The ABI3 gene is modulated by__ farnesylation and is involved in auxin signaling and lateral

TheABSCISICACIDINSENSITIVE3(ABI3)geneismodulatedbyfarnesylationandisinvolvedinauxinsignalingandlateralrootdevelopmentinArabidopsisSiobhanMaryBradyz,SaraF.Sarkarz,DarioBonettayandPeterMcCourtÃDepartmentofBotany,UniversityofToronto,25WillcocksSt,Toronto,CanadaM5S3B2Received29July2002;revised16December2002;accepted7January2003.ÃForcorrespondence(fax󰀄14169785878;e-mailmccourt@botany.utoronto.ca).yPresentaddress:DepartmentofPlantBiology,CarnegieInstitutionofWashington,StanfordUniversity,260PanamaStreet,Stanford,CA94305,USA.zEachoftheseauthorscontributedequallytothisstudy.SummaryGeneticscreenshaveidenti®edanumberofgenesthatregulateabscisicacid(ABA)responsivenessinArabidopsis.Usingacombinationofsuppressorscreensanddoublemutantanalysis,wehavedeterminedageneticrelationshipforanumberoftheseABAresponseloci.BasedongerminationinthepresenceofexogenousABA,theABI1andABI2phosphatasesactatorupstreamoftheERA1farnesyltransferaseandtheABI3andABI5transcriptionfactorsactatordownstreamofERA1.IncontrastwithABI3andABI5,theABI4transcriptionfactorappearstoactatorupstreamofERA1.Basedonreportergeneconstructs,theupstreamregulationofABI3byERA1occursatleastpartiallyattheleveloftranscription,suggestingthatthislipidmodi®cationisrequiredtoattenuateABI3expression.SimilarexperimentsalsoindicatethatABI3isauxininducibleinlateralrootprimordia.Relatedtothis,loss-of-functionabi3allelesshowreducedlateralrootresponsivenessinthepresenceofauxinandanauxintransportinhibitor,andera1mutantshaveincreasednumbersoflateralroots.Theseresultssuggestthepossibilitythatgenesidenti®edthroughABAresponsivegerminationscreenssuchasERA1andABI3havefunctionsinauxinactioninArabidopsis.Keywords:hormones,mutants,signalingcross-talk,rootdevelopment.IntroductionAbscisicacid(ABA)isahormonethatregulatesmanycomplexplantprocessesincludingtheinductionandmain-tenanceofseeddormancy,inhibitionofrootgrowth,con-trolofstomatalclosure,andprotectionoftheplantfromavarietyofenvironmentalstresses(Davies,1995).AlthoughthestructureandbiosynthesisofABAislargelyunder-stood,howABAregulatesthesemultipleplantresponsesisjustnowbeginningtoberevealedthroughgeneticana-lysisinArabidopsisthaliana.Todate,geneticscreensforABA-hypersensitivemutantshaveindicatedthatprocessesincludingfarnesylation(era1),inositolsignaling(fry1),andRNAmetabolism(abh1,sad1,hyl1)arerequiredtoattenuatetheABAsignal(forreview,seeFinkelsteinetal.,2002).AdditionalscreensformutationsthatconferareducedsensitivitytoABAhaveidenti®edhomologoustype2Cphosphatases(ABI1,ABI2)andthreedisparatetranscriptionfactors(ABI3,ABI4andABI5).Constructionofdifferentcombinationsofoverexpressingandloss-of-functionallelesofABI3±5indicatesthattheoverallABAß2003BlackwellPublishingLtd

responsivenessofthematureseedresultsfromacomplexinteractionbetweenthesetranscriptionfactors(Sodermanetal.,2000).Although®rstdescribedasseedspeci®cforalteredABAsensitivity,recentlymoredetailedphenotypicanalysisandexpressionstudiesofABI3,ABI4andABI5indicatethatthesegenesmayhavefunctionsoutsideofmodulatingABAseedsensitivity.Forexample,ABI4andABI5geneshavebeenshowntohavefunctionsinbothsugarandsaltresponsesandinearlyseedlinggrowthaftergermination(Gibson,2000;Lopez-Molinaetal.,2001;Quesadaetal.,2000).Bothgenesareexpressedoutsideofseeddevelop-ment,andloss-of-functionmutationsineachresultinal-teredlateralrootbranchinginresponsetonitrate(Signoraetal.,2001).InthecaseofABI3,itappearsthatthisgeneplaysrolesinplastiddevelopmentandbuddormancyandinteractswithgenesinvolvedinlightregulation(Rohdeetal.,2000,2002).ABI3expressionisdetectedinmanyquiescenttissuesincludingthereceptacleof¯owers,the67

68SiobhanMaryBradyetal.

