Silencing C-19-GA 2-oxidases induces parthenocarpic development and inhibits lateral branching in tomato plants

Jou nal o Expe imen al Bo any, Vol. 66, No. 19 pp. 5897–5910, 2015 doi:10.1093/jxb/e 300 Ad ance Access publica ion 19 June 2015 This pape is a ailable online ee o all access cha ges (see h p://jxb.ox o djou nals.o g/open_access.h ml o u he de ails) RESEARCH PAPER Silencing C19-GA 2-oxidases induces pa henoca pic de elopmen and inhibi s la e al b anching in oma oplan s LiliamMa ínez-Bello1, ThomasMo i z2 and IsabelLópez-Díaz1,* 1 Ins i u o de Biología Molecula y Celula de Plan as, Uni e sidad Poli écnica de Valencia (UPV)-Consejo Supe io de In es igaciones Cien í icas (CSIC), Ingenie o Faus o Elio s/n, 46022 Valencia, Spain 2 Umeå Plan Science Cen e, Depa men o Fo es Gene ics and Plan Physiology, Swedish Uni e si y o Ag icul u al Science, S-90183 Umeå, Sweden * To whom co espondence should be add essed. E-mail: [email p o ec ed].es Recei ed 7 July 2014; Re ised 1 May 2015; Accep ed 15 May 2015 Edi o : Ch is ine Be e idge Abs ac Gibbe ellins (GAs) a e phy oho mones ha egula e a wide ange o de elopmen al p ocesses in plan s. Le els o ac i e GAs a e egula ed by biosyn he ic and ca abolic enzymes like he GA 2-oxidases (GA2oxs). In oma o (Solanum lycope sicum L.) C19 GA2oxs a e encoded by a small mul igenic amily o i e membe s wi h some deg ee o edun- dancy. In o de o in es iga e hei oles in oma o, he silencing o all i e genes in ansgenic plan s was induced. Asigni ican inc ease in ac i e GA4 con en was ound in he o a ies o ansgenic plan s. In addi ion, he ansgenic un e ilized o a ies we e much bigge han wild- ype o a ies (abou 30 imes) and a ce ain p opo ion (5–37%) we e able o de elop pa henoca pically. Among he GA2ox amily, genes GA2ox1 and -2 seem o be he mos ele an o his pheno ype since hei exp ession was induced in un e ilized o a ies and ep essed in de eloping ui s, in e sely co ela ing wi h o a y g ow h. In e es ingly, ansgenic lines exhibi ed a signi ican inhibi ion o b anching and a highe con en o ac i e GA4 in axilla y buds. This pheno ype was e e ed, in ansgenic plan s, by he applica ion o paclobu azol, a GA biosyn hesis inhibi o , sugges ing a ole o GAs as ep esso s o b anching. In summa y, his wo k demons a es ha GA 2-oxidases egula e gibbe ellin le els in o a ies and axilla y buds o oma o plan s and hei silencing is esponsible o pa henoca pic ui g ow h and b anching inhibi ion. Key wo ds: B anching, GA 2-oxidases, gibbe ellins, pa henoca py, silencing, oma o. In oduc ion Gibbe ellins (GAs) a e phy oho mones ha egula e a wide ange o de elopmen al p ocesses in plan s such as ge mi- na ion, s em elonga ion, lea expansion, lowe ing, ui - se , and ui g ow h (Yamaguchi, 2008; Mu asa-Gö gens and Hedden, 2009; Seymou e al., 2013). GA biosyn hesis (Fig. 1) s a s wi h ge anyl ge anyl diphospha e (GGDP) which is con e ed o en -kau ene by he ac ion o wo enzymes: CPS (en -copalyl diphospha e syn hase), and KS (en -kau ene syn hase). The en -kau ene is hen con e ed o GA12, he i s GA o he pa hway, by he ac ion o wo mem- b ane-associa ed P450 enzymes (cy och ome P450 mono- oxygenases): en -kau ene oxidase (KO) and en -kau enoic acid oxidase (KAO). F om his s ep, he e a e wo pa allel pa hways: he non-13-hyd oxyla ion (leading o ac i e GA4) and he ea ly 13-hyd oxyla ion (leading o ac i e GA1), which depend on he ac ion o he cy och ome P450 enzyme, GA 13-oxidase (GA13ox) (Magome e  al., 2013). The enzymes ha ca alyse he las s eps o he biosyn hesis pa hways a e This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/3.0/), which pe mi s un es ic ed euse, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. © The Au ho 2015. Published by Ox o d Uni e si y P ess on behal o he Socie y o Expe imen al Biology. Abb e ia ions: GA, gibbe ellin; GA2ox, gibbe ellin 2-oxidase; GA3ox, gibbe ellin 3-oxidase; GA13ox, gibbe ellin 13-oxidase; GA20ox, gibbe ellin 20-oxidase; PCB, paclobu azol; RT–qPCR, e e se ansc ip ion-quan i a i e PCR; SE, s anda d e o ; shRNA, sho -hai pin RNA; WT, wild ype. 5898 | Ma ínez-Bello e al. GA 20-oxidases (GA20ox) and GA 3-oxidases (GA3ox) which a e 2-oxoglu a a e dependen dioxygenases (2ODDs) (Yamaguchi, 2008; Hedden and Thomas, 2012). The GA me abolic pa hway is igh ly con olled in o de o keep adequa e ho mone le els which a e necessa y o he ine egula ion o plan g ow h and de elopmen . Endogenous GA le els a e egula ed, mainly, by biosyn