Computer Simulations of Protein Aggregation

描述（由申请人提供）：正常可溶性蛋白的异常组装成有序聚集体，称为淀粉样原纤维，是许多不同人类疾病的原因或相关症状，包括阿尔茨海默病，帕金森病和亨廷顿病，朊病毒疾病和成人发病糖尿病。这些疾病统称为淀粉样病变，其特征是特定蛋白质缓慢沉积到淀粉样原纤维中，然后积聚成斑块，破坏受影响组织的功能，通常导致退行性并最终致命的后果。了解导致蛋白质聚集成淀粉样蛋白的分子水平机制对于发现潜在的治疗策略和诊断至关重要。作为美国国立卫生研究院资助的揭示控制蛋白质聚集的一般物理原理的努力的一部分，我们开发了一个中等分辨率的蛋白质模型，允许在快速工作站上使用不连续分子动力学（DMD）在几天内模拟多蛋白质系统。最近的一项突破使我们能够模拟48个16残基丙氨酸肽从随机线圈状态开始组装成纤维状结构。这表明中分辨率模型可以作为一种计算工具来探索短肽中的原纤维形成。该项目有三个具体目标：(1)通过使用DMD模拟含有聚丙氨酸链的多蛋白质系统，通过我们的中分辨率模型来学习控制蛋白质纤维形成的基本物理原理；(2)阐明负责聚谷氨酰胺聚集的分子水平机制，聚谷氨酰胺的纤维化与亨廷顿病有关；通过将中分辨率模型扩展到聚谷氨酰胺侧链的处理，然后进行DMD模拟；(3)通过将中分辨率模型扩展到所有20个残基的粗粒度表示，研究含有特定淀粉样蛋白生成肽的多蛋白系统的聚集和可能的纤化，特别是阿尔茨海默病肽Abeta（1-40）和Abeta(1-42)，进行DMD模拟；并与实验结果进行了比较。这项工作将最终形成细纤维化过程的分子水平图像，从而为直接参与开发治疗策略或抑制剂的医学研究工作者提供见解，以规避纤维化过程中最导致细胞损伤的那些步骤。

DESCRIPTION (provided by applicant): The aberrant assembly of normally soluble proteins into ordered aggregates, called amyloid fibrils, is a cause or associated symptom of many different human disorders including Alzheimer's, Parkinson's and Huntington's diseases, the prion diseases and adult-onset diabetes. Known collectively as the amyloidoses, these diseases are characterized by the slow deposition of a specific protein into amyloid fibrils, which then accumulate into plaques, destroying the function of the affected tissue, usually with degenerative and ultimately fatal consequences. An understanding of the molecular-level mechanisms that result in the aggregation of proteins into amyloid is essential for the discovery of potential therapeutic strategies and diagnostics. As-part of our NIH-sponsored effort to uncover the general physical principles that govern protein aggregation, we developed an intermediate-resolution protein model that allowed the simulation using discontinuous molecular dynamics (DMD) of multi-protein systems within days on a fast workstation. A recent breakthrough enabled us to simulate assembly of 48 16-residue alanine peptides into a fibrillar structure starting from the random coil state. This suggests that the intermediate-resolution model could be used as a computational tool to explore fibril formation in short peptides. The project has three specific aims: (1) to learn the basic physical principles governing protein fibril formation by using DMD to simulate multi-protein systems containing polyalanine chains modeled using our intermediate-resolution model, (2) to shed light on the molecular-level mechanisms responsible for the aggregation of polyglutamine, the protein whose fibrillization is linked to Huntington's disease, by extending the intermediate-resolution model to the treatment of polyglutamine side chains and then performing DMD simulations, and (3) to investigate the aggregation and possible fibrillization of multi-protein systems containing specific amyloidogenic peptides, particularly the Alzheimer's peptides Abeta(1-40) and Abeta(1-42), by extending the intermediate-resolution model to a coarse-grained representation of all 20 residues, performing DMD simulations, and comparing our results with experiment. This work should culminate in a detailed molecular-level picture of the fibrillization process, thereby providing insights to guide medical research workers directly involved in developing therapeutic strategies or inhibitors to circumvent those steps in the fibrillization process that are most responsible for cell damage.