Identification and Expression Analysis of Snf2 Family Proteins in Tomato (Solanum lycopersicum)

As part of chromatin-remodeling complexes (CRCs), sucrose nonfermenting 2 (Snf2) family proteins alter chromatin structure and nucleosome position by utilizing the energy of ATP, which allows other regulatory proteins to access DNA. Plant genomes encode a large number of Snf2 proteins, and some of them have been shown to be the key regulators at different developmental stages in Arabidopsis. Yet, little is known about the functions of Snf2 proteins in tomato (Solanum lycopersicum). In this study, 45 Snf2s were identified by the homologous search using representative sequences from yeast (S. cerevisiae), fruit fly (D. melanogaster), and Arabidopsis (A. thaliana) against the tomato genome annotation dataset. Tomato Snf2 proteins (also named SlCHRs) could be clustered into 6 groups and distributed on 11 chromosomes. All SlCHRs contained a helicase-C domain with about 80 amino acid residues and a SNF2-N domain with more variable amino acid residues. In addition, other conserved motifs were also identified in SlCHRs by using the MEME program. Expression profile analysis indicated that tomato Snf2 family genes displayed a wide range of expressions in different tissues and some of them were regulated by the environmental stimuli such as salicylic acid, abscisic acid, salt, and cold. Taken together, these results provide insights into the functions of SlCHRs in tomato.


