Katie Hasson recounts the recent failures to edit the human genome.
The “CRISPR babies” announced in headlines around the world recently turned two years old, and we still know nothing about their health or wellbeing. Debates continue about whether the societal risks of heritable genome editing are too great to proceed, as do calls for broadly inclusive public participation in such deliberations. In the meantime, we’ve been learning a lot about what can go wrong when using CRISPR to edit human embryos.
In October, the journal Cell published an article describing significant damage to human embryos edited with CRISPR. The experiments conducted in Dieter Egli’s lab at Columbia University found unintended rearrangements or deletions of large stretches of DNA at and around the targeted site. In some cases, the deletions were so large that an entire chromosome was lost. As one headline plainly put it: “In Embryos, Crispr Can Cut Out Whole Chromosomes—That’s Bad”.
The Cell article is the published version of a pre-print paper that, along with two others from Kathy Niakan’s and Shoukrat Mitalipov’s labs, received significant attention in June. At that time, Heidi Ledford reported on the trio in Nature under the headline “CRISPR gene editing in human embryos wreaks chromosomal mayhem”. The headline’s alarm reflected the startling findings: multiple experiments using CRISPR to edit the DNA of early embryos revealed large unwanted edits that could damage genes around the intended target. Significantly, these papers are from three of the best known research groups using CRISPR to modify human embryos.
Urnov: “There’s no sugarcoating this. This is a restraining order for all genome editors to stay the living daylights away from embryo editing.”
Musunuru: “Nobody has any business using genome editing to try to make modifications in the germline.”
Concerns about the significant technical challenges and risks of embryo editing with CRISPR are not new. Off-target edits and mosaicism are two known problems that scientists have been working to address.
The Egli, Mitalipov, and Niakan studies are the most recent to uncover previously unrecognized or underappreciated technical and safety issues in germline genome editing. The article from Egli’s lab even calls into question results of previous experiments that had claimed to show successful and safe embryo edits.
The idea that reducing technical errors in embryo editing experiments proves heritable genome editing safe continues to miss consideration of a broader set of problems. These include risks to any women who would carry gene-edited pregnancies, as well as risks to the health and well-being of any children born from edited embryos, and to any subsequent generations. Beyond these concerns lie urgent questions about the societal risks of altering the genes of future children and generations, which we have barely begun to discuss.
Ethical and policy discussions of heritable genome editing often start with some version of “assuming germline genome editing is safe and effective…,” making it seem like we are closer to this hypothetical state than we may ever be. Research findings of “chromosomal mayhem” show how far away we really are.
These results also place an inconvenient bump in the smooth progression to heritable genome editing envisioned by proponents who continually assure the public that scientists have nearly ironed out the technical wrinkles.
This tension was apparent in Heritable Genome Editing, the recent report from the International Commission on Clinical Use of Human Germline Editing, convened by the US National Academies of Science and Medicine and the UK Royal Society. The commission was tasked with “develop[ing] a framework for considering technical, scientific, medical, regulatory, and ethical requirements for germline genome editing”—in other words, a “translational pathway” to the use of heritable genome editing in the clinic—“should society find such applications acceptable.” That final caveat aside, “building an expensive bridge to a remote island” does seem to presume that it leads somewhere we want to go.
The report’s recommended pathway lays out intricate plans for moving from our current limited knowledge about human germline editing to using it for reproduction. To its credit, the report begins from a strong statement that this technology is not currently safe, then details the many incremental steps needed to establish that heritable genome editing would be effective and safe enough even to begin clinical trials. In doing so, it underscores how challenging, labour intensive, and expensive this “translational pathway” would be and how foolish it is to start our discussions by assuming these difficulties away.
Safety is far from the only thing we need to discuss in making decisions about whether to pursue heritable genome editing, but recent scientific findings confirm that the question of whether it will ever be safe is far from settled.