H2S Regulation of Metabolism in Cucumber in Response to Salt-Stress Through Transcriptome and Proteome Analysis

In a previous study, we found that H2S alleviates salinity stress in cucumber by maintaining the Na+/K+ balance and by regulating H2S metabolism and the oxidative stress response. However, little is known about the molecular mechanisms behind H2S-regulated salt-stress tolerance in cucumber. Here, an integrated transcriptomic and proteomic analysis based on RNA-seq and 2-DE was used to investigate the global mechanism underlying H2S-regulated salt-stress tolerance. In total, 11,761 differentially expressed genes (DEGs) and 61 differentially expressed proteins (DEPs) were identified. Analysis of the pathways associated with the DEGs showed that salt stress enriched expression of genes in primary and energy metabolism, such as photosynthesis, carbon metabolism and biosynthesis of amino acids. Application of H2S significantly decreased these DEGs but enriched DEGs related to plant-pathogen interaction, sulfur-containing metabolism, cell defense, and signal transduction pathways. Notably, changes related to sulfur-containing metabolism and cell defense were also observed through proteome analysis, such as Cysteine synthase 1, Glutathione S-transferase U25-like, Protein disulfide-isomerase, and Peroxidase 2. We present the first global analysis of the mechanism underlying H2S regulation of salt-stress tolerance in cucumber through tracking changes in the expression of specific proteins and genes.


INTRODUCTION
Salinization of soil is gradually becoming an earnest threat to world agriculture, affecting approximately 20% of arable irrigated land and consequently leading to the loss of US$ 27.5 billion per annum (Abdel Latef et al., 2019). Salt stress impedes plant growth and development, crop productivity, and geographic distribution by imposing ionic toxicity and perturbing cellular osmotic potential due to excessive accumulation of Na + (Dai et al., 2018). Soil salinity not only affects plant growth and development but also causes severe osmotic stress, toxicity from sodium and chloride ions, and oxidative damage . Therefore, plants may develop a sophisticated mechanism to exclude toxic ions to mitigate salt stress and lots of studies have demonstrated that salt-tolerant genotypes may accumulate less Na + through reducing Na + influx into the root, control of Na + xylem loading, Na + retrieval from the xylem, Na + recirculation in the phloem, Na + efflux from the root, intracellular compartmentation of Na + into the vacuoles, and Na + secretion from the leaf, which is a general rationale for combating salt stress (Zhu, 2003). These responses can be modified by small molecules such as plant growth regulators and signaling molecules (Mostofa et al., 2015a).
Sulfur metabolism is connected directly via the Met salvage cycle to ethylene-and polyamine-related responses to abiotic stresses (Sauter et al., 2013). Hydrogen sulfide (H 2 S) is a small gas molecule that resides in all kinds of plant tissues and cells and can serve both as physiological and signaling functions. (Rausch and Wachter, 2005). As a gaseous signal molecule, H 2 S functions in seed germination, growth and development, and stress adaptation in plants . L-cysteine desulfhydrase (L-CD, EC 4.4.1.1) and D-cysteine desulfhydrase (D-CD, EC 4.4.1.15) can catalyze the degradation of L-/D-cysteine to produce H 2 S, amine and pyruvate . The L-CD and D-CD were also identified as being mainly responsible for the degradation of cysteine in order to generate H 2 S (Jin et al., 2011).
A direct impact of Cys on ABA formation together with the herein identified co-regulation of genes related to ABA synthesis-and Smetabolism revealed a reciprocal regulatory network with the function of ensuring optimal supply with Cys for ABA synthesis to combat salt stress in Arabidopsis thaliana (Cao et al., 2014). H 2 S enhances tolerance to salinity during germination of Medicago sativa seeds and in barley seedlings (Wang et al., 2012;Chen J. et al., 2015). H 2 S interacts with the nitric oxide (NO) pathway, alters redox homeostasis, and prevents K + loss in seedlings of Medicago sativa (Wang et al., 2012;Lai et al., 2014), regulates the Na + /K + balance, mineral homeostasis, and oxidative metabolism in rice (Mostofa et al., 2015b) and maintains lower Na + concentrations in wheat seedlings via the regulation of nonselective cation channels (NSCCs) and the SOS1 pathway (Deng et al., 2016). Our previous study has shown that exogenous application of NaHS, an H 2 S donor, could ameliorate salt-induced growth inhibition by maintaining the Na + /K + balance and by regulating H 2 S metabolism and the oxidative stress response (Jiang et al., 2019). However, these reports about H 2 S regulation of salt-stress tolerance primarily focused on morphological, physiological, and biochemical aspects, and little is known about the molecular mechanism of H 2 S-regulated salt-stress tolerance at the transcriptomic and proteomic levels in cucumber.
The development and application of "omics" technologies, such as genomics, transcriptomics, proteomics, and metabolomics, have resulted in new widely used experimental methods to overcome limitations in crop quality and yield (Ghatak et al., 2017). Meng et al. (2019) revealed jasmonate ZIM-domain protein 7 (JAZ7)-mediated drought tolerance in Arabidopsis using the tandem mass tags (TMT)-based quantitative proteomics and targeted metabolomics. The more advanced proteomics technology, such as quantitative phosphoproteomics are used to elucidate the mechanism that ABAinduced1 identifies AT-Hook-Like10 phosphorylation required for stress growth regulation in Arabidopsis (Wong et al., 2019). Combining comparative transcriptomic and proteomic analyses is also an effective way to investigate differential activation of response and metabolism pathways at the global level . Transcriptional profiling based on total RNA sequencing is used to analyze gene expression changes in response to various environmental stresses (Chen Q. Z. et al., 2015) and to map the responding pathways . Since the expression of some genes is regulated only at the translational level, proteomics is also used to identify the specific proteins that are affected by different physiological conditions (Nogueira et al., 2012). Additionally, finding the correlation between transcript and protein levels helps to find the regular pattern of DEGs and DEPs. Even though two-dimensional gel electrophoresis (2-DGE) was introduced more than four decades ago, nowadays, it is still widely used for whole proteome analysis, comparative analysis of proteome changes, biomarker discovery, cancer research, as well as for the identification of protein isoforms and post-translational modifications (Goez et al., 2018).
Cucumber (Cucumis sativus L., 2n = 2x = 14) is a widely cultivated plant in the gourd family and also considered as a model species for the study of phloem characteristics, raffinose family oligosaccharide (RFO) metabolism, and sex determination (Jiang et al., 2019). The cucumber genome is relatively small, with an estimated size of 367 Mb (Arumuganathan and Earle, 1991). Our hydroponically based study aids in understanding how plants respond under NaCl-toxicity and how the use of H 2 S is capable of reverting this toxicity at the transcript and protein expression levels. In the present study, an integrated analysis based on RNA sequencing and 2-DE was used to investigate the global mechanism behind H 2 S-regulated salt-stress tolerance at the transcriptomic and proteomic levels. So far, this study is the first to combine transcriptomic and proteomic analysis of the global mechanism of H 2 S-regulated salt-stress tolerance. This study also contributes to the identification of salt-stress tolerance-related genes, particularly those responsive to H 2 S regulation.

