Screening macrocyclic peptide libraries by yeast display allows control of selection process and affinity ranking

Abstract

Macrocyclic peptides provide an attractive modality for drug development due to their ability to bind challenging targes, their small size, and amenability to powerful in vitro evolution techniques such as phage or mRNA display. While these technologies proved capable of generating and screening extremely large libraries and yielded ligands to already many targets, they often do not identify the best binders within a library due to the difficulty of monitoring performance and controlling selection pressure. Furthermore, only a small number of enriched ligands can typically be characterised due to the need of chemical peptide synthesis and purification prior to characterisation. In this work, we address these limitations by developing a yeast display-based strategy for the generation, screening and characterisation of structurally highly diverse disulfide-cyclised peptides. Analysis and sorting by quantitative flow cytometry enabled monitoring the performance of millions of individual macrocyclic peptides during the screening process and allowed us identifying macrocyclic peptide ligands with affinities in the low micromolar to high picomolar range against five highly diverse protein targets. X-ray analysis of a selected ligand in complex with its target revealed optimal shape complementarity, large interaction surface, constrained peptide backbones and multiple inter- and intra-molecular interactions, rationalising the high affinity and exquisite selectivity. The novel technology described here offers a facile, quantitative and cost-effective alternative to rapidly and efficiently generate and characterise fully genetically encoded macrocycle peptide ligands with sufficiently good binding properties to even therapeutically relevant targets.