基于線粒體dna的牡丹組種間關(guān)系研究

基于線粒體dna的牡丹組種間關(guān)系研究

與2000年和500年相比，遺傳因素被用作樣本變量?，F(xiàn)在，從這一過程中，我們可以充分地看到中國和世界上的遺傳因素，或名稱和生態(tài)價值（虎田，1999a；lanetal，2002）。TreepeonybelongstothesectionMoutanDC.othegenusPaeoniaL.(Paeoniaceae)(Stern,1946;Pan1979).Inadditiontothiswoodygroup,therearetwoherbaceoussectionsinthisgenus,sect.Paeoniaandsect.Onaepia(Stern,1946;Pan,1995).ThegenusPaeoniaconsistsofmorethan30species,distributedwidelyinthetemperateregionoftheworld(Pan1995).ThesectionOnaepiahasonlytwospeciesendemictoNorthAmerica,whereassect.Paeoniacomprisesabout22speciesfoundinEuropeNorth-westAfricaandAsia,spreadingfromPortugaandMoroccotoJapan(Stern,1946;Pan,1995).Thesect.Moutan,followingarecentclassificationsystemofHong&Pan(1999a),containseightspeciesandthreeofthemeachhavetwosubspecies,distributedinsouthwestern,centralandnorthernregionofChina(Hong&Pan,1999a).Inthepastdecade,importantprogresshasbeenmadeinthetaxonomyoftreepeony(Xi,1984;Hongetal.,1988;Hong&Pan,1999a;Zhou,2006).However,althoughtheinterspecificrelationshipsofwildtreepeonieshavebeenpreviouslyinvestigatedusingmorphologicaldata(Hong,1997a;Zhouetal.,2003)molecularmarker(Zouetal.,1999),andDNAfragmentsfromdifferentgenomes(Sangetal.,19951997a,b;Sang&Zhang,1999;Ferguson&Sang2001;Tank&Sang,2001;Linetal.,2004;Zhaoetal.,2004),thespeciesphylogenyofthesect.Moutanisstillpoorlyunderstood.Especially,theinterspecificrelationshipamongthesixspecies(P.suffruticosa,P.ostii,P.jishanesis,P.qiui,P.rockii,andP.decomposita)remainedunresolved(Hong&Pan,1999b;Linetal.,2004;Zhaoetal.,2004).UsingDNAmolecularmarkerandgenesequencesfrombothnuclearandcytoplasmicgenomeshashadanenormousimpactonstudiesofplantphylogeneticsandsystematics(Sang,2002).However,theproblemofthereconstructionofphylogenyoftaxaatlowertaxonomiclevel,especiallythecloselyrelatedspecies,remainsunresolved(Sang2002;Grobetal.,2004).Forexamples,nuclearribosomalDNAregions,especiallytheinternaltranscribedspacers(ITS1andITS2)areappliedextensivelyforphylogenyreconstructionatlowertaxonomiclevels.ITSsequences,however,havetheirownproblemswhenappliedtophylogeneticstudies,suchasextensivelengthvariationsbetweencopies,paralogyproblems,and/orlackofresolvingpower(Grobetal.,2004).Themitochondrialnad1intron2hasbeensuccessfullyusedinstudyingpopulationstructureandphylogeneticrelationshipsamongcloselyrelatedtaxa(Gugerlietal.,2001).Butfortheirlowsubstitutionratesandhighlevelsofgenerearrangements,plantmitochondrialgeneshavebeenregardedaslessusefulforsystematicstudiesthananimalmtDNA(Sanjuretal.,2002).ThechloroplastgenomesharesmanyfeatureswithanimalmtDNA,suchastheconservedgeneorderandtheirhighlevelsofsequencevariationinthenoncodingpartsofthegenome(Provanetal.,2001).Therefore,thecpDNAsequenceswerewidelyusedinsystematicstudies.Single-orlow-copynucleargeneshavehighratesofsubstitution.Thephylogeneticutilityofthiskindofgeneshasbeeninvestigatedinvariousplanttaxa,suchasgranule-boundstarchsynthase(GBSSIorwaxy)inPoaceae(Mason-Gameretal.,1998),vicilininSterculiaceae(Whitlock&Baum,1999),malatesynthaseinArecaceae(Lewis&Doyle,2001),andalcoholdehydrogenase(Adh)inOryza(Geetal.,1999;Guo&Ge,2005).Especially,thesingle-copynucleargeneglyceraldehyde3-phosphatedehydrogenase(G3pdh)wassuccessfullyusedinthephylogeographicstudyofCassvaanditscloserelatives(Olsen&Schaal,1999).Furthermore,justasZhang&Hewitt(2003)haspointedoutthatbecauseofitsgeneralhigherpolymorphismandtheeaseofdesigningprimerintheflankingcodingsequences(normallyconservedamongcloselyrelatedtaxa),intronwillbeprobablythemostusefulnuclearmarkersinthenearfuture.Thepotentialutilityoftheintronoflow-copynucleargenesinphylogeneticreconstructionatlowertaxonomiclevelshasbeenexplainedcompletely(Grobetal.,2004).Adhisametabolicenzymeresponsiblefortheinterconversionofethanolandacetaldehyde(Freeling&Bennett,1985).Adhgenefamilyusuallyhastwoorthreelociinabroadarrayofangiospermspecies(Cleggetal.,1997).TheDNAdivergenceandmolecularevolutionofthisgenefamilyhavebeeninvestigatedinabroadarrayofplants,suchasBrassicaceae(Kochetal.,2000),Gossypium(Smalletal.,1998),Paeonia(Sangetal.,1997b),Pinusbanksiana(Perry&Furnier,1996),andPoaceae(grass)(Gaut,1999;Geetal.,1999;Guo&Ge,2005).InthegenusPaeonia,Adhgenefamilyhasthreemembers,Adh1A,Adh1BandAdh2,amongthemAdh1Bwasmerelyfoundinsect.Moutan(Sangetal.,1997b).Sangetal.(1997b)reconstructedthephylogenyof11putativenonhybridspeciesandinvestigatedtheAdhgeneevolutioninthegenusPaeonia.TheirstudyshowedthatAdhgeneisbetterthanITSandmatKgenesinresolvingtheinterspecificrelationsamongthespeciesinthegenusPaeonia.However,inthisstudythespeciesofsectionMoutanwerepoorlysampled.SimilartoAdhgene,Glycerol-3-phosphateacyltransferase(GPAT)isanessentialenzymeandhasbeenusedforphylogeneticanalysis,whichisutilizedinthecatalysisoftheinitialstepofglycerolipidsynthesisinthecellsofallhigherorganisms(Nishidaetal.,1993).Inplantcell,therearethreetypesofGPATsthatdifferintheirsubcellularlocation(chloroplasts,mitochondriaandcytoplasm)andsubstratespecificity(Tank&Sang,2001).AmongthegenescodingfordifferentGPATs,thenuclear-encodedchloroplast-expressedGPATgenehasbeenfoundtobesingle-copyinseveraldistantlyrelatedangiospermfamilies(Tank&Sang,2001).InthegenusPaeonia,Tank&Sang(2001)investigatedtheDNAvariationinnuclearencodedchloroplast-expressedGPATgene,andsuccessfullyreconstructedthephylogenyof13Paeoniaspecies.TheirstudyrevealedthatthePaeoniaGPATgenecontainsalargeintronofmorethan2kbwithahighlevelofDNAvariationrate(Tank&Sang,2001).Althoughmanygeneshavebeenusedeffectivelytoaddressthesystematicproblems,theinformationfromsinglegeneusuallywasinsufficienttoresolvetheproblemofinterspecificrelationshipofsect.Moutan(Sangetal.,1997a,b;Sang&Zhang,1999;Linetal.,2004;Zhaoetal.,2004).Incontrast,combinedanalysisofmultipledatasetshasbecomeaneffectivewaytoincreasetheresolvingpowerandthereliabilityofphylogeneticreconstruction(Hillis&Huelsenbeck,1995;Wendel&Doyle,1998;Geetal.,1999;Cronnetal.,2002;Guo&Ge,2005;Jamesetal.,2006).Inthepresentstudy,wereconstructthephylogenyofthesect.Moutanusingfivegenefragmentsfrombothnuclear(Adh1A,Adh2,andGPAT)andchloroplast(trnS-trnGandrps16-trnQ)genomes,aswellasmorphologicalcharacters.OurobjectiveistoclarifyunresolvedproblemsconcerninginterpecificrelationshipwithinPaeoniasect.Moutan.1藝術(shù)與文化1.1u2004規(guī)范depochesinamesiza反應(yīng)Inthisstudy,twelveaccessionsrepresenteightwildspeciesofPaeoniasect.Moutan.Paeonialactiflorawassampledasoutgroup,whichbelongstoPaeoniasect.Paeonia.AllvouchersaredepositedattheHerbarium,InstituteofBotany,ChineseAcademyofSciences,Beijing,China(PE).Thedetailsofthesampledspecies,accessionnumbersorvouchersandoriginsarelistedinTable1.SomesequencesofGPATgene(AY016249),Adhgene(AF009042,AF009043,AF009044,AF009049,AF009058,AF009060,AF009061andAF009068),andrps16-trnQgene(DQ313804)weredownloadedfromGenBank.AllthesequencesproducedinthispaperaredepositedinGenBank(GenBankAccessionNumber:EF520815–EF520868).1.2我1.2.14細胞系統(tǒng)的構(gòu)建ThedatasetofmorphologywastakenfromthematrixpresentedbyZhouetal.(2003).Inthepresentstudy,weselected12samples,includingyinpingmudan-HNSX,yinpingmudan-AHCH,jishanensis-ShXYA,jishanesis-SXJS,decomposita-MEK,rotundiloba-MX,ostii-HNNX,qiui-HBBK,taibaishanica-ShXMX,rockii-HNNX,ludlowii-ML1,anddelavayi-XZLZ.1.2.2pcrcyclingparag/rps1-trng/trnq1.4和..3TotalDNAwasisolatedfromsilica-geldriedleavesusingtheCTABmethodasdescribedbyDoyleandDoyle(1987).AmplificationswereperformedinaPeltierThermalCycler(PTC-200,PE).ThenuclearAdhandGPATgenewereamplifiedandsequencedasdescribedbyLinetal.(2004)andZhaoetal.(2004),respectively.PCRcyclingparametersforthetwochloroplastgenes(trnS-trnGandrps16-trnQ)weresimilar,withaninitial4minat70℃,followedby2cyclesof1minat94℃,20sat52℃,and1.5minat72℃;afterthat,35cyclesof20sat94℃,20sat52℃,and1.5minat72℃wereconducted,withafinalextensiontimeof10minat72℃.AlltheamplifyingandinternalsequencingprimersarelistedinTable2.PCRproductswereelectrophoresedonandexcisedfrom1.5%agarosegel,thenpurifiedusingDNAPurificationKit(Pharmacia).SequencingwasdoneonanABI377(AppliedBiosystems,FosterCity,CA)orMegaBASE1000automaticDNAsequencer(AmershamBiosciences,Buckinghamshire,UK).1.2.3關(guān)于aain國際排他道案例,作品...3與randolrachigos...............................3.3.4與..........................MorphologicalcharactersweretreatedfollowingZhouetal(2003).Maximumparsimony(MP)analyseswereconductedusingPAUP*version4.0b10(Swofford,2002).Allcharacterswereequallyweighted,gapsweretreatedasmissing,andcharacterstatesweretreatedasunordered.HeuristicsearchwasperformedwithMULPARSoption,treebisection-reconnection(TBR)branchswapping,andrandomstepwiseadditionwith1000replicates.Topologicalrobustnesswasassessedbybootstrapanalysiswith1000replicatesusingsimpletaxonaddition(Felsenstein,1985).AnappropriatenucleotidesubstitutionmodelwasdeterminedusingModeltestversion3.06(Posada&Crandall,1998)foreachdataset.ThemodelswerechosenaccordingtotheHierarchicalLikelihoodRatioTest(LRT)andthenusedforsubsequentBayesiananalysis.Bayesianinference(BI)wasconductedusingMrBayesversion3.1.2(Huelsenbeck&Ronquist,2001).OnecoldandthreeincrementallyheatedMarkovChainMonteCarlo(MCMC)chainswererunfor1milliongenerations,withtreessampledevery100generation,usingrandomtreeasitsstartingpointandatemperatureparametervalueof0.2(thedefaultsettingofMrBayes).Foreachdataset,MCMCrunswererepeatedtwiceasasafeguardagainstspuriousresults.Thefirst1000treeswerediscardedasburn-in,andtheremainingtreeswereusedtoconstructBayesiantrees.Examinationofthelog-likelihoodsandtheobservedconsistencybetweenrunssuggestedthattheburn-inperiodsweresufficientlylongenoughforchainstohavebecomestationary.2產(chǎn)品系統(tǒng)2.1兩和3.4.3單最..............................................................3.3.3.Inthepresentstudy,theamplifiedsequencesofAdh1andAdh2includefourintronsandfiveexons.TheresultingfragmentofAdh1Arangedinlengthfrom1141to1167bpwithanalignedlengthof1191bp,amongwhich93sitesarevariableand23sitesareparsimonyinformative.ThesequencesofAdh2varyfrom1095to1149bpinlength.ThefinalalignmentofAdh2is1185bp,with135sitesvariableand31sitesinformative.TheamplifiedsegmentofGPATgeneisabigintronbetweenexon5and6.Thesequencelengthvariesfrom1848to1916bp,andthefinalalignmentofGPATis1962bp,including144variablesitesand54informativesites.Amongthethreenucleargenesegments,theinformativesiteofGPATisthehighest(2.75%),Adh2is2.62%,andAdh1Aisthelowest(1.93%).Chloroplastgenesrps16-trnQandtrnS-trnGhavelessvariableandinformativesitesthanthenucleargenes,withtheinformativesites1.06%and0.97%,respectively.ThecharacteristicsofeachsampledsequencearepresentedinTable3.EachdatasetwasidentifiedtofitacorrespondingmodelusingModeltestver.3.06(Posada&Crandall,1998).NucleargenesAdh1A,Adh2,andGPAThavethesameevolutionarymodelHKY+G,whilethemodelsofchloroplastgenesrps16-trnQandtrnS-trnGareF81.2.2關(guān)于兩共建性條款的相關(guān)數(shù)據(jù)和數(shù)據(jù)interityDuetolimitedinformationandlowresolutionofsinglegenedatasetormorphologicaldataset(datanotshown),inthepresentstudy,weonlyanalyzeddataincombineddatasets.Basedontheinformationofchloroplastgenes,thephylogeneticanalysisofthecombineddataset(rps16-trnQandtrnS-trnG,P=1.000forpartitionhomogeneitytests)suggeststhatthesespeciescanbedividedintotwoclades:onecladeincludesP.ludlowiiandP.delavayi;thesecondcladeconsistsofothersixspecies(datanotshown).TheBayesiananalysisgeneratesasimilartopology,withonlyafewdifferencesinbootstrappingsupport(BS)andBayesianposteriorprobabilityforsomeclades(datanotshown).Threenucleargenes(hereafterrepresentedasdata1,P<0.001),threenucleargenesplustwochloroplastgenes(hereafterrepresentedasdata2,P<0.001),andallDNAsequencedatawithmorphologicaldata(hereafterrepresentedasdata3,P<0.001),wereanalyzedonebyone(Table4).Thedata1generatedtwomostparsimonioustrees(Fig.1),thedata2generatedone(Fig.2),andonemostparsimonioustree(Fig.3)wasobtainedfromthedata3.Thetopologiesofthetreesgeneratedfromthedifferentcombineddatasetsaresignificantlycongruentwitheachother.Paeoniasect.Moutanisdividedintotwomajorclades,consistentwiththeaboveresultsbasedoncombinedchloroplastsequencedata.ThefirstcladeconsistsofP.ludlowiiandP.delavayi;thesecondcladeincludesothersixspecies.Withinthesecondclade,P.rockiiiscloselyrelatedtoP.decomposita;whileotherfourspeciesformanothersubclade,andP.jishanensis/P.qiuiandP.suffruticosa/P.ostiiwassistertoeachother.Althoughthetopologiesarecongruentbetweenthethreedifferentcombineddatasets,therearesomedifferencesamongthem:(1)inP.jishanensis/P.qiuiclade,twosamplesofP.jishanensisgroupedinthedata2anddata3,butnotindata1;(2)withinP.rockii/P.decompositaclade,twosubspeciesofP.rockiigroupedinthedata1andthedata3,butnotinthedata2.Inaddition,itisalsoapparentlythattheBSvalueisincreasingwiththeadditionofseparatedatasettocombineddataset.Foreachnodeofthedata1,theaverageBSis82%(Fig.1).TheaverageBSforeachnodeofphylogenyderivedfromthedata2was85%(Fig.2),whilethatofdata3getanaveragesupportof91.4%(Fig.3).ThetrendofBayesianposteriorprobabilityisthesameasthatofBS.3子階段3.1phyoleny相關(guān)specity.非織造材料,非織造材料,非織造材料,非織造材料,非織造材料,非價值specityspecityspecityspecity.specity.specity.specity.specityspecity.specity.specity.specity.specityspecity.specity.specity.specity.specity.specityspecity.specity.specity.specity.specitysq種植.Inpreviousstudies,duetotheincompletesamplingorlimitedinformativecharacters(Zouetal.,1999;Yuetal.,1998;Tank&Sang,2001),thephylogeneticrelationshipsofPaeoniasect.Moutanisambiguous.Recently,bothmorphological(Zhouetal.,2003)andmolecular(Linetal.,2004;Zhaoetal.,2004)evidencewasusedtoinvestigatethephylogeneticrelationshipofPaeoniasect.MoutanbasedonthenewclassificationsystemproposedbyHongandPan(1999a).Thesestudiesmadesomeimpressiveadvances,butinterspecificrelationshipofPaeoniasect.Moutanisstillnotresolvedbecauseofthelimitedphylogeneticinformation(Sangetal.,1997a,b;Zhouetal.,2003;Linetal.,2004;Zhaoetal.,2004).Forthefirsttime,inthisstudy,weusebothmultiplegenesandmorphologicalcharacterstoreconstructthephylogenyofPaeoniasect.Moutan.ThereareeightwildspeciesinPaeoniasect.Moutan,distributedinawiderangefromSouthwesttoNorthChina(Pan,1995;Hong&Pan,1999a).Withintheeightspecies,P.ludlowiiandP.delavayiaredistributedinthewestofMt.Daxueshan,whileothersixspeciesgrowintheeastofMt.Daxueshan(Hong&Pan,1999a).Basedonthemorphologicalcharacters,P.ludlowiiissimilartoP.delavayi,theirflowersareterminalandaxillary,discfleshyandshort.Incontrast,theflowersofothersixspeciesaresolitary,discleathery(Hong,1997b;Hong&Pan,1999a).Recently,phylogeneticanalysessuggestedthatPaeoniasect.Moutaniscomposedoftwoclades,onecladeincludesP.ludlowiiandP.delavayi,anothercladeincludesothersixspecies(Zhouetal.,2003;Linetal.,2004;Zhaoetal.,2004).TheseresultsarecorrespondenttothetreatmentoftwosubsectionsbyStern(1946).Inthepresentstudy,ourresults(Figs.1–3)supportthetwo-subsectiontreatment,namely,subsect.Delavayanaeandsubsect.Vaginatae.Thephylogeneticrelationshipofsubsect.Vaginataehasbeenindisputeinthepastseveraldecades.Afteranalyzingthemorphologicalcharacters,HongandPan(1999b)treatedfivespeciesasthecomplexP.suffruticosa,includingP.rockii,P.jishanensis,P.qiui,P.suffruticosa,andP.ostii,duetotheentiretydiscleathery,whereasP.decompositaisdistinguishedfromthembyitspartiallydiscleathery.Zhouetal.(2003)usedtwenty-fivemorphologicalcharacterstostudythephylogeneticrelationshipsofeightwildspeciesofPaeoniasect.Moutan,andtheresultindicatedthatP.decompositagroupedwiththecomplexP.suffruticosaandlocatedatthebasalpositionofthisclade,thussupportedthetreatmentofthecomplexP.suffruticosa.ButintherecentDNAsequenceanalyses,Linetal.