Chloroplastic thioredoxin-f and thioredoxin-m1/4 play important roles in brassinosteroids-induced changes in CO2 assimilation and cellular redox homeostasis in tomato

Summary Virus-induced gene silencing (VIGS) was used in this study to characterize the role of thioredoxin-f and thioredoxin-m1/4 in brassinosteroid-induced changes in CO2 assimilation and cellular redox homeostasis in tomato.


Introduction
Accumulating evidence supports the hypothesis that the cellular redox signalling and hormone signalling pathways form an integrated redox-hormone network that regulates many plant growth-and defence-related pathways (Bartoli et al., 2013). Redox regulation is often mediated by thioredoxins (TRXs), which are able to (de-)activate enzymes through the reversible reduction of disulphide bonds (Buchanan and Balmer, 2005;Montrichard et al., 2009).
There are five types of typical TRXs present in the chloroplasts. The f-and m-type TRXs function as messengers in the Fdx (ferredoxin)/TRX system by transmitting the redox signal from Fdx:TRX reductase to the target enzymes. TRXf has been shown to activate several chloroplast enzymes, including enzymes involved in the Benson-Calvin cycle, ATP synthesis, and fatty acid synthesis (Nishizawa and Buchanan, 1981;Wolosiuk et al., 1993;Sasaki et al., 1997;Schwarz et al., 1997;Zhang and Portis, 1999). TRX-m has been found to be involved in the inactivation of glucose-6-phosphate dehydrogenase, which is the first enzyme in the oxidative pentose phosphate cycle (Wenderoth et al., 1997). Both TRX-f and TRX-m can activate fructose-1,6-bisphosphatase (FBPase) and NADP-dependent malate dehydrogenase (NADP-MDH) in vitro, suggesting that there is some overlap in substrate specificity between the isoforms (Hodges et al., 1994;Geck et al., 1996). TRX-x, which is inactive toward FBPase and NADP-MDH, is the most efficient reductant of 2-Cys peroxiredoxin (2-CP). This property suggests that TRX-x functions specifically in resistance to oxidative stress . TRX-y exists in two isoforms and has been reported to act as an efficient electron donor for the chloroplast glutathione peroxidase (Navrot et al., 2006) and methionine sulphoxide reductase B2 (Vieira Dos Santos et al., 2007). Peroxiredoxin Q, which has been reported to function as an antioxidant (Rouhier et al., 2004;Lamkemeyer et al., 2006), was shown to be the best substrate of TRX-y (Collin et al., 2004). Lastly, TRX-z regulates plastid-encoded RNA polymerase-dependent transcription (Arsova et al., 2010). However, the biological functions of TRXs have been studied mostly in Arabidopsis through in vitro experiments, and detailed in vivo studies using different isoforms have not yet been conducted.
Brassinosteroids (BRs) are a group of plant steroid hormones that have been shown to play important roles in the growth, development, and stress response of plants (Clouse and Sasse, 1998;Müssig, 2005). For example, the tomato d ^im mutant with the block in C-6 oxidation in BRs biosynthesis results in a dwarf phenotype (Bishop et al., 1999). Previous studies showed that BRs can enhance stress tolerance against chilling, paraquat (PQ), and biotic stress, and this effect is largely dependent on BRs-induced hydrogen peroxide (H 2 O 2 ) accumulation in the apoplast (Xia et al., 2009b). Furthermore, BRs are involved in the regulation of CO 2 assimilation in several plants. In cucumber, BRs induce transient increases in RBOH1 (respiratory burst oxidase homolog 1) expression, NADPH oxidase activity, H 2 O 2 levels in the apoplast, and nitric oxide (NO) production (Xia et al., 2009b;Cui et al., 2011). Additionally, BRs can induce the expression and activity of genes and enzymes involved in the Benson-Calvin cycle, and this effect is attributed to the apoplastic H 2 O 2induced activation of photosynthesis-related redox-sensitive enzymes (Jiang et al., 2012a). Apoplastic H 2 O 2 can induce an increase in the ratio of reduced glutathione (GSH) to oxidized glutathione (GSSG), resulting in increased stability of redox-sensitive enzymes (Jiang et al., 2012b). Microarray analysis revealed that several TRX genes are targeted by BRs in Arabidopsis and rice (Müssig et al., 2002;Goda et al., 2004;Wu et al., 2008); however, their potential roles in BRsinduced cellular redox homeostasis and CO 2 assimilation remain unknown.
Tomato is an important horticultural crop distributed world-wide and has been widely used to study the stress response. Virus-induced gene silencing (VIGS) has been well established and widely used for the analysis of gene functions in Solanaceous species (Liu et al., 2002a, b;Li et al., 2006). In this study, the roles of the different chloroplastic TRXs in the BRs-induced changes in cellular redox homeostasis and CO 2 assimilation were examined. Accordingly, five chloroplast TRXs were partially silenced individually using a VIGS approach, and the changes in CO 2 assimilation as well as the activities of Calvin cycle and ascorbate (AsA)-GSH cycle enzymes and cellular glutathione redox homeostasis were subsequently determined in the presence or absence of exogenous BRs.

