Genome-wide methylation analysis unravels differential DNA methylation patterns between (prognostic) neuroblastoma patient subgroups

Neuroblastoma (NB) has been considered as a disease driven by genetic aberrations. However, recent studies have shown that the NB biology is also strongly determined by the epigenetic profile of the tumor, an important finding underscored by the fact that whole-genome sequencing of primary NB tumors revealed no highly recurrent mutations, apart from activating ALK mutations. By profiling the methylome of more than 100 primary NB tumors using methyl-CpG-binding domain (MBD) sequencing we reveal new insights into the NB epigenome and highlight DNA methylation changes as alternative targets of prognostic biomarker research. For MBD sequencing, primary tumor DNA was sheared and methylated fragments were captured using the high affinity of the MBD of human MeCP2 towards methylated cytosines. This methylated fraction was then paired-end sequenced on the Illumina GAIIx or HiSeq2000 system. After demultiplexing, performing QC and mapping of the raw sequencing reads, peak calling and absolute methylation quantification analyses were conducted in order to identify methylation enriched regions and differentially methylated regions (DMRs) between (prognostic) patient subgroups. We show remarkable different methylation patterns between tumors with and without MYCN amplification, stage 4S and other tumors, and different prognostic patient subgroups (low-risk survivors, high-risk survivors and high-risk non-survivors). Top ranking differentially methylated regions were subsequently selected for further validation on an independent cohort of more than 200 primary tumors using our previously established methylation-specific PCR (MSP) technology. In conclusion, this unique data set sheds new light on molecular alterations of the NB methylome and demonstrates the clinical utility of DNA methylation biomarkers in NB risk stratification.