Abstract Opportunistic fish pathogen Vibrio scophthalmi frequently infects olive flounder and turbot, which are primary marine species cultured for seafood production in Far East Asia. These infections cause substantial yield reductions and significant economic losses. Although quorum sensing (QS) genes were previously reported in V. scophthalmi, the impacts of QS on genome-wide gene expression and consequent behaviors and physiological traits have remained largely unexplored. In this study, we conducted genomic and transcriptomic analyses to uncover the global regulatory network governed by LuxR[Vs], a QS master regulator in V. scophthalmi. By comparing the wild-type strain and a luxR[Vs] deletion mutant strain, we found that LuxR[Vs] positively regulates biosynthetic genes for poly-hydroxyalkanoate (PHA) while negatively controlling genes for biofilm formation. Quantification of intracellular PHAs and biofilm biomass on borosilicate tubes confirmed these results. Gene set enrichment analyses further demonstrated that LuxR[Vs] also governs genes related to osmoprotection and defense against reactive oxygen species. Overall, these findings indicate that LuxR[Vs] acts as a global transcriptional regulator, controlling a wide range of physiological processes in V. scophthalmi. Targeting LuxR[Vs] could therefore be a promising strategy for improving seafood production by disrupting diverse physiological and pathogenic traits in this fish pathogen. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-024-78803-7. Keywords: Quorum sensing, Vibrio scophthalmi, Fish pathogen, Seafood production, LuxR regulon Subject terms: Pathogens, Bacterial genomics, Marine microbiology, Bacteriology, Biofilms Introduction The ongoing climate crisis significantly impacts on sustainable food production. Limited land for agriculture and stock-farming further worsens the existing problem^[28]1. Therefore, various strategies for sustainable food supplies are under intensive investigation, such as meat alternatives and cultured meats^[29]2. However, given that approximately 17% of edible meats currently come from various seafoods, it is crucial to reconsider the potential of aquaculture as a reliable alternative to land-derived food production. In fact, it has been predicted that global seafood production will increase from 59 megatons to 80–103 megatons by 2050^[30]3. Among the various species cultivated in aquaculture, the olive flounder (Paralichthys olivaceus) holds a significant position in Far East Asian countries, such as South Korea and Japan. Particularly in South Korea, it takes the lead as the primary cultured marine fish, reaching a production of 46,000 tons in 2022, constituting 50.5% of the total aquaculture production in the country^[31]4. Meanwhile, other countries, like the United States, are also recognizing the olive flounder as a promising candidate for aquaculture due to its high market value and advantageous traits for domestic cultivation. In fact, P. olivaceusexhibits an exceptional food conversion ratio compared to other flatfish species, and therefore requires a relatively short period of time for growth^[32]5. However, olive flounder cultivation often encounters issues that significantly impact its yields, with one of the major reasons being infectious diseases caused by various microbial agents, including bacteria, viruses, and parasites^[33]6–[34]8. Among these, Vibrio scophthalmi has been identified as an opportunistic bacterial pathogen, being the most prevalent within the intestines of diseased olive flounder and turbot^[35]9–[36]11. Indeed, when olive flounder is subjected to environmental stress, compromising its immunity, the presence of V. scophthalmiincreases the fish’s vulnerability^[37]12,[38]13. Similarly, multispecies infection by V. scophthalmi and another fish pathogen, Streptococcus parauberis, significantly raises the mortality rate of the fish compared to infections with either species alone^[39]13. Bacterial pathogens often communicate with each other to synchronize their gene expressions for group behaviors, such as the production of virulence factors^[40]14–[41]16. This phenomenon, known as quorum sensing (QS), involves the bacterial production and release of small molecules called autoinducers (AIs), which enable bacteria to detect their population density^[42]17,[43]18. In Vibrio harveyi and Vibrio cholerae, for example, multiple AIs are synthesized and detected by distinct transmembrane receptor proteins. At low cell density (LCD), these transmembrane receptors undergo autophosphorylation, transferring a phosphate group to the common response regulator, LuxO, via the phospho-relay protein LuxU. Subsequently, phospho-LuxO, together with σ-54, activates the transcription of small regulatory RNAs, which in turn down-regulates the expression of the quorum sensing master regulator, LuxR. At high cell density (HCD), AI binding to the receptors reverses the phosphate flow in the cell, leading to the translational production of LuxR proteins^[44]19–[45]21. It is noteworthy that LuxR homologues are often referred to by various names in different species, such as LuxR[Vh] in V. harveyi, HapR in V. cholerae, and SmcR in V. vulnificus^[46]22–[47]24. Previous studies have suggested that V. scophthalmiproduces two AIs and possesses a QS system similar to that of other vibrios^[48]25,[49]26. Indeed, a LuxR homologue in V. scophthalmi (hereafter referred to as LuxR[Vs]) exhibits 81.86% identity with V. vulnificusSmcR at amino acid sequence level^[50]26,[51]27. Biofilm formation has been proposed as a QS-regulated phenotype in V. scophthalmi, but scarcely genetic or molecular basis has been indicated^[52]26. In addition, although QS regulon have been explored in some clinically relevant Vibriospecies^[53]28–[54]31, no genome-wide analysis exploring LuxR-mediated QS regulon has been conducted for this important fish pathogen. Consequently, it remains ambiguous how diverse genes and physiologies are under the control of QS in V. scophthalmi, which significantly affecting olive flounder production. To address this gap, here we analyzed the chromosome of