四种发光甲虫的转录组测序和系统发育分析.pdf 11页

'中国科技论文在线http://www.paper.edu.cnTranscriptomesequencingandphylogeneticanalysisoffour#speciesofluminescentbeetles**5WANGKai,HONGWei,JIAOHengwu,ZHAOHuabin(CollegeofLifeSciences,WuhanUniversity,Wuhan430072)Abstract:Theevolutionofbioluminescencehaspromptedscientificattentiontoilluminatephylogeneticrelationshipsofluminescentbeetles.However,genomicresourcesarevirtuallylackinginrhagophthalmids(Rhagophthalmidae)andtheirrelatedfireflybeetleslampyrids(Lampyridae).Here,10weemployedtheIlluminaHiseq2000platformandsequencedthewhole-bodytranscriptomesofthefourluminescentbeetles:onerhagophthalmid(Rhagophthalmussp.)andthreefireflies(Asymmetricatacircumdata,Aquaticaficta,andPyrocoeliapectoralis).Weobtained55.4,43.4,38.6,and36.7millioncleanreadsforthefourspecies,respectively.Allreadswereassembledintocontigsfromwhichunigeneswerederived.Allunigeneswereannotatedbypubliclyavailabledatabases,andatotalof154325orthologousgeneswereidentified.Usingmultiplephylogeneticapproaches,ourtranscriptomedataconfirmedthedistinctivenessofRhagophthalmidaefromLampyridae.Together,thisstudyisthefirstreportofwholetranscriptomesequencingdatainRhagophthalmidaeandLampyridaespecies,representingavaluablegenomicresourceforstudyingtheoriginandevolutionofsomeremarkabletraitsinthesebeetlessuchasbioluminescence.Moreover,ourtranscriptomedataprovideuseful20phylogeneticinformationthatcouldbeofimportanceinfuturestudiesofphylogeneticinference.Keywords:Rhagophthalmidae,Lampyridae,Phylogenomics,Transcriptome0IntroductionBioluminescenceisamongthemostspectacularfeaturesinlivingorganisms,includingnumerous25speciesofmarinefishes,marineinvertebrates,terrestrialinvertebrates,fungi,bacteria,and[1,2]protists.Theusesofbioluminescenceinnatureinvolvearangeofvitalfunctions:camouflage,[2]attraction,defense,warning,communication,mimicry,andillumination.Thechemicalbasisof[3]thenaturallight-producingmoleculeshasbeenelucidatedinthelastcentury,permittingdiscoveriesofcountlessvaluableapplicationsinbiologyandmedicineusingluciferase-based[2]30systems.Inaddition,industrialdesignershavebeenambitiousinutilizingthenatural[4]light-producingsystemsinbioluminescentorganismsfordecorationandstreetlighting.Amongthesebioluminescentorganisms,luminescentinsectswereprimarilyfoundinmembersof[5]thethreeorders:Collembola(springtails),Diptera(flies),andColeoptera(beetles).Withinthe35superfamilyElateroideaoftheorderColeoptera,severalgroupsofbeetlesareabletoproduceandemitlightsuchasfireflies(Lampyridae),railroadworms(Phengodidae),clickbeetles(Elateridae),[5,6]andRhagophthalmidae.Theoriginandevolutionofbioluminescenceinthesebeetleshaspromptedanumberofstudiestoilluminatethephylogeneticrelationshipsofthesebioluminescent[7-16][10-12,14,16][7-9,13,15,17]beetles.Althoughbothmorphologicalandmolecularfeatureswere40involvedinthesestudies,molecularevidencehasbeenexpectedtoresolvetaxonomicstatusand[18]phylogeneticrelationshipsthathavebeencontentiousbasedonmorphologicalevidence,[18,19]becausetheevolutionaryhistoryofmorphologicalfeaturesisusuallycomplex.Unfortunately,theincongruenceofmolecularevidencehasalsobeenobservedinphylogeneticanalyses[7-9,13,15]concerningthephylogeneticrelationshipsofthesebeetles,possiblybecauseasingleora[20]45fewgeneslacksufficientphylogeneticsignals.Forinstance,Rhagophthalmidaehadpreviously[11]beenassignedtothefamilyPhengodidaeinferredfrommorphologicaldata;otherstudiesbasedonmorphological,embryologicalandmolecularevidencehadassignedRhagophthalmidaetobea[8,14,15,21]subfamilyorgenusinthefamilyLampyridae;butmorerecentevidencefrombothmorphologicalandmoleculardatahadsupportedthefamilialstatusof[7,9,10,13,16,22-24]50Rhagophthalmidae,whichisdistinctfrombothLampyridaeandPhengodidae.InFoundations:TheSpecializedResearchFundfortheDoctoralProgramofHigherEducation(20130141110066),andtheNationalNaturalScienceFoundationofChina(91331115).Briefauthorintroduction:WANGKai(1986-),Male,PhDcandidate,AnimalEvolutionCorrespondanceauthor:ZHAOHuabin(1980-),Male,Professor,AnimalEvolutionandEcologicalGenomics.E-mail:huabinzhao@whu.edu.cn-1- 