Journal of the American Chemical Society · 2000 · 128 citations · 33 references
PharmacyMedicinal ChemistryPharmacologyComputational ChemistryMolecular PharmacologyStructure-activity RelationshipsBinding Affinity PredictionComputational Drug DiscoveryLigand BindingMolecular DockingProtein-ligand DockingQuantitative Structure-activity RelationshipsDocking-based Drug DesignFree Energy PerturbationFree Energy Perturbation ModelingMolecular Docking Software DevelopmentProtein-ligand Complex Modeling
The origins of binding affinity and COX-2/COX-1 selectivity for analogues of celecoxib have been explored using an approach that combines docking with Monte Carlo (MC) simulations. These inhibitors are COX-2-selective nonsteroidal antiinflammatory drugs (NSAIDs) that are of current interest because the gastrointestinal irritation they cause is reduced compared to that caused by traditional NSAIDs. We report a novel docking method, based on a combined Tabu and Monte Carlo protocol, that determines starting conformations for MC simulations. Using the docking-predicted starting conformations, relative changes in binding free energies were computed for methyl, ethyl, hydroxymethyl, hydroxyl, thiomethyl, methoxy, trifluoromethyl, chloro, fluoro, and unsubstituted derivatives with the MC free energy perturbation (FEP) method. The computed free energies are in good accord with IC 50 values, and the structural information from the simulations can be used to explain the experimentally observed binding trends. In addition, the docking and FEP results have provided clarification of the binding conformation of the phenylsulfonamide moiety and the origin of COX-2/COX-1 selectivity. Namely, the COX-2 Val → COX-1 Ile subtitution is accompanied by an unfavorable conformational shift of the phenylsulfonamide ring.
33
Encyclopedia of Computational Chemistry
Anselm H. C. Horn, Harald Lanig · Journal of Molecular Modeling · 1999 · 2.5K citations
Natural Sciences, Chemistry, Physics +12