Let’s take the first three items on Yong’s list. Just as David Liu had asked JK directly in Hong Kong, what was the unmet medical need for this couple? Why did JK pick HIV and CCR5? Why was genome editing even necessary given that JK’s team performed sperm washing prior to IVF to remove any risk of HIV? JK could not give Liu a straight answer, so instead he painted the scourge of HIV in broader terms. “I truly believe that not just for this case but for millions of HIV children, they need this protection because a vaccine is not available. I have personally experienced ‘AIDS village’ where 30 percent people were infected. They had to give their children to relatives.” JK said that he was proud of his work; the twins’ father had been depressed but had pledged to work hard and take care of his new family.
AIDS has claimed an estimated 35 million lives. There are an estimated 37 million people worldwide infected with HIV, including more than 800,000 in China. The advent of a cocktail of antiretroviral drugs is one of the great recent medical success stories. These drugs can suppress the titer of HIV to undetectable levels and effectively abolish the risk of HIV transmission through sex. An advertisement from antiviral drugmaker Gilead, spotted in a Broadway theater playbill, said: “Dear HIV, We didn’t give up. XOXO, Science.”
Earlier discussions about which disorders might prove candidates for heritable gene surgery seldom considered HIV, which was stuck in the “definitely maybe” category. Inactivating CCR5 to enhance protection against HIV isn’t in the same category as correcting a disease-causing mutation. But JK had identified two major reasons for his choice: safety and “real world medical value.”2 He cited two decades of research on CCR5 and related clinical trials.3 He also considered his gene surgery as relatively simple, seeking to disrupt the CCR5 gene to scuttle the HIV receptor rather than trying to perform precision DNA surgery, swapping one letter in a gene for another, as would be needed if attempting to treat patients with muscular dystrophy or sickle-cell disease.
Worldwide about 100 million people carry the CCR5 Δ32 mutation, with the highest concentration among northern Europeans. All evidence suggests the deletion is safe. If JK had edited the CCR5 gene to mimic the naturally occurring Δ32 mutation, the scientific community’s concerns might have been muted. But the mutations JK engineered in Lulu and Nana did not produce the Δ32 deletion, raising serious questions about the impact of these man-made genetic alterations, never before seen in humans or tested in an animal model. JK targeted a spot in the CCR5 gene where he could land the Cas9 nuclease, like pointing a cursor on a particular word. But he couldn’t accurately control the extent of the edit itself. “These two lives are now an experiment, a matter of scientific curiosity, which is an outrageous way to relate to human lives,” said Ben Hurlbut.4
Sean Ryder, an RNA biologist at the University of Massachusetts Medical School, vented his fury on Twitter during the summit and subsequently in The CRISPR Journal. “Doudna warned us this might happen,” Ryder wrote. “I had trusted that my fellow scientists would ensure that human trials involving embryonic genome editing would be done transparently, ethically, and with strong moral purpose.”5 But no.
In Nana, both copies of CCR5 harbor so-called frameshift mutations: one copy carries a small deletion of 4 bases (-4), whereas the other copy actually has an insertion of a single base (+1); both mutations disrupt the natural phase of the genetic code.I It is possible, even likely, that these mutations would abrogate the function of CCR5 similar to the Δ32 mutation. But we don’t know for certain. These mutations are “never-before-seen variants of unknown significance,” said Ryder. “They might inactivate CCR5 activity and block HIV uptake, but they might also incur new risks.”
The situation in Lulu was even less clear-cut: one copy of the gene appeared to be untouched, while the other carried a fifteen-base deletion. The resulting loss of five amino acids in the middle of an otherwise intact protein might abolish the function of CCR5 but this has never been tested. Moreover, in both twins, the edits appeared to be mosaic, meaning that not every cell in the twins carried the gene edits. Mosaicism is a natural phenomenon in biology—think heterochromia, a condition that results in different colored eyes. But no doctor, certainly not JK, could say with conviction that he had first done no harm.
There were also serious questions about off-target effects, despite JK’s denials, including some evidence that such an effect had occurred in Lulu, and others could be uncovered if more rigorous whole genome sequencing were performed. Ryder closed his CRISPR Journal commentary with an impassioned plea:
It is my fervent hope that Lulu and Nana will lead long healthy lives. It is certainly possible, even plausible, that they will remain healthy based upon what is known about CCR5. However, there are enough uncertainties about the function of the mutations to raise clear scientific objections to the work, in addition to the moral and medical objections. I pray that Lulu and Nana are never exposed to HIV. I hope we never have a chance to find out if the experiment ‘worked.’ To me, the best outcome is that the babies are unaffected by the procedures of the He lab. Worse outcomes are possible.6
Even the assumption that deactivating CCR5 neatly thwarts HIV with no other health consequences started to crumble. The Δ32 mutation arose in northern Europe but as Lovell-Badge pointed out “there are almost no people with the Δ32 mutation living in China. Therefore, it is necessary to ask why?”7 Although it is likely that the deletion hasn’t had time to spread through Asia, another possibility is that the CCR5 mutation might have some other impact on health. We know Δ32 increases the risk of West Nile and possibly influenza. And some have even questioned the dogma that inactivating CCR5 protects against HIV. Apparently there are rare strains of HIV, including one called X4, that bypass CCR5 by using a different portal into the cell.8
Months after the twins’ birth, more disturbing scenarios hit the news. “China’s CRISPR twins might have had their brains inadvertently enhanced,” screeched a trademark Regalado headline.9 Alcino Silva, a UCLA neurobiologist, said the CCR5 mutations might damage cognitive function in the twins—based on his work in mice. His team’s latest data suggested that individuals lacking CCR5 recover more quickly from strokes, while carriers of a single inactive CCR5 gene perform better in school.10 Meanwhile, a pair of Berkeley researchers mining the UK Biobank—a charity- and government-funded database of genomic and medical information on more than 500,000 Brits—reported that individuals with two copies of the Δ32 mutation were 20 percent more likely to die before age seventy-six than people with one normal copy of CCR5.11 Media reports pounced on the supposed threat to the CRISPR babies’ life expectancy (ignoring that the twins’ edits were not the Δ32 deletion). But some excellent sleuthing by Sean Harrison, an epidemiologist at the University of Bristol in the UK, highlighted a systematic error in the study, leading eventually to its retraction.12
The CCR5 story spotlights a critical issue: how can we be certain that editing any gene will not have some unforeseen, knock-on, collateral damage? Our 10 trillion cells are sacks containing vast interconnected networks of protein-protein, RNA, and other biomolecular interactions laden with intracellular hubs and vesicles, membranes and molecular machines. When I studied biochemistry in college, we naïvely thought the key metabolic pathways of a cell could be displayed neatly on a wall poster. Today, we know that each of the more than 20,000 proteins in a cell interacts with dozens if not hundreds of other proteins. Predicting what happens next when editing any one of these important cogs and spokes may be possible one day with advances in artificial intelligence, but not today.