In vivo screening of TCR-based chimeric antigen receptors for improved anti-tumor functionality

Abstract

Chimeric antigen receptor (CAR) T-cell therapy has demonstrated remarkable efficacy in hematological malignancies, but its success in solid tumors remains limited. Conventional CAR-T designs do not integrate into the T-cell receptor (TCR) and rely on non-natural CD3{zeta} signaling. They thus often suffer from tonic signaling, rapid T-cell exhaustion, and antigen escape due to reduced sensitivity. To address these limitations, we explored alternative CAR architectures that take into account the evolutionary optimized TCR signaling machinery. Specifically, we employed T-cell receptor fusion constructs (TRuCs), where a single-chain variable fragment (scFv) targeting the Sialyl-Thomson-Nouveau antigen (STn) is attached to the CD3{varepsilon} subunit. We then systematically screened a library of costimulatory domains and their combinations, fused to the TRuC, using a novel in vitro and in vivo screening approach. This screen identified a potent TRuC variant incorporating both CD28 and CD27 costimulatory domains. This optimized construct exhibited increased in vitro and in vivo proliferation and enhanced 4-1BB and IFN{gamma} expression upon repeated antigen stimulation. It further showed improved antitumor efficacy compared to conventional second-generation CAR-T cells in a mouse melanoma model. Moreover, we validated the adaptability of this approach by targeting B7H3 in a sarcoma model. The TRuC with costimulatory domains again outperformed other CAR-variants. Our findings highlight the potential of incorporating combined CD28 and CD27 costimulatory domains into TCR-based CAR architectures to overcome limitations associated with conventional CAR-T therapy and improve efficacy against solid tumors.