It was logical that JK would select Nature as the prize venue for his CRISPR babies report. The journal had published the double helix of course, but also the seminal IVF papers of JK’s idol Robert Edwards, as well as the first (and only) two human embryo editing studies conducted outside China. For many scientists, a Nature paper is the ultimate validation, the Oscar of scientific achievement. (Chinese scientists are also handsomely rewarded with a cash incentive for publishing in top-tier journals.)
JK uploaded his manuscript in November 2018, about a week before the Hong Kong summit. He emailed Craig Mello on November 22 to say that the paper had been sent out for peer review. Although Nature staunchly preserves the confidentiality of the review process, extracts of the manuscript were eventually published by MIT Technology Review and by Kiran Musunuru, one of the experts originally contacted by the AP to vet JK’s claims.1 JK naïvely assumed his manuscript would be published expeditiously to worldwide acclaim. Months later, Fyodor Urnov trashed the manuscript as nothing more than “sets of pieces of paper.”2
Dissection of JK’s unpublished manuscript—“Birth of twins after genome editing for HIV resistance”—reveals a naïveté and arrogance on the part of the authors, as if assuming that the Nature editors would fall over themselves to publish this medical milestone. There were ten coauthors, JK positioned at the end of the list, signifying his senior status. The penultimate author was Michael Deem—indicating a major role in driving the work—although his lawyers later insisted that Deem had withdrawn his authorship name from the list of authors.
From the article’s opening summary, JK and his coauthors failed to accurately represent their work and reveal their true intentions. In reporting “the first birth from human gene editing,” JK wrote that his team had used CRISPR to “reproduce a prevalent genetic variant of the CCR5 gene,” when in fact the editing had failed categorically to reproduce the Δ32 deletion. JK asserted that his team had witnessed no off-target or cancer-causing mutations. And his revolutionary therapy would not only “control the HIV epidemic” but also “bring new hope to millions of families” desperate to avoid inheriting a genetic disease. Urnov could barely hide his disgust: “The profundity of the delusion and the hubris is overwhelming,” he said.3
JK reiterated his goal of trying to combat the spread of HIV worldwide, without addressing the overwhelming success of HIV antiviral drugs or the impracticality of addressing the epidemic via editing one embryo at a time. Nor did he dwell on the fact that during the IVF protocol, the father’s sperm were “washed thoroughly to remove infectious seminal fluid.” This was standard protocol, ensuring that any babies carried to term would not contract HIV.
From four healthy embryos, JK’s team analyzed the two that had been edited at the CCR5 gene. They reported just one off-target edit, unlikely to affect the activity of any genes. But brushed aside was the Catch-22 of embryo editing: the only cells tested from the blastocyst would, by definition, play no further role in the development of the embryo. The remaining cells of the embryo were not analyzed, making it impossible to measure what off-target effects might exist in those cells.
In the manuscript, JK said the babies were born in November 2018, even though reporting says the babies were born in late October. Perhaps this was a deliberate feint designed to throw off any amateur detectives trying to identify the CRISPR babies. JK wrote that the twins would undergo periodic medical testing at least until they were eighteen, including tests to ensure they were immune to HIV infection. Such experiments should have been conducted before Lulu and Nana became human guinea pigs. “This egregious violation of elementary norms of ethics and of research borders on the criminal,” Urnov fumed.
In his book The CRISPR Generation, Kiran Musunuru described the visceral emotion of being one of the first people outside JK’s group to read the manuscript, which he received from Marchione on the Monday after Thanksgiving. His dismay grew until he opened a figure that showed the DNA sequences of CCR5 obtained from the CRISPR-treated embryos. The first image (Lulu) should have shown two distinct traces—the normal sequence of CCR5 on one chromosome, a fifteen-base deletion on the other. This figure showed three. Musunuru knew that meant the embryo was mosaic, before emitting a guttural scream. It was a similar story with Nana’s embryo—three traces rather than the two edits JK had claimed.
The phrase that stuck in Musunuru’s mind was “hack job”—not an unusual occurrence among inexperienced researchers learning to use CRISPR, but unconscionable in a human embryo. Mosaicism happens naturally, the result of a random mutation occurring in one cell in the early embryo being passed down as that cell repeatedly divides. In principle, if applied early enough to a single-cell embryo, the CRISPR edits should be transmitted to all daughter cells. But Cas9 takes time to find the correct target sequence, so there is a possibility that the fertilized egg will have undergone cell division before the edits have taken place. If the CRISPR-Cas components are not shared evenly as the cell divides, genes will not be edited uniformly in the developing embryo.
Musunuru’s despair turned to anger as he saw signs of mosaicism at the off-target location in Lulu’s embryo, indicating that “both embryos were flawed.” The only silver lining: the mosaicism showed that Lulu and Nana were not a hoax. No scientist trying to invent such a sensational story would present such a mediocre editing job. He summed up: “The knowledge that somebody had gone ahead and made the first gene-edited babies—for better or for worse, a historic event for humankind—would have been distressing enough, even if it had been done perfectly. The fact that it had been done with flawed embryos, in such a careless fashion, made it a hundred times worse.”
Within a week, Nature closed the file as the scale of the scientific and ethical transgressions exploded around the world. Nature does not comment on the fate of specific articles but the guidelines were laid out by the editor of one of its sister journals:
Any manuscript reporting genetic modification of human embryos or gametes would need to follow strict scientific and ethical guidelines… On the basis of the available information, He’s research would not have met the editorial criteria adopted by Nature journals.4
JK decided to try again, resubmitting to the Journal of the American Medical Association (JAMA). In an extraordinary move, the JAMA editor sent the manuscript to eleven outside experts for review in early December, including George Church, before it too rejected the paper.
A year after being spotted on a guest house balcony, we finally learned JK’s fate. On December 30, 2019, the Nanshan District People’s Court of Shenzhen found JK and two colleagues guilty of “illegal medical practice.” (There is no law in China that expressly prohibits editing human embryos.) JK was sentenced to three years in prison, a fine of 3 million yuan (about $430,000), and a ban from further research in “assisted reproductive technologies.”I Qin Jinzhou (the first author on JK’s manuscript) and Zhang Renli were handed fines and suspended sentences of eighteen and twenty-four months, respectively. “The three accused did not have the proper medical certification to practice medicine, and in craving fame and wealth, deliberately violated national regulations in scientific research and medical treatment,” the court stated. “They have crossed the bottom line of ethics in scientific research and medical ethics.”5