The gastrointestinal tract (GIT) of mammals is colonized by a community of micro-organisms reaching levels as high as 1011 bacteria/g of contents in the colon. This microbiota is involved in intestinal homeostasis and epithelium structure through the control of proliferation and differentiation processes (Cherbuy et al., 2010). Bacteroides thetaiotaomicron and Faecalibacterium prausnitzii, dominant members of microbiota, are commensal model bacteria to study host-microbiota interactions. The objective of our work was to understand the effects of these two bacteria on colonic epithelium homeostasis in gnotobiotic rats and underlying this crosstalk using In Vitro models. Germ-free rats (GF) were orally inoculated with B. thetaiotaomicron either alone or with a mix of B. thetaiotaomicron and F. prausnitzii. Colonic epithelium morphology, crypt depth, cell cycle-related proteins and goblet cells markers were analyzed by histological staining, immunohistochemistry and Western blot in GF (n= 6), mono-associated (Ino-Bt; n=12) or di-associated (Ino-Bt+Fp; n=6) rats 30 days after inoculation. In Ino-Bt rats, no effect was observed neither on crypt depth nor on proliferation markers whereas Kruppel Like Factor 4 protein (KLF4), a transcription factor involved in goblet cells differentiation, was significantly increased (2.5 ±0.8-fold) compared to GF rats. In parallel with this KLF4 induction, a significant increase of goblet cells per crypt in the colon of Ino-Bt compared to GF rats was observed. In Ino-Bt+Fp rats, there was no induction of KLF4 protein, suggesting that the presence of F. prausnitzii impaired the mucus-stimulatory effect triggered by B. thetaiotaomicron. Short Chain Fatty Acids (SCFA) composition was determined on caecal contents by gas chromatograph to assess bacterial metabolic activity. In Ino-Bt rats, acetate was the major SCFA detected (5.8 μmol/g) with a low level of propionate (1.0 μmol/g) and no butyrate. In Ino-Bt+Fp rats, butyrate was the major SCFA (1.3 μmol/g) and acetate concentration was significantly decreased (two-fold) compared to Ino-Bt rats (3.0 μmol/g). Using HT29-MTX cells, a cell line producing mucus, we showed that acetate was able to increase significantly KLF4 protein whereas butyrate and propionate have no effect on KLF4 level. All of these results suggested that acetate, a bacterial metabolite produced by B. thetaiotaomicron, could mediate a bacterial signal to promote goblet cells differentiation pathway in the GIT of mono-associated rats. In Crohn's disease patients, a diminished prevalence of F. prausnitzii (Martinez-Medina et al., 2006) and an increase of mucus production (Gersemann et al., 2009) have been often reported. Our observations in gnotobiotic rats are coherent with reports made in patients and may provide mechanistic clues to understand the pathophysiology of inflammatory bowel diseases.