SlWRKY30 and SlWRKY81 synergistically modulate tomato immunity to Ralstonia solanacearum by directly regulating SlPR-STH2

Abstract Bacterial wilt is a devastating disease of tomato (Solanum lycopersicum) caused by Ralstonia solanacearum that severely threatens tomato production. Group III WRKY transcription factors (TFs) are implicated in the plant response to pathogen infection; however, their roles in the response of tomato to R. solanacearum infection (RSI) remain largely unexplored. Here, we report the crucial role of SlWRKY30, a group III SlWRKY TF, in the regulation of tomato response to RSI. SlWRKY30 was strongly induced by RSI. SlWRKY30 overexpression reduced tomato susceptibility to RSI, and also increased H2O2 accumulation and cell necrosis, suggesting that SlWRKY30 positively regulates tomato resistance to RSI. RNA sequencing and reverse transcription–quantitative PCR revealed that SlWRKY30 overexpression significantly upregulated pathogenesis-related protein (SlPR-STH2) genes SlPRSTH2a, SlPRSTH2b, SlPRSTH2c, and SlPRSTH2d (hereafter SlPRSTH2a/b/c/d) in tomato, and these SlPR-STH2 genes were directly targeted by SlWRKY30. Moreover, four group III WRKY proteins (SlWRKY52, SlWRKY59, SlWRKY80, and SlWRKY81) interacted with SlWRKY30, and SlWRKY81 silencing increased tomato susceptibility to RSI. Both SlWRKY30 and SlWRKY81 activated SlPRSTH2a/b/c/d expression by directly binding to their promoters. Taking these results together, SlWRKY30 and SlWRKY81 synergistically regulate resistance to RSI by activating SlPR-STH2a/b/c/d expression in tomato. Our results also highlight the potential of SlWRKY30 to improve tomato resistance to RSI via genetic manipulations.


Introduction
To defend against pathogen attack, plants have evolved sophisticated innate immune systems, which including pattern-triggered immunity (PTI) and effector-triggered immunity (ETI) [1,2]. Despite the distinct dynamics in ETI and PTI, they share common signaling components, such as Ca 2+ , various kinases, salicylic acid (SA), jasmonic acid (JA), and ethylene [3,4]. The immune signals accumulate in the nucleus, where they are integrated and translated by various transcription factors (TFs) to promote massive transcriptional reprogramming, thereby activating PTI and ETI [5]. As the transcription of immunity-related genes is largely determined by specific targeting of various TFs, these TFs are crucial for proper regulation of plant immunity to pathogens [6]. Several TF families, e.g. APETALA2 (AP2)/ethylene response factor (ERF), basic leucine zipper (bZIP)/TGACG motif-binding proteins (TGAs), basic helix-loop-helix (bHLH), WRKY, and NAC, control the activation of plant immunity [7,8].
WRKY TFs comprise one of the largest TF families; its members contain at least one conserved WRKY domain, which is a DNA-binding domain that preferably binds to the W-box motif (C/T)TGAC(C/T) in the promoter region of their target genes. WRKY TFs are divided into three distinct groups (I, II, and III) based on the number of WRKY domains and the amino acid sequence features of zinc-finger motifs [9]. WRKY TFs are involved in the regulation of plant immunity; in particular, the group III WRKY TFs positively or negatively modulate plant innate immunity [10,11]. In Arabidopsis (Arabidopsis thaliana), SA is sensed by two different classes of receptors, NONEXPRESSER OF PR GENES 1 (NPR1) and its homologs NPR3/NPR4, which function independently to modulate SA-mediated immunity through regulating WRKY70 (group III) expression [12,13]. During ETI, the group III WRKY54 and WRKY70, which are substrates of the NPR1-cullin-RING ligase 3 (CRL3) complex, promote SA-mediated cell survival [14]. Moreover, many WRKY members from different plant species are transcriptionally modified in response to pathogen attack; some positively [15,16] and others negatively [17] regulate plant immunity. Further, some WRKY TFs interact with each other at the transcriptional or post-translational level [16]. These functionally connected WRKY TFs may form transcriptional networks, with some having central roles in mediating rapid and efficient activation of host defense programs [18]. Nevertheless, how TFs are organized into networks and how they operate remain unclear.
Tomato (Solanum lycopersicum) is a major horticultural crop that has been adapted to a wide range of environmental conditions worldwide. Bacterial wilt, one of the most devastating soil-borne diseases, is caused by Ralstonia solanacearum, which constrains the production and geographical distribution of tomato [19,20]. Therefore, understanding how tomato responds to R. solanacearum infection (RSI) would facilitate the development of effective strategies to control this disease. Around 81-83 WRKY TFs are encoded in the tomato genome [21,22], among which SlWRKY3, SlWRKY31, SlWRKY33, SlWRKY39, SlWRKY45, SlWRKY72, SlWRKY73, SlWRKY74, and SlWRKY70 [22][23][24] function in plant immunity to biotic stresses, and SlWRKY23, SlWRKY46, SlWRKY53/SlWRKY54, SlWRKY80, and SlWRKY81 are transcriptionally modified during pathogen attack [22]. However, no SlWRKY TF has been implicated in the tomato response to RSI.
In this article, we investigated the behavior of the group III WRKY TFs during RSI and revealed that SlWRKY30 positively regulates tomato resistance to RSI. SlWRKY30 was strongly induced during RSI in tomato, and SlWRKY30-overexpressing tomato plants had increased resistance to RSI. Transcriptome analysis revealed that SlWRKY30 transcriptionally activates defenseresponse genes. In particular, expression of the pathogenesisrelated protein (SlPR-STH2) genes SlPR-STH2a, SlPR-STH2b, SlPR-STH2c, and SlPR-STH2d (hereafter SlPR-STH2a/b/c/d) was upregulated by SlWRKY30, which directly bound to the W-boxes in the SlPR-STH2a/b/c/d promoters. Furthermore, SlWRKY30 interacted with SlWRKY81 to synergistically activate SlPR-STH2a/b/c/d expression. Therefore, we revealed a genetic resource for breeding bacterial wilt-resistant tomato varieties.

