物理化学学报 >> 2010, Vol. 26 >> Issue (10): 2833-2839.doi: 10.3866/PKU.WHXB20100916

生物物理化学 上一篇    下一篇

A1腺苷受体的同源模建及其结构验证

柯艳蓉, 金宏威, 刘振明, 张亮仁   

  1. 北京大学医学部药学院, 天然药物及仿生药物国家重点实验室, 北京 100191
  • 收稿日期:2010-05-19 修回日期:2010-06-07 发布日期:2010-09-27
  • 通讯作者: 张亮仁 E-mail:liangren@bjmu.edu.cn

Homology Modeling and Structure Validation of the Adenosine A1 Receptor

KE Yan-Rong, JIN Hong-Wei, LIU Zhen-Ming, ZHANG Liang-Ren   

  1. State Key Laboratory of Natural and Biomimetic Drugs, School of Pharmaceutical Sciences, Health Science Center, Peking University, Beijing 100191, P. R. China
  • Received:2010-05-19 Revised:2010-06-07 Published:2010-09-27
  • Contact: ZHANG Liang-Ren E-mail:liangren@bjmu.edu.cn

摘要:

采用同源模建的方法构建了A1腺苷受体的三维结构, 并与拮抗剂分子DPCPX对接, 将得到的复合物结构进行5 ns的分子动力学模拟, 以最后2 ns 的平均结构和平衡后抽取的11 帧构象共12 个蛋白结构为研究对象,用包含52 个活性分子和1000个诱饵分子的测试库,分别通过DOCK、VINA 和GOLD 三种对接软件进行评价,最终得出合理的蛋白质模型.根据top10%的富集因子(EF)和ROC 曲线下面积(AU-ROC)的计算结果, 我们认为GOLD是最适合A1腺苷受体的对接软件, 而12 个蛋白质结构中F5 和Favg 的三维结构模型比较合理, 可以作为进一步大规模虚拟筛选的模型.

关键词: 分子动力学模拟, A1腺苷受体, 同源模建, GOLD, 虚拟筛选

Abstract:

A three dimensional structure model of the adenosine A1 receptor was built using homology modeling. The antagonist DPCPX was docked into the model protein to form a receptor-ligand complex. A molecular dynamics simulation over 5 ns was performed for this complex. We selected 12 protein structures, including the average structure obtained from the last 2 ns of the simulation and 11 frames extracted after equilibration for the study. A database comprising 52 active antagonists and 1000 decoys was then docked into the 12 protein models using DOCK, VINA, and GOLD software packages and these molecules were ranked by their docking scores. The best model protein with the highest enrichment factor (EF) and the largest area under the ROC (AU-ROC) was chosen for further study. The results from the enrichment factor at 10%of the ranked database (EF10) and AU-ROC calculations indicate that GOLD is the best virtual screening software for the adenosine A1 receptor. GOLD docking results suggest that optimized adenosine A1 receptor protein structures, Favg and F5, can be used for virtual screening and for novel design to discover more potent antagonists.

Key words: Molecular dynamics simulation, Adenosine A1 receptor, Homology modeling, GOLD, Virtual screening