enhancedABAsensitivity(era)ifitcannotgerminateon0.3mMABA.Toidentifysuppressormutationsofera1,mutagenizedM2seedsderivedfrom20000M1era1plantswerescreenedforgerminationon0.3mMABA.Ofthe124linesthatre-testedintheM3generationforsuppressionofera1,eightwereinsensitiveto3mMABA.BecausetheseeightlinesshowedstrongABAinsensitivity,theycouldrepresentnewallelesofknownabimutations.Partialalle-lismtestswereperformedwithaknownabi3allele(abi3-1),andthreesuppressorlineswereclassi®edasnewallelesofabi3(datanotshown).OneofthesealleleswassequencedandshowntocontainamissensemutationintheB3domainofABI3(glycinetoarginineatposition666).Tocorroboratesuppressioninteractionbetweenera1andabi3,F2plantsfromacrossofera1-2andaseverealleleabi3-6werescreenedforprotrudingcarpels,acommonphenotypeofhomozygousera1-2¯owers(Bonettaetal.,2000).Anumberoftheseplantssegregatedthecharacter-istichomozygousabi3-6underdevelopedgreenseedphe-notype.Plantsproducinggreenseedandprotrudingcarpelswerepropagatedandmolecularlygenotypedfortheera1-2andabi3-6mutation,andfromthese,50imma-turegreenseedswereplacedon0.3and3.0mMABA,respectively.AllseedsgerminatedontheseconcentrationsofABA,verifyingthatabi3-dependentABAinsensitivitywasepistatictoera1(Table1).Theotherera1-2suppressorlineswererecessiveforABAinsensitivityandnotallelictoeitherabi4orabi5(datanotshown).Thelackofanynewabi4orabi5allelesfromthesuppressorscreenwasnotsurprisingbecausethescreenwasnotsaturating.Hence,doublemutantsbetweenera1andknownallelesofabi4andabi5wereconstructedtotestepistasis.Among182F2seedsfromanera1-2byabi5-1cross,55germinatedon3mMABA,whichyieldedachar-acteristicratioforrecessiveepistasisforABAinsensitivityTable1ABAresponsivenessofsingleanddoublemutantseedsasmeasuredbygerminationonABAGenotypeMcolLerWSabi1-1abi2-1abi3-6abi4-1abi5-1era1-2abi1-1era1-2abi2-1era1-2abi3-6era1-2abi4-1era1-2abi5-1era1-2Minimalmedia170/172148/149103/104136/136112/112133/133127/127111/111120/125172/177204/212ND122/122274/2740.3mMABA163/167140/140134/134122/122125/125136/136159/163130/1301/1440/11/19950/501/1112/1153.0mMABA0/1540/1350/186171/171162/162130/133139/139167/1700/2800/1940/24550/50281/293262/270axilsofpedicelsandaxillary¯owerbracts,theabscissionzoneofsiliquesandrosetteleaves,andstipules(Parcyetal.,1994;Rohdeetal.,1999).Althoughthereasonsforthecomplexexpressionpat-ternsofABI3arenotclear,themaizeorthologofthisgene,VIVIPAROUS1(VP1),hasbeenshowntohavemultipleregulatorydomains(Hoeckeretal.,1995)suggestingthatthisgenehasmultiplefunctions.ComparisonsofABI3/VP1orthologshaveidenti®edfourhighlyconservedaminoaciddomains:A1,adomainintheacidicN-terminaloftheprotein;andthreebasicregions:B1±3(Giraudatetal.,1992).TheB1domainhasbeenshowntobecapableofprotein±proteininteractionswithABI5,whileB2andB3canbindDNA(Nakamuraetal.,2001;Suzukietal.,1997).TheB3DNA-bindingdomainisuniqueamongplantsandisconservedamongseveralgenefamilies(Ulmasovetal.,1997).Oneoftheseistheauxinresponsefactor(ARF)family.MutationsinanumberofARFgeneshavebeenshowntoalterauxinresponsesinArabidopsis(forreview,seeLiscumandReed,2002).Interestingly,inthepresenceofexogenousABA,plantsmisexpressingVP1areinsensi-tivetoauxin-inducedlateralrootformationsuggestingthattheVP1/ABI3transcriptionfactormayde®neaninteractionnodebetweenABAandauxinsignaling(Suzukietal.,2001).TofurtherunderstandtherolesofABI3andERA1inplanthormonesignaling,wehaveusedacombinationofsup-pressorandepistasisstudiesbetweenABAresponsemutantsinArabidopsistode®neageneticframework.AtthelevelofseedABAresponsiveness,ABI1andABI2actatorupstreamofERA1.Bycontrast,ABI3andABI5actatordownstreamofERA1.ABI4appearstoactatorupstreamofERA1andmayde®neaseparategeneticpathwayfromthatofABI3andABI5.ThesegeneticstudiessuggestthatERA1isanegativeregulatorofABI3andABI5,andthisissup-portedbytheincreasedexpressionofanABI3::GUStrans-geneinanera1mutantbackground.ThesestudiesalsosuggestthatABI3transcriptionmayberegulatedbyauxin,andthatauxinsignalinginlateralrootdevelopmentmayhaveanABI3-dependentcomponent.WediscussamodelforageneralroleofABI3inseedandrootdevelopment,andthepossibilitythatABA-sensitizingscreensusingseedgerminationasanassaymayalsoenrichforauxinresponsemutantsinArabidopsis.ResultsGeneticinteractionsbetweengenesaffectingseedABAresponsivenessUnderourassayconditions,wild-typeArabidopsisgermi-nationispermittedonconcentrationsofABAlowerthan1.0mM.AmutantlineisconsideredABA-insensitive(abi)comparedtowildtypeifitgerminateson3mMABAandhasNumbersinthenumeratorrepresentgerminatedseedoverthetotalnumberofseedsplated.ND:notdetermined.ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