hesis and inac i a ion. Se e al mechanisms o inac i a ing GAs ha e been iden i- ied, he mos p e alen being 2β-hyd oxyla ion. The enzymes esponsible o his ac i i y a e GA 2-oxidases (GA2oxs) which a e also 2ODDs, like he GA20ox and GA3ox enzymes (Thomas e al., 1999; Rieu e al., 2008; Hedden and Thomas, 2012). The GA2ox enzymes can be di ided in o wo majo g oups acco ding o subs a e speci ici y: one g oup ac s on he GAs wi h 19 ca bons (C19-GA2oxs), which include he bioac- i e GAs, GA1 and GA4, and hei immedia e p ecu so s GA20 and GA9 (Se ani e al., 2007a; Hedden and Thomas, 2012), whe eas he o he g oup ac s on GAs wi h 20 ca bons (C20- GA2oxs) GA12 and GA53 (Schombu g e  al., 2003). Toma o (Solanum lycope sicum L.) is one o he mos impo an c ops wo ldwide. I has been widely used, no only as a ood bu also as a model plan due o se e al de elopmen al ai s no ound in o he model plan s such as A abidopsis haliana, An i hinum majus, and ice (O yza sa i a L.). These ai s a e sympodial shoo b anching, compound lea es, and he o ma ion o leshy climac e ic ui s (Ranjan e  al., 2012). In addi ion, oma o has se e al cha ac e is ics ha make i a con enien model plan species, such as a ela i ely compac genome (950 Mb) combined wi h a ma ke -sa u a ed gene ic linkage map, ich ge mplasm collec ions (Toma o Gene ics Resou ce Cen e ), and highly e icien ans o ma ion p o ocols (Ca alho e al., 2011). The Mic o-Tom (MT) cul i a has been p oposed as a e y con enien oma o model sys em because o i s small size, apid li e cycle, and high- h oughpu capabili ies (Meissne e  al., 1997, Ca alho e  al., 2011). The pheno ype o his cul i a is he esul o poin mu a ions in he genes Dwa (D), Sel -p uning (SP), and In e node leng h educ ion (Il ) bu no in GA biosyn hesis (Ma í e al., 2006). Se ani e al (2007b) demons a ed ha his cul i a is an app op ia e model o esea ch on he ho monal egula ion o ui se and de elop- men because i esponds o GAs and auxins in he same way as o he oma o cul i a s. The isola ion and cha ac e iza ion o GA genes (Se ani e al, 2007a), GA mu an s (Ca e a e al, 2012), and GA ansgenic plan s (Ga cia-Hu ado e al., 2012) in his cul i a , is ano he a gumen in a ou o i s choice o he s udy o de elopmen al p ocesses media ed byGA. In oma o, he e a e i e GA2ox genes ha encode o enzymes ha belong o he g oup o C19-GA2oxs. Sequence analysis o amino acid composi ion showed ha GA2ox2, -4, and -5 belong o subg oup I and GA2ox1 and -3 belong o subg oup II (Se ani e  al., 2007a). The GA2ox genes a e edundan ly exp essed in ege a i e and ep oduc i e issues o oma o. In o de o unde s and he oles o GA2ox genes in oma o de elopmen , ansgenic plan s we e gene a ed in which he exp ession o all GA2ox genes was silenced. The ansgenic plan s showed he induc ion o acul a i e pa henoca py and a signi ican inhibi ion in he appea ance o la e al b anches and also a signi ican inc ease in he GA4 concen a ion in o a- ies and axilla y buds compa ed wi h wild ype (WT) plan s. This sugges s ha GA2ox genes a e impo an o he egula- ion o GA le els in he o a ies and axilla y buds, hus playing an impo an ole in egula ing hei de elopmen . Ma e ials and me hods Accession numbe s Accession numbe s o he sequences used in his s udy a e as ollows: GA2ox1 (EF441351), GA2ox2 (EF441352), GA2ox3 (EF441353) GA2ox4 (EF441354), GA2ox5 (EF441355), and SlAc in (U60482.1). Fig.1. GA biosyn hesis pa hway. GGDP, ge anylge anyl diphospha e; CPS, en -copalyl diphospha e syn hase; KS, en -kau ene syn hase; KO, en - kau ene oxidase; KAO, en -kau enoic acid oxidase; GA13ox, GA 13-oxidase; GA20ox, GA 20-oxidase; GA3ox, GA 3-oxidase; GA2ox, GA 2-oxidase; GA-ca , GA-ca aboli e. (This igu e is a ailable in colou a JXB online.) Toma o C19-GA2ox ole in pa henoca py and b anching | 5899 Plan ma e ial and g ow h condi ions Toma o (S.lycope sicum L.) c . Mic o-Tom (MT) was used. Plan s we e pu in a g ow h chambe a 26 ± 2°C unde 16/8 h ligh /da k (pho on luence o 115 μmol m–2 s–1) condi ions. In he g een- house, plan s (one pe po ) we e g own in 1 L po s wi h a mix u e o pea : e miculi e (1:1 / ) in 24/20ºC day/nigh condi ions and i iga ed daily wi h Hoagland’s nu ien solu ion (Hoagland and A non, 1950). Na u al ligh was supplemen ed wi h Os am lamps (Powe s a HQI-BT, 400 W) o ge a 16 h ligh pho ope iod. Gene a ion o cons uc s o silencing A sho -hai pin RNA (shRNA) cons uc was used o induce mul iple silencing o all he GA2ox genes. To ensu e i s e iciency, a