5 January 2021

January 2021

A cell cycle-dependent CRISPR-Cas9 activation system based on an anti-CRISPR protein shows improved genome editing accuracy

CRISPR is a modern and widely used genome editing tool, known for its fast and easy use. However, the efficiency with which precise modifications can be introduced into the DNA, by a pathway called homology-directed repair, is still limited. Homology-directed repair competes with the more efficient non-homologous end-joining pathway resulting in random mutations. Additionally, unintended mutations at DNA regions similar to the target site, called off-target effects, can occur and limit the safe use of CRISPR especially in clinical applications.

The homology-directed repair has been described to occur during the S- and G2- phase of the cell cycle, while non-homologous end-joining is active during all phases. Timed delivery of the CRISPR machinery to synchronized cells can increase precise genome editing. However, drugs synchronizing cells can have additional, potentially toxic, effects. Here the authors present a cell cycle-dependent CRISPR-Cas9 activation system, limiting genome editing activity in normally cycling cells to the S- and G2- cell cycle phase. The system consists of Cas9, the core protein of the CRISPR machinery, and an AcrIIA4-Cdt1 fusion protein. The anti-CRISPR protein AcrIIA4 naturally occurs in bacteriophages and inhibits Cas9 activity. The protein Cdt1 plays a role in controlling correct DNA replication and is degraded in the S/G2 phases of the cell cycle. Combining AcrIIA4 with a fragment of Cdt1 results in a fusion product which inhibits the gene-editing activity of the CRISPR machinery during the G1 phase. Degradation of the fusion protein during the S/G2 phases of the cell cycle prevents inhibition of Cas9 and thus allows active genome editing. As a result, CRISPR editing is limited to the S/G2 phase of the cell cycle, where homology-directed repair is most efficient.

The new cell cycle-dependent CRISPR-Cas9 activation system showed improved efficiency for precise genome editing, compared to the use of non-regulated CRISPR-Cas9 active in all cell cycle phases.  The homology-directed repair efficiency was increased in applications using double- as well as single-stranded DNA repair templates, with differences being more pronounced for the latter. Although the exact fold-changes depended on the genomic locus, positive effects could be demonstrated for all genes tested. Random mutations, introduced by non-homologous end joining, were decreased when Cas9 was co-expressed with AcrIIA4-Cdt1. Overall ratios of homology-directed repair to non-homologous end-joining were increased for all genes tested. Analysis of DNA regions similar to the target site showed decreased off-target activity for the cell cycle-dependent CRISPR-Cas9 activation system. It was also shown that off-target effects can be further reduced by combining the cell cycle-dependent CRISPR-Cas9 activation system with improved design of single-guide RNAs.

Overall, the authors show that the cell cycle-dependent CRISPR-Cas9 activation system improves precise genome editing, by increasing homology-directed repair and decreasing non-homologous end joining at the target site. Simultaneously, off-target activity is reduced. The authors also demonstrate that the cell cycle-dependent CRISPR-Cas9 activation system can be combined with other strategies to improve precise gene editing, such as the alternative design of sgRNAs.