Fusobacterium varium Exacerbates Neutrophil-driven Intestinal Inflammation Associated with Succinate–SUCNR1–NF-κB Signaling Authors: Qian Wang, Han Xiao, Wenxiu Liu, Xiangji Dang, Yanrui Bai, Ruihong Xiao, Liyu Tong, Yan Wang, Mi Li, Shuhan Wang, Sugui Pu, Duoji Pei, Dekui Zhang, Xiang Wang, Guang Hu, Jingjing Guo, Xiaojie Jin, Long Qin, Chunjiang Zhang, Yana Li, Tiansheng Zhang, Jing Yang, Qi Wang, Hui Sun Published: 2026 ABSTRACT AbstractDisruption of gut microbial homeostasis is a hallmark of ulcerative colitis (UC), yet the specific pathobionts and effector molecules driving mucosal inflammation remain unclear. In this study, metagenomic sequencing of fecal samples from 37 patients with UC and 30 healthy controls was performed to identify differentially enriched bacterial species. Fusobacterium varium (F. varium) was found to be significantly enriched in patients with UC and was therefore selected for further functional investigation. Germ-free mice colonized with F. varium developed more severe dextran sulfate sodium (DSS)-induced colitis, accompanied by enhanced mucosal inflammation. In addition, F. varium culture supernatants increased NF-κB reporter activity and inflammatory signaling at the protein level. Bioactivity-guided fractionation combined with mass spectrometry identified succinate as a major candidate bacterial-derived bioactive metabolite. Succinate exacerbated colonic inflammation in vivo and promoted neutrophil recruitment, whereas inhibition of CXCR2 signaling reduced neutrophil infiltration and alleviated disease severity. In vitro experiments further demonstrated that succinate activated NF-κB signaling in HL-60–derived neutrophils through succinate receptor 1 (SUCNR1) and induced the production of inflammatory mediators, including CXCL8. Collectively, these findings support a role for F. varium in exacerbating intestinal inflammation under colitic conditions, at least in part through succinate-associated neutrophil recruitment.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s10753-026-02547-x.Keywords: Fusobacterium varium, Ulcerative colitis, Neutrophils, NF-κB, Succinate Introduction Ulcerative colitis (UC) is a chronic, immune-mediated inflammatory disease characterized by continuous mucosal inflammation limited to the colon [1]. Clinically, patients typically present with abdominal pain, diarrhea, and rectal bleeding [2]. It is a lifelong, relapsing disorder that currently affects nearly five million people worldwide and imposes a substantial and growing global health burden [3, 4]. Although its etiology is multifactorial, UC is thought to arise from dysregulated interactions among the gut microbiota, host immunity, and environmental factors in genetically susceptible individuals [5]. The persistent relapsing nature of UC and its multifactorial pathogenesis highlight the need to dissect mechanisms that sustain chronic inflammation, particularly those arising from microbiota–immune crosstalk, to guide the development of more durable therapeutic strategies. Dysbiosis of the gut microbiota is a defining signature of UC. A key feature of UC is the decreased abundance of health-associated commensals, including Faecalibacterium prausnitzii, Christensenellaceae, Collinsella, and Roseburia, which produce short-chain fatty acids and other metabolites essential for epithelial integrity and immune homeostasis [6]. Conversely, UC is characterized by the expansion of mucin-degrading and pro-inflammatory taxa, particularly Proteobacteria such as adherent-invasive Escherichia coli, along with Fusobacterium spp. and Ruminococcus gnavus [7]. These community shifts rewire intestinal metabolic networks and disrupt mucosal immune signaling pathways. Together with mucus layer erosion and epithelial barrier dysfunction, these alterations create a self-sustaining inflammatory loop that fuels UC progression [8]. Given the expansion of Fusobacterium spp. in UC and their recognized pro-inflammatory potential, we focused on F. varium. This species has been implicated in CADASIL-associated systemic inflammation [9] and IBS-related visceral hypersensitivity [10], and is increasingly recognized as clinically relevant in UC. F. varium has been isolated from inflamed colonic mucosa of patients with UC [11–13], and its culture supernatant enriched in n-butyric acid induces cytotoxic effects and UC-like lesions in mice [12]. In addition, F. varium invades colonic epithelial cells and elicits robust pro-inflammatory cytokine responses [14]. Previous studies have implicated F. varium and its derived products in the pathogenesis of UC, however, the immune pathways engaged by F. varium–derived metabolites and the potential contribution of additional bioactive metabolites remain poorly understood. Moreover, geographic variation in gut microbial communities may lead to functional heterogeneity among F. varium strains, potentially influencing their metabolic capacity and pathogenic potential [15]. To address this gap, we sought to determine whether F. varium produces additional metabolites that drive mucosal inflammation and to identify the immune cell populations mediating this response. In this study, we investigated whether UC-enriched F. varium contributes to intestinal inflammatory amplification through metabolite-associated immune signaling. We identified F. varium as a significantly enriched species in fecal samples from patients with UC and isolated a UC-derived F. varium strain that exacerbated DSS-induced colitis under barrier-disrupted conditions. Through NF-κB reporter-based activity-guided screening, we identified succinate as a candidate bioactive metabolite associated with F. varium-related inflammatory activity. Subsequent analyses showed that exogenous succinate exacerbated DSS-induced colitis and was associated with neutrophil recruitment in vivo, while activating SUCNR1/NF-κB–related inflammatory signaling in neutrophil-like cells in vitro. These findings support a neutrophil-centered framework in which F. varium-associated metabolic activity may amplify intestinal inflammation under colitic conditions. Materials and Methods Sample Collection Fecal samples were obtained from 30 healthy controls (HCs) and 37 patients with UC. Detailed clinical characteristics of the study cohort are provided in Table S1 of the Supporting Information. Participants were excluded if they were unwilling to provide informed consent; had used antibiotics within three months prior to enrollment; had used probiotics, immunosuppressants, corticosteroids, or non-steroidal anti-inflammatory drugs within one month prior to enrollment; had acute gastrointestinal infections; were pregnant; had a history of gastrointestinal surgery; or had undergone major changes in dietary habits within the previous six months. Fecal samples were collected according to standardized protocols, immediately transferred into sterile centrifuge tubes, and stored at − 80 °C until further processing. The study protocol was reviewed and approved by the Ethics Committee of the Second Hospital and Clinical Medical School of Lanzhou University (approval number: 2025 A-1185). Animals Germ-free (GF) female C57BL/6J mice (6–8 weeks old) were purchased from the Gempharmatech Co., Ltd (Nanjing, China) and housed under sterile conditions at the Gempharmatech Laboratory Animal Research Center. Specific pathogen-free (SPF) female C57BL/6J mice (6–8 weeks old) were purchased from the Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences. SPF mice were bred and maintained under specific pathogen-free conditions and fed a standard chow diet ad libitum in the animal facility at Lanzhou University. All animal experiments were approved by the Ethics Committee of the Second Hospital and Clinical Medical School of Lanzhou University (approval number: D2020-38). Isolation and Identification of F. varium F. varium strain was isolated from fresh fecal pellets collected from patients with UC. Briefly, fecal samples were homogenized in 10 mL of sterile GAM medium (#05422, Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) supplemented with 10% DL-cysteine hydrochloride. The homogenates were plated on Columbia Blood Agar plates (#PB0123A, Oxoid Ltd., Basingstoke, UK) and incubated at 37 °C under anaerobic conditions (5% H₂, 5% CO₂, and 90% N₂). Individual colonies with distinct morphologies were picked and subcultured on Columbia Blood Agar plates for further purification and identification. The purified strain was identified by 16 S rRNA gene sequencing. Cell Culture HEK-293T, Henle-407, and NCM460 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% FBS, 100 U/mL penicillin, and 100 µg/mL streptomycin at 37 °C in a humidified atmosphere containing 5% CO₂. HL-60 cells were cultured in RPMI-1640 medium supplemented with 10% FBS at 37 °C in a humidified atmosphere containing 5% CO₂. To induce neutrophil-like differentiation, HL-60 cells were treated with 0.8% (v/v) dimethyl sulfoxide (DMSO) and 0.1 µM all-trans retinoic acid (ATRA; #R2625, Sigma-Aldrich) for 3 days. Neutrophil-like differentiation was confirmed by increased CD11b expression assessed by flow cytometry. Metagenomic Shotgun Sequencing Genomic DNA was extracted from human fecal samples using the E.Z.N.A.® Soil DNA Kit (#M5635-02, Omega Bio-Tek, USA). Sequencing libraries were prepared from DNA fragments with an average insert size of approximately 500 bp by end repair, adapter ligation, size selection, PCR amplification, purification, and sequenced on the Illumina NovaSeq 6000 platform (Sangon Biotech, Shanghai, China). Raw reads were quality-filtered and trimmed using Fastp (v0.36) to remove adapters and low-quality bases, and host-derived reads were subsequently removed using KneadData (v0.12.0), resulting in high-quality, decontaminated reads for downstream analyses. Taxonomic profiling was performed using MetaPhlAn4 (v4.0.6) and Kraken2 (v2.1.3), with genome-based reference taxonomy derived from the Genome Taxonomy Database (GTDB). Gene family and pathway abundances were quantified using HUMAnN3 (v3.7), and functional annotation was mapped to the KEGG, eggNOG, GO, EC, Pfam, and CAZy databases. Antibiotic resistance genes and virulence factors were additionally annotated against the CARD, SARG, and VFDB databases. DSS-induced Colitis Model To evaluate the effects of F. varium and a small-molecule metabolite in DSS-induced colitis, germ-free (GF) and specific pathogen-free (SPF) mouse models were employed. For the GF mouse model, mice were orally gavaged with F. varium (1 × 10⁹ CFU) resuspended in culture supernatant or vehicle control every other day for a total of eight administrations. Colitis was subsequently induced by administering 2.0% (w/v) dextran sulfate sodium (DSS; #60316ES60, Yeasen) ad libitum in the drinking water for 7 consecutive days. For the SPF mouse model, mice were provided with 3% (w/v) succinic acid (pH 6.5; #S9512, Millipore Sigma) or regular drinking water as a control, starting 3 days prior to DSS exposure and continuing throughout the experimental period. DSS-induced colitis was induced by administering 2.5% (w/v) DSS ad libitum in the drinking water for 7 consecutive days. For mice receiving succinate treatment, DSS was dissolved directly in succinate–supplemented drinking water. To inhibit CXCR2 signaling in vivo, mice were administered the selective CXCR2 inhibitor SB225002 (#SML0716, Sigma-Aldrich, USA) intraperitoneally (i.p.) at a dose of 2 mg/kg once daily during the DSS treatment period. Mice receiving vehicle served as controls. Mice were monitored daily for disease progression. Disease activity was evaluated using a disease activity index (DAI), which