Molecular Characterization of Three GIBBERELLIN-INSENSITIVE DWARF2 Homologous Genes in Common Wheat

F-box protein is a core component of the ubiquitin E3 ligase SCF complex and is involved in the gibberellin (GA) signaling pathway. To elucidate the molecular mechanism of GA signaling in wheat, three homologous GIBBERELLIN-INSENSITIVE DWARF2 genes, TaGID2s, were isolated from the Chinese Spring wheat variety. A subcellular localization assay in onion epidermal cells and Arabidopsis mesophyll protoplasts showed that TaGID2s are localized in the nuclei. The expression profiles using quantitative real-time polymerase chain reaction showed that TaGID2s were downregulated by GA3. The interaction between TaGID2s and TSK1 (homologous to ASK1) in yeast indicated that TaGID2s might function as a component of an E3 ubiquitin-ligase SCF complex. Yeast two-hybrid assays showed that a GA-independent interaction occurred between three TaGID2s and RHT-A1a, RHT-B1a, and RHT-D1a. Furthermore, TaGID2s interact with most RHT-1s, such as RHT-B1h, RHT-B1i, RHT-D1e, RHT-D1f, etc., but cannot interact with RHT-B1b or RHT-B1e, which have a stop codon in the DELLA motif, resulting in a lack of a GRAS domain. In addition, RHT-B1k has a frame-shift mutation in the VHIID motif leading to loss of the LHRII motif in the GRAS domain and RHT-D1h has a missense mutation in the LHRII motif. These results indicate that TaGID2s, novel positive regulators of the GA response, recognize RHT-1s in the LHRII motif resulting in poly-ubiquitination and degradation of the DELLA protein.


Introduction
Common wheat (Triticum aestivum L., AABBDD) is a primary food crop worldwide. One of the most valuable wheat breeding traits is dwarfism because semi-dwarf cultivars have greater resistance potential to lodging and have stable increased yields [1]. The extensive utilization of semi-dwarf cultivars resulted in unprecedented increases in world wheat yields, driving the "green revolution" in the 1960s and 1970s [2]. The two main "green revolution" genes are Rht-cetyltrimethylammonium ammonium bromide method [29][30]. Young leaves from Chinese Spring and the wild wheat diploid relatives Triticum urartu (AA), Aegilops speltoides (BB), and Aegilops tauschii (DD) were sampled for RNA extraction using the RNeasy Plant Mini Kit (QIAGEN, Hilden, Germany), and first-strand cDNA was synthesized using the Fas-tQuant RT Kit (with gDNase) (TIANGEN, Beijing, China) according to the manufacturer's instructions. Various wheat tissues, including young spikes, flag leaves, peduncles, the third and fourth internodes, and roots from Chinese Spring at the heading stage and seedlings of Chinese Spring treated with 30 μmolÁL -1 GA 3 or ddH 2 O (control) for 2 weeks were collected for RNA isolation.
Cloning the wheat GID2 and TSK1 genes To isolate the homologous GID2 genes in wheat, we selected a series of wheat expressed sequence tags (ESTs), which were highly similar to OsGID2 (AB100246), from the TIGR wheat EST database (http://blast.jcvi.org/euk-blast/index.cgi?-project=tae1) and designed a pair of Contig1.GID2.F/R primers (Table 1) by assembling the sequences. This primer pair was used to clone the GID2 genes in the genomic DNA and cDNA of Chinese Spring and the wild wheat diploid relatives. Amplifications were performed in a 20 μL reaction volume containing 100 ng DNA template, 0.4 μM forward and reverse primers, 1 U TaKaRa LA Taq (Takara, Biotechnology Co., Ltd, Dalian, China), 1× GC buffer I, and 0.4 mM of each dNTP under the following conditions: 94°C for 5 min, followed by 35 cycles of 94°C for 30 sec, 58°C for 30 sec, 72°C for 1 min, and a final extension at 72°C for 10 min. The polymerase china reaction (PCR) products were cloned into the pGEM-T easy vector (Promega, Madison, WI, USA), introduced into Escherichia coli, and 10 positive independent clones were sequenced commercially. DNAMAN and Clustal W were used for the sequence alignment and Clustal W and MEGA ver. 5.1 were used to draw the phylogenetic trees. In addition, to demonstrate whether TaGID2s play roles in the SCF ubiquitin-ligase complex, we designed the TSK1.F/R primers (Table 1) according to the complete coding sequence (CDS) of the T. aestivum SKP1/ASK1-like protein (AY316293), which was cloned by Li et al. (2006) [31]. PCR was carried out using Chinese Spring cDNA as the template, and we obtained a single clone called TSK1.

