Sugar Transporter Proteins (STPs) in Gramineae Crops: Comparative Analysis, Phylogeny, Evolution, and Expression Profiling

Sugar transporter proteins (STPs), such as H+/sugar symporters, play essential roles in plants’ sugar transport, growth, and development, and possess an important potential to enhance plants’ performance of multiple agronomic traits, especially crop yield and stress tolerance. However, the evolutionary dynamics of this important gene family in Gramineae crops are still not well-documented and functional differentiation of rice STP genes remain unclear. To address this gap, we conducted a comparative genomic study of STP genes in seven representative Gramineae crops, which are Brachypodium distachyon (Bd), Hordeum vulgare (Hv), Setaria italica (Si), Sorghum bicolor (Sb), Zea mays (Zm), Oryza rufipogon (Or), and Oryza sativa ssp. japonica (Os). In this case, a total of 177 STP genes were identified and grouped into four clades. Of four clades, the Clade I, Clade III, and Clade IV showed an observable number expansion compared to Clade II. Our results of identified duplication events and divergence time of duplicate gene pairs indicated that tandem, Whole genome duplication (WGD)/segmental duplication events play crucial roles in the STP gene family expansion of some Gramineae crops (expect for Hv) during a long-term evolutionary process. However, expansion mechanisms of the STP gene family among the tested species were different. Further selective force studies revealed that the STP gene family in Gramineae crops was under purifying selective forces and different clades and orthologous groups with different selective forces. Furthermore, expression analysis showed that rice STP genes play important roles not only in flower organs development but also under various abiotic stresses (cold, high-temperature, and submergence stresses), blast infection, and wounding. The current study highlighted the expansion and evolutionary patterns of the STP gene family in Gramineae genomes and provided some important messages for the future functional analysis of Gramineae crop STP genes.


Plant Treatments
Rice seedlings of different ages grown in greenhouses were used for cold, high-temperature, submergence, wound, and rice blast treatment. For heat and cold treatments, the temperatures for 12-day-old seedlings were changed to 40 • C and 4 • C, respectively, and then sampled (leaves) at 0, 1, 3, 6, and 12 h. As to submergence (Sub) treatment, 12-day-old seedlings were placed in water, with a depth of 5 cm from leaf top to water surface, and sampled (leaves) at 0, 12, 24, 48, and 72 h. For wound treatment, the daily photoperiodic cycle of 9-day-old seedlings was changed to a 24-h light condition. Then, 12-day-old seedlings were cut off nearly one third of leaf from the top, and sampled (leaves) at 0, 1/4, 1/2, 1, 3, 6, 12, and 24 h. As to rice blast treatment, the daily photoperiodic cycle of 9-day-old seedlings was changed to a 24-h dark condition. Next, 12-day-old seedlings were sprayed with 70 mL spore suspension (about 10 5 conidia mL −1 ), and sampled (leaves) at 0, 1/4, 1/2, 1, 3, 6, 12, 24, and 48 h. During the sampling time, ddH 2 O was sprayed to keep the humidity of the treated seedlings.
In this study, three biological replicates were produced for every treatment. Each replicate was collected from over 12 seedlings and pooled together. All samples were triturated immediately with liquid nitrogen, and stored at 80 • C for further use [4,31].
All identified STP proteins from these seven representative Gramineae crops were aligned by ClustalW [32] and a phylogeny tree was generated in MEGA 6.0 using the neighbor-joining (NJ) method with 1000 bootstrap replicates [4,32]. Names of putative STP genes were assigned based on chromosomal order in each genome in accordance with the previous rice STPs study [4,14].

Gene Duplication Events, Orthologous Groups' Analysis, Gene Structure, and Conserved Motifs
Various types of gene duplication events of the STP gene family were identified by the 'duplicate_gene_classifier' script in MCScanX with an E-value of 1 × e -5 in BlastP search [32]. The synonymous (Ks)/nonsynonymous (Ka) substitution (Ka/Ks) rates were analyzed using DnaSP 5.0 (http://www.ub.edu/dnasp/) for selective force analysis of a duplicate gene pair [33]. Moreover, the divergence time of all duplicate pairs was estimated by T = Ks/(2 × 9.1 × 10 −9 ) × 10 −6 million years ago (Mya) [4,32]. First, an all-vs-all BlastP result was obtained by diamond software with parameters: E value 1 × e −3 (https://ab.inf.uni-tuebingen.de/software/) as the input file for OrthoFinder v2.0 software (University of Oxford, Oxford, UK). Then, orthologous groups were analyzed according to the previously published paper [34]. The phylogenetic tree of all tested species was rebuilt based on the result of orthologous groups using STAG and STRIDE algorithms in OrthoFinder software [34]. Additionally, the selective forces on orthologous groups were evaluated by Tajima's D values and Tajima's D of each orthologous group was calculated by DnaSP 5.0 [33].
Exon/intron site and length data were extracted from seven respective genome annotation GFF files. The conserved motifs of all STP proteins were analyzed by the MEME program (http: //meme-suite.org/tools/meme) with parameters: the maximum number of 20 motifs with motif length between 6 wide and 300 wide amino acids. Then, TBtools was used to visualize the phylogenetic tree, gene structure, and conserved motifs of the STP gene family [35].

