Rational, Computational Design of Novel anti-AIDS Drugs

描述(由申请人提供)： 这项提议旨在通过改进和扩展我们在合成和临床测试之前预测药物活性的计算方法来解决艾滋病毒-1逆转录酶(NNRTI)的非核苷抑制剂的耐药性问题。我们的目标是在我们过去的成就的基础上，使用蒙特卡罗(MC)模拟结合线性响应(LR)方法来预测对当前所有抗艾滋病药物耐药的不同形式RT(Y181C、Y188C、L1001、V106A、K101E、K103N和可能的双突变)的活性。混合量子力学/分子力学(QM/MM)对MC方法的改进和蛋白质骨架近似的例程将被纳入以提高预测的准确性。采用组合计算方法的一种新的、快速的初步筛选技术将用于在完全计算评估之前评估候选结构。拟议的方法学变化预计将显著改善对抑制剂和蛋白质骨架的治疗，如果计算方法要成功地准确预测新系统中的主-客体相互作用，这两个问题是至关重要的。总体目标是提高设计耐药形式RT的有效抑制剂的速度和准确性。新的方法将被用于与NCI-Frederick的合作者共同开发的一类新型NNRTI-BPBrs的新成员的设计。新提出的BPBI类似物将被合成并测试广谱活性，这些类似物对不同形式的RT具有良好的预测活性。虽然该提案的一个具体和极其重要的目标是发现针对艾滋病患者中发现的NNIRT耐药病毒的新的抑制剂，但这里开发的通用计算方法在准确描述涉及蛋白质和有机分子的所有主-客体相互作用中具有广泛的应用。

DESCRIPTION (provided by applicant): This proposal seeks to address the problem of drug resistance to nonnucleoside inhibitors of HIV-1 reverse transcriptase (NNRTIs) through the refinement and extension of our computational methodology for predicting drug activity prior to synthesis and clinical testing. The goal is to build upon our past accomplishments in the use of Monte Carlo (MC) simulations coupled with a linear response (LR) method to predict activity against variant forms of RT (Y181C, Y188C, L1001, V106A, K101 E, K103N, and possible double mutants) that are resistant to all current anti-AIDS drugs. Hybrid quantum mechanics/molecular mechanics (QM/MM) refinements to the MC methodology and routines for protein backbone approximation will be incorporated to improve the accuracy of predictions. A new, rapid preliminary screening technique employing a combinatorial computational approach will be used to evaluate candidate structures prior to full computational evaluation. The proposed changes in methodology are anticipated to significantly improve the treatment of both the inhibitor and the protein backbone, problems of paramount importance if computational methods are to succeed in accurately predicting host-guest interactions in novel systems. The overall goal is to enhance the speed and accuracy of designing effective inhibitors of resistant forms of RT. The new methodology will be employed in the design of new members of a novel class of NNRTI's, BPBrs, developed in conjunction with collaborators at NCI-Frederick. Newly proposed BPBI analogs showing good predicted activity against variant forms of RT will be synthesized and tested for broad spectrum activity. While a specific and extremely important objective of the proposal is to discover new inhibitors targeted toward NNIRT resistant viruses found in AIDS patients, the general computational methodology developed herein has broad applications to the accurate description of binding in all host-guest interactions that involve proteins and organic molecules.