A clinical data- and AI-supported engineered fusion vesicle system for enhanced tumor-targeted drug delivery and immunochemotherapeutic synergy.
文献摘要
Cell- or bacteria-derived membrane vesicles (MVs) serve as platforms for delivering antitumor small-molecule drugs, characterized by targeted delivery, immunostimulatory activity, and ease of engineering. Despite extensive innovative research, clinical translation remains limited. Here, we report a clinically data- and AI- supported engineered fusion vesicle delivery system, ECMV RGD . The design was supported and guided by a systematic meta-analysis identifying clinically safe and potentially effective RGD peptides and AI molecular dynamics simulations to determine optimal insertion sites and copy numbers (one RGD [RGD4C] each at Loop 2 and Loop 3 of OmpA) on a previously validated safe and effective engineered membrane vesicle (EMV) platform, which was further fused with cancer cell-derived membrane vesicles (CMVs). This engineering strategy not only places translational potential at the forefront from the outset but also when loaded with doxorubicin (DOX), demonstrates comprehensive improvements over CMV in both tumor-targeted delivery and antitumor immune activation in vitro and in vivo. Mechanistic studies further revealed that ECMV RGD @DOX achieved complete recurrence-free antitumor immunochemotherapeutic efficacy (0/6 recurrence) by synergistically promoting dendritic cell (DC) activation and CD8 + T cell infiltration. This engineered fusion vesicle platform provides a novel strategy to advance the design of MVs-based delivery systems with improved translational potential.