Introduction
In eukaryotes, about 147 bp of DNA wrapping around a histone octamer forms nucleosome, the fundamental unit of chromatin. The reversible changes in chromatin structure alter the stability of the nucleosome, thereby facilitating regulatory factors access, such as transcription factor [1,2]. Thus, the precise chromatin structure is essential for the correct spatial and temporal gene expression in the eukaryotes [3,4]. The changes in chromatin involve histone modifications, DNA methylation, histone variants, and chromatin remodeling. Many proteins have been identified to mediate these processes, among which the Snf2 family proteins can affect gene expression by using the energy of ATP hydrolysis to alter the interactions between histones and DNA [5]. Indeed, most Snf2 proteins associated with other chromatin remodelers form large multisubunit complexes called chromatin-remodeling complexes, which most likely alter the activity of the core ATPase in vivo. The accessory subunits commonly contain additional domains that may affect the enzymatic activity of the complex, facilitate its binding to other proteins, and target the complex to DNA and/or modified histones [6]. The chromatin-remodeling complexes are conserved throughout eukaryotes with essential roles in many aspects of chromatin biology.
Based on the different protein compositions and functions, the chromatin-remodeling complexes can be divided into SWI/SNF, ISWI (imitation switch), INO80 (inositol requiring 80), and CHD (chromodomain, helicase, and DNA binding) groups [7]. The SWI/SNF complexes alter the position of the nucleosome at promoters, which can regulate transcription either positively or negatively [8]. The ISWI group complexes were essential for chromatin assembly and the formation of nucleosome arrays with well-ordered spacing, which might help to promote repression [9]. In yeast, the ino80 mutants are defects in homologous recombination during DNA repair, indicating that the INO80 complexes are involved in DNA repair [10]. Indeed, the INO80 complexes could be recruited to doublestranded breaks (DSBs) via direct binding of the complex subunits to phosphorylated H2AX or γ-H2AX [11,12], which facilitates nucleosome eviction at DSBs, allowing the recruitment of repair factors. In comparison, the CHD complexes have diverse functions. For instance, CHD1 is targeted to sites of active transcription through PHDmediated recognition of H3K4me3 [13,14] and associates with other preinitiation factors to facilitate transcriptional elongation and splicing [15]. In addition, CHD3 and CHD4 are incorporated into a large protein complex with histone deacetylases to repress transcription by binding to methylated DNA in an MBD2/3-dependent manner, remodeling the surrounding chromatin, and removing active histone marks [16,17].
SWI2/SNF2, the first Snf2 protein, was identified from Saccharomyces cerevisiae by the examination of mating type switching (SWI) and sucrose nonfermenting (SNF) mutants [18]. Further data indicated that the SWI2/SNF2 gene is homologous to a number of other ATP-binding helicases of the DEAD/H family [19]. The sequence similarity includes the catalytic ATPase domain and seven characteristic protein motifs [20]. Moreover, the conserved domain analysis indicated that the helicase-like region can be further divided into two domains: the SNF2-N and helicase-C domains [21,22]. Based on the helicase-like region, the Snf2 family proteins are grouped into six clades, including the Snf2-like, SWI/SN-F-related protein-like (Swr1-like), Rad54-like, Rad5/16-like, SSO1653-like, and Distant family [21].
Arabidopsis contains 41 Snf2 family proteins that fall into 18 subfamilies [22]. The function of Arabidopsis BRAHMA (BRM) and SPLAYED (SYD), the closest homologs of yeast and animal SWI2/SNF2 ATPase subunits (Snf2 subfamily), has been investigated. Mutations of SYD cause defects of the shoot apical meristem (SAM). Furthermore, SYD physically interacts with the promoter of WUSCHEL (WUS), a central regulator in SAM [23]. The expression profile showed that BRM was mainly expressed in the active cell division tissues, such as meristems and organ primordia [24]. The BRM mutants displayed multiple developmental defects, such as reduced plant size and root length [24,25], downward curling leaves [25], more sensitivity to abscisic acid (ABA) [26], and early flowering [27]. SYD and BRM were shown to interact with LEAFY and SEPALLATA3 proteins, which are essential for floral organ identity [28]. Indeed, the functions of BRM to modulate gene transcription are always through association with other nuclear proteins. For example, the plant-unique H3K27 demethylase, RELATIVE OF EARLY FLOWERING 6 (REF6), recruits BRM to its target genomic loci containing a CTCTGYTY motif [29]. Moreover, FORGETTER1 (FGT1), which is specifically required for the heat stress memory coactivator, maintains its target loci in an open and transcription-competent state by interacting with BRM near the transcriptional start site [30]. BRM also interacts with other transcription factors such as TEOSINTE BRANCHED1 CYCLOIDEA AND PCF-CODING GENE (TCP4), ANGUSTIFOLIA3 (AN3), and BREVIPEDICELLUS (BP) to regulate gene expression involved in leaf development and inflorescence architecture [31,32]. A recent report showed that BRM also interacts with PHY-INTERACTING FACTOR 1 (PIF1) to modulate PROTOCHLOROPHYLLIDE OXIDOREDUCTASE C (PORC) expression, which is essential for chlorophyll biosynthesis during the transition from heterotrophic to autotrophic growth [33]. Meanwhile, SUMOylation of BRM caused by METHYL METHANE SULFONATE SENSITIVITY 21 (MMS21) increases the BRM stability in root development [34]. Interestingly, more recent data demonstrated that microRNA precursors (pri-miRNAs) are the substrates of BRM. As a partner of the microprocessor component SER-RATE (SE), BRM accesses pri-miRNAs through SE and remodels their secondary structures, which prevents further downstream processing mediated by DCL1 and HYL1 [35].
Transgenic Arabidopsis plants overexpressing AtCHR12, a member of the Snf2 subfamily, exhibit growth arrest of primary buds and growth reduction of the primary stem under drought and heat stress [36]. Moreover, a Rad54-like family member, DEFECTIVE IN RNA-DIRECTED DNA METHYLATION1 (DRD1), and a member of Snf2-like protein, DECREASED DNA METHYLATION 1 (DDM1), are involved in DNA methylation [37,38]. In addition, DRD1 and DDM1 are also involved in leaf senescence, since drd1 and ddm1 mutants exhibit a delayed leaf senescence phenotype [39]. Furthermore, PHOTOPERIOD-INDEPENDENT EARLY FLOWERING 1 (PIE1), a Swr1 subfamily member, known to deposit histone H2A.Z, is also important for flowering and plant development [40,41], while the Mi-2 subfamily member PICKLE is a key regulator in brassinosteroid (BR), gibberellin (GA), and cytokinin (CK) signaling [42,43].
Compared with Arabidopsis, little is known about Snf2 proteins in other plant species. In rice, OsDDM1a and OsDDM1b, two genes homologous to Arabidopsis DDM1, are involved in DNA methylation [44], while rice CHR729, a member of the CHD3 family, plays an important role in seedling development via the GA signaling pathway [45]. A previous study has also analyzed the DRD1 and Snf2 subfamilies in tomato, which were reported to be involved in stress responses [46]. In addition, constitutively overexpressing a Snf2 gene (termed as SlCHR1, Solyc01g079690) caused reducing growth of transgenic tomato plants (cv. Micro-Tom) [47]. However, the largest and most diverse gene family, the Snf2 gene family, has not been systematically analyzed in the tomato genome. In this study, we identified and characterized 45 Snf2 family proteins from tomato. The expression profiles of the tomato Snf2 genes were also analyzed. The results provide a wealth of information for further exploring the developmental function of Snf2 family proteins in tomato, especially during fruit development.

