G protein ABA

GProteinRegulationofIonChannelsandAbscisicAcidSignalinginArabidopsis

GuardCells

Xi-QingWang,1HemayetUllah,2AlanM.Jones,2SarahM.Assmann1*

Thephytohormoneabscisicacid(ABA)promotesplantwaterconservationbydecreasingtheaperturesofstomatalporesintheepidermisthroughwhichwaterlossoccurs.WefoundthatArabidopsisthalianaplantsharboringtrans-ferredDNAinsertionalmutationsinthesoleprototypicalheterotrimericGTP-binding(G)protein␣subunitgene,GPA1,lackbothABAinhibitionofguardcellinwardKϩchannelsandpH-independentABAactivationofanionchannels.Stomatalopeningingpa1plantsisinsensitivetoinhibitionbyABA,andtherateofwaterlossfromgpa1mutantsisgreaterthanthatfromwild-typeplants.ManipulationofGproteinstatusinguardcellsmayprovideamechanismforcontrollingplantwaterbalance.HeterotrimericGproteinsarekeyregulatorsofionchannelsinanimalcells(1,2).Uponactivation,theGprotein␣subunit(G␣)bindsGTP,resultinginseparationofthe␣subunitfromthe␤␥subunitpair(G␤␥).G␣andG␤␥canbothinteractwithdownstreamcomponentsofsignalingpathways(2,3).AmongtheimportantdownstreameffectorsareKϩandCa2ϩchannels,whichareregu-latedbyGproteinsviabothcytosolicsignal-ingcascadesandmembrane-delimitedpath-ways(1,2).Gprotein–mediatedion-channelregulationisanintegralcomponentofvision,taste,smell,andhormonalsignalinginmam-maliansystems(2–4).

Inhigherplants,guardcellion-channelregulationcontrolsstomatalapertures.Sto-matalopeningreliesonincreasesinKϩ,Cl–,malate2–,andsucroseintheguardcellsym-plasttodrivewaterinfluxandcellswelling.Theseprocessesresultinanoutbowingoftheguardcellpairandanincreaseinporeaper-ture.Duringstomatalopening,KϩuptakeismediatedbyinwardlyrectifyingKϩchan-nels.Duringinhibitionofstomatalopeningbytheplanthormoneabscisicacid(ABA),thesechannelsareinhibited(5–7).Inguardcells,ABAactivatesphospholipasesCandD(8–10)andcanelevatecytosoliccalciumlev-elsviainositol1,4,5-trisphosphateorotherpathways(8,11–14).CytosolicCa2ϩeleva-tioninhibitsinwardlyrectifyingKϩchannels(6,7,15)andactivatesslowanionchannels

BiologyDepartment,PennsylvaniaStateUniversity,208MuellerLaboratory,UniversityPark,PA16802–5301,USA.2DepartmentofBiology,UniversityofNorthCarolina,ChapelHill,NC27599–3280,USA.

1*Towhomcorrespondenceshouldbeaddressed.E-mail:sma3@psu.edu

thatmediateCl–andmalate2–efflux(15,16).Inmammaliansystems,certainphospho-lipasesCandDareregulatedbyheterotri-mericGproteins.IntheArabidopsisthalianagenome,thereisonlyoneprototypicalG␣gene,GPA1(17,18),andthisgeneisexpressedinguardcells(Fig.1A)(19).Thus,wehypothesizedthatGPA1mayreg-ulateionchannels(20)andABAresponse(21)inthiscelltype.Weusedtwoinde-pendentArabidopsislinesharboringthere-cessivetransferredDNA(T-DNA)knock-outallelesgpa1-1orgpa1-2(22)totestthesehypotheses.

