Potent and sustained huntingtin lowering via AAV5 encoding miRNA preserves striatal volume and cognitive function in a humanized mouse model of Huntington disease

Abstract

Huntington disease (HD) is a fatal neurodegenerative disease caused by a pathogenic expansion of a CAG repeat in the huntingtin (HTT) gene. There are no disease-modifying therapies for HD. Artificial microRNAs targeting HTT transcripts for degradation have shown preclinical promise and will soon enter human clinical trials. Here, we examine the tolerability and efficacy of non-selective HTT lowering with an AAV5 encoded miRNA targeting human HTT (AAV5-miHTT) in the humanized Hu128/21 mouse model of HD. We show that intrastriatal administration of AAV5-miHTT results in potent and sustained HTT suppression for at least 7 months post-injection. Importantly, non-selective suppression of huntingtin was generally tolerated, however high dose AAV5-miHTT did induce astrogliosis. We observed an improvement of select behavioural and modest neuropathological HD-like phenotypes in Hu128/21 mice, suggesting a potential therapeutic benefit of miRNA-mediated non-selective HTT lowering. Finally, we also observed that potent reduction of wild type HTT (wtHTT) in Hu21 control mice was tolerated up to 7 months post-injection but may induce impairment of motor coordination and striatal atrophy. Taken together, our data suggests that in the context of HD, the therapeutic benefits of mHTT reduction may outweigh the potentially detrimental effects of wtHTT loss following non-selective HTT lowering.

Figure 1.

Overview of study design. (A) Cartoon representation of miHTT expression cassette and target sequence within exon 1 of human HTT. (B) Overview of study design highlighting timeline of AAV5-miHTT injections and endpoints. (C) Overview of cohort design with breakdown of Hu21 and Hu128/21 cohort animal numbers and sex distribution for each of the harvest timepoints.

Figure 2.

Widespread distribution and broad cell-type tropism of AAV5-GFP following bilateral intrastriatal injection in the Hu128/21 humanized model of HD. (A) A series of sections spanning the striatum was stained for GFP to evaluate viral distribution at 9 months of age (7 months post-injection). Tropism was evaluated qualitatively by co-localization of GFP from AAV-miScr treated animals with (B) NeuN, a general marker for neurons, (C) DARPP-32, a marker for medium spiny neurons (MSNs), (D) GFAP, a marker for astrocytes and (E) Iba1, a marker for microglia. (F) Quantification of the % overlap between GFP+ and either NeuN+, GFAP+, Iba1+ and DARPP-32+ cells in the striatum of AAV-miScr treated animals.

Figure 3.

AAV5-miHTT shows potent and sustained suppression of total HTT in Hu128/21 mice. (A) AAV5-miHTT viral vector genome copies in the cortex of Hu128/21 at 7 months post-injection (one way ANOVA P = 0.0007, Tukey's multiple comparison test **P = 0.0021 high dose compared to saline, ^##P = 0.0013 high compared to low dose, ^&P = 0.0446 high compared to medium dose. N = 7–9). LLOQ: Lower limit of quantification, LLOD: lower limit of detection. (B) Total picomoles of mature miHTT miRNA per gram of total RNA in the cortex of Hu128/21 mice at 7 months post-injection (one way ANOVA P< 0.0001, Tukey's multiple comparison test ****P< 0.0001 medium and high dose compared to saline, ^##P = 0.012 medium and ^####P<0.0001 high compared to low dose, ^&&P = 0.0047 high compared to medium dose. N = 7–9). Representative western blot showing dose-dependent HTT lowering in (C) striatum and (E) cortex at 7 months post-injection. Quantification of wt and muHTT levels in the (D) striatum (* for wtHTT and ^# for mHTT. 1 month post-injection two-way ANOVA dose P = 0.0002, HTT P = 0.6134, interaction P = 0.9767. Bonferroni post-tests: * P<0.05 low dose, ^#P<0.05 and **P<0.01 medium dose, ** and ^##P< 0.01 high dose compared to saline, N = 3. 4 months post-injection two way ANOVA dose P<0.0001, HTT P = 0.6882, interaction P = 0.9805. Bonferroni post-tests **P< 0.01 and ^#P< 0.05 medium dose, **P< 0.01 and ^##P< 0.01 high dose compared to saline, N = 4–6. 7 months post-injection two-way ANOVA dose P< 0.0001, HTT P = 0.1712, interaction P = 0.8258. Bonferroni post-tests *** and ^###P< 0.001 low, medium and high dose compared to saline, N = 6–9) and (F) cortex (* for wtHTT and ^# for mHTT, 1 month post-injection two-way ANOVA dose P = 0.0006, HTT P = 0.6582, interaction P = 0.9914. Bonferroni post ^# and * P< 0.05 medium dose, ** and ^##P< 0.01 high dose compared to saline, N = 3. 4 months post-injection two-way ANOVA dose P< 0.0001, HTT P = 0.4490, interaction P = 0.8036. Bonferroni post-tests ^#P<0.05 medium dose, *** and ^###P< 0.001 high dose compared to saline, N = 4–6. 7 months post-injection two-way ANOVA dose P< 0.0001, HTT P = 0.2641, interaction P = 0.8874. Bonferroni post-tests **P< 0.01 and ^#P< 0.05 low dose, ***P< 0.001 and ^##P< 0.01 medium dose, *** and ^##P< 0.001 high dose compared to saline, N = 6–9) at 1, 4 and 7 months post-injection. M.P.I. = months post-injection.

Figure 4.

