PhcX Is a LqsR-family response regulator that contributes to Ralstonia solanacearum virulence and regulates multiple virulence factors

ABSTRACT As one of the most notorious and successful phytopathogenic bacteria, Ralstonia solanacearum controls the transition between long-term survival and pathogenic modes through an intricate regulatory network, the understanding of which remains incomplete despite years of effort. In this study, we identified PhcX, a previously uncharacterized response regulator in R. solanacearum, and uncovered its essential functions in modulating virulence and metabolism. The phcX deletion mutant exhibited substantial phenotypic alterations, including slower initial growth, altered response to host extract, reduced motilities, polygalacturonase activity, and exopolysaccharide production, elevated biofilm formation, delayed hypersensitive response, and impaired virulence. Moreover, ~16% of all genes were differentially expressed in the mutant, among which the genes associated with virulence, nitrogen metabolism, and regulation were overrepresented (e.g., most T3SS/T3Es genes). Many of these traits and genes were regulated by PhcX and the global virulence regulator PhcA, but 81.4% of the genes were regulated in opposite directions. The functions of PhcX were largely conserved in R. solanacearum EP1 and GMI1000 strains. Apparent orthologs of PhcX are widely distributed in Proteobacteria, including the LqsR quorum sensing (QS) response regulator in Legionella pneumophilia. Notably, our data suggest that phcX was originally part of the Lqs QS system but was decoupled from Lqs in Ralstonia/Cupriavidus, physically linked to the phc QS genes, and connected with the virulence regulatory network in Ralstonia during its evolution. The findings of this study contribute to a better understanding of the virulence and metabolism regulation mechanisms in R. solanacearum and shed light on the evolution of its complex regulatory network. IMPORTANCE The bacterial wilt caused by the soil-borne phytopathogen Ralstonia solanacearum is one of the most destructive crop diseases. To achieve a successful infection, R. solanacearum has evolved an intricate regulatory network to orchestrate the expression of an arsenal of virulence factors and fine-tune the allocation of energy. However, despite the wealth of knowledge gained in the past decades, many players and connections are still missing from the network. The importance of our study lies in the identification of PhcX, a novel conserved global regulator with critical roles in modulating the virulence and metabolism of R. solanacearum. PhcX affects many well-characterized regulators and exhibits contrasting modes of regulation from the central regulator PhcA on a variety of virulence-associated traits and genes. Our findings add a valuable piece to the puzzle of how the pathogen regulates its proliferation and infection, which is critical for understanding its pathogenesis and developing disease control strategies.


IMPORTANCE
The bacterial wilt caused by the soil-borne phytopathogen Ralstonia solanacearum is one of the most destructive crop diseases.To achieve a successful infection, R. solanacearum has evolved an intricate regulatory network to orchestrate the expression of an arsenal of virulence factors and fine-tune the allocation of energy.However, despite the wealth of knowledge gained in the past decades, many players and connections are still missing from the network.The importance of our study lies in the identification of PhcX, a novel conserved global regulator with critical roles in modulating the virulence and metabolism of R. solanacearum.PhcX affects many well-characterized regulators and exhibits contrasting modes of regulation from the central regulator PhcA on a variety of virulence-associated traits and genes.Our findings add a valuable piece to the puzzle of how the pathogen regulates its proliferation and infection, which is critical for understanding its pathogenesis and developing disease control strategies.and other pathogenicity functions) (38,39), xpsR and epsR (EPS biosynthesis) (40,41), pehSR (polygalacturonase production and other virulence functions) (42), and flhDC (flagellar motility) (43).Furthermore, ralfuranones, a lectin protein, LecM, and EPS I have all been demonstrated to participate in the feedback of the QS signaling pathway (44,45).Although much progress has been made, there are still many missing pieces in the regulatory network of R. solanacearum, and new players continue to be discovered.For instance, anthranilic acid and its response regulator, RaaR, were recently reported to control important biological functions of R. solanacearum and the synthesis of QS signals, including 3-OH MAME and AHL signals (46,47).A putative sensor histidine kinase, PhcK, was found to be a positive regulator of phcA in R. solanacearum OE1-1 (48).
In this study, we set out to screen for additional virulence regulators in R. sola nacearum by conducting in vitro continuous passage experiments and obtaining an EPS-deficient mutant.Whole-genome sequencing of the mutant identified a mutation of the gene RSc2734 encoding a putative response regulator orthologous to Legionella LqsR.RSc2734 has an RSSC-wide conserved genomic location adjacent to the phcBSRQ operon, thus being designated as phcX.We generated phcX deletion mutants in both R. solanacearum EP1 and GMI1000, conducted growth assays, phenotypic characterization, pathogenicity tests, and transcriptome analyses, and compared the regulons of PhcX and PhcA.Our results suggest that the PhcX is a novel global regulator in R. solanacearum with broad regulatory functions in both virulence and metabolism.

Identification of a putative response regulator, PhcX, from spontaneous EPS-deficient mutants of R. solanacearum EP1
We conducted in vitro continuous passage experiments in R. solanacearum strain EP1 and screened for EPS-deficient mutants at each passage using visible EPS production on the solid medium as the indicator.One mutant from the 11th generation and two mutants from the 23rd generation were obtained and sequenced (Table S1A).The two EPS-deficient mutants from the 23rd generation contained the same missense mutations in pilT (AC251_RS04265) and the gene AC251_RS03885 as determined by whole-genome sequencing analysis (Fig. S1a).According to the NCBI RefSeq database, the AC251_RS03885 gene encodes a putative response regulator (RefSeq protein id: WP_011002641.1).AC251_RS03885 was predicted to contain no signal peptide or transmembrane domain and have a cytoplasmic localization.Despite its low levels of sequence similarity to other proteins in EP1, AC251_RS03885 carries a response regulator receiver domain (InterPro id: IPR001789) also found in PhcR and PhcQ.Information from the STRING database (49) indicated that AC251_RS03885 is strongly associated with the Phc QS system and other virulence regulators (e.g., VsrB and RcsC) (Fig. S1b).In the EP1 genome, the AC251_RS03885 gene is located immediately upstream of the phcBSRQ operon, with a 596-bp intergenic region in between.Notably, our analysis of the 517 RSSC genomes in the NCBI RefSeq database showed that only 2 lacked AC251_RS03885 (Table S2).Moreover, in strains containing phcX, this gene was located near the known phc cluster, although its location relative to the other phc genes varied by strain as this locus appears to have undergone rearrangements in many RSSC genomes.The AC251_RS03885 gene was thus named phcX due to its close proximity to the phc genes.

