Non-O1/non-O139 V. cholerae HN375 from the China Center for Type Culture Collection (CCTCCAB209168) was used as the wild-type (WT) strain (Luo et al., 2011). Cloning was carried out using Escherichia coli DH5 and DH5-λpir, and conjugation were carried out using WM3064. Each strain was grown on Luria-Bertani (LB) medium at 37°C until the stationary phase was achieved unless otherwise indicated. The culture medium was modified by adding 0.1% arabinose and 100 g/mL ampicillin depending on the situation. Details of all plasmids and strains used are presented in Supplementary Table 1 .

From the WT HN375 strain, deletion mutants were constructed with pWM91, a suicide plasmid (Wu et al., 2015). A list of the primers used can be found in Supplementary Table 2 . To construct complementary mutants, the entire arcA encoding region by cloning into the pBAD24 plasmid vector, which was then transformed into ΔdegSΔarcA by electroporation to obtain ΔdegSΔarcA:arcA. A similar method was used to construct ΔdegS-overexpressed icdh strains (ΔdegS+icdh). As previously described, we used the pBAD24-arcA plasmid as a template and constructed a point mutation in D54E of arcA using site-directed mutagenesis (Fisher and Pei, 1997). The pBAD24-arcA ^D54E plasmid vector was then transformed into ΔdegSΔarcA by electroporation to obtain ΔdegSΔarcA:arcA ^D54E. The complement and overexpression strains were grown in LB liquid medium with 0.1% arabinose for gene expression induction.

The WT and ΔdegS strains were added into sterile LB liquid medium and shaken at 220 rpm and incubated at 37°C until the logarithmic growth phase (optical density at 600 nm [OD₆₀₀] = 0.6). Both cultures were centrifuged at 10,000 g during 10 min at 4°C. After collection, the pellets were washed twice with 50 mM PBS, and used for untargeted metabolomic analysis. This analysis was performed by Biotech-Pack Scientific Co., Ltd. (Beijing, China). The Analysis Base File (ABF) converter software was used to convert the liquid chromatography-mass spectrometry (LC-MS) raw data into the ABF format (Zhang et al., 2021). The ABF format file was imported into MS-DIAL 4.10 software for preprocessing (D'Oria et al., 2022), including peak extraction, noise removal, inverse convolution, and alignment. The three-dimensional (3D) data matrix was exported in the CSV format (raw data matrix). Finally, the extracted peak information was searched against the MassBank, Respect, and Global Natural Product Social Molecular Networking (GNPS), for a full library comparison.

All strains were grown to the stationary phase (OD₆₀₀ = 1.2) in an LB liquid medium. Bacterial cultures were collected via centrifugation at 8000 rpm for 5 min. Total RNA was extracted with TRIzol reagent and reverse-transcribed into cDNA. qRT-PCR was performed using the TB Green Premix Ex TaqII (TaKaRa Bio, Shiga, Japan) (Huang et al., 2019). The 2^-ΔΔct method was used to calculate mRNA levels relative to each other (Livak and Schmittgen, 2001). For each qRT-PCR, two independent experiments were performed, each with three technical replicates.

To examine the NADH levels of bacteria grown in LB liquid medium or M9 liquid medium (containing 0.4% glucose) to stationary phase (OD₆₀₀ = 1.2), the concentration of bacteria was adjusted to approximately 1 × 10⁸ CFU/mL. The bacteria were ultrasonically lysed. NADH levels were measured using the Amplite™ Colorimetric NADH Assay Kit (AAT Bioquest, Pleasanton, CA, USA). Briefly, equal volumes of bacterial suspension and Amplite™ Colorimetric NADH Assay Kit working solution were mixed and dispensed in wells of a clear 96-well plate, followed by incubation at 26°C for 15 min to 2 h. Read the absorbance at 460 nm using an enzyme marker and construct a standard curve using the kit’s NADH standard (Xia et al., 2021). The experiment was repeated three times.

