AI Designed Conformation Locking Peptides Target STING to Restore Diabetic Wound Healing

Abstract

ABSTRACT Diabetic foot ulcers are a major complication of diabetes characterized by persistent inflammation and impaired tissue repair, in part driven by aberrant activation of the cGAS‐STING innate immune pathway. Precision immunomodulation in the protease‐rich wound microenvironment remains challenging because therapeutic efficacy requires both localized retention and responsiveness to pathological cues. Here, we developed an integrated AI‐to‐biomaterial strategy for diabetic wound repair by coupling generative AI‐guided peptide discovery with microenvironment‐responsive local delivery. A structure‐guided deep‐learning pipeline integrating RFDiffusion, ProteinMPNN, and AlphaFold2‐multimer identified SCP‐1, a conformation‐locking peptide designed to stabilize the inactive STING dimer. To enable therapeutic translation, SCP‐1 was incorporated into a dual‐responsive hydrogel (Gel‐SCP‐1) that provides in situ gelation and MMP‐9‐triggered release in the wound bed. Gel‐SCP‐1 suppressed STING‐TBK1‐IRF3 signaling, reduced inflammatory and oxidative stress, promoted reparative macrophage polarization, and enhanced angiogenic activity. In a full‐thickness excisional wound model in db/db diabetic mice, Gel‐SCP‐1 significantly accelerated wound closure and improved tissue regeneration, including enhanced re‐epithelialization and collagen remodeling. These findings establish an AI‐to‐biomaterial therapeutic paradigm for precision immunoregenerative therapy in chronic diabetic wounds.