TaClpS1, negatively regulates wheat resistance against Puccinia striiformis f. sp. tritici

Background The degradation of intracellular proteins plays an essential role in plant responses to stressful environments. ClpS1 and E3 ubiquitin ligase function as adaptors for selecting target substrates in caseinolytic peptidase (Clp) proteases pathways and the 26S proteasome system, respectively. Currently, the role of E3 ubiquitin ligase in the plant immune response to pathogens is well defined. However, the role of ClpS1 in the plant immune response to pathogens remains unknown. Results Here, wheat (Triticum aestivum) ClpS1 (TaClpS1) was studied and resulted to encode 161 amino acids, containing a conserved ClpS domain and a chloroplast transit peptide (1–32 aa). TaClpS1 was found to be specifically localized in the chloroplast when expressed transiently in wheat protoplasts. The transcript level of TaClpS1 in wheat was significantly induced during infection by Puccinia striiformis f. sp. tritici (Pst). Knockdown of TaClpS1 via virus-induced gene silencing (VIGS) resulted in an increase in wheat resistance against Pst, accompanied by an increase in the hypersensitive response (HR), accumulation of reactive oxygen species (ROS) and expression of TaPR1 and TaPR2, and a reduction in the number of haustoria, length of infection hypha and infection area of Pst. Furthermore, heterologous expression of TaClpS1 in Nicotiana benthamiana enhanced the infection by Phytophthora parasitica. Conclusions These results suggest that TaClpS1 negatively regulates the resistance of wheat to Pst. Supplementary Information The online version contains supplementary material available at 10.1186/s12870-020-02762-0.


Background
To ensure their survival in nature, plants must evoke many complicated mechanisms to cope with biotic and abiotic stresses. An increasing number of studies reveal the essential role that the degradation of intracellular proteins plays in plant responses to stressful environments. The protein degradation pathways include ubiquitin-26S proteasome system (UPS) and caseinolytic peptidase (Clp) proteases [1,2]. The two protein degradation machineries both consist of large multisubunit proteolytic complexes.
The first process in degradation of proteins by UPS is ATP-dependent ubiquitination, which involves the action of at least three main enzymes for selecting target substrates [3]: ubiquitin-activating enzymes (E1), ubiquitin-conjugating enzymes (E2), and ubiquitin ligases (E3). Studies reveal that ubiquitination during UPS is implicated in many biological processes in plants, including hormone signaling, growth and development, circadian rhythm control and cell cycle [4,5]. Recent studies have indicated that ubiquitination and E3 ubiquitin ligases are involved in plant immunity to pathogens. For instance, a RING finger E3 ubiquitin ligase, BLAST AND BTH-INDUCED1 (OsBBI1), was found to positively regulate resistance against Magnaporthe oryzae by modifying the rice cell wall [6]; SPL1, a rice U-box protein with E3 ubiquitin ligase activity, negatively regulates cell death and innate immunity against M. oryzae and Xanthomonas oryzae pv. oryzae (Xoo) [7]; SGT1, an ubiquitin ligase-associated protein, is required for induction of important defense mechanisms, including R genemediated defense mechanisms, systemic acquired resistance and basal defense [8]. These studies hinted that factors, which play an important role in selecting target proteins in protein degradation machineries, participate in regulating plant resistance to pathogens. The caseinolytic peptidase (Clp) protease-mediated protein degradation system initially discovered in bacteria consists of a proteolytic protein (mainly ClpP) and some regulatory AAA+ (ATPase associated with diverse cellular activities) proteins [9]. Importantly, regulatory AAA+ proteins use adaptor proteins to recognize and target specific substrates for degradation. For example, in Escherichia coli, the regulatory AAA+ protein ClpA uses the adaptor ClpS to recognize and target substrates for degradation by ClpAP protease [10][11][12]. These studies in E. coli indicated that the adaptor ClpS plays a central role in selecting target substrates for degradation by the ClpAP protease, which is the same as the function of E3 ubiquitin ligase in UPS. ClpS from E. coli was found to contain two conserved regions by analyzing the crystal structure of ClpS, and they were shown to be involved in the interaction with ClpA and substrates, respectively [13,14]. Intriguingly, the region involved in the interaction with substrates shared secondary structure homology with E3 ubiquitin ligases in UPS [15]. Moreover, previous phylogenetic analyses of ClpS proteins revealed evolutionary linkages among bacteria, cyanobacteria and plants [15,16]. Those studies raise the question: could ClpS proteins in plants participate in regulating plant resistance to pathogens just as E3 ubiquitin ligases function in plants?
The plant chloroplast Clp system comprises a heterooligomeric protease core complex consisting of five proteolytic subunits (ClpP1 and ClpP3-6) and four different subunits (ClpR1-4), ATP-dependent chaperones ClpC1/ 2 and ClpD, and an adaptor protein ClpS1, a redefinition of plant ClpS proteins based on subsequent phylogenetic analyses [17,18]. In Arabidopsis, several direct candidate chloroplast AtClpS1 substrates have been identified based on affinity purification methods, including glutamyl-tRNA reductase (GLUTR) and four enzymes in the shikimate pathway [17,19]. Moreover, the interaction of ClpS1 with the candidate substrates was strictly dependent on two conserved ClpS1 residues involved in recognizing and binding substrates, indicating that ClpS1 is a conserved substrate selector for the chloroplast Clp protease system [17]. In addition, AtClpS1 was reported to also interact with chloroplast chaperones ClpC1, 2 and adaptor ClpF, suggesting a model in which ClpS1 and ClpF form a binary adaptor for selective substrate recognition and delivery to ClpC [17,18]. However, these reports do not reveal that the plant ClpS1 protein participates in regulating plant resistance to pathogens.
Wheat stripe rust, caused by Puccinia striiformis f. sp. tritici (Pst), is one of the most widespread and destructive diseases of wheat worldwide [20]. In this study, the function of ClpS1 in plant responses to pathogens, based mainly on the interaction system of wheat and the stripe rust pathogen was studied. A ClpS1 gene from Triticum aestivum cv. Suwon 11, designated TaClpS1 was studied. The transcript level of TaClpS1 in wheat was induced by the Pst isolate CYR23 [21]. Knocking down TaClpS1 expression in wheat by virus-induced gene silencing (VIGS) attenuated Pst infection intensity and enhanced the accumulation of reactive oxygen species (ROS). Moreover, transient expression of TaClpS1 in N. benthamiana facilitated the infection of Phytophthora parasitica. These results suggested that TaClpS1 most likely serves as an enhancer of disease in plant which ultimately increases plant susceptibility to pathogen.

