Fig. 1

a sgRNAs were designed to target exon 10 of the MAPT. Protospacer sequences are underlined, PAM sequences are shown in green, the targeted nucleotide in red, and base edited nucleotides in blue. b Adenine base editing frequencies induced by tsAAV-NG-ABE8e-MAPT, which is controlled by the hSyn-1 promoter, in the hippocampus of PS19 mice at eight weeks after intracranial injection. Error bars indicate SEM (n = 4). c Genomic DNA isolated from the hippocampi of PS19 mice at eight weeks after injection of tsAAV-NG-ABE8e-MAPT was subjected to targeted deep sequencing. Mismatched nucleotides are shown in red, PAM sequences in blue, and DNA bulge on green. ON, on-target site; OT, off-target site. Error bars indicate SEM (n = 4). d Representative western blots using anti-tau antibodies and quantification of immunoblot staining. All antibodies used in this study are described in Table S4. *P < 0.05 vs mock control, Student’s t-test. Error bars indicate SEM (n = 6). e Representative images and quantification of anti-phospho-tau (AT8) staining in the hippocampus. *P < 0.05 vs mock control, Student’s t-test. Scale bar, 100 µm; n = 5–6. f Representative images of double-staining with anti-phospho-tau (AT8) and anti-HA-tag (which recognizes the NG-ABE8e fusion protein) antibodies in the hippocampus. Scale bars, 100 µm. g The escape latency during the training phase in the water maze test. #P < 0.05 vs WT, *P < 0.05 vs mock control; generalized estimating equation analysis. WT, n = 12; Mock, n = 11; MAPT, n = 10. h The latency to enter the dark compartment in the passive avoidance test (PAT). Maximum time was 300 s. #P < 0.05 vs WT, *P < 0.05 vs mock control, One-way ANOVA. WT, n = 12; Mock, n = 12; MAPT, n = 12