针对几种抗病毒、抗肿瘤靶标的活性化合物发现和作用机理研究

发布时间：2018-01-23 23:25

本文关键词： 分子对接 EV71 3C蛋白酶登革热 NS2B/NS3蛋白酶分子动力学 PRMT5　出处：《中国科学院大学(中国科学院上海药物研究所)》2017年博士论文　论文类型：学位论文

【摘要】：药物研发的过程中,分子对接对于发现和改造活性化合物非常重要。分子对接是通过计算模拟蛋白质与小分子相互作用方式的一种方法。随着计算方法的不断优化,分子对接的准确性也大大提高。因其快速和有效性,分子对接已被广泛运用到大规模筛选小分子的科学研究过程中。对于大部分的体系,分子对接都可以得到比较理想的结果。针对EV71 3C蛋白酶和PRMT5甲基转移酶两个靶标,我们使用分子对接方法成功地找到了骨架新颖的活性化合物。而登革热NS2B/NS3蛋白酶催化口袋非常平坦且极性较强,不适合使用分子对接方法进行筛选。因此,我们使用了基于配体的虚拟筛选方法。合理地选用虚拟筛选方法对于成功开发靶标抑制剂至关重要。本论文第一章首先介绍了分子对接的基本流程,接着介绍柔性对接和共价对接近期发展的进程以及需要改进的方向,最后介绍了打分函数,特别是机器学习方法在打分函数方面的应用,并将机器学习的打分函数与传统打分函数在打分能力、对接能力、排序能力以及筛选能力四个方面进行了比较。通过我们对目前分子对接方面的进展的介绍,将有助于药物设计者了解各对接软件的优缺点,根据实际需要挑选合适的对接软件。另外,也有利于进一步改进和发展分子对接软件。本论文第二章介绍了靶向EV71 3C蛋白酶的抑制剂设计以及相关生物实验验证过程。由EV71病毒引起的手足口病会使患者产生严重的神经系统并发症。EV71 3C蛋白酶是治疗手足口病的一个重要靶标。本章中,我们首先使用分子对接方法筛选得到了一个活性化合物DC07090,其IC50值为21.72μM。分子对接和动力学模拟预测了小分子的结合模式,并通过突变实验验证了结合模式的正确性。通过序列比对,我们发现与DC07090结合的氨基酸在小RNA病毒家族非常保守。抗病毒实验结果表明DC07090对EV71和CVA16两种病毒都有抗病毒活性,EC50值分别为22.09μM和27.76μM。化合物DC07090毒性比较小,CC50大于200μM。不同于现有的拟肽类和黄酮类抑制剂,我们发现的小分子DC07090骨架结构新颖,且毒性小,具有更好的研究前景。本论文第三章介绍了靶向登革热NS2B/NS3蛋白酶设计抑制剂的相关研究工作。登革热NS2B/NS3蛋白酶对病毒的复制至关重要,是研究抗登革热病毒抑制剂的主要靶标。NS2B/NS3蛋白酶催化口袋比较浅且极性较大,不适合分子对接虚拟筛选,所以采用了基于配体的虚拟筛选方法。首先进行了数据收集,将其分成训练集和测试集,建立了药效团模型ADRR261,并验证了该药效团模型的正确性。富集率测试结果表明该药效团模型对活性分子有很好的区分能力。接着,使用该药效团模型对SPECS数据库进行筛选,得到一个活性分子DC16,其IC50为14.8μM。根据结构相似性,找到了3个活性更好的化合物。实验结果表明,这四个活性化合物都能抑制登革热病毒的复制,其中DC16的抗病毒活性EC50为6.82μM。但是这类化合物具有一定的毒性,有待进一步优化改造。在本研究中,我们使用药效团模型,发现了一类具有新骨架的登革热NS2B/NS3蛋白酶抑制剂,研究成果有助于该类抗病毒药物的发展。本论文第四章介绍了靶向PRMT5抑制剂的设计及相关研究工作。蛋白质精氨酸甲基转移酶5(Protein arginine methyltransferase 5,PRMT5)在一些肿瘤细胞中过表达,抑制PRMT5成为治疗癌症的一种重要方法。首先,我们建立了分子对接的虚拟筛选模型,并测试了模型的区分能力。利用该模型,发现了一个新型骨架抑制剂DC_P33,其IC50为35.6μM。根据结构相似性,找到了14个类似物。通过分析结合模式和构效关系,进一步改造化合物。接着,合成了19个衍生物。活性最好的化合物为DC_C01,其IC50为2.8μM。实验发现DC_C01对PRMT5有较高的选择性。最后,细胞实验结果表明DC_C01对三种肿瘤细胞系Z-138、Maver-1和Jeko-1都具有抗增殖活性,其EC50分别为12.7μM、12.7μM、和10.5μM。因此,我们发现的小分子丰富了靶向PRMT5的特异性抑制剂骨架,为进一步优化和改造化合物提供了一些思路。
[Abstract]:The process of drug development, molecular docking for discovery and improvement of active compounds is very important. Molecular docking is through a simulation method of protein and small molecular interaction mode. With the continuous optimization calculation method, the accuracy of molecular docking is also greatly improved. Because of its rapid and effective, has been widely applied to molecular docking large scale screening of small molecules in scientific research. For most of the molecular docking system, can get ideal results. According to the EV71 3C protease and PRMT5 methyltransferase two target, we used molecular docking method successfully found the activity of novel compounds of the skeleton. The catalytic pocket dengue NS2B/NS3 protease is very flat and polar strong, not suitable for the use of molecular docking method for screening. Therefore, we use virtual screening method based on reasonable selection of ligands. The use of virtual screening methods for the successful development of target is very important. The first chapter of this thesis first introduces the basic process of molecular docking, and then introduces the recent development of flexible docking and covalent docking process and the direction of improvement, finally introduces the scoring function, especially the application of machine learning methods in the scoring function, and docking capability the scoring function of machine learning and traditional scoring function in scoring ability, the four aspects of ordering and screening ability were compared. Through our current progress in molecular docking of the drug will help designers to understand the advantages and disadvantages of the docking software, according to the selection of appropriate docking software actual needs. In addition, too is conducive to the further improvement