New Peptide Linker Design Makes Cancer-Killing Antibody Drugs More Effective at Destroying Nearby Tumor Cells

Tripeptide linkers with all-L or single D-alanyl residues effectively released cytotoxic maytansinoid metabolites inside cells. D-alanyl residues didn't impair activity unless directly attached to the self-immolative group. Extending the maytansinoid side chain increased bystander killing without affecting direct cytotoxicity. The best-performing ADCs showed improved in vivo efficacy compared to previous maytansinoid conjugate designs.

Conjugates were synthesized with varying peptide linker stereochemistry and maytansinoid side chain lengths. In vitro testing assessed direct cytotoxicity and bystander killing in target-positive and target-negative cells. Best candidates were evaluated in vivo in mouse tumor models and compared to previously described maytansinoid ADC types.

Improved bystander killing means ADCs can kill not just the tumor cells they bind to but also nearby cancer cells that may not express the target antigen — addressing a key limitation of conventional ADC therapy in heterogeneous tumors.

ADCs are one of the fastest-growing cancer drug classes (>15 approved), but tumor heterogeneity limits their effectiveness — not all cancer cells in a tumor express the antibody target. Bystander killing addresses this by allowing the released cytotoxic payload to kill neighboring target-negative cells. Optimizing the peptide linker that connects drug to antibody is critical for controlling when and how the drug is released. This study's systematic approach to linker design — testing stereochemistry and payload chemistry — advances the rational engineering of next-generation ADCs.

Mouse xenograft models may not predict human therapeutic responses. The study focuses on one antibody and one payload class (maytansinoids). Long-term safety and manufacturing feasibility at scale were not addressed. Bystander killing may also affect nearby healthy tissue, potentially increasing toxicity.