1 3

Plant Mol Biol (2014) 84:399–413DOI 10.1007/s11103-013-0139-8

Characterization of gibberellin‑signalling elements during plum fruit ontogeny defines the essentiality of gibberellin in fruit development

Islam El‑Sharkawy · Sherif Sherif · Walid El Kayal · Abdullah Mahboob · Kamal Abubaker · Pratibha Ravindran · Pavithra A. Jyothi‑Prakash · Prakash P. Kumar · Subramanian Jayasankar

Received: 1 April 2013 / Accepted: 3 October 2013 / Published online: 20 October 2013 © Springer Science+Business Media Dordrecht 2013

PslGID1 proteins structure, Y2H and BiFC assays indi-cated that plum GA-receptors can form a complex with AtDELLA-repressors in a GA-dependent manner. Moreo-ver, phenotypical-, molecular- and GA-analyses of various Arabidopsis backgrounds ectopically expressing PslGID1 sequences provide evidence on their role as active GA-sig-nalling components that mediate GA-responsiveness. Our findings support the critical contribution of GA alone or in association with other hormones in mediating plum fruit growth and development.

Keywords Fruit development · GA content · GA-signalling · Plum · PslGID1/DELLA interaction · Hormones cross-talk

Introduction

Fruit development involves a complex and coordinated interplay of cell division, differentiation and expansion of sporophytic and gametophytic tissues. Therefore, the fac-tors controlling the transition of fruit growth through vari-ous developmental stages are of primary importance. Plant

Abstract Fruit growth is a coordinated, complex inter-action of cell division, differentiation and expansion. Gib-berellin (GA) involvement in the reproductive events is an important aspect of GA effects. Perennial fruit-trees such as plum (Prunus salicina L.) have distinct features that are economically important and provide opportuni-ties to dissect specific GA mechanisms. Currently, very little is known on the molecular mechanism(s) mediating GA effects on fruit development. Determination of bioac-tive GA content during plum fruit ontogeny revealed that GA1 and GA4 are critical for fruit growth and development. Further, characterization of several genes involved in GA-signalling showed that their transcriptional regulation are generally GA-dependent, confirming their involvement in GA-signalling. Based on these results, a model is presented elucidating how the potential association between GA and other hormones may contribute to fruit development.

PslGID1b (JX569804); gPslGID1b (JX569802); PslGID1c (JX569805); gPslGID1c (JX569803); PslGA3ox1 (JX569806).

Electronic supplementary material The online version of this article (doi:10.1007/s11103-013-0139-8) contains supplementary material, which is available to authorized users.

I. El-Sharkawy · S. Sherif · K. Abubaker · S. Jayasankar (*) Department of Plant Agriculture, University of Guelph, 4890 Victoria Av. N., P.O. Box 7000, Vineland Station, ON L0R 2E0, Canadae-mail: jsubrama@uoguelph.ca

I. El-Sharkawy · S. Sherif Faculty of Agriculture, Damanhour University, Damanhour, Egypt

W. El Kayal Department of Biological Sciences, University of Alberta, Edmonton, AB, Canada

W. El Kayal Faculty of Agriculture, University of Alexandria, Alexandria, Egypt

A. Mahboob Department of Biological Sciences, Brock University, St. Catharines, ON, Canada

P. Ravindran · P. A. Jyothi-Prakash · P. P. Kumar Department of Biological Sciences, National University of Singapore, Singapore 117543, Singapore

400 Plant Mol Biol (2014) 84:399–413

1 3

hormones are the major regulators of fruit development (Crane 1964; Nitsch 1970; Gillaspy et al. 1993). Collec-tively, hormone application, endogenous hormone quantifi-cation and genetic studies support the hypothesis that fruit development is largely coordinated by hormonal interac-tions. Although the potential impact of GA in stone-fruit development has already been acknowledged (Jackson 1968), very little is known on its response pathway(s) that mediate fruit growth and ripening. This may be due to the unavailability of adequate information in its signalling machineries and the diversity of cross-talk between GA and other hormones, which are often species/organ/developmen-tal stage-dependent (Sun and Gubler 2004). Several lines of evidence point out the involvement of GA alone or in asso-ciation with other hormones in regulating fruit development. Earlier studies showed that the endogenous GA and auxin contents significantly increased during fruit maturity and ripening, concomitant with the production of climacteric ethylene (Jackson 1968; Miller et al. 1987). These findings coupled with the stimulatory effect of exogenous GA or auxin in the fruit growth of several species, including Pru‑nus spp., strongly support their mutual roles during fruit ontogeny (Bukvoac 1963; Lodhi et al. 1969; Augustí et al. 1999; Serrani et al. 2007). On the other hand, GA alone can have pronounced effects on fruit development. The scar-city of bioactive GA during plum fruit growth caused seri-ous developmental disorders, including malformed flowers, reduced fruit size and delayed ripening (El-Sharkawy et al. 2012b). The involvement of GA in a diverse array of critical physiological functions may well be mediated by an equally intricate network of signalling cascades. The discovery of GID1 protein (Ueguchi-Tanaka et al. 2005; Griffiths et al. 2006; Nakajima et al. 2006) has led to the recognition of sev-eral key events in GA perception and subsequent degrada-tion of GA-repressor DELLA (Ueguchi-Tanaka et al. 2007). GID1 acts as a soluble GA-receptor, displaying high affin-ity for active GA. In the absence of GA, DELLA represses the GA-response pathway. GA-binding to GID1 results in a conformational change that allows GID1 to interact with DELLA. The GA–GID1 complex can block DELLA activity by direct protein–protein interaction with DELLA domain (Ariizumi et al. 2008) or via assembling the stable complex GA–GID1–DELLA that promotes DELLA association with a SCFGID2/SLY1 (Harberd et al. 2009), resulting in the subse-quent DELLA degradation and releasing the GA-response (Hedden 2008; Shimada et al. 2008). Recently, we reported the consequences of lack of active GA on enhancing the defi-cits in overall plum tree growth (El-Sharkawy et al. 2012b).

In the present study, based on quantification of endoge-nous GA content during plum fruit ontogeny, we were able to show that GA1 and GA4 make significant contributions to fruit growth. Further, we characterized the transcript accumulation profile of two novel GA-receptors (PslGID1b

and 1c) throughout fruit development and in response to different hormones occurred during fruit growth. We next investigated the GA-perception machinery to provide evi-dence to support that plum GA receptors are responsible for regulating the GA-responsiveness. Protein structures, yeast two-hybrid (Y2H) and bimolecular fluorescence comple-mentation (BiFC) assays demonstrated that PslGID1 pro-teins are active GA-signalling components that can assem-ble GA–PslGID1–DELLA complex in a GA-dependent manner. Moreover, wild-type (WT) and gid1a-1/gid1c-1 (g1-ac) Arabidopsis backgrounds ectopically expressing PslGID1s demonstrated that both genes are able to enhance and recover the GA-responsiveness in WT and g1-ac loss-of-function mutant, respectively, presumably through pro-moting DELLA degradation.

Materials and methods

Plum tissues and post-harvest treatments

Flowers and fruits from different developmental stages were harvested from Japanese plum (Prunus salicina L.) cultivar ‘Early Golden’ (‘EG’) as described previously (El-Sharkawy et al. 2007). All plant materials were fro-zen in liquid nitrogen and stored at −80 °C. To investigate the effect of different hormones in PslGID1s expression, mature ‘EG’ fruits were harvested before fruit ethylene pro-duction had risen. The fruits were treated with gibberellin (GA4-100 μM), GA-inhibitor Paclobutrazol (PAC-10 μM), ethylene (ethephon-4 mM) and auxin (IAA-100 μM). Untreated fruits were used as control. After treatments, fruits were incubated at room temperature and collected at 0, 3, 6 and 9 days post-treatment. Mixed tissues from three fruits were frozen for further analysis. To evaluate the effect of GA in fruit development, ‘EG’ trees were sprayed biweekly with GA3 from early S2- till S3-stage.

Quantification of bioactive GA

Fruit tissues from various developmental stages of plum and the vegetative rosettes of Arabidopsis were frozen in liquid nitrogen, lyophilized and stored at −20 °C. Extrac-tion and quantification assay of GA were as described previously (Seo et al. 2011). Bioactive GA forms were identified by comparison of retention times with those of authentic standards. The data presented are mean ± SD of nine replicates.

