Comparative Transcriptome Profiling Provides Insights into Plant Salt Tolerance in Watermelon (Citrullus lanatus)

Salt stress seriously reduced the yield and quality of watermelon and restricted the sustainable development of the watermelon industry. However, the molecular mechanism of watermelon in response to salt stress is still unclear. In this study, 150 mmol·L−1 NaCl was used to deal with the seedlings of salt-tolerant and salt-sensitive watermelon varieties. Physiological characteristics showed that salt stress significantly reduced the biomass of watermelon seedlings and the accumulation of K+ in roots and leaves and significantly increased the content of Na+, Cl−, and malondialdehyde (MDA). Compared with the salt-sensitive variety, the salt-tolerant variety had higher K+ accumulation, lower Cl−, Cl− accumulation, and MDA content in roots and leaves. Then, RNA-seq was performed on roots and leaves in normal culture and under 150 mmol·L−1 NaCl treatment. A total of 21,069 genes were identified by RNA-seq analysis, of which 1412 were genes encoding transcription factors (TFs). In the comparison groups of roots and leaves, 122 and 123 shared differentially expressed genes (DEGs) were obtained, respectively. Gene ontology (GO) annotation and KEGG enrichment results showed that there were many identical GO terms and KEGG pathways in roots and leaves, especially the pathways that related to sugar or energy (ATP or NADP+/NADPH). In addition, some DEGs related to salt tolerance were identified, such as plant hormone indole-3-acetic acid (IAA) and gibberellin (GA) signal transduction pathway-related genes, K+/Na+/Ca2+-related genes, lignin biosynthesis-related genes, etc. At the same time, we also identified some TFs related to salt tolerance, such as AP2-EREBP, bZIP, bHLH, MYB, NAC, OFP, TCP, and WRKY and found that these TFs had high correlation coefficients with salt tolerance-related genes, indicating that they might have a potential regulatory relationship. Interestingly, one TCP TF (Cla97C09G174040) co-exists both in roots and leaves, and it is speculated that it may be regulated by miR319 to improve the salt tolerance of watermelon.


Introduction
Salt stress is an important factor affecting the sustainable development of global agriculture. Around 1 billion hm 2 of land worldwide are affected by salt stress, and soil salinization is expected to reach more than 50 percent of global arable land by midcentury [1]. Salt stress can cause primary stresses such as osmotic stress and ion toxicity, and high salt can also cause a series of secondary stresses such as oxidative stress and nutritional stress [2]. The accumulation of multiple stresses will affect cell growth and metabolism, thereby affecting seed germination, seedling growth, and crop yield [3,4]. In order to increase survival opportunities, plants have evolved complex mechanisms to cope with salt stress at the morphological structure, physiological metabolism, and molecular level, including the reduction of leaf number and leaf area, stomatal closure, accumulation of osmotic adjustment substances, efflux and compartmentalization of Na +

Plant Materials and Treatments
The seeds of the salt-tolerant watermelon variety ('Zhongshihong') and the saltsensitive watermelon variety ('PI186489') [13] were sown in clay after germination and then transplanted to Hoagland solution after cotyledons were fully expanded and grown in 16 h of light/8 h of dark. After 30 days of growth, seedlings were cultured in Hoagland solution containing 150 mmol·L −1 NaCl in intelligent greenhouses (25 • C, 16 h of light/8 h of dark) (Supplementary Figure S1). Roots of seedlings were sampled after 7 days of treatment. Some of the samples were stored at −20 • C for physiological and biochemical index determination. The other part was equally divided into two parts and the seedlings were frozen in liquid nitrogen immediately for RNA extraction. There were 30 seedlings of each treatment; when sampling, each sample contained 10 seedlings, with 3 biological replicates per treatment.

Measurements of Physiological and Biochemical Index
The biomass was the dry weight of the whole plant. Firstly, the surface of the seedling was dried with filter paper, and the seedling was dried in an oven until constant weight and then weighed dry weight.
All data were statistically analyzed using Office Excel 2016 software. SPSS 18.0 software was used to sort out the data for one-way ANOVA statistical analysis, and the significant difference was defined as p < 0.05 (n = 3).

