Identification of Novel Cannabinoid CB1 Receptor Antagonists by Using Virtual Screening with a Pharmacophore Model

Abstract

CB1 receptor antagonists have proven to be clinically effective in treating obesity and related disorders. We report here the identification of a novel class of azetidinone CB1 antagonists by using virtual screening methods. For this purpose, we developed a pharmacophore model based on known representative CB1 antagonists and employed it to screen a database of about a half million Schering-Plough compounds. We applied a stepwise filtering protocol based on molecular weight, compound availability, and a modified rule-of-five to reduce the number of hits. We then combined Bayesian modeling and clustering techniques to select a final set of 420 compounds for in vitro testing. Five compounds were found to have >50% inhibition at 100 nM in a CB1 competitive binding assay and were further characterized by using both CB1 and CB2 assays. The most potent compound has a CB1 K i of 53 nM and >5-fold selectivity against the CB2 receptor.