chime ic agmen was used composed o po ions iden ical o he GA2ox1, GA2ox3, and GA2ox4 genes and e y simila o GA2ox2 and GA2ox5 (mo e han 92% iden i y). This agmen was composed o a po - ion o 163 bp om GA2ox4, 188 bp om GA2ox1, and 123 bp om GA2ox3 (see Supplemen a y Table S1 a JXB online). Each agmen was ampli ied sepa a ely om cDNA clones om Se ani e al. (2007a) using p ime s desc ibed in Supplemen a y Table S2 a JXB online. These agmen s we e pu i ied and joined by an o e lapping PCR as shown in Supplemen a y Fig. S1 a JXB online. The inal agmen shRNA2ox was cloned in he pGEM-T Easy ec o (P omega) and sequenced o con i m ha he sequences we e co ec and had been well assembled. To c ea e he inal cons uc o oma o plan s ans o - ma ion, Ga eway echnology was used using he p ime s desc ibed in Supplemen a y Table S2 a JXB online. The agmen was ecombined in o he bina y ec o pK7GWIWG2 (II) (Ka imi e al., 2002), using LR Clonase (In i ogen) o ob ain he cons uc 35S::GA2ox/RNAi. This ec o ca ies he cons i u i e p omo e 35S om Cauli lowe mosaic i us (CaMV) and he np II gene o kanamycin esis ance. Isola ion o ansgenic oma oplan s Cons uc 35S::GA2ox/RNAi was used o ans o ma ion using he Ag obac e ium ume aciens s ain LBA4404. Toma o ans o - ma ion was ca ied ou as desc ibed in Ga cía-Hu ado e al. (2012). Ge mina ion condi ions Toma o seeds om ansgenic lines and WT plan s we e imbibed in 90 mm diame e plas ic Pe i dishes on h ee laye s o Wha man No. 1 il e pape (Fishe ) wi h double deionized wa e (ddH2O). Pe i dishes we e sealed wi h Pa a ilm (Fishe ) o p e en e apo a ion and pu in a g ow h chambe , as p e iously desc ibed. Ge mina ion was sco ed when he adicle p o uded om he seed coa . Hypoco yl and oo leng h measu emen s in seedlings Seeds om each WT and ansgenic line we e ge mina ed (no mo e han 20 seeds pe ja ) and cul u ed in ge mina ion medium (GM) [Mu ashige and Skoog (MS) sal s, 1% (w/ ) suc ose, and 0.8% (w/ ) aga ] in a g ow h chambe . A e 7 d, he seedlings we e pho- og aphed and he hypoco yls and oo s we e measu ed using he ImageJ so wa e (Na ional Ins i u es o Heal h, USA). B anching measu emen s The b anching pa e n o ansgenic 35S::GA2ox/RNAi and WT plan s was de e mined by measu ing he pe cen age o b anched axils, calcula ed as: % B anching=(no. o b anched axils/ o al no. o axils)×100. B anched axils a e hose axils wi h a b anch equal o longe han 0.5 cm. Ho mone applica ion inplan s Fo his expe imen , 27-d-old plan s g own in po s in he g eenhouse we e used. Paclobu azol (PCB) and gibbe ellic acid (GA3) om Duche a we e i s dissol ed in absolu e e hanol and hen dilu ed wi h he nu ien solu ion used o wa e he plan s. The GA3 con- cen a ion used was 10–5 M and he PCB concen a ion was 10–6 M.Plan s we e ea ed on al e na e days o 16 d, a o al o eigh applica ions. As a con ol g oup, plan s we e wa e ed wi h nu ien solu ion con aining he same amoun o absolu e e hanol used o dissol e PCB and GA3. Fo each line a o al o 15 plan s pe ea - men we e used. De e mina ion o pa henoca pic capaci y To de e mine pa henoca pic capaci y, h ee lowe s pe uss we e selec ed om he i s ou usses o he plan . Flowe emascula- ion was ca ied ou 2 d be o e an hesis o p e en sel -pollina ion and all non-selec ed lowe s om he uss we e emo ed. Only hose ui s bigge han 0.1 g, 20 d a e an hesis, we e conside ed o be pa henoca pic. RNA isola ion To al RNA was isola ed om ep oduc i e and ege a i e is- sues, using he RNeasy Plan Mini Ki (Qiagen) acco ding o he manu ac u e ’s p o ocol. To al RNA was ea ed wi h RNase-F ee DNAse (Qiagen), acco ding o he manu ac u e ’s ins uc ions. RNA concen a ion was measu ed using a Nanod op ND-1000 Spec opho ome e (Wilming on, Delawa e USA) and RNA in eg- i y was de e mined by aga ose gel analysis. Th ee biological epli- ca es o each sample we e used. Fo he o a ies, only hose om he i s ou in lo escences we e used and only h ee o a ies pe in lo escence. Quan i a i e Real-Time RT-PCR analysis (RT-qPCR) One mic og am o o al RNA was used o cDNA syn hesis using he High Capaci y cDNA Re e se T ansc ip ion Ki (Applied Biosys ems) wi h andom hexame s. Aliquo s (2 μl) o dilu ed cDNA solu ion (1:2 / ) we e used o qPCR eac ions in a inal ol- ume o 20μl. PCR eac ions we e pe o med in an op ical 96-well pla e on a 7500 Fas Real-Time PCR Sys em using Powe SYBR G een PCR Mas e Mix (Applied Biosys ems). The analyses we e ca ied ou using h ee biological eplica es and h ee echnical