Glutathione S-transferase interactions enhance wheat resistance to powdery mildew but not wheat stripe rust

Abstract

Wheat stripe rust and powdery mildew are important worldwide diseases of wheat (Triticum aestivum). The wheat cultivar Xingmin318 (XM318) is resistant to both wheat stripe rust and powdery mildew, which are caused by Puccinia striiformis f. sp. tritici (Pst) and Blumeria graminis f. sp. tritici (Bgt), respectively. To explore the difference between wheat defense response against Pst and Bgt, quantitative proteomic analyses of XM318 inoculated with either Pst or Bgt were performed using tandem mass tags technology. A total of 741 proteins were identified as differentially accumulated proteins (DAPs). Bioinformatics analyses indicated that some functional categories, including antioxidant activity and immune system process, exhibited obvious differences between Pst and Bgt infections. Intriguingly, only 42 DAPs responded to both Pst and Bgt infections. Twelve DAPs were randomly selected for reverse transcription quantitative polymerase chain reaction (RT-qPCR) analysis, and the mRNA expression levels of 11 were consistent with their protein expression. Furthermore, gene silencing using the virus-induced gene silencing system indicated that glutathione S-transferase (TaGSTU6) has an important role in resistance to Bgt but not to Pst. TaGSTU6 interacted with the cystathionine beta-synthase (CBS) domain-containing protein (TaCBSX3) in both Pst and Bgt infections. Knockdown of TaCBSX3 expression only reduced wheat resistance to Bgt infection. Overexpression of TaGSTU6 and TaCBSX3 in Arabidopsis (Arabidopsis thaliana) promoted plant resistance to Pseudomonas syringae pv. Tomato DC3000. Our results indicate that TaGSTU6 interaction with TaCBSX3 only confers wheat resistance to Bgt, suggesting that wheat has different response mechanisms to Pst and Bgt stress.

Proteomics and functional analysis showed that TaGSTU6/TaCBSX3 interaction plays an important role in wheat resistance to Bgt.

Introduction

Wheat (Triticum aestivum) is one of the main human food crops and is planted extensively around the world. Wheat stripe rust and powdery mildew, caused by the biotrophic pathogens, Puccinia striiformis f. sp. tritici (Pst) and Blumeria graminis f. sp. tritici (Bgt), respectively, are two of the most important foliar diseases worldwide. In China, these two diseases cause enormous losses of wheat yield and quality (Wellings, 2011). Reasonable breeding and use of resistant wheat cultivars are the most economical, effective, and environment-friendly methods of controlling stripe rust and powdery mildew (Chen, 2005; Nelson et al., 2018). However, long-term survival pressure and environmental stimuli endow pathogens with superior adaptability (Turrà et al., 2014), which results in high variation in Pst and Bgt isolates (Lei et al., 2017). Resistance genes of wheat also become less effective as pathogens evolve, thus stripe rust and powdery mildew continue to threaten global wheat production and food security (Huang et al., 1997). Therefore, it is of great importance to study and explore resistant genes to stripe rust and powdery mildew in wheat to devise long-term prevention and control of these diseases.

Transcriptome surveys showed that Bgt triggered a higher number of genes and pathways than Pst in the highly resistant wheat line N9134 (Zhang et al., 2014). The Chinese winter wheat cultivar XM318 is highly resistant to all the current prevalent Chinese Bgt and Pst races and pathotypes. However, whether XM318 responds to these two obligate biotrophic fungi by similar resistance mechanisms is unknown. Tandem mass tags (TMTs)-based proteomics analysis has proven to be an effective way to identify differentially expressed proteins between resistant and susceptible wheat cultivars in response to Fusarium pseudograminearum infection (Qiao et al., 2021). Although TMT technology has been used in qualitative and quantitative protein studies in wheat–Pst and wheat–Bgt interactions (Yang et al., 2016; Li et al., 2017), no thorough proteomics studies have been carried out to compare the interactions of wheat–Pst and wheat–Bgt. Therefore, in this study, proteomics analyses were performed to explore the response difference of XM318 to Pst and Bgt. Eventually, the glutathione S-transferase TaGSTU6 aroused our attention because it was downregulated in XM318–Pst but upregulated in XM318–Bgt.

Glutathione S-transferase (GST) represents a group of ubiquitous proteins in plants that comprise various functional proteins.Wang et al. (2019) categorized plant GSTs into eight classes: τ (GSTU), φ (GSTF), λ (GSTL), dehydroascorbate reductase, θ (GSTT), γ-subunit of translation elongation factor (EF1G), ζ (GSTZ), and tetrachlorohydroquinone dehalogenase (Wang et al., 2019). One of the most important functions of GSTs is their ability to inactivate toxic compounds. Soluble GSTs mediate degradation of reduced glutathione or its homologs that form complexes with herbicides (Jablonkai and Hatzios, 1993). Furthermore, many studies indicated that GSTs are involved in secondary metabolism (Mueller et al., 2000), growth and development (Gong et al., 2005), and biotic and abiotic stress response of plants (Ma et al., 2018).

In barley (Hordeum vulgare), GST interacts with the Blumeria effector BEC1054, possibly compromising well-known key players of defense and response to pathogen (Pennington et al., 2016). In addition, TaGSTU61 associates with the WRKY74 regulatory factor to mediate copper tolerance by affecting glutathione accumulation (Li et al., 2021). GST activity in emmer wheat leaves also contributes to decreased oxidative stress and increased plant resistance to herbicide agents (Karpenko et al., 2019). “Green islands” are areas of leaf tissue surrounding successful infection sites of biotrophic pathogens that display delayed senescence in comparison of the rest of the leaf tissue (Panstruga 2003), which enables the pathogen maximum access to nutrients at the early stages of the compatibility interaction (Walters et al., 2008). Interestingly, TaGSTU1 and TaGSTF6 are important for monocarpic senescence and drought stress (Secenji et al., 2009). However, whether TaGSTU6 is involved in green island formation during Pst and Bgt infection remains unclear. Evaluating the implications of TaGSTU6 in wheat resistance to Bgt and Pst will enable us to understand how wheat responds to different fungal infections.

In this study, based on quantitative proteomics analysis and DAP identification between wheat XM318–Pst and XM318–Bgt interactions, we have systematically analyzed the relationships between the defense response of wheat to Pst and Bgt, and have identified differences in protein regulation in these interactions. Furthermore, we have identified TaGSTU6, which is regulated at the protein and mRNA level during Pst and Bgt infections. Knockdown of TaGSTU6 expression reduced wheat resistance to Bgt, but had no obvious effect on Pst resistance. Further analyses showed that TaCBSX3, as a target of TaGSTU6, is a positive regulator of wheat resistance to Bgt. In addition, overexpression of TaGSTU6 or TaCBSX3 in Arabidopsis promoted plant resistance to Pseudomonas. syringae pv. tomato DC3000. The results broaden our understanding of the defense response of the wheat–Pst and wheat–Bgt interactions at the protein level and provide a theoretical basis for resistance breeding and sustainable control of stripe rust and powdery mildew.

Results

Histological observation of the resistance response of XM318 against Bgt and Pst

Through histological observation, we observed that the wheat cultivar XM318 was highly resistant to Pst and immune to Bgt. Full rust disease symptoms developed by 16-day postinoculation (dpi) in the susceptible control MX169, but XM318 displayed only slight necrotic flecks (Figure 1A). For Bgt infections, the leaves of XM318 failed to develop obvious signs of infection and exhibited remarkable powdery mildew resistance at 12 dpi compared to the abundant surface mycelia in the susceptible control JS16 (Figure 1A).

Figure 1.

Phenotypic observation and histological analysis of wheat cultivars XM318 inoculated with Pst or Bgt. A, Wheat cultivars XM318 and MX169 infected with Pst at 16 dpi, and XM318 and JS16 infected with Bgt at 12 dpi. Wheat MX169 and JS16 were the susceptible controls. B–M, The growth of fungi and accumulation of H₂O₂ in wheat leaves were observed via microscopy. Wheat leaves inoculated with Pst or Bgt were sampled at 24, 48, and 72 hpi. The samples were stained with WGA (B–M) and DAB (C2, E2, G2, I2, K2, and M2), respectively, and observed under a fluorescence microscope. Scale bars = 20 µm. N and O, Statistical analysis of necrotic cells and ROS accumulation in wheat leaves infected by Pst or Bgt. Vertical coordinates indicate the percentage of sites producing H₂O₂ or HR among all infested sites, respectively. Error bars represent ± se of three independent biological replicates. AP, appressorium; C, conidia; AGT, appressorium germ tube; PGT, primary germ tube; SV, substomatal vesicle; IH, infection hypha; NHC, necrotic host cell; H, haustorium or haustoria.

