Identification of down-regulated microRNAs in thyroid cancer and their potential functions.

Background: The mechanism of microRNAs (miRNAs) in thyroid cancer is still unclear. We identified miRNAs with differential expression in thyroid cancer versus normal tissues. Methods: Microarray datasets were obtained from the GEO and ArrayExpress databases, and from publications found via PubMed, EMBASE, and Web of Science. Differentially expressed miRNAs were identified using the limma package, and their targets predicted using miRWalk. Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and protein-protein interaction (PPI) network analyses were performed using these target genes to explore potential carcinogenic mechanisms. Correlations between target gene and miRNA expression levels were examined. Changes in target protein expression were confirmed using data from The Human Protein Atlas and the Cancer Genome Atlas. Results: We ultimately included five datasets, and further analyzed the four miRNAs that were down-regulated in at least four datasets (miR-7-2-3p, miR-138-5p, miR-144-5p, miR-486-5p). Predicted targets were enriched in GO terms including extracellular matrix organization, cell surface, and receptor binding, and in KEGG cancer pathways. PPI analysis identified 10 hub genes as key potential targets of these miRNAs. The expression levels of eight target genes were negatively correlated with those of their respective miRNAs. Furthermore, eight predicted target genes in cancer-related pathways showed up-regulated protein and mRNA expression in thyroid cancer. Conclusion: Low miRNA expression in thyroid cancer might influence tumorigenesis via critical pathways. The genes identified here may act as a starting point for further investigation of the carcinogenic mechanisms of these miRNAs.