Peripherally administered androgen receptor-targeted antisense oligonucleotide rescues spinal pathology in a murine SBMA model

Abstract

Degeneration of the neuromuscular system is a characteristic feature of spinal and bulbar muscular atrophy (SBMA), a CAG/polyglutamine (polyQ) expansion disorder caused by mutation in the androgen receptor (AR). Using a gene-targeted mouse model of SBMA, AR113Q mice, we demonstrate age-dependent degeneration of the neuromuscular system that initially manifests with muscle weakness and atrophy and progresses to include denervation of neuromuscular junctions and lower motor neuron soma atrophy. Using this model, we tested the hypothesis that therapeutic intervention targeting skeletal muscle during this period of disease progression arrests degeneration of the neuromuscular system. To accomplish this, AR-targeted antisense oligonucleotides were administered subcutaneously to symptomatic AR113Q mice to reduce expression of polyQ AR in peripheral tissues but not in the spinal cord. This intervention rescued muscle atrophy, neuromuscular junction innervation, lower motor neuron soma size, and survival in aged AR113Q mice. Single-nucleus RNA sequencing revealed age-dependent transcriptional changes in the AR113Q spinal cord during disease progression, which were mitigated by peripheral AR gene silencing. Our findings underscore the intricate interplay between peripheral tissues and the central nervous system in SBMA and emphasize the therapeutic effectiveness of peripheral gene knockdown in symptomatic disease.

Keywords: Neurodegeneration; Neuromuscular disease; Neuroscience; Therapeutics.

Figure 1. AR113Q males exhibit skeletal muscle atrophy and weakness at 26 weeks.

(A) TA muscle fibers from WT or AR113Q males were visualized by FITC–wheat germ agglutinin (WGA). Nuclei were stained with DAPI (blue). Scale bar: 50 μm. (B and C) Fiber size quantified as a histogram of frequency distribution (B) and box plot (C). n = 3 mice per genotype, >100 fibers per mouse. In C, the box is the interquartile range, the center line is the median, and the whiskers are the minimum and maximum values. ****P < 0.0001 by unpaired Student’s t test, F = 5.314, degrees of freedom (df) = 1,627. (D) Grip strength of WT and AR113Q mice (WT, n = 8; AR113Q, n = 9). Data are mean ± SEM. ****P < 0.0001 by unpaired Student’s t test, F = 11.07, df = 7. (E) Body weight of WT and AR113Q mice (WT, n = 9; AR113Q, n = 9). Data are mean ± SEM. ****P < 0.0001 by unpaired Student’s t test, F = 1.642, df = 8.

Figure 2. AR113Q males exhibit no significant spinal cord or NMJ pathology at 26 weeks.

(A) Immunohistochemistry of spinal cord lumbar enlargement. NeuN is shown in red, MBP in green. White line designates area of neuron quantification. Scale bar: 250 μm. (B) Large neurons in anterior spinal cord lumbar enlargement. NeuN is shown in red, DAPI in blue. Scale bar: 25 μm. (C) Density of large neurons in anterior spinal cord lumbar enlargement (WT, n = 6; AR113Q, n = 5). Data are mean ± SEM. ns, not significant by unpaired Student’s t test; F = 3.372, df = 5. (D) NMJs in TA muscle were visualized by immunofluorescence staining for α-bungarotoxin (red) and neurofilament plus synaptophysin (green). Representative images of NMJs exhibiting full innervation (colocalization ≥50%) and partial innervation (colocalization >20% to <50%). Scale bar: 10 μm. (E and F) NMJ innervation quantified as a box plot (E) or stacked bar graph (F). n = 3 mice per genotype, 100 NMJs per mouse. In E, the box is the interquartile range, the center line is the median, and the whiskers are the minimum and maximum values. NS, not significant by unpaired Student’s t test; F = 1.394, df = 299.

Figure 3. Peripheral ASO administration from 26 to 52 weeks rescues survival and ameliorates skeletal muscle atrophy.

AR113Q males at 26 weeks were given ASO (25 mg/kg body weight) or vehicle subcutaneously, once per week until 52 weeks. (A) TA muscle fibers from WT, AR113Q, or AR113Q plus ASO males were visualized by FITC-WGA. Nuclei were stained with DAPI (blue). Scale bar: 50 μm. (B and C) TA fiber size quantified as a histogram of frequency distribution (B) and box plot (C). In C, the box is the interquartile range, the center line is the median, and the whiskers are the minimum and maximum values. n = 3 mice per group, >100 fibers per mouse. ****P < 0.0001 by 1-way ANOVA with Tukey’s multiple-comparison test, F = 424.7, df = 2. (D) Survival curve of WT, AR113Q, and AR113Q plus ASO mice (WT, n = 15; AR113Q, n = 11; AR113Q + ASO, n = 15). NS, not significant; *P < 0.05, **P < 0.01 by log-rank test with Bonferroni’s correction, χ² = 16.07, df = 2. (E) Grip strength of WT, AR113Q, and AR113Q plus ASO mice at 52 weeks (WT, n = 7; AR113Q, n = 8; AR113Q + ASO, n = 8). Data are mean ± SEM. ***P < 0.001, ****P < 0.0001 by 1-way ANOVA with Tukey’s multiple-comparison test, F = 18.72, df = 2. (F) Body weight of WT, AR113Q, and AR113Q plus ASO mice at 52 weeks (WT, n = 7; AR113Q, n = 8; AR113Q + ASO, n = 8). Data are mean ± SEM. ****P < 0.0001 by 1-way ANOVA with Tukey’s multiple-comparison test, F = 39.40, df = 2.

