Enterocloster citroniae and related gut microbiome species modulate Vibrio cholerae biofilm formation through the production of bioactive small molecules Authors: Heidi Pauer, Saeideh Nasiri, Nathalia Santos Magalhães, Vivian T Nguyen, Nicole Victor Ferreira, Larissa D Silva Ferreira, Athena Brianne Bradshaw, Katherine E Kirby, Tyler Sabapathy, Chukwuma Great Udensi, Viktoriia Feofanova, Daniel Andrade Moreira, Thiago Estevam Parente, Jacob Wilde, David T Pride, Emma Allen-Vercoe, L Caetano M Antunes Published: 2026 ABSTRACT ABSTRACTCholera is a diarrheal disease that affects millions of people globally. Although the causative agent, Vibrio cholerae, has been extensively studied in isolation, investigation of its interactions with the gut microbiota started relatively recently. We and others previously showed that microbiota-derived metabolites significantly influence V. cholerae behavior. By investigating how an organic extract of human feces affects V. cholerae gene expression, we showed that gut metabolites strongly suppress swimming motility, a trait important for host colonization. Interestingly, extracts of pure cultures of a gut commensal, Enterocloster citroniae, recapitulated this inhibition. Here, we present a comprehensive examination of the effect of small molecules produced by E. citroniae and related species on V. cholerae behavior. We show that E. citroniae small molecules inhibit motility by various V. cholerae strains, and that several phylogenetically related species produce this activity, although the magnitude of the effect varies between strains. Using biofilm formation assays in static and flow conditions, we show that V. cholerae strongly induces biofilm formation in response to E. citroniae metabolites. Transcriptome and reporter analyses showed that several genes involved in the synthesis of an extracellular polysaccharide are induced by E. citroniae metabolites. Finally, we show that V. cholerae interactions with host cells are also modulated by this commensal. These findings advance our understanding of microbiome–pathogen interactions and how commensal bacteria influence V. cholerae virulence through the production of small molecules. In the future, this knowledge may be used to design novel microbiome-based therapeutic approaches to combat cholera and other infections.KEYWORDS: Chemical signaling, natural products, bioactive metabolites, microbe-microbe interactions, colonization resistance Introduction Vibrio cholerae is the causative agent of cholera, a diarrheal disease that continues to cause significant morbidity and mortality worldwide [1]. The behavior of V. cholerae, including how it employs multiple virulence factors to cause disease, has been extensively studied in pure cultures. Recently, studies have begun to explore V. cholerae interactions with host-associated microbes, but our understanding remains incomplete [2–5]. The gut microbiota, a complex community of microbes residing in the gastrointestinal tract of all animals, plays a vital role in protecting the host from infection, including by shaping host immune responses and by directly inhibiting pathogen growth and virulence [6]. Among the many ways in which the microbiota affects infection, one prominent mechanism is through the production of bioactive compounds that influence pathogen virulence and colonization [7–9]. Recent studies have described the role of microbiota-derived metabolites in influencing the virulence of V. cholerae. Bacteroides vulgatus has been shown to restrict V. cholerae growth during infection of antibiotic-treated mice through the production of short-chain fatty acids [3]. Blautia obeum, on the other hand, inhibits V. cholerae colonization by inhibiting virulence factor gene expression through the production of a quorum sensing signal [2]. However, interactions between microbiota members and V. cholerae are not limited to those that inhibit pathogen virulence. For instance, interactions between V. cholerae and Paracoccus aminovorans have been shown to enhance biofilm formation, which facilitates colonization and persistence in the host [10]. Therefore, complex interactions between V. cholerae and microbiome members likely occur during gut colonization. To investigate the impact of microbiome-derived small molecules on V. cholerae behavior, we previously determined the impact of an organic extract of human feces on V. cholerae global gene expression [8]. By doing so, we found that the fecal metabolome is a strong repressor of V. cholerae motility, a key virulence factor that facilitates host colonization and transmission of V. cholerae [11]. Prior to this report, our group had described that Enterocloster citroniae, a member of the human gut microbiome, produces small molecules that inhibit invasion gene expression by Salmonella enterica [7]. Therefore, we tested whether the E. citroniae strain with activity against S. enterica was also capable of recapitulating the effect of the human fecal extract on V. cholerae motility. Indeed, organic extracts of E. citroniae cultures strongly inhibit V. cholerae swimming motility [8]. In this study, we describe a comprehensive examination of the effect of small molecules produced by E. citroniae on various aspects of V. cholerae behavior. We examine the effects of E. citroniae extracts on V. cholerae global gene expression, biofilm formation, toxin production, and interactions with host cells. We also describe the effect of several gut commensals phylogenetically related to E. citroniae on V. cholerae. Altogether, our results expand our knowledge of the effects microbiome members impinge upon various aspects of V. cholerae virulence through the production of small molecules. Studying these phenomena will generate new insights into V. cholerae interactions with the microbiome and may uncover new targets for the development of microbiome-related therapeutics. Materials and methods Bacterial strains and growth conditions Strains used in this study are listed in Table S1. V. cholerae was routinely grown in Luria-Bertani (LB; Fisher Bioreagents, Waltham, MA, USA) or Brain Heart Infusion (BHI; MIDSCI, Fenton, MO, USA) medium, as indicated. BHI broth used for V. cholerae growth was supplemented with 0.5% yeast extract (MIDSCI) and hemin (5 µg/mL; Sigma-Aldrich, St. Louis, MO, USA). Cultures were incubated at 37°C with shaking at 200 rpm. Enterocloster spp. and Clostridium innocuum strains were cultured in BHI supplemented with 0.5% yeast extract, cysteine (1 g/L; Thermo Scientific, Waltham, MA USA), hemin (5 µg/mL), and menadione (0.1 mg/mL; Sigma-Aldrich) at 37°C in an anaerobic chamber (Coy Laboratory Products, Ann Arbor, MI, USA) under an anaerobic atmosphere composed of 80% N2, 5% H2, 15% CO2. For growth curve assays, an overnight culture of V. cholerae in BHI was subcultured 1:200 into 200 µL of BHI broth containing either an extract from one of the commensal strains studied (Enterocloster spp., C. innocuum), or a control extract, prepared as described below, in a 96-well polystyrene plate (TPP, Trasadingen, Switzerland). Cultures were incubated at 37°C and bacterial growth was monitored by measuring the optical density at 600 nm (OD600) using a SpectraMax i3 spectrophotometer (Molecular Devices, San Jose, USA) for 24 hours with readings taken every 30 minutes. Isolation of gut microbes Strains were isolated from donor stool samples by dilution and culture on fastidious anaerobe agar (LabM, Potters Bar, England) containing 5% defibrinated sheep’s blood (Hemostat Laboratories, Dixon, USA), and incubation at 37°C for up to 72 hours under an atmosphere of CO2:H2:N2 5:5:90 in a Baker Ruskinn Concept 500 anaerobic chamber (Baker, Sanford, USA). Isolated colonies were purified by repeated plate subculture under the same growth conditions and subsequently stored at −80°C. Genomic DNA was extracted from overnight plate cultured cell biomass using a Maxwell® 16 DNA purification kit (Promega) and identified by sequencing of the amplified 16S rRNA gene using Sanger sequencing (as a service through the University of Guelph Advanced