Transcriptional profiling reveals multiple defense responses in downy mildew-resistant transgenic grapevine expressing a TIR-NBS-LRR gene located at the MrRUN1/MrRPV1 locus

Grapevine downy mildew (DM) is a destructive oomycete disease of viticulture worldwide. MrRPV1 is a typical TIR-NBS-LRR type DM disease resistance gene cloned from the wild North American grapevine species Muscadinia rotundifolia. However, the molecular basis of resistance mediated by MrRPV1 remains poorly understood. Downy mildew-susceptible Vitis vinifera cv. Shiraz was transformed with a genomic fragment containing MrRPV1 to produce DM-resistant transgenic Shiraz lines. Comparative transcriptome analysis was used to compare the transcriptome profiles of the resistant and susceptible genotypes after DM infection. Transcriptome modulation during the response to P. viticola infection was more rapid, and more genes were induced in MrRPV1-transgenic Shiraz than in wild-type plants. In DM-infected MrRPV1-transgenic plants, activation of genes associated with Ca2+ release and ROS production was the earliest transcriptional response. Functional analysis of differentially expressed genes revealed that key genes related to multiple phytohormone signaling pathways and secondary metabolism were highly induced during infection. Coexpression network and motif enrichment analysis showed that WRKY and MYB transcription factors strongly coexpress with stilbene synthase (VvSTS) genes during defense against P. viticola in MrRPV1-transgenic plants. Taken together, these findings indicate that multiple pathways play important roles in MrRPV1-mediated resistance to downy mildew.


Introduction
Plants have evolved complex and sophisticated defense mechanisms to cope with the threat of pests and diseases. These defense mechanisms can be divided into two layers of immune responses, PAMP-triggered immunity (PTI) and effector-triggered immunity (ETI) 1 . PTI is the first layer of immune defense and is triggered the detection of conserved microbe-or pathogen-associated molecular patterns (MAMPs/PAMPs) or self-molecules (damage-associated molecular patterns, DAMPs) by pattern-recognition receptors (PRRs) 2 . ETI is activated by pathogen effector proteins via intracellular immune receptors, which typically possess central nucleotide binding and C-terminal leucine-rich repeat domains (NLRs) 3 . The intracellular immune receptors involved in ETI are also known as plant disease resistance (R) proteins. Immune receptors can be subdivided into two classes based on whether their Nterminal domain shares homology with cytosolic domains of Drosophila Toll or animal interleukin-1 receptors (TIR-NBS-LRR) or has a predicted coiled-coil domain (CC-NBS-LRR) 4 .
Studies examining the immune response to DM infection have been described in a number of DM-resistant grapevine species [20][21][22][23][24][25][26] . Most of these studies have focused on comparing the differences in the infection process between susceptible and resistant varieties and include morphological and histological differences, as well as differences in transcriptional responses. For example, leaf hairs of some grape varieties can form a natural physical barrier to prevent the invasion of the DM pathogen 27 . Transcriptional analysis of V. riparia 20 , V. amurensis 21,22,24 , V. pseudoreticulata 26 , M. rotundifolia 28 , V. labrusca, and V. vinifera hybrids containing DM resistance loci 23,24,29 has revealed that the induction and regulation of many defense-related genes may contribute to DM resistance. However, one observation that is common across many of these studies is that there is more rapid or stronger upregulation of stilbene biosynthesis pathway genes in DM-resistant genotypes than in DM-susceptible genotypes. For example, Wang et al. 28 showed that stilbene synthase (VvSTS) was induced earlier in DMresistant M. rotundifolia than in susceptible V. vinifera tissues following DM inoculation, while the resveratrol O-methyltransferase (ROMT) gene was upregulated in only the resistant genotype. More recent studies with V. vinifera hybrids containing either the Rpv3 or Rpv10 resistance locus have also strongly suggested a role for stilbenes in DM resistance mediated by these loci 23,24,29 . Similarly, in Rpv12-mediated DM resistance, transresveratrol was proposed to act as a signaling molecule in ROS formation and initiation of programmed cell death (PCD) triggered by a CC-NB-LRR gene product 24 .
One of the drawbacks of many of these studies is that they involve comparisons of DM-induced gene expression profiles between resistant and susceptible individuals, which not only differ in the presence/ absence of a DM resistance locus but also have different genetic backgrounds, i.e., they compare either different individuals from a segregating population 22,29 , different Vitis cultivars 23 or completely different grapevine species 20,25,28 . This makes it difficult to conclude with certainty that all of the differences in DM-induced gene expression observed between the resistant and susceptible genotypes are mediated by the DM resistance locus alone or are a function of other genotypic differences.
Feechan et al. 18 previously reported on the map-based cloning of seven putative resistance genes (RGAs) located at an MrRUN1/MrRPV1 locus introgressed from the wild North American grape species M. rotundifolia. Functional testing of these MrRGA genes for mildew resistance revealed that MrRGA8 conferred resistance to P. viticola, and MrRGA8 was designated MrRPV1, while MrRGA10 conferred resistance to Erysiphe necator and was designated MrRUN1. In this study, we used a transgenic DM-resistant Shiraz line expressing the MrRPV1 gene (S8-1) and wildtype plants of the same Shiraz clone, which lack the MrRPV1 gene, to investigate MrRPV1-mediated transcriptome responses within the first 36 h after DM inoculation. A weighted gene coexpression network analysis (WGCNA) was performed to identify the hub genes and key pathways involved in MrRPV1-mediated resistance. This information has allowed us to draw conclusions regarding the molecular mechanisms underlying DM resistance conferred strictly by the MrRPV1 gene without the complications associated with previous studies based on comparisons of transcriptional responses in plants with different genetic backgrounds.

