Alteration of SlYABBY2b gene expression impairs tomato ovary locule number and endogenous gibberellin content

Abstract

Tomato is an ideal model species for fleshy fruit development research. SlYABBY2b regulates the ovary locule number, which is increased by gibberellins, in tomato. However, the relationship between SlYABBY2b and endogenous gibberellin is poorly understood. In this study, SlYABBY2b-overexpressing and RNA interference (RNAi) transgenic tomato plants were used to elucidate the mechanism by which SlYABBY2b regulates the ovary locule number and endogenous gibberellin content in tomato. SlYABBY2b-overexpressing plants showed fewer locules and lower gibberellin content than the control plants. Contrasting results were found in the RNAi lines. Therefore, the SlYABBY2b gene negatively regulates tomato ovary locule number and endogenous gibberellin content. Furthermore, the expression of SlYABBY2b gene was remarkably higher than that of the wild type in the apical shoots of gibberellin-deficient mutants. This showed that the gibberellins can inhibit the expression of SlYABBY2b gene negative regulation. Further study revealed that SlYABBY2b suppressed the expression of SlGA20ox1 and SlGA3ox2, but increased that of SlGA2ox1 and SlGA2ox5 in the apical shoots of SlYABBY2b-overexpressing plants, thereby reducing gibberellin content. Contrasting results were found in the RNAi lines. Our results showed that the SlYABBY2b gene was located on gibberellin signal transduction pathways, fed back regulation of the synthesis of gibberellin, and felt exogenous gibberellin signal to further regulate the formation of tomato locule.

1. Introduction

Tomato is an ideal model species for fleshy fruit development research. Domesticated tomato fruit is enlarged 1000 times compared to their wild progenitors, which is an extreme case. Fruit size and weight are the primary characteristics of commercial tomato varieties. Increases in fruit weight and fruit size are controlled by multiple quantitative trait loci (QTL), and some of them have been cloned or identified, namely, six fruit weight QTLs and three locule number QTLs (fw2.2, fw3.2, fw1.1, fw3.3, fw6.1, fw11.2, lcn2.1, lcn2.4, lcn5.1) (Fukazawa et al., 2000; Foolad, 2007; Chakrabarti et al., 2013; van der Knaap et al., 2014; Fernández-Lozano et al., 2015; Hernández-Bautista et al., 2015; Illa-Berenguer et al., 2015). Locule number is associated with fruit size and weight, and almost all wild species and several small-fruited varieties produce fruits with only 2‒4 locules, whereas most cultivar varieties consumed today can develop fruits with up to 10 locules. Increased locule number contributes as much as 50% variance to fruit enlargement and is believed to represent the second major step in the tremendous increase in tomato fruit size in evolution (Lippman and Tanksley, 2001; Tanksley, 2004). Molecular genetic studies have so far in tomato identified two traits that govern fruit locule number, QTL fasciated and QTL lc, which control floral meristem size and ultimately the development of supernumerary locules (Lippman and Tanksley, 2001; Barrero and Tanksley, 2004; Barrero et al., 2006; Muños et al., 2011; van der Knaap et al., 2014; Xu et al., 2015). Both of these traits have an effect on the shape and fruit size in the tomato (Muños et al., 2011; Rodríguez et al., 2011). Regarding the lc locus, which has a weaker effect, located in a 1608-bp non-coding region, two single-nucleotide polymorphisms (SNPs) are responsible for the extreme high-locule-number phenotype (Muños et al., 2011). Two SNPs were proposed to disrupt the repression of SlWUS by TOMATO AGAMOUS-LIKE1 (TAG1), which is the homolog of the MADS-box gene of Arabidopsis flower development AGAMOUS (van der Knaap et al., 2014). SlWUS and peptide SlCLAVATA3 (SlCLV3) interact in a negative feedback loop (Schoof et al., 2000). Recent studies have shown that down-regulation of SlWUS affected tomato flower and locule development (Li et al., 2017). The fasciated mutation contains a 294-kb inversion with two breakpoints, named the first intron 1 of SlYABBY2b and upstream of the tomato SlCLV3 start codon (Huang and van der Knaap, 2011), resulting in reduced messenger RNA (mRNA) accumulation of the SlCLV3 gene (Xu et al., 2015). SlYABBY2b, one of the 9 Solanum lycopersicum SlYABBY genes (Huang et al., 2013), is the first found gene encoding a YABBY-like transcription factor that controls fruit locule number development in tomato (Cong et al., 2008).

