**Background:** Crohn's disease (CD) is a multifactorial inflammatory bowel disease driven by genetic susceptibility, environmental factors, and gut microbiota. Murine models of CD-like ileitis require microbial triggers. The TnfΔARE mouse model develops transmural ileal inflammation resembling human CD, and germ-free TnfΔARE mice remain disease-free. Previous work by this group showed expansion of unknown Clostridiales and filamentous structures near the epithelium in TnfΔARE mice, pointing to segmented filamentous bacteria (SFB), a commensal mucosal-adherent bacterium known to induce Th17 responses in mice. The role of SFB in CD-like inflammation and the mechanism by which exclusive enteral nutrition (EEN) exerts its protective effects remain unclear.
**Methods:** Germ-free TnfΔARE mice and wild-type (WT) littermates were colonized with SPF-derived WT cecal microbiota, SFB (by maternal transmission or oral gavage), individual bacterial strains (Alistipes, B. adolescentis L2-32, E. coli LF82, L. murinus), a 7-strain minimal bacterial consortium (MIBAC), or human CD patient fecal microbiota (with or without SFB co-colonization). Mice were fed chow diet, EEN-like purified diet (PD), or fiber-rich purified diet (FRD). Inflammation was assessed by histopathology scoring (0–12), qPCR for cytokine and antimicrobial peptide gene expression, flow cytometry for immune cell populations, immunofluorescence for Paneth and goblet cells, and FISH for SFB localization. SFB abundance was quantified by qPCR (Ct values). Human SFB screening used 468 ileal and colonic mucosal biopsies from three independent IBD cohorts (Spanish HSCT cohort: 44 ileal, 86 colonic biopsies from CD patients; French biotherapy cohort: 43 ileal, 143 colonic biopsies from CD and UC patients; German pediatric cohort: 72 ileal, 80 colonic biopsies from children with IBD) using three SFB-specific primer sets.
**Key Results:** In SPF-housed TnfΔARE mice, SFB abundance strongly correlated with ileitis severity (p < 0.0001, r = 0.5998). In SAMP/Yit mice, a similar correlation was observed (p < 0.01, r = 0.4827). SFB mono-colonization of germ-free TnfΔARE mice from birth induced severe ileo-colonic inflammation by 12 weeks of age, with elevated Tnf, Il-17A, and Ifng transcript levels, neutrophil infiltration, and significantly reduced Paneth cells (lysozyme-positive) and goblet cells (PAS/AB-positive). Antimicrobial peptide expression (Defa5, Ang4, Reg3b) was significantly reduced in inflamed mice. SFB failed to induce inflammation in Il10-/- mice despite comparable SFB loads. B. adolescentis mono-colonization did not trigger inflammation in TnfΔARE mice. In 468 human IBD mucosal biopsies screened with three SFB-specific primer sets (including primers based on the published human SFB genome), SFB was not detected in any sample (0/468). Human fecal microbiota transfer from CD patients failed to induce inflammation in TnfΔARE mice, but co-colonization with SFB reactivated inflammation, with reduced Paneth cell numbers and expansion of Alistipes and Bilophila. EEN-like purified diet completely prevented SFB colonization in mono-colonized TnfΔARE mice (Ct values ≥ 35 vs. 21.5 ± 2.5 on chow diet), abolished inflammation, and preserved Paneth cell numbers. Fiber-rich purified diet allowed partial SFB colonization (Ct values = 32.4 ± 3.5) but did not induce inflammation.
**Clinical Implications:** This study identifies SFB as a novel pathobiont capable of driving severe CD-like ileo-colonic inflammation in genetically susceptible murine hosts, characterized by Th1/Th17-driven immune responses and impaired epithelial antimicrobial defense. The absence of SFB in 468 human IBD biopsies suggests species-specific functionality, indicating that SFB-driven mechanisms in mice may not directly translate to human CD. However, the demonstration that an EEN-like purified diet completely prevents SFB colonization and inflammation provides a plausible mechanistic explanation for the clinical efficacy of EEN in pediatric CD: dietary modulation of pathobiont expansion. The finding that fiber re-introduction partially restores SFB colonization without causing inflammation suggests that fiber content may be a critical dietary determinant of SFB growth, though the relationship between fiber, SFB, and inflammation is complex and may depend on host genetics and microbial community context. These results support a direct causal link between diet, microbial community structure, and intestinal inflammation, and suggest that dietary interventions targeting specific pathobionts may represent a therapeutic strategy in IBD.