To determine the key time points of immune response of XM318 to the two obligate biotrophic parasitic fungi, the infection processes of Pst and Bgt were investigated histologically at 12, 24, 36, 48, 60, 72, 84, and 96-h postinoculation (hpi). Our results showed that the hyphal length of Pst was significantly (P < 0.05) shorter in XM318 than in the susceptible control MX169 at 24 hpi, indicating that XM318 developed resistance at 24 hpi (Figure 1, B and C1; Supplemental Table S1). At 48 hpi, XM318 showed a pronounced inhibition of mycelial extension with significantly reduced hypha length and area of infection in comparison with MX169, and the resistance was more pronounced at 72 h (Figure 1, D–G). Furthermore, the numbers of haustorial mother cells (HMCs) and hyphae branches increased more slowly after 48 hpi in the XM318 plant cells than in MX169 (Supplemental Figure S1; Supplemental Table S1).

For Bgt infection, although haustorial center bodies (HCBs) were observed at 24 hpi in both XM318 and the susceptible control JS16 (Figure 1, H and I1), the area of infection in XM318 was significantly smaller than that of susceptible control JS16 since 12 hpi (Supplemental Table S2), indicating that XM318 developed resistance at 12 hpi. At 48 hpi in the JS16 control cultivar, the number of mature haustoria and mycelial extension with the number of hyphal branches, hyphal length, and area of infection were significantly greater than that in the XM318 leaves (Figure 1, J and K1; Supplemental Table S2). Interestingly, in contrast to the susceptible control, the hyphae almost stopped growing, and the colonies about to abort as hypersensitive reactions appeared after 24 hpi in XM318 (Figure 1, K1 and M1; Supplemental Figure S2; Supplemental Table S2). In total, the onset of resistance in both XM318–Pst and XM318–Bgt appears to occur at 12–48 hpi, during which the growth of mycelia of Pst and Bgt was restricted by XM318. Therefore, the 12, 24, and 48 hpi time points merited attention in subsequent experiments.

XM318 responded differently to Bgt and Pst infection

To elucidate the immune response kinetics of XM318 to Pst or Bgt infection, the percentages of infection sites with hydrogen peroxide (H₂O₂) accumulation or with necrosis were measured. H₂O₂ accumulation could be detected microscopically in the leaves infected with Pst or Bgt at 24 hpi (Figure 1, C2, I2, and N), but the development of H₂O₂ in XM318–Bgt was more rapid than in XM318–Pst at 36 hpi. The percentage of infection sites with H₂O₂ accumulation reached a peak at 48 and 60 hpi for XM318–Bgt (82%) and XM318–Pst (88%) interactions, respectively, after which their prevalence was maintained until 72 hpi in both interactions before they began to decline (Figure 1N). In addition, hypersensitive necrosis of mesophyll cells were observed at 36 hpi in leaves infected with Pst or Bgt (Figure 1O;Supplemental Figures S1F2 and S2F2). Moreover, the percentage of infection sites with hypersensitive necrosis increased rapidly from 36 to 60 hpi and the rate of necrosis site development in XM318–Bgt was higher than that of XM318–Pst (Figure 1O). For the two susceptible cultivars, a much lower percentage of necrosis sites were observed upon Pst or Bgt infection (Figure 1O). In total, these results indicate that the response kinetics of XM318 to Bgt infection is different from its kinetics to Pst infection.

Overview of TMT data

To characterize the response difference between XM318 to Pst and XM318 to Bgt at the protein level, proteomics analysis were performed on wheat leaves infected with Pst or Bgt at 24 and 48 hpi, respectively. A total of 834,046 MS² mass spectra were obtained (Supplemental Table S3). MS² mass spectra were analyzed using the MASCOT software to obtain an MS² mass spectrum score for each, which showed that the distribution of the MS² mascot score conformed to normal distribution (Supplemental Figure S3A). The raw data were filtered and searched using the MASCOT engine and Proteome Discoverer to characterized, and the 171,645 total mass spectra were matched to a number of peptides. Finally, a total of 36,854 matched peptides, 22,159 matched unique peptides, and 7,311 proteins were identified. The qualitative proteins were 7,311, and the quantified proteins were 5,485 (Supplemental Data Set 1). In addition, ∼90% of the identified proteins had a molecular weight distribution between 10 and 110 kDa, which followed the normal distribution (Supplemental Figue S3B). A total of 741 proteins were significantly different between treatments (Pst or Bgt infected at 24 or/and 48 hpi) and identified as differentially accumulated proteins (DAPs). Moreover, all identified proteins and DAPs were investigated by heat map analysis, which provides a clear visual perception for a large amount of complex protein data (Figure 2, A and B).

Figure 2.

Analysis of DAPs. A, The profiles of all proteins identified by TMT in wheat infected by Pst or Bgt at indicated time points. B, Profile of all DAPs in wheat induced by Pst or Bgt. The blue color indicated upregulated protein and green color indicated downregulated protein. The color scales range from −3 to 3, and the value refers to log₂(fold change). Pst24hpi:CK indicates the protein levels in wheat infected with Pst at 24 hpi compared with the mock water inoculated. Bgt24hpi:CK indicates the protein levels in wheat infected with Bgt at 24 hpi compared with the mock inoculated. Pst24 hpi:Bgt24 hpi indicated the differences in protein levels between wheat infection with Pst and Bgt at 24 hpi. C, The overlap among DAPs identified by TMT analysis. Pst_-horizontal and Bgt_-horizontal represent DAPs identified from wheat interacted with Pst and Bgt, respectively. D, The overlap among HRPs identified by TMT analysis between different interactions at different infection time. HRPs means vertical or Pst24 hpi:Bgt24 hpi or Pst48 hpi:Bgt48 hpi.

Bioinformatics analysis of DAPs

All DAPs were derived from two categories: horizontal regulated proteins (HRPs: DAPs from comparisons of proteomic data from leaves infected with the same pathogen but collected at different time points) and vertical regulated proteins (VRPs: DAPs of comparative proteomic data from different pathogen-infected leaves at the same time points) (Figure 2C). There were 213 HRPs from Pst infections, among which 118 and 114 were upregulated at 24 and 48 hpi, respectively, but only 21 and 41 were downregulated at 24 and 48 hpi, respectively (Figure 2D;Supplemental Data Set 2). In contrast, there were more HRPs downregulated than upregulated from Bgt infections (0 h versus Bgt24h, 0 h versus Bgt48h) (Figure 2D). Also, the majority of the 345 HRPs caused by Bgt was from samples collected at 48 hpi, 124 upregulated and 212 downregulated (Figure 2D;Supplemental Data Set 3; Supplemental Figure S4). Therefore, the results of proteomic analysis further confirmed that XM318 responds differently to Pst and Bgt infections. Intriguingly, although there were 544 VRPs (Pst24h versus Bgt24h; Pst48h versus Bgt48h) (Supplemental Data Set 4), only 42 proteins (Supplemental Table S4; Supplemental Data Set 5) were regulated in both Pst and Bgt infections, which led to further investigation.

Gene ontology (GO) and OmicShare GO (OSGO) enrichment analyses showed that the HRPs were mainly categorized into cellular components and basal biological processes (Supplemental Figure S5, A and B). OSGO analyses displayed that the number of DAPs belong to antioxidant activity and the immune system process were substantially higher in the XM318–Bgt than in the XM318–Pst interaction, consistent with the observation that the accumulation of H₂O₂ was more rapid during XM318–Bgt than XM318–Pst, and that XM318 is immune to Bgt but highly resistant to Pst. Although the number of DAPs belong to response to stimulus are similar between XM318–Pst and XM318–Bgt interactions, protein kinases and hormonal response proteins included in response to stimulus categories were substantially different between XM318–Pst and XM318–Bgt at both 24 and 48 hpi (Supplemental Table S5). For example, the kinases, mitogen-activated protein kinase 5 (MAPK5), and calcium-dependent protein kinase 2 (CDPK2) were upregulated in XM318–Bgt but not in XM318–Pst interaction, and abscisic acid (ABA)- and salicylic acid (SA)-related proteins were more likely to be induced by Pst, and SA-, jasmonic acid (JA)-, and ABA-related proteins were mainly expressed in response to Bgt. The results indicated XM318 responded differentially to Pst and Bgt infections.

The Clusters of Orthologous Groups (COG) database was used to classify all DAPs (VRPs and HRPs). The main functional categories were involved in posttranslational modification (O, 90, 13.12%), carbohydrate metabolism and transport (G, 73, 10.64%), and other functional classification (Supplemental Figure S6; Supplemental Data Set 6). In summary, the COG analysis showed that the responses of XM318 to pathogen challenges comprise complex regulatory mechanisms.