MrRPV1-mediated transcriptional responses to DM infection
Feechan et al. 18 previously showed that transgenic versions of the V. vinifera cultivars Portan, Tempranillo, and Shiraz expressing the MrRPV1 gene displayed a 92-98% reduction in DM sporulation at 6 days postinoculation compared to the wild-type controls. The strongest resistance was observed in Shiraz transgenic line S8-1, which was associated with the induction of PCD in penetrated cells. The transcriptional responses of leaf discs of MrRPV1-transgenic Shiraz line S8-1 to P. viticola inoculation were thus compared with those of wild-type Shiraz leaf discs lacking the MrRPV1 gene. An examination of the timing of MrRPV1-mediated PCD induction in transgenic Shiraz, as determined by staining with trypan blue 18 , indicated that it was first observed 24 h post inoculation (hpi) (data not shown), and the transcriptional response within the first 36 hpi was therefore made a focus. Both the Shiraz transgenic line S8-1 and a wild-type Shiraz plant were propagated to produce six clonal copies of each, which were divided into three replicates of two plants each from which leaves were sampled to obtain leaf discs. Leaf discs were inoculated with DM sporangia or water and sampled at 0, 12, 18, 24, and 36 h. The remaining discs were incubated for an additional 4-5 days to check for sporulation; no sporulation was observed on the S8-1 leaf discs, whereas leaf discs from the wild-type Shiraz showed a dense covering of sporangia (~1.3 × 10 5 sporangia per disc).
For the convenience of description, S_Mock and TS_Mock are used to represent the wild-type Shiraz plants and MrRPV1-transgenic Shiraz plants under control conditions, whereas S_DM and TS_DM represent DMinfected plants, respectively. To compare transcription profiles between wild-type Shiraz plants and MrRPV1transgenic Shiraz plants in response to DM inoculation, we subjected the expression data to pairwise comparisons, i.e., S_DM vs. S_Mock and TS_DM vs. TS_Mock, at different time points (Supplemental Data Set 1). Figure 1A shows that the number of differentially expressed genes (DEGs) that were either upregulated or downregulated in the comparison of TS_DM vs. TS_Mock increased from 9 at 12 hpi to 1322 at 36 hpi. In contrast, in the comparison of S_DM vs. S_Mock, no DEGSs were observed at 12 hpi or 18 hpi, but the number increased to 45 at 24 hpi and 216 at 36 hpi. In addition, 180 DEGs were shared between S_DM vs. S_Mock and TS_DM vs. TS_Mock (Fig. S1). Furthermore, 85 out of these 180 common genes were present at 12 hpi and 18 hpi in MrRPV1-transgenic Shiraz plants (Supplemental Data Set 2). This finding indicates that DM effector recognition by MrRPV1, either directly or indirectly through interaction with an effector-modified guard or decoy protein 30 , results in a more rapid and extensive transcriptomic response to P. viticola infection in Shiraz cells with MrRPV1 than in Shiraz cells lacking MrRPV1.
To gain a better understanding of the mechanism of resistance mediated specifically by the MrRPV1 gene, we focused on a subset of DEGs that had to meet one of the following criteria: (1) the genes were differentially expressed in the comparison of TS_DM vs. TS_Mock but not differentially expressed in the comparison of S_DM vs. S_Mock or (2) the genes are differentially expressed in both S_DM vs. S_Mock and TS_DM vs. TS_Mock, but the differences in TS_DM vs. TS_Mock group were statistically significantly different from the S_DM vs. S_Mock comparison values (P-value of Fisher's test <0.05). This refinement resulted in the identification of a total of 1356 DEGs at four-time points (Supplemental Data Set 3). Almost all of the DEGs were differentially expressed in transgenic plants specifically in response to DM infection at at least one-time point. Only seven DEGs met the second criterion, i.e., were significantly induced in response to DM in both MrRPV1-transgenic and wildtype Shiraz leaf discs but were significantly more highly induced in the presence of MrRPV1. Furthermore, 94% of the DEGs were upregulated, with only 6% significantly downregulated (Fig. 1B). These findings indicate that the following activation in the presence of DM, MrRPV1 functions mainly as a positive regulator of gene expression to mediate DM resistance. Quantitative real-time PCR (qPCR) assays of a set of randomly selected DEGs confirmed that their expression was in accordance with the results of the transcriptome analysis (Fig. S2).
Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses of these DEGs were performed based on the timing of expression changes. No significant GO enrichment terms were obtained at 12 hpi because there were too few DEGs. The GO enrichment analysis of DEGs at later time points (18-36 hpi) revealed that genes involved in biological processes related to metabolic processes, including protein phosphorylation and protein modification, were significantly enriched (Table S4). Among the molecular function terms, most were related to plant immunity, including protein kinase activity, phosphotransferase activity, ATP or nucleoside binding, calcium ion binding, lipase activity, etc. (Supplemental Data Set 4). KEGG results showed that most of the DEGs in MrRPV1transgenic lines were enriched in biological pathways related to biotic stress (Fig. 2), including biosynthesis of secondary metabolites, plant-pathogen interaction, protein processing in the endoplasmic reticulum, flavonoid biosynthesis, phenylalanine metabolism, glutathione metabolism, and glycolysis/gluconeogenesis.

