Application of Computational Data Modeling to a Large-Scale Population Cohort Assists the Discovery of Inositol as a Strain-Specific Substrate for Faecalibacterium prausnitzii

Abstract

Faecalibacterium prausnitzii (F. prausnitzii) is a bacterial taxon in the human gut with anti-inflammatory properties, and this may contribute to the beneficial effects of healthy eating habits. However, little is known about the nutrients that enhance the growth of F. prausnitzii other than simple sugars and fibers. Here, we combined dietary and microbiome data from the American Gut Project (AGP) to identify nutrients that may be linked to the relative abundance of F. prausnitzii. Using a machine learning approach in combination with univariate analyses, we identified that sugar alcohols, carbocyclic sugar, and vitamins may contribute to F. prausnitzii growth. We next explored the effects of these nutrients on the growth of two F. prausnitzii strains in vitro and observed robust and strain-dependent growth patterns on sorbitol and inositol, respectively. In the context of a complex community using in vitro fermentation, neither inositol alone nor in combinations with vitamin B exerted a significant growth-promoting effect on F. prausnitzii, partly due to high variability among the fecal microbiota community from four healthy donors. However, the fecal communities that showed an increase in F. prausnitzii on inulin also responded with at least 60% more F. prausnitzii on any of inositol containing media than control. Future nutritional studies aiming to increase the relative abundance of F. prausnitzii should explore a personalized approach accounting for strain-level genetic variations and community-level microbiome composition.