BnGF14-2c Positively Regulates Flowering via the Vernalization Pathway in Semi-Winter Rapeseed

14-3-3s are general regulatory factors (GF14s or GRFs) involved in a variety of physiological regulations in plants, including the control of flowering time. However, there are poorly relevant reports in rapeseed so far. In this study, we identified a homologous 14-3-3 gene BnGF14-2c (AtGRF2_Like in Brassica napus) in rapeseed based on bioinformatic analysis by using the sequences of the flowering-related 14-3-3s in other plant species. Then, we found that overexpression of BnGF14-2c in the semi-winter rapeseed “93275” promoted flowering without vernalization. Moreover, both yeast two-hybrid and bimolecular fluorescence complementation analysis indicated that BnGF14-2c may interact with two vernalization-related flowering regulators BnFT.A02 and BnFLC.A10., respectively. qPCR analysis showed that the expression of BnFT (AtFT_Like) was increased and the expression of two selected vernalization-related genes were reduced in the overexpression transgenic plants. Further investigation on subcellular localization demonstrated that BnGF14-2c localized in the nucleus and cytoplasm. The results of RNA-seq analysis and GUS staining indicated that BnGF14-2c is ubiquitously expressed except for mature seed coat. In general, the interaction of 14-3-3 and FLC was firstly documented in this study, indicating BnGF14-2c may act as a positive regulator of flowering in rapeseed, which is worthy for more in-depth exploration.


Introduction
Flowering, the transition from vegetative to reproductive, as a crucial decision feature in the life cycle of higher plants, is switched by both external and internal signals [1][2][3]. To date, the underlying mechanisms of floral transition have been well-characterized in the model plant Arabidopsis, including photoperiod pathway, vernalization pathway, autonomous pathway, gibberellins pathway and age pathway [4,5]. These pathways act independently and involve extensive crosstalks and feedback loops with each other [6,7]. A process of low temperature induction during the vegetative growth period known as vernalization is required for flowering in many plants [8][9][10]. The vernalization-requiring monocarpic plants are called biennial or winter annual [11,12], and the rapid flowering types that can complete the life cycle in one season are named summer or spring annual [13].

Characterization and Sequence Alignment of BnGF14-2c in Rapeseed
BnGF14-2c (BnaA07g34180D, Darmor-bzh) is one of the 47 14-3-3 proteins in rapeseed identified by blastP and phylogenetic analysis (data were not shown). The full-length cDNA of BnGF14-2c is 1203 bp including a 897 bp open reading frame encoding a protein that with 299 amino acids. Multiple sequence alignment was performed with the 14-3-3 orthologs from other plants, which have been proved to be participating in the flowering regulation, and the result showed that BnGF14-2c is highly conserved with these 14-3-3s ( Figure 1a). All of the sequences have the conserved core region comprising nine antiparallel α-helices and divergent N/C terminis. The model image of 3D matching prediction showed that BnGF14-2c could form a homologous dimer with a horseshoe shape, typically of 14-3-3 proteins (Figure 1b). Therefore, we made a bold speculation that BnGF14-2c may also function in the flowering regulation in rapeseed. OsGF14-c is a 14-3-3 in rice; Gh14-3-3 is a 14-3-3 in cotton. 14-3-3 Proteins contain nine conserved α-helices. (b): Homodimer 3D structure modeling of BnGF14-2c. The letter indicated by the red dot is the amino acid corresponding to the start codon, '*' indicates positions which have a single, fully conserved residue, ':' indicates that one of the following 'strong' groups is fully conserved. AtGF14µ and AtGF14ν are 14-3-3s in Arabidopsis; LeGF14-2, LeGF14-5 and LeGF14-74 are 14-3-3s in tomato; OsGF14c is a 14-3-3 in rice; Gh14-3-3 is a 14-3-3 in cotton. 14-3-3 Proteins contain nine conserved α-helices. (b): Homodimer 3D structure modeling of BnGF14-2c. The letter indicated by the red dot is the amino acid corresponding to the start codon, '*' indicates positions which have a single, fully conserved residue, ':' indicates that one of the following 'strong' groups is fully conserved.

