Molecular Dynamical Study On HIV-1 Enzyme Systems With Allosteric Inhibitor

AIDS is caused by human immunodeficiency virus(HIV). Currently, the type of HIV-1 is the major epidemic in the world. People infected with HIV-1 would loss immune function and life is threatened. The copy of HIV-1 is a multi-stage process. Every step is the key to successful replication, so each step becomes a potential target for antiretroviral drugs. Reverse transcriptase(RT) and protease(PR) play the key roles in the replication process of HIV-1 and are the important target enzymes of inventing the new anti-HIV/AIDS drug.However, the long-term medication has made HIV-1 evolve into variation forms to resist the existing drugs. In order to improve the therapeutic effect for a long term, especially for those who are infected with multidrug resistance(MDR) mutant strains, new strategies are required. A strategy is to introduce allosteric inhibitors that bind to inactive sites, and by slightly changing the PR conformation, to make the substrate and acceptor bind more tightly. The advantage of allosteric inhibitor is able to avoid selective pressure for either the PR active site or the substrate to mutate. Thus, allosteric inhibitor is of a potential weapon to rescue the effectiveness of the current PR drugs and is a new strategy in the drug designing, and also a hot spot in the current subjects.RT is a dimmer and composed of two subunits, p66 and p51. It can catalyzes viral RNA genome and make it take place reverse transcript into a precursor of double-stranded DNA virus. Among the anti-AIDS drugs targeting RT, non nucleoside RT inhibitors(NNRTIs) which combine with inactive site in RT, so called allosteric inhibitors, having the characteristics of broad-spectrum resistance to HIV-1, is regarded as a kind of promising specific drug. Although there are many crystal data about the states of the RT, knowledge about and detailed understanding of the action mechanism of NNRTIs on RT is still lacking.Protease is a homogeneous dimmer and composed of two peptide chains, and functions to transform the polypeptide chain into the virus particles with infectivity. Currently, all FDA-approved AIDS drugs of PR target the active site. Perryman et al, through a systemic investigating of a fragment-based screen against wild-type PR, found two binding sites for small molecules, the exo-site and top-site. These results create a beginning point of developing larger and higher affinity molecules that bind to inactive sites, as a prospective allosteric inhibitor. However, how the allosteric fragments effect on the PR conformation and the binding free energy contributions to the active site inhibitor is still unknown.MD simulation method provides a powerful tool for understanding the properties and dynamics of biological macromolecules at atomic level. Nevertheless, the limitation of the method is also obvious. One question is about the sampling that simulation trajectories are unable to contain all useful molecular conformations within available timescales at the present computer level. Another is about the force field that the inaccuracy of the potential energy function may lead to incorrect conformations of biomolecules. In addition,the reliability of molecular dynamics simulation results also depends on the simulation methods and strategies, such as force field, time step length, simulation time and so on.The emergence of GPU parallel computing technology provides a more efficient platform for biological macromolecule simulation; the developing force field and suitable simulation strategy will help to improve the calculation results and get more reliable information.In this work, using GPU parallel computing technology, with ff12 SB force field and Amber12 package, we have carried out MD simulations on the two enzyme systems, RT and PR, i.e., RT+allosteric inhibitor and PR+allosteric molecules. And these simulations yield some important results.Firstly, in order to understand the allosteric modulation dynamics by NNRTIs, we performed 100 ns long molecular dynamics simulations on three RT systems, respectively. Analyses of influence of the EFV on conformation of the RT, flexibility of residues and dynamic behaviors of the systems were conducted. The simulations indicate that the EFV binding induces structural distortion of the RT, whereas configuration of the RT is more stable in dynamics along with a decreasing motion extent of residues; No conformational transition occurs throughout the whole simulations and the samples maintains their starting conformations, i.e., the free RT with closed conformation stays in the functional state and the EFV-bound RT remains open in conformation. However, the EFV-free RT with an initial open conformation exhibits an evident trend toward the close state. These results agree with the models from experiments, and present a useful insight into the allosteric inhibiting mechanism of NNRTIs. In addition, the simulation methodology has been contrastively discussed and will be of significance for simulation computing of large biological molecules.Secondly, the molecular dynamics simulations of 100 nanoseconds were performed on two systems of HIV-1 protease and inhibitor, respectively, one of which 2FX is attached to and another is not. Both types of new force fields, ff99 SBildn and ff12 SB were used for each of the two systems in the simulations. By investigating the influence of 2FX on the protease, it is found that the binding of 2FX induces a conformational change of the protease. The dynamics behaviors of the two systems under the different force fields are found different. The RMSD, the B-factors of the two systems et al indicate that the 2FX bound in the protease can stabilize the system. The resulting free energy values of 2FX-system being generally of greater magnitudes, also suggest the allosteric fragment stabilizes the binding of inhibitor and PR. The present results, despite the discrepancies between those from the two different force fields, demonstrate that the allosteric fragment 2FX unambiguously makes the inhibitor and PR combine more firmly. In addition, it is also found that, with ff12 SB force field, the dynamics processes are more stable, the calculated binding free energies are more favorable, and the results are better consistent with the experiments than with ff99 SBildn.Finally, to gain insight into the influences of two allosteric molecules 1F1 and 4D9 on PR we performed molecular dynamics simulation on 3 systems of inhibitor-PR. The simulations demonstrate, although 1F1 and 4D9 bind to different sites of top- and exo-site, that they unambiguously cause structural changes of the PR and make the inhibitor(3TL)-PR systems more stably. In particular, with 4D9, the dynamics processes of 3TL-PR are more stable and affinities are more favorable than with 1F1. Moreover, the action mechanisms of 1F1 and 4D9 are different.The results of the study provide a theoretical support for designing the allosteric inhibitor and the larger, higher affinity molecules that bind to inactive sites, as a prospective allosteric inhibitor. In addition, it is hoped that the simulation methodology and strategy which has been contrastively discussed will be of significance for simulation computing in large biological molecules.