To better understand the metabolism pathways involved in XM318–Pst and XM318–Bgt, plant visualization pathways and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed. The visualization pathways showed that glucose metabolism, lipid metabolism, and amino acid metabolism were main pathways in both interactions, but there are big differences in the way of protein regulation between two interactions (Supplemental Figure S7). KEGG analysis showed that lipid metabolism and plant–pathogen interactions were activated in XM318 upon Pst infection, suggesting their vital roles during XM318–Pst interaction. In the XM318–Bgt interaction, DAPs were significantly enriched in plant–pathogen interactions and MAPK signaling pathway, indicating their important roles (Supplemental Figure S8). These data show an association in the XM318 response to the Pst and Bgt pathogens, suggesting that differences exist in the main metabolic pathways.

Gene expression analysis of selected DAP genes by RT-qPCR

Reverse transcription quantitative polymerase chain reaction (RT-qPCR) analyses were performed to verify whether gene transcription profiles were consistent with the proteomic profiles of XM318 upon Pst or Bgt infections. We selected 12 genes encoding DAPs with significant differential patterns induced by Pst and/or Bgt. Among these DAPs, six proteins were regulated by both Pst and Bgt infections and the other six accumulated differently in Pst or Bgt infections (Supplemental Table S6). The transcriptional levels of 11 genes were consistent with their protein levels from proteomic data for at least one sampling time point (Supplemental Figure S9). However, the expression patterns of TaGSTU6 (UniProt database, A0A1D5RWE5) were consistent between the mRNA and protein levels in XM318–Bgt but inconsistent in XM318–Pst. In the proteomic data, TaGSTU6 expression was downregulated in Pst infection but upregulated in Bgt infection (Supplemental Table S6), whereas its transcript levels were induced by both Pst and Bgt (Supplemental Figure S9C). The discrepancy between the protein and transcript levels of TaGSTU6 after Pst infection may be due to degradation of its mRNA by an unknown mechanism. In conclusion, the transcriptional and proteomic analysis generally yielded similar results.

Identification of TaGSTU6

Among the above-mentioned 42 proteins regulated in both Pst and Bgt infections (Supplemental Table S4; Supplemental Data Set 5), the ribulose bisphosphate carboxylase/oxygenase activase A (A0A1D6S090), 14-3-3 protein (G5DFC5), 12-oxophytodienoate reductase 11 (A0A1D6AU24), and TaGSTU6 were downregulated in Pst infection but upregulated in Bgt infection, among which TaGSTU6 has the highest expression. Thus, TaGSTU6 was selected for further study. The transcription of TaGSTU6 was induced by Bgt and Pst in both compatible and incompatible interactions with lower levels in the compatible combinations (Figure 3, A and B; Supplemental Table S7), suggesting that TaGSTU6 may function in the resistant response of XM318 to Pst and Bgt infections. BLAST analysis on the hexaploid wheat genome databases revealed three TaGSTU6 copies located on chromosomes 1A, 1B, and 1D, respectively. Phylogenetic analysis showed that TaGSTU6 is closely related to HvGSTU6 (barley, H. vulgare; 92.27%) and OsGSTU6 (rice, Oryza sativa; 70.39%) (Supplemental Figure S10A). Sequence analyses indicated that TaGSTU6 contains a typical GST C-terminal domain and an N-terminal domain that bears a transmembrane helix (Supplemental Figure S10B).

Figure 3.

TaGSTU6 expression profiles and functional assessment by VIGS. A and B, The expression profile of TaGSTU6 during wheat–Bgt (A) or wheat–Pst (B) sampled at the indicated time points. C, Phenotypes of wheat leaves inoculated with BSMV:TaPDS, FES buffer (mock), BSMV:γ, and BSMV:TaGSTU6 at 10 dpi. D, Phenotypes of the fourth leaves inoculated with Bgt at 12 dpi. Plants were preinoculated with FES buffer (mock), BSMV:γ, and BSMV:TaGSTU6. E, Images of wheat leaves preinoculated with FES buffer (mock), BSMV virus, and BSMV:TaGSTU6 and challenged with Pst at 16 dpi. F, Relative transcript levels of TaGSTU6 in silenced plants inoculated with Bgt at 0, 24, 48, and 96 hpi. G, The relative expression levels of TaGSTU6 in silenced leaves infected with Pst. Leaves were collected at 24, 48, 96, and 120 hpi. Wheat leaves inoculated with BSMV:γ and sampled after inoculation with Bgt or Pst were used as controls and TaGSTU6 transcript level in the control group at every sampled time point was standardized as 1. All expression levels were normalized to TaEF-1a. Transcript level of the gene was calculated by the 2^−ΔΔCT method and differences were calculated by t tests. Mean and standard deviation were calculated with data from three independent biological replicates. Asterisks indicate significant differences at the same time points using Student’s t test (*P < 0.05 and **P < 0.01).

Silencing of TaGSTU6 attenuates wheat resistance to Bgt

To evaluate the function of TaGSTU6 during wheat–Pst or wheat–Bgt interaction, the barley stripe mosaic virus (BSMV)-virus-induced gene silencing (VIGS) system was used to silence TaGSTU6 in XM318. At 10 days postBSMV inoculation, BSMV:TaPDS plants exhibited a bleaching phenotype, and mild chlorotic mosaic symptoms were observed in almost all BSMV-inoculated plants (Figure 3C), indicating that the BSMV–VIGS system was functional and that virus inoculations were successful and stable.

The XM318 plants inoculated with BSMV:TaGSTU6 (TaGSTU6-1as and TaGSTU6-2as) were greatly reduced in their resistance to Bgt in comparison with the control (XM318 inoculated with BSMV:γ) (Figure 3D), whereas inoculation with BSMV:TaGSTU6 had no obvious effects on the resistance of XM318 to Pst (Figure 3E;Supplemental Figure S11). Comparing the transcript levels between the BSMV:TaGSTU6-inoculated leaves with the control leaves, the TaGSTU6 transcripts were reduced by 62%–75% for TaGSTU6-1as and by 67%–76% for TaGSTU6-2as (Figure 3, F and G), indicating that the reduced resistance of XM318 to Bgt could be due to the silence of TaGSTU6. Wheat germ agglutinin (WGA) staining and confocal microscopy observation revealed that, although no substantial hyphal growth differences were observed between the TaGSTU6-silenced and control plants at 24 hpi (Figure 4, A and B), Bgt progressed much faster in the TaGSTU6-silenced plants than in the control at 48, 96, and 12 dpi (Figure 4, C–H). The number of hyphal branches/infection, HMC, hyphal length, and total infection areas were significantly higher in the silenced leaves than in control leaves at 48 and 96 hpi (Figure 4, I–L). These results indicated that TaGSTU6 plays an important role in XM318 resistance to Bgt after 48 hpi. To further verify its function in wheat resistance to Bgt, TaGSTU6 was also silenced in the susceptible wheat–Bgt compatible interaction. Aggravated disease phenotypes and larger colony morphologies were observed in TaGSTU6-knockdown plants at 12 dpi (Supplemental Figures S12 and S13). In conclusion, TaGSTU6 plays a positive role in wheat resistance to Bgt, but has no obvious disease-resistance functions during wheat–Pst infection.

Figure 4.

Epifluorescence observations of fungal growth in the XM318 wheat cultivar inoculated with BSMV and infected with Bgt (incompatible combination). A–H, Fungal growth was examined under an epifluorescence microscope after staining with WGA (A1–H1), and accumulation of ROS were observed after staining with DAB (A2–H2) at 24, 48, 96 hpi and 12 dpi. The same letter indicates the photos were taken at the same infection site. I–L, The number of hyphal branches (I), haustoria (J), hyphal length (K), and total infection area (L) in BSMV:γ or BSMV:TaGSTU6 leaves were measured using DP-BSW software at 24, 48, and 96 hpi. All results were obtained from 50 infection sites. Data are the means ± se of three independent samples. Differences were assessed using Student’s t tests. *P < 0.05; **P < 0.01. A–G, scale bars = 20 µm; (H1and H2) scale bars = 50 µm.

TaGSTU6 interacts with the CBS domain-containing protein TaCBSX3

To identify the interacting proteins of TaGSTU6 during XM318–Bgt interaction, the cDNA of TaGSTU6 was used as the bait to screen yeast two-hybrid (Y2H) library constructed with RNA isolated from Bgt-infected XM318 leaves. A total of 11 clones representing 8 XM318 proteins were screened out as potential TaGSTU6-interacting proteins (Supplemental Table S8), among which the cystathionine β-synthase (CBS) domain-containing protein TaCBSX3 (XM_020317884.1) displayed the strongest interaction and therefore was selected for subsequent studies. Sequence analyses revealed that TaCBSX3 contains two conservative CBS domains with three copies in the wheat genome that are located on chromosomes 4A, 4B, and 4D, respectively (the sequence identity of the three copies of TaCBSX3 gene is 97%, 99%, and 100% when compared to the TaCBSX3 in chromosome 4D). Furthermore, when the TaGSTU6-GFP and TaCBSX3-mCherry constructs were coinfiltrated into Nicotiana benthamiana leaves, strong green and red fluorescence were co-localized in the plasma membrane and cytoplasm (Figure 5A), suggesting the possible interaction of TaGSTU6-GFP and TaCBSX3-mCherry.

