A molecularly cloned, live-attenuated japanese encephalitis vaccine SA14-14-2 virus_ a conserved sin

EncephalitisVaccineSA14-14-2Virus:AConservedSingleAminoAcidintheijHairpinoftheViralEGlycoproteinDeterminesNeurovirulenceinMice

Sang-ImYun1.,Byung-HakSong1.,Jin-KyoungKim1,Gil-NamYun2,Eun-YoungLee3,LongLi4,RichardJ.Kuhn4,MichaelG.Rossmann4,JohnD.Morrey1,Young-MinLee1*

1DepartmentofAnimal,Dairy,andVeterinarySciences;UtahScienceTechnologyandResearch,CollegeofAgricultureandAppliedSciences,UtahStateUniversity,Logan,Utah,UnitedStatesofAmerica,2DepartmentofMicrobiology,CollegeofMedicine,ChungbukNationalUniversity,Cheongju,RepublicofKorea,3DepartmentofAnatomy,CollegeofMedicine,ChungbukNationalUniversity,Cheongju,RepublicofKorea,4DepartmentofBiologicalSciences,PurdueUniversity,WestLafayette,Indiana,UnitedStatesofAmerica

Abstract

Japaneseencephalitisvirus(JEV),amosquito-borneflavivirusthatcausesfatalneurologicaldiseaseinhumans,isoneofthemostimportantemergingpathogensofpublichealthsignificance.JEVrepresentstheJEserogroup,whichalsoincludesWestNile,MurrayValleyencephalitis,andSt.Louisencephalitisviruses.Withinthisserogroup,JEVisavaccine-preventablepathogen,butthemolecularbasisofitsneurovirulenceremainsunknown.Here,weconstructedaninfectiouscDNAofthemostwidelyusedlive-attenuatedJEvaccine,SA14-14-2,andrescuedfromthecDNAamolecularlyclonedvirus,SA14-14-2MCV,whichdisplayedinvitrogrowthpropertiesandinvivoattenuationphenotypesidenticaltothoseofitsparent,SA14-14-2.Toelucidatethemolecularmechanismofneurovirulence,weselectedthreeindependent,highlyneurovirulentvariants(LD50,,1.5PFU)fromSA14-14-2MCV(LD50,.1.56105PFU)byserialintracerebralpassageinmice.Completegenomesequencecomparisonrevealedatotalofeightpointmutations,withacommonsingleG1708RAsubstitutionreplacingaGlywithGluatposition244oftheviralEglycoprotein.UsingourinfectiousSA14-14-2cDNAtechnology,weshowedthatthissingleGly-to-GluchangeatE-244issufficienttoconferlethalneurovirulenceinmice,includingrapiddevelopmentofviralspreadandtissueinflammationinthecentralnervoussystem.Comprehensivesite-directedmutagenesisofE-244,coupledwithhomology-basedstructuremodeling,demonstratedanovelessentialregulatoryroleinJEVneurovirulenceforE-244,withintheijhairpinoftheEdimerizationdomain.Inbothmouseandhumanneuronalcells,wefurthershowedthattheE-244mutationalteredJEVinfectivityinvitro,indirectcorrelationwiththelevelofneurovirulenceinvivo,buthadnosignificantimpactonviralRNAreplication.OurresultsprovideacrucialsteptowarddevelopingnoveltherapeuticandpreventivestrategiesagainstJEVandpossiblyotherencephaliticflaviviruses.

Citation:YunS-I,SongB-H,KimJ-K,YunG-N,LeeE-Y,etal.(2014)AMolecularlyCloned,Live-AttenuatedJapaneseEncephalitisVaccineSA14-14-2Virus:AConservedSingleAminoAcidintheijHairpinoftheViralEGlycoproteinDeterminesNeurovirulenceinMice.PLoSPathog10(7):e1004290.doi:10.1371/journal.ppat.1004290

Editor:R.MarkL.Buller,SaintLouisUniversity,UnitedStatesofAmericaReceivedAugust20,2013;AcceptedJune18,2014;PublishedJuly31,2014

Copyright:ß2014Yunetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited.

Funding:ThisworkwassupportedbyNationalInstitutesofHealthgrantAI1014,UtahScienceTechnologyandResearchfunds,andNationalResearchFoundationgrant532-2008-1-C00011(toSIY)fundedbytheKoreanMinistryofEducation,Science,andTechnology.YMLwassupportedinpartbytheNationalResearchFoundationofKorea(2007-0052563).Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofthemanuscript.

CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.*Email:youngmin.lee@usu.edu

.Theseauthorscontributedequallytothiswork.

Introduction

Japaneseencephalitisvirus(JEV)isthemostcommoncauseofviralencephalitisinAsiaandpartsoftheWesternPacific,with,60%oftheworld’spopulationatriskofinfection[1].WithinthefamilyFlaviviridae(genusFlavivirus),JEVbelongstotheJEserologicalgroup,whichalsoincludesmedicallyimportanthumanpathogensfoundoneverycontinentexceptAntarctica[2,3]:WestNilevirus(WNV),St.Louisencephalitisvirus(SLEV),andMurrayValleyencephalitisvirus(MVEV).Historically,theJEserologicalgroupmembershaveclusteredingeographicallydistinctlocations,buttherecentemergenceandspreadofJEVinAustralia[4]and

PLOSPathogens|www.plospathogens.org

1

WNVinNorthAmerica[5,6]havecausedgrowingconcernthatthesevirusescanspreadintonewterritory,posingasignificantchallengeforglobalpublichealth[3,7].IntheUS,whereWNVandSLEVareendemic,thesituationisparticularlyproblematicbecausethelikelihoodofJEVbeingintroducedisconsiderable[8,9].Worldwide,,50,000–175,000clinicalcasesofJEareestimatedtooccurannually[10];however,thisincidenceisundoubtedlyaconsiderableunderestimatebecausesurveillanceandreportingareinadequateinmostendemicareas,andonly,0.1–4%ofJEV-infectedpeopledevelopclinicaldisease[11,12].Onaverage,,20–30%ofpatientsdie,and,30–50%ofsurvivorssufferfromirreversibleneurologicaland/orpsychiatricsequelae

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

AuthorSummary

Agroupofmosquito-borneflavivirusesthatcausefatalencephalitisinhumansisamongthemostimportantofallemerginghumanpathogensofglobalsignificance.ThisgroupincludesJapaneseencephalitis(JE),WestNile,St.Louisencephalitis,andMurrayValleyencephalitisviruses.Inthiswork,wehavedevelopedareversegeneticssystemforSA14-14-2,aliveJEvaccinethatismostcommonlyusedinJE-endemicareas,byconstructinganinfectiousbacterialartificialchromosomethatcontainsthefull-lengthSA14-14-2cDNA.UsingthisinfectiousSA14-14-2cDNA,combinedwithamousemodelforJEVinfection,wehaveidentifiedakeyviralneurovirulencefactor,aconservedsingleaminoacidintheijhairpinadjacenttothefusionloopoftheviralEglycoprotein,whichregulatesviralinfectivityintoneuronswithinthecentralnervoussysteminvivoandneuronalcellsofmouseandhumaninvitro.Thus,ourfindingselucidatethemolecularbasisoftheneuroviru-lencecausedbyJEVandothercloselyrelatedencephaliticflaviviruses,amajorstepinunderstandingtheirneuro-pathogenesis.Fromaclinicalperspective,thediscoveryoftheviralneurovirulencefactoranditsrolewillhavedirectapplicationtothedesignofanovelclassofbroad-spectrumantiviralstotreatandpreventinfectionofJEVandothertaxonomicallyrelatedneurotropicflaviviruses.[13].Mostclinicalcasesoccurinchildrenunderage15inendemicareas,butinnewlyinvadedareas,allagegroupsareaffectedbecauseprotectiveimmunityisabsent[14].Thus,giventhecurrentdiseaseburdenandsignificantthreatoftheJEVemergence,resurgence,andspreadamongmuchlargergroupsofsusceptiblepopulations,controlofJEVremainsahighpublichealthpriority.

JEVcontainsanucleocapsidcomposedofan,11-kbplus-strandgenomicRNA,complexedwithmultiplecopiesofthehighly-basica-helicalCproteins[15,16].Thenucleocapsidissurroundedbyahost-derivedlipidbilayercontainingthemembrane-anchoredMandEproteins[17–19].Theinitialstepintheflavivirusreplicationcycleinvolvesattachmentofthevirionstothesurfaceofsusceptiblecells[20–24].TheviralEproteinisthenassumedtobindwithhighaffinityandspecificitytoanas-yetunidentifiedcellularreceptor(s),whichtriggersreceptor-mediated,clathrin-dependentendocytosis[25–27].TheacidicconditionsintheendosomeleadtoaconformationalchangeintheEprotein[28–32],whichtriggersfusionoftheviralmembranewithhostendosomalmembrane[33].Oncethegenomeisreleasedintothecytoplasm,thegenomicRNAistranslatedintoasinglepolyprotein,whichisprocessedco-andpost-translationallybyhostandviralproteasestoyieldatleast10functionalproteins[34]:threestructural(C,prM,andE)andsevennonstructural(NS1,NS2A,NS2B,NS3,NS4A,NS4B,andNS5).ThenonstructuralproteinsactivelyreplicatetheviralgenomicRNAinthereplicationcomplex[35–38]thatisassociatedwiththevirus-induced,ER-derivedmembranes[39–41].NewlysynthesizedgenomicRNAandCproteinsareinitiallyenvelopedbytheprMandEproteinstogenerateimmaturevirions[42,43]thatbudintothelumenoftheER[44].TheseimmaturevirionsarethentransportedviathesecretorypathwaytotheGolgiapparatus.Inthelow-pHenvironmentofthetrans-Golginetwork,thefurin-mediatedcleavageofprMtoMinducesthematurationoftheviralparticles[45],whichisalsoaccompaniedbysignificantstructuralrearrangementsoftheMandEproteins[42,46,47].Finally,maturevirionsarereleasedintotheextracellularspacebyexocytosis.

PLOSPathogens|www.plospathogens.org

2

JEVismaintainedinanenzooticcycleinvolvingmultiplespeciesofmosquitovectors(primarilyCulexspecies)andvertebratehosts/reservoirs(mainlydomesticpigs/wadingbirds).Humansbecomeinfectedincidentallywhenbittenbyaninfectedmosquito[48].Intheabsenceofantiviraltherapy,activeimmunizationistheonlystrategyforsustainablelong-termprotection.FourtypesofJEvaccinesareusedindifferentpartsoftheworld[49,50]:(i)themousebrain-derivedinactivatedvaccinebasedontheNakayamaorBeijing-1strain,(ii)thecellculture-derivedinactivatedvaccinebasedontheBeijing-3orSA14-14-2strain,(iii)thecellculture-derivedlive-attenuatedvaccinebasedontheSA14-14-2strain,and(iv)thelivechimericvaccinedevelopedonayellowfevervirus(YFV)17DgeneticbackgroundthatcarriestwosurfaceproteinsofJEVSA14-14-2.Ofthefourvaccines,theonlyonethatisavailableinternationallyisthemousebrain-derivedinactivatedNakayama[11].Unfortu-nately,theproductionofthisvaccinewasdiscontinuedin2006[51]becauseofvaccine-relatedadverseevents,short-termimmunity,andhighproductioncost[13,52].Todate,themostcommonlyusedvaccineinAsiaisthelive-attenuatedSA14-14-2[53],butthisvaccineisnotrecommendedbytheWHOforglobalimmunization[13,54].Inadditiontothedependenceofthedurationofimmunityonthenumberofdosesreceived,thereisatleastatheoreticalriskofvirusmutationandreversionofthevaccinevirustohighvirulence.Recently,theSA14-14-2vaccinevirushasbeenutilizedtoproduceanewVerocell-derivedinactivatedvaccinethathasbeenapprovedintheUS,Europe,Canada,andAustraliasince2009[51,55,56].IntheUS,thisvaccineisrecommendedforadultsaged$17yearstravellingtoJEV-endemiccountriesandatriskofJEVexposure[51,57],butnovaccineiscurrentlyavailableforchildrenunder17[58].Morerecently,theprMandEgenesofJEVSA14-14-2havebeenusedtoreplacethecorrespondinggenesofYFV17D[59],creatingalivechimericvaccine[60]thatisnowlicensedinAustraliaandThailand[61,62].Thus,theapplicationofJEVSA14-14-2tovaccinedevelopmentandproductioniscontinuouslyexpanding,buttheviralfactorsandfundamentalmechanismsresponsibleforitslossofvirulencearestillelusive.