ABI3signalintegrationinArabidopsis69

(w2󰀅2.;P>0.90).Theera1abi5doublemutantswereestablishedandtestedforgerminationon0.3and3mMABA(Table1).Thesedoublemutantswereinsensi-tivetobothconcentrationsindicatingthattheabi5muta-tionisepistatictotheera1mutationontheseconcentrationsofABA.Bycontrast,ofthe126F2seedsgeneratedfromanabi4-1byera1-2cross,only15oftheexpected31germinatedon3mMABA,suggestingthatabi4isnotepistatictotheera1mutation.Furthermore,35of87F2seedsdidnotgerminateon0.3mMABAsuggestingthatera1isepistatictoabi4.Thiswasveri®edbytheidenti®ca-tionoftheera1abi4doublemutantasbeingsensitiveto0.3and3mMABA(Table1).Takentogether,theseresultssuggestthatABI3andABI5areatorgeneticallydown-streamofERA1andthatABI4isatorgeneticallyupstreamofERA1.Therecessivenatureofera1,ABI3,ABI4andABI5muta-tionsmadetheidenti®cationofdoublemutantsbetweenthesemutationsrelativelyeasy.However,theidenti®cationofera1abi1-1andera1abi2-1doublemutantswasmorecomplicatedbecauseofthedominantnatureoftheabi1andabi2mutations(Leungetal.,1997).F2progenyfromacrossbetweenabi1-1andera1-2wereplatedon3mMABAand121ofthe187seedsgerminated.Thesenumbers®ta9:7ratio(w2󰀅1.;P>0.95)suggestingthatera1-2isepistatictoabi1-1.Tocon®rmthis,individualF3progenyweremolecularlyidenti®edthatwerehomozygousforboththeabi1-1andera1-2mutations.Thesedoublemutantsweresensitiveto0.3and3mMABAasmeasuredbyger-mination,indicatingthattheera1-2mutationisepistatictoabi1-1(Table1).ThecloselinkageofABI2andERA1genes(15mapunits,chromosome5)meantthatasimplegeneticratiointheF2progenywouldnotbeexpected.Consistentwiththis,theexpectedratioof9:7insensitivetosensitiveon3mMABAwasnotobserved(321:172,w2󰀅16.6).WethereforescreenedF2seedsforsensitivityto0.3mMABAtoenrichforera1-2homozygousmutants.Fromtheseseeds,31plantsweregenotypedforthemolecularpolymorphismthatidenti®edtheabi2-1mutation,and10plantswereidenti®edasheterozygousfortheabi2-1mutation.FouroftheseplantswereallowedtoselfandF2progenyweremolecularlygenotypedforabi2-1homozygousplants.AllprogenyfromthesedoublemutantlinesshowedABAsensitivityto0.3and3mMABA,indicatingthattheera1mutationisepistatictoabi2-1mutation(Table1).Together,theseresultsshowthatERA1actsatordownstreamofABI1andABI2.ERA1negativelyregulatesABI3transcriptioninvegetativetissuesLoss-of-functionmutationsinERA1conferanenhancedABAsensitivity,suggestingthatfarnesylationisanegativeregulatorofABAsensitivityinArabidopsis.AsshownßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

above,ABI3andABI5maygeneticallyinteractdownstreamofERA1,thus,farnesylationmayregulateseedABAsensi-tivitybymodulatingthesegeneproducts.Totestthis,theera1mutationwascrossedintoatransgenicplantcontain-ingtheABI3::GUSreporterconstruct,andb-glucuronidaseactivitywasmonitoredbyhistochemicalstaining(Parcyetal.,1994).Underourstainingconditions,era1mutanttissuesshowedincreasedGUSstainingcomparedtowildtype(Figure1).Unexpectedly,therootsofera1mutantsalsoshowedapunctuatestainingpatternthatwasnotobservedinthewildtypemacroscopically(Figure1).Usingdifferentialinterferencecontrastmicroscopy(DIC),how-ever,itwaspossibletodiscernweakGUSstaininginwildtypeinapatternthatcorrespondstospeci®cpericyclecellsoftherootthatareundergoingdifferentiationintolateralrootprimordia(Figure2).Lateralrootsdevelopwhenadifferentiatedpericyclecelladjacenttoxylembeginsanasymmetrictransversedivisionleadingtotheformationofaconvexrootprimordium.Thelateralrootprimordiumcontinuestogrowthroughtheoverlyingcelllayerstoeventuallyemergefromtheparentroot.Afteremergence,thelateralrootcontinuestogrowfromanautonomousfullyfunctionalmeristem(Laskowskietal.,1995;MalamyandBenfey,1997).Inwild-typeroots,weakGUSstainingis®rstvisualizedjustafterthepericyclecellhascommittedtobecomingalateralrootprimordiumcellatapproximatelytwo-cellstagewhenthelayerismidwaythroughtheendo-dermis.Thispatternoccursuptoapproximatelythe®ve-cell-layerstageatwhichtimetheexteriorborderofexpres-sionappearsasacrescentwithstainingattheouterborderoftheprimordiumanddecreasedstainingtowardsthemiddle(Figure2).Bythetimethelateralrootemergesfromtheepidermallayer,theGUSstainingpatterndissipatesandisnolongervisiblebyDIC.Therefore,itappearsthatinwild-typeplants,ERA1maydownregulatetranscriptionofABI3intheembryo,apicalmeristemandlateralroots,whichisconsistentwiththenegativeregulationbyERA1onABI3identi®edinthesuppressionstudies.LateralrootdefectsinERA1andABI3loss-of-functionmutationsTheaberrantABI3::GUStransgeneexpressionpatternsobservedinera1mutantrootssuggestthatABI3mayhavefunctionsinlateralrootdevelopment.However,thenum-beroflateralrootsobservedinabi3-6loss-of-functionmutantswassimilartothatseeninwild-typegrownonminimalmedia(Figure3a).Thisisincontrasttoera1-2,whichdoesshowanincreasednumberoflateralrootsversuswildtype(Figure3a).Thelackofalateralrootphenotypeinabi3mutantsversusera1maysimplymeanthattheroleofABI3onlateralrootdevelopmentisatbestsubtleundernormalgrowthconditions.We,therefore,decidedtotesttheresponseofabi3-6rootgrowthinthe70SiobhanMaryBradyetal.

