Structural and Informatics Basis of Channel Protein Interactions

采用归纳同源建模方法和低同源区域演绎分子模拟相结合的方法，建立通道蛋白的三级结构模型。特别是，我们发展了一种新的柔性环区域的建模方法，该方法基于基于广义Born近似的隐式溶剂模型的偏向力多正则分子动力学(FB-MCMD)。该方法即使在柔性环区域也能给出精确的蛋白质模型结构，并应用于实际通道蛋白质的建模。此外，利用我们的新数据库EF-SITE和HINTdb分别利用信息学技术对蛋白质的功能及其位点和蛋白质-蛋白质相互作用对进行了鉴定和预测。为了分析通道蛋白的分子识别，我们发展了一种新的核磁共振分析方法，其中复杂结构是通过分子动力学模拟产生的交叉饱和转移和剩余偶极耦合信号产生的。应用该方法建立了KCSA K~(++)通道与Aitoxin-2的络合模型。还发展了布朗动力学模拟方法来计算小离子和化合物通过通道蛋白的渗透。

Methods were developed for building tertiary structural models of channel proteins, by combining the inductive homology modeling method and the deductive molecular simulation for low-homologous regions. In particular, we developed a new model building method for flexible loop regions, based on the Force-Biased Multicanonical Molecular Dynamics (FB-McMD) with the implicit solvent model using the generalized Born approximation. This method could give us a precise protein model structure even for the flexible loop regions, and it was applied to modeling of actual channel proteins. In addition, protein functions with their sites and protein-protein interaction pairs were identified and predicted with the informatics technology using our new databases, eF-site and HINTdb, respectively.In order to analyze the molecular recognition of channel proteins, we have developed a new NMR analysis method, where the complexed structure is produced by molecular dynamics simulations using the cross saturation transfer and the residual dipolar coupling signals. This method was applied to build a complexed model between KcsA K^+-channel and agitoxin-2. Brownian dynamics simulation method was also developed for computing permeation of small ions and compounds through a channel protein.