GIBBERELLIN BIOSYNTHESIS: Enzymes, Genes and...

1040-2519/97/0601-0431$08.00

431

HEDDEN& KAMIYAGIBBERELLIN BIOSYNTHESISAnnu.Rev. Plant Physiol. Plant Mol. Biol. 1997. 48:431–60Copyright© 1997by AnnualReviewsInc. All rightsreserved

GIBBERELLIN BIOSYNTHESIS:Enzymes, Genesand Their Regulation

PeterHeddenIACR-Long Ashton Research Station, Departmentof Agricultural Science, Universityof Bristol, Bristol, BS189AF, UnitedKingdom

Yuji KamiyaFrontierResearchProgram, TheInstituteof Physical and Chemical Research (RIKEN),Hirosawa2-1,Wako-shi, Saitama351-01, Japan

KEY WORDS: terpene cyclases, monooxygenases, 2-oxoglutarate-dependent dioxygenases,feedback,light andtemperature regulation

ABSTRACT

Therecent impressiveprogressin researchongibberellin (GA) biosynthesishasresulted primarily from cloning of genesencoding biosynthetic enzymes andstudies with GA-deficient andGA-insensitive mutants. Highlights include thecloningof ent-copalyl diphosphatesynthaseandent-kaurenesynthase(formallyent-kaurenesynthasesA andB) andthedemonstrationthat theformer is targetedto theplastid; thefindingthattheDwarf-3 geneof maizeencodesacytochromeP450, althoughof unknown function; andthe cloning of GA 20-oxidaseand3β-hydroxylase genes. Theavailability of cDNA and genomicclones for theseenzymesis enabling themechanismsby which GA concentrationsareregulatedby environmental andendogenousfactors to bestudied at the molecular level.For example, it has been shown that transcript levels for GA 20-oxidaseand3β-hydroxylasearesubject to feedback regulationby GA action and, in thecaseof the GA 20-oxidase, are regulated by light. Also discussed is other newinformation, particularly from mutants, that has addedto our understanding ofthebiosynthetic pathway, theenzymes,andtheir regulationand tissuelocaliza-tion.

CONTENTSINTRODUCTION..................................................................................................................... 432ent-KAURENESYNTHESIS...................................................................................................... 433

ent-Copalyl DiphosphateSynthase...................................................................................... 436ent-KaureneSynthase.......................................................................................................... 437Subcellular Localization of ent-Kaurene Synthesis............................................................. 437

MONOOXYGENASES............................................................................................................ 439DIOXYGENASES.................................................................................................................... 440

7-Oxidase............................................................................................................................. 44113- and 12α-Hydroxylases.................................................................................................. 44120-Oxidases......................................................................................................................... 4423β-Hydroxylasesand RelatedEnzymes.............................................................................. 4442β-Hydroxylasesand RelatedEnzymes.............................................................................. 447

REGULATION OFGA BIOSYNTHESIS............................................................................... 448FeedbackRegulation........................................................................................................... 448RegulationbyLight ............................................................................................................. 449RegulationbyTemperature................................................................................................. 451

SITESOFGA BIOSYNTHESIS.............................................................................................. 452CONCLUDING REMARKS.................................................................................................... 453

INTRODUCTION

The gibberellin(GA) hormonesact throughoutthe life cycle of plants,influ-encingseedgermination, stemelongation,flower induction,antherdevelop-ment,andseedandpericarpgrowth.Furthermore,they mediateenvironmentalstimuli, which modify theflux throughtheGA-biosynthetic pathway.Regula-tion of GA biosynthesis istherefore of fundamentalimportanceto plantdevel-opmentand itsadaptationto theenvironment.

The last review of GA biosynthesis in this series,by Graebe(38), wasfollowed by severalreviewscoveringthis topic (68,75,121).Graebe’s articlefocusedon GA-biosynthetic pathwaysin cell-freesystems,characteristicsofbiosynthetic enzymes,andfactorsaffectingGA production.He predictedac-curately that the main topic of the next review in this serieswould be thecloning andcharacterizationof genesfor theGA-biosynthetic enzymes.Sev-eralof the enzymes havenow beencloned, and theavailability of theircDNAsis providing new and often unexpectedinformation on the natureof theseproteins.For example,someof the enzymescatalyzemultiple stepsin thepathway.It is alsonow possible to investigatethemechanismsby which GAbiosynthesis is regulatedin responseto environmentaland endogenoussig-nals.

In light of theseexciting advances,a new review on GA biosynthesis isappropriate.As is usualpractice,we discussthebiosyntheticpathway,shownin Figures1 and2, in threesectionsaccordingto the natureof the enzymes:terpenecyclasesinvolved in ent-kaurenesynthesis, monooxygenases,anddi-

432 HEDDEN& KAMIYA

oxygenases.We alsoincludediscussionson regulationandsitesof synthesis.Werestrictourselvesto higherplants,becauseGA biosynthesis in otherorgan-isms,principally fungi andferns,is not aswell understood.However,futurephylogeneticcomparisonsbetweenGA biosynthesisgenesin all organismsshouldbe instructivein determining the origin of GAs. Much of the currentprogresson GA biosynthesishascomefrom work with mutants, which hasbeencoveredcomprehensivelyin severalreviews (98, 99,101,105) andis notdealt with specifically here.Someof the better characterizedGA-deficientmutants,with the position of lesions,are given in Table 1, which also listscDNA clones for biosyntheticenzymes.

ent-KAURENE SYNTHESIS

ent-Kaurene is synthesizedby the two-step cyclization of geranylgeranyldiphosphate(GGDP) via the intermediate, ent-copalyl diphosphate(CDP).The enzymesthat catalyzethesereactionsare referred to as the A and Bactivities, respectively,of ent-kaurenesynthase(formerly ent-kaurenesyn-thetase).However, we adopt the more logical nomenclature proposedbyMacMillan (75). Thus, the conversionof GGDP toCDP is catalyzedbyent-copalyl diphosphate synthase(CPS)and of CDP to ent-kaureneby ent-kaurenesynthase(KS). The biosynthesis of GGDP from mevalonicacid iscommonto manyterpenoidpathwaysandis coveredin thereviewby Chappell

Figure 1 Early GA-biosynthetic pathwayto GA12-aldehyde.GGDPis producedin plastidsby theisoprenoidpathway, originating frommevalonic acidor,possibly, pyruvate/glyceraldehyde3-phos-phate.

GIBBERELLIN BIOSYNTHESIS 433

(13). Recently,a nonmevalonate pathwayto isoprenoids,involving pyruvateand glyceraldehyde-3-phosphate, has been proposedin green algae(112).Suchapathway may operatein plastidsof higherplants, giventhedifficulty indemonstrating theincorporationof mevalonateinto isoprenoids in theseorgan-elles.

Figure 2 Gibberellin-biosynthetic pathwayfrom GA12-aldehyde.

434 HEDDEN& KAMIYA

CPSand KS were first separatedby anion-exchangechromatographyonextractsof Marah macrocarpusendosperm(23). Therewerealsoindicationsfor the involvementof two enzymesfrom studieson GA-deficientmutants;work with cell-freeextractsof young fruits suggestedthat the dwarf tomatomutants,gib-1andgib-3,havelesions atCPSandKS, respectively(10).Usinganent-kaureneoxidaseinhibitor to estimate ratesof ent-kaurenebiosynthesis,a methodfirst usedwith germinatingbarley grain (40), Zeevaart& Talon(150) demonstratedthat the nonallelic ga1 and ga2 mutantsof Arabidopsis

Table 1 MutantsandcDNA clonesfor GA-biosynthetic enzymes

Enzyme Plant Mutant References cDNAcloning

Database

CPS ArabidopsisthalianaZeamaysPisum sativumLycopersiconesculentum

ga1An1ls-1gib-1

635312810

12492—

U11034L37750U63652

KS Cucurbita maximaA. thalianaZ. maysL. esculentum

—ga2d5gib-3

—1504810

147———

U43904

ent-Kaureneoxidase

P. sativumA. thalianaOryza sativa

lhi

ga3dx

12715089

———

Monoxy-genase

Z. maysP. sativum

d3na

2849

144—

U32579

GA 20-oxidase

C. maximaA. thaliana

A. thaliana

P. sativum

P. sativumPhaseolusvulgaris

O. sativaSpinaciaoleracea

—ga5

—

—

——

——

—129

—

—

——

——

71146

92

77317331

137145

X73314U20872U20873U20901X83379X83380X83381X91658U70471U58830U70530U70531U70532U50333U33330

GA 3β-hy-droxylase

A. thalianaZ. maysO. sativaP. sativumLathyrusodoratus

ga4dldylel

129285950106

14————

L37126

GA 2-oxidases P. sativum sln 108 —


were both defective inent-kaurene production,whereasga4 and ga5 areblockedlater inthe pathway(129).

ent-CopalylDiphosphateSynthase

Koornneefet al (63) constructeda fine structuregeneticmapof theArabidop-sisGA1locususingnineindependentga1alleles,threeof whichweremadebyfastneutronbombardmentandtherestby treatmentwith ethylmethanesulfon-ate(EMS).Amongthefast-neutron-generatedmutants,ga1-3containsa largedeletion(5 kb) and failed to recombinewith the EMS-treatedmutants(63).Sun et al (123) usedga1-3 to clone the GA1 locus by genomicsubtraction.Cosmid clonescontainingwild-type DNA insertsspanningthe deletion inga1-3 complemented the dwarf phenotype when integratedinto the ga1-3genomeby T-DNA transformation. GA1 cDNA containsa 2.4-kbopenread-ing frame, which was shown by functional analysisto encodeCPS (124).Escherichiacoli co-transformedwith a bacterialGGDP synthasegene(12),and the GA1 cDNA producedCDP, from which copalol was identified bycombined gas chromatography–massspectrometry (GC-MS) afteralkaline hy-drolysis. Although ga1-3 contains alarge deletion andgenomic Southernanalysisindicatedthat GA1 is a single-copygene,the mutantproduceslowamountsof GAs (150), suggesting that thereareGA1 homologues in Arabi-dopsisor thereis an alternativepathwayfor ent-kaurenesynthesis. A similarsituationexistsin maize,from which theAn1(Antherear-1) locuswasclonedby transposontagging(9). The predictedaminoacid sequenceof An1 cDNAshareshigh sequenceidentity (51%, without transit peptidesequence)withthatof theGA1 protein.A homozygousdeletionmutantof An1,an1-bz2-6923,accumulatedent-kaureneto 20%of thewild-type content,indicating thepres-enceof isoenzymes.Furthermore,a putative homologouscDNA, An2, wasclonedby RT-PCR (8). At least two different GA1 homologueshavebeenobtainedfrom tomato seedlingsby RT-PCR using oligonucleotideprimersbasedon ArabidopsisandmaizeCPSsequences(R Imai, personalcommuni-cation). It appears,therefore,that leakinessof the ga1-3 and an1 deletionmutantsis due tothepresenceof other CPSs.

TheLs locusof peawasshownto encodeCPS.The ls-1 dwarf mutanthadreducedCPSactivity in a cell-freesystemfrom immatureseeds(127).Confir-mation was obtainedafter cloning a CPSfrom peaby RT-PCR,its identitybeingconfirmedby expressionin E. coli of a glutathioneS-transferasefusionproteinwith CPSactivity (2). The ls-1 mutation, producedby EMS treatment,is dueto a G-to-A substitution at an intron-exonborderthat causesimpairedsplicinganda frameshiftin thetranscript(2).

