Identification and Physical Property Analysis of Putative G2-Like Proteins in Tomato
A total of 66 G2-like proteins (designated SlGlk1 to SlGlk66) were identified using the bioinformatics approach. Their basic physical information is presented in Table S2. The predicted molecular weights ranged from 20316.24 Da (SlGlk60) to 77633.43 Da (SlGlk63). Consistent with the molecular weights, SlGlk60 (174 aa) had the shortest sequence and SlGlk63 (708 aa) had the longest. The isoelectric point varied from 4.83 (SlGlk34) to 10.07 (SlGlk46), and the predicted aliphatic indices varied from 55.85 (SlGlk21) to 90.43 (SlGlk58). The instability indices varied from 33.02 (SlGlk8) to 63.18 (SlGlk61), and the instability indices of SlGlk8/9/22/30/43/62 were lower than 40, indicating that these proteins were relatively stable. The hydrophobicity indices ranged from -1.003 (SlGlk28) to -0.29 (SlGlk14).
Sequence Alignment And Phylogenetic Analysis
The phylogenetic relationships of the G2-like proteins of tomato were classified into five groups according to their groupings with maize G2-like proteins and motif analysis of tomato G2-like proteins (Text S1 and Text S2). Groups I to V contained 12, 6, 4, 18 and 26 G2-like proteins, respectively (Figure 1A). Multiple sequence alignments (Figure 1B) revealed that the genes were conserved across at least two regions of the HLH structure of the Myb-like domain: the first helix contained the initial sequence PELHRR (except in SlGlk46), and the second helix contained NI/VASHLQ. In addition, most SlGlks in group Ⅳ had a specialized conserved Myb-CC-LHEQLE domain. A separate multiple sequence analysis was also performed in each subgroup (Figure S1). In the members of groups Ⅰ and II, the form was mostly LHR/L/H/A; PELHRR and PDLHRR were the two types of conserved domains in group Ⅲ members; the form of PELHRR was XD/ELHD/EX in group Ⅳ members; and XQ/ELHXX was the main type in group V members. The VI/NASHLQ domain was more conserved than the PELHRR domain was in all the groups. Four main types were present in members of all the groups: LKSHLQ, VKSHLQ, IKSHLQ and VASHLQ. On the basis of the results of the analysis of each group, we summarized the types of G2-like conserved domains (Table S3).
Motif Analysis and Gene Structure of Tomato G2-like Genes
Fifteen putative motifs whose lengths varied from 13 to 50 amino acids were identified via the MEME website (Table S4) as representing the structure of all G2-like proteins. The results of the motif analysis are shown in Fig. 2B. Each protein sequence includes a different number of motifs (1~7), and each motif is present only once, with the exceptions of motif 13/motif 2 and motif 6, which were present twice in SlGlk6 and SlGlk8, respectively. Most G2-like proteins (83.33%) contained motif 1 and motif 2, which corresponded to the myb SHAQKYF domain based on Pfam database and Conserved Domain Database (CDD). We also found that the PLN03162 superfamily domain also contained motif 1 and motif 2. SlGlk62 and SlGlk46, which corresponded to the myb SHAQKYF domain and PLN03162 superfamily domain, contained only motif 2 and motif 11, respectively. The same motif compositions correspond to different domains, or different motif compositions correspond to the same domain, suggesting functional diversity. Details of the CDD results are showed in Table S5. The G2-like proteins clustered in the same group of the phylogenetic tree contained a similar motif composition besides SlGlk46/62, showing that they were highly conserved. For example, motif 15 was present only in group Ⅲ members, which had very similar sequences. SlGlk6/7/8/9, which were formed from tandem duplications, also presented similar motif compositions. The function of the majority of these motifs needs to be further studied.
Studying the introns and exons of tomato, which were determined by the alignment of G2-like genes, would give more insight into the evolution of the G2-like family members in tomato. Intron and exon predictions are shown in Fig. 2C, and the sequences of the G2-like genes are shown in Text S1. The number of exons varied from 1 to 11. More than half of the G2-like genes (62, 93.94%) had four or more exons, and only 4 genes (6.06%) had three or fewer exons. The conserved regions of all of the tandemly duplicated genes and segmentally duplicated genes presented similar exon distributions. Overall, the phylogenetic analysis results, motif composition and similar gene structure of the G2-like members in the same group provided reliable results for group classification.
