AAV-mediated base-editing therapy ameliorates the disease phenotypes in a mouse model of retinitis pigmentosa

Abstract

Base editing technology is an ideal solution for treating pathogenic single-nucleotide variations (SNVs). No gene editing therapy has yet been approved for eye diseases, such as retinitis pigmentosa (RP). Here, we show, in the rd10 mouse model, which carries an SNV identified as an RP-causing mutation in human patients, that subretinal delivery of an optimized dual adeno-associated virus system containing the adenine base editor corrects the pathogenic SNV in the neuroretina with up to 49% efficiency. Light microscopy showed that a thick and robust outer nuclear layer (photoreceptors) was preserved in the treated area compared with the thin, degenerated outer nuclear layer without treatment. Substantial electroretinogram signals were detected in treated rd10 eyes, whereas control treated eyes showed minimal signals. The water maze experiment showed that the treatment substantially improved vision-guided behavior. Together, we construct and validate a translational therapeutic solution for the treatment of RP in humans. Our findings might accelerate the development of base-editing based gene therapies.

Fig. 1. Design and optimization of the ABE8 system for rd10 SNV correction.

a Dual-AAV system of SpRY-ABE8e using a split-intein mechanism and sgRNA design targeting the rd10 pathogenic SNV. b Schematic representation of rd10 pathogenic SNV correction by SpRY-ABE8e. c BE efficiency of rd10 pathogenic SNV by dual-AAV plasmids of SpRY-ABE8e. Bars represent the average editing specificity, and error bars represent the s.d. of three independent biological replicates. Source data are provided as a Source Data file.

Fig. 2. Dual-AAV-mediated BE therapy in rd10 mice.

a Flowchart of in vivo ABE treatment. b In vivo genomic DNA and cDNA BE efficiencies in rd10 mice. c Comparison of average editing efficiencies of genomic DNA and cDNA. d Endogenous genome off-target efficiencies after SpRY-ABE8e treatment. Two-tailed unpaired t-tests; adjustment was not made for multiple comparisons. For all plots, the dots represent individual biological replicates, bars represent the average editing specificity, and error bars represent the s.d. of independent biological replicates. BSS treated (n = 6), NT treated (n = 6), and A7 treated (n = 7). Source data are provided as a Source Data file.

Fig. 3. Restoration of PDE6B protein in rd10 mice after ABE treatment.

a Western blot analysis of ABE (left) and PDE6B (right) expressions in mouse neuroretina lysate after treatment at P35. b Immunofluorescence analysis of representative eye cross-sections at P35. Blue indicates DAPI, green indicates HA tag (i.e., ABE), and red indicates PDE6B. OS, outer segments; IS, inner segments; ONL, outer nuclear layer; INL, inner nuclear layer. Source data are provided as a Source Data file.

Fig. 4. Photoreceptor preservation in rd10 mice after ABE treatment.

a Representative eye section of an A7-treated rd10 mouse with H&E staining at P35. ONL, outer nuclear layer; INL, inner nuclear layer; and GCL, ganglion cell layer. b Quantification of ONL thickness in DAPI nuclei-stained retinal cryosections of WT (n = 4 eyes), BSS treated (n = 4 eyes), NT treated (n = 4 eyes), and A7 treated (n = 4 eyes) mice at P35. ONH, optic nerve head. Data are presented as means ± s.d. c Immunofluorescence analysis of representative retinal sections at P35. Blue indicates DAPI and red indicates Rhodopsin. OS, outer segments. d Quantification of rod OS length at 1 mm temporal of the optic nerve head of WT (n = 4 eyes), BSS treated (n = 4 eyes), NT treated (n = 4 eyes), and A7 treated (n = 4 eyes) mice at P35. Means ± s.d. are shown. Source data are provided as a Source Data file.

Fig. 5. Rescue of retinal function in rd10 mice after ABE treatment.

a Schematic of in vivo electroretinography settings. b Representative scotopic ERG signals of WT, BSS treated, NT treated, and A7 treated eyes at P35. c Quantification of scotopic a-wave amplitudes from each group at P35. *P < 0.05, **P < 0.01, ***P < 0.001, two-way ANOVA tests with Tukey’s multiple comparisons. Asterisks indicate significant differences between A7 treated and NT treated mice. P values are 0.0292 at −0.5, 0.0024 at 0, and <0.0001 at higher light intensities. d Quantification of scotopic b-wave amplitudes from each group at P35. *P < 0.05, ***P < 0.001, two-way ANOVA tests with Tukey’s multiple comparisons. Asterisks indicate significant differences between A7 treated and NT treated mice. P values are 0.0294 at −2 and <0.0001 at higher light intensities. The numbers of eyes were as follows: WT, n = 10; BSS treated, n = 10; NT treated, n = 10; and A7 treated, n = 12. Means ± s.d. are shown. Source data are provided as a Source Data file.

Fig. 6. Improvement of vision-guided behavior of rd10 mice after ABE treatment.

a Representative swimming routes on day 4 from each group at P35. b Quantification of the success rate to locate the platform within 1 min from day 1 to day 4. ***P < 0.001, two-way ANOVA tests with Tukey’s multiple comparisons. Asterisks indicate a significant difference between A7 treated and NT-treated mice at day 4. P value is 0.0004. c Quantification of the escape latency from day 1 to day 4. ***P < 0.001, two-way ANOVA tests with Tukey’s multiple comparisons. Asterisks indicate a significant difference between A7 treated and NT-treated mice at day 4. P value is <0.0001. d Quantification of the total path length from day 1 to day 4. ***P < 0.001, two-way ANOVA tests with Tukey’s multiple comparisons. Asterisks indicate a significant difference between A7-treated and NT-treated mice at day 4. P value is <0.0001. The numbers of mice were as follows: WT, n = 5; BSS treated, n = 5; NT treated, n = 5; and A7 treated, n = 7. Each mouse completed 8 trials on day 1 to day 3 and 4 trials on day 4. Data are shown as the means ± sem. Source data are provided as a Source Data file.