SELECTION OF MICROBIOME-DERIVED PEPTIDES THROUGH MULTI-TECHNOLOGY META-ANALYSIS, BINDING, AND FUNCTIONAL ASSAYS, IDENTIFIES PEPTIDES WITH THERAPEUTIC POTENTIAL THAT MODULATE IMMUNE RESPONSES AND DISEASE

Abstract

Abstract Microbiome-driven drug discovery holds great promise for therapeutic development. However, identification of disease-associated microbiome-host interactions are challenging, owing largely to the complexity of a diverse community and the myriad of direct and indirect interactions through which the microbiome can influence its host. Moreover, inflammatory bowel disease (IBD) is a heterogeneous disease with a complex diagnosis, and multiple etiologies, pathologies, and treatment responses. Finding robust associations between IBD and the microbiome has proven difficult. Using a systematic computational pipeline, we identified bacterially-derived peptides that were consistently increased in non-IBD samples and improved colitis in mouse models. To identify concordant changes in 15 IBD cohorts with 21 datasets, we used a novel multi-technology meta-analysis (MTMA) procedure and a curated reference database (StrainSelect). From these strains’ genomes and from metagenomic assemblies, the encoded bacterial peptides were synthesized and screened to find those which interact with the human immune system. Using phage panning, peptide libraries were assessed in a large scale screen using primary human cells to identify those that bind to our cells of interest such as T cells. Peptides that demonstrated binding were then screened in functional assays to identify bioactive peptides that influence cytokine production. Peptides with strong effects and cross-reacted to mouse cells also showed efficacy in mouse models of IBD. Identification of bioactive peptides demonstrates that a workflow in which i) a standardized computational pipeline is employed to identify peptides that are consistently decreased in IBD patients, ii) peptides are first screened by their ability to bind to cells of interest, and iii) peptides are then screened for functional activity, results in bioactive peptides with therapeutic potential.