Deletion of phcX impairs the growth rate of EP1 during the early proliferation stage and alters the response to ground and filtered tomato stems
In order to investigate the function of PhcX and reveal its relationship with the Phc QS system, the phcX gene deletion mutant (ΔphcX) and its complementation strain (phcX-comp) were constructed from strain EP1 (Table S1B).The growth curve assay was conducted on wild-type EP1, ΔphcX, and phcX-comp under three different nutrition conditions.Compared with wild-type EP1, ΔphcX showed a lower growth rate at the early proliferation stage as it took a longer time for ΔphcX to reach an OD 600 of 0.3 in the CTG medium (a rich medium; delayed by ~6 h; Fig. 1a), the MP medium (39) (a minimal, nutrient-poor medium) containing 50 mM L-glutamine (delayed by ~2-3 h; Fig. 1b), and pure ground and filtered tomato stems (mimicking the in vivo nutrition condition; delayed by >9 h; Fig. 1c).We calculated the growth rates of strains in CTG medium and found that ΔphcX had a much longer lag phase duration (13.36 h) and a slightly lower exponential growth rate (0.20 h −1 ) than wild-type EP1 (7.82 h; 0.22 h −1 ) and phcX-comp (8.64 h; 0.22 h −1 ) (Fig. S2).In the ground and filtered tomato stems, the lag-phase of ΔphcX (21.67 h) was also substantially prolonged compared with wild-type EP1 (11.92 h) and phcX-comp (11.84 h) (Fig. S2).Interestingly, ΔphcX entered the stationary phase at a similar cell density as the wild-type EP1 in the rich medium (Fig. 1a) but could grow to higher OD 600 values in the minimal medium (Fig. 1b) and ground and filtered tomato stems (Fig. 1c).The growth curves of phcX-comp and the wild-type strain EP1 were similar under all three conditions.
Our results revealed a more pronounced difference in the growth curves of the wild-type EP1 and ΔphcX in ground and filtered tomato stems (Fig. 1c).To further investigate the role of phcX in the response of EP1 to ground and filtered tomato stems, we also examined the growth curves of the EP1 strains in CTG medium containing 0.1% ground and filtered tomato stems.Our results showed that, for each strain, adding 0.1% sap did not alter the lag phase growth curve but led to an increased log phase growth rate and a higher maximum cell density (Fig. 1d through f).Meanwhile, with the addition of 0.1% sap, the wild-type EP1 and phcX-comp started to exhibit significantly higher cell density at an earlier time point in the log phase (at the 19th h; t test, P < 0.0001; Fig. 1d and f) than ΔphcX (at the 31st h, t test, P < 0.0001; Fig. 1e).Therefore, the deletion of phcX delayed the induction of EP1 growth rate by stem sap.Overall, our results suggest that phcX plays a role in the early proliferation of EP1 and its response to host plant extract.Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analyses showed that genes related to nitrogen metabolism (narG/H/L) and cytochrome c oxidase (CcoO/N/P) were down-regulated in ΔphcX compared with the wild-type (Table S3A).

Multiple virulence-related phenotypes of EP1 are modulated by PhcX
We examined ΔphcX for several virulence-related phenotypes, including swimming and twitching motilities, polygalacturonase and cellulase activities, biofilm formation, EPS production, and hypersensitive response (HR).

Motility
Swimming and twitching motilities are important virulence factors of RSSC required for plant colonization and full virulence, especially during the early infection stages (10,43).Here, the swimming and twitching motilities of the EP1 strains were compared.The deletion of phcX substantially decreased the swimming and twitching motilities relative to the wild-type EP1, and the complementation of phcX fully restored the corresponding phenotypes (Fig. 2a and b).RT-qPCR tests demonstrated that the expression levels of genes related to swimming motility (e.g., fliC, flgK, flgN) were reduced in ΔphcX compared with the wild-type (Table S3A).

Plant cell wall degrading enzymes
Plant cell wall degrading enzymes are key virulence factors contributing to the virulence of R. solanacearum (50,51).Here, the polygalacturonase and cellulase activities of the EP1 strains were analyzed.Results indicated that ΔphcX exhibited substantially attenuated polygalacturonase activity (P < 0.0001) and slightly but significantly decreased cellulase activity (P < 0.01) compared with the wild-type EP1 and phcX-comp (Fig. 2c and d).An RT-qPCR test of the polygalacturonase-related gene pehB showed that it was down-regu lated in ΔphcX (Table S3A).

Biofilm
The polyvinyl chloride crystal violet assay and pellicle formation assay were used to quantify in vitro bacterial adhesion to surfaces and cohesion, respectively.The results showed that the adherent biofilm production of ΔphcX was substantially higher (P < 0.001) than that of the wild-type EP1 (Fig. 2e).In addition, after 36 h of static culture in CTG medium, ΔphcX formed considerably more floating pellicles at the air-liquid interface than the wild-type EP1 (Fig. 2e).phcX-comp produced slightly more adherent biofilm (P < 0.01) than the wild-type EP1 and showed no significant difference in pellicle formation (Fig. 2e).Furthermore, lecM, encoding a lectin, is reported to be involved in the biofilm formation in R. solanacearum (8) and was found to be significantly up-regulated in ΔphcX by our RT-qPCR assay (Table S3A).

EPS
In bacterial wilt diseases, the accumulation of pathogen-produced EPS is a major factor that blocks the water traffic within xylem vessels and ultimately causes plant wilting symptoms (25).The EPS production of the EP1 strains was analyzed by an adapted Elson and Morgan assay for EPS quantification in R. solanacearum (52).Each strain was inoculated at an OD 600 of 0.01 and cultured until an OD 600 of 1.0, and the EPS accumulated during the culturing process was measured.ΔphcX was cultured for 26 h in the experiment, which was ~40% longer than the wild-type EP1 (18 h) and phcX-comp (19 h).However, ΔphcX produced 18% less EPS than the wild-type EP1 (P < 0.01), while phcX-comp did not show significantly different EPS production from the wild-type (Fig. 2f).The EPS repressor gene epsR was up-regulated in ΔphcX according to RT-qPCR analysis (Table S3A).