Analysis Center). Species identifications were determined using NCBI BLAST (www.ncbi.nlm.nih.gov/BLAST/). Chemicals Short-chain fatty acids (SCFAs; Thermo Scientific) were prepared as concentrated stock solutions (1 M) of sodium acetate (99% pure), sodium propionate (99% pure), and sodium butyrate (≥98% pure) in ultrapure water, filter-sterilized (0.22 µm), and stored at −20°C until used. Stocks were diluted directly into BHI medium immediately before use to achieve final concentrations that varied based on the SCFA (acetate: 15, 30, 60 mM; butyrate: 5, 10, 20 mM; propionate: 5, 10, 20 mM), or mixed at the ratios specified in the text and figure legends. Antibiotics were obtained from Sigma-Aldrich and used at the following concentrations: kanamycin, 100 µg/mL; chloramphenicol, 10 µg/mL; polymyxin B, 20 µg/mL. Plasmids and DNA manipulation All plasmids used in this study are listed in Table S1. Escherichia coli S17-λpir harboring pCMW5 (kindly provided by Dr. Christopher M. Waters), which confers constitutive expression of the green fluorescent protein (GFP), was used as the donor strain to mobilize the plasmid into V. cholerae C6706 via conjugation. The vpsR reporter plasmid was constructed using pBBRlux [12]. The upstream sequence of vpsR (−495 to the start codon) was cloned into the BamHI and SpeI restriction sites of pBBRlux by GenScript (Piscataway, NJ, USA), to generate pLC010 (PvpsR-luxCDABE). The plasmid was first transformed into E. coli BW29427 and then transferred into V. cholerae C6706 by conjugation, as follows. All strains were grown overnight to the stationary phase at 37°C with shaking (200 rpm). V. cholerae C6706 was subcultured (1:100 dilution) and grown in LB medium at 37°C with shaking until reaching an OD600 of approximately 0.7. E. coli and V. cholerae cells were washed with LB medium by centrifugation (5 min at 13,000 xg) to remove antibiotics and were concentrated 5-fold and 20-fold in LB medium, respectively. Fifteen microliters of each culture were mixed and spotted onto LB agar plates without antibiotics. When using the BW29427 strain, an auxotrophic E. coli strain requiring 2,6-diaminopimelic acid (DAP), 300 µg/mL of DAP was added to the plates. Conjugation mixtures were incubated overnight at 37°C. The bacterial spots were scraped from the agar and resuspended in 1 mL of LB medium. One hundred microliters of the suspension were plated onto agar plates supplemented with the appropriate antibiotic corresponding to the resistance marker in the respective plasmid and incubated overnight at 37°C. Colonies were picked the day after, grown overnight, and stored at −80°C. Small molecule extraction Small molecule extraction was performed as described previously [13]. Enterocloster and C. innocuum strains were inoculated onto anaerobic blood agar plates (Remel, Lenexa, KS, USA) and incubated at 37°C under anaerobic conditions for 48 hours. A single isolated colony was then inoculated into a small volume of BHI broth (~5 mL) and incubated for 24 hours at 37°C under anaerobiosis. Subsequently, a subculture was prepared by diluting this culture 1:50 into a larger volume of BHI and incubating it anaerobically at 37°C for an additional 24 hours. Cultures of Enterocloster and C. innocuum were then extracted with ethyl acetate (Fisher Chemical, Waltham, MA). To do this, ethyl acetate was added to the culture at a 1:1 (v/v) ratio. The mixture was shaken vigorously and allowed to settle for 10 minutes; this process was then repeated once. The organic phase was then collected and transferred to a clean bottle. Control extracts were produced by extracting sterile culture media using the same procedure. Extracts were stored at −20°C for up to 4 weeks. Before experiments, the solvent was evaporated using an R-300 rotary evaporator (BÜCHI, Flawil, Switzerland) or a Savant SPD140DDA SpeedVac concentrator (Thermo Fisher Scientific, Waltham, MA, USA) at 40°C. Dried residues were resuspended in culture media, and the pH was adjusted to 7.2 before sterilization through 0.2-μm filters. The relative concentrations of extracts were defined based on the volume of organic phase recovered after extraction. Extracts resuspended in a volume of culture medium equal to the recovered organic phase were regarded as 1X. Bacterial culture and control extracts were used at a 2X relative concentration in all experiments. Motility assays To observe V. cholerae swimming motility, BHI plates containing 0.3% agar were prepared containing either control extract or extracts from Enterocloster or C. innocuum strains. The medium was poured into 100-mm Petri dishes and allowed to solidify overnight at room temperature. Plates were gently inoculated with a single colony picked from a plate containing fresh, overnight growth using a sterile toothpick and incubated at 37°C for 18 hours. Swimming was quantified by measuring the diameter of bacterial growth [8]. The effect of SCFA on V. cholerae motility was assessed under the same conditions. For the microscopic observation of swimming motility, a GFP-labeled V. cholerae strain (VC002; C6706 with pCMW5) was grown in BHI broth containing either a control extract or an extract from E. citroniae WAL17108 for 3 hours at 37°C with shaking (200 rpm). After incubation, the cultures were diluted to an OD600 of 0.8. A 1-µL aliquot of the culture was placed on a microscope slide, covered with a coverslip, and visualized using a Leica DM 5500 B microscope with a 60x objective (Leica Microsystems, Deerfield, USA). Videos were then recorded for 10 seconds. Four biological replicates were conducted, and, for each replicate, four fields were recorded. Bacterial swimming was analyzed using TrackMate, an ImageJ Fiji plugin [14]. Biofilm formation assay Biofilm formation assays (crystal violet staining) were conducted as described previously [13], with some modifications. Briefly, V. cholerae C6706 was grown overnight in LB broth at 37°C with shaking (200 rpm). Following overnight growth, the culture was diluted 1:200 in BHI broth containing either a control extract or an extract from E. citroniae WAL17108 and grown as static cultures in a 96-well polystyrene plate (TPP) at 37°C for 24 hours. After incubation, planktonic bacteria were removed from the microplate by aspiration using a micropipette, and the wells were washed three times with phosphate-buffered saline (PBS pH 7.4; Fisher Bioreagents). The biofilm was fixed at 60°C for 1 hour and stained with 200 µL of a 0.1% crystal violet (Sigma-Aldrich) solution for 15 minutes at room temperature. Excess stain was removed by rinsing the wells three times with PBS. The dye was then solubilized using 200 µL of a 95% ethanol solution, and the OD570 was measured in a Varioskan LUX Multimode Microplate Reader (Thermo Fisher Scientific). Biofilm formation was also analyzed under flow conditions using the microfluidic BioFlux 200 system (Cell Microsystems, Durham, NC). V. cholerae VC002 was grown overnight in LB broth at 37°C with shaking (200 rpm). Following overnight growth, the culture was diluted 1:100 in BHI broth containing either a control extract or an extract from E. citroniae WAL17108, and this bacterial suspension was used as the inoculum. Prior to inoculation, the channels of a 48-well Low Shear Plate (Cell Microsystems) were primed with 50 μL of PBS to fill the channels and remove air from the system. After priming, 20 µL of bacterial inoculum was added to the outlet well and perfused from the outlet to the inlet at 1 dyn/cm2 until cells filled the viewing windows. The plate was incubated statically for 1 hour at 37°C to allow for microbial attachment. One milliliter of BHI broth containing either a control extract or an extract from E. citroniae WAL17108 was then added to the inlet wells of each channel. To remove non-adherent cells, the channels were perfused from the inlet to the outlet at 1 dyn/cm2 for 1 minute. Then, the shear was reduced to 0.5 dyn/cm2 and channels were perfused with medium for 18 hours. To monitor biofilm formation, the plate was mounted on the stage of an Olympus IX81 Inverted Microscope (Evident, Waltham, USA). Biofilm formation was tracked over 18 hours, with images collected every hour using a 10x objective at two individual positions in multiple channels. After image acquisition, a representative region was selected per channel, and the average pixel intensity (mean gray value) was measured for each sample using ImageJ. The Corrected Total Cell Fluorescence (CTCF) was calculated using the following formula: CTCF = integrated density - (area of selected cell × mean fluorescence of background readings). 