Introduction

In their cropping environments, plants are continuously threatened by various pathogens. The first layer of plant innate immune response is achieved by the recognition of pathogen-associated molecular patterns (PAMPs) based on plasma membrane-localized pattern recognition receptors (PRRs). The PAMPs-triggered immunity (PTI) leads to the occurrence of various events that prevent pathogens from further invasion, including transcriptional induction of genes encoding pathogenesis-related (PR) proteins, burst of reactive oxygen species (ROS), deposition of callose at the cell wall, and signal transduction by mitogen-activated protein kinases (MAPKs) (Fritig et al. 1998; Jones and Dangl 2006; Newman et al. 2014).

The hypersensitive response (HR) triggered by the recognition of a pathogen avirulent (Avr) effector by a plant resistance (R) protein has been widely considered as the second layer of plant innate immune response. Currently, most of the cloned plant R genes have been found to encode nucleotide-binding site-leucine-rich repeat (NBS-LRR) proteins (Michelmore et al. 2013). Rapid accumulations of ROS and auto-florescent phenolic compound seem to be the earliest events in the HR (Kamoun et al. 1999; Wang et al. 2007). Then, a large number of programmed cell death (PCD)-related genes, including PCD regulators and protein caspases, are successively activated (Wang et al. 2012; Tang et al. 2015). In addition, such effector-triggered immunity (ETI) can also induce the systemic acquired resistance (SAR), an inducible form of plant defense conferring broad-spectrum immunity to secondary infection beyond the initial infection site. In Arabidopsis, SAR involves the generation of mobile signals, accumulation of salicylic acid (SA) hormone, and transcriptional activation of PR genes (Fu and Dong 2013).

Wheat leaf rust caused by Puccinia triticina (Pt) is one of the most severe fungal diseases threatening the global wheat production. Typically, the yield loss caused by wheat leaf rust can reach as high as 40% under favorable conditions. In China, several destructive epidemics of wheat leaf rust occurred in the 1970s, which had especially resulted in dramatic yield losses in the Yellow-Huai areas. Due to the trends of global warming and high-density wheat cultivation, leaf rust has expanded its area of infection to most of the wheat-planting regions in China. Recent epidemics of wheat leaf rust in China were reported in 2012 (Zhou et al. 2013a). Currently, the utilization of wheat leaf rust resistance (Lr) genes in breeding programs is still the most effective way to control this disease.

So far, more than 70 Lr genes have been designated, based on which two distinct types of resistance conferred by different Lr genes have been identified: seedling resistance and adult-plant resistance (APR). According to the gene-for-gene theory, seedling resistance is conferred by a single dominant R gene controlling the mentioned HR or ETI. Most of the cloned Lr genes in seedling resistance encode NBS-LRR proteins, including Lr1, Lr10, and Lr21 (Feuillet et al. 2003; Huang et al. 2003; Cloutier et al. 2007). On the other hand, APR is normally controlled by quantitative trait loci (QTL), and some of the APR genes have displayed the durable resistance to multiple pathogens, including wheat leaf rust (Lr), yellow rust (Yr), stem rust (Sr), and powdery mildew (Pm). At present, only three APR genes have been cloned in wheat, which are Lr34/Yr18/Pm38 encoding an ABC transporter, Yr36 encoding a kinase-START protein, and Lr67/Yr46 encoding a hexose transporter (Fu et al. 2009; Krattinger et al. 2009; Moore et al. 2015).

Until now, several transcriptome studies have been carried out to investigate the mechanism of wheat resistance to various fungal diseases. For instance, the genes associated with Lr28-mediated wheat resistance to Pt have been identified by the Serial Analysis of Gene Expression (SAGE) (Singh et al. 2017). Resistance responses in wheat upon infections by stripe rust, powdery mildew, and Fusarium head blight were investigated by large-scale transcriptomic comparisons (Xiao et al. 2013; Zhang et al. 2014). Both SSH library construction and transcriptome sequencing were conducted to understand the mechanism of APR in an elite wheat germplasm “Xingzi 9104” (Huang et al. 2013; Hao et al. 2016). The common genes identified from these related studies may be valuable genetic resources for the improvement of broad-spectrum resistance in wheat to various pathogens.

