Enzyme Isoselective Inhibition ? a Novel Computational Approach to Drug Design

描述（由申请人提供）：该项目的长期目标是开发和实施一种计算方法，用于虚拟筛选和设计新的共价过渡态（TS）模拟酶抑制剂。传统的计算机辅助药物设计（CADD）方法侧重于酶抑制剂非共价识别相互作用的优化。不幸的是，由于传染病和癌症的突变耐药病例的发展，这种体外识别非共价抑制剂的良好优化在体内迅速丧失。酶的催化残基不受突变的影响，这一事实使我们假设那些共价TS类似物抑制剂，通过与催化残基的相互作用具有显著的结合贡献，应该受到较少的突变抗性。因此，一个计算工具来处理和设计共价TS模拟抑制剂是非常重要的。我们考虑了一类特殊的等选择过渡态模拟抑制剂-具有相同的识别位点（RS）和不同的化学位点（CS）。提出的项目是基于一个假设，即通过一个高级量子力学模型-酶抑制剂趋势分析（EITA），可以预测一系列此类抑制剂与目标酶的结合亲和力趋势。这个假设是基于我们早期的研究，分析了各种因素对酶抑制剂复合物稳定性的能量贡献。我们以前在酶催化和抑制机制的研究中应用了跨学科的计算/实验方法。模型包括相关的化学反应中心和一个考虑蛋白质/水环境效应的特殊方案。利用实验数据对计算模型进行了校正，得到了真实、准确的结果。基于这些发现，我们建议开发一种设计新的可逆共价酶抑制剂的方法。我们的研究将集中在水解酶超家族酶的抑制剂上。首先，我们将用各种医学上重要的酶来检验和验证所提出的模型。该模型将通过高阶量子力学DFT计算来解释酶-抑制剂共价键和溶剂-抑制剂竞争反应，并将结果与实验动力学数据相关联。然后，我们将通过考虑CS的非共价相互作用来扩大筛选能力。研究结果将以数据库的形式呈现，可用作信息源或药物设计工具。在我们未来的项目中，成熟的算法和数据库最终将被整合到一个软件包中，强调简单明了，以便于实际药物设计的日常使用。该方法可用于设计新的先导化合物，以及开发不会因靶酶突变而失去活性的药物。公共卫生相关性：该项目旨在验证并进一步开发一种计算工具，以预测结合亲和力，并对新的共价过渡态（TS）类似酶抑制剂进行虚拟筛选。目前的项目将产生一个公开可用的算法和一个由各种参数组成的数据库，这些参数描述了医学上重要目标与其抑制剂之间的结合趋势。最终，我们的方法将被开发成一个软件包，以帮助研究人员设计可能不易受突变耐药性影响的新药。

DESCRIPTION (provided by applicant): The long-term goal of this project is to develop and implement a computational methodology for virtual screening and design of new covalent transition-state (TS) analog enzyme inhibitors. The conventional computer assisted drug design (CADD) methodologies focus on optimization of the enzyme-inhibitor noncovalent recognition interactions. Unfortunately, such a well optimized in vitro recognition of non-covalent inhibitors is rapidly lost in vivo due to the development of mutational drug resistance cases of infectious diseases and cancer. The fact that catalytic residues of enzymes are not subjected to mutations led us to hypothesize those covalent TS analog inhibitors, having significant binding contributions from interactions with catalytic residues, should suffer less from mutational resistance. Thus, a computational tool to handle and design covalent TS analog inhibitors is of major importance. We have considered a special class of isoselective transition state analog inhibitors a set with identical recognition site (RS) and different chemical site (CS). The proposed project is based on the hypothesis that the trend of binding affinity to a target enzyme in a series of such inhibitors can be predicted, by a high-level quantum mechanical model- Enzyme Inhibitor Trend Analysis (EITA). This hypothesis is based on our earlier studies that analyzed the energetic contribution of various factors to the stability of the enzyme-inhibitor complex. We have previously applied an interdisciplinary computational/experimental approach in the study of enzyme catalysis and inhibition mechanisms. The models included the relevant chemical reaction center and a special protocol accounting for the protein/water environmental effect. The calculated models were calibrated by experimental data, to provide realistic and accurate results. Based on these findings, we suggest developing a methodology for the design of new reversible covalent enzyme inhibitors. Our research will focus on inhibitors of enzymes of the hydrolases superfamily. First we will examine and validate the proposed model with various medicinally-important enzymes. The model will account for the enzyme-inhibitor covalent bond and for the solvent-inhibitor competing reaction by high-level quantum mechanical DFT calculations and the results will be correlated with experimental kinetic data. We will then expand the screening abilities by taking into account also non-covalent interactions of the CS. The results will be presented as a database that could be used as an information source or as a tool for drug design. The mature algorithm and the database will ultimately be incorporated, in our future project, into a software package, with emphasis on simplicity and clarity for routine use of practical drug design. The methodology can be beneficial in the design of new lead compounds, as well as in the development of drugs that will not lose their activity due to target enzyme mutations. PUBLIC HEALTH RELEVANCE: This project aims to validate and further develop a computational tool to predict binding affinities and to perform virtual screening of new covalent transition-state (TS) analog enzyme inhibitors. The current project will result in a publicly available algorithm and a database comprised of various parameters that characterize binding trends between medicinally important targets and their inhibitors. Ultimately our method will be developed into a software package that will assist researchers in the design of novel drugs that may be less susceptible to mutational drug resistance.