436 HEDDEN& KAMIYA

ent-KaureneSynthase

ent-Kaurenesynthase(KS) waspurified from endospermof pumpkin (Cucur-bita maxima) (110),which is a rich sourceof GA-biosynthetic enzymes(38).The enzyme,which hada predictedMr of 81,000,requireddivalentcations,suchasMg2+, Mn2+, andCo2+, for activity andhadan optimal pH rangeof6.8–7.5(110).TheKm for CDPwas0.35µM. Purificationof KS wasquicklyfollowed by its molecularcloning (147). PCRwasusedwith degenerateoli-gonucleotides,designedfrom aminoacidsequencesof thepurified protein,toproducea cDNA fragment for library screening.The isolated full-lengthcDNA was expressedin E. coli as a fusion protein, with maltose-bindingprotein,which converted[3H]CDP to ent-[3H]kaurene.The KS transcriptisabundantin growingtissues,suchasapicesanddevelopingcotyledons,andispresentin everyorganin pumpkin seedlings.Although it is difficult to com-pare mRNAabundance across species,it appears thatKS is expressed atmuchhigherlevelsthanis CPS.WhereasCPStranscriptsareundetectablein leavesof Arabidopsis(124) andpea(2) by northernblot analysis,requiringRNaseprotectionassays(A Silverstone& TP Sun,personalcommunication)or RT-PCR,respectively,KS mRNA, thoughof low abundance,canbe assayedinpumpkinleavesby northernhybridization(147).This is consistentwith strictregulationof thefirst stepof ent-kaurenesynthesis fromthe abundantGGDP.

Thededucedaminoacidsequenceof KS sharessignificanthomology withother terpenecyclases(Figure3A), with highesthomology (51% aminoacidsimilarity) with CPSfrom Arabidopsisand maize.It containsthe DDXXDmotif, which is conservedin casbenesynthase(81), 5-epi-aristolochenesyn-thase(24),andlimonenesynthase(17),andwhich is proposedto functionas abinding site for the divalentmetal ion-diphosphatecomplex(13). CPSlacksthe DDXXD motif, consistentwith its catalytic activitynot involvingcleavageof the diphosphate group.

SubcellularLocalizationof ent-KaureneSynthesis

Althoughtherehasbeenevidencefor ent-kaurenesynthesisin plastidsfor over20 years(reviewedin 38),unequivocalconfirmationof this hasbeenprovidedonly recently.By preciseuseof markerenzymesto assessplastidpurity andGC-MSto identify enzymeproducts,Aachet al (1) clearlydemonstratedthatCPS/KSactivity (GGDP to ent-kaurene)is localized in developingchloro-plasts from wheatseedlingsandleucoplasts from pumpkin endosperm. Maturechloroplastscontainedlitt le activity. Theseresults were supportedby therecentcDNA cloning of CPS(124) and KS (147). The first 50 N-terminalamino acids of the GA1 protein (ArabidopsisCPS) are rich in serineand


Figure 3 Phylogenetic trees,producedusing the PHYLIP package(J Felsenstein, University ofWashington, Seattle), for (A) terpenecyclases,including ent-copalyl diphosphatesynthase(CPS)andent-kaurenesynthase(KS), and(B) GA 20-oxidases.Referencenumbers areshown in paren-theses.(Sources:aR Croteauetal, unpublished data; bNEJ Appleford, JR Lenton, AL Phill ips & PHedden,unpublisheddata; cDA Ward, J MacMillan,AL Phill ips & PHedden,unpublisheddata.)

438 HEDDEN& KAMIYA

threoninewith an estimatedpI of 10.2 (124). Suchpropertiesare commonfeaturesof precursorsof manychloroplast-localizedproteins,suchasthesmallsubunitof Rubisco(54). Thetransitpeptideis cleaved on entryinto the plastidto producea functionalmatureprotein.Incubationof a 35S-labeledArabidop-sispre-CPSof 86 kDa with isolatedpeachloroplastsresultedin transportintothechloroplastsandprocessingto a 76-kDaprotein(124). Thededuced aminoacid sequenceof KS alsocontainsa putativetransitpeptide,althoughimportinto chloroplastscould notbe demonstrated (147).

Plastidsarethemajorsiteof productionof GGDP,andmostof theGGDPsynthases cloned inplantshavetransitpeptides forplastid transport(6). Local-izationof GGDPsynthasefrom Capsicumannumin plastids hasbeendemon-stratedimmunocytochemically (64). GGDPis a commonprecursorfor manyplastid-localizedterpenoids,including carotenoidsandthephytolside-chainofchlorophyll. Overexpressionof phytoenesynthase,which convertsGGDPtophytoene,in transgenictomato resultedin a lower chlorophyllcontentthaninwild-typeplantsandadwarfphenotypethatwaspartiallyreversedby applyingGA3 (27). The endogenousGA concentrationsin apicalshootsof the trans-genic plants were reducedto about 3% of that in wild-type shoots.If, assuggested(27), overproductionof phytoenehasdepletedGGDPcontent,re-sulting in reducedsynthesisof GA andchlorophyll, the threepathwaysmustshare the samepoolof GGDP and thusbe interdependent.

MONOOXYGENASES

The highly hydrophobic ent-kaurene is oxidized by membrane-boundmonooxygenasesto GA12. TheenzymesrequireNADPH andoxygenand,onthe basisof the early demonstrationthat ent-kaureneandent-kaurenaloxida-tion are inhibited by carbonmonoxidewith reversibility by light at 450 nm(86), areall assumedto involve cytochromeP450.The involvementof cyto-chromeP450in ent-kaurenoicacid 7β-hydroxylasein the fungusGibberellafujikuroi hasnowalsobeenshown(51).At leastoneof theenzymes(GA12-al-dehyde synthase)is associatedwith the endoplasmic reticulumin peaembryosandpumpkin endosperm(37), requiringtransportof ent-kaurene,or perhapsalater intermediate, from theplastid.

SeveralGA-deficient dwarf mutantsare defectivein ent-kaureneoxidaseactivity. In pea,the lhi (lh-2) mutation affectsstemelongationandseeddevel-opment(125–127).Cell-freeextractsfrom immaturelh-2 seedsweredeficientin ent-kaureneoxidase activityrelative towild-typeseeds; thethree steps froment-kaureneto ent-kaurenoicacid wereaffected,suggesting that a singleen-zymemight catalyzethesereactions(127).However,unequivocalverification


of this mustawait theavailability of pureenzymebecausea regulatoryfunc-tion for Lh cannotbeexcluded.TheTan-ginbozumutantof rice (dx) is prob-ably also deficient in ent-kaureneoxidaseactivity (89). Application of uni-conazole,anent-kaureneoxidaseinhibitor, mimics thephenotypeof Tan-gin-bozu and producesendogenousGA concentrationssimilar to those in themutant(89).Becauseof growthresponsesto appliedent-kaurene,Tan-ginbozu(85) and also lh (49) were thought previously to have lesionsbefore ent-kaurenesynthesis. Therecentfindings(89,127) indicatethatsuchapplicationexperimentsmaygive misleading results.

Although lessprogresshasbeenmadewith themonooxygenasesthanwiththeother enzymes, the recentcloning oftheDwarf-3 (D3) geneof maize (144)shouldenablerapidprogressin characterizingthis groupof enzymes.TheD3gene,obtainedby transposontagging,wasfound,on the basisof its deducedaminoacidsequence,to encodea memberof a newclassof cytochromeP450monooxygenases with closest homology to sterol hydroxylases. Unfortu-nately, there is uncertaintyabout the stepcatalyzedby the D3 protein (BOPhinney,personalcommunication),althoughthis shouldnow be revealedbyfunctionalexpressionof thecDNA in a suitable heterologoussystem.

DIOXYGENASES

The enzymesinvolved in the third stageof the pathwayaresolubleoxidasesthatuse2-oxoglutarateasa co-substrate.These2-oxoglutarate–dependentdi-oxygenasesbelongto a family of nonhemeFe-containingenzymesthat havebeenthe subjectof severalrecentreviews(20, 94, 95). The enzymesshowconsiderablediversity of function,but they areclearly relatedon thebasisofconserved aminoacidsequences.

The reactionsknown to be catalyzedby 2-oxoglutarate–dependentdioxy-genasesareshownasa networkof pathwaysin Figure2. Although theywereoriginally delineatedin developingseeds(38), both 13-hydroxylation andnon-13-hydroxylation pathwayshave now beendemonstrated in vegetativetissues(41, 60). The individual stepsbetweenGA12-aldehydeand GA3 andGA8 weredemonstratedin intact maizeshootsby applyingeachisotopicallylabeledintermediateandidentifying its immediatemetabolite by GC-MS(30,60).Both pathwayswere observedin a cell-freesystemfrom embryos/scutellaof two-day-oldgerminatingbarleygrain (41), although GA4 wasnot identi-fied. The dioxygenasesin GA biosynthesiswill now be discussedin detail,with particularemphasison neweraspectsnot coveredin thereviewby Lange& Graebe (68).

440 HEDDEN& KAMIYA

7-Oxidase

Oxidationat C-7 from an aldehydeto a carboxylicacid may be catalyzedbyeither dioxygenasesor monooxygenases.Pumpkinendospermcontainsboth7-oxidaseactivities (46), as doesbarley embryos/scutella(41). The dioxy-genaseactivity from pumpkin hasbeenpartially purified andshownto haveavery low pH optimum (72),whereasthemonooxygenaseis mostactiveabovepH 7 (46). Thetwo typesof activity alsodiffer in their substratespecificities;themonooxygenaseis specificfor GA12-aldehyde,whereasthesolubleactiv-ity oxidizesseveralhydroxylatedGA12-aldehydederivatives(46). The pres-enceof both types ofenzymein a single tissue may indicate subcellularcompartmentation of GA-biosynthetic pathways.

13- and12α-Hydroxylases

As for GA 7-oxidase,both dioxygenaseandmonooxygenaseforms of thesehydroxylaseshavebeendescribed.Thesoluble,2-oxoglutarate–dependent13-hydroxylasedetectedin cell-freeextractsfrom spinachleaves(36) is still theonly exampleof adioxygenasewith thisactivity. 13-Hydroxylasesin pumpkinendosperm(46, 69), developing pea embryos (52), and barley embryos/scutella(41) areof the monooxygenasetype. The preferredsubstratefor the13-hydroxylasesis probablyGA12, althoughother GAs are hydroxylatedtosomeextent.GA12-aldehydeis 13-hydroxylatedin embryocell-freesystemsfrom Phaseoluscoccineus(140)andP. vulgaris (128a),indicatingthatGA53-aldehydeis an intermediatein the13-hydroxylationpathwayin thesetissues.In thebarleyembryos/scutella system,GA15 andGA24 were13-hydroxylatedat very low ratescomparedwith GA12, while GA9 wasnot metabolized(41).This resultis similar to thatfoundpreviouslywith microsomes from immaturepeaembryos(52), but in this caseGA15 and GA9 were hydroxylatedonlywhentheir lactoneswereopenedby hydrolysis.Although“late” 13-hydroxyla-tion (on GA9 or GA4) canoften be demonstrated,it may be relatively ineffi-cient and accompaniedby hydroxylation at other positions on the C and Drings (56). However, in somespecies,such as Picea abies (84), it wouldappear to be themajorpathway.

Both forms of the 12α-hydroxylases arepresent in pumpkinseed, themonooxygenasehydroxylating GA12-aldehyde(GA12 is not a substrate)(46),whereasthe dioxygenaseusesa variety of GA tricarboxylic acid substrates(69, 70). The monooxygenasehasa low pH optimum andmay thuscatalyzepart of the same pathway as the soluble 7-oxidase, for which 12α-hy-droxyGA12-aldehydeis a substrate.Thesoluble12α-hydroxylaseis sensitiveto the presenceof phosphate(69) andwas,therefore,undetectedin previous


studiesin which phosphate,ratherthan Tris, buffer was usedto extract theenzymes(45). BecausephosphateremovesFe by precipitation, it would ap-pear that the12α-hydroxylasehas an unusually highrequirementfor Fe2+.

