TaWRKY10 transcription factor is a novel jasmonic acid signalling regulator involved in immunity against Septoria tritici blotch disease in wheat

Septoria tritici blotch (STB), caused by the fungus Zymoseptoria tritici , is currently the main threat to worldwide wheat production in temperate climates. Understanding the genetic mechanisms that underpin the Z . tritici– wheat interaction will be crucial for generating new control strategies against STB. Plant hormones are essential regula tors of growth and immunity. In particular, jasmonic acid (JA) plays a central role in de fence against necrotrophic plant pathogens, but its role in mediating immunity against Z . tritici is largely unknown. Here we identify the transcription factor TaWRKY10 that is taxonomically restricted to the grass subfamily Pooideae as a novel regulator of JA responses in wheat. We demonstrate by using virus- induced gene silencing that silencing TaWRKY10 leads to increased resistance against Z . tritici with an earlier onset of necrotic symptoms, but with reduced pathogen sporulation. Wheat plants silenced for TaWRKY10 show an up- regulated JA response. Transcriptional profiling of TaWRKY10 knock- down plants indicates that it is a negative regulator of the JA receptor TaCOI1 gene expression. Our data indicate that TaWRKY10 down- regulates JA responses, contributing to the susceptibility of wheat to Z . tritici . We postulate that manipulating TaWRKY10 may provide a strategy to boost STB resistance in wheat.

appearance of leaf necrosis indicates the initiation of the pathogen reproductive stage. In this phase, pycnidia (the reproductive organ of the fungus) filled with pycnidiospores appear as black lesions on leaf surfaces. The mature pycnidia release spores that allow Z. tritici to colonize the plant canopy, affecting flag leaf productivity by reducing its photosynthetic potential, which ultimately impacts grain yield (Ponomarenko et al., 2011). Resistance to STB is one of the major targets in wheat breeding programmes. Currently 21 distinct genetic loci (Stb genes) as well as a large number of quantitative trait loci (QTLs) that confer STB resistance have been identified in various wheat germplasm screening programmes (Kettles & Kanyuka, 2016;Saintenac et al., 2018). However, the molecular mechanisms underlying wheat resistance to STB are largely unknown.
Plant hormones are essential regulators of growth and immunity (Balbi & Devoto, 2008). In particular, jasmonic acid (JA) plays a central role during defence against necrotrophic pathogens (Okada et al., 2015). JA-dependent defence has been widely studied in the model system Arabidopsis thaliana challenged with the necrotrophic pathogen Botrytis cinerea (Smirnova et al., 2017)

and in Oryza sativa
against Magnaportha oryzae (reviewed by Nasir et al., 2017). JA and its derivatives are fatty acid-derived hormones ubiquitously present in the plant kingdom (Wasternack, 2007). They act as regulatory molecules in many plant developmental processes that include fertility, sex determination, root elongation, and fruit ripening (Mandaokar et al., 2006). JA is also involved in activating plant defences against herbivory, wounding, and abiotic stress (Balbi & Devoto, 2008).
In recent years JA has been found to have a critical role in mechanisms that allow plants to switch to defence rather than growth, therefore transcriptionally reprogramming plant cells to activate defence mechanisms and arrest cell cycle and growth (Pauwels et al., 2008).
JA-Ile, the bioactive form of JA, is formed through a conjugation of JA and isoleucine (Ile) under the action of a GH3 family of amido synthetases. JA-Ile is subsequently recognized by CORONATINE INSENSITIVE 1 (COI1)-JASMONATE ZIM DOMAIN (JAZ) coreceptor complexes and activates a signalling cascade for the induction of a series of defence genes that are also expressed in response to wounding, insect herbivory, and necrotrophic pathogens (Balbi & Devoto, 2008).
In comparison to Arabidopsis and rice, knowledge on JA biosynthesis and signalling in wheat is limited. The PR genes PR1.1 and PR1.2 were highly induced in wheat 3 weeks after emergence by JA application (Liu et al., 2016), and the same pattern of expression was detected on wheat infected with the fungal pathogens Tilletia tritici and Tilletia laevis (Lu et al., 2006). Similarly, using a transcriptomebased method, it was revealed that JA biosynthesis genes such as LOX, AOS, AOC, and OPR3 and JA signalling transduction genes, including COI1, JAZ, and MYC2, were induced in response to Fusarium head blight pathogen in a resistant wheat variety (Xiao et al., 2013).
Nevertheless, the mechanism underpinning the signal perception and transduction remains unclear.
The WRKY transcription factor gene family is one of the largest and plant-specific transcriptional regulators (Bakshi & Oelmüller, 2014). WRKYs are part of an intricate plant signalling web and regulate multiple pathways from stress responses to growth (Bakshi & Oelmüller, 2014). It has been established through multiple studies that WRKY transcription factors are central regulators of innate plant immunity in plants (Bakshi & Oelmüller, 2014). The characteristic feature of the WRKY transcription factor family is its highly conserved 60 amino acid long WRKY domain at the N-terminus and zinc-finger-like motif at the C-terminus (Bakshi & Oelmüller, 2014).
These motifs enable WRKY transcription factors to bind with high affinity to a specific DNA cis-acting element, (C/T)TGAC(T/C), named the W box, on target gene promoters. The regulatory networks controlled by WRKY genes allow plants to transduce environmental cues into differential expression of hundreds of genes, resulting in major physiological responses (Pandey & Somssich, 2009).
Although the role of JA in defence against necrotrophic plant pathogens is well established, its role in mediating immunity against Z. tritici is largely unknown. To reveal any role of JA in Z. tritici-wheat interaction, we studied the susceptibility of two wheat cultivars, KWS Lili and KWS Santiago, to STB, and investigated the role of the regulator TaWRKY10 in defence against Z. tritici using a gene silencing approach.

