血管内皮生长因子VEGF及其受体VEGFR2对于肿瘤血管生成起至关重要的作用。本文旨在研究VEGFR2的咪唑并哒嗪类抑制剂的三维定量构效关系及新抑制剂分子与VEGFR2的作用机制。构建的Topomer Co MFA模型具有较强的预测能力和拟合能力(q^2=0.809,r^2=0.968)以及外部预测能力(r_(pred)~2=0.571)。应用Topomer Search技术在含1304868个分子的ZINC数据库中进行了虚拟筛选,采用基于片段的药物设计方法设计了68个高活性的新VEGFR2抑制剂。最后借助Surflex-dock技术研究了新分子与VEFGR2的作用机制,发现新抑制剂与残基Glu885、Cys919、Asn923、Asp1046等作用显著。本研究为VEGFR2抑制剂分子的结构修饰、设计与合成提供了重要的理论指导。
Janus kinase 3（JAK3） is a member of Janus kinase（JAK） family, and it represents a promising target for the treatment of immune diseases and cancers. However, no highly selective inhibitors of JAK3 have been developed. For discovering the binding mechanism of JAK3 and these inhibitors, a molecular modeling study combining molecular docking, three-dimensional quantitative structure-activity relationships（3D-QSAR）, molecular dynamics and binding free energy calculations was performed on a series of pyrimidine-based compounds which could bind with the unique residue Cys909 of JAK3 kinase as the selective inhibitors of JAK3 in this work. The optimum Co MFA and Co MSIA models were generated based on the conformations obtained by molecular docking. The results showed that the models have satisfactory predicted capacity in both internal and external validation. Furthermore, a 50 ns molecular dynamics simulation was carried out to determine the detailed binding process of inhibitors with different activities. It was demonstrated that hydrogen bond interactions with Leu828, Glu903, Tyr904, Leu905 and Leu956 of JAK3 are significant for activity increase, and the Van der Waals interaction is mainly responsible for stable complex.
Enhancer of Zeste homolog 2（EZH2） is closely correlated with malignant tumor and regarded as a promising target to treat B-cell lymphoma. In our research, the molecular docking and three-dimensional quantitative structure-activity relationships（3D-QSAR） studies were performed on a series of pyridone-based EZH2 compounds. Molecular docking allowed us to study the critical interactions at the binding site of EZH2 protein with inhibitors and identify the practical conformations of ligands in binding pocket. Moreover, the docking-based alignment was applied to derive the reliable 3D-QSAR models. Comparative molecular field analysis（CoMFA） and comparative molecular similarity indices analysis（CoMSIA） provided available ability of visualization. All the derived 3D-QSAR models were considered to be statistically significant with respect to the internal and external validation parameters. For the CoMFA model, q^2 = 0.649, r^2 = 0.961 and r^2 pred = 0.877. For the CoMSIA model, q^2 = 0.733, r^2 = 0.980 and r^2 pred = 0.848. With the above arguments, we extracted the correlation between the biological activity and structure. Based on the binding interaction and 3D contour maps, several new potential inhibitors with higher biological activity predicted were designed, which still awaited experimental validation. These theoretical conclusions could be helpful for further research and exploring potential EZH2 inhibitors.
