CRISPR-Cas9 induces large structural variants at on-target and off-target sites in vivo that segregate across generations

Basics about changes to off-target and on-target sites

It is known from previous studies on mice and human cell lines that the use of CRISPR/Cas gene scissors can cause unintended changes in DNA at off-target sites. Off-target sites are very similar to the target sequence, which is why the gene scissors can also cut and induce alterations at these sites. Until now, only small changes, such as point mutations or short insertions and deletions (InDels) at off-target sites, have been observed. These small changes can also occur at the target sequence, sometimes called the on-target site. Larger structural changes have also been found at these locations, e.g. deletions of large parts of DNA sequences or newly inserted DNA sequences.

Results of the Höijer et al. study

Until now, it was unclear whether such large structural changes as described at on-target sites could also occur at off-target sites. The scientists from the present study tested this in zebrafish. They firstly analysed the genome of unmodified fish using modern sequencing methods in order to develop suitable guides (also called guide RNAs or gRNAs) for the gene scissors. Several different gRNAs were used. The overall aim of their experiments was to investigate possible structural changes at off-target sites, they therefore used gRNAs with the gene scissors that were very likely to cut at those sites. This allowed them to analyse changes caused at these DNA regions in more detail.

In various experiments, the scientists injected the gene scissors as a protein complex into zebrafish eggs, and examined both the fish that had developed from the eggs and their offspring using sequencing methods. It was found that the efficiency with which changes are obtained at the target sequence is high overall. Predominantly small mutations and, to a lesser extent, larger structural changes (large InDels) were induced.

Investigation of the DNA sequence at possible off-target sites revealed that unintended alterations had occurred at three of them. Their analysis showed that mainly small changes such as point mutations had been induced. However, they also observed larger changes in the DNA sequence. At one off-target site, for example, 903 base pairs (i.e. the letters of the DNA) were deleted, thus affecting a large part of a gene that was not supposed to be changed.

The major structural changes at on- and off-target sites were found in both the founder animals and their offspring.

Relevance of the results

The results are highly relevant for clinical applications of CRISPR/Cas to avoid unexpected effects with major consequences for patients. In plant applications, however, no one has yet investigated whether large structural changes occur at off-target sites. However, it is likely, as it has also been reported at on-target sites.

The design of CRISPR/Cas experiments is crucial for reducing unintended alterations at off-target sites. For example, this requires an appropriate gRNA design so that as few as possible off-target sites are affected, and a well-tuned concentration of the gene scissors used.

Overall, it is important that the genome of the target organisms at on- and off-target sites is examined in detail after CRISPR/Cas experiments to detect unintended changes that may affect other gene regions.