Previously designated seedling resistance genes Lr39 and Lr41 were transferred from Aegilops tauschii into common wheat and reported as the same or closely linked genes on chromosome 2DS, with an approximate distance of 1.9 cM to the SSR marker Xgdm35 (Singh et al. 2004). In a more recent investigation, wheat isogenic line carrying the Lr39/41 gene was found to have moderate to high resistance to most of the Pt pathotypes detected in China (Li et al. 2010). However, the Lr39/41 gene has not been cloned yet, and the transcriptional changes associated with the Lr39/41-mediated resistance remain largely unknown. In the current study, a typical HR was observed by microscopy in leaves of the Lr39/41 isogenic line inoculated with the avirulent Pt pathotype THTT from 48 h post-inoculation (hpi). The differentially expressed genes (DEGs) involved in the Lr39/41-mediated resistance were initially revealed by constructing SSH libraries and subsequent microarray hybridization. Using a qRT-PCR assay, the expression profiles of eight selected Lr39/41 resistance-related DEGs were validated.

Materials and methods

Plant materials, Pt inoculation, histological observation, and RNA extraction

Seedling plants of wheat isogenic line carrying the Lr39/41 gene in a Thatcher background and Thatcher susceptible line were maintained and inoculated with the Pt pathotype THTT following the procedures and conditions described earlier (Zhou et al. 2013b). Seedlings inoculated with sterile water were used as a control, and disease symptoms were scored at 10 days post-inoculation (dpi).

For histological observation, wheat leaves from the Lr39/41 isogenic and Thatcher susceptible lines were collected at 24, 48, and 120 hpi with Pt pathotype THTT. The samples were bleached using ethanol and acetic acid. Then, the rust fungal structures were specifically stained using Calcofluor White (Sigma, St. Louis, MO, USA) as previously described (Wang et al. 2014). The hyphal structure and accumulation of phenolic autofluorescence at the infection sites were observed and photographed by an Olympus DP70/DP30BW microscope camera (excitation filter, 485 nm; dichromic mirror, 510 nm; barrier filter, 520 nm).

Plant samples were collected for RNA isolation at 0 and 48 hpi for SSH libraries construction and further microarray hybridization. RNA samples were harvested at 0, 1, 2, 5, and 8 dpi for qRT-PCR assay. Four independent biological replicates were set at each time point. Besides, all samples were rapidly frozen in liquid nitrogen and stored at −80 °C.

Construction of SSH libraries

The frozen samples collected at 48 hpi were grinded in liquid nitrogen. Total RNA was extracted using an RNA Extraction Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instructions. Subsequently, mRNA was obtained by purification of total RNA from each sample using a ployATract mRNA Purification Kit (Clontech, Mountain View, CA, USA). Two SSH libraries were constructed using a BD PCR-selectTM cDNA Subtraction Kit (Clontech). In detail, SSH library “JD” was constructed using mRNA samples from Pt-inoculated Lr39/41 isogenic line as the tester (being enriched) and from non-inoculated Lr39/41 isogenic line as the driver (being suppressed). Another SSH library “BJ” was constructed with mRNA samples from non-inoculated Lr39/41 isogenic line as the tester and from non-inoculated Thatcher susceptible line as the driver. cDNAs synthesized from the tester and the driver were digested with RsaI, extracted by phenol/chloroform, precipitated by ethanol, and finally re-suspended in water. Initially, the tester cDNA was split into two pools that were then ligated with different adaptors. Two rounds of hybridization and PCR amplification were carried out to normalize and enrich the differentially expressed cDNA. The 1st PCR analysis was performed in 27 cycles (94 °C for 15 s; 66 °C for 30 s; and 72 °C for 90 s), whereas 12 cycles (94 °C for 30 s; 68 °C for 30 s; 72 °C for 90 s) were performed in the 2nd PCR. The PCR products were purified and cloned into the pGEM T-easy vector (Promega) and transformed into Escherichia coli DH5a competent cells.