Physical localization assay
Three sets of gene-specific primers, TaGID2-1LF/1LR, TaGID2-2LF/2LR, and TaGID2-3LF/ 3LR (Table 1), were used to amplify the genomic DNA of Chinese Spring, 21 nulli-tetrasomic lines, and 11 deletion lines of Chinese Spring to confirm the location of TaGID2s on the wheat chromosomes. The PCR products were resolved by 1.5% agarose gel electrophoresis.

Real-time PCR analysis
Quantitative real-time PCR (qRT-PCR) was performed to measure transcript levels of TaGID2s using the LightCycler 480 system (Roche, Indianapolis, IN, USA). SYBR Green I Master (Roche) was used according to the manufacturer's protocol. Three pairs of gene-specific primers, TaGID2-A-EF/ER, TaGID2-B-EF/ER, and TaGID2-D-EF/ER (Table 1), were used for the gene expression analysis, and the Ta4045 gene was used as the reference [32]. The expression profiles of the three TaGID2s in various wheat organs at the heading stage and their response to exogenous GA were investigated according to the description of Li et al. (2013a) [33]. All qRT-PCR experiments were performed independently using three biological and three technical replicates.

Subcellular localization assay
The subcellular localization assay was performed by transiently expressing the TaGID2-green fluorescent protein (GFP)/yellow fluorescent protein (YFP) fusion protein in onion epidermal cells and Arabidopsis protoplasts, respectively. The GFP fusion proteins of TaGID2A-GFP, TaGID2B-GFP, and TaGID2D-GFP were constructed using PCR combined with the restriction enzyme digestion method and the TaGID2-A-HindIII.F/BamHI.R, TaGID2-B-HindIII.F/ BamHI.R, and TaGID2-D-HindIII.F/BamHI.R primers ( Table 1). The TaGID2-GFP fusion protein was transformed into onion epidermal cells by particle bombardment with a PDS-1000/He system (Bio-Rad, Hercules, CA, USA) according to the method of Oh et al. (2008) using the pJIT-163-hGFP plasmid as a positive control [34]. The complete CDS regions of the TaGID2s were amplified with the TaGID2-A-attB.F/R, TaGID2-B-attB.F/R, and TaGID2-D-attB.F/R primers (Table 1) and cloned into the pEarleyGate101 vector using the Gateway- Table 1. Polymerase chain reaction primers used in this study.

Primer
Forward primer sequence Reverse primer sequence TaGID2-D1-NdeI.F/ EcoRI.R compatible vector cloning system to generate the TaGID2A-YFP, TaGID2B-YFP, and TaGID2D-YFP constructs [35]. Arabidopsis mesophyll protoplast cells were isolated and transformed using the PEG transformation method as described by Yoo et al. (2007) [36]. The pA7-YFP plasmid was introduced into the protoplasts as a positive control. After 18-24 h incubation, the GFP/YFP fluorescence emissions from living cells were observed under a confocal microscope (Zeiss LSM 710 NLO; Carl Zeiss, Oberkochen, Germany) at excitation wavelengths of 488/514 nm and emission wavelengths of 506-538/527 nm.
Yeast two-hybrid assay The yeast two-hybrid assay was performed to detect the interactions between TaGID2 and TSK1/ RHT-1s, as well as the interactions between TaGID1s and RHT-1s in vitro. The CDSs of TaGID1s, TSK1, TaGID2s, SLN1, and Rht-1s were ligated into the pGADT7 and pGBKT7 vectors (Clontech, Mountain View, CA, USA) [37][38], using the NdeI and EcoRI sites to generate the pGADT7-TSK1, pGADT7-SLN1, and pGADT7-RHT1s prey plasmids and the pGBKT7-TaGID1s, pGBKT7-TaGID2s bait plasmids, respectively. The gene-specific primers are shown in Table 1. The AD and BK fusion constructs were co-transformed into Saccharomyces cerevisiae strain AH109, and the transformants were screened on SD/-Leu-Trp media. The positive transformants were further streaked on SD/-Leu-Trp-His-Ade media with or without 100 μmolÁL -1 GA 3 . The plate assay was carried out according to the Yeast Protocol Handbook (Clontech).

