Identifying Novel Bioactive Microbial Gene Products in Inflammatory Bowel Disease

Microbial communities and their associated bioactive compounds are often disrupted in conditions such as inflammatory bowel diseases (IBD). However, even in well-characterized environments (e.g. the human gastrointestinal tract), more than one-third of microbial proteins are uncharacterized and often expected to be bioactive. Here, we introduce a method to identify putative bioactive gene products in any microbial community and use it to prioritize potentially bioactive novel proteins from the human gut during IBD, beginning with a catalog of 1,665,223 protein families assembled from 1,595 metagenomes in the Integrative Human Microbiome Project (HMP2). Remarkably, 1,157,695 (~70%) of these proteins were uncharacterized, including those with strong homology to known functionally uncharacterized proteins (24%), potentially novel proteins with weak homology (33%), and completely novel proteins without significant homology (12%). We assigned putative annotations to these uncharacterized families using a combination of guilt-by-association, taxonomic binning, secondary structure analysis, and environmental parameters or phenotypes, ultimately leaving only 30,567 protein families (2.6%) still functionally and taxonomically unannotated. >340,000 protein families were prioritized as potentially bioactive with respect to gut inflammation during IBD by integrating evidence from ecological properties and environmental or phenotypic properties. ~39% of the prioritized novel proteins expanded the pangenomes of common gut taxa, and >90% of the remainder with unclear taxonomy were assigned at least one putative biochemical annotation. To validate our prioritized microbial gene products, we used a combination of metagenomics, metatranscriptomics and protein structure analysis to provide evidence of bioactivity for a subset of proteins involved in the crosstalk processes between host and microbial communication, such as adherence/invasion processes and secreted proteins. This methodology is fully generalizable to other environmental microbial communities and human disease phenotypes, and we provide an open source implementation as MetaWIBELE (Workflow to Identify novel Bioactive Elements in the microbiome). Prioritized results here provide thousands of new candidate microbial proteins likely to interact with the host immune system in IBD, expanding our understanding of potentially bioactive gene products in chronic disease states and offering a rational and targeted compendium of potential therapeutic compounds.