Expression analysis of group III SlWRKY genes during R. solanacearum infection and exogenous application of salicylic acid
To identify group III SlWRKY genes potentially involved in the tomato response to RSI, we analyzed their transcriptional response to RSI by reverse transcription-quantitative PCR (RT-qPCR). Expression of SlWRKY30, 52, 59, 80, and 81 was significantly upregulated following RSI. In particular, expression of SlWRKY30 [12 hours post-inoculation (hpi)] and SlWRKY81 (6 hpi) was 5.2-and 5.8-fold higher than that of the other genes, respectively (Fig. 1).
Since SA is a crucial signaling molecule that is involved in plant defense against biotrophic [25] and hemibiotrophic pathogens, such as R. solanacearum [26], we therefore analyzed the response of the group III SlWRKY genes to exogenous SA application. SlWRKY30, 52, 59, 80, and 81 were upregulated by SA, similar to their response to RSI, while SlWRKY41, SlWRKY53, and SlWRKY54 were downregulated (Supplementary Data Fig. S1A). Next, we analyzed the expression of these SlWRKY genes in different tissues (stem, leaf, and root) under normal conditions using RT-qPCR and observed similar expression patterns (except for SlWRKY54), with relatively higher expression in the leaf and root than in the stem. Conversely, SlWRKY54 expression was higher in the stem than in the leaf and root (Supplementary Data Fig. S1B). Taken together, these results indicate that these eight group III SlWRKY TFs are involved in tomato immunity to RSI.

Subcellular localization and transcriptional activity of the eight group III SlWRKY transcription factors
The biological function of a TF is closely associated with its subcellular localization. We therefore investigated the subcellular localization of these eight group III SlWRKY TFs through transiently expressed SlWRKY-GFP (green f luorescent protein) fusion proteins in Arabidopsis mesophyll protoplasts. All eight SlWRKY-GFP fusion proteins localized exclusively to the nuclei, with a complete overlap with the red f luorescent protein (RFP) nuclear marker protein, indicating that these SlWRKY TFs function in the nucleus ( Fig. 2A, Supplementary Data Fig. S2A and B). Next, we analyzed the transcriptional activity of these SlWRKY proteins by a transcriptional activation assay in yeast strain AH109. The BD-SlWRKY fusions activated LacZ reporter gene expression (Supplementary Data Fig. S3A), indicating that these SlWRKY TFs have transcriptional activation activity. Protoplasts were incubated in WI buffer for 10 hours after transformation and imaged using a f luorescence microscope. Scale bar = 20 μm. (B) Phylogenetic tree of group III WRKY TFs from tomato, pepper, and Arabidopsis. Amino acid sequences labeled with red circles, dark blue diamonds, and inverted yellow triangles represent the group III WRKY TFs from tomato, pepper and Arabidopsis, respectively. The tree was constructed using MEGA 6.06.
Furthermore, we conducted a phylogenetic analysis to investigate the relationship among the group III SlWRKYs and their orthologs in pepper (Capsicum annuum), and constructed two unrooted phylogenetic trees based on amino acid sequence similarities using the neighbor-joining method. SlWRKY30 shared highest sequence similarity with CaWRKY41 (Fig. 2B, Supplementary Data Fig. S3B), which confers pepper resistance to RSI [42]. Together, these results indicate the eight group III SlWRKY members are TFs and may function in the tomato response to RSI.

