Phenotypic variability within isogenic population of lymphocytes

Division Of Basic Sciences - Nci

Investigators

Linked publications, trials & patents

Abstract

We expanded the project horizon by collaborating with Naomi Taylor (CCR) and Paul François (UMontréal) in order to usher new Chimetic Antigen Receptor (CAR) T cells. CAR T cell therapy has been transformative for blood cancers but remains limited in treating solid tumors due to on-target/off-tumor (OTOT) toxicity. To address this, we engineered T cells to co-express both a chimeric antigen receptor (CAR) and a neoantigen-specific T cell receptor (TCR), enabling "fuzzy logic" control of T cell activation. Unlike conventional Boolean logic systems, our approach allowed for graded responses shaped by the strength of TCR engagement. We discovered that TCR signaling modulated CAR activity in a non-binary, spectrum-like manner: weak TCR signals antagonized CAR activation, while strong signals enhanced it. Using high-throughput robotic assays and a revised adaptive kinetic proofreading (AKPR) model, we mapped the quantitative rules governing this crosstalk. This led us to define and implement an "antagonism-enforced braking system" (AEBS), whereby self-antigen recognition through the TCR dampened CAR function, thus reducing toxicity against healthy tissues. We validated this AEBS system in multiple mouse models of leukemia and solid tumors. Dual TCR/CAR T cells selectively eliminated tumor cells expressing strong neoantigens while sparing healthy cells expressing self-antigens. In humanized models, AEBS T cells preserved anti-tumor efficacy while avoiding damage to normal human lung tissue expressing the same CAR antigen. We further demonstrated that this antagonistic modulation was widespread: many altered peptide ligands (pAPLs) derived from tumor neoantigens acted as CAR antagonists via the TCR. Finally, we engineered AEBS T cells combining a HER2 CAR with TCRs specific for p53 and HHAT neoantigens, showing enhanced tumor specificity and reduced OTOT toxicity across diverse human cancer cell lines. Our work establishes a new therapeutic principle-TCR-controlled fuzzy logic-to enhance the precision of CAR T cell therapies (we filed a provisional patent and hope to start clinical trials soon). This approach opens a pathway for more effective and safer immunotherapies by encoding self-tolerance directly into engineered T cells.