Plant Culture and Treatments
Seeds of cucumber cultivar "Chunxiaqiuwang" were purchased from the Ningyang County Lu-Ming Seed Co., Ltd. (Shandong province). Seed germination and growth conditions followed those we previously described (Jiang et al., 2019). The seedlings of cucumber were maintained in a controlled growth chamber with a light/dark regime of 14/10 h, relative humidity of 70%, temperature of 25℃ and a photosynthetically active radiation (PAR) of 800 mmol.m −2 .s −1 . When the first true leaf emerged, uniform and healthy seedlings were selected and divided into three groups for the following treatments: (i) Hoagland's nutrient solution (as control, C); (ii) Hoagland's nutrient solution +200 mM NaCl (as salt treatment, S); (iii) Hoagland's nutrient solution + 200 mM NaCl +15 µM NaHS (as H 2 S treatment, H 2 S). In our previously studies, 5, 10, 15, and 20 µM NaHS was tested for alleviation of NaCl-induced damage, and morphological and physiological assessments showed that exogenous supplementation with 15 µM NaHS was the most effective in boosting cucumber tolerance to NaCl stress (Jiang et al., 2019). Thus, the control (C), 200 mM NaCl (S), and 15 µM NaHS (H 2 S) treatments were chosen to study the transcriptomic and proteomic responses to H 2 S alleviation of salt-stress in cucumber leaves. NaHS was purchased from Sigma (St. Louis, MO, United States) and used as H 2 S donor. NaHS is commonly used as an H2S donor since it dissociates to water produce HS − and Na + , and then, combination of HS − with H + produces H 2 S (Lin et al., 2012). NaHS is responsible for induction of stress tolerance, but not Na 2 S, Na 2 SO 4 , NaHSO 4 , Na 2 SO 3 , NaHSO 3 , or CH 3 COONa (Deng et al., 2016). Three independent biological replicates were conducted per treatment (8~10 plants per treatment) under the same experimental conditions. A 10-ml fresh treatment solution was irrigated every other day at 9:00 AM during the whole culture process, and the remaining treatment solution, which was not fully absorbed by plants, was removed. After 7 days of treatment, the leaves from at least eight seedlings of each treatment sample were harvested as three replicate, immediately frozen in liquid nitrogen and then stored at −80°C until RNA and protein extraction.

