Genome-Wide Identification and Expression Profiling of Sugar Transporter Protein (STP) Family Genes in Cabbage (Brassica oleracea var. capitata L.) Reveals their Involvement in Clubroot Disease Responses

Sugar transporter protein (STP) genes are involved in multiple biological processes, such as plant responses to various stresses. However, systematic analysis and functional information of STP family genes in Brassica oleracea are very limited. A comprehensive analysis was carried out to identify BoSTP genes and dissect their phylogenetic relationships and to investigate the expression profiles in different organs and in response to the clubroot disease. A total of 22 BoSTP genes were identified in the B. oleracea genome and they were further classified into four clades based on the phylogenetic analysis. All the BoSTP proteins harbored the conserved sugar transporter (Sugar_tr, PF00083) domain, and the majority of them contained 12 transmembrane helices (TMHs). Rates of synonymous substitution in B. oleracea relative to Arabidopsis thaliana indicated that STP genes of B. oleracea diverged from those of A. thaliana approximately 16.3 million years ago. Expression profiles of the BoSTP genes in different organs derived from RNA-Seq data indicated that a large number of the BoSTP genes were expressed in specific organs. Additionally, the expression of BoSTP4b and BoSTP12 genes were induced in roots of the clubroot-susceptible cabbage (CS-JF1) at 28 days after inoculation with Plasmodiophora brassicae, compared with mock-inoculated plants. We speculated that the two BoSTPs might be involved in monosaccharide unloading and carbon partitioning associated with P. brassicae colonization in CS-JF1. Subcellular localization analysis indicated that the two BoSTP proteins were localized in the cell membrane. This study provides insights into the evolution and potential functions of BoSTPs.


Introduction
Sugars (e.g., monosaccharides, sucrose, and polyols) act as carbohydrate molecules, main energy sources, precursors of cellular compounds, and signaling molecules for signal transduction as well as environmental stress responses, which are important for plant growth and development [1][2][3][4]. Sugars are mainly synthesized in leaves (source organs) and then translocated via phloem sap into the sink organs, such as modified leaves, roots, seeds, fruits, and other reproductive organs [5,6]. In plants, sugar transport is mediated by monosaccharide transporters (MSTs) and sucrose transporters (SUTs) and sugars will eventually be exported transporters (SWEETs) [7,8]. The sucrose can be transported from the phloem to sink cells via a symplastic pathway or an apoplastic pathway [6]. Apart from The TMHs of BoSTP proteins were predicted by the TMHMM Server v. 2.0 (http://www.cbs. dtu.dk/services/TMHMM/). The putative functional domains of the identified BoSTP proteins were predicted by hmmscan [39] and SMART [40]. The subcellular localization of STP proteins was predicted using Plant-mPLoc [50].

Expression Profile Analysis of BoSTP Genes
To analyze the expression profiles of BoSTP genes in different organs and in response to P. brassicae infection, RNA-Seq data of GSE42891 (GEO database) submitted by Liu et al. [37] and PRJNA453960 (BioProject accession) submitted by Chinese Academy of Agricultural Sciences (CAAS) were downloaded from the National Center for Biotechnology Information (NCBI, https: //www.ncbi.nlm.nih.gov/). The GSE42891 dataset included gene expression profiles of seven different organs (bud, callus, flower, leaf, root, silique, and stem) in cabbage homozygous line 02-12 [37]. The PRJNA453960 dataset contained gene expression profiles of roots in clubroot-resistant cabbage Xiangan336 (CR-XG336) and clubroot-susceptible cabbage Jingfeng No. 1 (CS-JF1) after the inoculation with P. brassicae spores suspended in water at 7 (primary infection stage) and 28 (clubroot formation stage) days after inoculation (DAI), respectively. Furthermore, the corresponding root samples at 7 DAI and 28 DAI without inoculation were sampled as mock controls. Raw RNA-Seq reads with three biological replicates from the PRJNA453960 dataset were processed to trim the adapter and low-quality Genes 2019, 10, 71 4 of 17 sequences using Trimmomatic [51]. The high-quality reads were aligned to the B. oleracea genome database using HISAT [52] allowing up to 3 edit distances. Following the alignments, raw counts for each gene were derived and normalized into FPKM (fragments per kilobase of exon model per million mapped fragments). Raw count data was then fed to DESeq2 [53] to identify differentially expressed genes with a cutoff of fold change >2 and FDR <0.05. The heatmap of BoSTPs expression profiles was generated using the R package pheatmap (http://rpackages.ianhowson.com/cran/pheatmap/) based on the log2 transformed FPKM values.