Isolation and in silico analysis of PslGID1 sequences

Based on the sequence similarity of various GIBBEREL-LIN-INSENSITIVE DWARF1 (GID1) sequences, a pair

401Plant Mol Biol (2014) 84:399–413

1 3

of degenerate primers (# 1 and 2; Table S1) was designed to amplify the P. salicina GID1 orthologs. The isolated fragments were cloned, sequenced and analyzed using BLAST (Altschul et al. 1997). Extension of the partial cDNA clones were carried out using the RACE kit (Inv-itrogen, Burlington, ON, Canada). Full-length amplifica-tion of cDNA sequences designated PslGID1b and 1c was carried out using the Platinum Taq DNA Polymerase High Fidelity (Invitrogen). Alignment of PslGID1 predicted proteins and the neighbor-joining tree construction were performed as described previously (El-Sharkawy et al. 2009). Genomic DNA was extracted from young leaves according to the DNeasy Plant Maxi Kit (Qiagen, Mis-sissauga, ON, Canada). Full-length genomic sequences including promoters of gPslGID1b and 1c were isolated using the Universal Genome Walker Kit (Clontech, Palo Alto, CA, USA). Promoter sequence analysis was per-formed using PLACE Signal Scan Search database (Higo et al. 1999).

Protoplast isolation and transient expression of PslGID1–GFP fusion proteins

The coding sequence of PslGID1b and 1c were cloned as a C-terminal fusion in frame with the GFP into the pGreenII vector using the BamHI site and expressed under the con-trol of the 35S promoter. Protoplasts from suspension cul-tured tobacco BY-2 cells were transfected with the different constructs with or without 100 μM GA3 and analyzed for GFP fluorescence using confocal microscopy as described previously (El-Sharkawy et al. 2009). All transient expres-sion assays were repeated three times.

RNA isolation and qRT-PCR

Total RNA extraction, DNase treatment, cDNA synthesis, and qRT-PCR reactions were performed as described previ-ously (El-Sharkawy et al. 2012b). Gene-specific primers were designed using Primer Express (v3.0, Applied Biosystems, Carlsbad, CA, USA) (primers # 3–18; Table S1). Three inde-pendent biological replicates for each reaction were run on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems) and each experiment was repeated three times. Transcript abundance was quantified using standard curves for both target and reference genes [PslAct (EF585293), AtAct (NM_121018)], which were generated from serial dilu-tions of PCR products from corresponding cDNAs. The data were present as an average of nine replicates (±SD).

Protein structure prediction

The two-dimensional (2-D) GA3/GA4–PslGID1b/1c images were generated using Poseview software. The

three-dimensional (3-D) crystal structures of Arabidopsis GA3- and GA4–GID1–GAI complexes (PDB ID: 2ZSH and ID: 2ZSI) were used as a template to obtain homology models by the MODELLER package. Resulting structures were optimized using the generalized born model for sol-vent of Amber12 software package. Parameters for GA3 and GA4 were developed using the general amber force-field. The binding energy of the respective DELLA proteins to individual GA3/GA4–PslGID1 complex was then calcu-lated through obtaining the electrostatic components of the thermodynamic cycle corresponding to a protein–protein binding event (Wang et al. 2004). The electrostatic energies were calculated using the numerical Poisson–Boltzmann solver algorithm APBS 1.3 (Baker et al. 2001).

Yeast two-hybrid assay (Y2H)

Y2H assays were performed with the Matchmaker Gold Yeast two-hybrid System (Clontech). PslGID1b and 1c full-length ORFs were inserted into the BamHI-PstI and NdeI-BamHI sites of the pGBKT7 bait vector (GAL4 binding-domain), respectively, to generate DBD: GID1b and DBD: GID1c. AtGAI and AtRGL1 cDNAs were fused into the NdeI-BamHI site of the pGADT7 prey vector (GAL4 acti-vation-domain) to generate AD: GAI and AD: RGL1. Bait and prey vectors (100 ng) with the corresponding fusion pro-teins were then transformed into Y2HGold and Y187 yeast strains, respectively, using Yeastmaker yeast transformation system 2. One positive colony was picked from each vector type, mixed together in 2X-YPDA. The mated culture was incubated overnight at 30 °C with shaking (200 rpm), har-vested and resuspended in distilled water. Five microliters of the resuspended yeast were spread in a well (96-well plates) containing DDO/X/A medium in the presence or absence of 100 μM GA3. The growth of yeast colonies was observed 3 days after incubation at 30 °C. All assays were repeated at least three times.

Bimolecular fluorescence complementation (BiFC) assay

For constructs used in the BiFC experiment, the N-ter-minal (pSAT1-N) and C-terminal (pSAT1-C) EYFP vec-tors were used. Full-length PslGID1b and 1c ORFs were inserted into the BglII-BamHI of the pSAT1-N vector. Consequently, AtGAI and AtRGL1 were fused into the SacI-BamHI site of the pSAT1-C vector. Various combi-nations of plasmids encoding N_ and C_EYFP at similar concentration were mixed and co-transformed into pro-toplasts obtained from suspension cultured tobacco BY-2 cells in the presence or absence of 100 μM GA3. Typi-cally, 0.3 ml of protoplast suspension (0.3 × 106 cells) was transfected with 50 μg of shared salmon sperm car-rier DNA and 30 μg of each plasmid DNA. Transfected

402 Plant Mol Biol (2014) 84:399–413

1 3

protoplasts were incubated at 25 °C for at least 16 h before assaying for YFP activity using confocal microscopy. All assays were repeated at least three times.

Plant transformation

Full-length PslGID1b and 1c ORFs were fused into the BglII/BstEII site of the pCambia1305.1 binary vector. The resulting vectors were transformed into A. tumefaciens and employed for Arabidopsis transformation as described previ-ously (El-Sharkawy et al. 2012b). The two PslGID1 cDNAs were introduced into several Arabidopsis backgrounds, includ-ing WT and g1-ac double mutant (Griffiths et al. 2006). T3 homozygous independent lines from each group were grown (24 plants/group/growth condition) under long day (LD)-conditions (16:8 h light/300 μmol m−2 s−1; 23:18 °C and 65 % relative humidity) or short day (SD)-conditions (8:16 h light and 23:18 °C). Plant materials were frozen in liquid N2 immediately after collection and stored at −80 °C until use.

Results

Analysis of bioactive GA content during fruit development

Plum fruit growth followed a typical double sigmoid curve (Fig. 1a), in which four developmental stages (S1–S4) can be clearly recognized (El-Sharkawy et al. 2007). In order to evaluate the contribution of GA on fruit growth, ‘EG’ trees were sprayed with GA3. GA application resulted in significant increase in fruit size (~62 %) and weight (~94 %) along with slight accelera-tion in fruit development, 3.3 ± 1.2 days ahead of the control (Fig. 1b). Determination of active GA content during plum fruit ontogeny revealed the presence of two out of the three active GA forms (GA1 and GA4); however, GA3 was either extremely low or out of detec-tion limit (Fig. 2a). GA1 and GA4 were high at bloom, but dropped drastically post-fertilization (~7 DAB). A clear divergence between the accumulation behaviors of the two GA forms occurred thereafter. GA1 sharply

Fig. 1 a Stages of ‘EG’ plum fruit development and growth curves. The x-axis represents the developmental stages indicated by number of days after bloom (DAB). The y-axis refers to the changes on fruit fresh weight (g) and diameter (cm) during development. The develop-

mental stages include flower (FL), fruit initiation (FI), and stages 1–4 (S1–S4) of fruit development. b Close-up views of ‘EG’ fruits before and after GA application along with the alterations in fruit size and weight due to GA treatment

403Plant Mol Biol (2014) 84:399–413

1 3

increased at ~10 DAB and progressively decreased after-ward through S2-stage (42–52 DAB). At S3 (~62 DAB), GA1 consistently increased, reaching maximal levels at climacteric stage (~82 DAB). However, GA4 gradually increased from fruit-set reaching maximal levels by S3 (~72 DAB) and then steadily decreased during ripening. The concentration of bioactive GAs is determined by the balance between their rates of biosynthesis and deac-tivation, in which GA3ox and GA2ox genes encode key enzymes of biosynthesis and inactivation of GA, respec-tively (Yamaguchi 2008). Previously, we have reported the accumulation of PslGA2ox in the same developmen-tal stages studied in the present work (El-Sharkawy et al., 2012b). To further assess whether the profile of active GA development is associated with the accumulation of