RNA Isolation, cDNA Library Preparation, and Sequencing
Total RNA was isolated from each treatment using the RNAplant Plus Reagent DP437 (Tiangen, Beijing, China) according to the manufacturer's instructions. The concentration and quality of RNA were determined using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Then, the RNA was sent to the Shanghai Personal Biotechnology Cp. Ltd. for RNA-Seq analysis (Personalbio, Shanghai, China).
A total of 20-mg RNA was equally isolated from the each sample for cDNA library construction. Beads with oligo (dT) were used to enrich poly(A) mRNA, and then, fragmentation buffer was added for interrupting mRNA to short fragment (200~300 nt) which were used as templates to synthesize the first-strand cDNA using random hexamer primer, and the second-strand cDNA was synthesized by adding buffer, dNTPs, RNaseH, and DNA polymerase I, respectively. The doublestrand cDNA was purified with QiaQuick PCR Purification Kit (Qiagen, Germany) and washed with EB buffer for end repair and single nucleotide. Finally, the cDNA fragments were connected with sequencing adaptors and gel-electrophoresis was used to select cDNA fragments of 200 bp in size as the templates for amplification with PCR. The library products were ready for sequencing analysis via Illumina HiSeq ™ 2000 (Illumina, San Diego, CA, USA).

Protein Extraction, 2-DE, and Gel Staining
Protein extraction was performed using a modified version of the trichloroacetic acid (TCA)-acetone precipitation method described by Wu et al. (2011) (Wu et al., 2011). Protein concentrations were quantified according to the Bradford method (Bradford, 1976). Isoelectric focusing (IEF) was carried out on 18-cm immobilized pH gradient (IPG) linear gradient strips at pH 4-7. IPG strips in a rehydration tray were loaded with 350 ml of protein sample containing 500-mg protein for 12-16 h at room temperature. The IPG strips were then run on an Ettan IPGphor 3 (GE Healthcare, USA) using the following voltage settings: 50 V for 1 h, 100 V for 2 h, 500 V for 1 h, 1,000 V for 2 h, and 2,000 V for 1 h, followed by a gradient of 5,000 V for 1 h and 10,000 V for 2 h, with a subsequent 10,000 V rapid focusing to reach 65,000 V.h. The temperature was maintained at 20°C, with a maximum current of 50 mA per strip. After running the first dimension, IEF strips were incubated for 15 min in 10ml equilibration buffer containing 6 M urea, 30% (v/v) glycerol, 2% (w/v) SDS, 1.5-ml Tris-HCl (1.5 M, pH 8.8), 0.002% (w/v) bromophenol blue, and 1% (w/v) DTT and then equilibrated with fresh equilibration buffer containing 2.5% iodoacetamide (IAM) for another 15 min.
The equilibrated strips were placed directly onto 12.5% polyacrylamide-SDS slab gels and sealed with 0.5% agarose solution containing bromophenol blue dye. The second dimensional SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted using the Ettan DALTSix electrophoresis system (GE Healthcare, USA). Electrophoresis was carried out at 15 mA per gel for 30 min, and then, 38 mA per gel until the bromophenol blue dye front was approximately 1 cm from the bottom of the gel. After 2-DE, protein spots were visualized using Coomassie Brilliant Blue (CBB) R-250.