HR
T3SS and T3Es are major pathogenicity determinants required for the full virulence of RSSC and can induce a strong defense response often associated with an HR (20).To determine if the deletion of phcX affects the ability of EP1 to elicit HR, we inoculated the EP1 strains on tobacco leaves and monitored the development of necrotic lesions.The results showed that tobacco leaves inoculated with ΔphcX developed necrotic lesions much slower than those inoculated with the wild-type EP1 or phcX-comp, particularly at the early time points (18 and 24 h) (Fig. 3a), and our RT-qPCR results also revealed that several genes encoding T3SS components (hrcC, hrpB, hrpG) and HR elicitor (popA) in ΔphcX were down-regulated (Table S3A).

phcX deletion impairs the virulence of EP1
As the deletion of phcX affected various virulence-related phenotypes of EP1, we inoculated tomato cultivar Zhongshusihao with the EP1 strains by stem injection and root dipping methods to assess the role of phcX in virulence.In the stem injection assay, tomato plants inoculated with ΔphcX wilted significantly slower than those inoculated with wild-type EP1 (P < 0.001, Fig. 3b).A similar result was also obtained by using the dipping method (P < 0.0001, Fig. 3c).In addition, all the plants inoculated with wild-type EP1 died within 2 weeks, whereas 12.5% and 44% of the plants inoculated with ΔphcX survived in the stem injection and root dipping assays, respectively.Together, our results suggest that phcX is important for the full virulence of EP1.

phcX deletion leads to broad down-regulation of genes underlying motility, chemotaxis, virulence, and nitrogen metabolism
To further investigate the regulatory roles of phcX, RNA-seq analysis of wild-type EP1 and ΔphcX (both cultured in CTG medium until an OD 600 of 0.3) was conducted.We obtained 7.5 to 7.9 million high-quality read pairs per sample from the wild-type and ΔphcX samples, mapped quality-filtered reads of each sample to the EP1 genome, and performed differential gene expression analysis.Principal component analysis (PCA) indicated high reproducibility among biological replicates (Fig. S3a).Compared with the wild-type, 845 differentially expressed genes {DEGs; false discovery rate (FDR) <0.05, |log 2 [fold change (FC)]| ≥1} were identified in ΔphcX, among which 473 and 372 genes were up-and down-regulated, respectively (Table S3B).The results of RNA-seq analysis were validated by RT-qPCR assays on selected genes, which revealed similar differential expression patterns (Table S3A).
To reveal the functional types of genes regulated by phcX, we performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses on DEGs.GO terms significantly enriched (FDR <0.05) among down-regulated genes include molecular transducer activity, cytoskeletal motor activity, organelle, extracellular region, external encapsulating structure, cell motility, program med cell death, signaling, and nitrogen cycle metabolic processes (Table S3C).The KEGG analysis (FDR <0.05) showed that the flagellar assembly, bacterial chemotaxis, two-component system, and nitrogen metabolism pathways were most enriched among down-regulated genes, while the steroid degradation pathway was enriched among up-regulated genes (Table S3D).These enriched GO terms and KEGG pathways are largely consistent with the phenotypes of ΔphcX.

phcX and phcA regulate multiple phc QS-dependent phenotypes and associated genes in opposite patterns
Our results demonstrated that growth, swimming motility, and polygalacturonase activity were positively regulated by PhcX.Interestingly, several studies have reported the negative regulation of these phenotypes by PhcA (4,31).To further confirm this observation, we generated a phcA deletion mutant (ΔphcA) in EP1 and compared the phenotypes of the wild-type EP1, ΔphcX, and ΔphcA side by side.Our results showed that ΔphcA grew faster than the wild-type EP1 in minimal media with one of sucrose, glucose, fructose, myoinositol, or proline as the sole carbon substrate, while ΔphcX grew at slower rates than the wild-type strain under the same conditions (Fig. 5a through d).ΔphcA also exhibited significantly increased swimming motility and polygalacturonase activity, as well as decreased adherent biofilm formation in vitro, opposite to the phenotypes of ΔphcX (Fig. 5e through g).
Similarly, our transcriptome analysis showed that the deletion of phcX led to the up-regulation of swimming-related genes and the solIR and rasIR QS systems, which were reported to be positively regulated by PhcA (30).To better understand the differences in the regulatory roles of PhcX and PhcA, we compared the DEGs identified in our RNA-seq analysis of ΔphcX with those reported in a previous transcriptome study of ΔphcA in GMI1000 (30).It should be noted that, due to differences in the gene contents, we focused our comparison on the 4,390 one-to-one orthologous genes between EP1 and GMI1000 (Table S4A and B).
opposite directions in ΔphcX and ΔphcA (Fig. 6a).In addition, the 484 DEGs detected in both ΔphcX and ΔphcA exhibited significantly higher degrees of differential expression than the DEGs specific to either ΔphcX or ΔphcA (Wilcoxon rank-sum test, P < 10 −9 ) (Fig. S3c and d).We also found a significant negative correlation (r = −0.43,P < 10 −15 ) between genes' log 2 FC values in ΔphcX and ΔphcA for all genes, and the negative correlation was relatively strong (r = −0.70,P < 10 −15 ) for the DEGs detected in both mutants (Fig. 6b).In particular, the genes related to nitrate reductase, swimming motility, chemotaxis, T3SS, hdf biosynthesis, adhesion, and QS all showed an overwhelming trend of opposite differential expression patterns in ΔphcX and ΔphcA (Fig. 6c; Table S4C).This pattern revealed by transcriptome data was independently validated by RT-qPCR assays of representative genes from each category (Fig. 5h and i).Overall, our results reveal a significant overlap between the genes regulated by PhcX and PhcA and contrasting modes of regulation between the two regulators.

Deletion of phcX results in similar phenotypes in R. solanacearum strains GMI1000 and EP1
To evaluate if the functions of phcX are conserved in other R. solanacearum strains, we generated the phcX-deletion mutant in R. solanacearum strain GMI1000 (GMI1000-ΔphcX) (Table S1B) and analyzed its phenotypes.Compared with the wild-type GMI1000, GMI1000-ΔphcX exhibited a much slower growth rate at the early stage, substantially attenuated swimming/twitching motilities and polygalacturonase activity, and significantly increased biofilm production, all consistent with the phenotypes of EP1-ΔphcX (Fig. S4a through f and h).At the same time, the deletion of phcX in GMI1000 had no significant influence on cellulase activity (Fig. S4g).Interestingly, the deletion of phcX led to substantially increased EPS production in GMI1000 (P < 0.0001), although it took twice as long for GMI1000-ΔphcX to reach an OD 600 nm of 1.0 compared with the wild-type strain (34 vs 17 h) (Fig. S4i).Transformation of native PhcX recovered the phenotypes of GMI1000-ΔphcX to those of the wild-type GMI1000.Our results suggest that functions of PhcX were largely conserved in these two phylotype I R. solanacearum strains, GMI1000 and EP1.