GA-signalling pathway genes, three PslGA3ox sequences were identified and their expression was studied during fruit development. The three PslGA3oxs exhibited gener-ally similar accumulation pattern (data not shown), thus only the expression PslGA3ox1 that showed the closest association with GA accumulation was selected for fur-ther study (Fig. 2b). In general, the expression profile of PslGA3ox1 was consistent with the GA-regulation model. Throughout various developmental stages, its transcrip-tion clearly displayed a negative correlation with active GA content. The strongest signal of PslGA3ox1 was detected immediately post-fertilization (~7 DAB), when the active GA levels were low. During various develop-mental points of S3 and S4, when the pulp (mesocarp) separates from seed (endocarp + embryo), PslGA3ox1 showed diverse accumulation profile between the two tis-sues. In the pulp, its transcription was initially high and gradually declined till post-climacteric stage (~83 DAB). However in seed, its mRNA levels were generally moder-ate and slightly increased with fruit maturity.

Isolation and structural characterization of PslGID1 sequences

Two putative GID1 orthologs with similarity to hormone-sensitive lipase were isolated from plum. Their deduced amino acid sequences were 74 % identical to each other and displayed high identity to Arabidopsis GID1b (78 %) and GID1c (80 %), thus they were designated PslGID1b and 1c, respectively. PslGID1b and 1c sequences are 1,167-bp and 1,264-bp in length, respectively, with pre-dicted open reading frame encoding a protein of 344-amino acids. Isolation of corresponding full-length genomic sequences, gPslGID1b and 1c, indicated that both genes have a single intron at the same position in the N-terminal portion of the predicted sequences (Fig. S1A). Both gPslGID1 promoter sequences (~1.6-kb) hold com-mon predictive auxin and gibberellin cis-acting regula-tory elements related to various physiological responses and presumed to be involved in PslGID1 transcription regulation (Table S2). Additionally, other important motifs were detected uniquely in one out of the two pro-moter sequences. For instance, several ethylene-elements were detected in gPslGID1b and lignin-biosynthesis ele-ments were detected in gPslGID1c promoters. Align-ments of PslGID1 deduced amino acid sequences with their homologues in Arabidopsis (Fig. S1B) highlighted a number of conserved motifs and structural similari-ties that are common within GID1 gene family, includ-ing all the amino acid residues involved in GA-binding and in interaction with DELLA (Nakajima et al. 2006; Ueguchi-Tanaka et al. 2007). Analysis of GID1 sequences from different plant species revealed the absence of any

Fig. 2 a Changes in GA1 and GA4 levels during plum fruit develop-ment, including flowers at bloom (~0 DAB), post-fertilization (~7 DAB), early fruit formation (~10 and 22 DAB), S1-fruit (~27 and 37 DAB), S2-fruit (~42 and 52 DAB), S3-fruit (~62 and 72 DAB), and S4-fruit (~78 and 82 DAB). During S1–S4 fruit stages, the GA assays were determined in the fruit pulp only. Values represent the mean ± SE derived from six independent replicates. Black and white symbols refer to GA1 and GA4, respectively. The gray arrows repre-sent values that were below the detection limit. The y-axis refers to the progression on GA1 and GA4 contents (ng g−1 dry weight). b Steady-state transcript levels of PslGA3ox1 assessed by qRT-PCR during ‘EG’ flowering and fruit development. During S1 and S2 stages, the expression was determined in the whole fruit; however, during S3 and S4 the expression was determined in pulp (Black filled bars) and in seeds (Gray filled bars). The experiments were car-ried out in three biological replicates. Standard curves were used to calculate the numbers of target gene molecules per sample, which were then normalized relative to PslAct expression. Error bars rep-resent standard deviation. The y-axis refers to the mean molecules of PslGA3ox1 per reaction/mean molecules of PslAct. The x-axis in each figure represents the developmental stages indicated by the num‑ber of days after bloom (DAB)

404 Plant Mol Biol (2014) 84:399–413

1 3

feasible targeting sequence that can signify the localiza-tion of such protein in the plant cell. Our results showed that both PslGID1 proteins exhibited similar localization compartment, where the GFP distribution was spread throughout the cytosol and nucleus (Fig. S2). Treatment of BY-2 cells by GA3 prior constructs transfection did not alter the localization compartment (data not shown). In order to classify PslGID1s, a phylogenetic tree that includes various GID1 orthologs was constructed (Fig. 3).

The dendrogram divided the dicot proteins into two main clades based upon sequence conservation. The two clades containing the plum and other plant GID1s are sup-ported by high bootstrap values. PslGID1b is located in an independent subgroup that has uniquely GID1b related proteins and PslGID1c is classified within the group of GID1ac subfamily. Analysis of sequences from different plant species indicated the presence of at least three dif-ferent GID1 orthologs within their genome. This could include two members of GID1ac subfamily and one mem-ber belonging to GID1b subfamily as in Arabidopsis and grape. However, the genomes of the majority of plants, analyzed so far, contain only one ortholog that belongs to GID1ac subfamily and two members classified within GID1b subfamily as in tomato, apple and strawberry. Sequence data mining in peach genome (Phytozome v8.0: Search), the closest genome to plum, identified the orthologs of GID1b and 1c as the only putative GID1-like genes within its genome.

PslGID1 expression during fruit ontogeny

To strengthen our understanding of GA’s role in fruit development, the relative expression levels of PslGID1b and 1c were assessed (Fig. 4). Both PslGID1s exhibited similar accumulation pattern until S2 and diverged at sub-sequent adult stages (S3 and S4). Both transcripts were initially abundant in flower buds (~ −4 DAB) followed by sharp decline of their transcription at bloom. The transcrip-tion of both genes sharply peaked soon after fertilization (~7 DAB). During early fruit formation (10–22 DAB), the accumulation of the two genes gradually decreased, reach-ing basal levels by S2-end (~52 DAB). During S3 and S4, PslGID1b and 1c were generally higher in developing seed than pulp. Through S3 (57–77 DAB), the expression pat-tern of the two transcripts was slightly different. PslGID1b accumulated in both pulp and seed similarly. In early S3 its signal was considerably high and gradually decreased along with maturity, reaching basal levels by S3-end. However, PslGID1c showed a clear divergence in its accu-mulation pattern among the two fruit tissues. In the pulp, PslGID1c was temporally quite constitutive; however in seed, it was high at the beginning of S3 and declined subsequently. During fruit ripening (S4, 78–83 DAB), PslGID1b was scarcely expressed in non-climacteric fruit (~78 DAB) and increased in abundance along with the pro-gression in fruit ripening, reaching relatively high levels in the pulp of post-climacteric fruit; however, in the seed, its transcription peaked at the climacteric-phase (~82 DAB). Conversely, high PslGID1c signal was detected in both tis-sues of non-climacteric fruit (~78 DAB) and consistently decreased to undetectable levels by post-climacteric-phase (~83 DAB).

