Nature Meets Nurture
Through the exploding field of epigenetics, scientists are pinpointing just how the lifestyle paths we take--and don't take--can change how our genes behave.
Date: May 15, 2010
As Andy Feinberg ’76, director of Johns Hopkins’ Epigenetics Center, takes a visitor through a tour of his lab facilities, he motions down each bay. “So-and-so is working on that project I mentioned to you here, and that new machine I told you about is here,” he points. “And, we abide by all child labor laws,” he says with a grin, gesturing toward a nearby office space with something you don’t see every day in a scientific setting: a red double stroller, surrounded by admirers.
“Would you like to meet the babies?” he asks.
Inside the stroller are two adorable toddlers, the twin sons of Feinberg’s colleague, researcher Karen Reddy (see “Zip Codes in the Gene,” at the bottom of this page). At 16 months old, the twins, though fraternal, live practically identical lives: They developed in the same womb, have grown up in the same house, eat about the same things at the same time, and spend approximately the same time napping. But as they mature, the two will probably embark on different paths, as siblings are wont to do. Perhaps one will become a marathon runner, pushing his body to the limits, while the other becomes an avid video gamer, spending much of his free time glued to the screen. One might opt to be a vegetarian, while the other sticks to a meat and potatoes diet. One might decide to take up smoking, while the other becomes a wine aficionado.
Each of these differences in lifestyle won’t just mark the siblings in a sociological way—their choices may also be marking their genomes, laying on chemical tags that tell some genes to turn on and others to stay off. These effects on gene expression seem to have both long-term and wide-ranging effects on health. And they might explain the correlations that researchers have found between lifestyle and risk of disease.
Since 2007, scientists at Johns Hopkins have been working together to study these marks and other types of “epigenetic” changes—so-called because they act “above the genome” to control gene activity—through the Epigenetics Center, a virtual hub for researchers interested in applying this new field to understand both normal development and diseases including schizophrenia, autism, cancer, and macular degeneration.
“When we’ve been trying to figure out what makes us who we are and the basis of diseases, scientists have long been saying that it can’t just be our genes—and it turns out that it’s not just our genes,” says Reddy. “Epigenetics really is where the rubber meets the road in terms of genes and the environment.”
Almost 30 years ago, the field of epigenetics got its start in cancer—and that’s also where Feinberg got his start in epigenetics. As a young Master of Public Health student at Hopkins who had received his medical degree here years before, Feinberg stopped on a walk through campus one day in 1981 to hear famed Hopkins urological oncologist Donald Coffey give a talk about the different types of cells present in cancerous tumors. The ability of cancers to change into different cell types within a tumor reminded Feinberg of the ability of a maturing slime mold to develop into different cells, something he’d studied earlier as a postdoctoral fellow at the University of California in San Diego.
“I approached Don after the talk, and I told him, ‘There must be a similar mechanism going on with cancer, too,’” says Feinberg. As a for-credit project, Feinberg proposed investigating the hypothesis that cancer initiation involves a genetic change, or mutation, and that cancer promotion or progression involves an epigenetic change that is not inherited and not correlated with mutation.
Playing scientific matchmaker, Coffey soon introduced Feinberg to Bert Vogelstein, the prestigious cancer geneticist at Hopkins who later discovered a single mutation found in almost all solid cancers; he had also been working on another slime mold species. When Feinberg shared what he calls “very theoretical” ideas with Vogelstein, the researcher invited Feinberg to come study at his lab.
For two years, the two scientists worked together to test the genetic and epigenetic hypotheses of cancer. Comparing normal and cancerous tissue, Vogelstein and Feinberg eventually discovered that cancer cells have a smattering of chemical methyl groups attached to their DNA in places that differ from normal cells. The methylation changes affected all tumors examined, even very early premalignant lesions, suggesting that the epigenetic changes occur much earlier than they had guessed.
Eventually, researchers found that where these methyl groups had taken residence, associated genes are often permanently shut off. Some of these affected genes are known tumor suppressors, which keep cell division in check. With these genes effectively turned off, cells are able to divide out of control—a hallmark of cancer.
