The wild type (WT) strain RIMD2210633 of V. parahaemolyticus was utilized in this study (Makino et al., 2003). Nonpolar acsS deletion mutant (ΔacsS), derived from the WT strain, was constructed by our previous study (Chang et al., 2024). The complementation strain ΔacsS/pBAD33-acsS (C-ΔacsS) was constructed by introducing pBAD33-acsS into ΔacsS (Chang et al., 2024). Control strains (WT/pBAD33 and ΔacsS/pBAD33) were generated by introducing pBAD33 into WT and ΔacsS, respectively. The acsS and tpdA double-gene mutant (ΔacsSΔtpdA) and the tpdA single-gene mutant (ΔtpdA) were generated via deletion of a 258-bp fragment (nucleotides 1305–1562) of tpdA from ΔacsS and WT, respectively, by homologous recombination using suicide plasmid pDS132 (Sun et al., 2012). All primers used in this study are listed in Table 1.

Unless stated otherwise, the cultivation of V. parahaemolyticus was conducted as previously described (Lu et al., 2021). Briefly, V. parahaemolyticus was grown in 2.5% (w/v) Bacto heart infusion (HI) broth (BD Biosciences, New Jersey, United States) at 37 °C with shaking at 200 rpm for 12 h. The resultant culture was diluted 50-fold into 5 mL HI broth, and then incubated under the same conditions until it reached to an optical density at 600 nm (OD₆₀₀) value of 1.4. This culture was referred to as the bacterial seed. Subsequently, the bacterial seed was diluted 1,000-fold into 5 mL of HI broth for a third round of growth and was harvested at an OD₆₀₀ value of 1.4. When necessary, the medium was supplemented with 50 μg/mL of gentamicin, 5 μg/mL of chloramphenicol and/or 0.1% (w/v) L-arabinose.

Crystal violet (CV) staining assay was performed similarly as previously described (Zhang et al., 2023). Briefly, the bacterial seed was diluted 50-fold into 2 mL of Difco marine (M) broth 2216 (BD Biosciences, New Jersey, United States), supplemented 5 μg/mL chloramphenicol and 0.1% L-arabinose, in a 24-well cell culture plate, and then incubated at 30 °C with shaking at 150 rpm for 24 h. Planktonic cells were collected for measurement of OD₆₀₀ values. Surface attached biofilm cells were washed with deionized water, and then stained by 0.1% CV. Bound CV was dissolved in 2.5 mL of 20% acetic acid, and then the OD₅₇₀ values were measured. The capacity for biofilm formation was expressed as the ratio of OD₅₇₀ to OD₆₀₀.

For the colony morphology assay (Zhang et al., 2023), the overnight bacterial culture was diluted 50-fold into 5 mL M broth, and it was then statically incubated at 30 °C for 48 h. After thorough mixing, 2 μL of the culture was spotted onto an HI plate, or an HI plate supplemented with 5 μg/mL chloramphenicol and 0.1% (w/v) L-arabinose, and incubated at 37 °C for 48 h.

The WT and ΔacsS strains were incubated under the same conditions as the CV staining assay, but without the addition of chloramphenicol and L-arabinose. Three technical replicates were conducted for each strain. Bacterial cells were harvested simultaneously from biofilms and planktonic fractions for the preparation of total RNA, which was extracted using TRIzol Reagent (Invitrogen, Massachusetts, United States) (Zhang et al., 2023). One RNA sample was prepared from each technical replicate. RNA concentration and integrity were determined by a Nanodrop 2000 and the agarose gel electrophoresis, respectively. rRNA removal and mRNA enrichment were performed using an Illumina/Ribo-Zero™ rRNA Removal Kit (bacteria) (Illumina, California, United States). All RNA-related manipulations including RNA extraction were performed in GENEWIZ Biotechnology Co. Ltd. (Suzhou, China). cDNA sequencing was performed on an Illumina Hiseq platform (Zhang et al., 2023; Zhang et al., 2022). Gene expression in ΔacsS (test group) was compared with that in WT (reference group). DESeq (v1.12.4) was used to identify the differentially expressed genes (DEGs), filtering for p ≤ 0.01 and absolute FoldChange ≥ 2. DEGs were further analyzed using the Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, and Cluster of Orthologous Groups of proteins (COG) database (Zhang et al., 2023; Zhang et al., 2022).