the V. scophthalmi 21FBVib0096, isolated from a diseased olive flounder. Subsequently, we conducted a genome-wide transcriptomic analysis comparing the wild-type and the luxR[Vs] deletion mutant (ΔluxR[Vs]). Through these analyses, we successfully identified numerous LuxR[Vs] regulon. By employing the gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses, we further gained valuable insights into the functional implications of LuxR[Vs]-mediated QS in V. scophthalmi. The results represent an initial step toward a deeper understanding of the pathogen’s physiology, which could be modulated to reduce its pathophysiological traits and thus improve seafood production. Results Genomic information of V. scophthalmi 21FBVib0096 The genome of V. scophthalmi 21FBVib0096, isolated from a diseased P. olivaceus by the National Institute of Fisheries Science in 2021, was sequenced and analyzed following the procedures summarized in the [55]Materials and Methods section. The genome consisted of two circular chromosomes [chromosome I (3,352,351 bp, GC content 44.95%) and chromosome II (1,419,321 bp, GC content 44.01%)], along with one plasmid (100,413 bp, GC content 44.13%) (Fig. [56]1). In chromosome I, a total of 3,193 genes were identified, including 3,060 coding sequences (CDSs), 95 tRNA genes, 37 rRNA genes, and 1 tmRNA. Chromosome II contained 1,300 genes with 1,290 CDSs and 10 tRNA genes. The plasmid comprised 124 genes with 123 CDSs and 1 tRNA gene. Overall, out of 4,473 CDSs in the whole genome, 2,807 (62.75%) were annotated as functional genes, while the remaining, 1,666 CDSs (37.25%) were categorized as hypothetical or uncharacterized proteins. Gene annotation of V. scophthalmi 21FBVib0096 revealed the presence of multiple QS components, including LuxU and LuxO, indicating that the pathogen likely possesses a functional QS system (Supplementary Table [57]S1). Among, Vs096_03041, a gene encoding QS master regulator LuxR[Vs], is in the chromosome I. Accordingly, this gene was targeted for deletion mutation, and the resulting ΔluxR[Vs] strain was used for subsequent transcriptomic analyses. Fig. 1. [58]Fig. 1 [59]Open in a new tab Genome of V. scophthalmi 21FBVib0096. Two chromosomes and one plasmid are shown. The outermost circle indicates positions in megabase (for chromosomes) or kilobase (for plasmid) scale. The second circle illustrates coding sequences on both the forward and reverse strands. The third circle denotes tRNA (blue) and rRNA (red) positions. The fourth circle displays GC-skew [(G-C)/(G + C)], with green and red bands denoting positive and negative values, respectively. The fifth circle represents the GC-content, where blue and yellow bands indicate values above or below 50%, respectively. Identification of LuxR[Vs ]regulon via RNA-sequencing To identify LuxR[Vs] regulon, we compared the whole transcriptomes of ΔluxR[Vs] and wild-type V. scophthalmi under HCD condition (MT-HCD vs. WT-HCD; in the Supplementary Figure [60]S1) via RNA-sequencing. In the meantime, we also compared the transcriptomes of the wild-type strain under both LCD and HCD conditions (WT-LCD vs. WT-HCD; Supplementary Fig. [61]S1) to evaluate the influence of growth status on gene expression patterns. Each sample was prepared in triplicate, generating an RNA-sequencing dataset comprising a total of 549,231,520 reads, of which 436,286,522 reads successfully mapped to the V. scophthalmi genome. Quality assessment indicated high Q20 and Q30 percentages, surpassing 98% and 94%, respectively (Supplementary Table [62]S2). Notably, the gene expression profiles of all three samples in each condition, except for MT-HCD3, exhibited a similar pattern in both multidimensional scaling (Supplementary Fig. [63]S2a) and hierarchical clustering analyses (Supplementary Fig. [64]S2b). Consequently, a total 8 transcriptome profiles (MT-HCD1 and 2; WT-HCD1-3; and WT-LCD1-3) were used in further analyses. In the comparison of MT-HCD vs. WT-HCD, 631 differentially expressed genes (DEGs; |fold change| ≥ 1.75 and p ≤ 0.05) were identified, comprising 366 up-regulated and 265 down-regulated genes in the MT-HCD compared to the WT-HCD (Fig. [65]2a, Supplementary Fig. [66]S3a, and Supplementary Dataset S1). In the comparison of WT-LCD vs. WT-HCD, a total of 2,121 genes exhibited differential expression, with 1,024 up-regulated and 1,097 down-regulated genes in the WT-LCD compared to the WT-HCD (Fig. [67]2a, Supplementary Fig. [68]S3b, and Supplementary Dataset S2). Further comparative analysis of the two comparisons revealed a set of 332 overlapping DEGs. Among these, 94 genes exhibited an up-regulated pattern in both comparisons, while 115 genes exhibited a down-regulated pattern in both comparisons (Fig. [69]2b). In the meantime, the remaining 123 genes showed a contrasting expression pattern in the two comparisons, where 85 genes exhibited an up-regulated pattern in the MT-HCD vs. WT-HCD but a down-regulated pattern in the WT-LCD vs. WT-HCD, and 38 genes exhibited vice versa (Fig. [70]2b). The list of overlapping DEGs for each category is provided in Supplementary Dataset S3. Fig. 2. [71]Fig. 2 [72]Open in a new tab Overview of differentially expressed genes (DEGs) identified by RNA-sequencing. (a) The total number of DEGs (gray), as well as up-regulated (red) and down-regulated (blue) DEGs revealed in the MT-HCD vs. WT-HCD comparison and the WT-LCD vs. WT-HCD comparison, are shown. (b) A Venn diagram illustrating the number of DEGs from both comparisons is shown. The overlapping DEGs were further categorized into coincidentally up- or down-regulated DEGs and contrastingly regulated DEGs, with their respective numbers depicted as bar graphs. (c and d) Validation of RNA-sequencing data via qRT-PCR analysis. The fold changes in expression of the indicated genes in either the MT-HCD (c) or WT-LCD (d) sample are presented relative to those in the WT-HCD sample. The recA gene was used as internal reference for qRT-PCR. To further validate the differential gene expression patterns observed, thirteen genes representing the aforementioned categories were selected for quantitative Real-Time PCR (qRT-PCR) analysis. Specifically, bccT3 