Endogenous sequence patterns predispose the repair modes of CRISPR/Cas9‐induced DNA double‐stranded breaks in Arabidopsis thaliana

Summary The possibility to predict the outcome of targeted DNA double‐stranded break (DSB) repair would be desirable for genome editing. Furthermore the consequences of mis‐repair of potentially cell‐lethal DSBs and the underlying pathways are not yet fully understood. Here we study the clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9‐induced mutation spectra at three selected endogenous loci in Arabidopsis thaliana by deep sequencing of long amplicon libraries. Notably, we found sequence‐dependent genomic features that affected the DNA repair outcome. Deletions of 1‐bp to 1 kbp (all due to NHEJ) and deletions combined with insertions between 5‐bp to >100 bp [caused by a synthesis‐dependent strand annealing (SDSA)‐like mechanism] occurred most frequently at all three loci. The appearance of single‐stranded annealing events depends on the presence and distance between repeats flanking the DSB. The frequency and size of insertions is increased if a sequence with high similarity to the target site was available in cis. Most deletions were linked to pre‐existing microhomology. Deletion and/or insertion mutations were blunt‐end ligated or via de novo generated microhomology. While most mutation types and, to some degree, their predictability are comparable with animal systems, the broad range of deletion mutations seems to be a peculiar feature of the plant A. thaliana. Significance Statement Little information is known of what effect the target sequence has on the result of genome editing. We found that the choice of repair pathways and the outcome regarding size and frequency of deletions and insertions at targeted double‐stranded breaks depends on the sequence context (e.g. the presence and arrangement of repeats), thus having the potential to predict the results of genome editing in Arabidopsis thaliana.