Image and Data Analysis
The 2-DE gels were scanned at 300 dpi using an Image Scanner III (GE Healthcare). Spot detection, quantification, and matching were conducted using ImageMaster ™ 2D Platinum version 7.0 analysis software (GE Healthcare). The optimized parameters were: smooth 2.0, min area 5.0, and saliency 3.0. The abundance of each spot was estimated by the percentage volume (%VOL), which was normalized as a percentage of the total volume of all the spots present in the gel to correct for variability caused by loading, gel staining, and destaining. Only spots with significant and biologically reproducible changes (abundance variation ≥1.3-fold or ≤0.67-fold, one way ANOVA, p ≤ 0.05) were considered to be differentially expressed proteins (DEPs).

In-Gel Digestion, MALDI-TOF/TOF-MS, and Database Searching
The selected DEPs were manually excised from the 2-DE gels, and gel pieces were digested overnight in trypsin at 37°C. The hydrolysates were then transferred to new tubes and disrupted by sonication for 15 min in a buffer containing 100-ml 60% acetonitrile (ACN)/0.1% trifluoroacetic acid (TFA) and desalinated in a ziptip (Millipore). Peptides in the supernatant were collected by centrifugation. The peptide pellets were dissolved with 1-ml matrix solution. After drying completely at room temperature, 0.6-ml over-saturated CHCA (50% ACN/ 0.1% TFA) was added. MS and MS/MS data for protein identification were obtained by using a MALDI-TOF-TOF instrument with an AB SCIEX MALDI TOF-TOF ™ 5800 Analyzer (AB SCIEX, Foster City, CA, USA) equipped with a neodymium (Nd): yttrium-aluminum-garnet (YAG) laser (laser wavelength: 355 nm). Instrument parameters were set using the 4000 Series Explorer software (Applied Biosystems). The MS spectra were recorded in reflector mode in a mass range from 800 to 4000 with a focus mass of 2000. The TOF/TOF calibration mixtures (AB SCIEX) were used to calibrate the spectrum to a mass tolerance within 10 ppm. The MS spectra were processed using TOF-TOF Series Explorer software (V4.0, AB SCIEX). At least 1,000 laser shots were typically accumulated with a laser pulse rate of 400 Hz in the MS mode, whereas in the MS/MS mode spectra, up to 2,000 laser shots were acquired and averaged with a pulse rate of 1,000 Hz. For MS calibration, autolysis peaks of trypsin ([M + H] + 842.5100 and 2, 211.1046) were used as internal calibrates, and the most intense ion signals (up to 10) were selected as precursors for MS/MS acquisition, excluding the trypsin autolysis peaks and the matrix ion signals. Peptide mass finger printing (PMF) and MS/MS queries were performed by using the MASCOT search engine 2.2 (Matrix Science, Ltd. UK) embedded into GPS-Explorer Software 3.6 (Applied Biosystems) on the database of uniprot_Eimeria acervulina with the following parameter settings: 100 ppm mass accuracy, trypsin cleavage one missed cleavage allowed, carbamidomethylation set as fixed modification, oxidation of methionine was allowed as variable modification, MS/MS fragment tolerance was set to 0.4 Da. All proteins with a confidence interval (CI) > 95% were considered to be identified successfully.

Bioinformatics Analysis
The DEPs were classified according to the biological processes in which they are involved based on UniProtKB (http://www.uniprot. org/) and NCBI (https://www.ncbi.nlm.nih.gov/) databases. Hierarchical clustering of protein expression patterns was performed using MultiExperiment Viewer (MeV) software version 4.9. The relative ratios of DEPs were subjected to log 2 transformation, the Euclidean distance similarity metric was used to define the similarity, and hierarchical clusters were assembled using the complete linkage clustering method. The clustering result was visualized through MeV software.

Statistical Analyses
Data for RNA data and protein expression levels was analyzed using SPSS software (version 21.0, SPSS Inc., USA). Differences in proteins expression levels were analyzed by Duncan's multiple range test at P ≤ 0.05 level of significance.

Data Deposition
Raw transcriptome data from this manuscript can be found in the NCBI BioProject database under accession numbers: SRP220380.

Screening Significant Differential Transcripts From Transcriptomes Based on RNA-Seq
Total RNA from three replicates each of the control (C), 200 mM NaCl (S), and 15 µM NaHS (H 2 S) treatment samples was extracted for RNA sequencing by Illumina Hiseq 2500. Nine libraries was constructed and searched for DEGs associated with H 2 S-regulated salt-stress tolerance. After filtering raw reads, a total of 382,202, 188 high-quality reads were obtained from the nine samples. The clean reads of the sequencing results for each cDNA library accounted for >97.42% of the raw reads (Q 30 > 86.22%). The standard deviation of clean reads among the three replicates for each treatment was~0.2% (Table S1). Among all nine samples, 85.75~90.78% of the reads could be mapped to the genome of cucumber. Of these mapped reads, 84.79~89.90% matched uniquely.

