CRISPR-Cas9 genome editing induces megabase-scale chromosomal truncations

• DOI: 10.1038/s41467-019-09006-2
• Publication Date: 2019-03-08T11:03:45Z
• Authors: Grégoire Cullot; Julian Boutin; Jérôme Toutain; Florence Prat; Perrine Pennamen; Caroline Rooryck; Martin Teichmann; Emilie Rousseau; Isabelle Lamrissi...; Véronique Guyonne...; Alice Bibeyran; Magalie Lalanne; Valérie Prouzet-M...; Béatrice Turcq; Cécile Ged; Jean-Marc Blouin; Emmanuel Richard; Sandrine Dabernat; François Moreau-G...; Aurélie Bedel
• Keywords: 570; Science; Porphyria, Erythropoietic; [SDV.BBM]Life Sciences [q-bio]/Biochemistry; Primary Cell Culture; 610; RNA, Guide, CRISPR-Cas Systems; Models, Biological; Article; 03 medical and health sciences; CRISPR-Associated Protein 9; Deoxyribonuclease I; Humans; [SDV.BBM]Life Sciences [q-bio]/Biochemistry, Molecular Biology; Clustered Regularly Interspaced Short Palindromic Repeats; DNA Breaks, Double-Stranded; Molecular Biology; Gene Editing; 0303 health sciences; Chromosomes, Human, Pair 10; Genome, Human; Q; High-Throughput Nucleotide Sequencing; DNA; Genetic Therapy; Fibroblasts; 3. Good health; HEK293 Cells; CRISPR-Cas Systems; Chromosome Deletion; K562 Cells

AbstractCRISPR-Cas9 is a promising technology for genome editing. Here we use Cas9 nuclease-induced double-strand break DNA (DSB) at the UROS locus to model and correct congenital erythropoietic porphyria. We demonstrate that homology-directed repair is rare compared with NHEJ pathway leading to on-target indels and causing unwanted dysfunctional protein. Moreover, we describe unexpected chromosomal truncations resulting from only one Cas9 nuclease-induced DSB in cell lines and primary cells by a p53-dependent mechanism. Altogether, these side effects may limit the promising perspectives of the CRISPR-Cas9 nuclease system for disease modeling and gene therapy. We show that the single nickase approach could be safer since it prevents on- and off-target indels and chromosomal truncations. These results demonstrate that the single nickase and not the nuclease approach is preferable, not only for modeling disease but also and more importantly for the safe management of future CRISPR-Cas9-mediated gene therapies.