The Sugar Transporter family in wheat (Triticum aestivum. L): genome-wide identification, classification, and expression profiling during stress in seedlings

Plant materials

The wheat cultivar Zhoumai 36 was germinated and grown in a tissue culture room at Zhoukou Normal University. The light conditions were set to dark for 12 h and light for 12 hours at 25–30 °C. Wheat seedlings were watered once every three days with Hoagland’s nutrient solution. The stress treatment was performed when seedlings grew to the second-true leaf stage. Wheat seedlings were divided into four groups with 100 seedlings per group. Seedlings from the first group to the third group were treated with 10% PEG6000 solution, 150 mmol L⁻¹NaCl, and 10% PEG6000 solution+150 mmol L⁻¹NaCl, respectively. The other group was left untreated as a control check (CK). Furthermore, different seedling tissues, such as leaves, stems, and roots, were collected to prepare the mixed cDNA template, which was used for TaSTP gene cloning. All fresh materials were quickly frozen in liquid nitrogen and stored in a −80 °C refrigerator for total RNA isolation.

Identification and characteristics of the TaSTP genes in wheat

The consensus protein sequences (PF00083) from the Sugar_tr hidden Markov model (HMM) were downloaded from Pfam 31.0 to identify the STP genes in wheat (https://pfam.xfam.org/family/PF00083). Then, the whole-genome data of wheat (IWGSC) and Brachypodium distachyon (BD) were obtained from the International Wheat Genome Sequencing Consortium and Phytozome v12.1 website. The HMM profile was used as a query to identify all STP-containing sequences in wheat and BD by searching against the downloaded genome with an E-value of <1e⁻⁵. Furthermore, all candidate TaSTPs were verified using Pfam and SMART (http://smart.embl-heidelberg.de/) to confirm that they contained the core domains. All potentially redundant TaSTP sequences and sequences lacking a part of STP domain or complete STP domain were discarded based on the sequence alignments generated by ClustalX software. Finally, a total of 81 STP protein sequences, which were encoded by 81 genes in the same subgroup as that in Arabidopsis thaliana, were identified in the wheat genome.

Phylogenetic analysis

For the phylogenetic tree of TaSTP, 14 Arabidopsis AtSTPs, eight rice STP protein sequences, six brachypodium distachyon STP protein sequences, eight foxtail millet STP protein sequences, 20 cassava STP protein sequences, 20 Ricinus communis STP protein sequences, and 81 full-length TaSTP protein sequences were aligned using ClustalX 2.1 (http://clustalx.software.informer.com/2.1/). A neighbour-joining (N-J) tree was constructed with the alignments using MEGA 7.0 software (Kumar, Stecher & Tamura, 2016) with bootstrap analysis from 1,000 replicates. The STP genes were named TaSTP1 to TaSTP81 according to their positions on the chromosomes.

Gene structure analysis and motif detection

The coding sequences (CDS) and genomic sequences of the STP genes were obtained from the International Wheat Genome Sequencing Consortium (http://plants.ensembl.org/Triticum_aestivum/Info/Index). By comparing the coding sequences and their chromosomal genomes, the exon-intron structures of the TaSTP family were generated using the Gene Structure Display Server website (http://gsds.gao-lab.org/). The MEME motifs of the predicted TaSTP proteins were identified using MEME suite 5.0.1 (http://meme-suite.org/tools/meme). Subsequently, these files were optimized using TBtools (Chen et al., 2018).

Homologous gene pairs and synteny analysis

The Ka/Ks ratios of gene pairs in duplication blocks were calculated using DnasP v6.0 and performed as described by Cao et al. (2016). The parameters of the sliding window analysis of the Ka/Ks ratios were set as follows: window size, 200 bp; step size, 25 bp. Moreover, TaSTP duplications were performed using the MCScanX program in the Bio-Linux system. In brief, with the BLAST outputs, the file of the gene positions, and the use of the MCScanX program, the TaSTP genes were classified into different types of duplications. We used Circos software to construct a schematic diagram (Krzywinski et al., 2009; Liu et al., 2018b) to exhibit the synteny relationship of the orthologous TaSTP genes obtained from wheat and other selected species. The syntenic analysis maps were constructed using the Dual Systeny Plotter software (https://github.com/CJ-Chen/TBtools).