H 2 S Treatment Changed DEGs Pattern in Cucumber Under Salt Stress
The similarity between the expression of genes between the cucumber leaf samples following C, S, and H 2 S treatment was determined by PCA. As shown in Figure 1, the genes identified in the C, S, and H 2 S samples were clearly separated from each other, and the difference was statistically significant. The expressed genes in the control and H 2 S samples clustered more closely to each other than those from the salt treatment, indicating that salt stress remarkably changes gene expression and that H 2 S may alleviate some of these changes at the transcript level. The principal coordinates PC1 and PC2 represented 71% and 25% of the variance, respectively, and the contribution of the cumulative variance of the two principal coordinates (PC1 and PC2) accounted for 96% ( Figure 1). The transcripts were compared between C and S, between H 2 S and S, and between C and H 2 S by DESeq. Transcripts with a P < 0.05, and |log 2 fold change | > 1 were defined as DEGs. All identified DEGs are shown in Table S2. A total of 11,761 (5,417 upregulated and 6,344 down-regulated) DEGs were detected from the C, S, and H 2 S samples ( Figure 1). Among these DEGs, 4,942 DEGs (2,629 up-regulated and 2,313 down-regulated) in S vs. C, 2,351 DEGs (741 up-regulated and 1,610 down-regulated) in H 2 S vs. S, and 4,468 DEGs (2, 047 up-regulated and 2, 421 downregulated) in H 2 S vs. C were detected (Figure 1). A Venn diagram of the genes show how many genes are specific to one treatment and how many genes transcriptionally respond to multiple treatments ( Figure 1). Of the 11,761 DEGs, 1,124 DEGs were unique to untreated plants, 366 genes were unique to salt treatment, and 734 were unique to H 2 S and S treatment ( Figure 1). Additionally, the C, S, and H 2 S treatment samples shared 519 DEGs as shown in Figure 1.

H 2 S Treatment Changed GO (Gene Ontology) Enrichment in Cucumber Under Salt Stress
After gene annotation, GO enrichment analysis of differentially expressed genes was carried out. As shown in Figure 2, photosystem took the first place in cellular components in the S vs. C and H 2 S vs. S, and intrinsic component of membrane took the first place in cellular components in the H 2 S vs. C. Catalytic activity was the largest category in molecular function in the S vs. C and H 2 S vs. C; however, tetrapyrrole binding was the largest category in molecular function in the H 2 S vs. S (Figure 2). Oxidation-reduction process occupied the largest proportion of biological processes in the S vs. C and H 2 S vs. S, and photosynthesis, light harvesting occupied the largest proportion of biological processes in the H 2 S vs. C ( Figure 2).

H 2 S Treatment Changed KEGG Enrich Pattern of DEGs in Cucumber Under Salt Stress
The DEGs were compared to known metabolic and signaling pathways to identify those that were more responsive to salt stress and H 2 S amelioration. In the transcriptome, there were 118 KEGG pathways differentially regulated in the S vs. C samples, 111 in the H 2 S vs. C, and 88 in the H 2 S vs. S ( Table  S3). The 20 most enriched pathways, based on the number of DEGs (including up-and down-regulated genes), are shown in Figure 2. The signal transduction, carbohydrate metabolism, energy metabolism, amino acid metabolism, environmental adaptation, folding, sorting and degradation, and secondary metabolites biosynthesis pathways are involved in the three pairwise comparisons (S vs. C, H 2 S vs. C, and H 2 S vs. S), and the detailed metabolic pathway are as follows: plant hormone signal transduction, MAPK signaling pathway, glycolysis/ gluconeogenesis, amino sugar and nucleotide sugar metabolism,  pentose and glucuronate interconversions, photosynthesis, tyrosine metabolism, glutathione metabolism, plant-pathogen interaction protein processing in endoplasmic reticulum, and phenylpropanoid biosynthesis (Figure 3). Salt stress significantly resulted in the changes of basal metabolism such as carbohydrate metabolism (starch and sucrose metabolism and glyoxylate and dicarboxylate metabolism), energy metabolism (carbon fixation in photosynthetic organisms), amino acid metabolism (arginine biosynthesis, cysteine and methionine metabolism, alanine, aspartate and glutamate metabolism and phenylalanine metabolism), and transport and catabolism (peroxisome) in S vs. C (Figure 3). However, the treatment of H 2 S regulated the lipid metabolism (alpha-Linolenic acid metabolism and glycerophospholipid metabolism), and energy metabolism (photosynthesis-antenna proteins), as well as amino acid metabolism in leaves of cucumber under salt stress (H 2 S vs. S) (Figure 3). In addition, the treatment of H 2 S alone regulated the genes involved in the amino acid metabolism, carbohydrate metabolism, and transport and catabolism (H 2 S vs. C) (Figure 3).

