This item is only available as the following downloads:

Zheng et al 2013 PLoS ONE ( PDF )

PAGE 1

MediatorSubunit18ControlsFloweringTimeandFloral OrganIdentityinArabidopsisZhenguiZheng1,2* ,HexinGuan1¤,FranciscaLeal1,ParisH.Grey1,DavidG.Oppenheimer1*1 DepartmentofBiology,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmerica, 2 HowardHughesMedicalInstitute,DepartmentofMolecularGeneticsand Microbiology,UniversityofFlorida,Gainesville,Florida,UnitedStatesofAmericaAbstractMediatorisaconservedmulti-proteincomplexthatplaysanimportantroleinregulatingtranscriptionbymediating interactionsbetweentranscriptionalactivatorproteinsandRNApolymeraseII.MuchevidenceexiststhatMediatorplaysa constitutiveroleinthetranscriptionofallgenestranscribedbyRNApolymeraseII.However,evidenceismountingthat specificMediatorsubunitsmaycontrolthedevelopmentalregulationofspecificsubsetsofRNApolymeraseII-dependent genes.AlthoughtheMediatorcomplexhasbeenextensivelystudiedinyeastandmammals,onlyafewreportsonMediator functioninfloweringtimecontrolofplants,littleisknownaboutMediatorfunctioninfloralorganidentity.Hereweshow thatin Arabidopsisthaliana ,MEDIATORSUBUNIT18(MED18)affectsfloweringtimeandfloralorganformationthrough FLOWERINGLOCUSC ( FLC )and AGAMOUS ( AG ).A MED18 loss-of-functionmutantshowedaremarkablesyndromeoflater floweringandalteredfloralorgannumber.Weshowthat FLC and AG mRNAlevelsand AG expressionpatternsarealteredin themutant.Ourresultssupportparallelsbetweentheregulationof FLC and AG anddemonstrateadevelopmentalrolefor Mediatorinplants.Citation: ZhengZ,GuanH,LealF,GreyPH,OppenheimerDG(2013) MediatorSubunit18 ControlsFloweringTimeandFloralOrganIdentityin Arabidopsis .PLoS ONE8(1):e53924.doi:10.1371/journal.pone.0053924 Editor: XiaoyuZhang,UniversityofGeorgia,UnitedStatesofAmerica Received September10,2012; Accepted December4,2012; Published January11,2013 Copyright: 2013Zhengetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited. Funding: ThisworkwassupportedbyNationalScienceFoundationgrantDBI-0115684toDGO.PublicationofthisarticlewasfundedinpartbytheUniversityof FloridaOpen-AccessPublishingFund.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decisiontopublish,orpreparationofth emanuscript. CompetingInterests: Theauthorshavedeclaredthatnocompetinginterestsexist. *E-mail:patrickz@ufl.edu(ZZ);oppenhe@ufl.edu(DGO) ¤Currentaddress:AgriculturalCollege,GuangxiUniversity,Nannin,People’sRepublicofChinaIntroductionMediatorplaysanimportantroleinregulatingRNApolymeraseII(PolII)transcription.TheMediatorcomplexcontains22–28 subunits,andmediatesinteractionsbetweentranscriptionalcoactivatorsandPolII[1,2].Mediatorisevolutionarilyconserved andhasanancienteukaryoticorigin[3];itisfoundinorganisms fromfungitomammalsandplants,althoughtheevolutionary conservationofindividualsubunitsismoderate[4–6].Morethan 30differentsubunitshavebeendescribedthatarepartofthe Mediatorcomplexindifferentorganisms,butonlyabout20 subunitsarefoundinalleukaryotes[6–8].Theothersconsistof eitherspecies-specificsubunitsorotherancillarysubunitsassociatedwithactivationofspecificgenes.Med18isonesubunitofthe Mediatorcomplexandacomponentoftheheadmodulethatis involvedinstimulatingbasalRNAPolIItranscriptioninyeast [4,9]. Arabidopsis Mediatorwasfirstfoundtocontain27subunits,and mostofthemareconservedineukaryotes[5].Untilnow,allthe knownyeast/metazoanMediatorcomponentshavebeenidentifiedinplants[10].PHYTOCHROMEandFLOWERING TIME1(PFT1),nowknownasMED25,wasidentifiedasafactor ofaPhytochromeB(phyB)signalingpathwaythatpromotes floweringandcontrolsfinalorgansize[5,11–14].STRUWWELPETER(SWP)/MED14wasreportedtobeanuclearproteinthat playsaroleindefiningthedurationofcellproliferation[15].Some MediatorsubunitslikeMED25,MED8,MED16,andMED21act asintegratorsinresponsetoenvironmentalcuesin Arabidopsis [16– 21].MED18isasubunitoftheheadsubmoduleoftheplant Mediatorcomplex[10].Recently,Kimetal.(2011)reportedthat severalMediatorsubunitsincludingMediator18(MED18)are requiredformicroRNAbiogenesis[22].Littleisknownabout Mediatorfunctionduringfloralorganformationoritsroleinthe regulationoffloweringtime. Thetransitionfromvegetativegrowthtoreproductivedevelopmentin Arabidopsis isregulatedbymultiplefloralinduction pathways.Geneticstudiesofthetimingoffloweringin Arabidopsis haverevealed5majorpathways[23].Thephotoperiodand vernalizationpathwaysintegrateexternalsignalsintothefloral decision,theautonomousandgibberellin(GA)pathwaysact independentlyofenvironmentalcues,whereastheendogenous pathwayaddsplantagetothecontroloffloweringtime.The floweringpathwaysareinterconnectedandconvergeonafew floralintegrators,suchas FLOWERINGLOCUST ( FT )and SUPPRESSOROFOVEREXPRESSIONOFCONSTANS1 ( SOC1 or AGL20 )[24–26].Oneimportantregulatoroffloralinitiationisthe MADS-boxtranscriptionfactorFLOWERINGLOCUSC(FLC), whichactsasanegativeregulatoroffloweringinresponsetoboth endogenousandenvironmentalsignals;itisalsoanintegratorof theautonomousandvernalizationpathways[23,27,28].The autonomousandvernalizationpathwaysbothsuppressthe expressionof FLC [29],resultinginadecreasedexpressionof FLC andconsequentlyincreasedexpressionof FT and/or SOC1 in alaterdevelopmentalstageorafteraprolongedexposureofplants PLOSONE|www.plosone.org1January2013|Volume8|Issue1|e53924