中国科技论文在线http://www.paper.edu.cnaddition,thetaxonomicstudiesfocusingonLampyridaeandRhagophthalmidaehavebeenscarce,[15,25,26]especiallyforspeciesdistributedinChina.Thescarcitycallsforadditionaldataoftaxonomicimportance.Phylogenomics,i.e.phylogeneticanalysisinvolvinggenome-scaledata,hasbeenbelievedtooutcompetesingle-genephylogenetics,whichfrequentlyyieldedconflicting[20]55resultscausedbystochasticerrorsfromsmall-scaledatasets.However,genomicresourcesareextremelylimitedinRhagophthalmidaeanditsrelatedbeetles,whichmaypreventtheirphylogeneticrelationshipsfromin-depthinvestigations.ToprovidenovelgenomicresourcesfromRhagophthalmidaeanditsrelatedfireflybeetles60(Lampyridae),weemployedtheIlluminaHiseq2000platformandsequencedthewhole-bodytranscriptomesofthefourbeetles:onerhagophthalmidbeetle(Rhagophthalmussp.)andthreerepresentativesoflampyridbeetles(Asymmetricatacircumdata,Aquaticaficta,andPyrocoeliapectoralis).OurtranscriptomeanalysisrecognizedthedistinctivenessofRhagophthalmidaefromLampyridae.651Materialsandmethods1.1EthicsstatementAllbeetlespeciesusedinthisstudyweresampledinthefield.Nospecificpermitswererequired,andnoendangeredorprotectedspecieswereinvolved.Allexperimentsonthebeetleswere70conformedtotherulesandguidelinesonanimalexperimentationinChina.1.2TaxonsamplingandDNAextractionLivelarvaeoradultsofbeetlesstudiedhereweresampledatfivelocationsinChina.Alltheseindividualswerestoredat−80°Cafterfreezinginliquidnitrogen.751.3Whole-bodytranscriptomesequencing(RNA-seq)Followingthemanufacturer’sprotocol,totalRNAswereisolatedusingTrizol(Invitrogen)fromwholebodiesoffourindividuals:oneadultindividualofRhagophthalmussp.andthreelarvaindividualsoffireflybeetles(A.ficta,A.circumdata,andP.pectoralis).Fourpaired-endlibraries80withaninsertsizeofapproximately200bpwereconstructedusingtheIlluminaTruseqRNAsampleprepkitaccordingtothemanufacturer"sprotocol.DetailsofconstructingRNA-seq[19]librarieswerepreviouslydescribed.Alllibrariesweresequencedcommerciallytogeneratepaired-endreadsofaveragelength101-bpontheIlluminaHiSeq2000sequencingplatform.851.4DenovoassemblyandunigeneannotationSequencequalityofeachRNA-seqsamplewasassessedbyFastQCversion10.1(www.bioinformatics.bbsrc.ac.uk/projects/fastqc).Trimmomaticversion0.32wasusedtotrimoutlowqualitysequences,ambiguoussequencesandartificialsequencessuchasresidualadaptorsandIlluminaspecificsequencesasdescribedelsewhere.Fourdenovotranscriptomeassemblieswere90constructedbytheTrinityprogramwithdefaultsettings.Becausedenovotranscriptomeassembliesinvolvemanycontigsinacontigcluster,werequiredaminimumexpressionfilterofonefragmentperkilobaseofexonpermillionfragmentsmapped(FPKM)andfilteredthecontigswithFPKMvaluesmallerthanone;thecontigwiththehighestexpressionlevelinacontigclusterwasselectedtobeaunigenefordownstreamanalyses.Theresultingunigeneswereusedfor95BLASTXsearchesandunigeneannotationsbasedontheNR,Swiss-protandCOGdatabases,withanE-valuecutoffof1e-5.TheGOannotationwasconductedbyBlast2GOsoftwarewithdefaultparametersusingtheNRblastresultsinXMLformat.1.5Orthologidentificationandphylogenomicanalysis100ThecDNAsequencesoftheredflourbeetle(Triboliumcastaneum3.0Assembly)were-2- 中国科技论文在线http://www.paper.edu.cndownloadedfromBeetleBase(http://www.Beetlebase.org),andcomparedwiththetranscriptome-derivedunigenesbyreciprocal(orbi-directional)TBLASTXsearches.Weidentifiedputativeorthologsbyexaminingreciprocalbestblasthits.WediscardedputativeorthologswithaBLASTscoreratioofthesecondbest-hittothefirstbest-hitgreaterthan0.8,105whichcanexcludepotentialparalogsinourphylogeneticanalysis.Allputativeorthologswere[28]alignedbyPRANKversion100802,andpoorlyalignedpositionsanddivergentregionswere[29]removedbyGBLOCKSversion0.91b.Inaddition,thealignmentswithanalignedregionshorterthan100-ntwerediscarded.110Inaddition,weuseddeducedproteinsequencesinourphylogenomicanalysis,aimingtoreducetheimpactofnucleotidecompositionalbias.Briefly,proteinsequencesofthenucleargeneswerededucedandalignedbyMEGAversion5.20.ProtTestversion3.41wasappliedtoselectthebest-fitmodelofproteinsequenceevolutionfollowingtheAkaikeinformationcriterion(AIC).TheLG+I+Gmodelwasdeterminedtoprovidethebestfittotheconcatenatedproteinsequences115ofnucleargenes.PhylogenetictreeswerereconstructedbyMrBayesversion3.2.6withasmanygenerationsasrequired.2Results2.1Transcriptomesequencingandassembly120TotalRNAwasisolatedfromeachofthefrozenwhole-bodiesofbeetles.mRNAwaspurifiedfromthetotalRNAusingoligo-dTattachedmagneticbeads.FollowingmRNAfragmentation,cDNAsynthesis,endrepair,3’adenylation,adaptorligation,andPCRenrichment,fourRNAsequencing(RNA-Seq)librarieswereconstructedusingtheIlluminaTruSeqRNAsamplepreparationkit.Therawsequencedatawerefilteredbyremovingadaptors,low-qualityreads,and125ambiguousreads.Weultimatelyobtained55.4,43.4,38.6,and36.7millioncleanreadsforRhagophthalmussp.,A.circumdata,A.ficta,andP.pectoralis,respectively(Table1).Thesecleanreadswereseparatelyassembledinto44883,57254,71424,and80017contigs,inwhich15418,24275,31520,and31356unigenes(i.e.uniqueputativegenes)werederived,respectively(Table1).Itappearsthatthetranscriptomedataingreatersizetendtohavelessassembledcontigs130orunigenes.Forexample,thelargestdata(Rhagophthalmussp.)containstheleastcontigs,whilethesmallestdata(P.pectoralis)consistsofthemostcontigs(Table1).Detailsofcleanreads,assembledcontigsandunigenesofthefourtranscriptomesgeneratedinthisstudyweregiveninTable1.Inaddition,wecomparedthe13PCGsinA.fictamitogenomeidentifiedfromtheIllumina-basedtranscriptomeassemblywiththesamegenesobtainedfromthetraditionalSanger135sequencingforanotherindividualofthesamespecies,andfoundthatatotalof99.3%sequencesfrombothsequencingapproachesareidentical.Thisfindingsuggeststhatourtranscriptomeassembliesarereliableandoursequencingcoveragesareacceptable.2.2Functionalannotationofunigenes140Atotalof10345(percentageofallunigenesinaspecies:67.10%),15352(63.24%),17480(55.46%),and15342(48.93%)unigenesofRhagophthalmussp.,A.circumdata,A.ficta,andP.pectoraliswererespectivelyannotatedbyatleastoneofthepubliclyavailabledatabases:Swiss-prot,NCBInon-redundantprotein(NR),ClustersofOrthologousGroupsofproteins(COG),andGeneOntology(GO)(Table2).WeusedallunigenesasqueryandrantheTBLASTX145programtosearchagainsttheSwiss-prot,NR,COG,andGOdatabaseswithanE-valueof1.0E-5.Forexample,weidentified10322bestblasthitsintheNRdatabasewhenannotatingthetranscriptomeofRhagophthalmussp.,comprising66.95%ofallunigenes(Table2);whileonly4845besthitsweredetectedintheGOdatabaseforthesamespecies,covering31.42%ofallunigenes(Table2).Indeed,theNRdatabasegenerallyannotatedthehighestpercentagesof150unigenes,whereastheGOdatabasetypicallyannotatedthelowest(Table2).Overall,Rhagophthalmussp.hasahigherpercentageofunigenesbeingannotatedcomparedtotheotherthreebeetles(Table2),possiblybecauseagreatersizeofsequencingdatacangiveabetterassemblyfortheRhagophthalmusbeetle(Table1).Inaddition,thedistributionsofbothGOterms-3- 中国科技论文在线http://www.paper.edu.cnandCOGclassesweresimilaramongthefourtranscriptomes,exceptA.circumdatashowinga155uniqueexpansioninasingleGOterm(Fig.1).AAsymmetricatacircumdataAquaticafictaPyrocoeliapectoralisRhagophthalmussp.transportregulationoftranscription,DNA-templatedBiologicalProcessoxidation-reductionprocessmetabolicprocessregulationofcyclin-dependentproteinSer/ThrkinaseactivityGTPbindingstructuralconstituentofribosometransferaseactivitycatalyticactivitycalciumionbindinghydrolaseactivityoxidoreductaseactivityMolecularFunctionDNAbindingnucleotidebindingnucleicacidbindingzincionbindingbindingmetalionbindingATPbindingcytoplasmintegralcomponentofmembraneCellularComponentmembranenucleusintracellular01234PercentageofunigenenumberB6000A:RNAprocessingandmodificationB:ChromatinstructureanddynamicsAsymmetricatacircumdataC:EnergyproductionandconversionAquaticafictaD:Cellcyclecontrol,celldivision,chromosomepartitioning5000PyrocoeliapectoralisE:AminoacidtransportandmetabolismRhagophthalmussp.F:NucleotidetransportandmetabolismG:CarbohydratetransportandmetabolismH:Coenzymetransportandmetabolism4000I:LipidtransportandmetabolismJ:Translation,ribosomalstructureandbiogenesisigenesK:TranscriptionunL:Replication,recombinationandrepairf3000M:Cellwall/membrane/envelopebiogenesisoN:CellmotilityberO:Posttranslationalmodification,proteinturnover,chaperonesmP:Inorganiciontransportandmetabolismu2000Q:Secondarymetabolitesbiosynthesis,transportandcatabolismNR:GeneralfunctionpredictiononlyS:FunctionunknownT:Signaltransductionmechanisms1000U:Intracellulartrafficking,secretion,andvesiculartransportV:DefensemechanismsW:ExtracellularstructuresY:Nuclearstructure0Z:CytoskeletonABCDEFGHIJKLMNOPQRSTUVWYZCOGfunctionalclassFigure1.FunctionalclassificationofunigenederivedfromthefourtranscriptomesbasedontheGO(GeneOntology)annotation(A)andtheCOG(ClusterofOrthologousGroups)annotation(B).Abbreviations:Rha,Rhagophthalmussp.;AC,Asymmetricatacircumdata;AF,Aquaticaficta;PP,Pyrocoeliapectoralis.1602.3OrthologidentificationGenesfromthereferencegenomeoftheredflourbeetleTriboliumcastaneumwererespectivelycomparedwithallunigenesfromeachofthefourtranscriptomeassembliesusingtheTBLASTX[27]searches.Byexaminingreciprocal(orbi-directional)bestblasthits,putativeorthologswere165identified.IftheBLASTscoreratioofthesecondbest-hittothefirstbest-hitisgreaterthan0.8,weexcludedtheputativeorthologforfurtheranalysis,becausethecandidateorthologispossiblyaparalogduetothehighlevelofsequencesimilarity.Theresultingputativeorthologsrangefrom6992to7612(Table2).Eachorthologfromtheredflourbeetleandthefourbeetleswith-4- 中国科技论文在线http://www.paper.edu.cntranscriptomedatawasalignedbyPRANKversion100802[28],andpoorlyalignedpositionsand[29]170divergentregionsinalignmentswerefilteredbyGBLOCKSversion0.91b.Afterdiscardingthealignmentswithanalignedregionshorterthan100nt(nucleotides),weobtained4325putativeorthologswithhigh-qualityalignmentsforsubsequentanalyses.Table1.Statisticsofcleanreads,assembledcontigsandunigenesofthefourtranscriptomes175generatedinthisstudy.Abbreviation:nt,nucleotide(s).SpeciesRhagophthalmussp.A.circumdataA.fictaP.pectoralisCleanreadsNumberofreads55,374,19443,437,50438,642,10236,663,932Numberofbases(nt)5,496,107,2763,991,132,6863,909,904,6303,532,502,346Meanlength(nt)99969696ContigsNumber44,88357,25471,42480,017Meanlength(nt)1,4611,0871,1661,283N50statistics(nt)2,9812,0162,4612,622UnigenesNumber15,41824,27531,52031,356Meanlength(nt)1,451958838923N50statistics(nt)2,1971,5561,4371,690Table2.Summaryofunigeneannotationsforthefourtranscriptomes.Percentagesofallunigenesinagivenspecieswereshowninparentheses.Abbreviations:NR,NCBInon-redundantprotein;COG,Clustersof180OrthologousGroupsofproteins;GO,GeneOntology.DatabaseRhagophthalmussp.A.circumdataA.fictaP.pectoralisSwiss-prot8,447(54.79%)11,171(46.02%)12,042(38.20%)11,263(35.92%)NR10,322(66.95%)15,315(63.09%)17,432(55.30%)15,304(48.81%)GO4,845(31.42%)6,403(26.38%)7,235(22.95%)6,428(20.50%)COG9,027(58.55%)12,197(52.25%)13,357(42.38%)12,415(39.59%)Total10,345(67.10%)15,352(63.24%)17,480(55.46%)15,342(48.93%)2.4PhylogeneticanalysisbasedonnucleargenesequencesPhylogeneticanalysiswasfirstundertakenwithtranscriptome-derivednucleargenesequences.GiventhatLampyridaewasdividedintotwomonophyleticclades,withonecladeconsistingof[13]185LampyrinaeandtheothercladecomprisingCyphonocerinae,Ototetrinae,Luciolinae,weselectedonespeciesP.pectoralis(Lampyrinae)fromthefirstcladeandtwospeciesA.ficta(Luciolinae)andA.circumdata(Luciolinae)fromthesecondcladetoperformwhole-bodytranscriptomesequencing,inadditiontothespeciesRhagophthalmussp..Theresultingtranscriptomedatawereassembledandannotated,andatotalof4325putativeorthologswere190identifiedaftercarefulfilteringofsequencingartifacts.Adatasetisconsideredtobemutationallysaturatedwhenmultiplesubstitutionsaredominating.Substitutionsaturationdecreases[30]phylogeneticinformation,andahighlysaturateddatasetwillproduceanincorrectphylogeny.Toassesstheimpactofsubstitutionsaturation,werandomlyselected100nuclearorthologsandperformedsubstitutionsaturationtests.Resultsshowedthatthethirdcodonsitesofmostnuclear195orthologshaveundergonesubstantialsubstitutionsaturation,wethusonlyusedthefirstandsecondcodonsitesofthe4325nuclearorthologsforsubsequentphylogeneticanalysiswiththe[31]concatenationmethod.Basedonthe1854774-ntconcatenatedalignmentofnuclearorthologsfromthefourspecieswithtranscriptomedataandtheredflourbeetlewithreferencegenomedata,weusedbothMaximumLikelihood(ML)andBayesianInference(BI)approachestoreconstruct-5- 