Overexpression of BnGF14-2c Showed an Early Flowering Phenotype without Vernalization in Semi-Winter Rapeseed
The full-length open reading frame of BnGF14-2c was cloned from Darmor and introduced into the semi-winter rapeseed "93275" to generate 35S::BnGF14-2c overexpression transgenic plants. A total of 12 independent transgenic lines exhibiting herbicide resistance in the T1 generation was confirmed by PCR with the specific primers (Table S1). Among them, four lines showed an early flowering phenotype without vernalization compared to the wild-type (WT), and two lines were selected for further analysis (OE-21 # , OE-22 # ) ( Figure 2a). Real-time quantitative PCR analysis showed that the expression levels of BnGF14-2c in transgenic lines were at least three-folds higher than WT (Figure 2b). Subsequent phenotypic observations found that the transgenic plants bolted and bloomed about 60 days after sowing, while wild-type plants had to be delayed for a long time (more than 170 days) to partially enter into the reproductive growth stage (some individual wild-type plants even took longer to flowering or did not bloom at all in the greenhouse) (Figure 2a,c). In addition, we found that the number of rosette leaves (when the first floral bud was visible) of wild-type was more than twice as much as the transgenic lines ( Figure 2d).

Overexpression of BnGF14-2c Showed an Early Flowering Phenotype without Vernalization in Semi-Winter Rapeseed
The full-length open reading frame of BnGF14-2c was cloned from Darmor and introduced into the semi-winter rapeseed "93275" to generate 35S::BnGF14-2c overexpression transgenic plants. A total of 12 independent transgenic lines exhibiting herbicide resistance in the T1 generation was confirmed by PCR with the specific primers (Table S1). Among them, four lines showed an early flowering phenotype without vernalization compared to the wild-type (WT), and two lines were selected for further analysis (OE-21 # , OE-22 # ) ( Figure 2a). Real-time quantitative PCR analysis showed that the expression levels of BnGF14-2c in transgenic lines were at least three-folds higher than WT (Figure 2b). Subsequent phenotypic observations found that the transgenic plants bolted and bloomed about 60 days after sowing, while wild-type plants had to be delayed for a long time (more than 170 days) to partially enter into the reproductive growth stage (some individual wild-type plants even took longer to flowering or did not bloom at all in the greenhouse) (Figure 2a,c). In addition, we found that the number of rosette leaves (when the first floral bud was visible) of wild-type was more than twice as much as the transgenic lines (Figure 2d).

BnGF14-2c May Interact with FT and FLC Respectively Both In Vitro and In Vivo
Previous studies have indicated that FT or FT_Like proteins can interact with 14-3-3 isoforms affecting flowering time in many plant species, and the interactions were observed both in the cytoplasm and the nuclei [48][49][50][51][52], but the FLC protein has not been reported to interact with 14-3-3s. As a heterotetraploid plant species, rapeseed has a high copy number of homologous genes, and these genes maybe subfunctionalized to different degrees. Previous reports have been identified that BnFT.A02 as a candidate gene for flowering in rapeseed [66][67][68], BnFLC.A02 and BnFLC.A10 were two floral integrator genes with the vernalizing conditions in winter rapeseed [15,69]. To testify the possible role of BnGF14-2c in the vernalization pathway, the florigen gene (BnFT.A02) [68] and a key gene in the vernalization pathway (BnFLC.A10) were selected as candidate interacting proteins [58,70].
Y2H was performed by cloning the coding sequences of BnGF14-2c into the prey vector pGBKT7, while BnFT.A02 and BnFLC.A10 into the bait vector pGADT7 (Figure 3). The plaque of SD/-LT+X (SD/-Leu/-Trp/X-α-gal) were all in good condition, indicating that the co-transformation of the prey vector and the bait vector was successful. The positive control grew well on the SD/-LTHA+X (SD/-Leu/-Trp/-His/-Ade/X-α-gal) medium no matter with or without 3-AT, while the negative control group was inhibited seriously. The auto-activation verification group could grow on the SD/-LTHA+X without 3-AT, but the plaque turned red and the growth was inhibited. The auto-activation group did not grow after the addition of 3 mM 3-AT, indicating that BnGF14-2c had a certain degree of self-activation activity, but the activity could be inhibited under the action of 3 mM 3AT. The test group of AD:BnFT.A02 and BK:BnGF14-2c was grown well on the SD/-LTHA+X medium, even with 3 mM 3-AT, suggesting that the interaction was strong. The interaction between AD:BnFLC.A10 and BK:BnGF14-2c was weak indicated by the red plaque that grew slowly. Taken together, our results suggested that BnGF14-2c possessed the abilities of interaction with BnFT.A02 and BnFLC.A10 in vitro.