PAGE 2

tolowtemperature[23,30].Theflorigen,FT,directlyregulates floralmeristemidentitygenessuchas APETALA1 ( AP1 )and initiatesfloralmorphogenesis[31,32]. Followingthevegetativetofloraltransition, Arabidopsis flowers developfourdifferentorgantypesthatarearrangedinconcentric whorls:thefirstwhorlcontainsfoursepals,thesecondwhorl containsfourpetals,sixstamensdevelopinthethirdwhorl,and twofusedcarpelsforminthefourthwhorl.Thecontroloffloral organidentityhasbeenintensivelystudiedinthepast25years. Analysisoffloralhomeoticmutantsledtotheproposalofasimple geneticmodel,explaininghowthreegroupsofregulatorygenes (classA,BandCgenes)aloneorbyinteractions,controltheorgan identityofthefourfloralwhorls[33,34].Theterminationofstem cellsinthefloralmeristemrequires AGAMOUS ( AG ),aMADSdomaintranscriptionfactor[35];AsaclassCfloralhomeotic gene, AG specifiesstamenidentitytogetherwiththeBclassand SEPALLATA ( SEP )genesandcarpelidentitytogetherwiththe SEP genes[34,36]. AG activates SPOROCYTELESS ( SPL ),which controlssporogenesisinbothstamensandcarpels[37].Clearly, AGisoneofthemostimportantregulatorsforthefloraltransition, floralorganidentity,andsporeformation. Inthisstudy,wedescribethefunctionof Arabidopsis Mediator subunit18(MED18),andshowthatitcontrolsfloweringtimeand floralorganidentitybytranscriptionalregulationof FLC and AG .MaterialsandMethods PlantMaterialsandGrowthConditionsAllthetransgenicandmutant Arabidopsis linesusedinthisstudy wereofecotypeColumbia(Col)exceptforthe ag-1 and pi-1 mutants,whichwereintheLandsberg erecta (L er )background. Plantsweregrowninatemperaturecontrolledgreenhouseat22– 24 u C,witharelativehumidityof60%–70%,anda16hourslight and8hoursdarkphotoperiod. Forshortdayexperiments,Plantsweregrowninagrowth chamberequippedwith40Wfluorescentlighttubeswitha9 hourslightand15hoursdarkphotoperiod.AnalysisofTranscriptsLevelsSemi-quantitativeRT-PCRwasusedtomeasurethetranscript levelsof MED18 andthefloralhomeoticgenesusing ACTIN4 (At5g59370)and GAPC (AT3G04120)asloadingcontrols.Total RNAsamplesweretreatedextensivelywithRNase-freeDNaseIto removeanycontaminatinggenomicDNA.First-strandcDNAwas synthesizedusing1mgoftotalRNAina20mlreactionvolume usingHighCapacitycDNAArchiveKitfromAppliedBiosystems (FosterCity,CA)accordingtothemanufacturer’sinstructions, followedbyphenol/chloroformpurification,andethanolprecipitation.ThecDNAweredissolvedin30mlTEbufferand1mlwas subjectedtoPCRina20mlreactionvolume.TheRT-PCRruns were20to30cycles,dependingonthelinearrangeofPCR amplificationforeachgene,withcycleparametersof94 u Cfor 0.5min,58 u Cfor0.5min,and72 u Cfor1minforeachcycle, withafinalincubationof72 u Cfor10min. AP1 AP2 AP3 PI ,and AG ,primersweredesignedaccordingtopublishedsequences[38]. AllotherprimersdesignedinthisstudywereinTableS3. QuantitativeRT-PCRwasmodifiedfromapreviously publishedmethod[39].Mutantandwildtypeseedlingorflowers weredissectedandpooled.TotalRNAwasextractedusing RNeasyplusmicrokit(Qiagen)andRNAquantity( 100ng/ml) andpurity(260/280 2.0,260/230 1.65)weredeterminedusing aNanodrop.RNAintegrity(RIN 8.5)and28S/18Sratio( 1.5) wasassessedusingaBioanalyzer2100(AgilentTechnologies).A quantityof500ngofhigh-qualityRNAforeachpooledsample wasconvertedintocDNAusingtheiScriptcDNASynthesisKit (Bio-Rad).GeneexpressionwasdeterminedusingtheCFX96TM RealTimesystem(Bio-Rad)with Act2/8 [40]asacontrol. FT SOC1 and FLC PrimersweredesignedaccordingtoZhouandNi [41].Theresultsweredeterminedusing DD Ctmethod[42],4 replicatesofpooledsampleswereusedforbothwildtypeand mutantseedlingsandflowers.InsituHybridizationWeusedpreviouslyestablishedmethodsfor insitu hybridization [43]withthefollowingmodifications.Togeneratetemplatesfor insitu probesynthesis,acDNAwasPCRamplifiedusingprimers thatcontainedthephageT7RNApolymeraseinitiationsequence. ThePCRproductwasusedfor invitro transcriptionofdigoxigenin-labeledprobesusingaDIG-RNAlabelingkit(Roche AppliedScience).DIG-labeledRNAprobeswerenothydrolyzed, andusedatafinalconcentrationof400ng/mlinthehybridizationsolution.Slideswerephotographedunderbrightfield illumination.StatisticalAnalysisAllgroupdifferencesinourdependentvariableswererevealed usingtwo-tailedStudent’sT-tests,and a -Levelsweresetat0.05.Results MED18 ControlsFloweringTimeandFloralOrgan IdentityThe MED18 (NP_565534;At2g22370)genewasoriginally identifiedthroughaphylogenomiccomparisonofsingle-copy genesconservedinangiosperms[44].Additionaldatabase searchesrevealedthatMED18wasaplanthomologofMediator subunit18(Fig.S1,ref[3,5]).Togainfurtherinsightintothe functionofthisgenein Arabidopsis ,weexaminedthephenotypesof plantshomozygousforT-DNAinsertionsinthe MED18 coding sequence.WeidentifiedT-DNAinsertionalleles(SAIL_889_C08, med18-1 ;SALK_027178, med18-2 )fromtheSALKT-DNA insertiondatabase(http://signal.salk.edu/cgi-bin/tdnaexpress) [45,46],andconfirmedthelocationoftheT-DNAinsertsby PCRusing MED18 andT-DNAspecificprimers(Fig.1A).We alsoexamined MED18 mRNAlevelsinplantshomozygousfor eachoftheallelesusingRT-PCR. MED18 mRNAwas undetectablein med18-1 plantscomparedtowildtype.Thisresult suggeststhat med18-1 isanullallele.Kimetal.[22]reportedweak expressionof MED18 mRNAin med18-2 plants,butwefailedto detectany MED18 mRNAusingprimerslocatedoneithersideof theT-DNAinsertion(Fig.S2). Wheneither med18-1 or med18-2 wascrossedwithwildtype plants,theF1showedawildtypephenotype,whichdemonstrates thatboth med18-1 and med18-2 arerecessive.Acomplementation testwasalsoperformedusingthe med18-1 and med18-2 alleles.The F1plantsfromacrossof med18-1andmed18-2 plantsshoweda med18 phenotype(TableS1),whichisdescribedbelow.Thisresult demonstratesthat med18-1 and med18-2 areallelic. Inadditiontothepreviouslydescribedphenotype[22], med18 mutationscauseasyndromeofrelatedphenotypesaffecting floweringtime,inflorescencestructure,andflowermorphology. Underlongdayconditions(16hourlight,8hourdark),both med18-1 and med18-2 mutantsdidnotflowerevenafter35days aftergermination(DAG)(Fig.1C).Incontrast,wildtypeplants floweredapproximately14DAG(12leaves)(Fig.1B).The med18-1 mutationalsoaffectsflowermorphology.Inwildtypeflowers,the sepalsfullyenclosethedevelopingflowersuntilshortlyafterthe beginningofanthesis(Fig.1D).Thesepalsof med18-1 flowersdidMediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org2January2013|Volume8|Issue1|e53924