DEGs related to reactive oxygen species
Reactive oxygen species (ROS) are widely produced in many plants with a transient burst as an early and rapid response to pathogen attack. Our results showed differential expression of several categories of respiratory burst genes involved in ROS production and balance, including genes encoding peroxidases, oxidases, and glutathione Stransferases (Supplemental Data Set 5). Within 12 h of DM inoculation in MrRPV1-transgenic plants, two genes encoding nonsymbiotic hemoglobin and peroxidase showed~2.5-fold and~2.9-fold increases compared with their expression levels in the mock control. At later time points (18-36 hpi), significant upregulation of the genes encoding six L-ascorbate oxidases (~2.7-to~103.9-fold), two respiratory burst oxidase homolog (RBOH) proteins (~2.1-fold and~8.2-fold) and 20 glutathione S-transferases (GSTs,~1.8-to~138.0-fold) was observed in MrRPV1transgenic leaf discs. For example, the expression of an Lascorbate oxidase (VIT_07s0031g01070) showed an~12.9fold increase at 18 hpi, peaked at an increase of~103.9-fold at 24 hpi and decreased~45-fold at 36 hpi compared to expression in mock control leaf discs.

DEGs related to calcium-and kinase-mediated signaling
Kinase-mediated signaling plays a vital role in plant innate immunity. In our study, more than two hundred protein kinases, including receptor-like protein kinases (RLKs), mitogen-activated protein kinases (MAPKs or MPKs), and wall-associated receptor kinases (WAKs), were more highly upregulated in the MrRPV1-transgenic lines than in the wild-type control lines within 36 h after inoculation with DM (Supplemental Data Set 6). These differentially expressed RLKs included three types: cysteine-rich receptorlike protein kinases (CRKs), L-type lectin-domain containing receptor kinases (LecRKs) and G-type lectin S-receptor-like serine/threonine-protein kinases. Among them, there were 47 RLKs with greater than a 10-fold change in expression. The two most highly upregulated RLK genes VIT_09s0002g03010 and VIT_19s0014g04190 were both with a~108.9-fold increase at 36 hpi in MrRPV1-transgenic RichFactor refers to the ratio of the number of genes located in the pathway entry in the DEGs to the total number of annotated genes located in the pathway entry. The greater RichFactor is, the greater the degree of enrichment. The q-value indicates the P-value after correction for the testing of multiple hypotheses. The range of the q-value is [0, 1], with values closer to zero indicating more significant enrichment plants after P. viticola infection. Recognition of PAMPs by their receptors triggers a series of downstream defense responses, including the activation of MAPKs or MPKs, calcium influx, and upregulation of defense genes. From 24 hpi, differential expression of MAPK was detected with slight changes (~1.6-to~3.8-fold). Significant upregulation of the genes encoding four types of calcium sensors, including calcium cyclic nucleotide-gated ion channel (CNGC), calcium/calmodulin-binding protein (CBP), calcium-dependent protein kinase (CDPK) and CBL-interacting protein kinase (CIPK), was observed in only the MrRPV1-transgenic line. In addition, one calcium-transporting ATPase (Novel00720) and three calcium-binding proteins, CML (VIT_01s0010g02930, VIT_01s0010g02940, VIT_01s0010g02970), were expressed only after DM infection in the MrRPV1transgenic line. During pathogen infection, WAKs trigger the innate immune response as candidate receptors of cell wall-associated oligogalacturonides 31 . The expression levels of two WAKs (VIT_03s0132g00340 and VIT_03s0132g00350) continuously increased compared with those of the control from 18 to 36 hpi. Another WAK gene, VIT_17s0000g04420, was expressed only after DM infection in the MrRPV1-transgenic line.
The genes encoding EDS, PAD4, and SAG101, which are important for signal transduction mediated by TNL-R-protein, were upregulated at 18 hpi (~2.4-fold) and 36 hpi (~1.6-to~6.7-fold) (Supplemental Data Set 7). The increased expression of ten PR genes, including PR-1, PR-4, and STH-21, was also observed in MrRPV1-transgenic plants in response to P. viticola attack. It is particularly noteworthy that the expression of PR1 (VIT_03s0088g00890) was increased~478.0-fold at 36 hpi compared with the mockinoculated transgenic plants (Table S3). The expression of another PR1 gene (VIT_03s0088g00700) was observed only after DM infection in MrRPV1-transgenic lines. Three SAR DEFICIENT 1 (SARD1) genes, which have been reported as key regulators for ICS1 induction and SA synthesis 36 , were upregulated~2.8-,~6.3-, and~12.4-fold at 18 hpi. Auxin is generally a negative regulator in plant disease resistance 37 . Consistent with this, the expression of the auxin-responsive gene SAUR22 was downregulated in MrRPV1-transgenic lines in response to DM inoculation.

RPV1-mediated activation of genes related to secondary metabolism
Many studies have shown that secondary metabolites, including phenolic compounds, terpenes, and alkaloids, may play a role in plant defense against pathogens 38 . In the current research, the pathways related to secondary metabolite biosynthesis, including flavonoid biosynthesis; tropane, piperidine, and pyridine alkaloid biosynthesis; isoquinoline alkaloid biosynthesis; stilbenoid, diarylheptanoid, and gingerol biosynthesis; ubiquinone and other terpenoid-quinone biosynthesis; and terpenoid backbone biosynthesis pathways, are represented (Fig. 2). A total of 18 stilbene synthases (STSs) were upregulated within 36 hpi in MrRPV1-transgenic leaves, with 17 of them showing increases ranging from~3.0-to~24.9-fold at 18 hpi (Supplemental Data Set 8). Seven phenylalanine ammonia-lyases (PALs) were also upregulated by 2.9-fold to~16.3-fold, and two ROMT genes were upregulated by 30.1-and 93.3-fold in MrRPV1-transgenic leaves at 36 hpi by DM compared to mock-inoculated leaves (Supplemental Data Set 8).

