RESUMO
PURPOSE: In this study, we aimed to identify a DNA methylation pattern suitable for prognosis assessment of muscle-invasive bladder cancer and to investigate metastasis-associated processes regulated by DNA methylation. METHODS: Genome-wide methylation analysis was performed on 23 muscle-invasive bladder tumors by microarray analysis. Validation was performed by the qAMP technique in two different patient cohorts (n = 32 and n = 100). mRNA expression was analyzed in 12 samples. Protein expression was determined using tissue microarrays of 291 patients. Bladder cancer cell lines T24 and 253JB-V were used for functional analyses. RESULTS: Microarray analyses revealed KISS1R, SEPT9 and CSAD as putative biomarkers with hypermethylation in node-positive tumors. The combination of the three genes predicted the metastatic risk with sensitivity of 73% and specificity of 71% in cohort 1, and sensitivity of 82% and specificity of 54% in cohort 2. mRNA expression differences were detected for KISS1R (p = 0.04). Protein expression of KISS1R was significantly reduced (p < 0.001). Knockdown of SEPT9v3 resulted in increased cell migration by 28% (p = 0.04) and increased invasion by 22% (p = 0.004). KISS1R overexpression resulted in decreased cell migration (25%, p = 0.1). CONCLUSIONS: We identified a methylation marker panel suitable to differentiate between patients with positive and negative lymph nodes at time of cystectomy. This enables a risk assessment for patients who potentially benefit from extended lymph node resection as well as from neoadjuvant chemotherapy and could improve the survival rates. Furthermore, we examined the impact of putative markers on tumor behavior. Hence, KISS1R and SEPT9 could represent a starting point for the development of novel therapy approaches.