PAGE 3

notfullyenclosethedevelopingflowerssuchthatthebuds appearedprematurelyopen(Fig.1E). Moststrikingofallwasthat med18-1 mutantsshowed dramaticallyalteredfloralorgannumbersascomparedto wildtype.Allfloralorganswereaffected(TableS1,Fig.1F–H), morethan40and80percentoftheflowerson med18-1 plants hadabnormalsepalsandpetalsnumberrespectively,while 80%oftheflowershadfewerthan6stamens.Approximately Figure1.PhenotypesofMediatorsubunit18(med18)mutants. (A)SchematicdiagramoftheMED18geneshowingthelocationsofT-DNA insertionsinthe med18 mutants.Blackrectanglesrepresentexons,linesrepresentintrons,andtrianglesrepresentT-DNAinsertions;the med18-1 mutationcorrespondstoinsertionlineSAIL_889_C08,whereasthe med18-2 mutationcorrespondstoinsertionline,SALK_027178.(B,C)35dayold wildtype(B), med18-1 (C)plants.(D,E)Inflorescenceofwildtype(D),and med18-1 (E)plants.(F–H) Arabidopsis wildtype(F), med18-1 (G),and dissected med18-1 (H)flower.(I,J)Transversesectionofwildtype(I),and med18-1 (J)carpels.(K,L)Anthersofwildtype(K)and med18-1 (L)(stained withKI/I2)attimeofflowering.Scalebars:1mminD,E,F,GandH;100mminI,J,KandL. doi:10.1371/journal.pone.0053924.g001 MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org3January2013|Volume8|Issue1|e53924

PAGE 4

25%of med18-1 mutantflowershadmorethantwocarpels (TableS1,Fig.1I,J).Inadditiontothealteredfloralorgan number, med18-1 mutantsalsoshoweddelayedstamendevelopmentandlatermaturationofpollen(Fig.1K,L),whichled toreduceseedset. Overexpressionof MED18 causedincreasedcarpelandstamen numbers(Fig.S3A),andreducedpetalnumbers(Fig.S3B), Carpel-likesepalswerealsoobservedin MED18 overexpressing plants(Fig.S3C),andthese MED18 overexpressingplants floweredearlierthanwildtype(datanotshown).MED18 Expressionduring Arabidopsis DevelopmentThefloweringtimeandfloralorganpatterningdefectsinthe med18 mutantssuggeststhat MED18 playsaroleinregulating genesimportantforfloweringandfloralorgandevelopment.To determineif MED18 expressioncoincideswiththephenotype,we examinedtheRNAexpressionpatternof MED18 inwildtype plantsusing insitu hybridization. MED18 transcriptscouldbe detectedindevelopingseeds(Fig.2A),theinflorescencemeristem, thefloralmeristem,andfloralorganprimordia(Fig.2BandC). Transcriptabundanceappearedhighestinthedevelopingstamens andpistils(Fig.2D).Inlaterstages, MED18 expressionwas abundantindevelopingovulesandpollen(Fig.2E)andin addition, MED18 mRNAwasweaklyexpressedinpetals,sepals, andthewallsofcarpelsuptostage12(Fig.2E).Thestrong expressionof MED18 intheinflorescencemeristemandfloral organsisconsistentwiththefloraldefectsobservedin med18-1 mutants.med18 MutantsAffectFloweringTimeinBothLongand ShortDaysInlongdaycondition(16hlight,8hdark),themutantplants didnotfloweruntilmorethan40leaves;andinshortday(9h light,15hdark),after95days(morethan60leaves), med18 mutantsstilldidnotflower,andmanyrosetteandcaulineleaves showedsenescence(Fig.3A).Comparedtowildtypeplants (12leavesunderlongdaysandabout40leavesundershortdays), med18 islatefloweringunderbothlongandshortdays(Fig.3A).med18 MutantsAffectFloweringTimethroughupregulationof FLC ExpressionTounderstandhowMED18regulatesfloweringtime,we comparedtheexpressionlevelsofkeygenesofdifferentpathways inmutantandwildtypeseedlingsusingRT-PCR. FT and FLC showexpressiondifferencesinthe med18 mutantsascomparedto wildtypeplants(Fig.S4).WeperformedQPCRandshowedthat FT mRNAwasdownregulatedapproximately4.7fold(Fig.3B, p # 0.004),whilethe FLC transcriptlevelwasmuchhigher (9.5fold)in med18-1 thaninthewildtypeplants(Fig.3B, Figure2.ExpressionpatternsofMED18. ( A–D )Longitudinalsectionsofwildtypeshow MED18 mRNAdetectedin(A)developingseed,(B) inflorescencemeristem,floralmeristemandsepals,(C,D)petal,stamen,andcarpelprimordia.( E )Transversesectionofawildtypeflowershows MED18 mRNAexpressioninallfloralorgans.( F ) MED18 mRNAsensestrandcontrolshowingnonon-specifichybridization.NumbersinB–DandF indicateflowerstages;IMinBandFindicatesinflorescencemeristem.Scalebars:100mminB–Fand25mminA. doi:10.1371/journal.pone.0053924.g002 MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org4January2013|Volume8|Issue1|e53924