RNA Extraction, Sequencing and Expression Profiling
Total RNA was isolated from the roots and leaves of 24 samples. Firstly, RNA was extracted using RNeasy Plant Mini Kit (Beijing Tiangen) according to the manufacturer's instructions. Secondly, total RNA was qualified and quantified using a Nano Drop and Agilent 2100 bioanalyzer (Thermo Fisher Scientific, Waltham, MA, USA). Thirdly, high quality and more than 200 ng RNA were amplified and reversely transcribed to cDNA based or polyA tail. The template was switched to the 5' end of RNA and full-length cDNA was generated by PCR. The cDNA fragments verified and purified in the previous step were segmented by PCR. The cDNA was quantitatively analyzed by an Agilent 2100 Bioanalyzer. The double-chain PCR product was thermally denatured and cycled by splinting oligonucleotide sequences through quality check steps. At last, single-strand circle DNA (ssCir DNA) was formatted into the final library. The final library was phi29 (Thermo Fisher Scientific, USA), which manufactured DNA nanospheres (DNB) by loading more than 300 copies of a molecule's DNA nanospheres into a patterned nanoarray. BGISEQ-500 system (BGI-Shenzhen, China) was used to sequence the cDNA library, and the read length was 100 bp.
Clean reads were obtained by screening the sequencing data, subsequently using Bowtie2 to map them to the genome of the watermelon in CuGenDB (http://cucurbitgenomics. org/organism/21, 13 January 2022). FPKM (fragments per kilobase million) was used to calculate gene expression levels. Gene annotation and function assignment were performed based on KEGG (http://www.genome.jp/kegg/, 27 February 2022) and GO (http://www. geneontology.org/, 27 February 2022) databases. DEGswere set as the following: gene fold change ≥ 2.00 and FDR (false discovery rate) ≤ 0.001. Through the GO enrichment and KEGG enrichment pathway, the metabolic pathways with significant enrichment were identified and compared with the genome-wide background. According to GO and KEGG annotation results and official classification, the DEGs were functionally classified. Under normal circumstances, FDR ≤ 0.01 is considered as significant enrichment.

RT-qPCR for DEGs
Total RNA was extracted using the Plant RNA Kit (Huayueyang Biotechnology Co., Ltd., Beijing, China). A total of 1.0 µg RNA was used for synthesizing cDNA by using the PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa) according to the manufacturer's protocol. All primers were synthesized by SunYa (Zhengzhou, China). Quantitative realtime PCR was performed on the Light Cycler480 Real-Time System (Bio-Rad Laboratories) in the following steps: 45 cycles of 95 • C for 5 min, 95 • C for 10 s, 58 • C for 10 s, and 72 • C for 10 s, followed by a melting curve analysis. Each reaction mixture (final volume: 20 µL) contained 1.0 µL previously diluted cDNA (1:5). All primers are shown in Supplementary  Table S1. Actin was used as the reference gene, and the primer sequences used were as follows: forward primer: 5 -GAACTTGGCACCTGTCCTGT-3 and reverse primer: 5 -GAACAGTGCAACAGCCTCAA-3 . Relative gene expression values were calculated using the 2 −∆∆Ct method [15].

Salt Stress Inhibits the Growth of Watermelon Seedlings
Under 150 mmol·L −1 NaCl stress, the growth of the salt-tolerant variety and the salt-sensitive variety watermelon seedlings was significantly inhibited, and the height of the plant was significantly lower than that of the control ( Figure 1A). For the two varieties, the growth of the salt-tolerant variety was better than that of the salt-sensitive variety; for example, leaves of the salt-tolerant variety grew normally under 150 mmol·L −1 NaCl treatment, while the leaves of the salt-sensitive variety withered or even died.
We further measured physiological indexes to analyze the difference between the two varieties under salt stress, including MDA content, Na + content, Cl − content, and K + content ( Figure 1B). For the four indexes, the value of the two varieties significantly increased after NaCl treatment both in leaves and roots, and the variation range of the saltsensitive variety was higher than that of the salt-tolerant variety. Moreover, the expression levels of genes Cla97C05G107320 (trehalose-6-phosphate synthase, TPS), Cla97C06G123400 (high-affinity K + transporter, HKT), and Cla97C09G180310 (K + uptake permease, KUP) related to salt tolerance in salt-tolerant variety roots were significantly higher than that of salt-sensitive variety roots ( Figure 1C). All the results suggested that salt-tolerant varieties absorbed less Na + and Cl − , absorbed more K + , and suffered much less damage than the salt-sensitive variety under 150 mmol·L −1 salt stress.

RNA-Seq for the Roots and Leaves of Watermelon under Salt Treatment
In order to analyze the related genes of watermelon response to salt stress, transcriptome sequencing was performed on watermelon seedling samples of the salt-tolerant variety and the salt-sensitive variety before and after 150 mmol·L −1 NaCl stress, including leaves and roots, with three biological replicates for each group, with a total of twenty-four samples. After data filtering, each sample produced an average of 6.33 Gb of data, of which data quality Q30 was greater than 91%, the average genome alignment rate was 93.26%, and the average gene set alignment rate was 66.65% (Supplementary Table S2). In general, the transcriptome data were of reliable quality and could be used for subsequent analysis.
A total of 21,069 genes were detected in all samples, including 20,432 known genes and 637 predicted genes, 1412 of which encoding transcription factors. Then, the correlation among the different samples was analyzed based on principal component analysis (PCA). PCA maps showed that the root and leaf samples were clearly divided into two groups (Figure 2), indicating that the gene expression profiles of these samples were reliable.

