Modulation of the BH3/BCL-xL Protein-Protein Interactions Using Peptidomimetic Approaches

Protein-protein interactions (PPIs) play essential roles in regulating cellular processes, thus emerging as an important class of targets for drug discovery. The B cell CLL/lymphoma-2 (BCL-2) family of proteins control the intrinsic (or mitochondrial) apoptotic pathway, and can be divided into two major classes: pro-apoptotic members and pro-survival members. In cancer cells, the overexpression of pro-survival proteins can block pro-apoptotic signalling by sequestering the BH3-only proteins (pro-apoptotic), thus allowing cancer to grow and spread. Selective inhibition of BCL-xL, a pro-survival member, is of great interest due to its functional roles in solid tumours and drug resistance. Using peptides as inhibitors of PPIs serves as a promising approach because of their larger sizes and intrinsic conformations mimicking native binding domains. However, peptides that have native helical structures do not readily adopt bioactive shapes when taken out of the corresponding parent proteins. The main aim of this project was to design potent and selective inhibitors of BCL-xL using peptidomimetic approaches. The generated hits were proposed to serve as starting points for development of therapeutics for treatment of BCL-xL-overexpressed cancer. To do this, two main methods, peptide constraining and coiled-coil grafting, were used with assistance of computational tools.
Peptide constraining is one of the notable methods for generating stabilised helices. Strategies and consequences of constraining are described in chapter 1 (published in Chem. Sci.). Two BH3 ligands, BAD and HRK, were chose as starting points for the design of peptidomimetics targeting BCL-xL. Chapter 2 describes efforts to constrain the BAD BH3 domain (published in Bioorganic Med. Chem. Lett.). CD spectroscopy and MD simulations revealed that the loss of potency after introducing a constraint might be due to disruption of the orientation of key hot-spot residues and a reduced helicity of the bound-state ligand or complex.
Efforts to construct HRK-derived constrained peptides led to more promising outcomes as described in chapter 3 (also published in Chem. Commun.). A combination of computational modelling and a maleimide-staple scan led to a constrained peptide, HRKS4, with slightly improved potency, enhanced helicity in comparison to the linear precursor, and improved proteolytic stability. Further sequence truncation alongside hybridisation led to shorter peptides with significantly improved potency. The corresponding constrained peptide, HRKS4H, had slightly reduced potency; but more importantly, was observed to have improved selectivity.
Finally, the well-fined coiled coil, CC-Di, was selected for grafting hot-spot residues from BH3 ligands, in order to develop coiled-coil-templated inhibitors targeting the BCL-xL. Using computational tools, suitable grafting sites in the sequence were successfully identified. Several designed peptides showed weak potency in the FA competition assays. Further studies will focus on the improvement of potency and investigation of selectivity.