Bimolecular fluorescence complementation (BiFC) assay
The BiFC assay was performed to detect the interactions between TaGID1-A1 and RHT-1s in vivo. The complete CDS regions of TaGID1-A1 and Rht-1s were amplified with the TaGID1-A-attB.F/R, Rht-B1-attB.F/R, and Rht-D1-attB.F/R primers (Table 1) and cloned into the pX-nYFP and pX-cCFP vectors using the Gateway technology to generate the TaGID1A-nYFP, TaGID1A-cCFP, RHT1s-nYFP, and RHT1s-cCFP constructs. Arabidopsis mesophyll protoplasts were isolated and co-transformed using the PEG transformation method as described by Yoo et al. (2007) [36]. The pA7-YFP plasmid was introduced into the protoplasts as a positive control. After 18-24 h incubation, YFP fluorescence emission in living cells was observed under a confocal microscope (Zeiss LSM 710 NLO) at an excitation wavelength of 514 nm and emission wavelength of 527 nm. Chloroplast autofluorescence was observed at an excitation wavelength of 488 nm and emission wavelengths of 664-696 nm.

Western blot analysis and co-immunoprecipitation
The fourth internodes were sampled at the heading stage, batch-frozen in liquid N 2 , and ground into powder using a mortar and pestle to extract total soluble protein. The powder (1 g) was resuspended in 1 mL extraction buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 1 mM DTT, 1 mM EDTA pH 8.0, 10% glycerol, 1% Triton X-100, and protease inhibitor cocktail; Roche), the suspension was vortexed continuously at 100 rpm for 30 min, and then centrifuged at 16,000 g for 30 min. The supernatant was collected and resolved by 10% acrylamide gel electrophoresis, and electro-transferred to PVDF membranes. The membranes were blocked for 2-3 h at room temperature in TBST (20 mM Tris-HCl, pH 8.0, 137 mM NaCl, 0.05% [v/v] Tween 20) containing 5% (w/v) nonfat dry milk. The membranes were incubated with TaGID2 polyclonal antibody (1:1,000; CWBIO Co. Ltd, Beijing, China) for 12 h at 4°C. All dilutions were done in TBST containing 5% (w/v) nonfat dry milk. The blots were washed five times in TBST for 6 min each before incubation with the secondary horseradish peroxidase-conjugated goat α-rabbit antibody (1:10,000; cat. no. CW0103 CWBIO). The blots were washed five times with TBST for 6 min each, the bound antibody was detected with the eECL Western Blot Kit (cat. no. CW0049A; CWBIO), and exposed to X-ray film (Kodak-Biomax Light; Kodak Inc., Rochester, NY, USA). Protein extracts were prepared as described above for the Western blot analysis. The resulting supernatants were incubated with DELLA monoclonal antibodies (1:100; Abmart Co. Ltd., Shanghai, China) for 12 h at 4°C. Then, 60 μl Protein G Agarose beads (Millipore, Billerica, MA, USA) were added to each tube and vortexed continuously at 70 rpm for 2 h. The antibody-antigen complexes were precipitated by centrifugation for 5 min at 2,000 g, washed four times with 1 mL PBS, and eluted from the beads by boiling in sodium dodecyl sulfate-polyacrylamide gel electrophoresis sample buffer for 5 min. Equal quantities of immunoprecipitated and unbound proteins were resolved on 10% acrylamide gels. Protein immunoblots were carried out as described for the Western blot analysis.

Yeast three-hybrid assay
The yeast three-hybrid assay was performed using pGADT7 and pBridge (Clontech) as expression vectors to detect the interaction between TaGID1-DELLA-TaGID2 in vitro. The pGADT7-SLN1 and pGADT7-RHT1s plasmids were used as prey. The pBr-TaGID1A-TaGID2A bait plasmids were generated by using the BamHI and EcoRI sites in MCSI for TaGID1s and the NotI and BglII site in MCSII for TaGID2 using the specific primers shown in Table 1. The AD and pBr fusion constructs were co-transformed into yeast strain HF7c, the transformants were screened on SD/-Met-Leu-Trp media, and the positive transformants were further streaked on SD/-Met-Leu-Trp-His media containing 20 mM 3-amino-1, 2, 4-triazole with or without 100 μmolÁL -1 GA 3 .

