY , but shows no change in RM ( Figure 4D , Figure S5D ) . The slower growth rate of BY-sup35 in comparison to wildtype BY is partially rescued by introduction of the RM allele of SKY1 and completely rescued by knocking out SKY1 ( Figure 4D , Figure S5D ) . Introduction of the BY allele of SKY1 into RM-sup35 reduces the growth rate of this strain but not to the extent seen in the BY background ( Figure 4D , Figure S5D ) . Thus , sup35C653R and the BY allele of SKY1 act together to lower growth rate on diamide and hydrogen peroxide , and this effect is accentuated by as yet unidentified factors in the BY genetic background . Given the evidence for cis-regulatory polymorphism in SKY1 that lowers expression of the RM allele , we investigated whether differences in transcript abundance could account for the allelic effect of SKY1 . We used quantitative RT-PCR to measure SKY1 mRNA levels in wildtype and sup35 BY and RM , as well as in the corresponding SKY1-replaced strains ( Materials and Methods ) . We found that SKY1 mRNA levels were independent of the growth condition used ( Figure 5 ) . As expected based on microarray data [19] we found that SKY1 expression is higher in wildtype BY than in wildtype RM . Moreover , we found that swapping SKY1 in wildtype BY and wildtype RM with the alternate allele changed SKY1 expression level to the alternative level . These results confirm the presence of cis-regulatory polymorphism that alters the expression level of SKY1 between BY and RM . Surprisingly , in the RM background , increasing readthrough from the wildtype level to the sup35 level resulted in roughly a ten-fold drop in SKY1 expression level , while no change was observed in BY-sup35 . This drop in SKY1 mRNA in RM-sup35 was largely reversed by swapping in the BY allele of SKY1 ( Figure 5 ) . We used ANOVA to model the effect of the measured SKY1 mRNA abundance on growth rate , and then used the residual growth rate to test whether the allelic effect of SKY1 was changed . The results suggested that the readthrough-dependent growth effects of SKY1 are not mediated by mRNA levels ( Table 2 , Table S3 ) . To gain further mechanistic insight into how sup35C653R leads to the differential allelic effects of SKY1 , we swapped just the downstream sequences of SKY1 in both wildtype and sup35 BY and RM with the alternative alleles . These replacement strains differ from the parent strains only by the polymorphism that extends the C-terminus . Swapping this polymorphism alone captured the allelic effect of SKY1 in both growth conditions ( Figure 6 , Figure S6 ) , consistent with a differential effect of readthrough on the downstream regions from the two strains . To test whether the allelic effect of SKY1 is directly related to translational readthrough , we then introduced a second stop codon immediately after the native stop codon in wildtype and sup35 BY and RM strains . Introducing a second stop codon in wildtype BY and RM did not affect growth rates of these strains in diamide or hydrogen peroxide ( Figure 6 , Figure S6 ) . In the presence of the sup35 mutation , the growth rate in the BY strain with the second stop codon rose to the same level as when the SKY1 allele is replaced with the RM version , while the second stop codon did not alter growth rate in RM ( Figure 6 , Figure S6 ) . These results support the hypothesis that increased readthrough of the BY downstream region due to the sup35 mutation causes reduced growth in diamide and hydrogen peroxide , perhaps because translation of this region stabilizes Sky1p .
Modifying translational readthrough in S . cerevisiae has been shown to affect yeast cells in various ways [4] . The prion [PSI+] provides one example of translational readthrough modification in yeast cells . Previous works have shown that [PSI+] can reveal hidden phenotypic variation among yeast strains , that this effect is largely recapitulated by the sup35C653R mutation , which increases translational readthrough , and that the resulting phenotypic differences may have a complex genetic basis [9] , [16] . Here , we have advanced our understanding of the genetic basis of readthrough-dependent phenotypes by identifying specific loci that underlie hidden variation revealed by sup35C653R . Using this partial-loss-of function allele of SUP35 allowed us to focus on distinct hidden phenotypes in BY and RM revealed by increased translational readthrough . Our growth rate measurements in diverse stressful conditions for wildtype and sup35 BY and RM showed that sup35-mediated differences in growth between these two strains were relatively modest compared to previous studies of [PSI+]-mediated effects [9] , [16] . This could potentially be explained by other [PSI+]-dependent phenotypic effects in yeast , such as ribosomal frame shifting [10] or the presence of Sup35 prion aggregates [20] , which are absent in our system . We consider it more likely , given the reported recapitulation of most [PSI+] strain-dependent phenotypic effects with the sup35C653R mutation [9] , that this difference in effect sizes is due to the different genetic backgrounds used . We found that sup35-mediated increase in readthrough had different effects on growth rates in BY and RM for about one-quarter of the growth conditions tested . Our mapping results lend additional support to the previously reported inference that some readthrough-dependent growth phenotypes are genetically complex based on their segregation patterns [9] , and further suggest that some of the underlying loci have small effect sizes . We showed that SKY1 is the causal gene underlying the strongest readthrough-dependent locus detected for growth in the presence of diamide and hydrogen peroxide . Our results suggest that translation of the BY downstream sequence of SKY1 is disadvantageous for growth in these conditions . We found that a complex interplay between the genetic background , SKY1 allelic state , and sup35 determines growth rate in these two conditions . SKY1 mRNA measurements showed that the readthrough-dependent effect of SKY1 on the growth differences between BY and RM is not mediated by mRNA levels . However , we observed a dramatic drop in the SKY1 mRNA level in sup35 RM relative to wildtype RM , while we did not see a drop in the SKY1 mRNA level in sup35 BY relative to wildtype BY . One mechanism that could explain the sup35-mediated drop in the SKY1 mRNA level in RM is Nonstop mRNA Decay ( NSD ) pathway [21] , which might be differentially active in BY and RM . This mRNA surveillance mechanism is initiated when the ribosome reaches the 3′ end of the mRNA , and therefore eliminates transcripts lacking stop codons or transcripts that have stop codons that were bypassed during translation . Ribosomes are more likely to reach the 3′ end of an mRNA after reading through one or more stop codons in sup35 strains than in wildtype strains . Therefore , NSD is also more likely to eliminate such mRNAs in the presence of sup35 . Importantly , even a single ribosome that reaches the 3′ end of an mRNA is predicted to trigger NSD , resulting in the reduction of the mRNA level [22] . The decrease in SKY1 mRNA caused by sup35 in the RM background is largely rescued by swapping in the BY allele of SKY1 , which suggests that the combination of increased readthrough and NSD can act in an allele-specific fashion . Sky1p is a protein kinase that phosphorylates SR proteins [23] , proteins with domains containing alternating serine and arginine residues which are components of the machinery for the processing [24] and nuclear export [25] of mRNAs . One of the known Sky1p targets , Npl3p , was shown to promote translation termination accuracy in yeast [26] . However , the same paper showed that the role of Npl3p in translation termination is independent of the posttranslational modifications mediated by Sky1p . Sky1p has also been shown to regulate cation homeostasis and salt tolerance [27] . Deletion of SKY1 confers resistance to several anticancer drugs , such as cisplatin and carboplatin [28] , and to polyamine toxic analogues [29] . Several hypotheses have been proposed to explain the role of Sky1p in resistance to these drugs , such as a Sky1p-mediated effect on splicing or transport of target mRNAs , regulation of membrane permeability , and regulation of drug uptake . However , the direct target ( s ) of Sky1p that mediate these effects are unknown . Our results demonstrate that deletion of SKY1 also confers a growth advantage in the presence of oxidative stress inducers diamide and hydrogen peroxide . A genetic interaction between SKY1 and SUP35 was previously reported in the S288c background in rich media [30] . Here we showed that a genetic interaction is present between SKY1 and SUP35 in the presence of diamide and hydrogen peroxide in the BY background but not in the RM background . These results support previous finding that readthrough-dependent phenotypes vary based on the genetic background [9] .
Cultures were grown in minimal medium containing 0 . 67% ( w/v ) yeast nitrogen base without amino acids ( Difco ) containing 2% ( w/v ) glucose ( SMD ) or rich medium , as specified . Additional nutritional supplements or drugs were added as required . YPD plates were made as described [31] . For sporulation , SPO++ was used ( http://www . genomics . princeton . edu/dunham/sporulationdissection . htm ) . We used pEF675 ( a kind gift from Eric Foss ) to replace SUP35 with sup35C653R via two-step allele replacement [32] in BY4724 ( Mata ura3Δ lys2Δ ) [33] and RM11-1a ( Mata ura3Δ his3Δ0::NatMX hoΔ::HphMX AMN1BY ) . pEF675 was made by sub-cloning Sup35 into a common URA3-marked integrating yeast plasmid ( pRS306 [34] ) and subsequently changing cysteine 563 ( TGT ) to an arginine ( CGT ) . Successful replacement for each strain was then confirmed by sequencing . We refer to strains with SUP35 as wildtype and strains with sup35C653R as sup35 . We then transferred sup35 into strains with suitable genetic markers for X-QTL . To do so , we crossed BY-sup35 and RM-sup35 strains into BY MATα can1Δ::STE2pr-SpHIS5 lyp1Δ his3Δ1 and RM MATα AMN1BY his3Δ0::NatMX hoΔ::HphMX [14] , respectively . After sporulating the obtained diploids and genotyping the dissected tetrads , we selected BY Matα his3Δ1 lyp1Δ can1Δ::STE2prSpHIS5 sup35C653R and RM Mata AMN1BY his3Δ0::NATMX hoΔ::HphMX sup35C653R as sup35 parental strains in X-QTL . We used strains form [14] as wildtype parental strains in X-QTL . SKY1 and MRE11 replacement strains were generated by a two-step replacement method [35] . Each gene was first replaced with URA3-KanMX cassette from pCORE in wildtype and sup35 BY and RM strains generating goiΔ::URA3-KanMX knockout strains . SKY1 and MRE11 alleles from the donor strains were amplified by PCR and introduced into recipient strains to replace URA3-KanMX cassette . Where mentioned , 400-base pair from the upstream regulatory region or 300-base pair from the downstream region is included in making replacement strains . Allele replacements were confirmed by sequencing . The RM SKY1 and MRE11 sequences were obtained from the whole genome-sequencing project at the Broad Institute ( http://www . broad . mit . edu/annotation/genome/saccharomyces_cerevisiae/Home . html ) . All sequencing was done using standard dideoxy methods . We inoculated strains under examination in a 96-well plate ( Costar; 3370 ) in rich medium and incubated the plate in 30°C until saturation . We then used a sterile 96-pin replicator ( Nunc; 62409-606 ) to inoculate Costar 96-well plate ( 3370 ) containing rich medium and the reagent of interest . We then used Synergy 2 Multi-Mode Microplate Reader ( BioTek Instruments ) set at the desired temperature ( 30°C unless mentioned otherwise ) and with continuous shaking at medium speed to collect OD600 at 30-minute intervals for up to 20 hours . We used data points corresponding to 0 . 05<OD600<0 . 5 ( logarithmic growth phase ) to calculate growth rate . Growth rate was calculated as the slope of a linear regression of the log transformed logarithmic growth phase data points . For each strain , unless otherwise specified , growth rate in rich media plus the reagent of interest is normalized by the strain's growth rate in rich media . Growth rate is shown as the mean ± the standard deviation of values obtained from at least eight independent growth measurements , including at least four biological replicates . We then performed t-test comparison between BY and RM growth rate ratios for all growth conditions . We used the obtained p-values to estimate the FDR at p-value<0 . 05 based on q-value calculation [36] in R ( http://www . r-project . org/ ) . Dual luciferase assay was performed as explained before [12] . Plasmids with the stop codon ( pDB691 ) or the sense codon ( pDB690 ) , kindly provided by David Bedwell ( University of Alabama at Birmingham ) , were transformed into the indicated yeast strains , and transformants were selected on SMD drop-out plates lacking uracil . Transformed strains were grown in liquid SMD medium to a cell density of 0 . 5–0 . 7 A600 units/mL as measured using Synergy 2 Multi-Mode Microplate Reader ( BioTek Instruments ) . The luciferase assay was performed using the Dual-Luciferase Reporter Assay System ( Promega; E1910 ) . Approximately 104 yeast cells from each strain expressing the indicated dual luciferase reporter were lysed using 100 µL of Passive Lysis Buffer in a 96-well plate ( Costar; 3370 ) . Two microliters of the lysate were added to 10 µL of the Luciferase Assay Reagent II in an opaque 96-well plate ( Costar; 3614 ) . Relative luminescence units ( RLUs ) produced by firefly luciferase activity were then measured for 10 seconds using Synergy 2 Multi-Mode Microplate Reader ( BioTek Instruments ) . 10 µL of Stop&Glo buffer was then added to quench the firefly activity and activate the Renilla luciferase activity . RLUs were again measured for 10 seconds to determine the Renilla luciferase activity . Negative controls that contained all the reaction components except cell lysates were used to determine the background for each luciferase reaction and were subtracted from the experimental values obtained . Percent readthrough is expressed as the mean ± the standard deviation of values obtained from at least eight independent dual luciferase assay including at least four biological replicates . For each growth condition , we performed X-QTL on two biological replicates for the wildtype BY×RM cross and two biological replicates for the sup35 cross . MATa haploid segregants from the indicated cross were selected as explained before [14] . To create the segregating pool , a single colony of the diploid progenitor of the mentioned cross was inoculated into 5 mL YPD and grown to stationary phase . The diploid culture was spun down and the supernatant was decanted . The diploid pellet was then resuspended in 50 mL SPO++ sporulation medium . The sporulation was kept at room temperature ( ∼22°C ) with shaking and monitored for the fraction of diploids that had sporulated . Once more than 50% of the diploids had sporulated , 10 mL of the sporulation were spun down and then the supernatant was decanted . The pellet was resuspended in 2 mL water . 600 µL β-glucoronidase ( Sigma; G7770 ) were added to the preparation , and the mixture was incubated at 30°C for one hour . Water was added to the sample to the total volume of 20 mL . The spore preparation was spread onto SMD+canavanine/thialysine plates ( Sigma; C9758 for canavanine ( L-canavanine sulphate salt ) ; A2636 for thialysine ( S- ( 2-aminoethyl ) - L-cysteine hydrochloride ) ) , with 100 µL of sample going onto each plate . The plates were incubated at 30°C for two days . Then 10 mL of water were poured onto each plate and a sterile spreader was used to remove the segregants from the plate . The cell mixtures from four plates were then pipetted off the plates into a container . The pool was spun down and the water decanted . Haploid segregants were then inoculated into 100 mL liquid YPD and were incubated in 30°C while shaking on a rotary shaker at 200 rpm for about 30 minutes to recover . After the recovery , 100 µL of the segregant pool was pipetted and spread on the selection plates ( YPD+reagent of interest ) and control plates ( YPD ) . For each condition/cross we used five selection plates to pool the resistant segregants . For each cross , we used three control plates to pool the whole population of segregants . Selection and control plates were then incubated at 30°C for two days . DNA was extracted from the grown cells using Genomic-tip 100/G columns ( Qiagen; 10243 ) . DNA was labeled using the BioPrime Array CGH Genomic Labeling Module ( Invitrogen; 18095-012 ) with the sample being labeled with Cy3 dUTP and the reference being labeled with Cy5 dUTP . We used a BY/RM diploid as the reference for all hybridizations . Labeled samples were then hybridized onto the allele-specific genotyping microarray with isothermal probes that assay ∼18 , 000 single nucleotide polymorphisms ( SNPs ) between BY and RM [14] . The array data have been deposited in NCBI's Gene Expression Omnibus ( GEO ) [37] and are accessible through GEO Series accession number GSE33817 ( http://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE33817 ) . Hybridization intensities were extracted and normalized using the rank invariant method in the Agilent Feature Extraction software package . For a given SNP , the difference in the log10 ratios of BY and RM-specific probes on a single array ( or log10 intensity difference ) was computed . Background allele frequency changes that occur during pool construction were removed from the selection by subtracting the log10 intensity difference obtained for the whole ( control ) population from the log10 intensity difference observed in the selections . To find readthrough-dependent peaks , we used a Savitzky-Golay filter to smooth the input data within sliding windows of 100 probes . The Savitzky-Golay method essentially performs a local polynomial regression on a series of values to determine the smoothed value for each point . This smoothing approach was used to preserve local maxima in the data . For each probe , we subtracted the average of the two wildtype X-QTL replicates from the average of the two sup35 X-QTL replicates for each growth condition and used this measure as the input data . Readthrough-dependent loci were called if the smoothed value surpassed the threshold for a 5% FDR , where the number of false discoveries at each threshold was determined by using the same algorithm on the control data , which were obtained by subtracting wildtype control X-QTL results from sup35 control X-QTL results for growth on YPD . To ensure that the results are robust to the statistical approach used , we also performed a student's t-test comparison of a moving window of six probes from the background-subtracted data for the two wildtype and the two mutant replicates for each selection . We set the threshold at p value<2 . 78×10−6 ( Bonferroni-corrected p<0 . 05 ) . Peak calling and all other statistical analyses were conducted in R ( http://www . r-project . org/ ) . For growth rate in diamide and hydrogen peroxide , an ANOVA of the formwas performed in R using the lm function . BG stands for genetic background , which can be either BY or RM . sup35 stands for the allelic status of SUP35 , which can be either wildtype or sup35C653R . SKY1 stands for the allelic status of SKY1 , which can be either BY allele or RM allele . To test whether SKY1 mRNA level could account for SKY1 allelic effect , we first used an ANOVA of the formwhere mRNA stand for SKY1 mRNA level and then used the residuals in an ANOVA of the form mentioned above . Each quantitative RT-PCR measurement represents data collected from three biological replicates . We harvested cells from the logarithmically growing strains in the mentioned growth conditions . We then used total RNA extraction kit ( Norgen; 17200 ) to extract total RNA from collected cells . To perform quantitative RT-PCR , we used TaqMan RNA-to-CT one-step kit ( Applied Biosystems; 4393463C ) and 7900HT Fast Real-Time PCR System ( Applied Biosystems; 4329001 ) . We used TaqMan TAMRA probes . We used a 6-FAM labeled probe for SKY1 detection and a VIC labeled probe for TDH2 detection ( internal control ) . The primers were selected so that there would be no polymorphisms in the sequence amplified for SKY1 ( 185-base pair fragment starting at 403rd nucleotide of SKY1 coding sequence ) and TDH2 ( 168-base pair fragment starting at 379th nucleotide of TDH2 coding sequence ) . The sequences of primers and probes used are as follows: SKY1-F: ATGTGACGAAAGGAACGAAGA SKY1-R: ACTAAAATGTAGCGTGCATCCTT SKY1-probe: TCTTTGAAAGATTACAGGCCGGGTG TDH2-F: AGGTTGTCATCACTGCTCCAT TDH2-R: GTGGTACAAGAAGCGTTGGAA TDH2-probe: CCAATGTTCGTCATGGGTGTTAACG | Proper termination is an important step in a successful mRNA translation event . Many factors , employing genetic and epigenetic mechanisms , are involved in modifying translation termination efficiency in the budding yeast , Saccharomyces cerevisiae . [PSI+] , the prion conformation of Sup35p , one of the translation termination factors in yeast , provides an example of such mechanisms . [PSI+] increases readthrough of stop codons . This has the potential to unveil hidden genetic variation that may enhance growth in some yeast strains in certain environments . The specific details of readthrough-dependent phenotypes , however , have remained poorly understood . Here , we used a partial loss-of-function allele of SUP35 , which increases readthrough of stop codons , and a recently developed linkage mapping technique , X-QTL , to map loci underlying readthrough-dependent growth phenotypes in two divergent yeast strains , BY ( a laboratory strain ) and RM ( a wine strain ) . We found that readthrough-dependent growth phenotypes are often complex , with multiple loci influencing growth . We also showed that variants in the gene SKY1 underlie one of the loci detected for readthrough-dependent growth phenotypes in the presence of two chemicals that induce oxidative stress . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology"
] | 2012 | Genetic Basis of Hidden Phenotypic Variation Revealed by Increased Translational Readthrough in Yeast |
RNA viruses induce specialized membranous structures for use in genome replication . These structures are often referred to as replication organelles ( ROs ) . ROs exhibit distinct lipid composition relative to other cellular membranes . In many picornaviruses , phosphatidylinositol-4-phosphate ( PI4P ) is a marker of the RO . Studies to date indicate that the viral 3A protein hijacks a PI4 kinase to induce PI4P by a mechanism unrelated to the cellular pathway , which requires Golgi-specific brefeldin A-resistance guanine nucleotide exchange factor 1 , GBF1 , and ADP ribosylation factor 1 , Arf1 . Here we show that a picornaviral 3CD protein is sufficient to induce synthesis of not only PI4P but also phosphatidylinositol-4 , 5-bisphosphate ( PIP2 ) and phosphatidylcholine ( PC ) . Synthesis of PI4P requires GBF1 and Arf1 . We identified 3CD derivatives: 3CDm and 3CmD , that we used to show that distinct domains of 3CD function upstream of GBF1 and downstream of Arf1 activation . These same 3CD derivatives still supported induction of PIP2 and PC , suggesting that pathways and corresponding mechanisms used to induce these phospholipids are distinct . Phospholipid induction by 3CD is localized to the perinuclear region of the cell , the outcome of which is the proliferation of membranes in this area of the cell . We conclude that a single viral protein can serve as a master regulator of cellular phospholipid and membrane biogenesis , likely by commandeering normal cellular pathways .
Myriad cellular mechanisms exist to thwart viral infection [1–4] . These mechanisms are triggered when a cellular pattern recognition receptor ( PRR ) engages a virus-associated molecular pattern , for example 5’-triphosphorylated RNA , the absence of 2’-O-methylation of the mRNA cap , double-stranded RNA , among many others [1–4] . PRRs are located at every portal of viral entry into a cell but are particularly abundant in the cytoplasm , the site of replication of most RNA viruses , especially positive-strand RNA viruses . RNA viruses have evolved multiple mechanisms to escape host innate immunity [1–4] . Some mechanisms are specific , for example the use of virus-encoded protein ( s ) to bind and/or to degrade a PRR [1–4] . One generic approach exploited by positive-strand RNA viruses may be the use of a replication organelle for genome replication , which limits surveillance by cellular antiviral defenses [5] , although the need to evade host defenses in cell culture may not be absolute [6] . Virus-induced replication organelles , also referred to as replication complexes , are apparent in cells infected by positive-strand RNA viruses within a few hours post-infection [7 , 8] . Some viruses remodel existing membranes . For example , Flaviviruses ( Dengue virus , West Nile virus and Zika virus ) induce invaginations of negative curvature in membranes of the endoplasmic reticulum ( ER ) that appear as vesicle packets or spherules [9] . Alphaviruses ( Sindbis virus and chikungunya virus ) induce similar structures but use membranes of endosomes or the lysosome instead [10] . In contrast , hepacivirus ( hepatitis C virus , HCV ) and picornaviruses ( poliovirus , PV; Coxsackievirus B3 , CVB3; human rhinovirus HRV; and foot-and-mouth disease virus , FMDV ) use organellar or vesicular membranes to induce protrusions of positive curvature that interact to form a distinct , virus-induced entity [11–14] . The creation of sites for genome replication that are only permeable to small molecules creates a dilemma for trafficking of viral proteins to these sites , given the expectation that viral proteins are produced in the cytoplasm . Therefore , production and/or trafficking of viral proteins and formation of the replication organelle need to be coordinated . For years , it was presumed that a combination of interactions between viral proteins and between viral and host proteins would be essential to this coordination [15] . However , it became clear several years ago that the phosphoinositide , phosphatidylinositol-4-phoshate ( PI4P ) , is enriched in the picornavirus and hepacivirus replication organelles [16] . This discovery inspired the hypothesis that PI4P contributes to recruitment of viral and cellular proteins to the replication organelle [16] . Phosphoinositides have a well-established role in cellular protein trafficking and in coupling activation of protein function to phosphoinositide binding [17] . The RNA-dependent RNA polymerases ( RdRps ) from PV and CVB3 have been reported to bind to PI4P , consistent with this role during infection [16] . PI4P is enriched in the Golgi apparatus ( Golgi ) [18] . A phosphatidylinositol ( PI ) -4 kinase ( PI4K ) produces PI4P from PI . At the Golgi , the type IIIβ and type IIα PI4Ks are used [18] . PI4KIIIβ ( referred to herein by its gene name , PI4KB ) is activated by a pathway in which the guanine-nucleotide exchange factor ( GEF ) , Golgi-specific brefeldin A-resistance GEF ( GBF1 ) , is recruited to Golgi membrane by an ill-defined mechanism and converts the GDP-bound form of ADP-ribosylation factor 1 ( Arf1 ) to its active GTP-bound form [19] . Arf1 activation is a requirement for activation of PI4KB; however , the steps between Arf1 and PI4KB activation are not clear [19] . Further complicating our understanding of steps succeeding Arf1 activation is the fact that numerous Golgi functions are controlled by effectors whose recruitment to membranes also rely on Arf1 activation using ostensibly different and incompletely elaborated mechanisms [20 , 21] . The mechanism ( s ) used by viruses to induce PI4P remains a topic of active research . There is good evidence for picornaviruses and hepaciviruses that either PI4KB or PI4KIIIα ( referred to herein by its gene name , PI4KA ) can be used [22] . In addition , studies of infected cells have implicated GBF1 and Arf1 as contributors to PI4P induction [16 , 23] . The most detailed mechanistic studies have been performed using PV and CVB3 . The earliest studies concluded that the picornaviral 3A protein recruits PI4KB to membranes [16 , 24 , 25] . However , more recent studies have suggested that PI4KB recruitment to membranes by 3A uses a mechanism independent of GBF1 and Arf1 [26–29] . These observations have been interpreted to mean that picornaviruses have evolved a mechanism to recruit PI4 kinases to membranes for production of PI4P that is unique relative to that used by the cell . Here we show that ectopic expression of a single picornaviral protein , the 3CD protein , is sufficient to induce PI4P in cells . PI4P induction appears to reflect the capacity of 3CD to hijack and constitutively activate the normal PI4P biogenesis pathway of the Golgi . GBF1 , Arf1-GTP and activated PI4KB are required for PI4P induction . Amino acid substitutions in the 3D domain of 3CD ( 3CDm ) interfere with steps pre-Arf1 activation; those in the 3C domain of 3CD ( 3CmD ) interfere with steps post-Arf1 activation . Co-expression of 3CDm and 3CmD restore PI4P induction , consistent with 3CD functioning in two discrete steps . The steps targeted by 3CD appear to be those understood least in the cellular pathway . Interestingly , 3CD binds to PI4P-containing membranes but not to membranes containing only phosphatidylcholine ( PC ) . Both 3CD derivatives retain the capacity to bind to PI4P-containing membranes . In addition to PI4P , expression of 3CD also induces phosphatidylinositol-4 , 5-bisphosphate ( PIP2 ) and PC in cells . Neither 3CDm nor 3CmD prevented induction of PIP2 or PC , suggesting distinct mechanisms for induction of these phospholipids relative to PI4P . Lipid induction by 3CD leads to a substantial proliferation of membranes in the perinuclear region of the cell . In summary , we have discovered that a single viral protein is sufficient to commandeer multiple phospholipid biosynthetic pathways and create new membranes for use by the virus in the genesis of its replication organelle . Elucidation of the mechanisms used by 3CD to induce membrane biogenesis may reveal non-genomic mechanisms used by the cell to regulate this process .
Although it is now quite clear that phosphoinositides are exploited by many RNA viruses and mark the sites of genome replication , the viral factor ( s ) contributing to synthesis of these lipids and corresponding mechanisms are largely unknown . The PV genome encodes a polyprotein that can be divided into structural ( P1 in Fig 1A ) and non-structural ( P2 and P3 in Fig 1A ) regions . The P3 region is processed by the protease activity residing in the 3C domain to yield 3AB and 3CD protein as the primary cleavage products ( Fig 1A ) . Several years ago , we reported a PV mutant ( GG PV ) that expresses 3ABC and 3D as the primary cleavage products [30] . This mutant exhibits a delay in the events that lead up to genome replication , as well as a delayed induction of PI4P biosynthesis [30 , 31] . This observation implicated 3AB and/or 3CD in the PI4P-induction process . Expression of PV proteins by DNA transfection was problematic , so we used transcription in vitro to produce RNA that we capped and polyadenylated in vitro to produce mRNAs for 3AB , 3CD , and enhanced green fluorescent protein ( EGFP ) , which served as a negative control . The protease activity of 3CD was inactivated by changing the codon for the catalytic Cys residue to one coding for Gly . We transfected these mRNAs into HeLa cells and used immunofluorescence ( IF ) to monitor protein expression and the fate of PI4P 4 h post-transfection . EGFP expression caused no change in PI4P levels or localization relative to mock-transfected cells ( compare EGFP to mock in Fig 1B ) . 3AB expression caused both a loss of PI4P localization at the Golgi and a reduction in the level of PI4P detected ( 3AB in Fig 1B and 1C ) . 3CD expression caused an increase in PI4P levels , as well as PI4P redistribution ( 3CD in Fig 1B and 1C ) . PI4P levels increased by an average of five fold ( Fig 1C ) . For completeness , we also evaluated 3A and 3CD with an active protease ( 3CDpro+ ) . Like 3AB , 3A expression caused a loss of PI4P localization at the Golgi ( 3A in Fig 1B ) . 3A caused a greater reduction in PI4P detected than 3AB ( compare 3A to 3AB in Fig 1C ) . The addition of protease activity to 3CD caused no discernible difference to induction or redistribution of PI4P ( 3CDpro+ in Fig 1B ) . PI4P is present at the highest level in the Golgi [18] . PV infection causes Golgi dissolution [32] . One mechanism of Golgi loss is inhibition of GBF1 by 3A protein which interferes with replenishment of Golgi membranes by blocking anterograde transport from ERGIC [24 , 33 , 34] . In order to determine if the observed changes to PI4P levels was a consequence of perturbed Golgi integrity , we repeated the experiment described above . We stained cells with an antibody to PI4P to show changes to PI4P and one to Giantin to image the Golgi . Expression of neither 3AB nor 3CD impacted Golgi organization ( Giantin in Fig 1D ) . Expression of 3A , however , caused apparent dissolution of the Golgi as previously reported [24 , 35] . The 3CD protein was clearly localized primarily to the cytoplasm ( Fig 1B ) ; however , it is known that 3CD can enter the nucleus and alter gene expression [36] . Perturbations to gene expression are thought to require 3C-protease activity , which was inactivated for the purpose of this study . Nevertheless , to formally rule out the possibility that 3CD expression perturbs transcription to induce PI4P , we transfected 3CD mRNA into cells treated with actinomycin D ( AMD ) . AMD interferes with transcription by all three cellular RNA polymerases [37] . Disruption of RNA polymerase I by AMD causes redistribution of Nucleolin in the nucleus [38]; we used this phenomenon as a positive control for AMD treatment ( compare anti-Nucleolin -/+ AMD in Fig 2A ) . The presence of AMD did not interfere with 3CD-mediated induction of PI4P ( Fig 2B ) . In order to determine if 3CD-mediated induction of PI4P used aspects of the normal pathway for PI4P production at the Golgi , we measured the impact of brefeldin A ( BFA ) , golgicide A ( GCA ) and PIK93 on PI4P induction . BFA targets at least three cellular GEFs including GBF1 [39] . GCA only inhibits GBF1 [39] . PIK93 inhibits PI4KB [40] . 3CD was unable to induce PI4P in the presence of BFA ( Fig 2C ) or GCA ( Fig 2D ) . The presence of PIK93 reduced the extent to which 3CD was able to induce PI4P ( Fig 2E ) by more than two fold ( Fig 2F ) . The preceding experiment suggests that 3CD hijacks the normal cellular pathway , which includes Arf1 activation by converting Arf1-GDP to Arf1-GTP . In order to measure intracellular levels of Arf1-GTP , we used a GGA3 ( Golgi associated , gamma-adaptin ear containing , Arf-binding protein 3 ) pull-down assay ( Fig 3A ) [41] . GGA3 is a clathrin adaptor protein that selectively binds to Arf-GTP [41] . We validated the assay using PV-infected cells . Infection caused an increase of Arf1-GTP ( S1A Fig ) of seven fold ( S1B Fig ) . Expression of 3CD alone caused an increase in Arf1-GTP ( Fig 3B ) of four fold ( Fig 3C ) . The biochemical and pharmacological studies point to the hijacking of Arf1 , GBF1 , and PI4KB for induction of PI4P biosynthesis . All three of these proteins are present in abundance at the Golgi ( anti-Arf1 , anti-GBF1 , and anti-PI4KB in column marked Mock in Fig 4A and 4B ) . Expression of 3CD caused redistribution of all three of these proteins ( anti-Arf1 , anti-GBF1 , and anti-PI4KB in column marked 3CD in Fig 4A and 4B ) . Arf1 and GBF1 now appeared as speckles in the perinuclear region of the cell ( anti-Arf1 and anti-GBF1 in column marked 3CD in Fig 4A and 4B ) . PI4KB was more diffusely distributed throughout the cytoplasm ( anti-PI4KB in column marked 3CD in Fig 4A and 4B ) in a pattern that closely resembled that of 3CD ( merge in column marked 3CD in Fig 4A ) . In no case did the co-localization of Arf1 , GBF1 , or PI4KB with PI4P appear as strong in the presence of 3CD as observed in its absence ( Fig 4B ) . Importantly , redistribution of Arf1 , GBF1 , and PI4KB was not caused by changes to Golgi integrity ( anti-Giantin in Fig 4C ) or ER integrity ( anti-Calnexin in Fig 4C ) . We compared the changes to Arf1 , GBF1 , and PI4KB localization caused by 3CD to corresponding changes caused by infection . Here we took advantage of our observation that infection in the presence of the replication inhibitor , guanidine hydrochloride ( GuHCl ) , is sufficient to induce to PI4P and membrane biogenesis [42] . Because viral proteins do not accumulate to a detectable level by the indirect immunofluorescence assay in the presence of GuHCl ( anti-3D in column marked PV in Fig 5A ) , we used relocalization of PI4P to distinguish uninfected cells from infected cells ( compare anti-PI4P in column marked Mock to column marked PV in Fig 5A and 5B ) . As observed for 3CD expression , PV infection caused redistribution of Arf1 , GBF1 , and PI4KB ( anti-Arf1 , anti-GBF1 , and anti-PI4KB in column marked PV in Fig 5A ) without an impact on integrity of the Golgi ( anti-Giantin in column marked PV in Fig 5B ) or ER ( anti-Calnexin in column marked PV in Fig 5B ) . For completeness , we repeated this experiment in the absence of GuHCl ( S2 Fig ) . These experiments reinforced the observations made above in the presence of GuHCl . For years , our laboratory has been interrogating the structure , function and dynamics of PV 3C and 3D proteins , creating a large bank of recombinant PVs with mutations in the corresponding genes . We screened this collection to identify 3CD genes that failed to induce PI4P and identified several . Changing both Leu-630 and Arg-639 in the 3D domain to Asp ( referred to throughout as 3CDm ) interfered with PI4P induction ( Fig 6A and 6B ) . Changing Lys-12 in the 3C domain to Leu ( referred to throughout as 3CmD ) also interfered with PI4P induction ( Fig 6A and 6B ) . If each domain functions independently , for example at distinct steps of the PI4P-induction pathway , then each may complement the other . To test this possibility , we co-expressed 3CDm and 3CmD . PI4P induction was restored ( Fig 6A and 6B ) . The pathway for PI4P induction is bisected at the Arf1-activation step . Therefore , we used the GGA3 pull-down assay to determine where in the process of PI4P induction each 3CD derivative was impaired . 3CDm failed to activate Arf1 , suggesting a defect at a step preceding Arf1 activation ( Fig 6C and 6D ) . In contrast , 3CmD activated Arf1 , consistent with a defect at a step succeeding Arf1 activation ( Fig 6C and 6D ) . Introduction of 3CDm and 3CmD into a PV subgenomic replicon that expresses firefly luciferase completely abolishes genome replication as measured by luciferase activity ( Fig 6E ) . If the observed replication defect is related to functions of 3CD required for formation of the RO , then we should be able to rescue this defect by providing a wild-type copy of 3CD in trans . However , if the defect is related to other genome-replication functions of 3CD , then we should be unable to rescue [43 , 44] . We delivered the “wild-type” 3CD in the form of a mutant in which the catalytic site of the RdRp has been inactivated by changing the signature GDD sequence to GAA . As expected , the GAA replicon was dead ( Fig 6E ) . Co-transfection of the replicons expressing 3CDm or 3CmD and GAA showed complementation of only 3CmD , although both the kinetics and yield of luciferase activity were reduced relative to the bona fide wild-type replicon ( Fig 6E ) . The inability of GAA to complement 3CDm was anticipated because the amino acid substitutions in the 3D domain of 3CDm interferes with recruitment of the polymerase to the site of genome replication [45] . The combination therefore should be unable to perform genome replication . Co-transfection of 3CDm and 3CmD should not have this problem . Each of these two defective replicons complemented the other ( Fig 6E and 6F ) , consistent with defects to induction of PI4P biosynthesis contributing to the inability of these replicons to replicate . While we only identified one 3CDm variant that incapacitated PI4P production , we identified two additional 3CmD variants incapable of PI4P induction ( S3A and S3B Fig ) . These 3CmD variants changed Arg-13 to Leu ( referred to as R13L ) or Arg-84 to Leu ( referred to as R84L ) . Both variants were capable of Arf1 activation ( S3C and S3D Fig ) and could be complemented to varying degrees by the GAA replicon ( S3E Fig ) . The efficiency with which replication was restored by the GAA replicon could be ordered as follows: R13L > K12L > R84L . Expression of both 3AB and 3A diminished the abundance of PI4P in cells ( Fig 1C ) . The mechanisms used by 3AB and 3A to interfere with PI4P levels may be distinct , given the inability of 3AB ( column 3AB in Fig 1D ) to interfere with Golgi integrity but the ability of 3A to do so ( column 3A in Fig 1D ) . Regardless of the mechanism , however , the capacity for 3CD to overcome this inhibition is paramount to the mechanism of 3CD activation of PI4P biosynthesis being relevant to the mechanism of PV activation of PI4P biosynthesis . Cells co-expressing 3CD and 3AB produced as much PI4P as cells expressing 3CD alone ( compare 3CD to 3CD+3AB in Fig 7A and 7B ) . In stark contrast , cells co-expressing 3CD and 3A failed to support induction of PI4P biosynthesis ( compare 3CD to 3CD+3A in Fig 7A and 7B ) . Together , these observations suggest unexpected functional differences between 3AB and 3A that give rise to different outcomes for PI4P biosynthesis . The ability of 3CD to overcome the action of 3AB to induce PI4P biosynthesis is consistent with the scenario in the infected cell , because 3AB is the predominant species [46] . Recently , we developed an assay to study the interaction of phosphoinositide-binding proteins with membranes using a supported-lipid bilayer prepared in the channels of a microfluidic device [47] . A fluorescence probe in the bilayer senses protein binding , causing fluorescence quenching for the proteins used to date . The system was validated using the Pleckstrin homology ( PH ) domain from phospholipase C δ1 that binds to PIP2 [47] and has now been validated for use with PI4P as well [48] . The first study demonstrating PI4P induction showed that the 3D domain alone was able to bind to immobilized PI4P [16] . It therefore seemed reasonable to expect that 3CD would also bind to PI4P . To test this possibility formally , we used the chip-based assay to evaluate 3CD binding to PC-based membranes in the absence or presence of PI4P . 3CD bound to the membrane only when PI4P was present ( Fig 8 ) . Interestingly , both 3CDm and 3CmD retained the capacity to bind to PI4P-containing membranes , suggesting that the ability to be recruited to the appropriate membranes remained intact . A statistically significant difference in the binding affinity for both derivatives relative to wild-type 3CD was noted ( Fig 8 ) . 3CmD exhibited two-fold higher affinity binding; 3CDm exhibited two-fold lower affinity binding ( Fig 8B and 8C ) . If and how these changes to PI4P-binding affinity correlate to the observed phenotype in cells requires further investigation . Given the completely unexpected capacity of 3CD alone to induce production of PI4P , we decided to evaluate the fate of other phospholipids in PV-infected cells and determine the extent to which 3CD contributes to their production . Immunological reagents are commercially available for the following phosphoinositides: PI3P , PI3 , 4P2 , PI3 , 5P2 and PI4 , 5P2 ( PIP2 ) . In addition to PI4P , PV infection also induces production of PC [49 , 50] . We therefore evaluated synthesis of PC here as well . Of the four additional phosphoinositides tested , PIP2 showed the most robust , reproducible difference in PV-infected cells relative to mock-infected cells ( Fig 9A ) . We observed a near 7-fold increase in PIP2 ( Fig 9A ) . In the uninfected cell , PIP2 localized predominantly to the plasma membrane . This staining can only be observed at 0°C [51] . We used staining of the plasma membrane to validate the specificity of our anti-PIP2 antibody ( S4 Fig ) . As previously reported , PV infection elevated PC levels by 3-fold relative to control ( Fig 9B ) . We evaluated PIP2 and PC levels in 3CD-transfected cells . We observed a 5-fold increase in PIP2 ( panel 3CD in Fig 9C ) and a 2-fold increase in PC ( panel 3CD in Fig 9C ) in the presence of 3CD . Like induction of PI4P , 3CD-mediated induction of PIP2 and PC occur by post-transcriptional mechanisms as AMD did not impact levels of these phospholipids ( S5A and S5B Fig , respectively ) . Next , we determined if the genetic requirements for PIP2 and PC induction were the same as those for PI4P . 3CDm caused a small but statistically significant ( P <0 . 0001 ) 1 . 7-fold decrease in PIP2 levels relative to that observed for wild-type 3CD ( compare 3CDm to 3CD in Fig 9C ) . None of the 3CmD variants caused a difference in PIP2 levels relative to wild-type 3CD ( Figs 9C and S5C ) . Neither the 3CDm nor the 3CmD variants impacted the outcome of 3CD-mediated induction of PC ( Figs 9D and S5D ) . Induction of so many different phospholipids by 3CD begs the question: where do they all go ? To address this question , we transfected cells with EGFP mRNA as a negative control or 3CD mRNA and compared the ultrastructure of the cell as observed by standard transmission electron microscopy . We observed a normal ultrastructure in EGFP-transfected cells ( panel EGFP in Fig 10 ) . The Golgi was easily identifiable in the perinuclear region and unremarkable ( enlargement of panel EGFP in Fig 10 ) . In contrast , we observed hypertrophy of membranes in the perinuclear region of 3CD transfected cells ( panel 3CD in Fig 10 ) , so much so that the Golgi stacks were no longer distinguishable from the layers of 3CD-induced membranes ( enlargement of panel 3CD in Fig 10 ) . Clearly , Golgi was present and intact based on Giantin staining ( Fig 1D ) . The ability of 3CDm and 3CmD to induce PC ( Fig 9D ) suggests that these derivatives also supported membrane biogenesis . This suggestion was confirmed by TEM of HeLa cells expressing each derivative ( panels 3CDm and 3CmD in Fig 10 ) . The ultrastructure of the membranes produced in 3CmD-expressing cells was different than observed for 3CD or 3CDm . In particular , the membranes exhibited a tubular organization and were not as clustered and organized in appearance ( compare enlargement of panel 3CmD to others in Fig 10 ) .
Phosphoinositides distinguish one organelle from another , couple activation of protein/enzyme function to appropriate cellular localization , and , along with small GTPases and their effectors , enable directional trafficking of proteins and membranes in the cell [17] . In the context of this long-established paradigm of cell biology , the discovery that picornaviruses use PI4P to target proteins to sites of genome replication was logical but also stunning [16] . Since this transformative discovery was made , many laboratories have strived to fill in the gaps between virus entry and formation of the PI4P-rich replication organelle . In the cell , PI4P biogenesis begins with GEF recruitment to the membrane to produce Arf-GTP , which , in turn , recruits effectors that ultimately lead to the recruitment and/or activation of a PI4 kinase [19] . In the specific case studied here , how GBF1 is recruited to membranes and the number and identity of the factors between Arf1-GTP and PI4KB is unclear . However , our studies invoke a requirement for 3CD instead of other P3-derived proteins ( Fig 11A and 11B ) . Conventional wisdom for picornavirology has been that recruitment of the kinase alone or in complex with an unconventional host factor by the picornaviral 3A protein is sufficient to remodel the lipid composition of the replication organelle , formation of which must occur by unrelated mechanisms and directed by other viral factors [16 , 25–29] ( Fig 11C and 11D ) . Our study challenges this paradigm by showing that a single viral protein , the 3CD protein , is sufficient to promote membrane biogenesis in the area of the cell between the ER and Golgi , the so-called ER-Golgi intermediate compartment ( ERGIC ) ( Fig 10 ) . These membranes contain the expected PC ( Fig 9B and 9D ) and are enriched in PI4P ( Fig 1 ) and PIP2 ( Fig 9A and 9C ) . Importantly , 3CD-induced changes to levels of PC , PI4P , and PIP2 appear to be a deliberate process instead of a transcriptional response to stress ( Figs 2A and S5A and S5B ) . The mechanism used by 3CD to induce PI4P is the GBF1•Arf1-GTP•PI4KB pathway ( Figs 2C–2E , 3 and 4 ) , the normal cellular pathway instead of a virus-contrived pathway . Connections between 3CD and this pathway were made years ago in cell-free systems [52]; however , the field was enthralled by the road less traveled and pursued mechanisms dependent on 3A . Our genetic analysis suggests that 3CD functions both before and after Arf1 activation ( Fig 6 ) . The 3D domain contributes to events before Arf1 activation; the 3C domain contributes to events after Arf1 activation ( Fig 6 ) . We propose that 3CD hijacks the pathway by infiltrating its two black boxes: ( 1 ) recruitment of GBF1 to membranes; ( 2 ) recruitment of the effector ( s ) responsible for PI4KB activation ( Fig 11B ) . Both of these recruitment functions of 3CD would benefit from the ability of 3CD to bind to membranes , an activity reported here to be dependent on PI4P ( Fig 8 ) . Identification of the 3CD-interacting factors responsible for PI4P induction may illuminate the mechanism and regulation of the cellular pathway . The ability of 3AB to promote loss of PI4P from the Golgi ( Fig 1B ) without interfering with Golgi integrity ( Fig 1D ) was a surprise . It is known that 3A interacts with GBF1 , and the suggestion has been made that the 3A-GBF1 interaction inhibits anterograde transport from ERGIC to Golgi , thus promoting dissolution of the Golgi ( Fig 11D ) [24] . For this outcome to occur , transport from ERGIC would have to be impacted instead of transport within the Golgi . Our assumption was that 3A and 3AB functions overlapped . They do not . 3A also promotes loss of PI4P from the Golgi ( Fig 1B ) , but this loss likely reflects disruption of Golgi integrity ( Figs 1D and 11D ) . It is known that 3A interacts with membranes differently than 3AB ( Fig 11C ) [53] . This physical difference may explain the functional difference observed here . Levels of 3AB in the infected cell greatly exceed levels of 3A ( Fig 11A ) ; therefore , functions of 3A may appear only at late times post-infection as the amount of 3A in the cell rises [46] . What we observe perhaps for 3AB is inhibition of GBF1 or PI4KB at the Golgi , the consequence of which may only be inactivation of the normal pathway for maintenance of PI4P levels ( Fig 11E ) . Such a mechanism would have the advantage of depleting all PI4P that could misdirect localization of viral proteins , and simultaneously make all GBF1-mediated PI4P synthesis strictly dependent on 3CD . Consistent with such a model is the observation that 3CD is able to overcome the inhibitory action of 3AB on PI4P levels ( Figs 7 and 11E ) , but not the action of 3A on PI4P levels ( Fig 7 ) . 3AB was much more efficient at clearing PI4P from the Golgi than the PI4KA/B inhibitor , PIK93 ( compare PI4P in the 3AB column of Fig 1B to PI4P in the +PIK93 column of Fig 2E ) . This difference may reflect the fact that PI4P levels at the Golgi are maintained not only by PI4KB but also by PI4KIIα [18 , 54] . Indeed , PI4KIIα is responsible for as much as 50% of the PI4P found in cells [54] . This PI4 kinase is not a reported target of PIK93 , but our data would suggest that it is a target of 3AB . Several studies have been published recently reporting that enteroviruses can adapt to growth in the presence of PIK93 [6 , 55 , 56] . The suggestion of one group is that induction of PI4P biosynthesis is not absolutely essential for enterovirus genome replication . As discussed above , type III PI4 kinases may not be the only kinases contributing to PI4P biosynthesis as PIK93 is not able to completely abolish induction of PI4P biosynthesis by 3CD ( Fig 2E and 2F ) . Resistance to PIK93 maps to 3A-coding sequence , and several alleles have been described [6 , 55 , 56] . These substitutions dysregulate processing of 3AB , leading to production of elevated levels of 3A [6 , 55 , 56] . Importantly , one group suggests that loss of 3AB correlates with increased levels of PI4P in infected cells in the presence of PIK93 [55] . It is possible that 3A and/or the 3AB derivatives with PIK93-resistance-conferring substitution are unable to interfere with PI4P biosynthesis as well as wild-type 3A or 3AB ( Fig 1B and 1D ) . If this is the case , then resistance to PIK93 does not demonstrate the absence of a need for PI4P but the ability of the virus to shift from utilizing a type III PI4 kinase to another class of PI4 kinases for PI4P biosynthesis . Both the observation that PV infection induced PIP2 ( Fig 9A ) and the observation that 3CD alone was sufficient for induction of PIP2 ( Fig 9C ) were surprises . The existence of plasma membrane-localized pools of PIP2 have been known and studied for a long time [57] . More recently , it has become clear that PIP2 is contained within many intracellular compartments , including ER , Golgi , endosomes and lysosome , and regulates intracellular membrane trafficking , including endolysosomal and autophagic vesicles [57] . Binding of PIP2 to ATG14 , a component of the phagophore-inducing PI3 kinase complex , promotes initiation of autophagy [58] . Autophagy contributes substantively to the multiplication of picornaviruses , especially enteroviruses like poliovirus [59–68] . This pathway is thought to be responsible for the non-lytic egress of virus particles . Perhaps 3CD-induced PIP2 during PV infection contributes to formation , function and/or trafficking of the phosphatidylserine-enriched phagophores that engulf PV virions to produce vesicles thought now to be important for viral transmission [69] . Much remains to be learned about PIP2 synthesis during PV infection . PIP2 can be produced from PI4P by a PI4P-5-kinase ( PIP5K ) or from PI5P by a PI5P-4-kinase ( PIP4K ) [17] . There are multiple genes for each class; alternative splicing of PIP5K transcripts creates even more isoforms [17] . PIP2-producing kinases are even less well understood than the PI4P-producing kinases [17] . However , PIP2-producing kinases appear to be activated by small GTPases that are regulated by GEFs . Therefore , it is reasonable to propose that the role of 3CD in the mechanism of PIP2 induction will be conceptually similar to that reported here for PI4P induction . 3CD may act upstream of a GEF and downstream of an activated small GTPase to recruit and/or activate the PIP2-producing kinase . The mechanism remains to be defined . The question of which class of kinases is used will need to be addressed by future experiments . Our observation that 3CD derivatives that fail to induce synthesis of PI4P ( Figs 6 and S3 ) retain the ability to induce synthesis of PIP2 ( Figs 9 and S5 ) suggest that 3CD-directed synthesis of PI4P is not required for PIP2 synthesis . If PI4P is the substrate , then a PIP5K ( also referred to as a type II PIPK ) would be required . The α and β isoforms of PIP5K reside in the nucleus , and the γ isoform resides in the ER [17 , 70] . Any or all of these isoforms are feasible candidates in the context of infection because PV infection actually leads to dysregulation of nuclear import and breakdown of the nuclear envelope [71] . In 3CD-transfected cells , it is conceivable that nuclear forms could also be hijacked . 3CD enters the nucleus and could establish sites of PIP2 production at sites of contact between the nucleus and ER [36] . PI4P is also required for multiplication of hepatitis C virus ( HCV ) [72–74] . Non-structural protein 5A ( NS5A ) is required for maintenance of high levels of PI4P in Huh-7 cells and sublines thereof replicating HCV RNA [75–77] . This activity is mediated by an interaction between NS5A and PI4KA [76] . Inhibitors disrupting the NS5A-PI4KA interaction are potent inhibitors of HCV multiplication in cell culture and in vivo [75 , 76] . PIP2 is required by other viruses , although there is a paucity of reports of virus-induced synthesis of PIP2 . Binding of PIP2 to NS5A has been shown to promote interactions between NS5A and a host factor required for genome replication [78] . In most other cases , PIP2 is used for some aspect of virion assembly . For example , PIP2 binding to the matrix protein domain of the retroviral Gag polyprotein sends the protein to the plasma membrane as a step in virion assembly [79] . PIP2 binding to the matrix protein ( VP40 ) of Ebola virus is an essential step in virion assembly at the plasma membrane [80] . PIP2 binding to the NP protein of influenza virus in the context of the viral ribonucleoprotein ( vRNP ) harboring the genome targets the vRNP to the plasma membrane for incorporation into virions [81] . In all of these cases endogenous levels of plasma-membrane associated PIP2 appear sufficient . Use of PIP2 by PV and related viruses for virus assembly fits nicely into this existing paradigm but a mechanism for induction of PIP2 in the cell appears to be required . In a recent study using representative members of the major supergroups of positive-strand RNA viruses of plants and animals , it was demonstrated that PC is induced by all of the viruses studied , including PV [50] . The mechanism of PC induction used by brome mosaic virus involved recruitment of one PC biosynthetic enzyme and activation of another but details are lacking [50] . Nothing is known about how PV might induce PC . Our inclusion of PC in this study initially was as a negative control . To the best of our knowledge , no pathway using a GEF or a small GTPase is known that leads directly to activation of the enzyme governing the rate-limiting step of PC biosynthesis , CTP: phosphocholine cytidylyltransferase ( CCT ) . Recruitment of CCT to membranes may be sufficient for activation of the enzyme , which could be accomplished by direct interaction with 3CD or a CCT-interacting protein [49] . The observation that 3CD expression caused PC induction motivated us to consider the possibility that phospholipid imbalances resulting from PI4P and/or PIP2 induction triggered a stress response that turned on transcription of genes required for membrane biogenesis . However , the resistance of 3CD-mediated induction of PC to AMD is inconsistent with a transcriptional response ( S5B Fig ) . We suggest the existence of a post-transcriptional or post-translational mechanism that permits cells to respond to an acute need for membrane biogenesis . Perhaps studies of 3CD-mediated induction of PC will illuminate the mechanism . How is it that a single , viral protein can have so many functions interfacing with host pathways required for phospholipid metabolism and membrane biogenesis , function at the heart of genome replication by binding to all of the cis-acting replication elements for initiation , and also cleave the viral polyprotein to activate viral function and cleave host factors to hijack host ribosomes and antagonize host innate defenses ? Our studies of 3CD structure and dynamics have revealed a panoply of conformations achieved by rotating the 3C domain relative to the 3D domain . Each of these conformations have been suggested to expand the proteome and provide unique surfaces for interaction with viral and/or host proteins and nucleic acids [82] . The competency to manipulate the 3CD proteome in a deliberate manner would facilitate elucidation of 3CD-interaction network and uncover mechanisms diluted by the ensemble of states . Understanding the mechanisms used by 3CD to induce these profound changes in the cell will likely illuminate regulatory hubs for membrane biogenesis that have undoubtedly not only been usurped by picornaviruses but other families of RNA viruses as well . Poliovirus and its cousins still hold many secrets of mammalian cell biology for future studies to expose .
HeLa cells ( CCL-2 ) were purchased from American type culture collection ( ATCC ) and cultured in Dulbecco’s Modified Eagle Medium: Ham’s nutrient mixture F-12 ( Gibco ) supplemented with 10% heat-inactivated fetal bovine serum ( Atlanta Biologics ) and 1% Penicillin-Streptomycin ( Corning ) . All experiments were performed at 37°C in 5% CO2 . Antibodies for staining PI4P and PI ( 4 , 5 ) P2 were purchased from Echelon Biosciences . Anti-Giantin , anti-Calnexin and anti-GBF1 were from Abcam . Anti-PI4KB and anti-Nucleolin were from BD Biosciences and Novus biologicals , respectively . Anti-Arf1 antibody was a generous gift from Dr . Sylvain Bourgoin at Laval University , Canada . Anti-phosphatidylcholine ( PC ) antibody was provided by Dr . Umeda at Tokyo University , Tokyo . Antibodies against PV 3D and 3AB were produced in Cameron lab . All secondary antibodies used for immunofluorescence were purchased from Molecular Probes . Secondary antibody for Western blotting was purchased from SeraCare . Inhibitors used in this study including brefeldin A ( BFA ) , golgicide A ( GCA ) , PIK93 and actinomycin D ( AMD ) were purchased from Sigma-Aldrich . Plasmids used for ectopic expression of proteins were constructed by amplifying the gene of interest using pRLuc-RA plasmid as template and using oligonucleotide pairs provided in S1 Table . The resulting PCR product was ligated into the pIRES vector ( Clonetech ) backbone between the EcoRI and NotI unique sites . This caused a loss of the IRES element from the pIRES . To avoid confusion with nomenclature , these cloned plasmids were designated pSB vectors . For making pSB-EGFP , the pEGFP-N1 plasmid was digested with EcoRI and NotI . The resulting EGFP fragment was then ligated into the pIRES backbone between the EcoRI and NotI unique sites . Quickchange mutagenesis was used to construct the mutant 3CD genes carrying point mutations in 3C and 3D using the oligonucleotide pairs provided in S1 Table . In order to inactivate 3C protease activity , the catalytic cysteine ( C147 ) was converted to glycine . The clones were verified by sending to the sequencing facility at The PSU Genomics Core Facility . Subgenomic luciferase-encoding replicons harboring point mutations were constructed by overlap extension PCR using the oligonucleotides listed in S1 Table . For both reactions , pRLuc-HpaI-SacII was used as the template [31] . The resulting overlap PCR fragment was digested and ligated in between the HpaI and SacII sites of the pRLuc-HpaI-SacII vector to produce the pRLuc plasmids with desired mutations . Linearization of cDNA and transcription was performed as described previously [30] . Briefly , the pSB- and pRLuc- plasmids were linearized with NotI or ApaI and purified with QIAEX II suspension ( QIAGEN ) following the manufacturer’s protocol . RNA was then produced using the linearized plasmid in a 20 μL reaction containing 350 mM HEPES pH 7 . 5 , 32 mM magnesium acetate , 40 mM dithiothreitol ( DTT ) , 2 mM spermidine , 28 mM nucleoside triphosphates ( NTPs ) , 0 . 025 μg/μL linearized DNA , and 0 . 025 μg/μL T7 RNA polymerase . The reaction mixture was incubated for 3 h at 37°C and magnesium pyrophosphate was removed by centrifugation for 2 min . The supernatant was transferred to a new tube and subjected to RQ1 DNase treatment for 30 min at 37°C . RNA quality was verified by agarose gel ( 0 . 8% ) electrophoresis . The in vitro transcribed pSB-RNAs were purified using the RNeasy kit ( QIAGEN ) using the manufacturer’s protocol and subjected to modification by adding a 5’-end cap and a 3’-end poly ( A ) tail using the T7 mRNA production kit ( Cellscript ) following the manufacturer’s protocol . The modified RNAs were further purified using the RNeasy columns and quality was evaluated on a 0 . 8% agarose gel . The addition of poly ( A ) tail was confirmed by comparing the modified RNA to unmodified RNA . 2 . 5 X105 HeLa cells were seeded on coverslips in 6-well plates and transfected 16 h post-seeding . For pSB-mRNAs , 2 μg of column-purified transcripts were transfected using Transmessenger transfection kit ( QIAGEN ) following the manufacturer’s protocol . For pRLuc-RNAs , transfections were performed using 2 . 5 μg of column purified RNA following the TransIT transfection reagent ( Mirus ) and corresponding protocol . In all cases , cells were incubated at 37°C and subjected to immunostaining 4 h post-transfection . For studies with inhibitors , transfections were performed in presence of the inhibitors: AMD ( 5 μg/mL ) ; BFA ( 2 μg/mL ) ; GCA ( 10 μM ) and PIK93 ( 15 μM ) and the inhibitors were kept on the monolayers for the entire 4 h period . 2 . 5 X 105 HeLa cells were seeded on coverslips in 6-well plates overnight . Cells were infected at confluency with PV at an MOI of 10 , in the presence or absence of 3mM GuHCl . The inhibitor was used on cells when adding the virus for a 30min period as well as the 4-hour incubation immediately after . Cells were fixed 4 h post-transfection using 4% formaldehyde in PBS for 20 min followed by washing in PBS and permeabilizing with 20 μM digitonin for 10 min . Digitonin was washed off with three washes of PBS . Cells were then blocked with 3% BSA in PBS for 1 h and incubated with primary antibodies for 1 h . Following washes cells were incubated with the secondary antibodies for 1 h . For PIP2 membrane staining , the protocol described in Hammond et al . 2009 . [51] The processed coverslips were mounted on glass slides using VECTASHIELD Mounting Medium with DAPI ( Vector Laboratories ) . The mounted coverslips were sealed with nail polish . Samples were imaged using Zeiss Axiovert 200 M epifluorescence microscope . Primary antibodies used were: anti-Giantin ( 1:400 ) for Golgi staining; anti-Calnexin ( 1:500 ) ; anti-Nucleolin ( 1:200 ) for nucleolar staining; anti-PI4P ( 1:200 ) ; anti-PI ( 4 , 5 ) P2 ( 1:200 ) ; anti-PC ( 1:40 ) ; anti-3AB ( 1:400 ) and anti-3D ( 1:100 ) ; anti-Arf1 ( 1:200 ) ; anti-GBF1 ( 1:200 ) ; anti-PI4KB ( 1:100 ) . All dilutions were made in the blocking buffer . For fluorescence intensity quantifications , each image was acquired as a Z-stack file with Z = 11 slices . The sum of intensities was measured using the ZEN 2012 ( blue edition ) software from Zeiss . The sum fluorescence intensity for each cell was normalized against the average of the sum fluorescence intensities of mock cells . BL21-DE3 cells were transformed using a GST-GGA3 expression vector [40] . Transformed cells were plated on NZCYM media containing ampicillin ( 100μg/ mL ) and incubated at 37°C overnight . BL21-DE3 cells harboring the GST-GGA3 expression vector were then grown overnight in 50 mL NZCYM/Amp at 37°C . The culture was back-diluted at a 1:25 ratio in 500 mL NZCYM/Amp media and grown at 37°C until the culture reached 0 . 4–0 . 6 OD600 . Expression of GST-GGA3 was then induced by addition of IPTG to 1 mM and incubated at 37°C for 6 h . The cell pellet was harvested by centrifugation at 6000 rpm for 10 mins . The pellet ( 1 . 8 g ) was suspended in 13 mL ice cold bacterial lysis buffer ( BLB ) containing 50 mM Tris , pH 8 . 0 , 20% sucrose , 10% glycerol , 2 mM DTT , 0 . 1 mM PMSF and 1 μg/mL each of pepstatin and leupeptin . Cells were lysed using a French press ( Glen Mills ) , and cell debris was removed by centrifugation at 25 , 000 rpm for 30 min at 4°C . Glutathione-sepharose beads ( 250 μL ) were equilibrated with BLB , by four sequential washes with 1 mL of BLB . Beads were collected after each wash by centrifugation at 400 X g for 2 min at 4°C . Beads were added to the clarified lysate and incubated at 4°C for 1 h on a rotating shaker . The beads were then pelleted by centrifugation at 200 X g for 30 min and washed six times in 15 mL of BLB . Finally , the beads were suspended in 400 μL of BLB with containing DTT ( 1 mM ) . The protein yield was estimated by Bradford assay using BSA as a standard . Beads were aliquoted ( 40 μg/ tube ) for single use and stored at -80°C . Activation endogenous Arf1 by transfection of mRNA or by infection was monitored using GST-GGA3 pull-down assay described in [41] with minor changes . Cells were plated at a density of 3 X 106 cells in 100 mm dish 24 h prior to transfection or infection . 3CD ( WT or mutant ) mRNA ( 15 . 5 μg ) was transfected into HeLa cells using TransIT-mRNA transfection kit ( Mirus ) . Four hours post-transfection or post-infection cells were lysed at 4°C in 0 . 5 mL of lysis buffer ( 50 mM Tris , pH 7 . 5 , 200 mM NaCl , 10 mM MgCl2 , 0 . 1% SDS , 0 . 5% sodium deoxycholate , 1% triton X-100 , 5% glycerol with 0 . 1 mM PMSF and 1 μg/mL each of pepstatin and leupeptin ) . Lysates were incubated on ice for 5 min in presence of 50 μL of CL-4B Sepharose beads ( per sample ) and then centrifuged for 15 min at 16 , 000 X g and 4°C . The clarified lysate ( 500 μg ) was incubated with 40 μg of GST-GGA3 bound to Glutathione Sepharose beads for 1 h on a rotating shaker at 4°C . The beads were then washed three times with 1 mL cold lysis buffer followed by a final wash in 1 mL cold PBS ( 1X ) . Liquid was removed beads by quick inversion followed by and a 30 s spin at 6 , 000 X g at 4°C and removal of residual buffer using a pipetman . Bound proteins were eluted from beads by adding 30 μL of 2X SDS-PAGE sample buffer and incubating at 65°C for 10 min . Beads were pelleted by centrifugation at 6 , 000 X g for 2 min at room temperature . Twenty μL of the supernatant was resolved by SDS polyacrylamide gel electrophoresis . Endogenous , activated Arf1 was detected using anti-Arf1 antibody . For Total Arf1 was detected using the same antibody , but the sample was prepared from 10 μg of the clarified lysate . Alkaline phosphatase activity associated with the secondary antibody was detected using the ECF reagent ( GE Healthcare ) and quantified by imaging the fluorescence signal using the Syngene G-box . Luciferase assays were performed as described previously [30] with the following modifications . Subgenomic replicon RNA ( 5 μg of in vitro transcribed RNA ) was electroporated into HeLa cells . The cells were incubated in normal growth media ( DMEM/F12 supplemented with 10% fetal bovine serum , 1% penicillin/streptomycin , 5 mL/1 × 106 cells ) and 1 × 105 cells were harvested and lysed using 100 μL of 1X cell culture lysis reagent ( CCLR , Promega ) at the indicated times post-electroporation . Luciferase activity was measured by adding equal volume of firefly luciferase assay substrate ( Promega ) to cell lysate and the reaction mixture was applied to a Junior LB 9509 luminometer ( Berthold Technologies ) to read relative light units ( RLU ) for 10 s . Relative light units ( RLU ) were then normalized based on the total protein concentration determined by Bio-Rad protein assay reagent ( BioRad ) for each sample . HeLa cells were transfected with mRNAs expressing PV-3CD or EGFP and 4 h post-transfection cells were fixed and embedded for TEM studies as described previously [31] . Briefly , cells were harvested and fixed with 1% glutaraldehyde , washed with 0 . 1 M cacodylate ( sodium dimethyl arsenate , Electron Microscopy Sciences ) twice for 5 min each , incubated in 1% reduced osmium tetroxide containing 1% potassium ferricyanide in 0 . 1 M cacodylate for 60 min in the dark with one exchange and washed two times with 0 . 1 M cacodylate again . En bloc staining was performed with 3% uranyl acetate in 50% ethanol for 60 min in the dark . Dehydration was carried out with different concentrations of ethanol ( 50 , 70 , 95 and 100% for 5–10 min ) and 100% acetonitrile . Embedding was performed overnight with 100% Epon at 65°C . The embedded sample was sectioned with a diamond knife ( DiATOME ) to slice 60–90 nm thickness by using ultra microtome ( Reichart-Jung ) . The sectioned sample was placed on copper grid ( Electron Microscopy Sciences ) , stained with 2% uranyl acetate in 50% ethanol followed by lead citrate staining for 12 min . The grid was washed with water and the grid was dried completely . The image was obtained using FEI Tecnai G2 Spirit BioTwin located in The PSU Electron Microscopy Facility . A microfluidic platform was employed to test the interaction between the viral protein and 7 . 5 mol% PI4P SLBs . SLBs containing 99 . 5 mol% POPC and 0 . 5 mol% oSRB-POPE , a pH-sensitive fluorescent probe conjugated to a lipid , served as a negative control . Synthesis of oSRB-POPE was described previously [44] . A PDMS block was fused to a borosilicate glass by use of a plasma-oxygen treatment . The fabrication of the PDMS device was described previously [44] . SLBs were formed on the borosilicate glass of each micro channel by spontaneous fusion and rupturing of the SUVs after treatment with 0 . 2 N HCl [47] . Running buffer ( 20 mM HEPES , 100 mM NaCl , pH 7 . 0 ) was flowed through each channel for 30 min remove excess vesicles . In order to equilibrate the SLBs to the experimental condition , running buffer containing the appropriate dilution of gel filtration buffer ( 20 mM HEPES , 20% glycerol , 1 mM BME , 500 mM NaCl , pH 7 . 0 ) was flowed through the channels . The fluorescence intensity obtained post-equilibration step served as a reference for each channel . PV-3CD dilutions were prepared using the running buffer . PV-3CD was flowed until the fluorescence intensity in each channel stabilized ( 30–45 min ) . Change in fluorescence intensity , normalized to the reference ( no protein ) channel , was plotted as a function of PV-3CD concentration and then fit to a Langmuir isotherm using the equation below . ΔF=ΔFmax∙[3CD]Kd , app+[3CD] where , Fmax represents the normalized fluorescence intensity value at saturation level and Kd , app represents the apparent dissociation constant . Graphs were generated by GraphPad Prism v . 6 software . Error bars represent the SEM . Images were taken with an Axiovert 200M epifluorescence microscope ( Carl Zeiss Microscopy ) equipped with a AxioCam MRm camera ( Carl Zeiss Microscopy ) and X-Cite 120 ( Excelitas Technologies Corp . ) light source was used to take fluorescence images . A 10X air objective was used for imaging along with Alexa 568 filter set ( Carl Zeiss Microscopy ) with an excitation and emission at 576 nm and 603 nm , respectively . Exposure time was set to 200 msec per exposure with minimal times of exposure throughout experiments . AxioVision LE64 v . 4 . 9 . 1 . 0 software ( Carl Zeiss Microscopy ) was used to process the images . pSUMO plasmids encoding 3CDm or 3CmD was transformed into Rosetta ( DE3 ) competent cells [83] . Cells were grown to an OD of 1 . 0 , harvested by centrifugation ( 5 , 400 x g , 10 min , 4°C ) , and washed with 30 mL of buffer ( 10 mM Tris and 1 mM EDTA , pH 8 . 0 ) per liter of bacterial culture , and re-centrifuged . Each gram of cell pellet was suspended in 5 mL of lysis buffer ( 20 mM HEPES , 10% glycerol , 5 mM imidazole , 10 mM β-mercaptoethanol [BME] , 500 mM NaCl , 1 mM EDTA , 1 . 4 μg/mL pepstatin A , 1 . 0 μg/mL leupeptin , pH 7 . 0 ) . Suspended cells were homogenized by a Dounce homogenizer and lysed by two passes through a French press at a pressure of 1 , 000 psi . Phenylmethylsulfonyl fluoride ( PMSF ) was added to the cell lysate at a final concentration of 1 mM . The suspension was clarified by centrifugation ( 74 , 000 x g , 30 min , 4°C ) . The supernatant was loaded onto a Nickel-nitrilotriacetic ( Ni-NTA ) resin ( Thermo Fisher Scientific Inc . ) , which was previously equilibrated with 10 column volumes ( CV ) of equilibration buffer containing 5 mM imidazole at 1 mL/min using a peristaltic pump . The protein load was passed through the equilibrated Ni-NTA resin at 1 mL/min . To remove contaminants , the loaded resin was washed with 50 CV and 4 CV of equilibration buffer ( 20 mM HEPES , 20% glycerol , 10 mM β-mercaptoethanol ( BME ) , 500 mM NaCl , pH 7 . 0 ) containing 5 mM and 50 mM imidazole , respectively . PV-3CD was eluted into multiple fractions using equilibration buffer containing 500 mM imidazole . Fractions were evaluated by Coomassie stained SDS-PAGE ( 10 . 0% ) gel . Concentrated fractions were pooled , treated with 1 μg ubiquitin-like-specific-protease 1 ( ULP-1 ) per 1 mg protein of interest , and dialyzed against gel filtration buffer ( 20 mM HEPES , 20% glycerol , 1 mM BME , 500 mM NaCl , pH 7 . 0 ) overnight at 4°C using a 12–14 kDa MWCO dialysis membrane ( Spectrum Laboratories ) . Dialyzed PV-3CD sample was centrifuged ( 74 , 000 x g , 30 min , 4°C ) to ensure non-aggregated protein , and subsequently concentrated to 2–5 mL for loading onto a gel filtration column ( GE Healthcare ) . The gel filtration column was run at 1 . 0 mL of gel filtration buffer per minute , and divided into 34 3-mL fractions . Fractions with the protein of interest were pooled and concentrated with a Vivaspin Turbo 15 ( Sartorius , 10 kDa MWCO ) . The protein concentration was determined by measuring the absorbance at 280 nm ( εmax = 0 . 08469 μM-1·cm-1 ) with a NanoDrop ( Thermo Fisher Scientific Inc ) . The ionic strength of the dialyzed protein solution was confirmed with a conductivity meter ( Control Company ) . To assess extraneous protein contamination , dynamic light scattering ( DLS ) experiments performed with Viscotek 802 DLS instrument at 20°C . Agarose gel electrophoresis was run to check for any contaminating nucleic acids . Flash frozen aliquots were stored at -80°C . For statistical analysis data were plotted as means ± SEM . The intensity comparisons between control and experimental measurements used an unpaired Student’s t-test in the GraphPad Prism software . The means and P values for pairwise comparisons of all experiments are provided in S2 Table , a subset of these are also presented in the appropriate figure legends . | Picornaviruses replicate their genomes in association with host membranes . Early during infection , existing membranes are used but remodeled to contain a repertoire of lipids best suited for virus multiplication . Later , new membrane synthesis occurs , which requires biosynthesis of phosphatidylcholine in addition to the other more specialized lipids . We have learned that a single picornaviral protein is able to induce membrane biogenesis and decorate these membranes with some of the specialized lipids induced by the virus . A detailed mechanism of induction has been elucidated for one of these lipids . The ability of a single viral protein to commandeer host pathways that lead to membrane biogenesis was unexpected . This discovery reveals a new target for antiviral therapy with the potential to completely derail all aspects of the viral lifecycle requiring membrane biogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"transfection",
"nuclear",
"staining",
"hela",
"cells",
"biological",
"cultures",
"microbiology",
"cell",
"cultures",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"lipids",
"vi... | 2018 | Hijacking of multiple phospholipid biosynthetic pathways and induction of membrane biogenesis by a picornaviral 3CD protein |
Peptide recognition domains ( PRDs ) are ubiquitous protein domains which mediate large numbers of protein interactions in the cell . How these PRDs are able to recognize peptide sequences in a rapid and specific manner is incompletely understood . We explore the peptide binding process of PDZ domains , a large PRD family , from an equilibrium perspective using an all-atom Monte Carlo ( MC ) approach . Our focus is two different PDZ domains representing two major PDZ classes , I and II . For both domains , a binding free energy surface with a strong bias toward the native bound state is found . Moreover , both domains exhibit a binding process in which the peptides are mostly either bound at the PDZ binding pocket or else interact little with the domain surface . Consistent with this , various binding observables show a temperature dependence well described by a simple two-state model . We also find important differences in the details between the two domains . While both domains exhibit well-defined binding free energy barriers , the class I barrier is significantly weaker than the one for class II . To probe this issue further , we apply our method to a PDZ domain with dual specificity for class I and II peptides , and find an analogous difference in their binding free energy barriers . Lastly , we perform a large number of fixed-temperature MC kinetics trajectories under binding conditions . These trajectories reveal significantly slower binding dynamics for the class II domain relative to class I . Our combined results are consistent with a binding mechanism in which the peptide C terminal residue binds in an initial , rate-limiting step .
Protein-protein interactions control numerous processes in the cell . Recently , it has been realized that a significant fraction of these interactions are mediated by the binding of flexible polypeptide segments to folded domains [1]–[3] . This realization is in part due to the discovery of many so-called peptide recognition domains ( PRDs ) , which function specifically by recognizing sets of short peptide sequences [4] , [5] . PRDs often interact with their ligand peptides in a reversible , transient manner , making them particularly well suited to mediate interactions in signaling and regulatory processes , which require fast response to initiated or ceased stimuli . A fundamental understanding of the detailed dynamics and binding free energy landscapes of these PRD-peptide interactions will therefore eventually be necessary in order to understand the finely tuned specificities and affinities which underpin many protein interaction networks . Achieving such an understanding may also be of practical importance , as it can be a starting point towards altering signaling networks in a controlled way [6] , [7] or designing small molecules to inhibit domain-peptide binding [8] , [9] . Modeling peptide binding in atomistic detail is a challenge . One reason for this is the inherent flexibility of a disordered peptide chain which necessitates a statistical mechanical approach . At the same time it is a major modeling opportunity because of the relatively small molecular interface and few amino acids involved , making the peptide binding process computationally accessible . Several docking methods designed specifically for peptide binding have been developed [10]–[16] , which aim to predict the correct peptide binding pose on a protein surface . Most of these methods require some prior knowledge of the peptide binding site , although true blind docking has also been attempted [17] , [18] . Other in silico methods seek to provide binding predictions for whole PRD families , including SH2 [19] , SH3 [20] , [21] , and PDZ [22] domains . These methods rely on structural models of domain-peptide complexes using an available experimental peptide-bound configuration as a template . Most PRD families , however , display significant diversity in how peptides interact with domains , which fundamentally limits this approach . In a recent effort to alleviate this problem , King et al [15] combined peptide docking and subsequent structure-based binding prediction using the Rosetta scoring function . Molecular Dynamics simulations of domain-peptide bound states have also been carried out , emphasizing the importance of dynamics and flexibility for understanding the molecular basis of peptide binding [23]–[25] . Our aim here is to go beyond docking and investigate the binding process from an equilibrium perspective . To this end , we use a recently developed Monte Carlo-based procedure for protein-peptide binding [26] and apply it to three different PDZ domains and their target peptide sequences . The approach combines a global conformational search of the peptide chain , as well as limited protein backbone flexibility around the native state , with an effective energy function inspired by protein folding studies [27]–[29] . Rather than relying on large numbers of docking attempts , we perform fewer but long simulations such that each run exhibits multiple binding and unbinding events , thereby providing an equilibrium picture of the binding process . In particular , this allows us to investigate and compare features of the global binding free energy landscape as determined by the interaction between the protein surface and the amino acid sequence of the peptide . The PDZ domain is an archetypical PRD existing in large numbers in many genomes [30]–[32] . It distinguishes itself from other PRDs in that it typically binds sequence motifs at the extreme C terminal end of proteins . The architecture is mostly conserved across the domain family with a typical core structure consisting of two -helices and six -strands . The PDZ fold includes a binding pocket between the second -helix ( ) and second -strand ( ) such that a ligand peptide can augment the -strand upon binding and pack its sidechains against the -helix . In addition , the peptide C terminus forms hydrogen bonds with the backbone amides of a highly conserved loop on the PDZ domain . Like many other PRD families , PDZ domains have been divided into different classes depending on which peptide sequences they preferentially bind . The most established division of PDZ domains is into classes I , II , and III , corresponding to the sequence patterns Ser/Thr-X--COOH , -X--COOH , and Asp/Glu-X--COOH , respectively , where is any hydrophobic amino acid , X is any amino acid , and COOH is the C terminus [32] . It can be pointed out that more fine-grained classifications are also possible [33] . We focus here on comparing the binding behavior of class I and II domains , which represent the majority of known PDZ domains [30] , [32] . An important aspect of any binding study is the ability to capture binding to free molecules , i . e . , to structures determined in the absence of a ligand . This is important not the least for PDZ domains , for which only domain-peptide complexes have been solved experimentally so far [32] , compared to the almost 200 free PDZ domain structures in the Protein Data Bank ( PDB ) [34] . We therefore start out by testing our computational procedure using two different structural forms of the domains , free and peptide-bound . Thereafter , we describe the conformational transitions of the peptides and the binding free energy landscapes for the domains . Finally , we perform a large number of Monte Carlo based kinetic simulations to obtain a deeper microscopic picture of the peptide binding process .
As class I and class II representatives we chose the 3rd PDZ domain of PSD-95 and the 6th PDZ domain of GRIP1 , respectively . These are typical class I and II PDZ domains in the sense that all known binding peptides fall within their respective ideal class motifs [30] , [35] . Free and peptide-bound X-ray structures have been determined for both domains [36] , [37] , and for PSD-95 the binding thermodynamics [38] as well as kinetics [39] , [40] have been particularly well characterized . The ligands present in the two peptide-bound structures were derived from the C termini of the proteins CRIPT ( PSD-95 ) [36] and human Liprin- ( GRIP1 ) [37] . We consider here the binding of these two ligands to both the bound ( b ) and free ( f ) structural forms and denote the systems by PSD95-Ib , PSD95-If , GRIP1-IIb , and GRIP1-IIf , respectively . In addition to these class I and II domains , we include in this study the PDZ domain of PICK1 which is one of the few known PDZ domains with dual class I and II specificity . The structure of PICK1 PDZ has been determined with class II peptides [41] , [42] . We consider binding of ligands taken from protein kinase ( , class I ) and AMPA receptor subunit GluR2 ( GluR2 , class II ) , which are known binders to PICK1 [43] , [44] , and denote the systems with PICK1-Ib and PICK1-IIb , respectively . The PDZ domains and peptide amino acid sequences under study are summarized in Table 1 . To simulate the domain-peptide binding process , we use the MC based approach developed in Ref . [26] . This simulation procedure is general in that it can in principle be applied to any protein-peptide pair as long as a protein structure is available . Briefly , it works in the following way . A relaxed protein domain structure is centered in a cubic box and joined by a peptide in a random conformation away from the protein surface . The peptide is entirely free to search conformational space , restricted only by periodic boundary conditions on the box . The protein , on the other hand , is kept close to its native structure using constraints on the -atoms , which allow limited backbone and in principle full sidechain flexibility . We combine this simple procedure with an implicit-solvent all-atom energy function based on effective hydrogen bond , electrostatic , and hydrophobic forces [26] . Here we improve the model by including a context-dependent desolvation effect for backbone atoms groups ( see Methods ) . We find , in particular , that including such a context-dependence improves the challenging case of simulating peptide binding to free domain structures . Energies E and temperatures T are given in dimensionless model units . The thermodynamic behavior of our systems is obtained using Simulated Tempering ( ST ) [45]–[47] , an expanded ensemble MC method in which T is treated as a dynamical parameter . The method is convenient both for finding global minimum-energy states and studying equilibrium behavior . For each PSD-95 and GRIP1 structure-peptide pair , we performed 5 independent ST runs . An example trajectory is shown in Figure S1 in Supporting Information . In addition , fixed-T MC simulations close to the midpoint , , i . e . , where bound and unbound populations are equal , were also performed to provide additional statistics for free energy surface calculations . 10 independent fixed-T runs were performed for each structure-peptide pair in Table 1 . Additional details on the computational model and simulation procedure are provided in Methods . A challenging test for our computational model , used also in guiding the development of our all-atom energy function , is the prediction of bound peptide conformations . Figure 1 shows the model conformations found with the lowest total energy , E , across all ST and fixed-T MC runs for each system , superimposed on the corresponding experimental structures . All 6 min-E conformations are bound at the PDZ peptide binding pocket and many of the finer atom-level details match the experimental structures . Of special interest is to compare the two sets of results obtained for the ligand-bound and ligand-free PSD-95 and GRIP1 PDZ domain structures . One of the most pronounced differences is due to the different sidechain orientations at P ( –2 ) between GRIP1-IIb and GRIP1-IIf docked peptides , such that the Tyr sidechain is pointing either out ( GRIP1-IIf ) or into ( GRIP1-IIb ) the peptide binding pocket ( residue positions on PDZ binding peptides are typically numbered P ( 0 ) for the C terminus residue , P ( –1 ) for the immediately preceding residue , and so on ) . This difference in orientation is likely related to a small shift in the helix between the ligand-free and ligand-bound structures of the GRIP1 domain [37] , such that the binding pocket is slightly wider in the bound structure . Having seen that the lowest-E states represent more or less correctly bound ligands , we turn to the equilibrium behavior of the domain-peptide interaction . Figure 2 shows the T dependence of inter-chain hydrogen bond and hydrophobic interactions for PSD95-If/b and GRIP1-IIf/b . Some general trends are immediately seen . At high Ts , only limited interactions between peptides and domains occur , consistent with a process dominated by entropic effects . As T is lowered , peptides and domains associate increasingly , making both favorable hydrogen bonds and hydrophobic interactions . While all binding curves are smooth , the precise behavior is seen to depend on which domain structure type is used . Particularly , we find that the free domain structures ( PSD95-If and GRIP1-IIf ) bind their ligands somewhat weaker than their respective bound structures ( PSD95-Ib and GRIP1-IIb ) . To investigate this difference quantitatively , we fit the binding curves in Figure 2 to a simple two-state expression with 4 free parameters . The fits are good for all binding curves and the fitted parameters are given in Tables 2 and 3 . Of particular interest are the parameters , the midpoint temperature representing equal populations of the two states , and , the energy difference which controls the sharpness of the transition . The midpoints obtained are and for PSD95-Ib and GRIP1-IIb , respectively . The corresponding for PSD95-If and GRIP1-IIf are roughly 4% lower . We also find differences in , as well as in the other 2 fit parameters , but the statistical errors for these parameters are larger ( see Table 2 and 3 ) . One statistically significant difference is a slightly sharper binding transition for PSD95-If compared to PSD95-Ib . This can also be seen as a relatively higher peak in the specific heat capacity curve ( ) for PSD95-If , as shown in Figure 3 . However , all curves exhibit single peak behavior and the T-values at the peaks correspond well to the found from the fits in Figure 2 . Hence , while we find differences in the binding behavior for bound and free domain structures , binding as an overall two-state process with a single transition appears to be a robust feature . The variations in binding behavior between bound and free structures obtained in our simulations reflect structural differences between liganded and unliganded PDZ domain forms . Some of these differences are likely preserved by our native state constraints . Previous simulation results indicate that overall receptor flexibility and dynamics can play a major role in PDZ peptide binding and selectivity [7] , [25] , [48] , [49] . Interestingly , structural differences in the binding pocket between bound and free form is significant for the GRIP1 domain [37] while quite negligible for PSD-95 [36] . Our results thus indicate that even subtle structural differences can impact binding significantly . Regardless of these differences between bound and free form our model predicts that the GRIP1 domain binds its peptide more strongly than PSD-95 , with ( see Figure 2 ) . Meaningful quantitative binding affinities cannot be directly obtained , however , because T is not matched to physical units . Experimentally , the dissociation constant of the PSD-95/CRIPT interaction has been measured to at 298 Kelvin , using isothermal titration calorimetry [38] . The binding affinity of the GRIP1 domain for the Liprin- peptide has to our knowledge not yet been determined . The binding curves in Figure 2 report on the overall character of the binding transition but do not provide any structural details , such as where on the protein surface binding preferentially occurs or how the peptide chain dynamics is influenced by binding . In defining a bound state , we use the root-mean-square-deviation between the atom coordinates of a model peptide conformation , , and those of the experimental ( native ) peptide structure , , i . e . , ( 1 ) where the sum goes over n peptide atoms , either all non-H or only -atoms ( indicated by superscripts ALL and , respectively ) . An advantage of the RMSD measure is that a small value indicates that binding has occurred both at the right surface area and with a native-like internal conformation . Any peptide with is considered correctly bound in the PDZ binding pocket . The choice of will be discussed later . In order to delineate the internal conformational dynamics of the peptide chain from its binding , we calculate also , where the minimization is over all rigid body translations and rotations of the peptide conformation . Hence , is the measure typically used in the analysis of folding trajectories and its notation is chosen merely to distinguish it from the “non-optimized” RMSD measure in Equation 1 . A small means that the peptide is native-like regardless of whether it is bound or not . For both the PSD-95 and GRIP1 domain-peptide pairs , the probability that the peptides occupy the bound state , , increases sharply as T is lowered ( see Figure 4 ) . It is notable that for PSD95-Ib , at the lowest T simulated , , indicating a very low probability for the peptide to bind parts of the domain surface other than the PDZ binding pocket . values for PSD95-If , GRIP1-IIb , and GRIP1-IIf are lower but the PDZ binding pocket is the dominating binding site in these cases , too , and will likely increase further at still lower Ts . Consistent with our results in Figure 2 , Figure 4 shows a higher peptide binding propensity for liganded ( PSD95-Ib and GRIP1-IIb ) compared to the unliganded structures ( PSD95-If and GRIP1-IIf ) . These shifts are smaller than the differences between the two PDZ domains , as noted above . When the peptides associate with the protein surfaces they not only bind to the peptide binding pocket , they also undergo internal conformational transitions such that they more closely resemble the native peptide structures . This is clear from the lower panel of Figure 4 , which shows that decreases with temperature T . Hence , the peptide-binding process also leads to increasingly native-like peptide conformations . By contrast , the peptide chains by themselves show little tendency to form any specific structure , at least over the temperatures studied , as indicated by a relative constant for isolated chains ( see Figure 4 ) . Moreover , the chain compactness is similarly only weakly dependent on T for both peptide sequences ( see Figure S2 in Supporting Information ) . In this sense , our peptides are intrinsically disordered and their interaction with the PDZ domains can be seen as a minimal example of coupled folding and binding . Direct observation of such coupled folding-binding behavior in atomistic simulations has been seen previously mainly for -helical peptides [50]–[54] . It must be pointed out that despite the indicated “folding , ” significant structural heterogeneity remains in the bound state . This diversity represents the conformational entropy of the bound state and is important to take into account since it can significantly contribute to ligand binding [55]–[57] . In fact , in defining the bound state , our aim was to choose large enough to comprise most of this diversity , but not too large such that incorrectly bound peptide conformations are included . To explore this tradeoff , we show in Figure 4 curves obtained also with and for PSD95-Ib and GRIP1-IIb . Increasing to 9 Å from 6 Å has a relatively small impact on the curves . Most of the structural diversity is therefore included with . At the other end , to see that is not too large , we superimposed representative sets of peptide conformations with . This ensemble is naturally diverse but do not include conformations that can be considered misdocked ( see Figure S3 in Supporting Information ) . Finally , we find it instructive to construct reference structures by rotating the experimental peptide structures by a half turn , such that the atoms of the first and last peptide amino acids exchange positions . These “flipped” peptides have and for the CRIPT ( PSD-95 ) and Liprin- ( GRIP1 ) peptides , respectively . Hence , peptide conformations of this nature would not contribute positively towards in our definition of the bound state ( and are not observed in our simulations ) . We turn now to the binding free energy landscapes of our PDZ domains , i . e . , the free energy as a function of a set of order parameters indicating the progress of binding . For this purpose we use , in addition to the total energy E , two standard [58] , [59] structural order parameters , and Q , defined as the distance between the centers-of-mass ( CM ) of model and experimental peptide conformations and the fraction of inter-chain native contacts , respectively . and Q are complementary in that each provide different perspective on the peptide binding process . The binding free energy surfaces for PSD95-Ib and GRIP1-IIb show bound and unbound states well separated with a single barrier ( the transition state , TS ) at 4–6 Å and 0 . 1–0 . 2 ( see Figure 5 ) . The binding landscapes do not exhibit any competing deep local minima representing misdocked conformations and therefore constitute almost ideal “binding funnels” [60] . This is reassuring in terms of the validity of the model and indicates that nonspecific binding between PDZ domain and peptide chains may be very limited . The one-dimensional free energy profiles in , Q and E reveal a more distinct free energy barrier between the bound and unbound states for GRIP1-IIb compared to PSD95-Ib , indicating a more cooperative binding process for the class II domain ( see Figure 5 ) . In the E parameter , a small barrier separates bound and unbound states for GRIP1-IIb while such a barrier is mostly absent for PSD95-Ib . In the structural parameters , Q and , the barriers are overall much higher but the trend remains . This can be seen , for example , in the free energy difference between the transition state and the native , bound state , , in the parameter . From Figure 5 , we find that and for PSD95-Ib and GRIP1-IIb , respectively . One could easily suspect that the relatively higher barrier for GRIP1-IIb is due to its longer peptide . This is however not the case . We re-made our simulations for GRIP1-IIb with a truncated , 5-amino acid version of Liprin- and found that in fact increases slightly to . Hence , the difference between the PSD-95 and GRIP1 systems is likely mainly related to differences in the amino acid sequences . The bound state for GRIP1-IIb is characterized by a single , deep minimum at , i . e . , with most of the native contacts formed . The PSD-95 domain , by contrast , exhibit a significantly wider distribution of Q-values in the bound state . In addition to a deep minimum , a second weaker minimum exists at . Visual inspection of the minimum reveals peptide conformations in which the C terminal Val of CRIPT is tethered to the PDZ binding pocket , kept in place mainly through hydrophobic interactions involving the Val and hydrogen bonding between the peptide C terminus and the PDZ carboxylate binding loop , leaving a floppy N terminal region . Such flexible , yet bound conformations are mostly absent for GRIP1-IIb . Instead , its peptide typically binds through both the Cys and Tyr sidechains at P ( 0 ) and P ( –2 ) . From the perspective of our model , we find that additional hydrophobic contacts provided by P ( –2 ) in class II domain-peptide binding give a more rigidly bound peptide ensemble , which in turn produces a higher free energy barrier for binding and a more cooperative binding process . A question that arises in comparing features of the free energy surfaces of PSD95-Ib and GRIP1-IIb is to what extent they can be controlled by the peptide sequence . In this regard , promiscuous PDZ domains which bind both class I and II peptides are of particular interest . We therefore apply our method to one such domain , the PDZ domain of PICK1 , and simulate the binding of both a class I ( PICK1-Ib ) and a class II ( PICK1-IIb ) peptide , as displayed in Table 1 . Despite that the two peptide sequences bind the same domain structure , their free energy surfaces are quite different ( see Figure 5C and D ) . Specifically , the PICK1-Ib landscape exhibits striking similarities with PSD95-Ib , particularly with regard to a broad Q-distribution of the bound state . PICK1-IIb , on the other hand , has a binding free energy landscape similar to GRIP1-IIb , with a single well-defined native basin of attraction . The binding free energy barriers for PICK1-Ib and PICK1-IIb are and , respectively , such that the class II peptide again shows a relatively stronger binding cooperativity . It is interesting to compare our results for PICK1-Ib and PICK1-IIb with those of Madsen et al . [44] . Using an assay based on fluorescence polarization , they found that the PICK1 PDZ domain showed a higher affinity for a class II than a class I peptide ( ) . This is in qualitative agreement with our results , as we find a higher for PICK1-IIb over PICK1-Ib ( see Figure 5 legend ) , although their class II ligand was not the same as ours . Madsen et al . also obtained docked peptide structures using homology modeling and found to be unusually displaced from at the N terminal end , somewhat reminiscent our local free energy minimum . However , for typical peptides in our simulations the N terminal ends have become almost entirely displaced from the -helix . One might think that this structural diversity is exaggerated by our model because , after all , PDZ specificity is in part obtained from interactions with P ( –2 ) . We therefore tested the PICK1 mutation Ala87Leu , which was introduced by Madsen et al . and meant to fill out the hydrophobic pocket normally occupied by the P ( –2 ) residue . The mutation was indeed found to essentially eliminate binding to both the class I and II peptides in their assay [44] . We find in our simulations that the Ala87Leu mutation drastically reduces from roughly 0 . 5 at in wild-type PICK1 to and 0 . 09 for the class I and II peptides , respectively . Hence , interactions involving P ( –2 ) are still crucial for proper binding in our model despite the local minimum . In this context , it is interesting to note that experimental PDZ domain-peptide complexes were recently obtained in which the interaction occurs mainly through the P ( 0 ) position , such that the peptides bind roughly perpendicular to the domain surface [61] . Above we have shown that , in our model , peptide binding can be seen roughly as a two-state process in which a single free energy barrier separates the bound and unbound states . How is this free energy barrier crossed during binding ? To address this question and further investigate the mechanism underlying peptide binding we perform a large number of fixed-temperature simulations where the peptide chains are , as previously , initiated in random positions and conformations . In contrast to above , the MC “kinetics” simulations are performed using only small-step updates for the peptide chain; global , unphysical pivot moves are excluded ( see Methods ) . A fraction of rigid body translation and rotation MC moves for the peptide chain is included . There are two processes for the peptide chain in these simulations , a search on the protein surface for the peptide-binding pocket and , subsequently , a conformational search for the correctly bound structure . Because of the inclusion of rigid body moves , we assume a dynamics in which the search process across the protein surface is fast . Relaxation towards equilibrium is therefore limited by a conformational reorganization of the peptide and protein chains during binding , which is the process we are primarily interested in . We find that the relaxation behavior for both PSD95-Ib and GRIP1-IIb systems is consistent with a single-exponential curve , as can be seen in Figure 6 . This indicates a single rate-limiting step in the peptide binding process , or , in other words , the free energy barrier is crossed without significantly populating an intermediate state . Only a handful kinetic experiments of PDZ domain-peptide binding have been performed so far but one such study has presented results for the PSD-95 system analyzed here . Using stopped-flow fluorescence spectroscopy , Jemth et al . [39] observed single-exponential binding traces for the PSD-95 PDZ domain and a dansylated CRIPT peptide . Our results are therefore consistent with these observations . However , it must be pointed out that the MC-based simulations performed here should not be seen as mimicking kinetic experiments , as chain diffusion effects are not rigorously taken into account . A more realistic comparison is likely achieved by focusing on relative kinetic effects between peptide binding systems . In this respect , we observe a significant difference in relaxation times between PSD95-Ib and GRIP1-IIb , such that , a prediction which may be tested experimentally . This difference in relaxation rate between the two domains is consistent with the larger free energy barrier seen for GRIP1-IIb over PSD95-Ib . We have developed a MC based procedure for exploring peptide binding processes and employed it to two typical PDZ class I and II domains and a dual class I–II domain . In combining the equilibrium and small-step , fixed-temperature kinetic simulation results , a picture emerges for the binding process in which there are overall similarities but also differences in the details . In all cases , binding is coupled to folding , and can be characterized as an overall two-state process with a free energy surface funneled towards the peptide bound state . Binding to the PSD-95 PDZ domain involves a lower free energy barrier than the GRIP1 PDZ domain , leading to significantly faster binding kinetics , at least for the peptide sequences studied . What is the origin of this difference ? The shape of the near-native free energy surface for the GRIP1 PDZ domain indicates a relatively coherent ensemble of bound peptide conformations , stabilized by hydrophobic interactions with P ( 0 ) and P ( –2 ) . As a class I domain , the PSD-95 domain lacks strong hydrophobic interactions at P ( –2 ) leading to a more conformationally diverse bound state , spanning a wider range of and Q values . In particular , we find a weak free energy minimum corresponding to peptides bound to the PDZ binding pocket mainly through the P ( 0 ) position , with a flexible N terminal tail . The population of such conformations are significantly smaller for the GRIP1 PDZ domain . Our results are therefore consistent with a binding mechanism in which the rate-limiting step is the initial binding of P ( 0 ) at the PDZ peptide binding pocket . This interpretation is also supported by recent experimental PDZ domain-peptide structures , including GRASP [61] and X11 [62] , where peptides are attached in a “perpendicular” mode . To what extent these results apply to other class I and II PDZ domains remains to be seen . However , the fact that an analogous behavior is found for the dual class I–II PICK1 domain indicates that it may have some generality .
All simulations are performed using essentially the model described in [26] , with a small improvement described in the following . Our original starting point was a model developed for peptide folding [27] , [28] which combines an all-atom protein representation with an effective energy function based mainly on hydrogen bonding , hydrophobicity , and electrostatic attractions . This model was then adapted for peptide binding [26] , where , in particular , we added a context dependence to the energy function such that electrostatic attractions between partial charges buried in the protein were made effectively stronger than those solvent exposed . This was accomplished by using a parameter , , indicating the “degree of buriedness” for any atom i . In this work , we add a context-dependent term describing desolvation effects on backbone atom groups , ( 2 ) in which the sum goes over all backbone NH and CO groups i . For “unsatisfied” NH and CO groups , i . e . , those not participating in any intra- or inter-chain hydrogen bond , , and for all others , . The quantity is calculated at a point , , which for a NH group is located 2 . 0 Å from the H atom in the NH direction , and for a CO group , 2 . 0 Å from the O atom in the CO direction . is thus found approximately in the space occupied by a potential solvent molecule hydrogen bonded to i . indicates that this space is available to a solvent molecule while indicates it is instead occupied by other protein atoms . Hence , “unsatisfied” NH and CO groups with ( i . e . also unlikely to participate in solvent hydrogen bonding ) are energetically penalized . The term therefore acts as a desolvation effect for backbone atoms . The strength chosen is . Including this energy term yields a crucially improved performance over the previous model [26] , most notably for peptide binding to free domain structures . Specifically , the PSD95-If domain-peptide pair exhibited almost no propensity for correct binding previously [26] while including yields reliable binding as detailed in this work . To obtain equilibrium conformational ensembles of our domain-peptide systems we used Simulated Tempering ( ST ) . [45]–[47] , in which conformational updates are alternated with updates in the temperature T . Initially , a set of discrete temperatures are selected . Changes between these discrete temperatures during simulations are then treated as ordinary MC updates . 8 different temperatures in suitable ranges are used for all domain-peptide systems . For updates in conformational space , we use a few different move types . For the protein domain , which is constrained close to its native structure , we use sidechain rotations , in which a single -angle is turned , and semi-local backbone moves , in which 8 consecutive - and -angles are turned in a coordinated way [63] . For the peptide chain , 3 additional moves are used: a pivot move which turns a single - or -angle , and rigid body translation ( ) and rotation ( ) moves . An effective peptide concentration is set by the box side L . For computational reasons , we use a small box such that , corresponding to an effective concentration of . We performed the following peptide binding simulations . For PSD95-Ib , PSD95-If , GRIP1-IIb , and GRIP1-IIf , 5 ST runs were performed with at least elementary MC steps . These runs were used to find the T dependence of various observables including the specific heat curves . 10 fixed-T MC runs at were performed for all of the 6 domain structure-peptide pairs in Table 1 , each with 2 or steps . These simulations were used for free energy surfaces calculations . The MC kinetic simulations differs from the equilibrium runs in the following ways . First , the global , unphysical pivot move was turned off , such that only small-step chain moves were allowed . Second , the translation step size was decreased from 5 Å to 1 Å . 200 independent binding runs were performed for PSD95-Ib and GRIP1-IIb consisting of elementary MC steps . The progress of binding is quantified using the two order parameters Q , the fraction of native inter-chain contacts , and , the distance between model and native peptide centers-of-mass ( CM ) . To calculate Q , we determined initially a set of inter-chain amino acid contacts for each experimental domain-peptide structure . Two amino acids are considered in contact if any two non-H atoms , one from each amino acid , have a distance . This yields sets of 28 , 30 , and 23 inter-chain native contacts for the domain-peptide structures 1BE9 ( PSD-95 ) , 1N7F ( GRIP1 ) , and 2PKU ( PICK1 ) , respectively . For PICK1-Ib , in the absence of an experimental ligand-bound structure for the peptide , we use our minimum-energy conformation ( see Figure 1F ) , which yields a set of 23 native contacts . In calculating Q for a peptide conformation , the fraction of native contacts formed is determined by applying the same contact definition . The CM distance is determined using , where and are the CMs of the model and native peptide conformations , respectively , calculated over the atoms of the last 4 amino acids . | The complex biological processes occurring in living organisms are enabled by numerous networks of interacting proteins . It is therefore of great interest to understand the physical interplay between proteins and , in particular , how this process gives rise to highly specific network connectivities . For a long time , the dominant molecular view of protein-protein interactions was the docking of more or less static folded structures , with specificity obtained from a complementarity in shape and charge distributions . Lately it has been realized that many of the links in protein networks are mediated by interactions between folded domains , on the one hand , and disordered polypeptide segments , on the other . We use an all-atom Monte Carlo based approach which attempts to capture this domain-peptide binding process in full and apply it to representative members of a common domain family . This allows us to examine and compare detailed aspects of the binding free energy landscapes which underlie specificity and affinity . Being able to model domain-peptide binding in a physically sound , yet computationally tractable way is essential for identifying molecular binding mechanisms and opens up possibilities for modifying interaction networks in a controlled way . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"physics",
"biomacromolecule-ligand",
"interactions",
"biophysic",
"al",
"simulations",
"biology",
"computational",
"biology",
"biophysics",
"simulations",
"biophysics"
] | 2011 | Binding Free Energy Landscape of Domain-Peptide Interactions |
Genetic engineering with luciferase reporter genes allows monitoring Trypanosoma brucei ( T . b . ) infections in mice by in vivo bioluminescence imaging ( BLI ) . Until recently , luminescent T . b . models were based on Renilla luciferase ( RLuc ) activity . Our study aimed at evaluating red-shifted luciferases for in vivo BLI in a set of diverse T . b . strains of all three subspecies , including some recently isolated from human patients . We transfected T . b . brucei , T . b . rhodesiense and T . b . gambiense strains with either RLuc , click beetle red ( CBR ) or Photinus pyralis RE9 ( PpyRE9 ) luciferase and characterised their in vitro luciferase activity , growth profile and drug sensitivity , and their potential for in vivo BLI . Compared to RLuc , the red-shifted luciferases , CBR and PpyRE9 , allow tracking of T . b . brucei AnTaR 1 trypanosomes with higher details on tissue distribution , and PpyRE9 allows detection of the parasites with a sensitivity of at least one order of magnitude higher than CBR luciferase . With CBR-tagged T . b . gambiense LiTaR1 , T . b . rhodesiense RUMPHI and T . b . gambiense 348 BT in an acute , subacute and chronic infection model respectively , we observed differences in parasite tropism for murine tissues during in vivo BLI . Ex vivo BLI on the brain confirmed central nervous system infection by all luminescent strains of T . b . brucei AnTaR 1 , T . b . rhodesiense RUMPHI and T . b . gambiense 348 BT . We established a genetically and phenotypically diverse collection of bioluminescent T . b . brucei , T . b . gambiense and T . b . rhodesiense strains , including drug resistant strains . For in vivo BLI monitoring of murine infections , we recommend trypanosome strains transfected with red-shifted luciferase reporter genes , such as CBR and PpyRE9 . Red-shifted luciferases can be detected with a higher sensitivity in vivo and at the same time they improve the spatial resolution of the parasites in the entire body due to the better kinetics of their substrate D-luciferin .
African trypanosomes pose a threat to millions of humans and animals in sub-Saharan Africa . Only two species readily infect humans and both are subspecies of Trypanosoma brucei ( T . b . ) . T . b . gambiense is responsible for the chronic form of human African trypanosomiasis ( HAT ) in West and Central Africa and accounts for more than 97% of the near 10 , 000 sleeping sickness patients who are diagnosed and treated annually [1] . T . b . rhodesiense causes a more acute form of HAT in South and East Africa , only differing from a third non human-infective subspecies , T . b . brucei , by a single gene , SRA , that confers resistance against human serum [2] . All T . b . subspecies are transmitted by the bite of tsetse flies ( Glossina spss ) . Upon injection into a vertebrate host , parasites multiply locally in the lesion and cause a blood , lymph and tissue infection , also called first stage disease . Later , the parasites invade the central nervous system ( CNS ) initiating the second stage of the disease . Untreated infections almost invariably have a fatal outcome that occurs after weeks to months in rhodesiense HAT and months to years in gambiense HAT [3] . Treatment is subspecies- and stage-specific [1] . First stage gambiense and rhodesiense HAT are treated with pentamidine and suramin respectively . The first line treatment for second stage gambiense HAT consists of nifurtimox-eflornithine combination therapy ( NECT ) while second stage rhodesiense HAT is still treated with melarsoprol . All current drugs used to treat HAT are toxic [4] . For obvious reasons , basic and applied research on trypanosomes and HAT , including drug resistance studies , highly benefit from the availability of T . b . rhodesiense and T . b . gambiense strains that have been recently isolated from patients with known treatment outcome and that along their isolation underwent only few in vivo and/or in vitro passages [5]–[8] . To this end , bloodstream form trypanosomes can be isolated from diverse patient specimens such as blood , lymph or cerebrospinal fluid ( CSF ) . Generally , T . b . brucei and T . b . rhodesiense can be easily isolated in classical laboratory rodents such as mice and rats [9] . T . b . gambiense , on the other hand , is very difficult to isolate and often requires either susceptible rodent species [9]–[12] or severely immune-suppressed or –deprived hosts [13] , [14] . Seldomly , isolation of bloodstream form T . b . gambiense parasites has been achieved by direct inoculation of in vitro medium containing feeder layer cells [5] , [15] . Apart from isolation of trypanosome strains , rodent models for HAT are considered to be informative because they can reproduce the invasion of the central nervous system ( CNS ) [16] , [17] . Thus , these rodent models are highly relevant for investigating drug discovery , drug resistance and treatment failure [18] , [19] . Much like the clinical diversity seen in HAT [1] , rodent models also reveal a very broad spectrum of pathology resulting from infection with diverse strains of the human pathogenic subspecies of T . b . [5] , [20]–[27] . Genetic engineering of parasites has made it possible to monitor infections in living animals using biophotonic techniques , such as in vivo fluorescence and bioluminescence imaging [28] , [29] . In vivo bioluminescent imaging ( BLI ) allows the tracking of luciferase-modified cells in living animals over time without the need to sacrifice them . This technique has been applied to study the spatio-temporal distribution of T . b . brucei and T . b . gambiense parasites in murine models and lead to the discovery of a testis tropism of T . b . brucei AnTaR 1 , where the parasites are less accessible for trypanocides [30] . BLI also revealed that different T . b . gambiense strains can induce a variety of infections in mice , ranging from chronic to silent , thus mimicking the clinical diversity that is observed in HAT [5] . The former models made use of RLuc as the bioluminescent reporter [31] , which , upon oxidation of its substrate coelenterazine , emits blue light ( peak luminescence at 480 nm ) that is readily absorbed by blood and other tissues . Alternative luciferases emitting light with longer wavelengths exist , such as luciferase enzymes from fireflies , click beetles and railroad worms that all use D-luciferin as substrate [32] , [33] . Firefly luciferase ( FLuc ) appeared to sort into the glycosome of T . b . possibly disturbing ATP/ADP equilibrium upon activation [34] . Recently the well characterised T . b . brucei GVR 35 strain was modified with a firefly luciferase variant ( LUC2 ) , that emits yellow light ( 560 nm ) , and proved to be useful to shorten the follow-up period in studies on drugs that reach the CNS during the infection [29] . Also , studies with LUC2 modified T . vivax have shown that this trypanosome species , like T . b . , invades tissues including the CNS [35] . However , luciferases that emit light beyond 600 nm are potentially even more useful for in vivo imaging due to the fact that transmission of light through animal tissue increases greatly above this wavelength [36] . The synthetic click beetle red luciferase ( CBR ) from Pyrophorus plagiophtalamus and the thermostabilised PpyRE9 luciferase from Photinus pyralis both emit light around 617 nm [37] . These red-shifted luciferases have shown a potential to better resolve signals from deeper tissues than the original FLuc or to emit more stable luminescence than the LUC2 variant [38] , [39] . Very recently it was shown that PpyRE9-tagged T . b . brucei GVR 35 parasites allow improved detection in BLI over LUC2-tagged T . b . brucei GVR 35 models [40] . In the present study we explored the use of red-shifted firefly luciferases CBR and PpyRE9 for bioluminescent imaging of all three subspecies of T . b . , including the human pathogens , in comparison with the existing RLuc strains in our collection . This collection consists of a set of four genetically and phenotypically very different trypanosome strains ( Table 1 ) . T . b . brucei AnTaR 1 is a pleomorphic strain from Uganda that causes a sub-acute infection in various mouse and rat strains [41] , [42] . T . b . rhodesiense RUMPHI is a recently isolated strain from Malawi , which underwent only a few in vivo and in vitro passages and originates from an area with less virulent T . b . rhodesiense than those circulating in Uganda [43] , [44] . T . b . gambiense LiTaR 1 is a well characterised virulent strain that is used for the production of diagnostic antigens in the card agglutination test for trypanosomiasis ( CATT ) and in the recently developed HAT-Sero-K-SeT [45] , [46] . This strain is extensively passaged in vivo and produces an acute monomorphic infection in rodents . In contrast , T . b . gambiense 348 BT was isolated in the HAT focus of Mbuji-Mayi in the Democratic Republic of the Congo ( DRC ) [7] , where a very high relapse rate after melarsoprol treatment has been observed that may be related to the presence of a chimeric aquaglyceroporin 2/3 gene [47] , [48] . We focus this report on the in vitro drug sensitivity , the in vitro growth characteristics and the in vivo virulence in mice , assessed through bioluminescence imaging of either RLuc , CBR and PpyRE9 luciferase activity , of the four bioluminescent T . b . strains .
This study was approved by the Veterinary Ethics Committee of the Institute of Tropical Medicine , Antwerp , Belgium ( protocol BM2012-1 and BM2013-5 ) and the Veterinary Ethics Committee of the University of Antwerp , Belgium ( protocol BPI-EAT ) . It adheres to the European Commission Recommendation on guidelines for the accommodation and care of animals used for experimental and other scientific purposes ( 18 June 2007 , 2007/526/EG ) and the Belgian National law on the protection of animals under experiment . The parasite strains included in this study belong to the cryobank of the World Health Collaboration Center for Research and Training on Human African Trypanosomiasis Diagnostics at the Institute of Tropical Medicine in Antwerp , Belgium . The axenic in vitro culture of monomorphic and pleomorphic bloodstream form trypanosome populations in HMI-9 has been described elsewhere [49] , [50] . The original host , the year and country of isolation , the number of in vivo passages and the medium for in vitro culture of T . b . brucei AnTaR 1 , T . b . rhodesiense RUMPHI and the T . b . gambiense strains LiTaR 1 and 348 BT are described in Table 1 . The propagation of the bloodstream form in vivo in rodents , the adaptation in vitro to an HMI-9 based culture medium and the molecular confirmation of their taxonomic identity have been described previously [7] , [51] , [52] . Iscove's modified Dulbecco's medium powder ( IMDM ) and foetal calf serum ( FCS; heat-inactivated; EU approved; South American origin ) were purchased from Invitrogen ( Carlsbad , USA ) . For in vitro assays , medium was prepared from IMDM without phenol red ( Invitrogen ) and without addition of antibiotics . All other culture media ingredients were from Sigma–Aldrich ( St . Louis , MO , USA ) . Briefly , strains were isolated from first peak parasitaemia in mice , cultured in HMI-9 based medium containing 1 , 1% methylcellulose and 15% foetal bovine serum with or without 5% heat-inactivated human serum until adaptation [51] . All strains were adapted to medium without methylcellulose before transfection as previously described [52] . Strains were cultivated in 500 µl of medium in a 48-well plate at densities between 103–106 cells ml−1 and maintained in logarithmic growth phase by subpassages at appropriate dilutions after 24 to 72 hours of incubation at 37°C and 5% CO2 . Cultures were monitored by phase contrast inverted microscopy . Cell counting was performed in disposable counting chambers ( Uriglass , Menarini Diagnostics , Belgium ) . For larger cell preparations , the cultures were stepwise scaled up to 40 ml in 25 cm2 flasks , by addition of four ( for T . b . gambiense ) to nine ( for T . b . rhodesiense/brucei ) volumes of fresh medium once the parasites reached a density of 5×105 cells ml−1 . For long term storage , cells were concentrated tenfold from log phase cultures in 90% medium with 10% glycerol and frozen stepwise to −40°C at 1°C/min using a programmable cryogenic freezing device ( MiniCool MP40 , Air Liquide , Belgium ) whereafter they were kept in liquid nitrogen . The promoterless vector pHD309 was used for constitutive expression of foreign genes in trypanosomes [5] , [30] , [52] . For overexpression of a single reporter gene , the pHD309 plasmid was cut with BamHI and HindIII and PCR products were fused using In-Fusion Cloning and transformed in Fusion Blue cells according to the manufacturer's recommendations ( Clontech , Takara Bio , Japan ) . The cDNA sequences of the reporter genes were amplified from their donor plasmids using gradient PCR and a proofreading polymerase ( Deep VentR , New England Biolabs , UK ) . All primers ( Biolegio , Nijmegen , The Netherlands ) contained a cDNA specific sequence and a 5′ extension of 15 nucleotides specific to the place of integration , containing the restriction site and sequence overlap with the vector as required for the In-Fusion Cloning reaction ( Table S1 ) . Trypanosomes that have been electroporated ( Gene Pulse Xcell , Bio-Rad , USA ) with a NotI linearised plasmid can afterwards be selected with hygromycin to obtain stable recombinants . The transfection and selection of trypanosome strains that express RLuc has been described earlier [30] , [52] . Recombinant populations were maintained in 1 µg ml−1 hygromycin for T . b . gambiense strains and 5 µg ml−1 for T . b . brucei and T . b . rhodesiense strains for over three weeks after transfection upon which the most resistant populations were cryopreserved and used for further analysis of luciferase activity . The selection antibiotic was no longer added to the in vitro cultures during luciferase activity and drug sensitivity testing . To measure the luminescent activity of the RLuc-modified strains , the EnduRen Live Cell assay ( Promega , Madison , USA ) was used with a final EnduRen concentration of 6 µM . Sixty mM EnduRen stock solution in DMSO was diluted 1∶1000 into HMI-9 medium without phenol red . Five µl of this solution were transferred to a well of a white opaque 1/2 area 96-well plate ( Perkin Elmer , Waltham , MA , USA ) and 45 µl of a trypanosome suspension were added . The plate was incubated for at least one hour in a 5% CO2 incubator at 37°C . After measurement of RLuc activity with EnduRen , the amount of ATP was measured in the same sample by adding 50 µl of CellTiter Glo reagent ( Promega ) , to create a luminescent multiplex viability assay as described previously [52] . To measure the luminescent activity of the firefly luciferases , the ONE-Glo Luciferase reagent ( Promega ) was reconstituted as described by the manufacturer and 20 µl of this assay solution were added to 20 µl of a trypanosome suspension in HMI-9 in an opaque white 1/2 area 96-well plate ( Perkin Elmer ) . A separate aliquot of 20 µl of the trypanosome suspension was used to measure ATP luminescence using the CellTiter-Glo reagent . No centrifugation or wash steps were required in any of the protocols . All luminescent measurements were performed after 2 minutes of shaking and the number of counts was integrated by sampling over a 1 second period ( CPS ) , every minute for 10 minutes using WorkOut software from Victor X3 plate reader ( Perkin Elmer ) . The CPS values were divided by the CPS value of HMI-9 medium ( fold change ) , plotted against the trypanosome cell density and a linear regression was calculated in GraphPad ( Prism ) . The threshold for detection was defined as a fold change >3 . The relative activity in each clone was calculated as the ratio of the CPS in the luciferase assay ( EnduRen , for RLuc or ONE-Glo , for CBR and P9 ) over the CPS in the luminescent cell viability assay ( CellTiter Glo ) . The means of the clones of the red-shifted luciferases with the highest relative activity and the mean doubling time of the wild-type and their luminescent population ( s ) were compared using one-way analysis of variance ( ANOVA ) with Bonferroni post-hoc test in GraphPad ( Prism ) . Eflornithine ( Sanofi Aventis , Paris , France ) and hygromycin B ( Sigma ) were prepared as 10 mg ml−1 stock solutions in distilled water . Melarsoprol ( Sanofi Aventis ) , suramin ( Bayer , Leverkusen , Germany ) , pentamidine isethionate ( Sanofi Aventis ) and nifurtimox ( Sigma ) were stored as 10 mg ml−1 stock solutions in DMSO . Dophanil powder ( Docpharma , Hoeilaart , Belgium ) , containing 455 mg diminazene diaceturate and 555 mg antipyrine per gram , was prepared as a 10 mg ml−1 diminazene diaceturate solution in DMSO . A method to measure the IC50 values of compounds in 96-well plates was performed as described elsewhere [53] . Threefold drug dilutions in duplicate were made in HMI-9 medium to allow testing in final drug concentrations ranging from 100 to 0 . 14 µg ml−1 for eflornithine and hygromycin , from 50 to 0 . 07 µg ml−1 for nifurtimox , from 10 to 0 . 014 µg ml−1 suramin and from 500 to 0 . 7 ng ml−1 for diminazene diaceturate , melarsoprol and pentamidine with 5×103 cells ml−1 in a total volume of 200 µl . Next , the plate was incubated for 72 hours at 37°C with 5% CO2 followed by addition of 20 µl of resazurin ( Sigma; 12 . 5 mg in 100 ml PBS ) . After a further 24 h incubation at 37°C with 5% CO2 , fluorescence was measured ( excitation λ = 560 nm; emission λ = 590 nm ) with a VictorX3 multimodal plate reader using top reading ( Perkin Elmer ) [54] . The results were expressed as the percent reduction in parasite viability compared to the parasite viability in control wells without drugs . The 50% inhibitory concentration ( IC50 ) was calculated using non-linear regression and compared between groups with one-way ANOVA and Bonferroni post-hoc test in GraphPad ( Prism ) . For experiments with T . b . brucei AnTaR 1 , T . b . gambiense LiTaR 1 and T . b . rhodesiense RUMPHI , female OF-1 mice ( 25±3 g ) in groups of 3 were infected intraperitoneally ( IP ) with 2×104 parasites ( from culture medium ) . Every group was tested at days 1 , 4 , 7 , 18 and 26 post-infection . For experiments with T . b . gambiense 348 BT , female OF-1 mice ( 30±5 g ) in groups of 3 , treated or not treated with 200 mg/kg cyclophosphamide ( CPA ) IP ( Endoxan , Baxter , Lessing , Belgium ) 2 days pre-infection , were infected with 2×105 parasites ( from infected mouse blood ) . These groups were tested at days 1 , 3 , 7 , 11 , 43 and , for CBR only , also on day 60 post-infection . Before each BLI recording , animals were weighed and anaesthetised by inhalation of 5% isoflurane ( Isoflo , USP ) for induction and 2% isoflurane for maintenance in 100% 02 at a flow rate of 1000 ml min−1 . While under anaesthesia , mice were injected IP with 10 ml kg−1 body weight of a 15 mg ml−1 D-luciferin ( ViviGlo , D-luciferin potassium salt , Promega ) in phosphate buffered saline pH 7 . 4 ( PBS ) or 1 mg ml−1 ViviRen ( Promega ) in PBS with 0 , 1% bovine serum albumin ( BSA ) [55] . Two to five minutes after injection of the substrate , a ten to fifteen minute image acquisition was made on an in vivo bioluminescence imager ( Photon Imager , Biospace , France ) . During the imaging session , the animal was placed on its back on a heated mat ( 39°C ) to maintain body temperature . After each session , the parasitaemia was estimated using the matching method on 30 fields , allowing a detection limit of 105 cells ml−1 in whole blood [56] . The BLI data were analysed by dividing the images of the mice in 3 rectangular shaped regions of interest ( ROI ) ; covering the abdomen ( 12 . 3 cm2 ) , the thorax ( 6 . 1 cm2 ) and the head ( 2 . 9 cm2 ) ( Figure 1 ) . The radiance in each ROI was obtained from a 30 to 60 second period within the plateau phase of luminescence and expressed in photons per second per square centimetre per steradian ( ph s−1 cm−2 sr−1 ) in M3 Vision ( Biospace , France ) . In the non-infected controls ViviRen was injected first , followed by a washout period of at least 4 hours before D-luciferin administration . The threshold for detection was defined as a >3 fold change in radiance compared to non-infected controls . At day 43 and day 60 for T . b . gambiense 348 BT and at day 26 for T . b . brucei AnTaR 1 and T . b . rhodesiense RUMPHI , mice were transcardially perfused , under Nembutal anaesthesia ( 60 mg kg−1 in PBS , IP ) , with 50 ml of phosphate buffered saline glucose ( PBSG; 10 mM phosphate pH 7 . 4 , 0 . 9% NaCl and 1% glucose ) at a flow rate of 5 ml min−1 to rinse the vascular compartments of trypanosomes . The spleen was excised and weighed while the brains ( without dura mater and arachnoid ) were removed , washed in 10 ml of PBSG and incubated in a 24-well plate in 1 ml of PBSG containing either 1 . 5 mg ml−1 D-luciferin or 0 , 1 mg ml−1 ViviRen . After 5 minutes delay , a BLI recording was made for 10 minutes . The BLI data were analysed by drawing a ROI around the circumference of the well . The radiance ( ph s−1 cm−2 sr−1 ) was expressed as fold change over the average values of the non-infected control brains for each substrate as described above .
RLuc luciferase expressing clones of T . b . brucei AnTaR 1 and T . b . gambiense 348 BT were available from previous studies [30] , [52] . CBR luciferase was integrated in these strains as well as in T . b . gambiense LiTaR 1 and T . b . rhodesiense RUMHPI . The PpyRE9 luciferase ( P9 ) , a promising red-shifted luciferase for in vivo imaging , was only integrated in T . b . brucei AnTaR 1 and was not yet tested in the other strains . Out of four clones of T . b . brucei AnTaR 1 transfected with pHD P9 ( AnTaR 1 P9 ) , ten clones of T . b . brucei AnTaR 1 transfected with pHD CBR ( AnTaR 1 CBR ) , 3 clones of T . b . rhodesiense RUMPHI transfected with pHD CBR ( RUMPHI CBR ) , 7 clones of T . b . gambiense LiTaR 1 transfected with pHD CBR ( LiTaR 1 CBR ) and 2 clones of T . b . gambiense 348 BT transfected with pHD CBR ( 348 BT CBR ) that were simultaneously tested , we identified for each strain and luciferase reporter combination , the clone with the highest relative luciferase activity ( Figure 2: A and Figure S1 ) . There was a significant difference in relative luciferase activity among these red luminescent clones ( ANOVA , F ( 5 , 39 ) = 751; p<0 . 0001 ) . Post-hoc analysis revealed that the relative luciferase activity of clone 4 of P9-modified T . b . brucei AnTaR 1 was significantly higher than the highest relative luciferase activity of the CBR-modified clones from T . b . brucei AnTaR 1 ( clone 19 ) , T . b . gambiense LiTaR 1 ( clone 14 ) , T . b . rhodesiense RUMPHI ( clone 51 ) and T . b . gambiense 348 BT ( clone 13 ) ( p<0 . 05 ) . When comparing only the CBR-tagged strains ( ANOVA , F ( 4 , 32 ) = 48 . 5; p<0 . 0001 ) , post-hoc analysis showed that the relative luciferase activity of T . b . gambiense 348 BT CBR was significantly lower than the others ( p<0 . 05 ) . We found that at least 105 cells ml−1 were necessary for detection of the most luminescent clones among the CBR-modified strains , while for T . b . brucei AnTaR 1 P9 5×103 cells ml−1 of clone 4 were sufficient ( Figure S2 ) . In the luminescent viability assay , at least 1×104 cells ml−1 were needed to obtain a >3 fold change for all tested strains ( Figure S3 ) . The in vitro growth rates ( expressed as doubling time ) were significantly different between the collection of strains ( ANOVA F ( 10 , 22 ) = 30 . 9; p<0 . 0001 ) . However , post-hoc analysis revealed no significantly different growth rates of the wild-type and the corresponding recombinant population ( s ) of each strain ( p>0 . 05 ) ( Figure 2: B ) . The IC50 values of the luminescent populations for hygromycin were significantly higher than those of the corresponding wild type populations ( ANOVA F ( 10 , 49 ) = 46 . 98 , p<0 . 0001 ) . Post –hoc analysis revealed no difference in IC50 values for hygromycin between the luminescent T . b . brucei AnTaR 1 clones ( p>0 . 05 ) , but the IC50 values of T . b . gambiense LiTaR 1 CBR , T . b . rhodesiense RUMPHI CBR and both CBR- and RLuc-modified T . b . gambiense 348 BT strains were significantly lower than those of the luminescent T . b . brucei AnTaR 1 populations ( p<0 . 05 ) ( Figure 2: C ) . The drug sensitivity profiles of all wild-type and luminescent strains were compared against a set of trypanocides ( eflornithine , nifurtimox , diminazene diaceturate , melarsoprol , suramin and pentamidine isethionate ) to test if the luminescent modifications induced differences in IC50 value . For each drug , ANOVA found differences between the IC50 values from the collection of strains ( for eflornithine F ( 10 , 66 ) = 183 , p<0 . 0001; for nifurtimox F ( 10 , 66 ) = 37 , p<0 . 0001; for diminazene diaceturate F ( 10 , 11 ) = 112 . 3 , p<0 . 0001; for melarsoprol F ( 10 , 63 ) = 33 , p<0 . 0001; for suramin F ( 10 , 68 ) = 26 . 22 , p<0 . 001 and for pentamidine F ( 6 , 7 ) = 105 , p<0 . 0001 ) . Post-hoc analysis did not reveal significant differences between the IC50 values of the wild-type and the corresponding luminescent population ( s ) of each strain ( for each drug , p>0 . 05 ) . However , there were substantial differences between the different strains for each of the different drugs as represented in Figure 3 . For eflornithine , all T . b . brucei AnTaR 1 populations had significantly higher IC50 values than all the other populations and all T . b . rhodesiense RUMPHI populations had significantly higher IC50 values than T . b . gambiense populations ( p<0 . 05 ) ( Figure 3: A ) . For nifurtimox , the IC50 values of all T . b . brucei AnTaR 1 populations were significantly higher , while all IC50 values of T . b . gambiense 348 BT were significantly lower than all IC50 values of T . b . rhodesiense RUMPHI and T . b . gambiense LiTaR 1 populations ( p<0 . 05 ) ( Figure 3: B ) . For diminazene diaceturate , IC50 values of all T . b . brucei AnTaR 1 and T . b . gambiense 348BT were significantly different and were significantly higher IC50 than of all T . b . rhodesiense RUMPHI and T . b . gambiense LiTaR 1 populations ( p<0 . 05 ) ( Figure 3: C ) . For melarsoprol , IC50 values of all T . b . brucei AnTaR 1 and T . b . gambiense 348 BT populations were higher than those of T . b . gambiense LiTaR 1 and T . b . rhodesiense RUMPHI ( Figure 3: D ) . Strikingly , the IC50 values of all T . b . gambiense 348 BT for suramin were about tenfold higher than of all the other strains ( p<0 . 05 ) ( Figure 3: E ) . Also for pentamidine , the IC50 values of all T . b . gambiense 348 BT populations were much higher than those of T . b . brucei AnTaR 1 , while the IC50 values of T . b . gambiense LiTaR and T . b . rhodesiense RUMPHI were below the lower threshold ( <0 . 70 ng ml−1 ) ( p<0 . 05 ) ( Figure 3: F ) . The sensitivity of RLuc , CBR and PpyRE9 luciferase detection during in vivo BLI of a murine infection was assessed with T . b . brucei AnTaR 1 . Throughout the infection , very high parasitemia between 107 and 108 . 4 cells ml−1 was observed for the wild-type strain and for the RLuc- , CBR- and P9-modified clones . Mice infected with wild-type and recombinant T . b . brucei AnTaR 1 parasites showed increased body weight gain . The spleen weight of T . b . brucei AnTaR 1 infected mice that survived until day 26 post-infection varied from 1 to 2 gram , roughly tenfold the spleen weight of uninfected mice , indicating severe splenomegaly . For the BLI experiments , the background luminescence of each luciferase substrate was measured in uninfected mice and was at least four times higher for ViviRen ( 16228±4826 ph s−1 cm−2 sr−1 ) than for D-luciferin ( 4622±927 ph s−1 cm−2 sr−1 ) in the abdominal ROI , with lesser differences in the thoracic ROI ( ViviRen: 4064±873 and D-luciferin: 3743±927 ph s−1 cm−2 sr−1 ) and in the head ROI ( ViviRen :3772±789 and D-luciferin: 3131±459 ph s−1 cm−2 sr−1 ) . The in vivo luciferase activities in function of ROI and days post-infection are represented in Figure 4 and visualised in Figure 5 . At day 1 post-infection , when parasitaemia is not yet detectable by microscopy , the BLI signal is already above threshold in the abdominal region of some mice infected with the CBR strain ( Figure 5: E ) and in all compartments of all mice infected with the P9 strain ( Figure 5: I ) . No signal above background could be detected in mice infected with RLuc-tagged parasites ( Figure 5: A ) . At day 4 post-infection , the RLuc-infected mice ( Figure 5: B ) were 100 fold more luminescent in the abdominal region and at least 10 fold more luminescent in the thorax and the head region than wild-type-infected mice ( Figure 4 ) . Mice infected with trypanosomes expressing red luciferases were much more luminescent with a fold change of 5000 in the abdomen and 1000 in the thorax and head for CBR-infected mice and a fold change of almost 100000 in the abdomen and of 10000 in the thorax and head for the P9-infected mice ( Figure 4 and Figure 5: F and J ) . Interestingly , the difference in sensitivity ( as defined by higher radiance values ) between CBR and P9 was not associated with a different distribution of the trypanosomes in the body . No more visual detail could be obtained from the higher luminescence of P9 compared to CBR-tagged parasites ( Figure 5: F and J , G and K ) . BLI pictures often revealed the contours of lymph nodes , but in most cases the signal consisted of a superimposed surface covering multiple organs in the body , especially in the abdomen . With RLuc-tagged parasites , less information was obtained . Luminescence was often only visible in the abdominal region ( Figure 5: B , C and D ) . After day 7 post-infection , data were too limited for analysis due to differences between the strains in survival of mice and parasitaemia . The median survival time was 26 days for mice infected with wild-type; 26 days with RLuc; 26 days with CBR and 11 days with P9 , with deaths in all groups occurring first at day 11 post-infection ( 3 animals ) and later between day 18 and 25 post-infection ( 4 animals ) . T . b . gambiense LiTaR 1 killed mice in 5 days without any overt signs of pathology . The trypanosomes were monomorphic and at day 4 post-infection , hours before death , the parasitaemia approached 108 , 7 cells ml−1 . In contrast , T . b . rhodesiense RUMPHI infection was not fatal up to day 26 post-infection , which marked the end of the experiment . Throughout the infection , the parasitaemia of T . b . rhodesiense RUMPHI varied between 2×106 and 5×107 cells ml−1 and was markedly lower than that of T . b . brucei AnTaR 1 and T . b . gambiense LiTaR 1 . One mouse showed a delay in reaching the first peak of parasitemia ( Figure 5: M–O ) . Although mice infected with RUMPHI did show signs of lethargy , splenomegaly was less pronounced than in mice infected with T . b . brucei AnTaR 1 . When we compared the in vivo luciferase activity in mice infected with CBR-tagged T . b . brucei AnTaR 1 , T . b . rhodesiense RUMPHI and T . b . gambiense LiTaR 1 , it appeared that on day 1 post-infection , when none of the infected mice showed microscopically detectable parasitaemia , all strains could be detected with BLI in the abdominal region of some infected mice ( Figure 5: E , M and Q ) . At day 4 post-infection , all three strains generated a comparable BLI signal in the abdomen and thorax ( Figure 4: A and B ) . The signal from the head was highest in mice infected with T . b . gambiense LiTaR 1 , followed by mice infected with T . b . brucei AnTaR 1 and with T . b . rhodesiense RUMPHI ( Figure 4: C and Figure 5: F , N and R ) . Later during the infection , all mice infected with T . b . gambiense LiTaR 1 and 2 of 3 mice infected with T . b . brucei AnTaR 1 died while all mice infected with T . b . rhodesiense RUMPHI survived until the end of the experiment at day 26 post-infection with steadily increasing BLI signals in the head ROI ( Figure 5: O and P ) . RLuc- and CBR-modified parasites of T . b . gambiense 348 BT were injected in BALB/c that underwent weekly CPA treatment . When these animals became parasitologically positive after 14 to 20 days ( a total of 2–5 trypanosomes in 30 fields ) , their blood was injected into CPA-treated OF-1 mice . The peak of parasitaemia occurred within one week and again , blood containing the parasites was injected in 2 CPA-treated OF-1 mice . The first peak of parasitaemia in the latter mice reached 108 , 1 cells ml−1 and blood was diluted with PBSG to 106 cells ml−1 and used to infect three CPA-treated OF-1 mice and three untreated OF-1 mice . In the untreated mice , parasitaemia remained undetectable in all RLuc-infected mice and in two CBR-infected mice; only one CBR-infected mouse was once positive at day 4 post-infection . In the CPA-treated mice , both RLuc and CBR infections gave rise to detectable parasitaemia that increased until 11 days post-infection where after the mice became only sporadically positive . At the end of the experiment we recorded mild splenomegaly in all mice that became parasitologically positive . The in vivo luciferase activity in function of ROI and days post infection are represented in Figure 6 and visualised in Figure 7 . As observed in the BLI experiment with T . b . brucei AnTaR 1 , the T . b . gambiense 348 BT RLuc-modified parasites were less informative than the CBR-modified parasites . With BLI , the signal of the RLuc-modified trypanosomes was below threshold during the whole infection period in untreated mice ( Figure 7: A–E ) , but in CPA-treated animals , the T . b . gambiense 348 BT RLuc-modified parasites were detectable at day 1 and day 3 post-infection in the abdomen ( Figure 7: F and G ) and at day 7 also in the thorax and the head of some mice ( Figure 7: H ) . In contrast , in both CPA-treated and untreated mice , T . b . gambiense 348 BT CBR-tagged parasites were detectable at day 1 post-infection in the abdomen , and at day 4 post-infection also in the thorax and the head ( Figure 7: K–V ) . At day 7 post-infection , the T . b . gambiense 348 BT CBR-modified parasites became undetectable in the CPA-untreated mice ( Figure 7: M–P ) . In the CPA-treated mice we were able to track the infection in all animals and in all compartments until day 43 post-infection , yet at day 60 post-infection the signal decreased below the detection threshold ( Figure 7: V ) . At day 26 post-infection for T . b . brucei AnTaR 1 and T . b . rhodesiense RUMPHI and at day 43 and day 60 post-infection for T . b . gambiense 348 BT , the surviving animals were sacrificed for ex vivo BLI quantification of the parasites in the brain . The ventral portion of the brain was the most informative for BLI data ( Figure 8 ) . The brains from the animals infected with CBR-tagged strains ( Figure 8: A–C , I ) showed 50 to 100 fold higher luminescence than brains of uninfected mice ( Figure 8: D , representative image measured with D-luciferin ) . In the brain of the mouse infected with the T . b . brucei P9-tagged strain ( Figure 8: J ) , the luminescence was even about 1000 fold higher . In the case of infection with T . b . gambiense 348 RLuc and T . b . brucei AnTaR RLuc , luminescence was detectable at the circumference of the brain rather than in the brain itself ( Figure 8: E , image of T . b . gambiense 348 BT RLuc measured with ViviRen ) and similar to the recording made from the brain of the T . b . brucei AnTaR P9 infected mouse ( Figure 8: J ) , light radiated into the PBSG medium . However , we did not check whether we could detect free trypanosomes in the surrounding liquid or on the dorsal portion of the brain . In case of infections with CBR and P9 tagged trypanosomes and with D-luciferin as substrate , BLI signals emanated from all over the brain , but the densest spots were often observed in the olfactory bulbs , in the ventral anterior hypothalamic region including the suprachiasmatic nucleus , and in the cerebellum , as well as in the pituitary gland . The brains of two CPA-untreated animals infected with the T . b . gambiense 348 BT CBR as well as all untreated mice infected with T . b . gambiense 348 BT RLuc remained negative in BLI . As stated above , these animals also remained aparasitaemic in blood .
In previous studies , we used RLuc as the reporter gene in bioluminescent models of T . b . [5] , [30] , [52] . For the current study , we opted to replace RLuc for in vivo imaging by a red-shifted firefly luciferase reporter for several reasons . Coelenterazine , used for RLuc activity detection , has rather unfavourable kinetics in vivo . For example , it is thought not to pass the blood-brain barrier due to the abundance of P-glycoprotein pumps in brain vascular endothelium , whose efflux activity restricts access of coelenterazine to the parenchyma unless there is severe dysfunction of the blood-brain barrier [57] . Although this might be interesting for studying advanced neurological trypanosomiasis models , it does not reflect the precise timing of CNS infection by the parasite since trypanosomes have been observed in CNS before tight junctions are disrupted [58] . Novel variants of coelenterazine , such as ViViRen , are more resistant against auto-oxidation in serum , but similarly to coelenterazine , the eventual BLI signal highly depends on the route of their administration [59] . Furthermore , all coelenterazine variants are more expensive than D-luciferin . Also , in contrast to coelenterazine variants , the distribution of D-luciferin in vivo is fairly well characterised and optimised protocols for administration and anaesthesia are available [60] . Therefore , we did not extend our research into red-shifted Renilla luciferases that have been described recently , neither did we compare different administration routes [61] . We prefer IP injection which is the most practical and appropriate for D-luciferin , the substrate of all firefly and beetle luciferases , including the red-shifted variants such as CBR and PpyRE9 [37] . When reporter genes are to be compared , one should use the same vector background and the same assay to standardise expression and activity measurement [62] . We expressed the different reporter genes , RLuc , CBR and PpyRE9 in the same trypanosomal expression vector , pHD309 . This vector can be integrated in the β-tubulin locus of trypanosomes and is one of the few known expression vectors that has been proven successful in T . b . gambiense through simple electroporation . Genome data show that this locus consists of multiple tandem repeats , thus allowing multiple integrations , while among different T . b . strains , a wide variation in β-tubulin copy numbers has been reported [63] , [64] . In our strains we did not assess the number of β-tubulin copy numbers or the number of reporter gene copies integrated in this locus but rather used hygromycin resistance to select modified clones of the same strain with equal IC50 values for comparison . Using this approach , our study confirms the higher catalytic activity of P9 compared to CBR in T . b . brucei AnTaR 1 , both in vitro and in vivo with 10 to 20-fold higher signals generated by P9 as previously described by Branchini et al [37] . However , it should be noted that higher expression of reporter genes in T . b . can also be obtained by modifying the expression vector for RNA polymerase I dependent transcription and by replacing the 5′ and 3′ untranslated regions flanking the reporter gene with sequences that flank highly expressed genes , as demonstrated in McLatchie et al [40] . One of the limitations of this study is that we did not correlate the in vivo BLI data with precise quantification of the parasites , e . g . by real-time RT-PCR . However , we microscopically estimated parasitaemia in blood to allow comparison between luminescent clones of the same strain early in infection ( day 4 to day 7 ) . Furthermore , for CBR-modified T . b . brucei AnTaR 1 , T . b . rhodesiense RUMPHI and T . b . gambiense LiTaR 1 , clones were selected with the same level of luciferase activity per living cell , making differences in the in vivo BLI responses dependent on parasitaemia and distribution in organs . However , we did not verify whether the expression of the reporter genes remained similar during the infection . At least in vitro testing did not reveal changes in hygromycin resistance or luciferase activity during a 6 week period in culture . With an inoculum of 2×104 to 2×105 parasites , the red luciferase trypanosomes can be tracked from day 1 post-infection , although the sensitivity of detection was higher with P9 than with CBR . This higher sensitivity of P9 is important to confirm infection prior to treatment where treatment is to be given very early after infection . Later in the infection , the added detail gained from higher sensitivity of P9 becomes less important since the low background signal inherent to in vivo luminescence imaging allows similar localisation of the trypanosomes within mice infected with CBR-modified trypanosomes , as previously reported by Close et al for FLuc and Lux comparison [65] . Compared to P9 and CBR , signals from RLuc modified trypanosomes are mainly from the abdominal region , which can be explained by the poor absorbance of ViviRen after IP administration . In contrast with their poor performance in vivo , protected coelenterazine compounds , like EnduRen ( Promega ) , perform very well in vitro as was demonstrated in a luminescent multiplexed viability assay developed to monitor the response of RLuc-modified trypanosomes for compound screening [52] . Our collection of luminescent trypanosomes contains strains with very diverse infection outcome . Although less virulent than T . b . gambiense LiTaR 1 , T . b . brucei AnTaR 1 strains induce a subacute infection . The sustained high parasitaemia in the mice might take its toll on the cardiovascular , hepatic or splenic physiology causing some animals to die early in infection . The occurrence of hepatosplenomegaly is an indication of high virulence in T . b . brucei and T . evansi strains [66] , [67] . Other studies record a less virulent phenotype of T . b . brucei AnTaR 1 that may be related to the number of passages in mice or in vitro or to the rodent species or breed . Our data suggest that the integration of the expression plasmid did not change the growth phenotype , as reflected by the in vitro doubling times of the wild-type and the recombinant strains . Because of this high lethality in mice , it would be advantageous to use rats as host , which are known to control the infection longer than mice [68] , [69] . Unfortunately , in vivo imaging on larger rodents does not yet seem very efficient [70] . The T . b . gambiense LiTaR 1 strain induces the same high parasitaemia levels and early lethality as the monomorphic T . b . brucei Lister 427 strain [30] . In our in vivo experiments , the mice did not survive the first peak of parasitaemia . The T . b . gambiense LiTaR 1 has undergone numerous passages in vivo and has become monomorphic and highly virulent and thus very different from wild-type strains that cause chronic gambiense sleeping sickness . The investigation of T . b . gambiense LiTaR 1 is important since laboratory accidents have shown that this strains is very virulent also in humans [71] . T . b . rhodesiense RUMPHI induces lower peak parasitaemia , does not show the survival bottleneck that characterises T . b . brucei AnTaR 1 , but still , survival is limited to approximately one month . This strain is fully tsetse transmissible in G . morsitans morsitans ( personal communication; Dr . Jan Van den Abbeele ) and under pressure of normal human serum in vitro , the serum resistance associated ( SRA ) protein expressing phenotype can easily be obtained ( data not shown ) . All experiments in the present study were performed without normal human serum pressure . It is not known if SRA expression would alter the virulence phenotype in the rodent model . In contrast to the other strains in our luminescent collection , T . b . gambiense 348 BT mimics very well the chronic phenotype of classical gambiense sleeping sickness in humans . Although both RLuc and CBR modified parasites were visible in BLI early after infection of immunosuppressed mice , the BLI signal strongly decreased later during the infection reflecting progression to a chronic infection . Such chronic infections , similarly to what is observed in humans , are characterised by the absence of detectable parasites in blood but by presence of the parasites in the CNS . With T . b . brucei , such chronic infections with parasites only present in the CNS can only be obtained by subcurative treatment of the mice [17] . In our in vivo experiments , some animals infected with luminescent T . b . gambiense 348 BT remained negative in BLI . This may correspond to the silent infection phenotype of T . b . gambiense , in which trypanosomes are present in very low numbers and can only be traced back by histopathological or molecular assays [5] . In addition , our experiments show that with the same parasite strain a variety of infection phenotypes can be encountered in different individuals from the same outbred mouse strain , in casu OF-1 [20] , [21] . With all pleomorphic parasite strains , the CNS infection was confirmed ex vivo . The sites that were found infected are not different from those previously described with other T . b . gambiense and T . b . brucei strains [5] , [29] , [41] , [72]–[74] . It should be noted that we cannot report on the presence of the trypanosomes in the dura mater or the subarchnoid space because this tissue was removed during brain extraction . Furthermore , our sample size and time of brain extraction did not allow description of differences in CNS invasion between the different trypanosome strains . Next to their diversity in virulence phenotype , the bioluminescent trypanosome strains display a wide diversity in drug sensitivity phenotype . T . b . brucei AnTaR 1 is less sensitive to eflornithine , nifurtimox and diminazene diaceturate than the human infective strains . Among the latter , T . b . gambiense 348 BT shows a particularly interesting phenotype . Compared to the other strains , it is less sensitive to melarsoprol , pentamidine and , surprisingly , to suramin . This strain originates from the HAT focus of Mbuji-Mayi in East Kasai , DRC where melarsoprol treatment failure rates of about 40% have been reported [47] . Other strains isolated from the same HAT focus have been found to be cross-resistant to melarsoprol and pentamidine , a phenotype that was linked to a chimaeric aquaglyceroporin 2/3 gene in their genome [48] . Sequencing confirmed the presence of the same mutation in T . b . gambiense 348 BT ( personal communication , Dr . Mäser Pascal ) . A lower susceptibility of T . b . gambiense than T . b . brucei to suramin has been described previously , but an even higher resistance has been described in T . evansi [53] , [54] . T . b . rhodesiense RUMPHI is less sensitive to eflornithine than the two T . b . gambiense strains in our collection but is fully sensitive to pentamidine . This finding is consistent with recent evidence that first stage rhodesiense HAT patients can be cured with pentamidine , a drug that is much less toxic than suramin [75] , [76] . Yet , suramin is still the first line treatment for first stage rhodesiense HAT [1] . We established a very diverse collection of bioluminescent T . b . brucei , T . b . gambiense and T . b . rhodesiense strains that is now available for in vitro and in vivo studies on trypanosomiasis research . For in vivo monitoring of murine infections , we recommend the use of trypanosome strains transfected with red-shifted luciferases , such as PpyRE9 , above the blue RLuc . Recently it was shown that PpyRE9 was also advantageous to monitor disease progression with T . b . brucei GVR35 . However , the study of T . b . gambiense field strains is far more relevant for sleeping sickness rodent models , because this parasite subspecies causes more than 95% of the cases of HAT and severely differs in virulence from T . b . brucei , causing more chronic and protracted infections in murine models . Furthermore , we modified a T . b . gambiense strain that harbours the AQP2/3 chimaeric gene , which was recently shown to be responsible for pentamidine and melarsoprol cross-resistance in field isolates and quite possibly contributes to the high treatment failure rates seen in the DRC and Southern Sudan . | Research on African trypanosomes heavily relies on rodent infection models . One way to reduce the number of laboratory rodents used in each experiment and effectively follow the progression of the infection in the same animals is to use genetically modified trypanosomes that allow monitoring of the infection over time with bioluminescence technology , without having to sacrifice the animals at multiple time points . In this study , we were able to establish a collection of bioluminescent strains of all three subspecies of Trypanosoma brucei ( T . b . ) , including T . b . gambiense and T . b . rhodesiense that cause human African trypanosomiasis ( HAT ) or sleeping sickness . Making use of bioluminescence assays , we demonstrate the diversity of our collection in terms of in vitro and in vivo growth , drug sensitivity and in vivo parasite distribution , including central nervous system tropism . Growth characteristics and drug sensitivity are not affected by the genetic modification with luciferase reporter genes . Trypanosome strains transfected with red-shifted luciferase reporter genes have several advantages compared to the corresponding blue luciferase modified strains . Red light is less absorbed in the blood than blue light , which should lead to higher sensitivity of detection . Furthermore , the substrates that drive the light reaction are better distributed through the body for the red luciferase than for the blue luciferase , which greatly improves spatial resolution of the infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonoses",
"medicine",
"and",
"health",
"sciences",
"african",
"trypanosomiasis",
"neglected",
"tropical",
"diseases",
"biology",
"and",
"life",
"sciences",
"vector-borne",
"diseases",
"tropical",
"diseases",
"trypanoso... | 2014 | A Panel of Trypanosoma brucei Strains Tagged with Blue and Red-Shifted Luciferases for Bioluminescent Imaging in Murine Infection Models |
The role of environmental factors in driving adaptive trajectories of living organisms is still being debated . This is even more important to understand when dealing with important neglected diseases and their vectors . In this paper , we analysed genetic divergence , computed from seven microsatellite loci , of 614 tsetse flies ( Glossina palpalis gambiensis and Glossina palpalis palpalis , major vectors of animal and human trypanosomes ) from 28 sites of West and Central Africa . We found that the two subspecies are so divergent that they deserve the species status . Controlling for geographic and time distances that separate these samples , which have a significant effect , we found that G . p . gambiensis from different landscapes ( Niayes of Senegal , savannah and coastal environments ) were significantly genetically different and thus represent different ecotypes or subspecies . We also confirm that G . p . palpalis from Ivory Coast , Cameroon and DRC are strongly divergent . These results provide an opportunity to examine whether new tsetse fly ecotypes might display different behaviour , dispersal patterns , host preferences and vectorial capacities . This work also urges a revision of taxonomic status of Glossina palpalis subspecies and highlights again how fast ecological divergence can be , especially in host-parasite-vector systems .
A capital step for species diversification is the initiation of some kind of disruptive selection , driving the newly diverged group of entities to some level of genetic adaptive divergence [1 , 2] . There has been a continuous debate on the respective role of geography and ecology in speciation , especially the speed at which these factors drive organisms to divergence [3] . These debates are important as they focus on key processes involved in evolution . For parasites and their vectors , the role of ecology and geography in driving divergence has important implications for management , as rapid evolution can occur in response to control practices or introductions to new environments [4] . This can have consequences on dispersal capacity [4] , behaviour [5] and vectorial capacities [6–8] . Tsetse flies ( Diptera: Glossinidae ) are the sole cyclical vectors of human ( HAT or sleeping sickness ) and animal ( AAT or nagana ) African trypanosomoses , two major plagues that are seriously impeding African development [9] . Among these , Glossina palpalis palpalis and Glossina palpalis gambiensis , which are major vectors of both HAT and AAT , have recently been the subject of several population genetics studies ( see [10] for a review ) . These studies , mainly based on spatio-temporal variation at microsatellite loci , have recurrently revealed some degree of genetic divergence , in some cases above the reasonable amount expected from geographically based population structure [11–14] . Because control programs against trypanosomoses often rely on tsetse eradication or suppression , it is important to specify the amount of such divergences and , if possible , if it could be linked to some ecological factors . Indeed , adaptive divergence may be correlated to variation in behaviour , host preference ( or attractiveness to trapping devices ) and vectoring ability . In this paper , we combined and synthesized published and unpublished microsatellite data sets of these two taxa from populations sampled in West Africa and central Africa . We analysed the whole data set in order to evaluate the genetic divergence between the two taxa as assessed with microsatellite markers and then we analysed separately G . p . gambiensis and G . p . palpalis in order to assess the respective role of geographic distance , date of capture ( time distance ) , landscape type and river basin in determining the level of genetic divergence of tsetse flies . The observed levels of divergence provide support for changes in the taxonomic status of these subspecies . Furthermore , based on both genetic and ecological criteria , we propose that several additional taxonomic groups should be recognized . The importance of these findings in developing novel control strategies and facilitating future research endeavours is discussed . Subspecies G . p . gambiensis and G . p . palpalis may have split no more than 13000 years ago [15 , 16] . The ecotypes evidenced in the present study necessarily are much younger and illustrate on how swift ecological divergence can be .