Expression profiles of the TaSTP genes in the RNA-Seq data

Public transcriptome data were downloaded from the website http://www.wheat-expression.com/. Subsequently, all TPM data were log10 transformed, which was detailed in a previous study (Yue et al., 2018), and a heat map was created using TBtools.

Network interaction and GO analysis

The protein interaction network analysis of the TaSTPs was conducted using the STRING v10.5 database (http://www.string-db.org) based on the orthologous genes between Arabidopsis and wheat. GO analysis was conducted by GOToolBox (http://genome.crg.es/GOToolBox/) and visualized with Tbtools.

RNA isolation and validation analysis

Total RNA was extracted using the Trizol method (TriQuick Reagent, Solarbio, China), denatured in agarose gel (1%), and stained with SolarGelRed (Solarbio, China) to check the quality of the mixed RNA. The first-strand cDNAs were synthesized using a PrimeScript RT reagent kit (Takara Bio, Tokyo) according to the manufacturer’s instructions and stored at −20 °C. Specific primers were designed using Beacon Designer software version 8.13 (Premier Biosoft International, Palo Alto, CA, USA) for the quantitative real-time polymerase chain reaction (qRT-PCR); the primer sequence details are provided in Table S1. Detailed information on the qRT-PCR protocol was described in a previous study (Liu et al., 2018a).

Identification of the STP gene family in wheat

To identify the STP family genes in wheat, the Sugar_tr HMM profile (Pfam: PF00083) was used as a query in a BlastP search against the wheat genome database (International Wheat Genome Sequencing Consortium: IWGSC). A total of 476 candidate TaSTP proteins were identified from the wheat genome with an E-value of <1e⁻⁵. To identify the members of the TaSTP protein family, the proteins were checked for the presence of the Sugar_tr domain by the Pfam database (Finn et al., 2010) and the SMART database (Letunic & Bork, 2018). A total of 86 candidate STP proteins were discarded in this study because of a lack of a part of STP domain or complete STP domain. Finally, a total of 390 putative TaSTP genes were identified in the wheat genome through a HMMER analysis. A phylogenetic tree was constructed according to the protein sequences of 390 putative TaSTP and 14 AtSTP. The evolutionary tree was divided into three subgroups, and only the second subpopulation contains all Arabidopsis sugar transporters proteins, named AtSTP1 to AtSTP14 (Fig. 1, Group 2). These are all monosaccharide transporters and our focus was on the analysis of these monosaccharide transporters in wheat. There were 81 genes in wheat in the same group as in Arabidopsis thaliana. According to the position of the genes on the chromosomes, the 81 TaSTP genes were unevenly located on the wheat chromosomes and thus were named TaSTP 1 to TaSTP 81. The general information on the 81 TaSTP members, including molecular weight, isoelectric point, CDS length, amino acid number, and transmembrane domain information, is summarized in Table 1. These genes were distributed on either the long or short arm of the A, B, and D chromosomes. Among them, chromosomes 1A, 2A, 3A, 4A, 5A, 6A and 7A had four, seven, two, three, eight, three, and two TaSTPs, respectively; chromosomes 1B to 7B had four, eight, three, four, five, two, and one TaSTPs, respectively; and chromosomes 1D to 7D had three, six, three, three, seven, one, and two TaSTPs, respectively (Table 1). We found that chromosomes A, B, and D had 29, 27, and 25 TaSTP genes, respectively. The open reading frames (ORFs) and the protein lengths of the TaSTP genes ranged from 1,080 to 1,848 bp in length and 359 to 615 amino acids in length, respectively (Table 1). The predicted molecular weights of the TaSTPs ranged from 39.55 kDa to 67.04 kDa. The theoretical isoelectric points (pIs) of the TaSTP proteins ranged from 8.42 to 10.05. Moreover, 77 of the 81 TaSTPs contained 10 to 12 conserved transmembrane domains (TMDs), and most had 11 or 12; however, TaSTP73, TaSTP57, and TaSTP6 carried only six, eight, and nine TMDs, and TaSTP24 carried 13 TMDs (Table 1).

