De Novo Design Peptide Probe Enables Super-Resolution Imaging of MFGE8−Integrin Nanoscale Signaling Platforms in Triple-Negative Breast Cancer

Abstract Super-resolution imaging based on conventional antibody labeling can introduce linkage amplification effects, while the resulting antibody complexes may also cause substantial steric hindrance. Therefore, nanoscale imaging of ligand proteins on the cell membrane under low-perturbation conditions remains challenging. In this study, we developed a structure-guided and artificial intelligence-assisted strategy for the de novo design of MFGE8-targeting peptide probes. Electrostatic potential analysis identified an exposed pocket around residues 121−126 of MFGE8, followed by RFdiffusion-based backbone generation and ProteinMPNN-mediated sequence optimization. The binding capability of the designed peptides was validated by surface plasmon resonance. The optimized Cy5-labeled peptide probe exhibited a smaller molecular size and shorter labeling distance, making it suitable for dSTORM super-resolution imaging, while showing a high spatial correlation with antibody labeling. Using this probe, we found that MFGE8 membrane nanoclustering increased with the malignant progression of breast cancer cells and was markedly enriched in triple-negative breast cancer cells. Two-color dSTORM imaging further revealed spatial coupling between MFGE8 nanoclusters and integrin αV signaling platforms. Treatment with the cyclic peptide cyclo(RGDyK) reduced the colocalization between MFGE8 and integrins, attenuated MFGE8 clustering, and reversed epithelial−mesenchymal transition-associated molecular features; these findings were further validated by MFGE8 knockdown experiments. Collectively, this study not only reveals the functional role of MFGE8 membrane nanoassembly in integrin signaling activation and EMT progression but also establishes an artificial intelligence-assisted strategy for developing super-resolution imaging probes, providing a new tool and insight for resolving the spatial architecture of membrane ligand−receptor complexes and exploring therapeutic intervention strategies for triple-negative breast cancer.