Fig. 3 Phylogenetic tree of different GID1s subfamilies. The tree was constructed by comparing the amino acid sequences of 21 GID1 proteins from different plant species that belong to monocots and dicots, including P. salicina PslGID1b, PslGID1c; A. thaliana AtGID1a (NP_187163), AtGID1b (NP_191860), AtGID1c (NP_198084); V. vinifera VvGID1a (AFG17072), VvGID1b (XP_002271700), VvGID1c (XP_002265764); M. domestica MdGID1b.1 (AFD32891), MdGID1b.2 (AFD32892), MdGID1c (AEC04638); S. lycopersicum genome database SlGID1b.1 (SL1.00sc07266), SlGID1b.2 (SL1.00sc03996), SlGID1c (SL1.00sc05858); F. vesca genome database FvGID1b.1 (scf0513173), FvGID1b.2 (scf0513123), FvGID1c (scf0513061); O. sativa OsGID1 (NP_001055520); Z. mays ZmGID1 (CAP64327); S. bicolor SbGID1 (CAP64327); H. vulgare HvGID1 (CAO98733) based on full-length amino acid sequence. Bootstrap confidence values from 1,000 replicates are indicated above branches. GID1s belong to monocots and dicots are specified

405Plant Mol Biol (2014) 84:399–413

1 3

Effect of different hormones on PslGID1s expression pattern

In view of the PslGID1s accumulation profile during fruit ontogeny, the question was raised whether either of the two PslGID1s is regulated by other hormones that are associ-ated with fruit development. To address this hypothesis, the expression profiles of the two genes were assessed in fruits treated with GA, PAC, C2H4 and IAA. Both PslGID1b and c mRNAs steadily declined along with time post GA-treat-ment. Contrarily, blocking GA-biosynthesis by PAC sig-nificantly enhanced the transcription of both genes within the same periods (Fig. S3A, B). Only PslGID1b transcripts increased due to ethylene treatment; whereas, PslGID1c did not visibly respond to the presence of ethylene (Fig. S3C). IAA treatment activated the transcription of both genes in the same manner (Fig. S3D).

Molecular modelling of PslGID1 proteins

The predicted 2-D structures of GA–PslGID1 complex demonstrated that both active GA forms (GA3 and GA4) bind to the PslGID1b and 1c in the same binding pocket, but in a slightly different manner (Fig. 5a, c). GA4 is some-what more hydrophobic and interacts with the Phe-27 residue of PslGID1. Both GA3 and GA4 interact with the sidechains of Ser-115, Ser-190 and Tyr-126; as well as the backbone of Phe-237 of PslGID1. The predicted 3-D struc-tures (Fig. 5b, d) as well as our calculations of the binding energies of the five different forms of Arabidopsis DELLA to the PslGID1b and 1c showed that both proteins bind DELLA in the presence of GA (GA3 and GA4) with no preference to a particular form of DELLA-binding, which

is mainly due to the conserved DELLA motif. The binding energies of AtDELLA to PslGID1 were found to be within the range of −10 kcal mol−1.

PslGID1: AtDELLA interactions is GA-dependent

In Arabidopsis, perception of GA by its receptors (GID1) promotes the direct interactions between GID1 and DELLA (Yasumura et al. 2007; Ariizumi et al. 2008). To determine whether PslGID1 possess a comparable func-tion as those of Arabidopsis, the interactions between plum GA-receptors (GID1b and 1c) and Arabidopsis GA-repressors (GAI and RGL1) were assessed in yeast system in the presence or absence of GA (Fig. 6a). The absence of GA completely abolishes PslGID1-interaction with AtDELLA. However, GA presence in the reaction illustrated the essential role of GA to assemble the GA–PslGID1–AtDELLA complex in yeast. To provide addi-tional evidence for yeast results, we attempted to visual-ize the direct GA-triggered interactions between PslGID1 and AtDELLA using BiFC approach (Lee et al. 2008). Tobacco protoplasts supplemented with 0 and 100 μM GA3 were co-transfected with the various combinations of NY: GID1b/1c and CY: GAI/RGL1 constructs (Fig. 6b). As in the yeast system, the YFP signal, caused by interac-tion between GID1b/1c and GAI/RGL1 was only detected in protoplasts pre-treated with GA. Although, the untreated cells should contain endogenous GA, no fluorescence sig-nals were observed in cells grown in GA-free medium (data not shown). This is probably due to the scarcity of active GA content that will not be sufficient to promote interaction between the fluorescence-labeled PslGID1 and AtDELLA.

Fig. 4 Steady-state transcript levels of PslGID1b and 1c assessed by qRT-PCR dur-ing ‘EG’ flowering and fruit development. Other details are as described in Fig. 2

406 Plant Mol Biol (2014) 84:399–413

1 3

Ectopic expression of PslGID1 in Arabidopsis demonstrates its involvement in the GA-signalling pathway

Previous results established that PslGID1s are active GA-receptors that can interact with AtDELLA in a GA-dependent

manner. However, they do not show the activation of down-stream signalling mechanism caused by assembling the stable GA–PslGID1–DELLA complex. In Arabidopsis, GA enhances plant growth characteristics due to stimulat-ing interactions between GID1 and DELLA, which triggers the degradation of DELLA and activates the overall GA-responses (Yasumura et al. 2007; Ariizumi et al. 2008). To test whether PslGID1 proteins possess similar GA functions in planta, transgenic Arabidopsis background (Col) express-ing PslGID1b (GID1b/w) or 1c (GID1c/w) were generated. A number of independent lines were obtained and only two representatives from each transformation were selected for further studies on the basis of the highest transgene levels (Fig. 7a). Increasing the amount of GA-receptors within the plant should activate the GA-signalling machinery, which consequently can trigger the overall growth (Yasumura et al. 2007; Ariizumi et al. 2008). Therefore, to better character-ize the consequential phenotypes, the expression of some Arabidopsis genes involved in GA-metabolism was assessed. GA-homeostasis in a variety of plant species has been found to be tightly linked to the activities of enzymes involved in a feedback mechanism that regulates GA-biosynthesis and catabolism (Sponsel and Hedden 2004; Sun and Gubler 2004). When GA-levels are high, genes encoding enzymes

Fig. 5 The predicted 2-D structures establishing the GA–PslGID1s interactions a and c. PslGID1s amino acid residues that interact with GA3 a and GA4 c molecules are indicated. Green, red and blue amino acids represent hydrophobic, negatively and positively charged PslGID1s residues, respectively. The dotted lines signify a hydrogen bond. The 3-D modelling structure of b GA3–PslGID1b/c–GAI based on 2ZSH template and d GA4–PslGID1b/c–GAI based on 2ZSI tem-plate. The GAI–DELLA domain and PslGID1b/1c are indicated in red and blue, respectively. The bound GA3 and GA4 molecules are repre-sented as a space-filling model with carbon in cyan and oxygen in red

Fig. 6 Interactions between PslGID1b/1c and AtGAI/AtRGL1 using a Y2H and b BiFC approaches. a Y2H assays were performed using PslGID1s as bait in Y2HGold yeast strain and AtDELLAs as prey in Y187 yeast strain. The mated yeast were grown in 96-well plates containing DDO/X/A liquid medium in the presence or absence of 100 μM GA3. b GA-treated tobacco cells were co-transfected with the indicated constructs. YFP fluorescence was visualized using confocal laser scanning microscopy. Empty BiFC vectors (NY/CY), (NY: PslGID1/CY) and (NY/CY: AtDELLA) were co-transfected into protoplast as negative controls. All experiments were repeated at least three times. Fluorescence (left panel) and merged fluorescence to light micrographs images (right panel) are showed to illustrate the location of interaction. Bar 10 μm

407Plant Mol Biol (2014) 84:399–413

1 3

for GA-biosynthesis (AtGA3ox and AtGA20ox) and enzymes for GA-inactivation (AtGA2ox) are subject to negative-feedback and positive-feedforward regulation, respectively (Thomas et al. 1999; Xu et al. 1999). Consistent with the GA-regulation model, the accumulation of AtGA3ox1 and AtGA20ox1 were lower than in WT, while AtGA2ox8 stead-ily increased in GID1/w plants (Fig. 7b). We further charac-terized the molecular basis of the GA-signalling enhance-ment in GID1/w plants by quantifying the levels of the three AtGID1 mRNAs (Fig. 7c). PslGID1-overexpression leads to negative-feedback regulation of AtGID1s orthologs.