Plant Materials and Growth Conditions.
In this study, the Solanum lycopersicum cultivar "Heinz 1706" was used as an experimental material. Surface-sterilized tomato seeds were grown in the Murashige and Skoog (MS) medium with 1.5% sucrose and 0.8% agar for 14 days in a controlled environment greenhouse with a long photoperiod (16 h light/8 h dark) at 23 ± 1°C.

Chromosome Location and Sequence Feature Analyses.
Chromosome location of SlCHR genes was determined by BLAST analysis of SlCHRs against SGN (http://solgenomics .net/organism/Solanum_lycopersicum/genome). The program DnaSP was used to carry out synonymous substitution (Ks) values of paralogous gene pairs [49]. The Compute pI/Mw tool on the ExPASy server (http://web.expasy.org/ compute_pi/) was used to predicted molecular weight (Mw) and theoretical isoelectric point (pI) of SlCHRs. The structures of SlCHR genes were predicted using the Gene Structure Display Server (http://gsds.cbi.pku.edu.cn/) [50].

Expression Data Visualization.
The expression data of tomato SlCHRs were extracted from publicly available RNA-seq datasets from the Tomato Genome Consortium [52] and visualized with Matrix2PNG (http://www.chibi .ubc.ca/matrix2png/bin/matrix2png.cgi) [53]. The RNA-seq data were obtained from transcriptome sequencing using three-week-old sand-grown seedlings, roots, leaves, buds (unopened flower buds), and flowers (fully open flowers) as well as fruits (at 1 cM, 2 cM, and 3 cM), MG (mature green), breaker (B, early ripening), and 10-day post-B (B10, red ripe) stages of tomato "Heinz 1706." 2.6. Gene Expression Analyses. For hormone and salt stress response test, 2-week-old tomato "Heinz 1706" seedlings grown in the MS medium were transferred to the liquid MS medium containing SA (1 mM), ABA (50 μM), and NaCl (200 mM) for 4 h, respectively. For cold stress test, the planes were transferred to a 4°C growth cabinet for 4 h. Total RNA from treated seedlings was extracted with TRIzol reagent (Invitrogen) according to the manufacturer's protocol and used to synthesize cDNA. Real-time PCR was performed with iTaq™ Universal SYBR® Green Supermix (Bio-Rad) using ABI 7500 Fast Real-Time PCR System. The gene-specific primers for real-time PCR were designed by Primer 3.0 [36] and listed in Supplemental Table 2. Tomato Actin (Solyc03g078400) was served as an internal control.

Identification of Snf2 Family Proteins in Tomato.
To uncover the complete family of genes for encoding Snf2 proteins in the tomato genome, iterative BLASTP researches using representative sequences from yeast (S. cerevisiae), fruit fly (D. melanogaster), and Arabidopsis (A. thaliana) were conducted against SGN (http://solgenomics.net/organism/ Solanum_lycopersicum/genome, ITAG3.20) genome annotation database. In total, 45 nonredundant putative Snf2 genes were identified in the tomato genome (Table 1).
According to the current used nomenclature in Arabidopsis and rice, we designated Snf2 proteins of tomato (Solanum lycopersicum) as SlCHRs. All of the deduced SlCHR proteins contained the conserved SNF2-N domain and helicase-C domain. The theoretical isoelectric point (pI) of SlCHR candidates ranged from 5.13 to 9.42, and the length of SlCHRs varied from 391 to 2500 amino acids. The molecular weight (Mw) and the number of introns varied from 44.3 to 274.2 kDa and 1 to 37, respectively (Supplemental Table 1). Mapping SlCHRs to the tomato genome showed that 45 SlCHRs were unevenly distributed on 11 chromosomes (except for chromosome 10). Among them, there were 9 SlCHRs on Chr1; 5 on each of Chr2 and Chr4; 3 on each of Chr6, Chr11, and Chr12; 4 on each of Chr3 and Chr7; 2 on Chr5; 6 on Chr8; and one on Chr9, respectively ( Figure 1). Most SlCHRs were located in the bottom regions of tomato chromosomes, and few were in the central regions of chromosomes ( Figure 1).
Moreover, 8 pairs of SlCHRs (Ks < 1 0) were evolved from intrachromosomal duplication (Supplemental Figure 1 and Table 2), indicating the importance of gene duplication for SlCHR gene expansion.

