A continuous evolution platform for cyclic peptide discovery

The University of Wisconsin–Madison received a $1.83M Project Grant award from the National Institute of General Medical Sciences (NIGMS) under the Biomedical Research and Research Training program (CFDA 93.859) effective September 1, 2025, through August 31, 2029. The award funds development of a phage-assisted continuous evolution (PACE) platform for cyclic peptide discovery that addresses significant limitations in cellular selection technologies. The research will deliver a validated platform capable of performing activity-based selections for protein-protein interaction modulators while incorporating negative selection mechanisms to eliminate promiscuous hits and false positives—capabilities not currently available in existing cellular cyclic peptide discovery methods.

The project will advance the platform through three primary deliverables: (1) expansion of activity-based selection methodologies for both inhibitors and inducers of protein-protein interactions, (2) development of continuous in vivo mutagenesis methodology to overcome library size constraints inherent in cellular selections, and (3) incorporation of non-canonical amino acid incorporation to enable side-chain and backbone-modified cyclic peptides with enhanced diversity and improved bioavailability. These research outputs are designed to bridge the gap between in vitro display methods—which identify potent binders but lack cellular context—and current cellular selection approaches, enabling discovery of cyclic peptides with complex inhibitory activities in physiological cellular environments.

Abstract

Cyclic peptides have attracted enormous interest as research tools and therapeutics because they can inhibit historically difficult to drug targets such as protein-protein interactions. In vitro display methods for cyclic peptide discovery have proven highly successful in identifying potent inhibitors but are generally restricted to the selection of binders to purified and immobilized target proteins. A second approach is to use cellular selections to discover new cyclic peptides, which can possess advantages that complement some drawbacks of in vitro methods, such as the ability to select for more complex inhibitory activities, the lack of requirement for purified target protein, and the ability to identify cyclic peptides active in the complex cellular environment. However, the current technology for cellular cyclic peptide selection has notable limitations such as smaller library sizes, lower HIT potencies, lack of counter-selection to weed out promiscuous hits or false positives, slow rate of selection, and limited non-canonical amino acid incorporation. To advance the state of the art of cellular cyclic peptide discovery, we propose a new platform for cyclic peptide selection based upon phage-assisted continuous evolution. We demonstrate a working system that uses an activity-based selection for inhibiting target protein aggregation in cells and can employ negative selection to purge promiscuous hits and false-positives. We will augment our platform with additional activity- based selections for both inhibitors and inducers of protein-protein interactions, develop methodology for continuous in vivo mutagenesis that will address library size limitations of cellular selections, and incorporate non-canonical amino acids that enable the biosynthesis of both side-chain and backbone modified cyclic peptides to improve the diversity and potentially bioavailability of cyclic peptides identified through our platform. If successful, this work will provide a powerful new platform for selecting structurally diverse cyclic peptides with a wide range of inhibitory activities.