Coexpression network analysis to identify key modules and hub genes
To gain insights into the regulation of gene networks by MrRPV1 in response to DM infection, rather than focusing on individual genes, coexpression network analysis with WGCNA was performed on the basis of correlation patterns of the expressed genes across all 54 samples, i.e., S_DM, S_Mock, TS_DM, and TS_Mock, to find the hub genes related to MrRPV1-mediated resistance to P. viticola infection. Modules are defined as clusters of highly interconnected genes. This analysis resulted in 18 distinct modules (MEs) with different expression patterns (Fig. 3, Supplemental Data Set 9). Nine MEs, designated MEblack, MEblue, MEcyan, MEgreen, MEivory, MElightsteelblue1, MEsienna3, MEsteelblue, and MEviolet, showed statistically significant differences in expression in DM-infected MrRPV1-transgenic leaf tissues compared with mock-inoculated tissues at least one-time point. Notably, the MEblack and MEsienna3 modules were specific to the MrRPV1-transgenic line. A total of 1299 genes were clustered in these two modules, of which 65.7% (854/1299) were DEGs between MrRPV1-transgenic vs wild-type Shiraz lines. This means that most of the DEGs induced by the MrRPV1 gene were concentrated in these two modules.
The expression patterns of the MEblack module genes in the mock treatment of leaf discs from MrRPV1-transgenic and wild-type Shiraz plants were relatively consistent across all time points. In wild-type Shiraz leaves, DM infection did not result in a significant change in the expression of MEblack module genes compared to that of the mock control. However, the expression of MEblack module genes increased within 12-18 hpi in the MrRPV1transgenic plants, which was followed by an even larger increase in expression 24-36 hpi. The genes in the MEsienna3 module showed a gradual downregulation in expression for the first 24 hpi, followed by upregulation between 24 and 36 hpi, which was much more pronounced in Shiraz plants containing the MrRPV1 gene.

Transcriptional regulatory modules associated with transcription factors
To investigate the regulatory network of TFs, we analyzed motif enrichment in the promoters of genes in each module compared with those of the genes in all the samples. We found that the different modules shown in Fig. 3 can be characterized by different sets of TF binding motifs. Interestingly, the MEblack cluster showed enrichment for genes with binding motifs of WRKY, MYB, and AHL transcription factors in their promoters, whereas MEsienna3 was enriched for genes with binding motifs for CMTA transcription factors. Furthermore, binding motifs for WRKY TFs were enriched only in the MEblack module, which is related to plant-pathogen interaction pathways. The WGCNA also identified a higher number of TFs, including WRKY, MYB, NAC, bHLH, and ERF, that were strongly coexpressed with STSs in the MEblack module (Fig. 4, Supplemental data set 11). All these findings suggest that transcription factors, especially WRKY TFs, play important roles in MrRPV1mediated resistance.
To further investigate connections between these TFs and the MrRPV1-induced immunity network, the kME values, a measure of correlations between the expression patterns of individual genes and those of MEs, were calculated using WGCNA. We selected differentially expressed MEs and WRKY transcription factors to construct their coexpression networks (|kME| > 0.8) (Fig. 5, Supplemental data set 12). Six modules, including MEblack and MEsienna3, were positively correlated with multiple WRKY transcription factors, while MEblue was negatively correlated with most WRKY transcription factors. This suggests that the change in expression of multiple WRKY transcription factors contributes to upregulation or downregulation of the genes in these modules.

