Genome-wide identification and expression analysis of the xyloglucan endotransglucosylase/hydrolase gene family in poplar

Xyloglucan endotransglucosylase/hydrolase (XTH) family plays an important role in cell wall reconstruction and stress resistance in plants. However, the detailed characteristics of XTH family genes and their expression pattern under salt stress have not been reported in poplar. In this study, a total of 43 PtrXTH genes were identified from Populus simonii × Populus nigra, and most of them contain two conserved structures (Glyco_hydro_16 and XET_C domain). The promoters of the PtrXTH genes contain mutiple cis-acting elements related to growth and development and stress responses. Collinearity analysis revealed that the XTH genes from poplar has an evolutionary relationship with other six species, including Eucalyptus robusta, Solanum lycopersicum, Glycine max, Arabidopsis, Zea mays and Oryza sativa. Based on RNA-Seq analysis, the PtrXTH genes have different expression patterns in the roots, stems and leaves, and many of them are highly expressed in the roots. In addition, there are11 differentially expressed PtrXTH genes in the roots, 9 in the stems, and 7 in the leaves under salt stress. In addition, the accuracy of RNA-Seq results was verified by RT-qPCR. All the results indicated that XTH family genes may play an important role in tissue specificity and salt stress response. This study will lay a theoretical foundation for further study on molecular function of XTH genes in poplar.


Background
As the external supporting structure of cells, cell wall largely determines the shape and size of cells in the process of plant growth and development. The main components of plant cell wall are cellulose, lignin, hemicellulose and pectin [1]. Xyloglucan is the most important hemicellulose in the primary cell wall of dicotyledons [2]. Xyloglucan endotransglucosylase/hydrolase (XTH) is widely present in plant cells, which can catalyze the cleavage and polymerization of xyloglucan molecules, and achieve cell remodeling by regulating the elasticity and ductility of cells [3]. It is a key enzyme in the process of cell wall reconstruction in plants.
XTH family genes contain a typical catalyze enzymatic reaction motif (HDEIDFEFLG), in which the first glutamic acid residue (E) is affinity site and the second is proton donor [4,5]. XTHs can be used to form a covalent glycosylase intermediate, which can be decomposed by water to carry out hydrolysis reaction, or enter the sugar substrate to produce transglycosylation [3]. According to the conserved motifs, the XTH family can be divided into three groups (I, II, and III), and the group III is divided into IIIA and IIIB group [6]. It has been found that the XTHs in groups I, II and IIIB have significant xyloglucan endohydrolase (XET) activity, while the XTHs in IIIA show xyloglucan endoglucosidase (XEH) activity [7]. Moreover, a small part of outlier group (early diverting) was found in detailed clade [7,8]. In summary, XTH family is divided into I / II, IIIA, IIIB and the early diverting groups [9].
The expression characteristics of many XTH family genes are different in different tissues during different developmental stages. For example, poplar PtxtXET16-34 was specifically expressed in the developing wood, and overexpression of the gene could increase xyloglucan content in the xylem of the primary wall and promote the growth of vessel element [10]. In Arabidopsis, AtXTH17, AtXTH18, AtXTH19 and AtXTH20 were specifically expressed in roots and play an important role in root elongation and root hair initiation [11,12]. AtXTH21 participates in the growth of primary roots by changing the cellulose deposition and the extensibility in the cell wall [13]. AtXTH31 can encode genes of root extension region expression and participates in root elongation and growth [14]. In addition, some XTH genes are active in the development of leaf [15], fruit [16] and flower [17]. Many studies have also indicated that the expression of XTHs was regulated by plant hormones. For example, the expression of rice OsXTH8 [18] and Arabidopsis AtXTH21 [13] was up-regulated by gibberellin treatment. The expression of banana MA-XET1 was induced by ethylene and participates in the ripening and softening of peel and pulp [19]. Under the treatment of brassinosteroid (BR), the expression of AtXTH22 and AtXTH24 genes increased significantly in Arabidopsis, thereby promoting the elongation of cell walls [20].
In particular, XTH genes play an important role in abiotic stress in plants. The content of xyloglucan was reduced in AtXTH31 mutant of Arabidopsis, which leads to a decrease in the content of absorbed Al 3+ , thereby improving the aluminum stress tolerance [14]. Similarly, Arabidopsis xth17 and xth15 mutants had higher aluminum tolerance compared with wild-type plants [21,22]. CaXTH3 gene in pepper can be induced by drought, high salt and cold, and overexpression of the gene leads to severe fold of Arabidopsis leaves, however, which improves the drought and salt tolerance of transgenic plants [23]. Under low temperature, the expression of persimmon DkXTH6 gene was inhibited, while DkXTH7 showed a high level of transcription [24]. Overexpression of DkXTH1 enhanced the tolerance of transgenic Arabidopsis to salt, ABA and drought stress, and affected the growth of roots and leaves [25].
The first XTH gene was discovered in cowpea [26], then the XTH family genes were established in other species. Among them, a total of 33, 29, 35, and 25 XTH genes were identified in Arabidopsis [11], rice [27], sorghum [28], and tomato [16], respectively. In poplar, Geisler-Lee [29] found 41 XTH genes, and Ye [30] found 42 genes. However, they both did not systematically analyze their expression pattern under salt stress in poplar. In the present research, we found 43 XTHs family genes from poplar. And we determined the evolutionary relationship of poplar XTH family by the analysis of gene structure and homologous collinearity. Meanwhile, the gene expression pattern of XTH family genes was profiled among different tissues under salt stress. This study provides a theoretical foundation for the research on salt-tolerance improvement of poplar plants.