AAV5-miHTT may ameliorate psychiatric and cognitive phenotypes in Hu128/21 mice. (A) Open field centre time at 1 and 7 months post-injection (mixed-effects model on Hu21 and Hu128/21 saline treated groups at both time points; genotype P = 0.9760, timepoint P< 0.0001, genotype × timepoint P = 0.4023; mixed-effects model on all Hu128/21 treated groups at both timepoints; treatment P = 0.2820, timepoint P<0.0001, treatment x timepoint P = 0.0365. Tukey multiple comparison test: Hu128/21 low versus high dose *P = 0.0227, Hu128/21 miHTT high versus Hu128/21 saline = not significant. 3 months: N = 26–32, 9 months: N = 18–25). (B) Novel object location test at 9 months of age comparing Hu21 saline to Hu128/21 treated groups (unpaired t-test of trial 1 versus trial 2 for Hu21 saline *P = 0.0227, medium dose P = 0.0742 and high dose * P = 0.0182. N = 10–15).

Figure 5.

AAV5-miHTT may prevent striatal volume loss and normalize striatal DARPP-32 levels in Hu128/21 mice. Stereological evaluation of (A) striatal (unpaired t-test between Hu21 and Hu128/21 saline treated P = 0.0517; one-way ANOVA on Hu128/21 treated animals P = 0.465. N = 12–14), (B) cortical (unpaired t-test between Hu21 and Hu128/21 saline treated P = 0.6380; one-way ANOVA on Hu128/21 treated animals P = 0.9741. N = 13–14) and (C) corpus callosum volumes (unpaired t-test between Hu21 and Hu128/21 saline treated P = 0.0623; one-way ANOVA on Hu128/21 treated animals P = 0.3810. N = 13–14) at 7 months post-injection. (D) DARPP-32 immunoreactivity was evaluated by integrated optical density (IOD) in a series of 4 mid-striatal sections at 2 months post-injection (unpaired t-test between Hu21 and Hu128/21 saline treated **P = 0.0316; one-way ANOVA on Hu128/21 treated animals P = 0.0002. Tukey multiple comparison test ^###P< 0.001 Hu128/21 saline versus high, medium and low doses. N = 3) and 7 months post-injection (unpaired t-test between Hu21 and Hu128/21 saline treated P = 0.5222; one-way ANOVA on Hu128/21 treated animals P = 0.7315. N = 13–14). Values were normalized to the mean value for saline treated Hu21 animals at each timepoint.

Figure 6.

AAV5-miHTT shows efficient transduction and potent suppression of wtHTT in Hu21 control mice. (A) AAV5-miHTT viral vector genome copies in the cortex of Hu21 at 7 months post-injection (One way ANOVA P<0.0001. Tukey multiple comparison test ***P< 0.0001 high dose compared to saline, ^###P = 0.0003 high compared to low dose, ^&&P = 0.0044 high compared to medium dose. N = 9–11). LLOQ: lower limit of quantification, LLOD: lower limit of detection. (B) Total picomoles of mature miHTT miRNA per gram of total RNA in the cortex of Hu21 mice at 7 months post-injection (one way ANOVA P< 0.0001. Tukey multiple comparison test *P = 0.0387 medium and ****P< 0.0001 high dose compared to saline, ^####P< 0.0001 high compared to low dose, ^&&&&P< 0.0001 high compared to medium dose. N = 9–11). Representative western blot showing dose-dependent HTT lowering in (C) striatum and (E) cortex at 7 months post-injection. Quantification of wtHTT levels in the (D) striatum (one-way ANOVA P< 0.0001. Tukey's multiple comparison test ****P< 0.0001 saline compared to low, medium and high dose, ^#P = 0.03 medium and ^####P< 0.0001 high compared to low dose, ^&&P = 0.0040 high compared to medium dose. N = 8–10) and (F) cortex (one-way ANOVA P< 0.0001. Tukey's multiple comparison test saline compared **P = 0.0082 medium and ****P< 0.0001 high dose, ^###P = 0.0004 high compared to low dose, ^&P = 0.0326 high compared to medium dose. N = 9–11) at 7 months post-injection.

Figure 7.

AAV5-miHTT treatment may induce motor but not psychiatric or cognitive abnormalities in Hu21 control mice. (A) Accelerating rotarod test of Hu21 treated groups at 2, 4, 6 and 8 months of age (mixed-effects model on all Hu21 groups treatment P = 0.1719, timepoint P< 0.0001, treatment × timepoint P = 0.7945. 2 months: N = 24–26, 4 months: N = 24–26, 6 months: N = 24–25, 8 months: N = 23–25). (B) Open field centre time at 1 and 7 months post-injection (mixed effects model on all Hu21 groups treatment P = 0.2531, timepoint P< 0.0001, treatment × timepoint P = 0.1642. 3 months: N = 24–26, 9 months: N = 17–25). (C) Novel object location test at 9 months of age comparing Hu21 treated groups (unpaired t-test of trial 1 versus trial 2 for saline *P = 0.0227, low dose **P = 0.0034, medium dose *P = 0.0126, high dose **P = 0.0081. Trial 1, N = 21–23; trial 2, N = 15–21).

Figure 8.

AAV5-miHTT treatment may induce striatal atrophy in Hu21 control mice. Stereological evaluation of (A) striatal (one-way ANOVA P = 0.533. N = 13–15), (B) cortical (one-way ANOVA P = 0.5855. N = 13–15) and (C) corpus callosum (one-way ANOVA P = 0.1176. N = 13–15) volumes was performed at 7 months post-injection. (D) IOD of DARPP-32 immunoreactivity was evaluated in a series of 4 mid-striatal sections at 2 months post-injection (one-way ANOVA P = 0.0351. N = 3) and 7 months post-injection (one-way ANOVA P = 0.5430. N = 13–15). Values were normalized to the mean value for saline treated Hu21 animals at each timepoint.