ABOVE: An artist’s rendering of the inside of a nucleus
© ISTOCK.COM, SELVANEGRA

Most cancer cells have genomes that are much less methylated than those of normal cells, but whether this loss of methylation, an epigenetic process, has any functional meaning for the cells has long been an unanswered question. Now researchers show that the loss of DNA methylation across the genome changes the timing of DNA replication and alters the shape of the 3-D compartmentalization of DNA, which helps steer gene expression.

The study, published September 21 in Cell Reports, is “an elegant dissection of the impact of DNA methylation on 3-D genome organization,” says Emma Bell, a bioinformatician at the Princess Margaret Cancer Center in Toronto who did not participate in the work. “It’s really important to show how extensively aberrant DNA methylation, which is a common artifact of cancer, impacts higher order genome organization and DNA replication.”

For the past 25 years or so, Susan Clark and her group at Garvan Institute of Medical Research in Australia have been interested in how epigenetics is involved in cancer. As most cancer cells lose DNA methylation throughout the genome, it was surprising how little was known about the global consequences of the hypomethylation, Clark explains in an email to The Scientist. 

Genome methylation is a kind of a fingerprint for a cell, and cell identity is also guided by genome organization into 3-D compartments that help determine what genes are expressed. Clark and colleagues hypothesized that DNA methylation plays a role in the maintenance of this genome architecture, and, in the new study, used colorectal cancer cells to investigate that hypothesis.

The researchers started with two versions of the same cells: one normal cancer cell line and one with knockouts of two DNA methyltransferase enzymes, which copy the methylation pattern of a DNA strand to daughter DNA strands generated during replication. The cells lacking DNA methyltransferases had reduced genome-wide methylation levels. Then, in individual cells from both cell lines, they mapped DNA replication throughout the genome and investigated the genomes’ 3-D organization. 

The team found that hypomethylation resulted in a shift in replication timing—that is, how early or late during replication a region was copied. Most of those shifts were moderate: regions were replicated slightly earlier or later in genomes that lacked methylation. But more than three percent of the hypomethylated genomes were replicated much earlier or later than in the cells with intact DNA methyltransferases. 

It is “striking that they were able to see these differences in replication timing when they got rid of the methyltransferases,” says Christine Cucinotta, who studies chromatin architecture at the Fred Hutchinson Cancer Research Center in Seattle and was not involved in the work. Replication timing, she adds, is “generally pretty robust outside of cell fate transitions.”

These shifts in replication timing also appeared to affect the 3-D organization of genomic regions, particularly in so-called partially methylated domains, which are places that in the regular cancer cells have lower levels of DNA methylation. This DNA architecture, in turn, plays a role in gene expression. For instance, where DNA replication shifted earlier in hypomethylated cells, there was a shift toward 3-D structures indicative of more active gene expression, which the researchers also observed. They determined that the loss of precision in replication and associated changes in 3-D compartmentalization and in gene expression especially affected cancer-related genes. 

“These findings highlight the role of epigenetics in cancer progression that may help explain how cancers can become more heterogeneous with each cell division,” Clark writes. 

Clark’s group also found evidence to suggest the cell can lessen the impact of reduced methylation to some extent. In the regions of hypomethylated cells where replication was delayed, the team found an increase in swaths of DNA in which histone H3 had been modified by both repressive chromatin marks as well as those associated with DNA that is more accessible to transcription. This new, dual layer of regulation may “suppress overall disruption of gene transcription across compartments,” Clark tells The Scientist. 

This work confirms the proposed link between DNA methylation, histone methylation, changes to transcription, and replication timing, says Susan Gasser, a molecular biologist affiliated with the University of Basel and the Friedrich Miescher Institute for Biomedical Research in Switzerland who did not participate in the study. “The important thing is not to say that DNA methylation is the trigger of the whole cascade,” she adds. “I think we see that there’s a balance of these four things, and that there are regions of the genome that are hypersensitive to . . . these fluctuations.” 

One next step would be “a comparison between healthy cells, precancerous cells, and cancer cells,” Bell tells The Scientist. “What we’ve got represented here is a picture of cancer and aberrant DNA methylation, but that doesn’t just happen all at once. I’d like to see what happens in DNA methylation, 3-D genome organization, and DNA replication timing during the process of oncogenesis.”