Phylogenetic analysis identifies PhcX as the ortholog of LqsR
To further investigate the distribution of PhcX in other bacteria, we queried the OrthoDB v11 database and found that PhcX belongs to an orthologous group (ID: 410377at1224) consisting exclusively of genes from Proteobacteria.We then retrieved all the 6,937 representative genomes (each representing a species) of Proteobacteria from the NCBI RefSeq database and compared each genome with the EP1 genome to identify potential orthologs of PhcX.In total, we found orthologs of PhcX in 310 representative genomes, including 16 Alphaproteobacteria, 206 Betaproteobacteria, 1 Deltaproteobacteria, and 87 Gammaproteobacteria (Table S5 and Fig. S5).Notably, PhcX was found to be orthologous to the LqsR of Legionella pneumophila, the causal agent of Legionnaires' disease (54).The orthology between PhcX and LqsR was further supported by our phylogenetic analysis of PhcX/LqsR homologs in Proteobacteria (Fig. S5).PhcX and LqsR have the same response regulator receiver domain and share a protein sequence similarity of 50.45%.In addition, the predicted structures of the two proteins were highly similar, with a root mean square deviation (RMSD) value of only 1.064 (Fig. 7a).
LqsR is the response regulator of the Lqs QS system, which also contains an auto inducer synthase, LqsA, and two sensor histidine kinases, LqsS and LqsT.The genes encoding LqsA, LqsR, and LqsS form a gene cluster in the L. pneumophila genome (55).The Lqs QS system is different from the Phc QS system in both their genes and the QS signal.The Phc QS system was not found in L. pneumophila, whereas the autoinducer synthase and sensor kinases of the Lqs QS system were all missing in RSSC, raising an intriguing question regarding the association of PhcX and the two different QS systems during evolution.
We then systematically investigated the distributions of gene clusters encoding the Phc and Lqs QS systems in Proteobacteria.Our results showed that the lqs gene cluster is widely distributed in Alpha-, Beta-, and Gammaproteobacteria, while the phc gene cluster has a more restricted distribution in Betaproteobacteria (Fig. 7b).The two QS systems coexist in many Betaproteobacteria species; in most cases, phcX/lqsR belongs to the lqs gene cluster.Interestingly, in Cupriavidus, phcX/lqsR is de-associated from lqsA and lqsS but relocated to the proximity of the phc gene cluster.Finally, phcX/lqsR becomes an immediate neighbor of the phc gene cluster in Ralstonia, where lqsA is lost in all its is represented by a different color.In the graphical representation of genomic arrangements, immediate neighboring genes are connected by a solid line; genes separated by a few (1 to 49) genes are connected with the number shown in between; genes separated by more than 50 genes are connected by a dotted line; and genes on different replicons (or assembled genomic sequences) are separated by a vertical line.The "+" and "−" signs in gene boxes indicate that the corresponding genes are on the forward and reverse strands, respectively.The three genera in which phcX/lqsR is located adjacent to the phc QS genes, including Hydrogenophaga, Cupriavidus, and Ralstonia, are highlighted in red.members except for R. insidiosa.Our phylogenetic results suggest that PhcX/LqsR was ancestrally part of the Lqs QS system and became associated with the Phc QS system gradually during the evolution of Ralstonia.

