Development and Application of Computational Drug Design Methods Against Microbial Pathogen Enzymes
MetadataShow full item record
The overall objective of this research is to develop new computational protocols that address some of the current challenges in computational drug design, and to use them in the identification of novel small molecules that inhibit enzymes essential for viral or bacterial replication and survival. In this dissertation, we target a cysteine protease from the Severe Acute Respiratory Syndrome Coronavirus (SARS-CoV) and a condensing enzyme from the B. anthracis fatty acid biosynthesis pathway. The SARS-CoV 3-Chymotrypsin-like protease (3CLpro), vital for SARS-CoV replication, and the condensing enzyme, beta-Ketoacyl-acyl carrier protein synthase III (FabH), essential for the B. anthracis fatty acid biosynthesis pathway, serve as selective targets for novel antiviral and antibacterial drug design, respectively. We have developed a structure-based virtual screening protocol to identify novel small molecules active against SARS CoV 3CLpro. Through the use of a tiered docking approach and consensus scoring function we have obtained considerable enrichment of true positives compared to the individual scoring functions. SARS-CoV 3CLpro shows large movements of active site loops as well as side chain flexibility of the binding pocket residues. In our screening protocol, we have incorporated a combination of docking and pharmacophore based approaches to overcome some of the challenges of protein flexibility. Nearly half a million compounds from the ZINC database were screened using the developed protocol, leading to the identification of several compounds with micromolar activity. The structure activity relationship using the active and inactive compounds was also developed to provide guidance for compound optimization. Recently, additional experimental inhibitory data against 3CLpro was obtained from experimental screening of an in-house compound collection in our laboratory, and the public release on PubChem of results from an HTS campaign through the NIH Molecular Libraries Program. We used this data to re-validate our structure-based screening protocol. Additionally, the availability of sufficient ligand data allowed us to use a ligand-based approach to find novel scaffolds active against 3CLpro. The structural determinants for FabH-inhibitor binding are poorly understood. We first characterized the FabH binding pocket using computational solvent mapping and identified hot spot residues making dominant contributions to the binding affinity. The results show good agreement with the protein-inhibitor interactions observed in the co-crystal structures. These insights can serve as a guideline for designing and optimizing lead compounds. We also developed a reliable and computationally efficient protocol for virtual screening against FabH. A comprehensive pharmacophore based screening protocol was developed taking advantage of all the structure and inhibitor information available in the literature. Three different approaches were used for pharmacophore query generation for screening the ZINC library. Three compounds were selected and their binding to the baFabH active site was confirmed using fluorescence quenching experiments. The structural and energetic analysis of the putative ligands suggest that they bind to the key hot spots identified in this work.
computational drug design
Severe acute respiratory syndrome