Phylogenetic tree analysis of XTH family
In this study, 43 XTH members were found in poplar genome and named as PtrXTH1 -PtrXTH43 according to their position in the chromosome. The phylogenetic tree of XTH family genes from poplar and Arabidopsis was constructed by MEGA6. As shown in the Fig. 1, the XTH family could be divided into 4 groups (GroupI/II, Group IIIA, Group IIIB and Early diverging group), which was similar to that in Brassica rapa [9]. And the GroupI/IIcontained the most members with 33 members. Group IIIA had 6 members, Group 3B had 3 members, and group 4 had only one member (Fig. 2a).
The physicochemical properties of XTH proteins were also analyzed (Supplementary Table 1). PthXTH2 encoded the largest amino acids (351) with the Fig. 1 Phylogenetic analysis of the XTH proteins from poplar and Arabidopsis. The phylogenetic tree was constructed by MEGA6 software, with neighbor-joining method. The evolutionary tree was divided to 4 groups, each color represents one group molecular weight (MW) of 40.77 kDa, while PthXTH11 had the smallest amino acids (117) with the MW of 13.44 kDa. The theoretical isoelectric point (PI) ranged from 4.53 to 9.67, while the instability coefficient ranged from 27.44 to 55.83. In addition, only one protein had grand average of hydropathicity (GRAVY) greater than 0, which was regarded as hydrophobic protein, and the others were less than 0, which were expressed as hydrophilic protein.
For subcellular localization, there were 8 genes predicted to be located in the extracellular, and the other 35 proteins in the plasma membrane, and 80% of XTH proteins had signal peptides. Then we randomly selected PtrXTH2, PtrXTH14 and PtrXTH37 genes to verify the predicted results (Fig. 3). The signal of positive control (35S-GFP) was detected in the Nicotiana benthamiana leaves nucleus and plasma membrane, while 35S-XTH2-GFP, 35S-XTH14-GFP and 35S-XTH37-GFP were detected in the plasma membrane. The results verified the accuracy of the subcellular localization prediction of XTH proteins.