SlWRKY30 positively regulates tomato resistance to R. solanacearum infection
Among the group III SlWRKY genes, SlWRKY30 was upregulated by RSI and SA treatment in tomato. SlWRKY30 was also induced by JA and ACC (the precursor of ethylene) (Supplementary Data Fig. S4A). Therefore, we considered SlWRKY30 a strong candidate gene for a role in the tomato immune response. To study the role of SlWRKY30 in tomato immunity to RSI, we generated nine independent SlWRKY30-overexpressing T 2 or T 3 tomato transgenic lines, which had higher SlWRKY30 expression levels compared with wild-type (WT) plants ( Supplementary Data Fig. S4B). We selected two independent lines (OE6 and OE8) for an RSI assay. OE6 and OE8 exhibited similar growth and development phenotypes to WT plants under normal conditions. Following RSI, all of the tomato plants exhibited wilt symptoms (Fig. 3A), but the SlWRKY30-overexpressing lines showed decreased susceptibility compared with WT plants at 4 days post-inoculation (dpi), as evidenced by a lower disease index and growth rate of R. solanacearum  Fig. S5C and D). Together, these results demonstrate that SlWRKY30 positively modulates tomato resistance to RSI.
R. solanacearum exclusively invades plant roots [19,20], colonizes the root cortex, and spreads through xylem vessels to systemically infect the whole plant [27]. To test if SlWRKY30 is expressed in vascular tissues to induce an immune response, we generated pSlWRKY30:GUS reporter lines in tomato and analyzed the SlWRKY30 tissue-specific expression pattern in T 2 plants upon RSI or SA treatment. Under normal conditions, we observed βglucuronidase (GUS) activity mainly in the leaf veins and root stele but not in the cortex, while both RSI and SA significantly promoted GUS activity (

