For the seed germination experiment, seeds from S. lycopersicum (cv. Micro-Tom; used here as the WT), and from the T2 generation transgenic lines (see below for details of generating the transgenic plants), were imbibed on Petri dishes with 0.74% agar and maintained at 28°C for 3 days in darkness, and then maintained at 25°C for 4 days under a photosynthetic photon flux density (PPFD) of 200 μmol m^-2 s^-1. For the phenotype evaluation, seeds were germinated at 28°C in Petri dishes lined with two layers of filter paper moistened with deionized water. The germinated seeds were then sown and plants were also grown in plastic pots (7 cm × 7 cm × 8 cm), filled with 105 g of a mixture of peat and vermiculite (v/v = 7:3) in trays. Seedlings and plants were grown in growth chambers, at a temperature 25/20°C, a PPFD of 500 μmol m^-2 s^-1, and a 12-h photoperiod. Flowers were tagged at the full-bloom stage to synchronize developmental comparisons. Fruits were identified as being at the breaker (B) stage when they displayed the first sign of external color change. Fruit from the WT and transgenic lines took 54 and 45 days, respectively, to develop from anthesis to the B stage. Fruit were considered to be at the mature green (MG) stage 3 days before B, while fruit at 3 days after B were considered to be at the P (pink) stage and those at 6 days after B were at the MR (mature red) stage.

The tomato SlBRI1 gene was PCR-amplified from tomato (S. lycopersicum cv. Micro-Tom) cDNA based on a UniGene sequence (Solyc04g051510), and a Flag peptide encoding sequence was added to the 3′ end. The SlBRI1 cDNA sequence was cloned into the KpnI and XbaI sites of the binary pBI121vector, which carries the kanamycin resistance gene for bacterial and plant selection, and drives transgene expression using the constitutive cauliflower mosaic virus 35S promoter. The construct was transformed into the Agrobacterium tumefaciens strain GV3101 by electroporation, which was then used to transformed tomato cotyledon explants (Park et al., 2003). Three homozygous transgenic tomato lines (BRI1: OX4, BRI1: OX6, and BRI1: OX37), exhibiting high SlBRI1 expression levels, were chosen for subsequent phenotypic and molecular characterization.

The CO₂ assimilation rate (Pn) was determined using a portable photosynthesis system (LI-6400-40, Li-Cor, Lincoln, NE, United States). All measurements were carried out at 400 μmol (CO₂) mol^-1, at 22°C, and 800 μmol m^-2 s^-1 light intensity.

Plant height, expansion diameter of plant, leaf area, length and width of the third leaf were determined at 30 days after sowing and of the sixth leaf at 60 days after sowing. The fresh weight (FW) of the shoot, leaf mass per area, leaf thickness, stigma length, and petal length were measured at 30 days after sowing. The third and sixth leaves were photographed with a digital camera (Canon G15, Japan) at 30 and 60 days after sowing. Leaf areas were calculated using ImageJ (Collins, 2007) software. Leaf thicknesses of WT and transgenic plants were measured using light microscopic observation of paraffin sections, generated as in Yang et al. (2011), with an Olympus microscope (BX53, Japan). The third leaf was selected and at least six sections were analyzed. The growth and fruit characteristics of 15 plants per transgenic line and of the WT genotype were observed.

To measure ethylene production, B, P, and R stage fruit of approximately the same size were harvested and kept in sealed plastic bags for 3 h at room temperature (each biological replicate corresponded to five fruits). One milliliter of gas was removed from each plastic bag with a syringe, and ethylene production was determined using a gas chromatograph, which main specifications were FID detector, DB-130 m, 0.32 μm, 0.25 μm chromatographic column (Agilent 7890B, United States). All samples involved three biological replicates.