TaGID2s are orthologous to OsGID2
To elucidate the function of F-box proteins in GA signal transduction, three homologous genes were cloned from the cDNA of the Chinese Spring wheat variety using PCR and the Contig1. GID2.F/R primers. The cDNA sequences of the three genes were 812, 815, and 815 bp long, respectively, and contained 723, 726, and 726 bp CDSs, respectively, and partial 5 0 -and 3 0untranslated regions ( Table 2). They encoded 240, 242, and 241 amino acid residues with molecular weights of 26.24, 26.26, and 26.69 KDa, with pIs of 10.36, 11.12, and 10.53, respectively. A BLAST search revealed that the predicted protein sequences were homologous with the F-box protein family [15]. Comparison of the primary structures of the three predicted proteins with OsGID2 and AtSLY1 revealed that three putative amino acid sequences shared 45.70%, 47.27%, and 45.53% sequence identity with OsGID2 and only 19.58%, 21.16%, and 21.58% sequence identity with AtSLY1, respectively, whereas their F-box domains had similarities of 80.85%, 89.36%, and 82.98% with OsGID2 and 48.94%, 48.94%, and 48.94% with AtSLY1, respectively, indicating that the three genes in wheat are orthologous to OsGID2 and AtSLY1. Thus, the three genes were named TaGID2-1, TaGID2-2, and TaGID2-3 (Genbank accession numbers KU857036-KU857038; Fig 1).
Analysis of the conserved domain (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) showed that all three TaGID2s have F-box and SANT conserved domains. The sequence alignment of OsGID2 and TaGID2s (Fig 1) indicated that TaGID2-1, TaGID2-2, and TaGID2-3 share five motifs, including the VR1, F-box, GGF, VR2, and LSL motifs as in OsGID2. The physicochemical property analysis using the ProtParam tool (http://web.expasy.org/ protparam/) showed that the three TaGID2s were unstable hydrophilic and alkaline lipid soluble proteins. The SignalP 4.1 Server (http://www.cbs.dtu.dk/services/SignalP/) and TMHMM Server v. 2.0 (http://www.cbs.dtu.dk/services/TMHMM/) results predicted that the TaGID2s were non-secretory, non-transmembrane cytoplasmic proteins. The phylogenetic tree of the GID2 proteins (S1 Fig) was divided into two major groups of dicotyledons and monocotyledons. The GID2 proteins from dicotyledons, such as Arabidopsis, soybean, grape, and sunflower, formed one cluster, and the GID2 proteins of monocotyledons, such as, rice, wheat, and maize, formed another cluster. The SmGID2a and SmGID2c were clustered separately. TaGID2 had a close evolutionary relationship with AetGID2 and OsGID2, whereas the relationship was more distant with AtSLY1 and BnSLY1, and even more distant with SmGID2, indicating that the evolutionary relationship among the GID2 proteins is consistent with their taxonomic distance. The corresponding genomic DNA sequences of the three TaGID2s were isolated from the Chinese Spring wheat variety (Genbank accession numbers KU857045-KU857047; S2 Fig) Their lengths were 1,085, 1,160, and 1,160 bp, and they shared high similarity with each other. All of the DNA sequences had two exons and one intron. The intron in TaGID2-1 occurred between nucleotides 509 and 510 in the CDS sequence, whereas both introns of TaGID2-2 and TaGID2-3 were situated at nucleotides 512 and 513 of the CDS sequence, and their lengths were 273, 345, and 345 bp, respectively.
Three TaGID2s were located on the short arms of group 3 chromosomes We amplified 21 nulli-tetrasomic lines of Chinese Spring (Fig 2A) and found that the TaGID2-1 gene specific primers amplified nothing in N3AT3B, the TaGID2-2 gene specific primers did not produce a product in N3BT3D, and the TaGID2-3 gene specific primers failed to produce a product in N3DT3A, indicating that the three TaGID2s were individually assigned to the 3A, 3B, and 3D wheat chromosomes. Thus, they were renamed TaGID2-A1, TaGID2-B1, and TaGID2-D1, respectively. In addition, we determined the location of the TaGID2 chromosome bin (Fig 2B) using 11 deletion lines from group 3 chromosomes [39]. Among the four 3A deletion lines, the TaGID2-A1 gene specific primers did not yield a product in 3AS-2 and 3AS-4, the TaGID2-B1 gene specific primers did not yield a product in 3BS-1 and 3BS-9 in four 3B deletion lines, and the TaGID2-D1 gene specific primers failed to amplify 3DS-4 and 3DS-5 in the three 3D deletion lines, demonstrating that the three TaGID2s are located in chromosome bins 3AS4-0.45-1.00, 3BS9-0.57-0.78, and 3DS4-0.59-1.00, respectively (Fig 2C).