RNA-Seq for the Roots and Leaves of Watermelon under Salt Treatment
In order to analyze the related genes of watermelon response to salt stress, transcriptome sequencing was performed on watermelon seedling samples of the salt-tolerant variety and the salt-sensitive variety before and after 150 mmol·L −1 NaCl stress, including leaves and roots, with three biological replicates for each group, with a total of twentyfour samples. After data filtering, each sample produced an average of 6.33 Gb of data, of In roots, the number of DEGs before salt treatment (TCR-vs-SCR, 1794) was significantly higher than that after salt treatment (TTR-vs-STR, 1152), and the number of DEGs in salt-sensitive variety before and after salt treatment (SCR-vs-STR, 8055) was significantly higher than that in salt-tolerant variety (TCR-vs-TTR, 6755) ( Figure 3A). In leaves, the number of DEGs before salt treatment (TCL-vs-SCL, 948) was significantly lower than Life 2022, 12, 1033 6 of 21 that after salt treatment (TTL-vs-STL, 5175), and the number of DEGs in the salt-sensitive variety before and after salt treatment (SCL-vs-STL, 7180) was also significantly higher than that in salt-tolerant variety (TCL-vs-TTL, 6094) ( Figure 3B). These results indicated that watermelon responded to salt stress through a series of gene changes, and the number of DEGs in the salt-sensitive variety was significantly higher than that in the salt-tolerant variety, which might be due to more severe damage under salt stress. In addition, there were 10,679 DEGs in roots and 122 shared DEGs in the four comparison groups (TCRvs-SCR, TTR-vs-STR, TCR-vs-TTR and SCR-vs-STR) ( Figure 3C); there were 10,679 DEGs in leaves and 123 shared DEGs in the four comparison groups (TCL-vs-SCL, TTL-vs-STL, TCL-vs-TTL, and SCL-vs-STL) ( Figure 3D) (Supplementary Table S3), suggesting that these genes might play important roles in watermelon response to salt stress. groups ( Figure 2), indicating that the gene expression profiles of these samples were reliable.
In roots, the number of DEGs before salt treatment (TCR-vs-SCR, 1794) was significantly higher than that after salt treatment (TTR-vs-STR, 1152), and the number of DEGs in salt-sensitive variety before and after salt treatment (SCR-vs-STR, 8055) was significantly higher than that in salt-tolerant variety (TCR-vs-TTR, 6755) ( Figure 3A). In leaves, the number of DEGs before salt treatment (TCL-vs-SCL, 948) was significantly lower than that after salt treatment (TTL-vs-STL, 5175), and the number of DEGs in the salt-sensitive variety before and after salt treatment (SCL-vs-STL, 7180) was also significantly higher than that in salt-tolerant variety (TCL-vs-TTL, 6094) ( Figure 3B). These results indicated that watermelon responded to salt stress through a series of gene changes, and the number of DEGs in the salt-sensitive variety was significantly higher than that in the salt-tolerant variety, which might be due to more severe damage under salt stress. In addition, there were 10,679 DEGs in roots and 122 shared DEGs in the four comparison groups (TCR-vs-SCR, TTR-vs-STR, TCR-vs-TTR and SCR-vs-STR) ( Figure 3C); there were 10,679 DEGs in leaves and 123 shared DEGs in the four comparison groups (TCL-vs-SCL, TTL-vs-STL, TCL-vs-TTL, and SCL-vs-STL) ( Figure 3D) (Supplementary Table S3), suggesting that these genes might play important roles in watermelon response to salt stress.

Verification of DEGs by RT-qPCR
To verify the accuracy of RNA-seq results and examine the expression patterns of DEGs, 11 DEGs (Cla97C09G167270, Cla97C01G018360, Cla97C05G107320, Cla97C09G169250, Cla97C06G127570, Cla97C06G109590, Cla97C01G022320, Cla97C08G149800, Cla97C10G205730, Cla97C11G221050, and Cla97C09G168720) were randomly selected to conduct qPCR ( Figure 4A). The results showed that the expression patterns of 11 DEGs were highly consistent with those of genes in RNA-seq data. In addition, the correlation analysis showed that the determination coefficient (R 2 ) of RT-qPCR and RNA-seq data was 0.9134 ( Figure 4B), which demonstrated that the RNA-seq data are reliable and can be used for further analysis.

Verification of DEGs by RT-qPCR
To verify the accuracy of RNA-seq results and examine the expression patterns of DEGs, 11 DEGs (Cla97C09G167270, Cla97C01G018360, Cla97C05G107320, Cla97C09G169250, Cla97C06G127570, Cla97C06G109590, Cla97C01G022320, Cla97C08G149800, Cla97C10G205730, Cla97C11G221050, and Cla97C09G168720) were randomly selected to conduct qPCR ( Figure 4A). The results showed that the expression patterns of 11 DEGs were highly consistent with those of genes in RNA-seq data. In addition, the correlation analysis showed that the determination coefficient (R 2 ) of RT-qPCR and RNA-seq data was 0.9134 ( Figure 4B), which demonstrated that the RNA-seq data are reliable and can be used for further analysis.