ATP levels were determined using an ATP Assay Kit (Beyotime, Shanghai, China) following the manufacturer’s instructions. Briefly, the bacteria were cultured to stationary phase in LB liquid medium or M9 liquid medium (containing 0.4% glucose) and the bacterial concentration was adjusted to 1 × 10⁸ CFU/mL. The bacteria were ultrasonically lysed. Equal volumes of the bacterial suspension and the working solution in the ATP reagent were mixed in a black opaque 96-well plate and incubated at 26°C for 5 min (Janet-Maitre et al., 2023). Luminescence was detected using a multifunctional enzyme labeler (Thermo Fisher Scientific Inc, Waltham, MA, USA). A standard curve was plotted using the standards provided in the kit. The experiment was repeated three times.

The His-tagged recombinant ArcA protein was constructed as previously described (Wang et al., 2023b). Briefly, primers were used to amplify the full-length ArcA-encoded open reading frames. The PCR product was ligated into the pET28a vector and transformed into E. coli BL21 (DE3). Transformed bacteria expressing ArcA were grown to OD₆₀₀ = 0.6 at 37°C and induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) at 18°C for 16 h in LB liquid medium. Recombinant proteins labeled with His were purified by nickel-nitrilotriacetic acid affinity chromatography. Eight 6-week-old CD1 female mice, provided by the animal center of Zunyi Medical University (Zunyi, China), were housed in a specific pathogen-free environment and used to prepare anti-ArcA serum for western blot analysis. Mice were subcutaneously inoculated with 30 µg of recombinant ArcA on days 0, 14, and 28, along with the same volume of aluminum adjuvant. Anti-ArcA antiserum was prepared from blood collected 2 weeks after the last immunization.

All strains were grown to stationary phase (OD₆₀₀ = 1.2) in an LB liquid medium. To prepare whole bacterial proteins, the culture was subjected to centrifugation, resulting in the separation of a supernatant layer. The precipitate was resuspended with 100 μL of distilled deionized water and 25 μL of 5× sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) protein sampling buffer was added and boiled. The bacterial proteins were resolved by SDS-PAGE and transferred to a polyvinylidene fluoride membrane. After blocking, the membrane was incubated with a primary antibody (anti-ArcA serum at a dilution of 1:1000, prepared in our laboratory) overnight at 4°C. The membrane was incubated with a 1:5000 dilution of horseradish peroxidase-conjugated sheep anti-mouse IgG as a secondary antibody for 2 h after three washes with 1× Tris-buffered saline containing 0.1% Tween-20 (TBST). Finally, the membrane was washed three times with TBST, and color developed after the addition of chemiluminescent reagents (Epizyme Biomedical Technology Co. Ltd, Shanghai, China). The experiment was repeated three times.

Growth curves were generated as described previously (Kovač et al., 2021) with certain modifications. Briefly, the bacteria were cultured in LB liquid medium at 37°C until the stationary phase (OD₆₀₀ = 1.2). Aliquots of the culture were inoculated (1:500 v/v) into M9 liquid medium containing 0.4% glucose and incubated at 37°C with shaking at 200 rpm. Measurements were taken hourly for absorbance at 600 nm. The experiment was repeated three times.

Six-day-old CD1 suckling mice which were randomized into the experimental and control groups (n = 8 per group). All animal experiments were approved by the Ethics Committee of Zunyi Medical University (No. ZMU21-2301-069). All strains were grown at 37°C to stationary phase (OD₆₀₀ = 1.2), and the bacteria were collected by centrifugation at 1,200 × g for 5 min. The bacterial concentration was adjusted to 1 × 10⁷ CFU/mL with PBS. Each suckling mouse in the experimental group was gavaged with 50 μL of the bacterial suspension. The same volume of 1× PBS was used in the negative control. In the 18th hour following gavage, mice were euthanized. The small intestinal tissue was subsequently dissected, weighed, and homogenized (Zou et al., 2023). After 100-fold dilution of this preparation, 100 μL aliquots were added to 0.5 mg/L gentamicin agar plates, and the colonies were counted after 18 h of incubation at 37°C. The final results are presented as the logarithm CFU/g.

Data are expressed as mean ± standard deviation. Non-paired two-tailed t-tests were used to analyze differences between two groups, and a one-way analysis of variance was used to analyze differences between multiple groups. SPSS version 29.0 (IBM Corp., Armonk, NY, USA) was used for the analyses. P<0.05 indicated statistical significance.