Guardcellsisolatedfromhomozygousgpa1-1andgpa1-2plants(23,24)failedtoexpressfull-lengthGPA1transcripts(Fig.1B),asexpected(19,22).Incontrasttotheresponseofwild-typeplants,stomataofgpa1mutantplantsshowednoinhibitionofstoma-talopening(25,26)byABA(Fig.2A).WenextusedpatchclamptechniquestotestwhethersensitivityoftheinwardKϩchan-nelstoinhibitionbyABAhadalsobeenalteredinthegpa1mutants(27).Justasforstomatalopening,theinwardKϩchannelsofthemutantplantswereABAinsensitive(Fig.2,BandC).Thus,ABAinhibitionofinwardKϩchannelsandstomatalopeningrequirethepresenceoffunctionalGPA1.

ABAactivationofslowanionchannels(28–30)isthoughttobeanothercomponentofABAinhibitionofstomatalopening.An-ioneffluxoccurringuponanionchannelopeningshoulddepolarizethemembraneandhinderKϩuptake.However,thehypothesisthatGproteinsregulateanionchannelsinplantshadnoexperimentalprecedence,andfewmammaliananionchannelsareGproteinregulated.Accordingly,wetestedwhether

ABAactivationofslowanionchannelswasalteredinthegpa1mutants.Weinitiallyusedanestablishedprotocolforrecordingthewhole-cellanionchannelresponsetoABA,whichusesacytosolicsolutionwithmoder-ateCa2ϩandstrongpHbufferingcapacities(29,31).WefoundthatGPA1doesregulateplantanionchannels:Inbothgpa1mutants,ABAactivationofanionchannelswasabol-ished(Fig.3,AandB).

Activationofanionchannelsalsopro-motesstomatalclosurebymediatinglossofanionicsolutesandmembranedepolarizationthatdrivesKϩefflux(28–30,32,33).Thus,weexpectedthatABApromotionofstomatalclosure(34)wouldalsobeeliminatedingpa1mutantplants,butitwasnot(Fig.3C)(35).TheuncouplingofABAinhibitionofstomatalopening(Fig.2A)andABApromo-tionofstomatalclosure(Fig.3C)demon-stratesthatthesetwoeffectsarenotsimplythereverseofoneanother(28).

Considerationoftheanionchannelandstomatalclosureexperimentsledtotheinter-pretationthataparallelorcompensatorypathwaymediatingABApromotionofsto-matalclosurewaspresentintheintactgpa1guardcells(Fig.3C)butwasnonfunctionalundertheconditionsofthepatchclampex-periments(Fig.3,AandB).OnecandidatememberofsuchapathwayiscytosolicpH.CytosolicpHwasstronglybufferedintheprotocolofFig.3,AandB,yetABAelevatescytosolicpHinintactguardcells(34,36).

Fig.1.RT-PCRanalysisofGPA1expressioninguardcellsofwild-typeandgpa1mutantlines.(A)GPA1ismorehighlyexpressedinguardcellprotoplasts(GCP)thaninmesophyllcellpro-toplasts(MCP)ofwild-typeArabidopsis.Differ-entcyclenumberswereperformedasindicatedtoassessGPA1expressionlevels.Expectedbandsizesof605bp(GPA1)and1kb(Actin)areobserved;theidentityofGPA1productswasconfirmedbysequencing.(B)GPA1isex-pressedinwild-typeguardcellsandnotintheguardcellsofgpa1knockoutmutants,asas-sessedbyRT-PCR.GPA1primers(lanes2,4,and6)bracketedtheT-DNAinsertionsite;thus,noPCRproductwasexpectedfromthegpa1lines.ActinbandsindicaterelativecDNAamounts.

207015JUNE2001VOL292SCIENCEwww.sciencemag.org

Therefore,wenextobtainedelectrophysio-logicalmeasurementsunderweakcytosolicpHbuffering(31)thatwouldpermitobserva-tionofanionchannelregulationbyABA-inducedalterationsinpH.NormalABAac-tivationofanionchannelswasobservedinallgenotypes(Fig.3,DandE),consistentwiththeobservationofABA-inducedstomatalclosureinepidermesofmutantplants(Fig.

