Computational Studies of Model Systems for Protein Unfolded States

描述（由申请人提供）：虽然以前认为蛋白质的未折叠状态在很大程度上是随机的，但特定结构的证据正在积累。即使在低种群中，这种类型的排序也可能直接影响蛋白质的稳定性、折叠机制和速率，以及与许多人类疾病有关的错误折叠和聚集。在大多数情况下，残留结构的弱性质和生理条件下错误折叠状态的低种群妨碍了使用实验技术直接和详细地表征这些集成。模拟已经开始为实验提供数据补充，因为它们可以在大多数实验无法达到的时间尺度上提供单分子分辨率的结构细节。然而，由于多种原因，模拟在许多重要情况下失败了。本提案概述了模拟算法和力场的持续发展，使其能够成功地应用于展开状态的研究，并通过建立合作获得的实验数据进行直接验证。重点放在构象取样上；该项目的一个主要组成部分涉及开发一种新的采样方法，该方法可以在显式包含溶剂的情况下改善集成数据的收敛性，同时大大降低了计算成本。模拟将在几个不同复杂程度的模型系统上进行。在每种情况下，特定的突变将探测相互作用，这些相互作用被假设参与决定天然褶皱的结构或稳定性。其他突变体探讨了我们在未展开状态下观察到的相互作用对稳定性的影响。这些研究将为这些重要的模型系统提供有用的见解，并将为我们的力场和采样方法的性能提供有价值和关键的反馈。研究的下一阶段包括对每个模型系统的未展开的集成进行表征，并对任何残余结构的性质产生实验可测试的假设。对于每个模型系统，最近的实验表明，在展开状态下可能存在残余结构，我们的模拟将为解释这些数据提供重要的模型。通过用丙氨酸取代柔性甘氨酸和刚性脯氨酸，进一步研究折叠中的展开态和熵的作用，研究折叠的展开态熵和自由能的影响。

DESCRIPTION (provided by applicant): While unfolded states of proteins were previously thought to be largely random, evidence of specific structure is accumulating. Even at low populations this type of ordering could have direct bearing on protein stability, folding mechanism, and rate, as well as the misfolding and aggregation that are implicated in many human disorders. In most cases, the weak nature of the residual structure and low populations of misfolded states under physiological conditions preclude direct and detailed characterization of these ensembles using experimental techniques. Simulations have begun to provide data complementary to experiments, since they can provide structural detail with single molecule resolution on a time scale inaccessible to most experiments. However, simulations fail in many important cases for multiple reasons. This proposal outlines continued development of simulation algorithms and force fields to enable successful application to the study of unfolded states, with direct validation against experimental data obtained through established collaborations. A strong emphasis is placed on conformational sampling; a major component of the project involves development of a novel sampling approach that provides improved convergence of ensemble data with explicit inclusion of solvent at a significantly reduced computational cost. Simulations will be performed on several model systems of varying complexity. In each case, specific mutations will probe interactions that are hypothesized to be involved in determining the structure or stability of the native fold. Other mutants probe the effect on stability of interactions that we have observed in the unfolded state. These studies will provide useful insight into these important model systems, and will also provide valuable and critical feedback on the performance of our force fields and sampling methods. The next phase of the research involves characterization of unfolded ensembles for each model system, and generation of experimentally testable hypotheses about the nature of any residual structure. For each of the model systems, recent experiments suggest that residual structure may exist in the unfolded state, and our simulations will provide important models for interpretation of this data. Additional studies of the unfolded state and the role of entropy in folding will be performed through replacement of flexible glycine and rigid proline with alanine, investigating the effect on unfolded state entropy and free energy of folding.