Our previous RNA-seq data suggested that the knockout of degS results in the downregulation of genes related to the TCA cycle of the aerobic respiration pathway (Huang et al., 2019). The finding implies that DegS may affect aerobic respiration in V. cholerae. To further test this hypothesis, we conducted untargeted metabolomics on degS knockout mutants (ΔdegS). The analysis identified a total of 109 metabolites ( Figure 1A ; Supplementary Table 3 ). Significantly (P<0.05) differentially expressed metabolites were screened according to fold changes >2 or <0.5. A combination of multidimensional and unidimensional analyses identified 19 significantly differentially expressed metabolites ( Table 1 ). One-dimensional statistical analyses were performed using multiplicity and t-tests. The resulting data were plotted as volcano plots ( Figure 1B ). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of enriched genes mainly revealed genes involved in purine and glutathione metabolism ( Figure 1C ). Glutathione metabolism is an important component of aerobic respiration and provides important redox buffers (Lushchak, 2012; Hatem et al., 2014). Purine metabolism provides purine nucleotides that are essential for ATP production from aerobic respiration (Gessner et al., 2023). These findings indicate that DegS has an impact on aerobic respiration in V. cholerae.

Given the transcriptome and metabolome results, we hypothesized that DegS may affect the production of the aerobic respiration pathway products NADH and ATP. To test this hypothesis, we determined the levels of NADH in WT and ΔdegS strains. The NADH levels of the WT strain were approximately twice as high as those of the ΔdegS mutants, whereas the NADH levels of the complemented strain ΔdegS::degS were able to restore NADH levels close to those of the WT strain ( Figure 2A ). The pBAD24 null plasmid was unable to recover the NADH levels of ΔdegS. Subsequently, we investigated ATP levels in the above strains, which is the final energy product of the aerobic respiration pathway. The ATP level of the WT strain was approximately thrice that of ΔdegS, whereas the ATP levels of ΔdegS::degS were partially restored, with no restoration using the pBAD24 empty plasmid ( Figure 2B ). The above data suggest that DegS influences positively NADH and ATP levels in V. cholerae.

DegS regulates stress response and motility through σ^E and DegS deficiency significantly reduces σ^E activity (Ades et al., 1999). To investigate whether DegS affects V. cholerae NADH and ATP levels via σ^E, we first examined the transcript levels of rpoE. The qRT-PCR results showed that the rpoE gene transcript level in the WT strain was approximately four times higher than that of ΔdegS, while the transcript level of the rpoE gene in ΔdegS::degS was almost the same as that of the WT strain ( Figure 3A ). Next, we performed experiments for the detection of NADH and ATP levels using the rpoE deletion mutant (ΔrpoE) and corresponding complemented strain (ΔrpoE::rpoE). The NADH level of the ΔrpoE mutant was not statistically different from the WT and ΔrpoE::rpoE strains ( Figure 3B ). The ATP levels of the rpoE mutation did not differ from that of the WT and ΔrpoE::rpoE strains ( Figure 3C ). We used qRT-PCR to screen for changes in the expression of some aerobic respiratory genes in different strains, and the expression levels of genes encoding type I glyceraldehyde-3-phosphate dehydrogenase (GAP), isocitrate dehydrogenase (ICDH), and phosphoenolpyruvate carboxykinase (PckA) were significantly reduced in the ΔdegS mutant as compared with the WT strain ( Figure 3D ). However, the expression of the above genes did not change after rpoE knockout. These results indicate that DegS positively affects NADH and ATP levels in V. cholerae independent of σ^E.