Figure1.Effectsoftheera1mutationontheexpressionofABI3::GUStransgeneatdifferentstagesofdevelopment.PlantsofdifferentdevelopmentalageswerestainedtodetermineGUSactivityfor12hat378C.Toppanel:6-day-oldseedlings.Onanaverage,6-day-oldwild-typeseedingsshowed5507Æ308(SD)nmolmgÀ1proteinhÀ1,andera1seedlingsofthesameageshowed72Æ557(SD)nmolmgÀ1proteinhÀ1.Middlepanel:12-day-oldseedlings.Bottompanel:rootsof12-day-oldseedling.presenceofindoleacidicacid(IAA)andtheauxintransportinhibitorN-1-naphthylphthalamicacid(NPA)toseeifanyphenotypescouldbeobservedunderconditionsthatper-turbedauxin-regulatedlateralrootdevelopment.Exogen-ousapplicationofauxinincreaseslateralrootinitiation,whereasinhibitionofauxintransportinhibitslateralrootinitiation.Underthesegrowthconditions,bothgenotypesrespondedtoIAAbyinitiatingnewlateralroots.However,abi3-6plantsrequiredhigherconcentrationsofIAAtoinitiatelateralrootnumberscomparabletowildtype(Figure3b).Similarly,applicationoftheauxintransportinhibitorNPAinhibitedlateralrootinitiationinbothwildtypeandinabi3-6,butagainmutantrootswerelessrespon-sivetotheNPA-inducedrepressionoflateralrootinitiation(Figure3c).Theabi3-6mutantrootsstillretainedthecharacteristicagravitropicresponsetoNPA,indicatingthatFigure2.ABI3::GUSreportergeneexpressioninwild-typelateralrootprimordia.(a)Two-cellstage.(b)Five-cellstage.(c)Midwaythroughtheepidermis.(d)Afteremergencefromtheprimaryroot.Allmagni®cationswere40Â.theinhibitorwasbeingperceivedbythemutantroots(datanotshown).Thus,itappearsthatABI3functionisrequiredatsomelevelforcorrectauxinresponsivenessinlateralrootsinArabidopsis.ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

ABI3signalintegrationinArabidopsis71

thedomainofGUSaccumulationwasexpandedinhor-mone-treatedrootswiththevasculartissueandtheendo-dermisalongwiththelateralrootprimordiashowingincreasedbluestaining(Figure4).Inlongerexposuresupto48h,muchoftheABI3::GUSexpressionpatterncorre-spondedtoaseriesoflateralrootsinitiatedadjacenttoeachotherinresponsetothehormonetreatment(Fig-ure4).Thesepatternsofexpressionwerealsoobservedforrootsexposedto1mMIAA,buttheintensityoftheGUSstainingwasreducedsubstantiallyandamuchlongerexposuretoIAAwasrequiredtoseesimilarresults(datanotshown).TheABI3transgenereporterisalsoresponsivetocon-centrationsofABAaslowas0.4mMinthelateralroot(Figure4).However,incontrasttotheauxininductionkinetics,1-dayincubationswererequiredtoseeaclearGUSprecipitateinthelateralroottissuemacroscopically.Furthermore,althoughABAapplicationincreasedGUSexpressionovertime,unlike2,4DapplicationABAapplica-tiondidnotleadtoanyincreaseinthedomainofGUSstainingoutsidetheregionoflateralrootdevelopment(Figure4).Inthisstudy,ABAdidnotinduceABI3::GUSexpressioninanyothervegetativetissue,whichisconsis-tentwithpreviousreportsusingthispromoterfusiontrans-genicplant(Parcyetal.,1994).AuxinandABAactsynergisticallytoinhibitgerminationAphysiologicalprocesssuchasgerminationispositivelyandnegativelyin¯uencedbymultipleenvironmentalandhormonalsignals.TheobservationthatABI3::GUSexpres-sioncanbeinducedbyauxinandthatloss-of-functionmutationsinABI3showauxin-relatedphenotypesinlateralrootdevelopmentsuggeststhatABI3maybeinvolvedinmultiplehormonalsignalingpathways.Thisbringsintoquestiontheplethoraofphenotypesobservedinsevereloss-of-functionabi3mutants.Forexample,althoughloss-of-functionmutantsinabi3andABAauxotrophsarebothnon-dormant,onlyabi3mutantsshowreduceddesiccationtolerance,lossofseedstoragereserves,andtheinabilitytocompletelateembryogenesis(Nambaraetal.,1992;Oomsetal.,1993).Possibly,theadditionalphenotypesseeninabi3mutantscouldbeauxinrelatedandeventhedecreasedABAsensitivityofabi3allelesmaybebecauseofachangeinauxinsensitivity.Totestthepossibilitythatauxincanin¯uenceseedABAresponsivenessinArabidopsis,wild-typeseedswereexposedtoincreasingconcentrationsofABAinthepresenceoflowconcentrationsofexogenousIAA.Asexpected,increasingconcentrationsofABApro-gressivelyinhibitedgermination,butthepresenceofauxinshiftedtheABAsensitivitysothatlessABAwasrequiredtoinhibitgermination(Figure5).Auxinonlyaffectsseedger-minationinthepresenceofABAanddoesnotdirectlyinhibitseedgermination.Figure3.Lateralrootsofwild-typeera1-2andabi3-6plantlets.(a)Numberoflateralrootsemergedfromwildtype,era1-2andabi3-6grownonminimalmediafor2weeks.(b)Three-day-oldseedlingsofwildtype(black)andabi3-6(open)wereexposedtoIAAfor10days.LateralrootswerecountedusingDIC.(c)Three-day-oldseedlingsofwildtype(black)andabi3-6(open)exposedtoNPAfor10days.LateralrootswerecountedusingDIC.Experimentsweredoneintriplicate,and15rootswerecountedforeachexperiment.ErrorbarsrepresentSD.ABI3::GUSexpressionisresponsivetobothauxinandABAapplicationintherootTofurtherexplorethisrelationshipofABI3toauxinandABA,theexpressionoftheABI3::GUStransgenewasmon-itoredinresponsetoexogenousconcentrationsofIAA,thesyntheticauxin2,4-dichlorophenoxyaceticacid(2,4D)andABA.GUSprecipitatewasvisiblewithin1hof2,4Dappli-cation,andwasatamaximumafter1dayofexposure(Figure4).Unlikewild-typerootsgrownonminimalmedia,ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

72SiobhanMaryBradyetal.

