Genomic Changes in Resynthesized Brassica napus and Their

GenomicChangesinResynthesizedBrassicanapusandTheirEffectonGeneExpressionandPhenotype

WOARobertT.Gaeta,aJ.ChrisPires,a,b,1FedericoIniguez-Luy,aEnriqueLeon,aandThomasC.Osborna,2aDepartmentbDepartment

ofAgronomy,UniversityofWisconsin,Madison,Wisconsin

ofBiologicalSciences,UniversityofMissouri,Columbia,Missouri

Manypreviousstudieshaveprovidedevidenceforgenomechangesinpolyploids,buttherearelittledataontheoverallpopulationdynamicsofgenomechangeandwhetheritcausesphenotypicvariability.Weanalyzedgenetic,epigenetic,geneexpression,andphenotypicchangesin;50resynthesizedBrassicanapuslinesindependentlyderivedbyhybridizingdoublehaploidsofBrassicaoleraceaandBrassicarapa.Apreviousanalysisofthefirstgeneration(S0)foundthatgeneticchangeswererare,andcytosinemethylationchangeswerefrequent.OuranalysisofalatergenerationfoundthatmostS0methylationchangesremainedfixedintheirS5progeny,althoughthereweresomereversionsandnewmethylationchanges.GeneticchangesweremuchmorefrequentintheS5generation,occurringineverylinewithlinesnormallydistributedfornumberofchanges.Geneticchangesweredetectedon36ofthe38chromosomesoftheS5allopolyploidsandwerenotrandomacrossthegenome.DNAfragmentlosseswithinlinesoftenoccurredatlinkedmarkerloci,andmostfragmentlossesco-occurredwithintensificationofsignalfromhomoeologousmarkers,indicatingthatthechangeswereduetohomoeologousnonreciprocaltranspositions(HNRTs).HNRTsbetweenchromosomesA1andC1initiatedinearlygenerations,occurredinsuccessivegenerations,andsegregated,consistentwitharecombinationmechanism.HNRTsanddeletionswerecorrelatedwithqualitativechangesintheexpressionofspecifichomoeologousgenesandanonymouscDNAamplifiedfragmentlengthpolymorphismsandwithphenotypicvariationamongS5polyploids.OurdataindicatethatexchangesamonghomoeologouschromosomesareamajormechanismcreatingnovelallelecombinationsandphenotypicvariationinnewlyformedB.napuspolyploids.

INTRODUCTION

Polyploidyisthoughttoplayasignificantroleinfloweringplantevolution,and30to80%ofextantspeciesareconsideredpolyploid(Masterson,1994;RamseyandSchemske,1998;OttoandWhitton,2000;MeyersandLevin,2006;RiesebergandWillis2007).Recentevidencesuggeststhatmostmajorlineagesofangiospermshaveundergonegenomedoublingoverevolution-arytime,suggestingthatmostoralldiploidsmaybeancientpolyploids(Cuietal.,2006).Polyploidscanoftenbedistin-guishedfromtheirpresumedprogenitorswithrespecttomor-phology,ecology,cytology,andphysiology(Levin,1983;RamseyandSchemske,2002).Variationintheseandotherpheno-typictraitsmaycontributetonicheexploitationandspeciation(RiesebergandWillis,2007)andmightalsoenhancetheutilityofpolyploidsforagriculture.Thisvariationisdifficulttostudyinnaturalpolyploidsthatoftenhaveextinctorunknownparents,whilenewlyresynthesizedallopolyploidswithknownparentspermitmorecriticalcomparisonsofpolyploidsandprogenitors.Thereisgrowingevidencefromsuchcomparisonsthatpoly-1Address2Current

correspondencetopiresjc@missouri.edu.

address:SeminisVegetableSeeds,36437StateHighway16,

Woodland,CA,95695.

TheauthorresponsiblefordistributionofmaterialsintegraltothefindingspresentedinthisarticleinaccordancewiththepolicydescribedintheInstructionsforAuthors(www.plantcell.org)is:J.ChrisPires(piresjc@missouri.edu).WOnlineversioncontainsWeb-onlydata.OAOpenAccessarticlescanbeviewedonlinewithoutasubscription.www.plantcell.org/cgi/doi/10.1105/tpc.107.054346

ploidshavenovelphenotypicvariationandthatpolyploidycanleadtochangesingeneexpressionthroughgenedosageef-fects,alteredregulatoryinteractions,andgeneticandepigeneticchanges(reviewedinSoltisandSoltis,1995;Guoetal.,1996;RamseyandSchemske,2002;Osbornetal.,2003b;Soltisetal.,2004;AdamsandWendel,2005;ChenandNi,2006).

Evidenceforgenomicchangesinnascentpolyploidtaxacomesfromobservationsofresynthesizedpolyploids(e.g.,Arabidopsissuecica,Brassicanapus,wheat[Triticumaestivum],cotton[Gossypiumhirsutum],Nicotianatabacum,andTriticale)andnaturalpolyploidspeciesthathavewell-documentedpar-entage(e.g.,Tragopogon,Senecio,Spartina,andGlycine).Amongtheseexamples,therehavebeenreportsofdeletionevents(Ozkanetal.,2001;Shakedetal.,2001;MaandGustafson,2006;Tateetal.,2006),geneconversionevents(Wendeletal.,1995;Kovariketal.,2004,2005),rDNAlocichanges(Jolyetal.,2004;Pontesetal.,2004),transposonactivation(Kashkushetal.,2002,2003;Madlungetal.,2005),chromosomalrearrangements(Kentonetal.,1993;Parkinetal.,1995;Limetal.,2004,2006;Piresetal.,2004;Udalletal.,2005),andepigeneticphenomena(ChenandPikaard,1997a,1997b;Adamsetal.,2003;Hegartyetal.,2005;Madlungetal.,2005;Salmonetal.,2005;Lukensetal.,2006;Wangetal.,2004).However,littleisknownregardingthemechanismsleadingtothesechanges,theirpopulationdynamics,andtheirimpactongeneexpressionandphenotypeinnewpolyploids.

Geneticchanges,includingchromosomalrearrangements,transpositions,anddeletions,canleadtoalteredgeneexpres-sioninallopolyploids(Osbornetal.,2003b;ChenandNi,2006).

3404ThePlantCell

Piresetal.(2004)detectedhomoeologousrearrangementsinresynthesizedB.napusthatalteredtheexpressionofparentalFLCgenes.InTragopogon,thelossofDNAfragmentswasshowntobethemajorcauseofcDNAamplifiedfragmentlengthpolymorphisms(AFLPs)(Tateetal.,2006).Epigeneticchanges,includingDNAmethylation,histonemodifications,andRNAinterference,alsomayleadtochangesingeneexpressioninnewlyformedpolyploids(ChenandPikaard,1997a;Comai,2000;LiuandWendel,2003;Osbornetal.,2003b;MadlungandComai,2004;AdamsandWendel,2005).StudiesinwheatandArabidopsisallopolyploidshavefoundthatchangesincytosinemethylationcorrelatewithalteredgeneexpression(LeeandChen,2001;Kashkushetal.,2002)andreactivationofsilenttransposons(Madlungetal.,2005).Geneticandepigeneticchangesarepredictedtoaffectphenotypicvariation,andphe-notypicstudiesinnewlyformedB.napusallopolyploidsprevi-ouslydemonstratedtheemergenceofheritabledenovovariationinfloweringtimeandotherlifehistorytraits(SchranzandOsborn,2000;Piresetal.,2004;SchranzandOsborn,2004).Studieshavesuggestedthathomoeologousrearrangementsmightberesponsibleforalteringvariationinseedyield(Osbornetal.,2003a),floweringtime(Piresetal.,2004),anddiseaseresistance(Zhaoetal.,2006).However,mostpreviousstudiesonpolyploidyhavelackedenoughindependentpolyploideventstounderstandthepopulationdynamicsofsuchchangesandhavebeenunabletodeciphermechanismsofgeneticchangeduetoalackofmapped,genome-wide,dosage-sensitivemolecularmarkers.Therehasalsobeenapaucityofintegratedgenotypic,geneexpression,andphenotypicdata.

Inthisstudy,weinvestigatedgenetic,epigenetic,transcrip-tional,andphenotypicchangesamong;50independentlyresynthesizedB.napuslinesderivedbyhybridizingdoublehaploidlinesofBrassicaoleraceaandBrassicarapa.Thisrela-tivelylargepopulationallowedforacomprehensiveanalysisofpopulation-wideandgenome-widechangesfollowingpolyploid-

ization.Previously,weanalyzedthispopulationattheS0gener-ationandfoundthatgeneticchangeswererareandcytosinemethylationchangeswerefrequent(Lukensetal.,2006).Here,weanalyzedthesamepopulationofresynthesizedpolyploidsattheS0andS5generationsusingsimplesequencerepeat(SSR),restrictionfragmentlengthpolymorphism(RFLP),cDNAsingle-strandconformationpolymorphism(SSCP),andcDNA-AFLPmarkers,andwemeasured13phenotypictraitsinareplicatedexperiment.WepresentevidenceforamechanisticlinkbetweengeneticchangesinnewlyresynthesizedB.napusandchangesingeneexpressionandphenotypicvariation.RESULTS

GeneticChangesinS5ResynthesizedB.napusAllopolyploids

Analysisof368HindIII,EcoRI,orDraIRFLPmarkersand65SSRmarkersin47S5linesofresynthesizedB.napusrevealedfrag-mentlossesineveryline,with33%oftheRFLPmarkersand71%oftheSSRmarkersshowingchangesamongtheS5lines.Approximately48%ofA-genome–specific(B.rapa)and34%ofC-genome–specific(B.oleracea)RFLPandSSRsshowedchanges.The47S5lineswerehighlyvariable(Figure1)andnormallydistributed(Shapiro-Wilktest;P¼0.3824)fortotalnumberoffragmentchanges.AllbuttwolineslostbothA-andC-genomemarkers.TherewasamarginalsignificantdifferencebetweentheaveragenumberofA-genomelossesperS5line(mean6.6,SD¼4.3)andtheaveragenumberofC-genomelossesperS5line(mean8.6,SD¼5.6;ttest,P¼0.048),andtheproportionofA-andC-genomefragmentlossesamongS5lineswasunequal(4.1and4.7%,respectively;x2¼3.92,P¼0.047;Table1).Therewerenodifferencesbetweenthemeannumberofchangespercolchicine-doubledS5line(mean15.6,SD¼7.1)andthemeannumberofchangesperspontaneouslydoubledS5line

Figure1.NumberofDNAFragmentLossesinEachS5LineofResynthesizedB.napus.

Thesumofallfragmentlosses(RFLPandSSR)isshownforeachof42colchicine-doubledand10spontaneouslydoubledlineages(asterisks).LineagesconnectedbyarrowswerederivedfromthesameS0plants,onebycolchicinedoubling(lines17and19)andonebyspontaneousdoubling(lines28and39).Linenumbers28and39wereconsiderednonindependentmeasuresandwerenotincludedinthestatisticalanalyses.ThreelinesdiedoutattheS1,S3,andS5generationsandarelabeledaccordingly.TheselineswerenotincludedinstatisticalanalysesoftheS5generation.TheportionofmarkerslostfromtheA(B.rapa)andC(B.oleracea)genomesofB.napusareshownwithgrayandblackbars,respectively.

