'Epigenetic' gene tweaks seem to trigger cancer

22:00 25 July 2014 by Helen Thomson
For similar stories, visit the Cancer and Genetics Topic Guides

You could be forgiven for thinking of cancer as a genetic disease. Sure, we know it can be triggered by things you do – smoking being the classic example – but most of us probably assume that we get cancer because of a genetic mutation – a glitch in our DNA. It turns out that this is not quite the end of the story.

We now have the first direct evidence that switching off certain genes – something that can be caused by our lifestyle or the environment we live in – can trigger tumours, without mutating the DNA itself. The good news is that these changes are, in theory, reversible.

All cells contain the same DNA, but individual genes in any cell can be switched on or off by the addition or subtraction of a methyl group – a process known as epigenetic methylation.

For years, researchers have known that mutations to our DNA – either those passed on at birth or those acquired as a result of exposure to radiation, for example – can cause cancer. But epigenetic changes have also been implicated in cancer because abnormal patterns of gene methylation are seen in virtually all types of human tumours.

For example, a gene called MLH1 produces a protein that repairs DNA damage. It is often mutated in colon cancer tumours, but in some tumour samples the gene is healthy, but appears to have been silenced by methylation.

Teasing it apart

The problem is that it has been difficult to test whether abnormal methylation occurs as a result of a tumour or is a cause of its growth.

"In genetics you can easily delete a gene and see what the consequence is, but it's much harder to direct methylation to specific regions of the genome," says Lanlan Shen of Baylor College of Medicine in Houston, Texas.

To get round this problem, Shen and her colleagues used a naturally occurring sequence of DNA, which draws in methyl groups to methylate nearby genes. They call it their "methylation magnet".

The team inserted this sequence next to the tumour suppressor gene, p16, in mouse embryonic stem cells. These embryos then developed into mice that carry the "methylation magnet" in all of their cells. The team focused on methylating p16 because it is abnormally methylated in numerous cancers.

Gene silencer

They monitored the rodents for 18 months – until they reached the mouse equivalent of middle age. Over this time, 30 per cent of the mice developed tumours around their body, including in their liver, colon, lungs and spleen. None of a control group of genetically identical mice developed tumours.

"Some tissues showed faster methylation than others, for example in the liver, colon and spleen, and that's exactly where we saw the tumours grow," says Shen. "It seems like methylation predisposed the tissue to tumour development."

She reckons that methylation silences p16, which lifts the break that it normally places on any abnormal cell division.

That's all well and good in the lab, but how might hyper-methylation be triggered in the real world? Stephen Baylin, deputy director of the Cancer Center at Johns Hopkins University in Baltimore, Maryland thinks that events that cause inflammation in the body – such as an infection or smoking – would be likely to play a role.

Sometimes you want methylation to temporarily turn off genes, for example, while cells undergo some kind of repair, says Baylin. But if tissues are repeatedly exposed to chemicals produced as a result of inflammation, that silencing signal might not get switched off.

Cause or consequence?

This is very interesting work, says Keith Brown, who researches epigenetics and cancer at the University of Bristol, UK. He says there is a lot of circumstantial evidence that methylation plays a causal role in cancer but this seems to be the first time that someone has shown that it can be a primary cause of tumours.

Adrian Bird, a geneticist at the University of Edinburgh, UK, is more cautious. He says it is still not clear from this experiment whether methylation is fully responsible for the tumours. "By inserting the new sequence into the DNA, perhaps they are altering the gene in some other way," he says.

Baylin agrees. "You don't know if the p16 triggered a series of events that were tied to a genetic mutation," he says. "So you can't rule out that it still requires collaboration with a genetic change."

Either way, it doesn't really matter, says Bird. The study shows definitively that methylation contributes to cancer. "So whether it's the primary cause or not, it may well be an Achilles' heel of cancer – one that we can reverse."

This is what Shen plans to do. "We can use different approaches to reverse methylation – turning back on silenced genes – and see whether we can prevent the tumours from occurring or treat the cancer after it has appeared".

Journal reference: The Journal of Clinical Investigation, DOI:10.1172/JCI76507

Food for thought

If epigenetic changes do drive cancer, there are several ways to potentially flip the switch in the opposite direction.

"The coolest thing from an environmental perspective is that we can ask whether diet can influence this epigenetic process," says Shen.

That's because methyl groups that switch genes on and off epigenetically are not synthesised by the body, they can come from folate in our diet, for example. There is some evidence that dietary changes can accelerate methylation or promote de-methylation.

Shen says her group's research includes testing whether diet has any effect on the epigenetic changes in their mouse cancer models. "Broccoli extract has been shown to have a de-methylation effect so we'd be interested to see if that has any effect on the development of tumours in our model," she says.

"Certain natural products have been shown to interfere with DNA methylation – green tea, for example," says Baylin. "These products are not that potent but certainly there may be ways you could help reduce your risk of cancer with dietary manoeuvres."

A more efficient way to reduce the amount of methylation might be to use drugs. Brown says that de-methylating agents already treat some blood cancers successfully, although it is not clear how they work.

Unfortunately drugs that reverse methylation are not specific to tumour cells – they switch off methylation in all dividing cells. So while they primarily affect fast-dividing tumour cells, they also cause other problems throughout the body.

"If you could reactivate p16 in cancer patients, no cancer would like that at all," says Baylin. "The problem is that it is difficult in patients because the doses needed are so often toxic. If you could find a way to switch these genes back on it would lead to magnificent therapies."