The wheat WRKY transcription factors TaWRKY49 and TaWRKY62 confer differential high-temperature seedling-plant resistance to Puccinia striiformis f. sp. tritici

WRKY transcription factors (TFs) play crucial roles in plant resistance responses to pathogens. Wheat stripe rust, caused by the fungal pathogen Puccinia striiformis f. sp. tritici (Pst), is a destructive disease of wheat (Triticum aestivum) worldwide. In this study, the two WRKY genes TaWRKY49 and TaWRKY62 were originally identified in association with high-temperature seedling-plant resistance to Pst (HTSP) resistance in wheat cultivar Xiaoyan 6 by RNA-seq. Interestingly, the expression levels of TaWRKY49 and TaWRKY62 were down- and up-regulated, respectively, during HTSP resistance in response to Pst. Silencing of TaWRKY49 enhanced whereas silencing TaWRKY62 reduced HTSP resistance. The enhanced resistance observed on leaves following the silencing of TaWRKY49 was coupled with increased expression of salicylic acid (SA)- and jasmonic acid (JA)-responsive genes TaPR1.1 and TaAOS, as well as reactive oxygen species (ROS)-associated genes TaCAT and TaPOD; whereas the ethylene (ET)-responsive gene TaPIE1 was suppressed. The decreased resistance observed on leaves following TaWRKY62 silencing was associated with increased expression of TaPR1.1 and TaPOD, and suppression of TaAOS and TaPIE1. Furthermore, SA, ET, MeJA (methyl jasmonate), hydrogen peroxide (H2O2) and abscisic acid (ABA) treatments increased TaWRKY62 expression. On the other hand, MeJA did not affect the expression of TaWRKY49, and H2O2 reduced TaWRKY49 expression. In conclusion, TaWRKY49 negatively regulates while TaWRKY62 positively regulates wheat HTSP resistance to Pst by differential regulation of SA-, JA-, ET and ROS-mediated signaling.

Introduction Plants initiate pathogen-associated molecular pattern (PAMP)-triggered immunity (PTI) responses to recognize molecular signatures of many pathogens via pattern recognition receptors (PRRs). Pathogens deliver virulence effectors to suppress plant PTI while the plant detects these effectors by resistance (R) proteins, activating effector-triggered immunity (ETI) responses [1]. R gene-mediated ETI involves complex defense processes, including production of reactive oxygen species (ROS) and salicylic acid (SA), rapid programmed cell death (hypersensitive responses, HR) and induction of host genes including pathogenesis-related (PR) genes [1]. These complex defense responses involve timely recognition of the invading pathogen, followed by defense responses through complicated signaling pathways [2]. These signaling pathways regulate the defense responses by fine-turning transcriptional activation of defense-related genes [3]. Increasing evidence has revealed that transcriptional regulation of gene expression in response to pathogen attacks is a crucial part of the plant defense system [3,4]. Defense-related gene expression is regulated by transcriptional factors (TFs) that alter gene expression by binding to target DNA-binding sites of genes, in cooperation with other proteins. WRKY proteins are zinc-finger-containing TFs that belong to a large family of related proteins in the plant kingdom [5]. There are 109, 72 and more than 160 WRKY family members identified in rice [6], Arabidopsis [7] and wheat [8,9], respectively. WRKY TFs are structurally classified into three main groups (I, II and III), and also multiple subgroups (e.g. IIa, IIb and IIc, etc.) that is dependent on the presence of one or two WRKY domains in a 60-amino acid region at the N-terminus, having the conserved heptapeptide sequence WRKYGQK, and a zinc-finger like motif at the C-terminus [5].
WRKY TFs have received increasing attention for their roles in regulating plant defense responses [6], including cell death [10]. The oxidative burst and production of reactive oxygen species (ROS), are among the earliest defense reactions that are activated in response to pathogen attack [11,12], which may lead to rapid programmed cell death preventing pathogen progression [13]. When Nicotiana tabacum NtWRKY1 is co-expressed with the salicylic acidinduced protein kinase (SIPK) there was significantly more host cell death than that observed with expression of SIPK alone in response to Ralstonia solanacearum [10]. Salicylic acid (SA), ethylene (ET) and jasmonic acid (JA) are the main signaling molecules involved in plant defense responses. In Arabidopsis, AtWRKY70 is linked to SA-mediated signaling in response to Erysiphe cichoracearum, a biotrophic fungus, and JA-mediated signaling against Alternaria brassicicola, a necrotrophic fungus [14]. Recently, we demonstrated that TaWRKY70 positively regulates high-temperature seedling-plant (HTSP) resistance to Puccinia striiformis f. sp. tritici (Pst) in wheat, probably through the SA-and ET-mediated signaling pathways [15]. WRKY TFs may also act as negative regulators of plant defense responses. Overexpressing OsWRKY62 in rice compromises the basal defense and Xa21 (receptor-like kinases)-mediated resistance against Xanthomonas oryzae pv. oryzae (Xoo), and suppresses the activation of defense-related genes [16].
