A High-Throughput Exploration of the CAR-T Signaling Landscape Identifies Novel Domains for Next-Generation Immunotherapies

Chimeric Antigen Receptors (CARs) are the first successful synthetic immune receptors for adoptive T cell therapy. CARs with different signaling domains (such as CD28 and 4-1BB) can modulate distinct T-cell functions including cytokine secretion, cytotoxicity, and overall persistence in the patient. Identification and optimization of new CAR signaling domains will help to extend their therapeutic applications to solid tumors and indications beyond cancer. Here we use a pooled screening approach which combines fluorescence-activated cell sorting and high-throughput sequencing to conduct a broad search of the CAR signaling landscape in primary human T cells. We measured CARs containing 40 different domains - including those not normally associated with T-cell costimulation - for activation, cytokine secretion, and proliferation in response to multiple weeks of repetitive antigen challenge. These domains differed in activity between CD4 and CD8 T cells and between early and late proliferators. We identified and further characterized several potent costimulatory domains beyond 4-1BB and CD28 which enhance proliferation (CD40) and cytotoxicity (BAFF-R and TACI) after repeated antigen exposure in vitro and in vivo. We also identified a potent co-inhibitory domain, KLRG1, which strongly attenuates multiple aspects of T cell function. Via single-cell RNA and surface protein analysis, we show that these domains are enriched in distinct functional states which differ in markers of T cell memory, metabolism, and innateness. In particular, CAR-T cells containing the BAFF-R costimulatory domain, normally associated with B-cell survival, exhibit improved cytotoxicity, proliferative capacity, and a functional program which is correlated with positive clinical outcomes in multiple studies. These new domains could enhance CAR clinical efficacy and their identification shows how mapping the costimulatory signaling landscape will enhance our ability to engineer diverse functions into therapeutic primary human T cells.