Traditional cancer driver gene discovery has generally focussed on coding mutations, which can affect protein function and structure. Mutations in noncoding genes, which make up over 98% of the genome, have been much less widely studied. A recent study published in The Proceedings of the National Academy of Sciences (PNAS) found that these noncoding mutations may play a key role in the development of medulloblastoma, the most common malignant brain tumour in children, and could be specifically targeted using personalised medicine.
The unchanged part of the genome
To understand the key genes responsible for medulloblastoma development, researchers at Uppsala University in Sweden studied mutations in conserved portions of the genome, those which have not altered even following millions of years of evolutionary changes. The group’s hypothesis was that mutations in conserved genome positions are more likely to be important and functional than those in positions with low evolutionary conservation. “We focused on mutations at the best-preserved positions, as these are likely the most critical ones for gene regulation. This way we can sort out the most important mutations, which would otherwise not be possible, and then test their functionality”, stated Karin Forsberg-Nilsson, principal investigator of the group.
Groupings by mutation status
Using whole-genome sequencing, medulloblastoma samples were compared to samples of pilocytic astrocytoma, a benign type of brain tumour, to identify genes enriched in noncoding constraint mutations (NCCMs) that were likely to be regulatory in nature. Over 81% of paediatric medulloblastoma patients were found to have at least one NCCM, with 39% of patients having NCCMs in more than one locus. Several loci with NCCMs were found to be associated with age of onset, such as the HOXB cluster in young medulloblastoma patients, and the WASF-2/AHDC1/FGR locus in adult patients.
From this analysis, the researchers were able to identify four key molecular subgroups of medulloblastoma, of which the WNT subgroup, which carries germline APC mutations, had the best prognosis in all age groups. Other subgroups, such as the SHH subgroup, which has mutations in PTCH, SMO and SUFU from the SHH pathway, were found to have different levels of mutational burden in adults compared to children. The remaining two groups, named Groups 3 and 4, were associated with metastatic disease at diagnosis. Group 3, which only occurs in infants and children, is characterised by MYC amplification and had the worst clinical outcomes, with 5-year survival rates under 40%. Group 4, characterised by MYCN and CDK6 amplifications and intermediate outcomes, was found to be the most common subgroup, accounting for 35% of cases.
Targets for personalised medicine
Following the grouping of patients by their mutations, the team were also able to identify some specific NCCMs which could alter the responsiveness of medulloblastoma cells to treatments. One such example led to increased expression of the SRC kinase FGR, resulting in increased responsiveness of medulloblastoma cells to dasatinib, a SRC kinase inhibitor.
These newly identified putative candidate driver mutations may aid in patient stratification. This would enable the use of personalised medicine in medulloblastoma patients, with treatment choices made depending on which mutational subgroup they fall in to. As whole-genome sequencing is now offered to all children diagnosed with brain tumours in Sweden, the ability to choose a treatment course based on the patient’s specific mutations may significantly improve outcomes.