Wheat stripe rust (or yellow rust), caused by Pst, is one of the most destructive wheat (Triticum aestivum) diseases worldwide [17,18]. Temperature change-induced wheat resistance to Pst is usually non-race-specific and durable, of which two types of resistance were reported: high-temperature adult-plant (HTAP) and high-temperature seedling-plant (HTSP) resistance. Wheat plants with only HTAP resistance are susceptible to Pst in the seedling stage and under low temperature, but become resistant as plants grow old and under high temperature [19,20]. In HTSP resistance, wheat seedlings are susceptible to Pst at low temperatures but become resistant when exposed to high temperatures for 24 h at the initial Pst symptom-expression stage (eight days after inoculation) [21][22][23][24]. There is little information whether WRKY TFs play a role in the HTSP resistance against Pst [15].
Recently, we obtained a cDNA library with RNA from wheat cultivar Xiaoyan 6 (possessing HTSP resistance) infected by the Pst CYR32 pre-exposure to a high temperature (HT) treatment [15˚C for first eight days post-inoculation, then 20˚C for 24 h, and then back to 15˚C] and sequenced the library. Preliminary studies identified 24 differentially regulated candidate WRKY TFs during the HTSP process, including TaWRKY70 [15]. In this study, we characterized two WRKY TFs in relation to their roles in the HTSP resistance to Pst: a WRKY62 homolog (the second most highly up-regulated TF, after TaWRKY70) named TaWRKY62, and a WRKY49 homolog (the only down-regulated TF), named TaWRKY49. We hypothesize that TaWRKY62 and TaWRKY49 play positive and negative roles in regulating the HTSP resistance, respectively. Functional analyses of these two WRKY genes by gene silencing experiments were performed, and further data were collected on the regulation of the genes when exposed to phytohormones and abiotic stressors.

Materials and methods
Plant materials, growth conditions, high-temperature resistance induction and stress treatments Wheat cultivar Xiaoyan 6 and Pst race CYR32 were used to study the wheat-Pst interaction. The seeds of Xiaoyan 6 were provided by Dr. Sanhong Fan in Northwest A&F University, and the Pst CYR32 urediniospores were provided by Prof. Qiuzhen Jia in Institute of Plant Protection, Gansu Academy of Agricultural Science. Wheat seeds (10)(11)(12)(13)(14)(15) were grown in a plastic pot (10×10×10 cm 3 ) filled with a potting mixture under rust-free conditions. The first leaves of seedlings at the two-leaf stage (approximately 10-14 days after planting) were uniformly brushed with a mixture of Pst urediniospores and sterile water at a ratio of approximately 1:10-15 (v/v). The seedlings were then placed in a dew chamber in the dark for 24 h (temperature, 10˚C; rh, 90-100%) and subsequently transferred to a growth chamber (Percival E-30B, Perry, IA, USA) and grown under 16 h of light at 15±1˚C (rh, 60-80%; supplemented with sodium lighting (505 μmol/m 2 /s photon flux density)) and 8 h of dark at 12±1˚C (rh, 60-80%). In parallel, control plants were brushed with sterile water. In the initial symptom-expression stage of rust development (8 dpi), the plants were divided into two groups for exposure to different temperature regimes. The first group was subjected to low-temperature (LT) treatment; i.e., Pst-inoculated wheat plants were incubated at a constant temperature (15±1˚C). The second group was subjected to high temperature (HT) treatment, i.e., Pst-inoculated wheat plants were incubated at 15±1˚C, at 8 days post inoculation transferred to a growth chamber set at 20 ±1˚C and incubated for 24 h and then moved back to and maintained thereafter at 15±1˚C, as the exposure to 20±1˚C for 24 h was previously shown to activate HTSP resistance to Pst [23]. The leaf tissues from LT and HT treatments were sampled at 0, 48,96,192,194,198,204,216,240,264 and 312 hours post inoculation (hpi); the 192-hpi time corresponded to the beginning of HT treatment. Leaves, stems and root tissues were sampled also from the two-leaf-stage seedlings. Three biological replicates were used for each assay.
In addition, experiments were conducted to study wheat responses to extreme temperatures and to hormone treatments. For extreme-temperature treatments, two-leaf stage seedlings were incubated under cold (4˚C) or hot (40˚C) temperatures. For hormone treatments, seedlings were sprayed at 15˚C with hydrogen peroxide (H 2 O 2 , 100 μM), methyl jasmonate (MeJA, 100 μM), ethylene (ET, 100 μM), salicylic acid (SA, 100 μM) or abscisic acid (ABA, 100 μM) [25]. In all of these treatments, seedlings in a similar state of growth (two-leaf stage) were used, and non-treated wheat seedlings were used as controls. All of the treated and non-treated seedlings were harvested at 0, 0.5, 2, 6, 12 and 24 h, frozen in liquid nitrogen and stored at -80˚C. Three biological replicates were performed independently for each time point.