and Vs096_04402 were selected from the LuxR[Vs] regulon, while cysK and hutU were chosen to represent growth status-dependent genes. From the overlapping DEGs, Vs096_01435, guaB, and Vs096_03709 were selected as coincidentally up-regulated genes, while degP, phaC, and phaB were selected as coincidentally down-regulated genes. For the contrastingly regulated DEGs, Vs096_00748, vpsT, and Vs096_04425 were chosen. As shown in Fig. [73]2c and d, the qRT-PCR results were highly consistent with the RNA-sequencing data, further confirming the reliability of our whole transcriptome sequencing analysis. Overall, the transcriptome analyses revealed the LuxR[Vs] regulon, comprising 631 genes, of which 332 are co-regulated by the cell growth status, whereas 299 are part of the growth independent LuxR[Vs] regulon. Functional profiling of LuxR[Vs ]regulon To elucidate the regulatory role of LuxR[Vs] and how cell growth status affects LuxR[Vs]-mediated physiologies, we functionally classified the identified DEGs through GO and KEGG pathway analyses. The GO enrichment analysis for DEGs from the MT-HCD vs. WT-HCD comparison revealed up-regulation of osmoprotection-related molecular process and membrane transporter activities in the MT-HCD. These included glycine betaine transport, dicarboxylic acid transport, symporter activity, carbohydrate transport, and carbohydrate transmembrane transporter activity (Fig. [74]3a). Notably, diguanylate cyclase activity was also significantly enriched as an up-regulated GO term in the MT-HCD compared to the WT-HCD (Fig. [75]3a), suggesting a suppression of cyclic bis-(3’, 5’)-dimeric guanosine monophosphate (c-di-GMP) biosynthesis by LuxR[Vs] in V. scophthalmi at HCD. In a meantime, the MT-HCD also exhibited significant enrichment of down-regulated biological processes and molecular functions compared to the WT-HCD. These down-regulated processes included unfolded protein binding, iron ion transport, oxidoreductase activity, and metabolic processes for tricarboxylic acids, dicarboxylic acids, and carbohydrate derivatives (Fig. [76]3b). Fig. 3. [77]Fig. 3 [78]Open in a new tab GO enrichment analysis for DEGs. Up-regulated (a and c) and down-regulated (b and d) DEGs from the comparisons of MT-HCD vs. WT-HCD (a and b) and WT-LCD vs. WT-HCD (c and d) were analyzed by GO enrichment. On the y-axis, GO terms are represented, while the x-axis shows the corresponding rich factor of each term. The color intensity and size of the points correspond to the p-value and the number of DEGs associated with each GO term, respectively. GO enrichment analysis for DEGs from the WT-LCD vs. WT-HCD comparison revealed the up-regulation of ribosome-related GO terms (Fig. [79]3c). These terms included cytosolic small/large ribosomal subunits and structural constituent of ribosome, suggesting relatively increased translational activity in early exponential phase cells (WT-LCD) compared to stationary phase cells (WT-HCD). Conversely, diverse GO terms, such as tricarboxylic acid cycle, oxidoreductase complex, respiratory electron transport chain, and peptide transport were down-regulated in the WT-LCD compared to the WT-HCD (Fig. [80]3d). Consistent with the GO analysis results, the KEGG pathway analysis identified relevant pathways. In the comparison of MT-HCD vs. WT-HCD, KEGG analysis revealed significant enrichment in a total of 140 pathways from 222 DEGs. Among these, the up-regulated genes in the MT-HCD compared to the WT-HCD significantly clustered into pathways involved in nitrogen metabolism, phosphotransferase system, ABC transporters, and biofilm formation (Fig. [81]4a). Conversely, the down-regulated genes in the MT-HCD compared to the WT-HCD were associated with pathways responsible for amino acid and carbohydrate metabolism (Fig. [82]4b). Fig. 4. [83]Fig. 4 [84]Open in a new tab KEGG pathway enrichment analysis for DEGs. Up-regulated (a and c) and down-regulated (b and d) DEGs from the comparisons of MT-HCD vs. WT-HCD (a and b) and WT-LCD vs. WT-HCD (c and d) were analyzed by KEGG pathway enrichment. On the y-axis, the pathway names are represented, while the x-axis shows the corresponding rich factor of each pathway. The color intensity and size of the points correspond to the p-value and the number of DEGs associated with each pathway, respectively. In the comparison between WT-LCD and WT-HCD, 883 DEGs were classified into 230 pathways. Significantly up-regulated pathways in the WT-LCD compared to the WT-HCD included ribosome and purine metabolism (Fig. [85]4c), while significantly down-regulated pathways encompassed the two-component system, quorum sensing, glyoxylate/dicarboxylate metabolism, and the citrate cycle (Fig. [86]4d). Notably, the pyruvate metabolism and butanoate metabolism pathways were identified as significantly down-regulated in both MT-HCD vs. WT-HCD (Fig. [87]4b) and WT-LCD vs. WT-HCD (Fig. [88]4d) comparisons. LuxR[Vs]-mediated up-regulation of poly-hydroxyalkanoate (PHA) biosynthesis Because the KEGG pathway enrichment analysis revealed significant down-regulation of butanoate metabolism in both comparisons (Fig. [89]4b and d), we further inspected the DEGs associated with this pathway. The butanoate metabolism is known to be interconnected with the citrate cycle (TCA cycle), glycolysis, and the metabolic pathways of amino acids and fatty acids^[90]32,[91]33. Consistent with this, genes encoding citrate cycle enzymes (e.g., fumarate reductase, Vs096_02777-80; succinate dehydrogenase, Vs096_01705-08 and Vs096_03353) and those involved in amino acid-related metabolic pathways (e.g., acetolactate synthase, Vs096_02022-23 and Vs096_02587) were significantly down-regulated in the MT-HCD and/or WT-LCD compared to the WT-HCD (Fig. [92]5a and Supplementary Datasets S1 and S2). Fig. 5. [93]Fig. 5 [94]Open in a new tab LuxR[Vs]enhances butanoate metabolism pathway and PHA production. (a) Schematic representation of the butanoate metabolism pathway with integrated PHA biosynthesis. Enzymes enriched in the KEGG pathway analysis are indicated with green boxes. Within these boxes, DEGs from the MT-HCD vs. WT-HCD and the WT-LCD vs. WT-HCD comparisons are annotated with yellow and pink dots, respectively, at the top-left corner. Among the overlapping DEGs from both comparisons, coincidentally down-regulated DEGs are indicated with blue frames. (Copyright Kanehisa laboratories; reproduced with permission^[95]55.) (b) A heat map illustrating the log[2] fold changes in the expression of key genes related to butanoate metabolism and PHA biosynthesis is presented for the MT-HCD vs. WT-HCD and the WT-LCD vs. WT-HCD comparisons. An asterisk (*) represents a statistically significant change in gene expression (|fold change| ≥ 1.75 and p ≤ 0.05). (c and d) PHA production in the wild-type, ΔluxR[Vs], and complemented strains were examined by Nile red plate assay (c) and fluorescence method (d). Normalized PHA production levels are presented as means ± standard deviation (SD) from three independent experiments. