Theoretical and computational modeling of amyloid aggregation

描述(申请人提供)：蛋白质聚集的实验研究使用的方案，相对于阿尔茨海默氏症等疾病的几十年的发病时间尺度，可以加速聚集形成许多数量级。体内和体外聚集时间尺度之间的差距需要聚集过程的详细理论，以便推断针对生理条件的实验观察。目前的理论工具不适合这项任务。分析理论目前缺乏预测序列扰动或环境条件所需的微观基础，而在电子计算方法中，即使是在体外聚集时间尺度也难以达到，而不牺牲必要的分辨率。申请人已经开发出一种关于纤维伸长的微观理论，该理论与关于 温度、变性剂和蛋白质浓度。这一理论认为，在氢键态上的构象搜索是聚集过程中最慢的步骤(这一观察结果与最近的模拟一致)，并表明这种搜索可以有效地模拟为在崎岖的一维势中的随机行走。拟议的工作将以两种方式扩展这一模型：1)通过开发新的淀粉样蛋白聚集的微观模型。这些模型将解决聚集途径中的关键步骤，包括成核和新分子结合时原纤端的模板效应。他们还将探索不同自结合模式之间的动力学竞争，包括低聚物和纤维。2)利用初步模型的启示，开发了一个多尺度的计算算法来模拟原子细节中的纤维生长和成核。在这 算法中，将使用大量的小型模拟来计算解析理论所预测的反应坐标空间中的系统扩散张量。然后，该扩散张量将被用于计算聚集过程的马尔可夫状态轨迹。这项工作的创新之处在于使用解析建模来推导微观动力学中目前无法通过实验或模拟解决的缓慢步骤，并利用这些见解来开发效率大大提高的模拟方法。结果将是解决小序列扰动的影响的能力，例如区分阿尔茨海默氏症相关多肽A的主要形式的两种氨基酸，以及允许合理操纵聚集途径并提供从体外实验推断生理后果的“动力学相图”。

DESCRIPTION (provided by applicant): Experimental studies of protein aggregation utilize protocols that accelerate aggregate formation by many orders of magnitude relative to the multi-decade timescales that characterize the onset of dis- eases like Alzheimer's. The gap between in vivo and in vitro aggregation timescales demands detailed theories of the aggregation process in order to extrapolate experimental observations toward physiological conditions. Current theoretical tools are not suitable for this task. Analytic theories presently lack the microscopic basis that is needed to make predictions about sequence perturbations or environmental conditions, and in silico methods struggle to reach even in vitro aggregation timescales without sacrificing necessary resolution. The applicants have developed a microscopic theory of fibril elongation that agrees with experiments with respect to the effects of temperature, denaturants, and protein concentration. This theory identifies the conformational search over H-bonding states as the slowest step in the aggregation process (an observation that is in agreement with recent simulations) and shows that this search can be efficiently modeled as a random walk in a rugged one-dimensional potential. The proposed work will expand on this model in two ways: 1) By developing new microscopic models of amyloid aggregation. These models will resolve key steps in the aggregation pathway including nucleation and the templating effect of the fibril end upon the binding of new molecules. They will also explore the kinetic competition between different modes of self-association, including oligomers and fibrils. 2) By using the insights from the preliminary model to develop a multi-scale computational algorithm to simulate fibril growth and nucleation in atomistic detail. In this algorithm, a large number of small simulations will be used to compute the system diffusion tensor in the reaction coordinate space predicted by the analytic theory. This diffusion tensor wil then be used to compute Markov state trajectories of the aggregation process. The innovation of this work is to use analytic modeling to deduce slow steps in the microscopic kinetics that are not presently resolvable by experiments or simulation, and to use these insights to develop simulation methods with greatly improved efficiency. The outcome will be the ability to resolve the effects of small sequence perturbations, such as the two amino acids differentiating the major forms of the Alzheimer's-related peptide A, as well as a "kinetic phase diagram" that will allow the rational manipulation of aggregation pathways and provide a means to infer physiological consequences from in vitro experiments.