Plasmid construction
The γRNA-based vector derived from BSMV was constructed as described previously [26]. cDNA fragments derived from the coding sequence (251 bp, nt 9-259) of TaWRKY49 complete cDNA were used to construct the recombinant TaWRKY49-as plasmids, and cDNA fragments (149 bp, 593-741) derived from TaWRKY62 were used to construct the recombinant TaWRKY62-as plasmids. All primers used for vector construction are listed in S1 Table. BSMV-mediated TaWRKY49 and TaWRKY62 silencing Plasmids used for barley stripe mosaic virus (BSMV)-mediated gene silencing were constructed according to Holzberg et al. [26]. The cDNA fragments of TaWRKY49 (251 bp) and TaWRKY62 (149 bp) were amplified with primer pairs TaWRKY49VIGS and TaWRKY62-VIGS respectively (S1 Table). The wheat PDS gene (TaPDS) was replaced with the cDNA fragment of TaWRKY49 or TaWRKY62 in BSMV:γ-PDS.
The inoculation and incubation conditions for the virus were as described previously by Scofield et al. [27] and Wang et al. [15] respectively. Rust symptoms and sporulation on the fourth leaves were assessed on 14 days post inoculation (dpi).

Reverse Transcription Quantitative PCR (RT-qPCR)
Total RNA from wheat tissue was extracted using the PureLink 1 Plant RNA Reagent (Invitrogen, Carlsbad, CA, USA), then genomic DNA contamination was removed using DNase I treatment (Thermo Fisher Scientific, Waltham, MA, USA), and 500 ng of poly(A)+ mRNA was reversed transcribed into cDNA using a PrimeScript 1 RT Reagent Kit (TaKaRa, Tokyo, Japan). RT-qPCR was conducted to quantify the expression of TaWRKY49 (primer TaWR-KY49Q, S1 Table) and TaWRKY62 (primer TaWRKY62Q, S1 Table) according to Wang et al. [15]. After a preliminary study (data not shown), wheat 26S gene (ATP-dependent 26S proteasome regulatory subunit) (Unigene No. Ta22845) (primer Ta26SQ, S1 Table) was used as the internal reference for each RT-qPCR assay. The efficiency and specificity of the primer pairs are presented in S1 Fig, and the primers are given in S1 Table. The relative expression of mRNA was calculated using the 2 -ΔΔCt method [28].
Silencing efficiency and expression levels of TaWRKY49 and TaWRKY62 under high-temperature treatment after Pst infection The fourth leaves of the seedlings for which the second leaves had been pre-inoculated with BSMV:00 and BSMV:WRKY49-as or BSMV:WRKY62-as were sampled separately at 0, 24, 48 and 120 hours post inoculation (hpi) with CYR32. RT-qPCR was performed to determine the silencing efficiency of TaWRKY49 and TaWRKY62 in each assay.
To assess whether the expression level of TaWRKY49 and TaWRKY62 on their corresponding silenced plants with Pst infection could be induced by high temperature (HT) treatment, plants from the low temperature (LT) treatment were divided into two groups. One group was immediately subjected to the HT treatment at 192 hpi of Pst, and another group was kept under the LT treatment. Each group included TaWRKY49 or TaWRKY62-silenced and BSMV:00 control plants. Samples were harvested at 0, 12, 24, 48, 72 and 120 h post-temperature treatment (hptt) from both LT and HT treatments to detect the temperature-induced expression of TaWRKY49 and TaWRKY62 via RT-qPCR. Stripe rust symptoms and Pst sporulation on leaves were assessed at 14 dpi. Additionally, the transcript levels of TaPR1.1 (Genbank Accession AJ007348) (primer TaPR1.1Q, S1 Table), TaAOS (AY196004) (primer TaAOSQ, S1 Table), TaPIE1 (EF583940) (primer TaPIE1Q, S1 Table), TaCAT (X94352) (primer TaCATQ, S1 Table) and TaPOD (TC303653) (primer TaPODQ, S1 Table) on BSMV:00 and TaWRKY49 or TaWRKY62 silenced leaves subjected to the LT and HT treatments were quantified and compared at 0, 12, 24, 48, 72 and 120 hptt through RT-qPCR. Three biological replicates were performed for each time point.

