Current or recent enteroviral infections show an association with type 1 diabetes. However, evidence for this has mainly been generated using a particular mouse monoclonal antibody (clone 5-D8/1) which binds the viral capsid protein VP1. Difficulty in confirming these findings using other independent methods has led to the concern that this might be artefactual. To address this, we examined the potential cross-reactivity of clone 5-D8/1 with normal islet proteins. Western blotting, two-dimensional gel electrophoresis, and mass spectrometry were used to identify human islet proteins bound by the clone 5-D8/1. We found a distinct reactivity with two mitochondrial proteins, creatine kinase B-type and ATP synthase beta subunit. Immunohistochemistry using the clone 5-D8/1 revealed a granular cytoplasmic staining pattern in mitochondria-rich cells, ie hepatocytes, ductal epithelial cells, vascular endothelial cells, skeletal muscle cells, and the neoplastic salivary gland oncocytoma cells, whereas connective tissue and infiltrating immune cells were negative. Staining on islets of Langerhans from subjects with recent-onset type 1 diabetes, but not on isolated human islets infected in vitro with enteroviruses, could be blocked after mixing the clone 5-D8/1 with the mitochondrial proteins. Collectively, our data show that the clone 5-D8/1 detects two human mitochondrial enzymes in addition to enteroviral VP1. The notion that the previously reported VP1 positivity in islets of recent-onset type 1 diabetes patients could reflect cross-reactivity to native islet proteins and not the presence of EV is supported by difficulties in demonstrating EV infection by independent techniques such as PCR or in situ hybridization. These findings call for revisiting the presence of enteroviruses in pancreatic islets of patients with type 1 diabetes.
Copyright © 2013 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.