20-Oxidases

Formationof theC19-GA skeletonrequiressuccessiveoxidationof C-20froma methylgroup,asin GA12 or GA53, throughthealcoholandaldehyde,fromwhich this C atom is lost as CO2 (Figure 2). As discussedbelow, a singleenzyme(GA 20-oxidase)can catalyzethis reactionsequence,althoughthenumberof enzymesthatareactuallyinvolvedin vivo is unknown.A 2-oxoglu-tarate-dependentdioxygenasethat convertedGA53 to GA44 and GA19 waspartially purified from 20-day-olddevelopingembryosof P. sativum (67).Although the proportionof the two productsremainedconstantthroughoutpurification, it was uncertainwhethera single enzymecatalyzedboth steps.Clear evidencethat GA 20-oxidasesare multifunctional was obtainedafterpurification of the enzymeto homogeneity from pumpkin endosperm(66).The enzymeconvertedGA12 to GA15, GA24 andGA25, andGA53 to GA44,GA19 and GA17, with a small amountof putative GA20 producedat highproteinconcentrations.GA12 wasconvertedmoreefficiently thanwasGA53.Theproductionof thetricarboxylicacids,GA25 andGA17, is characteristicofthepumpkin endospermcell-freesystem (69)andindicatesthatthe20-oxidasein this tissueis functionallydifferent from that encounteredin othersystems,which producepredominantly C19-GAs (52, 128a).

Thepurificationof theGA 20-oxidasefrom pumpkinwas quickly followedby the cloning of a cDNA that encodedthis enzyme(71). The cDNA wasselectedfrom an expressionlibrary, derivedfrom developingembryos,usingantiserumagainsta peptidesequencefrom the purified protein. As well asbinding the antibodies,the expressedfusion protein was functionally activeandcatalyzedthesamereactionsasthenativeenzyme.Thealdehydeinterme-diate,GA24, was convertedto both GA25 and GA9, althoughthe latter wasobtainedin lessthan1% yield. The presenceof hydroxyl groupsreducedtheefficiencyof conversion,by about50%in thecaseof GA19 (13-hydroxylated)and 95% for GA23 (3β, 13-dihydroxylated),althoughsmall amounts of theC19-GA productweredetectedin eachcase.Thederivedaminoacidsequencecorrespondsto a proteinof 43.3kDa, which is closeto that estimatedfor thenative enzymefrom gel filtration(44kDa),andcontains theconservedregionsfound inotherplantdioxygenases.

Thecloningof theGA 20-oxidasecDNA from pumpkin seedsenabledtheisolationof homologousclonesfrom otherspecies.The first examples,from

442 HEDDEN& KAMIYA

Arabidopsis,wereobtainedindependently in two laboratories(92, 146).TwoGA 20-oxidase cDNAs werecloned from thega1-3mutantutilizingPCR withdegenerate primersdesignedfrom the conservedaminoacidsequences; athirdcDNA clonewasfoundafterscrutinyof theDataBaseof ExpressedSequenceTags(92). Confirmation that all threecDNAs encodedGA 20-oxidaseswasobtainedby demonstratingthat the productsof heterologousexpressionin E.coli convertedGA12 to GA9 and GA53 to GA20, with GA12 the preferredsubstrate.Small amountsof the tricarboxylic acids,GA25 andGA17, respec-tively, wereformed but, in contrast tothepumpkin enzyme,theC19-GAs werethe major products.Thus,theseenzymesappearedto be involved in the bio-synthesisof activeGAs. It wasconfirmedthatat leastoneof the isozymesisactivein vivo whena genomic cloneencodingoneof theGA 20-oxidaseswasisolatedfrom Arabidopsisby probing a genomiclibrary with the pumpkin20-oxidasecDNA (146).The clonemappedtightly to theGA5 locus,mutationof which resultsin semidwarfism(62) anda reductionin theconcentrationsofC19-GAs (129).Expressionof theGA 20-oxidasegenesis tissue-specific,withtranscriptsdetected,respectively,in stems/floral apices,floral apices/siliques,and siliques (92). The silique-specific20-oxidasetranscript is much moreabundantthan the others.The stem-specificgenecorrespondsto the GA5locus,mutationof which,in ga5,is dueto aG to A substitution thatintroducesa prematurestopcodon(146). Although the mutantproteinwould be highlytruncatedandunlikely to becatalytically active,thega5plant is semidwarfedandcontainslow amountsof C19-GAs (129). It mustbe assumed,therefore,that other GA 20-oxidases,suchas that expressedin the floral apex,supplyGAs tothe stem.

Gibberellin 20-oxidasecDNAs havebeenclonedfrom at leastsevenspe-cies,with multiple genesfound in severalof them.Their encodedaminoacidsequencessharea relatively low degreeof sequenceconservation,with aminoacid identitiesrangingfrom 50–75%.Therelationshipbetweenthesequencesis shownin Figure3B. With theexceptionof theenzymefrom pumpkinseed,the proteinshave very similar functions, converting20-methyl GAs to thecorrespondingC19 lactones. Itis notablethattheenzymeclonedfrom develop-ing cotyledonsof Marah macrocarpus(J MacMillan & DA Ward, unpub-lished information) produces C19-GAs despiteits being mostclosely related tothe pumpkin enzymeon the basisof sequence(Figure 3B). The structuraldifferencesthat determinewhetherthe 20-oxo intermediatesare oxidized toC19-GAs orto tricarboxylic acids arelikely to besubtle. The20-oxidases frompumpkin, Marah, and Arabidopsisprefer nonhydroxylated substratesto the13-hydroxylatedanalogues.This is consistentwith thetypesof GAs found inthe tissues in which theseenzymesarepresent.For example,GA4, which is


not13-hydroxylated,is the majorGA in Arabidopsisshoots(129). In contrast,a GA 20-oxidaseclonedfrom shootsof rice, in which 13-hydroxyC20-GAsare the predominantforms (61), oxidizesGA53 moreefficiently than it doesGA12 (137).

Thereis evidencefor thepresencein shoottissuesof GA 20-oxidaseswithpropertiesdifferent from thosethat havebeenclonedso far. GA44 oxidaseactivity from spinachleaveswas separatedby anion-exchangechromatogra-phy from GA53 oxidaseand GA19 oxidaseactivities, which co-eluted(35).Theselasttwo activitiesareinducedby transferof plantsto longdays,whereasGA44 oxidaseactivity is not photoperiod-sensitive (36). The spinachGA44oxidaseconvertsthelactoneform of thisGA (36),asdocell-freesystemsfrompeashoots(38) and germinatingbarley embryos(41), whereasGA 20-oxi-dasesfrom immatureseedsrequirea free alcohol at C-20 for oxidation tooccur(45, 52, 128a).Detailedstudieswith recombinantGA 20-oxidase,pro-ducedby expressionof oneof theArabidopsiscDNAs in E. coli, revealed thatthis enzymealso requireda free alcohol function and that oxidation of thealcoholwasmuchslowerthanthatof themethylandaldehydesubstrates(47).It seemslikely, therefore,thata separateenzyme(s)with a highaffinity for the20-alcohols,perhapsasthe lactones,existsin shoottissues.An enzymewithsimilar propertiesto the ArabidopsisGA 20-oxidasehasbeencloned fromspinachleaves(145).Onthebasisof theactivity of theproteinafterexpressionin E. coli andits higherexpressionin long daysthanin shortdays,it wouldappearto correspondto theGA53 andGA19 oxidases that wereobserved intheleaf homogenates.

An unexpecteddifferencebetweenthe spinachGA44 oxidaseand the re-combinantArabidopsisGA 20-oxidaseis in the stereospecificremovalof ahydrogenatom during oxidation of the C-20 alcohol intermediates.It wasshown,using GA15 or GA44 labeled stereospecificallywith deuterium, thattheArabidopsisenzymeremovesthe pro-R H atom on conversionof the freealcohol to the aldehyde(142). In contrast,GA44, as the lactone,is oxidizedwith lossof thepro-SH, by cell-freeextractsof spinachleaves.This observa-tion providesfurther evidencefor the existenceof a distinct lactoneoxidase;the different stereochemistry of the reactionsis presumablydue to the fixedorientationof C-20 in the lactoneasopposedto it assumingan energeticallymore favored conformationas the free alcohol.

3β-Hydroxylases and Related Enzymes

3β-Hydroxylationresultsin theconversionof theC19-GAs GA20 andGA9 toGA1 andGA4, respectively,in the final stepin the formationof physiologi-

444 HEDDEN& KAMIYA

cally activeGAs. Thereis now increasingevidencethat,in commonwith GA20-oxidases,certainGA 3β-hydroxylasesmaybemultifunctional.An enzymepurified from developingembryosof P. vulgaris catalyzed2,3-desaturationand 2β-hydroxylation reactions,in addition to 3β-hydroxylation(115, 116).GA20 andGA9 wereaboutequally reactiveassubstrates.A 3β-hydroxylasefrom thesamesourcealsoepoxidizedGA5 to GA6 (65).Theenzymecouldusenon-2β-hydroxylatedC19 (γ-lactone)GAs or 19–20 δ- lactoneC20-GAs assubstrates(65), the latter presumablyacting as structuralanaloguesof theformer, which are the natural substrates.An enzymethat 3β-hydroxylatedGA15 to give GA37 waspartially purified from pumpkin endosperm(72). Itdid not possessdesaturase,althoughit was nottestedwith C19-GAs.

ThepumpkinGA 3β-hydroxylasehasthetypical propertiesof a 2-oxoglu-tarate-dependentdioxygenase(72). In particular, it was possible to demon-stratea1:1stoichiometrybetweentheformationof hydroxyGA andsuccinate,onceuncoupledoxidation of 2-oxoglutaratewas subtracted.In contrast,al-thoughtheP. vulgarisenzymerequires2-oxoglutaratefor activity, Smithet al(116) could find no evidencethat this compoundfunctionedas a substrate.They suggestedthat ascorbatemay serveasthe cosubstrate,asit doesin therelatedenzyme,1-aminocyclopropane-1-carboxylicacid (ACC) oxidase(22).However,2-oxoglutarateis essentialfor full 3β-hydroxylaseactivity, whereasit servesno function for ACC oxidaseactivity. The natureof the P. vulgaris3β-hydroxylaseis unresolved.

The desaturaseactivity of 3β-hydroxylasesprovidesthe first step in theproductionof GA3. After applying[3H]GA5 to immatureseedsof apricot,deBottini et al (19) obtainedchromatographicevidencefor thepresenceof GA1,GA3, andGA6 in theproducts.Unequivocalevidencefor conversionof GA20to GA3 via GA5 wasobtainedin shootsof Zeamays(30), thusestablishing anew biosynthetic pathway.There was no indicationthat GA5 was convertedtoGA1, a reductionthat is without precedentin GA biosynthesis.Theequivalentpathwayfor non-13-hydroxylatedGAs wasdemonstratedin cell-freesystemsfrom immatureseedsof Marah andapple(3). Enzymeactivity presentin bothendospermand developingembryosof Marah results in the conversionofGA9 to GA4 and2,3-dehydroGA9, which is furtheroxidizedto GA7 (75a).TheMarah andapplesystems havea markedpreferencefor non-13-hydroxylatedsubstrates;althoughbothsystemsconvertedGA5 to GA3, GA20 wasmetabo-lized to GA1 (3β-hydroxylation),GA29 (2β-hydroxylation),andGA60 (1β-hy-droxylation),but not to GA5, by theMarah systemandwasunmetabolizedbytheapplepreparation.The branchpathwayfrom GA20 to GA3 occursalsoinbarleyembryos(41) andmaybecommon,althoughnot ubiquitous,in higherplants.The conversionof GA5 to GA3 has also beendemonstratedin pea


shoots(93) andin a cell-freesystemfrom rice anthers(58), althoughbecauseGA5 is not formed in thesesystems(50, 57), the function of this activity isunclear.