GO Classification and KEGG Pathway Enrichment Analysis of DEGs
In order to analyze the main response pathways of watermelon seedlings to salt stress, the shared DEGs of the root and leaf were mapped to the GO database, respectively. All DEGs were classified on the basis of biological process, cell component, and molecular function ( Figure 5). In the roots, 122 DEGs were classified into 17 GO terms, including 6 terms of biological process, among which cellular process (39) and metabolic process (38) contained the most genes; 3 terms of cell component, among which cellular element contained the most genes (65); 8 terms of molecular function, of which catalytic activity (54) and binding (40) contained the most genes ( Figure 5A). In leaves, 123 DEGs were classified into 19 GO terms, including 11 terms of biological process, among which cellular process (31) and metabolic process (27) contained the most genes; three terms of cell component, among which cellular element (57) contained the most genes; 5 terms of molecular function, with catalytic activity (49) and binding (37) containing the most genes ( Figure 5B). By further analysis, we found that there were 14 shared GO terms in the root and leaf, accounting for 82.4% and 72.7% of the total number of GO terms in the root and leaf, respectively. Taken together, these results indicated that a large number of genes that relate to biological process, cell component, and molecular function are activated after salt stress treatment.

GO Classification and KEGG Pathway Enrichment Analysis of DEGs
In order to analyze the main response pathways of watermelon seedlings to salt stress, the shared DEGs of the root and leaf were mapped to the GO database, respectively. All DEGs were classified on the basis of biological process, cell component, and molecular function ( Figure 5). In the roots, 122 DEGs were classified into 17 GO terms, including 6 terms of biological process, among which cellular process (39) and metabolic process (38) contained the most genes; 3 terms of cell component, among which cellular element contained the most genes (65); 8 terms of molecular function, of which catalytic activity (54) and binding (40) contained the most genes ( Figure 5A). In leaves, 123 DEGs were classified into 19 GO terms, including 11 terms of biological process, among which cellular process (31) and metabolic process (27) contained the most genes; three terms of cell component, among which cellular element (57) contained the most genes; 5 terms of molecular function, with catalytic activity (49) and binding (37) containing the most genes ( Figure 5B). By further analysis, we found that there were 14 shared GO terms in the root and leaf, accounting for 82.4% and 72.7% of the total number of GO terms in the root and leaf, respectively. Taken together, these results indicated that a large number of genes that relate to biological process, cell component, and molecular function are activated after salt stress treatment.
Furthermore, we performed KEGG enrichment to analyze the functions of shared DEGs of the root and leaf from metabolic pathways, and 20 top enriched metabolic pathways were selected ( Figure 6). In roots, the main enriched pathways were phenylpropanoid biosynthesis, photosynthesis, galactose metabolism, photosynthesis-antenna proteins, pyruvate metabolism, citrate cycle (TCA cycle), glycolysis/gluconeogenesis, and biotin metabolism, etc., and phenylpropanoid biosynthesis pathway contained the most genes (9) ( Figure 6A). In leaves, the main enriched pathways were plant-pathogen inter- Furthermore, we performed KEGG enrichment to analyze the functions of shared DEGs of the root and leaf from metabolic pathways, and 20 top enriched metabolic pathways were selected ( Figure 6). In roots, the main enriched pathways were phenylpropanoid biosynthesis, photosynthesis, galactose metabolism, photosynthesis-antenna proteins, pyruvate metabolism, citrate cycle (TCA cycle), glycolysis/gluconeogenesis, and biotin metabolism, etc., and phenylpropanoid biosynthesis pathway contained the most genes (9) ( Figure 6A). In leaves, the main enriched pathways were plant-pathogen interaction, glyoxylate and dicarboxylate metabolism, sesquiterpenoid and triterpenoid biosynthesis, pyruvate metabolism, citrate cycle (TCA cycle), galactose metabolism, glycolysis/gluconeogenesis, and biotin metabolism, etc. ( Figure 6B). Interestingly, we found that there were five shared pathways in the root and leaf KEGG pathway enrichment, including pyruvate metabolism (ko00620), citrate cycle (TCA cycle, ko00020), galactose metabolism (ko00052), glycolysis/gluconeogenesis (ko00010), and biotin metabolism (ko00780), and most of the genes contained in the pathways were related to sugar or energy (ATP or NADP + /NADPH) ( Table 1). These results suggest that the metabolic process of watermelon seedlings under salt stress is complex and induces the expression of ATP-or NADPH-related genes.
there were five shared pathways in the root and leaf KEGG pathway enrichment, including pyruvate metabolism (ko00620), citrate cycle (TCA cycle, ko00020), galactose metabolism (ko00052), glycolysis/gluconeogenesis (ko00010), and biotin metabolism (ko00780), and most of the genes contained in the pathways were related to sugar or energy (ATP or NADP + /NADPH) ( Table 1). These results suggest that the metabolic process of watermelon seedlings under salt stress is complex and induces the expression of ATP-or NADPH-related genes.