RNA from various organs of WT and transgenic plants was extracted using the RNAiso Plus (TaKaRa, Japan) kit, according to the manufacturer’s instructions. Residual DNA was removed with DnaseI (Thermo, United States). One microgram total RNA was reverse-transcribed using a cDNA synthesis Kit (Roche, Switzerland), following the manufacturer’s instructions. The cDNA concentrations were normalized to the expression levels of the tomato ubiquitin gene (X58253.1) for subsequent RT-PCR analysis. The qPCR experiments were performed using a Real-Time PCR system (Bio-Rad) with SYBR Green Master Mix (Vazyme, Nanjing, China). Relative gene expressions were calculated using the 2^-ΔΔCT method (Livak and Schmittgen, 2001) and the tomato UBI3 gene was used as the internal control. The experiment had three independent biological samples per line and each sample had four technical replications. Gene-specific qPCR primers are listed in Supplementary Table S2. The WT and transgenic plants were represented by four replicates for each sample.

Carotenoids were extracted and analyzed as previously described (Liu et al., 2014) with modifications. Tomato pericarp (0.5 g) was ground to a fine powder in liquid nitrogen and extracted with 15 mL of hexane/acetone/ethanol (1:1:1) solution containing 0.01% butylated hydroxytoluene (BHT) on a magnetic stirrer for 30 min. The extracts were centrifuged at 4000 × g at 4°C for 10 min, and the supernatant transferred to 50 mL tubes. Fifteen milliliter of hexane/acetone/ethanol was added to the sediment and the extraction process was repeated. The supernatants from the two extractions were merged, before hexane was added up to a total volume of 25 mL. The absorbance of 5 mL of the mixture was measured at 450 nm (UV-1750; Shimadzu Corporation) and this value used to represent total carotenoid levels. The remaining mixture was frozen and evaporated using a vacuum evaporator (RE-52AA). The residue was dissolved in 1 mL ethyl acetate and 20 μL were injected onto a high pressure liquid chromatograph (HPLC, Shimadzu [LC 2010A, Japan]) using an autosampler. All the procedures were performed under dim light. A Nova-Pak C18 column (4.6 mm × 150 mm) and ultraviolet detector were used with a mobile phase of methanol/acetonitrile (45/55 by volume) at a flow rate of 1 mL/min. The total retention time was 25 min. Absorbance was detected at 450 nm. Standards (lutein, lycopene, and β-carotene; Sigma, United States) were prepared and used to identify and quantify the corresponding carotenoids. The carotenoid content was expressed as μg/g FW of tomato fruit.

Proteins was extracted from young leaves (0.2 g) of 25-day-old transgenic and WT plants with 2 × SDS gel loading buffer, as in Wang et al. (2016).

To measure fruit firmness, R stage fruits of the approximately same size and location were harvested (each biological replicate consisted of five fruits) and firmness was determined with a hand-held penetrometer (FHM-1, Japan). Data values are the means ± SD of six independent biological samples.

Soluble solid contents (^oBRIX) of total fruit juice from freshly harvested R stage fruits were determined with a hand-held refractometer (Chengdu Optical Instrument Factory, Chengdu, China). Data values are the means ± SD of six independent biological samples.

The soluble sugar contents were measured as previously described (Liu et al., 2014). The experiment had three independent biological samples per line and each sample had four technical replications.

Ascorbic acid contents were determined as previously described with minor modifications (Hu et al., 2016). Frozen pericarp (1 g) of R stage fruits was ground to a fine powder in liquid nitrogen and extracted with 5 mL of 6% (w/v) trichloroacetic acid (TCA). The homogenates were centrifuged at 4000 × g for 15 min at 4°C. The supernatant (0.4 mL) was transferred to a 10 mL tube containing 0.4 mL of 5 mM DL-dithiothreitol (DTT) in 0.4 M potassium phosphate buffer, pH 7.4. The 10 mL tube was incubated for 30 min at 37°C, and 0.2 ml 0.5% (w/v) N-ethylmaleimide (NEM) added, followed by incubation for 1 min at room temperature. Color reagent (1.6 mL) was then added to the mixture, which was incubated for 1 h at 37°C. The absorbance was recorded at 550 nm (Shimadzu UV-1750). The color reagent was prepared as follows: solution A, 31% orthophosphoric acid, 4.6% (w/v) TCA, and 0.6% (w/v) iron chloride (FeCl₃); solution B, 4% 2, 2-dipyridyl (w/v in 70% ethanol). Solutions A and B were mixed in a ratio of 2.75:1 before use. The experiment had three independent biological samples per line.