DISCUSSION
In this study, we report the identification and functional characterization of phcX, encoding a new regulator of metabolism and virulence in R. solanacearum.Results from in vitro experiments on R. solanacearum strains growing in culture showed that the phcX deletion mutant (ΔphcX) displayed multiple significant phenotypic alterations, includ ing a prolonged lag phase, substantially reduced swimming and twitching motili ties, decreased polygalacturonase activity and EPS production, and increased biofilm formation.In planta assays further demonstrated that phcX deletion leads to delayed HR and impaired virulence.PhcX is predicted to be a response regulator, and the deletion of phcX resulted in differential expression of a wide range of genes, many of which directly correspond to the phenotypic outcomes.Although phcX is located immediately upstream of the phc operon, it is transcribed independently, and the expression of phc genes was not significantly different in ΔphcX, suggesting that the phenotypes of ΔphcX were caused by disruption of PhcX functions.In addition, the ΔphcX phenotypes were successfully restored by phcX complementation.These lines of evidence all support the important regulatory functions of PhcX in R. solanacearum.
The ΔphcX mutant grew much slower than the wild-type strain at the early prolifera tion stage (Fig. 1), and RNA-seq analysis revealed the enrichment of GO and KEGG terms on nitrogen metabolism among down-regulated genes (Table S3C and D), suggesting a crucial role of PhcX in metabolism.The metabolism and energy generation processes are required by pathogens to thrive during their infection of host tissues.PhcA and EfpR, two of the most important regulators in R. solanacearum, play dual roles in regulating both virulence and metabolism (31,56).In ΔphcX, all nar genes related to the nitrate reductase pathway were significantly down-regulated, especially narG (log 2 FC = −5.8)(Table S4E).Several studies in R. solanacearum have highlighted the importance of nitrogen metabolism in root attachment, stem colonization, and virulence in the host xylem (21,57); in particular, ΔnarG mutant grows poorly on NO 3 − and is metabolically impaired compared to the wild-type.Another set of metabolic genes dramatically down-regula ted in ΔphcX was the cytochrome cbb3 oxidases operon (ccoSNOQPGH) (Table S4E), including ccoN, whose disruption was sufficient to abolish cbb3-type cytochrome c activity in Vibrio cholerae (58).In Proteobacteria, cytochrome cbb3 oxidases are required for the colonization of oxygen-limited environments and thus may be an important determinant of pathogenicity (59).For instance, cbb3-type cytochrome c oxidase was required for the full virulence and multiplication of R. solanacearum R3bv2 in microaero bic environments (22).These observations suggest that PhcX might control the growth of R. solanacearum positively in soil, rhizosphere, and plants by regulating nitrogen metabolism and cytochrome c oxidase activity.
Swimming and twitching motilities, chemotaxis, and polygalacturonase activity are all essential for plant colonization and full virulence in R. solanacearum at the early stages of infection, when bacterial population densities are low (60).The swimming motility mediated by flagella is activated by FlhDC, which is positively regulated by PehSR and VsrBC and negatively regulated by MotN and VsrAD (61).PehSR is also a positive regulator of twitching motility and polygalacturonase activity (62), while PhcA negatively regulates these traits by repressing PehSR (43).Interestingly, although ΔphcX exhibited substantially reduced motilities and polygalacturonase activity (Fig. 2a through  c), as well as a broad down-regulation of flagellar and polygalacturonase genes (Table S3E), most of the regulatory genes (e.g., flhD/C, pehS/R, vsrAB, vsrCD, and phcA) were not differentially expressed (Table S3B).motN, which encodes a repressor of motility, was even down-regulated (log 2 FC = −1.7)(Table S3B).Similarly, many chemotaxis-related genes were down-regulated in ΔphcX but are not known to be regulated by PhcA, PehSR, etc.Therefore, PhcX likely controls motility and polygalacturonase activity via a mechanism independent of PhcA and other unknown regulators.
Biofilm is a crucial component of virulence in R. solanacearum, and its formation involves adhesins and lectins.We found in ΔphcX an overall up-regulation of adhesionrelated genes, particularly rsl (also named lecF; log 2 FC = 8.4) and lecM (log 2 FC = 9.5) (Table S3B), two lectin-encoding genes.It has been shown in strain OE1-1 that the LecM positively regulates biofilm formation and facilitates the attachment of OE1-1 to the host cells in intercellular spaces (8).Consistent with this, both the floating pellicle at the air-liquid interface and adherent biofilm formation were significantly increased in ΔphcX (Fig. 2e).These findings indicate that PhcX exerts strong repressive control on biofilm formation.
EPS accumulates as R. solanacearum invades and multiplies in planta and blocks the flow of water and other nutrients, eventually leading to severe wilting symptoms.EPS production requires the eps gene cluster (epsABCDEF) and is regulated by multiple activators (e.g., XpsR, VsrAD, and VsrBC) and one repressor (e.g., EpsR).Deletion of phcX resulted in reduced EPS production (Fig. 2f) and a mild down-regulation of the entire eps gene cluster, which might be the integrated output of up-regulated expression of both epsR (log 2 FC = 2.67) and xpsR (log 2 FC = 1.33) (Table S3E and F).
After invading host intercellular spaces, R. solanacearum uses T3SS to secrete an arsenal of T3Es to subvert plant defense responses and promote virulence (20).It is proposed that the pathogens sense plant signals by PrhA and transduce the signals to PrhJ via PrhI/R; PrhJ affects the expression of HrpG, which positively regulates HrpB together with PrhG; HrpB directly controls the expression of T3SS and T3Es genes (39,63).The T3SS regulatory network also contains other regulators, such as PrhN, PrhO, PrhP, and CysB (64).PhcA represses the expression of PrhI/R directly, thus negatively regulating HrpG and, in turn, T3SS/T3Es (65).In our study, ΔphcX exhibited delayed HR development and attenuated virulence (Fig. 3a), accompanied by substantially lower expression of PrhJ/HrpG/HrpB (Table S3A and E) and a global down-regulation of T3SS and T3Es, as well as the HrpB-activated operon of six HDF genes responsible for the synthesis of a tryptophan derivative (Table S3E).Overall, these results suggest that PhcX plays a critical role in the plant-pathogen interaction and may activate the expression of T3SS/T3E genes via both PrhJ-and HrpG-mediated pathways.
Our results also provide insights into the relationships between PhcX and other regulatory mechanisms of R. solanacearum (Fig. 8).Several gene regulation-related GO terms (molecular transducer activity and signaling) and KEGG pathways (two-component system), as well as the gene set on QS, were enriched among genes differentially expressed in ΔphcX.Importantly, our transcriptome analyses revealed a significant overlap between the regulatory ranges of PhcX and the master regulator PhcA, including many of the abovementioned genes associated with ΔphcX phenotypes.Furthermore, PhcX and PhcA showed contrasting regulatory modes on most shared targets, and their mutants exhibited phenotypic alterations in opposite directions (Fig. 5 and 6).For instance, genes in the solIR and rasIR QS systems were downstream of both regula tors but were repressed by PhcX and activated by PhcA (Table S4C).Intriguingly, the expression of phcA and phc QS genes was not affected in ΔphcX (Table S4B), suggesting that PhcX and PhcA may execute their regulatory functions independently, or one may regulate the other at post-transcription levels.In particular, it has been proposed that the production of active PhcA is post-transcriptionally regulated by PhcR (25).The regulatory relationship between PhcX and PhcA remains to be further investigated, for instance, by examining the protein abundance of PhcA (or PhcX) in ΔphcX (or ΔphcA).
Interestingly, our RT-qPCR results (Fig. 5h and i) showed that the regulatory effects of PhcX were more evident at an OD 600 of 0.3 than at an OD 600 of 0.5.Given that ΔphcX exhibited a prolonged lag phase and the traits positively regulated by PhcX are generally important for the early stages of infection, it is possible that PhcX might function early on during the transition from survival and early invasion mode to high virulence status.It remains to be further investigated if PhcX acts after PhcA or responds to other QS signals.
On the other hand, although the phc QS pathway is normally activated at a much lower cell density (OD 600 <0.1), the |log 2 FC| values of most examined genes were greater at an OD 600 of 0.5 than at an OD 600 of 0.3 in ΔphcA.This pattern reflects the dynamic nature of gene regulation in R. solanacearum and suggests that the regulatory functions of genes like phcX and phcA might need to be investigated at multiple cell densities from low to high to achieve a more complete understanding.
Although phcX was first discovered and investigated in R. solanacearum EP1, we also showed in GMI1000 that the deletion of phcX led to largely similar phenotypes as in EP1 (Fig. S4).In addition, our genomic analysis revealed that phcX is present and clustered with the phcBSRQ operon in nearly all Ralstonia genomes and PhcX is highly conserved at the protein sequence level (average identity >96%) (Table S2).These lines of evidence all support the idea that the functions of phcX might be relatively conserved in RSSC.
Outside RSSC, phcX is also broadly present in more than 300 Proteobacteria genomes (Table S5; Fig. S5).Notably, we found that phcX was orthologous to lqsR, encoding the response regulator LqsR of the Lqs QS system in L. pneumophila, the causal agent of Legionnaires' disease.Lqs QS system plays important regulatory roles in L. pneumo phila to modulate its life cycle, pathogen-host cell interactions, motility, and natural competence (55).Besides LqsR, the Lqs system also consists of the pyridoxal-5-phos phate-dependent autoinducer synthase LqsA and two homologous membrane-bound sensor histidine kinases LqsS and LqsT (66).The Lqs and Phc QS systems have completely different signals and genes, and they coexist in numerous Betaproteobacteria genomes, including Cupriavidus spp.and Ralstonia insidiosa, which are closely related to RSSC.Interestingly, phcX/lqsR is clustered with lqsA and lqsS in most cases but becomes disassociated with lqs and linked to the phc gene cluster instead in Cupriavidus spp.and R. insidiosa, and lqsA is further lost in RSSC and most other Ralstonia (Fig. 7b).It was shown recently in Cupriavidus taiwanensis that both Lqs and Phc are functional (67); however, the Lqs system seems to have few influences on growth rate, swimming motility, biofilm formation, or EPS production, unlike the Phc system in C. taiwanensis or the R. solanacearum PhcX examined in this study.Altogether, the evolutionary dynamics of Phc and Lqs as well as the functional characterization of PhcX suggest that, during the evolution of Ralstonia, PhcX might have experienced network rewiring and acquired new regulatory roles along with the breakdown of the Lqs system.
Strikingly, the predicted structures of PhcX and LqsR are almost identical (Fig. 7a), indicating highly similar molecular functions.A recent structural analysis of mono meric LqsR suggested that the putative output domain of LqsR forms a five-stranded antiparallel β-sheet fold, similar to the diguanylate cyclase domain of PleD.LqsR also has a degenerate form of the conserved GG [DE][DE]F sequence pattern, which is a signature of diguanylate cyclases, suggesting that the output function of LqsR might be related to c-di-GMP metabolism and the signaling network (54).However, the study did not detect any diguanylate cyclase activity of LqsR in vitro.Therefore, it awaits further investigation to elucidate the regulatory mechanisms of PhcX and LqsR.

