Abstract
<jats:p>Macrocyclic peptides can bind protein surfaces and enzyme active sites that are difficult to address with small molecules, but their discovery from synthetic libraries remains limited by the challenge of sequencing intact cyclic structures by tandem mass spectrometry. Existing affinity selection–mass spectrometry workflows often require post-selection chemical linearization, which adds processing time and complicates hit identification. Here, we report a macrocyclic peptide discovery platform that allows for direct sequencing and leverages the collision-induced fragmentation chemistry of disuccinimidyl sulfoxide (DSSO) as a selective linearizable linker integrated with fast-flow peptide library synthesis. Optimization of stepped higher-energy collisional dissociation enabled in-instrument macrocyclic linearization and backbone fragmentation, providing high-confidence de novo sequence assignments without post-selection chemical treatment. Comparison of DSSO-, disulfide-, and pimelate-linked libraries across multiple scaffold lengths showed that selective linearization is essential for reliable sequencing; DSSO-linked libraries achieved sequencing coverage comparable to chemically linearized disulfide libraries. Affinity selection against the antihemagglutinin antibody 12ca5 validated the platform by recovering the expected binding motif and identifying macrocyclic binders with single-digit nanomolar affinities. By incorporating non-natural functionalities, application to matrix metalloproteinase-8 (MMP-8) enabled the discovery of macrocyclic peptide inhibitors from a 2.48-million-member library, including a lead candidate with an IC50 of 35 nM, up to 291-fold selectivity across tested MMP subtypes, and enhanced serum stability relative to its linear analogue. Co-crystal structures of the lead inhibitors in complex with MMP-8 revealed the molecular basis of inhibition and target selectivity. This integrated platform accelerates synthetic macrocyclic peptide discovery by combining rapid library preparation, direct de novo sequencing, and functional hit identification without post-selection chemical linearization.</jats:p>