The target molecules of the T-cell response in type 1 diabetes, despite their pathogenic importance, remain largely uncharacterized, especially in humans. Interestingly, molecules such as insulin and glutamic acid decarboxylase (GAD) have been shown to be a target not only of autoantibodies, but also of autoreactive T-lymphocytes both in man and in the non-obese diabetic (NOD) mouse. In the present study we aimed to determine the existence of a specific T-cell response towards the insulinoma-associated protein 2 (IA-2) islet tyrosine phosphatase, a recently identified autoantigen which is the target of autoantibodies strongly associated with diabetes development. Human recombinant IA-2 produced in Escherichia coli, was tested for its reactivity with peripheral blood lymphocytes obtained from 16 newly diagnosed type 1 diabetic patients and from 25 normal controls, 15 of whom were HLA-DR-matched. A T-cell proliferation assay was performed in triplicate employing freshly isolated cells in the absence or in the presence of the antigen to be tested (at two different concentrations: 2 microg/ml and 10 microg/ml). A specific T-cell proliferation (defined as a stimulation index (S.I.) >/=3) was observed against IA-2 used at a concentration of 10 microg/ml (but not of 2 microg/ml) in 8/16 diabetic patients, in 1/15 HLA-DR-matched control subjects (P<0.01 by Fisher exact test) and in 0/10 of the remaining normal individuals. A statistically significant difference (P<0.003 by Mann-Whitney U test) was also observed in S.I. values between patients (3.1+/-1.4) and HLA-DR-matched controls (1.7+/-0.54) employing IA-2 at a concentration of 10 microg/ml. However, when IA-2 was used at a concentration of 2 microg/ml, the difference in S. I. between patients (1.65+/-0.8) and controls (1.0+/-0.3) did not reach statistical significance. In conclusion, these data show the presence of a specific, dose-dependent T-lymphocyte response against the IA-2 islet tyrosine phosphatase at the onset of type 1 diabetes. Consequently, this molecule appears to be a target not only at the B-lymphocyte but also at the T-lymphocyte level, reinforcing the potential pathogenic role of this autoantigen in the islet destructive process.