Autoantibody responses to N-terminally truncated GAD65 (96-585) in children with newly diagnosed type 1 diabetes and in GADA-positive children from the general population

Background 

GAD antibodies (GADA) are used to identify patients with type 1 diabetes (T1D) and subjects at increased disease risk. They are, however, also detected in many healthy subjects without increased T1D risk. Radiobinding assays (RBAs) using N-terminally truncated GAD65 (96–585) instead of full-length GAD65 (1-585) identify more specifically at-risk relatives of T1D patients. We assessed whether the specificity of the RBA can be improved without losing sensitivity when using truncated ³⁵GAD₆₅ (tGAD) as well as identified GAD65 epitopes in single GADA positive, multipositive subjects and T1D patients. 

Methods 

The samples were derived from 101 children with newly diagnosed T1D and from 83 children recruited from the general population who tested positive for GADA at some time point during follow-up. All samples were analyzed for tGADA and GAD65 N-, M-and C-epitope immune responses using RBAs.

Results 

In patients with recent-onset T1D and in multiautoantibody positive non-diabetic subjects the immune response was directed primarily towards the M- and C-epitopes. Those subjects who tested positive for GADA but negative for tGADA were single GADA positive, N-epitope positive, and the immune response did not spread to other epitopes or other antigens, and none of these subjects has developed T1D during follow up. Out of 202 samples in 83 non-diabetic subjects, 120 (59%) were positive for GADA (1?585) and 82 (41%) for GADA (96?585) (P<0.001). Truncated GADA (96?585) positivity was more common in those children carrying HLA genotypes conferring increased risk vs. the remaining children (60% vs. 29.5%, P<0.001).  

Conclusions

Our results indicate that tGADA is more specific for T1D than full-length GADA. Isolated positivity for the N-epitope of GADA does not increase the risk of T1D.