Figure4.ABI3geneexpressioninwild-typelateralrootsinthepresenceof2,4DandABA.Fourteen-day-oldplantswereexposedfor1,24and48handthenstainedforGUSactivityfor12hat378C.Thetopsixpanelsare10Âmagni-®cationsandthebottomthreepanels(48h)are40Âmagni®cations.Figure5.Germinationofwild-typeseedsonincreasingconcentrationsofexogenousABAinthepresenceandabsenceof2mMIAA.(~)Minimalmedium,(*)2mMIAA.ExperimentswereperformedintriplicateanderrorbarsrepresentSD.DiscussionGeneticinteractionsbetweenABAresponsemutationsIfmutationsexistthatconferoppositesignalingstatesinaparticularprocess,doublemutantscanoftenbeusedtoordergenesintoageneticpathway(McCourt,1999).Byconstructingdoublemutantscontainingera1-enhancedABAsensitivitywithavarietyofABA-insensitiveabimuta-tions,wehavebeenabletoorderanumberofABA-respon-sivelociintoageneticframework(Figure6).ThisframeworksuggeststhatERA1geneticallyactsatordown-streamoftheABI1/2phosphatasesbutfunctionsatorupstreamoftheABI3andABI5transcriptionfactors.ThegeneticrelationshipbetweentheABI1/2andERA1muta-tionsareconsistentwiththeorderingofthesegenesusingFigure6.GeneticinteractionsbetweenABA-responsemutantsbasedondoublemutantepistasis.Arrowsrepresentpositiveregulationandbarsrepresentnegativeregula-tion.TheplacementofABI4atthesamelevelasERA1andactinginconcertwithABI3andABI5isbasedoninteractionsreportedinotherstudies(Sodermanetal.,2000).EpistaticinteractionsbetweenthesegenescouldalsoplaceABI4geneticallyaboveERA1inthispathway.Intheabovepathway,theABI1/2toERA1arrowisgivenapositivesignbasedonthebeliefthatABI1isanegativeregulatorofthepathway(ParcyandGiraudat,1997).However,trueloss-of-functionalleleshavenotbeenidenti®edinthisgeneandifABI1/2arepositiveregulators,thesignofthisinteractionwouldbenegative.stomatalconductanceandpatchclampanalysis(Peietal.,1998).TheassignmentofABI3andABI5inthesameplaceinthisgeneticpathwayre¯ectsrecentworkshowingthatthesetwotranscriptionfactorsinteractinayeasttwo-hybridassay(Nakamuraetal.,2001).Furthermore,thefactthatERA1isupstreamofABI3asobservedintheepistaticstudiesissupportedbytheobservationthatlossofERA1functioncausesincreasedtranscriptionoftheABI3::GUSreporterconstruct.Interestingly,theothertranscriptionfactorABI4thathasbeenimplicatedinseedABAsignalingßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

ABI3signalintegrationinArabidopsis73

throughgeneticscreens,appearstoactatorupstreamofERA1.Thegeneticrelationshipsseenherearerelativelysimplecomparedtootherstudiesinvolvinginteractionswiththesegenes(Finkelstein,1994;Sodermanetal.,2000).Thismayre¯ectthatweonlyfollowedoneresponseoutput,seedABAresponsiveness,whereasotherstudieshavealsoincludedotherphysiologicalresponsesandexpressionofmultiplegenes.Secondly,whilethisstudyinvolvedinter-actionsbetweenmutations,oftenotherstudiesinvolveusinglinesthataremisexpressingoroverexpressinggenesofinterestorconstructingdoublemutantswithsimilarphenotypes.Assignmentoforderusingover-ormisexpres-sioncanbedeceptivebecauseofthearti®cialityofthesystem(Ptashne,1988).Forexample,squelching,whichisthetitrationofonetranscriptionfactorbytheoverexpres-sionofapartnerprotein,oftenactsinadominantnegativemanneronthenormalfunctionofthetranscriptionfactorandcanoftencauseproteininteractionsthatdonotnor-mallyoccurinvivo.Constructionofdoublemutantsusingmutationsthatconfersimilarphenotypesrequiresthateachmutationisanull,arequirementthatisoftendif®culttoattaininhigherplants(AveryandWasserman,1992).ABI3andauxininteractionsinlateralrootinitiationandgerminationArelationshipbetweenlateralrootinitiationandABI3hasalsobeenuncoveredinstudiesinvolvingmisexpressionoftheABI3orthologVP1inArabidopsis(Suzukietal.,2001).ThatectopicexpressionofVP1suppressesauxin-inducedlateralrootformationarguesthatthisgeneproductmayde®neaninteractionnodebetweenmultiplehormonesig-nalssuchasABAandauxin.WhilebothABAandauxinhavebeenimplicatedincontrollingoverallrootlength,itisauxinthatappearstohaveadirecteffectonlateralrootinitiation(Boerjanetal.,1995;Reedetal.,1998).Thereducedresponseofabi3loss-of-functionmutantstoIAAorNPAasmeasuredbylateralrootinitiationindicatesthatthepresenceofABI3isrequiredforcorrectauxinsignalinginthelateralroot.Moreover,theinductionofABI3by2,4DapplicationpositionsABI3asbothagenethatin¯uencesauxinrootresponsesandagenethatrespondstoauxin.ThefunctionofABI3inauxinsignalingisatthistimeunclearbuttheparallelsofABI3tootherauxinresponsegenesareintriguing.TheB3-bindingdomainofABI3isconservedinmanyauxinresponse(ARF)genesandARFproteinsarethoughttointeractwithIAAgeneproductstoelicitacorrectauxinresponse(Tiwarietal.,2001;Ulmasovetal.,1997).Potentially,theABI3transcriptionfactorcouldinteractwithotherARForIAAproteinstomediateauxinresponses.AlthoughABI3was®rstdescribedasageneinvolvedinABAsignaling,abi3loss-of-functionmutantsshowaddi-ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