SlPR-STH2
To identify genes regulated by SlWRKY30 during the tomato response to RSI, we compared the transcriptome profiles of the OE6 line and WT plants at 24 hpi with R. solanacearum by transcriptome sequencing (RNA-seq). We identified the differentially expressed genes (DEGs) and filtered them based on the criteria of fold change ≥2 and false discovery rate <.01. The expression of 304 genes was significantly affected, among which 208 (F) GUS expression in pSlWRKY30:GUS tomato shoots and roots that were infected with R. solanacearum or received exogenous application of SA for 24 h. Scale bar = 150 μm. (G) GUS activity in pSlWRKY30:GUS tomato shoots and roots that were infected with R. solanacearum or received exogenous application of SA for 24 h. Relative GUS activity in treated plants was compared with activity in control plants, which was set to 1. Data represent the mean ± standard error.
were upregulated and 96 were downregulated ( Fig. 4A and B, Supplementary Data Table S2). The most enriched term in Gene Ontology (GO) enrichment analysis was 'defense response', which involved 12 genes (Fig. 4A, Supplementary Data Table S3), all of which had increased expression in OE6 (Fig. 4C).
Next, we performed transient expression assays in Nicotiana benthamiana leaves using dual-luciferase reporter plasmids (Supplementary Data Fig. S6B). Compared with the empty vector (p62-SK) control, co-expression of SlWRKY30 driven by the CaMV35S promoter with LUC driven by the SlPR-STH2a/b/c/d promoters enhanced LUC reporter gene expression and LUC luminescence intensity ( Fig. 5A and B). Furthermore, we identified W-boxes, which are preferentially bound by WRKY TFs [9], in the SlPR-STH2a/b/c/d promoters (Fig. 5C), implying that these SlPR-STH2 genes are directly activated by SlWRKY30. To test this, we performed an electrophoretic mobility shift assay (EMSA) using MBP (maltose-binding protein)-WRKY30 fusion protein and DNA probes (WT and mutant) designed from the SlPR-STH2a/b/c/d promoters containing the W-box. MBP-SlWRKY30, but not MBP alone, directly bound to DNA probes 1 and 2 (WT), were repressed by the competitors (Fig. 5D). Moreover, SlWRKY30 does not bind to DNA probes, in which the TGAC motifs were mutated to TaAa (Supplementary Data Fig. S6C). Together, these results indicate that SlWRKY30 positively regulates tomato immunity to RSI by directly activating SlPR-STH2a/b/c/d expression.
SlWRKY30 interacts with SlWRKY52, 59, 80, and 81 during the response to R. solanacearum infection WRKY TFs may be post-translationally modulated by interacting with regulatory proteins, such as themselves or other WRKY TFs, to cooperatively or antagonistically regulate gene transcription [16,31]. However, few studies have investigated the interacting partners of group III WRKY TFs in plant immunity, especially in non-model plants. To explore the interactions among group III SlWRKY TFs in the tomato response to RSI, we performed a yeast two-hybrid (Y2H) assay between SlWRKY30 and itself and the other seven group III WRKY TFs investigated in this study (SlWRKY41, SlWRKY52, SlWRKY53, SlWRKY54, SlWRKY59, SlWRKY80, and SlWRKY81). SlWRKY30 interacted with SlWRKY52, 59, 80, and 81 in yeast. Notably, SlWRKY30 did not form a homodimer with itself, and did not interact with SlWRKY41, 53, and 54 in yeast (Fig. 6A). To validate these results, we performed a bimolecular f luorescence complementation (BiFC) assay in Arabidopsis mesophyll protoplasts. Indeed, SlWRKY30 interacted with SlWRKY52, 59, 80, and 81 (Fig. 6B, Supplementary Data Fig. S7A), but not with itself or SlWRKY41, 53, or 54 ( Supplementary Data Fig. S7B). Consistent with this, we observed protein-protein interactions between SlWRKY30 and SlWRKY52, 59, 80, and 81 in N. benthamiana leaves by a split luciferase complementation imaging (LCI) assay (Fig. 6C), but did not observe interactions between SlWRKY30 and itself, SlWRKY41, SlWRKY53, or SlWRKY54 (Supplementary Data Fig. S7C). To further confirm the interaction of SlWRKY30 and SlWRKY52, 59, 80, and 81, we performed coimmunoprecipitation (Co-IP) assays using GFP-fused SlWRKY30 and HA-fused SlWRKY52, 59, 80, and 81. Our data showed that SlWRKY30 interacted with SlWRKY52, 59, 80, and 81 in N. benthamiana leaves (Fig. 6D). Collectively, these data demonstrate that SlWRKY30 interacts with SlWRKY52, 59, 80, and 81. As all of their corresponding genes were upregulated by RSI in tomato, we speculated that these interactions may modulate SlWRKY30 function during the tomato response to RSI.

SlPR-STH2a/b/c/d
To test if SlWRKY30 is modulated by SlWRKY52, 59, 80, and 81 during the response to RSI, we investigated the SlWRKY30-SlWRKY81 interaction, because SlWRKY81 was most strongly induced in tomato plants after 6 hpi with R. solanacearum (Fig. 1). First, to determine the function of SlWRKY81 in tomato immunity to RSI, we generated SlWRKY81-silenced tomato plants by VIGS using the TRV:Slwrky81 construct, and tested the silencing efficiency by RT-qPCR. SlWRKY81 was downregulated in TRV:Slwrky81 plants compared with plants transformed with the empty vector control (TRV:00) (Supplementary Data Fig. S8A). SlWRKY81 silencing enhanced tomato susceptibility to RSI compared with the empty vector control plants (Supplementary Data Fig. S8B), exemplified by an increased disease index and R. solanacearum growth rate (Supplementary Data Fig. S8C and D). These results indicate that SlWRKY81 positively regulates tomato immunity to RSI, and that SlWRKY30 and SlWRKY81 might synergistically regulate the expression of SlPR-STH2a/b/c/d. To validate this, we performed dual-luciferase assays with SlWRKY81 in N. benthamiana leaves. Similar to the results with SlWRKY30, transient overexpression of SlWRKY81 driven by the CaMV35S promoter significantly promoted the expression of LUC driven by the SlPR-STH2a/b/c/d promoters (Fig. 6E and 6F), indicating that SlWRKY81 positively regulates tomato immunity to RSI by cooperating with SlWRKY30 to regulate SlPR-STH2a/b/c/d expression.