Discussion
In this study, we utilized transgenic Shiraz plants carrying the MrRPV1 resistance gene from M. rotundifolia to investigate the mechanism by which this gene confers resistance to DM. This approach allowed the precise identification of defense genes and pathways specifically linked to MrRPV1 activation because of the high level of uniformity of the genetic background of the transgenic and wild-type plants, except for the presence of this resistance gene. As expected, significantly fewer DEGs were identified using this approach than using comparisons of transcription profiles of DM-susceptible and DMresistant grapevines with very different genetic backgrounds [20][21][22]25 . For example, we identified a total of 1356 DEGs that were differentially regulated in MrRPV1transgenic Shiraz leaves in comparison to wild-type Shiraz leaves in response to DM infection. In contrast, a previous transcriptomic analysis that compared susceptible V. vinifera and resistant V. riparia plants following DM inoculation found as many as 5550 and 6379 genes with statistically significant differences in expression at 12 and 24 hpi, respectively 20 .
PCD is a fundamental cellular process in animals and plants. In plants, two forms of PCD, developmentally controlled PCD (dPCD) and pathogen-triggered PCD (pPCD), have been described 40 . The plant hypersensitive response (HR) is a rapid localized PCD that occurs at the point of pathogen penetration and is generally associated with ETI 41 . This PCD is an effective defense strategy against biotrophic plant pathogens by restricting pathogen access to water and nutrients. However, the HR is not restricted to the ETI response 42 . More recently, ETI was shown to restore and potentiate PTI signaling components, leading to a robust immune response, and PTI coactivation was shown to enhance NLR-mediated HR cell death 43,44 . In our study, the strongest resistance associated with the induction of PCD in penetrated cells was observed in Shiraz transgenic line S8-1, and this MrRPV1-mediated PCD induction was first observed 24 h post inoculation. The MrRPV1 protein recognizes the AvrRPV1 effector protein that is secreted by P. viticola and mediates the activation of the HR, leading to leaf resistance to DM. Downstream from NLR activation, the HR involves a series of events that include calcium influxes, accumulation of SA, ROS, and transcriptional reprogramming 45 . Many DEGs related to most of these elements were identified in MrRPV1-transgenic Shiraz. Because most of these elements are shared between PRR and NLR signaling, the zigzag model of the plant immune system defined a threshold for the activation of HR 2 rather than support its being a highly regulated phenomenon. However, recent advances proposed a provocative model in which the funnel-like structure of a resistosome triggers HR cell death by translocating into the plasma membrane and perturbing membrane integrity, similar to the action of pore-forming toxins 46 . Plant hormones play central roles in the process of plant resistance to many pathogens. It is generally accepted that SA and JA/ET induce defense against biotrophic and necrotrophic pathogen attack, respectively 47 . However, studies in the past few years have proven that JA/ET-mediated defenses also contribute to resistance against some biotrophic pathogens 48,49 , although the signaling crosstalk between them is not clear. In the present study, the induction of LOX, AOS, ACO, ACS, and ERF genes involved in