Epigenetics: It's All in the Packaging

Roll over, Mendel. Watson and Crick? They are so your old man's version of DNA. And that big multibillion-dollar hullabaloo called the Human Genome Project? To some scientists, it's beginning to look like an expensive genetic floor pad for a much more intricate—and dynamic—tapestry of life that lies on top of it.

There's a revolution sweeping biology today—begrudged by a few, but accepted by more and more biologists—that is changing scientific thinking about the way genes work, the way diseases arise and the way some of the most dreadful among them, including cancer, might be diagnosed and treated. This revolution is called epigenetics, and it is not only beginning to explain some of the biological mysteries that deepened with the Human Genome Project. Because of a series of accidental events, it is already prolonging the lives of human patients with deadly diseases.

Over the past several years, and largely without much public notice, physicians have reported success using epigenetic therapies against cancers of the blood and have even made progress against intractable solid-tumor malignancies like lung cancer. The story is still preliminary and unfolding (dozens of clinical trials using epigenetic drugs are currently underway), but Dr. Margaret Foti, chief executive officer of the American Association for Cancer Research, recently noted that epigenetics is already resulting in "significant improvements" in cancer diagnosis and therapy. "It's really coming into its own now," she said. Leaping on the bandwagon, the National Institutes of Health made epigenetics the focus of one of its cutting-edge "Roadmap" initiatives announced last fall.

"I think we were all brought up to think the genome was it," says C. David Allis, a scientist at Rockefeller University whose research in the 1990s helped catalyze the current interest in epigenetics. "But even when the genome was a done deal, some people thought, 'Is that the whole story?' It's really been a watershed in understanding that there is something beyond the genome."

The emergence of epigenetics represents a fundamental rethinking of how molecular biology works. Scientists have learned that while DNA remains the basic text of life, the script is often controlled by stage directions embedded in a layer of biochemicals that, roughly speaking, sit on top of the DNA. These modifications, called epimutations, can turn genes on and off, often at inappropriate times. In other words, epigenetics has introduced the startling idea that it's not just the book of life (in the form of DNA) that's important, but how the book is packaged.

At one level, this higher order of control makes perfect sense. Biologists have long known that developing organisms—humans included—need a full complement of genes at the moment of fertilization, but that many genes subsequently get turned on and off as the embryo develops. In humans, this is a lifelong process. There are genes for a fetal version of hemoglobin, for example, and then an adult version that kicks in after birth; through epigenetic control, the fetal genes are permanently turned off at a certain stage of development, and the adult genes are permanently activated. As each one of us developed from a fertilized egg, stem cells in the early embryo matured into brain cells, liver cells and indeed several hundred specialized cells and tissues; at each step of that maturation process, our DNA was modified. When we entered puberty, quiescent genes were suddenly activated. And as we age, the dings of earlier life experiences seem to shape the activity of our DNA. Many if not most of those changes are epigenetic in nature, where the DNA itself remains unchanged, but the packaging has been dramatically perturbed; animal experiments suggest that environmental factors, from childhood diet and maternal care to stress, can play epigenetic havoc with our basic DNA hardware.

The interest in epigenetics has assumed critical mass in the past 10 years for several reasons. The Human Genome Project, often touted as "biology's moonshot," provided the basic text of life, in the form of the complete human sequence of DNA, but scientists have had a hard time linking specific genetic causes to many common illnesses. The role of "misspelled" DNA (in the form of both classic mutations and genetic variation, first teased out in the 19th century by the monk Gregor Mendel) has turned out to explain, in the words of a recent New England Journal of Medicine commentator, "only a small fraction of disease." "We were all raised on the Watson and Crick concept of DNA-driven inheritance," Allis says. "It turns out that epigenetics may be even more responsible for gene expression and disease than DNA alone, especially in more advanced multicellular organisms." In the 1990s, meanwhile, scientists like Allis reported basic but breathtaking discoveries that showed how several groups of enzymes, common to every cell, could create epimutations without ever changing the DNA script.

Basic research has shown that enzymes can tamper with genetic information in at least two distinct ways. In some cases, the on-off switch of a gene can be smothered when an enzyme attaches chemicals to the DNA; known as DNA methy-lation, this process essentially silences a gene that should be on. In other cases, a separate class of enzyme improperly disrupts the normal cellular packaging of DNA. Typically, the gossamer thread of DNA is wound around a spool of protein called histone; when this second class of enzymes strips away part of the packaging, the DNA becomes so tightly wound up that it can't loosen up enough to be read by the cell. In effect, the slip jacket for specific genes is so tight that it's impossible to crack open the spine and get a glimpse of the genetic text. Conversely, sometimes genes that should remain permanently interred in a tomb of histone suddenly come back to life, like some cellular version of Night of the Living Dead.

In the past five years, the evidence has become "absolutely rock solid" to cancer researchers that epigenetic changes play a fundamental role in cancer, according to Robert A. Weinberg, an elder statesman of cancer biology at the Whitehead Institute in Cambridge, Mass. DNA methylation, he adds, "may ultimately be far more important than gene mutation in shutting down tumor suppressor genes," one of the cell's main mechanisms to short-circuit an incipient cancer.