Microarray hybridization

A total of 6000 clones from these two SSH libraries were used as PCR templates. Inserts were amplified using the nested PCR primers 1 and 2R (Clontech, Table S1) that flanked each end of the insert region. A 50-μL PCR mixture containing 37.0 μL of sterile water, 5.0 μL of 10 × PCR buffer, 10 μL of 10 mM dNTP mix, 2.0 μL of each primer (10 μM), 1.0 μL of 50 × Advantage cDNA polymerase mix, and 1 μL of E. coli culture were prepared. The following PCR conditions were used: an initial denaturation at 94 °C for 25 s, followed by 12 cycles of 94 °C for 10 s, 66 °C for 30 s, and 72 °C for 90 s, and then a final extension was realized at 72 °C for 10 min. The PCR products were precipitated with anhydrous ethanol and subsequently re-suspended in 40 μL of sterile water. PCR products of 23 clones from each of these two SSH libraries were run on 1% agarose gel to check the size of inserts.

Then, PCR products were precipitated one more time with 100 μL of anhydrous ethanol and re-suspended in 15 μL of 50% dimethyl sulphoxide (DMSO) to a final concentration of 0.1–0.5 μg/μL. All the samples were spotted in three replicates onto amino-silane glass slides by the SmartArrayerTM Arrayer (CapitalBio, Beijing, China). Eight sequences derived from intergenic regions of yeast genome, showing no significant homology to any existing wheat sequences in GenBank (https://www.ncbi.nlm.nih.gov/), were spotted multiple times onto the slides as exogenous controls. In the following steps, wheat total RNA was spiked with a mixture of yeast exogenous control RNAs to verify the semi-quantitative microarray result. The slides were baked for 1 h at 80 °C and dried at room temperature. Prior to hybridization, the slides were rehydrated over 65 °C water for 10 s, briefly dried on a 100 °C heating block for 5 s, cross-linked by 250 mJ/cm2 UV light irradiation, washed with 0.5% SDS for 15 min at room temperature, dipped in anhydrous ethanol for 30 s, and spin-dried at 1000 g for 2 min.

RNA samples extracted from leaves of the Pt-inoculated and non-inoculated Lr39/41 isogenic lines, and from leaves of the Pt-inoculated and non-inoculated Thatcher susceptible lines, respectively, were used for the Cy5/Cy3-labeled microarray hybridizations (Guo et al. 2005). Briefly, cDNA was synthesized using 10 μg of total RNA extracted from each sample with a cDNA Synthesis Kit (Takara), which was then purified using a PCR Purification Kit (QIAGEN). Subsequently, cRNA was synthesized using a T7 RiboMAX Express Large-scale RNA Production System (Promega, Madison, WI, USA) and purified with a RNeasy Mini Kit (QIAGEN). cDNA was fluorescently labeled with Klenow enzyme after reverse transcription with Superscript II Reverse Transcriptase (Invitrogen, Waltham, MA, USA) and purified with a PCR Purification Kit (QIAGEN). Later, equal amounts of labeled control and test samples were mixed into a hybridization solution (3 × SSC solution, 0.2% sodium dodecyl sulfate, 25% formamide, and 5 × Denhart’s solution) based on the efficiency of Cy5-dCTP or Cy3-dCTP incorporation. The hybridization solution was denatured at 95 °C for 3 min prior to loading onto the PCR product-immobilized slides. Then, the slides were hybridized at 42 °C overnight and washed with two washing solutions (2 × SSC solution at 42 °C for 5 min and 0.2% SSC solution at room temperature for 5 min) consecutively. Each hybridization was comprised of two biological replicates and was performed in duplicate by dye swap.

Microarray data analysis

Microarray slides were scanned using a confocal laser scanner LuxScan 10 K-A (CapitalBio) and the resulting images were analyzed using the GenePix Pro v4.0 software. Spots with the signal pixels less than 50% of the local background value were removed. The ratio values were normalized according to the spatial and intensity-dependent (LOWESS) normalization method (Yang et al. 2002). Clones with less than eight data sets were removed. Differentially expressed clones were identified using the significance analysis of microarrays (SAM) method (Tusher et al. 2001). SAM plot was generated for each microarray assay. Mean and standard error (SE) of the fold change, SAM score, and local false discovery rate (FDR) for each clone were calculated. Clones with FDR less than 0.01 and a fold change greater than 1.5 were defined as significantly upregulated, whereas clones with FDR less than 0.01 and a fold change less than 0.5 were considered as downregulated.