Transcription Factors Respond to Salt Stress
Transcription factors regulate various biological processes by directly targeting downstream genes. Many transcription factors were identified from DEGs in the root and leaf under salt stress (Figure 7). In roots, 79 transcription factors were found to be differentially expressed in the four comparison groups, among which 5 were shared ( Figure 7A). In leaves, 88 transcription factors were found to be differentially expressed in the four comparison groups, of which 10 were shared ( Figure 7B). We counted the subfamilies of these shared transcription factors according to the annotations of each gene ( Figure 7C,D). Results showed that the five transcription factors in roots were divided into four subfamilies including 1 NAC (Cla97C08G155640), 1 AP2-EREBP (Cla97C08G155430), 1 MYB (Cla97C08G149140), and 2 TCP (Cla97C03G057940, Cla97C09G174040) ( Figure 7C); among them, the expression levels of Cla97C08G149140 and Cla97C08G155640 were significantly up-regulated by salt stress and were significantly up-regulated in the salttolerant variety ( Figure 7E). Ten transcription factors in leaves were divided into seven subfamilies including 1 AP2-EREBP (Cla97C11G208660), 1 OFP (Cla97C10G201530), 1 TCP (Cla97C09G174040), 1 bZIP (Cla97C08G155680), 1 bHLH (Cla97C06G112140), 2 MYB (Cla97C01G012490, Cla97C05G106570), and 3 WRKY (Cla97C06G113910, Cla97C07G133580, Cla97C10G206240) ( Figure 7D); among them, the expression levels of Cla97C05G106570, Cla97C09G174040, Cla97C06G113910, and Cla97C08G155680 were significantly up-regulated by salt stress, of which Cla97C05G106570 and Cla97C09G174040 were significantly upregulated in the salt-tolerant variety ( Figure 7F). It was worth noting that in these subfamilies, one TCP transcription factor (Cla97C09G174040) was shared in roots and leaves ( Figure 7G), indicating that TCP transcription factor significantly responds to salt stress. These results suggested that transcription factors, especially TCP, play a crucial role in the responses to salt stress of watermelon.

Key DEGs Related to Salt Stress in Watermelon
In order to further identify the key genes in response to salt stress in watermelon, we analyzed the shared DEGs through all the comparison groups ( Figure 9). The results showed that there were only three shared DEGs between the root and leaf ( Figure 9A), including one TCP transcription factor (Cla97C09G174040), one bromodomain-containing protein (Cla97C01G024070), and one SCO1-like protein (Cla97C04G073570). Among the three DEGs, both Cla97C01G024070 and Cla97C04G073570 were up-regulated in all eight comparison groups, indicating that they were induced by salt stress, but the expression levels of the salt-tolerant variety were lower than that of the salt-sensitive variety; however, for Cla97C09G174040, it was significantly up-regulated by salt stress in the leaf, and the expression levels of the salt-tolerant variety were higher than that of salt-sensitive variety both in the root and leaf ( Figure 9B), suggesting that TCP transcription factors play a crucial role in salt tolerance of watermelon. Then, we performed the co-expression network to explore the correlation between transcription factors and salt-tolerant related genes, and the selected correlation coefficient was all greater than 0.9. In roots, the co-expression network contained five transcription factors and 10 DEGs, including one HSP (Novel_G000299), one potassium-related gene (Cla97C06G123400), two cell wall-related genes (Cla97C07G138170, Cla97C09G182830), two calcium-related genes (Cla97C08G159250, Cla97C08G159420), and four lignin-related genes (Cla97C02G047340, Cla97C08G151810, Cla97C08G161100, Cla97C11G214540), re-  Table S4).

Key DEGs Related to Salt Stress in Watermelon
In order to further identify the key genes in response to salt stress in watermelon, we analyzed the shared DEGs through all the comparison groups ( Figure 9). The results showed that there were only three shared DEGs between the root and leaf ( Figure 9A), including one TCP transcription factor (Cla97C09G174040), one bromodomain-containing protein (Cla97C01G024070), and one SCO1-like protein (Cla97C04G073570). Among the three DEGs, both Cla97C01G024070 and Cla97C04G073570 were up-regulated in all eight comparison groups, indicating that they were induced by salt stress, but the expression levels of the salt-tolerant variety were lower than that of the salt-sensitive variety; however, for Cla97C09G174040, it was significantly up-regulated by salt stress in the leaf, and the expression levels of the salt-tolerant variety were higher than that of salt-sensitive variety both in the root and leaf ( Figure 9B), suggesting that TCP transcription factors play a crucial role in salt tolerance of watermelon.