The conversionsof GA5 to GA3, and of 2,3-didehydroGA9 to GA7, areunusualreactionsthatareinitiatedby lossof the1β-H (3). Hydrogenabstrac-tion is accompanied by rearrangement ofthe 2,3 double bondto the 1,2position andhydroxylation onC-3β. Thisenzymatic activitymayalsoresult inthe 1β-hydroxylation of GA20 and GA5, also observedin Marah (3). TheenzymethatconvertsGA5 to GA3 requires2-oxoglutarate,butnotaddedFe2+,for activity and, in contrastwith most other relateddioxygenases,it is notinhibitedby iron chelators(116).Its activity, however,is reducedby Mn2+ andother metal ions, the inhibition by Mn2+ being reversedby Fe2+. It wouldappear that Fe isboundvery tightly at theactive site.

GA3 andGA1 formation in maizeis apparentlycatalyzedby oneenzyme(122). As well as affecting the conversionof GA20 to GA1, the dwarf-1mutation reducesformation of GA5 and the conversionof GA5 to GA3. IfDwarf-1 is a structuralgene, a single enzyme mustcatalyzeall three reactions.In contrast,the le (3β-hydroxylation)mutationof peawas purportednot toaffecttheconversion of GA5 to GA3 (93). Thisconclusionwas based onequalgrowth responsesof Le and le plantsto appliedGA5, which wasassumedtohaveno intrinsic biological activity. However,GA5 is asactiveasGA1 andGA3 ondwarf-1maizeshoots,despite no metabolism toGA3 by thisgenotype(122).

Thega4 mutation of Arabidopsisresultsin low amountsof GA1 andGA4andanaccumulation ofGA20 andGA9 in flowering shoots, indicating reduced3β-hydroxylaseactivity (129). This conclusionwas supportedby an 85%reductionin theamountof GA1 producedfrom labeledGA20 in ga4seedlingscompared withthoseof Landsbergerectaor thega5(20-oxidase) mutant(55).TheGA4 locushasbeenclonedby T-DNA insertion(14).Thegeneencodesadioxygenasethathasrelatively low aminoacidsequenceidentity with theGA20-oxidases;it has30% identity (50% similarity) with the Arabidopsisstem-specific20-oxidase(GA5). Expressionof the ga4 cDNA in E. coli hascon-firmed thatit encodesa 3β-hydroxylase(J Williams,AL Phillips& P Hedden,unpublished information).The preferredsubstrate for therecombinantenzymeis GA9, for which theKm is tenfold lowerthanthatfor GA20. Thus, as with theArabidopsisGA 20-oxidases,the presenceof a 13-hydroxyl group reducessubstrateaffinity for theenzyme.Theenzymealsoepoxidizesthe2,3-doublebondin GA5 and 2,3-didehydroGA9, andhydroxylatescertainC20-GAs,albeitwith low efficiency. This activity could accountfor the presenceof 3β-hy-droxy C20-GAs in Arabidopsis(129). However,thereareundoubtedlyother

446 HEDDEN& KAMIYA

GA 3β-hydroxylasesactivein this species.The original EMS-inducedmuta-tion (ga4-1) resultsin semidwarfism(62),anda reductionto about30%of thenormalcontentof 3β-hydroxy GAs (129).The mutantenzymehasan aminoacid substitution, cysteineto tyrosine (14), that might allow a low level ofactivity. The mutantwith the T-DNA insertion(ga4-2) is also a semidwarf,phenotypically similar to ga4-1,with very littl e likelihood of themutantgeneencodingan active 3β-hydroxylase. Residual growthin this mutant must,therefore, resultfrom theaction ofother enzymes.

2β-Hydroxylases and Related Enzymes

Hydroxylation on C-2β resultsin the formation of inactiveproductsand is,therefore,importantfor turnoverof thephysiologically activeGAs. Thenatu-ral substratesfor theseenzymesarenormallyC19-GAs, although 2β-hydroxyC20-GAs arealsofound in plant tissues,particularlywherethe concentrationof C20-GAs is high (76). 2β-Hydroxylaseshavebeenpartially purified fromcotyledonsof P. sativum(118) and Phaseolusvulgaris (39, 117). There isevidencethat, for both sources,at leasttwo enzymeswith different substratespecificities arepresent.Two activitiesfrom cotyledonsof imbibedP. vulgarisseedswere separableby cation-exchangechromatographyand gel-filtration(39). The majoractivity, correspondingto an enzyme ofMr 26,000by size-ex-clusionHPLC, hydroxylatedGA1 andGA4 in preferenceto GA9 andGA20,while GA9 was the preferred substratefor the secondenzyme (Mr 42,000).

Formationof 2-keto derivatives(GA catabolites) by further oxidation of2β-hydroxy GAs (Figure 2) occursin severalspecies,but it is particularlyprevalentin developing seeds(119)androots(108)of pea.TheconversionofGA29 to GA29-catabolitein peaseedswasinhibitedby prohexadione-calcium,aninhibitor of 2-oxoglutarate–dependentdioxygenase(87), indicatingthatthereactionis catalyzedby an enzymeof this type (108). Although the slender(sln) mutation of peablocksboththeconversionof GA20 to GA29 andof GA29to GA29-catabolitein seeds,the inability of unlabeledGA20 to inhibit oxida-tion of radiolabeledGA29, andviceversa,indicatedthatthestepsarecatalyzedby separate enzymes(108). Furthermore, in shoot tissues,the slendermutationinhibits 2β-hydroxylation of GA20, but not the formationof GA29-catabolite.It was,therefore, proposedthatSlnencodes a regulatoryprotein(108). Forma-tion of theGA catabolitescouldbeinitiatedby oxidation eitheratC-1or C-2α.Althoughthereactionsequenceis unknown,it is of interestthat,afterapplica-tion of labeledGA20 to leavesof pea, labeledGA81 (2α-hydroxy GA20)accumulates in theroots,together with thecatabolitesof GA8 andGA29 (108).


GA81 is not formedfrom GA29 andmust, therefore,be formeddirectly asaresultof 2α-hydroxylaseactivity.

REGULATION OF GABIOSYNTHESIS

The role of GAs as mediatorsof environmental stimuli is well established.Factors,suchasphotoperiodandtemperature,canmodify GA metabolism bychangingtheflux throughspecificstepsin thepathway.More recentwork hasshownthatGA biosynthesis is modifiedby theactionof GA itself in a typeoffeedbackregulation.The mechanismsunderlyingtheseregulatoryprocessescannow be investigated asa resultof the currentadvancesin the molecularbiologyof GA biosynthesis.

FeedbackRegulation

Thepresenceof abnormallyhigh concentrationsof C19-GAs incertain GA-in-sensitivedwarf mutants, suchasRht3wheat(5), Dwarf-8 maize(29), andgaiArabidopsis(130), indicate a link betweenGA action and biosynthesis. Inmaize,thereis agene-dosageeffect,with a60-fold increasein theGA1 contentof homozygous Dwarf-8 shoots,comparedwith wild-type and a 33-fold in-creasein the heterozygote.It was suggestedthat GA action results in theproductionof a transcriptional repressorthat limits theexpressionof GA-bio-syntheticenzymes(111). Mutantswith impaired responseto GA would lackthis repressorandhaveelevatedratesof GA production.Suchplantsnormallycontainsemidominantmutations,which mayresultin a gainof function(91).It has beenproposedthat GA 20-oxidaseis a primary target for feedbackregulation(5, 18, 44). In additionto an elevatedC19-GA content,theGA-in-sensitivedwarfs often contain lower amountsof C20-GAs than their corre-sponding wild-types, suggesting increasedGA 20-oxidaseactivity. Con-versely,overgrowthmutants, suchasslender(sln) barleyandla crys pea,thatgrow asif saturatedwith GA, evenin its absence,containreducedamountsofC19-GAs and elevatedcontentof C20-GA GAs (18, 77). It appearsthat theslendermutation activatesthe GA signal transduction pathway,evenin theabsenceof GA, and may therebycauseconstitutive downregulation of GA20-oxidaseactivity.

Further support for feedbackregulationof GA 20-oxidaseactivity wasprovidedby work with GA-biosynthesismutants. ReducedconcentrationsofC20-GAs, as well as a highly elevatedGA20 content,are characteristicsofmost3β-hydroxylase-deficientmutants, including dwarf-1 maize(28), le pea(96), and l sweetpea(109).Treatmentof the maize(44) or pea(77) mutantswith 2,2-dimethylGA4, a syntheticand highly bio-active GA, restoredthe

448 HEDDEN& KAMIYA

concentrationsof GA20 andGA19 to thoseof wild types.Although thechangein GA contentsin the foregoingexamplesis associatedwith alteredgrowthrates,the effectof GA actionon GA metabolism is not a consequenceof thechangein growthrate.For example,therearenumerousGA-insensitive dwarfmutantswith normalGA contents,in which themutation, normally recessive,is likely to affect processesthat arenot part of the primary responseto GAs(99, 105).

With theavailability of GA 20-oxidasecDNA clones,it hasbeenpossibleto begina molecularanalysisof thefeedbackmechanism. Transcript levels foreachof the threeArabidopsisGA 20-oxidasegenesare much higher in thega1-3(CPS-deficient)mutantthanin Landsbergerectaandarevery substan-tially reducedby treating the mutantwith GA3 (92). The reductionoccurswithin 1–3 h, long beforea growth responseis discernible(AL Phillips, DValero& P Hedden, unpublishedinformation),confirming that itis not relatedto growth-rateand indicating thatthemessageis turnedoverrapidly. Thelevelof mRNA for the stem-specificGA 20-oxidaseis also higher in the ga5(GA-deficient)andgai (GA-insensitive) mutantsthanin the wild type (146).Treatmentof ga5and,to a lesserextent,wild-type with GA4, causeda reduc-tion in GA 20-oxidasetranscriptlevels,whereastreatmentof gai resultedin aslight increasein 20-oxidasemRNA. Strongdownregulationof GA 20-oxi-dase transcriptlevelsby GA hasalso beenobserved inpea (77)and rice(137).Low endogenousGA concentration,as in mutantsor after treatmentwith abiosynthesis inhibitor, consistentlyresultedin increasedmRNA levels.Con-versely,theselevelsweresubstantially reduced by application of GA. Further-more,leavesof theslender(la crys) peamutantcontainedonly smallamountsof 20-oxidasetranscript,consistent with strongdownregulationof geneex-pression(77).

Other enzymesin the pathway,particularly the GA 3β-hydroxylase,mayalsobesubjectto feedbackregulation.Treatmentof seedlingsof theGA-defi-cientna mutantof peawith 2,2-dimethylGA4 causeda slight reductionin theconversionof GA19 to GA20, but a muchgreaterreductionin GA20 metabo-lism (77).Furthermore, Chianget al(14) found much moreGA4(3β-hydroxy-lase)transcriptin rosetteleavesof ga4mutants thanin Landsbergerecta,andthat treatmentwith GA3 reducedthe amountof transcriptwithin 8 h. It ispossiblethatseveral enzymesare subjectto regulationby GA; the identities ofothers mayemerge whenmore enzymesof the pathwayhave been cloned.