Intracellular c-di-GMP was quantified as previously described (Zhang et al., 2023). Briefly, bacterial cells were harvested at an OD₆₀₀ value of 1.4, and then they were resuspended in 2 mL ice-cold phosphate buffered saline (PBS). The bacterial suspension was incubated at 100 °C for 5 min, sonicated for 15 min, and then centrifuged at 9,000 g for 5 min. Total protein and c-di-GMP levels in the supernatant were determined using a Pierce BCA Protein Assay kit (ThermoFisher Scientific, Massachusetts, United States) and a c-di-GMP Enzyme-linked Immunosorbent Assay (ELISA) Kit (Mskbio, Hubei, China), respectively. The c-di-GMP level was expressed as pmol/g of protein.

Bacterial cells were harvested at an OD₆₀₀ value of 1.4. Total RNA was extracted using TRIzol Reagent (Invitrogen, Massachusetts, United States). cDNA was generated from 1 μg of total RNA using a FastKing First Strand cDNA Synthesis Kit (Tiangen Biotech, Beijing, China). Real-time quantitative PCR (RT-qPCR) was performed using a LightCycler 480 (Roche, Basel, Switzerland) together with SYBR Green master mix (Tiangen Biotech, Beijing, China) (Gao et al., 2011). The relative expression levels of each target gene were determined using the 2^−ΔΔCt method, with the 16S rRNA serving as the internal control.

The regulatory DNA region of each target gene was cloned into pHRP309 harboring a promoterless lacZ gene and a gentamicin resistance gene (Parales and Harwood, 1993). Each recombinant plasmid was transferred into WT and its corresponding mutants. Transformants were cultured and lysed to measure the β-galactosidase activity of the cellular extracts using a β-Galactosidase Enzyme Assay System (Promega, Wisconsin, USA). Miller Units representing the β-galactosidase activity was calculated as previously described (Zhang et al., 2023). For the two-plasmid lacZ reporter assay (Zhang et al., 2023), the recombinant pHRP309 was transferred into Escherichia coli 100 λpir (EC100; Epicenter, Wisconsin, USA) bearing pBAD33-acsS or pBAD33. The transformants were cultured in Luria-Bertani (LB) broth containing 0.1% L-arabinose and 20 μg/mL chloramphenicol at 37 °C with shaking at 200 rpm. Bacterial cells were harvested at an OD₆₀₀ value of 1.2, and then lysed to measure the β-galactosidase activity in the cell extracts.

The coding region of acsS was cloned into pET28a (Novagen, Darmstadt, Germany). The recombinant pET28a plasmid was transferred into E. coli BL21λDE3 to express the His-tagged AcsS protein (His-AcsS). Expression and purification of His-AcsS were performed as previously described for His-OpaR (Sun et al., 2012). The purity of His-AcsS was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The concentration of His-AcsS solution was determined using a Pierce BCA Protein Assay kit. Purified His-AcsS was stored at −60 °C.

For electrophoretic mobility-shift assay (EMSA) (Zhang et al., 2023), the regulatory DNA region of each target gene was amplified by PCR. EMSA was performed in a 10 μL reaction volume containing 0.5 mM EDTA, 1 mM MgCl₂, 50 mM NaCl, 0.5 mM DTT, 10 mM Tris–HCl (pH 7.5), 0.625 μg/mL salmon sperm DNA, 100 ng target DNA, and a certain amount of His-AcsS. After incubation at room temperature for 20 min, the binding reactions were visualized on a native polyacrylamide gel, which was stained with ethidium bromide and analyzed using a UV transilluminator.

The CV staining, colony morphology assay, c-di-GMP measurement, LacZ fusion assay, RT-qPCR, and two-plasmid lacZ fusion assay were performed at least three times, with at least three technical replicates each time. EMSA for each target gene was performed at least two times, independently. The numerical results were expressed as the mean ± standard deviation (SD). To calculate statistical significance, Student’s t-tests or two-way ANOVA with Tukey’s post hoc corrections were applied, considering a p value of less than 0.05 as significant.

