Cancer-testis antigen specific chimeric antigen receptor T cells for cancer immunotherapy

Adoptive chimeric antigen receptor (CAR) T cell therapy has shown promising results in early phase clinical trials, in particular when treating hematological malignancies. In this novel immunotherapeutic approach, isolated autologous T cells are ex vivo engineered to express a CAR, expanded to obtain sufficient cell numbers and finally infused back to the patient. Such CAR T cells are redirected to a predefined target antigen by an antibody derived binding domain and have been shown to substantially reduce the tumor burden as long as the cancer cells present the target antigen. However, solid tumors do not express exclusive target antigens, thereby hampering therapeutic safety due to potential on-target/off-tumor toxicity. In this project, we developed CARs that contain phage display selected fully human antibody binding domains specific for the tumor associated antigens AKAP4, SP17 and SPAG9. In healthy tissues, the expression of these so called cancer-testis antigens is restricted to spermatocytes, however, it has been shown that their expression can be upregulated in certain tumor entities most probably due to deregulated epigenetic events. We successfully selected AKAP4, SP17 and SPAG9 antibody fragments (scFvs) from human B cell repertoires that were further characterized as soluble scFv-hFc fusion proteins. We demonstrated antigen specificity by ELISA and SPR spectroscopy as well as surface binding to several tumor cell lines representing colorectal, ovarian, plasma cell and B cell cancers. Importantly, we also observed binding of the scFv-hFc fusion proteins to primary plasma cells isolated from multiple myeloma patients, but neither to plasma cells from two 2 out of 3 patients showing unremarkable bone marrow morphology, nor to peripheral blood mononuclear cells derived from healthy donors. SP17 and SPAG9 specific scFv lead candidates were fused as antigen binding domain to five different CAR backbones that contained either CD8α or hFc extracellular spacer domains and intracellular second or third generation stimulatory domains. The second generation CAR constructs showed superior characteristics compared to the third generation CARs as indicated by higher transduction efficiencies of the T cells and a higher surface expression of the CAR molecule. In addition, T cells expressing the second generation CARs showed less signs of T cell exhaustion, preferentially differentiated toward the central memory phenotype and showed specific cytotoxicity towards ovarian cancer and multiple myeloma cell lines. We here provide the basis for a novel cellular therapy by redirecting CAR T cells to tumor specific cancer-testis antigens to safely treat various cancer entities.