epli- ca es. The p ime pai s used o PCR ampli ica ion we e he same as in Se ani e al. (2008) (see Supplemen a y Table S3 a JXB online). Fo each pai o p ime s he op imal concen a ion and e iciency in he PCR eac ion was de e mined. In e e y un a empla e- ee con- ol was included. To be able o compa e da a om di e en uns, he CT alue o e e y gene was no malized wi h he CT alue o SlAc in de e mined in he same un. Rela i e exp ession o mRNA was calcula ed by he 2–ΔΔCT me hod (Li ak and Schmi gen, 2001). GA quan i ica ion GAs we e quan i ied in ep oduc i e and ege a i e issues. Apical shoo s, s ems, and axilla y buds we e chosen as he ege a i e is- sues. Apical shoo s (apices and he h ee younges lea es) we e col- lec ed a e lowe ing bu be o e an hesis, when he bigges lowe in he i s in lo escence was no mo e han 4 mm. Fo his analysis he lowe s we e emo ed. Se en biological eplica es o i e plan s each we e used o WT plan s and ou biological eplica es o i e plan s each o ansgenic line L1. S ems we e aken om 27-d-old plan s om he i s o he i h in e node o e he co yledons, once he api- cal shoo and he lea es we e emo ed. Th ee biological eplica es o h ee plan s each we e used. Axilla y buds, which did no show any isible g ow h a he ime o ha es , we e aken om he 3 d, 4 h, and 5 h axil o he plan and h ee biological eplica es o 15 plan s each we e used. Fo ep oduc i e issue, 10-d-old hand-pollina ed o a ies we e chosen and i e biological eplica es o ou o a ies each we e used. The ma e ial was ozen immedia ely in liquid N2 be o e s o age a –70°C un il analysis. 5900 | Ma ínez-Bello e al. The ex ac ion p ocedu e o apical shoo s and o a ies was pe - o med as ollows. Aliquo s o 100 mg esh weigh we e homog- enized in cold 80 % ( / ) me hanol-H2O, s i ed o e nigh a 4°C and e-ex ac ed wice o 30 min wi h one-hal olume o me ha- nol. The me hanolic ex ac s we e combined and aken o d yness in a acuum and dissol ed in 0.5 ml o 10% me hanol solu ion. This ex ac was hen applied o a p e-equilib a ed 500 mg SAX column (BondElu SS-SAX; Va ian Scha lau). The column was washed wi h H2O pH 8.0 be o e being elu ed wi h 4 ml o 0.2 M o mic acid solu- ion. The o mic-acid elua e was un di ec ly on o a p e-equilib a ed 500 mg C18 column (BondElu C18; Va ian Scha lau) which was hen washed wi h H2O pH 3.0 and inally elu ed wi h 5 ml 80% me hanol. The me hanol elua e was d ied, dissol ed in 0.2 ml o 2-p opanol and hen me hyla ed wi h ime hylsilyl-diazome hane in hexane. A e me hyla ion, he esidue was d ied and he samples we e ime hyl- silyla ed using 7.5µl o py idine and 7.5µl o BSTFA solu ion con- aining 1% TCMS (N,O-bis( ime hylsilyl) i luo oace amide+1% ime hyl chlo osilane). Fo gas ch oma og aphy–mass spec om- e y analysis (GC–MS), he samples we e dissol ed in he mix o py idine/BSTFA and injec ed in o a gas ch oma og aph (7890A; Agilen Technologies) coupled o a mass spec ome e (7000 T iple Quad; Agilen Technologies). The amoun o GA in he samples was quan i ied by iso opic dilu ion, using he MassHun e Quan i a i e so wa e p o ided by Agilen . The ex ac ion p ocedu e o GA quan i ica ion in s ems and axil- la y buds was pe o med as ollows. Aliquo s (abou 100–200 mg esh weigh ) o ozen ma e ial we e ex ac ed wi h 80% me ha- nol–1% ace ic acid, and he ex ac s passed consecu i ely h ough HLB ( e e se phase), MCX (ca ionic exchange), and WAX (ionic exchange) columns (Oasis 30 mg, Wa e s), as desc ibed in Seo e al. (2011). The inal esidue was dissol ed in 5% ace oni ile–1% ace ic acid, and he ho mones we e sepa a ed using an au osample and e e se phase UPHL ch oma og aphy (2.6 μm Accuco e RP-MS column, 50 mm leng h×2.1 mm i.d.; The moFishe Scien i ic) wi h a 5–50% ace oni ile g adien con aining 0.05% ace ic acid, a 400μl min–1. The ho mones we e analysed by elec osp ay ioniza ion (neg- a i e mode, sp ay ol age 3.0 kV, hea e empe a u e 150ºC, shea h gas low a e 40μl min–1, auxilia y gas low a e 10μl min–1) and a ge ed-SIM using a Q-Exac i e spec ome e (O bi ap de ec o ; The moFishe Scien i ic). The concen a ions o ho mones in he ex ac s we e de e mined using embedded calib a ion cu es and he Xcalibu 2.2 SP1 build 48 and T aceFinde p og ams. The deu e ium-labelled ho mones (pu chased om P o esso L Mande , Canbe a, Aus alia, OlChemim L d, Olomouc, Czech Republic, and he Camb idge Iso ope Labo a o y, Ando e , USA) we e added o he ex ac s as in e nal s anda ds o quan i ica ion o each o he di e en GAs. In he case o