Since then, the study of how epigenetic mechanisms can affect human health has grown in a big way. For example, it’s well known now that what we eat, drink, and breathe, as well as how we behave, can all affect the methylation patterns on our cells. Abnormal methylation patterns have been linked to a number of diseases, including immune problems, cirrhosis, and muscular dystrophy. Additionally, scientists have identified several other epigenetic influences besides methylation. One of the best studied is the modification of histones, the molecular spools that wind up DNA to keep it organized inside the nucleus. Various chemical groups can stick to these spools, affecting whether nearby DNA is expressed, or used to make proteins.
Sean Taverna, an assistant professor of pharmacology and molecular sciences who studies how methyl and acetyl groups on histones affect associated genes, notes that the promise of epigenetics isn’t just an abstract way of understanding disease—it also offers new hope for therapy. “If you have a mutation in your genome, it can be difficult to correct. We haven’t been very successful with gene therapy. But if you have an epigenetic mutation, you might be able to reset your epigenetic state with a drug,” he says.
For example, he explains, some chemotherapy agents already being used to fight leukemia and other cancers have the capacity to change histone modification states, and may work to fight disease through that mechanism. Additionally, some Hopkins researchers, including Phillip Cole in the Department of Pharmacology and Molecular Sciences and Stephen Baylin in the Departments of Oncology and Medicine, are already developing drugs that change epigenetic marks in other ways to fight cancer.
In 2004, Feinberg, co-director Cynthia Wolberger, and other researchers interested in studying epigenetic phenomena joined together to form the Epigenetics Center. The idea was an integrated lab scheme—researchers in the basic and applied sciences could easily form new collaborations, strategically apply for large grants together, and meet regularly to discuss new ideas. “We have tea every Thursday afternoon—it’s very British,” says Feinberg. “That’s the time for all the young people to get together and talk about what they’re working on.”
Right now, the center encompasses 27 researchers scattered in departments across Hopkins. “It’s been an absolutely spectacular success,” says Feinberg.
One field that’s been particularly effective in harnessing epigenetics is psychiatry, and one researcher at the forefront of this movement is James Potash.
Potash, an associate professor of psychiatry, seemingly couldn’t avoid his destiny to become a psychiatrist here. As he points to a black-and-white framed photo of his father on his office wall, he divulges that he was born at Hopkins just after his father, a Hopkins psychiatrist, completed his residency here. “I guess I was just predisposed,” he says with a smile.
Potash says he became fascinated with the workings of the brain during medical school. Though other parts of the body are readily accessible and their ailments often easy to figure out, the brain remains particularly challenging, with many psychiatric disorders still difficult to explain and treat.
One of the most mysterious psychiatric illnesses is also the most common: clinical depression, which affects tens of millions of people worldwide and has a lifetime risk of 16 percent. Though medications and other treatments, such as electroconvulsive therapy, have had some measure of success in treating many people, depression remains persistently entrenched in others and can stick around episodically for a lifetime, surfacing repeatedly.
Searches for genes linked to depression have been disappointing, Potash explains. Studies looking at inheritance have estimated that the cause of depression is only 40 percent genetic, compared to 70 percent or so for bipolar disorder and schizophrenia. So what’s to blame for depression? A 2009 meta-analysis in the Journal of the American Medical Association confirmed that, unsurprisingly, stressful life events seem to be a big trigger.
The surprise factor, says Potash, is what these life events might be doing to the DNA of cells in the brain. He and his colleagues are currently working on a study that seems to suggest that cortisol, the hormone produced by bodies under stress, may somehow be methylating a gene that has been previously implicated in depression.
To simulate stress, Potash and his colleagues, including Gary Wand, a Hopkins endocrinologist, gave mice injections of corticosterone, the mouse version of cortisol. They then monitored which genes picked up the methyl mark. They found that a gene called Fkbp5 readily picked up methyl groups, which effectively altered the gene’s ability to make its protein.
