Bridging Disciplines to Study CRISPR-Induced Chromosome Destabilization

A collaboration between friends led to a cautionary finding about CRISPR’s effect on cells.

Aparna Nathan

Aparna is a freelance science writer pursuing a PhD in bioinformatics and genomics at Harvard University. Her writing has also appeared in The Philadelphia Inquirer, Popular Science, PBS NOVA, and more.

View full profile.

Learn about our editorial policies.

Apr 8, 2022

When David Pellman and Mitchell Weiss made plans to catch up over dinner, they did not expect it to lead to a scientific discovery. Because of this get-together, the pair published a study earlier this year in Nature Genetics showing that CRISPR can cause a cell’s genome to fall into disarray.1

Weiss, a physician scientist from St. Jude Children’s Research Hospital, was on a business trip to Boston where Pellman works at Dana-Farber Cancer Institute. The pair are friends and experts in their respective fields; Pellman studies genomic instability during cell division and Weiss is a hematologist who develops gene therapies for sickle cell disease.

“There's something particularly nice about having dinner with your old friend when you're a little relaxed and you're talking about science,” Weiss said. “That's extra special.”

The conversation veered toward CRISPR, a tool that Weiss has been testing to mitigate sickle-cell-causing mutations in patients’ hematopoietic stem cells. To do this, the Cas9 endonuclease snips through both strands of DNA at a genomic site that matches its guide RNA sequence. 

But outside of CRISPR applications, double-stranded breaks are something to avoid at all costs. They pose a danger to cells because they can destabilize the chromosome structures that organize DNA. Pellman studies chromosome breaks in the context of cancer, where this instability can trigger cells to divide uncontrollably. He previously showed that when cells divide, double-stranded breaks in the DNA can create fragments that float away from the nucleus and end up encapsulated in a separate membrane to create a micronucleus.2 

Double-stranded DNA breaks can cause fragments to break off and form micronuclei (shown in red).
Double-stranded DNA breaks can cause fragments to break off and form micronuclei (shown in red).
David Pellman

“It’s a little bit like musical chairs,” Pellman said. “As soon as mitosis is finished, chromosomes will get a nuclear envelope, no matter where they are.” Within these micronuclei, DNA can be shattered into many small pieces and shuffled around, a process called chromothripsis.

While some previous studies revealed unexpected chromosomal rearrangements after CRISPR-Cas9 editing, they had not systematically tracked these incidents to figure out how frequent they are or what caused them. Gaetan Burgio, a CRISPR researcher at Australian National University who was not involved in the study, noted that while safety is an important consideration, researchers are still wrestling with how to efficiently deliver the CRISPR machinery to make edits in the first place.

Pellman and Weiss decided that this was an important question to investigate, especially as CRISPR-based therapeutics advance toward clinical use in patients. They put together a team of scientists developing CRISPR-based therapeutics and experts in chromosomal abnormalities. The team used CRISPR-Cas9 to edit cells in the lab and imaged single cells to look for chromosomal rearrangements. The scientists confirmed their suspicions; the double-stranded cuts that CRISPR-Cas9 makes in the genome can lead to chromosomal instability, and at higher rates than expected. They credit their unique interdisciplinary collaboration as a key to making the study possible.

“I think there has been quite a surprising amount of interest in his story because it connects groups of researchers that have not really been in communication,” Pellman said. “Now it's obvious that they should be in communication to at least consider this possible type of outcome after clinical treatments.”

Reactions to the study have varied. The team received criticism that publishing these results might unnecessarily scare people away from using CRISPR therapeutically. Some scientists became alarmed that this might be the death knell for CRISPR, a reaction that Weiss said is unwarranted. “The point of writing this paper was not to say CRISPR is a bad thing or that it shouldn't be used for therapeutics,” Pellman said.

Others appreciate the study’s goals and are more cautious about its implications. Burgio noted that with the current low efficiency of CRISPR editing and the fact that many cells with rearranged genomes will die, the overall impact could be limited. However, he said that it is still important to see how prevalent chromosomal consequences of Cas9-induced double-stranded breaks can be. Pellman agreed, noting that even if an adverse event is rare, all it takes is one event among the millions of cells being edited to create a malignant cell.

Pellman said that responses to the study have mostly centered on finding ways to manage risk to ensure patients’ safety. One option—depending on the tissue that is being edited—is to monitor patients by regularly sequencing their DNA or measuring the number of copies of genomic segments. Other ways to avoid problems from chromosomal instability are to edit non-dividing cells or to use different technologies like base-editing that do not make double-stranded breaks.

Importantly, this study only describes the chromosomal phenomena in a lab setting, but stops short of making claims about clinical applications. CRISPR experiments in animals and humans, including Weiss’s own studies, have not yet reported the generation of cancerous cells. 

“We don't know if a problem is going to manifest in the clinic or not, so you don't want to jump ship,” Weiss said. “You just want to be aware.”


  1. M.L. Leibowitz et al., “Chromothripsis as an on-target consequence of CRISPR-Cas9 genome editing,” Nat Genet,” 53(6):895-905, 2021.
  2. C. Zhang et al., “Chromothripsis from DNA damage in micronuclei,” Nature, 522(7555):179-84, 2015.
Research Resources Logo