Featured Articles

Epigenetics may have many roles

By Sue Goetnick Ambrose, Dallas Morning News

2 November 2003: While the human DNA code is at the root of all body processes, scientists are starting to appreciate that the code is coated with extra biological instructions. Although studies involving this extra layer of information are in their infancy, there are tantalizing hints that this emerging science of epigenetics may reach into many aspects of human life and disease.

Press Photo

Some scientists believe that nutrition, cancer, in vitro fertilization – even autism – may be among the arenas where epigenetics plays a role.


You are what you eat. But you may also be what your mother, father and grandparents ate before you.

Besides passing on genes, mom, dad, grandma and grandpa can also give their kin another type of legacy – little bits of biological information – attached to those genes. And new research is showing that the nature of these "epigenetic" legacies can be dictated by what parents eat before their children are born.

"It's becoming unavoidable to conclude that there's something going on beside just genetics," said Robert Waterland, a nutritionist at Duke University Medical Center in Durham, N.C.

In people, a forebear's diet can influence obesity, heart disease, high blood pressure and diabetes. In mice, diet can influence body weight, blood pressure and even fur color. Because examples of this inheritance are only starting to emerge, scientists still don't know the full extent to which diet influences health. But they may be starting to figure out how.

These female mice are genetically identical, yet a change in the diet of the pregnant and nursing mothers is the reason the mouse on the right is brown instead of yellow and thin instead of obese.

For instance, a study published in August in Molecular and Cellular Biology by Dr. Waterland and Duke's Randy Jirtle showed that in a particular strain of mice, fortifying the pregnant and nursing mother's diet with four nutrients shifted babies' coat color from yellow to brown, and made the mice less obese.

The nutrients – vitamin B12, folic acid, choline and betaine – help the body add chemical flags called methyl groups onto genes. In the Duke study, the researchers were able to trace the color difference to an increase in methyl groups on a gene that controls fur color and body weight.

The researchers don't know which of the four nutrients caused the increase in methyl groups. But the study raises some interesting questions about nutrition, the scientists said. In the Duke University experiments, the nutrients helped the mice, making them less obese. However, the nutrients may have affected other genes too, but not for the good of the mouse.

It's possible that diet influences genes in people too, said Dr. Jirtle. Folic acid, for instance, is added to flour and other grains in the U.S. food supply to decrease the rates of certain birth defects. But it's possible, in theory, that the nutrient may have adverse effects too.

"I'm not saying folic acid is the problem. But we make the assumption that when we expose everybody it's innocuous," Dr. Jirtle said. "It could be that some people are more susceptible. The 99.9 percent who don't even need it could be harmed by it."

Other studies also suggest that nutrition may have effects that ripple through a family tree. A 2002 study by researchers at Umea University in Sweden found that food abundance just before a boy hit puberty affected the health of his son's children. Poor food availability protected against death from heart disease. Good availability increased the risk of diabetes in grandchildren.

The Swedish scientists can't say for sure why a paternal grandfather's diet affects his grandchildren's health. But the scientists suggest that eating too much or too little while sperm are beginning to mature may result in subtle chemical editing of DNA. These edited patterns may be passed on to new generations.

Researchers are just now developing the tools to study how nutrition may cause these epigenetic changes, Dr. Jirtle said.

"The whole field of nutrition," he said, "is going to be turned on its damned head."

In Vitro Fertilization

Since the first test tube baby was born in 1978, conceiving children outside the womb has become almost routine.

But recent research suggests that joining mother's and father's genes in a lab dish may carry a previously unappreciated risk.

In general, babies conceived through assisted reproductive techniques are healthy and normal. But a handful of studies have hinted that these techniques may increase the risk of certain conditions, which, in naturally conceived children, have been linked to errors in the usual programming of DNA.

A 2002 study, for instance, suggested a link between Angelman syndrome, a condition marked by mental retardation and other problems, and in vitro fertilization.

Three reports published this year suggested a similar connection for Beckwith-Wiedemann syndrome, which can cause increased organ growth and cancer.

"It's compelling, but not absolutely conclusive evidence, that there is an association," says Dr. Michael DeBaun, a pediatrician at Washington University School of Medicine in St. Louis who conducted one of the studies, published in January in the American Journal of Human Genetics.

Dr. DeBaun and Dr. Andrew Feinberg, of Johns Hopkins School of Medicine in Baltimore, were studying patients with Beckwith-Wiedemann syndrome. The syndrome is linked to a stretch of genes on human chromosome 11.

Normally, these genes are programmed with so-called epigenetic information, in this case a characteristic pattern of chemical flags called methyl groups attached directly to the genes.

But in children with the syndrome, this pattern of flags is lost. As a result, the genes don't work right, causing the syndrome's distinctive features.

Drs. Feinberg and DeBaun noticed that among a group of patients with Beckwith-Wiedemann syndrome, the rate of assisted reproductive techniques was almost 5 percent.

Because less than 1 percent of children born in the United States are conceived through these techniques, the researchers speculate that the technique could sometimes cause the condition.