Histological observations of TaWRKY49-silenced and TaWRKY62silenced wheat plants under LT and HT treatments
Wheat leaves pre-infected with BSMV and subjected to the LT treatment after Pst inoculation were sampled at 48 and 120 hpi to observe Pst development and the host response on TaWRKY49-silenced and TaWRKY62-silenced plants. At eight days following Pst inoculation, BSMV-pre-infected wheat leaves subjected to the LT and HT treatments were sampled at 0 (192 hpi), 24, 48, 72 and 120 h to examine changes in Pst development and host response, arising from the temperature-reduced expression of TaWRKY49 and temperature-induced expression of TaWRKY62. The staining and fixing of specimens were performed using Calcofluor M2R White New (Sigma, MO, US) staining method as described previously by Wang et al. [29]. After fading and fixation, cleared wheat leaf segments were analysed to determine the hyphal length and number of haustorial mother cells under a microscope. Autofluorescence of pathogen-induced host necrotic cells was observed under a fluorescence microscope (excitation filter, 485 nm; dichromic mirror, 510 nm; barrier filter, 520 nm). Given the lateness of initial Pst development, linear lengths of the fungal colonies and uredinia were chosen for the assessment of fungal development. Colony length was measured from the substomatal vesicle to the apex of the longest hypha, and uredinium length as the length of its long shaft. In addition, plant cell death was defined as the presence of autofluorescence under a fluorescence microscope associated with an infection unit. About 30-50 infection sites on 8-10 wheat leaf segments (length, 1.5 cm) from 8-10 randomly selected wheat plants were examined. All of the microscopic observations were performed using an Olympus BX-51 microscope (Olympus Corporation, Tokyo, Japan), and the data were analysed using DP-BSW software.

Detection of H 2 O 2 and O 2 − (ROS)
The production of H 2 O 2 in the TaWRKY49-silenced and TaWRKY62 , respectively. The percentage of DAB or NBT staining is based on 30 infection sites that were randomly selected from one leaf segment, and the mean value of percentages of DAB or NBT staining come from 3-8 leaf segments.

Statistical analyses
Analysis of variance was conducted using SAS v8.01 (SAS Institute Inc., Cary, NC, USA). Data relating to the number of necrotic cells and haustorial mother cells as well as length of hyphae in silenced and non-silenced wheat leaves under LT treatment were performed according to the Student's t-test at P = 0.05 or P = 0.01, under the assumption of homogeneous variance. Data on hormone treatments as well as cold and heat stresses were analyzed using the Duncan's multiple range test for different time points. Data relating to with or without Pst inoculation under LT or HT treatments, percentage of infection sites with DAB and NBT staining, number of necrotic plant cells, length of fungal colonies and uredinia for each time point were analyzed using the Duncan's multiple range test. The relative expression of TaWRKY49, TaPR1.1, TaAOS, TaPIE1, TaCAT and TaPOD in the four VIGS treatments (HT BSMV:00, LT BSMV:00, HT BSMV:TaWRKY49-as and LT BSMV:TaWRKY49-as) as well as the relative expression of TaWRKY62 and the TaPR1.1, TaAOS, TaPIE1, TaCAT and TaPOD genes in the four VIGS treatments were analyzed using Duncan's multiple range test for each time point.

Sequence analysis of TaWRKY49 and TaWRKY62
Twenty-four WRKY TFs with differential expression during the HT treatment post Pst infection were identified from the preliminary RNA-seq data analysis. Among these TFs, TaWRKY49 was the only one down-regulated (approximately 2.9 times), and TaWRKY62 was the second most highly up-regulated (rank only second to TaWRKY70 [15] (approximately 5.2 times). Two full-length 1221-bp cDNA and 798-bp cDNA were obtained respectively, from Pst-infected leaves. The cDNA of 1221-bp length encodes a 321-amino-acid polypeptide protein with a molecular mass of 34.17 kDa and a theoretical isoelectric point (pI) of 6.13. This protein is predicted to localize in the nucleus and shares 88% amino acid identity with Triticum urartu WRKY49 protein (EMS52311.1) and 45% amino acid identity with the Arabidopsis IIc-type WRKY protein WRKY49 (NP_199143.1). We named this gene TaWRKY49 (GenBank No. LC169122), the cDNA of which was 798-bp in length, and encodes a 258-amino-acid polypeptide with a molecular mass of 28.39 kDa and a theoretical pI of 5.83. The encoded protein, also predicted to localize in the nucleus, shares 95% amino acid identity with Aegilops tauschii WRKY62 protein (EMT18619.1) and 33% amino acid identity with the Arabidopsis III-type WRKY protein WRKY62 (NP_195810.2). Thus, the gene was named TaWRKY62 (GenBank No. LC169123). TaWRKY49 and TaWRKY62 were mapped onto wheat chromosome 3B and the long arm of 5B, respectively. TaWRKY49 was clustered in a large clade within the subgroup IIc of the WRKY family, with highest homology to AtWRKY49, while TaWRKY62 belonged to a large clade of the group III WRKY members (Fig 1). Both TaWRKY49 and TaWRKY62 have conserved WRKY domains that includes one WRKYGQK sequence at the N-terminus. However, there is a C 2 -H 2 (C-X 4 -C-X 23 -H-X 1 -H)-type zinc-finger motif at the C-terminus of TaWRKY49 (S2 Fig), and a C 2 -HC (C-X 7 -C-X 29 -H-X 1 -C)-type zinc-finger motif at the C-terminus of TaWRKY62   mock (HT but without Pst inoculation), and its expression level was lower (P < 0.05) than the low temperature Pst inoculation (LT) treatment at 204 hpi (Fig 2A). Under LT treatment, the expression level of TaWRKY49 was higher than the LT mock (LT but without Pst inoculation) at 48 and 204 hpi, while thereafter, it was lower than the LT mock at 216 and 312 hpi (Fig 2A). The relative expression of TaWRKY62 was elevated in the HT treatment in comparison with the HT mock at 48, 194, 264 and 312hpi, and was higher in the HT treatment than in the LT treatment at 264 hpi. In particular, at 264 hpi, the expression level of TaWRKY62 was highest at HT treatment among the four treatments ( Fig 2B). TaWRKY49  the surface of the second leaf of wheat seedlings at the two-leaf stage as the positive control. The empty vector (BSMV:00) was used as the negative control. At 9 days post-inoculation (dpi), photobleaching symptoms were observed for the BSMV:TaPDS inoculation (Fig 3A and  3D), and the typical striping mosaic symptoms were apparent on new leaves of plants for BSMV:00 inoculation (Fig 3A and 3D). The mock wheat plants (FES buffer inoculated) developed new normal leaves under the same conditions, suggesting that the specific silencing of TaPDS occurred on the BSMV:TaPDS-inoculated leaves (Fig 3A and 3D).

