Computational Studies of HIV-1 RT and Quinone Oxidoreductase -- Interaction, Activity, and Mechanism

HIV-1 reverse transcriptase (HIV-1 RT) plays an important role in the duplication of HIV-1. It has been a target in the development of drugs to inhibit HIV and cure AIDS. Non-nucleoside RT inhibitor (NNRTI) is one type of RT inhibitor, which binds in an allosteric site on p66 subdomain. Docking, free energy of binding (FEB), and MD simulations are used in the work to explore the binding structure, binding affinity and interaction of ligand/receptor of RT and QR1. Docking results show that it reproduces crystal structures well. The results of large set of other NNRTIs show that these NNRTIs dock in a similar position and orientation as known inhibitors. Also, hydrogen bonds around entrance of active site were observed and analyzed. The docking simulations using crystal structures gotten under different resolutions indicate that a coarse resolved crystal structure can be used in docking simulation. A FEB model was developed based on the docked structures of TIBOs using a linear response approach of Dock-MM-PB/GS, which has good correlation with the experimental activity. The promising approach in computer-aided rational drug design has good ability to estimate binding affinity of a large set of ligands within a reasonable computer time. Also 3-D QSAR models of TIBOs were developed using combination of ligand-based drug design approach - CoMFA and receptor-based -- docking by which the active conformation and alignment were performed. MD simulation and PCA were employed to explore the dynamic properties of an unliganded RT, RT/nevirapine and RT mutant/nevirapine complexes. It is shown that NNRTI affects dynamic properties of cleft formed by fingers and thumb through affecting some subdomains around active site in palm. The changes on cleft dynamics finally affect its ability to hold and synthesize DNA. The binding affinities calculated by MM-PBSA show that nevirapine has stronger interaction with RT than RT mutant. Meanwhile it is suggested that a good inhibitor should be one that can effectively interact with these subdomains and maximum occupy the active site