3’,5’-cyclic diguanylic acid (c-di-GMP) quantification assay Relative levels of 3’,5’-cyclic diguanylic acid (c-di-GMP) in V. cholerae cultures were determined using the method described by Roy and colleagues [15], with modifications. Briefly, V. cholerae C6706 was grown overnight in BHI broth at 37°C with shaking (200 rpm). The culture was diluted 1:200 in BHI broth containing either a control extract or an extract from E. citroniae WAL17108 (2X final concentration) and incubated at 37°C with shaking (225 rpm). After 4 hours of growth (late-log phase), cells were pelleted, washed in cold PBS, and extracted as described [15]. Extracts were analyzed on a Vanquish Flex UHPLC system (Thermo Fisher Scientific) using an Acclaim 120 C18 column (Thermo Scientific; 4.6 × 250 mm; 5 μm) and a flow rate of 0.2 mL/min using methanol and water as solvents. Cyclic di-GMP was analyzed by monitoring absorbance at 253 nm and calculating peak area. RNA sequencing and data analysis V. cholerae C6706 was grown overnight in LB medium at 37°C with shaking (200 rpm). The culture was diluted 1:200 in BHI broth containing either a control extract or an extract from E. citroniae WAL17108 and incubated at 37°C with shaking (200 rpm). After approximately 3 hours of growth (mid-log phase), bacterial RNA was stabilized using RNAprotect Bacteria Reagent (Qiagen, Germantown, MD, USA) according to the manufacturer’s instructions. Bacterial cells were pelleted by centrifugation, and total RNA was isolated using the SV Total RNA Isolation System (Promega, Madison, WI, USA), following the manufacturer’s recommendations. RNA samples were treated with Turbo DNase (2 U/μL; Life Technologies, Waltham, MA, USA) to eliminate residual DNA, followed by an additional purification step using the SV Total RNA Isolation System (Promega), according to the manufacturer’s recommendations. RNA concentration and purity were assessed using an ND-1000 NanoDrop spectrophotometer (Thermo Fisher Scientific). Absence of DNA contamination in the RNA was checked by quantitative Real-Time PCR, as described in the next subsection. RNA sequencing (RNA-seq) was performed on an Illumina NextSeq 2000 platform using a P1 flow cell, as single reads (100 bp), at The University of Kansas Genome Sequencing Core. After sequencing, the quality of reads was assessed using FastQC v0.12.1 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and MultiQC v1.14 [16]. Reads were then mapped against the V. cholerae C6706 reference genome (GCF_015482825.1) using STAR software v2.7.10b (parameter: --quantMode GeneCounts) [17]. Raw counts were normalized and the differentially expressed genes were determined with DESeq2 R package (Log2 fold-change ≥ 1, p-adjusted < 0.05) [18]. Quantitative real-time PCR (qRT-PCR) V. cholerae C6706 cultures were prepared under the same conditions described for RNA-seq. Bacterial RNA was stabilized using RNAprotect Bacteria Reagent (Qiagen) and extracted with the SV Total RNA Isolation System (Promega), according to the manufacturer’s protocols. Following DNase treatment and RNA cleanup (as described above), cDNA synthesis was performed using Go ScriptTM Reverse Transcription System (Promega, Madison, WI, USA). qRT-PCR was conducted on a QuantStudio™ 3 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) using PowerTrack™ SYBR™ Green Master Mix (Applied Biosystems, Waltham, MA, USA). The absence of DNA contamination in RNA samples was checked by performing cDNA synthesis reactions without reverse transcriptase in parallel to the ones containing enzyme and then checking for the absence of amplification during qRT-PCR. Gene expression levels were normalized to the housekeeping gene recA, and relative expression was calculated using the 2−ΔΔCt method [19]. Primer sequences are listed in Table S1. Transcriptional reporter assays Luminescence assays were performed using V. cholerae C6706 carrying the pLC010 vpsR transcriptional reporter plasmid (designated V. cholerae VC007). Overnight cultures were grown in LB medium supplemented with chloramphenicol (10 µg/mL) at 37°C with shaking (200 rpm). Cultures were diluted 1:200 into fresh BHI medium with chloramphenicol (10 µg/mL and containing either a control extract or an extract from E. citroniae WAL17108. Cultures were incubated at 37°C in white, 96-well plates (BRAND GmbH & Co KG, Wertheim, Germany) with intermittent shaking (5 seconds right before each reading). Luminescence and OD600 were measured every 30 minutes for 16 hours using a SpectraMax i3 plate reader (Molecular Devices). Luminescence is expressed as relative light units (RLU), calculated by dividing the luminescence signal by the corresponding OD600 to normalize for cell growth. Cell culture Human colonic epithelial cells (HT-29) were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA), and were routinely cultured in Dulbecco’s Modified Eagle Medium (DMEM; Gibco, Thermo Fisher Scientific) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; Cytiva, Marlborough, MA, USA), 1% (v/v) GlutaMAX (Gibco), 1% (v/v) non-essential amino acids (Gibco), and 1% (v/v) penicillin (100 U/mL) and streptomycin (100 μg/mL) (Gibco). Cells were maintained at 37°C in a humidified atmosphere containing 5% CO2 until reaching approximately 80–90% confluence. Cell viability and cytotoxicity assays Cell viability and cytotoxicity were evaluated using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) and lactate dehydrogenase (LDH) release assays following exposure of HT-29 cells to control extract or extracts of E. citroniae WAL17108. After 22–24 h of treatment under standard culture conditions (37°C, 5% CO2), cell viability was determined by the MTT colorimetric assay. Briefly, cells were incubated with an MTT solution (0.5 mg/mL; Sigma-Aldrich, St. Louis, MO, USA) for 2 hours. The medium was then removed and the resulting formazan crystals were solubilized in DMSO. Absorbance was measured at 570 nm using a Varioskan LUX microplate reader. Cell viability (%) was calculated as: cell viability (%) = (absorbance of treated cells × 100)/ absorbance of untreated cells. Cytotoxicity was quantified by measuring LDH release into the culture supernatant using the CytoTox 96® Non-Radioactive Cytotoxicity Assay (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Absorbance was measured at 490 nm. LDH release (%) was calculated relative to unexposed cells (0% cytotoxicity) and cells treated with a lysis solution (100% cytotoxicity). Increased LDH release was interpreted as an indicator of membrane damage. Adhesion and invasion assays V. cholerae adhesion and invasion assays were performed using HT-29 cells seeded in 24-well plates at 106 cells/well in supplemented DMEM. V. cholerae strains were grown overnight in LB broth at 37°C with shaking (200 rpm). Overnight cultures were subcultured 1:200 into BHI broth, containing either an extract from E. citroniae WAL17108 or a control extract, and grown at 37°C with shaking (200 rpm) to the mid-log growth phase (OD600 of approximately 0.5). Cultures were washed once with sterile PBS and resuspended in supplemented DMEM (no antibiotics). HT-29 monolayers were infected at a multiplicity of infection (MOI) of 200:1 and incubated for 1 hour at 37°C. For adhesion assays, following infection, monolayers were washed once with PBS to remove non-adherent bacteria. Cells were lysed with 0.025% Triton X-100 (Sigma-Aldrich) in PBS for 5 minutes at room temperature. Lysates were serially diluted and plated on LB agar to enumerate adherent bacteria by determining the number of colony-forming units (CFUs). For