Comparative Analysis Gene Structures of SlCHR and
AtCHR. Gene structure analysis of 45 SlCHR genes displayed that the number of introns varied from 1 (SlCHR21, SlCHR4, SlCHR5, and SlCHR21) to 37 (SlCHR41) (Figure 3 and Supplemental Table 1). By contrast, the intron number of 41 AtCHRs varied between 2 and 33 (Supplemental Table 1 and Supplemental Figure 3). The length of introns also varied significantly among the SlCHR subfamily including Snf2-like, Swr1-like, and Rad5/16-like genes ( Figure 3). Interestingly, the distribution of intron phases in SlCHRs was very similar to AtCHRs (Supplemental Figure 4). Next, we compared the internal exons and introns of SlCHRs with those of AtCHRs. The results showed that the exons of SlCHRs varied 18 to 3067 bp with the average of 215 bp, which was smaller than the average length of AtCHR exons (263 bp). Interestingly, most CHRs (about 86% of SlCHR and 83% of AtCHR) had an exon with a size below 300 bp (Figure 4(a)), while 56% of SlCHR exons and 53% of AtCHR exons were between 60 and 160 bp (Figure 4(b)).
Although the size distribution of SlCHR exons was similar to AtCHR exons, the size distribution of intron was more variable, ranging from 34 bp to 9.0 kb. There were 54 SlCHR introns (9.5%) with sizes > 1 5 kb; however, no such introns existed in AtCHRs (Figure 4(c)). About 61% of SlCHR and 89% of AtCHR introns had sizes below 300 bp, while 56% of SlCHR introns were between 60 and 160 bp and 53% of AtCHR introns were between 80 and 120 bp, respectively (Figure 4(c)). Meanwhile, the average sizes of SlCHR introns and AtCHRs were 595 bp and 153 bp, respectively. These results indicated that the exon and intron size distribution was different between SlCHRs and AtCHRs.

The Conserved Motifs in SlCHRs.
To investigate the conserved domains of SlCHRs, Pfam (http://pfam.xfam.org/) and Conserved Domain Database (CDD, http://www.ncbi .nlm.nih.gov/Structure/cdd/wrpsb.cgi) programs were used. The results showed that all the 45 SlCHRs contained a helicase-C domain with about 80 amino acid residues and a SNF2-N domain with more variable amino acid residues ( Figure 5).
Unlike the human Snf2 subfamily proteins hBRG1and hBRM, the conserved HSA (helicase-SANT-associated) domain was not found in all three Snf2 subfamily proteins (SlCHR8, SlCHR41, and SlCHR6) and only SlCHR8 contained bromodomain, an acetyl-lysine binding domain ( Figure 5). However, an alignment profile using the HSA domain of humans and the N-termini of SlCHR8, SlCHR41, and SlCHR6 showed that the conserved amino acid residues including E, H, and L were found in tomato Snf2 subfamily proteins (Supplemental Figure 5). Interestingly, the Swr1 subfamily SlCHR17 was highly homologous to Arabidopsis PIE1, containing the HSA domain at the N-terminus ( Figure 5). Furthermore, two members of the Iswi subfamily, SlCHR2 and SlCHR26, had the conserved domains HAND, SANT, and SLIDE located on the Cterminus ( Figure 5). The Mi-2 subfamily proteins, SlCHR27 and SlCHR33, contained two double chromodomains and an additional PHD domain at the N-terminal part of the proteins. All members of the Rad5/16-like family group except SlCHR29 had a RING-finger E3 ubiquitin ligase domain embedded between the SNF2-N and helicase-C domain in the C-terminal regions ( Figure 5). Furthermore, an additional HIRAN domain was found in the N-terminal region of SlCHR42 and SlCHR43 in this group. In general, the HIRAN domain was predicted to recognize features associated with damaged DNA or stalled replication forks, such as ssDNA stretches or DNA lesions [54].
In addition to these conserved domains, other conserved motifs were searched using the MEME program. 20 motifs for 45 SlCHRs were identified ( Table 3). The number of motifs in each SlCHR varied from 5 to 16 (Table 3). Motifs 10, 4, and 1 were actually the helicase-C domain (Supplemental Figure 6) that was found in most of the SlCHRs. In addition to the conserved motifs, several other motifs were also identified in SlCHR proteins, such as motifs 13, 16, 20, 7, 8, and 18 in the DRD1 subfamily as well as motifs 17, 14, 15, and 19 in the Rad5/16-like group (Table 3). Sequence analysis of helicase-C domains

Paralogous genes Ks
AtCHR6-PICKLE identified the conserved acid residues such as Asp, Gly, Arg, Gln, and Lue in the motifs 10, 4, and 1 (Supplemental Figure 6 and Supplemental Figure 7).