Discussion
Biomass is a key indicator for plant growth under salt stress [6]. Overall, salt stress reduces plant biomass, but the difference in the degree of decline was determined by species. Biomass of tomato [16] and sunflower [17] decreased significantly under 50 mmol·L −1

Discussion
Biomass is a key indicator for plant growth under salt stress [6]. Overall, salt stress reduces plant biomass, but the difference in the degree of decline was determined by species. Biomass of tomato [16] and sunflower [17] decreased significantly under 50 mmol·L −1 NaCl treatment, while the biomass of wheat, rice, and maize decreased significantly under 100-150 mmol·L −1 NaCl treatment [1]. However, biomass of halophyte Salicornia europaea began to decrease when NaCl concentration was higher than 400 mmol·L −1 [18]. In addition, Han [12] found that biomass, growth potential, and photosynthesis of watermelon seedlings decreased significantly under salt stress. In this study, the biomass of salt-tolerant and salt-sensitive watermelon seedlings were both significantly reduced, and the decrease degree of salt-tolerant watermelon varieties was less than that of salt-sensitive watermelon varieties, which decreased by 43.0% and 47.7%, respectively ( Figure 1); the results were similar to previous studies. Na + or Cl − concentration, K + content, and Na + /K + ratio in leaves and roots under salt stress are often used to reflect the degree of salt stress in plants [6]. Han et al. [19] proved that Na + accumulated in the roots and stems of watermelon seedlings, Cl − accumulated in roots, and K + content decreased significantly in stems. Guo et al. [20] found that wild watermelon had strong salt tolerance, possibly because its root had a stronger ability to intercept Na + and transport K + . In this study, salt stress significantly reduced K + content and K + /Na + ratio in roots and leaves, and increased Na + and Cl − accumulation. We also found that Cl − and Na + accumulated more in leaves, and K + accumulated in roots (Figure 1).
MDA is one of the main products of peroxidation of membrane-lipid, and also an important indicator to reflect the degree of plasma membrane damage. MDA content is an important index reflecting the damage degree of plasma membrane. Previous studies showed that the content of MDA in Carex and soybean increased significantly after salt stress treatment [21,22]. In addition, studies showed that with the increase of salt concentration, the cell membrane structure of watermelon seedlings was damaged, the MDA content and relative permeability of the plasma membrane were significantly increased, the metabolism of ROS was disordered, and the O −.
2 production rate of seedling leaves was significantly increased [21,23]. In our data, the MDA content also increased significantly after salt stress treatment both in leaves and roots, and the increased degree of salt-tolerant watermelon varieties was less than that of salt-sensitive watermelon varieties (Figure 1).
Under salt stress, various plant hormones respond to salt stress. For example, the reduction of plant growth and development under stress may be the result of changes in plant auxin accumulation and redistribution [24]. The primary expression genes induced by auxin were GH3 (gretchen hagen 3), Aux/IAA, and SAUR (small auxin up RNA) [25]. GH3 family encodes a class of enzymes that catalyze the coupling of auxin and amino acids, which can regulate plant growth and development to be adapted to external changes. GH3 can interact with ARF (auxin response factor) to induce the expression of downstream genes [26]. In this study, the expression of GH3 (Cla97C01G016350) increased in roots, which ensured the growth of watermelon seedlings under salt stress. Furthermore, AUX/IAA played a central role in the regulation of auxin response and inhibited the expression of auxin-related genes by binding to ARF as an inhibitor of auxin-induced genes [27]. However, Cla97C06G119080 (IAA16) was down-regulated, which indicated that it maintains the growth of watermelon seedlings under salt stress. Under high salt stress, the gibberellins (GA 1 and GA 4 ) content decreased, and the DELLA protein content increased, which enhanced tolerance to salt stress by inhibiting plant growth [28]. In our study, one GA 4related gene (Cla97C07G133720) was significantly down-regulated under salt stress, which may be the reason for the inhibition of watermelon seedling growth and adaptation to salt stress. Furthermore, GA can stimulate the H + -ATPase activity of the vacuole membrane and form the electrochemical gradient of H + across the vacuole membrane, which provides a driving force for the secondary active transport of various solute molecules (anions, cations, amino acids, and carbohydrates) across the vacuole membrane and increases the germination rate of seeds under salt stress [29]. In addition, the Na + /H + reverse transporter on the plasma membrane can use ATP on the plasma membrane to expel Na + from the cells. The sodium hydrogen pump on the vacuole membrane uses the energy provided by ATP on the vacuole membrane to compartmentalize the Na + region into the vacuole, thereby reducing the impact of excessive Na + accumulation on the cytoplasm [30]. In wheat, the high-density thylakoids of wheat under salt stress may produce more ATP and NADPH, thereby increasing the energy supply and improving the response rate of wheat to salt stress [31]. It was worth noting that many KEGG pathways enriched by DEGs in roots and leaves are related to energy metabolism (Table 1). For example, we screened two genes that encode alcohol dehydrogenase; one (Cla97C02G031490) of them was down-regulated and another (Cla97C02G031490) was up-regulated under salt stress. A previous study proved that overexpression of Sysr1 (a member of the alcohol dehydrogenase superfamily) in tobacco induced stress-related gene expression and increased tolerance to salt stress [32]. In this study, there was one down-regulated gene (Cla97C05G107020) that encodes pyruvate dehydrogenase E1 under NaCl stress, and in A.thaliana, mutation of IAR4 (a member of the pyruvate dehydrogenase E1 superfamily) led to an accumulation of greater Na + and exhibited a greater Na + /K + ratio under NaCl treatment by reducing SOS1 and SOS3 expression, resulting in increasing sensitivity to salt stress [33]. We also found two genes (Cla97C10G192070 and Cla97C11G220850) that encode phosphoenolpyruvate carboxykinase (PEPCK), and both of them were up-regulated under NaCl stress. In rice, the expression of PEPCK was down-regulated in early stages and tended to respond to salt stress during maturity under NaCl stress [34]. These results indicated that energy might be required in the Na + efflux process, and watermelon seedlings respond to salt stress by changing the expression of these genes.