Materials and methods
Plant materials and experimental design Experiment I Tomato seeds from wild-type (Solanum lycopersicum L. cv. Condine Red, CR) and its partially BRs-deficient d^i m mutant were obtained from the Tomato Genetics Resource Center (University of California, Davis, CA, USA). These seeds were germinated and grown in a mixture of peat and vermiculite (1:1, v:v) under a 16 h light (200 μmol m -2 s -1 ; at 25 °C), 8 h dark (at 20 °C) cycle. To determine the role of BRs in the regulation of CO 2 assimilation and the effects of BRs on the transcript levels of the chloroplastic TRX genes, plants at the four-leaf stage were sprayed with water or 0.2 μM EBR (24-epibrassinolide; Sigma, Santa Clara, CA, USA), one of the most active and stable forms of BRs. Twenty-four hours later, CO 2 assimilation was measured in the third leaf from the bottom. Additionally, leaf samples were harvested from the wateror EBR-treated CR and d^i m plants, frozen immediately in liquid nitrogen, and stored at -80 °C prior to gene expression analyses. Experiment II Tomato (CR) seeds were germinated and grown in a mixture of peat and vermiculite (1:1, v:v) under a 16 h light (200 μmol m -2 s -1 ; at 25 °C), 8 h dark (at 20 °C) cycle. To examine the roles of the chloroplast TRXs in the regulation of CO 2 assimilation and to determine whether they are involved in the BRs-induced increase in CO 2 assimilation, VIGS was performed when the cotyledonary leaves were fully expanded but the true leaves had not yet appeared. Partially gene-silenced plants at the six-leaf stage were treated with 0.2 μM EBR via foliar spraying on all of the leaves; 10 ml of this solution was applied per plant, and distilled water was used as a control. After ~24 h, CO 2 gas exchange was measured, and leaf samples were harvested for biochemical and gene expression analyses.
For virus infiltration, a mixed culture of A. tumefaciens carrying the pTRV1:pTRV2-target gene in a 1:1 ratio was infiltrated into fully expanded cotyledonary leaves of tomato plants (Ekengren et al., 2003). An Agrobacterium culture carrying the empty pTRV2 vector was also infiltrated into a set of plants, which were used as a control. The inoculated plants were maintained at 20-22 °C in a growth chamber with a 16 h daylength. After ~4 weeks, quantitative real time-PCR (qRT-PCR) was performed to determine the gene silencing efficiency before the plants were used in assays (Supplementary Fig. S1 available at JXB online).

Leaf gas exchange measurements
Gas exchange analysis was conducted on the third leaf for Experiment I and on the fifth leaf for Experiment II using an open gas exchange system (LI-6400; LI-COR, Lincoln, NE, USA). The light-saturated rate of CO 2 assimilation (A sat ) was measured under an ambient CO 2 concentration of 380 μmol mol -1 at a saturating photosynthetic photon flux density (1000 μmol m -2 s -1 ), at a leaf temperature of 25 ± 1.5 °C and a relative air humidity of 80-90%. An assimilation versus intercellular CO 2 concentration (A/C i ) curve was determined according to von Caemmerer and Farquhar (1981). Assimilation was first measured at the ambient CO 2 concentration under which the plants had grown. The atmospheric CO 2 concentration was decreased to 50 μmol mol -1 in a stepwise manner and then returned to the growth concentration to check that the original rate could be regained; the rate was then finally increased stepwise to 2000 μmol mol -1 to complete the response curve. The maximum ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) carboxylation rates (V c,max ) and maximum ribulose-1,5-bisphosphate (RuBP) regeneration rates (J max ) were estimated from the A/C i curves using the method described by Ethier and Livingston (2004).

Determination of lipid peroxidation
Lipid peroxidation was estimated by measuring malondialdehyde (MDA) equivalents, according to the method described by Hodges et al. (1999). Leaf samples (0.3 g) were ground in 3 ml of ice-cold 25 mM HEPES buffer (pH 7.8) containing 0.2 mM EDTA and 2% polyvinylpyrrolidone (PVP). The obtained homogenates were centrifuged at 4 °C for 20 min at 12000 g, and the resulting supernatants were used for analysis of MDA equivalents. The samples were mixed with 10% trichloroacetic acid (TCA) containing 0.65% 2-thiobarbituric acid (TBA) and heated at 95 °C for 25 min. The content of MDA equivalents was corrected for non-MDA compounds by subtracting the absorbance at 532 nm of a TBA-less solution containing the plant extract.