Study sites are located as represented in Fig . 1 . The species , country , landscape type , river basin , date of capture , GPS coordinates and sample sizes are presented in Table 1 . Raw data are available in S1 Table . Most of the samples studied in this paper were already used and genotyped for publications relating to other , though related purposes . These papers are cited in Table 1 and sites samples can be seen in the Fig . 1 . Folonzo sample was never published and was sampled during April 2007 following the same method as in [17] . Guekedou sample was never published and was sampled during May 2007 following the same procedure as in [18]; Senegal 1 and Senegal 3 samples were never published and were kindly provided by the Insect Pest Control Laboratory , Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture and sampling followed the same procedures as described in [9] . Azaguié sample was never published and was sampled and genotyped for another project of our team by S . Ravel in collaboration with Dr D . Kaba ( Pierre Richet / Institut National de Santé Publique , Abidjan , Ivory Coast ) and Dr G . Acapovi-Yao ( Laboratoire de Zoologie , Université d’Abidjan-Cocody , Abidjan , Ivory Coast ) ( Acapovi-Yao et al . , manuscript in preparation ) . Published papers are available at http://gemi . mpl . ird . fr/SiteSGASS/SiteTDM/ArtiPDF . html . These 28 samples summed to 614 genotyped individuals , with 9 unpublished samples . Genotyping of unpublished data followed the same protocol as described in [17] and [19] . For Mali 12 , Mali 8 , Senegal 1 and Senegal 3 , the genotypes of the flies were kindly provided by the Insect Pest Control Laboratory , Joint FAO/IAEA Program of Nuclear Techniques in Food and Agriculture and protocols were the same . Some of analyses undertaken do not tolerate missing data . For the sake of consistency between all analyses , only complete genotypes at seven loci were kept . These loci were: Gpg55 . 3 ( X linked ) [20]; B104 ( X linked ) , B110 ( X linked ) and C102 that were kindly supplied by A . Robinson , Insect Pest Control Laboratory ( formerly Entomology Unit ) , Food and Agricultural Organization of the United Nations/International Atomic Energy Agency [FAO/IAEA] , Agriculture and Biotechnology Laboratories , Seibersdorf , Austria; pGp13 ( X linked ) and pGp24 [21]; and GPCAG [22] . Protocols followed what was described in references cited above ( e . g . [18] ) . All genotyping were handled or supervised by the same person ( SR ) who ensured perfect calibration of allele sizes across sub-samples . A total of 614 tsetse flies from 28 sites displayed a full genotype at the seven microsatellite loci . All genotypic data were coded as they appeared , hence males were coded as homozygous at X-linked loci . Sex information was missing in samples from Gambia and assessed through genotypes found on X-linked loci . All data sets were built in appropriate text files and converted with Create V 1 . 1 [23] into the appropriate format as needed except for bootstrap analysis with Phylip for which we used Convert V 1 . 31 [24] . Genetic distances were computed with MSA 4 . 05 [25] . We used a Cavalli-Sforza and Edwards chord distance matrix [26] for dendrogram construction with a Neighbour-joining tree ( NJTree ) [27] and for regression analyses , as recommended [28 , 29] . The NJTree dendrogram showing relationships between all tsetse subsamples was built with Mega V 5 [30] . Robustness of nodes was assessed through 1000 bootstraps over loci with Phylip v 3 . 68 [31] . For that purpose , nodes in G . p . gambiensis were studied after rooting the tree with Malanga subsample ( DRC ) , while for G . p . palpalis nodes , tree was rooted with Banjul North subsample ( Gambia ) . Sample sizes are represented in Table 1 . Relationships between genetic distances and the other parameters were tested with partial Mantel tests ( for robustness ) and also explored with linear regressions ( for illustrations and strength of signals measures ) . Explanatory variables and factors were the subspecies distance ( whether the two compared samples contain the same subspecies or not ) , geographic distance ( in km ) , time distance ( in days ) , landscape and river basin ( same or not ) . The different landscapes and river basins are presented in Table 1 . For the Mantel tests , these factors were coded as 0 when the two sites compared shared the same value ( e . g . both G . p . gambiensis ) or 1 when different . For the linear regression , factors were coded as "Same" when similar in both sites and a combination of two modalities when different ( e . g . Savannah-Coast ) . Mantel test for global data was undertaken to test for the effect of sub-speciation between G . p . palpalis and G . p . gambiensis . Because there are probably interactions with this effect , other factors were then analyzed more precisely within each subspecies separately . Partial Mantel tests were undertaken under Fstat V 2 . 9 . 3 [32] ( updated from [33] ) with 10000 Monte-Carlo randomizations of genetic distance matrix items . We also undertook Principal Component Analyses ( PCA ) on each sub-species data set . For this we used PCAGen 1 . 2 . 1 [34] that works on allele frequencies in subsamples and reorganize the data into a multidimensional space the metric of which is equivalent to Wright's FST [35] , i . e . the part of inbreeding that is explained by population subdivision . The significance of the first axes was tested with the broken stick criterion [36] and also with 10000 permutations of individuals across subsamples . We then submitted subsample coordinates of each significant axis to multiple regressions . The general model to start with was always of the form Axisi ∼ Lat + Long + Day + Landscape + RiverBasin + Lat:Long where i identifies the axis number being investigated , Lat and Long mean the latitudinal and longitudinal GPS coordinates in degrees , Day means the number of days after the oldest sub-sample , Lanscape is as described above , RiverBasin is the name of the river basin as described above and ":" stands for interaction between two explanatory variables . Here the variables were weighted for subsample sizes . All multiple regressions were undertaken under R [37] using sample sizes as weights . For all linear regressions , the best ( minimum ) model was selected after a stepwise procedure , using the Akaike Information Criterion [38] , significance tested with a F test and multiple comparisons ( when useful ) were done with the Student-Newman-Keuls ( SNK ) test . Order of entry of explanatory variables matters both in Fstat ( Mantel ) and R analyses . We thus chose to enter these variables following an order of decreasing importance we thought they would have: geographic distance , time , landscape , basin and interactions ( if any ) . Ecological factors were entered last to make sure the response was controlled for the other parameters . Null alleles and X-linked loci produce an artificial excess of inbreeding in subsamples that should not affect the tests in any predictable direction but a decrease in power . These issues are thus relevant only in those cases where tests do not appear significant . A coming work involving one of the authors ( TDM ) will be devoted to the robustness of different genetic distances to such issues ( manuscript in preparation ) . NJTrees were also built without X-linked loci , on females only and on males only . This did not change the general aspect of the tree even if a few populations happened to branch in slightly different places . These NJTrees can be seen in S1 File . The complete data set is available in S1 Table .
Geographic locations , landscape types and genetic relationships between all subsamples are presented in Table 1 , Fig . 1 and Fig . 2 . It can be seen that the two subspecies are clearly separated . In G . p . gambiensis , distinction between Savannah , Niayes and Coastal populations , in some instances , overcome geographic differentiation . This is particularly clear for samples from Gambia and Senegal ( Fig . 1 ) . For instance , as can be seen from Figs . 1 and 2 , subsamples Senegal 1 and 3 are genetically closer to Burkina-Faso and Mali sites ( Savannah ) than from the geographically closer Dakar , Pout ( Niayes ) , Missira , Banjul North and South ( Coast ) . In G . p . palpalis , Central and Western African flies are clearly separated and geography seems to be the predominant factor within each of the two zones . Again , genetic distances are quite pronounced and bootstrap values relatively high . Results of partial Mantel test for the whole data set provided a highly significant contribution of subspecies ( partial R2 = 0 . 65 , P-value<0 . 0001 ) . For G . p . gambiensis , partial Mantel test highlighted two major factors that best explained genetic distance between subsamples ( Table 2 ) . The first is geographic distance , which explains 41% of the variance , followed by landscape distances that explain 10% of the variance of genetic distances and are highly significant . Other parameters ( time and river basin ) contribute little to the coefficient of determination R2 , though significantly so . Regarding the linear model , the stepwise procedure could not simplify the model . Nevertheless , river basin distances did not display consistent results since the response mainly was due to higher genetic distances between sites from the same basin as compared to other comparisons . This incoherence , which probably comes from interaction with geographic distance , led us to remove this factor from the analysis . Results are presented in Fig . 3 . The total R2 = 0 . 66 . Here the main factor is geographic distance , followed by landscape distance . Both explained not less than 62% of the total genetic variance ( which is quite big given the variation expected for genetic distances ) . Time explained very little of the variance though significant: the more time between subsamples , the more genetic divergence between them . For landscape distances , paired comparisons led to the conclusion that genetic distances between similar landscapes were smaller than any other comparison , and that Niayes subsamples were always genetically more distant from the other sites than any other comparison . In G . p . palpalis subsamples , the Mantel test , partial R2 and corresponding P-value are presented in Table 3 . Only geographic distance displayed a significant effect here , with 34% of the variance explained . For the multiple regression approach , only geography seemed to play a significant role ( partial R2 = 0 . 34 , P-value = 0 . 0002 ) ( Fig . 3D ) . In particular , very high bootstrap values are observed between Central and West Africa and between Cameroon and RDC subsamples . For PCA analysis of G . p . gambiensis sub-samples , the first two axes appeared significant both with the broken stick criterion and with permutation testing ( P-value≤0 . 0001 for Axis 1 and P-value = 0 . 0345 for Axis 2 , permutation test ) . Axes 1 and 2 represent 41% and 16% of total inertia respectively . After stepwise procedures , Axis 1 is explained by all initial variables but Day ( Table 4 ) . By far the two most important variables are the latitude and the landscape that explain respectively 66% and 29% of the total variance in axis 1 ( both P-values<0 . 0001 ) . For the second axis , the minimum model was Axis2 ∼ Lat + Long + Landscape + RiverBasin ( Table 5 ) . Here , the most important variables are Lanscape and Latitude that respectively explain 39% and 30% of the total variance in axis 2 ( P-values<0 . 00001 ) . For PCA analyses of G . p . palpalis sub-samples , the first three axes appeared significant both with the broken stick and permutation tests , with permutation P-value≤0 . 0001 for the two first axes and P-value = 0 . 011 for the third . They respectively represent 34 , 27 and 17% of total inertia respectively . Here , variable Landscape was not introduced as it does not vary in the sampled zones for G . p . palpalis . For axis 1 , no simplification of the model was possible ( Table 6 ) . The only significant effect comes from the latitude which explains 95% of the total variance on axis 1 ( P-value = 0 . 0038 ) . On axis 2 , only two variables stayed in the model after the stepwise procedure ( Table 7 ) . However only longitude really mattered and explained no less than 94% of axis 2 ( P-value≤0 . 0001 ) . Finally , for axis 3 , the minimum model was Axis3 ∼ Lat + Long + RiverBasin + Lat:Long and the most important explanatory variables were the river basin , explaining 81% of axis 3 variance ( P-value = 0 . 0053 ) , and the interaction between latitudinal and longitudinal coordinates that explained 14% of axis 3 variance ( P-value = 0 . 0291 ) ( Table 8 ) .
The importance of geographic distance for determining genetic relationships between tsetse populations has been recurrently reported [9 , 19 , 39] . Its predominant effect above the effect of river basin was an expected result , at least for G . p . gambiensis [9] and is newly demonstrated here for G . p . palpalis . The genetic distance that separates the two subspecies and the high bootstrap level obtained with microsatellite markers ( known for their homoplasic nature ) are advocating for a revision of the nomenclature of those taxa as different species . This is also in line with the biological definition of species , although the usefulness of such a concept is debatable [40] , since the heterozygous males of the F1 crossing between these taxa are completely sterile [41 , 42] which leads to a very sharp allopatry between them in Ivory Coast [43 , 44] . Moreover , these taxa can be discriminated on a morphological basis using the size of the palette of the inferior claspers ( larger in G . p . gambiensis ) and the length of hairs on the inferior claspers ( longer in G . p . gambiensis ) [45] . This is even more justified as we also find evidence in the present paper of the existence of subunits within these two taxa , some of which are of an ecological nature . The stronger impact of river basins on G . p . gambiensis than on G . p . palpalis is not surprising , taking into account that the savannah environment of the former makes it much more difficult to cross the interfluve than the dense forest environment of the latter . Time did not play a very pronounced effect on G . p . gambiensis and apparently had no effect on G . p . palpalis . For the latter , smaller sample sizes are probably the cause of this absence of detectable effect . For G . p . gambiensis , the significance of the effect is in line with genetic drift due to small effective population sizes that could be estimated in several studies in these taxa [9 , 10 , 12 , 17–19 , 39 , 46] but also in other tsetse taxa ( see [47] for review ) . It highlights the need to take into account this factor in population genetics studies and the necessity to avoid pooling individuals that do not belong to the same cohort , in particular to estimate population differentiation , isolation by distance and migration . In G . p . gambiensis , an important and significant effect of landscape where tsetse flies are found was evidenced . Interestingly , in several instances , genetic distances between subsamples from different landscapes are far above those between subsamples from the same landscape , even between the most remote ones . This strong impact of landscape was confirmed by the regression analyses where this variable explained as much , and sometimes more , the genetic composition of G . p . gambiensis sub-samples . Our study also confirms the genetic isolation of G . p . gambiensis from the Niayes [12 , 48] which has led to an eradication program in Senegal ( http://www . fao . org/news/story/en/item/211898/icode/ ) . It is clear from the different analyses that tsetse from the Niayes ( Senegal ) represent an objective subspecies , adapted to a specific environment [48 , 49] . This subspecies is able to reproduce in the complete absence of perennial hydrographic network . Moreover , tsetse from savannah and those from coastal landscapes also represent original diverged entities that can deserve the denomination of ecotypes , if not subspecies . There is however no pre- or post-mating barriers between these taxa , as evidenced by successful mating observed between tsetse flies from the Niayes and savannah tsetse flies from Mali and Burkina-Faso [50] . They can thus be considered as subspecies . It is the first time that such subspecies are evidenced . It has to be underlined that the discovery of these ecotypes may have important consequences . In particular , data from several studies made in the coastal part of Guinea have shown that the G . p . gambiensis ecotype caught in the sleeping sickness foci of this country do not display any infection with the pathogenic trypanosomes usually identified ( including human and animal trypanosomes ) in this species . Nevertheless , G . p . gambiensis is the only vector of sleeping sickness there [51 , 52] . To what extent the fact that they constitute a distinct ecotype can be linked to a different vector capacity remains to be documented , but may be of paramount importance for control programmes against both human and animal trypanosomoses . It was also demonstrated that Trypanosoma brucei gambiense from Guinea were genetically very different than those from Ivory Coast , and that this was probably due to the fact that they were not transmitted by the same tsetse taxa , i . e . G . p . gambiensis of the coastal landscape for T . b . gambiense from Guinea , and G . p . palpalis for the T . b . gambiense from Ivory Coast [53] . For G . p . palpalis , the very high bootstrap values observed between Central and West Africa and between Cameroon and RDC subsamples suggest subspeciation , if not more , in the ecological sense of it ( adaptively divergent but not necessarily sexually isolated entities , see [40 , 54] ) . The existence of three subspecies ( or even species ) separating flies from West Africa ( Ivory Coast ) , South of Cameroon , Equatorial Guinea and DRC has already been suggested , based on mtDNA ( COI ) [13] and there are probably more than that [55] . Here , our seven microsatellite loci provide a strong confirmation that G . p . palpalis is a strongly heterogeneous taxon . Moreover , [56] found significant differences in the morphology of the head between G . p . palpalis from West Africa and DRC . Regression analyses on PCA axes also highlighted the relevance of river basins . Nevertheless , many sites in the range of this species are missing ( Gabon , Nigeria , Benin , Togo and Ghana ) and other environmental measures are missing as well . Future studies , implying GIS approaches should bring more information and more precision on the mechanisms of ecotype and population delimitations in tsetse flies . These observations are not only of academic interest as they have important repercussion as regard to vector control . Such ecological entities might represent different cases as regard to control success and reinvasion probabilities . It is thus key that such newly defined entities be ecologically characterized in order to compare their respective ecology ( hygrometric and temperature preferences , host preferences , symbiotic flora and vector competences ) . Mating preferences or differential competitiveness may also alter the success of sterile insect technique ( SIT ) if inappropriate ecotypes are released in the wrong environment . This thus opens the gate to many and very productive new research topics on trypanosomes and their vectors . It also highlights how useful genetic markers can be in exploring the ecology of difficult organisms . Finally , our results call for an urgent taxonomic review of the status of G . palpalis subspecies . The split between G . p . gambiensis and G . p . palpalis was dated as old as 3 . 2 million years according to COI mtDNA assuming molecular clock [13] . Nevertheless , this result is based on a single mtDNA marker known to behave very oddly at the beginning of a split ( for less than 1 million year the divergence can vary from 4 to 20% ) [57] . Moreover , because of their lack of neutrality [58–60] , mtDNA markers might not be ideal to estimate divergence time . We thus prefer relying on experts of the life history of tsetse flies who dated the split between the two sub-species around 13000 years ago ( around 91000 tsetse fly generations ) when the initial forest was separated into two isolated masses by drought [15 , 16] . It is probably the most parsimonious interpretation of tsetse flies history . The mean genetic distance between the two taxa is 0 . 65 ( which is very high for a distance bonded to 1 ) . It is 0 . 48 between savannah and coast subsamples , 0 . 53 between savannah and the Niayes and 0 . 5 between coast and the Niayes . Assuming constant microsatellite divergence with time , we can extrapolate that the ecological split in G . palpalis gambiensis occurred around 10000 years ago ( around 70000 generations ) , hence at the end of last glaciation . These estimates probably correspond to considerable overestimates as divergence speed probably strongly decreased as the two sub-species increased in population size ( which tends to freeze genetic drift ) when meteorological constraints were progressively relaxed at the end of the Würm ice age . These results provide another powerful illustration on how swift ecological divergences can occur , in particular in host-parasite-vector systems [4 , 54] . | The role of environmental factors in driving adaptive trajectories of living organisms is still being debated . This is even more important to understand when dealing with important and /or neglected diseases and their vectors . In this paper , we analysed genetic divergence , computed from several genetic markers , of 614 tsetse flies ( Glossina palpalis gambiensis and Glossina palpalis palpalis , major vectors of animal and human trypanosomes ) from 28 sites of West and Central Africa . We found that the two subspecies are so divergent that they deserve the species status . We found that G . p . gambiensis from different landscapes ( Niayes of Senegal , savannah and coastal environments ) were significantly genetically different , and thus represent different adaptive entities or even subspecies . We also confirm that G . p . palpalis from Ivory Coast , Cameroon and DRC are strongly divergent . These results provide an opportunity to examine whether these different types of tsetse fly might display different behaviour , dispersal patterns , host preferences and vectorial capacities . This work also urges a revision of taxonomic status of Glossina palpalis subspecies and highlights again how fast ecological divergence can be , especially in host-parasite-vector systems . | [
"Abstract",
"Introduction",
"Material",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Ecotype Evolution in Glossina palpalis Subspecies, Major Vectors of Sleeping Sickness |
Rabies is a major public health problem with a fatality rate close to 100%; however , complete prevention can be achieved through pre- or post-exposure prophylaxis . The rapid fluorescent focus inhibition test ( RFFIT ) is one of the recommended testing methods to determine the production of neutralizing antibodies after vaccination . Here , we report the development of a new monoclonal antibody ( mAb ) designed to react specifically with Rabies virus ( RABV ) phosphoprotein ( P protein ) , and the evaluation of its applicability to the RFFIT and its effectiveness as a diagnostic reagent for human rabies . The mAb KGH P 16B8 was produced to target the P protein of the Korean KGH RABV strain . An indirect immunofluorescence assay ( IFA ) was conducted to detect various strains of RABV in various cell lines . Alexa-conjugated KGH P 16B8 ( 16B8-Alexa ) was developed for the RFFIT . The IFA test could detect RABV up to a 1:2 , 500 dilution , with a detection limit comparable to that of a commercial diagnostic reagent . The sensitivity , specificity , positive predictive value , and negative predictive value of the RFFIT using 16B8-Alexa in 414 clinical specimens were 98 . 67% , 99 . 47% , 99 . 55% , and 98 . 42% , respectively . The results of the RFFIT with 16B8-Alexa were strongly correlated with those obtained using an existing commercial diagnostic reagent ( r = 0 . 995 , p<0 . 001 ) . The mAb developed in this study shows high sensitivity and specificity , confirming its clinical utility with the RFFIT to measure the rabies neutralizing antibody titer and establish a diagnosis in human . Thus , 16B8-Alexa is expected to serve as an alternative diagnostic reagent that is widely accessible , with potentially broad applications beyond those of the RFFIT in Korea . Further studies with 16B8-Alexa should provide insight into the immunological mechanism of the P protein of Korean RABV .
Rabies is the oldest and most fatal viral zoonosis that has been known to humans for at least 2 , 300 years . This disease remains a significant public health burden , and it is estimated that more than 60 , 000 people in over 150 countries worldwide die from rabies every year . In particular , Asia and Africa account for more than 95% of the global prevalence of human rabies [1] . Rabies causes inflammation in the central nervous system of warm-blooded animals , including humans , and is usually transmitted to humans through bites or scratch wounds from animals infected with the virus . Approximately 99% of all cases of human rabies are caused by virus transmission from dogs [2] , although wild animals such as bats , raccoons , cats , and foxes may also transmit rabies to humans . Korean cases of rabies in animals were officially documented for the first time by Japanese researchers in 1907 during the Japanese occupation . Rabies cases were variously observed both in humans and animals until 1984 , and no case occurred until 1992 as a result of preventive measures taken . However , rabies cases reappeared in 1993 in animals , followed by continuous case reports since then . There was no incidence of rabies in humans from 1995 to 1998 , but it reappeared in 1999 , with six cases reported from then until 2004 [3] . Among recently isolated strains , the complete genome sequence of the Rabies virus ( RABV ) KGH strain has been determined . KGH was the first human RABV strain isolated in Korea from hair follicles of a rabies patient whose symptoms developed following a raccoon bite in 2001 . The entire KGH genome is 11 , 928 nucleotides in length . In comparison to the other 40 RABV strains whose complete genomes have been sequenced , there is one unique amino acid replacement in the KGH strain in a region of the phosphoprotein ( P protein ) , which is related to STAT1 control . Phylogenetic analysis showed that KGH is most closely related to the NNV-RAB-H strain isolated in India and the transplant RABV serotype 1 strain [4] . Clinical manifestations of rabies have two forms: the furious form and the paralytic form . Approximately 80% of patients with rabies exhibit the furious form ( also termed “encephalitic rabies” ) , which is characterized by prominent autonomic nervous system disorders , such as hypersalivation and sweating . As a classical sign , confusion , aggressiveness , hydrophobia or aerophagia may occur due to infection of the nervous system or hypersensitivity of the sensory organs; moreover , impairment of consciousness , paralysis and multisystem organ failure could result in death [5 , 6] . Approximately 20% of patients with rabies show the paralytic form , which present symptoms such as numbness and muscle weakness without accompanying brain inflammation or hydrophobia [5 , 6] . This variation makes it difficult to establish a differential diagnosis of rabies from other central nervous system diseases , which can result in a lack or delay of effective treatment , leading to death of the patient [7] . The lethality of rabies is nearly 100% upon presentation of symptoms; however , infection is completely preventable through pre-exposure prophylaxis ( PrEP ) or timely post-exposure prophylaxis ( PEP ) [8] . The World Health Organization ( WHO ) recommends PrEP for people with a high risk of exposure to rabies based on geography or occupation , including laboratory workers , as well as travelers planning to visit rabies high-risk areas and children living in or visiting rabies-affected areas [1] . PEP includes disinfection and treatment of the wound area as early as possible , and administration of a vaccine alone or with rabies immunoglobulin , depending on the type of exposure . The rapid fluorescent focus inhibition test ( RFFIT ) is considered the gold standard method for determining the level of RABV neutralizing antibodies . The WHO recommends 0 . 5 international units ( IU ) /mL by RFFIT as a neutralizing antibodies indicator of adequate vaccination in human at risk for rabies exposure [1] . A rabies antigen-specific antibody is required to detect infected cells through the RFFIT . While antibody reagents for this test are commercially available , these often pose an economic burden and are difficult to distribute widely to the areas that are the most in need . Since the amendments to the Enforcement Regulations of the Medical Device Act came into effect in 2011 , it has been increasingly difficult to import RABV antibodies in Korea . Therefore , there is an urgent need for a supply of an alternative antibody that is applicable in diagnostic testing for the detection of RABV antigen and measurement of the neutralizing antibody titer . In addition , although only RABV belonging to genotype 1 has been reported in Korea to date , bat species that are potential hosts of bat-related lyssavirus inhabit Korea , thereby posing an additional risk . Since P protein contains a broadly cross-reactive epitope for all lyssaviruses , monoclonal antibodies ( mAbs ) targeting P protein have been proven to be potentially useful for diagnosis and serotyping [9] . To this end , in this study , we developed an Alexa-conjugated monoclonal antibody ( mAb ) that specifically combines with P protein , using the first domestic KGH strain isolated from a human . The utility of this fluorescently labeled KGH P protein-specific mAb was evaluated through its application to the RFFIT , using 414 clinical specimens .
This study was approved by the Institutional Review Board of the Korea Centers for Disease Control and Prevention ( KCDC; IRB No . 2015-07-05-R-A ) and the data were analyzed anonymously . Total RNA of the KGH isolate was amplified from virus-infected mouse neuroblastoma ( N2a ) cells ( American Type Culture Collection [ATCC] CCL131 ) , using the RNeasy Plus Mini Kit ( Qiagen , Hilden , Germany ) . cDNA was synthesized from viral RNA by reverse transcription-polymerase chain reaction ( PCR ) and used as template to amplify the KGH P protein-coding sequence . The oligonucleotide was amplified with two primers ( P Ab_F: 5ʹ-CAC CAT GAG CAA AAT CTT TGT CAA-3ʹ , P Ab_R: 5ʹ-GCT GGA TAC ATA GCG ATT CAG ATC-3ʹ ) . The thermocycling parameters were 30 cycles of 30 s at 94°C , 30 s at 60°C , 1 . 5 min at 72°C , and a final extension for 5 min at 72°C . The 1353-bp PCR products were visualized by agarose gel electrophoresis and purified using the QIAquick Gel Extraction kit ( Qiagen , Hilden , Germany ) . The fragment was cloned into the pBAD202 Topo vector ( Invitrogen , Waltham , MA , USA ) and expressed in Escherichia coli . The pBAD202/KGHP clone insert was verified through sequencing . Protein expression and mAb production were conducted by Youngin Frontier Co . Ltd . To determine the optimum conditions for the most effective KGH P protein expression , competent-cell type , isopropyl β-d-1-thiogalactopyranoside ( IPTG ) concentration , temperature , and time were evaluated . The pBAD202/KGHP was transformed into BL21 ( DE3 ) , BL21 ( DE3 ) pLysS , BL21 codon-plus RIL , and Rosetta ( DE3 ) competent cells , respectively ( New England Biolabs , Ipswich , MA , USA ) . IPTG was used to induce protein expression , at concentrations of 0 . 1 mM , 0 . 4 mM , and 1 mM . The amounts of proteins expressed under incubation at 37°C for 4h at 200rpm and at 20°C overnight at 110rpm were verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) . Two buffers , 20 mM Tris and 50 mM NaH2PO4 , were tested to select the optimum buffer condition for purifying the expressed protein by Ni-NTA His-tag affinity chromatography . Protein expression was also verified through SDS-PAGE in this case . The verified protein was used to produce the specific mAb for KGH P protein ( KGH P 16B8 ) according to the standard protocol for mAb production outlined by Youngin Frontier . To allow for its utilization in the RFFIT , Alexa Fluor 488 fluorescent particles were conjugated with KGH P 16B8 ( 16B8-Alexa ) using the Alexa Fluor 488 protein labeling kit ( Molecular Probes Inc . , Eugene , OR , USA ) according to the manufacturer’s instructions . An indirect immunofluorescence assay ( IFA ) was conducted to test whether KGH P 16B8 can detect antigens from diverse RABV strains . Baby hamster kidney ( BHK ) -21 cells ( ATCC CCL10 ) infected with two reference RABV strains , challenge virus standard ( CVS ) -11 ( ATCC VR959 ) and Evelyn-Rokitnicki-Abelseth ( ERA ) , and the Korean isolate KGH were used to produce antigen slides of RABV . BHK-21 cells were grown in Dulbecco’s modified Eagle’s medium ( Life Technologies , Waltham , MA , USA ) supplemented with 10% fetal bovine serum ( Life Technologies ) inactivated at 56°C for 30 min , penicillin-streptomycin ( Life Technologies ) , and l-glutamine ( Life Technologies ) . To evaluate any effect of the cell line type used for the test , antigen slides were produced with N2a cells according to the method described above . The cells were grown in Dulbecco’s modified Eagle’s medium-high glucose ( Life Technologies ) supplemented with 10% inactivated fetal bovine serum and penicillin-streptomycin . Rabies antigen slides were prepared to contain 3 , 000 cells/well using each cell type infected with each virus strain for 48 h . Rabies direct fluorescent antibody ( DFA ) reagent ( Millipore , Billerica , MA , USA ) , which is commercialized as a diagnostic antibody and commonly used for the detection of RABV antigen , was used as a positive control . For the IFA , KGH P 16B8 was diluted with phosphate-buffered saline ( PBS ) including 1% bovine serum albumin at 1:100 , and was added to each well of the antigen slides . The slides were incubated in a humid chamber at 37°C for 30 min and were washed by soaking in PBS for 10 min . After air drying , fluorescein isothiocyanate ( FITC ) -labeled anti-mouse immunoglobulin G conjugate ( Jackson ImmunoResearch , West Grove , PA , USA ) was diluted with PBS including 0 . 00025% Evans blue at 1:300 and was added to each well of the slides . The slides were incubated in a humid chamber at 37°C for 30 min and washed by soaking in PBS for 10 min . After mounting with a cover glass , the slides were observed within 2 h under a fluorescence microscope Axioskope2 ( Zeiss , Oberkochen , Germany ) at a magnification of 400× . To evaluate the RABV detection capacity of 16B8-Alexa , its detection limit was measured using a RABV antigen slide with DFA reagent as the positive control . 16B8-Alexa was serially diluted five times from 1:100 to 1:12 , 500 with PBS including 0 . 00025% Evans blue . It was then placed on each well of the slides that were incubated in a humid chamber at 37°C for 30 min and washed by soaking in PBS for 10 min . After mounting with a cover glass , the slides were observed within 2 h under a fluorescence microscope at a magnification of 400× . The results obtained through the RFFIT using the total of 414 clinical specimens were analyzed according to the value of 0 . 5 IU/mL , which was established as an indication of adequate vaccination in humans at risk of rabies exposure [1] . Accordingly , for qualitative analysis , the RFFIT results were classified into two categories: a value of 0 . 5 IU/mL or higher was considered positive , and 0–0 . 49 IU/mL was considered negative . The applicability of 16B8-Alexa to the RFFIT was evaluated in terms of its diagnostic sensitivity , specificity , positive predictive value ( PPV ) , and negative predictive value ( NPV ) . The correlation coefficient between the results of the RFFIT using 16B8-Alexa and the DFA reagent was evaluated by Pearson's correlation analysis . A P-value <0 . 01 was regarded to indicate statistical significance . IBM SPSS Statistics , Version 21 . 0 was used for all statistical analyses .
The RFFIT was conducted using 414 clinical serum specimens to evaluate the utility of 16B8-Alexa as a diagnostic mAb ( S1 Table ) . The results of the RFFIT using 16B8-Alexa were compared to those obtained using the DFA reagent . Reference serum was used in each experiment to calculate the antibody titers ( IU/mL ) . High sensitivity and specificity values of 98 . 67% and 99 . 47% , respectively , were obtained in the RFFIT using 16B8-Alexa . In addition , there was very good ( 99 . 03% ) agreement between the results obtained with 16B8-Alexa and DFA; the PPV and NPV of RFFIT using 16B8-Alexa was 99 . 55% and 98 . 42% , respectively ( Table 1 ) . Discordant results were observed in four samples for which the titer ranged between 0 . 42 and 0 . 53 IU/mL . Owing to the nature of the experiments , the detection limit varied between experiments . In this study , the measurable minimum titer was 0 . 02 IU/mL . Therefore , as it was not possible to measure the IU value in cases without a neutralizing antibody titer , immeasurably low values were substituted with 0 . 01 IU/mL in correlation analysis to minimize any impact of this uncertainty . For the same reason , the measurable maximum in this study was set to 41 . 29 IU/mL , and data for samples with immeasurably high antibody titers were substituted with 41 . 30 IU/mL in correlation analysis to minimize any impact of this uncertainty . Pearson's correlation analysis showed a significantly high level of correlation ( r = 0 . 995 , p<0 . 001 ) between the DFA reagent assay results and the 16B8-Alexa assay results ( Fig 4 ) .
In this study , we developed a new mAb that reacts specifically with the RABV P protein and confirmed its applicability to the RFFIT . Numerous previous studies have reported antibodies developed as diagnostic reagents for rabies detection that target the N protein [12 , 13] or G protein [14 , 15] . N protein is related to the immune response and pathogenicity in the host . Because of its high inter-species conservation and its high-level production upon infection , N protein is frequently used as a target protein in rabies diagnosis . G protein is a representative membrane protein , and is the only antigen that induces virus-neutralizing antibodies . The amount of each type of protein in the virus varies slightly among studies . Nonetheless , N protein is generally reported to be present in the largest amount ( 1325 or 1800 copies ) , followed by P protein ( 691 or 950 copies ) [16] . P protein is a multi-functional protein interacting with N ( N-P ) and is a key component and regulatory protein of the P-L complex for viral genome replication [17] . P protein is detected in both the virus and virus-infected host cells , in two forms: a major hypo-phosphorylated 37-kDa form and a minor hyper-phosphorylated 40-kDa form [18] . mAbs targeting P protein have been developed in some studies [19 , 20] . In this study , 16B8-Alexa was developed using a KGH P protein-specific mAb . 16B8-Alexa detected the viral antigen in two types of cells infected with RABV , with no significant difference in detection ability as compared to existing diagnostic reagents . Specific granule morphology was also clearly observed in infected cells . These findings confirmed that 16B8-Alexa can be applied as a diagnostic reagent to specifically target the RABV P protein . In the preliminary experiments , the sensitivity of Alexa Fluor was found to be higher when used as a marker for labeling KGH P 16B8 than that of FITC . Although fluorescein is more frequently used for mAb development , Alexa Fluor exhibits stronger fluorescence and is less pH-sensitive; thus , it is a safety-certified fluorescent particle [21] , and was chosen as the fluorescent marker in this study . Moreover , the detection limit of 16B8 Alexa in the RFFIT was not inferior to that of existing diagnostic reagents . To date , 14 virus species , including RABV , have been identified in the genus Lyssavirus [22] . Based on phylogenetic analysis , these viruses are categorized into seven major genotypes , with the prototype RABV belonging to genotype 1 [23] . In general , the prototype RABV belonging to genotype 1 is the main causal agent of rabies , but the disease can also rarely be caused by other lyssaviruses [24] . For this reason , most of the commercial antibodies available for rabies diagnosis are designed to detect rabies-related lyssaviruses along with RABV most of the cases [25] . The majority of diagnostic laboratories utilize commercial reagents [26] , whereas some utilize in-house products [15 , 27 , 28] . Both polyclonal antibodies and mAbs , which are specific to the entire virus or to the nucleocapsid protein and are conjugated with a fluorophore , can be used for rabies diagnosis . The Office International des Epizooties suggests that the sensitivity and specificity of any produced fluorescent antibody used in the fluorescent antibody test ( FAT ) should be verified through thorough validation and should be able to detect other lyssaviruses as well [29] . In this study , the ability of 16B8-Alexa to detect the prototype RABV was confirmed; however , it was not possible to confirm its ability to detect rabies-related lyssaviruses owing to experimental limitations . This remains a constraint in applying 16B8-Alexa to the testing of all lyssaviruses that could give rise to rabies symptoms . The main testing method of focus in our study was the RFFIT , and 16B8-Alexa was confirmed to detect the CVS-11 strain , which is the only RABV strain recommended for use in the RFFIT . Therefore , 16B8-Alexa can be used in the RFFIT to measure the neutralizing antibody titer in the serum of vaccinated individuals . The applied significance of this work has been demonstrated , as the utility of 16B8-Alexa was evaluated using 414 clinical specimens . In reference to 0 . 5 IU/mL as the marker concentration for a sufficient vaccination effect suggested by the WHO [30] , 16B8-Alexa demonstrated high sensitivity and specificity , and the agreement between the two assay methods evaluated using 16B8-Alexa or DFA reagent was also significantly high . Furthermore , the PPV and NPV were both very high . Therefore , we suggest that 16B8-Alexa can be utilized in the RFFIT for the measurement and diagnosis of rabies neutralizing antibody titers . Discordant results using 16B8-Alexa and DFA were observed in four samples , all of which had titers close to the limit of 0 . 5 IU/mL and were thus considered to represent variations within the acceptable range . These discordant results could be associated with issues of experimental reproducibility rather than an inherent difference in the efficiency of the diagnostic reagent . The RFFIT applies relatively loose restrictions for standardization , as it is a cell-based bioassay using biological materials such as live viruses [31] . The tested clinical specimens included 22 serum samples of patients obtained from the KBN; seven of the patients had viral encephalitis and 15 had encephalitis , myelitis , and encephalomyelitis . All of these patients presented symptoms similar to those of rabies , including fever ( 68 . 18% ) , vomiting ( 45 . 45% ) , convulsion/numbness ( 40 . 90% ) , and headache ( 36 . 36% ) . Therefore , these samples were used to test the cross-reactivity of 16B8-Alexa . In the majority of patients with rabies , central nervous system infection symptoms such as encephalitis or encephalomyelitis are observed . Therefore , differential diagnosis is required using the neutralizing antibody titer of the sera of patients presenting such symptoms [32] . In the present study , the specificity using these specimens was 100% . The 16B8-Alexa developed in this study was shown to react strongly with both fixed and street RABV strains , and as it targets the P protein of RABV , the reagent is expected to be useful in Korea with high potential risk of bat-related lyssaviruses . Furthermore , as a reagent developed and available in the destination area , the following problems can be overcome: elimination of the economic burden associated with import; reduction in the time required for completing certain procedures by up to 25% of the previous level; and avoidance of any unnecessary efforts for passing through the complicated import procedures . As potential advantages , 16B8-Alexa can be applied to a diverse array of other testing methods for prototype RABV detection beyond those involving fluorescence microscopy , simply by replacing the marker substance with another . As 16B8-Alexa is a RABV P protein-specific mAb , its utility will also be valuable for conducting basic immunological research to better understand the functions of P protein . With future studies to evaluate its detection ability for other lyssaviruses and investigations into its robustness and stability , it is expected that 16B8-Alexa can resolve the current issues of the limited accessibility of existing diagnostic reagents , and will serve as a valuable alternative diagnostic reagent free of economic constraints . | Rabies represents a significant cause of fatality upon presentation of symptoms; however , pre- or timely post-exposure prophylaxis can provide complete protection to a population . The current reagents available for laboratory tests to determine the level of Rabies virus ( RABV ) neutralizing antibodies are not readily accessible in several regions , including Korea , and are associated with time and economic constraints owing to the import process . To resolve these issues , we developed a new monoclonal antibody with a fluorescent marker ( 16B8-Alexa ) that targets the RABV phosphoprotein from a recent Korean isolate . We show its good detection ability , specificity , and sensitivity , demonstrating highly concordant results with those obtained with a standard commercial reagent using over 400 clinical samples applied to a World Health Organization-recommended diagnostic test . Thus , 16B8-Alexa shows great potential to resolve the current limitations in rabies diagnosis and monitoring in Korea , and is expected to serve as a valuable research tool for immunological studies on RABV . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"post-exposure",
"prophylaxis",
"immunology",
"tropical",
"diseases",
"microbiology",
"light",
"microscopy",
"geographical",
"locations",
"viruses",
... | 2017 | Development and evaluation of an anti-rabies virus phosphoprotein-specific monoclonal antibody for detection of rabies neutralizing antibodies using RFFIT |
Plasmodium and dengue virus , the causative agents of the two most devastating vector-borne diseases , malaria and dengue , are transmitted by the two most important mosquito vectors , Anopheles gambiae and Aedes aegypti , respectively . Insect-bacteria associations have been shown to influence vector competence for human pathogens through multi-faceted actions that include the elicitation of the insect immune system , pathogen sequestration by microbes , and bacteria-produced anti-pathogenic factors . These influences make the mosquito microbiota highly interesting from a disease control perspective . Here we present a bacterium of the genus Chromobacterium ( Csp_P ) , which was isolated from the midgut of field-caught Aedes aegypti . Csp_P can effectively colonize the mosquito midgut when introduced through an artificial nectar meal , and it also inhibits the growth of other members of the midgut microbiota . Csp_P colonization of the midgut tissue activates mosquito immune responses , and Csp_P exposure dramatically reduces the survival of both the larval and adult stages . Ingestion of Csp_P by the mosquito significantly reduces its susceptibility to Plasmodium falciparum and dengue virus infection , thereby compromising the mosquito's vector competence . This bacterium also exerts in vitro anti-Plasmodium and anti-dengue activities , which appear to be mediated through Csp_P -produced stable bioactive factors with transmission-blocking and therapeutic potential . The anti-pathogen and entomopathogenic properties of Csp_P render it a potential candidate for the development of malaria and dengue control strategies .
The influence of the gut microbiota on the vector competence of disease vectors such as mosquitoes has gained increasing interest over the past decade [1]–[3] . Previous work has shown that co-infection of Anopheles mosquitoes with Plasmodium and with Serratia sp . or Enterobacter sp . bacteria leads to reduced Plasmodium infection [3] , [4] . Additionally , the presence of certain bacterial species in Aedes mosquito midguts leads to a lower intensity of dengue virus infection [5] . Studies have also shown that Anopheles and Aedes mosquitoes that have had their gut microbiota experimentally reduced via antibiotic treatment show higher Plasmodium and dengue virus infection levels , respectively , than do their untreated counterparts [6]–[8] . The anti-pathogen activity of mosquito midgut bacteria has been attributed to the elicitation of the mosquito immune system in some instances , and to direct anti-pathogenic activity of bacteria-produced molecules in others [9] . Activation of the IMD pathway , the major anti-P . falciparum immune pathway , has been shown to be mediated through an interaction between the pattern recognition receptor PGRP-LC and the midgut microbiota [10] . In turn , microbe-derived anti-pathogen factors have been characterized in some microbe-host interaction systems and include cytotoxic metalloproteases , hemolysins , antibiotics , haemaglutinins , proteases , prodigiosin pigments , and iron chelators ( siderophores ) [9] . In nature , bacteria commonly grow attached to surfaces in complex matrices of cells , proteins , polysaccharides , and DNA ( biofilm growth ) , rather than as single free-swimming cells ( planktonic growth ) [11] , [12] . Biofilm formation allows the bacteria to survive exposure to host-derived antimicrobial factors and other environmental stressors [11] , [12] . Furthermore , bacterial cells in a biofilm have quite different gene expression and metabolic profiles than do cells in a free-swimming planktonic state [11] . Studies of Pseudomonas aeruginosa colonization of the Drosophila melanogaster gut have shown that biofilm formation can dramatically affect dissemination in the hemolymph and fly mortality [13] . In this study , we show that a Chromobacterium sp . isolate , Csp_P , previously isolated from the midgut of field-collected Ae . aegypti mosquitoes [5] , exerts in vitro anti-Plasmodium and anti-dengue activity when grown under biofilm conditions . Csp_P can effectively colonize the intestines of the two most important mosquito disease vectors , An . gambiae and Ae . aegypti , where it blocks Plasmodium and dengue infection . It also exerts entomopathogenic activity against both larval and adult stages and could therefore be used for the development of a biocontrol agent . Csp_P's anti-pathogen activities appear to be mediated by stable secondary metabolites , suggesting that Csp_P is a source of potentially interesting candidates for the development of therapeutic and transmission-blocking drugs .
To assess the ability of Csp_P to colonize the mosquito midgut , we exposed antibiotic-treated mosquitoes to a sugar source containing106 colony forming units ( CFU ) /ml for Ae . aegypti or 108 CFU/ml for An . gambiae for 24 h and then dissected , homogenized and plated the midguts on LB agar plates at 3 days post-exposure . Treatment with antibiotics through the sugar meal was performed to remove the native microbial flora which can fluctuate in terms of load and species composition between individual mosquitoes of the same cage and generation , thereby complicating the interpretation of our data [7] . The presence of the native microbiota would also render it difficult to discriminate the Csp_P colonies from those of other species through visual inspection . Csp_P displayed an exceptional ability to rapidly colonize mosquito midguts , showing a prevalence of 80% in An . gambiae and 97% in Ae . aegypti cage populations at 3 days after exposure ( Fig . 1A ) . Average bacterial loads at this time point were approximately 105 and 104 CFU per midgut in Ae . aegypti ( Fig . 1B ) and An . gambiae ( Fig . 1C ) females , respectively . We also tested the ability of Csp_P to colonize the midguts of non-antibiotic treated mosquitoes . Because nearly all septic ( i . e . non-antibiotic treated ) An . gambiae mosquitoes had died two days after Csp_P introduction through sugar-feeding at 108 CFU/ml ( Fig . 2C ) , we were only able to assay prevalence and bacterial load of Csp_P at days one and two post feeding . At one day after Csp_P ingestion , we found that Csp_P was present in all sampled mosquitoes with an average bacterial load of 5 . 12×104 ( Figure 1D , E ) . At two days after Csp_P exposure , only 5% of Csp_P-fed An . gambiae were still alive ( Fig . 2C ) and Csp_P was detected in only one ( 12 . 5% ) of these remaining mosquitoes ( Fig . 1D , E ) . In septic Ae . aegypti mosquitoes that had fed on a 1010 CFU/ml Csp_P-containing sugar solution , we identified Csp_P in 79% of mosquitoes sampled on day 1 post feeding ( Fig . 1F ) . At three days after feeding on the Csp_P-containing sugar solution , approximately 30% of the Ae . aegypti were still alive ( Fig . 2D ) and Csp_P was detected in 15% of these mosquitoes ( Fig . 1F ) . These data suggest that Csp_P colonized the vast majority of An . gambiae and Ae . aegypti mosquitoes by day 1 post exposure and that Csp_P caused rapid mortality in most individuals . The small percentage that survived up to day 2 or 3 , post exposure , may have received a small dose of bacteria and succeeded in clearing it by the time they were dissected . It is difficult to compare the colonization efficiency between septic and antibiotic treated mosquitoes because the survival curves differ dramatically ( Fig . 2 ) . While it appears that Csp_P was better at colonizing the midgut of antibiotic treated An . gambiae ( Fig . 1A vs . 1D ) and Ae . aegypti ( Fig . 1A vs . 1F ) , our measurement does not take into account that individuals died much more rapidly in the septic population . This rapid mortality likely selected for mostly Csp_P negative individuals by day 2 and 3 post-feeding . We examined the influence of Csp_P midgut colonization on mosquito longevity by exposing antibiotic-treated An . gambiae and Ae . aegypti mosquitoes to a sugar source for 24 h containing Csp_P at a final concentration of 108 and 106 CFU/ml , respectively , and then monitoring survival . This treatment led to a decrease in the longevity of both species when compared to non-exposed control mosquitoes ( Fig . 2A , B ) . We repeated this experiment with septic ( i . e . not antibiotic treated ) An . gambiae and Ae . aegypti . We found that feeding on a sugar source containing Csp_P at a concentration of 108 CFU/ml resulted in rapid mortality of An . gambiae adult females ( Fig . 2C ) . Mortality of septic Ae . aegypti females was not increased after feeding on a sugar source containing Csp_P at a concentration of 106 CFU/ml but was dramatically increased when the sugar meal contained Csp_P at a concentration of 1010 CFU/ml ( Fig . 2D ) . These data suggest that Csp_P has strong entomopathogenic activity regardless of whether other microbes are present in the mosquito gut . We observed lower survival in septic An . gambiae and Ae . aegytpi after feeding on a blood meal containing Csp_P at a final concentration of 108 CFU/ml ( Fig . 2E , F ) . The stronger entomopathogenic effect upon Csp_P introduction through the blood meal was most likely because the mosquitoes received a large single bacterial dose upon bloodfeeding rather than the multiple low doses that would be expected during sugar feeding . It is also possible that Csp_P proliferated to high numbers in the nutritious blood . To study the influence of Csp_P on larval viability , we placed 2- to 4-day-old mosquito larvae in groups of 10 in pools containing 5 ml distilled water supplemented with 50 µl of a 1 . 0 OD600 liquid culture of Csp_P , and then monitored survival . This resulted in almost complete mortality of An . gambiae and Ae . aegypti larvae over a 3- and 2-day period , respectively , when compared to the control larvae that were exposed to the normal breeding water microbiota ( Fig . 2G , H ) . These studies suggest that Csp_P –mediated mortality may be the direct result of a mosquitocidal factor or systemic infection through dissemination into the hemolymph; alternatively , its colonization of the midgut ( or other tissues ) might cause mortality indirectly by interfering with vital functions of the mosquito . Studies of Pseudomonas aeruginosa colonization of the Drosophila melanogaster gut have shown that biofilm formation can dramatically affect both dissemination within the hemolymph and fly mortality [13] . Csp_P is capable of forming biofilms in vitro , though whether biofilm formation occurs within the mosquito midgut remains untested . C . violaceum produces cyanide at high cell density [17] , [18] via the cyanide-producing hcnABC operon , a behavior that is reportedly regulated by quorum sensing [18] , [19] . Cyanide production by bacteria has been shown to cause host mortality in both nematodes [20] and insects [21] . Chromobacterium subtsugae has previously been shown to exert oral toxicity in various insects of agricultural importance , but not in Culex mosquitoes [22] . To investigate whether the presence of Csp_P in the mosquito midgut could influence the infection of An . gambiae with P . falciparum and of Ae . aegypti with the dengue virus DENV2 , we assayed the infection of mosquitoes that had been exposed to Csp_P through sugar feeding 2 days prior to feeding on parasite- or virus-infected blood . Approximately one week after An . gambiae had fed on a P . falciparum gametocyte culture , parasite infection was assayed by counting oocyst-stage parasites on the basal side of the mosquito midgut . DENV2 infection of the midgut of Ae . aegypti was assayed through standard plaque assays 7 days after an infectious bloodmeal . All experiments were initiated using similar numbers of adult females for each treatment , but because Csp_P exposure causes high mortality in adults ( Fig . 2 ) , very few Csp_P-fed mosquitoes were still alive when the parasite and dengue infection assays were conducted . Nevertheless , we found that surviving mosquitoes exposed to Csp_P through sugar feeding prior to feeding on infectious blood displayed significantly increased resistance to P . falciparum infection and DENV infection ( Fig . 3 ) . The inhibition of P . falciparum infection was even greater when Csp_P was introduced through the gametocyte-containing blood meal at 106 CFU/ml ( Fig . 3B ) , an effect most likely attributable to the larger number of ingested bacteria . Csp_P may inhibit pathogen infection directly through physical interaction with the pathogens or the production of anti-pathogen molecules . Alternatively , Csp_P may indirectly inhibit Plasmodium or dengue by ( a ) altering the long-term physiology or health of the mosquito such that pathogen infection is inhibited , ( b ) triggering a mosquito anti-pathogen response or ( c ) selecting for individuals that are more fit to resist Csp_P as well as DENV and Plasmodium infection . However , Csp_P's in vitro anti-pathogen activity ( discussed below ) suggests it has the potential to directly inhibit pathogen survival in the mosquito gut . Further studies are necessary to elucidate the mechanism by which Csp_P inhibits the pathogens in vitro and in vivo . We have previously shown that the An . gambiae and Ae . aegypti midgut microbiota elicit basal immune activity by elevating the expression of several immune factors , including antimicrobial peptides and antipathogen factors [5] , [7] , [8] , [23] , [24] . To determine Csp_P's potency in inducing the mosquito's innate immune system , we exposed mosquito SUA-5B cells to various concentrations of Csp_P and assayed for changes in the activity of a Cecropin1 promoter driving the expression of a luciferase reporter gene . We exposed these same cells to Pseudomonas putida , a Gram-negative bacterium that belongs to a bacterial genus commonly found in mosquito midguts [25]–[27] . This experiment showed that Cec1 expression increased with increasing Csp_P exposure , providing evidence that Csp_P is a potent immune elicitor ( Fig . 4 ) . We also compared the transcript abundance of mosquito immune genes in midguts from antibiotic-treated naïve mosqutioes to those from mosquitoes that had been provided a sugar source spiked with Csp_P ( 108 CFU/ml for An . gambiae and 106 CFU/ml for Ae . aegypti ) 2 days earlier . We chose to assay gene expression at 2 days post exposure because this is the time at which increased mortality due to infection begins to occur . We hypothesized that infection levels and therefore any potential immune response would be high at this time . In Ae . aegypti , we found that cecropin E and G and defensin C displayed at least a 2-fold increase in transcript abundance in the midgut of Ae . aegypti colonized with Csp_P bacteria when compared to naïve controls ( Fig . S1A ) . In An . gambiae , we found non-significant trends toward increased transcript abundance of the Rel2 , FBN9 and cecropin genes and toward decreased transcript abundance of the defensin gene in the midgut tissue ( Fig . S1B ) . These data represent a single time point post-infection , and while it is possible that additional time points may reveal dynamic patterns of Csp_P-induced changes in gene expression , our results generally agree with the cell culture data , and as a whole show that Csp_P has an immune-eliciting capacity in the mosquito gut . To test whether Csp_P could exert a direct anti-Plasmodium or anti-dengue effect in vitro that is independent of the mosquito , we performed experiments in which parasite development and virus infectivity were assayed after exposure to various preparations of either planktonic or biofilm cultures of Csp_P . Planktonic-state Csp_P was obtained by culturing Csp_P in liquid LB at 30°C overnight on a platform shaker . Biofilm was produced by culturing Csp_P in LB without agitation in a polystyrene 24-well plate at room temperature for 48 h , unless otherwise indicated . The anti-dengue and anti-Plasmodium activity of the following five different preparations of Csp_P was then tested: ( a ) 1 ml ( 108 CFU/ml ) planktonic-state liquid culture , ( b ) 1 ml ( 109 CFU/ml ) biofilm supernatant consisting of liquid LB drawn off freshly cultured biofilm , ( c ) 5 mg ( 109 CFU/ml ) fresh biofilm resuspended in 1× PBS , ( d ) 5 mg dessicated biofilm prepared from biofilm collected 1–2 days prior to assay and allowed to completely dessicate at room temperature and then rehydrated in 1× PBS , ( e ) 5 mg heat-inactivated biofilm prepared by heating biofilm at 90°C for 24 h , collected 1 day prior to assay . Our in vitro assays showed that Csp_P exerts potent anti-Plasmodium activity against both asexual and sexual parasite stages . We exposed P . falciparum 3D7 asexual stage parasites to all five bacterial preparations in vitro . Because bacterial growth can interfere with determining parasite number , we removed bacterial cells by filtering all preparations though a 0 . 2-µm filter . We found that all filtrates from 36-h biofilm preparations ( fresh , supernatant , and dessicated ) possessed strong anti-Plasmodium activity , resulting in inhibition of asexual stage parasites at a level comparable to the chloroquine-treated positive control ( p<0 . 001 , Fig . 5A ) . We also detected moderate anti-asexual stage activity in planktonic Csp_P preparations ( p<0 . 001 ) , while heat-inactivated Csp_P biofilm and biofilm from another bacterial species , Comamonas sp . , had no inhibitory effect . We exposed an in vitro Plasmodium ookinete culture to all five filtered bacterial preparations to assess sexual-stage inhibition and found that the Csp_P 48-h biofilm ( fresh , p<0 . 001; and dessicated , p<0 . 05 ) and biofilm supernatant ( p<0 . 001 ) strongly blocked ookinete development ( Fig . 5B ) . Exposure of the ookinete culture to the filtered planktonic Csp_P liquid culture resulted in a moderate but non-significant inhibition of ookinete development , and exposure to heat-inactivated Csp_P biofilm or Comomonas sp . biofilm filtrate had no effect on ookinete development ( Fig . 5B ) . We also tested the effect of Csp_P bacterial preparations on P . falciparum gametocyte viability . Exposure to 42-h fresh biofilm filtrate resulted in 100% inhibition ( p<0 . 001 , Fig 5C ) and exposure to 42-h dessicated biofilm resulted in approximately 60% inhibition ( p<0 . 05 , Fig . 5C ) of P . falciparum gametocyte development . Gametocytemia could not be estimated for 42-h biofilm supernatant because this preparation caused hemolysis of RBCs ( Fig . 5C ) . However , 36-h biofilm supernatant ( which is not hemolytic ) caused approximately 60% gametocyte inhibition when compared to the LB+PBS control ( p = 0 . 06 , Fig . S2 ) . To test the inhibitory effect of Csp_P preparations on dengue virus infectivity in vitro , we mixed dengue virus ( 106 PFU/ml ) in MEM 1∶1 with each of the five bacterial preparations of Csp_P for 45 min . Samples remained unfiltered during intial exposure to dengue and were filtered through a 0 . 2-µm filter before proceeding with standard plaque assays to avoid contamination of host cells . We found that exposure of dengue virus to a planktonic Csp_P culture did not affect its infectivity in BHK21-15 cells , whereas exposure to Csp_P biofilm , dessicated biofilm , or biofilm supernatant did abolish dengue virus infectivity ( p<0 . 001 , Fig . 5D ) . To better replicate the effect that Csp_P biofilm might have on dengue virus in human blood , we exposed dengue virus in human blood to Csp_P fresh biofilm for 45 min . We then filtered the blood+virus/biofilm mixture and assessed dengue virus infectivity by standard plaque assay . We found that fresh Csp_P biofilm displayed strong anti-dengue activity when the virus was suspended in human blood ( Fig . 5E ) . Csp_P fresh biofilm was unique in its anti-dengue activity , since the biofilms of several other bacterial isolates from the guts of field-caught mosquitoes [5] did not exert any antiviral activity against dengue virus in human blood ( Fig . 5E ) . The anti-dengue activity of Csp_P was apparently dependent on biofilm maturation , since biofilm grown for 24 h showed only weak inhibition when compared to 48-h biofilm ( Fig . S3A ) . The Csp_P biofilm-associated anti-Plasmodium and antiviral activity was also heat-sensitive , since it could be inactivated through a 24-h incubation at 90°C ( Fig . 5A–D ) . Bacterial biofilms are composed of a matrix of extracellular polymeric substances containing polysaccharides , proteins , DNA , and secondary metabolites [12] . To investigate whether the anti-viral activity could simply be a result of virus particle sequestration by the biofilm , we mixed a dengue virus suspension with biofilm and incubated the mixture for a period of 45 min . Samples were then centrifuged , and viral RNA in the supernatants was quantified by RT-qPCR and compared between experimental ( biofilm+DENV ) and control ( LB+DENV ) treatments . Our results indicated that the dengue virus was not sequestered by Csp_P biofilm , since similar viral RNA copies were detected in the biofilm-exposed sample and the LB control sample ( Fig . S3B ) . To investigate whether the loss of dengue virus infectivity was due to a biofilm-mediated change in the pH of the medium , we measured the pH of a dengue virus-Csp_P biofilm mixture at the end of a 45-min incubation period . The pH measurements showed an increase in the pH of the medium from 7 . 6 to 8 . 3 ( Fig . S4A ) . A similar change in the pH was observed when we used the biofilms of other bacteria ( Pantoea sp . Pasp_P and Proteus sp . Prsp_P ) that do not affect dengue virus infectivity ( Fig . S4A ) . To further investigate the effect of pH on dengue virus infectivity , we adjusted the pH of the MEM medium with NaOH and HCl to pH values of 5 . 0 , 7 . 7 , 8 . 5 , and 10 . 0 prior to a 45-min incubation with the dengue virus . A decrease in virus infectivity was only observed after exposure to a pH of 5 . 0 , suggesting that the moderate increase in pH did not mediate the Csp_P biofilm's inhibition of virus infectivity ( Fig . S4B ) . We also showed that Csp_P biofilm does not exert a cytotoxic effect on insect or mammalian cells , as assessed by standard trypan blue cell staining ( Invitrogen ) ( Fig . S5 ) . We finally tested whether the Csp_P biofilm could influence the host cells' susceptibility to dengue virus infection by exposing C6/36 cells to Csp_P biofilm , then removing the biofilm through washes with PBS prior to dengue virus infection assays . This treatement did not influence the virus's ability to infect the host cells ( Fig . S6 ) , suggesting that the anti-DENV activity of Csp_P biofilm is not due to a reduced host cell susceptibility to the virus but is likely a direct anti-viral effect . To provide baseline information on the potential production of antibacterial factors by Csp_P , we performed a basic growth inhibition assay by investigating the ability of a number of other mosquito midgut-derived bacterial isolates ( Ae . aegypti-derived microbiota: Ps . sp = Pseudomonas sp . , Pr . sp = Proteus sp . , C . sp_P = Chromobacterium sp_P , C . viol = C . violaceum , Pn . sp = Paenobacillus sp . ; An . gambiae-derived microbiota: Co . sp = Comamonas sp . , Ac . sp = Acinetobacter sp . , P . pu = Pseudomonas putida , E . sp = Enterobacter sp . , Pa . sp = Pantoea sp . , Ps . sp = Pseudomonas sp . , S . sp = Serratia sp . , Ch . sp = Chryseobacterium sp . ) to grow in proximity to Csp_P on LB agar plates ( Fig . 6 ) . Csp_P was streaked on LB agar and allowed to grow for 48 hours . Midgut-derived bacterial isolates were then vertically streaked up to the Csp_P streak , and allowed to grow in the presence of Csp_P . This assay showed a prominent growth inhibition zone around the Csp_P streak , with inhibition of the growth of all the bacterial isolates that were derived from field-collected Ae . aegypti [5] and An . gambiae [3] , including a close relative known for its production of a variety of bioactive factors , C . violaceum ( Fig . 6A ) . Insect-bacteria associations can influence vector competence in multiple ways; these include shortening the insect's life span , blocking infection with human pathogens by the production of bioactive anti-pathogen factors , and eliciting the insect immune system . We have identified a Chromobacterium sp . ( Csp_P ) bacterium from the midgut of field-derived Aedes aegypti that exerts broad-spectrum anti-pathogen activity against Plasmodium and dengue virus . Specifically , Csp_P renders An . gambiae and Ae . aegypti more resistant to infection by the human malaria parasite Plasmodium falciparum and dengue virus , respectively . Csp_P inhibits the growth of a variety of other bacterial species found in the mosquito midgut and is capable of rapidly colonizing the mosquito midgut . Csp_P appears to exert entomopathogenic activity , since exposure of larvae to Csp_P in the breeding water and ingestion of Csp_P by adult mosquitoes result in high mosquito mortality . It is possible that Csp_P could be effectively used as a transmission blocking agent if it was delivered to mosquitoes through baited sugar traps [28] . Csp_P's ability to colonize the mosquito gut could be further enhanced through established selection procedures based on consecutive passages of the bacterium through the mosquito intestine [29] . Csp_P could be used alone in baited sugar traps or in combination with other microbes that have also been shown to either kill the mosquito or reduce pathogen infection , or both , when present in the mosquito gut [3] , [24] . The larvicidal activity of Csp_P also renders it interesting for potential use in mosquito population suppression . The anti-pathogen activities of Csp_P appear to be mediated by bacteria-produced metabolites that also inhibit parasite and virus infection in vitro , making them interesting as possible lead compounds for transmission blocking and therapeutic drug development . The entomopathogenic , anti-bacterial , anti-viral , and anti-Plasmodium properties of Csp_P make this bacterium a particularly interesting candidate for the development of novel control strategies for the two most important vector-borne diseases , and they therefore warrant further in-depth study .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . Mice were only used for mosquito rearing as a blood source according to approved protocol . The protocol was approved by the Animal Care and Use Committee of the Johns Hopkins University ( Permit Number: M006H300 ) . Commercial anonymous human blood , supplied from Interstate Blood Bank Inc . , was used for Plasmodium and dengue virus infection assays in mosquitoes , and informed consent was therefore not applicable . The Johns Hopkins School of Public Health Ethics Committee has approved this protocol . Mosquito collections were performed in residences after owners/residents permission . Aedes aegypti mosquitoes were from the Rockefeller strain , and Anopheles gambiae mosquitoes were from the Keele strain . Both were maintained on a 10% sugar solution at 27°C and 95% humidity with a 12-h light/dark cycle . Sterile cotton , filter paper , and sterilized nets were used to maintain the cages as sterilely as possible . For experiments utilizing aseptic mosquitoes , females were maintained on a 10% sucrose solution with 20 U penicillin and 20 µg streptomycin from the first day post-eclosion until 1–2 days prior to challenge . The effectiveness of the antimicrobial treatment was confirmed by colony forming unit assays prior to blood-feeding or bacterial challenge . In cases where mosquitoes were antibiotic treated , reintroduction of bacteria through a sugar meal was done by first treating mosquitoes with antibiotics for 2–3 days after emergence , then providing them with 10% sucrose ( for An . gambiae ) or sterile water ( for Ae . aegypti ) for 24 h post-antibiotic treatment . When mosquitoes were not antibiotic treated , they were maintained on 10% sucrose for 2–5 days post emergence . Ae . aegypti were given sterile water during the final 24 hours of this period . In all cases , mosquitoes were then starved overnight and fed for 24 h on cotton strips moistened with a 1 . 5% sucrose solution containing Csp_P at a final concentration of approximately 108 CFU/ml for An . gambiae and 106 CFU/ml for Ae . aegypti . In some experiments ( Figures 1 and 2 ) , Ae . aegypti mosquitoes were also fed Csp_P at a final concentration of 1010 CFU/ml . In antibiotic treated mosquitoes , midguts were dissected three days post ingestion of Csp_P , homogenized in 1× PBS and plated on LB agar . Colonies were then counted to estimate colony forming units ( CFUs ) per midgut as well as prevalence of Csp_P . In mosquitoes not treated with antibiotics , prevalence and/or bacterial load was estimated in one of two ways . For An . gambiae , midguts were dissected at one and two days post Csp_P ingestion , homogenized in 1× PBS and serial dilutions of the homogenate were plated on LB agar supplemented with ampicillin ( 10 , 000 ug/ml ) . Csp_P is highly resistant to ampicillin and grows readily even at this high concentration . We verified that Csp_P was the only bacterium growing on antibiotic treated plates by first confirming that all colonies that grew were similar in color , growth rate and colony morphology . 16s rDNA was then sequenced from a subset of colonies and verified to match the sequence of Csp_P from pure freezer stock . It was not possible to use this method for Ae . aegypti because their midguts commonly contained other highly ampicillin-resistant bacteria . These contaminants grew to very high numbers on the ampicillin-treated plates and interfered with the detection of Csp_P . DNA was therefore extracted using the ZR Soil Microbe DNA MicroPrep kit ( Zymo Research ) from samples dissected 1 and 3 days after feeding on a sugar meal containing either PBS or Csp_P ( 1010 CFU/ml ) . The manufacturer's protocol was altered in the following way: instead of using lysis buffer to disrupt cells , each midgut was put in 500 µl 1× PBS , 25 µl lysozyme ( 10 mg/ml ) and 7 . 5 µl mutanolysin ( 10 KU/ml ) were added and the samples were incubated at 37°C for 1 . 5 h . 15 µl proteinase K and 25 µl 10% SDS were then added , samples were incubated at 55°C for 1 h , and the standard protocol was then resumed . A diagnostic PCR was performed to assess the presence of Csp_P in each individual midgut . Primers were designed to amplify a 415 bp fragment of the Csp_P hydrogen cyanide synthase B gene and the primers were verified to be Csp_P-specific using Primer BLAST from NCBI ( Forward primer: 5′AGGGCGTAACCCTGGACTAT 3′ , Reverse primer: 5′ CCGAAGGAACTGGCTTCGTA 3′ ) . PCR was performed with the above primers using 10 ng DNA as template and Phusion High-Fidelity DNA Polymerase according to the manufacturer's instructions , with the following exceptions: 0 . 5 µl of each primer ( 10 µm ) was used , and 0 . 25 µl BSA was added to each reaction . Cycling conditions were as follows: 95°C for 30 seconds , [95°C for 30 s , 65°C for 30 s , 72°C for 45 s]×27 cycles , 72°C for 10 minutes . 8 µl of each sample was run on a 1% agarose gel and visualized at 400 ms exposure . A visible 415 bp band was considered positive evidence of Csp_P bacteria ( see Fig . S7 for a representative example ) . A very faint band was detected in one of 40 PBS samples , suggesting a minor contamination event or the presence of another bacterium with high sequence identity to Csp_P . This was an isolated incident and was not seen in any other PBS samples . Two independent PCR products were sequenced from Csp_P fed samples and verified to be a perfect match to the sequence obtained from Csp_P sequenced directly from freezer stock . To serve as a positive control and to allow estimation of the sensitivity of the diagnostic PCR , a standard curve was run in which a range of 107–101 copies of the Csp_P hcn B PCR product was used as template . In this way , it was possible to estimate the minimum detection threshold of this assay . Using the above mentioned PCR conditions , a band was detectable in wells containing 103 initial copies of the hcn B product but not in wells containing 102 initial copies , suggesting that this assay is capable of detecting a minimum of 103 copies of Csp_P/midgut . At 2 days prior to blood feeding , sucrose was removed , and the mosquitoes were given sterile water . They were then starved for 12 h prior to blood feeding . Csp_P was grown overnight in liquid LB at 30°C . The overnight culture ( 1 ml ) was then pelleted , washed with 1× PBS , and resuspended in 1× PBS to OD600 = 1 . 0 , which equals a concentration of approximately 108 CFU/ml . Mosquitoes were then allowed to membrane-feed on blood containing bacteria or 1× PBS as a control ( blood mixture: 50% 1 . 0 OD600 bacterial culture or 1× PBS , 40% blood , 10% human serum ) . Bacteria-fed adult females ingested approximately 105 CFU per mosquito . At 2–4 days post-hatching , larvae were placed in cell culture plates in groups of 10 per well . Each well contained 5 ml sterile water plus a small amount of larval food ( liver powder , tropical fish flake food , and rabbit food pellets mixed in a 2∶1∶1 ratio ) . We then added 50 µl of an overnight culture of Csp_P diluted to OD600 = 1 . 0 ( 108 CFU/ml ) to each well; 1× PBS was added to control wells , and mortality was monitored in all wells for a 5-day period . Dengue virus serotype 2 ( New Guinea C strain , DENV-2 ) was propagated in the C6/36 mosquito cell line according to previously published methods [8] . In brief , cell line infection was allowed to proceed for 5–7 days , at which time the cells were harvested with a cell scraper and lysed by freezing and thawing in dry CO2 and a 37°C water bath , then centrifuged at 800 g for 10 min . Dengue virus serotype 2 was isolated and mixed 1∶1 with commercial human blood and used for infections as described in [8] . Mosquitoes that had previously fed on Csp_P bacteria-sucrose solution were starved overnight prior to dengue virus infection . Infected mosquitoes were collected at 7 days post-infection and surface-sterilized by dipping them in 70% ethanol for 1 min and then rinsing them twice in 1× PBS for 2 min each . Midgut dissection was done in one drop of 1× PBS under sterile conditions , and the midgut was transferred to a microcentrifuge tube containing 150 µl of MEM . Midguts were homogenized using a Kontes pellet pestle motor , filtered , and stored at −80°C until ready for virus titration . Dengue virus titration of infected midguts was done as previously reported [8] , [30] . In brief , the infected midgut homogenates were serially diluted and inoculated into C6/36 cells in 24-well plates . After an incubation of 5 days at 32°C and 5% CO2 , the plates were fixed with 50%/50% methanol/acetone , and plaques were assayed by peroxidase immunostaining using mouse hyperimmune ascitic fluid specific for DENV-2 as the primary antibody and a goat anti-mouse HRP conjugate as the secondary antibody . In addition , where indicated , dengue virus plaque assays were conducted in BHK-21 cells . At 5 days post-infection , the 24-well plates were fixed and stained with crystal violet . Plaques ( formed by cells with cytopathic effect ) were counted and analyzed . P . falciparum strain NF54 was maintained in continuous culture according to the method described by Tragger and Jensen [31] . In brief , P . falciparum was grown in O+ red blood cells ( RBCs ) at 2% hematocrit and RPMI 1640 medium supplemented with glutamine , HEPES , hypoxanthine , and 10% O+ human serum . To maintain a microaerophilic environment , parasites were maintained in a candle jar at 37°C . Use of human erythrocytes to support the growth of P . falciparum was approved by the internal review board of the Bloomberg School of Public Health . Gametocytemia and exflagellation events were assessed after 18 days of P . falciparum culture . The gametocyte culture was centrifuged and diluted in a mixture of RBCs supplemented with serum . Mosquitoes were rendered aseptic via antibiotic treatment and then fed on membrane feeders for 30 min with blood containing P . falciparum gametocytes . Csp_P was either added directly to the infectious blood meal ( bacterial concentration = 106 CFU/mL ) or introduced via sugar meal as described above 3 to 4 days prior to the infectious blood meal . On the same day as the blood meal , mosquitoes were sorted , and the unfed mosquitoes were removed . At 7 to 8 days after blood feeding , the fed mosquitoes were dissected , and their midguts were stained with 0 . 1% mercurochrome . The number of oocysts per midgut was determined with a light-contrast microscope , and the median was calculated for the control and each experimental condition . More than three independent replicates were used per group . To grow bacteria in planktonic conditions , we spiked 5 ml sterile LB with 5 µl of bacterial freezer stock and allowed the culture to grow overnight at 30°C with shaking . We then diluted planktonic cultures to OD600 = 1 . 0 ( ±0 . 1 ) with additional sterile LB broth which , for Csp_P , results in a concentration of approximately 108 CFU/ml . To grow bacteria under biofilm conditions , we dispensed 1 ml of sterile LB into each well of a 24-well cell culture plate and spiked each well with 1 µl of bacterial freezer stock . We then allowed the culture to grow at room temperature without shaking for 48 h . Csp_P biofilm supernatant was harvested from single bacterial culture wells containing 48-h biofilm and was found to have an average bacterial concentration of approximately 109 CFU/ml . To harvest fresh biofilm , we removed the supernatant from five wells containing 48-h biofilm , resuspended the biofilm from each well in 100 µl 1× PBS and pooled the five wells . For Csp_P , this pooled biofilm solution contained approximately 109 CFU/ml and an average of 5 mg of biofilm ( dry weight ) . To obtain desiccated biofilm , we collected the fresh biofilm from five wells as indicated , centrifuged the biofilm at 5000 rpm for 2 . 5 min , removed the PBS supernatant , and allowed the biofilm to dry at room temperature . On the day of the experiment , we resuspended the five wells of desiccated biofilm in 500 µl 1× PBS to mimic the fresh biofilm treatment . To heat-inactivate the fresh biofilm , we collected fresh biofilm as indicated and incubated samples at 90°C for 24 h prior to the experiment . We prepared Csp_P bacterial cultures as described above and filtered all samples through a 0 . 2-µm filter . Asexual-stage assay: Inhibition of asexual-stage P . falciparum was assessed using a SYBR green I-based fluorescence assay as described earlier [32] . Csp_P biofilm was grown for 36 h for this experiment because 48-h biofilm causes hemolysis of RBCs ( Fig . S8 ) , which interferes with the assay . Parasites were synchronized using 5% sorbitol [33]; 5 µl of each bacterial preparation was dispensed in triplicate wells of 96-well microplates , followed by addition of 95 µl of synchronous ring-stage P . falciparum cultures at 1% hematocrit and 1% parasitemia . Chloroquine ( 250 nM ) was used as a positive control , and parasite growth medium was used as a negative control . After 72 h of incubation in a candle jar at 37°C , an equal volume of SYBR green-I solution in lysis buffer ( Tris [20 mM; pH 7 . 5] , EDTA [5 mM] , saponin [0 . 008%; w/v] , and Triton X-100 [0 . 08%; v/v] ) was added to each well and mixed gently , then incubated 1–2 h in the dark at room temperature . Plates were read on a fluorescence plate reader ( HTS 7000 , Perkin Elmer ) , with excitation and emission wavelengths of 485 and 535 nm , respectively . Percent inhibition was calculated relative to negative ( 0% inhibition ) and positive controls ( 100% inhibition ) . Three biological replicates were assayed . Ookinete-stage assay: To assess inhibition of ookinete-stage P . berghei parasites , female Swiss Webster mice ( 6–8 weeks old ) were infected with a transgenic strain of P . berghei that expresses Renilla luciferase . Starting at 3 days post-infection , exflagellation assays were performed until at least 20 exflagellation events were recorded in a 20× field . At this time , mice were bled by heart puncture using a heparinized needle , and the blood was diluted in 10 volumes of ookinete medium ( RPMI 1640 , 10% FBS , 50 mg/ml hypoxanthine , and 2 mg/ml NaHCO3 , pH 8 . 3 ) with 4% mouse RBC lysate . Samples ( 50 µl ) of each bacterial preparation were then mixed with the infected blood and incubated for 24 h at 19°C . Ookinete counts were determined using the Renilla luciferase assay system ( Promega , USA ) according to the manufacturer's instructions . The experiment was performed on two independent days , and each sample was assayed in triplicate on each day . Gametocyte-stage assay: Inhibition of gametocyte-stage P . falciparum by Csp_P was assessed as described previously [34] . To prevent hemolysis of RBCs , Csp_P biofilm was grown for 36 and 42 hours for this experiment . In brief , NF54 P . falciparum cultures were started at 0 . 5% asexual parasitemia and 4% hematocrit . Csp_P bacterial preparations were added 15 days after Plasmodium cultures were initiated , and gametocytemia was determined 18 days after culture initiation . At least three biological replicates were tested for each culture preparation . More than 500 erythrocytes were examined for gametocytes across Giemsa-stained blood films from each sample . We prepared Csp_P bacterial cultures as described above ( planktonic state , biofilm , biofilm supernatant , dessicated biofilm , and heat-inactivated biofilm ) , mixed 75 µl of each bacterial culture preparation with 75 µl of MEM containing dengue virus serotype 2 and incubated the mixture at room temperature for 45 min . Samples were then filtered through a 0 . 2-µm filter , serially diluted , and used to infect BHK21-15 cells . Plaque assays were conducted as described above to assess dengue virus infectivity . Percent inhibition was calculated as the percent decrease in PFU/ml relative to the PBS+LB control , which was standardized to 0% inhibition . The experiment was performed on two independent days , and each assay was performed in triplicate on each day . In experiments in which dengue was mixed with human blood before exposure to Csp_P , bacterial biofilms were not removed from the cell culture plate . Rather , dengue virus was mixed 1∶1 with human blood , and 150 µl of this mixture was added directly to each well containing Csp_P biofilm and incubated for 45 min at 30°C . Following this incubation period , the blood-dengue virus solution was mixed with the biofilm , and 50 µl of the mixture was then drawn from the well , diluted in MEM , and filtered through a 0 . 2-µm filter . The resulting filtrate solution was then serially diluted and used to infect C6/36 cells . To assess whether the anti-dengue activity of Csp_P was due to sequestration of DENV by the Csp_P biofilm , we mixed a dengue virus suspension with Csp_P 48 hr biofilm or LB broth and incubated it for a period of 45 min . Samples were then centrifuged at 5 , 000 rpm for 5 min . The supernatants were collected , and RNA was extracted from equal volumes ( 50 µl ) of experimental ( biofilm+DENV ) and control ( LB+DENV ) samples using the RNeasy kit ( Qiagen ) . Comparison of viral RNA loads in the extracted supernatant was done via RT-qPCR relative quantification , using 2 µl of the viral RNA in a 20-µl reaction volume . The pH of bacterial biofilms and supernatants was assessed with a micro-pH electrode ( Lazar Lab ) at room temperature . Effects of pH changes on dengue virus infectivity were assessed by adjusting the pH of the MEM with NaOH and HCl until the desired range of pH values was obtained: 5 . 0 , 7 . 7 , 8 . 5 , and 10 . 0 . The pH-adjusted MEM was then mixed with dengue virus-laden blood and incubated for 45 min prior to serial dilution and infection of C6/36 cells . Cell viability assays on the mosquito cell line C6/36 and the vertebrate cell line BHK-21 were performed via trypan blue staining ( 0 . 4% , Invitrogen ) according to the manufacturer's instructions . In brief , 50 µl of suspended cells were placed in a microcentrifuge tube and mixed with 10 µl of Csp_P filtered fresh biofilm or PBS as a control . C6/36 cells were incubated at 32°C and BHK-21 cells were incubated at 37°C+5% CO2 for 45 min . Cells were then mixed with 12 µl of 0 . 4% trypan blue stain . The mixture was allowed to stand for 5 min at room temperature and then loaded into a hemocytometer for cell viability assessment and counting under a microscope . To assess whether exposure to Csp_P biofilm changes the susceptibility of the host cell to DENV , we conducted assays exposing C6/36 cells to Csp_P-filtered biofilm prior to dengue virus infection . Cells were grown to 80% confluency; the cell medium was then removed , washed once with 1× PBS , and then overlaid with 100 µl of Csp_P biofilm that had been filtered using a 2-µm filter or with 1× PBS ( control ) for about 10 min . Plates were then washed three times with 1× PBS and then infected with 100 µl of dengue virus for about 45 min . Cells were assessed for plaque formation at 6 days post-infection . Human erythrocytes were washed with RPMI 1640 medium until the supernatant was visually free of hemoglobin pigment . The washed erythrocytes were suspended in malaria complete medium to yield a 1% hematocrit . Filtered Csp_P biofilm was mixed with erythrocytes and incubated up to 24 h at 37°C . To separate lysed RBC cytosol from whole RBCs , the suspension was centrifuged at 2000 rpm for 5 min . The resulting supernatant was carefully aspirated and plated in new 96-well microplates . Control erythrocytes without any bacterial material were used as a negative control ( blank ) , and freeze-thawed erythrocyte lysate was used as positive control ( 100% hemolysate ) . To determine the % lysis in test samples , plates were read at 405 nm in an ELISA plate reader ( HTS 7000 Perkin Elmer ) , and the reading was expressed as a fraction of the positive control . To conduct real-time PCR assays , RNA samples were treated with Turbo DNase ( Ambion , Austin , Texas , United States ) and reverse-transcribed using M-MLV Reverse Transcriptase ( Promega , USA ) . The real-time PCR assays were performed using the SYBR Green PCR Master Mix Kit ( Applied Biosystems , Foster City , California , USA ) in a 20-µl reaction volume; all samples were tested in duplicate . The ribosomal protein S7 gene was used for normalization of cDNA templates . Primer sequences used in these assays are given in Table S1 . The Mann-Whitney U test , one-way ANOVA with Dunnett's post-test and pairwise Log-Rank tests for survival analysis were conducted using the GraphPad Prism statistical software package ( Prism 5 . 05; GraphPad Software , Inc . , San Diego , CA ) . Data in Figure 5 were analyzed using an ANOVA , followed by a Tukey's test in R ( R Foundation for Statistical Computing ) . | The infectious agents that cause malaria and dengue are transmitted by Anopheles and Aedes mosquitoes , respectively . Bacteria found in the mosquito midgut have the potential to dramatically affect the susceptibility of the mosquito vector to the malaria parasite and dengue virus . In this work , we investigate one such microbe , Chromobacterium sp . ( Csp_P ) , a bacterium isolated from a field-caught Aedes aegypti mosquito . We show that Csp_P can effectively colonize the midguts of Anopheles gambiae and Aedes aegypti mosquitoes and can , when ingested by the mosquito , significantly reduce the mosquito's susceptibility to infection with the malaria parasite and dengue virus . We also show that exposure to , and ingestion of , Csp_P can reduce the lifespan of larval and adult mosquitoes , respectively . We show that Csp_P has anti-Plasmodium and anti-dengue activity independent of the mosquito , suggesting that the bacterium secretes metabolites that could potentially be exploited to prevent disease transmission or to treat infection . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"bacteriology",
"antimicrobials",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"emerging",
"viral",
"diseases",
"microbiology",
"animals",
"parasitology",
"viral",
"vec... | 2014 | Chromobacterium Csp_P Reduces Malaria and Dengue Infection in Vector Mosquitoes and Has Entomopathogenic and In Vitro Anti-pathogen Activities |
Animals’ exchanges are considered the most effective route of between-farm infectious disease transmission . However , despite being often overlooked , the infection spread due to contaminated equipment , vehicles , or personnel proved to be important for several livestock epidemics . This study investigated the role of indirect contacts in a potential infection spread in the dairy farm network of the Province of Parma ( Northern Italy ) . We built between-farm contact networks using data on cattle exchange ( direct contacts ) , and on-farm visits by veterinarians ( indirect contacts ) . We compared the features of the contact structures by using measures on static and temporal networks . We assessed the disease spreading potential of the direct and indirect network structures in the farm system by using data on the infection state of farms by paratuberculosis . Direct and indirect networks showed non-trivial differences with respect to connectivity , contact distribution , and super-spreaders identification . Furthermore , our analyses on paratuberculosis data suggested that the contributions of direct and indirect contacts on diseases spread are apparent at different spatial scales . Our results highlighted the potential role of indirect contacts in between-farm disease spread and underlined the need for a deeper understanding of these contacts to develop better strategies for prevention of livestock epidemics .
The structure of contact networks between individuals from human and animal populations is a key determinant of the dynamics of communicable diseases . In response to the Bovine Spongiform Encephalopathy crisis , the Council of the European Union implemented in 1997 a system of permanent identification of individual bovine animals enabling reliable traceability from birth to death . A fundamental part of this system consists in an extensive database to track movements of farmed animals . Information on between-farm animal movements have been used to reveal the existing contact network structure within livestock systems [1–5] , as it is considered the most effective route of disease transmission [6] . The study of the disease spread in livestock systems made it possible to fine tune surveillance systems , to address biosafety guidelines and control strategies aimed to reduce the risk of disease outbreaks , and to limit the impact on animal health and economic sustainability of farming systems . The vast majority of these studies , models and analyses mainly ( if not exclusively ) focuses on direct contacts , i . e . infections following the movement of diseased animals between farms . Alternative and often more cryptic transmission pathways have received much less attention , even though they can crucially affect the efficacy of disease surveillance and control systems . For instance , in 2001 foot-and-mouth Disease ( FMD ) epidemic in the UK , although animal movements were banned since late February , the on-set of newly infected farms was reported until the end of the summer [7 , 8] . Since further studies excluded a strong effect of aerial spread for the FMD virus strain in question , the only alternative route of transmission for the infection was represented by the spread through fomites [9]–such as contaminated operators , vehicles , and equipment–potentially capable of carrying pathogens from infected farms to susceptible ones [10] . In fact , the FMD epidemic stopped only after the implementation of stronger biosecurity measures in UK farms , which mostly targeted the movement of contaminated equipment and personnel [8] . Between-farm disease transmission through fomites , mediated by operators and personnel external to the farm , is usually defined as indirect transmission [11] . In recognition of its importance , epidemiological models of disease dynamics in livestock have started to include different pathways of between-farm transmission in addition to cattle movements ( see [12] , and references therein ) . Unlike animal movements , the role of indirect transmission of livestock diseases is still largely unknown . The reason for this knowledge gap is twofold: on the one hand , because of the subtle , highly diverse and complex nature of indirect contacts , it is intrinsically difficult to assess the relative importance of alternative transmission pathways for disease risk . On the other hand , because of privacy reasons , it is much easier to track livestock movements than that of farm operators and personnel . Indeed , collecting reliable and extensive quantitative data on indirect contacts on a temporal and spatial scale relevant for epidemiological modelling has proved to be very challenging . Information on farm operator movements has been generally gathered by using voluntary questionnaires on the number and the frequency of farm visits over a given time span [10 , 11 , 13–15] . Despite the usually low questionnaire response rate and the low number of farms involved in these studies , this approach has been crucial to provide a preliminary rank of categories of indirect contacts by potential disease risk . However , the information gathered has been often insufficient to fully investigate the network structure of indirect contacts in a given area , and to characterize the contact frequency between farms . This is why only few studies have applied network analysis techniques on questionnaire-based data ( see e . g . [10 , 16 , 17] ) . Alternatively , indirect routes of transmission in epidemic models have been represented using risk kernel functions [18 , 19] . These are functions assigning a probability of between-farm disease transmission on the basis of the inter-farm distance [19 , 20] . However , the approach based on kernel functions is unable to tease apart the relative importance of different networks of indirect contacts , such as those associated to the movement of different farm operators . The aim of this work is to present a novel quantitative analysis of the relative importance of indirect and direct contacts in a network of 1 , 349 dairy farms in the Province of Parma in the Emilia Romagna Region ( Italy ) . The analysis is based on a unique , high-resolution temporal and spatial database of between-farm movements of 50 public officers of the regional veterinary service and 203 private veterinary practitioners ( which represent the potentially infectious indirect contacts ) . The former visit a large number of farms usually only few times a year; the latter serve a small subset of farms each , and they visit them several times per year . We thus expected that the structures of their contact networks are substantially different and might result in different risks of disease transmission through fomites . We used network analysis to characterize the structure of the networks of indirect contacts and contrast them with the structure of the network of direct contacts ( i . e . the one associated to animal movement/trade ) . Then , we assessed the contribution of both direct and indirect contacts networks in explaining the observed spatial distribution of dairy farms infected by Mycobacterium avium subsp . paratuberculosis ( MAP ) . MAP is responsible for Johne’s disease , a chronic gastrointestinal inflammation affecting ruminants and it is endemic in the study area [21] . It is well documented that animal movements represent the primary route of MAP transmission between farms [22 , 23] . However , the role of fomites such as footwear [24] and shared farm and veterinary equipment [25] as secondary transmission routes has been highlighted . Finally , we used advanced techniques in network analysis to characterize the temporal network defined by direct and indirect contacts in order to understand the between-farm transmission for fast spreading diseases where the time scale of epidemics is similar to those of the evolution of the network , such as FMD [26] .
Our study system is represented by a network of 1 , 349 dairy farms operating in the Province of Parma ( Emilia-Romagna region , Italy ) in 2013 ( Fig 1 ) . For each farm , we extracted from the Italian National Bovine Database ( BDN ) a unique identification code and the related spatial coordinates . As we were interested in analysing the structure of the cattle movement network on a wider geographical scale and time window as well , we extracted from BDN also information on cattle movement from the 4564 dairy farms operating in the whole Emilia-Romagna region ( which includes also the province of Parma ) between 2010–2013 . Each individual cattle movement record contained: a unique identification code for the animal , identifier codes of the farms of origin and destination , codes for farm production sector ( dairy or mixed ) , and the movement date . As the Province of Parma is a strongly oriented dairy area , beef farms were not considered in the present study . In fact , they represented less than 25% of the total cattle farms area ( 473 over 1 , 822 ) , and the two systems are almost completely separated . The only unidirectional contact points consist in the shipment of surplus individuals from dairy , mostly male calves , to beef farms . Beef farms are less involved in veterinarians networks too . First , they do not receive frequent inspections because they are not included in surveillance plans for most diseases ( i . e . bovine tuberculosis [27] ) . Second , beef animals receive less care by practitioners , in part because individuals’ life span is shorter ( 2 vs . 5 years for dairies [27] ) , but also because the lower economic value of individuals does not justify intensive health assistance as for dairy cattle . The network of cattle movements was assembled by creating a directed edge between any two farms ( representing network nodes ) that exchanged animals during the observed period and setting a non-zero value in the corresponding adjacency matrix [28] . Among the many ways in which edges could be weighted in a time-aggregated cattle movement network [29] , we considered ( i ) the unweighted case , in which a link value is equal to 1 if at least one contact exists in a given period , and 0 otherwise; and ( ii ) the case in which links are weighted proportionally to the number of animals exchanged through each contact within a given period . Data on visits of veterinary officers ( VO ) were provided by the Local Health Unit of Parma Province ( LHU ) . These visits were scheduled by the LHU for various purposes , including animal health inspection and disease surveillance . We created a database including all visits on dairy farms during the year 2013 . For each visit we recorded the farm unique identifier , the identifier of the VO visiting the farm ( in an anonymous form ) , and the visit date . Data on veterinary practitioners ( VP ) were obtained from three datasets . The first was the list of drug prescriptions in 2013 , that is , documents that compulsorily need to be ( i ) signed by a registered practitioner , and ( ii ) delivered and kept by both the LHU and the farm . For this reason , each drug prescription corresponds to at least one on-farm visit by a veterinarian . This dataset contained the unique identifier of the farm where the drug was prescribed , the identifier code of the VP prescribing the drug ( in an anonymous form ) , and the prescription date . The second dataset was constituted by the records of diagnostic samples submitted to the Istituto Zooprofilattico Sperimentale della Lombardia e dell’Emilia-Romagna ( IZSLER ) , the local veterinary diagnostic laboratory . These samples are collected on farm by practitioners and delivered to the IZSLER for biological testing . This dataset contained the unique identifier of the farm where the sample was collected , the identifier of the VP collecting the sample ( in an anonymous form ) , and the date of collection . The third dataset included the list of on-farm inspections on order of the LHU , but subcontracted to VPs . For each dataset , the record contained the VP identifier , the farm unique identifier , and the visit date . To build the VO and the VP contact networks , we assigned a directed edge connecting a given farm to those later visited by the same veterinarian within a given time interval . This time interval represents the time span in which the veterinarian or her/his equipment can remain contaminated by the pathogenic agent . We defined it as contamination period , h , and it depends on the pathogen ability to survive in the environment and on the type of contaminated material . The order by which VO and VP operators visited the farms within a given day was not reported in the dataset . Thus , to define the contact chain generated by multiple farm visits occurring in any given day , we generated 50 networks with potential itineraries by randomly selecting , for each veterinarian , the first farm visited in that day , so to derive the itinerary that minimized travel distance ( see S1 Text for details on method and software ) . The analyses of network structure and disease risk were conducted by setting h = 0 days under the conservative assumption that only within day visits can result in infection transmission . We assessed the effect on network properties of higher values for h in a specific analysis reported in S1 Text . As for the cattle movement network , we also developed a weighted version of the veterinarian networks , where the weighting coefficients represent the number of contacts within the year 2013 . To evaluate the potential risk of disease spread in the three different networks ( CM , VO , and VP ) , as per other studies on farm networks [2 , 3 , 30] , we first derived three important metrics for each network , namely: ( i ) the link density [31] , which is defined as the fraction of observed links over the possible number of links; ( ii ) the giant strongly connected component ( GSCC ) [2] , which is defined as the biggest portion of the network in which each node is reachable from any other node; and , in the weighted version of the networks , ( iii ) the contact frequency , which is defined as the mean frequency of the observed links within the observation period . The node degree k , a central measure in network theory , is defined as the number of links for each node [28] . As our networks were characterised by directed links , we derived two node degrees for each farm in each contact network: the i-th farm in-degree , defined as the total number of farms from which farm i receives cattle or farms visited by VO and VP before visiting farms i ( kI ) ; and i-th farm out-degree , defined as the number of farms to which farm i sends cattle or farms visited by VO and VP after visiting farm i ( kO ) . Consequently , for each contact network , we computed the degree distributions P ( kI ) and P ( kO ) , respectively . In the case of VO and VP networks , we computed degrees and degree distributions assuming h = 0 days ( analyses based on different assumptions on h are shown in S1 Text ) . In addition , in the case of weighted networks , we computed the node strengths , which are defined as the sum of all incoming ( in-strength , SI ) and outgoing ( out-strength , SO ) links' weights [28] . As for the degrees , for each weighted contact network , we computed the strength distributions P ( SI ) and P ( SO ) ( see S1 Text for details ) . A fundamental assumption in the spatial analysis of epidemiological data for communicable diseases is that the contact network is an important driver of the observed disease dynamics and , accordingly , may be able to partially explain the spatial distribution of reported cases of the infectious disease under study . Here , we wanted to assess the relationship between direct and indirect contact networks and Mycobacterium avium subsp . paratuberculosis ( MAP ) positive farms at regional scale ( i . e . Emilia-Romagna , for cattle movements only ) and local scale ( i . e . Parma province , for veterinarians and cattle movements ) by using data on the infection status of 2 , 648 dairy farms in Emilia-Romagna region ( whereof 966 in Parma province ) as identified in Ricchi et al . [21] . The infection state of farms , positive or negative , was evaluated by ascertaining the presence of MAP in bulk tank milk by real-time PCR targeting insertion sequence IS900 , twice per farm [21] . To avoid detection bias , farms for which bulk tank milk had not been tested for MAP were excluded from the analysis . Since MAP has a long incubation period [32] and , consequently , a slow infection dynamics , static network measures as those underlying our analysis can be appropriately used to represent the between-farm disease transmission , as shown by [26] . To assess the relationship between MAP presence and direct/indirect contact structures , we used a network-based model approach similar to that developed in [16] . Specifically , we defined the mean exposure to infection of MAP positive farms ( EI ) as: EI=∑j=1n∑i≠jEijδj∑j=1nδj , ( 1 ) where δj = 1 [δj = 0] if farm j is MAP positive [negative]; and Eij represents exposure of farm j to farm i , defined as Eij = Aij ( where A represents the adjacency matrix of the weighted network ) if farm i is MAP positive; Eij = 0 otherwise . Consequently , ∑i ≠ jEij represents the total exposure for farm j . Analogously , we defined the mean exposure to infection of MAP negative farms ( ES ) as: ES=∑j=1n∑i≠jEijθj∑j=1nθj , ( 2 ) where θj = 1 [θj = 0] if farm j is MAP negative [positive] . Garcia Alvarez et al . [16] proposed that , if the infection was transmitted from the observed contact networks , we would expect that MAP positive farms were more exposed than MAP negative ones ( i . e . , EI > ES ) . In addition , we would expect that MAP positive farms were more exposed in the observed contact network than in a random one , while MAP negative farms were expected to be less or similarly exposed compared to a random network . To test these hypotheses , we generated random networks with the same number of nodes as in the observed contact networks , but randomly allocating the edges between the nodes . In order to investigate the impact of the network linking pattern ( instead of the in-degree distribution ) on the spread of the disease , the distribution of farm contacts was maintained through a rewiring process in the random networks , as suggested by Kiss et al . [33] . The node state with respect to MAP infection was maintained fixed in all random networks using the observed bulk tank milk data . Since MAP can persist in farms for years , we also tested whether our results were robust with respect to the specific year used to derive the contact networks . In particular , for the cattle movement network , for which data were available for multiple years , we tested whether contact networks built by using data from years 2010–2012 could explain the infectious state by MAP detected in 2013 . In addition , since in the absence of biosecurity measures MAP can persist in the environment , we tested whether our results at local and regional scale were robust with respect to different assumptions on the contamination period ( specifically , h = 0 and h = 7 days ) . Moreover , to determine the possible effect of spatial clustering as a driver of the differences in MAP positivity among farms , we used the q-nearest neighbours test . The q-nearest neighbours is a non-parametric test able to identify a potential spatial clustering in the distribution of cases , by computing the number of cases observed within the q neighbour farms of each positive case [39] . We set the number of neighbours q from 1 to 10 [16 , 34] . A major focus of our work was to identify which farms could act as super-spreaders in the studied networks . A super-spreader is defined as a highly connected individual farm able to potentially spread the infection to a very large number of farms in the system [35] . In the context of diseases spreading through a contact network , centrality measures are often used to identify the super-spreaders [36] . The most simple but still effective centrality measure is the degree centrality [28] . However , despite the simplicity and the usefulness of the degree centrality , it is well known that assuming a static network derived by all the contacts that occurred in a given period , 2013 in our case , and ignoring the temporal sequence of connections , can lead to largely overestimate the transmission risk for fast spreading diseases , such as FMD and influenza , where the time-scale of the epidemics is similar to that of the evolution of the network [4 , 26 , 37 , 38] . To overcome this limit , Dubé and colleagues [3] introduced a risk-based measure that accounts for the temporal sequence of contacts in animal trade networks , called the infection chains ( IC ) . In particular , for a given farm , the ingoing IC ( IIC ) and the outgoing IC ( OIC ) measure the maximum number of farms connected with it through a sequence of animal movements [3 , 30] . Konschake et al . [39] extended the infection chain concept by assuming that only contacts occurring within a finite infectious period ( γ ) may act as potential transmission events . Accordingly , we computed time-dependent ICs referred to a specified date of emergence of the infection in the farm system ( d ) , specifically IIC ( d ) and OIC ( d ) . From an epidemiological view point , the OIC ( d ) in the i-th farm , OICi ( d ) , provides an upper bound to the size of an epidemic emerging from the i-th farm in day d . Analogously , the IIC ( d ) in the i-th farm , IICi ( d ) , provides an upper bound for the probability for the i-th farm of getting infected on day d following an epidemic event occurred in any other farm in the network . This was computed as IICi ( d ) / ( N−1 ) , with N corresponding to the total number of farms in the system . Following these considerations , we defined the infection potential ρi ( d ) of the i-th farm in a given day d as: ρi ( d ) =IICi ( d ) N−1OICi ( d ) . ( 3 ) From expression ( 3 ) , we built two more general epidemiological indicators: the mean infection potential of i-th farm on a given time-period of m days as: ρi=∑d=1mρi ( d ) m ( 4 ) and the average infection potential of the system in day d as: ρ ( d ) =∑i=1Nρi ( d ) N . ( 5 ) In order to avoid boundary effects due to the limited period of data availability ( 1 year ) , we computed ρi ( d ) for a period of 245 days starting from the beginning of March to the end of October 2013 . We computed the infection potential for individual CM , VO and VP networks , for the veterinarians total network ( VT = VO + VP networks ) , and for the network of all transmission routes combined ( TN = CM + VT ) . For VO , VP , VT , and TN networks , the calculation was repeated for each of the 50 simulations . We defined as super-spreaders the farms in the highest 5th percentile of ρi value distribution . We assumed a farm infectious period length γ of 14 days . According to Konschake et al . [39] , this is the threshold value above which the IC measure is stable to variations in γ and , additionally , this infectious period is compatible with rapidly spreading diseases , such as FMD [40] . However , we performed a sensitivity analysis on the ICs for different values of γ ( from 3 to 28 days ) to assess whether a variation of γ has strong consequences on farm ranking in terms of γ and , thus , on the identification of super-spreaders . In order to assess the correlation between ρi calculated for CM and VT networks , we used the Kendall's τ , a non-parametric test .
There were 16 , 647 cattle moved in the province of Parma in 2013 , for a total of 1 , 433 between-farm directed links in the yearly aggregated CM network . Link-density was 0 . 0008 ( see Table 1 ) . The giant strongly connected component ( GSCC ) included 18 farms , corresponding to the 1 . 14% of the network , while the distribution of the yearly exchanged animal-per-contact was very heterogeneous , with an average of 11 . 62 animals . The mean farm degree was 1 . 06 , and both kI and kO ranged from 0 to 15 . The degree distributions P ( kI ) and P ( kO ) are showed in Fig 2 ( red line , a and b panels , respectively ) . The median of both kI and kO was equal to zero , and this was a consequence of the large number of farms with no incoming ( 688 ) or outgoing ( 719 ) cattle movements within the system during 2013 . The total number of moved cattle between dairy farms within the Emilia-Romagna region ranged from 50 , 186 to 57 , 276 over the period 2010–2013 . These movements originated from 4 , 624 to 5 , 094 between-farm contacts . The veterinary officer ( VO ) dataset included data on 6 , 524 on-farm visits performed by 50 officers . By setting the contamination period h equal to 0 ( corresponding to assuming possible indirect transmission only for consecutive visits on the same day ) , the median [5th-95th percentile among the 50 simulations] link density was 0 . 0049 [0 . 0047–0 . 0050]; the median [5th-95th percentile] giant strongly connected component ( GSCC ) included about 67% [35%-70%] of the network , corresponding to 918 [480–948] farms . The median [5th-95th percentile] yearly contact frequency was 1 . 19 [1 . 15–1 . 22] . The overall veterinary practitioner ( VP ) dataset included a total of 14 , 053 visits performed by 203 practitioners . This was the result of joining three data sources: the drug prescription list ( 11 , 611 prescriptions by 181 VP ) , the animal tissue-drop records ( 1 , 085 records by 108 VP ) , and the government subcontracted visits ( 1 , 426 visits performed by 12 VP ) . As for the VO network , we simulated the same day visit order 50 times . The median [5th-95th percentile] link density was 0 . 0029 [0 . 0029–0 . 0029]; the median [5th-95th percentile] GSCC included about 54% [53%-55%] of the networks , corresponding to 732 [721–740] farms; and the median [5th-95th percentile] yearly contact frequency was 1 . 19 [1 . 18–1 . 20] . VO and VP networks showed more widespread degree distributions with respect to the CM network ( Fig 2 ) . At h = 0 , the mean degrees were 6 . 55 and 3 . 90 for VO and VP , respectively , and 1 . 06 for CM . On the other hand , the strength distributions of all three networks showed to be more similarly distributed ( see Figure S1 . 3 . 1 and Table S1 . 2 . 2 in S1 Text ) . As expected from previous literature on between-farm transmission of Mycobacterium avium subsp . paratuberculosis ( MAP ) , we found evidence of association between cattle movements and the distribution of MAP positive farms within the Emilia-Romagna region ( see Fig 3 ) . Specifically , we found that the mean exposure of MAP positive farms ( Fig 3 upper segment: blue dot , EI ) derived by using the CM network at regional scale is: a ) higher than that of MAP negative farms ( Fig 3 upper segment: red dot , ES ) ; b ) higher than that in a randomly generated network ( Fig 3 upper segment: blue vertical bars , p = 0 . 005 ) . Conversely , within Parma province , we did not find a significantly higher mean exposure of MAP positive farms derived by using the CM network ( Fig 3 middle segment: blue dot ) compared to random networks ( Fig 3 middle segment: blue vertical bars , p = 0 . 456 ) . On the other hand , within Parma province , we found that the mean exposure of MAP positive farms in the veterinary network ( Fig 3 bottom segment: blue dot ) was significantly higher than in random networks ( Fig 3 bottom segment: blue vertical bars , p < 0 . 001 ) . In addition , we found that our results were robust with respect to the year considered for building the contact networks and the assumptions on the length of the contamination period ( see S1 Text ) . The spatial analysis showed that no clear spatial clustering could be detected among MAP-positive farms . Specifically , the q-nearest neighbours test was not statistically significant for neighbour farms from q = 1 to q = 10 ( see Table 2 ) . This result suggests that between-farm transmission of MAP was not associated to spatial proximity ( see Fig 4 ) . Infection potential ρi of VT network was poorly correlated with that of CM network ( Kendall's τ = 0 . 08 , p < 0 . 01 ) . The number of shared super-spreaders for the CM and VT networks ( defined as the 5% of farms with the highest ρi value ) ranged between 4 and 7 over 50 simulations of within day veterinary itineraries . By using the median value of ρi for each farm as a reference measure , the number of shared super-spreaders between direct and indirect contacts was 4 ( see Fig 5 ) . To test whether the number of observed shared super-spreaders was significantly higher than in the random case , we computed a permutation test by assigning the observed ρi values in each network to random farms . Upon 20 , 000 runs ( 1 , 000 for each of the 50 simulations ) , the test turned out to be not statistically significant ( p = 0 . 20 ) . Fig 6 shows the time trend of the average infection potential of the whole farm system , ρ ( d ) , during the 245-day period . The average infection potential ρ ( d ) was: 0 . 04 × 10−3 ( sd = 0 . 44 × 10−3 ) for the CM network , 0 . 14 × 10−3 ( sd = 0 . 37 × 10−3 ) for the VO network , and 0 . 16 × 10−3 ( sd = 0 . 69 × 10−3 ) for the VP network . Combining networks together , the average ρ ( d ) was 0 . 84 × 10−3 ( sd = 2 . 17 × 10−3 ) for the veterinarians total network ( VT ) , while ρ ( d ) was 3 . 13 × 10−3 ( sd = 39 . 23 × 10−3 ) for all the direct and indirect contact networks combined . Sensitivity analyses showed stable rankings for IIC , OIC and ρ with varying infectious period γ ( see S1 Text ) .
Prevention measures such as targeted surveillance and farm isolation , which are largely used to control livestock epidemics , have been shown to be more effective when directed to those higher risk farms [1 , 4] , also called super-spreaders . Understanding the structure of the contact network is therefore crucial to derive effective surveillance strategies . Identification of super-spreaders , however , should be performed by accounting for all transmission pathways and not only the direct ones . In fact , despite it is well known that indirect contacts are less efficient in transmitting infectious diseases compared to direct ones [6 , 40] , our results showed that they can substantially affect the ability of farms to potentially spread a disease within the network system . In particular , our analysis showed that direct and indirect transmission routes shared only a handful of super-spreader farms , indicating that direct and indirect transmission risks were independent from each other . A major consequence of this observation is the need to account for both routes in the definition of contingency plans for the control of potential epidemics where indirect contacts represent an effective route of disease transmission . By considering only one of these transmission routes , we would miss a substantial part of the spreading pattern . This conclusion is strongly supported also by the analysis of infectious risk in the direct and indirect contacts networks . The infection potential was substantially lower for cattle movement ( CM ) compared to the veterinarians total ( VT ) network . This reflects the fact that farms may receive or move cattle only few times in a year and a large fraction of the farms in our database did not trade any animal during the study period . On the contrary , each veterinary practitioner visits a small set of farms on a regular basis , while veterinary officers have frequent visits to farms but visit the same farm only once a year on average . Farm infection potentials derived by using the network of indirect contacts is not correlated with that derived by using the network of direct contacts , thus reinforcing the finding that the transmission pathways of the two contact networks are remarkably different . The most striking result of our study , however , is that the infection potential derived by combining the networks of direct and indirect contacts is considerably larger than the one computed by using only cattle movements or the veterinary network ( Fig 6 ) . This suggests a synergistic effect between the networks of direct and indirect contacts: despite being sparse , direct contacts act as a bridge joining different clusters of potentially infectious contacts due to veterinarians . Despite movements of infected animals usually play a primary role in the spread of livestock epidemics ( since they represent the most efficient transmission route between farms ) , our results highlighted the importance of considering indirect contacts to adequately model between-farm spread of infections . We showed that the combination of different pieces of information included in the infection potential metric is essential to understand the role of farms within the study system . In essence , the infection potential is the expression of the two fundamental components of the contact system , the contact structure and the temporal sequence of the potential infectious contacts [3 , 30 , 39] . In fact , as pointed out in many studies [4 , 37 , 38 , 41–43] , traditional network metrics based on a static representation of the contacts between nodes can hide significant temporal patterns in epidemic network structures of fast spreading diseases , such as FMD . Conversely , the temporal sequence of contacts is crucial in defining the real risk of epidemic spread in a network . For this reason , we applied the concept of infection chains proposed by Dubé et al . [3] and extended by Konschake et al . [39] , providing the estimate of the potential maximum size of epidemics in a network system . Here , as a measure of farm infection potential , we proposed the weighted product of in- times out-infection chains . This new metric aims to extend the concept of basic reproduction number for between-farm transmission provided by Woolhouse et al . [1] , defined as the product of in-times out-degrees . Since the duration of the farm infectious period ( γ ) varies across different diseases , the stability of our results is indicative of the validity of the analysis with regard to a wide range of diseases . However , Konschake and colleagues [39] showed that in German swine livestock , risk ranking of farms according to IC was not stable with respect to variations in γ , in particular for γ shorter than 14 days . The different outcome of our study could be due to the structural differences between dairy and swine livestock industries . We also analysed a weighted version of CM , VO and VP networks ( where the links were weighted proportionally to the number of animals exchanged , in CM , and to the number of on-farm visits , in VO and VP ) . Similarly to the results in the unweighted networks , the measures of strength showed to be only partially correlated with the infection potential . However , in contrast with the unweighted cases ( where VO and VP networks showed larger degree distributions than CM network ) , farms' strength distributions in the weighted networks showed similar patterns for direct contacts ( CM ) and for indirect contacts networks ( VO and VP ) . This might be important for low-prevalence or poorly contagious diseases , in which the number of shipped animals or the recurrence of personnel-mediated contacts might strongly influence the probability of infection spread between farms . Previous questionnaire-based works applied network analysis techniques on small networks of cattle exchanges and on-farm visits by veterinarians [10 , 16] . Similarly to their findings , our results showed that indirect contacts produced a more connected network compared to cattle movements . This finding was supported by the traditional static measures we applied to the investigated networks . The link density and the giant strongly connected component ( GSCC ) were more than one order of magnitude higher in VO and VP networks than in CM network , both in- and out-degree distributions showed longer tails , and the number of isolated farms was substantially greater in CM network , implying a higher number of farms reachable during an epidemics via indirect contacts than via direct contacts . For CM network , the observed value of GSCC was also lower than that observed elsewhere , such as in Scottish [29] and French herds [44] . This result is probably due to the spatial scale considered , since within-province animal exchanges only represent a fraction of the animals introduced in the dairy farms in Parma Province . On the other hand , contact frequency in CM network was slightly higher than in VO and VP networks , underlining a higher recurrence of these contacts . However , the VP network considered in our analysis might represent an underestimate of the real contact frequency , as the dataset used in the analysis only accounted for traceable visits in public datasets . In fact , the average number of monthly visits derived from our datasets was on the lower range than those observed in similar farm systems ( [11 , 14 , 45]; see S1 Text for more details ) . Our analysis suggested that the network of veterinarian contacts is associated to the distribution of MAP positive farms in the Province of Parma , while the network of cattle movements is associated to the distribution of MAP positive farms when the farm system of the whole Emilia-Romagna region is considered . These outcomes can be interpreted as the effect of the different spatial scales on the observed processes . The provincial level represents a spatial scale that could properly fit the process of disease spread due to veterinary practitioners , because of their limited range of activity , and to veterinary officers who operate into sub-provincial public health districts . Conversely , the CM network at provincial level , which only looks at within-province animal exchanges , cannot take into account for the role in disease transmission of animals introduced from outside the province , which represent a significant fraction of the exchanges . Consequently , the relationship between infection status and animal exchanges becomes apparent only at a wider spatial scale , such as at the regional level , where more nodes ( i . e . farms ) and links ( i . e . animal exchanged ) are intercepted by the network . We also showed that spatial clustering failed to explain the observed pattern of MAP infection ( Table 2 ) . Then , our analysis suggests that , at the investigated spatial scale , between-farm transmission was not significantly affected by spatial proximity among farms . The latter result is in agreement with the findings obtained by Ahlstrom et al . [46] on MAP distribution in Canada at intra-provincial scale by using single nucleotide polymorphisms identified through whole genome sequencing . From this finding , they suggested that human driven activities ( such as cattle movements ) are major drivers of MAP transmission at the herd level in contrast to spatially-localized transmission mechanisms ( e . g . wildlife ) , which are crucial in the transmission of other pathogens , such as Mycobacterium bovis [46] . In a similar way , we suggest that , in addition to direct contacts , also indirect contacts can contribute to between-farm MAP spread at local level . Similarly to our results , by analysing bulk tank milk samples , Garcia-Alvarez et al . [16] showed that the veterinary network could explain the difference in the infectious status of dairy farms for some strains of Staphylococcus aureus . However , unlike Garcia-Alvarez and colleagues [16] , we did not know the temporal pattern of MAP infection across farms . Therefore , to account for MAP introductions occurred before the time period considered in the study , we also tested the significance of the relationship between network structures and infectious status by using contact data from years before 2013 and maintaining the infectious status of 2013 . We found that the results were robust compared to the tested year , suggesting a high tendency of farms to maintain contacts with the same elements over time [41] . Our results suggested that between-farm transmission of MAP through fomites might be not negligible at local level . Consistent with our observation , Marcé et al . [32] mentioned the transfer of faeces , manure , slurry , and soiled forage as a viable route of pathogen introduction into farms . The transmission of MAP via contaminated environment has been already pointed out in several empirical works ( specifically , via footwear [24] , via shared equipment [25] , and via settled-dust [48] ) and theoretical models of within-farm transmission [49 , 50] . In this work , we defined the positive/negative infection status of farms for paratuberculosis using data obtained by Ricchi et al . [21] through real-time PCR on bulk tank milk samples . Real-time PCR techniques on bulk tank milk display only intermediate sensitivity in detecting MAP in cattle herds [47] , in particular because they do not detect not-excreting infected animals . It follows that the obtained prevalence represents an underestimate of the between-farm true prevalence in the area . However , for the purpose of our analysis , the major focus was the contamination status of farms; in that sense , the real-time PCR assessment of the presence of MAP in the herd environment is functional to estimate the ability of the farms to spread the infection , especially in the case of fomite contaminations and in the short run .
Our results highlight the urgency of further defining the indirect contacts between farms established by operators like veterinarians , milk , feed , and slaughterhouse trucks , and artificial insemination technicians , in addition to estimates of non-recordable movements such as neighbours and service providers , that even if less infectious , would be more frequent farm visitors [51] . The availability of personnel visits data would improve the surveillance and control of epidemics through modelling approaches . In particular , this could make it possible to effectively tackle diseases among which we recognize some of the most devastating threats to the modern livestock industry , above all , foot-and-mouth disease and avian influenza . | Farm-to-farm contacts due to shared operators and vehicles–such as veterinarians , hoof-trimmers , milk and rendering trucks–are generally considered important for the spread of many infectious diseases in livestock systems . These contacts are usually defined as indirect , as opposed to animal movements which are defined as direct contacts . The actual significance of indirect contacts is still poorly understood due to study limitations deriving from their highly diverse and complex nature and to privacy issues in data collection . Thanks to the availability of high-resolution data in space and time on veterinarian on-farm visits in a dairy farm network in Northern Italy , we showed through network analysis techniques that between-farm indirect contacts are more widespread and display significantly different patterns compared to direct contacts . Using infection data on paratuberculosis , we also found that indirect contacts can support disease spread at local scale , while direct contacts ( due to long distance animals exchanges ) can support disease spread on a larger spatial scale . Our results corroborate the conclusion that the role of indirect contacts in livestock disease spread is essential and deserves a deeper understanding . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusions"
] | [
"livestock",
"medicine",
"and",
"health",
"sciences",
"animal",
"diseases",
"ruminants",
"infectious",
"disease",
"epidemiology",
"vertebrates",
"animals",
"mammals",
"paratuberculosis",
"farms",
"veterinarians",
"zoology",
"veterinary",
"science",
"infectious",
"diseases",... | 2017 | The Potential Role of Direct and Indirect Contacts on Infection Spread in Dairy Farm Networks |
Unlike most bacterial species , Mycobacterium tuberculosis depends on the Clp proteolysis system for survival even in in vitro conditions . We hypothesized that Clp is required for the physiologic turnover of mycobacterial proteins whose accumulation is deleterious to bacterial growth and survival . To identify cellular substrates , we employed quantitative proteomics and transcriptomics to identify the set of proteins that accumulated upon the loss of functional Clp protease . Among the set of potential Clp substrates uncovered , we were able to unambiguously identify WhiB1 , an essential transcriptional repressor capable of auto-repression , as a substrate of the mycobacterial Clp protease . Dysregulation of WhiB1 turnover had a toxic effect that was not rescued by repression of whiB1 transcription . Thus , under normal growth conditions , Clp protease is the predominant regulatory check on the levels of potentially toxic cellular proteins . Our findings add to the growing evidence of how post-translational regulation plays a critical role in the regulation of bacterial physiology .
Our understanding of how bacteria regulate cellular processes has long focused on the role of transcription factors in the modulation of cellular responses . In eukaryotes , however , elucidation of the ubiquitin-proteasome pathway has illustrated that targeted degradation of functional proteins is often employed as a regulatory mechanism[1] , [2] . Like eukaryotes , bacteria possess an array of compartmentalized proteolytic complexes , capable of degrading proteins into smaller polypeptides and amino acids[3] , [4] . Initially , they were thought to maintain protein quality control through the recognition of misfolded , aberrant protein products . Several studies identified an array of endogenous proteins that were targeted for degradation in bacteria[5] , [6] . While this suggested an active role of proteolysis in the regulation of bacterial physiology , it has been difficult to determine the functional significance of protein degradation by these proteolytic machines in bacteria . Mycobacterium tuberculosis ( Mtb ) , the causative agent of tuberculosis that kills nearly 1 . 3 million people annually[7] , may provide unique insights into the importance of targeted protein degradation in bacteria . In most model prokaryotes , where the compartmentalized proteases have been extensively studied , they are largely dispensable for normal growth[8] , [9] . However , a genome-wide screen for essential genes in Mtb suggested that numerous proteolytic complexes ( namely Clp , FtsH , and HtrA ) were absolutely required for cell survival , providing evidence for their critical role in bacterial physiology[10] . Further studies on the Clp complex revealed that inhibition or depletion of the protease results in mycobacterial death both in vitro and in a mouse model of infection[11] , [12] . The ATP-dependent Clp proteolytic complex is composed of a serine proteolytic core that interacts with a set of regulatory ATPases . In mycobacteria , the ClpP proteolytic core , normally a homomeric complex in most bacteria , is actually comprised of two stacked heptameric rings of ClpP1 and ClpP2 multimers [13] . Targeted proteins enter through an axial pore that is regulated by the interaction of the ClpP1P2 complex with various AAA+ ATPases ( ClpC1 and ClpX in Mtb ) [14] , [15] , thus forming the full Clp complex . In Mtb , though endogenous protein substrates have yet to be identified , Clp has been implicated in the recycling of abnormal peptides stalled on the ribosome , through recognition of SsrA-tagged proteins[11] . In this study , we constructed a conditional ClpP1P2 protease mutant in Mtb and compared the proteomes of ClpP1P2-deficient cells to wildtype Mtb using recently developed MS3-based isobaric multiplexed quantitative proteomics . We identified one of the targets of the Clp protease as WhiB1 , an essential transcriptional repressor that contains an iron-sulfur cluster . Blocking Clp-dependent degradation of WhiB1 resulted in stabilization of WhiB1 in mycobacteria . This stabilized allele was functional but toxic even at physiological levels , suggesting that proteolysis is the primary regulatory check on the amount of WhiB1 present . These data establish a mechanism for the essentiality of Clp protease in mycobacteria , and provide critical evidence of the dominant role that protein turnover plays in regulating bacterial physiology .
The essentiality of Clp protease has been demonstrated in the non-pathogenic , fast growing model organism , Mycobacterium smegmatis ( Msm ) , but not in Mycobacterium tuberculosis ( Mtb ) . We constructed a Clp protease conditional mutant in Mtb that took advantage of complementary systems of promoter regulation and inducible protein degradation , recently developed for use in mycobacteria ( FIGURE 1A ) . Addition of anhydrotetracycline ( ATc ) to this strain , denoted P750-clpP1P2DAS , simultaneously repressed transcription of the clpP1P2 operon and led to the degradation of existing ClpP2 protein[16] . At low inoculums ( 5×105 CFU/mL ) , addition of ATc ( 1 . 5 µg/mL ) to P750-clpP1P2DAS had a bactericidal effect , demonstrating that ClpP1 and ClpP2 are essential in Mtb ( FIGURE 1B ) . At higher inocula ( 1×107 CFU/mL ) , depletion also inhibited growth ( FIGURE S1 ) . These higher inocula allowed us to harvest cellular material for protein and transcript expression analysis . Production of SspB resulted in profound depletion of ClpP2-DAS within 48 hours , or two replicative cycles ( FIGURE 1C ) . Furthermore , qPCR analysis of clpP1 and clpP2 mRNA revealed that by 48 hours , transcription at the clpP1P2 locus was significantly repressed in cultures exposed to ATc ( FIGURE 1D ) . The P750-clpP1P2DAS strain enabled us to conditionally deplete the ClpP1P2 proteolytic core and explore the mechanism of essentiality of Clp protease in Mtb . We hypothesized that growth inhibition observed in this strain resulted from an accumulation of Clp substrates that were either toxic to the cell or that repressed normal growth . To identify potential substrates , we utilized LC/MS/MS3-based multiplexed quantitative proteomics with isobaric tandem mass tags ( TMT ) to quantify and compare the proteomes of Clp deficient and wildtype Mtb[17] . Briefly , six isobaric TMT ( TMT126-131 ) tags ( Thermo Fisher ) have isobaric masses and are used to label the peptides from the proteomes of six different experimental conditions . When pooled , the same peptides from all conditions co-elute during fractionation . However , upon isolation and MS3 fragmentation of the initial single MS peptide peak , the six TMT molecules fragment differently between the 126–131 mass range . The resultant peak heights of the TMT ions represent the relative quantities of a given peptide between the different samples . We harvested P750-clpP1P2DAS grown from an initial inoculum of 1×107 CFU/mL for 48 hours either in the absence or presence of 1 . 5 µg/mL ATc , with biological triplicates ( a total of six conditions and proteomic samples ) . Immunoblot analysis demonstrated a significant knockdown of ClpP2 in each of the cultures exposed to ATc ( FIGURE 2A ) . Through TMT labeling followed by MS3-based quantitative proteomics[18] , we quantified 1564 proteins . Hierarchical clustering using Pearson correlational analysis revealed a strong correlation between the proteomes of the three biological replicates ( FIGURE 2B ) [19] . A total of 132 proteins were significantly over-represented in mutant bacteria . We defined significant over-representation ( or under-representation ) as an average change of two-fold or more between mutant and wildtype conditions , and a p-value of less than or equal to 0 . 01 across the biological replicates ( FIGURE 2C ) . Gene ontology ( GO ) analysis failed to reveal enrichment of any particular GO class among the over-represented proteins in Clp-depleted bacteria[20] . However , there were numerous transcriptional modulators with increased abundance in mutant bacteria ( TABLE S1 ) . Comparing protein intensity values between mutant and wildtype bacteria revealed a set of proteins that were highly over-represented ( >5-fold increase , 24 proteins ) , moderately over-represented ( 2–5-fold increase , 108 proteins ) , or under-represented ( >2-fold decrease , 23 proteins ) ( see examples in FIGURE 2D ) . ClpP1 and ClpP2 were the two most under-represented proteins in the screen , both depleted over 90% in mutant cells compared to wildtype . Protein accumulation upon Clp depletion could be the result of ineffective proteolysis due to reduced levels of Clp protease , but may also be due to a transcriptional upregulation of certain stress-induced proteins as a reaction to Clp depletion . The accumulation of numerous heat shock proteins suggested that , to some extent , this was the case . To determine the subset of over-represented proteins that were likely Clp substrates , we used quantitative PCR analysis to compare the transcript levels of putative substrates in Clp-deficient and wildtype bacteria . This analysis revealed three groups , one where increases in protein abundance upon Clp depletion could be explained by mRNA abundance , a second where there was a clear discordance between protein amount and transcript level change , and a third where the difference was less clear ( FIGURE 3A ) . We posited that the latter two groups were more likely to contain Clp substrates , as the changes in protein abundance were more likely due to protein-level regulation than transcriptional upregulation . To further validate potential substrates of the Clp protease , we turned to a conditional clpP2 mutant ( clpP2_ID Msm ) we had previously developed in Msm[11] . In this strain , an analogous protein degradation system leads to rapid loss of ClpP2 protein . We previously reported that this mutant allowed for rapid degradation of ClpP2 , and ClpP2 depletion resulted in the accumulation of a reporter substrate due to decreased turnover . We performed similar quantitative proteomic and transcriptional analysis on this strain comparing proteins both with and without ClpP2 depletion ( FIGURE S2 , TABLE S2 ) . Proteomic analysis revealed 107 proteins elevated in the Msm mutant compared to wildtype Msm , and a 9 . 3% overlap ( n = 10 proteins ) with the results from the Mtb screen ( TABLE S3 ) . Through our combined Mtb and Msm analysis , we identified two essential transcriptional effectors , CarD and WhiB1 , with increased protein abundance and insignificant changes in transcript level between Clp-deficient and wildtype bacteria . To assay WhiB1 and CarD degradation by Clp , we constructed GFP-fusion proteins by adding GFP to either the N- or C-terminus of each protein , and producing these fusions on an ATc-inducible promoter . These fusion proteins allowed us both to alter protease recognition by modifying a potential , terminal recognition sequence and follow the accumulation of the resultant protein . Additionally , by tightly regulating the transcription of these constructs on an inducible plasmid , we could prevent transcriptional modulation from confounding our results . Production in wildtype Msm revealed differential abundances , as measured by fluorescence , between N- and C-terminal fusions for each protein . ( FIGURE 3B , black bars ) . Despite the differential fluorescence between the two WhiB1 fusions , quantitative PCR analysis revealed that inducible production of GFP-WhiB1 and WhiB1-GFP led to similar amounts of transcript in the cell , suggesting that differential fluorescence observed was regulated at the protein level ( FIGURE S3A ) . To demonstrate that this discrepancy was specifically due to Clp protease , we introduced the fusions into clpP2_ID Msm , where addition of ATc simultaneously induced production of each fusion construct and depletion of ClpP2 , and assessed protein abundances . For both WhiB1 and CarD , depletion of ClpP2 resulted in an increase in the abundance of the N-terminal GFP fusion relative to wildtype Msm , as measured both by fluorometry and immunoblot ( FIGURE 3B , FIGURE S3B ) , presumably reflecting stabilization due to reduced turnover . Several other proteins exhibited different effects . For both RpL28 and DnaA , C-terminal GFP fusions were actually less stable than their respective N-terminal constructs . In the case of RpL28 , depletion of ClpP2 stabilized both fusions , suggesting that the motif for Clp recognition was internal and not dependent on an exposed terminus . DnaA fusions were not stabilized at all upon ClpP2 depletion suggesting that DnaA was either not a substrate or that both free ends were required for proteolysis ( FIGURE S4A ) . From these results , it appears that there may be numerous recognition motifs that lead to Clp-dependent degradation . Unfortunately , bioinformatics analysis did not reveal any common motifs among the proteins identified as putative Clp substrates in our proteomic screening . To test whether the C-terminus of WhiB1 was sufficient to confer destabilization and recognition by Clp protease we constructed a variety of fusions where a variable number of C-terminal WhiB1 residues were appended to the end of GFP . We found that the addition of the last fifteen , nine , and five amino acids of WhiB1 to GFP destabilized the protein with respect to wildtype GFP . Furthermore , wildtype levels of GFP were restored in these constructs upon ClpP2 depletion ( FIGURE 3C ) . Similarly , the C-terminal fifteen residues of CarD destabilized GFP ( FIGURE S4B ) . We noted that prolonged over-production of WhiB1-GFP inhibited the growth of mycobacteria and led to cell lysis ( FIGURE 4A and 4B ) . This effect appeared to be specific to the C-terminal GFP fusion . To determine if the protein was still functional , and toxicity was not due to a non-specific effect , we determined if the WhiB1 fusion , as has been previously shown with wildtype WhiB1[21] , could act as an auto-repressor . We used RT-PCR to determine transcription of the native protein in its normal chromosomal location , and found that the fusion protein was still able to serve as a repressor ( FIGURE 4C ) . To further test the functionality of the WhiB1-GFP allele , we built a more sensitive reporter of promoter activity by fusing the putative whiB1 promoter to luciferase . By introducing this construct into strains inducibly producing the GFP fusions , we could simultaneously monitor fluorescence for protein abundance and stability and luminescence for whiB1 promoter activity . For both fusions , the amount of repression appeared to correlate inversely with the amount of fusion protein present ( FIGURE 4D ) . These observations suggest that the toxicity observed for WhiB1-GFP could be due to increased protein abundance , a result of stabilization from lack of recognition of the blocked C-terminus by Clp protease . In all of the above experiments , the fusion whiB1 constructs were expressed on multi-copy plasmids , perhaps leading to dramatic overexpression . To test if physiological levels of the degradation-deficient WhiB1 protein would be lethal , we constructed integrative plasmids with the native whiB1 promoter upstream of each fusion gene . This construct would lead to physiological levels of WhiB1 under natively regulated conditions . Transformation of these plasmids into Msm resulted in significantly lower transformation efficiencies for the plasmid bearing WhiB1-GFP compared to those with GFP-WhiB1 and WhiB1wt ( FIGURE 4E ) . Colonies that resulted from WhiB1-GFP transformations plasmid were significantly smaller and took nearly twice as long to be seen than the GFP-WhiB1 or control WhiB1wt transformants . This suggests a considerable fitness cost associated with even a single copy of the stabilized allele is regulated by the native promoter .
Clp proteases have two primary functions . Like other degradative proteases , they play a role in protein quality control , degrading improperly synthesized or folded proteins . Indeed , this does appear to be the case in mycobacteria[11] . However , Clp proteases in other species also play an important regulatory role in degrading endogenous proteins . In other bacteria , identifying such Clp substrates has been facilitated by the use of in vitro systems in which Clp components have been inactivated so that binding could be assayed in the absence of proteolysis[22] . Thus far , due to the heteromeric composition of Clp in Mtb and its stringent requirements for normal growth , we have been unable to use an analogous approach . Instead , we relied on an in vivo assay for the accumulation of substrates taking advantage of recently developed , highly accurate proteomic methods . This approach is somewhat problematic as depleting Clp is quite toxic and it can be difficult to disentangle protein accumulation due to cellular responses from that resulting from lack of proteolysis of direct substrates . Nevertheless , using a combination of transcriptional analysis and sequence modifications to alter protein recognition , we were able to unambiguously define some new Clp protease substrates . Alternative approaches to identifying ClpP substrates have been highly successful . For example , using an inactive ClpP mutant in Staphylococcus aureus , Feng , et al . , [23] were able to affinity purify substrates . Unfortunately , despite many attempts , we were unable to use this method in mycobacteria . And , since protein abundance can be regulated both by transcription and degradation , using transcriptional regulation to screen for likely substrates will exclude some actual substrate proteins . For example , while CarD is not transcriptionally upregulated in Msm , it is mildly upregulated in Mtb; yet , we find that it is a substrate . Thus far , using any approach has not clearly identified strong consensus sequences for degradation . In E . coli , the N end rule governs degradation of some substrates[24] others have conserved C terminal di-alanines[22] but most have none . Certainly , among the proteins we have found to be potential substrates we cannot identify a consensus . Comparing degradation in two different mycobacterial species helps provide some additional confidence in recognized substrates . Although we used different regulatory approaches to deplete proteolytic subunits , we expect to have a good deal of biological concordance . ClpP1 and ClpP2 are both necessary for proteolysis to occur[11] . Depleting either or both should have similar effects . Our results suggest that there are significant similarities among substrates and , for example , both WhiB1 and CarD can be degraded by Clp in both species . Do these results help us to understand why Clp is essential for bacterial growth and survival ? Certainly , altering the sequence of WhiB1 so that it is no longer easily degraded results in cellular toxicity . Even when this stabilized allele is expressed at physiologic levels , it appears to be quite toxic . Thus , part of the function of Clp is to degrade the WhiB1 protein and keep its levels in check . This is reminiscent of the situation in Caulobacter crescentus , one of the few bacteria where Clp protease is also essential . In this organism , degradation of CtrA by Clp is absolutely necessary for the transition of a non-replicating , motile swarmer cell into a replicating , stalked cell[25] , and loss of Clp protease activity results in cellular growth arrest . In mycobacteria , we can further establish the importance of post-translational regulation in prokaryotes by showing that WhiB1 levels are coordinated by a mixture of transcriptional and protein-level regulation , but that it is the interruption of protein-level regulation through Clp inhibition that is essential . Why could the turnover of WhiB1 be required for normal growth ? In mycobacteria , WhiB proteins are capable of binding redox-sensitive [4Fe-4S] clusters , which can serve as redox-active co-factors or as switches that reflect the reductive and oxidative potential of a cell[26] . The WhiB proteins may be disulfide reductases[27] , but are certainly transcription factors capable of modulating cellular processes that are intimately tied to the redox state of the cell or perceived oxidative stress[28] . In Mtb , WhiB1 is an essential DNA binding protein capable of auto-repressing its own transcription[21] . ChIP-Seq to determine the WhiB1 regulon in Mtb has been undertaken , and preliminary data suggests the presence of 71 binding sites for WhiB1 , thirteen of which are essential[29] . Accumulation of WhiB1 might repress at several sites resulting in the lack of synthesis of critical metabolites such as heme and riboflavin . Alternatively , supraphysiologic levels of WhiB1 , due to absence of turnover , may result in binding of the transcription factor to low affinity sites with repression of other essential genes . Does stabilization of WhiB1 account for all of the essentiality of Clp ? Several lines of evidence suggest that this is unlikely to be true . Clp is the primary proteolysis system for degrading SsrA-tagged proteins , which result from trans-translation due to ribosomal stalling[30] . We have found that the small RNA , tmRNA , required for producing this tag , is itself essential[31] . In fact , this trans-translation system is at least one of the targets of the antimycobacterial drug pyrazinamide[32] . Loss of Clp would result in accumulation of these tagged proteins . Moreover , not only are the ClpP1 and ClpP2 protease subunits required for optimal in vitro growth but two of the alternate adapter proteins , ClpX and ClpC1 are essential as well . The fact that both these adapters are required supports the hypothesis that there are multiple substrates that need to be recognized ( at least one per ATPase ) by Clp protease to facilitate normal growth . For example , it is interesting to note that CarD has been implicated in the stringent response , and directly interacts with the beta-subunit of the RNA polymerase to down regulate transcription of the translational machinery and amino acid biosynthetic enzymes[33] . While stabilization of CarD alone is not sufficient to cause toxicity in mycobacteria , the transcriptional repression of enzymes important for vegetative growth facilitated by CarD may contribute to the growth inhibition observed upon depletion of Clp protease , and partially explain the essentiality of Clp protease in mycobacteria . Transcriptional regulation plays a critical role in bacterial adaptation to new environments . However , much of regulation is likely to be post-transcriptional and no less important for bacterial survival . In the case of WhiB1 , M . tuberculosis employs both transcriptional and post-transcriptional ( Clp-mediated proteolysis ) forms of regulation . Clearly , protein degradation mediated by Clp is required even in the absence of clear environmental stressors . Identifying the substrates for the Clp protease and other essential degradative proteases will help move us towards a more holistic understanding of how bacteria coordinate critical cellular activities through the integration of transcriptional and proteolytic regulation .
Msm mc2155 ( Msm ) or Mtb H37Rv were grown at 37°C in Middlebrook 7H9 broth with 0 . 05% Tween 80 and ADC ( 0 . 5% BSA , 0 . 2% dextrose , 0 . 085% NaCl , 0 . 003 g catalase/1 L media ) . Mtb was additionally supplemented with oleic acid ( 0 . 006% ) . For growth curves , overnight cultures were diluted into the appropriate media and growth was either measured by OD600 or colony forming units per mL . A complete list of plasmids and primers used in this study can be found in TABLE S4 and TABLE S5 . Detailed procedures on strain construction can be found in the Supplemental Experimental Procedures . In order to measure the abundance of GFP fusion proteins , cultures within one experiment were normalized based on OD600 values , spun down to remove media , and resuspended in 100 µL of PBS in a clear bottom 96 well plate . For luminescence , cultures were normalized based on OD600 values , and 150 µL of culture was used for measuring luciferase activity . 50 µL of Cell Culture 5X Lysis Reagent ( Promega ) was added to cultures , and samples were agitated for 10 min on an orbital shaker , at room temperature . Next , 75 µL of Luciferase Assay Substrate ( Promega ) was added to each sample and directly taken for measurement . Fluorescence was measured at 485/538 nm , and luminescence was measured at an exposure time of 10 milliseconds , by the Fluroskan Ascent FL plate reader ( ThermoScientific ) . Results represent the median +/− standard deviation of biological replicates . In Mtb and Msm , RNA was generated from equivalent of 20 mLs of cells at OD600 0 . 5 . Cultures were spun down , and subjected to bead beating ( 3X 45 sec each , 5 min on ice between pulses ) after resuspension in TRIzol . After chloroform phase separation , genetic material was precipitated with isopropanol , resuspended in dH2O , and RNA was purified using RNeasy Mini Kit ( Qiagen ) . To ensure no contamination from genomic DNA , purified RNA was subjected to an additional round of DNase digestion using the TURBO DNA-free Kit ( Invitrogen ) . cDNA was created from isolated equal concentrations of RNA with the SuperScript III First Strand Synthesis System ( Invitrogen ) . Quantitative PCR was performed with the SYBR FAST qPCR kit ( KapaBiosystems ) using the Applied Biosystems 7500 Fast Real-Time PCR System . All experiments were done using biological replicates , and representative experiments are depicted +/− standard error of mean of technical replicates . For Mtb proteomics , P750-clpP1P2DAS was diluted to a starting OD600 0 . 02 in 900 mLs of 7H9 media . This culture was split into six 150 mL cultures , and 1 . 5 µg/mL ATc was added to three batches , while three were left to grow without induction . After 48 hours , cultures were spun down ( 10 min , 4000 rpm , 4°C ) and washed 3X with PBS . Cultures were then resuspended in 1 mL Urea Lysis Buffer ( 8 M urea in 50 mM Tris pH 8 . 2 , 75 mM NaCl , 50 mM NaF , 50 mM β-glycerophosphate , 1 mM Na orthovanadate , Roche Complete EDTA-free Protease Inhibitor Cocktail tablets ) , and subjected to bead beating ( 3X 45 sec each , 5 min on ice between pulses ) . Cell lysates were spun down ( 10 min , 13 , 000 rpm , 4°C ) , and samples were reduced with DTT ( final concentration of 5 mM ) for 30 min at 37°C and cooled to room temp for 15 minutes . Samples were then alkylated with iodoacetamide ( final concentration of 14 mM ) for 30 min at room temperature in the dark . Adding an additional 5 mM DTT and incubating samples at room temperature in the dark for 15 min quenched excess iodoacetamide . To remove samples from the BL3-level facility , proteins were precipitated with 20% trichloroacetic acid and incubated on ice for a hour . Proteins were pelleted by centrifugation ( 30 min , 13 , 000 rpm , 4°C ) , and pellets were washed twice with acetone . Samples were resuspended in 8 M urea containing 50 mM TRIS pH 8 . 5 , diluted , and protein amounts were quantified using a BCA assay ( ThermoScientific ) . As the proteomics screen in Msm was performed prior to the development of MS3-based proteomics for TMT analysis , we performed MS2-based quantitation of TMT peptide signals . For Msm proteomics , Msm/pTet ( OR ) ::HIV2pr and clpP2_ID Msm were diluted to a starting OD600 0 . 05 in 450 mLs . ATc ( 100 ng/mL ) was added to each strain , and cultures were divided into 150 mL batches . Samples were harvested at 0 h , 5 h , and 11 h post addition of ATc , spun down ( 10 min , 4000 rpm , 4°C ) , and washed 3X with PBS . Protein samples were prepared in a similar fashion as above , except no TCA precipitation was needed , and protein quantitation by the BCA assay was done prior to sample reduction and alkylation . Proteins isolated from Mtb and Msm were digested overnight with Lys-C ( Wako ) in a 1∶100 enzyme:protein ratio in 4 M urea and 50 mM Tris-HCl ( pH 8 . 2 ) . Digests were acidified with formic acid to a pH of ∼2–3 , and subjected to C18 solid-phase extraction ( Sep-Pak , Waters ) . Isobaric labeling of the peptides was accomplished with sixplex TMT reagents ( Thermo Scientific ) . Reagents ( 0 . 8 mg ) were dissolved in 40 µl acetonitrile , and 20 µl of the solution was added to 200 µg of peptides dissolved in 100 µl of 50 mM HEPES ( pH 8 . 5 ) . After 1 h at room temperature , the reaction was quenched by adding 8 µl of 5% hydroxylamine for 15 minutes . Half of each of the labeled reactions was pooled into one vial , acidified with formic acid , diluted ACN to 5% final volume and subjected to C18 solid-phase extraction . Details on sample fractionation , liquid chromatography electrospray ionization tandem mass spectrometry , and data processing/analysis can be found in the Supplemental Experimental Procedures . | To date , studies on the regulation of physiology and virulence in Mycobacterium tuberculosis ( Mtb ) have focused on how transcriptional changes lead to adaptation . Interestingly , Mtb has numerous proteases that are essential for normal growth suggesting that protein turnover may also play an important regulatory role in the pathogen . We used novel methods to identify the set of proteins that are degraded by the essential Clp protease . The degradation of one protein , WhiB1 , was required for normal growth confirming that inhibiting turnover of certain substrates can have a lethal effect . The understanding of essential pathways in Mtb will be important for the discovery of novel drugs to aid in the global fight against tuberculosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"microbial",
"physiology",
"microbial",
"pathogens",
"biology",
"microbiology",
"bacterial",
"pathogens"
] | 2014 | Post-Translational Regulation via Clp Protease Is Critical for Survival of Mycobacterium tuberculosis |
Lynch syndrome ( LS ) is a hereditary cancer predisposition caused by inactivating mutations in DNA mismatch repair ( MMR ) genes . Mutations in the MSH6 DNA MMR gene account for approximately 18% of LS cases . Many LS-associated sequence variants are nonsense and frameshift mutations that clearly abrogate MMR activity . However , missense mutations whose functional implications are unclear are also frequently seen in suspected-LS patients . To conclusively diagnose LS and enroll patients in appropriate surveillance programs to reduce morbidity as well as mortality , the functional consequences of these variants of uncertain clinical significance ( VUS ) must be defined . We present an oligonucleotide-directed mutagenesis screen for the identification of pathogenic MSH6 VUS . In the screen , the MSH6 variant of interest is introduced into mouse embryonic stem cells by site-directed mutagenesis . Subsequent selection for MMR-deficient cells using the DNA damaging agent 6-thioguanine ( 6TG ) allows the identification of MMR abrogating VUS because solely MMR-deficient cells survive 6TG exposure . We demonstrate the efficacy of the genetic screen , investigate the phenotype of 26 MSH6 VUS and compare our screening results to clinical data from suspected-LS patients carrying these variant alleles .
Lynch syndrome ( LS ) is an autosomal-dominantly inherited predisposition to a variety of malignancies at a young age , mainly colorectal cancer ( CRC ) and endometrial cancer ( EC ) [1] . It is caused by inactivating germ-line mutations in the DNA mismatch repair ( MMR ) genes MLH1 , MSH2 , MSH6 or PMS2 , or a deletion in the 3’ region of the EPCAM gene that affects MSH2 expression [2–6] . The DNA MMR system is essential for the fidelity of DNA replication . Its primary function is the correction of base-base mismatches and insertion-deletion loops that may arise during DNA replication . Base-base mismatches are recognized by the MSH2-MSH6 heterodimer while MSH2-MSH3 detects loops of unpaired bases . Following mismatch binding , the MSH heterodimers recruit another heterodimer , MLH1-PMS2 , to coordinate removal and resynthesis of the error-containing strand [7–9] . A second function of the DNA MMR system is to mediate the toxicity of certain DNA damaging agents such as methylating agents and thiopurines . These DNA damaging agents create adducts in the genome that give rise to mismatches when replicated . The DNA MMR system recognizes the mismatches but will remove the incorporated nucleotide rather than the lesion itself , creating a repetitive cycle of nucleotide incorporation and deletion that ultimately leads to DNA breakage and cell death [10 , 11] . In the absence of MMR , cells tolerate methylation damage , but consequently show high levels of DNA damage-induced mutagenesis on top of a strongly elevated level of spontaneous mutagenesis [12] . LS patients inherit a functional and a mutant copy of one of the DNA MMR genes . For cells to become MMR-deficient and develop a mutator phenotype that accelerates carcinogenesis , somatic loss of the wild-type allele is required [13] . Microsatellite instability ( MSI ) , i . e . , length alterations of repetitive sequences like ( CA ) n or ( A ) n , and loss of immunohistochemical staining ( IHC ) for MMR proteins are considered hallmarks of LS tumors . Analysis of MSI and IHC on tumor tissue can identify patients who may suffer from LS . For a definitive LS diagnosis , however , sequence analyses must reveal a pathogenic germline mutation in one of the DNA MMR genes or the 3’ region of EPCAM [14 , 15] . Many LS-associated sequence variants are nonsense and frameshift mutations that clearly truncate the protein and unambiguously abrogate MMR activity . Missense mutations that only alter a single amino acid are also frequently identified in suspected-LS patients . The functional implications of these variants are less clear . Consequently , the diagnosis of suspected-LS patients carrying missense variants is difficult in the absence of clear segregation and functional data . As long as the phenotype of these variants of uncertain significance ( VUS ) is unclear , non-carriers cannot safely be discharged from burdensome surveillance programs [16] . Surveillance programs have proven to significantly reduce morbidity and mortality in LS patients [1 , 17 , 18] , but pose unnecessary psychological and physical stress on carriers of innocent VUS as well as pressure on preventive healthcare . Therefore , techniques that characterize MMR gene VUS and enable the identification of individuals at risk are urgently needed . While in the past primarily MSH2 and MLH1 were sequenced to identify LS-causing mutations , in recent years MSH6 has been gained fame for causing LS due to the advancement of DNA sequencing . However , MSH6 mutation carriers can be difficult to diagnose because they may not entirely fulfill the criteria for LS diagnosis: their age at cancer onset is often later than for MLH1 and MSH2 mutation carriers , and their tumors occasionally stain for MSH6 and have no or low MSI [19–21] . We therefore extended the applicability of the oligonucleotide-directed mutagenesis screen we recently described for the identification of pathogenic MSH2 variants to MSH6 variants [22] . The genetic screen uses oligonucleotide-directed gene modification ( oligo targeting ) [23] to introduce variant codons into the endogenous Msh2 gene of mouse embryonic stem cells ( mESCs ) and subsequently identifies pathogenic variants by selecting for cells that are resistant to the thiopurine 6-thioguanine ( 6TG ) . Here we present the applicability of this screen for the characterization of MSH6 VUS .