Figure 5.

TaGSTU6 interacts with TaCBSX3. A, Co-localization analysis of TaGSTU6 and TaCBSX3 in N. benthamiana. Leaves of N. benthamiana were infiltrated with Agrobacterium strains containing the TaCBSX3-mCherry and TaGSTU6-GFP construct pairs. The infiltrated N. benthamiana leaves were observed via fluorescence microscopy. B, Y2H analysis of the interaction between TaGSTU6 and TaCBSX3. Yeast cells of Y2H Gold strain transformed with the bait and prey vectors were assayed for growth on the selective medium SD/–Leu–Trp–His–Ade/X-α-gal/AbA. Yeast co-expressing TaGSTU6 and TaCBSX3 grew on the selective medium and had X-α-gal activity. Yeast cells containing pGADT7-T and pGBKT7-53 or pGBKT7-Lam vectors were used as positive and negative controls. C, A Co-IP assay was performed to detect the TaGSTU6–TaCBSX3 interaction in plant cells. Western blots of total proteins from N. benthamiana leaves transiently expressing the constructs carrying TaGSTU6-GFP and TaCBSX3-mCherry, and proteins eluted from GFP-Trap agarose beads were detected with anti-GFP or anti-mCherry antibodies. D, Confirmation of the interaction between TaGSTU6 and TaCBSX3 by LCI assay in N. benthamiana. N. benthamiana leaves were infiltrated with the Agrobacterium strain containing the indicated constructs. Signals were collected 48 h after infiltration. E, Interaction of TaGSTU6 and TaCBSX3 in planta using the fluorescence complementation assay. Leaves of N. benthamiana were infiltrated with Agrobacterium strains containing the following construct pairs, CE+TaCBSX3−NE, TaGSTU6−CE+NE, TaGSTU6−CE+TaCBSX3−NE, and TaCBSX3−CE+TaGSTU6−NE. The infiltrated N. benthamiana leaves were observed via fluorescence microscopy under yellow light excitation (excitation wavelength 490–500 nm) or red light excitation (excitation wavelength 550–580 nm). AtH2B, a histone protein, was used as a nuclear localization marker gene. Scale bars = 20 µm.

We further confirmed the TaGSTU6–TaCBSX3 interaction with Y2H, co-immunoprecipitation (Co-IP), firefly luciferase complementation imaging (LCI), and bimolecular fluorescence complementation (BiFC) assays. Only yeast cells expressing both TaGSTU6 and TaCBSX3 grew on the selective medium SD/–Leu–Trp–His–Ade/X-α-gal/AbA and exhibited beta-galactosidase activity (Figure 5B;Supplemental Figure S14). In Co-IP assays with N. benthamiana leaves infiltrated with the TaGSTU6-GFP and TaCBSX3-mCherry fusion constructs, bands representing TaGSTU6-GFP and TaCBSX3-mCherry were present in the western blots of the total proteins and the proteins eluted from anti-GFP beads (Figure 5C), indicating TaGSTU6 interacting with TaCBSX3 in vivo. In LCI assays, strong luciferase activity was observed with the nLuc–TaGSTU6 and cLuc–TaCBSX3 combinations, providing further evidence of the TaGSTU6–TaCBSX3 interactions in N. benthamiana (Figure 5D). In addition, BiFC assays were performed in N. benthamiana leaves to detect TaGSTU6–TaCBSX3 interactions. Strong yellow fluorescence signals were observed in cells co-transformed with nEYFP-TaGSTU6 and cEYFP–TaCBSX3 (Figure 5E). In summary, the results of Y2H, Co-IP, LCI, and BiFC provided strong evidence for a TaGSTU6–TaCBSX3 interaction.

TaCBSX3 contributes to wheat resistance to Bgt

To determine whether TaCBSX3 is involved in wheat resistance to Bgt and Pst, we first analyzed the expression patterns of TaCBSX3 in the incompatible and compatible combinations between Bgt/Pst and wheat by RT-qPCR. The transcripts of TaCBSX3 were upregulated in the incompatible Bgt–XM318 interaction, but downregulated in the compatible combination (Figure 6A;Supplemental Table S7), suggesting that TaCBSX3 may play a role in the resistance of XM318 to Bgt. Similar expression patterns but to a lower extent were observed in wheat–Pst interactions (Figure 6B), indicating that TaCBSX3 may not be involved in the resistance of XM318 to Pst.

Figure 6.

TaCBSX3 expression profiles and functional analysis by VIGS. A and B, The expression profile of TaCBSX3 were analyzed when the host leaves were inoculated with Bgt and Pst respectively. C, Phenotypes of wheat leaves inoculated with BSMV:TaPDS, FES buffer (mock), BSMV:γ, and BSMV:TaCBSX3 at 10 dpi. D, Bgt conidium on the fourth leaves examined at 12 dpi. The arrows indicated conidia of powdery mildew. E, Relative transcript levels of TaCBSX3 in silenced plants inoculated with Bgt at 0, 24, 48, and 96 hpi. F, The phenotype of mock leaves and leaves preinoculated with BSMV virus and all challenged with Pst at 16 dpi. G, The relative expression levels of TaCBSX3 in silenced leaves infected with Pst at 24, 48, 96, and 120 hpi. Wheat leaves inoculated with BSMV: γ and sampled after inoculation with Bgt or Pst were used as controls and TaCBSX3 transcript level in the control group at every sampled time point was standardized as 1. Expression levels were normalized to TaEF-1a. Data are the means ± se of three independent samples. Asterisks indicate significant differences at the same time points using Student’s t test (**P < 0.01).

Subsequently, the expression of TaCBSX3 was knocked down by BSMV–VIGS in the incompatible XM318–Bgt combination, and 10 days after virus inoculation, the plants displayed chlorotic mosaic symptoms (Figure 6C). After inoculation with Bgt, sporadic single spore foci appeared on leaves of the TaCBSX3-silenced plants at 12 dpi, whereas conidium failed to appear on the control and BSMV:γ unsilenced control plants (Figure 6D). The silencing efficiency confirmed that endogenous TaCBSX3 expression was reduced in the silenced plants by ∼50%–86% compared with the control plants (Figure 6E). However, after inoculation with Pst, BSMV:TaCBSX3-inoculated XM318 failed to exhibit obvious phenotypic differences compared with BSMV:γ-inoculated control plants (Figure 6, F and G; Supplemental Figure S15). Therefore, both the expression profile of TaCBSX3 and the results of BSMV–VIGS experiment suggested that TaCBSX3 played an important role in wheat resistance to Bgt but not to Pst.

Although no substantial histological differences between the TaCBSX3-knockdown and control plants were observed microscopically at 24 hpi (Figure 7, A–D), the resistance of the TaCBSX3-silenced plants to Bgt was reduced at 48, 72, and 96 hpi, in comparison with that of the control plants (Figure 7, E–P), indicating that TaCBSX3 functions at 48 hpi. Statistical analyses also confirmed that the number of Bgt hyphal branches, hyphal length, and the total infected areas remarkably increased in the BSMV:TaCBSX3-inoculated leaves at 48, 72, and 96 hpi (Figure 7, Q–S). Furthermore, the necrotic areas of the Bgt-infected leaves of the TaCBSX3-silenced plants were significantly smaller than the BSMV:γ-inoculated plants after 48 hpi (Figure 7T), suggesting that TaCBSX3 is important for the resistance of XM318 to Bgt. In order to verify the disease resistance effects of TaCBSX3, VIGS assays were performed to silence TaCBSX3 in the wheat–Bgt compatible interaction, and the BSMV:TaCBSX3-inoculated leaves also appeared to be more susceptible to Bgt (Supplemental Figures S16 and S17). Together, our results indicated that TaCBSX3 positively regulates wheat resistance to Bgt.

Figure 7.

Histological observations of Bgt growth in XM318 preinoculated with BSMV. A–P, Fungal structures of leaves were observed under an epifluorescence microscope after staining with WGA conjugated to Alexa-488. Leaves inoculated with BSMV:γ were not inoculated with Bgt (A, E, I, and M). Inoculation with BSMV:γ or BSMV:TaCBSX3-1/2as and Bgt at 24 hpi (B–D), 48 hpi (F–H), 72 hpi (J–L), and 96 hpi (N–P). Q–T, The number of hyphal branches per infection unit (Q), hyphal length (R), total infected area (S), necrosis area (T) in BSMV-infected plants inoculated with Bgt was calculated using SPSS software. All results were obtained from 50 infection sites. Data are the means ± se of three independent samples. Differences were assessed using Student’s t tests. *P < 0.05; **P < 0.01. B–D, scale bars = 20 µm; (A and E–P) scale bars = 50 µm. SH, secondary hyphae; S, stoma.

