Prediction of HLA class I-restricted T-cell epitopes of islet autoantigen combined with binding and dissociation assays

Abstract

Identification of cognate peptides recognized by human leucocyte antigen (HLA)/T cell receptor (TCR) complex provides insight into the pathogenic process of type 1 diabetes (T1D). We hypothesize that HLA-binding assays alone are inadequate metrics for the affinity of peptides. Zinc transporter-8 (ZnT8) has emerged in recent years as a novel, major, human autoantigen. Therefore, we aim to identify the HLA-A2-restricted ZnT8 epitopes using both binding and dissociation assays. HLA class I peptide affinity algorithms were used to predict candidate ZnT8 peptides that bind to HLA-A2. We analyzed 15 reported epitopes of seven β-cell candidate autoantigens and eight predicted candidate ZnT8 peptides using binding and dissociation assays. Using IFN-γ ELISpot assay, we tested peripheral blood mononuclear cells (PBMCs) from recent-onset T1D patients and healthy controls for reactivity to seven reported epitopes and eight candidate ZnT8 peptides directly ex vivo. We found five of seven recently reported epitopes in Chinese T1D patients. Of the eight predicted ZnT8 peptides, ZnT8_153–161 had a strong binding affinity and the lowest dissociation rate to HLA-A*0201. We identified it as a novel HLA-A*0201-restricted T-cell epitope in three of eight T1D patients. We conclude that ZnT8_153–161 is a novel HLA-A*0201-restricted T-cell epitope. We did not observe a significant correlation (P = 0.3, R = − 0.5) between cytotoxic T cell (CTL) response and peptide/HLA*0201 complex stability. However, selection of peptides based on affinity and their dissociation rate may be helpful for the identification of candidate CTL epitopes. Thus, we can minimize the number of experiments for the identification of T-cell epitopes from interesting antigens.

Keywords::

• T-cell epitope
• type 1 diabetes
• ZnT8
• autoantigen
• autoimmunity

Acknowledgments

We thank Prof. John C. Hutton (Barbara Davis Center for Childhood Diabetes, University of Colorado at Denver and Health Sciences Center, CO, USA) for providing autoantibodies to GAD, IA-2, and ICA. We are grateful to Prof. John C. Hutton and Liping Yu for their comments and assistance.

Declaration of interest: This work was supported by grants from the National Natural Science Foundation of China (grant numbers 30971405 and 30671010). The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.