axilla y buds and s ems, GA34 was no a ailable and could no be quan i ied. S a is ical me hods S a is ical ea men s o he da a we e made using he SPSS p o- g am, e sion 16.0 o windows, IBM. The analyses we e made by -s uden and one-way ANOVA o p<0.05. As pos -hoc es s we use Bon e oni and Tukey’ HSD. Resul s Isola ion o ansgenic oma olines The i e GA2ox genes o oma o a e edundan ly exp essed acco ding o Se ani e  al. (2007a) sugges ing unc ional edundancy which had p e iously been demons a ed in A abidopsis (Rieu e al., 2008). In o de o s udy he ole o GA2ox genes, he whole oma o GA2ox amily was silenced. As desc ibed in he Ma e ial and me hods, oma o lea explan s we e in ec ed wi h Ag obac e ium ume aciens LBA4404 ca ying he 35S::GA2ox/RNAi sho -hai pin cons uc . Six independen diploid lines we e ob ained: L1, L5, L13, L22, L25, and L81. These we e g own in po s and cul u ed in he g eenhouse. Thei p ogenies, ob ained by sel -pollina ion, we e es ed o kanamycin esis ance o de e mine he seg e- ga ion o he ansgene, bu only h ee o hem had a 3:1 p o- po ion, co esponding o a single locus o he ansgene (L1, L5, and L81). Homozygous lines om hese plan s, ca ying he 35S::GA2ox/RNAi ansgene, we e isola ed and lines L1 and L5 we e chosen o ex ensi e cha ac e iza ion. Silencing o GA2ox genes in ansgenicplan s To de e mine he e iciency o silencing o ansgene 35S::GA2ox/RNAi, he exp ession o he GA2ox genes was analysed in ege a i e and ep oduc i e issues. In ege a i e issues, signi ican silencing o genes GA2ox3, GA2ox4, and GA2ox5 was de ec ed in oo s and o GA2ox3 and GA2ox5 in hypoco yls and s ems (Fig.2). Genes GA2ox1 and GA2ox2 we e no silenced in any o he ege a i e issues analysed. In ep oduc i e issues, by con as , silencing o all he GA2ox genes was de ec ed (Fig. 3). These esul s indica e ha he 35S::GA2ox/RNAi cons uc is able o induce silencing o all a ge genes in a leas one issue. The ela i e abundance o GA2ox ansc ip s was compa ed in di e en issues and a posi i e co ela ion was ound be ween abundance and silencing (Fig.4). Genes mo e e icien ly silenced wi hin a is- sue we e hose mos highly exp essed. Gene al cha ac e iza ion o ansgenicplan s T ansgenic plan s om lines L1 and L5 we e cha ac e ized o a numbe o ege a i e and ep oduc i e ai s which we e known o be in luenced by GAs, such as ge mina ion, hypoco yl and s em elonga ion, and lowe ing. None o hese pa ame e s showed signi ican di e ences be ween ansgenic and WT plan s, including plan heigh (see Supplemen a y Figs S2 and S3 and Supplemen a y Table S4 a JXB online). Di e ences ound in he hypoco yl leng hs and ui weigh o line L5 we e no ep oduced in o he expe imen s. Asmalle numbe o seeds pe ui was only de ec ed in line L1. T ansgenic plan s show acul a i e pa henoca py GAs a e impo an o ui -se and de elopmen (Seymou e  al., 2013). Usually, un e ilized oma o o a ies do no g ow and de elop in o ui s al hough GA applica ion can induce pa henoca pic de elopmen (Se ani e al., 2007a). To de e mine i o a ies om ansgenic lines had he abili y o g ow in he absence o e iliza ion, lowe s we e emas- cula ed 2 d be o e an hesis, o p e en sel -pollina ion, and he o a ies we e collec ed a e 20 d.While, in WT plan s, he unpollina ed o a ies did no g ow, in ansgenic plan s a ce ain numbe o non-pollina ed o a ies ( om 5% o 37%, depending on he line and expe imen ) de eloped in o pa henoca pic ui s (Table1; Fig.5). The de elopmen o hand-pollina ed and emascula ed o a ies was also compa ed o 20 d in WT and ansgenic lines (Fig.6). The de elopmen Toma o C19-GA2ox ole in pa henoca py and b anching | 5901 o emascula ed o a ies was e y di e en be ween ansgenic and WT plan s (Fig.6A). A e an hesis, he weigh o WT emascula ed o a ies did no change signi ican ly while, in ansgenic plan s, o a y weigh inc eased d ama ically (mo e han 30 imes) al hough hey did no eached he weigh o pollina ed o a ies which we e 10 imes bigge . Pollina ed o a ies de eloped in a simila way in ansgenic and WT plan s al hough o a ies om ansgenic lines g ew as e ROOT 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,1 1 10 WT L1 L5 * *** A STEM 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 * * B APICAL PORTION 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,1 1 10 WT L1 C HYPOCOTYL 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,1 1 10 WT L1 L5 * * * * D LEAVES-Yp 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 L5 E LEAVES-Op 