Identification of the wheat TaClpS1
In this study, wheat TaClpS1 was identified using the protein sequence of Arabidopsis AtClpS1 (GenBank accession no. NP_564937.1) blasting the hexaploid wheat genome databases (http://plants.ensembl.org/index.html). The results showed that six homologous sequences of AtClpS1 in wheat were located on chromosomes 2A, 2B, 2D, 3A, 3B and 3D, respectively. Phylogenetic analysis of the ClpS1 proteins from various plant species showed that the TaClpS1-2A, TaClpS1-2B, and TaClpS1-2D proteins are closely related to ClpS1 proteins from other plants, including Arabidopsis, Zea mays and Oryza sativa. The TaClpS1H-3A, TaClpS1H-3B, and TaClpS1H-3D proteins are not included in this group (Fig. 1a). Therefore, in this study we focused on the function of TaClpS1-2A, TaClpS1-2B, and TaClpS1-2D.
To further confirm the localization of TaClpS1 in wheat cells, the fusion constructs pCaMV35S: TaClpS1-GFP and pCaMV35S: TaClpS1Δ-GFP were generated and transformed into wheat protoplasts by polyethyleneglycol (PEG)-calcium. GFP fluorescence signals of pCaMV35S: TaClpS1Δ-GFP and pCaMV35S: GFP appeared in the nucleus, cytomembrane and cytoplasm of wheat protoplasts. In contrast to the results observed in N. benthamiana leaves, GFP fluorescence signals of TaClpS1-GFP aggregated mainly in chloroplasts of wheat protoplasts (Fig. 2b).

Relative transcript levels of TaClpS1 at different stages during Pst infection
To explore the role of TaClpS1 during Pst infection, the qRT-PCR assay was performed to examine the relative transcript levels of TaClpS1 at different stages during infection by Pst isolate CYR23. qRT-PCR data showed that the transcript levels of TaClpS1 increased as early as 12 h post-inoculation (hpi), continued to increase to 24 hpi, and subsequently diminished at 48 hpi, before the transcript levels rose again at 96 and 120 hpi (Fig. 3). The qRT-PCR results clearly indicate that the transcript levels of TaClpS1 in wheat were induced during Pst infection, suggesting that TaClpS1 participates in the interaction between wheat and Pst.