Non-reducing SDS-PAGE and western blot analysis of 2-Cys peroxiredoxin
Total proteins were isolated from leaf tissues in a protein extraction buffer (100 mM HEPES, pH 7.5, 5 mM EDTA, 5 mM EGTA, 10 mM Na 3 VO 4 , 10 mM NaF, 50 mM β-glycerophosphate, 1 mM phenylmethylsulphonyl fluoride, 10% glycerol, and 7.5% polyvinylpolypyrrolidone) supplemented with 10 mM N-ethylmaleimide (NEM) (thiol-blocking reagent). After centrifugation at 13 000 g for 20 min, the supernatants were transferred to clean tubes, quickly frozen in liquid nitrogen, and stored at -80 °C. The protein concentration in the extracts was determined using the Bio-Rad protein assay kit (Bio-Rad, Hercules, CA, USA), and bovine serum albumin (BSA) was used as a standard. Reducing agents and boiling were omitted in this protocol, as it is important to maintain the remaining disulphide bonds (Muthuramalingam et al., 2010). Total protein samples (15 μg) supplemented with 5× loading buffer [225 mM TRIS-HCl, pH 6.8, 5 % (w/v) SDS, 50% glycerol, 0.05 % bromophenol blue] were separated via 12% SDS-PAGE, and the redox state of 2-CP was detected through western blot analysis with a polyclonal antibody against 2-CP (Beijing Protein Innovation, Beijing, China). After incubation with a horseradish peroxidase (HRP)-linked secondary antibody (Cell Signaling Technology, Boston, MA, USA), the complexes on the blot were visualized using an enhanced chemiluminescence kit (Perkin Elmer, Wellesley, MA, USA), according to the manufacturer's instructions. Band intensity was quantified using Quantity One software.
Determination of Rubisco, Rubisco activase (RCA), and FBPase activity Rubisco activity was measured spectrophotometrically by coupling 3-phosphoglyceric acid formation with NADH oxidation at 25 °C according to the method described by Lilley and Walker (1974), with some modifications. Total activity was assayed after the crude extract had been activated in a 0.1 ml activation mixture containing 33 mM TRIS-HCl (pH 7.5), 0.67 mM EDTA, 33 mM MgCl 2 , and 10 mM NaHCO 3 for 15 min. The initial measurements of Rubisco activity were carried out in 0.1 ml of reaction medium containing 5 mM HEPES-NaOH (pH 8.0), 1 mM NaHCO 3 , 2 mM MgCl 2 , 0.25 mM dithiothreitol (DTT), 0.1 mM EDTA, 1 U of glyceraldehyde 3-phosphate dehydrogenase (GAPDH), 0.5 mM ATP, 0.015 mM NADH 2 , 0.5 mM phosphocreatine, 0.06 mM RuBP, and 10 μl of extract. The change in absorbance at 340 nm was monitored for 90 s. RCA activity was determined using a Rubisco Activase Assay Kit (Genmed Scientifics, Washington, DC, USA). FBPase activity was determined by monitoring the increase in A 340 using an extinction coefficient of 6.2 mM -1 cm -1 (Scheibe et al., 1986). Total activity was assayed after the crude extract had been activated in a 0.1 ml activation mixture containing 100 mM DTT, 2 mM FBP, 10 mM MgCl 2 , and 0.1 M TRIS-HCl (pH 8.0). The initial activity was assayed immediately after homogenization. The assay mixture consisted of 0.1 M HEPES-NaOH (pH 8.0), 0.5 mM Na 2 EDTA, 10 mM MgCl 2 , 0.3 mM NADP + , 0.6 mM fructose-1,6-bisphosphate, 0.6 U of glucose-6-phosphate dehydrogenase from baker's yeast (Sigma, Santa Clara, CA, USA), 1.2 U of glucoe phosphate isomerase from baker's yeast (Sigma, Santa Clara, CA, USA), and 100 μl of enzyme extract in a final volume of 1 ml.