Subcellular Localization Analysis of BoSTPs in Tobacco
To investigate the subcellular localization of BoSTPs, transient transformation of translational GFP fusion vectors into tobacco leaf epidermal cells was used. The full-length CDS of BoSTP4b and BoSTP12 genes were amplified with the forward primer containing a Sma I restriction site and the reverse primer containing a Xba I restriction site. The primer sequences are listed in Supplementary File S1. The reaction mixture (50 µL) of the gene amplification consisted of 1 × PCR buffer (final concentrations, 10 mM Tris-HCl, 2.5 mM MgCl2, and 50 mM KCl [pH 8.3]), 0.2 mM dNTP, 0.5 µM forward and reverse primer, 50 ng of cDNA template and 2.5 U of Taq DNA polymerase (TaKaRa, Dalian, China). The amplification was performed at 95 • C for 3 min, followed by 35 cycles of 94 • C for 30 s, 56 • C for 30 s, and 72 • C for 50 s, with a final extension at 72 • C for 7 min and hold at 4 • C. The amplification products were digested with Sma I and Xba I, and ligated into the pCAMBIA1300-35S-GFP (35S-GFP) vector for constructing the translational GFP fusion constructs. The recombined plasmids were then transformed into Agrobacterium tumefaciens strain GV3101 [54]. The methods of preparing A. tumefaciens for transient expression and infiltrating into tobacco epidermal cells were carried out according to the previous protocols [55]. Leaves transformed with the vector of 35S-GFP (GFP alone) was used as the control. The fluorescence and bright-light images of transiently infected tobacco leaves were observed under a fluorescence microscope (BX41, Olympus, Rungis, France) after 48 h of infiltration.

Identification and Phylogenetic Analysis of STPs in Brassica oleracea, Brassica rapa, and Arabidopsis thaliana
In the present study, a total of 22 BoSTP proteins and 22 BrSTP proteins were identified in the B. oleracea and B. rapa genomes, respectively (Supplementary File S2). These proteins were named BoSTP1 to BoSTP14 and BrSTP1 to BrSTP14, respectively, according to their A. thaliana homologs (Table 1). An unrooted phylogenetic tree of 58 STP proteins from B. oleracea, B. rapa, and A. thaliana was constructed to elucidate their evolutionary relationships ( Figure 1). Phylogenetic analysis revealed that these STP proteins could be divided into four distinct clades (I-IV), and the BoSTPs were grouped with AtSTPs and BoSTPs in each clade. The clade I, II, III, and IV contained 14, 16, 15, and 13 members, respectively. Physical and chemical characteristics of BoSTP proteins were predicted. The length of the BoSTP proteins ranged from 345 amino acids (AA, BoSTP14) to 532 AA (BoSTP13b), with the corresponding open reading frames (ORFs) ranging from 1038 bp to 1599 bp ( Table 1). The molecular weights ranged from 38.33 kDa (BoSTP14) to 58.04 kDa (BoSTP13b), and the theoretical isoelectric points (pI) ranged from 5.79 (BoSTP11) to 9.31 (BoSTP13a) ( Table 1). In addition, the instability index, aliphatic index and GRAVY of BoSTP proteins ranged from 30.06 to 45.45, 97.50 to 112.94, and 0.45 to 0.69, respectively (Table 1).   (22), B. rapa (22), and A. thaliana (14) showing similar groups in the three species. Four clades were marked with different background colors.