Sequencing and annotations of differentially expressed clones

A total of 386 clones remarkably induced or suppressed in the microarray assay were further sequenced using T7/SP6 primers derived from the pGEM T-easy vector (Promega). All sequences were annotated with both the GenBank non-redundant protein database (Blastx, http://blast.ncbi.nlm.nih.gov) and URGI wheat genome database (Blastn, https://urgi.versailles.inra.fr). The redundant clones were removed. Snapshots of comparative expression changes were presented using MAPMAN software, which were achieved by assigning DEGs to the “biotic stress” pathway based on “Taes_AFFY_0709” mapping file from Triticum aestivum (Thimm et al. 2004; Barbierato et al. 2016). Additionally, sequences of all the DEGs were annotated with Gene Ontology (GO, http://amigo1.geneontology.org/cgi-bin/amigo/blast.cgi).

qRT-PCR assay

RNAs for qRT-PCR assay were extracted using a RNA Extraction Kit (QIAGEN) in accordance with the manufacturer’s instructions, and the first-strand cDNA was synthesized using a Reverse Transcription Kit (Takara). The primers utilized for qRT-PCR were listed in Table S1. Wheat Actin gene (TaActin, GenBank accession AB181991.1) was used as an internal reference for qRT-PCR analyses (Paolacci et al. 2009). The amplification efficiency of each pair of primers was calculated by preliminary qRT-PCR assay using six 2-fold diluted cDNA samples (1:1, 1:2, 1:4, 1:8, 1:16, and 1:32). Dissociation curves at temperatures ranging from 60 °C to 94 °C were generated to ensure that a single and specific product was amplified. The gene expression was quantified by the threshold values (Ct) generated from the Roche LightCycler 96 based on the Delta Ct method (Wang et al. 2014, 2016).

Data availability

The datasets during and/or analysed during the current study are available from the corresponding author on reasonable request.

Results

HR was observed in leaves of the Lr39/41 wheat isogenic line inoculated with the avirulent Pt pathotype THTT

In our previous research, wheat isogenic line carrying the Lr39/41 gene showed a moderate to high resistance to most of the Pt pathotypes detected in China, including FCST, TCGT, FGSQ, FHST, PHSS, FHGS, THTT, and FCTT. Particularly, THTT was the most prevalent pathotype in China, which had displayed high virulence to many of the designated Lr genes (Li et al. 2010). After inoculating wheat seedlings with the Pt pathotype THTT, necrotic spots phenotype (Infection Type; or 1) was observed on leaves of the Lr39/41 isogenic line, whereas heavy rust pustules (Infection Type 4) were observed on leaves of the Thatcher susceptible line (Fig. 1a).

Fig. 1
figure 1

HR was observed in leaves of the Lr39/41 wheat isogenic line inoculated with the avirulent Pt pathotype THTT. a Wheat isogenic line carrying the Lr39/41 gene had displayed a moderate to high resistance to Pt pathotype THTT, which was identified as the most prevalent pathotype in China. Necrotic spots phenotype (Infection Type; or 1) was observed on leaves of the Lr39/41 isogenic line, whereas heavy rust pustules (infection type 4) were observed on leaves of the Thatcher susceptible line at 10 dpi. b To observe the histological changes associated with the Lr39/41-mediated resistance, samples from both the Lr39/41 isogenic and the Thatcher susceptible lines were collected at 24, 48, and 120 hpi. Fungal structures were specifically stained using Calcofluor White. After successful penetration of the germ tube and formation of the stomatal vesicles, the production of Pt hypha was blocked by the HR generated in leaves of the Lr39/41 isogenic line from 48 to 120 hpi, whereas the Pt hypha extended, and mature haustoria were formed in leaves of the Thatcher susceptible line. GT, germ tube; SV, substomatal vesicle; IH, initial hypha; SH, secondary hypha; H, haustoria; and HR, hypersensitive response. Scale bars = 100 μm

Full size image

To observe the histological changes associated with the Lr39/41-mediated resistance, Pt-inoculated leaves from both the Lr39/41 isogenic and the Thatcher susceptible lines were collected at 24, 48, and 120 hpi, respectively. After successful penetration of the germ tube and formation of the stomatal vesicles, the production of Pt hypha was blocked by the HR generated in leaves of the Lr39/41 isogenic line from 48 hpi. In comparison, Pt hypha extended, and haustoria were formed in leaves of the Thatcher susceptible line (Fig. 1b).

