Abstract: Classical antibody and chimeric antigen receptor (CAR) T-cell therapies to treat cancer are limited to target cell surface proteins, which only constitute about 20-25% of all available tumor antigens. In addition, these surface antigens are typically not restricted to tumor cells. Targeting peptides presented in major histocompatibility complexes (pMHCs) offers great potential to exploit the intracellular reservoir of proteins associated with cancer. Soluble T cell receptor (TCR) proteins have been used to target pMHC-restricted tumor-specific antigens. However, their low affinity and substantial challenges related to stability and production compromise their developability as drugs. We developed a discovery platform for the selection of TCR-like antibodies with high target specificity and affinity. This provides access to the intracellular antigen repertoire analogous to TCRs but with much higher affinity as well as the clinical validation and format versatility of monoclonal antibodies (mAbs). As proof of concept, we used the HLA-A*02:01 (HLA-A2) restricted MAGE-A4 epitope GVYDGREHTV. A broad collection of antibodies was isolated from scFv phage libraries specifically designed to bind pMHC-complexes. Based on initial compound screening with ELISA and sequence analysis, a set of binders was selected for further SPR affinity characterization. Out of 1482 hits, several binders showed KD values in the picomolar range for the MAGE-A4/HLA-A2 complex and no binding against HLA-A2 in complex with unrelated peptides. Additionally, specific binding in a cellular context was verified for selected leads by flow cytometry using T2 cells pulsed with MAGE-A4 peptide while no binding was observed to T2 cells pulsed with control peptides with high similarity to MAGE-A4. The best performers were selected to generate CD3-based bispecific T cell engagers (TCE), for which potent killing was demonstrated in several MAGE-A4/HLA-A2 positive cancer cell lines of different histological origin, including lung cancer, melanoma, bladder cancer and osteosarcoma, as well as in a xenograft mouse model. No off-target activity was observed in MAGE-A4 negative, HLA-A2 positive control cells. Notably, cytotoxicity was significantly potentiated by concomitant blockade of the PD-1/PD-L1 axis supporting the concept of sensitizing solid tumors with TCEs to immune checkpoint inhibitor (ICI) therapy. In summary, we demonstrated the potential of our platform to deliver TCR-like antibodies with high affinity and specificity against intracellular antigens proven by efficient and specific killing of MAGE-A4/HLA-A2 positive cancer cells with bispecific TCEs. For the first time, we showed the synergistic effect of combining pMHC-specific TCEs with ICI treatment corroborating the potential of this treatment modality to overcome major limitations of current cancer immunotherapies.
This post was written by CDR Life Editor