Humoral autoimmunity against the extracellular domain of the neuroendocrine autoantigen IA-2 heightens the risk of type 1 diabetes.

The objective of this study was to determine whether antigenic determinants localized within the extracellular domain of the neuroendocrine autoantigen tyrosine phosphatase-like protein IA-2 are targets of humoral responses in type 1 diabetes (T1DM). Previous studies indicated that the immunodominant region of IA-2 is localized within its intracellular domain (IA-2ic; amino acids 601-979). We analyzed 333 subjects from the Children's Hospital of Pittsburgh study, 102 of whom progressed to insulin-requiring diabetes (prediabetics). Autoantibodies from these individuals were initially assayed for ICA512bdc (Barbara Davis Center amino acids 257-556; 630-979), IA-2ic (amino acids 601-979), and IA-2 full-length (amino acids 1-979) in addition to islet cell antibody (ICA), glutamic acid decarboxylase, 65-kDa isoform, and insulin autoantibodies. We identified an autoantibody response reactive with the extracellular domain of IA-2 that is associated with very high risk of T1DM progression. Relatives with no detectable autoantibodies against ICA512bdc (or IA-2ic) exhibited antibody responses against the IA-2 full-length peptide (log rank, P = 0.008). This effect was also observed in first-degree relatives who were positive for glutamic acid decarboxylase, 65-kDa isoform (log rank, P = 0.026) or at least two islet autoantibodies but were negative for ICA512bdc (log rank, P = 0.022). Competitive binding experiments and immunoprecipitation of the IA-2 extracellular domain (amino acid residues 26-577) further lend support for the presence of autoantibodies reactive with new antigenic determinants within the extracellular domain of IA-2. In summary, the addition of measurements of autoantibodies reactive with the IA-2 extracellular domain to assays geared to assess the progression of autoimmunity to clinical T1DM may more accurately characterize this risk. This has considerable implications not only for stratifying high diabetes risk but also facilitating the search for pathogenic epitopes to enable the design of peptide-based immunotherapies that may prevent the progression to overt T1DM at its preclinical stages.