Genome-Wide Identification of Wheat WRKY Gene Family Reveals That TaWRKY75-A Is Referred to Drought and Salt Resistances

It is well known that WRKY transcription factors play essential roles in plants’ response to diverse stress responses, especially to drought and salt stresses. However, a full comprehensive analysis of this family in wheat is still missing. Here we used in silico analysis and identified 124 WRKY genes, including 294 homeologous copies from a high-quality reference genome of wheat (Triticum aestivum). We also found that the TaWRKY gene family did not undergo gene duplication rather than gene loss during the evolutionary process. The TaWRKY family members displayed different expression profiles under several abiotic stresses, indicating their unique functions in the mediation of particular responses. Furthermore, TaWRKY75-A was highly induced after polyethylene glycol and salt treatments. The ectopic expression of TaWRKY75-A in Arabidopsis enhanced drought and salt tolerance. A comparative transcriptome analysis demonstrated that TaWRKY75-A integrated jasmonic acid biosynthetic pathway and other potential metabolic pathways to increase drought and salt resistances in transgenic Arabidopsis. Our study provides valuable insights into the WRKY family in wheat and will generate a useful genetic resource for improving wheat breeding.


INTRODUCTION
It is well known that wheat is one of the most important and widely cultivated cereals, supporting 60% of the world's population for their daily calorific and protein needs (Tadesse et al., 2017). The warranty of wheat production and safety thus has been the first-line mission for world development. However, the productivities of wheat in some regions are still very low due to various stress conditions such as drought, high salinity, heat, and cold. Drought and salinity are the major abiotic constraints for the yield of not only wheat but also other crops and will continue to act as major challenges to researchers and traditional crop breeders in the future (Bartels and Sunkar, 2005;Nezhadahmadi et al., 2013;Oyiga et al., 2016). Plant response to environmental stress is regulated by complex signaling pathways and networks which are coordinated by transcriptional factors (TFs) (Jiang et al., 2015). WRKY TFs comprise a large gene family and are involved in various abiotic stress responses, including drought and high salinity stresses (Phukan et al., 2016). WRKY TFs encode proteins harboring a WRKYGQK amino acid sequence that can bind to W-box motif in the promoter for the regulation of target gene expression (Eulgem et al., 2000;Ciolkowski et al., 2008). The WRKY gene family is a large group that is composed of multiple members. The Arabidopsis and rice (Oryza sativa) genomes contain 75 WRKY genes (Riechmann and Ratcliffe, 2000) and 102 putative WRKY gene members (Wu et al., 2005), respectively. In Arabidopsis, the WRKY gene family is classified into three groups: group I (two WRKY domains), group II (one WRKY domain and C2H2 type of zinc finger motif), and group III (one WRKY domain and C2HC type of zinc finger motif) (Eulgem et al., 2000;Bakshi and Oelmüller, 2014). Group II can be further categorized into IIa, IIb, IIc, IId, and IIe subgroups based on the additional short conserved structural motifs (Eulgem et al., 2000;Chen et al., 2018).
In wheat, some WRKY genes have been reported to be involved in drought and salt stresses (Mondini et al., 2011(Mondini et al., , 2012. TaWRKY1 and TaWRKY33 are induced by drought stress, and the introduction of these genes in Arabidopsis confer the plants drought and/or heat resistance (He et al., 2016). Meanwhile, another wheat WRKY1 gene (GenBank accession no. EU665424) mediates drought tolerance via an ABA-dependent pathway (Ding et al., 2016). Overexpression of wheat WRKY2 enhances drought stress tolerance in transgenic wheat (Gao et al., 2018), while overexpression of wheat WRKY44 and WRKY46 enhances both drought and salt stress tolerances Li et al., 2020). Besides this, two stress-responsive wheat WRKY genes, WRKY2 and WRKY19, are confirmed to play roles in drought, salt, and freezing stresses (Niu et al., 2012). Although these studies provided information on the function of some of the individual WRKY gene members in wheat, the systematic and comprehensive information of the wheat WRKY gene family is still lacking, partially due to the complexity of the wheat genome (Appels et al., 2018).
In order to uncover the detail and to facilitate the future research on WRKY TF family in wheat, we performed a genomewide identification of WRKY gene family members based on a conserved motif search from the recently released common wheat genome (Appels et al., 2018). In the current study, we present an overview of the gene number, classifications, and stress-induced expression patterns of wheat WRKY TF family members. We also demonstrate characteristics and molecular functions of TaWRKY75-A associated with drought and salt tolerance by analyzing the phenotype and RNA sequencing data sets of transgenic Arabidopsis expressing TaWRKY75-A. The results will serve as foundation to molecular study and genetic engineering for wheat against drought and salt stresses.

