Abstract
Point mutations cause many genetic disorders, but modelling them in organisms is technically challenging. Creating mouse models that mimic these mutations is crucial for establishing a causal relationship between mutations and disease phenotype, thereby supporting the development of therapeutic strategies. Adenine base editors (ABEs) can correct single-nucleotide variants (SNVs) in disease modelling without double-stranded breaks (DSBs) or donor DNA, achieving higher product purity than traditional Cas9 methods. Earlier ABE techniques faced issues like limited targetability, bystander editing, and off-target effects. By combining two editor advancements, we introduced and tested ABE9-SpRY, an improved ABE variant fused with a PAM-flexible SpRY-Cas9 nickase. Our results show that ABE9-SpRY effectively generates three out of four targeted A-to-G mutations in mouse embryos, achieving desired editing efficiencies of up to 96% in individual adult founder mice. Furthermore, we observe fewer off-target events at predicted DNA sites in mouse embryos and in an orthogonal R-loop assay compared with ABE8e-SpRY. ABE9-SpRY also enhances product purity in mouse embryos under pooled sgRNA injections and, as a proof-of-concept, at a single endogenous locus in human induced pluripotent stem cells (hiPSCs), relative to ABE8e-SpRY. Our findings support ABE9-SpRY's precision at the loci tested and PAM-flexible versatility. Although performance remains sequence-dependent, these data support ABE9-SpRY as a PAM-flexible tool for generating precise point-mutation models where bystander editing is a concern.