Quantitative Real-Time PCR (qRT-PCR)
All RNAs were extracted by the TRIzol Ragent (Invitrogen, Beijing, China) and reverse transcriptions of all RNAs were performed using HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Shanghai, China), according to their instruction manuals. The qRT-PCR reaction (10 µL) was formulated using the 2X SYBR Green qPCR Master Mix (US Everbright ® Inc., Suzhou, China). Primers of rice 19 STP genes were designed by Primer 5.0 and displayed in Table S1. All qRT-PCRs were carried out on a CFX96 Touch™ Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). The actin gene was used as an internal control [32]. Three biological replicates (from three independent RNA samples) of each treated point were used for qRT-PCR. For each biological replicate, three technical replicates were used. The average threshold cycle (Ct) from three biological replicates was used to calculate the gene expression fold change by the 2 −∆∆CT method [4,32].  Figures 1 and 2, Hv, Bd, Os, and Or have more STP genes than Si, Sb, and Zm. We speculated the number of differences of STP genes in these species as a result of gene expansion of different clades among these species. Thus, a phylogenetic analysis and clade classifications were conducted. We found that all STP proteins were clustered into four clades, which include Clade I, Clade II, Clade III, and Clade IV. The Clade I, Clade III, and Clade IV showed a clear number expansion compared to Clade II ( Figure 2). Clade III and Clade IV vary greatly among these tested species, while Clade I and Clade II showed more STP gene copies in Hv, Bd, Os, and Or than Si, Sb, and Zm ( Figure 2).

Homologous Relationship, Gene Structure, and Selective Forces
To understand the expansion mechanism of paralogues in these species, we investigated gene locations and duplication modes within each species via MCScanX. The identified STP genes among these tested species were unevenly distributed on the chromosomes (Chrs). Only the ZmSTP22 (gene ID: Zm00001d000183_T001) could not be mapped to any Chr conclusively ( Figure 3 and Table S2). In total, 29 duplicate gene pairs were identified containing two duplicate types: whole genome duplication (WGD)/segmental duplication and tandem duplication in these seven species (Table 1 and Figure 3). Further studies displayed that the numbers and types vary greatly among these six species (Table 1 and Figure 3), which indicates that gene expansion mechanisms of these tested species are different. For example, no duplicate gene pair was found in Hv, whereas nine, five, five, four, three, and three duplicate gene pairs were detected in Si, Or, Os, Zm, Sb, and Bd (Table 1 and Figure 3). Numbers and types of duplicate gene pairs were the same between Os and Or, which supported that wild rice (Or) is the ancestor of cultivated rice (Os) [36]. In addition, the number of WGD/segmental duplication was more than that of tandem duplication in Bd and Zm, while the number of tandem duplication was more than that of WGD/segmental duplication in Or, Os, Sb, and Si. Divergence time of all gene duplicate pairs ranged from 5.57 to 60.90 Mya and varied greatly among these tested species. For example, divergence time of Bd duplicate gene pairs were more than 25 Mya, that of Or and Os ranged from 5 to 31 Mya, that of Sb were more than 23 Mya, that of Si ranged from 15 to 50 Mya, and that of Zm was more than 29 Mya. However, Ka/Ks values of all identified duplicate pairs were less than 1.0 ( Table 2). These results suggested that multiple duplication events play essential roles in the gene expansion of Gramineae crop genomes during the long-term evolutionary process and the main duplication mode are different between these tested species. However, all duplicate genes are under purifying selective forces (all Ka/Ks values of all identified duplicate pairs <1.0).   Figure S1.
Note: Different colors in table represent numbers from lower (green color) to higher (red color). Red OG means single-copy OG among these seven species.  Figure S1.
Additionally, to understand the evolutionary pattern of STP genes in Gramineae crops, the orthologous groups (OGs) were identified by OrthoFinder software. A total of 174 STP genes were assigned into 19 OGs except for three unassigned STP genes ( Figure 1, Table 2, and Tables S3  and S4). We found that the number of genes in each OG is not identical to each other. For instance, OG1 contained 29 genes, while OG16 and OG17 had only four and two genes, respectively. Of these OGs, these seven tested species shared 12 OGs (OG1-OG12). Three single-copy OGs were found, which include OG10, OG11, and OG12. OG13-OG17 only existed in some species. When taking into account these results, we concluded that unequal loss or expansion of most OGs happened during the evolutionary process, but a few OGs were very conservative (e.g., OG10, OG11, and OG12). Additionally, all Tajima's D values of OGs were less than 0, which means the STP gene family was under strong purifying selection (Table 2). We also calculated Tajima's D values for four clades using the same methods and found four Tajima's D values were less than 0. These results indicated that the STP gene family in these tested species was under strong purifying selection. Next, conserved motifs analysis of clades and OGs showed that a total of 20 conserved motifs were found and most STP proteins had similar conserved motifs ( Figure S2B), which indicates the STP domain was strictly conserved in plants. However, we noticed that gene structures of four clades differed greatly, even in one OG ( Figure S2C), which implies STP genes might have been differentiated.

