Cas9 has no exonuclease activity resulting in staggered cleavage with overhangs and predictable di- and tri-nucleotide CRISPR insertions without template donor

Genome editing has found applications in almost every field in biology and medicine and has permeated many aspects of our society. In particular, the clustered regularly interspaced short palindromic repeats (CRISPR)—CRISPR-associated nuclease 9 (Cas9) system—has been widely used in genome editing because of its convenience, simplicity, efficiency, and low cost. It is generally thought that Cas9 cleavage generates blunt-ended double-strand breaks (DSBs) at the 3 base pairs upstream (−3 bp) of protospacer adjacent motif (PAM). In addition, the editing outcomes resulted from several competing cellular DSB repair pathways such as canonical or classic non-homologous end joining (cNHEJ) and microhomology-mediated end joining (MMEJ) are random, imprecise, and unpredictable. Only homologous recombination with a donor template (HDR, homology directed repair) is thought to be precise. However, recent studies revealed that the CRISPR-Cas9 genome editing system is precise and predictable even without donor templates for natural or synthetic targeting sites 1 , 2 , 3 , 4 , 5 , 6 . Specifically, it is firmly established that Cas9-mediated 1-bp insertions are predictable 1 – 3 , 7 , 8 , 9 because they are resulted from ligation and polymerase-fill-in of DSB ends with 1-bp overhang from the staggered Cas9 cleavage at the −4 bp position of the noncomplementary strand upstream of the NGG PAM site 1 . However, whether Cas9 mediates 2-bp or 3-bp insertions in a template-free manner and whether they are predictable remain obscure.