| Plant materials and growth conditions
Nicotiana benthamiana plants were used to propagate barley stripe mosaic virus (BSMV) to perform virus-induced gene silencing (VIGS), as previously described (Lee et al., 2015). These N. benthamiana plants were grown in environmentally controlled cabinets at 24 °C (day) and 22 °C (night) in long day conditions (16 hr light/8 hr dark).

| Preparation of pathogen inoculum and plant inoculation
Z. tritici isolate IPO323 was used for all plant assays, as described previously by Keon et al. (2007). Spores were grown on yeast extract peptone dextrose (YPD) agar (Sigma) for 7 days at 18 °C.
For plant inoculation, spores were suspended in water containing 0.1% (vol/vol) Tween 20 at a density of 7.5 × 10 6 spores/ml, then spray-inoculated on wheat leaves. Wheat plants were inoculated with spores 15 days after VIGS (GS 1; see below). Two replicates of five leaves each were used to determine spore production within pycnidia 28 days after inoculation for each experiment run. Spore counts were carried out using a light microscope and a haemocytometer after washing spores, as previously described by Lee et al. (2015). Area under the disease progress curve (AUDPC) was obtained from the percentage of leaf area covered by necrotic lesions and calculated using an AUDPC calculation algorithm in the R software package. Three independent experiments were performed to evaluate the phenotypes.
The time points considered for this experiment spanned from 11 to 17 days postinoculation (dpi).

| Quantitative reverse transcription PCR
Quantitative reverse transcription PCR (RT-qPCR) was employed to analyse total gene expression in response to Z. tritici infection or after hormonal treatment. The total RNA was extracted using Direct-zol RNA prep Kit (Zymo Research) as per the manufacturer's instructions. The RNA was quantified and 1 μg RNA was used for cDNA synthesis. cDNA synthesis was done using SuperScript II reverse transcriptase (Invitrogen) following the manufacturer's guidelines. Quantitative PCR was performed using Brilliant III Ultra-Fast SYBR QPCR master mix (Agilent) by using ABI One Step Plus instrument and the analysis was done with the software provided using comparative quantification methods. TaEF1 was used as internal control for normalization. The experiments were repeated three times. In each experiment, two pools of three leaves from three different plants were used as technical replicates. To ascertain gene expression changes in response to Z. tritici, leaves were collected immediately after inoculation of the pathogen for time point 0, and subsequently every 48 hr.
Gene expression analysis in response to JA treatment in roots was repeated three times. In each experiment, two pools of the whole root system from three different seedlings were used as a technical replicate. The ΔΔC t method (Lee et al., 2015) was used to analyse the raw data. All the primers used in this study are listed in Table S2.

| Virus-induced gene silencing
Reverse genetics studies in T. aestivum were carried out using the VIGS platform (Baulcombe, 1999) based on BSMV, as previously described (Lee et al., 2015). Two independent silencing fragments were designed for each gene, in order to minimize the off-target effect. BSMV inoculations were performed on 12-day-old seedlings.
Primers used to generate the BSMV construct are listed in Table S1. VIGS inoculation was performed as described by Lee et al. (2015).

