Featured Articles

Epigenetics Is Seen as Possible Key to Cloning

By Brian Vastag, JAMA

28 February 1997: Washington -- Sheep cloner Ian Wilmut, PhD, of the Roslin Institute in Scotland, recently said that for human cloning to succeed, "we need a step as big as the one that produced Dolly" (see accompanying story).

Press Photo

If that step ever comes, it will likely involve the nascent field of epigenetics, literally, "after genetics." With the Human Genome Project nearing completion, scientists at the edge of biology are surveying enigmatic cellular processes that are perhaps as fundamental to life as DNA itself.

Although many of the estimated 40 000 genes have been spelled out to the last A, C, T, or G, researchers have only a shaky understanding of how and why they switch on and off. Thanks to microarray analysis, they know that patterns of gene expression shift like the colors of a kaleidoscope. The spiraling hues depend on where the cell lies in timedevelopmentally or per the cell cycle, and spacewhether in the heart or the brain.

Strides made in the past few years point to proteins that bind and wrap DNA into chromatin, the scrunched-up threads located in cell nuclei. Enzymes and even RNA regularly interact with this complex. These chaperone molecules can tag genes needed by the cell, pry open the appropriate stretch of chromatin, and initiate protein production. They can stitch chromatin shut to silence a gene.

New research suggests that variations in this epigenetic machinery can be inherited, but not via DNA. No one knows how, but in a turn that would drive Mendel out of the pea patch, it seems that another means of inheritance exists. In a recent review, experts propose a system called the "histone code"histones being proteins involved in forming chromatin. According to the authors, this code "represents a fundamental regulatory mechanism that has . . . far reaching consequences for . . . both normal and pathological development" (Science. 2001;293:1074-1080). In other words, it may be just as important as DNA.

A Black Box

Epigenetics and cloning meet in a newly cloned zygote. There, the transferred nucleus of an adult cell undergoes a remarkable reprogramming. The black box of cloning, this cellular prestidigitation transmutes, say, an ordinary skin cell, like the millions that are shed every day, into a potential life.

Through an epigenetic process not yet understood, the egg resets genetic switches inside the donor nucleus, directing its 6 ft of coiled DNA to stop acting like that of a skin cell and start acting like that of an embryo. It's like rolling back the odometer of a used car to zeroand actually getting a new car.

With natural reproduction, this reprogramming transpires over months and years as sperm and egg, respectively, mature. After a sperm wriggles through the zona pellucida, some mysterious apparatus in the egg's cytoplasm completes the reprogramming. The job is crucial. Each stage of fetal development triggers a unique cascade of genes, switching on and off in precise choreography. If the reprogramming goes well, the zygote gracefully dances through development, growing into an embryo, a fetus, a person.

In nuclear transfer cloning, reprogramming is allotted precious little timeminutes to hours. When a micropipette injects a donor nucleus into a waiting egg, the donor's genome retains its old, worn configuration. It knows only how to make other cells of its kind, whether skin, muscle, or something else. Asking the nuclear material to engender an embryo is akin to telling a grandfather to repeat puberty. Unless reprogramming quickly succeeds, the opportunity for disaster is huge (see accompanying story).

Led by Wilmut and a handful of others, cloners are beginning to understand what goes awry in clonal embryonic wrecks. In particular, they're finding problems with so-called imprinted genes, which are exquisitely sensitive to epigenetic reprogramming errors.

These renegades defy biological convention. Their discovery nearly two decades ago shocked biologists. While the vast majority of genes operate with two alleles, one from each parent, imprinted genes function with just one allele. The other is silenced by epigenetic machinery, most likely during early embryo development. To date, some 50 imprinted genes have been found in mice and humans.

With just one working allele, a single mutation can trigger serious problems, as can errors in imprinting itself. Already researchers report strong links to severe congenital diseases such as Prader-Willi, Angelman, and Beckwith-Wiedemann syndromes. The latter is characterized by increased risk for rare cancers (in particular Wilms' tumor), fetal overgrowth, and enlarged tongues and organs.

Those problems sound familiar to animal cloners. Large offspring syndrome (LOS) appears in perhaps 30% to 40% of cloned animals, who manifest overgrowth traits similar to those seen in children with Beckwith-Wiedemann syndrome (Rev Reprod. 1998;3:155-163).

A Singular Path

Work on this imprinting link has taken Duke University Medical Center's Randy Jirtle, PhD, down a path that has resulted in his angering some prominent cloning experts. In August, Jirtle published an article in which he concluded that cloned humans would be unlikely to experience LOS. An accompanying press release claimed that people might be easier to clone than animals. The story attracted widespread media attention, leading to characterizations of Jirtle as pro-cloning. "I'm not advocating that we go out and clone people," Jirtle said in a phone interview, joking that he has two children made "the old-fashioned way."