Expression Profiling of Rice STP Genes in 14 Tissues
Expression analysis of one gene family can provide vital clues for functional differentiation [4,32]. We, thus, analyzed expression patterns of rice 19 STP genes during the flower organ development process by qRT-PCR. In the present study, we found that rice STP genes showed diversified expression patterns and most STP genes showed relatively higher expressions in some specific tissues (Figure 4). For example, more than 50% of STP genes showed the predominant expressions in Car and SC of different ages. OsSTP27 were highly expressed in Pan10 and P20, while OsSTP28 and OsSTP14 had relatively high expression in Pan10 and Pan20, separately. In addition, OsSTP17 showed higher expressions in An than other tissues. Our result also illustrated that most STP genes showed a relatively low expression level in SO and ImSe relative to other tissues.
the STP gene family in these tested species was under strong purifying selection. Next, conserved motifs analysis of clades and OGs showed that a total of 20 conserved motifs were found and most STP proteins had similar conserved motifs ( Figure S2B), which indicates the STP domain was strictly conserved in plants. However, we noticed that gene structures of four clades differed greatly, even in one OG ( Figure S2C), which implies STP genes might have been differentiated.

Expression Profiling of Rice STP Genes in 14 Tissues
Expression analysis of one gene family can provide vital clues for functional differentiation [4,32]. We, thus, analyzed expression patterns of rice 19 STP genes during the flower organ development process by qRT-PCR. In the present study, we found that rice STP genes showed diversified expression patterns and most STP genes showed relatively higher expressions in some specific tissues (Figure 4). For example, more than 50% of STP genes showed the predominant expressions in Car and SC of different ages. OsSTP27 were highly expressed in Pan10 and P20, while OsSTP28 and OsSTP14 had relatively high expression in Pan10 and Pan20, separately. In addition, OsSTP17 showed higher expressions in An than other tissues. Our result also illustrated that most STP genes showed a relatively low expression level in SO and ImSe relative to other tissues.  (panicles harvested before heading at lengths of 10/20 cm), SC1/2 (seed coat, 3/15 days after flowering), An (anther, 1 to 3 days before flowering), SO (stigma and ovary, 1 to 3 days before flowering), ImSe (immature endosperm, 15 days after flowering), and Car1/3/5/7/10/15/20 (caryopses, 1/3/5/7/10/15/20 days after flowering) [31]. The color scale represents relative expression levels from the 2 −∆∆CT method: red indicates a high level and green represents a low level of transcript abundance.

