Xanthomonas campestris VemR enhances the transcription of the T3SS key regulator HrpX via physical interaction with HrpG

Abstract VemR is a response regulator of the two‐component signalling systems (TCSs). It consists solely of a receiver domain. Previous studies have shown that VemR plays an important role in influencing the production of exopolysaccharides and exoenzymes, cell motility, and virulence of Xanthomonas campestris pv. campestris (Xcc). However, whether VemR is involved in the essential pathogenicity determinant type III secretion system (T3SS) is unclear. In this work, we found by transcriptome analysis that VemR modulates about 10% of Xcc genes, which are involved in various cellular processes including the T3SS. Further experiments revealed that VemR physically interacts with numerous proteins, including the TCS sensor kinases HpaS and RavA, and the TCS response regulator HrpG, which directly activates the transcription of HrpX, a key regulator controlling T3SS expression. It has been demonstrated previously that HpaS composes a TCS with HrpG or VemR to control the expression of T3SS or swimming motility, while RavA and VemR form a TCS to control the expression of flagellar genes. Mutation analysis and in vitro transcription assay revealed that phosphorylation might be essential for the function of VemR and phosphorylated VemR could significantly enhance the activation of hrpX transcription by HrpG. We infer that the binding of VemR to HrpG can modulate the activity of HrpG to the hrpX promoter, thereby enhancing hrpX transcription. Although further studies are required to validate this inference and explore the detailed functional mechanism of VemR, our findings provide some insights into the complex regulatory cascade of the HpaS/RavA‐VemR/HrpG‐HrpX signal transduction system in the control of T3SS.


| INTRODUC TI ON
Bacteria have developed various regulatory mechanisms to precisely control the expression of varied genes involved in different physiological processes in response to environmental changes. Twocomponent signalling systems (TCSs) are a major mechanism widely adopted by bacteria to sense environmental changes and modulate subsequently the expression of related genes (Buschiazzo & Trajtenberg, 2019;Gao et al., 2019). TCSs are composed of at least a histidine kinase (HK) sensor and a response regulator (RR). Generally, a membrane-associated HK sensor carries out autophosphorylation on a conserved histidine residue in the transmitter domain on sensing a specific environmental stimulus. The phosphoryl group is then transferred to the conserved aspartate residue in the receiver (REC) domain of a cognate RR, leading to a conformational change and the activation of its output domain, which in turn regulates the downstream targets (Buschiazzo & Trajtenberg, 2019;Gao et al., 2019).
Notably, some RRs contain only a phosphoryl-accepting REC domain but lack a dedicated output domain. Genomic analysis reveals that these RR proteins, referred to as single-domain response regulators (SD-RRs), are widespread in bacteria, constituting the second largest class of RR proteins (about 23%) (Gao et al., 2019). Given that the SD-RRs do not have an output domain such as a DNA-binding domain, they are thought to act by directly interacting with their downstream protein targets and allosterically regulating the activity of the targets. However, only a few SD-RRs have been studied for their cellular functions. So far, the best characterized SD-RR is the chemotactic protein CheY. The Escherichia coli CheY and its homologues in a number of other bacteria have been shown to regulate cell swimming motility via protein-protein interaction. When CheY is phosphorylated by the phosphoryl group from the phosphorylated HK sensor CheA, the phosphorylated CheY binds to FliM and FliN, the components of the flagellar motor switch complex, to modulate the flagellar motor rotation (Parkinson et al., 2015;Sarkar et al., 2010).
Xanthomonas campestris pv. campestris (Xcc) is a gram-negative bacterial phytopathogen that infects almost all of the cruciferous plants and causes a threatening disease (black rot) to brassica crops worldwide (Vicente & Holub, 2013). The virulence of Xcc toward hosts depends on a number of pathogenicity factors, including lipopolysaccharides, exopolysaccharides, exoenzymes (such as amylase, cellulase, protease, and pectate lyase) secreted by the type II secretion system (T2SS), effectors (T3Es) secreted by the type III secretion system (T3SS), cell motility, and biofilm .
The Xanthomonas T3SSs and their regulation have been subjected to extensive studies. It is clear that, like many other Xanthomonas T3SSs, the Xcc T3SS apparatus is encoded by the hrp (hypersensitive response and pathogenicity) gene cluster, which consists primarily of six operons (hrpA-hrpF) and more than 20 hrp or hrc (hrp-conserved) genes (Arlat et al., 1991;Gürlebeck et al., 2006;Huang et al., 2009;Qian et al., 2005). The activation of Xcc hrp/ hrc genes is mainly controlled by three key regulators, HpaS, HrpG, and HrpX (Alvarez-Martinez et al., 2020;An et al., 2020). Disruption of hpaS, hrpG or hrpX in Xcc resulted in loss of virulence in host plants and hypersensitive response (HR) in nonhost plants (Huang et al., 2009;Li et al., 2014). HrpX is an AraC-type transcriptional regulator. It induces the transcription of the hrp/hrc genes and some T3E-encoding genes by directly binding to their promoters (Huang et al., 2009;Koebnik et al., 2006;. HpaS is a membrane-bound HK sensor. It forms a TCS with HrpG, an OmpR family DNA-binding regulator, which activates the expression of hrpX (Li et al., 2014).
Interestingly, in addition to HrpG, HpaS can also co-opt the SD-RR VemR to form another TCS (Li et al., 2020). VemR is involved in extracellular polysaccharide (EPS) production, cell motility, and virulence of Xcc (Tao & He, 2010). It has been shown that after being phosphorylated by HpaS or RavA, VemR directly interacts with the flagellum rotor protein FliM or the flagellum biosynthesis regulator FleQ to control cell motility (Li et al., 2020;Lin et al., 2022). Recently, we found that the Xcc VemR is a global regulator that plays an important role in modulating the expression of the T3SS. Here, we demonstrate that VemR positively regulates the expression of the hrp gene cluster via HrpG and present evidence showing that VemR physically interacts with HrpG and enhances the transcriptional activation of hrpX by HrpG.

