In June 2019, a team of scientists published a study that sent ripples throughout the scientific community. The mutation that Chinese scientist He Jiankui had tried to engineer into embryos last year, the authors argued, could shorten their lives.
Last week, the team retracted their paper, originally published in Nature Medicine, citing flaws in its design. Subsequent efforts by independent research groups to repeat the study’s results also failed to find evidence that people carrying the mutation in question were more likely to die early.
Study author Rasmus Nielsen, a biologist at the University of California, Berkeley, first came forward about the paper’s validity on Twitter in September, STAT News’ Rebecca Robbins reported at the time.
“The one thing that all scientists fear the most is to find out that a major result they have published was based on erroneous data,” Nielsen wrote.
The original study was conducted in the wake of a series of controversial experiments by He Jiankui, who attempted to use the CRISPR gene-editing tool to manipulate a gene called CCR5 in twin girls born last year. A subset of the human population is born naturally carrying a version of CCR5 with a mutation called ∆32, which has been linked to resistance against certain strains of HIV. Though He’s strategy didn’t induce this exact mutation, it was intended to be comparable, and protect the infants—whose father is HIV positive—from succumbing to AIDS.
He was widely condemned by the scientific community, who criticized his work as, among other things, a serious and irresponsible breach of ethics—in part because the side effects of CRISPR editing and the mutation itself aren’t fully understood. The experiment also had mixed results: while He modified both copies of CCR5 in one of the twins, her sister carried edits in just one. And none of the changes were an exact match to naturally-occuring ∆32.
When published, Nielson’s study appeared to raise the stakes further. After sifting through genomes recorded in the UK Biobank database, Nielson and his team reported that carrying two copies of ∆32 was linked to earlier age of death.
In the months following, several groups (including Nielson’s) tried in vain to generate similar results with other genomic databases. With the help of David Reich, a geneticist at Harvard Medical School, Nielsen and his colleagues discovered a systematic error in the data they’d initially analyzed that had caused them to underestimate the number of people in the UK Biobank database with two copies of ∆32, reports Ewen Callaway for Nature News. This mistake had given the researchers the false impression that double-∆32-ers were unexpectedly underrepresented in the population—something that’s often related to higher mortality.
Adjusting for that error essentially erases the effect the team initially observed. “If there is an effect on mortality, it is certainly not as strong as we previously reported,” Nielson told Robbins at STAT.
That doesn’t mean ∆32 is entirely without its costs: Other studies have linked the variant with worse outcomes in flu and West Nile virus infections. On the flip side, CCR5 may still have valuable functions that haven’t yet been measured, Reich told Calloway, and seems “unwise” to edit out.
The retraction also doesn’t absolve He of his faulty scientific approach, which may still have yet untold repercussions. CRISPR remains a valuable tool for researchers worldwide, but it’s a tool that “needs to be used judiciously,” Dana Waring, education director and co-founder of pgEd, the Personal Genetics Education Project, told NOVA Next last year.
Importantly, the error doesn’t affect the rest of UK Biobank’s data, Nielson told Callaway. And it’s something the team should have caught the first time around, he said, adding, “We are very sorry for the confusion we might have spread by publishing these results.”