Suwon11 (Su11), a Chinese cultivar of wheat and two Pst races CYR23 (incompatible) and CYR31 (compatible) were used to study the wheat-Pst interaction. Su11 carries the resistance gene YrSu conferring resistance to CYR23 but not to CYR31 (Cao et al., 2002). Plant cultivation and Pst inoculation procedures and conditions were followed as described previously by Kang et al. (2002). After inoculation of the first leaves with freshly harvested urediospores of CYR23 and CYR31, leaf tissues were harvested at 0, 12, 24, 48, 72, and 120 h post-inoculation (hpi), and the control plants corresponding to each time point were treated with sterile water. The time points were selected based on the microscopic study of the wheat-Pst interaction (Wang et al., 2007). For chemical treatment assays, 2-week-old wheat seedlings were treated with 2 mM salicylic acid (SA), 100 mM methyl jasmonate (MeJA), 100 mM ethepon (ETH), or 100 mM abscisic acid (ABA) dissolved in 0.1% (v/v) ethanol. Control plants were treated with 0.1% (v/v) ethanol. Leaves of treated and control plants were harvested at 0, 2, 4, 12, and 24 h post-treatment (hpt). For different abiotic stresses, high salinity or drought stress, the roots of 2-week-old wheat seedlings were soaked in 200 mM NaCl (causing osmotic salt stress due to high salinity) or 20% PEG6000 (causing drought stress due to water deficiency), respectively. To assess the effects of wounding, the first leaves were scraped with a sterilized needle. For low-temperature treatment, wheat seedlings were transferred to a 4°C chamber. The first leaves treated with different chemicals and stress treatments along with leaves of the control plants were harvested at 0, 2, 4, 12, and 24 hpt. Intact tissues of different wheat organs from 2-week-old seedlings were collected for tissue-specific expression analysis. All freshly collected samples were rapidly frozen in liquid nitrogen and stored at −80°C until the extraction of total RNA or DNA. For each time point, three independent biological replications were performed.

The DNeasy Plant Mini Kit (Qiagen) was used for extraction of Genomic DNA from wheat leaves. Total RNA from wheat leaves that ware challenged with chemicals, abiotic stresses elicitors, Pst as well as different wheat tissues were extracted with the Trizol TM Reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. Contaminating DNA was removed by treatment with DNase I. First strand cDNA was synthesized using 2 μg of total RNA with the RT-PCR system (Promega, Madison, WI, USA) and Oligo (dT)18 primer.

PCR was performed with the primers TaDIR1-2(ORF)-S and TaDIR1-2(ORF)-AS (Supplementary Table S1) designed based on the sequences of DIR1-2 orthologous group (Supplementary Figure S1) to amplify TaDIR1-2. The amplified product was cloned into the pGEM-T Easy Vector (Promega, Madison, WI, USA) for sequencing. This cloned sequence was aligned with the wheat cv. Chinese Spring genome, based on the data of International Wheat Genome Sequencing Consortium (https://urgi.versailles.inra.fr/blast/). The chromosomal location and predicted related sequences were also obtained from this website. ORF Finder and the BLAST (https://www.ncbi.nlm.nih.gov/). programs ware used for the analyses of cDNA and amino acid sequences. Conserved domains were identified using Pfam (http://pfam.xfam.org/). and Inter ProScan (http://www.ebi.ac.uk/interpro/search/sequence-search). Multiple sequence alignment was performed with DNAMAN software (Lynnon Biosoft, USA), and Mega 5.0 software was used for phylogenetic tree construction (Tamura et al., 2011). si-Fi software v1.4.0 (http://labtools.ipk-gatersleben.de/) was used to identify off-target sequence.

Specific primers were designed (Supplementary Table S1) and used as described previously (Duan et al., 2011). The ABI PRISM 7500 software tool (Applied Biosystems, Foster City, CA, USA) was used to generate threshold values (CT) for the quantification of relative gene expression using the comparative 2^−ΔΔCt method (Livak and Schmittgen, 2001) and the data were normalized against expression of the wheat elongation factor TaEF-1α gene (GenBank accession no. Q03033) (Paolacci et al., 2009). All reactions were performed in triplicate. To ensure specific amplification, dissociation curves were generated for each reaction. The qRT-PCR analysis for respective experiment was replicated three times with similar results.