BnGF14-2c May Interact with FT and FLC Respectively Both In Vitro and In Vivo
Previous studies have indicated that FT or FT_Like proteins can interact with 14-3-3 isoforms affecting flowering time in many plant species, and the interactions were observed both in the cytoplasm and the nuclei [48][49][50][51][52], but the FLC protein has not been reported to interact with 14-3-3s. As a heterotetraploid plant species, rapeseed has a high copy number of homologous genes, and these genes maybe subfunctionalized to different degrees. Previous reports have been identified that BnFT.A02 as a candidate gene for flowering in rapeseed [66][67][68], BnFLC.A02 and BnFLC.A10 were two floral integrator genes with the vernalizing conditions in winter rapeseed [15,69]. To testify the possible role of BnGF14-2c in the vernalization pathway, the florigen gene (BnFT.A02) [68] and a key gene in the vernalization pathway (BnFLC.A10) were selected as candidate interacting proteins [58,70].
Y2H was performed by cloning the coding sequences of BnGF14-2c into the prey vector pGBKT7, while BnFT.A02 and BnFLC.A10 into the bait vector pGADT7 (Figure 3). The plaque of SD/-LT+X (SD/-Leu/-Trp/X-α-gal) were all in good condition, indicating that the co-transformation of the prey vector and the bait vector was successful. The positive control grew well on the SD/-LTHA+X (SD/-Leu/-Trp/-His/-Ade/X-α-gal) medium no matter with or without 3-AT, while the negative control group was inhibited seriously. The auto-activation verification group could grow on the SD/-LTHA+X without 3-AT, but the plaque turned red and the growth was inhibited. The auto-activation group did not grow after the addition of 3 mM 3-AT, indicating that BnGF14-2c had a certain degree of self-activation activity, but the activity could be inhibited under the action of 3 mM 3AT. The test group of AD:BnFT.A02 and BK:BnGF14-2c was grown well on the SD/-LTHA+X medium, even with 3 mM 3-AT, suggesting that the interaction was strong. The interaction between AD:BnFLC.A10 and BK:BnGF14-2c was weak indicated by the red plaque that grew slowly. Taken together, our results suggested that BnGF14-2c possessed the abilities of interaction with BnFT.A02 and BnFLC.A10 in vitro. Figure 3. Yeast two−hybrid analysis of the interactions between BnGF14-2c and the candidate interactors. Yeasts expressing the indicated combinations of bait and prey were spotted on the synthetic dropout medium without leucine and tryptophan (SD-Leu/-Trp/X-α-gal) and SD medium without leucine, tryptophan, histidine, and adenine (SD-Leu/-Trp/-His/-Ade/X-α-gal), both mediums were supplemented with X-α-Gal. Yeast strains expressing full length of BnGF14-2c (PGBKT7), BnFT.A02 (PGADT7) and BnFLC.A10 (PGADT7) were indicated in figures, respectively. All the experiments were repeated for three times. Yeast two−hybrid analysis of the interactions between BnGF14-2c and the candidate interactors. Yeasts expressing the indicated combinations of bait and prey were spotted on the synthetic dropout medium without leucine and tryptophan (SD-Leu/-Trp/X-α-gal) and SD medium without leucine, tryptophan, histidine, and adenine (SD-Leu/-Trp/-His/-Ade/X-α-gal), both mediums were supplemented with X-α-Gal. Yeast strains expressing full length of BnGF14-2c (PGBKT7), BnFT.A02 (PGADT7) and BnFLC.A10 (PGADT7) were indicated in figures, respectively. All the experiments were repeated for three times.
To further testify these interactions in vivo, the BiFC assays in the leaves of tobacco (Nicotiana benthamiana) were carried out. The fluorescence signals from NE:BnGF14-2c + CE:BnFT.A02 interaction were detected in the cytoplasm and the nucleus. The interaction of NE:BnGF14-2c + CE:BnFLC.A10 was mainly detected in the nucleus, and slightly lower in the cytoplasm, but no GFP signal was detected in the NE:BnGF14-2c + CE ( Figure 4). All these results suggested that BnGF14-2c may interact with BnFT.A02 and BnFLC.A10 in plant cells in vivo.