Measurements of glutathione contents and the activity of enzymes involved in the AsA-GSH cycle
For the measurement of reduced glutathione (GSH) and oxidized glutathione (GSSG), plant leaf tissue (0.3 g) was homogenized in 2 ml of 6% metaphosphoric acid containing 2 mM EDTA and centrifuged at 4 °C for 10 min at 12 000 g. After neutralization with 0.5 M phosphate buffer (pH 7.5), 0.1 ml of the supernatant was added to a reaction mixture containing 0.2 mM NADPH, 100 mM phosphate buffer (pH 7.5), 5 mM EDTA, and 0.6 mM 5,5'-dithio-bis (2-nitrobenzoic acid). The reaction was initiated by adding 3 U of glutathione reductase (GR) and was monitored by measuring the changes in absorbance at 412 nm for 1 min. For the GSSG assay, GSH was masked by the addition of 40 μl of 2-vinylpyridine to the neutralized supernatant, whereas 40 μl of water was added for the total glutathione assay. The GSH concentration was obtained by subtracting the GSSG concentration from the total concentration (Rao and Ormrod, 1995).
To determine the enzymatic activities of proteins involved in the AsA-GSH cycle, leaf tissue (0.3 g) was ground in 3 ml of ice-cold buffer containing 25 mM HEPES (pH 7.8), 0.2 mM EDTA, 2 mM ascorbic acid, and 2% PVP. The homogenates were centrifuged at 4 °C for 20 min at 12 000 g, and the resulting supernatants were used to determine the enzymatic activity. The ascorbate peroxidase (APX) and dehydroascorbate reductase (DHAR) activities were evaluated by measuring the decrease in absorbance at 290 nm and the increase in absorbance at 265 nm, as described by Nakano and Asada (1981). Monodehydroascorbate reductase (MDAR) activity was measured using 1 U of ascorbate oxidase, and the oxidation rate of NADH was followed at 340 nm (Hossain et al., 1984). GR activity was measured according to the method reported by Halliwell and Foyer (1976), which is based on the rate at which the absorbance of NADPH decreases at 340 nm. All spectrophotometric analyses were conducted in a SHIMADZU UV-2410PC spectrophotometer (Shimadzu Corporation, Kyodo, Japan).

Total RNA extraction and gene expression analysis
Total RNA was isolated from tomato leaves using the TRIZOL reagent (Sangon, Shanghai, China) according to the instructions supplied by the manufacturer. After extraction, the total RNA was dissolved in diethyl pyrocarbonate-treated water. The cDNA template for qRT-PCR was synthesized from 2 μg of total RNA using the ReverTra Ace qPCR RT Kit (Toyobo, Osaka, Japan).
For qRT-PCR analysis, PCR products were amplified in triplicate using iQ SYBR Green SuperMix (Bio-Rad, Hercules, CA, USA) in 25 μl qRT-PCR assays. PCR was performed using the iCycler iQ 96-well real-time PCR Detection System (Bio-Rad, Hercules, CA, USA), and the cycling conditions consisted of denaturation at 95 °C for 3 min, followed by 40 cycles of denaturation at 95 °C for 30 s, annealing at 58 °C for 30 s, and extension at 72 °C for 30 s. The tomato actin gene was used as an internal control. Gene-specific primers were designed according to expressed sequence tag (EST) sequences and were employed for amplification as described in Supplementary  Table S2 available at JXB online. Relative gene expression was calculated as described by Livak and Schmittgen (2001).

Statistical analysis
The experimental design was a completely randomized block design with four replicates. Each replicate contained 10 plants. Statistical analysis of the bioassays was performed using the SAS statistical package. The differences between the treatment means were separated using Tukey's test at a level of P<0.05.

BRs-induced changes in chloroplast TRX transcripts in tomato
A database (Tomato Genome Sequencing Project) search based on sequence similarity with the predicted chloroplast TRXs of Arabidopsis indicated five chloroplast TRX nucleotide sequences in Solanum lycopersicum: TRX-f (Solyc05g056300), TRX-m2 (Solyc10g006970), TRX-m1/4 (Solyc12g013810), TRX-x (Solyc01g008250), and TRX-y (Solyc04g071560). A phylogenetic tree built from the alignment of these five proteins with the previously identified Arabidopsis TRXs revealed the evolutionary distances between the sequences (Fig. 1). Among these sequences, SlTRX-f, -x, and -y showed 100% similarity to the sequences of AtTRX-f, -x, and -y, while SlTRX-m2 and -m1/4 showed low similarity to their orthologues in Arabidopsis. To examine how the transcription of these TRX genes is influenced by BRs, their transcript levels were determined following EBR treatment in CR and d ^im plants. Transcript levels for TRX-f, TRX-m2, TRX-m1/4, and TRX-x were reduced by 30-40% in d ^im plants compared with CR plants, respectively, with the exception of TRX-y (Fig. 2a). In contrast, TRX-f, TRX-m2, TRX-m1/4, and TRX-x transcript levels were upregulated ~1-fold following treatment with EBR in CR plants. Similarly, no significant change was observed in TRX-y transcript levels in the EBRtreated CR plants. Interestingly, the expression of those TRX genes was restored to a level similar to that in the water-treated CR plants when d ^im plants were treated with EBR (Fig. 2a).

