Analysis of widely targeted metabolites of the euhalophyte Suaeda salsa under saline conditions provides new insights into salt tolerance and nutritional value in halophytic species

Background Suaeda salsa L. (S. salsa) is an annual euhalophyte with high salt tolerance and high value as an oil crop, traditional Chinese medicine and vegetable. However, there are few comprehensive studies on the metabolomics of S. salsa under saline conditions. Results Seedlings of S. salsa were cultured with 0, 200 and 500 mM NaCl for two days. Then, widely targeted metabolites were detected with ultra performance liquid chromatography and tandem mass spectrometry. A total of 639 metabolites were annotated. Among these, 253 metabolites were differential metabolites. Salt treatment increased the content of certain metabolites, such as nucleotide and its derivates, organic acids, the content of amino acids, lipids such as α-linolenic acid, and certain antioxidants such as quercetin. These substances may be correlated to osmotic tolerance, increased antioxidant activity, and medical and nutritional value in the species. Conclusion This study comprehensively analyzed the metabolic response of S. salsa under salinity from the perspective of omics, and provides an important theoretical basis for understanding salt tolerance and evaluating nutritional value in the species. Electronic supplementary material The online version of this article (10.1186/s12870-019-2006-5) contains supplementary material, which is available to authorized users.


Background
Soil salinization has seriously affected agricultural production and food security. More than 800 million hectares of land worldwide are affected by soil salinization, which accounts for approximately 6% of the world's total land area, and this would result in reduced food production [1]. Meanwhile, the increasing amount of arable land lost to urban sprawl is forcing agricultural production into marginal areas [2]. Halophytes can be used as food, medicine and feed, and for restoring salinization and land contaminated with heavy metals.
Metabolite differences are directly correlated to the phenotype of an organism [3,4]. The total number of metabolites in plants is approximately 200,000 [5,6]. Changes in the type and amount of metabolites can show how the organism adapts to environmental changes [7,8]. Metabolomics is a method of qualitatively and quantitatively analyzing all metabolites in an organism [6,9]. Plant metabolism was disturbed under abiotic stress, and plants need to regulate metabolic levels to maintain basic metabolism and reach new homeostasis [10]. Hence, metabolomics was the most direct tool for studying this process [11,12]. During this process, the change in primary metabolism was most pronounced and also showed the general trend of plant response to abiotic stress. It involved the accumulation of compatible solutes such as sugars and sugar alcohols, amino acids, etc. to cope with osmotic stress [13]. However, changed in secondary metabolism were more specific for different species and stress conditions, such as modification and interaction of enzyme proteins [14], increase in asparagine [15], accumulation of flavonoids and other phenols [16], to scavenge ROS and act as a signal regulator substance [13]. At present, metabolomics has been playing an important role in the analysis of regional differences in wild rice [17], plant tolerance to abiotic stress [10], nitrogen metabolism [18], and the phenotypic variation analysis of Saccharomyces cerevisiae [19]. For example, under salt stress, co-induced glycolysis and sucrose metabolism and the co-reduction of the methylation cycle occur in Arabidopsis under salt stress [20]. Furthermore, small molecular organic solutes involved in the resistance to osmotic stress in Zea mays are significantly induced at high salinity, and this is found to be stronger in the shoots, when compared to the roots [21]. In salttolerant Hordeum vulgare L., the levels of hexose phosphate, TCA circulating intermediates, and metabolites involved in cell protection increases with the increase in salt concentration [22]. These studies have found pathways and metabolites that play important roles in salt tolerance. Therefore, metabolomics provides an important basis for salt response studies in plants. Suaeda salsa (S. salsa) has stronger salt tolerance, and its metabolic level changes under salt stress. Hence, it has high value for research.
S. salsa is an euhalophyte with high salt tolerance during germination [23][24][25][26], vegetative growth [27] and reproductive stages [28][29][30][31]. The species is considered to have potential as a vegetable and oilseed crop [32,33], and it is rich in protein, crude fiber, carotenoids and amino acids [34]. Furthermore, the species has high value as a medicine, and is a promising model for understanding salt tolerance [34]. The content of flavanols in S. salsa is much higher in July (98.8 mg g − 1 DW) than in the other months, and the extracts in July have the highest antioxidant activity in vitro [34]. S. salsa has high ability to maintain ion homeostasis. For example, SsNHX1 and SsSOS1 are involved in maintaining Na + homeostasis, while SsHKT1;1 is involved in maintaining K + and SsCAX1 is involved in maintaining Ca 2+ homeostasis under salinity in S. salsa [34]. Salinity upregulated expression levels in certain genes, such as choline monooxygenase (CMO), betaine aldehyde dehydrogenase (BADH) and catalase (CAT), elevate the activities of superoxide dismutase (SOD), peroxidase (POD), CAT and glutathione peroxidase (GPx) in S. salsa [35]. High salinity causes metabolic responses, such as depleted amino acids, malate, fumarate, choline and phosphocholine, and elevated betaine and allantoin in the shoots, as well as depleted glucose and fructose, and elevated proline, citrate and sucrose in the roots of S. salsa [35]. A series of metabolomics studies have also been conducted in S. salsa under heavy metal stress at ambient salt concentrations [36][37][38]. However, there are few comprehensive studies on the metabolomics of S. salsa in controlled saline conditions, such as determining how salinity affects the metabolites in the species. Widely targeted metabolomics based on multiple reaction monitoring (MRM), using multiple ion monitoring (MIM) survey scans to trigger enhanced product ion (EPI) acquisition to identify metabolites. Compared to nontargeted metabolomics, it is a more sensitive and accurate method for detecting metabolites [39]. Therefore, widely targeted metabolites of S. salsa under salt stress were investigated in the present study.