Chromosomal Distribution of BoSTP Genes
The BoSTP genes were distributed across the three sub-genomes (LF, the least fractionated blocks of Brassica; MF1, the medium fractionated blocks of Brassica; MF2, the most fractionated blocks of Brassica), with 12 BoSTP genes in LF, six in MF1, and four in MF2 (Table 1). Interestingly, all the STP genes were retained in B. oleracea after whole-genome triplication (WGT) and fractionation. In total, half of the BoSTP genes (7/14) were retained in one copy, while six BoSTP genes (BoSTP1, BoSTP2, BoSTP4, BoSTP7, BoSTP9, and BoSTP13) and only one (BoSTP6) from B. oleracea were retained in two and three copies, respectively ( Table 1). The retained gene copies had the same conserved collinear blocks. In addition, according to the chromosomal localization of BoSTP genes ( Figure 2), a total of 15 BoSTP genes (68.2%) were distributed across the nine chromosomes (C01-C09), with the exception of chromosomes C02 and C04, which had no members, and with the most (four) on chromosome C06 ( Figure 2). Furthermore, seven BoSTP genes (BoSTP4b, BoSTP5, BoSTP6a, BoSTP6b, BoSTP6c, BoSTP9b, and BoSTP10) were not anchored on any of the B. oleracea chromosomes. The identified orthologous and paralogous STP genes were used to analyze the syntenic relationship between BoSTP and BrSTP genes. A total of 24 orthologous STP gene pairs between B. oleracea and B. rapa were identified ( Figure 3). In addition, four and eight paralogous STP gene pairs were identified in B. oleracea and B. rapa, respectively.  (22), and A. thaliana (14) showing similar groups in the three species. Four clades were marked with different background colors.

Chromosomal Distribution of BoSTP Genes
The BoSTP genes were distributed across the three sub-genomes (LF, the least fractionated blocks of Brassica; MF1, the medium fractionated blocks of Brassica; MF2, the most fractionated blocks of Brassica), with 12 BoSTP genes in LF, six in MF1, and four in MF2 (Table 1). Interestingly, all the STP genes were retained in B. oleracea after whole-genome triplication (WGT) and fractionation. In total, half of the BoSTP genes (7/14) were retained in one copy, while six BoSTP genes (BoSTP1, BoSTP2, BoSTP4, BoSTP7, BoSTP9, and BoSTP13) and only one (BoSTP6) from B. oleracea were retained in two and three copies, respectively ( Table 1). The retained gene copies had the same conserved collinear blocks. In addition, according to the chromosomal localization of BoSTP genes ( Figure 2), a total of 15 BoSTP genes (68.2%) were distributed across the nine chromosomes (C01-C09), with the exception of chromosomes C02 and C04, which had no members, and with the most (four) on chromosome C06 ( Figure 2). Furthermore, seven BoSTP genes (BoSTP4b, BoSTP5, BoSTP6a, BoSTP6b, BoSTP6c, BoSTP9b, and BoSTP10) were not anchored on any of the B. oleracea chromosomes. The identified orthologous and paralogous STP genes were used to analyze the syntenic relationship between BoSTP and BrSTP genes. A total of 24 orthologous STP gene pairs between B. oleracea and B. rapa were identified ( Figure 3). In addition, four and eight paralogous STP gene pairs were identified in B. oleracea and B. rapa, respectively.

Ka and Ks Calculation of Orthologous STPs between Brassica oleracea and Arabidopsis thaliana
According to the 22 syntenic STP orthologous pairs between B. oleracea and A. thaliana, the Ks and Ka values of these homologous genes were calculated by DnaSP 6 software ( Table 2; Supplementary File S3). The Ka/Ks values of all the syntenic STP orthologous pairs were far less than 1, suggesting that these STP genes have undergone purifying selection. The Ks values of B. oleracea relative to A. thaliana ranged from 0.43 to 0.60 and concentrated at 0.49 (Figure 4), which indicated that these STP genes of B. oleracea diverged from A. thaliana approximately 16.3 MYA. We concluded that these STP genes diverged following the Brassica-specific WGT event that occurred approximately 13-17 MYA [38,56].

Ka and Ks Calculation of Orthologous STPs between Brassica oleracea and Arabidopsis thaliana
According to the 22 syntenic STP orthologous pairs between B. oleracea and A. thaliana, the Ks and Ka values of these homologous genes were calculated by DnaSP 6 software (  (Figure 4), which indicated that these STP genes of B. oleracea diverged from A. thaliana approximately 16.3 MYA. We concluded that these STP genes diverged following the Brassica-specific WGT event that occurred approximately 13-17 MYA [38,56].