H 2 S Treatment Changed Protein Expression Profiling in Cucumber Under Salt Stress
Proteomics analysis using 2-DE for protein separation was performed to examine changes in protein levels in cucumber leaves and representative gel maps are shown in Figure S1. The total numbers of protein spots reproducibly detected using ImageMaster ™ 2D Platinum software were 875 ± 93, 986 ± 11, and 776 ± 47 in the S vs. C, H 2 S vs. S, and H 2 S vs. C, respectively. Quantitative image analyses revealed a total of 69 protein spots showing significant changes in their abundance in pairwise comparisons (P < 0.05). Among the 69 DEPs observed in the gels, there were 29 DEPs in S vs. C, 25 DEPs in H 2 S vs. S, and 15 DEPs in H 2 S vs. C. Finally, 61 DEPs were chosen for further identification using MALDI-TOF/TOF MS ( Figure S1A) after removing eight very low abundance proteins. Of these 61 DEPs, 42 proteins were up-regulated and 19 proteins were downregulated, in pairwise comparisons ( Figure S1B).

MALDI-TOF/TOF-MS Analysis and Identification of DEPs
Altogether, 49 DEPs were successfully identified by MALDI-TOF/ TOF-MS. To investigate the functional information of these proteins, the Viridiplantae (taxonomy, Green plant) data in NCBInr and Uniprot Swissprot were searched using the Mascot search engine. The basic identification information for each of these DEPs was listed in Table S4 and summarized in Table 1. It was interesting to note that a chloroplast stem-loop binding protein of 41 kDa (spots 14 and 17) and an acetylornithine deacetylase-like protein (spots 31, 32, and 51) were identified in more than one spot on the same gel. This multiple observation of the same protein could result from the presence of different protein isoforms, posttranslational modification, or degradation. Forty-seven proteins out of the 49 identified MS-sequenced proteins (except spots 1 and 4) were matched with proteins in Cucumis sativus L.

H 2 S Treatment Changed Functional Categorization of the DEPs in Cucumber Under Salt Stress
Salt stress resulted in 22 DEPs, 17 of which were up-regulated and 5 of which were down-regulated in the S vs. C sample ( Figure S2A). However, H 2 S treatment of cucumber induced only 8 DEPs, 7 of which were up-regulated and only one of which was down-regulated in the H 2 S vs. C sample, and 19 proteins were significantly altered (P < 0.05) in the H 2 S vs. S sample, with 9 proteins showing up-regulation and 10 proteins showing downregulation ( Figure S2A). Functional categorization of the DEPs in S vs. C showed that 31.82% of the DEPs participated in photosynthesis and carbohydrate and energy metabolism, followed by protein folding, assembly and degradation (22.73%), amino acid biosynthesis (9.09%), and secondary metabolite biosynthesis (4.08%) ( Figure S2B). In the H 2 S vs. S, 21.05% of the DEPs belonged to stress and defense-related protein and another 21.05% to photosynthesis, followed by protein folding, assembly and degradation (15.79%), carbohydrate and energy metabolism (15.79%), and amino acid biosynthesis (10.53%). In the H 2 S vs. C, 25% of the DEPs participated in amino acid biosynthesis and protein folding, assembly and degradation, followed by photosynthesis, secondary metabolite biosynthesis, and carbohydrate and energy metabolism ( Figure S2B). Based on their putative physiological and biochemical functions, the majority of these DEPs belonged to one of seven categories, namely, photosynthesis (12, 24.49%), carbohydrate and energy metabolism (11, 22.45%), protein folding, assembly and degradation (10, 20.41%), amino acid biosynthesis (6, 12.24%), secondary metabolite biosynthesis (2, 4.08%), stress and defense (4, 8.16%), and other and unknown proteins (4, 8.16%) in all treatment samples ( Figure S2B).

