This article is from The Checkup, MIT Technology Review's weekly biotech newsletter. To receive it in your inbox every Thursday, sign up here.

This week, I’ve been working on a big story about a controversial treatment that creates babies with three genetic parents. The “three-parent baby” technique was thought to help parents avoid passing diseases on to their kids. But new evidence suggests it doesn’t always work—and could create babies at risk of severe diseases.

The evidence comes from two babies born after the procedure was used to help couples with a different problem: infertility. It’s lucky we found the problem in these cases—these babies didn’t have parents with disease-causing mutations, so they should be fine.

And there’s another silver lining to the results. They add to growing evidence that the “three-parent” technique might help treat infertility and shed light on why some people struggle to conceive.

For years, scientists have scoffed at the idea of using this technology for infertility. But now they are changing their minds. Let’s take a look at why.

First, a recap. The “three-parent” technology is so called because it uses genes from three people to create an embryo. Almost all of the DNA in our cells resides in the nucleus, but we have a miniature second genome—a string of 37 genes housed in our mitochondria.

Mitochondria are tiny organelles that supply our cells with energy. They float around in the cytoplasm, the fluid that surrounds the nucleus. Mitochondrial DNA (mtDNA) is only passed through the maternal line—all of your mtDNA comes from your genetic mother.

Sometimes these genes can carry mutations that cause diseases. Mitochondrial diseases, although rare, can affect multiple organs, and they can be severe. Some are fatal. People who carry mtDNA mutations in their eggs risk passing along disease to their children. Some of these children don’t survive long after birth.

In an attempt to avoid this, scientists developed mitochondrial replacement therapy (MRT). The idea is to create an embryo where the DNA comes from the nucleus of one would-be parent’s egg and the sperm of another, but the mtDNA comes from a donor. There are a few ways of doing this, but they all involve putting the parents’ nuclear DNA into the cytoplasm of a donor’s egg, which may or may not be fertilized. The result is an embryo with DNA from three people.

In 2016, I reported the birth of the first baby created using one of these approaches: it involved transferring the DNA of a woman’s nucleus into the egg of a donor, which had its own nucleus removed. The baby, a little boy, was born to a woman who carried mitochondrial genes for a disease called Leigh syndrome. Her first two children had died from the disease. But the boy was born healthy.

Since then, other clinics have started offering the treatment. A center in Newcastle in the UK is the only one in the world with regulatory approval to offer MRT to couples with mitochondrial diseases. The team launched a trial in 2017, but it hasn’t yet breathed a word of any results.

Mitochondria mess

Meanwhile, some scientists believe that mitochondria might play a role in infertility, which affects around 10% of people in the US, often without any clear explanation. After all, these organelles provide energy to cells. If they aren’t working, cells might not have enough energy to divide properly. What’s more, mitochondria in the eggs of women over 40 can look swollen and abnormal. Some have wondered if that might contribute to age-related infertility.

In 2014, a company called OvaScience began marketing a new technology that was designed to capitalize on this idea. The company developed a form of IVF that involved using mitochondria from a different source to power older women’s eggs. But in this case, the donated mitochondria came from the women’s own stem cells—cells that are thought to be “young.”

The company claimed that the treatment, called Augment, helped an infertile couple conceive a baby boy, who was born in 2015. A couple of very small trials suggested that it might work for others. But IVF is notoriously unpredictable, as is pregnancy itself. I’ve heard plenty of stories about people who couldn’t get pregnant for years, had multiple failed rounds of IVF, and then had an accidental pregnancy in their 40s. And when Augment was subjected to larger, controlled studies, it was found not to work.

This whole mess is one of the reasons why many scientists didn’t believe MRT would work for infertility. In 2020, one professor of obstetrics and gynecology described the doctors using MRT for infertility as “complicit … in providing unproven [fertility] technologies to desperate parents willing to pay and try.”

But the tide appears to be turning. A newly published study suggests the use of MRT for mitochondrial disease might carry significant risks. But its results are promising when it comes to infertility. Now, some scientists are changing their minds about how and when MRT should be used.

In the study, a team used MRT to treat 25 cisgender heterosexual couples who had been diagnosed with infertility. In all cases, the woman’s eggs seemed to be the problem. Between them, the women had previously undergone 159 treatments to stimulate the production of eggs that could be collected for IVF. They’d each been through an average of six IVF cycles. Despite all that, none of them had ever gotten pregnant.

Success stories

But MRT seems to have worked for them. The team was able to collect 112 eggs from the women, and used cytoplasm from another 112 donated eggs. These were fertilized and generated a good number of embryos—the same as you’d expect from people who don’t have fertility problems, says Dagan Wells, a reproductive biologist at the University of Oxford and a member of the team.