Phenotypes of
BSMV-inoculated plants displayed mild chlorotic mosaic symptoms at 9 dpi. At 14 dpi, the TaWRKY49-silenced (BSMV:TaWRKY49-as) leaves under the LT treatment showed fewer uredinia than the non-silenced (BSMV:00) (Fig 3B), and the typical HR phenotype was observed for the HT-treated TaWRKY49-silenced leaves, showing necrotic stripes with smaller pustules than the HT-treated non-silenced leaves (Fig 3C). Under the HT treatment, the TaWRKY62-silenced leaves exhibited stripes without chlorosis and abundant pustules. The leaves in which TaWRKY62 was not silenced displayed necrotic/chlorotic stripes with limited sporulation (Fig 3F).
Compared with the BSMV:00 vector inoculated leaves, the expression level of TaWRKY49 in silenced plants with BSMV carrying TaWRKY49-as was reduced (P < 0.01) for the LT treatment (S3 Fig) after inoculation with Pst. For the HT treatment, the transcriptional level of TaWRKY49 in TaWRKY49-silenced and non-silenced leaves were reduced, though not significantly (Fig 3G). Similarly, the expression level of TaWRKY62 in TaWRKY62-silenced leaves was reduced under the LT treatment (S3 Fig) after Pst inoculation. The levels were also reduced compared with the non-silenced leaves for HT treatment (P<0.05; Fig 3H), suggesting the silencing is effective under both temperature conditions. Changes in the fungal development and host responses in wheat leaves with TaWRKY49 or TaWRKY62 silenced The colony linear length and number of haustorial mother cells and host necrotic cells of Pst were assessed microscopically (Fig 4). At 120 hpi, hyphal length of Pst in TaWRKY62-silenced leaves was 58.13 μm more than those leaves where TaWRKY62 was not silenced, while the number of Pst haustorial mother cells in TaWRKY49-silenced leaves were fewer than those Functions of wheat TaWRKY49 and TaWRKY62 in HTSP resistance to Pst observed in the non-silenced leaves (Table 1). Necrotic cells were rarely seen on the silenced leaves.
The length of the Pst colonies in infected wheat leaves was observed in response to the two different temperature regimes (Fig 5A1 and 5A2). For the HT treatment, the length of colonies in the BSMV:TaWRKY49-as leaves was similar to that in the BSMV:00 leaves, except at 0 hptt ( Fig  5B). HT treated TaWRKY62-silenced leaves developed lengthier colonies (P < 0.05) than those in TaWRKY62 non-silenced leaves at 12 hptt (Fig 5C). From 24 hptt onwards, the pustules gradually formed, and some uredinia with sparsely scattered urediniospores were seen in the HT-treated leaves ( Fig 5A4); whereas urediospores in LT-treated leaves gathered closely (Fig 5A3). For the HT treatment, the length of uredinia in BSMV:TaWRKY49-as leaves was shorter than that in BSMV:00 leaves from 72 hptt onwards (P<0.05; Fig 5D). In contrast, the BSMV:TaWRKY62-as leaves developed lengthier uredinia than the BSMV:00 leaves from 48 hptt onwards (P < 0.05; Fig  5E). At 120 hptt, the number of dead cells in the HT-treated BSMV:TaWRKY49-as leaves was higher than the BSMV:00 leaves (Fig 5F). Fewer necrotic cells in the HT-treated BSMV:TaWR-KY62-as leaves were observed than the BSMV:00 leaves at 24 hptt onwards ( Fig 5G).

