Molecular characterisation of the broad-spectrum resistance to powdery mildew conferred by the Stpk-V gene from the wild species Haynaldia villosa

Plant material and fungal isolate

Yangmai158 is a powdery mildew susceptible wheat variety that is preserved in the Cytogenetics Institute of Nanjing Agricultural University (CINAU). T₅, the Stpk-V over-expressed transgenic wheat with high powdery mildew resistance, was produced by gene-gun bombardment in CINAU using cv. Yangmai158 as recipient and is a descendant of the T3-2 line used in Cao et al. (2011). Mixed isolates of Bgt were collected from Jiangpu Experimental Station, CINAU, and maintained on seedlings of susceptible wheat variety Sumai 3 in a spore-proof greenhouse. The mixed population of Bgh was collected from fields in the eastern provinces of China and maintained in the highly susceptible barley cultivar Hua30.

Secale cereale (2n = 14, RR), Aegilops longissima (2n = 14, S^lS^l), Hordeum vulgare (2n = 14, HH), Ae. geniculata (2n = 28, U^gU^gM^gM^g), Ae. peregrina (2n = 28, SSUU) and a series of alien chromosome addition lines were used. These addition lines used were: wheat–S. cereale DA1R, DA2R, DA3R, DA4R, DA6R and DA7R; wheat–Ae. longissima DA1S, DA2S, DA3S, DA4S, DA5S and DA6S and DA7S; wheat–H. vulgare DA2H, DA3H, DA4H, DA5H, DA6H and DA7H; wheat–Ae. peregrinum DA1U, DA2U, DA3U, DA4U, DA5U, DA6U and DA7U; wheat–E. elongatum DA1E, DA2E, DA3E, DA4E, DA5E, DA6E and DA7E; and wheat–Ae. geniculata DA1M^g, DA2M^g, DA3M^g, DA4M^g, DA5M^g, DA6M^g and DA7M^g in the background of wheat cv. Chinese Spring from the Wheat Genetic and Genomic Resources Center, Kansas State University, USA.

Sequencing of DGE with Illumina

Susceptible Yangmai158 and resistant T₅ seedlings were inoculated with Bgt at the two-leaf stage, then leaves from about 20 seedlings were harvested at 0, 6 and 24 hai for RNA extraction using the Trizol reagent (Invitrogen, Waltham, MA, USA). Each sample of 6 μg of total mRNA was purified with Oligo (dT) magnetic bead adsorption. A DGE library was built and fixed onto the Illumina Sequencing Chip (flowcell) for sequencing. Six DGE libraries were constructed to study gene expression profiles in the Pm resistant Stpk-V transgenic T₅ line and the susceptible recipient line Yangmai158. The libraries included non-inoculated Yangmai158 and T₅ (designated Y-0 and T-0), Bgt-inoculated Yangmai158 and T₅ at 6 and 24 hai (designated Y-6, Y-24, T-6 and T-24). Approximately 3.6 million raw tags were generated for each of the six samples, and after removing low quality reads, each library contained ~3.5 million clean reads. The number of mapped clean tags was calculated for each library and then mapped to the wheat unigene database. Then we performed cluster analysis of gene expression patterns with ‘cluster’ software (Eisen et al. 1998) and ‘Java Treeview’ software (Saldanha 2004) to construct the Hierarchical clustering (HCL) tree (using the Pearson correlation method with average linkage). To identify genes associated with resistance mediated by Stpk-V, powdery mildew-responsive genes in Bgt infected T₅ and Yangmai158 were compared at 6 and 24 hai (DEG tag analysis data in Table S1).

The SYBR green real-time RT-PCR assay

All selected genes were used for SYBR green real-time RT-PCR, primers are listed in Table S2. A real-time RT-PCR reaction (20 μl) included 20 ng cDNA, 0.2 μm of each primer, 1× SYBR Premix ExTaq (TaKaRa, Tokyo, Japan). Reactions were performed on a Bio-Rad IQ single-colour Real-Time PCR detection System (Bio-Rad, Hercules, CA, USA) under the following conditions: 94 °C for 30 s, 40 cycles of 94 °C for 5 s, 60 °C for 15 s and 72 °C for 20 s to calculate cycle threshold (Ct) values, followed by 95 °C for 15 s, 60 °C for 1 min, then 95 °C for 15 s to obtain melt curves to ensure primer specificity.