Identification and Phylogenetic Analyses of WRKY Genes in Wheat Reference Genome
The annotated coding sequences and protein sequences were downloaded from the International Wheat Genome Sequencing Consortium (IWGSC) website 1 . The seed file of the WRKY domain (PF03106) in Pfam 32.0 database 2 (El-Gebali et al., 2019) was used as bait (perform hmmbuild program to establish HMM profile) to search against the local reference genome database by performing hmmsearch program via HMMER software (ver. 3.2.1) 3 (Wheeler and Eddy, 2013). The threshold value was set as lower than 0.05. The gained proteins were then submitted to NCBI conserved domain search tool 4 (Marchler-Bauer et al., 2015) and Swiss-Model online tool 5 to further check the WRKY domain. All confirmed wheat WRKY protein sequences were used to carry out multiple sequence alignment using ClustalX 2.1 software (Larkin et al., 2007). The phylogenetic tree was established using MEGA 6.0 based on neighbor-joining (NJ) method with 1,000 bootstrap replicates (Tamura et al., 2011). Gene Structure Display Server 6 was used to determine and visualize the structures of wheat WRKY genes (Hu et al., 2015). The Gene Structure Display Server Program 7 was used to analyze the exon-intron structure of wheat WRKY genes. The MEME program was employed to identify conserved motifs in wheat WRKY proteins (Bailey et al., 2006).

Naming of Wheat WRKY Genes
In order to name all wheat WRKY genes sequentially, we utilized a consistent naming pattern based on their phylogenetic relationships and chromosome locations in A, B, and D subgenomes (Bai et al., 2019). First, each name of the wheat WRKY genes was started with the abbreviation of the species name Triticum aestivum (Ta). Second, based on the phylogenetic relationships, genes that located in A, B, and D subgenomes and clustered into the same branch were thought as the homeologous triplets of one gene. The serial number of each TaWRKY gene was arranged as the order of wheat chromosome complement, i.e., from chromosome complement 1-7. Third, the serial number of TaWRKY gene in one chromosome complement was decided as the physical location and number of TaWRKY gene copies. TaWRKY genes, which located on the chromosome that contained the most copies, were numbered preferentially. If the number of copies was equal, the name priority was assigned to chromosome A. Finally, the names of TaWRKY genes included an A, B, or D, denoting their subgenome loci .

Chromosome Locations and Classification of TaWRKY Family Members
Chromosome location information of TaWRKY genes was collected based on the genome annotation information. All TaWRKY genes were mapped to their respective locus in the wheat genome and visualized using shinyCircos (Yu et al., 2018). The classification of TaWRKY genes was performed through analyzing the phylogenetic relationship between TaWRKY genes and representative types (i.e., types I, IIa, IIb, IIc, IId, IIe, and III) of Arabidopsis WRKY genes. ClustalX 2.1 software and MEGA 6.0 were used to generate the NJ phylogenetic tree (1,000 bootstrap replicates).

Expression Analysis of TaWRKY Family in silico
The expression data under different stresses, including drought, heat, cold, and osmosis, were downloaded from the Wheat Expression Browser (WEB) 8 (Borrill et al., 2016;Ramírez-González et al., 2018). The transcripts per million values were log 2 -transformed to create a heat map by using WEB.

Plant Materials and Growth Conditions
The wheat cultivar Chinese spring (CS) and Arabidopsis thaliana (Col-0) were used as plant materials in this study. The CS seedlings were planted hydroponically in a greenhouse with continuous light cycle at 25 • C (Niu et al., 2012). Arabidopsis seeds were sterilized and sown on half Murashige and Skoog (MS) medium. These Arabidopsis seeds were then transferred to a growth chamber (continuous light cycle at 22 • C with 50% humidity) after 3 days of freezing (4 • C) treatment.
Total RNA Extraction, Reverse Transcription PCR, and Quantitative Real-Time PCR Total RNA was extracted using the RNA extraction kit (QIAGEN, Germany) following the manufacturer's instructions. The genome DNA in RNA samples were digested using the RNAse-Free DNase Set (QIAGEN, Germany). The quality of the RNA samples was checked using NanoDrop 2000c (Thermo Fisher Scientific, United States) and gel electrophoresis. Reverse transcription PCR (RT-PCR) was performed using the PrimeScript TM II 1st Strand cDNA Synthesis Kit (TaKaRa, Dalian, China). Quantitative real-time PCR (qRT-PCR) was carried out using SYBR Premix Ex Taq TM II (TaKaRa, Dalian, China) on an ECO real-time PCR system (Illumina, United States). The reaction program was described previously . The Arabidopsis ACTIN2 gene (AT3G18780) and wheat 18S-rRNA gene were used as endogenous control. All qRT-PCR primers were designed using Primer Premier 5.0 software based on the subsequent coding sequence sets downloaded from IWGSC archive, v. 1.0, or TAIR 9 . The primers are listed in Supplementary Table 1. The relative expression levels of genes detected in this study were calculated using the comparative threshold cycle method 2 − CT (Livak and Schmittgen, 2001). Each experiment was replicated three times.