Accumulation of reactive oxygen species (ROS) in TaWRKY49-and TaWRKY62-silenced wheat leaves
Accumulation of H 2 O 2 was induced in the mesophyll cells or the cell walls of the leaves, which was evident by the reddish-brown staining due to 3, 3-diaminobenzidine (DAB) polymerization (Fig 6A). Under the HT treatment, the percentage point of DAB staining in the BSMV: TaWRKY49-as leaves was 20% and 25% lower than that of BSMV:00 leaves at 0 and 12 hptt, but higher at 24 hptt (Fig 6B). DAB staining was apparent at more than 50% of the infection sites at 24 hptt for TaWRKY49-silenced leaves under the HT (Fig 6B), indicating the importance of BSMV:TaWRKY49-as in reactions to Pst development at 24 h. There were no significant differences in the accumulation of H 2 O 2 between TaWRKY62-silenced and non-silenced leaves under the HT condition (S4 Fig). Under the HT treatment, O 2 − production levels were similar between TaWRKY49-silenced and non-silenced leaves (Fig 6C). There were no significant differences between TaWRKY62-silenced and non-silenced leaves in the percentage of infection sites with NBT staining under the HT treatment (S4 Fig). Under the HT treatment, the expression level of TaCAT (catalase) in the infected leaves of BSMV:TaWRKY49-as was higher than in the BSMV:00 leaves at 0, 12 and 48 hptt (Fig 6D). In contrast, the expression of TaCAT did not differ between BSMV:TaWRKY62-as and BSMV:00 Functions of wheat TaWRKY49 and TaWRKY62 in HTSP resistance to Pst (Fig 6E). Under the HT treatment, the TaPOD (peroxidase) gene was induced in BSMV: TaWRKY49-as leaves before 12 hptt when compared with the BSMV:00 leaves. TaPOD was suppressed in the BSMV:TaWRKY49-as leaves, when compared to the BSMV:00 leaves at 24 hptt ( Fig 6F). The expression level of TaPOD in the HT-treated TaWRKY62-silenced leaves was higher than the HT-treated TaWRKY62 non-silenced leaves at 24 and 48 hptt (Fig 6G). Under the HT treatment, the expression level of a SA marker gene (TaPR1.1) in TaWRKY49-silenced leaves was higher than in the non-silenced leaves at 0, 6 and 48 hptt ( Fig  7A). Similarly in the HT treatment, the level of TaPR1.1 expression in TaWRKY62-silenced leaves was higher than in the non-silenced leaves at 0, 6, 12 and 48 hptt (Fig 7B). The expression of pathogen-induced ethylene response factor 1 (TaPIE1), an ET-responsive gene, was reduced in the TaWRKY49-silenced leaves (Fig 7C) and in the TaWRKY62-silenced leaves Functions of wheat TaWRKY49 and TaWRKY62 in HTSP resistance to Pst ( Fig 7D) compared with the non-silenced leaves under the HT treatment. The marker gene of JA signaling, allene oxide synthase (TaAOS), was induced rapidly from the TaWRKY49silenced leaves in HT treatment at 2 hptt when compared to the HT-treated non-silenced ( Fig  7E). The expression level of TaAOS did not differ between the HT-treated TaWRKY62silenced and non-silenced leaves at earlier stage ( Fig 7F); however, at the later stage, the level of TaAOS expression was lower in TaWRKY62-silenced leaves than in the non-silenced leaves at 24 and 120 hptt (Fig 7F).  (Table 2). Under the ET treatment, this gene was rapidly induced within 0.5 h post-treatment (hpt), and its expression level reached the peak at 6 hpt (the peak was about 15-fold higher than the control at 0 hpt) ( Table 2). The induction of TaWRKY49 by the application of SA and ABA only occurred at 0.5 and 2 hpt, respectively (Table 2). In contrast, methyl jasmonate (MeJA) application did not affect the expression of TaWRKY49, and H 2 O 2 treatment reduced the expression of TaWRKY49 (Table 2). TaWRKY62 was up-regulated by all five hormone treatments ( Table 2). Under ABA and ET treatments individually, the peak of the expression level of TaWRKY62 reached was at 2 hpt (Table 2). Under H 2 O 2 , MeJA and SA treatments individually, this gene was fastly induced and its expression level reached at a peak at 0.5 hpt (Table 2).
With regard to the extreme temperature treatments, the transcript levels of TaWRKY49 and TaWRKY62 were up-regulated by cold (4˚C) stress and reached the peak at 6 and 24 hpt respectively ( Table 2). And the transcript levels of both genes were down-regulated in response to heat (40˚C) stress (Table 2). TaWRKY49 was expressed predominantly in stem, but TaWRKY62 was expressed predominantly in roots ( Table 2).