PAGE 5

p # 0.001).For FT and SOC1 ,whicharedownstreamof FLC ,only FT mRNAwassignificantlysuppressedbythelossof MED18 function,whereas SOC1 mRNAlevelsshownosignificant difference(Fig.3B,p=0.106).med18 MutantsareResponsivetoVernalizationbutnot GATreatmentToexaminetheresponsivenessof med18 tovernalization,the med18-1 plantswereplantedandkeptat4 u Cfor4weeksinthe darkandthentransferredtonormalgrowthtemperature(23 u C). Thevernalization-treated med18 plantsfloweredmuchearlierthan theuntreatedplants,afterproducingabout14leaves,whichis comparabletothatofwildtypeplants(Fig.3C.Theuntreated plantsfloweredafterproducingcloseto40leaves,whichsuggests thatthefloweringofthemutantplantswasdecreasedby vernalization(Fig.3C).ToexaminetheeffectsofGAtreatment, aGAsolutionof20mMwassprayedtwiceaweekafter germinationuntilflowering.TheresultsshowthatGAhasno obviouseffectonthefloweringof med18 mutantsascomparedto wildtypeplants.TheGA-treated med18-1 plantsinitiatedflowering afterproducingmorethan35leaves,whichisnotsignificantly differentthanuntreatedmutantplants(p=0.115,Fig.3D).Our resultsareconsistentwithpreviousresultsthatshowthat vernalizationpromotesfloweringbyrepressing FLC andreleasing FT fromrepression[23,47].Theresponsivenessof med18-1 to vernalizationsuggeststhat MED18 regulatesfloweringtime throughthevernalizationpathway.MED18 AffectsFloralOrganFormationinallFourWhorlsBothmutationandoverexpressionof MED18 alteredthe numberoffloralorgans(Fig.1G–I,TableS1,Fig.S3).The increasednumberofpetalsandfewerthannormalstamensseenin the med18 mutantswasreminiscentofthefloralphenotypeof agamous ( ag )mutants[48,49].Tofurtherexplorethefloral developmentalpathwayinwhich MED18 isinvolved,we constructeddoublemutantswiththewell-studiedfloraldevelopmentalregulators, ag pistillata ( pi )and apetala2 ( ap2 ).Flowersonthe med18-1ag-1 doublemutantshowedthestrikingreiterationof sepalsandpetalscharacteristicof ag mutants(Fig.4A–C,Table S2).The med18-1pi-1 doublemutantflowershadabnormal carpels,butfewersepalsthanthe pi-1 singlemutant(Fig.4D,E, TableS2).Theflowersonthe med18-1ap2-5 doublemutant displayedamuchmorecomplexphenotype.Carpelsinboththe firstandfourthwhorlswerepresentasseeninthe ap2-5 single mutant.Inaddition,doublemutantflowersexhibitedpetaloid stamensandotherpetaloidstructuresinwhorls2and3(Fig.4F, G,TableS2).MED18 AffectsFloralOrganFormationthrough RegulationofFloralHomeoticGeneExpressionTheclearepistasisobservedinthe med18-1ag-1 doublemutants suggestedthat MED18 and AG functioninthesamepathwayto controlfloraldevelopment.Therefore,weexamined AG mRNA levelsaswellasthelevelsofmRNAforotherkeyfloralregulators in med18 mutantsusingsemi-quantitativeRT-PCR.Ourresults showedthatthemRNAexpressionlevelsof AG AP1 and PI were Figure3.MED18controlsfloweringtime. ( A ) MED18 controlsfloweringtimeunderbothlongdayandshortdayconditions.Arrowindicatesthat floweringwasnotachievedwhentheexperimentwasterminated.( B )Transcriptslevelsof FT and FLC arealteredin med18 mutants.( C,D )Flowering timeinvernalization(C)andgibberellin(D)treatment.Vm:vernalization,GA:gibberellin.Scalebarsindicatemean 6 s.e.;**p # 0.01. doi:10.1371/journal.pone.0053924.g003 MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org5January2013|Volume8|Issue1|e53924

PAGE 6

downregulatedin med18-1 mutantswhile AP2 and AP3 mRNA showednoobviouschange(Fig.S5).QPCRresultsrevealed AG mRNAwasdownregulatedupto4.8fold(Fig.5A,p # 0.007), AP1 and PI mRNAexpressionlevelswerealsoreducedinthe med18-1 mutant1.9(p # 0.014)and2.4fold(p # 0.011)respectively,while AP2 and AP3 transcriptlevelsshowednosignificantdifferences between med18-1 andthewildtype(Fig.5A,p=0.068and0.082). Thedecreased AG expressionislikelytocauseastamentopetal transitionaswellasabnormalcarpeldevelopment[49],as observedin med18-1andMED18 overexpressionflowers(Fig.1F– H,Fig.S3).Theseresultssupportthehypothesisthat MED18 controlsfloralorganidentitythroughitsregulationoffloral homeoticgeneexpression.Altered AG ExpressionPatternsareObservedin med18 MutantsTofurtherexaminetheexpressionof AG in med18-1 mutants, in situ hybridizationusingan AG antisenseRNAprobewas performedontissuesectionsfromwildtype(Fig.5B,D,F,H) and med18-1 plants(Fig.5C,E,G,I).Asdescribedpreviously [50,51], AG isnotexpressedintheinflorescencemeristem,norin stage1orstage2floralmeristemsinwildtypeplants.Strong AG expressionisfirstfoundinthecenterofstage3andstage4 wildtypeflowers,butnotintheemergingsepalprimordia(Fig.5B). Duringlaterstagesofwildtypeflowerdevelopment, AG expression ispresentinstamensandcarpels(Fig.5D,F,H).In med18-1 mutants,apatternof AG expressionsimilartothatseeninwildtype wasobservedinboththeinflorescencemeristemandinstage1 andstage2floralmeristems(Fig.5C).Inflowersfromstage4to stage7,weak AG expressionwasdetectedinthecenterofthe carpelprimordia,butstrong AG expressionwasobservedin stamenprimordia,althoughnoexpressionwasobservedinpetal andsepalprimordia(Fig.5C,E,G).Inlaterdevelopmentstages (stage11,Fig.5H,I), AG expressionwasobservedindeveloping pollensimilartothatobservedinwildtypeflowers(Fig.5H),butit wasdifficulttodetectanysignalinvascularbundlesofstamens andcarpels(Fig.5I).Thisresult,togetherwiththeresultsofthe QPCRanalysis,stronglysuggeststhat MED18 isrequiredto maintainthenormal AG expressionlevelandpatternduringearly stamenandgynoeciumdevelopment.DiscussionUndermanyconditionsMediatorappearstofunctionasa generaltranscriptionfactor[52].Nonetheless,expressionprofiling ofyeastMediatorsubunitmutantshasrevealedthedirect regulationofspecificsetsofgenesbyMediator[53],andanalysis ofMediatorin Arabidopsis hasshownthattheMediatorsubunits areimportantinregulatingspecificdevelopmentalprocesseslike earlyembryopatterning[54,55],cellnumberandorgansize [12,15],floweringtimecontrol[5,14,56]environmentalregulation anddefensegeneregulation[16,18–21,57]. MED18wasfirstcharacterizedasageneraltranscriptionfactor thatpromotesPolIItranscriptionthroughpromotionofthe transcriptionofmiRNA,andknockingdown MED18 expression causedabnormalcotyledonandsiliquedevelopmentaswellasa laterfloweringphenotype[22].However,themechanismsby whichMED18regulatesfloweringarepoorlyunderstood,and nothingisknownabouthowMED18regulatesfloralorgan identity.Inthisstudy,weidentifiedMED18asaregulatorofboth floweringtimeandfloralorganidentity.Ourfindingsshowthat MED18controlsfloweringtimebyup-regulating FLC expression, whichalsoaffectstheexpressionofthedownstreamgene, FT Figure4.Doublemutantofmed18-1andfloralhomeoticgenes. ( A ) ag-1 flower.( B ) med18-1ag-1 doublemutantflower.( C )dissectionof med18-1ag-1 doublemutantflower.( D ) pi-1 flower.( E ) med18-1pi-1 doublemutantflower.( F ) ap2-5 flower.( G ) med18-1ap2-5 doublemutant flower.Scalebars:1mm. doi:10.1371/journal.pone.0053924.g004 MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org6January2013|Volume8|Issue1|e53924