Data quality assessment
During the analysis, quality control (QC) samples prepared by mixing the sample extracts were inserted in every 10 test samples to monitor the reproducibility of the analysis process. The accuracy and reproducibility of metabolite detection could be determined using the superimposed display analysis of mass spectrometry total ion current (TIC). TIC is the spectrum obtained by continuously summing the intensity of all ions in the mass spectrum at each time point. The multi-substance extracted ion chromatogram (XIC) can be used determine the ion flux spectrum of each extracted substance in the multiple reaction monitoring mode (MRM). The mass spectral peaks for each color represents the different metabolites detected. The peak area represents the relative content of the corresponding substance. The integration and calibration of peaks were performed using the Mul-tiaQuant software (v 3.0.3).

Principal component analysis (PCA)
The PCA of the quality control and treatments revealed that the variability of each treatment of samples was small. The samples had similar metabolic characteristics, and the test results were stable and reproducible. In addition, the separation trend between treatments was obvious, indicating the significant metabolic differences between salt treatments (Fig. 1a). The metabolism of these three treatments was clearly separated in the first component (PC1), and the effect of salt treatment on the metabolism of S. salsa was obvious (Fig. 1b, c and d).

Orthogonal partial least squares-discriminant analysis (OPLS-DA)
The OPLS-DA decomposes the X matrix information into Y correlation and irrelevance by orthogonal signal correction (OSC) and partial least squares discriminant analysis (PLS-DA) [40]. The difference variables were filtered by eliminating the irrelevant differences. Compared with PCA, PLS-DA can maximize the distinction between treatments, and is more conducive to finding differential metabolites. The R 2 Y and Q 2 scores were all greater than 0.99 in the 200 mM NaCl vs. CK, 500 mM NaCl vs. CK, and 500 mM NaCl vs. 200 mM NaCl (Fig. 2a, b and c), demonstrating that the results of the salt treatment that led to the differential metabolism of S. salsa were correct.
The OPLS-DA model was verified using 200 alignment experiments. The horizontal line corresponded to the R 2 and Q 2 of the original model, while the red and blue dots represented the R 2 ' and Q 2 ' after replacement, respectively. These results show that the model was meaningful, and that the differential metabolites could be screened according to the VIP value analysis in the subsequent analysis (Fig. 2d, e and f).