invasion assays, after the 1-hour infection period, cells were washed once with PBS and incubated for an additional hour in DMEM containing gentamicin (100 µg/mL) to kill extracellular bacteria. Cells were then washed with PBS and lysed with 0.025% Triton X-100 for 5 min. Intracellular bacteria were quantified by plating serial dilutions of the lysates on LB agar. Bacterial intracellular survival assay Bacterial survival assays were performed to evaluate the persistence of V. cholerae C6706 within HT-29 cells over time in the presence or absence of E. citroniae WAL17108 extracts or a control extract. HT-29 cells were seeded into 24-well plates (106 cells/well) and grown for 18 hours to reach 80–90% confluence. Cells were maintained at 37°C in supplemented DMEM in a 5% CO2 atmosphere. V. cholerae cultures were prepared as described for adhesion and invasion assays. Bacteria were washed once with PBS and resuspended in antibiotic-free DMEM, containing either an extract from E. citroniae WAL17108 or a control extract. HT-29 monolayers were infected at an MOI of 200:1 in the continued presence or the respective extracts. After a 1-hour infection period at 37°C, monolayers were washed once with PBS to remove non-adherent bacteria. Fresh supplemented DMEM containing gentamicin (100 µg/mL) and the appropriate extract (control or E. citroniae) were added, and plates were incubated further. At 2 hours post-infection, cells were washed once with PBS and the medium was replaced with supplemented DMEM containing gentamicin (10 µg/mL) and either a control or E. citroniae extract. The infections continued until 6 hours. Monolayers were then washed once with PBS and lysed with 0.025% Triton X-100 for 5 minutes, and lysates were serially diluted and plated for CFU enumeration as described above. Statistical analyses Statistical significance was determined using unpaired t tests or multiple unpaired t tests. For multiple comparisons, analysis of variance (ANOVA) was used with Bonferroni’s, Dunnett’s, or Geisser-Greenhouse correction, as indicated. Mann–Whitney tests and Welch’s t tests were performed for biofilm, motility, adhesion, invasion, and reporter assays, as indicated. Statistical analyses were performed using Prism 10 (GraphPad Software). Specific details of statistical tests, statistical significance values (“p”), sample sizes (“n”), and replicates are indicated in figure legends. For all analysis, significance was considered as p < 0.05. Ethics statement Stool samples from which strains ET129A-DCM99, ET325A-NB6, ET129A-LIQR2, ET144A-DCM5, ET233A-FMU6 and ET325-BBE13 were isolated were collected in a study approved by the University of California San Diego Administrative Panel on Human Subjects in Medical Research under protocol #160524. This study did not require informed consent on behalf of the study subjects because it involved the use of existing records and specimens that were recorded in such a manner that the subjects could not be identified directly or through identifiers linked to the subjects. All human research described herein was performed in accordance with the principles stated in the Declaration of Helsinki. Data availability statement Raw RNA-seq data was deposited in Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/), under study accession number GSE313194. The data can be directly accessed through the following link: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc = GSE313194 [20]. A table containing the processed RNA-seq data can also be found at https://doi.org/10.6084/m9.figshare.31712785 [21]. Results Bioactive molecules from gut commensals inhibit V. cholerae motility We previously showed that a strain of E. citroniae, FM-V5-E, produces compounds that inhibit V. cholerae swimming motility [8]. For these original studies, we used VC833, a multidrug-resistant V. cholerae strain isolated from human feces during a cholera outbreak in Nigeria [22]. To determine whether this is a strain-specific phenomenon, we assessed the ability of E. citroniae extracts to modulate swimming motility of other V. cholerae strains, include 1 Classical and 3 El Tor biotype strains (Table S1). Swimming assays were performed on 0.3% BHI agar plates containing either a control extract or an extract from E. citroniae FM-V5-E. Exposure to the E. citroniae extract significantly reduced the motility diameter of all V. cholerae strains tested compared with the control (Figure 1). Next, to determine whether this effect is specific to E. citroniae FM-V5-E or prevalent among other E. citroniae strains as well as other genetically related species of gut commensals, extracts from additional E. citroniae strains, strains of other Enterocloster species (E. aldenensis, E. bolteae, E. clostridioformis), and strains of C. innocuum were tested against V. cholerae C6706. As can be seen in Figure 2, with the exception of E. clostridioformis, all species had at least one strain that displayed activity against V. cholerae swimming. However, the prevalence of bioactivity varied between species; while all E. citroniae [5] and E. aldenensis [5] strains showed activity, most E. bolteae (6/9) and all E. clostridioformis [2] strains were inactive. Interestingly, C. innocuum showed high variation, with some highly active strains and some completely inactive ones. In total, among the 32 strains examined, 20 significantly decreased V. cholerae motility (Figure 2). Bioactive molecules produced by E. citroniae inhibit swimming motility of multiple V. cholerae strains. Swimming motility of V. cholerae strains C6706, N16961, O395, and 1276 were assessed on 0.3% agar BHI plates containing either a control or E. citroniae (FM-V5-E) extract. The diameter of motility zones was measured after 18 hours of incubation at 37°C. Each data point represents an independent biological replicate (n = 9). Bars indicate the mean ± SEM. Statistical significance was determined by unpaired t-tests; ****p < 0.0001 compared with the corresponding control. Multiple bacterial species phylogenetically related to E. citroniae inhibit V. cholerae motility. Swimming motility of V. cholerae C6706 is expressed as a percentage of control after exposure to extracts from 32 gut isolates, including E. aldenensis, E. citroniae, E. bolteae, E. clostridioformis, and C. innocuum. Bars show mean ± SEM. Statistical significance was determined by one-way ANOVA with Dunnett’s multiple-comparisons test; **p < 0.005, ****p < 0.0001. Each data point represents an independent biological replicate (n ≥ 7). To directly observe the effect of microbiome-derived bioactive compounds on V. cholerae motility, at the cellular level, we tracked fluorescently labeled V. cholerae C6706 (VC002) in the presence or absence of a control extract or extracts from E. citroniae WAL17108 [Figure 3 ; Movies S1 (control) and S2 (E. citroniae extract)]. Consistent with the results obtained from motility assays in agar plates, V. cholerae grown in the presence of E. citroniae extracts exhibited reduced swimming distance (Figure 3(A)). Additionally, increased bacterial aggregation was observed in cultures treated with E. citroniae extracts (Figure 3(B)) compared to the control, suggesting that biofilm formation may also have been affected. Bioactive molecules produced by E. citroniae reduce swimming distance and promote aggregation of V. cholerae. (A) Quantification of swimming distances of fluorescently labeled V. cholerae VC002 tracked by live imaging in the presence or absence of an E. citroniae (WAL17108) or control extract. Data are mean ± SEM; ****p < 0.0001 (Welch’s t test). Four biological replicates were conducted, and for each replicate, four fields were recorded. (B) Representative motility tracks and fluorescence micrographs showing reduced motility and increased aggregation upon exposure to the extract. E. citroniae produces bioactive molecules that increase biofilm formation To determine whether E. citroniae metabolites influence V. cholerae biofilm formation, we performed a crystal violet staining assay in polystyrene, 96-well plates. Indeed, V. cholerae C6706 grown