Relative to WT, GID1b/w plants exhibited ~90-, 24- and 13-fold reduction in AtGID1a, 1b and 1c levels, respectively, while GID1c/w displayed ~150-, 28- and 12-fold reduction, respectively. Ectopic expression of PslGID1 in WT Arabi‑dopsis led to phenotypical growth performance consistent with altered GA-signalling under various growth conditions, LD or SD (Fig. 7d, e). Analysis of transgene profile showed that the level of PslGID1 is generally correlated positively with the severity of the phenotype; the higher transgene lev-els the more activation in growth performance. Furthermore, the stimulation of the various growth traits was somehow

Fig. 7 a PslGID1 transgene accumulation, b the accumulation lev-els of the GA-metabolism mRNAs and c GA-receptors in WT and the different transgenic plants overexpressing PslGID1b and 1c. The transcripts accumulation was determined using qRT-PCR on three biological replicates. Standard curves were used to calculate the num-bers of target gene molecules per sample, which were then normalized

relative to AtAct expression. ND non-detectable. Aerial portions of WT and the different transgenic mutants grow under LD- (d) and SD- (e) conditions. Bars 10 cm. f GA-responsiveness of WT, GID1b/w and GID1c/w plants grown under LD. The plants were divided into three groups: control, sprayed with GA4 (10 μM) or PAC (10 μM). The measurements are the means (±SD) of approximately 24 plants

408 Plant Mol Biol (2014) 84:399–413

1 3

dependent on the ortholog of PslGID1-introduced. Despite the transgene level and under different growth conditions, all GID1b/w plants exhibited faster germination, higher germi-nation rate, earlier silique maturity and longer siliques; how-ever, GID1c/w plants showed accelerated vegetative growth associated with notably longer internodes (Tables S3a and b). Other developmental traits such as transition to flowering, stamen length and fertility were not significantly altered due to transgene presence, except when the plants were exposed to stressful SD-conditions (Table S3b). Under these condi-tions the differences between WT and the transgenic plants were more distinguishable. Moreover, pistil length character was slightly altered, but only in association with the pres-ence of extremely high transgene levels despite the type of ortholog introduced as in case of L.5 and L.1 for GID1b/w and 1c/w, respectively. To confirm the GA-receptor func-tions, we compared the GID1/w plants’ responsiveness to GA by applying GA and PAC (Fig. 7f). All GID1/w trans-genics exhibited more pronounced growth than WT as deter-mined by plant height in response to GA or PAC treatments.

Overexpression of PslGID1 rescues g1‑ac double mutant

To further validate that plum GA-receptors biologically func-tion as those of Arabidopsis, we tested the ability of PslGID1 to recover AtGID1s absence through rescue of GA-respon-siveness in GA-insensitive g1-ac double mutant (Griffiths et al. 2006). A number of independent transgenic lines were generated and two representatives from each transforma-tion that exhibited the highest transgene level were selected for further studies (Fig. 8a). To assess whether PslGID1-overexpression can disturb the GA-response pathway, the alteration in GA-metabolism genes (AtGA2ox8, AtGA3ox1 and AtGA20ox1) was determined in untransformed g1-ac mutant as control against GID1/g plants. Ectopic expression of PslGID1 in g1-ac mutant considerably reduced GA-bio-synthesis-related transcripts and increased that of GA-inacti-vation-related mRNA in all GID1/g plants (Fig. 8b). To fur-ther confirm the molecular data, we assessed the alteration in active GA content (GA1 and GA4) in the rescued plants comparing with those of control g1-ac and WT to determine

Fig. 8 a PslGID1 transgene accumulation and b the accu-mulation levels of the different GA-metabolism mRNAs in g1-ac double mutant and the different transgenic plants over-expressing PslGID1b and 1c. The transcripts accumulation was determined using qRT-PCR on three biological replicates. Other details are as described in Fig. 7. c Active GA1 and GA4 contents in g1-ac, WT, GID1b/g and GID1c/g Arabidopsis backgrounds. Aerial portions of g1-ac double mutant and the different transgenic plants grow under LD (d) and SD (e) conditions. ND non-detectable. Bars 10 cm

409Plant Mol Biol (2014) 84:399–413

1 3

the level of recovery (Fig. 8c). It has been demonstrated that the loss of feedback regulation of GA-biosynthesis resulted in high accumulation of the active GA compared with WT (Ueguchi-Tanaka et al. 2005). Likewise, the concentra-tion of GA4 and GA1 were elevated by ~10- and 50-fold, respectively, in g1-ac mutant compared with WT. In case of GID1/g, GA level was consistent with the growth habit of transgenic plants. Compared to g1-ac, GID1b/g and 1c/g plants exhibited much less GA content, but still considerably higher than WT. Moreover, selected GID1b/g and 1c/g plants were phenotypically characterized for some well-known GA-related traits under LD and SD conditions against that of control (g1-ac) as well as WT. PslGID1-overexpression in g1-ac generally rescued various growth defects of the mutant. Under LD, g1-ac seeds exhibited low germination rate and longer period to germinate either with or without GA (~35 days). PslGID1-overexpression notably improved all germination characteristics, compared to WT (Fig. S4, Table S4a). Moreover, SD-photoperiod destructively reduced seed germination, which was more pronounced in control g1-ac (failure of germination). However, PslGID1 presence repressed the germination deficiency nature, which remained slightly less than WT. Further PslGID1 presence resulted in significant increase in plant height (Fig. 8d) estimated by ~1.9- and 2.5-times for GID1b/g and 1c/g, respectively. Gen-erally, SD-conditions drastically decreased stem elongation and shoot growth in all plants; however, GID1/g and WT were least affected (Fig. 8e, Table S4a).

GA is involved in key developmental events leading to flo-ral determination and reproductive growth (Cheng et al. 2004; Tyler et al. 2004). Relative to g1-ac, GID1/g plants exhib-ited considerable acceleration in all flowering related char-acteristics (Table S4b). Under LD, GID1b/g and 1c/g plants started flowering at least 17 ± 1.5 and 12 ± 0.9 days before control mutant, similar to WT flowering time. SD substan-tially retarded the transition to flowering in all Arabidopsis backgrounds; however, GID1b/g and 1c/g started flowering 15 ± 0.8 and 10 ± 1.9 days earlier than control g1-ac with no significant difference between GID1b/g and WT. Moreover, WT Arabidopsis flowers exhibit a coordinated flower structure that ensures proper self-pollination. Mutant of g1-ac displayed generally smaller flowers along with major defect in flower structure through producing stamens much shorter than pistils. In self-pollinated species, such variation between the stamens and pistil can cause a major reduction in fertility. Under LD, PslGID1-overexpression visibly suppressed all the defects in g1-ac flowers. Both GID1/g transgenics displayed larger flow-ers with accelerated stamen and pistil growth (Fig. 9a). SD generally reduced the size of various floral organs and that was more dramatic in g1-ac mutant. Although, GID1/g flow-ers under SD were slightly smaller than WT, they all exhibited proper floral patterning without defects. Under different con-ditions, pistil and stamen lengths of all GID1/g flowers were

restored to a comparable size with WT. The improvement in various growth aspects, especially flowering characteristics, in GID1/g resulted accordingly in enhancement of the vari-ous reproductive traits (Table S4b). In all GID1/g plants under both growth conditions, the periods to silique emergence as well as silique maturation were similar to WT and noticeably shorter than control g1-ac. Furthermore, defects in silique length (Fig. 9b) and fertility were considerably repressed in GID1/g compared to control mutant, but remained less than WT under all conditions. Griffiths et al. (2006) reported that GA-responsiveness in the dwarf phenotype of the g1-ac dou-ble mutant is perturbed, since GA application was not able to alter the mutant growth performance. However, increas-ing the amount of GA-receptors should re-establish the GA-responsiveness, which consequently can trigger the overall growth behavior (Yasumura et al. 2007; Ariizumi et al. 2008). In our study, g1-ac root length was at least ~45 % shorter than WT. Relative to g1-ac, GID1b/g and 1c/g root lengths were enhanced by ~54 and ~71 %, respectively (Fig. 10; Table S5). To determine whether the GA-responsiveness in GID1/g was re-established, we investigated the effect of GA on all Arabi‑dopsis backgrounds root growth behavior. However, due to the elevated GA levels in g1-ac mutant that may be saturated for some responses (Fig. 8c), we investigated the effect of GA in g1-ac seedlings with or without PAC. PAC inhibited the growth of g1-ac to some extent. When GA was added, the roots of WT, GID1b/g and 1c/g seedlings were significantly longer than their corresponding untreated plants. However,

Fig. 9 Recovery of floral organs and siliques growth due to PslGID1-overexpression on g1-ac. Close-up views of WT, GID1b/g, GID1c/g and g1-ac flowers (a) and siliques (b) from plants grow under LD and SD-conditions. Sepals and petals were removed from flowers to reveal the anthers and pistil. Siliques were collected from adult plants have approximately 10 % shattered siliques. Bar 1 mm

410 Plant Mol Biol (2014) 84:399–413

1 3

g1-ac did not significantly respond to GA, which confirms that both PslGID1b and 1c are able to restore the GA-respon-siveness in the mutant.