RegulationbyLight

The involvementof GAs in the photoperiod-inducedbolting of long-dayro-setteplantsis well documented(38). Transferof Silenearmeria plantsfrom


shortdays(SD) to long days(LD) causestheGA1 contentto increaseseveral-fold, particularly in the subapicalregion, with a decreasein GA53 contentconsistentwith increasedGA53 metabolism (131,133,134). In spinach(Spi-nacia oleracea), changesin GA concentrations(135) andenzymeactivity incell-freesystems(36) on transferfrom SD to LD areconsistentwith enhancedoxidationof GA53 andGA19 in LD. Theactivitiesof GA53 andGA19 20-oxi-dases,now known to be the sameenzyme(145), increasein the light anddecreasein the dark (36). Furthermore,therearehigheramountsof GA 20-oxidasemRNA in plantsgrown in LD thanthosein SD or in total darkness(145). It hasbeensuggestedthat, in LD, there is sufficient GA 20-oxidaseactivity to raisethe GA1 concentrationabovethe thresholdrequiredfor stemextension(135). In fact, light appearsto increasethe total flux throughthepathway,becauseent-kaurene synthesis is alsoenhancedin LD in spinachandin Agrostemmagithago (149).Although GA53 20-oxidaseactivity is regulatedby light, oxidation of GA44, in the lactoneform, remainsat high, constantlevels irrespectiveof light or dark treatment(36). As discussedearlier, thislatteractivity is probablybecauseof anotherenzyme,which is not underlightregulation.

Despitemanyattempts to implicateGA metabolismin phytochrome-medi-atedchangesin growth rate,supportingevidenceis sparse.EnhancementofGA20 3β-hydroxylation by far-red light has beenobservedin lettuce (138,139) and cowpeaepicotyls (25, 31, 78, 79). In the latter case,higher GA1concentrations in plants grown in far-redlight wereduealsoto reduced2β-hy-droxylationandwereaccompaniedby heightenedtissueresponsivenessto GA(78, 80). However, in peas,enhancedshoot elongationby treatmentwithfar-red-richlight wasnot associatedwith increasedGA1 content(100).Thereis alsono evidenceto suggestthat dark-grownpeas(34, 120, 143) or sweetpeas(109) contain more GA1 than light-grown plants. In fact, work withphytochrome-andGA-deficientmutantsof peaindicatesthat growth inhibi-tion by red light, which is mediatedby phytochrome B, is due to alteredresponsivenessto GA, rather than to changesin the concentrationof GA1(143).Severalphytochrome-deficientmutants,suchastheeinmutantof Bras-sicarapa (21) andthema3

R mutantof Sorghum(15,16), haveanovergrowthphenotypeandwereoriginally thoughtto containabnormallyhigh GA levels.Although the GA1 contentof theseplantsmay be elevated(7, 104), this isapparentlynot the causeof their alteredphenotype. PhytochromeB-deficientmutantsof cucumber(74)andpea(143)containcomparableamountsof activeGAs to the wild types,but showan enhancedresponseto GAs. In Sorghum,alteredGA contentis dueto a shift in thephaseof a diurnal fluctuationin GAconcentrations(26). It was proposedthat phytochromedeficiencydisrupted

450 HEDDEN& KAMIYA

diurnalregulationof theconversionof GA19 to GA20, resultingin a 12-hshiftin thepeaksof GA20 andGA1 concentrations,whereasthepatternof fluctua-tion in the levelsof GA12 andGA53 was unaffected.

Gibberellin metabolism is sensitive to light quantity. Whenpeaseedlingswere grown in low irradiance(40 µmol • m−2 • s−1), GA20 concentrationincreasedsevenfoldcomparedwith plantsgrownin high irradiance(386µmol• m−2 • s−1) (34), whereasin plantsgrown in thedark, theGA20 contentwasreducedto 25% of that in high irradiance.Moreover, the responseof theseedlingto exogenousGA1 washeightenedin thedark.Theseresultsindicatethat therateof GA 20-oxidation is sensitive to light fluence;with thecloningof GA 20-oxidasesnowreportedfor manyspecies,themechanismsunderlyingthis processas well as the diurnal fluctuation in GA biosynthesiscan beprobed.

RegulationbyTemperature

Induction of seedgermination(stratification) or of flowering (vernalization)by exposureto low temperaturesare processesin which GAs have beenimplicated.There are, however,few examplesin which GAs have been shownunequivocally tomediatethetemperaturestimulus.Themostextensively stud-ied systemis Thlaspi arvense,in which stem extensionand flowering areinduced by exposureto low temperaturesfollowed by a returnto highertemperatures(82). Thesameeffectcanbeobtainedwithout cold inductionbyapplicationof GAs, themostactiveof thosetestedbeingGA9 (83). In nonin-ducedplants,ent-kaurenoicacid accumulatesto high concentrationsin theshoottip, the site of perceptionof thecold stimulus,whereasaftervernaliza-tion andreturnto high temperatures,the level of this intermediatefalls withindaysto relativelylow values(43).Metabolismof labeledent-kaurenoicacid toGA9 could be demonstratedin thermo-inducedshoottips, but not in nonin-ducedmaterial(42). Furthermore,microsomesfrom inducedshootsmetabo-lized ent-kaurenoicacid and ent-kaurene,but microsomesfrom noninducedshootsweremuchlessactivefor both activities(43). Leaves,or microsomesextractedfrom leaves,from thermo-induced and noninducedplantsmetabo-lized ent-kaurenoicacid to the sameextent.Theseresultsareconsistentwithregulationof ent-kaurenoicacid 7β-hydroxylaseand,to a lesserdegree,ent-kaurene oxidaseby coldtreatmentin shoottips of Thlaspi.

A changein GA contentfollowing vernalizationwasfoundin shoottips ofBrassicanapus,in which GAs, including GA1 andGA3, accumulatedduringthecoldperiod(148).Thehigherratesof GA productionin vernalized,relativeto nonvernalized,plantsappearedto persistfor 1–2 weeksafter the returnto


high temperatures.In contrastto thesefindings, there was no evidencetosuggestthat GAs were the signal for cold-inducedflowering in Raphanussativus(88) orTulipagesneriana(97).

Although the mechanismfor thermo-induction of GA biosynthesisis notyetknown,it hasbeensuggestedthat cold treatment may allow increasedratesof geneexpression,possibly via demethylation of the promoters (11). Somecircumstantial support forthis theorywasobtainedby reducingDNA methyla-tion in Thlaspiandlate-floweringecotypesof Arabidopsisby treatmentwith5-azacytidine.Flowering times in noninduced plants were reducedin bothcases. Confi rmation of this theory must await the isolation and charac-terizationof the7β-hydroxylasegene.

SITES OF GABIOSYNTHESIS

Developmentalregulation of GA biosynthesis is determinedmainly bychangesin plant ontogenyduring developmentandthe tissuedistribution ofindividual enzymesof thepathway.Work with legumesindicatesthatGA-bio-synthesisoccursmainly in activelygrowingtissues,with leavesandinternodesimportantsites(113, 114). Gibberellin20-oxidasetranscriptlevelsaremuchhigher in pea leavesthan in internodes(32). The same20-oxidasegeneisexpressedin shoots, young seeds,and expandingpods(32), but a differentgeneis expressedin developingcotyledons(73). Orthologuesof the peaGA20-oxidases,with similar patternsof gene expression,were cloned fromFrenchbean(32), anda third gene,which is expressedin developingcotyle-dons,leaves, androots,was also detectedin this species.

The tissue-specificexpressionof the GA 20-oxidasegenes,alsonotedinArabidopsis(92), has not beenfound for CPS (2, 124), KS (147), or GA3β-hydroxylase(14). Whereasdifferent GA 20-oxidasegenesare expressedduring peaseeddevelopment,the sameCPSgeneis expressedin a biphasicmanner,correspondingwith the two stagesof GA production(2). The firstphase,whichresultsin theproduction ofGA1 andGA3 (33,103),is associatedwith seeddevelopment(126–128)andpodgrowth(33,103).Both endospermandtestaarepotential sitesof synthesis (103).Thesecondphaseoccursin thedevelopingembryoand hasno known physiological function. As seedsap-proachmaturity thereis oftenanincreasein 2-oxidation activitiesin embryosandtesta(4, 119).Thephysiological significanceof this deactivationmecha-nism is vividly demonstratedby the slender(sln) mutantof pea,which accu-mulateshigh concentrationsof GA20 in the matureseeddue to a lack of2-oxidaseactivities(102,107,108).On germination, theGA20 is 3β-hydroxy-latedto GA1, resultingin overgrowthof the first 10–12internodes(102).An

452 HEDDEN& KAMIYA

intriguing phenotypewasobtainedby crossingslenderwith na,aGA-deficientdwarf, blockedat an intermediate stepin the biosynthetic pathway,but notexpressedin the developingseed(49). The doublemutantis phenotypicallyslenderto thesix-leafstage,but thereafteris severelydwarfed(108).Crossingslenderwith le (3β-hydroxylase-deficient) produceda dwarf from emergence,asdid crossingwith lhi, which blocksat ent-kaureneoxidasein seeds(127)and preventsthe accumulation of GA20 (108).

Gibberellin productionin the pericarp of developing fruit may be importantfor fruit growth, particularly in the absenceof seeds(132). Work with peaindicatesthatthepresenceof seedsstimulatesGA biosynthesisin thepericarp(90, 141).Both seedsand4-chloroindole-3-aceticacid,theproposedseed-de-rived signal,stimulated the conversionof GA19 to GA20 in peapods.How-ever,the finding thatGA 20-oxidasetranscriptlevelsin pericarpfrom seededpeafruit aremuchlower thanin seedlessfruit (32) doesnot supporta regula-tory rolefor GA 20-oxidation in fruit developmentin this species.

CONCLUDING REMARKS

Researchon GA biosynthesis has reachedan exciting stage.cDNA andgenomiccloneshavebeenobtainedfor five typesof enzyme,including mem-bersof eachof the threeenzymaticclasses:cyclases,monooxygenases,anddioxygenases.Thecloningof eachenzymein thepathwayshouldbeachievedwithin the next three to five years.The availability of clones is enablingsignificant advancesin severaldirections.Enzymescanbe preparedby het-erologousexpressionin sufficientquantitiesfor detailedstudies on theirstruc-tureand function, andfor theproduction ofantibodies.Promoter-reportergenefusions,in situ hybridization,and immunolocalizationcan be usedto deter-mine cellular andsubcellularsitesof synthesis. It is possible to examinetheregulationof individual enzymesat the transcriptand protein levels. Suchstudiesare alreadyyielding important information on the regulationof GA20-oxidasetranscriptabundanceby GA actionandby light, andon thedevel-opmentalcontrolof CPS andKS gene expression.

From a practicalstandpoint, it will be possibleto manipulate GA produc-tion in transgenicplantsby alteringthe expressionof individual genes.Thistechnologyoffers an alternativeto the useof chemicalgrowth regulatorsforthe control of plant development. It alsoprovidesa meansto alter the abun-danceof specificenzymesandtherebydeterminetheircontributionsto thefluxthroughthe biosynthetic pathway.For example,overexpressionof GA 20-oxi-dasecDNAs in Arabidopsisresultsin acceleratedbolting, confirming that theactivity of this enzymeis rate-limiting for this developmentalprocess(53).


Experimentsof this type may provide new insights into the role of GAs inplantdevelopment.

ACKNOWLEDGMENTS

We thank colleaguesfor their critical readingof the manuscript.IACR re-ceivesgrant-aidedsupportfrom the Biotechnology and Biological SciencesResearch Councilof the UnitedKingdom.

