Coxsackievirus infections have been proposed as an environmental trigger for the development of T-cell-mediated autoimmune (type 1) diabetes by either providing a molecular mimic of the candidate pancreatic ␤-cell autoantigen GAD or inducing bystander inflammation in the pancreas. In this study in the NOD mouse model, we found that infection with a pancreatrophic coxsackievirus isolate can accelerate type 1 diabetes development through the induction of a bystander activation effect, but only after a critical threshold level of insulitic ␤-cell-autoreactive T-cells has accumulated. Thus, coxsackievirus infections do not appear to initiate ␤-cell autoreactive immunity but can accelerate the process once it is underway. These findings indicate that the timing of a coxsackievirus infection, rather than its simple presence or absence, may have important etiological implications for the development of T-cell-mediated autoimmune type 1 diabetes in humans. Diabetes 49:708-711, 2000 T -cell-mediated autoimmune (type 1) diabetes in both humans and NOD mice is controlled by multiple susceptibility genes whose pathogenic functions can be modulated by various environmental factors (1-3). Coxsackieviral infections may represent one environmental factor that could contribute to type 1 diabetes development in genetically susceptible individuals (4). An antigenic epitope derived from the 65-and 67-kDa isoforms of GAD and proposed to be one of the earliest targets of diabetogenic CD4 + T-cell responses in NOD mice is characterized by a PEVKEK sequence also found within a peptide consisting of amino acids 28-50 from the coxsackievirus P2C protein (Cox sequence similarity peptide [ssp]) (5). Thus, one mechanism by which a coxsackievirus infection could contribute to type 1 diabetes is by providing a molecular mimic during replication that triggers a cross-reactive CD4 + T-cell response against the candidate ␤-cell autoantigen GAD. Arguing against this molecular mimicry hypothesis was a report that GAD reactive T-cell responses were not enhanced in NOD mice after a coxsackievirus infection (6). Instead, this earlier study found that the coxsackievirus B4 Edwards strain (CVB4) can accelerate type 1 diabetes development in a T-cell receptor (TCR) transgenic stock of NOD mice in which virtually all T-cells are of the CD4 + BDC2.5 clonotype that recognizes a ␤-cell autoantigen other than GAD. This finding was interpreted to mean that rather than providing a molecular mimic of GAD, infection with the pancreatrophic CVB4 isolate contributes to type 1 diabetes development by stimulating a local inflammatory response that leads to subclinical levels of ␤-cell destruction and the subsequent release of normally sequestered antigens, which then trigger pathogenic autoreactive T-cell responses. However, this interpretation does not explain why, in that previous study, CVB4 infection failed to elicit type 1 diabetes development in standard nontransgenic NOD mice that are characterized by not only the presence of BDC2.5 clonotypic T-cells, but al...

Type 1 diabetes acceleration in nonobese diabetic (NOD) mice through coxsackievirus B4 (CVB4) infection requires a preexisting critical mass of autoreactive T cells in pancreatic islets

Investigations of humans and nonobese diabetic mice suggest that proinsulin and/or a fragment of the region spanning C-peptide and the B-chain of insulin (i.e., proinsulin peptide) may serve as key autoantigens in IDDM. Therefore, we analyzed cellular immune reactivities against these molecules in people with or at varying risks for the disease to clarify their role in the pathogenesis of IDDM. In vitro peripheral blood mononuclear cell (PBMC) responses against these antigens, a control antigen (tetanus toxoid), and phytohemaglutinin were determined in 60 individuals with newly diagnosed IDDM (< or = 1 day from diagnosis) in 34 islet cell cytoplasmic autoantibody- and/or insulin autoantibody-negative first-degree relatives of the IDDM subjects, and in 28 autoantibody-negative control subjects. Unlike previous reports suggesting diabetes-associated elevations in cellular immunity to other beta-cell antigens (e.g., GAD, IA-2, etc.), we observed equivalent levels of phytohemaglutinin stimulation and cellular proliferation in all groups against these antigens (all P values were not significant). The mean stimulation index +/- SD and frequency of reactivity to proinsulin for healthy control subjects and IDDM patients, respectively, were as follows: 1 microg/ml (1.5 +/- 1.0, 1 out of 17 [6%]; 1.9 +/- 1.4, 4 out of 33 [12%]); 10 microg/ml (1.7 +/- 1.3, 1 out of 17 [6%]; 1.2 +/- 0.6, 0 out of 28 [0%]); and 50 microg/ml (1.2 +/- 0.6, 1 out of 16 [6%]; 1.1 +/- 0.6, 1 out of 27 [4%]). The response in healthy control subjects, autoantibody-negative relatives, and IDDM patients, respectively, against the proinsulin peptide fragment were as follows: 1 microg/ml (0.9 +/- 0.4, 1 out of 12 [8%]; 1.3 +/- 1.1, 4 out of 34 [11%]; 1.1 +/- 0.3, 2 out of 28 [7%]); 10 microg/ml (0.9 +/- 0.6, 1 out of 12 [8%]; 1.2 +/- 0.6, 3 out of 34 [9%] 1.4 +/- 1.7, 2 out of 28 [7%]); and 50 microg/ml (1.0 +/- 0.7, 1 out of 12 [8%]; 1.2 +/- 0.5, 2 out of 34 [6%]; 1.3 +/- 0.5, 2 out of 28 [7%]). Taken together with previous studies reporting relatively infrequent occurrences of autoantibodies to proinsulin, the role of immunity to this molecule in the pathogenesis of IDDM in humans remains unclear.

Autoantibodies to the neuroendocrine protein insulinoma-associated protein 2 (IA-2), a member of the tyrosine phosphatase family, have been observed in individuals with or at increased risk for IDDM. Because this disease is thought to result from a T-cell-mediated autoimmune destruction of the insulin-producing pancreatic beta-cells, we analyzed humoral and cellular immune reactivity to this autoantigen to further define its role in the pathogenesis of IDDM. Peripheral blood mononuclear cells (PBMC) from individuals with newly diagnosed IDDM or at varying levels of risk for the disease were stimulated in vitro with the entire 42-kDa internal domain of IA-2 (amino acids 603-979), a series of control antigens (glutathionine-S-transferase, tetanus toxoid, Candida albicans, mumps, bovine serum albumin), and a mitogen (phytohemagglutinin). The frequency and mean stimulation index of PBMC proliferation against IA-2 was significantly higher in newly diagnosed IDDM subjects (14 of 33 [42%]; 3.8+/-4.5 at 10 microg/ml) and autoantibody-positive relatives at increased risk for IDDM (6 of 9 [66%]; 3.9+/-3.2) compared with autoantibody-negative relatives (1 of 15 [7%]; 1.8+/-1.0) or healthy control subjects (1 of 12 [8%]; 1.5+/-1.0). The frequencies of cellular immune reactivities to all other antigens were remarkably similar between each subject group. Sera from 58% of the newly diagnosed IDDM patients tested were IA-2 autoantibody positive. Despite investigations suggesting an inverse association between humoral and cellular immune reactivities against islet-cell-associated autoantigens, no such relationship was observed (rs=0.18, P=0.39) with respect to IA-2. These studies support the autoantigenic nature of IA-2 in IDDM and suggest the inclusion of cellular immune responses as an adjunct marker for the disease.