The HLA Ligandome Landscape of Chronic Myeloid Leukemia Delineates Novel T-Cell Epitopes for Immunotherapy

Chronic myeloid leukemia (CML) is characterized by the translocation t(9;22), which leads to the formation of the BCR-ABL fusion transcript. Several approved tyrosine kinase inhibitors (TKIs) target the resulting fusion protein, leading to an improved prognosis of CML patients. Currently, the main treatment goal is the achievement of a deep molecular response (MR), in which TKI therapy can be terminated. Several studies provide evidence that immunological control plays a major role for the course of CML and contributes to the achievement of deep MR in CML patients under TKI treatment (CML_TKI). This implies that reinforcing these immune responses might sustain long-term TKI-free survival or even cure for CML patients. Besides unspecific immunotherapy, such as interferon or immune checkpoint blocking antibodies, a more specific and minor side effect targeting of CML cells might be achieved by antigen-specific immunotherapy approaches. The prerequisite for such strategies is the identification of T-cell targets represented by tumor-associated human leukocyte antigen (HLA)-presented peptides on malignant cells.

In this study, we used a mass spectrometry-based approach to identify naturally presented, CML-associated peptides in primary CML samples (HLA class I, n=21, 11,945 peptides, 5,478 source proteins; class II, n=20, 5,991 peptides, 1,302 proteins). Comparative HLA peptidome profiling using a comprehensive dataset of various benign tissues (e.g. blood, bone marrow, spleen, and lung) revealed frequently presented and strictly CML-associated antigens. In detail, the benign tissue dataset comprises hematological benign samples (class I, n=108, 51,233 peptides, 11,437 proteins; class II, n=88, 42,753 peptides, 4,877 proteins) and non-hematological benign tissues (28 tissues, n=166; class I, 128,590 peptides, 16,405 proteins; class II, 143,652 peptides, 13,410 proteins). We identified 50 CML-associated, HLA class I-restricted peptides with HLA allotype adjusted representation frequencies of ≥38% presented on HLA-A*02, -A*03, -A*11, and -B*07. HLA class II comparative profiling delineated 36 peptides exclusively and frequently presented in the HLA peptidome of ≥20% analyzed CML patients.

For immunological characterization, we selected 8 HLA class I- and 6 class II-restricted highly CML-associated antigens. These peptides were analyzed in IFNγ ELISPOT assays using PBMCs from CML_TKI patients and healthy volunteers (HVs). Peptide-specific immune recognition was detected for 1/8 (13%) HLA class I peptides in 2/17 (12%) of CML_TKI patients. We hypothesized that this weak immune response might be due to an impaired CD8⁺ T cell function that reportedly is caused by TKI treatment. Thus, we compared T-cell responses against viral epitopes in IFNγ ELISPOT assays of CML_TKI patients with that of HVs and chronic lymphocytic leukemia (CLL) patients: in line with our hypothesis, we observed significantly reduced IFNγ release of T cells from CML_TKI patients compared to HVs and CLL patients (p<0.001), whereas CD8⁺ T-cell counts were not reduced. In contrast, no reduced IFNγ production was observed for HLA class II-restricted viral epitopes. These results were confirmed by memory T-cell responses detected for 6/6 (100%) HLA class II CML-associated peptides with frequencies up to 24% (4/17) of analyzed CML_TKI patients.

To assess the immunogenicity of all HLA class I peptides, we performed in vitro artificial antigen presenting cell-based priming experiments using CD8⁺ T cells of HVs and CML patients. Effective priming of T cells was observed for 8/8 CML-associated peptides in ≥70% of analyzed HVs with frequencies of 0.1-33.9% (mean 2.2%) of CD8⁺ peptide-specific T cells. Notably, peptide-specific CD8⁺ T cells with frequencies of 0.1-2.2% (mean 0.4%) could also be induced in samples of CML_TKI patients that had not displayed preexisting immune responses. For 6/8 peptides, we observed multifunctionality of peptide-specific T cells by IFNγ and TNF production as well as upregulation of the degranulation marker CD107a. Cytotoxicity assays with polyclonal peptide-specific effector T cells confirmed the capacity to induce antigen-specific lysis for 3/4 analyzed peptides.

Taken together, we here identified novel, naturally presented, CML-associated antigens and validated them as promising targets for tailored T cell-based immunotherapeutic approaches for CML patients.