TaClpS1 is a negative regulator of wheat resistance to Pst
To examine whether TaClpS1 is involved in regulating the wheat defense resistance against Pst, the Barley Stripe Mosaic Virus (BSMV)-induced gene silencing (VIGS) strategy was used. Two specific fragments (TaClpS1-1/2as) were designed to specifically silence all three copies of the endogenous TaClpS1 gene (TaClpS1-2A/2B/2D) in Su11 wheat (Additional file 1: Fig. S1). All the wheat leaves inoculated with BSMV: γ (negative control) or BSMV: TaClpS1-1/2as displayed mild chlorotic mosaic symptoms at 12 dpi (Fig. 4a). Subsequently, the fourth leaves of silenced wheat were inoculated with incompatible Pst isolate CYR23. The leaves inoculated with CYR23 displayed hypersensitive response (HR) symptoms in negative controls and BSMV: TaClpS1-1/2as silenced lines (Fig. 4a). qRT-PCR confirmed that the transcript levels of TaClpS1 were significantly reduced in BSMV: TaClpS1-1/2as silenced lines compared with BSMV: γ treated wheat at 0, 24 and 120 hpi (Fig. 4b). Pathogenesis-related (PR) genes, including PR1 and PR2, are generally considered as marker genes in HR and are necessary for resistance of plants to pathogens [21,22]. Then, the transcript levels of TaPR1 and TaPR2 were analyzed in TaClpS1-1/2as silenced lines and BSMV: γ treated wheat inoculated with CYR23 at 0, 24 and 120 hpi. Our qRT-PCR results showed that the transcript levels of TaPR1 (Fig. 4c) and TaPR2 (Fig. 4d) were notably increased in TaClpS1-1/2as silenced lines compared with that in BSMV: γ treated wheat. In addition, the areas of necroses and H 2 O 2 accumulation induced by inoculation with CYR23 in wheat leaves were measured at 24 hpi. As shown in Fig. 5, H 2 O 2 accumulation per infection site (Fig. 5a, b) and the necrotic area (Fig. 5c, d) in TaClpS1-1/2as silenced wheat were obviously greater than that in BSMV: γ treated wheat. Taken together, these results revealed that TaClpS1 stimulates the Pst infection in wheat, which finally increased plant susceptibility during wheat-Pst incompatible interaction.
Silencing TaClpS1 significantly inhibits the growth of Pst In addition to analyze necroses and H 2 O 2 accumulation, histological analysis of mycelial structures of Pst was performed in wheat leaves infected with CYR23 at 24 and 120 hpi (Fig. 6a). The numbers of haustoria (Fig.  6b), hyphal lengths (Fig. 6c) and hyphal infection area (Fig. 6d), which are indicators to assess fungal expansion ability, were strictly reduced in TaClpS1-1/2as silenced wheat compared with that in BSMV: γ treated wheat. These results indicated that silencing TaClpS1 diminished the growth of Pst.

TaClpS1 negatively regulates disease resistance of N. benthamiana to Phytophthora parasitica
To further verify the conclusion that TaClpS1 negatively regulates plant resistance against pathogens, firstly the interaction of the model plant N. benthamiana and oomycete Phytophthora parasitica was examined. In this experiment, A. tumefaciens carrying plasmid pCAMBIA1302: TaClpS1-GFP or pCAMBIA1302: GFP (negative control) were infiltrated, and then P. parasitica mycelial plugs were placed onto the infiltrated N. benthamiana leaves. As expected, compared with the pCAMBIA1302: GFP negative control, lesion diameters of leaves expressing pCAMBIA1302: TaClpS1-GFP were Fig. 3 Analysis of TaClpS1 expression in wheat leaves inoculated with Pst CYR23. Wheat leaves (Su11) inoculated with Pst CYR23 were sampled at various times, representing the different stages of Pst infection. Mock was sampled at 0 hpi. The comparative threshold (2 -ΔΔCT ) approach was used to analyze the relative transcript levels of TaClpS1. Data were normalized with the transcript level of wheat elongation factor, TaEF-1α, and visualized as the fold changes compared to control at 0 hpi. Each data represent the means ± standard errors of three independent biological replicates. Double asterisks denote the significant difference (P < 0.01) from 0 hpi calculated by unpaired two-tailed Student's t-test. Hpi, hours post inoculation significantly larger, demonstrating that ectopic expression of TaClpS1 in N. benthamiana can enhance P. parasitica infection (Fig. 7a, b). In addition, VIGS strategy was utilized to silence TaClpS1 and then inoculated with Pst virulent isolate CYR31. As shown in Fig. 7c, the uredia on TaClpS1 silenced leaves inoculated with CYR31 at 12 hpi were less than that on negative controls. Moreover, qRT-PCR confirmed that the transcript levels of TaClpS1 were significantly reduced in BSMV: TaClpS1-1/2as silenced lines compared with BSMV: γ treated wheat at 24 and 120 hpi (Fig. 7d). Taken together, these results indicate that TaClpS1 negatively regulates disease resistance of plants.