Gene Structures, TMHs, and Putative Functional Domains of BoSTP Proteins
The exon-intron organizations are thought to play critical roles in the evolution of multiple gene families [57]. To gain insights into the structural evolution of the BoSTP genes, the exon-intron boundaries were dissected. The results indicated that the number and distribution of exons and introns were highly conserved in BoSTP genes ( Figure 5). A total of 16 (72.7%) BoSTP genes were found to have four exons and three introns, while five BoSTP members possessed three exons and two introns, and only BoSTP14 had two exons and one intron ( Figure 5). The results suggested that the gain and loss of exons might result in functional diversity of closely related STP genes during the evolutionary process of the STP gene family. Due to the significance of the sugar transporter domain for the catalytic activity, the functional domains of 22 BoSTP proteins were predicted, which revealed that all the BoSTP proteins harbored a Sugar_tr domain (PF00083) belonging to the clan major facilitator superfamily (MFS; CL0015). In addition, all the BoSTP proteins except BoSTP14 harbored an MFS domain (MFS_1 domain; PF07690) (Supplementary File S4). Among these BoSTP proteins, 15 harbored 12 conserved TMHs, five harbored 11 TMHs, and BoSTP10 and BoSTP14 harbored 10 and eight TMHs, respectively (Figure 5c). The majority of BoSTP proteins contained 12 TMHs, which is a typical feature of MFS, suggesting the activities were related to substance transportation [58].

Gene Structures, TMHs, and Putative Functional Domains of BoSTP Proteins
The exon-intron organizations are thought to play critical roles in the evolution of multiple gene families [57]. To gain insights into the structural evolution of the BoSTP genes, the exon-intron boundaries were dissected. The results indicated that the number and distribution of exons and introns were highly conserved in BoSTP genes ( Figure 5). A total of 16 (72.7%) BoSTP genes were found to have four exons and three introns, while five BoSTP members possessed three exons and two introns, and only BoSTP14 had two exons and one intron ( Figure 5). The results suggested that the gain and loss of exons might result in functional diversity of closely related STP genes during the evolutionary process of the STP gene family. Due to the significance of the sugar transporter domain for the catalytic activity, the functional domains of 22 BoSTP proteins were predicted, which revealed that all the BoSTP proteins harbored a Sugar_tr domain (PF00083) belonging to the clan major facilitator superfamily (MFS; CL0015). In addition, all the BoSTP proteins except BoSTP14 harbored an MFS domain (MFS_1 domain; PF07690) (Supplementary File S4). Among these BoSTP proteins, 15 harbored 12 conserved TMHs, five harbored 11 TMHs, and BoSTP10 and BoSTP14 harbored 10 and eight TMHs, respectively (Figure 5c). The majority of BoSTP proteins contained 12 TMHs, which is a typical feature of MFS, suggesting the activities were related to substance transportation [58].

Prediction of cis-Acting Elements of BoSTP Genes
In this study, the 1500-bp promoter sequences of 14 BoSTP genes were obtained. For the remaining eight BoSTP genes, the obtained promoter sequences were shorter than 1500 bp due to the presence of other genes located <1500 bp upstream. The PlantCARE web tool (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) was used to predict the cis-acting elements. As shown in Table 3, a total of 17 cis-acting elements were predicted in the promoters of the 22 BoSTP genes. The common cis-acting elements included those responsive to distinct plant hormones (ABA, auxin, ethylene, gibberellin, meJA, and salicylic acid) and stress factors (anaerobic, drought, light, low temperature, and wound) ( Table 3). Among these elements, eight common cisacting elements, including ABRE, ARE, CGTCA-motif, ERE, G-box, GT1-motif, STRE, and TGACGmotif, were highly conserved.

Prediction of cis-Acting Elements of BoSTP Genes
In this study, the 1500-bp promoter sequences of 14 BoSTP genes were obtained. For the remaining eight BoSTP genes, the obtained promoter sequences were shorter than 1500 bp due to the presence of other genes located <1500 bp upstream. The PlantCARE web tool (http://bioinformatics.psb.ugent. be/webtools/plantcare/html/) was used to predict the cis-acting elements. As shown in Table 3, a total of 17 cis-acting elements were predicted in the promoters of the 22 BoSTP genes. The common cis-acting elements included those responsive to distinct plant hormones (ABA, auxin, ethylene, gibberellin, meJA, and salicylic acid) and stress factors (anaerobic, drought, light, low temperature, and wound) ( Table 3). Among these elements, eight common cis-acting elements, including ABRE, ARE, CGTCA-motif, ERE, G-box, GT1-motif, STRE, and TGACG-motif, were highly conserved.

