A computational study on cannabinoid receptors and potent bioactive cannabinoid ligands: homology modeling, docking, de novo drug design and molecular dynamics analysis.

When X-ray structure of a ligand-bound receptor is not available, homology models of the protein of interest can be used to obtain the ligand-binding cavities. The steroelectronic properties of these cavities are directly related to the performed molecular model coordinates. Thus, the use of different template structures for homology may result in variation of ligand-binding modes. We have recently reported the MD simulations of a potent CB ligand at bovine rhodopsin-based CB1 and CB2 receptors (Durdagi et al., Bioorg Med Chem 16:7377-7387, 2008). In this present study, a homology modeling study based on the beta2-adrenergic receptor for both CB1 and CB2 receptors was performed, and the results were compared with rhodopsin-based models. In addition, the role of membrane bilayers to the adopted conformations of potent AMG3 CB ligand has been analyzed for receptor-free and membrane-associated receptor systems. The performed MD trajectory analysis results have shown that gauche conformations at the terminal segment of the alkyl side chain of AMG3 are not favored in solution. Different adopting dihedral angles defined between aromatic and dithiolane rings at the active sites of the CB1 and CB2 receptors, which are adapted lead to different alkyl side chain orientations and thus, may give clues to the medicinal chemists to synthesize more selective CB ligands. The binding sites of receptors derived by rhodopsin-based models have been regenerated using the beta2-adrenergic based template receptors. The re-obtained models confirmed the ligand-binding pockets that were derived based on rhodopsin.