Discussion

Earlier studies reported the potential contribution of GA in the regulation of fruit development in different tree fruit spe-cies (Bukovac and Nakagawa 1967; Yamaguchi and Taka-hashi 1976; Bukvoac and Yuda 1979; Vargas et al. 2013; Mesejo et al. 2013). In continuation of our earlier studies on the role of GA in stone-fruit growth (El-Sharkawy et al. 2012b), we investigated its possible involvement in fruit development. In tree fruits, the proper establishment of fruit development is dependent on coordinated levels of GA at the appropriate developmental stages (Serrano et al. 2007). Application of GA on plum trees resulted in visible increase in fruit size and weight. On the other hand, inadequate quan-tities of GA resulted in a series of distortions in flowering and general reproduction (El-Sharkawy et al. 2012b). Com-paring the profiles of bioactive GA accumulation with the physiological events that characterize plum fruit develop-ment suggested that the requirement for specific GA form is dependent on the stage of fruit development. GA4 mediates fruit growth throughout fruit life, excluding ripening stage. However, GA1 plays more extensive roles during early fruit formation or later through adult phases (S3 and S4). The accumulation pattern of the two GA forms during fruit devel-opment is well organized to provide adequate GA during fruit ontogeny. The absence of one form is recovered by con-comitant increase of other form. Studies on the effect of GA on plant growth and development may be hindered by their low abundance and variation in forms, time and localization. Assessing the expression of genes involved in GA-signalling provides easier validation for such studies. Although tran-scripts of genes for GA-signalling intermediates are detected in all tissues throughout fruit development, their expression profile appears to be organ- and developmental stage-dependent. Therefore, a detailed analysis of the expression

pattern of these genes was necessary to elucidate their roles in modulating GA-response during fruit development. In plum, the transcript profiles of the various GA-signalling proteins tested so far were generally aligned with their regu-lation model. Usually, when GA levels are high, genes encoding for GA-biosynthesis and perception are subject to negative-feedback regulation (Griffiths et al. 2006; Yamagu-chi 2008). In general, comparing the expression data with the physiological aspects that characterize fruit development suggested that the transcript accumulation of the different GA-signalling intermediates is usually associated with the physiological signature events that are triggered in a GA-dependent manner, mainly during flowering. In contrast, PslGID1s exhibited abnormal expression profiles at some developmental points that are not consistent with GA accu-mulation. To understand the cause of this deviation, we investigated the influence of GA and other hormones—auxin and ethylene—on PslGID1s regulation. In flowers, appar-ently the transcription of PslGID1 is regulated at bloom and post-fertilization in GA- and auxin-dependent manners, respectively. Several lines of evidence, including (1) the find-ings of many GA/auxin cis-acting elements and transcription factor binding sites in PslGID1 promoters, (2) PslGID1 response to GA and auxin treatments, (3) and the accumula-tion profiles of several auxin perception elements in the same developmental stages studied in the present work (El-Sharkawy et al. 2010, 2012a) provide further credence to the hormonal cause of PslGID1s regulation. The scarcity of PslGID1 at bloom, which coincides with the inhibition of PslGA3ox1 and the abundance of PslGA2ox (El-Sharkawy et al. 2012b), is correlated well with the large accumulation of GA, particularly GA4 that is the most effective form lead-ing to proper florigenesis and reproductive growth (Dunberg and Odén 1983). However, the abundance of PslGID1s along with the sharp decline in GA content and the subsequent accumulation profile of the different GA-metabolism genes post-fertilization indicate that these collective patterns could be partially due to auxin that accumulates post-fertilization (Miller et al. 1987; El-Sharkawy et al. 2010, 2012a, b). The availability of auxin within the plant is usually associated

Fig. 10 GA-responsiveness of WT, GID1b/g, GID1c/g and the control g1-ac mutant plants grown under LD-conditions. Representative primary roots of 15 days-old seedling from different genotypes grown in the presence (+) or absence (−) of 0.5 μM GA4. The control mutant seedlings were pre-grown in the presence (g1-ac/P) or absence (g1-ac) of 1 μM PAC. Bar 10 mm

411Plant Mol Biol (2014) 84:399–413

1 3

with the induction of GA-biosynthetic transcripts and the reduction of GA-catabolic mRNAs (Ngo et al. 2002; Ross et al. 2002; Wolbang et al. 2004). However, the effect of auxin on GA-responses seems to be species/organ/develop-mental stage-dependent (Sun and Gubler 2004). Accord-ingly, it is tempting to speculate that in floral organs auxin might inhibit GA-responsiveness. Throughout fruit develop-ment, it is almost certain that the series of modifications that make the fruit proceed through the consequent developmen-tal stages involve many different metabolic pathways. Previ-ous studies suggested a cross-talk between GA and other hormones such as auxin and ethylene in the regulation of dif-ferent fruit development events such as fruit-set, -growth and -ripening (Ogawa et al. 2003; Fleet and Sun 2005; de Jong et al. 2009; Csukasi et al. 2011). In addition to the stimula-tory and inhibitory impact of IAA on the transcription of GA-biosynthetic and catabolic mRNAs, respectively, GA can mediate increased rates of polar IAA transport (Bjorklund et al. 2007). Thus, contrary to GA/auxin interac-tion pattern in flowers, apparently GA and auxin during immature stages (S1 and S3) along with ethylene during rip-ening (S4) act synergistically to coordinate the progression of fruit development, as proposed in other species (Serrani et al. 2007; Csukasi et al. 2011). This hypothesis is supported by our recent studies in Prunus, which showed the stimula-tory effect of exogenous GA and auxin on enhancing fruit development and ripening (El-Sharkawy et al. 2008, 2012a, b, this paper). Earlier study reported that the growth of tomato fruit is coordinated by a delicate balance between auxin and GA (de Jong et al. 2009), in which auxin is needed to mediate cell division and GA is required to organize cell expansion. The mutual effect of the two hormones was fur-ther validated in the development of parthenocarpic tomato fruit. Either auxin or GA application can promote partheno-carpic fruit growth, whereas neither of them alone was able to maintain the growth rate to the end of ripening. Only the joint application of both hormones resulted in parthenocarpic fruits similar to those obtained by pollination (Serrani et al. 2007). This is accurate for the different developmental stages in plum, excluding S2-phase. As mentioned previously, dur-ing S2 there is hardly any increase in fruit size, which coin-cides with a scarcity of auxin and ethylene (Miller et al. 1987; El-Sharkawy et al. 2007, 2008). However, the accu-mulation of only GA4 along with the expression profile of the different GA-signalling mRNAs during S2 seem to be associated with the lignification of the endocarp, which is the only developmental process occurring during this stage (El-Sharkawy et al. 2012b; this paper). Biemelt et al. (2004) demonstrated that GA mediates lignin formation and deposi-tion by polymerization of pre-formed monomers. Appar-ently, a mutual effect between GA and auxin occurs during maturation phase leading to accelerated cell division and expansion (El-Sharkawy et al. 2010, 2012a; this paper),

resulting in increased fruit size. This suggests that both GA1 and GA4 are directly involved in S3 events through promot-ing cell expansion. Thus, PslGID1 should be a component ofthe GA signal network that regulates S3-fruit growth. Dur-ing fruit ripening (S4), the accumulation of different GA-sig-nalling transcripts indicated continuous contribution of GA in fruit development, a role that is sustained mainly by GA1. During ripening, PslGID1 s showed a dramatic divergence in their expression profile. PslGID1c accumulated in a typical GA-dependent manner; however, that of PslGID1b followed closely the ethylene evolution profile of the fruit from this stage (El-Sharkawy et al. 2007). Accordingly, it is feasible to speculate that PslGID1b transcripts may be regulated during plum fruit ripening in an ethylene-dependent manner. This argument is supported by the presence of a remarkable com-bination of ethylene-responsive cis-acting elements found exclusively in PslGID1b promoter sequence along with the selective trigger of PslGID1b transcription in response to ethylene. Although GA content increased during ripening, the involvement of GA in the activation of ethylene-signal-ling, leading to climacteric ripening is not evident. Active ethylene-signalling usually results in increase of DELLA stability via declining GA content, supposedly through post-transcriptional control of GA20ox/GA3ox/GA2ox genes (Achard et al. 2003). Apparently, GA is needed at this stage mainly to promote cell expansion, leading to full fruit size.