ThevirulenceofJEVisdefinedbytwoproperties:(i)neuroinvasiveness,theabilityofthevirustoenterthecentralnervoussystem(CNS)wheninoculatedbyaperipheralroute;and(ii)neurovirulence,theabilityofthevirustoreplicateandcausedamagewithintheCNSwheninoculateddirectlyintothebrainofahost.Overthepast20years,manyinvestigatorshavesoughttounderstandthemolecularbasisofJEVvirulence,byusingcellandanimalinfectionmodelsystemstocomparethenucleotidesequencesofthegenomesofseveralJEVstrainsthatdifferinvirologicalproperties[63–74].Thesestudieshaveidentifiedalargenumberofmutationsscatteredessentiallythroughouttheentireviralgenome.Becauseofthecomplexityofthemutations,however,themajorgeneticdeterminant(s)criticalforeitherJEVneurovirulenceorneuroinvasivenessremainsunclear.Inparticu-lar,thesituationismorecomplicatedforthelive-attenuatedSA14-14-2virus,whichhasbeenreportedtohaveanumberofdifferentmutations,i.e.,47–nucleotidechanges(17–27aminoacidsubstitutions),whencomparedtoitsvirulentparentalstrainSA14;theexactnumberdependsonboththepassagehistoryofthevirusesandthetypeofcellsubstrateusedforviruscultivation[63–65].AmorecomprehensivesequencecomparisonwithanotherSA14-derivedattenuatedvaccinestrain,SA14-2-8,togetherwithtwoothervirulentstrains,hassuggestedsevencommonaminoacidsubstitutionsthatmaybeinvolvedinvirusattenuation:4inE,1inNS2B,1inNS3,and1inNS4B[].However,thegeneticcomponentdirectlyresponsiblefortheattenuationofSA14-14-2is

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

stillunknown.GiventhatSA14-14-2hasbeenadministeredto.300millionchildrenfor.20yearsinChinaandrecentlyinotherAsiancountries[53],itisstrikingthatthereisafundamentalgapinourknowledgeatthemolecularlevelabouthowSA14-14-2isattenuated.

HerewereportthedevelopmentofaninfectiouscDNA-basedreversegeneticssystemforJEVSA14-14-2thathasenabledtheanalysisofmolecularaspectsofitsattenuationinneurovirulence.Byinvivopassageofamolecularlydefined,cDNA-derivedSA14-14-2virus,wegeneratedthreeisogenicvariants,eachdisplayinglethalneurovirulenceinmice,withacommonsingleG1708RAsubstitutionthatcorrespondstoaGlyRGluchangeatposition244oftheviralEglycoprotein.Byinvitrosite-directedmutagenesisoftheinfectiousSA14-14-2cDNA,coupledwithconventionalvirologicandexperimentalpathologicmethodsandhomology-basedstruc-turemodeling,wehavedemonstratedanovelregulatoryroleinJEVneurovirulenceofaconservedsingleaminoacidatpositionE-244intheijhairpinadjacenttothefusionloopoftheEdimerizationdomain.ThesefindingsoffernewinsightsintothemolecularmechanismofJEVneurovirulenceandwilldirectlyaidthedevelopmentofnewapproachestotreatingandpreventingJEVinfection.

Results

Developmentofafull-lengthinfectiouscDNAofSA14-14-2,aliveJEvaccinevirus

AsaninitialstepininvestigatingthemolecularbasisforthevirulenceattenuationofSA14-14-2,wegeneratedafull-lengthinfectiousSA14-14-2cDNAtoserveasatemplateforgeneticmanipulationoftheviralgenome(Fig.1A).The10,977-nucleotidegenomeofSA14-14-2(GenBankaccessionnumberJN604986)wasfirstclonedasfourcontiguouscDNAsintothebacterialartificialchromosome(BAC)designatedpBAC/Frag-ItoIV(Fig.1B).pBAC/Frag-IwasmodifiedtohaveanSP6promoterimmediatelyupstreamoftheviral59-end,andpBAC/Frag-IVwasengineeredtocontainanartificialXbaIrun-offsitejustdownstreamoftheviral39-end,allowinginvitrorun-offtranscriptionofcapped,genome-lengthRNAsbearingauthentic59and39endsofthegenomicRNA.SincetheviralgenomealreadyhadaninternalXbaIsiteatnucleotide9131intheNS5protein-codingregion,thispre-existingsitewaseliminatedinpBAC/Frag-IIIbyintroducingasilentpointmutation,A9134RT(Fig.1B,asterisk),whichinturnservedasageneticmarkertoidentifythecDNA-derivedvirus.Inthelastcloningstep,apanelofthefouroverlappingSA14-14-2cDNAswassequentiallyassembledbyjoiningatthreenaturalrestrictionsites(BsrGI,BamHI,andAvaI)tocreatethefull-lengthSA14-14-2cDNA,pBAC/SA14-14-2(Fig.1C).

ThefunctionalityofpBAC/SA14-14-2wasanalyzedbydeterminingthespecificinfectivityofthesyntheticRNAstranscribedinvitrofromthecDNAafterRNAtransfectionintosusceptibleBHK-21cells(Fig.1D).TwoindependentclonesofpBAC/SA14-14-2werelinearizedbyXbaI,followedbymungbeannucleasetreatmenttoremovethe59overhangleftbytheXbaIdigestion.EachwasthenusedasatemplateforSP6polymeraserun-offtranscriptioninthepresenceofthem7G(59)ppp(59)Acapstructureanalog.TransfectionofthesyntheticRNAsintoBHK-21cellsgavespecificinfectivitiesof6.0–7.56105PFU/mg;thevirustitersrecoveredfromtheRNA-transfectedcellswere3.0–4.56105PFU/mlat22hpost-transfection(hpt)andincreased,10-foldto2.9–3.76106PFU/mlat40hpt(Fig.1D).Unequivocally,therecoveredviruscontainedthemarkermutation(A9134RT)thathadbeenintroducedinpBAC/SA14-14-2(datanotshown).OurresultsshowthatthesyntheticRNAsgeneratedfromthefull-length

PLOSPathogens|www.plospathogens.org

3

SA14-14-2cDNAarehighlyinfectiousinBHK-21cells,producingahightiterofmolecularlydefined,infectiousvirus.

Incellcultures[75,76],weassessedtheinvitrogrowthpropertiesofthemolecularlyclonedvirus(SA14-14-2MCV)rescuedfromtheinfectiouscDNA,ascomparedtothoseoftheunclonedparentalvirus(SA14-14-2)usedforcDNAconstruction.InhamsterkidneyBHK-21cells,whichareusedmostfrequentlyforJEVpropagationinlaboratories,SA14-14-2MCVreplicatedasefficientlyasSA14-14-2,withnonoticeabledifferenceintheaccumulationofviralgenomicRNA(Fig.2A)andproteins(Fig.2B)overthefirst24hafterinfectionatamultiplicityofinfection(MOI)of1plaque-formingunit(PFU)percell.Theseobservationswereconsistentwiththeirgrowthkinetics,whichwereessentiallyidenticalfor4daysfollowinginfectionatthreedifferentMOIs:0.1,1,and10PFU/cell(Fig.2Canddatanotshown).Similarly,therewasnodifferenceinfocus/plaquemorphologybetweenSA14-14-2MCVandSA14-14-2at4dayspost-infection(dpi)(Fig.2D);asexpected,theirfoci/plaqueswere,30%smallerthanthoseproducedbyCNU/LP2,avirulentJEVstrainusedasareference(Fig.S1).Also,theirgrowthpropertieswereequivalentintwoothercelllines,humanneuroblastomaSH-SY5YandmosquitoC6/36cells,whicharepotentiallyrelevanttoJEVpathogenesisandtransmission,respectively(Fig.S2).Thesedatasuggestthattheunclonedparentalandmolecularlyclonedvirusesareindistinguishableinviralreplicationandspreadinbothmammalianandinsectcells.

Inmice[75,77],weevaluatedinvivotheattenuationphenotypesofSA14-14-2MCVandSA14-14-2,withavirulentJEVCNU/LP2[78]inparallel.Groupsof3-week-oldICRmice(n=20)wereinfectedwithvariousdoses(1.5to1.56105PFU/mouse)ofeachvirus,viathreedifferentinoculationroutes:intracerebral(IC)forneurovirulence,andintramuscular(IM)andintraperitoneal(IP)forneuroinvasiveness.AswithSA14-14-2,the50%lethaldoses(LD50s)ofSA14-14-2MCV,regardlessoftherouteofinoculation,wereall.1.56105PFU(Figs.2EandS3).Specifically,allmiceinfectedwithSA14-14-2MCVorSA14-14-2remainedhealthyanddisplayednoclinicalsignsofJEVinfection(e.g.,ruffledfur,hunchedposture,tremors,orhindlimbparalysis)afterIMorIPinoculationwithanyofthetesteddoses;ontheotherhand,asmallfractionofthemiceinfectedwithSA14-14-2MCV(5–20%)orSA14-14-2(5–10%)developedtypicalsymptomsanddeathaftertheICinoculationwitharelativelyhighdoseof$1.56103PFU/mouse,butnotthelowdoseof#1.56102PFU/mouse(Fig.S3).Inalldeadorsurvivingmice,virustitrationconfirmedthepresence(1.8–4.16106PFU/brain)orabsence,respectively,ofviralreplicationintheirbraintissues.Asexpected[75,77],theLD50valuesofCNU/LP2[78],irrespectiveoftheinoculationroute,werealways,1.5PFU(Figs.2EandS3);thecontrolgroupsofmock-infectedmiceallsurvivedwithnosignsofdisease(Fig.S3).Thus,ourdataindicatethatSA14-14-2MCVdisplaysavarietyofbiologicalpropertiesidenticaltothoseofSA14-14-2,bothinvitroandinvivo.

GenerationofthreehighlyneurovirulentvariantsderivedfromSA14-14-2MCVAswastrueforSA14-14-2,directinoculationofarelativelyhighdoseofSA14-14-2MCVintomousebrainsinitiatedaproductiveinfectionintheCNSandcausedlethalencephalitis,albeitataverylowfrequency(Fig.S3).Intriguedbythisobservation,wedecidedtogenerateisogenicneurovirulentvariantsfromSA14-14-2MCVbyserialbrain-to-brainpassageinmice(Fig.3A).Atpassage1(P1),thecDNA-derivedSA14-14-2MCVwasdirectlyinoculatedintothebrainsof3-week-oldICRmiceat1.56105PFU/mouse(threegroups,n=10pergroup);oneortwoinfectedmicepergroupexhibitedclinicalsymptomsofJEVinfection.Attheonsetofhindlimbparalysis

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure1.Constructionofthefull-lengthinfectiousSA14-14-2cDNAasaBAC.(A)StructureoftheSA14-14-2genomicRNA(GenBankaccessionno.JN604986).NCR,non-codingregion;ORF,openreadingframe.(B)DiagramofapaneloffouroverlappingSA14-14-2cDNAscontainedinpBAC/Frag-ItoIV.SP6promoterandanartificialrun-offsiteareshown.Anasteriskindicatesapre-existingXbaIsiteatnucleotide9131thatwasinactivatedbyintroducingasilentpointmutation,A9134RT.(C)Structureofthefull-lengthSA14-14-2cDNA(pBAC/SA14-14-2).(D)FunctionalityofpBAC/SA14-14-2.AfterinvitrotranscriptionwithSP6RNApolymerase,RNAtranscriptswereelectroporatedintoBHK-21cells,andinfectiousplaquecentersweredetermined(RNAinfectivity).At22and40hpt,supernatantsfromRNA-transfectedcellswereharvestedforvirustitrationonBHK-21cells(Virusyield).

doi:10.1371/journal.ppat.1004290.g001

(6–10dpi),viruswasharvestedfromthebrainofamoribundmouseineachgroup(3total);ineachcase,abrainhomogenatewaspreparedforplaquetitrationandusedasaninoculumforthenextroundofpassage.Serialintracerebralpassagewascontinuedforthreeadditionalrounds,withagraduallydecreasinginoculuminordertoensurethestabilityofselectedmutationsandasufficientlypurepopulationofviruses:1,500(P2),to15(P3),to1.5PFU/mouse(P4).Usingthisapproach,weobtainedthreeindependentlyselectedvariants,SA14-14-2MCV/V1toV3(Fig.3A).

WefirstcomparedthebiologicalpropertiesofthethreeSA14-14-2MCVvariants,bothinvitroandinvivo,tothoseoftheparentalSA14-14-2MCV.Inthreecellcultures(BHK-21,SH-SY5Y,andC6/36),allthreevariantsexhibitedcharacteristicsofviralreplicationidenticaltotheparent,asdemonstratedby(i)quantitativereal-timeRT-PCRstomeasurethelevelofviralgenomicRNAproduction,(ii)immunoblottingwithapanelofJEV-specificrabbitpolyclonalantiseratoprobetheprofileandlevelofviralstructuralandnonstructuralproteinaccumulation,and(iii)one-stepgrowthanalysestoassesstheyieldofprogenyvirionsproducedduringasingleroundofinfection(datanotshown).In3-week-oldICRmice,however,therewasacleardifferencebetweentheparentandthethreevariantsinbothphenotypeandvirulencelevel(Fig.3B).Whenperipherallyinoculated(i.e.,IMandIP),neithertheparentnoritsthreevariantscausedanysymptomsordeathatamaximumdoseof1.56105PFU/mouse.Incontrast,wheninoculatedIC,thethreevariants,unliketheparent(ICLD50,.1.56105PFU),wereallhighlyneurovirulent(ICLD50s,,1.5PFU)(Fig.3BandTableS1).OurfindingsshowthatallthreevariantsstilllackedadetectablelevelofneuroinvasivenessbutgainedahighlevelofneurovirulenceafterserialICpassageinmice.