tionalphenotypescomparedtoABAauxotrophs(Nambaraetal.,1995;Oomsetal.,1993).Forexample,severeabi3allelesaredesiccationintolerant,failtocompleteseedmaturationandcannotaccumulatesomestoragereserves,allphenotypesnotdetectedinmutantsdefectiveinABAbiosynthesis.Ifabi3mutantshavereducedauxinrespon-siveness,thismayofferanexplanationforthenon-ABA-relatedphenotypesobservedintheembryo.ThefactthatexogenousapplicationofIAAtowild-typeArabidopsisseedsincreasesABAsensitivityasmeasuredbygermina-tionsuggeststhatchangesinauxinsynthesisoractioncantranslateintoalteredseedABAresponsiveness.Furthermore,asERA1isanegativeregulatorofABI3tran-scription,thehypersensitivityobservedinera1loss-of-functionallelesmaysimplybetheresultofincreasedproductionofABI3,whichcouldinturnincreaseauxinactiontherebyincreasingseedABAsensitivity.Indeed,theresponsecurveofwild-typeseedstoexogenousABAinthepresenceofauxinisrelativelysimilartotheincreasedsensitivityofera1mutantseedstoABA(Cutleretal.,1996).Hence,theincreasedlateralrootinitiationobservedinera1maysimplyre¯ecttheincreasedauxinsensitivityofthismutant.Onthisnote,otherphenotypesobservedinera1mutantssuchasincreasedapicalmeristemsize,semisteri-lityanddefectivelateralshootinitiationmayalsoberelatedtoalteredauxinaction(Bonettaetal.,2000;Runningetal.,1998).ThelackofafarnesylationsiteonABI3clearlypredictsthatafarnesylatedintermediatemustexistbetweenERA1andABI3.Interestingly,genomicanalysisindicatesanum-berofpotentialfarnesylationtargetgenesinvolvedinauxinresponsesandanumberoftranscriptionfactors(NambaraandMcCourt,1999).Theuseofgeneticenhancerandsuppressorscreensoftheera1hypersensitivephenotypesshoulduncovertheseintermediates.Finally,theseresultsfurthersuggestthattheuseofgerminationasanassayforABAresponsivenessingeneticscreensenrichesformuta-tionsthatalterotherhormoneresponsepathways.Todate,mutationsinGA,ethyleneandbrassinosteroidsynthesisoractionhavealsobeenshowntoalterseedresponsivenesstoexogenousABA(Ghassemianetal.,2000;Steberetal.,1998;SteberandMcCourt,2001).Thuscautionshouldbeexercisedinclassifyingmutationsintospeci®chormoneresponsepathwaysbasedsimplyontheirgerminationresponse.ExperimentalproceduresPlantmaterialandgrowthconditionsGerminationef®ciencyisdependentonseedageandthecondi-tionsinwhichtheseedsmaturedonthematernalplants.We®ndthatwild-typeandmutantstrainsgivemoreconsistentresultsiftheseedsareyoungerthan6monthspost-harvesting,andonly74SiobhanMaryBradyetal.

