Design and Applications of Encoded Peptide Libraries for Covalent Ligand Discovery Against Spike RBD and MRCKβ

Candidate: John Mulawka
Supervisor: Dr. Marc Adler

Abstract:

The recent resurgence of covalent drugs in the world of pharmaceutical design has prompted a growing need for novel screening methodologies capable of identifying ligands which bind targets previously considered undruggable. With approximately 85% of protein targets exhibiting shallow, structurally constrained, or dynamic binding surfaces which evade small molecule therapeutics, combinatorial peptide libraries offer a compelling solution by allowing a high-throughput method to screen complex structures. However, the widespread adoption of these screening methodologies has been challenged by synthetic encoding and sequence identification techniques. This thesis investigates two distinct encoding methodologies, spatial segregation and ladder synthesis, for the synthesis, screening an characterization of one-bead-one-compound peptide libraries tailored towards the identification of covalent ligand discovery.
Chapter two investigates the synthesis of a spatially segregated one-bead-two-compound hexapeptide library with five random positions. Using topological segregation, the library was synthesized to contain internal encoding sequences and external screening sequences. The library was screened against SARS-CoV-2 spike RBD with hit sequences later being functionalized with 3-(Fluorosulfonyl) benzoic acid. This two-step screening approach identified 25 covalent hits. Hit sequences were analyzed with partial Edman degradation and MALDI-TOF-MS, however characterization was limited through ion suppression and segregation efficiency.
Chapter three details a ladder-encoded peptide library where a 5% alanine cap is introduced during library synthesis. This method generated six discernable truncates during peptide elongation. The library was screened against MRCKβ identifying 28 covalent hits, and subsequently sequenced using MALDI-TOF-MS, HRESI-MS, and UHPLC-ESI-MS. Characterization of peptides following UPHLC separation prior to mass analysis significantly increased complete sequence deconvolution successfully identifying 77% of strong hits.
Together, these methodologies introduce a scalable and modular approach to the discovery of novel covalent ligands. The critical analysis of various encoding and characterization techniques offer a distinct advantage for hit-sequence identification and electrophile incorporation, addressing limitations in the current drug discovery efforts for difficult protein targets.