Genome-wide methylation profiling identifies novel methylated genes in neuroblastoma tumors. Issue 1 (2nd January 2016)

• Record Type: Journal Article
• Authors: Olsson, Maja; Beck, Stephan; Kogner, Per; Martinsson, Tommy; Carén, Helena
• Is Part Of: Epigenetics. Volume 11:Issue 1(2016)
• Page Start: 74
• Page End: 84
• Publication Date: 2016-01-02
• Subjects: CIMP -- DLX5 -- DNA methylation -- epigenetics -- neuroblastoma -- PCDHGA4 -- pediatric -- TERT -- 450K; Epigenesis -- Periodicals; Epigenetica; 572.86505
• DOI: 10.1080/15592294.2016.1138195
• Languages: English
• ISSNs: 1559-2294

Abstract

ABSTRACT: Neuroblastoma is a very heterogeneous tumor of childhood. The clinical spectra range from very aggressive metastatic disease to spontaneous regression, even without therapy. Aberrant DNA methylation pattern is a common feature of most cancers. For neuroblastoma, it has been demonstrated both for single genes as well as genome-wide, where a so-called methylator phenotype has been described. Here, we present a study using Illumina 450K methylation arrays on 60 neuroblastoma tumors. We show that aggressive tumors, characterized by International Neuroblastoma Risk Group (INRG) as stage M, are hypermethylated compared to low-grade tumors. On the contrary, INRG stage L tumors display more non-CpG methylation. The genes with the highest number of hypermethylated CpG sites in INRG M tumors are TERT, PCDHGA4, DLX5, and DLX6-AS1 . Gene ontology analysis showed a representation of neuronal tumor relevant gene functions among the differentially methylated genes. For validation, we used a set of independent tumors previously analyzed with the Illumina 27K methylation arrays, which confirmed the differentially methylated sites. Top candidate genes with aberrant methylation were analyzed for altered gene expression through the R2 platform (http://r2.amc.nl ), and for correlations between methylation and gene expression in a public dataset. Altered expression in nonsurvivors was found for the genes B3GALT4 and KIAA1949, CLIC5, DLX6-AS, TERT, and PIRT, and strongest correlations were found for TRIM36, KIAA0513, and PIRT . Our data indicate that methylation profiling can be used for patient stratification and informs on epigenetically deregulated genes with the potential of increasing our knowledge about the underlying mechanisms of tumor development.