GenomicChangesinBrassicanapus3405

Table1.SummaryofGeneticChangesamong47ResynthesizedS5B.napusAllopolyploids

MarkerType

Total

Totalno.DNAfragmentsexpectedbTotalno.DNAfragmentslost(%)No.Afragmentsexpectedb,dNo.Afragmentslost(%)dNo.Cfragmentsexpectedb,dNo.Cfragmentslost(%)dNo.ACfragmentsexpectedb,dNo.ACfragmentslostdaMethodbExpected

GenomeDoubling

RFLP17,140509(3.0)c6,336216(3.4)7,151293(4.1)3,6530

Colch.a16,733610(3.6)6,301254(4.0)7,129356(5.0)3,3030

Spont.a3,434103(3.0)1,28854(4.2)1,46849(3.3)6780

SSR3,027204(6.7)1,25392(7.3)1,446112(7.7)3280

20,167713(3.5)7,589308(4.1)8,597405(4.7)3,9810

ofallohaploidgenomedoubling.Colch.,colchicine;Spont.,spontaneously.

numberoffragmentsisequaltotheproductoftotalmarkersscoredandthenumberoflines,minusmissingdata,andisthenumber

expectedunderparentaladditivity.

cIn71%ofthe509casesinwhichanRFLPfragmentwaslost,itwasassociatedwithintensificationofafragmentfromtheotherparentalsubgenome,suggestingthataloss/duplicationhadoccurred.ThesedatawerenotcollectedforSSRmarkersbecausePCRproductsarenoteasilyquantifiable.dDesignationsA(B.rapa),C(B.oleracea),andAC(commontoboth)indicatethegenomeoforiginofthemarkersscored.

(mean12.9,SD¼5.0;ttest,P¼0.30)orintheproportionoffragmentlossesforcolchicineandspontaneouslydoubledlines(3.6and3.0%,respectively;x2¼3.30,P¼0.069;Table1).Althoughcolchicine-doubledlineshadalargervarianceforfragmentlossesthandidspontaneouslydoubledlines,thisdifferencewasnotsignificant(Ftest;P¼0.34).

DetectionofHomoeologousNonreciprocalTranspositionsApproximately71%ofRFLPmarkerfragmentlosses(Table1)occurredconcomitantlywithanintensificationofahomoeolo-gousRFLP,consistentwithlossofonelocusandduplicationoftheother(Figure2).SSRmarkersdetectedthelossofparentalmarkers,buttheycouldnotbeusedtodetecttheduplications.Patternsoflossandintensificationoftenoccurredatlinkedmarkersonhomoeologouslinkagegroups(Figure2),suggestingthatthechangesresultedfromexchangesbetweenhomoeolo-gouschromosomesleadingtohomoeologousnonreciprocaltranspositions(HNRTs)(Piresetal.,2004;Udalletal.,2005).ThelargenumberofmarkersavailableforA1andC1allowedanin-depthanalysisoftheHNRTspresentintheS5generation(A1isB.rapalinkagegroup1andisequivalentB.napuslinkagegroupN1;C1isB.oleracealinkagegroup1andisequivalentB.napuslinkagegroupN11;seeMethodsforexplanationofBrassicalinkagegroupnomenclature).Twenty-nineofthe47S5linesshowedgeneticchangesonthesechromosomes(Figure3).Changesinmarkerhybridizationpatternsoftenaffectedblocksofmultiplelinkedmarkers.Lossofmarkersinablockwastypicallyassociatedwithintensificationofmarkersontheho-moeologousblock(e.g.,bottomofA1/C1).ExceptionsincludeSSRmarkers(explainedabove)andsomeRFLPmarkers,forwhichthehomoeologsmaynothavebeendetectableduetogelmigrationpatternordivergenceorabsenceofthehomoeologoussequenceinoneoftheparents.Mostofthealteredsegmentsincludedmarkersattheendsofchromosomes(e.g.,bottomofC1infirstthreelinesofFigure3),althoughsomeinterstitialchangeswereobserved(e.g.,topofC1inline2andbottomofC1inline9ofFigure3).

DynamicsofGeneticChangesacrosstheB.napusChromosomes

Geneticchanges(fragmentlossesandlossduplications)weredetectedon36of38chromosomesofB.napusdespitethelimitationthatonly133RFLPandSSRs(31%ofallmarkers)couldbeassignedamaplocation(seeMethods).Thefrequencyofgeneticchangesacrosslinkagegroupswasnotequal(Waldx2¼172.7,P<0.0001).ThelinkagegroupwiththehighestfrequencyofchangeswasC1(20.8%),onelinkagegroupcouldnotbeassignedanymarkers(A4),andone(A8)showednochanges(seeSupplementalTable1online).HomoeologouslinkagegroupsA1/C1,A2/C2,andA3/C3areeachsyntenicalongtheirentirelength(Parkinetal.,1995,2005;Udalletal.,2005).NeithertheA2/C2norA3/C3homoeologouspairsshowedsignificantbiasesinthefrequencyofB.rapaorB.oleracea–derivedfragmentlosses.However,atA1/C1,therewasasignif-icantbiastowardC1losses(20.8%)overA1losses(7.9%)(x2¼17.52,P<0.0001).Noevidencefornullisomic-tetrasomicaneu-ploidswasfoundamonganyofthe47independentlines.However,wecannotruleoutthelossofachromosomeafteranexchangeoccurred.

TransgenerationandS1SegregantAnalysesSupportaRecombination-BasedMechanism

TodeterminewhenrearrangementsoccurredbetweentheS0andS5generations,weconductedatransgenerationalanalysisofsixallopolyploidlineagesformarkersonA1/C1.WedetectedgeneticallyfixedHNRTsinbulkedDNAsamplesasearlyastheS2generationandaslateastheS5generation(seeSupplemen-talFigures1and2online).Wethengenotyped14to16individ-ualsinthegenerationpriortoourdetectionofafixedHNRTanddetectedanaverageof3.6(range1to6)individualscarryingahomozygousrearrangement.RegardlessofwhetheranHNRTinalineappearedinterstitialorterminalizedintheS5analysis,wefrequentlyobservedvariablelengthsorformsoftheexpectedrearrangementsegregatinginthepriorgeneration(see

3406ThePlantCell

Figure2.DetectionofGenomicRearrangementsinResynthesizedB.napus.

DNAgelblotsshowinggeneticallylinkedRFLPmarkersdetectedbyprobespW225,pW239,andpX135usingEcoRI(leftpanel)andHindIII(rightpanel)restrictiondigests.GenomicDNAfromthediploidparentallinesandeightS5linesofresynthesizedB.napusareshown(numberedaccordingtoFigure1).DesignationsAandCindicateparentalgenomicfragmentsofB.rapa(lineIMB218)andB.oleracea(lineTO1000)origins,respectively,thatmaponB.rapaA1andB.oleraceaC1,respectively.SomeS5linesshowlossofA-genomeandintensificationofhomoeologousC-genomefragments(lines17and35)orlossofC-genomeandintensificationofhomoeologousA-genomefragments(lines3,5,and37).Loss-intensificationpatternsformarkerspW225andpW239wereperfectlycorrelatedamongallS5lines,whilethemoredistalmarkerpX135wasmoderatelycorrelated(r¼0.59)withpW225andpW239.

SupplementalFigure3online).WedidnotdetectHNRTsfixedfortheotherhomoeologsegregatinginthegenerationpriortoafixedrearrangementthatwouldhavederivedfromanintactreciprocalexchange;however,wedidinsomecasesdetectpartialreciprocalrearrangementssegregatingthatcouldhavederivedfromareciprocalexchangewithadditionalrecombina-tions(seeSupplementalFigure4online).

Inapreviousstudy,evidenceforchromosomalrearrange-ments(HNRTs)wasnotobservedintheS0generation(Lukensetal.,2006).Toseeifrearrangementsaroseafterthefirstmeiosis,weanalyzedfourmarkersthatspannedthreepairsofhomoeologouschromosomes(A1/C1,A2/C2,andA3/C3)in10to16S1progenyfromeachofeightS0allopolyploids.Ofthe118segregantsanalyzedamongtheeightlines,threehadevidenceofarearrangement(seeSupplementalFigure5online).

47independentS5lineswerenormallydistributedfornumberofmethylationchanges(Shapiro-Wilktest;P¼0.1054).Thepro-portionoftotalHpaIIfragmentchangesamongS5lines(lossandreversions;2.1%)wasgreaterthantheproportionofMspIfragmentchanges(0.1%)(x2¼249.8,P<0.0001;datanotshown).Thus,CpGmethylationchangeswereobservedatnearlya20-foldgreaterratethanCpNpGchangesintheS5generation.TheproportionoftotalA-genomefragmentchanges(1.8%)amongS5lineswassignificantlygreaterthanthepro-portionoftotalC-genomefragmentchanges(1.3%)(x2¼9.003,P¼0.003;datanotshown).Theproportionoffragmentchangesincolchicine-doubledlines(1.0%)wasnotdifferentfromtheproportionofchangesamongspontaneouslydoubledlines(1.4%)(x2¼2.765,P¼0.096;datanotshown).

DNAMethylationChangesinS5ResynthesizedB.napusAllopolyploids

Analysisof337HpaIIand320MspIRFLPmarkersrevealedmethylationchangesineveryS5line,with60%oftheHpaIIand54%oftheMspImarkersshowingchanges.ThesamemarkerswereanalyzedintheS0generation(Lukensetal.,2006),andtheresultswerecompared(seeSupplementalFigure6online).The

LossofSpecificDNAMarkersandTranscriptsamongS0andS5PolyploidLines

WeassayedfortranscriptionalchangesinthreegenetargetsforwhichwedetectedHNRTsamongS5polyploidlines.TheRFLPprobepW225detectedHNRTsbetweenlinkagegroupsA1andC1inmanypolyploids(Figures2and3).RearrangementatthislocuswasperfectlycorrelatedwiththelossandintensificationofhomoeologouspW225RNAtranscriptsamongS5polyploids

GenomicChangesinBrassicanapus3407

Figure3.IdeogramsofHomoeologousLinkageGroupsA1andC1forS5LinesofResynthesizedB.napus.

GeneticchangesfortheA1andC1chromosomecomplimentsareshownforeachof42colchicine-doubledand10spontaneouslydoubledlines(asterisksabovelinenumbers)inthesameorderasshowninFigure1.EachverticalbarrepresentstheA1(toppanel,mostlywhitebars)andC1(bottompanel,mostlygraybars)karyotypesofeachpolyploidline.ThepresenceofagraybaronlinkagegroupA1indicatesthatthelossofarapa-specificRFLPmarkerwasassociatedwithanintensificationoftheC1homoeologousmarker(seeFigure2).Conversely,thepresenceofawhitebaronC1indicatesthatthelossofanoleracea-specificRFLPmarkerwasassociatedwithintensificationofanA1homoeologousmarker.Checkeredbarsindicatelosses(allSSRsandsomeRFLPs),andblackbarsindicatemissingdata.MarkersarelistedtotheleftandwereassignedtolinkagegroupsA1andC1basedonmappingdata:*,markerswithidenticalfragmentsmappedinB.oleracea,B.rapa,andB.napuspopulations(ourunpublisheddata;Udalletal.,2005);**,markerassignedtolinkagegroupbasedonmappingstudieswiththesameprobeforwhichasinglelocuswaspredicted(ourunpublisheddata;Udalletal.,2005;D.LydiateandA.Sharpe,personalcommunication);***,markerassignedtolinkagegroupbasedonmappingdatawiththesameprobeforwhichtwoormorelociwerepredicted;thesemarkerswereassignedtothislinkagegroupbasedoncoincidentlossesinothermarkers.Somedifferencesexistbetweenmappingdataandthesemarkerorders(e.g.,pW145EfandpW220onC1areintheoppositeorderofthatsuggestedbyourunpublisheddata).Thesedifferencesmaybeexplainedbytherelativelyshortcentimorgandistancebetweenthesemarkersandbecausewehadsofewlinesandmarkersforanalysis.

(Figure4).EvidenceforchangesinCpGmethylationoccurredintwolinesprobedwithpW225,buttherewasnoevidenceforchangesinpW225transcriptsintheselines(datanotshown).RFLPandPCRanalysesoftheBrassicaFLC-3andFLC-5genomiclocialsoidentifiedHNRTsamongS5polyploids.FLC-3DNAmarkerlosseswerecorrelatedwiththelossofparentaltranscripts(seeSupplementalFigure7online).OnlyB.rapaFLC-5geneexpressionisdetectable,andlossoftranscriptionofthislocusmatchedlossesofB.rapaFLC-5DNA.Forthegenesanalyzed,wedidnotdetectanyquantitativedifferencesintheexpressionofhomoeologs,otherthanthosethatcorrespondedtoquantitativeDNAdifferencesandprobablyrepresentedvaria-tioninthedosageofhomoeologsfromunfixedrearrangements.