JA/ET signaling and biosynthesis was quantitatively greater in degree in MrRPV1-transgenic plants than in wild-type Shiraz plants from 18 hpi to 36 hpi. The results of other studies have also previously implicated a role for JA in the genetic and inducible resistance of various grapevine species and varieties to DM 20,[48][49][50][51][52][53][54][55] . PR1 gene expression is associated with the induction of disease resistance in plants. It is a molecular marker for disease resistance mediated by salicylic acid 56 . In our study, the expression of PR-1 genes increased up to 480-fold, suggesting that PR-1 genes may also play an important role in MrRPV1mediated resistance. In Arabidopsis, the EDS1-PAD4 complex predominantly contributes to TNL-based ETI 57 . Recent studies have shown that TNL-mediated resistance responses require EDS1 complexes incorporating a SAG101 isoform in Solanaceous species but not in members of the Brassicaceae 58,59 . Intriguingly, our results showed that the SAG101 genes were differentially expressed 18 h earlier than the PAD4 gene. Thus, we propose that SAG101 may be required for MrRPV1mediated immune signaling in grapevine. In Arabidopsis thaliana, the combined action of EDS1, PAD4, and SAG101 can also promote SA-mediated defenses to limit Fusarium graminearum infection 60 . Altogether, we speculate that multiple phytohormones participate in the MrRPV1-mediated defense against DM infection. A variety of secondary metabolites play a vital role in the host plant's immune process against pathogens, including terpenoids, alkaloids, flavonoids, and phenolics 38,61 . Previous studies have shown that stilbenes confer resistance to grape powdery mildew by efficient recruitment of SA signaling [62][63][64] . Multiple omics data revealed that Rpv3-1-mediated resistance to grapevine DM is associated with accumulation of stilbenes 23,24 . Genome-wide analysis of the grapevine stilbene synthase gene family identified three principal groups designated STS A, STS B, and STS C 65 . In our study, the expression of genes in groups STS A and STS C showed a continuous increase within 36 h after DM inoculation in MrRPV1transgenic leaf discs. In addition, upregulation of PAL and ROMT genes involved in the stilbenoid biosynthesis pathway was also observed in resistant plants. Our results strongly suggest an important role for stilbenes in MrRPV1-mediated defense against DM, as has been proposed previously for the DM-resistance loci Rpv3 23 and Rpv10 15,22 . The results of our gene network analysis are also consistent with the results of previous studies showing that WRKY and MYB transcription factors strongly coexpress with STS genes in grapevine tissues under biotic and abiotic stress [66][67][68] .

Plant and pathogen materials
Stably transformed transgenic Shiraz plants containing MrRPV1 were generated previously as described by Feechan et al. 18 . Untransformed and transgenic V. vinifera cv. Shiraz (clone BVRC12) plants were grown in a greenhouse (16 h light, 26°C/8 h dark, 22°C). Plasmopara viticola inoculum was collected from an experimental vineyard on the Waite Campus, Adelaide, Australia, and then maintained on detached leaves of V. vinifera cv. Cabernet Sauvignon. Discs 1-1.5 cm in diameter were cut from fully expanded leaves (collected from nodes 3-5), sprayed with a P. viticola sporangia suspension in water (5 × 10 4 sporangia/ml) and incubated in sealed petri dishes in a chamber at 22°C under a 16-h light/8-h dark cycle for 5-6 days.