Gene structure analysis of poplar XTH family genes
In general, the genes that are clustered in one group have a similar structure. In this study, the genes in GroupI/IIhave 2 to 3 introns except PtrXTH11 (1 intron) and PtrXTH1 (4 introns), and the genes in Group IIIA have 3 introns except PtrXTH5 (1 intron). In addition, there were 3 introns in the genes in Group IIIB and Early diverging group, respectively (Fig. 2b). We used PFAM data to analyze the conserved domain of the 43 XTHs. Only three proteins including PtrXTH11, PtrXTH43 and PtrXTH5 had one conserved domain (Glyco_hydro_16), while the other 40 genes had both Glyco_hydro_16 and XET_C domain (Fig. 2c).
In addition, we found 10 related motifs through MEME online software ( Fig. 2d and Supplementary Table 2). The longest one contained 50 amino acids (motif 3 and motif 4), and the shortest contained 7 amino acids (motif 7) (Fig. 2d). Glyco_hydro_16 contained the motifs 1, 4, 6, 7 and 8, but not all protein had the 5 motifs. For instance, XET_C contained the motif 5 and motif 10. The results were similar to those in the previous study [9].

Cis-acting elements in PtrXTHs
In this study, the 2000 bp upsteam sequences of the 43 PtrXTHs were analysed by PlantCARE. As shown in Fig. 4 and Supplementary Table 3, we found 16 types of cis-elements in the upstream promoter sequences of PtrXTHs. Among them, there were many elements related to abiotic stresses such as drought-inducibility, defense and stress responsiveness, low-temperature responsiveness, and wound-response. The elements including meristem expression, analytical induction, circadian control, seed-specific regulation, cell cycle regulation and light responsive were related to plant growth and development. The elements such as auxin responsiveness, MeJA-responsiveness, gibberellinresponsive, abscisic acid responsiveness and zeatin metabolism regulation were related to hormones response. In addition, there were multiple elements in response to  Table 3). For example, Ibox, AE-box, and G-Box were involved in light responsive. P-box, TATC-box and GARE-motif were involved in gibberellin-responsive. The results showed that many PtrXTHs were involved in plant growth and stress response.

Chromosomal distribution and synteny analysis of PtrXTHs
We mapped the poplar XTH family genes to poplar chromosomes through genome annotation. As shown in Fig. 5a, the 43 PtrXTHs were not centrally allocated to a certain chromosomes, while they were randomly allocated to as many chromosomes as possible in a scattered manner. For example, there was only one XTH gene allocated on chromosomes 7, 8, 10, 16, 19, and two on the chromosomes 1, 3, 4, 5, 9 and 13, and six on the chromosome 6, which had the most members. Interestingly, there was no PtrXTHs on the chromosomes 12, 15, and 17.
We analyzed the repetitive events of the PtrXTHs through MCscan. As shown in Fig. 5a and Supplementary Table 4, there were 6 pairs of tandem repetitive events among 43 PtrXTHs, of which PtrXTH18 shared two pairs of repetitive events. And 6 pairs of fragment duplication events were also identified in PtrXTHs (Fig. 5b), which were distributed on the different chromosomes.
In order to further study the evolutionary relationship of XTH family, we constructed systematic maps of the XTH genes between poplar and six other species, including four dicotyledonous (E. robusta, S. lycopersicum, G. max and Arabidopsis) and two monocotyledons (Z. mays and O. sativa). As shown in Fig. 6 and Supplementary Table 5, there were 8 homologous pairs in Arabidopsis, 14 pairs in G. max, 12 pairs in E. robusta, 11 pairs in S. lycopersicum, 1 pair in Z. mays, and 0 pairs in O. sativa. Among them, a total of 16 PtrXTHs showed high collinearity relationship with other 6 species (7 in Arabidopsis, 8 in G. max, 9 in E. robusta, 10 in S. lycopersicum, and 1 in Z. mays). In addition, the PtrXTHs shared homologous pairs in the different species (Fig.  S1). For example, PtrXTH10 was homologous with the gene from E. robusta, S. lycopersicum and Z. mays, and PtrXTH7, PtrXTH15 and PtrXTH37 were shared in G. max, Arabidopsis and S. lycopersicum.