Increases in fruit weight and fruit size are controlled by tomato fruit locule number. However, the tomato fruit locule number is strongly linked to malformation. Fruits with several ovary locules and large sizes are prone to malformation. Fruit malformation is affected by the environment and hormones (Asahira et al., 1982; Tomer et al., 1998; Li et al., 2008), and gibberellin (GA) content has also been associated with fruit malformation (Sawhney and Greyson, 1971; Sawhney and Dabbs, 1978; Liu and Li, 2012).

Fruit development is a complex and precise genetically regulated process (Seymour et al., 2013). Plant hormones play a key role in controlling fruit growth and development in the tomato. GAs are growth factors that participate in the process of plant growth and development (Olszewski et al., 2002; Kumar et al., 2014; Pesaresi et al., 2014), and the metabolic pathway of the plant is well understood. GA levels are influenced by GA biosynthesis and inactivation (Hedden and Phillips, 2000). These enzymes, named GA20ox, GA3ox, and GA2ox, involved in the GA metabolism pathway, are encoded by small gene families and regulate bioactive GA levels (Hedden and Phillips, 2000; Sakamoto et al., 2004; Yamaguchi, 2008). Many transcription factors, including those from the YABBY family, are also involved in regulating GA biosynthesis in plants (Fukazawa et al., 2000; Rosin et al., 2003; Gazzarrini et al., 2004; Ishida et al., 2004; Magome et al., 2004; Wang et al., 2004; Dai et al., 2007), and thus in controlling the growth and development of plants. YABBY transcription factors are closely linked to GA synthase genes in Arabidopsis and rice (Hay et al., 2002; Kumaran et al., 2002; Dai et al., 2007).SlYABBY2b, which was the first YABBY transcription factor identified in tomato, regulates fruit development by increasing locule number. GAs also play a similar role in fruit development. However, the relationship between SlYABBY2b and endogenous GA is poorly understood. In the present study, we analyzed the regulatory effects of SlYABBY2b on tomato ovary locule number and endogenous GA content through transgenic approaches.

2. Materials and methods

2.1. Plant materials

Tomato (Solanum lycopersicum L.) lines MLK1 and FL1 were used in the research. The “MLK1” line, which has multiple locules, and the “FL1” line, which has few locules, were obtained from Shenyang Agricultural University, Shenyang, China. The other agronomic characteristics are similar, except for the differences in locule number. The locule numbers of MLK1 and FL1 are about 14 and 2, respectively. All gib mutants and wild types were obtained from the University of California, Davis, USA. Plants were grown in soil under greenhouse growing conditions (25 °C day, 15 °C night) in September 2015 at Shenyang Agriculture University. gib mutants and wild types are listed in Table S1).

2.2. Construction of SlYABBY2b RNAi and overexpressing vectors, and tomato transformation

A 551-bp fragment of SlYABBY2b was amplified from “MLK1” by real-time polymerase chain reaction (RT-PCR). Specific primers were designed using the sequences of the SlYABBY2b gene from tomato (Gene Bank accession number EU557674) (Cong et al., 2008). The overexpression-and RNAi-specific primers used for the SlYABBY2b gene are as follows: SlYABBY2b 1: CACCTCCCCTTTGATCCATGTTCT (forward) and CGCTATTTGTTGCCCTCC (reverse); SlYABBY2b 2: TCCCCTTTGATCCATGTTCT (forward) and CGCTATTTGTTGCCCTCC (reverse). The sequence-confirmed amplified fragments were cloned into the Gateway-compatible vector pENTR/D-TOPO and pCR8⁄GW⁄TOPO entry vectors via a TOPO cloning reaction (Invitrogen), respectively. The nucleotide sequence was verified, and the fragments were then transferred to the binary vectors pB7WG2D and pB7GWIWG2 via LR recombinant reaction, respectively. Through sequence analysis and restriction enzyme digestion, the constructs containing the expected insert were introduced into Agrobacterium tumefaciens LBA4404 cells by electroporation.