For subcellular localization in wheat protoplasts, the TaDIR1-2 protein-encoding sequence was amplified and inserted into the PstI and XbaI sites of the pCaMV35S::GFP vector to generate the pCaMV35S::TaDIR1-2::GFP fusion vector by PCR using primers TaDIR1-2(163)-S and TaDIR1-2(163)-AS (Supplementary Table S1). The protoplasts were isolated from mesophyll tissue of 8–10-day-old wheat seedlings as previously described (Li et al., 2015). The fusion pCaMV35S::TaDIR1-2::GFP construct and the control plasmid pCaMV35S::GFP were transformed into wheat protoplasts by the PEG-mediated transformation system (Bio-Rad, Hercules, CA, USA), and then the PEG-transfected mesophyll protoplasts were submersed in W5 solution and incubated at 23°C for 18 h in a dark chamber. Fluorescent signals were observed with a Zeiss LSM510 confocal laser microscope (Zeiss, Germany) with a 480-nm filter as previously described (Ito and Shinozaki, 2002).

The recombinant vectors, PVX-TaDIR1-2, PVX-pBin19, PVX-Bax, PVX-Avr1b, and PVX-eGFP were constructed by inserting the protein-encoding sequence into the ClaI and SalI sites using the respective primers (Supplementary Table S1) and transformed individually into Agrobacterium tumefaciens strain GV3101 as described previously (Dou et al., 2008). A. tumefaciens cell suspensions carrying the respective plasmids were cultured in 5 ml of LB medium containing kanamycin, rifampicin and gentamycin, and then collected at an OD600 of 0.6–1.2 and re-suspended in 10 mM MgCl2 to a final density of 0.2–0.3 at OD600. The cells were infiltrated into 4–6-week-old tobacco (Nicotiana benthamiana) leaves as described by Wang et al. (2011). The same infiltration site was challenged with the Bax gene carrying A. tumefaciens cell suspensions at 16 h after f the 1st infiltration. PVX-empty vector (EV) and PVX-Avr1b were infiltrated as negative and positive controls, respectively. Subsequently, PVX-pBin19 was used to suppress gene silencing of PVX-TaDIR1-2 (Voinnet et al., 2003), and PVX-eGFP was used to assess the efficiency of the experiments. Green fluorescence was identified in eGFP-treated leaves at 4 days after the 2nd infiltration. Symptom development was examined and photos were taken 4 days after the 2nd infiltration. The experiment was replicated three times with similar results.

A small fragment with 155-bp was used to silence TaDIR1-2. The fragment of TaDIR1-2 with NotI and PacI restriction sites was obtained by reverse transcription PCR to modify the original BSMV:γ vector. Capped in vitro transcripts of BSMV RNAs were prepared from the linearized plasmids γ-TaPDS-as, γ-TaDIR1-2, γ, α, β using a Message T7 in vitro transcription kit (Ambion, Austin, TX) according to the manufacturer's protocol. During the inoculation, the RNA transcripts were diluted 4-fold, and 2.5 μL of each transcript (BSMV RNA α, β, γ; γ-TaPDS and γ-TaDIR1-2) were mixed with 42.5 μL of FES buffer (Pogue et al., 1998). The mixture was inoculated individually onto the second leaf of 2-leaf wheat seedlings by gently rubbing the leaf surface with a gloved finger as described previously (Scofield et al., 2005). The virus-infected wheat seedlings were incubated for 24 h in darkness and high humidity, and then incubated in a growth chamber at 25 ± 2°C with a 16-h photoperiod. In total, 40 seedlings were inoculated with each of the three viruses (BSMV:γ, BSMV:TaPDSas and BSMV:TaDIR1-2 as). In addition, 40 seedlings were treated with FES buffer as the control. To check the BSMV infection, BSMV:γ and BSMV: TaPDS were used as controls. After 10 dpi, the fourth leaves of BSMV-infected seedlings were further inoculated with freshly harvested urediospores of Pst race CYR23 or CYR31, and the Pst infected plants were consequently maintained in the condition described above. After 15 dpi, the infection types of Pst were examined on the basis of McNeal measurements scale (McNeal et al., 1971) and photos were taken. Three independent sets of plants were prepared for each assay. The Pst-infected fourth leaves were collected at 0, 24, 48, and 120 hpi for histological observation as well as RNA isolation. The silencing efficiency of TaDIR1-2 knockdown plants and the relative transcript levels of the pathogenesis related (PR) protein genes TaPR1 and TaPR2, asecondary metabolite gene TaPAL, and ROS related genes, including superoxide dismutase (TaSOD), catalase (TaCAT) and NADH-oxidase (TaNOX), were analyzed by qRT-PCR in comparison with the control plants in each assay as described above. The primers were used to perform qRT-PCR are listed in Supplementary Table S1.