Expression patterns of PtrXTH genes in different tissues
In order to explore the expression patterns of P. simonii × P. nigra XTH genes, the expression level of PtrXTHs in the different tissues was profiled by RNA-Seq. As shown in Fig. 7 and Supplementary Table 6, the XTH genes were divided into 4 groups according to their expression patterns. Among them, the genes in the group 2, group 3 and group 4 were mainly expressed in the leaves, roots and stems, respectively. However, the expression of the genes in the group 1 was mixed. Similarly, some duplicate gene pairs had similar expression patterns. For example, PtrXTH1 and PtrXTH9 were expressed in the leaves, and PtrXTH17 and PtrXTH18 were expressed in the roots. However, some tandem gene pairs showed different expression trends. For instance, the PtrXTH21 was expressed in the leaves, while the PtrXTH24 gene was mainly expressed in the stems. The PtrXTH26 gene was mainly expressed in the roots, while the PtrXTH27 gene was mainly expressed in the leaves.

PtrXTHs expression analysis under salt stress
We also analyzed gene expression patterns of PtrXTHs in different tissues under salt stress by RNA-Seq. As shown in Fig. 8 and Supplementary Table 7, there were 7 differential expression genes (DEGs) in the leaves (4 up-and 3 down-regulated), 9 in the stems (4 up-and 5 down-regulated), and 11 in the roots (5 up-and 6 downregulated). In addition, we analyzed the shared genes among different tissues. As shown in Fig. 8d, there were 3 shared genes (0 up-and 3 down-regulated) in the roots and stems,3 shared genes (2 up-and 1 downregulated) in the stems and leaves, and 2 shared genes (0 up-and 2 down-regulated) in the roots and leaves. Furthermore, only one shared genes (0 up-and 1 downregulated) were shared in the three tissues.
To verify the accuracy of RNA-Seq data, the DEGs in the roots, stems and leaves were selected for RT-qPCR. As showed in Fig. 9, the relative expression levels of DEGs verified by RT-qPCR were in consistent with the mRNA abundance profiled by RNA-Seq analysis generally.