DNAMarkerLossesCorrelatewiththeLossofcDNA-AFLPFragmentsamongS5PolyploidLines

Analysisof360cDNA-AFLPmarkersinresynthesizedB.napusrevealedthat4%ofallS0markersshowedchangesamongthelines.Approximately5%oftheA-genomecDNA-AFLPmarkersand6%ofC-genomecDNA-AFLPmarkersshowedchangesintheS0generation.Allchangeswerefragmentlossesconfirmedintwobiologicalreplicates(seeSupplementalFigure8online).Wedidnotobservereproduciblenonparental(novel)cDNA-AFLPmarkerfragments.IntheS0generation,theproportionofA-genomefragmentsthatwerelost(0.2%)wasnotdifferentfromtheproportionofC-genomelosses(0.3%)(x2¼0.31,P¼0.576;

3408ThePlantCell

Figure4.DNAandRT-PCRSSCPAnalysisofResynthesizedB.napusPolyploids.

ComparisonofpW225DNAandtranscriptsfromparentallinesofB.rapa(IMB218;A-genome)andB.oleracea(TO1000;C-genome),a1:1mixtureofparentalcDNAorDNA,andsamplesfrombothS0andS5generationsofsixsyntheticB.napuspolyploidlines(9,21,31,17,35,and49).TwobiologicalreplicatesofRNAforeachRT-PCRsampleandtwotechnicalreplicatesofeachDNAPCRsamplewereresolvedinadjacentlanes(1and2).TocontrolforgenomicDNAcontaminationinRT-PCRreactions,adjacentreactionswereconductedongenomicDNA,whichamplifiedlargergenomicfragmentsnotdetectedinRTreactions(notvisibleinfigure).pW225DNAandgeneexpressionwasadditiveforalllinesintheS0generation.ThelossofhomoeologousDNAandtranscriptswasobservedinmanyoflinesattheS5generation;twolinesshowndemonstrateevidenceforthelossofB.rapaDNAandtranscripts(17and35)andtwodemonstratethelossofB.oleraceaDNAandtranscripts(9and21).SixtotalbandsweredetectedbyDNAandcDNASSCPfromeachparent,indicatingthatthreehighlysimilartargetswereamplified(fragmentsconnectedbyarrowsrepresentcomplementarystrandsofthesameDNAsequence).Foreachlinethathadarearrangementatthislocus,thelossofaparentalcDNAwasassociatedwithanintensificationofthehomoeologouscDNAfragment,matchingtheHNRTpatternsobservedinbothRFLPandDNASSCPmarkers(seeFigures2and3).

Table2).TheaveragenumberofcDNA-AFLPfragmentlossesintheS0was0.63perline.Twenty-nineof31cDNA-AFLPfrag-mentsthathaddisappearedintheS0generationappearedintheS5.Therewasnodifferencebetweentheproportionoffragmentslostincolchicine(0.2%)andspontaneously(0.1%)doubledlines(x2¼0.10,P¼0.747;Table2).

Analysisof360cDNA-AFLPmarkersintheS5generationrevealedthat42%showedchangesamongthelines.Approxi-mately60%oftheA-genomemarkersand54%oftheC-genomemarkersshowedchangesintheS5generation.Onlytwomarkerscommontobothparentaltranscriptomes(2.1%oftotal)showedchangesintheS5generation.IntheS5generation,thepropor-tionofA-genomefragmentlosses(4.0%)wasnotsignificantlydifferentfromtheproportionofC-genomefragmentlosses(4.3%)(x2¼0.72,P¼0.395;Table2).ThetotalnumberofcDNA-AFLPfragmentlosseswasnormallydistributedamongS5lines(Shapiro-Wilktest;P¼0.0889).TheaveragenumberofcDNA-AFLPfragmentlossesintheS5generationwas10.6perline.Therewasnodifferenceinthefrequencyoffragmentlossesbetweencolchicine(3.0%)andspontaneouslydoubledlines(3.0%)(x2¼0.002,P¼0.964).TheproportionoftotalfragmentslostintheS0generation(0.2%)wassignificantlydifferentfromtheproportionoflossesintheS5generation(3.0%)(x2¼432.91,P<0.0001;Table2).

NorelationshipbetweengeneticorepigeneticchangesandcDNA-AFLPfragmentslosseswasdetectedintheS0genera-tion;however,intheS5generation,DNAfragmentlossesweresignificantlycorrelatedwithcDNA-AFLPfragmentlosses(r2¼0.55,P<0.0001;Table3;seeSupplementalFigure9online).Methylationchangeswerenotcorrelated,suggestingthatge-nomicrearrangementsinS5lineswerelargelyresponsibleforlossofcDNA-AFLPmarkers.

GenomicChangesinBrassicanapus3409

Table2.SummaryofcDNA-AFLPsbetween48S0and47S5ResynthesizedB.napusPolyploidLines

S0GenerationaTotal

Totalno.fragmentsexpectedbTotalno.fragmentslost(%)No.Afragmentsexpectedb,cNo.Afragmentslost(%)cNo.Cfragmentsexpectedb,cNo.Cfragmentslost(%)cNo.ACfragmentsexpectedb,cNo.ACfragmentslost(%)caThe

S5GenerationaColch.14,18227(0.2)5,35210(0.2)4,99017(0.3)3,8400

Spont.2,8384(0.1)1,0764(0.4)99407680

Total16,680504(3.0)6,298250(4.0)5,870252(4.3)4,5122(0.04)

Colch.13,838419(3.0)5,224197(3.8)4,870220(4.5)3,7442(0.05)

Spont.2,84285(3.0)1,07453(4.9)1,00032(3.2)7680

17,02031(0.2)6,42814(0.22)5,98417(0.3)4,6080

tableissplitbyfragmentsscoredintheS0orS5generationandisfurtherdividedintototalmarkerdataanddatacollectedfromcolchicine(Colch.)orspontaneously(Spont.)doubledlines.

bExpectednumberoffragmentsisequaltotheproductoftotalmarkersscoredandthenumberoflines,minusmissingdata,andisthenumberexpectedunderparentaladditivity.

cDesignationsA(B.rapa),C(B.oleracea),andAC(commontoboth)indicatethetranscriptomeoforiginofthecDNA-AFLPmarkersscored.

PhenotypicVariationamongSelfedPolyploidLineagesOverallphenotypicvariability(measuredasthesumofSDacrossalltraits)increasedbetweentheS0andS5generations(Figure5A),andmostindividualtraitsexhibitedasignificantincreaseinphenotypicvariation(examplesofafewtraitsshowninFigures5Band5C;seeSupplementalTable2online).OneS5allopoly-ploid(line9)displayedadwarf-likephenotype(Figure5D).TherewasalsovariationinleafmorphologyandcuticularwaxinessamongS5polyploidlines(Figure5E).

ThenumbersofDNAandcDNAfragmentlossesperlineintheS5generationwerebothsignificantlycorrelatedwithourmetricforoverallphenotypicvariabilityintheS5(Table3).Totalmeth-ylationchangeswerenotsignificantlycorrelatedwithphenotypicvariability.Amultivariateanalysisofvariance(MANOVA)jointtestconfirmedthatthenumberofDNAfragmentlosseshadanoveralleffectonthemeansofallphenotypesintheS5generation(Wilks’Lambda;F¼2.3;P¼0.0317).NosignificantoveralleffectsonphenotypesweredetectedforcDNAfragmentlossesortotalmethylationchangesineithergeneration.UnivariatetestsdetectedasignificanteffectofDNAfragmentlossesonthenumberofopenflowersatfirstflowerintheS0generation(P¼0.0249)andeffectsofcDNAfragmentlossesonthenumberofopenflowersatfirstflowerandracemeheightintheS5gener-ation(P¼0.032and0.017,respectively).

UsingRFLPmarkerdatafromFLCloci(FLC-1,FLC-2,FLC-3,andFLC-5),amultiplelinearregression(MLR)wasperformedtomeasurewhetherornotrearrangementsinvolvinganyofthesefloweringtimelociwereassociatedwithfloweringtimevariationamongthepolyploidlines(Figure5B).TheanalysisindicatedthatthepresenceorabsenceofB.oleraceahomoeoallelesofFLC-1andFLC-3togetherexplained28%ofthevariationinfloweringtimeamongadvancedpolyploids(P<0.01).

DISCUSSION

GeneticandEpigeneticChangesinB.napus

SincetheseallopolyploidlinesweredevelopedfromdoubledhaploidgenotypesofB.oleraceaandB.rapa,ournullhypothesiswasthatalllineswouldhaveidenticalgenomes.ThishypothesiswasrejectedintheS0generationduetoDNAmethylationchanges(Lukensetal.,2006).IntheS5generation,mostoftheearliermethylationchangesremainedfixed(seeSupplementalFigure6online),butasmallproportionrevertedandsomenewchangeswereobserved.IntheS5generation,thenullhypothesiswasrejectedprimarilyduetogeneticchanges.EverylinelostRFLPorSSRfragmentsbytheS5generation(Figure1).Althoughsomelossesmayrepresentdeletions,manyoftheRFLP

Table3.PearsonCorrelationsandPValuesforPairwiseComparisonsofMarkerandPhenotypicDataSetsamongtheS5GenerationResynthesizedB.napusPolyploids

No.DNALosses

No.DNAlossesaNo.cDNAlossesaNo.methylationchangesaPhenotypicvariabilityaaDNA

No.cDNALossesP<0.0001

No.MethylationChangesP¼0.7544P¼0.6716r¼0.192

PhenotypicVariabilityP¼0.0041P¼0.0134P¼0.2127

r¼0.742r¼0.048r¼0.424

r¼0.065r¼0.370

¼totalDNAfragmentlosses/line;cDNA¼totalcDNAfragmentlosses/line;methylation¼totalmethylationchanges(sumoffragmentloss,

fragmentreversions,novelfragmentsacrossHpaIIandMspIblots)/line;phenotypicvariability¼sumofSDfromthemeanofalltraits/line).

3410ThePlantCell

Figure5.ExamplesofPhenotypicVariationamongResynthesizedB.napusPolyploids.

(A)to(C)FrequencydistributionsofS0(graybars)andS5(blackbars)generationsfortotalphenotypicvariabilityrepresentedassummationofalltraitvaluesinSDunits(A),floweringtime(B),andthenumberoffertilesiliques(C).ThevaluesforthediploidparentsB.rapa(lineIMB218;A-genome)andB.oleracea(lineTO1000;C-genome)areindicatedwitharrowsin(B)and(C).NotethatthismeasureoffertilesiliqueswasconductedintheabsenceofpollinatorsandthatwhiletheTO1000isself-compatibleitdoesnotreadilyself-pollinate.

(D)PhotographsofoneS5line(no.9)displayingadwarfedgrowthhabitthatwaspronouncedbytheS4generation.

(E)Variationinshapeofthefourthleaffromdiploidparents(IMB218farleft,andTO1000atfarright)andsomeS5polyploidlines(shownbetweenparentalsamples).Bar¼20cm.

fragmentlosseswereassociatedwithanintensificationofahomoeologousfragment,suggestingHNRTs.Thelargesizeofourresynthesizedallopolyploidpopulationallowedustoaddressthedynamicsofgenome-widechanges.The47lineswerenormallydistributedforgenetic,epigenetic,andexpressionchanges,aswouldbeexpectedifthechangesoccurredinde-pendentlyamonglines.Althoughgeneticchangesoccurredonnearlyeverychromosome,theywerenotequallydistributedamongchromosomes,occurringathigherfrequenciesonchro-mosomeswithlargeregionsofhomoeologybetweentheA-andC-genomes(e.g.,A1/C1,A2/C2,andA3/C3).AdetailedanalysisofHNRTbetweenA1andC1demonstratedvariouscombina-tionsofnonreciprocaleventsontheendsofchromosomesandininterstitialregions(Figure3).

