Improving CAR-T cell immunotherapy in immunocompetent sarcoma models

Cancer immunotherapies have revolutionized how some tumor types are treated. However, evidence for their benefit in sarcoma patients remains limited with only small subsets of patients responding to treatments. Impressive results have been observed with T-cell receptor (TCR)-based adoptive T-cell therapy in synovial sarcoma. Still, not all patients respond and the development of improved adoptive T cell therapies that could potentially be applied to a wider range of sarcoma types is needed. Chimeric antigen receptor (CAR) T cells have been developed to bind directly to antigens present on the tumor cell surface. While sarcoma-associated antigens amenable to CAR-T cell treatment have emerged, CAR-T cell therapy in solid suffers from a number of hurdles, such as limited tumor infiltration or lack of efficacy. Recent efforts using genome-wide CRISPR/Cas9-based screens identified novel regulators of CAR-T cell function in other tumor types. However, most of these findings were made using in vitro assays or immunocompromised mouse models which cannot model the complex immunological signaling crosstalk that occurs in immunocompetent hosts. The tumor microenvironment (TME) harbors a variety of immunosuppressive cell types that dampen the effects of immunotherapy and can lead to an T cell exclusion phenotype or accumulation of dysfunctional T cells. In this proposal we will take advantage of immunocompetent sarcoma models we have recently developed using electroporation-based delivery of transposon vectors to the muscle. We will concentrate on two of the models we have characterized and represent sarcomas with varying degree of immune cell infiltration: synovial sarcoma (SS, SS18-SSX1-driven) with low immune cell infiltration, and undifferentiated pleomorphic sarcoma (UPS, driven by KRASG12V and p53 loss) with higher immune cell infiltration. We will use CD276 (B7-H3) as a prototype pan-cancer antigen that can serve as CAR-T cell target across different sarcoma models. We will first characterize the sarcoma tumor microenvironment in these models with and without CAR-T cell treatment, and identify the CAR-T baseline activity that can be achieved in these contexts. In a second aim we will apply loss of function CRISPR/Cas9 screens in CAR-T cells in vivo to identify candidate genes that improve T cell infiltration and activity in immunocompetent hosts. To this end tumor-bearing mice will be treated with a library pool of CAR-T cells and sgRNAs present in tumor-enriched CAR-T cell populations will be identified by next generation sequencing. High confidence candidates can then be validated in mouse and human models such as patient-derived xenografts. The results of this project will potentially lay the experimental basis to refine cell-based immunotherapies for sarcoma and other solid tumors, providing a much needed preclinical rational to develop new and innovative clinical cancer treatment trials in the future.