Grid Assisted Ensemble Molecular Dynamics Simulations of HIV-1 Proteases Reveal Novel Conformations of the Inhibitor Saquinavir

摘要

Drug resistant mutations have severely limited the success of HIV therapy. Here we provide insight into the molecular basis of drug resistance in HIV-1 protease with the inhibitor saquinavir. We employ protocols consisting of chained molecular dynamics simulations that allow preparation of desired mutants from an available wildtype structure. By conducting ensembles of molecular dynamics simulations we report differing frequencies of adoption of four stable conformations of the P2 subsite of saquinavir. The P2 subsite hydrogen bonds more frequently with the catalytic aspartic acid dyad in the wildtype, whilst preferring to bind with the flaps of the protease in three chosen mutants. Previously such simulations have been demanding to perform on computational grids due to the difficulty in tracking large numbers of simulations. Using the Application Hosting Environment, a lightweight grid middleware solution, we present a simple way to construct chained ensembles of simulations seamlessly across multiple grid resources.