Abstract
In the present study, we analyzed microRNA (miRNA) and gene expression profiles using 499 papillary thyroid carcinoma (PTC) samples and 58 normal thyroid tissues obtained from The Cancer Genome Atlas database. A pivotal regulatory network of 18 miRNA and 16 targets was identified. Upregulated miRNAs (miR-222, miR-221, miR-146b, miR-181a/b/d, miR-34a, and miR-424) and downregulated miRNAs (miR-9-1, miR-138, miR-363, miR-20b, miR-195, and miR-152) were identified. Among them, the upregulation of miR-424 and downregulation of miR-363, miR-195, and miR-152 were not previously identified. The genes CCNE2 (also known as cyclin E2), E2F1, RARA, CCND1 (cyclin D1), RUNX1, ITGA2, MET, CDKN1A (p21), and COL4A1 were overexpressed, and AXIN2, TRAF6, BCL2, RARB, HSP90B1, FGF7, and PDGFRA were downregulated. Among them, CCNE2, COL4A1, TRAF6, and HSP90B1 were newly identified. Based on receiver operating characteristic curves, several miRNAs (miR-222, miR-221, and miR-34a) and genes (CCND1 and MET) were ideal diagnostic indicators, with sensitivities and specificities greater than 90%. The combination of inversely expressed miRNAs and targets improved diagnostic accuracy. In a clinical feature analysis, several miRNAs (miR-34a, miR-424, miR-20b, and miR-152) and genes (CCNE2, COL4A1, TRAF6, and HSP90B1) were associated with aggressive clinical features, which have not previously been reported. Our study not only identified a pivotal miRNA regulatory network associated with PTC but also provided evidence that miRNAs and target genes can be used as biomarkers in PTC diagnosis and clinical risk evaluation.