mRNA Nuclear Clustering Leads to a Difference in Mutant Huntingtin mRNA and Protein Silencing by siRNAs In Vivo

Abstract

Huntington's disease (HD) is an autosomal dominant neurodegenerative disease caused by CAG repeat expansion in the first exon of the huntingtin gene (HTT). Oligonucleotide therapeutics, such as short interfering RNA (siRNA), reduce levels of huntingtin mRNA and protein in vivo and are considered a viable therapeutic strategy. However, the extent to which they silence huntingtin mRNA in the nucleus is not established. We synthesized siRNA cross-reactive to mouse (wild-type) Htt and human (mutant) HTT in a divalent scaffold and delivered to two mouse models of HD. In both models, divalent siRNA sustained lowering of wild-type Htt, but not mutant HTT mRNA expression in striatum and cortex. Near-complete silencing of both mutant HTT protein and wild-type HTT protein was observed in both models. Subsequent fluorescent in situ hybridization analysis shows that divalent siRNA acts predominantly on cytoplasmic mutant HTT transcripts, leaving clustered mutant HTT transcripts in the nucleus largely intact in treated HD mouse brains. The observed differences between mRNA and protein levels, exaggerated in the case of extended repeats, might apply to other repeat-associated neurological disorders.

Keywords: Huntington’s disease; mRNA aggregation; nuclear localization; siRNA.

FIG. 1.

HTT mutant and wild-type mRNA but not protein are differentially affected by di-siRNA treatment in BAC-CAG mice after 1 month. (A) Sixteen-week-old BAC-CAG mice were injected ICV with 10 µL of 20 nmol di-siRNA targeting either HTT or a NTC (n = 6). BAC-CAG mice express both transgenic human mutant HTT with 120 CAG repeats and endogenous mouse wild-type Htt with 7 CAG repeats. Mice were sacrificed 1 month later at 20 weeks old. After sacrifice striatum and cortical tissue were tested for changes HTT mRNA (Quantigene) and protein (automated simple western) expression. (B) Human mutant HTT and wild-type mouse Htt mRNA levels were measured by Quantigene and normalized to Hprt as a housekeeping gene. mRNA levels are displayed as a percentage of their respective NTC value. For comparisons between groups, a one-way ANOVA was used with multiple comparisons. For comparisons between siNTC- and siHTT-treated tissue, P < 0.01 and P ≤ 0.001 for wild-type Htt in the striatum and cortex, respectively. For mutant HTT, P = 0.18 in the striatum and cortex. (C) Human mutant and wild-type mouse HTT protein levels were measured by automated simple western and normalized to vinculin as a loading control. Protein levels are displayed as a percentage of their respective NTC value. For comparisons between groups, a one-way ANOVA was used with multiple comparisons. For comparisons between siNTC- and siHTT-treated tissue, P < 0.05 and P < 0.001 for wild-type HTT in the striatum and cortex, respectively. For mutant HTT, P < 0.05 and P < 0.001 in the striatum and cortex, respectively. ICV, intracerebroventricular; NTC, nontargeting control; siRNA, short interfering RNA.

FIG. 2.

HTT mutant and wild-type mRNA but not protein is differentially affected by di-siRNA treatment in YAC128 at 6 months postinjection. (A) Eight-week-old YAC128 mice were injected ICV with 10 µL of 20 nmol di-siRNA targeting either HTT or a NTC (n = 6). YAC128 mice express both transgenic human mutant HTT with 128 mixed CAA/CAG repeats and endogenous mouse wild-type Htt with 7 CAG repeats. Mice were sacrificed 24 weeks later at 32 weeks old. After sacrifice, striatum and cortical tissue were tested for changes HTT mRNA (Quantigene) and protein (automated simple western) expression. (B) Human mutant HTT and wild-type mouse Htt mRNA levels were measured by Quantigene and normalized to Hprt as a housekeeping gene. mRNA levels are displayed as a percentage of their respective NTC value. For comparisons between groups, a one-way ANOVA was used with multiple comparisons. For comparisons between siNTC- and siHTT-treated tissue, P < 0.0001 and P < 0.05 for wild-type Htt in the striatum and cortex, respectively. For mutant HTT, P < 0.05 and P = 0.32 in the striatum and cortex, respectively. (C) Human mutant and wild-type mouse HTT protein levels were measured by automated simple western and normalized to vinculin as a loading control. Protein levels are displayed as a percentage of their respective NTC value. For comparisons between groups, a one-way ANOVA was used with multiple comparisons. For comparisons between siNTC- and siHTT-treated tissue, P < 0.0001 for wild-type HTT in the striatum and cortex. For mutant HTT, P < 0.0001 in the striatum and cortex.

FIG. 3.

Nuclear mutant but not wild-type huntingtin mRNA fraction is resistant to siRNA-based silencing in striatum of BAC-CAG mice (A) 16-week-old BAC-CAG mice were injected ICV with 10 µL of 20 nmol di-siRNA targeting either HTT or a NTC (n = 6). BAC-CAG mice express both transgenic human mutant HTT with 120 CAG repeats and endogenous mouse wild-type Htt with 7 CAG repeats. Mice were sacrificed one month later at 20 weeks old. After sacrifice striatum and cortical tissue were tested for changes HTT mRNA (Quantigene), protein (automated simple western) expression (Fig. 1), and mRNA localization (RNA FISH). (B) Representative FISH images showing NTC-treated (upper) and siHTT-treated (lower) BAC-CAG striatum tissue. Probes measure mutant HTT (right, green) and wild-type Htt (middle, red). Left images show a merge of both probes plus DAPI staining cell nuclei (blue). White arrows denote intranuclear clustering. (C, D) Quantification of FISH images (n = 3 animals) for mutant HTT and wild-type Htt in different compartments. All statistical tests were performed using a mixed effects model and analyzed using pairwise comparisons contrasting the impact of siRNA treatment. (B) Counts of Htt foci in cellular compartments with siNTC or siHTT treatment (****P < 0.0001). (C) Counts of HTT foci in cellular compartments with NTC or siHTT treatment (*P < 0.05, ***P < 0.001). FISH, fluorescent in situ hybridization.