Cloning of TaWRKY75-A and Subcellular Localization
The coding sequence (CDS) of TaWRKY75-A was amplified using the special primers listed in Supplementary Table 1. The CDS of TaWRKY75-A was inserted into pENTR/D-TOPO vector (pENTR TM Directional TOPO R Cloning Kits, Thermo Fisher Scientific, United States) and confirmed by sequencing. The confirmed TaWRKY75-A CDS (no termination codon) was then inserted into the 16318GFP (pro35S:GFP) vector. Wheat mesophyll protoplast cell isolation and transformation were described previously . Green fluorescent protein (GFP) signal was detected by laser confocal fluorescence microscopy (ZEISS LSM 880, Germany).

Transcriptional Activation and Yeast One-Hybrid Assay
In order to test the transcriptional activity of TaWRKY75-A, the full-length coding sequence (222 amino acids) and interceptive sequences including 1-44 amino acids (N-terminal), 1-102 amino acids, 44-222 amino acids, and 102-222 amino acids (C-terminal) were cloned into the pGBKT7 vector, respectively. The empty pGBKT7 vector was used as the negative control. The generated constructs were transformed into the AH109 yeast strain. The transformants were diluted into different concentrations and selected on SD/-Trp and SD/-Trp/-His/-Ade/X-α-gal deficiency media. The transcriptional activities were evaluated according to their growth status at 30 • C for 3 days in darkness.
To analyze TaWRKY75-A binding to W-box motif, yeast one-hybrid assay was performed. The promoter sequence of Arabidopsis RD29A gene (AT5G52310), which contains triple tandem repeats of the TTGAC cis-acting element, was cloned into the yeast expression vector pHis2.1 as the reporter construct (pHis2.1-W-box). The full-length coding sequence of TaWRKY75-A was cloned into the pGADT7 vector and fused with the GLA4 domain to generate the effector pGADT7-TaWRKY75-A. Both recombinant vectors were transformed into yeast strain Y187 and selected on SD/-Trp/-Leu plates at 30 • C for 2 days. The surviving colonies were then transferred to SD/-Trp/-Leu/-His medium containing 35 mM 3-AT. The interaction between TaWRKY75-A and W-box element was evaluated by the transformants' growth performance. All the primers used here are listed in Supplementary Table 1.

Generation of Transgenic Arabidopsis Plants Overexpressing TaWRKY75-A
The coding sequence of TaWRKY75-A (no stop codon) was inserted into the pENTR/D-TOPO vector. Then, the gateway reaction was applied to pB7WG2 containing CaMV 35S promoter fused with GFP to generate TaWRKY75-A-pB7WG2 (Karimi et al., 2002). Agrobacterium-mediated transformation method was performed to generate the transgenic Arabidopsis lines (Zhang et al., 2006). Positive seedlings were selected by spraying Basta (1:1,000). Homozygous seedlings were established from T3 generation and used for the subsequent analyses.

Stress Treatments and Expression Profiles of Selected TaWRKY Genes
For polyethylene glycol (PEG) and salt stress treatments of wheat, 2-week-old CS seedlings were watered with 5% polyethylene glycol 6000 (PEG6000) and 100 mM NaCl. The leaf tissues were then collected at 0, 1, 2, 6, and 12 h after treatments. The samples were frozen in liquid nitrogen immediately. The qRT-PCR primers were designed based on the conserved region of each TaWRKY gene  and are listed in Supplementary Table 1.

Phenotype Observations and Physiological Index Measurements in Transgenic Arabidopsis
Germination rates and primary root growth assays of wild type (WT) and transgenic lines were performed as described previously . For drought tolerance assay, 4week-old plants of WT and transgenic lines grown in the soil mixture containing vermiculite and Metro-Mix were cultured without watering for 2 weeks and then were re-watered. The survival rates were calculated at 4 days after re-watering . Relative water content in the leaves was measured using the previous method (Schonfeld et al., 1988). The chlorophyll content of leaf tissues was measured, according to the method described previously (Veronese et al., 2006), from the seedlings after 2 weeks of drought stress. For the salt tolerance assay, 4week-old plants of WT and transgenic lines grown in soil mixture were watered with 100 mM NaCl continuously for 2 weeks. The survival rates were then calculated. The fresh weights of WT and transgenic lines under salt stress were measured from 50 1week-old plants of WT and transgenic lines. All the experiments mentioned above were replicated five times.

RNA-Seq Sample Collection and Sequencing
Three-week-old transgenic Arabidopsis plants (OE1 line) were cultured on plates containing half MS and then transferred in liquid medium containing half MS and 5% PEG6000 or 100 mM NaCl. The plants soaked in half MS solution were used as control. The samples of the control and treated groups were collected at 0, 2, 6, and 12 h after treatments. Total RNA extraction was carried out using the method mentioned above. Transcriptome library creation was created as described previously (Hu et al., 2016) and sequenced on Illumina Novaseq 6000 platform. All RNA-seq libraries had three biological replications, and 150-bp pair-ended reads were generated and used for further analyses.