GA downregulates TaGID2s expression levels in wheat
Similar expression patterns were observed for the three TaGID2s in various wheat organs ( Fig  3A) at the heading stage. TaGID2s were constitutively transcribed in young spikes, flag leaves, peduncles, the third and fourth internodes, and roots from Chinese Spring. However, their expression levels in six organs were different, and they were preferentially expressed in young spikes and elongating stems, such as the third and fourth internodes, but transcribed less in flag leaves and roots. Among the three TaGID2s, TaGID2-B1 was expressed at the lowest level, and TaGID2-D1 was transcribed the most in the six organs, except peduncles and flag leaves. Furthermore, TaGID2s were downregulated in seedlings treated with 30 μmolÁL -1 GA 3 ( Fig  3B), suggesting feedback regulation of TaGID2s by GA in wheat.

TaGID2s form an SCF complex by interacting with TSK1
The Arabidopsis ASK1 homologous gene TSK1 (Genbank accession number KU857044) was cloned from the Chinese Spring wheat variety to clarify the role of TaGID2 in formation of the SCF complex, which has only two single nucleotide polymorphisms (SNPs) with AY316293 in the CDS (Fig 4A) [31]. TSK1 and AY316293 could be the same gene or a different copy of the same gene. The yeast two-hybrid assay showed that TSK1 interacted with TaGID2s directly in a GA-independent way (Fig 4B), indicating that TaGID2s might function as a component of an SCF E3 ubiquitin-ligase complex, triggering poly-ubiquitination and degradation of DELLA proteins.

TaGID2s interact with RHT-1 in the nuclei
The yeast two-hybrid assay showed that GA-independent interactions occurred among the three TaGID2s and RHT-A1a, RHT-B1a, and RHT-D1a, demonstrating that RHT-1s are targets of TaGID2s. Moreover, among the nine TaGID2/RHT-1 combinations, TaGID2-A1 exhibited relatively stronger activities than those of TaGID2-B1 and TaGID2-D1 when interacting with the same RHT-1, suggesting it has stronger affinity to DELLA proteins (Fig 5A). To further verify the TaGID2 and RHT-1 interacting region, the CDS of SLN1 and 14 allelic variations of Rht-1 were constructed into pGADT7 and the three TaGID2s into pGBKT7 for yeast two-hybrid assays. The results showed that TaGID2s interacted with RHT-B1a, RHT-B1h, Characterization of TaGID2 Genes in Wheat RHT-B1i, RHT-B1j, RHT-B1o, RHT-D1a, RHT-D1e, RHT-D1f, RHT-D1g, and RHT-D1i in a GA-independent manner, but did not interact with RHT-B1b, RHT-B1e, RHT-B1k, or RHT-D1h (Fig 5B). The reason may be a stop codon in the RHT-B1b and RHT-B1e DELLA motif resulting in a lack of the GRAS domain, which would prevent interaction with TaGID2s. Although RHT-B1k and RHT-D1g frameshifts occurred in the GRAS domain, the RHT-B1k Yeast two hybrid assay between TaGID2 and TSK1, TSK1 was used as prey, and TaGID2-A1, TaGID2-B1, and TaGID2-D1 were used as baits, the AD and BK constructs were co-transformed into the yeast strain AH109, the transformants were grown on the SD/-Leu-Trp media (control) and the SD/-Leu-Trp-His-Ade media with or without 100 μM GA 3 , all experiments were conducted three times, and five clones were used each time.   TaGID2-D1 were used as baits, and RHT-A1a, RHT-B1a, and RHT-D1a were used as preys, the AD and BK constructs were co-transformed into the yeast strain AH109, the transformants were grown on the SD/-Leu-Trp media (control) and the SD/-Leu-Trp-His-Ade media containing 20 μg/mL X-α-Gal with or without 100 μM GA 3 , all experiments were conducted three times, and five clones were used each time. (B) TaGID2s interact with allelic variations of RHT-1 in the LHRII motif, Characterization of TaGID2 Genes in Wheat frameshift occurred in the VHIID motif, causing loss of the LHRII motif, which could prevent RHT-B1k from interacting with TaGID2s, whereas the RHT-D1g frameshift occurred in the PFYRE motif, giving rise to the interaction between RHT-D1g and TaGID2s. The RHT-D1e and RHT-D1h missense mutations both occurred in the LHRII motif, but only the RHT-D1h mutation resulted in loss of the interaction with DELLA-GID2, which was consistent with the evaluation by the SIFT program that RHT-D1h was predicted to have a damaging effect on the protein. As a consequence, the interaction between TaGID2 and DELLA occurred in the LHRII motif.
Western blot and co-immunoprecipitation analyses demonstrated that the TaGID2 protein was approximately 26 kDa (Fig 6). The Western blot analysis of total proteins with the TaGID2 antibody showed that RHT-1s existed in all samples, suggesting that TaGID2s interact with RHT-1s in wheat. In addition, a Western blot analysis using the TaGID2 antibody for proteins captured by co-immunoprecipitating total proteins with the DELLA antibody also confirmed interaction between TaGID2s and RHT-1s in wheat. Moreover, common wheat is hexaploid with A, B, and D genomes in which most genes have three copies, and blots also existed in the mutant samples, such as RHT-B1b, RHT-B1e, RHT-B1k, and RHT-D1h, illustrating that the Rht-1s-encoded DELLA proteins were functionally redundant.
RHT-1s were localized in nuclei. Therefore, TaGID2s-GFP/YFP fusion genes were constructed to identify whether the interaction between TaGID2 and RHT-1 occurred in nuclei. Transient expression of TaGID2s