Visit the Annual Reviews home pageat http://www.annurev.org

LiteratureCited

1. Aach H, Böse G, GraebeJE. 1995. ent-Kaurene biosynthesisin acell-freesystemfrom wheat (Trit icum aestivum L.) seed-lings and the localization of ent-kaurenesynthetase in plastids of three species.Planta 197:333–42

2. Ait-Ali T, Swain SM, Reid JB, Sun TP,Kamiya Y. 1997. The Ls locusof peaen-codesthe gibberellin biosynthesisenzymeent-kaurenesynthaseA. PlantJ.77:443–54

3. Albone KS, GaskinP, MacMillanJ, Phin-neyBO,Wil lisCL. 1990. Biosynthetic ori-gin of gibberellin A3 andgibberellin A7 incell-freepreparationsfrom seedsof Marahmacrocarpus andMalus domestica. PlantPhysiol. 94:132–42

4. Albone K, GaskinP, MacMillan J, SmithVA, Weir J. 1989. Enzymesfrom seedsofPhaseolus vulgaris L.: hydroxylation ofgibberellinsA20andA1 and2,3-dehydroge-nation of gibberellin A20. Planta 117:108–15

5. Appleford NEJ, Lenton JR. 1991. Gib-berellins and leaf expansion in near-iso-genic wheatlinescontainingRht1andRht3dwarfingalleles.Planta 183:229–36

6. Bartley GE,Scolnik PA, GiulianoG.1994.Molecular biology of carotenoid biosyn-thesisin plants. Annu. Rev. Plant Physiol.Plant Mol. Biol. 45:287–301

7. BealFD, Morgan PW, ManderLN, Mil lerFR,BabbKH. 1991. Genetic regulation ofdevelopment in Sorghum bicolor V. Thema3

R allele results in gibberellin enrich-ment. Plant Physiol. 95:116–25

8. BensenRJ,Crane VC, Wang X, Duvick J,BriggsSP. 1995. MaizeGA-mutants:isola-tion, characterization,andgenecloningandexpression. In Abstract of InternationalConferenceon Plant Growth Substances.

(Abstr. 096). Minneapolis: Int. PlantGrowth SubstancesAssoc.

9. BensenRJ,Johal GS,CraneVC, TossbergJT, Schnable PS,et al. 1995. Cloning andcharacterization of the Maize An1 gene.Plant Cell 7:75–84

10. BensenRJ,ZeevaartJAD.1990. Compari-sonof ent-kaurenesynthetaseA andB ac-tivitiesin cell-free extracts fromyoung to-mato fruits of wild-type andgib-1, gib-2,andgib-3 tomatoplants. J. Plant GrowthRegul. 9:237–42

11. Burn JE,Bagnell DJ, MetzgerJD, DennisES,PeacockWJ. 1993. DNA methylationandthe initiation of flowering. Proc.Natl.Acad. Sci. USA90:287–91

12. Chamovitz D, MisawaN, Sandmann G,Hirschberg J.1992. Molecularcloning andexpressionin Escherichia coli of a cyano-bacterial gene coding for phytoene syn-thase,a carotenoid biosynthesis enzyme.FEBSLett. 296:305–10

13. Chappell J. 1995. Biochemistry and mo-lecular biology of the isoprenoid biosyn-thetic pathway in plants. Annu. Rev. PlantPhysiol. Plant Mol. Biol. 46:521–47

14. Chiang HH, Hwang I, Goodman HM.1995. Isolationof theArabidopsisGA4 lo-cus.Plant Cell 7:195–201

15. Childs KL, Cordonnier-Pratt MM, PrattLH, MorganPW. 1992. Genetic regulationof development in Sorghum bicolor VII.Thema3

R flowering mutantlacks a phyto-chrome that predominatesin greentissue.Plant Physiol. 99:765–70

16. Childs KL, PrattLH, Morgan PW. 1991.Genetic regulation of development in Sor-ghumbicolor VI: thema3

R alleleresults inabnormal phytochrome physiology. PlantPhysiol. 97:714–19

454 HEDDEN& KAMIYA

17. Colby SM, Alonso WR, Katahira EV,McGarvey DJ, Croteau R. 1993. 4S-Li-monene synthase from the oil glands ofspearmint (Mentha spicata): cDNA isola-tion, characterization, andbacterial expres-sionof thecatalytically activemonoterpenecyclase.J. Biol. Chem.268:23016–24

18. Croker SJ,HeddenP, Lenton JR,StoddartJL. 1990. Comparison of gibberellins innormalandslenderbarleyseedlings.PlantPhysiol. 94:194–200

19. de Bottini GA, Bottini R, Koshioka M,PharisRP, CoombeBG.1987. Metabolismof [3H]gibberellinA5 by immatureseedsofapricot (Prunus armeniaca L.). PlantPhysiol. 83:137–42

20. DeCarolisE, DeLucaV. 1994. 2-Oxoglu-tarate-dependent dioxygenasesandrelatedenzymes: biochemical characterization.Phytochemistry 36:1093–107

21. Devlin PF, Rood SB,Somers DE, Quail P,Whitelam GC. 1992. Photophysiology ofthe elongated internode (ein) mutant ofBrassicarapa.Plant Physiol.100:1442–47

22. DongJG,Fernández-MaculetJC,YangSF.1992. Purifi cation andcharacterization of1-aminocyclopropane-1-carboxylate oxi-dasefromapple fruit.Proc.Natl. Acad. Sci.USA89:9789–93

23. DuncanJD, WestCA. 1981. Propertiesofkaurenesynthetasefrom Marah macrocar-pusendosperm: evidencefor theparticipa-tion of separatebut interacting enzymes.Plant Physiol. 68:1128–34

24. Facchini PJ,Chappell J.1992. Genefamilyfor an elicitor-induced sesquiterpene cy-clasein tobacco.Proc.Natl.Acad.Sci.USA89:11088–92

25. FangN, Bonner BA, Rappaport L. 1991.Phytochromemediationof gibberellin me-tabolism and epicotyl elongation in cowpea,Vigna sinensisL. SeeRef. 128b, pp.280–88

26. Foster KR, Morgan PW. 1995. Geneticregulation of developmentin Sorghumbi-color IX: the ma3

R allele disrupts diurnalcontrol of gibberellin biosynthesis. PlantPhysiol. 108:337–43

27. FrayRG,Wallace A,FraserPD,ValeroD,HeddenP, etal. 1995. Constitutiveexpres-sion of a fruit phytoene synthasegene intransgenic tomatoes causesdwarfism byredirectingmetaboli tesfromthegibberellinpathway. Plant J. 8:693–701

28. FujiokaS,YamaneH, SprayCR,Gaskin P,MacMillan J, et al. 1988. Qualitative andquantitative analyses of gibberellins invegetative shoots of normal, dwarf-1,dwarf-2, dwarf-3 anddwarf-5 seedlings ofZeamaysL. Plant Physiol. 88:1367–72

29. Fujioka S,YamaneH, Spray CR, KatsumiM, PhinneyBO, etal. 1988. Thedominantnongibberellin-responding dwarf mutant(D8) of maizeaccumulatesnativegibberel-lins. Proc. Natl. Acad. Sci. USA 85:9031–35

30. Fujioka S, YamaneH, Spray CR,PhinneyBO,GaskinP, etal.1990. GibberellinA3 isbiosynthesizedfrom gibberellin A20 viagibberellin A5 in shoots of Zea maysL.Plant Physiol. 85:9031–35

31. García-MartínezJL, Keith B, Bonner BA,Stafford A, Rappaport L. 1987. Phyto-chrome regulation of the response to ex-ogenousgibberellins byepicotylsof VignasinensisL. Plant Physiol. 85:212–16

32. García-Martínez JL, López-Díaz I,Sánchez-Beltrán MJ, Phil lips AL, WardDA, et al. 1997. Expression of gibberellin20-oxidasegenesin peaandbeanin rela-tion to fruit development. Plant Mol. Biol.In press

33. García-MartínezJL, SantesC, Croker SJ,HeddenP. 1991. Identification, quantita-tionanddistributionof gibberellinsin fruitsof Pisum sativum cv. Alaska during poddevelopment. Planta 184:53–60

34. GawronskaH,YangYY,FurukawaK, Ken-drick RE,TakahashiN, etal. 1995. Effectsof low irradiancestresson gibberellin lev-els in pea seedlings. Plant Cell Physiol.36:1361–67

35. Gilmour SJ,BleeckerAB, ZeevaartJAD.1987.Partialpurificationof gibberellin oxi-dasesfrom spinachleaves.Plant Physiol.85:87–90

36. Gilmour SJ,ZeevaartJAD, SchwenenL,GraebeJE. 1986. Gibberellin metabolismin cell-free extracts from spinachleavesinrelation to photoperiod. Plant Physiol. 82:190–95

37. GraebeJE.1982. Gibberellin biosynthesisin cell-freesystemsfrom higher plants. InPlant Growth Substances 1982, ed. PFWareing, pp. 71–80. London: Academic

38. GraebeJE.1987. Gibberellin biosynthesisandcontrol. Annu. Rev. Plant Physiol. 38:419–65

39. GriggsDL, HeddenP, LazarusCM. 1991.Partial purifi cation of two gibberellin 2β-hydroxylasesfrom cotyledons of Phaseolus vulgaris. Phytochemistry 30:2507–12

40. GrosselindemannE, GraebeJE,Stöckl D,Hedden P. 1991. ent-Kaurenebiosynthesisin germinating barley (Hordeum vulgareL., cv Himalaya)caryopsesandits relationto α-amylase production. Plant Physiol.96:1099–104

41. Grosselindemann E,Lewis MJ, HeddenP,


GraebeJE.1992. Gibberellin biosynthesisfromgibberellinA12-aldehydein acell-freesystemfrom germinating barley(HordeumvulgareL., cv. Himalaya)embryos.Planta188:252–57

42. Hazebroek JP, Metzger JD. 1990. Ther-moinductive regulation of gibberellin me-tabolism in Thlaspi arvensiL. I. Metabo-lism of [2H]kaurenoic acid and [14C]gib-berellin A12-aldehyde. Plant Physiol. 94:154–65

43. Hazebroek JP, Metzger JD,Mansager ER.1993. Thermoinductive regulation of gib-berellin metabolism in Thlaspi arvensiL.II. Coldinductionof enzymesin gibberellinbiosynthesis.Plant Physiol. 102:547–52

44. HeddenP, Croker SJ.1992. Regulation ofgibberellin biosynthesisin maizeseedlings.See Ref.52a,pp. 534–44

45. Hedden P, GraebeJE. 1982. Cofactor re-quirements for the soluble oxidasesin themetabolism of the C20-gibberellins. J.Plant Growth Regul. 1:105–16

46. HeddenP, GraebeJE,BealeMH, GaskinP,MacMillan J. 1984. The biosynthesisof12α-hydroxylated gibberellins in a cell-free systemfrom Cucurbita maxima en-dosperm.Phytochemistry 23:569–74

47. HeddenP, Phil lipsAL, JacksonGS, ColesJP. 1995. Molecular cloning of gibberellinbiosyntheticenzymes,aroutetothegeneticmanipulation of plant growth. In Proceed-ings of the Ninth Forum for Applied Biotechnology, pp. 1559–66.Ghent: Med. Fac.Landbouww. Univ. Gent

48. HeddenP, PhinneyBO.1979. Comparisonof ent-kaureneandent-isokaurenesynthe-sis in cell-free systems from etiolatedshoots of normal anddwarf-5 maizeseed-lings.Phytochemistry18:1475–79

