Discovery Of Inhibitors Targeting HIV-1 Integrase-LEDGF/p75 Interaction And Studies On RNA-Ligand Docking

With the rapid development of structural biology, more and more 3D structures of biomacromolecules have been experimentally determined, which drives Computer-Aided Drug Design (CADD) into the stage of rapid development. Nowadays, owing to CADD techniques integrated into the workflow of drug discovery, the drug research and development have been transferred from traditional random screening to structure-and mechanism-based drug design gradually. In chapter 1, we summarized the key role of CADD in the process of drug discovery and discussed some applications of core techniques of CADD on the main clue of molecular docking techniques.In chapter 2, using multiple CADD techniques comprehensively, we performed a task of virtual screening against the interface of integrase-LEDGF/p75. In this study, the molecular shape and structure-based pharmacophore techniques were used to pre-screened large small molecule databases, forming some focus databases. Afterward, combining molecular docking and GB/SA rescoring techniques, we further explored the molecules in these focus databases and selected some potential compounds for purchasing and biological test. Twety-eight compounds from NPD and NPLC compound libraries were validated by yeast two-hybrid assay. As the results, five of twety-eight compounds showed the biological activity of inhibiting integrase-LEDGF/p75 interaction. One most potent hit showed the IC50 of 40.11μM, which is equal to the biological activity of published CHIBA-3003.In chapter 3, comprehensive evaluations were first carried out on two widely used protein-based docking programs-GOLD and Glide-for their docking and virtual screening accuracies against RNA targets. Both GOLD 4.0 and Glide 5.0 could successfully reproduce 57% experimental binding modes for the 60 tested complexes. The results became much better when multiple ligand input conformations and crystal waters were considered. Applying different docking/scoring combinations, reasonable enrichments from the simulated virtual screening and fragment screening experiments were achieved against the ribosomal decoding region of the aminoacyl-tRNA acceptor (rRNA A-site). All the results demonstrated that current protein-based docking programs could fulfill general docking tasks against RNA, which would be very helpful for RNA-based drug discovery and design.