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparisons (***, p < 0.005; *, p < 0.05). Notably, the expression of PHA synthase gene and related genes, including acetyl-CoA acetyltransferase (Vs096_03255, phaA), acetoacetyl-CoA reductase (Vs096_03256, phaB), PHA synthase (Vs096_03253, phaC), and phasin (Vs096_03254, phaP), was also markedly suppressed in both MT-HCD and WT-LCD samples compared to the WT-HCD (Fig. [96]5a). Indeed, the fold change values of the phaBAPC genes in the MT-HCD were − 10.12, −9.00, −10.25, and − 6.30, and those in the WT-LCD were − 176.76, −156.99, −235.23, and − 70.29, when compared to the WT-HCD (Fig. [97]5b). These findings suggest that LuxR[Vs] and growth status mediate the regulation of PHA biosynthesis. To further investigate the role of LuxR[Vs] in PHA synthesis, we compared PHA levels in the wild-type, ΔluxR[Vs], and complemented strains. As shown in Fig. [98]5c, the wild-type and complemented strains, but not the ΔluxR[Vs] strain, exhibited intense pink fluorescent colonies on the Nile red containing plate under UV light, indicating substantial accumulation of PHA in the cells. Consistent with this result, quantitative analysis via Nile red fluorescence method confirmed that the ΔluxR[Vs] strain had the lowest PHA production level. Notably, complementation with the luxR[Vs] gene significantly enhanced PHA production over the wild-type level (Fig. [99]5d), likely due to the effects of multicopy complementation. When the smcR gene, a luxR homologous gene in V. vulnificus, was used instead of the luxR[Vs] gene, restoration of PHA level was also achieved, albeit with less PHA level than the luxR[Vs] complementation (Fig. [100]5c and d). Collectively, these findings demonstrate that LuxR[Vs] plays a positive role in enhancing PHA biosynthesis by activating genes involved in butanoate metabolism and PHA synthesis, and that LuxR[Vs] and SmcR are functional homologs of each other. LuxR[Vs]-mediated down-regulation of biofilm formation Because diguanylate cyclase activity and biofilm formation were significantly enriched in the GO term and KEGG pathway analyses, respectively (Figs. [101]3a and [102]4a), we also inspected the DEGs related to c-di-GMP synthesis and biofilm formation. In V. cholerae, a well-known Vibrio species recognized for its biofilm formation, a polar flagellum and mannose-sensitive hemagglutinin (MSHA) pili aid cells in scanning and attaching to surfaces, and the VPS (Vibriopolysaccharide) extracellular matrix is utilized in biofilm formation. During this process, c-di-GMP, synthesized by diguanylate cyclase, plays a critical role by binding to the key transcriptional regulator, VpsT, which in turn activates the vpsoperons and other biofilm-related genes^[103]16. Consistent with these previous reports, we found that the MSHA pilus gene mshC (Vs096_00555) and the diguanylate cyclase gene cdgA (Vs096_03429) in V. scophthalmi were significantly up-regulated in the MT-HCD (2.02-fold change for both genes) and WT-LCD (1.76- and 2.03-fold changes, respectively) compared to the WT-HCD (Fig. [104]6a and b). Notably, the expression of genes involved in the VPS production (e.g., glycosyltransferase vpsL, Vs096_03444; VPS biosynthesis/export protein vpsN, Vs096_03443; VPS-related tyrosine kinase vpsO, Vs096_3442; and VPS biosynthesis protein vpsQ, Vs096_03539) and the transcription activation for vps operons (i.e., transcriptional regulator vpsT, Vs096_03818) were significantly enhanced in the MT-HCD compared to the WT-HCD (Fig. [105]6a). Indeed, the fold change values of these genes in the MT-HCD were 2.93, 4.59, 3.86, 1.93, and 3.05, when compared to the WT-HCD (Fig. [106]6b). These findings suggest that LuxR[Vs] suppresses the VPS production and biofilm formation at HCD. Fig. 6. [107]Fig. 6 [108]Open in a new tab LuxR[Vs]negatively regulates biofilm formation. (a) Schematic representation of the biofilm formation pathway in vibrios. Proteins enriched in the KEGG pathway analysis are indicated with green boxes, in which the DEGs from the MT-HCD vs. WT-HCD and the WT-LCD vs. WT-HCD comparisons are marked with yellow and pink dots, respectively. Among the overlapping DEGs from both comparisons, up- or down-regulated DEGs from the MT-HCD vs. WT-HCD comparison are indicated with red or blue frames, respectively. (Copyright Kanehisa laboratories; reproduced with permission^[109]55.) (b) A heat map illustrating the log[2] fold changes in the expression of key biofilm formation pathway genes is presented for the MT-HCD vs. WT-HCD and the WT-LCD vs. WT-HCD comparisons. An asterisk (*) represents a statistically significant change in gene expression (|fold change| ≥ 1.75 and p ≤ 0.05). Since the luxR[Vs] gene is deleted in the mutant, the corresponding part is left as black. (c) Biofilm formation of the wild-type, ΔluxR[Vs], and complemented strains in borosilicate tube was assessed by crystal violet staining. The stained tubes were photographed before the dye elution with ethanol (upper image). In the lower bar graph, normalized biofilm formation levels are presented as means ± SD from three independent experiments. Statistical significance was assessed by one-way ANOVA with Tukey’s multiple comparisons (***, p < 0.005; **, p < 0.01). To validate this finding, we