The data were analyzed using SPSS version 17.0 and a Student’s t-test. Mean and standard error values were calculated for the variables comparison. Values of P < 0.05 and P < 0.01 were considered statistically significant.

SlBRI1 expression in different organs was determined using RT-PCR, and we observed that expression was very low in roots, higher in flowers and red fruits, and highest in stems and leaves (Figure 1A), suggesting that SlBRI1 is involved in regulating both vegetative and reproductive growth.

We generated transgenic tomato plants expressing SLBRI1 under the control of the constitutive CaMV 35S promoter. Transcript levels of SlBRI1 in the three transgenic lines BRI1:OX4, BRI1:OX6, and BRI1:OX37 were 6.0, 11.6, and 23.4 times higher, respectively, than in WT plants (Figure 1C). The expression levels of the BR biosynthetic gene, DWARF, in these transgenic lines were 3.7, 2.9, and 3.5 times lower, respectively, than in WT plants for DWARF (Figure 1D), while those of another BR biosynthetic gene, CPD, were 1.52, 1.23, and 1.51 times lower, respectively (Figure 1E). Similarly, Western blot analysis indicated that SLBRI1 protein levels in the BRI1:OX4 and BRI1:OX6 lines were lower than in the BRI1:OX37 line (Figure 1B), while the SLBRI1-Flag fusion protein was, as expected, not detected in the WT plants. Taken together, these results indicate that overexpression of SLBRI1 resulted in increased transcription and expression of SLBRI1 and caused a feedback-inhibition of BR biosynthetic genes, which consequently showed decreased expression. This is consistent with an increase in the intensity of endogenous BR signaling in the SLBRI1 overexpressing transgenic lines (He et al., 2005).

To test whether the SLBRI1 overexpression mediated increase in BR signaling affected seed germination seeds of WT, BRI1:OX4, BRI1:OX6, and BRI1:OX37 were germinated on agar and were monitored for 7 dyas. BRI1:OX4, BRI1:OX6, and BRI1:OX37 seeds germinated earlier than WT seeds (Figure 2A). Additionally, radicle lengths increased in proportion to the increase in SLBRI1 expression, such that the radicle lengths in the BRI1:OX4, BRI1:OX6, and BRI1:OX37 lines were 1.9-, 2.1-, and 2.7-fold greater, respectively, than in WT, 3 days after germination (Figure 2C). Compared with WT, the transgenic lines showed increased root and hypocotyl length, number of lateral roots and seedling weight after 7 days of germination (Figures 2B–G), highlighting the critical role of BR signaling in seed germination and early growth vigor.

To elucidate the effects of the SLBRI1 overexpression mediated increase in BR signaling on the growth of tomato plants, we observed the phenotypes of WT, BRI1:OX4, BRI1:OX6, and BRI1:OX37 plants grown in a growth chamber. The transgenic plants exhibited increased plant height, expansion diameter, internode length, and hypocotyl length 30 days after sowing (Figures 3A,C,E,G,H), and also 60 days after sowing in the case of the BRI1:OX37 line (Figures 3B,D,F). The transgenic plants had a larger leaf area than WT at the early growth stage (Figures 4A,C), and also longer leaves throughout development. The transgenic plants also exhibited a reduced leaf thickness and leaf mass per area compared with WT (Figure 4), as well as larger flowers with longer petals and style (Figure 5). Finally, gas exchange analyses indicated that the SLBRI1-overexpressing plants had an increased CO₂ assimilation rate and FW 30 days after sowing (Figures 3I,J). Taken together, these results indicated that the increase in BR signaling due to SLBRI1 overexpression promoted various aspects of growth and development, as well as photosynthetic capacity.

