RESUMO
PURPOSE: To estimate the roles of extracellular vesicles (EVs) in tears and to determine whether their profiles are associated with the type of ocular disease. STUDY DESIGN: Cross-sectional study. METHODS: Tear EVs were extracted from 14 healthy participants and from 21 patients with retinal diseases (age-related macular degeneration [AMD] or diabetic macular edema [DME]). The surface marker expression of tear EVs was examined, and microRNAs (miRNAs) were extracted and profiled by use of real-time PCR array. The stability of the expression of the miRNAs was determined, and their functions were assessed by network analyses. Classification accuracy was evaluated by use of a random forest classifier and k-fold cross-validation. RESULTS: The miRNAs that were highly expressed in tear EVs were miR-323-3p, miR-548a-3p, and miR-516a-5p. The most stably expressed miRNAs independent of diseases were miR-520h and miR-146b-3p. The primary networks of the highly stably expressed endogenous miRNAs were annotated as regulation of organismal injury and abnormalities. The highly expressed miRNAs for severe retinal disease were miR-151-5p for AMD and miR-422a for DME, suggesting potential roles of tear EVs in liquid biopsy. Nine miRNAs (miR-25, miR-30d, miR-125b, miR-132, miR-150, miR-184, miR-342-3p, miR-378, and miR-518b) were identified as distinguishing individuals with AMD from healthy individuals with a classification accuracy of 91.9%. CONCLUSIONS: The finding that tear EVs contain characteristic miRNA species indicates that they may help in maintaining homeostasis and serve as a potential tool for disease diagnosis.