included body weight loss, stool consistency, and fecal blood. Each parameter was scored independently from 0 to 4, and the final DAI score was calculated as the average of the three scores: DAI = (body weight loss score + stool consistency score + fecal blood score)/3. Body weight loss was scored as follows: 0, 0–1%; 1, 1–5%; 2, 5–10%; 3, 10–15%; and 4, > 15%. Stool consistency was scored as follows: 0, normal stool; 1, soft but non-sticky stool; 2, soft and sticky stool; 3, diarrhea; and 4, severe diarrhea. Fecal blood was scored as follows: 0, negative fecal occult blood; 1, weakly positive fecal occult blood; 2, positive fecal occult blood; 3, moderately positive fecal occult blood; and 4, hematochezia or strongly positive fecal occult blood. DNA Extraction and Real-time Quantitative PCR of F. varium Fecal DNA was extracted from approximately 100 mg of mouse feces using a fecal genomic DNA extraction kit (#DP712, TIANGEN, China) according to manufacturer’s instructions. DNA was eluted in 50 µL nuclease-free water, and 2 µL DNA was used as the qPCR template. Absolute quantification of F. varium was performed using a standard curve generated from a plasmid containing an 84-bp F. varium-specific fragment. The plasmid concentration was 9.3 ng/µL, corresponding to 1.01 × 10¹¹ copies/µL. Ten-fold serial dilutions from 10⁸ to 10³ copies/µL were prepared, and 1 µL of each dilution was used for standard curve construction. F. varium abundance was calculated from the standard curve and normalized to fecal weight. Data were expressed as log10 F. varium copies per mg feces. Succinate Levels in Mouse Feces and Bacterial Culture Supernatants Succinate levels in mouse fecal samples and bacterial culture supernatants were measured using a succinate assay kit (#S0529S, Beyotime, China) according to the manufacturer’s instructions. Briefly, fecal samples were homogenized and centrifuged to collect the supernatant, while bacterial cultures were centrifuged to obtain cell-free culture supernatants. The prepared samples were then subjected to succinate quantification using the assay reagents provided in the kit. Absorbance was measured according to the kit protocol, and succinate concentrations were calculated based on the standard curve. Data were normalized to fecal weight for fecal samples or expressed as concentration in bacterial culture supernatants. Isolation of Lamina Propria Immune Cells Lamina propria (LP) immune cells were isolated using a modified protocol based on previously described methods [16]. Briefly, mice were anesthetized with isoflurane, and colorectal tissues were excised and opened longitudinally. The tissues were placed into 50 mL conical tubes, washed three times with 20 mL phosphate-buffered saline (PBS) by gentle shaking, and cut into 1 cm segments. Tissue segments were incubated in 20 mL Buffer A at 37 °C with shaking (220 rpm) for 15 min. The supernatant was filtered through a 100 μm strainer, and the tissues were recovered for a second 15 min incubation in fresh Buffer A under the same conditions. Buffer A consisted of 5 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT), 14 mM 4-(2-hydroxyethyl)−1-piperazineethanesulfonic acid (HEPES; from a 10× stock containing 100 mM HEPES and 250 mM sodium bicarbonate (NaHCO₃)), 5% FBS, and 1× Hanks’ Balanced Salt Solution (HBSS), which was prepared by diluting a 10× stock solution containing 0.4% (w/v) potassium chloride (KCl), 0.06% (w/v) potassium dihydrogen phosphate (KH₂PO₄), 8% (w/v) sodium chloride (NaCl), 0.0475% (w/v) disodium hydrogen phosphate (Na₂HPO₄), and 1% (w/v) D-glucose. After removing the supernatant, tissue fragments were washed three times with PBS and transferred to 20 mL Buffer B digestion medium (0.04% (w/v) collagenase I, 1 mM MgCl₂, 1 mM CaCl₂, and 10% FBS in RPMI-1640 medium) in new tubes for 20 min at 37℃ on a shaker (220 rpm). The resulting cell suspensions were filtered through 100 μm strainers. Residual tissues were homogenized with a sterile pestle, resuspended in PBS containing 2% FBS, filtered into the same tube, and subjected to a second digestion under identical conditions. Cell suspensions were centrifuged at 800 × g for 10 min at 4 °C. Supernatants were discarded, and the cell pellets were resuspended in PBS containing 2% FBS for flow cytometry analysis. Flow Cytometry LP immune cells were isolated from the colorectal tissues as described above. For surface staining, the following antibodies were used: anti-CD45 (#103114, Biolegend, USA), anti-CD3 (#100236, Biolegend, USA), anti-CD19 (#152410, Biolegend, USA), anti-F4/80 (#565787, BD Biosciences, USA), anti-CD11b (#MA1-10081, ThermoFisher, USA), anti-Ly6G (#127639, Biolegend, USA), anti-CD11c (#117322, Biolegend, USA), anti-MHCII (#107624, Biolegend, USA). Live/dead cell discrimination was performed using Ghost Dye™ Red 780 Fixable Viability dye (#13–0865, Tonbo Biosciences, USA). After a 30-minute incubation at 4 ℃ in the dark, cells were washed, resuspended in PBS, and immediately analyzed by flow cytometry. For intracellular staining, cells were stimulated for 4 h at 37 °C with Leukocyte Activation Cocktail containing BD GolgiPlug™ (#550583, BD Biosciences, USA). After stimulation, cells were stained with live/dead dye, washed, and fixed/permeabilized in 100 µL fixation/permeabilization buffer (#554722, BD Biosciences, USA). Cells were then washed twice with PBS and stained for intracellular cytokines using the following antibodies: anti-CD45 (#103114, Biolegend, USA), anti-CD3 (#100237, Biolegend, USA), anti-CD4 (#100412, Biolegend, USA), anti-IFNγ (#505808, Biolegend, USA), anti-FOXP3 (#126410, Biolegend), anti-IL-17 (#506944, Biolegend, USA), anti-IL-4 (#11–7042-82, Biolegend, USA). After intracellular staining, cells were washed, resuspended in PBS, and analyzed by flow cytometry. Histology The collected intestinal tissues were cut into small pieces and fixed in 4% paraformaldehyde. The fixed tissues were dehydrated, embedded in paraffin, sectioned at 4 μm thickness, and stained with hematoxylin and eosin (H&E) following standard procedures. Histological damage was scored based on inflammatory severity (0–3: none, slight, moderate, and severe), depth of injury (0–4: none, mucosal, submucosal, muscularis propria, and serosa), extent of crypt damage (0–4: none, basal 1/3 damaged, basal 2/3 damaged, only surface epithelium intact, and entire crypt and epithelium lost), and lesion extent (0–4: none, 1%–25%, 26%–50%, 51%–75%, and 76%–100%) [17]. NF-κB Dual-luciferase Reporter Assay The NF-κB luciferase reporter construct was derived from the commercially available pGL4.32[luc2P/NF-κB-RE/Hygro] vector (#E8491, Promega) by introducing transposase recognition sequences to facilitate stable genomic integration. To establish a stable internal control, the Renilla luciferase (Rluc) gene was cloned into the lentiviral backbone lentiCRISPRv2 under the control of the HSV-TK promoter. The resulting lentiviral plasmid was co-transfected with the packaging plasmids psPAX2 and pMD2.G into HEK-293T cells to produce lentiviral particles. Viral supernatants were collected 48 h post-transfection and used to transduce HEK-293T cells. Transduced cells were selected with puromycin (3 µg/mL), and single-cell clones were isolated by limiting dilution. Clones exhibiting stable Renilla luciferase activity were expanded for subsequent experiments. The NF-κB luciferase reporter plasmid was then co-transfected with a transposase-expressing plasmid into the Renilla luciferase–expressing monoclonal cells, followed by selection with Hygromycin B (200 µg/mL), to generate double-stable cell lines expressing both Renilla and firefly luciferase. In dual-luciferase reporter assays, Renilla luciferase activity was used as an internal control to normalize for variations in cell number, cell viability, and assay efficiency. For luciferase assays, cells were seeded in 48-well plates and treated with varying concentrations of bacterial culture supernatants and small-molecule metabolites. Firefly and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (#E1960, Promega, USA) and quantified with a multimode microplate reader (Thermo Fisher Scientific, USA). Endotoxin Assessment Endotoxin levels in F. varium culture supernatants were assessed using the ToxinSensor™ Chromogenic LAL Endotoxin Assay Kit (#L00350, GenScript, USA) according to the manufacturer’s instructions. Samples and endotoxin standards were prepared under endotoxin-free conditions, and endotoxin concentrations were calculated based on the standard curve and expressed as EU/mL. To further evaluate the contribution of LPS, F. varium culture supernatants were pretreated with polymyxin B at 20 µg/mL for 30 min before stimulation of HEK-293T NF-κB reporter cells, followed by NF-κB dual-luciferase reporter assays. Construction of Stable Cell Lines Stable RELA-overexpressing cell lines were generated using a lentiviral expression system. For Henle-407 cells, the human RELA coding sequence was cloned into the lentiviral vector Lenti-EF1α−3×Flag-GFP-Puro. Lentiviral particles were produced in HEK-293T cells, and viral supernatants were collected 48 h post-transfection and used to infect Henle-407 cells in the presence of polybrene (8 µg/mL). Transduced cells were selected with puromycin (3 µg/mL) to establish a stable RELA-overexpressing Henle-407 cell line. A stable RELA-overexpressing HEK-293T cell line was generated in parallel by lentiviral transduction using a construct encoding human RELA. Stable cells were selected with puromycin and used for subsequent experiments. Immunofluorescence Cells were cultured on glass-bottom dishes and fixed with 4% paraformaldehyde for 15 min. After washing with PBS, the cells were permeabilized with 0.1% Triton X-100 for 10 min and blocked with QuickBlock™ Blocking Buffer (#P0260, Beyotime) for 1 h. Cells were incubated with anti-Flag antibody (1:200 dilution, #F1804, Sigma) overnight at 4 °C. DyLight 594-conjugated secondary antibody (1:300 dilution, #ab150116, Abcam) was applied for 1 h at room temperature in the dark. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; #P0131, Beyotime, China). Images were acquired using a laser scanning confocal microscope (Zeiss LSM880, Carl Zeiss, Germany). Real-time qPCR Total RNA was extracted from cells or mouse intestinal tissues using the FastPure® Cell/Tissue Total RNA Isolation Kit V2 (#RC112, Vazyme) and reverse-transcribed into cDNA using the iScript™ cDNA Synthesis Kit (#1708891, Bio-Rad) according to the manufacturer’s instructions. RT-qPCR was performed on a Bio-Rad CFX96 Touch Real-Time PCR Detection System using the following program: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Relative mRNA expression levels were calculated using the 2⁻ΔΔCt method, with mouse Gapdh or human GAPDH serving as the internal control. The sequences of the forward and reverse primers used for qRT-PCR are provided in Table S1. Protein Extraction and Western Blot Analysis Cellular proteins were lysed on ice for 30 min using Cell Lysis Buffer for Western and IP (#P0013, Beyotime, China) supplemented with the protease inhibitor phenylmethylsulfonyl fluoride (PMSF; #ST506, Beyotime, China). Lysates were then centrifuged at 15,000 × g for 15 min at 4 °C, and the supernatants were collected. Equal amounts of protein were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes (#IPVH00010, Millipore, USA). The membranes were blocked with 5% non-fat milk in Tris-buffered saline with Tween 20 (TBST) for 1 h at room temperature and incubated with primary antibodies overnight at 4 °C. The primary antibodies used included IκBα (1:1000, #4814, Cell Signaling Technology, USA), β-Actin (1:1000, #3700S, Cell Signaling Technology, USA), NF-κB p65 (1:1000, #6956S, Cell Signaling Technology, USA), α-Tubulin (1:1000, #66031-1-Ig, Proteintech, USA), and Lamin B1 (1:1000, #66095-1-Ig, Proteintech, USA). After washing with TBST, the membranes were incubated with HRP-conjugated secondary antibodies (1:10,000, #SA00001-1, Proteintech, USA) for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence (ECL) kit (#P10200, NCM Biotech, China), and densitometric analysis was performed using ImageJ software. Nuclear and cytoplasmic proteins were extracted using the Nuclear and Cytoplasmic Protein Extraction Kit (#P0028, Beyotime, China) according to the manufacturer’s instructions. Briefly, cells were harvested and resuspended in Cytoplasmic Extraction Reagent A, followed by incubation on ice for 15 min. Cytoplasmic Extraction Reagent B was then added, and the mixture was vortexed and centrifuged at 15,000 × g for 5 min at 4 °C to collect the cytoplasmic fraction (supernatant). The nuclear pellet was resuspended in Nuclear Extraction Reagent and incubated on ice for 30 min with intermittent vortexing. After centrifugation, the supernatant containing nuclear proteins was collected. Ultrafiltration-based Fractionation of Bacterial Supernatants Standardized 96-h cultures of F. varium or control GAM medium were centrifuged and filtered through 0.2 μm membranes to obtain supernatants. To fractionate the bacterial supernatants according to molecular weight, the filtrates were applied to ultrafiltration tubes with molecular weight cutoffs (MWCO) of 3, 10, 30, 50, and 100 kDa (Millipore, Billerica, MA, USA). Equal volumes of sterile GAM medium were processed in parallel as negative controls. Ultrafiltration was performed by centrifugation at 3,500 × g for 30 min at 4 °C. The filtrates collected from the lower chambers represented fractions of < 3 kDa, < 10 kDa, < 30 kDa, < 50 kDa, and < 100 kDa, respectively. To remove residual solutes, the retentates in the upper chambers were washed three times with sterile ultrapure water, and the retained solutions were collected as the > 3 kDa, > 10 kDa, > 30 kDa, > 50 kDa, and > 100 kDa fractions, respectively. The resulting fractions were applied to dual-luciferase reporter cells to assess their ability to activate the NF-κB signaling pathway and thereby determine the molecular weight range of the active components. Physicochemical Characterization of Bacterial Supernatants After determining the molecular weight range of the active components in the bacterial supernatants by ultrafiltration, the supernatants were subjected to heat treatment (100 °C for 30 min) or enzymatic digestion with deoxyribonuclease I (DNase I), ribonuclease A (RNase A), or proteinase K (10 µg/mL, 37 °C for 60 min). The enzymes were subsequently removed by ultrafiltration using appropriate MWCO membranes. The treated supernatants were filtered through 0.22 μm membranes and analyzed using a dual-luciferase reporter assay to evaluate their thermal and biochemical stability. Sequential Solvent Extraction of Bacterial Metabolites Bacterial cultures were centrifuged at 12,000 rpm for 10 min at 4 °C and filtered through 0.22 μm membranes to obtain clarified supernatants. The supernatants were sequentially extracted twice with equal volumes (1:1, v/v) of petroleum ether, ethyl acetate, and n-butanol. For each solvent, the combined organic phases from the two extractions were collected and lyophilized. The dried crude extracts were weighed and reconstituted in DMSO or water, depending on solubility, to a final concentration of 20 mg/mL for use as stock solutions. Chemical Isolation Purification of metabolites was performed using a Gilson GX-281 semi-preparative high-performance liquid chromatography (HPLC) system (Gilson, Inc., Middleton, WI, USA) equipped with a Phenomenex Luna C18 column (250 × 21.2 mm, 5 μm; Phenomenex, USA). The column was maintained at room temperature, and UV absorbance was monitored at 210 nm. The mobile phases consisted of water (A) and acetonitrile (B), delivered at a flow rate of 9 mL/min. The gradient elution program was as follows: 10% B (0–8 min), 10–70% B (8–38 min), 70% B (38–43 min), 70–100% B (43–53 min), and 100% B (53–58 min). The n-butanol extract was fractionated into 10 fractions under these conditions. The collected fractions were analyzed by high-performance liquid chromatography–mass spectrometry (HPLC–MS) using a Shimadzu LC-20 A system coupled with an LCMS-2020 single quadrupole mass spectrometer (Shimadzu Corporation, Kyoto, Japan). Separation was achieved on a Phenomenex Gemini NX-C18 column (150 × 4.6 mm, 5 μm; Phenomenex, USA) maintained at 40 °C. The mobile phases consisted of H₂O with 0.0375% trifluoroacetic acid (TFA) (v/v) (A) and acetonitrile with 0.0375% TFA (v/v) (B), delivered at a flow rate of 1 mL/min. The gradient elution program was as follows: 5% B (0–1 min), 5–9% B (1–4.5 min), 9–14% B (4.5–7.5 min), 14% B (7.5–9 min), 14–40% B (9–25 min), 40–100% B (25–25.01 min), and held at 100% B until 26 min. UV detection was performed at 210 nm. Ten fractions (F1–F10) were collected from the semi-preparative HPLC, dried under reduced pressure using a rotary evaporator, and weighed. The dried residues were reconstituted and subjected to a dual-luciferase reporter assay in HEK-293T cells to evaluate their NF-κB–activating activity. Primary active fractions were identified based on luciferase activity measurements. The active fractions (F1 and F9) were further purified. Fraction F1 was subjected to gel filtration chromatography on a Sephadex LH-20 column using 50% methanol in water as the eluent. The eluates were collected, concentrated to dryness under reduced pressure using a rotary evaporator, and designated as F1S1 to F1S5 for subsequent HPLC–MS analysis. Fraction F9 was further fractionated by semi-preparative HPLC using a Gilson GX-281 system equipped with a Phenomenex Luna C18 column (250 × 21.2 mm, 5 μm). The mobile phases were water (A) and acetonitrile (B). The gradient elution program was 50–100% B (0–20 min) followed by 100% B (20–35 min) at a flow rate of 9 mL/min. The column was maintained at room temperature, and UV absorbance was monitored at 210 nm. Fractions were collected every 7 min, concentrated to dryness under reduced pressure, and designated as F9S1 to F9S5. These subfractions were analyzed by HPLC–MS and assessed for NF-κB activity using the HEK-293T dual-luciferase reporter assay. Functional Screening and Chemical Identification of Active Compounds Metabolic profiling was performed using an AB TripleTOF 4600 high-resolution mass spectrometer (AB SCIEX, USA) equipped with an electrospray ionization (ESI) source and coupled to a Shimadzu LC-20ADXR HPLC system (Shimadzu, Japan). Chromatographic separation was achieved on a Shim-pack GIST C18-AQ HP column (2.1 × 100 mm, 3 μm; Shimadzu) maintained at 35 °C. The mobile phases consisted of 0.1% formic acid in methanol (A) and 0.1% formic acid in water (B), delivered at a flow rate of 0.2 mL/min. The gradient program was as follows: 5% A (0–3 min), 25% A (3–5 min), 80% A (5–24 min), and 100% A (24–30 min). Samples were analyzed in both positive and negative ionization modes. Background signal from the n-butanol extract of GAM medium was subtracted to minimize interference from medium-derived compounds (Table S2, Supporting Information). The MS scan range was m/z 100–1300, and the MS/MS scan range was m/z 50–1300. Mass calibration was performed during data acquisition. In MS/MS mode, nitrogen was used as the collision gas, and collision energies ranging from 20 to 60 eV were applied in an alternating voltage mode to obtain metabolite fragmentation spectra. Raw data were acquired using Analyst 1.7 software (AB SCIEX, USA) and processed with PeakView 2.2 (AB SCIEX, Canada). Compound identification in both positive and negative ESI modes was performed using the Natural Products HR-MS/MS Spectral Library 1.0.1 (AB SCIEX, USA), and identification parameters included a mass error < ± 5 ppm, isotope ratio difference < 10%, and a Formula Finder score > 0.4. Additional annotation was conducted using the Formula Finder function. Metabolite abundance was estimated based on relative LC–MS peak areas after background subtraction using the n-butanol extract prepared from uninoculated GAM medium as the blank control. All samples were processed using the same starting volume, extraction solvent volume, extraction procedure, and LC–MS injection volume; therefore, background-corrected relative peak areas were used directly for downstream comparative analyses without further normalization. For candidate compounds confirmed with authentic standards, metabolite identities were validated by matching both retention time and MS/MS fragmentation patterns to those of standards analyzed under the same LC–MS conditions. For NF-κB activity screening, LC–MS features were retained only if the Formula Finder Score, Combined Score, and Isotope Score each exceeded 70%. For LC–MS/MS features, the corresponding thresholds were set at 60% to account for the additional structural information provided by fragmentation spectra. Further quality control (QC) criteria included mass accuracy, isotopic pattern fidelity, signal-to-noise ratio, and retention time reproducibility. Detailed annotation metrics and compound parameters are provided in Table S3 and S4 (Supporting Information). Candidate compounds associated with gut microbial metabolism and meeting all QC criteria were subsequently screened for NF-κB–activating activity. For active compounds, metabolite identities were confirmed by retention time matching with authentic standards and validated by concordant MS/MS fragment ions. Bacterial Whole-genome Sequencing Genomic DNA was extracted using the E.Z.N.A.® Bacterial DNA Kit (#D3350, Omega Bio-Tek, USA). Whole-genome sequencing was performed using a hybrid approach combining Pacific Biosciences (PacBio) long-read sequencing and Illumina NovaSeq 6000 short-read sequencing (Sangon Biotech, Shanghai, China). Illumina raw reads were quality-filtered to remove adapters and low-quality bases, whereas PacBio data were processed to obtain high-quality subreads. Hybrid de novo assembly was conducted using Unicycler, followed by polishing with both long and short reads. Gene prediction and annotation were performed using Prokka and the NCBI Prokaryotic Genome Annotation Pipeline (PGAP), with functional annotation referenced against KEGG, COG, GO, CAZy, VFDB, CARD, and PHI-base databases. Genome completeness and quality were assessed with CheckM and BUSCO. Comparative Genomic Analysis of F. varium Comparative genomic analysis was performed to determine the phylogenetic relationships and orthologous gene conservation between the human-derived F. varium strain and other publicly available F. varium genomes. Whole-genome sequences of six F. varium strains (ATCC 27725, Fv113-g1, KR001_HAM_0062, NCTC10560, SB064, and SB065) were downloaded from the NCBI RefSeq database (https://www.ncbi.nlm.nih.gov/datasets/genome/). Comparative genome visualization, including GC content, GC skew, coding sequences (CDSs), and orthologous gene clusters, was performed using Proksee (https://proksee.ca/). Core genome-based phylogenetic analysis was performed using the VBCG pipeline (v1.3) with default parameters [18]. Whole-genome assemblies of all F. varium isolates were analyzed locally using the VBCG workflow. Single-copy core genes were automatically identified, aligned at the protein level, and concatenated to construct a maximum-likelihood phylogenetic tree with 1,000 bootstrap replicates. The resulting tree file in