Discussion
XTH family genes play an indispensable role in plants growth and development. In this study, we found 43 XTH genes from P. trichocarpa, which were named as PtrXTH1 to PtrXTH43 according to their location on the chromosomes. The 43 PtrXTHs were divided into 4 subgroups compared with Arabidopsis, which was similar with rice and soybean. Most XTH genes contain two main conserved domains (Glyco_hydro_16 and XET_C domain). However, we found that PtrXTH5 and PtrXTH11 were lack of the XET_C domain. We speculate that there was a loss of XET_C domain during the Fig. 4 Cis-elements analysis of the poplar XTH gene promoters. The left side represents the phylogenetic tree of the poplar XTH genes, and the rights with different colored ellipses represent different elements evolution of the poplar XTH genes. In addition, we predicted that most PtrXTH proteins are located in the plasma membrane, while a small are located in the extracellular, which similar with the tobacco [31]. In our study, the homeopathic transformation experiment of tobacco found that the PtrXTH2, PtrXTH14 and PtrXTH37 proteins are located in the plasma membrane, and possibly in the cell wall. In addition, we found that Arabidopsis AtXTH30 (a homologous protein of PtrXTH2) protein and soybeans GmXTH23 (a homologous protein of PtrXTH14 and PtrXTH37) protein are located in the plasma membrane [32,33], from which we speculate that they are localized in the plasma membrane.
Plants have experienced many whole-genome replication events in the process of evolution, including tandem repeat, fragment repeat and conversion events [34,35]. In this study, we analyzed the collinearity events of XTH family genes in poplar. Interestingly, we found that the number of tandem repeat gene pairs was same to that of fragment repeat gene pairs. We speculate that the two evolutionary mechanisms were co-regulated in the process of evolution. In addition, we found that 7 out of 12 repetitive gene pairs had different expression patterns. For example, the fragment repetitive gene pair PtrXTH10 / PtrXTH25 had same structure and conserved motif, however, they displayed obviously different expression patterns. PtrXTH10 was mainly expressed in the roots, while PtrXTH25 was mainly expressed in the leaves. Therefore, we speculate that PtrXTH genes participate in the replication process, which causes gene mutations, leading to the change of gene function and expression patterns. This phenomenon is also occurred in the ERF family [36] and NAC family [37]. In addition, the collinear relationship between PtrXTHs and other six species were analyzed. We found that PtrXTHs had XTH family genes are involved in the development of plant tissues and organs, and they have expression specificity of tissue and cell. In tobacco, NtXTH4, NtXTH5, NtXTH12, and NtXTH19 had high expression levels in 19 tissues [31], and a similar situation occurred in barley [38]. In our study, we found that the PtrXTHs in groupI such as PtrXTH16 and PtrXTH31 were not expressed in the leaves, roots or stems, which may function in other organizations. The other three groups of PtrXTHs were expressed in the different tissues. However, there were more PtrXTHs expressed in the roots and stems than that in the leaves, which indirectly proves that PtrXTHs may play an important role in the development of root and stem. The previous studies have proved it. For example, Arabidopsis XTH19 and XTH23 (a homologous gene of PtrXTH19) were regulated by BES1 to participate in the development of lateral roots through brassinosteroid signaling pathway, contributing to salt adaptation of lateral roots [39]. The deficiency of XTH9 (a homologous gene of PtrXTH40) in Arabidopsis regulated the secondary wall thickening by triggering integrity signal of cell wall and stimulating the production of xylem cells [40]. During the process of growth and development, plants are subject to multiple abiotic stresses such as high temperature, salinity and drought [41]. XTH genes play an important role in abiotic stress. In our study, we analyzed the expression pattern of poplar XTH family genes under salt stress by RNA-Seq. We found 11 differential expression genes in the roots, 7 in the leaves and 9 in the stems. In addition, RT-qPCR was used to verify the accuracy of RNA-Seq analysis, which indicated the results were consistent in general. Phylogenetic analysis also showed that those genes play an important role in salt stress. For example, Arabidopsis gene XTH30, a homologous gene of PtrXTH2 and PtrXTH8, was upregulated under salt stress, and the lack of the gene slowed down the decrease of crystalline cellulose content and microtubule depolymerization under salt stress, which had a negative impact on salt tolerance [32]. Overexpression of PeXTH from Populus euphratica (a homologous gene of PtrXTH28) could increase water holding capacity and reduce salt concentration in fleshy tissues and mesophyll cells, which improved salt tolerance of transgenic tobacco [42]. All evidences indicate that PtrXTH genes play an important role in regulating stress response in plants.

Conclusion
In this study, a total of 43 XTH genes were identified from poplar, which were divided into 4 groups. These genes were randomly distributed on the 16 chromosomes and 2 scaffolds in poplar. And as many as 12 pairs of duplication events among poplar XTH family genes were found, including 6 pairs of fragment duplication and 6 pairs of tandem duplication. In addition, we analyzed the homologous evolutionary relationship of XTH genes between poplar and other six species, which indicates the PtrXTHs had stronger homologous relationship with dicotyledonous plants, compared to monocotyledonous plants. Furthermore, we profiled the expression pattern of PtrXTHs in different tissues under salt stress through RNA-Seq and RT-qPCR. There were more PtrXTHs expressed in the roots and stems than that in the leaves, which indicates PtrXTHs may play an important role in the development of root and stem. All the results in this study provide a theoretical basis for function identification of poplar XTH genes.

