Prevalent integration of genomic repetitive and regulatory elements and donor sequences at CRISPR-Cas9-induced breaks

Abstract

Listen

Translate article

CRISPR-Cas9 genome editing has been extensively applied in both academia and clinical settings, but its genotoxic risks, including large insertions (LgIns), remain poorly studied due to methodological limitations. This study presents the first detailed report of unintended LgIns consistently induced by different Cas9 editing regimes using various types of donors across multiple gene loci. Among these insertions, retrotransposable elements (REs) and host genomic coding and regulatory sequences are prevalent. RE frequencies and 3D genome organization analysis suggest LgIns originate from randomly acquired genomic fragments by DNA repair mechanisms. Additionally, significant unintended full-length and concatemeric double-stranded DNA (dsDNA) donor integrations occur when donor DNA is present. We further demonstrate that phosphorylated dsDNA donors consistently reduce large insertions and deletions by almost two-fold without compromising homology-directed repair (HDR) efficiency. Taken together, our study addresses a ubiquitous and overlooked risk of unintended LgIns in Cas9 editing, contributing valuable insights for the safe use of Cas9 editing tools.