compared the biofilm formation activity of the wild-type, ΔluxR[Vs], and complemented strains. As shown in Fig. [110]6c, the biofilm biomass of the ΔluxR[Vs] strain was 1.45-fold higher than that of the wild-type strain. When this mutant strain was complemented with the luxR[Vs] gene, biofilm formation was reduced to an even slightly lower level than that of wild-type strain. As similar to the PHA production (Fig. [111]5c and d), the complementation with smcR gene also reverted the phenotype but to a lesser extent than that with the luxR[Vs] gene (Fig. [112]6c). Altogether, the results demonstrate that LuxR[Vs] negatively regulates the biofilm formation in V. scopthalmi by repressing the expression of MSHA pili MshC, diguanylate cyclase CdgA, and series of VPS biosynthesis genes. Discussion Because the QS system plays a role in regulating various bacterial physiologies, including virulence factor expression^[113]14,[114]15, we employed transcriptome analysis in this study to elucidate the global regulatory network controlled by LuxR[Vs], the QS master regulator in V. scophthalmi. In the comparison of ΔluxR[Vs] and wild-type strains (i.e. MT-HCD vs. WT-HCD), we found that LuxR[Vs] significantly influences approximately 13.7% of the V. scophthalmi genome (631 genes out of a total of 4,617 genes; Fig. [115]2a). While this is considerably less than the number of genes regulated by growth status (as revealed in the comparison of WT-LCD and WT-HCD, 45.9%; 2,121 genes out of a total of 4,617 genes; Fig. [116]2a), this result confirms that LuxR[Vs] acts as a global transcriptional regulator in this fish pathogen, similar to other clinically relevant Vibriospecies^[117]28,[118]29. Among the total of 631 LuxR[Vs] regulon, more than 52.6% (332 genes) were also identified as DEGs controlled by the growth status (i.e. from the WT-LCD vs. WT-HCD comparison; Fig. [119]2b). Additional analysis further categorized this subset into four different groups based on their co-regulation and contrasting regulation patterns in two comparisons. First, 94 genes were up-regulated in both MT-HCD and WT-LCD experimental samples than the WT-HCD control sample, indicating that they are repressed at HCD with the help of LuxR[Vs] (Fig. [120]2b). In contrast, the second group, comprising 115 co-downregulated genes, displays reduced expression in both experimental samples compared to the WT-HCD control, suggesting that they are highly expressed at HCD with the help of LuxR[Vs]-mediated activation (Fig. [121]2b). Considering that VibrioLuxR homologues are highly translated only at HCD^[122]17,[123]18, these co-regulations may represent a V. scophthalmi’s effective strategy to reinforce the cell density-dependent regulation of genes by incorporating the LuxR[Vs]-mediated regulation. For instance, the aphAgene (Vs096_02787), which encodes the LCD-specific QS master regulator^[124]18,[125]21, is repressed at HCD and by LuxR[Vs]. In contrast, the luxS gene (Vs096_02054), encoding the AI-2 synthase, is activated at HCD and by LuxR[Vs] (Supplementary Dataset S1 and Supplementary Fig. [126]S3). These reinforced regulations of QS components prevent V. scophthalmi from dysregulating the QS regulon, which could otherwise be affected by unintended AphA-mediated regulation and/or insufficient AI-2 production even at HCD. Third and fourth groups, containing 85 and 38 genes respectively, exhibited contrasting expression patterns in the two comparisons (i.e. MT-HCD vs. WT-HCD and WT-LCD vs. WT-HCD; Fig. [127]2b). In general, genes in microbes are regulated in response to various environmental cues, including changes in cell density, and this allows them to cope effectively with various challenges, ensuring their survival. Therefore, the observed contrasting regulations in third and fourth groups may represent another intelligent strategy employed by V. scophthalmi to precisely regulate important subsets of LuxR[Vs] regulon, not simply through growth phase-dependent metabolic and/or environmental changes, but more specifically via QS-mediated bacterial communication. Our study also revealed that V. scophthalmi activates genes in the biosynthetic pathway for PHA production through the LuxR[Vs] (Figs. [128]2c and [129]5). PHA is a bacterial intracellular semi-crystalline polymer that bacteria synthesize and store as granules when they encounter nutrient limitations or stress^[130]37. Therefore, our finding suggests that V. scophthalmi would experience nutrient starvation when crowded in a certain local environment, such as biofilms, and thus had evolved the QS-dependent PHA production mechanism. Alternatively, PHA biosynthesis at HCD might serve as a survival strategy for V. scophthalmias it sheds from the fish host into the sea during its pathogenesis cycle. Indeed, this fish pathogen would proliferate high in the diseased host’s intestine but is ultimately released into the nutrient-limited marine environment to initiate the next infection cycle^[131]34–[132]36. In this context, the pathogen likely evolved to synthesize PHA granules within the crowded niche (i.e. the host intestine) to prepare and ensure survival upon encountering the open marine environment. In fact, several other bacterial species, including Pseudomonas chlororaphis, Burkholderia thailandensis, and V. harveryi, have also been reported to regulate PHA biosynthesis via QS, although their specific QS systems and components differ from each other and those in V. scophtalmi^[133]37,[134]38. While details about V. scophtalmi pathogenesis remain to be elucidated, our findings, along with previous reports, contribute to a more comprehensive understanding of the bacterial survival strategies in marine environments by suggesting the role of QS in nutrient reservation. In the meantime, PHA is also considered as a sustainable alternative to petroleum-based plastics because of its biodegradability. In fact, a few Vibriospecies have been explored as potential candidates for PHA production, albeit the studies simply overexpressed the PHA biosynthesis genes or operon^[135]39,[136]40. Since the PHA biosynthetic pathway is intrinsically linked to the butanoate