| Sequence analysis
To retrieve genes and protein sequences used in this study, the publicly available databases GenBank, SwissProt, PLANTTFDB, NCBIBlast, Plant-Ensembl were used (Kersey et al., 2003). Domain sequences were aligned using Clustal-Omega. These sequence alignments were used to verify the presence of all the conserved domains that define a WRKY transcription factor. Gene expression in response to JA treatment analyses in roots were repeated three times. In each experiment, two pools of the whole root system from three different seedlings were used as a technical replicate. Sampling was performed by excising root tissue daily, at 8 hr into the 16 hr light cycle.

| Hormone treatments
To evaluate wheat response to Z. tritici infection under JA treatment, infected leaves were sprayed with 10 μM JA solution every 4 days starting at 2-week-old wheat plants. In order to maintain a standardized infection of plants, Lili and Santiago were inoculated at 27 days after germination even if not silenced. The pathotests were performed as described above.

| Root growth measurement
To evaluate the effect of TaWRKY10 silencing on root growth, wheat seeds were germinated on plates as previously described, and then moved to pots of sand. The nutrients were supplied using 1/2 × Hoagland's solution. Seven weeks after germination the sand was removed by washing with water and pictures were taken and analysed using ImageJ software.

| Anthocyanin relative quantification
Anthocyanin extraction and quantification was done by collecting 3 g of leaf tissue for each sample, then extraction and quantification was done as previously described by Neff and Chory (1998). Three biological replicates were performed, with three technical replicates each.

| Susceptibility of wheat cultivars to STB and TaWRKY10 gene expression
As a first step to discover the role of WRKYs in Z. tritici defence, we defined the WRKY gene family in bread wheat using WRKY protein sequence information from the dicot model A. thaliana, and the monocot model rice (Figure 1a). Using WRKY protein sequence information from these model species (Table S1), we interrogated publicly available wheat sequence databases (PLANTTFDB, GenBank, SwissProt) and identified 71 WRKY genes in the wheat genome (TaWRKYs) and generated a phylogenetic tree based on predicted amino acid sequence similarity amongst wheat WRKY domaincontaining proteins (Figures 1 and S1).
This evolutionary tree indicated that TaWRKY10 ( Figure 1b) is a Pooideae-specific WRKY transcription factor with no known homologs in other lineages except in Hordeum vulgare (HvWRKY10). To ascertain the role of TaWRKY10 in STB, we investigated its transcription in two cultivars that are known to show different STB suscepti-

