Abstract Chimeric antigen receptor (CAR) protein are artificial protein created by the fusion of an extracellular domain targeting one or several cell surface antigens, a transmembrane domain and an intracellular domain responsible for signal initiation and transmission. First-generation CARs consist of an intracellular signaling domain, often a cytoplasmic tail of the CD3ζ chain. The second and third generation include the addition of one or two co-stimulation domains (e.g. cytoplasmic domain of CD28 and/or 4-1BB). NKG2D is a receptor present on the cell surface of NK cells, NK T cells and subgroups of CD8+ T cells. The receptor exists as a dimer that interacts with a co-adaptor protein called DAP10. Upon interaction of NKG2D with one of its ligand (MICA, MICB, and ULBP1-ULBP6), the DAP10 cytoplasmic tail will induce downstream signaling. DAP10 has a similar signaling motif to CD28 and similarly gets phosphorylated and recruits kinases. NKG2D ligands are expressed on a vast majority of solid and liquid tumors and are thus a very attractive target for CAR T cells. In this study, multiple NKG2D constructs were created and compared to one another. The first CAR construct tested (chNKG2D.1) consists of a fusion of full-length wild-type NKG2D and the cytoplasmic domain of CD3ζ. This construct is currently tested in clinical settings. This CAR benefits from the co-stimulatory signal provided by DAP10 and can therefore be considered as a second generation CAR. Next to this, a new chNKG2D was constructed by combining DAP10 overexpression (chNKG2D.1.DAP10). Secondly classic second-generation CARs were created by adding CD28 (chNKG2D.2.1) or 4-1BB (chNKG2D.2.2) co-stimulatory domains to the chNKG2D. Co-expression of DAP10 increased chNKG2D expression at the cell surface of both CD4 and CD8 T cells (although the increase was more pronounced in CD4 T cells). Furthermore, co-cultures of T cells expressing this construct with K562 cells led to similar IFN-γ levels irrelevant of the CD4/CD8 ratio. Other CD4 specific cytokines were highly increased when the ratio was in favor of CD4 positive T cells. Interestingly, the addition of the co-stimulatory domains to the construct (chNKG2D.2.1 and .2.2) did not lead to higher levels of IFN-γ or any of the CD4 cytokines. Next to cytokine release, cytolytic activity was also assessed which interestingly showed no difference in any of the conditions even when the CD4/CD8 cell ratio was different. In conclusion NKG2D CAR T cells only showed increased cytokine release when DAP10 was overexpressed, all other conditions did not lead to any changes when compared to the ‘NKG2D CAR T cells currently tested in the clinic. This implies that NKG2D is optimally costimulated through the DAP 10 co-signaling, and that this co-signaling is a least as potent as traditional CD28 or 4-1BB based co stimulation. Citation Format: Jennifer BOLSEE, Eytan BREMAN, Fanny HUBERTY, Benjamin VIOLLE, Jerome MARIJSSE, Céline JACQUES-HESPEL, Celine MARCHAND, Nancy RAMELOT, Thuy NGUYEN, Julien HOUSSA, David E. GILHAM, Sophie AGAUGUE. NKG2D as a chimeric antigen receptor - DAP 10 provides optimal co stimulation for NKG2D based CARs [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3583.
Read full abstract