The oligonucleotide-directed mutagenesis screen takes a four step approach to the identification of pathogenic MSH6 mutations ( Fig 1 ) : 1 ) site-directed mutagenesis to introduce the variant of interest into a subset of Msh6+/- mESCs , 2 ) selection for cells that consequently lost MMR capacity , 3 ) PCR analysis to exclude cells that lost MMR capacity due to loss of the Msh6+ allele ( loss of heterozygosity events ) , 4 ) sequence analysis to confirm the presence of the planned mutation in the MMR-deficient cells . mESCs provide a good study model because the human and mouse MSH6 amino acid sequences share over >86% identity ( S1 Fig ) and mouse models can be made from these cells if VUS need to be studied in vivo . Msh6+/- mESCs only contain one wild type Msh6 allele ( Msh6+ ) ; the other allele was disrupted by a puromycin-resistance gene and therefore inactivated ( Msh6- ) [24] . Hence introduction of a specific mutation into the one active Msh6 allele will lead to expression of solely the variant protein and allow immediate investigation of its phenotype . To achieve this , Msh6 was site-specifically mutated by oligo targeting , a gene modification technique that uses short single-stranded locked-nucleic-acid-modified DNA oligonucleotides ( LMOs ) ( with either sense or antisense orientation ) to substitute a single base pair at a desired location . LMO-directed base-pair substitution can be achieved at an efficiency of 10−3; thus , about 1 in every 1000 LMO-exposed Msh6+/- mESCs will contain the desired mutation [23] . To determine whether the substitution abrogated Msh6 activity and this subset of cells consequently lost MMR activity , LMO-exposed mESCs were treated with 6TG . The thiopurine DNA damaging agent 6TG is highly toxic to MMR-proficient but only moderately toxic to MMR-deficient cells [11] . Therefore , the appearance of colonies that survived mild 6TG selection is indicative for loss of MMR capacity . Loss of MMR capacity may arise due to the introduced mutation or due to loss of heterozygosity events that caused loss of the functional Msh6 allele . To exclude the latter from further investigation , a PCR that detected the presence of both the disrupted and non-disrupted Msh6 alleles was performed [24] . 6TG-resistant colonies that maintained both Msh6 alleles were sequenced to confirm the presence of the planned mutation . To demonstrate the ability of the oligonucleotide-directed mutagenesis screen to distinguish pathogenic MSH6 mutations from polymorphisms , a proof of principle study was performed with MSH6 variants G1139S and L1087R that were previously proven to be pathogenic and not pathogenic , respectively [25] , as well as all classified pathogenic and not pathogenic missense variants described in the International Society for Gastrointestinal Hereditary Tumours ( InSiGHT ) colon cancer variant database ( http://insight-group . org/ ) . This database uses available clinical , in vitro and in silico data to categorize DNA MMR gene sequence variants according to a five-tiered classification scheme as: class 5 , Pathogenic; 4 , Likely pathogenic; 3 , Uncertain; 2 , Likely not pathogenic; and 1 , Not pathogenic [26] . Msh6+/- mESCs were first exposed to antisense oriented LMOs encoding the desired base-pair substitution . If subsequent 6TG selection did not reveal resistant colonies encoding the planned mutation , the screen was repeated with sense oriented LMOs . LMO-mediated introduction of both , pathogenic and not pathogenic variants led to the appearance of 6TG-resistant colonies . For each LMO , we picked and analyzed 18 colonies . The vast majority of 6TG-resistant colonies obtained with LMOs encoding polymorphisms had lost the wild-type Msh6 allele by loss of heterozygosity ( LOH ) events , as inferred from allele-specific PCR analysis . Sequencing of the few 6TG-resistant colonies that had retained both Msh6 alleles ( ±6% ) , did not detect any mutation ( Fig 2A ) . These background colonies apparently arose from cells that for unknown reasons survived 6TG exposure . Of the 6TG-resistant colonies that emerged following LMO-mediated introduction of pathogenic mutations , ±40% still contained both Msh6 alleles . Sequence analysis detected pathogenic mutations in all but one of these 6TG-resistant colonies ( Fig 2B; S2A Fig ) . Thus , the oligonucleotide-directed mutagenesis screen detected all 4 pathogenic mutations and not one of the 5 non-pathogenic variants , indicating it is capable of distinguishing pathogenic MSH6 mutations from polymorphisms . We used the oligonucleotide-directed mutagenesis screen to investigate the phenotype of 18 MSH6 VUS described in literature and the InSiGHT database as well as 8 MSH6 VUS detected in suspected-LS patients from the Erasmus Medical Center Rotterdam and the Radboud University Medical Center Nijmegen ( see S1 and S2 Tables for clinical data [27–38]; see S3 Fig for location of variants in MSH6 [39 , 40] ) . Of the 26 variants , 18 were not present in 6TG-resistant colonies and hence do not appear to affect MMR activity . Mutations R510G , A586P , G683D , F703S , L1060R , E1191K , T1217D and T1217I were identified in 6TG-resistant colonies by sequence analysis ( Fig 3A and 3B; S2B Fig ) . Of note , variants R510G and F703S were detected in only two colonies out of five and four , respectively , that had not resulted from LOH ( Fig 3B ) . Given the low frequency of LMO-mediated base-pair substitution , we consider the presence of a variant allele in two independent colonies indicative for pathogenicity . The MMR abrogating effect of all Msh6 variants conferring 6TG-resistance was further characterized by Western blot analyses , MSI assays and methylation-damage-induced mutagenesis assays . The effect of the identified MMR abrogating mutations on MSH6 and MSH2 protein levels was evaluated by Western blot analyses ( Fig 4 ) . MSH6 and MSH2 form a heterodimer; consequently , a drop in MSH6 levels is often associated with a slight decrease in MSH2 protein stability . Protein levels were quantified with respect to Msh6+/- mESCs , which maintain a functional MMR system with about two-third of the MSH6 level observed in Msh6+/+ mESCs [25] . Known pathogenic mutations V397E , L448P , G1137S and R1332Q reduced MSH6 levels to 7–33% of that seen in Msh6+/- mESCs . The R1332Q mutation is located in the splice donor site of exon 9 which may explain the appearance of a shorter protein . The drop in MSH6 levels seen for the known pathogenic mutations was mirrored by variants A586P , G683D , F703S and L1060R that reduced protein levels to 7–24% . Variants R510G , E1191K , T1217D and T1217I maintained relatively high MSH6 levels of 59–79% . MSI in MSH6 mutation carriers is largely restricted to mononucleotide markers [41] . To investigate the effect of the detected Msh6 variants on MSI we used a ( G ) 10-neo slippage reporter . The neomycin resistance gene ( neo ) in this reporter is rendered out of frame by a preceding ( G ) 10 repeat . When DNA polymerase slippage errors at the ( G ) 10 repeat such as the deletion of one G or insertion of two Gs remain unnoticed , the neo becomes in frame and generates Geneticin-resistant cells . Hence the number of Geneticin-resistant colonies is indicative of the frequency of neo-restoring slippage events and the MMR capacity of the cells [42] . The slippage rates , i . e . , the chance of a slippage event occurring during one cell division , in 6TG-resisant Msh6 VUS expressing mESCs ranged from 5 . 3x10-5 to 5 . 1x10-4; which is around the average rate of 1 . 9x10-4 observed for the known pathogenic mutations and 140 to 1340-fold higher than the slippage rate of 3 . 8x10-7 seen for Msh6+/- MMR-proficient mESCs ( Fig 5 ) . In addition to increased spontaneous mutagenesis events , MMR-deficient cells also experience increased methylation-damage-induced mutagenesis [43] . To study the influence of the detected MMR attenuating Msh6 variants on methylation-damage-induced mutagenesis , mESCs were exposed to the methylating DNA damaging agent N-methyl-N’-nitro-N-nitrosoguanidine ( MNNG ) and the number of cells that consequently attained mutations was quantified . In MMR-proficient cells , DNA replication across MNNG-induced O6-methylguanine lesions is impaired by futile cycles of MMR , ultimately leading to cell death and suppression of methylation-damage-induced mutagenesis . Under MMR-deficient conditions , however , the MNNG-induced mismatches are not recognized and remain in the genome leading to the accumulation of mutations . To provide a quick read out for the frequency of mutation accumulation , we measured the number of MNNG-exposed cells that became resistant to a high dose of 6TG for an extended period . Solely cells that carry an inactivating mutation in Hprt survive stringent 6TG treatment because HPRT is required for 6TG to behave as a DNA damaging agent . All detected Msh6 variant cell lines showed an elevated MNNG-induced mutator phenotype when compared to the MMR-proficient Msh6+/- mESCs ( Fig 6 ) . According to literature MSH6-G566R may be pathogenic [29 , 44] , yet our screen did not identify this variant in 6TG-resistant colonies . Hence we investigated whether the MMR abrogating effect of Msh6-G565R could have been missed by the screen due to technical difficulties . Rather than applying 6TG selection after oligonucleotide-directed mutagenesis , we purified Msh6G565R/- mESCs using a Q-PCR-based protocol [25] ( S2C Fig ) and subsequently examined their MMR capacity . Exposure of Msh6G565R/- cells to increasing doses of 6TG revealed that they were equally sensitive to 6TG as Msh6+/- cells ( Fig 7A ) . In the MSI assay , Msh6G565R/- mESCs did not experience significantly more slippage events than the MMR-proficient control ( Fig 7B ) . Thus , Msh6-G565R did not attenuate MMR consistent with the oligonucleotide-directed mutagenesis screening result .
The results of our study demonstrate the oligonucleotide-directed mutagenesis screen we previously described for the characterization of MSH2 VUS [22] can be extended to MSH6 VUS . Combining oligo targeting in Msh6+/- mESCs with 6TG selection and sequence analysis allows pathogenic MSH6 variants to be distinguished from polymorphisms . The efficacy of the genetic screen was established in a proof of principle study with 4 known pathogenic MSH6 mutations and 5 polymorphisms . This number was low because of the paucity of MSH6 variants that were classified with 100% certainty . Not one of the 5 non-pathogenic variants was identified as MMR abrogating . Also , among the 26 MSH6 VUS we subsequently analyzed , not one of the 4 variants classified as likely not pathogenic was identified as pathogenic by our screen . Finally , functional assays established that one of the VUS that was not detected as pathogenic by the screen indeed did not influence MMR activity ( G565R ) . Hence the false positive rate of the screen , i . e . , the chance the screen identified a VUS as MMR abrogating while it was a priori or a posteriori identified as ( likely ) non-pathogenic was <1/10 , giving a specificity >90 . 0% . The sensitivity of the genetic screen is a measure of the false negative rate; it is the likelihood that a pathogenic mutation is not detected . All 6 InSiGHT classified pathogenic and likely pathogenic variants as well as the previously proven pathogenic G1139S mutation were recognized as MMR abrogating by the screen , translating to a sensitivity of >85 . 7% . We used the oligonucleotide-directed mutagenesis screen to investigate the MMR capacity of 26 MSH6 VUS . Eight of these were found in suspected-LS patients from two medical centers in the Netherlands . From this clinical cohort , the mouse equivalents of mutations R511G , A587P and F706S were detected by our screen and shown to abrogate MMR . However , R510G and F703S were detected in only 2/5 and 2/4 6TG-resistant colonies , respectively , that had retained two Msh6 alleles , while the other pathogenic variants were present in virtually all colonies diploid for Msh6 ( Figs 2B , 3A and 3B ) . The poorer recovery of R510G and F7103S mutants may have been due to a lower success rate of LMO-mediated base-pair substitution . The pathogenic phenotype observed for these three variants is in line with clinical data: all three variants were detected in patients with MSI-H LS-related tumors and with a family history of LS-related tumors . In the case of VUS A587P and F706S , relatives with LS-related tumors carried the same mutation . IHC also demonstrated MSH6 was absent in the patients encoding MSH6-A587P and MSH6-F706S; the IHC data for MSH6-R511G were inconclusive . The other 5 variants in the clinical cohort , A25S , E221D , G670R , R922Q and c . 3438+6T>C , were not identified as MMR abrogating . VUS E221D , G670R and R922Q were found in patients who also harbored a second , known pathogenic mutation in one of the DNA MMR genes that was likely causative for the LS phenotype . E221D was also detected in a second patient who was 83 years old and did not have a family history suspicious for LS . MSH6-A25S was found in a typical LS tumor , i . e . , a colon tumor showing MSI , loss of heterozygosity of MSH6 , and loss of MSH6 protein expression . The patient however only had one relative with a colorectal tumor and this tumor was not MSI-high and stained positive for all MMR proteins . A previous in vitro study also suggested MSH6-A25S is not pathogenic [45]; it could be that the tumor arose due to a missed somatic mutation . VUS c . 3438+6T>C was found in a patient with a family history suspicious of LS . We however do not know if the relatives with LS-associated cancers also carried this specific MSH6 sequence variant . IHC failed in the index patient carrying the c . 3438+6T>C variant , therefore we cannot exclude that a somatic mutation or MLH1 hypermethylation caused the MSI in the tumor . Tumor tissue of one family member was tested and showed no MSI and normal IHC . It is also possible that the genetic screen was unable to identify c . 3438+6T>C as pathogenic due to differences between the human and mouse MSH6 sequences . While the MSH6 coding sequence is highly conserved , intron sequences are more variable between species ( S4 Fig shows human and mouse sequence around c . 3438+6 ) . Hence there is a chance that variant c . 3438+6T>C affects splicing in man but not in mice . According to several splice site prediction programs ( NNSPLICE , GeneSplicer , Human Splicing Finder ) , however , c . 3438+6T>C does not affect splicing . The other 18 MSH6 VUS we studied were attained from literature and the InSiGHT database . The genetic screen found 5 of these variants abrogate MMR: G686D , L1063R , E1193K , T1219D and T1219I . The detection of G686D and L1063R is in line with their InSiGHT classification , which describes the mutations as likely pathogenic . Variant E1193K has previously been suggested to cause LS in studies that identified the mutation in patients with ECs that were MSI and did not stain for MSH6 [27 , 28] . Not much clinical data is available for VUS T1219D but Msh6T1217D mice were demonstrated to have increased cancer susceptibility [46] . VUS T1219I has been described in a CRC patient who had a family history of CRC and a MSI tumor that stained positive for MSH6 , the latter being consistent with the high levels of this variant protein we observed in mESCs . Both clinical and in vitro data indicate MSH6-T1219I abrogates MMR activity [37 , 45] . MSH6 VUS R128L , R468H , V509A , Y556F , P623A , S666P , E983Q , R1095C , T1255M and R1304K were not identified as pathogenic in our screen . These sequence variants were classified as likely not pathogenic by InSiGHT , identified in patients with MLH1 promoter methylation or with MSS and MSH6 positive tumors , or observed in patients for whom little clinical data was available . VUS S285I , G566R and T1142M were also not detected as MMR attenuating by our screen , yet they seem suspicious for pathogenicity based on available data . MSH6-T1142M was previously suggested to be probably pathogenic based on clinical data describing the variant in a 27 year old patient with polyps who met the Bethesda guidelines , had a 61 year old mother with polyps , and did not carry pathogenic mutations in any other MMR gene nor showed MLH1 promoter methylation in the tumor [36] . VUS S285I and G566R were detected in CRC patients with MSI ( low and high , respectively ) tumors that had loss of heterozygosity of MSH6 [29] . Cyr and Heinen [44] investigated the effect of these two mutations on mismatch binding and processing: variant S285I was not found to have a specific MMR attenuating effect but variant G566R was suggested to abrogate MMR by interfering with the ATP-dependent conformational change that must take place to activate downstream repair pathways upon mismatch binding . We therefore purified Msh6G565R/- mESCs and assessed their MMR capacity . The Msh6G565R/- cells behaved like MMR-proficient Msh6+/- mESCs , confirming the result of the oligonucleotide-directed mutagenesis screen . Despite the good performance of our screen and the high amino acid conservation of MSH6 , we cannot exclude Msh6-G565R was not identified as pathogenic due to differences between mice and men . To fully dissuade this argument we will need to develop the oligonucleotide-directed mutagenesis screen in human cells . The oligonucleotide-directed mutagenesis screen presented here is a relatively simple tool that can be used to investigate the pathogenic phenotype of many MSH6 VUS in parallel . While the evolutionary conservation of MMR justifies the use of mouse cells for the majority of VUS , testing of splice-site and intronic mutations necessitates adaptation to human cells . Also , as long as uncertainty exists about its specificity and sensitivity , functional testing needs to be combined with clinical data and in silico estimations to arrive at a reliable classification of VUS . Conforming the updated American College of Medical Genetics and Genomics ( ACMG ) standards and guidelines for sequence variant interpretation , we are currently transferring our functional tests to certified Clinical Genetics laboratories and creating an infrastructure where test results are compared and interpreted taking into account all available data . In this way , LS mutation carriers can be identified with the highest certainty and enrolled in tailored surveillance programs while relatives without the mutation can be excluded from surveillance .
The genetic screen was developed in Msh6+/- mESCs , which contain one active Msh6 allele ( Msh6+ ) and one Msh6 allele that was disrupted by the insertion of a puromycin resistance marker ( Msh6- ) [24] . The MSH6 variants under investigation were introduced into the Msh6+/- mESCs by oligo targeting using LMOs [23] . 7x105 Msh6+/- mESCs were seeded in BRL-conditioned medium on gelatin-coated 6 wells and exposed to a mixture of 7 . 5 μl TransIT-siQuest transfection agent ( Mirus ) , 3 μg LMOs and 250 μl serum-free medium the following day . After 3 days , 1 . 5x106 LMO-exposed cells were transferred to gelatin-coated 10 cm plates and subjected to 6TG ( 250 nM ) ( Sigma-Aldrich ) selection . After 10 days the 18 largest 6TG-resistant colonies were picked . Cells that became 6TG-resistant due to loss of heterozygosity events were excluded from further analyses using a PCR specialized to detect the presence of both the disrupted and non-disrupted Msh6 alleles [24] . 6TG-resistant mESCs that maintained both Msh6 alleles were sequenced to confirm the presence of the planned mutation . Western blot analyses were performed as described in Wielders et al . [25] . Rabbit polyclonal antibodies against mMSH2 ( 1:500 ) [47] and mMSH6 ( 1:500 ) [24] as well as mouse polyclonal antibody against γ-Tubulin ( 1:1000; GTU-88 Sigma-Aldrich ) were used as primary antibodies . Protein bands were visualized using IRDye 800CW goat anti-rabbit IgG and IRDye 800CW goat anti-mouse IgG secondary antibodies ( Li-cor ) and the Odyssey scan . The infrared fluorescent signals measured by the Odyssey scan are directly proportional to the amount of antigen on the Western blots , allowing quantification of the protein bands . mESCs were electroporated with the ( G ) 10-neo Rosa26 targeting vector as described in Dekker et al . [48] . The ( G ) 10-neo Rosa26 targeting vector is composed of a promoterless histidinol resistance gene as well as a neomycin resistance gene ( neo ) that is rendered out of frame by a preceding ( G ) 10-repeat [42] . Once electroporated , 106 cells were seeded on gelatin-coated 10 cm plates in BRL-conditioned medium and exposed to Histidinol ( 3mM ) ( Sigma-Aldrich ) . Successful integration of the vector into the Rosa26 locus of the Histidinol-resistant colonies routinely occurs at a frequency of ±95% and was confirmed by Southern blot analyses . The individual successfully targeted colonies were subsequently expanded to 107 cells and transferred to gelatin-coated 10 cm plates at a density of 105 cells per plate for Geneticin selection ( 600 μg/ml ) ( Life Technologies ) . After 10 days , the number of Geneticin-resistant colonies was counted and the slippage rate of the variant mESCs calculated using the formula: 0 . 6 x Geneticintotal = N x p x log ( N x p ) , where Geneticintotal is the number of Geneticin-resistant colonies , N the number of cells to which the culture was expanded , and p the number of mutations per cell division . Experiments were performed in quadruplicate and statistical differences calculated using a one-tailed , unpaired t-test with Welch’s correction . The MNNG-induced mutagenesis assay was performed as described in Claij and te Riele [43] . 2 . 5x106 variant mESCs were seeded on an irradiated mouse embryonic fibroblasts feeder layer in 10 cm plates and exposed to 0 or 4μM MNNG ( Sigma-Aldrich ) for 1h the following day . 40 μM O6-benzylguanine was present in the medium from 1h prior to the MNNG treatment until 6 days after , at which point 1 . 5x106 cells were transferred to gelatin-coated 160 cm2 plates for 6TG selection ( 10 μg/ml ) . After two weeks of 6TG selection , the number of resistant colonies and hence the frequency of MNNG-induced Hprt mutants could be determined . Experiments were performed in duplo and the statistical difference between MNNG-treated Msh6+/- mESCs and MNNG-treated variant cell lines calculated using a one-tailed , unpaired t-test with Welch’s correction . Msh6G565R/- mESCs were made as described by Wielders et al . [25] . Variant G565R was introduced into Msh6+/- mESCs by oligo targeting and a pure Msh6G565R/- mESC clone was obtained by consecutive rounds of seeding and mutation specific PCR: oligonucleotide-exposed cells were expanded and subsequently seeded on a 96-well plate at a density of 5000 cells per well . A mutation-specific quantitative PCR was used to identify wells that contained Msh6G565R/- mESCs . Positive wells were reseeded at lower density and positive wells again identified by Q-PCR . A pure clone was finally obtained by seeding single cells per well . Sequence analysis confirmed the creation of Msh6G565R/- mESCs . The 6TG sensitivity of Msh6G565R/- mESCs was investigated by exposing the variant cell line to increasing doses of 6TG , as described in Wielders et al . [49] . MMR-deficient Msh6-/- and MMR-proficient Msh6+/- and Msh6+/+ mESCs were taken along for comparison . We investigated the pathogenic phenotype of MSH6 VUS that were found in suspected-LS patients at the Clinical Genetics departments of the Erasmus Medical Center Rotterdam and Radboud University Medical Center Nijmegen . We collected tumor characteristics , age at diagnosis , results of molecular diagnostics and germline mutation analysis , and family history from medical records . MSI analysis was performed with the Bethesda panel [50] or with the Promega pentaplex MSI analysis [51] . IHC for MLH1 , MSH2 , MSH6 and PMS2 protein was performed as described previously [52] . Germline mutation analysis of MSH6 was performed by sequencing and multiplex ligation dependent probe amplification . The in silico prediction model PolyPhen [53] was used to estimate the chance of a variant being deleterious . | The colorectal and endometrial cancer predisposition Lynch syndrome ( LS ) is caused by an inherited heterozygous defect in one of four DNA mismatch repair ( MMR ) genes . Deleterious mutations ( e . g . , protein-deleting or -truncating ) in DNA MMR genes unambiguously allow for the clinical diagnosis LS and hence enable appropriate surveillance measures to be taken to reduce cancer risk and ensure early detection of tumors . However , currently about one-third of detected MMR gene variants are subtle with less clear functional consequences: missense mutations affecting a single amino acid may be innocuous , hence not causing LS , or partially or fully destroy protein function . As long as uncertainty exists about their pathogenicity , such mutations are labeled ‘variants of uncertain ( clinical ) significance’ ( VUS ) . VUS hamper genetic counseling and therefore the need for functional testing of VUS is widely recognized . To functionally annotate MMR gene VUS , we have developed a high content cellular assay in which the VUS is introduced in a cell culture by oligonucleotide-directed gene modification . Should the VUS be deleterious for MMR , the modified cells survive exposure to the guanine analog 6-thioguanine ( 6TG ) and 6TG-resistant colonies appear . Should the mutation not affect MMR , no colonies appear . Here we present the adaptation and application of this protocol to the functional annotation of variants of the MMR gene MSH6 . Implementation of our assay in clinical genetics laboratories will provide clinicians with information for proper counseling of mutation carriers and treatment of their of tumors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"oligonucleotide-directed",
"mutagenesis",
"alleles",
"dna",
"damage",
"mutation",
"dna",
"molecular",
"biology",
"techniques",
"mutagenesis",
"and",
"gene",
"deletion",
"techniques",
"frameshift",
"mutation",
"research",
"and",
"analysis",
"methods",
"molecular",
"biolog... | 2017 | Suspected Lynch syndrome associated MSH6 variants: A functional assay to determine their pathogenicity |
Despite the existence of safe and effective vaccines , rabies disease still causes an estimated 59 , 000 human deaths a year in the endemic areas in Asia and Africa . These numbers reflect severe drawbacks regarding the implementation of PrEP and PEP in endemic settings , such as lack of political will and low priority given to rabies . Since these contextual factors have proven to be persistent , there is an urgency to improve current strategies or develop novel approaches in order to control rabies disease in the future . This study aimed to identify and systematically prioritize the research needs , through interviews and questionnaires with key-opinion-leaders ( KOLs ) . A total of 46 research needs were identified and prioritized . The top research needs are considered very high priority based on both importance for rabies control and need for improvement . KOLs agree that animal rabies control remains most important for rabies control , while research on human host , agent ( rabies virus ) and the environment should be prioritized in terms of need for improvement . A wide variety in perceptions is observed between and within the disciplines of virology , public health and veterinary health and between KOLs with more versus those with less experience in the field . The results of this study give well-defined , prioritized issues that stress the drawbacks that are experienced by KOLs in daily practice . The most important research domains are: 1 ) cheap and scalable production system for RIG 2 ) efficacy of dog mass vaccination programs and 3 ) cheap human vaccines . Addressing these research needs should exist next to and may reinforce current awareness and mass vaccination campaigns . The differences in perspectives between actors revealed in this study are informative for effective execution of the One Health research agenda .
Rabies is a neglected tropical disease causing an estimated 59 , 000 human deaths a year [1] . Human rabies is 100% preventable by either pre-exposure prophylaxis ( PrEP ) or post-exposure prophylaxis ( PEP ) which together effectively prevent approximately 372 , 000 deaths yearly [2] . In resource-poor settings , however , these prophylaxes are frequently not accessible , incomplete or delayed and consequently , almost 96% of all human cases occur in Africa and Asia despite the fact that rabies virus circulates worldwide [3] . Without treatment options and effective animal rabies control , human rabies will continue to be a social and economic burden . An important and cost-effective strategy in the control of human rabies is the prevention of infection . Transmission occurs via saliva of animal reservoirs and dogs are the major ( 90% ) source of infection to humans [4] . Mass vaccination of domestic dogs has resulted in effective control of both canine rabies and human rabies when a coverage ratio of 70% is achieved and maintained [5–7] . Due to the size and rapid turnover of dog populations [8] this requires long-term determination which poses a challenge for most of the developing world , due to e . g . a lack of resources , diagnostic capacity and in-country expertise [4 , 9 , 10] . Addressing these contextual factors could enable rabies elimination , as was achieved in Latin American and Caribbean countries [5] , but is hard to achieve given the low priority given to rabies [4] . In settings with little political commitment , prevention and clinical interventions are more feasible strategies to improve health [11] . For rabies , this can be translated in an urgency to improve current control tools and develop novel strategies for rabies control . However , a focus and direction in research and development in the field of rabies is lacking as literature publicizes different priorities [12–14] . Considering the limited resources available for NTDs and rabies in specific , such direction could accelerate the control of rabies in the future . Therefore , the aim of this research was to assess and prioritize the research needs that could improve current strategies or lead to the development of novel strategies to control rabies disease .
For this study , KOLs were defined as individuals with extensive knowledge in the field of virology , public health and/or veterinary ( public ) health in the context of rabies . KOLs were identified via a web search on representatives of rabies initiatives and ( keynote ) speakers at international conferences . Additionally , snowball sampling was employed , which allowed the researchers to approach a global network of rabies experts . To ensure a high level of expertise , rabies experts with at least an MSc degree or sufficient rabies-related work experience ( >5 years ) were approached to participate in this study . To ensure data richness , KOLs representing different fields of expertise , contexts of expertise and with different professions were selected for participation in the interviews and questionnaires . As rabies disease is a global problem , international KOLs were selected with a special focus on endemic settings in Africa and Asia . This included rabies experts working for non-profit seeking knowledge institutes , for-profit ( pharmaceutical ) companies , doctors ( MD , DVM ) , non-governmental organizations ( GARC ) , and regulatory and public health authorities ( FAO , WHO , OIE ) . To obtain a comprehensive overview of the research required to enable rabies control , the epidemiological triangle was used [20 , 21] . This framework covers the components important for disease transmission and consequently the targets for infectious disease control: the agent ( rabies virus ) , human and animal hosts and the environment [21] . Relevant research needs are defined as research that contributes to assaulting virulence of the agent ( e . g . antivirals and passive immunization ) , raising susceptibility of the host ( active immunization ) and/or diminishing the favorability of the ( sociocultural and physical ) environment . The multi-staged prioritization process started with the identification of research needs through interviews with KOLs . KOLs were prompted with semi-structured interview questions based on the components of the epidemiological triangle . The interviews pursued questions about the strengths and weaknesses of current strategies , followed by the question what research is needed to advance rabies control in order to prime the respondents to research on both novel strategies and improvements to current strategies . Probing was based on the concepts of the epidemiological triangle . Interviews were conducted by two researchers via phone or Skype and lasted , on average , 30 minutes . Interview invitations were sent until data saturation ( no new research needs mentioned in four subsequent interviews ) was reached in the interviews . Data from the interviews was analyzed through thematic coding by two independent researchers , leading to inductively derived research needs [22] . Subsequent discussion led to agreement regarding the final coding of the research needs , thereby making the formulations as clear , complete and concise as possible . The identified research needs formed the basis for the questionnaire . Research needs that were mentioned by only one KOL were not considered to have priority and were therefore excluded from the questionnaire . The anonymous questionnaire consisted of two parts: ranking of individual research needs and ranking of the components of the epidemiological triangle . During the ranking exercise , participants were asked to apply two criteria to encourage active and balanced prioritization [18]: importance for rabies control and need for improvement . This enabled respondents to distinguish clearly between those aspects that are already in place and research that may have a large impact on the unmet need [16 , 17] . The final questionnaire consisted of 30 questions . The questionnaire was pilot tested and distributed through the online web survey program SurveyMonkey . The questionnaire was distributed among 172 KOLs selected through a web search based on abovementioned criteria . This selection included the interview participants . Additionally , KOLs were encouraged to send the questionnaire to colleague rabies experts . For all respondents demographics were checked for compliance with the inclusion criteria . A reminder was sent after 7 and 11 days to increase the response-rate . A copy of the questionnaire is deposited with DANS ( see Data Availability Statement ) .
A total of 28 KOLs participated in the interviews , after which data saturation was reached ( Fig 1 ) . Other invitees either did not respond ( n = 27 ) , did not consider themselves experts based on our definition ( n = 8 ) , had no time to participate within the indicated time frame ( n = 7 ) or perceived a conflict of interest ( n = 3 ) . A total of 126 ( 73% of initially invited ) participants filled out the questionnaire , of which one response was excluded because the inclusion criteria for KOLs were not met . The age distribution of the KOLs ( n = 125 ) was: 55-up ( 44% ) , 40–55 ( 42% ) and 25–40 ( 14% ) . The highest obtained academic rank of KOLs are a PhD ( 39% ) , followed by professor ( 30% ) , Master of Science ( 27% ) , and Bachelor of Science ( 3% ) . The majority of the KOLs work as scientists ( 58% ) , but may also work as policy makers ( 25% ) , industry ( 13% ) and medical professionals ( 30% ) . The distribution of rabies related expertise is shown in Table 1 . Based on these demographic profiles , we consider our KOL sample to have significant expertise in rabies research and representative for the different disciplines . A total of 59 research needs emerged from the interviews , of which 46 were mentioned by more than one KOL . The research needs related to all components of the epidemiological triangle: 10 research needs addressed the animal host , 12 the human host , 13 the rabies virus and 11 the environment . A total of 125 KOLs assessed the research needs of their expertise: animal host ( n = 96 ) ; human host ( n = 88 ) ; agent ( n = 83 ) and; environment ( n = 78 ) . For all four components , several research needs were assigned to the high priority group . Fig 2 illustrates the average priority ( moderate-high-very high ) KOLs attributed to the need of improvement of each research need , divided over the different components . For all components , high priority research needs were identified . Very high priority was assigned to research needs linked to the human host , agent and environment . None of the identified research needs was assigned to the very low or low priority group . In addition , respondents were asked to rank the identified research needs on the importance of that aspect for the control of rabies disease . Based on the combined prioritization of importance and need for improvement ( Fig 3 ) , the research needs that should be given very high priority in rabies research , according to KOLs , are 1 ) developing a cheap alternative for rabies immunoglobulins ( RIG ) , followed by 2 ) developing an alternative for RIG that is easy to produce , 3 ) increasing knowledge on factors that hamper the efficacy of dog mass vaccination programs and 4 ) developing a cheap alternative for the human vaccine . A comparison of the results reveals modest differences between importance and need for improvement and 11 research needs with significant differences ( p<0 . 05 ) between the scores of the two criteria ( S1 Table ) . Only 3 out of these research needs could be assigned to different priority groups for importance compared to need for improvement ( Fig 3 ) and they were all given a higher need for improvement than importance for rabies control: the development of treatment options that can clear the virus from the CNS ( very high versus high ) , the development of immuno-contraceptives for dogs ( high versus moderate ) , and the development of treatment for animals ( moderate versus low ) . The other research needs were assigned to the same priority groups for need for improvement and importance . Table 2 shows the ranking of the four components , through distribution of 100 points , with the animal host having the highest mean for importance ( 37 . 2 ) , followed by agent ( 22 . 1 ) , human ( 22 . 0 ) and environment ( 19 . 5 ) ( p<0 . 001 ) . The component prioritization shows a high standard deviation , indicating a wide variety in the points allocated by the KOLs . The number of points allocated to agent and environment ranged from 0 to 70 , for human from 0 to 75 , and for animal host from 0 to 100 points . This illustrates that the KOLs have divergent views on the importance and need for improvement of the components for rabies control .
This paper provides a unique new dataset in canvassing and prioritizing research needs in rabies going further than mere control of the animal reservoir . KOLs assigned high or very high priority on need for improvement to a total of 26 research needs , and their urgency is amplified by the finding that these research needs are equal to the research needs that were considered important for rabies control ( Fig 3 ) . Research on the animal host is considered most important for rabies control , but top priorities reflect the invariable demand for improved preventive and therapeutic strategies for human application to decrease the burden of rabies disease on the short term . Taking into account the limited resources available , research efforts should focus on the research needs that are prioritized as highly important by KOLs , which could be reduced to the following domains: 1 ) cheap and scalable production systems for RIG; 2 ) efficacy of dog mass vaccination programs and; 3 ) cheap human vaccines . Importantly , literature shows that these unmet needs are addressable . Recent developments regarding production systems for cheap and scalable alternatives for RIG include monoclonal antibodies [23] and nanobodies [24] . The identification of barriers hampering the efficacy of current mass vaccination programs would require such programs to include a qualitative causal component . The few studies that report on barriers to low vaccination coverage show that the collection of these data can lead to the formulation of program-specific strategies to increase vaccination coverage [25–27] . Lastly , novel vaccines using adjuvants have shown encouraging clinical outcomes [28] . The costs of adjuvanted vaccines may exceed the costs of existing vaccines , however , the improved immunogenicity may reduce the total costs of PEP via reduction of doses and number of hospital visits . The characteristics of above-mentioned products , such as costs , scalability and regimen , would improve the accessibility of products in endemic countries of Asia and Africa and , hence , significantly decrease the burden of rabies disease [29] . It can be argued that addressing the research needs presented here could align stakeholders towards effective use and implementation of rabies control programs . Individuals at risk will be more likely to translate awareness in demand [29] , willingness to pay [30] and compliance [31] , if PrEP and PEP could be obtained and used more easily . The introduction of improved and novel products may , thus , increase the impact of education and awareness programs . Likewise , accurate data on the societal and economic burden of disease could increase political will and advocate funding for rabies control , which is currently suffering from low priority and a lack of resources [32] . Tackling the research needs presented here could , thus , introduce a mutually reinforcing cycle and accelerate the control of rabies disease . Besides sketching the contents of the rabies research agenda , the current study highlights differences between KOLs that should be used to inform its implementation . The control of rabies , like other zoonoses with a serious socioeconomic impact , could benefit from an One Health approach in which human , animal and environmental health are integrated [33] . This could be hampered by the observed variety in perception on the importance of different epidemiological components ( Table 2 ) . The variety between KOLs of different disciplines is not surprising considering the general preference of animal rabies control measures , which seems to have caused a protective tendency by the public health sector . The significant importance that more experienced KOLs ( >10 years of experience ) give to the agent indicates that the limited successes in the past led to disbelief in possible solutions without a better understanding of the pathogenesis of rabies virus [34] . This finding not only argues for the reintroduction of basic research on research agendas , which has been diminished by ‘impact assessments’ [35] , it also informs funders to shift their expectations for the development of novel medications against rabies from short-term to medium-long term . Overcoming these innovation barriers , that can stop research ( funding ) in a premature phase [34] , should coincide with the implementation of research priorities to optimize societal impact . Despite the differences in KOL’s perceptions , KOLs seem to acknowledge the importance of an One Health approach . It was observed that research needs for all the components were given high priority ( Fig 2 ) and that all components received a significant number of points ( Table 2 ) . Taken together , stakeholders are encouraged to converge their activities relatively to the One Health research agenda presented here , which serves as a uniting tool and may provide the necessary focus to achieve global rabies elimination . | Rabies is a 100% vaccine-preventable disease but invariably fatal once symptoms occur . Annually , tens of thousands of people die after being infected with rabies virus , predominantly through bites or scratches of infected dogs . The stable mortality rates highlight the limitations of current disease specific interventions , including prophylaxes , awareness campaigns and mass vaccination of dogs . Consequently , research is needed to develop improved and novel strategies that circumvent the barriers faced in implementation in endemic settings . Interest for rabies , however , is limited and to effectively allocate budgets the field would benefit from a more focused research agenda . This study prioritized research topics based on the importance for rabies control and the need for improvement . According to experts , research should focus on 1 ) cheap and scalable production systems for RIG; 2 ) efficacy of dog mass vaccination programs and; 3 ) development of a cheap human vaccine . By elucidating differences in perceptions of stakeholders between disciplines and between those with more versus those with less experience in the field , this study also provides practical insights to inform stakeholders concerned with the implementation of interdisciplinary collaboration in the field of rabies . The prioritization of rabies-specific research needs is a vital step in accelerating innovation necessary to decrease the burden of disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"animals",
"mammals",
"dogs",
"viruses",
"preventive",
"medicine",
"vaccines",
"rabies",
"rna",
"viru... | 2018 | A research agenda to reinforce rabies control: A qualitative and quantitative prioritization |
Plant virus movement proteins ( MPs ) localize to plasmodesmata ( PD ) to facilitate virus cell-to-cell movement . Numerous studies have suggested that MPs use a pathway either through the ER or through the plasma membrane ( PM ) . Furthermore , recent studies reported that ER-PM contact sites and PM microdomains , which are subdomains found in the ER and PM , are involved in virus cell-to-cell movement . However , functional relationship of these subdomains in MP traffic to PD has not been described previously . We demonstrate here the intracellular trafficking of fig mosaic virus MP ( MPFMV ) using live cell imaging , focusing on its ER-directing signal peptide ( SPFMV ) . Transiently expressed MPFMV was distributed predominantly in PD and patchy microdomains of the PM . Investigation of ER translocation efficiency revealed that SPFMV has quite low efficiency compared with SPs of well-characterized plant proteins , calreticulin and CLAVATA3 . An MPFMV mutant lacking SPFMV localized exclusively to the PM microdomains , whereas SP chimeras , in which the SP of MPFMV was replaced by an SP of calreticulin or CLAVATA3 , localized exclusively to the nodes of the ER , which was labeled with Arabidopsis synaptotagmin 1 , a major component of ER-PM contact sites . From these results , we speculated that the low translocation efficiency of SPFMV contributes to the generation of ER-translocated and the microdomain-localized populations , both of which are necessary for PD localization . Consistent with this hypothesis , SP-deficient MPFMV became localized to PD when co-expressed with an SP chimera . Here we propose a new model for the intracellular trafficking of a viral MP . A substantial portion of MPFMV that fails to be translocated is transferred to the microdomains , whereas the remainder of MPFMV that is successfully translocated into the ER subsequently localizes to ER-PM contact sites and plays an important role in the entry of the microdomain-localized MPFMV into PD .
Plasmodesmata ( PD ) , channels providing symplastic continuity of the ER and the plasma membrane ( PM ) between adjacent cells , play vital roles in intercellular communication in plants [1] . The ER and PM passing through PD are highly specialized to regulate PD permeability [2 , 3] . Plant viruses must pass through PD to establish systemic infection . To modify PD function and facilitate the cell-to-cell movement , viruses have PD-targeting proteins , the so-called movement proteins ( MPs ) [4] . Hence , understanding how MPs reach PD will provide insight into the mechanism underlying virus cell-to-cell movement . MPs have been frequently proposed to use a membrane trafficking pathway either through the ER or through the PM to reach PD . Several viruses possess MPs that reportedly use endomembrane trafficking through the ER . These MPs are apparently associated with the ER and PD in infected cells [5–8] or in cells transiently expressing only MPs [9] , even though the detailed mechanism by which these MPs traffic from the ER to PD is unclear . On the one hand , other MPs such as those of cauliflower mosaic virus and cowpea mosaic virus localize to the PM [10 , 11] . Inhibition by brefeldin A ( BFA ) , an inhibitor of COPII transport , showed that the secretory pathway is not involved in the PM localization [10 , 11] , but how these MPs traffic to the PM is still unknown . One recent study has proposed that cauliflower mosaic virus MP is transported from the PM to PD through the endocytic pathway [12] . Recent studies have shown that two membrane subdomains in the ER and PM are also involved in virus cell-to-cell movement . ER-PM contact sites , membrane subdomains connecting between the cortical ER and the PM , are known to play roles in intracellular Ca2+ homeostasis and signaling in mammalian cells [13] . Arabidopsis synaptotagmin 1 ( SYTA ) , a key component in connecting the cortical ER and the PM in plant cells [14 , 15] , substantially localizes to the nodes of the cortical ER [15] . SYTA interacts with several virus MPs , and knockout or dominant-negative inhibition of SYTA delays cell-to-cell movement of several viruses [16 , 17] . These facts indicate that the function of ER-PM contact sites is important for virus cell-to-cell movement . PM microdomains , small regions which have compositions and functions distinct from the surrounding PM , are another type of membrane subdomains that are involved in virus cell-to-cell movement . PM microdomains have been proposed to have lipid compositions that differ from the surrounding PM and to be detergent insoluble , although this is still a matter of debate [18] . The number and the biological roles of microdomains are largely unknown in plant cells , but certain proteins show patchy distribution in the PM and are recognized as microdomain-associated proteins [19] . One of the microdomain-associated proteins , remorin ( REM1 . 3 ) , suppresses cell-to-cell movement of potato virus X ( PVX ) [20] . Furthermore , REM1 . 3 localizes also to PD and interacts with triple gene block protein 1 ( TGBp1 ) , an MP of PVX . Thus , it has been suggested that PM microdomains as well as ER-PM contact sites are important for virus cell-to-cell movement . Considering that PD localization of virus MPs is necessary for facilitating virus cell-to-cell movement , these two membrane subdomains , microdomains and ER-PM contact sites , are speculated to be involved in MP traffic to PD [16 , 17 , 20] . However , there is no direct evidence that virus MPs use these subdomains to reach PD , and a functional relationship of these subdomains in MP traffic to PD is unclear . Signal peptides ( SPs ) are short sequences comprising approximately 7–30 aa that are frequently found in the N terminus of a diverse array of proteins . In general , proteins in eukaryotic cells with an SP are co-translationally recruited to the ER , and penetrate the ER membrane or are released into the ER lumen concomitant with the SP cleavage [21 , 22] . Some of these proteins play specific roles in the ER , whereas others are further transported to other organelles or secreted into the extracellular space . Thus , SPs are essential for proper localization and membrane targeting of proteins . Fig mosaic virus ( FMV ) is a negative-strand RNA virus in the genus Emaravirus . We showed previously that the MP of FMV ( MPFMV ) localized to the PM in addition to PD , and remarkably , MPFMV was predicted to possess an N-terminal SP [23] . To the best of our knowledge , no viruses , other than the members of the genus Emaravirus , have an MP possessing an SP . In this study , we analyzed the intracellular trafficking of MPFMV focusing on the SP function in hopes of determining how the recruitment of a virus MP to the ER is involved in PD targeting . As a result , we found that the SP of MPFMV has extremely low ER translocation efficiency compared with conventional SPs of plant proteins , thereby causing abortive ER translocation of MPFMV at a high frequency . A fraction of MPFMV was translocated to the ER , whereas the remainder of MPFMV , which was not translocated to the ER , was transported to the patchy microdomains in the PM . Moreover , the ER-translocated MPFMV specifically localized to the ER-PM contact sites and played an essential role in the entry of microdomain-localized MPFMV into PD . Taken together , these findings suggest that dual targeting to two distinct subdomains in the ER and PM is essential for PD localization of MPFMV .
MPFMV was fused to YFP ( MPFMV:YFP ) to investigate its subcellular localization in Nicotiana benthamiana . Consistent with our previous results [23] , transiently expressed MPFMV:YFP localized to the punctate structures along the PM in epidermal cells ( Fig 1Ai ) . Treatment with aniline blue , which stains callose structures including PD , showed that the punctate structures of MPFMV:YFP colocalized with PD ( Pearson correlation coefficient [PCC] = 0 . 53 ± 0 . 03 ) . Measurement of the fluorescent intensity across plasmodesma shows that the fluorescence signal of MPFMV:YFP coincided with that of aniline blue ( Fig 1Aii ) , showing that MPFMV:YFP localized to PD . MPFMV:YFP appeared to accumulate also on the PM . PM and PD localization can be easily distinguished in plasmolyzed cells because PM proteins are associated with Hechtian strands , which are stretched PMs connecting the retracted PM and the cell wall , whereas PD proteins are retained in PD even during plasmolysis [24] . In plasmolyzed cells , MPFMV:YFP fluorescence was observed in Hechtian strands and PD ( Fig 1Bi ) . On the other hand , fluorescence signal was observed only in PD , but not in Hechtian strands , when cells expressing tobacco mosaic virus MP as a GFP fusion ( 30K:GFP ) were plasmolyzed ( Fig 1Bii ) . These results indicate that MPFMV:YFP accumulated in the PM , in addition to PD . Observation of the cell surface revealed that MPFMV:YFP shows patchy distribution throughout the PM ( Fig 1Ci ) . This distribution pattern was different from the fluorescence pattern of the PM stained with FM4-64 , an amphiphilic styryl dye which is inserted into the outer layer of the PM ( Fig 1Cii ) . These images were taken in the abaxial surface of the abaxial epidermal cells , and PD were not contained in these patches . Since such uneven distribution in the PM is similar to the localization of microdomain-associated proteins remorines [19 , 20] , a YFP fusion with A . thaliana REM1 . 3 ( YFP:REM1 . 3 ) was co-expressed with MPFMV:CFP; however , YFP:REM1 . 3 did not substantially co-localize with MPFMV:CFP ( PCC = −0 . 06 ± 0 . 12; Fig 1Ciii ) . MPFMV probably localized to the PM subdomains distinct from those of REM1 . 3 . To corroborate the subcellular distribution of MPFMV observed by microscopy , we performed chemical treatment using 1% TritonX-100 . 1% TritonX-100 was expected to solubilize proteins associated with the cellular membranes excluding proteins localized to detergent-insoluble domains , including PM microdomains [25] . As expected , TritonX-100 treatment solubilized the majority of a PM marker H+ATPase and an ER marker BIP ( Fig 1D ) . Conversely , only a small proportion of transiently expressed MPFMV:FLAG was soluble in 1% TritonX-100 . In FMV-infected cells , almost all MPFMV was detected from insoluble fraction . MPFMV became more insoluble in FMV-infected cells probably due to other virus factors . Furthermore , this result was in agreement with that of YFP:REM1 . 3 , which localizes to the Triton-insoluble microdomains [19 , 20] . Given that MPFMV unevenly distributed in the PM and that did not colocalize with REM1 . 3 ( Fig 1Ciii ) , MPFMV may localize to the detergent-insoluble microdomains different from those of REM1 . 3 . An ER-directed SP was predicted at the N-terminus of MPFMV by SignalP software in our previous work [23] . The cleavage site was between G19 and M20 , and the length of the deduced SP was 19 aa ( Fig 2A ) . To verify the cleavage at the predicted site , MPFMV:FLAG expressed transiently in N . benthamiana was purified by immunoprecipitation using anti-FLAG antibody . Immunoblot analysis using anti-FLAG antibody and Coomassie Brilliant Blue ( CBB ) staining of the immunoprecipitated samples showed a protein band of approximately 37 kD , indicating that precipitated MPFMV:FLAG has a single molecular weight ( Fig 2B ) . Amino acid sequences beginning at M20 , but not at the N-terminal methionine , were found in the sequences determined by Edman degradation of the purified MPFMV:FLAG , which means that the N-terminal 19 aa was cleaved off ( S1 Fig ) . Combined with CBB staining and immunoblot analysis , the N-terminal 19 aa was suggested to be almost perfectly processed from MPFMV:FLAG . The N-terminal 19 aa had characteristics of SPs , a central hydrophobic region which forms an α-helix and a polar C-terminal region ( Fig 2C ) [21 , 22] . These results suggest that MPFMV has an N-terminal SP ( hereafter referred to as SPFMV ) , which is expected to translocate the nascent protein into the ER . To confirm whether or not SPFMV translocates a protein into the ER lumen as those of plant proteins , we constructed GFPs flanked with the N-terminal SPs derived from MPFMV or plant proteins ( sporamin A , calreticulin or CLAVATA3 ) and a C-terminal ER retention signal ( Fig 3A; SPFMVGFP:HDEL , SPspoGFP:HDEL , SPcalGFP:HDEL and SPclvGFP:HDEL , respectively ) . We selected these three SPs because the SP activity was empirically assessed in previous studies [26–28] . These four SPs including SPFMV have no sequence homology ( S2 Fig ) . As controls , we prepared free GFP and GFP:HDEL . Here , it is noted that GFP:HDEL does not have an N-terminal SP . Unexpectedly , transiently expressed SPFMVGFP:HDEL was distributed throughout the cytosol , but not to the ER ( Fig 3B ) . The fluorescence pattern appeared to be the same as that of free GFP or GFP:HDEL . In contrast , SPspoGFP:HDEL , SPcalGFP:HDEL and SPclvGFP:HDEL predominantly localized in the ER , as expected . These results indicate the functional difference in ER translocation between SPFMV and the other SPs . We further investigated the ER translocation ability of these SPs using N-glycosylation as an indicator . We constructed fusion proteins in which the 3× N-glycosylation sequon was C-terminally fused to GFP , SPFMVGFP , SPspoGFP , SPcalGFP and SPclvGFP ( Fig 3C; GFPglc , SPFMVGFPglc , SPspoGFPglc , SPcalGFPglc and SPclvGFPglc , respectively ) [29] . If an SP successfully translocates the GFP fusion into the ER lumen after SP cleavage by an ER membrane-bound SP peptidase on the lumenal side , glycosylation of asparagine residues in the sequon by an ER-resident enzyme , oligosaccharyl transferase , occurs and results in an increase in the molecular weight of the translocated GFPglc relative to that of the non-translocated GFPglc . Immunoblot analysis of total proteins extracted from leaves expressing GFPglc , SPFMVGFPglc , SPspoGFPglc , SPcalGFPglc or SPclvGFPglc using anti-GFP antibody showed that almost all of the SPcalGFPglc or SPclvGFPglc molecules were glycosylated ( 91 . 9% and 93 . 0% , respectively ) , and that more than half ( 70 . 9% ) of the SPspoGFPglc molecules were glycosylated ( Fig 3D and 3E ) . However , compared with these measurements , a dramatically lower proportion ( 4 . 3% ) of the SPFMVGFPglc molecules were glycosylated . No glycosylation was detected in GFPglc , which lacks an SP . Thus , SPFMV had much lower translocation efficiency compared with conventional SPs , suggesting that only a small proportion of MPFMV molecules were translocated to the ER . In the case of SPFMVGFP:HDEL ( Fig 3B ) , the fluorescence of ER-translocated GFP was thought to be masked by GFP fluorescence in the cytosol because SPFMV translocated only a small fraction of GFP molecules into the ER . To gain insight into the role of ER translocation in the intracellular trafficking of MPFMV , we constructed an MPFMV mutant whose SP was not expected to be cleaved ( ncMP ) . In this mutant , two substitutions , L7P and V11P , were introduced into the central α-helix region of SPFMV to break the helix [21] . We verified that an SP was no longer predicted in the ncMP sequence by SignalP ( S3A Fig ) . Transiently expressed ncMP as a YFP fusion ( ncMPFMV:YFP ) showed aberrant accumulation in the cytoplasm ( S3Bi Fig ) , and did not target to the ER , PM and PD ( S3Bii Fig ) . This result indicates that the cleavage of SPFMV is essential for MPFMV to localize properly . We assessed localization of an MPFMV mutant lacking the N-terminal 19 aa SP . The SP-deficient mutant was fused with YFP ( Trun:YFP ) and expressed in the same conditions as MPFMV:YFP in Fig 1 . Trun:YFP was distributed to the PM microdomains , similar to MPFMV:YFP ( Fig 4Ai , compared with Fig 1Ci ) . The PM localization of Trun:YFP was checked by the fluorescence in Hechtian strands of plasmolyzed cells as was the case for MPFMV:YFP ( Fig 4Av left panel ) . Co-expression with ER-CFP showed that Trun:YFP was not associated with the perinuclear or peripheral ER ( PCC = −0 . 01 ± 0 . 00; Fig 4Aii ) . Also , Trun:YFP did not specifically localize to aniline blue-stained PD ( PCC = 0 . 21 ± 0 . 05; Fig 4Aiii ) . The fluorescent signal of Trun:YFP was reduced in the region corresponding to the center of PD ( Fig 4Aiv ) , unlike MPFMV:YFP ( Fig 1Aii ) . Furthermore , we confirmed that Trun:YFP was not retained in PD in plasmolyzed cells ( Fig 4Av right panel ) . These results indicate that the SP-deficient mutant did not localize to PD and that the SPFMV is essential for the PD localization of MPFMV , but dispensable for targeting the PM microdomains . Next , SPFMV function was compared with SPs derived from plant proteins , sporamin A , calreticulin or CLAVATA3 , by analyzing localization of SP chimeras in which the SPs of these proteins were fused to the N-terminus of Trun:YFP ( SPspoTrun:YFP , SPcalTrun:YFP and SPclvTrun:YFP , respectively ) . SPspoTrun:YFP localized to the PM microdomains ( Fig 4Bi , 4Bii and 4Bv ) and PD ( PCC = 0 . 47 ± 0 . 08; Fig 4Biii and 4Biv ) , similar to MPFMV:YFP . However , unlike MPFMV:YFP , cytoplasmic aggregations were observed in the SPspoTrun:YFP-expressing cells ( Fig 4Bi ) . Co-expression with the ER marker ER-CFP showed that these aggregations were formed in the ER ( Fig 4Bii ) . On the other hand , SPcalTrun:YFP and SPclvTrun:YFP were associated with nodes in the ER network ( PCC: 0 . 55 ± 0 . 05 and 0 . 52 ± 0 . 06 , respectively; Fig 4Ci , Cii , 4Di and 4Dii ) . Although some of these punctate spots were located in close proximity to PD , many of them did not co-localize with PD ( PCC: 0 . 18 ± 0 . 03 and 0 . 27 ± 0 . 07 , respectively; Fig 4Ciii and 4Diii ) . Fluorescence intensity measurements confirmed that fluorescence signals of SPcalTrun:YFP and SPclvTrun:YFP did not coincide with that of aniline blue ( Fig 4Civ and 4Div ) . Hechtian strands were invisible when cells expressing SPcalTrun:YFP or SPclvTrun:YFP were plasmolyzed ( Fig 4Cv and 4Dv ) , indicating that these chimeras were not distributed to the PM . These observations suggest that SPFMV plays an essential role in MPFMV localization . Our previous study showed that the expression of MPFMV complements cell-to-cell movement of movement-deficient PVX mutant ( PVXΔTGBp1-GFP ) [23] . We investigated whether MPFMV mutants ( Trun , SPspoTrun and SPcalTrun ) facilitate virus cell-to-cell movement using this system . SPclvTrun was not used in this experiment because its transient expression for more than 3 days caused cell death . The fluorescence of PVXΔTGBp1-GFP spread to adjacent cells when MPFMV or SPspoTrun was co-expressed ( Fig 5A ) , indicating that MPFMV and SPspoTrun complemented cell-to-cell movement of PVXΔTGBp1-GFP . Quantitative analysis of fluorescence area suggested significant differences between MPFMV or SPspoTrun and β-glucuronidase ( GUS ) control ( Fig 5B ) . Categorizing the fluorescence area by the cell number per fluorescent spot also showed significant difference between MPFMV or SPspoTrun and GUS control ( p < 0 . 01 by Fisher's exact test ) . In contrast , the fluorescence of PVXΔTGBp1-GFP was almost confined to a single cell when co-expressed with Trun or SPcalTrun like GUS control ( Fig 5A ) . Comparable expression levels of MPFMV and its mutants were validated by Western blot analysis ( Fig 5C ) . Thus , SPspoTrun , an SP chimera which has the ability to reach PD , complements cell-to-cell movement of PVXΔTGBp1-GFP , indicating that PD localization of MPFMV is necessary to facilitate virus cell-to-cell movement . Most virus MPs are known to move to adjacent cells autonomously [30] . We assessed whether or not these MPFMV mutants were able to move to adjacent cells . MPFMV:YFP and SPspoTrun:YFP spread to adjacent cells when expressed in a single cell ( Fig 6A ) . Conversely , Trun:YFP , SPcalTrun:YFP and SPclvTrun:YFP , which do not have the PD-targeting ability , did not move to adjacent cells . Quantitative analysis suggested a significant difference between these two groups in the ability to move to adjacent cells ( Fig 6B ) . These results are in good accordance with virus cell-to-cell complementation assay ( Fig 5 ) . The experiments described above suggested that the translocation efficiencies determine the subcellular distribution and function of MPFMV . SPcalTrun and SPclvTrun , whose SPs have high translocation efficiencies , localized to the ER , whereas Trun , which does not have an SP , localized to the PM ( Table 1 ) . SPspoTrun , which have moderate translocation efficiency , were able to localize to PD in addition to the ER and the PM . Given that SPFMV has low translocation efficiency , a small fraction of MPFMV was probably recruited to the ER . These results allowed us to speculate that co-existence of ER-translocated and non-translocated MPFMV is required for PD localization; in other words , ER-translocated MPFMV and microdomain-localized MPFMV act cooperatively to reach PD . We tested this hypothesis by co-expressing SP-deficient MPFMV fused with CFP ( Trun:CFP ) and MPFMV:YFP , Trun:YFP , SPspoTrun:YFP , SPcalTrun:YFP or SPclvTrun:YFP . Localization pattern was not altered when Trun:YFP and Trun:CFP were co-expressed ( Fig 7B ) . By contrast , Trun:CFP was found to localize to punctate structures along the PM when co-expressed with MPFMV:YFP , SPspoTrun:YFP , SPcalTrun:YFP and SPclvTrun:YFP ( Fig 7A and 7C–7E ) . Aniline blue staining confirmed that these punctate structures formed by Trun when co-expressed with MPFMV and SP chimeras coincided with PD ( PCC: 0 . 50 ± 0 . 06 , 0 . 60 ± 0 . 06 , 0 . 61 ± 0 . 05 and 0 . 56 ± 0 . 15 , respectively; S4 Fig ) . YFP-fused MPFMV and SP chimeras did not alter their localization by the expression of Trun:CFP . Detailed views showed that SPcalTrun:YFP and SPclvTrun:YFP localized in close proximity to , but did not substantially colocalize with Trun:CFP ( Fig 7D and 7E; PCC: 0 . 31 ± 0 . 11 and 0 . 29 ± 0 . 03 , respectively ) . This localization pattern was similar to those observed when SPcalTrun:YFP or SPclvTrun:YFP was expressed alone ( Fig 4Ciii and 4Diii ) . These results suggest that Trun , which normally localized to the PM , was transported to PD by the function of MPFMV or SP chimeras , which were at least partially translocated to the ER . The fact that MPFMV or SP chimeras changed Trun localization raised the possibility of the physical interaction between microdomain-localized MPFMV and ER-translocated MPFMV . We have now investigated the interaction between MPFMV and Trun or SPclvTrun using bimolecular fluorescence complementation ( BiFC ) . We first co-expressed the basic leucine zipper transcription factor bZIP63 fused with N-terminal half of YFP ( bZIP63:NYF ) and with C-terminal half of YFP ( bZIP63:CYF ) as a control [31] . In this combination , strong fluorescence was observed in the nuclei ( S5 Fig ) . Next , we tested the interaction between MPFMV:NYF and MPFMV:CYF , Trun:CYF or SPclvTrun:CYF . Co-expression of MPFMV:NYF and MPFMV:CYF or Trun:CYF showed a weak fluorescence on the PM . This fluorescence can be ascribed to a background signal or a weak dimerization of microdomain-localized MPFMV . Co-expression of MPFMV:NYF and SPclvTrun:CYF showed no signal , and the physical interaction between microdomain-localized MPFMV and ER-translocated MPFMV was not suggested . Given that SPcalTrun:YFP and SPclvTrun:YFP localized in the close proximity to PD ( Fig 4Ciii and 4Diii and Fig 7D and 7E ) , translocated MPFMV appeared to localize in the specific region of the ER . We noticed that the distribution pattern of SPcalTrun:YFP and SPclvTrun:YFP was similar to that of the ER-PM contact site-associated protein , synaptotagmin1 ( SYTA ) [14 , 32] , which is known to be involved in virus cell-to-cell movement [15–17] . Prior to the co-expression with MPFMV:YFP , we first analyzed localization of SYTA . Co-expression of SYTA:CFP and ER-YFP showed that SYTA:CFP was predominantly distributed to nodes of the ER ( PCC = 0 . 61 ± 0 . 01; S6A Fig ) consistent with the previous reports [14 , 15] . Aniline blue staining showed that SYTA:CFP was in close proximity to PD ( PCC = 0 . 21 ± 0 . 01; S6B Fig ) , which was similar to those seen in SPcalTrun:YFP and SPclvTrun:YFP ( Fig 7D and 7E ) . The close proximity of ER-PM contact sites and PD was reported also in a previous study [32] . To assess whether translocated MPFMV localize to ER-PM contact sites , SYTA:CFP was co-expressed with SPclvTrun:YFP or MPFMV:YFP . SPclvTrun:YFP co-localized almost perfectly with SYTA:CFP , probably in the ER nodes ( PCC = 0 . 54 ± 0 . 08; Fig 8A ) . A large fraction of MPFMV:YFP was distributed to the PM and PD , but a small fraction of punctate structures colocalized with SYTA:CFP ( Fig 8B ) . These results show that ER-translocated MPFMV specifically localized to ER-PM contact sites . To obtain more information about the role of ER-PM contact sites in MPFMV trafficking , we constructed a c-myc tagged SYTAΔC2B ( SYTAΔC2B:myc ) , which is a dominant-negative form lacking the C-terminal 177 aa of SYTA [16] . When MPFMV:YFP was expressed together with SYTAΔC2B:myc , localization of MPFMV:YFP was apparently affected ( Fig 8C and 8D ) . Although a small fraction of MPFMV:YFP was still retained in PD , a large proportion of MPFMV:YFP excessively accumulated next to PD ( PCC = 0 . 20 ± 0 . 13; Fig 8C ) . MPFMV:YFP substantially colocalized with SYTA:CFP in the cortex; instead , fluorescence from MPFMV that localized in the PM microdomains became weaker ( PCC = 0 . 72 ± 0 . 10; Fig 8D ) . This result suggests that MPFMV:YFP that normally localized to the PM microdomains aberrantly accumulated in ER-PM contact sites by the expression of SYTAΔC2B:myc . We also investigated whether SYTAΔC2B affects cell-to-cell movement of MPFMV:YFP . Expression of SYTAΔC2B:myc inhibited MPFMV:YFP movement to adjacent cells compared with when expressed with GUS ( Fig 8Ei and 8Eii ) . Immunoblot analysis using anti-myc antibody confirmed the expression of SYTAΔC2B:myc ( Fig 8Eiii ) . Taken together , these data suggest that translocated MPFMV localized to ER-PM contact sites and played an essential role in cell-to-cell movement . To see whether MPFMV interacts with SYTA , BiFC was carried out . Fluorescent signal was not observed when MPFMV:NYF , Trun:NYF and SPclvTrun:NYF were co-expressed with SYTA:CYF ( S7 Fig ) . Thus , the interaction between SYTA and MPFMV or MPFMV mutants was not suggested by BiFC . As the MPs of several viruses use the secretory pathway [9 , 33] , involvement of COPII transport in MPFMV trafficking has been verified by BFA treatment or expression of a dominant-negative form of Sar1 [Sar1 ( H74L ) ] [34] . We first confirmed that BFA treatment and Sar1 ( H74L ) expression caused retention of the Golgi marker ManI:CFP [35] in the ER , as expected ( S8 Fig ) . PD localization in cells expressing MPFMV:YFP was not affected by BFA treatment or Sar1 ( H74L ) expression . Similarly , inhibiting COPII transport did not affect localization of Trun:YFP in the PM . These results suggest that COPII transport is not involved in the subcellular localization of MPFMV and that MPFMV uses pathways different from BFA-sensitive MPs [9 , 33] .
In this study , we analyzed the intracellular trafficking of MPFMV , focusing on SP function , and found that MPFMV targets two subdomains in the ER and PM as well as PD . MPFMV , which has an N-terminal SP , was distributed mainly in PD and patchy microdomains of the PM ( Fig 1 and Fig 2 ) . Investigation of ER translocation efficiency revealed that SPFMV has much lower translocation efficiency compared with those of SPspo , SPcal and SPclv ( Fig 3 ) . The SP-deficient mutant ( Trun ) exclusively localized to the PM microdomains ( Fig 4A ) , whereas two SP chimeras ( SPcalTrun and SPclvTrun ) exclusively localized to the ER ( Fig 4C and 4D ) . SPspoTrun was distributed in the ER , PM microdomains and PD similar to MPFMV , even though a portion of SPspoTrun aggregated in the ER ( Fig 4B ) . The results so far indicated that MPFMV dually targets the ER and PM due to the inefficient SP; a fraction of MPFMV was successfully translocated into the ER , whereas the remainder of MPFMV , which failed to be translocated , is transferred to the microdomains . This finding led us to speculate that both ER-translocated MPFMV and microdomain-localized MPFMV are necessary for PD localization . Consistent with this notion , the SP-deficient mutant entered into PD by the expression of MPFMV or the SP chimeras , which are able to , at least partially , translocate into the ER ( Fig 7 ) . Furthermore , we showed that translocated MPFMV specifically localized to ER-PM contact sites ( Fig 8A and 8B ) , and dominant-negative inhibition of SYTA affected PD localization and cell-to-cell movement of MPFMV ( Fig 8C–8E ) . These results suggest that MPFMV localized to ER-PM contact sites plays an essential role in the entry of microdomain-localized MPFMV into PD . PD localization of MPFMV is necessary to facilitate cell-to-cell movement as shown in the virus movement complementation assay ( Fig 5 ) and the cell-to-cell movement assay ( Fig 6 ) . Altogether , we propose a new model for the intracellular trafficking of a viral MP . A substantial proportion of MPFMV , which failed to be translocated , is directly transferred to the microdomains , whereas the remainder of MPFMV , which successfully translocated into the ER , subsequently localizes to ER-PM contact sites and functionally interact with PD to enable microdomain-localized MPFMV to enter into PD ( Fig 9 ) . Dual targeting of MPFMV to the ER and the PM is explained by the low translocation efficiency of the SPFMV ( Fig 3 ) . In general , proteins with an SP are co-translationally recognized by the signal recognition particle in the cytosol and recruited to the signal recognition particle receptor on the ER membrane . Then , the Sec61p complex , a channel integrated into the ER membrane , translocates these preproteins into the ER , and subsequently the SP is cleaved by a membrane-bound SP peptidase on the lumenal side [36] . Therefore , the finding that MPFMV was translocated to the ER at a lower rate ( approximately 5%; Fig 3 ) despite the cleavage of SPFMV ( Fig 2 ) is surprising . A substantial proportion of MPFMV molecules probably abort ER translocation after SP cleavage and are released into the cytosol , but further studies are needed to reveal the detailed mechanism of the abortion . In animal cells , it has been reported that an inefficient SP of an ER chaperone calreticulin regulates the ratio of translocated and nontranslocated populations [37] . However , the inefficiency of the animal calreticulin SP is quite moderate , and it generates only a small nontranslocated population . In this regard , to our knowledge , this is the first report of an SP with extremely low efficiency that controls the subcellular distribution of the nascent protein . This low-efficiency SPFMV generates only a small ER-translocated population , but it is probably sufficient for the entry of microdomain-localized MPFMV into PD considering that SPspoTrun:YFP , whose SP has medium translocation efficiency , excessively accumulated in the ER compared with the case of MPFMV:YFP ( Fig 4B ) . SPFMV likely regulates the distribution of MPFMV between the PM and ER in the appropriate proportion . How is MPFMV associated with two different types of the cellular membrane , the ER membrane and PM ? Curiously , a transmembrane domain in MPFMV has not been predicted by SOSUI ( http://harrier . nagahama-i-bio . ac . jp/sosui/ ) and TMHMM server v . 2 . 0 ( http://www . cbs . dtu . dk/services/TMHMM/ ) . Localization of the MPFMV to the PM microdomains is explained as a peripheral membrane protein . Given that the SP-deficient MPFMV was exclusively associated with the PM microdomains ( Fig 4A ) , even ER translocation is not required to localize to PM microdomains . Microdomain-associated proteins , remorins and flotillins , which do not have SPs and transmembrane domains similarly , are suggested to be peripherally associated with the PM [38–40] . In other words , a transmembrane domain and penetrating the membrane are dispensable for the localization to PM microdomains . The association with the ER membrane is explained by the abnormality of viral proteins . Recent studies on the topology of viral MPs revealed that MPs of tobacco mosaic virus and tomato spotted wilt virus are suggested to be associated with the ER membrane using unusual hydrophobic regions [41–43] . MPFMV may also establish such unconventional hydrophobic regions to be associated with the ER membrane . Our results raise the question of how two populations of MPFMV that localize to the ER and PM subdomains act cooperatively to gain access to PD . According to BiFC analysis ( S7 Fig ) , physical interaction of these two population is not likely . Although only limited information is available on these membrane subdomains , one possible explanation is that MPFMV in ER-PM contact sites functionally interact with PD , and this might allow the microdomain-localized MPFMV to access into PD ( Fig 9 ) . This hypothesis is corroborated by the facts that ER-PM contact sites are spatially close to PD ( S6B Fig ) [32] and that expression of SYTAΔC2B , a dominant-negative form of SYTA , affected PD and PM localization of MPFMV ( Fig 8C and 8D ) . In accord with our results , a previous study showed that , although the localization of 30K was not affected , the cell-to-cell movement of 30K was suppressed in an Arabidopsis syta mutant [16] . This study also showed that SYTAΔC2B inhibited the formation of endosomes , suggesting that SYTA regulates endocytosis [16] . From these facts , we suspect that MPFMV in the patchy microdomains is transported into PD through endosomal trafficking regulated by SYTA ( Fig 9 ) . The intimate relationship between PD and patchy domains in the PM is implied in this and in other studies . In our study , relocation of the microdomain-localized MPFMV to PD ( Fig 7 ) indicates a functional connection between the patchy microdomains and PD . One previous study about TMV 30K reported that dominant-negative inhibition of class VIII myosins affected PD localization of 30K and induced a patchy distribution in the PM , which did not merge with that of REM1 . 3 similar to the results from this study [44] . Originally , PM passing through PD is also recognized as a type of microdomain , as it is functionally and spatially distinguished from the surrounding PM [2 , 45] . Taken together , these two types of PM microdomains , the patchy domains and PM passing through PD , might be functionally connected by ER-PM contact sites . This study also presents a functional differentiation of MPFMV between the two populations , MPFMV in ER-PM contact sites and PM microdomains . The functional differentiation of a virus MP is reminiscent of the mechanism of cell-to-cell movement regulated by more than one protein . For example , triple gene block movement proteins , which are encoded by viruses belonging to the Virgaviridae , Alphaflexiviridae and Betaflexiviridae families , have specialized functions and perform different tasks for virus cell-to-cell movement: delivering viral factors to PD , interacting with host factors , and increasing PD permeability [46 , 47] . MPFMV plays multiple roles in cell-to-cell movement using the SPFMV with low translocation efficiency , probably to avoid splitting into modules . This concept may be true also in plant proteins . Although a number of studies have shown that a diverse array of plant proteins have N-terminal SPs , their ER translocation abilities have seldom been investigated . Our findings raise the possibility that SPs can potentially regulate subcellular distribution of the nascent proteins and contribute to protein function in plant cells .
Plasmids expressing GUS , MPFMV , MPFMV:YFP and MPFMV:CFP were prepared as described earlier [23] . Expression vectors of ER-CFP and ER-YFP were purchased from the Arabidopsis Biological Resource Center ( Stock numbers CD3-953 and CD3-957 , respectively ) . MPFMV mutant whose SP is not cleaved ( ncMPFMV ) was generated by an PCR using primers containing substitutions to introduce L7P and V11P mutations . ncMPFMV was cloned into pEarleyGate 101 ( ncMPFMV:YFP ) using Gateway technology [48 , 49] . Trun sequence , which lacks the N-terminal 19 aa of MPFMV , was amplified by PCR and cloned into pEarleyGate 100 ( Trun ) , pEarleyGate 101 ( Trun:YFP ) and pEarleyGate 102 ( Trun:CFP ) . SP chimera sequences were amplified by PCRs using primers containing each SP sequence , followed by cloning into pEarleyGate 100 ( SPspoTrun , SPcalTrun and SPclvTrun ) or 101 ( SPspoTrun:YFP , SPcalTrun:YFP and SPclvTrun:YFP ) . In these SP chimeras , the SP region of MPFMV was replaced with the N-terminal sequences of Ipomoea batatas sporamin A ( M16861; 24 aa ) [26] , Nicotiana tabacum calreticulin ( EU984501; 28 aa ) [27] or A . thaliana CLAVATA3 ( AF126009; 22 aa ) [28] , each of which contains an SP sequence and one aa downstream of the cleavage site . MPFMV:FLAG , in which MPFMV was fused to a FLAG epitope tag immediately downstream of its C terminus , was amplified by PCRs using primers containing FLAG sequence , and cloned into pEarleyGate 100 . A 3× N-glycosylation sequon [29] or an ER-retention signal [50] were introduced to the GFP sequence in their C terminus ( GFPglc and GFP:HDEL ) as was the case with MPFMV:FLAG . The GFP sequence was derived from pEarleyGate 103 . The SPs of MPFMV , sporamin A , calreticulin or CLAVATA3 were N-terminally added to GFPglc ( SPFMVGFPglc , SPspoGFPglc , SPcalGFPglc and SPclvGFPglc ) or GFP:HDEL ( SPFMVGFP:HDEL , SPspoGFP:HDEL , SPcalGFP:HDEL and SPclvGFP:HDEL ) as was the case with SP chimeras . The REM1 . 3 ( At4g36970 ) and SYTA ( At2g20990 ) sequences were amplified by PCR from total DNA of A . thaliana . REM1 . 3 was cloned into pEarleyGate 104 ( YFP:REM1 . 3 ) , and SYTA was cloned into pEarleyGate 101 ( SYTA:YFP ) and pEarleyGate 102 ( SYTA:CFP ) . SYTAΔC2B , a dominant-negative form of SYTA , was amplified by PCR according to a previous study [16] , and cloned into pEarleyGate Cmyc ( SYTAΔC2B:myc ) . pEarleyGate Cmyc is an in-house expression vector built from pEarleyGate 101 to introduce a myc tag at the C terminus of a cloned gene . Vectors for BiFC analysis were constructed as shown in the previous study [31] . The TMV MP:GFP sequence ( 30K:GFP ) was amplified from pTMV-MP:GFP [51] , and cloned into the pBI121 vector using SalI and BamHI sites . The upper leaves of four-week-old N . benthamiana plants were used for the transient expression assays . Transient expression was mediated by infiltration of Agrobacterium tumefaciens as described previously [23] . Cells expressing fluorescent protein fusions were imaged using a Leica TCS SP5 laser-scanning confocal microscope . An HCX PL Apo 63×/1 . 4–0 . 6 oil CS lens was used for imaging subcellular localization of fluorescent protein fusions and an HC PL Apo 10×/0 . 4 CS lens was used for imaging cell-to-cell movement of fluorescent protein fusions . Cells expressing CFP and/or YFP fusions were visualized as described earlier [52] . GFP was excited at the 488-nm argon laser line , and the emission was visualized at 500 to 600 nm . For PM staining , leaves were infiltrated with 50 μM FM4-64 in distilled water , and observed at 1 hpi . FM4-64 was excited at the 543-nm helium/neon laser line , and the emission was visualized at 580 to 650 nm . For PD staining , leaves were infiltrated with 0 . 1% ( w/v ) aniline blue in 50 mM sodium phosphate buffer ( pH 9 . 0 ) , and cells were observed at 2 hpi . Aniline blue was excited at the 405-nm laser line , and the emission was visualized at 425 to 480 nm . All the images were acquired at room temperature . Confocal images were processed with LAS AF software version 2 . 7 . 3 and Adobe Photoshop CS4 . For deconvolution image analysis , between 3 and 8 z-section images at 0 . 15 μm intervals were captured and processed with Leica Hyvolution system . Fluorescence intensity graphs were generated using the LAS AF quantify intensity tool . PCCs were measured using an Fiji Colocalization plugin [53] , and the mean values and standard deviations were calculated from three different images . Generally , PCC values of 0 . 2–0 . 4 indicate weak positive correlations and PCC values above 0 . 5 indicate strong positive correlations [54] . Leaves expressing fluorescent protein fusions were immersed in 4% ( w/v ) NaCl for 15 min for plasmolysis . For the inhibition assay , leaves were treated with 50 μg/ml BFA in 0 . 5% ( v/v ) dimethyl sulfoxide at 18 hpi and observed at 6 h after the treatment . A GFP-tagged movement-defective mutant of the PVX infectious clone ( PVXΔTGBp1-GFP ) [55] was used in the virus movement complementation experiment . Two Agrobacterium cultures harboring the binary plasmid expressing GUS , MPFMV or an MPFMV mutant and PVXΔTGBp1-GFP were resuspended and mixed to final concentrations of OD600 = 0 . 4 and 0 . 0002 , respectively . Leaves at 5 dpi were observed under an M165 FC fluorescence stereomicroscope ( Leica Microsystems ) with an ET GFP filter . Images were captured by a Leica DFC 310 FX camera and LAS software version 4 . 4 . 0 . The areas of fluorescent foci were measured using ImageJ software version 1 . 40 ( National Institutes of Health ) . In the assessment of cell-to-cell movement , an Agrobacterium culture harboring the binary plasmid expressing MPFMV or its mutants was resuspended and diluted to OD600 = 0 . 0002 . When cell-to-cell movement under condition of dominant-negative inhibition of SYTA was investigated , Agrobacterium cultures harboring the binary plasmid expressing SYTAΔC2B:myc or GUS and MPFMV:YFP were resuspended and mixed to final concentrations of OD600 = 1 . 0 and 0 . 0002 , respectively . Leaves at 3 dpi were observed under the laser-scanning confocal microscope as described above . FMV-infected fig leaves and N . benthamiana leaves transiently expressing MPFMV:FLAG or YFP:REM1 . 3 at 36 hpi were used for isolation of membrane-rich fraction ( P30 ) . The fractionation of P30 and chemical treatment were carried out according to the methods of Schaad et al . [56] with minor modifications as follows: Complete Mini ( Roche Diagnostics ) was added to buffer Q ( Lysis Buffer ) as a protease inhibitor instead of leupeptin , aprotinin and phenylmethylsulfonyl fluoride . P30 pellets were treated with buffer Q or 1% Triton X-100 ( 1%[v/v] TritonX-100 , 25 mM Tris-HCl [pH 7 . 5] , 150 mM NaCl and 5 mM EDTA ) . MPFMV:FLAG was immunoprecipitated from cell lysate of N . benthamiana leaves expressing MPFMV:FLAG with EZview red anti-FLAG M2 affinity gel ( Sigma-Aldrich ) [52] . Immunoblot analysis was performed [52] using anti-FLAG M2 antibody ( Sigma-Aldrich ) , anti-myc antibody ( EMD Millipore ) , anti-Bip antibody ( Santa Cruz Biotechnology , Inc . ) , anti-H+ATPase antibody ( Agrisera ) or anti-MPFMV antibody . Anti-MPFMV antibody , a polyclonal antibody against the mature region of MPFMV , was generated as described previously [52] . SignalP 4 . 1 software was used to predict SPs [57] . The accession number of MPFMV sequence is BAM13816 . Deglycosylation using endoglycosidase H ( Endo H; New England Biolabs ) was carried out according to the manufacturer's instructions . Leaves transiently expressing GFPglc , SPFMVGFPglc , SPspoGFPglc , SPcalGFPglc or SPclvGFPglc at 30 hpi were homogenized in 1× glycoprotein denaturing buffer ( 0 . 5% SDS and 40 mM DTT ) , and incubated at 65°C for 15 min . After removal of cellular debris by centrifugation , the supernatant was suspended in 1× GlycoBuffer 3 ( 50 mM sodium acetate , pH 6 . 0 ) followed by the addition of distilled water or Endo H . After incubation at 37°C for 1 h , Endo H was inactivated at 75°C for 10 min . These samples were analyzed by immunoblotting using anti-GFP antibody ( Roche ) . Signal intensity of each band was quantified using ImageJ software . | Intercellular trafficking of molecules through plasmodesmata ( PD ) is indispensable for plant development . Plant viruses also use the intercellular trafficking system to establish systemic infection . Virus movement proteins ( MPs ) , which have abilities to localize to PD and to move to the adjacent cells autonomously , play important roles in facilitating virus cell-to-cell movement . Hence , understanding how MPs reach PD has great significance for virology and plant cell biology . In this study , we analyzed the intracellular trafficking of fig mosaic virus movement protein ( MPFMV ) mainly based on its N-terminal signal peptide ( SP ) . SPs , short peptides directing proteins to the ER , are frequently found in a diverse array of proteins , but rarely found in plant virus proteins . We focused on the SP of MPFMV and investigated the relationship between ER translocation and PD localization . We showed that the SP of MPFMV had quite low translocation efficiency and contributes to generating two distinct populations . Each population localized to specialized subdomains of the ER and PM , and was essential for PD localization , indicating that these subdomains and PD are functionally related . Thus , our findings offer new insights into cell-to-cell movement in plants . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"physiology",
"fluorescence",
"imaging",
"chemical",
"compounds",
"plant",
"cell",
"biology",
"organic",
"compounds",
"luminescent",
"proteins",
"plant",
"science",
"immunoprecipitation",
"yellow",
"fluorescent",
"protein",
"chaperone",
"proteins",
"anilines",
"re... | 2017 | Dual targeting of a virus movement protein to ER and plasma membrane subdomains is essential for plasmodesmata localization |
The root knot nematode , Meloidogyne incognita , is an obligate parasite that causes significant damage to a broad range of host plants . Infection is associated with secretion of proteins surrounded by proliferating cells . Many parasites are known to secrete effectors that interfere with plant innate immunity , enabling infection to occur; they can also release pathogen-associated molecular patterns ( PAMPs , e . g . , flagellin ) that trigger basal immunity through the nematode stylet into the plant cell . This leads to suppression of innate immunity and reprogramming of plant cells to form a feeding structure containing multinucleate giant cells . Effectors have generally been discovered using genetics or bioinformatics , but M . incognita is non-sexual and its genome sequence has not yet been reported . To partially overcome these limitations , we have used mass spectrometry to directly identify 486 proteins secreted by M . incognita . These proteins contain at least segmental sequence identity to those found in our 3 reference databases ( published nematode proteins; unpublished M . incognita ESTs; published plant proteins ) . Several secreted proteins are homologous to plant proteins , which they may mimic , and they contain domains that suggest known effector functions ( e . g . , regulating the plant cell cycle or growth ) . Others have regulatory domains that could reprogram cells . Using in situ hybridization we observed that most secreted proteins were produced by the subventral glands , but we found that phasmids also secreted proteins . We annotated the functions of the secreted proteins and classified them according to roles they may play in the development of root knot disease . Our results show that parasite secretomes can be partially characterized without cognate genomic DNA sequence . We observed that the M . incognita secretome overlaps the reported secretome of mammalian parasitic nematodes ( e . g . , Brugia malayi ) , suggesting a common parasitic behavior and a possible conservation of function between metazoan parasites of plants and animals .
M . incognita can infect 1 , 700 plant species [1] . At the infective juvenile ( J2 ) stage of development , M . incognita enters the elongation zone of the root and burrows through the apoplast to the root tip where it enters the vascular cylinder , moving up to the zone of root differentiation . The nematode then inserts its stylet into the plant cell cytoplasm and induces nuclear division without cytokinesis , creating multinucleate giant cells that nurture the developing worm . Infection is associated with the reprogramming of plant cell development rather than host cell death [2] . M . incognita infection causes plant defense genes to become either promptly suppressed or transiently induced , in contrast to incompatible interactions , which immediately induce and sustain expression of defense genes [3] . The proteins and metabolites secreted from the esophageal glands ( subventral and dorsal glands ) of plant-parasitic nematodes are thought to be responsible for compatibility [4] . The two subventral gland ( SvG ) cells are biologically active during the J2 stage , while the dorsal gland cell is predominantly active on the second day post-infection through to the end of the nematode's life . In vivo observations of the root cyst nematode , Heterodera schachtii , revealed that the dorsal gland secretions are released through the stylet into the plant cell [5] . Other nematode tissues also secrete proteins that may be important for plant-pathogen interaction: two amphids localized in the anterior part of the worm , around the lip region , and two phasmids at the posterior part could be receptors for chemotaxis [6] . These two kind of organs contain socket cells that are highly secretory but functions of these secretions remain mostly unknown [6] , [7] . Following the establishment of compatibility , pathogen-produced effector molecules are the key to infection . These molecules have been found in well-characterized pathosystems where they modulate host signaling pathways to prevent defense responses [8] , but little is known about effectors that mediate plant-metazoan pathogenesis . Bird and Saurer ( 1967 ) characterized secreted molecules from the esophageal gland cells of Meloidogyne javanica [9] . They showed that the secretions were mainly proteins; no nucleic acids were detected . Antibodies have been used to monitor the expression of esophageal antigens from several plant nematode species [10] . In Globodera rostochiensis , antibodies recognized proteins present both in the subventral gland cells and on the surface of the nematode [11] . Other studies of Meloidogyne spp . showed that silencing of genes expressed in the SvG reduced pathogenicity [12] , [13] . Secretions of the animal-parasitic nematode , Trichinella spiralis , appear to reprogram the host cell into a nurse cell , and in vitro injection of collected secretions from T . spriralis into rat muscles mimicked cellular changes that occur in vivo [14] . Pathogen associated molecular patterns ( PAMPs ) are typically proteins or nucleic acids that are wide-spread in microbes and are shed during infection . Host receptors are activated by PAMPs . For example , flagellin from bacteria stimulates innate immunity from both plant and mammalian cells [15] , [16] . No PAMPs from metazoans have been reported . The identification of secreted proteins from M . incognita may facilitate the discovery of effectors and PAMPs . Effectors might account for how root knot nematodes reprogram plant cells to become giant cells and to form root knots . Several hypotheses have been proposed to explain how M . incognita establishes compatibility with its plant hosts . It may invade root tissues by first producing cell wall-degrading enzymes . Once established in the root it could produce detoxifying enzymes , followed by additional effectors that induce giant cell formation [2] . Discovery of secreted proteins by bioinformatics is possible for organisms with known genomic DNA sequences . Recently a secretome of Plasmodium falciparum comprising 200 proteins was predicted using bioinformatics [17] . A similar approach cannot be applied to M . incognita since its genomic sequence is not yet known . Instead , experimental approaches have been used . A transcript profile of the esophageal gland cells of M . incognita has been reported [18] . Based on bioinformatic analyses of cDNA sequences , secreted proteins were predicted to include cellulases , chitinases , extensins , proteases , and a superoxide dismutase ( SOD ) . In a recent study , Roze et al . ( 2008 ) analyzed the cDNA sequences of proteins putatively secreted by Meloidogyne chitwoodi [19] . They identified cDNAs corresponding to 398 putative proteins and confirmed by in situ hybridization seven that are specifically transcribed in the SvG , one in the dorsal gland , and one in the phasmids . We chose to directly identify secreted proteins based on the pioneering work of Jaubert et al . ( 2002 ) who used resorcinol to induce esophageal gland secretion by M . incognita . It was clear from their work that many more proteins were secreted than were identified [20] . To explore the M . incognita secretome in greater depth , we developed sensitive methods for high-throughput proteomics based liquid chromatography , nano-electrospray ionization and tandem mass spectrometry ( nanoLC ESI MS/MS ) [21] . This method requires both a protein database as well as algorithms to assign peptide sequences to mass spectra . The conservation of protein sequences between species enables a protein database from heterologous species to partially substitute for a database from the cognate species . To control for false positive identifications we reversed the amino acid sequence of the protein databases and filtered the search result so that our protein false discovery rate ( FDR ) was 0 . 4% . While use of heterologous databases precludes discovery of peptides that are unique to the organism , thereby reducing the number of proteins identified , it nevertheless opens a window on the proteome . In this study , we identified 486 proteins from the M . incognita secretome , including proteins that could play a role in root knot formation by regulating the plant cell cycle and plant growth .
We induced protein secretion by J2 stage M . incognita nematodes by treating first with filtered , low-molecular weight ( <3 , 500 Da ) tomato root exudates followed with resorcinol . A sample of nematodes was removed and stained with Coomassie Blue , which confirmed that treatment caused proteins to be secreted from the stylet region . Secreted proteins were extracted from the solution bathing unstained nematodes and were identified by nanoLC ESI MS/MS ( Figure 1 ) . To ensure the accuracy of protein identifications , the threshold for mass spectral quality was set at high stringency using very low peptide and protein false discovery rates ( FDRs ) . FDRs were determined by searching the MS/MS spectra against a concatenated 1∶1 forward-reverse database [22] . MS/MS spectra with peptide FDR less than 0 . 1% were considered valid . We set the protein FDR at 0 . 4% . Due to the multiple protein databases used in this study and the natural sequence redundancy in the protein databases , the same peptide sometimes appeared in multiple protein sequences . In order to address this protein redundancy issue , protein sequences containing the same set or subset of valid peptides were grouped together into protein groups with the best match listed first [23] . The numbers of proteins we report in this paper are protein group numbers . This is a conservative measure because more than one protein within a group may actually be detected . Only proteins with at least 2 valid MS/MS spectra were reported . Proteins with a single unique peptide but multiple spectra were manually validated . Observations of M . incognita and Heterodera glycines stylet activity with and without stimulation by neuroregulators has been extensively studied [24] . The authors reported that neuroregulators induce a dramatic stimulation of stylet pulsing frequency but they pointed out that even without stimulation , stylet pumping occurred . We observed by both mass spectrometry and silver stained gels that J2 nematodes secrete low but detectable levels of proteins ( less than 1% as much as after stimulation ) . Proteins identified in the absence of stimulation included 14-3-3b [listed as protein ( 4 ) in Table S1]; Hsp90 ( 9 ) ; SEC-2 ( 11 ) ; aldolase ( 20 ) ; glyceraldehyde-3-phosphate-dehydrogenase ( 14 ) ; protein with thioredoxin domain ( 52 ) ; and protein with glutathione S-transferase domain ( 43 ) . We identified 486 proteins from the M . incognita secretome after treatment using a protein FDR of 0 . 4% ( Figure 2 ) . These include all seven proteins reported by Jaubert et al . ( 2002 ) , indicating that our results both confirm and extend previous studies . The majority of the proteins ( 311; 64% ) were identified by the detection of 2 or more peptides . Of the 175 proteins identified by only one peptide , some were previously shown to be secreted . Proteins identified by several MS/MS spectra but only one peptide have been manually validated and the spectra are summarized in Table S2 . To serve as a control for potential contamination of the secretome by cellular debris , we examined proteins extracted from intact nematodes . Visual inspection of nematode preparations did not reveal any signs of damage or debris . We compared the relative abundance of proteins in the secretome to their abundance in extracts from intact nematodes . This revealed that many ( 19% ) of the secreted proteins are highly abundant in intact nematodes; these were removed from consideration out of concern that they may be contaminants , even though they were not observed in the water control . The normalized spectrum count ratio of each protein ( secretome/whole nematode proteome ) was used to calculate secretome enrichment . Most of the proteins identified in the solution bathing treated nematodes ( i . e . , the secretome ) were significantly less abundant or absent in the proteome of whole nematodes , providing further evidence that they are indeed secreted ( Table S1 , column 5 ) . Approximately 81% ( 394 ) of the secreted proteins are enriched and 60% ( 288 ) are at least 2-fold more abundant in the M . incognita secretome ( Figure 3 and Table S1 , column 5 ) . Due to the relatively large size of the SvGs and the number of dense granules in them , it would not be surprising to find secreted proteins in the whole nematode extract . The remaining 19% ( 92 ) of non-enriched proteins ( e . g . , actin ) may in fact be secreted , but to be conservative we do not consider them further . High-throughput nano-LC ESI MS/MS depends upon protein databases and is most useful when the entire annotated genome sequence of an organism is available . However , with the proliferation of genome projects , adequate sequence information has become available to enable protein identification using databases from other species . We used two M . incognita cDNA sequence databases with sequence databases from all nematodes and plants ( Figure 1 and Table S3 ) . Nearly all of the secreted proteins ( 481; 99% ) were identified by reference to the nematode protein sequences ( Figure 2 ) . Approximately half of the proteins ( 235; 48% ) were identified both by M . incognita sequence and by sequence from other nematodes . Only 20% ( 95 ) were identified by orthologous nematode sequence alone and 31% ( 151 ) from M . incognita sequence alone . A total 69% of the M . incognita secretome could have been identified without reference to the M . incognita DNA sequences ( Figure 2 ) . Table S4 shows full-length proteins with identified peptides derived from searching the M . incognita sequence database . Comparison of our observations with published reports of proteins secreted by M . incognita revealed extensive overlap and , in addition , we identified orthologs of proteins that are secreted by other parasitic nematodes ( Table S5 ) . Among the 10 most abundant proteins in our data ( Table S1 ) , 14-3-3b protein and calreticulin were previously shown to be produced and secreted by the SvG of M . incognita [25] , [26] . Comparison of our M . incognita secretome with that from the parasitic helminth , Brugia malayi , reveals significant overlap [27] . Of the 80 proteins known to be secreted by B . malayi , 26 are also secreted by M . incognita ( Table S5 ) . This conserved group includes proteins involved in detoxification ( e . g . SODs ) , cytosolic stress response ( e . g . 14-3-3-like proteins ) , cytosolic energy metabolism ( e . g . a triose phosphate isomerase ) , structure ( e . g . actin ) , protein turnover or folding ( e . g . ubiquitin-like proteins SMT3 and protein disulfide isomerases PDI ) , protease inhibitors ( e . g . Cystatin-type Cysteine Protease Inhibitor CPI-2 ) , and two transthyretin-like family proteins ( TTLs ) . The discovery of effectors from nematodes has lagged behind progress made with bacterial and oomycete pathogens , but recently phytopathogenic nematode effectors have been reported . We re-examined our mass spectra using sequence from members of the SPRYSEC protein family , which includes effectors from G . rostochiensis [28] , [29] . We also searched for Cg-1 , an M . incognita candidate effector gene acting in the Mi1 . 2 resistance pathway [30] , and for MAP-1 , a putative avirulence protein produced by amphids [31] . We did not identify peptides corresponding to any of these proteins in the M . incognita secretome nor in the extract of intact nematodes . We searched our mass spectra for peptides from proteins secreted by M . chitwoodi but could find none [19] . However , by doing a BLASTP search using proteins identified in our study , we were able to show that 4 proteins secreted by M . chitwoodi are also in the M . incognita secretome ( cysteine protease , beta-1 , 4-endoglucanase , VAP-1 and pectate lyase ) . The reason we initially missed them is because our search algorithms require exact amino acid sequence matches but the peptides identified in the M . incognita secretome have at least one amino acid difference compared to those deduced from M . chitwoodi ESTs . Using the euKaryotic Orthologous Groups ( KOGs ) classification scheme to annotate the secreted proteins [32] we found that 103 proteins catalyze post-translational modifications , protein turn-over or chaperone functions; 93 participate in protein synthesis or secretion; 88 trigger metabolic reactions; 48 interact with nucleic acid ( DNA or RNA ) ; 25 are involved in signal transduction and 33 interact with actin or microtubules . We performed a BLASTP search for each protein to refine their annotations ( Table S1 , column 8 and 9 ) . We combined the KOGs and BLASTP results to classify the M . incognita secretome into 9 subfamilies ( Tables S1 and S6 ) : Proteins interacting with actin or microtubules ( 33 proteins , family 1 ) ; Proteins interacting with nucleic acids ( 48 proteins , family 2 ) ; Post-translational modification , protein turnover , and chaperone functions ( 103 proteins , family 3 ) ; Metabolism ( 88 proteins , family 4 ) ; Signal transduction ( 25 proteins , family 5 ) ; Protein synthesis and secretion ( 93 proteins , family 6 ) ; Detoxification ( 17 proteins , family 7 ) ; Cell wall modification enzymes ( 8 proteins , family 8 ) ; and Other ( 94 proteins , family 9 ) . Nematode infection causes gene expression changes in the plant cell [33] . These changes could be due to indirect effects , but there is evidence for secreted nematode proteins interacting directly with plant transcription factors ( reviewed in references [2] , [34] ) . This was first suggested when putative secreted factors were observed to have nuclear localization signals ( NLSs ) [18] . Later , an mRNA was identified from the esophageal gland of H . schachtii and the capacity of its expressed protein to interact in planta with two putative plant SCARECROW-like transcription factors was reported [35] . To determine whether the secreted proteins we observed could be targeted to the plant nucleus and could potentially modify plant gene expression , we searched for NLSs and DNA or chromatin interaction motifs . We found 66 proteins that meet one or both criteria: 26 proteins with an NLS motif and 40 additional proteins with putative nucleotide binding activity . Of these , 8 proteins are predicted to have both an NLS and a nucleotide binding activity ( Table S7 ) . We identified 5 secreted proteins present only in the plant protein sequence database ( Table S1 ) . Among them was LeMir , a protease inhibitor known to be upregulated in plants during nematode infection . Low molecular weight tomato root exudates were used to induce nematode secretion so we examined as a control the water medium without nematodes for proteins and peptides that could potentially diffuse across the membrane that separated root exudates from nematodes . No proteins were detected in gels by silver staining ( data not show ) but , using mass spectrometry , we identified 4 peptides derived from 3 plant proteins ( Table S8 ) . Only one protein overlapped with the nematode secretome ( remorin 1 ) ; we could not detect LeMir or any of the other plant homologs in the nematode secretome indicating that they were not contaminants . Earlier reports identified other proteins with putative horizontal gene transfer ( HGT ) origins . We confirmed that several of these are in the secretome , including two pectate lyases [36] , a cellulose binding protein [37] , and two beta-1 , 4-endoglucanases [38] , [39] . McCarter et al . , ( 2003 ) reported cases of potential HGT from microbes; we confirmed their existence in the secretome , including a Rhizobacterial homolog of nodL ( CL221Contig1_1 ) and a polygalacturonase ( 221104r1 . 1_1 ) [40] . Two other putative HGT candidates were identified: a conserved hypothetical protein from Trichomonas vaginalis ( MI00116 ) and a putative Type IV secretory pathway VirB6 component from Ehrlichia canis str . Jake ( CL1842Contig1_1 ) ( Table 1 ) . We localized mRNA corresponding to a subset of secreted proteins using in situ hybridization to J2 stage nematodes ( Figure 4 ) . As a positive control , we localized transcripts for two previously characterized secreted proteins from the SvG: beta-1 , 4 endoglucanase and calreticulin ( Figure 4D and G respectively ) . We tested and confirmed that the following members of the M . incognita secretome are also expressed specifically in the SvG: CL312Contig1_1 ( protein with unknown function ) ; CL5Contig2_1 ( SEC2 ) ; CL2552Contig1_1 ( Transthyretin-like family protein homolog ) ; CL321Contig1_1 ( Translationally-controlled tumor protein homolog ) ; CL480Contig2_1 ( triosephosphate isomerase homolog ) . A BLASTX search revealed that CL312Contig1_1 encodes for a C . elegans homolog ( E value 9E-06 ) that is predicted to be a membrane protein with unknown function . We also found a transcript that encodes a putative CDC48 protein ( contig CL1191Contig1_1 ) that is enriched in phasmid organs .
We identified several chaperones that may be involved in protein secretion: thioredoxin , glutathione peroxidases , cyclophilins , and protein disulfide isomerases ( PDIs ) . PDIs have also been found in the secretion of the nematode , Ostertagia ostertagi , where their overexpression increases the yield of secreted proteins [42] . PDIs participate in actin filament polymerization , gene expression , cell-to-cell interactions and in the regulation of receptor functions [43] , [44] . Cyclophilins are associated with protein trafficking , protein folding , chromatin remodeling , and chaperone activity [45] . Coaker et al . ( 2005 ) showed that the Pseudomonas syringae cysteine protease , AvrRpt2 , requires activation by a plant cyclophilin before it can cleave RIN4 [46] . It is possible that M . incognita secretes cyclophilins to activate its effectors . The correct folding of secreted nematode proteins may be necessary for infection . It has been shown previously that the AVR9 peptide elicitor of Cladosporium fulvum contains three disulfide bridges and that its correct folding depends on the redox state of the environment , with folding rates greatly increased in the presence of PDI [47] . If AVR9 is even partially reduced , it loses all activity , illustrating the importance of disulfide bridges . SvGs are known to secrete a beta-1 , 4-endoglucanase in planta [48] , [38] , [49] , as well as a pectinase [50] and an expansin [51] . We observed these and other cell wall degrading enzymes in the M . incognita secretome indicating that the nematode may use these enzymes for moving through the root or for assisting with plant cell wall remodeling during root knot formation . One of the earliest plant responses to infection is the production of reactive oxygen species ( ROS ) [52] . Based on our study , the M . incognita secretome contains detoxification enzymes that may be able to degrade ROSs . This could assist the nematode to establish a successful feeding site . It was previously reported that M . incognita secretes proteins which protect it from ROSs [53] . In plant tissues , SODs exist in three main families containing Cu and Zn , Mn , or Fe in their active site . We found two putative cytosolic CuZnSODs in the M . incognita secretome . A CuZnSOD was highly expressed and active in emergent symbiotic Rhizobium nodules of Lotus japonicus suggesting that this enzyme could play an important role in the early stages of symbiosis [54] . Taking this into consideration it is possible that the nematode enzyme may play a role in establishing compatibility with the plant cell . ROSs are also scavenged by ascorbate peroxidases , cytochrome C-peroxidases , catalases , thioredoxins and glutathione peroxidases [55] . Two glutathione peroxidases and one thioredoxin were observed in the M . incognita secretome , as were several glutathione S-transferases . Normally these enzymes are induced in plants by H2O2 , where they act as calcium-dependent cellular protectants [56] , so perhaps the nematode enzymes also provide protection from ROS-catalyzed damage . A similar mechanism has been observed in the maize pathogen , Ustilago maydis , which overcomes host redox defenses by sensing peroxide with Yap1 . Once activated , Yap1 induces U . maydis peroxidase gene expression , leading to the successful establishment of infection [57] . Therefore , it is possible that M . incognita may have evolved enzymes to control the global oxidative status of the plant cell as a way to increase its virulence . Two of the most obvious consequences of nematode infection are distortion of the plant cell-cycle and cytoskeleton , leading to the formation of giant cells and the characteristic root knot [58] , [59] . We identified several secreted proteins that could be targeted to the plant cell nucleus , where they could regulate gene expression resulting in some of the morphological changes observed . The target of nematode effectors to the plant nucleus was first suggested by the presence of putative secreted proteins with nuclear localization signals [18] . Later , a small , secreted peptide was identified that interacts in planta with two plant SCARECROW-like transcription factors [35] . We identified 66 secreted proteins with putative nuclear localization , DNA binding , or chromatin modification domains . These include two helicases , several RNA and DNA binding proteins , histones and the Nucleosome Assembly Protein , NAP-1 ( Table S7 ) . NAP proteins move histones into the nucleus , assist with nucleosome assembly , and modulate transcription [60] . Several secreted proteins were identified that could potentially regulate plant cell proliferation including a CDC48-like protein ( VCP/CDC48 ) , SKP1 , TCTP , NAC protein , and a CDPK . We confirmed by in situ hybridization that the corresponding mRNA of the CDC48-like protein is specifically expressed in the nematode phasmid ( Figure 4A ) . A previous study using Coomassie Brillant Blue G-250 revealed that phasmids secrete proteins that take up the stain [61] . Phasmids are specialized pairs of sensory organs found in the posterior lateral field of most nematodes . The function of phasmids remains unclear although a role as receptors for female sex pheromone was proposed for Scutellonema brachyurum [62] . Most plant parasitic nematodes have phasmids [6] . Baldwin ( 1985 ) identified two types of phasmids in the J2 stage of H . schachtii: a larger type that secretes and a smaller one that does not [63] . In proliferating cells of Arabidopsis , AtCDC48 is highly expressed , but it's down-regulated in most differentiated cell types [64] . CDC48/VCP/p97 in Zebrafish has been shown to induce cell proliferation [65] . Based on this discovered we can add phasmids to the set of organs that may play a role in nematode parasitism . S-phase kinase-associated protein 1 ( SKP1 ) is a key component of the SCF complex that provides ubiquitin-protein ligase activity required for cell cycle progression . Gao et al . , ( 2003 ) identified a SKP1 homolog in the dorsal gland of Heterodera glycines [66] . The SKP1 homologue identified in our study has a nuclear localization signal , and therefore could be potentially targeted to plant nuclei . Translationally-controlled tumor proteins ( TCTPs ) are highly conserved and are implicated in several different cellular processes including growth , cell cycle progression , malignant transformation , and protection of cells against stress and apoptosis [67] . TCTP proteins are expressed in rapidly growing plant organs , such as the apical meristem , suggesting a role in cell proliferation [68] . Overexpression of TCTP in cultured tobacco cells resulted in faster regeneration and the induction of more calli following Agrobacterium infection [69] . We found that the mRNA for secreted TCTP is enriched in the SvG of M . incognita ( Figure 4F ) suggesting that TCTP could be actively secreted into the host plant cell . We observed one Calcium-Dependent Protein Kinase ( CDPK ) and several CaM proteins in the M . incognita secretome . Using RNA interference , Ivashuta et al . ( 2005 ) showed that in Medicago truncatula , CDPK1 is essential for root hair formation and cell elongation [70] . Inactivation of CDPK1 results in significant diminution of Rhizobial and mycorrhizal symbiotic colonization [70] . The CDPK family and signaling pathways are conserved across the plant kingdom [71] , so nematodes may have developed the ability to control this central and ubiquitous element of plant development . We identified several secreted proteins with established or suggested roles in the virulence of parasites . Anand et al . ( 2007 ) [72] used virus-induced gene silencing and an in planta tumorigenesis assay to identify plant genes involved in Agrobacterium-mediated plant transformation . They identified several genes that were required to produce the crown gall phenotype; we identified homologs in the M . incognita secretome . Among them were SKP1 , actin or actin-binding proteins , and histones H3 , H2a , and H2b . Histone H2a is required for T-DNA integration [73] and histone H3 has also been implicated [72] . We found a homolog of the Nodulin protein , NodL , in the M . incognita secretome , which is similar to the nodulin-like proteins ( NLP ) required for Agrobacterium-mediated transformation [73] . Root knot nematodes induce cytoskeletal changes that closely resemble those induced by Nod proteins [74] . MtENOD11 is expressed early following both arbuscular mycorrhizal infection and Meloidogyne infection of Medicago [75] . Therefore , it is possible that root knot nematodes use a Nod-like pathway to initiate giant cell formation . We were surprised to observe that plant and animal metazoan parasites secrete a common set of proteins . For example , B . malayi and M . incognita both secrete transthyretin-like protein ( TLP or TTL ) , which is a member of a growing family of transthyretin ( TTR ) -related proteins ( TRPs ) . TRPs are related to the vertebrate transthyretin , an extracellular thyroid hormone carrier protein [76] . TRPs may represent the ancestor of the vertebrate thyroid hormone carriers [77] . We found in the M . incognita secretome a TTL and confirmed that its corresponding transcript is specifically expressed in the SvG of J2 stage nematodes ( Figure 4E ) . Therefore , we reason that this TTL homolog is secreted into the plant cell where it regulates growth . A plant TTL is known to interact with the brassinosteroid receptor kinase to control plant growth [78] . Both cysteine ( CPI-2 ) and aspartyl ( API-2 ) protease inhibitor ( PI ) family members were observed in the M . incognita secretome . The function of PIs in nematodes is to protect their intestine from dietary proteases [79] . In plants , endogenous PIs are active against all four classes of proteinase ( cysteine- , serine- , aspartyl- , and metallo- ) . PIs accumulate following wounding or herbivory and they may provide protection [80] . PIs have also been shown to regulate programmed cell death ( PCD ) . For example , synthetic peptide inhibitors of caspases could suppress PCD induced by a Ps . syringae infection of tobacco [81] . Recently a cystatin CPI-2 protease inhibitor was identified in B . malayi secretions and it was proposed to inhibit host proteases required for antigen processing and presentation [27] . The M . incognita secretome contained metallopeptidases , aminopeptidases , a cysteine proteinase , proteasome components , and proteins involved in ubiquitination . Secreted proteases could have two obvious functions: either the destruction of plant defense proteins or nutritional pre-digestion . Cysteine proteinases are involved in both the initiation and execution of the cell death program [82] and intriguingly we found two kinds of cysteine proteinase inhibitors . G . rostochiensis has been shown to secrete metalloproteases [11] , as have other phytopathogenic nematodes and free living nematodes; a role in the hatching process was proposed for the latter [83] , [11] . Nematode metalloproteases could catalyze protein degradation in planta to enable uptake of proteins that are otherwise too large [84] . We identified several ubiquitin proteins in the M . incognita secretome . Tytgat et al . , ( 2004 ) [85] identified a ubiquitin extension protein secreted from the dorsal pharyngeal gland of root cyst nematodes . The ability of pathogens to manipulate the ubiquitination-proteasome system of animal immune systems is known ( for a review see Loureiro and Ploegh . , 2006 ) [86] . The ubiquitin pathway is required for innate immunity in Arabidopsis [87] . We identified 94 proteins that we were unable to classify ( Table S6 ) . Among them we had shown that the transcripts of two genes , SEC2 and CL312Contig1_1 , are enriched in the SvG of J2 stage M . incognita ( Figure 4C and 4B respectively ) . We identified several proteins with a putative function but we were unable to discern a role for them in pathogenicity . One example is a triosephosphate isomerase ( TPI ) homolog that is highly secreted ( ratio secreted/whole = 10 . 35 ) ; the corresponding transcript ( CL480Contig2_1; Figure 4H ) is enriched in the SvG of J2 stage M . incognita . BLASP and KOG annotation revealed a putative function of this protein in metabolism . However , a similar TPI was also found in the fungal mammalian pathogen Paracoccidioides brasiliensis where TPI localized to the cell wall and cytoplasmic compartments [88] . The authors suggested that TPI is required for interaction between P . brasiliensis and the extracellular matrix and could be important for fungal adherence to and invasion of host cells . A similar function could be postulated for the M . incognita TPI since after the mobile J2 stage , the parasitic nematode is sedentary and is in close contact with plant tissue . The development of sensitive proteomics methods has allowed us to significantly expand the known secretome of M . incognita . A rich set of candidates has been found that can now be functionally evaluated . Conservation of protein sequence allowed us to search our mass spectra using sequence databases from other nematode species and plants . Nearly half ( 48% ) of our identifications from heterologous sequence databases were confirmed by matches to the limited M . incognita sequence that is publicly available , suggesting that proteomics can be useful even with nematodes for which no sequence information is available . As more M . incognita DNA sequence becomes available , we can probably identify additional proteins by re-searching our mass spectra . We confirmed that most secreted proteins are produced by esophageal glands and we found direct evidence for one secreted by phasmids [2] , [34] . Twenty-six proteins overlap between the M . incognita and B . malayi secretomes ( Table S5 ) . These include proteins with potential functions in parasitic behavior ( e . g . , TCTP; Cystatin CPI-2 ) . This remarkable conservation of sequence raises the possibility that plant and animal parasitic nematodes share conserved mechanisms of infection .
Meloidogyne incognita was propagated from greenhouse-grown tomato plants ( Solanum esculentum cv . Rutgers ) . After 8 weeks of infection , eggs were recovered from tomato plants by shaking M . incognita-infected roots in 1∶9 dilution of bleach for 3 min in a flask . Eggs were collected onto a 25 µm mesh and were then bleached twice for 10 minutes with a 1∶5 . 7 dilution of bleach supplemented with 0 . 02% Tween 20 . Eggs were rinsed four times with sterile ddH2O . Twenty million eggs were hatched at room temperature for 3 days in 10 mM Tris pH 7 . 0 with 300 mg/l carbenicillin ( hatching buffer ) , and juvenile 2 stage ( J2 ) worms were allowed to crawl though five Kimwipe tissue layers into the same hatching buffer . Freshly hatched J2s were washed several times in sterile water and then collected on 8 µm sieves . Tomato seeds ( Solanum esculentum cv . Rutgers ) were placed above a plastic cylinder filled with cotton fiber and placed into an aerated hydroponics vessel constructed from a 2-liter flask . Hydroponic vessels were supplied with 250 ml sterilized solution of 0 . 5× Gamborg media basal salts medium complemented with 1× Gamborg vitamins , 0 . 5% sucrose and 200 mg/l carbenicillin . Tomato plants were maintained under a 16-h photoperiod for 6 weeks and root media was collected and filtered through a 0 . 22 µm syringe filter to give the “hydroponic tomato root culture solution” . Hatched J2s were stimulated for 4 hours by hydroponic tomato root culture solution separated from the nematodes by a 3 , 500 MW cutoff mini dialysis membrane ( Pierce , Rockford , USA ) . Then they were treated for 4 h with 0 . 4% resorcinol ( Sigma-Aldrich Chimie , St Quentin , France ) . Stylet secretions were filtered through a 0 . 22 µm syringe filter to remove nematodes . Secreted proteins were concentrated to ∼1 ml in a vacuum centrifuge at room temperature . Tris buffer was added to a final concentration of 20 mM ( pH 7 . 2 ) . Proteins were reduced and alkylated using 1 mM Tris ( 2-carboxyethyl ) phosphine ( Fisher , AC36383 ) at 65°C for 30 minutes and 2 . 5 mM iodoacetamide ( Fisher , AC12227 ) at 37°C in dark for 30 minutes , respectively . Proteins were then digested with 1 µg trypsin ( Roche , 03 708 969 001 ) at 37°C overnight . Whole M . incognita worms were lysed in 100 µL 2% ( w/v ) RapiGest ( Waters ) by sonicating in a Branson Sonifier 450 fitted with a high intensity cup horn ( Part No . 101-147-046 , Branson ) at 4°C for 2 minutes . Crude lysate was spun down at 16 , 100 g at 4°C for 5 min . Supernatant was collected and the pellet was discarded . RapiGest was diluted to 0 . 5% ( w/v ) by adding 300 µL of 20 mM Tris . Proteins were reduced and alkylated as described above . Protein concentration was measured using a Bradford assay . Protein ( 400 µg ) was digested with 10 µg trypsin ( Roche , 03 708 969 001 ) at 37°C overnight . TFA ( 0 . 5% v/v ) was added to each sample to a final pH of 1 . 8 to precipitate RapiGest after digestion . Samples were incubated at 4°C overnight and then centrifuged at 16 , 100 g at 4°C for 15 minutes . Supernatants were collected and centrifuged through a 0 . 22 µM filter to clear any solid particles . An Agilent 1100 HPLC system ( Agilent Technologies , Wilmington , DE ) delivered a flow rate of 300 nL min−1 to a 3-phase capillary chromatography column through a splitter . Using a custom pressure cell , 5 µm Zorbax SB-C18 ( Agilent ) was packed into fused silica capillary tubing ( 200 µm ID , 360 µm OD , 20 cm long ) to form the first dimension reverse phase column ( RP1 ) . A 5 cm-long strong cation exchange ( SCX ) column packed with 5 µm PolySulfoethyl ( PolyLC ) was connected to RP1 using a zero dead volume 1 µm filter ( Upchurch , M548 ) attached to the exit of the RP1 column . A fused silica capillary ( 100 µm ID , 360 µm OD , 20 cm long ) packed with 5 µm Zorbax SB-C18 ( Agilent ) was connected to SCX as the analytical column ( RP2 ) . The electro-spray tip of the fused silica tubing was pulled to a sharp tip with the inner diameter smaller than 1 µm using a laser puller ( Sutter P-2000 ) . The peptide mixtures were loaded onto the RP1 column using the custom pressure cell . Columns were not re-used . Peptides were first eluted from the RP1 column to the SCX column using a 0 to 80% acetonitrile gradient for 150 minutes . The peptides were then fractionated by the SCX column using a series of salt gradients ( from 10 mM to 1 M ammonium acetate for 20 minutes ) , followed by high resolution reverse phase separation using an acetonitrile gradient of 0 to 80% for 120 minutes . To avoid sample carry-over and keep good reproducibility , a new set of three columns with the same length was used for each sample . Spectra were acquired on LTQ linear ion trap tandem mass spectrometers ( Thermo Electron Corporation , San Jose , CA ) employing automated , data-dependent acquisition . The mass spectrometer was operated in positive ion mode with a source temperature of 150°C . As a final fractionation step , gas phase separation in the ion trap was employed to separate the peptides into 3 mass classes prior to scanning; the full MS scan range was divided into 3 smaller scan ranges ( 300–800 , 800–1100 , and 1100–2000 m/z ) to improve dynamic range . Each MS scan was followed by 4 MS/MS scans of the most intense ions from the parent MS scan . A dynamic exclusion of 1 minute was used to improve the duty cycle . MS/MS spectra were collected for secreted and whole M . incognita proteins ( 2 , 904 , 233 and 947 , 474 spectra , respectively ) . Raw data were extracted and searched using Spectrum Mill ( Agilent , version A . 03 . 02 ) . MS/MS spectra with a sequence tag length of 1 or less were discarded . MS/MS spectra were searched against the protein databases ( Table S3 ) . The enzyme parameter was limited to full tryptic peptides with a maximum mis-cleavage of 1 . All other search parameters were set to Spectrum Mill's default settings ( carbamidomethylation of cysteines , +/−2 . 5 Da for precursor ions , +/−0 . 7 Da for fragment ions , and a minimum matched peak intensity of 50% ) . To eliminate redundant protein identifications , proteins with one or more shared peptides were grouped . The numbers of proteins we report in this paper are protein group numbers . A concatenated forward-reverse database was constructed to calculate the in situ false discovery rate ( FDR ) [22] . We used an identification filtering criteria of 0 . 1% FDR at the peptide level for every search . A total of 486 secreted proteins from the forward database were identified , while 2 proteins ( 0 . 4% protein FDR ) from the reverse database were identified . Spectrum counting was used to determine the relative protein amounts in the secretome and the extract from intact nematodes . The number of valid MS/MS spectra from each protein was normalized to the total MS/MS spectra number of each dataset . The normalized spectrum count ratio of each protein ( secretome/intact nematode proteome ) was used to evaluate whether the protein was enriched in the secretome . The data associated with this manuscript may be downloaded from ProteomeCommons . org Tranche , http://tranche . proteomecommons . org , using the following hash ( without the quotes ) : “FXMi2Tyve1I0DfzhT9FN17TmpNTTDiggs7Njjoh7MYMouHYIx+xUoDILMXFl17RZrVjueXuCZc5c3005l9fdKISeVUEAAAAAAAB0ug = = ” . While this paper was under review , Abad et al . [89] reported the draft genome sequence of M . incognita . The 9 , 538 contigs resulting from the M . incognita genome assembly and annotation were deposited in the EMBL/Genbank/DDBJ databases under accession numbers CABB01000001–CABB01009538 for release at a future date . When these contigs become publicly available , further bioinformatics analysis of our mass spectra can be conducted to search for additional secreted proteins . Identified protein sequences were BLASTed against the non-redundant database at NCBI ( http://www . ncbi . nlm . nih . gov/ ) . euKaryotic Orthologous Group ( KOG ) annotations were assigned based on sequence similarity searches against the KOG annotated proteins ( http://www . ncbi . nlm . nih . gov/COG/grace/kognitor . html ) . Putative nuclear function was assigned based on homologous proteins found using BLASTP or on the identification of a Nuclear Localization Site ( NLS ) . The NLS search was performed using the PredictNLS search engine available at http://cubic . bioc . columbia . edu/predictNLS/ [90] . In situ hybridizations were performed on freshly hatched J2s as described in Rosso et al . [38] . Briefly , freshly hatched J2s were fixed in 2% paraformaldehyde for 16 h at 4°C and 4 h at room temperature . Nematodes were cut into sections and permeabilized with proteinase K , acetone , and methanol . The sections were then hybridized at 45°C with the sense or antisense probe . Clone and primers are listed in Table S9 . | Parasitic nematodes are microscopic worms that cause major diseases of plants , animals , and humans . Infection is associated with secretion of proteins by the parasite; these proteins suppress the immune system and cause other changes to host cells that are required for infection . Identification of secreted proteins has been difficult because they are released only in trace amounts . We have developed very sensitive methods that enabled the discovery of 486 proteins secreted by the root knot nematode , Meloidogyne incognita; prior to this , only a handful of secreted proteins were known . Several secreted proteins appear to mimic normal plant proteins , and they may participate in the process by which the nematode hijacks the plant cell for its own purposes . Meloidogyne species infect many crops , including corn , soybean , cotton , rice , tomato , carrots , alfalfa , and tobacco . The discovery of these secreted proteins could lead to new methods for protecting these important crops from nematode damage . We observed that the secretome of the human pathogen , Brugia malayi , overlaps that of M . incognita , suggesting a common parasitic behavior between pathogens of plants and animals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant-biotic",
"interactions",
"pathology/molecular",
"pathology"
] | 2008 | Direct Identification of the Meloidogyne incognita Secretome Reveals Proteins with Host Cell Reprogramming Potential |
In HIV-infected individuals receiving suppressive antiretroviral therapy , the virus persists indefinitely in a reservoir of latently infected cells . The proliferation of these cells may contribute to the stability of the reservoir and thus to the lifelong persistence of HIV-1 in infected individuals . Because the HIV-1 replication process is highly error-prone , the detection of identical viral genomes in distinct host cells provides evidence for the clonal expansion of infected cells . We evaluated alignments of unique , near-full-length HIV-1 sequences to determine the relationship between clonality in a short region and clonality in the full genome . Although it is common to amplify and sequence short , subgenomic regions of the viral genome for phylogenetic analysis , we show that sequence identity of these amplicons does not guarantee clonality across the full viral genome . We show that although longer amplicons capture more diversity , no subgenomic region can recapitulate the diversity of full viral genomes . Consequently , some identical subgenomic amplicons should be expected even from the analysis of completely unique viral genomes , and the presence of identical amplicons alone is not proof of clonally expanded HIV-1 . We present a method for evaluating evidence of clonal expansion in the context of these findings .
The HIV-1 virion carries two copies of a 9 . 7 kb RNA viral genome , which is reverse transcribed to DNA and integrated into the genome of a host cell during infection . Although combination antiretroviral therapy ( ART ) can suppress HIV-1 plasma viremia indefinitely to a level below the clinical limit of detection , the virus persists for decades in a latent reservoir composed of resting memory CD4+ T cells carrying integrated viral genomes , known as proviruses [1–4] . The development , composition , and plasticity of this latent reservoir , which presents a major barrier to the cure of HIV-1 infection , are all active areas of investigation [5 , 6] . An emerging body of research identifies the proliferation of latently infected CD4+ T cells as a possible mechanism for the persistence—and perhaps expansion—of the latent reservoir . Because viral replication is a low-fidelity process , expansion of the reservoir through cellular proliferation can be distinguished from expansion through de novo infection events by the presence of identical HIV-1 genomes with identical sites of integration into the host cell genome in distinct cells [7–9] . In part due to the substantial diversity of HIV-1 even within a single infected individual , generating full-length sequences of individual proviruses can be prohibitively expensive and labor intensive . Instead , it is common to sequence short PCR amplicons covering less than 3000 base pairs [7 , 8 , 10–18] . While these smaller , subgenomic amplicons contain sufficient sequence information to inform phylogenetic analysis , they do not capture the total diversity–in this case , defined as the proportion of non-clonal sequences—present in full-length viral genomes from the same sample . On the contrary , the sequence diversity in a subgenomic region is best understood as a minimum estimate of the total sequence diversity in the sample analyzed . Without comparing full-length HIV-1 genomes , it is impossible to determine whether two proviruses with identical sequence over a subgenomic amplicon are also identical over the remainder of the viral genome . This inherent overestimation of viral clonality when comparing short , subgenomic sequences is of critical importance in the investigation of clonal expansion of latently infected cells as a mechanism of HIV-1 persistence . A number of studies have identified identical viral sequences in independent samples from a single subject [7 , 8 , 13 , 17–20] . These identical sequences may reflect the expansion of latently infected CD4+ T cells in vivo . However , when the identical sequences analyzed cover only a short fragment of the viral genome , they may also represent distinct infection events with viral genomes that happen to differ only in areas of the genome that were not analyzed . This distinction underscores the importance of understanding the relationship between the sequence diversity in subgenomic PCR amplicons and that of full viral genomes . The goal of this study was to identify which short PCR amplicons—if any—capture the total diversity of full-length genomes in a sample . We analyzed near-full-length HIV-1 sequences available from previous studies; importantly , we specifically characterized intra-subject diversity . We considered how length , genomic position , and sample type contribute to the likelihood that a given subgenomic amplicon will include enough information to differentiate unique , full-length HIV-1 sequences . Given these findings , we characterized and evaluated eight PCR primer sets used in previously published studies of HIV-1 diversity for their ability to differentiate full-length viral sequences . We showed that subgenomic sequences are contextualized by the primer set used to generate them , and we present here a strategy for the evaluation of sequence data in the context of specific primer sets .
We analyzed data sets containing between 5 and 121 ( mean = 20 . 7 , median = 9 ) unique , near-full-length HIV -1 sequences from a total of 31 subjects . None of the sequences analyzed cover the 5’ LTR , and all sequences are fully characterized at a minimum from positions 2000 through 8000 of the HXB2 reference genome . These sequences represent five different sample types: 1 ) proviral DNA from the resting CD4+ T cells of subjects who initiated suppressive ART during acute HIV-1 infection , designated “Acute treated–DNA”; 2 ) proviral DNA from the resting CD4+ T cells of subjects who initiated suppressive ART during unspecified stages of chronic HIV-1 infection , designated “Chronic treated–DNA”; 3 ) proviral DNA isolated from quantitative viral outgrowth assay ( VOA ) [21] wells negative for p24 antigen , representing resting CD4+ T cells that were not induced to produce replication-competent virus after stimulation with phytohemagglutinin , designated “VOA–DNA”; 4 ) genomic viral RNA isolated from the plasma of viremic subjects over a series of longitudinal time points , designated “Longitudinal–RNA”; and 5 ) genomic viral RNA isolated from the plasma of subjects during acute HIV-1 infection , designated “Acute–RNA . ” The data sets were aligned to the HXB2 reference HIV-1 genome and processed to remove repeat sequences . Sequences were characterized as repeats if they were identical or differed only at ambiguously sequenced positions for all nucleotides sequenced . Thus , every full-genome sequence is unique in the alignments analyzed below . Details about sequence data sets and their sources are shown in Table 1 and S1 Table . We have shown that our results are not sensitive to the specific alignments; that is , the results are equivalent for different but equally probable alignments of the same sequences ( S1 Fig ) . For this study , only alignments including at least five unique , near-full-length HIV-1 genome sequences from the same individual were considered . Because our goal was to characterize intra-subject diversity , we analyzed the alignments from each subject individually . Fig 1A shows a schematic of our analytic procedure . We considered a series of hypothetical primer pairs defining subgenomic PCR amplicons . Each of these primer sets was evaluated against each subject sequence alignment . We assumed that a sequence could be amplified by a primer set if the sequence aligned to the HXB2 reference genome without any gaps in either primer site . For each sequence alignment , we considered the amplicons that would be produced by a PCR with a given hypothetical primer set , discarding the sequences that would not be amplified by that primer set due to insertions or deletions overlapping the primer binding sites . We chose to consider only PCR-amplifiable sequences to maximize the practical value of our metric for the analysis of sequence data sets generated using PCR-based protocols . We defined the clonal prediction score ( CPS ) of the primer set as the number of unique amplicons produced , divided by the total number of amplicons produced , and multiplied by 100 ( Fig 1A ) . In words , the CPS of a primer set with respect to an alignment is defined as the percentage of sequences in the alignment that would be correctly identified as unique using only the sequence region amplified by those primers . Each full genome in these alignments is unique , and primer sets that produce a unique amplicon for every amplifiable sequence in the alignment have a CPS of 100 . These parameters define a maximal CPS of 100 and a minimum possible CPS of 100/N , where N is the number of amplified sequences . When a primer set will not amplify any of the sequences in an alignment , the CPS is undefined for that primer set with respect to that alignment . Undefined CPS values are excluded from the figures described below . Importantly , the precision of the CPS is limited by the number of sequences amplified by a primer set . For example , a primer set that correctly distinguishes 5 of 5 amplicons would have the same perfect CPS = 100 as a primer set that correctly distinguishes 100 of 100 amplicons . If a new unique sequence were added to the alignment and incorrectly identified as clonal , the CPS values in these two cases would change to 83 and 99 , respectively . For this reason , the empirically perfect CPS = 100 may be more precisely described as CPS > 100*N/ ( N+1 ) . This example emphasizes why alignments containing more sequences lead to greater precision in CPS values . For clarity , we have plotted maximal CPS values at CPS = 100 in the figures described below . To investigate how CPS varies across the HIV-1 genome and across infected individuals , we calculated CPS values for hypothetical 1 kb amplicons spanning the HIV-1 genome at 10 bp intervals . These amplicons were defined by hypothetical 10 bp forward and reverse primers , and we have shown that the results of the following analysis are not sensitive to the choice of primer length ( S2 Fig ) . While the amplicons defined by these hypothetical primers have a length of 1 kb in the HXB2 reference genome , they may have different lengths in sequences containing insertions or deletions between the primer sites . Along with sequence differences , length polymorphisms can be used to differentiate unique sequences . The results of this analysis for six Acute treated–DNA samples ( Table 1 ) are shown in Fig 1B . There was substantial variation among subjects , suggesting that a primer set optimized to differentiate sequences in one sample may not be optimal in another sample from a different subject . For subject 2453 , almost every 1 kb amplicon contained sufficient variation to differentiate all amplified sequences , but the other five subjects had CPS values below 100 for amplicons spanning large portions of the genome . For subject 2521 , less than one percent of all possible 1 kb amplicons had a perfect CPS of 100 . That is , almost every 1 kb amplicon would incorrectly classify unique genomic sequences from subject 2521 as identical . There is no relationship between these patterns and the number of sequences analyzed for each subject ( S3 Fig and S1 Table ) . Plots of CPS across the genome for individual Chronic treated–DNA , VOA–DNA , Longitudinal–RNA , and Acute–RNA samples ( Table 1 ) are shown in S3 Fig . The perfect CPS of 100 is much more common for proviral DNA than for plasma RNA; in Chronic treated–DNA and VOA–DNA samples , more than half of subjects had perfect CPS values for every 1 kb amplicon across the viral genome . These perfect CPS values are often found at locations in the genome where one or more sequences in the alignment contain deletions and cannot be amplified . In these cases , the total number of sequences detected by a primer set ( the denominator in the CPS equation ) is lower , and there are fewer amplicons to be differentiated . Although alignments containing only a few sequences do not lead to bias or inaccuracy in the CPS , the precision of the CPS calculation is correlated with the number of sequences in the alignment being analyzed . In the compilation of our data set , we chose to include only subjects for whom at least five unique , near-full-length sequences had been characterized , but the precision of these results would be improved by the collection and inclusion of more sequences per subject . The dramatic variation in CPS among different subjects is seen in all sample types assayed except Longitudinal–RNA ( S3 Fig ) . For Longitudinal–RNA samples , CPS is consistent between subjects and across the viral genome . This result reflects the high diversity of the HIV-1 quasispecies during chronic infection . To determine whether optimized primer sets for the identification of clonal HIV-1 sequences should be located in specific regions of the viral genome , we plotted the CPS for hypothetical 1 kb amplicons spanning the HIV-1 genome , averaged over all of the subjects within each sample group ( Fig 1C ) . The top plot in Fig 1C shows the average of the six plots in Fig 1B , and the other plots in Fig 1C show averages of the plots in S3 Fig . Fig 1D is an overlay of the five plots in Fig 1C . The purpose of these averaged scores was to determine whether any region of the viral genome yields consistently higher or lower CPS than other regions across different subjects or sample types . CPS values averaged over several subjects are relatively consistent across the viral genome for all five sample types evaluated . In the Acute treated–DNA and Chronic treated–DNA samples , CPS values appear slightly higher for the 5’ half of the genome than for the 3’ half . In contrast , the CPS for VOA–DNA samples is highest at the 3’ end of the genome . Importantly , these general trends are not representative of individual subjects , e . g . , Acute treated–DNA subject 3693 has a perfect 100 CPS only for amplicons toward the 3’ end of the viral genome ( Fig 1B ) . The Acute–RNA samples stand out as having lower CPS values across the viral genome than the other sample types ( Fig 1D ) , indicating that the sequences in these alignments differ at relatively few places in the genome . This finding is consistent with the biological characteristics of the Acute–RNA sample type; these sequences contain low genetic diversity because they represent plasma collected during acute HIV-1 infection , before the viral quasispecies has expanded and developed the sequence diversity characteristic of chronic infection . In most individuals , HIV-1 infection is initiated by a single transmitted founder virus that expands into a diverse quasispecies over the course of untreated infection [22 , 23] . Importantly , these results show that there is no optimal region of the genome best suited for differentiating unique sequences in a subject- or sample type-independent manner . To determine the effect of amplicon length on CPS , we calculated the CPS for hypothetical amplicons of different lengths spanning the HXB2 genome . We summarized the analysis of all amplicons of a given length in three ways , by taking the average , median , and minimum CPS over all amplicons of that length across the genome ( Fig 2A ) . For example , all of the points plotted in the top chart from Fig 1B are averaged together to yield the subject 2286 , 1000 bp data point in the Acute treated–DNA Average CPS plot in Fig 2A . Individual results for the 31 subjects are grouped by sample type . For every subject , regardless of sample type , longer amplicons yield higher CPS . This is unsurprising , as a longer amplicon should be more likely than a short amplicon to contain sequence diversity . However , the precise relationship between amplicon length and CPS varies dramatically by sample type . In proviral DNA samples , CPS increases with amplicon length but varies among subjects . For very short amplicons between 100 bp and 500 bp , average and median CPS can range from about 40 to a perfect 100 . For very large amplicons of 6 kb , the average CPS is greater than 80 and the median CPS is 100 for almost every proviral DNA sample . Again , there is no correlation between CPS and the number of amplicons in a sequence alignment ( Fig 2 and S1 Table ) . In DNA from subjects treated during chronic infection , the relationship between amplicon length and CPS is similar for three of five subjects . CPS values are high for the three Chronic treated–DNA samples that cluster together in Fig 2; all three subjects have a perfect CPS of 100 across the viral genome for amplicons as short as 1 kb . Subject CP03 is believed to have initiated therapy early in chronic infection , which may explain the similarity of that sample to the Acute treated–DNA samples . Subject CP10 illustrates an important limitation of any sequence-based analysis . Two of the sequences in the CP10 alignment differ by a single nucleotide . These sequences likely represent identical genomes that differ because of an error generated during PCR , but they were considered to be unique for this analysis . Because the error rate of PCR using high-fidelity polymerases is of a much lower magnitude than HIV-1 sequence diversity , our results overwhelmingly reflect true viral diversity ( rather than technical diversity ) except in this exceptional case of PCR error in identical genomes . In plasma RNA samples , the relationship between amplicon length and CPS is direct and consistent across subjects . Unlike with the proviral DNA samples , the average and median CPS for plasma RNA can be lower than 20 for 100 bp amplicons . These low CPS values indicate alignments with many detectable sequences but insufficient sequence diversity to distinguish them using small amplicons . The relationship between amplicon length and CPS is direct and linear for Longitudinal–RNA samples . There is much less inter-subject variation in this relationship for Longitudinal–RNA than for any of the other sample types . Importantly , even the minimum CPS increases predictably with amplicon length for these samples . For all Longitudinal–RNA samples , the average CPS is greater than 95 for 2 kb amplicons , and even the minimum CPS for a 2 kb amplicon is greater than 90 . This contrasts with the other sample types , where average CPS increases predictably with amplicon length but minimum CPS is variable . Fig 2A shows that most short PCR amplicons are limited in their capacity to distinguish unique HIV-1 genomes , and that the goal of a single short amplicon with a perfect CPS of 100 in a variety of subjects and sample types is unattainable . Instead , we evaluated how many amplicons achieve high—but not necessarily perfect—CPS values . Fig 2B shows the proportion of amplicons of a given length with CPS > 80 . The relationship between amplicon length and the frequency of amplicons with CPS > 80 is largely consistent with the results shown in Fig 2A but emphasizes the inter-subject differences in CPS patterns . Importantly , CPS is undefined for a primer set that will not amplify any of the sequences in a given alignment . The summary statistics in Fig 2 describe only the primer sets with defined CPS for each subject . In other words , the CPS values shown in Fig 2 represent only the primer sets for which at least one sequence in the alignment is detectable . We calculated PCR coverage , or the fraction of sequences in each alignment that would be detectable by PCR , averaged over all amplicons of a given length between HXB2 coordinates 2000 and 8000 ( Fig 3 ) . We chose this region because it is fully characterized for every sequence in our data set , and so the data in Fig 3 reflect the presence of internal deletions in the sequences rather than missing sequence data . As described above , we considered amplicons defined by hypothetical primer sets spanning the region at 10 bp intervals . The PCR coverage reflects the overall percentage of the viral genome lost to internal deletions in each sample . This value is consistent across subjects in plasma RNA samples ( Fig 3 ) , for all of which the proportion of detectable amplicons approaches 100% , especially with very large 6 kb amplicons . Plasma RNA represents viral genomes capable of producing all of the viral proteins necessary to generate a functional virion and therefore does not contain large internal deletions . In contrast , proviral DNA samples are much less likely to be detected by PCR because of the high frequency of large internal deletions in archived proviral genomes [24] . There is substantial variation in PCR coverage for proviral DNA samples among different subjects and across sample types . On average , Acute treated–DNA sample sequences are more likely to be detected by PCR than Chronic treated–DNA or VOA–DNA sample sequences . The results shown in Fig 3 have implications for primer design independent of CPS . In proviral DNA samples , the results of PCR-based analyses may vary dramatically with the choice of primers because different sequences in the sample can be intact or deleted in different regions of the viral genome; bias may be introduced by the specific selection of only the sequences that happen to be intact in a particular region . We calculated the CPS for eight primer sets used in published studies of HIV-1 diversity , evolution , and clonal expansion ( Table 2 ) with respect to the sequence alignments in our data set . The results , averaged over all subjects for each sample type , are presented in Fig 4A . Consistent with the results in Fig 2 , larger amplicons typically have higher CPS values , and this relationship is most distinct in plasma samples . Importantly , in most of the samples shown , none of the published primer sets have a perfect CPS of 100 . Consequently , in any analysis of sequences derived using these primers , some unique full-length genomes will be misrepresented as identical amplicons . The analysis of subgenomic amplicons to characterize HIV-1 genomes cannot provide definitive evidence of clonality in a sample , even when identical sequences are detected . The CPS for a primer set indicates how much false-positive clonality is found by that primer set in alignments of unique genomes . Thus the CPS for alignments of unique genomes can be understood as an estimate of the background signal expected when using a given primer set to evaluate clonality . We used the CPS of published primer sets to estimate the presence or absence of true clonality in published phylogenetic trees . This analysis is described in detail in the Methods section . In short , we plotted the relationship between the total number of sequences analyzed and the number of unique sequences detected ( Fig 4B and 4C ) . The CPS of a primer set is represented as a black line on these plots , and red dashed lines indicate the CPS plus or minus one standard deviation . We then plotted phylogenetic trees from six previously published studies [7 , 8 , 13 , 16–18] as points on the same axes . Each plot in Fig 4B and 4C shows a different primer set , and each point represents a single phylogenetic tree generated using that primer set . Fig 4B shows results for trees composed of proviral DNA samples , and Fig 4C shows trees composed of plasma RNA samples . Trees containing both RNA and DNA were separated and evaluated individually; these trees are represented by separate points in Fig 4B and Fig 4C . The distance between the lines ( CPS , background signal estimate ) and the points ( sample data ) demonstrate the amount of true clonality likely to be present in a sample . Importantly , the black lines in Fig 4B and 4C do not lie along the x = y axis . For every primer set defining a subgenomic amplicon , even in a sample composed entirely of unique HIV-1 genomes , some false-positive clonal amplicons are likely to be detected , causing the slope of the black line to deviate from 1 . The expected number of false positive identical amplicons ( y-axis ) increases proportionally with the total number of amplicons sequenced ( x-axis ) . Points that lie on the black lines are consistent with the null hypothesis that there are no clonally expanded HIV-1 genomes in the sample . Points that fall below the black line suggest clonal expansion , i . e . , the presence of more identical amplicons than would be expected in a sample composed of unique viral genomes . We did not see evidence of points plotted well above the black line , which would have indicated greater sequence diversity than was present in the alignments used to generate this model . This observation serves as validation of the data sets used to train our model . Studies in which trees do not include hypermutated sequences are marked with an asterisk . Hypermutated sequences are easily distinguishable even using short amplicons because of their tremendous diversity ( S4 Fig ) . For this reason , the removal of hypermutated or otherwise defective sequences from an alignment will artificially deflate the ratio of unique sequences to total sequences detected . In other words , for trees that have been curated to remove hypermutated sequences , the true location of the points plotted in Fig 4B and 4C is expected to fall closer to the black line . Due to the mechanisms of APOBEC-mediated hypermutation and HIV-1 reverse transcription , the rate of hypermutation is not constant across the viral genome [25 , 26]; the relationship between CPS and genome location in hypermutated sequences is shown in S4B Fig . The analysis shown in Fig 4 provides a method for evaluating both primer sets and sequence data . A primer set with high CPS and minimal variation in CPS across samples provides the most powerful background estimate to enable the confident interpretation of sequence data . And regardless of the primers used to generate sequences , phylogenetic trees and sequence alignments can only be understood properly when interpreted in the context of the CPS for the primer set used to generate them .
In this study , we defined the CPS as a metric for how well a subgenomic amplicon differentiates unique HIV-1 genomes . We calculated CPS values for hypothetical amplicons of varying sizes across the HIV-1 genome to investigate the contribution of size and location to the capacity of an amplicon to distinguish unique genomes . Finally , we calculated the CPS values for commonly used primer sets and used them to estimate the background level of clonality in phylogenetic trees generated with those primer sets . The primary goal of this study was to identify PCR amplicon ( s ) best suited to differentiate unique , full-length HIV-1 genomes . We found compelling evidence that no single , subgenomic amplicon will be sufficient to distinguish HIV-1 genomes across a variety of sample types or subjects . However , we evaluated eight previously published primer sets and found that of those , a 1 . 5 kb amplicon in the p6-gag-pro region seems to be the best compromise between coverage , practicality , and CPS , a metric which describes the capacity of an amplicon to correctly identify unique sequences as unique . We also identified a variety of best practices to maximize the validity of future studies of HIV-1 clonality in any sample type . Most importantly , researchers should always emphasize which region is being sequenced and explicitly consider the CPS of the amplicon ( s ) used . This context is essential for evaluating claims of clonal expansion , as it provides an estimate of the expected background signal against which sequencing results should be compared . We recommend against highlighting identical amplicons as evidence of full-genome clonality without comparing those results to an appropriate background estimate , and we emphasize that the only way to definitively demonstrate clonal expansion is to corroborate results with full-genome and integration site sequencing . Importantly , all of the analyses presented here assume that the individual amplicons sequenced were collected independently using methods specifically designed for single-genome sequencing . Many studies are confounded by PCR resampling [27]; all sequences analyzed using the methods described here should be collected independently . Furthermore , although some analyses call for curated sequence data , every full , uncurated data set should be made available . Phylogenetic trees presented to highlight clonally expanded populations should not be curated to remove replication-incompetent sequences . As described above , the removal of hypermutated sequences from a data set artificially inflates the proportion of clonal sequences in that data set . When evaluating alignments or trees against the background CPS of the amplicon sequenced , it is necessary to analyze every sequence collected . To aid researchers in the analysis of new data sets with different amplicons than those described here , we have published a Web tool available at http://silicianolab . johnshopkins . edu/cps . This tool computes the CPS for user-specified primer sets and performs a comparison with user-entered values to characterize phylogenetic trees or alignments of amplicons in the context of appropriate background signal estimates . Whether using this CPS analysis or any other method , it is impossible to conclusively prove or disprove the clonality of full-length viral genomes using only the sequence of a subgenomic amplicon . However , because information about the region amplified is available , it is essential to consider this context when interpreting data . These implications are no less relevant for phylogenetic studies in fields beyond HIV persistence; whenever the results and interpretation of a phylogenetic analysis may be impacted by the choice of amplicon sequenced , it is critical to consider that impact explicitly in the evaluation and presentation of data . This CPS analysis is intended not as a definitive arbiter of full-sequence clonality but as a tool to quantify the context provided by primer location and inform the interpretation of sequence data . Some important limitations of this study suggest additional considerations for future research . We have considered amplicon sequences on a binary scale; either sequences are correctly identified as unique , or they are not . We have not considered the distribution of sequence clonality . For example , in an amplicon alignment with twenty total sequences and only ten sequences correctly identified as unique , we did not distinguish between cases where the ten clonal sequences are identical to each other and cases where they represent duplicates of several other sequences . The case where all clonal sequences are identical may or may not be more likely to represent full-genome clonality; the quantification of that likelihood is beyond the scope of this study . In this study , we took advantage of sequence alignments representing a variety of sample types . Importantly , the results of our analyses often varied dramatically , especially between plasma RNA HIV-1 genomes , which are typically intact , and proviral DNA genomes , which often contain large internal deletions . Although the CPS of a primer set is clearly correlated across sample types , the implications of our analysis for different sample types than those evaluated in this study are imprecise . We have also shown that even within a sample type , the variation in CPS among subjects can be extreme . Additionally , especially in the case of proviral DNA sequences containing deletions , the methods used to identify full-length genomic HIV-1 sequences in previous studies may have been more efficient for some sequences than for others . Any bias in the nature of the sequences used to define our CPS model may be reflected in our results . For these reasons , the model presented here and any future calculations of CPS would benefit from the inclusion of more full-length or near-full-length HIV-1 genome sequences and sample types collected in the future . In summary , identical HIV-1 sequence fragments must be validated to demonstrate full-genome clonality conclusively . This validation can be achieved in a variety of ways . When technical constraints permit , additional regions of the genome can be sequenced . To demonstrate the clonality of proviral DNA sequences , the best validation is to sequence the associated integration site . Unvalidated sequences may be described as “identical throughout the region sequenced” but without further corroboration should not be described as clonally expanded . The web tool available at http://silicianolab . johnshopkins . edu/cps can be used to calculate CPS for the evaluation of new data or primer sets not described in this study .
The online tool to calculate CPS can be found at http://silicianolab . johnshopkins . edu/cps . The tool is written in JavaScript , and the code is accessible at https://github . com/gitliver/HIVCPS . All sequences used for CPS analysis are available from GenBank . The CPS values in Fig 4A are similar across all three DNA sample types . To calculate the expected CPS of a given primer set as used to characterize a proviral DNA sample , we averaged the CPS of that primer set for the three DNA sample types . To calculate the expected CPS of a given primer set used to characterize a plasma RNA sample , we used the average CPS from the Longitudinal–RNA samples ( Fig 4A ) . We chose not to include the Acute–RNA samples in this average because plasma virus from acutely infected individuals contains much less diversity than any other sample type , and its relevance to the analysis of other samples is minimal . The appearance of the Acute–RNA sample type as an outlier is evident in Figs 1 , 2 and 4A . For each primer set in Table 2 , we used the average CPS values for proviral DNA and plasma RNA samples to calculate the proportion of false-positive clonal sequences that should be expected from sequencing subgenomic HIV-1 RNA or DNA amplicons . The relationship between the total number of amplicons sequenced and the expected number of correctly identified unique amplicons detected for each primer set is plotted as a black line in Fig 4B ( proviral DNA ) and 4c ( plasma RNA ) . The slope of this line is the CPS divided by 100 . The dashed red lines give a confidence interval representing one standard deviation in CPS values . We counted the total number of sequences and the number of unique sequences in 53 different phylogenetic trees from six previously published studies [7 , 8 , 13 , 16–18] . The color and shape of the points indicate the study in which each tree was published . | Although antiretroviral therapy effectively blocks HIV-1 replication , the virus persists indefinitely in a reservoir of latently infected cells . This reservoir is a major barrier to HIV-1 cure . Recent studies have identified the proliferation of latently infected cells as a mechanism that may contribute to the lifelong persistence of HIV-1 . In contrast with cellular proliferation , viral replication is highly error-prone; therefore , the detection of identical viral genomes in distinct host cells provides evidence that those cells are the progeny of an infected cell that underwent clonal expansion . For this reason , the accurate and reliable identification of identical HIV-1 genomes derived from distinct cells is critical for understanding HIV-1 persistence and advancing cure research . Studies of HIV-1 clonal expansion often present clonality of short , subgenomic sequences as evidence for the clonality of full viral genomes . In this study , we quantified the relationship between sequence identity in short regions and sequence identity of near-full-length genomes to demonstrate that no subgenomic region completely captures the diversity of the full viral genome . Consequently , identical subgenomic sequences are not proof of identical full-length HIV-1 genomes . In the context of these findings , we developed a method to evaluate identical subgenomic sequences as possible evidence for clonal expansion . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"retroviruses",
"viruses",
"immunodeficiency",
"viruses",
"rna",
"viruses",
"viral",
"genome",
"phylogenetic",
"analysis",
"mole... | 2016 | Evaluating Clonal Expansion of HIV-Infected Cells: Optimization of PCR Strategies to Predict Clonality |
Cellular gene expression measurements contain regulatory information that can be used to discover novel network relationships . Here , we present a new algorithm for network reconstruction powered by the adaptive lasso , a theoretically and empirically well-behaved method for selecting the regulatory features of a network . Any algorithms designed for network discovery that make use of directed probabilistic graphs require perturbations , produced by either experiments or naturally occurring genetic variation , to successfully infer unique regulatory relationships from gene expression data . Our approach makes use of appropriately selected cis-expression Quantitative Trait Loci ( cis-eQTL ) , which provide a sufficient set of independent perturbations for maximum network resolution . We compare the performance of our network reconstruction algorithm to four other approaches: the PC-algorithm , QTLnet , the QDG algorithm , and the NEO algorithm , all of which have been used to reconstruct directed networks among phenotypes leveraging QTL . We show that the adaptive lasso can outperform these algorithms for networks of ten genes and ten cis-eQTL , and is competitive with the QDG algorithm for networks with thirty genes and thirty cis-eQTL , with rich topologies and hundreds of samples . Using this novel approach , we identify unique sets of directed relationships in Saccharomyces cerevisiae when analyzing genome-wide gene expression data for an intercross between a wild strain and a lab strain . We recover novel putative network relationships between a tyrosine biosynthesis gene ( TYR1 ) , and genes involved in endocytosis ( RCY1 ) , the spindle checkpoint ( BUB2 ) , sulfonate catabolism ( JLP1 ) , and cell-cell communication ( PRM7 ) . Our algorithm provides a synthesis of feature selection methods and graphical model theory that has the potential to reveal new directed regulatory relationships from the analysis of population level genetic and gene expression data .
Network analyses are increasingly applied to genome-wide gene expression data to infer regulatory relationships among genes and to understand the basis of complex disease [1] , [2] . Probabilistic graphical techniques , which model genes as nodes and the conditional dependencies among genes as edges , are among the most frequently applied methods for this purpose . A diversity of such approaches have been proposed including Bayesian networks [3]–[5] , undirected networks [6]–[8] , and directed cyclic networks [9]–[11] . The popularity of these methods derives , in part , from the structure of these models that is well suited to algorithm development and because the network representation of these models can be used to construct specific biological hypotheses about the processes governing the activity of genes in a system [3] . As an example of this latter property , genes connected by an edge may indicate ( at least ) one of the genes is regulated by the other . In graphical network inference , a theoretical principle that is now well appreciated [5] , [10]–[17] is that ‘perturbations’ of the network can be leveraged to reduce the set of possible networks that can equivalently explain gene expression . In fact , since equivalent models can indicate conflicting regulatory relationships , perturbations are often necessary to extract regulatory relationships with any confidence . If the perturbations are controlled ( e . g . knockouts of single genes ) , then a network among n genes can be recovered very efficiently with n knockouts [12] . Alternatively , perturbations that arise from naturally segregating variants , or combinations of genetic variants produced from carefully designed crosses , can also be leveraged [5] , [10] , [11] , [13]–[19] . Perturbations of this type , caused by genetic polymorphisms in a population that alter the expression of genes across a population sample , are expression quantitative trait loci ( eQTL ) [15] . Despite the acknowledged importance of perturbations in network analysis , there has been little theoretical work concerning sets of perturbations that maximally limit the set of equivalent models for arbitrary directed networks . Limiting the set of equivalent models is of particular concern in cases where the true network has cyclic structure , where the set of statistically indistinguishable models may include drastically different topologies [20] . In this paper , we present theory concerning a minimally sufficient set of ( genetic ) perturbations to infer a maximally limited equivalent set of network architectures , which can subsequently be reconstructed using a single , convex optimization procedure . We demonstrate that for a specific type of network among both genes expression products and genotypes ( an interaction or conditional independence network [21] ) , when including an appropriate set of genetic perturbations for the genotypes , specifically locally occurring cis-eQTL [14] , the interaction network contains all the information necessary for directed network reconstruction . We can therefore estimate the regulatory relationships or features of a network directly from the interaction network with many different approaches [7] , [8] , [22]–[24] . Here , we use the adaptive lasso [25] , a convex optimization procedure , to efficiently solve this model selection problem . This approach allows us to avoid the reliance on computationally inefficient heuristics [3]–[5] , [10] , [11] , [16] , [18] , [19] with non-unique solutions , which can generate many possibly poor-fitting networks when considering sample sizes that are typical of experiments collecting genome-wide gene expression data . Our algorithm includes three steps . First , an association analysis is carried out to identify strong local ( cis-eQTL ) perturbations of gene expression . Second , we combine the gene expression data and genotypes for the cis-eQTL , and use an adaptive lasso regression procedure [8] , [25] to identify an interaction network [21] among gene expression products and cis-eQTL genotypes . The novel component of our algorithm is incorporated into this step , where we can immediately extract a unique , directed acyclic or cyclic network , given each gene in the network analysis has a unique cis-eQTL . Third , to ensure the edges in the interaction network correspond to the correct dependencies in the directed graph , we do a permutation test to ensure marginal independence between the cis-eQTL and the upstream gene based on the undirected edges recovered . We only use genetic perturbations that are cis-eQTL because of empirical evidence that local genetic polymorphism tends to have larger effects than trans-eQTL [26]–[28] , and are therefore statistically more likely to be linked to locally causal variants . If the true network is a directed cyclic graph and if one uses trans-eQTL to attempt to find the true model , there can still be a larger equivalence class of models , since there is no way to know which gene a trans-eQTL actually feeds into in a cyclic graph because of equivalence ( this is shown in the “Recovery” Theorem in the Methods ) . Our approach mirrors directed network inference approaches that seek to identify conditional independence and dependence relationships but avoids a computationally demanding step of iteratively testing for these relationships [11] , [20] , [29] , [30] . To test this algorithm , we explore performance for simulated data . Specifically , the simulations are designed to capture scenarios where the underlying network is relatively sparse , and the strength of both the cis-eQTL and regulatory relationships is strong enough to detect given a relatively small numbers of samples , on the order of the number of genes being tested . We investigated networks of modest size ( either 10 or 30 genes ) , since we wished to focus on cases where the set of genes being tested have strong cis-eQTL in linkage equilibrium , which in a typical eQTL genome-wide association study will be much smaller than the total number of genes being tested , [27] , [28] . As a benchmark , we compare the performance of our algorithm to the PC-algorithm [29] , [31] , the QDG algorithm [14] , the QTLnet algorithm [16] , and the NEO algorithm [18] . We find that our algorithm can outperform all of these approaches in terms of controlling the false-discovery rate , and having greater power ( given a large enough sample size ) for the recovery of directed acyclic graphs and directed cyclic graphs . To empirically assess our algorithm , we also analyze data from a well powered intercross study in yeast [27] . From this analysis , we identify 35 genes with strong , independent cis-eQTL , and leveraged these perturbations to identify novel interactions . While we analyze the data from an intercross , both the theoretical results as well as the algorithm itself can be applied to natural populations as well .
