NEW YORK — Using a new method, Swedish researchers have identified mutations in the noncoding regions of the human genome linked to the common pediatric brain cancer, malignant medulloblastoma (MB).
They said their findings expand the number of treatable mutations for MB, which could result in individualized therapies for patients in the long run.
In a paper published in the Proceedings of the National Academy of Sciences on Monday, researchers from Uppsala University in Sweden compared mutations in patients with MB and a benign brain tumor, pilocytic astrocytoma (PA), with positions in the human genome that have been relatively conserved over 100 million years of evolution.
According to the authors, 98 percent of the human genome belongs to non-protein coding regions, which may have essential gene regulatory elements. They wanted to systematically study cancer mutations in these noncoding regions, especially conserved ones. The rationale, they said, was that conserved sequences likely harbor essential gene regulatory functions.
To do so, they started by studying highly conserved regions of the human genome that have stayed constant throughout millions of years of mammalian evolution. For this, they turned to a series of papers by the Zoonomia consortium published in Science earlier this year, which compared 240 mammalian genomes to identify these conserved regions.
They then studied cancer mutations in these conserved regions of the human genome, comparing them with mutations identified in 146 MB and 89 PA patients, which came from whole-genome sequencing studies from the International Cancer Genome Consortium. The authors refer to these as noncoding constraint mutations (NCCMs).
Their analysis showed that MB had higher NCCM accumulation rates than PA. While NCCMs affected the BRAF locus, the most commonly mutated gene in PA, in MB more than 500 genes had high levels of NCCMs.
Of the 200,000 mutations in MB patients, 114 were found in conserved positions in the genome. The authors noted that many of these mutations occurred in genes previously unreported in these cancers and that different mutations were seen in other age groups and subgroups of MB. For instance, mutations in the HOXB cluster were seen in young MB patients and in WASF-2/AHDC1/FGR in adult patients. They also saw that NCCMs changed gene expression in MB cell culture.
"These newly identified novel putative candidate driver genes may aid in patient stratification in MB and could be valuable for future selection of personalized treatment options," the authors concluded.
In their final experiment, the researchers used CRISPR-Cas9 editing technology to introduce a commonly seen NCCM mutation in MB cells. They found that it alters the responsiveness of the cells to a common cancer drug. "Thus, we can show that our method works to identify mutations in noncoding parts of the genome," lead author Karin Forsberg-Nilsson, a researcher at Uppsala University, said in an email.
As the next steps, her team wants to follow up on the 114 mutations in MB by analysis in cell cultures to see which ones might be used for diagnosis and treatment. "The follow-up analysis is time-consuming, and we need to combine large-scale study with lots of lab work to pinpoint each mutation's exact function; in the future, we hope to develop AI tools to move faster when we do this analysis in other cancers," she added.