Each epigenetic change seems to leave a chemical flag, or "mark," on the DNA, and hence researchers are intensely cataloging these marks into "epigenomes" as a possible clue to diagnosis, prognosis and perhaps even prevention of disease. Unlike genetic markers, which reveal small "typographic" variations in the spelling of genes, epigenetic markers indicate places where entire genes have been silenced or activated. Paula Vertino of the Emory University School of Medicine, for example, has identified patches of DNA that seem especially prone to be inappropriately silenced or activated in breast and lung cancer; researchers at Johns Hopkins have used epigenetic markers in brain-cancer cells to predict which patients are likelier to benefit from chemotherapy. Recent laboratory findings suggest that deciphering the layers of genetic control modifying DNA has implications not just for cancer, but also for chronic diseases associated with aging, like heart disease and diabetes; for mental disorders like autism and depression; for stem-cell biology; and even for our notions of what constitutes an inherited disease. Everything is up for grabs.

"There's only one genome," says Wolf Reik, professor of epigenetics at the University of Cambridge in England, "but hundreds of epigenomes." And unlike string theory in physics, for example, epigenetics is neither an exotic intellectual idea nor a theory awaiting verificationby future data. The biology is real, and the practical effects have already reached the bedside.

In the 1990s, Stephen Baylin of Johns Hopkins University led the effort showing that epigenetic changes in DNA were associated with cancer; in fact, disruptions in tumor suppressor genes, which normally protect cells against cancer, are more often due to epigenetic silencing than outright mutation. In May, Baylin and Peter Jones of the University of Southern California received a three-year, $9.1 million grant to launch accelerated testing of epigenetic therapy in patients with lung, colon and breast cancer, with interim results promised within a year. The Hopkins group has presented preliminary results at recent meetings showing that a combination of two epigenetic drugs produced several responses (including one complete remission) in patients with advanced lung cancer. "The trials are still ongoing, and we don't know what percentage of patients will respond, if it will be 10 or 20 percent," says Baylin. "But we have had very robust responses, of both primary tumors and metastases, in non-small-cell lung cancer." "That's just extraordinary," says Foti of AACR, noting the poor prognosis for patients with these advanced cancers.

If the amount of clinical testing seems surprising, it's probably because the medical part of the epigenetics story is unfolding in reverse: doctors had the drugs long before they had a theory suggesting how to use them properly. Indeed, several of the drugs now being tested against cancer have been around for decades, but in the past were used in the wrong way for the wrong reason. Azacitidine, for example, was first discovered in Czechoslovakia in the 1960s as a traditional chemotherapy drug, and doctors used it to kill cancer cells the old-fashioned way: giving as much as patients could tolerate. Jones, a South African by birth who now heads the Norris Comprehensive Cancer Center at USC, discovered in the 1980s that the drug had another mode of action: it could turn genes back on by stripping away the "duct tape" of DNA methylation that muffled genes. This suggested a different kind of attack on cancer—not by killing cancer cells outright, but by reversing the epigenetic changes that make a cell cancerous in the first place.

In the 1980s, as a young oncology fellow at Mount Sinai School of Medicine in New York, Lewis Silverman proposed testing azacitidine as an epigenetic drug—that is, at lower doses than is typical for traditional chemotherapy, where it still might be effective reversing silenced genes. Silverman has since shown that low doses of the drug reduce the symptoms of a type of leukemia and allows patients to live longer. The Food and Drug Administration approved azacitidine in May 2004; the drug is now marketed as Vidaza.

A different class of epigenetic drug has emerged from work at Harvard, Columbia and Memorial Sloan-Kettering Cancer Center in New York. In addition to the silencing effect of methylation, genes can be turned on and off by enzymes that tighten or loosen the packaging of DNA. Paul Marks and Ronald Breslow at Columbia created a small molecule, called vorinostat, that blocks the action of the enzymes that tamper with DNA's packaging, thus turning inactivated genes back on. That drug was approved by the FDA in 2006 for a rare form of lymphoma and is now being tested against a number of other cancers; Merck markets the drug as Zolinza. Part of the current clinical excitement is that there are already hints that combinations of these and second-generation drugs may be more effective at reversing the epigenetic changes in cancer cells.

Researchers remain guarded in their optimism. Issa concedes that the first-generation epigenetic drugs have not included a home run like Gleevec, the molecular treatment for chronic myeloid leukemia that produces dramatic and lasting remissions. And it is not unusual for deleterious side effects to become more apparent as drugs are used more widely—a particular concern in the case of drugs that have the potential to modify gene expression broadly in normal cells. But people who have witnessed the explosion of promising results in the past year have difficulty suppressing their excitement. "The promise is staggering," says Allis.

The stakes in epigenetics go well beyond clinical therapies, however. There have been hints from laboratory experiments and epidemiological studies that epigenetic changes in one generation—caused, for example, by smoking or diet—can be passed on to children and even grandchildren. Reik, who is also associate director of the Babraham Institute in Cambridge, is investigating how the overlay of epigenetic changes is erased from DNA when mice make their germ cells—how all the epigenetic changes, like some microscopic version of duct tape, get stripped off the DNA that goes into the sperm in males and eggs in females. "People are now beginning to realize that there are probably things that don't get wiped out or erased in the germ cells," he says, "so these are so-called epimutations that can be passed on from parents to children and to grandchildren—not genetic changes passed on, like Mendel, but an epimutation.

"We don't know how common this might be," Reik adds, choosing his words carefully, "but it's potentially quite revolutionary. It's not only challenging Mendel, but potentially challenging even Darwin. We are very careful when we talk about these things."