To test whether SLBRI1 expression level is important for the regulation of ethylene production, we measured ethylene accumulation in the fruits of the transgenic lines and WT by gas chromatography. Ethylene production in BRI1:OX4, BRI1:OX6, and BRI1:OX37 were 65.1, 40.5, and 62.7% higher, respectively, than in WT at the B stage (Figure 6A). Ethylene production in the transgenic lines was still higher than in WT up to MR stage. We also measured the transcript abundance of the ethylene biosynthetic genes ACO1, ACS2, and ACS4, which are important for ethylene production in tomato fruit (Barry et al., 1996; Barry et al., 2000; Klee and Giovannoni, 2011). Compared with WT fruit, ACO1 and ACS2 transcript levels in the fruits of the three transgenic lines were markedly higher than in WT at the B and P stages (Figures 6B,C), with ACO1 expression being 69-, 50-, and 62-fold higher at the B stage. Similarly, ACS4 transcript levels were higher in the transgenic lines than in WT at the B and P stages (Figure 6D). These results showed that SLBRI1 overexpression mediated increases in BR signaling promote ethylene production during tomato fruit ripening by upregulating the expression of ethylene biosynthetic genes.

To test whether the SLBRI1 overexpression mediated increase in BR signaling affects carotenoid accumulation, total carotenoid levels were determined by spectrophotometry. We observed that total carotenoid accumulation was significantly greater in the SLBRI1-overexpressing lines (Figure 7A). Furthermore, when lycopene, β-carotene, and lutein accumulation in the pericarp of fruit at the MR stage from the transgenic lines and WT was measured, using HPLC, the SLBRI1 transgenic lines showed increased levels of both lycopene and lutein. The lycopene abundance was 120, 84, and 79% greater in the pericarp of BRI1:OX4, BRI1:OX6, and BRI1:OX37, respectively, than in WT (Figure 7B), while the lutein levels were 22, 140, and 182% greater, respectively (Figure 7C). We observed no significant difference in β-carotene levels between the transgenic lines and WT (Figure 7D). Taken together, these results indicated that carotenoid accumulation in tomato fruits correlated with the BR signaling strength in the SLBRI1 overexpression lines.

To investigate whether changes in the expression of genes in the carotenoid biosynthetic pathway correlated with changes in carotenoid content, we used qRT-PCR to determine the transcript levels of SlBRI1 (Figure 8A), SlDXS (Figure 8B), SlGGPS (Figure 8C), SlPSY1 (Figure 8D), and SlCYC-B (Figure 8E) in the pericarp of fruit at four stages. SlBRI1 expression in the three transgenic lines was significantly greater at all four stages, with SlBRI1 transcript abundance in BRI1:OX37 being 18-, 109-, 119-, and 159-fold higher than in WT at the MG, B, P, and MR stages, respectively (Figure 8A). Lycopene biosynthesis related genes (SlDXS, SlGGPS, and SlPSY1) were also all markedly up-regulated in the three transgenic lines at the B, P, and MR stages, with the exception of SlGGPS in the BRI1:OX4 and BRI1:OX6 lines at the R stage (Figures 8B–D). Additionally, SlDXS expression was also significantly up-regulated at the MG stage (Figure 8B). SlCYC-B encodes an enzyme that catalyzes the conversion of lycopene to β-carotene (Hirschberg, 2001), and its expression in the transgenic lines was higher in the B and P stages compared to WT, but was significantly lower at the MR stage (Figure 8E). These results are consistent with previous reports (Römer and Fraser, 2005; Dellapenna and Pogson, 2006), that the expression of genes encoding enzymes involved in carotenoid biosynthesis determining the total levels of carotenoids, and that SlBRI1 overexpression cause an increase in BR signaling, which is important for carotenoid accumulation.

We also investigated whether the SlBRI1 expression level affected fruit quality and yield. The transgenic lines exhibited a slight increase in fruit yield and number per plant compared with WT, but a decrease in individual fruit weight. SlBRI1 overexpression resulted in early flowering and a significant reduction in the ripening time (Supplementary Table S1). Levels of ascorbic acid in the fruit of the BRI1:OX4, BRI1:OX6, and BRI1:OX37 lines were 17, 17, and 59% higher, respectively, than in WT (Figure 9A), and the increase in total soluble solids was 27, 22, and 29%, respectively (Figure 9B). Furthermore, the contents of soluble sugars in fruit from the BRI1:OX4, BRI1:OX6, and BRI1:OX37 lines were 21, 30, and 22% higher than in WT (Figure 9C). The firmness of fruit from the BRI1:OX37 line was greater than WT, while no difference was observed between the BRI1:OX4 and BRI1:OX6 lines compared to WT (Figure 9D).