Hierarchical cluster analysis (HCA) and volcano plot
The HCA can evaluate differences in the characteristics of salt treatment that lead to metabolite accumulation, and comprises of intra-treatment homogeneity and inter-treatment variability. With the increase in NaCl concentration, the difference in the expression of  (Fig. 3a). The points in the volcano map represents the metabolites, and the abscissa and ordinate represents the logarithm and VIP values of the quantitative difference of metabolites in the two samples, respectively. Metabolites with a fold change of ≥2, a fold change of ≤0.5, and a VIP of ≥1 were selected. The metabolites screened under the above conditions had significant differences. There was no difference in the expression of most metabolites, and the number of increases in differentially expressed metabolites was close to the number of decreases (Fig. 3b, c and d).

Statistical analysis of differential metabolites
There were 253 metabolites with different content in all 639 metabolites (Additional file 1: Table S1). These were divided into 10 categories: amino acid and its derivatives, phenolamides and phenolic acids, nucleotide and its derivates, flavonoids, lipids, carbohydrates, vitamins, indole derivatives, organic acids, and others (Table 1). Compared to controls, there were 101 differential metabolites in the 200 mM NaCl treatment, in which 61 metabolites were decreased, while 40 metabolites were increased. Furthermore, there were 199 differential metabolites in A value closer to 1 means that the model is more stable and reliable, and when Q 2 is > 0.9, the model is excellent. The horizontal line corresponds to the R 2 and Q 2 of the original model, and the red and blue points represent the R 2 ' and Q 2 ' of the model after Y replacement, respectively the 500 mM NaCl treatment, in which 104 metabolites were decreased, while 95 metabolites were increased. In addition, there were 107 differential metabolites in the 500 mM NaCl treatment, when compared to the 200 mM NaCl treatment, in which 46 metabolites were decreased, while 61 metabolites were increased (Table 1 and Additional file 2: Table S2).

Functional annotation and enrichment analysis of differential metabolites
The results of these differentially significant metabolite annotations were classified according to the type of pathway in the KEGG database (http://www.genome.jp/kegg/). differentially expressed metabolites that were decreased, the red dots represent the increased differentially expressed metabolites, and the black color indicates the detected metabolites, but there was no significant difference; (e) the Venn diagram shows overlapping and specific differential metabolites from CK, 200 and 500 mM NaCl Differential metabolites were mainly involved in the metabolic pathways and biosynthesis of secondary metabolites, such as flavonoids, phenols and phenolic acids, amino acids and their derivatives, lipids, organic acids, and other small molecules (Fig. 4a, b and c, Additional file 4: Table S4). The changes in content of these metabolites may play an important role in cell membrane structure protection, maintaining cell osmotic potential and resisting the destruction of reactive oxygen species (ROS) (Fig. 5).