Chromosomal Location and Gene Duplication Events of G2-like Genes in Tomato
SlGlks were unevenly distributed across the 12 tomato chromosomes, and the locations of most SlGlks were on the proximal or distal ends of the tomato chromosomes. The number of SlGlks per chromosome ranged from 2 to 9 (Chr10 had 9 genes; Chr03 only had 2 genes). As shown in Fig. 3, Chr01, which is the longest chromosome, contained only 3 genes. There was no significant positive correlation between chromosome length and gene number.
Genome duplication events, which are usually divided into three types (tandem duplications, segmental duplications and transposition events), occurred during plant evolution [33–35]. Tandem duplications are defined as chromosomal regions that are less than 200 kb in length and contain two or more genes . There were 4 genes (SlGlk6/7/8/9) located on Chr02, which resulted from tandem duplications, that formed 1 tandem duplication event region. Using BLASTP and MCScanX, we also found 9 segmental duplications (18 G2-like genes in total) events (Figure 3 and Table S6). Taken together, these results showed that some SlGlks may have arisen via gene duplication.
Analysis Of Cres
To further determine the potential function of G2-like genes in response to abiotic stress, the CREs within the sequences 2 kb upstream from the translation start site of the G2-like genes were searched within the PlantCare database. Analysis of the promoters of SlGlks in tomato revealed that all family members contained light-responsive elements and two core elements-the CAAT box and TATA box. Detailed elements are displayed in Table S7 and Fig. 4. There are also significant differences in the number of CREs among the promoters of the different members of the G2-like gene family. As shown in Fig. 4, the promoters of SlGlk18 contained the most kinds of CREs (13), while SlGlk34 contained only three kinds of CREs. Only 20 SlGlks did not have any abiotic stress response elements, while the other SlGlks contained at least one abiotic stress element, which indicated that the expression of more than half of the G2-like genes was related to abiotic stress. In addition, we found that 52 SlGlks (78.79%) had two or more hormone induction elements and that SlGlk47 contained all five hormone induction elements, such as ABA-, IAA-, GA-, JA- and SA-induction elements. Regardless, analysis of the CREs showed that the number and distribution of CREs and that CREs in the same subgroup were not similar, which indicated that each SlGlks are regulated by different combinations of TFs and that the expression of SlGlks could be induced in response to different hormones and abiotic stresses.
G2-like Gene Expression Patterns in Response to Different Abiotic Treatments
To explore the G2-like genes that respond to three different abiotic stresses, we downloaded RNA-seq data from the NCBI database; data for SlGlk4/9/43/57/58 were not found in the RNA-seq database. Nonetheless, the information was presented in the form of a heatmap (Table S8 and Fig. 6). As shown in Fig. 5, there were 40 genes expressed in response to cold stress, 42 genes expressed in response to drought stress and 41 genes expressed in response to salt stress. We found that the expression of SlGlk22/44/25/59/29/56/24 was high in response to the three different stress treatments. The expression of SlGlk56 increased but then decreased after cold and drought stress treatment. Moreover, the expression pattern of SlGlk36 decreased but then increased after drought and salt stress treatment, while the opposite occurred after cold stress treatment. These results indicated that some G2-like genes are involved in abiotic stress responses. We defined G2-like genes DEGs whose expression levels changed more than or equal to twofold than 0 h (and when P<0.05). A total of 53.03% of G2-like genes, including 5 cold stress-related genes, 22 drought stress-related genes and 27 salt stress-related genes, were differently expressed. As shown in Fig. 6A, more than half of the DEGs (21 genes) responded to one stress treatment, and only 11 genes responded to two stresses. Four genes, SlGlklk11/20/26/62, were expressed in response to all the stresses. The detailed gene list is shown in Fig. 6B. The DEGs that responded to only one abiotic stress or that differed from the other two genes, are shown in Fig. 6C. The results showed that the expression of half of the drought specific-related genes was upregulated. In contrast, the expression of most salt specific-related genes was downregulated. Interestingly, we also found that the expression of SlGlk44/21/1 and SlGlk46/22/65/36 was up-/down-regulated under both drought and salt stresses respectively.
