A QUANTUM MECHANICAL APPROACH FOR EXPLORING HIV DRUG RESISTANCE

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The majority of drugs that are effective against HIV infection interfere with viral reverse transcriptase (RT). These drugs include nucleoside reverse transcriptase inhibitors (NRTI) that directly interfere with the polymerase catalytic site in RT and non-nucleoside reverse transcriptase inhibitors (NNRTI) that influence polymerase activity through an allosteric mechanism [1]. Recently drugs that inhibit the RNA removal function (RNH) of RT without affecting polymerase activity have also been discovered. Unfortunately, drug resistance develops rapidly to all these agents due to the high mutation rate of the HIV virus. Residue changes may eliminate favorable binding interactions or they may block drug access through steric effects. They may also interfere with flexibility preventing "induced fits" at the binding site or they may alter allosteric effects. A mathematical model that could predict and quantify the local and remote effects of mutations on drug binding and catalytic activity could lead to new strategies for combating drug resistance. Pilot studies on HIV-1 RT bound to the inhibitor dihydroxy benzoyl napthyl hydrazone (DHBNH) indicate this is feasible. The basic approach involves the application of quantum mechanical (QM) calculations to analyze selected regions of interest (QMROI). The main idea is to create a "quantum mechanical laboratory" that can be perturbed in silico to model the effects of mutations on drug binding and catalytic sites. Previous attempt to use QM for this purpose have treated drug binding as the sum of the interactions between drugs and isolated amino acid residues [2]. The QMROI approach seeks to create a more realistic local binding environment with complete polypeptide chains. Such an environment has a better chance for identifying the conformational changes leading to drug resistance. The QMROI is centered on the binding site and includes the bound drug and all residues containing atoms within 9 ¿ of the center of the site. Residues are added as necessary to create a set of short continuous polypeptide chains defining the binding site. The ends of these chains are capped with hydrogens to saturate the open valences. This is accomplished on the N-terminus by mutating the amino nitrogen to hydrogen. On the C-terminus, the carbonyl carbon is mutated to hydrogen. The positions of these hydrogen "cap" atoms are fixed during geometric optimization to lock in the conformational state imposed on the QMROI by the surrounding protein. All remaining atoms in the QMROI are unconstrained. The electrostatic effect of the surrounding protein is simulated by optimizing at set of point charges distributed on a surface surrounding the QMROI. In the case of RT, the QMROI contains ~400-500 atoms. This QMROI is large enough to include all the atoms in the bound drug and all the residues with polarizable atoms that are close enough to influence the drug binding site. It is also large enough to capture the highly conserved tyrosine-methionine-aspartate-aspartate (YMDD) motif in the polymerase catalytic site. The geometry of each QMROI structure is determined by numerical solution of its molecular wavefunction at a density function theory (DFT) level (b3lyp/6-31g(d,p)) of QM theory [3]. All calculations are carried out using the Gaussian'03" suite of programs. Binding energies are determined by applying frequency and single point energy studies to the drug and protein components of the optimized QMROI. The binding energy is calculated as the difference between the total energy of the protein with bound drug and the total energies of the protein and drug by themselves. Frequency calculations are carried out to obtain zero point energy corrections and thermodynamic functions. The effects of mutations on drug binding are studied by replacing the residue sidechains in silico followed by new QMROI calculations. The conformational states available to the QMROI atoms are simulated by varying the positions of the fixed hydrogen cap atoms that anchor the ends of the set of polypeptide chains that define the QMROI. The allowable variance in the pairwise positions between these fixed cap atoms is determined by the positional variation observed in different crystallographic structures, molecular dynamics (MD) simulations or coarse grained models such as the anisotropic elastic network model (ANM). The QMROI model provides a means for determining the effect of any mutation on drug binding using electronic structure calculations. Measurement of the distortion created in key amino acid motifs in catalytic binding sites provides a measure of the "fitness" of a given mutant to