A Comparison of Co-stimulatory HVEM Domains in Second Generation CAR-T Cells

Abstract

Chimeric antigen receptor T cells (CAR-T) are T cells that have been genetically modified to express a receptor that recognizes a specific target molecule. When T cells bind to their target molecule, usually a surface antigen expressed on a tumor cell, they exhibit effector functions similar to T cells, including proliferation, cytokine secretion, and target cell killing. First-generation CAR-T cells contained only a synthetic receptor coupled to the CD3ζ signaling domain of the T cell receptor complex. Although first-generation CAR-T cells functioned well in vitro, they were generally unable to kill tumor cells effectively in mouse models. Second-generation CAR-T cells, which contain an intracellular co-stimulatory domain in tandem with the CD3ζ signaling domain, significantly improved the function of CAR-T cells and are now the basis for several approved therapies targeting lymphomas and leukemias. The choice of a costimulatory molecule (co-stim domain) is critical in ensuring CAR-T persistence and clinical response. As relapse rates in CAR-T recipients are still considerable, identifying new potential co-stim domains may be essential to improve clinical performance. The costimulatory domain of Herpes Virus Entry Mediator (HVEM) is a promising costimulatory domain in CAR-T cell models and might strike a “Goldilocks balance” between performance characteristics of the only co-stim domains in approved CAR-T therapies: CD-28 and 4-IBB. Mutated forms of the HVEM costimulatory domain were created to identify which components are critical to its function by generating HVEM variants with mutations in three key regions of the HVEM cytoplasmic domain (referred to as M83): deletion of a membrane-proximal α -helix, deletion of a small tail domain, and a point mutation in a potentially critical TRAF signaling domain.

Upon coculture of anti-CD19 CAR-T cells containing modified M83 domains with target cell lines, altered CAR-T activation responses emerged as a function of specific co-stim domain modifications. The loss of the membrane-proximal α-helix severely ablated CAR activation. Mutating the M83 TRAF binding domain reduced the expression of the activation marker relative to wild-type M83 and CAR-T cells without a co-stim domain. Deleting the C-terminal tail domain produced a minimum impact on activation. However, interesting trends in the general responses of CAR-T culture to target cells can potentially indicate a change in the dynamics of TRAF signaling, suggesting increased non-canonical NF-κB (ncNFκB) signaling, which could encourage longer persistence of CAR-T cells. Although more work is needed to confirm these observations, this work strongly suggests that using a truncated M83 as the CAR-T co-stim domain may improve the persistence and efficacy of CAR-T cells.