SA accumulation is important for TaGSTU6 and TaCBSX3-mediated resistance to Bgt infection

Plant hormones have been demonstrated to be critical to immune signaling processes (Vlot et al., 2009). Many hormone-related proteins in XM318 showed differential expression upon infection by Pst and Bgt in proteomic analysis. Therefore, we speculated that the endogenous hormone levels in XM318 varied between XM318–Pst and XM318–Bgt. To test this hypothesis, the endogenous levels of ABA, JA, SA, and the ethylene precursor 1-aminocyclopropane-1-carboxylic acid (ACC) in XM318 in response to Pst or Bgt were measured. The concentrations of ABA and SA in XM318 leaves were significantly upregulated by Pst infection at 8 and 12 hpi compared with the control (Figure 8A). However, SA was increased significantly in Bgt-infected leaves at 12 hpi, and ACC was increased significantly at both 8 and 12 hpi (Figure 8B). These results suggest that SA accumulation is important for wheat resistance against both Pst and Bgt infections.

Figure 8.

Expression profiles of TaGSTU6 and TaCBSX3 in response to hormones and the content of endogenous hormones in leaves in response to Bgt and Pst infections. A and B, Analysis of endogenous hormone content in wheat leaves inoculated with Pst or Bgt. JA, jasmonate. C and D, RT-qPCR analysis of TaGSTU6 and TaCBSX3 expression after treated with hormones exogenously. The wheat seedlings treated with 0.1% (v/v) water were used as a mock control. Expression levels were normalized to TaEF-1a. ETH, ethylene. E and F, Hormones accumulation in the nonsilenced and TaGSTU6- or TaCBSX3-silenced leaves after inoculated with Bgt. Data are the means ± se of three independent samples. Differences were assessed using Student’s t tests. *P < 0.05; **P < 0.01. FW, fresh weight.

To determine whether expression levels were regulated by plant hormones, the transcription profiles of TaGSTU6 and TaCBSX3 in response to exogenous hormones were analyzed in resistant and susceptible cultivars by RT-qPCR, respectively. The transcript levels of TaGSTU6 significantly increased in wheat leaves after SA or ethylene（ETH） treatments, and also responded positively to JA and ABA treatment (Figure 8C;Supplemental Figure S18A). The expression of TaCBSX3 also increased after SA and ETH treatment, and the abundance of the TaCBSX3 transcripts decreased after JA treatment (Figure 8D;Supplemental Figure S18B). Therefore, the transcription of TaGSTU6 and TaCBSX3 is regulated by exogenous applied hormones.

To further explore the involvement of hormones in TaGSTU6- and TaCBSX3-related resistance to Bgt infection, hormone accumulation in TaGSTU6- and TaCBSX3-silenced XM318 leaves upon Bgt infection was measured. Surprisingly, only SA accumulation was significantly reduced in the TaGSTU6- or TaCBSX3-silenced leaves at 12 hpi, whereas the other hormones remained relatively stable or were minimally impacted (Figure 8, E and F). Together, these results imply that SA has an important role in TaGSTU6- and TaCBSX3-mediated resistance to Bgt infection.

Overexpression of TaGSTU6 or TaCBSX3 enhances Arabidopsis resistance to P. syringae pv. tomato (P. syringae) DC3000

To further verify the function of TaGSTU6 in plant resistance to pathogens, TaGSTU6 overexpression constructs were generated and introduced into the Arabidopsis ecotype Columbia-0 (Col-0; wild-type [WT]). BLASTp analysis showed that the AtGSTU17 (Arabidopsis thaliana, AY091102.1) is homologous to TaGSTU6. Therefore, the AtGSTU17 mutants were ordered from Arabidopsis Biological Research Center and named SK88, and the homozygous atgstu17 mutants were obtained by screening subcultures from atgstu17-F1 (SK88). Then, the TaGSTU6 overexpression lines (TaGSTU6-OE), AtGSTU17 deletion mutant atgstu17, and WT Col-0 plants were challenged with P. syringae DC3000. Obvious chlorotic symptoms were observed on the leaves of two atgstu17 lines (atgstu17-9 and atgstu17-13) but not on the WT leaves at 2 dpi, indicating AtGSTU17 plays a role in the resistance of Arabidopsis to P. syringae DC3000. In contrast, chlorotic symptoms were rarely observed on the leaves of the three TaGSTU6-OE lines (TaGSTU6-OE6/13/20) (Figure 9A;Supplemental Figure S19A), suggesting that overexpression of TaGSTU6 increased the resistance of Arabidopsis to P. syringae DC3000. RT-qPCR analysis revealed that the transcript levels of TaGSTU6 in TaGSTU6-OE20 were the most elevated, followed by TaGSTU6-OE13 and TaGSTU6-OE6, which is consistent with their levels of resistance (Figure 9B), suggesting a positive association between Arabidopsis resistance and TaGSTU6 expression levels. Furthermore, pathogenic bacterial growth was significantly inhibited in the TaGSTU6-OE lines compared with the atgstu17 and WT plants (Figure 9C). All these results strongly indicated that TaGSTU6 increased the resistance of Arabidopsis to P. syringae DC3000.

Figure 9.

Overexpression of TaGSTU6 and TaCBSX3 promote Arabidopsis resistance to P. syringae pv. tomato DC3000. A and D, Bacterial defense phenotypes of TaCBSX3-OE, TaGSTU6-OE, atgstu17 mutant, and WT at 2 dpi with P. syringae DC3000. TaGSTU6 overexpression lines were more resistant than wild and atgstu17 mutant type. TaCBSX3 overexpression lines were more resistant than WT. B and E, Relative transcript levels of TaGSTU6 and TaCBSX3 in TaGSTU6-OE and TaCBSX3-OE lines respectively by RT-qPCR. C and F, Bacterial titer (log10) of TaCBSX3-OE, TaGSTU6-OE, and atgstu17 mutant at 2 dpi with P. syringae DC3000 (OD₆₀₀ = 0.01). Bacterial growth is expressed as mean values of viable bacteria per gram of leaf tissue ± sd. Error bars indicate SD. Data are shown as mean ± sd (n = 6, where n indicates biological replication). Differences were assessed using Student’s t tests. *P < 0.05; **P < 0.01.

We then examined whether overexpression of TaCBSX3 also increases the resistance of Arabidopsis to P. syringae DC3000. Compared with WT lines, chlorotic symptoms were obviously reduced on leaves of the three TaCBSX3-OE lines (Figure 9D;Supplemental Figure S19B). Moreover, the anti-bacterial effects were consistent with the expression level of the TaCBSX3-OE lines (Figure 9E). As shown in Figure 9F, pathogenic bacterial growth was also significantly inhibited in the TaCBSX3-OE lines compared with WT plants. In conclusion, overexpression of TaGSTU6 or TaCBSX3 enhanced disease resistance to P. syringae DC3000 in Arabidopsis.

Discussion

Stripe rust and powdery mildew are important worldwide diseases of wheat. Transcriptome surveys demonstrated that different pathways and genes were activated in wheat leaves upon Pst and Bgt infection (Zhang et al. 2014). In this study, we have extended this work by performing quantitative proteomics to analyze the mechanisms of wheat resistance interactions with Pst and Bgt. Our results show that proteins in different functional categories are activated in response to Pst and Bgt, and that there are obvious differences and association between the DAPs in Pst and Bgt infection.

Different protein kinases are involved in wheat responses to Pst and Bgt

MAPKs, a group of highly conserved proteins, are key signaling enzymes involved in the regulation of various plant processes, including metabolism and abiotic and biotic defense signaling processes (Xu and Zhang, 2015; Liu et al., 2017). Accumulation of MPK3 and MPK6 is critical for stimulating full induction of plant defense during induced resistance (Gawroński et al., 2014). TaMAPK4 was induced during the wheat–Pst incompatible interaction and was regulated by miR164, which indicated that MAPK signaling plays an important role in wheat defense against the rust pathogen (Wang et al., 2018). Li et al. (2017) found that MAPK5 transcripts were reduced in Bgt-infected susceptible wheat leaves. Therefore, they postulated that MAPK5 may have a positive role in wheat resistant to powdery mildew. In this study, consistent with the results of Li et al. (2017), MAPK5 protein level was significantly increased in the incompatible wheat–Bgt interaction, but not in the incompatible wheat–Pst interaction, suggesting that MAPK5 plays a role in wheat resistance to Bgt but not to Pst.