This difference faded away a week after the scientists stopped giving corticosterone to the mice. However, Potash wonders whether the effect may be cumulative, with more doses of stressful hormone leading to greater persistence for the methyl groups. Speculating further, Potash says that a series of stressful events starting in childhood might be the perfect setup for a lifetime of depression.
Epigenetics isn’t just an abstract way of understanding disease—it also offers new hope for therapy, says Sean Taverna, assistant professor of pharmacology and molecular sciences.
“Freud thought that what happens in childhood shapes mental life forever,” says Potash. “Many of his theories were off the mark, but it would be interesting to see whether a series of stressors starting in early life might lead to methylation changes that last a lifetime.”
Future medicines might eventually be able to change the course even after a lifetime of stressors and resulting depression. Sodium valproate, a drug that has long been used to treat bipolar disorder, is known to change acetyl groups that attach to the histone spools that wind DNA, showing that it’s possible to affect epigenetics with psychiatric medications. “If we know that stress affects epigenetic marks, a major goal would be to figure out how we can reverse that change. Drugs that accomplish this goal could be an important new class of psychiatric medications,” Potash explains.
Dani Fallin, an associate professor at the Bloomberg School of Public Health, is using epigenetics to ferret out the origins of a different psychiatric disease: autism. “I call myself a neuropsychiatric genetic epidemiologist,” she says.
Fallin, whose training lies mainly in epidemiology and statistical genetics, did her first work at Hopkins in Alzheimer’s disease genetics. Now, the question of what causes autism, a disease estimated to affect one of every 110 children born in the U.S., takes up most of her time.
Fallin’s latest project involves painstakingly examining the environments of pregnant mothers to understand which factors might lead to autism in their children. She and her colleagues recently began recruiting pregnant mothers who already have a child with autism, a factor that significantly increases the risk that the new baby will also get the disease.
“If you want to understand how genes cause disease, you really need to understand the epigenetics portion of it—which we know strongly depends on the environment,” Fallin says.
Starting before the mothers reach 20 weeks of gestation, the researchers begin making home visits, surveying for cleaning products and potential toxins, even swiping for dust to measure for contaminants. Periodically, the mothers give blood samples, and at birth and beyond, the babies do as well.
This blood, beside being another tool to use in searching for exposures to health-harming agents such as heavy metals, serves as a vector for looking at epigenetic effects over time. Using a method that surveys practically the whole genome for methylation, Fallin and her colleagues will test the blood cells for changes in their methylation marks, looking for temporal correlations between exposures and when marks appear. Eventually, if any of the children develop autism, these correlations might offer clues to which environmental agents might be causative.
Cancer and psychiatric diseases aren’t the only targets for epigenetic research. Shannath Merbs, an associate professor of ophthalmology at the Wilmer Eye Institute, is taking epigenetics into uncharted territory: the eye. She and her colleagues are studying DNA methylation in genes in the retina and optic nerve, looking for clues to understand the development of diseases including macular degeneration and glaucoma.
“No one is doing anything like this at all,” says Merbs.
Merbs explains that while researchers have identified much of the genetic risk for age-related macular degeneration, the leading cause of vision loss in Americans age 60 and older, these genes are associated with increased risk but don’t seem to cause the disease directly. In the case of glaucoma, very little of the genetic risk has been identified. However, for both diseases, environmental factors appear to play heavily into risk. For example, smokers have as much as a threefold increase in risk of developing macular degeneration compared to nonsmokers.
To understand how these environmental factors might affect the risk of glaucoma and macular degeneration, Merbs and her colleagues are conducting a large comparative study looking at tissue from eyes of people who had either disease compared with healthy eyes. The eyes themselves come from autopsies, as well as a national repository for human tissues, which also supplies health and epidemiological information about the eye donors, including environmental factors that may have contributed to their diseases. Merbs says she can make specific requests to get exactly what she wants.
“I can say, ‘I want eyes with glaucoma, removed from the body within six hours of death, and shipped to me in 24 hours.’ Then, when eyes like that come in to the repository, we’re on call,” she explains. Her unusual research subjects arrive in a rather mundane fashion: packed in ice in a Fed Ex box, marked “Human Eyes” on the side.
