The use of structure-based drug design to optimise protein-ligand interactions

Awarding institution
  • University of Strathclyde
Date of award
  • 2015
Thesis identifier
  • T14735
Person Identifier (Local)
  • 201053223
Qualification Level
Qualification Name
Department, School or Faculty
  • High attrition rates in drug candidates represent a key problem facing the pharmaceutical industry. Several studies have shown that controlling the physicochemical properties of molecules can reduce the likehood of toxicity and, thus, the risk of attrition. Fragment-based drug discovery (FBDD) is a new approach to discovering high quality drug candidates using structure-based design with inherent focus on physicochemical properties. This approach was applied to two biological targets, mitochondrial branched-chain aminotransferase (BCATm) and phosphoinositide 3-kinase delta (PI3Kδ). BCATm, which catalyses the first step of branched-chain amino acids catabolism, has been identified as a potential novel target for an anti-obesity drug. FBDD screening using a combination of biochemical and biophysical assays successfully identified a variety of starting points for BCATm inhibition. From this exercise, 11 protein-ligand structures were solved by X-ray crystallography displaying significant diversity in their binding mode. Two fragment hits were optimised using structure-based design and fragment merging techniques to identify a high quality lead molecule for this novel target. In addition, all of the data generated contributed to the discovery of a novel series of inhibitors suitable for use in critical in vivo experiments to further understand the biological role of BCATm. PI3Kδ, a lipid kinase critical in immune cell signalling, has been confirmed as a promising target for respiratory diseases. Structure-based design was initially employed to improve the selectivity profile of an existing lead series. Subsequently, novel and selective PI3Kδ inhibitors were identified by targeting a specific interaction with the protein demonstrating an advantage over previously disclosed inhibitors, which retain potency at a related kinase. In addition, FBDD and structure-based design were employed to optimise potency and selectivity of a fragment hit whilst keeping the physicochemical properties in good range for oral dosing. Overall, these novel inhibitors demonstrate a promising profile in relation to the discovery of a suitable oral PI3Kδ inhibitor for clinical development.
Resource Type
  • Previously held under moratorium in Chemistry department (GSK) from 2015 until 12th October 2017.
  • The confidentiality statement on each page of this thesis DOES NOT apply