Wheat CDPKs are crucial sensors of calcium concentration changes that occur in response to various biotic and abiotic stressors, including H₂O₂, salt, drought, and Bgt (Li et al., 2008). Freymark et al. (2007) found that at least nine CDPK putative paralogs are expressed during the early phases of powdery mildew infection of barley. Similarly, in our study, CDPK2 expression was upregulated in the wheat–Bgt interaction. Therefore, we speculate that CDPK2 has important roles in wheat resistance to Bgt. In addition, other classes of protein kinases were identified in the wheat interactions with Pst and Bgt (Supplemental Table S5), but no protein kinase was found to be affected in both Pst and Bgt infections, suggesting that wheat utilizes different kinases in response to Pst and Bgt infection.

Pathogenesis-related proteins highlight differences in wheat resistance to Pst and Bgt

Pathogenesis-related (PR) proteins are types of proteins produced by plants under various biological circumstances and are closely related to the hypersensitive response and systemic acquired resistance (Van Loon and Van Strien, 1999). To date, PRs have been divided into 17 families (Sels et al., 2008), and different types of PRs have different and extremely important roles in various plant stress responses. In this study, we have identified many types of disease-resistant proteins, including PR-1, PR-2, and PR-3 in the wheat–Bgt interaction, and PR-1, PR-2, and Wrab17 in the wheat–Pst interaction. These results suggest that wheat has complex disease resistance systems that elicit differences in responses to the Pst and Bgt pathogens.

Plant lipid transfer proteins (LTPs, PR-14) also take part in broad-spectrum resistance to pathogens. Several studies have found that particular plant LTPs have various functions in inhibiting the growth of different pathogens or enhancing host resistance to various pathogens (Molina et al., 1993; García-Olmedo et al., 1995; Isaac Kirubakaran et al., 2008). TaDIR1-2, a recently characterized wheat putative ortholog of LTP, has been confirmed as a negative regulator in wheat resistance to Pst by modulating reactive oxygen species (ROS) (Ahmed et al., 2017). In this study, some LTPs were differentially expressed in wheat during Pst and Bgt infections, and more LTPs were downregulated in wheat inoculated with Bgt than Pst (Supplemental Table S5). Therefore, different LTPs may be involved in the resistance response to Pst and Bgt. In addition, the wheat cultivar XM318 used in this study is resistant to both Bgt and Pst, and Bgt development in wheat is more rapid than Pst. Therefore, there is a possibility that the identified differential expression of PR proteins may be a consequence of differential pathogen infection stages, which requires more intensive proteomics monitoring at different infection time points.

Hormonal responses differ in wheat–Pst and wheat–Bgt interactions

Plant hormones play important roles in plant defense responses to a variety of pathogens. SA has been demonstrated to be critical to immunity signaling processes and elevated SA can protect plants from a wide range of pathogens (Vlot et al., 2009). SA and its analogs also induce systemic acquired resistance and expression of PR proteins in plants (Kohler et al., 2002). ABA regulates abiotic stress tolerance and negatively regulate plant disease resistance pathways (Asselbergh et al., 2008). The ET and JA pathways, also known as the JA/ET signaling pathway, are essential for induced systemic resistance of plants (Pieterse and Van Loon, 1999), and TaWRKY62 positively regulates wheat high-temperature seedling resistance to Pst by regulating SA-, ET-, and JA-mediated signaling (Wang et al., 2017). These signaling pathways interact with each other through positive and negative networks, and ultimately, plants appear to use the most reasonable defense response combinations against specific pathogens (Flors et al., 2008).

In this study, many hormone-related proteins were identified in wheat–Pst and wheat–Bgt interactions (Supplemental Table S5), but few were simultaneously induced in both interactions. Among these hormone-related proteins, ABA- and SA-related proteins were activated in response to Pst, and SA-, JA-, and ABA-related proteins were mainly expressed in Bgt responses. Huai et al. (2019) found that Pst infection resulted in increased ABA accumulation, and our determination of endogenous hormone contents by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) further confirmed it. In wheat–Pst interaction, the ABA and SA contents in leaves were significantly upregulated. In contrast, the SA and ACC levels increased significantly in Bgt-infected wheat leaves, but the JA concentrations were downregulated (Figure 8, C and D). These results show that hormonal responses are regulated differently in wheat–Pst and wheat–Bgt interactions. Further analyses indicated that SA accumulation decreased significantly in TaGSTU6- or TaCBSX3-silenced leaves, whereas other hormones were almost unchanged or had minor changes after inoculation with Bgt (Figure 8, E and F). Therefore, SA plays an important role in TaGSTU6 and TaCBSX3 conferring resistance to Bgt infection. Taken together, the ABA or SA pathway or both appear to be involved in wheat defenses against Pst, whereas single SA, JA, or ACC pathways or a combination of these pathways participate in defense responses against Bgt.

TaGSTU6–TaCBSX3 interaction promotes wheat resistance to Bgt

GSTs constitute an ancient, rich class of multifunctional proteases and are encoded by a multi-gene family. A genome-wide comprehensive analysis has identified 346 GST genes and 87 TaGSTU members in common wheat (Wang et al., 2019; Hao et al., 2021). RT-qPCR assays show that the TaGSTU6, TaGSTU4, and TaGSTU7 are substantially upregulated in the incompatible wheat–Bgt interaction (Supplemental Figure S20). However, the transcripts of TaGSTU4 and TaGSTU7 were lower than TaGSTU6 during Bgt infection.

Previous research indicated that Blumeria effector BEC1054 interacted with barley GST to confer plant defense with ROS accumulation (Pennington et al., 2016). Li et al. (2017) reported that TaGSTU61 was involved in oxidative burst in response to Bgt infection, suggesting that the high levels of antioxidants may protect wheat from oxidative damage. Therefore, we hypothesized that TaGSTU6 is involved in wheat resistance to Bgt in an ROS-dependent manner.

In this study, mass spectrometry and RT-qPCR results showed that TaGSTU6 is significantly regulated in both Pst and Bgt infections. However, the VIGS assay showed that TaGSTU6 acts as a resistance-associated protein in the wheat response to Bgt infection, but not to Pst. In summary, TaGSTU6 has different roles in wheat resistance to Pst and Bgt. There are six CBS domain-containing proteins predicted in the Arabidopsis genome, and these contain only a single CBS pair, but no protein domains were designated as CBSX (Ok et al., 2012). CBSXs are speculated to be sensor relay proteins that function in regulation of many enzymes (Seok et al., 2020). CBSX1 and CBSX2 directly regulate activation of TRXs as redox regulators in chloroplasts. Overexpression of CBSX1 and CBSX2 reduces ROS levels, and subsequently affects the expression of cell wall thickening genes (Yoo et al., 2011; Jung et al., 2013). Interactions of CBSX3 with Thioredoxins (Trx-o2) provide a ROS mitochondrial regulator in Arabidopsis, and ROS accumulation was shown to increase markedly during CBSX3-overexpression (Seok et al., 2020). The interaction of the transcription factors TaNAC069 and TaCBSX3 may have positive regulatory roles in wheat resistance to P. triticina (Zhang et al., 2021). Mou et al. (2015) determined that OsCBSX3 acts as a positive regulator in rice resistance to Magnaporthe oryzae by working synergistically through SA- and JA-mediated signaling pathways. OsCBSX4 was developmentally regulated by sense alterations in ionic and energy balance (Singh et al., 2012). Therefore, TaCBSX3 may directly activate Trx and regulate the production of ROS.

Our results indicated that TaCBSX3 associates with TaGSTU6 during wheat–Bgt incompatible combinations. Additional BMSV–VIGS and transgenic Arabidopsis inoculation assays showed that TaCBSX3 can positively regulate resistance to Bgt and to the bacterial P. syringae DC3000 pathogens, respectively. Therefore, we speculate that wheat senses Bgt signals after infection, and that TaGSTU6 may interact with effector proteins secreted by Bgt. After infection, TaGSTU6 interacts with TaCBSX3 to pass signals into mitochondria, and TaCBSX3 associates with Trx to activate downstream genes that regulate ROS production and inhibit Bgt growth. The accumulation of ROS and SA are important signaling factors in plant disease resistance responses, and are involved in regulating and inducing the expression of a series of disease resistance and systemic acquired resistance genes (Liu, 2020). Increased SA levels were detected in Arabidopsis plants with persistent H₂O₂ production by peroxisomes (Oa et al., 2010). SA-mediated defense against P. syringae was impaired in Arabidopsis plants with insufficient peroxidase production ROS (Mammarella et al., 2015). Thus, TaGSTU6 interacting with TaCBSX3 is involved in the regulation of ROS concurrently, so that the plant produces a disease resistance response while the plant accumulates SA signaling molecules at the pathogenic infestation site, which induces the plant to produce SAR. As a regulator of Bgt resistance, TaGSTU6 may participate in numerous ways to inhibit Bgt infection. Verification of these activities requires further studies of disease resistance mechanisms.