Expression Profiling of BoSTP Genes in Different Organs and in Response to P. brassicae Infection
Organ-specific gene expression can provide valuable information about the function of BoSTPs in different organs. Thus, the transcript levels of the BoSTP genes in the bud, callus, flower, leaf, root, silique, and stem of cabbage derived from the RNA-Seq dataset (GSE42891) were analyzed. The heatmap revealed that a large portion of BoSTP genes were organ-specific (Figure 6a; Supplementary File S5). For example, several BoSTP genes were expressed in only one or two organ types, such as BoSTP2a in buds, BoSTP13a in leaves, BoSTP2b in buds and calli, and BoSTP6a and BoSTP10 in buds and flowers (Figure 6a). A total of seven BoSTP genes were expressed in all the organs, including BoSTP1a, BoSTP1b, BoSTP4a, BoSTP4b, BoSTP7a, BoSTP12, and BoSTP13b. The diversity of the expression profiles of BoSTP genes suggested a broad range of biological functions during the growth and development of cabbage.

Expression Profiling of BoSTP Genes in Different Organs and in Response to P. brassicae Infection
Organ-specific gene expression can provide valuable information about the function of BoSTPs in different organs. Thus, the transcript levels of the BoSTP genes in the bud, callus, flower, leaf, root, silique, and stem of cabbage derived from the RNA-Seq dataset (GSE42891) were analyzed. The heatmap revealed that a large portion of BoSTP genes were organ-specific (Figure 6a; Supplementary File S5). For example, several BoSTP genes were expressed in only one or two organ types, such as BoSTP2a in buds, BoSTP13a in leaves, BoSTP2b in buds and calli, and BoSTP6a and BoSTP10 in buds and flowers (Figure 6a). A total of seven BoSTP genes were expressed in all the organs, including BoSTP1a, BoSTP1b, BoSTP4a, BoSTP4b, BoSTP7a, BoSTP12, and BoSTP13b. The diversity of the expression profiles of BoSTP genes suggested a broad range of biological functions during the growth and development of cabbage. We also generated expression profiles of BoSTP genes in roots of CR-XG336 (clubroot-resistant line) and CS-JF1 (clubroot-susceptible line) infected by P. brassica, and observed the expression variation of BoSTP genes between the two different cultivars at two different infection stages ( Figure  6b; Supplementary File S6). A total of five BoSTP genes (BoSTP2a, BoSTP6a, BoSTP9b, BoSTP10, and BoSTP11) were undetectable in both CR-XG336 and CS-JF1 under the two infection stages. In CS-JF1, compared with mock-inoculated plants, two BoSTPs (BoSTP4b and BoSTP12) and two BoSTPs (BoSTP1a and BoSTP1b) were significantly up-and down-regulated at 7 DAI and 28 DAI, respectively. No BoSTP genes were significantly up-or down-regulated in CR-XG336 at 7 DAI compared with mock-inoculated plants (Figure 6b; Supplementary File S6). Moreover, only the BoSTP3 gene was significantly up-regulated in CR-XG336 after infection at 28 DAI (Figure 5b; Supplementary File S6).

Subcellular Localization Analysis of BoSTP Proteins in Tobacco
In silico subcellular localization prediction using Plant-mPLoc [50] suggested that all the BoSTP proteins were localized in the cell membrane. To further characterize the subcellular localization of BoSTP proteins, BoSTP4b and BoSTP12 translational GFP fusion proteins were heterologously and transiently expressed in tobacco leaf epidermal cells. The GFP signal in epidermal cells transfected with the 35S-GFP was mainly found in the cell membrane, cytoplasm, and nucleus (Figure 7), whereas the GFP signals in epidermal cells transfected with 35S-BoSTP4b-GFP and 35S-BoSTP12-GFP We also generated expression profiles of BoSTP genes in roots of CR-XG336 (clubroot-resistant line) and CS-JF1 (clubroot-susceptible line) infected by P. brassica, and observed the expression variation of BoSTP genes between the two different cultivars at two different infection stages (Figure 6b; Supplementary File S6). A total of five BoSTP genes (BoSTP2a, BoSTP6a, BoSTP9b, BoSTP10, and BoSTP11) were undetectable in both CR-XG336 and CS-JF1 under the two infection stages. In CS-JF1, compared with mock-inoculated plants, two BoSTPs (BoSTP4b and BoSTP12) and two BoSTPs (BoSTP1a and BoSTP1b) were significantly up-and down-regulated at 7 DAI and 28 DAI, respectively. No BoSTP genes were significantly up-or down-regulated in CR-XG336 at 7 DAI compared with mock-inoculated plants (Figure 6b; Supplementary File S6). Moreover, only the BoSTP3 gene was significantly up-regulated in CR-XG336 after infection at 28 DAI (Figure 5b; Supplementary File S6).