RNA-Seq Data Analysis and Gene Validations
The adapter sequences and low-quality sequences were removed using the NGS QC Toolkit (v 2.3.3) (Chen et al., 2015), and clean reads were mapped to the Arabidopsis reference genome sequences using TopHat (Kim et al., 2013). Differential expression analysis of two conditions/groups was performed using the DESeq R package (1.10.1). Genes with an adjusted P-value < 0.05 found by DESeq were assigned as differentially expressed. Genes with a fold change ≥1 and false discovery rate <0.05 were considered as differentially expressed genes (DEGs). The fragments per kilobase of transcript per million fragments mapped reads method (E et al., 2014) was used to calculate the expression abundances of genes. Gene Ontology (GO) enrichment analysis of DEGs was implemented by the GOseq R packages based on Wallenius non-central hyper-geometric distribution (Young et al., 2010). A Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis was performed (P ≤ 0.05) using BlastX searches against the KEGG pathway databas 10 .

Statistical Analysis and Data Availability
One-way ANOVA followed by Tukey honest significant difference test was performed to detect differences as required. The RNA-seq data sets have been submitted to the National Genomics Data Center 11 Genome Sequence Archive database under accession number CRA003799.

Identification of WRKY Family Members in Wheat
In order to examine the WRKY gene family in wheat, we searched the wheat genome using hmmsearch and WRKY domain seed file as query. In total, we obtained 353 protein sequences that possessed conservative WRKY domains (Supplementary Table 2). Among these candidate WRKY proteins, 47 proteins had two or more spliced isoforms (Supplementary Table 2), and for these the first isoform of each candidate WRKY protein was selected in the following analyses. After removing the spliced isoforms, a total of 294 WRKY coding sequences were confirmed (Supplementary Table 3). It is well known that common wheat is allohexaploid and contains A, B, and D subgenomes. Therefore, each wheat gene principally have three homeologous gene copies (Yu et al., 2020). Then, we generated an unrooted phylogenetic tree to confirm the homeologous gene copies of each WRKY gene. As shown in Figure 1, 294 gene copies, respectively, belonged to 124 WRKY genes (Figure 1). Because the WRKY gene nomenclature in wheat is currently not consistent, we renamed all genes based on their subgenome association (see "Materials and Methods" section for details). Then, we analyzed homeologous types of TaWRKY genes. It appears that over a half of these 294 TaWRKY genes have three homeologous copies (60.48%), and 16.14% of TaWRKY genes lost one homeolog ( Table 1). Besides this, we found that 23.38% of the TaWRKY genes have only one gene copy, indicating that over one-fifth of the members are orphans/singletons ( Table 1).

Gene Structure, Conserved Domain, and Chromosomal Location
To gain further insights into the TaWRKY gene members, we surveyed the gene structure, conserved domain components, and chromosomal location of each WRKY gene copy. The TaWRKY genes showed a wide range in sequence lengths, with TaWRKY76-D having the longest (20.8 kb) and TaWRKY2-D having the shortest (0.5 kb) genomic sequence (Supplementary Figure 1). Besides this, the intron-exon structures of TaWRKY genes indicated normal distributions of introns and exons (Supplementary Figure 1). Most of the TaWRKY gene copies harbored two (58 gene copies) or three (139 gene copies) exons. Twenty-six gene copies had no intron, and 71 gene copies contained 4-7 exons (Supplementary Figure 1). Meanwhile, the conserved protein motifs in the TaWRKY family were predicted using the MEME tool. Ten types of motifs, which were named motif 1 to motif 10, were identified in each TaWRKY gene copy. The number of conserved motifs in each TaWRKY protein varied from 3 to 8; all members had motifs 1, 2, and 3, and most members had motifs 4, 5, and 6 (Supplementary Figure 2), indicating that the TaWRKY proteins are relatively conserved. Furthermore, we investigated the chromosomal locations of each of the 294 TaWRKY gene copies using the recently released IWGSC wheat genome. Chromosome 1 contained the highest number of TaWRKY genes (24 genes), and chromosome 6 had the lowest number of TaWRKY genes (five genes) (Figure 2 and  Supplementary Figure 3). Meanwhile, we found that subgenome D had the highest number of TaWRKY gene copies (109; 37.07%), while subgenome B had the lowest number (87; 29.59%) (Figure 2).