AcsS is involved in promoting the swimming and swarming motility of V. parahaemolyticus, which are required for mature biofilm formation (Chang et al., 2024; Enos-Berlage et al., 2005). We therefore investigated the potential regulatory role of AcsS in biofilm formation by V. parahaemolyticus. As depicted in Figure 1a, the ΔacsS/pBAD33 strain displayed remarkably reduced CV staining compared to both the WT/pBAD33 and C-ΔacsS strains (p < 0.01). Notably, the C-ΔacsS strain demonstrated a restored CV staining pattern indicative of biofilm formation. Additionally, the colony of ΔacsS appeared smoother than that of WT (Figure 1b). Moreover, the colonies of both WT/pBAD33 and C-ΔacsS appeared smoother than that of WT. This observation may be attributed to the significant influence of chloramphenicol and L-arabinose on this particular phenotype (Zhang et al., 2023; Zhang et al., 2023). While the colony morphology of C-ΔacsS and WT/pBAD33 appeared quite different, both exhibited a more wrinkled appearance compared to ΔacsS/pBAD33. Since mutation of acsS does not affect the growth of V. parahaemolyticus (Chang et al., 2024), these results indicate that AcsS positively regulates biofilm formation in V. parahaemolyticus.

To determine the regulatory mechanism of AcsS on biofilm formation in V. parahaemolyticus, RNA sequencing (RNA-seq) analysis was performed comparing the ΔacsS (test) and WT (reference) strains. As shown in Figure 2a and detailed in Supplementary Table S1, 235 genes were identified as regulated by AcsS under biofilm growth conditions. Of these, 78 genes were upregulated and 157 genes were downregulated, in ΔacsS compared to WT. GO term enrichment analysis indicated that DEGs were associated with molecular functions (7 GO terms, 29 DEGs), cellular components (5 GO terms, 42 DEGs) and biological processes (13 GO terms, 41 DEGs) (Figure 2b). KEGG pathway enrichment results revealed that 181 DEGs mapped to pathways including metabolism, human diseases, genetic information processing, environmental information processing, and cellular processes (Figure 2c). COG enrichment analysis categorized DEGs into 19 functional groups, with the most significant enrichment in function unknown and metabolism-related categories (Figure 2d). These findings suggested that AcsS regulates global gene expression in V. parahaemolyticus.

As listed in Table 2, several DEGs implicated in biofilm formation were identified, including tpdA, flgD, flgE, motY, mshG, capF, and VP0226. Specifically, tpdA encodes a trigger PDE involved in biofilm formation and c-di-GMP degradation (Martínez-Méndez et al., 2021). The genes flgD, flgE, and motY are associated with the polar flagellar system, while mshG belongs to the type IV pili gene cluster. Additionally, capF and VP0226 contribute to capsular polysaccharide (CPS) synthesis. However, AcsS is unlikely to promote biofilm formation solely through polar flagella, type IV pili, and CPS, given the extensive gene networks governing these structures in V. parahaemolyticus (Makino et al., 2003). The cpsA-K gene cluster, directly associated with the wrinkled colony phenotype and regulated by TpdA (Liu et al., 2022), further underscores this complexity. Consequently, tpdA and cpsA (VPA1403) were selected as focal genes for subsequent experiments.

The RT-qPCR results showed that the mRNA level of tpdA significantly increased in ΔacsS and ΔacsSΔtpdA but decreased in ΔtpdA compared to WT (Figure 3a). Specifically, tpdA mRNA levels were significantly lower in ΔacsSΔtpdA than in ΔacsS and significantly higher than in ΔtpdA (p < 0.05) (Figure 3a). Furthermore, the mRNA level of cpsA significantly decreased in ΔacsS and increased in ΔtpdA, whereas no significant change was observed in ΔacsSΔtpdA compared to WT (Figure 3a). Compared to ΔtpdA, the cpsA mRNA level was significantly elevated in ΔacsSΔtpdA but significantly reduced relative to ΔacsS (p < 0.05) (Figure 3a). As further determined by LacZ fusion assay (Figure 3b), the promoter activity of tpdA significantly increased in ΔacsS and ΔacsSΔtpdA but significantly decreased in ΔtpdA compared to WT (p < 0.01). Additionally, the promoter activity of tpdA was significantly lower in ΔacsSΔtpdA than in ΔacsS and higher than in ΔtpdA (p < 0.01) (Figure 3b). For cpsA, promoter activity significantly decreased in ΔacsS and ΔacsSΔtpdA but increased in ΔtpdA compared to WT (p < 0.01). Notably, cpsA promoter activity in ΔacsSΔtpdA was elevated relative to ΔacsS but reduced compared to ΔtpdA (p < 0.01) (Figure 3b). Both assays demonstrated that tpdA expression in ΔacsSΔtpdA exceeds WT levels, suggesting that AcsS’s negative regulatory effect on tpdA outweighs TpdA’s positive regulation. Regarding cpsA expression, a discrepancy was observed between the RT-qPCR and LacZ fusion results: RT-qPCR revealed no significant difference in cpsA mRNA levels between ΔacsSΔtpdA and WT, whereas LacZ assays showed significantly lower cpsA promoter activity in ΔacsSΔtpdA than in WT. This inconsistency warrants consideration. Possible explanations include: (Baker-Austin et al., 2018) the lacZ fusion construct might lack regulatory elements present in the native chromosomal context that modulate mRNA stability or post-transcriptional processing, leading to a discrepancy between promoter activity measured by the reporter and steady-state mRNA levels; or (Sharan et al., 2022) post-transcriptional regulatory mechanism (e.g., affecting mRNA stability or translation efficiency) could differentially influence the endogenous cpsA mRNA measured by RT-qPCR versus the heterologous lacZ mRNA transcribed from the cpsA promoter fusion. Collectively, despite this discrepancy for cpsA in the double mutant, the results consistently indicate that AcsS suppresses tpdA expression but activates cpsA, independent of TpdA. Conversely, TpdA promotes its own expression while repressing cpsA, regardless of AcsS.