BnGF14-2c Negatively Affect the Expressions of Two Vernalization-Related Genes and Positively Influence the BnFT Expression
To study the molecular mechanisms of BnGF14-2c in the flowering vernalization pathway, we then tested the expression levels of two vernalization-related genes (BnFLC, BnFRI) and a florigenic gene (BnFT) by qPCR. The results showed that BnFT expression was dramatically increased in BnGF14-2c overexpressing transgenic lines than that in the wild type. While, the expression levels of BnFLC and BnFRI, were decreased in those transgenic lines compared with that of wild-type plants ( Figure 5). Therefore, our results implied that BnGF14-2c may negatively affect the expression levels of the two selected To study the molecular mechanisms of BnGF14-2c in the flowering vernaliz pathway, we then tested the expression levels of two vernalization-related genes (B BnFRI) and a florigenic gene (BnFT) by qPCR. The results showed that BnFT expre was dramatically increased in BnGF14-2c overexpressing transgenic lines than that wild type. While, the expression levels of BnFLC and BnFRI, were decreased in transgenic lines compared with that of wild-type plants ( Figure 5). Therefore, our r implied that BnGF14-2c may negatively affect the expression levels of the two se vernalization-related genes and positively influence the BnFT expression. These r might explain why overexpression of BnGF14-2c in semi-winter rapeseed pro flowering without vernalization.

BnGF14-2c Is Ubiquitously Expressed
To characterize the expression patterns of BnGF14-2c in detail, we constructed map to exhibit the expression levels of BnGF14-2c across tissues in two rapeseed cul ("ZS11" and "862") according to the transcriptome data obtained previously in our ( Figure 6). The result indicated that BnGF14-2c was generally expressed at relatively levels in most tissues and stages of these two varieties. The gene was particularly h expressed in the stamen, stem, and flower bud.
Promoter determines the spatiotemporal pattern of gene expression and usua cates upstream of the initial codon about 1000-2000 bp in length. To check the pro activity of BnGF14-2c, a 1495 bp upstream sequence of BnGF14-2c from the starting was cloned as the promoter sequence and fused with the reporter gene GUS to con the plant expression vector proBnGF14-2c::GUS, and then transformed into the wild Arabidopsis by the floral-dip method. Homozygous transgenic plants were obtain continuous self-crossing in the T3 generation. Further GUS staining results demons that GUS activity could be detected in almost all the tissues focused (seedling, leaf stem, flower, and silique), except for the mature seed coat (Figure 7). All these r implied that the BnGF14-2c promoter may be a constituent promoter driving u tously gene expression in all tissues and the BnGF14-2c gene may be involved in va aspects of plant growth and development.