Newick format was downloaded and visualized using MEGA11. Bootstrap support values provided by VBCG were retained to indicate branch reliability. Transcriptome Sequencing Total RNA from cells was extracted using AG RNAex Pro Reagent (#AG21101, Accurate Biotechnology, Hunan, China) and assessed for quality using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Transcriptome libraries were prepared using the VAHTS Universal V5 RNA-seq Library Prep Kit according to the manufacturer’s instructions. RNA sequencing and subsequent bioinformatic analyses were performed by OE Biotech Co., Ltd. (Shanghai, China). Briefly, mRNA was enriched using oligo (dT) magnetic beads and fragmented into short fragments, which were then used as templates for cDNA synthesis. Following end repair, adapter ligation, and PCR amplification, the libraries were sequenced on an Illumina NovaSeq 6000 platform to generate paired-end 150 bp reads. Raw sequencing reads were processed to remove adapters and low-quality bases using fastp [19], Clean reads were aligned to the reference genome using HISAT2. Gene expression levels were quantified as Fragments per kilobase of transcript per million mapped reads (FPKM) using featureCounts [20]. Gene-level read counts were obtained with HTSeq-count [21], and differentially expressed genes (DEGs) were identified with DESeq2 [22]. Statistical Analysis GraphPad Prism v10.0 (GraphPad Software, San Diego, CA, USA) was used for statistical analyses. Differences between two groups were assessed using unpaired, two-tailed Student’s t-tests. Comparisons among three or more groups were performed using one-way or two-way analysis of variance (ANOVA), followed by Tukey’s multiple-comparisons test, Dunnett’s multiple-comparisons test or Šidák multiple-comparisons test, as appropriate. Differences were considered statistically significant for P values < 0.05. Unless otherwise stated, data are presented as the mean ± SEM. Metagenomic data visualization and analyses were performed in R v4.3.3 (R Foundation for Statistical Computing, Vienna, Austria). Results Increased Abundance of F. varium in Patients with UC A total of 37 patients with UC and 30 healthy controls (HCs) were included in the study, with similar sex ratios and age distributions (Fig. 1SA and B). To characterize the gut microbial signatures in UC, we conducted metagenomic sequencing of fecal samples from patients with UC and HCs (Fig. 1A). The gut microbiome of patients with UC exhibited significantly reduced richness and diversity compared to HCs, as measured by Chao1 and Shannon indices, respectively (Fig. 1B). In addition, beta diversity analysis based on Bray–Curtis dissimilarity revealed a significant difference in microbial community structure between patients with UC and HCs (Fig. 1C). These data indicate reduced microbial diversity and a global shift in community structure in UC. Interrogating the composition of this dysbiotic microbiota, genus-level profiling revealed a shift, characterized by a loss of commensal bacteria with known beneficial functions (e.g., Agathobacter, Roseburia, Lachnospira) and an enrichment of genera frequently associated with pathogenesis (e.g., Citrobacter, Eggerthella, Fusobacterium) in patients with UC (Fig. 1D). To pinpoint specific bacterial species that drive this association, we performed differential abundance analysis by the Wilcoxon rank-sum test. This revealed F. varium as the most significantly enriched species in patients with UC, displaying the largest effect size among UC-associated species (Fig. 1E). Consistent with our findings, prior studies have shown that F. varium can colonize the intestinal niche in patients with UC [11]. The prominent association of F. varium with UC was further supported by linear discriminant analysis effect size analysis (LEfSe, LDA score ≥ 3), which identified both the Fusobacterium genus and the F. varium species as key discriminating taxa in UC (Fig. 1F and G). Given its consistent enrichment within the dysbiotic UC gut microbiome and the established link between Fusobacterium species and mucosal inflammation, we hypothesized that F. varium may contribute to UC-associated intestinal inflammation. We therefore focused our subsequent investigations on evaluating the potential pro-inflammatory role of F. varium in colitis. F. varium is enriched in patients with UC. (A) Fecal samples were collected from patients with ulcerative colitis (UC; n = 37) and healthy controls (HCs; n = 30) and subjected to whole-metagenome shotgun sequencing to characterize gut microbial community composition and identify UC-associated microbial alterations. (B) Alpha diversity of the gut microbiota assessed by Chao1 (richness) and Shannon (diversity) indices. Violin plots depict the distribution of values for each group, with embedded boxplots indicating the median and interquartile range (IQR). Statistical significance between groups was determined using a two-tailed Wilcoxon rank-sum test. (C) Principal coordinates analysis (PCoA) based on Bray–Curtis dissimilarity showing differences in microbial community structure between the UC and HC groups. Each dot represents a single sample; red indicates patients with UC, and blue indicates HCs. Statistical significance was assessed using PERMANOVA based on Bray–Curtis dissimilarity, which showed a significant difference between the two groups (R² = 0.046, P = 0.001). (D) Relative abundance of the top 50 genera in patients with UC and HCs, shown as stacked bar plots. (E) Differentially abundant species between patients with UC and HCs identified using the Wilcoxon rank-sum test. The left panel shows the mean relative abundance of significantly different species in each group, and the right panel depicts the between-group differences with 95% confidence intervals. (F) Linear discriminant analysis effect size (LEfSe) taxonomic cladogram highlighting taxa that differ significantly between patients with UC and HCs. Taxa enriched in HCs are shown in green, whereas taxa enriched in patients with UC are shown in red. The size of each node corresponds to the relative abundance of the taxon, and taxonomic levels from phylum to species are displayed from the center outward. (G) LDA scores from LEfSe analysis for taxa significantly differentiating patients with UC and HCs. Significance was assessed using the Kruskal–Wallis and Wilcoxon tests (α = 0.01), with an LDA score cutoff of 3. Bars are colored according to enrichment in HCs (green) or patients with UC (red) F. varium Exacerbates DSS-induced Colitis in Mice To investigate the pro-inflammatory potential of F. varium in colitis, we isolated a strain from the feces of patients with UC and performed whole-genome sequencing. Phylogenetic analysis revealed that this isolate is closely related to the pathogenic strain F. varium Fv113-g1 (Fig. 2A and S2A), suggesting it might share pro-inflammatory properties [23]. This prompted us to hypothesize that this strain could contribute to intestinal inflammation. To test this, germ-free (GF) mice were colonized with the isolated F. varium suspension or vehicle control every other day for two weeks. Successful intestinal colonization was supported by fecal qPCR quantification of F. varium abundance and agarose gel verification of the expected 84-bp qPCR product (Fig. S2B and C). Subsequently, mice were challenged with 2% dextran sulfate sodium (DSS) in drinking water for seven days to induce colitis, or were left untreated (Fig. 2B). Upon DSS challenge, F. varium–colonized mice developed significantly more severe DSS-induced colitis, as evidenced by greater weight loss, higher disease activity scores, and increased colon shortening and histopathological damage (Fig. 2C–H). In contrast, mice colonized with F. varium alone showed no signs of colonic inflammation. These data indicate that F. varium does not initiate inflammation on its own but rather synergizes with DSS-induced epithelial barrier injury to amplify colitis. F. varium exacerbates colitis in mice. (A) Circular genome map of the F. varium strain isolated and sequenced in this study, shown together with comparative genomic alignments to other publicly available F. varium strains. From the inner to outer rings: GC content (black), GC skew (positive, green; negative, magenta), coding sequences (CDSs, pink), tRNA (light green), pseudogenes (blue), rRNA (orange), and noncoding RNA (cyan). The six outer rings show orthologous gene alignments with F. varium strains ATCC 27,725, Fv113-g1, KR001_HAM_0062, NCTC10560, SB064, and SB065, each indicated by a distinct color; gaps denote regions lacking significant sequence similarity. (B) Experimental schematic of the germ-free (GF) mouse model of DSS-induced colitis. GF mice were gavaged with vehicle control or F. varium every other day for 2 weeks, followed by administration of 2% DSS in drinking water for 7 days. (C–H) F. varium colonization aggravated DSS-induced colitis, as evidenced by greater body weight loss (C), increased disease activity index (DAI) scores (D), shortened colon length with representative gross images (E and F), and enhanced histopathological damage shown by representative H&E-stained colon sections and corresponding histopathology scores (G and H). Scale bars, 400 μm–50 μm. (I and J) Flow cytometric analysis showing increased proportions of neutrophils (CD11b⁺Ly6G⁺) in the colonic lamina propria (cLP) of F. varium–colonized mice. n = 6 mice per group. Data are presented as mean ± SEM. Statistical analyses were performed using two-way ANOVA followed by Šidák’s multiple comparisons test (C and D) and one-way ANOVA followed by Tukey’s multiple comparisons test (F, H, and J). **P < 0.01; ***P < 0.001 We reasoned that this synergy likely involved altered immune responses. To test this, we examined the changes in cytokines and immune cells in the colonic tissues of mice. The transcription levels of the chemokines Cxcl1 and Cxcl2 were significantly upregulated in the colons of mice colonized with F. varium following DSS treatment compared with DSS-treated control mice (Fig. S2P). Consistent with this, profiling of colonic immune cells revealed selective recruitment of neutrophils and Th17 cells in DSS-treated mice colonized with F. varium (Fig. 2I and J, Fig. S2L and M). In contrast, the proportions of dendritic cells (DCs), macrophages, Th1, Th2, and Treg cells (Fig. S2D–K, S2N–O) remained unchanged between the two groups, suggesting a preferential association with neutrophil- and Th17-related inflammatory responses in the DSS model. Collectively, these data suggest that human gut–derived F. varium can amplify DSS-induced intestinal inflammation in association with enhanced recruitment of neutrophils and Th17 cells. F. varium Secreted Factors Trigger Pro-inflammatory Responses via NF-κB Activation Having observed that F. varium exacerbated DSS-induced intestinal inflammation in vivo, we next asked whether its culture supernatant could activate pro-inflammatory signaling in host cells. To address this, we established an NF-κB reporter cell line in HEK-293T cells and found that F. varium culture supernatant induced NF-κB activity in a dose-dependent manner (Fig. 3A). To assess whether endotoxin contamination contributed to this activity, endotoxin levels were measured using a chromogenic LAL assay, and polymyxin B neutralization experiments were performed. Polymyxin B reduced detectable endotoxin levels in the culture supernatant (Fig. S3A), but polymyxin B-treated F. varium culture supernatant retained NF-κB–activating activity comparable to untreated supernatant (Fig. S3B), suggesting that the observed reporter activation was unlikely to be