The cell wall is the primary line of defense for plants to resist salt stress [35]. When plant roots are subjected to external salt stress, the cell wall sensing system takes the lead to receive salt stress signals [36,37]. Cellulose is the main component of the cell wall, which is synthesized by cellulose synthases (CESAs) guided by microtubules on the plasma membrane [38]. A previous study found that the content of plant cellulose decreased under salt stress [39]. In roots, we found two genes encoding CESA, of which Cla97C07G138170 was down-regulated under salt stress, whereas Cla97C09G182830 was up-regulated, suggesting that these CESAs have different expression patterns in response to salt stress. Many ion transporters related to salt tolerance have been discovered in recent twenty years. Among them, HKT plays a key role in K + /Na + absorption, long-distance transport, and redistribution [40,41]. Research on barley showed that Na + accumulated in roots and leaves when HvHKT1:1 was knocked out, while overexpression of HvHKT1:1 in salt-sensitive Arabidopsis mutant hkt1-4 and sos1-12 caused Na + content to decrease significantly in roots and shoots, indicating that HvHKT1:1 plays an important role in Na + transport in roots [42]. Recent studies have found that HvHKT1 significantly reduced Na + transport from roots to shoots and increased K + /Na + ratio compared with wild type [43]. Kader et al. [44] found that OsHKT1 and OsHKT2 were induced under salt stress, and Na + was reduced by regulating the Na + /K + ratio. In this study, we found one HTK (Cla97C06G123400) in roots, which was down-regulated under salt stress, but its expression in the salt-tolerant variety was significantly higher than that in salt-sensitive variety, indicating that HTK plays a key role in Na + transport in roots of salt-tolerant variety. CYP450s are a class of monooxygenases encoded by the supergene family, which mainly exist in animals and plants and microorganisms, and play an important role in biological defense [45]. For example, secondary metabolites produced by amino acids catalyzed by CYP71A are related to plant defense and stress resistance [46]. In leaves, there were three genes encoding CYP450 (Cla97C03G054270, Cla97C10G198010, and Cla97C11G219620), among which Cla97C03G054270 belonged to CYP71A1 and was markedly up-regulated under salt stress. Lignin is the second most abundant biopolymer in plants, which mainly exists in the secondary wall of vessel molecules [47]. Under salt stress, the degree of lignification of the plant root cell wall will increase, which can not only effectively prevent the intracellular ion absorption, but also enhance the structural rigidity and firmness of the conduction tissue and improve the salt tolerance of plants. Salt stress can change many enzymes involved in lignin biosynthesis, generally by changing their expression patterns, thereby regulating lignin synthesis and responding to salt stress [48]. For example, the expression of tomato peroxidase gene TPX1 was up-regulated under salt treatment, and the lignin content in transgenic tomato leaves with high expression of TPX1 increased, indicating that TPX1 may be involved in the lignification of root and aboveground parts [49]. In roots, we found nine lignin biosynthesis-related DEGs, five of which were peroxidase genes (PERs), suggesting that they were involved in the response of watermelon seedlings to salt stress.
Under salt stress, transcription factors regulate their expression levels by changing the expression of different genes, such as bZIP, WRKY, AP2/ERF, MYB, bHLH, NAC, etc. In soybean, overexpression of GmbZIP2 in hairy roots enhanced the expression of stress response genes GmMYB48, GmWD40, GmDHN15, GmGST1, and GmLEA [50]. Transcription factors bHLH and WRKY enhanced salt tolerance of Arabidopsis by regulating the expression of salt response gene AtKUP2 [51]. AtMYB20 in Arabidopsis influenced plant resistance to salt stress by regulating the ABA signaling pathway. Salt stress and ABA stress induced AtMYB20 gene expression. Overexpression of AtMYB20 gene can enhance the resistance of transgenic Arabidopsis to salt stress and reduce the expression levels of ABA signal negative regulatory genes ABI1, ABI2, and AtPP2CA under salt stress [52]. In addition, overexpression of SlMYB102 in tomato increased the K + /Na + ratio and the activity of active oxygen scavenging enzyme, as well as the transcription abundance of salt stress-related genes in the two overexpression lines was up-regulated [53]. Ju et al. [54] reported that overexpression of VvNAC17 in Arabidopsis enhanced drought and salt tolerance and upregulated the expression of ABA and stress-related genes ABI5, AREB1, COR15A, COR47, and P5CS. In roots and leaves, we screened five and ten differentially expressed TFs, respectively ( Figure 7C), and found a strong correlation between them and salt tolerancerelated genes, most of which were greater than 0.9 ( Figure 8C,D), indicating that these TFs may improve salt tolerance of watermelon seedlings by regulating the expression of corresponding genes. Among the TFs, TCP (Cla97C09G174040, TCP10, Supplementary Figure S2) was most noteworthy, which was found in roots and leaves. Although it had different expression patterns in roots and leaves, its expression levels in the salt-tolerant variety were significantly higher than those in the salt-sensitive variety, indicating that it plays a very important role in the salt tolerance of watermelon seedlings.
TCP is a plant-specific transcription factor, which belongs to bHLH TF. TCP activates or inhibits gene expression through its bHLH domain interacting with other proteins [1]. For example, TCP4 binds to the promoter region of functional genes and regulates their expression, thus playing a key role in multiple growth and development processes [55]. In Phyllostachys edulis, PeTCP10 was induced by drought and ABA, and PeTCP10 directly targeted stress-/ABA-responsive gene BT2 to improve drought resistance [56]. In Betula platyphylla, BpTCP10 was involved in the response to salt stress, and the transgenic Betula platyphylla lines showed salt sensitivity after inhibiting the expression of BpTCP10 [57]. Micro RNAs are non-coding single-stranded RNAs that widely exist in the biological world and regulate the expression of their target genes to change the growth and development process of plants. In Arabidopsis, microRNA319 (miR319) has been demonstrated to target TCP genes to influence their growth and development. MiR319-TCP4 bound to the cis-acting element of YUCCA5, directly activated its transcription and expression, and promotes cell elongation of Arabidopsis hypocotyl [1]. In Agrostis stolonifem, the target gene AsPCF5/6/8/14 in miR319a overexpression plants was down-regulated, and the drought resistance and salt tolerance of transgenic lines were enhanced [58]. Overexpression of miR319 gene in Solanum habrochaites can also improve the drought resistance and high temperature resistance of tomato [59]. In addition, the overexpression of miR319 in switchgrass can promote ethylene synthesis and enhance the ability to resist salt stress [60]. In watermelon, ClTCP10 and AtTCP10 have high homology, while AtTCP10 is regulated by miR319 [61], suggesting that ClTCP10 may be the target gene of miR319, which has an extremely important regulatory effect on the development of watermelon, leading to improvement of the salt tolerance of watermelon.