in the presence of an E. citroniae extract (WAL17108) displayed significantly higher biofilm biomass compared with the control (Figure 4(A)). To further characterize this phenotype, biofilm formation was analyzed in real time using the BioFlux microfluidic system, which allows dynamic visualization of biofilm development under controlled shear flow. V. cholerae VC002 successfully formed biofilms under a shear stress of 0.5 dyne/cm2 in both conditions; however, fluorescence quantification after 18 hours revealed a substantial increase in biofilm accumulation in the presence of E. citroniae extracts (Figure 4(B)). Representative fluorescence images confirm the enhanced surface coverage of V. cholerae biofilms when exposed to the E. citroniae extract (Figure 4(C)). Time-lapse movies of GFP-labeled biofilms can be seen in Movie S3. Because biofilm formation is affected by c-di-GMP signaling in V. cholerae and many other bacteria [23], we set out to determine the levels of this secondary messenger in V. cholerae cultures grown in the absence or presence of E. citroniae extracts. To do this, V. cholerae BHI overnight cultures were subcultured in BHI with or without the addition of E. citroniae extracts at 2X. Once cultures reached the late-log growth phase, c-di-GMP was extracted as described [15]. Relative levels of c-di-GMP were then determined by HPLC. As can be seen in Figure 4(D), growth in the presence of E. citroniae extracts caused a small (~11%) but statistically significant increase in c-di-GMP levels. Bioactive molecules produced by E. citroniae enhance biofilm formation by V. cholerae. (A) Biofilm biomass was quantified by crystal violet staining after 18 h of static incubation at 37°C in BHI medium containing either an E. citroniae (WAL17108) or control extract. Optical density was measured at 570 nm; ****p < 0.0001 (Mann–Whitney test). Three biological replicates were performed, each consisting of five technical replicates. (B) Representative fluorescence images of GFP-labeled V. cholerae VC002 biofilms formed in the BioFlux system after 18 h under 0.5 dyne/cm2 flow, showing increased biofilm density in the presence of an E. citroniae (WAL17108) extract. (C) Quantification of fluorescence intensity from the experiment in panel B; ***p < 0.001 (Mann–Whitney test). Five biological replicates were performed, with two images collected per replicate. (D) Relative levels of c-di-GMP in V. cholerae C6706 cultures grown in the presence of a control or E. citroniae (WAL17108) extract. Data shown represent peak area divided by the OD600 of cultures. **p < 0.005 (Welch’s test). Microbiome-derived extracts do not markedly affect V. cholerae growth To determine whether gut microbiota members produce compounds that significantly affect V. cholerae growth, bacterial cultures were incubated with or without extracts from Enterocloster spp. or C. innocuum strains, and V. cholerae growth was assessed through measurements of the OD600 of cultures overtime, in a 96-well plate using a plate reader. As shown in Figure 5, V. cholerae growth was only slightly delayed in the presence of extracts from E. citroniae WAL17108. Other E. citroniae and E. aldenensis strains also only slightly delayed V. cholerae growth (Figure S1A), particularly during the exponential phase of growth. In contrast, extracts from E. clostridiformes, E. boltae, and C. innocuum did not affect bacterial growth (Figures S1B and S1C). These results suggest that the effects observed on motility and biofilm formation are not due to general effects on bacterial growth. E. citroniae extracts do not significantly impact V. cholerae growth. Cultures of V. cholerae C6706 were grown in BHI broth containing either a control or an E. citroniae (WAL17108) extract at 37°C, and growth was measured by absorbance (OD600) for 12 h in a microplate reader. Results represent the average of three independent cultures, each consisting of two technical replicates (n = 6), and bars show the SEM. Short-chain fatty acids do not account for the bioactivity of E. citroniae extracts Many members of the gut microbiome are known to produce short-chain fatty acids (SCFA), which have been extensively shown to display bioactivity [reviewed in [24]. Because SCFAs are among the most abundant and well-characterized metabolites produced by the gut microbiota that are known to modulate bacterial physiology and host–pathogen interactions, we further assessed whether the phenotypic effects of E. citroniae extracts on V. cholerae could be attributed to SCFAs [25]. Therefore, we tested the effects of acetate, propionate, and butyrate, individually and in combination, across a range of physiologically relevant concentrations, on V. cholerae swimming motility. As shown in Figure 6(A), none of the tested SCFAs significantly altered V. cholerae motility compared with the control condition. Similarly, biofilm formation was not enhanced by SCFA treatment (Figure 6(B)), in contrast to the marked induction observed in the presence of E. citroniae extracts. Also, growth curve analyses showed butyrate and propionate, as well as SCFA combinations that included these two SCFA, caused marked reductions in V. cholerae growth (Figure 6(C)). These findings indicate that SCFAs alone are not responsible for the inhibition of swimming motility and promotion of biofilm formation by V. cholerae caused by E. citroniae extracts, and that other bioactive metabolites must be responsible for modulating V. cholerae behavior. Short-chain fatty acids are not responsible for the effects of E. citroniae extracts on V. cholerae motility and biofilm formation. (A) Swimming motility of V. cholerae C6706 was assessed on 0.3% BHI agar plates supplemented with SCFAs at the indicated concentrations. The diameter of motility zones was measured after 18 h of incubation at 37°C. Each data point represents an independent biological replicate (n = 6). Bars indicate the mean ± SEM. One-way ANOVA determined statistical significance; (B) Quantification of biofilm biomass by crystal violet staining (OD570 nm) after 18 h of static growth at 37°C in BHI supplemented with SCFAs at the indicated concentrations. ***p < 0.001 compared with control (one-way ANOVA). (C) Growth curves of V. cholerae in BHI supplemented with SCFAs at the indicated concentrations, monitored by OD600 over 12 h. E. citroniae molecules modulate V. cholerae global gene expression To assess the impact of E. citroniae metabolites on V. cholerae transcriptional activity, we compared the transcriptome of cells grown to the mid-log growth phase in the presence of an E. citroniae extract (WAL17108) vs a control extract. Sequencing produced approximately 126 million high-quality reads, with a unique mapping rate over 92.7% to the reference genome (Table S2). A total of 576 genes (approximately 15% of the genome) were differentially expressed in response to the E. citroniae extract, including 238 upregulated and 338 downregulated genes (Table S2). Functional categorization of differentially expressed genes using EggNOG [26] revealed that, among upregulated genes, the most affected categories were energy production and conversion (13.4%), amino acid transport and metabolism (12.4%), and cell wall/membrane/envelope biogenesis (10.6%) (Figure 7). In contrast, downregulated genes were enriched for functions related to carbohydrate transport and metabolism (8.7%), nucleotide transport and metabolism (5.6%), replication/recombination/repair (5.6%), and cell motility (4%) (Figure 7). Notably, genes associated with biofilm formation and nitrate reductase activity were upregulated. Conversely, several virulence-associated loci - including components of the type VI secretion system (T6SS), RTX toxin, and toxin-coregulated pilus (tcp) cluster – were significantly downregulated, indicating that E. citroniae metabolites suppress key virulence pathways in V. cholerae (Table 1). To confirm the transcriptomics results, relative mRNA levels of selected genes were quantified by qRT-PCR during V. cholerae growth in the absence or presence of an E. citroniae extract. The