In S. typhimurium, ArcA may negatively regulate ATP and NADH levels by inhibiting gene transcription levels of the pyruvate dehydrogenase complex (PDH) in the TCA cycle (Morales et al., 2013). Using qRT-PCR, we observed that the transcript level of arcA in ΔdegS was approximately six times higher than that of the WT strain ( Figure 4A ), suggesting that DegS is a negative regulator of ArcA. Therefore, we hypothesized that DegS influences the NADH and ATP levels in V. cholerae through ArcA. To assess the hypothesis, we constructed ΔdegSΔarcA and ΔdegSΔarcA::arcA and measured the levels of NADH and ATP. Both levels in ΔdegSΔarcA could be partially restored compared to ΔdegS ( Figures 4B, C ). Next, we detected the expression level of ArcA protein in each strain. ArcA protein expression was almost the same in WT strains, ΔdegS, and ΔdegS::degS ( Figure 4D ).ArcA is a response factor in a two-component system that can activate downstream genes in a phosphorylated form. Meanwhile, in E. coli, ArcA is an important inhibitor, and its phosphorylated form directly inhibits the expression of some genes in the TCA cycle, such as citrate synthase (GltA) and malate dehydrogenase (MDH) (Park et al., 2013). Therefore, we speculated whether its phosphorylation modifications are involved in this regulatory process. Next, we constructed a point mutation model (ΔdegSΔarcA::arcA^D54E ) to mimic dephosphorylation (Jeon et al., 2001) to explore whether ArcA phosphorylation is associated with DegS affecting NADH and ATP levels in V. cholerae. Both NADH and ATP levels were decreased in the ΔdegSΔarcA::arcA^D54E strain compared to the ΔdegSΔarcA strain. The ΔdegSΔarcA::arcA^D54E strain had a smaller decrease in NADH and ATP levels than the ΔdegSΔarcA::arcA strain ( Figures 4E, F ). These results suggest that DegS affects NADH and ATP levels, which are partially dependent on ArcA phosphorylation.

ICDH is a key rate-limiting enzyme of the TCA cycle; the knockdown of ICDH leads to a decrease in bacterial NADH and ATP levels (Kabir and Shimizu, 2004a). Our qRT-PCR results revealed that the transcription level of icdh in ΔdegS strains was approximately five times lower than that of WT strains and that the transcriptional level of icdh was recovered in part in the ΔdegSΔarcA strain ( Figure 5A ). These findings suggest that DegS may control the transcription of icdh through the ArcA pathway. To determine whether ICDH is involved in regulating V. cholerae NADH and ATP levels in DegS, we overexpressed ICDH based on the ΔdegS strain and measured NADH and ATP levels. Both levels were partially restored in the ΔdegS+icdh strain, but not to the level of the WT strain ( Figures 5B, C ). Collectively, these results suggest that DegS is required for high levels of ATP and NADH because it indirectly increases ICDH expression.

Many enzymes and metabolites associated with bacterial energy metabolism have direct regulatory roles in bacterial growth (Kabir and Shimizu, 2004a; Weart et al., 2007; Hill et al., 2013; Sperber and Herman, 2017). Our experiments showed that DegS affects ATP and NADH levels in V. cholerae through the ArcA-ICDH signaling pathway. To confirm whether DegS affects growth in V. cholerae through this pathway, we conducted growth curve experiments in the M9 medium. The growth rate of the ΔdegS strain was lower than that of the WT strain during the logarithmic growth phase ( Figure 6A ). Compared to the ΔdegS strain, the ΔdegSΔarcA strain grew faster during the logarithmic growth period. The ΔdegS+icdh strain had a faster growth rate than the ΔdegS strain during the logarithmic growth phase ( Figure 6B ). Concurrently, the trends of NADH and ATP levels of the ΔdegSΔarcA strain and the ΔdegS+icdh strain in M9 medium corresponded to the trends of their growth rates in the logarithmic growth phase ( Figures 6C, D ). These evidences demonstrate that DegS affects the growth of V. cholerae through the ArcA-ICDH pathway.

Inhibition of NADH and ATP production in bacteria affects their colonization (Jones et al., 2007; Schurig-Briccio et al., 2020). To examine whether the regulation of V. cholerae NADH and ATP levels mediated by DegS is critical for bacterial colonization, we used a suckling mouse model of intestinal colonization. The in vivo results showed that compared with the WT strain, the colonization capacity of the ΔdegS strain was significantly reduced, while the colonization capacity of the ΔdegS::degS strain was similar to that of the WT strain, and the colonization ability of the ΔdegS+icdh strain was partially restored ( Figure 7 ). However, the colonization ability of ΔdegSΔarcA was comparable to that of ΔdegS, and the colonization ability of ΔdegSΔarcA::arcA was stronger than that of ΔdegSΔarcA strain.