Controlling Your Genes

The promise—and the hype—of changing your DNA through behavior.

Of all the discoveries about how experiences can reach into a creature's very DNA to turn genes on or off, I confess that my favorite is still one of the earliest, from 2004. That summer, scientists unveiled experiments in which they showed that how a mother rat treats her newborns—specifically, whether she attentively licks and grooms them, in the rodent version of mother love—affects whether certain genes in the brain are silenced or activated, with dramatic consequences for the rats' later behavior. But despite a lot of hope and hype, studies showing how experience alters genes have been few and far between—which is why a new one on smoking and diet caught my eye.

The study of these kinds of changes in genes is called epigenetics. Crucially, the changes do not involve alterations of gene sequences, those famous A's, T's, C's, and G's that the Human Genome Project figured out. When I first wrote about epigenetics, in 2003, it was pretty much a backwater, with a lot of scientists annoyed that (after all their hard work on sequencing) this upstart field—which claimed that DNA sequences are merely the beginning when it comes to understanding genetic traits—was raining on their parade. Now we're in danger of going too far in the other direction. Dr. Mehmet Oz's new book on pregnancy, YOU: Having a Baby, plays up epigenetics, which Oz has also been blogging on and discussing on his new TV show, saying that what a woman eats while pregnant can turn some genes on and some off. Time magazine's cover story this week promises that epigenetics will show how "the choices you make can change your genes."

True. But like so much else in medicine, epigenetics is at real risk of being overhyped. Although we know in principle that what Mom eats while pregnant can turn genes on or off, no one has any idea how to put that to practical use. Sure, you can eat lots of foods that contain the "off" switch—it's called a methyl group and consists of one carbon atom and three hydrogens—and hope that you turn off, say, cancer- or diabetes- or obesity-causing genes in Junior. But what if flooding the fetus with methyl groups turns off healthy genes?

The good news is that scientists are now making specific, actionable discoveries in epigenetics. This week, for instance, researchers are reporting that eating leafy green vegetables, folate (found in these veggies as well as in some fruits and in dried beans and peas), and multivitamins can affect the epigenetics of genes involved in lung cancer in a way that could reduce the risk of getting the disease, especially from smoking.

Steven Belinsky of the Lovelace Respiratory Research Institute in Albuquerque and colleagues examined cells coughed up by 1,101 smokers and ex-smokers. They then looked at the extent of methylation—the on/off status—of eight genes in the cells. In previous research, methylation of all eight genes had been linked to a higher risk of lung cancer; that is, gene off = higher likelihood of getting the disease.

That made sense, given the function of the genes. P16, for instance, is a tumor-suppressor gene. When its on switch (called the promoter) is methylated and thus turned off, p16 is silenced and the cell is denied p16's tumor-suppressing function, as Belinsky and colleagues reported in 1998. Cigarette smoke, for instance, contains carcinogens that methylate p16. Smoke also methylates the gene called MGMT, which repairs DNA. Without MGMT, damaged DNA can make a cell cancerous. "Aberrant gene methylation is a known mechanism in the development of cancer from cigarette smoke carcinogens," said Jacob Kagan of the National Cancer Institute in a statement. As Belinsky and his team describe in a new paper posted online in the journal Cancer Research, they found that eating a lot of leafy green vegetables, folate, and taking multivitamins containing vitamin C, carotenoids, lutein, folic acid, and vitamins A and K was strongly associated with lower methylation of these cancer-related genes.

A mere 12 servings a month of leafy greens reduced DNA methylation in these genes about 20 percent; taking a multivitamin reduced it almost 50 percent, says Belinsky. As a result of lower methylation, these beneficial genes—suppressing cancer and repairing DNA—remained on. So it seems that these foods and vitamins can counteract the effect of cigarettes on DNA: although the carcinogens in smoke turn off the beneficial genes, these foods keep them in the game.

And it isn't just cancer. A new study from scientists at Cambridge University found a connection between heart disease and DNA methylation. In a nutshell, they found that particular regions of the DNA in the diseased hearts contained DNA-silencing marks—methylation—while healthy hearts did not. "There is already good evidence that these [methylation] marks are strongly influenced by environment and diet," said Cambridge's Roger Foo. "We found that this process is different in diseased and normal hearts. Linking all these things together suggests this may be the 'missing link' between environmental factors and heart failure."

I'd say that epigenetics will turn out to be more useful in understanding disease than plain old genetics, which focuses on what genes people have. Having a gene is only the first step. Unless the gene is turned on, it's as meaningless and as silent as a CD you never play. Take the 2004 study that I find so intriguing. In rat pups whose moms smother them with attention, a gene that makes molecules in the brain that serve as landing sites for stress hormones gets turned on. When a mother rat ignores and neglects her pups, rarely licking and grooming them, the gene remains silenced, its status at birth. The more stress-hormone receptors the brain has, the fewer stress hormones the body releases. As a result, the offspring of attentive moms—with the receptor gene on, receptors galore, and stress hormones in check—grow up to be mellow, curious, and laid back, scientists led by Michael Meaney of McGill University reported in Nature Neuroscience. But the pups of neglectful mothers—with the receptor gene off, receptors scarce, and stress hormones flooding their brains—grow up to be skittish, fearful, and neurotic. Yet the mice's genes for this receptor are identical; all that differs is whether it is turned on or off, which in turn is determined by their mothers' behavior.

Epigenetics could also solve the longstanding mystery of why identical twins are different in so many ways. For instance, one study followed an identical twin brother who developed schizophrenia at age 22 while the other has remained healthy. Like all identical twins, they carry the same DNA sequences. But their epigenetic status differed dramatically, scientists led by Arturas Petronis of the Centre for Addiction and Mental Health found. In the healthy brother, a gene linked to schizophrenia had somehow been silenced by methylation. So although he had "a schizophrenia gene," he was spared. In the twin with schizophrenia, the gene was roaring its disease-causing message at full volume.

With any luck, the discovery of how leafy greens and multivitamins keep beneficial genes turned on, reducing the risk of lung cancer, will be only the first of many on how the way we live can reach into our double helix, for good or for ill.