extract significantly upregulated genes associated with biofilm formation (vpsR, vpsT, vpsL, rbmA, bap1), whereas virulence-related genes, including tcpA and rtxC, were markedly downregulated (Figure 8(A)). Additionally, we confirmed the effects of E. citroniae extracts on several genes involved in motility and chemotaxis. As can be seen in Figure 8(B), the expression of flrA, motB, flgK, mlp37, acfB, and VC1248 was significantly repressed by growth in the presence of an E. citroniae extract. Altogether, these qRT-PCR findings are consistent with the RNA-seq data, further supporting that E. citroniae-derived metabolites repress motility and virulence pathways while promoting biofilm-associated gene expression in V. cholerae. Functional categorization of V. cholerae genes differentially expressed during growth in the presence of an E. citroniae extract. Genes showing ≥ 2-fold changes in expression (based on RNA-seq analysis) were grouped according to predicted functions using EggNOG (v5.0.0). Black bars indicate the percentage of upregulated genes, whereas gray bars represent the percentage of downregulated genes in cultures grown with the E. citroniae extract. Table 1.V. cholerae genes involved in biofilm formation, virulence, and other relevant pathways that were differentially expressed in response to small molecules produced by E. citroniae.AnnotationGeneFold changeaUpregulated periplasmic nitrate reductase subunit alphanapA72.0chaperone NapD 61.8nitrate reductase cytochrome c-type subunit 52.0ferredoxin-type protein NapFnapF51.6NapC/NirT family cytochrome c 28.9hypothetical protein 18.9exopolysaccharide biosynthesis glycosyltransferase VpsLvpsL13.4UDP-N-acetyl-D-mannosamine dehydrogenase VpsBvpsB9.56hypothetical proteinrbmA9.43UDP-N-acetylglucosamine 2-epimerase (non-hydrolyzing) VpsAvpsA8.97exopolysaccharide biosynthesis beta-barrel protein VpsMvpsM8.70exopolysaccharide export protein VpsNvpsN7.83beta-prism lectin domain-containing proteinbap17.13exopolysaccharide biosynthesis acetyltransferase VpsCvpsC6.37beta-prism lectin domain-containing proteinrbmC6.05exopolysaccharide regulatory protein-tyrosine phosphatase VpsUvpsU5.58O-antigen ligaserbmD4.05exopolysaccharide biosynthesis acetyltransferase VpsGvpsG3.71exopolysaccharide biosynthesis glycosyltransferase VpsDvpsD3.68LuxR family transcriptional regulator VpsTvpsT3.38exopolysaccharide regulatory tyrosine autokinase VpsOvpsO3.28exopolysaccharide biosynthesis protein VpsHvpsH3.11exopolysaccharide biosynthesis glycosyltransferase VpsKvpsK3.05exopolysaccharide biosynthesis flippase VpsEvpsE3.00exopolysaccharide biosynthesis glycosyltransferase VpsIvpsI2.80glycosyl hydrolase family 28-related proteinrbmB2.78exopolysaccharide biosynthesis protein VpsJvpsj2.57cyclic-di-GMP-binding transcriptional regulator VpsRvpsR2.46exopolysaccharide biosynthesis protein VpsFvpsF2.37Downregulated toxin-coregulated pilus major pilin TcpAtcpA−7.18RTX toxin-activating lysine-acyltransferase RtxCrtxC−6.12toxin-coregulated pilus methyl-accepting chemotaxis protein TcpItcpI−5.79toxin-coregulated pilus pre-pilin peptidase TcpJtcpJ−5.60MARTX multifunctional-autoprocessing repeats-in-toxin holotoxin RtxArtxA−4.31pilus/toxin transcriptional regulator ToxTtoxT−3.69type VI secretion system baseplate subunit TssFtssF−3.32type VI secretion system baseplate subunit TssGtssG−3.23toxin-coregulated pilus minor pilin TcpBtcpB−2.57type VI secretion system contractile sheath small subunittssB−2.51type VI secretion system contractile sheath large subunittssC−2.46RTX toxin T1SS ABC transporter subunit RtxBrtxB−2.10aExpression levels observed in the presence of the E. citroniae extract were divided by expression levels obtained in the absence of the extract. Negative signs indicate repression by the extract. V. cholerae genes involved in biofilm formation, virulence, and other relevant pathways that were differentially expressed in response to small molecules produced by E. citroniae. aExpression levels observed in the presence of the E. citroniae extract were divided by expression levels obtained in the absence of the extract. Negative signs indicate repression by the extract. Validation of V. cholerae transcriptional responses to E. citroniae extracts by quantitative PCR. (A) Relative mRNA levels of selected virulence and biofilm-related genes were quantified in V. cholerae C6706 grown containing either a control or an E. citroniae (WAL17108) extract. (B) Relative expression levels of motility and chemotaxis-related genes under the same conditions. The expression of vpsR is also shown in panel B as a positive control of regulation by E. citroniae extracts. In both panels, expression was normalized to the housekeeping gene recA and presented as Fold change relative to the control. Bars represent means ± SEM from three independent biological replicates. Each qRT-PCR reaction was performed in duplicate. Statistical significance was assessed using unpaired t-tests (*p < 0.05, **p < 0.005, ***p < 0.001, ****p < 0.0001). Whole fecal extracts and E. citroniae culture extracts display distinct effects on the V. cholerae transcriptome To evaluate how E. citroniae metabolites influence V. cholerae transcription compared to the effects of the more chemically complex human fecal matrix, we compared our RNA-seq data with our previously published transcriptomic dataset of V. cholerae grown in the presence of a human fecal extract [8]. A scatter plot comparing the log2 fold changes under both conditions is shown in Figure 9. This analysis revealed a subset of genes with shared regulation across both conditions, as well as many other genes that respond specifically to E. citroniae metabolites. A subset of genes (e.g. the vps cluster) that were strongly upregulated in the E. citroniae extract remained unchanged or were mildly downregulated in the fecal extract, suggesting specific regulation of biofilm pathways by E. citroniae signals. Together, these data indicate that while both fecal and E. citroniae-specific metabolites markedly modulate V. cholerae global transcription, E. citroniae metabolites elicits a distinct and more targeted regulatory response compared to the complex and diverse signals present in the whole fecal extract. Correlation between transcriptional profiles of V. cholerae grown in the presence of a fecal extract or an E. citroniae extract. Each point represents a gene, plotted by its log2 fold change in V. cholerae exposed to an E. citroniae extract (x-axis) vs its log2 fold change in a fecal extract (y-axis). Genes are color coded by regulation pattern: purple (downregulated in both conditions), orange (upregulated in both), light blue (downregulated in E. citroniae extract only), dark blue (upregulated in E. citroniae extract only), green (upregulated in fecal extract only), yellow (downregulated in fecal extract only), and gray (not significantly changed in either condition). Vibrio polysaccharide (vps) gene expression is activated by E. citroniae culture extracts To confirm the effect of metabolites produced by E. citroniae on vps gene expression, we created a reporter system by fusing the promoter region of vpsR to the lux operon in pBBRlux [12], creating pLC010. The vpsR gene encodes a cyclic-di-GMP-binding transcriptional regulator that is critical for the expression of other vps genes involved in the synthesis of Vibrio polysaccharide, known to be important for biofilm formation [27,28]. In our RNA-seq data, vpsR expression was induced 2.46-fold when V. cholerae was cultured in the presence of an E. citroniae extract. Therefore, we grew our V. cholerae vpsR reporter in the presence of a control extract or an E. citroniae extract (WAL17108), and measured growth and light production. As can be seen in Figure 10, light production was significantly increased when V. cholerae was grown in the presence of an E. citroniae extract, confirming that vps gene expression is activated by compounds produced by E. citroniae. Expression of vpsR in a V. cholerae reporter strain in response to E. citroniae extracts. Light production by a reporter strain containing a