SlWRKY30 and SlWRKY81 directly and synergistically regulate SlPR-STH2a/b/c/d
Since SlWRKY30 and SlWRKY81 interact, and both positively regulate tomato immunity to RSI by regulating SlPR-STH2a/b/c/d expression, and because SlPR-STH2a/b/c/d are directly targeted by SlWRKY30, we speculated that SlWRKY30 and SlWRKY81 synergistically and directly regulate SlPR-STH2a/b/c/d expression.
To test this, we performed an EMSA with the MBP-SlWRKY81 fusion protein and SlPR-STH2a/b/c/d promoter DNA probes. MBP-SlWRKY81, but not MBP alone, directly bound to DNA probes 1 and 2, were repressed by the competitors (Supplementary Data Fig. S9A and B), indicating that SlWRKY81 directly regulates SlPR- To examine if the SlWRKY30-SlWRKY81 interaction affects SlWRKY30 or SlWRKY81 targeting to SlPR-STH2a/b/c/d, we performed another EMSA. Both MBP-SlWRKY30 and MBP-SlWRKY81 bound to DNA probe 1, and the mobility of MBP-SlWRKY30-bound DNA probe 1 was altered by addition of MBP-SlWRKY81, while addition of MBP alone did not alter mobility (Fig. 7A), indicating that SlWRKY30 interacts with SlWRKY81 and they synergistically bind the SlPR-STH2a/b/c/d promoters. In addition, transiently expressed SlWRKY30 or SlWRKY81 alone activated the expression of LUC driven by the SlPR-STH2a/b/c/d promoters in N. benthamiana leaves (Fig. 7B). However, co-expression of SlWRKY30 and SlWRKY81 significantly increased LUC expression compared with SlWRKY30 or SlWRKY81 alone, indicating that SlWRKY81 promotes SlWRKY30mediated transactivation of the SlPR-STH2a/b/c/d promoters. Collectively, these results demonstrate that SlWRKY30 interacts with SlWRKY81 and that they directly and synergistically activate the expression of SlPR-STH2a/b/c/d.

Discussion
Bacterial wilt, caused by R. solanacearum, is one of the most devastating diseases in tomato. WRKY TFs are implicated in plant immunity to various pathogens, but their functions in immunity to R. solanacearum in tomato remain poorly understood. Here we report that several group III SlWRKY TFs participate in tomato immunity to RSI; in particular, SlWRKY30 positively regulates immunity, and its function is modulated by other group III SlWRKY TFs through protein-protein interactions. This article provides novel insight into the molecular mechanisms whereby group III SlWRKY TFs regulates immunity, as well as candidate genes for improving tomato resistance to RSI.

SlWRKY30 and several other group III SlWRKY transcription factors participate in tomato immunity to R. solanacearum infection
Genes that are transcriptionally modified upon pathogen attack are often implicated in plant immunity, which is the rationale for many immunity-related transcriptome analyses. Moreover, many group III SlWRKY TFs participate in plant immunity to various pathogens [10,14]. Therefore, we explored the transcriptional changes of group III SlWRKY TF genes using RT-qPCR upon RSI in tomato, as these TFs might participate in tomato immunity to RSI. Five group III SlWRKY genes, particularly SlWRKY30 and SlWRKY81, were significantly induced by RSI (Fig. 1), and SlWRKY30 and SlWRKY81 increased tomato immunity to RSI (Fig. 3A-C, Supplementary Data Figs S5 and S8), indicating that several group III SlWRKY TFs positively regulate tomato resistance to bacterial wilt. In addition, phylogenetic analysis revealed that the homolog of SlWRKY30 in pepper is CaWRKY41 (Fig. 2B, Supplementary Data Fig. S3B), which confers R. solanacearum resistance in pepper. Considering that tomato and pepper are attacked by similar pathogens, such as R. solanacearum, the resistance loci containing SlWRKY30 and CaWRKY41 may have been selected during tomato and pepper domestication, respectively. The induction of SlWRKY30 in leaf veins and the root stele during RSI further supports its role as a positive regulator of tomato immunity (Fig. 3F, Supplementary Data Fig. S4C), since R. solanacearum exclusively invades roots [19,20] and then spreads through the vasculature [27]. Furthermore, the positive role of SlWRKY30 in tomato immunity might be due to regulation of the H 2 O 2 accumulation and hypersensitive response (HR) cell death ( Fig. 3D and E), as H 2 O 2 promotes HR cell death [32]. HR cell death frequently accompanies ETI [2,33], and sometimes also PTI [34]; therefore, SlWRKY30 may function in ETI and possibly also PTI. In addition, SlWRKY30 and several other group III SlWRKY genes were upregulated by exogenous application of SA (Supplementary Data Fig. S1A). SA signaling is involved in the RSI response in pepper [26], and HR cell death is also promoted by SA signaling [35], indicating that SlWRKY30 and other group III SlWRKY TFs might function in an SA-dependent manner.