The role of chloroplast TRXs in BRs-induced CO 2 assimilation
BRs levels are closely related to CO 2 assimilation in cucumber plants (Yu et al., 2004;Xia et al., 2009a). Here, it was found that the light-saturated rate of CO 2 assimilation (A sat ) was reduced by 24.7% in BRs-deficient d ^im mutant plants compared with wild-type CR plants. However, exogenous application of 0.2 μM EBR increased the A sat value in CR plants 24 h after EBR application. Furthermore, exogenous EBR increased the A sat of the d ^im plants to the level of untreated CR plants (Fig. 2b).
To determine the role of these TRXs in BRs-induced CO 2 assimilation, the five TRX genes were partially silenced individually using a VIGS method. Transcript analysis of the leaflets in the middle of the fifth fully expanded leaves revealed that the transcripts for these genes were reduced by 65-85% in the respective silenced plants ( Supplementary Fig. S1 available at JXB online). To investigate the underlying molecular mechanisms of BRs-induced CO 2 assimilation, the effect of EBR on CO 2 assimilation were analysed in partially TRXf-silenced (pTRV-TRX-f,), TRX-m2-silenced (pTRV-TRX-m2), TRX-m1/4-silenced (pTRV-TRX-m1/4), TRX-x-silenced (pTRV-TRX-x), and TRX-y-silenced (pTRV-TRX-y) plants. As shown in Fig. 3a, partially silencing the TRX-f, TRX-m2, TRX-m1/4, and TRX-y genes resulted in decreases in the A sat of 18.9, 26.0, 20.5, and 22.8%, respectively, compared with the pTRV control plants. However, partially silencing TRX-x did not result in significant changes in A sat . The application of 0.2 μM EBR increased A sat in the pTRV, pTRV-TRX-m2, and pTRV-TRX-y plants by 48.0, 30.9, and 31.6%, respectively, but had little effect on the pTRV-TRX-f, pTRV-TRX-m1/4, and pTRV-TRX-x plants. Similar to the observed changes in A sat , the V c,max and the J max values were significantly decreased in the pTRV-TRX-f, pTRV-TRX-m2, pTRV-TRX-m1/4, and pTRV-TRX-y plants but were not changed in the pTRV-TRXx plants (Fig. 3b, c). Accordingly, exogenous application of EBR increased V c,max and J max only in the pTRV, pTRV-TRX-m2, and pTRV-TRX-y plants and not in the pTRV-TRX-f, pTRV-TRX-m1/4, and pTRV-TRX-x plants. All of these results indicated that chloroplastic TRX-f, TRX-m2, TRX-m1/4, and TRX-y are involved in the regulation of CO 2 assimilation, whereas only TRX-f and TRX-m1/4 play a role in the EBR-induced increase in CO 2 assimilation.

Involvement of glutathione redox homeostasis in BRs-induced CO 2 assimilation
TRX can regulate the activity of the AsA-GSH cycle, which plays a critical role in maintaining the cellular redox status.
In this study, the changes in the activities of APX, MDAR, DHAR, and GR, which are four important enzymes involved in the AsA-GSH cycle, were analysed. As shown in Fig. 5a, partial silencing of TRX-f, TRX-m2, TRX-m1/4, and TRXy decreased the activity of APX, MDAR, DHAR, and GR. However, no significant effects on the activities of these enzymes were observed in the pTRV-TRX-x plants. EBR treatment significantly increased the activity of the AsA-GSH cycle in the pTRV control, pTRV-TRX-m2, and pTRV-TRXy plants, but had no effect in the pTRV-TRX-f, pTRV-TRX-m1/4, and pTRV-TRX-x plants. To analyse the EBR-induced changes in glutathione redox homeostasis further, the effects of EBR on the contents of GSH and GSSG and their ratios in the pTRV and various pTRV-TRX plants were compared. There was generally little change in GSH content observed in the pTRV and partially TRX-silenced plants (Fig. 5b). However, partial silencing of TRX-f, TRX-m2, TRX-m1/4, and TRX-y resulted in 31.5, 27.5, 30.2, and 30.1% increases in the GSSG content, respectively, leading to a decrease in the GSH/GSSG ratio. In contrast, EBR induced a slight increase in GSH content and decrease in GSSG, resulting in a significant increase in the GSH/GSSG ratio. Similar to the observed changes in the activities of the AsA-GSH cycle enzymes, the  EBR-induced increases in the glutathione pool and GSH/ GSSG ratio were abolished by partially silencing TRX-f and TRX-m1/4 but not by the silencing of other TRXs (Fig. 5a, b).