Expression Changes of STP Genes in Rice Seedling under Cold, High Temperature, and Submergence Stress
Our result demonstrated that rice STP genes respond differently to submergence, cold temperature, and high-temperature stress, which hints at functional differentiation of STP genes ( Figure 5). In the present result, more genes respond to submergence and high-temperature stress than cold stress ( Figure 5, Table S5). For example, OsSTP14, OsSTP28, OsSTP1, and OsSTP3 were up-regulated under submergence stress, whereas OsSTP11, OsSTP8, OsSTP20, and OsSTP21 showed clear up-regulation under high-temperature stress. It should be noted that OsSTP14 showed apparent up-regulation in all tested time points of submergence and high-temperature stress, as well as two time points of cold temperature stress. than cold stress ( Figure 5, Table S5). For example, OsSTP14, OsSTP28, OsSTP1, and OsSTP3 were up-regulated under submergence stress, whereas OsSTP11, OsSTP8, OsSTP20, and OsSTP21 showed clear up-regulation under high-temperature stress. It should be noted that OsSTP14 showed apparent up-regulation in all tested time points of submergence and high-temperature stress, as well as two time points of cold temperature stress. Figure 5. Expression changes of the rice seedling under cold stress, high-temperature stress, and submergence stress by qRT-PCR. Sub0, Sub1, Sub3, Sub6, and Sub12 represent 0 hours, 1 hours, 3 hours, 6 hours, and 12 hours after the submergence stress. C0, C1, C3, C6, and C12 mean 0 hours, 1 hour, 3 hours, 6 hours, and 12 hours after cold stress (4 °C). H0, H3, H6, and H12 represent 0 hours, 1 hour, 3 hours, 6 hours, and 12 hours after high-temperature stress (40 °C). The color scale represents fold changes (2 −ΔΔCT method): the red color indicates high up-regulation and the green color represents low up-regulation.

Expression Changes of STP Genes in Rice under Blast Infection and Wounding Stress
In this study, most STP genes were down-regulated after rice blast infection. However, several STP genes were up-regulated ( Figure 6). For example, two STP genes (OsSTP8 and OsSTP16) showed significant up-regulation during 0-1 hours, while OsSTP3, OsSTP11, OsSTP14, OsSTP19, OsSTP1, OsSTP10, OsSTP4, and OsSTP28 were up-regulated significantly at various time points during 1-48 hours ( Figure 6, Table S5). Based on our knowledge, the functions of rice STP genes under wounding stress were poorly documented to date. Our results showed that two STP genes (OsSTP11 and OsSTP16) had up-regulation after wounding stress, especially OsSTP11 (Figure 7, Table S5). Moreover, OsSTP28, OsSTP14, and OsSTP4 were up-regulated at various time points after wounding stress (Figure 7, Table S5).

Expression Changes of STP Genes in Rice under Blast Infection and Wounding Stress
In this study, most STP genes were down-regulated after rice blast infection. However, several STP genes were up-regulated ( Figure 6). For example, two STP genes (OsSTP8 and OsSTP16) showed significant up-regulation during 0-1 h, while OsSTP3, OsSTP11, OsSTP14, OsSTP19, OsSTP1, OsSTP10, OsSTP4, and OsSTP28 were up-regulated significantly at various time points during 1-48 h ( Figure 6, Table S5). Based on our knowledge, the functions of rice STP genes under wounding stress were poorly documented to date. Our results showed that two STP genes (OsSTP11 and OsSTP16) had up-regulation after wounding stress, especially OsSTP11 (Figure 7, Table S5). Moreover, OsSTP28, OsSTP14, and OsSTP4 were up-regulated at various time points after wounding stress (Figure 7, Table S5).