SlWRKY30 functions in tomato immunity to R. solanacearum infection by directly targeting and regulating SlPR-STH2 genes
The RNA-seq data of SlWRKY30-overexpressing tomato lines indicated that SlWRKY30 functions by upregulating multiple genes involved in a broad range of processes, such as 'defense response', 'chitin binding', and 'chitin catabolic process', and also by downregulating processes such as 'chloroplast thylakoid membrane', 'photosystem II', and 'chlorophyll binding' (Fig. 4A and B),  indicating that SlWRKY30 overexpression activated immunityrelated genes and repressed genes related to growth and other biological processes, to ensure mobilization of resources for the immune response. Defense-response genes were enriched in the GO analysis of the SlWRKY30-overexpression line, including several SlPR-STH2 genes (SlPR-STH2a, b, c, and d) that exist as a gene cluster (Fig. 4C, Supplementary Data Fig. S6A). Some SlPR-STH2 orthologs in pepper are activated by RSI and positively regulate pepper immunity [26]. These SlPR-STH2 genes were also directly activated by SlWRKY81, which also positively regulated tomato immunity ( Fig. 6D and E, Supplementary Data Figs S8 and S9), indicating that these PR genes have important roles in tomato resistance to bacterial wilt. Furthermore, genes in the abscisic acid (ABA) signaling pathway were also activated by SlWRKY30. Studies on the role of ABA in plant immunity have yielded conf licting results. For example, ABA signaling positively regulates Arabidopsis resistance to bacterial wilt [36] but negatively regulates pepper immunity to RSI under room temperature. Therefore, the biological relevance of SlWRKY30-mediated upregulation of ABA signaling genes in tomato during the response to RSI needs further investigation.

SlWRKY30 is post-translationally modified by other group III SlWRKY transcription factors
We found that eight group III SlWRKY TFs were transcriptionally modulated by RSI in tomato. Among them, SlWRKY30, 52, 59, 80, and 81 were upregulated by RSI and by exogenous application of SA, which is implicated in plant immunity to hemibiotrophic pathogens such as R. solanacearum [26]. Y2H, BiFC, LCI, and Co-IP analyses revealed that SlWRKY30 interacts with SlWRKY81 (Fig. 6), which also positively regulated tomato immunity to RSI (Supplementary Data Fig. S8), SlWRKY52, 59, and 80, indicating that these TFs and their interaction are involved in tomato immunity to RSI. Protein-protein interactions between group III WRKY TFs are not limited to plant immunity, since Arabidopsis AtWRKY30 interacts with AtWRKY53, 54, and 70, which negatively regulate leaf senescence in Arabidopsis [37]. Moreover, interactions among group IIa WRKY members AtWRKY18, CaWRKY40, and AtWRKY60 are involved in basal defense [38], indicating that WRKY interactions allow for plasticity in their regulation of different biological processes. We revealed that the promoters of SlPR-STH2a/b/c/d are targeted and regulated by SlWRKY30 and SlWRKY81 via two closely spaced W-boxes (Fig. 7). The interaction between SlWRKY30 and SlWRKY81 may further promote the transcriptional expression of their shared target genes over that of either TF alone, as demonstrated in our dual-luciferase assay, which is distinct from the interactions between CaWRKY17 and CaWRKY40 and between CaWRKY27b and CaWRKY40, in which both CaWRKY17 and CaWRKY27b function by physically interacting with CaWRKY40 to promote CaWRKY40 binding to and activation of its immunity-related target genes [16,39]. Although SlWRKY52, 59, and 80 also interacted with SlWRKY30, their precise roles in the tomato immune response to R. solanacearum need further study. SlPR-STH2a/b/c/d share ∼70% identity in their deduced amino acid sequences, and form a SlPR-STH2 tandem duplication (Supplementary Data Fig. S6A). In soybean (Glycine max), the Rps11 locus, which harbors a cluster of large NLR genes from a single origin and results in promoter fusion and leucine-rich repeat (LRR) expansion, confers broad-spectrum resistance to Phytophthora sojae [40]. However, whether the tandem duplication of the SlPR-STH2 genes is responsible for the increased tomato resistance to R. solanacearum needs more study. Moreover, it will be interesting to elucidate the molecular mechanism by which the SlWRKY30-SlWRKY81 module promotes the SA-mediated signaling pathway.