TaClpS1 interacts with TaHEMA1 by yeast two-hybrid assay
In Arabidopsis, GluTR, encoded by gene HEMA1 (AT1G58290), was identified as a candidate substrate of AtClpS1 [17]. To confirm whether glutamyl-tRNA reductase is a substrate of TaClpS1 in wheat, yeast two-hybrid (Y2H) technique was used. Firstly, a The comparative threshold (2 -ΔΔCT ) approach was used to measure the relative transcript levels of TaClpS1, TaPR1 and TaPR2. Data were normalized with the transcription level of TaEF-1α, and visualized as the fold changes compared to control at 0 hpi. BSMV: γ infected wheat leaves were used as a control. Each data point represents the means ± standard deviation of three independent biological replicates. Single asterisks (P < 0.05) and double asterisks (P < 0.01) denote significant differences from BSMV: γ treatment calculated by unpaired two-tailed Student's t-test BLAST search using the protein sequence of Arabidopsis HEMA1 as a query showed that the hexaploid wheat genome contains three homologous sequences of AtHEMA1, which were located on chromosomes 1A, 1B and 1D, respectively. Subsequently, the three homologous genes were named TaHEMA1 based on phylogenetic tree constructed with HEMA1 proteins from various plant species (Additional file 2: Fig.  S2A). Considering that the three copies of TaHEMA1 are highly conserved in amino acid sequence with 98.37% identity (Additional file 2: Fig. S2B), TaHEMA1 in chromosomes 1B was selected as a representative to perform Y2H assay. In Y2H assay, only yeast cell carrying TaHEMA1 and TaClpS1 could grow normally on SD-Leu-Trp-His-Ade medium containing X-α-gal and appeared blue (Additional file 2: Fig. S2C), indicating that TaClpS1 interacts with TaHEMA1. Overall, these results suggest GluTR encoded by TaHEMA1 could be a candidate substrate of TaClpS1 in wheat.