Expression Patterns of Tomato Snf2
Family Genes. In order to explore the possible role of tomato snf2s, we analyzed their expression profiles ( Figure 6). SlCHR2, SlCHR26, SlCHR8, and SlCHR41 (Snf2-like family) had similar expression profiles and were expressed mainly in roots and fruits from 1 cM to B stages (Figure 6(a)), suggesting that these genes may play redundant roles in root and fruit development. Swr1-like SlCHRs, SlCHR17, and SlCHR19 were strongly expressed in roots and B+10 stage fruits, while SlCHR7 was mostly expressed in roots (Figure 6(b)). Interestingly, SlCHR10 was expressed in roots and in the early stages of fruit development (Figure 6(b)). Most of the SSO1653-like and Distant SlCHR genes accumulated in the early stages of fruit development and roots (Figures 6(d) and 6(f)). According to the expression profile of Rad54-like and Rad5/16-like SlCHRs, these SlCHRs could be categorized into two groups: high expression and low expression (Figures 6(c) and 6(e)). However, some SlCHR genes showed specific expression peaks. For example, SlCHR9, SlCHR38, and SlCHR40 were strongly expressed in fruits at the 3 cM stage, SlCHR4 and SlCHR5 in buds, while SlCHR28 and SlCHR43 in roots (Figures 6(c) and 6(e)). In contrast, SlCHR30, SlCHR34, and SlCHR35 were not detected in all tissues analyzed ( Figure 6). We further investigated the expression pattern of SlCHR genes responding to environmental stimuli including hormones, salt, and cold by qRT-PCR. All of the genes analyzed were clearly repressed by SA and cold treatment, especially SlCHR27 (Figure 7). Most of the genes analyzed except SlCHR14 were induced by ABA and salt treatments. In particular, SlCHR7 and SlCHR17 were strongly induced by ABA and salt treatment, respectively (Figure 7). These results revealed that these SlCHR genes may be involved in response to different environmental stimuli in tomato.

