A paper published in Nature that suggested babies born following treatment with CRISPR would have shorter lifespans has been retracted by the authors due to ‘technical errors’ made during analysis that have undermined the researchers’ conclusions.
In November 2018, Chinese geneticist He Jiankui announced he was part of a team that had used CRISPR-Cas9 to edit the genomes of human embryos which had been implanted in a patient and had reached full term. Twin girls had been born who were healthy and were doing well.
The researchers had used CRISPR to disable the CCR5 pathway that encodes a protein that enables HIV to enter cells, thus preventing HIV infection. Genetic studies revealed the treatment had worked and that the genetic pathway used by HIV to infect cells had been disabled.
The researchers artificially recreated a genetic defect in the CCR5 gene known as delta-32, as studies had shown that individuals with the naturally occurring delta-32 mutation were resistant to HIV and appeared to have no observable health defects.
Concerns were raised about how this largely untested gene editing technique would affect the future health of the babies. The study in Nature suggested individuals with two copies of the CCR5 mutation were more likely to die early. One of the twins had two copies of the mutation and therefore was likely to have a shortened lifespan. The other twin only had the 32-DNA letter deletion in one copy of the CCR5 gene.
The authors of the study, Rasmus Nielsen and Xinzhu Wei at Berkeley, had reviewed data in the UK Biobank, which contains genome data of 500,000 British people. They estimated approximately 1% of those individuals had two copies of the naturally occurring CCR5 mutation, and that they were more likely to die at the age of 76 than others who did not have the mutation. They found the UK Biobank data included fewer individuals with the mutation than was expected, which suggested individuals with two copies of the CCR5 mutation were dying earlier.
Several researchers attempted to replicate the researchers’ findings, but the findings could not be reproduced. No evidence was found to suggest individuals with two copies of CCR5 mutation were dying earlier than those without the mutation.
Harvard Medical School population geneticist David Reich, who was also working on CCR5, worked with Nielsen to reevaluate the study and found that errors had been made identifying the delta-32 mutation which had resulted in the number of individuals with the CCR5 mutation in the UK Biobank data being undercounted. This was due to the method used to track the mutation, which did not always identify the targeted sequence. That explained why there was an apparent lack of individuals carrying the double mutation in the UK Biobank data. The error was not due to the technique used as it was effective at identifying other mutations. It just did not always correctly identify the gene variant they were searching for.
“I feel I have a responsibility to put the record straight for the public,” says Nielsen, lead author of the retracted study. “There were checks we could have done and should have done that we didn’t do. We missed the fact that there was a genotyping error.” Consequently, there is no evidence that the CCR5 mutation results in a shortened lifespan.