An allogeneic gamma/delta armored Dual-Targeting CAR with Split Costimulatory Domains to Target Epithelial Ovarian Carcinoma

Abstract CAR T cell therapies have produced robust remissions in several hematological malignancies. However, they have not achieved the same success against solid tumors. These treatments fail for multiple reasons, including antigen escape and poor T cell persistence in the immunosuppressive tumor microenvironment. Additionally,

many patients receiving T cell therapy experience serious, sometimes life-threatening side effects. Here, we propose a CAR T product – an armored split CAR – that combines 2 innovative strategies specifically tailored to address the unique challenges presented by solid tumors, as well as an inducible suicide switch to enhance

safety. Moreover, we are developing this product on our proprietary allogeneic γδ T cell platform. These cells display innate anti-tumor properties and do not mediate GVHD. We use a “Split CAR” approach, wherein two CAR molecules, each targeting a different antigen, are expressed by a single T cell and dimerize upon antigen

binding. By targeting 2 antigens, we reduce the potential for antigen escape. Each CAR molecule contains either a CD28 or 4-1BB costimulatory domain which are independently activated to cooperatively enhance cytokine release, cell survival, and proliferation. The heterodimerized CARs trigger activation through a single shared

CD3ζ domain incorporated on one of the CAR chains. In addition, our engineered T cells express a binding molecule to sequester and disable the tumor-secreted immunosuppressive cytokine, TGFβ. Lastly, we are including a tEGFR safety switch so that the clinical product can be rapidly eliminated in vivo in case of severe

side effects. Efficient gene delivery of such a large cassette is made feasible by using the Tc Buster transposon system, which can stably integrate transgenes as large as 10 kb with an acceptable copy number and a superior safety profile compared to viral transduction. Transposon-based gene delivery also circumvents the long

manufacturing timelines and bottlenecks that beset viral methods. We are adapting the armored split CAR to treat epithelial ovarian carcinoma by targeting the tumor antigens mesothelin (MSLN) and chondroitin sulfate proteoglycan-4 (CSPG4). These antigens are overexpressed in at least 40% of ovarian cancer patients and are

associated with poor outcomes. We will test our armored split CAR in vitro using a co-culture target cell killing assay with ovarian cancer lines that are single- or double-positive for MSLN and CSPG4. We will test how well the engineered cells resist exhaustion in a serial killing assay in which T cells are exposed to antigen in co-

culture over ~14 days and multiple passages, then assessed for relevant activation and exhaustion markers. We will also test the tEGFR safety switch by culturing the cells with the inducer, cetuximab, and measuring selective lysis of engineered cells. Lastly, we will assay tumor killing by the armored split CAR in a mouse model of ovarian

cancer. These experiments are designed to collectively demonstrate efficacy of our product under stress conditions and will lead into anticipated Phase II preclinical studies to assess toxicity and safety.