Résumé
Hydrolysis of β-lactam drugs, a major class of antibiotics, by serine or metallo-β-lactamases (SBL or MBL) is one of the main mechanisms for antibiotic resistance. New Delhi Metallo-β-lactamase-1 (NDM-1), an acquired metallo-carbapenemase first reported in 2009, is currently considered one of the most clinically relevant targets for the development of β-lactam-β-lactamase inhibitor combinations active on NDM-producing clinical isolates. Identification of scaffolds that could be further rationally pharmacomodulated to design new and efficient NDM-1 inhibitors is thus urgently needed. Fragment-based drug discovery (FBDD) has become of great interest for the development of new drugs for the past few years and combination of several FBDD strategies, such as virtual and NMR screening, can reduce the drawbacks of each of them independently. Our methodology starting from a high throughput virtual screening on NDM-1 of a large library (more than 700,000 compounds) allowed, after slicing the hit molecules into fragments, to build a targeted library. These hit fragments were included in an in-house untargeted library fragments that was screened by Saturation Transfer Difference (STD) Nuclear Magnetic Resonance (NMR). 37 fragments were finally identified and used to establish a pharmacophore. 10 molecules based on these hit fragments were synthesized to validate our strategy. Indenone 89 that combined two identified fragments shows an inhibitory activity on NDM-1 with a K i value of 4 µM.