In vivo precision base editing to rescue mouse models of disease.

CRISPR base editing enables precise, irreversible base conversions without inducing double-stranded breaks (DSBs) and has gained significant attention in recent years. By converting cytosine to thymine (C→T) or adenine to guanine (A→G), base editors (BEs) efficiently correct pathogenic single-nucleotide variants (SNVs). This review examines in vivo mouse disease models-assessing editing efficiency, phenotypic rescue, and therapeutic potential across 66 studies. A key challenge in base editing is optimizing delivery. Most studies rely on split-intein dual adeno-associated virus (AAV) vectors due to BEs exceeding AAV packaging limits, though lipid nanoparticle (LNP) delivery is emerging. Editing efficiencies vary widely, influenced by enzyme design, delivery method, and sequence context. Many studies show significant functional gains, including extended survival in severe models such as FAH-deficient tyrosinemia type I and Hutchinson-Gilford progeria, restored dystrophin in Duchenne muscular dystrophy, and cognitive improvement in neurodegenerative models. Despite advantages such as reduced indels and increased precision, base editing is restricted to SNV correction and targets only a limited editing window relative to a protospacer adjacent motif (PAM) site. Advances in enzyme engineering, delivery strategies, and hybrid approaches incorporating prime editing could broaden its applications. As base editing evolves, its success in preclinical models positions it as a key player in next-generation gene therapies.