PAGE 7

Afterthefloweringtransition,MED18playsaroleinfloralorgan identitybyregulatingthe AG expressionlevelandpatternaswell as AP1 expressionlevels(Fig.5,6).Theseresultssuggestthat MED18isimportantintheintegrationofkeysignalingpathways inplantsbycontrollingtargetgenestranscription(Fig.6). The med18 mutantsdisplayvariousdevelopmentaldefects, suggestingthatMED18regulatesmultiplepathways.Thisisin agreementwiththefunctionofanimalandyeastMediatorproteins, whichhavebeensuggestedtoregulatebothbasicandspecific transcription[8,58,59].WehavefocusedonastudyofMED18 functionduringfloweringandfloralorgandevelopment.Wefound that MED18 mRNAislocalizedmainlytotheprecursorcellsof inflorescencemeristem,thefourfloralwhorlsandstronglyinthe pollenandovuleprimordia.ThissuggestsaroleforMED18inthe controloffloralandreproductiveorganinitiationanddevelopment. Floralprimordiaarisefromtheinflorescencemeristem,andfloral organsarethenformedinthefloralmeristem.TheABCmodelof flowerdevelopmentexplainshowthreeclassesofgenescontrol sepal,petal,stamen,andcarpelidentity[33].Furthermore,the modelindicatesthatclassAandCgenesaremutuallyantagonistic [60].Thepreviouslyisolatedfloralhomeoticgenesallseemtocode forpotentialtranscriptionfactors.AP1,AP3,PI,AG,andtheSEP proteinscontaintheMADSdomainknowntobindtoDNA[36]. AP2containsanotherDNAbindingdomain,theAP2domain[61]. GeneralregulatorslikeMediatorarelikelytoregulatethesegenes. Ourstudieshaveshownthatmutationsin MED18 causedown regulationofseveralhomeoticgenes,suchas AP1 PI and AG ,but notallofthem;thisfindingsuggeststhattheeffectofMED18on homeoticgeneexpressionisgenespecific. BecauseMED18affectstheexpressionofmultiplegenes,the flowersof med18 mutantsshowvariablephenotypes,suchas4–14 petals,2–6stamens,0–6sepalsand1–3carpels.Inoneofthe flowersthathad14petals,fourstamensareclearlyvisible(Fig.S6). Therefore,theeffectofthe med18 mutationcannotbeasimple homeotictransformationofthefloralorgansintopetals.Itis interestingtonotethatthe med18 mutationcausesadown Figure5.MED18regulatesfloralhomeoticgeneexpression. (A) Relativetranscriptlevelsofsixfloralhomeoticgenesdeterminedby real-timeRT-PCRinwildtypeand med18-1 mutants(**p # 0.01, *p # 0.05).(B,D,F,H) AG expressionpatternasdeterminedby insitu hybridizationinwildtypeflowers.(C,E,G,I)The AG expressionpattern in med18-1 mutantflowers.NumbersinB–Ishowflowerstages,IMinB andCindicatesinflorescencemeristem.Arrowsin(H)and(I)indicate expressioninvascularbundlesofstamensandcarpels,in(F)and(G) showingdevelopingpedals.ScalebarsinB–I:50mm. doi:10.1371/journal.pone.0053924.g005 Figure6.ModelofMED18controlfloweringtimeandfloral organidentity. GrowthsignalsaretransmittedtotheMediator complexbydirectactionorthroughHEN3onMED18toregulatethe transcriptionoftargetgenes.BothfloweringtimeintegratorFLCand floralorganidentityorganizerAGtranscriptslevelsaredeterminedby MED18.Inaddition,MED18affectsthetranscriptionofothertarget genes,whicharefloweringtimeandfloralorganidentityregulators, suchas FT AP1 and PI .Theseregulatorsworktogetherandwithother factorstocontrolfloweringandfloralorganformation. doi:10.1371/journal.pone.0053924.g006 MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org7January2013|Volume8|Issue1|e53924