Histochemical detection of fungal structures or host cells

Yangmai158 and T₅ seedlings were treated with exogenous chemicals at the two-leaf stage: 10 mm dimethylthiourea (DMTU, H₂O₂ scavenger) and 7 mm H₂O₂ administered as a 0.05% Tween-20 solution, with 0.05% Tween-20 used as mock treatment (Garretón et al. 2002). The chemically treated leaves were then inoculated with Bgt and the first leaves harvested at the indicated time points. H₂O₂ production in the Bgt interaction host cells was detected using the DAB staining method (Thordal-Christensen et al. 1997). Cell death in the Bgt interaction sites were detected by in situ histochemical staining (Peterhänsel et al. 1997). After DAB or Trypan blue staining, the leaves were bleached in boiling 95% ethanol for 10 min then stored in 50% glycerol. Before observation, the stored leaves were stained in Coomassie blue [0.6% (w/v) Coomassie Brilliant Blue R 250 in methanol] for 30 s, rinsed in distilled water twice, and mounted on glass slides in 50% glycerol for microscopy (Olympus, Tokyo, Japan). For each time point, four leaf pieces and at least 200 infection sites on each leaf piece were observed. Standard deviations and t-test statistical analyses were performed using the SAS software and bars represent ±SD.

Homologous analysis

For the evolutionary study of Stpks in the grass family, all PCRs were performed using the primers Stpk-F (AGATCCAACACCAGTTCAAG) and Stpk-R (ATGTTAAGGAGGCTTGTGTC), covering the fifth and sixth exon, and amplification was performed at 94 °C for 3 min; 30 cycles of 94 °C for 30 s, 55–58 °C for 30 s, and 72 °C for 2 min; followed by 10 min at 72 °C.

Production and screening of Arabidopsis transgenic plants

To obtain the transgenic plasmid, the pAHC25 plasmid was digested with EcoRI and HindIII, then the transcript unit Ubi promoter::GUS::NOS was maintained and inserted into the pCAMBIA1301 digested with the same two restriction endonucleases. The Stpk-V gene was amplified with the forward SmaI-containing primer (5′-TCCCCCGGGATGTCAATTTCCAACATA-3′) and the reverse SacI-containing primer (5′-CGATCGCTCACTCACGCTCTGATATTGC-3′) from the clone vector p18MD-T- Stpk-V plasmid as template, then the PCR product and the plasmid pCAMBIA1301-AHC were digested with SmaI and SacI enzymes, following by a ligation reaction to construct the recombinant vector pCAMBIA1301-AHC-Stpk-V. The recombinant plasmid was introduced into Agrobacterium tumefaciens C58C1 for Stpk-V transformation in Arabidopsis as described previously (Xiang et al. 1999). The hygromycin B resistant plants were harvested and used to reproduce the T₂ generation. Positive plants of the T₂ generation were screened for hygromycin B resistance, X-gluc staining and target gene analysis. PCR and RT-PCR were performed to analyse the presence and expression level of Stpk-V transgenic Arabidopsis using the gene-specific primers Stpk-F (5′-GATCCAACACCAGTTCAAG-3′) and Stpk-R (5′-ATGTTATGGAGGCTTGTGTC-3′). The expression level of the Tubulin gene was used as control with the primers Tubulin-F: 5′-CTCATCACAGGCAAGGAAGAT-3′ and Tubulin-R: 5′-CAGAACCTCAGGGCAACG-3′. PCR and RT-PCR were then used to verify positive plants for the powdery mildew evaluation.

Single-cell transient expression assay

The single-cell transient expression assay was performed according to Shirasu et al. (1999). Reporter plasmids containing β-glucuronidase (GUS) genes and the plasmids pAHC:Stpk-V were mixed before coating the particles (molar ratio of 1:1; 1 μg total DNA). The Bgh susceptible cv. Hua 30 was used for the transient leaf assay. Bombarded barley leaves were transferred to 1% agar plates supplemented with 85 μm benzimidazole and incubated at 18 °C for 8 h before high-density inoculation with single Bgh spores. Leaves were stained for GUS to identify Stpk-V transformed cells at 48 h after spore inoculation. The haustorium index (percentage GUS-staining cells with haustoria in the total GUS-staining cells attacked by Bgh) is indicated from four independent experiments, each of at least 70 interactions.

