IMMU-74. Novel screening method to identify potent immune-inspired inhibitors for NOT-gated CAR T cells to suppress on-target-off-tumor toxicity in brain tumors

Abstract CAR T cell therapy is limited in its translation to solid tumors—including brain tumors—in part because of on-target-off-tumor toxicity (OTOT), wherein tumor-associated antigens such as EGFR, B7-H3, and GD2 are also expressed on nearby healthy tissue. NOT-gated CAR T cells address this issue by pairing a tumor-targeting CAR with an inhibitory receptor that recognizes antigens on normal cells (inhCAR). When the inhCAR binds its target, it suppresses CAR T cell activation, mitigating normal cell cytotoxicity. Previous studies—including our own—have identified antigen pairs differentially expressed between tumor and normal tissues that are suitable for such NOT-gated strategies. In brain tumors, where HLA expression is often down-regulated, incorporating inhCARs that recognize HLA or brain-specific markers presents a unique opportunity to enhance tumor specificity. However, most existing immune-inspired inhibitory constructs show incomplete or delayed suppression of CAR T cell activity. To overcome this, we are developing a high-throughput pooled screening platform to identify more potent inhCARs. Our system uses Jurkat T cells engineered to co-express tumor-targeting second- and third-generation CARs along with a library of normal cell-targeting inhCARs. The platform integrates synthetic biology tools: an activation-responsive fluorescent reporter linked to inducible cell death, and an effector-to-target binding read-out. Effective inhibition is indicated by the absence of reporter activation and cell death when both tumor and normal antigens are engaged. We validated this read-out with previously characterized inhCARs and are now constructing a library of >60 single and >3,600 dual/tandem immune-inspired inhibitory constructs for testing on relevant brain-tumor antigen pairs (e.g., EGFR/HLA). Screen hits will undergo validation and functional testing in primary T cells. This platform enables rational pairing of robust inhibitory domains with defined CAR architectures, paving the way for next-generation NOT-gated CAR T cells designed for enhanced safety and precision in solid-tumor immunotherapy.