| VemR is a global regulator modulating the expression of a large number of genes involved in various cellular processes
VemR is encoded by the open reading frame (ORF) XC_2252 in the genome of the Xcc wild-type strain 8004 (Qian et al., 2005). Previous studies showed that deletion of vemR resulted in a severe reduction in virulence, EPS production, and cell motility (Li et al., 2020;Tao & He, 2010). To gain a better understanding of the regulatory role of VemR in various cellular processes of Xcc, we determined the transcriptome of the vemR deletion mutant strain ΔvemR by RNAsequencing (RNA-Seq). The mutant strain as well as the wild-type strain 8004 (Table S1) were grown to the mid-exponential phase (OD 600 = 0.6) in nutrient-yeast-glycerol (NYG), a medium widely used in physiologic studies of Xcc (Daniels et al., 1984). Total RNA was extracted from two independent biological replicates for each strain and analysed by RNA-Seq as described previously (Cui et al., 2018).
The results displayed that out of the 4617 annotated protein-coding genes in the genome of the Xcc strain 8004 (Luneau et al., 2022;Qian et al., 2005), 438 were differentially expressed by two-fold or more in the vemR deletion mutant (Table S2). Among these 438 differentially expressed genes (DEGs), 193 were up-regulated and 245 were down-regulated ( Figure 1a and Table S2). To verify the transcriptome changes, 15 DEGs related to virulence and adaption were selected to quantify their expression level using reverse transcription-quantitative real-time PCR (RT-qPCR). The total RNAs used for RT-qPCR were prepared from the bacterial cells grown in the same conditions as those for RNA-Seq. The RT-qPCR results showed that the transcriptional levels of all the tested genes were obviously changed in the mutant ΔvemR compared to the wild type, and the expression patterns of these selected genes are all consistent with that observed in the data obtained from the transcriptome analysis (Figure 1b and Table S2).
To get an insight into the functions of the genes regulated by VemR, we conducted a functional clustering analysis according to the genome annotation of the Xcc strain 8004 (He et al., 2007;Qian et al., 2005). Based on the clusters of orthologous groups (COG), out of the 438 DEGs in ΔvemR mutant, 272 were assigned to 15 various functional categories and 166 were predicted to encode hypothetical proteins or proteins that have not been given a functional category ( Figure 1a and Table S2). The dominant functional categories are "pathogenicity and adaption" and "cellular processes". In total, 57 and 52 genes were assigned to these two categories, respectively ( Figure 1a). Notably, 105 genes were assigned to "cell envelope and cell structure" (36), "transport" (33), "translation" (21), and "energy and carbon metabolism" (15) ( Figure 1a).
Consistent with the finding that VemR regulates cell motility (Li et al., 2020;Lin et al., 2022;Tao & He, 2010), the identified transcriptional profiles reveal that VemR has a crucial impact on a number of genes that contribute to cell motility. All of the 52 DEGs assigned to the category "cellular processes" are related to chemotaxis and flagella-dependent motility (Table S2). Additionally, out of the 22 down-regulated genes assigned to cell envelope and cell structure, 18 (XC_0937, XC_0938, XC_0939, XC_0940, XC_0941, XC_1059, XC_1183, XC_1184, XC_1185, XC_1186, XC_1187, XC_1358, XC_1359, XC_1621, XC_1622, XC_1624, XC_1626, and XC_3823) are involved in the type IV pilus biogenesis/fimbrial assembly. Given that in xanthomonads and other plant-pathogenic bacteria the type IV pili were reported to play a role in cell motility (Köhler et al., 2000;Mhedbi-Hajri et al., 2011), the motility deficiency in the ΔvemR mutant might be partially due to the reduced pilus biogenesis. It has been shown that the mutant ΔvemR produced less type II-secreted plant cell wall-degrading enzymes, such as extracellular amylase, cellulase, and protease, relative to the wild-type strain (Tao & He, 2010).
Scanning revealed that the DEGs assigned to the category "pathogenicity and adaption" include two genes (XC_0741 and XC_0742) encoding T2SS components and eight genes (XC_0125, XC_0126, XC_0783, XC_1120, XC_2483, XC_3379, XC_3590, and XC_3591) encoding extracellular amylase, cellulase or protease, which were all down-regulated in vemR mutant (Table S2). This result supports that VemR also plays a role in the regulation of exoenzyme production in Xcc.
Importantly, in addition to the genes involved in bacterial cell motility and T2SS-secreted enzyme production, which are known to be regulated by VemR, the transcriptome analysis uncovered that VemR also modulates the expression of genes encoding the T3SS, T4SS (type IV secretion system), and TonB-dependent receptors.
VemR acts as a global regulator affecting a number of genes involved in a variety of cellular processes cultured in nutrientyeast-glycerol (NYG) medium. (a) Functional categories of the 438 differentially expressed genes (DEGs) in the vemR-deletion mutant of Xanthomonas campestris pv. campestris (Xcc). The transcriptomes of Xcc strains cultured in NYG medium were investigated by RNA-Seq. Among the 438 DEGs, 193 and 245 were up-regulated and down-regulated, respectively. These genes were broadly categorized according to their biological function. (b) Reverse transcription-quantitative PCR assay of the expression level of several virulence-related genes regulated by VemR in Xcc strains. RNA was isolated from the cultures of the Xcc wild-type strain 8004 and the vemR deletion mutant ΔvemR grown in NYG medium to a concentration of OD 600 of 0.6. Relative gene expression with respect to the corresponding transcript levels in the wild-type strain was calculated. Values given are the means ± SD of triplicate measurements from a representative experiment. Differences were evaluated using Student's t test (**p < 0.01). Genes were considered to be differentially expressed if |log 2 (fold change)| ≥ 1 compared to the wild type. Similar results were obtained in two other independent experiments XC_1637/virB2, XC_1638/virB3, XC_1639/virB4, and XC_2016/virB6) ( Table S2). All of these vir genes were down-regulated in the vemR mutant. The T4SS is a large multiprotein nanomachine composed of a core set of 12 different proteins (VirB1 to VirB11 plus VirD4) in many gram-negative bacterial species. Xanthomonads use the T4SS to kill other bacterial cells in a contact-dependent manner, thereby obtaining a growth advantage in interbacterial competition (Souza et al., 2015). Fourteen of the DEGs encode TonB-dependent receptors: XC_0124, XC_0756, XC_0759, XC_1165, XC_1222, XC_1241, XC_1546, XC_1644, XC_2296, XC_2484, XC_2485, XC_2983, XC_3209, and XC_0988 (Table S2). It has been demonstrated that in Xanthomonas spp. TonB-dependent receptors are involved in efficient nutrition uptake for bacterial growth under in planta infection conditions (Blanvillain et al., 2007).
However, the DEGs do not harbour any gum genes that code for EPS production. This observation is inconsistent with the previous finding that deletion of vemR significantly reduced EPS production (Li et al., 2020;Tao & He, 2010). In Xcc, the gum cluster consisting of 12 genes (gumB to gumM) is responsible for the assembly, polymerization, and export of EPS (Ielpi et al., 1993;Katzen et al., 1998).
It has been demonstrated that in Xcc EPS synthesis is initiated in the late-exponential growth phase and reaches maximal production during the stationary growth phase, and that the expression of the gum genes mirrors the time course of EPS production (Harding et al., 1995;Vojnov et al., 2001). As described above, our transcriptome analysis used the bacterial cells in the mid-exponential growth phase. Therefore, this inconsistency is probably due to the employment of bacterial cells from different growth phases in the experiments.
Taken together, the results from the transcriptome analysis reveal that VemR has a broader regulatory role than previously observed. It acts as a global regulator involved in many cellular processes in Xcc.