Discussion
WRKY TFs have been implicated in many plant defense processess including responses to biotic and abiotic stresses [5]. Many WRKY genes are responsive to infections, including those initiated by fungi [14,32], bacteria [33,34], and a virus [35], suggesting that WRKY TFs play critical roles in plant defense responses to pathogens. In this present study, in wheat HTSP resistance against Pst, TaWRKY49 and TaWRKY62 were down-and up-regulated, respectively. Inducing of silencing constructs targeting TaWRKY49 led to the enhanced HTSP resistance. In contrast, the silencing of TaWRKY62 led to reduced HTSP resistance. These results suggest a negative regulatory role of TaWRKY49 TF and a positive regulatory role of TaWRKY62 TF in wheat HTSP resistance against Pst.
Many WRKY TFs, such as AtWRKY4 [36], AtWRKY8 [37], AtWRKY27 [38] and AtWRKY48 [39] in Arabidopsis; HvWRKY1 and HvWRKY2 in barley [40]; CaWRKY1 [41] and CaWRKY58 [34] in pepper; and OsWRKY62 in rice [16], function as negative regulators of disease resistance in plant-pathogen interactions. In the present study, the group-IIc WRKY gene, TaWRKY49, negatively regulated wheat HTSP resistance against Pst. In contrast, TaWRKY62 functioned as a positive regulator in the wheat resistance. Thus, the two WRKY TFs affected plant resistance in opposite ways, and this may be due to the distinct roles played by the two WRKY proteins in regulating the crosstalk between defense signaling pathways. This may require that these WRKY TFs act as transcriptional activators or repressors in a gene-specific manner. The expression levels were determined through RT-qPCR analyses and were relative to the levels observed at 0 hpi. Three biological replicates were performed independently for each treatment. Functions of wheat TaWRKY49 and TaWRKY62 in HTSP resistance to Pst Furthermore, the WRKY TFs could have opposite effects against different types of pathogens. For example, loss-of-WRKY70 function in Arabidopsis increased plant susceptibility to Pseudomonas syringae, Erwinia carotovora, E. cichoracearum and Botrytis cinerea, but increased the resistance to A. brassicicola [14,42,43]. Overexpression of OsWRKY62 in rice compromises the resistance against Xanthomonas oryzea pv. oryzea [16]; while in this study, the silencing of TaWRKY62 compromised HTSP resistance to Pst.
High temperature stimuli induced HTSP resistance to Pst, and hence the expression levels of relevant WRKY TFs were responsive in high temperature treatment. The molecular basis of the effects of high temperature on TaWRKY49 and TaWRKY62 in response to Pst is unknown, although Pst resistance genes affected by temperature have been reviewed recently [44]. In leaf rust resistance testing, the Thatcher wheat line with Lr23 was susceptible to all isolates at all temperatures except for one isolate which was avirulent at 30˚C and 10~30˚C, indicating temperature specificity is necessary for wheat-leaf rust genetic interaction for Lr23 in some cases [45]. During adult plant resistance to leaf rust on wheat cv. Thatcher (carries the Lr22b gene, and its nearisogenic lines Lr34 and Lr37), Lr22b is inefficient; while Lr34 slows down the disease development at a mean daily temperature below 16˚C, but is poorly efficient at temperatures above 20˚C; Lr37 provides high resistance under all conditions [46], showing that some resistance genes were correlated with the temperature-dependent resistance. Two quantitative trait loci (QTLs) were detected from a population of 188 F 2:3 families (from wheat cross Fundulea 900/'Thatcher'), and the two QTLs were designated as QLr.hebau-1BL that was Lr46 and QLr.hebau-7DS that was Lr34, respectively [47]. These genes have minor effects, conferring partial, durable resistance to leaf rust [48,49]. Therefore, the character of TaWRKY49/ TaWRKY62 in HTSP resistance to Pst is similar with above genes-temperature sensitive and exerting minor effect. Yr36 functions as a regulatory gene, and its START domain is postulated to bind lipids from Pst at high temperature and change its conformation, which might cause the kinase domain to initiate a signaling cascade leading to programmed cell death [50]. In rice, OsWRKY62 interacts with Xa21-cleaved intracellular domain exclusively in the nucleus, mediating immune responses [51]. Although OsWRKY62 itself is localized in an unknown intracellular structure/organelle, OsWRKY62 and OsWRKY76 can form a hetero-complex in the nucleus, suggesting that these two TFs may function collaboratively [52]. TaWRKY62 at high temperature may either interact with other proteins or directly regulate the transcriptional reprogramming of defense-related genes. Negative regulators of disease resistance, such as AtWRKY17, AtWRKY11 [53], CaWRKY1 [41], HvWRKY1/2 [40] and OsWRKY62 [16], are inferred to prevent the inappropriate activation of defense responses at suboptimal concentrations of signal molecules, or to turn off the activated defense reaction generated by positive regulators once the pathogen infection has been halted, since the defense responses against disease-causing microbes are energy consuming processes [54]. Thus, the negative role of TaWRKY49 in HTSP resistance against Pst may lie in preventing the inappropriate activation or in turning off excessive defense responses.
