Few-Shot Identification of Active Therapeutic Peptide Networking with Gradient Dynamics against Cancer.

The treatment of osteosarcoma remains challenging due to limitations of chemotherapy. Anticancer peptides present a promising avenue as alternative therapeutic agents; however, they encounter significant drawbacks, including susceptibility to proteolytic degradation and constrained selectivity toward tumor. Herein, a few-shot peptide screening strategy is developed, yielding identification of RT2 (tumor-lytic peptide) and E16 (ionic-responsive self-assembling peptide). Their fusion E16-RT2 (ER) is engineered, exhibiting 175% enhanced tumor-selective cytotoxicity and cancer membrane targeting. By leveraging the dual functionality of ER, an injectable hydrogel encapsulating doxorubicin is developed, enabling gradient dynamics for tumor-specific drug release and sustained therapeutic efficacy (>9 d). Moreover, there is no systemic/organ toxicity in murine models dosed with the therapeutic network. The potential mechanism is proposed: charge-guided ER localization disrupts cancer membrane integrity, enabling therapeutics influx while sparing healthy cells, coupled with network degradation kinetics ensuring prolonged drug retention. This rational design paradigm establishes a clinically translatable platform combining tumor-perturbing peptides with smart biomaterials, effectively overcoming limitations of conventional agents. The demonstrated "safety-efficacy" profile positions this approach as a viable strategy for cancer treatment.