Conclusions

This study reveals that wheat differentially responses to the obligate biotrophic parasitic fungi Pst and Bgt, and that differences exist in hormonal responses and protein kinase activities. We also identified TaGSTU6 as conferring wheat resistance to Bgt but not to Pst. Furthermore, TaGSTU6 interacting with TaCBSX3 positively regulates wheat resistance to Bgt, possibly through the SA signaling pathways. Our results provide insights into the wheat resistance mechanisms and further resources for disease-resistance breeding and generating transgenic resistance wheat lines.

Materials and methods

Biological materials, fungal inoculation, and chemical treatments

The Chinese winter wheat (T. aestivum) line XM318, developed by Shaanxi Xingmin Seed Company, shows high resistance to all current Chinese predominant Pst and Bgt races and pathotypes. The Pst isolate CYR32 and the Bgt isolate E09, which are predominant races in China (Wan et al., 2007; Xia et al., 2020), were used in this study. CYR32 and E09 were maintained and propagated on the susceptible Chinese wheat cultivars Mingxian169 (MX169) and Jingshuang16 (JS16), respectively. Nicotiana benthamiana and Arabidopsis (A. thaliana) (Col-0 background) were used in this study, and plant cultivation was performed as previously described (Huai et al., 2019).

For inoculation, XM318, MX169, and JS16 seeds were planted at the same time under identical and fungal-free conditions. When the first seedling leaves had fully expanded, half of the XM318 seedlings were brushed with Pst, and the other half of the XM318 seedlings were inoculated with Bgt. MX169 and JS16 seedlings were inoculated with Pst and Bgt, respectively, as susceptible controls. Leaves were collected at 12, 24, 36, 48, 60, 72, 84, and 96 hpi for histological observations and RT-qPCR analyses. In addition, the leaves were sampled at 0, 24, and 48 hpi for proteomics.

For chemical treatments, XM318 (incompatible) and JS16 (compatible) seedlings were treated with SA, ET, JA and ABA, as previously described (Wang et al., 2020). Leaves were sampled at 0, 0.5, 2, 6, 12, and 24-h posttreatment. For endogenous hormone accumulation, TaGSTU6-silenced and TaCBSX3-silenced leaves inoculated with Pst or Bgt (incompatible combination) were sampled at 8 and 12 hpi for HPLC–MS/MS analysis. For gene silencing assays, the wheat cultivars XM318 and “Suwon 11” BSMV-silenced leaves were inoculated with Pst or Bgt as incompatible and compatible combinations, respectively.

All of the sampled leaves were immediately frozen in liquid nitrogen and stored at −80°C. Three biological replicates were performed for each treatment.

Histological staining

For histological observations, sampled leaves were stained with 3,3'-diaminobenzidine hydrochloride (MP Biomedicals, USA; Xiao et al., 2003) and WGA (Invitrogen, Waltham, MA, USA) as described previously (Ayliffe et al., 2011). Necrotic areas were detected by auto-fluorescence of mesophyll cells in infected leaves by epi-fluorescence microscopy. For each treatment, 50 random infection sites were examined microscopically with an Olympus BX-53 microscope (Olympus, Tokyo, Japan). ROS areas generated by H₂O₂, hyphal branches, hyphal length, and areas of infection were observed and measured.

Total protein extraction and TMT labeling

For proteomics, total proteins were extracted from samples using the trichloroacetic acid/acetone method as previously described (Zhang et al., 2016). For each sample, 200 μg of proteins were digested with 4-μg solid trypsin according to a filter-aided sample preparation procedure (Wiśniewski et al., 2009). Then peptide mixtures of each sample were labeled using the TMTsixplex reagent according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). Samples were labeled with the TMT tags, and three biological repeats had three sets of TMT tags as detailed in Supplemental Table S9.

LC–MS/MS analysis and proteomic data analysis

The Pierce High pH Reversed-Phase Fractionation Kit (Thermo Scientific) was used to separate the TMT-labeled digested samples into 15 fractions. LC–MS/MS analysis of each fraction was performed with a Q Exactive mass spectrometer (Thermo Scientific) as described by Zhang et al. (2017). Raw MS/MS data were searched using the MASCOT engine (Matrix Science, London, UK; version 2.2) embedded into Proteome Discoverer version 1.4 (Thermo Scientific). After the raw data (.raw) were converted to a .mgf file, it was then run against the protein database: uniprot_Triticum_aestivum_146090_20170302.fasta (Total number of sequences: 146090, download link: http://www.uniprot.org). The following parameters were set: the peptide FDR was set to ≤0.01; the variable modification choice was oxidation (M) and TMT 6plex (Y); for fixed modification, Carbamidomethyl (C), TMT 6plex (N-term), and TMT 6plex (K) were selected. Other search parameters were set according to Guo et al. (2020). For protein quantification, the mock inoculation sample at 0 h was used as a reference. The quantitative protein ratios were weighted and normalized by the median ratio in Mascot. DAPs were analyzed according to the ratios of the abundance of the proteins identified and a one-sample t test. Proteins with P-values (t test) <0.05 and fold change ratios ≥1.3 or ≤0.77 were considered to be significantly upregulated or downregulated, respectively.

Bioinformatics analysis

Functional annotations of the identified proteins were conducted using the Blast2GO program against the UniProtKB database. Level 2 GO analyses of multi-group data were performed with OmicShare tools, a free online platform for data analysis (http://www.omicshare.com/tools). The expression levels of all DAPs were converted into log2 fold changes. Then, R language was used to perform heatmap analyses of all identified proteins and DAPs. DAPs were classified into different functional categories with the COG database (http://eggnogdb.embl.de/#/app/emapper?jobname=MM_FnYQKJ). Analysis of plant visualization pathways was performed using MapMan software. KAAS (KEGG Automatic Annotation Server) was used to retrieve their Kos, which were subsequently mapped to KEGG pathways, and TBtools software was used to draw the KEGG enrichment chart.

Total RNA extraction and RT-qPCR

With reference to the Plant RNA Reagent instructions (TransGen Biotech, Beijing, China), total RNAs were isolated and reverse transcribed into cDNA following the manufacturer’s directions (Thermo, USA), and genomic DNA contaminants were removed by DNase I treatment. Specific primers (Supplemental Table S10) were designed using Primer Premier version 5.0 software to assess the transcriptional expression levels of selected protein genes by RT-qPCR. The wheat elongation factor TaEF-1α was chosen as the internal control gene. All RT-qPCR was performed according to the manufacturer’s instructions (TransGen Biotech, China). Cycle threshold values were generated using the QuantStudio 5 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) to quantify relative gene expressions by use of the comparative 2^−ΔΔCt method (Livak and Schmittgen, 2001). All PCR analyses were repeated 3 times.

Sequence analysis of TaGSTU6 and TaCBSX3

TaGSTU6 and TaCBSX3 alleles from chromosomes were obtained from the wheat UGRI genome database (https://urgi.versailles.inra.fr/). The National Center for Biotechnology (NCBI) conserved domain database (https://www.ncbi.nlm.nih.gov/cdd/) was used to identify domains. Phylogenetic trees were constructed with MEGA version 7.0 software. Physicochemical properties of proteins were predicted using the “ProtParam” tool of “ExPASy” (http://www.expasy.org).

BSMV-mediated gene silencing

The TaGSTU6 and TaCBSX3 genes were silenced using the BSMV–VIGS system. Two pairs of special gene fragments (TaCBSX3: 255 and 362 bp; TaGSTU6: 199 bp and 282 bp) were generated and cloned into BSMV as previously described (Holzberg et al., 2002). The infectious BSMV RNAs were transcribed in vitro from linearized plasmids γ-TaPDS, γ-TaCBS, γ-TaGSTU6, γ, α, and β (Petty et al., 1989) using the RiboMAX Large Scale RNA Production System-T7 kit (Promega, Madison, WI, USA) and the Ribo m7G Cap Analog (Promega). Each group of mixed BSMV RNA was inoculated onto the second leaf of wheat seedlings at the two-leaf stage as described previously (Scofield et al., 2005). The seedlings were maintained in a growth chamber at 25°C and examined for symptoms. BSMV:TaPDS acted as the positive control, BSMV:γ was used as the viral control, and mock plants were inoculated with FES (viral inoculation buffer) as the negative control. At 10–12 dpi with BSMV, at the four-leaf stage of seedlings, the third and fourth leaves of each plant were inoculated with Pst or Bgt, and leaves were sampled at 0, 24, 48, 96, and 120 hpi to determine the silencing efficiencies and histological observations. The stripe rust and powdery mildew phenotypes were evaluated at 16 and 12 dpi. All primers used are listed in Supplemental Table S10, and the experiments were repeated 3 times.