metabolism pathway (Fig. [137]5a), metabolic engineering targeting the entire butanoate metabolism pathway could warrant higher production of PHA or PHA derivatives. In this aspect, our finding that the QS system activates many parts of the butanoate pathway via LuxR[Vs]-mediated regulation in V. scophthalmi (Fig. [138]5a) provides an important foundation for future engineering of this or related Vibrio species as a next-generation host for PHA fermentation. Another key finding in this study is that the V. scophthalmi QS system negatively regulates biofilm formation via LuxR[Vs]-mediated regulation of various biofilm-related genes (Fig. [139]6). In fact, a previous study reported exactly contradictory results, showing that deletion of the luxR or luxS gene in the V. scophthalmiA102 strain resulted in reduced biofilm formation^[140]26. The strain and medium (marine broth) used in the previous study differ from those in our study, which might explain the opposing results. Nonetheless, our transcriptome analysis, qRT-PCR validation, and phenotypic comparison, including the complemented strains (Figs. [141]2c and [142]6), clearly demonstrate that V. scophthalmi represses biofilm formation through the LuxR[Vs]-mediated QS. Since c-di-GMP molecules governing biofilm formation are synthesized from two GTP molecules, the significant up-regulation of two genes in purine metabolism pathway [i.e., IMP dehydrogenase (Vs096_01797, guaB) and GMP synthase (Vs096_01798, guaA)] as well as the diguanylate cyclase gene (Vs096_03429, cdgA) in the ΔluxR[Vs] strain compared to the wild-type strain (Figs. [143]2c and [144]6b, Supplementary Fig. [145]S3a, and Supplementary Dataset S1) further supports our finding. Consistent with these, some other Vibrio species, such as V. vulnificus and V. parahaemolyticus, have also been reported to down-regulate biofilm formation through LuxR homologues^[146]29,[147]41. Notably, genes related to osmoprotection were also found to be negatively regulated by LuxR[Vs] in V. scophthalmi (Fig. [148]3a). Indeed, biosynthesis genes for ectoine (Vs096_04335–04336, ectAB), one of the important compatible solutes in bacteria^[149]42, were significantly up-regulated in the MT-HCD compared to the WT-HCD (Supplementary Dataset S1). Similarly, genes encoding betaine-carnitine-choline transporter 3 and ABC-family ProU transporter (Vs096_03282, bccT3; Vs096_04334 − 04332, proVWX), which are responsible for the uptake of compatible solutes^[150]42, exhibited extreme up-regulation in the MT-HCD compared to the WT-HCD (Supplementary Fig. [151]S3a and Supplementary Dataset S1). As suggested previously, this LuxR[Vs]-mediated negative regulation of osmoprotection genes may represent a reduced necessity for compatible solutes in V. scophthalmiwhen this pathogen enters the HCD, a state for resting, not active adapting and replicating^[152]43. Importantly, we were unable to find any known toxins or type 3/4 secretion system apparatus in the V. scophthalmi genome (Fig. [153]1). Nonetheless, any enzymes that contribute to in vivo fitness, promoting the survival and proliferation of this opportunistic pathogen, could be potential virulence determinants. This is because fish possess various defense mechanisms against pathogen infection, including the production of reactive oxygen species (ROS) by macrophages. Consequently, invading pathogens must efficiently neutralize ROS to survive and proliferate in the face of host immune responses^[154]44. Consistently, a previous study has demonstrated that the activity of protective enzymes against host-derived ROS and free radicals is greater in hyper-virulent V. scophthalmistrains compared to less virulent strains^[155]10. Notably, genes encoding such enzymes, namely superoxide dismutase (Vs096_03760, sod1), alkyl hydroperoxide reductase subunit C (Vs096_02795, ahpC), and catalase (Vs096_03799, katE), were significantly down-regulated in the MT-HCD compared to the WT-HCD (Supplementary Fig. [156]S3a and Supplementary Dataset S1), indicating LuxR[Vs]-mediated activation. Since V. scophthalmi would proliferate to a much higher concentration within the host than in the open marine environment, the LuxR[Vs]-mediated activation of such protective enzymes is not a simple cell-density dependent regulation but rather a form of spatiotemporal regulation. Therefore, it is clear that V. scophthalmi employs the QS system to precisely regulate these virulence traits, which are crucial for its survival and pathogenesis in the fish host. In conclusion, this study provides critical insights into the global regulatory roles of the QS master regulator LuxR[Vs] in the fish pathogen V. scophthalmi. Our genomic and transcriptomic analyses reveal that LuxR[Vs] mediates the regulation of diverse physiological processes, including nutrient reservation (PHA biosynthesis), biofilm formation, osmoprotection, and virulence (Fig. [157]7). These functions are of particular relevance to the pathogen’s ability to survive and proliferate in aquatic environments, where they significantly impact seafood production, as demonstrated in olive flounder cultivation. Importantly, the identification of LuxR[Vs] as a central global regulator highlights its potential as a target for mitigating the virulence of V. scophthalmi without compromising its viability (Supplementary Fig. [158]S1), thus avoiding the risk of resistance development. The use of QS inhibitors, such as the previously characterized QStatin^[159]29, presents a promising strategy for disrupting LuxR[Vs]-mediated QS pathways, offering an innovative approach to managing bacterial infections in aquaculture. Given the growing concerns over microbial infections and antibiotic resistance in aquaculture worldwide^[160]45, our findings offer a foundation for the development of more sustainable and effective measures to enhance seafood production. Fig. 7. [161]Fig. 7 [162]Open in a new tab LuxR[Vs]-mediated regulation of group behaviors in V. scophtahlmi. As cell density increases, LuxR[Vs] directly or indirectly regulates the expression of multiple downstream genes involved in various group behaviors. Nutrient