ReferencesAvery,L.andWasserman,S.(1992)Orderinggenefunction:theinterpretationofepistasisinregulatoryhierarchies.TrendsGenet.8,12±16.Bonetta,D.,Bayliss,P.,Sun,S.,Sage,T.andMcCourt,P.(2000)FarnesylationisinvolvedinmeristemorganizationinArabidop-sis.Planta,211,182±190.Boerjan,W.,Cervera,M.T.,Delarue,M.,Beeckman,T.,Dewitte,W.,Bellini,C.,Caboche,M.,VanOnckelen,H.,VanMontagu,M.andInze,D.(1995)Superroot,arecessivemutationinArabidopsis,confersauxinoverproduction.PlantCell,7,1405±1419.Cutler,S.,Ghassemian,M.,Bonetta,D.,Cooney,S.andMcCourt,P.(1996)AproteinfarnesyltransferaseinvolvedinABAtrans-ductioninArabidopsis.Science,273,1239±1241.Davies,P.J.(1995)Theplanthormones:theirnature,occurrenceandfunctions.InPlanthormones;physiology,biochemistryandmolecularbiology.(Davies,P.J.,ed.).Dordrecht:KluwerPress.Finkelstein,R.R.(1994)MutationsattwonewArabidopsisABAresponselociaresimilartotheabi3mutations.PlantJ.5,765±771.Finkelstein,R.R.,Wang,M.L.,Lynch,T.J.,Rao,S.andGoodman,H.M.(1998)TheArabidopsisabscisicacidresponselocusABI4encodesanAPETALA2domainprotein.PlantCell,10,1043±1054.Finkelstein,R.R.,Gampala,S.S.L.andRock,C.D.(2002)Abscisicacidsignalinginseedsandseedlings.PlantCell,S15±S45.Giraudat,J.,Hauge,B.M.,Valon,C.,Smalle,J.,Parcy,F.andGoodman,H.M.(1992)IsolationoftheArabidopsisABI3genebypositionalcloning.PlantCell,4,1251±1261.Ghassemian,M.,Nambara,E.,Cutler,S.,Kawaide,H.,Kamiya,Y.andMcCourt,P.(2000)RegulationofabscisicacidsignalingbytheethyleneresponsepathwayinArabidopsis.PlantCell,12,1117±1126.Gibson,S.I.(2000)Plantsugar-responsepathways.Partofacom-plexregulatoryweb.PlantPhysiol.124,1532±1539.Hoecker,U.,Vasil,I.K.andMcCarty,D.R.(1995)IntegratedcontrolofseedmaturationandgerminationprogramsbyactivatorandrepressorfunctionsofViviparous-1ofmaize.GenesDev.9,2459±2469.Jefferson,R.A.,Kavanagh,T.A.andBevan,M.W.(1987)GUSfusion:b-glucoronidaseasasensitiveandversatilegenefusionmarkerinhigherplants.EMBOJ.6,3901±3907.Laskowski,M.J.,Williams,M.E.,Nusbaum,H.C.andSussex,I.M.(1995)Formationoflateralrootmeristemsisatwostagepro-cess.Development,121,3303±3310.Leung,J.,Merlot,S.andGiraudat,J.(1997)TheArabidopsisABSCISICACID-INSENSITIVE2(ABI2)andABI1genesencodehomologousproteinphosphatases2Cinvolvedinabscisicacidsignaltransduction.PlantCell,9,759±771.Liscum,E.andReed,J.W.(2002)GeneticsofAux/IAAandARFactioninplantgrowthanddevelopment.PlantMol.Biol.49,387±400.Lopez-Molina,L.,Mongrand,S.andChua,N.H.(2001)Apost-germinationdevelopmentalarrestcheckpointismediatedbyabscisicacidandrequirestheABI5transcriptionfactorinAra-bidopsis.Proc.Natl.Acad.Sci.USA,98,4782±4787.Malamy,J.E.andBenfey,P.N.(1997)Organizationandcelldiffer-entiationinlateralrootsofArabidopsisthaliana.Development,124,33±44.McCourt,P.(1999)Geneticsofhormonesignaltransduction.Annu.Rev.PlantPhysiol.Mol.Biol.50,219±243.Nakamura,S.,Lynch,T.J.andFinkelstein,R.R.(2001)PhysicalinteractionsbetweenABAresponselociofArabidopsis.PlantJ.26,627±635.ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75

seedswithinthisagewereusedinallexperimentsreportedhere.Seedsforallgenotypeswereharvestedfromplantsthathadallbeengrownunderthesameconditionsofconstantlightat228C.Beforeplating,seedsweresurfacesterilizedin95%ethanolfor5minandvacuumdriedandthenplatedonto0.5ÂMSminimalmediaplatesbufferedwith500mMMEStopH5.7.Plantswereimbibedinthecoldfor3daysandthengerminatedat228Cinconstantlight.Forgerminationstudies,plantsweregrownfor7daysandgerminationwasscoredaspositivebyradicalemer-genceandcotyledonexpansion.FortheGUShistochemicalstu-dies,plantsweregrownfor7±14daysandthenstained,ortransferredtotheappropriateconcentrationofhormoneforupto3daysandthenstained(seebelow).Forthelateralrootdevelop-mentandinitiationstudies,plantsweregrownfor14daysanddevelopmentwasscoredaccordingtoZhangetal.(1999).FortheIAAandNPAstudies,plantsweregrownfor3daysandthenNPAorIAAwasappliedtotheroot±shootjunctionaccordingtoReedetal.(1998),followingwhichtheplantswereallowedtogrowfor10days.GeneticanalysisFortheepistaticanalysis,thefollowingstrainswereused:abi1-1,abi2-1,abi3-6,abi4-1,abi5-1,andera1-2.Theera1-2isanullallelewhileabi3-6isasevereallelethatcontainsaninternaldeletion.Theabi1-1andabi2-1allelesareinaLandsbergerectageneticback-groundwhileabi4-1andabi5-1allelesareinaWSbackground.AllotherlinesareinaColumbiageneticbackground.ThePABI3:GUSlineisintheC24accession(Parcyetal.,1994).Forveri®cationofmanyofthegenotypesinvolvingabi1-1andabi2-1crosses,PCRampli®cationwithprimersthatidentifytheabi1-1orabi2-1muta-tionwasused(Leungetal.,1997).Genotypingtheera1-2mutationwasperformedbySouthern(DNA)blotanalysisusingtheERA1cDNAasaprobe(Cutleretal.,1996).Often,protrudingcarpelscouldalsobeusedtoidentifytheera1homozygotes.Theidenti-®cationofabi4-1andabi5-1requiredbackcrossingofpotentialdoublestotheabi4andabi5parents.F1progenythatshowedreducedABAsensitivitywerescoredashomozygousfortheabimutations.Histochemicalb-glucoronidase(GUS)assayandmicroscopyDetectionofGUSactivitywascarriedoutusing5-bromo-4-chloro-3-indolyl-b-D-glucoronide(X-Gluc)asasubstrate.TissuewasplacedinaX-Glucbuffersolution(100mMNaPO4(pH7),0.5mMK3Fe3(CN)6,0.5mMK4Fe(CN)6,10mMEDTA,0.1%TritonX-100,20%methanol)undervacuumfor30minandallowedtoincubateovernightat378C.Formostanalyses,sampleswerethenclearedinethanolandviewedunderadissectingmicroscope.ForviewingunderDIC,materialwasmountedinachloralhydrate:-glycerol:waterclearingsolution(8:2:1).GUSactivitymeasure-mentswereperformedon40±506-day-oldseedlings,asdescribed,using4-methylumbelliferyl-b-D-glucuronide(MUG)asasubstrate(Jeffersonetal.,1987).AcknowledgementsWewouldliketothankthemembersoftheMcCourtlaboratoryandJ.Mattsonforstimulatingdiscussions.WewouldalsoliketothankDrR.Finkelsteinfortheabi4andabi5allelesusedinthisstudy.P.M.holdsanNSERC-IndustrialChair,andtheworkwasfundedbyagrantfromNSERCtoP.M.ABI3signalintegrationinArabidopsis75

