论文部分内容阅读
作为第二代共价类蛋白酶体抑制剂,硼酸肽类抑制剂由于代谢稳定且具较高的蛋白酶体抑制活性和选择特异性,已成为当前新型抗肿瘤药物的重点研究内容。本文以最近合成的44个Tyropeptin硼酸三肽类蛋白酶体抑制剂为研究对象,采用共价距离约束对该类分子与蛋白酶体进行了分子对接研究。结果表明,该类抑制剂与蛋白酶体的非共价相互作用主要以氢键和疏水作用为主,具体表现为:(1)硼酸基团的羟基与ARG19、LYS33和GLY47的氢键作用;(2)肽链骨架原子与THR21、GLY47和ALA49的氢键作用;(3)R1基团与S1口袋的疏水作用和氢键作用。基于对接构象,对上述分子进行了骨架叠合及随后的比较分子场分析(CoMFA)和比较分子相似性形状指数分析(CoMSIA)。最优模型为包含氢键供体场、疏水场、氢键受体场以及立体场的CoMSIA模型,其最佳主成分数、决定系数r2、标准差S、交互验证系数q2以及外部预测r2pred分别为3、0.882、0.188、0.494以及0.756。在以上研究基础上,以活性最高的3C和6号分子为模板,采用基于遗传算法的结构搜索方法结合分子相似性评价函数对其侧链进行了优化设计。结合Lipinski“5规则”和最优CoMSIA模型活性预测结果,最终得到5个目标分子,其预测活性均达到纳摩尔水平。
As the second generation of covalent class of proteasome inhibitors, boronic acid peptide inhibitors have become the focus of the current novel anti-tumor drugs because of their stable metabolism and high proteasome inhibitory activity and selective specificity. In this paper, 44 newly synthesized Tyropeptin boric acid tripeptide proteasome inhibitors were selected as the research objects, and the covalent distance constraints were used to study the molecular docking of these molecules with the proteasome. The results showed that the non-covalent interactions between these inhibitors and the proteasome mainly dominated by hydrogen bonds and hydrophobic interactions, which were as follows: (1) the hydrogen bonding of the boronic acid groups with ARG19, LYS33 and GLY47; ( 2) Hydrogen bonding between peptide chain backbone atoms and THR21, GLY47 and ALA49; (3) Hydrophobic interaction and hydrogen bonding between R1 group and S1 pocket. Based on the docked conformation, the above-mentioned molecules were scaffolds and subsequent comparative molecular field analysis (CoMFA) and comparative molecular similarity shape index analysis (CoMSIA). The optimal model is the CoMSIA model which includes the donor hydrogen field, the hydrophobic field, the hydrogen bond acceptor field and the stereo field. The optimal principal components, the determination coefficient r2, the standard deviation S, the mutual verification coefficient q2 and the external prediction r2pred 3,0.882,0.188,0.494 and 0.756. Based on the above studies, the optimal design of the side chains was carried out by using the structural search method based on genetic algorithm and the molecular similarity evaluation function using the most active 3C and 6 as templates. Combined with the Lipinski “5 rules ” and the optimal CoMSIA model activity prediction results, eventually obtained five target molecules, the predicted activity reached nanomolar levels.