REPORTS

3C).Thus,thepH-independentpathwayofABAaction(Fig.3,AandB)requiresGPA1,whereasthepH-dependentpathway(Fig.3,DandE)doesnot.Consistentwiththismod-el,blockofthepH-dependentpathwayinepidermalpeelsbyacytosolicpHclampwithbutyricacid,amembrane-permeantweakacid(34,35),opposedABA-inducedstoma-talclosureinallthreegenotypes(Fig.3F).Stomatalaperturesinleavestakendirectlyfromthegrowthenvironment(24)wereϳ20%greateringpa1mutantlinesthaninwild-typeplants,suggestingthatstomatalsensitivitytobackgroundlevelsofendoge-nousABAisshiftedinthegpa1mutants.Waterloss(37,38)isgreaterfromgpa1leavesthanfromwild-typeleaves(Fig.4),confirmingthateliminationofGPA1impactswaterrelationsinplanta.Thisinformationmaycontributetoeffortstoengineerplantswithimprovedstomatalregulation.

PreviouspharmacologicalstudiesofguardcellfunctionwithGproteinactiva-torssuchasguanosine5Ј-O-(3Ј-thio-triphosphate)(GTP-␥-S),andinactivatorssuchasguanosine5Ј-O-(2Ј-thiodiphos-phate)(GDP-␤-S)(18),suggestedthatac-tiveGproteinsinactivatetheinwardKϩchannelsviabothcytosolicandmembrane-delimitedpathwaystoinhibitstomatalopening(21,39–41).Suchdatawerechal-lenged,however,byotherpharmacologicalstudiessuggestingthatGproteinactivationcouldstimulatestomatalopening(42)orhadbothstimulatoryandinhibitoryeffectsoninwardKϩcurrents(43).TheseresultscanbeclarifiedbytheprecisionaffordedbyT-DNAmutagenesis(44),inwhichonespecificproteiniseliminatedwithretentionofanotherwisewild-typegeneticcomple-ment.Reversegeneticsalsohasallowedustodirectlytestandsupportthehypothesis

thatABAsignalinginguardcellsusesGpro-teinactivation(21).ThereisevidencethatplantGproteinsareinvolvedinresponsestolight(45),pathogens(46),andseveralhormones(18,47),includingABAasshownhere.ItwillbeofinteresttounravelhowplantGproteinpathwayscancouplereceptorswiththeircog-natedownstreameffectsforsuchdiverseandmultiplesignals,giventhattheArabidopsisge-nomecontainsonlyGPA1asaprototypicalG␣subunitgene.

1.A.M.Brown,L.Birnbaumer,Annu.Rev.Physiol.52,197(1990).

2.K.Wickman,D.E.Clapham,Physiol.Rev.75,865(1995).

3.J.R.Hepler,A.G.Gilman,TrendsBiochem.Sci.17,383(1992).

4.G.M.Sheperd,Cell67,845(1991).5.M.R.Blatt,Planta180,445(1990).

6.F.Lemtri-Chlieh,E.A.C.MacRobbie,J.Membr.Biol.137,99(1994).

7.X.-Q.Wang,W.-H.Wu,S.M.Assmann,PlantPhysiol.118,1421(1998).

8.Y.S.Leeetal.,PlantPhysiol.110,987(1996).

9.T.Jacob,S.Ritchie,S.M.Assmann,S.Gilroy,Proc.Natl.Acad.Sci.U.S.A.96,12192(1999).

10.I.Staxenetal.,Proc.Natl.Acad.Sci.U.S.A.96,1779

(1999).

11.S.Gilroy,N.D.Read,A.J.Trewavas,Nature346,769

(1990).

12.M.R.McAinsh,C.Brownlee,A.M.Hetherington,

Nature343,186(1990).

13.C.P.Leckie,M.R.McAinsh,G.J.Allen,D.Sanders,

A.M.Hetherington,Proc.Natl.Acad.Sci.U.S.A.95,15837(1998).

14.E.A.C.MacRobbie,Proc.Natl.Acad.Sci.U.S.A.97,

12361(2000).