Sequence structure features of the TaSTP s

According to the neighbour-joining phylogenetic tree of the TaSTPs gene sequences, which was constructed using 1,000 bootstrap replicates by MEGA v7.0 (Pennsylvania State University, Philadelphia, PA, USA), the 81 TaSTP genes could be classified into three distinct groups according to the similarity of their gene sequences, and the bootstrap value ranged from 331 to 1,000 (Fig. 2). The largest group contained 42 TaSTP genes, and the other two groups contained 16 and 23 TaSTP genes, respectively. The TaSTP gene structures commonly had two, three, or four CDS divided by one, two, or three introns, except for TaSTP43, TaSTP77, and TaSTP80 with one CDS. Furthermore, most of the TaSTP genes that were classified into the same subgroup exhibited similar gene structures, such as TaSTP42,69 and TaSTP70, TaSTP6,11 and TaSTP4, TaSTP71,76, and TaSTP74 (Fig. 2).

To understand their functional regions, the conserved motifs of the TaSTP proteins were identified through the MEME website. Twenty motifs with minimum and maximum motif width set from six to 200 in the 81 proteins were analysed using the MEME tool. The results from MEME agreed with those from the phylogenetic tree (Fig. 3). It was found that most of the TaSTP proteins contained 15 or 16 motifs. TaSTP2,15,22 and 29 had 14 motifs without motif 15 or another motif; TaSTP23 and TaSTP35 had 12 motifs without motifs six, eight, and 15; and TaSTP73 had only nine motifs without motifs one, two, seven, nine, 14, and 15 (Fig. 3). We used Pfam to further analyze the 20 motifs and found that motifs one, two, three, five, and seven contained the Sugar_tr region. These five motif widths were 94, 72, 41, 36 and 45, respectively (Fig. 4, Table S2).

Phylogenetic analysis of the TaSTP genes in seven species of monocotyledons and dicotyledons

To investigate the evolutionary relationships of the TaSTP gene family in monocotyledons of graminaceae (wheat, foxtail millet, brachypodium distachyon, and rice) and dicotyledons such as Arabidopsis, cassava, and Ricinus communis, we used the STP genes from wheat (81), brachypodium distachyon (6), foxtail millet (8), rice (8), cassava (20), Ricinus communis (20), and Arabidopsis (14) to construct a N-J phylogenetic tree usng MEGA 7.0 software with 1,000 bootstraps (Table S3). In the phylogenetic tree, we use different-colored clades to represent different species: wheat (black), brachypodium distachyon (gray), foxtail millet (green), rice (azure), Arabidopsis (red), cassava (blue), and Ricinus communis (purple). The results of the phylogenetic tree indicated that it can be roughly divided into five groups: Group I was one small branch with seven cloned wheat STP genes (TaSTP3,5,10,46,47,57,58), and the other large branch; Group II to Group V mainly contained the other genes, characterized by blue, purple, sky blue, and green areas, respectively. The branches of Group II in the phylogenetic tree show that eight STP genes (TaSTP7, 8, 33, 35, 38, 43, 77, 80) and AtSTP3 were brought together. Group III consisted of twenty-three TaSTPs (TaSTP1, 2, 9, 14-18, 21-26, 28-32, 45, 49, 56, and 72), two RcSTPs, one MeSTP, BdMST1, AtSTP3 and SiMST1, and OsMST1 (Fig. 5). Nineteen TaSTPs (TaSTP12, 13, 19, 20, 27, 34, 36, 37, 39, 40, 51, 59, 64, 71, 74, 76, 78, 79, and 81) clustered with six AtSTPs (AtSTP1, 4, 9-12), seven RcSTPs, five MeSTPs, four BdMSTs, and six SiMSTs in Group IV (Fig. 5). Group V consisted of twenty-three TaSTPs (TaSTP4, 6, 9, 11, 41, 42, 44, 48, 52-55, 60-63, 65, 66-70, 73, and 75), six AtSTPs (AtSTP2, 6, 7, 8, 13, and 14), twelve MeSTPs, ten RcSTPs, BdMST4, SiMST4, and OsMST4 (Fig. 5).

