Exosomes in cancer drug resistance: dual roles in therapy failure and emerging precision therapeutics
摘要
Exosomes play a key role in cancer, functioning both as drivers of drug resistance and as tools for therapy. Tumor-derived exosomes facilitate intercellular communication through selective transfer of bioactive cargo, including proteins (e.g., P-gp, PD-L1), nucleic acids (e.g., miR-21, lncRNA H19), lipids, and metabolites. These cargos remodel the tumor microenvironment, activate oncogenic pathways such as PI3K/AKT and MAPK, and promote drug resistance, immune evasion, and metabolic reprogramming. This work presents a conceptually integrated translational perspective, introducing the ‘exosome paradox’ to unify the dual roles of exosomes in therapy resistance and therapeutic application. The review incorporates cancer-specific insights, highlighting how resistance mechanisms vary across pancreatic, breast, lung, colorectal, and ovarian cancers, underscoring the need for context-dependent strategies. Mechanistically, we integrate emerging advances in exosome heterogeneity, selective cargo loading (including SUMOylation-dependent pathways), and the metabolic-epigenetic interface, particularly the lactate-lactylation axis. We also explore the growing role of artificial intelligence in biomarker discovery, cargo profiling, and rational design of engineered exosomes. Therapeutically, two complementary strategies are emphasized: inhibition of pathogenic exosome biogenesis, secretion, and uptake, and development of engineered exosomes as delivery vehicles for chemotherapeutics, RNA-based therapies, and CRISPR/Cas9 systems. Despite promising preclinical advances, clinical translation remains limited by challenges in targeting specificity, pharmacokinetics, scalable manufacturing, and regulatory standardization. This review provides a unified framework combining suppression of pathogenic exosome signaling with strategic engineering of therapeutic exosomes, offering new directions for overcoming drug resistance and advancing precision oncology.