PvpsR-lux fusion in the absence and presence of E. citroniae extracts (2X concentration). Statistical significance was assessed using the Welch’s test comparing each treatment to the corresponding untreated control. *p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.0001. E. citroniae extracts modulate V. cholerae interactions with host cells Due to the observed effects of E. citroniae extracts on the expression of several traits that are important for host interactions, we set out to determine if interactions of V. cholerae with host cells would be affected by E. citroniae culture extracts. Before doing so, we sought to determine if the compounds produced by E. citroniae affect host cells, by performing toxicity and viability assays using a human colon epithelial cell line (HT-29). After 24 hours of exposure of HT-29 cells to E. citroniae extracts (WAL17108), a concentration-dependent cytotoxic effect was observed, as shown by MTT and LDH assays. MTT results (Figure 11(A)) revealed a significant reduction in cellular metabolic activity when cells were incubated in the presence of E. citroniae extracts, as compared to the DMEM control, with a decrease to approximately 5% of the control at the highest concentrations used (10X and 5X), 35% with 2X, and 45% with 1X, remaining significantly lower than the control at all tested concentrations. Interestingly, however, a control extract produced from bacterial culture medium also caused a significant, although less pronounced, reduction in metabolic activity, suggesting that medium components are at least partially responsible for this effect. Nevertheless, it is important to note that a significant difference between the E. citroniae and control extracts was still observed at a 2X concentration, the concentration used for most experiments described herein. Similarly, the LDH assay indicated increased membrane damage at higher E. citroniae extract concentrations when compared to the DMEM control (Figure 11(B)), with LDH release reaching approximately 50% at the 10X concentration, and decreasing progressively at 5X, 2X, and 1X. In this assay, however, control extracts did not induce any toxicity. Of note, no significant differences were observed between the E. citroniae and control extracts at a 2X concentration, the concentration used for most experiments described herein. These findings confirm that, although the effect is somewhat moderate, E. citroniae extracts induce dose-dependent cytotoxicity toward HT-29 cells. Cytotoxic activity of E. citroniae extracts in HT-29 cells. (A) Cell viability was assessed using the MTT assay after treatment with E. citroniae extracts at concentrations of 10X, 5X, 2X, and 1X. Data are presented as means ± SEM from five replicates. (B) Cytotoxicity was evaluated using an LDH release assay in HT-29 cells exposed to E. citroniae extracts at concentrations of 10X, 5X, 2X, and 1X, and expressed as percent cytotoxicity. LDH release from cells treated with a lysis solution served as positive control. Statistical significance was determined by one-way ANOVA (*p < 0.05, ***p < 0.001, ****p < 0.0001). During infection, V. cholerae must interact with host cells in the intestinal tract. Although V. cholerae has historically been considered a noninvasive pathogen, others have shown that it is capable of invading intestinal epithelial cells in vivo and in vitro [29,30]. To determine if bioactive molecules produced by E. citroniae modulate these interactions, we performed cell adhesion and invasion assays using V. cholerae cells that had been grown in the presence of a control extract or E. citroniae extracts (WAL17108). To do this, V. cholerae was grown under both conditions until mid-log growth was achieved. Cells were then washed and used to infect HT-29 cells. This was done so the epithelial cells were not exposed to E. citroniae extracts, avoiding the issue of toxicity. As can be seen in Figure 12, there was a modest, yet noticeable reduction in bacterial counts in both the adhesion and invasion assays in the presence of E. citroniae extracts compared to the control. In our adhesion assays, the control group exhibited approximately 2.1 × 105 CFU/mL, whereas treatment with the E. citroniae extracts reduced adhesion to approximately 1.2 × 105 CFU/mL, representing a significant decrease (Figure 12(A)). Similarly, invasion was significantly reduced in the presence of E. citroniae extracts, with CFU counts decreasing from approximately 1.4 × 103 CFU/mL in the control condition to approximately 8 × 102 CFU/mL in the group treated with E. citroniae extracts (Figure 12(B)). These results indicate that E. citroniae extracts inhibit both adhesion and invasion of HT-29 cells by V. cholerae. Lastly, in order to evaluate the impact of E. citroniae extracts on V. cholerae survival following host cell invasion, an overnight V. cholerae culture was subcultured in the presence of control or E. citroniae extracts until reaching the mid-log growth phase (OD600 of approximately 0.5). This time, however, both HT-29 cells and V. cholerae were exposed to either condition (control extracts or E. citroniae extracts) at the time of infection. However, exposure occurred for only 6 hours to minimize potential cytotoxic effects. At this time point, a significant decrease in V. cholerae intracellular counts was still observed in the presence of E. citroniae extracts (Figure 12(C)). Also, to ascertain that this result was not a consequence of differences in the numbers of epithelial cells in both conditions, we normalized the number of bacterial CFU to the number of HT-29 cells. This still resulted in significantly reduced numbers of intracellular V. cholerae in the presence of E. citroniae extracts (Figure 12(D)). Effect of E. citroniae extracts on the adhesion, invasion, and survival of V. cholerae in epithelial cells. Adhesion to (A) and invasion of (B) HT-29 cells by V. cholerae grown in the presence or absence of E. citroniae extracts (2X concentration). (C) Intracellular survival of V. cholerae was also assessed by CFU counting at 6 hours post-infection. (D) Data in panel C was also analyzed by dividing the number of CFUs observed by the number of eukaryotic cells in the wells, in order to normalize the data for potential variations in cell numbers between conditions. For these assays, bacteria were grown with E. citroniae extracts prior to infection and were also exposed to the extracts throughout the infection period. Statistical significance was assessed using unpaired t-tests comparing each treatment to the corresponding untreated control (*p < 0.05, **p < 0.01, ***p < 0.001). Discussion Our study demonstrates that metabolites produced by E. citroniae profoundly influence the physiology of V. cholerae, revealing the role of metabolites produced by the gut microbiota in shaping pathogen behavior. We show that exposure to E. citroniae extracts triggers a coordinated shift from a motile, virulent phenotype toward a sessile, biofilm state. This transition is supported by phenotypic assays and transcriptomic data, which together indicate that microbiota-derived metabolites can reprogram V. cholerae behavior in ways that may impact colonization and persistence within the host. One of the most striking findings was the strong induction of biofilm formation in the presence of E. citroniae extracts, observed under both static and flow conditions. RNA-seq analysis revealed upregulation of the entire vps gene cluster, as well as biofilm matrix components such as rbmA and bap1. These genes are essential for Vibrio polysaccharide (VPS) synthesis and biofilm maturation, processes that enhance environmental survival and host colonization [27,28,31]. The activation of vpsR and vpsT, key regulators of cyclic-di-GMP signaling, further supports the notion that E. citroniae metabolites promote a high-c-di-GMP state, favoring biofilm development over motility. This is consistent with our motility assays, which showed a marked reduction in swimming behavior, and aligns with previous reports that microbiota signals can modulate virulence factors in