A total of 19 embryos were transferred into 16 women. Seven of them got pregnant. And while one miscarried, the other six had healthy babies. For women who have struggled to conceive for years, it’s a significant result. “[MRT] really seems to have corrected any underlying problem there was,” says Wells.

This trial represents some of the first evidence that MRT could actually work for infertility—but maybe not in the way we once thought it might. While we call the technique “mitochondrial replacement therapy,” embryologists are really swapping the entire cytoplasm of an egg, which contains much more than just mitochondria. There are thousands of proteins floating around in there, for a start. We just don’t yet know how they might influence fertility.

The trial also found something somewhat surprising. All of the embryos that were transferred into the volunteers had mitochondria from a donor. Less than 1% of the mitochondrial DNA was from the mother. By the time they were born, five of the babies still had very low levels of mtDNA from their mothers.

But in one baby, the levels had changed dramatically. At birth, only around half of the child’s mtDNA came from the donor. The other half came from its mother. This phenomenon, called reversion, has also been seen in another child born using MRT in a clinic in Ukraine.

For people who don’t carry genes for mitochondrial diseases, this isn’t a problem. But if the same thing happens in a couple using MRT to avoid such a disease, they could end up with a severely ill baby. Heidi Mertes, a medical ethicist at Ghent University in Belgium, says she is “relieved that this trial was not in patients with mitochondrial disorders.” Me too.

“These patients were deliberately chosen such that they wouldn’t have a risk of mitochondrial disease,” says Wells. “We considered that it was likely to be a safer approach.”

Other scientists now agree that it is probably better to explore MRT in people with infertility before using it to avoid mitochondrial diseases—at least until we understand what’s going on, and can maybe figure out how to avoid any potentially dangerous cases of reversion.

Eight years ago, Björn Heindryckx of Ghent University was one of many influential scientists arguing that MRT should not be used for infertility and should only be used for mitochondrial disease. “But our insight into the technology has changed a little bit,” he says. He now believes the opposite: that MRT should be explored for infertility before it is used for mitochondrial disease.

We can’t draw any firm conclusions about MRT for infertility from the trial conducted by Wells and his colleagues. For a start, it was quite small. And, importantly, there was no control group. We’d need to directly compare the MRT results with those achieved using standard IVF in a similar group of people.

Shoukhrat Mitalipov, an embryo biologist at Oregon Health & Science University, who is collaborating with Wells, plans to run a larger trial in 400 volunteers to get a better idea of how well MRT might treat infertility, if at all.

The takeaway is a bit of a mixed bag. It’s worrying that MRT might not prevent mitochondrial diseases and could create babies at risk of severe illness. But if MRT trials in people struggling to conceive can tell us more about how infertility works and how to treat it, it still has a lot of potential.

Read more from Tech Review's archive

You can read more about the MRT trial, and the two cases of reversion, in this piece, which was published on Thursday.

Karen Weintraub has covered the rise and fall of OvaScience’s Augment technique. Both of these pieces were published in the same month, which gives you some idea of how quickly this field moves.

MRT is also being explored as a way to help trans men use their eggs to have babies. One early study suggests the approach might help generate more healthy embryos from their eggs, as I reported last year.

Babies born from MRT technically have three genetic parents. There are other technologies on the horizon that could allow us to create babies with four genetic parents, or none at all. I explored what this means for our understanding of parenthood in a previous edition of The Checkup.

While fertility clinics are trying to find ways to create healthy embryos to be used in IVF, a biotech company is finding ways to generate synthetic embryos for research, as my colleague Antonio Regalado reported in August. The embryos are being grown in “mechanical wombs,” in case you were wondering.

From around the web

Did the coronavirus that triggered a deadly pandemic leak from a lab? The theory lives on, despite being repeatedly contested by scientists. And US federal agencies can’t agree on where they stand either. (The Atlantic)

Eli Lilly, one of the largest manufacturers of insulin, has finally bowed to public pressure and reduced the cost of this drug. The price is being lowered from $82 to $25 a vial. The move has been described as “long overdue.” (STAT)

Premature births fell during some covid lockdowns, possibly because pregnant people were exposed to less air pollution and fewer viruses. The author of this piece hints that some people had unusually calm pregnancies, but that won’t have been the experience of frontline workers (and certainly wasn’t what I experienced during my own lockdown pregnancy!). (The New York Times)

We’re hearing about more outbreaks of treatment-resistant infections. The US Centers for Disease Control and Prevention is warning of “extensively drug-resistant” eye infections linked to artificial tears, and stomach infections that can spread between people. (CDC) 

Can AI treat mental illness? Welcome to the world of algorithmic psychiatry. (The New Yorker)