and development of molecular docking software. The second chapter of this paper introduces the design of EV71 protease inhibitors targeting 3C and related Biological experiments. Caused by the EV71 virus of HFMD patients will produce neurological complications of severe.EV71 3C protease is an important target for the treatment of HFMD. In this chapter, we first use molecular docking method and screened a DC07090 active compounds, the IC50 predicted binding modes of small molecules the 21.72 M. molecular docking and dynamics simulation, and the correctness of the binding mode is verified by experiment. Mutations by sequence alignment, we found that the amino acid combined with DC07090 is very conservative in the small RNA virus family. The experimental results show that the DC07090 virus has antiviral activity against EV71 and CVA16 two kinds of viruses, EC50 respectively 22.09 M and 27.76 M. compound DC07090 toxicity is relatively small, CC50 more than 200 M. is different from the existing peptidomimetic and small molecule inhibitors of flavonoids, we found that the DC07090 skeleton node Novel structure, and little toxicity, with better prospects for research. The third chapter of this paper introduces the related research work targeting NS2B/NS3 protease inhibitors. The design of dengue dengue NS2B/NS3 protease is essential for viral replication, is the main target of the catalytic pocket.NS2B/NS3 protease inhibitor anti dengue virus is shallow and large polarity, not suitable for molecular docking virtual screening, so the use of virtual screening method based on ligand. First data collection, divided into training set and test set, established the ADRR261 pharmacophore model, and verifies the correctness of the pharmacophore model. The enrichment rate test results show that the pharmacophore model has a good ability to distinguish between activity molecular. Then, using the pharmacophore model of SPECS database screening, a reactive molecule DC16, its IC50 was 14.8 M. according to the structural similarity, find 3 active compounds to better. The experimental results show that the four compounds can inhibit dengue virus replication, the antiviral activity of EC50 DC16 6.82 M. but the compounds have certain toxicity, needs to be further optimized. In this study, we use the pharmacophore model, found inhibitors NS2B/NS3 a new dengue protease framework, research results contribute to the development of this kind of antiviral drugs. The fourth chapter of this paper introduces the design of targeted inhibitors of PRMT5 and related research work. The protein arginine methyltransferase 5 (Protein arginine 5 methyltransferase, PRMT5) is over expressed in some tumor cells, inhibition PRMT5 has become an important method for the treatment of cancer. Firstly, we established the model of virtual screening of molecular docking, and tested the ability to distinguish model. Using this model, found A new type of skeleton inhibitor DC_P33, its IC50 is 35.6 M. according to the structural similarity, found 14 analogues. By analyzing the mode and structure-activity relationship, further transformation of compounds. Then, 19 derivatives were synthesized. The activity of the compounds for the best DC_C01, IC50 2.8 M. DC_C01 has found high selectivity for PRMT5. Finally, the experimental results show that DC_C01 cells of three tumor cell lines Z-138, Maver-1 and Jeko-1 have antiproliferative activity, the EC50 were 12.7 M, 12.7 M and 10.5 M., therefore, we found a rich small molecule targeting specific inhibitor PRMT5. The paper provides some ideas for further optimization and improvement of compounds.

【学位授予单位】：中国科学院大学(中国科学院上海药物研究所)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：R91

【相似文献】