SSH libraries construction and Lr39/41 resistance-associated DEGs

Two Lr39/41 resistance-associated SSH libraries were established. The first one, containing 3456 clones, was constructed using mRNA samples from the Pt-inoculated Lr39/41 isogenic line as the tester and the non-inoculated Lr39/41 isogenic line as the driver. The second, with 2544 clones, was developed using mRNA samples from the non-inoculated Lr39/41 isogenic line as the tester and the non-inoculated Thatcher susceptible line as the driver. Twenty-three clones randomly selected from each of the SSH libraries were amplified by PCR with an average size of the inserts of 350 bp (Fig. S1).

A total of 6000 clones from these two SSH libraries were further used for microarray hybridization. RNA samples extracted from leaves of the Pt-inoculated (Pt pathotype THTT at 48 hpi) and non-inoculated Lr39/41 isogenic lines, as well as from leaves of the Pt-inoculated and non-inoculated Thatcher susceptible lines, were employed for Cy5/Cy3-labelled microarray assays, respectively. Through conducting SAM analysis, clones with “FDR < 0.01 and a fold change > 1.5” were defined as significantly upregulated, whereas clones with “FDR < 0.01 and a fold change < 0.5” as downregulated (Fig. 2). All these differentially expressed clones were further sequenced and gene annotations were performed using both GenBank non-redundant protein database and URGI wheat genome database. A total of 34 upregulated and 2 downregulated DEGs were identified through comparing the Pt-inoculated Lr39/41 isogenic line with the non-inoculated one as “Lr39/41 resistance-related DEGs” (Table 1), whereas 11 upregulated and 18 downregulated DEGs by comparing the Pt-inoculated Thatcher susceptible line with the non-inoculated one as “Thatcher susceptible-related DEGs” (Table 2). The DEGs shared among different groups are presented in Fig. S2.

Fig. 2
figure 2

Differentially-expressed clones in the microarray assay. A total of 6000 SSH clones were used for microarray hybridization. Each hybridization was comprised of two biological replicates and was performed in duplicate by dye swap. Differentially expressed clones were analyzed according to SAM. Mean and SE of the fold change, SAM score, and FDR for each clone were calculated. Clones with “FDR < 0.01 and a fold change > 1.5” were defined as significantly upregulated, whereas clones with “FDR < 0.01 and a fold change < 0.5” as downregulated. a SAM plot for the distribution of all the SSH clones in the microarray using RNA samples from the Pt-inoculated and non-inoculated Lr39/41 isogenic lines. A total of 134 differentially expressed clones were identified. b SAM plot for the distribution of all the SSH clones in the microarray using RNA samples from the Pt-inoculated and non-inoculated Thatcher susceptible lines. A total of 160 differentially-expressed clones were identified

Full size image
Table 1 Lr39/41-resistance-related DEGs in the microarray assay
Full size table
Table 2 Thatcher-susceptible-related DEGs in the microarray assay
Full size table

Pathway analysis and GO category for Lr39/41 resistance-related DEGs

A total of 15 Lr39/41 resistance-related DEGs were annotated in the “biotic stress” pathway using the MAPMAN software (Fig. 3a and Table S2). On the other hand, 10 Thatcher susceptible-related DEGs were annotated (Fig. S3a and Table S3). All the DEGs were categorized according to their GO annotations into functional groups of three main categories (Fig. 3b and Fig. S3b). For the molecular function category, the majority of Lr39/41 resistance-related DEGs were annotated with “catalytic activity” (69.4%, Fig. 3b), whereas most of the Thatcher susceptible-related DEGs were predicted to be involved in “binding” (62.1%, Fig. S3b). In terms of the biological process category, a much higher ratio of Lr39/41 resistance-related DEGs were annotated with “response to stimulus” (55.6%, Fig. 3b) than that of Thatcher susceptible-related DEGs (37.9%, Fig. S3b).