BnGF14-2c Is Ubiquitously Expressed
To characterize the expression patterns of BnGF14-2c in detail, we constructed a heat map to exhibit the expression levels of BnGF14-2c across tissues in two rapeseed cultivars ("ZS11" and "862") according to the transcriptome data obtained previously in our group ( Figure 6). The result indicated that BnGF14-2c was generally expressed at relatively high levels in most tissues and stages of these two varieties. The gene was particularly highly expressed in the stamen, stem, and flower bud.  Promoter determines the spatiotemporal pattern of gene expression and usually locates upstream of the initial codon about 1000-2000 bp in length. To check the promoter activity of BnGF14-2c, a 1495 bp upstream sequence of BnGF14-2c from the starting code was cloned as the promoter sequence and fused with the reporter gene GUS to construct the plant expression vector proBnGF14-2c::GUS, and then transformed into the wild-type Arabidopsis by the floral-dip method. Homozygous transgenic plants were obtained by continuous self-crossing in the T3 generation. Further GUS staining results demonstrated that GUS activity could be detected in almost all the tissues focused (seedling, leaf, root, stem, flower, and silique), except for the mature seed coat (Figure 7). All these results implied that the BnGF14-2c promoter may be a constituent promoter driving ubiquitously gene expression in all tissues and the BnGF14-2c gene may be involved in various aspects of plant growth and development.