49. IngramTJ,ReidJB.1987. Internodelengthin Pisum gene na may block gibberellinsynthesis between ent-7α- hy-droxykaurenoic acid andgibberellin A12-aldehyde.Plant Physiol. 83:1048–53

50. IngramTJ, ReidJB, Murfet IC, Gaskin P,Wil lis CL, et al. 1984. Internode length inPisum. The Le gene controls the 3β-hy-droxylation of gibberellin A20 to GA1.Planta 160:455–63

51. Jennings JC, Coolbaugh RC, Nakata DA,WestCA. 1993. Characterization andsolu-bilization of kaurenoic acid hydroxylasefrom Gibberella fujikuroi. Plant Physiol.101:925–30

52. Kamiya Y, GraebeJE. 1983. The biosyn-thesis of all major pea gibberellins in acell-freesystemfrom Pisumsativum.Phy-tochemistry22:681–89

52a. KarssenCM,VanLoonLC,Vreugenhil D,

eds.1992. Progressin Plant GrowthRegu-lation. Dordrecht: Kluwer

53. KatsumiM. 1964. Gibberellin-like activi-tiesof certain auxins andditerpenes.PhDthesis.Univ. Calif., LosAngeles

54. KeegstraK, OlsenLD, Theg SM. 1989.Chloroplasticprecursorsandtheirtransportacrosstheenvelopemembranes.Annu.Rev.Plant Physiol. Plant Mol. Biol. 40:471–501

55. KobayashiM, GaskinP, Spray CR, Phin-neyBO, MacMillanJ. 1994. The metabo-lismof gibberellin A20 to gibberellin A1 bytall anddwarfmutantsof OryzasativaandArabidopsis thaliana. Plant Physiol. 106:1367–72

56. KobayashiM, Gaskin P, SprayCR, SuzukiY,PhinneyBO,etal.1993.Metabolismandbiological activity of gibberellin A4 invegetative shoots of Zeamays,Oryzasa-tiva, and Arabidopsis thaliana. PlantPhysiol. 102:379–86

57. Kobayashi M, Kamiya Y, Sakurai A, SakaH, Takahashi N. 1990. Metabolismof gib-berellins in cell-free extracts of anthersfrom normal and dwarf rice. Plant CellPhysiol. 31:289–93

58. Kobayashi M, Kwak SS, Kamiya Y,Yamane H,Takahashi N, et al. 1991. Con-versionof GA5 toGA6 andGA3 in cell-freesystems from Phaseolus vulgaris andOryza sativa. Agric. Biol. Chem. 55:249–51

59. KobayashiM, Sakurai A, SakaH, Taka-hashiN. 1989. Quantitativeanalysisof en-dogenousgibberellins in normalanddwarfcultivars of rice. Plant Cell Physiol. 30:963–69

60. Kobayashi M, Spray CR, Phinney BO,GaskinP, MacMillan J. 1996. Gibberellinmetabolism in maize—the stepwise con-version of gibberellin A12-aldehydeto gib-berellinA20. Plant Physiol. 110:413–18

61. KobayashiM, Yamaguchi I, MurofushiN,OtaY, Takahashi N. 1988. Fluctuationandlocalization of endogenous gibberellins inrice.Agric. Biol. Chem.52:1189–94

62. KoornneefM, vanderVeenJH. 1980. In-duction and analysis of gibberellin sensi-tive mutants in Arabidopsis thaliana (L.)Heynh. Theor. Appl. Genet.58:257–63

63. KoornneefM, vanEdenJ, Hanhart CJ,deJongh AMM . 1983. Genetic fine-structureof theGA-1locusin thehigherplantArabi-dopsis thaliana (L.) Heynh. Genet. Res.41:57–68

64. Kuntz M, RömerS, Suire C, Hugueney P,Weil JH, et al. 1992. Identification of acDNA for the plastid-locatedgeranylger-anyl pyrophosphate synthase from Cap-

456 HEDDEN& KAMIYA

sicumannuum: correlative increasein en-zyme activity and transcript level duringfruit ripening. Plant J. 2:25–34

65. KwakSS,KamiyaY, SakuraiA, TakahashiN,GraebeJE.1988.Partial purif icationnadcharacterization of gibberellin 3β-hy-droxylase from immature seeds ofPhaseolus vulgaris L. Plant Cell Physiol.29:935–43

66. Lange T. 1994. Purification and partialamino-acidsequenceof gibberellin 20-oxi-dase from Cucurbita maxima L. en-dosperm.Planta 195:108–15

67. Lange T, GraebeJE.1989. The partial pu-rif ication and characterization of a gib-berellin C-20 hydroxylasefrom immaturePisum sativum L. seeds. Planta 179:211–21

68. Lange T, GraebeJE. 1993. Enzymes ofgibberellin synthesis.In Methods in PlantBiochemistry, ed.PJLea,9:403–30. Lon-don:Academic

69. LangeT, HeddenP, GraebeJE.1993. Bio-synthesisof 12α- and13-hydroxylatedgib-berellinsin acell-freesystemfrom Cucur-bita maximaendosperm andthe identif ica-tion of new endogenous gibberellins.Planta 189:340–49

70. LangeT, HeddenP, GraebeJE.1993. Gib-berellin biosynthesis in cell-free extractsfrom developing Cucurbita maxima em-bryos andthe identifi cation of new endo-genous gibberellins.Planta 189:350–59

71. Lange T, HeddenP, GraebeJE.1994. Ex-pression cloning of a gibberellin 20-oxi-dase,amultifunctional enzymeinvolvedingibberellin biosynthesis.Proc. Natl. Acad.Sci. USA91:8522–66

72. LangeT, SchweimerA, WardDA, HeddenP, GraebeJE.1994. Separation andcharac-terization of three2-oxoglutarate-depend-entdioxygenasesfrom Cucurbita maximaL. endosperm involved in gibberellin bio-synthesis.Planta 195:98–107

73. LesterDR, Ross JJ, Ait-Ali T, Martin DN,Reid JB. 1996. A gibberellin 20-oxidasecDNA (AccessionNo. U58830) from pea(Pisum sativum L.) seed(PGR 96-050).Plant Physiol. 111:1353

74. López-JuezE, KobayashiM, Sakurai A,Kamiya Y, Kendrick RE. 1995. Phyto-chrome,gibberellin, andhypocotyl growth.A study using cucumber(CucumissativusL.) long hypocotyl mutant. Plant Physiol.107:131–40

75. MacMillan J. 1997. Biosynthesis of thegibberellin planthormones.Nat.Prod.Rep.In press

75a.MacMillan J, Ward DA, Phill ips AL,Sánchez-BeltránMJ,GaskinP,etal.1997.

Gibberellin biosynthesisfrom gibberellinA12-aldehyde in endosperm and embryosof Marah macrocarpus. Plant Physiol. Inpress

76. Mander LN, OwenDJ, Croker SJ,GaskinP, Hedden P, et al. 1996. Identification ofthreeC20 gibberellins:GA97 (2β-hydroxy-GA53), GA98(2β-hydroxy-GA44andGA99(2β-hydroxy-GA19). Phytochemistry 43:23–28

77. Martin DN, Proebsting WM, ParksTD,DoughertyWG,LangeT, etal.1996. Feed-backregulation of gibberellin biosynthesisandgene expressionin Pisum sativum L.Planta. 200:159–66

78. Martínez-García JF, García-Martínez JL.1992. Phytochrome modulation of gib-berellin metabolism in cow peaepicotyls.See Ref.52a,pp. 585–90

79. Martínez-García JF, García-Martínez JL.1992. Interaction of gibberellins andphy-tochromein thecontrol of cowpeaepicotylelongation. Physiol. Plant. 86:236–44

80. Martínez-García JF, García-Martínez JL.1995. An acylcyclohexanedione retardantinhibitsgibberellin A1 metabolism,therebynull ify ing phytochrome-modulation ofcowpeaepicotyl explants. Physiol. Plant.94:708–14

81. MauCJD,WestCA. 1994. Cloningof cas-benesynthasecDNA: evidence for con-servedstructural featuresamong terpenoidcyclasesin plants. Proc. Natl. Acad. Sci.USA91:8497–501

82. Metzger JD. 1985. Role of gibberellins inthe environmental control of stemgrowthin Thlaspi arvensi L. Plant Physiol. 78:8–13

83. MetzgerJD.1990. Comparisonof biologi-cal activities of gibberellins andgibberel-lin-precursors native to Thlaspi arvensiL.Plant Physiol. 94:151–56

84. Moritz T. 1995. Biological activity, identi-fication andquantif ication of gibberellinsin seedlings of Norway spruce (Piceaabies) grown underdifferentphotoperiods.Physiol. Plant. 95:67–72

85. MurakamiY. 1972. Dwarfinggenesin riceandtheir relation to gibberellin biosynthe-sis.In Plant Growth Substances,1970, ed.DJCarr, pp. 166–74. Berlin: Springer-Ver-lag

86. Murphy PJ, West CA. 1969. The role ofmixed function oxidasesin kaurenemeta-bolismin EchinocystismacrocarpaGreeneendosperm.Arch. Biochem.Biophys.133:395–407

87. NakayamaI, KamiyaY, Kobayashi M, AbeH, Sakurai A. 1990. Effects of a plantgrowthregulator, prohexadione,onthebio-


synthesisof gibberellins in cell-free sys-temsderived from immature seeds.PlantCell Physiol. 31:1183–90

88. NakayamaM, YamaneH, Nojiri H, YokotaT,YamaguchiY,etal.1995.Qualitativeandquantitative analysis of endogenous gib-berellins in Raphanus sativus L. duringcold treatment andthe subsequent growth.Biosci.Biotech. Biochem.59:1121–25

89. OgawaS,ToyomasuT, YamaneH, Muro-fushi N, IkedaR, etal. 1996. A stepin thebiosynthesisof gibberellins that is control-ledby themutation in semi-dwarf ricecul-tivar Tan-ginbozu. Plant Cell Physiol. 37:363–68

90. Ozga JA, Brenner ML, Reinecke DM.1992. Seedeffectson gibberellin metabo-lism in peapericarp.Plant Physiol. 100:88–94

91. PengJ, HarberdNP. 1993. Derivative al-lelesof the Arabidopsis gibberellin-insen-sitive (gai) mutation confer a wild-typephenotype.Plant Cell 5:351–60

92. Phill ipsAL, WardDA, UknesS,ApplefordNEJ, Lange T, et al. 1995. Isolation andexpressionof threegibberellin 20-oxidasecDNA clones from Arabidopsis. PlantPhysiol. 108:1049–57

93. Poole AT, RossJJ,LawrenceNL, ReidJB.1995. Identification of gibberellin A4 inPisumsativumL. andtheeffectsof appliedgibberellins A9, A4, A5 andA3 on the lemutant. Plant Growth Regul. 16:257–62

94. Prescott AG. 1993. A dilemmaof dioxy-genases(or wherebiochemistry and mo-lecular biology fail to meet).J. Exp. Bot.44:849–61

95. Prescott AG, John P. 1996. Dioxygenases:molecular structureandrole in plantmeta-bolism. Annu. Rev. Plant Physiol. PlantMol. Biol. 47:245–71

96. Proebsting WM, Hedden P, Lewis MJ,CrokerSJ,ProebstingN. 1992. Gibberellinconcentrationandtransport in geneticlinesof pea:effectsof grafting. Plant Physiol.100:1354–60

97. RebersM, VermeerE,Knegt E,SheltonCJ,vanderPlasLHW.1995. Gibberellin levelsandcold-induced floral stalkelongation intulip. Physiol. Plant. 94:687–91