GID1 and DELLA proteins are bona fide components of GA-signalling given that they display GA-dependent interaction both in vitro and in vivo (Ueguchi-Tanaka et al. 2005, 2007; Griffiths et al. 2006; Nakajima et al. 2006). Several lines of experiments were applied to investigate the function of the PslGID1 proteins in this study. These were divided into two main groups. The first one verified the potential of GA-enhanced interaction between PslGID1 and AtDELLA. Predicted 2-D protein structures illus-trated the ability of PslGID1 to bind active GA with high efficiency. Also, the 3-D protein modelling showed the structural capacity of plum GA-receptors to assemble the GA-PslGID1s–AtGAI complex. The previous theoretical results were further confirmed by the Y2H and BiFC assays that illustrated the ability of PslGID1 to function like the Arabidopsis and rice GID1 (Ueguchi-Tanaka et al. 2005; Griffiths et al. 2006) via binding GA and interacting with AtDELLA, where GA is a prerequisite to start the interac-tion. Without GA-binding, the N-terminal extension (N-Ex) of GID1 has a flexible structure that is highly sensi-tive to protease treatment. Binding of GA to the GID1 C-terminal domain induces a conformational switch of its N-Ex to cover the GA-binding pocket, as well as creates hydrophobic surfaces for DELLA-binding (Sun 2010). The second line of functional study was targeted via demon-strating the ability of PslGID1 to activate the downstream signalling machinery occurred due to assembling the stable

412 Plant Mol Biol (2014) 84:399–413

1 3

GA–PslGID1–DELLA complex in planta. Therefore, the function of the PslGID1 was further established by evaluat-ing the phenotypes of WT and g1-ac Arabidopsis mutants ectopically expressing PslGID1 and the consequent altera-tion on the active GA content as well as expression of genes involved in GA-signalling intermediates. Overexpression of PslGID1 in WT resulted in phenotypical changes consistent with altered GA-responsiveness; however, the phenotypic deviations from WT were more distinguishable under SD conditions. Furthermore, the presence of PslGID1 in g1-ac double mutant was able to suppress the various growth defects in the mutant, indicating that PslGID1 analogously function as AtGID1. At endogenous level, PslGID1-overex-pression in g1-ac was able to re-establish the regular GA accumulation profile and consequently recover the stand-ard expression of the various GA-metabolism genes. It is reasonable to speculate that the reduced levels of GA in GID1/g compared with the control g1-ac is due to activa-tion of GA-signalling, resulting from PslGID1-overexpres-sion, which induces feedback mechanisms that repress the GA-biosynthesis pathway and stimulate GA-inactivation. Compared to the PslGID1 transgene accumulation with the phenotypical characteristics of the resultant plants indicated that the severity of growth behavior correlate positively with the transgene level. However, the relative acceleration of growth is dependent on the ortholog of PslGID1-introduced. The PslGID1b and 1c pathways seem to fulfill different regulatory roles during fruit develop-ment and to control their downstream targets distinctly. In all Arabidopsis backgrounds used in this study, PslGID1b-overexpression had the strongest effect in germination, flowering and reproductive growth characteristics; how-ever, PslGID1c displayed more impact in vegetative growth events. Apparently, there are other factors that can selec-tively control the physiological roles of the various GA-receptors within the plants. Two factors that may influence GID1 roles have been suggested by Griffiths et al. (2006) to be either due to the different affinities of GID1 for binding various GA forms or the GID1 preference for interaction via a specific DELLA protein. In Arabidopsis, although AtGID1b has the highest GA-binding affinity, the lack of obvious phenotypic defects in the single gid1b-1 mutant, makes the significance of this increased affinity not evident (Griffiths et al. 2006). On the other hand, it is known that Arabidopsis has five DELLAs, which exhibit both overlap-ping and distinct roles in regulating GA-responsive growth (Dill and Sun 2001; Cheng et al. 2004; Tyler et al. 2004). It is conceivable that individual GID1 display specificity for the DELLA that target, although neither our binding energy results nor previous studies using Y2H approach (Griffiths et al. 2006; Nakajima et al. 2006) were able to support this model, since there are no clear differences in binding speci-ficity of particular GID1 to the different DELLAs. Further

studies are required to determine the affinity of PslGID1 to bind the various GA forms along with investigating the interaction specificity of individual PslGID1 to different PsDELLA. Nevertheless, our findings establish the critical role of GA in fruit growth and strongly support PslGID1 contribution as active GA-receptors for mediating GA-responsive actions during fruit development.

Acknowledgments We thank Dr. Stephen G. Thomas for provid-ing gid1a-1/gid1c-1 double mutant, Dr. Stanton Gelvin for providing the EYFP vectors. We also thank the Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA) and Early Researchers Award of the Ontario Ministry of Innovation for financial assistance. We thank Dr. Bouzayen for help with subcellular localization experiment.

References

Achard P, Vriezen WH, Van Der Straeten D, Harberd NP (2003) Ethylene regulates Arabidopsis development via the modula-tion of DELLA protein growth repressor function. Plant Cell 15:2816–2825

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W et al (1997) Gapped BLAST and PSI-BLAST: a new gen-eration of protein database search programs. Nucleic Acids Res 25:3389–3402

Ariizumi T, Murase K, Sun TP, Steber CM (2008) Proteolysis-inde-pendent downregulation of DELLA repression in Arabidop‑sis by the gibberellin receptor GIBBERELLIN INSENSITIVE DWARF1. Plant Cell 20:2447–2459

Augustí M, Almela V, Andreu I, Juan M, Zacarias L (1999) Syn-thetic auxin 3,5,6-TPA promotes fruit development and cli-macteric in Prunus persica L. Batsch. J Hortic Sci Biotech 74: 556–560

Baker NA, Sept D, Joseph S, Holst MJ, McCammon JA (2001) Elec-trostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci USA 98:10037–10041

Biemelt S, Tschiersch H, Sonnewald U (2004) Impact of altered gib-berellin metabolism on biomass accumulation, lignin biosynthe-sis, and photosynthesis in transgenic tobacco plants. Plant Phys-iol 135:254–265

Bjorklund S, Antti H, Uddestrand I, Moritz T, Sundberg B (2007) Cross-talk between gibberellin and auxin in development of Pop-ulus wood: gibberellin stimulates polar auxin transport and has a common transcriptome with auxin. Plant J 52:499–511

Bukovac MJ, Nakagawa S (1967) Comparative potency of gibberel-lins in inducing parthenocarpic fruit growth in Malus sylvestris Mill. Experientia 23:865

Bukvoac MJ (1963) Induction of parthenocarpic growth of apple fruits with gibberellins A3 and A4. Bot Gaz 124:191–195

Bukvoac MJ, Yuda E (1979) Endogenous plant growth substances in developing fruit of Prunus cerasus L. Plant Physiol 63:129–132

Cheng H, Qin L, Lee S, Fu X, Richards DE, Cao D et al (2004) Gib-berellin regulates Arabidopsis floral development via suppression of DELLA protein function. Development 131:1055–1064

Crane J (1964) Growth substances in fruit setting and development. Annu Rev Plant Physiol 15:303–326

Csukasi F, Osorio S, Gutierrez JR, Kitamura J, Giavalisco P, Naka-jima M et al (2011) Gibberellin biosynthesis and signalling during development of the strawberry receptacle. New Phytol 191:376–390

de Jong M, Mariani C, Vriezen WH (2009) The role of auxin and gib-berellin in tomato fruit set. J Exp Bot 60:1523–1532

413Plant Mol Biol (2014) 84:399–413

1 3

Dill A, Sun TP (2001) Synergistic de-repression of gibberellin signal-ling by removing RGA and GAI function in Arabidopsis thali‑ana. Genetics 159:777–785

Dunberg A, Odén PC (1983) Gibberellins and conifers. In: Crozier A (ed) The biochemistry and physiology of gibberellins. Praeger, New York, pp 221–295