PLOSPathogens|www.plospathogens.org

4

Next,wedeterminedthecompletenucleotidesequenceofthegenomeofthethreeSA14-14-2MCVvariantstoidentifythenucleotide(s)and/oraminoacid(s)inspecificviralloci/genesthatis(are)potentiallyresponsibleforthedrasticincreaseinneurovir-ulence.Accordingtoourprotocol[77],theconsensusgenomesequenceofeachvariantwasgeneratedbydirectsequencingofthreeoverlapping,unclonedcDNAampliconscoveringtheentireviralRNAgenomeexceptthe59-and39-termini;theremainingconsensussequencesofthe59-and39-terminalregionswereobtainedby59-and39-RACEreactions,eachfollowedbycDNAcloningandsequencingof10–15independentclones.Inallthreevariants,whentheconsensusgenomesequencewascomparedtothatoftheparent,asinglenucleotideG-to-Atransitionwasalwaysfoundatnucleotide1708,changingaGly(GGG)toGlu(GAG)codonataminoacid244oftheviralEglycoprotein(Fig.3C).Inaddition,eachofthethreevariantsalsocontainedasmallnumberofuniquesilentpointmutationsscatteredoverthegenome,confirmingtheywereindeedindependentvariants(Fig.3C):oneinSA14-14-2MCV/V1(U2580C),twoinSA14-14-2MCV/V2(G317AandU8588C),andfourinSA14-14-2MCV/V3(U419C,C3215U,C5987U,andG6551A).TheseresultssuggestthattheG1708Asubstitution,theonlymutationobservedinallthreevariants,maycontributetotheviralneurovirulenceinmice.

IdentificationofasingleGly-to-GluchangeatpositionE-244thatisresponsibleforthereversiontoneurovirulence

Toidentifyakeypointmutation(s)inthreevariantsofSA14-14-2thatleadstotheacquisitionofneurovirulence,wegeneratedeightderivativesofSA14-14-2MCV,eachcontainingoneofthe

MCVJuly2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure2.CharacterizationofbiologicalpropertiesofthemolecularlyclonedvirusSA14-14-2MCVinvitroandinvivo.(A-D)BHK-21cellsweremock-infectedorinfectedatanMOIof1withthemolecularlyclonedvirus(SA14-14-2MCV)ortheoriginalparentalvirus(SA14-14-2).Atthetimepointsindicated,cellswerelysedtoanalyzetheaccumulationlevelsofviralgenomicRNAbyNorthernblotting(A)andviralproteinsbyimmunoblotting(B),andculturesupernatantswereharvestedtoexaminetheproductionlevelsofprogenyvirionsbyplaquetitrationonBHK-21cells(C).At4dpi,cellmonolayerswerefirstimmunostainedwithamousea-JEVantiserumtovisualizetheinfectiousfoci,andthesamemonolayerswerethenrestainedwithcrystalviolettoobservetheinfectiousplaques(D).Theaverageplaquesizes(mean6SD)wereestimatedbycounting10representativeplaques.(E)Groupsof3-week-oldICRmice(n=20pergroup)wereinfectedIC,IM,orIPwithserial10-folddilutionsofeachvirusasindicated.TheLD50values(inPFU)werecalculatedbytheReedandMuenchmethod[119].CNU/LP2,avirulentJEVstrainusedasareference.doi:10.1371/journal.ppat.1004290.g002

eightpointmutationsfoundinourthreevariants,bycloningthemindividuallyintotheinfectiousSA14-14-2cDNAandtransfectingthesyntheticRNAsderivedfromeachmutantcDNAintoBHK-21cells.Inallcases,themutantRNAwasasinfectiousastheparentRNA,withaspecificinfectivityof6.4–8.36105PFU/mg;thesizesofthefoci/plaquesproducedbyeachmutantRNAwereindistinguishablefromthosegeneratedbytheparentRNA,parallelingtheirlevelsofvirusproduction,withanaverageyieldof2.1–4.56105PFU/mlat22hpt(Fig.4A).Inagreementwiththeseresults,nodifferencewasobservedintheprofileorexpressionleveloftheviralproteins,i.e.,threestructural(C,prM,andE)andonenonstructural(NS1),asdeterminedbyimmunoblottingofRNA-transfectedcellsat18hpt(Fig.4B).Allthemutantvirusesgrewasefficientlyasdidtheparentalvirusoverthecourseof96hafterinfectionatanMOIof0.1inBHK-21cells(Fig.4C).Thus,therewasnoapparenteffectofanyoftheeightintroducedgeneticchangesonvirusreplication.

Inmice,weexaminedtheneurovirulenceoftheseeightmutantviruses.Groupsof3-week-oldICRmice(n=10pergroup)were

PLOSPathogens|www.plospathogens.org

5

infectedbyICinoculationwithvariousdoses(1.5to1.56105PFU/mouse)oftheparentoreachmutantvirus.OneoftheeightmutantscontainingtheG1708AsubstitutionhadanICLD50of,1.5PFU,makingitcapableofkillingallmicewithin,7dpiwithaminimumdoseof1.5PFU/mouse;theothersevenmutantshadICLD50valuesall.1.56105PFUandbehavedliketheparentalvirus,withonly,20%ofinfectedmicedevelopingclinicalsymptomsanddeathatamaximumdoseof1.56105PFU/mouse(Fig.4DandTableS2).Inalldeadorsurvivingmice,virustitrationconfirmedthepresence(1.4–3.56106PFU/brain)orabsence,respectively,ofproductiveviralreplicationinthebraintissues;asexpected,allmock-infectedmicesurvivedwithnosignsofdisease(datanotshown).Thus,ourfindingsshowedthatoftheeightpointmutations,asingleG1708Asubstitution,replacingaGlywithGluataminoacidresidue244oftheviralEglycoprotein,issufficienttoconferlethalneurovirulenceinmice.

TodeterminewhetherthemutantG1708A,unliketheparentSA14-14-2MCV,isabletoreplicateandspreadintheCNS,weimmunohistochemicallystainedforJEVNS1antigeninmouse

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure3.Developmentofthreeindependent,highlyneurovirulentvariantsfromSA14-14-2MCVbyserialintracerebralpassageinmice.(A)DiagramillustratingtheinvivopassageofSA14-14-2MCV.(B)VirologicalpropertiesofthreeSA14-14-2MCVvariantsinmice.Groupsof3-week-oldICRmice(n=10pergroup)wereinoculatedIC,IM,orIPwithdosesofthevirusstockseriallydiluted10-fold.TheLD50valuesweredeterminedinPFU.(C)ComparisonofthecompletegenomesequenceoftheSA14-14-2MCVparentalvirusanditsthreevariantviruses.doi:10.1371/journal.ppat.1004290.g003

brainsafterICinoculation(Fig.4Eshowshippocampalslides,andFig.S4presentsslidesofotherbrainareas,i.e.,amygdala,cerebralcortex,thalamus,hypothalamus,andbrainstem):(i)InbrainsinfectedwithavirulentJEVCNU/LP2(control)[75,77,78],alargenumberofNS1-positiveneuronswereobservedat3dpiinallareaswestained;thisnumberwasincreasedsignificantlyat5dpi.Inthehippocampus,mostinfectedneuronswerefoundintheCA2/3regionat3dpiandhadspreadtotheCA1regionby5dpi.(ii)InbrainsinfectedwiththeparentSA14-14-2MCV,almostnoNS1-positivecellswerefoundinanybrainregionduringtheentire7-daycourseoftheexperiment.Inafewatypicalcases,asmallnumberofNS1-positiveneuronswerenotedat5–7dpiinthehippocampalCA2/3region,butnottheCA1region(datanotshown).(iii)InbrainsinfectedwiththemutantG1708A,aconsiderablenumberofNS1-positiveneuronswereobservedat3dpi,mainlyinthehippocampalCA2/3region,andonlyafewinotherareas(amygdala,cerebralcortex,thalamus,andbrainstem);overall,thenumberofinfectedneuronswasmuchlowerthaninbrainsinfectedwithJEVCNU/LP2.At5–7dpi,thenumberofNS1-positiveneuronswasnoticeablyincreasedinthehippocam-pus(nowintheCA1)andamygdala,butnotinotherbrainregions.Ourfindingsshowthat,inmice,asingleG1708AsubstitutionchangingaGlywithGluatpositionE-244promotessusceptibilitytoSA14-14-2MCVinfectionofneurons.

Understandingthenovelregulatoryrolein

neurovirulenceofE-244,locatedintheijhairpinoftheviralEglycoprotein

ToprobethefunctionalimportanceoftheaminoacidsidechainatpositionE-244fortheviralreplicationandneuroviru-lenceofSA14-14-2MCV,weperformedsite-directedmutagenesis,replacingG244with14otheraminoacidsofsixdifferentclasses:(1)aliphaticA,V,andL;(2)hydroxylSandT;(3)cyclicP;(4)aromaticFandW;(5)basicRandK;and(6)acidicandtheiramidesD,E,N,andQ.WefirsttestedtheviabilityofsyntheticRNAstranscribedinvitrofromthecorrespondingmutantcDNAs

PLOSPathogens|www.plospathogens.org

6

bymeasuringtheirinfectivityaftertransfectionofBHK-21cells.Inallcases,themutantRNAwasasviableastheparentRNA,withaspecificinfectivityof6.5–8.26105PFU/mg(Fig.5A,RNAinfectivity).However,threemutants(G244K,G244F,andG244W)werenoticeablydifferentfromtheparentandtheother11mutants,asdemonstratedbya,10-folddecreaseintheyieldofprogenyvirionsreleasedintoculturemediumduringthefirst22hpt(Fig.5A,virusyield)anda,2-2.5-foldreductioninthesizeoffoci/plaquesproducedat96hpt(Fig.5A,foci/plaques),althoughnosignificantdifferencewasobservedinthelevelofviralproteins(i.e.,C,prM,E,andNS1)accumulatedinRNA-transfectedcellsat18hpt(Fig.S5).AscomparedtoG244K,themutantG244Rexhibitedabarelymarginaldecreaseinfocus/plaquesizeandnodetectablechangeinvirusproduction(Fig.5A).Overall,thesefindingsweremoreevidentwhenallmutantviruseswereevaluatedinmultistepgrowthassaysoverthecourseof96hafterinfectionatanMOIof0.1,assessingtheirabilitytogrowandestablishaproductiveinfection(Fig.5B).OurfindingsindicatethatinBHK-21cells,theaminoacidsidechainatpositionE-244hasnoeffectontheviabilityofthemutantRNAs,althoughithasanegativeimpactontheproductionandspreadofinfectiousvirionsinthecaseofthethreemutantsG244K,G244F,andG244W.

Inmice,wedeterminedtheneurovirulenceofour14mutantvirusesbyICinoculatinggroupsof3-week-oldICRmice(n=10pergroup)withvariousdosesrangingfrom1.5to1.56104or105PFU/mouseoftheparentoreachmutantvirus.AccordingtotheirICLD50values,the14mutantvirusesareclassifiedintothreegroups(Fig.5CandTableS3):(i)group1(sixmutants),neurovirulent,withanICLD50of#1.5to31PFU,exemplifiedbyreplacingG244withE,D,T,S,Q,andP;(ii)group2(sixmutants),non-neurovirulentorneuroattenuated,withanICLD50of.1.56104or105PFU,behavingliketheparentSA14-14-2MCVandexemplifiedbyexchangingG244withR,K,F,W,N,andL;and(iii)group3(twomutants),withanintermediatephenotypeandanICLD50of1.2–5.86103PFU,exemplifiedbysubstitutingG244withAandV.Weconfirmedthepresenceorabsenceofviral

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure4.DiscoveryofasinglelocusthatleadstothereversionofSA14-14-2MCVtolethalneurovirulence.(A-C)Invitroreplicability.BHK-21cellsweremock-transfectedortransfectedwithRNAstranscribedfromtheparentoreachmutantcDNA,asindicated.RNAinfectivity(inPFU/mg)at4dptwasdeterminedbyinfectiouscenterassay,combinedwithstainingofcellmonolayersusingana-JEVantiserum,andvirusyield(inPFU/ml)at22hptwasmeasuredbyplaquetitrationonBHK-21cells(A);viralproteinaccumulationat18hptwasexaminedbyimmunoblottingofcelllysateswithapanelofantibodiesasindicated(B).Forviralgrowthanalysis,BHK-21cellswereinfectedatanMOIof0.1withtheparentoreachmutantvirusobtainedfromthecorrespondingRNA-transfectedcells.Attheindicatedtimepoints,culturesupernatantsharvestedforvirustitrationonBHK-21cells(C).(D-E)Invivoneurovirulence.Groupsof3-week-oldICRmice(n=10pergroup)wereinoculatedICwithdosesofthevirusstockseriallydiluted10-fold,andtheLD50valuesweredetermined(D).Forimmunohistochemicalstaining,groupsofthemice(n=15pergroup)weremock-infectedorinfectedICwith103PFUofeachvirus.At3,5,and7dpi,fivemicewereprocessedforbrainsectionstainingwithana-NS1antiserum.Shownarerepresentativehippocampalslides(E).doi:10.1371/journal.ppat.1004290.g004

replicationinthebraintissuesofalldeadorsurvivingmice,respectively;allmock-infectedmicesurvivedwithnosignsofdisease(datanotshown).Also,themutationandphenotyperelationshipwascorroboratedbysequenceanalysisofrecoveredvirusesfrombraintissuesofmoribundordeadmicefollowingICinoculation.Weanalyzedallofthe14mutantsexceptforfourgroup2mutants(G244R,G244F,G244W,andG244L),whichfailedtoproducealethalinfection.Ineachcase,thecomplete2,001-nucleotidecodingregionoftheprMandEgeneswasamplifiedfromeachoffourrandomlyselectedbrainsamples,followedbycloningandsequencingofatleastsevenindependentclonesperbrainsample.Inallsixgroup1andtwogroup3mutants,wefoundthattheinitialmutationsintroducedattheG244codonweremaintainedwithnosecond-sitemutations,consistentwiththehighandintermediatelevelsoftheirneurovirulentphenotype(Table1).Intheremainingtwogroup2mutants(G244KandG244N),however,amajorityofthesequencedclonescontainedapointmutationinthesamecodonthatledtoanaminoacidsubstitution(i.e.,KRE/TandNRD,respectively),convertingbothmutantsintoneurovirulentvirusesandhighlightingthebiologicalimportanceoftheaminoacidatpositionE-244forneurovirulence(Table1).