| VemR positively regulates hrp gene expression and HR induction in planta
Our transcriptomic data show that the hrp genes were up-regulated in the mutant ΔvemR cultured in the nutrition-rich medium NYG, suggesting that VemR negatively affects the expression of hrp genes under the test conditions. To clarify whether VemR regulates hrp gene expression in minimal medium in the same manner as in nutrition-rich medium, we determined by RT-qPCR analysis the transcript levels of several operons in the hrp cluster (hrcC, hrcN, hrpB2, hrcU, hrcQ, and hrpF), two key regulators (hrpX and hrpG), and two T3E genes (avrAC The results revealed that the ΔvemR mutant induced a delayed and weakened HR compared to the wild-type strain. As shown in Figure 2b, the ΔvemR mutant did not induce any visual HR symptoms until 16 h postinoculation (hpi), while the wild-type strain triggered typical HR symptoms. Moreover, the complemented strain CΔvemR (the mutant ΔvemR harbours an entire vemR coding sequence cloned into the vector pLAFR3; Table S1) could stimulate wild-type HR symptoms. The effect of VemR mutation on HR induction was further substantiated using the electrolyte leakage assay. Here, leaf tissues within the infiltration areas were collected To get a clue about the hierarchy of VemR and HrpG in the regulation of T3SS expression, we expressed constitutively hrpG in the vemR-deleted mutant ΔvemR and determined its phenotypes.
To achieve this, the entire open reading frame (ORF) sequence of the hrpG gene was cloned into the plasmid pLARF3 to generate the recombinant plasmid pR3G (Table S1), which was then introduced into the mutant strain ΔvemR. The obtained cross-complemented strain, named ΔvemR/pR3G (Table S1), was tested for HR induction in the nonhost plant pepper ECW-10R. The result showed that ΔvemR/pR3G induced a wild-type HR symptom, while the control strain ΔvemR/pLAFR3 produced a delayed and weakened HR similar to that caused by the ΔvemR mutant ( Figure 3a). The effect of constitutively expressing hrpG in ΔvemR on HR induction was further substantiated using the electrolyte leakage assay. As shown in Figure 3b, the electrolyte leakage values induced by ΔvemR/pR3G and the wild type were at a similar level at all of the tested time points. These combined data indicate that the HR-eliciting capability of the ΔvemR mutant can be completely restored by constitutively expressing hrpG. Although it is unclear whether constitutively expressing HrpG in the vemR deletion mutant could completely recover the production of the effectors responding for HR induction, the above results suggest that the vemR mutant with constitutive expression of HrpG could produce the HR induction effector enough for a wild-type HR stimulation.
The virulence of the mutant strain ΔvemR/pR3G was also determined by the leaf clipping inoculation method in the host plant Chinese radish. Ten days after inoculation, ΔvemR/pR3G caused disease symptoms with a mean lesion length of 8.7 mm, which was significantly more severe than that caused by the mutant ΔvemR (mean lesion length 3.2 mm) but significantly less severe compared to the wild-type strain (mean lesion length 13.8 mm) and the complemented strain CΔvemR (mean lesion length 13.2 mm) ( Figure 4a). This indicates that constitutive expression of hrpG could only partially restore the virulence of the VemR-deficient mutant.
In parallel, whether constitutively expressing hrpG in the mutant ΔvemR restores the transcriptional level of the T3SS genes in planta was investigated by RT-qPCR. The strains 8004, ΔvemR, CΔvemR, and ΔvemR/pR3G were inoculated into Chinese radish leaves by infiltration. RNAs were extracted from the infected leaves at 24 hpi and subjected to RT-qPCR analysis using specific primers designed for determination of some hrp/hrc (hrcC, hrcN, hrcQ, hrcU, hrpF, hrpG, and hrpX) and T3E genes (avrAC and xopN) ( Table S3).
As shown in Figure 4b, the mRNA levels of all the tested genes in ΔvemR/pR3G were fully recovered to the wild-type level, even to a higher level than the wild type. As expected, the mRNA levels of the determined genes were obviously decreased in ΔvemR compared to the wild-type strain and completely restored to the wild-type level in CΔvemR ( Figure 4b). These results suggest that constitutive expression of hrpG in the mutant lacking vemR can restore the expression of hrp genes in planta. Taken together, the above combined data indicate that VemR positively regulates hrp genes during infection and controls virulence partially by regulating the expression of T3SS.

