We performed a quantitative analysis of the methylation levels of the VIM, KRT18, RASSF1A, NKX3.1, CDH1, and PTEN genes in prostate cancer cell lines, prostate cancer tumor tissues and adjacent normal tissues. As a result, we found that the cell lines differed in the level of methylation of the studied genes, in particular, the methylation of the tumor suppressor RASSF1A in LNCaP cells was lower than in the other two cell lines. Very low levels of methylation were detected for CDH1 in all cell lines, and the VIM gene, on the other hand, was completely methylated, with a similar phenomenon observed in patients' tissues. Correlation analysis of the expression of these genes, their methylation and clinical and pathological characteristics of patients revealed a negative correlation between the level of methylation of KRT18 and the Gleason score in prostate adenocarcinomas. We also found that the levels of methylation of RASSF1A were higher in the group with a TMPRSS2: ERG fusion transcript.
Methylation of all of studied genes has been shown to be more common in patients with prostate cancer than in healthy donors, but was also common for people with inflammatory diseases of the genitourinary system. The best combination to differentiate between patients with PCa and those without cancer is a combination of methylated genes PTEN, CDH1, NKX3.1, RASSF1A and GDF15.
After analyzing the methylation of VIM, TMEFF2, RASSF1A, NKX3.1, MYO3A and GDF15 genes in the urine of patients with bladder cancer, we found that the MYO3A gene methylation was observed in all patients with inflammation and although its methylation did not occur in healthy donors, it is not able to distinguish between cancer patients and non-cancer individuals. It was determined that the cancer patients from non-cancer individuals can be distinguished by the presence of methylation of VIM, TMEFF2, GDF15, RASSF1A and NKX3.1.
We analyzed the relative expression of Aurora kinase genes in adenoma, adenocarcinoma, and paired adjacent prostate cancer tissues. We found a statistically significant difference in the relative expression of the AURKA gene in malignant tumors and adenomas with lower expression in malignant neoplasms. There was a positive correlation of the relative expression levels of the AURKA and AURKB genes, the correlation of the expression of these genes with MKI67, NKX3.1 gene. AURKA expression positively correlated with ARisof1 expression and AURKC expression positively correlated with expression of CYP17A1.
In urine-precipitated cells, AURKA gene expression was significantly higher in patients with PCa than in non-cancer patients. The expression of BRAF in patients with PCa strongly correlates with the expression of the AURKC, but with the expression of AURKA and AURKB in cells, precipitated from urine of people without cancer.
The results of bioinformatics analysis revealed, that PANDAR lncRNA can interact with a number of miRNAs involved in the regulation of oncogene expression. In the LNCaP cell line, the level of PANDAR in cells was the highest among other cell lines, while the lowest relative expression of PANDAR was in PC3 cells, which, at the same time, demonstrated the high levels of circulation PANDAR in the conditioned medium of these cells. We did not find a statistically significant difference in the levels of PANDAR lncRNA in the cell-free urine of patients with PCa and relatively healthy individuals. We found positive correlations between PANDAR levels and AURKC and BRAF expression in cells from these samples.
We found a difference in levels of PANDAR lncRNA in the urine of patients with bladder cancer and relatively healthy individuals. We hypothesized that by adding PANDAR levels to methylation-based markers we could obtain better results than for each alone. Indeed, simple logistic regression analysis and ROC analysis have made it possible to distinguish bladder cancer patients from cancer-free individuals quite accurately by combining these two types of potential markers.