reservation and virulence traits-related genes are positively regulated while biofilm formation and osmoprotection-related genes are negatively regulated by LuxR[Vs]. Materials and methods Bacterial strains and culture conditions All bacterial strains and plasmids used in this study are listed in Table [163]1. Unless otherwise specified, Escherichia coli and V. scophthalmi strains were grown in Luria-Bertani medium (LB; BD Difco, Sparks, MD, USA) at 37 °C and in LB supplemented with 2.0% (w/v) NaCl (LBS) at 27 °C, respectively. Autoinducer bioassay (AB) medium (300 mM NaCl, 50 mM MgSO[4], 0.2% vitamin-free casamino acids, 10 mM potassium phosphate, 1 mM L-arginine, pH 7.5)^[164]46supplemented with 66.7 mM glycerol^[165]47 was used for biofilm formation. When necessary, antibiotics were used at the following concentrations: for E. coli, 20 µg/ml chloramphenicol, 100 µg/ml ampicillin, and 100 µg/ml kanamycin; for V. scophthalmi, 3 µg/ml chloramphenicol, 100 µg/ml ampicillin, and 100 µg/ml kanamycin. Table 1. Bacterial strains and plasmids used in this study. Bacterial strain or plasmid Relevant characteristics ^a Reference or source Strains V. scophthalmi 21FBVib0096 Wild-type; isolated from diseased fish This study; NIFS ^b 21FBVib0096 ΔluxR[Vs] 21FBVib0096 with ΔluxR[Vs] This study E. coli DH5α supE44 ΔlacU169 (Φ80 lacZ ΔM15) hsdR17 recA1 endA1 gyrA96 thi-1 relAI Laboratory collection S17-1 λpir λpir lysogen; thi pro hsdR hsdM^+recA Tc::Mu-Km::Tn7;Tp^r Sm^r; host for π-requiring plasmids; conjugal donor ^[166]68 Plasmids pDS132 R6K γori sacB; suicide vector; oriT of RP4; Cm^r ^[167]69 pBK2301 pDS132 with ΔluxR[Vs]; Cm^r This study pJK1113 pKS1101 with nptI; Ap^r Km^r ^[168]59 pBK2302 pJK1113 with luxR[Vs]; Ap^r Km^r This study pBSS-WT pBAD24 with smcR; Ap^r ^[169]27 pBK2303 pJK1113 with smcR; Ap^r Km^r This study [170]Open in a new tab ^a Tp^r, trimethoprim-resistant; Sm^r, streptomycin-resistant; Ap^r, ampicillin-resistant;. Cm^r, chloramphenicol-resistant; Km^r, kanamycin-resistant; Tc^r, tetracycline-resistant. ^b NIFS, National Institute of Fisheries Science, South Korea. Whole genome sequencing, assembly, and annotation Genomic DNA of the V. scophthalmi 21FBVib0096 strain was extracted using the G-spin^™ Genomic DNA Extraction Kit (iNtRON Biotechnology, South Korea) and then sequenced using the MiSeq (Illumina Inc., San Diego, CA, USA) and the GridION (Oxford Nanopore Technologies, Oxford, UK) platforms at Sanigen Co. Ltd. (South Korea). Raw reads were subjected to de novoassembly using Unicycler (v0.4.8)^[171]48with circularization. Annotation procedure was conducted at CJ Bioscience, Inc. (South Korea). Briefly, prediction of CDSs was carried out using PRODIGAL (v2.6.2)^[172]49. Transfer RNAs were identified using tRNAscan-SE (v1.3.1)^[173]50. Ribosomal RNAs and non-coding RNAs were predicted using covariance model search with Rfam 12.0 database^[174]51. Clustered regularly interspaced short palindromic repeats (CRISPRs) were identified using Piler-CR (v.1.06) program^[175]52and the CRISPR Recognition Tool (CRT) (v.1.2)^[176]53. Functional annotation was performed through homology search against databases, including EggNOG (v 4.1), KEGG, and SEED subsystems, utilizing the USEARCH program (v. 8.01517)^[177]54–[178]57. The complete genome sequence of V. scophthalmi 21FBVib0096 strain has been deposited in the GenBank under accession numbers, [179]CP134277 (chromosome 1), [180]CP134278 (chromosome 2), and [181]CP134279 (plasmid). Construction of luxR[Vs ]deletion mutant and complemented strains Oligonucleotides used in this study are listed in the Supplementary Table [182]S3. To construct an isogenic luxR[Vs] mutant, about 80% of the internal region of luxR[Vs] open reading frame was deleted in vitro employing the PCR-mediated linker-scanning method. In brief, pairs of primers LuxR[Vs]-F1 and -R1 or LuxR[Vs]-F2 and -R2 were used for amplification of the 5’ and 3’ amplicons, respectively (Supplementary Table [183]S3). The 481-bp-deleted luxR[Vs] gene was then amplified by PCR using a mixture of both amplicons as templates and primers LuxR[Vs]-F1 and LuxR[Vs]-R2. The resulting DNA fragment was treated with SphI and SacI and then ligated into the same enzyme-treated pDS132 to generate the pBK2301 (Table [184]1). Subsequently, the plasmid was transformed into E. coli S17-1 λpir and then transferred into V. scophthalmi21FBVib0096 through conjugation. Exconjugants were isolated on selective plates (TCBS agar plates supplemented with chloramphenicol) and subjected to the sucrose challenge to facilitate the second homologous recombination, as previously described^[185]58. The desired mutation was verified by PCR and sequencing of the engineered luxR[Vs] gene region using primers LuxR[Vs]-F and LuxR[Vs]-R (Supplementary Table [186]S3). For complementation of the mutation, coding regions of the V. scophthalmi luxR[Vs] and V. vulnificus smcR were amplified from the V. scophthalmi21FBVib0096 genome and pBSS-WT vector^[187]27, respectively, using specified primer sets (Supplementary Table [188]S3). Each amplified product was cloned into NcoI-XbaI-digested pJK1113, an arabinose-inducible expression vector^[189]59, resulting in the creation of pBK2302 and pBK2303, respectively (Table [190]1). The plasmids were then transferred into the V. scophthalmi ΔluxR[Vs] by conjugation as described above. RNA isolation and transcriptome analysis The wild-type and ΔluxR[Vs] strains were cultured in LBS at 27 °C for 9 h to ensure a HCD of the cultures, reaching stationary phase (Supplementary Fig. [191]S1). Accordingly, the harvested samples from wild-type and mutant strains were designated as WT-HCD and MT-HCD, respectively. Total RNAs were isolated from the harvested samples using RNeasy Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer’s instructions. The RNA samples were then treated with DNase I to eliminate any genomic DNA and subsequently purified using RNeasy Minelute Cleanup Kit (QIAGEN). To prepare an LCD sample, wild-type cells grown to early exponential phase (Supplementary Fig. [192]S1) were harvested, and their total RNA was purified. This sample was designated as