Conformation-selective rather than avidity-based binding to tumor associated antigen derived peptide-MHC enables targeting of WT1-pMHC low expressing cancer cells by anti-WT1-pMHC/CD3 T cell engagers.

T cell engagers, a category of T cell-retargeting immunotherapy, are rapidly transforming clinical cancer care. However, the lack of tumor-specific targets poses a significant roadblock for broad adaptation of this therapeutic modality in many indications, often resulting in systemic on-target off-tumor toxicity. Though various tumor-derived intracellular mutations provide a massive pool of potential tumor-specific antigens, targeting them is extremely challenging, partly due to the low copy number of tumor associated antigen (TAA)-derived pMHC on tumor cell surface. Further, the interplay of binding geometry and format valency in relation to the capacity of a T cell engager to efficiently target low density cell-surface pMHC is not well understood. Using the Wilms' tumor 1 (WT1) oncoprotein as a proof-of-principle TAA, combined with an array of IgG-like T cell engager modalities that differ in their anti-TAA valency and binding geometry, we show that the ability to induce an immunological synapse formation, resulting in potent killing of WT1 positive cancer cell lines is primarily dependent on the distinct geometrical conformations between the Fab arms of anti-WT1-HLA-A*02:01 and anti-CD3. The augmented avidity conferred by the binding of two anti-WT1-HLA-A*02:01 Fab arms has only minimal influence on cell killing potency. These findings demonstrate the need for careful examination of key design parameters for the development of next-generation T cell engagers targeting low density TAA-pMHCs on tumor cells.