Identification of poplar XTH genes
All amino acid sequences of poplar XTH proteins were obtained from the genome of P. trichocarpa v3.0 in Phy-tozome12 (https://phytozome.jgi.doe.gov/pz/portal. html). The hidden markov models of two typical XTH family protein structures (PF00722 and PF06955) were downloaded from Pfam database (http://pfam.xfam.org/ ). The HMMER3.0 [43] and SMART database (http:// smart.embl-heidelberg.de/) were used to search and filter all potential XTH proteins in poplar. The ExPASy website (http://web.expasy.org/protparam/) was used to calculate physical and chemical parameters of the XTH proteins. SignalP v4.1 server (http://www.cbs.dtu.dk/ services/SignalP/) was used to predict the signal peptide cleavage site, and ProtComp 9.0 (http://linux1.softberry. Fig. 7 Expression pattern of the poplar XTH genes in different tissues. Red represents high expression, green represents low expression. The left represents four gene clusters, R, L, and S represents roots, leaves and stems without salt stress, respectively com) was used to predict the subcellular localization of the XTH proteins.

Subcellular localization of PtrXTH2, PtrXTH14 and PtrXTH37 proteins
The full-length coding sequences of the PtrXTH2, PtrXTH14 and PtrXTH37 genes without stop codon were fused with pBI-121-GFP vector driven by 35S promoter, respectively. The fusion vector was transferred into GV3101 agrobacterium by liquid nitrogen freezethaw method. One-month-old N. benthamiana seedlings were used for transient transformation. The agrobacterium containing the fusion vector (35S-XTH2-GFP, 35S-XTH14-GFP, 35S-XTH37-GFP and 35S-GFP) was injected into the tobacco leaves with 1 ml syringes. Incubate in the dark for 36 h and observe tobacco cells under LMS800 laser confocal microscope.

Gene structures and conserved motif analyses of PtrXTHs
The poplar genomic sequences were downloaded from the Phytozome12 database (https://phytozome. jgi.doe.gov/pz/portal.html). We used GSDS (http:// gsds.cbi.pku.edu.cn/) online software to analyze the gene structure of the XTH family genes. MEME (http://meme-suite.org/tools/meme) online software was used to analyze the conserved motifs with default parameters, and the motifs were visualized with TBtools software [46].

Cis-acting element analysis of PtrXTHs
The upstream 2000 bp sequences of the XTH genes were obtained from the Phytozome12 database The ciselements in the all sequences were predicted through PlantCRAE (http://bioinformatics.psb.ugent.be/webtools/ plantcare/html/), and the cis-elements were screened and visualized by TBtools software [46].

Chromosomal locations, synteny and duplications analyses of PtrXTHs
The genomic data of the poplar was downloaded from phytozome12, and each gene was mapped to the corresponding chromosome position through the position information of the poplar XTH genes. We used the MCscan [47] to calculate the repetitive events of the PtrXTHs family genes among different species. Dual Synteny Plotter was drew to calculate the repetitive events of PtrXTHs between poplar and the other six species (E. robusta, S. lycopersicum, G. max, Arabidopsis, Z. mays and O. sativa) by TBtools software [46].

Plant material and stress treatments
The experimental material used in this study was dihaploid P. simonii × P. nigra and N. benthamiana, which were cultured in the experimental forest of Northeast Forestry University, Harbin, China. The poplar seedlings were grown in 1/2 MS medium, and onemonth-old seedlings were stressed with 0 and 150 mM NaCl for 0 and 12 h, respectively. The roots, stems and leaves at different time points were collected for RNA extraction. Meanwhile, the samples were sent to GENE-WIZ Company for RNA-Seq with Illiumina platform [48]. The expression patterns of the PtrXTH genes in different tissues, and the DESeq2 [49] was used to screen the differentially expressed genes (DEGs) with absolute value of log 2 (fold change) ≥1 and the adjusted p-value ≤0.05. The heatmap was drawn by TBtools [46]. In addition, we used RT-qPCR to verify the accuracy of DEGs screened by RNA-Seq. And the detailed information of RT-qPCR were referred to our previous studies [50]. All the samples were prepared with three biological replicates. All the primers were list in Supplementary  Table 8.