Biologically , our goal is to identify relationships between the expression of multiple genes , such as the case depicted in Figure 1 . In this figure we see that the expression level of Gene A has an effect on the expression level of Gene B , mediated through some biological process ( i . e . unobserved factors ) . Even though we do not directly observe all the factors involved in the regulatory interaction , we still want to be able to detect that there is a regulatory effect , including the relative magnitude , the presence , and direction of the effect . To resolve these relationships uniquely , we need perturbations of expression , which in this case arise from genetic polymorphisms affecting expression . Therefore , both gene expression and genotype data needs to be collected on the same set of individuals , for all genes of interest , as well as all genotypes that will possibly act as perturbations of expression . Overall , one can consider our model selection process as acting on the joint covariance between and within the gene expression products and genotypes identified as being strong QTL . In our algorithm we further focus on cis-eQTL , because of recent studies indicating that there are widespread genetic polymorphisms local ( i . e . cis ) to genes that cause significant changes in expression [26]–[28] . We want to identify the genes with strong cis-eQTL ( ) with linear effects on gene expression ( ) parametrized by genetic effect parameters ( ) , and then identify unique regulatory relationships among gene expression products parametrized by . For measured gene expression phenotypes and loci for which we have genotypes , the directed graphical model of the network has nodes and possible edges , representing possible regulatory relationships among the genes , and possible perturbation effects of loci ( eQTL ) on each of the expression phenotypes . Written in matrix notation , the network model for a sample of individuals can be represented as: ( 1 ) where is a matrix of gene expression measurements , is a matrix of regulatory effects , is a matrix of observed perturbations , is a matrix of genetic effect parameters , and , where is a diagonal matrix . Non-zero elements of and are edges representing regulatory relationships and eQTL effects , respectively , where the size of the parameter indicates the strength of the resulting relationship , as shown in Figure 1 . Versions of this model are used regularly in analysis of networks [3] , [8] , [10] when assuming that gene expression measurements are taken from independent and identically distributed ( iid ) samples , where the regulatory relationships can be approximated by a system of linear equations , and the distribution of expression traits across samples is well modeled with a multivariate normal distribution . Another common assumption we make use of in our algorithm is that most detectable eQTL effects will have a significant linear component , especially for cis-eQTL [27] , [28] , where the polymorphism has simple switch-like behavior , such as determining whether transcription of the gene is up or down regulated . A potential pitfall of modeling expression traits using directed networks of the type in Equation ( 1 ) is the problem of likelihood equivalence between models . Figure 2 presents a simple example that illustrates the problems raised by equivalence for network inference . In this example , the true model , which is a linear pathway between four genes , is indistinguishable from three other equivalent models . Each of these equivalent models has a very distinct implication for regulatory relationships among these genes but they are indistinguishable , regardless of the sample size . To be able to distinguish between these models , one needs to either collect time-course data to determine the temporal sequence in which regulation occurs [32] , or alternatively , perturb the expression level of these genes in some fashion . Our goal is to identify a unique network underlying the observed expression and genotype data , especially when the sample size is at most 1 , 000 ( a large , biologically realistic sample size ) . To accomplish this , in the Methods we prove a set of theorems to show that if each gene being considered has its own , unique eQTL , then one can go from the sample covariance among gene expression phenotypes and genotypes ( defined as S in the Methods , see Figure 3a ) , to the inverse covariance ( i . e . precision matrix or undirected network defined as in the Methods , see Figure 3b ) , then subsequently to a directed cyclic network underlying the expression data ( defined as , see Figure 3c ) , where the last step makes use of our “Recovery” Theorem . In the algorithm , we begin with a screening process to identify a set of expression traits with putative strong cis-eQTL ( Step 1 ) . We then make use of the adaptive lasso function for reconstruction of conditional independence networks ( i . e . the structure of the inverse covariance matrix , Figure 3b ) ( Step 2 ) to identify genes with strong induced dependencies among cis-eQTL genotypes and gene expression phenotypes and reconstruct the unique directed acyclic or cyclic network that is a result of these induced edges . Finally , for each putative strong induced dependency , we further filter the induced edges based on a permutation test ( Step 3 ) , to ensure marginal independence between the upstream gene and the downstream cis-eQTL: To benchmark the performance of our algorithm , we compared it to the PC-algorithm [29] , [31] , the QDG algorithm [11] , the QTLnet algorithm [16] , and the NEO algorithm [18] . The other previously proposed cyclic algorithms either do not scale well ( e . g . the approach of Li et al . [9] ) or have prohibitively complex implementations ( Richardson's cyclic recovery algorithm [20] or the algorithm of Liu et al . [10] ) . The PC-algorithm is designed to recover directed acyclic graphs using iterative tests of conditional dependence and independence , is a computationally efficient algorithm ( scales to thousands of genes for sparse networks ) , and has competitive performance with other directed acyclic graph reconstruction algorithms [29] , [36] . Additionally , the PC-algorithm forms the backbone of the QDG algorithm where it is used to construct an undirected graph ( the skeleton of the directed acyclic graph ) among expression phenotypes before orienting these edges using known QTL [11] . The QTLnet algorithm proposes a full Markov chain Monte Carlo approach to network inference , but does not scale above twenty phenotypes because of convergence rates of the Markov chain , and does not explicitly model directed cyclic graphs [16] . We also compared our algorithm to the NEO algorithm [18] , and found that our approach controlled the false-discovery rate much better and had higher power for small networks ( , results not shown ) , but the implementation of the NEO algorithm available from the author was not stable for our simulations of larger networks ( ) , and so we did not include it in a larger comparison . To compare the performance we simulated data from the model presented in Equation ( 1 ) with strong cis-eQTL , low sample variances , and different topologies , representing a scenario where there are strong eQTL , and few direct interactions between genes , with sample networks illustrated in Figure 4 . The four different classes of simulations included directed acyclic graphs for 10 phenotypes , with sparse and dense topologies ( Figure 4a , 4b ) , and directed cyclic graphs for dense ( Figure 4c ) and intermediate topologies ( Figure 4d ) , with 10 and 30 phenotypes respectively , for a total of 160 distinct network topologies generated across all the simulations . This simulation is biologically motivated by the need for strong , statistically independent cis-eQTL and interactions among genes , as observed in previous studies [26]–[28] . We simulated a set of either 10 or 30 expression phenotypes and genotypes for sample sizes of for both directed acyclic graphs and directed cyclic graphs . We simulated an F2 cross with the R package QTL [37] , with either 10 or 30 independent known unique cis-eQTL of constant effect ( ) , and error variances of . The regulatory effects ( ) were sampled from a uniform distribution with parameters or with equal probability . The network topologies were generated by randomly connected variables with equal probability , where the expected number of edges for each variable was either one , two , or three . Five replicate simulations were performed , sampling a new network topology and parameterization each time , and the power and false-discovery rate were computed for the adaptive lasso , PC-algorithm , QDG algorithm , and QTLnet algorithm for 10 expression traits , and all except QTLnet for 30 expression traits ( because of the scaling of QTLnet ) . In addition , because we simulate the QTL independently , with no trans effects , we do not perform the third step of our adaptive lasso algorithm . We compared the performance for both directed acyclic graphs as well as directed cyclic graphs . In Figure 5 and Figure 6 we show the power and false discovery rate for recovering the correct set of directed edges using these methods . While some of the power and false-discovery rate curves show large fluctuations with increasing sample size in Figure 5 and Figure 6 , this is due to elevated sampling variability due to each replicate simulation having a unique topology and parameterization . For two of these scenarios , we show that our algorithm using the adaptive lasso can outperform the PC-algorithm , the QDG algorithm , and QTLnet in terms of statistical performance ( see Figure 5c , 5d and Figure 6a , 6b ) with similar computational scaling . In general , only the QDG algorithm has competitive performance with the adaptive lasso ( see Figure 6c , 6d ) . This indicates that the necessary sample size to have a significant performance gain over the QDG algorithm may be much larger than is biologically realistic for larger more complex networks . These are significant results in two ways , the first being that we show that a feature selection method using linear regression can 1 ) identify directed regulatory architecture ( given sufficient perturbations ) and 2 ) it can also outperform state of the art network reconstruction algorithms , given a sufficient samples size and appropriate model dimension . The adaptive lasso approach appears to work the best for smaller problems ( i . e . 10 phenotypes ) with denser topologies ( i . e . Figure 4b , 4c ) and performs better than other approaches in such cases ( see Figure 5c , 5d and Figure 6a , 6b ) . This may be because smaller dimensional problems behave asymptotically at a faster rate . Unfortunately , this suggests that for larger problems ( e . g . hundreds to thousands of phenotypes ) , unless the true topology is relatively sparse , the adaptive lasso , and perhaps all of these approaches will have poor performance without unrealistically large sample sizes ( e . g . thousands ) for both directed acyclic and cyclic graphs . We also performed a simulation for a small network ( e . g . 10 phenotypes and 10 cis-eQTL ) , with dense directed acyclic topology and 200 or 1000 individuals with random variances and eQTL effects simulated from a distribution . We found a uniform reduction in power ( 10–20% ) across all methods , as well as a modest increase in false discovery rate ( 5–10% ) . Increased sample size appeared to correct for this additional randomness in the parameterization ( results not shown ) . We used our algorithm to reconstruct network structure for genome-wide gene expression data and genetic markers assayed in 112 segregants of a cross between two strains of Saccharomyces cerevisiae , reported by Brem and Kruglyak [27] . This cross was between a lab strain ( BY4716 ) and a wild strain ( RM11-1a ) , with 2 , 957 genetic markers genotyped and expression levels for 5 , 727 genes measured . While the sample size is relatively small , the study was well powered , with many strong cis-eQTL and interactions among genes [27] . An individual marker analysis was run around the cis region of each gene ( 25 kb around the start site of the gene ) to identify a set of gene expression products with strong cis-eQTL ( ( p-value ) ) , which identified 262 genes . We further filtered this set to remove cis-eQTL genotypes with high linkage , by filtering for a set with pairwise between any two cis-eQTL genotypes . Additionally , we tested the robustness of the inferred edges by randomly sampling the flanking genetic markers 20 times for all cis-eQTL and refitting the model . The percentage recovery for the top six recovered directed edges for the 20 resamplings are shown in Table 1 . All missing data for a given genotype or phenotype was set to the sample mean of the respective variable . After the additional filtering described above , we were left with 35 genes with unique , independent cis-eQTL , with an undirected network shown in Figure 7a , and possibly directed network shown in Figure 7b . Performing the adaptive lasso procedure on these 35 gene expression phenotypes and 35 genotypes identified 91 possibly directed edges among these genes , and 145 undirected edges among the genes . These hits were further filtered to ensure they represented induced dependencies , leaving six edges with relatively strong evidence of directionality ( see Table 1 and Figure 7b ) . These include four edges feeding out of the TYR1 gene , a gene involved in tyrosine biosynthesis [38] . Since TYR1 is also a hub in the undirected network ( see Figure 7a ) , this suggests that amino acid biosynthesis , and perhaps anabolism in general is driving the expression of many of this particular subset of genes . The genes in which TYR1 appears to have direct effects on have diverse molecular and biological functions including endocytosis ( RCY1 ) , sulfonate catabolism ( JLP1 ) , cell-cycle checkpoint ( BUB2 ) , and cell-cell communication ( PRM7 ) [39]–[42] . Additionally PRM7 feeds into POC4 , a proteasome chaperone protein [43] , representing possible cross-talk between cell-cell communication response and protein processing . Finally , SEN1 , a helicase indicated in RNA polymerase 2 termination [44] , appears to robustly directly affect MST27 an integral membrane protein implicated in vesicle formation [45] . In the implied undirected graph , there were striking topological features , including an average degree of 8 . 28 ( relatively dense ) , and four genes appeared to be major hubs of a sort , TYR1 , NUP60 , RDL1 , and POC4 . These hub genes may represent major axes of variation driving the expression of this subset of genes including processes such as amino acid biosynthesis , information transfer across the nuclear envelope [46] , and protein degradation . While most of the edges in the network were not orientable , there still appeared to be many dependencies ( even with a possibly high false-discovery rate ) , indicating a potentially complex set of regulatory interactions , projected on this subset of genes , driving variation in expression . Additionally , there were many edges from eQTL that would appear to be trans associations ( i . e . with large marginal correlations ) , demonstrating that many of the pathways that mediate these trans genetic effects are not captured in the observed sets of genes . Based on the simulation study , and the complexity of the recovered network ( which most likely indicates a high false discovery rate ) , a much higher sample size would need to be collected to definitively resolve this possible set of regulatory interactions , and have increased confidence in the directional interpretation of the induced edges .
Our algorithm represents a novel approach to directed network recovery by making use of a convex optimization approach for regulatory feature selection when analyzing gene expression products and cis-eQTL . This is the first algorithm that makes use of sufficient sets of cis-eQTL to infer unique directed cyclic networks from gene expression data with a feature selection methodology . Our use of the adaptive lasso procedure for feature selection has significant computational and theoretical advantages , since the underlying optimization program is convex ( ensuring a computationally efficient , unique solution ) , is model selection consistent , and has the oracle property ( asymptotically , the estimates of the non-zero regression coefficients behave as if the model was known a priori ) [25] . There have not been many algorithms proposed for genome-wide cyclic regulatory network recovery , [9]–[11] , [20] and they all have either computational or theoretical challenges associated with them , including heuristic searches through regulatory network space with no guarantee to reach networks with the strongest evidence given the data [10] , [11] , [18] , or lack sufficient perturbations to allow unambiguous regulatory inference [9] , [20] . With respect to directed acyclic network recovery , we see in the simulations that our feature selection approach with sufficient perturbations outperforms the PC-algorithm , the QDG algorithm , and the QTLnet algorithm for dense , small scale problems as shown in Figure 5c , 5d and Figure 6a , 6b . This increase in performance is a direct function of the adaptive lasso procedure correctly identifying the children of a given node , which will then force an edge to appear between the additional co-parents of that node , and its unique cis-eQTL . Once all these induced edges are identified , the structure of the directed network can be elucidated , since all the expression parents of each gene will be known . Our algorithm also does this all in a single optimization procedure , avoiding sets of iterative tests , where type-I and type-II errors can build up at each stage , such as in the PC-algorithm . Alternatively for larger more complex graphs the performance appears to be similar to that of the QDG algorithm Figure 6c , 6d , perhaps because the asymptotic properties take much larger sample sizes to be practically realized . For the analysis of the yeast data the topology of the identified network included many undirected cycles , with the few orientable edges being acyclic , as shown in Figure 7 . In addition there were a set of genes which appeared to be hubs ( the most connected being TYR1 , NUP60 , RDL1 , POC4 , and SEN1 , PCD1 , and SAN1 to a lesser extent ) . This phenomena is probably in part due to an inflation in false-positives because of the small sample size , and a complex underlying model with many unobserved variables . Yet a subset of these edges may represent hub genes capturing different broad patterns of variation across this entire sub-network . Even though most of the edges in this network are not orientable , an experiment could be devised where each of these hubs was perturbed , and given the topology it would produce a prediction about how a relatively large set of other genes in the hub's neighborhood would behave . More strongly , in the case of the TYR1 gene which had the most orientable edges , it suggests that if the process driving that gene's expression was stopped , many other genes would also be affected , but not vice-versa . A number of assumptions concerning biological networks are implicit to our algorithm . These include assumptions that are common to most graphical modeling techniques , such as sparsity , faithfulness , linearity of regulatory relationships , and normally distributed error , as well as an assumption that is specific to our algorithm: the presence of known , independent perturbations from cis-eQTL . The common assumptions are reasonable when constructing a first approximation to regulatory network structure . Sparsity and faithfulness ( i . e . the true network does not contain pathological parametrizations where there is parameter cancellation ) are essential assumptions that are implicit in algorithms for both directed and undirected network inference algorithms [5] , [6] , [11] , [16] , [20] , [29] , [30] . Regulatory relationships are not linear , but linearity is the simplest approximation that provides biologically relevant information , i . e . there is a detectable relationship between two genes , or no relationship . An assumption of normality is conservative in terms of being the most ‘random’ distribution that could have generated the data , since given an observed covariance structure , normal distributions have maximum entropy [47] . Given the absence of knowledge about the specific biological process generating the distribution of expression measurement error , and barring any clear evidence of non-normality in data , such a conservative approximation is appropriate . The assumption of independent , detectable cis-eQTL effects is the most restrictive assumption . Other methods have proposed to use trans-eQTL directly to increase the power to detect causal relationships and reduce the space of equivalent models [5] , [9]–[11] , [16] , [18] , [19] . We require the assumption of only cis-eQTL , because without it , there is no longer the exact isomorphism between the undirected graph among genotypes and phenotypes and the directed cyclic graph among phenotypes . This occurs because in the case of directed cyclic graphs , it is statistically impossible to know which phenotype in a network a trans-eQTL directly feeds into , unless their is prior knowledge about the true causal structure of the system , as with the assumption we make about cis-eQTL . This statistical degeneracy arises as a result of the “Recovery” Theorem , where when there is a set of equivalent models with independent , unique perturbations , that contains reversals of cycles , each equivalent directed cyclic graph will have an alternative perturbation topology ( i . e . the mapping between unique eQTL and gene expression phenotypes , determining which eQTL causally affects which gene expression product ) . Alternatively , as we show in real data , even if there do appear to be many trans-eQTL we can still detect a subset of edges from the cis-eQTL that behave how we would like ( by using Step 3 of the algorithm ) . While this may reduce our power to detect directed cycles in practice , it ensures that for real data-analysis we are more confident in the edges we reconstruct . Another possible solution to the incorporation of trans-eQTL would be to use the adaptive lasso to generate the initial undirected graph among genotypes and phenotypes , then to orient the edges in the graph using an iterated testing approach , as in the NEO algorithm [18] , the algorithm of Millstein et al . [19] , or the QDG algorithm [11] . We do not expect the requirement of unique cis-eQTLs to be a good approximation for all regulatory modeling situations . However , this assumption also seems reasonable , given recent biological observations of strong local polymorphism associations with gene expression ( eQTL ) which are often not in linkage disequilibrium [[26]- , [28] , [48] , [49]] . What is more , due to the structure of linkage disequilibrium in outbred populations ( the correlation structure among genotypes ) it is often possible to identify a large set of cis-eQTL that are uncorrelated and each have unique expression phenotypes , e . g . a set of eQTL that are present on different chromosomes or are far away from one another in terms of genetic map distance [28] . As a final comment , the theory of sufficient perturbations that maximize regulatory resolution , which is used as the foundation of our algorithm , is quite general , and could be used to integrate multiple data types to make predictions about putative causal regulators underlying complex phenotypes , such as disease [1] , [2] . The “Recovery” Theorem defines a class of perturbation architectures where there is a direct isomorphism between two very different types of networks: the inverse covariance structure ( an undirected network ) with perturbations and a directed cyclic graph representing a regulatory network . The theory does not require perturbations to be cis , just that there be an appropriate set of perturbations that provide resolution . More complex perturbation sets , which include sufficient perturbations as a subset , can also provide maximum resolution . One could therefore construct algorithms similar to the algorithm presented in this paper , without the local cis perturbation restriction . Moreover , the specific topology of eQTL effects need not be known , if one is willing to accept the cost of larger network equivalence classes and therefore less total regulatory resolution . With this restriction lifted , it would be possible to jointly infer the genetic perturbation architecture simultaneously with regulatory architecture , although such a joint reconstruction would require much larger sample sizes .
The network model is presented in equation ( 1 ) . For this model , we make the assumption that in the true network model , is sparse . In addition , we assume that , the error covariance matrix of expression products , is diagonal , and , where the constraint on the diagonal of ensures model identifiability . This constraint corresponds to a lack of self-loops , since the parameters representing self-loops are confounded with the error variance parameters specified by . These latter assumptions on and ( i . e . no error covariance or self-loops ) are standard , and used by all popular graphical network inference algorithms , directed and undirected , proposed to date [3] , [6] , [9]–[11] , [14] , [20] , [29] , [31] . The model depicted by Equation ( 1 ) is a completely observed structural equation model ( SEM ) [50] . The conditional log-likelihood of the model defined by Equation ( 1 ) can be written as: ( 4 ) where the full precision matrix and empirical covariance matrix are: ( 5 ) ( 6 ) with the data matrices and re-centered . We can define a fully parametrized model matrix : ( 7 ) since by definition , and , both and can be rescaled by the positive square root of the error precision matrix . From Equation ( 5 ) , Equation ( 6 ) , and Equation ( 7 ) the relationship between the fully parametrized model matrix , and the full precision matrix is ( 8 ) This defines a system of homogeneous polynomials of degree two which exactly specifies the relationship between the directed graph , which may contain no cycles ( a directed acyclic graph or DAG ) or may contain cycles ( a directed cyclic graph or DCG ) , and the moralized undirected graph . Given the importance of having as small a set of equivalent models as possible for making meaningful inference , and the necessity of perturbations for minimizing equivalence classes , it is of interest to know what will constitute a sufficient set of perturbations , i . e . to shrink the size of arbitrary equivalence classes as much as possible . In the following section we provide proofs of three theorems that describe such a set . We note that it should also be possible to use the work of Richardson on cyclic causal discovery [20] to arrive at the same theoretical condition concerning a set of sufficient perturbations , though it is beyond the scope of this work to show this connection . Here , we use an independent and simpler proof based on normal theory and matrix algebra . Our theory also provides a generalization of the work of Chaibub Neto et al . [11] , which shows that sets of unique ( or “driving” ) QTL for each phenotype can be used to uniquely orient edges in a directed cyclic network . Our approach allows us to represent the problem of directed network inference as a model selection problem within a regression equation for each phenotype . This allows us to avoid the reliance on computationally inefficient heuristics [3] , [10] , [11] , which can generate many possibly poor-fitting networks depending on how the algorithm is run , when considering sample sizes that are typical of experiments collecting genome-wide gene expression data . The “Recovery” Theorem demonstrates how the set of equivalent DCGs can be recovered from the precision matrix between expression phenotypes and loci ( the matrix ) . This last result is incorporated into our algorithm for inferring sparse network structure with a sufficient perturbation ( eQTL ) set . Note that while the algorithm depends on sparsity for efficient network recovery , the results of these theorems are general and do not require such a constraint . In addition , we note in a further Lemma that even in the case of directed cycles , if we know which phenotype a perturbation feeds into , we can further reduce the size of the equivalence class to a unique directed cyclic graph . Theorem 1: Given two distribution equivalent directed cyclic graphs , with equivalent parametrizations and , any matrix A which satisfies , must be orthonormal ( i . e . ) . Proof of Theorem 1: Since , and from the definition of equivalence , if and are equivalent , then . Therefore , . Left multiply by and right multiply by , then , where is a positive definite invertible matrix of rank . Left and right multiply by , and . The matrix can be thought of as a linear operator that allows transformations between models which produce the same covariance ( and inverse covariance ) structure ( even between models which are not faithful ) . We use this operator to prove the following theorem after rescaling the network and perturbation parameters as in Equation ( 7 ) : , : Theorem 2: If there exists an ordered set of rows of the perturbation graph parametrized by such that , where is a diagonal matrix of rank and is a signed permutation matrix , then 1 ) if parametrizes a DAG , then for any parametrization of any DAG , there does not exist an alternative equivalent DAG or DCG , and 2 ) if parametrizes a DCG , then for any parametrization of any DCG , there exists a finite set of equivalent DCGs , where each equivalent DCG contains a reversed directed cycle with reference to the original DCG . Proof of Theorem 2: Given exists , assume there exists an alternative equivalent model parametrized by and . Then , by Theorem 1 , there exists an orthonormal matrix where , , and . Because and are invertible , we have: . This implies that . Since is diagonal for any parametrization , and must also be diagonal for all equivalent parametrizations . If there does not exist a signed permutation matrix such that , with diagonal , then there always exists a parametrization of where is not diagonal , and therefore not equivalent ( since all non-zero elements of are free to vary ) . Therefore is either an identity matrix or a signed permutation matrix . Now consider . Because in this parametrization , , the only allowable equivalent model transformations must have positive non-zero elements along the entire diagonal . Therefore , if parametrizes a DAG , then , and if parametrizes a DCG , then where is any signed permutation matrix which ensures non-zero positive elements along the diagonal of . This corresponds directly to reversing the order of any set of directed cycles in the graph . This theorem allows us to understand constraints on possible equivalent models in the specific case when each node has at least one unique perturbation . In the next theorem , we focus on the structure of the moralized graph ( i . e . the precision matrix ) for these models , and see how it maps back to the set of possible unmoralized directed graphs that generated the moralized graph . We define the set of parents of a particular node , , from the directed graph as , and the set of all nodes in an undirected graph that have edges to node as . “Recovery” Theorem: If in there exists an independent perturbation vertex set and a response vertex set where and , then the only equivalent directed cyclic graphs among that could have generated contain permutations of cycles , and can be recovered from . Proof of the “Recovery” Theorem: The existence of an independent perturbation vertex set and response vertex set that satisfies these conditions corresponds directly to a perturbation topology and parametrization specified by from Theorem 2 . Given this observation , Theorem 2 ensures the constraint on possible equivalent models . Finally , the reason the structure can be recovered from is apparent from Equation ( 5 ) and ( 7 ) , where , and therefore Since is diagonal it won't change which elements of are zero or non-zero . In the case of DAGs , a generalization of this theorem is trivial to prove for graphs defined over arbitrary probability measures , since the process of moralization of a graph connects all the parents of a given node . Since in this specific perturbation case , each node has at least one unique parent ( from the perturbations ) , then a connection will be induced between the unique perturbation parent and each of its co-parents , indicating exactly what the unique set of parents are for that given node . Alternatively , as we saw in Theorem 2 , the assumptions of normality and linearity are key to showing that even for directed cyclic graphs that have unique perturbations , there still exists multiple equivalent models . In the “Recovery” Theorem we see that we can still determine these ‘minimal’ equivalence classes from the moralized graph . It is interesting to observe that the perturbation topology can completely change among equivalent directed cyclic graphs , whereas it cannot for directed acyclic graphs . If one knows which node each perturbation feeds into , then the following is true: Lemma: If the underlying perturbation topology , , is known , then the cardinality of all directed cyclic equivalence classes is reduced to one . This further reduction of the equivalence relationships is apparent when one considers that each equivalent perturbation topology specifies exactly one member of the equivalence class ( from the “Recovery” Theorem ) . Therefore , if one knows the true perturbation topology , then one knows the true regulatory model . This allows us to infer a unique directed cyclic graph in the case where we know which phenotype each genetic perturbation feeds into . Hence , the reason behind making our major biological assumption: to only consider the genetic effects of cis-eQTL and assume that the cis-eQTL feeds directly and uniquely ( i . e . non-pleiotropically ) into the local gene . With trans-eQTL , unless there is prior knowledge about exactly which gene each trans-eQTL affects ( i . e . about the pathways in question ) , there is no way to reduce this equivalence class to a unique directed cyclic graph . Adaptive lasso . For Step 1 of the algorithm , we perform an individual marker analysis of each genetic polymorphism in a window around the start site of the gene , and only include the markers that are significant given a Bonferroni correction for multiple testing . We then filter these sets of cis-eQTL such that they are effectively independent given the linkage disequilibrium structure of the data . For the analysis of the yeast data , we found that a maximum pairwise between cis-eQTL genotypes was a very conservative threshold given a resampling test of random markers across the genome ( results not shown ) . For Step 2 of the algorithm , the lasso problems from Equation ( 2 ) and ( 3 ) are solved using the cyclic coordinate descent method of Friedman et al . [52] , as implemented in the ‘glmnet’ package , called by the ‘parcor’ package [8] . While this method is an approximation to solving the adaptive lasso for the log-likelihood defined in Equation ( 4 ) , there are theoretical connections between an exact solution to the problem , and this approximate solution which suggest that in some cases the approximation will not perform much worse than the exact solution ( i . e . highly penalized cases ) [23] . For Step 3 of the algorithm , we performed a permutation test to very conservatively ensure that the induced edge found between an upstream gene , and the cis-eQTL , did not arise from a trans-effect of the cis-eQTL . To do this we randomly resampled the genotype data 10 , 000 times for each induced edge , and determined the proportion of the time the absolute value of the marginal correlation between upstream gene and cis-eQTL under the empirical null model was greater than the absolute value of the observed marginal correlation . We only treated induced edges as representing a directed relationship between a pair of phenotypes if the probability of observing a greater value under the empirical null model was greater than 0 . 90 . PC-algorithm . While this is only designed to reconstruct directed acyclic graphs , it has been used in a combined gene expression and genotype context to reconstruct directed cyclic graphs [11] . The PC-algorithm reconstructs the skeleton ( i . e . set of edges regardless of edge orientation ) of a partially directed acyclic graph ( PDAG ) by performing forward tests of conditional independence . It first starts by constructing a correlation graph ( i . e . a conditional independence graph where one conditions on the empty set ) , then in a forward step-wise manner , removing edges in the neighborhood of each node by increasing the size of the conditioning set based on the neighborhood of each node . Once the cardinality of the conditioning set is equal to or larger than the neighborhood for all nodes , the algorithm terminates . While this is being done , all identified v-structures ( co-parents of a common child ) are being tabulated , so that afterwards these edges can be oriented . Then , there is a set of rules , based on the seed v-structures which orient a small initial set of edges , which orient many additional edges in the network , by propagating the implications of the few initial oriented edges , with respect to the d-separation criterion defined for directed acyclic graphs [29] , [31] . We applied the PC-algorithm by giving it the entire set of gene expression products with cis-eQTL as well as all of the cis-eQTL genotypes as well . There is one tuning parameter , , for the implementation ‘PCalg’ , which represents the level of significance each test of conditional independence has to pass to correspond to removing an edge from the skeleton of the network . We used a conservative value of , based on simulation results presented in Kalisch et al . [29] . For directed acyclic graphs , the PC-algorithm will also use the cis-eQTL to orient each of the edges in the network correctly and uniquely . For directed cyclic graphs , the PC-algorithm will try to orient the edges to form a directed acyclic graph , but often will fail , and draw a random DAG instead . We also apply the PC-algorithm to directed cyclic network recovery by having it identify both the skeleton with perturbations , and then have it attempt to orient as many edges as possible , given that every regulatory relationship should be orientable with the PC-algorithm when there are sufficient , unique perturbations . While in some cases this will fail , especially as the sample size grows and it becomes more sensitive to variations away from the assumption of no cycles , in practice it is able to orient many edges correctly in a directed cyclic graph . QDG algorithm . The default settings were used for the QDG algorithm , as provided by the authors [11]: for the PC-algorithm skeleton reconstruction step , the skeleton reconstruction method based on the PC-algorithm , and the number of random restarts of iterative testing of different global edge orientations was set to ten . The QDG algorithm uses either the PC-algorithm or UDG algorithm [53] to generate a skeleton among phenotypes [12] . Then , the QDG algorithm orients edges between phenotypes based on a LOD score computed by leveraging each phenotype's known QTL . To find a globally optimal orientation of edges , an iterative search over orientations is performed to find all possibly cyclic networks which fit the data well [11] . We tried both methods in the QDG algorithm to generate the skeleton , and did not see a significant difference in performance for our simulations ( results not shown ) . QTLnet algorithm . The default settings were used for the QTLnet algorithm , as provided by the authors [16]: we ran it for 20 , 000 iterations , sampling every 20 iteration after a burn-in of 2 , 000 iterations . The QTLnet algorithm uses a fully Bayesian Markov chain Monte Carlo approach to solve the problem of joint phenotype genotype network inference , constraining the proposed graph transitions to directed acyclic graphs [16] . In our analyses , we use the Bayesian model averaged output of the QTLnet algorithm , and include an edge only if its posterior probability of inclusion is greater than 0 . 50 . NEO algorithm . We used the default settings for the NEO algorithm , based on the code available from the author's website: http://www . genetics . ucla . edu/labs/horvath/aten/NEO/ [18] . The NEO algorithm uses multiple QTL to orient edges between an arbitrary pair of phenotypes based on different structural equation model based statistics [18] , but has no mechanism to remove edges among phenotypes by conditioning on other phenotypes , and will therefore often have high false-discovery rate for recovery of the network generating the data among phenotypes . This was another justification , aside from the scaling of the algorithm , for why we did not include it in our broader comparison of alternative methods . | Determining a unique set of regulatory relationships underlying the observed expression of genes is a challenging problem , not only because of the many possible regulatory relationships , but also because highly distinct regulatory relationships can fit data equally well . In addition , most expression data-sets have relatively small sample sizes compared to the number of genes measured , causing high sampling variability that leads to a significant reduction in power and inflation of the false positive rate for any network reconstruction method . We propose a novel algorithm for network reconstruction that uses a theoretically and empirically well-behaved method for selecting regulatory features , while leveraging genetic perturbations arising from cis-expression Quantitative Trait Loci ( cis-eQTL ) to maximally resolve a network . Our algorithm has good performance for realistic samples sizes and can be used to identify a unique set of acyclic or cyclic regulatory relationships that explain observed gene expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/systems",
"biology",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2010 | Gene Expression Network Reconstruction by Convex Feature Selection when Incorporating Genetic Perturbations |
Through full genome analyses of four atypical Bacillus cereus isolates , designated B . cereus biovar anthracis , we describe a distinct clade within the B . cereus group that presents with anthrax-like disease , carrying virulence plasmids similar to those of classic Bacillus anthracis . We have isolated members of this clade from different mammals ( wild chimpanzees , gorillas , an elephant and goats ) in West and Central Africa ( Côte d’Ivoire , Cameroon , Central African Republic and Democratic Republic of Congo ) . The isolates shared several phenotypic features of both B . anthracis and B . cereus , but differed amongst each other in motility and their resistance or sensitivity to penicillin . They all possessed the same mutation in the regulator gene plcR , different from the one found in B . anthracis , and in addition , carry genes which enable them to produce a second capsule composed of hyaluronic acid . Our findings show the existence of a discrete clade of the B . cereus group capable of causing anthrax-like disease , found in areas of high biodiversity , which are possibly also the origin of the worldwide distributed B . anthracis . Establishing the impact of these pathogenic bacteria on threatened wildlife species will require systematic investigation . Furthermore , the consumption of wildlife found dead by the local population and presence in a domestic animal reveal potential sources of exposure to humans .
Bacillus anthracis has been classically defined as a clade with low genomic diversity within the B . cereus sensu lato group , whose members carry two virulence plasmids , pXO1 and pXO2 , and exhibit a set of known phenotypic characteristics [1–3] . Found in many parts of the world , the organism still causes significant losses in domestic and wild animal populations and sometimes fatal infections in humans [4] . However , cases of anthrax-like disease caused by non B . anthracis members of the B . cereus group have also been identified , affecting both animal and human populations [5–8] . Early descriptions involved fatal anthrax-like infections in wild western chimpanzees ( Pan troglodytes verus ) in Côte d’Ivoire in 2001 and 2002 followed by wild central chimpanzees ( P . t . troglodytes ) and a western lowland gorilla ( Gorilla gorilla gorilla ) in Cameroon in 2004 and 2006 [9–11] . Strains isolated have been shown to carry plasmids almost identical to pXO1 and pXO2 ( designated pCI-XO1 or pBCXO1 and pCI-XO2 or pBCXO2 , respectively , in previous publications [12 , 13] ) . As these atypical B . cereus types share various properties but differ significantly from B . anthracis [14] , they were aptly named B . cereus biovar ( bv ) anthracis . On phenotypic level , the strains combined features of B . anthracis ( lack of haemolyis and phospholipase C activity ) and B . cereus ( resistance to the diagnostic gamma phage , motility ) . Like B . anthracis , the strains from Côte d’Ivoire were sensitive to penicillin , but the strains from Cameroon were resistant [14] . Multi-locus sequence typing [15] showed the same sequence type for all B . cereus bv anthracis strains . Whole genome sequencing of one isolate from Côte d’Ivoire revealed the presence of six genomic islands ( 12–22 kb in size ) and a small , 14 kb plasmid ( pCI-14 ) with unknown function [12] . With a few exceptions , the sequences of these islands were only detected in B . cereus bv anthracis , whereas Island III , a putative prophage , was distributed among further strains of the B . cereus group [12] . Island VI was absent from the Cameroon isolates , and pCI-14 was only detected in some isolates from Côte d’Ivoire . The pleiotropic regulator PlcR , known to control expression of multiple genes including those related to virulence within the B . cereus group [16] is inactive both in B . anthracis due to a nonsense mutation [17] and in B . cereus bv anthracis due to a frameshift mutation which results in an altered C-terminus of the protein [12] . To some degree , virulence and virulence gene regulation are similar in B . cereus bv anthracis and classic B . anthracis . Small animal models indicate a similar level of virulence ( assessed by determination of lethal doses ) after infection with wild type strains , and regulation of the toxin and capsule genes by the global regulator AtxA was shown [13] . Differences in virulence account for the fact that besides the typical polyglutamate capsule of B . anthracis , an additional capsule composed of the polysaccharide hyaluronic acid was detected in B . cereus bv anthracis . This capsule , which is encoded by the hasACB operon on plasmid pXO1 , is lacking in B . anthracis due to a mutation in the hasA gene [13] . Here we report the isolation , characterization and genome sequencing of additional atypical B . cereus group members isolated from wild and domestic animals sampled in Cameroon ( hereafter referred to as CAM strain ) , the Central African Republic ( RCA strains ) and the Democratic Republic of Congo ( DRC strain ) . Phylogenomic analyses provide evidence that these strains and those from Côte d’Ivoire ( CI strains ) belong to a single chromosomal clade clearly distinct of the B . anthracis clade .
All sampling sites are indicated in Fig 1 . In March 2012 , 40 blood and tissue samples were collected from livestock ( goats ( n = 9 ) , pigs ( n = 23 ) and sheep ( n = 1 ) ) in Luebo , a forested town within the Kasai district of the Democratic Republic of Congo ( DRC ) . Among the animals sampled , one goat had recently died while another was described by residents as being sick . The samples were tested on-site during a joint training project led by the Public Health Agency of Canada and the Institut National de Recherche Biomédicale , DRC . In September 2012 , eco-guards discovered a forest elephant ( Loxodonta cyclotis ) carcass in the Dzanga-Ndoki National Park , part of the Dzanga Sangha Protected Areas ( DSPA ) complex , in the Central African Republic ( RCA , approximately 1500 miles away from DRC ) . The carcass was still intact and no signs of poaching were visible . A necropsy was performed and samples taken the following day in the course of a joint World Wide Fund for Nature / Robert Koch-Institute mission to investigate causes of death amongst wildlife in the area using full personal protective equipment [11] due to a history of highly pathogenic microorganisms in the area and species affected [18] . At this point the carcass had been partly opened , presumably by scavengers or humans . Five days later , an ape carcass ( later confirmed by genetic analyses as a central chimpanzee ) was discovered in a tree nest in the same area . Bone and skin samples were collected from the carcass , which was in an advanced stage of decomposition . Finally , in January 2013 , a western lowland gorilla was found dead in the same area . The three-year-old male was part of a closely monitored group habituated to humans , and had not shown any signs of illness the previous day . Since no veterinarian was on site at this point , only deep nasal swabs were taken from the carcass by specifically instructed and protected biologists . Samples were preserved in tubes with and without preservative RNAlater ( Ambion/Life Technologies , Darmstadt , Germany ) . All three RCA carcasses were found within a radius of five kilometres . Samples have not been collected in the course of research projects ( and therefore no permit numbers exist ) , they have been collected on request and as part of collaboration between the field site and the according wildlife authority of RCA ( Ministère d’Eaux et Fôret , Chasse et Peche and the Ministère de l’Education Nationale , de l’Alphabetisation , de l’Enseignement Superieur , et de la Recherche ) to investigate causes of death in wildlife of the region . The finding of the carcasses and later of according results of analyses have been communicated immediately to the authorities and been used to warn the local population . Samples from domestic animals from DRC have been collected in the course of collaboration with the INRB , no special permission for such sampling is required . All wildlife samples have been exported under permission of the according CITES ( Convention on International Trade in Endangered Species of Wild Fauna and Flora ) . The local veterinary authorities of DRC and RCA , provided a certificate of origin as requested by the German veterinary authorities ( Senatsverwaltung für Justiz und Verbraucherschutz Abteilung V—Verbraucherschutz Referat V A—Lebensmittel- und Veterinärwesen , Gentechnik Stellenzeichen—V A VET 0 . 2 , Berlin , Germany ) for import of samples . DNA of the RCA samples was extracted from various tissues following the protocol of the NucleoSpin RNA II Kit ( Macherey-Nagel GmbH & Co . KG , Düren , Germany ) using the NucleoSpin RNA/DNA Buffer Set for parallel purification of genomic DNA , excluding step 7 for DNA digestion . DNA was prepared on separate days and samples were treated from one animal at a time . To test for the presence of B . cereus bv anthracis real-time PCR assays were performed targeting the pagA gene for pXO1 [19] , the capB gene for pXO2 , and a marker specific for genomic island IV of B . cereus bv anthracis [12] . The primers and TaqMan-probes for capB and IslandIV were designed using the Primer Express V2 . 0 software ( Applied Biosystems , Darmstadt , Germany ) and ordered from Metabion ( Martinsried , Germany ) . Primer and probe sequences are shown in S3 Table . Real-time PCR conditions were applied as described before [19] . Genetic identification of the great ape skin samples was performed ( by tissue extraction as above , followed ) by a pan-mammal assay as described by Calvignac-Spencer et al . 2013 [20] . Blood samples from two goats from DRC were tested for B . anthracis using real-time PCR screening in the field . The field testing comprised of DNA extraction via use of the Qiagen ViralAmp ( Qiagen , Hilden , Germany ) kit as per manufacturer’s directions , followed by real-time PCR using an assay targeting both the pXO1 and pXO2 plasmids ( in-house developed ) in addition to a chromosomal region with the gyrase gene [21] . Liver samples ( conserved in RNAlater ) and untreated fat tissue of the elephant from RCA were cut into small pieces and used for cultivation on different selective ( blood trimethoprim agar , Cereus Ident agar , Cereus selective agar ) and non-selective ( sheep blood agar ) media [14] . To test for the presence of spores , a small piece of each tissue was additionally placed into 500 μl of saline and vegetative bacteria were inactivated by heating at 65°C for 30 min . The nasal swab ( also conserved in RNAlater ) taken from the gorilla in RCA was transferred into 900 μl saline and one half was directly spread on different agar plates and the remaining half was heat inactivated . After heat treatment , tissue material and supernatant was spread on different media as describe above . Heat-inactivated blood from the DRC was also cultured for the presence of Bacillus-like organisms on non-selective blood agar . Any colonies suspicious for B . cereus bv anthracis were dispensed in water , heat inactivated at 95°C for 30 min and used directly for real-time PCR assays as described above . If the presence of B . cereus bv anthracis was confirmed , the corresponding bacterial isolates were tested for motility by growth on a semi-solid tryptic soy agar ( 0 . 3% agar ) as well as for susceptibility to penicillin G and the diagnostic gamma phage as described before [14] . DNA was extracted from bacterial isolates using the DNeasy Blood & Tissue Kit ( Qiagen ) for the RCA isolates and the Epicentre MasterPure kit ( Madison , Wisconsin , USA ) for the DRC isolate . DNA was tested for the presence of genomic islands I to VI and plasmid pCI-14 using primers targeting appropriate genes . Primer sequences are listed in S3 Table . Gene fragments were amplified by PCR under the same conditions as described before [12] . For whole genome sequencing , DNA from B . cereus bv anthracis CI ( same isolate as sequenced previously using the Sanger method , [12] ) , CAM ( isolate from chimpanzee ) , RCA ( isolates from elephant and gorilla ) and DRC was processed using the Nextera DNA Library Preparation Kits ( Illumina , Munich , Germany ) as per manufacturers’ instructions . Miseq reagent kits with v2 chemistry ( 500 cycle ) were used on the Miseq platform ( Illumina ) to generate sequence data and fed through an in-house bioinformatic pipeline described below . MLST [15] was performed to confirm that also this method would be capable of differentiating B . cereus bv anthracis from B . anthracis and we investigated the existence of the frameshift mutation in the plcR gene and the integrity of the hasA gene by PCR and sequencing using standard methods . Strains were assembled with Spades version 2 . 5 . 1 with recommended parameters for 2 x 250 bp Illumina reads . Contigs were filtered against lengths < 200 bp [22] . Average Nucleotide identity ( ANI ) was determined with Jspecies blast option [23] using default parameters based on spades contig assemblies and refSeq records from NCBI ( S4 Table ) . Canonical SNPs were confirmed manually by visual inspection with Tablet from generated pileups and alignment with samtools and SMALT respectively [24 , 25] ( http://www . sanger . ac . uk/resources/software/smalt/; version 0 . 7 . 5 ) . Core pipeline analyses were generated with an in-house pipeline available at github ( https://github . com/apetkau/core-phylogenomics; commit version 0317413ba9 ) . To ensure consistent data across all strains , in silico error free illumina reads were generated using Wombac from contigs available on NCBI . This was only done for strains without any publicly available raw reads http://www . vicbioinformatics . com/software . wombac . shtml ( version dated Oct 3 , 2013 ) . All public accessible Bacillus strain sequences on NCBI ( S5 Table ) were downloaded on June 25 , 2015 and used in an initial round of phylogenetic analyses for each reference . An iterative approach was used to exclude strains based on their core percentage to the reference strain B . anthracis Ames Ancestor . Strains with less than 50% homology were removed from final tree ( s ) analyses . Core pipeline criteria for high quality SNPs ( hqSNPs ) were minimum base pair and mapping quality > = 30 phred score with 25 read coverage with 75% of consensus . SNPs were concatenated together to create multiple meta-alignments; one for each plasmid and chromosome . Model selection was performed in a maximum likelihood framework using jModelTest v2 . 1 . 3 [26] . Phylogenetic analyses were performed in PhyML using the GTR model; branch support was estimated with Shimodaira-Hasegawa-like approximate likelihood ratio tests [27] . ML trees were rooted with TempEst v1 . 5 [28] . Homoplasy indices were calculated using Paup * version 4 . 0 [29] . Figures were generated using FigTree v1 . 4 . 1 ( http://tree . bio . ed . ac . uk/software/figtree/ ) .
DNA extracted from tissues , bones , nasal swabs and blood of the elephant , chimpanzee , gorilla from RCA and livestock from DRC were tested by real-time PCR assays targeting pXO1 and pXO2 . All samples from wildlife and the samples of the sick and dead goats were positive . The presence of B . cereus bv anthracis was further confirmed by a specific real-time PCR targeting Island IV [12] ( Table 1 ) . Three new bacterial colony isolates with a typical B . cereus bv anthracis morphology were cultured from fat tissue of the elephant ( specimen number A-363/2 ) , the nasal swab of the gorilla ( specimen number A-364/1 ) and blood of the dead goat ( specimen number 14-0024-1 ) . The presence of spores in these samples was confirmed by cultivation of bacteria after heat treatment of clinical material at 65°C for 30 min . No culture was obtained from the bones and skin of the chimpanzee as well as from any tissue from the other goat . The bacteriological properties of the three new isolates resembled those described previously for the CI and CAM strains ( Table 1 ) . Like B . anthracis , the isolates were non-haemolytic , lacked the phosphatidylinositol-specific phospholipase C activity and showed a weak lecithinase ( phosphatidylcholine-specific phospholipase C ) activity . Unlike B . anthracis , the isolates were resistant to the gamma-phage [14] . Most colonies also showed a strong mucoid appearance , indicating the presence of encapsulated bacteria . In addition , like the CAM strain , the three new isolates were resistant to penicillin G . Like the CI and CAM strains , the forest elephant and the gorilla isolates from RCA both exhibited motility , whilst the goat isolate from DRC was non-motile . Molecular analyses were performed on bacterial DNA that was extracted from suspicious colonies isolated from the RCA and DRC samples . The presence of six genomic islands and the small plasmid pCI-14 was examined by PCR [12] . Isolates from RCA and DRC harboured genomic islands I to V , but like the CAM strain lacked genomic island VI and plasmid pCI-14 ( Table 1 ) . Analyses of key functional genes revealed that all strains shared the frameshift mutation in the plcR regulator gene already described for the CI strain [12] . In addition , and unlike B . anthracis , they possessed a functional hasACB operon on pXO1 allowing synthesis of a second capsule type composed of hyaluronic acid [13 , 30] . Interestingly , from the whole genomic sequencing data ( see below ) a pre-mature stop codon was detected in one gene ( fliP ) of the flagella gene cluster of the DRC strain , thereby explaining the lack of motility observed for this strain . Isolate characteristics are summarized in Table 1 . We sequenced the genomes of the three new isolates ( RCA elephant and gorilla , DRC ) , concurrently with the CAM and CI strains and compared their chromosomal and plasmid sequences ( unclosed ) to a collection of complete genomes of B . cereus , B . thuringiensis and B . anthracis . Common genes distributed across the chromosomes of all genomes analysed represent the chromosomal core , and altogether , 3 , 091 , 761 core coding positions were detected . Phylogenomic analyses performed in a maximum likelihood ( ML ) framework using variable coding positions ( single nucleotide polymorphisms; SNPs ) unambiguously identified 169 , 492 positions in all genomes for chromosomal sequences . Similar analyses were performed to identify the core regions of the two plasmids of B . anthracis , B . cereus bv anthracis and a few other members of the B . cereus group which also possess pXO1 . There were 697 and 250 variable positions determined for pXO1 and pXO2 plasmid sequences , respectively ( most members of the B . cereus group do not possess these plasmids ) . Despite the reduced size of these data sets , unique SNPs were observed for all strains . Maximum likelihood trees derived from the analysis of chromosomal sequences strongly supported the existence of a clade comprising all B . cereus bv anthracis strains ( Fig 2A ) . Of note , this clade was not closely associated to the B . anthracis clade . Instead , the closest relative of B . cereus bv anthracis appeared to be another member of the B . cereus group , i . e . B . cereus ISP3191 [31 , 32] , and , conversely , a number of B . cereus and B . thuringiensis strains appeared more closely related to B . anthracis than B . cereus bv anthracis . B . cereus bv anthracis strains did not differ amongst each other at the chromosome level at more than 450 positions ( ~0 . 01% difference ) , which was markedly lower than the maximum ~1 , 000 SNPs observed amongst B . anthracis chromosomes ( ~0 . 04% ) . However , the number of SNPs among B . cereus bv anthracis strains will probably increase with larger sampling numbers . Furthermore , approximately 39 , 000 SNPs separated isolates within the B . cereus bv anthracis and B . anthracis clades; in comparison , B . cereus ISP3191 was only separated from B . cereus bv anthracis by an average of ~12 , 000 SNPs . Within the B . cereus bv anthracis clade , isolates from Central Africa ( RCA and CAM ) were more closely related to each other than to the DRC and CI isolates ( Fig 2B ) . ML trees derived from pXO1 and pXO2 data sets also supported the monophyly of B . cereus bv anthracis isolates ( Fig 3 ) . B . anthracis isolates also formed a clade using sequence data from pXO1 , but not pXO2 . In contrast , B . anthracis strains A1055 , 2000031052 , 2000031021 and 2002013094 belonging to the rare C lineage [3] appeared more closely related to B . cereus bv anthracis isolates than to other B . anthracis isolates . This pattern , however , entirely depends on the placement of the root which , in the absence of an appropriate outgroup , was determined with TempEst by minimizing the variance of root-to-tip distances ( i . e . assuming the tree mostly behaves in a clocklike manner [28] ) . This observation should therefore be regarded as preliminary and certainly deserves further investigation . For both plasmids , the low number of polymorphic sites prevented the full determination of the branching order within B . cereus bv anthracis . The clade comprising CAM and RCA isolates was , however , satisfactorily resolved . The lack of support for the second deepest node in the clade did not allow addressing the question of the branching order of CI , DRC and CAM+RCA . We finally determined homoplasy indices ( HI ) in a parsimony framework for all complete trees as well as for the B . cereus bv anthracis clade . The complete chromosomal tree was the only one to reveal significant homoplasy levels ( HI = 0 . 47 ) which suggests a measurable contribution of recombination to the evolutionary history of the Bacillus cereus clade . In contrast , HI were very low for the complete plasmid trees ( pXO1 0 . 01 , pXO2 0 . 02 ) and for the B . cereus bv anthracis clade ( chromosome 0 . 00 , pXO1 0 . 00 , pXO2 0 . 00 ) . To determine whether easier PCR-based methods would also identify the B . cereus bv anthracis clade , we performed multilocus sequence typing ( MLST ) and showed that sequences of the alleles of the seven housekeeping gene fragments for MLST [15] resulted in a sequence type , ST 935 , which is unique for all isolates of B . cereus bv anthracis ( Table 1 ) . The profile has been submitted to the corresponding database ( http://pubmlst . org/bcereus/ ) . NCBI accession numbers for sequences of B . cereus bv anthracis CI , CAM , RCA elephant ( A-363/2 ) , RCA gorilla ( A-364/1 ) and DRC goat ( 14-0024-1 ) are from SAMN03610233 to SAMN03610237 , respectively .
Of note , after diagnoses of anthrax-like disease , the protected area authorities were informed and actions were put in place to monitor for further cases in DSPA . At present , no human cases have been reported in these regions . | Anthrax has historically been attributed to a single cluster within the Bacillus cereus complex denoted as B . anthracis . Here , we demonstrate the existence of a distinct clade of B . cereus isolates causing anthrax-like disease in a range of wild and domestic mammals in tropical Africa . These strains , designated B . cereus biovar anthracis , combine bacteriological and molecular features of B . cereus and B . anthracis . Many questions about the epidemiology , biology and impact of this cluster of anthrax causing B . cereus still remain open . On the technical side it will be important to adapt diagnostic methods for the detection of such atypical B . cereus strains–through the inclusion of molecular tools for the detection of the B . anthracis virulence plasmids that appear to be the prerequisite to cause disease . Through reliable detection of a broad range of B . cereus group isolates causing anthrax-like disease it will be possible to assess the distribution and diversity of these pathogens and their impact on public health and wildlife populations . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"animal",
"types",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"bacillus",
"microbiology",
"vertebrates",
"animals",
"mammals",
"primates",
"phylogenetic",
"analysis",
"molecular",
"biology",
"techniques",
"mammalian... | 2016 | Bacillus cereus Biovar Anthracis Causing Anthrax in Sub-Saharan Africa—Chromosomal Monophyly and Broad Geographic Distribution |
The establishment of correct neurotransmitter characteristics is an essential step of neuronal fate specification in CNS development . However , very little is known about how a battery of genes involved in the determination of a specific type of chemical-driven neurotransmission is coordinately regulated during vertebrate development . Here , we investigated the gene regulatory networks that specify the cholinergic neuronal fates in the spinal cord and forebrain , specifically , spinal motor neurons ( MNs ) and forebrain cholinergic neurons ( FCNs ) . Conditional inactivation of Isl1 , a LIM homeodomain factor expressed in both differentiating MNs and FCNs , led to a drastic loss of cholinergic neurons in the developing spinal cord and forebrain . We found that Isl1 forms two related , but distinct types of complexes , the Isl1-Lhx3-hexamer in MNs and the Isl1-Lhx8-hexamer in FCNs . Interestingly , our genome-wide ChIP-seq analysis revealed that the Isl1-Lhx3-hexamer binds to a suite of cholinergic pathway genes encoding the core constituents of the cholinergic neurotransmission system , such as acetylcholine synthesizing enzymes and transporters . Consistently , the Isl1-Lhx3-hexamer directly coordinated upregulation of cholinergic pathways genes in embryonic spinal cord . Similarly , in the developing forebrain , the Isl1-Lhx8-hexamer was recruited to the cholinergic gene battery and promoted cholinergic gene expression . Furthermore , the expression of the Isl1-Lhx8-complex enabled the acquisition of cholinergic fate in embryonic stem cell-derived neurons . Together , our studies show a shared molecular mechanism that determines the cholinergic neuronal fate in the spinal cord and forebrain , and uncover an important gene regulatory mechanism that directs a specific neurotransmitter identity in vertebrate CNS development .
The choice of neurotransmitter is one of the most fundamental aspects of neuronal fate decision . Cholinergic neurons are located in diverse regions of the CNS , which do not share the developmental origin , and regulate complex behaviors . In the spinal cord , cholinergic motor neurons ( MNs ) control locomotion , whereas in the forebrain , cholinergic neurons regulate cognitive processes [1] , [2] . Defects in function or survival of cholinergic neurons result in severe human pathologies , including spinal cord injuries , diseases associated with impaired motor function and cognitive disorders resulting from the loss of forebrain cholinergic neurons ( FCNs ) [3] . Despite the crucial roles of cholinergic neurons in human physiology and pathology , the mechanisms that specify cholinergic neuronal cell fate throughout the CNS during vertebrate development remain largely unknown . The cholinergic neurotransmission system requires the function of several key factors that are highly expressed in all cholinergic neurons , termed cholinergic pathway genes ( Fig . 1A ) [4] , [5] . Understanding the gene regulatory mechanisms that control the expression of cholinergic pathway genes in different groups of cholinergic neurons will provide crucial insights into the process of cholinergic fate specification in CNS development . Given that each of the cholinergic pathway genes is essential for efficient cholinergic neurotransmission , it is probable that they are up-regulated in a coordinated fashion as neurons acquire cholinergic neuronal identity during vertebrate development . Supporting this possibility , the vesicular acetylcholine transporter ( VAChT , also known as Slc18a3 ) gene is encoded within an intron of the choline acetyltransferase ( ChAT ) gene in all metazoans examined thus far , including C . elegans , Drosophila and mammals [6] . This unique genomic arrangement suggests that the ChAT and VAChT genes are co-regulated by a single set of transcription factors . Furthermore , in a subset of cholinergic MNs of C . elegans , an Ebf-type transcription factor UNC-3 regulates a battery of cholinergic genes via a shared UNC-3-response motif [7] . Two critical questions remain to be answered . First , is a battery of cholinergic pathway genes coordinately regulated by a common transcription factor in vertebrate CNS , similar to UNC-3-directed control of cholinergic genes in C . elegans ? Second , could there be a transcription factor ( s ) that determines cholinergic fate across different types of cholinergic cells in the vertebrate CNS ? While very limited information is available for the first question , it is interesting to note , for the latter question , that a LIM homeodomain ( LIM-HD ) transcription factor Isl1 is expressed in several cholinergic neurons in the spinal cord , hindbrain , forebrain and retina , such as spinal MNs , hindbrain MNs , some FCNs , and starburst amacrine cells [8] , [9] , [10] , [11] . Deletion of Isl1 gene results in a loss of MNs in the spinal cord and hindbrain [12] . Conditional deletion of Isl1 gene using a Six3-Cre transgene led to a reduction of restricted FCNs in the brain and cholinergic amacrine cells in the retina [13] . These findings point to the possibility that Isl1 may function as a cholinergic fate determinant in vertebrate CNS . However , it remains unknown whether Isl1 directly control the cholinergic phenotype and , if so , how Isl1 controls the fate of distinct cholinergic cell types whose gene expression patterns and functions are vastly different despite the shared property of cholinergic neurotransmission . In the developing spinal cord , Isl1 directs motor neuron fate specification by cooperating with another LIM-HD factor Lhx3 [12] , [14] , [15] , [16] . In differentiating MNs , Isl1 binds to Lhx3 and a LIM-interactor NLI ( also known as Ldb ) , thereby forming the Isl1-Lhx3-hexamer complex , also termed MN-hexamer ( Fig . S1A ) [14] , [17] . The combinatorial expression of Lhx3 and Isl1 , resulting in the formation of the Isl1-Lhx3-hexamer , is capable of triggering MN specification in chick spinal cord , embryonic stem cells ( ESCs ) , and induced pluripotent stem cells [14] , [17] , [18] , [19] , [20] . However , it is unclear whether the Isl1-Lhx3-hexamer directly controls cholinergic neuronal identity , an essential characteristic of MNs . In the developing forebrain , FCNs are derived from the medial ganglionic eminence ( MGE ) in the ventral telencephalon [21] , [22] . A LIM-HD protein Lhx8 is highly expressed in the MGE [21] , [23] . The formation of FCNs is severely disrupted in Lhx8-deficient mice [24] , [25] , [26] . Lhx8 appears to function in combination with Isl1 in driving the differentiation of cholinergic striatal interneurons [27] , but the mechanisms by which Lhx8 and/or Isl1 control cholinergic fates in the developing forebrain remain unclear . In this study , we found that the Isl1-Lhx3-hexamer directly activates the expression of a suite of cholinergic genes by binding to cholinergic gene enhancers that were discovered via ChIP-seq experiments . We also found that Isl1 is co-expressed with Lhx8 and NLI in the embryonic ventral forebrain and forms a hexamer complex with Lhx8 and NLI , named Isl1-Lhx8-hexamer . Interestingly , like the Isl1-Lhx3-hexamer in the spinal cord , the Isl1-Lhx8-hexamer directly controls cholinergic pathway gene expression via the same cholinergic gene enhancer in the forebrain . These findings imply that , despite distinct developmental histories and locations within the nervous system , MNs and some FCNs employ a common molecular mechanism that determines their cholinergic neuronal identity .
Given that MNs acquire cholinergic neuronal characteristics as they become specified , we considered the possibility that the Isl1-Lhx3-hexamer , a determinant of the MN fate , regulates expression of a battery of cholinergic genes by directly binding to the enhancer of each cholinergic gene . Intriguingly , our ChIP-seq analysis , which mapped the genomic binding sites of the Isl1-Lhx3-hexamer in mouse embryonic stem cells [20] , revealed Isl1-Lhx3-bound peaks in the key cholinergic pathway genes; ChAT , VAChT , high affinity choline transporter ( CHT , also known as Slc5a7 ) , a transporter that regulates the uptake of choline from the synaptic cleft into cholinergic neurons , and ATP-citrate lyase ( Acly ) , an enzyme that synthesizes acetyl-CoA ( Fig . 1A–C ) . ChAT has a strong peak within an intronic region that lies downstream of the VAChT gene , which is itself encoded within the intron of the ChAT gene . The Acly gene has two strong peaks in intronic regions and one upstream peak , while the CHT gene has a peak ∼100 kb downstream of its coding region . All the peaks have at least one hexamer response element ( HxRE ) ( Fig . 1C , Fig . S1B ) [20] . To test whether the Isl1-Lhx3-hexamer is recruited to Isl1-Lhx3-bound peak regions of the cholinergic genes in vivo , we purified genomic DNA bound by the Isl1-Lhx3-hexamer from E12 . 5 embryonic spinal cords using ChIP assays with α-NLI , α-Isl1 , and α-Lhx3 antibodies . All three components of the Isl-Lhx3-hexamer bound to the peaks in the cholinergic genes , while they did not bind to the genomic regions without the peaks ( Fig . 1D ) , indicating that the endogenous Isl1-Lhx3-hexamer is recruited to the cholinergic pathway genes in the developing spinal cord . Together , our unbiased , genome-wide ChIP-seq data , along with in vivo ChIP results , strongly suggest that the cholinergic pathway genes are directly activated by the Isl1-Lhx3-hexamer during MN fate specification . To test whether the Isl1-Lhx3-hexamer is capable of inducing the expression of multiple cholinergic genes in embryonic spinal cord , we misexpressed Isl1 and/or Lhx3 in the chick neural tube and monitored the expression of cholinergic genes . Co-electroporation of Isl1 and Lhx3 triggered the ectopic expression of a panel of cholinergic genes , including ChAT , VAChT , Acly and CHT , in the dorsal neural tube , while electroporation of Isl1 or Lhx3 alone did not ( Fig . 2A , Fig . S2 , data not shown ) . These data indicate that the Isl1-Lhx3-hexamer is capable of upregulating the cholinergic pathway genes in the developing spinal cord . To test whether Isl1 is needed for the cholinergic neuronal differentiation in the developing CNS , we deleted the Isl1 gene in neural progenitors using nestin-Cre [28] , [29] . In E12 . 5 Isl1f/f;nestin-Cre mice , Isl1 expression in MNs in the ventral spinal cord was greatly reduced ( Fig . 2B ) . In this condition , expression of cholinergic genes , such as Acly , ChAT , VAChT and CHT , is drastically downregulated ( Fig . 2B ) . The weak signal of VAChT was detected only in the remaining Isl1-expressing cells of Isl1f/f;nestin-Cre mice ( Fig . 2B ) . These results support a role of Isl1 in controlling cholinergic fate decision in the spinal cord . To test whether the Isl1-Lhx3-hexamer binding sites in the cholinergic genes act as enhancers to activate the cholinergic pathway genes in the embryonic spinal cord , we first examined whether the Isl1-Lhx3-hexamer activates the transcription of a reporter gene linked to each cholinergic gene peak , referred to here as ChAT-enh , Acly-enh1 and CHT-enh ( Fig . 3A–E ) , using luciferase reporter assays in mouse embryonic P19 cells . As NLI is expressed endogenously in P19 cells , co-expression of exogenous Isl1 and Lhx3 leads to the formation of the Isl1-Lhx3-hexamer [17] , [30] . The co-expression of Isl1 and Lhx3 strongly activated the Acly:LUC , ChAT:LUC and CHT:LUC reporters , but not LUC vector alone , in P19 cells , whereas the expression of Isl1 or Lhx3 alone did not ( Fig . 3A–C ) . The Acly:LUC reporter with point mutations in the HxRE was not activated by the co-expression of Isl1 and Lhx3 ( Fig . 3A ) . The DNA-binding defective forms of Lhx3 or Isl1 failed to synergize to activate Acly:LUC ( Fig . 3A ) , indicating that DNA-binding activity of both Isl1 and Lhx3 is needed to activate the reporter . To further test the role of the HxRE in each enhancer for the potent transcription response to the combination of Isl1 and Lhx3 , we generated luciferase reporters that are linked to multiple copies of the HxRE found within Acly-enh1 , ChAT-enh , and CHT-enh , respectively . These minimal HxRE reporters were also highly activated by co-expression of Isl1 and Lhx3 ( Fig . 3D , E , data not shown ) , establishing that the HxRE motif mediates activation of the cholinergic enhancers by the Isl1-Lhx3-hexamer . These data establish that the Isl1-Lhx3-hexamer is capable of activating each cholinergic enhancer in the Acly , ChAT/VAChT and CHT genes in heterologous cell types . To identify in vivo cell types in which the cholinergic enhancers activate gene expression in the developing spinal cord , we electroporated the neural tube of chick embryos with the GFP reporters linked to each cholinergic enhancer in ovo at a time when MNs are being specified . Interestingly , the Acly-enh1 drove strong GFP expression in MNs within the developing spinal cord ( Fig . 3F , S3A ) . In contrast , GFP was not expressed in non-MN cell types , despite efficient transfection of those cells following in ovo electroporation ( Fig . 3F , Fig . S3A ) , suggesting that only MNs have the transcriptional machinery that allows activation of Acly-enh1 . Additionally , Acly-enh1 with point mutations in the HxRE motif failed to activate target gene expression in MNs ( Fig . 3G , S3B ) , demonstrating that the HxRE motif is responsible for the MN-specific enhancer activity of the Acly-enh1 . Furthermore , the multimerized HxRE motifs from the Acly-enh1 were sufficient to drive GFP reporter expression in MNs ( Fig . 3H , Fig . S3C ) . Thus , the HxRE motif is necessary and sufficient for the MN-specific enhancer activity of the Acly-enh1 , suggesting that the endogenous Isl1-Lhx3-hexamer is responsible for Acly enhancer activity . Consistent with this idea , co-expression of Isl1 and Lhx3 , which assembles the Isl1-Lhx3-hexamer with endogenous NLI and triggers the formation of ectopic MNs in the dorsal spinal cord [14] , ectopically activated both Acly-enh1:GFP and Acly-HxRE:GFP reporters , but it failed to activate Acly-enh1 with mutations in HxRE motif ( Fig . 3F–H ) . These data indicate that the Isl1-Lhx3-hexamer is able to activate the Acly enhancer in the dorsal neural tube . The expression of Isl1 or Lhx3 alone failed to increase the transcriptional activity of the Acly-enh1 or Acly-HxRE ( Fig . S3D , E ) . Similarly to the Acly enhancer , the ChAT enhancer also directed gene expression specifically to MNs and became ectopically activated by the co-expression of Isl1 and Lhx3 ( data not shown ) . Together , these data demonstrate that the endogenous Isl1-Lhx3-hexamer binds to the cholinergic enhancers via the HxRE motifs and triggers the transcription of their target cholinergic genes as MNs are specified in embryonic spinal cords , establishing the Isl1-Lhx3-hexamer as a critical determinant of cholinergic neuronal identity in MNs . The coordinated upregulation of cholinergic pathway genes by the Isl1-Lhx3-hexamer in differentiating MNs , along with the previous loss-of-function studies suggesting that Isl1 and Lhx8 play important roles in the generation of FCNs [13] , [24] , raises the possibility that Isl1 and Lhx8 might form a complex similar to the Isl1-Lhx3-hexamer that drives cholinergic neuronal fate in the developing forebrain . To understand the role of Isl1 and Lhx8 in cholinergic gene expression in the developing forebrain , we determined the expression pattern of Isl1 , Lhx3 , and and NLI , which might form an Isl1-Lhx3-hexamer-like complex , using double immunohistochemistry analyses . All FCN precursors arise from the Nkx2 . 1-expressing MGE and preoptic area ( POA ) in the ventral telencephalon , and take two distinct migratory pathways; tangential migration to form striatal interneurons in the caudate-putamen ( CPu ) and radial migration to generate projection neurons in the basal forebrain ( Fig . 4A ) [21] , [31] . In E12 . 5 forebrain , Lhx8 expression is largely confined within the subventricular zone ( SVZ ) and mantle zone ( MZ ) of the MGE , but a few Lhx8+ cells were found in the MZ of the lateral ganglionic eminence ( LGE ) , which are likely the cells tangentially migrating from the MGE ( Fig . 4B , C ) . In contrast , Isl1 is more abundantly expressed in the SVZ and MZ of the LGE , but is also expressed in the SVZ and MZ of the MGE ( Fig . 4C , D ) . NLI is highly expressed in both MGE and LGE ( Fig . 4D , E ) . Thus , Isl1 , Lhx8 and NLI are co-expressed in a substantial fraction of cells in the MGE and LGE at E12 . 5 . A similar expression pattern for Isl1 , Lhx8 , and NLI was observed in E13 . 5 forebrain ( data not shown ) . By E16 . 5 , VAChT+ cholinergic neurons were readily detectable in the CPu and basal meganocellular complex ( BMC ) ( Fig . 5A , 6 ) . Although a majority of CPu cells are derived from the LGE , Nkx2 . 1+ progenitors in the MGE produce distinct subtypes of striatal interneurons in the CPu , including cholinergic interneurons [21] , [31] . Most cells in BMC , where a subset of cholinergic projection neurons is located , are generated from the Nkx2 . 1+ MGE [21] , [31] . In E16 . 5 brains , Isl1+ cells were much more abundant in the CPu than in the BMC , whereas Nkx2 . 1+ and Lhx8+ cells were more abundant in the BMC than in the CPu ( Fig . 5B , C , F , G ) , correlated with their expression at earlier developmental time points ( Fig . 4 ) . Despite this distinct pattern of gross expression , a number of Isl1+ cells co-expressed Nkx2 . 1 and Lhx8 in both the CPu and BMC , as shown by double immunohistochemistry analyses ( Fig . 5B , C , F , G ) . Given that Nkx2 . 1+ and Lhx8+ striatal interneurons are originated in the MGE [22] , [31] , a subset of Isl1+ cells in the CPu , which co-express Nkx2 . 1 and Lhx8 , is likely interneurons that are produced from the MGE . Isl1f/f;nestin-Cre mice die soon after E12 . 5 , precluding us from observing cholinergic neuronal differentiation in the forebrain . To understand the role of Isl1 in FCN specification in the forebrain , we generated Isl1f/f;Nkx2 . 1-Cre mice , in which Isl1 gene is deleted in cells derived from the Nkx2 . 1-expressing MGE [31] . As expected , in E16 . 5 Isl1f/f;Nkx2 . 1-Cre mice , the number of Isl1+ cells was greatly reduced in the BMC , but not in the CPu where most of Isl1+ cells were derived from LGE and thus did not express Nkx2 . 1-Cre ( Fig . 5B–I ) . In the CPu and BMC of the Isl1-conditional mutants , neither Isl1/Nkx2 . 1-double positive cells nor Isl1/Lhx8-co-expressing cells were found ( Fig . 5B–I ) , indicating that Isl1 is deleted in cells produced from Nkx2 . 1+ MGE and that Isl1/Lhx8-co-expressing cells in the CPu and BMC are derived from the MGE . Interestingly , in the CPu of Isl1f/f;Nkx2 . 1-Cre embryos , Lhx8+ and Nkx2 . 1+ interneurons were significantly reduced by ∼62% and ∼43% , respectively ( Fig . 5J , K ) , suggesting that Isl1 is required for specification of a subset of Nkx2 . 1+/Lhx8+ striatal interneurons . To monitor cholinergic neuronal differentiation , we performed immunostaining assays with VAChT antibodies . At E16 . 5 , cholinergic neurons were detected in the CPu and BMC , and co-expressed Isl1 and Lhx8 in both areas ( Fig . 6A–D ) . In E16 . 5 Isl1f/f;Nkx2 . 1-Cre mice , however , cholinergic neurons were almost eliminated in the CPu ( Fig . 6A , B ) . While ∼44% Lhx8+ cells and ∼35% Nkx2 . 1+ cells were cholinergic in the CPu of control embryos , almost all of Lhx8+ and Nkx2 . 1+ neurons did not express VAChT in the CPu of Isl1f/f;Nkx2 . 1-Cre mice ( Fig . 6E , F ) . The number of cholinergic neurons in the BMC area of Isl1f/f;Nkx2 . 1-Cre embryos appeared to be reduced compared to that in the littermate controls , but the heterogeneity of cholinergic neurons in the BMC made the quantification very challenging ( Fig . 6C , D ) . The remaining cholinergic neurons in the BMC expressed Lhx8 ( Fig . 6D ) . Similar to E16 . 5 , the number of cholinergic neurons remained markedly decreased in the CPu of E17 . 5 and P2 mice ( Fig . S4 ) . Together , our data indicate that Isl1 and Lhx8 are co-expressed in at least two different populations of FCNs in the CPu and BMC , and that Isl1 function in the MGE-derived cells is required for the specification of cholinergic interneurons in the CPu during forebrain development . The co-expression of Isl1 and Lhx8 in FCN precursors and FCNs and requirement of Isl1 and Lhx8 for the specification of a subset of FCNs ( Fig . 5 , 6 , S4 ) [13] , [24] support the possibility that Isl1 and Lhx8 cooperate for the FCN specification by forming a hexamer complex ( Fig . 7A ) , similar to the Isl1-Lhx3-hexamer ( Fig . S1 ) . The Isl1-Lhx3-hexamer assembly is dependent on the ability of Lhx3 to interact with Isl1 [14] . Another LIM-HD factor Lhx1 interacts with NLI , a common cofactor of the LIM-HD transcription factors , but does not bind to Isl1 , thus forming only a typical LIM tetramer complex consisting of 2NLI:2Lhx1 [14] . Thus , we investigated whether Lhx8 can interact with Isl1 , like Lhx3 , using in vitro GST-pull down assays ( Fig . 7B ) . As previously shown , Lhx3 interacted with Isl1 as well as NLI , whereas Lhx1 bound only to NLI . Interestingly , Lhx8 strongly associated with both Isl1 and NLI in vitro . Lhx8 also interacted with both Isl1 and NLI in HEK293 cells ( Fig . 7C ) . Combined with the notion that NLI strongly self-dimerizes [32] , our data supports a model by which Lhx8 , Isl1 and NLI can form a hexameric complex consisting of two NLIs , two Isl1s and two Lhx8 molecules ( Figure 7A ) . We refer to this complex as the Isl1-Lhx8-hexamer . To further test the formation of Isl1-Lhx8-hexamer complex , we examined whether Lhx8 interacts with NLIDD-Isl1ΔLIM , in which the dimerization domain ( DD ) of NLI is fused to LIM domains-deleted Isl1 ( Figure 7D ) . As NLIDD-Isl1ΔLIM lacks the LIM-interaction domain ( LID ) of NLI , it cannot bind to LIM-HD factors via typical interaction interfaces between NLI-LID and LIM-domains of LIM-HD factors , which lead to tetramer formation . Co-immunoprecipitation assays revealed that NLIDD-Isl1ΔLIM associated with Lhx8 in cells despite the lack of NLI-LID in this fusion ( Fig . 7D ) , further supporting the formation of the Isl1-Lhx8-hexamer in cells . Together , along with the fact that Lhx8 , Isl1 and NLI are co-expressed in FCN precursors in the ventral telencephalon , these results suggest that Lhx8 , Isl1 , and NLI form the Isl1-Lhx8-hexamer complex in the ventral telencephalon during development . To investigate whether the Isl1:Lhx8 dimer , the DNA-binding unit of the Isl1-Lhx8-hexamer , recognizes specific DNA sequences , we performed the unbiased screening method SELEX ( for systematic evolution of ligands by exponential enrichment ) assay with Isl1 , Lhx8 , or an Isl1-Lhx8 fusion , in which full-length Isl1 and Lhx8 proteins were linked by a flexible short linker ( Fig . 7E ) . Isl1-Lhx8 highly enriched a 15 nucleotide-long Isl1:Lhx8-binding motif after the third round of SELEX reaction , while Isl1 or Lhx8 failed to enrich any specific DNA sequences . The same motif was also isolated by SELEX with the mixture of Isl1 and Lhx8 , which were translated from Isl1-T2A-Lhx8 construct in vitro ( Fig . 7F ) , indicating that the Isl1-Lhx8-binding motif is not an artifact caused by use of the Isl1-Lhx8 fusion protein . These data indicate that Isl1:Lhx8 dimer in the Isl1-Lhx8-hexamer has high affinity to the specific DNA motif . Notably , Isl1-Lhx8-binding motif has a resemblance to the previously identified Isl1:Lhx3-site [17] , [20] , such as TAAT sequences , but also has unique features ( Fig . S5 ) . Considering the shared function of Isl-Lhx3-hexamer and Isl1-Lhx8-hexamer in inducing cholinergic genes and the similar features of their binding motifs , it is possible that they bind to the same enhancer regions of cholinergic pathway genes . To test whether , in the developing forebrain , the endogenous FCN-hexamer is recruited to the same cholinergic enhancers identified as targets of the Isl1-Lhx3-hexamer in our ChIP-seq analysis , we performed ChIP assays for the Isl1-Lhx8-hexamer using the dissected E15 . 5 embryonic forebrains and found that Isl1 , Lhx8 , and NLI , all components of the Isl1-Lhx8-hexamer , bound to the cholinergic enhancers ( Fig . 8A ) . These results suggest that the cholinergic enhancers recruit the Isl1-Lhx8-hexamer in the embryonic forebrain . To test the effect of the Isl1-Lhx8-hexamer on the transcriptional activity of cholinergic enhancers , we performed luciferase reporter assays in P19 cells using the Acly-HxRE:LUC and ChAT-HxRE:LUC reporter constructs . The co-transfection of Isl1 and Lhx8 activated each cholinergic enhancer , while expression of Isl1 or Lhx8 alone had minimal effect ( Fig . 8B , C ) . These results suggest that the Isl1-Lhx8-hexamer complex triggers the transcriptional activity of cholinergic enhancers . To investigate the activity of the Acly enhancer in the ventral forebrain , we injected the Acly-HxRE:GFP reporter along with the expression vectors encoding LacZ , Isl1 or Lhx8 into the ventral regions of E15 . 5 brain slices . The brain slices were then electroporated , cultured in vitro for four days , and examined for GFP expression ( Fig . 8D ) . Among many LacZ+ electroporated cells , only a small number of basal forebrain cells expressed GFP ( data not shown ) . The co-electroporation of Isl1 and Lhx8 along with the Acly-HxRE:GFP reporter drastically increased the number of GFP+ cells and the levels of GFP expression in the ventral forebrain , whereas expression of Lhx8 or Isl1 alone did not exhibit potent effects on Acly-HxRE:GFP . Likewise , the transfection of cortical progenitors using in utero electroporation revealed that the expression of Isl1-Lhx8 , but not Isl1 or Lhx8 alone , strongly activates Acly-HxRE in the developing cortex ( Fig . S6A ) . These results indicate that the combinatorial expression of Lhx8 and Isl1 promotes Acly-HxRE enhancer activity in the developing forebrain . Together , these data suggest that the Isl1-Lhx8-hexamer is sufficient to activate the cholinergic enhancers in heterologous cells and the developing forebrain . The binding and activation of cholinergic enhancers by Isl1-Lhx3 and Isl1-Lhx8 in the spinal cord and forebrain , respectively , prompted us to ask whether both complexes are capable of inducing the cholinergic gene battery irrespective of rostro-caudal positions within the CNS . To address this question , we misexpressed LacZ , Isl1 , Lhx8 , Lhx3 , Isl1-Lhx8 or Isl1-Lhx3 , along with EF1 promoter driven-GFP vector to mark the electroporated cells , in the E13 . 5 mouse cortex using in utero electroporation , and compared the expression levels of cholinergic genes between electroporated and control cerebral hemispheres at E18 . 5 using quantitative RT-PCR ( Fig . 9A ) . The expression level of transgenes was higher in electroporated sides than in control sides , as expected ( Fig . S7A ) . The expression of Isl1-Lhx8 substantially induced expression of ChAT , VAChT , and CHT in the cortex , compared to expression of Isl1 or Lhx8 alone ( Fig . 9B ) , indicating that Isl1 and Lhx8 function in combination to induce expression of cholinergic genes in the developing forebrain . Interestingly , Isl1-Lhx3 did not trigger cholinergic gene expression in the forebrain ( Fig . 9B ) , despite its potent activity to induce cholinergic pathway genes in the developing spinal cord ( Fig . 2A ) . Moreover , unlike the spinal cord , Isl1-Lhx3 failed to upregulate MN genes , Isl2 , Hb9 , and chodl [33] , in the forebrain ( Fig . S7B , data not shown ) , suggesting that the Isl1-Lhx3-hexamer is unable to turn on the MN gene program in the forebrain . Together , our results strongly support a model whereby the Isl1-Lhx8-hexamer orchestrates upregulation of a battery of cholinergic pathway genes in the developing forebrain . Our results that Isl1-Lhx3 failed to upregulate MN genes and cholinergic genes raise the question of whether Isl1-Lhx8-hexamer is functional in the spinal cord . To address this question , we expressed Lhx8 , Isl1 , Isl1-Lhx8 , or Isl1 plus Lhx8 in chick spinal cord using in ovo electroporation , and monitored cell differentiation and cholinergic gene expression three days post-electroporation . Lhx8 triggered ectopic generation of Chx10+ V2a interneurons in the dorsal spinal cord ( Fig . 9C , S7C ) like Lhx3 [14] , underlining the similarity between Lhx8 and Lhx3 . Co-expression of Isl1 with Lhx8 blocked Lhx8 from inducing V2a interneurons , suggesting that Isl1 binds to Lhx8 and changes the target gene specificity of Lhx8 as it does with Lhx3 [14] ( Fig . 9C , D , S7C ) . Interestingly , however , co-expression of Isl1 and Lhx8 induced neither ectopic Hb9+ MNs nor cholinergic genes in the dorsal spinal cord ( Fig . 9C , S7C ) . Likewise , Isl1-Lhx8 rarely triggered MN formation or cholinergic gene expression in the dorsal spinal cord ( Fig . 9C , S7D ) , indicating that the Isl1-Lhx8-hexamer is ineffective in activating cholinergic gene expression in the spinal cord . Together , our data highlight that the proper cellular context is critical for the Isl1-Lhx3-hexamer and Isl1-Lhx8-hexamer complexes to function in target gene regulation . Given that the Isl1-Lhx8-hexamer directly regulates the expression of cholinergic gene battery in the developing forebrain , it is possible that the Isl1-Lhx8-hexamer triggers cholinergic neuronal fate in stem cells . To test this possibility , we generated ESCs , in which the expression of Isl1-Lhx8 is induced by doxycycline ( Dox ) , namely Isl1-Lhx8-ESCs ( Fig . 10A , B ) . Isl1-Lhx8 forms the Isl1-Lhx8-hexamer with endogenous NLI in Dox-treated Isl1-Lhx8-ESCs ( data not shown ) . The expression of cholinergic pathway genes , ChAT , VAChT , CHT and Acly , but not a MN gene Hb9 , were readily induced by Isl1-Lhx8 under monolayer culture condition ( Fig . 10C , D ) , suggesting that the Isl1-Lhx8-hexamer controls the expression of cholinergic pathway genes in ESCs . We also monitored the cholinergic gene expression in floating culture of embryoid bodies ( EBs ) , which acquire the characteristics of forebrain neural precursors [34] . In the absence of Dox , many TuJ1+ neurons were observed in EBs , but VAChT+ neurons were hardly detected ( Fig . 10E , F ) . Dox treatment markedly induced VAChT+TuJ1+ cholinergic neurons in EBs ( Fig . 10E , F ) , suggesting that Isl1-Lhx8 triggers the cholinergic neuronal fate in stem cells . Likewise , RT-PCR also revealed that Isl1-Lhx8 significantly induced the expression of ChAT , VAChT and CHT in EB culture conditions ( Fig . 10G ) . In the same conditions , Isl1-Lhx8 did not induce the expression of MN genes , such as Hb9 , Isl2 , and Chodl ( Fig . 10G , data not shown ) . In contrast , Isl1-Lhx3 induced Hb9 as well as the cholinergic genes in both monolayer culture and floating embryoid bodies treated with retinoic acid and sonic hedgehog agonist ( Fig . 10H , data not shown ) [19] . Together , these results indicate that the Isl1-Lhx8-hexamer is capable of triggering the cholinergic neuronal fate , but not MN fate , in stem cells .
Establishment of correct neurotransmitter characteristics is an essential step of neuronal fate specification , but very little is known about how a battery of genes involved in a specific chemical-driven neurotransmission is coordinately regulated during vertebrate development . In this study , we report that Isl1 directly regulates a battery of genes establishing a cholinergic neurotransmitter characteristic in two developmentally unrelated cell types in vertebrate CNS ( Fig . 11 ) . Furthermore , we show that Isl1 does not do this alone , but performs its actions by forming two distinct cell type-specific transcription complexes , the Isl1-Lhx3-hexamer in the spinal cord and the Isl1-Lhx8-hexamer in the forebrain , both of which target common enhancer regions in each of the cholinergic pathway genes . In C . elegans , a set of dopamine pathway genes , which encode dopamine synthesizing enzymes and dopamine transporters , are co-regulated through a specific cis-regulatory element that is activated by the ETS transcription factor AST-1 [35] . Likewise , cholinergic pathway genes are co-regulated by a single transcription factor UNC-3 via UNC-3-binding motif in cholinergic MNs of C . elegans [7] . Does the vertebrate CNS with much more complex circuits utilize a similar strategy in establishing a particular neurotransmitter identity in multiple types of neurons sharing a neurotransmission system ? In vertebrate genome , gene regulatory motifs could occur far away from each gene transcription unit . Thus , identification of a common motif in a battery of neurotransmission-involved genes in vertebrates is much more difficult than in the nematode genome in which the regulatory sequences typically reside in proximity to the transcription start sites . While genome-wide unbiased ChIP-seq approaches could provide a solution to this challenging task , transcription factor ( s ) controlling a suite of neurotransmission genes need to be identified first to permit ChIP-seq analyses . Expression of Isl1 in multiple cholinergic cell types throughout CNS [8] , [9] , [10] , [11] suggests Isl1 as a good candidate factor to control cholinergic pathway genes . Loss-of-function studies established that Isl1 is required for cholinergic fate specification in spinal MNs , a subset of FCNs , and retinal amacrine cells ( this study ) [13] . Our study suggests that , to trigger cholinergic neuronal fate , Isl1 functions in combination with other proteins by forming cell type-specific transcription complexes; the Isl1-Lhx3-hexamer in the spinal cord and the Isl1-Lhx8-hexamer in the forebrain . Our ChIP-seq and subsequent analyses revealed that the core set of cholinergic pathway genes shares the binding motif , which recruits the Isl1-Lhx3-hexamer and the Isl1-Lhx8-hexamer in the embryonic spinal cord and forebrain , respectively , and is activated by these complexes . In addition to ChAT , VAChT , CHT and Acly , our ChIP-seq also uncovered the hexamer-binding peaks in other cholinergic pathway genes , such as acetylcholine esterase and a cluster of nicotinic acetylcholine receptors Chrna5/a3/b4 ( data not shown ) . An important area for future study is whether similar Isl1-containing complexes exist to control cholinergic fate decision in other areas of the CNS , such as retina and hindbrain . Isl1 is co-expressed with Phox2a , a paired-like homeodomain transcription factor , in the cranial MNs of the hindbrain [36] . Interestingly , a recent report shows that Isl1 associates with Phox2a and binds to the same cholinergic enhancer in the ChAT gene , which we identified in this study , when co-expressed with Phox2a [37] , raising a possibility that Isl1 forms a complex with Phox2a in the hindbrain MNs to control cholinergic gene expression . Together , these results strongly support a model in which the cholinergic pathway genes are concomitantly activated by cell type-specific Isl1-containing complexes during cholinergic neuronal differentiation in the developing CNS ( Fig . 11 ) . While the concept that a defined transcription factor controls the cholinergic gene battery is shared between nematodes and vertebrates , a clear difference is also noteworthy . In C . elegans MNs , a single transcription factor UNC-3 serves as a key regulator of the cholinergic pathway genes , whereas , in vertebrate CNS , Isl1-containing cell type-specific transcription complexes control the cholinergic gene battery . The combinatorial utilization of transcription factors is beneficial to generate massively divergent cell types in development . It is possible that regulation of the cholinergic genes by a single transcription factor in ancestral species has been diversified to a transcription complex in vertebrates , as the CNS circuitry becomes more complex . Another possibility is that the hexamer complexes and Ebf transcription factors , vertebrate UNC-3 orthologs , function cooperatively and/or redundantly to control cholinergic genes in the vertebrate CNS . Several findings support a possibility that the Isl1-containing hexamers act together with Ebf . Ebf proteins are expressed in differentiating MNs and ventral forebrain during embryonic development [38] , [39] , [40] . We found that Ebf1 associates with both types of hexamers in cells ( data not shown ) . Finally , our de novo motif analysis of the Isl1-Lhx3-hexamer-bound ChIP-seq peaks uncovered that the Ebf-binding site is enriched in a subset of the peaks ( data not shown ) . Thus , in the future , it will be interesting to investigate if Ebf factors collaborate with the hexamers in regulating cholinergic genes and other hexamer-targets . Our study demonstrated co-expression of Isl1 and Lhx8 in cholinergic neurons in the embryonic CPu and BMC . Interestingly , while Isl1 is required for cholinergic neuronal differentiation in the CPu , it is dispensable for differentiation of at least a subset of cholinergic neurons in the BMC . Lhx8 alone might be sufficient for the acquisition of the cholinergic phenotype in the remaining FCNs in the BMC . In addition , considering that some cholinergic neurons are still formed in Lhx8-deficient mice [24] , the Lhx8-independent pathway may be present to trigger cholinergic gene expression in the basal forebrain . Our finding that both Isl1-Lhx3-hexamer in spinal MNs and Isl1-Lhx8-hexamer in the forebrain bind to the HxRE motif in the cholinergic genes prompts the question of whether these two complexes share other target genes . Given the differences between MNs and FCNs in their functions , synaptic partners , and patterns of cell migration and axon trajectory , it is highly probable that the two complexes have largely separate sets of target genes , which establish MN- or FCN-specific characteristics , while sharing the cholinergic pathway genes as common targets . The Isl1-Lhx3-hexamer and Isl1-Lhx8-hexamer likely bind to similar but distinct sequences , and the HxREs in the cholinergic genes might have characteristics to be recognized by both Isl1-Lhx3-hexamer and Isl1-Lhx8-hexamer . In this respect , it is notable that the most optimal binding motifs for Isl1:Lhx3 or Isl1:Lhx8 identified by the SELEX methods exhibit unique features as well as shared sequences ( Fig . S5 ) [17] . The HxRE motifs in the cholinergic genes show variations from both Isl1:Lhx3- and Isl1:Lhx8-binding sequences ( Fig . S1B ) [17] , [20] . Isl1-Lhx8 failed to bind to the MN-specific enhancer of Hb9 , which recruits the Isl1-Lhx3-hexamer [30] , [41] ( data not shown ) , further suggesting that the Isl1-Lhx8-hexamer and the Isl1-Lhx3-hexamer have unique genomic binding sites . This idea is consistent with the recent finding that the genome occupancy of Isl1 substantially changes depending on whether Isl1 is expressed alone , or co-expressed with Lhx3 or Phox2a , each of which binds Isl1 [37] . Future studies to identify the genome-wide binding sites for the Isl1-Lhx8-hexamer and to compare the target genes and motifs among Isl1-containing cell type-specific complexes will provide important insights into one of fundamental questions of developmental biology; how a single transcription factor directs fates of multiple neuronal types with a common trait . Additional mechanisms likely operate for the Isl1-Lhx3 and Isl1-Lhx8 complexes to choose distinct sets of targets , given that the ability of Isl1-Lhx3-hexamer and Isl1-Lhx8-hexamer to activate target genes is highly dependent on the cellular context . Cholinergic genes were induced only by Isl1-Lhx3 in the spinal cord and only by Isl1-Lhx8 in the forebrain . Moreover , Isl1-Lhx3 readily activated MN genes in the developing spinal cord , but not in the forebrain . First , collaborating transcription factors or cofactors could contribute to the cell context-specific activation of the target genes for each hexamer complex ( Fig . 11 ) . The Isl1-Lhx3-hexamer has been shown to cooperate with Neurog2 ( Ngn2 ) and Stat3 in MN gene regulation [20] , [30] . It will be interesting to test whether the Isl1-Lhx8-hexamer interacts with other transcription factors , such as Mash1 , Olig2 , Dbx1/2 or Gbx1/2 , to control FCN differentiation in the ventral forebrain . Second , the in vivo chromatin context may play a role in cell type-specific gene expression . For instance , MN genes , such as Hb9 and Isl2 , may possess transcription-permissive chromatin environment in the spinal cord and transcriptionally inactive chromatin in the forebrain , thus allowing the gene activation by the Isl1-Lhx3-hexamer only in the developing spinal cord , but not in the forebrain . In this regard , it is noteworthy that the activation of Acly-HxRE:GFP reporter gene , which is free from chromain-mediated regulation , is cell context-independent . Both Isl1-Lhx3 and Isl1-Lhx8 was capable of activating the Acly-HxRE:GFP reporter in both the developing spinal cord and forebrain ( Fig . 3 , S6 ) . Together , our study provides key insights into the gene regulatory logic of cholinergic neuronal differentiation , which would be useful to generate cholinergic neurons for therapeutic or drug screening purposes .
All animal procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee ( IACUC ) at OHSU . Rat Isl1 , Isl1-N230S , and mouse Lhx3 , Lhx3-N211S , Lhx8 , Lhx1 , Isl1-T2A-Lhx3 , Isl1-Lhx3 fusion , Isl1-Lhx8 fusion , NLI , and LacZ genes were cloned into pCS2 , pcDNA3 ( Invitrogen ) containing a HA , Flag or myc-epitope tag , or pCIG for expression in mammalian cells and chick embryos and for in vitro transcription and translation reactions . All of these vectors except Lhx8 were previously described [14] , [30] , [41] . NLIDD-Isl1ΔLIM is a fusion of 1-298aa of NLI containing the self-dimerization domain of NLI and 111-349aa of Isl1 , which is a C-terminal region of Isl1 that does not include the LIM domains . Isl1-N230S and Lhx3-N211S are missense mutatnts , which are deficient in their ability to directly bind DNA [14] . Isl1 , Lhx3 , Lhx8 , Isl1-Lhx3 and Isl1-Lhx8 were also cloned into the pCIG-2 vector for electroporation of mouse brains . Isl1 and NLI were cloned into the bacterial expression vector pGEX4T-1 ( Amersham ) for in vitro GST-pull down experiments . Lhx8 and NLI were cloned into the mammalian GST expression vector pEBG for GST-pull down experiments in cell lines . Isl1-Lhx8 was generated by fusing Isl1 full-length and Lhx8 full-length via flexible linker GGSGGSGGSGG . Isl1-T2A-Lhx8 was generated by inserting T2A sequences between full-length Isl1- and full-length Lhx8-coding sequences . The location of Isl1-Lhx3-bound ChIP-seq peaks for cholinergic genes in mouse genome ( mm9 ) is the following; ChAT/VAChT , chr14:33256618–33257117; CHT , chr17:54298028–54298480; Acly , chr11:100381966–100382465 , chr11:100379377–100379876 , and chr11:100395147–100395645 . Mouse genomic regions covering the Acly-enhancer , ChAT-enhancer and CHT-enhancer were amplified using PCR , and two or three copies of these enhancers were cloned into TK-LUC or synthetic TATA-GFP reporter vectors . Primers to amplify these genomic enhancers are Acly-enahncer1 , forward 5′- GA AGA TCT TGA TAG CAC ACT ACT TTG CTC TGG , reverse 5′- CG GGA TCC CAG TGA CGC ACG GCG AGC GGG AAG; ChAT-enhancer , forward 5′- GA AGA TCT TAC TAA TTG GAT TAA TTG ATT TGC , reverse 5′- CG GGA TCC GGG AAT TAA TAA CTT AGA ATT TGA; CHT-enhancer , forward 5′- GA AGA TCT TGA GCA GCC TAT GCC ACA AGG ACA , reverse 5′-CG GGA TCC AGG AAT CCA TCA CAA AGC TAA GAC . AAGCTGATTA sequences in Acly-enh1 were mutated to CCGCGCGGCC to generate the Acly-enh1-HxRE-mt reporter . Acly-HxRE:LUC , Acly-HxRE:GFP , ChAT-HxRE:LUC and ChAT-HxRE:GFP reporters were created by cloning multiple copies of the following duplex oligonucleotides into synthetic TATA-GFP or TK-LUC vectors . Acly-HxRE , 5′- CAG AGC TAAT CAG CTTG AGTG GGT-3′; ChAT-HxRE 5′- TGG TAC TAAT TGG ATTA ATTG ATT-3′ . The generation of Isl1f/f , Nestin-Cre , and Nkx2 . 1-Cre mice has been described previously [28] , [29] , [31] . Isl1f/f mice were crossed with Isl1f/+;NesticCre mice or Isl1f/+;Nkx2 . 1Cre mice to generate Isl1f/f;NesticCre or Isl1f/f;Nkx2 . 1Cre embryos , respectively , for analyses . Mouse embryos were collected at the indicated developmental stages , and fixed in 4% paraformaldehyde , embedded in OCT and cryosectioned in 12 µm thickness for immunohistochemistry assays or 18 µm thickness for in situ hybridization with digoxigenin-labeled probes . These assays were performed as described [14] , [42] . In chick electroporation assays , DNAs were injected into a ∼ Hamburger and Hamilton ( HH ) stage 13 chick neural tube . The embryos were harvested 3 days post-electroporation and fixed in 4% paraformaldehyde , embedded in OCT and cryosectioned in 12 µm thickness for immunohistochemistry assays or 18 µm thickness for in situ hybridization with digoxigenin-labeled probes . Each set of chick electroporation experiments was repeated independently three to six times with at least three embryos injected with the same combination of plasmids for each experimental set . Representative sets of images from reproducible results were presented . For immunohistochemistry assays , the following antibodies were used; rabbit anti-Hb9 [43] , mouse anti-Mnr2/Hb9 ( 5C10 , DSHB ) , rabbit anti-Isl1/2 [9] , guinea pig anti-Chx10 [43] , rabbit anti-Lhx3 [15] , guinea pig anti-VAChT ( AB1588 , Millipore ) , goat anti-ChAT ( AB144P , Millipore ) , rabbit anti-GFP ( A6455 , Molecular Probes ) , rabbit α-Nkx2 . 1 ( Santa Cruz ) , guinea pig α-Lhx8 ( generated using mouse Lhx8 211–367aa region as antigen ) , rabbit anti-NLI [44] , TuJ1 ( Covance ) and mouse anti-Flag ( Sigma ) . For in situ hybridization analyses , cDNA for mouse ChAT , Acly and CHT and chick CHT , Acly , VAChT and ChAT were cloned to pBluescript vector and these vectors were used to generate digoxigenin-labeled riboprobes . HEK293T cells were seeded onto 10 cm tissue cultures dishes , cultured in DMEM media supplemented with 10% fetal bovine serum , and transfected using Superfect ( Qiagen ) . 48 hours after transfection , cells were harvested and lysed in IP buffer ( 20 mM Tris-HCl , pH 8 . 0 , 0 . 5% NP-40 , 1 mM EDTA , 150 mM NaCl , 2 mM PMSF , 10% Glycerol , 4 mM Na3VO4 , 200 mM NaF , 20 mM Na-pyroPO4 , and protease inhibitor cocktail ) . In these studies , precipitations were performed with either α-Flag antibody ( Sigma ) or glutathione sepharose beads ( GE-Healthcare ) . The interactions were monitored by western blotting assays using α-Flag ( Sigma ) and α-HA ( Babco ) antibodies . Following western blotting with fluorescence-labeled secondary antibodies , the bound fractions of proteins were scanned by the Odyssey imaging system ( Li-Cor ) following western blotting with fluorescence-labeled secondary antibodies . In vitro GST-pull down assays were performed as described [45] . BL21 E . coli were transformed with pGEX vector alone , pGEX-Isl1 , or pGEX-NLI to express the GST-fusion proteins and lysed by sonication . The GST-fusion proteins were purified by incubating the lysates with glutathione sepharose beads ( GE-Healthcare ) . The beads were then washed and incubated with the putative interacting partners Lhx8 , Lhx3 and Lhx1 , which were generated in vitro by using the TnT T7 Quick Coupled transcription/translation system ( Promega ) . Bound proteins were eluted by boiling , and were monitored by western blotting assays using α-HA ( Babco ) antibodies and Odyssey imaging system ( Li-Cor ) . SELEX was performed as described [46] with proteins in vitro transcribed and translated from the following vectors; Flag- tagged Isl1-Lhx8 fusion , Flag-Isl1 , Flag-Lhx8 , and Isl1-T2A-Lhx8 which produce both Flag-Isl1 and HA-Lhx8 proteins . The proteins , which were generated by using the TnT T7 Quick Coupled transcription/translation system ( Promega ) , were incubated with a pool of double-stranded oligonucleotides containing a central core region of 22 random nucleotides with identical 5′- and 3′-flanking regions . For each SELEX reaction , ∼30 clones were randomly selected and sequenced . The motif analysis was conducted using Multiple Em for Motif Elicitation ( MEME ) [47] . P19 embryonic carcinoma cells were cultured in α-minimal essential media supplemented with 2 . 5% fetal bovine serum ( FBS ) and 7 . 5% bovine calf serum . For luciferase assays , P19 cells were seeded and incubated for 24 hours , and transient transfections were performed using Lipofectamine 2000 ( Invitrogen ) . An actin promoter-β-galactosidase plasmid was cotransfected for normalization of transfection efficiency , and empty vectors were used to equalize the total amount of transfected DNA . Cells were harvested 36–40 hours after transfection . Cell extracts were assayed for luciferase activity and the values were normalized with β-galactosidase activity . Data are presented as means of triplicate values obtained from representative experiments . All transfections were repeated independently at least four times . Luciferase reporter data are shown in relative activation fold ( mean +/− standard deviation ) . The overall procedures for ventral forebrain electroporation and organotypic slice culture were previously described [48] . E15 . 5 mouse embryos were harvested and brains were dissected and embedded in 3% low melting point agarose dissolved in complete Hanks Balanced Salt solution ( cHBSS ) . 250 µm thick slices of the brains were generated using a Leica VT1200 vibratome . Slices containing the appropriate regions of the ventral forebrain were focally injected with combinations of plasmids . The slices were then mounted on the anode above a 1 mm agarose slice and cHBSS was used to gap the cathode , and electroporated using ECM 830 electroporator ( BTX ) under the following condition; 60 mV , 5 ms interval pulse , 500 ms delay , and 5 pulses . Immediately after the electroporation , the slices were transferred to transwell inserts ( 0 . 4 µm pore size ) and cultured for three to five days in vitro with slice media containing 5% heat inactivated horse serum added below the insert at 37°C with 5% CO2 . Slices were fixed in 4% paraformaldehyde , washed in PBS and analyzed post-fix using immunofluorescence histochemistry . The overall procedures for ex vivo brain electroporation and organotypic slice culture were previously described [49] . E15 . 5 mouse embryos were harvested and then the heads were removed and placed in cHBSS . Each combination of DNA constructs mixed with 0 . 5% Fast Green ( Sigma ) were injected into the lateral ventricles of isolated E15 . 5 mouse heads using a Picospritzer III microinjector . The electroporation was carried out on whole heads using ECM 830 electroporator ( BTX ) under the following condition; 30 mV , 100 ms intervals , 4 pulses , and 100 ms delay . For organotypic slice culture , brains were dissected immediately following electroporation , and embedded in 3% low melting point agarose dissolved in cHBSS . 250 µm thick slices of the brains were generated using a Leica VT1200 vibratome and transferred to transwell inserts ( 0 . 4 µm pore size ) . The slices were then cultured for three to five days in vitro with slice media containing 5% heat inactivated horse serum added below the insert at 37°C with 5% CO2 . Slices were fixed in 4% paraformaldehyde and analyzed for GFP expression . Each set of mouse brain electroporation experiments was repeated independently three to six times . For each set of mouse brain electroporation , three to four brain slices were electroporated per condition . Reproducible results were presented in the figures . Confocal images were acquired using a Nikon Eclipse Ti inverted microscope with perfect focus and a motorized stage coupled to a 4 laser line A1 scanning confocal system . Representative sets of images were presented . For in utero electroporation , timed-pregnant C57BL/6N females were anesthetized at stage E13 . 5 with isoflurane ( 4% during induction , 2 . 5% during surgery ) , and the uterine horns were exposed by way of laparotomy . 1 µℓ of the expression vector in PBS containing 0 . 05% fast green ( Sigma-Aldrich , St Louis , MO , USA ) was injected into the lateral ventricle of the embryo using a 100 mm glass capillary ( 1B100-4 , World Precision Instruments , Inc . , USA ) . Electroporation was performed using Tweezertrodes ( diameter , 5 mm; BTX , Holliston , MA , USA ) with 5 pulses of 45 V for 50 millisecond duration and 950 millisecond intervals using a square-wave pulse generator ( ECM 830; BTX ) . The uterine horns were then returned to the abdominal cavity , the wall and skin were sutured , and embryos were allowed to continue their normal development and collected for the further analyses at indicated stages . Total RNAs were extracted using the Trizol ( Invitrogen ) and reverse-transcribed using the SuperScript III First-Strand Synthesis System ( Invitrogen ) . For quantitative PCR of ChAT , VAChT , Acly and CHT , the following probes and primers predesigned by the TaqMan Gene Expression Assay ( Applied Biosystems ) for each gene were used with TaqMan Universial Master MixII and 7500 ABI qPCR machine ( Applied Biosystems ) ; ChAT ( Assay ID-Mm01221882_m1 ) , VAChT ( Assay ID-Mm00491465_s1 ) , Acly ( Assay ID-Mm01302282_m1 ) , CHT ( Assay ID- Mm00452075_m1 ) and Eukaryotic 18S rRNA Endogenous Control ( FAM Dye/MGB Probe , Non-Primer Limited ) . In addition , the following primers were used with the SYBR green kit ( 11762-500 , Invitrogen ) and Mx3000P ( Stratagene ) . Hb9 , 5′-GTT GGA GCT GGA ACA CCA GT , 5′-CTT TTT GCT GCG TTT CCA TT; ACLY , 5′-GAA GCT GAC CTT GCT GAA CC , 5′-CTG CCT CCA ATG ATG AGG AT; ChAT , 5′-CCT GCC AGT CAA CTC TAG CC , 5′-GGA AGC CTT TAT GAT GAG AA; CHT , 5′-GTG GTC TAG CTT GGG CTC AG , 5′-GGC AAT GAG TGC AGA GAC AA; VAChT , 5′-TTG ATC GCA TGA GCT ACG AC , 5′-CCA CTA GGC TTC CAA AGC TG; Hb9 , 5′-GTT GGA GCT GGA ACA CCA GT , 5′-CTT TTT GCT GCG TTT CCA TT; Isl2 , 5′-GCA AAC TCG CTG AGT GCT TTC , 5′-ACC ATA CTG TTG GGG GTG TC; Chodl , 5′-CAG TGG AAT GAC GAC AGG TG , 5′-GGT TCC CAA AGC AAC CAG TA; Isl1 , 5′-GAC ATG ATG GTG GTT TAC AGG C , 5′- GCT GTT GGG TGT ATC TGG GAG; Lhx3 , 5′-AGA GCG CCT ACA ACA CTT CG , 5′-GGC CAG CGT CTT TCT TCA GT; Lhx8 , 5′-CAG TTC GCT CAG GAC AAC AA , 5′-AGC CAT TTC TTC CAA CAT GG; GAPDH , 5′-ACC ACA GTC CAT GCC ATC AC , 5′-TCC ACC ACC CTG TTG CTG TA; and Cyclophilin A , 5′-GTC TCC TTC GAG CTG TTT GC , 5′-GAT GCC AGG ACC TGT ATG CT . RT-PCR experiments were performed with three or four independent sets of samples . Data are represented as the mean of duplicate or triplicate values obtained from representative experiments . Error bars represent standard deviation . The ChIP-seq data used for the analysis in this paper has been deposited in the Gene Expression Omnibus ( GEO ) database ( assession no . GSE50993 ) [20] . To perform the ChIP assays with mouse embryonic tissues , we dissected E12 . 5 spinal cords or E15 . 5 forebrains . The microdissected spinal cords from five E12 . 5 embryos or the forebrains of three E15 . 5 embryos were combined together for each ChIP reaction with a specific antibody . The tissues were dissociated completely , fixed by 1% formaldehyde for 10 min at room temperature , and quenched by 125 mM glycine . Next , cells were washed with Buffer I ( 0 . 25% Triton X-100 , 10 mM EDTA , 0 . 5 mM EGTA , 10 mM Hepes , pH 6 . 5 ) and Buffer II ( 200 mM NaCl , 1 mM EDTA , 0 . 5 mM EGTA , 10 mM Hepes , pH 6 . 5 ) sequentially . Then , cells were lysed with lysis buffer ( 0 . 5% SDS , 5 mM EDTA , 50 mM Tris-HCl , pH 8 . 0 , Protease inhibitor cocktail ) and were subjected to sonication for DNA shearing . Next , cell lysates were diluted 1∶10 in ChIP buffer ( 0 . 5% Triton X-100 , 2 mM EDTA , 100 mM NaCl , 50 mM Tris-HCl , pH 8 . 0 , Protease inhibitor cocktail ) and , for immunoclearing , were incubated with IgG and protein A agarose beads for one hour at 4°C . Supernatant was collected after quick spin and incubated with appropriate antibodies and protein A agarose beads to precipitate the hexamer/chromatin complex for overnight at 4°C . After pull-down of the hexamer/chromatin complex/antibody complex with protein A agarose beads , the beads were washed with TSE I ( 0 . 1% SDS , 1% Triton X-100 , 2 mM EDTA , 20 mM Tris-HCl , pH 8 . 0 , 150 mM NaCl ) , TSE II ( same components as in TSE I except 500 mM NaCl ) and Buffer III ( 0 . 25M LiCl , 1% NP-40 , 1% deoxycholate , 1 mM EDTA , 10 mM Tris-HCl , pH 8 . 0 ) sequentially for 10 minutes at each step . Then the beads were washed with TE buffer three times . The hexamer/chromatin complexes were eluted in elution buffer ( 1% SDS , 1 mM EDTA , 0 . 1M NaHCO3 , 50 mM Tris-HCl , pH 8 . 0 ) and decross-linked by incubating at 65°C overnight . Eluate was incubated at 50°C for more than two hours with Proteinase K . Next , DNA was purified with Phenol/chloroform and DNA pellet was precipitated by ethanol and resolved in water . The purified final DNA samples were subjected to quantitative PCR reactions using the SYBR green kit ( 11762-500 , Invitrogen ) and Mx3000P ( Stratagene ) . The total input was used for normalization . All ChIP experiments were repeated independently at least three times . Data are represented as the mean of duplicate or triplicate values obtained from representative experiments , and error bars represent standard deviation . The following primers were used for ChIP-PCR . ChAT-enhancer forward 5′-TAC TAA TTG GAT TAA TTG ATT TGC reverse 5′-GGG AAT TAA TAA CTT AGA ATT TGA ChAT-negative forward 5′- CTG TGG CTC ATA ACG CTC ATT TTG reverse 5′- AGT TTG TGG TGG GCC GAG ATG GCA Acly-enh1 forward 5′- TGA TAG CAC ACT ACT TTG CTC TGG reverse 5′-CAG TGA CGC ACG GCG AGC GGG AAG CHT-enhancer forward 5′-TGA GCA GCC TAT GCC ACA AGG ACA reverse 5′- CAT TAG GAG AGC TTG TTC CAG TGA The following antibodies were used for ChIP-PCR; mouse/rabbit IgG ( Santa Cruz ) , rabbit anti-Isl1 [9] , rabbit anti-Lhx3 [15] , rabbit anti-NLI [44] , and goat anti-Lhx8 ( sc-22216 , Santa Cruz ) . The generation of Isl1-Lhx3-ESCs was described previously [19] . To generate Isl1-Lhx8-ESCs , the A172LoxP ES cell line [50] was maintained in an undifferentiated state on 0 . 1% gelatin-coated dishes in the ESC growth medium that consisted of Knockout DMEM , 10% FBS , 0 . 1 mM non-essential amino acids , 2 mM L-glutamine , 0 . 1 mM β-mercaptoethanol and recombinant leukemia inhibitory factor ( LIF , 1000 units/ml , Chemicon ) . Flag-tagged Isl1-Lhx8 fusion was inserted into Tet-inducible plasmid p2Lox . The Isl1-Lhx8 vector was co-transfected with pSALK-Cre into the A172LoxP ES cell line using Lipofectamine 2000 ( Invitrogen ) . Stable transfectants were isolated by selection with neomycin ( G418 , 400 µg/ml ) for seven days . Dox-dependent induction of Flag-Isl1-Lhx8 expression was monitored by western blotting and immunohistochemical analyses using α-Isl1 , α-Lhx8 and α-Flag antibodies . To induce cell differentiation , Embryoid bodies ( EBs ) were formed and cultured for 2 days using the hanging drop method ( 1×103 ESCs per 20 µℓ drop ) . Hanging drops were transferred to suspension culture in 6 well low attachment dishes and cultured . EBs were cultured without or with doxycycline ( 2 µg/ml ) for 2–5 days in the ESC medium without LIF or in the differentiation medium that contains KnockOut serum replacement ( Life technologies ) . Then , EBs were collected for either RT-PCR or immunohistochemical analyses . | Neurons utilize various chemicals to transmit signals to a target cell . Distinct types of neurons in the spinal cord and forebrain , collectively termed cholinergic neurons , utilize the same chemical , acetylcholine , for signal transmission . These neurons play critical roles in controlling locomotion and cognition . In this study , we have found that the Isl1 gene orchestrates the process to generate cholinergic neurons in the spinal cord and forebrain . Isl1 forms two different types of multi-protein complexes in the spinal cord and forebrain . Both complexes bind the same genomic regions in a group of genes critical for cholinergic signal transmission , and promote their simultaneous expression . These cholinergic genes include enzymes that synthesize acetylcholine and proteins required to package acetylcholine into vesicles . The Isl1-containing multi-protein complexes were able to trigger the generation of cholinergic neurons in embryonic stem cells and neural stem cells . Our study reveals crucial mechanisms to coordinate the expression of genes in the same biological pathway in different cell types . Furthermore , it suggests a new strategy to produce cholinergic neurons from stem cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience",
"biology",
"and",
"life",
"sciences",
"developmental",
"biology"
] | 2014 | Isl1 Directly Controls a Cholinergic Neuronal Identity in the Developing Forebrain and Spinal Cord by Forming Cell Type-Specific Complexes |
Protein design algorithms enumerate a combinatorial number of candidate structures to compute the Global Minimum Energy Conformation ( GMEC ) . To efficiently find the GMEC , protein design algorithms must methodically reduce the conformational search space . By applying distance and energy cutoffs , the protein system to be designed can thus be represented using a sparse residue interaction graph , where the number of interacting residue pairs is less than all pairs of mutable residues , and the corresponding GMEC is called the sparse GMEC . However , ignoring some pairwise residue interactions can lead to a change in the energy , conformation , or sequence of the sparse GMEC vs . the original or the full GMEC . Despite the widespread use of sparse residue interaction graphs in protein design , the above mentioned effects of their use have not been previously analyzed . To analyze the costs and benefits of designing with sparse residue interaction graphs , we computed the GMECs for 136 different protein design problems both with and without distance and energy cutoffs , and compared their energies , conformations , and sequences . Our analysis shows that the differences between the GMECs depend critically on whether or not the design includes core , boundary , or surface residues . Moreover , neglecting long-range interactions can alter local interactions and introduce large sequence differences , both of which can result in significant structural and functional changes . Designs on proteins with experimentally measured thermostability show it is beneficial to compute both the full and the sparse GMEC accurately and efficiently . To this end , we show that a provable , ensemble-based algorithm can efficiently compute both GMECs by enumerating a small number of conformations , usually fewer than 1000 . This provides a novel way to combine sparse residue interaction graphs with provable , ensemble-based algorithms to reap the benefits of sparse residue interaction graphs while avoiding their potential inaccuracies .
Frequently , the goal of a protein design problem is to find the lowest energy sequences or conformations . Most protein design algorithms use pairwise energy functions to score protein conformations . Any such protein design problem can be represented by a residue interaction graph , where the nodes represent residues , and edges represent the interaction between residues . The energy functions usually consist of distance-dependent terms to model van der Waals and electrostatic interactions between residue pairs , and the interaction energy decreases with increasing distance between residues . Therefore , it is possible to neglect interaction energies between distant residues and not add them to the overall energy of a conformation . This eliminates edges between these negligibly-interacting residues from the residue interaction graph to construct a sparse residue interaction graph , with the corresponding GMEC called the sparse GMEC . We will refer to the residue interaction graph with no edges eliminated as the full residue interaction graph , and the corresponding GMEC as the full GMEC . Whether explicitly described or not , the concept of sparse residue interaction graphs is ubiquitous in the field of protein design . Many design algorithms apply appropriate distance cutoffs implicitly in the energy function ( using different cutoffs for different kinds of energies calculated ) [24 , 45–52] , while others develop new algorithms to take explicit advantage of the sparseness of the residue interaction graph [53–60] . By using sparse residue interaction graphs , the number of interacting residue pairs is fewer than all pairs of mutable residues , and this reduces the effective search space considerably . However , the energies omitted by deleting the edges between negligibly interacting residues can add up , causing differences in sequence ( such as amino acid identity ) of the GMEC returned or the rankings of the top sequences ( specific examples are discussed in detail in the Sections entitled “Results” and “Discussion” ) . While small sequence differences might not be consequential , larger differences in energies and sequences can lead to design algorithms returning a protein sequence that may not have the desired function . Therefore there is a potential tradeoff between reducing the search space and guaranteeing the accuracy and quality of the computed GMEC . Despite the widespread use of sparse residue interaction graphs in protein design , the effects of this tradeoff have not been previously analyzed . In this paper , we present the results of our analysis of using sparse residue interaction graphs in protein design . We implemented a variation of the A* search algorithm in the protein design software osprey [43] , for design with sparse residue interaction graphs to return the corresponding GMEC . osprey has been used in many successful designs in vitro [11 , 13–15 , 17 , 18 , 20] , and even in vivo [14 , 15 , 18 , 20] . We ran computational experiments on a total of 136 protein design problems , involving core , boundary , and surface residues . We used different energy and distance cutoffs to generate the sparse residue interaction graph , and analyzed the sequence and energy differences between the different GMECs returned . Our results show that commonly used distance cutoffs can return a GMEC whose sequence is different than that of the GMEC returned without those cutoffs . The underlying assumption when using distance and energy cutoffs is that neglecting long-range interactions do not have an effect on local interactions . We show that , contrary to this assumption , neglecting long-range interactions can alter favorable local interactions . Changes to the sequence and loss of favorable interactions between residues can both result in structural and functional changes to the predicted protein . Next , in order to study if the sequence differences between the full and the sparse GMEC lead to functional differences , we performed retrospective validation on 6 protein design problems for which experimentally determined thermal stability data was available , and analyzed the sequences differences between the GMECs returned with and without distance cutoffs . Our analysis shows that across all 6 design problems , the sparse and full GMEC predicted different amino acid identities at 13 residues . Out of these 13 residues that have a different amino acid identity in the two GMECs , the more thermostabilizing mutation is predicted by the GMEC of the sparse residue interaction graph for 7 residues , and by the GMEC of the full residue interaction graph ( without using distance cutoffs ) for the remaining 6 residues . This indicates that there is no clear trend on which of the two GMECs will predict mutations with the desired function in vitro . Moreover , it can be difficult to correctly choose between the GMEC of the full residue interaction graph and its less computationally expensive sparse equivalent . Therefore it is beneficial to compute the GMECs for both the full and the sparse residue interaction graph , and to do so efficiently , while still taking advantage of the computational benefits of the reduced search space induced by the sparse residue interaction graph . To achieve this goal , we provide a novel approach , called Energy-bounding enumeration , to combine sparse residue interaction graphs with provable , ensemble-based algorithms to generate both the GMECs efficiently . The gap-free list of low-energy conformations returned by an ensemble-based provable algorithm is guaranteed to contain the GMEC for the full residue interaction graph [59] . From this list , we prove that this GMEC can be found in additional O ( kn2 ) time , where n is the number of mutable residues , and k is the number of conformations generated . We show that in practice , the full GMEC is almost always found within the first 1000 conformations returned . Because the number of conformations required to capture the GMEC is usually small , protein designers can henceforth combine sparse residue interaction graphs with provable , ensemble-based algorithms to exploit the reduced search space and still compute the GMECs for both the full and the sparse residue interaction graph . In short: sparse residue interaction graphs induce substantial differences in predicted sequences , conformations , and energies , with no way of telling which model will best predict the desired function . But provable , ensemble-based algorithms rescue computational protein design from these difficulties by providing a way to compute both GMECs efficiently . In particular , this paper makes the following contributions:
Each protein design problem is defined by its input model , namely , the input protein structure , the mutable residues , the allowed amino acids at each mutable residue , allowed side-chain conformations , and energy function . Given this input model , the interaction energy between the mutable residues can be represented as an undirected graph , where vertices represent mutable residues , and edges represent pairwise interactions . An edge is present between two vertices when the pairwise energies between the two corresponding residues are included in the energy function . When the input energy function models all interactions between all mutable residues as pairwise energies , the residue interaction graph representing these interactions is the complete graph . We will refer to the complete residue interaction graph as the full graph . Every edge in the graph corresponds to a pairwise interaction between two mutable residues . By applying distance or energy cutoffs , the pairwise interactions between some mutable residues are omitted from the energy function . For every pair of residues ( vertices ) whose interactions are omitted , the corresponding edge between that pair is deleted from the residue interaction graph . This sparse graph , whose omitted edges correspond to the pairwise interactions omitted by the energy function , is called the sparse residue interaction graph . We will refer to the GMEC ( which encodes both the conformation and sequence ) of the full graph as the full GMEC and the GMEC of the sparse graph as the sparse GMEC . We will show that the full GMEC and sparse GMEC can be different , in both conformation and sequence . For any given conformation , we will refer to its computed energy with respect to the full graph as its full energy , and its energy with respect to the sparse graph as its sparse energy . For convenience , we will use δ and α to refer to distance and energy cutoffs , respectively: Fig 1 shows the full and sparse residue interaction graphs for a protein design problem with 8 mutable residues . In this section , we have given high-level intuition for sparse residue interaction graphs , and their corresponding energy functions . For the proofs of Lemma 1 and Lemma 2 in S1 Text , however , precise definitions are useful to provide a mathematical basis for our claims . Hence , in S2 Text , we provide formal definition of a residue interaction graph , the sparse residue interaction graph , and the corresponding GMECs computed using such interaction graphs . We also provide a mathematical model for the cutoff criteria used to prune pairwise interactions from a residue interaction graph . To study the sequence differences between the full and the sparse GMEC caused due to neglecting some pairwise energies , we need to compute the sparse GMEC first . We use the protein redesign package developed by the Donald lab , osprey for this study [43] . This section describes the changes made to osprey to compute the sparse GMEC , and the computational experiments designed to study the differences between the full and the sparse GMEC . osprey uses dead-end elimination ( DEE ) followed by the A* search algorithm [41 , 42 , 61 , 62] to provably return the GMEC and enumerate conformations in order of increasing energy . DEE prunes a significant portion of the possible conformations , and following that , A* search explores the unpruned search space to ensure that the first conformation returned is the GMEC . After the GMEC is returned , A* continues to enumerate conformations in increasing order of energy , until either all conformations within an energy window Ew of the GMEC are enumerated , or the number of conformations returned reaches a user defined number . A* guarantees that all conformations within Ew of the GMEC are returned , and the list of conformations enumerated is gap-free . To generate the sparse GMEC , we modified the A* search algorithm used by osprey to calculate the sparse energy of a conformation ( Eq . 2 in S2 Text ) . We will refer to this variant of the A* search algorithm as Sparse A* . Sparse A* evaluates conformations based on sparse energy ( as opposed to full energy in traditional A* search ) , and can now be used for protein design with sparse residue interaction graphs . As Sparse A* retains all the guarantees provided by the A* search algorithm ( because the algorithm is unmodified ) , the first conformation returned by Sparse A* is guaranteed to be the sparse GMEC , and it can also return a gap-free list of conformations within Ew of the sparse GMEC in increasing order of sparse energy . This property of Sparse A* is used to prove a surprising result , namely , that Sparse A* can efficiently generate not only the sparse GMEC , but also the full GMEC . This will be discussed later in the Section entitled “Discussion” . Computational experiments were performed on the following design problems to generate the full and the sparse GMEC , and subsequently the energy , conformational , and sequence differences were analyzed: In all experiments , the DEE pruning stage was followed by either A* to get the full GMEC , or the following two steps to generate the sparse GMEC and gap-free list of conformations: 1 ) sparse residue interaction graph generation using a user-defined distance cutoff δ or energy cutoff α , and 2 ) Sparse A* run to generate the sparse GMEC . For each design problem , Sparse A* was run four times using the following distance or energy cutoffs: To further investigate how the differences in predicted sequence and conformation between the full and the sparse GMEC correlate with experimental measurements , we performed retrospective validation against 6 full-sequence designs from the literature , for which the designed mutants were experimentally determined to have improved thermal stability over the wild type [64–67] . Each example taken from the literature consisted of a protein redesign with an input structure together with experimentally measured melting point measurements showing a more thermostable designed mutant compared to the wild type sequence . We then performed computational redesign on the input structure . Consecutive residues of one or more adjacent secondary structures were allowed to either retain the wild-type identity or mutate to the amino acid identity of the designed , thermostabilized mutant . To compute the full GMEC , DEE pruning was followed by A* search . To compute the sparse GMEC , DEE pruning was followed by the generation of the sparse residue interaction graph with distance cutoff 7 Å , and then Sparse A* search . The number of mutable residues varied from 10-19 residues . The full and the sparse GMEC were then correlated against the measured melting point data . The input model consists of a rigid backbone , rigid , discrete side-chain rotamers , and a pairwise energy function . All designs were done keeping the backbone fixed and modeling side-chain flexibility using the modal values of rotamers from the Penultimate rotamer library [22] . The energy function consisted of the amber van der Waals and electrostatic terms and the EEF1 pairwise implicit solvation model , as described in [11 , 43] . Protein design formulations that consider additional side-chain flexibility [29 , 32] , backbone flexibility [28 , 30 , 31] , free energy calculations [44 , 61] , or more accurate energy functions [68] have been developed . Nevertheless most of them call as a subroutine the simplified model discussed in this paper , which can be viewed as a core calculation common to most protein design software . Hence , the accuracy of this computation bounds the accuracy of the overall design . For more details on the input model , such as protein structures , mutable residues , and design protocol , please refer to S3 Text .
For 62 core designs , Sparse A* with both distance cutoffs returned the sparse GMEC identical to the full GMEC for all 62 design problems , and in 59 design problems with energy cutoff α = 0 . 1 kcal/mol , and in 58 design problems with energy cutoff α = 0 . 2 kcal/mol . For the four core problems where the full GMEC was different from the sparse GMEC , the full GMEC was the second conformation returned by Sparse A* , and the energy difference between the full and the sparse GMEC was less than the energy cutoff of 0 . 2 kcal/mol . Out of these four design problems , two had sequence differences between the full and the sparse GMEC: the human sulfite oxidase cytochrome b5 domain ( PDB id: 1MJ4 ) and bacterial iron-sulfur protein ( PDB id: 3A38 ) had single amino acid differences between the full and the sparse GMEC ( residue 50 for 1MJ4 and residue 26 for 3A38 ) . In both cases , serine in the full GMEC was replaced by alanine in the sparse GMEC . Except for these two cases , distance and energy cutoffs did not have sequence-changing effects on the GMEC returned for core designs . Interestingly , using an energy cutoff of 0 . 2 kcal/mol results in omitting a large fraction of residue pairs for most of the core design problems , between 45% to 80% ( S1 Fig ) . Despite the tightly packed nature of the protein core , the energy interactions were less than 0 . 2 kcal/mol , which is less than the typical van der Waals interaction energy of 0 . 5-1 kcal/mol . Unlike core designs , the number of boundary design problems where the sparse GMEC was identical to the full GMEC was larger for energy cutoffs than distance cutoffs , as shown in Fig 3 ( a ) . The energy cutoff of α = 0 . 1 kcal/mol gave the best results , returning the sparse GMEC identical to the full GMEC in 19 out of the 21 boundary design problems . For problems where the full GMEC and the sparse GMEC were different , the full energy and sequence difference between the full and the sparse GMEC are larger for distance cutoffs than for energy cutoffs , ( Fig 3 ( b ) and 3 ( c ) ) , with the distance cutoffs introducing sequence differences in a total of 24 residues as compared to 11 residues with energy cutoffs ( S1 Table ) . For a single design problem , this number can be as high as 6 ( C-terminal domain of the Rous Sarcoma Virus capsid protein , PDB id: 3G21 ) , which is more than one-third of the 15 mutable residues for that design problem . Similar to boundary design problems , energy cutoffs returned a sparse GMEC which was identical to the full GMEC in more cases than distance cutoffs did . The sparse GMEC is identical to the full GMEC in 10 out of the 12 surface design problems for energy cutoff α = 0 . 1 kcal/mol , and only for 1 out of 12 for distance cutoff δ = 7 Å ( Fig 4 ( a ) ) . Sequence differences between the full and the sparse GMEC occur even though the energy differences between these GMECs are small . Unlike boundary designs , where in a few cases the energy cutoffs introduced more sequence differences between the full and the sparse GMEC , the energy cutoff of α = 0 . 1 kcal/mol has a smaller or equal number of sequence differences than distance cutoffs in all 12 surface design problems , as shown in Fig 4 ( c ) . Overall , distance cutoffs introduced sequence differences in a total of 25 residues , as compared to 9 residues for energy cutoffs ( S2 Table ) . Since distances between residues on the surface of the protein are larger as compared to boundary or core regions , Sparse A* was run for the 12 surface designs problems again with a distance cutoff δ = 10 Å . This resulted in the sparse GMEC being identical to the full GMEC for 5 out of 12 design problems , as compared to 3 out of the 12 design problems for δ = 8 Å . For one additional case ( bacterial oxidized ferredoxin protein , PDB id: 1IQZ ) , the number of amino acid sequence differences was reduced . However , this still only increases the number of design problems for which the full GMEC and the sparse GMEC are identical to half of the 10 when using the energy cutoff of α = 0 . 1 kcal/mol . Overall , increasing the distance cutoff from 8 Å to 10 Å decreased amino acid differences in only 3 out of the 9 design problems . In human CD59 glycoprotein ( PDB id: 2J8B ) , the number of amino acid sequence differences is 5 with distance cutoffs of 7 Å , 8 Å , and 10 Å , and increasing the distance cutoff led to no change in the sequence difference between the full and the sparse GMEC whatsoever . The above results suggest that using distance cutoffs can neglect long range interactions between residue pairs , and cause significant sequence differences between the GMECs returned ( S2 Fig , S1 and S2 Tables ) . Using a larger distance cutoff ( 10 Å ) did little to improve the results . Neglecting these long range interactions tends to have a larger effect on boundary and surface designs than on core designs . To investigate this further , the maximum energy ( in absolute value ) contributed over all rotamer pairs by each of the edges deleted using distance and energy cutoffs were analyzed . To eliminate any uncertainties , the results of core , boundary , and surface designs on the same protein structure were used for this analysis . Fig 5 shows the distribution of omitted pairwise energies of sparse graphs generated with either distance or energy cutoffs for 36 different protein design problems over 6 different structures . For the 6 protein structures shown in Fig 5 , using distance cutoffs in boundary and surface designs can delete edges from the residue interaction graph with larger interaction energies , as compared to using energy cutoffs . The opposite occurs for core designs . This is consistent with the fact that both the distance cutoff δ = 7 Å and energy cutoff α = 0 . 2 kcal/mol had similar results for core designs , but for boundary and surface designs , the number of residues with different amino acids between the full and the sparse GMEC is larger for the distance cutoff than for the energy cutoff , except for bacterial cytochrome C-553 protein ( PDB id: 1C75 ) . By definition , energy cutoffs only delete an edge when its maximum absolute energy contribution is smaller than the specified limit . By contrast , distance cutoffs delete edges whose energy contributions can vary arbitrarily from being very small ( 0 . 05 kcal/mol ) to being very large ( almost 0 . 9 kcal/mol ) . As such , our results foreground a key difference between distance cutoffs and energy cutoffs: in terms of the energy contributed by each edge , precomputed energy cutoffs are more precise . While distance cutoffs omit any sufficiently distant pairwise interaction , providing limited control over the energy contributions of the omitted edges , energy cutoffs will never omit any pairwise interaction that can exceed the specified energy cutoff . Therefore energy cutoffs allow greater precision in selection of low-energy pairwise interactions . One reason distance cutoffs are widely used is the assumption that sufficiently distant interactions do not affect local interactions . Our results indicate that this is not always true . Figs 6 and 7 show the sequence differences between the full GMEC and the sparse GMEC with distance cutoff δ = 8 Å in two such examples from bacterial cytochrome C-553 protein ( PDB id: 1C75 ) and a domain of pneumococcal histidine triad A protein ( PDB id: 2CS7 ) . In both cases , neglecting the interaction between the distal residues ( red ) and the two proximal residues ( cyan ) leads to a missing hydrogen bond . In Fig 6 , the amino acids of the full GMEC are replaced with entirely different amino acids . In the sparse GMEC , residue 17 is an arginine instead of a lysine , and residue 32 is a histidine instead of a glutamic acid . While residue 17 and residue 32 of the full GMEC form a hydrogen bond , in the sparse GMEC the arginine at residue 17 forms hydrogen bonds with the backbone instead . In Fig 7 , the amino acids at residues 56 and 59 are swapped between the full and the sparse GMEC . These examples illustrate two cases in which neglecting long-range interactions can disrupt favorable local interactions . In general , for the design problems analyzed in this paper , the disruption of local interactions is more common in surface designs . For 4 surface design problems , the interaction between two mutable residues is different in the full and the sparse GMEC . In 3 of these cases a favorable local interaction is lost by using distance cutoff 7 Å . To study how well the sequence differences between the full and the sparse GMEC correlate with experimental measurements , we conducted retrospective design experiments . As described in the methods section ( entitled “Computational experiments” ) , we took examples of wild-type proteins from the literature which were computationally redesigned to improve thermostability ( the designed protein had higher Tm than the wild-type protein ) [64–67] , and redesigned a subset of each protein , allowing mutable residues to either retain their wild-type identity or mutate to the corresponding amino acid identity of the more stable designed mutant . We computed the full and the sparse GMEC for all 6 design problems and then compared the GMECs to identify differences in sequence , energy , and conformation . We correlated the difference in sequence with experimentally measured melting temperature data . Amino acid identities were restricted in the redesign procedure to ensure that any difference in sequence between the full and the sparse GMEC would correspond to either the less or the more stable protein sequence , and hence could be directly validated against the melting temperature data . The designed search space ranged from 6 . 65 × 1015 to 6 . 13 × 1025 conformations ( see the section entitled “Rank of full GMEC in practice” ) . For 5 of the 6 protein design problems , amino acid differences between the full and sparse GMEC were found . In the sixth design problem , the sparse GMEC and full GMEC predicted the same sequence , but have different side-chain conformations . Table 1 lists the residue numbers for the 5 problems in which the full and sparse GMEC differed in sequence . For the two residues that have different amino-acid identity between the full and the sparse GMEC for the B1 domain of protein L ( PDB id: 1HZ5 ) , the full GMEC predicts the amino acid identities in the more stable designed protein for both Lys 4 and Glu 26 . In contrast , for the U1 nuclear ribonucleoprotein A ( PDB id: 1URN ) the sparse GMEC predicts the amino acid identity in the more stable protein for both Gln 93 and Asp 97 . For the designed engrailed homeodomain dimer ( PDB id: 2MG4 ) , the sparse GMEC predicts the amino acid identity of the more stable protein for both Arg 52 and Glu 54 . For the remaining two protein design problems , at some residues the full GMEC predicts the amino acid identity of the more stable designed protein , and for other residues the sparse GMEC predicts the amino acid at the more stable designed protein instead . Comparison of the complete sequences of the sparse and full GMEC can be found in the S3 Table . In summary , for the 5 design problems , there are 13 residues where the sparse and full GMEC predicted different amino acids . For these 13 residues , the amino acid identity of the more stable designed protein is predicted by the full GMEC for 6 residues , and by the sparse GMEC for the other 7 residues . These results suggest that when using a rigid backbone , rigid rotamer , GMEC-only input model and sparse or full pairwise energy function , it is unclear which of the two GMECs ( sparse or full ) will correspond to the desired protein function ( in this case , improved thermostability ) . We then analyzed the sparse residue interaction graph to determine the significance of the omitted edges . In particular , we identified pairwise interactions whose omission would change the sequence of the computed GMEC . Table 2 lists these omitted pairwise interactions , the minimum distance ( closest Euclidean inter-residue distance between any two atoms when all rotamer combinations for the two residues are considered , see the Section entitled “Sparse residue interaction graphs and protein design” ) between the interacting residue pair , the total difference in energy between the full and sparse GMEC , and the difference in energy contributed by these omitted edges to the sparse and full GMEC . The omission of the pairwise interactions listed in Table 2 alone was large enough to change the sequence of the computed GMEC , and even lead to changes in experimental measurements . For example , in the case of acyl phosphatase ( PDB id: 2ACY ) , Fig 8 shows the sequence differences and key high-energy long-range interactions omitted in the sparse residue interaction graph . In the full GMEC the minimum distance between residues Glu 12 and Lys 76 is 7 . 6 Å , and its high-energy long-range interaction of -0 . 34 kcal/mol is omitted in the sparse residue interaction graph . The minimum distance between residues Asp 10 and Arg 77 is 8 . 6 Å , and its high-energy long-range interaction of -0 . 26 kcal/mol is omitted in the sparse residue interaction graph . The corresponding pairwise energy between residues Glu 12 and Asp 76 in the sparse GMEC is 0 . 318 kcal/mol , and the corresponding pairwise energy between residues Glu 12 and Asp 76 in the sparse GMEC is 0 . 314 kcal/mol . This amounts to a total difference of 1 . 24 kcal/mol . The energies of the sparse and full GMEC differ by only 0 . 75 kcal/mol . As can be seen in panel ( c ) of Fig 8 , the sparse GMEC neglects the energetically unfavorable interactions between both the negatively charged glutamic acid at residue 12 and aspartic acid at residue 76 , and the positively charged lysine at residue 10 and arginine at residue 77 . These unfavorable long-range electrostatics are not found in the full GMEC , as seen in panel ( d ) . Omitting these two pairwise interactions alone would change the sequence of the corresponding GMEC . Note that in this design problem , large , favorable electrostatic pairwise interactions in the full GMEC are replaced with large , unfavorable electrostatic interactions in the sparse GMEC . The cumulative difference is greater than 1 . 2 kcal/mol , which is large enough to be biophysically relevant . Even when the sparse energy function predicts amino acid identities of the more thermostable designed mutant , the difference in energy between the sparse GMEC and full GMEC can manifest as a difference in backbone coordinates and side-chain conformations . We analyzed an example of this , where the full GMEC predicts a rotamer that closely resembles the wild type while the sparse GMEC predicts a rotamer which has a χ1 angle difference of 90 . 6° . For human procarboxypeptidase A2 ( PDB id: 1AYE ) , the sequences of the full and sparse GMECs were identical , but the residue conformations differed . To test if sparse residue interaction graphs could be a source of conformational difference between the predicted and experimentally observed residue conformations , we performed side-chain placement using osprey on the designed mutant of human procarboxypeptidase A2 ( PDB id: 1VJQ ) . The structure of the designed mutant was used as input to rule out backbone changes as an additional source of error . We analyzed the conformations of the sparse GMEC , the full GMEC , and the crystal structure , and found the predicted conformations of sparse and full GMEC differed at two residues . At residue Glu 18 , the rotamer of the full GMEC coincides closely with the crystal structure , whereas the rotamer of the sparse GMEC had a 90 . 6° difference in its χ1 angle , differing significantly from crystal structure . At residue Lys 22 , both the sparse and the full GMEC preserve the overall direction of the charged amine group , but their χ-angles differ from the crystal structure . In the side-chain placement problem for the designed mutant of human procarboxypeptidase A2 ( PDB id: 1VJQ ) , two pairwise interactions contributed the most to the energy difference between the sparse and full GMEC: the pairwise interactions between residues Glu 12 and Lys 22 , and residues Glu 18 and Asp 24 . The minimum distance between residues Glu 12 and Lys 22 is 9 . 9 Å , and in the full GMEC their long-range interaction of -0 . 42 kcal/mol is omitted in the sparse residue interaction graph . The corresponding pairwise energy between residues 12 and 22 in the sparse GMEC is -0 . 32 kcal/mol . The minimum distance between residues Glu 18 and Asp 24 is 8 . 6 Å , and in the full GMEC its long-range interaction of 0 . 21 kcal/mol is omitted in the sparse residue interaction graph . The corresponding pairwise energy between residues Glu 18 and Asp 24 in the sparse GMEC is 0 . 32 kcal/mol . The energy difference from these two edges account for a cumulative difference of 0 . 21 kcal/mol . The energies of the sparse and full GMEC differ by 0 . 16 kcal/mol . Omitting these two pairwise interactions alone would change the conformation of the computed GMEC .
As described in the Section entitled “Definitions related to sparse residue interaction graphs” , sparse residue interaction graphs are generated by omitting distant or low-energy pairwise interactions from the energy function . These sparse graphs correspond to modified energy functions . Naturally , modifying the energy function can change the GMEC . The conformation and sequence of the GMEC of the full graph may be different from the GMEC of the sparse graph . However , for any pair of mutable residues , the maximum and minimum contributions of their pairwise interactions bound their total contribution to any conformation . Furthermore , their maximum and minimum contributions are efficient to compute . Thus , the contribution of any pairwise interaction can be efficiently bounded , and the cumulative maximal and minimal contribution of all omitted pairwise interactions can be efficiently bounded as well . These cumulative maxima and minima can be mathematically combined to bound the energy difference between the sparse and full GMEC ( Lemma 1 , S1 Text ) . Although the bounds given by Lemma 1 are often loose , it is important to know that the energy difference between the full and the sparse GMEC can always be bounded . This energy bound guarantees that a gap-free list ( enumerated by a provable algorithm such as Sparse A* ) , that contains the sparse GMEC and all conformations within that energy bound of the sparse GMEC , will contain both the sparse and full GMEC . If we simply compute the full energies of all conformations in this list , then the conformation with the lowest full energy is guaranteed to be the full GMEC . Furthermore , computing the full energies for all conformations in the list is efficient ( Lemma 2 , S1 Text ) . The fact that Sparse A* can enumerate both the sparse and the full GMEC means that we no longer have to worry about which of the two sequences ( the full or the sparse GMEC ) will predict the desired functional mutations . One can generate the sparse residue interaction graph , calculate the upper bound on the energy difference , and run Sparse A* to return a gap-free list of conformations that is guaranteed to return the full GMEC . The list of conformations returned by Sparse A* can then be re-ranked based on the full energy to get the full GMEC . Once both the full and the sparse GMEC are found , they can be evaluated based on more sophisticated methods for energy calculation [68 , 70] , or any other method that seems pertinent to the designer . Note that both the full and the sparse GMEC are guaranteed to be found only when using provable ensemble-based algorithms that are guaranteed not to miss any conformation within the specified energy window . The re-ranking can be done relatively quickly ( Lemma 2 in S1 Text ) , when the number of conformations that need to be generated by Sparse A* to find the full GMEC is not large . Both the full and the sparse GMEC were found for most of the design problems used in this study , and this is discussed in the Section entitled “Rank of full GMEC in practice” . In this section , we have provided high-level intuition showing how Sparse A* can be used to compute both the sparse and full GMEC . These statements are supported by mathematical guarantees , which show that the full GMEC is contained in the gap-free list enumerated by Sparse A* , and that it is efficient to compute the full GMEC from that list . In S1 Text we provide two Lemmas and their proofs . Lemma 1 proves an upper bound on the absolute difference in sparse energy between the sparse and full GMEC , and Lemma 2 gives the time complexity to compute the full GMEC from a gap-free list guaranteed to contain the full GMEC . We then describe how these two proofs are sufficient to compute both the sparse and full GMEC using Sparse A* . Finally , we briefly describe a recent provable algorithm [59] , which uses concepts from dynamic programming to exploit the optimal substructure induced by sparse graphs , and achieve asymptotic time complexity significantly better than the worst-case time complexity of any algorithm using the full graph . Fig 9 plots the calculated upper bounds vs . the actual full energy difference between the full and the sparse GMEC for the core , boundary , and surface designs . It is evident from the difference in the scale of the two axes that the actual energy difference between the full and the sparse GMEC ( ranges from 0 . 05 kcal/mol to 1 . 6 kcal/mol ) is an order of magnitude smaller than the computed upper bound , which can be as high as 40 kcal/mol . As a result , Sparse A* returned the full GMEC relatively early , well before all conformations within the energy bound ( calculated using Lemma 1 in S1 Text ) of the sparse GMEC were enumerated . The full GMEC was found within the first 20 conformations of the sparse GMEC for all 21 boundary design problems with energy cutoffs , and for 19 design problems with distance cutoffs . For the two remaining problems , ClpS protease adaptor protein ( PDB id: 3DNJ ) and bacterial ferredoxin protein ( PDB id: 1IQZ ) , the full GMEC was the 168th and 5062nd conformation returned by Sparse A* respectively with distance cutoff δ = 7 Å . The rank of the full GMEC in the gap-free list of conformations enumerated by Sparse A* for all 21 boundary design problems are given in Table 3 . For the 12 surface design problems , the full GMEC is within the first 30 conformations of the sparse GMEC for energy cutoffs , but for distance cutoffs , this number is on the order of a few hundred for some of the design problems . The rank of the full GMEC in the gap-free list of conformations enumerated by Sparse A* for all 12 surface design problems are given in Table 4 . Table 5 shows the rank of the full GMEC in the gap-free , in-order list enumerated after applying a distance cutoff δ = 7 Å for the six retrospective design problems discussed in the Section entitled “Retrospective validation against experimental data” ( Table 1 and Fig 8 ) , along with the search space of each design problem . Note that even after constraining the mutable residues to only two allowed amino acids , the search space size for these designs can be large . The search space of the largest retrospective design problem , a 19-residue design of the engrailed homeodomain dimer ( PDB id: 2mg4 ) , was 6 . 13 × 1025 conformations . Even in this case the rank of the full GMEC was merely 19 in the gap-free , in-order list , and our provable enumeration algorithm efficiently computes both . ( For these experiments , computing 20 additional conformations after computing the sparse GMEC took less than 7 . 5 minutes . ) The search space of the design problem that required the most enumeration , an 18-residue design of acyl phosphatase , was 6 . 64 × 1024 , and enumerating 240 conformations to recover the full GMEC took less than 48 minutes . Across the six design problems , the median time to compute the sparse GMEC was 8 . 81 minutes , and the median time to compute the top 1000 conformations was 46 . 2 minutes . These times show how computing the top 1000 conformations is far less costly than computing 1000 GMECs . Therefore , using a provable algorithm to enumerate the 1000 lowest-energy conformations is in most cases very practical . Overall , the energy difference between sparse and full GMEC was small , and therefore the rank of the full GMEC in the gap-free list enumerated by Sparse A* was low . For all but one design problem ( including the core , boundary , surface , and retrospective designs ) , the full GMEC was found by enumerating only the first 1000 conformations returned by Sparse A* ( with both distance and energy cutoffs ) . This shows that even when limited time and memory prevent Sparse A* from provably enumerating the full GMEC ( because of loose energy bounds ) , in practice the number of conformations that must be enumerated before Sparse A* returns the full GMEC can be small . This provides useful information that can be used to compute the full GMEC using Sparse A* for design problems where A* fails . This is highlighted by the three boundary and one surface design problem for which Sparse A* with distance cutoff δ = 7 Å ( the cutoff which deleted the most of edges ) returned the sparse GMEC , whereas A* ran out of 30 GB of memory before returning the full GMEC ( orange points in Fig 2 ) . Sparse A* returned the sparse GMEC along with a gap-free list of conformations for three boundary designs ( heterogeneous nuclear ribonucleoprotein K ( PDB id: 1ZZK ) , Beta-elicitin cinnamomin ( PDB id: 2AIB ) , Dihydrofolate reductase type 2 ( PDB id: 2RH2 ) , and for one surface design of scorpion toxin protein ( PDB id: 1AHO ) ) . The number of conformations enumerated by Sparse A* was 47 for 1ZZK , 3029 for 2AIB , 46 for 2RH2 , and 10 , 000 for 1AHO . Given the results that with distance cutoff δ = 7 Å the full GMEC can be found almost always within the first 30 conformations for boundary designs , and within 1000 conformations for surface designs , the gap-free list computed by Sparse A* for the above four protein design problems almost certainly contains the full GMEC . Because the number of conformations that must be enumerated by Sparse A* to find the full GMEC is usually small , the GMECs of both the full and the sparse residue interaction graphs can be computed by enumerating a gap-free , in-order list of conformations . Note that this study relies critically on provable algorithms that are guaranteed to enumerate the GMEC followed by a gap-free list of conformations in order of increasing energy . Without these algorithms it would be difficult and perhaps even unsound to compare the results of computational protein design with and without sparse residue interaction graphs , since differences induced by the sparse model can not be deconvolved from differences stemming from undersampling or inadequate stochastic optimization . Moreover , the provable guarantees of Lemma 2 ( S1 Text ) would not be possible if the enumeration algorithm missed any low-energy conformations within the calculated energy window of the sparse GMEC . It has been previously argued that crucial improvements to the energy function and input model ( e . g . side-chain flexibility , backbone flexibility , and entropy ) should , for reasons of computational complexity , be accompanied by novel algorithmic enhancements [39] . Hence , it is also important to distinguish design algorithms that only apply distance cutoffs to the energy function [24 , 45–52] vs . algorithms that exploit the optimal substructure induced by sparse residue interaction graphs ( via techniques such as dynamic programming ) [53–60] . While algorithms that only modify the energy function and algorithms that effectively exploit the optimal substructure both benefit from the reduced effective search space of sparse residue interaction graphs , significant large-scale gains in computational efficiency ( including reduced asymptotic time complexity ) are not achieved by the former , whereas they are guaranteed by the latter . By developing new , efficient methodologies and algorithms ( such as Energy-bounding enumeration , used in the Section entitled “Discussion” ) , larger , harder problems built on more sophisticated biophysical input models become tractable without any increase in hardware capability . So even though physical hardware power may not improve quickly enough to relieve the computational costs of protein design , algorithmic improvements can reduce previously difficult or even intractable tasks in protein design to well-understood problems for which a wealth of efficient algorithms already exist . One example is a recent paper [70] , which reduces the problem of protein design with continuous side-chain flexibility to the well-understood problem of protein design with discrete rotamers , enabling designers to perform designs with continuous side-chain flexibility as efficiently as they could previously perform discrete rigid rotamer designs . In this paper , we implemented a variant of the A* search algorithm in our lab’s open source protein design package osprey , for protein design with sparse residue interaction graphs . We ran A* and Sparse A* on 136 different protein design problems involving core , boundary , and surface residues and analyzed the effects of using various distance and energy cutoffs . We compared the energies and sequences of the full GMEC returned by A* vs . the sparse GMEC returned by Sparse A* , and found that distance cutoffs , especially in surface and boundary designs , can lead to significant sequence differences between the full and the sparse GMEC . Our analysis indicates that the effects of distance cutoffs range from introducing no sequence differences in core designs , to sequence differences in almost all surface designs . By comparison , the effects of energy cutoffs are similar across core , boundary , and surface designs . In addition , we show examples of protein design problems in which neglecting long-range interactions alters local interactions . Furthermore , we performed retrospective designs for proteins with experimentally measured data , and our analysis of sequence differences between the full and the sparse GMEC indicates that it is not readily apparent if the sparse or full GMEC predicts mutations that perform better in vitro . The sequence differences between the full and the sparse GMEC occur even though the energy differences between these GMECs are small . While these errors are severe , we provided a way to overcome these discrepancies . We used a provable , ensemble-based algorithm and showed that the full GMEC was found within the first 1000 conformations returned for all but one of our design problems . Because the number of conformations that must be enumerated to find the full GMEC is usually small , we can take advantage of the reduced search space provided by sparse residue interaction graphs and still efficiently compute both the full and the sparse GMEC . To do this , we compute a gap-free , in-order list of conformations . The gap-free list of low-energy conformations returned by Sparse A* is guaranteed to contain the full GMEC , and we show it takes only polynomial additional time to find the full GMEC ( Lemma 2 in S1 Text ) . For 3 boundary and 1 surface design problem where A* failed to return even a single conformation , Sparse A* not only computed the sparse GMEC , but also enumerated a gap-free list of conformations that almost certainly contains the full GMEC . This provides a novel way for provable , ensemble-based algorithms and sparse residue interaction graphs to compute not only the sparse GMEC , but also the full GMEC for previously intractable design problems . Previous studies have found that computational structure-based protein design protocols are often susceptible to forcefield inaccuracies ( particularly hydrogen bonding and electrostatics ) [71] . Distance cutoffs are widely used because interaction energy decreases with distance , which is relatively inexpensive ( compared with energy cutoffs ) to calculate . The underlying assumption is that beyond a certain distance , the interaction energy is negligible . Our results show that this is , in fact , not always true: these seemingly negligible interaction energies can add up , leading to significant sequence differences and changes in local residue interactions that can alter not only the structure , but also the function of the predicted protein . Therefore by using distance cutoffs , protein designers have been neglecting significant pairwise interactions , which may compromise the accuracy of their predictions . Our paper is the first large scale study of the magnitude and consequences of distance cutoffs and their effects . On the positive side , we showed that by combining sparse residue interaction graphs with provable , ensemble-based algorithms , we provide a way to overcome this inaccuracy . Therefore , by using provable algorithms in the manner we described , protein designers can continue to reap the benefits of distance cutoffs without worrying about loss in accuracy . The gap-free list of conformations generated will include the sequence of both the sparse and full GMEC , allowing both sequences to be inspected and tested . We believe this is a notable improvement over traditional protocols , which require designers to commit beforehand to either the sparse or full interaction model . Our work simultaneously exposes potentially significant experimental inaccuracies in the input model and provides novel methodology to address these inaccuracies directly . | Computational structure-based protein design algorithms have successfully redesigned proteins to fold and bind target substrates in vitro , and even in vivo . Because the complexity of a computational design increases dramatically with the number of mutable residues , many design algorithms employ cutoffs ( distance or energy ) to neglect some pairwise residue interactions , thereby reducing the effective search space and computational cost . However , the energies neglected by such cutoffs can add up , which may have nontrivial effects on the designed sequence and its function . To study the effects of using cutoffs on protein design , we computed the optimal sequence both with and without cutoffs , and showed that neglecting long-range interactions can significantly change the computed conformation and sequence . Designs on proteins with experimentally measured thermostability showed the benefits of computing the optimal sequences ( and their conformations ) , both with and without cutoffs , efficiently and accurately . Therefore , we also showed that a provable , ensemble-based algorithm can efficiently compute the optimal conformation and sequence , both with and without applying cutoffs , by enumerating a small number of conformations , usually fewer than 1000 . This provides a novel way to combine cutoffs with provable , ensemble-based algorithms to reap the computational efficiency of cutoffs while avoiding their potential inaccuracies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"biotechnology",
"infographics",
"sequencing",
"techniques",
"chemical",
"bonding",
"protein",
"interactions",
"engineering",
"and",
"technology",
"applied",
"mathematics",
"synthetic",
"biology",
"synthetic",
"bioengineering",
"simulation",
"and",
"modeling",
"algorithms",
... | 2017 | A critical analysis of computational protein design with sparse residue interaction graphs |
Adaptation to different nutritional environments is essential for life cycle completion by all Trypanosoma brucei sub-species . In the tsetse fly vector , L-proline is among the most abundant amino acids and is mainly used by the fly for lactation and to fuel flight muscle . The procyclic ( insect ) stage of T . b . brucei uses L-proline as its main carbon source , relying on an efficient catabolic pathway to convert it to glutamate , and then to succinate , acetate and alanine as the main secreted end products . Here we investigated the essentiality of an undisrupted proline catabolic pathway in T . b . brucei by studying mitochondrial Δ1-pyrroline-5-carboxylate dehydrogenase ( TbP5CDH ) , which catalyzes the irreversible conversion of gamma-glutamate semialdehyde ( γGS ) into L-glutamate and NADH . In addition , we provided evidence for the absence of a functional proline biosynthetic pathway . TbP5CDH expression is developmentally regulated in the insect stages of the parasite , but absent in bloodstream forms grown in vitro . RNAi down-regulation of TbP5CDH severely affected the growth of procyclic trypanosomes in vitro in the absence of glucose , and altered the metabolic flux when proline was the sole carbon source . Furthermore , TbP5CDH knocked-down cells exhibited alterations in the mitochondrial inner membrane potential ( ΔΨm ) , respiratory control ratio and ATP production . Also , changes in the proline-glutamate oxidative capacity slightly affected the surface expression of the major surface glycoprotein EP-procyclin . In the tsetse , TbP5CDH knocked-down cells were impaired and thus unable to colonize the fly’s midgut , probably due to the lack of glucose between bloodmeals . Altogether , our data show that the regulated expression of the proline metabolism pathway in T . b . brucei allows this parasite to adapt to the nutritional environment of the tsetse midgut .