Chromosome location and gene duplication analysis

The approximate locations of the TaSTP genes on the chromosomes of wheat were determined by MapDraw software according to http://plants.ensembl.org/Triticum_aestivum/Info/Index. We found that the 81 genes were located separately on chromosomes A, B, and D. Moreover, TaSTP14-TaSTP18, TaSTP21-TaSTP26, TaSTP28-TaSTP32, TaSTP53-TaSTP55, TaSTP61-TaSTP63, and TaSTP66-TaSTP68 formed a gene cluster with more than three genes on chromosomes 2A, 2B, 2D, 5A, 5B, and 5D, respectively (Fig. 6). We also investigated the tandem duplications and whole-genome duplications of the 81 TaSTP genes. As shown in Fig. 6, the TaSTP genes were differentially distributed on 21 wheat chromosomes. Among these TaSTP genes, 159 pairs of genes exhibited whole-genome duplications/segmental duplications (Table S4). A total of 22, 23, and 21 TaSTP genes were distributed in the A, B and D sub-genome, respectively (B¿A¿D) as shown with the tandem replication events of multiple genes in Fig. 7. Initial gene loss likely occurred in the A genomes following tetraploidy and D genomes following hexaploidy to decrease functional redundancy and define the core wheat genes. There were nine, eleven, eight, nine, eighteen, six, and give genes in groups one through seven of the chromosomes, which showed three obvious gradients between groups two and five, and one, three, four and the other two groups. Specifically, chromosomes 5A, 5B and 5D had six, five, and seven genes, respectively, whereas chromosome 7B had only one TaSTP gene. These results indicated that the distribution of the TaSTP genes was not random in chromosomes and that gene replication events might have occurred in chromosomes two and five and may involve gene functions.

We constructed three comparative syntenic maps of wheat associated with three representative species, including rice, foxtail millet, and brachypodium distachyum to further infer the phylogenetic mechanisms of the wheat TaSTP family. A total of 25 orthologous TaSTP gene pairs were found between wheat and rice and brachypodium distachyum (Table S5; Fig. 8A). These TaSTP genes were found to be associated with at least three syntenic gene pairs (particularly between rice and brachypodium distachyum genes), such as TaSTP12, TaSTP4,1 and TaSTP48 (shown as different color lines in the figure). These genes may have played an important role in the STP gene family during evolution. Moreover, 38 orthologous TaSTP gene pairs were found between wheat and foxtail millet (Table S5; Fig. 8B). We calculated the Ka/Ks ratios of the STP gene pairs to better understand the evolutionary constraints acting on STP gene family. All orthologous STP gene pairs had Ka/Ks < 1, suggesting that the wheat STP gene family might have experienced strong purifying selective pressure during evolution (Table S6). These results suggested that the STP genes of monocot have strong relationships.

Strong purifying selection for the TaSTP gene pairs in wheat

In general, a nonsynonymous (Ka)/synonymous (Ks) ratio > 1 indicates positive selection, Ka/Ks ¡ 1 indicates purifying selection with functional constraints, and Ka/Ks = 1 indicates neutral selection. In our study, the Ka/Ks ratios of most of the TaSTP paralogous pairs were less than 1, and the ratio of only 1 paralogous pair was more than 1, namely, TaSTP1 & TaSTP72. Remarkably, of these Ka/Ks ratios <1, 147 TaSTP paralogous pairs were less than 0.27 and appeared to be under strong purifying selection, 11 were between 0.27 and 0.98 and appeared to be under purifying selection, and one appeared to be under positive selection (Fig. 9, Table S6).