PLOSPathogens|www.plospathogens.org

7

WenextperformedhomologymodelingtogaininsightintothestructuralbasisofE-244function.The3DmodeloftheE

˚monomerofJEVSA14-14-2wasconstructedusingthe3.0-A

crystalstructureoftheEmonomerofWNVNY99[79]asatemplate,with75.5%sequenceidentity.Themodelwasthenfittedintotheouterlayerofthecryo-electronmicroscopy(EM)structureofWNVNY99[18],therebyvisualizingthreemonomersplacedintoanicosahedralasymmetricunitontheviralmembrane.IneachEmonomerofSA14-14-2containingthreedomains(DI,DII,andDIII),wenotedthatE-244lieswithintheijhairpinadjacenttothefusionloopatthetipofDII,withitsaminoacidsidechainpointingtowardtheviralmembrane(Fig.5D).WealsoconfirmedthelocationofE-244inthecrystalstructureoftheEectodomainofJEVSA14-14-2[80]thathasbeendescribedrecently(Fig.S6).

ElucidationofthefunctionalroleofE-244duringJEVinfectionofneuronalcellsinvitro

WehypothesizedthatE-244,locatedattheijhairpinoftheviralEglycoprotein,playsanimportantroleinJEVinfectionofneuronalcells.Totestthishypothesis,weperformedmultistep

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure5.E-244:AkeyneurovirulencefactorlocatedintheijhairpinoftheviralEglycoprotein.(A)RNAinfectivityandreplicability.BHK-21cellsweremock-transfectedortransfectedwithRNAstranscribedfromtheparentoroneofthe14E-244mutantcDNAs.RNAinfectivity(inPFU/mg)at4dptwasestimatedbyinfectiouscenterassay,coupledwithstainingofcellmonolayersusingana-JEVantiserum,andvirusyield(inPFU/ml)at22hptwasdeterminedbyplaquetitrationonBHK-21cells.(B)Viralgrowth.BHK-21cellswereinfectedatanMOIof0.1withtheparentoroneofthe14E-244mutantviruses.Attheindicatedtimepoints,culturesupernatantswereusedforvirustitrationonBHK-21cells.(C)Neurovirulence.Groupsof3-week-oldICRmice(n=10pergroup)wereinfectedICwithserial10-folddilutionsofeachvirusstock,andtheLD50valuesweredetermined.(D)Homologymodel.ThepredictedmodeloftheEectodomainofJEVSA14-14-2wasbuiltbasedonthecrystalstructureoftheEectodomainofWNVNY99[79],andthemodelwasthenfittedintothecryo-EMstructureofWNVNY99[18].IllustratedisanicosahedralasymmetricunitofthethreeEmonomersontheviralmembrane.HighlightedintheinsetisthecriticalresidueGluatE-244intheijhairpinadjacenttothefusionloopoftheviralEDII.

doi:10.1371/journal.ppat.1004290.g005

growthassaysintwoneuronalcells,NSC-34(mousemotorneuron)andSH-SY5Y(humanneuroblastoma),byinfectingatanMOIof0.1withthenon-neurovirulentparentSA14-14-2MCVandeachofthefourrepresentativeE-244mutantviruses,i.e.,twoneurovirulent(G244EandG244D)andtwonon-neurovirulent(G244RandG244K).Inparallel,thenon-neuronalBHK-21cellswerealsoinfectedforcomparisonwiththesamesetoffiveviruses.InNSC-34cells,whilethetwoneurovirulentvirusesgrewrapidlyandreachedtheirmaximumtitersof1.8–2.46105PFU/mlat72–96hourspost-infection(hpi),thethreenon-neurovirulentviruses,includingtheparent,allreplicatedpoorly,withpeaktitersonlyapproaching1.0–2.06103PFU/ml,,100-foldlowerthanthoseofthetwoneurovirulentviruses(Fig.6A).InSH-SY5Ycells,asimilardefectinviralgrowthwasalsoobserved,witha,50-to100-folddifferenceinmaximumvirustitersbetweentheneuro-virulentandthenon-neurovirulentviruses.Inaddition,wenotedadifferentialgrowthdefectinthethreenon-neurovirulentviruses,withG244RreplicatingmorepoorlythantheparentbutbetterthanG244K(Fig.6B).IncontrasttothepatternofviralgrowthobservedinNSC-34andSH-SY5Ycells,wefoundthatinBHK-21cells,onlyG244Khadanoticeabledefectinviralgrowth,with,20-foldlowerpeaktitersthanthoseoftheotherfourvirusesthatgrewwelltomaximumtitersof0.8–2.56106PFU/mlat48–72hpi(Fig.6C).ThesedataindicatethatE-244playsacrucialroleintheproductiveinfectionofJEVinneuronalcells.

PLOSPathogens|www.plospathogens.org

8

Subsequently,weexaminedtheinfectivity/replicabilityoftheparentanditsfourE-244mutantviruses/RNAsinNSC-34andSH-SY5Ycells,inparallelwithBHK-21cellsforcomparison.First,virusinfectivitywasquantifiedusingflowcytometrybyinfectingthethreecelltypesatanMOIof1witheachofthefivevirusesandcountingthenumberofcellsstainedwithamousea-JEVantiserumat12–15hpi.InNSC-34cells,thetwonon-neurovirulentmutants(G244RandG244K)exhibitedinfectivitiesnearlyidenticaltothatofthenon-neurovirulentparent(Fig.6D),whereasthetwoneurovirulentmutants(G244EandG244D)showedinfectivities,16-to20-foldhigherthanthatofthenon-neurovirulentparent.Similarly,theE-244mutationalsoalteredvirusinfectivityinSH-SY5Ycells.Specifically,thetwoneurovirulentmutantshad,3-to4-foldhigherinfectivitiesthanthenon-neurovirulentparent;ontheotherhand,thetwonon-neurovirulentmutantsdisplayedevenlowerinfectivitiesthantheparent(,3-foldforG244Rand,10-foldforG244K)(Fig.6E).Incontrast,nosignificantdifferenceinvirusinfectivitywasobservedamongallfivevirusesinBHK-21cells(Fig.6F).

Next,thereplicationefficiencyoftheviralgenomicRNAwasquantifiedbydirectlytransfectingthethreecelltypeswitheachofthefivesyntheticRNAstranscribedinvitrofromtherespectiveJEVcDNAsandestimatingthenumberofinfectiousfocistainedwiththemousea-JEVantiserumat4dayspost-transfection(dpt).Ineachofthethreecelltypes,therewasnodetectabledifferencein

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Table1.Geneticstabilityof14E-244mutantsrecoveredfrommousebrainsafterICinoculation.

Inputvirus

Virus

ANNAA

SA14-14-2MCV(Parent)G

244Recoveredvirus

Initialcodon

1707(aminoacid)(Gly)(Glu)(Asp)(Arg)(Lys)

Nucleotidechange(aminoacidchange)No.ofindependentclones

GGG1709EGAGGACAGAAAG

NoneNoneN.A.None

AAGRGAG(LysRGlu)AAGRACG(LysRThr)

28/2830/30

G244DG244RG

244K1/3025/304/30

AANNA

G

244FUUCUGGACGAGCAAC

(Phe)(Trp)(Thr)(Ser)(Asn)

N.A.N.A.NoneNoneNone

AACRGAC(AsnRAsp)

30/3030/308/3022/3030/30

G244WG244TGG

244244SN

NANII

GG

244244QL

CAGCUGCCGGCCGUC

(Gln)(Leu)(Pro)(Ala)(Val)

NoneN.A.NoneNoneNone

G244PG

24424428/2830/3030/30

AV

G

A,Attenuated;I,Intermediate;N,Neurovirulent.N.A.,Notavailable.

doi:10.1371/journal.ppat.1004290.t001

thespecificinfectivityofthefiveRNAs(NSC-34,Fig.6G;SH-SY5Y,Fig.6H;andBHK-21,Fig.6I).Also,quantitativereal-timeRT-PCRsindicatedthattheleveloftheviralgenomicRNAsaccumulatedintheRNA-transfectedcellsoverthefirst15hoftransfectionwasindistinguishablebetweentheparentandthefourdifferentE-244mutants(datanotshown).Regardlessofcelltype,however,theG244KmutantwasdifferentfromtheparentandtheotherthreeE-244mutants,asdemonstratedbya,1-logdecreaseintheyieldofinfectiousvirionsreleasedintoculturemediumduringthefirst20hpt(NSC-34,Fig.6J;SH-SY5Y,Fig.6K;andBHK-21,Fig.6L).Overall,theseresultsshowthattheE-244mutationaltersJEVinfectivityinaneuronalcell-specificmanner,inagreementwiththeneurovirulencephenotypeobservedinmice,anditalsoaffectsinfectiousparticleproductioninacelltype-nonspecificmanner.

Comparisonoftheaminoacidsequencesoftheijhairpininencephaliticandnon-encephaliticflaviviruses

Weinitiallygeneratedamultiplesequencealignmentusingall154full-lengthJEVgenomesavailablefromtheGenBanksequencedatabase.OfnoteisthefactthatSA14andSA14-14-2havebeenfullysequencedbythreeandfourindependentgroups,respectively;theirnucleotideanddeducedaminoacidsequencesarenotidentical[63–65,77,81].ThesequencealignmentshowedaGluresidueatpositionE-244intheijhairpinofallJEVstrainsisolatedfrominfectedmosquitoes,pigs,orhumans,exceptfortheGln-encodingmosquito-derivedK94P05andthreeGly-encodingSA14-derivedattenuatedstrains(i.e.,SA14-2-8,SA14-12-1-7,andallfourdifferentversionsofSA14-14-2)(Fig.S7).IncaseofSA14,itisintriguingtonotethatoneversionhasaGluresidueatpositionE-244,buttheothertwoversionshaveaGlyresidueatthatposition(Fig.S7);thisdiscrepancyislikelyduetovariationsinthe

PLOSPathogens|www.plospathogens.org

9

cultivationhistoryofthevirus[77].Wenextperformedthestructure-based,ij-hairpinaminoacidsequencealignmentwithsixrepresentativeflaviviruses(14strainstotal),includingfourencephalitic(JEV,WNV,SLEV,andMVEV)andtwonon-encephalitic(YFVandDENV)flaviviruses.InadditiontotheimportanceoftheE-244aminoacid,wenoted(i)theevolutionallyconservedresiduesintheijhairpinanditsflankingregioninallsixflaviviruses,i.e.,W233,F242,H246,A247,V252,L255,G256,Q258,E259,andG260;(ii)thesequencesimilaritiesinthefourencephaliticflaviviruses,particularlyina,15-aaij-hairpin-containingregion;and(iii)thesequencedifferencesbetweenthefourencephaliticandtwonon-encephaliticflaviviruses,e.g.,the4-aaYFV-specificmotifandthe3-aaDENV-specificmotif(Fig.7).Overall,ourfindingssuggestthattheijhairpinoftheEDIIplaysakeyroleindeterminingencephaliticflavivirusneurovirulence,anditsfunctionisregulatedbythechemicalpropertiesoftheaminoacidatpositionE-244inthathairpin.

Discussion

Inthiswork,wehavedevelopedareversegeneticssystemforSA14-14-2,alive-attenuatedJEvaccine[53,82],byconstructinganinfectiouscDNAandrescuingmolecularlyclonedvirusfromthecDNA.ThisreversegeneticssystemoffersusauniqueopportunitytoelucidatethegeneticandmolecularbasisofJEVneurovirulence.UsingourinfectiousSA14-14-2cDNAtechnology,we(i)generatedthreeisogenicSA14-14-2variantsthatunlikeitsparent,displayedlethalneurovirulenceinmice;(ii)identifiedasinglepointmutation,G1708RA,causingaGlyRGluchangeatposition244oftheviralEglycoproteinthatissufficienttoconferafullneurovirulencebypromotingviralinfectionintoneuronsinthemouseCNSinvivoandmouse/humanneuronalcellsinvitro;

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure6.E-244:Amajordeterminantofviralinfectivityinmouseandhumanneuronalcellsinvitro.(A-C)Viralgrowth.NSC-34(A),SH-SY5Y(B),orBHK-21(C)cellswereinfectedatanMOIof0.1withtheparentoroneofthefollowingfourE-244mutantviruses:G244E,G244D,G244R,andG244K.Attheindicatedtimepoints,culturesupernatantswereharvestedforvirustitrationonBHK-21cells.(D-F)Viralparticleinfectivity.NSC-34(D),SH-SY5Y(E),orBHK-21(F)cellswereinfectedatanMOIof1witheachofthefiveviruses,asindicated.At12–15hpi,thenumberofinfectedcellswasmeasuredbyflowcytometryusingmousea-JEVantiserum.Theresultsaretheaverageoftwoindependentexperiments,presentedas-foldchangesininfectivityrelativetotheparent(infectivityvalueof1).(G-I)ViralRNAinfectivity/replicability.NSC-34(G),SH-SY5Y(H),orBHK-21(I)cellsweretransfectedwithRNAstranscribedfromtheparentoreachmutantcDNA,asindicated.At4dpt,RNAinfectivitywasquantifiedbyinfectiouscenterassay,coupledwithstainingofcellmonolayersusingana-JEVantiserum.(J-L)Virusyield.At20hafterRNAtransfection,culturesupernatantsfromRNA-transfectedNSC-34(J),SH-SY5Y(K),orBHK-21(L)cellswereharvestedforvirustitrationonBHK-21cells.doi:10.1371/journal.ppat.1004290.g006

and(iii)demonstratedthestructure-functionrelationshipforneurovirulenceofE-244intheijhairpinadjacenttothefusionloopatthetipoftheviralEDII.Thus,ourfindingsrevealfundamentalinsightsintotheneurotropismandneurovirulenceofJEVandothertaxonomicallyrelatedencephaliticflaviviruses,includingWNV,SLEV,andMVEV.Intriguingly,ourresultsalsoprovideanewtarget,theijhairpin,forthedevelopmentofnovelantiviralsforthepreventionandtreatmentofinfectionwiththeencephaliticflaviviruses.