The study of the metabolic interactions between parasites and insect vectors is critical to understanding their biology and evolution , as well as to aid the design of control strategies that aim to prevent transmission of vector-borne pathogens . Parasites of the Trypanosoma brucei sub-species cause sleeping sickness and Nagana disease in sub-Saharan Africa , and are exclusively transmitted by tsetse ( Glossina spp . ) flies [1–3] . When T . b . brucei bloodstream forms ( BSF ) are ingested by a fly , the replicative ‘slender’ trypanosomes rapidly die within the insect midgut ( MG ) , whereas the pre-adapted ‘stumpy’ trypanosomes differentiate into the procyclic form ( PF ) within 24h [4] . Establishment of a trypanosome infection in the tsetse MG involves parasite colonization of the ectoperitrophic space ( a cavity between the peritrophic matrix and the gut epithelium ) and subsequent migration to the proventriculus ( PV ) [5] , where the parasite is confined and further differentiates [6] . After multiple morphological and biochemical changes ( reviewed in [7 , 8] ) , the parasites then migrate to the salivary glands ( SG ) , where they remain attached to the epithelial cells as epimastigotes ( [9] and reviewed in [7] ) . After colonizing the SG , epimastigotes differentiate into infectious metacyclic forms , which are then released into the fly’s saliva and transmitted to another vertebrate host during a subsequent feed [4] . Unlike most Dipterans , tsetse flies do not store carbohydrates for ATP production [10] . Furthermore , glucose does not seem to constitute a relevant source of energy , is rapidly metabolized ( ~1h ) after the bloodmeal is ingested , and is also found in low amounts in the fluids of these insects [11] . The use of minute amounts of glucose seems to be restricted to the production of other metabolites , such as non-essential amino acids in anabolism-requiring situations , e . g . pregnancy [12] . Thus , tsetse flies are adapted to efficiently metabolize amino acids and , more specifically , to catabolize proline to accomplish ATP biosynthesis [13 , 14] , a characteristic that is associated to obligatory blood feeding dipterans [15] . Additionally , proline is important in lactation , it is the metabolite that energetically supports the flight process and it is preferentially utilized by sarcomeres ( flight muscle cells ) , yielding alanine as the main product . In this context , proline is a critical metabolite for tsetse biology [16] . Amino acid metabolism requires a robust transamination network that allows the transfer of amino groups ( -NH2 ) to different acceptors , mainly ketoacids . In the specific case of glutamate , -NH2 is preferentially transferred to pyruvate , and yields alanine and oxoglutarate , which are the main intermediate products of proline catabolism . In tsetse flies , alanine is produced from proline by muscle cells and is further delivered into the hemolymph , which is then taken up into the fat body cells , for proline production [17] . This newly synthesized proline is , in turn , delivered to the hemolymph and taken up by flight muscle cells [13 , 18] . This cycle allows the continuous supply of proline to flight muscles by keeping high proline levels in the hemolymph , which fuels insect flight [19] . During the T . b . brucei life cycle , the parasite goes through a deep metabolic reprogramming; this process allows the parasite to optimize its nutritional requirements according to the available metabolic resources in each environment . This is the case when trypanosomes transit from glucose-rich environment ( in the bloodstream of the mammal ) to one rich in amino acids ( tsetse midgut ) , which requires a profound metabolic switch ( reviewed in [4 , 20] ) . Among the amino acids catabolized , L-proline plays a major role in the bioenergetics of trypanosomes [21–24] . In particular , the procyclic stage of T . b . brucei uses L-proline as a major carbon and energy source [23] , which is actively taken up [25] and catabolized inside the mitochondrion into succinate , alanine and acetate with production of intermediate metabolites , reduced cofactors and ATP [26 , 27] . Conversion of proline into glutamate is mediated by two enzymatic steps and one non-enzymatic step . First , proline is oxidized into Δ1-pyrroline-5-carboxylate ( P5C ) by a FAD-dependent proline dehydrogenase ( TbProDH ) [EC 1 . 5 . 99 . 8] [23] . Second , the cyclic P5C ring is spontaneously opened through a non-enzymatic reaction to produce glutamate-γ-semialdehyde ( γGS ) . Third , the carbonyl moiety of γGS is further oxidized to glutamic acid by a P5C dehydrogenase ( TbP5CDH ) [EC 1 . 5 . 1 . 12] with a concomitant reduction of NAD ( P ) + into NAD ( P ) H [28] . Unlike Trypanosoma cruzi , there are no genomic or biochemical data supporting the existence of a proline biosynthetic pathway in T . b . brucei [29] , which suggests it is auxotrophic for this amino acid . Moreover , in PFs it was reported that proline degradation is downregulated in the presence of glucose [24] , and the importance of Ca2+ regulation of TbProDH activity in the energy metabolism of trypanosome insect stages was recently suggested [30] . Collectively , both proline oxidation to glutamate and further oxidation through a part of the tricarboxylic acid cycle ( TCA ) are able to produce reduced equivalents , as well as fuel oxidative phosphorylation , and thus contribute to fulfilling the parasite’s energy requirements [31] . The relevance of proline metabolism for both T . b . brucei and the tsetse led us to address the long-standing question on the role of this amino acid in the parasite´s ability to infect flies . While the importance of TbProDH to the parasite’s biology has previously been studied , little is known on the specific role of TbP5CDH , besides its participation in the complete oxidation of proline . In this work we addressed this issue by studying the role of TbP5CDH in the bioenergetics of T . b . brucei as well as its importance during a tsetse infection . Our data show that in the absence of glucose , T . b . brucei PFs rely on the proline provided by the fly and on a fully functional proline catabolic pathway to successfully survive within the tsetse midgut .
In order to understand the role ( s ) of TbP5CDH in T . b . brucei biology , we first characterized its expression during the in vitro growth of both procyclic cultured forms ( PCFs ) and BSFs . Parasites were cultured in complete SDM79 and HMI9 media , respectively , and their growth followed up for 72h ( although the analyses were made at 24 and 48h depending on the different parasite doubling times; Fig 1A ) . To analyze the expression profile of TbP5CDH and its influence on proline metabolism , TbP5CDH mRNA and protein levels were determined by qPCR and western blot , respectively . While both the mRNA and protein levels remained almost constant over time in PCFs , no TbP5CDH protein was detected in BSFs ( Fig 1B and 1C ) . This indicates that , at least in vitro , expression of this enzyme is tightly regulated between different trypanosome stages . This observation is consistent with previous data showing that proline catabolism seems to be repressed in T . b . brucei BSFs [32] . We then investigated whether TbP5CDH expression is developmentally regulated during tsetse infection by isolating parasites from different infected organs; i . e . MG , PV and SG . TbP5CDH mRNA was detected in parasites collected from the PV and MG but not from SG-derived forms ( Fig 1D ) . No significant changes in the expression levels were observed between PV and MG forms , but there was a strong reduction ( 60-fold change , p<0 . 05 ) in mRNA levels in SG forms . Notably , it was not possible to examine TbP5CDH protein expression by western blotting due to strong cross-reactivity with the Glossina P5CDH protein . Collectively , these results suggest that both PV and MG trypanosome forms express the proline-oxidizing pathway , which would be necessary to fulfill the energy requirements for cell proliferation , although the enzyme is downregulated as the infection progresses towards the SGs . To determine the subcellular location of TbP5CDH , T . brucei PCFs were submitted to digitonin fractionation and the enzyme was detected by western blotting . As shown in Fig 2A , TbP5CDH was released together with the mitochondrial markers TbASCT and TbProDH , while the cytosolic marker enolase was released at much lower digitonin concentrations ( 20 μg compared to 350 μg of digitonin mg-1 of protein ) [33] . Under these assay conditions , we also detected TbProDH but at low amounts , which is consistent with its possible association with the mitochondrial inner membrane ( Fig 2A ) [23] . Furthermore , immunofluorescence of fixed PCFs showed co-localization of TbProDH and TbP5CDH ( Fig 2B ) , thus confirming the results obtained by digitonin fractionation . To determine the importance of TbP5CDH in the bioenergetics of trypanosomes , we downregulated its expression by RNAi using a tetracycline-inducible system [34] . After 72h of tetracycline-induction ( RNAiTbP5CDH tet+ ) , no TbP5CDH was detected by western blotting ( Fig 3A ) . However , when we assayed its enzymatic activity , we observed ~16% remaining activity compared to non-induced cells ( RNAiTbP5CDH tet- ) ( Fig 3B ) . No changes in the levels of TbP5CDH were observed in wt cells supplemented or not with tetracycline ( wt tet-/+ ) , which showed that addition of this antibiotic had no direct effect on TbP5CDH expression ( Fig 3A and 3B ) . As previously shown , wt PCFs are able to replicate in standard SDM79 supplemented ( or not ) with glucose ( SDM79 and SDM79 glc- , respectively ) [23] . In standard SDM79 , glucose is the preferred carbon source for PCFs , whereas in the absence of glucose , the parasites mainly use proline as a carbon source and for ATP production . In the case of TbP5CDH , the enzyme was essential when proline was the major carbon source . However , the phenotype was not lethal most likely because of the remaining enzymatic activity in the RNAiTbP5CDH cell line ( Fig 3C and 3D ) . These findings prompted us to evaluate the main mitochondrial functions ( i . e . ΔΨm , O2 consumption rates and ATP levels ) in RNAiTbP5CDH cells energized with proline . In digitonin-permeabilized cells , downregulation of TbP5CDH caused a diminished capacity to retain the mitochondrial dye safranin and to respond to the addition of ADP compared to non-induced cells . This profile reflects a partial depolarization of mitochondria from RNAiTbP5CDH tet+ cells when proline is the electron source for the oxidative phosphorylation ( OxPHOS ) process ( Fig 4A ) . No changes were observed for the same parameters when succinate was used as a mitochondrial substrate ( S1A and S1B Fig ) . In addition , ADP failed to induce the proton flux into the matrix space through the Fo/F1 ATP synthase complex and did not decrease ΔΨm to the same levels shown by non-induced cells . Moreover , addition of oligomycin , an inhibitor of ATP synthase , also resulted in a slight increase in ΔΨm , and reestablished the resting levels , which were significantly lower than control . This is likely due to the diminished electron flux from proline degradation to the respiratory complexes in RNAiTbP5CDH tet+ parasites , which seem to be insufficient to sustain physiological levels of OxPHOS ( Fig 4A ) . Interestingly , addition of Ca2+ to these mitochondrial preparations did not affect the ΔΨm of wt and RNAiTbP5CDH cells , which suggests that variations in the electrochemical potential using proline are due exclusively to mitochondrial electron transfer chain ( mt-ETC ) capacity rather than mitochondrial Ca2+ influx ( Fig 4B ) . Observations made at the ΔΨm level are consistent with the diminished ability of the mutant cell line to consume O2 when proline and ADP were present at high concentrations ( respiratory state 3 ) , and the high respiration rates are limited by respiratory chain activity [35] . Moreover , the maximal oxygen reduction capacity was dramatically affected in the RNAiTbP5CDH tet+ cells when FCCP ( which collapses the mitochondrial membrane potential ) was added to the mitochondrial preparations ( Fig 4C , Table 1 ) , and the respiratory control ratio significantly decreased to 1 . 44 ± 0 . 02 ( Table 1 ) . When succinate was used as the respiratory substrate in control and RNAiTbP5CDH tet-/+ parasites , no differences in ΔΨm or O2 consumption rates were observed ( S1A–S1C Fig ) . The ATP levels in parasites cultivated in either SDM79 and SDM79 glc- media were also determined . As expected , the absence of TbP5CDH did not affect ATP levels when glucose was present ( Fig 4D ) . Conversely , when ATP synthesis relied on proline oxidation ( cells grown in SDM79-glc- ) , the capacity of RNAiTbP5CDH tet+ cells to produce ATP was diminished ( Fig 4D ) . Given that the T . b . brucei genome does not appear to contain genes that encode putative P5C/γGS metabolizing enzymes ( with the exception of TbP5CDH ) , it is assumed that the proline-glutamate pathway has no branches . On this basis , it is expected that TbP5CDH-knocked down cells would produce elevated quantities of intracellular P5C , which has been described as a toxic metabolite in several cell types [36 , 37] . Thus , the deleterious effect observed in TbP5CDH knockdown cells could be due not only to a diminished efficiency in ATP synthesis but also due to P5C accumulation . To evaluate this , RNAiTbP5CDH tet-/+ parasites were incubated in vitro under different metabolic conditions ( i . e . PBS supplemented with L-proline , glucose , proline plus glucose , or P5C/γGS ) , and their viability was assessed over a 3h period . Controls consisted of RNAiTbP5CDH tet-/+ cells incubated with either SDM79 ( 100% viability ) or PBS ( which yielded a 3% viability compared to cells incubated in SDM79 alone ) . PCFs incubated in the presence of either proline or proline plus glucose showed a viability of 65% versus SDM79-treated cells , and no significant differences were found for these treatments between induced or not-induced cells ( Fig 5A ) . The addition of P5C/γGS to the RNAiTbP5CDH tet- cells resulted in almost the same viability as proline treatment ( 50% ) . Notably , incubation of RNAiTbP5CDH tet+ cells with P5C/γGS reduced their viability by more than 90% ( Fig 5A ) . In addition , non-induced and RNAi-induced procyclics were treated with proline or P5C for 1 or 3h , and P5C toxicity was indicated based on loss of plasma membrane integrity . Only in RNAiTbP5CDH tet+ cells P5C but not proline treatment resulted in 15% and 57% of Propidium Iodide ( PI ) -positive cells after 1h and 3h challenge , respectively ( Fig 5B ) . These data were compatible with observed mitochondrial and morphological alterations ( Fig 5C ) . Interestingly , in spite of its deleterious effect , P5C was able to support MitoTracker accumulation ( a process that is dependent on the mitochondrial inner membrane potential ) and to maintain higher ATP levels in wt or RNAiTbP5CDH tet- , when compared to RNAiTbP5CDH tet+ cells . These results , along with previous published evidence [36 , 37] , suggest that i ) P5C is able to reach the mitochondrial matrix; ii ) the only metabolic fate for P5C/γGS is to be oxidized to glutamate via TbP5CDH; and iii ) the intracellular accumulation of P5C/γGS has a detrimental effect on PCFs viability . The increased susceptibility of RNAiTbP5CDH tet+ when exogenous P5C/γGS is added is indicative of the inability of T . brucei PCFs to reduce it to proline . Thus , we then evaluated whether proline biosynthesis from glutamate or from P5C could happen in T . b . brucei . To address this question , parasites were grown in defined media supplemented or not with proline . When PCFs were grown in either complete SDM79 or SDM79 glc- media no differences were found in the cells doubling time ( 19 . 3 ± 1 . 1 h and 20 . 3 ± 1 . 4 h , respectively ) . After proline deprivation of the media ( SDM79 pro- glc- ) , PCFs showed a delay in doubling time ( 48 . 4 ± 6h ) ( Fig 6A ) . This diminished capability for proliferation under proline-depleted media strongly suggests that T . brucei is auxotrophic for this amino acid . Furthermore , when the T . b . brucei genome was interrogated for putative genes that encode P5C-synthase ( P5CS; converts glutamate into P5C/γGS0 ) and P5C-reductase ( reduces P5C/γGS into proline ) , using T . cruzi sequences as queries , only a protein sequence with 65% similarity with T . cruzi P5CR was found ( TritrypDB accession number: Tb927 . 7 . 2440 ) . No significant hits were found for P5CS ( TritrypDB accession number: TCSYLVIO_005298 ) . The presence of a putative P5CR ortholog in T . b . brucei prompted us to evaluate its enzymatic activity by measuring the reduction of P5C to proline in PCF cell-free extracts . The enzymatic test for P5CR revealed activities of 8 . 6 ± 0 . 5 versus 60 ± 9 nmol NADPH/min/mg of protein in T . b . brucei PCF and T . cruzi epimastigote cell-free extracts , respectively ( Fig 6B ) . Furthermore , P5CS protein was not detected in T . b . brucei lysate using antibodies raised against its T . cruzi ortholog ( Fig 6C ) . To evaluate the possible occurrence of a proline biosynthetic pathway in T . b . brucei PCFs , the levels of this amino acid were measured in proline–deprived parasites ( after 1h incubation in PBS ) . The cells were then incubated with different substrates that would restore proline levels , i . e . via uptake ( proline ) , reductive biosynthesis ( P5C/γGS , glutamate , glutamine ) , or through the connection between the urea cycle and proline-glutamate pathway ( arginine or alanine ) as occur in other organisms . Collectively , the demonstration that the only metabolite capable of restoring the normal intracellular levels of proline in PCFs after starvation was proline ( Fig 6D ) and the lack of genetic and biochemical evidence for a proline biosynthetic pathway in T . b . brucei further corroborate its auxotrophic nature for this amino acid . As no proline biosynthetic pathway or ornithine transaminase activity could be evidenced in T . b . brucei , TbP5CDH should be the only enzyme capable of metabolizing intra-mitochondrial P5C in these cells . In order to unambiguously evaluate the occurrence of this enzymatic activity we kinetically characterized TbP5CDH from PCF lysates . Our data revealed that these cells were able to reduce NAD+ upon the addition of P5C in a concentration dependent manner with apparent KM values of 92 . 7 ± 14 μM and 0 . 38 ± 0 . 04 mM for its substrate ( P5C/γGS ) and cofactor ( NAD+ ) , respectively , and Vmax values of 0 . 15 ± 0 . 01 and 0 . 19 ± 0 . 01 μmol/min/mg of protein for P5C and NAD+ , respectively ( S2 Fig ) . To further determine the metabolic perturbations caused by downregulation of TbP5CDH , end products excreted from the catabolism of proline and [U-13C]-glucose were analyzed by proton-NMR spectroscopy . We used a previously-developed metabolite profiling assay based on the ability of proton NMR spectroscopy to distinguish 13C-enriched from 12C molecules [38] . Cells were incubated in PBS with equal amounts ( 4 mM ) of non-enriched proline and of [U-13C]-glucose in order to perform a quantitative analysis of proline-derived and glucose-derived acetate production by proton NMR . For instance , [13C]-acetate derived from metabolism of [U-13C]-glucose ( annotated 13C in Fig 7 ) is represented by two doublets , with chemical shifts at around 2 . 0 ppm and 1 . 75 ppm , respectively , while the central resonance ( 1 . 88 ppm ) corresponds to proline-derived [12C]-acetate . As expected , the amounts of [U-13C]-glucose-derived end products ( 13C-enriched succinate , acetate and alanine ) are similar in the RNAiTbP5CDH tet+ mutant and wt cells ( 2081 versus 2057 nmol excreted/h/mg of proteins ) , whereas the amounts of excreted end products from proline degradation ( non-enriched succinate , acetate and alanine ) were 2 . 2-reduced in the RNAiTbP5CDH tet+ cell line ( Fig 7 and Table 2 ) . The remaining production of end products excreted from proline metabolism ( 44% compared to wt cells ) was probably due to a 16% residual TbP5CDH activity in the tetracycline-induced RNAiTbP5CDH mutant . Notably , reduction of succinate and acetate production from proline is compensated by an increased production of these molecules from glycolysis ( Fig 7 , Table 2 ) . Such flux redistribution towards glucose-derived acetate production was also previously observed in the threonine dehydrogenase procyclic mutant incubated with threonine and [U-13C]-glucose [38] . Altogether these metabolic data demonstrate that TbP5CDH is involved in the proline degradation pathway of procyclic trypanosomes . After observing differences in the expression levels of TbP5CDH during parasite development in the fly ( Fig 1D ) , we then analyzed its essentiality for parasite survival in the tsetse midgut . Flies were infected with a bloodmeal supplemented with either wt or RNAiTbP5CDH PCFs , which were either previously induced or not with tet . At 9 days post-infection ( dpi ) , the flies were dissected and midgut infections were determined . Flies fed with either wt or RNAiTbP5CDH tet- cells had infection rates of 82% ( Fig 8A , S3 Fig ) . Furthermore , there were no differences in the number of parasites in the midguts of wt tet- , wt tet+ or RNAiTbP5CDH tet- infected flies ( Fig 8A , S3 Fig ) . However , after downregulation of TbP5CDH , the midgut infection rates dropped significantly to 58% ( p<0 . 01 ) and , importantly , only a few parasites were visible ( Fig 8A ) . Furthermore , under normal TbP5CDH expression ( i . e . wt tet- , wt tet+ or RNAiTbP5CDH tet- ) , the infected midguts had a much higher number of parasites ( >1000 cells per field ) compared to flies infected with RNAiTbP5CDH tet+ cells ( ≤10 cells per field ) ( Fig 8A , S3 Fig ) . Parasites were probably present in the latter group due to residual expression of TbP5CDH and/or to the transient utilization of glucose present in subsequent bloodmeals . Altogether , these data demonstrate that TbP5CDH activity , a key enzyme in the parasite proline metabolism pathway , is crucial for trypanosome survival within the tsetse fly midgut . EP- and GPEET-procyclins are the most abundant GPI-anchored surface glycoproteins on the surface of T . b . brucei PCFs . The C-terminus of all EP-isoforms contains abundant ( up to 30 ) repeats of glutamate ( E ) and proline ( P ) dipeptides [55] . Likewise , GPEET-procyclin is also rich in E and P because of its 5–6 GPEET C-terminal repeats . We investigated whether alterations in the proline-glutamate oxidative flux interfere with the expression of all procyclin isoforms . Western blotting analysis showed a slightly decreased in EP-procyclin expression in RNAiTbP5CDH tet+ cells compared to wt tet- , wt tet+ ( Fig 8B ) . Interestingly , perturbations in the number of EP-positive cells were found after four days of RNAi-induction for TbP5CDH . Two different cell populations were observed , which were named as EP-pop1 and EP-pop2 ( Fig 8B , right panel ) . The EP-pop1 displayed similar values of fluorescence intensity versus controls ( wt tet-/+ or RNAiTbP5CDH tet- ( Fig 8C ) , whereas the EP-pop2 population showed a 10-fold reduction in the mean of fluorescence . However , when the repertoire of procyclins was analyzed by MALDI-TOF ( S4A–S4D Fig ) mainly EP1-2 and EP3 isoforms ( containing 25 and 22 EP repeats , respectively [55] ) were detected in either induced or non-induced cells . This suggests that although the overall expression of EP-procyclins appears to be slightly compromised when the proline metabolism pathway is altered ( Fig 8C ) , these cells do not seem to compensate the slight EP deficit by re-expressing GPEET-procyclin .
Some Dipterans ( including the genus Glossina ) are well adapted to use amino acids for energy production . In fact , due to the scarce carbohydrate reserves in tsetse , glycolytic activity is negligible within this insect [11 , 41] . Three characteristics make proline a readily mobilizable energy source in tsetse: i ) its highly reduced state , which is related to its high yield in terms of metabolic energy production ( i . e . 5-fold more efficient than carbohydrates ) ; ii ) its high solubility ( allowing its transport in high concentrations , thus permitting an efficient distribution in the entire fly body ) ; and iii ) its low nitrogen content limiting the amount of energy required for nitrogen detoxification ( reviewed in [42] ) . Einar Bursell concluded that "proline constitutes the only effective substrate for flight metabolism" for species in which both sexes are obligatory blood-feeders ( i . e . Glossina spp . ) [15] . In fact , proline represents ~4% of the total amino acid content in the tsetse hemolymph and is efficiently burnt during the flight process [43 , 44] . It is first oxidized to glutamate , and further converted into oxoglutarate by either an alanine transaminase or a glutamate dehydrogenase ( Fig 9 , left panel ) . Consequently , flight time in tsetse ( which is limited to about three minutes ) is likely to be determined by the amount of proline available in the hemolymph at the outset [43] . On the other hand , the synthesis of proline from alanine in tsetse takes place in the fat body . It is a complex process that comprises an alanine-glyoxylate transaminase , a pyruvate dehydrogenase and part of the TCA cycle . Part of the oxoglutarate produced is converted into glutamate in the same transamination reaction in which a new alanine molecule is converted into pyruvate to feed again the TCA cycle ( Fig 9 , right panel ) [17 , 18 , 45] . Thus , there is a strong interdependence between proline/alanine metabolism between the fat body and flight muscles of tsetse , which are metabolically connected through the hemolymph . Notably , this metabolic system does not work at the steady state: the release of CO2 by the flight muscles creates a deficit of carbon . This deficit is possibly compensated by using Acetyl-CoA from the β-oxidation of lipids in the fat body for proline de novo biosynthesis [17] . During a trypanosome infection , parasite colonization of different tsetse organs may alter the fly’s proline-alanine cycle . Such an alteration would not only have an impact on the activity of flight muscles , but also affects tsetse reproduction [45] . The crosstalk in the utilization of proline in trypanosome-infected flies becomes even more complex with the dependency on Wigglesworthia glossinidia ( the obligate tsetse bacterial symbiont ) for the production of vitamin B6 , which is essential for activity of alanine-glyoxylate aminotransferase ( involved in proline regeneration in fat body ) ( reviewed in [16] ) . It has previously been shown that TbProDH is essential when parasites are grown in the presence of proline as the main energy source ( SDM79 glc- ) [23] . In the present work , a different phenotype was observed: TbP5CDH knockdown cells did not die in vitro . However , the essentiality of TbP5CDH for survival in the fly was evident by the low midgut infection phenotype in knockdown cells . This discrepancy between the in vitro phenotypes could be due to the residual activity ( ~16% ) of TbP5CDH in the tetracycline-induced RNAiTbP5CDH mutant , which would be able to maintain a low but significant metabolic flux thus allowing proline oxidation in the TCA cycle/ETC . Alternatively , the activity of TbProDH , a FAD dependent enzyme , might also be able to directly transfer electrons to the ubiquinone pool at the ETC ( similarly to succinate dehydrogenase ) , as already proposed for T . cruzi [21] . Both possibilities , individually or combined , could explain parasite survival by partially fulfilling the energy requirements of these cells . In T . b . brucei , proline conversion into P5C/γGS produces FADH2 , which can transfer 2e- to the UQ pool with further reduction of cytochromes [21] . γGS is then converted into glutamate to produce NADH . In addition , four additional reactions downstream to the proline-glutamate conversion produce NADH in T . b . brucei ( reviewed in [29] ) . T . b . brucei expresses a mitochondrial NADH: ubiquinone oxidoreductase ( which is rotenone insensitive ) , uses FMN as cofactor , transfers one e- to the ubiquinone ( UQ ) pool and can reduce O2 to O2-• anion [46] . This enzyme is likely to be involved in the reoxidation of NADH , thus reducing UQ and driving proton pumping at level of C-III and C-IV in the mt-ETC [47] . Then , both proline oxidation steps generate reducing equivalents that feed the OxPHOS , thus driving the ATP synthesis through the FoF1/ATP synthase . The intramitochondrial glutamate produced from proline can either be: i ) deaminated into oxoglutarate or ii ) transaminated to pyruvate forming alanine . Oxoglutarate can be converted into succinyl-CoA and then into succinate , constituting two points of ATP generation , by substrate level phosphorylation ( at succinyl-CoA synthetase level ) and OxPHOS ( via succinate dehydrogenase complex ) , respectively [48] . Succinate can also be excreted as an end product . In the absence of glucose , alanine is also excreted by T . b . brucei PCFs as the end product of proline degradation . It may be possible that tsetse also utilizes trypanosome-excreted alanine for further conversion into proline , especially in highly infected tissues . An intriguing phenotype evidenced in our RNAiTbP5CDH cell line , was the cell toxicity displayed when exogenous P5C/γGS was added to tet-induced parasites . In most eukaryotes , P5C/γGS can be synthesized by proline oxidation , glutamate reduction or by loss of the -NH2 group at the δ-carbon of ornithine through an ornithine transamination reaction . In turn , P5C/γGS can be decreased by its oxidation to glutamate by P5CDH , its reduction to proline , or by its amination to form ornithine [49] . Thus , the amount of free P5C/γGS mainly results from the balance between all these enzymatic activities . T . b . brucei lacks a functional urea cycle , which eliminates any connection between this pathway and P5C/γGS [50] . Furthermore , our results show that neither relevant enzymatic activities related to proline biosynthesis nor genes encoding the putative enzymes for these pathways are present in the T . b . brucei genome . In addition , there was a cytotoxic effect for externally added P5C/γGS to RNAiTbP5CDH tet+ , which supports the oxidation to glutamate as the only fate for this metabolite in PCFs . It should be noted that these results differ to those obtained when TbP5CDH-knockdown cells were treated with proline ( resulting in the intracellular accumulation of P5C/γGS due to the TbProDH activity ) , which indicated the cells remained viable ( although non-replicative ) . This was also consistent with the viability shown by cells with increased levels of P5C/γGS by overexpressing a mitochondrial carrier ( TbMCP14 ) [51] . Altogether our results showed that , unlike T . cruzi –whose energetic metabolism also relies on proline consumption [21 , 28]- , T . b . brucei PCF is auxotrophic for proline , this being an essential metabolite and a main carbon source during this stage . As a consequence , P5C/γGS levels depend exclusively on the balance between its formation from proline oxidation and its depletion by oxidation to glutamate . In addition , we confirmed that proline deprivation dramatically affects cell proliferation as previously suggested [52] . Altogether , our data provide evidence that T . b . brucei has a strict requirement of a complete proline to glutamate oxidation pathway to successfully colonize tsetse midguts . The relevance of EP-procyclin expression for the successful development of T . b . brucei within the fly has been widely reported [53 , 54] . Expression of procyclins ( GPEET and EP ) varies according to the parasite stages in the tsetse [40] . There seems to be a correlation between the expression of such molecules in midgut forms and elevated mitochondrial activities [55] . More specifically , GPEET expression ( normally in early stages ) can be reactivated in late forms when mitochondrial activities such as the ASCT cycle or alternative oxidase are inhibited [55] . It was also stated that glycolytic activity , disrupted by RNAi-silencing of the trypanosome hexokinase gene , produces a switch in the surface expression from EP- to GPEET-procyclin [56] . We observed herein that alterations in the proline-glutamate pathway slightly affects the levels of EP-procyclin expressed at the surface . However , this alteration did not induce a change in the type of procyclins these cells expressed and no evidence of GPEET re-expression was observed . Given the specificity of the anti-EP mAb 247 for the glu-pro dipeptides [57] , it is likely that the alterations in EP-procyclin expression after down-regulation of TbP5CDH , could be simply due to a reduction in the overall levels of intracellular glutamate available for making the C-terminus glu-pro repeats . Furthermore , it is unlikely that such a small reduction in the surface expression of EP-procyclins accounts for the inability of RNAiTbP5CDH tet+ cells to colonize the tsetse midgut , although EP-procyclin-null trypanosomes are less efficient in establishing midgut infections [53 , 54] . Thus , these results further confirm that the fly phenotype observed in knocked down cells appears to be mainly a direct consequence of an interrupted proline metabolism pathway in these parasites .
Animal experiments in this work were performed in accordance with the local ethical approval requirements of the Liverpool School of Tropical Medicine and the UK Home Office Animal ( Scientific Procedures ) Act ( 1986 ) under license number 40/2958 . BSF of T . b . brucei TSW196 strain [58] , which is a fully fly-transmissible , was used for gene expression studies and proline determination in infected flies . BSF T . b . brucei from 2T1 strain ( kindly provided by David Horn , University of Dundee , UK ) were cultured in HMI-9 medium supplemented with 15% ( v/v ) FCS ( Gibco ) at 37°C and 5% CO2 [59] . The initial cell density was 5x104 cells/ml , which was sub-cultured each 48h . Parasite densities were determined by cell counting using a hemocytometer . PCFs of T . b . brucei Lister 427 ( 29–13 clone ) ( T7-RNAp+ NEO+ TET+ HYG+ ) , which expresses the T7 RNA polymerase under the control of the tetracycline ( tet ) promoter , was cultured in vitro in SDM79 media ( Gibco ) supplemented with GlutaMAX ( Gibco ) , 7 . 5 μg/ml hemin and 10% ( v/v ) heat-inactivated fetal calf serum ( FCS ) [60 , 61] at 26°C . For RNAi experiments , parasites were grown in SDM79 media supplemented with 25 μg/ml G418 ( G ) and 12 . 5 μg/ml hygromycin ( H ) as indicated [62] . To grow PCFs in defined medium , we used SDM79 base media without sodium bicarbonate , glucose , glutamine , glutamate , proline , pyruvate , threonine and acetate ( SDM79-CGGGPPTA ) ( PAA laboratories , Pasching , Austria ) and then supplemented , except for glucose ( SDM79 glc- ) or proline ( SDM79 glc- pro- ) as indicated [23] . In both cases , the preparation was supplemented with 10% ( v/v ) tet-free FCS ( Clontech Laboratories ) and an excess of 50 mM N-acetyl-D-glucosamine ( GlcNAc ) to inhibit uptake of glucose presented in serum ( about 1 . 5 mM ) [63] . The initial cell density was 106 cells/ml and sub-culturing was done every 72h [60] . Glossina morsitans morsitans flies were maintained in a laboratory colony at the Liverpool School of Tropical Medicine ( LSTM ) at 26°C and 65–70% relative humidity . Teneral ( 12-24h post-emergence ) flies were fed on sterile defibrinated horse blood ( TCS Biosciences Ltd . , Buckingham , UK ) . Fly-derived trypanosomes were isolated from the MG , PV and SG of infected flies , as described . Parasites were resuspended in SDM79 medium and midgut debris were removed by filtration through cytometer-tubes filter ( Becton Dickinson ) . Cells were harvested by centrifugation ( 2 , 000 g for 10 min at 4°C ) , washed twice with cold PBS ( 137 mM NaCl , 2 . 7 mM KCl , 8 mM Na2HPO4 , 1 . 5 mM KH2PO4 adjusted to pH 7 . 3 ) , counted and stored at -80°C until RNA or protein analysis . Total RNA extractions from fly-derived parasites ( ~5x106 cells ) were performed with TriZol reagent ( Sigma ) following standard procedures [64] . Then , 300 ng total-RNA were used for cDNA synthesis with oligo ( dT ) 20 and SuperScript III Reverse Transcriptase ( RT ) ( Invitrogen ) . Resultant cDNA samples were diluted ( 1:4 ) in nuclease-free water ( NFW ) for use in quantitative RT-PCR ( qPCR ) . Based on DNA-sequences for TbGAPDH and TbP5CDH ( TritrypDB accession numbers: Tb927 . 6 . 4300 and Tb427 . 10 . 3210 , respectively ) , specific primers were designed ( S1 File for oligonucleotides sequences ) . qPCR reactions were performed in 96-wells ( Stratagene , Agilent Technologies , La Jolla , TX , USA ) using 3 . 2 pg of each primer , 5 μl fast SYBR green master mix ( Applied Biosystems , Life Technologies , CA ) and 5 μl of cDNA samples to a final volume of 20 μl per well . Reactions were run in a Mx3000P qPCR-system ( Stratagene ) followed by a dissociation curve . Samples from naïve tissues were also used to verify primer specificity . The pZJM vector , which contains a cloning site between two opposing T7 promoters , was used to silence TbP5CDH expression ( Tb427 . 10 . 3210 ) [34] . A 5ʹ DNA fragment ( 480 bp ) corresponding to TbP5CDH was amplified by conventional PCR using specific primers ( see S1 File for oligonucleotides sequences ) , cloned into the pZJM vector ( pZJM/RNAiTbP5CDH ) and the resulting construct was confirmed by sequencing . Plasmid preparation was done using the QIAGEN plasmid Maxi Kit according to the manufacturer’s instructions ( QIAGEN ) . For transfections , 10 μg pZJM/RNAiTbP5CDH was linearized by digestion with the restriction endonuclease NotI ( Thermo Scientific ) , precipitated by standard procedures and dissolved in NFW . PCF trypanosomes ( 2x107 cells maintained in mid-log phase in SDM79 H/G medium ) were transformed using a Nucleofector transfection system II/2b , following the manufacturer’s instructions ( Lonza ) . Parasites were seeded into 24-well plates ( <10 cells/well ) and cloned by limiting dilution in SDM79 H/G supplemented with 2 . 5 μg/ml phleomycin as a selection marker . The obtained parasite lineages were referred to as wt ( parental Lister 427 29–13 strain ) or RNAiTbP5CDH , as the RNAi-TbP5CDH cell line , in the presence or absence of tetracycline ( tet-/+ ) . RNAi was induced by adding 0 . 5 μg/ml tetracycline disodium salt ( freshly dissolved in PBS ) to the selective media ( at 26°C ) . Teneral flies were infected with bloodmeal preparations that contained either wt or RNAiTbP5CDH parasites . Briefly , non-induced ( tet- ) or tetracycline-added ( tet+ ) parasites were added to sterile horse blood at a density of 5x105 parasites/ml . RNAi induction was maintained by adding 25 μg/ml tetracycline to the bloodmeal , and 24h after receiving an infectious blood meal , the flies were sorted and only fed flies were used . After nine days , flies were dissected and the number and intensity of infected midguts was determined by microscopy . A score was attributed to each infection as previously described [65] . Enzymatic determinations for both P5C reduction to proline or P5C oxidation to glutamate were performed . The substrate of TbP5CDH , a racemic mixture of DL-Δ1-pyrroline-5-carboxylate ( DL-P5C ) and its ring-open form gamma-glutamate semialdehyde ( γGS ) , was synthesized from peroxidation with NaIO4 ( Sigma ) , and eluted in acidic medium ( 1 M HCl ) as previously described [66] . The steady-state activity for TbP5CDH was measured in cell-free homogenates from PCFs , as previously described for T . cruzi [28] . The TbP5CDH reaction mixture contained: 0 . 3 mM P5C/γGS ( freshly prepared ) , 1 mM nicotinamide adenine nucleotide disodium salt ( NAD+ ) and 90 mM potassium phosphate buffer pH 7 . 2 , made up to a final volume of 3 ml with distilled water . The reaction was started after adding 200 μg cell-free homogenates from PCFs and the linear rate was determined by following the increase in absorbance ( λ340nm ) over 5 mins at 28°C with constant stirring . A blank without substrate ( P5C/γGS ) was used as a control . Readings of samples and controls were made in parallel in a double-beam Thermo Evolution 300 spectrophotometer ( Thermo Scientific ) . Kinetic parameters for P5C and the cofactor of TbP5CDH were also determined in PCF homogenates . Substrate dependence was assayed by varying the P5C/γGS concentrations over the range of 20–600 μM ( freshly prepared ) and 1 mM of NAD+ as saturating concentration . Cofactor dependence was assayed by varying the NAD+ concentrations over the range of 0 . 01–2 . 5 mM and 600 μM P5C/γGS as saturating concentration . The P5C-reductase reaction mixture contained: 500 μM P5C/γGS ( freshly prepared ) , 50 μM NADPH and 100 mM Tris-HCl pH 7 . 0 , and was made up to a final volume of 1 ml with distilled water . The reaction was started by adding different concentrations of PCF homogenates . The linear rate was determined by following the decrease in absorbance ( λ340nm ) over 3 min at 28°C with constant stirring . P5C-reductase enzymatic activity determinations from T . cruzi homogenates were used as controls under the same conditions . Parasites ( Lister 427 29–13 strain ) were incubated in PBS ( for 1h ) to diminish the intracellular pool of free proline . Parasites were then incubated for 40 min in the presence of different carbon sources and cofactors ( S2 File for detailed mix composition ) to determine which combination was able to restore the intracellular proline levels . Additional treatments consisted of parasite incubation with PBS supplemented with 5 mM L-proline ( positive control ) or non-supplemented PBS ( negative control ) . Parasites were then washed with cold PBS and centrifuged ( 3 , 000 g for 5 min at 4°C ) . Pellets were resuspended in 100 μl lysis buffer ( 100 mM Tris-HCl pH 8 . 1 , 0 . 25 M sorbitol , 1 mM EDTA , 1% ( v/v ) Triton X-100 , 1 mM phenylmethanesulfonylfluoride ( PMSF ) , 4 μg/ml aprotinin , 10 μg/ml tosyl-L-lysyl-chloromethane hydrochloride ( TLCK ) and 10 μM E-64 ) , and submitted to two cycles of snap freezing in liquid nitrogen thawing . Crude extracts were clarified by centrifugation ( 15 , 000 g for 15 min at 4°C ) and 100 μl supernatant was mixed ( in a separate reaction ) with 1 volume 20% ( w/v ) trichloroacetic acid for deproteinization . Samples were precipitated by centrifugation ( 20 , 000 g for 30 min at 4°C ) and 200 μl of the resultant supernatants were used for the Bates assay , as described elsewhere [67] . Non-induced and RNAi-induced ( tet-/+ ) PCFs from wt and RNAiTbP5CDH cell lines were grown for three days in SDM79 media at 26°C . Then , parasites were harvested by centrifugation and resuspended in either PBS or PBS supplemented with 5 mM L-proline , 1 . 5 mM P5C/γGS and 5 mM D-glucose , or with 5 mM L-proline + 5 mM D-glucose , and further incubated for 4h at 26°C . Cell viability was evaluated after incubation with 3- ( 4 , 5- dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ( MTT ) [68] . Results were obtained from three biological replicates ( n = 3 ) . Comparisons between non-induced and RNAi-induced cells were calculated using the one-way ANOVA test in GraphPad Prism v5 . 0a for Mac OS X ( GraphPad Software , USA ) . PCFs ( Lister 427 strain , 29–13 clone ) cultivated at the late logarithmic phase of growth ( 3x109 total cells ) in SDM79 medium were harvested by centrifugation [1 , 000 g for 10 min at room temperature ( RT° ) ] and washed twice with PBS buffer . Total protein concentration was determined by the Bradford method [69] and the final pellet was resuspended in STEN buffer ( 250 mM sucrose , 25 mM Tris-HCl pH 7 . 4 , 1 mM EDTA , 150 mM NaCl , 1 mM DTT and sigma-protease inhibitor mixture ) adjusted to a final concentration of 1 mg/protein in 200 μl . Cells were treated with variable concentrations of digitonin ( dissolved in STEN + dimethylformamide 40 mg/ml ) in a final volume of 300 μl for each treatment , incubated for 4 min at 25°C and centrifuged ( 2 min at max speed ) , as previously described [70] . Supernatants corresponding to solubilized fractions were mixed with 1x SDS Laemmli buffer and analyzed by western blotting . The presence of TbP5CDH , TbP5C-synthase , acetate:succinyl-CoA transferase ( ASCT ) , enolase , TbProDH and EP-procyclins was determined by antibody detection in parasite homogenates . Briefly , parasites were harvested as described above and resuspended in lysis buffer that contained: 20 mM Tris-HCl pH 7 . 9 , 1 mM EDTA pH 8 . 0 , 0 . 25 M sucrose , 50 mM NaCl , 5% ( v/v ) glycerol , 1% ( v/v ) Triton X-100 , 1 mM PMSF , 10 μg/ml aprotinin and 10 μg/ml leupeptin . Samples were chilled on ice ( for 40 min ) and clarified by centrifugation ( 15 , 000 g for 15 min at 4°C ) . Protein concentration was determined by the Bradford method using bovine serum albumin ( BSA ) as a standard [69] . Samples were submitted to protein electrophoresis ( SDS-PAGE ) and an equal amount of protein ( 30 μg ) was loaded per lane . Proteins were transferred into 0 . 2 μm PVDF membranes ( Amersham , GE , Life Sciences ) , blocked with PBS buffer plus 0 . 3% ( v/v ) Tween-20 ( PBST ) supplemented with 5% ( w/v ) skimmed milk powder and probed ( 16h at 4°C ) against specific sera . The enzyme TbP5CDH was probed with a polyclonal specific serum ( 1:4 , 000 ) raised against its T . cruzi ortholog ( TcP5CDH , TritrypDB accession number: Tc00 . 1047053510943 . 50 ) [28] . The enzyme P5C-synthase was probed with a polyclonal serum ( 1:3 , 000 ) produced in mouse against its close species T . cruzi , ( TcP5CS , access code: TCSYLVIO_005298 ) exactly as previously described [28] . For digitonin assays , extracted fractions were probed with rabbit polyclonal antibodies against T . brucei ASCT ( 1:1 , 000 ) , enolase ( 1:10 , 000 ) , PPKD ( 1:1 , 000 ) and ProDH ( 1:500 ) . EP-procyclins were probed with the monoclonal mAb-247 ( 1:1 , 500 ) , which recognizes the EP-repeats of T . brucei procyclins ( generous gift from Dr Terry W . Pearson , University of Victoria , Canada ) [57] . As loading controls , two different polyclonal antisera were used: the mouse anti-TcGAPDH ( 1:3 , 000 ) and anti-HSP60 ( access code: Tb427 . 10 . 6400 ) ( 1:2 , 000 ) , dissolved in PBST-skim milk . Membranes were washed three times and incubated with goat anti-mouse IgG horseradish peroxidase ( Sigma ) diluted in PBST ( 1:50 , 000 ) . Developing was done by using SuperSignal West Pico Chemiluminescent ECL substrate ( Thermo Scientific ) following the manufacturer’s instructions . PCFs were cultured up to mid-exponential growth phase in SDM79 . After this , parasites were washed with Voorheis’s modified PBS buffer ( vPBS: 137 mM NaCl , 3 mM KCl , 16 mM Na2HPO4 , 3 mM KH2PO4 , 46 mM sucrose , 10 mM glucose ) and harvested by centrifugation ( 850 g for 10 min at 4°C ) . Fixation , permeabilization and blocking were performed on poly-lysine coated glass slides , as previously described [71] . For antibody staining , polyclonal antisera produced against TbProDH ( 1:200 ) and TcP5CDH ( 1:250 ) [28] were dissolved in vPBS containing 20% ( v/v ) FBS and incubated for 2h at room temperature . Slides were washed five times with PBS and then incubated with AlexaFluor488-coupled goat anti-mouse IgG ( Invitrogen ) secondary antibody ( 1:600 ) plus TexasRed-X conjugated goat anti-mouse IgG ( H+L ) ( Invitrogen ) ( 1:400 ) for 1h . DNA staining was performed by adding 10 μg/ml of Hoechst probe ( Invitrogen ) and incubated for 5 min . Next , 2μl Fluoromount-G ( GE , Healthcare ) was added and a cover slip was mounted . Trypanosomes were visualized in a Leica DMi8 fluorescence microscope ( Leica Microsystems ) under an apochromatic 40x magnification lens . Image overlaying was done in imageJ software ( NIH , Bethesda , MA , USA ) . Wild type and RNAiTbP5CDH ( tet-/+ ) PCFs ( 5x106 parasites ) grown ( 3 days ) in complete SDM79 , as described above , were harvested ( 2 , 000g for 10 min at 4°C ) , washed twice with cold PBS , resuspended in 500 μl fixing solution [2% ( v/v ) formaldehyde and 0 . 05% ( v/v ) glutaraldehyde in PBS] and incubated for 20 min , as described before [72] . After fixation , parasites were washed twice and blocked in 200 μl PBS plus 2% ( w/v ) BSA ( PBS-BSA ) for 1h . Then , the cells were incubated with 200 μl monoclonal anti-EP procyclin solution ( mAb 247 diluted 1:500 in PBS-BSA ) for 2h [57] . After three washings with PBS , cells were incubated with a secondary antibody solution that contained goat anti-mouse IgG AlexaFluor-488 ( Invitrogen ) ( 1:1 , 000 in PBS-BSA ) for 1h and were protected from light . Flow cytometry analysis was performed in a FACSCalibur flow cytometer ( Becton Dickenson ) . FACS-acquired data were normalized using the unstained cells and only secondary antibodies provided as controls in the FlowJo v10 software ( Tree Star , Inc . ) . To analyze the mitochondrial functions in wt and RNAiTbP5CDH PCF cells , three parameters were taken into account: mitochondrial inner membrane potential ( ΔΨm ) , control of respiration and total ATP levels . After culture , cells were prepared as follow: parasites were harvested by centrifugation ( 1 , 000 g for 7 min at RT° ) and dissolved in buffer A with glucose ( BAG: 116 mM NaCl , 5 . 4 mM KCl , 0 . 8 mM MgSO4 , 50 mM HEPES-KOH , pH 7 . 2 and 5 . 5 mM D-glucose ) , as previously described [30] . Final densities were adjusted to 109 parasites/ml in BAG and kept on ice until further use . Parasite aliquots of 50 μl ( 5x107 cells ) of each group were used for measurements . The ΔΨm determinations was made spectrofluorometrically in parasites dissolved in cell respiration medium ( CRM: 125 mM sucrose , 65 mM KCl , 10 mM HEPES-KOH pH 7 . 2 , 1 mM MgCl2 , 2 . 5 mM potassium phosphate ) supplemented with 5 mM L-proline , 10 μM EGTA , 20% ( w/v ) non-fatty acids BSA ( NFA-BSA ) ( Sigma ) and 10 μM of the safranin-o dye ( Sigma ) , as previously described [73] . Changes in the fluorescence were recorded on a Hitachi 2500 spectrofluorometer ( λexi496nm , λemi586 nm ) at 28°C under constant stirring . Oxygen consumption was determined using a high-resolution oxygraph ( O2k , OROBOROS Instruments , Innsbruck , Austria ) , under constant stirring in a 2 . 1 ml final volume at 28°C . The reaction buffer was supplemented with NFA-BSA and EGTA as mentioned above . Assays were initiated by adding 5x107 parasites to the oxygraph chamber . After adding the cells to the tightly closed oxygen-chamber , preparations were supplemented with 5 mM succinate or 5 mM proline , as indicated in each experiment . In order to measure parameters at mitochondrial levels , parasite suspensions were further permeabilized by adding 40 μM digitonin . Data were recorded using DatLab software ( O2k , OROBOROS ) . In both measurements , additions of uncoupler or respiratory complex inhibitors were made as detailed in each experiment . ATP levels were determined using a luciferase bioluminescence assay ( Sigma ) according to the manufacturer’s indications . Briefly , the cells were harvested by centrifugation ( 2 , 000 g for 10 min at 4°C ) , washed twice with cold PBS and resuspended in the kit lysis buffer according to manufacturer´s instructions ( Sigma ) . The intracellular ATP contents were extrapolated from a standard curve with known concentrations of ATP disodium salt . Results were obtained from four separate biological replicates ( n = 3 ) . Statistical analysis was performed using a one-way ANOVA test in GraphPad Prism v5 . 0a for Mac OS X ( GraphPad Software , USA ) . PCFs use of glucose and proline as carbon sources was evaluated by nuclear magnetic resonance ( proton-NMR ) for the excreted end-products . Wt and RNAiTbP5CDH ( tet-/+ ) PCFs ( 106 parasites/ml ) were grown in complete SDM79 medium for 72h . Then , parasites were harvested by centrifugation ( 1 , 300 g for 10 min at 4°C ) and washed twice with PBS . Then , 5x108 parasites were transferred to 5 ml PBS supplemented or not with 4 mM L-proline + 4 mM D-[U-13C]-glucose . After 6h incubation at 26°C , cell suspensions were centrifuged and supernatants were submitted to NMR analysis , after adding 50 μl maleate ( 20 mM ) as an internal reference to a 500 μl aliquot of the collected supernatant . 1H-NMR spectra were performed at 125 . 77 MHz on a Bruker DPX500 spectrometer equipped with a 5 mm broadband probe head . Measurements were recorded at 25°C with an ERETIC ( Electronic REference To access In vivo Concentrations ) method , which provides an electronically-synthesized reference signal . Acquisition conditions were as follows: 90° flip angle , 5000 Hz spectral width , 32 K memory size and 9 . 3 s total recycle time . Measurements were performed with 256 scans for a total time of almost 40 min . Before each experiment , the phase of the ERETIC peak was precisely adjusted . Protons linked to acetate carbon C2 generate by 1H-NMR five resonances , a single peak ( unenriched acetate ) flanked by two doublets ( [13C]-acetate ) . | Bloodsucking insects play a major role in the transmission of pathogens that cause major tropical diseases . Their capacity to transmit these diseases is directly associated with the availability and turnover of energy sources . Proline is the main readily-mobilizable fuel of the tsetse fly , which is the vector of sub-species of Trypanosoma brucei parasites that cause human sleeping sickness and are partly responsible for animal trypanosomiasis ( Nagana disease ) in sub-Saharan Africa . Once trypanosomes are ingested from an infected host by the tsetse , the parasites encounter an environment that is poor in glucose ( as it is rapidly metabolized by the fly ) but rich in proline , which then becomes the main carbon source once the parasite differentiates into the first insect ( procyclic ) stage . In this work , we provide evidence on the essentiality of T . b . brucei proline catabolism for procyclic survival within the tsetse’s digestive tract , as this organism is unable to synthesize this amino acid and strictly depends on the proline provided by the fly . We also show that parasites deficient in TbP5CDH , a mitochondrial enzyme involved in the proline degradative pathway , failed to proliferate in vitro , showed a diminished respiratory capacity , and showed compromised maintenance of energy levels and metabolic flux when proline was offered as the main carbon source . Thus , the integrity of the trypanosome proline degradation pathway is needed to maintain essential functions related to parasite bioenergetics , replication and infectivity within the insect host . Our observations answer a long-standing question on the role of parasite proline metabolism in tsetse-trypanosome interplay . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"carbohydrate",
"metabolism",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"chemical",
"compounds",
"aliphatic",
"amino",
"acids",
"carbohydrates",
"parasitic",
"diseases",
"organic",
"compounds",
"neuroscience",
"glucose",
"glucose",
"metabolism",
"amino",
... | 2017 | Proline Metabolism is Essential for Trypanosoma brucei brucei Survival in the Tsetse Vector |
Voltage-sensitive dye imaging experiments in primary visual cortex ( V1 ) have shown that local , oriented visual stimuli elicit stable orientation-selective activation within the stimulus retinotopic footprint . The cortical activation dynamically extends far beyond the retinotopic footprint , but the peripheral spread stays non-selective—a surprising finding given a number of anatomo-functional studies showing the orientation specificity of long-range connections . Here we use a computational model to investigate this apparent discrepancy by studying the expected population response using known published anatomical constraints . The dynamics of input-driven localized states were simulated in a planar neural field model with multiple sub-populations encoding orientation . The realistic connectivity profile has parameters controlling the clustering of long-range connections and their orientation bias . We found substantial overlap between the anatomically relevant parameter range and a steep decay in orientation selective activation that is consistent with the imaging experiments . In this way our study reconciles the reported orientation bias of long-range connections with the functional expression of orientation selective neural activity . Our results demonstrate this sharp decay is contingent on three factors , that long-range connections are sufficiently diffuse , that the orientation bias of these connections is in an intermediate range ( consistent with anatomy ) and that excitation is sufficiently balanced by inhibition . Conversely , our modelling results predict that , for reduced inhibition strength , spurious orientation selective activation could be generated through long-range lateral connections . Furthermore , if the orientation bias of lateral connections is very strong , or if inhibition is particularly weak , the network operates close to an instability leading to unbounded cortical activation .
Horizontal connections link cells separated within the same cortical area , over a distance of a few millimeters ( mm ) covering several iso-functional columns [1 , 2] and spatially distributed into regular clusters [3–5] . In V1 , the lattice-like pattern of connectivity is reminiscent of the spatial regularity observed in orientation maps and has therefore been proposed to link neurons with similar preferred orientation [6] . Functional mapping combined with retrograde labeling [7] has shown that pyramidal cell horizontal axons have a bias to preferentially connect iso-orientation loci ( in cat [8 , 9] , tree shew [10] and monkey [11] ) . However , this result depends on cell type and location since neurons in layer 4 ( L4 ) [12] , pinwheel centers ( [13] but see [14 , 15] since this result may depend critically on the distance to the pinwheel center ) and inhibitory cells [16 , 17] connect without orientation bias . As a consequence of this cellular heterogeneity , it is not trivial to predict the selectivity of stimulus-driven horizontal activation . More information was recently gained from population measures of orientation selectivity at mesoscopic scales . Techniques such as optical imaging has led to the description of cartographic organization of many cortical structures [18 , 19] . The development of voltage-sensitive dye [20] has further allowed for investigation of dynamic features and computations arising within these maps , see reviews in [21] and [22] . In [23] voltage sensitive dye imaging ( VSDI ) was used to study the retinotopic activation with localized oriented inputs in cat V1 . A characteristic plateau of activity , coinciding with the retinotopic extent of the stimulus , was independent of stimulus orientation . Within the plateau several peaks of activation would appear , with location strongly dependent on orientation . [24] further explored orientation selectivity outside the retinotopic activation using localized stimuli . Over several hundred milliseconds activation gradually propagated outwards , extending several mm beyond the feedforward footprint ( FFF ) . The dynamic and the spatial range of this activation beyond the FFF is presumably generated by long-range excitatory connections in L2/3 of V1 . Note that , although less plausible because their spatial and temporal properties are not in the appropriate range ( see discussion in [25] and [24] ) , alternative circuits such as intra-thalamic , thalamo-cortical divergence and feedback loops cannot be entirely dismissed ( but see [26] ) . Interestingly , only a local component of this activation , circumscribed within the FFF , was found to be orientation selective ( Fig 1A ) , at first glance a surprising finding given the numerous studies showing an iso-orientation bias for long-range connections . However , as discussed in [24] and introduced above , this result may be expected from the anatomy known from studies already published at that time . First , because the iso-orientation connection bias is small and has been quantified only for short horizontal distances ( mostly <1–1 . 5 mm , a distance for which a similar small bias is also reported by [24] ) . Second , because it can be seen in the few existing intracellular labeling studies that the iso-orientation bias tends to decrease with distance [17 , 14] . Thus , the results from [24] , although in contradiction with a strict “like-to-like connectivity” principle , call for modification depending on cortical distance: from a like-to-like bias in functional connectivity at short range toward no bias at long range . Importantly , two follow-up studies consolidated these findings with an optogenetics functional confirmation ( [27] , see also [28 , 29] ) and a timely anatomical clarification using quantitative and statistical analysis of intracellular labeled neurons [15] . The aim of this modelling study is to explore the relationship between lateral connectivity properties ( structure ) and the way activity spreads across cortex as evoked by localized , oriented stimuli ( function ) . It remains to reconcile the studies of iso-orientation bias of anatomical connectivity in V1 [7 , 9 , 14] and the properties of cortical activation observed in other experiments [24 , 27 , 15] . To achieve this , we have developed a dynamical model to investigate stimulus driven activity in 2D cortical space with a discrete representation of orientation . The neural field equation gives a coarse-grained description of average membrane potential on a continuous domain [30 , 31] . This framework has been widely used to model visual cortex ( e . g . [32–35] ) and cortical dynamics in general ( see reviews [36] and [37] ) . This level of description is well-suited for comparison with VSDI [38–40] . The model’s design allows for connectivity properties such as the spatial profile of excitation and inhibition to be investigated . Model parameters were constrained to be consistent with the level of orientation bias reported in [14] . Our study shows that realistic cortical connectivity schemas [14 , 15] govern a spatio-temporal dynamics of cortical activation in accordance with the observed functional dynamics [24 , 27] . Our approach allowed us to further elaborate on the non-trivial link between anatomical constraints and predictions of population level activation patterns . To probe the structure-function link , we used our model to make experimentally tractable predictions on the specific role of excitatory-inhibitory balance .
Our neural field equation model [30 , 31] gives a mesoscopic , continuous description of neural activity on a 2D plane ( x , y ) parallel with the cortical surface in layer 2/3 of V1 . Neural activity is represented by an average membrane potential ui ( x , y , t ) evolving with time t in four sub-populations , each associated with a different orientation i = {0° , 45° , 90° , 135°} ( the ° notation will be dropped in the remainder of the manuscript ) . The decomposition into sub-populations is a convenient abstraction for modelling purposes; these sub-population outputs ui ( x , y , t ) will be combined into activity variables that can be compared with experimental data in Conversion of model output to VSD-like signal . The following integro-differential equation describes the dynamics of the ui population: τ ∂ ∂ t u i ( x , y , t ) = − ∑ j ρ j u j ( x , y , t ) ( 1 ) + ∑ j k j I j ( x , y ) ( 1 + β inp J j ( x , y ) ) ( 2 ) + S ( u i ( x , y , t ) ) ⋆ ( x , y ) [ g ex w E loc ( x , y ) − g in w I ( x , y ) ] ( 3 ) + ( 1 + β rec J i ( x , y ) ) S ( u i ( x , y , t ) ) ⋆ ( x , y ) g ex w E lat ( x , y ) . ( 4 ) The cortical timescale is τ = 10 ms , and the rate of change of ui is proportional to the right hand side of this equation with the following terms ( 1 ) Decay of the population activity to resting potential ( 2 ) Stimulus driven input with feedforward footprint I , modulated by the orientation map J ( 3 ) Non-selective intra-cortical interactions gated by sigmoidal threshold function S ( 4 ) Orientation-selective interactions , modulated by the orientation map J Term ( 1 ) describes a decay back to an resting potential of ui = 0 ( the membrane potential has arbitrary units ) . Within the sub-population the decay term has strength ρi=j = 1 ( an arbitrary choice for this first constant ) . Between sub-populations the decay term has strength ρi≠j = 0 . 1 ( local , linear cross inhibition ) . For ρ > 0 . 2 this cross inhibition can generate undesired above-threshold activation in non-stimulated sub-populations , therefore , a value 50% below this was selected . Term ( 2 ) describes localised circular inputs with one of four orientations ( e . g . Fig 1B ) . For an input with say orientation 0 , I0 is active and I45 , I90 and I135 are zero . The input weighting for a sub-population’s associated orientation ki=j = k1 is larger than the weighting for other orientations ki≠j = k2 , with k1 = 2k2 . Inputs are modulated by the orientation map Ji with strength βinp = 0 . 25 ( Fig 1D and 1I ) . For βinp > 0 . 2 the spatial phase of multi-bump patterns of activity will match the orientation preference map as required , increasing it further has little effect . Term ( 3 ) describes non-selective lateral connections within the sub-population ui via a convolution ⋆ over ( x , y ) ( spatial integrals , see ( 6 ) ) with the components of the spatial connectivity profile . This profile is radially symmetric and describes the average connectivity at any point in cortex . It is broken down into local excitatory w E loc and inhibitory wI components , see Fig 1G and Details and equations for connectivity profile . In the convolutions ui is processed through the sigmoidal threshold function S , which transforms membrane potential into a normalized firing rate . Any locations above threshold will influence their neighbors through lateral connections . Weighting constants gex and gin are determined by theoretical constraints , see Details and equations for connectivity profile and ( 17 ) . Term ( 4 ) describes orientation-selective excitatory connections as modulated by the preference map Ji with strength βrec . An example of the simulated sub-population activity is shown in Fig 1F for an input with orientation 0 . The main parameters of interest in this study are RWex , which modifies the spatial width of excitatory peaks ( Fig 1G ) and βrec ( taking values between 0 and 1 ) , which controls the amplitude modulation of long-range excitation by the orientation preference map ( Fig 1H ) . We will also study a parameter C controlling the strength of inhibition ( equivalently the balance between excitation and inhibition ) as defined by ( 16 ) in Details and equations for connectivity profile . The firing rate ( or threshold ) function is given by a sigmoid S ( u ) = 1 1 + e - μ u + θ - 1 1 + e θ , μ , θ > 0 , ( 5 ) where μ = 2 . 3 is a slope parameter and θ = 5 . 6 the threshold . Input strengths k1 = 2 . 8 and k2 = 1 . 4 are set such that inputs to the stimulated sub-population are above threshold in ( 5 ) μk1 > θ , whilst inputs to other sub-populations are below threshold μk2 < θ . The particular form of S is chosen such that ui = 0 ( all populations at resting potential ) is a always a solution to the model equations and values of μ and θ are chosen such that this is the only stable solution with no inputs [41] . The spatial convolution terms in ( 3 ) – ( 4 ) are expressed as integrals , e . g . Term ( 4 ) can be computed as one integral wElat ( x , y ) ⋆ ( x , y ) [ S ( ui ( x , y , t ) ) ( 1+βrecJi ( x , y ) ) ]=+∫x , ywElat ( x−x′ , y−y′ ) [ S ( ui ( x′ , y′ , t ) ) ( 1+βrecJi ( x′ , y′ ) ) ]dx′dy′ , ( 6 ) with dummy variables x′ and y′ . In this case the modulation by Ji is introduced whilst still preserving the convolutional structure , allowing for the numerical methods described in [41] , exploiting Fast Fourier Transforms to be applied . The radial inputs Ii ( r ) , r = x 2 + y 2 , as plotted in Fig 1I ( left ) , are given by I i ( r ) = { 1 , r < 0 . 7 Λ h ( r , 0 . 7 Λ , 0 . 3 Λ ) r ≥ 0 . 7 Λ . ( 7 ) where h is the radially shifted Gaussian ring given below in ( 10 ) . The extent of Ii and its decay at the stimulus border were chosen to match cortical point spread functions measured in [24] . In simulations the input amplitude ramps up linearly from 0 to full amplitude in the interval t ∈ [20 , 120] ms , see [40] . For each sub-population a finite differencing scheme for the domain [−L , L] × [−L , L] with L = 30 and N = 128 evenly distributed gridpoints in each spatial direction is used , as described in [41] . Model simulations were run on a domain much larger ( 2L = 60 , relative to Λ = 2π ) than the localized patterns of activation studied here , justifying the use of periodic boundary conditions . A standard Runge-Kutta time stepper in Matlab was used for model simulations with default tolerances . The source code for the full implementation of the model has been made available in the Supplemental Information ( S1 Code ) . Earlier theoretical works characterized connectivity constraints that lead to stable localized patterns of activation in 1D [42 , 43] and in 2D [44 , 45] . A common choice of connectivity function is a so-called Mexican hat ( e . g . a difference of Gaussians ) featuring a broader footprint for inhibition than for excitation [36 , 37] . [41] suggested a role for longer-range excitation serving to stabilize larger patterns of activation and equating the separation between peaks in excitation with Λ ( hypercolumn separation , [46] ) The formulation presented here is further inspired by the way [14] quantified their results so that we can link model parameters with their quanitification or orientation bias for long-range connections ( see next section ) . The model’s connectivity profile is broken down into local excitatory w E loc ( a local Gaussian bump ) , lateral excitatory w E lat ( Gaussian rings centered at Λ and 2Λ ) and inhibitory wI ( a broad local Gaussian bump ) components , which were plotted in Fig 1G . The use of Gaussian rings was inspired by [16] ( see their Fig 16 ) , noting the important features that excitatory connections 1 ) drop in number at a range Λ/2 , 2 ) have a peak at a range Λ and 3 ) can extend several mm across cortex . The following details give full definitions of these components and their relative scaling with particular attention to the global balance between excitation and inhibition , which will be controlled by a parameter C . When C = 0 excitation and inhibition are balanced ( the area under the 2D radial versions of the blue and red curves in Fig 1G would be equal ) . In this study C is taken to be negative ( net inhibition ) whilst remaining close to the balanced condition . We note that when C is larger ( or even positive ) localized patterns of activity are more likely to destabilize and spread across cortex [41] , which is investigated in Reduced inhibition leads to orientation selective activation outside stimulus footprint . The length scale Λ is the mean hypercolumn separation . Radially symmetric functions for the connectivity components are defined in terms of a radial coordinate r = x 2 + y 2 . We define a 2D Gaussian function with spatial decay rate σ: g ( r , σ ) =12πσ2e ( −r22σ2 ) , ( 8 ) where the pre-factor 1 2 π σ 2 normalizes the area . The number of inhibitory connection , based on a diverse class of inhibitory neurons [16] , is assumed to have a Gaussian decay with distance from the origin: w I ( r ) = g ( r , R W in ) , ( 9 ) with RWin = 0 . 55Λ ( a cross-section of this function is plotted red in Fig 1G ) . This value gives the qualitative feature that there is more excitation than inhibition at ranges Λ and above . The results are not contingent on the exact value chosen , but varying RWin has a similar effect to varying C ( inhibition strength , defined below ) , which is investigated in Reduced inhibition leads to orientation selective activation outside stimulus footprint . Assuming that there are peaks in the number of excitatory connections ever Λ-distance from the origin up to 2Λ , there are rings of excitation at distances {0 , Λ , 2Λ} . We assume that the amplitude of the peaks centered at these distances decay within an exponential envelope χ ( r , ζ ) = e - r ζ , where ζ = 0 . 625Λ . The exact value chosen is not critical , varying ζ by ±20% does not significantly affect the results . We define a radially shifted 2D Gaussian h ( r , r0 , σ ) =e ( − ( r−r0 ) 22σ2 ) , ( 10 ) which describes a ring that is maximal at a radius r = r0 and decays away with spatial scale σ ( note that 1 2 π σ 2 h ( r , 0 , R W ex ) = g ( r , R W ex ) ) . The local excitatory component ( a Gaussian bump centered at 0 ) is given by w E loc ( r ) = h ( r , 0 , R W ex ) , ( 11 ) where RWex is a free parameter . For RWex < 0 . 1 the number of excitatory connections at Λ/2 would drop to 0 , which is not realistic . Further RWex should be less than RWin; for our parameter exploration , we therefore consider a smaller range [0 . 1 , 0 . 4] based on the anatomical constraints introduced later . The long-range excitatory component ( Gaussian rings centered at Λ and 2Λ ) is given by w E lat ( r ) = χ ( Λ , ζ ) h ( r , Λ , R W ex ) + χ ( 2 Λ , ζ ) h ( r , 2 Λ , R W ex ) . ( 12 ) The overall profile of excitation w E loc + w E lat is plotted in Fig 1G . Noting the following analytic expression for the zero-mode of the Fourier transform of ( 10 ) H ( 0 , r 0 , σ ) = 2 π σ 2 e - r 0 2 2 σ 2 + π σ r 0 2 π ( 1 + erf r 0 2 σ ) , ( 13 ) where erf is the standard error function , we write the normalisation pre-factor for the combined excitatory components w E loc + w E lat B E = 1 / [ 1 2 π R W ex 2 + H ( 0 , Λ , R W ex ) + H ( 0 , 2 Λ , R W ex ) ] . ( 14 ) We now define the normalized combined excitatory profile as w E = B E ( w E loc + w E lat ) , ( 15 ) which by design has zero-order Fourier mode of 1 . The complete connectivity function is w ( r ) = P [ w E ( r ) + ( C - 1 ) w I ( r ) ] , ( 16 ) where C is a constant controlling the relative strength of excitation and inhibition . When C = 0 excitation and inhibition are balanced and when C is negative there is net inhibition globally . We introduce a constant P that matches the value of W’s largest Fourier mode with the connectivity used in [41] . This final scaling of the overall connectivity allows us to manipulate any of the connectivity parameters , whilst keeping all non-connectivity parameters constant ( e . g . input and threshold function parameters ) . Failing to do this means that the correct operating region of the model would shift each time , say , RWex was modified . Two constants gex and gin in ( 3 ) – ( 4 ) are given by g ex = B E P , and g in = P ( C - 1 ) . ( 17 ) Retinotopic space is mapped to the surface of V1 [47] and different locations show preference for a variety of low-level visual features such as spatial frequency , ocular dominance , orientation and direction of motion [48 , 49] . Cortical neurons exhibiting similar preference for low-level features are organized in a columnar fashion [50 , 51] . Orientation preference varies incrementally along the cortical surface [52] , defining a quasi-periodic organization characterized by linear zones and pinwheels , singularities about which all orientation preferences are present along a circular path [18] . The orientation maps have a quasi-periodic organisation with a regular length scale of around 0 . 5–1 mm ( depending on species and cortical areas ) , which can be measured as the mean distance between iso-orientation domains . This length scale , which we denote Λ , is reflected as a peak in the Fourier spectrum of the preference map represented as polar argument [53 , 46] . Orientation preference can be deduced from single-electrode recordings [54] or with optical imaging of intrinsic signals [18 , 10] . The characteristics , artificial generation and biological development of orientation preference maps have been investigated in various modelling studies [55 , 56 , 46 , 57] . [56] showed the importance of long range connections in generating the quasi-periodic repetition of key map features in a canonical pattern forming system with spatially extended complex representation of orientation . [46] showed that when normalized by the regular length scale , orientation preference maps show a constant pinwheel density; see also [58 , 59] . [57] explored mechanisms for the stable development using a Hebbian learning method for connections in a two-stage ( LGN , V1 ) model and found an important role for adaptation and normalisation . In this study we generated realistic maps specifically for our planar model with discrete representation of orientation ( in four-sub-populations ) . The maps were generated specifically for the periodic domain used in our model using a spatial Hebbian-like learning rule operating on the converged model output u i , n fin from a series of localized inputs Ii with random orientations and at random locations . After each simulation the Ji were updated via the following rule: Ji , n+1=Ji , n+HaIi ( 1−〈 | Ji , n |⋆G 〉[ 0 , 1 ] ) ui , nfin , Ji , n+1=〈 Ji , n+1 〉[ −1 , 1 ] . Ha is the learning rate , G a smoothing kernel ( 8 ) with σ = 0 . 6Λ , and 〈 . 〉 rectification on the given interval . The smoothing ensures that regions that already have local structure are modified less than regions without local structure . Learning was initiated from a homogenous initial set of Ji = 0 ( where the model produces localized multi-bump states in the stimulated area and sub-population ) . The learning process converged ( e . g . pinwheel density stabilizes ) after around 1600 steps as the map takes on structure across the whole domain ( after this the term 1 − 〈|Ji , n| ⋆ G〉[0 , 1] remains close to 0 across the whole domain , although small changes in the map continue ) . Final maps ( at 6400 steps ) were high-pass and low-pass filtered as described in [46] . The maps used in this study are included as part of the simulation code ( see Supplemental information S1 Code ) , thus allowing the results to be independently reproduced . Full details of the method will be the subject of a separate study . Alternatively , one could use maps obtained experimentally directly with our model , but this would not have any significant effect on the results presented . For the purposes of the present study it is sufficient to show that the maps used in our model show characteristic features of realistic maps including linear zones , pinwheels and the regular length scale Λ . Fig 2A shows the component maps Ji , each associated with a different orientation , that in ( 1 ) – ( 4 ) modulate inputs with strength βinp and long range excitatory connections with strength βrec . In conjunction , the composite maps combine to give the orientation preference map shown in Fig 2B , which has a regular length scale Λ characterized by a sharp peak in the map’s spectral power curve ( Fig 2B ) . The effective tuning of lateral connections is computed across ranges of the connectivity parameters RWex ( width of peaks in excitation ) and βrec ( orientation bias of long-range lateral connections ) . These parameters were chosen as they have a strong effect on the anatomical measure of interest over ranges where the model is well defined . Another choice could be ζ ( controlling the decay of peaks in excitation ) , but this would be redundant with ζ having a similar effect as varying RWex . These computations allow for a direct comparison with the anatomical-data-based model analysis presented in [14] , which reported the orientation bias of long-range lateral connections in V1 L2/3 . Optical imaging was used to find orientation preference maps and combined with intracellular labeling of lateral projections of pyramidal cells to identify target locations relative to the preference map . The orientation bias was quantified by tuning curves of the orientation preference at axon terminals relative to orientation preference in a region local to the originating cell’s body . Tuning curves were quantified by parameter-fits to a von-Mises distribution ( circular normal distribution ) . Our model was developed to be able to provide a direct point of quantitative comparison with these measures . The distribution is defined on a circular domain and parametrized by a tuning coefficient κ ≥ 0 ( 0 is untuned , larger is more tightly tuned ) and a preferred orientation μ ∈ [0 , 180 ) : f ( x ; μ , κ ) = e κ cos ( x - μ ) 2 π I 0 ( κ ) . ( 18 ) I0 ( κ ) is the modified Bessel function of order 0 . [14] reported values of κ in the range 0 . 7–1 . 2 for population-level tracing of lateral connections . We perform a similar computation for the connectivity function in our model . The computation of the effective orientation tuning of lateral connections , at one specific map location ( Fig 3A ( top ) ) , is illustrated in Fig 3A–3E . For the orientation preference map one can compute an orientation tuning curve by counting the number of pixels falling within equally sized orientation bins , as shown in Fig 3A ( bottom , circular markers ) . All orientations are represented with equal probability so the profile is untuned and a best-fit von-Mises distribution ( solid curve ) has κ ≈ 0 ( fit determined using a least-squares minimisation with Matlab’s lsqcurvefit for κ and μ in ( 18 ) ) . A spatial weighting for the pixel count can be introduced in order to find the tuning of orientations in some local region . A 2D Gaussian function ( Fig 3B ( middle ) ) was used to give the local weighting shown in Fig 3B ( top ) . In most regions of the map this results in a sharp tuning ( note large κ ) centered at one specific orientation as shown in Fig 3B ( bottom ) . Introducing the radial profile of excitatory connections from the model as a weighting function ( Fig 3C ( middle ) , see ( 15 ) ) one can compute the effective tuning of the lateral connections . With βrec = 0 the weighting function is radially symmetric , nevertheless there is an expected weak tuning of the weighted connections ( Fig 3C ( bottom ) ) due to the structure of the orientation preference map; there is a slight bias toward finding similar orientations at a range Λ ( annular ring ) relative to the origin ( star ) . Increasing βrec introduces an orientation-specific bias in the weighting profile for the long-range connections ( Fig 3D–3E ( middle ) ) . This results in locations with similar orientations to the origin being targeted specifically by long-range connections ( Fig 3D–3E ( top ) ) and a corresponding increase in the tuning strength κ ( Fig 3D–3E ( bottom ) ) . As one might expect the tuning strength of connections ( κ ) increases monotonically with increasing βrec or decreasing RWex , which allows us to define an operating range for the model before running simulations . Fig 3F shows κ ( average value from 50 randomly selected map locations ) computed at combinations of RWex ∈ [0 . 1 , 0 . 4] and βrec ∈ [0 , 1] . Map locations close to pinwheels ( 7/50 ) identified by having a local tuning ( computed as in Fig 3B ) with κ < 1 were excluded ( this exclusion had a very minor effect ) . The map shows that κ increases with βrec ( as expected ) and decreases with RWex . Solid white contours show a band of values for RWex and βrec where the tuning of connections in the model is consistent with the anatomical data ( κ ∈ [0 . 7 , 1 . 2] ) . These contours are later replotted in Orientation selective activation is restricted to stimulus footprint in the anatomical parameter range and Reduced inhibition leads to orientation selective activation outside stimulus footprint for comparison with model simulations . For RWex much beyond 0 . 4 the anatomical data cannot be matched as βrec must be <1 . The process of converting the model output from individual simulations into a VSD-like signal , and the method for computing the general and orientation selective activation , is illustrated in Fig 4 . For a given oriented simulation , the Ii are weighted toward the specific orientation ( Fig 1A ) . The input is further modulated by Ji if βinp > 0 ( Fig 1B ) and recurrent connections in each sub-population are modulated by Ji if βrec > 0 . The sub-population corresponding to the input orientation responds above threshold in a multi-bump pattern with the location of the bumps determined by Ji , whilst the other sub-populations have a sub-threshold response ( Fig 4A ) . We first transform the sub-population variables ui into a VSD-like signal following a similar method to the one proposed in [38] . The sub-population membrane potentials in Fig 4A are converted into a firing rate , processed through weighted excitatory and inhibitory connectivity profiles , summed across the sub-populations and diffused with a Gaussian profile ( representing the attenuation and diffusion of the signal in cortical tissue ) . The contribution from inhibition is assumed to be in the range 15–20% [60 , 61] . The resulting optical imaging signal OI0 in Fig 4B ( top left ) was computed from the four sub-population responses in Fig 4A . For different orientation inputs , the other OI signals can be computed in a similar fashion ( Fig 4B , other panels ) . The general activation Act is computed as an average across the responses to inputs with four different orientations; the response in Fig 4C was computed as an average of the four responses in Fig 4B . The preference Pref and selectivity Sel ( Fig 4D ) are computed by transforming the four OI signals into polar coordinates from difference maps ( OI0 − OI90 and OI45 − OI135 ) where the angular coordinate ( argument ) is the preference and the radial coordinate ( magnitude ) is the selectivity [62] . Thresholds delineating the general activated area and orientation selective area are defined as a fraction of the average activation or selectivity inside the FFF . The radial profiles ( average radial decay ) of Act and Sel are characterized by the parameters of best-fit Naka-Rushton functions [63] ( a widely-used , smooth , monotonically decaying function ( 29 ) , which has been used to fit , e . g . , contrast response data [64] ) . Further details and equations of all the processing steps are given below . Following the method described in [38] , to account for the known optical diffusion of the VSD signal ( light ) in V1 L2/3 , we convolved the signal with a Gaussian distribution ( 8 ) with σOI = 0 . 075Λ . Note that the results in this study are not contingent on this specific value , there being little effect of either increasing or decreasing σOI by a factor of 2 . The unattenuated imaging signal ui for each sub-population is assumed proportional to the post-synaptic membrane potential . Hence we first computed the dynamics of mean pre-synaptic membrane potential for each sub-population ui as given by ( 1 ) – ( 4 ) , which is converted to a firing rate of the pre-synaptic neurons through S . The postsynatpic population response was then computed via a convolution of the presynaptic firing rate with the connection profile . There is an 85% contribution from excitation and a 15% contribution from inhibition , leading to a weighting for inhibition of pI = 0 . 177 . These are summed across the sub-populations i to give the total unattenuated signal ( expression in large parentheses ) . Finally the optical imaging ( OI ) signal is computed as a convolution of the total unattenuated signal with a Gaussian: O I ( x , y ) = ( ∑ i S ( u i ( x , y ) ) ⋆ ( x , y ) [ w E loc ( x , y ) - p I w I ( x , y ) + w E lat ( x , y ) ( 1 + β rec J i ) ] ) ⋆ ( x , y ) g ( x , y , σ OI ) . ( 19 ) This equation converts the model’s state variables ui for an input stimulus with specific orientation ( e . g . 0° as in Fig 4A ) into a VSD-like signal ( e . g . OI0 in Fig 4B , top left ) . We are interested to explore how cortical activation spreads over time . One potential issue with the temporal dynamics in the model , and ( 19 ) , is the assumption that activity generated through long range lateral connections ( w E lat ) propagate instantaneously . Although the model converges to the correct final state , the transient dynamics may not be captured exactly . To solve this , whilst avoiding the introduction of say delay terms in ( 1 ) – ( 4 ) , we assume that there is a slower timescale τlat = 240 ms for the portion of the OI-signal generated through w E lat: O I ( x , y , t ) = ( ∑ i S ( u i ( x , y , t ) ) ⋆ ( x , y ) [ w E loc ( x , y ) - p I w I ( x , y ) + ( 1 - e - t / τ lat ) w E lat ( x , y ) ( 1 + β rec J i ) ] ) ⋆ ( x , y ) g ( x , y , σ OI ) , ( 20 ) where t is the time after stimulus onset . The introduction of τlat is done at the post-processing stage only and its value was chosen to match data from [24] . For four sequential simulations , each with an input with different orientation , the OI signal can be computed ( Fig 4B ) . The general activation is the average of these signals: Act ( x , y , t ) = 1 4 ∑ j O I j ( x , y , t ) , j = 0 , 45 , 90 , 135 , ( 21 ) as shown in Fig 4C . Before computing the preference Pref and selectivity Sel , the VSD signals are normalized by a scale factor that accounts for differences in the maximum value over ( x , y ) across the four simulations with different orientations . Two difference maps between the normalized VSD signals from simulations with orthogonal inputs are computed: D 1 ( x , y , t ) = O I 0 ( x , y , t ) − O I 90 ( x , y , t ) , ( 22 ) D 2 ( x , y , t ) = O I 45 ( x , y , t ) − O I 135 ( x , y , t ) . ( 23 ) The orientation preference of the activation is given by Pref ( x , y , t ) = Arctan ( D 1 ( x , y , t ) , D 2 ( x , y , t ) ) , ( 24 ) where Arctan is the four quadrant inverse tangent , and the selectivity strength is given by Sel ( x , y , t ) = D 1 ( x , y , t ) 2 + D 2 ( x , y , t ) 2 . ( 25 ) In the results section , all plots of Sel ( x , y , t ) and Act ( x , y , t ) are scaled by 1 . 1× their values at the final time point tfinal in the simulation , thus showing the time-evolution relative to the final state . The thresholds contours Tact and Tsel for activation and selectivity ( and corresponding areas Aact and Asel ) are set as a fraction of the mean activation and selectivity within the FFF at the final time point , Act ¯ FF = mean r < r FF ( Act ( x , y , t final ) ) , ( 26 ) Sel ¯ FF = mean r < r FF ( Sel ( x , y , t final ) ) , ( 27 ) where rFF is the FFF boundary . The thresholds are given by T a c t= η Act Act ¯ FF , T s e l= η Sel Sel ¯ FF , ( 28 ) where ηAct = 0 . 2 and ηSel = 0 . 5 . These thresholds were chosen ad hoc . In order to ensure the results obtained weren’t contingent only on these choices , we further characterized the radial decay rates of the general and orientation selective activation . We also look at radial profiles of the general and selective activation patterns ( radial decay of Act and Sel ) . Radial profiles Act ( r , t ) and Sel ( r , t ) are computed by re-meshing Act ( x , y , t ) and Sel ( x , y , t ) on radial coordinate system ( r , θ ) centered at the stimulus center and averaging in the angular coordinate θ . The decreasing Naka-Rushton function NR used here decays to zero as r increases: N R ( r ) = R max ( 1 - r n r n + r 50 n ) , ( 29 ) where the exponent n > 0 describes the steepness , the maximal value is Rmax and the half-max r-value is r50 > 0 . Best-fit Naka-Rushton functions were determined by minimising , for example the least-squares distance between Act ( r , tfinal ) and NR ( r ) varying the parameters n , Rmax and r50 using Matlab’s lsqcurvefit function .
The spatio-temporal dynamics are investigated for a specific case with inputs modulated by the orientation preference map ( βinp = 0 . 25 ) but with radially symmetric lateral connections ( βrec = 0 , no orientation bias in lateral connections ) . The dynamics produced by the model will be illustrated for two values of the width of the peaks of excitatory connections RWex . When RWex is small , excitatory connections cluster on tightly on rings at distances Λ and 2Λ away from the origin . When RWex is larger the clustering at these specific ranges is more diffuse ( Fig 1G ) . Simulations were run , sequentially for four different orientations , with a radial input centered at a specific location in the orientation preference map ( Fig 5A ) . Fig 5B and 5E ( top ) show that the general ( averaged ) activation spreads outward from the center of the stimulated region ( note that simulations are computed on a domain around 3 times larger than the window shown here ) . Only some portion of this general spread of activation is orientation selective as shown in the bottom panels ( area inside white contour is significantly smaller than area inside grey contour ) . When , RWex is small the selective activation extends outside the FFF ( Fig 5B ( bottom , right ) ) , and when RWex is larger it is confined to the FFF ( Fig 5E ( bottom , right ) ) . The temporal dynamics of the general and orientation selective areas ( Fig 5C and 5F ) shows that the rate of the area increase slows down after roughly the first 100 ms . Nevertheless , when RWex is small the spread of activation persists after 550 ms as shown by the blue/green curves still increasing in Fig 5C ( eventually converges at ∼ 700 ms ) . When RWex is larger the orientation selective activation converges after around 200 ms although the general activation continues to increase slowly ( Fig 5F ) . This latter behavior is compatible with the observations made using VSDI in [24] . The radial profile of the general and selective activation shows the decay with distance from the center of the stimulated region ( Fig 5D and 5G ) . When RWex is small , the profile of the general and selective activation exhibits a similar decay profile as marked by similar values of the exponent n in a best-fit Naka-Rushton function for the two profiles ( Fig 5D ) . For a larger RWex the exponent for the orientation selective activation is substantially larger ( nSel = 14 . 47 ) , indicating a steep transition from high to low selectivity , than for the general activation ( Fig 5G ) . Even with lateral connections not modulated by the orientation preference map ( no orientation bias in lateral connections ) , orientation selective activation can be generated outside the FFF of the stimulus due to convergent excitation generated from regions directly stimulated within the FFF . Indeed , activated locations with the same preference mutually excite each other at a range Λ . These activated regions can further generate overlapping excitation at Λ-equidistant points , either inside the FFF , or potentially outside the FFF . Orientation selective activation generated in this way , outside the FFF , does not necessarily agree with the orientation preference map ( Fig 5B , bottom right , region inside the white contour but not the yellow ) . This occurs if the range of peaks in excitation are highly specific ( small RWex ) as in Fig 5B–5D . These findings illustrate that with RWex excessively small , the ringed connectivity can lead to undesired behaviour that is inconsistent with activation observed in experiments . We note that later in section Reduced inhibition leads to orientation selective activation outside stimulus footprint we find that spurious activation ( not agreeing with the preference map ) can be generated by another mechanism ( destabilization of activity ) . When the peaks in excitation are broader ( RWex > 0 . 2 ) , the orientation selective activation decays quickly at the border of the stimulated region and no such miss-tuned activation ( as generated by convergent excitation ) is observed outside the FFF . We have seen how our modelling results allow us to distinguish between patterns of activation that are consistent with imaging studies in terms of 1 ) the area and range of general ( grey ) and selective ( white ) activation relative to the FFF ( red ) in Fig 5B and 5E ( bottom ) ; 2 ) whether activation reflects the correct orientation with respect to the underlying preference map ( difference between yellow and white contours in Fig 5B and 5E ( bottom , right ) ; and 3 ) the rate of decay of general and selective activation in Fig 5D and 5G as quantified by n ( larger is a steeper decay ) . The simulation shown in Fig 5E–5G is consistent with [23] and [24] because 1 ) in E ( bottom , right ) the grey contour extends much further than the white , 2 ) in E ( bottom , right ) the yellow and white contours agree closely , 3 ) in G the exponent n is larger for the selective activation ( blue ) than for the general activation ( green ) . Further , this steep decay of selectivity is consistent with there being a plateau of highly selective activation inside the FFF that transitions sharply to low selectivity outside the FFF . The effect of introducing an orientation bias to the long-range lateral connections ( βrec > 0 ) is shown in Fig 6 . For another location in the orientation preference map ( Fig 6A ) the orientation selective activation is shown for different values of βrec in Fig 6B–6D with the radial profile shown in F–I . The parameter values for panels Fig 6B and 6F are the same as those for Fig 5E–5G ( RWex = 0 . 25 and βrec = 0 ) , the only difference is the location in the orientation preference map . These two examples show a qualitatively similar spatial profile , and in general , the resulting spatio-temporal dynamics are independent of the location in the orientation preference map ( five locations were randomly chosen for the results in this paper ) . Across the five locations there is a large range of nSel ( between 6 and 15 ) when RWex = 0 . 25 and βrec = 0 . However , this variability in nSel across locations is much less ( between 5 and 7 ) when βrec is increased to say 0 . 5 . As βrec is increased , more orientation selective activation is observed increasing from small isolated patches outside the FFF at intermediate values ( Fig 6C–6D ) to many larger for βrec = 0 . 9 ( Fig 6E ) . In contrast with the orientation selective activation outside the FFF observed in Fig 5B ( bottom ) , which does not necessarily reflect the local orientation from the preference map , the activation outside the FFF in Fig 6E reflects the preference map ( white and yellow contours agree ) as it is generated by orientation-biased long-range connections . The increased selective activation outside of the FFF coincides with a reduced slope ( smaller exponent n ) in the radial decay of selective activation shown in Fig 6F–6I . Again , the results shown here for βrec in the range [0 , 0 . 5] are consistent with imaging studies [23 , 24] , as the region of selective activation is predominantly confined to the FFF ( limited small selective regions outside the red circle ) , the preference agrees with the underlying map ( yellow and white contours are very similar ) and the exponent n is larger for the selective activation . The measures of the orientation selective component of the lateral spread of activation as studied in individual simulations in Figs 5 and 6 are now quantified over a range of RWex and βrec . The ranges of these parameters consistent with anatomical data is indicated in Figs 7–9 ( between the white iso-κ contours reproduced from Fig 3 ) . The agreement of model simulations with qualitative features from imaging data will be assessed in this and the following section . Fig 7A shows a map of normalized area ( relative to the FFF area ) of the selective activation , where each value is an average from simulations at 5 randomly selected map locations . Within the red contour ( darker regions ) the selective area is roughly equal to or smaller than the FFF of the stimulated region ( e . g . Fig 7E , top ) . Outside this region the selective region is larger than the FFF ( e . g . Fig 7D , top ) . Fig 7B shows a greyscale map of the proportion of the selective area with the correct orientation with respect to the underlying orientation preference map . Within the white contour ( darker regions ) the majority of the selective region has the correct orientation . For small RWex or small βrec a significant proportion ( more than 15% ) of the selective activation is spurious ( not in agreement with the preference map ) . Fig 7C shows a map of the ratio of the slopes ( Naka-Rushton exponent ) of the selective activation nSel and the general activation nAct . When this ratio nSel/nAct is large there is steeper radial decay of the selective activation ( relative to the general activation ) and a sharp transition to non-selectivity at the border of the FFF . Note that nAct and nSel are independent of the choice of ηSel and ηAct defined in ( 28 ) . In each case ( Fig 7A–7C ) there is a significant overlap between the regions with orientation selectivity confined to the FFF , with this selective activation having the correct orientation , with a characteristically steep decay of this region , and the anatomical range for the lateral connections in the model . In the green triangular region ( overlayed in Fig 7A–7C ) , all of these characteristics are satisfied . We highlight that the operating range matching anatomy and functional data covers more than 20% of the ( conservative ) permissible range of RWex ( 0 . 1 < RWex < 0 . 55 = RWin ) . The extent of the operating region for the model depends on values of the thresholds used in Fig 7A–7C . A value of 1 . 05 for normalized selective area is consistent with observations from [24] , that the selective area is close to or marginally larger than the FFF area . A value of 85% for the proportion correct orientation was chosen heuristically , for smaller values isolated spurious selectivity occurs outside the FFF , e . g . compare Fig 7E and 7F . The value nSel/nAct = 1 . 3 , also chosen heuristically , gives a sharp drop-off in the selectivity close to the FFF border . Reducing the threshold on the selective area , or the threshold on nSel/nAct , by 10% would give a reduced by still existing operating region for the model , but increasing the proportion correct orientation threshold much beyond 85% rapidly reduces the operating region . Fig 8 illustrates the dependence of the operating region limits on each of the thresholds . The specific choice ηSel = 0 . 5 could also affect the size of the model’s operating region for the normalised selective area . For example , increasing ηSel could potentially increase the operating region for a fixed value of the threshold on the normalized selective area . However , the constraint on the ratio nSel/nAct ( also providing an upper bound on the green region ) would be unaffected as it is independent of the choice of ηsel . Although an anatomical bias towards iso-orientation in the model’s connectivity profile was introduced through βrec , the bias is not necessarily evident at the population functional level ( in the patterns of activity observed in simulations ) . This may originate from the non-linearities in the functional expression that combines excitatory and inhibitory activation . To further explore the role of the relative balance between E/I in this functional expression , we manipulate the strength of inhibition in the model ( controlled by C in ( 16 ) ) . Recall that when C = 0 there is global E/I balance and for the reduced inhibition case here we reduce C toward 0 ( from its standard value C = −0 . 4 to C = −0 . 2 ) . Fig 9 ( computed in the same way as Fig 7A ) shows that with reduced inhibition , the region of the ( RWex , βrec ) -plane with orientation selective activation confined to the FFF footprint ( inside the red contour ) is smaller . In general , maintaining the same values of other parameters , but reducing inhibition leads to a wider spread of orientation selective activation . Compare Fig 9B ( bottom ) where the stable pattern of activation extends outside the FFF with Fig 7B ( bottom ) where the activation is confined within the FFF . Furthermore , with strong orientation bias in the model’s lateral connections ( large βrec ) , the spread of activation can destabilize and continue indefinitely , see Fig 9B ( top ) . This activation outside the stimulated region has spurious orientation preference and is due to destabilized activity that spreads far beyond the boundary of the stimulated region . Note that this a distinct mechanism from the convergent excitation that generates activation with incorrect orientation when RWex is too small ( see Fig 5B ) . For the case illustrated in Fig 9B ( top ) , of the four stimulus presentations with I0 , I45 , I90 and I135 , the condition I0 resulted in an continual spread of activation down and to the right of the FFF ( blue-green regions ) , I45 above the FFF ( purple regions ) , whilst I90 and I135 remained constrained to the FFF . These different responses across different orientations arise from a local imbalance in the preference map , which becomes exaggerated when input drives the model close to spatial instability , as is the case with reduced inhibition . With stronger inhibition , small imbalances in the spread of activity can occur across different orientations , but these don’t impact the validity of tuning of the response ( e . g . see the slightly larger response to 135° in Fig 4B ) . We note that the destabilization is not contingent on there being local imbalance in the preference map , in general , moving in parameter space to the top left of Fig 7A , the unbounded activation would occur for all input orientations . In general , with reduced inhibition , there is no more overlap in the ( RWex , βrec ) -plane between the anatomical operating range of the model ( between the white contours ) and the range where orientation selective activation is confined to the FFF . With increased inhibition strength ( C = −0 . 6 ) the operating region for the model increases slightly shifting to lower RWex and extending to larger βrec values ( Fig 9C ) . This illustrates how inhibition constrains the spread of activation to the FFF , even with larger orientation bias of lateral connections . Fixing RWex = 0 . 25 , the operating region in terms of C and βrec is illustrated in Fig 9D , with the largest extent in βrec occurring at C = −0 . 4 . For large values of C the model responses become predominantly feed-forward ( input driven ) and the profile of decay for the general and selective activation becomes similar ( nSel/nAct tends to 1 ) . As discussed above , without sufficient balance from inhibition activity can spread unbounded across cortex ( top left region of Fig 9D ) .
Our modelling results predict a qualitative change in the spread of orientation selective activation for localized , oriented inputs if global inhibition strength is reduced . Under those conditions , we should observe selective activation outside of the retinotopic footprint , however , the exact orientation preference may not be preserved . Activation with spurious orientation preference is generated via spatial destablization of the localized activation generated by the input . These predictions could be addressed by manipulating the inhibitory cells pharmacologically [66–68] or optogenetically [27] . To confirm our model’s prediction though , one may have to go close to pathological epileptic conditions [69 , 70] . Other approaches could be used by simply comparing anesthetized and awake conditions . Recently , [71] have indeed shown this to be a valuable approach for investigating E/I balance in the integrative properties of the cortical populations . Comparison of the dynamics of propagation of orientation-selective activity in awake or anesthetized conditions could hence provide the appropriate non-pathological test to probe our model’s prediction . Spurious activation can also be generated by another mechanism . Activated regions inside the stimulated region can generate excitation at an equidistant range outside the stimulated region . The specific range is associated with the peak in the radial excitatory profile . This happens in the model when peaks in excitation are highly specific , in a parameter range that was ruled out from the operating regime of the model . This could be seen as an undesired consequence of the model’s design . However , the fact that , under particular circumstances , the preferred orientation of the horizontal propagation may be at odds with the underlying orientation preference map could unravel some new unexpected computational capacities of the horizontal network , which may be present in visual areas beyond V1/area 17 . For instance , the ability to link information of position and orientation for non co-circular filters that could be of importance for processing objects with sharp angles . Inline with this hypothesis , [24] showed that the spread of orientation selective activity is not fixed but can increase when increasing spatial summation . The transition to an unbounded spread across the network , as contingent on a spatial modulation parameter ( like βrec controlling orientation bias in our model ) , has been observed in a one dimensional theoretical study of the neural field equation with purely excitatory connectivity [72] . A more common choice of connectivity function ( e . g . difference of Gaussians ) features a broader footprint for inhibition than for excitation [36 , 37] . In our model excitation extends further than inhibition with additional peaks in excitation away from the origin [44] . The distance between excitation peaks fixes a regular length scale that stabilizes multi-bump patterns . In [41] , without an explicit representation of orientation , localized inputs were shown to produce multiple bumps within a stimulated region . It was proposed that the connectivity’s excitatory peak separation could be equated with Λ ( hypercolumn separation ) and the spatial phase of multi-bump patterns governed by an orientation preference map . In the present work , we have shown that the localized patterns of orientation selective activation observed in [23] and [24] are well described by a superposition of these intrinsic multi-bump states . Local models of orientation selectivity [32 , 33] have been more widely studied than spatial models capturing interactions across columns . Spatial models of map development [55 , 73 , 46 , 74 , 57] do not focus on dynamics on sub-second timescales . Indeed , [74] does not consider dynamics or long-range connections within V1; see also commentaries in [75 , 76] . The self-organizing map ( SOM ) model in [57] included , but did not show , the dynamics of stimuli responses . The focus of the study was on the long-timescale dynamics of map development , however , this class of model could be another candidate to investigate functional activation in the future . Integrate-and-fire neuron models of earlier thalamic and cortical processing stages proved successful for capturing the dynamics of orientation selectivity and tuning , but were restricted to a small patch of cortex without considering superficial layers accessible to imaging [77 , 78] . [79] explored the role of patchy long-range connections in cortex in a general setting , whilst [80] looked at larger regions of cortex in an integrate-and-fire network , reporting complex spatio-temporal dynamics , but did not model orientation . The neural field framework used here is an ideal level of description for comparison with VSDI recordings [38] that allowed us to simulate a large spatial domain extending beyond the local region of interest , thus avoiding issues with boundary effects . The four sub-population implementation used here was chosen because it allows for efficient computation of connectivity integrals using convolutions . In turn , this allowed for a comprehensive multi-parameter investigation of the model’s dynamics . Another choice would be to consider a single mixed population , where connection weights depend directly on orientation difference as has been considered in other spatial models [73 , 81] , including those posed in an elastic net framework [82 , 83] . This would have the advantage of allowing direct implementation of the heterogeneity of long-range connections , however , the increased computational burden may not allow such a broad parameter investigation . [27] used optogenetic stimulation in combination with visual inputs and observed a non-orientation specific linear addition of the two inputs , which could be explored in our modelling framework . The model could also be used to investigate selective recruitment and spatial summation in regions between localized oriented stimuli . Two further properties were described in [24] that can be further investigated . First , as demonstrated with intracellular recordings , the lack of orientation selectivity observed at the mesoscopic level ( VSDI ) was due to a diversity of microscopic rules: some cells received untuned presynaptic input , but others a tuned presynaptic input with preferred orientation either agreeing with or different from the recorded cell . Such diversity in local cellular rules is to be linked to our observations that , with narrow excitatory peak widths or strong orientation bias ( Fig 7B ) or reduced inhibition ( Fig 8 ) , our model can easily produce spurious selective propagation . Adding diversity in the connectivity rules can thus easily lead to natural diversity in the tuning of the propagation . Second , increasing spatial summation increases the slope of selectivity decay at the stimulus boundary , whilst selective propagation reaches further across cortex; the model could also account for this property . More generally , the model could be used to make predictions to decipher the selective functional connectivity rules that link position and orientation in cortical space . For example , the model could be extended to differentiate inhibitory cell sub-classes as reported in [16] . As such it could generate functional predictions on e . g . the role of long-range basket cell connections that preferentially target cross orientations . A lumped description of inhibition and excitation was used here with local cross-orientation inhibition . By separating out inhibitory and excitatory populations , one could consider different profiles for cross-orientation interactions including their spatial profile . This may change the model’s behavior and would be important if it is extended to consider inputs with multiple orientations that overlap spatially . Earlier stages of cortical and thalamic processing could be incorporated with differing properties of orientation tuning and bias of connections across V1 layers [84] . The four sub-population implementation used here can be viewed as a coarse discretization from a continuous representation of orientation , which could be considered in future work . For example , a recent paper studied spatio-temporal patterns with continuous orientation , but only on a 1D spatial domain [85] . Theoretical work characterising localized states in 2D space plus orientation would be an important first step . An extended feature space including spatial frequency ( SF ) could be used to investigate lateral connections in light of recent work showing interesting interactions between orientation and SF maps [86] . Our model of orientation selectivity in V1 is the first of its kind , capturing the spatio-temporal dynamical spread of localized patterns of activation with a representation of orientation . Our study addressed an apparent conflict between the orientation bias of lateral excitatory connections in L2/3 of V1 , as characterized in anatomical studies , and imaging studies on the lateral propagation of cortical activity for localized oriented visual stimuli . Simulations with the neural field model illustrated that observed levels of orientation bias in anatomical studies actually predict long-range activation outside of the retinotopic footprint of the stimulus , but with a sharply decaying profile of orientation selectivity , as observed in imaging studies . Without this sharp decay , which might occur with excessive orientation bias or diminished inhibition strength , the network could destabilize leading to unbounded spread of cortical activation . | Optical imaging techniques can reveal the dynamical patterns of cortical activation that encode low-level visual features like position and orientation , which are shaped by both feed-forward projections , recurrent and long-range intra-cortical connections . Anatomical studies have characterized intra-cortical connections , however , it is non-trivial to predict from this data how evoked activity might spread across cortex . Indeed , there remains an apparent conflict between the reported orientation bias of cortical connections , and imaging studies on the propagation of cortical activity . Our study reconciles structure ( anatomy ) with function ( evoked activity ) using a dynamic neural field model that predicts the dynamics of cortical activation in a setting both inspired by and parametrically matched to the available anatomical data . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"nervous",
"system",
"membrane",
"potential",
"brain",
"electrophysiology",
"neuroscience",
"simulation",
"and",
"modeling",
"convolution",
"neuroimaging",
"research",
"and",
"analysis",
"methods",
"sensory",
"physiology",
"imaging... | 2017 | Neural field model to reconcile structure with function in primary visual cortex |
Members of the family Trypanosomatidae infect many organisms , including animals , plants and humans . Plant-infecting trypanosomes are grouped under the single genus Phytomonas , failing to reflect the wide biological and pathological diversity of these protists . While some Phytomonas spp . multiply in the latex of plants , or in fruit or seeds without apparent pathogenicity , others colonize the phloem sap and afflict plants of substantial economic value , including the coffee tree , coconut and oil palms . Plant trypanosomes have not been studied extensively at the genome level , a major gap in understanding and controlling pathogenesis . We describe the genome sequences of two plant trypanosomatids , one pathogenic isolate from a Guianan coconut and one non-symptomatic isolate from Euphorbia collected in France . Although these parasites have extremely distinct pathogenic impacts , very few genes are unique to either , with the vast majority of genes shared by both isolates . Significantly , both Phytomonas spp . genomes consist essentially of single copy genes for the bulk of their metabolic enzymes , whereas other trypanosomatids e . g . Leishmania and Trypanosoma possess multiple paralogous genes or families . Indeed , comparison with other trypanosomatid genomes revealed a highly streamlined genome , encoding for a minimized metabolic system while conserving the major pathways , and with retention of a full complement of endomembrane organelles , but with no evidence for functional complexity . Identification of the metabolic genes of Phytomonas provides opportunities for establishing in vitro culturing of these fastidious parasites and new tools for the control of agricultural plant disease .
Flagellated protists of the family Trypanosomatidae , class Kinetoplastea , infect a large variety of organisms including animals , plants and humans [1] . While African and South-American trypanosomes are responsible for sleeping sickness [2] and Chagas' disease [3] , respectively , different Leishmania spp . cause visceral , cutaneous and mucocutaneous manifestations of leishmaniasis in many tropical and subtropical regions [4] . Various eukaryotes , particularly filamentous microorganisms like oomycetes and fungi have acquired the capacity to infect and grow inside the plant tissues . While some of these organisms could influence plant growth positively , in most cases they can cause major diseases in plants of economic importance [5] . The genomes of numerous of these filamentous plant pathogens have already been sequenced , unveiling an amazing variety of genome sizes and organization [6] . Certainly , a great number of these plant pathogens were molded into larger genomes by repeat-driven expansions , with the genes coding for proteins involved in host interactions located within repeat-rich regions [6] . In contrast , some filamentous plant pathogens have fairly small genomes , as a consequence of intron or gene loss , like U . maydis; ( 21 Mb ) [7] and Albugo laibachii; ( 37 Mb ) [8] , or abridged transposon content as in Sclerotinia sclerotiorum; ( 38 Mb ) [9] . Like fungi and oomycetes , trypanosomatids also infect plants , but using a radically different strategy to colonize and propagate inside the host [10]–[12] . Multiple insect species of the order Heteroptera act as the natural vectors of plant trypanosomatids , both in the transmission to lactiferous hosts [10] , [13] , and for infection by intraphloemic plant trypanosomes [14]–[16] . Phytomonas is the arbitrary genus name proposed for all trypanosomatids specific to plants [17]; however this rather restricted taxonomic description fails to fully capture the wide diversity of trypanosomatids encountered in plants , both with respect to their biological properties and their impact on the host [18]–[22] . Indeed , Phytomonas spp . infect more than 100 plant species , distributed primarily in tropical and subtropical zones , by multiplying in latex tubes , fruits and seeds or colonizing the phloem sap inside the sieve tubes . Phytomonas infection can occur without apparent pathogenicity , but conversely it can cause lethal disease in plants of substantial economic value , including the coffee tree , coconut and oil palms [10] , [23] . This results in important economic losses in Latin America and the Caribbean [24]–[27] . Ten distinct subgroups of plant trypanosomatids have been defined using the internal transcribed spacer region of the ribosomal RNA locus [22] . Only Group H , encompassing the Latin American intraphloemic trypanosomatids responsible for severe wilts , can be distinguished both by rRNA markers as well as biological and serological properties [19] . A full definition of the diversity of trypanosomatids within the overarching Phytomonas genus is still outstanding . The whole genome sequences of Trypanosoma cruzi , Trypanosoma brucei and Leishmania major were released in 2005 [28]–[31] . Since then , the genomes of several additional trypanosomatids , including several pathogens of mammals , have been completed and described [32]–[35] . These databases have provided an essential platform for investigations of basic biology and mechanisms of pathogenesis and facilitated the exploration of novel therapies . However , to date genome level analysis of Phytomonas spp . is limited . The biology of these parasites is reasonably well described [11] , [36] , but little information exists on their effective control by chemicals or , most critically , on their specific adaptations to the plant host and the mechanisms underpinning pathogenesis . Moreover , few genes are available in sequence databases , and little is known about genome size , chromosomal organization and ploidy [37] , [38] . We describe here the genome sequences of two plant trypanosomatids , one phloem-restricted pathogenic isolate from a diseased coconut from Guiana ( HART1 from Group H ) and the other a non-symptomatic latex isolate from Euphorbia ( EM1 from Group D ) [22] . The comparison of these two plant parasite genomes with each other and with those of other trypanosomes reveals a common simplified genome organization for the plant trypanosomes . Identification of the genes involved in Phytomonas metabolism is an important step for improving in vitro culture protocols and for development of new and better tools for the control and diagnosis of Phytomonas-mediated diseases .
Recently , the molecular karyotype of several different latex plant symbiont-like ( i . e . not associated with apparent pathology in the host ) Phytomonas isolates were analyzed by pulsed-field gel electrophoresis ( PFGE ) , showing 21 chromosomal bands for EM1 ( group D ) [37] . Similar analysis performed on phloem-restricted trypanosomatids allowed the identification of 7 chromosomal bands for the Hartrot wilt pathogen isolate ( HART1 , Group H ) [38] . A systematic genome sequencing project of these two Phytomonas isolates was initiated , as they represent two distinct phenotypes in terms of impact on the host . Both EM1 and HART1 genomes were assembled using 10× 454-technology and 0 . 1× Sanger reads , together with deep coverage Illumina sequencing reads for correction of sequencing errors [39] ( European Nucleotide Archive accession numbers CAVQ010000001-CAVQ010001400 for EM1 and CAVR010000001-CAVR010002560 for HART1; details in Text S1 ) . Ninety percent of the EM1 genome assembly was placed in 45 scaffolds longer than 100 kb , with one third in the size range of the Phytomonas EM1 chromosomes previously observed by PFGE [37] . In the case of EM1 , the scaffold N50 ( the scaffold size above which 50% of the total length of the sequence assembly can be found ) was 429 kb ( Figure 1A ) . Meanwhile , the scaffold N50 for HART1 isolate was 1 . 2 Mb , with 90% of the genome located in 15 of the scaffolds , again in the size range previously estimated for the HART1 chromosomes [38] ( Figure 1A ) . These assembly statistics indicate that majority coverage of both EM1 and HART1 genomes was achieved . A striking feature of these two plant parasite genomes is their small size ( 18 . 1 Mb for HART1; 17 . 8 Mb for EM1 ) , when compared to that of the human pathogenic trypanosomatids ( 26 . 3 Mb for T . brucei; 32 . 5 Mb for T . cruzi and 32 . 9 Mb for L . major ) [31] . Phytomonas EM1 and HART1 are likely fundamentally diploid with some supernumerary chromosomes , with an unknown level of polymorphism between the two haplotypes [37] , [38] , a feature they have in common with other trypanosomatids . The massively parallel sequencing strategy provided important read depth coverage across both EM1 and HART1 assemblies ( Table S1 ) , which was used to establish ploidy for both Phytomonas genomes ( details in Text S1 ) . Median read depth analysis revealed an even depth across both Phytomonas assemblies ( Figure S1 ) , pointing towards an underlying euploidy of diploid for both Phytomonas isolates . The use of allele frequency for heterozygous single nucleotide polymorphisms ( SNPs ) across the scaffolds revealed a consistent diploid distribution of frequencies ( Figure S2; method as described by [40] ) . Clusters of duplicated genes were found to be biased towards disomic scaffolds using a Monte Carlo simulation ( p = 7×10−4 hypergeometric distribution ) . These results were similar to the distribution of multicopy genes observed in Leishmania spp . chromosomes [40] , suggessting the existence of separate mechanisms for gene duplication and chromosome ( scaffold ) duplication in Phytomonas spp . Nonetheless , read depth reached values greater than twofold in some cases ( scaffolds 24 and 25 in Figure S1 A; scaffolds 13 and 22 in Figure S1 B ) , probably indicating , as for other parasite genomes , aneusomy of certain chromosomal regions ( Figure S3 and S4 ) [40] , [41] . Possible aneusomy was already envisaged for the Phytomonas HART1 isolate after study of its molecular karyotype [38] . This increase in read depth is not likely due to the amplification of specific regions of the scaffolds , since read depth was constant along the whole of both the disomic and tetrasomic regions ( Figure S5 ) . Both assemblies were annotated using a combination of evidence ( Table S2; for details , see Text S1 ) , with the major features of the genome annotation presented in Figure 1 A . The reference annotation of the Phytomonas EM1 and HART1 genomes ( European Nucleotide Archive Accession HF955061–HF955198 for EM1 and HF955199–HF955282 for HART1 ) harbor 6 , 381 and 6 , 451 putative protein-coding genes , covering 57 . 9 and 53 . 7% of the genome respectively ( Figure 1 A ) . The total number of predicted genes in both Phytomonas isolates is lower than in other sequenced trypanosomatids ( EuPathDB-TriTryp 4 . 2: T . cruzi , CL Brener Esmeraldo-like 10 , 342; T . cruzi CL Brener Non-Esmeraldo-like 10 , 834; T . brucei TREU927 , 10 , 533; Figure 1 A ) , but slightly closer to the Leishmania spp . ( EuPathDB-TriTryp 4 . 2: L . braziliensis , 8 , 357; L . infantum , 8 , 241; L . major , 8 , 412 , Figure 1 A ) , as expected by the close phylogenetic relationship of Phytomonas with Leishmania [42] . Such a decrease in predicted gene numbers is the consequence of an almost complete absence of tandemly-linked duplicated genes in both Phytomonas genomes as observed when compared to other sequenced trypanosomes [43] , [44] . Indeed , the genomes of T . brucei , T . cruzi and L . major contain a high percentage of repetitive genes ( Figure 1A; 27% for T . cruzi , 9 . 6% for T . brucei and 6 . 7% for L . major ) , whereas both Phytomonas isolates only possess a very low percentage of such genes ( Figure 1 A; EM1 and HART1 ) . This is the case for the NADH-dependent fumarate reductase , arranged in several copies in the T . brucei ( 6 copies ) , T . cruzi ( 7 copies ) and L . major genomes ( 4 copies ) but only detected as a single-copy gene in both Phytomonas isolates ( Table S3 ) . The uniform read depth coverage observed all along the Phytomonas EM1 and HART1 scaffolds overrules a collapse of multiple tandem repeats into fewer copies during assembly as an explanation for the Phytomonas gene copy number observed ( Figure S5 ) . A small fraction of EM1 genes were observed in multiple copies on the genome: only 99 clusters of paralogous protein-coding genes ( corresponding to 171 genes; for details see Methods ) were identified , constituting 2 . 6% of the Phytomonas EM1 putative genes . Typical cases are those of the chaperonin HSP60 ( 32 copies ( on average ) in the T . cruzi CL Brener genome ) and the thioredoxin peroxidase , both identified in three copies in the EM1 assembly . Excluding a multigene family ( six genes ) with a histone-fold domain , most of the “duplicated” genes were present in only two copies . A similar situation in which the genome was almost exclusively comprised of single-copy genes was observed in HART1 , with the exception of a gene family homologous to a major surface metallopeptidase of Leishmania promastigotes [45] . The metalloprotease gp63/leishmanolysin ( EC 3 . 4 . 24 . 36 ) was originally described as the most abundant surface protein of Leishmania spp , but has been subsequently demonstrated to be pan-eukaryotic . A massive expansion in the gp63 family is evident in HART1 with over 20 members , while EM1 has only two . Both expansions are lineage-specific . GP-63 has been implicated in interactions with both vertebrate and insect hosts of Leishmania , and there is preliminary evidence for it playing a role in insect interactions in P . serpens and other lower trypanosomatids [46] , [47] . In P . serpens gp63 is present in many endomembrane compartments; significantly expression levels can be reduced by exposure to fetal calf serum , suggesting an ability to respond to alterations in the environment , and/or potential for degradation of specific proteins or peptides [48] . Unlike the majority of eukaryotes , mRNA transcription in trypanosomatids is polycistronic . These genomes are organized into large polycistronic transcription units ( PTUs ) , with tens –to -hundreds of protein-coding genes arranged head-to-tail on the same DNA strand and apparently transcribed from a single upstream RNA pol II entry site , or promoter [28]–[30] , [49] . This unusual gene organization was observed in both Phytomonas isolates as well , where genes are organized into 298 ( EM1 ) and 334 ( HART1 ) putative PTUs with an average of 21 ( EM1 ) and 19 ( HART1 ) genes per cistron ( Figure S6 ) . Protein-coding genes in Phytomonas appear to lack conventional introns , similar to the structure of genes in other trypanosomatids [1] , [50] . Classical cis-splicing introns are documented only in the poly ( A ) polymerase and an ATP-dependent DEAD/H RNA helicase genes from T . brucei , T . cruzi [51] , and Leishmania spp . This striking feature is not conserved in the Phytomonas EM1 and HART1 isolates . Contraction in both plant parasite genomes is also reflected by the short length of the intergenic regions ( on average 1 , 140 bp for EM1; 1 , 280 for HART1 ) and a relatively low frequency of repeated sequences ( 0 . 9% and 1 . 2% for EM1 and HART1 , respectively ) ( Figure 1A ) . No significant difference in overall gene sizes was observed between these isolates ( 1 , 614 bp and 1 , 507 bp on average for EM1 and HART1 , respectively ) . These data suggest that the EM1 and HART1 genomes are compact and might lack many of the expansions of both coding and non-coding sequences that have been described for other trypanosomes [30] , [43] . Members of the order Kinetoplastida display an impressive number of structural and biochemical peculiarities . The acquisition of foreign genes through lateral gene transfer is a possible explanation of the trypanosome-specific evolution of novel processes and organization [52] . A systematical search for candidate bacterial horizontal gene transfer ( HGT ) events ( Material and Methods ) allowed us to identify 87 HGT candidates in these Phytomonas isolates , all shared between the two isolates , with eight of them specific to Phytomonas ( i . e . absent from Leishmania and Trypanosoma ) ( Table S4 ) . Several genes of bacterial HGT origin already identified in Leishmania were also found in Phytomonas , specifically sugar kinases and other genes involved in carbohydrate metabolism , which probably reflects their life cycle in plants and phytophagous insects [52] , [53] . All HGT events were common to EM1 and HART1 , but a metallocarboxypeptidase of potential bacterial origin was found in only one copy in EM1 and 11 copies in HART1 . In other trypanosomatids , the tRNA genes tend to occur in clusters with a synteny often conserved among different genera ( Figure S7; details in Text S1 ) . Most of the tRNA genes predicted for EM1 and HART1 corresponded to those identified previously in T . brucei , L . major and T . cruzi ( Table S5 ) . Interestingly , Phytomonas isolates possess two tRNAs not found among the animal pathogens , and present in the plant trypanosome branch: they are Asn ( ATT ) -tRNA ( in HART1 ) and Ser ( GGA ) -tRNA ( in EM1 ) ( Table S5 , highlighted in green ) . In all Trypanosomatidae the mitochondrial genome consists of a single network of kinetoplast ( k ) DNA , one of the most complex organellar genomes known . It is composed of dozens of maxicircles that carry protein-coding and mitoribosomal genes , and thousands of minicircles that encode guide ( g ) RNAs . The EM1 maxicircle could not be assembled , but a single maxicircle contig of 12 , 099 bp was recovered for HART1 . A homologous 10 , 478-bp region was sequenced previously for Phytomonas serpens [54] , and the identity over the matching region of 9 , 816 bp between the two Phytomonas isolates is 76 . 8% . Similar to the P . serpens maxicircle , the maxicircle of HART1 is characterized by a complete absence of cytochrome c oxidase subunits I–III ( COI , COII , COIII ) , and cytochrome b ( Cyb ) of the bc1 complex . Other maxicircle-encoded genes typical for trypanosomatids , 12S and 9S rRNAs , ND1 to ND5 , ND7 to ND9 , subunit 6 of ATP synthase ( A6 ) , ribosomal protein subunit 12 ( RPS12 ) , maxicircle unknown reading frames ( MURF ) 2 and 5 , and unidentified cryptogenes G3 and G4 , are present ( Figure S8 ) . Since PCR and limited sequencing data indicated that the same deletions are present in EM1 and in three P . serpens strains [54] , these deletions likely became established at the base of the Phytomonas clade . Some maxicircle-encoded transcripts are known to undergo extensive RNA editing via the insertion and/or deletion of four to hundreds of uridylate residues [55] . Information for the editing process is provided by hundreds of heterogeneous minicircle-encoded gRNAs . The extent of editing is reflected by the sequence identities of individual maxicircle-encoded genes . Although we lack RNA sequence data for HART1 , DNA sequence alignments with other kinetoplastids allow determination of the extent of editing for a given gene ( Table S6 ) . Genes that are pan-edited in almost all trypanosomatids studied [56] ( ND3 , ND8 , ND9 , RPS12 , G3 , and G4 ) show no reduction of the edited region in HART1 as compared to P . serpens ( Figure S8 ) . When all maxicircle-encoded genes are considered , HART1 and P . serpens are more divergent from each other than L . donovani is from . L . tarentolae , but less so than T . cruzi is from T . brucei . Furthermore , the HART1 maxicircle genes have slightly lower identity to L . tarentolae , T . brucei and T . cruzi genes , than the genes from these species have among themselves ( Table S6 ) . These facts reflect the relatively long branch of the Phytomonas clade observed in the SSU rRNA- and glycosomal GAPDH-based phylogenies and deep separation between individual branches of this clade [57] , [58] . Recovered full-length kDNA minicircles differ between both Phytomonas EM1 and HART1 . In HART1 the minicircles range in length from 1 , 626 to 1 , 652 bp and contain one conserved region , as does P . serpens [59] . The EM1 minicircles are longer ( 2 , 791 to 2 , 819 bp ) , and carry two conserved sequences opposite each other . These variations are not unprecedented , as the size of minicircles as well as the number of conserved regions are typically uniform within a species , but variable among species [60] , [61] . Extensive bioinformatics analyses have been performed for all known transposable elements ( TEs ) present in the trypanosomatid genomes . While both LTR-retrotransposons ( also called retrotransposons ) and non-LTR retrotransposons ( also called retroposons ) were described in the genome of T . brucei , T . congolense , T . vivax , T . cruzi , and Leishmania spp . ( ∼3% of nuclear genome ) , no transposons have been identified to date [31] , [32] , [62]–[67] . Significantly , there is evidence for involvement of non-autonomous TEs in the regulation of gene expression [65] . Leishmania spp . ( ∼2 , 000 copies per haploid genome ) , but not trypanosomes , have domesticated and expanded these small TEs , named SIDER ( Short Interspersed DEgenerated Retroposon ) and co-opted them as part of the gene expression machinery . All trypanosome species analysed so far contain at least one putative functional TE family of the ingi clade ( Tbingi , Tvingi , Tcoingi , L1Tco , L1Tc ) that may have the capability to be mobilized , but all members of the ingi clade are degenerate and non-functional in the Leishmania species sequenced to date . Two questions were considered important to address in the analysis of TEs in these Phytomonas isolates due to their relatively close phylogenetic position to Leishmania spp . : when , in the course of trypanosomatid evolution , did domestication and expansion of SIDER occur ? and when was the loss of TE functionality from the ingi clade ? As observed for Leishmania spp . , both Phytomonas genomes are missing potentially active ingi-like TEs , but contain a few non-functional TEs of the retroposon ingi clade . Two types of TEs belonging to the retroposon ingi clade ( PhDIRE , for Phytomonas Degenerated Ingi-Related Element , and PhSIDER , Table S7 ) were identified , with no evidence of functional elements , since all are likely to be inactivated by the accumulation of deletions , point mutations and/or frame shifts . PhDIRE belongs to the ingi1 subclade , considered as an early diverging ingi subfamily also present in Leishmania spp . , T . cruzi and T . congolense [66] , as shown by phylogenetic reconstruction ( Figure 2 ) and analysis of the conserved motif upstream of the retroposons . PhSIDERs are short elements that were probably derived from PhDIRE by deletion , as previously proposed for other potentially active ingi-like TEs [62] , [65] , [66] , [68] ( see Text S1 for details ) . No sequences related to other trypanosomatid TEs were detected in the Phytomonas genomes ( details in Text S1 ) . The EM1 genome was found to contain 41 DIREs , similar to all other trypanosomes and Leishmania spp . ( L . major: 52 and L . braziliensis: 65 ) ( Table S7 ) , however the seven SIDER copies was low in comparison to Leishmania spp . that carry around 2000 copies . Thus , the enormous expansion and domestication of SIDER in Leishmania spp . [65] is not observed in these Phytomonas isolates , and exaptation of SIDER was likely a Leishmania-specific event in the trypanosomatid lineage . The HART1 genome is depleted of TEs . Forty-eight retroposons were identified in the EM1 genome , while two PhDIREs were found in the HART1 genome , a 24-fold difference ( Table S7 ) . Indeed , both the un-annotated contigs and the non-assembled reads showed very low coverage of PhDIRE/PhSIDER in HART1 , confirming the low number of retroposons in this Phytomonas isolate . The majority of Phytomonas genes are shared between both isolates , as shown by independent approaches used for ortholog detection ( see Materials and Methods ) . The combination of both Best Reciprocal Hits ( BRH ) and orthoMCL strategies identified 5 , 210 ( 82% ) genes from EM1 with orthologs in HART1 , and 5 , 108 ( 79% ) genes from HART1 with counterparts in EM1 , similar in gene size ( Figure S9 A ) . The Phytomonas EM1 and HART1 orthologs were more closely related to each other than to their trypanosome orthologs with an average percentage of identity of 70 . 5% ( Figure 3 ) . The small nucleolar RNA ( snoRNA ) repertoires of HART1 and EM1 also showed higher similarity to each other than to T . brucei or L . major ( Table S8 ) . The genes for which no orthologs could be detected by this preliminary approach are excellent candidates for understanding Phytomonas spp . behaviors . After eliminating genes for which orthologs were not detected because of annotation or assembly issues , as well as suspected annotation artifacts , 13 genes remained in EM1 and 4 in HART1 that could be confidently considered as lacking an ortholog in the other isolate ( Figure S10 and Table S9 , see Materials and Methods for details ) . The vast majority of Phytomonas genes are shared between both isolates , highlighting the high level of conservation of the gene repertoire between these two trypanosomatids . We analyzed synteny between EM1 and HART1 using dot-plots ( Figure 4 ) . Synteny was conserved between EM1 and HART1 , with most of the synteny breaks corresponding to scaffold boundaries in one of the two isolates . Only five bona fide synteny breaks with HART1 were found in the EM1 assembly , and 10 in the HART1 assembly . The syntenic blocks are large ( average of 60 genes , median of 35 genes ) and usually include several hypothetical PTUs ( average 20 ORFs , median of 10 ) ( Figure S11 ) . There is good conservation between PTUs in EM1 and HART1 , with at least one boundary in common between EM1 and HART1 for all PTUs ( Figure S12A and Figure S13 ) . Significantly , synteny breaks tend to correspond to the boundaries between putative PTUs ( Figure S12B ) , and intergenic distances are well conserved ( Figure S9B ) . To identify putative insertions in one isolate compared to the other , we searched for gene number differences between successive pairs of BRH in syntenic PTUs ( Materials and Methods ) . After filtering possible annotation artifacts ( genes missed , splits/fusions , etc ) and genes with strong sequence similarity elsewhere in the genome ( Table S10 ) , we retained ten genes in EM1 absent at the syntenic position in HART1 , including three already identified as lacking a HART1 ortholog . Furthermore , three genes in HART1 lack a syntenic equivalent in EM1 , with two already identified as having no ortholog in EM1 ( Table S9 ) . The two strategies did not identify the same sets of genes because of slight differences in the very conservative quality controls applied ( see Material and Methods ) . Significantly , ten and three genes in EM1 and HART1 respectively , displayed weak hits in the syntenic region , suggesting that they have diverged in the other isolate; ten and two genes had no evidence for sequence homology , and could thus correspond to insertions or complete deletions . Combining the two approaches , 20 genes from EM1 were confidently determined to be absent from HART1 and 5 genes from HART1 were found to be absent from EM1 ( Table S9 ) . Since we could only compare assembled and annotated genes with confidence , these numbers may be underestimates of the true number of non-conserved genes between both isolates , but they are representative of the overall level of synteny and gene repertoire conservation between these two phylogenetically remotely related Phytomonas isolates ( Table S9 ) . OrthoMCL comparisons [69] were performed between Phytomonas EM1 and HART1 , and four other trypanosomatids: L . major [29] , T . brucei [28] , T . cruzi [30] and Trypanosoma vivax [70] ( Materials and Methods ) . This predicted 22 , 706 clusters of orthologous genes . Their conservation profiles ( i . e . the list of species in which they are found ) are shown in Table 1 . A core of 2 , 869 genes was conserved between all six species ( Table 1 ) . Indeed , expert examination of this group of genes showed that 80 . 6% of the identified protein kinases shared by both Phytomonas isolates are also present in T . brucei and L . major . This subgroup contained major regulators , including up to 11 cdc2-related kinases ( CRKs ) , WEE1 , aurora kinase AUK1 , glycogen synthase kinase 3 ( GSK3 ) and casein kinases CK1 and CK2 , expected to be present in all eukaryotes ( Table S3C ) . Putative amino acid transporters conserved in all four mammalian parasites were also identified in these Phytomonas isolates . Interestingly , both isolates contained the same repertoire of amino acid transporters ( AAPs ) , but with differing copy numbers ( Table S3E; details in Text S1 ) . Several genes with similarity to calmodulin and genes annotated as calmodulin-like in T . cruzi [71] were also present in both Phytomonas genomes . Manual inspection of Phytomonas gene families highlighted many examples of gene conservation within these plant parasites . Four conserved Phytomonas EM1 and HART1 kinases were absent in both T . brucei and L . major: These Phytomonas-specific kinases were one calcium/calmodulin regulated kinase-like , one UNC-51-like kinase , and two unique kinases that do not fall into any defined kinase group ( Table S3C; details in Text S1 ) , suggesting that these enzymes could be important for infection of , or survival in , plants . Conservation of the phosphatase complements was also observed in these two isolates; only slight differences were detected between both tyrosine and serine/threonine-specific complements ( Table S11; details in Text S1 ) . The Phytomonas isolates have more genes in common with Leishmania than with the three Trypanosoma spp . : 317 orthoMCL clusters are shared between at least one Phytomonas isolate and L . major but none of the Trypanosoma spp . , and only 111 clusters are common to at least one Phytomonas isolate and one Trypanosoma spp . but not L . major . The number of BRH , as well as their percentage of identity , was also significantly higher between Phytomonas and Leishmania than between Phytomonas and trypanosomes ( Figure 3 ) . However , the presence of two types of clusters conserved only in Trypanosoma or Leishmania suggests independent secondary losses from an ancestral organism with a substantially larger gene complement . Significant synteny was observed between Phytomonas and Leishmania ( Figure S14 ) , as well as between Phytomonas and trypanosomes ( Figure S15 , Figure S16 and Figure S17 ) . As expected from the closer phylogenetic relationship of Phytomonas with Leishmania ( Figure 1B ) [37] , [38] , [42] , more syntenic breaks were observed between the Phytomonas isolates and trypanosomes than Leishmania ( Figure S12B ) . Syntenic blocks usually include several PTUs ( Figure S13 ) . We compared the number of synteny breaks that occur at PTU boundaries with what would be expected by chance ( Materials and Methods ) : for all pairs of species , the synteny breaks tended to coincide with PTU boundaries ( Figure S12B ) . The high synteny conservation between trypanosomatids might thus be the result of a selective pressure against intra-PTU rearrangements . The topological complexity of the kDNA network has fascinated replication specialists for decades . The process is not fully understood , but many of the players have been identified . In the model flagellate T . brucei , the machinery is extremely complex , requiring the combined activity of several mitochondrial DNA polymerases , ligases , endonucleases , helicases and topoisomerases [72] . Using a database of 26 genes encoding the kDNA replication machinery of T . brucei , all orthologs have been identified in the Phytomonas EM1 and HART1 isolates . The transcripts of many maxicircle genes undergo RNA editing in order to be translatable on mitochondrial ribosomes . Editing and processing of these mRNAs require the participation of several dozen proteins . A list of 28 T . brucei orthologs that are confirmed components of the RNA editing core complex or predicted to interact transiently with the complex [73] revealed that both EM1 and HART1 have the same composition , with substantial similarity to T . brucei . With the exception of KREP4 , KREP5 and the oligoU-binding protein that have likely been lost or divergent as in L . major , all of the remaining orthologs are present . In both Phytomonas isolates , KREPB7 is duplicated . The available data is compatible with the existence of another complex involved in RNA editing , mitochondrial RNA binding complex 1 ( MRB1 ) being composed of transiently interacting sub-complexes , with up to 32 components [74] . While only recently identified , MRB1 and associated proteins are conserved , as EM1 and HART1 contain all of its known orthologs . Trypanosomatid flagellates are well known for their uniquely complex kDNA and kRNA . All in all , the gene order , editing patterns , as well as proteins that participate in the metabolism of these organellar nucleic acids , mostly identified in model species T . brucei , L . tarentolae and/or C . fasciculata , are conserved in these Phytomonas isolates . Analysis of the Phytomonas genome sequences provided a global view of the metabolic potential of plant trypanosomatids . Comparison of the gene repertoires from both isolates to other sequenced trypanosomatids revealed a simplified genome , coding for a minimal system with a clear lack of complexity for each isolate . Indeed , both EM1 and HART1 genomes presented diminutive gene sets when compared to T . cruzi , T . brucei and L . major ( Table 2 , for more details see Table S3 ) , retaining only the most essential functions for the parasite , and often including a considerable fraction of genes that could serve the hosts . Furthermore , both gene repertoires are reduced as a result of both the loss of entire gene families and the reduction of the numbers of paralogs within gene families . The protein kinase contents of the Phytomonas isolates provide a good example of genome contraction in these plant parasites: eukaryotic protein kinase ( ePKs ) genes were identified in both isolates ( 160 and 161 in EM1 and HART1 , respectively ) , but in smaller numbers than in the TriTryp kinomes ( Table 2 ) [31] , [75] . Twenty four protein kinases , conserved in T . brucei and L . major , were not present in either of the Phytomonas draft kinomes . ( Table S3C ) . Furthermore , nine T . brucei-only kinases and 24 L . major-only kinases were also absent from both Phytomonas draft kinomes . Even though it is possible that fewer ePKs are required for infection of plants compared to mammals , the similar number of ePKs in the pathogenic isolate HART1 was somewhat unexpected , as it could be considered that additional protein kinases might be required to coordinate virulence factor expression . The less investigated partners of the phosphorylation-dephosphorylation regulatory cascades are the protein phosphatases , organized into four major groups , depending on substrate preferences and catalytic signature motifs . Three of these groups corresponds to Ser/Thr specific phosphatases ( STP ) : metallo-dependent protein phosphatases ( PPM ) , phosphoprotein phosphatases ( PPP ) and aspartate based phosphatases with a DxDxT/V motif . The fourth group corresponds to the protein tyrosine phosphatases ( PTP ) [76] . The completion of the genome sequences of L . major , T . brucei and T . cruzi [31] has permitted a deeper analysis of the protein phosphatases , showing that the main protein phosphatase groups ( Tyr , Ser/Thr and dual specific protein phosphatases ) are present in these parasite genomes , as in higher eukaryotes [77] . The Phytomonas phosphatome provides another illustration of the genome reduction observed in these parasites . Comparing the two plant trypanosomes' phosphatomes to the TriTryp phosphatome [78] , the main differences were found in the PTP complements: the eukaryotic-like PTPs were absent from both EM1 and HART1 phosphatomes , and no orthologs of PTENs and CDC14s [76] have been identified ( Table S11A ) . PTENs and CDC14s ( dual specific phosphatase group ) are present in the phosphatomes of all three other kinetoplastids , where they can be grouped into two distinct families , the eukaryotic-like and kinetoplastid-like PTENs , depending on their sequence homology to other eukaryotic PTENs . One kinetoplastid-like PTEN enzyme has been found in the three kinetoplastids T . cruzi , T . brucei and Leishmania [79] . While four eukaryotic-like PTENs have been identified in T . cruzi , only one enzyme was found in L . major . Interestingly , no T . brucei ortholog was identified , thus suggesting a possible role of these enzymes in intracellular parasitism . When we compared the STP complements of the Phytomonas isolates , we detected a 20% decrease in the total number of phosphatases as compared to the TriTryps , mainly due to the reduced number of type 1 protein phosphatases . The number of PP1s has been augmented in the genomes of T . brucei , T . cruzi and L . major by a gene duplication process ( 8/7/8 ) [78] . Still , the functions associated to these apparently higher number of resembling genes have not been characterized . Both in EM1 and HART1 , four genes encoding PP1 catalytic subunits were identified , a similar number to those described in other eukaryote PP1 complements . We have also found a two-fold reduction in the number of the bacterial-like phosphatases , Alphs and Shelps [80] in the plant trypanosomatids compared to the TriTryp phosphatomes ( Figure 5 , Table S11B ) . The reduction in the number of members of ABC transporters ( Table S12 ) and amino acid transporter families in these Phytomonas isolates represents another relevant example of genome retrenchment . A unique family of amino acid transporter ( AAP ) genes from members of the trypanosomatid family ( 25 in Leishmania , 17 in T . brucei and 19 in T . cruzi ) has been identified , based on the existence of amino acid permease pfam domains [81] , [82] . This trypanosomatid-specific group of amino acid transporters corresponds to a distinct clade within the amino acid/auxin permease ( AAAP ) super family [83] , [84] . The analysis of these gene families revealed 15 and 16 AAP genes in EM1 and HART1 respectively , fewer than in the mammalian trypanosomatid genomes ( Table 2 and Figure S18 , details in Table S3E and Text S1 ) . Eukaryotic cells regulate their cytosolic calcium concentration using numerous channels and transporters located in the mitochondria , the plasma membrane and the endoplasmic reticulum . Additionally , calcium binds to an extensive collection of signaling and regulatory proteins in these eukaryotic cells . In trypanosomatids , acidic organelles known as acidocalcisomes , which have been identified in Phytomonas françai [85] , act as the major stock of the intracellular calcium , and are implicated in processes such as calcium homeostasis , osmoregulation and polyphosphate metabolism [71] . Hence , both Phytomonas EM1/HART1 genomes were investigated for the presence of orthologs to trypanosomatid genes known to be involved in calcium and polyphosphate metabolism . The trypanosomatid genome projects revealed a vast diversity of Ca2+-binding proteins ( as an example for T . cruzi see Table S3A ) , many of which are not characterized and have little or no homology with non-kinetoplastid proteins . Regulation of cytosolic Ca2+ concentration in Phytomonas isolates EM1 and HART1 appears similar to that of other trypanosomatids . Yet , several differences allow to clearly distinguish these organisms ( Table S3A ) . Though the inositol phosphate/diacylglycerol pathway is present in pathogenic trypanosomatids , no evidence of either a phospholipase C , or a protein kinase C was found in these Phytomonas isolates . However , there are orthologs to the putative InsP3 receptor in both Phytomonas EM1 and HART1 isolates . Another interesting difference is the lack of Phytomonas counterparts to calreticulin , a Ca2+ storage protein located in the endoplasmic reticulum of T . cruzi [86] , and the recently characterized polyphosphate kinase ( vacuolar transporter chaperone 4 ) of yeast , pathogenic trypanosomatids , and Apicomplexan . To predict both the level of intracellular organellar complexity and the surface composition of Phytomonas , the open reading frame complement of HART1 and EM1 were scanned for around 300 genes involved in membrane trafficking . Both isolates of Phytomonas share essentially identical membrane transport systems , with only one clear example of specialization ( Table S3D and Figure S19 ) . Overall , the endomembrane systems are the simplest yet described amongst trypanosomatids; for example the Rab GTPase repertoire , a primary determinant of specificity and organelle identity [87] , retains the basic core exocytic and endocytic functions and the trypanosome-specific Rab-like X1 and X2 [88] ( Figure 6 ) . However , the system is substantially simpler , with only 12 Rab/Rab-like proteins compared to 16 in T . brucei or 17 in L . major [28] , [89] . Given that the losses here are Rab21 , 28 and 32 , this reduction represents sculpting of the system by secondary loss from the common ancestor and hence is an adaptive streamlining [90] . This simplification is also seen in the secondary loss of the AP4 adaptin sorting complex from both Phytomonas genomes ( Table S3D ) , and in a rather simpler ARF GTPase family compared with other trypanosomatids . Further , these data likely suggest a simplified late endocytic system , to which Rab21 , Rab28 and AP4 are all assigned . Overall the view is of a minimal endomembrane system , which conserves the major complexes and pathways , indicating retention of all major organelles , but with an apparent lack of complexity or innovation; adaptation has been via minimization rather than invention . As befits the position of Phytomonas as basal within the trypanosomatid lineage , the surface appears to be rather similar to Leishmania spp . , and there is no evidence for mucin-like or variant surface glycoprotein-related protein coding genes , or a dominant , highly expressed , surface antigen as no predicted GPI-anchored protein was encoded by transcripts in the most abundant RNAseq percentiles . The surface system includes full glycosylphosphatidylinositol ( GPI ) anchor and glycolipid biosynthetic pathways , the enzymatic apparatus for synthesis of a lipophosphoglycan ( LPG ) -like molecule and evidence for the GPI-anchored gp63 protein ( Table S3D , Figure S19 ) . Analysis of the genomes of these two plant trypanosomes provided a global view of the metabolic capacity of Phytomonas . As a consequence of an almost complete absence of tandemly-linked duplicated genes , most of the metabolic genes in Phytomonas were identified as one haploid copy ( Figure 7 , Figure 8 , Figure S20 and Figure S21; for details see Table S3B ) . As part of its carbohydrate metabolism ( Figure 7; details in Table S3B ) , Phytomonas not only utilize the plant's sucrose but also its polysaccharide stores as major energy substrates , as confirmed by the identification of genes coding for glucoamylase , alpha-glucosidase and , only in the HART1 isolate , many copies of invertase ( beta-fructofuranosidase ) homologs ( Table S3B ) . The presence of an alpha , alpha-trehalose phosphorylase in both isolates suggested that Phytomonas is also capable of using the abundant plant disaccharide trehalose for its carbohydrate needs . The presence of this bacterial-type enzyme illustrates that the adaptation of the plant parasite to their sojourn in their specific hosts may have been facilitated by HGT events . In agreement with previous studies on the carbohydrate metabolism of Phytomonas [91] , [92] , genome analysis revealed the presence of a complete set of glycolytic enzymes , the majority of which seem to be sequestered inside glycosomes , similar to other trypanosomatids . The existence of glycosomes in Phytomonas , previously demonstrated , was now confirmed by the presence of peroxisomal targeting signals at either the C- or N-termini of the encoded glycolytic enzymes as well as by the identification of a number of genes for peroxisome biogenesis proteins or so-called peroxins . Besides the horizontal alpha , alpha-trehalose phosphorylase transfer event described here , other HGT events were previously described for other Phytomonas isolates . A zinc-containing alcohol dehydrogenase from a trypanosomatid isolated from the lactiferous plant Euphorbia characias , previously identified as an isopropanol dehydrogenase of bacterial origin , was also acquired by an event of lateral gene transfer from a strictly aerobic bacterium to an ancestral trypanosomatid [93] . The addition of this gene could explain a selective advantage for a plant colonizing-flagellate living in the phloemic or lactiferous tubes of infected plants , supported by the fact that this enzyme was only identified in all plant trypanosomes analyzed thus far , while absent from the rest of the trypanosomatid family . This zinc-containing alcohol dehydrogenase , together with a glycosomal malate dehydrogenase ( Table S3 ) , allowed us to assume that EM1 and HART1 would be able to produce small amounts of lactate , as observed for other Phytomonas isolates [94] . Almost nothing is known about the amino acid metabolism in Phytomonas . Amino acid metabolism of Phytomonas resembles that of the other trypanosomatids . The so-called non-essential amino acids can either be degraded and utilized as energy sources , or be formed from other metabolites . However , Phytomonas lacks the capacity to oxidize aromatic amino acids and is predicted to require an external supply of most of the essential amino acids . The absence of a fatty acid beta-oxidation pathway and of ETF predicts that Phytomonas is unable to oxidize both long chain and side chain amino acids ( Results in Figure S21 , for details see Table S3 ) . An arginine kinase was detected as a single copy gene in both isolates . This enzyme may have been acquired by horizontal gene transfer from the arthropod vector during evolution , as previously shown for Phytomonas Jma [95] . The genomes revealed that overall the interconversion and breakdown of amino acids is very similar to what has been described for the other trypanosomatids . However , while amino acids serve as the most important source of energy for the other trypanosomatids inside their insect vector , this cannot be the case in Phytomonas because of its limited mitochondrial capabilities [91] . Owing to the fact that their insect vector ( s ) feed exclusively on plant juices that are rich in carbohydrates , the switch from plant to insect host would probably not require a metabolic switch from carbohydrate to amino acid metabolism as occurs in the mammalian trypanosomes . The absence of such a switch may have allowed the irreversible loss of a number of mitochondrial functions such as a respiratory chain required for beta oxidation of fatty acids and the complete oxidation of amino acids . Indeed , no genes coding for any of the mitochondrial cytochromes could be found . The enzymes of the hexose monophosphate pathway , as well as the ones involved in gluconeogenesis are present in Phytomonas , even though no evidence for the synthesis of glycogen has been detected . Few genes were found for the formation of storage polysaccharides . However , several mannosyl transferases , possibly involved in the synthesis of mannan polysaccharides , were detected , suggesting that mannans rather than glycogen may serve as a polysaccharide store . Protein glycosylation differs in the two Phytomonas isolates ( Figure S20 , Table S3B ) . The genes required for the incorporation of glucose , mannose , galactose , N-acetylglucosamine , glucuronic acid , xylose and fucose into glycoproteins , but not for sialic acid , were identified in the genome of the EM1 isolate . The HART1 isolate seems to lack the genes necessary for the incorporation of N-acetylglucosamine and fucose . With respect to lipid metabolism , fatty acyl dehydrogenase or , oxidase , multifunctional enzyme and thiolase were absent in both parasite isolates , indicating that Phytomonas is not capable of oxidizing any fatty acids via the beta oxidation pathway . On the other hand , Phytomonas should be capable of fatty acid biosynthesis , since the genes coding for the responsible enzymes have been identified in both parasite genomes ( Type II fatty acid synthesis in mitochondrion , and Type I fatty acid synthesis absent but synthesis taking place by a set of elongases ) ( Figure 8 ) . Oxidant stress protection in trypanosomatids is based on trypanothione , an adduct of one spermidine and two molecules of glutathione [96] . Thus the Phytomonas proteome was searched for the presence of enzymes involved in this metabolism . Phytomonas has a trypanothione reductase as well as a homolog of glutathionylspermidine synthase , or trypanothione synthase , as well as the enzymes thioredoxin ( tryparedoxin ) , several thioredoxin ( tryparedoxin ) peroxidases , peroxiredoxin , and trypanothione peroxidase . Several mitochondrial and cytosolic superoxide dismutases and an iron/ascorbate oxidoreductase , but no catalase , were identified . The reducing equivalents in the form of NADPH are provided by the enzymes NADP-dependent isocitrate dehydrogenase in the mitochondrion and by the hexose-monophosphate pathway enzymes glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase . A plant-like ascorbate peroxidase , as described for T . cruzi and Leishmania , was not detected ( Table S3B ) . Phytomonas lacks the capacity for RNAi , since the argonaut AGO1 ( Tb10 . 406 . 0020 ) and the two dicer proteins DCL1 ( Tb927 . 8 . 2370 ) and DCL2 ( Tb927 . 3 . 1230 ) present in both T . brucei and in L . brasiliensis but not in T . cruzi and L . major , two organisms that lack RNAi , were also absent in both EM1 and HART1 genomes ( see Table S3B ) . In fact , the lack of these gene products agrees with the presence of a double stranded RNA virus reported in the phloem-restricted isolates [97] that could serve as an indication for the absence of defense mechanisms against invasion by foreign RNA . Similar viruses have been reported for Leishmania spp . as well [98] , [99] . Virtually no information is available about the existence of effectors of pathogenicity in Phytomonas spp . and their possible role in the interaction with the host . We investigated the secretome of Phytomonas EM1 and HART1 isolates for potential virulence factors , by selecting those sequences having a secretion signal peptide , no transmembrane domains and no glycophosphotidylinositol ( GPI ) anchors . We detected 282 putative secreted proteins in both HART1 and EM1 ( Table S13 ) . Among these proteins , only 43 proteins in HART1 and 44 in EM1 had a PFAM domain annotation . The secretome was classified into molecular function and biological process using the Gene Ontology annotation ( Figure S22 ) . However , we noted the presence of numerous false positives in the set of predicted secreted proteins . This is due to the high divergence between the trypanosomatid sequences and the one used by SignalP for learning , mostly from fungi , animals , plants or bacteria origin . In the set of putative secreted proteins , we looked for proteins involved in plant carbohydrate degradation . One protein having a glycoside hydrolase family 31 domain was present in both HART1 and EM1 isolates , but the EST data did not show any expression of the two corresponding genes . We also found a secreted protein in HART1 ( GSHART1T00001406001 ) coding for glycosyl hydrolase family 32 that corresponded to one of the beta-fructofuranosidases ( see Metabolism of HART1 and EM1 section ) ; other beta-fructofuranosidases harbored a signal peptide and GPI anchor . We did not identify any secreted proteins that were supported by expression data and likely to be involved in plant cell wall degradation . This finding is consistent with the fact that Phytomonas is directly injected in the host phloem by an insect vector , thus it does not need to degrade the plant cell wall to penetrate into the host and gain access to the phloem sap . We screened for secreted proteins having a proteolytic activity that may lead to degradation of host proteins . Three genes were found in EM1 coding for an S24 serine peptidase , an M3A metallo-peptidase and an A1 aspartyl protease ( AP ) ; one AP was also found in the secretome of HART1 . Cathepsin D-like A1 family AP genes have not been found in other known trypanosomatid genomes such as Leishmania and Trypanosoma . However , APs are known to be secreted and involved in the virulence of several pathogenic fungi . In the case of the fungal animal pathogen , Candida albicans , ten APs that contribute to the dissemination of the pathogen in mice are present [100] . Fourteen APs are also present in the genome of the ascomycete plant pathogen Botritys cinerea , including BCap8 , which was found to constitute up to 23% of the total secreted proteins [101] . Since secreted Leishmania proteins with proteolytic activities may contribute to pathogenesis [102] , [103] , we looked for other AP coding genes in the HART1 and EM1 genome . EM1 did not have any extra APs , while HART1 harbors a cluster of five APs located in scaffold 1 ( Table S3 ) . These five tandem genes , absent in the syntenic region of EM1 , were not detected by the “synteny” approach because of the stringency of filtering ( see Material and Methods ) . The “true” first methionine of each protein of the cluster was located in intercontig gaps . When extending the N-terminal region of each of these proteins , a signal peptide could only be detected for the most extended gene ( GSHART1T00000177001 ) . For the four remaining APs , the N-terminal extension was not long enough to detect a probable signal peptide , and none of the five APs harbored a GPI anchor . The phylogenetic analysis ( Figure 9A ) revealed that these Phytomonas APs evolved from a common gene that branched deeply in the tree with high aLRT support ( aLRT support = 96 ) . This result suggested the existence of an ancestral AP gene in the trypanosomatid lineages that may have been lost in Leishmania and Trypanosoma . The integration of the APs genomic positions on scaffold 1 and the topology of the HART1 clade allowed the reconstruction of the events that led to the creation of a pathogenicity gene cluster in HART1 ( Figure 9B ) . HART1 and EM1 had initially one copy of the gene coding for a secreted AP . Then , the HART1 gene duplicated once from scaffold 5 to scaffold 1 . The cluster of five genes was created in the scaffold 1 of HART1 by four successive tandem duplications . The presence of a signal peptide in the AP from EM1 , the AP in scaffold 5 and one AP in the cluster of scaffold 1 let us speculate about the presence of a signal peptide in the other four APs , but their sequences were too short to detect it . The scaffold gaps between the five AP genes may correspond to repeated elements that may have mediated the AP tandem gene duplication . The EST data provided evidence for the expression of the five AP genes which comprise the AP cluster in the Phytomonas HART1 isolate ( Figure S23 ) , suggesting that , similarly to the function of the AP family in the fungi Candida and Bothrytis [101] , [104] , the Phytomonas HART1 AP gene cluster could be involved in virulence , an example of convergent evolution between distant organisms . The genus Phytomonas encompasses flagellates that differ substantially in their pathogenic potential . Despite most genes being shared between EM1 and HART1 isolates with respect to both gene count and content , several differences are still present ( Table S3 , Table S9 ) . Among the members of the Phytomonas EM1 and HART1 kinase repertoires , only two specific genes were identified: a CMGC/DYRK EM1-specific kinase , absent from the HART1 genome , and an AGC/RSK only present in the pathogenic HART1 isolate . Both specific kinases have no orthologs in T . brucei or L . major . Since the function of these kinases has not been studied in any trypanosomatid , and little is known about protein kinase signaling pathways in the TriTryps , the biological implications are not clear at present . While two CDC25 phosphatases were also identified in EM1 , no orthologs were found in HART1 , as in the case of T . brucei . The CDC25/CDD25-like phosphatases were identified in Leishmania spp . and T . cruzi , suggesting distinct roles for the protein phosphatases only present in the two intracellular trypanosomatids . Phytomonas is adapted specifically to infect and live in plants , where an abundant and diverse supply of carbohydrates is available for the parasite . Surprisingly , genomes of both EM1 and HART1 isolates contained only one sugar transporter , an ORF encoding a GT2 homolog ( Table S3E ) . The presence of only a single sugar transporter is intriguing . It would suggest that both EM1GT2 and HART1GT2 have a broader substrate specificity than the mammalian trypanosome GT2 , and would be in agreement with a much more simplified metabolic life cycle . The pathogenic HART1 isolate seems to be specialized in metabolizing sucrose , as is suggested by the presence of many copies of an invertase ( fructofuranosidase ) homolog only detected in this strain . The main difference between the pathogenic and asymptomatic Phytomonas isolates resides in their specific location inside the host: EM1 multiplies in latex tubes , while HART1 colonizes the phloem sap inside the sieve tubes . It is not yet clear whether this difference in habitat is related to the presence , or not , of multiple invertase genes . The presence of an alpha , alpha-trehalose phosphorylase in Phytomonas may be the explanation of why this plant parasite can survive in the insect hemolymph by using trehalose , a disaccharide of glucose , as an energy substrate , rather than amino acids , as is the case in midgut-dwelling trypanosomatids of hematophagous insects . Trehalose , originally regarded as a sugar characteristic of certain lower plants , is also a major blood sugar of insects [105] . Since the phosphorylase is present in all trypanosomatids for which the genome has been sequenced so far , it is unlikely that the enzyme would be involved in the parasite's energy metabolism when dwelling in the plant host . It is more likely that it fulfills a major role in the passage of trypanosomatids through their insect vector rather than to their survival in the widely different types of mammalian and plant hosts , where only some plants have high concentrations of trehalose . Both isolates seem to differ in the make-up of their surface glycoproteins ( examples in the Text S1 and Table S3D ) . Most significantly , HART1 and EM1 have radically distinct gp63 repertoires , with only two genes detected in EM1 but over 20 in HART1 . These are clearly derived from a single common precursor , with multiple expansions in HART1 , and suggesting a more complex surface for HART1 than EM1 that potentially facilitated adaptation to a greater range of conditions , host species or tissue spaces . Perhaps the pathogenic effects of HART1 are primarily due to their location in the sap , containing the products of photosynthesis and essential for plant growth . The death of palms , coffee trees and Alpinia may be the result of competition for essential metabolites that are more efficiently scavenged by Phytomonas . Biological inoculation experiments using the isolate EM1 in palms would address this hypothesis . The specific relationships among Phytomonas , its vector , and the host make this experiment hard to endeavour . The intraphloemic trypanosomatids associated with wilts form a very distinct group , especially for their cultivation [15] . Parasites could not be isolated without the help of feeder cells in the cultures , while the cultures of latex isolates like EM1 or fruit isolates were obtained in an axenic medium . Further comparative analysis of the two Phytomonas genomes may reveal the source of these differences . The Phytomonas genomes consist essentially of single copy genes for the bulk of their metabolic enzymes , whereas Leishmania and Trypanosoma possess numerous duplicated genes or large gene families . While such gene duplications may have helped some trypanosomatids to adapt to multiple , widely different hosts , i . e . poikilothermic insects and warm-blooded mammals , their absence in the two Phytomonas genomes analyzed here suggests that plant trypanosomatids have been confronted much less with strikingly different metabolic environments and temperatures , and have hence lost or never needed these additional paralogs . The unlimited availability of sugars in both plant and insect hosts is a situation that normally leads to suppression of mitochondrial activities , such as cyanide-sensitive respiration and oxidative phosphorylation . Eventually this may result in an irreversible loss of the genes coding for all of the above functions . The irreversible suppression observed in Phytomonas resembles the adaptation of some African trypanosomes to a permanent stay in the bloodstream of their mammalian hosts , without the possibility for cyclic transmission via insects . This also leads to the loss of mitochondrial genes and results in the appearance of dyskinetoplastic or akinetoplastic trypanosomes , unable to survive in the tsetse fly . Interestingly , Phytomonas spp . possess orthologs to the mitochondrial calcium uniporter recently described [106] suggesting that , as bloodstream forms of T . brucei , they also utilize the mitochondrial ATPase in reverse to maintain a membrane potential that drives Ca2+ uptake through the uniporter . Some Phytomonas genes likely gained via HGT may have permitted increased flexibility of genome expression , enabling the successful adaptation of Phytomonas spp . Significant genome reduction has been identified in microbial lineages living in selective environments . Selection plays a key role during the initial phases of such adaptation removing “accessory” genes [107] . High gene density bear witness of genome contraction in several obligate intracellular parasites . In the case of microsporidia , genome-size variation resulted from varying frequencies of repeat elements without affecting gene density . Furthermore , Phytomonas shows important host dependency pictured by considerable gene losses . These adaptations combined with genome compaction led to gene size reduction and simplification of certain cellular processes [108] . Also phytoplasmas , specialized bacteria living as obligate parasites of plant phloem tissue and transmitting insects [109] , have suffered extreme genome shrinkage , which resulted in a gene repertoire that is specific for survival in plant hosts [110] . In the case of phytoplasmas , this adaptation was made possible thanks to the presence of repeated DNAs , which allowed survival in different environments . Here also adaptation is particularly important , as their host environments , including phloem tissues of plants , and guts , salivary glands , and other organs and tissues of the insect host , are extremely variable [110] . Phytomonas spp . are highly specialized trypanosomes , with central differences in life history and infection strategy compared to eukaryotic plant pathogens like fungi and oomycetes . Leaf , fruit and stem are some of the surfaces colonized by plant pathogens . Wind-blown rain , fog and any plant visitor are some of the mechanisms by which phytopathogens like filamentous fungi and oomycetes are disseminated to the host plants . Still , these phytopathogens should penetrate by themselves in order to colonize and circulate inside the host . In this process , several biological mechanisms are triggered to colonize and propagate into the host , by the use of enzymes ( cutinase , cellulase , pectinase ) , hormones , toxins and frequently by the interaction with metabolites produced by the plant in response to the invasion [111] . Contrary to these phytopathogens , plant trypanosomatids do not need to degrade cell walls to settle inside the plant since they are deposited into very specialized tissues or cells in the host thanks to insects that acts as their natural carriers . Yet , the discovery of a Phytomonas HART1 AP gene cluster , known to be secreted and involved in the virulence of several pathogenic fungi [100] , [101] but missing in animal parasitic trypanosomids and the non pathogenous EM1 isolate , could be described as a good example of convergent evolution between these distant phytopathogen organisms . The genome completion is the first step towards development of effective chemical control agents against Phytomonas spp . , which is not only of economic interest , but may have relevance for other Trypanosomatidae pathogenic to humans and animals , since they share similar metabolic routes and many other biological mechanisms [112] , [113] . Comparative studies between plant , human and animal pathogenic trypanosomatids as well as free living species will assist in the identification of gene cohorts specifically linked to various pathogenesis mechanisms . These comparisons will also contribute towards better and safer control methods for trypanosomatid diseases of animals , plants and humans and provide better insights into the evolution of parasitic and pathogenic mechanisms .