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 L5 * F Fig.2. Silencing o GA2ox1–5 in ege a i e issues o ansgenic 35S::GA2ox/RNAi plan s. Exp ession o GA2ox genes in ege a i e issues o wild- ype (WT) plan s and ansgenic lines L1 and L5. (A) Roo s ( om 90-d-old plan s), (B) s ems (27-d-old plan s, app oxima ely 1 week be o e an hesis), (C) apical shoo s (apical po ions consis ing o shoo apices and h ee apical young lea es be o e an hesis), (D) hypoco yls (7-d-old seedlings), (E) lea es-Yp ( e minal lea om young lowe ing plan s), and (F) lea es-Op (lea es om he apical po ion o 90-d-old plan s). Values a e means o h ee biological eplica es ±SE and a e ela i e o he alue o exp ession in he WT plan s (1.0). *, Signi ican ly di e en om he WT plan s (P <0.05). 5902 | Ma ínez-Bello e al. han WT o a ies be o e day 20, a which ime he di e ences disappea ed (Fig.6B). GA2ox genes exp ession in pollina ed and non-pollina ed o a ies The pa henoca pic ui de elopmen ound in ansgenic plan s sugges s a ole o GA 2-oxidases in he con ol o o a y g ow h. The e o e, he exp ession o he i e GA2ox genes was compa ed be ween e ilized and un e ilized o a- ies. T ansc ip le els o GA2ox genes we e measu ed in hand- pollina ed and emascula ed o a ies om 2 d be o e an hesis un il 5 d la e when ui -se had al eady been es ablished, as shown in Fig.7A. Pollina ed o a ies a e 5 d o an hesis a e 10- imes bigge han non-pollina ed ones. Two o he genes (GA2ox1 and -2) a e di e en ially exp essed in e ilized and non- e ilized o a ies a e an hesis (Fig.7B, C). The exp es- sion o GA2ox1 and GA2ox2 dec eased a e pollina ion bu inc eased signi ican ly in non-pollina ed o a ies by up o 30 imes. The o he h ee genes GA2ox3,-4, and-5 did no dis- play di e ences in he exp ession in e ilized and non- e - ilized o a ies (Fig.7D–F). A e an hesis, ansc ip le els o GA2ox3 dec eased while GA2ox4 and GA2ox5 inc eased (Fig. 7D–F), al hough no signi ican di e ence was ound be ween e ilized and un e ilized o a ies. The pa e ns o exp ession sugges ha GA2ox1 and GA2ox2 could play a ole in ui de elopmen as nega i e egula o s o o a y g ow h. T ansgenic plan s show inhibi ion o b anching As men ioned be o e, he main aspec s o ege a i e de el- opmen o ansgenic lines we e simila o WT plan s excep o he b anching pa e n (see Supplemen a y Fig. S3 and Supplemen a y Table S4 a JXB online). Aclea educ ion in he numbe o la e al b anches was ound in ansgenic lines compa ed wi h WT plan s (Fig.8A). To s udy his phe- no ype in mo e de ail, he de elopmen o axilla y buds in o b anches was ollowed in ansgenic lines L1 and L5 and in WT plan s o 85 d (Fig.8B). A e 40 d om sowing, mo e han 60% o he axils in WT plan s had la e al b anches while in ansgenic lines he pe cen age o la e al b anches was app oxima ely 45% o L5 and less han 20% o L1. O e ime he di e ences om he WT we e educed, mainly in L5. By he end o he expe imen (85 d), he pe cen age o b anching in WT plan s was abou 95%, while in L5 i was 80% and in L1 i was55%. d0 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,1 1 10 WT L1 L5 * * A P+5 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 L5 ** ** B P+10 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 L5 * * * * * * * * * C E+5 2ox1 2ox2 2ox3 2ox4 2ox5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 WT L1 L5 * * *** * ** * D Fig.3. Silencing o GA2ox1–5 in ep oduc i e issues o ansgenic 35S::GA2ox/RNAi plan s. Exp ession o GA2ox genes in ep oduc i e issues o wild- ype (WT) plan s and ansgenic lines L1 and L5. (A) d0 (o a ies a he ime o an hesis), (B) P+5 (pollina ed o a ies 5 d pos -an hesis), (C) P+10 (pollina ed o a ies 10 d pos -an hesis) and (D) E+5 (emascula ed o a ies 5 d pos -an hesis). Values a e means o h ee biological eplica es ±SE and a e ela i e o he alue o exp ession in he WT plan s (1.0). *, Signi ican ly di e en om he WT plan s (P <0.05). Toma o C19-GA2ox ole in pa henoca py and b anching | 5903 B anching inhibi ion is media ed byGAs To es i b anching inhibi ion was ela ed o a modi ica ion o he GA con en , GA le els we e al e ed by he applica ion o PCB (a GA biosyn hesis inhibi o ) and by GA3 applica ion. In addi ion, plan s we e also ea ed simul aneously wi h GA3 and PCB o de e mine i he PCB ac ion on b anching was media ed by GAs. Figu e9 shows he pe cen age o b anch- ing o 52-d-old plan s ea ed wi h a GA3, PCB o GA3+PCB solu ion o 16 d.As seen be o e, un ea ed ansgenic plan s (L1 and L5) had signi ican