The fungal development in TaDIR1-2-knockdown plants or control plants challenged with Pst was examined microscopically. For histological observation, collected leaf samples were cut into 1.5 cm long segments and treated with ethanol: acetic acid (1:1 v/v) for decolorization, and saturated chloral hydrate was used to clarify the leaf tissue. To examine the necrotic cell area, the auto-fluorescence of the pathogen-attacked necrotic cells was observed under a fluorescence microscope (excitation filter 485 nm, dichromic mirror 510 nm, and barrier filter 520 nm) and the necrotic area was measured using DP-BSW software. To measure the area of H₂O₂ accumulation in the infection sites, leaf samples were stained with 3,3′-diamino benzidine (DAB; Amresco, Solon, OH, USA) as previously described (Wang et al., 2007). H₂O₂ accumulation was observed under differential interference contrast optics and the area was measured with DP-BSW software. The hyphal length, haustoria, haustorial mother cell and infection area were determined under an Olympus BX-53 microscope (Olympus Corp., Tokyo) and the length and area were calculated with DP-BSW software. According to Bozkurt et al. (2010), after entering the Pst urediniospores in the plant, a substomatal vesicle forms within the stomatal cavity, from which three infection hyphae grow, an infection peg develops and breaching the mesophyll cell wall, were considered as successful penetrations and were microscopically examined to assess the infection hyphae, haustoria, haustorial mother cell and infection area. For each treatment, 50 infection sites were examined. Standard deviations and Student's t-test were used for statistical analysis.

For the quantification of SA, approximately 100–200 mg of fresh leaf tissue of each sample was used to extract SA, and analyzed with HPLC-MS (API 2000; AB SCIEX, Framingham, USA) by the protocol previously described (Segarra et al., 2006).

Microsoft Excel software was used to calculate mean values and standard errors. A one-way analysis of variance (ANOVA) was performed using the SPSS 16.0 (SPSS Inc., Chicago, Illinois, USA) statistical software to determine the significant differences between control and treatment or between time-course points. The probability (P) value < 0.05 was used to measure the significant change with unequal variance.

To systematically identify DIR1 orthologs from T. aestivum, Aegilops tauschii (A. tauschii) and Triticum urartu (T. urartu), a genome-wide search using DIR1 protein sequences from Arabidopsis and rice as query revealed a total of 25 putative DIR1 loci in wheat (3, 7, and 15 were obtained from the Aegilops tauschii, Triticum urartu and Triticum aestivum genome databases, respectively). Out of 15 putative DIR1 protein sequences from the Triticum aestivum genome database, loci 5, 8, and 2 were located in sub-genomes A, B, and D, respectively. To better understand the evolutionary relationship of putative DIR1 orthologs in wheat with other plant DIR1 proteins, a phylogenetic tree of wheat, rice, Arabidopsis DIR1 proteins was constructed based on the full-length amino acid sequences (Supplementary Figure S1). The resulting dendrogram indicated that putative DIR1 orthologs from wheat were clustered into six major groups, including DIR1-1, DIR1-2, DIR1-3, DIR1-4, DIR1-5, and DIR1-6. DIR1-1 orthologs showed highest orthology with AtDIR1 in Arabidopsis and OsDIR1-A (RICE-A) in rice. However, the DIR1 orthologs in the other five groups showed much higher similarity with OsDIR1-B (RICE-B) than OsDIR1-A, AtDIR1, and AtDIR1-like.

According to DIR1 sequences identified from the wheat genome, we designed primers and cloned one DIR1 orthologous gene in the DIR1-2 group from wheat cv. Su11, designated TaDIR1-2. Sequence alignment indicated that TaDIR1-2 has high identity with DIR1 proteins from monocot plants, 57–99% (60–94% nucleotide) with other DIR1 orthologs in wheat (Supplementary Figure S2), 66% (60% nucleotide) with rice OsDIR1-B and 60% (64% nucleotide) with Brachypodium BdDIR1, respectively. In contrast, TaDIR1-2 showed only 45% identity with Arabidopsis AtDIR1. The protein conserved domain analysis showed that TaDIR1-2 contains a nsLTP region from 29 to 100 amino acids residues, an eight cysteine signature with a central motif (C68-X-C70), and a proline rich domain (Figure 1A).