El-Sharkawy I, Kim WS, El-Kereamy A, Jayasankar S, Svircev AM, Brown DCW (2007) Isolation and characterization of four ethyl-ene signal transduction elements in plums (Prunus salicina L.). J Exp Bot 58:3631–3643

El-Sharkawy I, Kim WS, Jayasankar S, Svircev AM, Brown DCW (2008) Differential regulation of four members of ACC synthase gene family in plum. J Exp Bot 59:2009–2027

El-Sharkawy I, Sherif S, Mila I, Bouzayen M, Jayasankar S (2009) Molecular characterization of seven genes encoding ethylene-responsive transcriptional factors during plum fruit development and ripening. J Exp Bot 60:907–922

El-Sharkawy I, Mila I, Bouzayen M, Jayasankar S (2010) Regulation of two germin-like protein genes during plum fruit development. J Exp Bot 61:1761–1770

El-Sharkawy I, Sherif S, Mahboob A, Abubaker K, Bouzayen M, Jayasankar S (2012a) Expression of auxin-binding protein1 dur-ing plum fruit ontogeny supports the potential role of auxin in initiating and enhancing climacteric ripening. Plant Cell Rep 31:1911–1921

El-Sharkawy I, El Kayal W, Prasath D, Fernández H, Bouzayen M, Svircev AM et al (2012b) Identification and genetic characteriza-tion of a gibberellin 2-oxidase gene that controls tree stature and reproductive growth in plum. J Exp Bot 63:1225–1239

Fleet CM, Sun TP (2005) A DELLAcate balance: the role of gibberel-lin in plant morphogenesis. Curr Opin Plant Biol 8:77–85

Gillaspy G, Ben-David H, Gruissem W (1993) Fruits: a developmen-tal perspective. Plant Cell 5:1439–1451

Griffiths J, Murase K, Rieu I, Zentella R, Zhang ZL, Powers SJ et al (2006) Genetic characterization and functional analysis of the GID1 gibberellin receptors in Arabidopsis. Plant Cell 18:3399–3414

Harberd NP, Belfield E, Yasumura Y (2009) The angiosperm gibberel-lin-GID1-DELLA growth regulatory mechanism: how an “Inhibi-tor of an inhibitor” enables flexible response to fluctuating envi-ronments. Plant Cell 21:1328–1339

Hedden P (2008) Plant biology: gibberellins close the lid. Nature 456:455–456

Higo K, Ugawa Y, Iwamoto M, Korenaga T (1999) Plant cis-acting regulatory DNA elements (PLACE) database. Nucleic Acids Res 27:297–300

Jackson DI (1968) Gibberellin in the growth of peach and apricot fruits. Aust J Biol Sci 21:209–215

Lee LY, Fang MJ, Kuang LY, Gelvin SB (2008) Vectors for multi-color bimolecular fluorescence complementation to investigate protein–protein interactions in living plant cells. Plant Meth 4:24

Lodhi F, Bradley MV, Crane JC (1969) Auxins and gibberellin-like substances in parthenocarpic and non-parthenocarpic syconia of Ficus carica L., cv. King. Plant Physiol 44:555–561

Mesejo C, Yuste R, Martínez-Fuentes A, Reig C, Iglesias DJ, Primo-Millo E, Agustí M (2013) Self-pollination and parthenocarpic ability in developing ovaries of self-incompatible Clementine mandarins (Citrus clementina). Physiol Plant 148:87–96

Miller AN, Walsh CS, Cohen JD (1987) Measurement of indole-3-acetic acid in peach fruits (Prunus persica L. Batsch cv. Red-haven) during development. Plant Physiol 84:491–494

Nakajima M, Shimada A, Takashi Y, Kim YC, Park SH, Ueguchi-Tan-aka M et al (2006) Identification and characterization of Arabi‑dopsis gibberellin receptors. Plant J 46:880–889

Ngo P, Ozga JA, Reinecke DM (2002) Specificity of auxin regulation of gibberellin 20-oxidase gene expression in pea pericarp. Plant Mol Biol 49:439–448

Nitsch JP (1970) Hormonal factors in growth and development. In: Hulme AC (ed) The biochemistry of fruits and their products. Academic Press, London, New York, pp 427–472

Ogawa M, Hanada A, Yamauchi Y, Kuwahara A, Kamiya Y, Yamagu-chi S (2003) Gibberellin biosynthesis and response during Arabi‑dopsis seed germination. Plant Cell 15:1591–1604

Ross JJ, O’Neill DP, Wolbang CM, Symons GM, Reid JB (2002) Auxin-gibberellin interactions and their role in plant growth. J Plant Growth Regul 20:346–353

Seo M, Jikumaru Y, Kamiya Y (2011) Profiling of hormones and related metabolites in seed dormancy and germination studies. Methods Mol Biol 773:99–111

Serrani JC, Sanjuán R, Ruiz-Rivero O, Fos M, García Martínez JL (2007) Gibberellin regulation of fruit set and growth in tomato. Plant Physiol 145:246–257

Serrano M, Martínez-Romero D, Zuzunaga M, Riquelme F, Valero D (2007) Calcium, polyamine and gibberellin treatments to improve postharvest fruit quality. In: Dris R, Jain SM (eds) Production practices and quality assessment of food crops. Vol. 4. Posthar-vest treatment and technology. Kluwer Academic Publishers, Dordrecht, pp 55–68

Shimada A, Ueguchi-Tanaka M, Nakatsu T, Nakajima M, Naoe Y, Ohmiya H et al (2008) Structural basis for gibberellin recognition by its receptor GID1. Nature 456:520–523

Sponsel VM, Hedden P (2004) Gibberellin biosynthesis and inactiva-tion. In: Davies PJ (ed) Plant hormones: biosynthesis, signal trans-duction, action!. Kluwer Academic Publishers, Dordrecht, pp 63–94

Sun TP (2010) Gibberellin-GID1-DELLA: a pivotal regulatory mod-ule for plant growth and development. Plant Physiol 154:567–570

Sun TP, Gubler F (2004) Molecular mechanism of gibberellin signal-ling in plants. Ann Rev Plant Biol 55:197–223

Thomas SG, Philips AL, Hedden P (1999) Molecular cloning and functional expression of gibberellin 2-oxidases, multifunctional enzymes involved in gibberellin deactivation. Proc Natl Acad Sci USA 96:4698–4703

Tyler L, Thomas SG, Hu JH, Dill A, Alonso JM, Ecker JR et al (2004) DELLA proteins and gibberellin-regulated seed germination and floral development in Arabidopsis. Plant Physiol 135:1008–1019

Ueguchi-Tanaka M, Ashikari M, Nakajima M, Itoh H, Katoh E, Kob-ayashi M et al (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:693–698

Ueguchi-Tanaka M, Nakajim M, Katoh E, Ohmiya H, Asano K, Saji S et al (2007) Molecular interactions of a soluble gibberellin Receptor, GID1, with a rice DELLA protein, SLR1, and Gibber-ellin. Plant Cell 19:2140–2155

Vargas AM, Le Cunff L, This P, Ibanez J, de Andres MT (2013) VvGAI1 polymorphisms associate with variation for berry traits in grapevine. Euphytica 191:85–91

Wang T, Tomic S, Gabdoulline RR, Wade RC (2004) How optimal are the binding energetics of barnase and barstar? Biophys J 87:1618–1630

Wolbang CM, Chandler PM, Smith JJ, Ross JJ (2004) Auxin from the developing inflorescence is required for the biosynthesis of active gibberellins in barley stems. Plant Physiol 134:769–776

Xu YL, Li L, Gage DA, Zeevaart JAD (1999) Feedback regulation of GA5 expression and metabolic engineering of gibberellin levels in Arabidopsis. Plant Cell 11:927–936

Yamaguchi S (2008) Gibberellin metabolism and its regulation. Ann Rev Plant Biol 59:225–251

Yamaguchi I, Takahashi N (1976) Change of gibberellin and abscisic acid content during fruit development of Prunus persica. Plant Cell Physiol 17:611–614

Yasumura Y, Crumpton-Taylor M, Fuentes S, Harberd NP (2007) Step-by-step acquisition of the gibberellin-DELLA growth-regulatory mechanism during land-plant evolution. Curr Biol 17:1225–1230