Subcellular Localization Analysis of BoSTP Proteins in Tobacco
In silico subcellular localization prediction using Plant-mPLoc [50] suggested that all the BoSTP proteins were localized in the cell membrane. To further characterize the subcellular localization of BoSTP proteins, BoSTP4b and BoSTP12 translational GFP fusion proteins were heterologously and transiently expressed in tobacco leaf epidermal cells. The GFP signal in epidermal cells transfected with the 35S-GFP was mainly found in the cell membrane, cytoplasm, and nucleus (Figure 7), whereas the GFP signals in epidermal cells transfected with 35S-BoSTP4b-GFP and 35S-BoSTP12-GFP fusion proteins were detected in the cell membrane ( Figure 6), suggesting that the two BoSTPs are membrane proteins, consistent with the in silico prediction results.
Genes 2019, 10, x FOR PEER REVIEW 12 of 17 fusion proteins were detected in the cell membrane ( Figure 6), suggesting that the two BoSTPs are membrane proteins, consistent with the in silico prediction results.

Discussion
The availability of the B. oleracea genome sequence has provided great opportunity to explore the STP gene family members, and to investigate their phylogenetic relationships and potential functional roles. The STP gene family is widely distributed in the plant kingdom and plays critical roles in sugar transport, plant growth, and plant responses to pathogen attack and wounding [12,25,59]. To our knowledge, no systematic investigations on the STP gene family of B. oleracea have been reported, and the expression profiles and functional significances of the BoSTP genes should be further investigated.
In this study, we found that the number of STP genes among B. oleracea, B. rapa, and A. thaliana was variable, with 14 genes in A. thaliana and 22 in both B. oleracea and B. rapa. The number of BoSTP and BrSTP genes were more than that of AtSTP genes, possibly due to the Brassica-specific WGT and fractionation events [38]. Gene duplication is considered as a primary driving force for evolution, resulting in functional divergence and diversification [60,61]. In this study, we found evidence of gene replication and gene loss during polyploid speciation in the B. oleracea genome. After the split from A. thaliana, there were seven BoSTP genes with one separate orthologous gene, six BoSTP genes with two separate orthologous genes, and only one BoSTP6 gene with three separate orthologous genes (Table 1; Figure 1). These results may indicate functional redundancy among several BoSTP genes.

