Genetic-epigenetic interactions in medulloblastoma development

Abstract

Medulloblastoma is the most common malignant brain tumour of childhood. Transcriptomic profiling has revealed the existence of four core molecular subgroups (SHH, WNT, Group 3 and Group 4) with distinct clinical, pathologic and molecular characteristics. However, the specific molecular events associated with tumour development in these groups are poorly understood. DNA methylation plays a key role in epigenetic transcriptional regulation, and promoter hypermethylation leading to gene silencing is a common feature of medulloblastoma. DNA methylation profiling has identified distinct methylomic profiles associated with the four subgroups of medulloblastoma, and the wider role of DNA methylation in medulloblastoma now requires investigation. Using two high-throughput screening approaches, this project therefore undertook a comprehensive investigation into the potential role of specific DNA methylation events in the development of the distinct subgroups of medulloblastoma. Using DNA methylation profiles, which were generated for 216 medulloblastomas using the GoldenGate methylation array, the first approach identified 73 CpG methylation markers (encompassing 63 genes) which significantly distinguished Group 3 and/or Group 4 medulloblastomas. Subgroup-specific differential gene expression analysis showed that, for the majority of the methylation markers identified, there was no clear inverse association between methylation and gene expression. One gene (RHOH) was identified which showed strong evidence of epigenetic dysregulation in medulloblastomas. RHOH methylation represented a potential epigenetic event in Group 4 tumours; 51% of Group 4 medulloblastomas showed aberrant hypomethylation of multiple RHOH promoter-associated CpG residues, which was associated with upregulated RHOH expression in Group 4 tumours. RHOH was re-expressed in 4 out of 6 methylated cell lines following treatment with the demethylating agent 5’-aza-2’-deoxycytidine (5-azaCdR). This study has thus identified a novel putative oncogenic role for RHOH in Group 4 medulloblastoma development. In the second approach, a functional epigenomics screen identified 283 genes which were upregulated in 2 or more cell lines investigated (n=10) following 5-azaCdR treatment. Assessment of DNA methylation using the Illumina 450K methylation array identified 160 CpG residues (encompassing 21 of the 283 genes) whose methylation status was consistent with expression alterations observed after 5-azaCDR, and methylation-dependent gene regulation, in cell lines. 9/160 CpG residues (6%) showed evidence of subgroup-specific differential methylation which was concordant with differential gene expression and potential epigenetic gene regulation in medulloblastoma subgroups. These 9 sites represented 5 candidate genes (ACTC1, ANXA2, FAM46A, PRPH and S100A4). Aberrant hypermethylation of multiple gene body CpG residues was associated with FAM46A silencing in non-SHH tumours, while aberrant hypermethylation of multiple promoter-associated residues was associated with ACTC1 silencing in Group 3 and Group 4 medulloblastomas. Single site v hypomethylation events associated with upregulated expression in WNT tumours were identified for ANXA2, PRPH and S100A4. This study has identified six genes with putative oncogenic or tumour suppressor roles in the development of distinct subgroups of medulloblastoma through their epigenetic dysregulation. Further work is now required to validate these findings and to assess their functional significance in medulloblastoma subgroups, as well as their potential relevance in medulloblastoma sub-classification and prognostication.