Merbs’ team removes the macula and optic nerve, parts affected by the two diseases, and isolates DNA from these tissues. They then run this genetic material through a genomewide evaluation of its methylation.
The scientists hope to eventually identify genes affected by methylation in each disease. Ultimately, researchers could identify ways to revert the eyes of glaucoma or macular degeneration patients back to normal by altering these methylation patterns. Merbs’ and her colleagues’ efforts may help with prevention as well. If they can discover how environmental factors, such as smoking, directly affect methylation patterns, doctors may eventually have even more wherewithal to steer patients toward healthier habits.
From Vogelstein’s lab to now, Feinberg is still making strides in cancer epigenetics—he and his colleagues published a paper last year showing that methylation in cancer cells seems to mimic the normal methylation patterns in healthy cells of other types.
This new finding could lead to a new understanding of why cancer cells behave so differently from their normal brethren, says Feinberg’s former graduate student Christine Ladd-Acosta, now a postdoc in Dani Fallin’s lab. Ladd-Acosta explains that a colon cancer cell whose epigenetic marks look more like healthy liver cells might be inclined to migrate to where it’s surrounded by cells that share similar epigenetic marks. Indeed, one of the most common sites of metastasis for colon cancer is the liver.
“Cancer might go to where it feels more at home,” she explains. “If that’s the case, it could change our current thoughts on why primary cancer cells migrate and take up residence in specific normal tissues and not other tissues instead.”
Feinberg and his colleagues are now following up on this finding. However, his lab has also branched out to explore the role of epigenetics in a variety of other diseases. Carolina Montano, one of Feinberg’s graduate students, is currently leading a study to explore the role that a famine between 1959 and 1961 might have played in driving up the incidence of schizophrenia in China. Epidemiologists have long observed that the rate of schizophrenia is double among people born to mothers pregnant during the famine.
Montano is collaborating on the project with Mary-Claire King, a researcher at the University of Washington in Seattle who has collected blood samples from 35 sibling pairs: one exposed to the famine during his or her mother’s pregnancy, and one born before or after the famine struck. King sends the samples to Montano, who runs them through a comprehensive scan of methylation patterns across the genome, looking for differences between the two siblings.
“Once we have that information, we can start asking which genes are affected. Are they important for neurodevelopment? Are there unexpected pathways that might lead to this disease? This information can help us understand what part of prenatal development might predispose people to this disease,” Montano says.
She explains that the current theory is that vitamins such as B12 and folate, which provide the raw materials for methylation, were depleted from pregnant women’s bodies during the famine. Consequently, a state of starvation may have led to abnormal methylation patterns that set up their babies for mental illness.
Such information may eventually help Reddy’s twins’ generation as they plan their own families. Understanding prenatal nutrition in the context of epigenetics might eventually cut the risk of schizophrenia and a multitude of other diseases.
“Once we’re up to speed on epigenetic mechanisms,” says Montano, “we’ll have so much more power to modify our health.”
Zip Codes in the Gene
Not every epigenetic effect relies on chemical marks. Karen Reddy, an assistant professor in the Department of Biological Chemistry, is studying how a gene’s real estate, the position it occupies within the nucleus, affects which genes are on or off.
“If you look at individual chromosomes, they’re not floating around in the nucleus like a big bowl of spaghetti—they occupy distinct territories. It suggests a level of organization in the nucleus,” explains Reddy.
While she was a postdoctoral fellow at the University of Chicago, Reddy and her colleagues found that genes filed away at the edges of the nucleus, near its membrane, were typically inactive. To understand whether the membrane plays a role in keeping these genes turned off, she designed an experiment where an active gene was placed in the middle of the nucleus, then dragged to the membrane. Sure enough, as the gene neared the membrane, its activity drastically decreased.
Her recent work suggests that genes might be tagged with molecular zip codes that send them to various parts of the nucleus, determining by position whether they’re turned on or off.
“We don’t know what the zip codes are or who the postman might be,” she says. “We’re investigating these questions now.”