gene editing
© MS TECH Peggy Peterson
The unpublished research paper by He Jiankui about the creation of the babies shows proof of attempted gene editing gone awry.

It's been a year since He Jiankui announced that he'd made the world's first gene-edited human babies, twin girls with the pseudonyms Lulu and Nana. Widespread condemnation of his actions followed the announcement. But the facts of the case remain unclear, because he has not been transparent about his work.

In his single public appearance following his announcement, at the Second International Summit on Human Genome Editing in Hong Kong in November 2018, He presented his work by racing through about 60 slides in just 20 minutes. Although he showed data about what he had done to the twins' genes, it was blink-and-you'll-miss-it, and not enough to convince anyone of his claim that he'd safely edited the genomes of the human IVF embryos that became Lulu and Nana.

At the summit, He did say he'd just submitted a manuscript describing this work to a scientific journal. Twelve months later, however, the manuscript has remained unpublished and its contents mysterious.

He was asked at the summit why he hadn't posted his manuscript to a preprint server such as bioRxiv or on a public website — something scientists frequently do to invite feedback on early drafts. He claimed that he'd intended to do so, but colleagues had advised him to allow the manuscript to go through peer review by other scientists before posting it. (Normally, formal peer review takes place only when an academic journal is considering publishing a paper.)

By deciding not to release his manuscript right away, He has made it difficult for other scientists to figure out exactly what he did and how he did it. We already know that there were profound ethical problems with He's work in germline gene editing, which refers to genetic alterations to embryos — or to egg or sperm cells — that can be passed down through the generations. But its scientific merit, and especially its safety, have remained in question.

When I first had the opportunity to look through a complete manuscript from He last November, I immediately realized there were problems.

The most serious was rampant "mosaicism." This means that the gene edits He made to the embryos didn't take effect uniformly: different cells showed different changes. Evidence of mosaicism is present in both Lulu's and Nana's embryos, as well as in Lulu's placenta, making it likely the twins themselves are mosaic. Some parts of their bodies may contain the specific edits He said he made, other parts may contain other edits he didn't highlight, and yet other parts may contain no edits at all. This would mean that the purported benefit of He's editing — HIV resistance — may not extend to the twins' entire bodies, and they could still be fully vulnerable to HIV.

When judging whether the embryos had edits, He took a few cells from the 200 to 300 present in an IVF embryo and analyzed their DNA. But it was the remaining cells went on to multiply to make up the full body. So it is possible that some parts of the twins' bodies have edits that He didn't intend to make ("off-target" edits) and never had a chance to see. Such off-target edits could cause problems such as cancer and heart disease, and could be passed on to Lulu's and Nana's future children.

He apparently didn't realize that his own data revealed extensive mosaicism in the embryos, since he made no note of it in the manuscript I saw. Some have wondered if the CRISPR twins were actually a hoax, but to me, the flaws evident in the data make it clear that they weren't. Rather, He's work was a graphic demonstration of attempted gene editing gone awry. Two living human beings, and potentially their descendants too, will bear the consequences.

You shouldn't have to take my word for any of this. You should be able to judge for yourself, or at least hear what other scientists have to say about it.

However, it seems increasingly unlikely that He will be publishing in a peer-reviewed journal. For one thing, I doubt that any respectable journal would seriously consider publishing research with such ethical problems. And even if one did, and sent the manuscript for peer review, He would be in no position to respond to any technical criticisms with further experimental work. He has been under house arrest, and his laboratory was shut down shortly after his announcement about the twins last year.

The only reason to continue keeping He's work under wraps would be to preserve his ability to publish it someday in a peer-reviewed journal and earn the imprimatur of scientific quality. The community is under no obligation to grant him this privilege. Indeed, it owes him no professional courtesy at all, any more than it would have owed such courtesy to the doctors responsible for the medical experiments in Nazi Germany or the American scientists in charge of the Tuskegee syphilis study.

Rather, in light of the egregious scientific and ethical lapses inherent in He's cavalier and secretive efforts to make the world's first gene-edited babies, it is he who owes all of us a full accounting of his actions. Since he has shirked his responsibility to make his work public, it's up to others to step in.

Why must the information be public? It's because He's work reveals serious, unresolved safety concerns. It's not clear that any effort to directly edit human embryos, even if done ethically and with full social approval, can reliably avoid these problems.

International committees convened by the World Health Organization, the US National Academies of Medicine and Sciences, and the Royal Society are currently working to propose regulatory frameworks for doing clinical germline gene editing safely, if it is to be done at all. How can the committees properly do their work without fully understanding all the scientific problems with the single real-world application of clinical germline gene editing that's been attempted to date?

Most worrying is that scientists like Denis Rebrikov in Russia aspire to follow in He's footsteps. Rebrikov has said he'll be able to edit the human germline safely. But how can Rebrikov credibly claim to be able to do better than He if the nature of the problems with He's work aren't widely known? How can the Russian authorities properly evaluate the safety of his proposals without being able to refer to He's work for guidance?

It's time for the scientific community to fully understand what happened with Lulu and Nana, and to avoid stumbling down a path toward further ill-starred experiments with clinical germline gene editing.

About the author

Kiran Musunuru is an associate professor of cardiovascular medicine and genetics at the Perelman School of Medicine at the University of Pennsylvania and the author of The CRISPR Generation, a book about the history of gene editing and the Chinese twins.