Y2H assay

The cDNA libraries for the Y2H assay were constructed using the Gateway system (Invitrogen, USA) by OE Biotech (China). The cDNAs were constructed from total RNAs extracted from Bgt -infected XM318 leaves. The pGADT7 (AD)-cDNA library vector was used as the prey. The bait vectors were generated by inserting TaGSTU6 DNA fragments into pGBKT7 (BD) plasmids. The bait and prey vectors were co-transformed into yeast strain Y2H Gold according to the Yeast Protocols Handbook (Clontech), diluted on SD/–Leu–Trp/AbA for selection of positive colonies, which were transferred to SD/–Leu–Trp–His–Ade/X-α-Gal/AbA for further selection. To confirm TaGSTU6 interaction targets, full-length cDNAs of the candidate target of interest were cloned into pGADT7 vectors and transformed into the yeast strain Y2H Gold, containing the bait vector, and grown on SD/–Leu–Trp–His–Ade/X-α–Gal/AbA.

BiFC assay

For the BiFC assays, TaGSTU6 cDNAs were cloned into pSPYNE to make nEYFP-TaGSTU6 constructs, and the cDNA from TaCBSX3 was ligated into pSPYCE to form cEYFP-TaCBSX3. The constructed plasmids were transformed into Agrobacterium tumefaciens GV3101. The Agrobacterium strains were infiltrated into the leaves of 4-week-old N. benthamiana in different combinations. Two days after inoculation, images were observed at 490–500-nm wavelength laser via Olympus BX-53 microscope (Japan). The H2B-mcherry fusion protein was used as a nuclear marker.

Firefly LCI assays

For LCI assays, TaGSTU6 and TaCBSX3 were cloned into pCAMBIA1300-NLuc and pCAMBIA1300-CLuc to confirm TaGSTU6 interactions with TaCBSX3. Primers used for vector construction are shown in Supplemental Table S10. LCI assays were performed in N. benthamiana leaves as previously described (Chen et al., 2008). Three independent replicates were performed.

Co-IP and western blot analysis

For preparation of Co-IP assays, protein samples, and colocalization of TaGSTU6 and TaCBSX3, full-length TaGSTU6, and TaCBSX3 cDNAs were amplified to construct pCAMBIA1302-TaGSTU6-GFP and pYBA1132-TaCBSX3-mCherry plasmids. The primers used are listed in Supplemental Table S10. Agrobacterium strains containing TaGSTU6-GFP and TaCBSX3-mCherry plasmids were transformed into N. benthamiana leaves at an optical density (OD₆₀₀) of 0.5. At 48 h after agroinfiltration, fluorescence signals were observed with an Olympus BX-53 microscope (Olympus, Japan), and N. benthamiana total proteins were extracted with a native lysis buffer (Solarbio).

For Co-IP assays, 15 μL of GFP-Trap agarose beads (Chromotek, Germany) were incubated with 2 mL of crude protein at 4°C for 3 h and collected as described previously (Yang et al., 2019).

For western blot analyses, sodium dodecyl sulfate polyacrylamide gel electropheresis (SDS–PAGE), electroblotting, and immunodetection analysis were carried out as described previously (Yang et al., 2020). Precipitated proteins were detected with an anti-GFP antibody (#A02020; Abbkine), an anti-mCherry antibody (#A02080; Abbkine), and a goat anti-mouse secondary antibody (ab6789; Abcam, Cambridge, UK). Protein bands were visualized with a western ECL substrate kit (Bio-Rad, Hercules, CA, USA).

Arabidopsis transformation and inoculation

The recombinant plasmid pCAMBIA2300-TaGSTU6 and pCAMBIA2300-TaCBSX3 were constructed, transferred into A. tumefaciens strain EHA105, and subsequently transformed into Arabidopsis ecotype Col-0 as described (Zhang et al., 2006). More than 30 independent transgenic lines were screened based on resistance to the antibiotics G418 and kanamycin. The increased transcript levels of TaGSTU6 and TaCBSX3 were verified by RT-qPCR. Fully expanded leaves of 4-week-old transgenic Arabidopsis were infected with suspensions of P. syringae DC3000 cells (OD₆₀₀ = 0.01) in 10-mM MgCl_2. Bacterial growth in plant leaves was determined at 2 dpi. Inoculation and conidiophore counting methods were as described by Xiao et al. (2011).

Accession numbers

All mass spectrometry proteomics data in this study have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the Integrated Proteome Resources database partner repository with the dataset identifier PXD032182. Sequence data from this article can be found through the NCBI database (http://www.ncbi.nlm.nih.gov/) with the accession number TaGSTU6 (XM_044472374.1), TaCBSX3 (XM_020317884.1), and AtGSTU17 (AY091102.1).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. Histological changes in Pst growth, H₂O₂ accumulation, and necrotic cell areas in wheat XM318 and MX169 cultivars.

Supplemental Figure S2. Histological changes in Bgt growth, H₂O₂ accumulation, and necrotic cell areas in wheat XM318 and JS16 cultivars.

Supplemental Figure S3. The abscissa shows scoring of the MASCOT peptide fragment and the relative molecular weight of the identified protein.

Supplemental Figure S4. Classification of DAPs.

Supplemental Figure S5. GO functional annotation of DAPs.

Supplemental Figure S6. Functional analysis of DAPs.

Supplemental Figure S7. Analysis of pathways of wheat interactions with Pst or Bgt at different infection times.

Supplemental Figure S8. KEGG pathway enrichment analysis of DAPs.

Supplemental Figure S9. Relative expressions of twelve selected DAPs genes in Wheat–Pst or -Bgt (compatible and incompatible interactions) at the indicated time points.

Supplemental Figure S10. TaGSTU6 identification.

Supplemental Figure S11. Fungal growth rates in incompatible interactions of TaGSTU6-knockdown plants.

Supplemental Figure S12. Functional assessment of TaGST the roles of TaGSTU6 in compatible wheat–Bgt interactions.

Supplemental Figure S13. Epifluorescence observations of Bgt growth in wheat in compatible interactions after inoculation with BSMV.

Supplemental Figure S14. Interactions of TaGSTU6 with TaCBSX3 in Y2H assays.

Supplemental Figure S15. Fungal growth rates in incompatible interactions of TaCBSX3-knockdown plants.

Supplemental Figure S16. Functional assessment of the role of TaCBSX3 determined by BSMV-mediated gene silencing in compatible wheat–Bgt interactions.

Supplemental Figure S17. Epifluorescence observations of Bgt compatible growth in susceptible wheat after BSMV inoculation.

Supplemental Figure S18. TaGSTU6 and TaCBSX3 induction after exogenous hormone treatments.

Supplemental Figure S19. The phenotypes of TaCBSX3-OE and TaGSTU6-OE transformed Arabidopsis lines after P. syringae DC3000 infection.

Supplemental Figure S20. Expression profiles of TaGSTUs.

Supplemental Table S1. Numbers of HMC, hyphal branches, linear lengths, and infection foci in Pst-infected wheat.

Supplemental Table S2. Numbers of haustoria, hyphal branches, linear length, and areas of infection per infection site in the Bgt–wheat interaction.

Supplemental Table S3. Results of protein identification statistics.

Supplemental Table S4. DAPs regulated during Pst and Bgt infections.

Supplemental Table S5. Differentially expressed protein groups in wheat infected with Pst or Bgt.

Supplemental Table S6. Regulation of multiple DAPs selected for RT-qPCR.

Supplemental Table S7. TaGSTU6 and TaCBSX3 transcription profiles and protein expression level in response to Bgt and Pst infection.

Supplemental Table S8. Description of potential TaGSTU6 targeted by the Y2H system.

Supplemental Table S9. Sample tag information in TMT analysis.

Supplemental Table S10. Primers used for PCR and plasmid construction.

Supplemental Data Set 1. All identified proteins in the wheat–Pst and wheat–Bgt interaction.

Supplemental Data Set 2. Details of differentially expressed proteins in the wheat–Pst interaction.

Supplemental Data Set 3. Details of differentially expressed proteins in the wheat–Bgt interaction.

Supplemental Data Set 4. Details of differentially expressed proteins by comparing wheat–Pst and wheat–Bgt interactions at 24 and 48 hpi.

Supplemental Data Set 5. Details of differentially expressed proteins in wheat interactions with both Pst and Bgt.

Supplemental Data Set 6. Lists of proteins in different COG categories.

 

Q.W., Q.L., J.G., and B.T.W. conceived and designed the research plans. Q.W., P.F.J., and M.Y.G. performed the experiments. Q.W. and J.G. analyzed the data. Q.W., P.C., and Q.L. completed the writing.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is Baotong Wang (wangbt@nwsuaf.edu.cn) and Qiang Li (qiangli@nwsuaf.edu.cn).