中国科技论文在线http://www.paper.edu.cn200phylogenetictrees.BothmaximumlikelihoodbootstrapvaluesandBayesianinferenceposteriorprobabilityvalues(shownaspercentages)were100%atallnodesofthephylogenetictree(Fig.2).Specifically,bothspeciesfromLuciolinaewerefirstclusteredintoasisterclade,whichnextgroupedwiththesinglespeciesfromLampyrinae(Fig.2).Allthethreefireflybeetles(Lampyridae)formedamonophyleticgroup,andthesinglespeciesfromRhagophthalmidae205appearedtolocateoutsideofLampyridae(Fig.2).Toreducetheinfluenceofgenetreeheterogeneity,wealsoundertookphylogeneticanalysiswith[32]thecoalescentmodelusingtheMP-ESTmethod.Basedonthealignmentofeachorthologousgenewithallthreecodonpositions,theMLgenetreeswith100bootstrapreplicateswere[33]210respectivelyreconstructedbyPhyMLversion3.0usingthebest-fitmodelsgeneratedby[34]jModelTestversion2.1.4.Becausepoorlysupportedindividualgenetreescouldreduce[35,36]phylogeneticsignalsininferringspeciestree,only2,555geneswithaveragebootstrapsupportvaluesgreaterthan70%wereusedforMP-ESTtreereconstruction.ResultsshowedthatthetreetopologyandbootstrapvaluesoftheMP-ESTtreeusingthecoalescentmethodwere215identicaltothoseoftheMLandBItreesreconstructedusingtheconcatenationmethod(Fig.2).Forcomparison,weadditionallyreconstructedaphylogenyusingabioluminescencerelatedgene.Becausebioluminescenceinbeetlesisproducedbycatalyzingtheoxidationofluciferinby[37]luciferase,theluciferasegeneispreferableforthisanalysis.Weusedtheluciferasegeneof220Lampyrisnoctiluca(Genbankaccession:X89479)asthequerytosearchagainsttheassembledtranscriptsofeachtranscriptomeusingTBLASTN.Thebesthitfromeachtranscriptomewasselectedasacandidate,andcandidateluciferasegenesofthefourbeetlesweresearchedagainstNR(non-redundantdatabase)toensureaccuracyingeneidentification.TakingtheluciferasegeneofTriboliumcastaneumastheoutgroup,wegeneratedanalignmentusingthefiveluciferase[19,38-43]225genes,andphylogenetictreeswerebuiltusingsimilarapproachesasdescribedelsewhere.Theresultingtreetopologieswereidenticaltothoseinferredfrom4325genes(Fig.2).Asymmetricatacircumdata100/100/100LuciolinaeAquaticafictaLampyridae100/100/100PyrocoeliapectoralisLampyrinae100/100/100Rhagophthalmussp.RhagophthalmidaeTriboliumcastaneumTenebrionidae2300.02Figure2.PhylogeneticrelationshipsbetweenRhagophthalmidaeandLampyridaeinferredfromtheconcatenated4325nucleargenemarkerswiththethirdcodonpositionsremoved,Triboliumcastaneumwaschosentobetheoutgroup.NumbersateachnodearetheMLbootstrapvalues/Bayesianposteriorprobabilities/MP-ESTbootstrapvalues,shownaspercentages.235-6- 中国科技论文在线http://www.paper.edu.cn3DiscussionInthisstudy,wepresentthefirstwholetranscriptomeshotgunsequencingdatainRhagophthalmidaeandLampyridaespeciesusingmassiveparallelmRNAsequencing(RNA-seq),providingvaluablegenomeresourcesforstudyingtheevolutionofintriguingtraitsinthesebeetles240suchasbioluminescence.Usingvariousphylogeneticapproaches,ourtranscriptomedataunambiguouslydemonstrateRhagophthalmidaebeingdistinctfromLampyridaeandrejectprevioushypothesessupportingRhagophthalmidaeasasubgroupwithinLampyridae.OurmolecularphylogeneticstudysupportsthefamilialstatusofRhagophthalmidae.245ThephylogeneticpositionofRhagophthalmidaewascontroversialduetoconflictingresultsinferredfrommolecularandmorphologicaldata.Forexample,Rhagophthalmidaehadbeen[11]assignedtothefamilyPhengodidaeinferredfrommorphologicaldata;RhagophthalmidaehadbeenconsideredtobeasubfamilyorgenusinthefamilyLampyridaebasedonmorphological,[8,12,14,15]embryologicalandmolecularevidence.However,theoverwhelmingmolecularevidence250hasconsistentlyrecognizedthedistinctivenessofRhagophthalmidae,whichappearedtobe[7,9,13,17,23,24,26,44-46]differentfrombothLampyridaeandPhengodidae.Asaresult,thefamilialstatusofRhagophthalmidaehasbeenwellestablished.Inthisstudy,weused4325nucleargenesderivedfromtranscriptomedatatoconductphylogeneticinference.Ouranalysesbasedonvariousanalyticalapproachesandseveraldatasetsconsistentlyrecognizedthedistinctivenessof255Rhagophthalmidae(Fig.2).Althoughourstudydidnotinvolvemultipleindividualsorpopulationsfromeachspecies,intraspecificgeneticvariationshouldnotimpactthephylogeneticresolutionatthefamilyorsubfamilylevel.Althoughoursamplesarelimitedduetothelackofgeneticmaterial,weincludedtaxathatmayrepresentbasallineages.Moreimportantly,ourphylogeneticresolutionofRhagophthalmidaeremainedconsistentafterusingdifferentanalytical260approachesandmultipledatasets(Fig.2).Insupportofourfindings,morphologicalevidence[10,16]hasidentifiedthedistinctivenessofRhagophthalmidaefromLampyridaeandPhengodidae,[9,13]andmolecularevidencehasrevealedthatRhagophthalmidaeisasistergrouptoPhengodidae.Incontrasttoourfindings,twomolecularstudiesarguedthatRhagophthalmidaeshouldbeplaced[8,15]withinLampyridae.However,thisargumentwasinferredfromashort16SrRNAgene[8,15][9,13]265segmentwithalengthof506-ntandcannotbesupportedbymulti-locusdatasets.[8,15]Althoughbroadsamplingoftaxaisimportantinmolecularphylogeneticstudies,sufficient[35,47]numberofgenelociisalsorequiredtobuildareliablephylogenetictree.Thisfindingisconsistentwitharecentphylogeneticanalysisbasedonmitogenomes[9]butinconsistentwith[13]anotherstudybasedonacombinationofmitochondrialandnucleargenes.Thecontrasting270resultsbetweenmitochondrialandnucleargenescallforanin-depthstudybyaddingmoretaxaandmoregenome-scaledata,whichwillhelpreconstructareliablephylogenyandmakeaconclusiveinferenceontheevolutionofbioluminescenceinbeetles.Phylogenomicsreferstophylogeneticanalysisinvolvinggenome-scaledata.Phylogenomicshas275beenbelievedtooutcompetesingle-genephylogenetics,whichfrequentlyyieldedconflicting[20,47]resultscausedbystochasticerrorsfromsmall-scaledatasets.However,systematicerrorsare[48]stillpresentafteraddingmoredata.Therearethreemajorchallengesthatcouldgeneratestrongincongruenceinphylogenomicanalysis:substitutionsaturation,nucleotidecompositionalbias,[20]anddifferenttreereconstructionmethods.Wetookthefollowingstepstoovercomethese280challenges.First,weundertooksubstitutionsaturationtestsandremovedthethirdcodonpositionsofallgenesthatmayhaveundergonesubstantialsubstitutionsaturation.ThenucleargenedatasetwiththefirstandsecondcodonpositionsconsistentlysupportedRhagophthalmidaetobeanindependentgroupthatisdistinctfromLampyridae(Fig.2).Second,weusedthededucedproteinsequencestoreducenucleotidecompositionalbias.OurresultshowedthattheBItreetopologies285inferredfromproteinsequenceswereidenticaltothoseinferredfromnucleotidesequences(Fig.2),suggestingthatnucleotidecompositionalbiasisnotamajorfactoraffectingourphylogeneticanalysis.Third,weusedbothMLandBImethodswithnucleargenedataset,andidentifiednoincongruencebetweenthetwotreereconstructionmethods(Fig.2).Inaddition,giventhe[49-51]prevalenceofgenetreeheterogeneity,wealsoconductedphylogeneticanalysiswiththe290coalescentmethod,whichhasbeenprovedtogenerateaccurateandcongruentphylogeniesinthe[52,53]presenceofheterogeneousgenetrees.ThecoalescentmethodimplementedintheMP-EST[32]programwasappliedtoourtranscriptome-derivednucleargenedatasetof2555putative-7- 中国科技论文在线http://www.paper.edu.cnorthologousgenes,andyieldedaphylogenetictreeidenticaltotheconcatenation-basedMLandBItrees(Fig.2).Indeed,afterexaminingtheproportionsof14possibletopologiesinferredfrom295the2555orthologousgenes,wefoundthatmostgenes(75.7%)supportedthetopologyshowninFig.2andtheaverageproportionofother13topologiesis1.9%(Fig.3),suggestingalowlevelofgenetreeheterogeneityamongbeetlesstudiedinthiswork.Thisstudyprovedusefulnessofourgenome-scaledatainphylogeneticanalysis,andwillhelptoilluminatetheoriginandevolutionofbioluminescenceinLampyridaeandotherluminescentbeetles.3000.800.750.700.65oportion0.15rP0.100.050.00Figure3.Ahistogramdepictingtheproportionsof14treetopologieswithaveragebootstrapvaluesabove70%,inferredfrom2555orthologousgenesintheMP-ESTanalysis.Rha,Rhagophthalmussp.;AC,Asymmetricatacircumdata;AF,Aquaticaficta;PP,Pyrocoeliapectoralis.3054ConclusionInthispaper,wepresentthefirstwholetranscriptomeshotgunsequencingdatainRhagophthalmidaeandLampyridaespeciesusingmassiveparallelmRNAsequencing(RNA-seq),310providingvaluablegenomeresourcesforstudyingtheevolutionofintriguingtraitsinthesebeetlessuchasbioluminescence.Usingvariousphylogeneticapproaches,ourtranscriptomedataunambiguouslydemonstrateRhagophthalmidaebeingdistinctfromLampyridaeandrejectprevioushypothesessupportingRhagophthalmidaeasasubgroupwithinLampyridae.OurmolecularphylogeneticstudysupportsthefamilialstatusofRhagophthalmidae.315Acknowledgements(Optional)SpecialthanksshouldgotoDr.XinhuaFu(HuazhongAgriculturalUniversity)whokindlyprovidedmostbeetlesamples.WethankDr.ChengquanCao(LeshanNormalUniversity)forcollectingthesampleofLamprigerayunnana.WealsothankFeng-JuanMu,YiWang,andLiang320Ao(WuhanUniversity)fortechnicalassistanceinthelaboratory.ThisworkwassupportedinpartbytheSpecializedResearchFundfortheDoctoralProgramofHigherEducation(20130141110066),andtheNationalNaturalScienceFoundationofChina(91331115).