Katrich SEVA

COMPUTATIONAL MODELING OF GPCRS: UNDERSTANDING RECEPTOR SIGNALING FOR BIOLOGY AND DRUG DISCOVERY

Katrich SEVA
University of Southern California

Recent advances in crystallography and cryo-EM have led to structure determination for more than fifty G-protein coupled receptors (GPCRs), covering most branches of this therapeutically important 7-transmembrane superfamily of more than 800 proteins. The structures provide a solid 3D framework for computational studies of GPCR molecular interactions with ligands and allosteric modulators, characterization of their functional states and complexes with downstream effectors, as well as a platform for structure-based ligand discovery. In this talk I will describe several new computational approaches to GPCR modelling and design developed in my lab. To predict stabilizing mutations in GPCRs we have developed a CompoMug tool as combination of sequence-based analysis, structural information, and a derived machine learning predictor1. Novel insights into the GPCR functional mechanisms can be also obtained by structural bioinformatics and molecular dynamics characterization of the protein and its allosteric co-factors, including highly conserved sodium ion and water cluster in the center of the 7TM bundle in Class A GPCRs2,3. Finally, structural pharmacology analysis of the orthosteric and allosteric ligand binding sites helps to explain selectivity and functional profiles of known drugs and drug candidates4,5. Applied to the key receptors in inflammation and pain perception pathways, including opioid, cannabinoid and angiotensin receptors, these approaches give a powerful tool for virtual screening and rational design of novel GPCR ligands with desirable functional features as molecular probes and lead candidates..