bilities: KWS Lili and KWS Santiago.
Under control conditions, messenger RNA from 2-week-old seedlings was used as a template for RT-qPCR analysis. As indicated in Figure 1c, Lili showed lower TaWRKY10 expression than Santiago, with the latter displaying 3-fold higher TaWRKY10 transcription.
Using STB susceptibility scores developed by the Agriculture and Horticulture Development Board (AHDB), an industry levy board in the UK that provides annual information on the productivity of different cereal varieties under field conditions, yield losses and foliar damage were reflected at the adult plant stage. However, we F I G U R E 1 Characterization of TaWRKY10. (a) Multiple sequence alignment of TaWRKY10 amino acid sequence with other known WRKY domain proteins (Arabidopsis, rice, maize, barley, and brassica). Alignment was generated using Clustal Omega. WRKY consensus sequence and zinc finger motifs are in yellow and green, respectively. TaWRKY10 belongs to the subfamily of type II WRKY transcription factors. (b) Phylogenetic tree generated using Clustal Omega. TaWRKY10 displays only one homolog in Hordeum vulgare. (c) Quantitative reverse transcription-PCR performed on leaves of 2-week-old wheat seedlings of cultivars Lili and Santiago. Error bars, ± SEM of raw data for three biological replicates observed differential TaWRKY10 gene expression at the 2-week-old growth stage, and therefore to reconcile these differences and to ascertain if we could observe the same susceptibility phenotypes in 4-to 6-week-old plants as observed in field conditions, we performed Z. tritici infection assays in wheat varieties under controlled conditions in the laboratory.
Both varieties were infected with virulent Z. tritici isolate IPO323 (Arraiano & Brown, 2006). White light pictures were taken daily, with pycnidia and spore counts performed at 30 dpi, as previously described (Lee et al., 2015).
Lili leaves started to show the first symptoms of STB necrotic lesions between 14 and 15 dpi, while for Santiago these necrotic lesions were observed earlier, at 12 dpi ( Figure 2a). To provide a more accurate quantification of STB symptoms, the percentage of blotched area in the leaves was determined from white light pictures using ImageJ software, with the data analysed by exploiting the AUDPC method (Madden et al., 2007). The AUDPC quantification indicated a significant difference in disease severity and progression between Lili and Santiago at 11 to 17 dpi, with a value of 251.84 for Lili and 413.25 for Santiago (Figure 2b,c). These data indicated that the spread of the lesions was slower in Lili compared to Santiago. Z.
tritici sporulates within necrotic lesions containing pycnidia on the surface of leaves (Keon et al., 2007). Therefore, we quantified sporulation by performing pycnidia and spore counts via the spore wash technique (Lee et al., 2015). Z. tritici in Santiago produced on average 43 pycnidia per leaf and 250 spores/µl, while on Lili it only produced on average 33 pycnidia per leaf and 180 spores/µl (Figure 2d,e). The data demonstrate that Lili produced significantly fewer pycnidia and spores per leaf, indicating that Z. tritici is less successful in causing STB in Lili compared to Santiago (Figure 2d,e).
F I G U R E 2 Susceptibility to Septoria tritici blotch (STB) at the seedling stage in wheat cultivars Lili and Santiago. (a) White light picture of a single leaf of Lili and Santiago infected with Zymoseptoria tritici over a time-course from day 9 to day 17. Area under the disease progress curve (AUDPC) corresponding to Lili (b) and Santiago (c) (average of three biological replicates), calculated by evaluating the percentage of blotched leaf area of each cultivar. The mathematical calculation was performed using the appropriate R package. (d) Average number of pycnidia per leaf in Lili and Santiago; Lili was significantly more resistant to STB than Santiago (Student's t test, p = 3.3 × 10 −4 ). (e) Spore wash was performed 32 days after initial Z. tritici inoculation. Student's t test showed significance difference between Lili and Santiago sporulation p = 1.2 × 10 −9 . Level of expression of TaWRKY10 gene over 22 days of STB time-course, in cultivars Lili (f) and Santiago (g). Error bars, ± SEM of raw data for three biological replicates A time-course experiment indicated that TaWRKY10 gene expression was down-regulated in both cultivars in response to the pathogen, 2 days prior to the switch to the necrotrophic phase ( Figure 2f,g). Taken together with the differential susceptibility to STB in Lili and Santiago, our data suggest a potential involvement of TaWRKY10 in wheat defence against STB.

| Effect of silencing TaWRKY10 gene expression on resistance to STB
Because Santiago showed the greatest expression of TaWRKY10, and was the most susceptible variety to STB, we wanted to ascertain if TaWRKY10 gene activity could contribute to the differences in susceptibilities between Lili and Santiago.
Due to the large hexaploid wheat genome, isolating single gene knockouts is currently not feasible. Therefore, VIGS (Baulcombe, 1999) was employed to silence TaWRKY10. BSMV has previously been modified to effectively silence the expression of many wheat genes (Matthew, 2004). Gene silencing vectors based on BSMV (Lee et al., 2012), carrying the 5′ untranslated region (UTR) and 3′ UTR of TaWRKY10 mRNA, were used to induce sequence-specific Fourteen days after silencing treatment, the third to fifth wheat leaves were subjected to Z. tritici inoculation (7.5 × 10 6 spores/ml) and the symptom development was recorded for 28 days (Figure 3c Pycnidia formation was significantly reduced on TaWRKY10-silenced leaves, with 45% reduction on BSMV:TaWRKY10_1-and 43% reduction on BSMV:TaWRKY10_2-infected plants compared to empty vector BSMV:00 control ( Figure 3d). As expected from the reduced number of fruiting bodies, spore production was also reduced in the TaWRKY10-silenced wheat leaves (Figure 3e), with a 50% reduction on BSMV:TaWRKY10_1-and a 46% reduction on BSMV:TaWRKY10_2infected plants when compared to the BSMV:00 control.
BSMV:TaWRKY10_1 was more efficient at silencing TaWRKY10 than BSMV:TaWRKY10_2. Strikingly, the spore production in the BSMV:TaWRKY10-silenced Santiago lines was equivalent to spores produced from BSMV:00 control Lili plants (Figure 3e). These data indicate that TaWRKY10 may play a role in determining the susceptibility phenotypes of cvs Lili and Santiago. Furthermore, silencing TaWRKY10 enhanced Lili resistance to STB.

