Target the More Druggable Protein States in a Highly Dynamic Protein–Protein Interaction System
详细信息 查看全文
- 作者：Zuojun Guo ; Atli Thorarensen ; Jianwei Che ; Li Xing
- 刊名：Journal of Chemical Information and Modeling
- 出版年：2016
- 出版时间：January 25, 2016
- 年：2016
- 卷：56
- 期：1
- 页码：35-45
- 全文大小：752K
- ISSN：1549-960X
文摘
The proteins of the Bcl-2 family play key roles in the regulation of programmed cell death by controlling the integrity of the outer mitochondrial membrane and the initiation of the apoptosis process. We performed extensive molecular dynamics simulations to investigate the conformational flexibility of the Bcl-xL protein in both the apo and holo (with Bad peptide and ABT-737) states. The accelerated molecular dynamics method implemented in Amber 14 was used to produce broader conformational sampling of 200 ns simulations. The pocket mining method based on the variational implicit-solvent model tracks the dynamic evolution of the ligand binding site with a druggability score characterizing the maximal affinity achievable by a drug-like molecule. Major movements were observed around the α3-helical domain and the loop region connecting the α1 and α2 helices, reshaping the ligand interaction in the BH3 binding groove. Starting with the apo crystal structure, which is recognized as “closed” and undruggable, the BH3 groove transitioned between the “open” and “closed” states during equilibrium simulation. Further analysis revealed a small percentage of the trajectory frames (～10%) with a moderate degree of druggability that mimic the ligand-bound states. The ability to attain and detect by computer simulation the most suitable conformational states for ligand binding in advance of compound synthesis and crystal structure solution is of immense value to the application and success of structure-based drug design.