Of Mice and MYC : Modelling Medulloblastoma

Brain tumours are the leading cause of cancer-related paediatric deaths, with medulloblastoma (MB) being the most common malignant paediatric brain tumour. MB is stratified into four major subgroups – WNT, SHH, Group 3, and Group 4, nomenclature defined by key pathways and drivers involved within each subgroup. Group 3 MB is the most moribund subgroup of MB with a 5-year overall survival of less than 50%. It is commonly termed the ‘MYC’ subgroup due to focal high-level amplification of the MYC gene in 20% of these patients. To investigate the involvement of MYC in MB development, in paper I we generated a transgenic murine model of MYC-driven MB (GMYC) where aberrant MYC expression is constitutively driven from the Glutamate-transporter-1 (Glt1) promoter. Tumours develop spontaneously 3-6 months postnatally and recapitulate tumour histology seen in patients. Suppression of MYC in tumour-bearing mice led to clearance of cancerous cells, indicating a necessity for MYC in maintenance of GMYC tumours. Our novel GMYC model was compared to our previous GTML model (a transgenic model of Group 3 MB driven by MYCN), as well as to clinical patient data. GSEA revealed significant differences in the genetic pathways driving both mouse models. The Cdkn2a tumour suppressor gene was expressed at significantly higher levels in our GTML model compared to our GMYC model. Subsequent investigation of this gene revealed a methylation signature seen only on Group 3 patients with high MYC expression. Treatment using the HSP90 inhibitor, Onalespib, restored ARF in vitro and promoted increased survival in our animal model, suggesting its therapeutic potential for children affected with MYC-driven, ARF-silenced brain cancer. In paper II, we next investigated the putative cell-of-origin for GMYC tumours by crossing the GMYC mouse model with a fluorescent reporter system to track tumour development and comparison to normal, developing brains. In this paper, we show GMYC tumours arise from a stem/progenitor cell that likely develops in an extra-cerebellar location prior to clonal expansion and the full extent of tumour formation. Lastly, in paper III, we established transcriptional networks specific for Group 3 and 4 MB with a focus on gene interactions involving chromosome 17q genes. KIF18B was identified as one such gene located on Chr17q that may have a role in Group 4 MB pathogenesis. These transcriptional networks demonstrate a promising means of identifying novel cancer-related genes and their link to other regulatory genes known to be involved in tumourigenesis