Seeds of “MLK1” and “FL1” were surface-sterilized with 70% (v/v) alcohol for about 30 s, sown on 1/2 Murashige and Skoog (MS) medium, and cultured in a 16-h day/8-h night regime at 25 °C for 6–8 d. The explants from tomato seedling cotyledons were cut and pre-cultured on a pre-medium in darkness at 25 °C for 48 h. After pre-culturing, A. tumefaciens LBA4404 cells were used to infect the explants for 4 min with slow shaking. After infection, the explants were cultured on a cocultivation medium for 2 d in darkness at 25 °C. Then, the explants were cultured on selective medium, which contained MS salts, 3% (0.03 g/ml) sucrose, 7 g/L agar, 2 mg/L 6-benzylaminopurine (6-BA), 0.2 mg/L indole-3-acetic acid (IAA), 400 mg/L cephalosporin, and 0.5 mg/L glufosinate-ammonium to induce regeneration. Explants that regenerated plantlets were transplanted on a rooting medium including MS salt, 3% (0.03 g/ml) sucrose, 5 g/L agar, 400 mg/L cephalosporin, and 0.05 mg/L 1-naphthlcetic acid (NAA). The plants were grown in soil under greenhouse growing conditions.

2.3. Total RNA extraction and RT-PCR analysis

For quantitative real-time PCR (qRT-PCR), total RNA was extracted using TRIzol® reagent followed by the DNA-free™ kit (Ambion) in accordance with the instructions. Complementary DNA (cDNA) samples were synthesized from 1 µg of RNA using cDNA Archive kit (Life Technologies, USA). qRT-PCR was performed following the method described by Jain et al. (2006). RT-PCR amplification was conducted on an Applied Biosystems 7500 real-time system using SYBR Green PCR Master Mix (Life Technologies, USA) as described in accordance with the instructions. Three independent qRT-PCR analyses were carried out for each cDNA preparation. Data were normalized to actin and the specificity of the reactions was verified by the ΔΔC _T calculation method. The sequence of specific primers used for RT-PCR is listed in Table S2). Significant difference analysis was tested using Duncan’s multiple range test, with P<0.05 being considered statistically significant. The data were analyzed using Origin 8.0. Analysis of variance (ANOVA) was performed using SPSS 13.0 software.

2.4. Seedling sampling

Stem apices of seedlings were selected at the same stage. The stem apices of the transgenic and wild-type plants were observed under a microscope during the flower bud differentiation phase. The stem apices of the transgenic and control plants were dissected (30 per generation, 0.3 cm long) in triplicate. The freshly collected samples were immediately frozen in liquid nitrogen and then stored at −80 °C for RT-PCR analysis.

2.5. ELISA analysis of gibberellins

Extraction, purification, and measurement of GAs were performed using an enzyme-linked immunosorbent assay (ELISA) kit as described previously (Zheng and Zhou, 1995; Chen et al., 1998). The apical shoot samples at different developmental stages were collected in liquid nitrogen and extracted in cold 80% (v/v) ethanol solution. The samples were ground in ethanol solution with 80% (v/v) cold methanol three times in a cold mortar and moved to a suitable container in dark conditions at 4 °C overnight. Then, the supernatant extract was filtered through a C18 Sep-Pak cartridge (Waters, Milford, MA, USA). The extracts (ethyl acetate phases) were collected and dried under nitrogen, and the residue was re-dissolved in phosphate buffer saline (PBS; 0.01 mol/L, pH 7.4). Antibodies against GA were purchased from the Abmart Company (China). A solution of antibodies was coated on microtitration plates and incubated at 37 °C for 90 min. The wells were washed five times with PBS. Each well was measured for absorbency at 490 nm to determine the content of GAs.