To study the evolutionary relationships of DIR1 orthologs, putative DIR1 orthologous proteins from eudicots and monocots plants, respectively, were selected to construct a maximum likelihood tree. The result showed that the putative DIR1 from different organisms formed a monophyletic group, NsLTP-Type II containing monocot and eudicot subgroups. OsLTP2B, TuNSLTP, and HvLTP4.2 were clustered into NsLTP-Type I as an out group (Figure 1B). TaDIR1-2 that clustered in the monocot subgroup showed closer relationships with SbDIR1, SiDIR1, ZmDIR1, BdDIR1, and OsDIR1-B than OsDIR1-A (Figure 1B).

In Arabidopsis LTP3 protein was localized to the cytoplasm (Guo L. et al., 2013). In wheat TaLTP3 protein was shown to be localized in the cell membrane and cytoplasm of tobacco epidermal cells (Wang et al., 2014). In this study, subcellular localization of TaDIR1-2 in wheat cells was determined by the transfection of recombinant pCaMV35S::TaDIR1-2::eGFP into mesophyll protoplasts, and the empty pCaMV35S::eGFP vector was used as the control. By laser-scanning confocal microscopy, the green fluorescence of fusion eGFP was detected throughout the cell, including in the nucleus. In contrast, pCaMV35S::TaDIR1-2::GFP fusion proteins were localized in the cytoplasm, perhaps in the cell membrane, of wheat mesophyll protoplasts (Figure 2).

In Arabidopsis, DIR1 is expressed constitutively to low levels in all the tissues tested including seedlings, roots and flowers and in the living cells of leaves (Champigny et al., 2011). To examine the expression patterns of TaDIR1-2, the transcript level of TaDIR1-2 in different tissues of 2 weeks old wheat plants was measured by qRT-PCR analysis. The results revealed that TaDIR1-2 transcripts were detectable in all tested wheat tissues (root, stem and leaf) and the abundance of TaDIR1-2 transcripts in stem and leaf were significantly higher than in root (Figure 3A).

During infection with Pst, TaDIR1-2 expression was induced to different extents in compatible (CYR31) and incompatible (CYR23) interactions (Figure 3B). During the compatible interaction, theTaDIR1-2 transcript level was induced at all-time points except at 24 hpi, when its transcript level slightly decreased. At 72 hpi, TaDIR1-2 transcripts reached a peak which was 6.5-fold higher than that in the control plants. However, in the incompatible interaction, no significant up- or down-regulation (P < 0.05) for TaDIR1-2 transcript level was observed (Figure 3B). Thus, it is reasonable to assume that TaDIR1-2 may contribute to negative regulation of wheat resistance against Pst infection. To support this supposition, further experiments were performed subsequently.

To determine whether the transcription of TaDIR1-2 was affected by exogenous hormone elicitors and abiotic stresses, we performed the qRT-PCR analysis. As shown in Figure 4A, TaDIR1-2 was mainly induced by SA treatment but showed no significant response to the other treatments. In the SA treatment, the expression of TaDIR1-2 was increased at all-time points and peaked at 2 hpt with 4.5-fold expression. These results suggested that TaDIR1-2 may function in the SA-dependent signaling pathway.

For abiotic stresses, wheat seedlings were treated with NaCl, low temperature (4°C), PEG6000 (Khodarahmpour, 2011) and wounding. The results showed that the low-temperature (4°C) significantly up-regulated the expression of TaDIR1-2. In contrast, salt stress, drought stress, or wounding did not cause any significant changes in the expression of TaDIR1-2 (Figure 4B).