| Sensitivity of wheat cultivars to JA
It is well established that JA is a growth regulator as well as a key hormone in defence signalling (Chini et al., 2016;Goossens et al., 2016).
In order to assess if JA could have a role during the immune response against STB, we treated Lili and Santiago with JA and infected them with Z. tritici. Our data indicate that both cultivars showed increased resistance to STB when treated with JA (Figure 4a; Figure S2).
JA is a pivotal hormone in the control of root growth (Staswick et al., 1992). To ascertain if the JA response was different in Lili compared to Santiago, we germinated seeds of both cultivars on MS agar plates, then transferred 2-day-old seedlings to plates that were supplemented with 10 μM JA to monitor root growth daily. As shown in Figure 4b,c, seedlings of Lili showed root growth inhibition after 24 hr (40% root growth inhibition) compared to Santiago.
Lili roots barely elongated at all for the first 24 hr of JA treatment, indicating dramatic JA-mediated root growth inhibition response.
At 48 hr after JA treatment, there was similar root growth inhibition in both cultivars. Santiago roots remained unresponsive to JA and elongated at rates comparable to untreated seedlings in the first 24 hr of JA treatment, after which they exhibited growth inhibition at 48 hr. Our data indicates that Lili was more sensitive to JA than

Santiago.
JA treatment is also known to induce anthocyanin accumulation in plants (Chini et al., 2016) and this was observed to a much greater extent in Lili compared to Santiago, with 3-fold and 1.8-fold more accumulation, respectively, compared to their controls (Figure 4d,e).
Our data demonstrate that Lili is more responsive to JA, resulting in greater root growth inhibition and with more pronounced anthocyanin accumulation compared to Santiago. The evidence indicates that the observed sensitivity to JA between the cultivars might explain the greater resistance to STB in Lili.
To ascertain whether TaWRKY10 silencing leads to any changes in morphology, we analysed plant height and root length in the silenced lines and compared it to empty vector controls. Plants from both varieties were pregerminated on filter paper for 3 days and subsequently moved to sand-filled pots, and grown for 10 days before BSMV inoculation. Seven weeks after germination, the sand was washed away, the roots and the shoots were photographed, and the length quantified via ImageJ software. No significant difference was observed in plant height in 7-week-old TaWRKY10silenced plants compared to controls (data not shown). However, in cv. Santiago we observed a root growth inhibition of 25% ( Figure   S3) compared to empty vector controls. A similar phenotype was also observed in Lili ( Figure S4). This evidence suggests a role for TaWRKY10 in regulating root growth as well as immunity against STB.
Downstream response markers were monitored by analysing expression of genes encoding plant defensin 1.1 (TaPDF1) and tyrosine aminotransferase (TaTAT1).
Tissue samples from leaves and roots of JA-treated or mocktreated Lili and Santiago were used to extract mRNA at 0, 24, and 48 hr time points, where we observed the greatest differential JA response in root growth. We observed a down-regulation in the expression of genes that affect JA biosynthesis between the two cultivars after JA treatment, possibly due to a negative feedback mechanism (Figure 5a-d). Intriguingly, gene expression of JA receptor TaCOI1 was up-regulated dramatically in both shoot and root in Lili at 24 hr after the treatment, by 6-fold and 8-fold, respectively (Figure 5e), while in Santiago TaCOI1 was not up-regulated even 48 hr after JA treatment (Figure 5f). The JA signal transduction marker gene TaMYC2 was not up-regulated significantly in either cultivar. The JA signalling repressor TaJAZ was up-regulated 14-fold and 10-fold in Lili shoot and root tissues, respectively, but only 7-fold and 4-fold in Santiago shoot and root tissues 48 hr after JA treatment (Figure 5g,h,i,l). The downstream JA response gene TaTAT1 was more active in Lili at 24 hr after JA treatment. Moreover, a second JA marker gene, TaPDF1, showed enhanced transcription in Lili at 48 hr after JA treatment (Figure 5m,n,o,p). Our data indicate that JA F I G U R E 4 Wheat cv. Lili is more sensitive to jasmonic acid (JA) compared to cultivar Santiago. (a) Wheat leaves were treated with either JA or mock treated every 4 days and spores counted at the end of the infection. Treated leaves displayed a statistically significant decrease of Zymoseptoria tritici sporulation compared to their mock-treated control (Lili p = 1.6 × 10 −3 ; Santiago p = 4.6 × 10 −3 ). Biological and technical replication was performed similarly to previously described experiments in this work. (b) White light pictures of wheat seedlings of cvs Santiago and Lili exposed to JA compared to their controls. Seeds were germinated on Murashige and Skoog medium (MS) for 2 days, imaged at 0 hr and then moved either to MS or MS supplemented with 10 µM JA. Pictures were taken 24 and 48 hr after treatment. (c) Quantification of root growth during JA treatment. Data were obtained analysing white light pictures with ImageJ software. Student's t test did not show a statistical difference of Lili root length from 0 to 24 hr of JA treatment (p = 0.95) whereas a statistically significant increase in root length was observed for Santiago in the same time frame (p = 2.2 × 10 −3 ). (d) White light picture of seedlings after 5 days of JA treatment. (e) Relative quantification of anthocyanin accumulation in seedlings via spectrophotometer analysis. Error bars, ± SEM of raw data for three different biological replicates. Student's t test showed significant difference for Lili treated compared to its own control (p = 3.6 × 10 −5 ); Santiago did not show any significant increase in anthocyanin content upon JA treatment (p = 0.9) perception and response is dampened in Santiago compared to Lili.
Taken together with expression differences in TaWRKY10 between the cultivars, our data suggest that there may be a link between JA responses and TaWRKY10 gene expression.