15.J.I.Schroeder,S.Hagiwara,Nature338,427(19).16.G.J.Allen,K.Kuchitsu,S.P.Chu,Y.Murata,J.I.

Schroeder,PlantCell11,1785(1999).

17.H.Ma,M.F.Yanofsky,E.M.Meyerowitz,Proc.Natl.

Acad.Sci.U.S.A.87,3821(1990).

18.H.Ma,PlantMol.Biol.26,1611(1994).

19.RNAwasextractedfromguardcell(puritypercent-ageof99:1,guardcellprotoplasts:mesophyllcellprotoplasts)andmesophyllcellprotoplastprepara-tions.Reversetranscriptionfollowedbypolymerasechainreaction(PCR)wasperformed[SuperScriptfirst-strandsynthesissystemforreversetranscriptase(RT)–PCR,GibcoBRL,LifeTechnologies].Theforwardprimerstartedat662basepairs(bp)intheGPA1cDNAclone(forward;5Јggctgctgaaatcgaaagac3Ј);thereverseprimerstartedat1266bp(reverse;5Јgtccacccacgtcaaacaat3Ј).

20.G.S.Aharon,A.Gelli,W.A.Snedden,E.Blumwald,

FEBSLett.424,17(1998).

21.S.M.Assmann,inSignalTransductioninPlantGrowth

andDevelopment,D.P.S.Verma,Ed.(Springer,Vi-enna,1996),pp.39–61.

22.H.Ullahetal.,Science292,2066(2001).

23.L.A.Romano,H.Miedema,S.M.Assmann,PlantCell

Physiol.39,1133(1998).

24.Weused4-to6-week-oldplantsgrownundershort

days.Seedsweregerminatedinastandardmedium(23)in7.5%(w/v)phytagel(Sigma),withkanamycin(50mg/liter)formutants.After10days,robustseedlingsweretransferredtoMetromix250pottingmixture(Scotts,Marysville,OH).

25.J.Leymarie,G.Lasceve,A.Vavasseur,Aust.J.Plant

Physiol.25,785(1998).

26.Light-inducedstomatalopeningwasassayedaccord-ingto(25).Normalizationofaperturewidthsbystomatallength(16)mirroredtheresultsshownhereforwidthsalone(48).

27.Guardcellprotoplastswereisolatedfromrosette

leavesof4-to5-week-old,nonboltingplants[after(7)],exceptthatthesecondenzymesolutioncon-tained1.3%(w/v)CellulaseRS(YakultHonsha,To-kyo)and0.0075%(w/v)PectolyaseY-23(Seishin

Pharmaceutical,Tokyo).Solutionsusedinpatch-clampexperimentsonKϩchannelswereasfollows:10mMMES(tris,pH6.0),10mMK-glutamate,4mMMgCl2,and1mMCaCl2[osmolalityof0mmol/kg(bathsolution)]and10mMHepes-tris(pH7.8),80mMK-glutamate,and20mMKCl[osmolalityof560mmol/kg(pipettesolution)].Freshadenosinetriphosphate(ATP)(5mMfroma0.5MMgATPstocksolutionin0.5Mtris)wasaddeddailytothepipettesolution.Finalosmolalitieswereobtainedbyadditionofsorbitol.Time-activatedcurrentswerecalculatedasthedifferencebetweenaveragesteady-statecur-rentbetween3000and3800msandinstantaneouscurrentat20ms.

J.Li,X.-Q.Wang,M.B.Watson,S.M.Assmann,Science287,300(2000).

Z.-M.Pei,K.Kuchitsu,J.M.Ward,M.Schwarz,J.I.Schroeder,PlantCell9,409(1997).

A.Grabov,J.Leung,J.Giraudat,M.R.Blatt,PlantJ.12,203(1997).

