Abstract
Among several distinct mechanisms used by bacteria to circumvent antibiotic stress, a predominant form of resistance to ribosome-targeting compounds is the methylation of their ribosomal RNA (rRNA) binding sites. The acquisition of aminoglycoside-resistance methyltransferases that modify 16S rRNA nucleotides in the ribosome decoding center, for example, results in exceptionally high-level aminoglycoside resistance and poses a major threat to their future clinical utility. Here, we report the discovery of a first-in-class panel of small-molecule inhibitors that target a previously unexploited composite "Y-shaped" binding pocket that is unique to the 30S subunit (substrate)-bound form of the 16S rRNA (m(1)A1408) methyltransferase NpmA. This Y-shaped pocket, formed by the conserved S-adenosyl-l-methionine binding site and a channel in which A1408 is positioned for modification, was predicted by molecular dynamics simulations to be accessible and potentially druggable in the free enzyme. We therefore conducted high-throughput virtual screening of over 2 million compounds, followed by precision docking and chemoinformatics to select lead scaffolds for initial testing. Iterative experimental analysis and docking of analogs to top hits led to the discovery of three compounds with comparable NpmA inhibitory activity and other similar analogs unable to inhibit the enzyme. Structure-activity relationship analysis highlighted the importance of stereoselectivity, halogen-π interactions, and water-mediated binding. Our strategy provides a new model for methyltransferase inhibitor development, targeting conformationally adaptive and composite binding sites and could be applied to efforts to develop inhibitors of other clinically prevalent resistance determinants such as the aminoglycoside-resistance m(7)G1045 methyltransferases (e.g., RmtB).