The cancer community has long accepted conventional dogma that cancer is a disease of mutation.

“This model applies to the vast majority of cancer types,” said Mathieu Lupien, PhD, Princess Margaret Cancer Center and University of Toronto. “However, there are exceptions.”

Lupien explored some of those exceptions during the major symposium Metabolism and Chromatin Deregulation in Cancer and Cancer Heterogeneity and related panel discussion. Closed or compacted chromatin is associated with DNA methylation and forms more densely packed nucleosomes. The closed architecture makes it more difficult for transcription factors to access DNA.

More open chromatin forms nucleosomes that are more accessible to RNA polymerase and other transcription factors. These low-density nucleosomes are more likely to contribute to the resulting phenotype, he explained. Chromatin variants underlie the differential contribution of endogenous DNA sequences across tissues.

“Cancer is a genetic disease that relies on misused endogenous DNA sequences as much as it may rely on mutations,” he said. “It is possible to use chromatin variants to identify genetic determinants of intra-tumor phenotypic heterogeneity.”

In glioblastoma, chromatin variations can confer stemness to some cancer cells. Depending on chromatin characteristics, stem cells group in three clusters. Every tumor contains all three stem cell types in different ratios. Scoring tumors based on the presence of more invasive stem cells can identify patients with the worst outcomes.

“We must acknowledge the differences in these stem cells in designing effective therapies,” Lupien said.

Histone modification such as methylation can also lead to tumorigenesis. Mutations in polycomb repressive complex 2 (PRC2) can lead to misregulated methylation which can bridge and condense nucleosomes, which makes them refractory to activation by transcription factors.

Pediatric ependymomas are the third most common brain tumor in children. An uncharacterized protein, CXorf67, appears to modulate PRC2 activity in a distinct subgroup of pediatric ependymomas, called posterior fossa A ependymoma (PFA EP). Patients with PFA EP have a five-year survival rate around 3 percent.

PFA EP expresses high levels of the protein, renamed enhancer of zeste homologs inhibitor protein (EZHIP), which is associated with poor prognosis. EZHIP associates with PRC2 and leads to massive reductions in trimethylation of H3K27 in PFA EP, noted Peter W. Lewis, PhD, University of Wisconsin School of Medicine and Public Health.

“EZHIP is the first described and characterized oncohistone,” Lewis said. “It may be possible to detoxify EZHIP therapeutically. It is subject to numerous post-translational modifications that may be targetable.”

Another approach may be to target the epigenome or metabolism. PFA EP requires a hypoxic environment and cannot survive even a brief exposure to 21 percent oxygen.

“This requirement for a hypoxic microenvironment probably shouldn’t be a surprise, as PFA EP arises before birth, possibly as early as the first trimester,” said Michael D. Taylor, MD, PhD, University of Toronto and Labatt Brain Tumor Research Centre. “It exhibits diminished methylation of H3K27, which is reversed following a transient exposure to atmospheric levels of oxygen.”

Hypoxia drives S-adenosyl-L-methionine (SAMe) deficiency, which drives H3K27 hypomethylation through histone lysine demethylase activity, which fuels reverse TCA cycle activity, Taylor said.

“This is a complex mechanism, but it is druggable at multiple points,” Taylor said. “Once we can safely deliver either metabolic or epigenetic drugs to the tumor, they may form the backbone of a therapy to effectively and non-toxically treat PFA EP.”