Discussion
The availability of various Gramineae crops' genomes has provided great opportunity to explore the short-term evolutionary dynamics of the STP gene family in Gramineae crops. In this study, 27,26,28,26,25,23, and 22 STP genes were identified in Hv, Bd, Os, Or, Si, Sb, and Zm, respectively. We found no direct relevance between genome sizes and the number of STP genes. For example, there were 26 STP genes in Bd (genome size: 355 Mbp), while there were 23 STP genes in Sb (genome size: 730 Mbp) (Figure 1). Z. mays has reportedly undergone one specific WGDs more than other Gramineae plants [37]. However, we found that maize did not have more STP genes than other species. Instead, maize had the fewest STP genes in the tested species in this study (only 22). We speculated that a large number of gene losses occur after the gene duplication events. This speculation was supported by OGs results in which the lesser gene number was found in Zm than other tested species among key OGs, such as OG1 and OG2 (Table 2). Additionally, OG15, OG16, and OG17 were lost and 7 OGs containing only one member were retained, which included OG7-OG12, and OG13 (Table 2). Earlier studies reported that tandem, WGDs, segmental duplication events are the main driving forces for the expansion of one gene family [11,32,38]. Our results of identified duplication events and divergence time of duplicate gene pairs also indicated that tandem and WGDs/segmental duplication events play essential roles in the gene expansion of some Gramineae crops (expect for Hv). Notably, of these tested species, expansion mechanisms of the STP gene family were different. Furthermore, Hv had the most STP genes relative to other tested species. However, no tandem, WGDs/segmental duplication events were found. This result suggested that the expansion of the Hv STP gene family may result from other duplication types, such as proximal, dispersed, and replicative transposition.
Our clade classifications supported previous classification results that STP genes can be divided into four main clades [5,11]. We found most STP proteins had similar conserved motifs, while gene structures of four clades were greatly different, even in one OG ( Figure S2), which implies the function of STP genes has been differentiated. Our gene expression profiling of STP genes also confirmed the functional differentiation of STP genes due to various expression patterns. Moreover, Ka/Ks results of all duplicate gene pairs and Tajima's D values of all OGs or four clades both suggested that the STP gene family in Gramineae crops was under purifying selective forces. In addition, we found that a total of 17 OGs were identified. Further studies showed that OG1 contained 29 genes, while OG16 and OG17 had only four genes and two genes, respectively. These results indicated that different OGs have occurred at different degrees of expansion and loss. For example, OG1 and OG2 contained many duplication events, while OG16 and OG17 did not contain any duplication event. We hypothesized that the current difference in gene numbers for all OGs may be the result of natural or artificial selection. Beneficial genes to plants gained more copies, while redundant or pseudo-genes were lost in evolution. This speculation needs further verification in future work.
Previous studies reported orthologous with the same expression patterns, gene structure, and motifs [11,16,39]. We, thus, detected STP gene expression levels in multiple tissues and under several stresses for preliminary insights into the functional analysis of rice STP genes and orthologous in other tested species. In our expression result, STP genes have different expression patterns in diverse tissues and some STP genes with tissue-specific expressions. These results were consistent with previous findings on other species [11,13,18]. We found most STP genes showed the predominant expressions in Car and SC of different ages, which indicates STP genes play essential roles in Car and SC development. Previous studies reflected that STP genes play crucial roles in stresses and disease responses [4,11,13,26]. For example, wounding induces AtSTP3 [26]. Our result showed several STP genes with strong responses to cold, high temperature, and submergence stresses, which indicates that STP genes are associated with these stresses. We noted that OsSTP14 could be one good candidate gene for abiotic stress resistance breeding. In addition, OsSTP11 showed strong up-regulated responses to wounding, but the actual mechanism needs to be further explored in the future.

Conclusions
In the present study, 27,26,28,26,25,23, and 22 STP genes were identified in H. vulgare (Hv), B. distachyon (Bd), O. sativa ssp. japonica (Os), O. rufipogon (Or), S. italica (Si), S. bicolor (Sb), and Z. mays (Zm), respectively. Then, a systemic analysis, including phylogenetic relationship, chromosomal location, gene structure, protein motifs, duplication events, orthologous groups (OGs), selective forces, and expression patterns of rice STP genes were conducted. Our result revealed different expansion mechanisms of the STP gene family among these tested species and reported differential selective pressures of different clades and OGs. The expression result from qRT-PCR suggested that STP genes play essential roles in plant development and stress responses. We believe that OsSTP14 and its orthologous may be a good candidate gene for plant resistance breeding.
Supplementary Materials: Supplementary Materials are available online at http://www.mdpi.com/2073-4409/8/6/ 560/s1, Table S1: Primers of rice STP genes were used for qRT-PCR in this study. Table S2: The detailed information on rice monosaccharide STP genes in seven tested species. Table S3: Orthologous groups among these seven species. Table S4: Statistical result of orthologous groups among these seven species. Table S5: Statistical analysis (Fold change, P value, Benjamini-Hochberg corrected p-value) of Figures 5-7. Figure S1: A high resolution version of Figure 3. Figure S2: The phylogenetic tree (A), conserved motif compositions (B), and exon/intron structure of the STP genes in these seven species (C). A NJ phylogeny tree of STP protein sequences from H. vulgare, B. distachyon, O. sativa ssp. japonica, O. rufipogon, S. italic, S. bicolor, and Z. mays. Different colors of branches represent different clades. Different shaped markers indicated the different species. The numbers in blue boxes means different OGs. The widths of the grey bars showed the relative lengths of genes and proteins. Yellow boxes and grey lines displayed the exons and introns, respectively.