Discovery of antibacterial biotin carboxylase inhibitors by virtual screening and fragment-based approaches

Igor Mochalkin,J Richard Miller,Lakshmi Narasimhan,Venkataraman Thanabal,Paul Erdman,Philip B Cox,J V N Vara Prasad,Sandra Lightle,Michael D Huband,C Kendall Stover

DOI: https://doi.org/10.1021/cb9000102

2009-06-19

Abstract:As part of our effort to inhibit bacterial fatty acid biosynthesis through the recently validated target biotin carboxylase, we employed a unique combination of two emergent lead discovery strategies. We used both de novo fragment-based drug discovery and virtual screening, which employs 3D shape and electrostatic property similarity searching. We screened a collection of unbiased low-molecular-weight molecules and identified a structurally diverse collection of weak-binding but ligand-efficient fragments as potential building blocks for biotin carboxylase ATP-competitive inhibitors. Through iterative cycles of structure-based drug design relying on successive fragment costructures, we improved the potency of the initial hits by up to 3000-fold while maintaining their ligand-efficiency and desirable physicochemical properties. In one example, hit-expansion efforts resulted in a series of amino-oxazoles with antibacterial activity. These results successfully demonstrate that virtual screening approaches can substantially augment fragment-based screening approaches to identify novel antibacterial agents.