Amélioration de l'efficacité des cellules CAR-T in vivo par l'édition de gènes cibles sélectionnés

In the last few decades, cancer has imposed itself as the first cause of mortality in developed countries, a trend that is poised to confirm itself in the coming years. Cancer is a complex disease, many therapeutic options exist today, but despite that, roughly a third of cancer patients is not cured and ultimately dies of their cancer. Among the therapies against cancer, immunotherapy, which relies on strategies to activate the immune system to fight cancer, has recently achieved great successes against previously uncurable tumors. CAR-T cells are a type of immunotherapy that relies on equipping T cells from a patient with a chimeric antigen receptor to allow better tumor cell recognition and expand them ex-vivo prior to reinfusion. These therapies have achieved remarkable results in the treatment of refractory and relapsing hematologic malignancies, obtaining complete responses in patients that previously had no other therapeutic option. Nonetheless, many patients do not achieve durable responses, and these therapies remain generally ineffective against solid tumors despite recent promising advances. Therapeutic failure is closely associated with a poor expansion and persistence of CAR-T cells in patients, highlighting the need to identify strategies to promote in vivo expansion. Several strategies are being evaluated to improve CAR-T cell efficacy, such as improvements in the CAR construct signaling or in the CAR-T cell manufacture pipeline and others. Gene editing approaches, especially by gene inactivation, are among the most promising of these amelioration strategies, and several different approaches have been proposed to date. During my PhD thesis, I could participate in different studies from my host laboratory, broadly addressing different approaches to increase the efficacy of CAR-T cell therapy. Notably, addressing the contribution of metabolism and motility in CAR-T cell efficacy, studying the synergistic effects of interactions of CAR-T cells with other immune cells, as macrophages, and investigating the role of the retinoid-inducible nuclear factor, RINF, for the amelioration of CAR-T cells. Nonetheless, the focus of my project was to investigate, by a comparative approach, the efficacy of the most promising gene targets for CAR-T cell amelioration by gene inactivation, in a model of human lung adenocarcinoma. Using a focused CRISPR-KO library targeting 50 top candidate genes, I developed an in vivo competitive screening that revealed REGNASE-1, SOCS1, PTPN2, and P16INK4A as the most robust targets to improve persistence of EGFR CAR-T cells in human lung tumor-bearing mice. Surprisingly, I could also observe that disruption of other genes, MED12, PRDM1 and BATF, previously reported to improve CAR-T cell efficacy had a detrimental effect in the context of our study. I could also perform this study in the context of a subcutaneous, metastasizing tumor model and applying it to improve other characteristics of CAR-T cells, such as metabolic and migratory features. Overall, these results indicate that some gene-editing strategies can have beneficial, neutral, or even deleterious effects on CAR-T cell persistence and on other features depending on specific conditions. Altogether, these findings validate the pertinence of our approach as versatile pre-clinical tool to prioritize among different gene editing strategies, those most promising for the amelioration of specific CAR-T products, in precise tumor models. Furthermore, they highlight the importance of a context-specific evaluation of genetic modifications, to accelerate the clinical translation of the most promising editing strategies in specific conditions.