Discussion
In this study, the TaClpS1 was isolated from T. aestivum Suwon11 leaves. TaClpS1 is homologous with ClpS1 proteins from various plants, indicating a high sequence conservation of ClpS1 among different plant species. The numbers of uredia, haustoria, hyphal lengths and hyphal infection area, which are indicators to assess fungal expansion ability, were strictly reduced in TaClpS1-1/2as silenced wheat compared with that in BSMV: γ treated wheat, suggesting that TaClpS1 promotes susceptibility of wheat to Pst. Phytophthora belongs to hemibiotrophic filamentous pathogens, the disease spot size of which in host tissue visually reflects the degree of its infection and suggests the regulation role of host gene to resistance against Phytophthora. For instance, the lesion diameter of catalase-deficient N. benthamiana leaves inoculated with P. capsici was greater than control plants, indicating that catalase positively regulated host resistance against P. capsici [22]. Overexpression of AtRTP5 in N. Each data denote the means ± standard errors of three biological repetitions. Double asterisks (P < 0.01) denote significant differences from BSMV: γ at the same time points calculated by unpaired two-tailed Student's t-test benthamiana leaves resulted in the increase of lesion diameters when inoculated with P. infestans, indicating that AtRTP5 plays a negative role in regulating plant resistance against Phytophthora [23]. Our results showed that, transient overexpression of TaClpS1 in N. benthamiana enhanced infection of P. parasitica (Fig.  7b), suggesting that TaClpS1 promotes susceptibility of N. benthamiana to P. parasitica.
H 2 O 2 accumulation area and the average necrotic area per infection site in TaClpS1-1/2as silenced wheat were significantly increased, implying that disease resistance was enhanced in TaClpS1 silenced wheat plants. PR proteins are generally considered as marker genes in HRs and are necessary for resistance [24,25]. Herein, the transcript levels of TaPR1 and TaPR2 were detected in TaClpS1 silenced wheat plants and negative control infected with the avirulent Pst CYR23. At the 0 hpi, there were no significant differences in the transcript levels of TaPR1 and TaPR2 in TaClpS1 silenced wheat plants and negative control, revealing that the transcript levels of TaPR1 and TaPR2 were not constitutively induced in non-infected TaClpS1 silenced wheat plants. At 24 hpi, the transcript levels of TaPR1 and TaPR2 were significantly induced in TaClpS1 silenced wheat and negative Fig. 6 Silencing of TaClpS1 hinders the expansion of Pst on wheat leaves. a Wheat germ agglutinin (WGA) was used to stain fungal structures in wheat leaves infected with BSMV and Pst. The fungal structures were observed under an autofluorescence microscope. SV, sub-stomatal vesicle; IH, infection hypha; H, haustoria; HMC, haustorial mother cell. b The haustoria number per infection site was counted in BSMV-infected wheat leaves inoculated with CYR23 at 24 hpi. c The length of IH was measured in BSMV-infected leaves inoculated with CYR23 at 24 hpi. Hyphal length is the distance from the intersection of the sub-stomatal vesicle and the hypha to the tip of the hypha, which was measured by DP-BSW (units in μm) program. d The infection area was measured in BSMV-infected leaves inoculated with CYR23 at 120 hpi using DP-BSW program (units in μm 2 ). Results shown here were derived from three biological replicates and each replication included 40 individual infection sites. Each data denote the means ± standard deviation of three biological replicates. Asterisks (P < 0.05) or double asterisks (P < 0.01) indicate significant differences from BSMV: γ treatment at the same time points calculated by unpaired two-tailed Student's t-test controls compared to that at 0 hpi, while the transcript levels of TaPR1 and TaPR2 were remarkably greater than that in controls. Considering the fact that the accumulation of PR proteins following a pathogen attack is closely associated with the accumulation of salicylic acid (SA) [26], we speculate that the increase of TaPR1 and TaPR2 was accompanied by an increase of SA in TaClpS1 silenced wheat inoculated with Pst CYR23. Additionally, SA, as the representative immune signal, is synthesized in Arabidopsis chloroplasts [27,28], in which TaClpS1 was localized. Taken together, the findings suggest that TaClpS1  TaClpS1-silenced plants  inoculated with Pst isolate CYR31. The comparative threshold (2 -ΔΔCT ) approach was used to measure the relative transcript levels of TaClpS1. Data were normalized against the transcription level of TaEF-1α, and visualized as the fold changes compared to control at 0 hpi. BSMV: γ infected wheat leaves were used as control. Each data represent means ± standard deviation of three independent biological replications. Single asterisks (P < 0.05) or double asterisks (P < 0.01) denote significant differences from BSMV: γ treatment calculated by unpaired two-tailed Student's t-test could play a negative role in the SA mediated resistance of wheat to Pst.
Glutamyl-tRNA reductase in Arabidopsis chloroplast has been identified by affinity purification as a candidate substrate of AtClpS1 [17,29]. Based on the result that glutamyl-tRNA reductase TaHEMA1 in wheat was shown to interact with TaClpS1 using yeast two-hybrid technique, we inferred that TaHEMA1 may function as a substrate of TaClpS1 in wheat. Glutamyl-tRNA reductase is a control point for tetrapyrrole synthesis [30], and increasing research efforts have revealed that tetrapyrrole biosynthesis is involved in the defense response. For example, tetrapyrrole is the main source of singlet-oxygen generation, and singlet oxygen most likely mediates various biological responses, such as host immunity to pathogens [31]. Taken together, it is reasonable that TaClpS1 interrupts tetrapyrrole synthesis to negatively regulate the response of wheat to Pst via selecting glutamyl-tRNA reductase for Clp degration. Future work will be performed to test our hypothesis.