Conclusions
Discovery of genes that confer resistance to RSI is crucial to preventing bacterial wilt outbreaks in tomato production. We identified two group III SlWRKY TFs, SlWRKY30 and SlWRKY81, that were upregulated by RSI and positively regulated tomato immunity by directly targeting and regulating SlPR-STH2a/b/c/d. The function of SlWRKY30 might be modulated via proteinprotein interactions with SlWRKY52, 59, 80, and 81. Based on these results, we propose a model of the mechanism by which SlWRKY30 regulates immunity to RSI (Fig. 8).

Plant material and growth conditions
The tomato S. lycopersicum L. cv. Micro-Tom was used as the WT genotype for all tomato lines generated in the experiments. The recombinant SlWRKY30-OE and pSlWRKY30-GUS plasmids were individually transformed into Agrobacterium tumefaciens (strain GV3101), and the WT tomato plants were transformed through A. tumefaciens-mediated infection. Plants were grown on media to select for hygromycin B resistance, and nine SlWRKY30-OE and eight pSlWRKY30:GUS independent transgenic lines were generated and screened using PCR/RT-qPCR analysis and GUS assay, respectively. Then, the SlWRKY30-OE6 (∼14-fold) and SlWRKY30-OE8 (∼12-fold) T 2 or T 3 seeds were used for the experiments. The tomato seeds were germinated at 25 ± 2 • C in darkness for 48 hours on moistened sterile filter paper, then sown in a steamsterilized soil mixture, which included peat moss and vermiculite (2:1 by volume). Subsequently, tomato seedlings were transferred to a temperature (25 ± 2 • C)-controlled growth room under a light intensity of ∼100 μmol photons m −2 s −1 and a 16-hour light/8hour dark cycle.

Molecular cloning and plasmid construction
To generate the SlWRKY30-OE construct, the coding sequence (CDS) of full-length SlWRKY30 was PCR-amplified from tomato cDNA, cloned into the pDONR207 vector, then recombined with the binary vector pGWB2 [41]. To generate the pSlWRKY30:GUS construct, the 2.0-kb SlWRKY30 promoter was PCR-amplified, cloned into pDONR207, and recombined with the binary vector pMDC163 [42]. To generate the DNA constructs for transcriptional activity and subcellular localization analyses, the CDSs of SlWRKY30, 41,52,53,54,59,80, and 81 were PCR-amplified from tomato cDNA and cloned into vectors pGBKT7 and 35S:GFP (35S-CDS-NOS Terminator) [43]. PCR primers used in the DNA constructs are listed in Supplementary Data Table S1.

Pathogens and inoculation procedures
For the pathogenicity test, 5-week-old SlWRKY30-OE6, SlWRKY30-OE8, and WT tomato plants were infected with a suspension of R. solanacearum (strain FJ190401). The mean number of colony forming units (CFU) of five leaves from various tomato genotypes was recorded at 0, 24, and 48 hpi. Then, the disease index (from 0 to 5) of the R. solanacearum-infected tomato plants was scored daily as described previously [42].

Subcellular localization and transcriptional activity analyses
For subcellular localization, the 35S-SlWRKY-GFP plasmids of the different group III WRKYs were isolated and purified as described previously [44]. Subcellular localization of SlWRKY-GFP fusion proteins was conducted in A. thaliana mesophyll protoplasts (200 μl) transfected with 20 μg plasmid DNA and incubated in WI buffer [45] for 10 hours. Then GFP f luorescence was observed using a f luorescence microscope (Carl Zeiss, Germany).
For transcriptional activity analysis, the BD-SlWRKY constructs were transformed into AH109 yeast cells and grown on SD/−Trp medium at 30 • C for 48-60 hours, and then the positive transformants were transferred to SD/−Trp medium containing X-α-Gal (5-bromo-4-chloro-3-indolyl-α-d-galactopyranoside) as substrate for blue color development.

Virus-induced gene silencing
For VIGS assays [46], the SlWRKY30 and SlWRKY81 gene fragments were identified by BLAST searching in the tomato genome sequences. These fragments were cloned into a tobacco rattle virus VIGS vector (pYL-279) using the Gateway system (Invitrogen). Vectors TRV:Slwrky30, TRV:Slwrky81, and TRV:00 (empty vector) were transformed into A. tumefaciens, respectively; subsequently, they were mixed with A. tumefaciens harboring pYL-192 at a 1:1 ratio, their concentration was adjusted to OD 600 = 0.5, and then they were injected into fully expanded tomato seedling leaves.

