Understanding protein folding and function via molecular simulation

通过分子模拟了解蛋白质折叠和功能

• 批准号: 9357218
• 负责人: Robert Best
• 金额: $ 163.9万
• 依托单位: NATIONAL INSTITUTE OF DIABETES AND DIGESTIVE AND KIDNEY DISEASES
• 依托单位国家: 美国
• 资助国家: 美国
• 起止时间: 至
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/9357218
• 关键词: Active Sites; Address; Affinity; Amyloid fibers; Area; Base Pairing; Binding; Cereals; Chemicals; Collaborations; Computers; Cytoplasmic Granules; Data; Development; Diffusion; Disease; Dissociation; Enzymes; Equilibrium; Fiber; Fluorescence Resonance Energy Transfer; Free Energy; Friction; Gases; Generations; Goals; Height; Heterogeneity; Hydrogen Bonding; Hydrogenase; Investigation; Kinetics; London; Manuscripts; Membrane Lipids; Membrane Proteins; Methodology; Methods; Modeling; Molecular; Mutation; Oxygen; Peptides; Phase; Predisposition; Property; Proteins; Publishing; RNA; Resolution; Sampling; Small RNA; Structure; Study models; Surface; Testing; Time; Torsion; Universities; Viscosity; Water; Work; amyloid fibril formation; base; college; connectin; dimer; follow-up; improved; interest; models and simulation; mutant; novel; protein aggregation; protein folding; protein function; protein misfolding; research study; simulation; solute; stem; theories

The project has addressed the following areas in the past year: 1. Interpreting all atom folding simulations. It has recently become possible to fold proteins on a computer, but this in itself is insufficient to guarantee that the mechanism of folding is correct. However, there are very few experiments that can be used to test the mechanism. One which can is phi-value analysis. We have developed a method to compute phi-values from simulations and used to to interpret recent all-atom simulations of protein folding. We find that in most cases the computed phi-values are close to the experimental ones, but there are some discrepancies (Ref. 1). We show how heterogeneity in folding mechanism can be revealed by the sensitivity of the phi-value to the type of mutation made. A second line of investigation has been testing the assumptions made by folding theories. We had previously shown that the approximation that native contacts determine folding mechanism was valid in the context of the all-atom simulations. This year, we also showed that the dynamics can be simplified to one-dimensional diffusion along an appropriately chosen coordinate (Ref. 2). 2. Misfolding in multidomain proteins. We have developed a model for the susceptibility of different protein folds to intramolecular domain-swapped misfolding using coarse-grained simulations. The model can identify which proteins are known to misfold and which are known not to do so, the distinction between them being based mainly on the topology of the folded state (Ref 3). In collaboration with Ben Schuler (University of Zurich), we have also worked to interpret new time-resolved FRET experiments on misfolding in titin domains. Our coarse-grained simulations provided a plausible molecular mechanism for the intermediate states identified by FRET (Ref. 4). 3. Folding of membrane proteins. We are focussing our efforts on membrane protein folding using both coarse-grained and atomistic simulations, which is very challenging due to the high viscosity of lipid membranes. We find, using two-helix dimers as a testbed, that while current coarse-grained models yield reasonable values for dissociation constants, they often predict incorrect structures for the folded state. We have just completed a manuscript on this topic. We are also looking at all atom simulations, where the sampling problem is even more severe, but we have overcome this to a large extent by using a solute tempering replica exchange approach which we are now using to study the same problems at an all-atom resolution, with promising results. 4. Structure of unfolded and intrinsically disordered proteins in chemical denaturant. One of the outstanding controversies in protein folding concerns the effects of chemical denaturants on the structure of unfolded or disordered proteins, i.e. do chemical denaturants cause the unfolded chain to expand, or not? FRET and SAXS experiments had given different answers to this question. We have addressed the problem from two directions: first, a bottom-up approach in which a molecular simulation force field for the denaturant was carefully optimized (Ref 5). Simulations with this force field showed that increased denaturant concentration caused the protein to expand, but were also consistent with FRET and SAXS data from our collaborators Ben Schuler (University of Zurich) and Alex Grishaev (NIST) respectively (Ref 6). Thus, the two types of experiment are not necessarily in contradiction. We then addressed the inverse problem, determining structural data directly from the experiments, and showed that the apparent differences were essentially due to the models used in interpreting the data (Ref. 7). 5. A key unresolved aspect of protein folding dynamics is the contribution of interactions within the chain to slowing folding, known as 'internal friction'. Building on our previous results where we showed that a common origin for internal friction in all proteins is crossing of local torsional barriers in the energy landscape, we have developed a simple model to explore whether there is any contribution from the height of the global free energy barrier to the internal friction. We find that the internal friction is explained essentially by the local torsion barriers only (Ref. 8). We are now working on interpreting internal friction in unfolded and disordered chains. 6. In collaboration with David de Sancho (San Sebastian), and Jochen Blumberger (University College London) we have been investigating the diffusion of gas molecules to the active sites of hydrogenase enzymes. We have now applied the methodology to an FeFe hydrogenase, and combined the results with estimates of the rate for the chemical step from ab initio calculations to obtain a description of the overall kinetics. The results have been used to make mutants to alter the oxygen sensitivity of the enzyme, which have been tested by our experimental collaborator, Christophe Leger. (Ref. 9) 7. We are working in collaboration with Tuomas Knowles in Cambridge in order to describe the formation of amyloid fibers, and in particular the molecular mechanism of secondary nucleation. We are taking two-approaches. The first is to build a simple coarse-grained model based on the known structure of the fiber. The second is to characterise the affinity, association rate and binding mode of the peptides with the surface of an existing fiber, which can be checked against experimental data from the Knowles and Buell labs. 8. Force field development. Accurate energy functions, or force fields, are essential in order to obtain useful results from molecular simulations. In work published last year, we showed the critical importance of balancing protein-water interactions in force fields to obtain reliable results on the equilibrium properties of unfolded states. In follow-up work, we have shown that this balance is also critical for reproducing dynamics in the unfolded state (Ref. 10). In work with collaborators, we have looked at the suitability of current RNA force fields to fold small RNA tetraloops. We found that although the latest generation of RNA force fields is promising, there are significant shortcomings, most likely in the strength of hydrogen bonds between the base pairs in the stem (Ref. 11).

