Prime Editing of Alzheimer's Disease High-Risk APOE4 Allele by Brain-Directed Adeno-Associated Virus Vectors

Common variants of the apolipoprotein E (APOE) gene have a major impact on the risk of developing Alzheimer's disease (AD). Relative to homozygotes with the common E3 allele, the APOE4 variant (C112R) increases risk by 3.5-fold in E3/E4 heterozygotes and 15-fold in E4 homozygotes. Since the E3 and E4 alleles differ only by a single nucleotide, gene editing of E4 to E3 is a potential strategy to reduce AD risk in E4 homozygotes. Because the APOE pool in the brain is separate from systemic APOE, editing to treat AD would ideally be directed to the brain. Following in vitro optimization of prime editing guide RNAs, efficient prime editing expression cassettes were inserted into the adeno-associated virus (AAV) split-intein system and packaged into pairs of AAV vectors for in vivo editing. The AAV vectors were administered to human homozygous APOE4-targeted replacement mice (TRE4), and APOE4 to APOE3 editing efficiency was assessed after 4 weeks. The prime editing construct designated APOE3/4-3_10 was the most efficient at APOE4 to APOE3 conversion, both in liver following intravenous delivery and in brain following intrahippocampal delivery. To assess brain-wide editing, two AAV capsids were compared, including AAVrh.10 with administration either directly to the hippocampus or to the cerebrospinal fluid via the cisterna magna and AAV-CAP.B10 administered intravenously. Other than minor differences in APOE4/3-3_10 mediated E4 to E3 editing in the cerebellum, the different capsids and routes yielded similar editing efficacy throughout the brain. This may represent a candidate treatment to reduce the risk of AD.