TheflavivirusglycoproteinEmediatesreceptor-mediatedendocytosisandlowpH-triggeredmembranefusion[33,83,84].Ontheviralmembrane,180Emonomersarepackedinto30protein‘‘rafts’’,eachcomposedofthreeEhead-to-tailhomodi-mers[17–19].EachEmonomeriscomposedofthreeparts:(i)anelongatedectodomainthatdirectsreceptorbindingandmem-branefusion;(ii)a‘‘stem’’regioncontainingtwoamphipathic

PLOSPathogens|www.plospathogens.org

10

a-helicesthatliesflatontheviralmembraneunderneaththeectodomain;and(iii)amembrane‘‘anchor’’regioncontainingtwotransmembraneantiparallelcoiled-coils.TheEectodomainfoldsintothreeb-barreldomains[85]:(i)DI,astructuraldomaincentrallylocatedinthemolecule;(ii)DII,anelongateddimeriza-tiondomaincontainingthehighlyconservedfusionloopatitstip[86];and(iii)DIII,anIg-likedomainimplicatedinreceptorbinding[20,87,88]andantibodyneutralization[–92].Basedonpre-andpost-fusioncrystalstructuresoftheectodomain[28,30,31,47,79,80,85]andbiochemicalanalyses[29,32,93],acurrent,detailedmodelforflavivirusmembranefusionhasbeendeveloped.Inthismodel,thefusionisinitiatedbyalowpH-induceddissociationoftheantiparallelEhomodimersthatleadstotheexposureofthefusionloopsandtheirinsertionintothehostmembrane,followedbyalarge-scalestructuralrearrangementintoaparallelEhomotrimer[33,83,84,94].Intheparallelconformation,DIIIfoldsbacktoward

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

Figure7.Structure-based,ij-hairpinaminoacidsequencealignmentforsixrepresentativeflaviviruses(14strainstotal):JEV,WNV,SLEV,MVEV,YFV,andDENV.AllsequenceinformationwasretrievedfromtheGenBankdatabasewithaccessionnumbersindicated.MultiplesequencealignmentswereperformedusingClustalX[116].Highlightedarethe,15-aaij-hairpin(yellow),the3-aaDENV-specificmotif(pink),andthe4-aaYFV-specificmotif(green).Theconsensussequenceoftheijhairpinanditsflankingregionispresentedontop,andonlydifferencesfromthatsequenceareshown.Deletionsareindicatedbyhyphens.TheaminoacidresidueisnumberedbasedontheJEVSA14-14-2(GenBankaccessionno.JN604986).

doi:10.1371/journal.ppat.1004290.g007

DII,presumablywiththestemextendedfromtheC-terminusofDIIIalongDIIandtowardthefusionloop(‘‘zipping’’),drivingthefusionoftheviralandhostmembranes[95–98].Despiteourdetailedknowledgeaboutthefusionprocess,thereislittleavailablestructuralinformationabouthowflavivirusesbindtotheircellularreceptors.Inencephaliticflaviviruses,thepresenceofanRGDmotifinDIIIandcarbohydratemoietiesontheviralsurfacesuggestsamechanisminvolvinginteractionwiththeRGDmotif-recognizingintegrinsandsugar-bindinglectinsonthecellsurface,respectively.However,blocking/alterationofeithertheRGDmotiforglycandoesnotabolishviralentry[22,99–101].Thus,theviralfactorsandtheinteractingcellularcounterpartsrequiredforviralentryarestillelusive.

Inflaviviruses,theijhairpinisastructuralmotifthatiscloselyassociatedwiththefusionloopatthetipoftheviralEDII,butitsroleisthusfarunknown.InJEV,wenowreportthatasingleaminoacidintheijhairpin,E-244,servesasakeyregulatortocontrolthelevelofneurovirulenceofSA14-14-2inmice.Thisaminoacidwasalsocorrelatedwithadifferentialabilitytoinfectneurons,theprimarytargetcellsintheCNS.Consistentwiththisfinding,wefoundthatsite-directedmutagenesisofthecodonforE-244inSA14-14-2createdapanelof14recombinantvirusesofvaryingneurovirulence:(i)non-neurovirulentviruses,producedbysubstitutionsofpositively-charged(R,K),aromatic(F,W),polar(N),oraliphatic(L)residues;(ii)neurovirulentviruses,producedbysubstitutionsofnegatively-charged(E,D),hydroxyl(T,S),polar(Q),orcyclic(P)residues;and(iii)virusesintermediateinneurovirulence,producedbysubstitutionsofaliphatic(A,V)residues.TheseresultshighlighttheroleofE-244inneuroviru-lence,whichwasdirectedbyacombinationofthreemajorpropertiesofitsaminoacidsidechain:(i)charge(R/Kvs.E/D);(ii)size(Nvs.QandLvs.A/V);and(iii)functionalgroup(Nvs.D).OurdatasuggestthattheijhairpinactsasaviralfactorthatpromotesJEVinfectionofneuronswithintheCNS,likelythroughitsroleinoneofthreemajorstepsinvolvedinviralentry:binding,

PLOSPathogens|www.plospathogens.org

11

endocytosis,ormembranefusion[33,83,84].Alternatively,itispossiblethatthelatestepsintheviruslifecycleinneurons,suchasassembly,maturation,andrelease,couldbeaffected.ForJEV,WNV,andtick-borneencephalitisvirus,theassembly/releaseofinfectiousvirionsorvirus-likeparticleshasbeenshowntobeaffectedbytheN-glycosylationoftheviralprMand/orEproteininnon-neuronalcells[44,75,102–104].Moreover,aconservedsingleN-glycosylationsiteintheJEVprMproteinhasbeenshowntobeimportantforviralpathogenicityinmice[75].

Overtheyears,thevirulenceofJEVhasbeenanactiveareaofresearch.Initially,comparisonofthegenomicsequencesofseveralJEVstrainswithadifferentdegreeofpathogenicityhadpredictedanumberofpotentiallociintheviralgenomethatareinvolvedinvirulence[63–74].Duetothecomplexityandvariationofthemutations,however,theidentityofthemajorviralfactorthatiscriticalforJEVvirulenceremainsunclear.Inparticular,SA14-14-2hasbeenreportedtohaveatotalof47–nucleotidechanges(17–27aminoacidsubstitutions)whencomparedtoitsvirulentparentalstrainSA14;thenumberofmutationsvariesanddependsonthecultivationhistoryoftheviruses[63–65].Ofthetenviralproteins,theEproteinhasbeentheprimarytargetofgeneticstudiesinvirulence,mainlybecauseitisinvolvedincell/tissuetropismandpathogenesis.SeveralaminoacidresiduesintheEproteinhavebeensuggestedtocontributetotheneurovirulenceand/orneuroinvasivenessofJEVinvivo:(i)E-123,illustratedbyanS123Rsubstitutionthatiscapableofenhancingtheneuroinva-sivenessoftheMie/41/2002strainin3-week-oldddYmice[105];(ii)E-279,exemplifiedbyanM279KmutationthatisabletoincreasetheneurovirulenceofChimeriVax-JE(achimericvirusthatcarriestheprMandEgenesofJEVSA14-14-2onaYFV17Dgeneticbackground)insucklingmiceandrhesusmonkeys[106];and(iii)E-138,indicatedbytworeciprocalmutations:(1)aK138Esubstitution,whencombinedwithatleasttwoothermutations,whichelevatestheneurovirulenceoftheChimeriVax-JEvirusin4-week-oldICRmice[107],and(2)anE138Ksubstitution,which

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

lowerstheneurovirulenceandneuroinvasivenessofthreedifferentJEVs(theJaOArS982strainin2-to5-week-oldSwissICRmice[73]andtheAT31andNT109strainsin3-week-oldBALB/cmice[108,109]).ThesedataindicatethatmultipleaminoacidresiduesintheEproteinofJEVfunctioninamorecoordinatedwaytoachievethemaximallevelofneurovirulenceand/orneuroinvasiveness[53,107].ThisnotionisconsistentwithourfindingthatalthoughasingleG244EmutationintheEproteinofSA14-14-2issufficienttoconferlethalneurovirulencein3-week-oldICRmice,thespreadofthevirusinthebrainsisstillslowandlimited,ascomparedtothehighlyvirulentCNU/LP2strain.TheseandpreviousfindingssuggestthatinadditiontoE-244,otheraminoacidresiduesintheEproteinplayaroleindeterminingtheneurovirulenceofSA14-14-2.Inaddition,JEVNS1’(aproductofribosomalframeshifting[110,111])isreportedtobeproducedincellsinfectedwithSA14butnotwithSA14-14-2,anditslackofexpressionisshowntocontributetotheattenuationphenotypeofSA14-14-2inmice[112].Similarly,theexpressionofNS1’isalsosuggestedtobeassociatedwiththeneuroinvasivenessofWNV[110].

TheneuroattenuationphenotypeofSA14-14-2hasbeentestedinseverallaboratoryanimals,includingmiceandmonkeys[53].In2-to4-week-oldICRandddYmice,nomorbidityormortalityhasbeenobservedaftersubcutaneousandintracerebralinoculationswith104-106PFUofSA14-14-2[53,63,113];inararecase,however,theviruswasfoundtobeabletocausethedeathofthemicefollowingICinoculation[113].Inlinewiththesepreviousresults,wealsofoundthatnoneofthe3-week-oldICRmiceinoculatedIMorIPwithupto,105PFUofSA14-14-2showedclinicalsignsordeath;ontheotherhand,althoughtherewassomevariabilityamongthegroupsofmiceandthedosesofvirusinoculum,,5–30%ofthemiceinoculatedICwithadoseof103–105PFUdevelopedJEV-specificsymptomsanddeath.ThislowbutunexpectedmorbidityandmortalityaftertheICinoculationofSA14-14-2islikelycausedbyacombinationoffactorsandconditionsimposedonourinfectionexperiments,particularlytheageandstrainofmice:(i)Age-dependentsusceptibilityofflaviviruses,includingJEV,inthemurinemodelhasbeendocumentedpreviously,althoughitsmolecularmechanismsremainunclear[106,114,115].(ii)Anoticeablevariabilityinmortalityhasalsobeenreportedwhentwodifferentlineagesoftheage-matchedoutbredICRmiceareinoculatedICwithamutantofChimeriVax-JEthatcontainstwoaminoacidsubstitutions(F107LandK138E)intheSA14-14-2Eprotein-codingregion,suggestingthatdifferencesinthegeneticbackgroundofmicemayaccountforthevariableneurovirulence[107].Moreimportantly,itshouldbepointedoutthatinourstudy,alltherevertantsrecoveredfromthemiceinoculatedICwithSA14-14-2appearedtohavetheG244Emutation,whichissufficienttoconferlethalneurovirulencetothevirus,corroboratingthattheparentalSA14-14-2virusishighlyattenuatedinneurovirulence.Furthermore,itisintriguingtonotethattheG244EsubstitutionhasbeenintroducedintoChimeriVax-JE,inwhichnomortalityoccurswheneight4-week-oldICRmiceareinjectedICwith104PFUofthemutantvirus[107];therefore,itappearsthatneurovirulencemaydependonthegeneticbackgroundofthepathogen.FurtherinvestigationisneededtofullyelucidatetheneurovirulenceandneuroinvasivenessofJEV.

Insummary,weshowforthefirsttimethatE-244intheijhairpinoftheviralEDIIisakeyregulatordeterminingtheneurovirulenceofSA14-14-2,andwealsoprovidedirectevidencethatviralEcancontributetotheneurovirulenceofJEVandpossiblyothercloselyrelatedencephaliticflavivirusesviaitsroleintheearlyorlatestageofviralreplicationinneurons.Adetailed,

PLOSPathogens|www.plospathogens.org

12

completeunderstandingoftheevolutionallyconservedviralijhairpinanditsfunctionintheviruslifecyclewillhavedirectapplicationtothedesignofanovelandpromisingclassofbroad-spectrumantivirals(e.g.,ligandsandsmallmolecules)toexpandthecurrentlyavailablepreventiveandtherapeuticarsenalagainstinfectionwithencephaliticflaviviruses.

MaterialsandMethodsVirusesandcells

AnoriginalstockofJEVSA14-14-2wasretrieveddirectlyfromabatchofcommercialvaccinevials(ChengduInstituteofBiologicalProducts,China)forviralgenomesequencingandcDNAconstruction,toavoidanypotentialconcernthatitsadaptationcouldoccurduringpropagationincellculture.ThisvirusstockwaspropagatedtwiceinBHK-21cellstogeneratehigh-titerviralpreparationsforcellandmouseinfectionexperiments.StocksofJEVCNU/LP2werederivedfromtheinfectiouscDNApBACSP6/JVFLx/XbaI[78].BHK-21cellsweregrowninalphaminimalessentialmedium(a-MEM)containing10%fetalbovineserum(FBS),2mML-glutamine,vitamins,andpenicillin-strepto-mycinat37uCin5%CO2[78].SH-SY5Ycellswerecultivatedina1:1mixtureofMEMandHam’sF-12nutrientmixsupplementedwith10%FBS,0.1mMnonessentialaminoacids,andpenicillin-streptomycinat37uCin5%CO2[75].NSC-34cellsweremaintainedinDulbecco’smodifiedEagle’smediumcontaining10%FBSandpenicillin-streptomycinat37uCin5%CO2.

