Background:
The field of genome editing has been revolutionized by the development of an easily programmable editing tool, the CRISPR–Cas9. Despite its promise, off-target activity of Cas9 posed a great disadvantage for genome editing purposes by causing DNA double strand breaks at off-target locations and causing unwanted editing outcomes. Furthermore, for gene integration applications, which introduce transgene sequences, integration of transgenes to off-target sites could be harmful, hard to detect, and reduce faithful genome editing efficiency.

Method:
Here we report the development of a multicolour fluorescence assay for studying CRISPR–Cas9-directed gene integration at an endogenous locus in human cell lines. We examine genetic integration of reporter genes in transiently transfected cells as well as puromycin-selected stable cell lines to determine the fidelity of multiple CRISPR–Cas9 strategies.

Result:
We found that there is a high occurrence of unwanted DNA integration which tarnished faithful knock-in efficiency. Integration outcomes are influenced by the type of DNA DSBs, donor design, the use of enhanced specificity Cas9 variants, with S-phase regulated Cas9 activity. Moreover, restricting Cas9 expression with a self-cleaving system greatly improves knock-in outcomes by substantially reducing the percentage of cells with unwanted DNA integration.

Conclusion:
Our results highlight the need for a more stringent assessment of CRISPR–Cas9-mediated knock-in outcomes, and the importance of careful strategy design to maximise efficient and faithful transgene integration.