primarily attributable to LPS contamination. To further characterize the host cellular responses triggered by F. varium–derived factors, we performed transcriptomic profiling of HEK-293T cells treated with the bacterial culture supernatant. Global transcriptomic profiling revealed extensive remodeling of host gene expression programs. Sequencing depth and detected gene numbers were comparable across samples (Fig. S3C), and principal component analysis showed clear segregation between F. varium supernatant–treated cells and control cells (Fig. S3D), indicating substantial molecular divergence between the two groups. Differential expression heatmaps further highlighted widespread transcriptional shifts induced by F. varium supernatant (Fig. S3E). Consistent with these findings, volcano plot analysis revealed more than 2,000 differentially expressed genes, including prominent upregulation of key inflammatory mediators (Fig. 3B). Notably, KEGG analysis identified the NF-κB signaling pathway as one of the most enriched pathways (Fig. 3C). Together, these findings suggested that F. varium culture supernatant can activate NF-κB-associated inflammatory signaling. F. varium culture supernatant activates NF-κB signaling. (A) Dual-luciferase reporter assay showing dose-dependent activation of NF-κB signaling in HEK-293T reporter cells following 5 h treatment with F. varium supernatant, with GAM medium as the vehicle control. (B) Volcano plot of differentially expressed genes in HEK-293T cells treated with F. varium culture supernatant compared with GAM medium control. Significantly upregulated genes are shown in red, downregulated genes in blue, and non-significant genes in gray. (C) KEGG pathway enrichment analysis of genes upregulated (log₂ fold change ≥ 1, q < 0.05) following treatment with F. varium supernatant. (D) Relative mRNA expression levels of CXCL1, CXCL3, CXCL8, and TNF in HEK-293T cells after stimulation with F. varium supernatant, as determined by RT-qPCR. (E) Western blot analysis of IκBα protein levels in HEK-293T cells treated with increasing concentrations (2.5%, 5%, 10%, and 20%) of F. varium supernatant for 30 min. Band intensities were quantified using ImageJ and normalized to β-actin. (F) HEK-293T stable cell lines overexpressing Flag-tagged RelA were treated with 20% F. varium supernatant for 30 min, followed by western blot analysis of NF-κB p65 in nuclear and cytoplasmic fractions. Band intensities were quantified using ImageJ and normalized to Lamin B1 (nuclear) and α-tubulin (cytoplasmic). (G) Representative immunofluorescence images showing p65 nuclear translocation in Henle-407 cells stably overexpressing Flag-tagged RelA after 30 min treatment with F. varium supernatant or GAM medium control. Nuclear localization of p65 was quantified using ImageJ. Data with error bars represent the mean ± SEM of at least three independent biological replicates. Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple comparisons test (A and E) and unpaired two-tailed t-tests (D, F, and G). *P < 0.05; **P < 0.01; ****P < 0.0001; ns, not significant We validated the RNA-seq findings by measuring the expression of key NF-κB target genes (CXCL1, CXCL3, CXCL8, TNF) using RT-qPCR, all of which were robustly induced (Fig. 3D). We next probed the core biochemical events of pathway activation. Western blot analysis demonstrated a dose-dependent degradation of the NF-κB inhibitor IκBα (Fig. 3E), the definitive upstream trigger for NF-κB nuclear translocation. Consistently, cellular fractionation assays revealed a substantial accumulation of the p65 subunit in the nucleus, with no appreciable change in its cytoplasmic levels (Fig. 3F). Finally, immunofluorescence microscopy visually confirmed the enhanced nuclear localization of p65, which was quantified by a significant increase in nuclear fluorescence intensity (Fig. 3G). Collectively, these data indicate that F. varium secreted factors can activate the canonical NF-κB signaling cascade in HEK-293T cells, as reflected by IκBα degradation, nuclear translocation of p65, and induction of pro-inflammatory gene transcription. Identification of the NF-κB–activating Secreted Factors in F. varium Culture Supernatant To define the size of the active secreted factors, we first employed stepwise ultrafiltration and evaluated the NF-κB–inducing activity using dual-luciferase reporter assays (Fig. 4A). Strikingly, all filtrates (ranging from < 100 kDa to < 3 kDa) retained robust NF-κB–inducing activity, indicating that the active component(s) can pass through a 3 kDa membrane (Fig. 4B). Conversely, none of the corresponding retentates (> 3 kDa to > 100 kDa) showed any detectable activity (Fig. 4C), indicating that the NF-κB–activating activity was mainly associated with molecules smaller than 3 kDa. We next probed the biochemical nature of the < 3 kDa active fraction. The < 3 kDa fraction of F. varium supernatant was subjected to enzymatic digestion with DNase I, RNase A, and Proteinase K, as well as heat treatment at 100 °C (Fig. 4D). Notably, the NF-κB–inducing activity remained intact following all treatments (Fig. 4E). Together, these findings suggest that the NF-κB–activating activity is unlikely to be primarily attributable to proteins or nucleic acids, and instead implicate heat-stable small-molecule metabolites as candidate active components. Stepwise fractionation identifies NF-κB–activating metabolites from F. varium culture supernatant. (A) Schematic workflow illustrating molecular weight–based fractionation of F. varium culture supernatant and GAM medium control using ultrafiltration. (B) NF-κB–activating activity of F. varium supernatant fractions (< 100 kDa to < 3 kDa), assessed using a HEK-293T dual-luciferase reporter assay. (C) NF-κB–activating activity of F. varium supernatant fractions (> 3 kDa to > 100 kDa), assessed using a HEK-293T dual-luciferase reporter assay. (D) Schematic workflow illustrating enzymatic and heat-based characterization of the NF-κB–activating fraction. (E) NF-κB activity of the < 3 kDa fraction following treatment with DNase I, RNase A, proteinase K, or heat (100 °C), as measured by a HEK-293T dual-luciferase reporter assay. (F) Schematic workflow illustrating organic solvent extraction of F. varium culture supernatant and subsequent assessment of NF-κB–activating activity. (G) NF-κB–activating activity of organic solvent fractions evaluated using HEK-293T dual-luciferase reporter cells. (H) Preparative HPLC was used to isolate fractions F1–F10 from the n-butanol extract, which were further analyzed by analytical HPLC–PDA chromatograms (λ = 210 nm; intensity in mAU). (I) Preparative HPLC was used to isolate subfractions F1S1–F1S5 from fraction F1, which were subsequently analyzed by analytical HPLC–PDA chromatograms (λ = 210 nm; intensity in mAU). (J) Dual-luciferase reporter assays identified fractions F1 and F9 as possessing significant NF-κB–inducing activity. (K) Dual-luciferase reporter assays further identified F1S3 and F1S4 as subfractions derived from F1 with significant NF-κB–inducing activity. Data with error bars represent the mean ± SEM. Each panel is a representative experiment of at least three independent biological replicates. Statistical analyses were performed using two-way ANOVA followed by Šidák’s multiple comparisons test (B, C, E and G), and either one-way ANOVA followed by Dunnett’s multiple comparisons test or unpaired two-tailed t-tests, as appropriate (J and K). ns, not significant; ***P < 0.001; ****P < 0.0001 To further enrich and identify candidate active metabolites, we employed a polarity-based fractionation strategy. Sequential liquid-liquid extraction of the culture supernatant revealed that the NF-κB-activating activity partitioned into both the n-butanol and aqueous fractions, but not into the petroleum ether or ethyl acetate fractions (Fig. 4F and G). Given its consistently higher bioactivity and lower matrix complexity, which favored downstream analysis, we selected the n-butanol extract for further purification. We then scaled up F. varium cultivation and subjected the n-butanol extract to two rounds of activity-guided fractionation using preparative and analytical HPLC. Screening of ten primary fractions (F1–F10; Fig. 4H) identified F1 and F9 as the only ones possessing significant NF-κB-activating activity (Fig. 4J). Subsequent sub-fractionation of F1 yielded five subfractions (F1S1–F1S5) (Fig. 4I), among which F1S3 and F1S4 retained the bioactivity (Fig. 4K). In contrast, sub-fractionation of F9 failed to yield any active subfractions (Fig. S4A and B). These data collectively suggest that the NF-κB–activating activity of F. varium culture supernatant is likely associated with specific small-molecule metabolites enriched in the n-butanol fraction, with F1S3 and F1S4 emerging as candidate active fractions for subsequent structural characterization. Identification of Succinic Acid as a Candidate NF-κB–activating Metabolite To identify candidate microbial metabolites responsible for NF-κB activation, we focused on the F1S3 and F1S4 fractions obtained through HPLC, which displayed robust activity in our initial screen. We performed HPLC–Q/TOF–MS/MS to characterize the metabolite composition of these active fractions (Fig. 5A). From this analysis, we selected several candidate compounds commercially available for validation using a dual-luciferase reporter assay. Among these, succinic acid, present in F1S3, produced a significant, dose-dependent increase in NF-κB activation (Fig. 5B). Kynurenic acid, identified in F1S4, also induced reporter activity but did not exhibit a consistent dose–response relationship (Fig. S5N). The remaining candidates showed no measurable activity compared to vehicle controls (Fig. S5A–M). Based on these observations, we selected succinic acid for further investigation as a candidate NF-κB–activating metabolite. Identification and functional validation of succinic acid as the major NF-κB–activating metabolite from F. varium. (A) Schematic workflow of HPLC–Q/TOF–MS/MS–based screening and identification of bioactive metabolites. (B) Dual-luciferase reporter assay showing dose-dependent NF-κB activation by succinic acid compared with vehicle control. (C) Total ion chromatogram (TIC) of the F1S3 fraction acquired in negative ESI mode, displaying a peak at 0.97 min consistent with succinic acid. (D) TIC of the succinic acid standard showing a peak at 0.85 min, closely aligning with the retention time of the corresponding peak detected in F1S3. (E) MS/MS mirror plot comparing the fragmentation spectra of the candidate compound in F1S3 (red) with the authentic succinic acid standard (blue). Fragment ion m/z values (Δ < 5 ppm) and relative intensity distributions exhibited excellent spectral concordance. (F) Western blot analysis of IκBα protein levels in HEK-293T cells treated with increasing concentrations of succinic acid (125 µM, 250 µM, and 500 µM) for 30 min. Band intensities were quantified using ImageJ and normalized to β-actin. (G) HEK-293T cells stably overexpressing Flag-tagged RelA were treated with 500 µM succinic acid for 30 min, followed by nuclear and cytoplasmic fractionation and western blot analysis of NF-κB p65. Lamin B1 and α-tubulin served as nuclear and cytoplasmic loading controls, respectively. Band intensities were quantified using ImageJ. Data with error bars represent the mean ± SEM. Each panel is a representative experiment of at least three independent biological replicates. Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple comparisons test (B and F) and unpaired two-tailed t-tests (G). ns, not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 To verify the presence of succinic acid in the active F1S3 fraction, we compared its chromatographic and mass spectrometry fragment features to those of an authentic standard. The standard eluted at approximately 0.85 min, aligning closely with the retention time of the candidate peak in F1S3 (Fig. 5C and D). High-resolution MS/MS analysis further confirmed this match, with fragment ion m/z values (Δ < 5 ppm) and intensity distributions highly consistent between the sample and the standard (Fig. 5E), such as the fragment ions m/z 99.0073 and m/z 73.0289. Together, these data confirm the presence of succinic acid in the bioactive F1S3 fraction. Having identified succinic acid as a candidate NF-κB–inducing metabolite, we next asked whether it could activate the canonical NF-κB cascade in HEK-293T cells. Indeed, western blot analysis revealed that succinic acid stimulation led to a dose-dependent loss of IκBα (Fig. 5F), consistent with its degradation. Furthermore, succinic acid treatment specifically increased the accumulation of p65 in the nucleus without altering its cytoplasmic levels (Fig. 5G), demonstrating the translocation event central to NF-κB pathway activation. Together, these data indicate that succinic acid (succinate), a metabolite detected in the bioactive F. varium fraction, can activate canonical NF-κB signaling in HEK-293T cells. CXCR2-dependent Neutrophil Recruitment Contributes to Succinate- and F. varium-associated Colitis Having shown that succinate activates NF-κB signaling in vitro, we next investigated whether elevated succinate availability could exacerbate DSS-induced intestinal inflammation in vivo. We supplemented mice with succinate prior to and during DSS exposure (Fig. 6A). Compared with DSS controls, succinate supplementation markedly exacerbated colonic inflammation, as indicated by greater body weight loss (Fig. 6B), significantly higher DAI (Fig. 6C), shorter colon length (Fig. 6D and E), and histopathological scores (Fig. 6F and G). Analysis of colonic chemokine expression revealed a selective and significant upregulation of Cxcl1 and Cxcl2 in the succinate-supplemented DSS group, whereas Tnf expression remained unchanged. (Fig. 6H). Flow cytometric analysis of the colonic lamina propria (cLP) revealed a marked increase in neutrophil frequency in mice receiving succinate (Fig. 6I and J), while the frequencies of dendritic cells, macrophages, and major T helper cell subsets (Th1, Th2, Th17, and Tregs) showed no significant differences compared with DSS controls (Fig. S6A–L). Collectively, these data indicate that elevated succinate availability can exacerbate experimental colitis and is associated with enhanced neutrophil recruitment in the colon. CXCR2 is a chemokine receptor predominantly expressed on neutrophils and plays a crucial role in neutrophil recruitment to sites of inflammation [24]. To determine whether neutrophil migration contributes to succinate-exacerbated colitis, we employed the selective CXCR2 inhibitor SB225002 to block neutrophil recruitment in vivo (Fig. 7A). Mice treated with the CXCR2 inhibitor SB225002 developed milder colitis in the succinate-supplemented DSS model, exhibiting less body weight loss (Fig. 7B), lower DAI scores (Fig. 7C), and improved colon morphology and length (Fig. 7D and E). Furthermore, flow cytometric analysis revealed that SB225002 treatment markedly reduced neutrophil accumulation in the colonic lamina propria (Fig. 7F and G). Succinate promotes neutrophil recruitment and exacerbates colitis in mice. (A) Schematic of the experimental design. C57BL/6J SPF mice were divided into four groups: Ctrl, Ctrl + DSS, Suc, and Suc + DSS. Succinate was administered in drinking water from day − 3 to day 7, and colitis was induced by 2.5% DSS in drinking water from day 0 to day 7. (B and C) Changes in body weight (B) and DAI (C) monitored daily during the course of DSS treatment. (D and E) Representative images of colons (D) and quantification of colon length (E). (F and G) Representative H&E-stained colon sections (F) and histopathological scores of colons (G). Scale bars, 400 μm and 50 μm (insets). (H) Relative mRNA expression levels of Cxcl1, Cxcl2 and Tnf in colon tissues. (I and J) Representative flow cytometry plots (I) and quantification of neutrophil frequencies (J) in the colonic lamina propria (cLP). Data are presented as mean ± SEM; n = 6 mice per group. Statistical analyses were performed using two-way ANOVA followed by Šidák’s multiple comparisons test (B and C), one-way ANOVA followed by Tukey’s multiple comparisons test (E, G, and J), and unpaired two-tailed t-tests (H). *P < 0.05; **P < 0.01; ***P < 0.001 Inhibition of neutrophil migration by SB225002 alleviates colitis driven by succinate. (A) Schematic of the CXCR2 inhibition experimental design. SPF C57BL/6J mice were pretreated with 3% succinate in drinking water for 3 days, followed by co-administration of 2.5% DSS and 3% succinate for an additional 7 days. Mice received daily intraperitoneal injections of SB225002 (2 mg/kg), a CXCR2 inhibitor, and were sacrificed at the indicated time points. (B-E) Colitis severity was evaluated by body weight change (B), DAI (C), representative colon images (D), and colon length quantification (E). (F) Representative flow cytometry plots showing neutrophils (CD11b⁺Ly6G⁺) in the colon. (G) Quantification of colonic neutrophils among CD45⁺ immune cells. Data are presented as mean ± SEM (n = 6 mice per group). Statistical analyses were performed using two-way ANOVA with Šidák’s multiple-comparisons test (B and C) and the Mann–Whitney test (E and G). *P < 0.05; **P < 0.01 To further establish whether CXCR2-dependent neutrophil recruitment contributes to F. varium-associated colitis, we next treated F. varium-colonized germ-free mice with the CXCR2 antagonist SB225002 during DSS exposure (Fig. 8A). Fecal F. varium levels and succinate concentrations were comparable between PBS- and SB225002-treated mice (Fig. 8B and C), suggesting that SB225002 did not markedly affect F. varium persistence or succinate accumulation. However, SB225002-treated mice exhibited significantly attenuated DSS-induced disease, as shown by reduced body weight loss (Fig. 8D), lower DAI scores (Fig. 8E), and improved colon morphology and colon length (Fig. 8F and G). Flow cytometric analysis further showed that SB225002 markedly reduced the frequency of colonic neutrophils in F. varium-colonized DSS-treated mice (Fig. 8H and I). Together, these findings demonstrate that CXCR2 blockade mitigates F. varium-associated colitis by limiting neutrophil recruitment, thereby supporting a link between F. varium and CXCR2-dependent neutrophil-mediated intestinal inflammation. CXCR2 blockade attenuates F. varium-associated DSS-induced colitis. (A) Schematic of the experimental design. Germ-free C57BL/6J mice were orally gavaged with F. varium every two days from day − 16 to day − 2. From day 0 to day 7, mice received 2% DSS and were intraperitoneally injected daily with PBS or SB225002 (2 mg/kg). (B) Fecal F. varium abundance was quantified by qPCR. (C) Fecal succinate concentrations in F. varium-colonized mice treated with PBS or SB225002. (D − G) Colitis severity was evaluated by body weight change (D), DAI (E), representative colon images (F), and colon length quantification (G). (H − I) Representative flow cytometry plots and quantification of colonic neutrophils (H), defined as Ly-6G⁺CD11b⁺ cells among CD45⁺ immune cells (I). Data are presented as mean ± SEM (n = 6 mice per group). Statistical analyses were performed using two-way ANOVA with Šidák’s multiple-comparisons test (D and E) and the Mann–Whitney test (G and I). *P < 0.05; **P < 0.01; ****P < 0.0001 To examine whether the capacity for succinate production was relatively enriched in F. varium rather than reflecting a general feature of UC-associated dysbiosis, we selected F. varium and two additional representative taxa, Bacteroides uniformis (B. uniformis) and Enterococcus lactis (E. lactis), based on our UC fecal microbiome analysis (Fig. 1G). We then measured succinate concentrations in their culture supernatants. F. varium supernatant contained markedly higher succinate levels than the other tested bacterial supernatants (Fig. S7A). SUCNR1 is a known G protein–coupled receptor for extracellular succinate [25], whereas CXCR2 mediates neutrophil recruitment. We therefore treated HL60-derived neutrophil-like cells with F. varium culture supernatant to link F. varium-derived soluble factors with SUCNR1/CXCR2-associated responses. The selected treatment conditions showed no marked cytotoxicity (Fig. S7B–D). F. varium supernatant increased SUCNR1 and CXCR2 expression, and these responses were attenuated by pretreatment with 10 µM SB225002 or with the SUCNR1 inhibitor NF56-EJ40 at 2 µM (Fig. S7E–G). Together, these results suggest that CXCR2-dependent neutrophil recruitment contributes to both succinate- and F. varium-exacerbated DSS-induced colitis, and further support a link between F. varium-derived soluble factors and SUCNR1/CXCR2-associated neutrophil responses. Succinate Activates NF-κB-associated Inflammatory Signaling in Neutrophil-like Cells To further validate the host cellular mechanisms underlying succinate-mediated inflammation, we next examined its effects in relevant host cell types in vitro. To assess whether succinate directly triggers inflammatory responses in intestinal epithelial cells, we examined the expression of inflammatory mediators in multiple epithelial cell lines (NCM460, Caco-2, and HCT116). Succinate treatment did not consistently induce inflammatory gene expression in these epithelial cells (Fig. 9A–C). In contrast, immune cell lines (HL60 and U937) exhibited robust induction of CXCL chemokines and TNF upon succinate stimulation (Fig. 9D and E). We therefore performed transcriptomic profiling of HL60-derived neutrophils and U937 monocytic cells to characterize succinate-associated inflammatory programs in immune cells. Succinate activates the NF-κB signaling pathway in HL-60–derived neutrophil-like cells. (A-E) Heatmaps showing the relative mRNA expression levels of CXCL1, CXCL2, CXCL3, CXCL8, and TNF in human colonic epithelial cell lines NCM460 (A), Caco-2 (B), and HCT116 (C), as well as immune cells, including HL60-derived neutrophil-like cells (D) and the U937 cells (E). following Suc treatment compared with control (Ctrl). (F) Principal component analysis (PCA) of HL-60–derived neutrophil-like cells treated with succinate or vehicle control. (G) Volcano plot showing DEGs in succinate–treated HL-60 cells compared with control, with upregulated genes (fold change ≥ 1.5, q < 0.05) shown in red–orange and downregulated genes (fold change ≤ − 1.5, q < 0.05) shown in green. (H) Protein–protein interaction (PPI) network analysis of DEGs and identification of hub genes. Genes with fold change ≥ 1.5 or ≤ − 1.5 and q value < 0.05 were imported into the STRING database to construct a PPI network at a confidence score of 0.7. The network was visualized using Cytoscape, and hub genes were identified with the CytoHubba plugin. (I) KEGG pathway enrichment analysis of DEGs in HL-60–derived neutrophil-like cells treated with succinate or control. (J and K) Gene set enrichment analysis (GSEA) of the NF-κB signaling pathway (J) and inflammatory bowel disease (IBD)–related gene sets (K) in HL-60 cells treated with succinate To systematically assess the impact of succinate on immune cell responses, we performed transcriptomic profiling of HL60-derived neutrophils and U937 monocytic cells following succinate stimulation. This analysis revealed substantial