JEVreversegenetics

Asavector,weusedtheBACplasmidpBeloBAC11[78].First,fourcDNAfragmentscoveringtheentireviralgenomewereclonedintothevectorindividually,thenjoinedsequentiallyatthreenaturalrestrictionsites(BsrGI,BamHI,andAvaI)togenerateasingleBACclonethatcontainedthefull-lengthSA14-14-2cDNA,namedpBAC/SA14-14-2(Fig.1).TheSP6promotersequencewaspositionedjustupstreamoftheviral59-end,andanartificialXbaIrun-offsitewasengineeredjustdownstreamoftheviral39-end.Apre-existingXbaIsiteatnucleotide9131wasremovedbyintroducingasilentpointmutation(A9134RT);thismutationalsoservedasarescuemarkertoidentifythecDNA-derivedSA14-14-2.AllmutationswerecreatedbyoverlapextensionPCR.AllPCR-generatedfragmentsweresequenced.DetailedcloningproceduresaredescribedinSupportingInfor-mation.

RNAtranscriptionandtransfection

AllBACplasmidswerepurifiedbycentrifugationusingCsCl-ethidiumbromideequilibriumdensitygradients.TheclosedcircularplasmidswerelinearizedbyXbaIandmungbeannucleasedigestiontoproduceDNAtemplatesforinvitrorun-offtranscription.RNAwastranscribedfromalinearizedplasmidwithSP6RNApolymeraseasdescribed[78].TheresultingRNAwasstoredat280uCuntilneeded.RNAyieldwasmeasuredonthebasisoftheincorporationrateof[3H]UTP,andRNAintegritywasevaluatedbyagarosegelelectrophoresis.RNAwastransfectedbyelectroporationintocellsunderouroptimizedconditions(980V,a99-mspulselength,andfivepulsesforBHK-21cells;and760V,a99-mspulselength,andfivepulsesforNSC-34andSH-SY5Ycells)[78].RNAinfectivitywasdeterminedbyinfectiouscenterassayasreported[78].Theinfectiouscentersoffociweredetectedbydecoratingofcellswithamousea-JEVantibody(AmericanTypeCultureCollection[ATCC],1:500)andahorseradishperoxidase-conjugatedgoata-mouseIgG(Jackson

July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

ImmunoResearch,1:1,000),followedbystainingwith3,39-diaminobenzidine(Vector).described[75,77].Inallmice,viralreplicationinbraintissuewasconfirmedbyplaquetitrationand/orRT-PCR[75].

Northernblots

TotalRNAwasextractedwithTRIzolreagent(Invitrogen).Northernblotanalysiswasperformedasdescribed[78].JEVgenomicRNAwasdetectedwithanantisenseriboprobethatbindstoa209-bpregion(nt9143–9351)intheNS5protein-codingregion.Theprobewassynthesizedwith[a-32P]CTPbyusingtheT7-MEGAscriptkit(Ambion).Theblotswereprehybridized,hybridized,andwashedat55uC.Autoradiographswereobtainedbyexposuretofilmfor24–48h.

Ethicsstatement

AllanimalstudieswereconductedinstrictaccordancewiththeregulationsintheGuidefortheCareandUseofLaboratoryAnimalsissuedbytheMinistryofHealthandWelfareoftheRepublicofKorea.TheprotocolwasapprovedbytheInstitu-tionalAnimalCareandUseCommitteeoftheChungbukNationalUniversityMedicalSchool(PermitNumber:LML08-73).AllmicewerehousedinouranimalfacilitylocatedattheChungbukNationalUniversityMedicalSchool,andeveryeffortwasmadetominimizesuffering.

Immunoblots

Cellswerelysedinsamplebuffer(80mMTri-HCl[pH6.8],2.0%SDS,10%glycerol,0.1Mdithiothreitol,0.2%bromophe-nolblue).EqualamountsofthelysateswererunonSDS-polyacrylamidegels,transferredtopolyvinylidenedifluoridemembranes,andsubjectedtoimmunoblottingasdescribed[78].Thefollowingpolyclonalantiserawereusedasprimaryantibodies[75,76]:a-JEV(mouse,1:1,000),a-C(rabbit,1:1,000),a-pr(rabbit,1:4,000),a-E(rabbit,1:500),a-NS1(rabbit,1:1,000),anda-GAPDH(rabbit,1:10,000).Analkalinephosphatase-conjugatedgoata-mouseora-rabbitIgG(JacksonImmunoResearch,1:5,000)wasusedforthesecondaryantibody,asappropriate.Thespecificsignalswerevisualizedbychromogenicmembranestainingwithamixtureof5-bromo-4-chloro-3-indolyl-phosphateandnitrobluetetrazolium(Sigma-Aldrich).

Cells(56105)wereharvestedbytrypsinization,washedwithphosphate-bufferedsaline(PBS),andcollectedbycentrifugationat1,0006gfor5min.Thecellswereresuspendedin250mlofCytofix/Cytopermsolution(BDBiosciences)andincubatedat4uCfor20mininthedark.AllsubsequentwashandstainingstepswereperformedinPerm/Washbuffer(BDBiosciences).Thecellswerewashedtwiceandincubatedin200mlofmousea-JEVantiserum(ATCC,1:500)for1hat4uC.Subsequently,thecellswerewashedtwiceandincubatedin200mlofAlexaFluor488goata-mouseIgG(MolecularProbes,1:1,000)for1hat4uC.Thecellswerethenwashedtwiceandresuspendedin200mlofPerm/Washbuffer.ThesampleswereanalyzedonaFACSAriaIIIcellsorterwithDiva6.1.3software(BDBioscienc-es).Foreachsample,50,000eventswerecollectedwithinthelinearrangeofdetection.

Immunohistochemistry

Groupsof3-week-oldfemaleICRmice(n=15pergroup)wereinfectedICwith103PFUofvirusin20mlofa-MEM;10controlmicewereinoculatedICwithanequivalentvolumeofsupernatantfromuninfectedcontrolBHK-21cellculturesatcomparabledilution.At3,5,and7dpi,fiverandomlyselectedmiceweretranscardiallyperfusedwithice-coldPBS,followedby4%paraformaldehyde(PFA).Brainswerefixedin4%PFA,embeddedinparaffin,andcutinto6-mmsections.Brainsectionsweretreatedinmicrowaveforantigenretrievalandincubatedwith1%H2O2inice-coldmethanolfor30mintoblockendogenousperoxidase.Theywerethenblockedwith1%normalgoatserumandincubatedwithrabbita-NS1antiserum(1:200)for12hat4uC,followedbyincubationwithbiotinylateda-rabbitIgGplustheavidin-biotin-peroxidasecomplex(Vector).Signalswerevisualizedbystainingwith3,39-diaminobenzidinesolutioncontaining0.003%H2O2andcounterstainingwithhematoxylin.

Flowcytometry

Homologymodeling

ThesequenceandstructureoftheEectodomainofWNVNY99(PDBaccessioncode2HG0)wasusedastemplateforthehomologymodeling.ThesequencealignmentwasdoneusingtheonlineversionofClustalW2[117].ProteinstructurehomologymodelingwasperformedusingtheSWISS-MODELWorkspace,accessibleviatheExPASywebserver[118].ThegeneratedmodelwasvisualizedusingUCSFChimera1.5.3.ThemodelisinagreementwitharecentcrystalstructureoftheEectodomainofJEVSA14-14-2(PDBaccessioncode3P54)[80].

SupportingInformation

ofSA14-14-2MCV.BHK-21cellsweremock-infectedorinfectedwithoneofthefollowingthreeJEVs:SA14-14-2MCV,SA14-14-2,orCNU/LP2(avirulentstrainusedasareference).Afterinfection,cellswereoverlaidwithagarosetoexaminefocus/plaquemorphologies.At4dpi,cellmonolayerswerefirstimmunostainedwithamousea-JEVantiserumtovisualizetheinfectiousfoci,andthesamemonolayerswerethenrestainedwithcrystalviolettoobservetheinfectiousplaques.Theaverageplaquesizes(mean6SD)weredeterminedbycounting10representativeplaques.(PPT)

ViralgrowthpropertiesofSA14-14-2MCVin

SH-SY5YandC6/36cells.CellswereinfectedatanMOIof1withthemolecularlyclonedvirus(SA14-14-2MCV)rescuedfromthefull-lengthinfectiouscDNAortheoriginalparentalvirus(SA14-14-2)usedforcDNAconstruction.Culturesupernatants

FigureS2

13

July2014|Volume10|Issue7|e1004290

Sequenceanalysis

ThefullgenomesequencesofSA14-14-2anditsneurovirulentvariantsweredeterminedasdescribed[77].SequencingoftheprM-EcodingregionoftheE-244mutantswasdoneasfollows:(i)amplificationofa2,069-bpcDNAbyRT-PCRusingasetofthreeprimers(prMErt,prMEfw,andprMErv;seeTableS4);(ii)cloningofa2,057-bpXhoI-SacIIfragmentintothepRS2vector;and(iii)sequencingof,30randomlypickedindependentclonescontain-ingtheinsert.MultiplesequencealignmentswereperformedusingClustalX[116].

FigureS1Representativefocus/plaquemorphologies

Mouseinfection

Female3-week-oldICRmice(CharlesRiver)wereused.Groupsof10or20micewereinoculatedIC(20ml),IM(50ml),orIP(50ml)with10-foldserialdilutionsofvirusstockina-MEM.MiceweremonitoredforanyJEV-inducedclinicalsignsordeathevery12hfor24days.TheLD50valuesweredeterminedas

PLOSPathogens|www.plospathogens.org

JEVNeurovirulence

werecollectedatthehourpostinfection(hpi)indicated,andvirustitersweredeterminedbyplaqueassaysonBHK-21cells.(PPT)

mice.Groupsof3-week-oldfemaleICRmice(n=20pergroup)weremock-inoculatedorinoculatedintracerebrally(IC),intra-muscularly(IM),orintraperitoneally(IP)withserial10-folddilutionsofSA14-14-2MCV,SA14-14-2,orCNU/LP2(avirulentJEVstrainusedasareference).MicewereobservedforanyJEV-inducedclinicalsignsanddeathevery12hfor24days.SurvivalcurveswereplottedbytheKaplan-Meiermethod.(PPT)

bilitytoSA14-14-2MCVinfectionofneuronsintheCNS.Groupsof3-week-oldfemaleICRmice(n=15pergroup)weremock-infectedorinfectedICwith103PFUofSA14-14-2MCV(Parent),G1708A,orCNU/LP2(avirulentJEVstrainusedasareference).Ontheindicateddaysafterinfection,fivemiceweresubjectedforimmunostainingofJEVNS1antigeninfixedbrainsliceswithana-NS1antiserum.Presentedarerepresentativeslidesofamygdala,cerebralcortex,thalamus,hypothalamus,andbrainstem(notethathippocampalslidesareshowninFig.4E).ArrowheadsindicatetheNS1-positivecells.(PPT)

FigureS5LevelsofJEVproteinaccumulationinBHK-21cellstransfectedwith14E-244mutantRNAs.BHK-21cellsweremock-transfectedortransfectedwithRNAstranscribedfromSA14-14-2MCV(Parent)oreachofthe14E-244mutantcDNAsasindicated.At18hpt,viralproteinaccumulationwasanalyzedbyimmunoblottingofcelllysateswithapanelofJEV-specificantisera.Inparallel,GAPDHproteinwasusedasaloadingandtransfercontrol.(PPT)

FigureS6ThelocationofE-244onthecrystalstructureFigureS4Asinglepointmutationpromotessuscepti-FigureS3VirologicalpropertiesofSA14-14-2MCVin

retrievedfromtheRCSBProteinDataBank(PDBaccessioncode3P54).(PPT)

Aminoacidsequencealignmentof154fully

sequencedJEVstrainsattheconservedijhairpinofviralEglycoprotein.Multiplesequencealignmentswereperformedusingtheaminoacidsequenceof154fullysequencedJEVgenomes,includingSA14(red),SA14-14-2(green),andtwootherSA14-derivedattenuatedstrains,SA14-2-8(orange)andSA14-12-1-7(blue).NotethatSA14andSA14-14-2havebeensequencedbythreeandfourindependentresearchgroups,respectively.Theconsensussequenceoftheijhairpinanditsflankingregionispresentedontop,andonlydifferencesfromthatsequenceareshown.Highlightedarethe,15-aaij-hairpinandthepositionE-244inthathairpin.(PPT)

FigureS7

SA14-14-2MCVanditsthreevariantsin3-week-oldICRmice.(PPT)

mutantsin3-week-oldICRmice.(PPT)

TableS3

TableS1Neurovirulenceandneuroinvasivenessof

TableS2NeurovirulenceofSA14-14-2MCVanditseight

NeurovirulenceofSA14-14-2MCVandits14E-244mutantsin3-week-oldICRmice.(PPT)

TableS4Oligonucleotidesusedforligation,cDNA

synthesis,andPCRamplification.(PPT)

Acknowledgments

WethankDr.DeborahMcClellanforeditingthemanuscript.

oftheEectodomainofJEVSA14-14-2.TheEectodomainofJEVSA14-14-2:DI(coloredred),DII(yellow),DIII(blue),andthefusionloop(green).ThecriticalresidueGlyatE-244intheijhairpinadjacenttothefusionloopoftheviralEDIIisshown.ThecrystalstructureoftheEectodomainofJEVSA14-14-2was

AuthorContributions

Conceivedanddesignedtheexperiments:SIYBHSYML.Performedtheexperiments:SIYBHSJKKGNYEYL.Analyzedthedata:SIYBHSRJKMGRYML.Contributedreagents/materials/analysistools:LLRJKMGRJDM.Wrotethepaper:SIYBHSYML.

