There’s much we need to learn about it, though, and a new Nature Medicine study highlights that fact. Led by Novartis Institutes for Biomedical Research in Cambridge, a pharmaceutical company based in Switzerland, their work with CRISPR found that it tends to kill off embryonic stem cells, those that can differentiate into any cell type in the human body.
It appears that the gene-editing techniques double-strand breaks – snips in both strands of DNA’s double helix – are, as the authors of the paper term it, “toxic,” which leads to a die-off of most stem cells. They report that this effect was less apparent in previous studies because the techniques’ efficiency was lower.
What’s particularly concerning is that in the remaining stem cells can contain a malfunctioning gene named p53. When this gene is working, it helps cells to deal with stressors, and instructs damaged cells to die or stop replicating, which makes it a powerful anti-cancer gene. As a new, separate study focusing on p53 elucidates, a malfunction of this gene causes at least half of all cancers.
The problem is that p53 reduces the efficiency of CRISPR gene-editing. CRISPR is inadvertently going for cells without a functioning p53 gene because it’s better at fixing them, but this essentially leaves behind edited cells with cancer-prone mutations in them.
This study therefore suggests that there’s a risk that the CRISPR editing of embryonic stem cells may up the risk of cancer down the line, but precisely what that risk may be remains unclear.
As it happens, yet another CRISPR/p53 Nature Medicine paper by a different team – this time led by Stockholm’s Karolinska Institute – has also been published this week. Although these researchers were probing cells found in the human retina, the outcome was the same as the aforementioned work: CRISPR leads to a selection of cells with a dysfunctional p53 gene.
Dr Alena Pance, a senior staff scientist at the Wellcome Trust Sanger Institute – and who wasn’t involved in the latter study – explained to Science Media Centre that these results were certainly important, but there's a huge caveat.
She cautioned that “the results are preliminary,” noting that they mainly show “the effect of p53 in one particular cell line.” At this point, it’s unclear whether this happens to these cells exclusively, or other types.
Professor Robin Lovell-Badge, group leader at The Francis Crick Institute, added that the need to reduce p53 activity was known previously through the study of pig cell lines.
Noting that the retinal cells in the second study were likely suffering from cellular stress, he suggests that the issues raised by these two studies may be relatively specific. He also stressed that cells used in CRISPR should always be checked for p53 mutations and not used if found to contain them.
“It is therefore not apparent that the problems identified by these two papers, and certainly the scary press releases, are entirely justified,” he concluded.
Researchers who are keen to investigate CRISPR’s potential aren’t blindly unaware that we still need to understand what side-effects snipping parts of an organism’s genetic code may have. One oft-cited worry is that of mosaicism, which refers to edited embryos that, as they divide, still contain unedited DNA.
The other frequently discussed problem is that of the accidental triggering of cancer. New Scientist recently suggested that this was becoming less of a concern over time, but this new study moves it back into the spotlight. It appears that investors in CRISPR are taking a hit on this news, and MIT Technology Review suggests this research may be why a human CRISPR study in the US was canned by the Food and Drug Administration back in May.
The second papers' authors emphasized that they aren't saying CRISPR is “bad or dangerous.” Affirming that it will clearly be a vital biomedical tool, they are merely pointing out that safety concerns must be investigated thoroughly.
As Prof. Darren Griffin, a professor of genetics at University of Kent, put it, this work “provides reason for caution, but not necessarily alarm.”