Figure 4. Spinal cord and NMJ pathology of 52-week AR113Q mice is rescued by peripheral ASO administration.

(A) Large neurons in anterior spinal cord lumbar enlargement. NeuN is shown in red, DAPI in blue. Scale bar: 25 μm. (B) Density of large neurons in anterior spinal cord lumbar enlargement from WT (n = 5), AR113Q (n = 5), and AR113Q plus ASO (n = 5) mice. Data are mean ± SEM. ns, not significant; *P < 0.05, **P < 0.01 by 1-way ANOVA with Tukey’s multiple-comparison test, F = 9.38, df = 2. (C–E) NMJs in TA muscle were visualized by immunofluorescence staining for α-bungarotoxin and neurofilament plus synaptophysin. Innervation quantified as a box plot (C) or stacked bar graph (D). In C, the box is the interquartile range, the center line is the median, and the whiskers are the minimum and maximum values. One hundred NMJs per mouse. ns, not significant; ****P < 0.0001 by 1-way ANOVA with Tukey’s multiple-comparison test, F = 82.00, df = 2. (E) Representative images of NMJs exhibiting full innervation (colocalization ≥50%) from WT muscle, denervation (colocalization ≤ 20%) from AR113Q muscle, and partial innervation (colocalization >20% to <50%) from AR113Q plus ASO muscle. Scale bar: 10 μm. (F and G) Relative Ar mRNA by quantitative PCR in TA muscle (F) and spinal cord (G) of WT, AR113Q, and AR113Q plus ASO mice (n = 4 per group). Data are mean ± SEM. NS, not significant; ****P < 0.0001 by 1-way ANOVA with Tukey’s multiple-comparison test. In E, F = 45.43, df = 2; in F, F = 4.347, df =2.

Figure 5. Single-nucleus RNA sequencing of AR113Q lumbar cord.

(A and B) UMAP embeddings of 75,870 nuclei by time point, genotype, and ASO treatment (A) and by cell type (B). Identified cell types include Chat⁺ motor neurons (MN_Chat), 7 families of dorsal excitatory neurons (DE_Cpne4, DE_Prkcg4, DE_Maf, DE_Reln, DE_Rreb1, DE_Sox5, and DE_Megfr11), medial excitatory neurons (ME_Lmx1b), ventral excitatory neurons (VE_Lhx2), 5 families of dorsal inhibitory neurons (DI_Rorb, DI_Adamts5, DI_Cdh3, DI_Pdyn, and DI_Npy), medial inhibitory neurons (MI_Gad2), ventral inhibitory neurons (VI_Slc6a5), astrocytes 1 (AS1), astrocytes 2 (AS2), oligodendrocyte progenitor cells (OPC), oligodendrocytes (OL), microglia (MG), macrophages (MP), T cells (T), meningothelial cells (MEN), ependymal cells (EP), pericytes (PER), and endothelial cells (END). (C) UMAP embedding showing the relative expression of Ar. (D) Box plots showing normalized expression of Ar in each cell type in the 52-week groups. Boxes indicate the interquartile range, center lines are the median, and the whiskers and individual points were drawn by the Tukey’s method.

Figure 6. Temporal-specific transcriptional dysregulation in AR113Q spinal cord cells.

(A and B) Number of downregulated, upregulated, and total differentially expressed genes (DEGs; imputed |EMD| ≥ 0.1 and P_corrected < 0.01) in neurons (A) and glia and other populations (B) at 26 and 52 weeks. (C) UpSet plot showing the number of upregulated and downregulated DEGs in MN_Chat at each time point, along with genes containing androgen-responsive element (ARE) as determined by AR-ChIP-Seq (27) and their overlap among gene sets. Vertical bar graphs display total genes in each set, and horizontal bar graphs display unique and overlapping genes with separate and linked dots on each row. (D) Heatmap of Gene Ontology (GO) biological process analysis of downregulated (top) and upregulated (bottom) DEGs in MN_Chat (left), DE_Cpne4 (middle), and DI_Rorb (right) at 26 and 52 weeks. The top 3 significantly enriched GO terms for each gene set are displayed. Asterisks indicate statistically significant enrichment (adjusted P < 0.05).

Figure 7. Effect of peripheral AR-targeted ASO injection on spinal cord cell types.

(A–C) UMAP embeddings of MN_Chat (A), DE_Cpne4 (B), and DI_Rorb (C). (D–F) Heatmaps with dendrogram showing normalized expression of the respective total DEGs (rows) in MN_Chat (D), DE_Cpne4 (E), and DI_Rorb (F) between 52-week AR113Q and 52-week WT. (G) Heatmap of GO biological process analysis of downregulated (top) and upregulated (bottom) DEGs in MN_Chat, DE_Cpne4, and DI_Rorb between 52-week AR113Q ASO and other groups. The top 3 significantly enriched GO terms of each gene set are displayed. Asterisks indicate statistically significant enrichment (adjusted P < 0.05).