173 An in vivo CRISPR/Cas9 screening platform to identify T cell enhancing edits in distinct solid tumor microenvironments

Background Chimeric antigen receptor (CAR)-based T cell therapy and other forms of adoptive cell therapies (ACTs) have shown remarkable success in the treatment of hematologic malignancies; however, reports of clinical activity in solid tumors are limited to date. One key therapeutic challenge presented by solid tumors is the immunosuppressive tumor microenvironment (TME). Adding to the complexity, it is becoming increasingly clear that TMEs are heterogeneous (broadly classified as ‘inflamed,’ ‘immune excluded’ and ‘immune dessert’), utilizing different mechanisms of immunosuppression. Instrumental to overcoming the barriers presented by solid tumors will be the development of T cells with immune- enhancing edits that improve penetration, potency and persistence, while also preventing exhaustion in hostile TMEs. T cells with these properties may help in the development of ACTs in solid tumors. Methods CRISPR/Cas9-based functional genetic screens in T cells can enable prioritization of known targets and uncover novel targets to improve the design of genetically reprogrammed cell therapies, in an unbiased fashion. Most CRISPR screens to date have been performed in vitro with tumor cells due to the complexity of setting up CRISPR screens in primary T cells, particularly for in vivo target discovery. Here, we describe the development and careful optimization of an in vivo mouse CRISPR-screening platform to identify knock-out targets in primary T cells, with the goal of increasing T cell abundance and persistence in tumors with different TMEs. Using a mouse retroviral system to express single-guide RNA (sgRNA) libraries in T cells from Cas9 transgenic mice, we performed in vivo screens in syngeneic, fully immune-competent mouse tumor models. Results We identified both known and potential novel regulators of T cell activation and persistence. Importantly, we have discovered knock-out targets that accumulate in multiple, distinct TMEs and other targets that are TME-specific. The use of sub-genomic- focused libraries allowed us to rapidly screen in multiple tumor model systems and reproducibly identify hits across individual mice. Conclusions We have developed a fully optimized an in vivo genetic screen, which could be a rich source for target discovery, and can enable identification of functional regulators of T cells for rapid incorporation into CRISPR-engineered T cell therapies for different solid TMEs.