2.6. Phenotypic analysis

The stem height and internode length of 12-week-old tomato plants grown in greenhouse conditions were measured. Stem height was measured from the bottom to the top of the tomato plant. Internode length was calculated by dividing the stem height by the leaf number. The numbers of total flowers, sepals, and petals in the first, second, and third inflorescences were counted, and the fruits were weighed. The first, second, and third clusters of three fruits from the transgenic and non-transgenic plants were harvested, and their locule numbers were determined. Significant difference analysis was tested using Duncan’s multiple range test, with P<0.05 being considered statistically significant. ANOVA was performed using SPSS 13.0 software.

3. Results

3.1. Phenotypic characterization of transgenic plants

Gene-specific PCR analysis revealed that six independent SlYABBY2b-overexpressing transgenic plants contained the desired insert (data not shown). All SlYABBY2b-overexpressing plants were morphologically distinguishable from the non-transgenic control plants. They exhibited dwarf phenotypes due to the decreased internode lengths (Fig. 1g; Table 1). Both shoot and internode lengths in all transgenic lines were significantly shorter than those of the control (Fig. 1h; Table 1). However, the stem diameter was markedly greater in the SlYABBY2b-overexpressing lines than in the control, except at the four-leaf stage (Table 1). The number of flowers in the first and second inflorescences was significantly higher in the SlYABBY2b-overexpressing lines than in the control, whereas the number of sepals was significantly lower in the SlYABBY2b-overexpressing lines than in the control (Figs. 1a and 1b; Table 1). The number of petals decreased in the transgenic lines (Table 1). The locule number significantly decreased in the overexpressing transgenic plants, especially in the fruits of the first inflorescence (Figs. 1c and 1d; Table 1). Fruit weight was also reduced with decreasing locule number (Figs. 1e and 1f). Five SlYABBY2b RNAi transgenic lines were obtained and identified by PCR analysis (data not shown). For further investigation, we selected one line because they have similar phenotypic features within the group. The SlYABBY2b RNAi plants had significantly greater plant heights and internode lengths but smaller stem diameters than the control plants (Fig. 2e; Table 2). The number of flowers in every inflorescence was significantly higher than that in the control. The numbers of sepals and petals increased (Table 2). In addition, the locule number and fruit weight of the transgenic plants significantly increased (Figs. 2a–2d; Table 2).

Fig. 1.

Phenotypic characteristics of representative wild-type and SlYABBY2b-overexpressing (OE) transgenic tomato plants

(a) Flower from MLK1 plants. (b) Flower from SlYABBY2b-OE lines. (c) MLK1 locules. (d) SlYABBY2b-OE locules. (e) MLK1 fruit size. (f) SlYABBY2b-OE fruit size. (g) MLK1 (left) and SlYABBY2b-OE (right) plants at same leaf stage. (h) MLK1 (left) and SlYABBY2b-OE (left) stems

Table 1.

