Allosteric Transitions of Supramolecular Systems Explored by Network Models

Most proteins are biomolecular machines. They perform their function by undergoing changes between different structures. Understanding the mechanism of transition between these structures is of major importance to design methods for controlling such transitions, and thereby modulating protein function. However, exploring the transition between conformations is difficult, both experimentally and computationally, due to the transient nature of the intermediate, high energy conformers crossed over as the molecule undergoes structural changes. In many cases, only the two ending structures are known from experiments. Furthermore, the passage between the two end points does not necessarily involve a single pathway, but multiple pathways in the multidimensional energy landscape associated with the macromolecular structures. To bridge between structure and function, a molecular understanding of the most probable transition pathways between the two end structures is required.While there are many computational methods for exploring the transitions of small proteins, the task of exploring the transition pathways becomes prohibitively expensive in the case of supramolecular systems. Coarse-grained models that lend themselves to analytical solutions appear to be the only possible means of approaching such cases. Motivated by the utility of elastic network models for describing the collective dynamics of biomolecular systems, and by the growing theoretical and experimental evidence in support of the intrinsic accessibility of functional substates, we introduce a new method, adaptive anisotropic network model (aANM) for exploring functional transitions.As described by aANM, a series of intermediate conformations along the transition pathways between the initial and final conformations were generated by successive deformations of both end structures that were iteratively updated. The directions of deformations were determined by implementing the deformations along the directions of dominant ANM modes accessible to the intermediate states. The recruitment of the particular subsets of modes results from a tradeoff between minimizing the path length and selecting the direction of the lowest increase in internal energy. To calculate the ANM modes, please visit our related websites.ANM: http://ignmtest.ccbb.pitt.edu/cgi-bin/anm/anm1.cgi GNM: http://ignm.ccbb.pitt.edu/GNM_Online_Calculation.htmPCA_NEST: http://ignm.ccbb.pitt.edu/oPCA_Online.htmThe bacterial chaperonin GroEL is a supramolecular machine that has been broadly studied in recent years using both experimental and theoretical or computational methods. Yet, a structure-based analysis of the transition of the intact chaperonin between its functional forms has been held back by the large size of the chaperonin. The aANM method is proposed as a first approximation toward approaching this challenging task. The application of aANM to GroEL, not only elucidated the highly probable pathways and the hierarchic contribution of modes to achieve the transition; but also provided us with biologically significant information on critical interactions and sequence of events occurring during the chaperonin machinery and key contacts that make and break at the transition.On a practical side, the major utility of the method lies in its application to the transitions of supramolecular systems beyond the range of exploration of other computational methods. The computing time in the present method is several orders of magnitude shorter than that required in regular molecular dynamics or Brownian dynamics simulations. *Figure above: Snapshots of the protein chaperonin GroEL in its transition pathway, evolving from open (upper left) to closed form (lower right). The color scheme was inspired by Wassily Kandinsky and his artwork “Squares and Concentric Rings”.For more information, please visit:http://www.ccbb.pitt.edu/Faculty/bahar/index.phphttp://www.ccbb.pitt.edu/Faculty/bahar/publications/YZResearch/Coupling.htmlhttp://www.ccbb.pitt.edu/Faculty/bahar/publications/YZResearch/Transitions.html <br/><br/>This project includes the following software/data packages: <br/> <ul> <li> <a href="https://simtk.org/frs?group_id=531#pack_873">aANM </a> : Adaptive Anisotropic Network Model(aANM) source codes. For details, please see the enclosed tutorial. </li> <li> <a href="https://simtk.org/frs?group_id=531#pack_874">Pathway Trajectories </a> : The pathway trajectories, in PDB file format, for both single GroEL chain and intact complex. PDB files can be opened with UCSF Chimera or PyMOL.For details, please see enclosed "Readme.txt" file. </li> </ul>

大多数蛋白质均为生物分子机器。它们通过在不同结构之间发生转变以执行其功能。理解这些结构之间转变的机制对于设计控制此类转变的方法至关重要，从而调节蛋白质的功能。然而，由于中间态、高能量构象在分子经历结构变化时跨越的短暂性，探索构象之间的转变在实验和计算上都具有难度。在许多情况下，实验仅揭示了两个终态结构。此外，两个终点之间的过渡并不一定涉及单一的途径，而是涉及与宏分子结构相关的多维能量景观中的多个途径。为了在结构和功能之间架起桥梁，需要理解两个终态结构之间最可能的转变途径。尽管存在许多用于探索小蛋白质转变的计算方法，但在超分子系统中探索转变途径的成本却变得过高。粗粒度模型，其自身适用于解析解决方案，似乎是接近此类情况的唯一可能途径。受弹性网络模型在描述生物分子系统集体动力学中的效用以及支持功能亚态内在可及性的理论和实验证据日益增多所启发，我们引入了一种新的方法，即自适应各向异性网络模型（aANM），用于探索功能转变。根据aANM的描述，通过迭代更新两端结构，通过连续变形生成了一系列位于初始和最终构象之间转变途径上的中间构象。变形的方向是通过实施变形沿中间态可访问的占主导地位的ANM模式的方向来确定的。特定模式子集的征召源于在最小化路径长度和选择最低内部能量增加的方向之间的权衡。为了计算ANM模式，请访问我们的相关网站。ANM: http://ignmtest.ccbb.pitt.edu/cgi-bin/anm/anm1.cgi GNM: http://ignm.ccbb.pitt.edu/GNM_Online_Calculation.htm PCA_NEST: http://ignm.ccbb.pitt.edu/oPCA_Online.htm细菌伴侣蛋白GroEL是一种广受研究的超分子机器，近年来，人们已通过实验和理论或计算方法对其进行了广泛研究。然而，由于伴侣蛋白的大尺寸，对其在功能形式之间完整转变的结构分析一直受到阻碍。aANM方法被提出作为接近这一挑战性任务的第一近似。将aANM应用于GroEL不仅阐明了高度可能的途径和模式对实现转变的层次贡献；还为我们提供了关于伴侣蛋白机器中关键相互作用和事件序列以及形成和断裂的关键接触的生物意义信息。从实际应用的角度来看，该方法的主要效用在于其应用于超分子系统的转变，这些转变超出了其他计算方法的探索范围。与常规分子动力学或布朗运动模拟相比，本方法所需的计算时间短几个数量级。*上图：蛋白质伴侣蛋白GroEL在其转变途径上的快照，从开放形式（左上角）演变为闭合形式（右下角）。色彩方案灵感来源于瓦西里·康定斯基及其作品“正方形与同心圆”。有关更多信息，请访问：http://www.ccbb.pitt.edu/Faculty/bahar/index.php http://www.ccbb.pitt.edu/Faculty/bahar/publications/YZResearch/Coupling.html http://www.ccbb.pitt.edu/Faculty/bahar/publications/YZResearch/Transitions.html <br/><br/>本项目包括以下软件/数据包： <br/> <ul> <li> <a href="https://simtk.org/frs?group_id=531#pack_873">aANM </a> : 自适应各向异性网络模型(aANM)源代码。有关详细信息，请参阅附带的教程。 </li> <li> <a href="https://simtk.org/frs?group_id=531#pack_874">Pathway Trajectories </a> : 单个GroEL链和完整复合物的途径轨迹，格式为PDB文件。PDB文件可以用UCSF Chimera或PyMOL打开。有关详细信息，请参阅附带的“Readme.txt”文件。 </li> </ul>