PAGE 8

regulationof AG expression,yetsomeflowershavethreecarpels. AccordingtotheABCmodel,downregulationofanAclassgene (suchasAP1)willcauseectopiccarpelformation. Thelaterfloweringphenotypeobservedin med18 mutantsis causedbyup-regulationof FLC mRNAexpression.FLCalso containsaMADSdomain,andinterestingly,both FLC and AG expressionlevelsarealteredin med18 mutants,butinopposite directions.TherearesimilaritiesbetweentheDNAstructureof FLC and AG .Mostofthegenesinthe Arabidopsisthaliana genome haveshort( 1kb)introns[62],butintron1in FLC andintron2 in AG arebothgreaterthan3.0kbinlength,andbothofthese intronshavecis-elementsthatareimportantfortranscriptional control[27,63,64]. HUA1 HUA2 HUAENHANCER2 [ HEN2 ] and HEN4 wereshowntopositivelyregulate AG expression,either byinhibitingprematurepolyadenylationwithin AG intron2orby promotingthesplicingofthisintron,and hua1hua2 double mutantshavereducedlevelsof FLC mRNA[65].Alltheseresults andourdatathat MED18 regulates FLC and AG mRNA expressionlevels,stronglysupporttheideaofparallelregulation of FLC and AG [66]. Theresultofup-regulationof FLC anddown-regulationof AG suggestthat MED18 isnotageneraltranscriptionrepressor,but ratheritplaysdifferentrolesdependingontheidentityofitstarget, and/orthedevelopmentalstage.Themechanismbywhich MED18affects FLC and AG expressioniscurrentlyunknown. Highlyspecificgeneregulationisthoughttobedeterminedby activatorsandcombinatorialuseofcofactors.Inyeast,Med18acts downstreamofCDK8,andmayactasadirectprocessorof signalingpathwaysfordeterminingspecificgeneexpression[53]. Med18wasalsoreportedtoberequiredforproperterminationof transcriptionofasubsetofgenesduringyeastbudding[67].In Arabidopsis ,theCDK8homologis HUAENHANCER3 [ HEN3 ], whichalsocontrolsorganidentityandshowsimilarloss-of-Cfunctionphenotypes(WangandChen,2004),suggestingthat HEN3 mayregulateorganidentitythrough MED18in Arabidopsis (Fig.6).Ourresultssuggestaconservedmechanismmayexistin yeastandplants.SupportingInformationFigureS1Phylogenetictreeofeukaryotic MEDIATOR SUBUNIT18 ( MED18 ). TheBayesianinferenceanalysiswas derivedfrom406aminoacidpositionsofMED18indifferent speciesofEukaryotes.Thebestmodelaminoacidreplacementfor MED18sequenceswasJTTandgammamodelforsubstitution rateheterogeneitybetweensites.Bayesianphylogeneticinference wasperformedwithMrBayesVersion3.0usingfourchainsand 2,000,000generations.Numbersatnodeindicateposterior probabilities;scalebarshows0.2aminoacidsubstitutionspersite. (TIF)FigureS2RT-PCRresultof MED18 showingthat MED18 mRNAwasnotdetectedin med18-1 med18-2 and med18-1/med18-2 T1plantsusingprimersdesigned forthePCRproductincludingbothinsertionsites. The GAPC genewasusedasacontrol. (TIF)FigureS3Phenotypeofoverexpressionof MED18 in transgenicplantstransformedwitha 35S::MED18 construct. (TIF)FigureS4RT-PCRresultsofselectedfloweringtime regulatorsinwildtypeand med18-1 seedlings. In med18-1 plants, FLC isup-regulated, FT isdown-regulated,butothersshow noobviousdifference. (TIF)FigureS5RT-PCRresultsofselectedfloralorgan identitygenesinwildtypeand med18-1 flowers. In med18-1 plants, AP2 PI and AG aredown-regulated,butothers shownoobviousdifference. (TIF)FigureS6A med18-1 mutantflowerthatshows14petals and4stamens. (TIF)TableS1Numberofdifferentorgantypesin med18 flowers.aa200flowerswerecounted.bmeanof200lowers.c Organnumberis6ormore.dOrgannumberis4orless.eOrgan numberis1or1.5(somecarpelsonlyhavehalforonethirdof normalsizelocatedatthetipofpistil).fOrgannumberisbetween 2and3(samease).Mutantplantsweregrowninthegreenhouse (16hlight,23 6 2 u C),wildtypeplantsweregrownunderthesame conditions.Thefloralorganson200wildtypeplantswerealso counted,andallflowersshowed4sepals,4petals,6stamensand2 fusedcarpels,except2flowersshowed5petals.The med18-1 allele isastrongalleleandallfourfloralorgansshowsignificant differencesfromwildtype(**p 0.01), med18-2 isaweakerallele, andonlypetalsandstamensshowsignificantdifferencesfrom wildtype.TheF1plantsfromacrossof med18-1 with med18-2 ( med18-1 6 med18-2 )alsoshowobviousfloralorgannumber changes. (DOCX)TableS2Numberofdifferentorgantypesinfloral homeoticmutantsanddoublemutantswith med18-1 flowers.aa,atleast15flowerswerecountedforeachmutant.b, thefirstwhorlofeachorganshowedthenormalorgannumberfor thatwhorl.c,manyflowersshowed1connectedcarpel-likesepal. (DOCX)TableS3Primersdesignedinthisstudy. (DOCX)AcknowledgmentsTheauthorsgratefullyacknowledgetheSALKInstituteGenomicAnalysis LaboratoryandSyngentaforsupplyingtheT-DNAinsertionalleles.AuthorContributionsConceivedanddesignedtheexperiments:ZZDGO.Performedthe experiments:ZZHGPG.Analyzedthedata:ZZFLDGO.Wrotethe paper:ZZDGO.References1.BlazekE,MittlerG,MeisterernstM(2005)ThemediatorofRNApolymerase II.Chromosoma113:399–408. 2.CasamassimiA,NapoliC(2007)Mediatorcomplexesandeukaryotic transcriptionregulation:anoverview.Biochimie89:1439–1446. 3.BourbonHM(2008)Comparativegenomicssupportsadeepevolutionaryorigin forthelarge,four-moduletranscriptionalmediatorcomplex.NucleicAcidsRes 36:3993–4008. 4.KornbergRD(2005)Mediatorandthemechanismoftranscriptionalactivation. TrendsBiochemSci30:235–239. 5.BackstromS,ElfvingN,NilssonR,WingsleG,BjorklundS(2007)Purification ofaplantmediatorfromArabidopsisthalianaidentifiesPFT1astheMed25 subunit.MolCell26:717–729.MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org8January2013|Volume8|Issue1|e53924

