Tomato is an important and nutritious vegetable crop worldwide. However, cultivated tomatoes are susceptible under a wide range of environmental pressures (Zhu et al., 2014; Zushi et al., 2014; Sasidharan and Voesenek, 2015). High salinity, for example, can have harmful effects, such as reducing germination, inhibiting growth and reducing fruit production (Xie et al., 2000; Cuartero et al., 2006).

Salt stress can adversely affect the growth and productivity of plants. In the course of growth and development, plants have developed many physiological and biochemical mechanisms to adapt to the stress of the environment (Zhu et al., 2014), and a large number of physiological and metabolic reactions are produced in plants to accommodate this change (Yamaguchi-Shinozaki and Shinozaki, 2006; Hirayama and Shinozaki, 2010; Lata and Prasad, 2011; Tang et al., 2012). By studying the patterns of gene expression in different stress conditions, the adaptive mechanism of different stresses was analyzed (Zhu, 2002).

Salt stress has been observed that amino acids and carbohydrates increase greatly (Fougère et al., 1991). Haitao et al. (2013) demonstrated that arginase expression modulates abiotic stress tolerance in Arabidopsis. It was suggested by Tattini et al. (1996) that in olive leaves, mannitol and glucose play an active role in the osmotic adaptation of plants to salinity. These studies clearly demonstrate that the protein metabolism and carbohydrate metabolism play very important roles in the stress response. We can more fully understand gene function by studying the plant transcriptome. The RNA sequence is a new, efficient and rapid way to conduct transcriptional research and explore the mechanisms that exist in cells (Gal, 1980; Sultan et al., 2008; Trapnell et al., 2010). Despite progress in understanding how protein metabolism and carbohydrate metabolism participate in the adaptation to salt stress, the key genes and signaling pathways involved in these processes remain unclear.

Therefore, we use the RNA-seq screen for obvious pathways and key genes and protein metabolism and carbohydrate metabolism based on the Web gene ontology (GO) (http://wego.genomics.org.cn/cgi-bin/wego/index.pl) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis (http://www.funnet.ws/). TOPSIS (the Technique for Order Preference by Similarity to Ideal Solution) also test and determine the salt concentration and processing time, to provide a base for other researchers. Our aim is to explore the mechanism of protein metabolism regulation and carbohydrate metabolism regulation during the salt stress response in tomato and to provide valuable information to enhance the ability to resist pressure, and improve the quality of the plant. The results of the study can also be used as a reference for the selection of candidate genes in tomato for further functional characterization.

The adaptation of plant cells in salt stress is closely relative to various metabolic processes. It has been suggested that proline metabolism (Khedr et al., 2003), ascorbic acid (ASA) metabolism (Zushi et al., 2014), and other pathways play important roles during the salt stress response. In these metabolic processes, amino acid and carbohydrates are the main players in various metabolic and regulatory pathways. They are responsible for biological cell adaptation. Therefore, amino acid metabolism and carbohydrate metabolism are crucial to the salt stress response. In our study, we found that in the adaptation process, the differentially expressed genes were significantly enriched in 7 amino acid metabolic pathways and in 6 carbohydrate metabolic pathways.

Many studies have found that amino acid metabolism is closely related to abiotic stress tolerance (Fougère et al., 1991; Zörb et al., 2004; Haitao et al., 2013). We identified 7 significant amino acid metabolic pathways and verified 8 key genes. In our study, the ARG2 gene encoding arginase 2 (Solyc01g091170.2.1), involved in arginine and proline metabolism, was significantly enriched. Arginine is an vital amino acid to transport and storage the nitrogen and occupy a precursor to synthesis other amino acids or polyamines (Flores et al., 2008; Brauc et al., 2012). The arginase 2 genes variations associate with steroid response, which probably plays an important role in asthma development, severity and progression (Vonk et al., 2010). The AtARG2-knockout Arabidopsis could enhance environmental stress tolerance compared with the wild type, and increased tolerance was suggested by changes in physiological parameters, containing electrolyte leakage, water loss, porosity and surial rate (Haitao et al., 2013). However, in the past, most attention has focused on the function of proline which occupy a compatible osmolyte (Yancey et al., 1982) and osmoprotectant (Serrano and Gaxiola, 1994), It is much less concerned with the further role of stress tolerance. Our study suggested that proline metabolic pathways exhibit significant differences during the salt stress response (P < 0.05). Consistently, Khedr et al. (2003) reported that proline induces the expression of salt-stress-responsive proteins, which may increase the adaptation of Pancratium maritimum L. to response salt stress. The gene encoding beta-glucosidase 11-like (Solyc09g075070.2.1) was down-regulated in our study, which is involved in Cyanoamino acid metabolism. The role of beta-glucosidase is the hydrolysis of terminal, non-reducing beta-D-glucose residues with the increase of beta-D-glucose. Compared to the control group, the enzyme increased in the corn roots and buds (Zörb et al., 2004). Structure of acid beta-glucosidase with pharmacological chaperone provides insight into Gaucher disease (Raquel et al., 2007). López-Berenguer et al. (2008) demonstrated that cysteine and methionine decreased significantly in broccoli after salt application. Furthermore, the expression of the ACO5 gene encoding the 1-aminocyclopropane-1-carboxylate oxidase (Solyc07g026650.2.1), which is involved in cysteine and methionine metabolism, was observed to be down-regulated throughout the salt stress response. An increase in ethylene biosynthesis of arabidopsis is releated to the transcript of ACO5 under waterlogging stress (López-Berenguer et al., 2008; Sasidharan and Voesenek, 2015). ACO5 osmoregulation in cotton could probably response to water stress (Tschaplinski and Blake, 1989). ACO5 is the specific target gene of the NAC (for no apical meristem [NAM], Arabidopsis transcription activation factor [ATAF], and cup-shaped cotyledon [CUC2]) transcription factors), and many studies have found that NAC is involved in the plant abiotic stressresponse (Xie et al., 2000; Nuruzzaman et al., 2010; Rauf et al., 2013). Hernández et al. (2000) found that the activity of glutathione reductase (GR) accumulated in the salt-tolerant pea compared with the salt-sensitive cultivar after long-term NaCl treatment, indicating that glutathione metabolism is participated in the adaption to salt stress. In our study, the gene encoding a probable glutathione S-transferase (Solyc12g011300.1.1) involved in glutathione metabolism was significantly down-regulated. Cho et al. (2001) reported that Glutathione S-Transferase Mu modulates the stress-activated signals by suppressing apoptosis signal-regulating kinase 1. Roxas et al. (1997) found that hyperexpression glutathione s-transferase enhanced the growth of genetically modified tobacco seedlings suffering stress. Moreover, the gene encoding peroxidase 24 (Solyc01g067860.2.1), which is involved in phenylalanine metabolism, was observed to be significantly down-regulated. Huang et al. (2009) reported that a rice peroxidase 24 precursor was down-regulated and could exhibit peroxidase activity ofscavenging H₂O₂ under salt treatment. The total peroxidase activity improved under salt stress, and elevation activity depend on NaCl concentration (Sreenivasulu et al., 1999). SE-glutathione peroxidase play the important role for cell survival against oxidative stress (Carine et al., 1994).