Classification of the TaWRKY Gene Family
In Arabidopsis, the WRKY gene family is classified into three groups based on the number of WRKY domains and the types of zinc-finger-like motif (Eulgem et al., 2000). Among these groups, group II is further divided into subgroups from IIa to IIe (Eulgem et al., 2000;Bakshi and Oelmüller, 2014;Chen et al., 2018). To classify the TaWRKY gene members, we selected the typical WRKY genes of each group from Arabidopsis and performed the phylogenetic analysis with all TaWRKY genes (Supplementary Figure 4). Chromosomes 1 and 5 had six groups, and chromosomes 2, 3, 4, 6, and 7 contained five groups of TaWRKY members (Figure 3A). Among the 294 TaWRKY gene copies, 94 belonged to group III, which possessed a CCHC type of zinc finger motif ( Figure 3B). A total of 52 gene copies were classified into group I, which had two WRKY domains (Figure 3B). The rest of the gene copies belonged to group II, which was separated into IIa, IIb, IIc, IId, and IIe subgroups and contained one WRKY domain and C2H2 type of zinc figure motif ( Figure 3B).

The Expression Pattern of the TaWRKY Genes Under Abiotic Stresses
In order to characterize the expression profiles of the TaWRKY gene family, we analyzed the RNA-seq data downloaded from expVIP (Borrill et al., 2016;Ramírez-González et al., 2018). We specially tracked the expression patterns of the TaWRKY family members after 2 weeks of cold , 6 h of drought and heat (Liu et al., 2015), and 12 h of PEG treatments (Supplementary Table 4). Notably, there were 68 TaWRKY gene copies that could not be detected or that can be detected under only one of the cold, drought, heat, and PEG stresses (Supplementary Figure 5). The rest of the TaWRKY gene copies could respond to at least two types of stresses (Supplementary Figure 5). Under the drought and PEG treatments, TaWRKY10,34,35,41,43,55,66,74,75,78,81,83,86,88,89,93, and 112 displayed relatively high expression levels, and their gene copies showed similar expression trends under drought and PEG treatments (Supplementary Figure 5). Due to TaWRKY10, 34, and 43 having been cloned and studied (Bahrini et al., 2011;Wang et al., 2013;Kumar et al., 2014;He et al., 2016), we excluded these three genes out of our later analyses.
To investigate the expression profiles of the rest of the 14 TaWRKY genes, we monitored their expression trends under PEG (5% PEG6000) and high salinity (100 mM NaCl) stresses using qRT-PCR assay. Under PEG stress, TaWRKY41 and 55 were down-regulated compared with mock (0 h) at each time point, and TaWRKY78 remained almost unchanged ( Figure 4A). The rest of the 11 TaWRKY genes responded to PEG stimulation after 1 h of treatment and displayed an increasing expression trends afterward ( Figure 4A). Among these, TaWRKY75 showed the highest expression level after 1 h of PEG treatment compared with the other members ( Figure 4A). Under a high salinity stress condition, the expression levels of TaWRKY35,41,55,86,88,93,and 112 were inhibited at each time point (Figure 4B). No expression changes occurred in TaWRKY78 as it did after PEG treatment. The expression of TaWRKY81 increased after 1 h but was reduced subsequently compared to mock (0 h) ( Figure 4B). Inversely, TaWRKY66, 74, 75, 83, and 89 were induced at each time point compared with mock (0 h). Similar as in drought stress, TaWRKY75 exhibited the highest fold change in the expression level at 1 h after treatment ( Figure 4B). Notably, the physiological function of TaWRKY75 has not been reported to date. Therefore, we hypothesized that TaWRKY75 may serve as an important stress-related gene and conducted functional analyses as the following.

Characterization of Transcription Activity of TaWRKY75-A
TaWRKY75 harbored one WRKY domain and a CCHC zinc finger motif, and the encoding gene had three copies on chromosome 4 ( Figure 3A). To assess the molecular function of TaWRKY75, we cloned the nucleic acid sequence of one of its copies, designated TaWRKY75-A. TaWRKY75-A contained 669 base pairs and encoded the protein which had 222 amino acids. To observe subcellular localization, TaWRKY-A was fused with GFP under 35S promoter and transiently expressed in wheat protoplast cells. The TaWRKY75-A-GFP was detected in the nucleus (Figure 5A), indicating that TaWRKY75-A was a nucleus-localized protein. To evaluate the trans activity of TaWRKY75-A, the complete coding sequence and special fragments shown in Figures 5B,C were fused with the GAL4 DNA binding-domain coding sequence of pGBKT7 vector, and this recombinant construct was transformed into the yeast strain AH109. It showed that the N-terminal region without WRKY domain (1-44 aa) was minimally and sufficiently required for   its trans activity ( Figure 5C). It is well known that WRKY TFs bind the W-box sequence (C/T)TGAC(T/C) (Ulker and Somssich, 2004;Chen et al., 2018). Indeed the yeast-onehybrid assay showed that TaWRKY75-A binds W-box with a conserved sequence ( Figure 5D). Taken together, the results mentioned above showed that TaWRKY75-A was a typical WRKY transcription factor.