Expression of Tomato G2-like Genes in Response to Abiotic Stress and Hormone Treatments
To verified the expression data from RNA-seq and explore whether the expression of SlGlks is affected by hormone treatments, we randomly selected 11 tomato G2-like genes from among the 36 DEGs to investigate through qRT-PCR the transcript levels of these 11 genes in response to different treatments. Detailed expression patterns of these G2-like genes are shown in Fig. 7. The expression of these 11 genes generally corresponded to RNA-seq results, indicating the RNA-seq results was reliable. We also found that the expression of these SlGlks could be induced by at least 4 different abiotic stress and hormones treatments after 24 h and 12 h. Interestingly, we also found that the expression of some SlGlks was induced/repressed by one treatment. For instance, the expression levels of all SlGlks except SlGlk36 were upregulated under JA and IAA treatment, while the expression level of most SlGlks were downregulated under drought and ABA treatment. The expression levels of seven SlGlks (SlGlk61/16/38/55/53/54/64) were first upregulated but then were downregulated under GA treatment, and the expression of SlGlk16 exhibited the same dynamic pattern under abiotic stresses. Several genes, such as SlGlk38/55, exhibited opposite expression patterns under different treatments. Taken together, these results showed that SlGlks were influenced by most of the applied abiotic stress and hormone treatments.
GO Enrichment and KEGG Enrichment Analysis of G2-like DEGs
To further determine the function of the 36 DEGs, GO enrichment and KEGG enrichment analysis was performed. The results of GO analysis can be generally divided into three categories: BP, CC and MF. We found that 87 GO terms were enriched for 31 DEGs, excluding SlGlk22/23/38/14/34 and the detailed information was displayed in Figure S2. Biological regulation and cellular process were dominant in the BP category. CC clusters contained two subcategories, cell and organelle, with 31 DEGs, and binding was the only subcategory among the MF category. The top 20 terms for enrichment analysis with the whole genome as the background according to P value are shown in Fig. 8A. With the exception of DNA binding, the remaining of 19 terms belonged to BP groups. Thirty-five DEGs (SlGlk62 was excluded) were mapped to 3 KEGG pathways (P value<0.05): plant hormone signal transduction; arginine biosynthesis; and alanine, aspartate and glutamate metabolism (Figure 8B and Table S9). Therefore, we divided the annotated DEGs into two groups: 31 DEGs in the plant hormone signal transduction pathway were assigned to group 1 while 4 DEGs (SlGlk25/44/45/50) related to amino acid metabolism were assigned to group 2. We further statistically analyzed the expression patterns of these genes under each abiotic stress in the two groups. We found that the expression patterns of DEGs in the same group were different. In group 1, the proportion of genes whose expression was downregulated was higher than the genes whose expression upregulated under drought and salt stress treatment, while the number of genes whose expression was upregulated was far greater than the number of genes whose expression was downregulated under cold stress treatment. In group 2, the proportions of genes whose expression was downregulated and the genes whose expression was upregulated were equal, and the expression of only one gene was upregulated. However, the expression of none of them changed under cold stress.
Silencing SlGlk16 reduced drought tolerance in tomato plants
Based on the previous study (the SlGlks expression under different treatments and RNA-seq databases under abiotic treatments), SlGlk16 which is defined as drought-specific related DEG was selected for further analysis under drought stress. Figure S3 showed the expression level of SlGlk16 in the infected tomato seedlings and we chose the plants (4 seedlings) whose SlGlk16 expression level was less 50% than CK for further test. As shown in Fig. 9A, the growth status of all plants was in similar before abiotic treatment, and the wilting of all plants was increased with increasing drought treatment time. Noteworthy, the leaves of SlGlk16-silenced plants began to wilt and the stem began to curl after 3 h drought treatment, while other plants only exhibited a slight curled stem. After 12 h later, the leave of SlGlk16-silenced plants showed wilting and drying more severe than the other two groups, although the above phenomena increased in severity in all the plants. The activity of SOD, POD and the content of MDA and Pro were measured for all plants under normal and drought treatment to explain the decreased drought tolerance in SlGlk16-silenced plants. After drought treatment, the activity of SOD and POD and the content of MDA increased in all plants. Under drought treatment, the activities of SOD and POD in SlGlk16-silenced plants were lower than that in the other two groups (Figure 9B, C). However, the MDA contents in SlGlk16-silenced plants were higher than others after drought treatment (Figure 9D). Interestingly, we also found that the Pro contents in detached leaves of SlGlk16-silenced plants were decreased after drought treatment and lower than CK and CK-TRV2.