The sequencing strategy used for both Phytomonas genomes corresponds to a mix of three technologies: 454/Roche for most coverage; Solexa/Illumina for automatic corrections of low-quality regions ( especially around homopolymers ) ; and classical Sanger sequencing at low coverage with large-insert clones ( 10 kb-insert containing plasmids and fosmids ) to organize the contigs into scaffolds . The assembly method is described in detail in Text S1 . Illumina reads were mapped to the corresponding Phytomonas genome using SOAP version 1 . 10 [114] , under the guidance of a custom perl script . The number of bases mapping to each position in each scaffold was recorded , and used to determine the total number of read bases mapping to each scaffold and the median read depth for each scaffold . Observing that a majority of the scaffolds displayed similar median read depths , and interpreting this as a nominal ‘ploidy’ , a within-genome normalisation was performed by setting the average of the read depth of the four longest ‘euploidic’ scaffolds to 2 . The read depth for each scaffold was subsequently normalized to this value . Results of the scaffolds “somy” are shown in Table S1 . Protein-coding genes are predicted by combining ab initio gene model predictions ( already trained on manually annotated genes ) and homology searches , using collections of expressed sequences - full length cDNAs , ESTs or massive-scale mRNA sequences from the same or closely related organisms – proteins or other genomic sequences . Details on the pipeline are given in the Text S1 . Moreover , tRNA-Scan [115] was used to detect tRNAs in both Phytomonas assembled sequences . After a final integration of all gene evidence using GAZE [116] , the final proteome was delivered with computed annotation data , such as ortholog and paralog associations , functional domains and ontology relationships . Phytomonas proteins were used against the protein nr database ( blastx , [117] ) , with the parameters “-f 100 -X 100 -e 0 . 00001 -E 2 -W 5” , and the best hits were retained using the following criteria: only BLAST scores greater than 90% of the best score outside kinetoplastids ( so that horizontal gene transfers shared between kinetoplastids could be detected ) and above 100 were retained . Then , the proteins with all their best hits in bacteria or archaea were retained as candidates to have arisen from bacterial/archeal horizontal gene transfers . All the candidates were then manually inspected , which provided 87 final candidate HGT genes , 80 of which have orthologs in other trypanosomatids , and 8 have no orthologs in Leishmania spp . nor Trypanosoma spp . and might thus be specific of Phytomonas ( Table S4 ) . We identified orthologous genes between Phytomonas EM1 and HART1 , and 4 other trypanosomatids: L . major [29] , T . brucei [28] , T . cruzi [30] and T . vivax [70] ( Tritryp release 2 . 1 ) . Each pair of annotated genes was aligned with the Smith-Waterman algorithm , and alignments with a score higher than 300 ( BLOSUM62 , gapo = 10 , gape = 1 ) were retained . Orthologs were defined as best reciprocal hits ( BRH ) , i . e . two genes , A from genome GA and B from genome GB , were considered orthologs if B is the best match for gene A in GB and A is the best match for B in GA . Indeed , 5006 gene pairs ( representing 77 . 6% for HART1 and 78 . 4% for EM1 genes ) , similar in gene size and intergenic length , were detected using this approach ( Figure S9A and B ) . The number of BRH for each comparison and their average and median %id are displayed in Figure 3; the distribution of %id for these BRH between different pairs of species designated both Phytomonas isolates as being much closer to each other than to other trypanosomes ( average %id of 70% for EM1 and HART1; 56 to 57 . 6% between the different pairs of trypanosomes ) . The results of the pairwise alignments between all protein sequences of the 6 genomes were then inputted to the orthoMCL software V1 . 4 [69] , in order to assemble clusters of orthologous genes between both Phytomonas EM1 and HART1 , and other trypanosomatids . This approach was complementary to what was observed by the BRH strategy , since it permitted us to ascertain orthologs for multigenic families , not seen by the BRH strategy alone . This procedure provided 7 , 694 clusters of orthologs genes , gathering 5 , 188 EM1 and 4 , 643 HART1 genes in clusters containing genes from both isolates ( regardless of the presence or absence of genes from other trypanosomes ) ( Table 1 ) . We also ran orthoMCL on the subset of genes from EM1 and HART1 that have strong support ( i . e that are overlapping uniprot genewise hits , or cDNA reads as well as ab initio predictions ) : EM1 contains 5 , 237 such genes , and HART1 5 , 247 genes ( Table 1 ) . An in-house perl script was used to draw the dot plots and build syntenic blocks between species . The clustering was performed by single linkage clustering using the euclidian distance between genes . Those distances were calculated with the gene index in each scaffold rather than the genomic position . The minimal distance between two orthologous genes was set to 10 on both counterparts and we only retained clusters that were composed of at least 5 pairs of paralogous genes . The boundaries of the syntenic clusters were then filtered in order to eliminate those occuring at the end of scaffolds and corresponding to “assembly breaks” rather than synteny breaks . As a consequence , for genomes with a more fragmented assembly , the number of synteny breaks detected is lower because some real syntenic breaks occur at scaffolds boundaries and are discarded . This is the case for T . cruzi ( 41 scaffolds ) that appears to have less synteny breaks with Phytomonas compared to T . vivax and T . brucei ( 11 chromosomes ) . We performed a simulation to distribute randomly the same number of synteny breaks as observed for each scaffold ( 1000 iterations ) and counted the number of randomly distributed synteny breaks that coincided with operon boundaries . In all cases , the observed number of synteny breaks at operon boundaries was significantly higher than expected randomly ( Figure S12B ) . Both genomic EM1 and HART1 assemblies were queried using sequence probes encompassing selected Interpro domains , by a series of reciprocal sequence comparisons using the BLAST server , accessed through the SeqTryplant Genome Browser or directly on a secure web site . Likewise , the reads not included in the assembly as well as the contigs smaller than 5 kb and so excluded from the assembled sequence , were scanned with the same probes . The results obtained were subsequently examined by the experts of the Phytomonas consortia in order to validate the gene models . Details on the probes and procedure used for each gene family can be found in the Text S1 file . The same gene probes used to search for gene families in both Phytomonas genomes were later employed to query the TriTrypDB 4 . 0 Released , in order to obtain the corresponding genes in the T . brucei , T . cruzi and L . major genome annotations . Later on , these sequences were applied to query the OrthoMCL DB ( version 5 ) , and copy number , as automatically defined by the OrthoMCL approach was reported . Moreover , T . brucei , T . cruzi and/or Leishmania spp . gene copy number for members of certain families ( e . g . kinases and transporters ) was obtained from the literature or human expertise when available ( Table 2 and Table S3 ) . Proteins with a signal peptide were detected with SignalP version 3 . 0 [122] , transmembrane domains were detected with TMHMM 2 . 0 [123] and GPI anchors with KOHGPI version 1 . 5 ( http://gpi . unibe . ch/ ) . Proteins harboring a signal peptide , not containing transmembrane domains nor GPI anchors were considered to be secreted by Phytomonas , and their annotation was performed using BLASTp against the non-redundant NCBI database , Interproscan and Gene Ontology . We retrieved aspartic proteases from others clades ( amoebae , plants , chromalveolates , fungi and animals ) using the Phytomonas aspartic proteases amino acid sequences as queries with BLASTp on the NCBI nr database [121] . Phylogenetic analysis was executed on the Phylogeny . fr platform [124] as described in [125] , with the parameters “minimum length of a block after gap cleaning: 5 , no gap positions were allowed in the final alignment , all segments with contiguous non conserved positions bigger than 8 were rejected , minimum number of sequences for a flank position: 85%” for Gblocks v0 . 91b [126] . | Some plant trypanosomes , single-celled organisms living in phloem sap , are responsible for important palm diseases , inducing frequent expensive and toxic insecticide treatments against their insect vectors . Other trypanosomes multiply in latex tubes without detriment to their host . Despite the wide range of behaviors and impacts , these trypanosomes have been rather unceremoniously lumped into a single genus: Phytomonas . A battery of molecular probes has been used for their characterization but no clear phylogeny or classification has been established . We have sequenced the genomes of a pathogenic phloem-specific Phytomonas from a diseased South American coconut palm and a latex-specific isolate collected from an apparently healthy wild euphorb in the south of France . Upon comparison with each other and with human pathogenic trypanosomes , both Phytomonas revealed distinctive compact genomes , consisting essentially of single-copy genes , with the vast majority of genes shared by both isolates irrespective of their effect on the host . A strong cohort of enzymes in the sugar metabolism pathways was consistent with the nutritional environments found in plants . The genetic nuances may reveal the basis for the behavioral differences between these two unique plant parasites , and indicate the direction of our future studies in search of effective treatment of the crop disease parasites . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"sequencing",
"genome",
"analysis",
"tools",
"plant",
"science",
"plant",
"pathogens",
"plant",
"pathology",
"comparative",
"genomics",
"biology",
"genomics",
"gene",
"prediction",
"computational",
"biology"
] | 2014 | The Streamlined Genome of Phytomonas spp. Relative to Human Pathogenic Kinetoplastids Reveals a Parasite Tailored for Plants |
The theory of phase oscillators is an essential tool for understanding population dynamics of pacemaking neurons . GABAergic pacemakers in the substantia nigra pars reticulata ( SNr ) , a main basal ganglia ( BG ) output nucleus , receive inputs from the direct and indirect pathways at distal and proximal regions of their dendritic arbors , respectively . We combine theory , optogenetic stimulation and electrophysiological experiments in acute brain slices to ask how dendritic properties impact the propensity of the various inputs , arriving at different locations along the dendrite , to recruit or entrain SNr pacemakers . By combining cable theory with sinusoidally-modulated optogenetic activation of either proximal somatodendritic regions or the entire somatodendritic arbor of SNr neurons , we construct an analytical model that accurately fits the empirically measured somatic current response to inputs arising from illuminating the soma and various portions of the dendritic field . We show that the extent of the dendritic tree that is illuminated generates measurable and systematic differences in the pacemaker’s phase response curve ( PRC ) , causing a shift in its peak . Finally , we show that the divergent PRCs correctly predict differences in two major features of the collective dynamics of SNr neurons: the fidelity of population responses to sudden step-like changes in inputs; and the phase latency at which SNr neurons are entrained by rhythmic stimulation , which can occur in the BG under both physiological and pathophysiological conditions . Our novel method generates measurable and physiologically meaningful spatial effects , and provides the first empirical demonstration of how the collective responses of SNr pacemakers are determined by the transmission properties of their dendrites . SNr dendrites may serve to delay distal striatal inputs so that they impinge on the spike initiation zone simultaneously with pallidal and subthalamic inputs in order to guarantee a fair competition between the influence of the monosynaptic direct- and polysynaptic indirect pathways .
The basal ganglia ( BG ) are a collection of forebrain nuclei involved in various aspects of motor control and habit formation . The substantia nigra pars reticulata ( SNr ) is one of the main output nuclei of the BG , innervating the ventral thalamus , superior colliculus and reticular formation [1–3] . SNr GABAergic neurons receive thousands of synaptic inputs . Most of them are inhibitory inputs arising from direct pathway spiny projection neurons ( dSPNs ) in the striatum [4 , 5] , or from the external segment of the globus pallidus ( GPe ) in the indirect pathway [6 , 7] . Excitatory inputs originate from the subthalamic nucleus ( STN ) [8 , 9] , the pedunculpontine nuclei ( PPN ) [10] , and , to a lesser extent , from the cerebral cortex [11] . Thus , in the SNr , inputs from the direct and indirect pathways converge onto the same neuron . The vast majority of SNr GABAergic neurons are autonomously active and discharge continuously and regularly at 6–30 spikes/s in rodents in vitro [12–15] , even when their synaptic inputs are completely blocked and at higher rates in vivo [16–18] . A rhythmically firing neuron–also called a pacemaker neuron–can be represented by a phase variable that advances from 0 to 1 as the somatic voltage of the pacemaker neuron advances from one action potential threshold to the next . Such a phase oscillator is sensitive to the timing of impinging inputs . Thus , in the case of a pacemaker neuron , each individual synaptic input is sufficiently small so that its only lasting effect is a finite change in the phase of the pacemaker that either postpones or advances the next spike . The sign and amplitude of this phase change depend on the phase of the pacemaker’s cycle at which the synaptic input arrived . Thus , the complex neuronal dynamics of synaptic integration by the pacemaker are reduced to phase change as a function of the timing of inputs [19–22] . This reduction vastly simplifies the treatment of the collective dynamics of populations of pacemakers . The quantity that describes the susceptibility of the pacemaker neuron’s phase to small voltage perturbation as a function of the timing of the perturbation is called the neuron’s phase response curve ( PRC ) [19–22] . Although the phase reduction and the use of PRCs are theoretically valid only for weak perturbations , they prove to be very useful in practice . The PRC in response to somatic stimulations can be easily measured experimentally , and has been used successfully to predict neurons’ responses to arbitrary patterns of somatic current injections [23–30] . Thus , the phase reduction is computationally simple , can capture the essence of the neuronal dynamics , and can be readily implemented experimentally on real neurons . The use of phase reductions and PRCs has been applied both experimentally and theoretically [23 , 24 , 27 , 29 , 30] , by and large , only to somatic voltage perturbations . However , neurons can possess elaborate dendritic arbors . Previous theoretical and experimental work has suggested that PRCs need to be redefined and generalized to accommodate inputs arriving at the dendritic arbors [31–37] . These studies predicted that taking the dendritic localization of inputs into consideration could have a measurable effect on collective neurons dynamics . Here , we put this prediction to test for the first time , and do so for actual neuronal pacemakers with dendrites . SNr neurons in rodents possess dendrites that can extend up to 0 . 75 mm long , usually branching only once or twice [38 , 39] . Excitatory inputs from the STN are distributed along the entire length of the dendrites [40 , 41] . In contrast , the two major inhibitory inputs are differentially distributed: inputs from the GPe , belonging to the indirect pathway , are located at the soma and on proximal dendrites [42–45] , while direct pathway striatal synapses impinge upon smaller distal dendrites and terminal tufts [45] . The functional relevance of the differential spatial organization of direct and indirect pathway inputs is unknown . We combine theory and electrophysiological experiments in acute brain slices from transgenic mice that express channelrhodopsin-2 ( Thy1-ChR2 mice ) in GABAergic SNr neurons [46 , 47] to investigate the impact of dendritic localization of synaptic inputs on the control of spike timing in these pacemakers . This assay allows us to directly perturb the membrane voltage at various locations on the somatodendritic tree . In order to compare stimulation of proximal vs . distal locations on the dendritic arbor , we illuminate either a small region containing the soma and proximal dendrites or a wider field of view containing the entire dendritic tree , while measuring somatic currents and voltage . This approach does not take into account the complex morphology of the cells . However , eliminating the effect of the variability arising from the shape of individual dendritic arbors would hopefully generate more robust results . Moreover , it allows us to use a simple cable theoretic model , in which the entire dendritic tree is collapsed to a single cable , to fit the experimental results [48] . We begin by constructing a model that accurately predicts the somatic current response to inputs arising from stimulating various portions of the dendritic arbor . Next , we show that the choice of whether to illuminate only the soma and proximal dendrites or the entire dendritic arbor generates measurable and systematic differences in the PRCs corresponding to each of these conditions . Finally , we show that the divergent PRCs correctly predict differences in two major features of the collective dynamics of SNr neurons . The fidelity and latency of population responses to sudden changes in inputs , on the one hand , and the phase latency at which SNr neurons are entrained by rhythmic stimulation , on the other , are both determined by the extent of the dendritic tree illuminated–in close agreement with predictions arising from their respective empirical PRCs . This is the first experimental demonstration of the impact of the electrotonic length of the region of the dendritic arbor being activated on the collective dynamics of pacemaker neurons . Our results suggest a possible role for the differential spatial distribution of inputs onto SNr GABAergic neurons , where SNr dendrites may serve to delay distal striatal inputs so that they impinge on the spike initiation zone simultaneously with pallidal and subthalamic inputs . Coherent rhythmic inputs to SNr neurons arise under various physiological ( e . g . , sleep , [49] ) and pathophysiological ( e . g . , Parkinsonism , [18 , 50] ) conditions . Thus , the way in which these neurons follow their inputs , or become entrained to them , will depend on the spatial distribution of their inputs throughout their dendritic arbor ( e . g . , distal striatal vs . proximal pallidal inputs ) .
In order to investigate the impact of the dendritic location of inputs on the SNr pacemakers , we employed our previously published model of a semi-infinite cable attached to a pacemaking axosomatic compartment [32] . Active dendritic conductances are incorporated into the model by linearizing them in the vicinity of the dendrite’s resting membrane potential , resulting in a quasi-linear cable u ( x , t ) ( Fig 1A ) . The cable’s Green’s function ( or impulse response ) describes the spatio-temporal voltage response to a brief perturbation applied at the origin at time t = 0 ( with vanishing boundary conditions at ±∞ ) and provides a method to solve the cable’s partial differential equation . The Green’s function is given by G ( x , t ) =∫−∞∞dω2π∫−∞∞dk2πei ( kx+ωt ) λ2k2+α ( ω ) +iβ ( ω ) Where λ is the characteristic length of the cable , k is the wave number and ω is the angular frequency in the spectral representation of G . α ( ω ) and β ( ω ) are the contributions from the passive cable properties as well as terms that arise from linearizing the nonlinear conductances in the cable [32 , 33] . In particular , for a linear cable α ( ω ) = 1 and β ( ω ) = ωτ [32] , where τ is the time constant of the cable . Because we are interested in the current injected into the soma by the cable , we assume that it is semi-infinite where the somatic boundary at x = 0 is kept at a constant holding potential u ( 0 , t ) = u0 for all t . In the voltage clamp experiments this assumption is enforced experimentally . Thus , the current injected into the soma is proportional to the spatial derivative of u ( x , t ) at x = 0 , with a proportionality factor κ: I ( t ) =κ∂∂x|x=0u ( x , t ) The Green’s function for the semi-infinite ( si ) cable with these boundary conditions is given by Gsi ( x , y , t ) =G ( x−y , t ) −G ( x+y , t ) and describes how a voltage perturbation at location y propagates to location x . The optogenetic stimulation of the cell is modeled as a band of illumination that starts at distance r from the soma and ends at distance R , where the stimulus has the temporal waveform cos ( ωt ) , with ω = 2πf where f is the driving frequency . To calculate the current injected into the soma , it is necessary to integrate the input from y = r to y = R and over all time . Thus , I ( t ) =κ∂∂x|x=0u ( x , t ) ==κ∂∂x|x=0∫rRdy∫0∞dscos ( ωs ) ∫−∞∞dΩ2π∫−∞∞dk2πeikx+iΩ ( t−s ) λ2k2+α ( Ω ) +iβ ( Ω ) ( e−iky−eiky ) ==κ[F ( r , t;ω ) −F ( R , t;ω ) ] Where F ( r , t;ω ) =exp ( −p ( ω ) rλ ) λp ( ω ) 2+q ( ω ) 2cos ( ωt−q ( ω ) rλ−arctan ( q ( ω ) p ( ω ) ) ) With p ( ω ) =α ( ω ) 2+β ( ω ) 2+α ( ω ) 2andq ( ω ) =α ( ω ) 2+β ( ω ) 2−α ( ω ) 2 This means that illuminating a band of the cable with a sinusoidal temporal envelope generates a somatic current injection with two sinusoidal contributions , each with an amplitude that decays exponentially with distance of the boundaries of the band from the soma , and each with two contributions to the phase shift relative to the phase of the input . One shift that scales with the distance and another that does not . The spatial decay and all phase shifts increase with the driving frequency f . To measure the resulting phase shift between the injected current cos ( ωt ) and the somatic current response I ( t ) , we calculate the location of the peak of the cross-correlation function ( CCF ) between them by averaging their product over all time . The CCF is given by: C ( Δ ) =〈cos ( ω[t−Δ] ) I ( t ) 〉=κ[F ( r , Δ;ω ) −F ( R , Δ;ω ) ]/2 Differentiating with respect to Δ and comparing to zero results in the following equation: sin ( ωΔ−q ( ω ) Rλ−arctan ( q ( ω ) p ( ω ) ) +q ( ω ) δRλ ) =e−p ( ω ) δRλsin ( ωΔ−q ( ω ) Rλ−arctan ( q ( ω ) p ( ω ) ) ) where δR≡R-r>0 . This equation can be used to find the phase shift Φ = ωΔ corresponding to the peak in the CCF . We consider two cases: In this case , after linearization we get the equation ΦR=ωΔ=Rλq ( ω ) mod ( 2π ) [1] At this limit , the phase behaves like the stimulation of a single point at distance R from the soma ( Fig 1A , purple dot ) . In this case ( Fig 1A , green band ) , we get Φ0=ωΔ=arctan ( q ( ω ) p ( ω ) ) −arctan ( sin ( Rλq ( ω ) ) eRλp ( ω ) −cos ( Rλq ( ω ) ) ) [2] To summarize , the model allows us to calculate the current injected by the cable into the soma in response to a sinusoidal current perturbation with frequency f , given at some location R in the dendrite ( case I ) , or applied to a region of the dendrite beginning at the soma ( case II ) . Our analysis shows that the current arriving at the soma is also sinusoidal , but it lags by a phase delay of ΦR/2π or Φ0/2π , respectively . Note that the phase delay following a field stimulation is bounded in the case of a linear dendrite [e . g . , α ( ω ) = 1 and β ( ω ) = ωτ] , as the distance from the soma grows ( R→∞ , Eq 2 ) . In contrast , the phase shift in response to point stimulation ( Eq 1 ) , continues to increase linearly with the distance from the soma ( Fig 1B ) . To measure the effect of the dendritic localization of inputs on the somatic current response , we used transgenic mice that express ChR2 throughout the entire somatodendritic arbor of SNr neurons [46 , 47] . We chose to compare two extreme cases: illuminating either the soma and proximal dendrites ( a diameter of ~130 μm around the soma , see Materials and Methods ) or the entire dendritic tree . This choice was motivated by three considerations . First , we reasoned that even though the most straightforward comparison would be to compare a point-like proximal illumination to a distal one , the strategy of illuminating the entire region is likely to create a more consistent and robust effect across diverse geometries of dendritic arbors . While illumination of a larger portion of the dendritic arbor should result in a larger somatic current , we opted to adjust the illumination so that the amplitude of the somatic current is similar under both conditions because we wanted to remain within the weak perturbation regime . Second , the cable theoretic model we use to fit the data is more applicable under these conditions , as illuminating a band of cable in the model is comparable to illuminating an annulus in the plane , centered around the soma . Note that in our model the edges of the stimulated region are sharp , while our empirical stimulation scheme generates some falloff between the center of the illuminated area and its boundaries . However , the transition region is small compared to the difference between the diameters corresponding to the two conditions , adding a negligible error to the estimate generated by our model . Finally , both situations could conceivably be physiologically relevant . Coherent rhythmic input could arise under physiological conditions such as sleep [16 , 49] and pathophysiological conditions , such as Parkinsonism [18 , 50] . Coherent pallidal input would preferentially target the soma and proximal dendrites [42 , 43 , 45] , whereas coherent subthalamic input could activate the entire dendritic arbor [40 , 41] . To ensure that our proximal stimulation condition indeed activates a portion of the dendritic arbor , we employed two-photon laser scanning microscopy and measured the diameters of SNr somata . GABAergic SNr cells tend to be elongated , and the diameters of their somata measured 17 . 7–33 . 74 μm ( with a mean of 24 . 99±SD of 4 . 88 μm , n = 15 ) along the larger axis . Thus , both stimulation conditions activate dendrites as well as the cell’s soma and they differ in the extent of the dendritic arbor that they stimulate . Hence , we optogenetically stimulated the ChR2-expressing SNr neurons using 470 nm LED light that was sinusoidally modulated at various temporal frequencies ( 0 . 25 to 16 Hz , 3–4 seconds of stimulation per frequency ) , while blocking all glutamatergic and GABAergic inputs . The cell was held at –70 mV to prevent spiking , and the current injected to the soma was measured in whole-cell voltage clamp . This was repeated under the two stimulation conditions: a ) proximal stimulation targeting the soma and proximal dendrites; and b ) full-field illumination exciting the soma as well as the entire dendritic arbor ( Fig 2A ) . The current measured with a patch electrode at the soma in response to the sinusoidal optogenetic stimulation was a phase shifted sinusoidal waveform of the same temporal frequency ( Fig 2B ) . The phase shift between the periodic stimulation and the somatic response was determined by the location of the peak in the CCF between the two traces ( Fig 2C ) . This was used to generate a plot of the phase shift as a function of stimulation frequency for each illumination condition ( Fig 2E ) . Amplitudes of somatic current responses were similar for proximal and full-field illumination ( Fig 2F ) , and we saw no relationship between current response amplitude and the magnitude of the phase shift . However , the two spatial conditions give rise to distinct curves ( n = 6; p = 1 . 2·10−6 , ANCOVA ) . As expected , the phase increases with frequency in both conditions . The phase shift between the stimulation and the somatic current response is larger when the entire dendritic field is illuminated compared to only the soma and proximal dendrites ( Fig 2E ) . Thus , the response of SNr GABAergic neurons to subthreshold rhythmic stimulation shows a spatial effect , indicating that optogenetic methods can reveal the contribution of the electrotonic properties of the dendrites to somatic currents . Importantly , the experiment was repeated with holding voltages of –60 mV and –50 mV and yielded similar results , indicating that there is no voltage dependence of the phase shifts in the subthreshold range ( S1 Fig ) . In other words , under these stimulation conditions the dendrites do not exhibit substantial subthreshold nonlinearities . In order to estimate the electrotonic properties of SNr dendrites , we fit the parameters τ and ρ≡R/λ of our cable theoretic model to our data [48] . The empirical curves describing the dependence of the phase shift on frequency were fit to the appropriate case II of the model ( Eq 2 ) . Because the electrode does not measure the current injected by the dendrite into the soma directly , but rather the current at the patch pipette tip , we need to add to the dendritic phase shift ( predicted by the model ) the phase introduced by the somatic membrane ( Fig 2D ) . Assuming an isopotential passive membrane ( which is a good approximation when holding at subthreshold voltages ) yields the additional angular phase of ΦS = arctan ( 2πfτ ) ( Fig 2D ) . Even with the simplifying assumption of a strictly passive linear dendrite–which is in line with the voltage independence of the phase shift we described above–we attain a good agreement between the predictions of our model and the empirically observed average frequency dependent phase shift at the soma ( Fig 2E ) . Fitting our theoretical curve to the empirical data for the two conditions results in similar and physiologically plausible [51] values for the membrane time constants ( i . e . , τproximal = 9 ms , τfull-field = 11 ms , Table 1 ) and the characteristic length λ . The ratio ρ of the boundary of the stimulated region to the dendritic characteristic length was also extracted ( ρproximal = 0 . 05 , ρfull-field = 0 . 44 , Table 1 ) . From ρproximal we can deduce that the effective characteristic length of the SNr dendrites is approximately 1 . 3 mm ( = 65 μm/0 . 05 ) [51] . Moreover , the model estimates that the value corresponding to illumination of the neuron’s entire morphology is ~9 times larger than the value for illumination of only the soma and proximal dendrites . Therefore , we can deduce that on average dendrites are light-activated out to ~580 μm from the soma . This is a reasonable estimate , considering that in acute slices regions beyond that are often cut or are too deep to be affected by the illumination . The experiment repeated at -60 mV ( S1 Fig ) gave rise to a different estimate of ρfull-field = 0 . 22 , but because the estimates at -70 mV and -50 mV were similar we trusted these . If we were to choose an average of the three estimates it would lower our estimate of the diameter up to which the dendrites were activated but it would still be several hundred micrometers out . In summary , our theoretical and experimental results demonstrate that distal and proximal inputs undergo different phase shifts by the time they reach the soma . We therefore hypothesize that these different phase shifts will also differentially affect the suprathreshold behavior of these pacemaking neurons . To address this hypothesis , we estimated proximal and full-field PRCs by incorporating our two illumination conditions into a previously described method for PRC measurement using optogenetic pulses [47] . The neurons were subjected to Poisson-like processes composed of barrages of 0 . 5–1 ms long light pulses , separated by exponentially distributed inter-pulse intervals ( Fig 3A , see Materials and Methods ) . In order to examine the ChR2 currents generated by the barrage stimulation , cells were hyperpolarized to prevent spiking and steady-state currents ( 4–9 seconds after beginning of stimulation to avoid any effects of ChR2 deactivation ) were measured in the whole-cell voltage clamp configuration ( Fig 3A ) . For each cell ( n = 11 ) , a curve portraying the average current response to a light pulse was generated ( Fig 3B ) . As previously reported , average responses were well fit by a double exponential , chosen to accommodate ChR2 deactivation kinetics [47] ( Eq 9 , see Materials and Methods ) , and rise and decay time constants ( τrise , τdecay ) were extracted ( Table 1 ) . Decay time constants were not significantly different for the two conditions ( S2 Fig ) . However , rise time constants were significantly larger when the entire dendritic tree was illuminated compared to only the soma and proximal dendrites . This is evident in the distribution of rise time constants ( p = 8 . 15·10−5 , RST ) ( Fig 3C ) , as well as in the average current response curves for the different conditions ( Fig 3B ) . The delayed peak of the somatic response when the entire dendritic tree is illuminated is consistent with the delayed full-field response that we observed in the previous section . Thus , the barrage stimulation induces disparate somatic current profiles under the proximal and full-field stimulation conditions . We have shown previously that the more distal a dendritic perturbation is , the more leftward a shift it will generate in the dendritic PRC compared to the somatic one [32] . Thus , the shift we observed in the current profile generated by the full-field illumination relative to the proximal one buttresses the hypothesis that the empirical PRC measured under the full-field illumination should exhibit a leftward shift relative to that measured under proximal illumination . To test this hypothesis , we applied the barrage stimulation to spontaneously firing SNr cells recorded in the cell-attached configuration ( Fig 3A ) and estimated the two PRCs corresponding to the proximal and full-field illumination conditions . For each stimulation condition and for every cell , the PRC was estimated using a multiple linear regression method [47 , 52] ( see Materials and Methods ) . Our estimations revealed that the PRCs of SNr GABAergic pacemakers have a type I structure ( i . e . , do not contain a negative lobe meaning that the excitatory perturbations can only advance the next spike ) . This is consistent with previous PRC measurements which uncover a triangular form—the curve slopes upwards before peaking and falling to zero as the phase approaches 1 [47] . We therefore fit a triangle ( parameterized by a parameter 0<θ<1 for the location of the peak , an amplitude A and an offset C ) to our experimental data ( see Materials and Methods ) , in order to extract the location of the peak ( Table 1 ) . Our empirical data include negative values around phases 0 and 1 ( Fig 3D ) due to jitter in the pacemaker’s unperturbed period in the course of the recording . Due to the refractory period of the action potential , phase theory dictates that the PRC should pass through the points ( 0 , 0 ) and ( 0 , 1 ) . Thus , the empirical negative values are inconsequential . However , we focus on the phase of the PRC’s peak for the two stimulation conditions ( θproximal , θfull-field ) , and these values remain almost unchanged when the curve is forced to be zero at the beginning and end of the period . Our analysis results in two distinct average PRCs ( n = 19; p = 2 . 13·10−7 , ANCOVA ) , with the full-field PRC peaking before the proximal one ( Fig 3D ) . The steep negative slope at late phases represents the cells approaching the causal limit , where the input elicits an immediate spike [53] . According to our theoretical model , as well as the empirical results , the somatic current generated by a stimulation delivered in the full-field condition would be phase delayed compared to a proximal stimulation , which explains why the peak of the full-field PRC is shifted leftward relative to the proximal PRC . Hence , differentially located inputs arrive at the soma at different times , thereby shifting the full-field PRC leftward relative to the proximal one . Next , we turn to measure properties of the collective activity of SNr pacemakers , which theory predicts are dependent on the shape of the PRC [53–56] . First , we measure the fidelity and latency of the population response of SNr neurons to sudden changes in their input , and we then consider the propensity of rhythmic inputs to entrain SNr pacemaker neurons . Barrage stimulation changed the firing frequency of SNr GABAergic cells ( Fig 4A ) . The perturbation significantly increased the pacemaker’s mean firing rate by 33% for proximal ( n = 19; p = 1 . 32·10−4 , SRT ) and by 19% for full-field ( p = 1 . 5·10−3 , SRT ) illumination . Importantly , the distributions of baseline ( p = 0 . 82 , RST ) and perturbed ( p = 0 . 45 , RST ) firing rates were not significantly different across the two stimulation conditions ( Fig 4B ) . In a population of cells , the response to a step stimulation consists of two stages—an initial increase in firing and a relaxation to a steady-state rate . The rapidity at which the PSTH can track the change in inputs is a measure of the fidelity of the neurons’ coding of the input . The shape of the PSTH has been shown to be theoretically related to that of the PRC [56] . The Fokker-Planck formalism enables one to predict this relationship ( See Eq 8 in Materials and Methods ) [54 , 55] . Interestingly , this connection demonstrates that the initial rise in the PSTH should reflect the mirror image of the falling phase of the PRC [54] . Thus , the leftward shift in the full-field PRC should generate a rightward shift in the rise time of the full-field PSTH relative to the proximal PSTH . Estimation of the empirical PSTHs under both conditions confirmed this prediction , with the peak in the proximal PSTH appearing approximately 12 ms before that of the full-field PSTH ( Fig 4C ) . Moreover , fitting the empirical PSTH to the theoretical equation for the PSTH with an underlying triangular PRC , enables us to extract independent estimates of the peak of the triangular PRC based on the rise time of the PSTH ( θp = 0 . 901 , θf = 0 . 733 , Table 1 ) . Indeed , this estimation gave rise to a larger ( but comparable ) estimate of the phase shift between the peaks of the underlying proximal and full-field PRCs . A caveat in PRC estimation is the difficulty to estimate it empirically near the causal limit [53] . Thus , for weak perturbations , values of PRC peaks extracted from the PSTHs may provide a more precise estimation of the true shape of PRC’s late downswing , which may be occluded by the causal limit . The discharge of a regularly active neuron can be entrained by small amplitude sinusoidal currents delivered to the soma , when the driving frequency is close to the neuron’s natural firing rate ( or to integer multiples of it ) . Within this range of frequencies , the rhythmic stimulation reshapes the neuron’s firing pattern—the firing rate changes to match that of the periodic stimulation and becomes phase locked to it [57] . SNr neurons were allowed to fire spontaneously and their spiking activity was recorded in the perforated patch current clamp configuration . The perforated patch configuration was chosen to allow us to see subthreshold responses and injected somatic currents , while avoiding the disruption of pacemaking activity by the disruption of the neurons intracellular milieu . A 30–60 second long , 10–20 pA cosine shaped current of frequency f was injected into the soma of spontaneously firing cells ( example in Fig 5A ) . We studied the effect of the oscillatory input on the cell’s firing pattern based on a previously published method [57] . For each spike , we measured the duration of the next perturbed period , denoted Tp , of the neuron’s spiking as a function of its effective phase , the phase ψ in the period of the sinusoidal input in which the spike occurred , and fit a periodic function to the effective phase comprised of 3 Fourier modes , resulting in the function Tp ( ψ ) ( Fig 5B ) . Given an initial action potential at the effective phase ψn , we can predict the effective phase ψn+1 of the next action potential using the following iterative map: ψn+1=ψn+Tp ( ψn ) Tmod1 [3] Therefore , the evolution of the phase of spiking relative to the oscillatory input can be represented using an iterative map . An intersection between the map and the diagonal ( where ψn+1 = ψn ) represents a fixed point of the dynamics . When the slope of the iterative map at the fixed point is smaller than unity , the fixed point is stable and the neuron is phase-locked with the stimulus . Thus , the sequence of phases visited by the neuron over a series of spikes can be predicted via this iterative map [57–59] . The fitted curve for the measurements of Tp as a function of ψ ( Fig 5B ) was used to generate an iterative map of effective phases ( Eq 3 ) . Note that the map closely traces the empirical scatter plot of ψn+1 versus ψn ( Fig 5C ) . In the example presented in Fig 5 , the cell was spiking spontaneously at a frequency f0≈7 spikes/s throughout the recording . When the stimulation frequency f is larger than the cell’s natural firing frequency f0 ( Fig 5C ) , the map passes mostly above the diagonal and does not transect it , which is consistent with the lack of phase locking in the raw data ( Fig 5A ) . The lack of phase locking is evident in the chaotic dynamics exhibited by the trajectory in the iterative map of effective phases ( Fig 5C ) , as well as in the wide probability distribution of effective phases covering all phases between 0 and 1 ( Fig 5D ) . When f < f0 ( Fig 5E ) , the map passes below the diagonal . Again , the neuron’s firing is not entrained by the periodic stimulation and the effective phase map gives rise to chaotic dynamics ( Fig 5G ) . Nevertheless , the distributions of phases in both cases ( Fig 5D–5H ) show an accumulation of phases in the area where the map passes closest to the diagonal ( Fig 5C–5G ) , demonstrating the effect of the sinusoidal input . When f ~ f0 , the firing of the cell is phase-locked to the stimulus at a phase close to 0 . 5 ( Fig 5I ) . In this case , the pacemaker only visits a limited range of effective phases . Therefore , we cannot fit a curve to Tp measurements ( Fig 5J ) and an iterative map ( Eq 3 ) cannot be estimated . Nevertheless , the entrainment of the neuron’s firing to the periodic input is clear from the narrow distribution of points in the scatter plot of ψn+1 versus ψn and their accumulation on the diagonal ( Fig 5K ) , as well as the sharp peaks around the phase 0 . 5 in the probability distribution of effective phases ( Fig 5L ) . Next , we incorporate optogenetic excitation into this scheme . The experiment above was repeated , but the temporal cosine waveform stimulation was applied optogenetically using the two spatial conditions of proximal ( Fig 6A ) and full-field ( Fig 6E ) illumination . Depending on the ratio between the stimulation frequency f and the natural firing frequency of the cell f0 , the stimulus either entrained the spiking activity of the cell , or was ineffective in causing entrainment . The efficacy in phase locking also differed somewhat between the two conditions , with full-field illumination inducing entrainment more readily . In order to examine the dendritic effect on entrainment , we focused on instances where phase locking was obtained . To obtain a deeper understanding of the effect of the portion of the dendritic arbor being activated on entrainment , we performed a simulation . When phase locking occurs , the pacemaker samples a narrow range of effective phases around the phase of entrainment , and it is thus impossible to estimate an empirical curve relating the perturbed period to the effective phase . However , the perturbed period can be described using the following theoretical equation Tp ( ψ ) =T0 ( 1−Δϕ ( ψ ) ) [4] Where T0 is the unperturbed period of the oscillating neuron , and Δϕ is the change in the intrinsic phase of the neuron induced by the stimulus . The evolution of intrinsic phase in response to a cosine shaped stimulation depends on the pacemaker’s PRC , denoted by Z , and is given by dϕdt=f0−Acos ( 2π[ft+ψ] ) Z ( ϕ ) [5] We evaluated the change in intrinsic phase Δϕ by numerically integrating Eq 5 until the intrinsic phase ϕ reaches 1 and measuring the resulting change Δϕ for various values of the effective phase ψ . PRCs were modeled as triangles peaking at θp = 0 . 9 and θf = 0 . 75 ( values similar to those extracted from PSTH estimates ) for proximal and full-field illumination , respectively . Both f and f0 were set to 7 Hz ( A = 5 ) . Simulated Δϕ values were plugged into Eq 4 , generating a numerical curve of Tp as a function of ψ predicted by the phase dynamics and the PRC ( Fig 6B–6F ) . This curve was then used to construct a map for the evolution of effective phases ( Eq 3 ) under each illumination condition ( Fig 6C–6G ) . Both simulated maps intersect with the diagonal achieving stable fixed points , which represent the effective phases of locking between the neuron’s spiking activity and the oscillatory stimulation . When a large portion of the dendritic field is stimulated , phase locking is predicted by the phase dynamics to occur at the phase 0 . 6 ( Fig 6G ) . In contrast , the predicted phase of locking corresponding to proximal stimulation is 0 . 537 ( Fig 6C ) . Thus , our simulation predicts that the effective phase of entrainment will occur slightly after the peak of the driving sinusoid and that this locking phase will be delayed when the entire dendritic arbor is stimulated compared to when only the soma and proximal dendrites are stimulated . This is consistent with the relative locations of the peaks in the probability distributions of empirical effective phases under the two spatial conditions ( Fig 6D–6H ) . This experiment was repeated using various stimulation frequencies ( 2 . 5-21Hz ) in several cells ( n = 6 ) . Oscillatory stimulation was delivered either optogenetically using proximal or full-field illumination , or as a somatic current injection . For each trial , a probability distribution of effective phases was generated ( Figs 6D–6H ) , and was used to calculate a circular variance vector [60] ( see Materials and Methods ) . The argument of this complex vector represents the phase of locking between firing activity and the periodic stimulus , and its amplitude expresses the strength of locking . Because we are interested in instances of entrainment , only trials with circular variance amplitudes above a threshold determined by bootstrapping were included in the analysis ( see Materials and Methods ) . While somatic current injection was more effective at inducing entrainment than proximal illumination ( Fig 6I ) , both stimulation conditions generated similar ranges of locking phases ( averaging at 0 . 555 and 0 . 587 for current injection and proximal illumination , respectively ) . This suggests that the proximal stimulation condition , in which only the soma and proximal dendrites are activated optogenetically , is comparable to somatic current injections . In contrast , the typical phase of locking in the full-field illumination condition ( averaging at 0 . 6585 ) is significantly delayed compared to the proximal illumination ( p = 9 . 2·10−4 , see Materials and Methods ) or somatic current injections ( p<<10−4 , see Materials and Methods ) . Thus , the effective phase of the entrainment ( shortly after the peak of the stimulation ) and the difference between locking phases corresponds to those predicted by our simulation ( Fig 6C–6G ) , which was based on the empirical PRCs for the two illumination conditions ( Fig 3D ) . The divergence of the PRCs , in turn , is explained by the electrotonic properties of the dendrites of the SNr GABAergic neurons ( Figs 2 & 3 ) .
To capture the entire dynamical repertoire of a neuron requires a detailed and high-dimensional model [61–63] . However , pacemaker neurons ( and their corresponding models ) offer an opportunity for simplification . Pacemakers traverse a low-dimensional closed contour called a limit cycle , so the high-dimensionality of the individual neuron can usually be shed in favor of a simpler dynamical system that captures the essence of the limit-cycle . The phase reduction method is a powerful formalism that does precisely that [19–22 , 64 , 65] . It reduces the pacemaker to a phase oscillator represented by its intrinsic frequency and its PRC , while maintaining the robustness and universality of pacemakers and oscillators . Pacemakers are attractive because the core of their dynamical properties can be measured and quantified in a single experiment . Many studies have capitalized on this and successfully measured empirical PRCs in order to characterize neuronal pacemakers and their collective dynamics [22–24 , 27 , 29 , 30] . However , reducing pacemakers to a phase oscillator with a single PRC , overlooks the contribution of their morphology to the phase dynamics . Previous studies investigated pacemaking using elaborate models that take into account detailed morphologies and ionic currents [34–37] . In the current study , we worked within the formalism of phase reduction theory , but still preserved the essential contribution of the dendritic arbor to the pacemaker’s dynamics . We combined previous results from cable theory [32 , 48 , 51] , phase reduction theory [19–22 , 64 , 65] , statistical physics [55 , 66 , 67] and nonlinear dynamical theory [57 , 59] to elaborate how the spatial extent of dendritic activation dictates the structure of the PRC and how that in turn characterizes the effect of dendrites on the population rate response and entrainment of pacemaker neurons . We aimed to generate the simplest model that would provide a plausible translation of the electrotonic length of the activated dendritic region to temporal delays . While more detailed models can certainly describe our empirical data , remarkably , the very simple cable model in combination with these powerful and generic theories gave rise to predictions that were then corroborated experimentally . The advantage of the phase reduction formalism is that it reduces high-dimensional oscillators to a one-dimensional limit-cycle represented by a single-phase variable . This formalism does not need to be abandoned once a complex dendritic structure is included . Rather , with a relatively simple extension of the formalism developed for point neurons , one can add the effects of dendrites . This may be valuable for neuroscientists who study and want to simulate large populations of neurons . With this formalism , they can relatively easily add the main impact of the dendrite due to its effective dendritic length , which , as we show is experimentally accessible . We began our combined theoretical and experimental study by generating a model that accurately fits the somatic current response to inputs arising from illumination of the soma and various portions of the dendritic field . Although our formalism can detect dendritic nonlinearities , our analysis showed that SNr GABAergic pacemakers and their dendrites are mostly linear under our stimulation conditions . This is supported by the good fit we obtained between the empirical data and the model under the assumption of a linear dendrite ( Fig 2D ) , as well as the fact that the effect was not dependent upon holding potential ( S1 Fig ) . However , it is well established that dendrites express various active conductances that can produce amplifying [68] or restorative [69] effects , and our findings cannot entirely rule out the contribution of these active dendritic conductances to the spatial effects that we detect . Our novel method generated measurable and physiologically meaningful spatial effects . Fitting the model to the data yielded an estimate of the space constant of SNr dendrites of 1 . 3 mm , which is of the correct order of magnitude and consistent with the diameter of the dendrites [51] . It would be important to compare this estimate to an estimate drawn from a model with a morphologically accurate dendritic arbor . However , our model is a vast simplification of the dendritic arbor and does not take into account the morphologies of individual cells . Moreover , this simplified model provides predictions regarding localized perturbations that could not be tested using our optogenetic stimulation method , which consisted of illuminating an entire portion of the dendritic field , beginning at and including the soma . The spatial effects that we detect using this method are robust but small . Interestingly , our theoretical model of the dendrite predicts that the spatial effect induced by a localized input would be significantly more dramatic ( Fig 1B ) . Taken together , this calls for a further investigation of the effect of the dendritic electrotonic structure using localized stimulation , which takes into account the structure of each specific dendritic tree . The stimulation conditions that we employ could also be physiologically relevant . Because GPe targets the proximal somatic region [42 , 43 , 51] , while the STN targets the entire dendritic tree uniformly [40 , 41] , the proximal illumination could reflect a coherent activation of afferent GPe inputs , while the full-field illumination could reflect the coherent activation of afferent STN inputs . Coherent rhythmic input could impinge on SNr neurons both under physiological conditions , such as sleep [49] or anesthesia [16 , 18] , and pathophysiological conditions , such as Parkinsonism [50] . Thus , the spatial effect that we demonstrate could provide insight into how rhythmic pallidal and STN inputs may interact to drive and entrain SNr neurons , and possibly drive pathological oscillatory activity in Parkinson’s disease ( PD ) [70] . In this pathological context , it may be important to consider that the oscillatory inputs from the GPe and STN to SNr may maintain a specific phase relationship because they are reciprocally connected [9 , 71 , 72] . An interesting future direction would thus be to further examine the physiological significance of our findings by optogenetically activating natural afferents to SNr cells arising from the GPe , the STN and the striatum both in vitro and in vivo . A fundamental question in the field of neuroscience is how the properties of the single cell manifest themselves in a network of neurons . Both theoretical and experimental work have demonstrated that the large-scale network dynamics of neurons often depend on a simple dynamical property or characterization of the individual neurons , rather than their high-dimensional and morphologically complex description . For example , a large body of work , conducted over the past decade by Wolf and collaborators , has shown that the high-fidelity of the population rate response of cortical neurons arises from and depends on the rapidity of action potential initiation . Because cortical neuronal dynamics are governed by a balance between excitation and inhibition that keeps these neurons near threshold , when an abrupt change occurs in their shared input , sufficiently many neurons can cross threshold and follow the input . Therefore , the only rate limiting factor , in these fluctuation-driven neurons , is how fast each one can generate an action potential [55 , 66 , 67] . This mechanism of stochastic resonance provides a good description of population responses in the cortex , the hippocampus and many other brain regions . However , in the BG , many neurons are not driven by noise but rather by their autonomous pacemaking . In this case , the framework for simplifying the description of neurons is the phase reduction method [19 , 21 , 64] where each neuron is characterized by its mean intrinsic rate and its PRC . Here too the high-fidelity of the population response can be derived from the PRC [54 , 55] . It turns out that the limiting factor on the fidelity of the response is the final descending part of the PRC . The initial rising phase of the PSTH , which is a reflection of the rapidity with which the population can respond , scales like the intrinsic period of the pacemaker and is roughly a mirror image of the final descending region of the PRC ( see Eq 3 ) . Here we provide an empirical demonstration of this previously known relationship , and show that a single global parameter—the effective electrotonic length of the dendrite ( ρ = R/λ ) being stimulated—can capture measurable dendritic effects on the PRCs and PSTHs of SNr pacemakers . Unlike in the case of fluctuation driven neurons , the pacemakers whose phase is closest to action potential threshold are the least responsive to input ( the PRC vanishes there ) and therefore cannot contribute to the population response . In order for populations of pacemakers to attain a high-fidelity representation of their input they need to discharge faster . In primates and humans , SNr GABAergic cells exhibit very high discharge rates of up to 145 spikes/s [73] , enabling SNr projection neurons to transmit information rapidly . According to our analysis , SNr GABAergic cells and their dendrites are linear and do not exhibit subthreshold resonances . However , they do display spiking resonances—the firing of SNr pacemakers can be entrained to rhythmic inputs , if the stimulation frequency is close enough to the neuron’s intrinsic frequency–its natural discharge rate . We demonstrated this first using periodic somatic current injection . We then incorporated optogenetic stimulation into this scheme and demonstrated that the tendency of the SNr cells to be entrained to rhythmic inputs is significantly affected by the dendrite , with inputs arriving at different locations inducing distinct phases of locking . The spiking resonance acts as a filter—signals containing frequencies near the natural discharge frequency of the neuron will be transmitted more effectively [57 , 74 , 75] . It is therefore possible that deep brain stimulation is only therapeutically effective at frequencies in the 120–140 Hz range [76] , because it needs to successfully entrain BG output neurons ( e . g . , SNr ) whose spiking resonances are in that range due to these pacemakers’ exceptionally high intrinsic firing rates [73] . The three stimulation methods that we employed—somatic current injection , proximal and full-field illumination , were not equally effective in entraining SNr neurons . Comparing the effects of current injection and optogenetic stimulation is complicated , as the two methods perturb the cell in inherently different ways . We deal with this issue by defining a bootstrapping based threshold ( see Materials and Methods ) and only considering instances where entrainment did occur , focusing on the dendritic effect on the phase of locking rather than the efficacy in entraining the cell . Interestingly , somatic current injection and proximal illumination induced similar phases of locking , while current injection was more similar to full-field illumination in the potency to entrain the spiking of SNr projection neurons . This strengthens the view that the two features—entrainment efficacy and phase of locking—are independent . A recent theoretical study has argued that the impedance load of the dendritic arbor should affect the rapidness of spike initiation [77] , which should invariably impact the response fidelity of fluctuation driven neurons such as pyramidal neurons . Similarly , but in the case of pacemakers , our study demonstrated that dendrites affect two aspects of the collective dynamics of SNr GABAergic pacemakers . First , we examined the fidelity of the population rate response and showed that the peak in the response induced by a stimulation of the entire dendritic arbor is delayed compared to stimulation of the soma and proximal dendrites . Next , we showed that when the firing of an autonomously active neuron is entrained by a periodic input , locking tends to occur at a later effective phase for activation of a larger portion of the dendritic arbor . Importantly , our study of the dendritic impact on currents arriving at the soma allowed us to relate–for the first time empirically–these population effects to the transmission properties of SNr dendrites . Because SNr neurons are actively decorrelated by recurrent connections [49 , 78] they are perfectly fit to function as a population readout of the integrated activity of the direct and indirect pathways . Thus , the rapidity and fidelity of transmission in SNr projection neurons are highly important . The intracellular response latency of SNr GABAergic cells is typically attributed to synaptic delays [5 , 12] . However , our results suggest that the delay in the response of a population of neurons can originate from additional sources . We show , as previously suggested [54 , 56] , that the fidelity of response in a population of oscillating neurons is determined by the shape of their PRCs , and that the PRC is in turn affected by the dendrite . The earlier peak in the response to proximal inputs , implies that inputs arising from the GPe , which impinge upon the soma and proximal dendrites , would generate a faster response than inputs originating from striatal spiny projection neurons and arriving at distal dendrites and terminal tufts . STN projections onto SNr neurons are distributed along the entire length of the dendrite , and would thus induce a delay that is longer than that generated by proximal GPe inputs , but considerably shorter than the delay in response to distal striatal inputs . The spatial organization of inputs can therefore have a significant effect on the latency in the response of a population of SNr neurons . SNr GABAergic projection neurons integrate inputs from the direct and indirect BG pathways , with afferents arising from striatal dSPNs , the GPe and the STN converging onto the same cell . GABAergic inputs from direct pathway dSPNs are thought to promote movement by inhibiting SNr neurons , while glutamatergic inputs transmitted by indirect pathway STN neurons block behavior by exciting the same neurons . GABAergic GPe inputs are inhibited by indirect pathway spiny projection neurons resulting in the disinhibition of SNr cells and the further depressing of movement . Thus , an input originating in the striatum and activating both pathways simultaneously generates a competition between ‘go’ and ‘no-go’ signals . However , because the ‘go’ signal is a monosynaptic input whereas the ‘no-go’ signal is a polysynaptic input , the ‘go’ signal would have an inherent advantage in the race to inhibit or entrain SNr neurons . Here we show that dendritic delays are manifested in the PSTH rise time in SNr pacemakers , and thus have a direct impact on the relative timing between the two competing signals . This may act as a compensation mechanism , effectively delaying direct pathway inputs originating from the striatum so that they impinge on the axosomatic region of the SNr neuron together with the indirect pathway inputs that coursed through the polysynaptic route via the STN or GPe , enabling a fair competition between ‘go’ and ‘no-go’ signals . A similar race is believed to underlie a core function of BG circuits—the decision to interrupt and cancel actions [79–81] . According to this view , the successful interruption of an action depends on the outcome of a competition between a dSPN ‘go’ signal and a PPN induced STN ‘stop’ signal . Evidence of this race was observed in the PSTHs of SNr and STN neurons , and suggests that the fast glutamatergic STN signal will fail in inhibiting the action , if the slower to rise GABAergic dSPN ‘go’ signal arrives at the SNr early enough to shunt away its effects [17] . This race occurs at a significantly slower time scale than the dendritic delays that we detect . However , a fast SNr response to STN inputs is critical for effective action inhibition . Our findings suggest that the spatial organization of inputs onto SNr neurons gives the uniformly distributed STN inputs an advantage over distal striatal inputs in quickly affecting SNr projection neurons . Conversely , because the ‘go’ signal arising from dSPNs in the race model peaks in the SNr PSTH at approximately 100 ms after the ‘stop’ cue , it does not require SNr neurons to respond particularly fast , which makes the distal location of dSPN synapses appropriate [17] .
Experimental procedures adhered to and received prior written approval from the Hebrew University Institutional Animal Care and Use Committee . Experiments were conducted with 3-12-week old male and female homozygous transgenic Thy1-ChR2 mice [B6 . Cg-Tg ( Thy1-COP4/EYFP ) 18Gfng/1] . These mice express ChR2 under the Thy1 promoter [46] in SNr GABAergic neurons . In SNr GABAergic cells , ChR2 is expressed in the soma as well as all parts of the dendritic field [47] . Mice were deeply anesthetized with ketamine ( 200 mg/kg ) –xylazine ( 23 . 32 mg/kg ) and perfused transcardially with ice-cold-modified artificial cerebrospinal fluid ( ACSF ) bubbled with 95% O2–5% CO2 , and containing ( in mM ) 2 . 5 KCl , 26 NaHCO3 , 1 . 25 Na2HPO4 , 0 . 5 CaCl2 , 10 MgSO4 , 0 . 4 ascorbic acid , 10 glucose and 210 sucrose . The brain was removed , and 240 μm thick sagittal slices containing the SNr were cut in ice-cold-modified ACSF . Slices were then submerged in ACSF , bubbled with 95% O2–5% CO2 , containing ( in mM ) 2 . 5 KCl , 126 NaCl , 26 NaHCO3 , 1 . 25 Na2HPO4 , 2 CaCl2 , 2 MgSO4 and 10 glucose , and stored at room temperature for at least 1 h prior to recording . The slices were transferred to the recording chamber mounted on a Zeiss Axioskop fixed-stage microscope and perfused with oxygenated ACSF at 31°C . To guarantee that the effects we measured were generated post-synaptically , the ACSF solution contained 10M 6 , 7-Dinitroquinoxaline-2 , 3-dione ( DNQX ) to block AMPA receptors , 50M D- ( - ) -2-Amino-5-phosphonopentanoic acid ( D-AP5 ) to block NMDA receptors , 10M 6-Imino-3- ( 4-methoxyphenyl ) -1 ( 6H ) -pyridazinebutanoic acid hydrobromide ( SR ) to block GABAA receptors , and 2M ( 2S ) -3-[[ ( 1S ) -1- ( 3 , 4-Dichlorophenyl ) ethyl]amino-2-hydroxypropyl] ( phenylmethyl ) phosphinic acid hydrochloride ( CGP ) to block GABAB receptors . An Olympus 60X , 1 . 0 NA water-immersion objective with a 26 . 5 mm field number ( FN ) was used to examine the slice using standard infrared differential interference contrast video microscopy . Patch pipette resistance was typically 4–5 MΩ when filled with recording solutions . The junction potential estimated at 7–8 mV was not corrected . For both whole-cell and cell-attached recordings the intracellular solution contained ( in mM ) 135 . 5 KCH3SO3 , 5 KCl , 2 . 5 NaCl , 5 Na-phosphocreatine , 10 HEPES , 0 . 2 EGTA , 0 . 21 Na2GTP , and 2 Mg1 . 5ATP , pH 7 . 3 with KOH ( 280–290 mOsm/kg ) . For perforated patch recordings 2g/ml of gramicidin B was added to the intracellular solution . In whole-cell voltage clamp recordings , cells were held between -70 to -80mV to avoid spiking , or at -50mV and -60mV to search for possible voltage dependencies in the subthreshold range . In cell-attached and perforated patch current clamp recordings , neurons discharged spontaneously , and figures depict the temporal derivative of the voltage . Electrophysiological recordings were obtained with a MultiClamp 700B amplifier ( Molecular Devices , Sunnyvale , CA ) . Signals were filtered at 10 kHz online , digitized at 20 kHz and logged onto a personal computer with the Signal 6 software ( Cambridge Electronic Design , Cambridge , UK ) . Optogenetic stimulation was achieved with blue-light ( 470 nm ) LED illumination via the objective ( Mightex , Toronto , ON , Canada ) . Field illumination was applied under two conditions , in all experiments conducted in this study: a ) proximal stimulation illuminating a ~130 μm diameter around the soma ( achieved by placing an opaque disk with a central pinhole at the 60X objective’s back focal plane ) , thereby targeting the soma and proximal dendrites ( proximal ) ; and b ) full-field illumination stimulating the entire SNr ( and beyond ) with a 5X objective which excites the soma and the entire dendritic field ( full-field , see S3 Fig ) . Given the small size of SNr GABAergic cell somata reported in the Results , both illumination conditions include dendritic activation . In all experiments , LED light intensity at the back plane of the objective was chosen such that stimulation generated comparable current and voltage responses for the two conditions . Because the opaque disk with the pinhole blocks out ~85% of the illuminated area , significantly higher LED intensities were required to induce comparable somatic responses in the proximal stimulation condition . For PRC measurements , cells were recorded in the cell-attached , current clamp mode ( to preclude any feedback from the amplifier that can alter the cells’ firing pattern ) . Barrages of light pulses were delivered based on a previously described method [47] . Pulses were 0 . 5–1 ms long and were separated by random , exponentially distributed inter-pulse intervals , with means of 6 ms and 2 . 17 ms ( from pulse onset until the onset of the next pulse ) for the full-field and proximal illumination conditions , respectively . LED light intensity at the back plane of the objective was 15 mW for proximal illumination through the disk with the pinhole and 0 . 03–0 . 06 mW for full-field illumination . Stimulation consisted of 3 seconds of baseline recording followed by 9 seconds of barrage stimulation , and this was repeated 25 times . Each of the 25 repetitions was a different realization of exponentially distributed inter-pulse intervals , but all cells received the same set of stimulation barrages . This stimulus was also used for measurements of ChR2 current responses . To avoid the effect of ChR2 deactivation , only data collected during the last 5 seconds of stimulation were used in the analysis . In experiments probing the dendrites’ electrotonic properties , cells were stimulated with a temporal sinusoid at different frequencies ( 0 . 25–16 Hz , 3–4 seconds per frequency ) . LED light intensity at the back plane of the objective was 1 . 5mW for proximal illumination and 1 . 5–15 μW for full-field illumination . In entrainment experiments , cells were recorded in the perforated patch configuration and allowed to fire spontaneously as 30-60-second-long sinusoidal waveforms at different frequencies around the natural firing frequency of the cell ( typically ~7 Hz ) were delivered . Stimulation was applied either optogenetically , in the proximal or full-field illumination configurations , or as a current injection with a 10–20 pA amplitude . LED light intensity at the back plane of the objective was 0 . 15–0 . 9 mW for proximal illumination and 15–30 μW for full-field illumination . Neurons were patch clamped with 100 μm Alexa Fluor 568 ( Molecular Devices ) in the patch pipette . The two-photon excitation source was a Chameleon Vision II tunable Ti:Sapphire pulsed laser ( Coherent , Santa Clara , CA , USA ) tuned to 820 nm . The images were collected with the Femto2D system ( Femtonics , Budapest , Hungary ) which includes two 3 mm galvo-scanners and a multi-alkaline non-descanned photomultiplier tube for imaging Alexa Fluor . Z stacks of optical sections ( spaced 2 μm apart ) were collected using 0 . 2 μm pixels and 15 μs dwell times . The image in Fig 2A is a montage of several collapsed Z stacks . Measurements of the long axis of SNr GABAergic cells somata were based on these images ( n = 15 ) . Data were analyzed and curve fitting was performed using custom made code in MATLAB ( MathWorks , Natick , MA , USA ) . | The substantia nigra pars reticulata ( SNr ) is a main output nucleus of the basal ganglia ( BG ) , where inputs from the competing direct and indirect pathways converge onto the same neurons . Interestingly , these inputs are differentially distributed with direct and indirect pathway projections arriving at distal and proximal regions of the dendritic arbor , respectively . We employ a novel method combining theory with electrophysiological experiments and optogenetics to study the distinct effects of inputs arriving at different locations along the dendrite . Our approach represents a useful compromise between complexity and reduction in modelling . Our work addresses the question of high fidelity encoding of inputs by networks of neurons in the new context of pacemaking neurons , which are driven to fire by their intrinsic dynamics rather than by a network state . We provide the first empirical demonstration that dendritic delays can introduce latencies in the responses of a population of neurons that are commensurate with synaptic delays , suggesting a new role for SNr dendrites with implications for BG function . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"optogenetics",
"engineering",
"and",
"technology",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"probability",
"distri... | 2019 | Population dynamics and entrainment of basal ganglia pacemakers are shaped by their dendritic arbors |
We develop a unified model accounting simultaneously for the contrast invariance of the width of the orientation tuning curves ( OT ) and for the sigmoidal shape of the contrast response function ( CRF ) of neurons in the primary visual cortex ( V1 ) . We determine analytically the conditions for the structure of the afferent LGN and recurrent V1 inputs that lead to these properties for a hypercolumn composed of rate based neurons with a power-law transfer function . We investigate what are the relative contributions of single neuron and network properties in shaping the OT and the CRF . We test these results with numerical simulations of a network of conductance-based model ( CBM ) neurons and we demonstrate that they are valid and more robust here than in the rate model . The results indicate that because of the acceleration in the transfer function , described here by a power-law , the orientation tuning curves of V1 neurons are more tuned , and their CRF is steeper than those of their inputs . Last , we show that it is possible to account for the diversity in the measured CRFs by introducing heterogeneities either in single neuron properties or in the input to the neurons . We show how correlations among the parameters that characterize the CRF depend on these sources of heterogeneities . Comparison with experimental data suggests that both sources contribute nearly equally to the diversity of CRF shapes observed in V1 neurons .
The dependence of the neuronal response amplitude on stimulus contrast , the contrast-response function ( CRF ) , typically displays a sigmoidal shape in the visual cortex: it accelerates at low contrast and saturates at high contrast [1]–[8] . This major nonlinearity appears to be accentuated in cortex , as ganglion cells in the retina and relay cells in the LGN saturate at higher contrast and show shallower slopes [3] , [7] , [9]–[15] . In the extreme , some parvocellular neurons in primate LGN display a quasi-linear contrast-response function [13] , [14] , [16] . A large fraction of neurons in the primary visual cortex ( V1 ) respond in a manner that is selective to the stimulus orientation [17] , [18] . The dependence of the spike rate on stimulus orientation ( the orientation tuning curve ) is well described by a Gaussian whose amplitude varies substantially with stimulus contrast . Although this might be less true in primate [19] , [20] , it has been shown in carnivore and rodents that the width of the tuning curves does not change when the contrast is modified [2] , [7] , [15] , [21]–[25] This property is referred to as “contrast invariance” of orientation tuning . The membrane potential response of cortical neurons displays an orientation tuning width that is typically 1 . 5 times larger than that of the spiking response [15] , [26]–[30] . This tuning width is also contrast invariant [23] . These contrast-invariant properties constitute strong constraints for understanding the mechanisms underlying the response of V1 neurons to visual stimuli . The models that have been proposed to explain orientation selectivity in V1 can be broadly classified in two groups ( reviewed in [31] , [32] ) : feedforward models , in which orientation selectivity emerges mainly from the spatial arrangement of ON and OFF receptive fields of the LGN cells that form the input to V1 neurons , and recurrent network models in which the orientation selectivity emerges mainly from the recurrent connectivity within V1 . Both classes of models have limitations . Although recurrent models can account for contrast invariance in the spiking response [33]–[37] , they appear incompatible with the fact that V1 recurrent inputs seem to have , at best , a very weak effect on the voltage tuning width [38] , [39] . Recurrent models have further been questioned given the peculiar responses they generate in the presence of pairs of oriented contours [40] and given strong interaction between spatial frequency and orientation selectivity [36] . Furthermore , in contradiction to the experimental results , the response of neurons in such models either display contrast invariance of orientation selectivity , or CRFs saturation , but not both simultaneously [41] , [42] . The feedforward model , in its original formulation [17] , cannot account simultaneously for the fact that orientation tuning of the spike response is sharper than the tuning of the voltage and for the contrast invariance of the spike response tuning width . Nevertheless , including feedforward anti-phase inhibition [43] or broadly tuned inhibition [44] , [45] in the feedforward model permits contrast-invariance of orientation tuning in the membrane potential response . Anderson et al . [23] further showed that contrast invariance of orientation tuning for the spiking response , in addition to that of the membrane potential response , can be achieved in the feedforward model if membrane potential fluctuations ( “synaptic noise” ) are taken into account . This is because these fluctuations smooth the threshold non-linearity [15] , [46] , [47] . This smoothing effectively transforms the transfer function of the neurons to a power-law voltage-rate relationship . This is exactly what is needed to obtain contrast invariance for both the voltage and the firing rate , provided membrane potential fluctuations amplitude scales with contrast [15] . However , the feedforward model may account for sigmoidal CRFs only if the LGN input saturates sufficiently strongly . Yet the contrast at which saturation occurs in V1 is lower than for the LGN input . This implies that additional mechanisms are required to account for the co-occurrence of contrast-invariant orientation tuning and of CRFs typical of V1 neurons . Some of the models used to examine the mechanisms responsible for the saturation of the CRF also display contrast-invariant stimulus selectivity . In the “normalization model” [48]–[51] , saturation results from feedback shunting inhibition from a pool of inhibitory neurons . Because this pooling includes inhibitory neurons with a wide range of preferences , this model also accounts for cross-orientation inhibition as well as for contrast-invariance of orientation tuning . However , this model has been questioned due to membrane time constants requirements [32] . Alternatively , synaptic depression has been proposed as one mechanism to explain saturation at high contrast [52] , [53] . In these models however , contrast-invariance of orientation tuning does not depend on synaptic depression but depends on the push-pull arrangement of inhibition and excitation , as in the model proposed by Troyer et al . [43] . In another recent model , Banitt et al . [54] examined how contrast-invariance of orientation tuning may depend on thalamocortical synaptic depression , but they did not explore the mechanisms underlying contrast saturation . Models based on synaptic depression are able to explain not only the static properties of the behavior of V1 neurons , but also dynamical aspects , such as contrast-dependent phase advances and frequency-dependent contrast saturation . Nevertheless , recent experimental studies showed that synaptic depression in the thalamocortical pathway may be rather weak in vivo , especially in the presence of spontaneous activity that generates a steady state of synaptic depression [55]–[57] . Thus the question is: can one formulate , without resorting to synaptic depression , a model in which cortical neurons display contrast-invariant tuning-width for both membrane potential and spike responses , as in the feedforward model in the presence of synaptic noise , while at the same time intracortical interactions induce a saturation of the CRF of cortical neurons at lower contrast than their LGN afferents ? To examine this question , we investigated a rate model of a hypercolumn in the visual cortex with neurons whose transfer function nonlinearity was described by a power-law . This allowed us to find conditions for getting both a sigmoidally shaped CRF and contrast invariant orientation tuning width when both feedforward and feedback inputs were included . We then tested whether our results hold in a less idealized network model made of conductance-based ( CBM ) neurons . Using numerical simulations in this later model , we investigated the robustness of the results obtained in our rate model . We analyzed the respective contributions of the feedforward input , of the recurrent intra-cortical input , and of neuronal intrinsic properties in shaping the CRF . In particular , we studied the differences between inhibitory and excitatory neurons , and how these relate to differences in their intrinsic properties . Finally , we explored possible explanations for the broad diversity of CRFs shapes observed in V1 neurons: although typically sigmoidal , CRFs are characterized by parameter values that vary widely at the population level [1] , [3] , [8] . For this purpose , we compared the predictions from our model with experimental data obtained in area V1 of the marmoset monkey . Our results suggest that substantial heterogeneities in the intrinsic properties of the neurons as well as heterogeneities in the CRFs of LGN neurons are required to account for the diversity of CRFs shapes observed in the primary visual cortex . Part of this work has been presented at the 34th and 36th annual meeting of the Society for Neuroscience ( San-Diego , Oct 2004; Atlanta , Oct 2006 ) .
In this section , we investigate to what extent the results we have obtained in our simplified rate model still hold in a more realistic conductance-based model , in which neuronal dynamics is governed by voltage-dependent conductance channels and synaptic interactions are mediated by conductance changes ( see Eqn . 23 ) . We also investigate in this framework how far diversity in the intrinsic cell properties or in the connectivity can account for the heterogeneity in the CRFs observed experimentally ( present study and [1]–[6] , [8] ) . We simulated a network model of V1 made of these conductance-based neurons . The effect of a visual stimulus is modeled by adding an input to the neurons . We take the connection widths such that they satisfy the condition: ( see Eqns . ( 7 , 8 ) ) where are given by the best fit of the f-I curves ( see above and Fig . 5 ) . The maximal LGN input , , depends on the contrast : ( 11 ) The width of the LGN input to the excitatory and inhibitory populations is and respectively . Fig . 6 shows the orientation tuning curves for the firing rate and average voltage of both neuron types . The firing rate tuning curve is well fitted by a Gaussian for both types of neurons . The width of the optimal Gaussian changes by less than 10% when the contrast increases from 1 to 64% . For this contrast range , the effective leak conductance , , increases from 0 to 0 . 19 mS/cm for the excitatory neurons and from 0 to 0 . 13 mS/cm for the inhibitory ones . We have also plotted in Fig . 6 the predictions given by the effective rate model . For the excitatory population , the simulations results differ substantially from the prediction of the rate model; the rate model underestimates the peak of the tuning curve of the excitatory neurons by as much as 30% ( Fig . 6A ) . The discrepancy is less substantial for the inhibitory population ( Fig . 6B ) . It may be surprising that the discrepancy is larger for the excitatory neurons than for the inhibitory neurons , whereas the deviations from a power-law in Fig . 5 is bigger for the former than for the later . However , this can be explained as follows . According to Fig . 5 , the inhibitory rate should be lower in the spiking network than in the effective rate model . This , however , also decreases the inhibitory feedback to the population . This decreased inhibitory feedback cancels the effect of the deviation from power law of the f-I curve to a large extent . For the excitatory neurons the fit to a power-law is good for the whole input range , but the population also receives less inhibitory feedback than predicted from the effective rate model . This leads to a substantial increase in the firing rate of the excitatory neurons , compared to what one would expect from the effective rate model ( Fig . 6A ) . In contrast to the height of the tuning curves , there is surprisingly little discrepancy between the numerical simulations and the predictions of the effective rate model for what concerns the width of the tuning curves . This also stems from the corrective effect of the inhibitory feedback . The inhibitory feedback to the inhibitory populations suppresses the broadening of the output tuning curve implied by the deviation of the power-law . As a result , the width of the inhibitory feedback to the excitatory cells is close to that predicted by the effective rate model . Hence the excitatory tuning width is also close to the predicted one . Because in the CBM the average voltage varies almost linearly with the input , the tuning curve of the voltage follows the tuning curve of the net input . Since the latter is close to a Gaussian with a contrast independent width , the voltage tuning curves are well approximated by Gaussians and have a close to contrast-invariant tuning width , as shown in Fig . 6C , D . Note that voltage tuning width is substantially broader than the tuning width of the spike response . Fits to the H-ratio function of the CRFs of V1 neurons reveal a large diversity in the parameters , and [1]–[6] , [8] . Can this diversity be accounted for in the framework of our model ?
All the findings of the present paper rely on the fact that , in the presence of noise , the effective input-output transfer function is accelerating and can be fitted by a power-law over the physiological range of neuronal responses to visual stimuli [15] , [23] , [46] , [47] , [68] , [69] . The noise in the input influences the neuron's transfer function by effectively smoothing the effect of the spiking threshold . The mean input current and voltage are also non-linearly related , such that the rate-voltage transfer function is well fitted by a power-law , but with an exponent that is larger than the one of the rate-current transfer function . In the present model , the exponent , , of the input-output transfer functions of the neurons must be larger than 1 to insure spike tuning curves sharper than the tuning curves of the LGN input . For neurons in vivo , the transfer function for voltage vs . firing rate is well approximated by a power-law , with an exponent , ranging between 2 and 5 [15] , [23] , [68] . Under the assumption that the input noise is on the same order for different neuron types , the input-output transfer function of our model inhibitory neurons accelerate more than that of excitatory neurons . This is because inhibitory neurons have higher gain and show less firing rate adaptation ( e . g . , [63] , [70] ) . Thus , the fit of the spiking rate to a power-law reveals different exponents , for the different neuron types in our model . That the exponent tends to be higher in the inhibitory cells than in the excitatory ones has been reported in recent experimental studies [69] . A major difference between the rate model and the conductance-based model is that , in the later , synaptic inputs increase the effective leak conductance , an effect that was not taken into account in the former . Nevertheless , we have shown here that an increase , , of the leak conductance , if not too large ( increasing the effective up a factor of 2 ) has the same effect on the transfer function as an additional negative current , . This current is proportional to , . This is similar to what was found by Shriki et al . [59] for the transfer of conductance based neurons in the absence of noise . As we have shown , this allows for the derivation of an effective rate model , which replicates the steady state behavior of the CBM . Noise , as inferred from voltage traces , has been reported to be independent of stimuli contrast and orientation [23] ( but see [15] , [71] ) . Such a noise in the input current effectively results in a power law transfer function [46] , [47] . It has been shown that , in the absence of recurrent cortical interactions and with feedforward inputs alone , the power-law transfer function leads to an approximate contrast invariance of the orientation tuning curve width [46] , given contrast invariant input width , as they emerge from the spatial arrangement of LGN ON and OFF cells [17] , [61] . Due to the nonlinearity of the transfer function the outputs are more tuned than the inputs by the factor . Here we extended these results to take into account recurrent cortical interactions . We showed that they remain true provided that the synaptic distributions have an appropriate spatial extent , namely that the conditions expressed by Eqns . ( 7 , 8 ) are satisfied . When the conditions for the width of the feedback , expressed by Eqns . ( 7 , 8 ) are satisfied , the feedback interactions do not contribute to the sharpening of the tuning . The latter is determined by the tuning of the LGN input , together with the sharpening effect of the power-law transfer function . This is in sharp contrast to the role of recurrent interactions in network models of V1 studied previously [33]–[37] , [72] . Recurrent interactions , however , appear essential for explaining the shape of the CRFs ( see below ) . In the absence of recurrent interactions , the CRF of the cortical neurons is shifted toward higher contrast compared to the CRF of their feedforward inputs . This means that to achieve a reasonably large response at low contrast the parameter of the LGN input must be quite large . This implies that , at maximum contrast , the response of the cortical neurons is large too . However , beyond a critical value , the response amplitude would fall in a range where the transfer function of the neurons deviates substantially from a power-law . In our conductance-based model , this deviation becomes appreciable above . In turn , this deviation from power-law implies substantial deviations from contrast-invariance of the tuning-width at high contrast . Therefore , the strong inhibitory feedback in the recurrent network model we have studied plays a crucial role , which is to regulate the high contrast responses , relative to the responses at intermediate and low contrast . As a result , both feedforward and excitatory recurrent inputs can be relatively strong , resulting in a consistent response for both low and intermediate contrast , yet the response at high contrast does not reach values beyond which contrast invariance is lost . We have demonstrated this role in our conductance-based model . The saturation due to the feedback which keeps the response within the power-law range for high contrast also causes a decrease of the and an increase in the slope of the CRF relative to the LGN input . We have modeled the LGN input as a Gaussian , with a width that is independent of contrast . This represents a simplification , which is nevertheless justified given previous theoretical studies on contrast invariance of orientation tuning in simple cells . A well known problem in this context [31] , [43] is that , in simple cells , the LGN input generates an untuned DC component in the membrane potential response , which grows faster with contrast that the tuned AC component . A solution to this problem consists in canceling this DC component by including either anti-phase or broadly tuned inhibition in the models [43] , [45] , [72] . This was not explicitly incorporated in our model . We rather simplified it with a tuned LGN input that one should view as a net input into the cells which combines both the actual LGN input and the feedforward inhibition . The conditions expressed by Eqns . ( 7 , 8 ) imply specific range for the synaptic connections between sub-populations of neurons . They show that , if the orientation tuning width of inhibitory neurons is broader than that of excitatory neurons as reported experimentally [28] , [30] the synaptic projection from inhibitory to excitatory neurons should be narrower than the projection width from excitatory to excitatory neurons . This is compatible with anatomical data , which show that the spatial extent of inhibitory connections is usually less than that of excitatory connections [73] , [74] . Note that these conditions were obtained under the assumptions of Gaussian inputs and outputs , which are in line with experimental data ( e . g . , [75] ) . Here an important caveat should be made . We showed that contrast invariance of the tuning width is robust to violations of conditions Eqns . ( 7 , 8 ) . If the range of the synaptic feedback , both excitatory and inhibitory , is changed by as much as 50% , contrast invariance is still nearly achieved with a relative error of less than 10% . Thus the model predictions about the relative extent of the excitatory and inhibitory feedback should not necessarily be taken as quantitative . The parameters we used generated relatively narrow tuning curves ( see Results ) , in accordance with the tuning width reported for layer 4 simple cells in some studies ( e . g . , [15] , [69] ) . However , others studies reported a large heterogeneity of tuning width , including broadly tuned cells and cells showing a non-negligible response at the orthogonal orientation [7] , [19] , [20] , [28] , [30] , [76] . We therefore checked whether our results were valid for parameter regimes different from the one we initially used . We simulated networks with broader tuning curves ( ) , for which the response at the orthogonal orientation was approximately one tenth of that at the preferred orientation . For such networks , we found that the orientation tuning width did not change significantly with contrast . However , the ratio of the response at the orthogonal orientation versus the preferred orientation decreased slightly with contrast . Interestingly , this departure from strict contrast-invariant orientation tuning has been observed experimentally for broadly tuned cells in some studies [7] , [19]; but see [20] . However , this should not be taken too seriously because , as Fig . 8 shows , deviations from Eqns . ( 7 , 8 ) for the feedback width can have a substantial effect on the response at the orthogonal orientation , which could result in the reverse effect . The CRF of the spike response can be well fitted by an H-ratio function in a large fraction of V1 neurons . However , the parameters of the function are highly diverse across neurons [1]–[4] , [7] . Most studies that aim to explain contrast invariance or the shape of the CRF ignore this heterogeneity and usually do not indicate whether the proposed mechanism can accommodate a large diversity of responses . Whether the excitatory neurons saturate or not is determined by the strength of the feedback connections , particularly from the inhibitory cells . This implies that some degree of fine-tuning of these strengths is necessary if we impose that the average excitatory CRF saturates at 100% contrast . Because of this sensitivity , relatively small variability in the feedback strengths for individual neurons leads to rather large changes in the CRFs . This can contribute to the large variability in CRFs , with non-saturating , saturating and super-saturating cells observed in the primary visual cortex of the same animal . Here we have investigated other possible sources for this diversity , focusing on the contribution of variability in single neuron intrinsic properties , and on the contribution of heterogeneities in the CRFs of LGN neurons . We have demonstrated that these two sources of variability can both account for the diversity observed in experiments . In addition , our model predicts a correlation between the parameters and , which is either negative or positive , depending on the source of heterogeneities . The strength of the correlation is further predicted to be reduced when both sources are mixed , in proportion to the relative contribution of each . We examined CRFs for neurons in the primary visual cortex of marmoset monkeys . The parameters and obtained in these experimental data were at best weakly negatively correlated . This suggests that heterogeneity in the LGN input may contribute slightly more than the neurons' intrinsic properties to the diversity of CRFs shape . Another possible source of heterogeneity we did not examine is heterogeneity in the recurrent feedback inputs . We assumed that these are uncorrelated . Then , given their large number comparatively to LGN inputs , heterogeneities in feedback inputs would cancel each others and this would result in an “averaged” CRF input to all neurons . However , some studies showed that subset of excitatory and inhibitory neurons may form specific connections with other neurons [77]–[80] , and in many cases the connections are not reciprocal . This would lead to heterogeneity in the feedback input , that we expect to have the same effect on the correlations between and as the diversity in the feedforward input . Other studies [81] , [82] , however , suggest that inhibitory fast spiking cells establish a dense network with other neurons , as assumed in the present study . Two major weaknesses of our model is that we have to add external noise to the system to obtain voltage fluctuations that are biologically plausible and that it does not exhibit heterogeneity in the orientation tuning curves . One way to obtain input fluctuations intrinsically is to use a model that operates in the balanced regime [83] , [84] . In this regime , heterogeneity in the response naturally arises from the strongly amplified effect of randomness in the connectivity . However , in their current formulation , balanced network models cannot explain the shape of the CRF as observed experimentally . This is because in such networks the population averaged response should scale linearly with the external input [83] , [84] , so that on average the of both the excitatory and inhibitory populations should be the same as the of the LGN input , in contrast to what is observed experimentally . It is our hope that development of such models , in which recurrent connections are responsible for the synaptic noise which is so essential to contrast-invariance of tuning width , will help further integration of feedforward and feedback models for a better understanding of the mechanisms at work in cortical processing .
The protocol for the experiments which are reported here is in accordance with guidelines of the French ministry of agriculture ( décret 87/848 ) and the European Union ( directive 87/609 ) . Our rate model consists of excitatory and inhibitory neurons . The firing rate of excitatory neuron and inhibitory neuron , denoted by and respectively satisfy ( 12 ) where is the membrane time constant for population , is the total , noise averaged , input into the neuron , and is the effective , noise averaged , transfer function . Following recent experiments [23] , [68] and theoretical studies [46] , [47] , we assume that the transfer function is a threshold power-law function , . Here denotes the half rectified linear function , for and for . The exponent of the power law function is and sets its scale . Our model network represent a hypercolumn in V1 and has the geometry of a ring [33] . Neuron in population is characterized by an angle , defined as the orientation of the visual stimulus for which the LGN input it receives is maximum . We model this input as ( 13 ) where is the orientation of the stimulus , is the -periodic Gaussian with width , defined as . gives the overall strength of the LGN input and depends on the stimulus contrast . As we will see , for , not only the LGN input to neuron is maximum but so is also of its spike response . Therefore , is also the preferred orientation of the neuron . We assume that varies with the contrast , , of the visual stimulus as where is in percents . This logarithmic dependence , which does not saturate , was chosen to facilitate the analysis of the cortical network . The preferred orientations of the neurons are uniformly distributed over the interval . The feedback input from the network to neuron , , is given by ( 14 ) where the synaptic strengths , , depend on the difference in preferred orientations between neurons and and falls off with this difference as a periodic Gaussian with width ( 15 ) Note that we have scaled the synaptic strength by the density of neurons . The number of neurons in population with preferred orientation between and is equal to , which explains the factor in Eqn . ( 14 ) . In the limit of large , we can replace by , and by , where is a continuous variable . The rates satisfy the dynamics ( 16 ) Due to the rotation symmetry of the network , the response of the neurons depends on the stimulus orientation only through the difference , , between this orientation and the neurons preferred orientation , . Thus we need only to consider the case where . In the conductance-based network , neurons are point-like and the dynamics of their membrane potential , , is: ( 21 ) where . The first term on the right-hand side of Eqn . ( 21 ) is the leak current . The next five terms correspond to a sodium current , , a delayed rectifier potassium current , , responsible for the up and down-stroke of the action potential respectively , a slow potassium current , , inducing spike adaptation , an A-type potassium current , , which becomes active during the hyper-polarization period and affects the length of the inter-spike interval , and a persistent sodium current , , which tends to amplify small depolarizations . The gating variables , , , follow the dynamics: ( 22 ) For , the functions and , with the parameters and as given in Table 1 and for the functions and are given in Table 2 . The maximal conductances of the ionic channels of the excitatory and inhibitory neurons are given in Table 3 . They are chosen to reproduce qualitatively the frequency-current transfer functions of regular spiking excitatory neurons and fast spiking inhibitory neurons , such that excitatory neurons have a lower threshold [85] and stronger spike frequency adaptation than inhibitory neurons ( e . g . , [63] , [70] ) . The terms left on the right-hand side of Eqn . ( 21 ) are the synaptic inputs , , the neuron receives because its recurrent interactions with the other neurons in the network , a current , , representing the feedforward inputs from the LGN to V1 , and the noise . The synaptic current received by neuron in population , is ( 23 ) where mV and mV are the reversal potentials of excitatory and inhibitory synapses respectively . The strength of a synapse connecting the presynaptic neuron in population , to postsynaptic neuron in population , is characterized by , where is given by Table 4 . Note the normalization to the neuronal density . The term describes the contribution of the th spike of neuron in population , which occurred at time , to the synaptic conductance at time . We take ( 24 ) with rise time constant msec and decay time constant msec for excitatory as well as for inhibitory synapses . The current , , is a Gaussian white noise with zero mean . Its standard deviation , , is chosen such that the standard deviation of the membrane potential of the neurons is approximately 3–4 mV , as measured experimentally in V1 [23] . The LGN input is modeled as in Eqn . ( 13 ) with , where the values of and are taken in accordance with experimental data for magnocellular cells [3] and is such that the activity of the neurons are similar to those measured in V1 during visual stimulation [3] , [6] . In the rate model we simulated networks with 100 neurons for each of the populations , using a second order Runge Kutta integration scheme with a time step of 1 msec . After verifying that this discretization was sufficiently fine , we used these simulations to find the fixed points in the rate equations and to verify the stability of steady state . The conductance-based model dynamics of networks consisting of 400 excitatory and 400 inhibitory neurons was simulated using a second order Runge-Kutta integration scheme with a time step msec . For each contrast , ten trials with different noise realizations were simulated and the responses were averaged over a time window of 1 . 5 sec after elimination of a transient . The orientation tuning curves of the neurons were fitted with Gaussians parametrized as: . For the rate model , we set the offset , , to zero . For the CBM , was in general non-zero because the noise induced a non-zero activity at cross-orientation . The peak amplitude of these Gaussians estimated for different contrast , , yielded the CRFs of the neurons , which were subsequently fitted with the H-ratio function [1]: ( 27 ) where is the maximum firing rate , is the contrast ( in ) for and the exponent , , is a measurement of the function's steepness . In the case of the CBM , we additionally computed the relative error of the estimated values of the CRF parameters ( the relative error on is its SD divided by its mean , ) . Good fits were defined as those with relative errors smaller than 0 . 15 for all the parameters . Experimental data for the CRF was obtained from marmoset monkeys ( Callithrix Jacchus , ) . Details about the experimental protocol can be found in [20] . One half hour before anesthesia induction , the animals were tranquilized with diazepam ( Valium , Roche ) ( i . m . , 3 mg/kg ) and atropine ( 0 . 05 mg/kg ) was given at the same time to reduce secretions and to prevent bradycardia . Anesthesia was induced with Alphadalone/Alphaxalone acetate ( Saffan , Essex Pharma , 1 . 2 ml/kg ) injected intramuscularly and maintained during surgery by i . v . injection ( 0 . 17 ml/kg every 10–15 minutes ) . Synthetic corticoids were given to prevent brain edema . Animal's body temperature was maintained at C using a heating pad controlled by a rectal thermistor . EKG recording was performed through metallic pliers . The surgical procedure consisted first in placing a catheter in the femoral vein . Next , a tracheotomy was performed to allow artificial ventilation . The marmoset was then set in a stereotaxic frame . Two holes were drilled over the frontal cortex and Ag wires inserted for epidural EEG recording . A craniotomy was made to gain access to area V1 . A head post was sealed to the skull and fixed to the stereotaxic apparatus . Following surgery , the animal was artificially ventilated with / ( 50%/50% ) . Anesthesia and analgesia were supplemented by a continuous infusion of sufentanil citrate ( Sufenta , Janssen , 4–6 g/kg/hr ) after a loading dose of 1 g/kg . The infusion vehicle was made of the mixture of 2 ml glucose 30% , 15 ml of amino-acid perfusion solution ( Totamin , Baxter ) and included synthetic corticoids ( 0 . 4 mg/kg/hr ) ; NaCl was added to a final volume of 50 ml . We waited for 1–2 hours of infusion with this solution to ensure adequate depth of anesthesia . The animal was then paralyzed by adding pancuronium bromide ( Pavulon , Organon , 0 . 1 mg/kg/hr ) to the solution described above . Mydriasis and cycloplegia were induced with ophthalmic atropine sulfate ( 1% , Alcon ) . Gas permeable contact lenses were used to protect the eyes . The heart rate , rectal temperature and expiratory concentration were monitored throughout the experiment and maintained at 250–350 bpm , 37–C and 3–5% , respectively . The EEG and the absence of reaction to noxious stimuli were regularly checked . Action-potentials were recorded extracellularly in area V1 using tungsten-in-glass microelectrodes . Spike-sorting was performed using Spike2 ( Cambridge Electronic Design , Cambridge , UK ) system . Appropriateness of single-unit isolation was based on the refractory period of the neuron . Visual stimuli were presented onto a computer monitor placed at 114 cm from the animal's eyes . We first determined the preferred orientation using square-wave drifting gratings . Optimal spatial frequency was then determined using sine-wave drifting grating . The CRF was then established using sine-wave drifting grating with optimal orientation and spatial frequency , presented at 12 different levels of contrast increasing geometrically for 2 to 90% . All visual stimuli were presented in a circular patch of 2–6 degrees diameter , centered on the receptive field . Drift velocity was between 0 . 5 and 2 cycles/sec . To avoid transient responses , the contrast was incremented in a 1 sec duration ramp , maintained at steady level for 3 or 4 sec , then decreased back to 0% in a 1 sec duration ramp , then maintained at 0% contrast for 1 sec . The measurement of mean firing rates was restricted to the 3–4 sec plateau period . The fits of the CRF to a H-ratio function was performed as with the simulations data ( see above ) . The quality of the fit was good , ( ) except one supersaturating cell ( ) but there was no good reason to exclude this cell . The mean was ( S . D . ) and the median ( interquartile ) . Receptive fields were classified as “simple” or “complex” on the basis of the relative modulation ( F1/F0 [86] ) in their response to gratings at the optimal spatial frequency . In our data set , the distribution of F1/F0 was bimodal , with a gap at 1 . Cells were considered as simple when the relative modulation was and complex when it was [86] . A cell was considered to display saturating response when the response extrapolated to 100% contrast was equal to . It was considered as non-saturating when the extrapolated response was less than 0 . 95 and as super-saturating if the response to at least one of the test contrast below 90% was larger than 1 . 05 . | Both the response and membrane potential of neurons in the primary visual cortex ( V1 ) are selective to the orientation of elongated stimuli . The widths of the tuning curves , which characterize this selectivity , hardly depend on stimulus contrast whereas their amplitude does . The contrast dependence of this amplitude , the contrast response function ( CRF ) , has a sigmoidal shape . Saturation of the spike response is substantially lower than the neurons' maximal firing rate . These well established facts constrain the possible mechanisms for orientation selectivity in V1 . Furthermore , the single neuron CRFs in V1 display a broad diversity in their shape . This adds other constraints . Many theoretical works have tried to elaborate mechanisms of orientation selectivity that are compatible with the contrast invariant tuning widths . However , these mechanisms are usually incompatible with sigmoidal CRFs . We propose a mechanism which accounts simultaneously for contrast invariant tuning width for both rate and voltage response and for the shape and diversity of the CRFs . This mechanism relies on the interplay between power-law frequency-current transfer functions of single neurons , as measured in vivo in cortex , and on the recurrent interactions in the cortical circuit . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience/natural",
"and",
"synthetic",
"vision",
"neuroscience/theoretical",
"neuroscience",
"neuroscience/sensory",
"systems"
] | 2011 | Power-Law Input-Output Transfer Functions Explain the Contrast-Response and Tuning Properties of Neurons in Visual Cortex |
Single-cell RNA sequencing ( scRNA-seq ) has been widely applied to discover new cell types by detecting sub-populations in a heterogeneous group of cells . Since scRNA-seq experiments have lower read coverage/tag counts and introduce more technical biases compared to bulk RNA-seq experiments , the limited number of sampled cells combined with the experimental biases and other dataset specific variations presents a challenge to cross-dataset analysis and discovery of relevant biological variations across multiple cell populations . In this paper , we introduce a method of variance-driven multitask clustering of single-cell RNA-seq data ( scVDMC ) that utilizes multiple single-cell populations from biological replicates or different samples . scVDMC clusters single cells in multiple scRNA-seq experiments of similar cell types and markers but varying expression patterns such that the scRNA-seq data are better integrated than typical pooled analyses which only increase the sample size . By controlling the variance among the cell clusters within each dataset and across all the datasets , scVDMC detects cell sub-populations in each individual experiment with shared cell-type markers but varying cluster centers among all the experiments . Applied to two real scRNA-seq datasets with several replicates and one large-scale droplet-based dataset on three patient samples , scVDMC more accurately detected cell populations and known cell markers than pooled clustering and other recently proposed scRNA-seq clustering methods . In the case study applied to in-house Recessive Dystrophic Epidermolysis Bullosa ( RDEB ) scRNA-seq data , scVDMC revealed several new cell types and unknown markers validated by flow cytometry . MATLAB/Octave code available at https://github . com/kuanglab/scVDMC .
In recent years , single-cell RNA sequencing ( scRNA-seq ) has emerged as the dominant method for quantifying transcriptome-wide mRNA expression in individual cells . While traditional bulk RNA-seq ignores the differences between individual cells and treats the population of cells as homogeneous , scRNA-seq identifies sub-populations of single cells and can be useful for characterizing sub-population structure , mechanisms of transcription regulation , and understanding disease progression [1] and immunology [2] . A typical scRNA-seq protocol consists of several steps: isolation of single cells and RNA , reverse transcription , amplification , library generation , and sequencing . In addition to the noise and bias that exist in bulk RNA-seq experiments , issues unique to scRNA-seq include those from biological sources , such as cell-cycle stage or cell size , as well as from technical/systematic sources , such as capture inefficiency , material degradation , sample contamination , amplification biases , GC content , and sequencing depth . These experimental biases and limitations cause uneven coverage of the entire transcriptome and result in an abundance of zero-coverage regions [3 , 4] . Typically , the cost of scRNA-Seq is much higher than bulk RNA-Seq per sample , and thus , scRNA-Seq of a large patient cohort is still prohibitively costly . When a large number of single-cells from multiple samples are sequenced , more complex batch effects might be introduced . Finally , some poorly sampled cell populations might only contain very few cells for the analysis . To address all these challenges , proper integration of multiple scRNA-Seq datasets generated from different experiments is important . When multiple single-cell populations from biological replicates or related samples such as a patient cohort are analyzed to discover the common and sample-specific cell types , technical biases and irrelevant biological variance among independent samples cannot be easily identified and removed from the signal before clustering the single cells . For example , when the scRNA-seq profiles from multiple patients are pooled together for clustering , the clusters will highly overlap with the division of the single cells by the sample origins rather than similar types such as pathogenic cells vs normal cells . In this paper , we introduce a multitask learning method with embedded feature selection to simultaneously capture the differentially expressed genes among cell clusters and across all cell populations to achieve better single-cell clustering . The key advantage of multitask clustering is the use of multiple single-cell populations to leverage the sample size limitation in each individual dataset while allowing dataset-specific variations among the same cell types across the datasets . To illustrate the objective , Fig 1 shows a simulation example of scRNA-seq data of 100 single cells from three cell populations ( n = 33 , 33 and 34 ) with 1000 expressed genes . Among the 1000 genes , gene A and gene B are the hidden markers that are differentially expressed across the four cell types ( indicated by four different colors ) . In the ideal scenario , there is no technical bias and the marker genes are known as shown in the ground truth in Fig 1 ( A ) . Fig 1 ( B ) shows the single-cell datasets after biological variation , technical biases , and noise are introduced . The data distributions are very different across the three cell populations after the rotation , re-scaling and addition of noise . It is also challenging to identify the true marker genes with a limited number of samples in each population . Simply pooling the single-cell data from the three populations together will confuse the clustering , even with the correct marker genes identified ( Fig 1 ( C ) ) . Conversely , separated clustering on each single-cell population suffers more from the biological variation as the number of single cells is not sufficient in each individual analysis to identify the true maker genes ( Fig 1 ( D ) ) . As shown in Fig 1 ( E ) , variance-driven multitask clustering of single-cell RNA-seq data ( scVDMC ) utilizes expression patterns of different single-cell populations with shared cell-type markers and corresponding similar clusters for better integration .
Assume a total of D domains with each domain representing a single-cell population for clustering . Let matrix X ( d ) ∈ R m × n ( d ) denote RNA-seq gene expression values from domain d ∈ {1 , 2 , … , D} , where m is the number of features ( genes ) and n ( d ) is the single-cell sample size of domain d . Let U ( d ) ∈ R m × k denote the cell-type cluster centers , vector Y i , j = [ U i , j ( 1 ) , U i , j ( 2 ) , … , U i , j ( D ) ] T stack the ( i , j ) -th entry of every U ( d ) and the binary matrix V ( d ) ∈ {0 , 1}n ( d ) ×k denote the assignments of each single-cell to the clusters , where k is the number of cell types ( clusters ) . With the binary vector B ∈ {0 , 1}m denoting the indicators of feature selection ( 1: selected and 0: not selected ) and DB denoting the diagonal matrix with B on the diagonal , scVDMC model outlined in Fig 2 is defined as: minimize U ( d ) , V ( d ) , B 1 2 ∑ d = 1 D | | D B ( X ( d ) - U ( d ) V ( d ) T ) | | F 2 - w ∑ d = 1 D B T Var ( U ( d ) ) + α ∑ i , j B i Var ( Y ( i , j ) ) subject to ∑ B = λ , ∑ j V i , j ( d ) = 1 , ∀ i = 1 , 2 , … , n ( d ) , ∀ d = 1 , 2 , … , D , ( 1 ) where w and α > 0 are hyper-parameters to balance the three error terms: the reconstruction error , the cluster center separation in each cell population , and the variance of the cluster centers across the different single-cell populations . λ ∈ Z + is the predefined number of features to be selected . | | D B ( X ( d ) - U ( d ) V ( d ) T ) | | F 2 in Eq ( 1 ) denotes the reconstruction error of the classic k-means clustering as matrix factorization with DB selecting marker genes by B , i . e . the reconstruction error is only measured on the marker genes by ignoring the irrelevant ( non-selected ) genes . The second term BTVar ( U ( d ) ) is introduced to maximize the separation of the cluster centers , where Var ( U ( d ) ) is defined as a vector in which each element is the variance of the vector U i , : ( d ) ∈ R k × 1 [5] . The third term Var ( Y ( i , j ) ) denotes the variance of the vector Y ( i , j ) , which is introduced to require similar gene expression centers across different single-cell populations . Note that the reconstruction error encourages selection of low expression genes since the errors are usually smaller on smaller values while the second variance term encourages selection of high expression genes since the variances tend to be larger on larger values . Together as the sum over all the domains , the cost function provides a balanced error on the compactness and separation of the clusters of the cell types tuned by feature selection across all the domains . The unique but similar cluster centers in each domain preserves the unique expression patterns while the features are selected as common marker genes for different cell types . For the three hyper-parameters in Eq ( 1 ) , λ ( the number of marker genes ) is typically a small number based on prior knowledge of the cell types , and the selection of balancing weight w and α is discussed later in this section . Algorithm 1 scVDMC algorithm 1: Input: X ( d ) , α , k , w , λ , d = 1 , 2 , … , D 2: output: U ( d ) , V ( d ) , B 3: Initialize U ( d ) and V ( d ) . 4: repeat 5: compute B with integer linear programming in Eq ( 7 ) 6: for d = 1 , 2 , … , D do 7: solve V ( d ) by Eq ( 2 ) 8: solve U ( d ) by ( 6 ) 9: end for 10: until U ( d ) , V ( d ) and B converge 11: return U ( d ) , V ( d ) and B The full scVDMC algorithm is shown in Algorithm 1 . The goal is to minimize the cost function in Eq ( 1 ) to obtain the optimal U ( d ) , V ( d ) and B . We employ an alternating update strategy to solve the optimization problem . First , we fix the feature selection B , all the cluster centers U ( i ) , i = 1 , 2 , … , D and all other V ( i ) , i ≠ d , to obtain a certain V ( d ) . minimize V ( d ) 1 2 | | D B ( X ( d ) - U ( d ) V ( d ) T ) | | F 2 subject to ∑ j V i , j ( d ) = 1 , ∀ i = 1 , 2 , … , n ( d ) . ( 2 ) This is equivalent to assigning samples to the nearest centers U ( d ) by the Euclidean distance in the features selected by B , where each column of DB X ( d ) is a sample and each column of DB U ( d ) is a center . Then the distance of a sample to every center is calculated and the nearest center is chosen to assign 1 to the corresponding V ( d ) . The time complexity for assigning each sample to one of the k clusters over the λ marker genes will be O ( n × k × λ ) , where n is the total number of samples in all the domains . Next , we fix the feature selection B , all clustering assignments V ( i ) , i = 1 , 2 , … , D , and all other U ( i ) , i ≠ d , to solve a certain U ( d ) , rewritten as: minimize U ( d ) 1 2 ∑ i = 1 m B i | | ( X i , : ( d ) - U i , : ( d ) V ( d ) T ) | | 2 2 - w ∑ i = 1 m B i Var ( U i , : ( d ) ) + α ∑ i , j B i Var ( Y ( i , j ) ) , ( 3 ) where Var ( U i , : ( d ) ) is the variance of vector U i , : ( d ) defined as Var ( U i , : ( d ) ) = 1 k ( U i , : ( d ) - U i , : ( d ) 1 k 1 k T k ) ( U i , : ( d ) - U i , : ( d ) 1 k 1 k T k ) T = 1 k U i , : ( d ) ( I k - 1 k 1 k T k ) ( I k - 1 k 1 k T k ) T U i , : ( d ) T = 1 k U i , : ( d ) ( I k - 1 k 1 k T k ) U i , : ( d ) T , ( 4 ) where Ik denotes the identity matrix of size k and 1k is a length k column vector of all ones . Similarly , we have Var ( Y ( i , j ) ) = 1 d Y ( i , j ) T ( I d - 1 d 1 d T d ) Y ( i , j ) . ( 5 ) As shown in S1 Appendix , the analytical solution of Eq ( 3 ) when Bi = 1 is Ui , : ( d ) T= ( V ( d ) TV ( d ) −2wkΨ+2αkΦd , ddIk ) −1 ( V ( d ) TXi , : ( d ) T−2αkd∑l≠dΦdlUi , : ( l ) T ) . ( 6 ) The time complexity is O ( k3 ) for the matrix inversion and O ( n × k2 ) for computing V ( d ) T V ( d ) . Since the matrix inversion is common to all the genes and only needs to be computed once , the total time complexity is only O ( n × k × λ ) . Finally , to update binary vector B , we fix all U ( d ) and V ( d ) to optimize minimize B ∑ i = 1 m B i ( 1 2 ∑ d = 1 D | | ( X i , : ( d ) - U i , : ( d ) V ( d ) T ) | | 2 2 - w ∑ d = 1 D Var ( U i , : ( d ) ) + α ∑ j = 1 k Var ( Y ( i , j ) ) ) subject to ∑ B = λ , ( 7 ) which is a standard constrained linear binary integer programming problem that can be easily solved by sorting the coefficients of B and taking the top λ entries . The time complexity is O ( m × n × k ) for computing the construction error terms , O ( D × m × k ) for computing the variances and O ( m log m ) for sorting the coefficients . The overall time complexity is O ( m × n × k ) assuming n × k > log m . Thus , the total time complexity of each iteration in Algorithm 1 will be O ( ( m + λ ) × n × k ) , which is comparable to k-means when λ <<m . There are four hyper-parameters to tune for the scVDMC algorithm , α and w: weights of the two variance terms , k: the number of clusters and λ: the number of marker genes . Below we describe our strategies for tuning α , w and k assuming that λ can be approximately informed by prior knowledge of the cell types . Tuning α: The role of α is to weight the cost term on the cross-domain variance of the cluster centers . The larger the α the more similar the cluster centers are across the domains . Ideally , α should be relatively small to allow smaller reconstruction error but yet meet the consistency requirement across the domains . The strategy is to start with a small α and measure the total difference between the cluster centers of the corresponding cluster across the domains , and then increase α to reduce the difference until the total difference does not change much . This selection can also be achieved by visualization of the cluster centers with Principle Component Analysis ( PCA ) or other dimension reduction methods . After clustering , we can project the data in each domain into the first two PCs . The distance between the cluster centers of the same cluster in each domain can be compared for choosing an appropriate α . Several examples are shown later in the experiments . Deriving the upper bound of w: Eq ( 3 ) is a sum of a few quadratic terms of variable U i , : ( d ) . The global minimum of U i , : ( d ) can be solved in closed-form if the Hessian below is positive semi-definite , H = V ( d ) T V ( d ) - 2 w k Ψ + 2 α k Φ d , d d I k . ( 8 ) In the following , we show that an upper bound on w will guarantee that H is positive semi-definite . By Gershgorin circle theorem ( For any eigenvalue δ of matrix H , |δ − Hii| ≤ ∑j≠i |Hij| for ∀i ⇔ Hii − ∑j ≠ i |Hij| ≤ δ ≤ Hii + ∑j ≠ i |Hij| . ) , the sufficient condition of H ≽ 0 is Hii − ∑j ≠ i |Hij| ≥ 0 for ∀i . This is equivalent to stating that H is diagonally dominant and only has non-negative diagonal entries . H can be rewritten as follows , H i i = c i + 2 w ( 1 - k ) k 2 + 2 α k ( d - 1 ) d 2 , ∀ i = 1 , . . . , k H i j = 2 w k 2 , ∀ i ≠ j , where ci is the ith diagonal entry of matrix V ( d ) T V ( d ) , i . e . , the size of cluster i . Then we have c i + 2 w ( 1 - k ) k 2 + 2 α k ( d - 1 ) d 2 ≥ 2 w ( k - 1 ) k 2 and thus , w ≤ k 2 c m i n 4 ( k - 1 ) + α k 3 ( d - 1 ) 2 d 2 ( k - 1 ) where cmin is the minimum of ci , ∀i = 1 , … , k . Since cmin ≥ 1 ( no empty cluster ) , we obtain a loose upper bound of w = k 2 4 ( k - 1 ) + α k 3 ( d - 1 ) 2 d 2 ( k - 1 ) . In all the experiments , we set w to be smaller than the upper bound for feasible implementation . Determining the number of clusters k: The number of clusters k is selected by an “elbow” plot of the within-clusters sum of squares Ts computed as follows: T s = ∑ d = 1 D | | D B ( X i , : ( d ) - U i , : ( d ) V ( d ) T ) | | 2 2 . ( 9 ) Ts represents the amount of variance to minimize for better clustering . Larger k will lead to smaller Ts . By plotting Ts under different options of k , we can select the best k at the so-called “elbow” of the curve . S6 and S7 Figs show the “elbow” plot on two datasets in the experiments . In addition , when an empty cluster is created , the calculation of cluster center variance will be invalid . To address the possible issue , we use a simple splitting procedure to handle empty clusters . Specifically , if there is an empty cluster in V ( d ) ( i . e . the whole column is 0 ) we randomly split the largest cluster into two clusters . This procedure is repeated until there are exactly k clusters . This strategy is similar to commonly used k-mean rerun when a cluster center is collapsed on a single data point or no data point . To identify sub-populations producing homing signals that could attract bone marrow-derived cells to injured skin , we captured single dermal fibroblasts from six patients with severe generalized RDEB and their HLA-matched healthy siblings using the Fluidigm C1 system . The demographics information of the patients and donors are shown in S1 Table . Cell culture: Dermal fibroblasts from patients with severe generalized RDEB and their human leukocyte antigen ( HLA ) matched healthy siblings were obtained from skin biopsies and cultured in DMEM high glucose ( Thermo Fisher Scientific ) containing 10% fetal bovine serum ( MilliporeSigma ) , 1% Pen/Strep ( Thermo Fisher Scientific ) , 1% L-glutamine ( Thermo Fisher Scientific ) , and 1% MEM NEAA ( Thermo Fisher Scientific ) . For sub-culture , the medium was removed and cells were washed with 1X PBS ( Thermo Fisher Scientific ) and detached using Trypsin/EDTA ( Thermo Fisher Scientific ) . Experiments were performed with fibroblasts at passages 4-9 . Single-cell capture and RNA-seq: Fibroblasts were collected by trypsinization and resuspended in 5 μL of fibroblast medium for loading into the capture chip . The medium- ( 10-17 μm diameter ) and large-size ( 17-25 μm diameter ) chips were used to capture cells with the C1 system ( Fluidigm ) . Cells were loaded at a concentration of 2 . 5 x 105 per μL and stained with the Live/Dead Viability/Cytotoxicity kit ( Thermo Fisher Scientific ) . Cells were imaged with phase-contrast and fluorescence microscopy to assess cell number and viability at each capture point . Capture sites with single , live cells were selected while capture sites with multiple , no , or an unclear number of cells were excluded from further analysis . Images for each single-cell used in this study are available upon request . In total , 295 patient cells and 248 sibling cells were selected . On the device , cDNA was created from the selected cells using the SMARTer Ultra Low RNA kit designed for the C1 system ( Clontech ) . mRNA libraries were constructed using the Nextera XT kit ( Illumina ) according to the manufacturer’s protocol . The libraries were sequenced on an Illumina MiSeqv3 with 75bp paired-end reads to a depth of 19-22 million reads per lane . Target sequencing depth for each library was 200K reads . The RNA-seq data have been deposited in NCBI’s Gene Expression Omnibus and are accessible through GEO series accession number GSE108849 ( https://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE108849 ) . Processing of RNA-seq data: Paired-end 75bp reads were mapped to the UCSC human transcriptome ( hg19 ) using Bowtie2 ( version 2 . 2 . 4 ) and Tophat ( version 2 . 0 . 9 ) . Gene expression levels were calculated using Cuffquant ( Cufflinks version 2 . 2 . 1 with parameters -u -max-bundle-frags 10000000 ) and Cuffnorm ( Cufflinks version 2 . 2 . 1 ) . FPKM values as estimated by Cufflinks were added a value of 1 ( to avoid zeros ) and log2 transformed . We removed nine single-cell samples with low read counts ( < 50K ) and sub-sampled two single-cell samples sequenced as population controls with high read counts ( > 1 . 5M ) ( random sub-sampling , 10% of total reads ) . 11 single-cell samples were excluded as outliers . We excluded lowly expressed genes ( average log2 ( FPKM ) < 1 . 5 ) from further analysis . The remaining 543 single-cell samples met the requirement of expressing at least 2 , 000 of these remaining 5 , 196 genes . For each individual , the number of single-cells used in the analysis and the average number of reads for those single-cells is summarized in Table 1 . The total number of the reads and the number of aligned reads in each cell are also shown in S3 Fig . Flow cytometry: Fibroblasts were collected by trypsinization ( as above ) and resuspended in fibroblast medium . A BD Cytofix/Cytoperm™ kit ( BD Biosciences ) was used to prepare the cells for intracellular staining . Cells were fixed for 15 min with 150 μl Fixation/Permeabilization solution before being resuspsended in 300 μl 1X BD Perm/Wash Buffer and incubated at 4°C for 20 min . Primary antibodies ( S2 Table ) were diluted in 100 μl 1X BD Perm/Wash Buffer and cells were resuspended in this for 20-30 min at 4°C , followed by one wash with 500 μl 1X BD Perm/Wash Buffer . Secondary antibodies ( S3 Table ) were diluted in 300 μl 1X BD Perm/Wash Buffer and cells were resuspended in this for 20-30 min at 4°C , followed by one wash with 500 μl 1X BD Perm/Wash Buffer , and resuspension in 300 μl 1X BD Perm/Wash Buffer . Flow cytometry experiments were carried out on a BD LSRII system equipped with FACsDiva 8 . 0 software ( BD Biosciences ) and analyzed using FlowJo ( Tree Star Inc . ) . Most existing methods focus only on sub-population clustering and differential gene expression detection among the learned cell clusters with one ( pooled ) cell population . Some of these methods were directly adopted from traditional bulk RNA-seq analysis and/or classical dimension reduction algorithms such as Principal Component Analysis [6–8] , hierarchical clustering [9] , t-SNE [10–12] , Independent Component Analysis [13] and Multi-dimensional Scaling [14] . Other methods focus on special properties of scRNA-seq data , such as high variance and uneven expressions . For example , SNN-Cliq [15] uses a ranking measurement to get reliable results on high dimensional data; [16] proposed a special dimension reduction method to handle the large amount of zeros in scRNA-seq; [17] proposed a Latent Dirichlet Allocation model with latent gene groups to measure cell-to-cell distance; CellTree method [17] clusters single cells by a detected tree structure outlining the hierarchical relationship between single-cell samples to introduce biological prior knowledge; Seurat [18] was proposed to infer cellular localization by integrating single-cell RNA-seq data with in situ RNA patterns; and more recently a consensus clustering approach SC3 [19] was proposed to improve the robustness of clustering through combining multiple clustering solutions by consensus . Mixed multiple batch strategies [9 , 20] have been proposed to reduce the technical variance , which does not directly improve clustering . To the best of our knowledge , multitask clustering with an embedded feature selection has not been previously applied to scRNA-seq data analysis . All patients gave consent for samples to be taken per the Declaration of Helsinki . This research was approved by the University of Minnesota’s Institutional Review Board: IRB 1301M26601: MT2013-02R ( Establishment of a Cell and Tissue Repository for Human Cell Reprogramming and Derivation of iPS Cell Lines to Investigate Mechanisms and Treatment of Human Disease ) .
scVDMC was compared with six baseline methods: ( 1 ) k-means clustering on each domain separately , ( 2 ) pooling all domains and applying k-means clustering , ( 3 ) SNN-Cliq [15] , ( 4 ) CellTree [17] , ( 5 ) Seurat [18] and ( 6 ) SC3 [19] . Pooled k-means ( 2 ) was used to obtain the initialization for scVDMC . To apply the SNN-Cliq method [15] , we used the provided MATLAB code to transform the data into the SNN graph , then used the Python code to produce the clustering result by ranking measurement . There are three hyper-parameters: k ( size of the nearest neighbor list ) , r ( parameter for quasi-clique finding , range ( 0 , 1] ) , and m ( parameter for cluster merging range ( 0 , 1] ) . We tested multiple combinations of the three hyper-parameters using k = 3 , 5 , 7 , r = 0 . 1 , 0 . 2 , … , 0 . 9 and m = 0 . 1 , 0 . 2 , … , 0 . 9 . We also required the program to annotate all the data instead of leaving singletons unlabeled ( −n ) . Since SNN-Cliq identifies the number of clusters automatically , we only reported the results with the correct number of clusters . In all experiments with SNN-Cliq , we further removed genes with low expression and log-transformed the data , as recommended in [15] . To apply the CellTree method [17] , we used the provided R package to first fit a Latent Dirichlet Allocation ( LDA ) model with the default method ( joint MAP estimation ) to choose the number of topics followed by learning a pair-wise distance for all cells . Then we ran hierarchical clustering with four different methods for computing cluster distance ( ‘ward’ , ‘complete’ , ‘single’ , ‘average’ ) and selected the best clustering results . To apply Seurat [18] , Seurat v2 . 0 R package was downloaded from SATIJA LAB . The scRNA-seq data were converted into the required format ( gene index | cell index | gene expression ) as the input . The parameter “Resolution” tunes the granularity of the downstream clustering , with increased values resulting a larger number of clusters . We tested a range [0 . 5 , 1 . 5] to get the exact number of clusters for comparison with other methods . The reported result of Seurat is computed with the resolution parameter that gives the exact number of clusters and the lowest error . To apply the SC3 [19] we downloaded SC3 v1 . 7 . 2 R package from Bioconductor . All parameters in SC3 are set to default . In the experiments with more than 5000 instances for clustering , the SVM mode will be trigged to run a second stage supervised learning to improve the scalability . To further test separated cluster , pooled clustering and SC3 combined with feature selection , we chose the genes with larger variance as the marker genes . Since the other three baselines use a different strategy for clustering and do not provide marker-gene selection , we only focused on the clustering result for these three baselines . The true cluster labels are obtained as the validated clusters with high confidence in the mESC data [21] and Lung data [22] , and the known PBMC populations from donor A sorted with FACS analysis [23] . We downloaded the single-cell expression data for 250 mESCs [21] from the European Bioinformatics Institute’s ( EBI ) ESpresso database . These 250 mESCs cultured in serum conditions were captured using the Fluidigm C1 on three different days from three different passages ( biological replicates , n = 81 , 90 , and 79 ) . After removing genes expressed uniformly within a single replicate , 12 , 114 genes remained . To tune α for scVDMC , we examined the positions of the cluster centers across the domains and show the visualization by PCA in S4 ( A ) & S4 ( B ) Fig . Based on the visualization , α = 0 and 1 are chosen since the relative positioning of cluster centers are similar in all the three domains . For the SNN-Cliq method , we further removed genes with log-transformed average expression less than 20 . Fig 3 ( A ) shows the clustering results . Compared with the six baselines , scVDMC shows a consistently lower error with different choices of λs . Within a reasonable range of λ , such as from 20 to 200 , scVDMC shows significant improvement compared with the baseline methods . When λ is too small , such as 10 genes selected , there are not enough markers to capture the difference among the cell types such that the error is larger . When λ is too big , scVDMC will consider almost all the genes and the variance selection will not play a role . As such , scVDMC will eventually degrade into separated k-means and the error will also increase . As shown in S1 ( A ) Fig , it is worth noting that the results are less sensitive to the choice of the parameter w , for which the upper bound for w is 9 8 in this case . It is also interesting that the CellTree method performed better than both pooled and separated k-means , while SNN-Cliq and SC3 performed better than separated k-means but worse than pooled k-means . Under various tuning of the parameters , Seurat still performed poorly on this dataset . Both separated k-means and pooled k-means performed much worse with the feature selection by variance , indicating that simple feature selection strategies will not identify correct markers in this dataset . Running scVDMC with α = 1 performed the best when 20 marker genes are selected but the overall performance is very similarly as running with α = 0 , indicating that the control of the cross-domain variance could play a role in improving the results . However , since the cluster centers are already not very different when running with α = 0 , the improvement will only be marginal . Fig 3 ( B ) shows the detailed clustering errors by scVDMC , pooled k-means and separated k-means . Compared with the pooled k-means and separated k-means , scVDMC captures relatively high variance in the leading principle components and achieves improved clustering in every domain ( fewer mixed-color dots ) . In S2 ( A ) Fig , we also show the convergence of scVDMC by the number of iterations . Analysis of the mESC transcriptome data using scVDMC yielded comparable results on marker gene selection in the original paper [21] as well as pooled and separated k-means . Both analyses were able to detect and highly rank the known markers of differentiation Krt8 , Krt18 , Anxa1 , Anxa3 , and Acta1 . Further , scVDMC detected several additional genes that pooled k-means , separated k-means and the original paper did not . These included Cav1 , which is required for normal lung development [24] and Dsp , variants of which are associated with idiopathic pulmonary fibrosis [25] . We downloaded the single-cell expression data for 80 embryonic mouse lung epithelial cells [22] . These 80 single-cell samples were taken from three different mice ( biological replicates , n = 20 , 34 , and 23 ) and contained five cell types: ciliated , Clara , AT1 , and AT2 cells , as well as a bi-potential progenitor ( BP ) . Since only one replicate contained ciliated cells , we removed these from the analysis , leaving 77 single-cell samples . After removing genes expressed uniformly within a single replicate , 7 , 357 genes remained . To tune α for scVDMC , we examined the positions of the cluster centers across the domains and show the visualization by PCA in S4 ( C ) & S4 ( D ) Fig . α = 1 is chosen as the optimal parameter to achieve similar relative positioning of cluster centers in all the three domains . For the SNN-Cliq method , we further removed genes with log-transformed average expression less than 2 . With the limited number of single-cell samples in this dataset , scVDMC still much improved clustering over the baselines in the range of λ ∈ [30 , 80] shown in Fig 3 ( C ) . In Fig 3 ( D ) , PCA plots of the top 50 genes show a trend similar to the ESC dataset , where scVDMC’s top genes capture more variance and show less clustering error . Both SNN-Cliq and CellTree performed better than pooled k-means and separated k-means , with SNN-Cliq leading CellTree by a very small margin . Similarly , Seurat also performed poorly while SC3 performed well on the dataset with only 5 mistakes . It is also interesting to observe that running scVDMC with α = 1 performed significantly better than running with α = 0 , indicating that the control of the cross-domain variance played an important role in improving the results . Since the cluster centers are very different when running with α = 0 , the improvement is significant . Another interesting observation is that the clustering performance is more sensitive to the number of marker genes to select by scVDMC . In particular , selection of 20-80 genes with scVDMC ( α = 1 ) will give the optimal clustering results while selection of more than 90 genes will give much higher error . This is due to the small clusters in this dataset ( e . g . purple cluster in domain 2 and yellow cluster in domain 1 ) , which could be sensitive to the number of selected genes in low-read-coverage samples . Thus , the error will be more sensitive to the gene selection in this small dataset . On this dataset , both separated k-means and pooled k-means performed better with the feature selection by variance but never achieved zero clustering error as scVDMC does . As shown in S1 ( B ) and S2 ( B ) Figs , scVDMC behaved similarly by the choices of the w parameters and the convergence . Analysis of the mouse lung epithelial transcriptome data using scVDMC yielded comparable results in the original paper [22] as well as pooled and separated k-means . Both analyses were able to detect and highly rank the known marker genes of the different cell types: Clara ( Scgb1a1 ) , AT1 ( Pdpn , Ager ) , and AT2 ( Sftpc , Sftpb ) . Further , scVDMC detected several additional genes that pooled k-means , separated k-means and the original paper did not . These included two components of the Notch signaling pathway ( Notch1 and Nrarp ) previously shown to be critical for the development of lung alveolar spaces , with AT2 cells being major sites of Notch activation [26] . We downloaded the peripheral blood mononuclear cells ( PBMC ) data generated by [23] from the 10xGenomics website . In the original data , there are 10 bead-enriched subpopulations of PBMC from a fresh donor ( Donor A ) with 93802 cells in total . In addition , there are also PBMC from two other frozen donors ( Donor B and C ) with 7783 and 9519 cells , respectively . A massive droplet-based method was applied to count the mRNAs in the tens of thousands of cells in parallel . To better evaluate the multitask learning setting , we sampled from each of the 10 subpopulations of Donor A in proportion to the sizes of the populations to obtain four subsets of cells from Donor A with 200 , 500 , 1000 and 10000 cells by sampling . We repeated the sampling procedure five times to generate the mean and variance of Adjusted Rand index ( ARI ) [19] . We kept all the cells in Donor B and C . We removed the genes expressed in less than 3 cells which results in 17647 genes remained . To determine the number of clusters in the PBMC data , we examined the “elbow” plot in all the three cell populations shown in S6 Fig . The plots show consistent patterns in the three cell populations that the “elbow” is observed to start around k = 10 verifying that there are indeed around 10 cell types in the data . To tune α for scVDMC , we examined the positions of the cluster centers across the domains and show the visualization by PCA in S4 ( E ) & S4 ( F ) Fig . α = 5 is chosen since the relative positioning of cluster centers are also relatively similar in the three domains . The baseline methods k-means and SC3 are tested on the pooled data ( mixture of Donor A , B and C ) and separated data ( Donor A only ) . For SC3 , the hybrid approach ( consensus clustering + SVM ) with its default parameters is applied on the pooled data due to the scalability issue [19] . Clustering performance is measured using Adjusted Rand index ( ARI ) [19] by comparing the predicted labels with the true labels from sampling the ten subpopulations of PBMC in Donor A . Fig 4 shows the clustering results . Compared with pooled k-means and SC3 , scVDMC shows a consistently higher ARI with different choices of λs . scVDMC also shows a significant improvement compared with separated k-means and SC3 when there are 200 , 500 and 1000 cells from Donor A . The improvement by scVDMC becomes only marginal when there are 10000 cells sampled from Donor A . The observation is common since larger dataset often benefit less from multitask learning , i . e . as the sample size in donor A increases , less additional information carried in the data of donor B and C can inform a better clustering of donor A data . On this dataset , we also observed that the clustering performance of scVDMC does not rely on the parameter α . This is likely because the agreement among the 10 clusters in the three domains is already high when α = 0 as shown in S4 ( E ) Fig . Therefore enforcing stronger agreement by increasing α will not lead to big improvement as shown in S4 ( F ) Fig . Overall , scVDMC performed well on the large-scale data showing the advantage of applying multitask learning . SC3 did not over-perform separated k-means indicating the consensus clustering is less effective on this dataset . Recessive Dystrophic Epidermolysis Bullosa ( RDEB ) is an inherited blistering disorder caused by loss-of-function mutations in the COL7A1 gene that codes for type VII collagen ( C7 ) [27] . C7 forms the anchoring fibrils that attach the epidermis to the dermis [28] . When C7 is missing , the skin becomes extremely fragile , eroding at the slightest touch . From birth , patients with this disease must undergo intensive bandaging and daily wound care . They are also susceptible to a highly aggressive form of squamous cell carcinoma [29–32] . It has been shown that allogeneic hematopoeitic cell transplant ( HCT ) can partially rescue the RDEB phenotype . Cells from the bone marrow home to the skin and deposit C7 at the dermal-epidermal junction , greatly improving skin integrity in a subset of patients [33] . However , the molecular mechanism by which this occurs remains unknown . To determine the number of clusters in the RDEB data , we examined the “elbow” plot in all the six cell populations shown in S7 Fig . The plots show consistent patterns in all six cell populations that the “elbow” starts from k = 4 , which was chosen as the number of clusters for clustering in all the experiments on the RDEB data . The convergence of scVDMC on RDEB data is shown in S2 ( D ) Fig . Applying scVDMC to the RDEB single-cell dataset revealed quite different cell population structures for the six patient-sibling pairs . As shown in Fig 5 , the agreement among the cluster centers across the six populations varies under different choices of α . When α = 0 , no agreement among the cluster centers are required . The arrangement of the four cluster centers are very different in the six populations ( Fig 5 ( A ) ) . With larger values of α , the arrangement of the cluster centers becomes more similar . When α = 20 , the structure of the four cluster centers is almost identical for the six populations ( Fig 5 ( C ) ) . The visualization in Fig 5 clearly illustrates the effect of imposing variance constraint on the cluster centers across the populations to account for the population specificity and commonality . For comparison , we also applied SC3 on the pooled cell populations and the individual cell populations . SC3 failed to detect any cluster structures in the pooled cell populations by simply clustering the cells based on the sample origin as shown in S5 Fig . SC3 also only detected inconsistent clusters across the six populations as shown in Fig 5 ( D ) as expected since SC3 unlike scVDMC only clusters the cell populations independently . scVDMC identified several marker genes previously known to be involved in RDEB ( Fig 6 ) . These included CXCL12/SDF1 , the ligand for CXCR4 , which directs cells of the bone marrow to damaged tissue including skin [34] and HMGB1 , which has shown to be positively correlated with RDEB severity [35] and also mediates recruitment of bone marrow-derived cells to injured tissue [36] . Note that we empirically removed confounding cell cycle genes from the top 100 predicted markers and repeated scVDMC until there were no selected cell cycle genes . We also identified several genes as markers not previously associated with RDEB . These included COL11A1 , a minor fibrillar collagen shown to mark activated cancer-associated fibroblasts ( CAFs ) that is not typically expressed in fibroblasts associated with inflammation and fibrosis [37] . scVDMC also revealed GREM1 , a BMP antagonist associated with renal and pancreatic fibrosis [38 , 39] and MFAP5 , which promotes attachment of cells to micro-fibrils of the extracellular matrix and interacts with TGBβ growth factors [40] . We performed flow cytometry on the same RDEB patient and matched sibling fibroblasts to validate the expression levels of these genes at the single-cell level and found the results similar to our RNA expression data shown in Fig 7 . To further investigate the expressions of the these markers among the cells in the six populations , we plot the distribution of the cells with highly expressed markers in the six pairs in Fig 8 . In the plots , the expression patterns of GREM1 and MFAP5 are very consistent among the cells in all the six pairs with more enrichment in RDEB cells ( GREM1 ) or WT cells ( MFAP5 ) . The expression pattern of COL11A1 is consistent in five of the pairs with enrichment in WT cells except RDEB-WT pair 3 . Since the markers are selected to capture cell types rather than RDEB vs WT , there might be some discrepancy in the expression patters in each individual cell populations depending on the proportion of the cell types . As top hits , these genes potentially mark sub-populations of stromal cells that contribute to the transformation of the overlying epithelium and the development of squamous cell carcinoma in RDEB patients .
In this research , we demonstrated multitask learning is useful in analysis across multiple single-cell populations . It is also possible to apply other multitask learning or transfer learning methods [41] for the clustering tasks . scVDMC is a multitask clustering method specifically designed for scRNA-seq data for selection of a smaller set of cell-type markers and allows large variability in gene expression across the cell populations . Other methods are often built using different assumptions of the data that might not be applicable to the characteristics of scRNA-seq populations [42–44] . The amount of variation across multiple scRNA-Seq datasets depends on the nature of the datasets for the integrative analysis . For example , while we expect little variances among technical replicates and slightly more variances among biological replicates such that the variances do not play a major role in the pooled analysis , much larger variances might exist among samples of different tissue types or samples from different patients as those in the RDEB data . The key hypothesis of scVDMC is the existence of a common set of a small number of marker genes in every dataset that can partition each dataset into the same clusters . While the hypothesis is quite independent of the amount of variation across the datasets , scVDMC formulation accounts for the variation by tuning the parameter α to weight the variances . In theory , scVDMC is applicable to the general integration of scRNA-Seq datasets if the variances calculated among the cluster centers across the datasets well represent the underlying variations . However , in real applications , it is difficult to assess if the variations are captured by the computation of the variances . Thus , more careful practice of parameter tuning and validation of the results are necessary after the application of scVDMC . There are limitations in the scVDMC method . In multitask clustering , assuming a global k as the number of clusters in each cell population dataset does not always hold true as for some rare cell types , the corresponding cells may only be present in some populations . scVDMC might incorrectly split a cluster of one cell type because no empty cluster is allowed . One possible improvement is to model each domain with an individual k ( d ) with a more adaptive strategy for choosing k ( d ) . In this case , the overall balance between within-cluster distance and the variance will need to be more carefully weighted . In addition , cell-cycle-associated genes could be a large source of confounders . Unless the stages of cell cycle are the biological signal under study , cell cycle-related variation could obscure biological signals of interest . It is possible to model the confounders directly in the scVDMC method with more complex modeling . Alternatively , we could pre-process the scRNA-seq data to remove the cell cycle signals . For example , a Gaussian processes-based latent-variable model [45] was used to account for confounding variations due to the cell cycle in scRNA-seq data sets and then linear regression was applied to remove them . In this approach , a clearly defined cell cycle gene set is necessary to avoid removing true signals unexpectedly . Combined with the pre-precessing , scVDMC might achieve further improvement in clustering multiple cell populations . For a better interpretation of scRNA-seq data , CellTree [17] based on Latent Dirichlet allocation also provides soft cluster assignment as opposed to the hard one-cluster assignment and more recently , a new method [46] was introduced for visualizing the cluster membership of single cells by the soft cluster assignment known as “grades of membership” . It is also possible to extend scVDMC method to perform soft cluster assignment by relaxing V to contain positive real numbers rather than binary 0/1 in Eq 2 . The relaxation will require solving many least-squares problems and increase the computational time complexity . We plan to investigate better solutions of scVDMC in the future for soft cluster assignment and handling cell-cycle-associated gene signatures . | scRNA-seq enables detailed profiling of heterogeneous cell populations and can be used to reveal lineage relationships or discover new cell types . In the literature , there has been little effort directed towards developing computational methods for cross-population transcriptome analysis of multiple single-cell populations . The cross-cell-population clustering problem is different from the traditional clustering problem because single-cell populations can be collected from different patients , different samples of a tissue , or different experimental replicates . The accompanying biological and technical variation tends to dominate the signals for clustering the pooled single cells from the multiple populations . In this work , we have developed a multitask clustering method to address the cross-population clustering problem . The method simultaneously clusters each individual cell population and controls variance among the cell-type cluster centers within each cell population and across the cell populations . We demonstrate that our multitask clustering method significantly improves clustering accuracy and marker discovery in three public scRNA-seq datasets and also apply the method to an in-house Recessive Dystrophic Epidermolysis Bullosa ( RDEB ) dataset . Our results make it evident that multitask clustering is a promising new approach for cross-population analysis of scRNA-seq data . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"flow",
"cytometry",
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"cell",
"cycle",
"and",
"cell",
"division",
"applied",
"mathematics",
"cell",
"processes",
"marker",
"genes",
"fibroblasts",
"simulation",
"and",
"modeling",
"algorithms",
"devel... | 2018 | A multitask clustering approach for single-cell RNA-seq analysis in Recessive Dystrophic Epidermolysis Bullosa |
The combination of large-scale population genomic analyses and trait-based mapping approaches has the potential to provide novel insights into the evolutionary history and genome organization of crop plants . Here , we describe the detailed genotypic and phenotypic analysis of a sunflower ( Helianthus annuus L . ) association mapping population that captures nearly 90% of the allelic diversity present within the cultivated sunflower germplasm collection . We used these data to characterize overall patterns of genomic diversity and to perform association analyses on plant architecture ( i . e . , branching ) and flowering time , successfully identifying numerous associations underlying these agronomically and evolutionarily important traits . Overall , we found variable levels of linkage disequilibrium ( LD ) across the genome . In general , islands of elevated LD correspond to genomic regions underlying traits that are known to have been targeted by selection during the evolution of cultivated sunflower . In many cases , these regions also showed significantly elevated levels of differentiation between the two major sunflower breeding groups , consistent with the occurrence of divergence due to strong selection . One of these regions , which harbors a major branching locus , spans a surprisingly long genetic interval ( ca . 25 cM ) , indicating the occurrence of an extended selective sweep in an otherwise recombinogenic interval .
Strong selection during the evolution of crop plants has resulted in dramatic phenotypic differentiation . Undoubtedly , these same selective pressures will also have produced significant genomic consequences . Indeed , genomic regions that have been targeted by selection during crop evolution are expected to exhibit characteristic changes in their levels and/or patterns of nucleotide diversity . For example , under strong directional selection , one would expect a marked decrease in genetic variation in and near the targeted loci ( e . g . , tb1 [1]; waxy [2]; GIF1 [3] ) . However , under divergent selection , an increase in population genetic differentiation would be expected between the divergently selected lineages , coupled with localized decreases in genetic variation ( e . g . , [4] , [5] ) . The extent of these effects will be jointly determined by the strength of selection and the local recombination rate [6] , with stronger selection and/or reduced recombination affecting diversity across a larger chromosomal region . The use of large-scale population genomic analyses , especially when coupled with trait-based mapping approaches , thus has the potential to provide novel insights into the evolutionary history of crop plants and their genomes . Traditional quantitative trait locus ( QTL ) mapping analyses have provided considerable insight into the genetic basis of the phenotypic changes that have occurred during crop evolution ( e . g . , [7]–[9] ) . This general approach is , however , somewhat limited in terms of both mapping resolution and the amount of diversity assayed . Association mapping , which involves the correlation of molecular polymorphisms with phenotypic variation in a diverse assemblage of individuals , solves both of these problems , and is thus a useful alternative to standard QTL mapping approaches [10] . Because association populations typically capture many generations of historical recombination , linkage disequilibrium ( LD; the non-random association of alleles between loci ) is expected to be substantially lower than in family-based mapping populations , resulting in much higher mapping resolution . Moreover , the high level of diversity in a typical association mapping population allows for the simultaneous investigation of the effects of a broad spectrum of alleles across multiple genetic backgrounds . The downside of such analyses is that structure in the focal population can produce spurious marker/trait correlations in the absence of physical linkage [11] , [12] . Statistical advances have , however , made it possible to minimize the likelihood of false associations by accounting for relatedness amongst individuals ( i . e . , kinship ) and population structure ( e . g . , [13]–[15] ) . Ultimately , detailed insights into standing levels of nucleotide diversity , background patterns of LD , and relatedness amongst individuals within the focal population are critically important for the successful application of association mapping approaches . The use of high-density single nucleotide polymorphism ( SNP ) data derived from known genomic locations can facilitate the development of these insights and thus has the potential to enable the genetic dissection of important phenotypes in crop plants . Moreover , the outcomes of such analyses have potential downstream applications in marker-assisted breeding programs [16] . Here , we investigate SNP diversity , population differentiation , and the structure of LD across the 3 . 6 Gbp genome of sunflower ( Helianthus annuus L . ) and perform association analyses of plant architecture and flowering time in this valuable crop species . Cultivated sunflower is a globally important oilseed crop that was domesticated from the wild , common sunflower ( also H . annuus ) approximately 4 , 000 years ago by Native Americans . Following its domestication , sunflower was originally used as a source of edible seeds and for a variety non-food applications ( e . g . , as a source of dye for textiles and for ceremonial purposes ) [17]–[19] . The transformation of sunflower into an oilseed crop began in 18th century in Eastern Europe where breeding efforts increasingly focused on improving oil yield in a subset of the available germplasm . Commercial production commenced in North America in the mid-20th century , along with a focus on the development of sunflower as a hybrid crop . Modern sunflower is maintained in two primary breeding pools: the unbranched female ( A ) lines ( differing only in cytoplasm from paired maintainer or B lines ) , and the typically recessively-branching , multi-headed male restorer ( R ) lines that are crossed to generate the unbranched , fertile hybrids grown by producers . In this study , we used an Illumina Infinium 10 k SNP array [20] , [21] to genotype a diverse collection of publicly-available sunflower lines . We then used these data , along with phenotypic data collected from multiple locations , to analyze genome-wide patterns of genetic variation , characterize the extent of LD and population differentiation , and investigate the genetic basis of variation in plant architecture and flowering time .