WRKY TFs are nodes for cross-talk between SA, JA and ET signalling pathways and involved in plant defense through these signalling parhways [14,55,56]. Crosstalk among SA, JA and ET has emerged as an important regulation switch in plant disease resistance [57,58]. Usually, SA signaling is specific to resistance responses against biotrophic pathogens, whereas JA/ET signaling against necrotrophic pathogens [59,60]. Negative regulators of disease resistance, such as AtWRKY17, AtWRKY11 [48], CaWRKY1 [41], HvWRKY1/2 [40] and OsWRKY62 [16], are transcriptionally induced in response to pathogen infection, as well as in response to SA or JA [16,17,40,41,48]. However, the expression of TaWRKY49 was downregulated by Pst infection in the HT treatment, up-regulated by ET and SA and was not affected by MeJA. We speculate that the relatively higher temperatures may change the transcription level of TaWRKY49 that is induced by SA and ET signaling. In Arabidopsis, systemic acquired resistance (SAR) is strongly correlated with the coordinate expression of a set of genes encoding proteins that include the pathogenesis-related (PR) proteins, including PR1-1a [61,62]. In this study, TaPR1.1 was induced in both TaWRKY49-and TaWRKY62silenced wheat leaves under the HT treatment. These results suggest that the increased disease resistance of silenced-TaWRKY49 and increased susceptibility of silenced-TaWRKY62 wheat leaves at high temperature may partially depend on a common SA signaling pathway that is invovled in SAR.
The ROS network is essential for the induction of disease resistance [63,64]. H 2 O 2 is one of the primary ROS species at the site of pathogen infection [65]. In addition to being a direct protective agent, the oxidative burst functions as a threshold trigger for hypersensitive cell death [11,[66][67][68]. ROS-induced HR can act as a defense reaction against pathogens [69,70]. In HT-treated TaWRKY49-silenced leaves after Pst inoculation, H 2 O 2 and O 2 were produced rapidly from 24 hptt and the accumulation level of these chemicals was higher than the HTtreated non-silenced, accompanied by the rapid induction of antioxidant enzyme gene TaPOD. There was a positive correlation between POD and disease resistance in plants: after inoculation of pathogen, the POD activity was rapidly increased in a resistant cultivar but not enhanced or delayed in a susceptible cultivar [71]. The expression of TaPOD was induced early in the TaWRKY49-silenced leaves at the beginning of HT treatment, which might contribute to the induction of disease resistance. In the earlier stage of ROS production, that was induced by pathogen attacks, CAT can decompose H 2 O 2 into O 2 -, acting to trigger benzoic acid to form SA and leading to systemic acquired resistance (SAR) [72]. Thus, it is reasonable to speculate that up-regulation of TaCAT in the HT-treated silenced-TaWRKY49 leaves may play a role in triggering SA signaling. In contrast, in the HT-treated TaWRKY62-silenced leaves, the TaCAT expression level did not vary noticeably, and TaPOD was induced only at the later stage (after 12 hptt), which is consistent with the increased susceptibility at high temperature due to silencing TaWRKY62. These results suggested that ROS signaling is likely involved in the WRKY gene-regulated HTSP resistance to Pst via independently or interacting with other signaling pathways. The overexpression of a canola MYB gene-BnaMYB78 in Nicotiana benthamiana modulates ROS-dependent cell death through regulating the transcription of a few ROS-and defense-related genes [73]. In addition, colony and uredinium formation was also delayed, and the number of necrotic cell, representing the level of HR [74], was increased significantly in HT-treated TaWRKY49-silenced wheat leaves. Thus, TaWRKY49 negatively regulates the HTSP resistance to Pst partially via ROS signaling dependent HR cell death. On the other hand, the accumulation of H2O2 and O 2 in HT-treated TaWRKY62silenced leaves after Pst infection did not vary significantly. The longer uredinium length and fewer necrotic cells were observed in HT-treated TaWRKY62-silenced leaves, suggesting that TaWRKY62 positively participates in the HTSP resistance probably through enhancement of cell death independent of ROS. Although ROS production is an essential component for the onset of local cell death in micro-hypersensitive response and induction of host resistance [75], it may not be the only factor that leads to cell death. Further research is needed to understand the relationship between ROS and cell death.
In response to multiple stresses, the wheat WRKY genes exhibited similar responses. For example, they were induced by multiple phytohormones and cold stress (4˚C), and repressed by heat stress (40˚C). We recently showed that TaWRKY70 positively regulates the HTSP resistance to Pst via enhancing the expression of both TaPR1.1 and TaPIE1 genes [15]. These results together indicate that SA-, ET-and JA-related genes as well as ROS-related genes could be differentially regulated by TaWRKY49, TaWRKY62 and TaWRKY70 during the wheat HTSP resistance against Pst infection (Fig 8). Further understanding on the crosstalk between phytohormone-and ROS-mediated signaling will provide new insights into how these WRKY TFs regulate HTSP resistance against Pst.
Supporting information S1