Ectopic Expression of TaWRKY75-A Improved Drought and Salt Resistances in Transgenic Arabidopsis
The expression results indicated that TaWRKY75-A responded timely to drought and salinity stresses and pertained its high expression along continued stress input (Figures 1-4). To understand the function of TaWRKY75-A under drought and salinity stresses, we generated eight transgenic Arabidopsis lines by introducing TaWRKY75-A under constitutive 35S promoter and selected three lines (OE1, OE2, and OE3) for characterizing a drought-and salt-induced phenotype (Supplementary Figure 6). No significant differences were observed under normal condition between the WT and transgenic Arabidopsis lines in seed germination rate and length of the primary roots (Supplementary Figure 7). PEG-induced stress greatly reduced WT germination within 7 days, while this suppressive effect was largely abrogated in the transgenic plants. The seed germination rates in WT and transgenic lines were eventually kept at 89 and 100%, respectively (Figures 6A,B). The primary root growth of WT was also inhibited in the presence of PEG and partially recovered in transgenic lines (Figures 6C,D and Supplementary Figure 7). Under a longterm drought stress condition (2 weeks without watering), the transgenic plants retained larger and greener leaves than those of WT. After re-watering treatment, the transgenic plants further exhibited better performance in survival rate, relative water content, and chlorophyll content than WT (Figures 6E-H). On the medium containing a high level of salt (100 mM NaCl), the seed germination rate of WT was reduced to 73%, whereas the transgenic plants germinated at a higher level up to 83% (Figures 7A,B). Salt also reduced primary root growth and fresh weights in WT but to a lesser extent in transgenic lines (Figures 7C-F  and Supplementary Figure 7). Under a soil condition with long-term salt stress, the transgenic lines also showed a remarkably higher survival rate than the WT (Figures 7G,H).
Our data altogether showed that the ectopic expression of TaWRKY75-A increased the drought and salt tolerances in Arabidopsis.

Comparative Analysis of Transcriptome Data
To better understand the mechanism by which the ectopic expression of TaWRKY75-A increases drought and salt tolerances, we conducted the comparative transcriptome analysis in TaWRKY75-A transgenic plants treated with PEG or NaCl. A total of 3,655, 196, and 393 genes were differentially expressed at 2, 6, and 12 h after PEG treatment. Upon NaCl treatment, a total of 2,839, 1,189, and 96 genes were differentially expressed at 2, 6, and 12 h, respectively ( Figure 8A). There were 19 and 27 genes that displayed expression changes throughout the PEG and salt-treated periods, respectively ( Figure 8A). The GO and KEGG analyses of these DEGs indicated that they shared the same GO terms such as "membrane" and "membrane part" and KEGG pathways including "lipid metabolism" and "biosynthesis of other secondary metabolites" (Supplementary  Figure 8), indicating that the ectopic expression of TaWRKY75-A might activate the same pathways via regulating the same genes to increase PEG and salt tolerance.
Furthermore, DEGs formed in PEG and salt treatments shared seven genes in common ( Figure 8A). The GO function analysis results showed that six common DEGs participated in "metabolic process, " and five and four took part in "membrane" and "membrane part" terms, respectively (Supplementary Figure 9 and Supplementary Table 5). Meanwhile, the KEGG pathway analysis results revealed that three and four common DEGs were enriched in "α-linolenic acid metabolism" and "biosynthesis of secondary metabolites' (q-value < 0.05), respectively ( Figure 8B). Interestingly, these DEGs showed a similar expression pattern under PEG and NaCl treatments (Figure 8C), which was further confirmed by qRT-PCR assay ( Figure 8D).

DISCUSSION
WRKY TFs, as a highly conserved protein family, are found in all plants and other eukaryotic lineages, such as fungi, Amoebozoa, diplomonads, and slime molds (Zhang and Wang, 2005;Liu et al., 2014;Rinerson et al., 2015). Progress in gene identification and functional evaluation of WRKY family members uncovered their important roles in plant adaptation to various environmental cues . In comparison to other crops, knowledge on WRKY gene family has remained relatively scarce in the wheat field due to the relatively backward release of complete and high-quality genome. Hence, a genome-wide systemic understanding of WRKY TFs would facilitate the evolution and function of particular WRKY genes in wheat development and stress response.