GeneticchangesintheS5occurredmorefrequentlyintheC-subgenome(althoughthedifferencewasmarginallysignifi-cant),whilecytosinemethylationchangesoccurredmorefrequentlyintheA-subgenome.Inapreviousstudywithresyn-thesizedB.napus,Songetal.(1995)analyzedreciprocalallo-polyploids(AACCandCCAA)anddidnotdetectsignificantdirectionalchangesamongtheA-andC-subgenomes;however,theyincludedonlyafewlineagesderivedfromasinglehybridofeachreciprocalcross.WeanalyzedonlyCCAApolyploids,butweincludedmanyindependentlyderivedpolyploidevents.Im-portantly,weobservedlargevariationamongindependentlinesforboththeextentanddirectionofgenomicchange(Figure1),highlightingtheimportanceofanalyzingmanypolyploideventstoaccuratelyassessthedynamicsofgenomechange.Genomicchangesdirectedtowardoneparentalgenomehavebeenob-servedinBrassicajuncea,inwhichthedirectionwasdependentonthelineageanalyzed(Songetal.,1995),andinTriticalepolyploids,inwhichtherye(Secalecereale)subgenometendedtobepreferentiallymodified(Maetal.,2004;MaandGustafson,2006).ThecauseofvariationinnumberofrearrangementsobservedamongourS5linesisnotknown;however,itappearstobeunrelatedtovariationinmethylationchangesintheS0orS5generationssincetheseparameterswerenotcorrelated(r¼0.09and0.05,respectively).

PotentialMechanismsforChromosomalRearrangementbyHNRT

RFLPmarkersallowforthedetectionofboththelossandtheduplicationofhomoeologousmarkers(Parkinetal.,1995;Sharpeetal.,1995;Piresetal.,2004;Udalletal.,2005).Whenloss-intensificationpatternsareobservedamonglinkedhomoe-ologousmarkers,chromosomalrearrangements(HNRTs)canbediscriminatedfromdeletions.ThemechanismleadingtoHNRTsmostlikelyinvolvesrecombination.Inourpreviousstudy,wefoundthatgenomicrearrangementswererareintheS0gener-ationoftheseresynthesizedallopolyploidlineswhenbulkedDNAsamplesof8to12individualswereanalyzed(Lukensetal.,2006).Inthisstudy,weanalyzedindividualS1plantsanddetectedrearrangementssegregatingintheS1progenyatalowfrequency(3/118segregantswereobservedinthreeofeightlinesanalyzed),indicatingthatmeioseswererequiredfortheirfrequentoccurrencebytheS5.Ifareciprocalexchangeoccurredamongapairofhomoeologsduringmeiosis,thenoppositeforms(fixedforonehomoeologortheother)mighteachbeexpectedatafrequencyof1/16segregants.Thisestimateisconsistentwiththeobservationthat1/14ofthesegregantsderivedfromline17(S0)carriedtheoppositeformofaHNRTfixedinlatergenera-tionsforthatline(datanotshown)andthat1/16ofthesegregantsderivedfromline48(S0)carriedanHNRT(seeSupplementalFigure5online).

Furtherevidenceforarecombinationmechanismforchromo-somerearrangementswasdetectedinourtransgenerationalanalysisofancestorsoflinesgeneticallyfixedforA1/C1HNRTs(seeMethods).Theresultssuggestedthatanindividualpickedforselfingtwogenerationspriortofixationwasheterozygousforthenonreciprocaltransposition.Rearrangementsofvariablelengthwereobservedinsomesegregatingindividualsandmostlikely

GenomicChangesinBrassicanapus3411

resultedfromrecombinationamonghomoeologsfollowingsuc-cessivemeioses(seeSupplementalFigure3online).Thisissupportedbyourdetectionofpartialreciprocalrearrangementsamongsomesegregants(seeSupplementalFigure4online)andinterstitialrearrangements(Figure3).Therefore,thefixationofHNRTsobservedinthelatergenerationssuggeststhatstochas-ticforcesandsegregationduringselfingcanfixbothformsofagivenrearrangement.ThisisillustratedinFigure3,whichshowsthatbothformsofHNRTshavesegregatedfromreciprocalexchangesandbecomefixedamong52lineages.Insum,ourdatasupporttheroleofarecombination-basedmechanismforchromosomalrearrangements(HNRTs).

Geneticevidencesuggeststhatchromosomalrearrange-ments(HNRTs)occurinbothnaturalandresynthesizedB.napus(Sharpeetal.,1995;Songetal.,1995;Piresetal.,2004;Udalletal.,2005;Leflonetal.,2006;Liuetal.,2006;Nicolasetal.,2007).HomoeologousexchangeswereobservedcytologicallyforA7/C6reciprocaltranspositionsinpreviousstudieswithB.napusallopolyploids(Osbornetal.,2003a).B.oleraceaandB.rapashareextensivesimilarityamonghomoeologouschro-mosomes(Parkinetal.,2005),whichmaypromoterecombina-tion.ThisissupportedbystudiesinB.napushaploids(AC)andtriploids(AAC),whichhavedemonstratedthatpairingandex-changeamonghomoeologsoccurredatasignificantrateandwasunderthecontrolofgeneticloci(Jenczewskietal.,2003;Leflonetal.,2006;Liuetal.,2006;Nicolasetal.,2007).

Althoughdeletionshavebeenreportedinallopolyploidwheat,Triticale,andTragopogon(Ozkanetal.,2001;Shakedetal.,2001;Hanetal.,2005;MaandGustafson,2006;Tateetal.,2006),noneofthesestudiesreportduplicationofhomoeologousregionsassociatedwithdeletionevents.SomeofthesestudieshavereliedonPCR-basedmarkersystemsthatarenotdosagesensitive;thus,loss-duplicationeventsmayoccurinotherpoly-ploidspeciesbuthavenotbeendetected.Dosage-sensitiveRFLPmarkerswereusedfortwostudiesonresynthesizedB.junceapolyploids(Songetal.,1995;Axelssonetal.,2000),andneitheroneprovideddirectevidenceforexchangesbetweentheA-andB-subgenomes,whicharemoredivergentthantheA-andC-subgenomes.Axelssonetal.(2000)foundnoevidenceforanygenomicchangesinamappingstudy.Songetal.(1995)observedhighfrequenciesofDNAfragmentlossesandgainsinself-pollinatedlineages,butmarkerdataconsistentwithHNRTs(lossintensificationofhomoeologousfragments)werenotpre-sented.Thus,othermechanismsofgenomicchangemaypre-dominateinB.juncea.

GenomeRearrangementsAlterGeneExpressioninB.napusPolyploids

Wedetectedacorrelationbetweengenomerearrangementsandchangesintheexpressionoftargetedgenesandanonymoustranscriptmarkers.ForpW225transcripts,multipleSSCPmarkersweredetectedthatmaintainedconsistentsignalinten-sitywhileasinglepairofhomoeologousproductsrevealedlossandintensificationpatternsthatmirroredRFLPpatterns(Figures2and4).Thesepatternsmighthavebeenpredictediftheleveloftranscriptderivedfromeachhomoeologoussequencewasproportionaltoitscopynumber.AlthoughFLC-3andFLC-5

3412ThePlantCell

hadHNRTs,wecouldonlyinferthelossofhomoeologousFLC-3andB.rapaFLC-5transcriptsusingRT-PCR(seeSupplementalFigure7online).Nevertheless,genomerearrangementsalteredthecontributionofparentalallelestothepolyploidtranscrip-tomesandwereanimportantmechanismgeneratingallelicvariation.

Ourgenome-wideexpressionsurveyoftheS0generationrevealedasimilarfrequencyofcDNA-AFLPmarkersdisplayingchangesaswasdetectedintheearlygenerationsofresynthe-sizedwheatandArabidopsis(Comaietal.,2000;LeeandChen,2001;Kashkushetal.,2002).Neithergeneticchangesnorcyto-sinemethylationchanges(detectablebyHpaIIandMspI)werecorrelatedwithcDNA-AFLPchangesintheS0generation;how-ever,itispossiblethatsomeoftheselosttranscriptsresultedfromotherepigeneticmodifications.Thisissupportedbyourob-servationthatsomemissingcDNA-AFLPmarkerfragmentsamongS0polyploidsappearedintheS5generation,whichwouldpresumablyoccuronlyifthesilencingmechanismwasreversible.ThecDNA-AFLPmarkerlossesintheS5generationwerehighlycorrelatedwithDNAfragmentlosses,suggestingthatgenomerearrangementswerethemajorcauseofgeneexpres-sionchanges(Table3).StudiesinwheathavealsodetectedthelossofcDNA-AFLPfragmentsfollowingintergenerichybridiza-tionandconfirmedthatsequencepolymorphismswererespon-sibleforsomeexpressionchanges(Kashkushetal.,2002).InTragopogonallopolyploids,themajorityofcDNA-AFLPmarkerlossesresultedfromgeneticchanges(Tateetal.,2006).AsintheS0generation,wedidnotdetectarelationshipbetweencytosinemethylationchangesandcDNA-AFLPfragmentlossesintheS5generation.Thisresultsuggeststhatcytosinemethylationchangewasnotanimportantmechanismonagenome-widescalecausingqualitativechangesingeneexpressioninresyn-thesizedB.napus,ashasbeenobservedforspecificlociinstudieswithArabidopsisandwheat(ChenandPikaard,1997a;Kashkushetal.,2002).However,itispossiblethatcytosinemeth-ylationchangescausedquantitativetranscriptionalchangesorledtoqualitativesilencingofonlyafewloci.ItisalsopossiblethatotherDNAandchromatinmodificationsmightexplainadditionalvariationinthelossofcDNA-AFLPfragments(Osbornetal.,2003b;ChenandNi,2006).WedetectednooverallbiasinthefrequencyofA-orC-genome–specificcDNA-AFLPchangesamongS5polyploids.Incotton,organ-specificbiasesinthesilencingofhomoeologshasbeenobserved;however,thereisnooverallbiasintheexpressionoftranscriptsfromoneparentalcottongenomeovertheother(Adamsetal.,2003,2004;AdamsandWendel,2004).Conversely,inArabidopsissuecicaallopoly-ploids,silencingofA.thalianahomoeologsoccurredmorefre-quentlythanthosederivedfromArabidopsisarenosa(Wangetal.,2006).Aswithgenomicchanges,weobservedlargevariationamongourlinesfortheextentanddirectionofgenelossorsilencing,againhighlightingtheimportanceofanalyzingalargenumberoflineages.

