Ligand‐based pharmacophore modeling of 5‐HT2A receptor biased agonism

Abstract

The central dogma of pharmacology posits that ligands that function as agonists of a protein receptor will activate all effectors coupled to that receptor in equal proportions. A new conceptualization of ligand-receptor interactions has emerged, showing that some receptor agonists can differentially activate distinct effector systems. A single ligand can therefore function as a full agonist of one effector while being a partial agonist, neutral antagonist, or inverse agonist of another effector. Although this new conceptualization has profound implications for drug function, this area have been slowed by our poor understanding of the structural properties of ligands that determine biased activation of distinct effector systems. In these studies, we studied biased agonism of serotonin 2A (5-HT2A) receptors because there are several known 5-HT2A receptor biased agonists, extensive work has determined the effector systems activated by 5-HT2A receptor biased agonists, the downstream signaling cascades triggered by 5-HT2A receptor biased agonists are known, and the in vivo consequences of biased 5-HT2A receptor agonism are well characterized and profound. We used ligand-based computational approaches, including molecular modeling tools for pharmacophore modeling (Phase), docking (Glide), property prediction (QikProp), and QSAR z95-8. The results of these models were tested using in vitro and in vivo approaches. Our efforts have revealed distinct molecular interactions that are important determinants of biased agonism of 5-HT2A receptors and have successfully predicted several compounds that bind to these receptors.