Features of TaWRKY Gene Family During Evolution
In crop plants, the WRKY gene family has been well investigated. Until now, there are 103 and 119 WRKY genes that have been isolated in the rice (O. sativa) and maize (Zea mays) genome (Ramamoorthy et al., 2008;Wei et al., 2012), respectively. Our study revealed that the wheat genome possesses 124 WRKY genes. This number is similar as that of the rice and maize. Because wheat is hexaploid and has three subgenomes, the number of TaWRKY homeologous copies should be threefold of the WRKY genes in rice and maize in theory. We identified 294 WRKY homeologous gene copies in wheat, which is 2.85 and 2.47-folds compared to that in rice and maize, respectively. These indicate that the TaWRKY gene family is evolutionarily conserved in wheat in the context of gene number. The current and widely accepted classification system shows that the structures of WRKY genes classified in seven types are well FIGURE 6 | Overexpression of TaWRKY75-A enhances drought tolerance in transgenic Arabidopsis. (A) Observation of seed germination in wild type (WT) and OE lines at 4 days after germination. (B) Seed germination rate of WT and OE lines within a week. One-way ANOVA followed by Tukey honest significant difference (HSD) test was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was replicated five times. (C) Observation of the primary root elongation of WT and OE lines at 7 days after germination. Scale bar: 1 cm. (D) Statistics of lengths of the primary root of WT and OE lines at 7 days after germination. One-way ANOVA, followed by Tukey HSD test, was performed. * * p < 0.01, n = 20. Bars show the maximum and minimum values. (E) Observations of WT and OE seedlings during drought test. (F) Statistics of the survival rate of WT and OE lines at 4 days after re-watering. One-way ANOVA, followed by Tukey HSD test, was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was performed five times. (G) Quantification of the relative water content of the leaves of WT and OE lines after 2 weeks of drought treatment. One-way ANOVA, followed by Tukey HSD test, was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was performed five times. (H) Statistics of the chlorophyll content of the leaves of WT and OE lines after 2 weeks of drought treatment. One-way ANOVA, followed by Tukey HSD test, was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was performed five times.
defined in wheat and several other plants (Figure 3; Chen et al., 2018). In wheat, group III had the highest number of WRKY genes (Figure 3), which was similar to that in rice and maize (Ramamoorthy et al., 2008;Wei et al., 2012). This result suggested that the classification of TaWRKY is also conserved. Moreover, 10 different motifs are present in different TaWRKY proteins, with each protein comprising conserved motifs ranging from 3 to 8 motifs. In addition, all proteins have motifs 1, 2, and 3, demonstrating that the TaWRKY proteins have a conserved motif distribution profile.
As an allohexaploid plant, the percentage of wheat genes in homeologous groups for all configurations is highly similar across the three subgenomes: 63% (subgenome A), 61% (subgenome B), and 66% (subgenome D) (Appels et al., 2018). Among 294 TaWRKY gene copies, we found 109 copies (37.08%) on subgenome D, 87 copies (29.59%) on subgenome B, and 98 copies FIGURE 7 | Overexpression of TaWRKY75-A increases salt tolerance in transgenic Arabidopsis. (A) Observations of seed germination in wild type (WT) and OE lines at 1 week after germination. (B) Seed germination rate of WT and OE lines within a week. One-way ANOVA followed by Tukey honest significant difference (HSD) test was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was replicated five times. (C) Observation of primary root elongation of WT and OE lines during 7 days after germination. Scale bar: 1 cm. (D) Quantification of primary root lengths of WT and OE lines at 7 days after germination. One-way ANOVA followed by Tukey HSD test was performed. * * p < 0.01, n = 20. Bars indicate means ± SD. (E) Observations of the growing status of WT and OE seedlings under normal and salt conditions at 7 days after germination. (F) Quantification of fresh weight of 50 seedlings of WT and OE lines under normal and salt conditions at 7 days after germination. One-way ANOVA followed by Tukey HSD test was performed. * * p < 0.01. This experiment was replicated five times. (G) Observations of WT and OE seedlings during salt stress. The salt stress was applied for 2 weeks. (H) Quantification of the survival rate of WT and OE lines after 2 weeks of salt treatment. One-way ANOVA followed by Tukey HSD test was performed. * * p < 0.01. Bars indicate means ± SD. This experiment was performed five times.
(33.33%) on subgenome A. These percentages were similar with the configurations of homeologous groups. Among 124 TaWRKY genes, 75 genes (60.48%) contained triads with a single gene copy per subgenome (an A:B:D configuration). Besides that, in common wheat genome, the three subgenomes exhibit similar levels of loss of individual homeologous gene, showing 10.7% (B:D), 10.3% (A:D), and 9.5% (A:B) of the homeologous types in the A, B, and D subgenomes, respectively (Appels et al., 2018). However, we found only one homeologous type that lost the homeolog on D subgenome (0.81%) in the TaWRKY family. Furthermore, we found that 10 (8.06%), four (3.23%), and 15 (12.09%) TaWRKY genes, which located on A, B, and D subgenome, respectively, belonged to orphan genes. These results demonstrated that the TaWRKY gene family did not undergo duplicate events other than gene loss during the long evolutionary process.