GenomeRearrangementsanddeNovoPhenotypicVariationPhenotypicvariationincreasedwithgenerationadvancementandwassignificantlycorrelatedwithDNAfragmentlosses(Table3),suggestingthatchromosomalrearrangementsmaybroadlyimpactphenotypesinresynthesizedB.napuspolyploids.Phe-notypicvariationwasalsosignificantlycorrelatedwithcDNAfragmentlosses,althoughitwasnotashighasthecorrelationwithDNAloss,andtherewasnosignificantoveralleffectofcDNA-AFLPlossesonphenotypes(MANOVAresults).TheweakerassociationofphenotypicvariationwithtranscriptlosscomparedwithDNAlossmaybeduetosamplingvariationforDNAandcDNAlosses.ItisalsopossiblethatsincetheDNAlossesmayreflectrearrangementsthatincludelargenumbersofgeneticallylinkedtranscriptchanges,theymoreaccuratelyreflecteffectsonphenotypethandocDNAlosses.Inadditiontociseffects,homoeologouschromosomalrearrangementsmayhavetranseffectsthatcausechangesintheexpressionofgeneslocatedoutsideofrearrangedchromosomalsegments.Thiswouldbeparticularlytrueofgeneslyingwithinrearrangementsthatarecomponentsofregulatorycomplexesandthathavedivergedsignificantlyintheprogenitorspecies,suchthatageneticlossorlossduplicationleadstoaltereddosageandregulatoryinteractions(Osbornetal.,2003b;RiddleandBirchler,2003;Birchleretal.,2005).Thus,theinclusionofgenome-wideestimatesofquantitativevariationingeneexpression,inadditiontoestimatesofqualitativetranscriptloss,maybetterreflecttheoveralleffectsofexpressionchangesonphenotypicvariation.SchranzandOsborn(2000,2004)previouslydescribeddenovophenotypicvariationinresynthesizedB.napuspolyploidsandconcludedthatonlyafewgeneticchangesmightbenec-essaryforitsinduction,particularlyiftheyinvolvedimportantgenes.Geneticvariationgeneratedbythereplacementofonehomoeoallelewithanothermightresultinnovelphenotypeswhenthetwohomoeoalleleshavedifferentmagnitudesofeffectsonatraitandactinanadditivemannerorresultintheunmaskingofrecessivealleles.StudiesinB.rapacharacterizedFLCalleleswithadditiveeffects,suggestingthatallelicvariationamongreplicatedFLClocimaycontributetovariationinfloweringtimeinBrassica(Schranzetal.,2002).Piresetal.(2004)detectedrearrangements,includingFLC,thataffectedfloweringtimeinacosegregationanalysiswithresynthesizedB.napuspolyploids,aresultsupportedbyourfindingsthatHNRTsatFLC-3andFLC-1wereassociatedwithflowering-timevariationamongS5poly-ploids.Bothreciprocalexchangesandnonreciprocaltransposi-tionshavebeenobservedinnaturalpopulationsofB.napus,suggestingthatintergenomicrearrangementmaybeanimpor-tantmechanismgeneratingphenotypicvariationinthisspecies(Parkinetal.,1995;Sharpeetal.,1995;Osbornetal.,2003a;Osborn,2004;Piresetal.,2004;Udalletal.,2005).Infact,studiesinnaturalB.napushavereportedHNRTswithphenotypiceffectsonseedyield(Osbornetal.,2003a)andSclerotiniaresistance(Zhaoetal.,2006).Importantly,someofthechromo-somesalteredbyHNRTsinthesestudieswithnaturalB.napusalsoincurredrearrangementsinourstudyofresynthesizedB.napusallopolyploids(e.g.,A1/C1,A2/C2,andA3/C3),andsomeRFLPmarkersthathavedetectedHNRTsinpreviousstudiesdetectedtheminourpopulationofpolyploids(i.e.,pW239;seeUdalletal.,2005).Therefore,itislikelythatgenomichotspotsforchromosomerearrangementexistinB.napus,someofwhichhavespecificphenotypiceffects.Variationmightalsobecreatedifgenesaredeletedorduplicatedintheregionsofanexchangeduetomisalignmentduringrecombination.

AlthoughnaturalB.napushaslowerfrequenciesofdenovorearrangementsthanresynthesizedlines(Parkinetal.,1995;Sharpeetal.,1995;Songetal.,1995;Piresetal.,2004;Udalletal.,2005),itispossiblethatchromosomepairinginnaturalB.napuswasinitiallymoreunstableandthatstrongerdisomicpairingevolvedwithselectionforimprovedseedset.Inourstudy,weintentionallymaintainedalllinesduringtheadvance-mentofgenerations,includinglinesthathaddecreasedfertilityorpollenviability(Figure5C),andtheselineswouldhavebeenquicklyselectedagainstinanaturalenvironment.WecannotruleoutthepossibilitythatnaturalB.napusinitiallyacquiredstrongergeneticcontroloverchromosomepairingfromoneofitsdiploidprogenitors.ItisalsopossiblethatparentsofnaturalB.napusdifferedforhomoeoallelesatlocicontrollingpairingandthathomoeologousexchangescontributedtovariationforpairingcontrolandtheevolutionofmorestabilizedpairing.Regardless,ourresultsshowthatthismechanismcanbeparticularlyprev-alentintheearlygenerationsafterpolyploidformation,whennovelphenotypicvariationmaybecriticalforthesurvivalandpropagationofanewspecies.Furtherresearchwillbenecessarytodeterminetowhatextenthomoeologouschromosomeex-changescontributetonovelvariationinotherpolyploidspecies.

METHODSPlantMaterial

Morethan50resynthesizedBrassicanapusallopolyploidplants(CCAA)weredevelopedbyhybridizingdoubledhaploidBrassicaoleracealineTO1000(eggdonor;C-genome)withdoubledhaploidBrassicarapalineIMB218(pollendonor;A-genome)asdescribedpreviously(Lukensetal.,2006).Forty-twoamphidiploidlinesweregeneratedbycolchicinetreat-ment,andtheremaining10linesunderwentspontaneouschromosomedoubling.Lineageswereadvancedbysingleplantdescent.Ateachgenerationfollowingself-fertilization,sixseedswereplantedforeachgenotypetoensuregermination.Twoofthesixplantswerechosenatrandomandself-fertilizedtoensureadequateseedsetforeachgeno-type.Ifbothplantsproduced>50seeds,onewaschosenatrandomforadvancementtothenextgeneration;otherwise,theplantwiththehigherseedsetwasselected.LinenumberdesignationsusedinallfiguresarebasedonthoseshowninFigure1(seeSupplementalDataSet1onlineforlines,primers,probes,andgenotypenumbers).Twolinesweremosaicsofspontaneousallopolyploid(CCAA)andhybridallohaploidtissue(CA).Cuttingstakenfromthehybridtissuesweretreatedwithcolchicine,generatinglines(19)and(17)(Figure1).Thespontaneousallohaploidcuttingsgaverisetolines28and39,whichwereexcludedfromallS5analysesasdescribed(Lukensetal.,2006).ThreelinesthatdidnotreachtheS5generationwereexcludedfromthedataanalysis(Figure1).BulksofS1plantsandS6plantswereusedtoestimategeneticvariationintheS0andS5generations,respectively.Atotalof47allopolyploidplantswereanalyzedforgeneticandepigeneticchangesattheS5generationandincludedtheadditionalline45(seeSupplementalDataSet1onlineforlines,primers,andprobes).FortestingFLCmarkerphenotypeassocia-tions,S4line51wasusedalongwithallotherlinesadvancedtotheS5.Forty-fourindependentlineswerecommontoalldatasets(seeSupple-mentalDataSet1onlineforlines,primers,andprobes).DNAGelBlottingandMicrosatelliteAnalysis

GenomicDNAwasextractedfromleaftissuefromyoungplantsusingthecetyl-trimethyl-ammoniumbromidemethod(KidwellandOsborn,1992).

GenomicChangesinBrassicanapus3413

Eachsamplewascomprisedofabulkofeightto12S6plants.DNAfromeachsamplewasdigestedwiththefollowingrestrictionendonucleases:EcoRI,HindIII,DraI,MspI,andHpaII.DNAgelblotting,probelabeling,andhybridizationswereperformedusing76BrassicacDNAandgenomicDNAprobesasdescribedinpreviousstudies(Ferreiraetal.,1994;Butruilleetal.,1999;Udalletal.,2005;Lukensetal.,2006).SomeprobeswerenotusedineithertheS0orS5analyses(seeSupplementalDataSet1onlineforlines,primers,andprobes).Probenomenclatureofpreviousstudieswasused(Parkinetal.,1995;Sharpeetal.,1995;Udalletal.,2005).Thirty-sevenSSRprimerpairswereusedtoassaychangesinmicrosatellites.Twenty-eightoftheseprimerpairsweredevelopedbyD.LydiateandA.SharpeandwereamplifiedbyPCRandresolvedbypolyacrylamideelectrophoresisaspreviouslydescribed(Lukensetal.,2006).NineSSRsusedinarecentmappingstudywerePCRamplifiedandresolvedbyelectrophoresisin3%agarosegels(ourunpublisheddata;UzunovaandEcke,1999;Suwabeetal.,2002).ForSSRs,allpoly-ploidsampleswererunadjacenttoreactionscontainingparentalDNAs,amixofparentalDNAs,andareactionwithnoDNA.

GenomeFragmentAnalysis

Toassayforgeneticchanges,fragmentlosseswerescorediftheywereconfirmedinatleasttwoofthefollowingmethylation-insensitiveen-zymes:EcoRI,HindIII,andDraI.Insomecases,changeswereobservedinonlyoneoftheseenzymesbutwerealsopresentinbothMspIandHpaIIdigestswithnoevidenceforaDNAmethylationchangeandwerescoredasgeneticchanges.Anestimateofthetotalnumberofmarkersforeachprobewasdeterminedbytherestrictionenzyme(EcoRI,HindIII,orDraI)thatgeneratedthelargestnumberofmarkerfragments.Markersthatgeneratedidenticalpatternsforagivenprobewerescoredasasinglemarkerfragment.Fragmentlosses(SSRorRFLP)wererecorded,andHNRTwasinferredifthelossofoneRFLPfragmentoccurredwithanintensification(asestimatedbyeye)ofsignalatahomoeologousRFLPfragment.Thistypeofloss/intensificationdosagepatternhaspreviouslybeeninferredtoresultfromHNRTs(Udalletal.,2005).Lossesthatwerenotconfirmedinmultipleenzymes(asstatedabove)andunreadablefragmentswerecodedasmissingdata.

MappingGenomeRearrangements

Mappositionsof133markerswereassignedusingdatafromotherB.rapa,B.oleracea,andB.napusmappingpopulations(ourunpublisheddata;Udalletal.,2005;D.LydiateandA.Sharpe,personalcommunica-tion).Formostofthesemarkers(94),thefragmentfromourresynthesizedB.napuswasconfirmedtobeequivalenttoafragmentmappedinpop-ulationsusingthesameparentallinesandprobes(TO1000orIMB218).For27ofthemarkers,onlyonelocusoronepairofloci(A-andC-genomes)hadbeenmappedwiththesameprobe,andtheseassign-mentswereused.For12markerloci,mappingdatapredictedtwoormorepotentiallociwithinadiploidgenome,andweassignedthesemarkerlocitothelinkagegroupcontaininglinkedmarkersthatshowedcoincidentlossesamongourpolyploidlines.ForhomoeologouslinkagegroupsA1andC1,JoinMapwasusedtoapproximatetheorderofassignedmarkersusingmarkerpolymorphismdataderivedfromthepolyploidylines.TheBrassicalinkagegroupnomenclatureisundergoingachange:B.rapa,B.oleracea,andB.napuslinkagegroupswerepreviouslydesignatedR1-R10,O1-O9,andN1-N19,respectively.ThisnomenclatureisbeingreplacedbylinkagegroupsthatrefertotheA(B.rapa)andC(B.oleracea)genomes.Thus,thenewnomenclatureisA1-A10(¼R1-R10),C1-C9(¼O1-O9)fordiploids,andinallopolyploid,B.napusN1-N19isreplacedbyA1-A10andC1-C9designations(seehttp://www.brassica.info/information/lg_assigments.htm).

3414ThePlantCell

MethylationAnalysis

DNAmethylationchangeswerescoredbycomparingHpaIIandMspIblots(for67probes)aspreviouslydescribed(Lukensetal.,2006).TheabsenceofaparentalMspIorHpaIIfragmentinanallopolyploidwasinterpretedasachangeinCpCpGorCpGmethylation,respectively.ChangesobservedinMspIorHpaIIblotsthatwereobservedinanymethylation-insensitiverestrictiondigestsandchangesobservedinbothMspIandHpaIIblotsnotobservedintheotherrestrictiondigestswerescoredasmissingdata.DNAmethylationchangesintheS5generationwerecategorizedasfixed(havingoccurredintheS0generationandremainedunchanged),reversionsofpreviousmethylationchanges,ornewfragmentchanges(fragmentlossesþnovelfragments).ThetotalnumberofS5methylationchangeswastakenasthesumofreversionsandnewfragmentchangesintheS5.OnlylinesandprobesincommonbetweentheS0andS5generationswereanalyzedformethylationdatainthisarticle.