Subcellular Localizations of BnGF14-2c, BnFT.A02 and BnFLC.A10
BnGF14-2c was firstly predicted to localize to the nucleus, but some studies have pointed out that the 14-3-3 may also locate in the cytoplasm. In order to verify this result, GFP was fused to the C-terminus of BnGF14-2c and the fusion construct was introduced into tobacco epidermal cells for transient expression. Fluorescence observations showed that BnGF14-2c was localized in the nucleus and cytoplasm, and the fluorescence signal was particularly strong in the nucleus (Figure 8). The subcellular localizations of BnFT.A02 and BnFLC.A10 were also detected according to the method mentioned above. Our results indicated that BnFT.A02 and BnFLC.A10 may both be localized in the nucleus and cytoplasm (Figure 4). into tobacco epidermal cells for transient expression. Fluorescence observations showed that BnGF14-2c was localized in the nucleus and cytoplasm, and the fluorescence signal was particularly strong in the nucleus (Figure 8). The subcellular localizations of BnFT.A02 and BnFLC.A10 were also detected according to the method mentioned above. Our results indicated that BnFT.A02 and BnFLC.A10 may both be localized in the nucleus and cytoplasm (Figure 4). BnFLC.A10-GFP, respectively. 35S::GFP, free GFP signals were located in the cytoplasm. Merged, this micrograph was a combination of the three diagrams described below. 488 nm/GFP, GFP fluorescence. 561 nm/mCherry, FIB2-mCherry specifically localized to the nucleolus was used as a marker. DIC, bright-field. All the experiments were repeated three times.
The result of multiple sequence alignment demonstrated that the isoforms of GF14s possess a conserved core region comprising nine antiparallel α-helices and divergent N/C termini (Figure 1a). The conserved antiparallel α-helices are capable of binding a separate phosphorylated target protein to participate in diverse pathways. The primary diversity occurs at the N and C termini, which are related to dimerization and target binding, respectively. Intriguingly, the N terminal sequence of BnGF14-2c is unimaginable longer than any of the other family members (Figure 1a). After careful comparison, we found BnFLC.A10-GFP, respectively. 35S::GFP, free GFP signals were located in the cytoplasm. Merged, this micrograph was a combination of the three diagrams described below. 488 nm/GFP, GFP fluorescence. 561 nm/mCherry, FIB2-mCherry specifically localized to the nucleolus was used as a marker. DIC, bright-field. All the experiments were repeated three times.
The result of multiple sequence alignment demonstrated that the isoforms of GF14s possess a conserved core region comprising nine antiparallel α-helices and divergent N/C termini (Figure 1a). The conserved antiparallel α-helices are capable of binding a separate phosphorylated target protein to participate in diverse pathways. The primary diversity occurs at the N and C termini, which are related to dimerization and target binding, respectively. Intriguingly, the N terminal sequence of BnGF14-2c is unimaginable longer than any of the other family members (Figure 1a). After careful comparison, we found that the 40th amino acid at the N-terminal may be another start codon. If removing the 39 amino acids at the N-terminal will have no effect on the conserved sequence at all, and the homodimer model was also unaffected. Based on these results, we speculated that this may be a part of UTR (Untranslated Regions) or a subfunctionalization of BnGF14-2c.
As a leading consumer of edible oil in the world, China is short of edible vegetable oil and relies heavily on imports. If the planting area of rapeseed can be expanded, this situation will be greatly alleviated. The double-cropping rice area is the area with the largest potential resources of rapeseed in China. However, after the harvest of doublecropping rice, numerous winter idle fields are abandoned due to the lack of rapeseed varieties that can be sown late and harvested early. Therefore, the selection and creation of special early-maturing rapeseed varieties with a shorter duration period can make full use of winter idle fields in double-cropping rice areas and relieve the pressure on China's edible oil imports. Although the flowering traits are significant for rapeseed, but there is still no report about the involvement of 14-3-3 in regulating rapeseed flowering. Wild-type "93275" as a typical semi-winter rapeseed is widely adapted to the climatic conditions of the middle and lower reaches of the Yangtze River. In this study, overexpressing BnGF14-2c in "93,275" could promote flowering without vernalization, both in the greenhouse and field experiments (Figure 2a and Figure S1). Furthermore, the transgenic plants growing in the field showed dwarf and multi-branches with inconspicuous main stems. It is possibly due to the precocious bolting and flowering of the transgenic plants before winter comes, and then the low-temperature stress in winter prevents the growth of the main inflorescence by suppressing the apical dominance, which led to the plants to form relative more side branches in the next springtime.
Although rapeseed and Arabidopsis share common ancestor species, B. napus (n = 19, AA = 10, CC = 9) is an allotetraploid hybrid of B. rapa (AA) and B. oleracea (CC) under natural conditions, leading to many more gene copies within the B. napus by a series of whole-genome duplication. For example, the FT gene has six homologous in rapeseed scattering on different chromosomes, such as A2, A7, C2, C4, and C6 [68,71]. Likewise, nine copies of BnFLCs are located on A2, A3, A10, C2, C3, and C9 [72]. Except for the divergence of intron and promoter regions, the coding region sequences of these homologous genes are highly conserved, but the gene expression patterns show great differences [68,[71][72][73]. In Arabidopsis, vernalization suppresses the floral repressor gene FLC and upregulates floral promotion genes FT [74]. Previous reports have identified BnFT.A02 as a candidate gene for flowering in rapeseed [66][67][68]. The studies on vernalization requirement and response have also found that the allelic variations of BnFLC.A02 and BnFLC.A10 were two floral integrator genes with the vernalizing conditions in European winter rapeseed [15,69]. Here, we selected BnFT.A02 and BnFLC.A10 as the two candidate proteins to check their interactions with BnGF14-2c, and our results from Y2H and BiFC both verified the interactions.
The qPCR showed that the expression of BnFT was up-regulated and two genes (BnFLC and BnFRI) participating in the vernalization pathway were down-regulated in the rosette leaves before the bolting stage in the BnGF14-2c overexpressing lines (Figure 2a). Therefore, we speculated that BnGF14-2c may be involved in the fine-tuning of the rapeseed flowering via at least the vernalization pathway. Whether BnGF14-2c possibly involved in the other flowering pathways is still a question needed to be answered in the future. In conclusion, we discovered that BnGF14-2c, a conserved 14-3-3 gene, may act as a positive regulator of flowering in rapeseed possibly via the vernalization pathway in semi-winter rapeseed.