-8- 中国科技论文在线http://www.paper.edu.cn325References1Herring,P.J.Systematicdistributionofbioluminescenceinlivingorganisms.J.Biolumin.Chemilumin.1,147-163(1987).2Haddock,S.H.,Moline,M.A.&Case,J.F.Bioluminescenceinthesea.Ann.Rev.Mar.Sci.2,443-493(2010).3303Poisson,J.RaphaelDubois,frompharmacytobioluminescence.Rev.Hist.Pharm.(Paris)58,51-56(2010).4Callaway,E.Glowingplantssparkdebate.Nature498,15-16(2013).5Lloyd,J.E.Insectbioluminescence,pp.241–272,In:BioluminescenceinAction,HerringP.(ed.).(AcademicPress,1978).3356Viviani,V.R.Theorigin,diversity,andstructurefunctionrelationshipsofinsectluciferases.Cell.Mol.LifeSci.59,1833-1850(2002).7Bocakova,M.,Bocak,L.,Hunt,T.,Teravainen,M.&Vogler,A.P.MolecularphylogeneticsofElateriformia(Coleoptera):evolutionofbioluminescenceandneoteny.Cladistics23,477-496(2007).8Li,X.Y.,Yang,S.,Xie,M.&Liang,X.C.Phylogenyoffireflies(Coleoptera:Lampyridae)inferred340frommitochondrial16SribosomalDNA,withreferencestomorphologicalandethologicaltraits.Progr.Nat.Sci.16,817-826(2006).9Amaral,D.T.,Mitani,Y.,Ohmiya,Y.&Viviani,V.R.OrganizationandcomparativeanalysisofthemitochondrialgenomesofbioluminescentElateroidea(Coleoptera:Polyphaga).Gene586,254-262(2016).34510Branham,M.A.&Wenzel,J.W.Theevolutionofbioluminescenceincantharoids(Coleoptera:Elateroidea).Fla.Entomol.84,565-586(2001).11Crowson,R.A.AreviewoftheclassificationofCantharoidea(Coleoptera)withthedefinitionoftwonewfamilies,CneoglossidaeandOmethidae.Rev.Univ.Madrid21,35–77(1972).12Kobayashi,Y.,Suzuiki,H.&Ohba,N.Formationofasphericalgermrudimentintheglow-worm350RhagophthalmusohbaiWittmer(Coleoptera:Rhagophthalmidae)anditsphylogeneticimplications.Proc.Arthropod.Embryol.Soc.Jpn.36,1-5(2001).13Kundrata,R.,Bocakova,M.&Bocak,L.ThecomprehensivephylogenyofthesuperfamilyElateroidea(Coleoptera:Elateriformia).Mol.Phylogenet.Evol.76,162-171(2014).14McDermott,F.A.ThetaxonomyoftheLampyridae(Coleoptera).Trans.Am.Entomol.Soc.90,1-72355(1964).15Stanger-Hall,K.F.,Lloyd,J.E.&Hillis,D.M.PhylogenyofNorthAmericanfireflies(Coleoptera:Lampyridae):Implicationsfortheevolutionoflightsignals.Mol.Phylogenet.Evol.45,33-49(2007).16Wittmer,W.&Ohba,N.NeueRhagophthalmidae(Coleoptera)ausChinaundbenachbartenLändern.Jpn.Entomol.62,341–355(1994).36017Kundrata,R.&Bocak,L.ThephylogenyandlimitsofElateridae(Insecta,Coleoptera):isthereacommontendencyofclickbeetlestosoft-bodiednessandneoteny?Zool.Scr.40,364-378(2011).18Nei,M.&Kumar,S.MolecularEvolutionandPhylogenetics.(OxfordUniversityPress,2000).19Lin,G.H.etal.TranscriptomesequencingandphylogenomicresolutionwithinSpalacidae(Rodentia).BMCGenomics15,32(2014).36520Jeffroy,O.,Brinkmann,H.,Delsuc,F.&Philippe,H.Phylogenomics:thebeginningofincongruence?TrendsGenet.22,225-231(2006).21Kobayashi,Y.,Suzuiki,H.&Ohba,N.ormationofasphericalgermrudimentintheglow-wormRhagophthalmusohbaiWittmer(Coleoptera:Rhagophthalmidae)anditsphylogeneticimplications.Proc.Arthropod.Embryol.Soc.Jpn.36,1-5(2001).37022Hunt,T.etal.Acomprehensivephylogenyofbeetlesrevealstheevolutionaryoriginsofasuperradiation.Science318,1913-1916(2007).23Timmermans,M.J.etal.Whybarcode?High-throughputmultiplexsequencingofmitochondrialgenomesformolecularsystematics.NucleicAcidsRes.38,e197(2010).24Bocak,L.etal.BuildingtheColeopteratree-of-lifefor>8000species:compositionofpublicDNAdata375andfitwithLinnaeanclassification.Syst.Entomol.39,97-110(2014).25Li,X.,Ogoh,K.,Ohba,N.,Liang,X.C.