Conclusions
This study reports for the first time that cloning, localization analysis, and functional characterization of a ClpS1 homolog from wheat of AtClpS1. Expression of TaClpS1 in wheat was induced during Pst infection. Moreover, silencing TaClpS1 led to a decreased susceptibility of wheat to Pst. In addition, heterologous expression of TaClpS1 in N. benthamiana enhanced the infection of P. parasitica. These results suggest that TaClpS1 negatively regulates the plant resistance to pathogens.

Strains, plant materials and growth
In this study, Pst isolates CYR23 and CYR31 were used to investigate the transcript levels of TaClpS1 and the VIGS assay of TaClpS1 according to the procedure described previously [32]. Fresh Pst urediospores were collected from wheat infected with Pst. P. parasitica strain ZQ-1 used in this study was routinely maintained on 10% V8 juice medium at 25°C in the dark [33].

Plasmid constructs
The full length of TaClpS1 was cloned into T-simple19 vector to generate TaClpS1-T construct from wheat cultivar Suwon11 cDNA with TaClpS1-specific primers TaClpS1-F/R (Additional file 3: Table S1). To create the constructs for examining the subcellular localization of TaClpS1, full-length TaClpS1 and TaClpS1Δ were amplified from the above TaClpS1-T construct and inserted into pCAMBIA1302 or pTF486 vector [36] to generate pCAM-BIA1302: TaClpS1-GFP, pCAMBIA1302: TaClpS1Δ-GFP and pCaMV35S: TaClpS1-GFP, pCaMV35S: TaClpS1Δ-GFP respectively. Primer sequences are reported in Additional file 2: Table S1. For VIGS assay, two approximately 150-bp specific silencing fragments were designed based on the combination of Primer5 and NCBI. Barley stripe mosaic virus (BSMV), is a positive-sense RNA virus with a tripartite genome consisting of RNAs α, β and γ. The two designed fragments were cloned with NotI and PacI restriction sites and inserted into original BSMV: γ vector to prepare recombinant plasmids BSMV: TaClpS1-1as and BSMV: TaClpS1-2as using specific primers shown in Additional file 3: Table S1 [37]. Above all constructs were obtained from the Prof. Zhensheng Kang's Lab (Northwest A&F University, China) [35].

Phylogenetic analysis
For phylogenetic analysis of TaClpS1, ClpS1 proteins from other plants were obtained using the protein sequence of Arabidopsis AtClpS1 (GenBank accession no. NP_564937.1) to blast NCBI databases. For TaHEMA1, the copies and other related sequences were obtained from the Ensemble Plant database. The phylogenetic tree was constructed using maximum-likelihood method in MEGA5 software. DNAMAN v.7.0 software (Lynnon-Biosoft, USA) was used to perform multiple sequence alignments and the conserved ClpS domain was analyzed using Pfam online (http://pfam.xfam.org/).

RNA extraction and analyses of transcript levels
The second leaves of the two-leaf stage wheat seedlings were inoculated with Pst isolate CYR23. After inoculation, three independent wheat leaves were sampled at 0, 12, 24, 48, 72, 96, 120 hpi for extracting RNA. Total RNA was extracted with the Quick RNA isolation Kit (Huayueyang Biotechnology, China, Beijing). About 3 μg of the total extracted RNA was used for reverse transcription to cDNA with RevertAid First Strand cDNA Synthesis Kit. For RNA extraction and reverse transcription in VIGS assay of TaClpS1, the methods were as described above. LightCycler SYBR Green I Master Mix was used for the qRT-PCR assay, and the transcript levels of genes were normalized to the internal control gene TaEF-1α. The primers used in qRT-PCR assay are listed in Additional file 3: Table S1. The statistical significance was evaluated by unpaired two-tailed Student's t-test.

Subcellular localization analysis
To determine the subcellular localization of TaClpS1 in N. benthamiana leaves, A. tumefaciens carrying pCAM-BIA1302: TaClpS1-GFP, pCAMBIA1302: TaClpS1Δ-GFP or pCAMBIA1302: GFP vector at a final OD 600 of 0.5 was infiltrated into N. benthamiana leaves. Vectors pCAM-BIA1302: TaClpS1Δ-GFP and pCAMBIA1302: GFP were used as negative controls. The infiltrated N. benthamiana were maintained in growth rooms at 21-25°C with a 16-h/ 8-h light/dark cycle. At 48 h after agroinfiltration, confocal images were obtained with an Olympus IX83 confocal microscope (Japan) using excitation wavelength of 488 nm and emission wavelength of 520 nm for GFP, and excitation wavelength of 561 nm and emission wavelength of 640 nm for chloroplast autofluorescence, respectively.
For testing the localization of TaClpS1 in wheat cells, Triticum aestivum Suwon11 seedlings were grown in the glasshouse at 25°C for 2-3 weeks. The fusion constructs pCaMV35S: TaClpS1-GFP, pCaMV35S: TaClpS1Δ-GFP and pCaMV35S: GFP were independently transformed into wheat protoplasts by polyethyleneglycol (PEG)-calcium method as described previously [38,39]. The mixtures containing pCaMV35S: TaClpS1-GFP or pCaMV35S: GFP and wheat protoplasts were incubated at 22°C. Images were obtained with an Olympus IX83 confocal microscope (Japan) at 24 h after incubation.

