Structural insights into frog skin-derived cyclic peptides as selective matriptase inhibitors

Matriptase is a type II transmembrane serine protease implicated in epithelial integrity and cancer progression, making it a promising therapeutic target. However, the development of highly selective matriptase inhibitors remains a major challenge due to its structural homologies and similar proteolytic specificities with related proteases such as TMPRSS6. In this study, we investigated a series of cyclic peptides derived from the frog-skin Bowman-Birk inhibitor HV-BBI, focusing on their inhibitory activity and selectivity toward matriptase. Among them, compound 6, a 13-residue monocyclic peptide with a C-terminal amide, emerged as a standout candidate, demonstrating nanomolar potency and remarkable 1000-fold selectivity over TMPRSS6. Unexpectedly, when the C-terminal amide was replaced with a carboxylate group, the compound exhibited a significant loss of potency against matriptase. Moreover, structure-activity relationship studies, including alanine scanning and C-terminal residue substitutions, revealed that residues Lys5, Ser6, Ile7, and Arg10 are critical for matriptase binding, while the C-terminal Phe13 is less essential. Molecular dynamics simulations highlighted key inter- and intramolecular interactions that stabilize the peptide-enzyme complex and explain the superior potency of compound 6 compared to its analogues. Importantly, compound 6 retained its inhibitory activity in a cellular model without inducing cytotoxicity. These findings establish HV-BBI-derived peptides as promising scaffolds for the development of highly selective matriptase inhibitors and provide valuable structural insights for future therapeutic design.