Mitochondrial diseases, which affect as many as one in 4,000 people, can impair muscles, nerves, even entire organ systems, and have no known cure. Now, in a breakthrough study, Oregon researchers replaced defective mitochondrial DNA with that from a healthy donor. The first subjects, four baby monkeys, are pushing the envelope on the ethical debate that surrounds bioengineering.

Mitochondria are often called the cell’s power plants–the tiny organelles are responsible for energy production, and there can be hundreds to thousands of them in a single cell. They also contain their own DNA. Unlike nuclear DNA, which is a unique combination of both parents’ genomes, mitochondrial DNA (or mtDNA) is passed down through the mother, is derived almost exclusively from her egg, and typically remains unchanged from one generation to the next. Mutations in a woman’s mtDNA are inherited by her child, and so far there has been no way to cure these conditions or stop their transmission.

Now, Shoukhrat Mitalipov and his colleagues at Oregon Health & Science University in Beaverton, OR, have found a way to get rid of mutant mtDNA. Using a process similar to cloning, they first harvested a fertile egg. Then, when the egg was undergoing cell division, they removed a set of its chromosomes and inserted them into an egg harvested from another female, one that already had its nucleus removed. In essence, the enucleated egg provided a set of mitochondrial chromosomes, while the transferred nuclear chromosomes provided the main genetic material for development. Other researchers have attempted similar processes, but previous efforts couldn’t prevent mutant mitochondria from tagging along to the new egg.

The researchers avoided this problem by carefully isolating chromosomes during a very specific and segregated process of cell division, in which nuclear DNA is tied up into an elliptical spindle. “Our whole technique comes to efficiently separating the two different types of DNA that [mammals] carry, and to separate them very cleanly,” Mitalipov says. “We believe this can be used to prevent transmission of mutated mitochondrial DNA…[and] correct for mitochondrial DNA mutations in children even before they’re born.”

To date, there are 200 to 250 known disease-causing mutations in mitochondrial DNA, and they occur in as many as one in 4,000 people. The syndromes vary in severity, with symptoms ranging from muscle weakness and loss of motor control to diabetes, liver disease, and developmental delays. Many die before ever reaching adulthood. “The patients carrying these types of mutations don’t have the same options for genetic counseling,” Mitalipov says, since any mutation a woman has will be passed to her egg. “Currently, her only options are using donated eggs or adopting a child.”

“It’s an important study, and it’s the only approach that I can think by which you could render a family free of risk of their offspring developing a mitochondrial DNA disease,” says Douglas Wallace, a mitochondrial DNA researcher at the University of California, Irvine. Because mitochondrial DNA is self-replicating, the technique allows for a way to “swap” healthy versions for mutant ones without genetic alterations.

But therein also lies the rub. Many researchers and ethicists alike balk at the idea of making genetic changes to the germline, ones that fundamentally affect an egg or sperm and will be passed along to the next generation. While swapping out mitochondrial DNA may not qualify as the kind of germline engineering people have in mind when they worry about made-to-order babies–with certain traits like intelligence or eye color specifically engineered–it edges toward that shaky ethical ground.

“The technique that they used, transferring a chromosomal spindle to get new mitochondria to power the egg, seems completely ethical and defensible,” says Arthur Caplan, a bioethicist at the University of Pennsylvania. “But while this technique doesn’t have much use outside of fixing problems in the mitochondria, it does open the door a tiny bit on germline engineering.” Because it’s using a self-contained part of the cell, he notes that it’s not what people typically have in mind when they talk about tinkering with germline genetics. “But by cracking open the door, it puts the principle of never doing germline engineering into dispute.”

David Magnus, who heads Stanford University’s Center for Biomedical Ethics, agrees that most of society’s germline engineering concerns don’t apply in this case. But he does point out that the procedure would lead to, essentially, three parents instead of two, “making legal and social arrangements more complicated,” he says. “What happens if the mitochondrial donor decides, down the road, that she should have some parental rights to the offspring?”

This is, of course, getting way ahead of the science itself. Much more must be done before the procedure is approved. The new technique has only been applied in nine rhesus macaques, three of which became pregnant (one with twins)–a 33 percent success rate that appears to mirror that of regular in vitro fertilization in human patients. And since the seemingly healthy offspring have not yet reached reproductive age, Mitalipov and his colleagues don’t yet know whether the procedure has genetic implications they’ve not yet uncovered. The procedure will also need to be refined, tested in more than nonhuman primates and at other research facilities before human trials can begin. (The Oregon lab is known for very high success rates that other labs can rarely duplicate.)

Researchers in Britain, at the University of Newcastle upon Tyne, have reportedly done something similar in human embryos, but have yet to publish their results and would not comment on Mitalipov’s research.

The Oregon researchers believe they may be ready to apply for clinical trials in two to three years, but much depends on funding and government approval. “This points the way to a technique–it doesn’t provide a therapy,” says UC Irvine’s Wallace, who was the first researcher to discover disease-causing mutations in mitochondrial DNA. “It shows that the concept can work as one approach to treating mitochondrial DNA disease. And that’s an incredible advance, since we have very little to offer these families right now.”