TaWRKY Family Members Undertake Multiple Functions
Functional characterization of the individual WRKY gene in wheat has been carried out, primarily through interspecies gene expression. TaWRKY2 (corresponding to TaWRKY1 in our study) and TaWRKY19 (corresponding to TaWRKY35 in our study) are induced by salt, drought, and cold stresses. In TaWRKY10 (corresponding to TaWRKY47 in our study) transgenic tobacco, drought and salt stress tolerance is enhanced . Besides this, TaWRKY45 (corresponding to TaWRKY43 in our study) is upregulated in response to Fusarium graminearum, and the overexpression of TaWRKY45 confers increased resistance against F. graminearum in transgenic wheat (Bahrini et al., 2011). It has been reported that TaWRKY51 (corresponding to TaWRKY38 in our study) negatively regulates ethylene production and then promotes lateral root formation . These evidence show us that different TaWRKY family members have different functions and also uncover the potential roles of TaWRKY genes in regulating abiotic and biotic resistances and developmental process.
Given that expression patterns can lead to estimation of gene functions (Xiao et al., 2018), we monitored the expression profiles of TaWRKY family members under PEG, cold, heat, and drought conditions. A total of 68 TaWRKY genes (54.84% of 124 TaWRKY genes) did not respond to these stresses, indicating that these genes do not respond to the relevant abiotic stress responses. A considerable number of TaWRKY genes (14 genes) are shown to respond to PEG and drought stresses from our analysis, highlighting that these genes are promising regulators for drought and salt responses in wheat. While more studies would be required for evaluating the physiological function of these individual genes, our study provides that core candidate genes can be targeted for improvement of wheat tolerance against drought and salt in the future.

TaWRKY75-A Is the Positive Regulator of Drought and Salt Tolerance via Jasmonic Acid Biosynthesis and Membrane Metabolism
Lipoxygenase (LOX) and allene oxide cyclase (AOC) are two key enzymes that play decisive effects in jasmonic acid (JA) biosynthetic process (Farmaki et al., 2007). JA is one of the important phytohormones that regulate a plant's response to abiotic stresses , and several JArelated genes including JAZ, AOS1, AOC, LOX, and COI have been reported to be upregulated upon various stresses (Robson et al., 2010;Hu et al., 2017). Drought condition also upregulates several JA biosynthesis and metabolism-related genes in wheat , and the exogenous application of JA was shown to enhance the activities of peroxidase, which, in turn, increases salt tolerance (Qiu et al., 2014). In addition, TaAOC1, which encodes an allene oxide cyclase involved in the a-linolenic acid metabolism pathway, has been demonstrated to enhance salt tolerance in wheat . Our study reveals that TaWRKY75-A, when ectopically expressed in Arabidopsis, enhances drought and salt tolerance, most likely by upregulating JA biosynthetic genes AtLOX3 and AtAOC1, and this mechanism may be conserved in Arabidopsis and wheat.
In addition to JA-related genes, we found that a lecithin/cholesterol acyltransferase family protein, PDAT2, a calcium-dependent phosphotriesterase superfamily protein, SSL7, and an undefined gene AT4G36010, which are all clustered into "membrane" and "membrane part" GO terms, are upregulated under PEG and salt stresses in TaWRKY75-A transgenic Arabidopsis. These results indicate that TaWRKY75-A might enhance drought and salt resistance by altering cell membrane homeostasis via these genes. Besides that, we found that a nudix family protein NUDT8 and a farnesoic acid carboxyl methyltransferase (FAMT) are induced after PEG and salt treatments. In the JA biosynthesis defective mutant aos, the expression of NUDT8 is decreased significantly (Yan et al., 2007), suggesting that JA is required for NUDT8 expression in Arabidopsis. The expression of FAMT is promoted when exogenous methyl jasmonate is applied (Yang et al., 2006), revealing that this gene is a downstream component of the JA signaling pathway. This is in line with the previous finding that lipid-derived hormone JA plays key functions on the direct regulation of plant metabolism (Savchenko et al., 2019). Thus, JA might regulate metabolic processes in our transgenic plants. Hence, based on these data, we propose that TaWRKY75-A is a key stress-resistant gene in wheat for adaptation to drought and salt stress by modulating the JA biosynthetic pathway as well as other metabolic pathways ( Figure 8E).
In summary, our study provides comprehensive insights into the WRKY gene family in wheat. Classification of all TaWRKY TFs from whole genome survey will inevitably contribute to further molecular research for the identification and understanding of key TaWRKY members against various biotic and abiotic stresses in wheat.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are publicly available. This data can be found here: National Genomics Data Center, accession number: CRA003799 (https://bigd.big.ac.cn/ bioproject/browse/PRJCA004312).

AUTHOR CONTRIBUTIONS
YW and JB designed the research. HY, LQ, HG, LG, and FR performed the experiments and analyzed the data. YW and HY wrote the manuscript. All authors reviewed and approved the final manuscript.  Statistics of seed germination rate of WT and OE lines at 7 days after germination. One-way ANOVA, followed by Tukey honest significant difference (HSD) test, was performed. There are no significant differences between WT and each OE line. Bars show means ± SD. This experiment is replicated five times. (C) Observations of primary root elongations of WT and OE lines at 7 days after germination. Scale bars: 1 cm. (D) Statistics of primary root elongations of WT and OE lines at 7 days after germination. One-way ANOVA, followed by Tukey HSD test, was performed. There are no significant differences between WT and each OE line. Bars show means ± SD, n = 20. This experiment is replicated five times.