CRISPR caused unintended changes to the genetic code in 16% of human embryo cells in a new study, and its authors say the findings highlight that more research is needed before the gene-editing tool is safe for human fertility applications.

CRISPR-Cas9 is often touted for its precise gene-editing abilities, but a handful of recent studies have shown that the technique can cause unwanted deletions or rearrangements of the genome in animal cells. Now, the paper published April 9 in PNAS shows that CRISPR can also cause these types of unexpected and potentially harmful changes in human embryos.

"There are groups that want to apply CRISPR in the clinic to edit the germ line or edit embryos and then implant them, for example, in women undergoing [in vitro fertilization]," said lead author Gregorio Alanis-Lobato, a principal computational biologist at pharmaceutical company Boehringer Ingelheim who did the work as a postdoctoral researcher at The Francis Crick Institute, in London. 

"What our study shows," he continued, "is that we cannot bring CRISPR to the clinic yet, and the reason is that there are unintended effects that could lead to diseases like cancer or have other serious consequences."

Since CRISPR-Cas9 burst onto the molecular biology scene in the mid-2000s, it has revolutionized gene editing in medicine and agriculture. The tool works like a pair of molecular scissors, allowing researchers to precisely remove or edit pieces of genetic code. The Cas9 protein unzips double-stranded DNA, prompting the cell's natural repair pathways to fix the break. 

In doing so, these pathways often add or delete a few letters of genetic code, which inactivates the target gene. But these patch-ups may cause modifications beyond the gene of interest, and such changes can be overlooked if researchers check only small regions of DNA above and below the gene, according to Alanis-Lobato.

To evaluate broader nontarget modifications after CRISPR, the researchers used whole genome sequencing to evaluate the entire genetic code of cells. Using extra human embryos donated by people undergoing IVF, the researchers used CRISPR-Cas9 to target POU5F1, a gene found on chromosome 6. 

They found that four out of 25 CRISPR-modified cells, or 16%, had abnormal additions or removals of DNA, most of which were between 4,000 and 20,000 letters of genetic code. The effects of these unintended modifications to DNA are unknown, but they could cause developmental abnormalities or lead to disease later in life, if the embryos were to continue developing, Alanis-Lobato said.

For comparison, the researchers did not detect chromosomal abnormalities in uninjected control cells.

"The most important message is that we cannot apply CRISPR in the clinic until we understand how the embryo reacts to CRISPR. So we need more basic research, we need more data, not only for POU5F1, but also in many other genes," Alanis-Lobato told The Academic Times. "Once we understand that, probably we can start talks about bringing this to the clinic — but definitely not yet, and I don't think this will happen in the near future."

The other "very important message" for people using CRISPR for research is that "they have to check for these events," Alanis-Lobato added. 

"If you're studying the function of a gene in a certain biological context, and you knock out a gene with CRISPR, you may also knock out other genes as well because of these unintended events," he said. "These events may go unnoticed, and they can confound interpretation of the results."

The researchers developed and made public a computational pipeline to simplify analysis of such unintended modifications by CRISPR. 

The study, "Frequent loss-of-heterozygosity in CRISPR-Cas9–edited early human embryos," published April 9 in PNAS, was authored by Gregorio Alanis-Lobato, Boehringer Ingelheim; Jasmin Zohren, Afshan McCarthy, Emily Hardman, Maria Greco and James M. A. Turner, The Francis Crick Institute; Norah M. E. Fogarty, The Francis Crick Institute and King's College London; Nada Kubikova, University of Oxford; Dagan Wells, University of Oxford and Juno Genetics; and Kathy K. Niakan, The Francis Crick Institute and University of Cambridge.