Phenotype of the wild-type and SlYABBY2b-overexpressing transgenic plants

Type	Plant height (cm)			Stem diameter (cm)
	4-leaf stage	8-leaf stage	12-leaf stage	4-leaf stage	8-leaf stage	12-leaf stage
MLK1	5.1±0.2bB	9.6±0.6bB	43.4±3.3bB	0.224±0.017aA	0.395±0.040aA	0.812±0.065aA
SlYABBY2b-OE	4.6±0.4aA	8.3±0.5aA	34.4±2.2aA	0.264±0.038aA	0.474±0.041bB	0.901±0.129bB
Type	The first internodal length (cm)			The second internodal length (cm)
	4-leaf stage	8-leaf stage	12-leaf stage	4-leaf stage	8-leaf stage	12-leaf stage
MLK1	0.6±0.1aA	1.8±0.4bB	2.9±0.3bB	0.6±0.2aA	1.9±0.3bB	2.8±0.3bB
SlYABBY2b-OE	0.6±0.2aA	1.4±0.3aA	2.1±0.2aA	0.6±0.1aA	1.5±0.2aA	2.0±0.3aA
Type	The first flower of the first truss			The second flower of the first truss
	First fruit	Second fruit	Third fruit	First fruit	Second fruit	Third fruit
MLK1	16.8±1.5bB	13.2±1.6bB	12.9±1.7bB	14.2±2.5bB	12.6±2.8bB	14.3±1.9bB
SlYABBY2b-OE	10.9±1.7aA	9.9±1.5aA	10.4±1.4aA	11.4±2.9aA	10.6±1.4aA	12.3±1.9aA
Type	The third flower of the first truss			The first inflorescence
	First fruit	Second fruit	Third fruit	Total flowers	Sepals	Petals
MLK1	14.5±1.7bB	12.1±1.6bB	11.4±1.7aA	5.3±1.5aA	8.6±0.7bB	8.9±0.5bB
SlYABBY2b-OE	11.0±1.5aA	10.0±1.5aA	11.8±1.9aA	12.6±5.4bB	7.8±0.7aA	8.0±0.7aA
Type	The second inflorescence			The third inflorescence
	Total flowers	Sepals	Petals	Total flowers	Sepals	Petals
MLK1	5.7±1.6aA	9.3±0.8bB	9.3±0.6aA	5.7±1.0aA	9.5±0.8bB	9.4±0.7bB
SlYABBY2b-OE	7.2±2.5bB	8.6±0.5aA	9.3±0.3aA	5.6±0.8aA	8.8±0.3aA	9.1±0.8aA

Values are expressed as mean±standard deviation (n=3). Values in the same column followed by different letters are statistically different between different plants: ^a,b P<0.05, ^A,B P<0.01. OE: overexpressing

Fig. 2.

Phenotypic characteristics of representative wild-type and SlYABBY2b RNAi transgenic plants

(a) FL1 locules. (b) SlYABBY2b RNAi locules. (c) FL1 fruit size. (d) SlYABBY2b RNAi fruit size. (e) FL1 (left) and SlYABBY2b RNAi plants (right) at same leaf age. CK: FL1 locules; ZS: SlYABBY2b RNAi locules

Table 2.

Phenotype of the wild-type and SlYABBY2b RNAi transgenic plants

Type	Plant height (cm)				Stem diameter (cm)
	4-leaf stage	8-leaf stage		12-leaf stage	4-leaf stage	8-leaf stage		12-leaf stage
FL1	3.9±0.2aA	7.7±0.4aA		32.2±2.2aA	0.256±0.024aA	0.390±0.057bB		0.839±0.068bB
SlYABBY2b RNAi	4.9±0.3bB	9.8±0.3bB		38.6±2.8bB	0.248±0.016aA	0.368±0.034aA		0.785±0.056aA
Type	The first internodal length (cm)				The second internodal length (cm)
	4-leaf stage	8-leaf stage		12-leaf stage	4-leaf stage	8-leaf stage		12-leaf stage
FL1	0.5±0.1aA	1.0±0.2aA		2.1±0.2aA	0.5±0.1aA	1.1±0.3aA		2.0±0.1aA
SlYABBY2b RNAi	0.6±0.0bB	1.5±0.2bB		2.5±0.2bB	0.6±0.0bB	1.6±0.2bB		2.4±0.3bB
Type	The first flower of the first truss				The second flower of the first truss
	First fruit	Second fruit		Third fruit	First fruit	Second fruit		Third fruit
FL1	2.3±0.5aA	2.1±0.3aA		2.6±0.5aA	2.5±0.5aA	2.2±0.4aA		2.4±0.5aA
SlYABBY2b RNAi	2.9±0.3bB	3.0±0.0bB		3.1±0.3bB	3.4±0.5bB	3.3±0.7bB		3.1±0.7bB
Type	The third flower of the first truss				The first inflorescence
	First fruit	Second fruit		Third fruit	Total flowers	Sepals	Petals	Average fruit weight (g)
FL1	2.3±0.5aA	2.5±0.5aA		2.2±0.4aA	5.4±0.6aA	5.3±0.2aA	5.1±0.1aA	44.5±11.7aA
SlYABBY2b RNAi	3.2±0.6bB	3.2±0.4bB		3.1±0.7bB	6.0±1.1bB	5.4±0.2aA	5.1±0.2aA	54.4±18.4bB
Type	The second inflorescence				The third inflorescence
	Total flowers	Sepals	Petals	Average fruit weight (g)	Total flowers	Sepals	Petals	Average fruit weight (g)
FL1	5.6±0.6aA	4.9±0.2aA	5.0±0.0aA	79.4±16.7aA	5.8±0.5aA	5.1±0.4aA	5.1±0.2aA	81.7±17.8aA
SlYABBY2b RNAi	6.0±1.8bB	5.1±0.2bB	5.3±0.5bB	87.9±23.4bB	6.0±0.9bB	5.6±0.4bB	5.5±0.5bB	82.2±7.5bB