H 2 S Treatment Changed KEGG Enrich Pattern of DEPs in Cucumber Under Salt Stress
There were 49 KEGG metabolic pathways that were enriched for DEPs responding to S and H 2 S treatments (Table S5-a), and most of DEPs were enriched in 13 pathways (Table S5-b, Figure  S3). The primary metabolism-related pathways, such as photosynthesis (ko00710, 7), energy metabolism (ko00010, 6), and amino acid biosynthesis (ko00940, 4) were enriched in H 2 S vs. S. It is noteworthy that several DEPs were enriched in the sulfur and cysteine metabolism pathways (ko00270, 3). These pathways are responsible for the synthesis of endogenous H 2 S, suggesting that exogenous H 2 S treatment might cause changes in endogenous H 2 S production along with altered regulation of other signaling pathways. However, very few other DEPs were enriched in other primary metabolism-related pathways, only nitrogen metabolism (ko00910, 1 DEP), lipid metabolism     Figure S1. b Accession number from the NCBInr and SwissProt database. c Protein score Confidence Interval. d TMr and TpI are theoretical molecular mass and theoretical isoelectric point, respectively. e EMr and EpI are experimental molecular mass and experimental isoelectric point, respectively. f The sequence coverage of identified proteins. g Data represents the mean of two independent replicates with standard error as indicated in Table S4.

The Correlation Analysis of Differently Expressed Genes and Proteins
To correlate transcript and protein expression profiles, accession numbers from the proteome were extracted and compared with the annotated RNA-Seq libraries ( Table S6). The expression levels of 17 out of the 47 proteins were consistent with the mRNA levels, indicating that only a few proteins were regulated directly at the transcriptional level. Cluster analyses of the 47 differentially expressed genes and proteins were conducted ( Figures 4A). Then, comparative analyses of 12 genes (27.66%) that showed similar regulation at both the transcript and protein levels and 4 genes (8.51%) that showed opposite at the transcript and protein levels was conducted ( Figures 4C). The Pearson Correlations (r) were calculated to be 0.399 (Figure 4), 0.839 (Figure 4), and −0.872 (Figure 3).

Integrated Pathway Analysis During H 2 S Regulation of Salt-Stress Tolerance in Cucumber
Forty DEG/DEPs could be mapped to just two metabolic pathways within cucumber ( Figure 5). Among them, 13 DEGs (6 upregulated genes and 7 down-regulated genes) and 7 DEPs (7 upregulated proteins) were mapped onto carbon fixation in photosynthetic metabolism using the KEGG database ( Figure  5). The correlation analysis of differential expression proteins and genes, which had the same changing trend. (D) The correlation analysis of significant differential expression both proteins and genes, which had the opposite changing trend. For the correlation analysis diagram (B-D), the level of differential expression proteins was labeled on the horizontal axis, and the level of gene expression was labeled on the vertical.

DISCUSSION
Emerging evidence suggests that the versatile signal molecule H 2 S can ameliorate some of the adverse effects induced by salt stress. We have previously reported that exogenous application of H 2 S alleviates the NaCl-induced toxicity in the leaves and roots of cucumber through analysis of morphology, photosynthesis, stomatal responses, ROS accumulation, and H 2 S homeostasis at the physiological and biochemical levels (Jiang et al., 2019). The purpose of this study was to explore the potential regulatory mechanism underlying H 2 S involvement in response to salt stress in cucumber. Our analyses revealed significant changes in transcriptome and proteome levels and identified numerous DEGs and DEPs as well as the metabolic and signaling pathways that might be related to H 2 S regulation of salt-stress tolerance.

Proton and Ion Transport Under Salt Stress
In this study, V-type proton ATPase subunit a2 (Csa_2G364570) and cation/H + antiporter15/18 (Csa_3G563310; Csa_4G050260) were both detected in the transcriptomic data, indicating that H 2 S might regulate proton and ion transport under salt-stress (Table S2). Removing cytosolic Na + from plant cell generally through the activities of Na + /H + antiporters is critically important for plants to survive under salt stress. Additionally, previous research has shown that the shaker AKT1-like channels are likely involved in both high-and low-affinity K + uptake and correlate with salt tolerance in plants (Duan et al., 2015). Our transcriptome data showed that potassium channel AKT1 (Csa_1G303700) was involved in H 2 S regulation of salt-stress tolerance in cucumber (Table S2).