Many studies have shown that carbohydrate metabolism occupies a vital function in abiotic stress tolerance (Fougère et al., 1991; Zörb et al., 2004; Haitao et al., 2013). We identified 6 significant carbohydrate metabolic pathways and verified 7 key genes. Downton (1977) found that NaCl-stressed grapevine leaves included decreased sucrose and starch but improved the reducing sugars levels. In this study, Wiv-1 (Solyc10g083290.1.1), encoding the acid invertase involved in starch and sucrose metabolism, was down-regulated during the response to salt stress. Consistently, Dubey and Singh (1999) reported that acid invertase activity decreased on the shoots of NaCl-tolerant rice but increased on NaCl-sensitive rice. It is suggested that the increasion of sugars and other compatible solutes, contributes to osmotic adjustment under salt stress. In addition, Kim et al. (2000) reported that a maize vacuolar invertase is induced by water stress. In our study, the gene encoding fructose-bisphosphate aldolase 1 (Solyc01g110360.2.1), which is participated in the pentose phosphate pathway, was significantly up-regulated. Chaves et al. (2009) also demonstrated that fructose-bisphosphate aldolase was differently affected by salt stress. Lu et al. (2012) reported that fructose 1,6-bisphosphate aldolase genes in Arabidopsis response to abiotic stresses. The gene encoding citrate synthase 3 (Solyc07g055840.2.1) was up-regulated in our study, which is participated in the Citrate cycle (TCA cycle) pathway. Citrate synthase activity is an indicator of metabolic potential in mitochondria and should increase if there is a net increase in the volume of mitochondria in the tissue. However, there was no difference in citrate synthase activity among freshwater- and seawater-acclimated fish (Marshall et al., 1999). The gene encoding myo-inositol oxygenase 1 (Solyc06g062430.2.1) was up-regulated in our study, which is participated in the Ascorbate and alarate metabolism. The vital function of myo-inositol is showed to be relative to osmotic balance and transport the Na + from roots to shoots (Nelson and Bohnert, 1999). The activity of inositol and inositol oxygenase has been found to catalyze the oxidation of the d-glucuronate free inositol (Lorence et al., 2004). Cotsaftis et al. (2011) reported that myo-inositol oxygenase was down-regulated in salt-tolerant rice.

The gene encoding glutamine synthetase-like (Solyc05g051250.2.1) was up-regulated in our study and is involved in both amino acid and carbohydrate metabolism. Glutamine synthetase is beneficial to nitrogen assimilation and has been founded in prokaryotes and eukaryotes (Doskočilová et al., 2011). Viégas and Silveira (1999) found that salinity imposed sensitivity on nitrogen assimilation. The higher the sensitivity imposed, the more severe the salt-injurious affects on plant development. Meanwhile, NO3− uptake and its assimilation were suggested to limit the nitrogen assimilation under salt stress (Silveira et al., 2001). Furthermore, genes involved in starch and sucrose metabolism (Solyc09g075330.2.1), cysteine and methionine metabolism (Solyc06g060070.2.1), glycine, serine and threonine metabolism (Solyc09g008670.2.1) and other types of o-glycan biosynthesis (Solyc01g094380.2.1) were observed to exhibit significant expression after salt stress, indicating that they were involved in the response to salt stress, but their impact on stress adaptation has not yet been reported. Therefore, these specific regulatory mechanisms need to be explored in future.

In conclusion, 200 mM and 19d are the optimum salt treatment conditions for the identification of salt tolerance in tomato. Based on RNA-Seq, we analyzed the metabolic pathways and identified some of the relevant critical genes participating in amino acid metabolism and carbohydrate metabolism during the response to salt stress in tomato. These metabolic pathways and key genes play vital roles to response salt stress, which provides valuable information to enhance the ability to resist pressure, improve the quality of the plant and lay a solid foundation for future research.

XK planned the research. ZZ, CM, and ZS conducted experiments and data acquisition. ZZ wrote the manuscript closely with all authors. XK and CM modified the manuscript. All the authors were involved in many discussion and revised the manuscript.