TransgenerationandS1SegregantAnalysis

Sixlines(2,26,8,22,17,and48)thatrevealedinterstitialorterminalizedrearrangementsonhomoeologouslinkagegroupsA1andC1(seeFigure3)weregenotypedwithlinkedSSCPandSSRmarkers(pW225,pX135,SORC20,andSORD38)acrossallgenerations.SSCPmarkerspW241,pW225,andpX135weredesignedfromRFLPprobesequencesavailableonlineatGenBank(seeSupplementalDataSet1onlineforSSCPprimersequencesandRFLPprobes).BulkedDNAsamples(of16individualplantsperline)weregenotypedtoidentifyinwhichgenerationrearrange-mentswerefixedintheselines,andthenthe14to16individualplantDNAswereanalyzedinthepriorgeneration.

TodetermineifrearrangementshadoccurredfollowingmeiosisoftheS0generation,10to16individualS1plantsderivedfromeachofeightS0plants(26,17,22,18,48,2,8,and27)weregenotypedusingacombinationofSSRs(SORC20)andSSCPmarkersdesignedfromRFLPprobes(FLC3,pW241,andpW225)thatspannedhomoeologouslinkagegroupsA1/C1,A2/C2,andA3/C3(seeSupplementalDataSet1onlineforlines,primers,andprobes).cDNA-AFLPAnalysis

SeedwassowninMetroMixSoilinflatscontaining14rowsofsevencellseach(cellswere;3.81cmindiameterand;20cmdeep).Flatswereorganizedintotworeplicateblocksina1002footindoorgrowroomattheUniversityofWisconsinBiotronFacility.Outerrowsofeachflatwereseededwithborderplants,andeachoftheremaininginner12rowswereseededinrandomorderwithanindividualS0,S5,orparentallineofB.rapaorB.oleracea(eachcellwasdoubleseededandthinnedtooneplant).Eachblockcontained10flats,andeachlinewasrepresentedonceperblock.Withineachblock,thepositionsofflatswererandomizeddaily.Plantswerewatereddailyandfertilizedeveryotherdaywithdilute(1tableapoon/20liters)PetersProfessionalPeatLiteSpecial20-10-20.Temperaturewasmaintainedat218C,humidityat60%,andlightmain-tainedat;450mmol/m2/sÀ1for16heachday.

Allplantswereharvestedatthesamedevelopmentalstage,whenthesecondandthirdtrueleaveswereoutstretchedfromthemeristem.Plantswereharvestedatthesametimeofday(11:00AMto12:00PMCST).Leavestwoandthreewerebulkedfromthesevenplantscomprisingagivenlinereplicateandwereflash-frozeninliquidN2andstoredatÀ808C.TotalRNAwasextractedusingTri-Reagent(MolecularResearchCenter)accordingtothemanufacturer’sprotocol.RNAsampleswereDNasetreatedwithAmbionTurboDNA-freeDNaseandquantifiedusingaBio-RadSmartSpec3000.mRNAwaspurifiedfrom;50mgoftotalRNA/sampleusingtheQiagenOligotexmRNApurificationkitaccordingtothemanufacturer’sprotocol.EachmRNAsamplewaselutedwith40mLof

elutionbuffer.First-strandcDNAsynthesiswasconductedasfollows:20mLofelutedmRNAwascombinedwith1mLofoligo(dT)primer(20mM)andheatedto708Cfor5min;samplesweresnap-cooledoniceandaddedto18mLofdistilleddeionizedwater,5mLof103RTbuffer,5mLof10mMdeoxynucleotidetriphosphate(dNTP),and1mLofM-MLVreversetranscriptase(NewEnglandBiolabs);andreactionswereincubatedat378Cfor45min.Second-strandcDNAsynthesiswasperformedasfollows:50mLoffirst-strandcDNAwascombinedwith38.5mLofdistilleddeionizedwater,0.5mLofRNaseH,1mLofEscherichiacoliDNAPoly-meraseI,and10mLofNewEnglandBiolabsbuffer2;andreactionswereincubatedat168Cfor2h.Thirty-fourmicrolitersofdistilleddeionizedwater,1mLofE.coliDNAligase,and15mLofDNAligasebuffer(NewEnglandBiolabs)wereadded,andreactionswereincubatedatroomtemperaturefor15min.Double-strandedcDNAswerepurifiedusingphenol:chloroformextractionandresuspendedin20mLofdistilleddeionizedwater.Tenmicrolitersofdouble-strandedcDNAwasdigestedwithTaqIandAseIrestrictionenzymesaccordingtothemanufacturer’sprotocol(NewEnglandBiolabs).Primersandadaptersweredesignedusingpublishedprotocols(Bachemetal.,1996,1998)andcontainedeitheraTaqIorAseIrestrictionsite.Alladapter,preamplification,andselectiveprimersequencescanbefoundinSupplementalDataSet1online.Adapterswereligatedtorestriction-digesteddouble-strandedcDNAsamplesusingT4DNALigase(NewEnglandBiolabs),sampleswerediluted1:20,and1mLwasusedastemplatesinpreamplificationreactions.PCRreactionconditionswereasfollows:1mLoftemplatewascombinedwith35mLofdistilleddeionizedwater,5mLof103PCRBuffer,4mLof25mMMgCl2,2.5mLof2mMdNTP,1mLofeachprimer(5mM),and0.5mLofTaqpolymerase.Reactionswereplacedinathermocylerunderthefollowingprogram:(1)728Cfor30s,(2)948Cfor3min,(3)30cyclesof948Cfor1min,558Cfor1min,728Cfor2min,and(4)728Cfor5min.Forselectiveamplifications,TaqI-selectiveprimerswerelabeledwithP33usingT4polynucleotideKinase(NewEnglandBiolabs)accordingtothefollowingprocedure:10.8mLofTaq1primer(5mM)wascombinedwith7.8mLofdistilleddeionizedwater,3mLof103polynucleotidekinasebuffer,6mLofg-dCTP33,and2.4mLofpolynucleotidekinaseandwasincubatedat378Cfor30min(seeSupplementalDataSet1onlineforlines,primers,andprobes).SelectivePCRamplificationsweresetupbycombining2mLof1:20dilutedpreamplification,1mLof103PCRbuffer,0.8mLof25mMMgCl2,1mLofdNTP(2mM),1mLofAseIprimer(5mM),0.25mLoflabeledTaqIprimer,0.5mLofTaqpolymerase,and3.5mLofdistilleddeionizedwater.ReactionswerethermocycledaccordingtothefollowingPCRprogram:(1)948Cfor3min,(2)11cyclesof948Cfor1min,658Cfor1min(À0.78C/cycle),and728Cfor2min,(3)22cyclesof948Cfor1min,558Cfor1min,728Cfor2min,and(4)728Cfor5min.PAGEwasconductedat65Wofconstantpoweraccordingtostandardprotocols.AllcDNA-AFLPreactionswereseparatedwithsamplesofparentalandparentalmixcDNAandawatercontrol.TocontrolforDNAcontamination,DNA-AFLPswereconductedinparallelonparentalsamplesandtworandompolyploidlinesandwereusedascontrolsbesidecDNA-AFLPsforthefiveprimersetsanalyzed.DNA-AFLPcontrolsgeneratedmanymorebandsthancDNA-AFLPsandhaduniqueprofiles.Additionally,(RT-)controlswereperformedbyscreeningsamplesofmRNAbyPCRusingtwogene-specificprimersets(FLC-3andFLC-5;seebelow).cDNA-AFLPresultsthatwerenotrepeatableinbothbiologicalreplicateswerescoredasmissingdata.Sixteenprimercombinationsgeneratedatotalof360transcriptloci,ofwhich137wereofB.rapaorigin,127wereofB.oleraceaorigin,and96werecommontobothparentaltranscriptomes(seeSupplementalDataSet1onlineforlines,primers,andprobes).RT-PCRandDNASSCPAnalysis

ThesamemRNApreparationsusedforcDNA-AFLPanalysiswereusedforcDNAsynthesisinRT-PCRs.First-strandsynthesiswasconductedusing250ngofmRNA,M-MLVreversetranscriptase(NewEngland

Biolabs),andoligo(dT)primersaccordingtothemanufacture’sprotocol.Threespecificprimersets(FLC3,FLC5,andpW225)wereusedinRT-PCRSSCPanalyses(seeSupplementalDataSet1onlineforprimersequencesandPiresetal.,2004forastudythatpreviouslyusedourFLCprimers).pW225primersweredesignedfromtheRFLPprobepW225(accessionnumberCZ906459),whichwashighlysimilartotheArabi-dopsisgeneAT4G32551[p(N)¼4.3e-66,TheArabidopsisInformationResourceWU-Blast].Primersweredesignedinexons13and14basedontheAT4G32551annotation(CDSsequence1941to2136bp)andBrassicanucleotidesequence.pW225PCRproductsweresequencedtoverifytheiridentity.PCRamplificationsweresetupbycombining1.5mLof1:10dilutedcDNA,1mLof103PCRBuffer,0.8mLof25mMMgCl2,1mLofdNTP(2mM),1mLofFprimer(5mM),1mLofRprimer(5mM),0.5mLofTaqpolymerase,0.1mLofa-dCTP32,and3.1mLofH20.ReactionswerethermocycledaccordingtothefollowingPCRprogram:(1)968Cfor4min,(2)30cyclesof948Cfor30s,588Cfor45s,and728Cfor1min,(3)728Cfor10min,and(4)48Chold.a-dCTP32wasincorporatedduringPCRreactionsusingTaqpolymerase,andthesampleswerere-solvedbyMDEelectrophoresis(CambrexBioScience)usingBio-Radse-quencinggelsaccordingtothemanufacturer’sprotocol.Gelswererunfor17to25hat7Wconstantpower.AllRT-PCRSSCPreactionswererunbesidesamplesofparentalandparentalmixcDNAandDNA,awatercontrol.PhenotypicAnalysis

Seedwassownin6-inchpotsinMetroMixsoil.Potswereplacedonfour14.535-ftebbandflowtablesinanindoorgreenhouseroomattheUniversityofWisconsinBiotronFacility.Arandomizedcompleteblockdesignwasused,andeachoffourebbandflowtableswasconsideredablock.Withineachblock,eachoftheS0andS5polyploidlines,parentalgenotypes,andfournaturalB.napusgenotypeswererepresentedonce,andevery2weeksplantpositionswithinblockswererandomized.Temperaturewasmaintainedat218Candacombinationofnaturallightandsupplementallightingprovidedanaverageintensityof678mmole/m2/sÀ1acrossthegreenhouse.Eachtablehadtwo600-WHPSlightscenteredoverit,whichoperated16haday.Ebbandflowtableswerefloodedtwiceperdayfor10minwithhalf-strengthHoaglandfertilizer.Thirteenphenotypiccharactersweremeasuredinananalysisofvariationamongthepolyploidlines.Floweringtimewasmeasuredatthetimethefirstflowerbudopened.Atthistime,anumberofothertraitswasmeasured,includingleafnumber,thetotalnumberofopenflowersatfirstflower,plantheight,racemeheight,numberofserrationsonthemarginofthefourthtrueleaf,widthofthefourthtrueleafatitswidestpoint,lengthofthepetioleofthefourthtrueleaf,lengthoftheleafbladeofthefourthtrueleaf,andthelengthofpetioledisplayingleafwingswhenpresent.Twoweeksafterfloweringinitiated,weestimatedvariationinflowersizebymeasuringthelengthandwidthoftheopenfaceoftheseventhflowerfromthetopofeachraceme(flowerpetalswerespreadopenandmeasuredfromendtoend).Atharvest,thenumberoffertilizedsiliques(havingatleastonedevelopingseed)wascountedoneachplant,andthenumberofsecondarybrancheswascountedasthenumberthatbrokebud.

