Introduction: MicroRNAs (miRNAs) are small, non-coding RNAsthat can in?uence numerous genes expression and cellular functions.Aberrant miRNAs expressions are often happened in cancer. Herein,we temp to identify miRNAs altered in oral squamous cell carcino-mas (OSCC) and its contribution to oral carcinogenesis.Methods: We used a custom microarray platform to screen miR-NA expression pro?le on 40 OSCC and normal adjacent tissue sam-ples. The differentially expressed miRNAs were identi?ed with concordant fold-change by microarray analysis and subsequently validated by TaqMan RT-qPCR using a total of 46 OSCC tissue pairs. Results: Fifty-six miRNAs were differentially expressed in their expression between tumor and non-tumor parts. Twenty-seven highly differentially expressed miRNAs were chosen and validated using 20 (training samples) and another 46 (testing samples) tissue pairs. A comparison of selected miRNAs in RT-qPCR and microarray shows a very good correlation (R = 0.9327) between the two plat- forms. Furthermore, using Bayesian Binary Regression analysis, we identi?ed a miRNAs expression signature showed high discrimina-tory potential for disease prediction. We also found twenty down-regulation miRNAs in microarray analysis locate in DLK-MEG3 imprinted domain of chromosome 14q-arm. The array CGH data from clinical samples showed that no DLK-MEG3 region deletion among these 40 OSCC patients. We further analyzed these miRNAs, located around CpG islands, to identify tumor-suppressive miRNAs silenced through aberrant DNA methylation. The expression of those miRNAs was restored by DNA methyltransferase inhibitor (50-AZA-dC) treatment in OSCC cells lacking their expression. In addition, expression levels of these miRNAs were inversely correlated with their DNA methylation status in the OSCC cell lines. Conclusions: Our study showed that (a) tumor-speci?c hyperme-thylation in OSCC was an important mechanism causing the down-regulation of miRNAs; (b) a miRNAs expression signature in this study in an independent test population should be evaluated further as diagnostic biomarkers for differentiating OSCC tumor from non-diseased epithelia.