Bacterial strains, plasmids, and growth conditions
The bacterial strains and plasmids used in this study are presented in Table S1B.All R. solanacearum strains were routinely grown in CTG-rich medium (casein acid hydrolysate 1 g/L, tryptone 10 g/L, glucose 5 g/L) (68) at 28°C.CTG medium was supplemented with 0.005% 2,3,5-triphenyltetrazolium chloride (final concentration) in agar plates.Escheri chia coli strains were grown in Luria-Bertani medium at 37°C.Antibiotics were added at the following final concentrations (µg/mL) when necessary: kanamycin (50), gentamicin (50), and rifampicin (25).The competent cells of R. solanacearum strains for homologous recombination were generated by growing in MM broth with 10% glycerol.Bacterial growth was determined by measuring the optical density at a wavelength of 600 nm.

In vitro passage experiment and detection of mutation
For each passage, a single colony of the EP1 strain was grown overnight in a liquid CTG medium at 28℃.One hundred microliters of the suspension was transferred onto a CTG agar plate and spread uniformly.The plate was inoculated for 48 h at 28℃.Freshly grown bacterial populations were then washed down with 1 mL sterile water.Next, 100 µL of the suspension was spotted onto a new CTG agar plate for the next round as described above, and another 100 µL was used for serial dilutions to detect EPS-deficient mutants.Dark red granular colonies with reduced EPS production were chosen as EPS-deficient mutants and further confirmed on new CTG agar plates.To detect the mutation(s) causing the EPS-deficiency, genomic DNA of the ancestral wild-type strain and mutants was extracted using the EasyPure Bacteria Genomic DNA Kit following the manufactur er's instructions, and genome sequencing was conducted on the HiSeq X Ten system to generate 150 bp pair-end reads at more than 200× depth.Delly v0.8.5, SAMtools v1.1, and BCFtools v1.10.2 were used to detect structural variants, copy number variations, and single nucleotide variations, respectively (69).

Generation of in-frame deletion mutants and complementation
The phcX and phcA deletion mutants were generated from R. solanacearum EP1 and GMI1000 following the methods described previously (70).Briefly, the upstream and downstream flanking regions of their target genes and the gentamicin resistance gene sequence were PCR amplified from the R. solanacearum genome and plasmid pK18mob sacB, respectively.Then, these three fragments were conjugated with joint PCR, in which the gentamicin resistance gene sequence was added between the upstream and downstream flanking fragments by using the overlap primers.DNA fragments were mixed with the corresponding competent wild-type R. solanacearum strains and plated on CTG medium agar plates supplemented with gentamicin for mutant selection.The transformants were confirmed by PCR using primers and DNA sequencing.Genome resequencing was conducted for further verification to ensure that the target gene was deleted without introducing any other mutations in the genome.For complementa tion analysis, the native promoter sequence and coding regions of target genes were amplified and cloned into the HindIII (5′ end) and EcoRI (3′ end) restriction sites available in the multiple cloning region of the pBBR1-MCS2 plasmid.The HindIII restriction site is 92 nucleotides upstream of the start codon of the lacZ gene.The complementary plasmid was electroporated into the corresponding competent mutants.Primers used for gene deletion and complementation are listed in Table S1.

Bacterial growth analysis under different nutrition conditions
R. solanacearum was cultured in CTG medium at 28°C until it reached the logarithmic growth phase.For growth analysis in CTG medium, the bacterial suspension was diluted to an OD 600 nm of 0.1 with fresh CPG medium and then inoculated into new CTG medium with a ratio of 1:100.For growth analysis in ground and filtered tomato stems or MP medium (39), the bacterial suspension was washed twice with sterilized water.Bacterial cells were harvested by centrifugation at 4,000 rpm for 10 min at 4°C.Then, the bacterial pellet was resuspended into ground and filtered tomato stems or MP medium containing 50 mM sole carbon source (one of glutamine, glucose, proline, fructose, or myoinositol) and adjusted to an OD 600 nm of 0.01.Ground and filtered tomato stems were obtained by grinding the fresh tomato stem and filtering through a 0.25-µm membrane after centrifugation at 15,000 rpm for 2 min.The growth curves were recorded using the Bioscreen-C automated growth curve analysis system or detected manually by UV-visible spectroscopy at periodic intervals.The corresponding media without bacteria were used as negative controls.The exponential growth rate and lag phase duration were determined using the R package growthrates v0.8.4 (https:// github.com/tpetzoldt/growthrates).Three independent experiments were performed with at least five replicates each time.

Swimming and twitching motility
Swimming motility was assessed on a semisolid medium plate with 0.3% agar.Then, 1 µL of bacterial suspension was stabbed into the center of plates containing 0.5% tryptone, 0.3% yeast extract, and 0.3% agar.Photos were taken, and the swimming-area diameters were measured after 24 h of incubation at 28°C.For twitching motility, it was performed as previously reported with some modifications (10).In brief, 2 µL of bacterial suspension adjusted to an OD 600 of 0.1 was inoculated onto the surface of CTG plates with 1.6% agar.Colonies were examined, and photos with four times magnification were taken under a Leica stereomicroscope system at 24 h post-inoculation.Three independent experiments were performed with at least three replicates each time.

Cellulase activity and polygalacturonase activity
Cellulase activity and polygalacturonase activity were tested in cellulase detection medium (1 g carboxymethyl ethyl cellulose, 3.8 g Na 3 PO 4 , 8 g agarose, pH 7.0, per liter) and polygalacturonase detection medium [5.0 g polygalacturonic acid, 2.0 g sucrose, 2.0 g (NH 4 ) 2 SO 4 , pH 5.5, per liter], respectively.The bacterial culture was incubated to an OD 600 of approximately 1.0, and 20 µL of the bacterial suspension was inoculated into the detection plates.The plates were incubated at 28°C for 48 h.The cellulase detection plates were stained with 0.5% Congo red for 30 min and then washed three times with 1 M NaCl.The polygalacturonase detection plates were stained with 1 M HCl for 15 min and then washed with 1 M NaCl.Then, the diameter of the transparent ring was measured, and photos were taken.Three independent experiments were performed with at least three replicates each time.