StatisticalAnalysesofGeneticandTranscriptionalMarkersandPhenotypicData

ForRFLP,SSR,andcDNA-AFLPfragmentdataanalyses,theprop.testfunctionofRstatisticalsoftware(RDevelopmentTeam,2006)wasusedfortwosampletestsofequalproportionswhereapplicable.MicrosoftExcelwasusedforsimplettestsandhistograms.SASVersion9.1forWindows(SASInstitute,2006)wasusedforalltestsofnormality(PROCUNIVARIATENORMAL).ThePROCLOGISTICandPROCGENMODfunctionswereusedtotestthenullhypothesesofequalproportionofgeneticchangesacrossalllinkagegroups.SASAnalystwasusedfor

GenomicChangesinBrassicanapus3415

Ftestsofequalvariance,correlationanalyses,andMLRs.MLRswereconductedusingtotaltranscriptmarkerslost/lineasthedependentvariableandthenumberoftotalDNAfragmentlosses/lineandtotalnumberofDNAmethylationchanges/lineasindependentvariables.AnMLRwasalsoperformedusingFLCmarkers(FLC1-oleracea,FLC3-oleracea,FLC5-oleracea,FLC2-rapa,FLC3-rapa,andFLC5-rapa)asindependentvariablesanddaysuntilfloweringasadependentvariable.ForallMLRs,modelswerebuiltusingstepwiseselection,andthecriteriaforenteringandstayinginamodelwerebothsetatP<0.01.ToestimateoverallphenotypicvariabilityamongsyntheticpolyploidsacrossboththeS0andS5generations,thefollowingmetricwascalculated:foreachphenotype,least-squaresmeansforallpolyploidlines(S0andS5)wereusedtocalculateagrandmeanandSDforthattrait.Eachline’sdeviationfromthegrandmeanwascalculatedinSDunits.Standarddeviationsweresummedacrosseveryphenotypiccharacteranalyzed,givingrisetoameasureofhowphenotypicallyvariableeachlinewasrelativetotheentirepopulationofpolyploidsacrossbothgenerations.MANOVAwascon-ductedusingSASVersion9.1andwasusedtotestwhetherornottotalDNAfragmentlosses,cDNAfragmentlosses,andmethylationchangeshadeffectsonthemeansofallphenotypictraitsjointly.Ourmodeltreatedthe13phenotypictraitsasdependentvariablesandthethreemarkerdatasetsasindependentvariables.Univariatetestsforeverytraitwereperformedandthenthemanovah¼_all_functionwasusedtotesttheoveralleffectoftheindependentvariablesonalltraitssimultaneously.AccessionNumbers

SequencedatafromthisarticlecanbefoundintheGenBank/EMBLdatalibrariesunderaccessionnumbersCZ906459,CZ906477,CZ906395,DT469125,DT469153,CZ906400,CZ906385,DT469132,CZ906392,CZ906429,DT469129,CZ906379,CZ906384,AY115673,AY115675,AY115678,andAY115676.AcompletelistofaccessionnumbersforallRFLPprobescanbefoundinSupplementalDataSet1online.SupplementalData

Thefollowingmaterialsareavailableintheonlineversionofthisarticle.SupplementalFigure1.AnalysisofDNASamplesacrossGenerations.SupplementalFigure2.DNASSCPandSSRGelAnalysesofDNASamplesacrossGenerationsforTwoLineages.

SupplementalFigure3.DNASSCPandSSRAnalysesofLine2(S2)Segregants.

SupplementalFigure4.DNASSCPandSSRAnalysesofLine26(S1)Segregants.

SupplementalFigure5.DNASSCPAnalysisofFLC-3inLine48(S0)Segregants.

SupplementalFigure6.SummaryofMethylation-SensitiveRFLPChangesDetectedamongS0andS5PolyploidsRelativetotheParents.SupplementalFigure7.RT-PCRAnalysisofFLC-3andFLC-5.SupplementalFigure8.cDNA-AFLPDisplayofResynthesizedBrassicanapusPolyploids.

SupplementalFigure9.RelationshipbetweenTotalNumberofGenomeFragmentLossesandTotalNumberofcDNAFragmentLossesinResynthesizedBrassicanapusPolyploids.

SupplementalTable1.SummaryoftheDistributionofMarkerChangesamongBrassicaLinkageGroups.

SupplementalTable2.FTestsforEqualPhenotypicVarianceamongLinesofResynthesizedBrassicanapusPolyploidsintheS0andS5Generations.

SupplementalDataSet1.Lines,Primers,andProbes.

3416ThePlantCell

ACKNOWLEDGMENTS

WeacknowledgefundingfromNationalScienceFoundationPlantGe-nomeProgram(DBI-0077774andDBI-0501712).WethankDerekLydiateandAndrewSharpeforsharingSSRprimersandmapinformation,RobertVogelzangforassistancewithDNAgelblotpreparations,andMaqsoodRehmanandPatrickEdgerforDNAextractionsforthestudiesinsupple-mentaldata.CommentsfromBrianDilkesandanonymousreviewershelpedimprovethemanuscript.WealsothankRebeccaDoerge,PabloQuijada,Ting-LiLin,andNathanTefftforstatisticalsuggestionsand/orusefuldiscussionsondataanalysisandUniversityofWisconsinBiotronFacilitypersonnelforassistancewithsettingupcontrolledgrowthcon-ditionsforourplantexperiments.

ReceivedJuly18,2007;revisedOctober5,2007;acceptedOctober21,2007;publishedNovember16,2007.

REFERENCES

Adams,K.L.,Cronn,R.,Percifield,R.,andWendel,J.F.(2003).Genesduplicatedbypolyploidyshowunequalcontributionstothetran-scriptomeandorgan-specificreciprocalsilencing.Proc.Natl.Acad.Sci.USA100:4649–4654.

Adams,K.L.,Percifield,R.,andWendel,J.F.(2004).Organ-specificsilencingofduplicatedgenesinanewlysynthesizedcottonallotet-raploid.Genetics168:2217–2226.

Adams,K.L.,andWendel,J.F.(2004).Exploringthegenomicmysteriesofpolyploidyincotton.Biol.J.Linn.Soc.Lond.82:573–581.

Adams,K.L.,andWendel,J.F.(2005).Novelpatternsofgeneexpres-sioninpolyploidplants.TrendsGenet.21:539–543.

Axelsson,T.,Bowman,C.M.,Sharpe,A.G.,Lydiate,D.J.,andLagercrantz,U.(2000).AmphidiploidBrassicajunceacontainscon-servedprogenitorgenomes.Genome43:679–688.

Bachem,C.W.,vanderHoeven,R.S.,deBruijn,S.M.,Vreugdenhil,D.,Zabeau,M.,andVisser,R.G.(1996).VisualizationofdifferentialgeneexpressionusinganovelmethodofRNAfingerprintingbasedonAFLP:Analysisofgeneexpressionduringpotatotuberdevelopment.PlantJ.9:745–753.

Bachem,C.W.B.,Oomen,R.,andVisser,R.G.F.(1998).TranscriptimagingwithcDNA-AFLP:Astep-by-stepprotocol.PlantMol.Biol.Rep.16:157–173.

Birchler,J.A.,Riddle,N.C.,Auger,D.L.,andVeitia,R.A.(2005).Dosagebalanceingeneregulation:Biologicalimplications.TrendsGenet.21:219–226.

Butruille,D.V.,Guries,R.P.,andOsborn,T.C.(1999).LinkageanalysisofmolecularmarkersandquantitativetraitlociinpopulationsofinbredbackcrosslinesofBrassicanapusL.Genetics153:949–964.Chen,Z.J.,andNi,Z.F.(2006).Mechanismsofgenomicrearrange-mentsandgeneexpressionchangesinplantpolyploids.Bioessays28:240–252.

Chen,Z.J.,andPikaard,C.S.(1997a).EpigeneticsilencingofRNApolymeraseItranscription:AroleforDNAmethylationandhistonemodificationinnucleolardominance.GenesDev.11:2124–2136.Chen,Z.J.,andPikaard,C.S.(1997b).Transcriptionalanalysisofnucleolardominanceinpolyploidplants:Biasedexpression/silencingofprogenitorrRNAgenesisdevelopmentallyregulatedinBrassica.Proc.Natl.Acad.Sci.USA94:3442–3447.

Comai,L.(2000).Geneticandepigeneticinteractionsinallopolyploidplants.PlantMol.Biol.43:387–399.

Comai,L.,Tyagi,A.P.,Winter,K.,Holmes-Davis,R.,Reynolds,S.H.,Stevens,Y.,andByers,B.(2000).Phenotypicinstabilityandrapid

genesilencinginnewlyformedArabidopsisallotetraploids.PlantCell12:1551–1567.

Cui,L.Y.,etal.(2006).Widespreadgenomeduplicationsthroughoutthehistoryoffloweringplants.GenomeRes.16:738–749.

Ferreira,M.E.,Williams,P.H.,andOsborn,T.C.(1994).RFLPmappingofBrassicanapususingdoubledhaploidlines.Theor.Appl.Genet.89:615–621.

Guo,M.,Davis,D.,andBirchler,J.A.(1996).Dosageeffectsongeneexpressioninamaizeploidyseries.Genetics142:1349–1355.

Han,F.P.,Fedak,G.,Guo,W.L.,andLiu,B.(2005).Rapidandrepeatableeliminationofaparentalgenome-specificDNArepeat(pGcIR-1a)innewlysynthesizedwheatallopolyploids.Genetics170:1239–1245.

Hegarty,M.J.,Jones,J.M.,Wilson,I.D.,Barker,G.L.,Coghill,J.A.,Sanchez-Baracaldo,P.,Liu,G.Q.,Buggs,R.J.A.,Abbott,R.J.,Edwards,K.J.,andHiscock,S.J.(2005).Developmentofanony-mouscDNAmicroarraystostudychangestotheSeneciofloraltranscriptomeduringhybridspeciation.Mol.Ecol.14:2493–2510.Jenczewski,E.,Eber,F.,Grimaud,A.,Huet,S.,Lucas,M.O.,Monod,H.,andChevre,A.M.(2003).PrBn,amajorgenecontrollinghomoe-ologouspairinginoilseedrape(Brassicanapus)haploids.Genetics164:645–653.

Joly,S.,Rauscher,J.T.,Sherman-Broyles,S.L.,Brown,A.H.D.,andDoyle,J.J.(2004).Evolutionarydynamicsandpreferentialexpressionofhomoeologous18S–5.8S–26SnuclearribosomalgenesinnaturalandartificialGlycineallopolyploids.Mol.Biol.Evol.21:1409–1421.Kashkush,K.,Feldman,M.,andLevy,A.A.(2002).Geneloss,silenc-ingandactivationinanewlysynthesizedwheatallotetraploid.Genetics160:1651–1659.

Kashkush,K.,Feldman,M.,andLevy,A.A.(2003).Transcriptionalactivationofretrotransposonsalterstheexpressionofadjacentgenesinwheat.Nat.Genet.33:102–106.

Kenton,A.,Parokonny,A.S.,Gleba,Y.Y.,andBennett,M.D.(1993).CharacterizationoftheNicotianatabacumL.genomebymolecularcytogenetics.Mol.Genet.Genomics240:159–169.

Kidwell,K.K.,andOsborn,T.C.(1992).SimpleplantDNAisolationproceduresinplantgenomes:Methodsforgeneticandphysicalmapping,J.S.BeckmanandT.C.Osborn,eds(Dordrecht,TheNetherlands:KluwerAcademicPublishers),pp.1–13.

Kovarik,A.,Matyasek,R.,Lim,K.Y.,Skalicka,K.,Koukalova,B.,Knapp,S.,Chase,M.,andLeitch,A.R.(2004).Concertedevolutionof18–5.8–26SrDNArepeatsinNicotianaallotetraploids.Biol.J.Linn.Soc.Lond.82:615–625.

Kovarik,A.,Pires,J.C.,Leitch,A.R.,Lim,K.Y.,Sherwood,A.M.,Matyasek,R.,Rocca,J.,Soltis,D.E.,andSoltis,P.S.(2005).RapidconcertedevolutionofnuclearribosomalDNAintwoTragopogonallopolyploidsofrecentandrecurrentorigin.Genetics169:931–944.Lee,H.S.,andChen,Z.J.(2001).Protein-codinggenesareepigenet-icallyregulatedinArabidopsispolyploids.Proc.Natl.Acad.Sci.USA98:6753–6758.