succinate-induced transcriptional remodeling in both cell types. Notably, succinate elicited a more pronounced inflammatory transcriptional response in HL60-derived neutrophils, with treated and control samples clearly separating into distinct transcriptional states (Fig. 9F and S8A–C). Given this stronger response in HL60-derived neutrophil-like cells, together with our in vivo observation that succinate supplementation increased neutrophil accumulation in the colon, we focused subsequent analyses on neutrophils. Integrated analyses—including differential expression, network analysis, and pathway enrichment—converged on NF-κB signaling as a central inflammatory pathway (Fig. 9G–I). Consistent with these findings, GSEA further revealed significant enrichment of NF-κB signaling and inflammatory bowel disease–associated gene signatures (Fig. 9J and K), suggesting that succinate induces inflammatory transcriptional programs potentially relevant to mucosal inflammation. We next tested whether SUCNR1 mediates succinate-induced NF-κB activation in neutrophils. Transcriptomic profiling revealed that succinate robustly induced the expression of multiple inflammatory mediators, including PTGS2, IL1A, CXCL8, TNF, IL1B, and CCL2. Importantly, SUCNR1 inhibition with NF56-EJ40 markedly reduced these responses in neutrophil-like cells (Fig. 10A), suggesting that SUCNR1 mediates succinate-induced inflammatory activation in neutrophil-like cells. At the protein level, ELISA further demonstrated that succinate promoted CXCL8 secretion, a key neutrophil chemoattractant, and this effect was significantly blunted by NF56-EJ40 (Fig. 10B). Immunoblot analysis further showed that succinate induced IκBα degradation and promoted nuclear translocation of NF-κB p65, both of which were largely abolished by SUCNR1 inhibition (Fig. 10C and D). These results indicate that succinate activates canonical NF-κB signaling in HL60-derived neutrophil-like cells via SUCNR1. Collectively, these findings support a model in which succinate engages SUCNR1 to activate NF-κB–linked inflammatory programs in neutrophil-like cells, providing a potential link between microbial metabolism and neutrophil-associated mucosal inflammation. Succinate–SUCNR1 signaling drives NF-κB–dependent inflammatory responses in HL60-derived neutrophils. (A) HL-60–derived neutrophil-like cells were pretreated with the pharmacological SUCNR1 antagonist NF56-EJ40 (1 µM) for 1 h prior to stimulation with succinate (500 µM). Relative mRNA expression of CXCL8, CCL2, TNF, IL1B, IL1A, and PTGS2 was measured by RT-qPCR. (B) The expression of CXCL8 was quantified by ELISA. (C) Representative immunoblots and quantitative analysis of IκBα protein levels in cells treated with succinate, NF56-EJ40 + succinate, or NF56-EJ40 alone. β-actin was used as a loading control. (D) Immunoblot analysis and quantification of NF-κB p65 levels in the nuclear and fraction cytoplasmic under the indicated treatments. Lamin B1 served as a nuclear loading control, α-tubulin was used as a cytoplasmic loading control. Data with error bars represent the mean ± SEM. Each panel is a representative experiment of at least three independent biological replicates. One-way ANOVA followed by Tukey’s multiple comparisons test (A and B) and One-way ANOVA followed by Dunnett’s multiple comparisons test (C and D). ns, not significant; *P < 0.05; **P < 0.01; ****P < 0.0001 Together, our study identifies F. varium as a colitis-associated pathobiont linked to succinate accumulation and neutrophil-driven intestinal inflammation. Elevated succinate activates NF-κB signaling in neutrophil-like cells in vitro, and promotes neutrophil recruitment to the inflamed colon and exacerbates intestinal inflammation in vivo. These findings support a working model in which UC-enriched F. varium is linked to succinate-associated inflammatory signaling and CXCR2-dependent neutrophil recruitment, thereby amplifying mucosal inflammation (Fig. 11). Proposed model linking UC-enriched F. varium to succinate-associated neutrophilic inflammation Discussion Alterations in the gut microbiota and the expansion of disease-associated pathobionts are increasingly recognized as important contributors to mucosal inflammation in IBD [26–28]. However, despite dysbiosis being a consistent hallmark of UC, the identification of specific pro-inflammatory taxa and the mechanisms through which they contribute to intestinal inflammation remain incompletely defined. In line with many previous UC microbiome studies, we observed reduced alpha diversity in the UC group, supporting a loss of microbial richness and/or evenness as a feature of UC-associated dysbiosis [29, 30]. Nevertheless, the magnitude of alpha-diversity changes may vary across cohorts because of differences in medication exposure [31], clinical phase [32], sample type and sampling location [33], sequencing depth, and bioinformatic pipelines [34]. Fusobacterium species have attracted particular attention, although most studies have focused on Fusobacterium nucleatum [35], with comparatively limited investigation of F. varium. Early studies from Japan reported a markedly increased prevalence of F. varium in the colonic mucosa of patients with UC and showed that F. varium culture supernatants could induce colonic inflammation in mice [11, 12]. Subsequent work from Korea further detected F. varium and related Fusobacterium species in UC tissues [13]. Nevertheless, these studies were largely restricted to mucosal sampling and did not define the bioactive components responsible for inflammatory activation, nor did they delineate the host immune pathways involved. As such, whether F. varium contributes to UC-associated inflammation through a specific metabolite–receptor signaling axis has remained unresolved. In the present study, we extend previous observations by demonstrating significant enrichment of F. varium in fecal samples from patients with UC, thereby broadening its association beyond geographically restricted mucosal findings. Importantly, through activity-guided metabolite analysis, we identify succinate as a candidate low-molecular-weight effector that activates NF-κB signaling and is associated with neutrophil-centered inflammatory amplification. These findings extend previous correlative observations and provide a mechanistic basis for understanding how a UC-enriched pathobiont-associated metabolite may amplify mucosal immune activation. Collectively, our data support the concept that F. varium may represent a functionally relevant UC-associated pathobiont and highlight the broader importance of resolving microbe-associated metabolites as contributors to mucosal immune dysregulation. Thus, the main advance of this study is not the general observation of UC-associated dysbiosis, Fusobacterium enrichment, or succinate immunometabolism, but the functional connection between a UC-derived F. varium strain, F. varium-associated metabolic activity, and neutrophil-centered inflammatory amplification under colitic conditions. Interactions between commensal bacteria and host immune cells are essential for maintaining intestinal homeostasis. While many commensal species promote immune tolerance, increasing evidence suggests that pathobionts can subvert these regulatory networks and elicit pro-inflammatory immune responses [36]. In this study, colonization with a UC-derived F. varium strain aggravated DSS-induced colitis, accompanied by pronounced neutrophil accumulation in the colonic lamina propria. Neutrophils represent key innate effector cells in IBD and can amplify tissue injury through the release of reactive oxygen species, proteases, and neutrophil extracellular traps [37–40]. Moreover, inflammatory chemokine circuits, including the Cxcl1/Cxcl2 axis, can reinforce neutrophil recruitment and shape downstream adaptive immune responses [41, 42]. Notably, F. varium colonization also promoted Th17 cell expansion in vivo, consistent with reported reciprocal crosstalk between neutrophils and Th17 cells. Activated neutrophils can facilitate Th17 recruitment via chemokines such as CCL2 and CCL20, whereas Th17-derived IL-17 A/F further amplifies neutrophil responses by inducing chemokine production in epithelial and stromal compartments [42]. Supporting this framework, we observed robust upregulation of Cxcl1 and Cxcl2 in F. varium–colonized mice, particularly following DSS challenge, consistent with enhanced neutrophil recruitment. Although Th17 cells were also increased in F. varium-colonized mice following DSS challenge, the present study focused mechanistically on neutrophils because neutrophil accumulation and neutrophil-associated chemokines were consistently observed across the F. varium and succinate models, whereas the mechanistic contribution of Th17 responses to F. varium-associated inflammatory amplification will require further investigation. It should also be noted that the DSS model primarily reflects chemically induced epithelial barrier injury. Because F. varium colonization alone did not induce overt colonic inflammation but aggravated DSS-induced colitis, our findings support an inflammation-amplifying rather than disease-initiating role of F. varium. In the absence of comparison strains, whether the inflammation-amplifying effect observed in the DSS model is specific to F. varium or reflects a broader response to microbial stimulation remains to be determined. Phylogenetic analysis revealed that our UC-derived F. varium isolate clustered most closely with strain Fv113-g1 [23], forming a distinct clade separated from other publicly available reference genomes, including ATCC 27,725. Previous comparative genomics and transcriptomic analyses of Fv113-g1 reported an expanded accessory gene repertoire, including multiple type V secretion systems and several fadA paralogs, with increased expression under host-mimicking conditions, supporting niche-specific adaptation and enhanced mucosal persistence. The close genomic relatedness between our isolate and Fv113-g1 suggests that these UC-derived F. varium strains may share similar adaptive features. However, this interpretation remains preliminary. Moreover, the clinical sources of publicly available F. varium genomes are limited or incompletely annotated, and the number of UC-derived and non-UC comparator isolates remains insufficient to define disease-specific genomic signatures. Therefore, larger comparative genomic studies including additional UC-derived isolates and non-UC clinical or commensal isolates are needed to determine whether specific F. varium lineages are associated with UC. NF-κB signaling is a central hub linking microbial cues to mucosal immune activation in IBD [43]. To enable activity-guided identification of pro-inflammatory bacterial products, we employed an NF-κB–driven dual-luciferase reporter assay as a sensitive transcriptional readout. Compared with monitoring individual proximal signaling events, reporter-based assays provide an integrated measure of NF-κB–dependent transcriptional activation and are well suited for activity-guided screening of microbial supernatants and metabolites [44–46]. We utilized HEK-293T cells as a standardized low-background platform with high transfection efficiency [47], allowing robust detection of NF-κB activation across fractionation steps. This approach rapidly pinpointed F. varium culture supernatant as a potent inducer of NF-κB signaling and facilitated subsequent identification of succinate as a key bioactive metabolite. Importantly, these findings were further validated in immune cells and in vivo models, supporting a mechanistic link between F. varium metabolic activity