However, Dr. Feinberg noted that the reverse was also possible. It might be, he said, that some couples are infertile because they can't place the chemical flags properly on their DNA. In that case, the joining of sperm and egg in the lab dish would not be causing the syndrome.

Other researchers have noted that babies and animals born through in vitro fertilization have other differences from those conceived naturally. In vitro babies have, on average, a lower birth weight. Mice born through the same technique also weigh less than normal, while sheep and cattle tend to weigh more.

Many genes that affect growth are naturally flagged with methyl groups. So it could be, scientists have speculated, that when the very young embryo incubates in the lab dish rather than the womb, the decoration of DNA with chemical flags doesn't happen normally.

The next step, Dr. DeBaun said, is to scrutinize the records documenting the in vitro fertilization procedures that led to the children with Beckwith-Wiedemann syndrome. The scientists hope to find a common denominator that could explain the condition.

And scientists also need to investigate the link between in vitro techniques and general health more thoroughly, Dr. Feinberg said. Since the first in vitro baby is only 25 years old, researchers still don't know how the technique may impact health later in life.


For decades, cancer was blamed on broken genes. But in the past few years, scientists have found that broken genes aren't the only thing that lets tumor cells grow out of control.

Simply painting tiny chemical groups on genes can also encourage a cancer to flourish. This realization, scientists say, may make it easier to discover drugs that halt tumors in their tracks, or to detect minute traces of cancerous cells.

Normal body cells have a system of checks and balances to make sure they grow and die appropriately. When this system fails, cells replicate too much or don't die off as often as they should, resulting in a tumor.

Scientists know that sometimes this system failure is due to genetic mutations – misspellings or gaps that pop up in the genes that normally control a cell's growth. If enough mutations piled up in a cell, the theory held, a tumor would be well on its way.

But now scientists realize that something besides genetic mutations can disrupt a gene. Chemical modifications, termed epigenetic changes, may also be at work.

"Genetics and epigenetics are partners in driving the cancer process," says Dr. Steven Baylin, a tumor biologist at Johns Hopkins School of Medicine in Baltimore. "We're in an early understanding of all the epigenetic changes that play a role."

These epigenetic changes come in the form of tiny chemical assemblies called methyl groups that get attached to DNA. Or, the natural protein packaging that cradles genes inside cells may also acquire tiny chemical flags. Sometimes these flags can silence genes. And if that happens on a gene that normally prevents a cell from replicating, then the stage is set for the cell to divide rampantly.

The good news, Dr. Baylin says, is that these epigenetic changes may be easily reversed with drugs. Scientists have been unable to repair genetic mutations in cancer cells. But it may be easier to find medications that remove chemical flags from genes.

Some drugs are already in the testing phase, Dr. Baylin says.

"Right now it looks very promising," he says. "But it remains to be seen what all this is going to mean for the control and prevention of cancer."

Scientists are also trying to exploit these epigenetic changes to pick out a small number of cancer cells in a sea of normal ones.

Dr. Adi Gazdar, a cancer researcher at the University of Texas Southwestern Medical Center at Dallas, is investigating whether methyl groups on particular genes can indicate cancerous cells in the blood or bladder.

A test, or assay, to measure methyl groups on genes is technically easier than searching for misspellings or gaps in the genetic code, Dr. Gazdar says.

"Once you've got an assay going, it's very, very sensitive," he says.


Reported incidence rates have been rising for autism, a neurological disorder that impairs communication and other social skills. Researchers do not know the reason for the rise, but because autism runs in families, scientists have concluded that genes play a role in the disease.

But genetics may not tell the whole story.

Researchers at the Baylor College of Medicine in Houston and their colleagues are proposing that epigenetic information – chemical changes layered on top of the genetic code – may also be a factor.

The epigenetic idea centers on chemical flags, called methyl groups, which cells attach to their DNA. These flags signal cells to activate or silence genes.

Folic acid, a supplement found in prenatal vitamins and added to flour, bread and other grain products, aids in adding these methyl groups to DNA.

Folic acid supplements decrease the risk for neural tube defects in developing babies. A report scheduled for presentation this week at the annual meeting of the American Society of Human Genetics proposes that too much folic acid in a pregnant woman's diet may increase the risk for autism.

The scientists, led by Baylor geneticist Dr. Arthur Beaudet, have proposed that in autism, a gene on human chromosome 15 may not acquire the appropriate pattern of methyl group flags.

This gene has been previously linked to Angelman syndrome, a condition marked by mental retardation, certain traits of autism and other symptoms.

Other researchers stress that the Houston team's idea is only a hypothesis.

"I think it's a very interesting idea," said James Sutcliffe, a human geneticist at Vanderbilt University in Nashville.

But, he noted, "it's an intellectual construct right now for which there's really not much proof."

Dr. Sutcliffe said he favors the idea that autism occurs when someone inherits a set of genes that, together, increases the risk for the condition.

Nonetheless, other research has hinted at a link between autism and epigenetic problems, he said.

For instance, about 2 percent of children with autism inherit an extra copy of a particular stretch of DNA from their mother. This stretch is saddled with epigenetic information that influences how cells read the DNA