Conclusions
In conclusion, this study found that salt stress significantly reduced K + content and K + /Na + ratio in roots and leaves, and increased Na + , Cl − accumulation, and MDA content in watermelon seedlings of salt-tolerant variety and salt-sensitive variety. Furthermore, based on our results, we proposed a hypothetical model describing the molecular mechanism of watermelon seedling response to salt stress ( Figure 10). We suggest that the expression of some salt tolerance-related DEGs, such as plant hormone signal transduction pathway related genes, K + /Na + /Ca 2+ -related genes, and lignin biosynthesis-related genes, promoted the salt tolerance of watermelon. In addition, some DEGs related to sugar or energy pathways play a major role in ion transport. Moreover, transcription factors such as AP2-EREBP, bZIP, bHLH, MYB, NAC, OFP, TCP, and WRKY may also be associated with salt tolerance of watermelon by regulating above DEGs, especially a TCP TF (Cla97C09G174040).
Life 2022, 12, x FOR PEER REVIEW 19 of 22 Figure 10. A putative schematic model of salt tolerance mechanism in watermelon seedlings.

Supplementary Materials:
The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Phylogenetic tree of TCPs; Figure S2: Phylogenetic tree of TCPs; Table S1: Primers for q-PCR.; Table S2: Basic information about transcriptomic data; Table S3: Correlation coefficient between TFs and DEGs in roots; Table S4: Correlation coefficient between TFs and DEGs in leaves; Table S5: Correlation coefficient between TFs and DEGs in leaves. Institutional Review Board Statement: Not applicable. Figure 10. A putative schematic model of salt tolerance mechanism in watermelon seedlings. The ↑/↓ resprents the gene is up-regulated or down-regulated, respectively.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/life12071033/s1, Figure S1: Phylogenetic tree of TCPs; Figure S2: Phylogenetic tree of TCPs; Table S1: Primers for q-PCR.; Table S2: Basic information about transcriptomic data; Table S3: Correlation coefficient between TFs and DEGs in roots; Table S4: Correlation coefficient between TFs and DEGs in leaves; Table S5: Correlation coefficient between TFs and DEGs in leaves.