PAGE 9

6.BoubeM,JouliaL,CribbsDL,BourbonHM(2002)Evidenceforamediatorof RNApolymeraseIItranscriptionalregulationconservedfromyeasttoman.Cell 110:143–151. 7.BourbonHM,AguileraA,AnsariAZ,AsturiasFJ,BerkAJ,etal.(2004)A unifiednomenclatureforproteinsubunitsofmediatorcomplexeslinking transcriptionalregulatorstoRNApolymeraseII.MolCell14:553–557. 8.ConawayRC,SatoS,Tomomori-SatoC,YaoT,ConawayJW(2005)The mammalianMediatorcomplexanditsroleintranscriptionalregulation.Trends BiochemSci30:250–255. 9.LariviereL,GeigerS,HoeppnerS,RotherS,StrasserK,etal.(2006)Structure andTBPbindingoftheMediatorheadsubcomplexMed8-Med18-Med20. Naturestructural&molecularbiology13:895–901. 10.MathurS,VyasS,KapoorS,TyagiAK(2011)TheMediatorcomplexinplants: structure,phylogeny,andexpressionprofilingofrepresentativegenesinadicot (Arabidopsis)andamonocot(rice)duringreproductionandabioticstress.Plant physiology157:1609–1627. 11.CerdanPD,ChoryJ(2003)Regulationoffloweringtimebylightquality.Nature 423:881–885. 12.XuR,LiY(2011)ControloffinalorgansizebyMediatorcomplexsubunit25in Arabidopsisthaliana.Development138:4545–4554. 13.OuB,YinKQ,LiuSN,YangY,GuT,etal.(2011)Ahigh-throughput screeningsystemforArabidopsistranscriptionfactorsanditsapplicationto Med25-dependenttranscriptionalregulation.MolPlant4:546–555. 14.InigoS,AlvarezMJ,StrasserB,CalifanoA,CerdanPD(2012)PFT1,the MED25subunitoftheplantMediatorcomplex,promotesfloweringthrough CONSTANSdependentandindependentmechanismsinArabidopsis.PlantJ 69:601–612. 15.AutranD,JonakC,BelcramK,BeemsterGT,KronenbergerJ,etal.(2002)Cell numbersandleafdevelopmentinArabidopsis:afunctionalanalysisofthe STRUWWELPETERgene.EMBOJ21:6036–6049. 16.KiddBN,EdgarCI,KumarKK,AitkenEA,SchenkPM,etal.(2009)The mediatorcomplexsubunitPFT1isakeyregulatorofjasmonate-dependent defenseinArabidopsis.PlantCell21:2237–2252. 17.DhawanR,LuoH,FoersterAM,AbuqamarS,DuHN,etal.(2009)HISTONE MONOUBIQUITINATION1interactswithasubunitofthemediatorcomplex andregulatesdefenseagainstnecrotrophicfungalpathogensinArabidopsis.The Plantcell21:1000–1019. 18.ElfvingN,DavoineC,BenllochR,BlombergJ,BrannstromK,etal.(2011)The ArabidopsisthalianaMed25mediatorsubunitintegratesenvironmentalcuesto controlplantdevelopment.ProcNatlAcadSciUSA108:8245–8250. 19.ChenR,JiangH,LiL,ZhaiQ,QiL,etal.(2012)TheArabidopsismediator subunitMED25differentiallyregulatesjasmonateandabscisicacidsignaling throughinteractingwiththeMYC2andABI5transcriptionfactors.PlantCell 24:2898–2916. 20.CevikV,KiddBN,ZhangP,HillC,KiddleS,etal.(2012)MEDIATOR25acts asanintegrativehubfortheregulationofjasmonate-responsivegeneexpression inArabidopsis.PlantPhysiol160:541–555. 21.WathugalaDL,HemsleyPA,MoffatCS,CremelieP,KnightMR,etal.(2012) TheMediatorsubunitSFR6/MED16controlsdefencegeneexpression mediatedbysalicylicacidandjasmonateresponsivepathways.NewPhytol 195:217–230. 22.KimYJ,ZhengB,YuY,WonSY,MoB,etal.(2011)TheroleofMediatorin smallandlongnoncodingRNAproductioninArabidopsisthaliana.EMBOJ 30:814–822. 23.SrikanthA,SchmidM(2011)Regulationoffloweringtime:allroadsleadto Rome.Cellularandmolecularlifesciences:CMLS68:2013–2037. 24.ChouML,YangCH(1999)Late-floweringgenesinteractwithearly-flowering genestoregulatefloweringtimeinArabidopsisthaliana.PlantCellPhysiol40: 702–708. 25.LeeH,SuhSS,ParkE,ChoE,AhnJH,etal.(2000)TheAGAMOUS-LIKE20 MADSdomainproteinintegratesfloralinductivepathwaysinArabidopsis. GenesDev14:2366–2376. 26.KardailskyI,ShuklaVK,AhnJH,DagenaisN,ChristensenSK,etal.(1999) ActivationtaggingofthefloralinducerFT.Science286:1962–1965. 27.HeY,MichaelsSD,AmasinoRM(2003)Regulationoffloweringtimeby histoneacetylationinArabidopsis.Science302:1751–1754. 28.BastowR,MylneJS,ListerC,LippmanZ,MartienssenRA,etal.(2004) VernalizationrequiresepigeneticsilencingofFLCbyhistonemethylation. Nature427:164–167. 29.SheldonCC,RouseDT,FinneganEJ,PeacockWJ,DennisES(2000)The molecularbasisofvernalization:thecentralroleofFLOWERINGLOCUSC (FLC).ProcNatlAcadSciUSA97:3753–3758. 30.HelliwellCA,WoodCC,RobertsonM,JamesPeacockW,DennisES(2006) TheArabidopsisFLCproteininteractsdirectlyinvivowithSOC1andFT chromatinandispartofahigh-molecular-weightproteincomplex.PlantJ46: 183–192. 31.CorbesierL,VincentC,JangS,FornaraF,FanQ,etal.(2007)FTprotein movementcontributestolong-distancesignalinginfloralinductionof Arabidopsis.Science316:1030–1033. 32.AbeM,KobayashiY,YamamotoS,DaimonY,YamaguchiA,etal.(2005)FD, abZIPproteinmediatingsignalsfromthefloralpathwayintegratorFTatthe shootapex.Science309:1052–1056. 33.CoenES,MeyerowitzEM(1991)Thewarofthewhorls:geneticinteractions controllingflowerdevelopment.Nature353:31–37. 34.MeyerowitzEM,BowmanJL,BrockmanLL,DrewsGN,JackT,etal.(1991)A geneticandmolecularmodelforflowerdevelopmentinArabidopsisthaliana. DevSuppl1:157–167. 35.LenhardM,BohnertA,JurgensG,LauxT(2001)Terminationofstemcell maintenanceinArabidopsisfloralmeristemsbyinteractionsbetween WUSCHELandAGAMOUS.Cell105:805–814. 