Curvature-sensing peptide inhibits tumour-derived exosomes for enhanced cancer immunotherapy

Tumour-derived exosomes (T-EXOs) impede immune checkpoint blockade therapies, motivating pharmacological efforts to inhibit them. Inspired by how antiviral curvature-sensing peptides disrupt membrane-enveloped virus particles in the exosome size range, researchers at Sungkyunkwan University devised a broadly useful strategy that repurposes an engineered antiviral peptide to disrupt membrane-enveloped T-EXOs for synergistic cancer immunotherapy. The membrane-targeting peptide inhibits T-EXOs from various cancer types and exhibits pH-enhanced membrane disruption relevant to the tumour microenvironment. The combination of T-EXO-disrupting peptide and programmed cell death protein-1 antibody-based immune checkpoint blockade therapy improves treatment outcomes in tumour-bearing mice. Peptide-mediated disruption of T-EXOs not only reduces levels of circulating exosomal programmed death-ligand 1, but also restores CD8⁺ T cell effector function, prevents premetastatic niche formation and reshapes the tumour microenvironment in vivo. These findings demonstrate that peptide-induced T-EXO depletion can enhance cancer immunotherapy and support the potential of peptide engineering for exosome-targeting applications.

Repurposed antiviral AH-D peptide disrupts T-EXOs
with enhanced activity under tumour pH conditions

a, Illustration of T-EXO depletion strategy. b, NTA-tracked change in normalized number concentration of T-EXOs from human and murine cancer cells following T-EXO incubation with 1 μM AH-D or NH-D peptide for 10 min. WM-266, MDA-MB-231, B16F10 and 4T1 are human melanoma, human breast cancer, murine melanoma and murine breast cancer cell lines, respectively. c,d, Corresponding NTA results for B16F10-derived T-EXOs following incubation with 1 μM AH-D or NH-D peptide for different time intervals (c) or for 5 min under varying pH (d). b–d, Results reported as mean ± s.d. (n = 3 biological replicates, one-way analysis of variance (ANOVA)). e,f, Maximal QCM-D changes in resonance frequency (e) and energy dissipation (f) signals following the addition of 32 μM AH-D peptide to a layer of surface-adsorbed liposomes under varying pH. e,f, Results reported as mean ± s.d. (n = 4 biological replicates, one-way ANOVA). g, Binding cooperativity of AH-D peptide-induced liposomal membrane rupture under varying pH. Results reported as best-fit values ± s.d. from least-squares regression (n = 16 independent experiments, one-sided extra sum-of-squares F-tests between groups). h, Change in Gibbs free energy for membrane partitioning of AH-D peptide under varying pH. Modelling based on the Wimley–White interfacial hydrophobicity scale; dashed lines correspond to all charged (top) or all neutral (bottom) Asp residues in the peptide.