Prioritization and annotation of novel bioactive small molecules from the microbiome

Microbial communities, and the human gut microbiome, in particular, are a rich source of novel bioactive small molecule metabolites. Tens of thousands of metabolites have been assayed from stool, many in association with millions of microbial enzymes, but as yet with minimal biochemical characterization or knowledge of their therapeutic potential. This is especially true for microbially-derived or -associated small molecule immunomodulators in conditions such as the inflammatory bowel diseases (IBD), in which gut microbial alterations have been implicated in induction of or protection from inflammation. Here, we have developed and validated a new approach for identifying and prioritizing potentially bioactive novel metabolites from the gut microbiome, which we initially applied to ~82k compounds spanning 546 metabolomes from 106 IBD patients and controls in the Integrative Human Microbiome Project (HMP2). We assigned putative biochemical annotations and prioritization of potential bioactivity was done by integrating epidemiological properties (e.g. IBD pathogenesis) with ecological ones (e.g. covariation and prevalence) Top-ranked features were enriched for bile acid derivatives and short-chain fatty acid precursors, among other classes of chemical compounds that have been previously implicated in IBD-related dysbiosis, as well as modestly explored classes such as medium-chain fatty acids, putrescine metabolites and B vitamins. These results point to new potentially microbially-derived and -associated compounds for immunomodulation in inflammatory conditions. The general method can be applied to integrate knowledge of standard compounds from any microbial community metabolomics with phenotypic or environmental indicators of bioactivity. We provide an open source implementation as MACARRoN (Metabolome Analysis and Combined Annotation Ranks for pRediction of Novel bioactives). The metabolites prioritized in this study of IBD expand our understanding of interactions between the microbiota and host during gut inflammation and offer new candidate small molecules with therapeutic potential.