The sunflower association mapping population utilized in this study was composed of 271 lines that have previously been shown to capture nearly 90% of the allelic diversity present within the cultivated sunflower gene pool [22] . This population is composed of accessions from the collections held by both the USDA North Central Regional Plant Introduction Station ( NCRPIS ) and the French National Institute for Agricultural Research ( INRA ) ( Table S1 ) . These accessions include numerous inbred lines and historically important open-pollinated varieties ( OPVs; including high-oil Eastern European cultivars ) , as well as oilseed and confectionery ( non-oil ) accessions from elsewhere in the world . Where necessary , accessions were advanced via single-seed descent for one or two generations to minimize residual heterozygosity . All accessions were assigned to one of ten categories based on their origin ( USDA or INRA ) , breeding history ( maintainer [B] lines = HA , typically unbranched; restorer [R] lines = RHA , typically branched ) , and agronomic use ( oil vs . non-oil ) . Note that an oilseed vs . confectionery designation was not available for the INRA accessions; therefore , these were divided into INRA-derived B and R lines ( denoted INRA-HA and INRA-RHA , respectively ) . For the USDA accessions , the following categories were defined: HA non-oil , HA oil , RHA non-oil , RHA oil , introgressed , OPVs , other non-oil , and other oil . Accessions designated ‘non-oil’ were either confectionery types , or could not be clearly defined as being oil types . The ‘introgressed’ category included accessions with a recent history of introgression from wild Helianthus species as indicated by the available pedigree information ( e . g . , [23] , [24] ) . The OPV category included named sunflower accessions that represent open-pollinated varieties of the pre-hybrid era of sunflower breeding , including Jupiter , Manchurian , Jumbo , VIR 847 , Mammoth , etc . ( BS Hulke , USDA-ARS , pers . comm . ) along with two Native American landraces , Hopi and Mandan . The ‘other oil’ and ‘other non-oil’ categories included accessions of each type for which a B vs . R designation could not be made . In the spring of 2010 , we planted our association mapping population in replicate at three locations: the Plant Sciences Farm in Watkinsville , GA , USA , the North Central Regional Plant Introduction Station in Ames , IA , USA , and the University of British Columbia Campus in Vancouver , BC , Canada ( 12 seeds/plot x 271 lines x 2 replicates x 3 locales ) ( Figure S1 ) . Replicates were planted in an alpha lattice design constructed using the computer module ALPHA 6 . 0 , available from Design Computing ( http://www . designcomputing . net/ ) . Total DNA was extracted from bulked tissue collected from four individuals of each line using a CTAB extraction protocol [25] . Total DNA was quantified using Picogreen ( ABI ) , and the quality of DNA was inspected using a Nanodrop 1000 spectrophotometer . All lines were then genotyped on an Illumina Infinium 10 k SNP array designed for cultivated sunflower . The array was designed from a large collection of sunflower ESTs and included no more than one SNP per gene [20] . Genotyping was performed according to the manufacturer's recommendations on the Illumina iScan System ( Illumina Inc . , San Diego , CA ) at the Emory University Biomarker Service Center . Prior to hybridization of the Beadchips , DNA was diluted to 50 ng/µl and quality was assessed via UV spectrophotometry and agarose gel electrophoresis . All SNP data analyses were performed using the raw intensity data from the Illumina Beadchip and Genome Studio ver . 2011 . 1 ( Illumina ) following the methods outlined in Bowers et al . [21] . Note that only those SNPs that showed clearly interpretable clustering patterns were used in this study , thereby eliminating probes that hybridized to multiple gene copies from our dataset . Map positions were obtained from the sunflower consensus map [21] . Population-wide estimates of genetic diversity , including allele frequencies , observed heterozygosity , and unbiased gene diversity [26] , were calculated using GenAlEx v . 6 . 4 [27] . Population structure was investigated using the Bayesian , model-based clustering algorithm implemented in the software package STRUCTURE [28] . For this analysis , we used only polymorphic SNPs with a minor allele frequency ( MAF ) ≥10% . Briefly , individuals were assigned to K population genetic clusters based on their multi-locus genotypes . Clusters were assembled so as to minimize intra-cluster Hardy-Weinberg and linkage disequilibrium and , for each individual , the proportion of membership in each cluster is estimated . We employed the admixture model without the use of prior population information ( i . e . , USEPOPINFO was turned off ) . For each analysis , we evaluated K = 1–12 population genetic clusters with 5 runs per K value and averaged the probability values across runs for each cluster . For each run , the initial burn-in period was set to 50 , 000 with 100 , 000 MCMC iterations . The most likely number of clusters was then determined using the DeltaK method of Evanno et al . [29] . Genetic relationships amongst the cultivated sunflower accessions were also investigated graphically via principal coordinates analysis ( PCoA ) using GenAlEx and the same set of polymorphic SNPs ( MAF ≥10% ) that were used for the STRUCTURE analyses . A standard genetic distance matrix [26] was constructed based on the multi-locus genotypes . This distance matrix was then used for the PCoA , and the first two principal coordinates were graphed in two-dimensional space . A relative kinship matrix was then estimated from this set of SNPs using the program SPAGeDi [30] . Negative values between pairs of individuals , indicating that there was less relationship than that expected between two randomly chosen individuals , were set to 0 in the resulting matrix . To investigate the extent of linkage disequilibrium ( LD ) across the genome , a correlation matrix of r2 values , the squared allele frequency correlations , was constructed between all possible pairs of polymorphic loci with MAF ≥10% . Following the methods of Macdonald et al . [31] , we summarized the observed r2 values using the k-smooth function in the statistical programming language R ( http://www . R-project . org/ ) . We also visualized the extent of LD and genetic variation across the genome by averaging the r2 and UHe values , respectively , in a 5 cM sliding window across each linkage group ( LG ) . We calculated FST for all polymorphic SNPs between the two major heterotic groups ( including 125 RHA lines and 100 HA lines ) and plotted the results as a function of map position . We also performed outlier analyses on these data using the software BayeScan [32] . The program utilizes the Bayesian model from Beaumont and Balding ( 2004 ) and a reversible jump Markov chain Monte Carlo method to identify outlier loci that are putatively under selection based on FST estimates . Ten pilot runs of 5 , 000 iterations and an additional burn-in of 50 , 000 iterations were first performed . We then used 100 , 000 iterations to identify loci under selection based on locus-specific Bayes factors . Strong evidence for selection is indicated by a Bayes factor above 10 , or log10 = 1 . 0 [32] . In order to visualize genome-wide haplotypic structure in the association population , graphical genotypes were constructed by defining haplotypic blocks of 25 or more consecutive SNPs ( based on the map order from [21] ) that were identical between two or more cultivars . To do this , each accession was compared to all other accessions in the dataset to determine the percentage of the SNPs contained in shared haplotypic blocks . The accession with the highest fraction of the genome shared with all other accessions in the data set was set as the “template” for genotype #1 ( G1 ) . The raw data for the template accession and all matching haplotypic blocks in other accessions were converted into G1 from the raw scores . This process was repeated for 24 additional cycles , masking the data that had previously been assigned to haplotypic blocks to produce G2 , G3 , … , G25 . The most common genotypes were then color-coded and visually presented using spreadsheet software . The number of days to flower ( DTF; calculated from the planting date ) was recorded at the R-5 . 1 reproductive stage . The R-5 stage commences at the onset of flowering , and is divided into substages according to the percentage of disc florets that have opened; R-5 . 1 corresponds to the stage at which 10% of the disc florets have opened . The total number of branches per plant ( hereafter referred to as “branching” ) was measured in the field at the R-9 reproductive stage . This stage is regarded as physiological maturity and is characterized by the presence of yellow/brown bracts on the back of the sunflower head . Four plants per accession were scored for each replicate at each of three locations ( 4×271×2×3 = 6 , 504 plants scored ) . Data were analyzed using SAS software version 9 . 3 ( SAS Institute , Cary , NC ) . We calculated Pearson pairwise correlation coefficients for the branching data and the average DTF across locations using PROC CORR and corrected the resulting significance levels for multiple tests using a sequential Bonferroni correction ( Holm 1979 ) . We also analyzed our data using the GLM procedure of SAS . Because our initial analyses revealed highly significant ( P<0 . 001 ) genotype x environment ( G×E ) interactions ( data not shown ) , all subsequent analyses were performed separately by location . At each location , the entry ( i . e . , genotype ) was treated as a fixed effect and blocks and reps were treated as random effects . For branching , the block and rep effects were not significant in the model and thus raw means were used for association testing ( below ) . Because there were significant block and rep effects for DTF , the least-squares means ( LS means ) were used for the association mapping of this trait . Finally , variance components using the VARCOMP function in SAS were calculated for branching and DTF and were used to estimate broad-sense heritabilities ( H2 ) as the total genotypic variance divided by the total phenotypic variance . Association mapping of branching and DTF were performed in the software package TASSEL v . 3 . 0 [33] using all SNPs with MAF ≥10% . Three different association mapping models were run for each trait including a mixed linear model ( MLM ) accounting for kinship ( i . e . , familial relatedness; K-matrix ) and two MLMs using kinship and population structure as estimated via either principle component analysis ( PCA ) ( P-matrix ) or the program STRUCTURE ( Q-matrix ) [28] , [29] . Model effects for individual SNPs were output from TASSEL for each MLM [33] . Linear model testing was performed by plotting the observed P-values from the association test against an expected ( cumulative ) probability distribution . These quantile-quantile ( q-q ) plots indicate the extent to which the analysis produced more significant results than expected by chance . Models that follow the expected line more closely are assumed to have produced fewer false positives . Given that non-independence of linked makers in the dataset could lead to overly conservative significance thresholds [34] , we used the multiple testing correction method of Gao et al . [35] to evaluate the significance of our results . This approach accounts for correlations amongst markers while controlling the type I error rate ( alpha = 0 . 05 ) . Using both real and simulated data , this correction has been shown to be an efficient and accurate method of minimizing false positives in the presence of inter-marker LD . Where possible , to enable the identification of novel genetic effects , we also compared the genetic map positions of significant associations to those of previously mapped branching and flowering time QTL . This was done by projecting QTL onto the sunflower consensus map ( which , as noted above , was also used for ordering the SNPs employed in the present study ) based on shared markers . Note that some previous QTL results could not be included in this comparison due to a lack of shared markers and/or differences in linkage group nomenclature . The linkage group names were standardized by Tang et al . [36] , though the new naming scheme was not immediately adopted by all researchers .
The total number of readily scorable , bi-allelic SNPs in the focal population was 5 , 788 . The number of SNPs with a MAF ≥10% was 5 , 359 . Expected heterozygosity , or Nei's unbiased gene diversity , averaged 0 . 404±0 . 005 ( mean ± standard error ) , and ranged from 0 . 007 to 0 . 5 . Observed heterozygosity averaged 0 . 034±0 . 0044 , and ranged from 0 to 0 . 38 . Gene diversity and observed heterozygosity for each line classification grouping are found in Table 1 . Our STRUCTURE results using the full set of SNPs with MAF ≥10% indicated that K = 3 ( hereafter referred to as Q = 3 corresponding to the Q matrix for the association testing results below ) , providing support for the existence of three genetically distinct clusters in our association panel . STRUCTURE results are grouped and graphed according to the line classifications ( see Methods ) in Figure 1 . DeltaK and the mean likelihood values are plotted in Figure S2 . Clusters one and three largely consist of the maintainer ( HA ) lines whereas the majority of the restorer-oil ( RHA-oil ) lines exhibit substantial membership in cluster two . The PCoA analysis was largely consistent with the STRUCTURE results ( Figure S3 ) . In order to simplify the PCoA plot , we combined categories by grouping the lines into either HA , RHA-nonoil , RHA-oil , or other ( this category contained all remaining lines/accessions ) . The RHA-oil lines are generally separated from the balance of the cultivated germplasm along the first and second axes , while the HA lines are generally distinct along the first axis . Relative kinship was also estimated using the full set of markers ( MAF ≥10% ) . Approximately 60% of the pairwise kinship estimates were near zero ( i . e . , less than 0 . 005 ) , indicating the lines were essentially unrelated ( Figure S4 ) . The remaining estimates ranged from 0 . 05 to just less than 1 , with a rapidly decreasing number of sunflower pairs exhibiting higher levels of relatedness . The results of our genome-wide analysis of linkage disequilibrium ( LD ) are summarized in Figure 2 and Figure 3 and Figure S5 . Figure 2 displays an LD matrix of the squared allele frequency correlations ( r2 ) plotted for the ordered markers ( MAF ≥10% ) ; the x- and y-axes correspond to the 17 LGs in sunflower . Regions of the genome where LD extends for considerable genetic map distance are visible as yellow to red squares on the figure ( e . g . , on LGs 5 and 10 ) . Figure S5 shows plots of the squared allele frequency correlations ( r2 ) for 10 , 000 random pairs of SNPs within 50 cM as a function of genetic map distance between SNPs for the 17 LGs . Looking across chromosomes , different overall patterns of LD are apparent . For example , LG 10 exhibits a relatively slow overall decay in LD , largely due to strong haplotypic structure across a portion of the chromosome ( see also Figure 2 and Figure 3 ) , whereas LG 11 shows a much more rapid decay of LD . Following the methods of Macdonald et al . [31] , the red line on each graph in Figure S5 summarizes the observed r2 values as a function of map distance using the ksmooth function in the statistical software R ( http://www . R-project . org/ ) . On a per chromosome basis , the average genetic distance at which r2 dropped below 0 . 1 ranged from 6 . 95 cM to 12 . 6 cM with the medians ranging from 3 . 93 cM to 10 . 1 cM . The sliding window analysis of r2 further illustrates the variability in LD across the genome ( Figure 3 ) . In some cases , entire chromosomes show very low levels of LD . In other cases , elevated LD is visible in specific chromosomal regions , including portions of LGs 1 , 5 , 8 , 10 , and 13 . The sliding window analysis of genetic diversity likewise revealed variation in UHe across the genome ( Figure S6 ) . We also estimated FST between the two primary breeding pools ( i . e . , RHA vs . HA lines ) for the full set of markers ( MAF ≥10% ) , plotted the results against genetic map position ( Figure 4 ) , and tested for selection using BayeScan . Genomic regions exhibiting significantly elevated differentiation are visible as spikes in FST ( with individually significant markers being colored in red ) on several chromosomes , including portions of LGs 8 , 10 , and 13 , and a single marker on LG 14 . Our association mapping population exhibited substantial phenotypic diversity for both plant architecture and flowering time , expressed here in terms of branching and DTF . Significant positive correlations were found for both branching and flowering time across locations ( i . e . , more or less branched lines tended to behave similarly across locations , and the same was seen for DTF in terms of earlier vs . later flowering lines ) , whereas there was an overall significant negative correlation between branching and DTF ( i . e . , more highly branched plants tended to flower earlier; Figure S7 ) . As noted above , there was a significant G×E interaction ( P<0 . 01 ) for both traits studied . An important consequence of such G×E interactions is that different associations may be detected across environments; thus , we performed the association analyses separately for each location . The estimates of broad-sense heritability for branching and DTF were 0 . 861 and 0 . 124 , respectively . The number of plants that showed a complete lack of branching was 79 , 110 , and 70 in Georgia , Iowa , and British Columbia respectively . The number of branches per genotype averaged 7 . 3 , 7 . 2 , and 7 . 6 and ranged from 1–29 , 1–19 , and 1–30 in Georgia , Iowa , and British Columbia respectively . DTF averaged 57 . 1 ( range 41–77 ) , 68 . 7 ( 45–95 ) , and 80 . 4 ( 63–104 ) in Georgia , Iowa , and British Columbia respectively . In terms of branching , our analyses revealed significant associations on LGs 2 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 12 , 13 , 14 , and 17 ( Figure 5; Table 2; Table S2 ) . In general terms , each of these LGs had a single peak of significant marker-trait associations . The exceptions were: LG 7 , which had two peaks in IA ( one near 10 cM and another near 53 cM ) ; LG 10 , which had a primary peak centered near 25 cM in all locations and a secondary peak near 74 cM in IA and BC; LG 13 , which had a primary peak near 41 cM and a secondary peak near 64 cM in all locations , as well as a third peak near 5 cM in BC; and LG 17 , which had a primary peak near 41 cM in GA and IA and a secondary peak near 21 cM in IA . For DTF , our analyses revealed significant associations on LGs 1 , 3 , 4 , 9 , 10 , 12 , 13 , and 17 ( Figure 6; Table 3; Table S2 ) . Here again , each of these LGs typically had a single peak of significant associations . The exception was LG 13 , which had a primary peak near 70 cM in GA , a secondary peak near 3 cM in GA , and a third peak near 21 cM in IA . Table 2 and Table 3 summarize these results across locations for the three mixed models . Figure 5 and Figure 6 ( upper panels ) show the Manhattan plots for branching and DTF , respectively , in each of the three locations . For each location , the three mixed models are plotted as kinship ( K , red ) , population structure as measured by PCA plus kinship ( P+K , blue ) , and population structure as measured by STRUCTURE plus kinship ( Q+K , dark grey ) . Points above the dashed threshold line are significant after correcting for multiple tests , as detailed above . Figure 5 and Figure 6 ( lower panels ) also show the quantile-quantile ( q-q ) plots of the observed P-values versus the expected for each of the three models as well as a naive model that does not account for population structure or kinship . As can be seen from the q-q plots , the distribution of observed P-values in the naive model greatly deviated from the expected distribution whereas the other models followed the expected distribution much more closely . This result reflects the potential confounding effects of population structure and relatedness in the dataset . Note , however , that for DTF in BC , two of the models ( P+K and Q+K ) provided fewer significant results than expected by chance , suggesting that these models may be overly conservative ( e . g . , [37] ) . The full set of results , including functional annotations for the genes from which the SNPs were derived ( see [20] ) , significance values , and SNP effects for all individual loci at each location and using all three models , are provided in Table S2 . The graphical genotypes for all 17 LGs across the full population of 271 accessions are presented in Figure S8 . The genotype with the highest fraction of shared haplotypes across the genomes of the 271 lines ( G1 in red ) corresponds to the accession HA89 , a line of great historical importance in sunflower breeding . HA89 accounted for an average of 16 . 2% of the genomes of the 271 lines . Overall , the 25 most common genotypes accounted for an average of 63 . 5% of the genomes of the 271 lines examined . For ease of interpretation , only the top nine genotypes are color-coded ( beyond this point , each additional genotype individually accounted for ca . 1% or less of the genome ) ; in total , these nine genotypes accounted for an overall average of 50 . 7% of the genomes of the 271 lines ( see Table S3 ) . White regions either correspond to non-major genotypes or reflect stretches with fewer than 25 consecutive , homozygous SNPs . Figure 7 depicts the results for LG 10 with the data sorted by the average number of branches produced by plants of each accession at all three locations ( see heat map along the top ) . As noted above , the upper portion of this LG exhibits strong haplotypic structure and elevated LD across an extended region along with a major effect on branching at all three locations . See below for a detailed discussion of the historical and biological significance of these results .
The work presented herein represents the largest and most comprehensive analysis of population genomic diversity in cultivated sunflower to date . In terms of overall levels of SNP diversity , our data indicate that sunflower exhibits considerable molecular variation , on par with estimates derived from large-scale SNP surveys of other crops ( e . g . , maize [38]–[40] , barley [41] , and rice [42] ) . We also documented substantial phenotypic variation in terms of both plant architecture and flowering time , ranging from a complete lack of branching to whole plant branching and including accessions that reached reproductive maturity over a period spanning greater than 30 days . Thus , despite the population genetic bottlenecks that are known to have occurred during domestication and improvement , the cultivated sunflower gene pool harbors substantial variability . In terms of overall population structure , our results are largely in agreement with our prior analyses based on a much smaller set of simple-sequence repeat markers ( SSRs ) [22] . This general agreement between our current findings and the earlier , SSR-based work suggests that any possible ascertainment bias during SNP discovery and selection had minimal effects on our population genetic results . In fact , the preferential usage of SNPs with high heterozygosity would be expected to result in an underestimate of the magnitude of structure [43] but the FST estimates between the B and R lines is virtually identical between the SNPs ( FST = 0 . 049 ) and the SSRs ( FST = 0 . 047 ) . The much larger number of markers in the present study has , however , allowed us to refine our earlier findings . Notably , we found evidence for the presence of three genetically distinct groups within the germplasm collection . One of these groups was primarily composed of the RHA lines , while a second group consisted of a large number of HA lines , and the third included a more diverse assemblage of lines . The PCoA likewise demonstrated a split between the RHA and HA lines , with an even clearer division between the RHA-oil and HA lines . This genetic distinction between B and R lines is expected given the breeding history of sunflower , which involves the maintenance of distinct gene pools to maximize heterosis in hybrid crosses [44] , [45] . Taken together , these results underscore the need to account for population structure when performing association analyses in sunflower . Our analysis of LD revealed considerable variability across the genome . In most regions , LD declined quite rapidly as a function of genetic distance and the correlation between most pairs of SNPs fell to negligible levels ( i . e . , r2≤0 . 10 ) within 3 cM . In some instances , however , LD remained elevated ( on average ) over greater distances ( e . g . , LG 10; see Figure S5 ) . Inspection of Figure 2 and Figure 3 reveals the existence of a number of localized islands of LD , including blocks on LGs 1 , 5 , 8 , 10 , 13 , 16 , and 17 . In looking more closely at these localized regions exhibiting elevated LD , it is apparent that many of them occur in close proximity to genes or QTL underlying traits that have been targeted by selection during sunflower domestication and/or improvement . Given the history of breeding for resistance to diseases in sunflower [46] it is noteworthy that the four of these spikes in LD ( on LGs 1 , 5 , 8 , and 13 ) co-localize with QTL and/or candidate genes for resistance to several important diseases . Note that co-localization of QTL and LD spikes/associations ( below ) is based upon concordance of shared genetic markers ( most often microsatellites ) between the previous QTL map ( s ) and the map of Bowers et al . [21] . More specifically , QTL and/or genes for resistance to downy mildew ( Plasmopara halstedii; [47]–[52] ) co-localize with spikes in LD on LGs 1 and 8 . Similarly , the block of elevated LD observed on LG 5 co-localizes with a QTL for resistance to black stem ( Phoma macdonaldii , [53] ) . Finally , the spike in LD on LG 13 co-localizes with sunflower rust resistance genes [50] , [52] , [54] . It thus appears that selection for disease resistance may have played a role in shaping genome-wide patterns of LD in sunflower , though we cannot rule out the possibility of selection on other traits ( see below ) . The important role that selection on plant architecture played during the domestication and subsequent improvement of sunflower also appears to have shaped patterns of genetic diversity across the sunflower genome , especially with respect to LG 10 . During the initial domestication of sunflower , unbranched , monocephalic landraces were preferentially propagated and the domesticated lineage moved away from the intensely branched , multi-headed phenotype that is characteristic of wild sunflower [17] , [55]–[57] . Until the mid-20th century , modern cultivars were thus typically unbranched; however , beginning in the late 1960s , the transition to hybrid breeding and the associated desire for a prolonged flowering period in male lines resulted in the re-introduction of branching into the sunflower gene pool [45] . This resulted in selection favoring the fixation of a recessive branching allele at the so-called B-locus in R lines . The B-locus has since been mapped to approximately 27 cM from the top of LG 10 [58] . In viewing the graphical genotypes , the B-locus is visible as differentiated haplotypic blocks that span this region on LG 10 and which clearly correlate with the extent of branching ( Figure 7 ) . Interestingly , the re-introduced branching haplotype ( in dark blue ) spans ca . 25 cM , whereas the unbranched haplotype ( primarily in red ) appears to span ca . 10 cM . Thus , the effects of the very recent re-introduction of branching to the cultivated gene pool can be visualized as a large haplotypic block presumably resulting from a recent , strong selective sweep in the branched R ( RHA ) lines . This pattern can also be seen in the sliding window analyses of LD ( r2 ) and the plots of population differentiation ( FST ) vs . genetic map position . In both cases , spikes are clearly visible in that same region along LG 10 ( Figure 3 and Figure 4 ) . Notably , our BayeScan analysis indicated that this region of the genome , along with portions of LG 8 and 13 and a single marker on LG 14 , exhibits significantly elevated FST ( Figure 4 ) . This finding is consistent with the notion that this differentiation – which spans a remarkably long ( at least in genetic terms ) and otherwise recombinogenic genomic interval – was driven by strong selection . Not surprisingly , this same region of LG 10 harbored highly significant associations for branching in all three locations , as well as for DTF in GA . Overall , we found significant associations for branching in 17 genomic regions on 12 of the 17 LGs in sunflower . In five cases , these associations overlapped with previously identified QTL for branching in sunflower on LGs 10 ( a ) , 12 , 13 ( a and b ) and 17 ( Figure S9 ) [9] , [59]–[61]; the remainder were novel effects for number of branches that have not previously been documented . Similarly , we found significant associations for DTF in 10 genomic regions located on 8 of the 17 sunflower LGs ( Figure S10 ) [9] , [62]; the remainder ( on LGs 1 , 3 , 4 , 10 , 12 , 13 ) were novel effects . It is noteworthy that the detected associations generally spanned much smaller intervals than are typical of traditional QTL studies . This result is consistent with our finding that LD decays relatively rapidly across much of the genome ( see above ) and suggests that even higher marker densities would be desirable for developing a complete picture of trait variation in sunflower . In terms of the relationship between marker-trait associations and both LD and population differentiation , the branching and DTF associations co-localized with spikes in r2 and FST ( on LGs 8 , 10 , and 13; Figure 3 and Figure 4 ) . As noted above , all three of these regions , along with a single marker on LG 14 , exhibited significantly elevated FST values , suggestive of selective divergence . Given the historical importance of the B-locus and the clear correspondence of the observed haplotypes to sunflower breeding groups ( and thus branching architecture; see above ) , it seems likely that the driving force behind the pattern observed on LG 10 was selection on branching . The situation on LGs 8 and 13 is less clear; the observed patterns may have been driven be selection on branching , flowering time , disease resistance , or some combination thereof . It must be kept in mind , however , that the branching associations that we detected on LG 8 were most apparent in the kinship-only model , and disappeared almost entirely when we controlled for population structure . As such , this result in particular may have been a byproduct of population structure as opposed to a true functional association . It is particularly noteworthy that the majority of significant associations identified herein are located in the aforementioned islands of LD . Given that our analyses relied on a single SNP in each of ca . 5 , 300 genes , we are almost certainly missing out on associations in regions of low LD . In fact , nearly 50% of the sliding windows analyzed had an average r2<0 . 1 and over 85% had an average r2<0 . 2 . As such , future analyses aimed at assaying genetic variation at a much higher density ( e . g . , using genotyping-by-sequencing [63] or even whole genome re-sequencing ) seem warranted and are likely to facilitate a much more detailed characterization of the molecular basis of phenotypic variation in sunflower . | Selection during the evolution of crop plants has resulted in dramatic phenotypic differentiation , and these same selective pressures are expected to have had a significant impact on underlying genomic diversity . Population genomic analyses , especially when coupled with trait-based mapping approaches , thus have the potential to provide unique insights into the evolution of crop plants and their genomes . In this study , we performed a genome-wide analysis of genetic variation in cultivated sunflower and used the resulting data to genetically dissect variation in plant architecture ( i . e . , branching ) and flowering time . We found substantial variation in levels of linkage disequilibrium ( LD ) across the genome , with islands of elevated LD generally corresponding to genomic regions underlying traits that have been targeted by selection during the evolution of cultivated sunflower . A number of these same regions also exhibited strong population genetic differentiation across the sunflower gene pool , suggesting that they may harbor genes underlying adaptation following domestication . Our analyses also identified numerous genomic regions underlying variation in both plant architecture and flowering time , many of which fall in genomic regions that have not previously been shown to influence these traits using more traditional quantitative genetic approaches . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"genome-wide",
"association",
"studies",
"genomics",
"plant",
"biology",
"crops",
"genetics",
"population",
"genetics",
"biology",
"evolutionary",
"biology",
"genomic",
"evolution",
"genetics",
"and",
"genomics",
"agriculture"
] | 2013 | Association Mapping and the Genomic Consequences of Selection in Sunflower |
The Epstein-Barr virus ( EBV ) is a B lymphotropic virus that infects the majority of the human population . All EBV strains transform B lymphocytes , but some strains , such as M81 , also induce spontaneous virus replication . EBV encodes 22 microRNAs ( miRNAs ) that form a cluster within the BART region of the virus and have been previously been found to stimulate tumor cell growth . Here we describe their functions in B cells infected by M81 . We found that the BART miRNAs are downregulated in replicating cells , and that exposure of B cells in vitro or in vivo in humanized mice to a BART miRNA knockout virus resulted in an increased proportion of spontaneously replicating cells , relative to wild type virus . The BART miRNAs subcluster 1 , and to a lesser extent subcluster 2 , prevented expression of BZLF1 , the key protein for initiation of lytic replication . Thus , multiple BART miRNAs cooperate to repress lytic replication . The BART miRNAs also downregulated pro- and anti-apoptotic mediators such as caspase 3 and LMP1 , and their deletion did not sensitize B-cells to apoptosis . To the contrary , the majority of humanized mice infected with the BART miRNA knockout mutant developed tumors more rapidly , probably due to enhanced LMP1 expression , although deletion of the BART miRNAs did not modify the virus transforming abilities in vitro . This ability to slow cell growth could be confirmed in non-humanized immunocompromized mice . Injection of resting B cells exposed to a virus that lacks the BART miRNAs resulted in accelerated tumor growth , relative to wild type controls . Therefore , we found that the M81 BART miRNAs do not enhance B-cell tumorigenesis but rather repress it . The repressive effects of the BART miRNAs on potentially pathogenic viral functions in infected B cells are likely to facilitate long-term persistence of the virus in the infected host .
The Epstein-Barr virus ( EBV ) is a strongly B lymphotropic virus that infects the majority of the world human population and is associated with the development of malignant tumors , mainly lymphomas and carcinomas of the nasopharynx ( NPC ) and of the stomach [1] . Shortly after infection , B cells start to divide and generate continuously growing cell lines , commonly termed lymphoblastoid cell lines ( LCLs ) [2] . Infected cells express a set of latent proteins ascribed to subfamilies known as Epstein-Barr virus nuclear antigens ( EBNA ) and latent membrane proteins ( LMP ) , most of which are essential or strongly potentiate the B cell transformation process [1] . EBV also encodes 44 miRNAs that are divided into two clusters located around the BHRF1 gene ( BHRF1 miRNAs ) or within the introns of the BART gene ( BART miRNAs ) [3–5] . Viruses devoid of the BHRF1 miRNA locus are less transforming than their wild type counterparts [6–9] . Indeed , a recombinant virus that lacks this cluster retains only 1/20th of the wild type transforming capacity [8] . The BART miRNAs are present in all EBV-infected cells , but their expression level is up to hundred times higher in epithelial cells than in infected B lymphocytes , suggesting that they exert their main function in the former type of cells [5 , 10] . EBV-associated carcinomas produce only a restricted number of latent proteins but also the BART miRNAs , making these non-coding RNAs prime suspects in the transformation process [11] . Indeed , miR-BART9 and miR-BART7-3p have been found to promote metastasis of NPC cells [12 , 13] . Reciprocally , anti-miR-BART7-3p reduced tumor growth in an animal model [13] . In the same vein , miR-BART3* was found to target the tumor suppressor gene DICE1 in NPC cells [14] . In the primary B cell system , Vereide et al . used a B95-8 EBV strain virus with a reconstituted BART locus to show that the BART miRNAs also improve the transforming abilities of the virus [9] . The BART miRNAs have been found to regulate apoptosis by targeting pro-apoptotic proteins such as PUMA and BIM in epithelial cells [15 , 16] . A photoactivatable ribonucleoside-enhanced crosslinking and immunoprecipitation ( PAR-CLIP ) screening of the EBV-positive NPC cell line C666 revealed that the antiapoptotic properties of the BART miRNAs can be ascribed to the direct targeting of 10 pro-apoptotic proteins in these cells [17] . The BART miRNAs have also been suggested to act as repressors of EBV lytic replication in B cell or epithelial cell lines induced with drugs such as TPA . MiR-BART6 , through its ability to target DICER , and miR-BART20-5p through targeting of BZLF1 , control entry into the EBV lytic replication phase [18 , 19] . MiR-BART18-5p also controls the onset of replication in anti-Ig-treated Akata Burkitt’s lymphoma cell lines and in LCLs induced by TPA through its ability to target the expression of MAP3K2 [20] . This is in agreement with the view that EBV-infected B cells hardly replicate the virus and are mainly latent , whilst infected epithelial cells are the main sites of replication [2] . These data have frequently been collated in tumor cells or in LCLs infected by viruses that carry a partial deletion of the BART miRNAs and might not extend to strains with an intact locus , in particular in vivo . Furthermore , the view that LCLs are primarily latent and must be stimulated to produce virus is restricted to viral strains close to B95-8 . We have recently shown that M81 , a viral strain that carries a high degree of homology with viruses found in NPC and that infects a substantial proportion of the Chinese population , induces a high degree of spontaneous virus replication upon infection of primary B cells [21] . Importantly , this virus carries an intact BART locus . Both features render M81 a suitable experimental system to study the function of the BART locus in infected primary B cells . Here we report the phenotypic traits of a set of viruses that are devoid of different subsets of the BART miRNA cluster in the fully permissive M81-based replication system .
We set out to determine the role of the BART miRNAs by infecting B cells with a virus devoid of this locus . To this end , we sequentially deleted the BART subcluster 1 ( M81/ΔC1 ) , subcluster 2 ( M81/ΔC1C2 ) and miR-BART-2 ( M81/ΔAll ) ( S1 Fig ) . Importantly , this mutant retains intact RPMS1 , LF1 and LF2 exons . We then generated a revertant virus by reintroducing the complete BART locus in M81/ΔAll ( M81/ΔAll/rev ) . We also individually deleted the subcluster 2 or miR-BART2 in the wild type genome to generate M81/ΔC2 and M81/Δb2 ( S1 Fig ) . All these recombinants were stably introduced into 293 cells to generate producer cells lines that carry intact copies of the mutants , and were accordingly termed 293/M81/ΔAll et cetera ( S1 Fig ) . We began by infecting B cells isolated from the peripheral blood with M81 , M81/ΔAll and M81/ΔAll/rev to generate a panel of lymphoblastoid cell lines ( LCLs ) . We measured BART miRNA expression in these cells for miR-BART1-3p , miR-BART7* and miR-BART2-5p that are expressed at good levels in these LCLs and are representative of each BART subcluster [22] . This confirmed that the cells infected by the mutants did not express the miRNAs that had been deleted ( S2A Fig ) . Although the construction of the ΔAll mutant left the LF1 , LF2 and LF3 genes intact , we also quantified expression of the BART mRNA by RT-PCR in LCLs generated with M81/ΔAll or wild type M81 . This assay revealed that the transcript is produced in cells infected by the ΔAll mutant , on average at marginally higher levels relative to wild type ( S2B Fig ) . However , the difference was not statistically significant ( S2B Fig ) . We then assessed lytic replication in the LCL panel as we previously reported that M81 spontaneously replicates in B cells [21] . We first gauged BZLF1 by western blot in 18 donors ( Fig 1A and S3 Fig ) . These assays revealed that expression of this protein is enhanced by an average 3 . 4 fold in the absence of the BART miRNAs , compared to LCLs infected with wild type EBV or the M81/ΔAll revertant . In two of these cases ( samples F and I ) , the BZLF1 expression was very close or even higher in the LCL infected by the wild type virus ( S3A Fig ) . We then attempted to complement the phenotype of LCLs transformed by M81/ΔAll by transfecting them with a plasmid that encodes all BART miRNAs with the exception of miR-BART-b2 , as well as a truncated nerve growth factor receptor ( NGFR ) that is expressed at the surface of transfected cells , or with a control vector . We first used qPCR to confirm that NGFR antibody-purified cells transfected with the BART miRNAs express them , as shown in Fig 1B , and then performed a western blot with the same cells . Cells transfected with the BART expression vector exhibited lower levels of BZLF1 protein than those transfected with control plasmids , confirming that expression of the BZLF1 protein is modulated by the BART miRNAs . We then performed immunofluorescence stains ( IF ) with the same antibody at different time points that showed that the number of BZLF1-positive cells was , on average , 2 to 3 times higher in cells infected with the BART miRNA KO mutant , relative to wild type controls ( p<0 . 02 ) ( Fig 1C ) . However , the intensity of the staining at the single cell level was not significantly stronger in BZLF1-positive cells . To confirm this observation , we performed a western blot with protein extracts from LCLs infected with the M81/ΔAll mutant or wild type virus , normalized by the number of BZLF1-positive cells in the LCL infected by the wild type control , as determined by IF ( Fig 1D ) . Thus , these samples contained approximately the same number of BZLF1-positive B cells . Both samples generated BZLF1 signals of similar intensity , confirming that the replicating cells from LCLs infected by wild type or ΔAll mutants produce similar amounts of the BZLF1 protein . A RT-qPCR with BZLF1-specific primers performed on the LCL panel revealed that mRNA BZLF1 expression was higher in the LCLs infected with the miRNA mutant , particularly 40 days after infection but did not exceed a factor of 2 . 5 over time ( Fig 1E ) . Taking into account that excision of the BART miRNAs results in an increase of the number of BZLF1-positive cells but not in an increase in protein expression at the single cell level , the likeliest explanation for the enhanced BZLF1 mRNA production is that an increased number of cell produced it , although a direct effect of the BART miRNAs on RNA stability could have contributed to this process . Altogether , we conclude that the BART miRNAs contribute to the repression of spontaneous reactivation of the lytic cycle in infected cells , but that their absence does not substantially increase BZLF1 protein expression in the individual replicating cells . We then studied expression of the BZLF1 protein over time . B cells infected by a virus that lacks BZLF1 and BRLF1 ( ΔZR ) served as a negative control [21] . We performed immunoblots at day 22 , 43 and 84 dpi that showed an overall decrease in the expression of this protein . However , the LCLs infected with the M81/ΔAll remained longer strongly BZLF1-positive ( Fig 1F ) . We assessed the consequences of increased BZLF1 expression by staining the infected LCLs with antibodies against gp350 . Both the western blot and the IF stains showed a higher number of gp350-positive cells in cells infected with the ΔAll mutant than in those infected with the wild type virus ( Fig 2A and 2B ) . It is important to note that the large majority of the viruses that are produced by the replicating cells bind to their neighbor B cells , some of which become covered with gp350-positive signals , making it difficult to distinguish them from producer cells [21] . Thus , the results presented here include both types of gp350-positive cells . We also measured the viral copy numbers in supernatants from LCLs by qPCR and found them clearly increased in those from LCLs generated with M81/ΔAll ( Fig 2C ) . We then infected primary B cells with these LCL supernatants at low cell density to prove that the increased viral titers were a consequence of enhanced virus production . Indeed , this assay showed an increase in the number of outgrowing clones after treatment with supernatants from the M81/ΔAll LCL , although the efficiency of transformation widely varied between different supernatants ( Fig 2D ) . MiRNAs typically modulate expression of target genes by forming the RNA-induced silencing complex ( RISC ) that includes miRNAs , their mRNA targets and a member of the Argonaute family of proteins . Therefore , we tested whether the deletion of the BART miRNAs modifies the recruitment of BZLF1 mRNA to the RISC . We began by measuring the expression levels of DICER in cells lacking the BART miRNAs , as this protein has been suggested as a target of the BART miRNAs [19] . A western blot performed on three pairs of LCLs generated from three different blood donors and infected with either M81 or M81/ΔAll showed an increase in DICER protein expression ( Fig 3A ) . However , we gauged the expression levels of three cellular miRNAs expressed in B cells but could not identify any differences between LCLs transformed with wild type EBV or with the ΔAll mutant ( Fig 3B ) . We conclude that the impact of the BART miRNAs on DICER in infected B cells has no generalized and pronounced functional consequences . We performed RISC immunoprecipitations in couples of LCLs infected with wild type or ΔAll mutant using an antibody directed against Ago2 ( S4A Fig ) and measured mRNA expression by qPCR in the precipitates . We assessed the efficacy of this protocol by comparing expression of the EBV miR-BHRF1 miRNAs in the Ago2 antibody precipitate with the expression in untreated LCLs . This assay showed more than 10000-fold enrichment of this miRNA after immunoprecipitation ( S4B Fig ) . An immunoprecipitation with an antibody directed against BrdU performed in parallel measured the non-specific background mRNA recovery . We measured the expression levels of GAPDH , HPRT , IPO7 , LMP1 and BZLF1 mRNAs in the RISC of LCLs transformed with M81 or M81/ΔAll . GAPDH expression levels were used to normalize for mRNA recovery , HPRT has previously found not to be recruited in the RISC by BARTs , whereas the IPO7 mRNA is a previously validated BART target [9 , 23–25] . We also quantified expression of these mRNAs in infected cells . Comparison with the expression levels obtained after immunoprecipitation determines the efficacy of recruitment to the RISC . A prerequisite for this analysis is that the number of mRNA molecules than can be recruited to the RISC does not vary too much between the cells , so as to exclude saturation effects . However , we have seen that the increase in BZLF1 protein expression results from an increased number of replicating cells that all produce the protein at approximately the same level . Thus , it is unlikely that the BZLF1 mRNA expression level and recruitment to the RISC will vary significantly between different replicating cells . The results of these experiments are given in Fig 3C and 3D , respectively . They first show the relative amounts of mRNAs in the RISC after subtraction of the background generated by the BrdU antibody . This experiment shows that BZLF1 , LMP1 , and to a lesser extent IPO7 mRNAs are more abundant in the RISC of LCLs infected with M81 wild type virus . In contrast , the levels of HPRT mRNAs in the RISC were similar in LCLs infected with either type of virus ( Fig 3C ) . We then calculated the level of enrichment of these mRNAs into the RISC by comparing expression levels after Ago2 immunoprecipitation or in untreated cells ( Fig 3D ) . This figure clearly shows that the relative recruitment in the LCLs infected with wild type viruses was more efficient for BZLF1 , LMP1 and IPO7 than for HPRT . We conclude that the BART miRNAs recruit the first three mRNAs to the RISC . However , these genes are still detectable in the RISC of LCLs infected with the ΔAll mutant . Therefore , other miRNAs , presumably miRNAs of cellular origin , recruit these mRNAs to the RISC . We wished to confirm these results by performing luciferase reporter assays . To this end , we constructed a pGL4 . 5-based luciferase reporter plasmid that carries the BZLF1 3’UTR . We also fused the luciferase gene to the BALF5 3’UTR that has previously been reported to be directly targeted by BART-2 miRNA [26] . The luciferase expression plasmid devoid of 3’UTR was used as a negative control . These plasmids were cotransfected with a rat-CD2 expression plasmid into two independent sets of LCLs generated with either of M81 and M81/ΔAll to determine the transfection efficiency and allow normalization . We measured the luciferase activity in transfected cells and found that the luciferase activity for both plasmids carrying the BZLF1 or the BALF5 3’UTR was reduced by 40 to 60% in the wild type LCL , relative to the LCL generated with the ΔAll mutant ( Fig 3E ) , a result previously reported for BALF5 [26] . The control plasmid pGL4 . 5 showed no major difference in expression ( Fig 3E ) . These results suggest that M81 LCLs recruit more BZLF1 mRNA to the RISC than M81/ΔAll LCLs by targeting the BZLF1 3’UTR . However , the low transfection rates achieved in LCLs ( 1 to 2% ) indicate that we should exercise caution when interpreting these results The data gathered so far indicated that the BART miRNAs repress lytic replication in primary B cells . However , B cells infected with wild type EBV undergo lytic replication , although these cells can express the BART miRNAs . This paradox could be resolved if the expression levels of the BART miRNAs differed in replicating and non-replicating cells . To address this question , we constructed a virus that encodes a truncated version of the CD2 molecule , whose expression is driven by the viral EA-D promoter . Thus , infected B cells undergoing lytic replication express CD2 at their cell surface and can be immunocaptured by a specific antibody . We quantified BZLF1 expression in the CD2-positive and CD2-negative populations using quantitative RT-PCR ( Fig 4A ) . As expected , we found that only CD2-positive cells produced BZLF1 at the RNA and protein level . This implies that cells that expressed the BZLF1 mRNA also expressed the BZLF1 protein , ie these mRNAs are not subjected to massive miRNA interference . We then quantified the expression profile of some viral and cellular miRNAs in these 2 cell populations . We found that replicating cells expressed approximately 2 times less BART miRNAs and 3 times less BHRF1 miRNAs than non-replicating cells ( Fig 4B ) . Thus , there is an inverse relationship between EBV miRNA and BZLF1 expression . However , the cellular miRNAs were expressed at the same level irrespective of the replication status of the infected cells , suggesting that these cells did not globally downregulate miRNA synthesis ( Fig 4C ) . Altogether , the previous results demonstrated that the BART miRNA repress spontaneous expression of BZLF1 in vitro and that their deletion enhances full productive lytic replication with virion production . However , the BART cluster is very large and we wished to learn the respective contribution of its subclusters . Therefore , we quantified BZLF1 expression in LCLs transformed with M81/ΔC1 , M81/ΔC2 and M81/Δb2 from 2 independent blood samples at two different time points ( Fig 5A , 5B and S5 Fig ) . These experiments showed that BZLF1 expression is higher in LCLs obtained by infection with M81/ΔC1 than in controls . Although we found no evidence for increased BZLF1 expression in M81/ΔC2 , both M81/ΔC1 and M81/ΔC2 LCLs expressed BZLF1 less strongly than M81/ΔAll and M81/ΔC1C2 , and this effect remained visible after 101 days of culture . LCLs infected with M81/Δb2 were indistinguishable from the wild type controls in terms of BZLF1 expression , although this miRNA has been suggested to control the onset of lytic replication [26] . However , we measured the expression of BALF5 in LCLs infected with M81/ΔAll or M81/Δb2 or wild type controls and found that the expression of BALF5 is indeed increased in LCLs generated with the mutant , particularly in those that carry M81/ΔAll ( Fig 5C and 5D ) . However , the increased expression of BALF5 in the LCLs infected with the M81/Δb2 virus does not result from an increased replication in these cells as shown by the unchanged BZLF1 expression in LCLs generated by a virus that lacks miR-BART-2 ( Fig 5A and 5B ) . We wished to complete the phenotypic analysis of the M81/ΔAll mutant by infecting humanized mice . We infected 7 mice intraperitoneally with the mutant and 5 with its wild type counterparts and measured blood viral titers 5 weeks post-infection ( Fig 6A ) . The titers were much higher in 4 out of 7 mice infected with M81/ΔAll than in the positive controls at early time point ( Fig 6A ) . Furthermore , animals infected with the mutant showed signs of wasting ( loss of weight , apathy , food refusal , ruffled hair ) and four of them had to be euthanized approximately at week 6 , 2 weeks before the planned termination of the experiment . This phenomenon was not seen in mice infected with wild type virus and is statistically significant ( 0/5 in M81-infected versus 4/7 in M81/ΔAll-infected mice; p = 0 . 019 by one-tailed Chi-square test ) . To allow comparison with wild type infected mice , we euthanized one of these mice at the same time . Whilst gross examination of the organs showed large neoplastic nodules in the spleen of infected with the M81/ΔAll mutant , there were only a few interspersed EBER-positive cells in the spleen of the animal infected with wild type virus . The remaining mice survived until week 8 without signs of animal suffering at which time they were euthanized . Both mice infected with wild type M81 and those infected with the ΔAll mutant showed tumors in the spleen or tumors in the pancreas for 3 animals ( Fig 6B and Table 1 ) . Although the mice were not all investigated at the same time , we found that 3 out of 5 mice infected with wild type M81 and 7 out of 7 mice infected with M81/ΔAll had macroscopic tumors and the difference between the 2 groups of animals is statistically significant ( Fig 6H and Table 1 ) . Animals with tumors in the pancreas tended to have lower virus titers than those with tumors in the spleen , possibly because the tumor cells in this case have more restricted access to the blood circulation . Histological examination of the above-described neoplastic infiltrates readily confirmed the presence of activated lymphoid cells that proved to be EBER-positive ( Fig 6B ) . We also found histological evidence of EBV-positive diffuse B cell infiltration of variable intensity after infection with M81/ΔAll and wild type EBV in other organs such as the liver , or the pancreas ( Fig 6B and Table 1 ) . The density of EBER-positive cells in these organs was similar in mice infected with wild type or with the mutant , although there was great variation in the concentration of EBV-positive cells in the lymphoid tissues between animals infected with the same virus strain ( Fig 6B–6E ) . We then stained histological sections of the tissues infiltrated with tumor cells for BZLF1 and gp350 . All infected tissues contained cells expressing both the early and the late marker of lytic replication but the ratio between EBER and BZLF1 or EBER and gp350 proved to be globally higher in the mice infected with the virus devoid of the BART miRNAs . However , the differences in BZLF1 expression between the two viruses , that proved to be statistically significant , were more pronounced than for gp350 that indeed failed to reach statistical significance ( Fig 6D and 6E ) . One possible explanation for this result is that dead gp350-positive cells are rapidly eliminated in vivo but not in vitro . We also attempted to generate LCLs with the serum from euthanized animals . However , we could generate only one LCL with the serum from a mouse infected with WT virus that did not display the highest viral titers . We conclude that the viral DNA measured in the serum does not derive from infectious virions , but rather probably from decayed infected cells . This observation is concordant with the fact that free virions are captured by B cells [21] . We also stained the tissues for LMP1 and EBNA2 expression and found that , although immunohistochemistry is not an entirely reliable quantitative assay , mice infected with ΔAll express LMP1 at much higher levels than mice infected with wild type viruses ( Fig 6B ) . However , the percentage of infected cells that expressed LMP1 or EBNA2 among the EBV-infected population showed no difference between wild type- and ΔAll-infected mice ( Fig 6F and 6G ) . We then attempted to confirm these observations in another set of non-humanized immuno-compromised mice . The rationale behind this experiment was to evaluate cell growth in the absence of a functional immune system that might reduce cell growth , particularly in mice infected with the ΔAll mutant that produces more lytic antigens . To this end , we injected i . p . two sets of independent peripheral B cells exposed to wild type or ΔAll viruses . In that case , we terminated the experiment at week 5 to exclude tumor overgrowth and allow more direct comparison between tumor burdens . We found that 7 out of 7 animals infected with ΔAll developed macroscopically visible tumors , mainly in the pancreas and to a lesser extent in the kidney and liver , compared to 3 out of 7 in animals infected with wild type virus ( S6A Fig and S1 Table ) . Importantly , the tumor burden was much higher in mice infected with the ΔAll knockout and these differences were statistically significant ( S6B Fig ) . Whilst the tumor burden concentrated in the spleen in humanized mice , this organ was spared in the non-humanized counterparts , as was expected in the absence of human hematopoietic transplantation . There was no difference in the incidence of tumors between the mice groups infected with the 2 different B cell samples . As previously observed , the tumor cells expressed EBER , EBNA2 , LMP1 , BZLF1 and gp350 ( S6C–S6H Fig ) . Here again there was no statistical difference in the density of EBER-positive cells within the tumors . In the same vein , the percentage of EBNA2-positive cells was similar in either type of virus infection and lytic replication was clearly stronger in mice infected with M81/ΔAll . There was , however , a difference in the proportion of EBER-positive cells expressing LMP1 , the latter being higher in mice infected with M81/ΔAll , suggesting that these cells were selected against in immunocompetent mice . As was the case in humanized mice , the proportion of strongly LMP1-positive cells was much higher in mice infected with M81/ΔAll ( S6C Fig ) . In summary , animals infected with a virus that lacks the BART miRNAs showed increased spontaneous lytic replication and frequently accelerated tumor formation in vivo , accompanied by increased LMP1 production . The accelerated tumor growth in animals infected with the ΔAll mutant could be explained by a shorter doubling time of B cells infected with this virus . Therefore , we assessed the transforming ability of M81/ΔAll at low cell density and low MOI . We infected primary B cells from 5 independent peripheral blood donors and did not find any difference in transforming ability between wild type and mutant viruses ( S7 Fig ) . We also used western blot to quantify expression of latent genes implicated in B cell growth . We found no difference in EBNA2 , EBNA3A , or LMP2 expression . However , EBNA3B and EBNA3C were expressed at mildly higher levels in 3 out of 4 LCLs infected with the virus devoid of the BART miRNAs ( Fig 7A ) . Western blots performed with a LMP1-specific antibody showed a clearly increased expression of this protein in 3 out of 4 cell samples infected with ΔAll mutant , relative to wild type controls . We also assessed LMP1 expression in 3 additional LCL triplets that were generated with wild type M81 , the ΔAll mutant and the ΔAll revertant and found that LMP1 expression was identical in both wild type and revertant , but strongly increased in the LCL generated with the ΔAll mutant in 2 out of 3 cases , the remaining case showing a minor LMP1 increase in the ΔAll mutant ( S8A Fig ) . It is interesting to note that LMP1 expression varied in LCLs infected with wild type M81 varied markedly between blood samples , with a minority of LCLs showing a much stronger expression . In these latter cases only , the absence of BART did not or only hardly increased LMP1 expression that was already high in the wild type LCL , suggesting that cellular polymorphisms modulate LMP1 expression . Highly variable LMP1 transcription rates in LCLs infected with the same virus were previously reported [27] . We also addressed the relationship between LMP1 expression and lytic replication by infecting 2 additional blood samples with wild type M81 , the ΔAll mutant as well as with a mutant that lacks BZLF1 and BRLF1 and is replication-deficient ( M81/ΔZR ) . We confirmed that BZLF1 expression was higher in the LCLs generated with the ΔAll mutant than in the wild type control . As expected the LCLs generated with M81/ΔZR did not replicate at all . In both samples LMP1 expression was identical in the LCLs generated either with wild type virus or with M81/ΔZR and was weaker than in the LCLs infected with ΔAll ( S8B Fig ) . Altogether , we found that LMP1 expression was stronger after infection with ΔAll in 7 out of 9 independent LCLs . The BART miRNAs were previously found to increase resistance to apoptosis though inhibition of caspase 3 expression [9] . Therefore , we gauged caspase 3 expression at the protein level using a western blot analysis in infected cells from 4 different donors . This assay showed a 2 to 5-fold increase in caspase 3 expression in all cases . We also assessed the ability of LCLs to withstand an apoptotic stress by incubating the cell lines with a panel of apoptosis-inducing drugs including Ionomycin , Staurosporine , Simvastatin and Etoposide [28] . The levels of apoptosis were assessed by TUNEL or caspase 3 cleavage assays . We could not identify significant differences between LCLs infected with the mutants or with the controls after treatment with Etoposide or Simvastatin ( Fig 7B and 7C ) . However , whilst treatment with Ionomycin gave rise to more apoptosis in LCLs infected with wild type viruses in TUNEL assays , staining with caspase 3 revealed increased apoptosis in wild type LCLs treated with Staurosporine .
The study of spontaneous EBV lytic replication has been hampered by the propensity of the virus to enter latency in infected cells . LCLs initiate some degree of lytic replication after treatment with chemicals such as TPA or butyrate [29] . Some non-lymphoid cell lines such as 293 cells can support lytic replication after transfection with BZLF1 [30] . Infection of primary epithelial cells gives rise to spontaneous lytic replication but the efficiency of infection remains low and these cells are difficult to grow in large numbers [31 , 32] . Thus , tractable experimental systems have not been available for a long time . However , M81 , a virus isolated from an NPC patient replicates strongly in primary B cells isolated from any individual tested so far [21] . Furthermore , M81 is amenable to a genetic analysis after its cloning as a bacterial artificial chromosome [21] . We addressed the function of the BART miRNAs by constructing viruses that evince partial or complete deletions of this locus , as well as a revertant thereof . We found that the BART miRNAs negatively regulate spontaneous lytic replication in B cells , as their excision from the M81 viral genome gives rise to an increase in spontaneous lytic replication in vitro and in vivo in humanized mice . This phenotypic trait disappears in the revertant virus or upon complementation . The BART miRNAs seem to target BZLF1 directly as its mRNA is recruited more efficiently to the RISC in cells infected by wild type virus than in LCLs generated with the BART miRNA knockout virus and expression of a luciferase gene fused to BZLF1 3’UTR is lower in LCLs generated with wild type virus relative to LCLs generated with the ΔAll virus . However , the difficulties to transfect primary LCLs with high efficiency somehow qualifies the latter result . Typically , miRNAs and their cognate targets are expressed in the same cells and the miRNA down-regulate protein expression in all cells that express them . In LCLs , the BART miRNAs are expressed in latently infected cells that do not express the BZLF1 . Therefore , the BART miRNAs can only exert their function on BZLF1 in the minority of cells that initiate lytic replication in a given LCL . Thus , they do not directly control lytic replication but come into play only in cells that have already initiated BZLF1 synthesis . Such a scenario fits with the observation that the number of spontaneously replicating cells in LCLs infected with M81/ΔAll does not exceed 15% . The remaining 85% are devoid of BART miRNAs but nevertheless remain BZLF1-negative . However , in cells that have already initiated lytic replication through expression of BZLF1 , the expression of the BART miRNAs apparently needs to be lower than in non-replicating cells . This , combined with the observation that individual replicating cells in LCLs infected with wild type or with M81/ΔAll express the BZLF1 protein at the same level suggests that the halved level of BART miRNAs present in replicating cells infected with wild type viruses is too low to efficiently down-regulate BZLF1 protein . This would mean that the expression of the BART miRNAs needs be lower than in latent cells , but does not need necessarily need to be completely extinguished . It remains unclear at this point whether the BART miRNAs are actively downregulated in replicating cells through an unknown active molecular mechanism , or whether the expression of the BART miRNA is stochastically distributed within the different cells of a LCL , with replication taking place in cells that happen to express low levels of these non-coding small RNAs . We previously observed that B cells infected with M81 sustain lytic replication over long periods of time , exceeding 3 months of continuous cell culture growth [21] . However , even in these cells , lytic replication eventually stopped . It is interesting to note that the BART miRNAs , probably independently of their effects on BZLF1 , also interfere with the mechanisms that control the long-term ability of a lymphoid cell to support lytic replication , as LCLs generated with M81/ΔAll keep producing virus when LCLs established from the same patient with wild type EBV do not anymore . However , replication in LCLs generated with M81/ΔAll also decreases with time , suggesting that the BART miRNAs accelerate this long-term mode of control of lytic replication but are independent of it . Thus , the BART miRNAs could interfere with lytic replication in several ways , negatively modulating lytic replication at its onset through its effect on BZLF1 , but also influencing the long-term ability of the LCL to support lytic replication by interfering with the selection process that favors latency . We then turned our attention to the miRNAs within the cluster that affect BZLF1 expression . Because the cluster encodes 22 miRNAs , we addressed the role played by the subcluster 1 , subcluster 2 and miR-BART2 in this process . We found that viruses devoid of the subcluster 2 or of miR-BART2 do not differ from wild type viruses in their ability to control the expression of BZLF1 . However , a virus that lacks both subcluster 1 and subcluster 2 expressed more BZLF1 than those infected with a subcluster 1 deletion mutant . This demonstrates that subcluster 1 plays a predominant role in the control of lytic replication but also that subcluster 2 contributes to this process . We did not identify any difference between the ΔC1C2 and the ΔAll viruses in terms of BZLF1 protein expression . Thus , miR-BART-2 does not seem to be implicated in the onset of lytic replication although it clearly modulates BALF5 expression as previously shown and confirmed in the presented study [26] . Two BART miRNAs that belong to subcluster 2 have been proposed to control lytic replication through modulation of BZLF1 and BRLF1 expression . Jung et al . demonstrated that miR-BART20-5p suppressed lytic replication through direct targeting of BZLF1 and BRLF1 mRNAs in a variety of EBV-positive epithelial cell lines [18] . MiR-BART20-5p is not or barely expressed in EBV-transformed LCLs and is thus unlikely to play a substantial role in spontaneous lytic replication in B cells [10] . MiR-BART18-5p was found to repress lytic replication in anti-Ig-treated Akata Burkitt’s lymphoma cell lines and in LCLs induced by TPA through its ability to target MAP3K2 [20] . These models display obvious differences with the spontaneous replication of LCLs that could explain the relatively minor role that we ascribed to the subcluster 2 . Therefore , it is possible that miR-BART18-5p plays an essential role in induced but not in spontaneous lytic replication . Our data point to a control of BZLF1 expression shared by multiple BART miRNAs whose individual contribution might be very limited and undetectable in viruses lacking a single miRNA . In such a case , only deletion of a subset of BART miRNAs could reveal their effect on BZLF1 expression and lytic replication . We also evaluated the role played by the BART miRNAs in the control of EBV-mediated transformation . We found that the deletion of the BART cluster does not influence the transformation abilities of the virus in vitro . Its impact in humanized mice was more complex to assess . Mice infected with ΔAll or with wild type viruses developed similar tumor burdens . However , this tumor load was already present after 5 weeks of infection in 4 out of 7 mice infected with ΔAll , whilst it took two additional weeks in the remaining animals . Importantly , mice infected with the BART miRNA knockout experienced a higher level of lytic replication that should theoretically allow infection of a larger number of B cells , thereby increasing the ensuing tumor mass . However , higher lytic replication might also boost the immune response against replicating cells . Therefore , we turned to non-humanized immunocompromized mice that cannot mount an immune response and do not have a large reservoir of EBV-negative resting B cells that can be infected by the viruses produced by a replicating B cell . We find that injection of freshly isolated peripheral blood B cells exposed to viruses and injected intraperitoneally gives rise to lymphoid tumors . This experiment allowed direct comparison of the transforming capacities of the virus and of the mutant . It showed that the tumor incidence is more than twice as high in animals treated with B cells infected by ΔAll and that the tumor burden was on average 5 times higher relative to B cells exposed to wild type virus . We conclude that the absence of BART miRNAs efficiently supported the growth of EBV-transformed cells . This apparently contradicts numerous studies that implicated the BART miRNAs in EBV-induced epithelial carcinogenesis . Here again , the much higher expression levels in EBV-associated carcinomas needs to be considered . Along the same line , a recently established xenograft model in NSG mice that clearly implicates the BART miRNAs in tumorigenesis uses cells that express the BART miRNAs at even higher levels than in NPC [33] . The expression levels of some EBV latent proteins were also moderately increased in some LCLs transformed by the M81/ΔAll mutant . Although this might have contributed to the increased transforming abilities of the mutant , this increase was inconstant and lower in intensity than observed for BZLF1 or LMP1 . This suggests that the effect on these latent proteins was indirect . LMP2A was suggested to be a target of the BART miRNAs but we could not confirm this observation in our experimental system [34] . We also evaluated the impact of the BART miRNAs on apoptosis or more generally cell death in infected B cells . We could not observe any increase in apoptosis in LCLs generated with the BART miRNA-negative virus , but to the contrary a moderate protective effect after treatment with some pro-apoptotic drugs such as Staurosporine . Ionomycin induced a higher level of cell death as assessed by an increased number of cells in TUNEL assay but did not modify the level of cleaved caspase 3 . Therefore , it is unlikely to reflect an increased level of apoptosis . Reciprocally , Staurosporine increased the percentage of cleaved caspase 3 but not the percentage of positive cells in TUNEL assays , suggesting that the cells entered the apoptosis pathway but could not complete it . We then evaluated the influence of BART miRNAs on some proteins involved in the regulation of apoptosis in EBV-infected cells . We found that the caspase 3 protein is increased after excision of the BART miRNA cluster in LCLs infected with M81/ΔAll . Caspase 3 has been found as a direct target of the BART miRNAs in Burkitt’s lymphoma cells but a more recent study performed on the NPC C666 cell line could not confirm these results [9 , 17] . This raises the intriguing possibility of a different mode of interaction between the caspase 3 mRNA and the BART miRNAs in different cell lineages . Our own results cannot distinguish between a direct and an indirect effect of the BART miRNAs on caspase 3 expression . Importantly , the LCLs that showed increased caspase 3 protein production also expressed the LMP1 protein at higher levels than the controls . This confirms previous studies that identify this viral oncoprotein as a direct target of the BART miRNAs [25 , 35 , 36] . Importantly , although immunohistochemistry is not an accurate quantitative method , we found that LMP1 is expressed at clearly higher levels in mice infected with ΔAll and this event is likely to have boosted cell growth . LMP1 has been found to facilitate extrinsic apoptosis through its ability to increase the expression of CD95 [37] . Furthermore , the LMP1 transmembrane domain activates apoptosis through activation of the unfolded protein response [38] . However , LMP1 can also protect against apoptosis through induction of BCL2A1 , a member of the BCL2 family of proteins [38] , or though inactivation of the p53 protein [39] . As EBV-infected cells express LMP1 , the anti-apoptotic effect of this viral protein seems to predominate in infected cells [38] . In LCLs that carry the M81/ΔAll mutant , the increase in LMP1 might predominate over the induction of pro-apoptotic proteins caused by the absence of BART miRNAs . Interestingly , the BART miRNAs were found to have anti-apoptotic properties in Burkitt’s lymphoma cells in which apoptosis was induced by the loss of the EBV genome with the help of a dominant negative version of EBNA1 . In that case , LMP1 was not present in cells transfected with the BART miRNAs . It is interesting to note that LMP1 is rarely expressed in EBV-associated carcinomas and this might reflect the repressive effects of the BART miRNAs that are produced at much higher levels in this context [2 , 5] . The BART cluster offers a good example of how the expression level influences the functions of miRNAs . They also show that miRNAs can have a marked influence on cellular functions even if expressed at seemingly low levels . However , it is important to note that the BART miRNAs are expressed on average at slightly higher levels than the BHRF1 miRNAs in LCLs or even than some crucial cellular miRNAs such as those belonging to the let7 family in hematopoietic stem cells in which miRNAs play a crucial role in cell differentiation [10 , 40] . In conclusion , we used recombinant viruses to reveal functions of the BART miRNA locus that result from multiple , sometimes even contradictory alterations of viral and cellular functions in cells infected in vitro and in vivo . This obviously reflects their high number , but also the fact that they collaborate to downregulate targets such as BZLF1 as shown in the present paper , or NDGR1 as previously reported [41] . Virus knockouts that lack single BART miRNA or a subset of them will provide useful tools to dissect their multiple and intricate molecular functions .
All human primary B cells used in this study were isolated from anonymous buffy-coats purchased from the Blood Bank of the University of Heidelberg . No ethical approval is required . All animal experiments were performed in strict accordance with German animal protection law ( TierSchG ) and were approved by the federal veterinary office at the Regierungspräsidium Karlsruhe , Germany ( Approval number G156-12 ) . The mice were housed in the class II containment laboratories of the German Cancer Research and handled in accordance with good animal practice with the aim of minimizing animal suffering and reducing mice usage as defined by Federation of European Laboratory Animal Science Associations ( FELASA ) and the Society for Laboratory Animal Science ( GV-SOLAS ) . HEK293 cell line is a neuro-endocrine cell line obtained by transformation of embryonic epithelial kidney cells with adenovirus ( ATCC: CRL-1573 ) [42 , 43] . DG-75 is an EBV-negative cell line that was established from a pleural effusion of a patient with a primary abdominal lymphoma that resembled Burkitt’s lymphoma ( ATCC: CRL-2625 ) [44] . Peripheral blood mononuclear cells from buffy coats purchased from the blood bank in Heidelberg were purified on a Ficoll cushion and CD19-positive primary B-lymphocytes were isolated using M-450 CD19 ( Pan B ) Dynabeads ( Dynal ) and were detached using Detachabead ( Dynal ) . WI38 are primary human lung embryonic fibroblasts ( ATCC: CCL-75 ) . All cells were routinely cultured in RPMI-1640 medium ( Invitrogen ) supplemented with 10% fetal bovine serum ( FBS ) ( Biochrom ) , and primary B cells were supplemented with 20% FBS until establishment of LCLs . All synthesized oligonucleotides used for cloning or PCR are listed in S2 Table . The wild type EBV strain M81 is available as a recombinant BACMID [21] . The viral genome was cloned onto a prokaryotic F-plasmid that carries the chloramphenicol ( Cam ) resistance gene , the gene for green fluorescent protein ( GFP ) , and the Hygromycin resistance gene ( B240 ) . All PCR primers used for PCR cloning or chromosomal building are listed in S2 Table and are based on the M81 EBV sequence ( GenBank accession number KF373730 . 1 ) . Deletion of the miR-BART subcluster 1 ( deletion from nt 139133 to nt 140132 ) generated ΔC1; deletion of the miR-BART subcluster 2 ( deletion from nt 145492 to nt 148777 ) gave rise to ΔC2 . These mutations were achieved by homologous recombination of the recombinant virus with a linear DNA fragment that encodes the kanamycin resistance gene , flanked by Flp recombination sites , and short DNA regions homologous to the regions immediately outside of the deletion to be obtained , as described [45] . The double knockout BART miRNA subcluster 1 plus subcluster 2 ( ΔC1C2 ) was obtained by excising the kanamycin cassette present in ΔC1 with FLP recombinase , followed by a deletion of BART miRNA subcluster 2 via linear targeting with the PCR product that yielded ΔC2 . The ΔAll mutant that lacks all BART miRNAs was obtained by exchanging the miR-BART-2’s seed region with an unrelated sequence in the ΔC1C2 recombinant virus using chromosomal building [45] . This mutagenesis generated an additional AclI restriction site and digestion with this enzyme allows distinction between the mutant and the wild-type sequences ( S1 Fig ) . We applied the same strategy to the wild type M81 BAC to obtain a recombinant EBV that lacks mir-BART-2 only ( Δb2 ) . We constructed a revertant of ΔAll using chromosomal building on the basis of the original parental M81/ΔAll Bacmid before passaging in 293 . Here the complete BART miRNA locus , from the miR-BART subcluster 1 to miR-BART2 , was cloned from the M81 BAC and reintroduced into the M81/ΔAll BAC genome . We introduced the rat CD2 gene under the control of an EA-D promoter into the BXLF1 gene of the M81 genome ( nt 131044 to nt 133362 ) by homologous recombination using a linear vector that included the kanamycin resistance cassette as a selection marker . The disruption of BXLF1 gene does not interfere with the growth of LCLs [46 , 47] . Upon induction of the lytic replication , CD2 is expressed at the surface of replicating cells . CD2-positive cells can be pulled down with a specific monoclonal antibody ( OX34 ) coupled with anti-mouse IgG Dynabeads and submitted to protein or RNA extraction . Recombinant EBV plasmids were lipotransfected into HEK293 cells using Metafectene ( Metafectene , Biontex ) and the selection of stable 293 cell clones carrying the recombinant EBV plasmid was achieved by adding hygromycin to the culture medium ( 100 μg/ml ) as previously described [48] . To assess the genome integrity of recombinant EBV within the stable clones , the circular EBV genomes present in these cells were extracted using a denaturation-renaturation method [49] and transferred into the E . Coli strain DH10B by electroporation ( 1000V , 25μF , 200 Ohms ) . The transformed E . Coli clones were further assessed by restriction enzyme analysis of plasmid minipreparations . 293 cells stably transfected with recombinant EBV-BACs were transfected with expression plasmids encoding BZLF1 ( p509 ) and BALF4 ( pRA ) using the liposome-based transfectant Metafectene ( Biontex ) . Three days after transfection , virus supernatants were collected and filtered through a 0 . 4 μm filter . B1124 is a plasmid that contains a bi-directional tetracycline-inducible CMV promoter that encodes a truncated nerve growth factor receptor ( NGFR ) on one site , and the BART miRNA subcluster 1 and 2 without the PstI repeats and the LF3 gene on the other . B1034 contains only NGFR and was used as a negative control . M81/ΔAll LCLs were electroporated with either B1034 or B1124 and cultured with 1 μg/ml doxycycline for 30 days . NGFR-positive cells were isolated with specific antibodies and used for protein and RNA analyses . To evaluate EBV genome equivalents per milliliter of supernatant , viral supernatants were treated with DNase I . Following a subsequent treatment with proteinase K , we used quantitative real-time PCR analysis ( qPCR ) with primers and probe specific for the non-repetitive EBV BALF5 gene sequence to measure the EBV copy numbers in the supernatants [50] . The quantification of EBV DNA genome copies per milliliter in the blood of infected mice was performed from genomic DNA extracted from total blood by using regular phenol extraction . RNA extracted with Trizol from LCLs was reverse transcribed with AMV-reverse transcriptase ( Roche ) using a mix of random primers The primers and probes used to detect BZLF1 are listed in the S1 table . The PCR and data analysis was carried out using the universal thermal cycling protocol on an ABI STEP ONE PLUS Sequence Detection System ( Applied Biosystems ) . All samples were run in duplicates , together with primers specific to the human GAPDH gene to normalize for variations in cDNA recovery . BART miRNAs extracted from cells with Trizol were reverse transcribed using specific stem-loop primers and TaqMan miRNA reverse Transcription Kit ( Applied Biosystems ) , as described elsewhere [8] . The sequences of the stem-loop primers , primers and probes are listed in S2 Table . Reverse transcription and amplification of the cellular snoRNA RNU48 was performed in parallel to normalize for cDNA recovery ( Assay ID 001006; Applied Biosystems ) . Real-time PCR was performed on an ABI STEP ONE PLUS Sequence Detection System ( Applied Biosystems ) . B cells purified from peripheral blood were exposed to viral supernatant for two hours , then washed once with PBS and cultured with RPMI supplemented with 20% FBS in the absence of immunosuppressive drugs . For transformation assays , the percentages of EBNA2 positive cells in the infected samples were evaluated by immunostaining with a specific antibody at 3 days post-infection ( dpi ) . Cell populations containing 3 or 30 EBNA2-positive cells per well were seeded into 48 wells of U-bottomed 96-well plates that contained 103 gamma-irradiated WI38 feeder cells . Non-infected B cells were used as a negative control . Outgrowth of lymphoblastoid cell clones ( LCLs ) was monitored at 33 dpi . We also incubated 105 primary B cells with 25ml of cell-free LCL culture supernatants for 2 hours . These cells were plated on 96 well cluster plates at a concentration of 103 cells per well , together with the same number of gamma-irradiated feeder cells . We stained infected cells with mouse monoclonal antibodies against BZLF1 ( Clone BZ . 1 ) , gp350/220 ( Clone 72A1 ) , EBNA2 ( Clone PE2 ) and a Cy-3-conjugated goat-anti-mouse secondary antibody ( Dianova , Invitrogen ) . We performed western blots with mouse monoclonal antibodies against BZLF1 ( Clone BZ . 1 ) , gp350 ( Clone OT6 ) , DICER ( clone F10 , Santa Cruz Biotechnology ) , LMP1 ( clone CS1-4 ) , EBNA2 ( clone PE2 ) , and Actin ( clone ACTN05 , C4 , Dianova ) . We also used rabbit polyclonal antibodies against caspase-3 ( Cell Signalling Technology ) and rat monoclonal antibodies ( kindly provided by Dr . E . Kremmer Helmholtz Zentrum Munich and Dr . F . Grässer , University of Homburg ) against Ago2 ( clone 11A9; ) , EBNA3A ( clone E3AN-4A5 ) , EBNA3B ( clone 6C9 ) , EBNA3C ( clone A10 ) , LMP2A ( clone 4E11 ) , BALF5 ( clone 4C12 ) . Mouse monoclonal antibodies specific to LMP1 ( clone S12 , BD Pharmingen ) , BZLF1 ( Clone BZ . 1 ) and gp350 ( Clone OT6 ) were used for immunohistochemical staining against EBV proteins in infected murine tissues . Cells were fixed with 4% paraformaldehyde in PBS for 20 min at room temperature and permeabilized in PBS 0 . 5% Triton X-100 for 2 min except for samples stained for viral glycoproteins ( gp350 ) . Cells were incubated with the first antibody for 30 min , washed in PBS three times , and incubated with a secondary antibody conjugated to Cy-3 for 30 min before embedding in 90% glycerol . Proteins were extracted with a standard lysis buffer ( 150 mM NaCl , 0 . 5% NP-40 , 1% Sodium deoxycholat , 0 . 1% SDS , 5 mM EDTA , 20 mM Tris-HCl pH7 . 5 , proteinase inhibitor cocktail ( Roche ) ) for 15 min on ice followed by sonication to shear the genomic DNA . Up to 20μg of proteins denatured in Laemmli buffer for 5 min at 95 degree were separated on SDS-polyacrylamide gels and electroblotted onto a nitrocellulose membrane ( Hybond C , Amersham ) . Proteins extracted to assess gp350 expression were prepared in Laemmli buffer without 2-mercaptoethanol . After pre-incubation of the blot in 3% milk dry powder in PBST ( PBS with 0 . 2% Triton-X100 ) , the antibody against the target protein was added and incubated at room temperature for 1 hr . After extensive washings in PBST , the blot was incubated for 1 hr with suitable secondary antibodies coupled to horseradish peroxidase ( goat anti-mouse ( Promega ) , goat anti-rabbit ( Life technologies ) , or rabbit anti-goat ( Santa Cruz ) IgG ) . Bound antibodies were revealed using the ECL detection reagent ( Pierce ) . 6 × 108 cells were washed twice in ice-cold PBS and subsequently lysed in 5 ml lysis buffer containing 25 mM Tris HCl ( pH 7 . 5 ) , 150 mM KCl , 2 mM EDTA , 0 . 5% NP-40 , 0 . 5 mM DTT , 200 u/ml RNAse inhibitor and protease inhibitor cocktail ( Roche ) . Lysates were incubated for 30 min on ice and clarified by centrifugation at 20 , 000 g for 30 min at 4°C . To estimate the recovery of miRNAs after RISC-IP , total RNA was prepared from 10% of the cell lysates using the TRIzol RNA Isolation Reagents ( Life technologies ) following the manufacturer's instructions . 6 μg of purified Rat-monoclonal anti-hAgo2 antibody ( 11A9; Helmholtz Zentrum Munich ) or of monoclonal anti-BrdU-antibody ( Abcam ) was mixed with 20 μl of Dynabeads Protein G ( Dynabeads Protein G Immunoprecipitation Kit , Life technologies ) and subsequently incubated with 2 . 5 ml of cell lysates for 4–6 hours under constant rotation at 4°C . The beads were then washed four times with IP wash buffer ( 300 mM NaCl , 50 mM Tris HCl pH 7 . 5 , 5 mM MgCl2 , 0 . 1% NP-40 , 1 mM NaF ) and once with PBS to remove residual detergents . The beads were resuspended with 300μl of proteinase K buffer ( 100 mM Tris-HCl PH 7 , 4 / 50 mM NaCl /10 mM EDTA ) in the presence of proteinase K ( 0 . 33 mg/ml ) and incubated for exactly 30 min at 37°C with shaking at 600 rpm and immediately transferred onto ice . Total RNA present in complexes after RISC-IP was purified by using TRIzol RNA Isolation Reagents and dissolved in 50 μl RNase-free water . The 3’UTR regions of the M81 BZLF1 and BALF5 genes were PCR amplified with the pair of primers listed in S2 table . The PCR products were digested with EcoRI and XhoI and ligated into the firefly luciferase expressing vector , pGL4 . 5 ( Promega ) , which had firstly been modified to insert EcoRI and XhoI cutting sites behind the luciferase coding region . The luciferase reporter assays were performed by electroporation of 10 million LCL cells with 5μg of a pcDNA3 . 1-CD2 plasmid that encodes a truncated rat CD2 protein and 10μg of the different luciferase expression plasmids . 48 hours post electroporation , cells were washed twice with PBS and the luciferase activity was measured by the Beetle-Juice firefly luciferase assay system ( PJK ) . The transfected cells were also immunostained with an antibody specific to rat CD2 to evaluate the electroporation efficiency . Total RNA isolated from LCLs or equal volumes of RNA post RISC-IP from different samples were reverse transcribed with AMV-reverse transcriptase ( Roche ) using a mix of random hexamers . The mRNAs of interest were quantified with the Power SYBR green PCR Mix ( Life technologies ) using primer pairs specific to the gene of interest . Data analysis was carried out with the universal thermal cycling protocol of the ABI STEP ONE PLUS Sequence Detection System ( Applied Biosystems ) . We generated humanized mice by intrahepatical injection of human CD34-positive hematopoietic progenitor cells ( HPCs ) in irradiated ( 1 Gy ) newborn NSG-A2 mice ( NOD . Cg-PrkdcscidIl2rgtm1WjlTg ( HLA-A2 . 1 ) 1Enge/SzJ ) ( huNSG-A2 mice ) . CD34-positive HPCs were isolated from human fetal liver tissue ( Advanced Bioscience Resources , Alameda , CA , USA ) using a human CD34 purifying Microbead kit ( Miltenyi Biotec ) . We used 2 liver samples to generate the 12 humanized mice used in this study . The percentage of human CD45 positive cells was evaluated 12 weeks after HPC transplantation and only mice with more than 40% of human CD45-positive cells were infected . We performed a titration of viral stocks by infecting primary B cells with increasing dilutions . The infected cells were stained at 3 dpi to determine an EBNA2-positive B cells/ml virus titer . In all experiments , we injected intra-peritoneally in each mouse enough viruses to generate 5x106 EBNA2-positive cells as described [21] . Peripheral blood samples were drawn 5 weeks post-infection . Mice were euthanized at week 8 except if signs of animal suffering became apparent and we examined their blood and tissues for signs of viral infection . We isolated human CD19+ B cells from buffy coats and exposed to virus supernatants at a moi sufficient to generate 20% of EBNA2-positive cells for 2 hours at room temperature under constant agitation . The infected cells were collected by centrifugation and washed twice with PBS for two times . 2*10ˆ6 infected primary B cells , equivalent to 4*10ˆ5 infected cells were injected intraperitoneally into NSG mice . The mice were euthanized at 5 weeks post injection , autopsied and their organs were subjected to macroscopic and microscopic investigation . The organs from the studied mice were fixed in 10% formalin overnight and embedded in paraffin blocks . 3-μm-thin continuous sections were prepared and submitted to antigen retrieval at 98°C for 40 min in a 10 mM sodium citrate , 0 . 05% Tween20 pH 6 . 0 solution . Bound antibodies were visualized with the Envision+ Dual link system-HRP ( Dako ) . In parallel , adjacent sections were stained with hematoxylin and eosin ( H&E ) . The presence of EBV was detected by in situ hybridization with an EBER-specific PNA probe , in conjunction with a PNA detection kit ( Dako ) . Pictures were taken with a camera attached to a light microscope ( Axioplan , Zeiss ) . We induced apoptosis in LCLs ( 5*105 cell per well of a 48-well-plate ) transformed by ΔAll or wild type viruses at 40–60 dpi by adding Etoposide ( 4μg/ml , Sigma Aldrich ) or Staurosporine ( 4μg/ml , Sigma-Aldrich ) to the culture medium for 20 hrs . Cells were also treated with Ionomycin ( 4μg/ml , Sigma Aldrich ) for 48 hrs or Simvastatin ( 2μM , Calbiochem ) for 5 days . We included DMSO-treated cells and ethanol-treated cells as controls . Cells were then washed twice with ice-cold PBS , dried on glass slides and fixed with 4% paraformaldehyde in PBS to perform a TUNEL assay that labels apoptotic cells with DNA breaks ( Cell Death Detection Kit , TMR red , Roche ) following the instruction of manufacturer . Cells were also stained with a rabbit antibody specific for cleaved caspase 3 ( Cell signal technology ) . All results obtained in in vitro studies with LCLs generated by EBV wild type or mutants with B cells from the same blood donors were paired and analyzed by paired student t-test . Unpaired student t-test was applied for analyzing the grouped humanized or non-humanized NSG mice infected by either M81 or M81/ΔAll virus . All p-values were analyzed as 2-tailed and the values equal to 0 . 05 or less were considered significant unless indicated . We used a Chi square test to analyze the tumor incidence in humanized and non-humanized mice . The statistical analyses were performed with the GraphPad Prism 5 software . | The Epstein-Barr virus ( EBV ) infects more than 90% of the human adult population . Although EBV usually causes an asymptomatic infection , it is oncogenic in a small proportion of infected individuals . EBV produces a large number of microRNAs , a type of RNA that controls the production of their proteins though multiple mechanisms . We addressed the role played by the BART microRNAs , a subgroup of the EBV microRNAs , by generating a virus that lacks them and by comparing the characteristics of this modified virus with those of the unmodified virus . We found that the BART microRNAs cooperate to curb EBV multiplication , both in infected cells and in humanized mice . Furthermore , the BART miRNAs did not potentiate EBV’s ability to form tumors in different types of mice , some of which are unable to mount an immune reaction against the virus , as could have been expected from the literature . This can be explained at the molecular level by the ability of the BART microRNAs to downregulate the synthesis of multiple cellular and viral proteins , among which caspase 3 and LMP1 , two essential modulator of cell death and cell proliferation , are likely to play an important role in the outcome of the virus infection . Thus , the BART microRNAs negatively impact on two essential viral functions , probably to maintain a balance between the virus and its host . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The Epstein-Barr Virus BART miRNA Cluster of the M81 Strain Modulates Multiple Functions in Primary B Cells |
Hookworm-related cutaneous larva migrans ( HrCLM ) is a neglected parasitic skin disease , widespread in resource-poor communities in tropical and subtropical countries . Incidence and risk factors have never been investigated in a cohort study . To understand the seasonal epidemiology of HrCLM , an open cohort of 476 children in a resource-poor community in Manaus , Brazil was examined for HrCLM monthly over a period of 6 months . Monthly prevalence and intensity of infection were correlated with the amount of monthly precipitation . Multivariable Cox regression analysis indicated male sex ( hazard ratio [HR] 3 . 29; 95% confidence interval [CI] 1 . 95–5 . 56 ) , walking barefoot on sandy ground ( HR 2 . 30; 95% CI 1 . 03–5 . 16 ) , poverty ( HR 2 . 13; 95% CI 1 . 09–4 . 17 ) and age between 10 and 14 years ( HR 1 . 87; 95% CI 1 . 01–3 . 46 ) as predictors of HrCLM . Monthly incidence rates ranged between 0 . 21 and 1 . 05 cases per person-year with an overall incidence of 0 . 52 per person-year . HrCLM is a frequent parasitic skin disease in this resource-poor community . Every second child theoretically becomes infected during one year . Boys , 10 to 14 years old , belonging to the poorest households of the community , are the most vulnerable population group . Even in the tropical monsoonal climate of Amazonia there is a considerable seasonal variation with monthly incidence and number of lesions peaking in the rainy season .
Hookworm-related cutaneous larva migrans ( HrCLM ) is a neglected tropical skin disease caused by hookworm larvae of cats and dogs such as Ancylostoma braziliense , Ancylostoma caninum and Uncinaria stenocephala [1] . In humans , these larvae are unable to cross the basal membrane of the epidermis and hence cannot continue their normal development to adult worms . By consequence they haphazardly migrate in the epidermis , producing an elevated linear or serpiginous track . The intense itching leads to important pruritus-associated morbidity such as excoriations and bacterial superinfection of the lesions [2–5] . HrCLM is endemic in many tropical and subtropical countries worldwide [6–9] . Prevalence reached up to 8% in population-based studies with significant variation between sexes and age-groups [2] . Children are affected in particular [2 , 3] . In semi-arid climates such as in North-eastern Brazil , there is significant seasonal variation in prevalence from 0 . 2% in the middle of the dry season to 3 . 1% in the rainy season [4] . Known risk factors are male sex , young age , barefoot walking , poverty and presence of animal faeces on the compound [2 , 4 , 10] . In order to determine hazard ratios for previously identified risk factors and verify whether there is seasonal variation of incidence and morbidity in the tropical monsoonal climate of Amazonia , we conducted a longitudinal study with a cohort of 476 children living in a resource-poor community in the outskirts of Manaus , the largest city in Amazonia .
The study was conducted in the resource-poor neighbourhood Nova Vitoria in Manaus , capital of Amazonas State , Brazil . Manaus is situated at 03°06' south latitude and has a tropical monsoonal climate following the Köppen-Geiger classification with a dry season ( less than 60 mm precipitation/month ) usually in August and heavy monsoon rains during the rest of the year [11] . As many other resource-poor communities in Brazil , the study area was built up without permission of public authorities . This explains the lack of public infrastructure such as health facilities , childcare , paved streets or a sewage disposal system . In consequence many stray dogs and cats roamed through the streets and children were playing unattended on the sandy ground . The population was poor , one third of the households experienced food shortage in the past 12 months . Most of the families had several children . The study area and population have been described in detail previously [2] . At baseline all households in the study area were visited . Based on stringent inclusion criteria 92% of the households were admitted to the cross-sectional study . Methods of the cross-sectional study have been published previously [2] . In brief , all household members were examined clinically for HrCLM and environmental , socio-economic and behaviour-related risk factors were documented using pre-tested , structured questionnaires . All children of the 262 households were assessed for eligibility for the cohort and then monitored monthly for the presence of new HrCLM ( Fig 1 ) . The examination took place in a room where privacy was guaranteed . The whole body surface was examined , only the genital area was spared in case itching was absent . HrCLM was diagnosed clinically when the characteristic slow-progressing , elevated linear or serpiginous track was present [12–14] . Inclusion criteria for the cohort were: residence in the study area , absence of HrCLM-infection at baseline , age under 18 and provision of an informed written consent . Children who were found infected at baseline were included on the day they received treatment with ivermectin or topical thiabendazole . In all but one child , HrCLM lesions had resolved by the next follow-up . Due to an expected important drop out rate caused by out- and in-migration of whole families and temporary residence of family members in the countryside outside Manaus , we chose an open cohort design . Siblings of participants and children of newly encountered families replaced children who were lost for follow-up in order to keep the number of cohort members stable ( Fig 1 ) . Participants who entered the cohort after baseline were examined and interrogated in an identical manner . In case of temporary absence , only the time of presence in the study area between two consecutive follow-ups was analysed . Participants lost to follow-up before the second examination ( n = 29 ) were excluded from the analysis . To avoid selection bias 12 individuals , who were not identified by active case finding but presented to the team for procurement of treatment , were excluded . One patient was excluded because he refused to be treated with ivermectin ( Fig 1 ) . In order to gather information on reinfections , individuals censored after HrCLM diagnosis were observed further after treatment ( S1 Fig ) . These data were included in the analysis of monthly incidence and prevalence , and for analysis of clinical characteristics . Episodes of HrCLM at recruitment were documented and included in the analysis of clinical characteristics and monthly prevalence , if the prerequisite of residency for more than 2 months in the study area was fulfilled . For the calculation of monthly incidence rates only participants without HrCLM or participants that had been treated in the preceding month were included . The study started in April 2009 ( rainy season ) and ended in September 2009 ( which in 2009 was still part of the dry season ) . The study was approved by the Ethical Committee of the Fundação de Medicina Tropical–Amazonas ( FMT-AM ) . Informed written consent was obtained from each participant’s legal guardian . Every individual with HrCLM was offered free treatment independently whether he or she participated in the study . Treatment consisted of ivermectin ( Ivermec , Uci-farma , São Paulo , Brazil ) given as single oral dose ( 200 μg/kg ) . In the case of children <5 years or <15 kg , thiabendazole 5% ( Tiadol , Bunker Indústria Farmacêutica Ltda . , São Paulo , Brazil ) was topically applied 3 times a day for one week [2] . Data were entered in Microsoft Office Access 2007 , cleaned for entering errors and analysed with PASW Statistics Version 18 . 0 ( SPSS Inc . , Chicago , USA ) . Missing data were assumed to be missing at random and were marked in the analysis . A wealth score was formed out of household assets using principal component analysis as previously described [2 , 15] . Income and age were categorized similar to previous population-based studies to allow comparison [2 , 4 , 10 , 16] . Time to infection was defined as the time between the date of inclusion in the cohort and the date of diagnosis of HrCLM , date of last examination before lost to follow-up , or date of last examination at the end of the study ( right censoring ) , whichever occurred first . In case of temporary absence , only the time between consecutive follow-ups was included . Lost to follow-ups were compared to those who were censored by event or end of the study using chi-square-test or likelihood ratio where appropriate . For bivariable risk factor analysis , hazard ratios ( HR ) were calculated together with 95% confidence intervals ( 95% CI ) using a Cox proportional hazards model . For multivariable risk factor analysis , all variables that showed weak evidence of an association with HrCLM ( p<0 . 1 ) were entered into a stepwise Cox regression . Only significant variables ( p<0 . 05 ) remained in the model . We calculated standard errors and 95% CI to identify multicollinearity and removed variables where necessary . We used log-minus-log function to check if proportional hazard assumption was satisfied . Kaplan-Meier curves were used to visualize event-free periods and differences in incidence between sexes , age groups , frequency of barefoot walking and wealth strata . We calculated Population Attributable Fraction ( PAF ) for independent risk factors practically amenable to intervention . As only one risk factor was amenable to intervention , we used Levins unadjusted equation: {pe ( RR-1 ) }/ {pe ( RR-1 ) +1} with pe being the proportion of the population , which is exposed to the risk factor [17 , 18] . RR was calculated as a rate ratio . Incidence rate and prevalence were determined monthly . Seasonal changes in prevalence , clinical presentation ( superinfection , number of lesions , site of infection ) and total monthly precipitation in Manaus were correlated using Spearman´s rank correlation with a two-tailed significance level of p<0 . 05 . Climate data were obtained from the International Institute of Meteorology of Brazil ( INMET ) . The overall incidence rate was extrapolated from the number of patients divided by the accumulated time to HrCLM infection . Correspondingly , reinfection rate was calculated for the subgroup of patients who were infected at the first examination and were included after treatment . Monthly incidence rates were obtained by dividing the number of HrCLM-infections by the number of followed children during the corresponding time period multiplied by 12 assuming an exact 1-month-period between two follow-ups . Corresponding 95% CI were calculated by a Poisson regression .
During the 6 months of the study , a total of 476 children living in 209 households were included with a median time of follow-up of 149 days ( range 13–166 ) , which amounted to 54 , 938 person-days at risk . The median age at entry was 6 years ( range 0–15 ) and the majority of the children were girls ( 52% ) . After a median time of 65 . 5 days ( range 22–132 ) , 68 children ( 14 . 3% ) were lost to follow-up . Comparing their characteristics , there was no significant difference except for the proportion of barefoot walking on sandy ground ( Table 1 ) . The Kaplan-Meier estimated HrCLM-free proportion after 90 days and after 166 days ( at the end of the study ) was 84% and 82% , respectively ( Fig 2 ) . Mean time to HrCLM infection was 146 . 4 days ( 95% CI 142 . 3–150 . 5 ) . There was no difference in mean time to HrCLM infection between recruitment at baseline or later ( 145 . 6 days , 95% CI 141 . 1–150 . 0 days and 139 . 3 days , 95% CI 128 . 5–150 . 1 days , respectively , p = 0 . 462 ) . A total of 161 episodes of HrCLM were identified . Only 27 children ( 5 . 7% ) accounted for 66 ( 41 . 0% ) of all episodes by having several ( up to 4 ) episodes of HrCLM . During follow-up , 78 children developed HrCLM resulting in an overall incidence rate of 0 . 52 per person-year . Reinfection rate of children who were infected before inclusion in the cohort ( n = 64 ) was more than twice as high with 1 . 08 per person-year . Monthly incidence rates and prevalence decreased every month from the rainy to the dry season ( Table 2; Fig 3 ) . Clinical characteristics showed seasonal differences . During the rainy season in April , every second affected child had multiple lesions and lesions were found all over the body . During the dry season in contrast , only single lesions were encountered and lesions predominantly occurred on hands and feet . Superinfected lesions were only found from April to June ( Table 3 and Fig 3 ) . The decreases in prevalence and infection intensity ( multiple infections per person ) were correlated with the decreasing amount of monthly precipitation ( rho = 0 . 928 , p = 0 . 008 , and rho = 0 . 941 , p<0 . 001 respectively ) . Bivariable risk factor analysis revealed 10–14 years of age , a low wealth score , practicing soccer ( soccer is usually played barefoot ) , walking barefoot on sandy ground and HrCLM-infection at recruitment as predictors for HrCLM-infection during the follow-up period . The most potent predictor was male sex ( Table 4 ) . In the adjusted multivariable model age , sex , low wealth and barefoot walking remained as independent risk factors visualized in Kaplan-Meier curves ( Fig 4 ) . The PAF of walking barefoot on sandy ground was 45% .
So far , an incidence rate of HrCLM has never been determined in an appropriately designed study . Calculated monthly it showed a seasonal variation between a maximum of 1 . 05 per person-year during rainy season and a minimum of 0 . 21 per person-year during dry season . The overall incidence rate was 0 . 52 cases per person-year . That means that on the average every second child will get infected with HrCLM within one year , if the observed six months are representative for the whole year . The reinfection rate was even higher ( 1 . 08 per person-year ) . HrCLM-infections are not distributed homogenously throughout the child population; but obviously , children with a certain high-risk profile get infected several times a year . Considering the possible implications of such an infection including pruritus , sleep disturbance , impaired school performance and superinfection , this is a massive problem within the child population[5 , 21] . Most of the known risk factors identified in our previous cross-sectional study in Manaus and other studies in Northeast Brazil , could be confirmed by the present cohort study [2 , 4 , 10] . Male sex was the strongest predictor for HrCLM-infection with a more than three times higher risk in boys than in girls . Children between 10 and 14 years had a nearly two-fold higher risk to get infected than children of the youngest age group . Both might be due to gender-related behaviour or the extent of parental surveillance , as hypothesised earlier [4] . Generally , boys may spend more time outdoors , have more contact with the soil when playing and are less attended by their parents . These characteristics may be more present when children grow older , which might explain in part the age-related differences in hazard ratios . Walking barefoot especially on sandy ground may also be a risk factor in travellers [6] . In endemic areas , the frequency of using protective footwear clearly influences the risk of infection [2] . The estimated PAF related to barefoot walking on sandy ground was 45% , meaning that nearly half of the infections with HrCLM could have been prevented if all children wore shoes [17] . Even plastic sandals , which are the typical footwear in this area , can prevent infection [2] . HrCLM is a poverty-related disease . Poverty-related living conditions with stray dogs and cats , unpaved streets and many unattended children playing in the streets create a beneficial environment for the transmission of hookworm larvae . More than half of the study households had only one minimum wage or less to their disposition [2] . But even within this poor community , poverty was an independent predictor for the acquisition of HrCLM . This corroborates the results of our cross-sectional study as well as observations from other neglected tropical diseases [2 , 22 , 23] . The presence of animal faeces on the compound was the only independent risk factor identified in the previous cross-sectional study that could not be confirmed by this study . This might be due to the different composition of the study population with only children , who might spent more time outside the compound while playing with other children . Accordingly , the only deducible measure for disease control in the study area is prevention of barefoot walking . However , just providing shoes might not be sufficient similar to what we previously observed in a study on tungiasis control [24] . The underlying causes are much more complex . Even good parental knowledge about disease etiology couldn´t prevent HrCLM-infection [2 , 21] . Mothers feel unable to look out for their children [21] . Incidence was highest in boys aged 10–14 who probably are the less attended children in a poor household . Furthermore , the disease burden was highest among the poorest of the poor . In this way , HrCLM is a neglected disease acquired by the most neglected parts of the population . Only by changing social circumstances and health education of parents and children , preventive behaviour may be established . There may have been a selection bias in favour of younger children because of school attendance and work during visiting hours in the daytime . For security reasons visiting hours could not be extended beyond 6 p . m . Due to an exhaustive sampling strategy and high participation rate , we did obtain , however , a representative sample of the paediatric daytime population , which most likely is predominantly exposed in the study area . Although children carry the biggest part of disease burden , the results cannot be translated to the whole population as no adults participated . [2 , 4 , 10] . We observed a differential loss of participants ( Table 1 ) . This might have led to an overestimation of events in the group exposed to “walking barefoot on sandy ground” . Otherwise we have no reasons to believe that missing data biased the results . The absence of treatment in the study area may have led to more HrCLM-infections in April and biased therefore the changes in disease intensity ( multiple lesions per person ) and severity ( superinfection ) . However , the decrease in disease intensity continued over the whole study period and the correlation with monthly precipitation was significant . Incidence rate was calculated with person-time and referred to a whole year even though the probability of disease is not constant during the study period . The observation period , however , included three months of the dry and three months of the raining season and may therefore be representative for the whole year . In conclusion , the prevalence of HrCLM showed seasonal variation and was correlated with precipitation and disease intensity . Overall incidence rate among children was as high as 0 . 52 per person-year . Independent risk factors were male sex , age , walking barefoot on sandy ground and extreme poverty . | Hookworm-worm related cutaneous larva migrans ( HrCLM ) is a parasitic skin disease caused by hookworm larvae of cats and dogs occurring in many countries with a tropical or subtropical climate . Humans are a biological impasse for these helminths as the larvae cannot pass the basal membrane of the epidermis and hence migrate haphazardly in the skin causing local inflammation and intense itching . In scientific literature HrCLM is generally described as a disease of travellers returning from endemic areas . In contrast , epidemiological data is scanty . In a previous study , we had examined an entire resource-poor neighbourhood in Manaus ( Brazil ) and showed , that HrCLM is an important individual and public health problem , affecting up to 8% of the population , in particular children . In this study , we followed a cohort of children for six months . We found a significant seasonal variation in incidence and morbidity between dry and rainy season . Extrapolated , every second child in this population will be affected at least once within one year . The longitudinal study design enabled us to validate previously identified risk factors . Children aged 10–14 years , in particular boys , and those walking barefoot on sandy ground had the highest infection rates . Children from the poorest families in the resource-poor community were most vulnerable to HrCLM . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"recreation",
"children",
"medicine",
"and",
"health",
"sciences",
"salaries",
"sports",
"cutaneous",
"larva",
"migrans",
"geographical",
"locations",
"tropical",
"diseases",
"social",
"sciences",
"parasitic",
"diseases",
"age",
"groups",
"seasons",
"developmental",
"bi... | 2018 | Epidemiology and morbidity of hookworm-related cutaneous larva migrans (HrCLM): Results of a cohort study over a period of six months in a resource-poor community in Manaus, Brazil |
Despite intensive efforts using linkage and candidate gene approaches , the genetic etiology for the majority of families with a multi-generational breast cancer predisposition is unknown . In this study , we used whole-exome sequencing of thirty-three individuals from 15 breast cancer families to identify potential predisposing genes . Our analysis identified families with heterozygous , deleterious mutations in the DNA repair genes FANCC and BLM , which are responsible for the autosomal recessive disorders Fanconi Anemia and Bloom syndrome . In total , screening of all exons in these genes in 438 breast cancer families identified three with truncating mutations in FANCC and two with truncating mutations in BLM . Additional screening of FANCC mutation hotspot exons identified one pathogenic mutation among an additional 957 breast cancer families . Importantly , none of the deleterious mutations were identified among 464 healthy controls and are not reported in the 1 , 000 Genomes data . Given the rarity of Fanconi Anemia and Bloom syndrome disorders among Caucasian populations , the finding of multiple deleterious mutations in these critical DNA repair genes among high-risk breast cancer families is intriguing and suggestive of a predisposing role . Our data demonstrate the utility of intra-family exome-sequencing approaches to uncover cancer predisposition genes , but highlight the major challenge of definitively validating candidates where the incidence of sporadic disease is high , germline mutations are not fully penetrant , and individual predisposition genes may only account for a tiny proportion of breast cancer families .
Around one in six women who develop breast cancer has a first degree relative with the condition [1] . In the mid 1990s , a classical linkage approach identified germline mutations in two genes , BRCA1 and BRCA2 , which are associated with a high risk of developing both breast and ovarian cancer [2] , [3] . Although BRCA1 and BRCA2-specific genetic testing is rapidly evolving in the clinical setting , mutations in these genes are successful at explaining only around half of the dominant multi-case breast cancer only families [4] , and their contribution to the heritable risk of breast cancer has been estimated to be no more than around 20% of the total [5] , [6] . Importantly , the identification and management of individuals with high-risk breast cancer predisposition gene mutations is now well accepted in clinical practice . Although evidence-based risk management is only possible in a relatively small group of families , as it is limited by the identification of an underlying genetic mutation , the benefits for those individuals are well established [7] . Through a candidate gene approach , mutations in other high and moderate penetrance cancer-susceptibility genes have been identified in a further small proportion of families but the underlying etiology of the increased susceptibility to breast cancer in the majority of multi-case breast cancer families remains unknown . Recent advances in massively parallel sequencing technology have provided an agnostic means by which to efficiently identify germline mutations in individuals with inherited cancer syndromes at the individual family or cancer-specific level [8] , [9] . The aim of this study is to identify through a whole exome sequencing approach , the underlying familial predisposition to breast cancer in multiple multi-generational breast cancer families in whom no BRCA1 or BRCA2 mutation was identified ( BRCA1/2 negative families ) , and to assess the candidate genes identified by this means in a cohort of familial BRCA1/2 negative breast and ovarian cancer patients .
We performed intra-family exome sequence analysis of multiple affected relatives from 15 high-risk , trans-generational breast cancer families in whom full BRCA1 and BRCA2 mutation analysis had been performed and was uninformative in at least one breast cancer-affected family member ( Table 1 ) . Sequencing was performed on GAIIx or HiSeq instruments ( Illumina ) . The average read depth achieved for target regions was 83 . 19 and at least 80% ( average 89 . 12% ) of the capture target regions were covered by 10 or more sequence reads for all samples ( Table S1 ) . Following data filtering , an average of 35 overtly deleterious and 284 non-synonymous mutations were identified per individual ( Table S1 ) . To identify candidate predisposition genes we only considered those with overtly deleterious mutations that were shared by multiple affected relatives and/or were targeted in more than one family and further priority was given to genes with a role in mechanistically well-established breast cancer–associated DNA repair . A list of all overtly deleterious mutations identified in among the 33 individuals sequenced is provided in Table S2 . Two of the fifteen families were found to carry independent heterozygous truncating mutations in the Fanconi Anemia ( FA ) gene , FANCC . Neither family was reported to be of Ashkenazi Jewish ancestry and the mutations are different to those commonly reported among this ethnic group . Family 1 carried a novel nonsense mutation ( FANCC c . 535C>T , p . Arg179* ) that was present in the youngest affected individual ( breast cancer at age 37 ) and in her mother who had ovarian cancer at age 66 , but not in her breast cancer-affected sister who was diagnosed at age 46 ( Figure 1 ) . Family 2 was found to harbor a known pathogenic FA mutation ( FANCC c . 553C>T , p . Arg185* ) [10] which was present in two sisters who developed breast cancer aged 36 , and bilateral breast cancer aged 46 and 53 , respectively . A third family analyzed by exome sequencing was found to carry a heterozygous c . 1993C>T mutation in the BLM gene which is predicted to truncate the protein at codon 645 ( p . Gln645* ) . This known pathogenic Bloom syndrome mutation [11] co-segregated with cancer in the family ( Figure 1 ) , being present in all three sisters diagnosed with breast cancer aged 39 , 39 and 41 years respectively and absent in the two unaffected sisters . Although retrospective likelihood segregation analysis of these limited pedigrees did not reach significance ( see Text S1 ) , overall , co-segregation of FANCC and BLM mutations in these families appears consistent with that expected for moderately penetrant breast cancer alleles . Mutation analysis of all coding exons of FANCC and BLM was extended to the index cases from a further 438 BRCA1/2 negative breast cancer families ( from kConFab ) . This approach identified one further family with a heterozygous , known pathogenic FANCC mutation , ( c . 67delG , p . Asp23Ilefs*23 , rs104886459 ) [12] and one with a heterozygous pathogenic BLM mutation ( c . 2695C>T , p . Arg899* ) [11] . For FANCC , mutation hotspot exons 2 , 5 , 7 , 14 and 15 were screened in the index cases from an additional 957 BRCA1/2 uninformative breast cancer families attending familial cancer services ( including 561 obtained from the Peter MacCallum Cancer Centre Familial Cancer Centre and a further 396 from kConFab ) . One further family with a heterozygous FANCC c . 1661T>C ( p . Leu554Pro , rs104886458 ) missense variant , which is a functionally validated pathogenic FA mutation , was identified [13] . The index case in the FANCC c . 67delG family developed breast cancer at age 60 but independent clinical testing subsequently identified a deleterious mutation in BRCA2 ( c . 8297delC , p . Thr2766Asnfs*11 ) in other breast cancer-affected family members ( Figure 1 ) . Genotyping of both mutations within this family suggests that different individuals may carry risk conferred by one or both of these family mutations . The index case of the FANCC c . 1661T>C family developed bilateral breast cancer at age 44 and 55 , but DNA from other family members was not available for segregation analysis . All FANCC variants detected in index cases or controls are summarized in Table S3 . The index case of the BLM c . 2695C>T family developed breast cancer at age 33 but segregation analysis showed the mutation was inherited from her father rather than her mother whose reported family history of breast cancer had initiated their recruitment into kConFab ( Figure 1 ) . Interestingly , breast cancer was diagnosed much earlier in the index case compared to her maternal relatives ( 33 years versus 58 to 73 years ) possibly indicating a different genetic etiology . Unfortunately data regarding family history on the paternal side are limited . Neither the father nor the paternal grandparents were reported to have developed cancer but no further information regarding number or cancer status of other relatives is available . All BLM variants detected in index cases or controls are summarized in Table S4 . No pathogenic BLM mutations were detected in 464 healthy controls and none have been reported in the 1000 Genomes data ( 20100804 release , n = 1 , 092 ) [14] compared to 2/438 breast cancer families with BLM mutations . Likewise , no known pathogenic or overtly deleterious FANCC mutations were identified among the 464 controls or the 1000 Genomes data or among 654 healthy controls examined in an independent study [15] . The Exome Variant Server ( EVS ) , NHLBI Exome Sequencing Project , Seattle , WA , does report deleterious mutations in FANCC and BLM in 3/3 , 510 and 4/3 , 510 individuals of European decent , respectively . However , this cohort includes extreme tail sampling of traits relating to heart , lung and blood disorders . The latter group in particular may be expected to show enrichment for mutations in DNA repair machinery including FA genes . Excluding the Exome Variant Server frequency data , a total of 4/1 , 395 breast cancer families screened for all or at least the mutation hot spot exons carried overtly deleterious FANCC mutations compared to none among the combined control population ( n = 2 , 210 ) . While this is indicative that overtly deleterious mutation in FANCC and BLM are likely to be very rare in the population this must be considered a crude measure as the controls were drawn from diverse populations the majority of which were not matched to the index cases . However , it is possible that more families in our breast cancer family cohort may be explained by FANCC and BLM mutations since , for both genes , private non-synonymous variants were identified that are predicted to be damaging by in silico algorithms . One such variant , for which there was DNA available for segregation analysis , was FANCC p . Arg185Gln . This variant closely segregated with disease in this family , which included four female blood relatives with breast cancers diagnosed at ages 34 , 51 , 47 and 62 ( Figure 1 ) . The p . Arg185Gln variant was identified in 1/1 , 395 breast cancer families but not in any of 464 controls and has not been reported in the 1000 Genomes project or EVS database . Homozygous mutations in FANCC and BLM are responsible for FA ( complementation group C ) and Bloom syndrome , respectively , and individuals diagnosed with these syndromes have a high risk of cancer . Functionally , the FA and Bloom syndrome pathways play important roles in homologous recombination ( HR ) based repair of double-stranded DNA breaks [16] , [17] . Constitutional inactivating mutations in genes integral to error-free HR and responsible for FA have been clearly associated with an increased susceptibility to both breast and ovarian cancer [16] , and include the genes BRCA1 , BRCA2 ( FANCD1 ) , FANCN ( PALB2 ) , FANCJ ( BRIP1 ) , RAD51C ( FANCO ) and RAD51D . Thus , in addition to the direct genetic evidence that we have described here , FANCC and BLM are strong candidates for breast cancer susceptibility genes due to their role in the precise regulation of HR and some of its associated functions . Although there is limited data , heterozygous FANCC mutations have previously been linked to an increased incidence of breast and early onset pancreatic cancer [15] , [18] , [19] , however , no excess breast and ovarian cancer was observed among Ashkenazi Jews carrying the FANCC c . 711+4A>T mutation [20] . While another previous study failed to identify overtly pathogenic FANCC mutations in breast cancer , the study cohort size was small ( n = 88 ) [21] . In keeping with our data , two recurrent truncating mutations in the BLM gene were shown in a case control study to be associated with increased breast cancer risk in Russia [22] . Gruber et al reported an elevated risk of colorectal cancer in Ashkenazi Jews carrying the common BLMASH mutation and a non-significant excess of breast cancer [23] although a later study failed to confirm these findings [24] . Further to the germline mutations in FANCC and BLM , exome sequencing identified mutations in the breast cancer predisposition genes , PTEN and BRCA2 in an additional three of the original 15 families ( Figure S1 ) . The truncating PTEN mutation ( c . 217G>T , p . Glu73* ) was identified in only one branch of the family suggesting another susceptibility gene may explain the extended family history . Prior to this finding , the treating familial cancer centre reported no PTEN-associated clinical features within the family . In family 5 , exome sequencing identified a deleterious BRCA2 mutation ( c . 5722_5723delCT , p . Leu1908Argfs*2 , rs80359530 ) in two of the three family members tested ( Figure S1 ) . The mutation is present in a male diagnosed with breast cancer but not in the youngest affected female relative in the family , who had been offered the original clinical BRCA1 and BRCA2 mutation test in the clinic setting . Similarly in family 6 , exome sequencing identified a deleterious BRCA2 mutation ( c . 26delC , p . Pro9Glnfs*16 , rs80359343 ) in a female diagnosed with breast cancer at age 30 , but not in her cousin who was diagnosed at age 36 and was the only family member to have undergone full diagnostic BRCA1 and BRCA2 gene sequencing ( Figure S1 ) . These families are interesting in a clinical context since they were designated as unresolved on the basis of best clinical practice and demonstrate the need for targeted sequencing of all proven breast and ovarian cancer susceptibility genes to obtain maximum information in the clinical setting ( as previously demonstrated [25] ) . Our data also highlights the major challenge confounding genetic studies of common adult onset familial disease; the presence of ‘phenocopies’ in families with an inherited genetic predisposition and/or the convergence of pedigrees with different genetic causes ( e . g . PTEN family 4 ) . Among the remaining nine breast cancer families there were numerous genes that were recurrently targeted that warrant further investigation . It is noteworthy that in one family , one individual harbored a known FA pathogenic truncating mutation in FANCL . Mutation of this gene is responsible for a very small fraction of FA families and only three pathogenic mutations in FANCL are recorded in the Fanconi Anemia Mutation Database . In conclusion , we describe two potential breast cancer susceptibility genes FANCC and BLM both of which have functional roles in the regulation of HR . The heterozygous mutation carrier rate in Caucasians for these genes is extremely low ( for FANCC it is estimated at 1/3 , 000 [15] , whilst the carrier frequency of BLM mutations is unknown since the syndrome is exceedingly rare ) and notwithstanding the possibility of the “winners curse” [26] , the exome sequencing data is strongly suggestive that FANCC and BLM represent breast cancer predisposing genes . Together with the recently identified association of RAD51 paralogues with cancer predisposition [27] , [28] , our findings suggest that the number of unidentified moderate to high-risk susceptibility genes is very much larger than previously expected and the number of families explained by each gene is likely to be much less than 1% ( cf . RAD51C [27] , [29] ) . Consequently , providing definitive evidence for a causative role for novel breast cancer genes will be challenging and will require validation of rare mutations in thousands rather than hundreds of families . We predict that this will be a generic problem associated with identifying causative mutations in common diseases such as breast cancer and that validation rather than the technical exercise of exome sequencing is where the real challenge lies .
This study was approved by the Peter Mac Ethics Committee ( project numbers 09/62 and 11/50 ) . Informed consent was obtained from all participants . Fifteen high-risk breast cancer families with at least four cases of multi-generational breast cancer including at least one additional high-risk feature ( such as bilateral , early onset or male breast cancer , or ovarian cancer ) and at least two available blood specimens from breast cancer-affected individuals , were selected for whole exome sequencing from among approximately 800 BRCA1 and BRCA2 mutation negative families from the Kathleen Cunningham Foundation Consortium for Research into Familial Breast Cancer ( kConFab ) , which has been collecting biospecimens and clinical and epidemiological information from families recruited through Familial Cancer Centres in Australia and New Zealand since 1997 [30] . DNA from two or three breast cancer-affected individuals were obtained from each family for analysis ( as shown in Table 1 ) , at least one of whom had previously been screened for BRCA1 and BRCA2 mutations ( by sequencing of all coding exons and Multiplex Ligation-dependent Probe Amplification ) . Blood DNA from index cases from a further 834 mutation negative kConFab families and 561 mutation negative families obtained from the Peter MacCallum Cancer Centre Familial Cancer Centre were obtained for mutation analysis of candidate genes . Of those index cases obtained through the Familial Cancer Centre , individuals were breast cancer-affected , had a strong family history and been assessed for the probability of harboring a BRCA1 or BRCA2 mutations using BRCAPRO [31] and had been found on the basis of a verified family and personal history of having a 10% or greater probability . The index cases had undergone full diagnostic BRCA1/2 mutation search and no mutation was identified . However , it should be noted that the majority of these families did not fulfill the very stringent family history criteria that was required for recruitment to kConFab , the research cohort from which the families for the initial exome sequencing were taken [30] . Non-cancer control DNA samples were obtained from kConFab ( 226 age- and ethnicity-matched best friend controls ) and from the Princess Anne Hospital , UK ( 238 Caucasian female volunteers , as described previously [32] ) . DNA for candidate gene mutation analysis underwent whole genome amplification ( WGA ) using Repli-G Phi-mediated amplification system ( Qiagen ) prior to mutation analysis . 2–3 µg of DNA was fragmented to approximately 200 bp by sonication ( Covaris ) and used to prepare single- or paired-end libraries using the SPRIworks Fragment Library System I for Illumina Genome Analyzer on the SPRI-TE Nucleic Acid Extractor ( Beckman Coulter ) . Exome enrichment was performed using the NimbleGen Sequence Capture 2 . 1 M Exome Array , EZ Exome Library ( Roche NimbleGen ) or SureSelect Human All Exon version 2 or 50 Mb libraries ( Agilent Technologies ) according to the recommended protocols . Sequencing was performed on GAIIx or HiSeq instruments ( Illumina ) . Library preparation and sequencing details for each sample are provided in Table S1 . We did not observe any significant differences in performance of the different exome capture platforms . Paired-end sequence reads were aligned to the human genome ( hg19 assembly ) using the Burrows–Wheeler Aligner ( BWA ) program [33] . Local realignment around indels was performed using the Genome Analysis Tool Kit ( GATK ) software [34] . Subsequently , duplicate reads were removed using Picard and base quality score recalibration performed using GATK software . Single nucleotide variants ( SNVs ) and indels were identified using the GATK Unified Genotyper and variant quality score recalibration . Variants were annotated with information from Ensembl release 62 using Ensembl Perl Application Program Interface ( API ) including SNP Effect Predictor [35] , [36] . Single-end sequence reads were aligned as above except duplicate reads were flagged prior to base quality score recalibration and included in variant calling . Variants were first filtered for confident calls originating from bidirectional sequence reads using a quality threshold of ≥30 , read depth of ≥10 and allele frequency ≥0 . 15 . Prior to further filtering , variants were assessed for overtly deleterious mutation in known breast cancer associated genes [25] . Then , all variants present in the dbSNP database v132 , except those also reported in the public version of the Human Gene Mutation Database ( HGMD ) [37] were removed , as were all common variants detected in >10 out of 33 exomes . Next , variants with functionally deleterious consequences ( nonsense SNVs , frameshift indels , essential splice variants and complex indels ) were identified for evaluation [35] . Functionally deleterious variants were evaluated in each individual as well as pairwise between relatives . Primers flanking the BRCA2 , PTEN , FANCC and BLM mutations identified by whole exome sequence analysis were used to amplify germline DNA from affected index cases and all available relatives . The purified products were directly sequenced using BigDye terminator v3 . 1 chemistry on a 3130 Genetic Analyzer ( Applied Biosystems ) . High resolution melt ( HRM ) analysis was performed on duplicate PCR products amplified from 15 ng WGA DNA . Primer sequences and PCR conditions are provided in Table S5 . Melt analyses were performed on a LightCycler 480 Instrument using Gene Scanning Software ( Roche ) . Duplicate PCR products exhibiting variant DNA melt curves were Sanger sequenced to identify sequence variations . All novel sequence variants were confirmed by Sanger sequencing an independent PCR amplified from non-WGA DNA . The functional effect of missense variants were evaluated using in silico prediction tools SIFT and PolyPhen-2 [38] , [39] . The following GenBank reference sequences were used for variant annotation: FANCC , NM_000136 BLM , NM_000057; PTEN , NM_000314 and BRCA2 , NM_000059 . 1000 Genomes Browser , http://browser . 1000genomes . org/; Ensembl , http://www . ensembl . org/index . html; The Genome Analysis Toolkit , http://www . broadinstitute . org/gsa/wiki/index . php/The_Genome_Analysis_Toolkit; HGMD , http://www . hgmd . org/; Picard , http://picard . sourceforge . net; HGVS nomenclature for the description of sequence variants , http://www . hgvs . org/mutnomen/; NCBI SNP database , http://www . ncbi . nlm . nih . gov/projects/SNP/; The Fanconi Anemia Mutation Database , http://www . rockefeller . edu/fanconi/; BLMbase mutation registry , http://bioinf . uta . fi/BLMbase/; SIFT , http://sift . jcvi . org/; PolyPhen-2 , http://genetics . bwh . harvard . edu/pph2/ . Exome Variant Server , http://evs . gs . washington . edu/EVS/ . | Currently , we know that a woman who inherits a fault in one of two genes , BRCA1 or BRCA2 , has a high risk of developing both breast and ovarian cancer . However , such faults account for only half of all families with a strong family history of breast cancer . In this study , we planned to identify new genes that may be associated with an increased risk of developing breast cancer by looking for faults in every gene in the blood DNA of multiple women with breast cancer from large families with a strong family history of the condition over multiple generations . We can then track which gene fault is present in all the women with breast cancer in that family and in other families , but is not found in the women who did not develop breast cancer or have no family history . Using this approach , we identified faults in two genes , Fanconi C and Bloom helicase , in six families . Faults in these genes appear to increase the risk of developing breast cancer . Both these genes work in a similar way as BRCA1 and BRCA2 , and this highlights the importance of these functions in preventing breast cancer . Further studies need to be done to confirm our results . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"cancer",
"genetics",
"genome",
"sequencing",
"genome",
"analysis",
"tools",
"genetic",
"testing",
"biology",
"trait",
"locus",
"heredity",
"genetic",
"association",
"studies",
"genetic",
"screens",
"genetics",
"genomics",
"genetics",
"of",
"disease",
"genetics",
"and... | 2012 | Exome Sequencing Identifies Rare Deleterious Mutations in DNA Repair Genes FANCC and BLM as Potential Breast Cancer Susceptibility Alleles |
The clinical spectrum of human disease caused by the roundworms Toxocara canis and Toxocara cati ranges from visceral and ocular larva migrans to covert toxocariasis . The parasite is not typically recovered in affected tissues , so detection of parasite-specific antibodies is usually necessary for establishing a diagnosis . The most reliable immunodiagnostic methods use the Toxocara excretory-secretory antigens ( TES-Ag ) in ELISA formats to detect Toxocara-specific antibodies . To eliminate the need for native parasite materials , we identified and purified immunodiagnostic antigens using 2D gel electrophoresis followed by electrospray ionization mass spectrometry . Three predominant immunoreactive proteins were found in the TES; all three had been previously described in the literature: Tc-CTL-1 , Tc-TES-26 , and Tc-MUC-3 . We generated Escherichia coli expressed recombinant proteins for evaluation in Luminex based immunoassays . We were unable to produce a functional assay with the Tc-MUC-3 recombinant protein . Tc-CTL-1 and Tc-TES-26 were successfully coupled and tested using defined serum batteries . The use of both proteins together generated better results than if the proteins were used individually . The sensitivity and specificity of the assay for detecting visceral larval migrans using Tc-CTL-1 plus Tc-TES-26 was 99% and 94% , respectively; the sensitivity for detecting ocular larval migrans was 64% . The combined performance of the new assay was superior to the currently available EIA and could potentially be employed to replace current assays that rely on native TES-Ag .
The roundworms Toxocara canis and Toxocara cati cause a broad spectrum of clinical disease in humans ranging from visceral and ocular larva migrans to covert and common toxocariasis . Children are at particular risk of toxocariasis when they play in areas , potentially contaminated with Toxocara eggs , such as playgrounds or sandboxes and ingest embryonated roundworm eggs . After ingestion , the eggs hatch in the gut and larvae disseminate hematogenously to the lungs , liver , muscle , brain , and/or eyes . Once in the tissues , the larvae are unable to continue their normal life cycle , and a local inflammatory response to larvae leads to the varied symptoms of toxocariasis ( visceral , ocular larva migrans , and covert toxocariasis ) , and can lead to cerebritis and eosinophilic meningitis when larvae enter the central nervous system [1–6] . A third so-called “covert” form of toxocariasis has been linked to more subtle pulmonary and cognitive dysfunctions [7–9] , and even educational deficits [10] . Currently , diagnosis for toxocariasis relies on clinical signs , history of exposure to puppies or kittens , laboratory findings ( including eosinophilia ) , and the detection of antibodies to Toxocara antigens . The enzyme immunoassay ( EIA ) using T . canis excretory secretory antigens ( TES-Ag ) from infective-stage larvae is the most useful diagnostic test for toxocaral visceral larva migrans ( VLM ) and ocular larva migrans ( OLM ) and is the preferred assay used by most laboratories in the U . S . and worldwide [4 , 11] . The TES-Ag EIA has proven to be robust and reliable , although questions about specificity and reduced sensitivity leave ample room for improvement in laboratory diagnosis of toxocariasis [4 , 6 , 11] . In temperate countries , TES-Ag EIA and TES-western blot can provide sufficient support for clinical suspicion; however , testing is not widely available because of the limited availability of antigen made from T . canis larvae . The proven TES-Ag cross-reactivity with antibodies from other common helminth infections of humans also reduces the usefulness of native , unfractionated TES-Ag-based serodiagnosis in regions where poly-parasitism is endemic [6] . Recent efforts have focused on identification of recombinant proteins to improve sensitivity and specificity and to reduce reliance on native parasite materials [4] . Several groups have cloned , expressed , and developed EIAs based on recombinant antigens of the assay [12–16] . Ultimately , a diagnostic method that utilizes one or more recombinant diagnostic antigens could result in improved assays that are more widely accessible to health care providers . Our study aim was to identify one or more immunoreactive proteins found in the TES product from T . canis infective larvae and to develop at least one of these recombinant proteins as a diagnostic reagent in a multiplex bead format assay ( Luminex ) that could replace TES-Ag .
All clinical samples used in this study were collected in previous studies with specific permission for future use of stored samples ( CDC Study Protocol Number 3580 ) . Samples were anonymized and the study was performed in compliance with protocols approved by the ethical review boards of the CDC . Four sets of defined sera were used: ( 1 ) Two hundred and four sera from cases with presumed Toxocara spp . visceral larval migrans ( VLM ) , based on the presence of clinical symptoms and signs and reactivity in the in-house TES-Ag EIA [11] and TES-Ag Western blot; ( 2 ) Fifty sera from cases of presumed Toxocara spp . ocular larval migrans ( OLM ) based on the presence of clinical symptoms and signs only ( positivity with TES-Ag Western blot is not required to define OLM ) ; ( 3 ) A control group consisting of two hundred and eighty eight sera from healthy U . S . residents; ( 4 ) A convenience panel for cross-reactivity evaluation consisting of one hundred and twenty U . S . patients infected with various infections other than Toxocara spp . and fourteen sera from Egyptians negative to Toxocara spp . Each sample was not exhaustively tested for all other parasites . Sets number 3 and number 4 were tested with TES-Ag Western blot . Five serum samples collected from people with no travel history outside the United States were pooled and used as the ‘normal’ control serum pool . For TES 2D gel , we used a positive anti-Baylisascaris procyonis serum from a baboon that developed severe neural larval migrans after experimental infection with embryonated B . procyonis eggs . A Toxocara-positive serum pool was prepared by combining in equal volume of 10 EIA—positive serum samples . For 2D gel electrophoresis and mass spectrometry analysis we utilized TES-Ag provided by Dr . Steven Kayes’ Laboratory at the University of South Alabama , Mobile , AL . Briefly , larvae from artificially de-shelled embryonated T . canis eggs were cultured in RPMI 1640 at 37°C with saturated humidity and gassed with 5% CO2 . The culture supernatants containing TES-Ag were collected , pooled , and concentrated [17] . Protein concentrations were determined using the Bradford Protein Assay and BCA Protein Quantification Assay ( Pierce Biotechnology Rockford , IL ) . The 2D gel electrophoresis was performed following a protocol used previously [18] . Fifty μg of TES-Ag sample was separated on 11 cm , non-linear , pH 3–10 gradient Immobilized pH Gradient ( IPG ) strips ( Bio-Rad , Hercules , CA ) and after isoelectric focusing , the second dimension was carried out using Criterion XT 4–12% Bis-Tris pre-cast sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) gels . Three of the 2D gels were transferred to nitrocellulose membranes and blotted and probed with a 1:50 dilution of a strong EIA positive Toxocara human sera pool , negative human serum sample , or a B . procyonis positive serum; the fourth gel was stained using a mass-spectrometry compatible silver stain . Proteins were chosen based on the comparative reactivity seen in the blots and were manually excised from silver-stained 2D gels . The target proteins were digested with trypsin and the resulting peptides analyzed by electrospray ionization mass spectrometry [18] . The Mascot program was used to identify proteins from peptide sequence databases from the mass spectrometry data . TES-Ag proteins were electrophoretically separated using Criterion TGX ( Bio-Rad , Cat . # 567–1092 ) at 10 ng/mm and then transferred to nitrocellulose membrane ( Whatman Protran BA83 , Cat . # 10 541 103 , 0 . 2m pore size ) . The blots were cut into 2 . 5 mm strips and stored in PBS + 0 . 1% NaN3 at 4°C prior to use . Sera were tested and specific antibodies were detected as described previously [29 , 30] . A serum was considered positive if reactivity occurred with any bands at 24 , 28 , 30 , or 35 kDa [31 , 32] . Data were tabulated and analyzed using Microsoft Excel . Determination of the cut-off value and assay performance was obtained by using R statistical software version 3 . 0 . 1 ( R Foundation for Statistical Computing , Vienna , Austria ) and pROC package [33] . To combine the results of diagnostic tests , we used a method which finds optimal linear combination of multiple antigens[27] . In short , this procedure searches for the coefficients a and b in the equation below: =a*MFITc-CTL-1+b*MFITc-TES-26 ( MFI minus background is shortened to MFI in the equation for simplicity ) . In this formula , a and b are chosen to maximize the area under the ROC curve ( AUC ) and y is the combined value of the two MFI values . In order to avoid bias in the computation of the AUC , the AUC is estimated via cross-validation . Further details on the calculations can be found elsewhere [27] .
From the 2D gel electrophoresis and mass spectrometry analysis , we identified 24 hits/spots that were reactive to T . canis positive serum , and lacked reactivity with both a B . procyonis positive baboon serum and a normal human serum ( Fig 1A–1C ) . The corresponding protein spots were excised from the silver-stained gel and prepared for MS analysis ( Fig 1D ) . An initial MASCOT search revealed 3 hits from spots 17 , 22 , and 23 . The excretory/secretory mucin , MUC-3 , from T . canis , was identified in spot 17 , and excretory/secretory C-type lectin , Tc-CTL-1 , from T . canis , at spots 22 and 23 ( Table 1 ) . When the MS data were used to search a T . canis Expressed Sequence Tag ( EST ) database , from 7 spots ( 2 , 15 , 17 , 20 , 22 , 23 , 24 ) we found nine mRNAs ( Histone H4 , Actin containing A3R Repeat , Actin , Tc-MUC-3 , Fibrinogen beta and gamma chains , Glyceraldehyde 3-Phosphate Dehydrogenase , Elongation Factor , Tc-TES-26 , Tc-CTL-1 ) with acceptable scores and percent sequence coverage . However , we identified only 3 proteins of interest ( Tc-MUC-3 , Tc-TES-26 , and Tc-CTL-1 ) ( Table 2 ) . Tc-TES-26 from T . canis was identified as the protein in spot 20 . The remaining 17 spots were determined to be human protein contaminants . We expressed Tc-CTL-1 , Tc-TES-26 , and Tc-MUC-3 in bacterial expression system . These three antigens were expressed as fusion proteins with GST tags to allow easy recombinant protein purification and also for improving the coupling possibility of the proteins to the magnetic beads . The purity and antigenic potential of the proteins could be seen in Fig 2 . Among the three proteins , Tc-MUC-3 showed weaker reactivity when tested against the T . canis positive serum ( Fig 2C ) . The three antigens , Tc-CTL-1 , Tc-TES-26 and Tc-MUC-3 , were selected for further evaluation in the Luminex assay based on the MS data . The Tc-MUC-3 antigen , after coupled to the magnetic beads , did not show differentiation between positive and negative sera although the protein performed well in the immunoblot format . No further analysis of Tc-MUC-3 was performed . Tc-CTL-1 and Tc-TES-26 were coupled to the MagPlex Magnetic Beads and Luminex based assays were developed . Each set of beads was tested using the defined serum batteries . The Tc-CTL-1 Luminex assay performed significantly better than the Tc-TES-26 ( p < 0 . 001 ) for diagnosis of visceral toxocariasis ( Fig 3 , Table 3 ) . In comparison to the TES-Ag EIA , the Tc-CTL-1 Luminex assay also performed well for detecting visceral toxocariasis , however , only 54% of OLM cases were detected using the Tc-CTL-1 Luminex compared to 100% using the TES-Ag EIA ( Table 3 ) . We used the TES-Ag Western blot to better define the sera used from the OLM cases; only 70% of the OLM sera were reactive in the TES-Ag Western blot . When we restricted the OLM case definition to include only TES-Ag Western blot positive sera , the sensitivity and specificity increased 1% and 2% for Tc-CTL-1 and 3% and 3% for Tc-TES-26 . To establish the specificity of the Luminex based assays , we used the sera from infections other than Toxocara spp . The Tc-TES-26 Luminex assay showed more cross-reactivity than the Tc-CTL-1 Luminex; only amebiasis and E . nana infections cross-react with Tc-CTL-1 ( Table 4 ) . Assay performance was improved when a combination of Tc-CTL-1 and Tc-TES-26 was used ( Fig 3 and Table 3 ) . The combination of the two antigens increased the sensitivity to 99% but lowered the specificity to 94% although this improvement was not statistically significant compared to Tc-CTL-1 ( p = 0 . 10 ) . The coefficients of variation ( CV ) for intra-plate assays were 5% for Tc-CTL-1 and 4% for Tc-TES-26 . For inter-plate variation , the CV for the low positive control was quite large at 23% for Tc-CTL-1 and 31% for Tc-TES-26 , probably because of the low value of the observations; but for medium calibrator , the CV was 7% for Tc-CTL-1 and 12% for Tc-TES-26 .
The study confirmed previous studies on immunodominant TES antigens [1 , 13 , 16 , 19 , 20 , 23 , 24] . Briefly , we identified three major antigenic proteins from the TES-Ag , and based on these findings , expressed those antigens for further analysis as diagnostic reagents . One antigen , Tc-MUC-3 , after coupled to Luminex beads , did not produce a functional assay . We have no explanation for this , as we tried several methods of coupling including titration of antigens , different buffers , and pH conditions . It is possible that the low reactivity of this protein to positive sample causes the failure for coupling the protein to the beads . The performances of Tc-CTL-1 and Tc-TES-26 based on total IgG responses are comparable to the reported performances of the same antigens . Yamasaki et al . ( 2000 ) [16] , based on 11 subjects with toxocariasis , reported a sensitivity of 100% for Tc-CTL-1 and a specificity of 98% ( 3 out of 142 cross-reactors sera ) . From the study of Mohamad et al . ( 2009 ) [15] , and Norhaida et al . ( 2008 ) [34] , the sensitivity and the specificity of Tc-CTL-1 is 92–93% and 94–90% , respectively . The Tc-CTL-1 in our study has a sensitivity of 90% and a specificity of 99% . The differences of the performances of the assays due to a possibility that Yamasaki study used small number of positive sera and also soil-transmitted helminths were not prevalent in Japan . A Low prevalence of soil-transmitted helminths might contribute to less background or reactivity to the assay and will lower the cut-off points , thereby improving the sensitivity . In Norhaida study , the group used IgG4 , instead of total IgG [34] . For Tc-TES-26 , Mohamad et al . ( 2009 ) [15] reported the assay has a sensitivity of 80% and specificity of 96% ( similar to Tc-TES-26 Luminex performance ) , but this performance was based on IgG4 responses . Our assay was not based on specific IgG4 detection and it is possible that we could improve the sensitivity and specificity of the assay if we were to use the IgG4 specific responses [15 , 34] . Recombinant-based assays to determine antibody responses usually have lower sensitivity than those based on native/crude antigens because reactivity with multiple antigens can occur when crude mixtures are used . As expected , the assays developed here using Tc-CTL-1 and Tc-TES-26 , are less sensitive than the TES-Ag EIA or TES-Ag Western Blot , which utilize a crude parasite antigen . However , the combination of the two antigens resulted in an improved sensitivity of 99% for VLM detection that equaled the sensitivity of the TES-Ag Western Blot . Not unexpectedly , when sensitivity was maximized , the assay specificity decreased to 94% when the combination of two antigens was used . This specificity is not optimum , but it is still much better than the reported specificities of TES-Ag EIAs which are approximately 85% [4] . Although the Luminex based assays all represented improvements in detecting VLM , all were less sensitive than the TES-Ag EIA or even the TES-Ag Western blot method for detecting OLM cases . While additional antigens might improve the sensitivity for detection of OLM , we wonder how many individual antigens might be needed since the TES-Ag western blot , which is a complex mixture of multiple antigens failed to detect 30% of OLM subjects . Compared to the reference TES-Ag EIA , the use of the Luminex platform has advantages: more samples could be tested concurrently and downstream , a multiplex assay could be developed to distinguish human larval migrans syndromes caused by other helminths such as B . procyonis , another important cause of larval migrans in the U . S . , as well as other larval helminths species [2 , 35] . Although the Luminex-based assay offers advantages against the EIA ( in term of performance and multiplexing capability ) , the capital cost is much higher than for an EIA [36] . If the system is only used for detecting responses against single antigen , the Luminex-based assay is not cost effective and does not offer advantage on time-saving . For each additional antigen , the time and cost savings for running the Luminex-based assay are much better than running several EIA . The capital cost for running Luminex-based assay even could be reduced further by using a more robust , field-friendly system of MagPix that has similar performance to the Luminex platform . The only disadvantage of MagPix is that the capability for multiplexing is only up to 50 antigens ( not 100 antigens as for the Luminex system ) . In the term of skill needed , Luminex-based assay is similar to the requirement of the skill for running an EIA .
Recombinant antigen based assays in the Luminex platform for visceral toxocariasis perform similarly to the existing TES-Ag Western Blot and better than the TES-Ag EIA method . The utilization of the Luminex assay significantly diminishes the need for native parasite materials which can be expensive and cause data variability . | The roundworms Toxocara canis and Toxocara cati cause a broad spectrum of clinical disease in humans . Children are at particular risk of toxocariasis when they play in areas potentially contaminated with Toxocara eggs , such as playgrounds or sandboxes and ingest embryonated roundworm eggs . Currently , diagnosis for toxocariasis relies on clinical signs , history of exposure to puppies or kittens , laboratory findings ( including eosinophilia ) , and the detection of antibodies to Toxocara antigens . The enzyme immunoassay using T . canis excretory secretory antigens from infective-stage larvae is the most useful diagnostic test for toxocaral visceral larva migrans ( VLM ) and ocular larva migrans ( OLM ) and is the preferred assay used by most laboratories in the U . S . and worldwide . Although the EIA has been robust and reliable , improvement should be made in the specificity of the assay and the availability of a consistent antigen source . The crude TES-Ag shows cross-reactivity with antibodies from other common helminth infections of humans which reduces the usefulness of native , unfractionated TES Ag-based serodiagnosis in regions where poly-parasitism is endemic . To improve the assay performance , target antigenic proteins from T . canis excretory secretory antigens were identified using 2D gel electrophoresis . Three antigenic proteins sequences were found , expressed , and developed into Luminex bead-based assays . The combined use of two recombinant antigens ( Tc-CTL-1 and Tc-TES-26 ) represents an improvement over the existing immunodiagnostic methods that rely on native parasite materials . In the future , additional antigens could be added from other parasites that cause larval migrans to form a single method for detecting larval migrans syndromes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion",
"Conclusions"
] | [] | 2015 | Development of a Luminex Bead Based Assay for Diagnosis of Toxocariasis Using Recombinant Antigens Tc-CTL-1 and Tc-TES-26 |
TDP-43 and FUS are RNA-binding proteins that form cytoplasmic inclusions in some forms of amyotrophic lateral sclerosis ( ALS ) and frontotemporal lobar degeneration ( FTLD ) . Moreover , mutations in TDP-43 and FUS are linked to ALS and FTLD . However , it is unknown whether TDP-43 and FUS aggregate and cause toxicity by similar mechanisms . Here , we exploit a yeast model and purified FUS to elucidate mechanisms of FUS aggregation and toxicity . Like TDP-43 , FUS must aggregate in the cytoplasm and bind RNA to confer toxicity in yeast . These cytoplasmic FUS aggregates partition to stress granule compartments just as they do in ALS patients . Importantly , in isolation , FUS spontaneously forms pore-like oligomers and filamentous structures reminiscent of FUS inclusions in ALS patients . FUS aggregation and toxicity requires a prion-like domain , but unlike TDP-43 , additional determinants within a RGG domain are critical for FUS aggregation and toxicity . In further distinction to TDP-43 , ALS-linked FUS mutations do not promote aggregation . Finally , genome-wide screens uncovered stress granule assembly and RNA metabolism genes that modify FUS toxicity but not TDP-43 toxicity . Our findings suggest that TDP-43 and FUS , though similar RNA-binding proteins , aggregate and confer disease phenotypes via distinct mechanisms . These differences will likely have important therapeutic implications .
Amyotrophic lateral sclerosis ( ALS ) , also called Lou Gehrig's disease , is a devastating neurodegenerative disease . It is a rapidly progressing motor neuron wasting disorder that leads to paralysis and death typically within 2–5 years of onset . There are no cures or effective treatments . Given the similarities in presentation and pathology of familial and sporadic disease , study of genes mutated in familial disease can shed light on mechanisms of both familial ALS and the more common sporadic form . The first familial gene associated with ALS was SOD1 [1] , and much research over the past 10–15 years has focused on mechanisms by which mutant SOD1 may cause motor neuron dysfunction and loss [2] . Insight into ALS changed dramatically in 2006 when the 43 kDa TAR-DNA-binding protein ( TDP-43 ) was identified as a protein that accumulates abnormally in the ubiquitinated pathological lesions that characterize brain and spinal cord tissue of almost every non-SOD1 ALS patient [3]–[5] . Similar TDP-43 inclusions were also identified in degenerating neurons in a subset of frontotemporal lobar degeneration ( FTLD-TDP ) cases [3]–[5] . TDP-43 is an RNA-binding protein with two RNA recognition motifs ( RRMs ) and a glycine rich domain [6] . In 2008 , several groups independently reported the identification of over 30 different mutations in the TDP-43 gene ( TARDBP ) in various sporadic and familial ALS patients [6]–[10] . TDP-43 mutations were subsequently identified in various FTLD-TDP cases [11] , [12] . Taken together , these studies strongly suggest that TDP-43 is a new human neurodegenerative disease protein . Wild-type ( WT ) TDP-43 accumulates abnormally in cytoplasmic , ubiquitinated inclusions in degenerating neurons of ALS and FTLD-TDP patients , and mutations in the TDP-43 gene are linked with disease in rare familial and sporadic cases . Despite these advances , how TDP-43 contributes to disease , which domain of TDP-43 drives aggregation , and how ALS-linked mutations affect TDP-43 function and aggregation remained unclear . To address these deficits , we investigated the pathogenic properties of TDP-43 in yeast . The yeast system is simple and fast and has highly conserved fundamental pathways that allow powerful insights into complex human neurodegenerative diseases such as Parkinson's disease , Alzheimer's disease , and ALS [13] . Therefore , we developed a yeast model of TDP-43 to study TDP-43 biology as well as the mechanisms of TDP-43 aggregation and toxicity . Expression of human TDP-43 in yeast resulted in cytoplasmic aggregation and toxicity , thus modeling key aspects of human TDP-43 proteinopathies . These studies revealed that RRM2 and the C-terminal domain of TDP-43 ( Figure 1A ) are required for aggregation and toxicity [14] . Notably , all but one of over 30 ALS-linked mutations reside in the C-terminal domain , which the yeast system defined as critical for toxicity . Moreover , a combination of pure protein studies and in vivo analyses in yeast demonstrated that ALS-linked TDP-43 mutations render TDP-43 more aggregation-prone and enhance toxicity [15] . These studies demonstrated that the aggregation propensity and severity of toxicity of TDP-43 variants observed in ALS could be recapitulated in yeast . Moreover , we have discovered a potent genetic modifier of TDP-43 toxicity in yeast , Pbp1 , which is connected with ALS in humans [16] . The human homolog of Pbp1 , ataxin 2 , harbors a polyglutamine tract that is greatly expanded ( >34 glutamines ) in spinocerebellar ataxia type 2 [16] . Importantly , intermediate-length polyQ expansions ( ∼27–33 glutamines ) in ataxin 2 are a significant genetic risk factor for ALS in humans [16] . Clearly , the power of yeast genetics can be exploited to define basic disease mechanisms of fundamental importance to human neurodegenerative disease . Shortly following the identification of mutations in TDP-43 in ALS , mutations in another gene encoding an RNA-binding protein , FUS ( fused in sarcoma; also known as TLS , translocated in liposarcoma ) , were connected to familial ALS [17] , [18] . Additional mutations in FUS have recently been identified in sporadic ALS cases and in a subset of frontotemporal lobar degeneration ( FTLD-FUS ) cases [19] , [20] . FUS is normally a nuclear protein , but ALS patients harboring FUS mutations exhibit prominent neuronal cytoplasmic FUS accumulations that appear devoid of TDP-43 [18] . Several other examples of neurodegenerative disease are beginning to emerge where the predominant disease phenotype is the cytoplasmic aggregation of wild-type FUS . These include some cases of juvenile ALS [21] , basophilic inclusion body disease [22] , as well as the majority of tau- and TDP-43-negative frontotemporal lobar degeneration cases [23] . Moreover , FUS is also aggregated in Huntington's disease; spinocerebellar ataxia ( SCA ) type 1 , 2 , and 3; and dentatorubropallidoluysian atrophy [24] , . These findings extend the spectrum of disorders associated with FUS aggregation beyond ALS and FTLD-FUS and suggest the importance of understanding mechanisms of aggregation of WT as well as mutant FUS . FUS was initially discovered as part of a chromosomal translocation associated with human myxoid liposarcomas [26] . Subsequent studies have revealed roles for FUS in transcription , RNA processing , and RNA transport [27]–[29] . In neurons , FUS is localized to the nucleus but is transported to dendritic spines at excitatory post-synapses in a complex with RNA and other RNA-binding proteins [30] . In further support of a role of FUS in maintaining neuronal architecture , primary hippocampal neurons cultured from FUS knockout mouse embryos display defects in spine morphology and decreased spine density [31] . It remains unclear , however , how loss of this function of FUS or perhaps a novel toxic gain-of-function associated with FUS mutations contribute to ALS . Importantly , it is also uncertain whether FUS is intrinsically aggregation-prone . Indeed , FUS might simply be a marker of disease that is sequestered by other aggregated components . FUS and TDP-43 possess a similar domain structure . Like TDP-43 , FUS has an RRM and a glycine-rich domain ( Figure 1A ) . Moreover , using a bioinformatic algorithm designed to identify yeast prion domains [32] , we recently identified novel “prion-like” domains in the N-terminal domain of FUS ( amino acids 1–239 ) and in the C-terminal domain of TDP-43 ( amino acids 277–414 ) ( Figure 1A , Figure S1 ) [33] . Similar to prion domains found in yeast prion proteins such as Sup35 , Ure2 , and Rnq1 , this domain is enriched in uncharged polar amino acids ( especially asparagine , glutamine , and tyrosine ) and glycine [32] , [34] . This type of domain encodes all the information necessary to form a prion in yeast [27]– . It should be noted , however , that this type of domain is not found in all prion proteins , including HET-s from Podospora anserina and mammalian prion protein ( PrP ) [33] , [34] . Remarkably , by using this bioinformatic algorithm [32] to score and rank the human proteome ( 27 , 879 human proteins ) for prion-like properties , FUS and TDP-43 ranked 15th and 69th , respectively [33] . Our findings raise the intriguing possibility that RRM proteins with predicted prion-like domains may be particularly relevant to ALS [15] , [33] , [35] , [36] . Virtually all the ALS-linked mutations in TDP-43 lie in its prion-like domain [33] . By contrast , only a few of the ALS-linked mutations in FUS lie in its prion-like domain [33] . Indeed , the majority of ALS-linked FUS mutations reside at the extreme C-terminal region [37] . The identification of two RNA-binding proteins with a similar domain architecture that aggregate and are sometimes mutated in ALS and FTLD gives rise to the emerging concept that RNA metabolic pathways may play a major role in ALS and FTLD pathogenesis [38] . Despite these similarities between TDP-43 and FUS , it is unknown whether TDP-43 and FUS aggregate and cause toxicity by similar mechanisms . Here , we address this issue and establish , for the first time , two vital weapons in the fight against FUS proteinopathies , which have been critical in advancing our basic understanding of various other protein misfolding disorders , including Parkinson's disease , Huntington's disease , and TDP-43 proteinopathies [13]–[16] , [39]–[45] . First , we establish a simple yeast model of FUS aggregation and toxicity . Second , we reconstitute FUS misfolding and aggregation using pure protein . These two approaches have served as important foundations for understanding mechanistic aspects of numerous neurodegenerative disorders and have empowered countless advances . We establish that , as for TDP-43 , the RRM and the prion-like domain of FUS are required for aggregation and toxicity in yeast . However , in contrast to TDP-43 , we find that additional determinants within the first RGG domain ( Figure 1A ) are also critical for FUS aggregation and toxicity . Importantly , we demonstrate that pure FUS is inherently aggregation-prone in the absence of other components and this behavior requires determinants in the prion-like domain and first RGG domain of FUS ( Figure 1A ) . Aggregates formed by pure FUS are filamentous and resemble those formed by FUS in degenerating motor neurons of ALS patients . ALS-linked TDP-43 mutations can promote aggregation in vitro with pure proteins and in yeast [15] . By contrast , we find that ALS-linked FUS mutations do not promote aggregation per se . Finally , using two genome-wide screens in yeast , we identified several genes and pathways as potent modifiers of FUS toxicity . Many of the genes that we discovered in the yeast screens have human homologs . Thus , they are likely to provide insight into the specific cellular pathways perturbed by FUS accumulation and may ultimately suggest novel avenues for therapeutic investigation . Surprisingly , almost all of the genetic modifiers had no effect on TDP-43 toxicity in yeast . These key differences between FUS and TDP-43 will help guide the design of therapeutic interventions aimed at mitigating FUS aggregation in disease .