Interaction between TaGID1-RHT1-TaGID2
Our previous study isolated three GA receptor genes called TaGID1-A1, TaGID1-B1, and TaGID1-D1 from hexaploid wheat, and a yeast two-hybrid assay demonstrated that TaGID1s interact with the DELLA proteins RHT-A1a, RHT-B1a, and RHT-D1a in the presence of GA 3 [33]. Yeast two-hybrid and BiFC assays were carried out to detect the interactions between TaGID1 and 14 RHT-1 mutants in vitro and in vivo and identify the TaGID1 and RHT-1 SLN1 and RHT-1s were used as preys, TaGID2-A1, TaGID2-B1, and TaGID2-D1 were used as baits from left to right in each block, and the AD and BK constructs were co-transformed into the yeast strain AH109, the transformants were grown on the SD/-Leu-Trp media (control) and the SD/-Leu-Trp-His-Ade media with or without 100 μM GA 3 , all experiments were conducted three times, and five clones were used each time. interacting region. The yeast two-hybrid (S5 Fig) and BiFC (S6 Fig) assays showed that TaGID1 interacted with most RHT-1s in a GA-dependent manner, except RHT-B1b, RHT-B1e, RHT-B1k, and RHT-D1g. The interaction between TaGID1 and RHT-B1b/ RHT-B1e failed because of a premature termination mutation in the DELLA domain, indicating that the interaction between TaGID1 and RHT-1 occurs in the DELLA domain. TaGID1 could not interact with RHT-B1k or RHT-D1g due to a frameshift mutation in the VHIID and PFYRE motifs leading to deficiency of the GRAS domain, demonstrating that the GRAS domain is required for DELLA protein function.
Because the yeast two-hybrid assay showed that TaGID1-A1 and TaGID2-A1 exhibited stronger affinity for DELLA proteins, the pBr-TaGID1A-TaGID2A expression vector was generated as the bait plasmid for the yeast three-hybrid assay with pGADT7-RHT1s as the prey plasmids. The pBr-TaGID1A-TaGID2A bait plasmid interacted with most RHT-1s, except RHT-B1b, RHT-B1e, RHT-B1k, RHT-D1g, and RHT-D1h in the presence of GA 3 (Fig 7), Interaction among TaGID1, TaGID2, and RHT-1 in yeast cells. SLN1 and RHT-1s were used as preys, pBr-TaGID1A-TaGID2A was used as baits, and the AD and pBr constructs were co-transformed into the yeast strain HF7c, the transformants were grown on the SD/-Met-Leu-Trp media (control) and the SD/-Met-Leu-Trp-His media containing 20 mM 3-AT with or without 100 μM GA 3 , all experiments were conducted three times, and five clones were used each time.
doi:10.1371/journal.pone.0157642.g007 demonstrating that TaGID1 was the GA receptor. TaGID1 and RHT-1 interacted with the RHT-1 protein DELLA domain after TaGID1 received the GA signal. The interaction induced TaGID2 to recognize the RHT-1 protein through the LHRII motif of the RHT-1 protein, and the GA-TaGID1-DELLA-TaGID2 complex was subsequently degraded via the ubiquitin-proteasome pathway, which was triggered from the interaction between TaGID2 and TSK1.