| VemR physically interacts with a subset of proteins including HrpG
As mentioned above, previous works demonstrated that VemR physically interacts with FleQ, FliM, and RavA to control bacterial motility (Li et al., 2020;Lin et al., 2022). As deletion of VemR influences many other cellular processes in addition to cell motility, the scope of regulation by VemR in Xcc cannot be completely accounted for by FleQ, FliM, and RavA. To identify other proteins interacting with VemR, we employed co-immunoprecipitation (Co-IP) coupled with liquid chromatography tandem-mass spectrometry (LC-MS/ MS) as previously described (Li et al., 2022). To this end, we constructed the strain 8004/VemR::3 × FLAG, which chromosomally expresses VemR protein fused with a 3 × FLAG-tag (VemR::3 × FLAG) (Table S1). A western blot assay confirmed that the VemR::3 × FLAG fusion protein could be eluted from the strain 8004/VemR::3 × FLAG, but not the wild-type strain 8004 ( Figure S1). Protein complexes with VemR::3 × FLAG in Xcc cells were purified and analysed by LC-MS/MS. This analysis identified a number of putative interacting proteins for VemR (Table S4). Among these proteins, 17 (XC_0638, XC_1060, XC_1186, XC_1187, XC_1358, XC_1359, XC_1625, XC_2284, XC_2243, XC_2245, XC_2266, XC_2267/FliM, XC_2268/ F I G U R E 2 VemR plays a role in Xanthomonas campestris pv. campestris (Xcc) type III secretion system (T3SS). (a) The transcription level of T3SS-related genes in vemR deletion mutant was estimated by reverse transcription-quantitative PCR. Xcc strains were grown in XVM2 medium to a concentration of OD 600 of 0.6, and RNA was isolated. Relative gene expression with respect to the corresponding transcript levels in the wild-type strain 8004 was calculated. Values given are the means ± SD of triplicate measurements from a representative experiment. Differences were evaluated using Student's t test (**p < 0.01). Similar results were obtained in two other independent experiments. (b) Hypersensitive response (HR) symptoms observed on pepper leaves after infiltration. Bacterial cells from overnight cultures of the Xcc wildtype strain (8004), ΔvemR, CΔvemR, and ΔhrcV (negative control) were collected, washed with, and resuspended in 10 mM sodium phosphate buffer to a density of 10 7 cfu/ml. Approximately 5 μl of bacterial resuspension was infiltrated into the leaf mesophyll tissue with a blunt-end plastic syringe, then the plants were placed under a bright light for 24 h. The HR symptoms caused by Xcc strains were recorded at 8, 16, and 24 h postinoculation (hpi). Three replications were done in each experiment, and each experiment was repeated three times. The results presented are from a representative experiment, and similar results were obtained in all other independent experiments. (c) Electrolyte leakage from pepper leaves inoculated with Xcc strains. The conductivity of the infiltration spots was measured at 0, 8, 16, and 24 hpi, with four 0.4 cm 2 leaf disks collected from the infiltrated area for each sample. Three samples were taken for each measurement in each experiment. Data are shown as the mean ± SD of three replicates from a representative experiment, and similar results were obtained in two other independent experiments. Statistical differences denoted by asterisks at data points were determined by analysis of variance and Dunnett's post hoc test for comparison to the wild type at each time point. *p < 0.05, **p < 0.01 The reason why FleQ was absent is unknown. Importantly, the T3SS key regulator HrpG was identified as a potential VemR-interacting protein in the experiment.
Given that VemR affects the expression of T3SS genes, we further validated its interaction with the T3SS key regulator HrpG using the bacterial two-hybrid (B2H) assay as previously described (Li et al., 2020). To do this, vemR and hrpG were amplified by PCR using the corresponding primer sets (Table S3) and cloned into the bait vector pBT and the prey vector pTRG, respectively, resulting in the recombinant plasmids pBTvemR and pTRGhrpG (Table S1). The obtained recombinant plasmids were cotransformed into the reporter Escherichia coli XL1-Blue MRF′ (Table S1), and the resulting strain was tested on double-selective indicator plates containing 3-amino-1,2,4-triazole and streptomycin. As shown in Figure 5a, similar to the positive control strain expressing Mip and PrtA proteins which have been shown to interact with each other (Meng et al., 2011), the reporter strain expressing VemR and HrpG could grow on the selective plates. However, the strains expressing either HrpG or VemR could not grow (Figure 5a). These data suggest that HrpG and VemR can physically interact with each other. This was F I G U R E 3 Constitutively expressing hrpG in the vemR mutant restores its ability of hypersensitive response (HR) induction in nonhost plant. The Xanthomonas campestris pv. campestris (Xcc) wild-type strain 8004, vemR deletion mutant ΔvemR, cross-complemented strain ΔvemR/pR3G (ΔvemR constitutively expressing hrpG), and ΔvemR/pLAFR3 (ΔvemR harbouring an empty vector pLAFR3) were cultured in nutrient-yeast-glycerol (NYG) medium overnight. Bacterial cells were collected, washed with, and resuspended in 10 mM sodium phosphate buffer to a density of 10 7 cfu/ml. Approximately 5 μl of bacterial resuspension was infiltrated into the leaf mesophyll. (a) The HR symptoms recorded at 8, 16, and 24 h postinoculation (hpi). Three replications were performed in each experiment, and the experiment was repeated three times. The results presented are from a representative experiment, and similar results were obtained in all other independent experiments. (b) The electrolyte leakage from pepper leaves inoculated with Xcc strains. The conductivity of the infiltrating spots was measured at 0, 8, 16, and 24 hpi, with four 0.4 cm 2 leaf disks collected from the infiltrated area for each sample. Three samples were taken for each measurement in each experiment. Data are shown as mean and standard deviation. Significance was determined by analysis of variance and Dunnett's post hoc test for comparison to the wild type. *p < 0.05, **p < 0.01; n.s., not significant. The experiment was repeated three times with similar results further substantiated by a pull-down biotinylated protein-protein assay. To do this, recombinant 6 × His-tagged proteins HrpG and VemR were overproduced. After purification, the recombinant proteins were subjected to the pull-down assay. The 6 × Histagged HupB, which has been shown to interact with HrpG (Zhang et al., 2020), and the 6 × His-tagged McvR (Li et al., 2022) were F I G U R E 4 Constitutively expressing hrpG in the vemR mutant partially restores its virulence in host plant. (a) Virulence test of Xanthomonas campestris pv. campestris (Xcc) strains. The Xcc strains were cultured in nutrient-yeast-glycerol (NYG) medium overnight and bacterial cells were collected and resuspended to a concentration of 10 7 cfu/ml (OD 600 of 0.01). Chinese radish (Raphanus sativus) leaves were cut with scissors dipped in the bacterial suspensions. Lesion lengths (disease symptoms) were scored at 10 days postinoculation. Values given are the mean and SD from 15 inoculated leaves in one experiment. Analysis of variance (ANOVA) and Dunnett's post hoc test were used to identify significant differences which are indicated with asterisks. *p < 0.05, **p < 0.01; n.s., no significant. The experiment was repeated three times with similar results. (b) The expression levels of hrp genes in the Xcc wild-type strain 8004, the vemR deletion mutant ΔvemR, and the cross-complemented strain ΔvemR/pR3G in the host plant. Xcc strains were inoculated into Chinese radish leaves by infiltrating with a syringe. The infiltrated leaf part was collected 24 h after inoculation, and total RNA was extracted and reverse transcription-quantitative PCR was performed. Values given are the mean ± SD of triplicate measurements from a representative experiment. Differences were determined by ANOVA and Dunnett's post hoc test. **p < 0.01; n.s., not significant. Similar results were obtained in two other independent experiments F I G U R E 5 Physical interaction test between VemR and the type III secretion system (T3SS) key regulatory protein HrpG of Xanthomonas campestris pv. campestris (Xcc). (a) Bacterial two-hybrid experiment. The reporter strain Escherichia coli XL1-blue MRF′ with different plasmid pairs was grown on nonselective plates (inoculated with a cell concentration of OD 600 = 1.0) and double-selection indicator plates (inoculated with cell concentrations of OD 600 = 1.0 and 0.1) containing 3-amino-1,2,4-triazole (3-AT) and streptomycin (Sm). Protein-protein interactions activate the expression of addA and HIS3 genes within the reporter gene cassette of the reporter strain, resulting in resistance to 3-AT and Sm. The reporter strain with the plasmid pair pBTPrtA-pTRGMip was used as a positive control. Three independent experiments showed similar results. (b) Pull-down assay. 6 × His-tagged HrpG protein and other 6 × His-tagged proteins were overexpressed and purified. Bait protein HrpG was biotinylated and 50 μl of 0.6 μg/μl biotinylated HrpG protein was immobilized to streptavidin Sepharose beads. The potential prey protein VemR was mixed with the bait protein and incubated. The HupB and McvR proteins were used as positive and negative controls, respectively, in addition to the negative control that 6 × His::VemR was mixed with streptavidin Sepharose beads. After elution, samples were separated on 12% SDS-PAGE and visualized by Coomassie blue staining. Lane 1, 6 × His::VemR was mixed with streptavidin Sepharose beads; lane 2, pull-down of 6 × His::HupB by biotinylated HrpG; lane 3, biotinylated HrpG was mixed with 6 × His::McvR; lane 4, pull-down of 6 × His::VemR by biotinylated HrpG; M, molecular mass marker. Three independent experiments showed the same result used as a positive and a negative control, respectively. As shown in Taken together, the above data from the Co-IP coupled with LC-MS/MS, B2H, and pull-down assays indicated that a physical interaction exists between the single-domain response regulator VemR and the T3SS key regulator HrpG.