| TaWRKY10 silencing and JA signalling
Given that TaWRKY10 silencing led to altered immune responses in wheat, and cultivars that showed different TaWRKY10 levels had different JA responses, we wanted to explore the possibility that TaWRKY10 expression could be linked to JA signalling. We  Figure 6 demonstrates that there was no significant difference in expression of TaLOX1 genes encoding components of JA biosynthesis between the TaWRKY-silenced lines and nonsilenced control ( Figure 6a). Surprisingly, we detected a 4-fold increase in TaCOI1 transcript in the TaWRKY10-silenced lines (Figure 6b). We also noticed that TaCOI1 transcript was present at 10-fold higher level in Lili compared to Santiago in the empty vector controls.
We also observed an up-regulation of the JA-responsive transcriptional activator TaMYC2  in the TaWRKY10silenced lines (Figure 6c). Not surprisingly, TaPDF1, a downstream target gene of TaMYC2, was also up-regulated in the TaWRKY10knockdown plants (Figure 6d). Our data indicate that TaWRKY10 acts as a repressor of TaCOI1 gene expression. We confirmed the presence of W-boxes cis-elements in the TaCOI1 promoter region ( Figure S5).
We demonstrated that manipulating TaWRKY10 gene expression in wheat cultivars led to altered JA perception via changes in the JA receptor COI1 gene expression. Because JA is a major hormone that regulates defence against necrotrophic pathogens, our data provide a plausible mechanistic explanation for the differential Z. tritici susceptibility phenotypes observed in the wheat cultivars with altered The WRKY domain-containing transcription factor gene family is universally present in plants and is an integral part of plant defence signalling pathways (Bakshi & Oelmüller, 2014). Several recent studies confirm the role of WRKY transcription factors as major regulators of immune responses in crop plants Satapathy et al., 2018).
Here, we identified TaWRKY10 as a novel component of wheat immunity against STB. Among the TaWRKYs identified to date in wheat, TaWRKY10 shows homology to only a single WRKY factor from H. vulgare, indicating that TaWRKY10 is a Pooideae-specific gene in T. aestivum, and is only the second such gene to be characterized (Perochon et al., 2015). Interestingly, the other Pooideaespecific gene, TaFROG, is also reported to be involved in immunity against a wheat pathogen, suggesting that these genes may have evolved to play a common role in wheat.
Two different cultivars, Santiago and Lili, were selected to understand the role of TaWRKY10 in more detail. Initially, we demonstrated that these varieties show different degrees of resistance to STB. Lili (lower TaWRKY10 transcript levels) is more resistant to STB than Santiago (higher TaWRWKY10 transcript levels). To assess if TaWRKY10 expression level decrease has direct correlation with wheat defence against Z. tritici, we performed gene silencing experiments to down-regulate TaWRKY10 expression.
Silencing TaWRKY10 in both Lili and Santiago cultivars led to reduced sporulation but an earlier onset of the Z. tritici necrotrophic phase, similar to that observed for another negative regulator of immunity, TaR1, in wheat (Lee et al., 2015). Thus, for both of the cultivars the resistance to STB increased as a consequence of silencing TaWRKY10, suggesting that TaWRKY10 may be a contributing factor during STB response.
TaWRKY10 mRNA levels were down-regulated in response to Z. tritici approximately 2-3 days prior to the onset of the necrotrophic stage. Silencing TaWRKY10 enhanced TaCOI1 expression. It has been previously shown that up-regulating COI1 expression enhances JA signalling . Therefore, our data indicates that TaWRKY10 is a negative regulator of JA signalling, as silencing TaWRKY10 enhances TaCOI1 expression. Because TaWRKY10 is a negative regulator of JA signalling, this suppression of gene F I G U R E 6 TaWRKY10 gene silencing affects jasmonic acid (JA) signalling in wheat cultivars Lili and Santiago. Quantitative reverse transcription-PCR data shows the expression of (a) TaLOX1, a gene involved in JA biosynthesis pathway; (b) the JA receptor TaCOI1; and marker genes (c) TaMYC2 and (d) TaPDF1. Error bars, ± SEM of raw data for three biological replicates expression in response to Z. tritici indicates a response by wheat to up-regulate JA-mediated defence.