Discussion
The availability of the B. oleracea genome sequence has provided great opportunity to explore the STP gene family members, and to investigate their phylogenetic relationships and potential functional roles. The STP gene family is widely distributed in the plant kingdom and plays critical roles in sugar transport, plant growth, and plant responses to pathogen attack and wounding [12,25,59]. To our knowledge, no systematic investigations on the STP gene family of B. oleracea have been reported, and the expression profiles and functional significances of the BoSTP genes should be further investigated.
In this study, we found that the number of STP genes among B. oleracea, B. rapa, and A. thaliana was variable, with 14 genes in A. thaliana and 22 in both B. oleracea and B. rapa. The number of BoSTP and BrSTP genes were more than that of AtSTP genes, possibly due to the Brassica-specific WGT and fractionation events [38]. Gene duplication is considered as a primary driving force for evolution, resulting in functional divergence and diversification [60,61]. In this study, we found evidence of gene replication and gene loss during polyploid speciation in the B. oleracea genome. After the split from A. thaliana, there were seven BoSTP genes with one separate orthologous gene, six BoSTP genes with two separate orthologous genes, and only one BoSTP6 gene with three separate orthologous genes (Table 1; Figure 1). These results may indicate functional redundancy among several BoSTP genes.
To investigate the extent of gene fractionation in sub-genomes of B. oleracea relative to A. thaliana, three sub-genomes of B. oleracea were established [37]. In this investigation, a total of 12 BoSTP genes belonged to the LF sub-genome, six belonged to the MF1 sub-genome, and four belonged to the MF2 sub-genome (Table 1). The results indicated that the LF sub-genome contains the most BoSTP genes, consistent with that the LF sub-genome retains 70% of the genes found in A. thaliana [38]. All the BoSTP proteins harbored the Sugar_tr domain, and the majority of BoSTP proteins contained 12 TMHs, a typical feature of the major facilitator superfamily (MFS). The 12-TMH structure of STP proteins has evolved from an ancestral six-TMH transporter following gene duplication and fusion [58]. The STP family members belonging to MFS transporters generally possess a large central loop within the two transmembrane domains, the N-domain (TMH1-TMH6) and the C-domain (TMH7-TMH12) [11,14], and transmembrane transport is driven by the proton motive force [62]. A total of 17 cis-acting elements responsive to plant hormones and various stresses were predicted in the upstream sequences of the 22 BoSTP genes. Furthermore, the identified cis-acting elements, such as G-box, GT1-motif, and MRE, were transcriptionally regulated under light responsiveness, consistent with the functions of STP genes in sugar allocation between sink and source organs [17]. The STP genes play critical roles in the distribution of monosaccharides, which were found to be involved in various metabolic processes during plant growth and development [63].
For the BoSTP gene members, we were particularly interested in those that might play crucial roles in clubroot disease responses. Previous studies revealed that P. brassicae obtains sugars from hosts to complete its life cycle, involving the development and formation of galls, which act as an additional sink [64]. In A. thaliana, it has been found that sucrose accumulates in uninfected leaves, but not in P. brassicae infected leaves because sucrose is exported from leaves into the clubroot galls [65]. The expression of sugar transporter genes may therefore influence plant-P. brassicae interactions. In this study, the RNA-Seq dataset (PRJNA453960) of cabbage were further analyzed for dissecting the expression profiles of the BoSTP genes. The expression of BoSTP4b and BoSTP12 were up-regulated in roots of the susceptible CS-JF1 upon P. brassicae inoculation at the clubroot formation stage (28 DAI) compared with mock-inoculated plants ( Figure 5; Supplementary File S6). We inferred that the two BoSTP genes might play roles in sugar partitioning of clubroot-induced gall development and formation in CS-JF1. In Arabidopsis, the expression of the hexose transporter gene, AtSTP4, and the cell wall invertase gene, Atβfruct1, are induced after the powdery mildew infection. At the meantime, the uptake of glucose in host organs is enhanced substantially. The coordinated expression of AtSTP4 and Atβfruct1 might functionally interact for the supply of sink organs with hexoses [25]. In addition, the expression of AtSTP8 (transporting hexose) and AtSTP13 (transporting galactose) are up-regulated after infection. AtSTP8 and AtSTP13 are associated with programmed cell death in pathogen defense and cell wall remodeling [20,64]. We also noted that the expression of the BoSTP3 gene was 3.68-fold higher in CR-XG336 at 28 DAI than in the mock-inoculated plants (Supplementary File S6). Whether this increased expression of BoSTP3 contributes to CR-XG336 resistance to P. brassicae infection remains to be established.

Conclusions
In this study, a total of 22 BoSTP genes were identified in the B. oleracea genome and they were further classified into four clades based on a phylogenetic tree of 58 STP homologs from B. oleracea, B. rapa, and A. thaliana. The length of the BoSTP proteins ranged from 345 to 532 AA, and the Ks values of the orthologous STP genes in B. oleracea relative to A. thaliana ranged from 0.43 to 0.60 and concentrated at 0.49, suggesting that the estimated time of B. oleracea diverged from A. thaliana was approximately 16.3 MYA. RNA-Seq data analysis of seven organs in cabbage indicated that a large number of BoSTP genes exhibited organ-specific expression. The expression of two BoSTP genes (BoSTP4b and BoSTP12) were up-regulated in CS-JF1 at 28 DAI with P. brassicae infection compared with mock-inoculated plants, and we hypothesized that they might be involved in monosaccharide unloading and partitioning for the clubroot development and formation during P. brassicae colonization.
These results help to understand the characteristics of BoSTP genes and their potential functions in regulating monosaccharide distribution during the clubroot disease responses in cabbage.