Gene expression and enzyme activities involved in the Benson-Calvin cycle influenced by the TRXs and BRs
Previously, it was reported that BRs-induced CO 2 assimilation is associated with increased expression of photosynthetic genes in cucumber plants (Yu et al., 2004;Xia et al., 2009a). In this study, the transcript levels of eight Benson-Calvin cycle-related genes in the leaves of plants lacking different chloroplastic TRX genes were analysed. These tested photosynthesis-related genes included the genes encoding Rubisco activase (RCA), Rubisco large subunit (rbcL), Rubisco small subunit (rbcS), glycerate-3-phosphate kinase (PGK), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), fructose-1,6-bisphosphatase (FBPase), sedoheptulose-1,7-bisphosphatase (SBPase), and ribulose-5-phosphate kinase (PRK). As shown in Fig. 6, the transcripts of these genes were all up-regulated upon EBR treatment in the pTRV plants. In contrast, partially silencing the chloroplastic TRX genes resulted in down-regulation of these photosynthesis-related genes. Importantly, EBR-induced transcripts were detected only in the pTRV-TRX-m2 and pTRV-TRX-y plants and not in the other plants.
BRs levels are closely related to the activation of a subset of enzymes involved in the Benson-Calvin cycle, particularly the redox-sensitive enzymes (Yu et al., 2004;Xia et al., 2009a). Here, the changes in the activities of RCA, Rubisco, and FBPase were analysed in VIGS plants with and without exogenous application of EBR. As shown in Fig. 7, partial silencing of these genes did not result in significant changes   in total Rubisco activity but significantly decreased the initial Rubisco, initial FBPase, and Rubisco activase activities as well as the Rubisco activation rate, except in the pTRV-TRX-x plants, where these parameters were not significantly altered. For example, initial Rubisco activities decreased by 31.9, 47.4, 35.6, and 39.6%, while the Rubisco activation rate decreased by 32.1, 42.0, 29.8, and 41.3% in the pTRV-TRX-f, pTRV-TRX-m2, pTRV-TRX-m1/4, and pTRV-TRX-y plants, respectively, compared with the pTRV plants. Furthermore, exogenous EBR application increased the activity or the activation rate only in the pTRV-TRX-m2 and pTRV-TRX-y plants. The activity and the activation rates of these enzymes in the EBR-treated VIGS plants were much lower than in the pTRV control plants.

TRX-f, TRX-m2, TRX-m1/4, and TRX-y are involved in the regulation of cellular redox homeostasis via the AsA-GSH cycle and 2-CP
TRXs are involved in the regulation of ROS metabolism and CO 2 assimilation. The results of the present study demonstrated that TRXs can differentially regulate oxidative stress in tomato plants. In rice, Ostrxm RNAi plants exhibit increased H 2 O 2 accumulation in the leaves (Chi et al., 2008). Histochemical staining demonstrated that more O 2 ·− and H 2 O 2 accumulates in the leaves of the VIGS-TRX-f/TRX-m pea plants compared with control plants, confirming that TRX-f and TRX-m are important for the metabolic balance of ROS in plant cells (Luo et al., 2012). Here, it was found that partially silencing TRX-f, TRX-m2, TRX-m1/4, or TRXy resulted in increased oxidative stress, as indicated by the increased MDA content in the leaves, while partially silencing TRX-x did not induce significant changes in the MDA content of the leaves (Fig. 4a). These findings revealed that chloroplastic TRX-f, TRX-m2, and TRX-m1/4 as well as TRX-y are involved in ROS metabolism. It seems likely that TRX-x plays an insignificant role in ROS metabolism, as partially silencing TRX-x did not induce oxidative stress in the leaves.
ROS accumulation is closely related to the ROS scavenging capacity of cells, which is largely dependent on antioxidant enzymes, such as APX, MDAR, DHAR, and GR, and nonenzymatic antioxidants, such as AsA and GSH. Proteomic studies in Arabidopsis have determined that several key enzymes involved in the AsA-GSH cycle, such as GR, are redox sensitive (Ströher and Dietz, 2008;Wang et al., 2012), and biochemical and genetic assays have identified the NADPH-dependent thioredoxin system as a backup system for GR1 (Marty et al., 2009). Until now, few studies have investigated the role of TRXs in the regulation of antioxidant activity in plants. Here, it is demonstrated that partially silencing the TRX-f, TRX-m2, TRX-m1/4, and TRX-y genes resulted in decreased activity of the enzymes APX, MDAR, DHAR, and GR, while partially silencing the TRX-x gene did not induce significant changes in the activity of these enzymes (Fig. 5a). Additionally, partially silencing TRX-f, TRX-m2, TRX-m1/4, and TRX-y did not alter the total glutathione (GSH+GSSG) or GSH content but did induce an increase in the GSSG content, leading to a decrease in the GSH/GSSG ratio (Fig. 5b). These results indicated that TRX-f, TRX-m2, TRX-m1/4, and TRX-y do not influence GSH biosynthesis but affect glutathione homeostasis by regulating GR activity, as observed in both the glutathione-deficient cad2 and GR-deficient gr1 Arabidopsis mutants (Meyer et al., 2007;Marty et al., 2009). However, such an effect was not observed in the pTRV-TRX-x plants. All of these results suggested that the enzymes involved in the AsA-GSH cycle are largely modified by TRX-f, TRX-m2, TRX-m1/4, and TRX-y,  and that the oxidative stress induced in partially TRX-f, TRX-m2, TRX-m1/4, and TRX-y-silenced plants is at least in part attributable to the reduced activity of the antioxidant enzymes. The finding that the cellular glutathione redox status could be regulated by chloroplastic TRXs further revealed the existence of cross-talk between the TRX system and the glutathione system. In addition to the AsA-GSH cycle, peroxiredoxin, which is activated by TRX, acts as an alternative pathway to remove H 2 O 2 from chloroplasts (Foyer and Shigeoka, 2011). The 2-CP enzyme exhibits broad substrate specificity, showing activity toward both hydrogen peroxides and complex alkyl hydroperoxides. During the peroxide reduction reaction, 2-CP is alternatively oxidized and reduced as it catalyses the electron flow from an electron donor to peroxide. In vitro tests have demonstrated that TRX-x is by far the most efficient in Arabidopsis . However, only the oxidized dimeric form of 2-CP was detected in the Ostrxm RNAi plants, suggesting that Ostrxm could also regulate the catalytic activity of 2-CP by reducing redox-active cysteine residues in plants (Chi et al., 2008). The present study revealed increased accumulation of oxidized 2-CP in all of the partially TRX-silenced plants, with the partial silencing of TRX-x, TRX-m1/4, and TRX-m2 being more significant, suggesting that these TRXs could differentially regulate 2-CP (Fig. 4b, c).