Hormone signalling pathways play a pivotal role in plant response to environmental stimuli. JA is one of the main hormones involved in plant-pathogen interactions, and the presence of JA receptor COI1 is necessary to propagate the hormonal signal and thus the plant response (Chini et al., 2016;Goossens et al., 2016). Despite the wealth of knowledge of hormone signalling pathways in the model dicot A. thaliana, little is known about hormone signalling in monocots, especially in wheat. The current study sheds new light into regulation of JA signalling in wheat.
It has been previously reported that different WRKY transcription factors play opposite roles in the JA-mediated defence against the same necrotrophic pathogen. For example, overexpressing a wild grape WRKY transcription factor in Arabidopsis led to downregulation of JA response after B. cinerea inoculation, leading to increased susceptibility (Wang et al., 2017), whereas overexpressing Populus trichocarpa WRKY40 in poplar trees led to increased JA level and increased resistance to B. cinerea (Karim et al., 2015). This evidence indicates species specificity and functional diversity within the WRKY transcription factor gene family.
The inhibitory effect of JA on root growth is well established (Staswick et al., 1992). We demonstrated that silencing TaWRKY10 led to root growth inhibition, mimicking the effect of enhanced JA signalling on root growth inhibition. Taken together with TaWRKY10silenced plants exhibiting increased defence, our data demonstrate a negative role in JA signalling both in defence and also in growth control for TaWRKY10.
Moreover, we demonstrated that the down-regulation of TaWRKY10 transcript level led to the up-regulation of the JA receptor TaCOI1 gene transcription, establishing an inverse relationship between the expression patterns of these two genes. COI1 transcript abundance is directly correlated with JA sensitivity in rice (Yang et al., 2012). In the current study we demonstrated that silencing TaWRKY10 resulted in the up-regulation of TaCOI1 transcription, which might lead to the activation of the JA signalling pathway as demonstrated in other studies (Goossen et al., 2015) and in our observation of both root growth inhibition and increased resistance to STB. Therefore, it is likely that TaWRKY10 exerts a negative effect on JA signalling in wheat by suppressing TaCOI1 gene expression.
Because TaWRKY10 only exists in the subfamily Pooideae, this may be a novel mechanism for regulating JA signalling in wheat, but further work is needed to understand the mechanistic basis of how TaWRKY10 regulates TaCOI1 gene expression.
Until now there have been no studies that examine the sensitivity of different wheat varieties to phytohormones. We have been able to demonstrate that a more rapid JA perception may be regulated by TaWRKY10-mediated suppression of the JA receptor gene expression and is coupled with best disease resistance performance. Our data suggest that breeding programmes that target improved JA perception mechanisms could lead to novel wheat germplasm better suited to withstand Z. tritici infection, resulting in greater yield.

ACK N OWLED G EM ENTS
We acknowledge Fenella Croley for critical reading of the manuscript. A.C., L.M., J.L., and A.S. were funded by Biotechnology and Biological Research Council of the UK.

CO N FLI C T O F I NTE R E S T
The authors declare there is no conflict of interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available in the supplementary material of this article.