Prediction of binding for a kind of non-peptic HCVNS3 serine protease inhibitors from plants by molecular docking and MM-PBSA method

In the study, molecular dynamics simulations combined with MM-PBSA (Molecular Mechanics and Poisson-Boltzmann Surface Area) technique were applied to predict the binding mode of the polyphenol inhibitor in the binding pocket of the HCV NS3 serine protease for which the ligand-protein crystal structure is not available. The most favorable geometry of three candidates from molecular docking had a binding free energy about 3 and 6 kcal/mol more favorable than the other two candidates, respectively, and was identified as the correct binding mode. In the mode, the correlation of the calculated and experimental binding affinities of all five polyphenol compounds is satisfactory indicated by r(2) = 0.92. The most favorable binding mode suggests that two galloyl residues at 3 and 4 positions of the glucopyranose ring of the inhibitors interact with SER139, GLY137, ALA157, and ASP81 by hydrogen bond interaction and with ALA156 and HIE57 by hydrophobic interaction and are essential for the activities of the studied inhibitors. (c) 2006 Elsevier Ltd. All rights reserved.