To model aspects of FUS pathology in yeast , we first transformed yeast cells with a high-copy 2 micron ( 2 µ ) plasmid containing human FUS fused to the yellow fluorescent protein ( YFP; Figure 1B ) . Because TDP-43 was toxic to yeast [14] , we placed FUS-YFP expression under the control of a tightly regulated galactose-inducible promoter ( Figure 1B ) to prevent deleterious effects during routine passage . After growing transformants in non-inducing conditions ( raffinose media ) , we induced expression of FUS-YFP in galactose-containing media . Overexpression is a common tool to study the aggregation and toxicity of numerous proteins ranging from alpha-synuclein to TDP-43 [14] , [39] , [45] . It provides a method to elicit protein misfolding by increasing protein concentration and exceeding proteostatic buffers [46] . Moreover , overexpression is likely to yield key information because an established cause of several human neurodegenerative diseases is increased expression of aggregation-prone proteins , such as alpha-synuclein , amyloid precursor protein , and TDP-43 [47]–[49] . Following 4–6 h of induction , we visualized FUS-YFP localization by fluorescence microscopy ( Figure 1C ) . Whereas the control , YFP alone , was localized diffusely throughout the cytoplasm and nucleus , FUS-YFP localized to the cytoplasm where it formed numerous foci ( Figure 1C ) . FUS-YFP showed a similar cytoplasmic localization pattern when expressed from a low-copy galactose-inducible CEN plasmid ( unpublished data ) . The FUS localization pattern was strikingly similar to that of TDP-43 in yeast ( Figure 1C and [14] ) , in terms of size , shape , and quantity of foci in the cytoplasm ( Figure 1C ) . Indeed when co-expressed in the same cell , FUS-YFP and TDP-43-CFP co-localized to the same cytoplasmic foci ( Figure S2 ) . Thus , TDP-43 and FUS inclusions partition to a similar compartment in yeast . Next , we employed a weaker promoter ( glyceraldehyde-3-phosphate dehydrogenase ( GPD ) promoter ) to express FUS at lower levels . Here , FUS-YFP localized to both the nucleus and cytoplasm , where it was diffusely distributed ( Figure S3 ) . Similar results were seen with even weaker yeast promoters ( CYC1 and NOP1; unpublished data ) . Thus , the FUS expression level plays a key role in FUS localization and aggregation in yeast . These data predict that sequence variants or copy number variants in the FUS gene that increase FUS expression might also contribute to ALS , FTLD-FUS , and other FUS proteinopathies . Indeed , a variant in the 3′UTR of the TDP-43 gene increases TDP-43 expression and contributes to FTLD-TDP [47] . Moreover , motor neurons express higher levels of FUS than other tissues , which might render them more vulnerable to FUS misfolding events [50] . In mammalian cells , FUS is normally restricted to the nucleus [51]–[55] . By contrast , in yeast , FUS is mostly localized to the cytoplasm . This difference suggests that the non-canonical FUS nuclear localization signal ( NLS; amino acids 500–526 ) might not be very efficient in yeast . Indeed , in an accompanying manuscript , Ju et al . present data that directly support this hypothesis [56] . Alternatively , FUS might require post-translational modifications to localize to the nucleus , which do not occur in yeast . In an effort to restrict FUS to the nucleus , we fused a strong heterologous NLS ( the SV40 NLS [57] ) to the N-terminus of FUS . The SV40 NLS was sufficient to largely restrict FUS to the nucleus , but some cytoplasmic localization was also observed ( Figure 1F ) . Importantly , restricting FUS to the nucleus eliminated aggregation ( Figure 1F ) . Thus , FUS accumulation in the cytoplasm contributes to its aggregation . Despite the differences between FUS localization in yeast and mammalian cells , we can clearly use the genetically tractable yeast system to model FUS cytoplasmic aggregation , a critical pathological event in ALS and FTLD [17] , [18] . Furthermore , defective nuclear import of FUS might be a key upstream event in ALS [52] . Having established that FUS , like TDP-43 , forms cytoplasmic inclusions when expressed in yeast , we next asked if cytoplasmic aggregation of FUS was toxic . To assess FUS toxicity , we performed spotting assays on galactose media . Expressing FUS-YFP or untagged FUS inhibited growth , whereas YFP had no effect ( Figure 1E ) . Thus , as for TDP-43 , FUS expression in yeast was cytotoxic . Cytotoxicity correlated positively with cytoplasmic aggregation . First , expressing FUS at lower levels from the GPD promoter did not induce cytoplasmic FUS inclusions ( Figure S3 ) and did not confer toxicity ( unpublished data ) . Second , restricting FUS to the nucleus with the SV40 NLS ( Figure 1F ) greatly reduced toxicity ( Figure 1G ) . These data suggest that cytoplasmic FUS aggregation is a critical pathological event in ALS and that neurodegeneration might be caused by a toxic gain of function in the cytoplasm . Importantly , not every human RNA-binding protein aggregates and is toxic when expressed at high levels in yeast . Indeed , we expressed 132 human proteins containing RRMs in yeast . Of these , 35 ( including TDP-43 and FUS ) aggregated and were toxic ( A . D . G . unpublished observations; Figure 1C ) . It will be important to determine whether any of these RRM-bearing proteins , aside from FUS and TDP-43 , are connected to neurodegenerative disease . Moreover , it will be important to define whether common sequence determinants among these 35 RRM-bearing proteins promote aggregation and toxicity . One striking feature of FUS and TDP-43 , as well as at least seven other human RNA-binding proteins that are toxic and aggregate in yeast , is the presence of a prion-like domain ( Figure 1A; A . D . G . unpublished observations ) [33] . We noticed that FUS-YFP cytoplasmic accumulations in yeast are highly dynamic under various growth conditions ( Z . S . , X . D . F . , and A . D . G . , unpublished observations ) . This dynamic behavior was reminiscent of RNA processing bodies ( P-bodies ) and stress granules . P-bodies and stress granules play important roles in regulating the translation , degradation , and localization of mRNAs . The pathways regulating the incorporation of RNAs and RNA-binding proteins into these structures are highly conserved from yeast to human [58] . Under various stress situations , including heat shock and oxidative stress , TDP-43 and FUS localize to these transient subcellular compartments and sites of RNA processing [59]–[62] . Moreover , even under normal conditions some ALS-linked FUS mutants localize to stress granules [51]–[55] . Thus , we tested whether FUS could induce stress granule or P-body formation in yeast and whether FUS localized to these structures . We expressed FUS-YFP or YFP alone in yeast cells harboring RFP- or CFP-tagged stress granule or P-body markers ( Figure 2 ) . To detect stress granules we used Pbp1-CFP and to detect P-bodies we used Dcp2-RFP [63] . Expressing YFP alone did not affect the localization of the P-body or stress granule components , which were diffuse under normal conditions ( Figure 2A , B; unpublished data ) . However , FUS expression induced the formation of P-bodies and stress granules and FUS-YFP colocalized with both of these structures ( Figure 2A , B ) . Thus , FUS localizes to and induces the formation of RNA granules in yeast as it does in human cells [51] , [53]–[55] . These RNA granule assembly pathways are highly conserved from human to yeast . Thus , yeast provides a powerful system to dissect how FUS associates with these structures and to identify genetic and chemical modifiers of this process . To determine sequence features of FUS that were sufficient and necessary for aggregation and toxicity in yeast , we next performed a structure-function analysis . We recently used a similar approach for TDP-43 and determined that the C-terminal prion-like domain was required for aggregation and toxicity [14] . Underscoring the power of this approach , similar results have been reported for the C-terminal domain of TDP-43 in mammalian cells and in animal models [64] , [65] . Moreover , all but one of the recently identified human ALS-linked TDP-43 mutations are located in this same C-terminal region [6] . We generated a series of FUS truncations ( Figure 3A ) . We expressed each of the truncated FUS constructs as YFP-fusions and determined their subcellular localization ( Figure 3B ) and toxicity ( Figure 3D ) . Immunoblotting confirmed that all of the fusion proteins were expressed at comparable levels ( Figure 3C; unpublished data ) . Full-length FUS formed multiple cytoplasmic inclusions in yeast ( Figures 1C , 3B ) . Interestingly , removing the last 25 residues of FUS , which harbor most of the ALS-linked mutations [37] , did not affect aggregation ( Figure 3B , construct 1–501 ) . This result is consistent with a recent report that a similar FUS truncation mutant ( R495X ) is connected with a severe ALS phenotype [51] . A larger C-terminal deletion also had little effect on cytoplasmic aggregation ( Figure 3B , construct 1–453 ) . Thus , C-terminal portions of FUS are not essential for cytoplasmic aggregation . For TDP-43 , the C-terminal prion-like domain is necessary but not sufficient for cytoplasmic aggregation [14] . TDP-43 also requires a portion of RRM2 ( Figure 1A ) [14] . However , for FUS , the N-terminal prion-like domain and the RRM resulted in an entirely nuclear localized protein ( Figure 3B , construct 1–373; Figure S4 ) . Adding back the first RGG domain ( amino acids 371–422 ) was sufficient to restore cytoplasmic aggregation ( Figure 3B , construct 1–422 ) . Thus , in contrast to our findings with TDP-43 [14] , the prion-like domain and the RRM of FUS ( Figure 3B , construct 1–373 ) were insufficient to confer cytoplasmic aggregation . Additional C-terminal determinants within the first RGG domain are required to confer cytoplasmic aggregation ( Figure 3B , construct 1–422 ) . Next , we asked if deletion of portions of the N-terminal prion-like domain of FUS , which spans the QGSY-rich domain and a portion of the Gly-rich domain ( amino acids 1–239 ) ( Figure 1A ) , prevented aggregation . Indeed , the generation of large cytoplasmic inclusions required most of the N-terminal QGSY-rich domain ( Figure 3B , compare constructs 1–501 , 50–526 , 100–526 , and 165–526 ) ( Figure 3A , B ) . Deletion of the entire N-terminal QGSY-rich domain ( construct 165–526 ) yielded mostly diffuse cytoplasmic staining with occasional small foci ( Figure 3A , B ) . However , shorter N-terminal constructs comprising just the N-terminal QGSY-rich domain or this domain plus the Gly-rich domain did not aggregate and were localized in the nucleus ( Figure 3B , constructs 1–168 and 1–269; Figure S4 ) . Thus , the N-terminal prion-like domain of FUS is necessary but not sufficient for aggregation . Rather , FUS requires sequences in both the N-terminal region and the C-terminal region for robust formation of large cytoplasmic inclusions . Accordingly , large N-terminal deletions were diffusely localized within the cytoplasm , with only occasional small cytoplasmic puncta ( Figure 3B , constructs 165–526 , 267–526 , 285–526 , and 368–526 ) . Thus , in distinction to TDP-43 , which requires its C-terminal prion-like domain and a portion of RRM2 ( Figure 1A ) to aggregate in yeast [14] , FUS requires its N-terminal prion-like domain , RRM , and first RGG domain to aggregate in yeast ( Figures 1A , 3A ) . This key difference will have important implications for the design of therapeutic strategies aimed at preventing or reversing aggregation . Our domain mapping experiments in yeast indicate that the first RGG domain of FUS ( amino acids 371–422 ) is important for driving aggregation ( e . g . , Figure 3B , compare constructs 1–373 and 1–422 ) and that sequences in the N-terminal prion-like domain ( amino acids 1–239 ) are also important ( e . g . , Figure 3B , compare constructs 50–526 and 165–526 ) . To test these predictions in mammalian cells , we transfected several of these deletion constructs ( as C-terminal V5 epitope tag fusions ) in COS-7 cells . In contrast to yeast cells , where full-length FUS ( construct 1–526 ) forms cytoplasmic inclusions , and consistent with previous reports in mammalian cells [17] , [18] , [51] , [52] , full-length FUS localized almost exclusively to the nucleus , forming occasional cytoplasmic foci ( Figure 4 ) . This difference between the localization of full-length FUS in yeast ( almost entirely cytoplasmic and forms inclusions ) versus mammalian cells ( almost entirely nuclear and diffuse ) is also seen with TDP-43 ( e . g . , compare [66] and [14] ) and might reflect differences in the efficacy of the FUS and TDP-43 nuclear localization signals in yeast and mammals . Indeed , Ju et al . demonstrate that the FUS NLS ( amino acids 500–526 ) is ineffective in yeast [56] . Consistent with our yeast data , FUS constructs 1–269 and 1–373 localized almost exclusively to the nucleus in a diffuse pattern , although there was more cytoplasmic staining with 1–373 ( Figure 4 ) . These results were surprising since these constructs lack the C-terminal NLS defined in other studies [52] , [54] . However , these results are consistent with those of Kino et al . , who find that FUS 1–278 is localized to the nucleus and FUS 1–360 is localized to the nucleus as well as the cytoplasm [54] . These data suggest that additional determinants of nuclear localization exist in the FUS primary sequence . Indeed , scanning the FUS primary sequence using NLStradumus [67] revealed three NLS sequences in FUS comprising residues 241–251 , 381–395 , and 480–521 . These two additional NLS sequences ( 241–251 and 381–395 ) might help explain why all of the FUS constructs in Figure 4 have some ability to localize to the nucleus . Strikingly , as we observed in yeast , addition of the first RGG domain ( construct 1–422 ) resulted in prominent cytoplasmic FUS aggregation in COS-7 cells ( Figure 4 ) . FUS construct 50–526 aggregated in yeast ( Figure 3B ) and mammalian cells ( Figure 4 ) . However , the morphology of the 50–526 inclusions was distinct from those formed by 1–422 ( one or two large tight inclusions per cell with 1–422 versus numerous amorphous inclusions with 50–526 ) . These data indicate that the domains of FUS required for aggregation in yeast ( especially the first RGG domain ) are also critical for FUS aggregation in mammalian cells . Moreover , these data validate the yeast system as a useful platform for interrogating mechanisms and genetic modifiers ( see below ) of FUS aggregation and toxicity . Having determined the regions of FUS required for aggregation in yeast , we next determined which regions of FUS contributed to toxicity ( Figure 3D ) . As with FUS aggregation , the last 25 amino acids of FUS , where many of the ALS-linked mutations occur [37] , were not required for toxicity ( Figure 3D , construct 1–501 ) . Indeed , 1–501 was slightly more toxic than full-length FUS ( Figure 3D ) . This finding is consistent with the severe ALS phenotype linked to FUS R495X [51] . Similar to TDP-43 , the prion-like domain of FUS was required but not sufficient for toxicity ( Figure 3D , compare constructs 1–526 , 1–168 , 1–269 , and 267–526 ) . As for aggregation , most of the N-terminal prion-like domain of FUS ( amino acids 1–239 ) was needed for toxicity ( Figure 3D , compare constructs 1–501 , 50–526 , and 100–526 ) and larger N-terminal deletions were not toxic ( Figure 3D , compare constructs 165–526 , 267–526 , 285–526 , and 368–526 ) . However , unlike TDP-43 [14] , adding back the RRM to the prion-like domain did not restore toxicity ( Figure 3D , construct 1–373 ) . Rather , for toxicity the RRM and the first RGG domain were required in addition to the prion-like domain ( Figure 3D , compare constructs 1–373 and 1–422 ) . However , 1–422 was not as toxic as full-length FUS , and additional C-terminal sequences were required to confer full toxicity ( Figure 3D , compare constructs 1–422 , 1–453 , and 1–501 ) . These findings are consistent with a pathogenic FUS truncation mutant ( amino acids 1–466 ) connected with sporadic ALS [68] . Next , we tested whether FUS must bind RNA and aggregate to be toxic in yeast . Thus , we mutated conserved phenylalanine residues within the FUS RRM to leucine ( Phe305 , 341 , 359 , 368Leu ) that would disrupt RNA binding [69] . These mutations were sufficient to mitigate toxicity but had no effect on cytoplasmic aggregation ( Figure 3E , F ) . Analogous mutations to the RRMs of TDP-43 disable RNA binding [69] and also mitigate toxicity in yeast [16] . Taken together , these data indicate that the N-terminal prion-like domain , first RGG domain , and RRM ( likely via RNA binding ) of FUS contribute to toxicity . Identifying the specific RNA targets of FUS ( for example , see [70] ) will provide key insights into mechanisms of toxicity associated with FUS aggregation in disease . Overall , compared to TDP-43 , FUS aggregation and toxicity in yeast is a more complex multi-domain process . Importantly , our studies define the prion-like FUS N-terminal domain and first RGG domain as potential targets to prevent or reverse FUS aggregation and toxicity . To determine whether FUS is intrinsically prone to aggregation , we purified bacterially expressed recombinant FUS as a soluble protein under native conditions . However , expression of various constructs including N- and C-terminal His-tagged FUS in various bacterial strains failed to yield soluble protein . The solubility of various proteins , including TDP-43 and polyglutamine , can be enhanced by the addition of a glutathione-S-transferase ( GST ) tag [15] , [41] , [71] . Even so , FUS bearing a C-terminal GST-tag was also insoluble in various bacterial strains . Fortunately , an N-terminal GST-tag allowed FUS to be purified as a soluble protein under native conditions . GST-FUS remained soluble for extended periods and was competent to bind RNA in mobility shift assays ( Figure 5A ) . To study FUS aggregation , we added tobacco etch virus ( TEV ) protease to cleave at a single unique site and specifically remove the N-terminal GST-tag ( Figure 5B ) . This strategy has been utilized successfully to study the aggregation of extremely aggregation-prone proteins , including polyglutamine [41] , [43] . Upon addition of TEV protease , FUS aggregated extremely rapidly ( Figure 5C ) . By contrast , GST-FUS remained predominantly soluble ( Figure 5C ) . Under identical conditions neither GST nor TEV protease aggregated ( Figure 5C ) . Aggregation was dependent on FUS concentration in three ways: at higher FUS concentrations , the maximum amplitude or endpoint of turbidity was increased , the length of lag phase was reduced and the rate of aggregation during assembly phase was accelerated ( Figure 5C ) . Sedimentation analysis revealed that after addition of TEV protease , FUS entered the pellet fraction , whereas GST-FUS remained largely soluble ( Figure 5D ) . Indeed , there was very little FUS in the supernatant fraction at any time , indicating that aggregation occurred rapidly after proteolytic liberation of FUS from GST ( Figure 5D ) . The aggregates formed by FUS did not react with the amyloid-diagnostic dye Thioflavin-T and were SDS-soluble , in contrast to those formed by NM , the prion domain of yeast prion protein Sup35 ( Figure 5E , F ) . Thus , pure FUS forms aggregates that are likely non-amyloid in nature , just like the aggregated species of FUS observed in ALS and FTLD patients [5] , [72] , [73] . The rapid aggregation of FUS occurred without agitation of the reaction ( Figures 5C , 6A ) . Remarkably , under these conditions , even TDP-43 did not aggregate ( Figure 6A ) . TDP-43 requires many hours to aggregate unless the reaction is agitated [15] . Agitation had little effect on the rate of FUS aggregation ( Figure 6A , B ) , indicating that under these conditions FUS aggregation is energetically favorable . Even when the reaction was agitated , TDP-43 aggregation was still considerably slower than FUS aggregation ( Figure 6B ) . In particular , the lag period prior to aggregation was longer for TDP-43 than for FUS ( Figure 6B ) . This extended lag period was not due to different rates of FUS or TDP-43 cleavage by TEV protease , which were extremely similar ( unpublished data ) . Rather , nucleation of aggregation is apparently more rate limiting for TDP-43 than it is for FUS . Collectively , these data suggest that , even in comparison to TDP-43 , FUS is extremely aggregation prone . These data are also in keeping with the higher prion-like domain score of FUS compared to TDP-43 [33] . In vivo , such rapid FUS aggregation is most likely precluded by the proteostasis network [46] . However , FUS likely escapes these safeguards in disease situations where proteostatic buffers may have declined with age or because of environmental triggers . Irrespective of the factors that may elicit FUS aggregation in disease , pure protein assays akin to the one we report here have been powerful tools to dissect the mechanisms underlying the aggregation of various disease-connected proteins , including TDP-43 and polyglutamine [15] , [41] , [43] . Next , we determined how the N- and C-terminal domains of FUS contribute to aggregation of the pure protein . Consistent with observations in yeast ( Figure 3B ) , deletion of the N-terminal prion-like domain of FUS yielded protein ( 267–526 ) that remained soluble over the time course of the assay as determined by turbidity and sedimentation analysis ( Figure 7A , B ) . These data suggest that the prion-like domain of FUS is required for aggregation . Curiously , however , but also consistent with observations in yeast , a protein bearing the prion-like domain and adjacent C-terminal sequences ( 1–373 ) did not aggregate under these conditions ( Figure 7A , B ) . Even at higher concentrations ( 20 µM ) , neither FUS 267–526 nor FUS 1–373 aggregated . Moreover , if the reaction was subsequently agitated at 700 rpm for an additional 60 min neither FUS 267–526 nor FUS 1–373 aggregated . Next , we tested FUS 1–422 , a minimal fragment of FUS able to confer toxicity and aggregation in yeast ( Figure 3B , D ) . FUS 1–422 aggregated with similar kinetics to full-length FUS as determined by sedimentation analysis ( Figure 7B ) . Curiously , however , at these concentrations ( 2 . 5–5 µM ) FUS 1–422 aggregates did yield a signal by turbidity ( Figure 7A ) . Higher concentrations of FUS 1–422 ( 20 µM ) were required to generate aggregates detectable by turbidity ( Figure 7A ) . These concentration differences in the turbidity measurements for full-length FUS and FUS 1–422 suggest that there are large disparities in the sizes of the aggregates formed by these two proteins because turbidity readily detects large but not small aggregates [74]–[76] . A similar finding has been made with PrP , where deletion of the N-terminal domain reduces the formation of larger turbid aggregates , without affecting the formation of smaller aggregates [74] . These data suggest that the C-terminal region , comprising amino acids 423–526 , while dispensable for aggregation per se ( Figure 7B ) , promotes the formation of large macroscopic aggregates of FUS that are detected by turbidity ( Figure 7A ) . Electron microscopy ( EM ) confirmed that pure FUS 1–373 and FUS 267–526 do not form aggregated species in isolation ( Figure 8A , B ) . Rather , these proteins persist as small oligomeric particles ( Figure 8A , B ) . In the absence of TEV protease , both FUS and FUS 1–422 did not aggregate but remained as small oligomeric species ( Figure 8C , D ) . After addition of TEV protease , FUS and FUS 1–422 rapidly populated oligomeric forms , which adopted a pore-like conformation reminiscent of pathological oligomers formed by TDP-43 , α-synuclein , and Aβ42 ( Figure 8E ) [15] , [42] . FUS 1–422 rapidly aggregated in an ordered manner to generate separated filamentous structures ( Figure 8C ) . Likewise , full-length FUS also rapidly formed linear polymers ( Figure 8D ) . In both cases , these filaments were approximately 15–20 nm in diameter and could extend several micrometers in length ( Figure 8C , D ) . Consistent with turbidity measurements , the polymers formed by full-length FUS became tangled and stacked against one another to form extremely large and complex macroscopic networks ( Figure 8D , F ) . FUS 1–422 polymers remained more separated with limited lateral interaction ( Figure 8C , F ) . These ultrastructural observations explain why FUS 1–422 aggregates are more difficult to detect by turbidity . Importantly , the filamentous structures formed by both FUS and FUS 1–422 bear striking resemblance to the FUS aggregates observed in the degenerating motor neurons of ALS patients [21] , [77] . In motor neurons of patients with juvenile ALS , FUS forms filamentous aggregates with a uniform diameter of 15–20 nm , which are often associated with small granules [21] , [77] . The filamentous structures formed by FUS and FUS 1–422 in isolation ( Figure 8C , D , F ) are extremely similar to those observed in spinal motor neurons in Figure 3C of Huang et al . [21] . In vitro , small FUS or FUS 1–422 oligomers are often found clustered up against the filamentous structures ( Figure 8C , D , F ) . These oligomers may correspond to the granular structures observed in association with filamentous FUS aggregates in motor neurons of ALS patients [21] , [77] . In sum , these observations suggest that in isolation FUS is intrinsically capable of forming the aggregated structures observed in motor neurons of ALS patients . Taken together , the biochemical and EM data suggest that FUS aggregation requires multiple domains in both N- and C-terminal regions . Specifically , determinants in the N-terminal prion-like domain ( 1–239 ) and the first C-terminal RGG domain ( 374–422 ) are essential for the formation of filamentous structures . More C-terminal regions ( 423–526 ) are then required for the formation of large macroscopic aggregates detected by turbidity . FUS mutations have been connected with some familial and sporadic ALS cases [37] . We next used the yeast model to test the effects of some of these mutations on FUS aggregation and toxicity ( Figure 9A ) . For TDP-43 , we have used this approach to determine that ALS-linked mutations increase TDP-43 aggregation and toxicity [15] . This increased toxicity of mutant TDP-43 in yeast has been supported by independent studies in mammalian cells and animal models [9] , [78]–[80] . To assess aggregation , we expressed YFP-tagged fusions of WT FUS and 12 different ALS-linked FUS mutants in yeast . These FUS variants were all expressed at similar levels ( Figure 9B ) . Moreover , comparison of the number of proportion of yeast cells with three or more foci revealed that ALS-linked FUS mutations do not promote FUS aggregation in yeast ( Figure 9C , D ) . Indeed , FUS aggregation was slightly reduced in various ALS-linked FUS variants , although this reduction was not statistically significant ( Figure 9C , D ) . Consistent with these observations , the ALS-linked FUS variants—H517Q , R521C , and R521G—aggregated with very similar kinetics to WT in pure protein aggregation assays , although aggregation was slightly retarded in these mutants ( Figure 9E ) . Collectively , these data suggest that this set of ALS-linked FUS mutations , clustered in the extreme C-terminal region of FUS , do not promote FUS aggregation per se . Furthermore , we did not observe any significant difference in toxicity between WT and ALS-linked FUS variants ( Figure 9F ) . These data are in contrast to TDP-43 , where several ALS-linked mutations promote aggregation and toxicity [15] . It seems likely that in disease , these C-terminal ALS-linked FUS mutations promote pathological events that are primarily upstream of aggregation and toxicity . One obvious upstream event is mislocalization to the cytoplasm . Indeed , studies in mammalian cells suggest that ALS-linked FUS mutations can disrupt nuclear import [52] . In yeast , FUS is already localized predominantly to the cytoplasm ( Figures 1C , 9C ) , so in this setting the ALS-linked mutants are no more toxic than WT ( Figure 9C , D , F ) . Thus , even though FUS and TDP-43 are related RNA-binding proteins , the mechanisms by which ALS-linked mutations contribute to disease might be different for each protein [52] . Consequently , different therapeutic strategies might be needed for FUS and TDP-43 proteinopathies . To examine this idea further , we performed two genome-wide screens in yeast to ( 1 ) identify genetic modifiers of FUS toxicity and ( 2 ) determine whether genetic modifiers of FUS toxicity also affected TDP-43 toxicity . Of the many experimental benefits afforded by the yeast system [13] , the chief advantage is the ability to perform high-throughput genetic modifier screens . Therefore , to provide insight into cellular mechanisms underpinning FUS toxicity , we performed two unbiased yeast genetic modifier screens to identify genes that enhance or suppress FUS toxicity . We reasoned that the genes identified by these screens would illuminate cellular pathways perturbed by abnormal FUS accumulation and suggest potential novel targets for therapeutic intervention . Similar approaches have elucidated modifiers of the Parkinson's disease protein α-synuclein [39] , [40] , [45] , [81] , a mutant form of the Huntington's disease protein huntingtin [44] , [45] , and more recently , the ALS protein TDP-43 ( [16]; A . Elden and A . D . G . unpublished ) . In the latter example , the yeast system allowed definition of a common genetic risk factor for ALS in humans [16] . First , we performed a plasmid overexpression screen ( Figure 10A ) . We individually transformed 5 , 500 yeast genes , which comprise the Yeast FLEXGene plasmid overexpresssion library [82] , into a yeast strain harboring an integrated galactose-inducible FUS expression plasmid . We then identified yeast genes that suppressed or enhanced FUS toxicity when overexpressed ( Figure 10B ) . We repeated the screen three independent times and only selected hits that reproduced all three times . Genes from the screen that enhanced FUS toxicity , but also caused toxicity when overexpressed in WT yeast cells , were eliminated because these were unlikely to be specific to FUS . We also eliminated certain genes involved in carbohydrate metabolism or galactose-regulated gene expression because , based on previous screens with this library , we have found that they simply affect expression from the galactose-regulated promoter and are unlikely to relate to FUS biology . Indeed , most of these were also recovered as hits in screens with a galactose-regulated toxic huntingtin , α-syn or TDP-43 ( [16] , [39] , [44] , [45] , [81]; A . Elden and A . D . G . unpublished ) . Finally , we retested 10 random plasmids ( six suppressors and four enhancers ) by transforming them into a fresh yeast strain harboring the integrated FUS expression plasmid and performed spotting assays and all 10 of these were confirmed ( Figure S5 ) . Following the above validation and filtering procedures , we identified 24 genes that suppressed and 10 genes that enhanced FUS toxicity when overexpressed ( Table 1 ) . The largest functional class enriched in the screen included RNA-binding proteins and proteins involved in RNA metabolism ( Figure 10C ) . Thus , RNA metabolic pathways play a key role in FUS pathogenesis . Importantly , of 71 genes from this library that modify α-synuclein toxicity in yeast [39] , [40] , only two ( Cdc4 and Tps3 ) affected FUS toxicity . This lack of overlap underscores the specificity of the screen for FUS biology and pathobiology . Moreover , this specificity indicates that the screen does not simply identify generic cellular responses to misfolded proteins . Even more remarkably , out of the 40 yeast genes that we have found to modify TDP-43 toxicity when overexpressed ( [16] and A . Elden and A . D . G . unpublished observations ) , only two ( Fmp48 and Tis11 ) affected FUS toxicity . Thus , despite being similar RNA-binding proteins , the mechanisms by which FUS and TDP-43 contribute to disease are likely to be very different . Several of the yeast genes that modified FUS toxicity have human homologs . Thus , pathways involved in FUS toxicity in yeast are likely conserved to man . Interestingly , FUS has recently been shown to co-localize with stress granules in transfected cells [51] , [52] . Furthermore , cytoplasmic FUS-positive inclusions in ALS and FTLD-U patients contain stress granule markers [51] , [52] . Stress granules and P-bodies are transient cytoplasmic structures containing RNAs and RNA binding proteins , including translation initiation factors and the polyA-binding protein ( PABP-1 ) , which are sites where cells sequester mRNAs , during situations of stress , to inhibit translation initiation [83] . Notably , we identified two translation initiation factors ( Tif2 and Tif3 ) and Pab1 , the yeast homolog of human PABP-1 , which is involved in stress granule assembly in yeast , as suppressors of FUS toxicity ( Table 1 ) . Thus , in addition to being markers of FUS-positive inclusions in disease , stress granule components might play an important role in mediating FUS toxicity . Approaches aimed at manipulating stress granule assembly might be an effective therapeutic approach . As an initial step to extend our findings from yeast to mammalian cells , we selected genes from our overexpression screen for further analysis in a mammalian cell culture FUS toxicity model . We tested two distinct suppressor genes , FBXW7 and EIF4A1 , which are the human homologs of yeast Cdc4 and Tif2 , respectively ( Table 1 ) . We transfected HEK293T cells with WT FUS or two ALS-linked FUS mutants , R521C and R521H . The FUS mutants were more toxic than WT FUS , which only slightly reduced viability ( Figure 10D ) . Co-transfection with FBXW7 or EIF4A1 suppressed toxicity of WT FUS as well as the ALS-linked FUS mutants ( Figure 10D ) . Similar results were observed in COS-7 cells ( unpublished data ) . The FUS toxicity modifier genes and pathways identified in our yeast screens will have to be validated in neuronal cells and eventually animal models . However , the ability of FBXW7 and EIF4A1 to suppress toxicity in human cells , which are separated from yeast by ∼1 billion years of evolution , provides evidence that highly conserved genetic interactions involving FUS , discovered in yeast , can be highly relevant to mammalian cells . To complement the yeast overexpression screen , we also performed a deletion screen . The yeast genome contains ∼6 , 000 yeast genes and ∼4 , 850 of these are non-essential [84] , [85] . We used synthetic genetic array ( SGA ) analysis [86] , [87] to introduce a FUS expression plasmid into each non-essential yeast deletion strain by mating ( Figure 11A ) . Following sporulation , we selectively germinated meiotic progeny containing both the FUS plasmid and the gene deletion . We compared growth of each strain on glucose ( FUS expression “off” ) to that on galactose ( FUS expression “on” ) . We identified some yeast deletions that enhanced FUS toxicity ( aggravating interaction ) and others that suppressed toxicity ( alleviating interaction ) ( Figure 11B ) . As for the overexpression screen , we repeated the deletion screen three independent times and only selected hits that reproduced all three times and filtered out deletion strains that grew poorly on galactose-containing media , even in the absence of FUS ( using published data on yeast deletion strain fitness on galactose and in house measurements of the yeast deletion collection grown on galactose ) . Genetic interactions were further confirmed by random spore analysis and the integrity of the deletions verified by sequencing the deletion specific bar codes . We also independently confirmed six random hits by remaking the deletions , confirming the deletions by PCR , and then transforming those deletion strains with the FUS expression plasmid and performing spotting assays . We indentified 36 deletions that suppressed FUS toxicity and 24 that enhanced toxicity ( Table 2 ) . Deletions of yeast genes involved in RNA metabolic processes , ribosome biogenesis , and cellular stress responses were enriched as hits ( Figure 11C ) . Many of these genes have human homologs ( Table 2 ) . One interesting deletion suppressor was Sse1 , a member of the Hsp70 chaperone family , which promotes Sup35 prion formation [88] and might also promote FUS aggregation . Two other notable deletion suppressors were Pub1 ( TIAL1 in human ) and Lsm7 ( LSM7 in human ) , components of stress granules and P-bodies , respectively . Furthermore , TIAL1 ( and Pub1 ) contains a prion-like domain [32] , [89] , which can template the aggregation of the polyQ protein huntingtin [90] , suggesting that FUS aggregation and cytoplasmic sequestration might be templated by similar mechanisms [24] , [91] . Again , as for the plasmid overexpression screen , genetic manipulations that affect stress granule components are sufficient to mitigate FUS toxicity . And , as for the overexpression screen , there was little overlap between the FUS and TDP-43 modifier genes . In a broader sense , the collection of deletion suppressors of FUS toxicity is an interesting class , because these could represent attractive therapeutic targets for small molecule inhibitors or RNA interference . Taken together , the genetic modifiers uncovered by the yeast overexpression and deletion screens provide insight into the pathways affected by FUS . The way is now open to develop therapeutic strategies that target these pathways .
We have established a pure protein aggregation assay and a yeast model to gain insight into how FUS contributes to disease pathogenesis . We have recently used a similar approach to define mechanisms underpinning TDP-43 aggregation and toxicity [14] , as well as the pathogenic mechanism of ALS-linked TDP-43 mutants [15] . Using the yeast system we have also identified potent modifiers of TDP-43 toxicity [16] . One such modifier is ataxin 2 , which can harbor intermediate-length polyQ expansions that are associated with increased risk for ALS in humans [16] . Like TDP-43 , we find that , in isolation , FUS is an intrinsically aggregation-prone protein . FUS rapidly assembles into pore-like oligomeric species and filamentous structures that closely resemble the ultrastructure of FUS aggregates in degenerating motor neurons of ALS patients . Thus , all the information needed to assemble these structures is encoded in the primary sequence of FUS . Like TDP-43 , expression of FUS in yeast results in cytoplasmic FUS aggregation , colocalization of these inclusions with stress granules and toxicity , modeling key features seen in human disease [17] , [18] , [21] , [23] , [52] . In further similarity to TDP-43 , disabling the RNA binding activity of FUS reduced toxicity . Thus , we propose that the misfolded forms of FUS likely cause toxicity by binding to and sequestering essential RNAs or perhaps by interfering with the normal shuttling , stability , or metabolism of RNA . Importantly , FUS immunoreactive cytoplasmic inclusions now appear to characterize ALS and FTLD broadly , not only rare cases linked to FUS mutations [21] , [23] , [92] . Together these advances make it clear that FUS is a key aggregated protein in ALS , just as α-synuclein is in Parkinson's disease and huntingtin is in Huntington's disease [33] . Despite these similarities , we have uncovered key differences in the regions of the proteins that dictate aggregation and toxicity . For TDP-43 , pure protein data and results from yeast and other model systems suggest that the C-terminal prion-like domain ( Figure 1A ) [33] plays a major role in driving aggregation [14] , [15] , [66] , [93] . For FUS , we find that the N-terminal region , containing a predicted prion-like domain ( Figure 1A ) [33] , is also important for aggregation in vitro and for aggregation and toxicity in yeast cells . However , C-terminal regions in FUS , particularly the first RGG domain , are also critical . Intriguingly , the first RGG domain also contains a short region ( amino acids 391–407 ) that is detected by an algorithm designed to isolate prion-like domains [32] , [33] but does not quite reach significance ( Figure S1 ) . The requirement for two specific , disparate portions of FUS for the ordered formation of filamentous structures raises the possibility that communication between the N-terminal prion-like domain ( amino acids 1–239 ) and first RGG domain ( amino acids 374–422 ) might mediate a self-organizing assembly process . This process might even involve an intermolecular domain swap: a common mechanism that usually involves domains at the N- and C-terminal ends of proteins and can promote the polymerization of filamentous structures in various designed and natural proteins [94]–[96] . Thus , strategies aimed at targeting either the appropriate N- or C-terminal portions of FUS could be effective at mitigating FUS aggregation in disease . Indeed , our in vitro and yeast models could open up new therapeutic avenues and provide the basic screening system to isolate specific molecules able to antagonize and reverse FUS aggregation and toxicity . With regard to toxicity , the minimal toxic FUS fragment comprises the N-terminal prion-like domain , RRM , and the first RGG domain ( 1–422 ) . These findings contrast with TDP-43 , where the prion-like domain plus RRM2 are sufficient to drive aggregation and toxicity [14] . Indeed , a proteolytic fragment corresponding to these portions of TDP-43 is a pathogenic signature of ALS and FTLD-TDP [3] . By contrast , a similar pathogenic FUS fragment has not been identified in ALS or FTLD-FUS patients , which likely reflects the fact that the equivalent regions of FUS ( 1–373 ) are insufficient for aggregation and toxicity . Mutations in the C-terminal domains of FUS and TDP-43 have both been linked to ALS [6] , [37] . Interestingly , whereas some ALS-linked mutations in TDP-43 can increase stability , aggregation , cytoplasmic accumulation , and toxicity in yeast , mammalian cells , and animal models [15] , [16] , [78] , [80] , [97] , the mechanisms by which FUS mutations contribute to disease appear to be distinct . Our results in yeast and with pure protein show that C-terminal FUS mutations do not promote aggregation per se . Instead of enhancing aggregation , these mutations , especially those in the extreme C-terminal region of the protein ( amino acids 502–526 ) , disrupt a NLS , leading to increased cytoplasmic accumulation of FUS [52] . Interestingly , the severity of the effects of the mutations on FUS localization in cells correlate well with age of onset of ALS in humans , with stronger mutations resulting in earlier disease onset and more cytoplasmic FUS accumulation [52] . These results suggest distinct mechanisms by which ALS-linked FUS and TDP-43 mutations contribute to disease . Despite these differences , both TDP-43 and FUS have been shown to re-localize to stress granules and P-bodies , transient sites of RNA processing that assemble during cellular stress or injury and are conserved from yeast to man [59] , [60] , [62] , [98] . Both TDP-43 and FUS have been purified in a complex with one another and with various components of the RNA processing machinery , including stress granules and P-bodies [62] , [97] . Moreover , stress granule markers , including PABP-1 , are present in disease-associated cytoplasmic FUS accumulations [52] . ALS-linked FUS mutants appear more prone to entering stress granules [51] . However , it remains unclear whether stress granule assembly contributes to FUS toxicity or is simply a downstream consequence of cellular stress associated with degeneration . Our identification of several key P-body and stress granule components as potent genetic modifiers of FUS toxicity suggests a mechanistic connection that , if validated in animal models , represents a potentially tractable new therapeutic angle . We also note that for the majority of overexpression or deletion suppressors that we have examined so far , we do not see a major difference in FUS aggregation . This suggests that these genes act downstream or in parallel to FUS aggregation . Alternatively , these modifiers may affect FUS aggregation ( e . g . , composition or dynamics of FUS inclusions ) in subtle ways that we have so far not been able to visualize . Curiously , there was a conspicuous lack of overlap between genetic modifiers of FUS toxicity and TDP-43 toxicity . These genetic data suggest two interesting possibilities . On one hand , targeting the modifiers in common between TDP-43 and FUS might have broad therapeutic utility for ALS . On the other hand , defining the key differences between FUS and TDP-43 pathogenic mechanisms will empower a more accurate understanding of how these seemingly similar proteins might contribute to disease in different ways . What is the connection between TDP-43 , FUS , and ALS ? Does each protein contribute separately to the disease , or do they share a common disease pathway ? The lack of overlap in genetic modifiers suggests that the precise mechanism of TDP-43 and FUS toxicity may be subtly different . Moreover , initial reports suggested FUS cytoplasmic accumulations were specific to rare cases of ALS , owing to FUS mutations , and that these inclusions were devoid of TDP-43 aggregates [17] . However , in one study , using optimized antigen-unmasking methods , FUS cytoplasmic immunoreactivity has recently been detected broadly in sporadic and familial ALS , including cases with TDP-43 aggregates , as well as cases without FUS mutations [92] . Further , FUS and TDP-43 have been found to physically associate in a complex [97] , indicating that both TDP-43 and FUS , even in the WT state , likely contribute broadly to ALS pathogenesis . Therefore , defining mechanisms by which WT versions of these proteins are toxic to cells , as we report here for FUS and in previous studies for TDP-43 [14]–[16] , will likely be informative to not only rare familial cases but to the much more common sporadic forms as well . The discovery of RNA-binding proteins TDP-43 and FUS in ALS has re-invigorated the focus on RNA processing pathways in ALS [5] , [37] , [99] . Our identification of potent genetic modifiers of FUS toxicity in yeast , including a large number of conserved RNA metabolism genes , as well as key stress granule components , will provide a toehold for future studies aimed at elucidating the mechanisms by which FUS interfaces with these RNA processing pathways in disease . However , our study also suggests caution in assuming , based on sequence and structural similarity , that both TDP-43 and FUS contribute to disease via the same or similar mechanisms [38] . While there are clear similarities between the two proteins , there are also important differences , which we have defined here . Furthermore , the fact that genetic modifiers uncovered in screens for TDP-43 and FUS proteotoxicty are surprisingly distinct argues further that there are likely different underlying pathogenic mechanisms for FUS and TDP-43 proteinopathies . This conceptual framework we have established will aid the development of novel therapeutic approaches .
Yeast cells were grown in rich media ( YPD ) or in synthetic media lacking uracil and containing 2% glucose ( SD/-Ura ) , raffinose ( SRaf/-Ura ) , or galactose ( SGal/-Ura ) . A FUS Gateway entry clone was obtained from Invitrogen , containing full-length human FUS in the vector pDONR221 . A Gateway LR reaction was used to shuttle FUS into Gateway-compatible yeast expression vectors ( pAG vectors , [100] , http://www . addgene . org/yeast_gateway ) . To generate C-terminally YFP-tagged FUS constructs , a two-step PCR protocol was used to amplify FUS ( or truncated versions ) without a stop codon and incorporate the Gateway attB1 and attB2 sites along with a Kozak consensus sequence . Resulting PCR products were shuttled into pDONR221 using a Gateway BR reaction . The entry clones ( FUSnostop ) were then used in LR reactions with pAG426Gal-ccdB-YFP to generate the 2 micron FUS-YFP fusion constructs and pAG416Gal-ccdB-YFP to generate the CEN FUS-YFP constructs . Primer sequences are available upon request . To generate the integrating FUS construct , the FUS entry clone was used in an LR reaction with pAG303Gal-ccdB . Expression constructs for TDP-43 have been described previously [14] , [15] . ALS-linked point mutations , based on [38] , were introduced into FUS using the QuickChange Site-Directed Mutagenesis Kit ( Agilent ) according to the manufacturer's instructions . Mutations were verified by DNA sequencing . To disable FUS RNA binding , we mutated four conserved phenylalanine residues ( aa 305 , 341 , 359 , 368 ) within the FUS RNA recognition motif ( RRM ) to leucine . Two micron plasmid constructs ( e . g . , pAG426Gal-FUS-YFP ) were transformed into BY4741 ( MATa his3 leu2 met15 ura3 ) . The FUS integrating strain was generated by linearizing pAG303Gal-FUS by Nhe I restriction digest , followed by transformation into the w303 strain ( MATa can1-100 , his3-11 , 15 , leu2-3 , 112 , trp1-1 , ura3-1 , ade2-1 ) . To introduce the SV40 NLS to the N-terminus of FUS , we used PCR , incorporating DNA sequences encoding the SV40 NLS ( PPKKKRKV ) , optimized for yeast translation ( CCA CCA AAA AAA AAA AGA AAA GTT ) into the forward primer , following a start codon ( ATG ) and in frame with FUS . We verified the construct by DNA sequencing . Yeast procedures were performed according to standard protocols [101] . We used the PEG/lithium acetate method to transform yeast with plasmid DNA [102] . For spotting assays , yeast cells were grown overnight at 30°C in liquid media containing raffinose ( SRaf/-Ura ) until they reached log or mid-log phase . Cultures were then normalized for OD600 , serially diluted and spotted onto synthetic solid media containing glucose or galactose lacking uracil and were grown at 30°C for 2–3 d . Yeast lysates were subjected to SDS/PAGE ( 4%–12% gradient , Invitrogen ) and transferred to a PVDF membrane ( Invitrogen ) . Membranes were blocked with 5% nonfat dry milk in PBS for 1 h at room temperature . Primary antibody incubations were performed overnight at 4°C or at room temperature for 1–2 h . After washing with PBS , membranes were incubated with a horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature , followed by washing in PBS+0 . 1% Tween 20 ( PBST ) . Proteins were detected with Immobilon Western HRP Chemiluminescent Substrate ( Millipore ) . Primary antibody dilutions were as follows: anti-GFP monoclonal antibody ( Roche ) , 1∶5 , 000; Phosphoglycerate Kinase 1 ( PGK1 ) antibody ( Invitrogen ) , 1∶500; glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) , 1∶5 , 000; FUS rabbit polyclonal antibody ( Bethyl ) , 1∶10 , 000 . HRP-conjugated anti-mouse and anti-rabbit secondary antibodies were used at 1∶5 , 000 . For fluorescence microscopy experiments , single colony isolates of the yeast strains were grown to mid-log phase in SRaf/-Ura media at 30°C . Cultures were spun down and resuspended in the same volume of SGal/-Ura to induce expression of the FUS constructs . Cultures were induced with galactose for 4–6 h before being stained with DAPI to visualize nuclei and processed for microscopy . Images were obtained using an Olympus IX70 inverted microscope and a Photometrics CoolSnap HQ 12-bit CCD camera . Z-stacks of several fields were collected for each strain . The images were deblurred using a nearest neighbor algorithm in the Deltavision Softworx software and representative cells were chosen for figures . To assess differences in aggregation between wild-type and mutant FUS , yeast cultures were grown , induced , and processed as described above after having normalized all yeast cultures to OD600nm = 0 . 2 prior to galactose induction . After 6 h of induction , the identities of the samples were blinded to the observer before being examined . Several fields of cells were randomly chosen using the DAPI filter to prevent any bias towards populations of cells with increased amounts of aggregation in addition to obtaining the total number of cells in any given field . At least 200 cells per sample were counted for each replicate . Only cells with greater than three foci under the YFP channel were considered as cells with aggregating FUS . Plasmids of 5 , 500 full-length yeast ORFs ( Yeast FLEXGene collection , [82] ) were dried in individual wells of 96-well microtiter plates and transformed into a strain expressing FUS integrated at the HIS3 locus . A standard lithium acetate transformation protocol was modified for automation and used by employing a BIOROBOT Rapidplate 96-well pipettor ( Qiagen ) . The transformants were grown in synthetic deficient media lacking uracil ( SD-Ura ) with glucose . 48 h later , the cultures were inoculated into fresh SRaf-Ura media and allowed to reach stationary phase . Then the cells were spotted on to SD-Ura + glucose and SD-Ura + galactose agar plates . Suppressors and enhancers of FUS were identified on galactose plates after 2–3 d of growth at 30°C . The entire screen was repeated three times and only hits that reproduced all three times were selected for further validation . Toxicity enhancers were further tested in WT yeast cells to eliminate genes that were simply toxic when overexpressed . Immunoblotting was performed to test all modifiers for their effect on FUS expression . This screen was performed as described in [86] , [87] , [103] , with some modifications , using a Singer RoToR HDA ( Singer Instruments , Somerset , UK ) . The galactose-inducible FUS expression construct ( pAG416Gal-FUS-YFP ) was introduced into MATα strain Y7092 ( gift from C . Boone ) to generate the query strain . This query strain was mated to the yeast haploid deletion collection of non-essential genes ( MATa , each gene deleted with KanMX cassette ( confers resistance to G418 ) ) . Haploid mutants harboring the FUS expression plasmid were grown in the presence of glucose ( FUS expression “off” ) or galactose ( FUS expression “on” ) . Following growth at 30°C for 2 d , plates were photographed and colony sizes measured by ImageJ image analysis software , based on [104] . The entire screen was repeated three times and only hits that reproduced all three times were selected for further validation by random spore analysis on DNA sequencing of deletion strain bar codes . Deletion strains that grew poorly on galactose were eliminated based on published data on deletion strain fitness on galactose as well as in house measurements using the yeast deletion collection . FUS and FUS deletion mutants were expressed and purified from Escherichia coli as GST-tagged proteins . FUS constructs were generated in GV13 to yield a TEV protease cleavable GST-FUS protein , GST-TEV-FUS , and overexpressed in E . coli BL21 DE3 cells ( Agilent ) . Protein was purified over a glutathione-sepharose column ( GE ) according to the manufacturer's instructions . Proteins were eluted from the glutathione sepharose with 50 mM Tris-HCl pH 8 , 200 mM trehalose , and 20 mM glutathione . After purification , proteins were concentrated to 10 µM or greater using Amicon Ultra-4 centrifugal filter units ( 10 kDa molecular weight cut-off; Millipore ) . Protein was then centrifuged for 30 min at 16 , 100 g to remove any aggregated material . After centrifugation , the protein concentration was determined by Bradford assay ( Bio-Rad ) and the proteins were used immediately for aggregation reactions . GST-TEV-TDP-43 was purified as described [15] . RNA-binding assays were performed as described [105] . Briefly , FUS RNA probe was transcribed by T7 polymerase from DNA template ( 5′-GTAATACGACTCACTATAGGGGAAAATTAATGTGTGTGTGTGGAAAATT-3′ ) with 32P-labeled UTP . Probes were gel-purified and adjusted to 104 c . p . m . /µl specific activity . Standard binding reactions were carried out in 10 µl , with a final concentration of 4 mM MgCl2 , 25 mM phosphocreatine , 1 . 25 mM ATP , 1 . 3% polyvinyl alcohol , 25 ng of yeast tRNA , 0 . 8 mg of BSA , 1 mM DTT , 0 . 1 µl Rnasin ( Promega , 40 U/ml ) , 75 mM KCl , 10 mM Tris , pH 7 . 5 , 0 . 1 mM EDTA , 10% glycerol , and 0 . 15 µM to 5 µM GST-FUS or GST . Binding reactions were incubated for 20 min at 30°C with 32P-labeled probe . After binding , heparin was added to a final concentration of 0 . 5 µg/ml; reactions were analyzed on a 4 . 5% native gel ( Acrylamide/Bis 29:1 , BioRad ) . Aggregation was initiated by the addition of TEV protease ( Invitrogen ) to GST-TEV-FUS ( 2 . 5–5 µM ) in assembly buffer ( AB ) : 100 mM TrisHCl pH 8 , 200 mM trehalose , 0 . 5 mM EDTA , and 20 mM glutathione . Aggregation reactions were incubated at 22°C for 0–90 min with or without agitation at 700 rpm in an Eppendorf Thermomixer . No aggregation occurred unless TEV protease was added to separate GST from FUS or TDP-43 . Turbidity was used to assess aggregation by measuring absorbance at 395 nm . For sedimentation analysis , reactions were centrifuged at 16 , 100 g for 10 min at 25°C . Supernatant and pellet fractions were then resolved by SDS-PAGE and stained with Coomassie Brilliant Blue , and the amount in either fraction determined by densitometry in comparison to known quantities of FUS . For electron microscopy ( EM ) of in vitro aggregation reactions , protein samples ( 20 µl of a 2 . 5 µM solution ) were adsorbed onto glow-discharged 300-mesh Formvar/carboncoated copper grid ( Electron Microscopy Sciences ) and stained with 2% ( w/v ) aqueous uranyl acetate . Excess liquid was removed , and grids were allowed to air dry . Samples were viewed using a JEOL 1010 transmission electron microscope . We used fluorescent markers of P-bodies and stress granules and live cell imaging to monitor stress granule and P-body formation in yeast , based on standard protocols [63] . First , we transformed yeast strain BY4741 with pAG423GAL-FUS-YFP . This strain was then transformed with plasmids encoding P-body markers ( Lsm1-mCherry , LEU2 or Dcp2-RFP , LEU2 ) or stress granule markers ( Pub1-RFP , URA3 or CFP-Pbp1 , URA3 ) separately . Transformants were grown overnight to mid-log phase in raffinose-containing media . To induce expression of FUS-YFP , galactose was added to 2% and cells were incubated at 30°C for 4 h and then processed for microscopy . We used a spinning disk confocal microscope to monitor the YFP , CFP , and RFP signals in live cells . For each channel , 60 z-sections were acquired at 0 . 1 µm increments at 23°C . Figures display the maximum projection of each channel . HEK293T cells were plated in 96-well format and transfected with FuGene ( Roche ) according to the manufacturer's instructions . 72 h post-transfection , MTT ( 3- ( 4 , 5-Dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide ) ( Sigma ) was added to each well and incubated for 3 h at 37°C . Acidic Isoproponal ( 40 mM HCl ) was then added to each well to solubilize the blue formazan crystals . Absorbance of each well was read with a Tecan Safire II plate reader using 570 nm for absorbance and 630 nm as a reference wavelength . Absorbance measurements were normalized to the absorbance of untransfected cells and used to calculate a percent viability for each condition . | Many human neurodegenerative diseases are associated with the abnormal accumulation of protein aggregates in the neurons of affected individuals . Amyotrophic lateral sclerosis ( ALS ) , also known as Lou Gehrig's disease , is a fatal human neurodegenerative disease caused primarily by a loss of motor neurons . Recently , mutations in a gene called fused in sarcoma ( FUS ) were identified in some ALS patients . The basic mechanisms by which FUS contributes to ALS are unknown . We have addressed this question using protein biochemistry and the genetically tractable yeast Saccharomyces cerevisiae . We defined the regions of biochemically pure FUS protein that contribute to its aggregation and toxic properties . We then used genome-wide screens in yeast to identify genes and cellular pathways involved in the toxicity of FUS . Many of the FUS toxicity modifier genes that we identified in yeast have clear homologs in humans , suggesting that these might also be relevant for the human disease . Together , our studies provide novel insight into the basic mechanisms associated with FUS aggregation and toxicity . Moreover , our findings open new avenues that could be explored for therapeutic intervention . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neurological",
"disorders/movement",
"disorders",
"neurological",
"disorders",
"genetics",
"and",
"genomics/disease",
"models",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration"
] | 2011 | Molecular Determinants and Genetic Modifiers of Aggregation and
Toxicity for the ALS Disease Protein FUS/TLS |
A major challenge in developing vaccines for emerging pathogens is their continued evolution and ability to escape human immunity . Therefore , an important goal of vaccine research is to advance vaccine candidates with sufficient breadth to respond to new outbreaks of previously undetected viruses . Ebolavirus ( EBOV ) vaccines have demonstrated protection against EBOV infection in nonhuman primates ( NHP ) and show promise in human clinical trials but immune protection occurs only with vaccines whose antigens are matched to the infectious challenge species . A 2007 hemorrhagic fever outbreak in Uganda demonstrated the existence of a new EBOV species , Bundibugyo ( BEBOV ) , that differed from viruses covered by current vaccine candidates by up to 43% in genome sequence . To address the question of whether cross-protective immunity can be generated against this novel species , cynomolgus macaques were immunized with DNA/rAd5 vaccines expressing ZEBOV and SEBOV glycoprotein ( GP ) prior to lethal challenge with BEBOV . Vaccinated subjects developed robust , antigen-specific humoral and cellular immune responses against the GP from ZEBOV as well as cellular immunity against BEBOV GP , and immunized macaques were uniformly protected against lethal challenge with BEBOV . This report provides the first demonstration of vaccine-induced protective immunity against challenge with a heterologous EBOV species , and shows that Ebola vaccines capable of eliciting potent cellular immunity may provide the best strategy for eliciting cross-protection against newly emerging heterologous EBOV species .
The Ebolavirus genus of the family Filoviridae was thought previously to consist of four species , ZEBOV , SEBOV , Reston ( REBOV ) , and Cote d'Ivoire ( CIEBOV ) [1] . Of these , ZEBOV and SEBOV have been associated with the majority of Ebola virus hemorrhagic fever ( EHF ) cases in humans [2] . Within the last decade , the frequency of EBOV outbreaks in Africa has increased , probably due to human encroachment on the natural habitat of animal reservoir ( s ) and/or improved surveillance [3] . Due to the aggressive nature of EHF symptoms , the rapid spread of infection to other persons in close contact with the infected individual , resultant high mortality rate and threat of bioterrorism , vaccine development against EBOV virus is a high priority . EHF vaccines based on recombinant adenovirus serotype 5 ( rAd5 ) vectors encoding the ZEBOV and SEBOV envelope glycoproteins , GP ( Z ) and GP ( S/G ) , respectively , have shown protective efficacy in NHP [4] , [5] , [6] and hold promise as vaccine candidates for human use [7] . In addition to rAd vaccines , other viral-vectored and virus-like particle ( VLP ) vaccines have exhibited protective efficacy against EBOV infection in NHP [8] , [9] , [10] . Though each of these vaccines generates potent immune responses in NHP , protection is achieved only when the vaccine immunogen and the EBOV species used for infectious challenge are matched , and data show a lack of cross protection against antigens not contained in the vaccine [8] , suggesting that existing vaccines may not provide coverage against newly emerging EBOV species . An outbreak of HF in Western Uganda in late 2007 led to the identification of a fifth species in the genus Ebolavirus [11] . Complete genome sequence comparison of all EBOV species revealed that the virus from Western Uganda , the Bundibugyo species , differed from the previously characterized four EBOV species by 32–42% , as is characteristic for divergence between other members in the genus . Current human vaccine candidates encode GP from SEBOV and ZEBOV , whose sequences differ from BEBOV by 38–47% at the amino acid level . The lack of cross protection of existing vaccines against heterologous species with sequence divergence in the same range suggests that vaccines currently in development will not protect against emerging Ebola viruses . We have shown previously that a prime-boost vaccine strategy priming with DNA vectors and using rAd vectors to provide the boost generates broad immune responses across both T- and B-cell immune compartments [6] . This immunization regimen has been demonstrated to generate antigen-specific immune responses at least one log higher than those observed with either DNA or rAd alone [12] . Therefore , we hypothesized that a DNA prime/rAd5 EBOV vaccine strategy would be the most likely candidate to induce cross-protection against BEBOV . We demonstrate herein that potent responses induced by prime-boost vaccination can provide immune protection against newly emerging EBOV species and show for the first time vaccine-induced species cross-protection against EBOV infection .
It has been demonstrated previously that NHP immunized with a vaccine consisting of EBOV GP DNA followed by boosting with rAd5 GP were uniformly protected when challenged with a lethal dose of wild-type ZEBOV , Mayinga strain [6] . Four cynomolgus macaques were injected at 4–6 week intervals with GP ( Z ) and GP ( S/G ) DNA , followed by a rest period , and boosted after one year with rAd5 vectors containing the EBOV matched insert according to the schedule depicted in Figure 1A . Although sequence divergence between genes coding for BEBOV GP and the inserts contained within the previously used vaccine is substantial , homology is displayed within the N- and C-terminal regions of GP that contain structural elements critical for virus replication [13] . This genetic relatedness between species was the basis for selection of vaccine inserts with the goal of broad coverage against multiple species . Phylogenetic analysis demonstrates that ZEBOV shares genetic ancestry with CIEBOV and BEBOV , while SEBOV is closest to REBOV ( Figure 1B ) . To assess whether ZEBOV and SEBOV gene inserts were likely to provide cross protection against heterologous infectious challenge with BEBOV , GP antigen-specific immune responses in humoral and T cell compartments were assessed by ELISA and intracellular cytokine staining ( ICS ) , respectively , three weeks after delivery of the rAd5-GP ( Z ) boost immunization . Studies performed previously have shown an absence of neutralizing antibody in vaccinated macaques and a lack of correlation with protection from infection [14] , [15] . In contrast , there is a strong association between GP-specific ELISA IgG titers in serum or plasma of immunized animals and protection from EBOV infection in NHP [6] , and subsequent analysis has illustrated that vaccine-induced ELISA titers correlate with protection by rAd5 based vaccines [16] . Therefore , to assess DNA/rAd vaccine immunogenicity in the current study , anti-GP ELISA IgG responses specific for the ZEBOV vaccine insert were measured at the end of the rest period following DNA immunization ( pre-boost ) and compared to ELISA IgG titers after boosting with rAd5-GP ( post-boost ) ( Figure 2A ) . DNA priming alone induced modest plasma antibody titers , averaging an effective concentration ( EC90 ) of 1/900 for all subjects in the vaccine group . Subsequent immunization with rAd5-GP boosted plasma titers by at least an order of magnitude in all subjects ( p = 0 . 02 , pre-boost vs . post-boost titers ) and by two logs in subject V2 . These data confirm the potency of the prime boost vaccine regimen and demonstrate that significant boosting of DNA-primed humoral immunity can be achieved even one year after the final priming immunization . As an initial assessment of potential vaccine-induced cross-species immunity against BEBOV , subject plasma samples were evaluated using the same ELISA format described above except the capture antigen used was BEBOV GP . Comparison of subject antibody responses against BEBOV and ZEBOV ( Figure 2B ) shows that the ZEBOV DNA/rAd vaccine did not generate antibodies cross-reactive with BEBOV GP . Anti-BEBOV reactivity for all four vaccine subjects overlapped the average background level of antigen binding displayed by samples from unvaccinated control subjects ( n = 4 ) . The absence of cross-specific antibody reactivity suggests that immunoglobulins elicited by the DNA/rAd vaccine were directed against linear amino acid sequences not contained within BEBOV GP or against conformational epitopes dependent on protein tertiary structure . In earlier work , DNA/rAd5 vaccine-induced EBOV cellular immunity was assessed by measuring in vitro antigen-activated cell proliferation in PBMC obtained from immunized subjects . While proliferation assays provide a useful measure of T-cell immunity , important effector cell activity , especially within the CD8 T-cell compartment , may not be captured in these measurements [17] . Therefore , we evaluated PBMC from immunized macaques using intracellular cytokine staining to assess memory and effector CD4+ and CD8+ T-cell functions . PBMC samples collected from vaccinated animals four weeks after the rAd5 GP vaccine boost were isolated by density gradient centrifugation and stimulated with peptides spanning the ZEBOV or BEBOV GP reading frame . Intracellular expression of TNFα , IFNγ , and IL-2 induced in the CD8+ and CD4+ memory T cell subsets was evaluated in PBMC samples and quantified after gating on CD95 and CD45RA memory markers ( Figure 3A ) . DNA/rAd prime-boost EBOV immunization generated antigen-specific CD4+ T cell immunity against proteins expressed by the vaccine insert ( Figure 3B ) . The magnitude of antigen-specific CD4+ T cells was uniform across the four immunized macaques and exceeded that observed with a single-shot rAd vaccine [5] , [12] , [18] , [19] , [20] , [21] , demonstrating the potency of DNA priming for augmentation of CD4+ T cell immunity seen by others [18] . In contrast to the species specificity demonstrated for antibody responses in this study , CD4+ T cells elicited by the vaccine gene inserts were cross-reactive with BEBOV GP . Intracellular cytokine secretion was stimulated by BEBOV GP in each of the vaccinated macaques , suggesting that dominant T-cell epitopes are contained within the few highly conserved GP regions of sufficient length ( 10–12 residues ) for MHC class II presentation and TCR recognition . Overall , the composite cellular immune response elicited by the prime-boost immunization was skewed toward CD8+ T cell activity ( Figure 3C ) . For ZEBOV , GP activation of CD8+ T cells was several fold higher ( p = . 05 ) than the corresponding CD4+ T cell responses as a percentage of the lineage memory population . Subject A01088 was a notable outlier whose CD8+ T responses were markedly lower than in other subjects . As observed for CD4-based immunity , CD8+ memory cell cytokine responses were higher than those obtained with a rAd-only vaccine , and the results showed overall that prime-boost immunization with DNA/rAd5 elicited potent antigen-specific humoral and cellular immunity . Examination of the proportions of single- , double- , and triple-positive cytokine producing cells did not reveal a dominant phenotype among the vaccinated subjects . However , the distribution of responses represented by each cytokine revealed that IL-2 positive cells were a minor component of the overall CD8+ T-cell responses , but contributed a greater proportion of the overall CD4+ T cell response as expected for this population which must be fit to undergo proliferation in response to pathogen exposure . DNA/rAd5 immunization of cynomolgus macaques protects against infection when animals are challenged with a virus species homologous to the vaccine inserts . Although there was no serological cross-reactivity between ZEBOV and BEBOV species , sequence alignment demonstrated several regions of sequence identity sufficient to comprise conserved CD4+ and CD8+ T-cell epitopes ( not shown ) , and immunized macaques exhibited robust cellular immunity against the GP from both species . To test whether immunity was sufficiently conserved to provide protection against heterologous virus infection , ZEBOV-immunized animals were challenged with a lethal dose ( 1000 TCID50 ) of BEBOV . Infection was monitored using traditional measures of filovirus infection including the appearance of maculopapular rash and damage to hepatocytes [22] . The effect of infection on hepatocytes was evaluated by measuring the liver enzymes AST and ALT . By day 10 post-infection , all control subjects exhibited severe maculopapular rash ( not shown ) and elevated liver enzymes , and viral RNA ( Figure 4 ) , characteristic of filovirus infection in macaques . In the case of control subject C1 , AST and ALT subsequently decreased to normal and near normal levels , respectively ( Figure 4A , B ) . Three out of the four control animals succumbed to infection with BEBOV between days 12 and 13 ( Figure 4D ) . One unvaccinated control animal ( C3 ) survived challenge but exhibited the full constellation of EHF symptoms , suggesting that this animal was infected but successfully cleared the infection . The time course for lethal infection of control animals was somewhat longer than a comparative infectious challenge with ZEBOV which causes death in cynomolgus macaques on average within one week following challenge [23] . Among the vaccinated animals , AST levels remained normal or near normal at all tested time points ( Figure 4B ) . Subject V4 exhibited a mild , transient increase in the serum levels of ALT , which was lower in magnitude than that observed in the control animals . Additionally , viral RNA was detected in this subject on day 6 post infection and returned to undetectable levels by the next blood draw on day 10 ( Figure 4C ) . Thus , the four macaques that received the DNA prime/rAd5 GP boost vaccine regimen generated immunity sufficient to prevent or control BEBOV infection ( p = 0 . 04 vs . controls ) . These data demonstrate that a DNA/rAd5 vaccine containing ZEBOV and SEBOV antigens provides cross-protective immunity against heterologous challenge with BEBOV .
Until recently , there were four known species of EBOV , with the most virulent being the ZEBOV and SEBOV species [24] , [25] . While there is evidence pointing to fruit bats as a possible natural reservoir for EBOV , this has not yet been definitively proven . Therefore , it is difficult to successfully implement public health measures to prevent EHF outbreaks , and the potential use of EBOV as a weapon of bioterrorism also necessitates the development of medical countermeasures to prevent and/or treat infection . The absence of effective therapies to mitigate EHF symptoms and mortality reinforces the urgent need to develop an effective vaccine against EBOV . Prior studies [5] , [6] demonstrated that rAd or DNA/rAd genetic vaccines against ZEBOV provide protection against challenge with an otherwise lethal dose of the homologous virus species . Towner , et al . , described a fifth Ebola species in 2008 , BEBOV , which was responsible for a hemorrhagic fever outbreak in Uganda with a case fatality rate of approximately 36% [11] . The Bundibugyo species has 63% , 58% and 68% sequence similarity to ZEBOV , SEBOV and CIEBOV species , respectively , and there is little serological cross reactivity between most species [26] . Since all vaccines shown to protect NHP are targeted to ZEBOV and SEBOV , it is important to determine whether new species should be incorporated as additional components of vaccine formulations against EBOV or if current vaccines may provide adequate coverage against emerging viruses such as BEBOV . The data presented here demonstrate that a vaccination strategy targeting structural proteins from ZEBOV and SEBOV was able to provide cross-protective immunity against infectious challenge with a heterologous EBOV species . This may have been due in part to the ability of DNA/rAd prime-boost vaccination to generate more robust immune responses than single-shot vaccines . The time to death for the BEBOV controls ( 12–14 days ) was somewhat longer than what we have observed for ZEBOV ( 6–12 days ) [5] suggesting it may also be possible that BEBOV is less pathogenic than other EBOV species and therefore inherently more sensitive to host immunity . The observed 100% protection against BEBOV infection in NHP would not be predicted given the divergence in GP sequence between BEBOV GP and the vaccine inserts but suggests that sufficient conservation of immunogenic regions exists between the different species . Immunization with rAd-GP one year after the final DNA prime boosted antigen-specific antibody responses to an average titer of 1/40 , 000 which is an order of magnitude higher than the titer predicted to correlate with survival from challenge with homologous virus , and demonstrates both potency and durability for this vaccine platform . It is noteworthy that rAd vectors demonstrated efficient boosting of the antigen-specific immune response when administered even a year after the final DNA prime . This result is not altogether surprising since it has been reported previously that longer prime-boost intervals may actually enhance immune responses induced by the boosting immunization [27] , possibly because memory cells have had sufficient time to undergo complete contraction and development of a central memory phenotype . The strong boosting effect of rAd5 vectors in DNA primed subjects may also help to overcome the reduced potency of these vectors when administered to subjects with pre-existing immunity to the vaccine vector . The ability of prime-boost vaccination to generate cellular responses in the form of CD4+ T-cell help as well as CD8+ T-cell effector immunity likely accounts for the protection observed after challenge with BEBOV , since vaccinated macaques in the present study lacked antibodies reactive with BEBOV but were fully protected against disease . Although the anti GP immunoglobulins did not cross-react with BEBOV , the presence of high-titer antibodies is indicative of strong underlying antigen-specific immunity induced by the vaccine , comprising antibody , CD4+ and CD8+ T-cell functions . DNA immunization has been shown to elicit antigen-specific immunity biased toward the generation of CD4+ T-cell memory responses that are necessary for long term memory and potentiate CD8+ T-cell functions , while rAd5 vaccine boosting elicits strong antibody and CD8+ T-cell responses [28] . The high magnitude of CD8+ T-cell activity exhibited here is consistent with those findings and suggests an important role for this T-cell subset in the observed protective immunity . It is noteworthy that the animal with the lowest CD8+ T-cell response , V4 , exhibited a transient increase in clinical markers of disease . T-cell-mediated protection from EBOV infection is supported not only by the robust CD4 and CD8 responses generated by the DNA/rAd vaccine but also by previous experiments demonstrating that passive transfer of anti-Ebola neutralizing monoclonal antibody ( KZ52 ) into naïve rhesus macaques had no significant effect on survival when the recipients were exposed to a lethal dose of ZEBOV [15] . The findings reported herein demonstrate a mechanistic basis for vaccine-induced immune protection against EBOV infection and will therefore inform the design of next-generation vaccines . Furthermore , this study shows that it is possible to protect against EBOV species whose antigens are not present in the vaccine formulation . This suggests that current vaccines capable of eliciting robust T-cell immunity will have the greatest potential to protect against other newly emerging pathogenic EBOV species .
The vaccine vectors used in this study have been described previously [6] . Replication-defective rAd5 GP vectors were cloned and purified as described previously [29] . Eight 3–5 year old cynomolgus macaques ( Macaca fascicularis ) weighing between 2–3 kg were obtained from Covance for this immunization and challenge study . All animal experiments were conducted under protocols approved by NIH and USAMRIID Animal Care and Use committees . All experiments involving the use of BEBOV in animals were performed in USAMRIID's BSL-4 laboratory . Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals , National Research Council , 1996 . The facility where this research was conducted is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International . Animals were housed individually , and given enrichment regularly as recommended by the Guide for the Care and Use of Laboratory Animals ( DHEW number NIH 86–23 ) . Subjects were anesthetized with ketamine prior to blood sampling or vaccination . The vaccine and control groups each contained four cynomolgus macaques . After immunization , all the animals were transferred to the Maximum Containment Laboratory ( BSL-4 ) at Ft . Detrick , MD for infection with BEBOV , and remained there through study completion . The monkeys were fed and checked at least daily according to the protocol . DNA immunizations were administered by Biojector IM injection , bilateral deltoid , with a mixture of 2 mg each of two plasmid vectors encoding GP ( Z ) and GP ( S/G ) . DNA immunizations were administered at 0 , 4 , 8 , and 14 weeks . Each subject received a boost with 1011 particle units ( PU ) of rAd5 GP ( Z ) at 12 months following the final DNA priming immunization . All animals were challenged by the intramuscular route with 1 , 000 TCID50 of BEBOV , 7 weeks post rAd5 GP boost . The challenge virus used in this study was isolated from blood specimen #200706291 from a fatal case infected during the 2007 EBOV outbreak in Bundibugyo district , Uganda . The virus was isolated on Vero E6 cells and passaged twice prior to initiating these experiments . Liver enzyme levels for serum alanine aminotransferase ( ALT ) and aspartate aminotransferase ( AST ) were determined on days 0 , 3 , 6 , 10 , 14 , 21 and 32 using a Piccolo Point-Of-Care blood analyzer ( Abaxis , Sunnyvale , CA , USA ) . Methods for the GP IgG ELISA have been described previously [5] . Briefly , polyvinyl chloride ELISA plates ( Dynatech , Vienna , VA , or Nunc , Rochester , NY ) were coated with Ebola GP , washed , and incubated with serial dilutions of 1∶50-1∶50 , 000 of subject sera or plasma . Bound IgG was detected using goat anti-human IgG ( H+L; Chemicon/Millipore , Billerica , MA ) conjugated to horseradish peroxidase and Sigma Fast o-phenylenediamine dihydrochloride ( Sigma , St . Louis , MO ) . The conformation-dependent antibody , KZ52 , is used as a control to ensure native conformation of the capture antigen , GP . ELISA titers are expressed as effective concentration 90% ( EC90 ) reciprocal dilution values , which represent the dilution achieving a 90% reduction in antigen binding . Peripheral blood mononuclear cells ( PBMC ) were isolated from cynomolgus macaque whole blood samples by separation over Ficoll , stained , and analyzed by flow cytometry essentially as described previously [4] . Briefly , PBMC were stimulated with anti-CD28 and -CD49d antibodies ( BD Biosciences ) , Brefeldin-A ( Sigma-Aldrich , St . Louis , MO ) , and either dimethylsulfoxide ( DMSO ) or a pool of 15-mer peptides overlapping by 11 spanning the ZEBOV or BEBOV GP open reading frame . Cells were stained with a mixture of antibodies against lineage markers; CD3-Cy7-APC , CD4-QD605 ( BD Biosciences ) , CD8-TRPE , and memory markers CD95 Cy5-PE ( BD Biosciences ) and CD45RA QD655 , fixed and permeabilized with Cytofix/Cytoperm ( BD Biosciences ) followed by intracellular staining with antibodies against cytokines TNFα-APC , IFNγ-Cy7-PE , and IL-2 PE . The viability dye ViViD ( Invitrogen ) was included to allow discrimination between live and dead cells [30] . Samples were acquired on an LSR II cytometer ( BD Biosciences ) and analyzed using FlowJo 8 . 8 . 5 and SPICE 5 . 0 software ( Tree Star ) . Cytokine-positive cells are expressed as a percentage within CD4+ and CD8+ T cell memory subsets after subtraction of non-specific background responses that were measured in parallel for each sample . Total RNA was isolated by mixing in a ratio of 1 to 4 . 85 plasma sample to TRI Reagent BD ( Sigma ) . Samples were decontaminated with 3% Lysol and then transferred from the high containment-level laboratory to a BSL3 room . RNAs was extracted with the RNAqueous kit ( Ambion ) and tested for BEBOV by a qRT-PCR assay . Primers and TaqMan probe for qRT-PCR were designed using the Primer Express software v2 . 0 ( Applied Biosystem , Foster , CA ) . The primers/probe were: BEBOV Fw NP 5′-TGGAAACCAAGGCGAAACTG-3′; BEBOV Rv NP 5′-ACTTGTGGCATTGGCTTGTCT-3′; BEBOV Probe 5′ FAM-CCACGGGTAGCCCCCAACCAATACA- BHQ1-3′ . Samples were prior tested in control PCR runs with either no RNA template or reverse transcriptase enzyme . One step qRT-PCRs were performed in triplicate using Iscript One-step RT-PCR kit ( Bio-rad , Hercules , CA ) in 25 µl volumes , containing 6 ng total RNA , 12 . 5 µM each primer , 5 µM probe and 0 . 25 µl reference dye BD636 ( Megabase , Inc , USA ) . To make a standard curve for the absolute quantification , a BEBOV synthetic NP RNA was generated . The fragment was amplified from a virus containing -RNA sample with the primers BEBOV Fw NP 5′-AAACGATGGTGGGTATAATA-3′ and BEBOV Rv NP 5′-AGCGGGAGGTGCAGTGGCAGGCT-3′ and then cloned in the bidirectional transcriptional vector PCR II-TOPO ( Invitrogen , Carlsbad , CA ) . Sequence and orientation of the cloned DNA was confirmed by sequencing reaction . After in vitro transcription using MAXIscript SP6/T7 Kit ( Ambion ) , the RNA was treated with DNAse-RNAse free ( Ambion ) , run onto 6%-urea acrylamide gel and purified by gel-excision followed by elution at 65°C for 4 hrs . For each run , a standard curve was generated from triplicate samples of dilution of the purified RNA , ranging from 107 to 1×101 nominal copy equivalent/reaction . Copy number of test samples was determined by interpolation of the experimentally determined CT value for the test sample onto the control standard regression curve . Calculated copy equivalent per reaction values was then normalized and expressed as copy equivalents per milliliter of starting plasma . Assay was accepted for r2 value of the standard curve being >0 . 98 . Multiple alignment of Ebola glycoprotein ( GP ) sequences ( Zaire 1976 , GenBank Accession No . NC_002549; Bundibugyo , Accession No . FJ217161; Sudan 2000 , Accession No . NC_006432 ) was performed with the program ClustalW2 available at the EBI server ( http://www . ebi . ac . uk/Tools/clustalw2/ ) . To model the molecular relationship between the glycoprotein of the vaccine strain ( Zaire Mayinga 1976 ) and of the virus infecting strain ( Bundibugyo 2007 ) , an alignment was generated by using the program MAFFT [31] and improved manually . The GP amino acid sequences included in this alignment were Ebola Zaire Mayinga 1976 ( Accession No . Q05320 ) , Zaire Ekron 1976 ( Accession No . P87671 ) , Sudan Boniface 1976 ( Accession No . Q66814 ) , Sudan Maleo 1979 ( Accession No . Q66798 ) , Gabon 1994/1997 ( Accession No . AAC57989 and O11457 ) , Sudan Gulu ( Accession No . Q7T9D9 ) , Zaire Kikwit 1995 ( Accession No . P87666 ) , Reston ( Accession No . Q66799 ) , Cote d'Ivoire ( Accession No . Q66810 ) , and Bundibugyo 2007 ( Accession No . ACI28624 ) . A distance-based phylogenetic analysis was performed using the programs in the PHYLIP 3 . 68 package [32] . The distance matrix was calculated using the Jones-Taylor-Thornton ( JTT ) substitution model [33] . These distances were clustered with the neighbor-joining ( NJ ) algorithm [34] . Five hundred nonparameteric bootstrap replicates were performed to assess support for individual clades by the data . In the figure the numbers at the nodes of the tree are the bootstrap percentages , where any value greater than 0 . 70 indicates strong support for that grouping in the data . Branch lengths are measured in substitutions per site . The multiple sequence alignment was also analyzed using parsimony with the program PAUP . Five heuristic searches , each with an independent random starting tree , were performed and a consensus of the most parsimonious tree or trees from these searches was calculated . Five hundred nonparametric bootstrap replicates were performed to assess the support of the trees by the data . Bootstrap percentages are indicated at the nodes of the tree and are interpreted as for the distance analysis . Differences in survival outcome were compared by log rank test using GraphPad Prism 5 . 0 software . Averaged data values are presented as mean ± SEM . Comparison of anti-Ebola GP antibody titers ( EC90 ) and intracellular cytokine production by T cell memory subsets were done using one-tailed T-test in GraphPad software . The authors have declared that no competing interests exist . All animal experiments were conducted under protocols approved by NIH and USAMRIID Animal Care and Use committees . All animal experiments were conducted under protocols approved by NIH and USAMRIID Animal Care and Use committees . All experiments involving the use of BEBOV in animals were performed in USAMRIID's BSL-4 laboratory . Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to principles stated in the Guide for the Care and Use of Laboratory Animals , National Research Council , 1996 . The facility where this research was conducted is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International . Animals were housed individually , and given enrichment regularly as recommended by the Guide for the Care and Use of Laboratory Animals ( DHEW number NIH 86-23 ) . | Ebola virus causes death , fear , and economic disruption during outbreaks . It is a concern worldwide as a natural pathogen and a bioterrorism agent , and has caused death to residents and tourists of Africa where the virus circulates . A vaccine strategy to protect against all circulating Ebola viruses is complicated by the fact that there are five different virus species , and individual vaccines provide protection only against those included in the vaccine . Making broad vaccines that contain multiple components is complicated , expensive , and poses challenges for regulatory approval . Therefore , in the present work , we examined whether a prime-boost immunization strategy with a vaccine targeted to one Ebola virus species could cross protect against a different species . We found that genetic immunization with vectors expressing the Ebola virus glycoprotein from Zaire blocked infection with a newly emerged virus species , Bundibugyo EBOV , not represented in the vaccine . Protection occurred in the absence of antibodies against the second species and was mediated instead by cellular immune responses . Therefore , single-component vaccines may be improved to protect against multiple Ebola viruses if they are designed to generate this type of immunity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases",
"virology/vaccines",
"immunology/immune",
"response",
"infectious",
"diseases/viral",
"infections",
"immunology/immunity",
"to",
"infections"
] | 2010 | Demonstration of Cross-Protective Vaccine Immunity against an Emerging Pathogenic Ebolavirus Species |
Positive-sense RNA virus intracellular replication is intimately associated with membrane platforms that are derived from host organelles and comprised of distinct lipid composition . For flaviviruses , such as West Nile virus strain Kunjin virus ( WNVKUN ) we have observed that these membrane platforms are derived from the endoplasmic reticulum and are rich in ( at least ) cholesterol . To extend these studies and identify the cellular lipids critical for WNVKUN replication we utilized a whole cell lipidomics approach and revealed an elevation in phospholipase A2 ( PLA2 ) activity to produce lyso-phosphatidylcholine ( lyso-PChol ) . We observed that the PLA2 enzyme family is activated in WNVKUN-infected cells and the generated lyso-PChol lipid moieties are sequestered to the subcellular sites of viral replication . The requirement for lyso-PChol was confirmed using chemical inhibition of PLA2 , where WNVKUN replication and production of infectious virus was duly affected in the presence of the inhibitors . Importantly , we could rescue chemical-induced inhibition with the exogenous addition of lyso-PChol species . Additionally , electron microscopy results indicate that lyso-PChol appears to contribute to the formation of the WNVKUN membranous replication complex ( RC ) ; particularly affecting the morphology and membrane curvature of vesicles comprising the RC . These results extend our current understanding of how flaviviruses manipulate lipid homeostasis to favour their own intracellular replication .
Cellular lipids play a vital role in the replication of flaviviruses; forming the membranous microenvironments surrounding the replication complex ( RC ) , structural components of the virus particle , and providing a source of metabolic precursors for ATP synthesis in the host cell [1–7] . Not surprisingly , modulation of lipid biosynthesis and distribution is a hallmark of flavivirus intracellular replication [3 , 6] . Previously , it has been observed that lipid droplets are an important source of fatty acids and energy and that the host enzyme fatty acid synthase plays an important role in the generation of fatty acids for dengue virus ( DENV ) , and West Nile virus ( WNV ) replication [4 , 8 , 9] . We have also previously shown a strict requirement for cholesterol and ceramide during WNV strain Kunjin virus ( WNVKUN ) replication [3 , 10] , although the utilisation of ceramide was different to that we observed for DENV . Extending these studies further other groups have performed lipidomic analyses on DENV-infected mosquito cells and whole WNV virions , and identified discrete changes and requirements of specific lipid groups during infection [5 , 6] It is evident from multiple previous studies that the intimate interactions between flaviviruses and membrane platforms within the endoplasmic reticulum ( ER ) are the governing connections that establish and facilitate efficient virus replication . During the flavivirus replication cycle , characteristic membrane structures are formed ( termed paracrystalline arrays and convoluted membranes ( PC/CM ) ) that are derived from the ER and intermediate compartment and are thought to be a site for viral protein translation and proteolytic processing [11] . Additionally , small 70-100nm vesicles are formed via invagination of the ER membrane that house the flavivirus replicative machinery and the RNA intermediate double-stranded ( ds ) RNA [11–17] . The biogenesis of these vesicles is believed to provide an efficient microenvironment for viral RNA replication and to hide immune-stimulatory molecules ( such as single stranded and dsRNA ) from host surveillance . Furthermore , the ER appears to be the site of virion assembly [18] , with arguably most flaviviruses also activating and regulating the ER and unfolded protein response that is triggered during these replicative events [19–24] . Thus , there is a need to interrogate the critical requirements and interactions that occur on these membranes platforms as the development of interventions that can diminish this interface may severely restrict and hamper efficient virus replication . For many years the role of membrane platforms during virus replication has not been well understood nor investigated , primarily due to the lack of robust tools and reagents . With the advent of lipidomics approaches and a greater biochemical understanding of lipid properties , it is clear that lipid platforms or microdomains serve as a structural scaffold for the viral polymerase complex and subsequent replication . In particular , the combination of specific lipid species and both host and viral protein induce membrane curvature to bend and shape the membrane bilayer . Depending on their orientation within the membrane lipids can induce both negative and positive curvature [25] . Notable examples are ceramide that can induce negative curvature , and lyso-phoshatidylcholine ( lyso-PChol ) that induces positive curvature , due to their cone-like structure . In contrast the precursor to lyso-PChol , phoshatidylcholine ( PChol ) , and phoshatidyletholamine ( PE ) produce a more planar or linear membrane array due to their cylindrical structure [25] . Thus , identifying virus-induced recruitment of host lipids can provide a basis for the mechanistic biogenesis of the replication complex . In this study we have undertaken a whole cell lipidomics approach to determine the alterations in phospholipid profiles of WNVKUN-infected Vero cells in order to identify key phospholipids that are either up- or down-regulated during the infectious cycle . We have identified that the host enzyme superfamily Phospholipase A2 ( PLA2 ) and the phospholipid moieties lyso-PChol play critical roles in generating the membranous WNVKUN replication complex that facilitates efficient virus replication .
Our previous studies have investigated the structure , function and composition of the WNVKUN RC [11 , 17] . These analyses have revealed that biogenesis occurs via invagination of the ER membrane , a process that appears dependent on the cellular lipid cholesterol [3] . We observed that both chemical and biological depletion of cholesterol had deleterious effects on the ability of WNVKUN to replicate and survive in cells [3] . To further those studies , we aimed to determine the roles or contribution of cellular phospholipids in facilitating WNVKUN replication . To this end , we undertook a whole-cell lipidomics approach to determine the changes in selective phospholipids within cells infected with WNVKUN , when compared to uninfected cells . Each of the phospholipid moieties were extracted in chloroform:methanol ( 2:1 v/v ) and quantified by LC-MS and levels were compared to internal standard controls and then ratios between infected and uninfected cells were determined . As can be observed in Fig 1 , we observed some variation in the influence of WNVKUN on phospholipid homeostasis with most of the phospholipid groups showing both up- and down-regulation of distinct moieties . In particular , lyso-PChol showed the highest increase in fold-change with a maximum of 2 . 9-fold increase and was the lipid species most significantly elevated overall during infection ( Fig 1A ) . Levels of phosphatidylinositol ( PI , Fig 1E ) were also upregulated , albeit to a much lesser degree than lyso-PChol with a maximum fold change of 1 . 3 . Single lipid species of PChol ( Fig 1B ) and PE ( Fig 1C ) were also elevated , with a maximum fold increase of 1 . 6 for PChol and 1 . 9 for PE . However , overall the majority of the lipid species for both phospholipid classes were decreased . Detectable species of phosphatidylserine ( PS , Fig 1D ) were all downregulated . All statistical significant upregulated lipid moieties species belonged to the lyso-PChol phospholipid class , whereas all other species were downregulated and belonged to the PChol and PE phospholipid class ( see Fig 1F ) . An overall fold-change across each phospholipid class is summarised in Fig 1G Our overall evaluation of the modulation of phospholipid homeostasis revealed that a majority of the phospholipid groups were down-regulated , with the exception of lyso-PChol and some moieties of PI and PE . This data suggests a role of the Lands cycle during WNVKUN replication where phospholipase A2 ( PLA2 ) removes fatty acids at the sn-2 position of PChol to form lyso-PChol . Acyl remodeling in the Lands cycle functions as a route to modify the fatty acid composition of phospholipids derived from the Kennedy pathway ( de novo synthesis pathway of PC and PE ) . This membrane lipid remodeling process is evolutionarily conserved among eukaryotes [26 , 27] and may be utilised by flaviviruses to induce the rearrangement of host membranes in order to house viral replication . Our lipidomics analysis had revealed that there was an associated increase in lyso-PChol with the decrease in PChol . This change in lipid biosynthesis is associated with increased activity of the host enzyme PLA2 ( Fig 2A ) and as such we investigated whether the activity of this enzyme class was increased in the WNVKUN-infected cells . We utilized a commercially available assay kit to detect cellular Phospholipase A2 activity in mock- versus WNVKUN-infected tissue cultured cells of mammalian and mosquito origin ( Fig 2B ) . Bee venom was used as a positive control to indicate assay reproducibility . We tested tissue culture cells from the lineages Vero ( African green monkey kidney ) , HEK-293T ( Human embryonic kidney ) , LN-18 ( Human brain/cerebrum ) and C6/36 ( Aedes albopictus larva ) . We found a consistent elevation of PLA2 activity in WNVKUN-infected cells in comparison to mock-infected cells ( Fig 2B ) , providing further evidence for a role of PLA2-mediated host membrane remodeling during WNV infection . As we had observed a crucial role for the enzymatic activity of PLA2 to generate lyso-PChol during WNVKUN replication , we aimed to determine if the viral RC was enriched with lyso-PChol during biogenesis . Thus , Vero cells were infected with WNVKUN and 5μM fluorophore-tagged lyso-PChol ( lyso-PChol488 ) was added to live cells and its localisation determined at 24 h . p . i . As can be observed in Fig 2C ( images i-x ) , a vast majority of the exogenously added lyso-PChol488 was observed to accumulate in viral NS3 protein-packed compartments ( images xii-x , indicated by arrows ) . Consistent with this data , we also visualised partial sequestering of fluorophore-tagged lyso-PChol lipid to viral replication complexes visualised as dsRNA intermediates in Fig 2D ( images xi-xx ) , where colocalisation of lipid moieties with dsRNA is indicated by arrows ( see images xvii-xx ) . It is interesting to note that an administration of higher concentrations of lyso-PChol488 ( e . g . 20μM ) still resulted in partial sequestering of lyso-PChol to the viral RC ( see arrows in S1 Fig , images i-xii ) , however the majority of tagged lipid was located in large aggregates in the cytoplasm of both mock- ( image vii ) and WNVKUN-infected cells ( image viii ) , most likely representing storage of Lyso-PChol in lipid droplets bodies as suggested by the increased size of lipid bodies in lyso-PChol488-administered cells ( indicated by arrow heads in image xi , compared to image v ) . In addition to subcellular localisation of phospholipids , we also aimed to determine whether exogenously added lyso-PChol could rescue the viral replication deficiencies induced by drug treatment with ACA ( N- ( P-amylcinnamoyl ) anthranilic acid ) , a broad spectrum inhibitor of PLA2 activity . Thus WNVKUN-infected Vero cells were treated with ACA alone or co-treated with ACA and lyso-PChol and virus replication was assessed by plaque assay , and western blot analysis ( Fig 2E–2G ) . We observed a significant reduction in WNVKUN virus production and the amount of viral protein produced . Upon co-administration of the ACA-treated cells with lyso-PChol we observed a recovery of virus replication back to , or even slightly more , than the untreated cells . These results indicate that the production of lyso-PChol is an important requirement for WNVKUN replication . These observations also suggest that the exogenous addition can also slightly enhance the replicative ability of WNVKUN . Overall these results indicate that lyso-PChol lipid moieties are enriched within the WNVKUN RC and most likely serves as a structural component within this membrane platform . We have also shown that it is the direct role of the production of lyso-PChol , rather than any additional function of PLA2 , that is important for WNVKUN replication . Phospholipase A2s are a family of enzymes that are divided into several main groups with different chemical structures and physiological functions . These groups are secretory PLAs ( sPLA2 ) , cytosolic ( cPLA2 ) , calcium-independent ( iPLA2 ) , platelet activating factor-acyl hydrolase ( PAF-AH ) , lysosomal PLA2 ( LPLA2 ) and adipose-specific PLA2 ( Ad-PLA2 ) [28] . In order to pinpoint a potentially responsible type of PLA2 isoform , we tested the impact of two different PLA2 inhibitors; ACA , a broad-spectrum inhibitor of calcium-dependent PLA2 ( most cPLA2 ) , and PACOCF3 ( palmityl trifluromethyl ketone ) , a reversible inhibitor of calcium-independent PLA2 ( iPLA2 ) . We also tested the impact of the fatty acid synthase inhibitor C75 as a positive control on inhibition on fatty acid synthesis ( Fig 3 ) . All inhibitors were used below their cytotoxic threshold assessed with the Promega CytoTox 96 Non-Radioactive Cytotoxicity kit . We observed that C75 , ACA and the combined treatment of PACOCF3 and ACA all had a significant impact on viral production after repeated Western blot analysis for the WNVKUN viral proteins Envelope and NS5 ( Fig 3A and quantification in Fig 3B ) . Interestingly , PACOCF3 did not appear to affect protein production significantly suggesting a requirement for calcium-dependent PLA2 , rather than calcium-independent PLA2 . The results observed for the western blots were correlated with the amount of amplified genomic RNA , as assessed by qRT-PCR ( Fig 3C ) . However , we did not observe a significant effect with C75 although an overall decrease in RNA was observed . The effects of inhibition of fatty acid synthesis and PLA2 activity on viral RNA and protein were also reflected in the significant decrease in amount of secreted infectious WNVKUN virions ( Fig 3D ) . This was also emphasized by the decreased numbers of infected cells with the inhibitor treatments ( Fig 3E ) . Overall these results suggest that calcium-dependent PLA2 are more likely to play a role during the WNVKUN replication cycle with a small synergist effect of additional calcium-independent PLA2 inhibition . However , PACOCF3 did not significantly affect WNVKUN replication alone . Our results are consistent with previous observations suggesting that fatty acid synthesis and degradation are critical for flavivirus replication [1 , 4–7] . To elucidate the influence of PLA2 inhibition on WNVKUN replication the changes in the ultrastructure of inhibitor-treated and WNVKUN-infected cells were assessed by transmission electron microscopy ( Fig 4 ) as previously described [11] . DMSO- , C75- and PACOCF3-treated cells showed numerous equally sized vesicles ( 82±12 , 88±16 and 82±12nm , respectively , Fig 4K ) close to the induced paracrystalline array ( PC ) structures ( Fig 4A–4D , 4G and 4H ) , although the CM and PC were less frequent and visually smaller in the PACOCF3-treated cells ( Fig 4G and 4H ) . In contrast , ACA- and ACA+PACOCF3-treatment resulted in significantly more elongated vesicles , some >300nm in length , ( 106±44 and 94±29nm respectively , Fig 4K ) within the VP ( arrows in Fig 4E , 4F , 4I and 4J ) , and in fact we observed very few infected cells upon the ACA+PACOCF3 treatment . The morphology of these elongated vesicles was quite striking and obviously visually different to the other drug treated samples . Intriguingly the vesicles still appeared to be tethered to the enclosing membrane of the VP . One other notable feature of these elongated vesicles was the absence of “threads” representing the viral RNA . Thus , it is possible that these vesicles are not functional for replication . These observations suggest that WNVKUN recruits lyso-PChol to contribute to the biogenesis of the viral RC and that it appears that the role of lyso-PChol is to aid in membrane curvature to generate the 70-100nm vesicular invaginations within the ER membrane . In order to confirm the involvement of PLA2 as well as defining specific enzyme isoforms that could be crucial for WNVKUN replication , we targeted two out of twenty currently known PLA2 genes via siRNA-mediated silencing ( Fig 5 ) . Based on information available on genome databases ( e . g . KEGG ( Kyoto Encyclopedia of Genes and Genomes ) pathway database , Gene cards and Reactome ) , we chose our targets according to tissue expression levels , subcellular localisation and PLA2 subgroup ( Fig 5A ) . Cytosolic phospholipase A2-alpha ( cPLA2-α or PLA2G4A ) is a well characterized and ubiquitously expressed isoform , which is believed to be involved in multiple cellular processes such as differentiation , inflammation and cytotoxicity [29 , 30] . PLA2G4A is calcium-dependent and present throughout the cytoplasm including along the ER and mitochondrial membranes , where flaviviral activity occurs [17] . In contrast , phospholipase A2-gamma ( cPLA2-γ or PLA2G4C ) belongs to a different sub group of PLA2s , the calcium-independent isoforms ( iPLA2 ) . PLA2G4C is however also present in the cytosol and has previously been implicated to play a role in replication for the related Hepatitis C virus ( HCV ) [31] . Western blot analysis showed a small but significant increase in the phosphorylation status of PLA2G4A ( p-PLA2G4A ) in infected cells compared to mock-infected cells ( Fig 5B ) . The extent of phosphorylation in the virus-infected cells was not as robust as in cells stimulated with the chemical PLA2 activator N-Ethylmaleimide ( NME ) ( Fig 5B and 5C ) . However , we observed a distinct translocation of subcellular PLA2G4A enzyme in both , NME-activated as well WNVKUN-infected cells , indicating PLA2 activation ( [32]; Fig 5D and 5E ) . We saw a striking redistribution of PLA2G4A proteins from a cytoplasmic localization in mock-infected cells ( Fig 5D , image i ) to the cell periphery ( plasma membrane or Golgi membranes ) of WNVKUN-infected ( Fig 5D , see arrowheads in image iii ) analogous to the NME-stimulated cells ( Fig 5D , image vii ) . Based on translocation of PLA2G4A proteins , approximately 77% of infected cells were identified with this phenotype ( Fig 5E , left panel ) . Furthermore , in mock-infected cells we hardly observed phosphorylated PLA2G4A ( p-PLA2G4A ) , except in dividing cells ( Fig 5D , image iv , indicated by arrow ) . However , we detected a highly significant number of phosphorylated PLA2G4A in infected cells ( Fig 5D , images iv-ix , see arrowheads ) . Approximately 63% of all WNVKUN-infected cells were also positive for phosphorylated PLA2G4A ( Fig 5E , right panel ) confirming earlier findings of activated PLA2 in WNVKUN-infected cell populations . In addition to enzyme translocation , we also observed the sequestering of a subpopulation of PLA2G4A to viral NS3 protein-packed compartments ( Fig 5F , images x-xx ) . When protein expression of PLA2G4A and PLA2G4C was transiently knocked down by siRNA-mediated gene silencing ( Fig 5G , left panel: PLA2G4A only , right panel: PLA2G4A and PLA2G4C ) , we noticed a significantly greater reduction in protein level for the viral envelope protein in comparison to the non-structural proteins NS1 and NS5 . A similar trend was noticeable in PLA2G4C-silenced infected cells , albeit not statistically significant ( Fig 5H ) . We can only speculate that due to the gene silencing of the PLA2 enzymes , a potentially altered phospholipid composition could have influenced the embedding/membrane association of the envelope proteins influencing antibody recognition . However , blotting and probing tissue culture supernatants containing virus particles for envelope proteins , revealed no difference between PLA2 gene-silenced and siRNA control-treated infected cells ( Fig 5G , left panel ) , which is consistent with unaffected virus particle production in gene-silenced versus control-treated preparations ( Fig 5I ) . Interestingly , while positive-sense viral genomes were also unaffected , we observed an increase of negative-sense viral genome intermediates ( Fig 5J ) , suggesting the ratio between positive and negative sense RNA was affected in PLA2-silenced infected cells . Overall , these results suggested that in the absence of individual PLA2 gene ( at the least for the two tested genes ) , virus replication was only marginal affected . Intriguingly , we noticed in our Western blot analysis that when the PLA2G4C gene was silenced , PLA2G4A-gene expression was significantly increased ( Fig 5G , right panel , see arrow and Fig 5K , left panel ) , whereas PLA2G4C expression remained unaltered in PLA2G4A-silenced cells ( Fig 5K , right panel ) . We could observe this specific change in gene expression consistently for both , WNVKUN-infected preparations ( as shown ) as well as in mock-infected cell preparations , suggesting a compensation mechanism amongst certain PLA2 isoforms that was not virus-induced . If that was the case , such compensation mechanism could explain why the knockdown of PLA2 genes had a marginal effect on virus replication , whereas broad range chemical inhibition of PLA2 enzymes did affect virus production . Overall these observations have shown that the host enzyme PLA2 is activated in WNVKUN-infected cells and that this activation is the likely reason for the increased production of lyso-PChol observed in the lipidomics analyses . These observations combined also suggest that activation of PLA2 is a key step during the intracellular replication cycle of WNVKUN .
Many groups , including our own , have investigated and identified key cellular and viral components that initiate and regulate intracellular flavivirus replication at the membrane interface . We were the first to observe that the flavivirus RC was comprised of small 70-100nm vesicles , that were associated with invagination of the ER [11–13] , observations that have been supported by others [14–16 , 33] . We additionally showed that the formation of the RC is dependent on the host lipids cholesterol and ceramide [3 , 10] . More advanced lipidomic studies have also revealed an altered lipid homeostasis in ZIKV- or DENV-infected mosquito cells and that sphingolipids play a major role in the assembly of WNV [5 , 6 , 34] . In this study , we performed a lipidomics analysis of Vero cells infected with WNVKUN . Our results indicate that the replication of WNVKUN is dependent on and appears to up-regulate the activity of PLA2 to produce lyso-PChol from PChol , observations that are consistent with a recent lipidomic study of dengue-infected mosquito cells [6] . Interestingly , we also observed a significant increase in some PE species . PE is a phospholipid that also contributes to the production of PChol via the CDP-DAG pathway . Thus , we propose that WNVKUN may also induce an increase in some PE species to promote PChol production that is subsequently converted to lyso-PChol via the activity of PLA2 . Perhaps not surprisingly , both phospholipids PE and PChol have been observed to contribute significantly to the replication cycles of other ( + ) RNA viruses . PChol enrichment within viral RCs has been observed for Picornaviruses , Brome Mosaic Virus ( BMV ) and Hepatitis C Virus ( HCV ) [35 , 36] . Additionally , infection of cells with HCV and Tomato Bushy Stunt virus ( TBSV ) results in the increased production of PE [37 , 38] , and the biogenesis of the TBSV RC appears dependent on the production of PE . It is becoming increasingly evident that the lipid composition within membrane microdomains is critical to viral replications cycles with lipid requirements additionally shared amongst virus families . However , it is still to be elucidated how this composition directly promotes virus replication . The recruitment of PE to facilitate TBSV replication is mediated via the viral protein p33 that interacts with and influences the activity of key cellular proteins involved in lipid synthesis [39] . It is thus of great importance to elucidate the viral proteins encoded by the other virus families that can equally contribute to changes in lipid homeostasis and distribution . It is plausible to suggest that viral proteins associated with membrane alterations are most likely to contribute in this regard . For the flaviviruses , this would be the NS4A protein that is observed to induce membrane alterations but also to modulate the ER stress response , in particular Xbp-1 [19 , 40 , 41] . Thus , as many groups have suggested , the modulation of lipid synthesis and subsequent recruitment of specific phospholipids is a general strategy for the biogenesis of viral RCs on membrane platforms . We have observed that exogenously added lyso-PChol488 was recruited and sequestered in the viral-induced membrane structures identified by antibodies against WNVKUN NS3 and dsRNA . This observation would suggest that lyso-PChol is required as a structural component and this assumption is in fact supported by our EM studies where we observed an altered morphology in the WNVKUN RC when PLA2 activity was inhibited . One of the major properties of lyso-PChol is its ability to invoke a positive membrane curvature due to its large head-group to acyl chain ratio [25] . This type of structure would certainly benefit the formation of a vesicle after the neck of the invagination has occurred ( depending on the side of the leaflet incorporated ) . We observed that the vesicles within the RC were more elongated and cylindrical in shape ( sometimes upwards of 300-350nm ) than in untreated cells ( at ~80nm ) . One other addition possibility is that due to the unavailability of lyso-PChol , WNVKUN utilises membranes that are now enriched with PChol . As PChol is a cylindrical lipid this would restrict membrane bending and result in a planar ( or straight ) membrane [25] , in this case elongating the forming VP . It would thus be of interest to determine if the VP are in fact enriched in PChol to explain this phenomenon . Interesting , similar elongated vesicles have been observed in tick cells persistently infected with Langat virus [33] . The authors suggest that the formation of these altered vesicles may be due to the production of defective particles present in persistently infected cells . However , as we have performed our EM studies on acutely infected cells we can discount this explanation in our study . Intriguingly , PE is a cone shaped lipid that is the complete reverse of lyso-PChol and can induce negative membrane curvature . As we also observed an increase of some PE species in our lipidomic study this could indicate that WNVKUN induces phospholipids with both negative and positive membrane curvature properties to invoke the construction of the invaginated vesicle on the ER membrane . One of the most interesting observations we have made in this study is the apparent strict requirement for lyso-PChol during WNVKUN replication . This is evidenced by the apparent compensation of host PLA2 species contributing to the production of lyso-PChol . We were able to successfully silence PLA2G4C gene expression , however we observed that the associated ER-located PLA2G4A protein was significantly increased ( Fig 5G ) . Thus , we speculate that the replication of WNVKUN under these conditions still induced the production of lyso-PChol providing an efficient environment for replication . This compensation phenotype could be effectively overcome by chemically treating infected cells with a broad spectrum inhibitor of PLA2 ( Fig 2 ) , invoking a deleterious effect of virus replication . In conclusion , we have shown that the activity of the host enzyme PLA2 in generating lyso-PChol is a requirement for efficient WNVKUN replication . We have shown that WNVKUN appears to utilise this phospholipid in the biogenesis of the viral RC and that perturbations to the availability of lyso-PChol is detrimental for virus replication . Thus these results also offer a potential therapeutic target in the treatment and management of flavivirus infection and add to the increasingly role of PLA2 during replication of the Flaviviridae family of viruses [6 , 42 , 43] .
Cells were infected with WNVKUN strain MRM61C at a multiplicity of infection ( m . o . i . ) of 0 . 5 , 2 or 3 as has been described previously ( 34 ) . Vero and BHK cells ( both from ATCC ) were maintained in DMEM supplemented with 5% FBS ( Lonza , Basel , Switzerland ) at 37°C with 5% CO2 . Vero , HEK-293T and LN-18 cells for gene silencing experiments and enzyme activity measurements ( cPLA2 assay kit ) were maintained in EMEM , supplemented with 1mM non-essential amino acids and 10% FBS . C6/36 mosquito cells were maintained in L-15 ( Leibovitz ) media buffered by phosphates and supplemented with 10% FBS at 28C . Chemicals C75 ( final concentration of 30μM ) , Palmityl trifluoromethyl ketone ( PACOCF3; final concentration of 15μM ) , N- ( p-Amylcinnamoyl ) anthranilic acid ( ACA; final concentration of 20μM ) and N-Ethylmaleimide ( NME; final concentration of 100μM ) were all purchased from Sigma . The lyso-PChol mix ( final concentration of 1 . 5μM of 500nM 18:0 , 500nM 18:1 and 500nM 20:0 lyso-PChol ) and fluorophore tagged lyso-PChol ( lyso-PChol488; final concentration of 5μM or 20μM ) were purchased from Avanti Polar Lipids Inc . Cytotoxicity was assessed by serial dilution of each chemical on Vero cells and viability assessed with the CytoTox 96 Non-Radioactive Cytotoxicity Assay ( Promega ) . WNVKUN specific anti-NS1 ( clone 4G4; ( 12 ) ) , anti-NS5 ( clone 5H1 . 1 ) and anti-Envelope monoclonal antibodies were generously provided by Dr Roy Hall ( University of Queensland , Brisbane , Australia ) . WNVKUN-specific rabbit anti-NS3 polyclonal antisera has been described previously ( 18 ) . Rabbit anti-Calnexin was purchased from Epitomics . Rabbit anti-Actin and anti-GRP78 were purchased from Sigma . Rabbit anti-PLA2G4A and anti-phosphorylated PLA2G4A were purchased from Cell signaling Technology , and anti-PLA2G4C from Novus Biologicals . Mouse anti-dsRNA ( clone J2 ) antibodies were purchased from English & Scientific Consulting Bt . ( Hungry ) . Alexa Fluor 488- and 594-conjugated anti-rabbit and anti-mouse specific IgG were purchased from Molecular Probes ( Invitrogen , Leiden , The Netherlands ) . Lipid extraction and analysis was performed as described previously [44] . Briefly , mock- or WNVKUN-infected Vero cells were thawed and treated with butylhydroxytoluene in ethanol . Total lipid extraction was performed by a single-phase chloroform:methanol ( 2:1 ) extraction . Lipid analysis was performed by liquid chromatography , electrospray ionization–tandem mass spectrometry using an Agilent 1200 liquid chromatography system combined with an Applied Biosystems API 4000 Q/TRAP mass spectrometer with a turbo-ionspray source ( 350°C ) and Analyst 1 . 5 data system . Precursor ion scans and neutral loss scans were performed on plasma extracts from healthy individuals to identify the major lipid species of the following phospholipid groups: phosphatidylinositol ( PI ) , phosphatidylethanolamine ( PE ) , phosphatidylcholine ( PChol ) , lyso-phosphatidylcholine ( lyso-PChol ) , and phosphatidylserine ( PS ) . Vero cell monolayers on coverslips were infected with WNVKUN and incubated at 37°C for 24h . The cells were subsequently washed with PBS and fixed with 4% paraformaldehyde ( Sigma Aldrich , St . Louis , Mo . ) and permeabilised with 0 . 1% Triton X-100 as previously described ( 20 ) . Primary and secondary antibodies were incubated within blocking buffer ( PBS containing 1% BSA ) and washed with PBS containing 0 . 1% BSA between incubation steps . After a final wash with PBS the coverslips were drained and mounted onto glass slides with a quick dry mounting medium ( United Biosciences , Brisbane , Australia ) . Images were collected using a Zeiss LSM710 confocal microscope and Zen software before processing for publication using Adobe Photoshop software . Vero cells were seeded in DMEM complete media in 6-well plates and incubated at 37°C overnight . Tissue culture supernantant containing virus particles was diluted 10-fold in 0 . 2% BSA/DMEM and cells were infected with 200 μL of stock dilutions ( in duplicate ) and incubated at 37°C for 60 min . 2 mL of a semi-solid overlay containing 0 . 3% w/v low-melting point agarose , 2 . 5% w/v FCS , P/S , Glx , HEPES and NaCO3 was added to cells and solidified at 4°C for 30 min . Cells were incubated at 37°C for 3 days , fixed in 4% v/v formaldehyde ( in PBS ) for 1 hour and stained in 0 . 4% crystal violet ( with 20% v/v methanol and PBS ) at RT for 1 hour . Plaques were manually counted and plaque-forming units per ml ( Pfu/mL ) calculated . WNVKUN-infected cells were aspirated in PBS then lysed in SDS lysis buffer ( 0 . 5% SDS , 1 mM EDTA , 50 mM Tris-HCl ) containing protease inhibitors leupeptin ( 1 μg/mL ) and PMSF ( 0 . 5 mM ) and phosphatase inhibitors sodium orthovanadate ( 25 mM ) , sodium fluoride ( 25 mM ) and β-glycerophosphate ( 25 mM ) ( Sigma ) . Lysates were diluted in SDS loading buffer ( Invitrogen ) , heated at 70°C for 5 min and separated on a 10% Tris-Glycine polyacrylamide gel . Proteins were transferred to Hi-Bond ECL nitrocellulose membrane ( Amersham Biosciences ) and the membrane was blocked with 5% w/v skim milk ( Diploma ) or 3% BSA ( Sigma ) in TBS with 0 . 05% Tween ( PBS-T ) . Primary antibodies were incubated at 4°C with membrane overnight in blocking solution as above . Following primary incubation , the membrane was washed in TBS-T then incubated with secondary antibodies conjugated to Alexa Fluor 647 or Alexa Flour 488 ( Invitrogen ) in TBS-T at RT for 2 hours . The membrane was washed twice in TBS-T then TBS , and proteins visualised on a Pharos FX Plus Molecular Imager ( Biorad ) . Cells were lysed in Trizol Reagent ( Life Technologies ) and RNA extracted as indicated by the manufacturers . 1 μg of total RNA was treated with RQ1 DNase ( Promega ) at 37°C for 30 mins to remove any contaminating cellular DNA , and cDNA generated with the Sensifast cDNA synthesis kit ( Bioline ) using both oligo d ( T ) primers and random hexamers . Gene-specific cDNAs were amplified using primers to the WNVKUN genome and the internal control RPL13A ( as previously published [19] ) with ITaq Universal Sybr Green ( Bio-Rad ) on an MX3000 real-time PCR machine ( Agilent ) . Fold induction of the WNVKUN genome was calculated by comparing threshold cycle values ( CT ) to the internal control RPL13A , as previously described [45 , 46] . Cells were fixed with 3% glutaraldehyde in 0 . 1 M cacodylate buffer for 2 h at room temperature . Cells were washed several times in 0 . 1 M cacodylate buffer followed by fixation with 1% OsO4 in 0 . 1 M cacodylate buffer for 1 h . After washing of the cells in 0 . 1 M cacodylate buffer , specimens were dehydrated in graded acetones . Subsequently , samples were infiltrated with EPON resin and polymerised for 2 days at 60°C . 50–60 nm thin sections were cut on a Leica UC7 ultramicrotome using a Diatome diamond knife and collected on formvar-coated copper mesh grids . Before viewing in a Technai TF30 transmission electron microscope cells were post-stained with 2% aqueous uranyl acetate ( UA ) and Reynold’s lead citrate . HEK293T cells were reverse transfected with 0 . 25 μM siRNA ( Bioneer: PLA2G4A ( Gene ID: 5321 ) and PLA2G4C ( Gene ID: 8605 ) , homo sapiens ) using RNAiMAX ( Invitrogen ) . Cells were incubated at 37°C , 5% CO2 for 24 hours and once again transfected with 0 . 5 μM siRNA . 24 hours post transfection , cells were infected with WNVKUN , and incubated for a further 22 hours . RNA was extracted from cells with Trizol Reagent ( Life Technologies ) by following the manufacturers’ protocol . SuperScript III Reverse Transcriptase ( Invitrogen ) and strand specific primers for WNVKUN was used to generate cDNA for viral positive and negative sense RNA , and GAPDH as the internal control . qRT-PCR was performed with ITaq Universal SybrGreen ( Bio-Rad ) , 10 μM forward and reverse primers and various templates . The in vitro colorimetric assay measures phospholipase activity present in cell lysates , based on the conversion of synthetic arachidonoyl thio-PChol substrate and subsequent detection by DTNB ( 5 , 5’-dithio-bis ( 2-nitrobenzoic acid ) . Cells were grown in minimal media in tissue culture flasks , mock- or WNVKUN-infected for 24 h and harvested via cell scraping to avoid using proteolytic enzymes . Cells were collected in ice-cold buffer ( 50mM Hepes pH 7 . 4 , 1mM EDTA ) , counted and lysed via repetitive freeze-thaw cycles and centrifugation . Lysates were assayed with the substrate for 60 min and absorbance of DTNB measured at 414nm . PLA2 activity was calculated in μmol/min/ml as per manufacturer’s formula and normalized to equal cell numbers between the different cell lineages . | Positive-sense RNA viruses remodel the host cytoplasmic membrane architecture to induce the formation of membranous organelles termed viral replication complexes . These complexes aid the virus in providing a more efficient microenvironment for replication but additionally shield immune-stimulatory molecules from the immune response . In this report we have performed whole cell lipidomic approaches to identify a key role for the host phospholipase A2 enzyme family in generating lyso-phospholipids to remodel cellular membranes and shape the West Nile virus ( WNV ) replication complex . We observed elevated PLA2 activity levels in WNV-infected cell cultures from mammalian as well as arthropod origins suggesting a generic requirement of phospholipid hydrolysis for flavivirus replication . Furthermore , we found that chemical inhibition of these enzymes severely affected the ability of WNV to replicate in cells , and we could attribute this defect to an altered ultrastructural morphology of the viral replication complex . This study provides evidence for a mechanism for the biogenesis of the flavivirus replication complex and the specific utilisation of a host lipid to invoke specific membrane curvature , generating a crucial membrane organelle required for efficient virus replication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"vero",
"cells",
"medicine",
"and",
"health",
"sciences",
"vesicles",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"biological",
"cultures",
"microbiology",
"viruses",
"rna",
"viruses",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"a... | 2018 | Phospholipase A2 activity during the replication cycle of the flavivirus West Nile virus |
Accurate modelling of biological systems requires a deeper and more complete knowledge about the molecular components and their functional associations than we currently have . Traditionally , new knowledge on protein associations generated by experiments has played a central role in systems modelling , in contrast to generally less trusted bio-computational predictions . However , we will not achieve realistic modelling of complex molecular systems if the current experimental designs lead to biased screenings of real protein networks and leave large , functionally important areas poorly characterised . To assess the likelihood of this , we have built comprehensive network models of the yeast and human proteomes by using a meta-statistical integration of diverse computationally predicted protein association datasets . We have compared these predicted networks against combined experimental datasets from seven biological resources at different level of statistical significance . These eukaryotic predicted networks resemble all the topological and noise features of the experimentally inferred networks in both species , and we also show that this observation is not due to random behaviour . In addition , the topology of the predicted networks contains information on true protein associations , beyond the constitutive first order binary predictions . We also observe that most of the reliable predicted protein associations are experimentally uncharacterised in our models , constituting the hidden or “dark matter” of networks by analogy to astronomical systems . Some of this dark matter shows enrichment of particular functions and contains key functional elements of protein networks , such as hubs associated with important functional areas like the regulation of Ras protein signal transduction in human cells . Thus , characterising this large and functionally important dark matter , elusive to established experimental designs , may be crucial for modelling biological systems . In any case , these predictions provide a valuable guide to these experimentally elusive regions .
Many features of biological systems cannot be inferred from a simple sum of their components but rather emerge as network properties [1] . Organisms comprise systems of highly integrated networks or ‘accelerating networks’ [2] in which all components ( proteins , lipids , minerals , water , etc . ) are integrated and coordinated in time and space . Given such complexity , the gaps in our current knowledge prevent us from modelling complete living organisms [3] , [4] . Therefore , the development of bio-computational approaches for identifying new protein functions and protein-protein functional associations can play an important role in systems biology [5] . The scarce knowledge of biological systems is further compounded by experimental error . It is common for different high-throughput experimental approaches , applied to the same biological system , to yield different outcomes , resulting in protein networks with different topological and biological properties [4] . However , errors are not restricted to high-throughput analysis . For example , it has been demonstrated that high-throughput yeast two-hybrid ( HT-Y2H ) interactions for human proteins are more precise than literature-curated interactions supported by a single publication [6] . There has been a great deal of work analysing biological networks across different species , giving insights into how networks evolve . However , many of these publications have yielded disparate and sometimes contradictory conclusions . Observation of poor overlap in protein networks across species [7] and divergence amongst organisms [8] suggest fast evolution . Significant variation in subunit compositions of the functional modules has also been observed in protein networks across species [9] . However , in contrast to these observations , recent work using combined protein-protein interaction data suggests high conservation of the protein networks between yeast and human [10] . This approach , based on data combination , stresses the importance of integrating different data sources to reduce the bias associated with errors in functional prediction , and to increase the coverage in network modelling , and has been demonstrated in numerous studies [11]–[14] . Increasing the accuracy of networks by integrating different protein interaction data relies on the intuitive principle that combining multiple independent sources of evidence gives greater confidence than a single source . For any genome wide computational analyses , we expect the prediction errors to be randomly distributed amongst a large sample of true negative interactions ( i . e . the universe of protein-protein interactions that do not take place ) . Hence , it is unlikely that two independent prediction methods will both identify the same false positive data in large interactomes like yeast or human . In general , we expect the precision to increase proportionally to the number of independent approaches supporting the same evidence . From the available list of well-known integration methods specifically designed to integrate diverse protein-protein interaction -PPI- datasets ( e . g . Naïve-Bayes; SVM; etc . [15]–[19] ) , we chose the Fisher method [12] in order to have a predictor that is independent from the experimental data used to validate it . Fisher integration method is not a trained or supervised method as , for example , Naive Bayes or SVM methods . The Fisher method presumes a Gaussian random distribution of the prediction datasets' scores as a null hypothesis and the Fisher integrated score calculation is based on Information Theory statistics [20] , [21] . Therefore , the Fisher integration score is completely independent of the experimental datasets used in this study to validate and compare the predictions . In this work , we significantly increase the prediction power of binary protein functional associations in yeast and human proteomes by integrating different individual prediction methods using the Fisher integration method . Three different untrained methods are implemented: GECO ( Gene Expression COmparison ) ; hiPPI ( homology inherited Protein-Protein Interactions ) ; and CODA ( Co-Occurrence Domain Analysis ) run with two protein domain classifications , CATH [22] and PFAM [23] ( see the section: Ab initio methods used for building the Predictograms ) . The four different prediction datasets obtained by these methods ( GECO , hiPPI , CODAcath and CODApfam ) , were integrated using simple integration and Fisher's method as examples of untrained methods ( see the section: Integrating the prediction data ) . Similarly ab-initio prediction datasets from STRING [14] were also integrated using Fisher integration and compared against the integrated prediction datasets from our methods . Results from the Fisher integration of our prediction datasets were benchmarked and compared against the individual prediction methods and the results from the integrated STRING methods . In all cases we demonstrate increased performance for the integrated approach ( assessed by prediction power ) with the Fisher integration of GECO , hiPPI , CODAcath and CODApfam datasets yielding the best results . Protein pairs identified by significant Fisher integration p-values were used to build a protein network model for yeast and human proteomes referred to as the Predictogram ( PG ) . Additionally , all the protein-protein associations from several major biological databases , including Reactome [24] , Kegg [25] , GO [26] , FunCat [27] , Intact [28] , MINT [29] and HRPD [30] were retrieved and combined into a network referred to as a Knowledgegram ( KG ) . As implemented in other pioneering studies [31] , we built predicted ( PG ) and experimental ( KG ) models for further comparison . Different network topology parameters were calculated and compared between KG and PG models for two test species Homo sapiens ( human ) and Sacharomyces cerevisae ( yeast ) . We observe how the networks change as the cut-off on the confidence score of the predictions is varied . Results of this PG and KG network comparison demonstrate that PG networks resemble KG networks in many of the major topological features and model a substantial fraction of real protein network associations , as previously observed in some bacterial predicted networks [32] , [33] . There have been frequent observations of low overlaps between different experimental high-throughput approaches [34] . Our comparison of the PG and KG models also show that the intersection between the two models is small and that the majority of predictions in the PG are “novel predictions” . However , the overlap between PG and KG is significantly higher than expected by random in both species supporting a correspondence between the PG and KG screenings of PPI space . This PG and KG data overlap is significantly larger in yeast than in human , pointing to a better functional characterization of the yeast PPI network and the presence of larger dark areas in the human PPI network still hidden from current experimental knowledge . We suggest that this novel prediction set may be a valuable estimation of the relative differences in “dark matter” of uncharacterised protein-protein associations between both specie , and we show that this dark matter contains key elements , such as hubs , with important functional roles in the cell . By analogy [35] , “dark matter” in protein network models refers to predicted protein-protein associations , whose existence has not yet been experimentally verified . In this study , we suggest that dark matter involves functional associations difficult to characterise by current experimental assays making any network modelling of organisms highly incomplete and therefore inaccurate . The results are divided into four main sections in which the predicted and experimental PPI models of human and yeast are compared . The first section analyses the performance of the single and integrated methods predicting the protein associations and determines the correlation between the prediction scores and the degree of accuracy and noise in the predictions . The second chapter compares the topological network features of the predicted and experimental PPI models at equivalent levels of accuracy and noise . The third section searches for functional differences between the predicted and experimental models looking for specific functional areas which appear to be illuminated by the prediction methods but elusive to the experimental approaches . Whilst the final fourth section explores whether the predicted PPI network graphs contain additional context-based information on protein associations beyond the sets of predicted protein pairs used to build the networks .
The different methods for predicting protein-protein interactions and functional associations were run on the whole yeast and human proteomes , generating four prediction datasets for each organism , GECO , CODAcath , CODApfam and hiPPI ( see the section: Running the PG methods on the human and yeast proteomes and section 1 in Text S1 ) . Each of these methods produces an untrained score value , which was normalized to a p-value , reflecting the reliability of the predictions ( see the section: P-value calculation ) . Benchmark datasets for each organism , comprising reliable protein pairs based on Gene Ontology Semantic Similarity scores ( referred to as the Goss refined – Gossr datasets; see the section: The GO Semantic Similarity refined dataset ( Gossr ) used for validating the prediction methods ) , were used to assess performance ( note that the performance measured will depend on the quality of the validation dataset; see section 2 in Text S1; [4] ) . Precision values are estimated by comparing the methods performance in predicting true PPI versus a random predictor , used to calculate the FP ( False Positive ) rates ( see the section: Precision and Recall calculation ) . We find that for all methods the p-values correlate inversely with the precision scores , in both proteomes , as expected if genuine functional information is linked to the prediction score ( Figures 1a and b ) . It is possible that a randomly selected PPI could be a TP by chance . However , this is likely to be a rare event and in any case it will mean that we tend to underestimate the performance of the methods as it would mean we are overestimating FPs , from our random model ( see section 2 in Text S1 ) . The mutual information scores demonstrate the independence of the 4 different prediction datasets ( see section 3 in Text S1 ) . The p-values from the 4 prediction datasets were integrated using Fisher and Simple integration , both of which are untrained integration methods ( see the section: Integrating the prediction data ) . Precision ( TP/TP+FP ) versus Recall ( Recall considered as the number of predicted hits ) is plotted for yeast and human Gossr validation ( Figures 1c and d ) , for all the individual and integrated methods in order to compare their statistical prediction power ( prediction power equals the area under the Precision vs . Recall curves ) . The prediction powers of all of the integrated methods outperform any individual method . Increase of the prediction power following integration is especially pronounced in human . Whilst less pronounced , the increase in yeast remains significant above 80% precision ( zoom over Figure 1c ) . At these higher precision levels differences in the predictive powers become very significant with the Fisher integration methods approximately doubling the recall for a given precision over the best single or simple integration methods ( around 90 , 000 predictions with Fisher compared to around 60 , 000 predictions with simple integration , see the abscissas axis in zoom of Figure 1c ) . We have performed additional validation of the Fisher method using a set of physical interacting pairs as gold standards in yeast and human ( see section 4 in Text S1 ) . Validation with the Int ( physical interaction ) dataset in yeast ( Figure S3a in Text S1 ) assigns a higher precision ≥90% to the same predicted dataset of around 90 , 000 top ranked Fisher predictions , which were calculated with a precision ≥80% in the Gossr validation ( Figure 1a ) . Whilst in human the Int and Reactome_int ( physical interaction ) validations ( Figure S3b in Text S1 ) yielded precisions of ≥76% and ≥82% for the same top ranked Fisher dataset that was assigned a precision ≥80% in the Gossr validation ( Figure 1b ) . All these validations indicate that Fisher p-values scores are also linked to physical protein-protein interactions with a similar consistent reliability of around 80% precision as shown in the Gossr validation . Fisher was also implemented to integrate similar datasets of individual STRING ab-initio predictions ( gene neighbourhood , co-occurrence , fusion , and co-expression ) in yeast and human . FisherW integration of the STRING datasets showed a significantly lower performance compared to the GECO , CODAcath , CODApfam and hiPPI Fisher integration ( see section 5 in Text S1 ) . Using Gossr as the training dataset , GECO , CODAcath , CODApfam and hiPPI prediction datasets were also integrated by Bayes ( see section 6 in Text S1 ) . Bayes integration produced uneven results in yeast and human compared to Fisher , whilst Fisher outperforms Bayes for the highest levels of precision in yeast ( see left side of the Figure S5a in Text S1 ) , in human Bayes performs better ( see Figure S5b in Text S1 ) . From these results we observed that the Fisher integration yields a good performance compared to using a trained method ( i . e . Bayes ) , despite the fact that the latter has benefit of learning from the experimental ( KG ) information to predict PPIs . In all cases ( yeast and human ) Fisher integration of the GECO , CODAcath , CODApfam and hiPPI predictions was shown to be a powerful combination which significantly increases the prediction power without using any KG trained or supervised algorithms . This premise is crucial if we aim to detect genuine similarities between the PG and KG models , unbiased by overlap between supervised predictions and their training sets ( as would occur by using a Bayes integration ) . Because of this the Fisher weighted predictions were chosen for generating the PG network models used in subsequent analyses of the networks . To test whether the PG networks based on the binary predictions share features with networks built on reliable KG evidences , different topological parameters ( see section 7 in Text S1 ) were calculated and compared between PG and KG networks . This analysis was carried out at different levels of significance in the yeast and human proteomes . Different PG networks were constructed from the binary predictions by varying the link ( edge ) p-value cut-off . This was done for a range of p-values from p-value≤0 . 001 ( PG0 . 001 ) to p-value≤1 . 0 ( PG1 . 0 ) . KG network models were also tested at different levels of confidence based on the number of KG evidences supporting the same protein-protein associations . Mutual information calculation on the KG data showed broad independence except for the Goss and Foss ( FunCat semantic similarity ) datasets , therefore Goss and Foss evidences were summed and considered as a single dataset of KG evidences . Different KG networks were constructed by varying the minimum number of independent evidences required to form an edge/link . Random models were also generated for all the PG and KG networks as described in the section: Network randomisation . The PG , KG and their corresponding randomised networks , built at different significance levels , provide comparable frameworks for examining the topological properties of biological networks . Real biological networks have been shown to have a scale-free topology with a high degree of clustering [36] . Scale free networks show , amongst other characteristics , a power law distribution in the frequencies of connectivity of their nodes ( ki ) with values for the exponent between 2 and 3 [37] , [38] . When the frequency distribution for node connectivity ( ki ) is plotted for the PG and KG networks , constructed at different confidence levels for yeast and human , we observe a trend towards higher exponents in the fitted power law functions as the network reliability increases ( Figures 2a , 2b , 3a and 3b; and Figure S6a–Figure S8a in Text S1 ) . The trend toward scale-free organisation is more significant in yeast than in human KG and PG models , with exponent values that get close to 2 for the most reliable network levels ( see PG0 . 01 and KG≥3 evid . distributions in Figures 2a and b ) , whilst in human KG and PG models the exponents are systematically lower than in yeast , at equivalent levels of significance ( Figures 3a and b ) . Yeast and human KG and PG models show non-random distributions of their degree ( ki ) frequencies for all levels of network reliability tested , except for the lowest level ( Figures 2 and 3 , compare plots a and b , c and d , e and f; and Figure S6 , Figure S7 and Figure S8 in Text S1 compare plots a , b and c ) . In contrast to the real PG and KG models the adjacency randomised networks in yeast and human show a Gaussian distribution with node degree ( Figures 2 and 3 , plots c and d versus a and b ) . A Gaussian distribution is also observed in the p-value random models but for high node degree only ( Figures 2 and 3 , plots e and f ) . A Gaussian distribution , typical of random behaviour , is also observed for the KG and PG networks built at the lowest level of statistical reliability ( compare PGtotal and KGtotal in Figures 2 and 3 , plots a , b , e and f ) . However , this Gaussian random distribution disappears when edges with weaker statistical weight are removed in increasingly more significant PG and KG networks , indicating the correlation between edge statistical weight ( p-values and # of evidences ) and the non-random scale free topology expected in real biological networks . Power-law degree distribution is a necessary but not a sufficient characteristic of scale free networks . Therefore , other topological features of the KG and PG networks were measured in order to give more support to the hypothesis of scale free tendency for our models . These included: average clustering coefficient; assortativity; or network hierarchy amongst other parameters described in the section: Network topology structure characterisation and the section 7 in Text S1 . The trend of increasing average clustering coefficient with increasing network reliability ( KG and PG network models built at more highly significant p-values and # evidences levels ) lends further support to the scale-free organization of the KG and PG networks in yeast and human ( see Figure S6d–Figure S8d in Text S1 ) . Node assortativity ( or preferential attachment of the nodes ) is another topological parameter that supports the scale-free trend of the KG and PG models in yeast and human , ( see section 11 in Text S1 , and Figure S9 and Figure S10 in Text S1; [36] ) . The assortativity observed in KG and PG models indicates a network organization close to a real network in stark contrast to the random models [39] , [40] . Network hierarchy is another topological feature that can be considered by using the logarithmic distribution of the clustering parameter [41] . For all our KG and PG networks we observed a flat distribution ( no correlation between clustering coefficient and connectivity –ki- ) implying a non-hierarchical organization , since hierarchical organization exhibits a power-law distribution of these two parameters ( see section 12 and Figure S11 in Text S1 ) . This result , taken together with the observations on clustering distribution , indicates a modular organisation of the KG and PG networks [41] . This would explain why these networks tend towards , but never completely reach a scale-free distribution exponent [36] . The modularity of the KG and PG networks ( see section 13 and Figure S12 in Text S1 ) is also supported by other conventional network parameter values measured for these networks and presented in Table S3 and Table S4 in , such as: network density; cluster average ( triangle formation likelihood ) ; characteristic path length; network radius; and diameter . Radius and diameter are only measured for the largest connected component of the network [40] , [42] . KG models represent the known ( experimentally determined ) protein associations while PG models represent sets of associations predicted by ab-initio methods . We wanted to estimate the extent of ‘dark matter’ in the yeast and human networks by comparing how much of the predicted network space was not covered by experimental evidence in both specie . We also investigated the presence of hubs in the PG dark matter and the functional characteristics of these dark ( hidden ) hubs . We used the most reliable ( precision≥80% ) PG models ( PG0 . 01 in yeast – about 90 , 000 pairs - and PG0 . 014 in human – about 106 pairs; see section 15 and Table S5 in Text S1 ) to estimate the intersection with the KG models for the two organisms ( Table 1 ) . In yeast , the percentage of edges ( 18% ) overlapping between the KG and PG models is larger than for the human models ( 1 . 34%; Table 1 ) . That is , 18% of the predicted protein-protein associations in yeast PG0 . 01 model are backed by experimental evidences in the KG set , which is a highly significant figure compared to any of the random models ( 18 . 22/1 . 34 = 13 . 60 times higher than R . 1 model , and 18 . 22/4 . 26 = 4 . 28 times higher than R . 2 model; see Table 1 ) . The percentage of edge predictions , backed by experiments , drops considerably for human . Only 1 . 4 percent of the predicted protein associations ( PG ) were also present in the KG model ( Table 1 ) . Although , the percentage of experimentally backed predictions ( 1 . 34% ) is significantly higher than expected by random ( 1 . 34/0 . 08 = 16 . 75 times higher than R . 1 model , and 1 . 34/0 . 29 = 4 , 62 times higher than R . 2 model; see Table 1 ) . The density of the overlap between PG and KG in yeast ( #Ed . /#Nod . = 7 . 6 in Table 1 ) is double the human value ( 3 . 5 ) , and in both cases is considerably more than the expected random density ( Table 1 ) . Additionally , the percentage of proteins ( nodes ) in the PG model without known experimental association in the KG model is about 30 times smaller in yeast ( 105 nodes/4374 nodes = >2 . 4% see Table S5 in Text S1 ) compared to the human PG model ( 13 , 961 nodes/19 , 618 nodes = >71 . 2% see Table S5 in Text S1 ) . These statistical analyses of the PG and KG intersections indicate that about 82% , in yeast , and 98% of predicted protein associations in human are not backed by experimental evidence in the KG model , giving an estimate of dark matter in the yeast and human protein networks . Only 2 . 4% of the PG proteins in yeast are dark nodes ( proteins without experimental association in the KG model ) , whilst dark nodes constitute 71% in human . Although PG and KG models explore significantly different regions of protein binary association space , interestingly , given the small intersections and density values , the PG and KG overlap is still significantly larger than expected by random ( Table 1 ) , indicating the overall coherence of the PG and KG models despite the presumably huge size of real protein network space . It is likely that protein network space is much larger in human than in yeast , given their respective proteome sizes , which presumably explains the higher proportion of dark matter in the human compared to the yeast PG networks . Enrichment of the degree of a node in the PG model ( PGki_er ) was calculated in order to measure the difference in the connectivity ( ki ) values for a protein in the PG and KG networks ( see the section: Calculating the PGki enrichment ratio and the PG functional enrichment ) . A high PGki_er value indicates the presence of a dark ( experimentally hidden ) hub , a protein with many predicted associated proteins in the PG model and few , if any , experimentally validated KG associations . Proteins in the yeast and human PG models were ranked using their PGki_er value , retrieving the top 10 ranked proteins for both organisms ( see Table 2 ) as the most likely representatives of predicted dark hubs . A common interesting feature of dark hubs , shown in Table 2 , is that almost all of them correspond to predicted proteins with only electronically inferred or unknown functions in Uniprot . This is expected for proteins which are absent from the KG model and therefore have no associated functional evidence . This overrepresentation of functionally unknown proteins in the set of dark hubs is also supported by extensive functional annotation searches using the DAVID algorithm [43] in yeast and human ( see section 16 in Text S1 ) . Although enrichment in predicted datasets of uncharacterised proteins has also been observed in earlier studies by other groups [31] , it was not used to identify sets of dark hub proteins , as in our study . Here , we identify highly connected and therefore topologically important nodes in the PPI networks currently lacking direct experimental information . We analysed the top 10 dark hubs in the yeast PG network using functional annotation inferred by homology , these proteins correspond mainly to membrane embedded proteins , although there are also proteins related to other disparate functions , such as: transcription factors , RNase ( probably involve in siRNA degradation processes ) , sporulation , and various enzymes ( see Table 2 ) . Enrichment bias in “integral to membrane proteins” is statistically significant in the yeast dark hubs dataset comparing the extremes of the PGki_er ranked list with the DAVID algorithm ( see section 16 in Text S1 ) . Functions for the top 10 dark hubs in humans are even broader than in yeast including proteins with Fibronectin domains , kinases with presumably sensor or motor functions , an Ecto-5′-nucleotidase probably involved in extracellular nucleotide catabolism [44] , a transcription factor , and a matrix metallopeptidase amongst other proteins of completely unknown function ( see Table 2 ) . In order to study possible bias in the functional niches highlighted by the PG predictions but absent in the KGs , functional enrichment in the yeast and human PGki_er ranked lists was estimated using the GOrilla server [45] and the annotations of the respective proteomes in the GO database examined ( see Table 3 ) . Functional enrichment at the top of the ranked lists implies the existence of dark functional niches which are more accessible to ab-initio predictions than to experiments . GOrilla did not find any significant functional enrichment bias ( P-value>E-9 ) in the yeast ranked list , but detected enrichment of some GO terms in the human ranked list associated with particular biological processes and molecular function categories in GO ( see Table 3 ) . Dark ( or experimentally hidden ) functional niches in the human PG models correspond to key biological processes such as kinase driven regulation through protein amino acid phosphorylation and the regulation of GTPase mediated signalling , including the regulation of Ras protein signal transduction . The ATP binding GO molecular function enrichment is mainly associated with enrichment of kinases . If the reliable PG0 . 01&0 . 014 pairwise predictions capture a significant percentage of true functional relationships and the PG0 . 01&0 . 014 networks show most of the topological properties of KG networks , it is reasonable to expect that the topology associations in these PG0 . 01&0 . 014 networks will resemble real biological networks . In other words , we should be able to exploit information on the context of a protein ( i . e . connections in the network ) to predict associations it has with other proteins sharing a similar context . In order to test this hypothesis , functional predictions were generated for additional protein pairs , by comparing the interactions of the respective proteins in these pairs , in the PG networks . The results were then validated using the gold standard KG protein pairs' datasets . This context analysis of the PG networks [32] , which involves making predictions based on predictions , is what Mathematical Logic terms a second order analysis . The PG0 . 001&0 . 0014 pair-wise predictions' datasets used to build the networks in first place are considered the first order predictions in this work ( see section 17 in Text S1 ) . Comparison of the association profiles identified 1 , 668 , 584 protein pairs in yeast and 49 , 117 , 115 protein pairs in human sharing at least one third of their interacting proteins in the PG0 . 01&0 . 014 network matrices . The similarity scores of the profiles were validated using the different KG datasets i . e . Int , Kegg , Goss , Foss , Reactome , and Reactome_int ( see Figure S14 , Figure S15 , Figure S16 , Figure S17 , Figure S18 , Figure S19 , and Figure S20 in Text S1 ) and the integrated and refined KG≥2 evidences dataset ( Figure 4 ) . Bits and specific bits similarity scores ( see the section: Second order predictions from the PG networks: Measuring the similarity of protein interaction profiles ) positively correlate with an increase in precision for all the KG datasets ( see Figure S14 , Figure S15 , Figure S16 , Figure S17 , Figure S18 , Figure S19 , and Figure S20 in ) and the refined KG≥2 dataset ( Figure 4 ) . Bits and specific bits scores show very similar behaviour in all the KG datasets most probably due to the large set of potential random interactions in both PG matrices that make it very unlikely that two proteins would share a significant number of interactions by chance ( see section 18 in Text S1 ) . First order predictions based on Fisher scores yielded about 90 , 000 predictions in yeast with a precision≥80% ( see Figure 1 ) , while second order predictions only yielded 95 predictions at the 80% precision level in the KG≥2 validation dataset ( Figure 4b ) and 8 , 390 predictions maximum in the single evidence KG datasets ( Kegg validation recall in yeast; see Figure S18b in Text S1 ) . The same observation is valid for human with about 1 , 000 , 000 hits at 80% precision level in the Fisher first order predictions and only 889 second order predictions at 80% precision in the KG≥2 validation ( Figure 4c ) and a maximum of 118 , 800 predictions in the single KG datasets ( Reactome validation recall in Figure S19f in Text S1 ) . Since second order predictions are predictions performed over first order predictions , there is likely to be an accumulation of second order error over the primary error , lowering the general performance . Nevertheless , a common observation in all the validations is that the PG0 . 001&0 . 0014 networks have second order functional information of real biological value absent in the first order predictions . Although , using context does not predict many more interactions , this analysis is important because it confirms that the topology of our predicted network has real biological meaning .
The scoring functions of the three ab-initio methods ( GECO , CODA and hiPPI ) , showed close correlation with precision in predicting true , functionally associated proteins ( Figure 1 ) . The correspondence of the p-value scoring functions with prediction reliability and also the complementary nature of the prediction datasets , suggested by their independent mutual information , enabled the Fisher integration method to perform well . Fisher meta-statistic , untrained , integration of the four datasets ( GECO , CODAcath , CODApfam and hiPPI ) yielded a significant increase in prediction power within the yeast and human proteomes , adding value beyond any single method or the sum of all of them . Fisher integration thus allowed us to build comprehensive PG0 . 01&0 . 014 integrated models independent from the KG data , and at highly reliable precision levels ( 80% ) for yeast and human . While the KG network models contain much of the current knowledge on protein functional associations provided by disparate experimental resources , in yeast and human , the PG models represent sets of predictions inferred by the integration of different ab-initio ( non-experimental ) methods . Experimental ( KG ) and predicted ( PG ) networks share all of the main topological features explored in this work . In summary the node ki degree distribution , assortativity , clustering distribution , and clustering average coefficient for each of the PG and KG networks demonstrate a trend towards a scale-free organization as network confidence increases . KG and PG are both non-random networks , both in the connectivity and in the statistical weight distributions of their edges ( Figures 2 and 3; sections 8–13 in Text S1 ) . Different data integration methods are applied for reducing noise ( error ) in the KG and PG models , thereby generating analogous frameworks for the KG and PG models built at different reliability levels . In the KG models the associated error is inversely correlated to the number of evidences supporting a given protein-protein association . Reducing error by summing evidences is analogous to the repetition of experiments carried out in standard experimental protocols [3] . In the PG models , the Fisher method reduces noise by integrating the weighted ( p-value ) evidences within a probability space which has finer resolution than the presence/absence binary space used in the KG models . Gaussian distributions , typical of random network topologies , appear in the high node ( ki ) connectivity part of the plots for the least reliable KG and PG models , disappearing in the KG and PG models built at higher levels of reliability ( see Figures 2 and 3 ) . This indicates that errors in determining true protein associations are common to both KG and PG models and that the KG and PG network topologies respond in the same way to analogous methods for reducing noise ( data integration ) . We also observed that the topology of the PG0 . 01&0 . 014 models have functional information of real biological networks beyond first order binary predictions ( see Figure 4; and section 18 in Text S1 ) . Since one of the prediction methods , hiPPI , exploits available experimental data by inheriting experimentally validated interactions between homologous proteins there may be some concern that the dependency of the hiPPI predictions on some of the KG datasets could bias the PG network models so that the features resemble those of experimental KG networks . Addressing this possibility we repeated the main analyses of this work excluding the hiPPI predictions and demonstrated that the similarity of the PG and KG models remained and is therefore not due to any circular information or bias . This confirms our previous observations and conclusions of our work ( see section 21 in Text S1 ) . Coverage of reliable PG0 . 01&0 . 014 predictions by KG datasets appears much higher in yeast ( 18% ) than in human ( 1 . 34% ) for all the analysed cases ( Table 1 ) , highlighting the better network characterisation in the yeast proteome network . 82% of the predicted associations in yeast were not backed by any KG data indicating considerable dark matter in the yeast PG network . These figures are even higher for human; where a 98 . 5% of predictions are absent from the KG databases . Although low overlaps between high-throughput experimental datasets is not a surprising observation , the relative differences in the amount of dark matter found for yeast and human hint at important differences in the progress of our knowledge of these two organisms' PPI networks . The dark matter in the PG models of yeast and human contains hubs ( i . e . dark hubs ) which are key for network integration and functioning and which seem to be involved in disparate functions in both organisms ( Table 2 ) . In yeast dark hubs include many membrane embedded proteins with unknown functions . Membrane proteins are usually more poorly characterised than soluble proteins , due to the current design of experimental techniques , and therefore prediction methods could assist in characterising associations for these proteins . For human , the top ranked dark hub dataset ( see n column in Table 3 ) is significantly enriched in kinases and GTPase/Ras regulatory proteins associated with important biological regulatory pathways . These results reveal the existence of key regions ( i . e dark functional regions ) belonging to protein network functional space that are poorly characterised by experimental sciences but highly represented in the PG models . As for the membrane proteins in yeast , predictions of these proteins would be helpful in identifying associations which currently elude experimental approaches . It is quite well known that current experimental high-throughput datasets show limitations with respect to coverage and also systematic errors . For example , Y2H does not perform well on membrane-associated proteins and transient interactions tend to be under-reported [34] . This observation agrees with our analyses , which shows that dark hubs are particularly enriched in integral membrane proteins and transient interactions such as those involved in kinase mediated regulation , a mechanism over-represented in the Ras signalling pathway . Dark matter may even be more extensive than suggested by the initial comparison of PG and KG models . KG and PG models both show a non-hierarchical structure , as shown by the clustering parameter distribution ( Figure S11 in Text S1 ) , whilst preserving a highly modular structure ( Table S3 and Table S4 in Text S1 ) . Since all the functional modules must ultimately be integrated within a functioning organism , the high modularity and non-hierarchical structure suggests that our PG and KG models are incomplete lacking proteins ( nodes ) , and protein-protein associations ( edges ) still uncharacterised in our KG and PG models . Since much of the PG network is dark matter containing hubs and other important functional regions not easily reached by current experimental designs ( especially in more complex organisms like human ) , and since the PG models show the most important properties of real , biological networks , resembling the properties observed in the KG models , we can conclude that the yeast and the human PG networks are valuable models , akin to the currently more accepted KG models , for investigating the properties of real biological networks , complementing and completing experimental studies in Systems Biology .
Overview of the Methods . Homology inherited Protein-Protein Interaction ( hiPPI ) method , scores potential protein-protein interactions based on their homology to known interacting protein pairs; Co-Occurrence Domain Analysis ( CODA ) method , looks for and scores protein pairs in a given target genome ( e . g . yeast or human ) found as fused ( Co-Occurring ) domain architectures in homologues from genomes of 575 different species; Gene Expression COmparison ( GECO ) method , measures the correlation of gene expression profiles between protein pairs ( detailed explanation of the ab initio methods in section 1 in Text S1 ) . The GECO , hiPPI , CODAcath and CODApfam methods were run against all sequences in the human ( Homo sapiens ) and yeast ( Sacharomyces cerevisiae ) proteomes ( detailed datasets information in section 19 in Text S1 ) . Proteome files were downloaded from the Integr8 database June 2007 ( section 19 in Text S1 ) . GECO retrieved 26 , 292 , 126 protein pairs of predictions for human and 10 , 371 , 735 for yeast with total sequence coverage of 21% and 81 . 5% respectively . hiPPI yielded 86 , 099 protein pairs of predictions for human and 12 , 070 for yeast , with total protein sequence coverage of 31% and 56 . 6% respectively . CODAcath yielded 32 , 259 , 881 and 678 , 928 predictions ( coverage 39% and 36 . 4% ) for human and yeast respectively . Whilst CODApfam generated 24 , 984 , 943 and 336 , 781 predictions ( coverage 57% and 58 . 4% ) , for human and yeast respectively . We benchmarked our predictions using the highest quality annotations of yeast and human proteomes in the Gene Ontology ( GO ) database [26] . GO provides annotation codes which enable the selection of protein annotations based on quality and evidence source ( see further details in section 14 in Text S1 ) . The GO terms' Semantic Similarity ( Goss ) scores were calculated for all versus all protein pairs in human and yeast proteomes as described by Lord et al . 2003 [49] , using the GO relational graph implicit in the GO ontology file ( GO ontology files; OBO v1 . 0 format 30th-October-2008; http://www . geneontology . org/ ) . Sets of protein pairs with significant Goss score ( Goss≥4 . 0; [50] ) in the refined sets of GO annotations were selected as validating datasets for the yeast and human protein pair predictions . These protein pair sets are referred to Goss refined ( Gossr ) datasets . Precision was calculated as the ratio of accumulative TP/TP+FP at different prediction p-values , where TP ( True Positives ) is the rate of hits predicted within the validation dataset of true protein binary associations ( e . g . Gossr , see section above ) , and FP ( False positive ) is the average rate of hits predicted from 1000 random models of the same validation dataset . The FP are the randomly selected PPIs above different scoring thresholds ( i . e . prediction p-values ) . The FPs are calculated as an average of 1000 random validation iterations to estimate the errors ( deviations ) associated with the calculation . We then compare the relative differences in the TP and FP rates in the ranked prediction list , obtained by using our predictor and a random approach . For example , a precision ≥90% associated with a p-value≤0 , 001 means we find 9 times more TPs in the set of predictions with p-value≤0 , 001 than a random predictor does by chance . In our analyses the precision ( ie TP/TP+FP ) will always tend to 50% because we select the same number of FPs from our random predictors as given by the integrated prediction method . Using a random model for benchmarking it is possible that a randomly selected PPI could be a known TP , by chance , although the probability is expected to be very low since the space of known PPI ( TPs ) is much lower than the space of random PPIs pairs considering all possible combinations . It is also likely that any of the gold standard datasets , or combinations of them , do not contain all the true PPIs taking place in nature . Therefore it is not possible to correctly estimate FPs in the ranked predictions , based on pairs absent in the validating datasets ( ie many of these FPs may be currently uncharacterised TPs ) . In any case , the consequence of considering TPs as FPs in the random validation model used in this work is conservative , giving an underestimate of the performance of our predictor ( see section 2 in Text S1 ) . Although recall is usually defined as the TP/ ( TP+FN ) ratio , since not all the true PPI are known in our validation model , we can not reliably estimate the FN rates . Therefore , in this work Recall is calculated as the accumulated number of predicted hits by a given method , at different p-value levels . Yeast and human PG protein networks were built based on the binary protein prediction data selected at different discrete Fisher_W p-value statistical significance levels . Fisher_W predictions were chosen because these gave the best results from the benchmarking . Various PG networks were generated over a range of predicted p-value cut-offs . The p-value cut-off used to generate a given PG network is specified in the subscript of its name . For example if a p-value cut-off ≤0 . 01 was used the PG network was termed PG0 . 01 . The construction of KG protein networks for human and yeast proteomes was based on the existence of protein functional links . For the interaction databases HRPD [30] , MINT [29] , Intact [28] evidence of a protein interaction gave an evidence score of 1 . For the pathway resources , Reactome [24] and Kegg [25] , shared pathway membership was sufficient for an evidence score of 1 . An extra Reactome_Int dataset was built based on physical protein interaction evidence in Reactome . Binary protein associations in GO and FunCat were identified using the Semantic Similarity score calculated using the ontology association graphs in GO and FunCat [27] respectively . For sets based on GO ( Goss datasets ) we used all the annotations in GO in order to maximise coverage of GO functional space within the KG networks . These Goss datasets are therefore expected to contain more noise that the refined Gossr datasets used to validate the methods ( see PG methods' validation section above ) . Semantic similarity values were calculated with the Resnik method [49] , [51] as described in the section above: The GO Semantic Similarity refined dataset ( Gossr ) used for validating the prediction methods . Sets of protein pairs with significant functional associations ( Goss≥4 . 0; [50] ) gave a score of 1 . A Foss ( FunCat semantic similarity ) significant set was obtained from the FunCat [27] database . Foss score was calculated using the same process as GO with the Resnik method . Int dataset was generated by the union of all the above datasets excluding Goss , Foss , Kegg and Reactome . A cumulative score was associated with each edge ( functional link ) to represent the number of independent resources with evidence of the functional link between the two proteins . The KG models statistics are shown in Table S5 in Text S1 . Two different randomisation procedures were implemented . The first method randomised the p-values associated with edges in the PG network and the # of evidences associated with edges in the KG models , whilst keeping the same pairs of connected nodes in the matrices . These models are referred to as p-values random models and they were built to analyse the distribution of the statistical weights associated with protein edges ( p-values and # of evidences related to edges ) compared to random behaviour . The second randomised model , referred to as the adjacency random model , was generated by randomly distributing all nodes , p-values and # of evidences in the PG and KG pairs-wise datasets . Any new self-associations in the PG network datasets produced by the randomisations were removed . The adjacency random models were built to analyse the distribution of edges and p-values in the KG and PG models compared to random behaviour . Both models went through 1000 randomisation iterations . In order to compare the PG/KG networks generated by this study several different network statistical features were calculated . Topological parameters included the node degree connection ( ki ) [37] , [38] , degree correlation ( assortativity ) [37]–[40] , clustering distribution [41] , [52] and average clustering coefficient [37] . Distance based metrics to characterise the networks included the characteristic path length ℓ [37] , radius , diameter and eccentricity [42] ( see section 7 in Text S1 ) . In order to determine whether some nodes had elevated degree connections in the PG , the relative enrichment of the node degree connection ( ki ) for nodes in the PG network compared to the KG network was calculated for all the nodes ( proteins ) using the following formula: PGki_er ( pi ) = ( PGki−KGki ) / ( KGki+1 ) where PGki_er is the PGki enrichment ratio of the protein pi , PGki is the ki value of the protein pi in the KG network and KGki is the ki value of the protein pi in the KG network . Yeast and human proteins were ranked using the PGki_er parameter values and the ranked lists were used as input for the GOrilla web server ( http://cbl-gorilla . cs . technion . ac . il/ ) . GOrilla is a tool for identifying and visualizing enriched GO terms in ranked lists of genes ( Eden et al . 2009 , [45] ) . For each protein pair , the vectors of interacting proteins , within the PG0 . 01 in yeast and the PG0 . 014 in human network matrices ( 0 , 01 and 0 , 014 cut-offs relate to 80% precision in yeast and human respectively ) , were compared using different similarity measures , such as: bits , specific bits and congruence . These similarity scores , which are calculated over the PG network matrices , are termed second order predictions ( see section 17 in Text S1 ) . The bits score formula is b ( p1 , p2 ) = b1 , where p1 and p2 are the two proteins compared and b1 is the number of shared interacting proteins between the two proteins' interaction vectors in a given PG network matrix . The specific bits score was calculated using the following formula: s ( p1 , p2 ) = b1·[−log ( b1/ ( b1+b2 ) ) ] , where p1 and p2 are the two proteins compared , b1 is the number of shared interacting proteins , and b2 is the number of non-shared interacting proteins between the two compared proteins in the PG networks . Congruence is a similarity measure between pairs of protein interacting vectors that was calculated as described in Lehner [53] . Bits and specific bits scores were calculated for the yeast and human PG networks; whilst congruence calculation was only performed for yeast since the size of the human PG0 . 014 network matrix ( 13 , 961×13 , 961 , see Table S5 in Text S1 ) was too large to make it feasible to implement the combinatorial calculation implicit in the congruence measure . Second order predictions were ranked based on the different similarity score values ( see section above ) from the most significant to the least significant . Validation was performed using as true positives ( TP ) protein pairs from the KG matrices in yeast and human respectively ( Int , Goss , Foss and Kegg in Yeast and Goss , Foss , Kegg , Int , Reactom_Int , and Reactome in Human; see the section above: Knowledgegram ( KG ) construction ) mapped to pairs in the ranked lists . An extra gold standard dataset of mapped true positive hits was built using those pairs present in two or more KG datasets ( KG≥2 ) . False positive ( FP ) sets were obtained by mapping the same KG gold standard datasets on randomised lists of second order predictions ranked lists , with 1 , 000 random iterations in yeast and 500 in human ( fewer times in human balancing the sample size against computational cost ) . Precision and recall parameters were calculated as described above , the precision mean and error ( standard deviation ) values were calculated based on the TP and the different accumulated random FP distributions . In order to present representative results values with standard deviations more than 1/3 of the mean were ignored , as they were due to the small size of the TP and FP samples at the beginning of the accumulated distributions ( for further details see section 18 in Text S1 ) . | To model accurate protein networks we need to extend our knowledge of protein associations in molecular systems much further . Biologists believe that high-throughput experiments will fill the gaps in our knowledge . However , if these approaches perform biased screenings , leaving important areas poorly characterized , success in modelling protein networks will require additional approaches to explore these ‘dark’ areas . We assess the value of integrating bio-computational approaches to build accurate and comprehensive network models for human and yeast proteomes and compare these models with models derived by combining multiple experimental datasets . We show that the predicted networks resemble the topological and error features of the experimental networks , and contain information on true protein associations within and beyond their constitutive first order binary predictions . We suggest that the majority of predicted network space is dark matter containing important functional areas , elusive to current experimental designs . Until novel experimental designs emerge as effective tools to screen these hidden regions , computational predictions will be a valuable approach for exploring them . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"computational",
"biology/systems",
"biology",
"computational",
"biology/genomics",
"computational",
"biology"
] | 2010 | Finding the “Dark Matter” in Human and Yeast Protein Network Prediction and Modelling |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.