Theoretical Studies of Antibody-Antigen Binding

这项拟议的研究旨在了解特异性抗体-抗原结合的物理和化学基础。将研究的实验体系是3个抗体/蛋白质复合体和4个抗体/半抗原复合体，它们的高分辨X射线结构和结合数据，以及氨基酸突变和半抗原类似物结合差异的数据。重点将集中在两个关键方面：1)溶剂化的作用，它包括静电和疏水相互作用。2)缔合熵效应，指的是缔合时平移/转动熵的损失，伴随着络合物振动熵的增加，以及参与分子间接触的基团的构象迁移率的变化。将使用理论方法的组合，目的是开发在计算上可行的但定量的方法来计算结合能差和绝对结合能。这将使特定的生物学重要性问题得以解决，包括确定结合的总体驱动力、对特异性的贡献、静电互补的作用、诱导适配是否重要、表面基团的内在抗原性以及氨基酸突变和取代对结合能的影响。静电学将使用具有分子的原子细节表示的溶剂的连续统处理，使用有限差分泊松-玻尔兹曼(FDPB)方法。动态方面将由FDPB方法和分子力学(FDPB/MD)相结合的方法来处理。疏水相互作用将使用表面自由能关系式来处理，修正后的关系式将考虑形状效应，并根据小分子溶剂转移数据进行校准。平移和转动熵效应将从结合时转动和平移配分函数的变化中估计出来。生物体产生的分子能与某些“靶”分子或“靶”分子的某些部分特异结合，而不能与其他分子结合，从而在分子水平上产生生物识别现象。这种分子识别过程是许多基本生物学过程的基础，包括催化、基因转录和免疫反应。药物的设计还包括创造或修饰分子，以识别给定的生物靶分子。从物理角度讲，识别之所以发生，是因为一个分子与其“靶子”的结合比与其他分子的结合更紧密。众所周知，结合的紧密程度取决于当两个分子结合在一起时释放的能量(结合能)。人们还知道，对结合能的贡献来自两个分子之间的相互作用，每个分子与其周围环境，特别是水的相互作用，以及每个分子在结合时形状和流动性的变化。然而，即使知道这两个分子的结构，也不可能准确地计算结合能。这既阻碍了我们对识别所必需的分子性质的理解，也阻碍了我们设计识别给定目标的分子的能力。抗体对外来抗原的识别是免疫系统的关键特性之一，也是分子识别研究最多的例子之一：至少有7个抗原/抗体复合体的结构在原子水平上是已知的，并测量了它们的结合能。因此，我们选择这些体系作为结合的详细理论研究对象。这项拟议研究的目的是将最近开发的模拟分子行为的方法应用于计算抗体-抗原结合能的不同贡献的问题。目标是确定对紧密结合和特异性至关重要的抗体的性质，并提高计算结合能的能力。

The proposed research is directed at understanding the physical and chemical basis of specific antibody-antigen binding. The experimental systems to be studied are three antibody/protein complexes and four antibody/hapten complexes for which high resolution X-ray structures and binding data re available, plus data on binding differences for point amino acid mutations and hapten analogue binding. The focus will be on two key aspects: 1) The role of solvation, which involves both electrostatic and hydrophobic interactions. 2)'Association entropy' effects, meaning the loss of translational/rotational entropy upon association, with the concomitant gain in vibrational entropy of the complex, and the change in conformational mobility of groups involved in the intermolecular contact. A combination of theoretical approaches will be used, the aim being to develop computationally feasible yet quantitative methods for calculating both differences in binding energy, and absolute binding energies. This will enable specific questions of biological importance to be addressed, including identifying the overall driving force for binding, contributions to specificity, the role of electrostatic complementarily, whether induced fitting is important, the intrinsic antigenicity of surface groups, and the effect of amino acids mutations and substitutions on binding energies. Electrostatics will be treated using a continuum treatment of solvent with an atomic detail representation of the molecule, using the Finite Difference Poisson-Boltzmann (FDPB) method. Dynamic aspects will be handled by a method which combines the FDPB method with molecular mechanics (FDPB/MD). Hydrophobic interactions will be treated using surface free energy relationships, modified to account for shape effects, calibrated on small molecule solvent transfer data. Translational and rotational entropies effects will be estimated from changes in the rotational and translational partition function upon binding. %%% The ability of an organism to produce molecules that bind specifically to certain 'target' molecules or parts of 'target' molecules, but not to other molecules, gives rise to the phenomenon of biological recognition at the molecular level. This process of molecular recognition underlies many fundamental biological processes including catalysis, gene transcription and the immune response. The design of drugs also involves creating or modifying molecules to recognize given biological target molecules. In physical terms recognition occurs because a molecule binds more tightly to its 'target' than to other molecules. It is known that the tightness of binding is determined by how much energy is released (the binding energy) when two molecules are brought together. it is also known that contributions to the binding energy come from the interaction of the two molecules with each other, the interaction of each molecule with its surroundings, especially water, and from the change in shape and mobility of each molecule upon binding. However it is not yet possible to calculate the binding energy accurately, even if the structure o the two molecules is known. This impedes both our understanding of what properties of molecules are necessary for recognition and the ability to design molecules to recognize a given target. The recognition of foreign antigens by antibodies is one of the key properties of an immune system, and one of the most studied examples of molecular recognition: The structures of at least seven antigen/antibody complexes are known at the atomic level, and their binding energies have been measured. Therefore these systems have been chosen for a detailed theoretical study of binding. The aim of the proposed research is to apply recently developed methods for simulating the behavior of molecules to the problem of calculating the different contributions to the antibody-antigen binding energy. The goal is to identify the properties of the antibodies important for tight binding and specificity, and to improve the ability to calculate binding energies.