Values are expressed as mean±standard deviation (n=3). Values in the same column followed by different letters are statistically different between different plants: ^a,b P<0.05, ^A,B P<0.01

3.2. Expression levels of SlYABBY2b in transgenic tomato plants

To determine whether the difference in carpel number in our transgenic plants was caused by SlYABBY2b, we analyzed the expression levels of SlYABBY2b in the apical shoots using qRT-PCR. The expression of SlYABBY2b in the apical shoots of the overexpressing plants was significantly higher than that of the non-transgenic plants at every stage. In contrast, the expression of SlYABBY2b was significantly lower in the leaves and apical shoots of the RNAi plants than in those of the control (Fig. 3). In addition, we analyzed the expression of SlYABBY2b in the GA deficient mutants. In the shoot apexes of gib mutants (LA2893 and LA2895) and the control (LA2706), the expression of SlYABBY2b showed a tendency of up-down-up. The expression of SlYABBY2b reached its highest level during the first sampling period, and the lowest was in the second sampling period. As expected, the expression of SlYABBY2b in gib mutants was significantly higher than that in the control, and compared to LA2895, the expression of LA2893 was higher (Fig. 4).

Fig. 3.

qRT-PCR analysis of the expression of SlYABBY2b in the apical shoots of wild-type (WT) and SlYABBY2b transgenic plants

FI, floral bud differentiation initial stage; FD, floral bud differentiation stage; SP, sepal-petal formation stage; CF, carpel formation initial stage; OF, ovary locule complete formation stage; SlYABBY2b-OE, SlYABBY2b-overexpressing. The PCR levels were normalized to those of actin. The data represent the mean±SD of three biological samples. Values followed by the same letter (a or b) are not significantly different (P>0.05) at the same period

Fig. 4.

qRT-PCR analysis of SlYABBY2b expression in the apical shoots of wild-type (WT) and gib mutants

FI, floral bud differentiation initial stage; FD, floral bud differentiation stage; SP, sepal-petal formation stage; CF, carpel formation initial stage; OF, ovary locule complete formation stage; LA2706, moneymaker; LA2893, gib-1 mutants; LA2895, gib-3 mutants. The PCR levels were normalized to those of actin. The data represent the mean±SD of three biological samples. Values followed by the same letter (a or b) are not significantly different (P>0.05) at the same period

3.3. Endogenous GA content analysis in transgenic tomato plants

The content of endogenous GA in the apical shoots of the transgenic plants during development was determined by ELISA. The endogenous GA content in the apical shoots initially decreased and then increased, whereas those in the non-transgenic plants initially increased, decreased, and finally increased. The GA content in the SlYABBY2b-overexpressing lines was markedly reduced compared with that in the non-transgenic lines. In contrast, the endogenous GA content in the apical shoots of the RNAi transgenic plants and non-transgenic plants initially increased and then decreased. The endogenous GA content was also markedly higher in these plants than in the control (Fig. 5).

Fig. 5.

