Diagnostic sensitivity of immunodominant epitopes of glutamic acid decarboxylase (GAD65) autoantibodies in childhood IDDM.

The prevalence and titre of epitope-specific autoantibodies to glutamic acid decarboxylase (GAD65) in 155 insulin-dependent diabetic (IDDM) and 9 GAD65 antibody (Ab)-positive healthy children were determined using four GAD65/67 chimaeric molecules which discriminate among the N-terminal (N), middle (M) and C-terminal (C) epitopes of GAD65. Radioligand binding assays for IgGAb used immunoprecipitation of in vitro translated 35S-GAD. We found autoantibodies to GAD65 in 116 of 155 (75%), to GAD67 in 19 of 155 (12%) (p < 0.0001) and to the GAD65-N-67 chimaera in 25 of 155 (16%) (p < 0.0001) IDDM sera. GAD67Ab were found almost exclusively (17 of 19, 89%) in GAD65Ab-positive sera and the levels of GAD67Ab correlated with those of GAD65Ab (r2 = 0.5913; p = 0.009). GAD65Ab directed to GAD65-M were found in 104 of 155 (67%), to GAD65-C in 104 of 155 (67%) and to GAD65-M + C in 116 of 155 (75%) of IDDM sera, and indicated reactivity to at least two distinct epitopes. Among the nine GAD65Ab-positive healthy children, two (22%) were also positive with GAD67, nine (100%) with GAD65-M + C, seven (78%) with GAD65-M, eight (89%) with GAD65-C and two (22%) with GAD65-N-67. Titres of GAD65Ab (p = 0.007), GAD65-C-Ab (p = 0.002) and GAD65-C + M-Ab (p = 0.003), but not of GAD65-M-Ab (p = 0.101) were significantly higher in IDDM than in healthy children. We conclude that GAD65Ab in IDDM and healthy children are directed to middle and C-terminal epitopes, and propose that levels of antibodies specifically directed to the carboxy-terminal end of GAD65 may distinguish IDDM from healthy children.