Leflon,M.,Eber,F.,Letanneur,J.C.,Chelysheva,L.,Coriton,O.,Huteau,V.,Ryder,C.D.,Barker,G.,Jenczewski,E.,andChevre,A.M.(2006).PairingandrecombinationatmeiosisofBrassicarapa(AA)xBrassicanapus(AACC)hybrids.Theor.Appl.Genet.113:1467–1480.Levin,D.A.(1983).Polyploidyandnoveltyinfloweringplants.Am.Nat.122:1–25.

Lim,K.Y.,Matyasek,R.,Kovarik,A.,andLeitch,A.R.(2004).GenomeevolutioninallotetraploidNicotiana.Biol.J.Linn.Soc.Lond.82:599–606.

Lim,K.Y.,Souckova-Skalicka,K.,Sarasan,V.,Clarkson,J.J.,Chase,M.W.,Kovarik,A.,andLeitch,A.R.(2006).AgeneticappraisalofanewsyntheticNicotianatabacum(Solanaceae)andtheKostoffsynthetictobacco.Am.J.Bot.93:875–883.

Liu,B.,andWendel,J.F.(2003).Epigeneticphenomenaandtheevolutionofplantallopolyploids.MolPhylogenetEvol.29:365–379.Liu,Z.Q.,Adamczyk,K.,Manzanares-Dauleux,M.,Eber,F.,Lucas,M.O.,Delourme,R.,Chevre,A.M.,andJenczewski,E.(2006).MappingPrBnandotherquantitativetraitlociresponsibleforthecontrolofhomoeologouschromosomepairinginoilseedrape(Bras-sicanapusL.)haploids.Genetics174:1583–1596.

Lukens,L.N.,Pires,J.C.,Leon,E.,Vogelzang,R.,Oslach,L.,andOsborn,T.(2006).Patternsofsequencelossandcytosinemethyl-ationwithinapopulationofnewlyresynthesizedBrassicanapusallopolyploids.PlantPhysiol.140:336–348.

Ma,X.F.,Fang,P.,andGustafson,J.P.(2004).Polyploidization-inducedgenomevariationinTriticale.Genome47:839–848.

Ma,X.F.,andGustafson,J.P.(2006).TimingandrateofgenomevariationinTriticalefollowingallopolyploidization.Genome49:950–958.Madlung,A.,andComai,L.(2004).Theeffectofstressongenomeregulationandstructure.Ann.Bot.(Lond.)94:481–495.

Madlung,A.,Tyagi,A.P.,Watson,B.,Jiang,H.M.,Kagochi,T.,Doerge,R.W.,Martienssen,R.,andComai,L.(2005).GenomicchangesinsyntheticArabidopsispolyploids.PlantJ.41:221–230.Masterson,J.(1994).Stomatalsizeinfossilplants:Evidenceforpolyploidyinmajorityofangiosperms.Science264:421–424.

Meyers,L.A.,andLevin,D.A.(2006).Ontheabundanceofpolyploidsinfloweringplants.EvolutionInt.J.Org.Evolution60:1198–1206.Nicolas,S.D.,etal.(2007).HomoeologousrecombinationplaysamajorroleinchromosomerearrangementsthatoccurduringmeiosisofBrassicanapushaploids.Genetics175:487–503.

Osborn,T.C.(2004).ThecontributionofpolyploidytovariationinBrassicaspecies.Physiol.Plant.121:531–536.

Osborn,T.C.,Butrulle,D.V.,Sharpe,A.G.,Pickering,K.J.,Parkin,I.A.P.,Parker,J.S.,andLydiate,D.J.(2003a).DetectionandeffectsofahomoeologousreciprocaltranspositioninBrassicanapus.Ge-netics165:1569–1577.

Osborn,T.C.,Pires,J.C.,Birchler,J.A.,Auger,D.L.,Chen,Z.J.,Lee,H.S.,Comai,L.,Madlung,A.,Doerge,R.W.,Colot,V.,andMartienssen,R.A.(2003b).Understandingmechanismsofnovelgeneexpressioninpolyploids.TrendsGenet.19:141–147.

Otto,S.P.,andWhitton,J.(2000).Polyploidincidenceandevolution.Annu.Rev.Genet.34:401–437.

Ozkan,H.,Levy,A.A.,andFeldman,M.(2001).Allopolyploidy-inducedrapidgenomeevolutioninthewheat(Aegilops-Triticum)group.PlantCell13:1735–1747.

Parkin,I.A.P.,Gulden,S.M.,Sharpe,A.G.,Lukens,L.,Trick,M.,Osborn,T.C.,andLydiate,D.J.(2005).SegmentalstructureoftheBrassicanapusgenomebasedoncomparativeanalysiswithArabi-dopsisthaliana.Genetics171:765–781.

Parkin,I.A.P.,Sharpe,A.G.,Keith,D.J.,andLydiate,D.J.(1995).IdentificationoftheAandCgenomesofamphidiploidBrassicanapus(oilseedrape).Genome38:1122–1131.

Pires,J.C.,Zhao,J.W.,Schranz,M.E.,Leon,E.J.,Quijada,P.A.,Lukens,L.N.,andOsborn,T.C.(2004).FloweringtimedivergenceandgenomicrearrangementsinresynthesizedBrassicapolyploids(Brassicaceae).Biol.J.Linn.Soc.Lond.82:675–688.

Pontes,O.,Neves,N.,Silva,M.,Lewis,M.S.,Madlung,A.,Comai,L.,Viegas,W.,andPikaard,C.S.(2004).Chromosomallocusrearrange-mentsarearapidresponsetoformationoftheallotetraploidArabidop-sissuecicagenome.Proc.Natl.Acad.Sci.USA101:18240–18245.RDevelopmentTeam(2006).R:ALanguageandEnvironmentforStatisticalComputing.(Vienna,Austria:RFoundationforStatisticalComputing).

Ramsey,J.,andSchemske,D.W.(1998).Pathways,mechanisms,andratesofpolyploidformationinfloweringplants.Annu.Rev.Ecol.Syst.29:467–501.GenomicChangesinBrassicanapus3417

Ramsey,J.,andSchemske,D.W.(2002).Neopolyploidyinfloweringplants.Annu.Rev.Ecol.Syst.33:589–639.

Riddle,N.C.,andBirchler,J.A.(2003).Effectsofreuniteddivergedregulatoryhierarchiesinallopolyploidsandspecieshybrids.TrendsGenet.19:597–600.

Rieseberg,L.H.,andWillis,J.H.(2007).Plantspeciation.Science317:911–914.

Salmon,A.,Ainouche,M.L.,andWendel,J.F.(2005).GeneticandepigeneticconsequencesofrecenthybridizationandpolyploidyinSpartina(Poaceae).Mol.Ecol.14:1163–1175.

SASInstitute(2006).SAS/STATUsers’Guide.(Cary,NC:SASInstitute).Schranz,M.E.,andOsborn,T.C.(2000).NovelfloweringtimevariationintheresynthesizedpolyploidBrassicanapus.J.Hered.91:242–246.Schranz,M.E.,andOsborn,T.C.(2004).Denovovariationinlife-historytraitsandresponsestogrowthconditionsofresynthesizedpolyploidBrassicanapus(Brassicaceae).Am.J.Bot.91:174–183.Schranz,M.E.,Quijada,P.,Sung,S.B.,Lukens,L.,Amasino,R.,andOsborn,T.C.(2002).CharacterizationandeffectsofthereplicatedfloweringtimegeneFLCinBrassicarapa.Genetics162:1457–1468.Shaked,H.,Kashkush,K.,Ozkan,H.,Feldman,M.,andLevy,A.A.(2001).Sequenceeliminationandcytosinemethylationarerapidandreproducibleresponsesofthegenometowidehybridizationandallopolyploidyinwheat.PlantCell13:1749–1759.

Sharpe,A.G.,Parkin,I.A.P.,Keith,D.J.,andLydiate,D.J.(1995).Frequentnonreciprocaltranslocationsintheamphidiploidgenomeofoilseedrape(Brassicanapus).Genome38:1112–1121.

Soltis,D.E.,andSoltis,P.S.(1995).Thedynamicnatureofpolyploidgenomes.Proc.Natl.Acad.Sci.USA92:8089–8091.

Soltis,D.E.,Soltis,P.S.,andTate,J.A.(2004).Advancesinthestudyofpolyploidysinceplantspeciation.NewPhytol.161:173–191.

Song,K.M.,Lu,P.,Tang,K.L.,andOsborn,T.C.(1995).RapidgenomechangeinsyntheticpolyploidsofBrassicaanditsimplicationsforpolyploidevolution.Proc.Natl.Acad.Sci.USA92:7719–7723.

Suwabe,K.,Iketani,H.,Nunome,T.,Kage,T.,andHirai,M.(2002).IsolationandcharacterizationofmicrosatellitesinBrassicarapaL.Theor.Appl.Genet.104:1092–1098.

Tate,J.A.,Ni,Z.F.,Scheen,A.C.,Koh,J.,Gilbert,C.A.,Lefkowitz,D.,Chen,Z.J.,Soltis,P.S.,andSoltis,D.E.(2006).EvolutionandexpressionofhomoeologouslociinTragopogonmiscellus(Astera-ceae),arecentandreciprocallyformedallopolyploid.Genetics173:1599–1611.

Udall,J.A.,Quijada,P.A.,andOsborn,T.C.(2005).Detectionofchromosomalrearrangementsderivedfromhomoeologousrecombi-nationinfourmappingpopulationsofBrassicanapusL.Genetics169:967–979.

Uzunova,M.I.,andEcke,W.(1999).Abundance,polymorphismandgeneticmappingofmicrosatellitesinoilseedrape(BrassicanapusL.).PlantBreed.118:323–326.

Wang,J.L.,Tian,L.,Lee,H.S.,Wei,N.E.,Jiang,H.M.,Watson,B.,Madlung,A.,Osborn,T.C.,Doerge,R.W.,Comai,L.,andChen,Z.J.(2006).GenomewidenonadditivegeneregulationinArabidopsisallotetraploids.Genetics172:507–517.

Wang,J.L.,Tian,L.,Madlung,A.,Lee,H.S.,Chen,M.,Lee,J.J.,Watson,B.,Kagochi,T.,Comai,L.,andChen,Z.J.(2004).Sto-chasticandepigeneticchangesofgeneexpressioninArabidopsispolyploids.Genetics167:1961–1973.

Wendel,J.F.,Schnabel,A.,andSeelanan,T.(1995).Bidirectionalinterlocusconcertedevolutionfollowingallopolyploidspeciationincotton(Gossypium).Proc.Natl.Acad.Sci.USA92:280–284.

Zhao,J.W.,Udall,J.A.,Quijada,P.A.,Grau,C.R.,Meng,J.L.,andOsborn,T.C.(2006).QuantitativetraitlociforresistancetoSclerotiniasclerotiorumanditsassociationwithahomoeologousnon-reciprocaltranspositioninBrassicanapusL.Theor.Appl.Genet.112:509–516.

Genomic Changes in Resynthesized Brassica napus and Their Effect on Gene Expression and

Phenotype

Robert T. Gaeta, J. Chris Pires, Federico Iniguez-Luy, Enrique Leon and Thomas C. Osborn

PLANT CELL 2007;19;3403-3417; originally published online Nov 16, 2007;

DOI: 10.1105/tpc.107.054346

This information is current as of February 25, 2010

Supplemental Data References Permissions eTOCs

CiteTrack Alerts Subscription Information

http://www.plantcell.org/cgi/content/full/tpc.107.054346/DC1 This article cites 71 articles, 35 of which you can access for free at: http://www.plantcell.org/cgi/content/full/19/11/3403#BIBL

https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

Sign up for eTOCs for THE PLANT CELL at: http://www.plantcell.org/subscriptions/etoc.shtml Sign up for CiteTrack Alerts for Plant Cell at: http://www.plantcell.org/cgi/alerts/ctmain

Subscription information for The Plant Cell and Plant Physiology is available at: http://www.aspb.org/publications/subscriptions.cfm

© American Society of Plant Biologists

ADVANCING THE SCIENCE OF PLANT BIOLOGY