TRX-f and TRX-m1/4 play important roles in BRs-induced changes in cellular redox homeostasis
As observed in previous research, foliar application of EBR relieved oxidative stress (Xia et al., 2009b). In addition to increasing the activity of the AsA-GSH cycle, EBR also  induced an increase in the ratio of the reduced monomers to the oxidized dimers of 2-CP (Fig. 4c). Importantly, partially silencing TRX-f and TRX-m1/4 compromised the EBRinduced decrease in MDA content and increase in the activity of antioxidant enzymes, as well as the associated changes in glutathione metabolism (Figs 4a, 5a, b), suggesting that EBR partially alleviates oxidative stress by regulating TRX-f and TRX-m1/4. It was found that the TRXs are actively involved in the regulation of photosynthesis-related gene transcripts, enzyme activity, and, ultimately, the CO 2 assimilation capacity. Ostrxm plants exhibit abnormal chloroplast development and growth inhibition in rice plants, displaying decreased levels of several chloroplast proteins that are critical for photosynthesis and biogenesis (Chi et al., 2008). Silencing TRX-f and TRX-m (VIGS-TRX-f/TRX-m) in pea plants results in a significant reduction in Mg chelatase activity and the 5-aminolaevulinic acid synthesizing capacity as well as decreased transcript levels of RBCS and chlorophyll biosynthesis-related genes and a decreased photosynthetic capacity (Luo et al., 2012). In the present study, it was found that partially silencing the chloroplastic TRXs, except for TRX-x, resulted in a reduced photosynthetic capacity in tomato leaves (Fig. 3), supporting the hypothesis that TRX-f, TRX-m2, TRX-m1/4, and TRXy are actively involved in the regulation of CO 2 assimilation. In Arabidopsis, the maximal (F v /F m ) and effective quantum yields of photosystem II (PSII; Ф PSII ) in the trx f1.1 and trx f1.2 mutant plants are not different from those in wild-type plants ( Thormählen et al., 2013). In agreement with this finding, it was observed that only the pTRV-TRX-m1/4 plants showed a decrease in F v /F m , by 15.4% ( Supplementary Fig.  S2 available at JXB online). All of these results suggested that the TRXs do not affect CO 2 assimilation by modulating PSII electron transport.