The results of EMSA showed that His-AcsS dose-dependently binds to the regulatory DNA fragment of cpsA but does not bind to the regulatory DNA region of tpdA or the coding region of 16S rRNA (used as a negative control) (Figure 4a). Additionally, a two-plasmid lacZ reporter assay demonstrated that expressing acsS from pBAD33-acsS in EC100 significantly decreased tpdA promoter activity while increasing cpsA promoter activity (Figure 4b). Although the two-plasmid lacZ fusion assay is a widely used method to validate direct regulatory interactions (Zhang et al., 2021; Ante et al., 2015; Bina et al., 2016), results obtained in heterologous hosts like EC100 must be interpreted cautiously. Potential confounding factors include regulator overexpression and incompatibility or unintended interactions between the regulator and the host’s cellular machinery. Collectively, these findings confirm that AcsS directly activates the transcription of cpsA and indirectly represses tpdA transcription.

A previous study demonstrated that deletion of tpdA led to a 33% increase in c-di-GMP levels compared to WT during exponential growth (Martínez-Méndez et al., 2021). The data from this study also showed that the c-di-GMP level in ΔtpdA was significantly higher than in WT (p < 0.05) (Figure 5). Additionally, the c-di-GMP levels in ΔacsS were significantly reduced compared to WT, ΔtpdA and ΔacsSΔtpdA (p < 0.05) (Figure 3). Furthermore, the c-di-GMP level in ΔacsSΔtpdA was significantly lower than that in ΔtpdA (p < 0.05) (Figure 3). However, no significant differences were detected when comparing ΔacsSΔtpdA with WT (p > 0.05) (Figure 5). These findings indicate that TpdA degrades c-di-GMP in V. parahaemolyticus independently of AcsS, while AcsS stimulates c-di-GMP synthesis regardless of TpdA’s presence.

To determine whether AcsS-dependent biofilm formation is mediated by TpdA, we compared the biofilm-forming abilities of the WT, ΔacsS, ΔtpdA and ΔacsSΔtpdA strains. As depicted in Figure 6a, the ΔacsS and ΔacsSΔtpdA strains displayed significantly reduced CV staining compared to the WT and ΔtpdA strains, respectively (p < 0.05). In contrast, the ΔacsSΔtpdA strain exhibited significantly enhanced CV staining relative to the ΔacsS strain (p < 0.01). However, no significant difference was observed between the ΔtpdA and WT strains (p > 0.05). Additionally, the colonies of WT and ΔtpdA were more wrinkled than those of the ΔacsS and ΔacsSΔtpdA strains (Figure 6b). The colonies of ΔtpdA and ΔacsSΔtpdA were slightly wrinkled compared to those of the WT and ΔacsS strain, respectively (Figure 6b). These results suggest that AcsS-dependent biofilm formation is independent of TpdA, while TpdA appears to partially suppress biofilm formation in the ΔacsS genetic background.

To determine whether TpdA regulates acsS, we analyzed acsS mRNA levels by RT-qPCR. As shown in Figure 7, the mRNA levels of acsS were significantly elevated in the ΔtpdA strain compared to the WT strain (p < 0.01), suggesting that the expression of acsS was under the negative control of TpdA in V. parahaemolyticus.