Biofilm formation
The biofilm formation assay was performed in 24-well polystyrene plates as reported previously, with some modifications (8).In brief, bacterial suspensions were first adjusted to an OD 600 of 0.1 with fresh CTG medium and then inoculated 1:20 with new CTG medium in 24-well polystyrene plates.The samples were statically incubated at 28°C for 36 h.The suspension was first mixed well, measured at OD 600 , and carefully removed.Then, 1.5 mL of 0.1% crystal violet was added to each well, and the wells were stained for 30 min.The crystal violet-stained bound cells were repeatedly rinsed with water three times and then dissolved in 95% ethanol.The absorbance at 570 nm of the solution was measured and normalized with OD 600 to quantify biofilm formation.Three independent experiments were performed with six replicates each time.

EPS production
EPS production in culture broth supernatant of R. solanacearum cells was determined by the Elson-Morgan method as described by Peyraud et al. (52).R. solanacearum was cultured in CTG medium at 28°C until it reached the logarithmic growth phase.The bacterial suspension was washed twice with MP medium, inoculated at an OD 600 of 0.01, and cultured in MP minimal medium supplemented with L-glutamine (20 mM) until an OD 600 of 1.0 at 28°C.The culture broth was then filtered with a 0.22-µm filter, and the supernatant was collected in a 2.0-mL microcentrifuge.Then, 0.4 mL of supernatant was mixed with 0.008 mL of 5 M NaCl solution and fourfold volumes of acetone.The mixture was stored at 4°C for 12 h to precipitate exopolysaccharide.Precipitated EPS pellets were dissolved in 0.4 mL H 2 O and heated at 65°C on a heat block for 10 min.Then, 0.15 mL concentrated HCl (37%) and 0.05 mL H 2 O were added.The tubes were closed tightly by adding bridges and mixed with a vortex.Tubes were transferred to the dry bath heater and maintained in the dry bath heater at 115°C for 30 min.Samples were cooled to room temperature.Then, 0.4 mL of 2 M Na 2 CO 3 was added and homogenized, followed by 0.5 mL of 2% acetyl acetone in 1.5 M Na 2 CO 3 .The samples were then heated for 20 min with a heat block at 100°C.The samples were cooled to room temperature and transferred into 10 mL tubes.Then, 1.0 mL of 99.8% ethanol was then added and mixed well.In addition, 0.5 mL of Erlich's reagent solution was added slowly, and the tubes were left to stand for about 30 min in the dark.The optical density of the solution was measured at 530 nm.Quantification of hexosamine was carried out by running a standard curve of N-acetyl-galactosamine that followed the overall process except precipitation.

HR assay
The HR assay was performed on tobacco leaves with leaf infiltration.The bacterial suspension cultured in CTG medium was washed twice and adjusted to an OD 600 of 0.1 with sterilized water.Then, 10 µL of bacterial suspension was inoculated into leaves with blunt syringes, and the development of necrotic lesions was recorded periodically.

Virulence assays
About 6-week-old tomato plants of cultivar Zhongshusihao were inoculated with R. solanacearum strains according to stem injection and root dipping methods.The bacterial suspension cultured in CTG medium was washed twice and adjusted to an OD 600 of 0.1 with sterilized water.Toothpicks were used to create a wound on each tomato stem about 3 cm above the root.Then, 2 µL of bacterial suspension was inoculated into the wound.Sterilized water without bacteria was used as the nega tive control.For the root dipping method, tomato seedlings were soaked in bacterial suspensions (OD 600 of 1.0) for 30 min and then exposed to air for 15 min on dry paper.Inoculated plants were grown in a plant incubator with a 12-h light and 12-h dark cycle at 28°C with 80% relative humidity.The number of dead plants was recorded daily.

RT-qPCR and RT-PCR
Total RNA was extracted from R. solanacearum strains grown in CTG medium of specific OD 600 by the use of the Eastep Super Total RNA Extraction Kit (Promega).Thereafter, cDNA synthesis and RT-qPCR analysis were performed with the ChamQ Universal SYBR qPCR Master Mix (Vazyme) on the Applied Biosystems QuantStudio 6 Flex Real-Time PCR System (Thermo Fisher Scientific) following the manufacturer's instructions.recA was used as an internal standard for each cDNA sample, and the relative level of gene expression was calculated using the ΔΔCt method.Primers used for RT-qPCR analysis were designed by the online tool PrimerQuest (https://sg.idtdna.com/PrimerQuest/Home/Index) and listed in Table S1.As for RT-PCR, once the total RNA was extracted, DNA contaminant elimination and cDNA synthesis were performed following the instructions (TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix).PCRs were performed using standard procedures with an RNA sample without reverse transcriptase, genomic DNA, and cDNA as templates.Primers used for RT-PCR were listed in Table S1C.

RNA-seq analysis
Total RNA was extracted from the wild-type EP1 and the phcX deletion mutant strains growing in CTG medium with a comparable OD 600 of 0.3 using the Eastep Super Total RNA Extraction Kit (Promega).For each strain, three biological replicates were analyzed.Genomic DNA contamination was removed using RNasefree DNase I (Trans gen) according to the manufacturer's instructions.Total RNA was sent to Novogene (Beijing, China) for sequencing on the Illumina (CA, USA) NovaSeq 6000 platform.The raw reads were quality filtered with fastp v0.21.0 (71), and the resulting clean reads were mapped to the reference genome with bowtie2 v2.3.5.1 (72).FeatureCounts v2.0.2 was used to quantify the gene expression level (73).PCA and differential expression analysis were carried out with DESeq2 v1.28.1 (74).Genes with |log 2 FC| ≥1 and FDR <0.05 were considered significantly differentially expressed.For functional annotation, each protein sequence of EP1 was searched against the Uniport database using BLASTP v2.9.0+, and the GO annotations of close homologs (identity >80% and qcovhsp >80%) were transferred to the query sequence.KEGG annotation was performed by eggnog-mapper v2.1.4-2(75).ClusterProfiler was used to perform gene over-representation analysis and GSEA (76).

Statistical analysis
Statistical analyses were performed with Prism 9 software (GraphPad).Results are displayed as the mean ± standard deviation.An unpaired Student's t-test was carried out to compare the significance between R. solanacearum strains, and the P-value was used to indicate the statistical significance of the differences.* indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001, and **** indicates P < 0.0001.

