ObjectiveA decade of microbiome studies has linked IBD to an alteration in the gut microbial community of genetically predisposed subjects. However, existing profiles of gut microbiome dysbiosis in adult IBD patients are inconsistent among published studies, and did not allow the identification of microbial signatures for CD and UC. Here, we aimed to compare the faecal microbiome of CD with patients having UC and with non-IBD subjects in a longitudinal study.DesignWe analysed a cohort of 2045 non-IBD and IBD faecal samples from four countries (Spain, Belgium, the UK and Germany), applied a 16S rRNA sequencing approach and analysed a total dataset of 115 million sequences.ResultsIn the Spanish cohort, dysbiosis was found significantly greater in patients with CD than with UC, as shown by a more reduced diversity, a less stable microbial community and eight microbial groups were proposed as a specific microbial signature for CD. Tested against the whole cohort, the signature achieved an overall sensitivity of 80% and a specificity of 94%, 94%, 89% and 91% for the detection of CD versus healthy controls, patients with anorexia, IBS and UC, respectively.ConclusionsAlthough UC and CD share many epidemiologic, immunologic, therapeutic and clinical features, our results showed that they are two distinct subtypes of IBD at the microbiome level. For the first time, we are proposing microbiomarkers to discriminate between CD and non-CD independently of geographical regions.

Gut bacterium Faecalibacterium prausnitzii activates a newly identified set of human IL-10-producing Treg cells (CD4CD8αα lymphocytes), revealing a mechanism by which commensal microbes contribute to host immunity.

The human colonic mucosa contains regulatory type 1-like (Tr1-like, i.e., IL-10-secreting and Foxp3-negative) T cells specific for the gut Clostridium Faecalibacterium prausnitzii (F. prausnitzii), which are both decreased in Crohn's disease patients. These data, together with the demonstration, in mice, that colonic regulatory T cells (Treg) induced by Clostridium bacteria are key players in colon homeostasis, support a similar role for F. prausnitzii-specific Treg in the human colon. Here we assessed the mechanisms whereby F. prausnitzii induces human colonic Treg. We demonstrated that F. prausnitzii, but not related Clostridia, skewed human dendritic cells to prime IL-10-secreting T cells. Accordingly, F. prausnitzii induced dendritic cells to express a unique array of potent Tr1/Treg polarizing molecules: IL-10, IL-27, CD39, IDO-1, and PDL-1 and, following TLR4 stimulation, inhibited their up-regulation of costimulation molecules as well as their production of pro-inflammatory cytokines IL-12 (p35 and p40) and TNFα. We further showed that these potent tolerogenic effects relied on F. prausnitzii-induced TLR2/6 triggering, JNK signaling and CD39 ectonucleotidase activity, which was induced by IDO-1 and IL-27. These data, together with the presence of F. prausnitzii-specific Tr1-like Treg in the human colon, point out to dendritic cells polarization by F. prausnitzii as the first described cellular mechanism whereby the microbiota composition may affect human colon homeostasis. Identification of F. prausnitzii-induced mediators involved in Tr1-like Treg induction by dendritic cells opens therapeutic avenues for the treatment of inflammatory bowel diseases.