Homology modeling in tandem with 3D-QSAR analyses: a computational approach to depict the agonist binding site of the human CB2 receptor.

CB2 receptor belongs to the large family of G-protein coupled receptors (GPCRs) controlling a wide variety of signal transduction. The recent crystallographic determination of human beta 2 adrenoreceptor and its high sequence similarity with human CB2 receptor (hCB2) prompted us to compute a theoretical model of hCB2 based also on beta 2 adrenoreceptor coordinates. This model has been employed to perform docking and molecular dynamic simulations on WIN-55,212-2 (CB2 agonist commonly used in binding experiments), in order to identify the putative CB2 receptor agonist binding site, followed by molecular docking studies on a series of indol-3-yl-tetramethylcyclopropyl ketone derivatives, a novel class of potent CB2 agonists. Successively, docking-based Comparative Molecular Fields Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) studies were also performed. The CoMSIA model resulted to be the more predictive, showing r(ncv)(2) = 0.96, r(cv)(2) = 0.713, SEE = 0.193, F = 125.223, and r(pred)(2) = 0.78. The obtained 3D-QSAR models allowed us to derive more complete guidelines for the design of new analogues with improved potency so as to synthesize new indoles showing high CB2 affinity.