Loss of TET2 Creates a Distinct Immunopeptidome in Acute Myeloid Leukemia

Background:

Antibodies directed against leukemia-specific antigens have been an early form of immunotherapy. Chimeric antigen receptor T-cell therapy (CART) has shown tremendous success in lymphoid neoplasia, but both CART cells and antibodies are limited due to their restriction to cell surface-expressed proteins. In contrast, T cell responses recognizing leukemia cells via tumor-specific peptides have a broader spectrum of targets and are believed to mount the strongest tumor surveillance and anti-tumor response.

One major roadblock in developing effective immune therapies is the inability to identify tumor cell-specific neoantigens that can distinguish normal cells from cancer cells. Neoantigens can induce anti-tumor effects through inducible T-cell responses, thereby activating cellular immunotherapy with the potential to eradicate Acute Myeloid Leukemia (AML). Selected studies have shown that neoantigens encoded by recurrent genetic aberrations in AML can be targeted using immunotherapy.

Loss of function TET2 mutations (TET2^MT) have been frequently identified in myeloid neoplasia (MN), contributing to disease pathogenesis and clonal expansion as indicated by high prevalence of TET2^MT in clonal hematopoiesis of indeterminate potential (CHIP). Targeting TET2^MT in AML may not only effectively eradicate leukemia, but also pre-malignant hematopoietic stem cells, thereby disrupt clonal proliferation at its source. Here we report the identification of neoantigens associated with TET2^MT that will help to develop novel therapeutic strategy.

Methods:

We utilized CRSPR-Cas9 edited clonal isogenic TET2^WT and TET2^KO cells derived from THP1 (Guan et al., Blood Cancer Discov, 2021). Three independent biological replicates of TET2^ko and TET2^WT were used for immunoaffinity purification of cell surface HLA bound peptides. The solubilized HLA complexes were dissociated in acid and filtered through 5kDa NMWCO ultrafiltration to remove non-peptide small molecules. Resulting peptide fractions were analyzed on an Orbitrap mass spectrometer. The raw data files were searched with Protein discovery with 5% FDR and peptide with 7-15 amino acids in length. HLA-A binding peptides were identified by mapping peptide FASTA sequences and HLA class alleles in the Net MHC pan 4.1 database.

Results:

Through optimized immunopeptidome workflow, we found that the amount of antibody used in immunoprecipitation was critical for maximum enrichment of immunopeptides. Over 5000 HLA-A binding peptides were enriched for quality analysis. From all identified peptides, about 30% were filtered out as non-immunopeptides due to their lengths not being within 7-15 amino acids and derivation from most abundant cellular proteins. The majority of identified peptides were 9mers.

We used unsupervised alignment and clustering of peptide sequences based on HLA-A binding motifs. About 98% of the eluted peptides clustered into five motifs with high affinities for HLA-A binding, typically featuring leucine (L) at positions 2 and 9. We also calculated the immunogenicity of eluted peptides in silico using the Immune Epitope Database (IEDB) based on T-cell preferences for amino acids. 21.8% of eluted peptides had an immunogenicity score above 0.2, suggesting they are likely T-cell epitopes.

Protein pathway analysis revealed the top contributing functional groups, including antigen processing, separation of sister chromatids, cell cycle checkpoints, and DNA repair.

TET2 deficiency created a distinct immunopeptidome, with about threefold more eluted peptides identified in isogenic TET^KO compared to TET2^WT, suggesting higher possible T-cell immunogenicity. A total of 389 peptides were either exclusively identified in TET^KO cells or exhibited a 2-fold higher mass intensity compared to TET2^WT. From these enriched in TET^KO cells, 10 peptides were selected based on their higher immunogenicity and signal intensity for further evaluation as potential targets for immunotherapeutic approaches.

Conclusions:

Our study revealed that TET2 loss leads to a unique peptide repertoire, suggesting enhanced T-cell immunogenicity. Our findings offer promising avenues for the development of novel diagnostic tools and therapeutic strategies targeting TET2-associated neoantigens in AML, with the potential to improve patient outcomes by reducing the risk of relapse and eradicating both malignant and pre-malignant cells.