| VemR enhances the activation of hrpX transcription by HrpG in vitro
The fact that VemR physically interacts with HrpG implies that VemR may affect the regulatory efficiency of HrpG in Xcc. To validate this possibility, we determined whether VemR affects the transcriptional level of hrpX gene activated by HrpG in vitro.
Previous studies showed that hrpX is a direct target of HrpG (Ficarra et al., 2015;Zhang et al., 2020). In this study, we performed several in vitro transcription assays using a 311-bp template DNA fragment containing the hrpX promoter of the Xcc wild-type strain 8004 and the RNA polymerase (RNAP) holoenzyme from E. coli. In the experiments, we first tested the influence of different amounts of purified HrpG or VemR on hrpX transcriptional level. As shown in Figure 6a, when 0.5 U RNAP was used, a certain amount of hrpX transcripts could be generated without addition of HrpG or VemR protein. The hrpX transcript level was significantly increased when HrpG protein was added to the reaction and the transcript level increased along with the addition of increasing amount of HrpG (from 5 to 20 nM) (Figure 6a-i). However, there was no detectable variety in the hrpX F I G U R E 6 VemR enhances hrpX transcription activated by HrpG in vitro. RNA was generated from a 311-bp template DNA fragment extending from −131 to +179 relative to the transcriptional initiation site (TIS) of the hrpX promoter using Escherichia coli RNA polymerase (RNAP) holoenzyme. Transcription products were run on a 5% denatured polyacrylamide gel containing 7 M urea in Tris-borate-EDTA electrophoresis buffer. (a) VemR enhances the activation of HrpG on the transcription of hrpX. Template DNA was incubated with various amounts of HrpG (i), VemR (ii) or HrpG with various amounts of VemR protein (iii), 0.5 U of RNAP was added prior to the initiation of the reactions. Transcription products (2 μl) were then run on the gel. The amounts of RNAP, VemR, and HrpG used are indicated at the top. All experiments were replicated more than three times with similar results and representative results are shown. (b) Phosphorylation of VemR is important for its function. Reactions were carried out with DNA fragments of the hrpX and series of proteins. The amounts of proteins used are indicated above the photographs. The experiment was repeated more than three times and similar results were obtained transcript level along with the increase of VemR (from 5 to 20 nM) ( Figure 6a-ii), suggesting that unlike HrpG, VemR itself does not directly regulate hrpX. Given that VemR does not harbour a DNAbinding motif and can physically interact with HrpG, it is probable that VemR can enhance the efficiency of HrpG activating hrpX transcription by an unknown mechanism. To verify this probability, an in vitro transcription assay was performed by adding varied amounts (from 5 to 20 nM) of VemR in the reactions containing the same concentration (5 nM) of HrpG. The results showed that addition of VemR enhanced obviously the transcript level of hrpX and the enhancement was more conspicuous when the amount of VemR was increased from 5 to 20 nM (Figure 6a-iii), implying that VemR indeed acts as an enhancer for hrpX transcription activated by HrpG.

| The phosphorylation of VemR is crucial for its function in enhancing hrpX expression
It is known that in VemR the aspartyl residues at positions 11 and 56 are required for phosphorylation and the phosphorylation is essential for VemR to regulate cell motility and EPS production (Li et al., 2020;Tao & He, 2010). Here, we determined whether the aspartyl residues are also important for VemR to modulate hrpX expression. A variant of VemR, named VemR D11/56A , was constructed by replacement of the aspartyl residues at positions 11 and 56 with alanine in VemR. Then, the effect of the variant together with HrpG on the transcription of hrpX was estimated by in vitro transcription assay. As shown in Figure 6b, the variant VemR D11/56A plus HrpG (lane 5) produced a hrpX transcript level similar to HrpG alone (lane 1), which was significantly lower than that produced by VemR with HrpG (lane 2). Our previous work has revealed that VemR can be phosphorylated by the acetyl phosphate (AcP), which specifically phosphorylates the acceptor aspartyl residues of certain TCS RRs (Li et al., 2020). In the transcription assay, AcP was added in the reactions with the wild-type VemR or the variant VemR D11/56A . As shown in Figure 6b, addition of AcP in VemR (lane 3) could visibly improve hrpX transcript level compared to VemR (lane 2), while the hrpX transcript levels were similar in the reactions of VemR D11/56A with and without AcP (lanes 5 and 6).
In addition, no transcript was observed when a 121-bp internal fragment of the hrpX gene (from +1 to +121 relative to the transcriptional initiation site) was used as the template instead of the 311-bp fragment containing hrpX promoter (data no shown). Taken together, these data suggest that the phosphorylation at the aspartyl residues at positions 11 and 56 of VemR is probably important for its regulatory function in the enhancement of hrpX transcription by HrpG.

T3Es production
Our in vitro transcription results reveal that VemR can enhance the activation of hrpX by HrpG. As described above, HrpX is a crucial regulator for T3SS. In the study, we further validated the function of VemR by measurement of the T3Es produced by the vemR deletion mutant. It is known that in addition to activating the transcription of hrpX, HrpG can also bind to its own promoter to repress its own expression (Ficarra et al., 2015;. The transcriptome and RT-qPCR analyses showed that mutation of vemR significantly affected the expression of hrpG. To gain a better insight into the regulatory mechanism of VemR, we detected the T3Es production using a vemR deletion mutant with constitutive expression of hrpG to exclude the potential influence of hrpG self-regulation. To achieve this, we constructed a hrpG deletion mutant, named ∆hrpG (Table S1), and a hrpG as well as vemR double-deletion mutant, named ∆hrpG∆vemR (Table S1). The mutant strains were further modified by fusing a 3 × FLAG-tag to the T3E AvrAC or XopN (AvrAC::3 × FLAG or XopN::3 × FLAG) by in-frame insertion of the 3 × FLAG-coding sequence into the 3′ end of the AvrAC-or XopN-coding sequence in the genome of ∆hrpG and ∆hrpG∆vemR, generating strains named ∆hrpGAvrAC::3 × FLAG and ∆hrpG∆vemRAvrAC::3 × FLAG, or ∆hr-pGXopN::3 × FLAG and ∆hrpG∆vemRXopN::3 × FLAG (Table S1), respectively. We chose AvrAC and XopN because the expression of their coding genes is positively regulated by HrpX Xu et al., 2008). In Xcc, the effector responding for HR induction in the pepper plant used in this study is AvrBs1 (Xu et al., 2008).

