As creator of ‘CRISPR babies’ nears release from prison, where does embryo editing stand?


Biophysicist He Jiankui, having served a 3-year sentence for creating the world’s first genetically engineered babies, may be released from a Chinese prison this week, Science has learned. He’s largely secret use of the genome editor CRISPR to alter the DNA of human embryos and implant them into two women led to three births, sparking ethical outrage and fears for the babies’ health (about which little is known). It did not, however, bring an end to basic research on human embryo editing.

The response to He’s November 2018 announcement was “severe and vibrant,” says Fyodor Urnov, who studies CRISPR-based genome editing at the University of California, Berkeley. For now, Urnov sees no circumstance that would justify efforts to genetically modify babies. But he strongly supports using CRISPR to fix disease-causing mutations after birth, without causing heritable changes to a human genome, and regrets that “we have poured a jar of tar on gene editing.” And Urnov and others believe that, used responsibly and safely, embryo editing could eventually prove a powerful tool against disease in rare circumstances. In laboratory studies, they continue to explore possible avenues—and the many hurdles.

The work has proceeded with little notice. “The pandemic has pushed this topic out of people’s primary attention,” says Alta Charo, an emeritus bioethicist and lawyer at the University of Wisconsin, Madison, who notes that oversight measures intended to stop rogue experiments like He’s have stalled, including a proposed global registry of preclinical heritable genome-editing research.

This type of registry might have noted a study reported last week in which a research team working with surplus human embryos from in vitro fertilization (IVF) clinics showed how CRISPR could rid a newly fertilized egg of an extra copy of a chromosome—a problem that can lead to Down syndrome and other medical conditions. Other groups are exploring how to introduce heritable genetic changes via human sperm or eggs. There are “quite a lot of people pushing boundaries” in that regard, says Robin Lovell-Badge, a developmental geneticist at the Francis Crick Institute—although few if any think the work is ready for the clinic. “We’re still waiting for some better tools,” says developmental biologist Shoukhrat Mitalipov of Oregon Health & Science University.

The original concerns about designer babies centered on CRISPR’s sloppiness. The DNA-cutting enzyme that is one of its two components occasionally slices unintended spots, and even if the cut is on target, the cell’s gene repair equipment may scramble adjacent DNA by inserting or deleting bases, potentially creating new harm. Indeed, a study of CRISPR-altered human embryos found 16% had these “unintended editing outcomes” at the targeted DNA, a group led by Kathy Niakan of the Crick reported last year in the Proceedings of the National Academy of Sciences.

Genetic screening of edited IVF embryos might not catch these errors. Although CRISPR is introduced right after fertilization, at the single-cell stage, its action is not necessarily immediate. “The edit may occur at the two-cell or four-cell stage, so not all the cells are identical,” Lovell-Badge says—a phenomenon called mosaicism. Both incorrectly altered and unaltered cells can easily go undetected because an embryo is screened by taking a sample of its cells at the 5-day stage, when it contains about 100 cells. “If you have any mosaicism, then you don’t know what you’ve got in the rest of the embryo,” Lovell-Badge says.

Stem cell researcher Dietrich Egli at Columbia University hopes to find a way to start and stop CRISPR at the embryo’s single-cell stage, preventing mosaicism. In the meantime, his group has found a specific kind of CRISPR edit for an embryo that vastly reduces the risk of unintended DNA changes.

One of the most common abnormalities found when IVF clinics screen embryos, especially those made with eggs of older people, is the presence of either one or three copies of certain chromosomes rather than the normal two. In a preprint posted on bioRxiv on 10 March, Egli’s group demonstrated a strategy for trisomy, an errant third chromosome. The scientists showed they could target an extra paternal or maternal chromosome copy with a CRISPR cut at or near its centromere, the DNA-protein structure that holds the different arms of a chromosome together. The extra chromosome then falls apart during cell division. Unintended on- or off-target edits theoretically wouldn’t matter because CRISPR would, in effect, destroy the entire DNA sequence.

Mosaicism might still be a problem if CRISPR does not correct the trisomy in all of the cells in an early embryo, but Egli notes that when such embryos have a mix of cells with normal and abnormal chromosomes, a natural “rescue mechanism” usually seems to eliminate the abnormal cells. “There are still multiple obstacles,” he stresses. “We could have given this a different title, ‘Correction of Trisomy 16 in Human Embryo,’ and we might have created more buzz and news articles, but we didn’t think it was appropriate because it conveys that you’re going to do this clinically tomorrow, which is absolutely not the case.”

Researchers studying CRISPR in human embryos face obstacles beyond the science. In the United States, Congress forbids government funding of research with human embryos, forcing Egli, Mitalipov, and others to rely on foundations, academic institutions, or companies. Legislation also prevents the U.S. Food and Drug Administration from even evaluating therapies that edit human embryos.

Editing the DNA of egg or sperm precursor cells may avoid some of these hurdles. It also gets around what Kyle Orwig, a reproductive biologist at the University of Pittsburgh, calls “a numbers problem.” Even under the best circumstances, IVF clinics could only create, edit, and test a small number of embryos for a given couple, giving them few chances to get it right.

Editing the cells that give rise to sperm could improve the odds. Researchers have already removed these spermatogonial stem cells from mice and grown millions in culture. This allows for a rigorous quality control of CRISPR edits: Scientists can screen for the stem cells that have the correct edit, with no unintended DNA changes, and clone them en masse, again checking for errors. Then, they can transplant those cells into the testes where they should produce mature sperm, Orwig says. Indeed, rodents with edited sperm stem cells have been used to create offspring with a desired DNA edit.

Turning that basic research into a way to help potential parents won’t be easy. “The barrier is that we don’t yet know how to maintain human spermatogonial stem cells in culture,” Orwig says. His team is exploring a different route to creating edited sperm stem cells: “reprogramming” adult human cells into a stem cell state and trying to coax them partly down the pathway that creates sperm. Other groups are hoping that reprogrammed adult cells could one day produce human eggs, which could then be altered in large numbers.

Discouragingly, in mice, spermatogonial stem cells only survive when placed into newborn animals, which isn’t a realistic option for humans. As a first step to exploring whether the scheme could work in people, Orwig’s team is now recruiting men who became infertile because of cancer treatment and had testicular tissue or cells frozen before chemotherapy or radiation. The team plans to isolate spermatogonial stem cells from the thawed tissue and then inject them, unedited, into the owner’s testis to see whether that produces viable sperm.

The 3 years since He went to prison have seen glimmers of progress in heritable human genome editing, but many scientists say the increased awareness of CRISPR’s shortcomings has underscored the recklessness of transplanting edited embryos with the technology available today. An exception is Russian geneticist Denis Rebrikov, one of the few scientists after the He scandal to openly advocate implanting edited embryos into people. “We’ve done a lot of validation experiments, and now we’re confident that we can move on to real clinical use,” Rebrikov says.

Lovell-Badge speaks for most researchers when he says such confidence is unwarranted. Stick to lab work on embryo editing for now, he advises. “People should do as much preclinical research as they can, and let’s find out whether it’s feasible.”

Science, 21 March 2022