该项目在过去一年中解决了以下领域的问题： 1. 解释所有原子折叠模拟。最近，在计算机上折叠蛋白质已经成为可能，但这本身并不足以保证折叠机制的正确性。然而，可用于测试该机制的实验却很少。其中之一就是 phi 值分析。我们开发了一种通过模拟计算 phi 值的方法，并用于解释最近蛋白质折叠的全原子模拟。我们发现，在大多数情况下，计算出的 phi 值与实验值接近，但也存在一些差异（参考文献 1）。我们展示了如何通过 phi 值对突变类型的敏感性来揭示折叠机制的异质性。第二个研究方向是检验折叠理论所做的假设。我们之前已经证明，原生接触决定折叠机制的近似在全原子模拟的背景下是有效的。今年，我们还表明，动力学可以简化为沿着适当选择的坐标的一维扩散（参考文献 2）。 2. 多结构域蛋白的错误折叠。我们使用粗粒度模拟开发了一个模型，用于研究不同蛋白质折叠对分子内结构域交换错误折叠的敏感性。该模型可以识别哪些蛋白质已知会错误折叠，哪些蛋白质不会错误折叠，它们之间的区别主要基于折叠状态的拓扑结构（参考文献 3）。我们与 Ben Schuler（苏黎世大学）合作，还致力于解释有关 titin 结构域错误折叠的新时间分辨 FRET 实验。我们的粗粒度模拟为 FRET 识别的中间态提供了一个合理的分子机制（参考文献 4）。 3.膜蛋白的折叠。我们的工作重点是使用粗粒度和原子模拟进行膜蛋白折叠，由于脂质膜的高粘度，这非常具有挑战性。我们发现，使用双螺旋二聚体作为测试平台，虽然当前的粗粒度模型产生合理的解离常数值，但它们经常预测折叠状态的错误结构。我们刚刚完成了关于这个主题的手稿。我们也在研究所有原子模拟，其中采样问题更加严重，但我们通过使用溶质回火副本交换方法在很大程度上克服了这个问题，我们现在使用该方法在全原子分辨率下研究相同的问题，并取得了有希望的结果。 4. 化学变性剂中未折叠和本质上无序的蛋白质的结构。蛋白质折叠中的突出争议之一涉及化学变性剂对未折叠或无序蛋白质结构的影响，即化学变性剂是否会导致未折叠链扩展？ FRET和SAXS实验对这个问题给出了不同的答案。我们从两个方向解决了这个问题：首先，采用自下而上的方法，其中仔细优化了变性剂的分子模拟力场（参考文献 5）。该力场的模拟表明，变性剂浓度的增加导致蛋白质膨胀，但也与我们的合作者 Ben Schuler（苏黎世大学）和 Alex Grishaev（NIST）分别提供的 FRET 和 SAXS 数据一致（参考文献 6）。因此，这两种类型的实验并不一定矛盾。然后，我们解决了逆问题，直接从实验中确定结构数据，并表明明显的差异本质上是由于解释数据时使用的模型造成的（参考文献 7）。 5. 蛋白质折叠动力学的一个关键的未解决的方面是链内相互作用对减缓折叠的贡献，称为“内摩擦”。基于我们之前的结果，我们表明所有蛋白质内摩擦的共同根源是穿越能量景观中的局部扭转势垒，我们开发了一个简单的模型来探索全局自由能势垒的高度是否对内摩擦有任何贡献。我们发现内部摩擦基本上只能由局部扭转屏障来解释（参考文献 8）。我们现在正致力于解释展开且无序的链条中的内部摩擦。 6. 与 David de Sancho（圣塞巴斯蒂安）和 Jochen Blumberger（伦敦大学学院）合作，我们一直在研究气体分子向氢化酶活性位点的扩散。现在，我们已将该方法应用于 FeFe 氢化酶，并将结果与​​从头开始计算的化学步骤速率估计值相结合，以获得整体动力学的描述。研究结果已被用来制造突变体来改变酶的氧敏感性，我们的实验合作者 Christophe Leger 对此进行了测试。 （参考文献 9） 7. 我们正在与剑桥大学的 Tuomas Knowles 合作，描述淀粉样纤维的形成，特别是二次成核的分子机制。我们正在采取两种方法。首先是基于已知的纤维结构建立一个简单的粗粒度模型。第二个是表征肽与现有纤维表面的亲和力、关联率和结合模式，这可以根据 Knowles 和 Buell 实验室的实验数据进行检查。 8.力场发展。为了从分子模拟中获得有用的结果，准确的能量函数或力场至关重要。在去年发表的工作中，我们展示了平衡力场中蛋白质-水相互作用的至关重要性，以获得关于展开态平衡特性的可靠结果。在后续工作中，我们已经证明这种平衡对于在展开状态下再现动态也至关重要（参考文献 10）。在与合作者的合作中，我们研究了当前 RNA 力场折叠小 RNA 四环的适用性。我们发现，虽然最新一代的 RNA 力场很有前景，但也存在显着的缺点，最有可能的是茎中碱基对之间氢键的强度（参考文献 11）。