| DISCUSS ION
Previous studies have shown that the SD-RR VemR is crucial for the virulence, cell motility, and production of EPS and exoenzymes of Xcc (Li et al., 2020;Tao & He, 2010). In the present study, we uncovered by transcriptome analysis that VemR is a global regulator modulating the expression of more than 10% of the annotated protein-encoding genes in Xcc. Phylogenetic classification reveals that, in addition to the previously known phenotypes, many other cellular processes including the essential pathogenicity determinant T3SS are also under the control of VemR. VemR is a highly conserved protein in all Xanthomonas species and it was reported that mutation of the vemR in the citrus canker pathogen Xanthomonas citri subsp.
citri and the rice bacterial leaf streak agent Xanthomonas oryzae pv.
oryzicola also resulted in a reduction of EPS, motility, virulence, and HR (Cai et al., 2022;Wu et al., 2019). It is possible that the VemR proteins among Xanthomonas species have similar functions.
Our data showed that deletion of VemR led to an increase in hrpX expression in the rich medium NYG but a decrease in hrpX expression in the minimal medium XVM2 and in planta, indicating that VemR regulates hrpX expression negatively in nutrient-rich conditions but positively in nutrient-deficient conditions. Notably, in vitro transcription assay demonstrated that phosphorylated VemR significantly enhanced the activation of hrpX transcription by HrpG, which is consistent with VemR positively regulating hrpX expression in nutrient-deficient condition but inconsistent with VemR negatively regulating hrpX expression in nutrient-rich conditions. As previously reviewed (Mole et al., 2007;Tang et al., 2006), it is well established that in xanthomonads the expression of the hrp genes is repressed in rich media but induced in planta or in certain minimal media. However, the regulatory mechanism controlling such an expression manner is currently unknown. To date, a number of regulators involved in hrp gene induction, such as HpaS, HrpG, and HrpX, have been identified and characterized, providing some insights into the complex regulation mechanism of hrp gene activation (Alvarez-Martinez et al., 2020;An et al., 2020;Mole et al., 2007). However, only a few regulators have been reported to repress Xanthomonas hrp gene expression and their regulatory mechanism remains to be explored Lu et al., 2011). Similar to VemR, HpaR1, a GntR family transcriptional regulator in Xcc, was previously demonstrated to influence hrp gene expression with opposite regulations in response to different environments .
HpaR1 regulates hrp gene expression via hrpG negatively in standard media but positively in plant tissues . Recently, Zhang et al. (2020) found by ChIP-seq that in Xcc HrpG could bind to the vemR promoter region in vivo and Cai et al. (2022) showed that in X. oryzae pv. oryzicola the transcription of vemR is repressed by HrpG in nutrient-rich conditions. Taken together, these data suggest that VemR and HrpG play important roles in both induction and repression of hrp gene expression in different growth conditions. This work focused on the function of VemR in positive regulation of hrp genes.
The mechanism by which VemR represses hrp gene expression in nutrient-rich conditions needs to be further investigated.
Our Co-IP coupled with LC-MS/MS, B2H and pull-down assays confirm that VemR physically interacts with HrpG ( Figure 5). Furthermore, the vemR deletion mutant is deficient in full HR induction and constitutive expression of hrpG restores full HR to the mutant (Figures 2 and 3). These combined data suggest that the influence of VemR on HR induction is probably through HrpG. As described above, HrpG is an OmpR family DNA-binding regulator that directly binds to the promoter region of hrpX to activate its transcription. In this study an in vitro transcription experiment was conducted to validate the suggestion. The results showed that HrpG indeed could activate the transcription of hrpX in vitro. More importantly, the experiment showed clearly that VemR alone did not alter the hrpX transcript level, but it could obviously increase the hrpX transcripts initiated by HrpG. These results support the finding that F I G U R E 7 Western blot assays revealed that mutation in VemR reduces the level of type III effectors (T3Es) production. Strains ∆hrpGAvrAC::FLAG/pR3G, ∆hrpG∆vemRAvrAC::FLAG/pR3G, ∆hrpGXopN::FLAG/pR3G, and ∆hrpG∆vemRXopN::FLAG/pR3G were cultured in XVM2 medium for 12 h and total proteins were prepared. Thirty micrograms of protein for each strain was electrophoresed in SDS-PAGE and transferred to a PVDF membrane. The presence of AvrAC protein in strains ∆hrpGAvrAC::FLAG/pR3G and ∆hrpG∆vemRAvrAC::FLAG/ pR3G (a) or XopN protein in strains ∆hrpGXopN::FLAG/pR3G and ∆hrpG∆vemRXopN::FLAG/pR3G (b) was detected by anti-FLAG-tag mouse monoclonal antibody. As a loading reference, the blot was also probed with an anti-RNA polymerase β-antibody (low element). Experiments were replicated three times with similar results and representative results are shown VemR acts as an enhancer to improve the transcriptional activation of hrpX by HrpG. Notably, the hrpX transcript level could be significantly increased along with raising the HrpG concentration from 5 to 20 nM, without addition of VemR. In addition, the hrpX transcript level generated in the reaction with 5 nM HrpG and 10 nM VemR was similar to that generated in the reaction with 20 nM HrpG only.
These data may explain why constitutively expressing hrpG could restore full HR of the VemR-deficient mutant. As expected, constitutive expression of hrpG could only partially return virulence to the VemR-deficient mutant (Figure 4a) because that in addition to the T3SS regulated by HrpG/HrpX, VemR also plays an important role in cell motility and the production of EPS and exoenzymes, which are not regulated by HrpG/HrpX but important for Xcc virulence.
Currently, we do not know the detailed mechanism by which VemR improves the function of HrpG. The in vitro transcription assay demonstrated that in the reaction with HrpG plus VemR, the hrpX transcripts were significantly increased after addition of AcP, and that addition of AcP to the reaction with HrpG or HrpG plus VemR D11/56A did not alter the transcriptional level of hrpX. In addition, our previous work has confirmed that VemR can be phosphorylated by AcP in vitro and its aspartyl residues at positions 11 and 56 are essential for the phosphorylation (Li et al., 2020). These combined results suggest that AcP maybe improves the phosphorylation of VemR, thereby enhancing the activation of hrpX by HrpG, but cannot improve the activity of HrpG. It is possible that AcP has no effect on HrpG phosphorylation under the test conditions. Based on these results, we assume that a possible regulatory mechanism for VemR in hrpX expression may be that the binding of phosphorylated VemR to HrpG could modulate the activation activity of HrpG to hrpX promoter, thereby increasing the transcription level of hrpX. Thus, VemR acts as an accessory element to HrpG protein involved in controlling T3SS in Xcc. Notably, there are several TCS members and non-TCS member DNA-binding proteins that might physically interact with VemR, revealed by the Co-IP coupled with LC-MS/MS assay, implying that the phosphorylated VemR may interact with these regulators in multilayered control of diverse cellular processes. To further verify whether VemR binds to these proteins to alter their impact on target gene expression is of great value to broaden our understanding of the function of VemR.
Alternatively, given that HrpG and HpaS compose a TCS, VemR may act as an intermediate in the phosphorelay between HrpG and its cognate sensor kinase HpaS to control hrpX expression in vivo.
It is also possible that binding of VemR to HrpG can increase the stability of HrpG protein or modulate the binding strength of HrpG to hrpX promoter, thereby increasing the transcription of hrpX.
Further works are required to test these hypotheses.