RNA-seq and data analysis
A total of six clonal copies of untransformed Shiraz and six clonal copies of MrRPV1-transgenic Shiraz line S8-1 18 were divided into three groups of two "duplicates" each. Two leaves from each duplicate plant were selected and individually bagged, producing three control groups and three transgenic groups. From each group of leaves, a total of 60 discs were cut and placed onto moist filter paper with the abaxial side facing up. Prior to inoculation, six discs from each of the six groups (three controls, three transgenic) were selected at random from the 60 and snap frozen in liquid nitrogen to act as zero-time controls. Within each group, half of the discs were sprayed with water, and the other half were sprayed with a P. viticola suspension. Six discs were harvested at random from the mock and P. viticola-inoculated treatments from the three control and three transgenic groups at 12, 18, 24, 36 hpi, snap frozen in liquid nitrogen and stored at −80°C until use. Total RNA was isolated with the Spectrum Plant Total RNA purification kit (Sigma Aldrich) following the manufacturer's instructions.
A total of 3 μg of total RNA per sample was used as input material for RNA sample preparations. Sequencing libraries were generated using the NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (NEB, USA) following the manufacturer's recommendations, and index codes were added to attribute sequences to each sample. Library construction and RNA sequencing were performed by the Novogene Bioinformatics Institute (Beijing, China). Indexing of the reference genome was undertaken using Bowtie v2.0.6, and paired-end clean reads were aligned to the reference Grapevine genome PN40024 (GenBank assembly accession No: GCA_000003745.2) using TopHat v2.0.9. HTSeq v0.6.1 was used to count the read numbers mapped to each gene. Differential expression analysis of two conditions/groups (two biological replicates per condition) was performed using the DESeq R package (1.10.1). The resulting P-values were adjusted using Benjamini and Hochberg's approach 69 to control the false discovery rate. Genes with an adjusted P-value < 0.05 were assigned as differentially expressed. GO enrichment was performed using the GOseq R package 70 . KOBAS software was used to test the statistical enrichment of differentially expressed genes in KEGG pathways 71 .

WGCNA
Coexpression networks were constructed using the R package WGCNA 72 . Gene expression values of all samples were used to construct a single hybrid network. After excluding genes with FPKM > 1 across 50% of the samples, expression values of 19,732 genes were imported into WGCNA to construct coexpression modules using the automatic network construction function blockwiseModules with default settings, except that the power was 6. Genes were clustered into 18 correlated modules (correlation >0.8). The networks were visualized using Cytoscape 73 . For motif enrichment analysis using AME 74 , the 1000 nt upstream of the transcription start sites of the members of 18 modules were tested for enrichment of known cis-elements with the set of genes from all the modules as the control. The eigengene-based gene connectivity, kME, was calculated using the signedKME function in the WGCNA package and was used to visualize relationships between transcription factors and expression patterns of module eigengenes using Cytoscape. Genes with high positive kME values were referred to as intramodular hub genes and centrally located in their respective modules.

Gene expression analysis
Total RNA isolated from leaves of the untransformed and MrRPV1-transgenic Shiraz plants, mock-inoculated or inoculated with P. viticola, was used for cDNA synthesis using HiScript® II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme Biotech Co. Ltd. Nanjing, China). Real-time PCR was performed using a Light-Cycler® 480 II Real-time PCR Instrument (Roche, Switzerland) with a 10 μl PCR mixture that included 1 μl of cDNA, 5 μl of 2× QuantiFast® SYBR® Green PCR Master Mix (Qiagen, Germany), 0.2 μl of forward primer, 0.2 μl of reverse primer and 3.6 μl of nuclease-free water. Reactions were incubated in a 384-well optical plate (Roche, Switzerland) at 95°C for 5 min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. Each sample was run in triplicate. Specific primers for the selected genes are listed in Supplemental Table 1. The expression levels of mRNAs were normalized to that of EF1α and were calculated using the 2 −ΔΔCt method 75 .

Accession numbers
Sequence data from this article can be found at the NCBI repository under the following accession number: PRJNA706058.