Proposed resistance mechanism of Stpk-V from analysis of DEGs in T₅ and Yangmai158

The Stpk-V was identified as a key member of the powdery mildew resistance gene Pm21. In the present study, we attempted to find the critical genes or pathways responsible for powdery mildew resistance mediated by Stpk-V through analysis of DEGs. Several disease-resistance-related genes involved in different signalling pathways were chosen to confirm the DGEs results with qRT-PCR (Fig. 1a–f).

Involvement of reactive oxygen species in Stpk-V-mediated resistance

In our previous study (Cao et al. 2011) transgenic Stpk-V plants showed accumulation of reactive oxygen species (ROS) and hypersensitive response cell death after Bgt infection, indicating that ROS production might be crucial in the resistance mediated by over-expression of Stpk-V. Among the unigenes in the six DGE libraries, we detected 12 unigenes related to the ROS pathway: two oxalate oxidases (OxO), three germin-like proteins (GLP), five peroxidases (POD) and two glutathione peroxidases (GPX) (Table S3). One NADPH oxidase (Nox) was additionally selected and identified as a marker gene for the ROS pathway. This transcriptome expression result was confirmed by qRT-PCR as shown in Fig. 1a1–a6. Similar to the DGEs sequencing results, the ROS-producing gene Nox (Ta#S52545338) was up-regulated after Bgt inoculation only in T₅, and was unchanged in Yangmai158. The GLP gene (Ta#S52129023) was up-regulated after Bgt inoculation in both T₅ and Yangmai158, however, the expression levels were higher in T₅ than that in Yangmai158. POD (Ta#S26021080) and OxO (Ta#S12923252) genes were slight up-regulated in T₅ after Bgt inoculation but down-regulated or unchanged in Yangmai158. Two other ROS-scavenging superoxide dismutase (SOD, Ta#S17986579) and catalase (CAT, Ta#S58904542) genes were up-regulated only in Yangmai158 after Bgt inoculation, and were unchanged in T₅. These results indicate that ROS-generated genes were activated in both genotypes at an early stage, however, these genes were activated continuously only in T₅. So, the Stpk-V mediated ROS burst in T₅ was from both sustained expressed genes activated at an early stage and newly activated gens in the later stage, e.g. at 24 hai.

Salicylic acid (SA), jasmonic acid (JA) and ethylene (ET) pathways

To determine which hormone pathway was related to powdery mildew resistance mediated by Stpk-V, expression patterns of pathway-associated genes were analysed and compared in Yangmai158 and the T₅ line.

Seven genes associated with the SA pathway were identified: one EDR1 gene, one SA binding protein and five PR1 genes (Table S3). The SA binding protein gene SBP2 (Ta#S52542789) and one of the PR1.2 genes (Ta#S18006174) were induced only in T₅ after Bgt inoculation. Another PR1.1 gene (Ta#S58905285) and the NPR1 gene (Ta#S52545363) showed induced transcription in both genotypes, but the expression level of the gene was higher in the T₅ line than that in Yangmai158 (Fig. 1b1–b4). This suggests that the SA pathway was activated in T₅ and was involved in the defence response to Bgt infection, especially at late Bgt infection stage.

Fifteen genes related to the JA pathway were identified, 10 genes responsible for JA biosynthesis, including lipoxygenase (LOX), allene oxide synthase (AOS), allene oxide cyclase (AOC) and 12-oxo-phytodienoic acid reductase (OPR3), and five genes involved in JA signalling, including CORONATINE INSENSITIVE1 (COI1) and MYC (Table S3). Among these genes, one LOX gene (Ta#S52544755) and one MYC gene (Ta#S26028947) exhibited increased accumulation in both Yangmai158 and T₅ at 6 hai; however, only the LOX gene was up-regulated continuously at 24 hai in T₅. The OPR3 gene (Ta#S17889436), AOS gene (Ta#S16058046) and COI1 gene (Ta#S52543991) showed increased accumulation only in T₅ after Bgt inoculation (Fig. 1c1–c5). These results indicate that the JA pathway is activated in both genotypes during early Bgt infection stage, but was only activated continuously in T₅ through participation of Stpk-V at late Bgt infection stage.

