The long term goals of this work are to understand the recognition process in beta, mu and kappa opioid receptors to determine the basis of receptor selectivity and the mechanism of signal transduction for the rational design of opioid ligands that have high selectivity. This proposal is directed towards the development of models to simulate the structure and function of the opioid receptors using both computational methodologies and experimental data derived from ligand binding and functional studies and site directed mutagenesis. The project is an interdisciplinary effort that combines molecular biology, medicinal chemistry and biophysics to address several challenging goals. These include modeling the structure and transmembrane domains of the receptors, the determination of the conformation of the loops and terminal domains, and the prediction of ligand binding modes and function using a combination of modeling, point mutations and chimeras of the receptor. The models built are initially formulated from the most recent structural information on G-protein coupled receptors using interactive computer graphics and predictive techniques based on homology. The structures are refined using a number of new methods and techniques developed specifically for receptor modeling. In particular, lipid-protein interactions have been added to the model building methods to account for hydrophobic effects on the packing of the helices in the transmembrane domain. The new methodology embeds the helices in an explicit lipid bilayer to simulate the membrane environment In model building software. Dynamics simulations are also developed to account for this effect as well. Two innovative approaches are described that combine Langevin dynamics with force fields that include lipid """"""""solvation"""""""" terms to produce accurate physical properties during simulations. To model the loops and terminal domains, aqueous solvation effects have been added to the conformational search strategy. The methodology developed reduces the computational problem through loop-closure constraints based on helix-helix packing distances in the transmembrane domain. The resulting model is applied for ligand docking studies of opiate-derived structures. Dynamics simulations are performed in combination with site directed mutagenesis studies to determine the specific amino acid residues involved in binding ligands in the cavity and extracellular domains of the opioid receptors. Based on these results, simulations are performed to differentiate the functional role of agonists and antagonists in signal transduction.