Nambara,E.,Naito,S.andMcCourt,P.(1992)AmutantofArabidopsiswhichisdefectiveinseeddevelopmentandstorageproteinaccumulationisanewabi3allele.PlantJ.2,435±441.Nambara,E.,Keith,K.,McCourt,P.andNaito,S.(1995)Aregu-latoryrolefortheABI3geneintheestablishmentofembryomaturationinArabidopsisthaliana.Development,121,629±636.Nambara,E.andMcCourt,P.(1999)Proteinfarnesylationinplants:agreasytale.Curr.Opin.PlantBiol.2,388±392.Ooms,J.J.J.,Leon-Kloosterziel,K.M.,Bartels,D.,Koornneef,M.andKarssen,C.M.(1993)AcquisitionofdesiccationtoleranceandlongevityinseedsofArabidopsisthaliana.PlantPhysiol.102,1185±1191.Parcy,F.andGiraudat,J.(1997)InteractionsbetweentheABI1andtheectopicallyexpressedABI3genesincontrollingabscisicacidresponsesinArabidopsisvegetativetissues.PlantJ.11,693±702.Parcy,F.,Valon,C.,Raynal,M.,Gaubier-Comella,P.,Delseny,M.andGiraudat,J.(1994)RegulationofgeneexpressionprogramsduringArabidopsisseeddevelopment:rolesoftheABI3locusandofendogenousabscisicacid.PlantCell,6,1567±1582.Pei,Z.M.,Ghassemian,M.,McCourt,P.M.,Kwack,C.M.andSchroeder,J.I.(1998)RoleoffarnesyltransferaseinABAregula-tionofguardcellanionchannelsandplantwaterloss.Science,282,287±290.Ptashne,M.(1988)Howeukaryotictranscriptionalactivatorswork.Nature,355,683±6.Quesada,V.,Ponce,M.R.andMicol,J.L.(2000)Geneticanalysisofsalt-tolerantmutantsinArabidopsisthaliana.Genetics,154,421±436.Reed,R.C.,Brady,S.R.andMuday,G.K.(1998)InhibitionofauxinmovementfromtheshootintotherootinhibitslateralrootdevelopmentinArabidopsis.PlantPhysiol.118,1369±1378.Rohde,A.,VanMontagu,M.andBoerjan,W.(1999)TheABSCISICACID-INSENSITIVE3(ABI3)geneisexpressedduringvegetativequiescenceprocessesinArabidopsis.PlantCellEnviron.22,261±270.Rohde,A.,DeRycke,R.,Beeckman,T.,Engler,G.,VanMontagu,M.andBoerjan,W.(2000)ABI3affectsplastiddifferentiationindark-grownArabidopsisseedlings.PlantCell,12,35±52.Rohde,A.,Prinsen,E.,DeRycke,R.,Engler,G.,VanMontagu,M.andBoerjan,W.(2002)PtABI3impingesonthegrowthanddifferentiationofembryonicleavesduringbudsetinpoplar.PlantCell,14,1885±1901.Running,M.P.,Fletcher,J.C.andMeyerowitz,E.M.(1998)TheWIGGUMgeneisrequiredforproperregulationof¯oralmer-istemsizeinArabidopsis.Development,125,2545±2553.Signora,L.,DeSmet,I.,Foyer,C.H.andZhang,H.(2001)ABAplaysacentralroleinmediatingtheregulatoryeffectsofnitrateonrootbranchinginArabidopsis.PlantJ.28,655±662.Soderman,E.M.,Brocard,I.M.,Lynch,T.J.andFinkelstein,R.R.(2000)RegulationandfunctionoftheArabidopsisABA-insensi-tive4geneinseedandabscisicacidresponsesignalingnet-works.PlantPhysiol.124,1752±1765.Steber,C.M.,Cooney,S.E.andMcCourt,P.(1998)IsolationoftheGA-responsemutantsly1asasuppressorofABI1-1.Arabidopsisthaliana.Genetics,149,509±521.Steber,C.M.andMcCourt,P.(2001)AroleforbrassinosteroidsingerminationinArabidopsis.PlantPhysiol.125,763±769.Suzuki,M.,Kao,C.Y.andMcCarty,D.R.(1997)TheconservedB3domainofVIVIPAROUS1hasacooperativeDNA-bindingactiv-ity.PlantCell,9,799±807.Suzuki,M.,Kao,C.Y.,Cocciolone,S.andMcCarty,D.R.(2001)MaizeVP1complementsArabidopsisabi3andconfersanovelABA/auxininteractioninroots.PlantJ.28,409±418.Tiwari,S.B.,Wang,X.J.,Hagen,G.andGuilfoyle,T.J.(2001)AUX/IAAproteinsareactiverepressors,andtheirstabilityandactivityaremodulatedbyauxin.PlantCell,13,2809±2822.Ulmasov,T.,Hagen,G.andGuilfoyle,T.J.(1997)ARF1,atran-scriptionfactorthatbindstoauxinresponseelements.Science,276,1865±1868.Zhang,H.,Jennings,A.,Barlow,P.W.andForde,B.G.(1999)Dualpathwaysforregulationofrootbranchingbynitrate.Proc.Natl.Acad.Sci.USA,96,6529±6534.ßBlackwellPublishingLtd,ThePlantJournal,(2003),34,67±75