and host inflammatory responses. Nevertheless, because our activity-guided screening strategy was based on an NF-κB reporter system, this study primarily captured NF-κB–activating metabolites, whereas other NF-κB-independent immune pathways or cellular responses were not comprehensively investigated. Our fractionation and stability assays revealed that the NF-κB–activating activity of F. varium was confined to a heat-stable, low–molecular-weight (< 3 kDa) fraction, indicating that the responsible factors are small metabolites rather than proteinaceous virulence determinants. The chemical stability of these molecules suggests that they may persist within the intestinal lumen despite exposure to digestive enzymes and fluctuating physicochemical conditions, thereby acting as durable mediators of host–microbe signaling in the inflamed gut. Beyond previously described virulence-associated features of F. varium, including FadA adhesins and secretion systems [23], our findings identify secreted metabolites as an additional class of host-interacting factors capable of activating NF-κB signaling and contributing to mucosal inflammation. This mechanism aligns with accumulating evidence that microbial metabolites, such as cadaverine [48], bile acid derivatives [49], and indolimine [50] can profoundly modulate epithelial integrity and immune activation. Together, these results support a broader paradigm in which gut bacteria shape host immunity not only through structural components or classical virulence factors but also through chemically stable bioactive metabolites. Our study places F. varium within this framework and highlights metabolite-mediated signaling as a potential pathway through which UC-associated pathobionts exacerbate intestinal inflammation. Nevertheless, bacterial culture supernatants are complex mixtures, and potential synergistic effects between multiple molecules cannot be excluded. In addition, other heat-stable low-molecular-weight metabolites present in the supernatant may also contribute to the observed inflammatory activity, either independently or together with succinate. Thus, although succinate was identified as an active component in this system, the overall activity of the culture supernatant cannot be attributed solely to succinate. In this study, we identified succinate as a key NF-κB–activating metabolite through activity-guided fractionation of F. varium culture supernatants. Although kynurenic acid was also detected, its weak and non–dose-dependent activity suggests that succinate may represent a candidate bioactive metabolite contributing to F. varium–associated NF-κB activation. Traditionally viewed as a TCA cycle intermediate, succinate is now recognized as an immunometabolite that accumulates in inflamed tissues and signals through SUCNR1 to modulate immune activation. Elevated luminal succinate in active UC and increased fecal succinic acid during UC exacerbation have been reported, supporting the relevance of succinate metabolism to mucosal immune dysregulation [51, 52]. Our findings extend this paradigm by showing that exogenous succinate exacerbates DSS-induced colitis and is associated with enhanced neutrophil recruitment, suggesting that succinate may contribute to neutrophil-associated inflammation in the gut. This is consistent with our observation that gavage administration of F. varium culture supernatants exacerbated neutrophil accumulation in the colon during DSS-induced colitis, although further studies are needed to directly define the in vivo contribution of F. varium-derived succinate to this process. However, these findings do not establish succinate as the primary mediator of F. varium-induced inflammatory effects in vivo. Additional approaches, such as selective depletion of succinate, inhibition of bacterial succinate production, or genetic/metabolic manipulation of F. varium, will be required to strengthen the causal link between F. varium-associated succinate production and inflammatory amplification in vivo. Notably, succinate–SUCNR1 signaling has been reported to exert context-dependent immunoregulatory effects. In macrophages, SUCNR1 activation can support anti-inflammatory programs under type 2 cytokine–rich conditions [53], whereas in other cellular and tissue contexts, including intestinal inflammation, succinate–SUCNR1 signaling has been linked to pro-inflammatory outputs [51, 54]. Together, these observations suggest that the consequences of succinate signaling are shaped by the responding cell type and local microenvironment. Further delineation of the spatial and cellular dynamics of succinate production and sensing in the gut will be important for evaluating succinate–SUCNR1 signaling as a potential therapeutic target in UC. This study has several limitations. Because the relevant clinical information was obtained from routine medical records, we could not confirm whether all UC patients had undergone endoscopic assessment or had been evaluated using standardized disease activity indices or endoscopic scoring systems. In addition, the moderate sample size may have limited the statistical power for further clinical subgroup analyses. Moreover, although this study focused on F. varium and its metabolite succinate, the gut microbiome is a complex ecosystem involving community-level compositional changes, microbial interactions, and functional alterations in IBD [55]. Succinate is not uniquely produced by F. varium, and other gut bacteria, such as Bacteroides thetaiotaomicron [56] and Prevotella copri [57], may also contribute to succinate production or metabolism. Therefore, investigation of a single bacterial species and its metabolite may not fully capture the broader microbiome complexity associated with UC. Future studies comparing succinate production across additional UC-associated and non-UC-associated bacterial taxa, together with longitudinal measurement of luminal and tissue succinate levels in vivo, will be needed to define the quantitative contribution of F. varium-derived succinate to intestinal inflammation and to validate these findings within the broader context of UC-associated dysbiosis. Furthermore, although germ-free mice allowed us to assess the effect of F. varium colonization in the absence of pre-existing microbiota, we did not further validate these findings in antibiotic-treated SPF mice. Future studies using this complementary model are needed to determine whether F. varium exerts similar pro-inflammatory effects in a microbiota-depleted but not germ-free intestinal environment. It should also be noted that succinate supplementation in the DSS model represents an exogenous intervention designed to establish sustained succinate availability before and during DSS-induced epithelial injury, rather than to reproduce the endogenous temporal dynamics of succinate production or accumulation during colitis. Because luminal and tissue succinate levels were not measured longitudinally, future time-course analyses will be required to clarify the temporal and spatial relationship between endogenous succinate accumulation, SUCNR1 signaling, and intestinal inflammation. In addition, only female mice were used in the present study, which may limit the generalizability of our findings. This design was adopted to improve model consistency and reduce potential housing-related confounding factors associated with group-housed male mice, such as aggression, fighting, and social hierarchy-related stress [58]. However, sex as a biological variable may influence mucosal immune responses, gut microbiota composition, susceptibility to DSS-induced colitis, and responses to therapeutic or hormonal interventions [59, 60]. Future studies including both male and female mice are warranted to further validate the generalizability of these findings. Another limitation is that some mechanistic experiments used HEK-293T cells and HL60-derived neutrophil-like cells rather than primary intestinal or immune cells. Thus, the in vivo relevance of these findings should be interpreted with caution and further validated using primary cells, patient-derived organoids, and additional in vivo models. Our study extends current understanding of microbiota–metabolite interactions in intestinal inflammation by identifying F. varium-associated succinate as a candidate mediator linked to enhanced neutrophil recruitment in colitis. Together with the observation that succinate activates NF-κB signaling in neutrophil-like cells, these findings suggest that microbial metabolites may contribute to neutrophil-associated inflammatory responses during intestinal inflammation. Taken together, our findings suggest that microbiota-derived succinate contributes to neutrophil recruitment and intestinal inflammation, at least in part through SUCNR1-NF-κB–related signaling. By highlighting the potential contribution of microbial metabolites to neutrophil-associated inflammatory responses, this study suggests that targeting microbial metabolic products or their downstream signaling pathways may represent a promising strategy for the treatment of UC. From a therapeutic perspective, future strategies could include limiting bacterial succinate production or accumulation, modulating succinate-associated pathobionts, or targeting succinate-mediated host inflammatory signaling. However, these approaches remain exploratory and require further validation given the complexity of microbial metabolism and host–microbiome interactions. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (XLSX 12.8 KB) Supplementary Material 2 (XLSX 31.4 KB) Supplementary Material 3 (XLSX 43.3 KB) Supplementary Material 4 (XLSX 30.2 KB) Supplementary Material 5 (DOCX 11.3 MB) Supplementary Material 6 (DOCX 1.28 MB) Acknowledgements We thank all members in the Cuiying Biomedical Research Center, and the Second Hospital & Clinical Medical School, Lanzhou University. Author Contributions Qian Wang performed the experiments, conducted statistical analyses, curated and visualized the data, and drafted and revised the manuscript. Han Xiao, Wenxiu Liu and Xiangji Dang performed experiments, analyzed and interpreted the data, and contributed to manuscript revision. Yanrui Bai, Ruihong Xiao and Liyu Tong carried out animal experiments and contributed to data curation. Yan Wang, Mi Li and Shuhan Wang contributed to data curation and visualization. Sugui Pu, Duoji Pei, Dekui Zhang and Xiang Wang collected clinical samples and performed data curation. Guang Hu, Jingjing Guo and Xiaojie Jin contributed to the interpretation of the results. Long Qin, Chunjiang Zhang, Yana Li, Tiansheng Zhang, Jing Yang and Qi Wang contributed to the discussion of the results. Hui Sun conceived and designed the project, supervised the study, acquired funding, and revised the final manuscript. Funding This work was funded by National Natural Science Foundation of China (82572004), the Fundamental Research Funds for the Central Universities (lzujbky-2023-eyt04), Key Research and Development Program of Gansu Province (25YFFA067), Gansu Provincial Major Science and Technology Special Project Plan (24ZDNA003), Gansu Province key talent project (2023RCXM65), Lanzhou Talent Innovation and Entrepreneurship Project (2021-RC-115), Talent Introduction Plan of the Second Hospital of Lanzhou University (yjrckyqdj-2020-01, yjrckyqdj-2022-01), and the Longyuan Youth Talent Program of Gansu Province (to J.Y.). Data Availability The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Qian Wang, Han Xiao, Wenxiu Liu and Xiangji Dang contributed equally to this work. References Associated Data Data Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.