Discussion
Seeds of S. salsa contain high oil, and are rich in unsaturated fatty acids, especially linoleic acid. Hence, this can be used as a source of high-quality edible oil [32,33]. Fatty acid methyl esters extracted from seedlings and seeds, especially 9, 12-octadecandienoic acid methyl ester and γlinolenic acid methyl ester, can inhibit the formation of inflammatory factors, and have obvious anti-inflammatory effects [34]. Furthermore, S. salsa can enrich the heavy metals in the soil, such as Cu, Zn, Pb and As, in order to restore contaminated land, and has high value as the modal for understanding plant salt tolerance [34]. However, the comprehensive metabolic response of S. salsa under saline conditions has not been investigated.
Euhalophytes decrease their water potential by accumulating organic substances with small molecules and inorganic ions to resist external osmotic stress [34,[41][42][43][44]. Soluble sugars and organic acids play important roles in decreasing water potential for plants under osmotic stress [45]. In the present study, salinity induced the increase of certain kinds of soluble sugars and organic acids, when compared to controls (Table 1). Under osmotic stress, the study conducted by Cao et al. (2004) revealed a significant increase in the content of certain free amino acids, including proline (Pro), aspartic acid (Asp), phenylalanine (Phe) and alanine (Ala), in maize [46]. In the present study, high salinity (500 mM) induced the increase of amino acids, when compared to low salinity (200 mM) ( Table 1). This means that S. salsa under high salinity can accumulate more amino acids, as well as soluble sugars and organic acids, which can decrease the osmotic potential, and help the species deal with osmotic stress. Furthermore, these amino acids can also increase the nutritional value in species such as vegetables. It has been shown that 3methylcrotonyl glycine is an acyl glycine that has been studied in animals, but this is not been reported in plants [47]. Interestingly, 3-methylcrotonyl glycine was detected in S. salsa at 200 mM NaCl, but was barely detected in controls and at 500 mM NaCl (Additional file 1: Table S1). Hence, the physiological role and molecular mechanism for the metabolism of the substance in S. salsa can be further investigated.
Lipids mediate some important mechanisms to deal with salt stress, such as the process of participating in cell metabolism and maintaining the stability of the cytoskeleton [48][49][50][51]. In the present study, 500 mM NaCl induced the significant differential expression of lipid metabolites, and more unsaturated fatty acids were increased, when compared to those in controls and at 200 mM NaCl (Table 1). Monoacylglyceride (MAG), lysophosphatidylcholine (LysoPC), octadecatrienoic acid's derivates, α-linolenic acid (ALA) and punicic acid are unsaturated fatty acids that are good for human health, especially ALA [52]. The main role of ALA may be as a precursor to long-chain n-3 PUFAs, such as EPA and DHA [53]. ALA may have a function of preventing heart disease and sudden cardiac death [54,55]. The increase of α-linolenic acid content plays an important role in regulating intracellular fatty acid unsaturation to resist salt stress [56]. In the present study, 500 mM NaCl increased the content of α-linolenic acid, and the products related to metabolic pathways also significantly increased (Additional file 2: Table S2). Hence, the increased content of ALA may play a role in salt tolerance. Meanwhile, this means that planting S. salsa under high salinity should increase its nutritional value as edible vegetables.
Polyphenols are generally divided into two categories: flavonoids (e.g. flavones, flavanones, flavonols and catechins) and phenolic acids (e.g. hydroxybenzoic acid, hydroxycinnamic acid and quinic acid) [57][58][59]. Polyphenols have high antioxidant activity, and can prevent cardiovascular diseases and cancers [60,61]. Therefore, these have been applied in functional foods, cosmetics and medicine [62]. In the present study, different contents of 60 flavonoids and 28 phenolic acids under salt treatment were detected (Additional file 2: Table S2). The content of 5-O-p-coumaroylquinic acid, N-acetyl tryptamine, 2′-hydroxygenistein, morin, 2,4-dihydroxybenzoic acid, N′, N″, N″'-p-coumaroyl-cinnamoyl-caffeoyl spermidine, butin, quercetin, isorhamnetin, homoeriodictyol, hesperetin, naringenin, protocatechuic aldehyde and phloretin significantly increased with salinity. Among these, certain substances have special functions. For example, quercetin is an anti-oxidative flavonoid widely distributed in plants, and is a promising agent for cancer prevention [63][64][65][66]. However, more antioxidants, such as flavonoids and polyphenols, were decreased by salinity in the present study (Table 1). In S. salsa, the activities of Mn-SOD and several isoforms of Fe-SOD and CuZn-SOD increased under salinity [67]. The overexpression of Ss.sAPX (a gene of the stromal APX in S. salsa) increased the salt tolerance of transgenic Arabidopsis plants during both the germination and vegetative growth stages [68]. S. salsa is highly salt tolerant, and grows as well with 400 mM NaCl, when compared to 10 mM NaCl [69]. The distinctive trait of salt tolerance in euhalophytes, such as S. salsa, is to maintain ion homeostasis, including Na + accumulation in the vacuoles [70], Na + and Cl − exclusion through the roots [70,71], and K + [72] and Ca 2+ [73] homeostasis. Scavenging ROS may be the subsequent mechanism in salt tolerance only when ion homeostasis is destroyed in S. salsa. In addition, S. salsa may scavenge ROS, and mainly relies on enzymatic components, rather than on non-enzymatic components. The role of enzymatic and non-enzymatic components in scavenging ROS should be further investigated. Based on the present results, this study demonstrated that the reconstitution of the metabolic homeostasis of S. salsa under salt stress included: 1) the accumulation of primary metabolites (e.g. amino acids, soluble sugars, organic acids, lipids) exerts resistance to osmotic stress and maintains the osmotic potential and metabolism of cells; 2) secondary metabolites may play an important role as antioxidants and regulatory substances including quercetin, 2,4-dihydroxybenzoic acid, isorhamnetin, 2′-hydroxygenistein, and other metabolites that were significantly inhibited or induced by salt. These metabolites have special nutritional and medicinal value, and also greatly increase the application value of S. salsa (Fig. 6).