Fig. 3
figure 3

Pathway analysis and GO annotations for all the Lr39/41 resistance-related DEGs. a A total of 15 Lr39/41 resistance-related DEGs were annotated in the “biotic stress” pathway using MAPMAN software (detailed information was available in Table S2). b All the Lr39/41 resistance-related DEGs were categorized based on their GO annotations into 32 functional groups of three main categories, respectively: biological process, cellular component, and molecular function. The y-axis indicates the percentage of genes in a category

Full size image

Validation of microarray data by a qRT-PCR assay

The expression levels of eight DEGs with GO annotation of “response to stimulus” were further measured by a qRT-PCR assay. RNA samples from leaves of the Lr39/41 isogenic line inoculated with the Pt pathotype THTT were collected at 0, 1, 2, 5, and 8 dpi. Mock inoculation with sterilized water served as a control. From 2 to 8 dpi, the expression levels of a wheat gene encoding pathogenesis-related protein 1 (TaPR1, GenBank accession FJ815169.1) in the Pt-inoculated Lr39/41 isogenic line were relatively higher than those in the mock inoculation (Fig. 4a), which indicates that a basal plant resistance response has been activated.

Fig. 4
figure 4

Validation of DEGs by a qRT-PCR assay. The transcript levels of eight selected DEGs during the Lr39/41-mediated wheat resistance to Pt pathotype THTT at 0, 1, 2, 5, and 8 dpi were monitored by a qRT-PCR assay. Mock inoculation with sterilized water served as a control. The transcript levels for all genes were expressed as linearized fold-TaActin levels, which were calculated according to the formula 2(ACTIN CT – TARGET CT). a TaPR1 was used as a positive control. b-f The expression levels of TaLTP4, TaBCI6, TaTLP8, TaChit, and TaWIR1 showed significant upregulation during the Lr39/41-mediated resistance to Pt pathotype THTT, which were consistent with the microarray results. g-i The expression levels of TaSRP, TaZF19, and TaPADD2 presented significant downregulation. Data were expressed as mean ± SE from four biological replicates. The asterisk indicated a significant difference (p < 0.05) between the mock and Pt-inoculated samples upon t-test

Full size image

During the Lr39/41-mediated resistance, the expression levels of five DEGs showed upregulation, including wheat genes encoding for lipid transfer protein 4 (TaLTP4, homolog of PR14, Fig. 4b), from 2 to 8 dpi; inactive purple acid phosphatase 27 (TaBCI6, also known as barley chemical induced gene 6, Fig. 4c), at 1 dpi; thaumatin-like protein 8 (TaTLP8, homolog of PR5, Fig. 4d), from 1 to 2 dpi; endochitinase (TaChit, homolog of PR3, Fig. 4e), at 2 and 8 dpi; and wheat induced resistance 1 (TaWIR1, Fig. 4f), at 8 dpi. In contrast, three DEGs were suppressed during the Lr39/41-mediated resistance containing wheat genes encoding for stress responsive protein (TaSRP, Fig. 4g) and zinc finger CCCH domain-containing protein 19 (TaZF19, Fig. 4h), at 2 and 8 dpi, respectively; and Phospho-2-dehydro-3-deoxyheptonate aldolase 2 (TaPADD2, Fig. 4i), at 5 dpi.

Discussion

Currently, most of the cloned genes in wheat for seedling resistance to leaf rust, including Lr1, Lr10, and Lr21, have been recognized to encode proteins with the NBS-LRR structure and function through a cell death-mediated pathway (Feuillet et al. 2003; Huang et al. 2003; Cloutier et al. 2007; Michelmore et al. 2013). Based on the HR observed during the Lr39/41-mediated resistance in the present study (Fig. 1), we speculate that the Lr39/41 gene also encodes a NBS-LRR protein. However, no DEGs encoding such type of protein were detected in our SSH and microarray assays. Nevertheless, several HR-related and PR genes identified in the current study seem to be crucial for the Lr39/41-mediated resistance.

HR during the Lr39/41-mediated resistance

Accumulations of ROS and phenolic-based secondary metabolites are common features of the HR during incompatible interaction between wheat and rust (Moldenhauer et al. 2006; Wang et al. 2007). Researchers have concentrated on discovering HR-related genes involved in this biological process (Tang et al. 2015). In the current study, at least four DEGs have been reported to be involved in the HR triggered by the Lr39/41 resistance gene, including TaZF19, TaWIR1, TaSRP, and TaPADD2.

