Search

MicroRNAs (miRNAs) are small non-coding RNAs that play a crucial role in modulating gene expression and are enriched in cell-derived extracellular vesicles (EVs). We investigated whether miRNAs from human islets and islet-derived EVs could provide insight into β cell stress pathways activated during type 1 diabetes (T1D) evolution, therefore serving as potential disease biomarkers. We treated human islets from 10 cadaveric donors with IL-1β and IFN-γ to model T1D ex vivo. MicroRNAs were isolated from islets and islet-derived EVs, and small RNA sequencing was performed. We found 20 and 14 differentially expressed (DE) miRNAs in cytokine- versus control-treated islets and EVs, respectively. Interestingly, the miRNAs found in EVs were mostly different from those found in islets. Only two miRNAs, miR-155-5p and miR-146a-5p, were upregulated in both islets and EVs, suggesting selective sorting of miRNAs into EVs. We used machine learning algorithms to rank DE EV-associated miRNAs, and developed custom label-free Localized Surface Plasmon Resonance-based biosensors to measure top ranked EVs in human plasma. Results from this analysis revealed that miR-155, miR-146, miR-30c, and miR-802 were upregulated and miR-124-3p was downregulated in plasma-derived EVs from children with recent-onset T1D. In addition, miR-146 and miR-30c were upregulated in plasma-derived EVs of autoantibody positive (AAb+) children compared to matched non-diabetic controls, while miR-124 was downregulated in both T1D and AAb+ groups. Furthermore, single-molecule fluorescence in situ hybridization confirmed increased expression of the most highly upregulated islet miRNA, miR-155, in pancreatic sections from organ donors with AAb+ and T1D.