T cell and autoantibody recognition of nucleus-associated islet autoantigens in individuals with type 1 diabetes.

Aims/hypothesis: There is a progressive loss of self-tolerance in type 1 diabetes, manifested by the appearance of various autoantibodies. Array-based screening identified antibodies that recognise nucleus-associated proteins in individuals with type 1 diabetes, but the role of these antigens in the disease is poorly understood. Antibodies against MutL homologue 1 (MLH1) and nucleoporin 50 (NUP50) are enriched in DR4-positive and DR3-positive individuals, respectively. Therefore, we sought to investigate CD4⁺ T cell recognition of these antigens and to assess whether cellular and humoral recognition of these autoantigens are linked.

Methods: We used a systematic discovery process to identify CD4⁺ T cell epitopes within MLH1 and NUP50. We synthesised peptides derived from these antigens and then measured their ability to bind to recombinant DRB1*04:01 or DRB1*03:01 protein, our two HLA class II types of interest. We assessed peptide immunogenicity by expanding peripheral blood T cells in vitro and visualising peptide-specific T cells using HLA class II tetramers. We then performed direct tetramer staining of samples from individuals with type 1 diabetes and HLA-matched control individuals to enumerate MLH1- or NUP50-reactive CD4⁺ T cells and characterise their cell surface phenotype. Responses were also characterised using islet-derived T cells from pancreatic organ donors with type 1 diabetes using cytokine release as a readout. Antibody responses against both antigens were measured in matched serum samples using a previously published ELISA assay.

Results: Our discovery process revealed three novel DRB1*03:01-restricted NUP50 epitopes and four novel DRB1*04:01-restricted MLH1 epitopes that are present within the peripheral blood of individuals with type 1 diabetes and among pancreatic islet infiltrates. T cells specific for these epitopes were significantly more frequent in individuals with diabetes than in HLA haplotype-matched control individuals (p=0.0012 and 0.030 for NUP50 and MLH1, respectively). Variable levels of antibody responses were observed: elevated levels of MLH1 and NUP50 antibodies were present in individuals with type 1 diabetes, especially those with the HLA-DR types with previously reported associations, but high titres did not always directly correlate with elevated T cell frequency.

Conclusions/interpretation: The observation that T cell and antibody responses can target nucleus-associated self-antigens confirms and extends previously published studies. Disease-associated recognition of a class of proteins that are not exclusively expressed in pancreatic islets implies a systemic autoimmune component to the disease process. Linked antibody recognition does not appear to be a general phenomenon, suggesting a subtle relationship between humoral and cellular responses to these self-antigens.