Ten genes associated with the ET pathway were identified, four genes responsible for ET biosynthesis [including S-adenosylmethionine decarboxylase (SAM), ACC oxidase (ACO)] and six genes responsible for ET signalling [including ETHYLENE INSENSITIVE 3 (EIN3), ETHYLENE RESPONSE FACTOR1 (ERF)], however, expression of all these genes were unchanged or down-regulated in both genotypes. The results were confirmed by qRT-PCR for the ERF gene (Ta#s56463152) and SAM gene (Ta#S58902923) (Fig. 1d1, d2). This indicates that the ET pathway is perhaps not activated in the defence response to Bgt infection, especially at 6 or 24 hai.

Pathogenesis-related proteins

The 21 unigenes that encode PR proteins were identified in this study: five pathogenesis-related proteins 1.1 or 1.2 (PR1), two β-1-3-glucanases (PR2), eight chitinases (PR3), two vacuolar defence proteins (PR4), two thaumatin-like proteins (PR5) and three non-specific lipid transfer proteins (PR14) (Table S3). Among these, one PR3.2 gene (Ta#S17888136), two PR4 genes (Ta#S22379469 and Ta#S37933362) and one PR14.1 gene (Ta#S26027104) were only up-regulated in T₅. One PR2 gene (Ta#S25791161), one PR3.1 gene (Ta#S12923205), one PR5 gene (Ta#S26028075) and one PR14.2 gene (Ta#S26025120) were up-regulated in both Yangmai158 and T₅, but the expression level was higher in T₅ than in Yangmai158 after Bgt inoculation over time (Fig. 1e1–e8). In total, 10 PR genes were up-regulated only in T₅, among which nine were detected with high expression levels at 24 hai, indicated these genes were up-regulated by Stpk-V at the later stage.

Calcium signalling pathway

In our research, one calmodulin gene (Ta#S12923163) and one CDPK gene (Ta#S43837852) were differentially expressed in T₅, and these genes showed up-regulation only after Bgt inoculation (Fig. 1f1, f2), indicating that the calcium signalling pathway is involved in the defence response to Bgt infection.

Analysis of ROS signal in resistance mediated by Stpk-V using histochemical methods

In this study, the Stpk-V transgenic T₅ lines and Yangmai158 were treated with H₂O₂ and its antagonist DMTU, respectively, before Bgt inoculation to examine the influence of H₂O₂ on cell death and Bgt development. The percentages of H₂O₂ burst cells decreased in the DMTU-treated T₅ and Yangmai158 at 6, 12 and 24 h after Bgt inoculation compared to the MOCK controls. Correspondingly, the percentages of H₂O₂ burst cells increased in the H₂O₂-treated samples compared to the MOCK controls (Fig. 2a–c). As to the percentage of cell death in the resistant T₅, this decreased when treated with DMTU, but increased when treated with H₂O₂ (Fig. 2d–f). There was similar response trends in Yangmai158 and Stpk-V transgenic lines, however, the percentages of H₂O₂ burst cells and the percentages of cell death in Stpk-V transgenic lines were higher than in Yangmai158 at the three time points.

Papillae, built by the host cells surrounding the appresorium penetration pegs of Bgt, were considered important defence structures to suppress the attack of Bgt. In this study, a high percentage of papillae formation was found in both the susceptible Yangmai158 and resistant T₅ line, indicating that papillae may be involved in basal resistance (Fig. 2d). In the resistant T₅ line, slight changes in percentage of papillae formation were detected between the DMTU- and H₂O_2-treated samples after Bgt inoculation (Fig. 2g), suggesting that H₂O₂ might make a small contribution to papillae formation. However, the percentage of papillae formation was higher in the T₅ line than in Yangmai158, indicating that another signal transduction pathway induced by Stpk-V, other than the H₂O₂, was responsible for the increased papillae formation in the T₅ line.