Metabolic Pathways of Sulfur-Containing Compounds
Sulfur-containing compounds, including cysteine, methionine, glutathione, thioredoxin, and H 2 S, are crucial for the survival of plants under biotic and abiotic stress. Cysteine exists at the cross pathway of primary metabolism, protein synthesis, and the formation of low M r sulfur-containing defense compounds (Rausch et al., 2005). Our data revealed that H 2 S treatment resulted in significant changes in cysteine and methionine metabolism (21 DEGs) and glutathione metabolism (20 DEGs) at the transcriptomic level (H 2 S vs. C, Figure 3). Sulfur-containing compounds impact plant defense, and their formation saturates at a higher S-supply than does plant growth (Bloem et al., 2005). Our transcriptome analysis showed that numerous DEGs related to metabolism of sulfur-containing compounds, including probable sulfate transporter 3.5 (Csa_1G378520), thioredoxin H-type/M-typ (Csa_3G865340; Csa_6G343710), thioredoxin-like 3-2 (Csa_ 5G622670), cysteine-rich repeat secretory protein 3 (Csa_ 2G171840), and methionine S-methyltransferase (Csa_ 3G113310), responded to the addition of H 2 S during salt stress in cucumber (Table S2). In addition, H 2 S homeostasis in plant cells is closely related to cysteine metabolism. Similarly, application of H 2 S also caused significant changes at proteome level in sulfur, cysteine and methionine metabolic pathways (3 DEPs) and glutathione metabolism (1 DEP) ( Figure S3). Cysteine synthase 1 (XP_004138495.1), glutathione S-transferase U25-like (XP_004147762.1), and protein disulfide-isomerase (XP_ 004149798.1) were observed through proteomic analysis ( Table  1). H 2 S homeostasis is closely regulated by L-/D-cysteine desulfhydrase and L/D-cystiene desulphhydrase can catalyze the degradation of L-/D-cysteine to produce H 2 S, amine and pyruvate. It is surprising that L/D-cystiene desulphhydrase is not detected through transcriptome and proteome sequencing. However, Jiang et al. (2019) demonstrated that the activities of D/L-cysteine desulfhydrase significantly increased under excess NaCl and the endogenous H 2 S level also increased (Jiang et al., 2019). The metabolism related to sulfur-containing compounds, especially those that control the endogenous H 2 S level by regulating H 2 S synthetic and decomposition enzymes, plays a very important role in cucumber during H 2 S regulation of salt-stress tolerance.

Plant Hormone Signal Transduction-Related Pathways
Perception and transmission of signals would be the first steps of a response within cucumber leaves to H 2 S regulation of saltstress tolerance, and these interactions form a network of regulation leading to a variety of physiological responses (Sang et al., 2016). Here, several signal transduction-related pathways, such as plant hormone signal transduction (32 DEGs) and cAMP signaling pathway (6 DEGs), significantly responded at the transcriptomic level with application of H 2 S (H 2 S vs. S, Figure  3). Although no plant hormone signal transduction-related protein was found among the differentially accumulated proteins, numerous plant hormone signal transduction-related genes, such as auxin-responsive protein IAA13 (Csa_3G134550), auxin-binding protein ABP19a (Csa_7G450510), ethyleneresponsive transcription factor ERF017 (Csa_1G597730), salicylic acid-binding protein 2 (Csa_6G490080), abscisic acid receptor PYR1-like (Csa_3G902310), and gibberellinregulated protein (Csa_3G826660), were detected in the DEG data (Table S2). Ethylene-, abscisic acid-, and salicylic acidinduced signal transduction pathways are frequently reported to be closely related to plant resistance to various stresses. Additionally, calcium signals are involved in plant responses to various stimuli, including abiotic and biotic stresses, and regulate a wide range of physiological processes (Harper, 2001). Our data showed that calcium sensing receptor (Csa_1G041560), calciumdependent protein kinase 32 (Csa_4G172540), putative calciumbinding protein CML19 (Csa_1G002110), calcium-binding protein PBP1-like (Csa_1G071800), calcium uniporter protein 3, mitochondrial-like (Csa_6G487650), calcium permeable stressgated cation channel 1 (Csa_2G416110), calcium-transporting ATPase4 (Csa_5G310810), and cation/calcium exchanger 2-like