(2004)andZhaoetal.(2004)foundthatP.decompositawasgroupedwithP.rockii,andotherfourspeciesformedacladewithhigherBS.Otherpreviousstudiesbasedonmolecularmarkersalsogotthesimilarresults(Sangetal.,1997a,b;Sang&Zhang,1999;Zouetal.,1999;Tank&Sang,2001).Inthepresentstudy,thecombinedanalysisofmolecularandmorphologicaldatasupportsthatP.decompositaisgroupedwithP.rockii,andotherfourspeciesformedanotherclade(Fig.3).Therefore,wecanconcludethatP.decompositaiscloselyrelatedtoP.rockiicomparingtootherfourspeciesinsubsect.Vaginatae.PaeoniadecompositaandP.rockiihavetwosubspecies,respectively.Inpreviousstudies,therelationshipsamongthefoursubspeciesconflictedwitheachother(Hongetal.,1996;Zhouetal.,2003;Linetal.,2004;Zhaoetal.,2004).OurresultssuggestthatP.decompositassp.rotundilobawassistertoP.decompositassp.decompositaandtwosubspeciesofP.rockii(Figs.1–3).Therefore,wecanconcludethatP.decompositassp.decompositaiscloselyrelatedtoP.rockii.Comparingthemorphologicalcharacters,P.decompositassp.rotundilobahas2–5carpels,usually3or4,whileP.decompositassp.decompositahasastablenumber(5)ofcarpel(Hong&Pan,1999a).Therefore,bothmorphologicalandmolecularevidencesupportsthatP.decompositassp.decompositaiscloselyrelatedtoP.rockii.Insummary,Paeoniasect.Moutancanbedividedintotwosubsections,thesubsect.Delavayanaeandsubsect.Vaginatae.SubsectionDelavayanaeconsistsoftwospecies,P.delavayiandP.ludlowii;subsect.Vaginataeincludesothersixspecies.WithinthesubsectionVaginatae,P.rockiiiscloselyrelatedtoP.decomposita;whileotherfourspeciesformaclade,andP.jishanensis/P.qiuiandP.suffruticosa/P.ostiiweresistertoeachother.3.2trend梁ue,kha根據(jù)其文Theprimaryaimofphylogeneticanalysisistoinfertherelationshipamongspecies,namelyspeciestree.Asweknow,genetreederivedfromsinglegeneisdifficulttogiveacorrectspeciestreeforthelimitedinformationatlowertaxonomicgroup(Doyle&Gaut,2000).Inaddition,differentdatasetsusuallygeneratedifferentphylogeneticrelationships(Doyle&Gaut,2000),suchasamongdifferentmoleculardatasetsorbetweenmoleculardatasetandmorphologicalcharacterdata.Therefore,usingthecombinedgenesdatasettoinferphylogeneticrelationshiphasbecomeawidelyacceptedtrend(Hillis&Huelsenbeck,1995;Wendel&Doyle,1998;Geetal.,1999;Cronnetal.,2002;Guo&Ge,2005;Jamesetal.,2006).Uptonowtherearethreecommonlyusedmethodstodealwiththemultipledatasets.Firstly,alwayscombinethedata(Kluge,1989).Inparticular,alloftheavailabletaxa(e.g.fossilandliving)shouldbecombinedinaphylogeneticanalysis,aswellasalloftheavailablecharacters.Kluge(1989)believesthetotalevidencesolutionissoughtbecauseitmaximizesthe“informativeness”and“explanatorypower”ofthecharacterdatausedintheanalysis.Secondly,analyzedifferentdatasetseparatly(Miyamoto&Fitch,1995).Thatmeansanalyzingthedifferentdataindependently.Thirdly,conditionaldatacombination.Namely,partitionsweresubjectedtoastatisticaltestof“homogeneity”:Heterogeneousdatapartitionsarethosethatresultinsignificantlydifferentestimatesaboutphylogeny(differencesbeyondthatexpectedfromsamplingerror)whenanalyzedseparately(Farrisetal.,1995).Ifthetestresultisnon-sign