Mori Toshifumi, Saito Shinji
Institute for Molecular Science, Okazaki, Aichi 444-8585, Japan and School of Physical Sciences, The Graduate University for Advanced Studies, Okazaki, Aichi 444-8585, Japan.
J Chem Phys. 2015 Apr 7;142(13):135101. doi: 10.1063/1.4916641.
Molecular dynamics simulations have become an important tool in studying protein dynamics over the last few decades. Atomistic simulations on the order of micro- to milliseconds are becoming feasible and are used to study the state-of-the-art experiments in atomistic detail. Yet, analyzing the high-dimensional-long-temporal trajectory data is still a challenging task and sometimes leads to contradictory results depending on the analyses. To reveal the dynamic aspect of the trajectory, here we propose a simple approach which uses a time correlation function matrix and apply to the folding/unfolding trajectory of FiP35 WW domain [Shaw et al., Science 330, 341 (2010)]. The approach successfully characterizes the slowest mode corresponding to the folding/unfolding transitions and determines the free energy barrier indicating that FiP35 is not an incipient downhill folder. The transition dynamics analysis further reveals that the folding/unfolding transition is highly heterogeneous, e.g., the transition path time varies by ∼100 fold. We identify two misfolded states and show that the dynamic heterogeneity in the folding/unfolding transitions originates from the trajectory being trapped in the misfolded and half-folded intermediate states rather than the diffusion driven by a thermal noise. The current results help reconcile the conflicting interpretations of the folding mechanism and highlight the complexity in the folding dynamics. This further motivates the need to understand the transition dynamics beyond a simple free energy picture using simulations and single-molecule experiments.
在过去几十年里,分子动力学模拟已成为研究蛋白质动力学的重要工具。微秒到毫秒量级的原子模拟正变得可行,并被用于从原子层面详细研究前沿实验。然而,分析高维长时间的轨迹数据仍是一项具有挑战性的任务,而且有时根据分析方法的不同会得出相互矛盾的结果。为了揭示轨迹的动力学方面,我们在此提出一种简单方法,该方法使用时间相关函数矩阵,并应用于FiP35 WW结构域的折叠/去折叠轨迹[Shaw等人,《科学》330, 341 (2010)]。该方法成功地刻画了对应于折叠/去折叠转变的最慢模式,并确定了自由能垒,表明FiP35不是一个初始的下坡折叠蛋白。转变动力学分析进一步揭示,折叠/去折叠转变具有高度的异质性,例如,转变路径时间变化约100倍。我们识别出两种错误折叠状态,并表明折叠/去折叠转变中的动力学异质性源于轨迹被困在错误折叠和半折叠的中间状态,而非热噪声驱动的扩散。当前结果有助于调和对折叠机制的相互矛盾的解释,并突出了折叠动力学的复杂性。这进一步促使人们需要通过模拟和单分子实验,超越简单的自由能图景来理解转变动力学。