The overall goal of this project is to develop a selective antibiotic against Clostridium difficile, the major agent responsible for antibiotic-induced diarrhea and colitis. The pathogen can be carried in healthy individuals, and is kept at bay by gut symbionts. Antibiotic treatment suppresses the normal flora, but spores of C. difficile survive, germinate and cause disease. Treatment with metronidazole, vancomycin, or fidaxomycin kills not only the pathogen, but also the symbionts, which may result in relapse. Ideally, one would like to have a therapeutic which is selective against C. difficile, as it would allow the normal flora to restore in the course of treating the pathogen, preventing relapse. Based on genomic studies, bacteria share a core of ~200 genes, and in addition to those, there may be up to 100-200 essential genes specific to a given species/genus. We reasoned that the existence of a large number of specific targets presents an opportunity to discover antimicrobials acting selectively against C. difficile. The pathogen is an anaerobe, while HTS normally requires aerobic conditions. As a result, current therapeutics for treating the infection come from other programs, and are not selective. We developed an anaerobic HTS and performed a first direct screen against C. difficile. Given the historically high success rate of discovering antibiotics from natural products, we screened a library of extracts from soil bacteria. This library comes from an untapped resource, uncultured bacteria, and is rich in novel compounds. The library had been screened against S. aureus and E. coli, producing hit rates of 30% and 0.5%, respectively. We reasoned that the remaining "inactive" strains may harbor compounds active against particular species that were missed in the primary screen. A pilot screen of ~5,000 extracts resulted in a hit rate of 1.3% against C. difficile. Preliminary analysis of these extracts showed that two were selective against C. difficile when tested against a small panel of gut commensals, and contained compounds with novel masses. In the proposed project, we will follow up on these hits, and screen 50,000 additional extracts from the "inactive" library. Hits will be dereplicated by LC/MS, which will indicate the degree of novelty, and tested against representatives of the main groups of gut symbionts. Clostridium-selective compounds passing dereplication will be tested for potency, resistance frequency, cytotoxicity and penetration into intestinal epithelial cells. Compounds that are not absorbed and thus retained at the site of infection will be given priority. Structure determination will confirm novelty and provide information on the suitability of compounds for further development. Whole genome sequencing of resistant mutants will indicate the likely target. Maximum tolerated dose and bioavailability of leads will be determined in hamsters, which will inform the design of a C. difficile efficacy study. The goal of Phase I is to identify two to three lead compounds with efficacy in a hamster model of C. difficile infection. This will form a solid basis for a Phase II application aimed at preclinical development towards an IND.

Narrative C. difficile infection (CDI) has become a major cause of infectious diarrhea and colitis in healthcare settings and in the community. It is widely agreed that new antibiotics are needed to treat this disease, and the proposed study will address this need.