WT-LCD. RNA-sequencing was conducted as previously described^[193]29 with slight modification at CJ Bioscience, Inc. Total RNA quantity and quality were assessed using ND-2000 spectrophotometer (Thermo Fisher Scientific, Madison, USA) and Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), respectively. Before the construction of strand-specific complementary DNA (cDNA) library, mRNA was firstly enriched by depleting ribosomal RNAs using NEBNext rRNA Depletion Kit (Bacteria) (New England BioLabs, Hitchin, UK). Libraries were prepared using the TruSeq^® Stranded mRNA library prep kit (Illumina). Subsequently, sequencing was performed on the Illumina NovaSeq6000 platform using 150 bp paired-end sequencing. The raw sequencing reads were deposited in the NCBI Short Read Archive (SRA) under accession numbers SRR28019963 to SRR28019971. After preprocessing steps, including sequencing quality check and adapter trimming, the sequencing reads were mapped to the V. scophthalmi21FBVib0096 genome. By using the EdgeR package^[194]60, raw read counts for each gene were normalized using the TMM algorithm to account for any systematic biases across the samples^[195]61. To elucidate expression patterns among sample groups, a multidimensional scaling (MDS) plot was created using the plotMDS function of EdgeR package^[196]60. Following this, hierarchical clustering was executed using the heatmap package with a squared Euclidean distance metric as the second analysis. Pairwise comparison analysis was then statistically performed to identify DEGs (|fold change| ≥ 1.75 and p ≤ 0.05) in the MT-HCD or WT-LCD sample relative to the WT-HCD sample. For functional analysis of the DEGs, the transcriptome was aligned to the nonredundant (nr) database of the NCBI via blastx search with an E-value of < 0.001 and then assigned the GO terms and KEGG pathways using the OmicsBox software (BioBam Bioinformatics, Valencia, Spain; [197]https://www.biobam.com/omicsbox/)^[198]62. In functional enrichment analysis, each set of DEGs was selected as the test set while all other annotated transcripts were considered the reference set. Significantly over-enriched GO terms and KEGG pathways were identified from test set by using one-tailed Fisher’s exact test with p-value of < 0.05. qRT-PCR analysis To validate the RNA-sequencing data, five LuxR[Vs]-regulon genes were selected and analyzed by qRT-PCR. Total RNAs were extracted from WT-HCD, WT-LCD, and MT-HCD cells as described above. cDNA was synthesized from 1 µg of the total RNA using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) in a final volume of 20 µl. qRT-PCR was performed with pairs of specific primers (Supplementary Table [199]S3) and SYBR Green (Bio-Rad) by using the QuantStudio 1 Real-Time PCR System (Thermo Fisher Scientific). The recA housekeeping gene was utilized as an internal control gene for normalization, and the relative expression values were calculated using the comparative threshold cycle method (2^-ΔΔCt)^[200]63. The obtained results were presented as means ± standard deviation (SD) from triplicate. Quantitative comparison of PHA production To assess PHA biosynthesis, the Nile red agar plate assay was employed as described previously^[201]64. In the complex medium with 2% agar, which consisted of 10 g of NaCl, 3.7 g of Na[2]HPO[4], 1.0 g of KH[2]PO[4], 0.5 g of (NH[4])[2] HPO[4], 0.2 g of MgSO[4]·7H[2]O, 2 ml glycerol, 5 g of bactotryptone, 0.5 g of yeast extract per liter^[202]65 and appropriate antibiotics and arabinose [0.01% (w/v); for the induction of complemented genes], a solution of Nile red in dimethyl sulfoxide was added to achieve a final concentration of 0.5 µg/ml. Wild-type, ΔluxR[Vs], or the complemented strains were spotted on the plates, and then incubated for 72 h at 27 °C. Subsequently, the intensity of fluorescence from the colonies was observed under UV light. For the quantitative comparison of PHA biosynthesis, the Nile red fluorescence method was used^[203]66. In brief, the cultures were centrifuged at 12,000 × g for 5 min, and the supernatants were discarded. Remaining bacterial cell pellets were then resuspended in 1 ml of distilled water and a solution of Nile red was added to achieve a final concentration of 3.1 µg Nile red per ml of suspension. After 30 min incubation at dark, second centrifugation was performed and the pellets were resuspended again in 1 ml of distilled water. Finally, the samples were transferred to a 96-well microtiter plate and the fluorescence intensity was measured using the CLARIOstar^® Microplate Reader (BMG Labtech, Ortenberg, German) at an excitation wavelength of 535 nm and an emission wavelength of 605 nm. PHA production was quantified by normalizing the fluorescence intensity by optical density at 600 nm (OD[600]). Biofilm formation assay Wild-type, ΔluxR[Vs], and the complemented strains were freshly inoculated into AB-glycerol medium with appropriate antibiotics and arabinose in borosilicate tubes. After static incubation at 27 °C for 72 h, planktonic cells were removed, and their cell density was determined via spectrometric measurement at OD[595]. Subsequently, the remaining biofilms in the tube were carefully rinsed with phosphate-buffered saline (PBS, pH 7.4) and then stained with 1% (w/v) crystal violet solution for 30 min. After brief washing with distilled water, the biofilm-associated crystal violet was eluted using 100% ethanol and quantified by measuring OD[550]. The degree of biofilm formation was normalized by dividing OD[550] by the respective planktonic cell density (OD[595])^[204]67. Statistical analysis Unless noted otherwise, the data presented as means with SD from at least three experiments. Statistical analysis was performed using GraphPad Prism ver. 8.4.3 (GraphPad Software, San Diego, CA, USA; [205]https://www.graphpad.com/), as described in each figure legend. Electronic supplementary material Below is the link to the electronic supplementary material. [206]Supplementary Material 1^ (430.5KB, pdf) [207]Supplementary Material 2^ (90KB, xlsx) [208]Supplementary Material 3^ (310.7KB, xlsx) [209]Supplementary Material 4^ (40.8KB, xlsx) Acknowledgements