Discussion
The specificity of ubiquitination, which is an essential post-translational modification that regulates signaling and protein turnover in eukaryotic cells, is driven by SCF ubiquitin E3 ligases. As a core component of the SCF complex, F-box proteins, named for their highly conserved Fbox motif, recognize and bind to the substrate protein [13]. The F-box proteins SLY1 and SNE in Arabidopsis and GID2 in rice are involved in the GA signaling pathway [22, 24-25, 27-28, 40]. However, little information is available about homologous F-box proteins in wheat. Consequently, this study was carried out to genetically characterize and describe the functions of Fbox proteins in wheat.
The F-box gene is conserved in many plant species, so identifying the AtSLY1 and OsGID2 genes greatly facilitated homologous cloning of TaGID2 [22,24,[27][28]. As common wheat is a hexaploid plant containing three sets of genomes, three genes, called TaGID2-A1, TaGID2-B1, and TaGID2-D1, were isolated from the Chinese Spring wheat variety by PCRbased cloning. A phylogenetic analysis of the GID2 proteins showed that TaGID2s had a close evolutionary relationship with OsGID2, suggesting that TaGID2s may play similar roles to those of OsGID2 in wheat. The deduced amino acid sequences of the three TaGID2s were highly consistent with OsGID2, particularly the conserved F-box domain, which functions as a protein-protein interaction site during formation of the E3 ubiquitin-ligase SCF complex [12][13]. Besides the N-terminal F-box motif, many F-box proteins contain another proteinprotein interaction domain in their C-terminal, such as leucine-rich repeats (LRRs), WD-40 repeats, or Kelch. Hence, F-box proteins are grouped into three categories in humans and mice: FBXL denotes a protein containing an F-box and LRRs, FBXW denotes a protein with an F-box and WD repeats, and FBXO denotes a protein with an F-box and either another or no other motif [41][42]. Furthermore, Gagne et al. (2002) divided 694 potential Arabidopsis Fbox proteins into five distinct families and 20 subfamilies by phylogenetic analysis [15]. However, we did not find any conserved motifs, such as FBXL, FBXW, or FBXO, outside the Fbox domain in the structures of TaGID2s, but the structures of TaGID2s share the three conserved domains of the F-box, GGF, and LSL domains, with AtSLY1 and OsGID2. In addition, TaGID2 and OsGID2 both have a VR1 domain, which is absent in AtSLY1 [22,28]. Gomi et al. (2004) demonstrated that the F-box, GGF, and LSL domains are essential for OsGID2 function, whereas the VR1 domain is not necessary, and the N-terminus region may not be important for OsGID2 function; thus, the domain differences between GID2 and SLY1 are negligible [28]. TaGID2s shared all domains with OsGID2, further supporting that they may play similar roles.
The three TaGID2s shared a similar expression pattern during the wheat heading stage, were expressed in all tested organs, and exhibited higher expression levels in young spikes and elongating stems, but lower expression levels in flag leaves and roots in accordance with OsGID2, which is highly expressed in stems, the shoot apex, and unopened flowers [28]. They were similar to the expression pattern of the SD1 gene, which encodes a key enzyme in GA biosynthesis [43]. The expression analysis indicated that TaGID2s are involved in the GA response pathway. Moreover, GA downregulated the expression of TaGID2s, suggesting that TaGID2s are positive regulators of GA signaling, as described for OsGID2 and AtSLY1 [20][21][22][24][25][26][27][28].
