Efficient base editing with expanded targeting scope using an engineered Spy-mac Cas9 variant

Dear Editor, The clustered regularly interspaced short palindromic repeat (CRISPR) system, including newly developed base editing technology, has exhibited powerful genome manipulation capability. Base editors that can achieve targeted C-to-T (CBE) or A-to-G (ABE) conversions without generating DNA double-strand breaks (DSBs) or requiring a donor template represent significant advances in both disease modeling and gene therapy. However, the conventional Streptococcus pyogenes Cas9 (Spy Cas9) requires a protospacer adjacent motif (PAM) of NGG, which limits the applicability of base editors that are highly dependent on the PAMs suitably adjacent to target bases. In addition, although some natural and engineered Cas9 variants with different PAM specificities have been utilized in base editors, such as representative Cpf1 (TTTV (V=A/G/C)) and SpCas9-NG (NG), their targeting scope is still limited to genomic regions rich in G or T bases. Recently, Spy-mac Cas9 was generated by rationally exchanging the PAM-interacting (PI) region of the conventional Spy Cas9 with that of the newly discovered Streptococcus macacae Cas9 (Smac Cas9), showing 5′-NAA-3′ PAM specificity and possessing efficient gene editing in human cells. In this study, we demonstrated the effectiveness of the Spy-mac Cas9assisted cytidine and adenine base editors Spy-mac BE4max and Spy-mac ABEmax, and found that 5′-TAAA-3′ is the only high-efficiency PAM for Spy-mac Cas9 observed in this study. To obtain the best efficiency of base editing, the Spymac Cas9 system was combined with the current optimal version of the base editors BE4max and ABEmax to generate Spy-mac BE4max and Spy-mac ABEmax, respectively (Fig. 1a). To fully evaluate the PAM specificity and editing efficiency, we first tested the Spy-mac BE4max system in rabbit embryos at 16 target sites, including all NAAN PAMs, as a proof of concept (Supplementary Table S1). Base editing was conducted in rabbit pronuclear-stage embryos by microinjecting of Spy-mac BE4max-encoding mRNA and single-guide RNA (sgRNA). Base editing frequencies were evaluated by Sanger sequencing and T-A cloning. Notably, the average C-to-T editing frequency at the Tyr-1 site with TAAA PAM was high at 86.00 ± 8.72%, while much lower efficiencies ranging from 13.33 ± 6.67% to 26.67 ± 11.74% were observed at the other five sites with AAAT, GAAG, CAAG, AAAC, and CAAC PAMs (Fig. 1b and Supplementary Fig. S1). However, no obvious base editing events were observed at most tested sites (10/16), consistent with variations in the targeting efficiencies of Spy-mac Cas9 at different targeting sites in human cells (Fig. 1b and Supplementary Fig. S1). Encouraged by the results of the pilot study, we examined whether the Spy-mac Cas9 variant may primarily target TAAA PAM in rabbit embryos. Therefore, another three sites (Tyr-5, Mstn-3, and Dmd-2) with TAAA PAMs were designed to verify our hypothesis (Supplementary Table S1). Remarkably, all sites showed efficient C-to-T conversions, with average editing frequencies ranging from 28.00 ± 13.93% at Dmd-2 to 100.00 ± 0.00% at Tyr14 (Fig. 1c and Supplementary Fig. S2). Moreover, targeted C·G to T·A conversions were successfully achieved to induce stop codons at all four sites, as expected (Fig. 1d, g). In particular, the homozygous p.W178Stop-