Nature 578, 582–587 (2020)

Natural product gene clusters often encode resistance genes to protect the producing organism against its own bioactive compound, and these can also hint at the product’s mechanism of action. Focusing on these resistance determinants in glycopeptide antibiotic (GPA) biosynthetic gene clusters, Culp et al. used clusters lacking known GPA resistance genes as a guide to find compounds with potentially novel mechanisms of action. Two compounds encoded by such clusters, complestatin and corbomycin, exhibited activity against vancomycin-resistant Gram-positive bacteria, unlike typical GPAs. By examining their effects on various steps of peptidoglycan metabolism and genome sequencing of resistant mutants, the authors identified the peptidoglycan-remodeling autolysin enzymes as the target for both complestatin and corbomycin. Both compounds inhibit a range of autolysins through directly binding to their peptidoglycan substrate and were effective as topical antibacterials in a mouse MRSA skin infection model. These GPAs demonstrate the utility of focusing on genetic resistance determinants for antibiotic discovery through genome mining.

Credit: Nature