The F-box domain is essential for the interaction between OsGID2 and OsSkp15, resulting in formation of the SCF GID2 complex [28]. The Arabidopsis ASK1 homologous gene TSK1, as described by Li et al. (2006), was cloned from the Chinese Spring wheat variety to confirm that TaGID2s actually encode the F-box protein and to clarify the role of TaGID2 in formation of the SCF complex. The TSK1 gene we obtained had two SNPs compared with AY316293, which was cloned from T. aestivum cv. Jingdong 1. They may be the same gene in different cultivars or they could be different copies of the same gene because of the hexaploidy of wheat [31]. The yeast two-hybrid assay demonstrated that TaGID2s interacted directly with TSK1 in a GAindependent manner, as other F-box proteins; thus, TaGID2s formed an E3 ubiquitin-ligase SCF complex by interacting with TSK1 [15][16][17].
Transient expression of GFP and YFP confirmed that TaGID2s are localized in the nuclei, as described for AtSLY1 [24]. GA-independent interactions occurred between the three TaGID2s and RHT-A1a, RHT-B1a, and RHT-D1a, demonstrating that RHT-1s are TaGID2 targets. However, Gomi et al. (2004) revealed that SCF GID2 only targets the SLR1 protein under a phosphorylated state. This distinction needs further investigation of the GA signaling mechanism in wheat. The yeast two-hybrid assay, Western blot analysis, and co-immunoprecipitation results showed that TaGID2s interacted with RHT-1s. The yeast two-hybrid assay for TaGID2s and the 14 allelic variations of Rht-1 showed that RHT-B1b and RHT-B1e cannot interact with TaGID2s due to their premature termination mutation in the DELLA motif. Dill et al. (2004) also demonstrated that SLY1 interacts directly with RGA and GAI via the C-terminus of the GRAS domain, suggesting that the GRAS domain is essential for RHT-1 protein-protein interactions. The frameshift mutation in RHT-B1k prevented the interaction with TaGID2s, but RHT-D1g, possessing another frameshift mutation, was able to interact with TaGID2s, and the missense mutation of RHT-D1h in the LHRII motif resulted in the loss of the DELLA-GID2 interaction, indicating that the TaGID2 and RHT-1 interaction occurs in the LHRII motif of the DELLA protein.
We demonstrated previously that the wheat TaGID1 GA receptor interacts with the RHT-A1a, RHT-B1a, and RHT-D1a DELLA proteins in the presence of GA 3 [33]. In this study, we employed yeast two-hybrid, BiFC, and yeast three-hybrid assays to investigate the interactions between TaGID1-RHT1-TaGID2. The results showed that RHT-B1b and RHT-B1e did not interact with TaGID1 because of the stop codon, indicating that the TaGID1 and RHT-1 interaction occurred in the DELLA domain. However, as a result of the frameshift mutation in the VHIID and PFYRE motifs, TaGID1 did not interact with RHT-B1k or RHT-D1g, verifying that the GRAS domain is required for DELLA protein function.
RHT-1 interacted with the GA receptor TaGID1 in the DELLA domain of the RHT-1 protein after TaGID1 received the GA signal. This interaction induced TaGID2 to recognize the RHT-1 protein through the LHRII motif of the RHT-1 protein, and the E3 ubiquitin ligase SCF complex recognized the GA-TaGID1-RHT1-TaGID2 complex through the TaGID2 and TSK1 interaction, triggering degradation of RHT-1 via the ubiquitin-proteasome pathway. These findings are concordant with the well-known GA-GID1-DELLA model for the GA signaling pathway [6][7][8][9][10][11]. The identification and functional analysis of another GA receptor reported by Yano et al. (2015) and new proteins that interact with RHT-1s, will help us better understand the molecular mechanism of GA signal transduction in wheat [44]. TaGID1s were, from left to right, TaGID1-A1, TaGID1-B1, and TaGID1-D1 used as baits; SLN1 and RHT-1s were used as preys. The AD and BK constructs were co-transformed into the yeast strain AH109, then transformants grew on the SD/-Leu-Trp media (control) and the SD/-Leu-Trp-His-Ade media which supplemented with or without 100 μM GA 3 . All experiments were conducted three times, and five clones were used each time.