Concentration of endogenous GA in the apical shoots of wild-type and SlYABBY2b transgenic plants

FI, floral bud differentiation initial stage; FD, floral bud differentiation stage; SP, sepal-petal formation stage; CF, carpel formation initial stage; OF, ovary locule complete formation stage; WT, wild type; SlYABBY2b-OE, SlYABBY2b-overexpressing; FW, fresh weight. The PCR levels were normalized to those of actin. The data are shown as the mean of three independent experiments. Bars indicate the standard errors (n=3). Values followed by the same letter (a or b) are not significantly different (P>0.05) at the same period

3.4. Expression levels of GA metabolic genes in transgenic tomato plants

Changes in the GA content in the SlYABBY2b-overexpressing and RNAi lines may be attributed to the altered expression of those genes encoding GA metabolic enzymes. We then examined the gene expression levels of SlCPS, SlGA3ox, SlGA20ox, and SlGA2ox, in the apical shoots of the transgenic and wild-type plants at the floral bud differentiation stages using qRT-PCR. The expression levels of SlGA20ox1 and SlGA3ox2 were lower in the apical shoots of the SlYABBY2b-overexpressing plants than in those of the wild-type plants. In contrast, the transcript levels of SlGA2ox1 and SlGA2ox5 were higher in the transgenic plants than in the control. The expression levels of SlGA20ox1 and SlGA3ox2 were higher in the SlYABBY2b RNAi lines than in the wild-type lines, whereas the transcript levels of SlGA2ox1 and SlGA2ox5 were lower in SlYABBY2b RNAi plants than in the wild-type plants (Figs. 6 and 7).

Fig. 6.

Relative transcript levels of SlGA20ox, SlGA2ox, and SlGA3ox in the apical shoots of MLK1 and SlYABBY2b-overexpressing transgenic plants by qRT-PCR

FI, floral bud differentiation initial stage; FD, floral bud differentiation stage; SP, sepal-petal formation stage; CF, carpel formation initial stage; OF, ovary locule complete formation stage; WT, wild type; SlYABBY2b-OE, SlYABBY2b-overexpressing. The PCR levels were normalized to those of actin. The data are shown as the mean of three independent experiments. Error bars are standard errors (n=3). Values followed by the same letter (a or b) are not significantly different (P>0.05) at the same period

Fig. 7.

Relative transcript levels of SlGA20ox, SlGA2ox, and SlGA3ox in the apical shoots of FL1 and SlYABBY2b RNAi transgenic plants by qRT-PCR

FI, floral bud differentiation initial stage; FD, floral bud differentiation stage; SP, sepal-petal formation stage; CF, carpel formation initial stage; OF, ovary locule complete formation stage. The PCR levels were normalized to those of actin. The data are shown as the mean of three independent experiments. Error bars are standard errors (n=3). Values followed by the same letter (a or b) are not significantly different (P>0.05) at the same period

4. Discussion

In the tomato, the shape and size of fruit are affected by the locule number. The carpels in the flower developed into locules in the fruit. SlYABBY2b, which encodes a YABBY-like transcription factor, regulates fruit development by increasing locule number. High locule number is attributed to the level of SlYABBY2b mRNA accumulation (Cong et al., 2008). SlYABBY2b is a major regulator that can increase the number of locules from two to more than six (Lippman and Tanksley, 2001; Barrero and Tanksley, 2004). In the present study, we produced and characterized transgenic plants overexpressing and silencing the SlYABBY2b gene from tomato. Morphological alterations were observed in our transgenic plants. Overexpression of the SlYABBY2b gene in the multi-locule “MLK1” line decreased locule number (Figs. 1c and 1d). In contrast, RNAi silencing of the SlYABBY2b gene in the “FL1” line, which has few locules, increased locule number (Figs. 2a and 2b). Moreover, the plant height and internode length decreased in the SlYABBY2b-overexpressing plants (Fig. 1g; Table 1). In contrast, the plant height and internode length increased in the SlYABBY2b RNAi plants (Fig. 2e; Table 2). These results indicate that SlYABBY2b not only decreases the number of locules but also inhibits the growth of the tomato.