PCD is an important element for plant immunity against biotic and abiotic stress as well as for plant development and proliferation (Gadjev et al., 2008), including vacuolar cell death, hypersensitive cell death and necrosis which is induced by abiotic stresses through mitochondrial dysfunction that include uncoupling of respiration, the production of ROS and RNS, a drop in ATP level and mitochondrial membrane permeabilization (Van Doorn et al., 2011). Members of the diverse BAX family proteins in animals, have emerged as important regulators of PCD, which act as either activators or suppressors of PCD in animal as well as in transgenic plants (Kroemer, 1997; Mitsuhara et al., 1999). VAD1 (vascular associated death1), containing a lipid binding signaling domain and ACD1 (accelerated cell death1) gene encoding a protein homologous to a mammalian glycolipid transfer protein (GLTP), are expected to control the cell-to-cell propagation of PCD in Arabidopsis (Brodersen et al., 2002; Lorrain et al., 2004). To determine whether TaDIR1-2 could induce cell death or suppress PCD induced by the pro-apoptotic protein BAX, TaDIR1-2 was transiently overexpressed in tobacco leaves using potato virus X (PVX) delivery in combination with the Bax system by an A. tumefaciens-mediated transient overexpression assay. Agrobacterium strains carrying the recombinant genes were infiltrated into N. benthamiana leaves 16 h prior to infiltration with Bax-carrying strains. The green fluorescence was detected at 4 dpi of 2nd infiltration with eGFP treatment (Supplementary Figure S3). When leaves were infiltrated with empty vector (EV), Avr1b, (TaDIR1-2+pBin19) or eGFP, no cell death phenotype was observed (Figures 5A,B; circles 1, 2, 3, and 4). However, tobacco leaves infiltrated with EV+Bax, (TaDIR1-2+p19)+Bax or eGFP+Bax (Figures 5A,B; circles 5, 7, and 8) showed the Bax-triggered PCD phenotype at 4 dpi of 2nd infiltration. In addition, leaves infiltrated with Avr1b+BAX, suppressed the Bax-induced PCD and no cell death symptoms appeared (Figure 5; circle 6). These results indicated that TaDIR1-2 by itself is unable to induce PCD or suppress Bax-triggered PCD.

The function of TaDIR1-2 in the wheat defense response against Pst was examined by reducing the expression of TaDIR1-2 through the BSMV-mediated VIGS system (Scofield et al., 2005). To increase the silencing efficiency on TaDIR1-2, a 155-bp fragment was selected and designed from a region of the TaDIR1-2 coding sequence which was relatively conserved in other 11 wheat DIR1 orthologs identified in this study (Supplementary Figure S2; Supplementary Tables S2, S3). In the following experiments, the second leaves of Su11 seedlings at the two-leaf stage were inoculated with BSMV:TaPDSas, BSMV: TaDIR1-2, and BSMV:γ. A similar number of seedlings were also mock-inoculated with a buffer that did not include BSMV. The photo-bleaching phenotype was observed when TaPDS (wheat phytoene desaturase gene) was silenced, which provides a visual indication of the occurrence of gene silencing in the leaves after virus inoculation. All the BSMV: TaDIR1-2 and BSMV:γ inoculated plants displayed mild chlorotic mosaic symptoms 10 dpi which did not appear to affect further leaf growth, confirming that the BSMV-mediated gene silencing system functioned correctly and could be used in further experiments. In contrast, the leaves in the mock-inoculated wheat plants developed normally during the observation (Figure 6A). To clarify whether TaDIR1-2 was successfully silenced, transcript levels of TaDIR1-2 were examined by qRT-PCR. The results demonstrated that the expression levels of TaDIR1-2 were decreased at 0, 24, 48, and 120 hpi by approximately 80, 85, 70, and 85% for the incompatible interaction and by 70, 75, 70, and 70% for the compatible interaction, respectively (Figure 6B), indicating that TaDIR1-2 was successfully silenced.

The fourth leaves in wheat were then inoculated with CYR23 (incompatible) or CYR31 (incompatible). In the incompatible interaction, obvious HR was expressed on mock-inoculated plants and on leaves that were previously infected with BSMV:γ or BSMV:TaDIR1-2 (Figure 6C). In contrast, in the compatible interaction, all leaves from the mock-inoculated plants and leaves that were previously infected with BSMV:γ, revealed a fully susceptible phenotype. Leaves of the TaDIR1-2-knockdown plants also exhibited a susceptible phenotype, but clear necrotic cell death was observed accompanied by reduced sporulation at 15 dpi (Figure 6D). After 15 dpi, 40 leaves infected by Pst were examined for each treatment and they were scored as 10 out of 15 leaves which showed reduced disease symptom. McNeal measurements scale were used to estimate the difference between the phenotypes of BSMV: γ and BSMV:TaDIR1-2 under the interaction with Pst (McNeal et al., 1971) Our results indicated that knockdown of TaDIR1-2 expression increased the resistance of wheat against Pst.