98. Reid JB. 1990. Phytohormone mutants inplant research.J. Plant Growth Regul. 9:97–111

99. ReidJB. 1993. Planthormonemutants. J.Plant Growth Regul. 12:207–26

100.ReidJB,HasanO,RossJJ.1990. Internodelength in Pisum.Gibberellins and the re-sponse to far-red-rich light. J. PlantPhysiol. 137:46–52

101.Reid JB, Ross JJ. 1993. A mutant-based

approach, using Pisum sativum, to under-standing plant growth. Int. J. Plant Sci.145:22–34

102.Reid JB, RossJJ,SwainSM. 1992. Inter-node length in Pisum. A new, slendermu-tant with elevatedlevels of C19 gibberel-lins.Planta 188:462–67

103.Rodrigo MJ, García-Martínez JL, SantesCM, GaskinP, HeddenP. 1997. Theroleofgibberellins A1 andA3 in fruit growth ofPisumsativumL. andthe identifi cation ofgibberellins A4 and A7 in young seeds.Planta. In press

104.Rood SB, Wil liamsPH, PearceD, Muro-fushiN, ManderLN, etal. 1990. A mutantgene that increasesgibberellin productionin Brassica. Plant Physiol. 93:1168–74

105.RossJJ.1994.Recentadvancesin thestudyof gibberellin mutants. Plant GrowthRegul. 15:193–206

106.RossJJ,DaviesNW, Reid JB, Murfet IC.1990. Internodelength in Lathyrus odora-tus.Effectsof mutants l andlb ongibberel-lin metabolism andlevels.Physiol. Plant.79:453–58

107.RossJJ,ReidJB,SwainSM.1993. Controlof stemelongation by gibberellin A1: evi-dence from genetic studies including theslendermutantsln. Aust. J. Plant Physiol.20:585–99

108.Ross JJ, Reid JB, Swain SM, HasanO,PooleAT, etal.1995. Genetic regulationofgibberellin deactivation in Pisum. Plant J.7:513–23

109.Ross JJ, Wil lis CL, Gaskin P, Reid JB.1992. Shoot elongationin Lathyrusodora-tus L.: gibberellin levels in light and dark-growntall anddwarfseedlings.Planta187:10–13

110.Saito T, Abe H, YamaneH, Sakurai A,MurofushiN, et al. 1995. Purifi cation andpropertiesof ent-kaurenesynthaseB fromimmatureseedsof pumpkin.Plant Physiol.109:1239–45

111.Scott IM. 1990. Plant hormone responsemutants.Physiol. Plant. 78:147–52

112.Schwender J, SeemannM, LichtenthalerHK, RohmersM. 1996. Biosynthesisofisoprenoids (carotenoids, sterols, prenylside-chains of chlorophylls andplastoqui-none)viaanovel pyruvate/glyceraldehyde3-phosphatenonmevalonatepathwayin thegreen alga Scenedesmusobliquus. Bio-chem.J. 316:73–80

113.Sherriff LJ, McKay MJ,RossJJ,Reid JB,Wil lis CL. 1994. Decapitation reducesthemetabolism of gibberellin A20 to A1 inPisumsativumL., decreasingtheLe/le dif-ference.Plant Physiol. 104:277–80

114.SmithVA. 1992. GibberellinA1 biosynthe-

458 HEDDEN& KAMIYA

sis in Pisumsativum L. II. Biological andbiochemicalconsequencesof the le muta-tion. Plant Physiol. 99:372–77

115.Smith VA, AlboneKS,MacMillanJ.1990.Enzymatic 3β-hydroxylation of gibberel-l ins A20 and A5. See Ref. 128b, pp.62–71

116.Smith VA, Gaskin P, MacMillan J. 1990.Partial purification and characterization ofthe gibberellin A20 3β-hydroxylase fromseedsof Phaseolusvulgaris.Plant Physiol.94:1390–401

117.Smith VA, MacMillanJ.1984. Purifi cationandpartial characterization of agibberellin2β-hydroxylasefrom Phaseolus vulgaris.J. Plant Growth Regul. 2:251–64

118.Smith VA, MacMillanJ.1986. The partialpurif ication and characterization of gib-berellin 2β-hydroxylases from seedsofPisumsativum.Planta 167:9–18

119.SponselVM. 1983. Thelocalization, meta-bolismandbiological activity of gibberel-lins in maturing andgerminating seedsofPisum sativum cv. ProgressNo. 9. Planta159:454–68

120.Sponsel VM. 1986. Gibberellins in dark-and red-light-grown shoots of dwarf andtall cultivarsof Pisumsativum:the quanti-fication,metabolismandbiologicalactivityof gibberellins in Progress No. 9 andAlaska.Planta 168:119–29

121.SponselVM. 1995. Gibberellin biosynthe-sis and metabolism. In Plant Hormones:Physiology, Biochemistry and MolecularBiology, ed. PJ Davies, pp. 66–97. Dor-drecht: MartinusNijhoff

122.Spray CR, Kobayashi M, Suzuki Y, Phin-ney BO, Gaskin P, et al. 1996. Thedwarf-1 (d1) mutant of Zea maysblocks threestepsin the gibberellin biosynthetic path-way. Proc. Natl. Acad. Sci. USA 93:10515–18

123.SunTP,GoodmanHM, AusubelFM. 1992.Cloning the Arabidopsis GA1 locus bygenomic subtraction. Plant Cell 4:119–28

124.SunTP, Kamiya Y. 1994. The ArabidopsisGA1locusencodesthecyclaseent-kaurenesynthetaseA of gibberellin biosynthesis.Plant Cell 6:1509–18

125.SwainSM,ReidJB.1992. Internodelengthin Pisum: a new allele at the Lh locus.Physiol. Plant. 86:124–30

126.SwainSM, Reid JB, RossJJ.1993. Seeddevelopment in Pisum. The lhi allele re-duces gibberellin levels in developingseeds,andincreasesseedabortion. Planta191:482–88

127.Swain SM, Ross JJ,Kamiya Y, Reid JB.1995. Gibberellin and seeddevelopment.Plant Cell Physiol. 36:S110

128.Swain SM, Ross JJ,Reid JB, Kamiya Y.1995. Gibberellin and peaseeddevelop-ment. Planta 195:426–33

128a.Takahashi M, Kamiya Y, TakahashiN,GraebeJE.1986. Metabolism of gibberel-lins in a cell-free systemfrom immatureseedsof Phaseolusvulgaris L. Planta 168:190–99

128b.TakahashiN, PhinneyBO, MacMillanJ,eds.1991. Gibberellins.NewYork: Sprin-ger-Verlag

129.Talon M, Koornneef M, ZeevaartJAD.1990. Endogenous gibberellins in Arabi-dopsisthalianaandpossible stepsblockedin thebiosynthetic pathwaysof thesemid-warf ga4 and ga5 mutants. Proc. Natl.Acad. Sci. USA87:7983–87

130.Talon M, Koornneef M, ZeevaartJAD.1990. Accumulationof C19-gibberellins inthe gibberellin-insensitive dwarf mutantgai of Arabidopsis thaliana (L.) Heynh.Planta 182:501–5

131.Talon M, TadeoFR, ZeevaartJAD. 1991.Cellular changes induced by exogenousandendogenous gibberellins in shoot tipsof the long-day plant Silene armeria.Planta 185:487–93

132.Talon M, ZacariasL, Primo-Mil lo E.1992.Gibberellins andparthenocarpic abili ty indeveloping ovariesof seedlessmandarins.Plant Physiol. 99:1575–81

133.Talon M, ZeevaartJAD. 1990. Gibberellinandstemgrowth asrelatedto photoperiodin Silene armeria L. Plant Physiol. 92:1094–100

134.Talon M, ZeevaartJAD. 1992. Stemelon-gationandchangesin thelevelsof gibberel-lins in shoot tips induced by differentialphotoperiodic treatments in the long-dayplantSilene armeria. Planta 188:457–61

135.Talon M, ZeevaartJAD, GageDA. 1991.Identificationof gibberellinsin spinachandeffectsof light anddarknessontheir levels.Plant Physiol. 97:1521–26

136.Deletedin proof137.ToyomasuT, KawaideH, Sekimoto H, von

NumersC,Phill ipsAL, etal.1997.Cloningandcharacterization of a cDNA encodinggibberellin 20-oxidasefromrice(Oryzasa-tiva L.) seedlings. Physiol. Plant. 99:111–18

138.Toyomasu T, Tsuji H, Yamane H,NakayamaM, Yamaguchi I, et al. 1993.Light effectson endogenouslevelsof gib-berellins in photoblastic lettuce seeds.J.Plant Growth Regul. 12:85–90

139.Toyomasu T, YamaneH, Yamaguchi I,Murofushi N, Takahashi N, et al. 1992.Control by light of hypocotyl elongationand levels of endogenous gibberellins in


seedlings of Lactucasativa L. Plant CellPhysiol. 33:695–701

140.Turnbull CGN, Crozier A, Schwenen L,GraebeJE.1985. Conversion of [14C]gib-berellinA12-aldehyde toC19- andC20-gib-berellins in a cell-freesystemfrom imma-ture seeds of Phaseolus coccineus L.Planta 165:108–13

141.van Huizen R, Ozga JA, Reinecke DM,Twitchin B, Mander LN. 1995. Seedand4-chloroindole-3-aceticacid regulation ofgibberellin metabolism in pea pericarp.Plant Physiol. 109:1213–17

142.Ward JL, JacksonGS,BealeMH, GaskinP, HeddenP, etal.1997. Stereochemistryofoxidation of the gibberellin 20-alcohols,GA15 andGA44, to 20-aldehydesby gib-berellin 20-oxidases.Chem.Commun, pp.13–14

143.Weller JL, RossJJ, Reid JB. 1994. Gib-berellins and phytochrome regulation ofstemelongationin pea.Planta 192:489–96

144.Winkler RG, Helentjaris T. 1995. Themaizedwarf-3 geneencodesacytochromeP450-mediated early step in gibberellinbiosynthesis.Plant Cell 7:1307–17

145.Wu K, Gage DA, ZeevaartJAD. 1996.Molecular cloning and photoperiod-regu-

latedexpression of gibberellin 20-oxidasefrom the long-day plant spinach. PlantPhysiol. 110:547–54

146.Xu YL, Wu K, PeetersAJM, Gage D,ZeevaartJAD. 1995. The GA5 locus ofArabidopsisthalianaencodesamultifunc-tional gibberellin 20-oxidase: molecularcloning and functional expression. Proc.Natl. Acad. Sci. USA92:6640–44

147.Yamaguchi S, Saito T, Abe H, YamaneH,Murofushi N, et al. 1996. Molecular clon-ing andcharacterization of acDNA encod-ing the gibberellin biosynthetic enzymeent-kaurenesynthaseB frompumpkin (Cu-curbita maximaL.). Plant J. 10:203–13

148.Zanewich KP, Rood SB. 1995. Vernaliza-tion and gibberellin physiology of wintercanola.Endogenousgibberellin (GA1) con-tent and metabolism of [3H]GA1 and[3H]GA20. Plant Physiol. 108:615–21

149.Zeevaart JAD, Gage DA. 1993. ent-Kaurene biosynthesisis enhancedby longphotoperiods in the long-day plants Spi-naciaoleraceaL. andAgrostemmagithagoL. Plant Physiol. 101:25–29

150.ZeevaartJAD, Talon M. 1992. Gibberellinmutants in Arabidopsis thaliana. seeRef.52a, pp. 34–42

460 HEDDEN& KAMIYA

GIBBERELLIN BIOSYNTHESIS: Enzymes, Genes and...

Documents

Transcript of GIBBERELLIN BIOSYNTHESIS: Enzymes, Genes and...