Discussion
Snf2 family proteins are the catalytic subunit of the ATPase chromatin-remodeling complexes and contain highly conserved SNF2-N (DEAD) and helicase-C (HELICs) domains involved in many aspects of DNA events such as transcription, replication, homologous recombination, and DNA repair [6,7]. In this study, we systematically identified 45 genes encoding Snf2 proteins (SlCHRs) in tomato (Solanum lycopersicum), which are distributed on 11 chromosomes (Table 1 and Figure 1). Eight pairs of SlCHR intrachromosomal duplication were identified, indicating that gene duplication may play an important role in SlCHR gene expansion in tomato (Table 2 and Supplemental Figure 1). Similar results were also reported in other organisms such as human and Arabidopsis [55,56]. The intron phases were similar in SlCHRs and AtCHRs (Supplemental Figure 3), indicating that plant Snf2 genes originate from a common ancestor.
Previously, a number of genes encoding Snf2 proteins have been identified in Arabidopsis [22], rice [57], and tomato [46]. Nevertheless, only the members of DRD1 and Snf2 subfamilies were identified in tomato previously [46]. Consistent with the previous report, 3 members of Snf2, SlCHR8 (Solyc01g079690), SlCHR41 (Solyc11g062010), and SlCHR6 (Solyc01g094800), were identified. In addition, other three members, SlCHR34, SlCHR35, and SlCHR40, belonging to the DRD1 subfamily, were also found ( Table 1). Sequence comparative analysis of tomato SlCHRs and Arabidopsis AtCHRs revealed some conserved features. For example, all deduced CHRs contained the highly conserved helicase-C domain with about 80 amino acid residues ( Figure 5, Supplemental Figure 6 and Supplemental Figure 7). Unlike the members of the human Snf2 subfamily, the three Snf2 subfamily proteins (SlCHR8, SlCHR41, and SlCHR6) in tomato lack the conserved HSA domain ( Figure 5). Nevertheless, like the HSA domain of yeast and human Snf2 proteins, the N-terminal of tomato Snf2 CHRs also has the conserved amino acid residues E, H, and L (Supplemental Figure 5). As the primary binding platform for nuclear actin-related proteins (ARPs) and actin, the HSA domain is important for the activity of chromatin-remodeling ATPases in yeast and animals [58]. Indeed, the ARPs are conserved subunits of the SWI/SNF and INO80 chromatin-remodeling complexes that associate directly with the ATPase via the HSA domain [58]. The bromodomain was first identified in BRM, the Drosophila homolog of SWI2/SNF2, binding acetylated residues on histone tails [59]. Therefore, SlCHR8 may be the ATPase of at least one of the putative SWI/SNF complexes in tomato. Additional domains such as HAS and SANT that facilitate interaction with the other proteins, as well as bromodomain, chromodomain, and PHD domains that modified histones, were also found in SlCHRs ( Figure 5). Previous reports showed that AtCHRs played key roles in a variety of developmental processes in Arabidopsis. For example, the AtCHR2 (BRM) was involved in morphological traits of leaves and roots as well as reproduction [27,28,31,32,60]. The stem cell pool maintenance of the apical meristem was controlled by AtCHR3 (SYD) [23]. The brassinosteroid and gibberellin signaling pathways were regulated by AtCHR6 (PICKLE) during skotomorphogenic growth [42]. Furthermore, AtCHR2 (AtBRM) also acts as a positive regulator in GA biosynthesis, which regulates GA-responsive genes in a DELLA-independent manner [61]. AtCHR13 (PIE) and AtCHR1 (DDM1) are involved in DNA repair and DNA methylation [62,63]. A recent study in tomato showed that constitutively overexpressed SlCHR8 caused significantly shorter roots and hypocotyls with reduced cotyledon size in transgenic tomato plants (cv. Micro-Tom) [47]. In this study, we found that many protein motifs such as motifs 13, 16 20, 7, 8, and 18 in the DRD1 subfamily and motifs 17, 14, 15, and 19 in the Rad5/16-like family are unique to or mainly exist in one group of SlCHRs (Table 3), indicating that the same group SlCHRs may play similar roles as their Arabidopsis counterparts. The expression profiles of SlCHRs indicated that some SlCHRs may play different roles compared with their homologs in Arabidopsis. For instance, SlCHR8 was mainly expressed in roots and fruits (Figure 6(a)), indicating that it may function in root and fruit development, which is consistent with the report that overexpression of SlCHR8 in tomato resulted in considerably compacter growth including significantly shorter roots and hypocotyls as well as reduced cotyledon and fruit size [47]. In contrast, its Arabidopsis homolog BRM (AtCHR2) functions in leaf and flower development [27,31,60]. Furthermore, functional divergence was observed between SlCHR41 and its homolog AtCHR3 (SYD), since SlCHR41 is poorly expressed in flowers while AtCHR3 is highly  In addition, we found that some SlCHRs respond to environmental stimuli. For instance, the expression of most SlCHRs is repressed by SA but enhanced by ABA  SICHRs are involved in tomato development and hormone signaling pathways.

Conclusions
In this study, a total of 45 full-length SlCHRs were identified in tomato, which are clustered into 6 groups. Most SlCHRs within a group are highly conserved in sequence features, gene structures, and motifs, suggesting the functional conservation of SlCHRs within a group. Furthermore, diversities in the specific domains identified in different groups indicate that some SlCHRs may have undergone functional diversification. The expression profiles suggest that most SlCHRs are expressed constitutively in tomato organs, and RT-qPCR analyses show that the expression of some SlCHRs is modulated by the exogenous stimuli, suggesting that SlCHRs may play important roles in plant development and stress responses.

Data Availability
The original data of Snf2-like family proteins are available from ChromDB (http://www.chromdb.org). The sequences of tomato CHR proteins are available from the International Tomato Genome Sequencing Project (https://solgenomics .net/organism/Solanum_lycopersicum/genome).

Supplementary Materials
Supplemental Table 1: sequence features of SlSnf2s in tomato and Arabidopsis. Gene IDs, protein length, intron number, pI, and molecular weight of Snf2s in A. thaliana and tomato are shown. Supplemental Table 2: primers used in this study. Supplemental Figure 1: phylogenetic tree of SlCHR proteins. Maximum likelihood phylogenetic tree of predicted proteins from tomato. Bootstrap values higher than 50% are shown. Supplemental