Histochemical staining
For 3,3 -diaminobenzidine (DAB) and trypan blue staining, 4week-old tomato plants were inoculated with a suspension of R. solanacearum. At 48 hpi, leaves were harvested and stained with DAB or trypan blue solution, as described previously [42].
For GUS staining, leaves and root tissues were immersed for 12 hours in GUS staining solution, as described previously [41], at 37 • C. Then, the leaves were destained several times using 75% (v/v) ethanol. The samples were photographed with a stereo microscope (Leica, Germany).

Transient expression dual-luciferase assay
To generate the pSlPR-STH2ap:LUC, pSlPR-STH2bp:LUC, pSlPR-STH2cp:LUC, and pSlPR-STH2dp:LUC constructs, the SlPR-STH2a/b/c/d promoter fragments were cloned into pGreenII-0800-LUC vector [47], which was digested with HindIII-and BamHI. To generate the p62-SlWRKY30 and p62-SlWRKY81 constructs, the SlWRKY30 and SlWRKY81 CDSs were inserted into pGreenII 62-SK vector [47], which was digested with BamHI and HindIII. The transient expression dual-luciferase assay was performed in N. benthamiana leaves as described previously [47]. To determine the LUC/REN ratio, the values of LUC reporter expression were normalized to the values of REN luciferase expression, which was driven by the 35S promoter.

Electrophoretic mobility shift assay
According to the EMSA protocol, the SlWRKY30 and SlWRKY81 CDSs were amplified by PCR using primers MBP-SlWRKY30-F/R and MBP-SlWRKY81-F/R and cloned into pMAL-c4X (NEB, http://www.neb-china.com/). Then, MBP-SlWRKY30 and MBP-SlWRKY81 fusion proteins were induced at 16 • C for 16-20 hours by 0.4 mM β-d-1-thiogalactopyranoside (IPTG), and purified with Amylose Resin (NEB). The concentration of MBP-SlWRKY30 and MBP-SlWRKY81 was determined by bovine serum albumin (BSA) quantitative analysis. Then, 2 μg of purified MBP-SlWRKY30 or MBP-SlWRKY81 was incubated with Cy5-labeled probes in 20 μl reaction mixtures at 25 • C for 30 minutes and then separated in Tris-glycine buffer with 12% native polyacrylamide gels. Unlabeled WT probes were used as competitors. The gel image was visualized through a LI-COR Odyssey Infrared Imaging System (LI-COR, USA) to detect the Cy5labeled probes. The sequences of Cy5-labeled probes are listed in Supplementary Data Table S1.

RNA sequencing
Four-week-old WT and SlWRKY30-OE (OE6) plants were infected using R. solanacearum suspensions. At 24 hpi, samples were harvested from three biological replicates, and total RNA was extracted using the TRIzol kit (Invitrogen) following the user manual. RNA quality was tested by a NanoDrop spectrophotometer, and verified by an Agilent 2100 Bioanalyzer. Then, cDNA libraries were constructed for sequencing by Biomarker Technologies (China). The DEGs were identified and filtered based on the following criteria: fold change ≥2 and false discovery rate <.01. The raw sequencing reads (RNA-seq data) have been deposited in the China National GeneBank DataBase (CNGBdb, https://db.cngb. org/cnsa/) under accession number CNP0003958. A summary of the sequencing data is shown in Supplementary Data Table S2. The top 20 enriched functions of the DEGs in the RNA-seq data sets were selected by GO term enrichment analysis. A summary of the analysis is shown in Supplementary Data Tables S3 and S4.

Gene expression analysis
Gene expression data were analyzed using RT-qPCR, as described previously [42]. In brief, total RNA was extracted and then used as a template to synthesize cDNA using TaKaRa PrimeScript RT-PCR Kit (TaKaRa) following the user manual. Quantitative PCR was conducted as previously described [42]. The 2 -Ct method was used to calculated the relative expression of target genes, and normalized to that of the tomato internal reference gene SlACTIN2. For each sample, three independent biological replicates were performed.

Statistical analysis
Differences were analyzed by Student's t test (two groups) or Tukey's multiple comparisons test (four groups). The significances of differences were determined using the P value. Statistically significant differences are indicated in the figures with asterisks ( * P < .05 or * * P < .01) or different letters (lowercase letters, P < .05; uppercase letters, P < .01, Tukey's multiple comparisons test). All experiments were performed and analyzed with at least three biological replicates.

Data availability
All relevant data supporting our findings are available in the manuscript file or from the corresponding author upon request.