GA increases tomato ovary locule number (Sawhney and Greyson, 1971; Sawhney and Dabbs, 1978; Liu and Li, 2012). SlYABBY2b, which is the main regulator of tomato ovary locule number, was the first YABBY transcription factor found in tomato (Cong et al., 2008). YABBY transcription factors are closely associated with GA synthase genes in Arabidopsis thaliana and rice (Hay et al., 2002; Kumaran et al., 2002; Dai et al., 2007). In this study, the expression of SlYABBY2b was significantly higher in the apical shoots of the overexpressing plants than in those of the non-transgenic plants at every stage (Fig. 3). In addition, the endogenous GA content in the overexpressing plants was markedly lower than that in the control (Fig. 4). Conversely, the expression of SlYABBY2b was significantly lower in the apical shoots of the RNAi plants than in those of the wild-type plants (Fig. 3). The endogenous GA content was markedly higher in the RNAi plants than in the control (Fig. 5). Therefore, the SlYABBY2b gene plays a negative regulatory role in endogenous GA biosynthesis. As the first cloned YABBY transcription factor in tomato, SlYABBY2b is also involved in regulating GA biosynthesis, which is consistent with the existing findings in A. thaliana and rice (Hay et al., 2002; Kumaran et al., 2002; Dai et al., 2007). Furthermore, the expression of SlYABBY2b gene was markedly higher than that of the control in the apical shoots of GA-deficient mutants, and this showed that the GAs can inhibit the expression of SlYABBY2b gene negative regulation (Fig. 4).

The metabolism of GA has been comprehensively investigated (Sponsel and Hedden, 2010). These enzymes (GA20ox, GA3ox, and GA2ox) play a key role in controlling the bioactive GA levels (Hedden and Phillips, 2000; Sakamoto et al., 2004). The concentrations of active GA were altered in GA20ox, GA3ox, and GA2ox overexpressing or RNAi transgenic plants, which suggests that GA content is regulated by these genes (Hedden and Phillips, 2000). In the present study, the GA content was altered in the SlYABBY2b-overexpressing and RNAi lines (Fig. 4). Thus, we examined the expression levels of the genes encoding four enzymes, SlCPS, SlGA3ox, SlGA20ox, and SlGA2ox (Imai et al., 1996; Rebers et al., 1999; Serrani et al., 2007, 2008), in the apical shoots of the transgenic and wild-type plants at the floral bud differentiation stage. The present data showed that SlYABBY2b suppressed the expression of SlGA20ox1 and SlGA3ox2 but increased the expression of SlGA2ox1 and SlGA2ox5 in the apical shoots of the overexpressing plants (Figs. 6 and 7), thereby reducing GA content. Opposite results were found in the RNAi plants. These results indicate that SlYABBY2b can modulate GA synthesis. The expression levels of SlGA20ox1, SlGA3ox2, SlGA2ox1, and SlGA2ox5 were significant or extremely significant, suggesting that these genes are important in the regulatory effects of SlYABBY2b on GA content.

In conclusion, we studied the regulatory effects of SlYABBY2b on tomato ovary locule number and endogenous GA content using transgenic approaches. We found that SlYABBY2b-overexpressing plants showed fewer locules and lower GA content than the control plants, but SlYABBY2b RNAi plants had more locules and higher GA content than the control plants, which suggests that the SlYABBY2b gene negatively regulates locule number and GA content. Furthermore GAs can also inhibit the expression of the SlYABBY2b gene negative regulation. In addition, the expression levels of SlGA20ox1, SlGA3ox2, SlGA2ox1, and SlGA2ox5 were significant or extremely significant, suggesting that these genes are important in regulatory effects of SlYABBY2b on GA content, consequently affecting GA content and tomato ovary locule number. These results suggest that the SlYABBY2b gene was located on GA signal transduction pathways, fed back regulation of the synthesis of GA, and felt exogenous GA signal to further regulate the formation of the tomato locule.

List of electronic supplementary materials

Table S1

Gibberellin mutants and wild types

Table S2

RT-PCR primers used to amplify gene-specific regions