As mentioned above, previous studies have shown that HpaS and
HrpG form a TCS that regulates the expression of hrpX (Li et al., 2014), and that HpaS can also form another TCS with VemR to control EPS production and cell motility (Li et al., 2020). Notably, constitutively expressing VemR in the hpaS deletion mutant could partially restore HR induction and T3SS expression ( Figure S2). Furthermore, deletion of both hpaS and vemR caused a more severe reduction in T3SS expression and HR induction than deletion of hpaS or vemR only ( Figure S2). The data obtained in this study combined with the previous results suggest that HpaS can modulate directly HrpG activity and indirectly via VemR. As HpaS is a membrane-bound sensor kinase, it might play a distinct role in sensing environmental stimuli and transducing signals to regulate T3SS expression. In addition, it has been recently reported that the TCS sensor RavA can also interact with and phosphorylate VemR, and that VemR can interact with the transcriptional activator FleQ to repress the transcription of flagellum-related genes (Lin et al., 2022). These findings indicate that VemR may function as an intermediate in signal transduction pathways involved in multiple physiological processes (Figure 8).
In conclusion, this work shows that the regulon of the TCS SD-RR member VemR comprises a large portion of Xcc genes involved in many cellular processes. In addition to the previously known important virulence factors such as EPS and motility, VemR also controls the T3SS, a critical pathogenicity determinant. Taken together with previous studies, our results suggest that the phosphorylation of VemR might be crucial for its function and that VemR enhances the activation of hrpX by HrpG, probably via its interaction with HrpG ( Figure 8). To the best of our knowledge, this is the first report that a SD-RR very likely executes its function by directly interacting with a typical TCS response regulator with DNA-binding ability to regulate gene expression.
Although further studies are required to explore in depth the detailed functional mechanism of VemR, our findings provide some insights into the complex regulatory cascade of the HpaS/RavA-VemR/HrpG-HrpX signal transduction system in the control of T3SS expression.
More importantly, our work suggests VemR probably plays a vital role in the signalling/regulatory network controlling virulence (Figure 8).

| DNA and RNA manipulations
DNA manipulations followed the procedures previously described (Sambrook et al., 1989). Conjugations between Xcc and E. coli strains were performed as previously described (Turner et al., 1985). The restriction endonucleases T4 DNA ligase and Pfu DNA polymerase were provided by Promega. Total RNA extraction from bacterial culture or infected plant samples was carried out using RNAiso Plus Kit (TaKaRa) and cDNA was generated using a cDNA synthesis kit (Invitrogen).
For measuring the transcription level of certain genes, RT-qPCR was carried out as previously described (Li et al., 2022) using ChamQ universal SYBR qPCR master mix (Vazyme) with corresponding primers (Table S3)

| Deletion mutant construction and complementation
vemR (XC_2252) single-deletion mutant ΔvemR and the complemented strain CΔvemR have been constructed in previous work (Li et al., 2020). To construct the hrpG (XC_3077) deletion mutant (or hrpG and vemR double deletion mutant), 724-bp upstream and 698-bp downstream fragments of the hrpG coding region were amplified using the corresponding primers (Table S3). The two fragments were cloned together into the vector pK18mobsacB (Schäfer et al., 1994). The resulting recombinant plasmid pK18mobsacBhrpG was introduced into Xcc 8004 (or ΔvemR mutant) by triparental conjugation. The obtained mutant was named ΔhrpG (or ΔhrpGΔvemR) (Table S1). For cross-complementation of the ΔvemR mutant, the recombinant plasmids pR3G (Table S1), which derived from an 899-bp fragment containing promoterless hrpG gene cloned into the vector pLAFR3 (under the control of the lacZ promoter which expresses constitutively in Xcc), was transferred into the ΔvemR mutant by triparental conjugation, resulting in strain ΔvemR/pR3G.