36.PelazS,DittaGS,BaumannE,WismanE,YanofskyMF(2000)BandCfloral organidentityfunctionsrequireSEPALLATAMADS-boxgenes.Nature405: 200–203. 37.ItoT,WellmerF,YuH,DasP,ItoN,etal.(2004)Thehomeoticprotein AGAMOUScontrolsmicrosporogenesisbyregulationofSPOROCYTELESS. Nature430:356–360. 38.YuH,ItoT,ZhaoY,PengJ,KumarP,etal.(2004)Floralhomeoticgenesare targetsofgibberellinsignalinginflowerdevelopment.ProcNatlAcadSciUSA 101:7827–7832. 39.SeifertAW,ZhengZ,OrmerodBK,CohnMJ(2010)Sonichedgehogcontrols growthofexternalgenitaliabyregulatingcellcyclekinetics.Nature communications1:23. 40.NiW,XieD,HobbieL,FengB,ZhaoD,etal.(2004)Regulationofflower developmentinArabidopsisbySCFcomplexes.Plantphysiology134:1574– 1585.41.ZhouY,NiM(2009)SHB1playsdualrolesinphotoperiodicandautonomous flowering.Developmentalbiology331:50–57. 42.SuzukiE,ZhaoY,ItoS,SawatsubashiS,MurataT,etal.(2009)AberrantE2F activationbypolyglutamineexpansionofandrogenreceptorinSBMA neurotoxicity.ProcNatlAcadSciUSA106:3818–3822. 43.ParkSO,ZhengZ,OppenheimerDG,HauserBA(2005)ThePRETTYFEW SEEDS2geneencodesanArabidopsishomeodomainproteinthatregulates ovuledevelopment.Development132:841–849. 44.Leebens-MackJH(2006)DevelopmentalGeneticsoftheFlowerDouglasES, DouglasE,Leebens-Mack,J.H.,Soltis,P.S.&Callow,J.A,editor:Academic Press. 45.SessionsA,BurkeE,PrestingG,AuxG,McElverJ,etal.(2002)AhighthroughputArabidopsisreversegeneticssystem.PlantCell14:2985–2994. 46.AlonsoJM,StepanovaAN,LeisseTJ,KimCJ,ChenH,etal.(2003)GenomewideinsertionalmutagenesisofArabidopsisthaliana.Science301:653–657. 47.MichaelsSD,AmasinoRM(2001)LossofFLOWERINGLOCUSCactivity eliminatesthelate-floweringphenotypeofFRIGIDAandautonomouspathway mutationsbutnotresponsivenesstovernalization.PlantCell13:935–941. 48.BowmanJL,SmythDR,MeyerowitzEM(1989)Genesdirectingflower developmentinArabidopsis.PlantCell1:37–52. 49.ChuangCF,MeyerowitzEM(2000)Specificandheritablegeneticinterference bydouble-strandedRNAinArabidopsisthaliana.ProcNatlAcadSciUSA97: 4985–4990. 50.DrewsGN,BowmanJL,MeyerowitzEM(1991)Negativeregulationofthe ArabidopsishomeoticgeneAGAMOUSbytheAPETALA2product.Cell65: 991–1002. 51.BowmanJL,DrewsGN,MeyerowitzEM(1991)ExpressionoftheArabidopsis floralhomeoticgeneAGAMOUSisrestrictedtospecificcelltypeslateinflower development.PlantCell3:749–758. 52.TakagiY,KornbergRD(2006)Mediatorasageneraltranscriptionfactor.JBiol Chem281:80–89. 53.vandePeppelJ,KettelarijN,vanBakelH,KockelkornTT,vanLeenenD,etal. (2005)Mediatorexpressionprofilingepistasisrevealsasignaltransduction pathwaywithantagonisticsubmodulesandhighlyspecificdownstreamtargets. MolCell19:511–522. 54.ItoJ,SonoT,TasakaM,FurutaniM(2011)MACCHI-BOU2IsRequiredfor EarlyEmbryoPatterningandCotyledonOrganogenesisinArabidopsis.Plant CellPhysiol52:539–552. 55.GillmorCS,ParkMY,SmithMR,PepitoneR,KerstetterRA,etal.(2010)The MED12-MED13moduleofMediatorregulatesthetimingofembryopatterning inArabidopsis.Development137:113–122. 56.GonzalezD,BowenAJ,CarrollTS,ConlanRS(2007)Thetranscription corepressorLEUNIGinteractswiththehistonedeacetylaseHDA19and mediatorcomponentsMED14(SWP)andCDK8(HEN3)torepress transcription.MolCellBiol27:5306–5315. 57.KloseC,BucheC,FernandezAP,SchaferE,ZwickE,etal.(2012)The mediatorcomplexsubunitPFT1interfereswithCOP1andHY5inthe regulationofArabidopsislightsignaling.PlantPhysiol160:289–307. 58.BjorklundS,GustafssonCM(2005)TheyeastMediatorcomplexandits regulation.TrendsBiochemSci30:240–244. 59.ConawayRC,ConawayJW(2011)FunctionandregulationoftheMediator complex.CurrOpinGenetDev21:225–230. 60.BowmanJL,SmythDR,MeyerowitzEM(1991)Geneticinteractionsamong floralhomeoticgenesofArabidopsis.Development112:1–20. 61.OkamuroJK,CasterB,VillarroelR,VanMontaguM,JofukuKD(1997)The AP2domainofAPETALA2definesalargenewfamilyofDNAbindingproteins inArabidopsis.ProcNatlAcadSciUSA94:7076–7081. 62.YuJ,YangZ,KibukawaM,PaddockM,PasseyDA,etal.(2002)Minimal intronsarenot‘‘junk’’.GenomeRes12:1185–1189. 63.SheldonCC,ConnAB,DennisES,PeacockWJ(2002)Differentregulatory regionsarerequiredforthevernalization-inducedrepressionofFLOWERING LOCUSCandfortheepigeneticmaintenanceofrepression.PlantCell14: 2527–2537.MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org9January2013|Volume8|Issue1|e53924

PAGE 10

64.HongRL,HamaguchiL,BuschMA,WeigelD(2003)Regulatoryelementsof thefloralhomeoticgeneAGAMOUSidentifiedbyphylogeneticfootprinting andshadowing.PlantCell15:1296–1309. 65.ChengY,KatoN,WangW,LiJ,ChenX(2003)TwoRNAbindingproteins, HEN4andHUA1,actintheprocessingofAGAMOUSpre-mRNAin Arabidopsisthaliana.DevCell4:53–66. 66.SimpsonGG(2004)Theautonomouspathway:epigeneticandpost-transcriptionalgeneregulationinthecontrolofArabidopsisfloweringtime.CurrOpin PlantBiol7:570–574. 67.MukundanB,AnsariA(2011)NovelroleformediatorcomplexsubunitSrb5/ Med18interminationoftranscription.TheJournalofbiologicalchemistry286: 37053–37057.MediatorSubunit18 ControlsFlowerDevelopment PLOSONE|www.plosone.org10January2013|Volume8|Issue1|e53924