Hypermethylated Long Non-Coding RNA Genes KCNK15-AS1, MAGI2-AS3, and SSTR5-AS1 in Ovarian Cancer and Their Diagnostic Potential.

Abstract

We identified new diagnostic markers for ovarian cancer based on the analysis of aberrant methylation in a group of long non-coding RNA (lncRNA) genes. Using quantitative methyl-specific PCR with the nonparametric Mann-Whitney test, a significant (p < 0.01) increase in the methylation levels of three lncRNA genes (KCNK15-AS1, MAGI2-AS3, SSTR5-AS1) was demonstrated in a representative cohort of 56 clinical ovarian cancer samples. Using quantitative reverse transcription PCR on the same sample set, a significant decrease in the expression of these three lncRNA genes (p < 0.01) and a statistically significant inverse correlation between changes in methylation and expression levels (r < -0.5) were demonstrated, which confirms the functional role of methylation in suppressing their expression in tumor tissues. ROC analysis was used to assess the diagnostic potential of these genes. It was shown that the detection of methylation in at least one of the three genes in tissue DNA samples allows for identification of ovarian cancer with 98% sensitivity and 100% specificity (test > 0, AUC = 0.989). The developed marker panel may be of interest for the molecular assessment of at-risk patients when determining a preliminary diagnosis. However, the method requires validation on a larger, more representative sample set.