Robust genome editing with broad-targeting CRISPR enzymes

Abstract

Programmable CRISPR enzymes are powerful and versatile tools for genome editing. They, however, require a specific protospacer adjacent motif (PAM) flanking the target site, which constrains the accessible sequence space for position-specific genome editing applications, such as base editing and precise gene insertion. For example, the standard Cas9 from Streptococcus pyogenes (SpyCas9) requires a PAM sequence of 5'-NG̲G̲-3' downstream of its RNA-programmed target, which limits genome editing applications to around 10% of all DNA sequences. To broaden the targeting range of CRISPR, we first bioinformatically discover and characterize a highly similar SpyCas9 homolog from Streptococcus canis (ScCas9) with a more minimal 5'-NNG̲-3' PAM specificity. Furthermore, we employ motifs from closely-related Streptococcus orthologs to engineer an optimized variant of Sc- Cas9 (Sc++) that simultaneously exhibits broadened targeting capability, robust DNA cleavage activity, and minimal off-targeting propensity. Next, we recombine the PAM-interacting domain of Streptococcus macacae Cas9 (SmacCas9) with SpyCas9, and subsequently introduce enhancing mutations to generate iSpyMac with altered and efficient 5'-NA̲A̲-3' PAM preference. Together, these efforts expand the range of CRISPR nucleases to over 70% of DNA sequences, allowing for targeting of genomic loci that were previously inaccessible, including sequences within candidate genes for denser CRISPR screens and disease-related mutations that can now be fixed with genome editing architectures expressing our engineered variants.