Ameliorating off-target toxicities of CAR T cells by engineering NOT gates

Project Summary/Abstract Engineered therapeutic T cells have shown transformative success in treating B cell cancers but applying this approach to solid tumors has proven far more difficult. There do not appear to be absolutely tumor-specific single antigen targets for solid cancers, and thus, CAR T cells that attack most tumor-associated antigens have led to toxic cross-reaction with normal organs that also express the antigen. If we are to successfully and safely treat solid tumors with CAR T cells, it will be essential to mitigate toxic cross-reaction with normal tissues. To prevent off-tumor toxicity of therapeutic T cells, we propose to engineer multi-receptor T cell circuits that can recognize a tumor based on a multi-antigen profile. In this proposal, we focus specifically on engineering NOT gate circuits -- circuits that can override and inhibit CAR T cell activation and killing upon detecting an antigen that is uniquely indicative of a cross-reactive normal tissue (i.e., antigen is absent in the tumor). Our recently published bioinformatic analysis shows that there are numerous tissue-specific antigens that could be used as signals to induce T cell inactivation in common cross-reactive tissues like the brain and lung. Nonetheless, there is currently a lack of robust NOT-gate circuits demonstrated to work well in tumor models. Thus, we will develop and test new NOT circuits that can inactivate a CAR T cell in an antigen-induced manner. Our specific aims are: Aim 1. Engineer, prototype and optimize new NOT gate circuits that use diverse mechanisms to block therapeutic T cell activation in antigen-induced manner Aim 1.1. T cell NOT gates using transcriptional repressors of CAR expression. Aim 1.2. T cell NOT gates that inhibit T cell proliferation by antigen-induction of cell death effectors. Aim 1.3. T cell NOT gates that locally induce production of secreted immunosuppressive factors (paracrine) Aim 2. Applying NOT gate circuits to prevent anti-GD2 CAR T cross-reaction with brain/CNS tissue. Aim 2.1. in vitro prototyping of NOT gate circuit targeting the brain antigen MOG to turn off anti-GD2 CAR Aim 2.2. Test if brain NOT gates block CNS toxicity of anti-GD2 CAR T cells in vivo, while maintaining efficacy against murine neuroblastoma xenograft models (GD2+). Aim 2.3. Test in vivo safety & efficacy of NOT gate CAR T cells in an immunocompetent model of neuroblastoma. This work should provide important general capabilities for engineering CAR T cells that selectively turn themselves OFF when they are in the wrong, cross-reactive tissue. These are much needed tools that are currently missing in the toolbox for T cell engineering, but which will be critical for engineering T cells that safely treat solid cancers.