BSMV-mediated gene silencing
Plasmids BSMV: TaClpS1-1as, BSMV: TaClpS1-2as and BSMV: γ were linearized followed by transcribing and capping in vitro using the RiboMAX Large-Scale RNA Production System-T7 and the Ribom7G Cap Analog (both by Promega) according to the manufacturer's instructions. Wheat leaves were inoculated with the capped BSMV transcripts and Pst isolate CYR23 or CYR31 according to the procedure described previously [40]. BSMV: γ was used as the negative control. The wheat leaves infected with CYR23 were sampled at 0, 24, and 120 hpi for estimating the transcript levels of TaClpS1 and TaPR1/2, H 2 O 2 detection, measuring necrotic areas and histological observations in VIGS assay of TaClpS1. The symptoms on the wheat leaves were photographed at 12 d after inoculation with Pst CYR23 and CYR31. These experiments were repeated at least two times.

DAB staining for H 2 O 2 detection, measuring necroses
The wheat leaves inoculated with CYR23 in VIGS assay were sampled and stained in 1 mg/ml 3,3-diaminobenzidine (DAB) solution for 6 h at 16°C under light. After staining, the leaves were clarified in the destaining solution (absolute alcohol: acetic acid glacial, 1:1) for about one week. Then, the decolored wheat leaves were further clarified with chloral hydrate for two weeks. Subsequently, H 2 O 2 accumulation in the transparent leaves was detected with an Olympus BX-51 microscope under bright-field. Alongside detection of H 2 O 2 accumulation, necrotic areas were measured under UV channel. The results were obtained from 40 infection sites. The samples were collected from three independent leaves. The experiments were repeated three times. The statistical significance was evaluated by unpaired two-tailed Student's t-test.

Histological observations of Pst growth
For histological observations of Pst growth, the wheat leaves inoculated with CYR23 in VIGS experiments were destained at 24 and 120 hpi in absolute alcohol: acetic acid glacial, 1:1 for about one week. Then, clarified wheat samples were stained with wheat germ agglutinin (WGA) conjugated to Alexa-488 (Invitrogen, USA) as described previously [41,42]. For each biological replicate, 40 infection sites of each sample from three separate leaves were recorded to assess the number of haustoria, hyphal length and infection area. The statistical significance was evaluated by unpaired two-tailed Student's t-test.

P. parasitica inoculation
N. benthamiana leaves infiltrated with A. tumefaciens carrying pCAMBIA1302: TaClpS1-GFP, pCAM-BIA1302: GFP were detached at 36 h after infiltration and challenged with P. parasitica by placing mycelial plugs (5 mm diam). The inoculated leaves were maintained in a growth room at 25°C in darkness. At 36 hpi, the inoculated leaves were stained with trypan blue as previously described [43]. Stained leaves were photographed and the diameters of the lesion area were measured.

Yeast two-hybrid assay
The recombinant BD-TaHEMA1 vector was constructed by cloning TaHEMA1 full length sequence into pGBKT7 with primers TaHEMA1-BD-F/R, and recombinant AD-TaClpS1 vector was constructed by cloning TaClpS1 full length sequence into pGADT7 with primers TaClpS1-AD-F/R (Additional file 3: Table S1). For interaction assay, BD-TaHEMA1 and AD-TaClpS1 were cotransformed into yeast strain AH109 by the lithium acetate method following Yeast Protocols Handbook (Clontech), and grown on the SD/−Trp-Leu or SD/ −Trp-Leu-His selection medium. Colonies from SD/ −Trp-Leu-His were picked on SD/−Trp-Leu-His-Ade medium containing X-α-gal again for further selection.