Plant Materials and Phenotypic Measurements
The semi-winter rapeseed cultivar "93275" was used as recipient material for genetic transformation. The greenhouse was under the condition (

Vectors Construction and Transformation
According to the published rapeseed genomic sequence of "Dormar", we designed the cloning primers specific for BnGF14-2c and obtained the sequence by PCR cloning (the primers were listed in Table S1). The plant binary overexpression vector 35S::BnGF14-2c was constructed by the TOPO recombination method and pEarleyGate100 [75] as the skeleton vector. Then, the hypocotyl segments of rapeseed "93275" were impregnated by Agrobacterium tumefaciens to obtain transgenic plants as described previously [76].

RNA Extraction and Quantitative Real-Time PCR
Total RNA was extracted from the leaves at the vegetative stage of rapeseed using the Plant RNA Extraction Kit (TaKaRa, Maebashi, Japan) with PrimeScript™ RT reagent Kit (TaKaRa, Japan) that was used for cDNA synthesis. Quantitative Real-Time PCR was conducted by Hieff ® qPCR SYBR Green Master Mix (Low Rox Plus) according to the methods established in our previous studies [77]. Three replicates were performed for the analysis of each gene. The rapeseed BnTMA7 gene was used as an internal control. The primers used in this study were listed in Table S1.

Bimolecular Fluorescence Complementation
The sequences of BnGF14-2c, BnFT, and BnFLC were cloned from the leaf cDNA of Darmor. BiFC vectors (pSPYNE-35S, pSPYCE-35S) were constructed as NE-BnGF14, CE-BnFT, CE-BnFLC (the primers were listed in Table S1), and these constructed vectors were transiently transformed into the tobacco leaves as described by Katia [78]. The nucleolus mCherry marker was used simultaneously and its signals were observed as mentioned above.

Yeast Two-Hybrid
The yeast two-hybrid vector system pGBKT7/pGADT7 was used to test the interactions between BnGF14-2c and the other candidate proteins. The vectors were constructed as pGBKT7-BnGF14-2c, pGADT7-BnFT, and pGADT7-BnFLC (the primers were listed in Table S1). The plasmids were then transformed into AH109 strain and selected on the control media (SD-Leu/-Trp/X-α-gal) plate by incubation for three days at 30 • C. The interactions were detected on the selective plates (SD-Leu/-Trp/-His/-Ade/X-α-ga). Grew on the SD-Leu/-Trp/X-α-gal indicative of a successful transformation with both bait and prey plasmids. The yeast plaques grow on the SD-Leu/-Trp/-His/-Ade/X-α-gal with the blue color indicating an interaction between both proteins. Plaque cannot grow or turn red, indicating no interaction. AD-largeT and BK-p53 were the positive control, AD-largeT and BK-laminC were the negative control, AD and BK-BnGF14-2c were the negative controls to eliminate the self-activating activity.

Histochemical GUS Staining
A 1495 bp fragment upstream of the BnGF14-2c start codon (ATG) was cloned into binary vector pDX2181G to fuse with the GUS reporter gene (the primers were listed in Table S1) and then transformed into the Arabidopsis. After several rounds of hygromycin selection and PCR verification, a total of 16 positive and homozygous lines were obtained. Then, the tissues (7-day-after-sowing seedling, leaf, root, stem, flower, and 15-day-afterflowering silique) were collected for GUS staining according to the method described by Jefferson [79].

Statistical Analysis
All the experiments were repeated three times. The data were analyzed with SPSS 17.0 (SPSS Inc., Chicago, IL, USA) and was presented as the mean standard error. The mRNA expression levels were analyzed with a comparative 2 −∆∆Ct method. For comparison of two groups of data, the two-sided Student's t-test was used, thereinto, asterisks "*" p < 0.05 and "**" p < 0.01 was considered to be statistically significant.