基于PDB卷曲库研究迷你蛋白侧链特性对其结构的影响

The major driving force of protein folding depends on their inter-residue interactions, especially for sidechain-sidechain interactions. Although the Trp-cage is a fast-folding mini-protein containing merely 20 amino acid residues, it exhibits an extremely complex interaction network between residues with different types of sidechain orientation. Because current force fields exhibit accuracy inadequacies in the descriptions of atomic interactions, it still remains a matter of debate at what stage of folding the helix, hydrophobic cluster and salt-bridge is being conclusively formed in its possible folding mechanisms. Based on statistical analysis of intrinsic conformational features of coil regions in protein structures (PDB coil library), we have improved the current OPLS-AA/L force field by adding some special non-bonded interactions, and the modified force field (OPLS-AA/C) can fold a series of peptides and proteins with various secondary structures to their experimental structures. Herein, the Trp-cage folding simulations are conducted by employing conventional molecular dynamics (MD), replica exchange MD (REMD), temperature-controlled MD (TCMD) and distance-restrained MD (DRMD). We applied these two force fields and four different conformational sampling methods to investigate their on-pathway and off-pathway folding/unfolding events and compared them with respect to their folding accuracy and efficiency. The single-site mutations are performed for all the residues of the Trp-cage, except three glycines. Compared with the folding thermodynamics and kinetics of the wild-type (WF) protein, the respective role of each residue can be readily determined for structural stability and folding behaviors. In addition, the mutations are synchronously performed on the homologous residues, and then simulated for exploring their cooperative interactions and conformational entropy on the Trp-cage folding. These simulation results are further verified by combining with circular dichroism (CD) experiments. This systematic study provides useful insight for rapid folding of Trp-cage mini-protein, further improving its structure and more efficient polypeptide rational design.

蛋白质折叠驱动力主要取决于结构上残基间相互作用，尤其是其侧链间作用。迷你蛋白Trp-cage仅有20个氨基酸组成，但其结构中残基间的相互作用极为复杂且侧链类型多样。由于现有力场参数对其相互作用描述的精度问题，其折叠模型仍有争议。基于PDB构建的局部构象偏好卷曲库，我们已优化现有OPLS-AA/L力场，新的OPLS-AA/C力场能更准确快速折叠模型蛋白到实验结构。本项目拟将在这两种力场下通过四种构象取样方法对该蛋白从伸展态到折叠态转变过程进行模拟，比较不同力场和构象取样方法对其折叠路径推测的影响；然后采用残基点突变方法化解相互作用复杂性，确定各个残基侧链对其结构的影响，利用残基协同突变，阐明不同特性侧链作用协同性以及C-端连续脯氨酸构象熵效应对折叠过程的影响，并结合圆二色谱实验对模拟计算结果予以验证。本项目研究有助于揭示该蛋白快速折叠的驱动力成因，为进一步优化蛋白结构和多肽设计提供有益参考。

蛋白质结构中残基作用、折叠机理及其二者关系一直是结构生物学的研究重点，而理论计算研究该问题的难点主要体现在力场参数的准确性与构象采样的充分有效性。本项目采用不同力场和不同构象采样方法对迷你蛋白Trp-cage及其突变体的结构和折叠深入研究，得到以下结果：. 1. 在OPLSAA/L力场下，采用常温常规MD模拟Trp-cage折叠发现，结构在0.5 和1.0 μs内均无法准确折叠形成天然折叠态结构，而是形成错误折叠态结构M和M’，且其折叠/解折叠路径为D ↔ TS ↔ I1 ↔ I2↔ M；而通过提高模拟温度加快构象速率或两端位置原子逐步缩短并固定可准确快速结构平衡等方式可极大增强构象采样的充分有效性，结构在0.52 μs内能准确折叠形成近天然折叠态结构N（RMSD = 2.3 Å）；而基于PDB卷曲库构建的局部构象偏好的OPLSAA/C力场下，常规MD在1.0 μs内就可快速折叠形成天然折叠态结构F（RMSD = 0.45 Å）；. 2. 比较各个突变体与野生态自由能曲面发现突变体折叠差异显著，可分为三类：1）突变体P17G 与P19G 可折叠形成近天然折叠态结构（Near-native Folded State，N），且与野生态的最终折叠态结构相对自由能极为相似（-2.9，-2.9和-2.8 kBT）。2）突变体N1G、I4G、Q5G、L7G、P12G、S14G和P18G不同程度地发生解折叠，最终形成解折叠态结构（Unfolded State, U）。3）其他突变体L2G、Y3G、W6G、K8G、D9G、S13G、R16G和S20G结构扭转形成错误折叠态结构（Mis-folded State, M）。. 3. 残基侧链对于结构稳定性和折叠动力学的影响并不完全相同，这与残基侧链的位置和特性密切相关。如碳氮两端上Asn1、Leu2、Ser20等和α-螺旋区域上Lys8、Asp9等残基侧链外置不稳，对其结构稳定度贡献较小，这是由于残基侧链较长且带有极性和电荷，易与结构表面的溶剂水发生氢键作用，而无法参与结构疏水核心形成作用；但这些残基对其折叠动力学的影响却显著，这是因为结构折叠驱动力并不只是疏水塌缩作用，而是残基相互协同作用共同完成，尤其是（D9-R16）盐桥和氢键作用网络。这些残基作用效果的阐明为优化蛋白结构和多肽设计提供重要理论依据。