References

1.EndyTP,NisalakA(2002)Japaneseencephalitisvirus:ecologyandepidemiology.CurrTopMicrobiolImmunol267:11–48.

2.MackenzieJS,BarrettAD,DeubelV(2002)TheJapaneseencephalitisserologicalgroupofflaviviruses:abriefintroductiontothegroup.CurrTopMicrobiolImmunol267:1–10.

3.MackenzieJS,GublerDJ,PetersenLR(2004)Emergingflaviviruses:thespreadandresurgenceofJapaneseencephalitis,WestNileanddengueviruses.NatMed10:S98–109.

4.MackenzieJS,JohansenCA,RitchieSA,vandenHurkAF,HallRA(2002)Japaneseencephalitisasanemergingvirus:theemergenceandspreadofJapaneseencephalitisvirusinAustralasia.CurrTopMicrobiolImmunol267:49–73.

5.GublerDJ(2007)ThecontinuingspreadofWestNilevirusinthewesternhemisphere.ClinInfectDis45:1039–1046.

6.NashD,MostashariF,FineA,MillerJ,O’LearyD,etal.(2001)TheoutbreakofWestNilevirusinfectionintheNewYorkCityareain1999.NEnglJMed344:1807–1814.

7.WeaverSC,BarrettAD(2004)Transmissioncycles,hostrange,evolutionandemergenceofarboviraldisease.NatRevMicrobiol2:7–801.

8.NettRJ,CampbellGL,ReisenWK(2009)PotentialfortheemergenceofJapaneseencephalitisvirusinCalifornia.VectorBorneZoonoticDis9:511–517.9.WeaverSC,ReisenWK(2010)Presentandfuturearboviralthreats.AntiviralRes85:328–345.

10.CampbellGL,HillsSL,FischerM,JacobsonJA,HokeCH,etal.(2011)

EstimatedglobalincidenceofJapaneseencephalitis:asystematicreview.BullWorldHealthOrgan:766–774,774A–774E.

11.MonathTP(2002)Japaneseencephalitisvaccines:currentvaccinesandfuture

prospects.CurrTopMicrobiolImmunol267:105–138.

12.SolomonT,VaughnDW(2002)PathogenesisandclinicalfeaturesofJapanese

encephalitisandWestNilevirusinfections.CurrTopMicrobiolImmunol267:171–194.

13.SolomonT(2006)ControlofJapaneseencephalitis-withinourgrasp?

NEnglJMed355:869–871.

14.MisraUK,KalitaJ(2010)Overview:Japaneseencephalitis.ProgNeurobiol91:

108–120.

15.DoklandT,WalshM,MackenzieJM,KhromykhAA,EeKH,etal.(2004)

WestNileviruscoreprotein:tetramerstructureandribbonformation.Structure12:1157–1163.

16.MaL,JonesCT,GroeschTD,KuhnRJ,PostCB(2004)Solutionstructureof

dengueviruscapsidproteinrevealsanotherfold.ProcNatlAcadSciUSA101:3414–3419.

17.KuhnRJ,ZhangW,RossmannMG,PletnevSV,CorverJ,etal.(2002)

Structureofdenguevirus:implicationsforflavivirusorganization,maturation,andfusion.Cell108:717–725.

18.MukhopadhyayS,KimBS,ChipmanPR,RossmannMG,KuhnRJ(2003)

StructureofWestNilevirus.Science302:248.

19.ZhangW,ChipmanPR,CorverJ,JohnsonPR,ZhangY,etal.(2003)

Visualizationofmembraneproteindomainsbycryo-electronmicroscopyofdenguevirus.NatStructBiol10:907–912.

20.ChenY,MaguireT,HilemanRE,FrommJR,EskoJD,etal.(1997)Dengue

virusinfectivitydependsonenvelopeproteinbindingtotargetcellheparansulfate.NatMed3:866–871.

PLOSPathogens|www.plospathogens.org14July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

21.DavisCW,NguyenHY,HannaSL,SanchezMD,DomsRW,etal.(2006)

WestNilevirusdiscriminatesbetweenDC-SIGNandDC-SIGNRforcellularattachmentandinfection.JVirol80:1290–1301.

22.Navarro-SanchezE,AltmeyerR,AmaraA,SchwartzO,FieschiF,etal.(2003)

Dendritic-cell-specificICAM3-grabbingnon-integrinisessentialfortheproductiveinfectionofhumandendriticcellsbymosquito-cell-deriveddengueviruses.EMBORep4:723–728.

23.PokidyshevaE,ZhangY,BattistiAJ,Bator-KellyCM,ChipmanPR,etal.

(2006)Cryo-EMreconstructionofdenguevirusincomplexwiththecarbohydraterecognitiondomainofDC-SIGN.Cell124:485–493.

24.TassaneetrithepB,BurgessTH,Granelli-PipernoA,TrumpfhellerC,FinkeJ,

etal.(2003)DC-SIGN(CD209)mediatesdenguevirusinfectionofhumandendriticcells.JExpMed197:823–829.

25.ChuJJ,NgML(2004)InfectiousentryofWestNilevirusoccursthrougha

clathrin-mediatedendocyticpathway.JVirol78:10543–10555.

26.vanderSchaarHM,RustMJ,WaartsBL,vanderEnde-MetselaarH,Kuhn

RJ,etal.(2007)Characterizationoftheearlyeventsindengueviruscellentrybybiochemicalassaysandsingle-virustracking.JVirol81:12019–12028.27.vanderSchaarHM,RustMJ,ChenC,vanderEnde-MetselaarH,WilschutJ,

etal.(2008)Dissectingthecellentrypathwayofdenguevirusbysingle-particletrackinginlivingcells.PLoSPathog4:e1000244.

28.BressanelliS,StiasnyK,AllisonSL,SturaEA,DuquerroyS,etal.(2004)

Structureofaflavivirusenvelopeglycoproteininitslow-pH-inducedmembranefusionconformation.EMBOJ23:728–738.

29.LiaoM,Sanchez-SanMartinC,ZhengA,KielianM(2010)Invitro

reconstitutionrevealskeyintermediatestatesoftrimerformationbythedenguevirusmembranefusionprotein.JVirol84:5730–5740.

30.ModisY,OgataS,ClementsD,HarrisonSC(2004)Structureofthedengue

virusenvelopeproteinaftermembranefusion.Nature427:313–319.

31.NayakV,DessauM,KuceraK,AnthonyK,LedizetM,etal.(2009)Crystal

structureofdenguevirustype1envelopeproteininthepostfusionconformationanditsimplicationsformembranefusion.JVirol83:4338–4344.32.StiasnyK,KosslC,LepaultJ,ReyFA,HeinzFX(2007)Characterizationofa

structuralintermediateofflavivirusmembranefusion.PLoSPathog3:e20.33.HarrisonSC(2008)Viralmembranefusion.NatStructMolBiol15:690–698.34.LindenbachBD,ThielHJ,RiceCM(2007)Flaviviridae:thevirusesandtheir

replication.In:KnipeDM,HowleyPM,GriffinDE,LambRA,MartinMAetal.,editors.Fieldsvirology.5thed.Philadelphia,PA:LippincottWilliams&WilkinsPublishers.pp.1101–1152.

35.BrintonMA(2002)ThemolecularbiologyofWestNileVirus:anewinvaderof

thewesternhemisphere.AnnuRevMicrobiol56:371–402.

36.ParanjapeSM,HarrisE(2010)Controlofdenguevirustranslationand

replication.CurrTopMicrobiolImmunol338:15–34.

37.VillordoSM,GamarnikAV(2009)Genomecyclizationasstrategyfor

flavivirusRNAreplication.VirusRes139:230–239.

38.WestawayEG,MackenzieJM,KhromykhAA(2002)Replicationandgene

functioninKunjinvirus.CurrTopMicrobiolImmunol267:323–351.

39.GillespieLK,HoenenA,MorganG,MackenzieJM(2010)Theendoplasmic

reticulumprovidesthemembraneplatformforbiogenesisoftheflavivirusreplicationcomplex.JVirol84:10438–10447.

40.HsuNY,IlnytskaO,BelovG,SantianaM,ChenYH,etal.(2010)Viral

reorganizationofthesecretorypathwaygeneratesdistinctorganellesforRNAreplication.Cell141:799–811.

41.WelschS,MillerS,Romero-BreyI,MerzA,BleckCK,etal.(2009)

Compositionandthree-dimensionalarchitectureofthedenguevirusreplicationandassemblysites.CellHostMicrobe5:365–375.

42.LiL,LokSM,YuIM,ZhangY,KuhnRJ,etal.(2008)Theflavivirusprecursor

membrane-envelopeproteincomplex:structureandmaturation.Science319:1830–1834.

43.ZhangY,CorverJ,ChipmanPR,ZhangW,PletnevSV,etal.(2003)

Structuresofimmatureflavivirusparticles.EMBOJ22:2604–2613.

44.LorenzIC,AllisonSL,HeinzFX,HeleniusA(2002)Foldinganddimerization

oftick-borneencephalitisvirusenvelopeproteinsprMandEintheendoplasmicreticulum.JVirol76:5480–5491.

45.StadlerK,AllisonSL,SchalichJ,HeinzFX(1997)Proteolyticactivationof

tick-borneencephalitisvirusbyfurin.JVirol71:8475–8481.

46.YuIM,ZhangW,HoldawayHA,LiL,KostyuchenkoVA,etal.(2008)

StructureoftheimmaturedenguevirusatlowpHprimesproteolyticmaturation.Science319:1834–1837.

47.ZhangY,ZhangW,OgataS,ClementsD,StraussJH,etal.(2004)

ConformationalchangesoftheflavivirusEglycoprotein.Structure12:1607–1618.

48.GublerDJ,KunoG,MarkoffL(2007)Flaviviruses.In:KnipeDM,Howley

PM,GriffinDE,LambRA,MartinMAetal.,editors.Fieldsvirology.5thed.Philadelphia,PA:LippincottWilliams&WilkinsPublishers.pp.1153–1252.49.BeasleyDW,LewthwaiteP,SolomonT(2008)Currentuseanddevelopment

ofvaccinesforJapaneseencephalitis.ExpertOpinBiolTher8:95–106.

50.Wilder-SmithA,HalsteadSB(2010)Japaneseencephalitis:updateonvaccines

andvaccinerecommendations.CurrOpinInfectDis23:426–431.

51.FischerM,LindseyN,StaplesJE,HillsS,CentersforDiseaseControland

Prevention(2010)Japaneseencephalitisvaccines:recommendationsoftheAdvisoryCommitteeonImmunizationPractices(ACIP).MMWRRecommRep59:1–27.52.PlesnerAM(2003)AllergicreactionstoJapaneseencephalitisvaccine.

ImmunolAllergyClinNorthAm23:665–697.

53.YuY(2010)PhenotypicandgenotypiccharacteristicsofJapaneseencephalitis

attenuatedlivevaccinevirusSA14-14-2andtheirstabilities.Vaccine28:3635–31.

54.HalsteadSB,ThomasSJ(2010)Japaneseencephalitis:newoptionsforactive

immunization.ClinInfectDis50:1155–11.

55.JelinekT(2009)Ixiaro:anewvaccineagainstJapaneseencephalitis.Expert

RevVaccines8:1501–1511.

56.KollaritschH,Paulke-KorinekM,Dubischar-KastnerK(2009)IC51Japanese

encephalitisvaccine.ExpertOpinBiolTher9:921–931.

57.CentersforDiseaseControlandPrevention(2011)Recommendationsforuse

ofaboosterdoseofinactivatedverocellculture-derivedJapaneseencephalitisvaccine:advisorycommitteeonimmunizationpractices,2011.MMWRMorbMortalWklyRep60:661–663.

58.CentersforDiseaseControlandPrevention(2011)UpdateonJapanese

encephalitisvaccineforchildren:UnitedStates,May2011.MMWRMorbMortalWklyRep60:6–665.

59.ChambersTJ,NestorowiczA,MasonPW,RiceCM(1999)Yellowfever/

Japaneseencephalitischimericviruses:constructionandbiologicalproperties.JVirol73:3095–3101.

60.MonathTP,LevenbookI,SoikeK,ZhangZX,RatterreeM,etal.(2000)

Chimericyellowfevervirus17D-Japaneseencephalitisvirusvaccine:dose-responseeffectivenessandextendedsafetytestinginrhesusmonkeys.JVirol74:1742–1751.

61.HalsteadSB,ThomasSJ(2011)NewJapaneseencephalitisvaccines:

alternativestoproductioninmousebrain.ExpertRevVaccines10:355–3.62.AppaiahgariMB,VratiS(2012)ClinicaldevelopmentofIMOJEV(R)-a

recombinantJapaneseencephalitischimericvaccine(JE-CV).ExpertOpinBiolTher12:1251–1263.

