
NEW YORK – Researchers from the Dana-Farber Cancer Institute and St. Jude Children's Research Hospital have found that using CRISPR-Cas9 to generate double-strand breaks (DSBs) in DNA in order to edit a gene can lead to the mutational process chromothripsis, which can lead to human congenital disease and cancer.
In a study published on Monday in Nature Genetics, the researchers said they experimented with model cells, single-cell whole-genome sequencing, and editing in clinically relevant cells to demonstrate that CRISPR-Cas9 editing generated structural defects of the nucleus, micronuclei, and chromosome bridges, thereby initiating extensive chromosome rearrangements, or chromothripsis. In fact, they found that genome editing with Cas9 in actively dividing cells caused about a twentyfold increase in the formation of micronuclei and/or chromosome bridges.
Their data suggests that chromothripsis is a previously unappreciated on-target consequence of CRISPR-Cas9-generated DSBs and that the potential for extensive chromosomal rearrangements should be considered and monitored as CRISPR is therapeutically applied in the clinic, the researchers said. They further noted that while chromothripsis is implicated in tumorigenesis, the carcinogenic potential of CRISPR-Cas9-induced chromothripsis would likely depend on the set of genes on the targeted chromosome arm and whether rearrangements occurring after the initial cut caused those genes to be deleted, fused, or amplified.
Cas9 generates a DSB that cleaves the targeted chromosome into two segments, according to the researchers — one with the centromere region called the centric fragment and one without it called the acentric fragment. If the DSB is not repaired before cell division, the acentric fragment lacking a functional centromere can form a micronucleus. They evaluated this process in genetically stable human retinal pigment epithelial cells using single-guide RNAs targeting unique genomic sites on four different chromosomes.
They found that CRISPR-Cas9 cutting at individual target sites induced micronucleation at frequencies of 4 percent to 7.5 percent, or 10.2-fold to 19.3-fold, higher than in controls. Using fluorescence in situ hybridization (FISH), they established that 81 percent to 92 percent of the micronucleation contained the chromosome arm targeted by the specific gRNA species. Most micronuclei contained two copies of the targeted chromosome segment. Further FISH probes confirmed that the micronucleations were mostly acentric chromosome fragments.
In an experiment on allele-specific genome editing, the researchers used gRNAs that targeted only one allele due to a heterozygous polymorphism in the protospacer adjacent motif (PAM). They detected editing events exclusively on the targeted homolog, and these editing events were associated with a 2.7-fold and a twelvefold increase in micronucleation frequency for chr1p- and chr5q-targeting gRNA species, respectively. Further, they noted that allele-specific gRNAs primarily generated micronucleation with two copies of the targeted chromosome, meaning that allele-specific guides did not eliminate genome editing-induced micronucleus formation in actively dividing cells.
To directly test whether chromothripsis might be an unrecognized, on-target consequence of CRISPR-Cas9 genome editing, the researchers then used a process they called Look-Seq, which combined long-term live-cell imaging with single-cell whole-genome sequencing of the imaged cells. They found that on-target Cas9 genome editing generated micronuclei, which, in turn, induced chromosome arm-level DNA copy number alterations as well as copy number-neutral loss of heterozygosity.
The researchers did note that malignant transformation or abnormal clonal cell expansion following genome editing has so far not been observed in animal studies, including non-human primate models, nor has it been shown in the small number of humans who have participated in clinical trials. Therefore, the clinical risks associated with nuclease-based genome-editing therapies in human participants remain unclear, and the rates of forming micronuclei or chromosome bridges followed by chromothripsis are as yet unknown for any therapeutic application of CRISPR.
However, the authors added, these results do have several practical implications. First, because efficient Cas9-mediated homology directed repair requires cells to be actively dividing, therapeutic genome editing via non-homologous end-joining in non-dividing cells, such as retinal photoreceptors, should not produce micronuclei. Further, they said, screening for micronucleation and/or chromothripsis in clinical protocols is expected to become more feasible as high-throughput and low-cost methods for single-cell genome sequencing are developed.
"Finally, our study further motivates the development of genome-editing strategies that do not generate double-stranded DNA breaks, which, in principle, should minimize the potential for inducing chromothripsis," the authors wrote.
"A number of newer methods, such as base editing, act through mechanisms that are independent of DSBs and therefore offer certain safety advantages for either repairing or shutting down genes," St. Jude's Mitchell Weiss, the study's co-corresponding author, said in an email. "However, every method for genetic modification carries some risk for genotoxicity. Defining the risk versus benefit of different gene modification strategies used to treat different genetic diseases requires extensive new laboratory and clinical research."