To determine whether the phenotypic changes in wheat leaves are associated with fungal growth and development, when TaDIR1-2 was knocked down, leaf segments inoculated with CYR23 and CYR31 were examined microscopically (Figures 7A–Ff, Supplementary Figures S4A–Ff). In the compatible interaction, hyphal branches, hyphal length, number of haustorial mother cells, and infected area were not significantly (P < 0.05) affected at 24 hpi (Figures 7G–J). However, silencing of TaDIR1-2 significantly decreased hyphal branches, hyphal length, number of haustorial mother cells, and infected area by 120 hpi (Figures 7G–J), respectively, at 48 hpi, the hyphal length and number of haustorial mother cells of Pst were significantly reduced in TaDIR1-2-knockdown plants compared to BSMV:γ infected plants (Figures 7H,I), indicating that knocking down the TaDIR1-2 transcription in wheat plants reduces susceptibility in response to virulent CYR31 infection. In the incompatible interaction, compared with BSMV:γ infected leaves, the number of hyphal branches, hyphal length and number of haustorial mother cells in BSMV:TaDIR1-2 infected leaves were significantly decreased at 24 and 48 hpi, respectively (Supplementary Figures S4G–I). In addition, the infection area in TaDIR1-2-knockdown plants was significantly reduced only at 120 hpi (Supplementary Figure S4J). These histological results indicated that suppression of TaDIR1-2 expression restricts fungal growth and thus enhances host resistance to avirulent and virulent Pst.

An oxidative burst and the accumulation of ROS with the occurrence of HR are important for resistance against rust fungi in wheat (Dmochowska Boguta et al., 2013). To further clarify the host resistance response in TaDIR1-2-knockdown plants, the induction of H₂O₂ during Pst infection was analyzed by histochemical observations after DAB staining. In the compatible interaction, H₂O₂ accumulated in TaDIR1-2-knockdown plants to levels about 2.5- and 1.5-fold higher than in the control plants at 24 hpi and 48 hpi (Figures 8A–E), respectively. Similarly, in TaDIR1-2-knockdown plants, the necrotic cell death area observed by auto-fluorescence also exhibited a significantly increased pattern at 24 hpi and 48 hpi (Figures 8a–d,F), respectively. In the incompatible interaction, compared with the control plants, the accumulation of H₂O₂ and necrotic cell death areas per infection site were significantly increased both at 24 hpi (Supplementary Figures S5Aa,Bb,E) and 48 hpi (Supplementary Figures S5Cc,Dd,F), respectively.

To determine whether knockdown of TaDIR1-2 expression can affect SA balance in wheat and thereby lead to increased resistance to Pst, we further quantified the SA content in TaDIR1-2-knockdown plants. As shown in Figure 9, the SA concentration in TaDIR1-2-knockdown plants was significantly increased and reached 300 ng/mg FW, which is 2-fold higher than that in the control plants. Our results suggested that silencing of TaDIR1-2 enhances wheat resistance against Pst by regulating ROS and the SA signaling pathway.

To confirm the increased resistance and ROS accumulation in TaDIR1-2-knockdown plants, the expression of three ROS-related genes, two PR protein genes (PR1 and PR2), as marker genes for SA as well as in HR (Van Loon, 1997) and a defense-related secondary metabolite gene phenylalanine ammonia-lyase (PAL) were examined by qRT-PCR (Figure 10, Supplementary Figure S6). In compatible and incompatible interactions, the transcript levels of ROS-scavenging genes, both TaCAT (Figure 10A, Supplementary Figure S6A) and TaSOD (Figure 10B, Supplementary Figure S6B), were significantly decreased at 24, 48 and 120 hpi in TaDIR1-2-knockdown plants compared with the controls plants. TaNOX, a ROS-generating gene, exhibited a significantly increased expression at 24 and 48 hpi in TaDIR1-2-silenced plants compared to the control plants in compatible (Figure 10C) and incompatible (Supplementary Figure S6C) interactions. In addition, in compatible interaction the transcription of TaPR1 and TaPR2, were significantly induced at 48 and 120 hpi in TaDIR1-2 knockdown plants, respectively (Figures 10D,E), and in incompatible interactions significant induction were showed at 24 hpi (Supplementary Figures S6D,E). The transcript levels of the TaPAL were dramatically induced as early as 24 hpi in TaDIR1-2-knockdown plants in both the compatible (Figure 10F) and incompatible (Supplementary Figure S6F) interactions. In addition, at 0 hpi, which was considered as without a pathogen attack stage, no change in transcript levels of ROS related genes, PR protein genes or PAL gene were observed in TaDIR1-2-knockdown plants compared with the control plants (Figure 10 and Supplementary Figure S6), indicating that viral infection was not involved in defense response and the knockdown of TaDIR1-2 expression resulted in H₂O₂ as well as SA accumulation during Pst infection. These results further suggested that knockdown of TaDIR1-2 expression enhances the resistance of wheat to the stripe rust fungus in ROS- and SA-dependent manner.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.