PROJECT SUMMARY Bicyclic peptides are conformationally constrained peptides comprised of two macrocyclic rings. Owing to their conformational rigidity, bicyclic peptides are highly resistant to proteolysis, and can bind to protein targets with antibody-like affinity and selectivity. As a result, these molecules are highly desirable scaffolds for the development of peptide-based therapeutics. Phage display is a laboratory evolution technique that enables the discovery of high-affinity peptide ligands from large, combinatorial peptide libraries. Although originally limited to linear peptides, phage display was recently adapted for the discovery of novel bicyclic peptide ligands. Most often, phage-displayed bicyclic peptide libraries are prepared by chemically modifying linear peptides using cysteine-reactive small molecules; however, this method is time consuming and technically challenging. As a result, phage-displayed bicyclic peptide technology has not been widely adopted. Recently, several studies have used genetic code expansion to install cysteine-reactive noncanonical amino acids (ncAAs) into phage-displayed peptides to produce libraries of cyclic peptides. This strategy has significant advantages over the chemical cyclization approach, but is currently limited to monocyclic peptides. The overarching objective of this proposal is to develop technology that enables phage display of bicyclic peptides using genetic code expansion. Our central hypothesis is that bifunctional ncAAs, i.e. ncAAs containing two cysteine-reactive functional groups, can be used to generate ribosomally synthesized bicyclic peptides by intramolecular reaction with two cysteine residues. To realize our objective, we will pursue three Specific Aims. In Aim 1 we will engineer an aminoacyl-tRNA synthetase that recognizes bifunctional ncAAs containing two cysteine-reactive moieties. This will be accomplished using traditional and state-of-the-art methods of directed evolution. In Aim 2 we will develop a phage display system that is compatible with co-translational installation of bifunctional ncAAs and we will optimize this system for efficient bicyclic peptide formation. We will then validate this system by selecting and characterizing bicyclic peptide ligands for two model targets. In Aim 3 we will use our phage-displayed bicyclic peptide libraries to identify peptides that bind to the spike protein of human coronaviruses and inhibit virus-host membrane fusion. By targeting spike proteins from diverse coronaviruses, we will strive to identify inhibitors with broad-spectrum antiviral activity. The proposed work will provide a facile route for generating bi- cyclic peptide libraries thereby greatly accelerating the discovery of therapeutic peptide leads.