Expression profiles of the TaSTP genes

The RNA-seq data from http://www.wheat-expression.com/ were used to explore the expression profiles of the TaSTP genes in different tissues and under different abiotic stresses. According to the heat map results, the expression level of 76 TaSTP genes (genes with the same transcript were deleted) could be classified into A (yellow), B (blue), and C (pale red) pattern groups (Fig. S1). In the pattern group A (yellow), most genes in the different tissues showed low or no expression, and the genes TaSTP28, TaSTP21, TaSTP1, TaSTP68, TaSTP55, and TaSTP63 were expressed in the roots of Chinese spring wheat. In the B pattern group (light blue), no gene expression was detected. Conversely, in the C pattern group (pale red), most genes, such as TaSTP27 and TaSTP48, in the different tissues showed positively-regulated expression and, specifically, a higher expression level in the flag leaf blades and roots. Some genes were highly expressed in the seed coat, such as TaSTP52, TaSTP53, TaSTP60, TaSTP61, TaSTP65, and TaSTP6 6. In the stressed seedling leaves, the expression pattern was roughly divided into two groups: one group showed negative regulation with low expression compared with that of the control, such as the expression patterns in heat, drought stress or heat and drought combined stress at one hour and six hours, and the other group was positively or negatively regulated by PEG6000 stress at two hours and 12 h in nine-day-old seedlings of Giza168 and Gemmiza 10. Furthermore, compared with the gene expression in the no stress control, most genes were upregulated in shoots under cold stress at 4 °C for two weeks, except for the genes TaSTP44, TaSTP 53, TaSTP56, and TaSTP61 (Fig. S1).

Bioinformatic-based protein interaction network analysis of the TaSTPs using STRING 10.5 and GO analyses

To explore the relationship between all the identified TaSTPs, we created a protein interaction network by searching the 81 proteins against the TAIR (The Arabidopsis Information Resource) database in STRING. We found 17 proteins with hits in the STRING database. These proteins interactions were divided into three groups: the first group (blue) was transglycosidase (DIN10), which is involved in the synthesis of raffinose, a major soluble carbohydrate in seeds, roots and tubers, and the second group (green) was trehalose phosphate synthase and trehalose-6-phosphate phosphatase (which removes the phosphate from trehalose 6-phosphate to produce free trehalose that accumulates in plants and may improve abiotic stress tolerance). The third group (red) was beta-fructofuranosidase (ATBFRUCT1, that can use sucrose and 1-kestose), glutamate synthase, and glutamate dehydrogenase, which are required during photorespiration (Fig. S2).

To further identify the functions of the putative TaSTP proteins in wheat, molecular functions, cellular components, and biological pathway categories were predicted using GO annotation analysis. Based on amino acid similarity, the results of the GO analysis showed that 81 TaSTP proteins were categorized into eight functional categories (Fig. 10). Among the molecular functions, transporter activity and catalytic activity were predominant. The analysis of the cellular component annotations revealed that most of the TaSTP proteins were predominantly localized to the plasma membrane, membrane and vacuole. Furthermore, 81 TaSTP proteins were assigned to transport, response to abiotic stimulus, and response to stress based on biological process analyses.

