EAE

Given the contribution of the gut microbiota to pathogenesis of autoimmune diseases, microbiota characteristics could potentially be used to predict disease susceptibility or progression. Although various gut commensals have been proposed as risk factors for autoimmune disease development, predictions based on microbiota composition alone remain unreliable. Here, we evaluated a common approach to identify a potential microbial risk factor from complex communities, followed by in-depth evaluation of its risk factor properties using an autoimmune neuroinflammation disease model in mice harboring several distinct, defined microbiota compositions. We found that the relative abundances of commensal taxa across distinct communities are poorly suited to assess the disease-mediating property of a given microbiota. Instead, the presence of certain microbial risk factors allowed us to determine the probability of severe disease, but failed to predict the individual disease course. We investigated multiple other microbiota-associated characteristics by applying 16S rRNA gene sequencing, metatranscriptomic and metabolomic approaches, as well as in-depth analysis of host immune responses and intestinal barrier integrity-associated readouts. By leveraging gnotobiotic mouse models harboring six defined compositions, we identified the IgA coating index of Bacteroides ovatus as a reliable individual disease risk predictor before disease onset, due to its ability to reflect autoimmunity-mediating properties of a given gut microbial network. In summary, our data suggest that common taxonomic analysis approaches should be refined by taxonomic network analyses or combined with microbiota function-related readouts to reliably assess disease predisposition of a given host-microbiota combination.