Theinitialpipettesolutionforanionchannelrecord-ingwasasfollows:150mMCsCl,2mMMgCl2,6.7mMEGTA,3.35mMCaCl2,10mMHepes-tris(pH7.5),withosmolalityof0mmol/kg(29).Fresh5mMMg⅐ATPand5mMtris-GTPwereaddeddaily(29).WhenwestudiedtheroleofpHinanionchannelregulationbyABA,thepHbufferinthepipettesolutionwas0.1mMHepes-tris(pH7.5).Thebathsolutionwas30mMCsCl,2mMMgCl2,5mMCaCl2,10mMMes-tris(pH5.6),withosmolalityof560mmol/kg.Protoplastswerepretreatedforatleast1.5hourswithABA,and50␮MABAwasaddedtobathandpipettesolutions(29).Anioncurrentswererecorded11to14minafterachievingthewhole-cellconfiguration.Steady-statecurrentswereaveragedbetween52.5and62.5s.

E.A.C.MacRobbie,J.Exp.Bot.48,515(1997).R.Hedrich,H.Busch,K.Raschke,EMBOJ.9,38(1990).

M.R.Blatt,F.Armstrong,Planta191,330(1993).ForFig.3C,stomatalclosurewasassayedaccordingto(29),exceptthatafterABAapplication(20␮M),leaveswereincubatedfor3hoursbeforeaperturemeasurement.ForFig.3F,weassayedstomatalclo-surewiththeprotocolof(34),althoughwithminormodification.Leaveswereincubatedfor2hoursun-der450␮molmϪ2sϪ1oflight,followedbytheadditionof1mMNa-butyrateand20␮MABA.Aperturesweremeasured2hourslater.Thewild-typebutyrateresponse(48)wassaturatedwith1mMNa-butyrate.

H.R.Irving,C.A.Gehring,R.W.Parish,Proc.Natl.Acad.Sci.U.S.A.,1790(1992).

J.Leung,S.Merlot,J.Giraudat,PlantCell9,759(1997).

Onthebasisof(37),5to6youngleaveswereexcisedfromanindividualplant,andtheirtotalfreshweightwasmeasuredduringincubationinachamberthroughwhichdryairwaspassed.

K.Fairley-Grenot,S.M.Assmann,PlantCell3,1037(1991).

W.-H.Wu,S.M.Assmann,Proc.Natl.Acad.Sci.U.S.A.91,6310(1994).

F.Armstrong,M.R.Blatt,PlantJ.8,187(1995).H.J.Lee,E.B.Tucker,R.C.Crain,Y.Lee,PlantPhysiol.102,95(1993).

W.K.Kelly,J.E.Esser,J.I.Schroeder,PlantJ.8,479(1995).

P.J.Krysan,J.C.Young,M.R.Sussmann,PlantCell11,2283(1999).

K.M.Warpeha,H.E.Hamm,M.M.Rasenick,L.S.Kaufman,Proc.Natl.Acad.Sci.U.S.A.88,25(1991).

A.Gelli,V.J.Higgins,E.Blumwald,PlantPhysiol.113,269(1997).

M.Ueguchi-Tanakaetal.,Proc.Natl.Acad.Sci.U.S.A.97,11638(2000).

X.-Q.Wang,S.M.Assmann,unpublishedobservations.SupportedbyUSDAgrants98-35304-6681and2001-35304-09916andNSFgrantMCB-9874438toS.M.A.WethankZ.-M.Peiforadviceonpatch-clamp-ingArabidopsisguardcells.

16January2001;accepted10April2001

28.29.30.31.

ReferencesandNotes

32.33.34.35.

36.37.38.

39.40.41.42.43.44.45.46.47.48.49.

Fig.4.Waterlossisgreaterfromgpa1mutantleaves.Waterlossisexpressedasthepercent-ageofinitialfreshweight(37).ValuesaremeanϮSE(errorbars)ofmeasurementswiththreeindividualplantspergenotype.Oneofthreeindependenttrialsisshown.Regressionanalysisconfirmedthatthewild-typecurvedifferssignificantlyfromthegpa1-1andgpa1-2responsesatPՅ0.01.Circles,wildtype;squaresgpa1-1;triangles,gpa1-2.

207215JUNE2001VOL292SCIENCEwww.sciencemag.org