The percentage of secondary hyphae formation was also compared between the DMTU-treated and MOCK controls of Yangmai158 and Stpk-V. As to the T₅ line, the percentage of secondary hyphae formation at 24 and 48 hai was significant higher in the DMTU-treated samples than that in the MOCK controls and H₂O₂ treatment (Fig. 2h, i), which indicates that H₂O₂ had an important role in suppression of the normal development of Bgt. Also, there was a large difference between the H₂O₂-treated Yangmai158 and the T₅ line, once again indicating that resistance of T₅ was mediated through a complex network that included both H₂O₂ and other signals.

Evolution of the Stpk-V gene in the grass family

To investigate the conservation of Stpk-V in other grass species related to wheat, the partial sequence of Stpk were amplified from the disomic addition lines of the wheat variety ‘Chinese Spring’ and from corresponding wheat relatives. The tested wheat relatives were S. cereale, Ae. longissima, H. vulgare, Ae. geniculata, Aegilops peregrina. The PCR results showed that the same specific bands were amplified only from the wheat disomic addition lines containing the homologous group 6 alien chromosomes as that from the corresponding wheat relatives. This indicates that genes orthologous to Stpk-V are all located on homologous group 6 chromosomes (Fig. 3). Other genes orthologous to Stpk-V with high homology were also located to the syntenic chromosomal regions in rice, Brachypodium, sorghum and maize. Thus, Stpk is highly conserved in the grass family. However, homology of Stpk-V to genes in dicot species, such as Arabidopsis, was relatively low. Thus, Stpk is highly conserved in the grass family.

Resistance evaluation of Stpk-V to powdery mildew in barley and Arabidopsis

The Stpk-V confers broad spectrum resistance to the powdery mildew pathogen in wheat. To determine whether Stpk-V confers resistance against powdery mildew pathogens in other species, resistance was evaluated in the barley–Bgh and Arabidopsis–Golovinomyces orontii host–pathogen interaction systems. The successful formation of a mature haustorium, a structure for nutrient extraction during Bgh development, is usually considered as a compatible interaction between barley and Bgh (Bushnell et al. 1998). It was found that Stpk-V could prevent haustorium formation, so it mediated the before-haustorium formation resistance (Cao et al. 2011). In this study, a single-cell transient expression assay was conduced to evaluate the resistance effect of Stpk-V in the barley–Bgh system using the ‘haustorium index’ at 48 hai as parameter. The haustorium index was 79.7% in epidermal cells of the susceptible barley variety ‘Hua30’ transformed with the GUS gene. However, when the cells were co-transformed with GUS and Stpk-V, the haustorium index decreased to 42.7%, indicating that over-expression of Stpk-V could increase resistance of barley to Bgh (Fig. 4).

In the Arabidopsis–G. orontii host–pathogen system, we examined the effect of Stpk-V through stable transformation. Transgenic plants expressing Stpk-V were produced by introducing pCAMBIA1301:Stpk-V into the susceptible accession Col-0 via Agrobacterium-mediated transformation. Molecular screening and GUS staining in T₁ and T₂ generation plants were used to identify positive transgenic plants. Positive transgenic T₃ generation plants were used to evaluate the resistance effect of Stpk-V to G. orontii infection. The phenotype of the leaf and growth of hyphae after pathogen inoculation were compared between the three homozygous transgenic lines and the wild-type Arabidopsis for powdery mildew resistance evaluation. Both wild-type Arabidopsis and the transgenic lines were totally covered in hyphae. To further confirm the results, we performed microscopy observations of the inoculated leaves, and no difference was detected in the development of spores between the wild-type Arabidopsis and the transgenic lines (Fig. 5). The results indicate that expression of Stpk-V in Arabidopsis could not improve its resistance to powdery mildew.

Different effects of Stpk-V on powdery mildew resistance in monocots and dicots

The cereal powdery mildew fungus Blumeria graminis is classified into four formae speciales (f. sp.) based on their strict host specialisation: B. graminis f. sp. hordei on species of Hordeum, f. sp. tritici on Triticum, f. sp. secalis on Secale and f. sp. Avenae on Avena. Bgt and Bgh diverged about 10 MY ago, which was 2 MY later than the divergence of wheat and barley, based on analysis of the gene space region and transposons in the genomes of Bgt and Bgh (Oberhaensli et al. 2011). It has been proposed that there was originally a common, temporary Blumeria graminis formae that was able to infect both wheat and barley, which then evolved into forms specific to wheat or barley.