ly ewe b anches han WT, espe- cially L1. In e es ingly his pheno ype was supp essed by PCB ea men in bo h ansgenic lines while GA3 applica ion was able o inhibi b anching in he WT (Figs 9, 10). Mo eo e , b anching s imula ion induced by PCB ea men was com- ple ely supp essed by simul aneous GA3 applica ion, indica - ing ha he PCB e ec on b anching was due o a educ ion in GA le el and no o in e ac ion wi h o he pa hways (Figs 9, 10). GA con en in ansgenicplan s GA 2-oxidases ca alyse he deac i a ion o ac i e GAs and hei p ecu so . In o de o e alua e he e ec o GA2ox silencing on GA me abolism, GAs le els we e quan i ied in o a ies and ege a i e issues. Fe ilized o a ies we e com- pa ed because hey de elop in o ui s in bo h ansgenic and WT plan s. On he o he hand, un e ilized o a ies o ans- genic plan s de elop in o pa henoca pic ui s while o a- ies do no g ow in he WT plan s (Fig.5). Ahighe con en o ac i e GA4 was ound in he o a ies o ansgenic plan s Fig.4. Rela ion be ween abundance o mRNA le els o GA2ox genes and silencing in ege a i e and ep oduc i e issues o ansgenic plan s. Ba s ep esen he ela i e abundance o GA2ox genes in di e en issues and hose labelled wi h as e isks a e hose genes silenced in a leas one ansgenic line. The SlAc in gene was used as an in e nal con ol o each issue and alues a e means o h ee biological eplica es ±SE. All da a a e exp essed ela i e o he alues o GA2ox1 exp ession in oo s, se as 1.0. **, The gene is silenced in wo ansgenic lines; *, he gene is silenced in only one ansgenic line. The ege a i e issues we e aken om Roo s ( om 90-d-old plan s), S ems (27-d-old plan s, app oxima ely 1 week be o e an hesis), Apical shoo s (apical po ions consis ing o shoo apices and h ee apical young lea es be o e an hesis), Hypoco yls ( om 7-d-old seedlings), Lea es-Yp (apical young lea es o 35-d-old plan s), and Lea es-Op (apical young lea es o 90-d-old plan s). The ep oduc i e issues we e aken om o a ies: d0 (o a ies a he ime o an hesis), P+5 (pollina ed o a ies 5 d pos -an hesis), P+10 (pollina ed o a ies 10 d pos -an hesis), and E+5 (emascula ed o a ies 5 d pos -an hesis). Table1. Pa henoca pic capaci y o wild ype (WT) and ansgenic 35S::GA2ox/RNAi plan s Pa henoca pic capaci y was de e mined, in wo sepa a e expe imen s, by emascula ing lowe s 2 d be o e an hesis o p e en sel -pollina ion ( ou usses pe plan , h ee lowe s pe uss) and collec ing o a ies 20 d a e an hesis. Pa henoca pic ui s we e hose bigge han 0.1 g. Values in pa en heses a e he numbe o pa henoca pic ui s/ o al o emascula ed o a ies. Weigh da a a e he a e age o pa henoca pic ui s ±SE a 20 d a e an hesis. Line % F ui -se Pa henoca pic ui weigh d20 (g) Expe imen I WT 0 (0/69) – L1 37 (11/30) 0.62 ± 0.1 L5 15 (7/47) 0.81 ± 0.15 L81 34 (22/64) 0.59 ± 0.13 Expe imen II WT 0 (0/101) – L1 5 (4/79) 0.52 ± 0.2 L5 34 (35/102) 0.70 ± 0.05 5904 | Ma ínez-Bello e al. (mo e han 3- old in L1 and 2- old in L5) and a 60% dec ease was ound in he le els o inac i e GA34 compa ed wi h WT plan s (Table2). The ac i e GA1 con en was no al e ed in any o he ansgenic lines al hough GA20 was signi ican ly highe and GA8 signi ican ly lowe inL1. No di e ences in GA le els we e obse ed be ween ansgenic and WT plan s in apical shoo s and s ems in acco dance wi h he absence o s em elonga ion di e ences (see Supplemen a y Fig. S3 and Supplemen a y Tables S4 and S5 a JXB online). The b anching pheno ype ound in ansgenic plan s and he b anching s imula ion induced by PCB ea men sugges ed ha GA con en could be enhanced in axilla y buds o ans- genic plan s. The GA con en o axilla y buds in WT plan s and L1, which showed he mos ex eme pheno ype, was com- pa ed,. The le el o ac i e GA4 (bu no GA1) inc eased abou 2- old in he axilla y buds om axils 4 and 5 o ansgenic plan s (Table3). In addi ion, he le els o inac i e GA29 and GA51 we e educed o unde ec able le els in ansgenic plan s, as expec ed om he educed ac i i y o he GA 2-oxidases. Discussion E iciency o gene silencing The le els o ac i e GAs a e egula ed by biosyn he ic and ca abolic enzymes like he GA 2-oxidases (GA2oxs) which, in oma o (Solanum lycope sicum L.), a e encoded by a small mul igenic amily o i e membe s wi h some deg ee o edun- dancy (Se ani e al., 2007a). In o de o in es iga e he oles o GA2ox enzymes in oma o de elopmen , he silencing o all i e GA2ox genes was induced using a chime ic sho - hai pin cons uc . The 35S::GA2ox/RNAi ansgene educed GA2ox mRNA