63.AiharaS,RaoCM,YuYX,LeeT,WatanabeK,etal.(1991)Identificationof

mutationsthatoccurredonthegenomeofJapaneseencephalitisvirusduringtheattenuationprocess.VirusGenes5:95–109.

.NiH,ChangGJ,XieH,TrentDW,BarrettAD(1995)Molecularbasisof

attenuationofneurovirulenceofwild-typeJapaneseencephalitisvirusstrainSA14.JGenVirol76:409–413.

65.NitayaphanS,GrantJA,ChangGJ,TrentDW(1990)Nucleotidesequenceof

thevirulentSA-14strainofJapaneseencephalitisvirusanditsattenuatedvaccinederivative,SA-14-14-2.Virology177:541–552.

66.CeciliaD,GouldEA(1991)Nucleotidechangesresponsibleforlossof

neuroinvasivenessinJapaneseencephalitisvirusneutralization-resistantmutants.Virology181:70–77.

67.ChambersTJ,DrollDA,JiangX,WoldWS,NickellsJA(2007)JENakayama/

JESA14-14-2virusstructuralregionintertypicviruses:biologicalpropertiesinthemousemodelofneuroinvasivedisease.Virology366:51–61.

68.HasegawaH,YoshidaM,ShiosakaT,FujitaS,KobayashiY(1992)Mutations

intheenvelopeproteinofJapaneseencephalitisvirusaffectentryintoculturedcellsandvirulenceinmice.Virology191:158–165.

69.LeeE,HallRA,LobigsM(2004)CommonEproteindeterminantsfor

attenuationofglycosaminoglycan-bindingvariantsofJapaneseencephalitisandWestNileviruses.JVirol78:8271–8280.

70.NiH,BarrettAD(1996)Moleculardifferencesbetweenwild-typeJapanese

encephalitisvirusstrainsofhighandlowmouseneuroinvasiveness.JGenVirol77:1449–1455.

71.NiH,BarrettAD(1998)AttenuationofJapaneseencephalitisvirusbyselection

ofitsmousebrainmembranereceptorpreparationescapevariants.Virology241:30–36.

72.ShahPS,TanakaM,KhanAH,MathengeEG,FukeI,etal.(2006)Molecular

characterizationofattenuatedJapaneseencephalitislivevaccinestrainML-17.Vaccine24:402–411.

73.SumiyoshiH,TignorGH,ShopeRE(1995)Characterizationofahighly

attenuatedJapaneseencephalitisvirusgeneratedfrommolecularlyclonedcDNA.JInfectDis171:1144–1151.

74.WuSC,LinCW,LeeSC,LianWC(2003)Phenotypicandgenotypic

characterizationoftheneurovirulenceandneuroinvasivenessofalarge-plaqueattenuatedJapaneseencephalitisvirusisolate.MicrobesInfect5:475–480.75.KimJM,YunSI,SongBH,HahnYS,LeeCH,etal.(2008)AsingleN-linked

glycosylationsiteintheJapaneseencephalitisvirusprMproteiniscriticalforcelltype-specificprMproteinbiogenesis,virusparticlerelease,andpathoge-nicityinmice.JVirol82:7846–7862.

76.YunSI,ChoiYJ,SongBH,LeeYM(2009)3’cis-actingelementsthat

contributetothecompetenceandefficiencyofJapaneseencephalitisvirusgenomereplication:functionalimportanceofsequenceduplications,deletions,andsubstitutions.JVirol83:7909–7930.

77.SongBH,YunGN,KimJK,YunSI,LeeYM(2012)Biologicalandgenetic

propertiesofSA14-14-2,alive-attenuatedJapaneseencephalitisvaccinethatiscurrentlyavailableforhumans.JMicrobiol50:698–706.

78.YunSI,KimSY,RiceCM,LeeYM(2003)Developmentandapplicationofa

reversegeneticssystemforJapaneseencephalitisvirus.JVirol77:50–65.79.NybakkenGE,NelsonCA,ChenBR,DiamondMS,FremontDH(2006)

CrystalstructureoftheWestNilevirusenvelopeglycoprotein.JVirol80:11467–11474.

80.LucaVC,AbiMansourJ,NelsonCA,FremontDH(2012)Crystalstructureof

theJapaneseencephalitisvirusenvelopeprotein.JVirol86:2337–2346.

PLOSPathogens|www.plospathogens.org15July2014|Volume10|Issue7|e1004290

JEVNeurovirulence

81.NiH,BurnsNJ,ChangGJ,ZhangMJ,WillsMR,etal.(1994)Comparisonof

nucleotideanddeducedaminoacidsequenceofthe5’non-codingregionandstructuralproteingenesofthewild-typeJapaneseencephalitisvirusstrainSA14anditsattenuatedvaccinederivatives.JGenVirol75:1505–1510.

82.WorldHealthOrganization(2005)Globaladvisorycommitteeonvaccine

safety,9–10June2005.WklyEpidemiolRec80:242–247.

83.MukhopadhyayS,KuhnRJ,RossmannMG(2005)Astructuralperspectiveof

theflaviviruslifecycle.NatRevMicrobiol3:13–22.

84.StiasnyK,HeinzFX(2006)Flavivirusmembranefusion.JGenVirol87:

2755–2766.

85.ReyFA,HeinzFX,MandlC,KunzC,HarrisonSC(1995)Theenvelope

glycoproteinfromtick-borneencephalitisvirusat2Aresolution.Nature375:291–298.

86.AllisonSL,SchalichJ,StiasnyK,MandlCW,HeinzFX(2001)Mutational

evidenceforaninternalfusionpeptideinflavivirusenvelopeproteinE.JVirol75:4268–4275.

87.LeeE,LobigsM(2002)Mechanismofvirulenceattenuationofglycosamino-glycan-bindingvariantsofJapaneseencephalitisvirusandMurrayValleyencephalitisvirus.JVirol76:4901–4911.

88.LeeJW,ChuJJ,NgML(2006)QuantifyingthespecificbindingbetweenWest

NilevirusenvelopedomainIIIproteinandthecellularreceptoraVb3integrin.JBiolChem281:1352–1360.

.BeasleyDW,BarrettAD(2002)Identificationofneutralizingepitopeswithin

structuraldomainIIIoftheWestNilevirusenvelopeprotein.JVirol76:13097–13100.

90.CrillWD,RoehrigJT(2001)MonoclonalantibodiesthatbindtodomainIIIof

denguevirusEglycoproteinarethemostefficientblockersofvirusadsorptiontoVerocells.JVirol75:7769–7773.

91.KaufmannB,NybakkenGE,ChipmanPR,ZhangW,DiamondMS,etal.

(2006)WestNilevirusincomplexwiththeFabfragmentofaneutralizingmonoclonalantibody.ProcNatlAcadSciUSA103:12400–12404.

92.WuKP,WuCW,TsaoYP,KuoTW,LouYC,etal.(2003)Structuralbasisof

aflavivirusrecognizedbyitsneutralizingantibody:solutionstructureofthedomainIIIoftheJapaneseencephalitisvirusenvelopeprotein.JBiolChem278:46007–46013.

93.StiasnyK,BressanelliS,LepaultJ,ReyFA,HeinzFX(2004)Characterization

ofamembrane-associatedtrimericlow-pH-inducedformoftheclassIIviralfusionproteinEfromtick-borneencephalitisvirusanditscrystallization.JVirol78:3178–3183.

94.ReyFA(2003)Denguevirusenvelopeglycoproteinstructure:newinsightinto

itsinteractionsduringviralentry.ProcNatlAcadSciUSA100:69–6901.95.KaufmannB,ChipmanPR,HoldawayHA,JohnsonS,FremontDH,etal.

(2009)Capturingaflaviviruspre-fusionintermediate.PLoSPathog5:e1000672.

96.LinSR,ZouG,HsiehSC,QingM,TsaiWY,etal.(2011)Thehelicaldomains

ofthestemregionofdenguevirusenvelopeproteinareinvolvedinbothvirusassemblyandentry.JVirol85:5159–5171.

97.PangerlK,HeinzFX,StiasnyK(2011)Mutationalanalysisofthezippering

reactionduringflavivirusmembranefusion.JVirol85:8495–8501.

98.SchmidtAG,YangPL,HarrisonSC(2010)Peptideinhibitorsofdenguevirus

entrytargetalate-stagefusionintermediate.PLoSPathog6:e1000851.

99.ChuJJ,NgML(2004)InteractionofWestNileviruswithavb3integrin

mediatesvirusentryintocells.JBiolChem279:54533–54541.

100.HurrelbrinkRJ,McMinnPC(2001)AttenuationofMurrayValleyencephalitis

virusbysite-directedmutagenesisofthehingeandputativereceptor-bindingregionsoftheenvelopeprotein.JVirol75:7692–7702.

101.LeeE,LobigsM(2000)Substitutionsattheputativereceptor-bindingsiteofan

encephaliticflavivirusaltervirulenceandhostcelltropismandrevealaroleforglycosaminoglycansinentry.JVirol74:8867–8875.102.GotoA,YoshiiK,ObaraM,UekiT,MizutaniT,etal.(2005)RoleoftheN-linkedglycansoftheprMandEenvelopeproteinsintick-borneencephalitisvirusparticlesecretion.Vaccine23:3043–3052.

103.HannaSL,PiersonTC,SanchezMD,AhmedAA,MurtadhaMM,etal.

(2005)N-linkedglycosylationofWestNilevirusenvelopeproteinsinfluencesparticleassemblyandinfectivity.JVirol79:13262–13274.

104.LorenzIC,KartenbeckJ,MezzacasaA,AllisonSL,HeinzFX,etal.(2003)

Intracellularassemblyandsecretionofrecombinantsubviralparticlesfromtick-borneencephalitisvirus.JVirol77:4370–4382.

105.TajimaS,NeromeR,NukuiY,KatoF,TakasakiT,etal.(2010)Asingle

mutationintheJapaneseencephalitisvirusEprotein(S123R)increasesitsgrowthrateinmouseneuroblastomacellsanditspathogenicityinmice.Virology396:298–304.

106.MonathTP,ArroyoJ,LevenbookI,ZhangZX,CatalanJ,etal.(2002)Single

mutationintheflavivirusenvelopeproteinhingeregionincreasesneuroviru-lenceformiceandmonkeysbutdecreasesviscerotropismformonkeys:relevancetodevelopmentandsafetytestingoflive,attenuatedvaccines.JVirol76:1932–1943.

107.ArroyoJ,GuirakhooF,FennerS,ZhangZX,MonathTP,etal.(2001)

Molecularbasisforattenuationofneurovirulenceofayellowfevervirus/Japaneseencephalitisviruschimeravaccine(ChimeriVax-JE).JVirol75:934–942.

108.ChenLK,LinYL,LiaoCL,LinCG,HuangYL,etal.(1996)Generationand

characterizationoforgan-tropismmutantsofJapaneseencephalitisvirusinvivoandinvitro.Virology223:79–88.

109.ZhaoZ,DateT,LiY,KatoT,MiyamotoM,etal.(2005)Characterizationof

theE-138(Glu/Lys)mutationinJapaneseencephalitisvirusbyusingastable,full-length,infectiouscDNAclone.JGenVirol86:2209–2220.

110.MelianEB,HinzmanE,NagasakiT,FirthAE,WillsNM,etal.(2010)NS1’of

flavivirusesintheJapaneseencephalitisvirusserogroupisaproductofribosomalframeshiftingandplaysaroleinviralneuroinvasiveness.JVirol84:11–17.

111.FirthAE,AtkinsJF(2009)AconservedpredictedpseudoknotintheNS2A-encodingsequenceofWestNileandJapaneseencephalitisflavivirusessuggestsNS1’mayderivefromribosomalframeshifting.VirolJ6:14.

112.YeQ,LiXF,ZhaoH,LiSH,DengYQ,etal.(2012)Asinglenucleotide

mutationinNS2AofJapaneseencephalitis-livevaccinevirus(SA14-14-2)ablatesNS1’formationandcontributestoattenuation.JGenVirol93:1959–19.

113.EckelsKH,YuYX,DuboisDR,MarchetteNJ,TrentDW,etal.(1988)

Japaneseencephalitisviruslive-attenuatedvaccine,ChinesestrainSA14-14-2;adaptationtoprimarycaninekidneycellculturesandpreparationofavaccineforhumanuse.Vaccine6:513–518.

114.SigelMM(1952)Influenceofageonsusceptibilitytovirusinfectionswith

particularreferencetolaboratoryanimals.AnnuRevMicrobiol6:247–280.115.MonathTP(1986)Pathobiologyoftheflaviviruses.In:SchlesingerS,

SchlesingerMJ,editors.TheTogaviridaeandFlaviviridae.NewYork,NY:Plenum.pp.375–440.

116.ThompsonJD,GibsonTJ,PlewniakF,JeanmouginF,HigginsDG(1997)The

CLUSTALXwindowsinterface:flexiblestrategiesformultiplesequencealignmentaidedbyqualityanalysistools.NucleicAcidsRes25:4876–4882.117.GoujonM,McWilliamH,LiW,ValentinF,SquizzatoS,etal.(2010)Anew

bioinformaticsanalysistoolsframeworkatEMBL-EBI.NucleicAcidsRes38:W695–699.

118.KieferF,ArnoldK,KunzliM,BordoliL,SchwedeT(2009)TheSWISS-MODELRepositoryandassociatedresources.NucleicAcidsRes37:D387–392.

119.ReedLJ,MuenchH(1938)Asimplemethodofestimatingfiftypercent

endpoints.AmJHyg27:493–497.

PLOSPathogens|www.plospathogens.org16July2014|Volume10|Issue7|e1004290