In our qRT-PCR assay, the expression levels of TaZF19 gene were suppressed at 2 and 8 dpi (Fig. 4h). A previous study showed that the overexpression of its ortholog, GhZFP1 from Gossypium hirsutum, in transgenic tobacco plants enhanced plant tolerance to salt stress and resistance to Rhizoctonia solani. Moreover, lesions generated by Rhizoctonia solani in the GhZFP1-OE transgenic tobacco plants were much smaller in size and lower in density than those in the wild type ones (Guo et al. 2009). Consequently, we speculate that this gene may function as a negative regulator of the Lr39/41-mediated HR.

TaWIR1 and its ortholog in barley have been well-characterized during the resistance of Triticeae crops to various fungal pathogens, including Magnaporthe oryzae, Blumeria graminis, and Giberella zeae (Diethelm et al. 2011; Douchkov et al. 2011; Tufan et al. 2012). It was reported to be co-segregated with the resistance QTL, contributing only to the post-penetration resistance. Recent transcriptome analysis demonstrated that TaWIR1 was also involved in wheat resistance to stripe rust at the adult-plant stage (Hao et al. 2016). Induction of TaWIR1 at 8 dpi observed in our qRT-PCR assay (Fig. 4f) indicates a possible involvement of this gene during the late stage of the Lr39/41-mediated resistance.

Regarding TaSRP, its rice ortholog showed induction by abscisic acid (ABA) and salt stress (Moons et al. 1995, 1997). Although the role of ABA in plant defense had been intensively studied during the last decade (Adie et al. 2007; Lim et al. 2015; Vos et al. 2015), there is no specific report on its involvement in wheat resistance to rust. Further studies on TaSRP may provide an initial clue for understanding the role of ABA in the Lr39/41-mediated resistance.

We have observed an inconsistent expression profile of TaPADD2 between our microarray (upregulation, Table 1) and qRT-PCR assays (downregulation at 5 dpi, Fig. 4i). In a previous study, its homolog was reported to be involved in wheat resistance to Magnaporthe oryzae by mediating the shikimate pathway during the production of phenolic-based secondary metabolites (Tufan et al. 2009). The elucidation of the molecular function of TaPDDA2 in the Lr39/41-mediated wheat resistance to Pt requires further investigation.

Involvement of the SA pathway-related genes during the Lr39/41-mediated resistance

Four DEGs or their orthologs in other plant species, including TaLTP4, TaBCI6, TaChit, and TaTLP8, were previously reported to be involved in SA-related plant defense response (Caruso et al. 1999; Beßer et al. 2000; Wang et al. 2010; Li et al. 2015). PR genes are normally considered to be the downstream responsive genes of both the PTI and the SA-mediated SAR. Several PR genes were recently reported to be responsible for the Sr13-mediated high temperature resistance in wheat to stem rust (Zhang et al. 2017). In a previous transcriptome analysis, PR genes, including those encoding thaumatin-like protein (PR5) and lipid transfer protein (PR14), were highly induced during Fhb1-mediated wheat resistance to Fusarium head blight (Xiao et al. 2013). The constant upregulation of TaLTP4, TaTLP8, and TaChit genes observed in our qRT-PCR assay (Fig. 4b, d, and e) suggests that these PR genes may be functional downstream of the Lr39/41-mediated resistance.

TaBCI6 belongs to a group of genes sensitive to SA (or its orthologs BTH and DCINA) treatment in Triticeae crops (Beßer et al. 2000). BTH treatment in wheat protects the crops effectively against powdery mildew and leaf rust by affecting multiple steps in the life cycle of the pathogens (Görlach et al. 1996; Hafez et al. 2014). In our microarray analysis, we identified TaBCI6 in both Lr39/41-upregulated and Thatcher-downregulated DEGs (Tables 1 and 2). Along with its significant induction at 1 dpi in the qRT-PCR assay (Fig. 4c), we speculate an involvement of this gene in the basal defense response during the Lr39/41-mediated resistance.

The endogenous SA level during plant defense response is largely dependent on the severity of pathogen-triggered cell death (Fu and Dong 2013). In a recent study, endogenous SA levels during YrSu-mediated wheat resistance to yellow rust were measured using high performance liquid chromatography-mass spectroscopy (HPLC-MS), and a significant elevation of SA accompanying with the HR was observed at 48 hpi (Wang et al. 2017). Therefore, we speculate that SA and SA-mediated pathway may play key roles in the Lr39/41-mediated resistance.