enteric pathogens [2,7,8]. Interestingly, besides genes involved in biofilm formation, genes encoding nitrate reductases were among the most highly upregulated in response to E. citroniae metabolites, despite the use of aerobic growth conditions. Nitrate respiration is typically associated with anaerobic environments such as the gut lumen and has been linked to enhanced colonization in murine models [32,33]. The activation of this pathway under oxygen replete conditions may suggest that specific metabolites produced by this gut commensal may mimic gut-associated cues, priming V. cholerae for survival in niches where oxygen is limited. This metabolic reprogramming, coupled with increased biofilm formation, may represent an adaptive strategy that facilitates long-term persistence within the host or environmental reservoirs. Besides the aforementioned transcriptional changes, E. citroniae metabolites strongly repressed genes that encode major virulence determinants, including the toxin-coregulated pilus (tcpA), RTX toxin (rtxA, rtxC), and components of the type VI secretion system (T6SS). These systems are critical for host colonization, cytotoxicity, and interbacterial competition [34–37]. Downregulation of ToxT, the master regulator of virulence gene expression, suggests that E. citroniae metabolites interfere with upstream regulatory cascades, potentially reducing the pathogen’s ability to cause acute disease. This dual effect, biofilm promotion and virulence suppression, may reflect a trade-off between transmission and persistence, favoring a less aggressive state under microbiota influence. Conversely, it is possible that the effect of cues produced by E. citroniae on V. cholerae virulence is more nuanced, and may be involved in regulating the timing, rather than the intensity of the response. Similar antagonistic regulation of motility and virulence by gut metabolites has been reported for other pathogens, but our findings highlight a specific and potent effect of E. citroniae metabolites on V. cholerae physiology [2,7,8]. During V. cholerae host colonization and pathogenesis, two early and essential steps are the ability of the bacteria to actively traverse the intestinal mucus layer and adhere to the epithelial cell surface. Our analysis of the E. citroniae extracts demonstrated that these metabolites significantly impaired not only swimming motility but also the ability of V. cholerae to adhere to epithelial cells, indicating a direct inhibitory effect on another key virulence-associated behavior. Although V. cholerae is not considered an invasive organism, our data confirmed earlier reports by showing that V. cholerae can invade epithelial cells in vitro [29,30]. Interestingly, the ability to invade host cells was also repressed by compounds produced by E. citroniae, contributing to the multifaceted nature of the effects elicited by this commensal. Regarding additional aspects of host–pathogen interactions, one caveat of our study is that we did not evaluate the impact of E. citroniae bioactivity in a more complex in vivo model, such as a mouse model. Although such experiments can be informative, no good in vivo model that truly allows the study of the contributions of the microbiota to V. cholerae infection currently exists. Although the suckling mouse model has been used for decades, it relies on animals whose immune systems (and microbiomes) are not fully developed, and therefore it does not provide a good representation of the infection dynamics in humans. A more recent model was also described, where the gut microbiome of adult mice is depleted by antibiotic treatment prior to infection, increasing susceptibility to V. cholerae infection [3]. Although we have been able to recapitulate this finding in our lab, this model is also inappropriate to determine the impact of E. citroniae bioactivity on resistance to V. cholerae, since antibiotic treatment may also impact E. citroniae survival and growth in the gut. In our view, the most appropriate way to determine the impact of E. citroniae bioactivity on V. cholerae infection would be to colonize germ-free animals with complex, defined communities of bacteria that differed from one another solely with regards to the genetic status of the E. citroniae strain used. By colonizing animals with either a wildtype, bioactive E. citroniae vs an isogenic mutant that does not produce bioactivity, one could irrefutably attribute the potential effects on V. cholerae colonization to bioactivity. At the moment, this experiment is impossible, as no genetic tools have been developed for E. citroniae. Even once such tools are available, this would be a significant undertaking, as the communities used should be as complex as possible to recapitulate the functions of an intact microbiome, but at the same time defined in their content so that strain composition can be manipulated. Another caveat of our study is the apparent discrepancy between the regulation of toxin co-regulated pilus genes without the corresponding regulation of cholera toxin, as these virulence factors are known to be co-regulated by ToxR [38]. Although we currently lack an explanation for this, one possibility is that the chemical complexity of E. citroniae extracts offers varied environmental cues that may uncouple transcriptional regulation of these two genetic loci. Alternatively, compounds produced by E. citroniae may differentially affect transcript levels through post-transcriptional mechanisms. Both these hypotheses need experimental verification. Nevertheless, limitations notwithstanding, the observations presented herein demonstrate the complexity of microbiota–pathogen interactions during cholera infection. While previous studies have shown that certain commensals inhibit V. cholerae growth or virulence through short-chain fatty acids or quorum-sensing signals [2,3], our data indicate that E. citroniae employs distinct metabolites to modulate biofilm formation and metabolic pathways. Importantly, the ability of microbiota-derived molecules to suppress virulence while promoting biofilm formation could influence disease outcomes by reducing acute pathogenicity yet enhancing environmental persistence, potentially affecting transmission dynamics. These findings underscore the role of microbiota molecules as environmental cues that reprogram V. cholerae toward persistence rather than acute infection, potentially reshaping disease dynamics and transmission. By promoting biofilm formation and metabolic adaptations while suppressing virulence, these metabolites may influence both host–pathogen interactions and environmental survival. However, further studies are essential to identify the specific compounds responsible and elucidate their mechanisms of action. Such insights will deepen our understanding of the microbiota-pathogen interplay and pave the way for microbiome-based strategies to mitigate cholera and related enteric infections. Supplementary Material Supplemental Material Acknowledgements We thank Chris Waters for the kind gift of V. cholerae C6706, pBBRlux, and pCMW5. We thank Neta Sal-Man for the kind gift of V. cholerae 1276. We also thank Carolina Tropini for kindly sharing Enterocloster bolteae strains BAA-613 and CC43_001B, as well as Enterocloster clostridioformis strains 2_1_49FAA and WAL7855. We are also grateful to Sarah Vancuren and Jerry He for assistance with growing, preparing, and shipping gut microbiome isolates between laboratories. Funding Statement Research reported in this publication was supported by start-up funds from the University of Kansas to LCMA. Funders had no role in the design of the research, data collection, analysis, interpretation, or writing of the manuscript. This work also used computational resources provided by the RPT04A Bioinformatics Core Facility at Fiocruz, Rio de Janeiro. Disclosure statement No potential conflict of interest was reported by the author(s). Supplementary material Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2026.2712696. References Associated Data Supplementary Materials Supplemental Material