Due to the relatively close evolutionary relationship between the two hosts, wheat and barley, and between the two pathogens, Bgt and Bgh, there should be some common resistance pathways shared by the two hosts and activated by similar factors shared by the two pathogens. Broad-spectrum resistance can be obtained from genes that recognise conserved effectors produced by different isolates of the same pathogen (Dong 1998). The Stpk-V gene was shown to be located on chromosome arm 6VS, and the three orthologous genes Stpk-A, Stpk-B and Stpk-D were located to chromosomes 6AS, 6BS and 6DS of wheat, respectively (Cao et al. 2011). In this study, the orthologous genes of Stpk-V were located on the homoeologous group 6 chromosomes in many cereals, and the Stpk-V gene conferred resistance not only to different isolates of Bgt but also to the mixed population of Bgh. It is proposed that the Stpk-V gene can perceive the upstream signals induced by the conserved factors of Bgt and Bgh, and then activated downstream signals leading to the effective resistance to powdery mildew in the two hosts. It is also suggested that the resistance mechanism of the ancient gene Stpk-V to B. graminis established before the evolutionary divergence of Bgt and Bgh. It will be interesting to see whether the Stpk-V gene confers resistance to other formae speciales of B. graminis.

Golovinomyces orontii is a powdery mildew pathogen specific to Arabidopsis and other dicot species. The evolutionary relationship of Bgt and G. orontii is relatively more distant than that of Bgt and Bgh, and the genetic relationship of wheat and Arabidopsis is also far from that of wheat and barley. So, some factors shared by Bgt and Bgh might not be shared by G. orontii, and the resistance pathway activated by some genes in wheat and barley might not be established in Arabidopsis. It was found that overexpression of Stpk-V conferred no resistance to G. orontii in Arabidopsis. Stpk-V may be specific for monocot signalling or Blumeria resistance, based on results that the current construct (Maize Ubi Promoter:: Stpk-V) could not improve resistance of Arabidopsis.

Broad-spectrum resistance mediated by Stpk-V

The ROS pathway plays an important role in Stpk-V-mediated resistance

The ROS bursts occur at early stages after pathogen infection and later induce HR, resulting in rapid cell death in tissue immediately surrounding the infection sites to limit development of the pathogen (Torres et al. 2006). In our previous study, the most up-regulated gene in response to Bgt at 24 hai in Haynaldia villosa was an oxalate oxylase gene (OxO), a ROS-generating enzyme that can oxidise oxalate and thus generate H₂O₂. In this study, we also found that some ROS producing genes were activated continuously or specifically only in the T₅ line. It was revealed from histochemical analysis and DGE information that the ROS burst was activated in the early basal resistance stage, both in susceptible Yangmai158 and resistant T₅ transgenic plants; however, a high percentage of the H₂O₂ burst in the later specific resistance stage occurred only in resistant T₅ plants. This implies that the early and transient ROS burst in the basal resistance stage made no long-term contribution to powdery mildew resistance in Yangmai158, while the continuous production of ROS is critical for regulation of HR leading to Pm resistance in Stpk-V transgenic plants, and the disease resistance conferred by Stpk-V may partially result from its constitutive regulation of the ROS-producing pathway (Fig. 6).

In our previous study, Stpk-V was characterised as a key member of the Pm21 gene, which could be induced by Bgt and exogenous H₂O₂, but also mediated an increase of endogenous H₂O₂, leading to cell death and powdery mildew resistance, when the plant was attacked by Bgt (Cao et al. 2011). Therefore, we examined whether ROS pathway was blocked by DMTU or activated directly by H₂O₂ in the wild-type transgenic recipient Yangmai158 and the Stpk-V transgenic lines, whether the ROS response or resistance to Bgt changed and how the related pathway gene acts. The results help us to further understand the relationship between the Stpk-V-mediated resistance and the H₂O₂-mediated ROS pathway. The results showed that there were similar response trends in Yangmai158 and Stpk-V transgenic lines, however, the percentages of H₂O₂ burst cells and the percentage of cell death in Stpk-V transgenic lines was higher than that in Yangmai158, which implied that Stpk-V was a critical gene for H₂O₂ accumulation and took part in regulation of the hypersensitive responses.