Preclinical development of universal SNAP-CAR T cell therapy

Abstract

Abstract Chimeric antigen receptors (CARs) are artificial T cell receptors that re-target patients’ T cells to specifically recognize and kill tumor cells. Despite remarkable success of anti-CD19 CAR therapy against refractory B cell leukemias, there are several limitations to CAR T cell therapies including toxicities and antigen-loss leading to relapse. To address these issues, we previously developed a “universal” CAR, SNAP-CAR, for which antigen-specificity is directed by co-administered tumor-targeted antibodies. Instead of directly recognizing a tumor antigen, the SNAP-CAR carries out an enzymatic reaction to fuse with antibodies conjugated to a benzylguanine (BG) tag. Activation and effector functions of SNAP CAR T cells can be re-targeted by antibody-BG conjugates to several antigens including: CD20, CD19, HER2, and EGFR. We are now developing SNAP-CAR T cells for potential clinical translation including optimization of the CAR expression construct and functional characterization in a human tumor xenograft mouse model. SNAP-CAR-T2A-LNGFR was cloned into a gamma-retroviral expression system, yielding a 10-fold greater expression level in primary human T cells compared to our previous lentiviral vector. SNAP-CAR T cells were potently activated and lysed tumor cells in a target antigen-specific manner in vitro. Challenging NSG mice with CD20+ Raji leukemia cells, significant tumor regression was observed with SNAP-CAR T cells + Rituximab-BG and anti-CD20-CAR positive control T cells but not SNAP-CAR T cells without antibody, as compared to untransduced T cells + Rituximab-BG. The SNAP-CAR provides a powerful system to program the antigen-targeting capability of human T cells with promise for cancer therapy.