Exploiting HLA-e restricted T cell receptors to develop universal cellular therapies for virus-associated malignancies

Abstract

Abstract Adoptive T cell receptor-engineered therapies (TCR-Ts) enable targeted redirection of T cells to tumor antigens presented by human leukocyte antigen (HLA) molecules on cancer cells. While these therapies have demonstrated clinical promise, their broader application is constrained by the extreme polymorphism of classical HLA genes. This necessitates HLA allele-specific TCR-T designs and limits population-scale accessibility. To address this, we investigated HLA-E, a non-classical, functionally monomorphic HLA molecule that presents self- and pathogen-derived antigens. This characteristic positions HLA-E as a potential platform for universal TCR-T development for virus-associated cancers. HLA-E has been shown to present antigenic peptides and elicit T cell responses in cells infected with human immunodeficiency virus, Epstein-Barr virus (EBV), influenza, and cytomegalovirus (CMV) (Vietzen et al. Blood. 2023, Yang et al. Sci Immuno. 2021). Virus-associated cancers are often regionally endemic affecting genetically diverse populations and highlighting the need for broadly applicable therapies. One example, human T lymphotropic virus 1 (HTLV-1)-associated T cell leukemia/lymphoma (ATLL), is an aggressive malignancy distributed in South/Central America, Africa, and Japan with no standard curative treatment. As a proof-of-concept for HLA-E–restricted TCR-T therapy, we isolated a TCR specific for the VL9 peptide (VMAPRTLLL), derived from the CMV UL40 gene and presented by HLA-E. Short-term T cell lines were generated from peripheral blood mononuclear cells (PBMCs) of CMV-seropositive donors using artificial antigen-presenting cells (aAPCs) expressing HLA-E-VL9. VL9-specific T cells were detected in 5 of 12 donors. We used tetramer sorting and single-cell TCR sequencing to identify the dominant VL9-reactive TCR from the donor with the greatest T cell expansion. To assess the function of this TCR, we generated a T cell activation reporter line using CRISPR to delete endogenous HLA-E in the TCRnull NFAT-luciferase Jurkat cell line (Grailer et al. J of Immunotherapy. 2023) followed by transduction of the VL9-TCR. This system enabled high-throughput, luminescence-based screening against a panel of aAPCs presenting various HLA-E–peptide complexes. Unexpectedly, the VL9-TCR exhibited cross-reactivity, recognizing multiple unrelated HLA-E–peptide ligands. Structural modeling with AlphaFold3 supported TCR-peptide-HLA-E interaction interfaces across these complexes. We next screened a panel of HLA-E–positive cancer lines and healthy tissues for susceptibility to this TCR compared to HLA-Enull K562 cells. The reporter line showed robust activation (>20-fold) in the HTLV-1+ ATLL cell line, 84C and the EBV+ Burkitt's lymphoma line, Raji. Moderate activation (>5-fold) was also observed in human papilloma virus positive (HPV+) cervical cancer lines SiHa and HeLa. Healthy primary PBMCs and lung endothelial cells also moderately activated the reporter line, indicating potential off-tumor reactivity. We next examined if primary CD8 T cells redirected against HLA-E-VL9 (VL9-TCR-T) had cytotoxic potential. CD8 T cells from healthy donors were activated with α-CD3 and α-CD28 antibodies and stimulated with cytokines. The cells then underwent endogenous TCR and HLA-E deletion using CRISPR methods and transduced with the VL9-TCR. VL9-TCR-Ts were co-cultured with 84C and healthy donor PBMCs at a 1:1 effector-to-target ratio for 6 hours. VL9-TCR-Ts exhibited potent and preferential killing of 84C cells compared to PBMCs (49.6% vs. 70.1% viable cells). Expression of T cell activation markers INF-γ and CD107a by intracellular cytokine staining and flow cytometry and use of HLA-E KO control targets confirmed CD8 T cell mediated cytotoxicity and HLA-E dependence. These findings demonstrate the feasibility of HLA-E–restricted TCR-Ts for targeting virus-associated cancers while underscoring the challenges of off-target cross-reactivity. While alone, the VL9-TCR-T has limited clinical applicability, we are incorporating logic-gated control systems, such as synthetic Notch receptors that recognize tumor-associated markers like CD30 (expressed on the 84C line), to further restrict TCR expression against malignant cells and enhance therapeutic precision (Roybal et al. Cell. 2016). We are also using mass spectrometry to identify novel tumor-specific viral peptides, which may enable the development of more selective, HLA-E–based therapies.