FIG 1
FIG 1 Growth curves of R. solanacearum wild-type EP1, ΔphcX, and phcX-comp strains.(a) Growth curves in CTG medium.(b) Growth curves in MP medium with L-glutamine to a final concentration of 50 mM.(c) Growth curves in ground and filtered tomato stems.Growth curves of wild-type EP1 (d), ΔphcX (e), and phcX-comp (f) in CTG medium and CTG medium with ground and filtered tomato stems to a final concentration of 0.1% (vol/vol).Rectangular lines in panels d, e, and f show the point where each strain started to exhibit significantly higher cell density in CTG medium with 0.1% tomato xylem sap than in pure CTG medium.The experiments were repeated three times, with five technical replicates per experiment.The data were presented as mean ± standard deviation.Asterisks indicate that the cell density values at the same time point were significantly different (****P < 0.0001, multiple unpaired t-test).

FIG 2
FIG 2 Phenotypic characterization of R. solanacearum wild-type EP1, ΔphcX, and phcX-comp.The phenotypes examined include (a) swimming motility, (b) twitching motility, (c) polygalacturonase activity, (d) cellulase activity, (e) biofilm formation, and (f) quantitative measurements of EPS production.Swimming and twitching motility assays were conducted on 0.3% and 1.6% agar plates, respectively.Representative optical microscope images (4× magnification) of twitching motility assays were taken.Polygalacturonase and cellulase activities were detected on bioassay plates, and the final results of ΔphcX and phcX-comp were normalized to those of the wild-type EP1, which was set to a value of 100% for comparison.Biofilm formation was assessed in 24-well polystyrene plates, quantified with absorbance at 570 nm, and normalized with the cell density as measured by OD 600 .EPS production in the culture broth supernatant of R. solanacearum cells at an OD 600 of 1.0 was determined by the Elson-Morgan method, and results were shown in the concentration of N-acetyl-galactosamine in mM.The numbers (in h) on top of panel (f) indicate the time it took for each strain to reach an OD 600 of 1.0 from an initial OD 600 of 0.01.All experiments were repeated at least three times, with at least three technical replicates per experiment.The data were presented as mean ± standard deviation.Asterisks indicate values significantly different from those of wild-type EP1 (ns, no significance; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001, unpaired t-test).

FIG 3
FIG 3 HR development and virulence of R. solanacearum strains.HR development (a) induced by R. solanacearum wild-type EP1, ΔphcX, and phcX-comp was tested on tobacco leaves.Photographs were taken at the indicated times after inoculation.Double-distilled H 2 O was used as the negative control.Tomato plants were inoculated with R. solanacearum wild-type EP1 and ΔphcX via stem injection (b) and root dipping (c) methods.For stem injection, a 2-µL bacterial suspension was inoculated into the stem wound created by toothpicks.For root dipping inoculation, tomato seedlings were soaked in bacterial suspensions (OD 600 of 1.0) for 30 min and then exposed to air for 15 min on dry paper.The number of dead plants was recorded daily, and Kaplan-Meier survival analysis of plants inoculated with R. solanacearum strains was carried out (***, P < 0.001; ****, P < 0.0001, Gehan-Breslow-Wilcoxon test).

FIG 4
FIG 4 Differential expression patterns of genes involved in various virulence-related functional categories in ΔphcX.GSEA was carried out based on the gene sets of R. solanacearum compiled from multiple sources and the differential gene expression data generated in the RNA-seq analysis of the wild-type EP1 and ΔphcX (both cultured in CTG medium until an OD 600 of 0.3).

FIG 5
FIG 5 phcX and phcA regulate multiple phc QS-dependent phenotypes and associated genes in opposite patterns.Phenotypic characterizations of R.solanacearum wild-type EP1, ΔphcX, and phcA include growth curves in MP medium with 50 mM glucose (a), proline (b), fructose (c), or myoinositol (d) as the sole carbon substrate, swimming motility (e), polygalacturonase activity (f), and biofilm formation (g).RT-qPCR analysis of selected genes in multiple functional categories in ΔphcA and ΔphcX cultured in CTG medium until an OD 600 of 0.3 (h) and an OD 600 of 0.5 (i).The recA gene was used as an internal control for RT-qPCR.The differential expression results of mutant strains were shown in log 2 (fold change) compared with the wild-type strain.RT-qPCR assays were conducted at two cell densities to validate our RNA-seq results of ΔphcX (OD 600 of 0.3) and the reported transcriptome data of ΔphcA (OD 600 of 0.5), respectively.The experiments were repeated at least three times, with three technical replicates per experiment.The data were presented as mean ± standard deviation.

FIG 6
FIG 6 Comparison between the differential gene expression patterns of ΔphcX and ΔphcA.(a) The Venn diagram depicts the numbers of DEGs (|log 2 FC| ≥ 1, FDR <0.05) shared by both ΔphcX (blue) and ΔphcA (yellow) or unique to either mutant.Upward/downward arrows indicate up-/down-regulated genes, respectively.(b) The scatter plot shows, for all one-to-one orthologous genes between EP1 and GMI1000, the relative expression patterns of each gene in ΔphcX (x-axis) and ΔphcA (y-axis).Dots in red stand for DEGs detected in both ΔphcX and ΔphcA, and the dashed line shows the correlation (r = −0.70,P < 10 −15 ).(c) Heat maps show, for genes in selected virulence-related functional categories, the differential expression patterns of each gene in ΔphcX and ΔphcA.The blue and yellow colors correspond to down-and upregulation, respectively.

FIG 7
FIG 7 Genomic analysis suggests that PhcX likely originated as part of the Lqs QS system.(a) Predicted protein structures of the PhcX homodimer in R. solanacearum EP1, the LqsR homodimer in L. pneumophila C9_S, and their structural alignment.The protein structures were predicted using AlphaFold2 and visualized and aligned using PyMOL.The yellow dashed-line circle highlights the diguanylate cyclase (GGDEF) output domain of PhcX and LqsR in their homodimer structure models.(b) The presence/absence patterns and genomic arrangements of phc and lqs genes in Proteobacteria.The genus-level phylogeny shown on the left was extracted from Genome Taxonomy Database.For each genus, the representative genomic arrangement was shown.Each phc or lqs gene

FIG 8
FIG 8 Schematic representation of the PhcX and PhcA regulatory networks in R. solanacearum EP1.Many genes associated with virulence (e.g., genes related to motility, chemotaxis, and genes encoding polygalacturonases, T3SS and T3Es, lectins, and adhesins), metabolism (e.g., genes encoding nitrate reductases and cbb3-type cytochrome c oxidases), and regulation (e.g., genes in AHL QS systems) are regulated by both PhcX and the global virulence regulator PhcA, but in opposite directions.PhcX might interact indirectly with the phc QS system through the down-regulation of lecM, which positively affects the stability of 3-OH MAME, the phc QS signal.It remains to be investigated if other indirect or direct regulations exist between PhcX and PhcA.Dashed lines indicate potential indirect regulation.Lines ending with an arrowhead indicate positive regulation, while lines ending with a crossbar indicate negative regulation.