Selection of endogenous control and identification of significant microRNA deregulations in cervical cancer

Cervical cancer is a malignant disease that causes around 350 000 deaths annually. One of the essential things in enhancing the survival rates of cervical cancer is the availability of high-quality, sensitive, and specific diagnostic tests capturing the early stages of the disease. In this study, we analyzed the differential expression of microRNA in cervical cancer FFPE tissue samples. The comprehensive miRNA expression profile was detected using modern small-RNA sequencing technology, and the results were validated by real-time PCR. The relative expression of selected miRNAs was determined using the 2-ΔΔCt method. Moreover, two strategies for endogenous control selection were compared. miRNA profiling by small-RNA sequencing and a commercially supplied microfluidic card with 30recommended endogenous controls predesigned by the manufacturer. The RefFinder algorithm and coefficient of variation were used for endogenous control selection. A combination of miR-181a-5p and miR-423-3p was chosen as the most optimal normalizer. Statistically significant upregulation ofmiR-320a-3p, miR-7704, and downregulation of miR-26a-5p was detected, so we propose the combination as a new potential miRNA cervical cancer diagnostic panel. Using ROC curve analysis, the proposed panel showed 93.33% specificity and 96.97% sensitivity with AUC = 0.985. Inconclusion, our study suggests an optimal combination of two endogenous miRNAs for data normalization of miRNA expression studies in the cervical tissue (miR-181a-5p and miR-423-3p) and a novel diagnostic marker panel of three miRNAs (miR-320a-3p, miR-7704, miR-26a-5p) for cervical cancer. Two pools of extracted RNA from malignant and non-malignant samples were prepared for sequencing analysis. The RealSeq®-Biofluids, Plasma/Serum miRNA Library Kit for Illumina® sequencing (RealSeq Biosciences, Santa Cruz, CA, USA) was used for library preparation.