Phenotype comparison of seedling roots, stems, and leaves under stress

Stress treatment was performed when seedlings grew to the second-true leaf stage. Wheat seedlings were divided into four groups with 100 seedlings per group and were treated with 150 mmol L⁻¹NaCl, 10% PEG6000 solution, 10% PEG6000 solution+150 mmol L⁻¹NaCl, or the control. Investigations were carried out after three, six, and nine days of stress, and the length and width of the roots, stems and leaves were accurately measured using vernier callipers. The phenotypes of the seedling roots, stems, and leaves under stress were different from those of the normal seedlings, and the most obvious difference was in root length (Figs. S3A–S3L). Specifically, at the three time points of stress, the root length of control was slightly elongated, while NaCl stress and PEG6000+NaCl stress did not change significantly in the root length. Interestingly, the root lengths of plants under PEG6000 stress continued to increase and were significantly different from that of the control on the sixth day and ninth day of stress (P < 0.01) (Fig. S4A). In contrast, the root width of the control, NaCl stress, and PEG6000+NaCl stress plants gradually decreased as the plants grew. The root width of PEG6000 tended to be thinner on the sixth day and then thicker on the ninth day, but it was significantly thinner than that of the control (Fig. S4B). In the stems, there was no significant change in length among the four types of seedlings, but the diameter of the seedlings treated with stress was thinner than that of the control on the sixth and ninth days of treatment and showed significant (P < 0.05) or extremely significant (P < 0.01) differences. (Figs. S4C and S4D). The lengths of first true leaves of control were slightly shorter than those of the stressed leaves, but the width was wider than that of the stressed leaves (Figs. S4E and S4F). The length and width of the second true leaves of the NaCl-stressed and PEG6000+NaCl-stressed seedlings were significantly shorter than those of the control leaves (P < 0.01). However, the PEG6000-stressed leaves had approximately the same length and width as the control leaves (Figs. S4G and S4H).

Expression characteristics of wheat TaSTP genes using qRT-PCR analysis

We investigated the expression levels of the TaSTP genes during stress at the sixth day of wheat development to further explore TaSTP gene functions in wheat. First, we randomly selected four genes, TaSTP12, TaSTP41, TaSTP48, and TaSTP65, for qRT-PCR analysis, and their expression is quantified in Fig. 11. The results showed that these genes were expressed in the roots, stems, and leaves. Interestingly, the expression levels of the individual genes in the leaves were not the same but the expression of the same genes in the leaves was not affected by stress. For the gene TaSTP 12, the expression levels in the stressed wheat stems were similar but they were lower than those in the control stems; the difference was significant (P < 0.05). The expression of TaSTP 12 was significantly higher in the stressed roots than in the control roots, and the differences were extremely significant, with a 1.4-fold change and 2.11-fold change in the roots of the NaCl-stressed and PEG6000+NaCl-stressed roots, respectively (P < 0.01) (Fig. 11A). Compared with TaSTP 41 expression in the control stems, TaSTP 41 expression was significantly downregulated in the PEG6000+NaCl-stressed (2.17-fold change, P < 0.01) and PEG6000-stressed stems (1.18-fold change, P < 0.01), but the NaCl-stressed did not show a difference. TaSTP 41 expression was significantly upregulated in the PEG6000+NaCl-stressed (1.16-fold change, P < 0.01) and NaCl-stressed (1.4-fold change, P < 0.01) roots relative to the control roots. However, the expression was significantly downregulated (1.26-fold change, P < 0.01) in the PEG6000-stressed roots relative to the control roots (Fig. 11B). A difference in expression was not detected for the gene TaSTP 48 between the control stems and stressed stems. In the roots, the expression level of TaSTP 48 increased in the stressed roots compared with that in the control roots, and this was especially the case with PEG6000 stress, which showed more than a 10.04-fold change (P < 0.01) (Fig. 11C). TaSTP 65 expression was significantly downregulated in the PEG6000+NaCl-stressed (8.62-fold change, P < 0.01) and NaCl-stressed (7.07-fold change, P < 0.01) stems compared with that in the control stems. In the roots, significantly elevated expression levels were seen with sodium chloride stress, specifically, a 1.15-fold change and a 2.57-fold change with NaCl stress and PEG6000+NaCl stress, respectively, (P < 0.01). Moreover, TaSTP 65 showed a lower expression with a 1.26-fold change (P < 0.01) in the PEG6000-stressed roots compared to that in the control roots (Fig. 11D).