Jirtle's journey extends back a decade, to early research on imprinted genes. One of the first discovered was a receptor for insulinlike growth factor II (IGF2R). In 1991, Denise Barlow, PhD, then at the Research Institute for Molecular Pathology, Vienna, Austria, reported that IGF2R was imprinted in mice (Nature. 1991;349:84-87). Researchers then went on to make knockout mice lacking IGF2R. Like children with Beckwith-Wiedemann syndrome, the mice displayed traits eerily similar to those now being seen in LOS clones. "They grow about 50% larger; they're structurally abnormal," said Jirtle. "The lungs just don't develop. The heart becomes massive so you have cardiovascular problems." It appeared that insufficient IGF2R led to an excess of available insulinlike growth factor II, which led to excessive growth.

In people, insufficient IGF2R often leads to cancer. A radiation oncologist by training, Jirtle delved into imprinting during his research on IGF2R's role as a tumor suppressor gene. "I had no idea what it was," he said. But as he learned more, he devoted more and more time and laboratory space to the effort. Today, he runs a 40-person laboratory outfitted exclusively for imprinting research.

Conflicting Reports

As Jirtle's interest grew, so did a debate about IGF2R. In mice, there was no question that it was imprinted. But in people, there were conflicting reports, which Jirtle said stemmed from problems with the use of the new and unfamiliar laboratory methods being developed. A few articles reported that IGF2R was imprinted in perhaps 50% of the population. "That was very problematic," said Jirtle. "If a lot of people were imprinted, you would predict them to be exquisitely susceptible to tumor formation."

So he directed his laboratory to help resolve the debate. Using techniques springing from the Human Genome Project, the team found no evidence of imprinting in various tissues from 75 human fetuses and 12 full-term placentas (Hum Mol Genet. 2001;10:1721-1728). Because every other imprinted gene found in the mouse was also imprinted in humans, Jirtle's interest peaked. He wanted to know at what point the species diverged.

After collecting tissue samples from a Noah's Ark of mammals and birds, he and his team concluded that IGF2R imprinting first appeared about 175 million years ago, when live-bearing mammals split from egg layers on the evolutionary tree. Just three of these monotremes exist today, the platypus and two species of spiny anteater. The platypus, at least, does not carry the imprinted gene, a result that helped resolve one of the biggest debates in mammalian evolution (Mamm Genome. 2001;12:513-517).

Drawing on help from Andrew Hoffman, PhD, of Stanford University School of Medicine, and with tissue from the Duke University Primate Center, Jirtle then showed that opossums, pigs, cows, sheep, rats, and mice did carry the imprinted version, but primates did not. From the most primitive prosimian, the colugo, through tree shrews, lemurs, and people, he found no evidence of imprinting. "Mother nature took care of this overgrowth problem 70 million years ago," said Jirtle, when primates split from other mammals.

Why? Jirtle's pet theory is that primate mothers ran into a fetal overgrowth problem. As animals got smarter and brains became larger, pelvises stayed the same size. "If you look at humans," he said, "we're very close to not being able to deliver. Dying during childbirth was common until recently." In early human ancestors, then, fetuses that lost the imprinting, perhaps through some epigenetic mutation, would have had twice as much IGF2R as normal, preventing overgrowth. More of those that kept the imprint would perish during birth. Natural selection would take care of the rest. "There's no way to prove it," said Jirtle with a chuckle. His current research is comparing snippets of platypus and opossum DNA in a quest to find out why opossums, the most primitive live-bearing mammals, carry the imprint, while platypuses do not. He speculates that imprinting allowed live birth to become a viable evolutionary strategy.

He is also convinced that the lack of IGF2R imprinting in people decreases the odds of LOS during cloning. Support comes from an article Wilmut published earlier this year (Nat Genet. 2001;27:153-154). It showed that reduction in IGF2R leads to LOS in sheep fetuses. That reduction stems directly from epigenetic errors and imprinting defects that occur during embryo culturing. Wilmut concluded that his study shows a "causative role for IGF2R loss in sheep."

Troubled Researchers

While Jirtle is confident in his interpretation of the data, his announcement to the media disturbed both Wilmut and Rudolf Jaenisch, PhD, of the Whitehead Institute at the Massachusetts Institute of Technology. Jaenisch said in a phone interview, "He argues that this is a key gene for cloning. That is nonsense, it's not a key gene."

In his many mouse cloning experiments, Jaenisch said he has not seen any problems with IGF2R. Earlier, at the National Academy of Sciences cloning workshop, Jaenisch implored the panel to recognize that it is foolhardy to "focus on gene imprinting. It's part of the problem but not all of it." He said that preliminary evidence from his laboratory showed faulty regulation of nonimprinted genes in clones as well. Jaenisch also disputes the claim that IGF2R has been proven to cause LOS. He wants much more evidence.

Jirtle agreed that there could be other problems that lead to overgrowth. But he thinks the chances are slim. "If you look at the [IGF2R] knockout animals, it's incredible. If you look at the clones, it's incredible. They're absolutely identical" in overgrowth traits, he said.

While debate on scientific points like this will continue, and while research on epigenetic problems in cloning will steam ahead, Jaenisch said he was angered by what he sees as Jirtle's manipulation of the media. He's worried that those who want to clone humanspeople he calls "nuts"will point to Jirtle's research and say that it gives them the green light.

Jirtle's response: "It doesn't mean we're at green, but we're not at red either. We're at amber."