carry out its catalytic function. Such distortions can be quantified in terms of atomic displacements, changes in the dihedral angles of peptide backbone atoms or alterations in hydrogen bonding patterns. The QMROI model represents the first quantum mechanical approach to the problem of HIV drug resistance that addresses drug binding energy, local and global conformational change and the electrostatic effect of the surrounding protein and solvent environment. The QMROI model provides quantitative information about the steric alterations in drug binding sites induced by mutations. In many cases, this information is not available through purely experimental approaches. Detailed information about geometric relationships in drug binding sites is essential for rational drug design. Even though the QMROI model is intense from the calculation standpoint, this approach is suitable for mass production using parallel processing in modern clusters. The experimental design for the initial phase of the project focuses on two regions of interest. The first is the binding site for the NNRTI inhibitor nevirapine (PDB 1vrt). The second is the binding site for the RNH inhibitor DHBNH (2i5j). Both of these binding sites are adjacent to the RT polymerase catalytic site and both binding sites have overlapping components. Binding in both instances also involves an "induced fit". More importantly, the QMROI regions both overlap the critical YMDD motif in the polymerase catalytic site. The geometry of each QMROI will initially be optimized with no mutations. Two conformational states defined by the position of the fixed cap atoms in the QMROI will be studied for both drugs. These states will represent the maximum and minimum pairwise separation between fixed atoms estimated from a survey of the available crystallographic structures in the protein data bank (PDB). The YMDD motif between the two states will be compared. If distortion of this motif is the basis for NNRTI inhibition, it should be at a maximum in the NNRTI set and absent or minimal in the RNH set. Baseline QMROI regions will also be studied for each binding site without the presence of the inhibitor drugs. This will be accomplished using the conformations available from a 25 ns MD simulation of 2i5j in explicit water without DHBNH. This simulation was carried out as part of the pilot studies exploring the feasibility of this approach. When analysis of the binding energies and YMDD distortion is complete for both drugs and both baseline regions, the complete set will be restudied with seven different point mutations. The mutations will be selected from the list of mutations that are known to confer nevirapine resistance. Mutations conferring DHBNH resistance have not yet been identified. The mutations considered will be L100I, K103N, V106A, V108I, Y181C, Y188H and G190S [1]. Binding energies, geometric alterations and changes in the critical YMDD motif calculated for each mutant will be compared with the corresponding parameters calculated for the wild type. The geometric alterations in the YMDD motif in the bound and unbound states will also be analyzed to determine the influence of drug binding on the polymerase catalytic site. Analysis will provide insight into the mechanism of resistance conferred by each mutation. More importantly, it will provide geometric information about the drug and the binding site that can be used for the rational design of drug analogs. The second phase of the project will combine in silico QMROI studies with experimental approaches. This will be accomplished through collaborations with the laboratory of Michael Parniak at the University of Pittsburgh. This laboratory is focused on the development of new RT inhibitors that target the RNH site [4]. The QMROI approach will be used to study the effects of potential new inhibitory compounds, to guide the design of such compounds and to judge the potential effects of mutations that have not yet been observed. Such studies will also be used to guide site-directed mutagenesis studies of HIV-1 drug resistance. 1. Ilina T, Parniak MA: Inhibitors of HIV-1 Reverse Transcriptase. Advances in Pharmacology, 56:121-167, 2008. 2. He X, Mei Y, Xiang Y, Zhang DW, Zhang JZ: Quantum Computational Analysis for Drug Resistance of HIV-1 Reverse Transcriptase to Nevirapine through Point Mutations. Proteins: Structure, Function and Bioinformatics, 61:423-432, 2005. 3. Kohn W, Sham LJ: Quantum Density Oscillations in an Inhomogeneous Electron Gas. Phys. Rev., 137(6A):1697- 1705, 1965. 4. Himmel DM, Sarafinos SG, Dharmasina S, Parniak MA, et al: HIV-1 Reverse Transcriptase Structure with RNase H Inhibitor Dihydroxy Benzoyl Naphthyl Hydrazone Bound at a Novel Site. ACS Chemical Biology, 1:702-711, 2006.