Abstract
Group 3 medulloblastoma (MB) carry the worst prognosis of all MB. The transcription factors MYC and MYCN are suggested drivers for a subset of these tumours; with MYC amplifications (17–20%) representing the most common genetic alteration in Group 3 tumours, while MYCN amplifications (4–6%) are less frequent. To improve current treatment options for these patients, it is of crucial importance to decipher differential features of MYCN- and MYC-driven MB and to establish accurate animal models for these patients. By driving MYC from the hindbrain-specific Glutamate transporter 1 (Glt1) promoter using a Tet-OFF system, we have established a novel murine model of MYC-driven MB (GMYC) which accurately recapitulates aggressive Group 3 MB. GMYC tumours develop without any p53 mutations and with ~70% penetrance. Tumour-prone GMYC mice can further be cured by MYC-depletion through dox treatment. Comparison of transcriptional profiles between GMYC and our MYCN-driven GTML tumours revealed that both models accurately represent Group 3 MB, while showing differential expression of key features of MYC- or MYCN-driven tumours. CDKN2A was identified as one of the top upregulated genes in our GTML model as compared to our GMYC model. CDKN2A encodes two tumour suppressors, p16INK4A and p14ARF, which are key regulators of cell cycle progression and activation of p53. Similar enhancement of CDKN2A is observed in MYCN-amplified as compared to MYC-amplified Group 3/4 patients. Tumour formation following partial or complete knockout of CDKN2A significantly increased tumour penetrance in GTML as compared to GMYC animals. Similarly, CDKN2A levels significantly correlate with poor prognosis in MYCN amplified MB patients, while CDKN2A levels are low in MYC amplified patients. This suggests that MYC is regulating and suppressing CDKN2A during MB formation and further advocates that pharmacological restoration of CDKN2A would be a potential future therapy for this group of high-risk MB patients.