&Ohmiya,Y.Mitochondrialgenomesoftwoluminousbeetles,RhagophthalmuslufengensisandR.ohbai(Arthropoda,Insecta,Coleoptera).Gene392,196-205(2007).26Timmermans,M.J.etal.Family-LevelSamplingofMitochondrialGenomesinColeoptera:CompositionalHeterogeneityandPhylogenetics.GenomeBiolEvol8,161-175(2015).38027Altschul,S.F.,Gish,W.,Miller,W.,Myers,E.W.&Lipman,D.J.Basiclocalalignmentsearchtool.J.Mol.Biol.215,403-410(1990).28Loytynoja,A.&Goldman,N.Phylogeny-awaregapplacementpreventserrorsinsequencealignmentandevolutionaryanalysis.Science320,1632-1635(2008).29Castresana,J.Selectionofconservedblocksfrommultiplealignmentsfortheiruseinphylogenetic385analysis.Mol.Biol.Evol.17,540-552(2000).30Jeffroy,O.,Brinkmann,H.,Delsuc,F.&Philippe,H.Phylogenomics:thebeginningofincongruence?TrendsGenet.22,225-231(2006).-9- 中国科技论文在线http://www.paper.edu.cn31William,J.&Ballard,O.Combiningdatainphylogeneticanalysis.TrendsEcol.Evol.11,334(1996).32Liu,L.,Yu,L.&Edwards,S.V.Amaximumpseudo-likelihoodapproachforestimatingspeciestrees390underthecoalescentmodel.BMCEvol.Biol.10,302(2010).33Guindon,S.etal.Newalgorithmsandmethodstoestimatemaximum-likelihoodphylogenies:assessingtheperformanceofPhyML3.0.Syst.Biol.59,307-321(2010).34Darriba,D.,Taboada,G.L.,Doallo,R.&Posada,D.jModelTest2:moremodels,newheuristicsandparallelcomputing.Nat.Methods9,772-772(2012).39535Song,S.,Liu,L.,Edwards,S.V.&Wu,S.Y.Resolvingconflictineutherianmammalphylogenyusingphylogenomicsandthemultispeciescoalescentmodel.Proc.Natl.Acad.Sci.USA109,14942-14947(2012).36Zhong,B.,Liu,L.,Yan,Z.&Penny,D.Originoflandplantsusingthemultispeciescoalescentmodel.TrendsPlantSci.18,492-495(2013).40037Marques,S.M.&EstevesdaSilva,J.C.Fireflybioluminescence:amechanisticapproachofluciferasecatalyzedreactions.IUBMBLife61,6-17(2009).38Hong,W.&Zhao,H.Vampirebatsexhibitevolutionaryreductionofbittertastereceptorgenescommontootherbats.Proc.Biol.Sci.281,20141079(2014).39Feng,P.,Zheng,J.S.,Rossiter,S.J.,Wang,D.&Zhao,H.Massivelossesoftastereceptorgenesin405toothedandbaleenwhales.GenomeBiolEvol6,1254-1265(2014).40Zhao,H.etal.Theevolutionofcolorvisioninnocturnalmammals.ProcNatlAcadSciUSA106,8980-8985(2009).41Wang,K.&Zhao,H.Birdsgenerallycarryasmallrepertoireofbittertastereceptorgenes.GenomeBiolEvol7,2705-2715(2015).41042Zhang,Y.,Jiao,Y.,Jiao,H.,Zhao,H.&Zhu,Y.X.Two-stepfunctionalinnovationofthestem-cellfactorsWUS/WOX5duringplantevolution.Mol.Biol.Evol.,Forthcoming(2017).43Lu,Q.,Wang,K.,Lei,F.,Yu,D.&Zhao,H.Penguinsreducedolfactoryreceptorgenescommontootherwaterbirds.ScientificReports6,31671(2016).44Hunt,T.etal.Acomprehensivephylogenyofbeetlesrevealstheevolutionaryoriginsofasuperradiation.415Science318,1913-1916(2007).45Mckenna,D.D.etal.ThebeetletreeofliferevealsthatColeopterasurvivedend-PermianmassextinctiontodiversifyduringtheCretaceousterrestrialrevolution.Syst.Entomol.40,835-880(2015).46Yuan,M.L.etal.High-levelphylogenyoftheColeopterainferredwithmitochondrialgenomesequences.Mol.Phylogenet.Evol.104,99-111(2016).42047Philippe,H.,Delsuc,F.,Brinkmann,H.&Lartillot,N.Phylogenomics.Ann.Rev.Ecol.Evol.Syst.36,541-562(2005).48Felsenstein,J.Casesinwhichparsimonyorcompatibilitymethodswillbepositivelymisleading.Syst.Zool.27,401–410(1978).49Edwards,S.V.IsaNewandGeneralTheoryofMolecularSystematicsEmerging?Evolution63,1-19425(2009).50Liu,L.,Yu,L.,Kubatko,L.,Pearl,D.K.&Edwards,S.V.Coalescentmethodsforestimatingphylogenetictrees.Mol.Phylogenet.Evol.53,320-328(2009).51Belfiore,N.M.,Liu,L.&Moritz,C.MultilocusphylogeneticsofarapidradiationinthegenusThomomys(Rodentia:Geomyidae).Syst.Biol.57,294-310(2008).43052Kubatko,L.S.&Degnan,J.H.Inconsistencyofphylogeneticestimatesfromconcatenateddataundercoalescence.Syst.Biol.56,17-24(2007).53Degnan,J.H.&Rosenberg,N.A.Genetreediscordance,phylogeneticinferenceandthemultispeciescoalescent.TrendsEcol.Evol.24,332-340(2009).435四种发光甲虫的转录组测序和系统发育分析汪凯，洪炜，焦恒武，赵华斌（武汉大学生命科学学院，武汉430072）440摘要：生物发光的起源和演化具有重要的科学意义和应用潜能，因此，发光甲虫的系统发育研究引起了很多学者的关注。然而，发光甲虫的基因组数据非常稀少，限制了其系统发育研究。本文利用二代测序技术，对4种发光甲虫（1种雌光萤、3种萤火虫）进行了全身转录组测序；对转录组进行组装和注释后，共鉴定出4325个严格的直系同源基因，可用于后续的系统发育重建。使用最大似然法和贝叶斯推断等方法重建系统发育关系，证实了雌光445萤与萤科的萤火虫具有显著不同的系统位置，支持将雌光萤独立成一个科。总之，本研究第一次报道了雌光萤科和萤科昆虫的转录组数据，为研究发光甲虫的起源和演化提供了重-10- 中国科技论文在线http://www.paper.edu.cn要的基因组资源，也证实了这些转录组数据可以为将来的系统发育研究提供支撑。关键词：转录组；系统发育；萤火虫；甲虫中图分类号：Q951+.3450-11-'