| Construction of the strains chromosomally encoding 3 × FLAG fused form protein
Xcc strains expressing the proteins fused with a 3 × FLAG-tag at the C-terminus were constructed using the method previously F I G U R E 8 A model indicating the roles of the single domain response regulator VemR in Xanthomonas campestris pv. campestris (Xcc). As it does not have an output domain, VemR is supposed to recruit certain proteins to execute its regulatory functions. This study demonstrates that phosphorylated VemR regulates the type III secretion system (T3SS) by directly interacting with the two-component signalling system (TCS) response regulator HrpG and enhancing its activation activity on hrpX expression. The TCS HK sensor HpaS can phosphorylate HrpG and VemR. In addition to HpaS, the TCS HK sensor RavA can also phosphorylate VemR. Moreover, in addition to interacting with FliM to modulate cell swimming motility, phosphorylated VemR can also interact with the transcriptional activator FleQ to repress the expression of the flagellar genes. However, how VemR regulates extracellular polysaccharide (EPS) and exoenzyme production is unclear. Whether the unphosphorylated VemR playing a regulatory role is unknown. Red arrow, positive regulation; red line with endbar, negative regulation described (Li et al., 2022). Briefly, to construct a strain chromosomally encoding VemR::3 × FLAG for Co-IP assays, a 428-bp DNA fragment, which was composed of a 381-bp VemR-coding sequence and a 47-bp FLAG-coding sequence, was generated by PCR amplification using the genomic DNA of strain 8004 as template and the primer set LvemR-FlagF/R (Table S3).
Simultaneously, a 415-bp DNA fragment, which was composed of a 39-bp FLAG-coding sequence, the 3-bp stop codon of vemR, and 373-bp downstream of the vemR stop codon, was generated by PCR amplification using the primer set RvemR-FlagF/R (Table S3). The two fragments were joined using overlap extension PCR, and the resulting recombinant fragment was cloned into the suicide plasmid pK18mobsacB (Schäfer et al., 1994). The resulting recombined plasmid named pKvemR::flag (Table S1) was introduced into Xcc strain 8004 by conjugation, and the transconjugants chromosomally encoding VemR::3 × FLAG protein were screened and confirmed by the procedure described previously (Liu et al., 2019). The obtained variant strain was named 8004/ VemR::3 × FLAG (Table S1). Similarly, to construct a strain chromosomally encoding AvrAC::3 × FLAG or XopN::3 × FLAG for western blotting, a recombinant fragment encoding an in-frame 3 × FLAG peptide at the C-terminus of AvrAC or XopN was obtained by PCR amplification with the corresponding primer sets (LavrAC-FlagF/R and RavrAC-FlagF/R for AvrAC::3 × FLAG, and LxopN-FlagF/R and RxopN-FlagF/R for XopN::3 × FLAG; Table S3). These obtained fragments were cloned into the suicide vector pK18mobsacB, and the resulting recombinant plasmids, named pKavrAC::flag and pKxopN::flag (Table S1), were introduced into the Xcc hrpG deletion mutant ∆hrpG and the hrpG/vemR double deletion mutant ∆hrpG∆vemR, respectively.

| Co-immunoprecipitation and LC-MS/MS analysis
Co-IP combined with LC-MS/MS analysis was employed to identify the VemR-interacting partners as previously described (Li et al., 2022

| Western blotting
Western blot assays were performed followed the procedure described by Sambrook et al. (1989)

| Bacterial two-hybrid assay
The BacterioMatch II two-hybrid system (Stratagene) was carried out to detect the VemR-HrpG interaction as previously described (Li et al., 2014). Briefly, the 381-bp vemR gene, obtained by PCR using the primer set vemR-BTF/R (Table S3), was cloned into the bait vector pBT, generating the plasmid pBTvemR (Table S1). The 789bp hrpG coding sequence was PCR-amplified from the Xcc strain 8004 with the primer set hrpG-TRGF/R (Table S3) and cloned into the target vector pTRG, resulting the plasmid pTRGhrpG (Table S1).
The plasmid pairs were cotransformed into the reporter strain E. coli XL1-Blue MRF′. The cells of the resulting strains were resuspended in M9 medium (67.8 g Na 2 HPO 4 , 30 g KH 2 PO 4 , 5 g NaCl, 10 g NH 4 Cl per litre) and adjusted to a concentration of OD 600 of 1.0 and 0.1. The bacterial suspensions were spotted on the nonselective plates and double-selective indicator plates containing 5 mM 3-amino-1,2,4-triazole and 12.5 μg/ml streptomycin, and then incubated at 28°С for 24 h.

| Overproduction and purification of proteins
To overproduce HrpG protein, the 789-bp hrpG coding sequence was PCR-amplified from the Xcc strain 8004 using the primer set hrpG-OF/R (Table S3). The obtained DNA fragment was cloned into the expression vector pGEX-4T-1. The resulting recombinant plasmid named pGEX-HrpG (Table S1) (Table S1) were grown and induced by IPTG. The cells were collected and the 6 × His-tagged fusion proteins were purified using Ni-NTA resin (Qiagen).
Then beads were washed and prey protein was eluted using elution buffer (pH 2.8). Twenty microlitres of the eluted sample were electrophoresed on 12% SDS-PAGE and visualized by Coomassie brilliant blue staining.

| In vitro transcription assays
In vitro transcription assays were performed as previously described (Su et al., 2016). Promoter sequence fragments (311 bp) of hrpX were acquired using PCR with the primer set hrpXivt-F/R (Table S3). The obtained hrpX promoter sequence fragments and HrpG or/and VemR protein were incubated for 30 min at room temperature in transcription buffer. Then, a NTP mixture (250 μM each of ATP, CTP, and GTP, 250 μM biotin-16-UTP) and 0.5 U of E. coli RNA polymerase holoenzyme (New England BioLabs) were added to initiate transcription. After incubation at 28°C for 30 min, the reactions were terminated and the transcription products were analysed by electrophoresis. The transcripts obtained were visualized using a phosphor imager screen (GE AI600).

| Transcriptome analysis
Transcriptome analysis was performed as previously described (Cui et al., 2018). The Xcc wild-type strain 8004 and the vemR deletion mutant ΔvemR were cultured in NYG medium to an OD 600 of 0.6, and RNA was then prepared. After the quantity determination and quality assessment, total RNA was sent to Novogene for library construction and strand-specific RNA sequencing. Sequencing libraries were generated using a NEBNext Ultra Directional RNA Library Prep Kit for DEGs were selected randomly to perform RT-qPCR analysis.

| Plant assays
The virulence of Xcc strains to Chinese radish (Raphanus sativus) was tested by the leaf-clipping method (Dow et al., 2003) with minor modification. Bacterial cells from overnight culture were collected, washed, and resuspended to a cell density of OD 600 of 0.01 (approximately 10 7 cfu/ml) in sterile ultrapure water. Leaves were cut with scissors dipped in the bacterial suspensions. Lesion length was measured 10 days after inoculation. Three replications were done in each experiment, and the experiment was repeated three times.
For RT-qPCR analysis, leaves of Chinese radish were inoculated by infiltrating with the bacterial suspension with a concentration of OD 600 of 0.1. The infiltrated leaf part was collected 24 h after inoculation, snap frozen in liquid nitrogen, and stored at −80°C immediately.
HR was tested in pepper leaves (Capsicum annuum 'ECW-10R') as previously described (Castañeda et al., 2005;Li et al., 2014). Briefly, Xcc strains were cultured in NYG overnight and cells were collected, washed with 10 mM sodium phosphate buffer, and resuspended in the same buffer to a cell density of OD 600 of 0.01. Then, the bacterial suspension was infiltrated into the abaxial side of the pepper leaves. The inoculated plants were kept in the greenhouse to observe HR symptoms and to gauge conductivity at 0, 8, 16, and 24 hpi.
For conductivity measurements, samples (leaf discs of 0.4 cm 2 ) were collected and soaked in 10 ml of ultrapure water with shaking at 200 rpm for 30 min. The leaf discs were then removed and the conductivity of the water was measured.

ACK N O WLE D G E M ENTS
This work was supported by the National Natural Science Foundation of China (32160036; 31371263).

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.