Conclusion
The present study used widely targeted metabolites based on the UPLC-MS/MS detection platform to analyze the metabolic differences of S. salsa under different NaCl concentrations. The present study comprehensively analyzed the metabolic response of S. salsa under salt stress from the perspective of omics, providing an important theoretical basis for understanding salt tolerance and evaluating the nutritional value of S. salsa.

Samples
The seeds of S. salsa were obtained from Dongying Research Academy of Agriculture Science, China. Dry seeds were stored in a fridge at < 4°C before use.
In late March 2018, the brown seeds of S. salsa were planted in each plastic pot, which had 2 kg of rinsed river sand. The seedlings were cultured in a glasshouse under natural light. The temperature was 24 ± 4°C during the day and 18 ± 4°C at night in the glasshouse. The seedlings were watered every day with 1 mM of NO 3 − -N nutrient solution [70]. The pH of the solution was adjusted to 6.2 ± 0.1 with KOH and H 2 SO 4 .
After the seedlings were pre-cultured for 50 days, 15 seedlings in each pot were left and treated with 0 (as a control, which is indicated as CK in the figures), 200 and 500 mM NaCl (three pots for each concentration of NaCl), which was prepared with 1 mM NO 3 − -N nutrient solution [70]. In order to avoid osmotic shock, NaCl was added in increments of 50 mM per day. After the highest salinity concentration was obtained for two days, the matured fresh leaves in the same position were harvested and frozen in liquid nitrogen. Then, the samples were stored in dry ice and mailed to MetWare for metabolite testing. For each concentration of NaCl, the matured fresh leaves from each pot were harvested for metabolite testing. That is, three replicates were set for each NaCl treatment.

Sample extraction process
The leaves were first freeze-dried and then grounded to powder using a grinder (MM 400; Retsch, Germany) at 30 Hz for 1.5 min. Then, 0.1 g of the powder was placed in 1 ml of 70% aqueous methanol at 4°C overnight, and vortexed for three times during the period to increase the extraction efficiency. Next, the extract was centrifuged at 10,000 g for 10 min, and the supernatant was filtered through a microporous membrane (0.22-μm pore size) before LC-MS/MS analysis.

Qualitative and quantitative determination of metabolites
Based on the public metabolite database (e.g. MassBank or KNApSAcK) and the self-built database MetWare database (MWDB), the material was qualitative by secondary spectral information, while the isotope signal and the repetitive signal are removed during the analysis.
The metabolites were quantified using multiple reaction monitoring (MRM) of triple quadrupole mass spectrometry. The ions corresponding to other molecular weight substances were excluded, and the precursor ions of the target substance were screened. Meanwhile, in the collision cell, the precursor ions were ionized to break and form fragment ions, and the characteristic fragment ions were selected by triple quadrupole filtration. This makes the quantitative results more accurate and repeatable. Peak area integration was performed on the obtained metabolite mass spectral peaks, and the mass spectral peaks of the metabolites in different samples were integrated [74].

Statistical analysis
Multivariate statistical analysis methods, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA), were used. PCA was used to recombine original variables into new, mutually independent variables through orthogonal transformations, revealing the internal structure of multiple variables through a few principal components [75]. Based on the results of the OPLS-DA, a multivariate analysis of variable importance in the project (VIP) in the OPLS-DA model could be used to initially screen for metabolites with differences. At the same time, differential metabolites could be further screened by combining the P-values or fold changes in the univariate analysis. Finally, the differential metabolites were precisely mined. Hierarchical cluster analysis (HCA) was performed on the accumulation patterns of metabolites between different samples using the R software (www.r-project.org/). The KEGG database was used to annotate the differential metabolites [76].

Additional files
Additional file 1: Table S1. All detected metabolites in the leaves of S. salsa treated with different concentrations of NaCl. (XLSX 109 kb) Additional file 2: Table S2. All differential metabolites in the leaves of S. salsa treated with different concentrations of NaCl. (XLSX 68 kb)