le els in oma o plan s, al hough la ge di e - ences in he e iciency o gene silencing we e ound (Figs 2, 3). The e iciency o gene silencing has gene ally been associa ed wi h he deg ee o iden i y be ween a ge and igge RNA sequences (Ka och and Thaku , 2013). Howe e , in his case, he di e ences in silencing do no co ela e wi h he deg ee o sequence iden i y. On he o he hand, a di ec ela ionship be ween silencing and a ge abundance was ound as p e i- ously desc ibed in o he wo ks (Miki e al., 2005; Gallego- Gi aldo, 2008; Gil-Humanes e al., 2010; Shigemi su e al., 2012). Wi hin a issue, he silenced genes we e always among he mos abundan ly exp essed (Fig.4). Howe e , a gene ha is exp essed wi h a simila abundance in wo di e en issues may be silenced in only one o hem. This disc epancy migh be ela ed o di e ences in he cellula componen s o he silencing machine y o o di e ences in he exp ession o he angene be ween issues. E ec o GA2ox silencing in GA me abolism GA 2-oxidase enzymes ca alyse he addi ion o a hyd oxyl g oup in he C-2 ca bon o ac i e GA1 and GA4 and hei immedia e p ecu so s, GA20 and GA9, espec i ely (Fig.1), causing he loss o biological ac i i y. Silencing o GA2ox genes is expec ed o p oduce an accumula ion o ac i e GAs and a dec ease in he inac i e p oduc s (GA29, GA8, GA51, and GA34). GA le els we e quan i ied in ou di - e en issues bu in only wo o hem (o a ies and axil- la y buds) could signi ican al e a ions o GA me aboli es be ound. In he apical shoo s, whe e none o he GA2ox genes we e silenced, he e we e no changes in GA le els (see Supplemen a y Table S5 a JXB online). In s ems, whe e only he exp ession o GA2ox3 and GA2ox5 was educed, no di e ences in he GA con en we e de ec ed (see Supplemen a y Table S5 a JXB online). This ac may be WT L1 Fig.5. Pho og aph o 20-d-old un e ilized o a ies om WT and 35S::GA2ox/ RNAi plan s (Line L1). (This igu e is a ailable in colou a JXB online.) d-2d0d+5 d+10 d+20 OVARY WEIGTH (mg) 1 10 100 1000 WT L1 L5 L81 DAYS EMASCULATION ANTHESIS A d-2 d0 d+5d+10d+20 1 10 100 1000 10000 WT L1 L5 L81 OVARY WEIGTH (mg) DAYS EMASCULATION ANTHESIS B Fig.6. De elopmen o ui s om emascula ed o pollina ed lowe s in WT and 35S::GA2ox/RNAi plan s (Lines L1, L5, and L81). The s aminal cones we e emo ed wo days be o e an hesis (d-2) o p e en sel pollina ion ( ou usses pe plan , h ee lowe s pe uss). Some o hem we e hand- pollina ed 2 d a e emascula ion (d0) and o he s we e le unpollina ed. O a ies we e collec ed on days –2, 0, 5, 10, and 20. Values a e he mean o a leas 15 o a ies pe s age ±SE. (A) Non pollina ed o a ies. (B) pollina ed o a ies. Toma o C19-GA2ox ole in pa henoca py and b anching | 5905 due o unc ional edundancy. In o a ies and axilla y buds, as expec ed om a de iciency in GA 2-oxidases, signi ican dec eases in inac i e p oduc s we e de ec ed (Tables 2, 3) in bo h GA pa hways (Fig.1). The e we e, howe e , some excep ions, o ins ance, GA8 in o a ies o L5 did no show any signi ican educ ion. The inac i a ion o GAs by GA 2-oxidases may include a second oxida ion s ep (Thomas e  al., 1999) p oducing GA ca aboli es (GA-ca ; Fig. 1). Since hese GA ca aboli es could no be quan i ied and hei le els also depend on he ac i i y o GA 2-oxidases, he simila i ies in inac i e GAs do no p eclude di e ences in GA 2-oxidase ac i i y. Wi h espec o ac i e GAs, while GA4 le els inc eased in bo h issues o ansgenic plan s, GA1 le els emained unal e ed, which was unexpec ed since bo h GA4 and GA1 a e subs a es o GA 2-oxidases. This may be due o he complexi y o he pa hway (Fig.1). The A d-2d0d5 OVARY WEIGHT (mg) 1 10 100 Pollina ed Emascula ed GA2ox1 d-2d0d5 RELATIVE mRNA LEVEL 0,1 1 10 100 1000 Pollina ed Emascula ed B GA2ox2 d-2d0d5 RELATIVE mRNA LEVEL 0,1 1 10 100 Pollina ed Emascula ed C GA2ox3 d-2d0d5 RELATIVE mRNA LEVEL 0,01 0,1 1 10 Pollina ed Emascula ed D GA2ox4 d-2d0d5 RELATIVE mRNA LEVEL 0,1 1 10 100 Pollina ed Emascula ed E GA2ox5 d-2d0d5 RELATIVE mRNA LEVEL 0,1 1 10 100 1000 Pollina ed Emascula ed F Fig.7. Exp ession o GA2ox genes in hand-pollina ed and unpollina ed o a ies in WT plan s. The SlAc in gene was used as an in e nal con ol o each issue and alues a e means o h ee biological eplica es ±SE. Resul s a e exp essed ela i e o he ansc ip le el, a d-2 s age (2 d be o e an hesis), se as 1.0. (A) Weigh o o a ies om pollina ed and emascula ed lowe s used in his expe imen ; (B) GA2ox1; (C) GA2ox2; (D) GA2ox3; (E) GA2ox4; (F) GA2ox5.