Exploring The Drug Resistance Mechanisms Of HCV NS3/4A Protease Inhibitors

The approval of direct-acting antiviral agents(DAAs) has marked a milestone in the treatment of the hepatitis C. However, the selection of the virus allows the HCV to mutate and then resistant to the drugs, and resulting in subsequent treatment failure. Therefore, exploring the drug resistance mechanism of HCV to the inhibitors is quite valuable for the development of the future anti-resistant inhibitors.In this thesis, we chose the HCV NS3/4A protease, an important drug target for the DAA treatment, as the research object. First, molecular dynamics(MD) simulations and MM/GBSA(Molecular Mechanics/Generalized Born Surface Area) approach were employed to investigate the interactions between the substrates(4A4B, 4B5 A, 5A5B) and the inhibitors(telaprevir, boceprevir, simeprevir, ITMN191 and BI201335) of the NS3/4A protease. By comparing the binding profiles between the substrates and inhibitors, we observed that most residues involved in drug resistance form stronger interactions with the inhibitors than with the substrates, and the critical resistant mutations impair more on the bindings of inhibitors than those of the substrates. Furthermore, the predictions illustrate that the natural substrates of the NS3/4A protease form balanced interactions with the strands 135-139 and 154-160 whereas the inhibitors cannot. Therefore, to overcome drug resistance, it may be necessary to restore the interaction balance between the two strands and the drug candidates.In the second part, MD simulations(MD), MM/GBSA free energy calculations and the umbrella sampling(US) simulations were used to explore the interaction profiles of MK5172 with the wild-type(WT) and four mutated(R155K, D168 A, D168 V and A156T) HCV NS3/4A proteases. The binding free energies(ΔGbind) predicted by MM/GBSA have good linear correlation with the potentials of mean force(PMFs) predicted by US, and they are both consistent with the experimental data. The US simulations demonstrate that the binding/dissociation pathways of MK5172 escaping from the binding pockets of the WT and mutants are quite different. The resistant mutants may change the protein-ligand interaction profiles of MK5172, resulting in the unbalanced energetic distribution amongst the catalytic triad, the strand 135-139 and the strand 154-160 and that leads to the pathway changes. We predict that MK5172 may be hard to get access to the correct binding sites of the mutated proteases after the pathway changing, thereafter leading to drug resistance. The levels of resistance conferred by the mutations are closely related to the change of the binding/dissociation pathways.