TRX-f and TRX-m1/4 play important roles in BRsinduced changes in CO 2 assimilation
Previous studies have shown that chloroplastic TRXs participate in the regulation of the Calvin cycle and associated processes (Balmer et al., 2003). In addition to the down-regulation of genes involved in the Benson-Calvin cycle, partially silencing TRX-f, TRX-m2, TRX-m1/4, and TRX-y also resulted in a significant decrease in the initial Rubisco and FBPase activities as well as RCA activity (Fig. 7). An in vitro assay revealed that AtTRX-f is able to activate FBPase and NADP-MDH most efficiently, followed by AtTRX-m1, AtTRX-m2, and AtTRX-m4, whereas AtTRX-m3, AtTRX-x, and AtTRX-y cannot efficiently activate NADP-MDH from sorghum or from Chlamydomonas (Collin et al., , 2004Lemaire et al., 2003Lemaire et al., , 2005. The simultaneous decrease in V c,max and initial Rubisco activity, with little change in total Rubisco activity, indicates that chloroplastic TRX-f, TRX-m2, TRX-m1/4, and TRX-y mainly regulate the activation state of Rubisco via the action of Rubisco activase (Figs 3b, 7). The regeneration of RuBP is dependent on both the photosynthetic electron transport chain and the enzymes downstream of Rubisco in the Calvin cycle (Long et al., 2006). In addition to the decrease in RCA activity, initial FBPase activity was also decreased in the plants in which TRX-f, TRX-m2, TRX-m1/4, and TRX-y were partially silenced (Fig. 7). Accordingly, the lack of direct activation of RCA or FBPase was involved in the decrease in J max . Interestingly, the observed changes in the activities of these enzymes were in agreement with cellular glutathione homoeostasis (Fig. 5b). In a previous study, it was found that several redox-sensitive Benson-Calvin cycle enzymes, such as RCA, can undergo disulphide bond interchange, leading to changes in the activities of these enzymes (Jiang et al., 2012b). It is likely that the oxidized cellular environment induced by the partial silencing of TRX-f, TRX-m2, TRX-m1/4, and TRX-y can directly modify the structure of these redox-sensitive enzymes. Previously, it was found that inhibition of the biosynthesis of BRs using an inhibitor decreased CO 2 assimilation, while exogenously applied BRs increased CO 2 assimilation in cucumber plants (Yu et al., 2004;Xia et al., 2009a). Here, it was demonstrated that exogenous EBR application resulted in enhanced CO 2 assimilation in tomato plants, while BRsdeficient d ^im mutant plants exhibited decreased CO 2 assimilation compared with wild-type CR plants (Fig. 2b), providing further evidence that BRs levels are closely related to the CO 2 assimilation capacity of plants.
BRs can differentially induce TRX transcripts in Arabidopsis and rice (Müssig et al., 2002;Goda et al., 2004;Wu et al., 2008). In the present study, it was found that BRs were able to up-regulate TRX-f, TRX-m2, TRX-m1/4, and TRX-x but   did not affect the transcription of TRX-y (Fig. 2a). In contrast, partially silencing TRX-x did not induce any changes in CO 2 assimilation, the activities of antioxidant enzymes, redox homeostasis, or the expression and activities of photosynthesisrelated genes and enzymes (Figs 3a, 5a, 5b, 6, 7), suggesting that TRX-f, TRX-m2, and TRX-m1/4 are potentially involved in the BRs-induced increase in CO 2 assimilation. However, the partial silencing of TRX-f, TRX-m2, and TRX-m1/4 resulted in decreases in CO 2 assimilation, the expression of Benson-Calvin cycle-related genes, and the activities of associated enzymes, while EBR induced these changes only in pTRV-TRX-m2 plants and not in pTRV-TRX-f or pTRV-TRX-m1/4 plants (Figs 3a,  6, 7). These results suggest that TRX-f and TRX-m1/4 are the TRXs responsible for BRs-induced CO 2 assimilation. Many processes involved in plant growth, development, and metabolism are under the tight control of the TRXs (Schürmann and Buchanan, 2008). The results of this study demonstrated that the chloroplastic TRXs of tomato play different roles in cellular redox homeostasis and CO 2 assimilation, and BRs regulate cellular redox homeostasis and CO 2 assimilation in a TRX-f-and TRX-m1/4-dependent manner (Fig. 8). Among the examined TRXs, TRX-f, TRX-m2, TRX-m1/4, and TRX-y were all able to activate the antioxidant system by up-regulating the AsA-GSH cycle and the Benson-Calvin cycle, while TRX-x is not involved in the regulation of these processes. In addition to their direct effects on gene transcription, BRs also increased the antioxidant capacity, leading to a reduced cellular environment, which could directly reduce specific disulphide bridges in photosynthesis-related redox-sensitive enzymes, leading to the activation of these enzymes and therefore playing an important role in the regulation of normal photosynthesis.

Supplementary data
Supplementary data are available at JXB online. Figure S1. Relative mRNA abundance of TRX-f,  in respective virus-induced gene silencing (VIGS) plants. Figure S2. Changes in maximum quantum yield of PSII (F v /F m ) in the leaves of pTRV and various partially TRXsilenced plants. Table S1. PCR primers designed for vector construction. Table S2. Gene-specific primers designed for qRT-PCR.