2.1 Progressive glycogen accumulation and muscle weakness in the Gaa^−/− mouse model

We utilised the Gaa^−/− mouse model to study the effects of lowering Gys1 expression on Pompe myopathy in vivo. The Gaa^−/− mouse was generated and studied by Raben et al. (1998).⁴ To inform the design of appropriate prevention and reversal treatment paradigms, we characterised the progressive accumulation of glycogen in the skeletal and cardiac muscle of the Gaa^−/− mice from our colony. We utilised two methods to measure glycogen content in the tissues: a colorimetric biochemical assay, and Periodic acid-Schiff (PAS) histochemical staining.⁴ We observed higher amounts of glycogen in the skeletal muscle of Gaa^−/− mice compared to wild-type mice (Figure 1A). Remarkably, the glycogen levels in quadriceps muscle were already elevated in 1-month-old Gaa^−/– mice, suggesting that the mutant mice start to accumulate glycogen in skeletal muscle at an early age.

The accumulation of glycogen increased progressively over time, from 260 µg (95% confidence interval (C.I.): 167–353 µg), to 595 µg (95% C.I.: 467– 722), and 784 µg (95% C.I.: 575–992) glycogen per 100 mg tissue in quadriceps muscle homogenate from 1-, 2-, and 3-month-old Gaa^−/− mice, respectively (Figure 1A). Similar to previous reports,^{4, 30, 31} the excessive glycogen accumulation in the skeletal muscle was associated with a progressive decline in motor function in aging Gaa^−/− mice, as assessed by the rotarod test (Figure 1B).

2.2 Different levels of ASO-mediated Gys1 knockdown lead to the proportional amelioration of aberrant glycogen accumulation in the skeletal muscle of young Gaa^−/− mice

Through in vitro and in vivo screening, we identified two ASOs targeting mouse Gys1 that were potent and well tolerated in mice (Figures S1 and S2). When dosed systemically to wild-type mice via subcutaneous injection for 6 weeks, Gys1 ASO#1 and ASO#2 resulted in the efficient, dose-dependent knockdown of murine Gys1 mRNA. ASO#1 had an ED₅₀ of 45.2 mg/kg (95% C.I.: 32.3–63.2 mg/kg) in the quadriceps muscle, whereas ASO#2 was more potent, with an ED₅₀ of 24.2 mg/kg (95% C.I.: 22.4–26.1 mg/kg) (Figure S1A). Consistently, Gys1 ASO#2 achieved a robust reduction of GYS1 protein in skeletal muscle of wild-type and Gaa^−/− mice (Figure S1B–D).

To determine if ASO-mediated Gys1 knockdown attenuates the pathological accumulation of glycogen in vivo, we dosed young Gaa^−/− mice with the newly identified Gys1 ASO#1 and ASO#2. A starting age of 1 month was selected because at this time the Gaa^−/− mice have not developed severe myopathy yet, and the glycogen is only starting to accumulate in the skeletal muscle (Figure 1A). The 1-month-old Gaa^−/– mice (Cohort 1) were dosed via subcutaneous injection with 25 mg/kg Gys1 or control ASO, once a week for 16 weeks (Figure 2A). Compared to control ASO, Gys1 ASO#1 and ASO#2 resulted in 83.64% (95% C.I.: 77.40–88.58) and 88.58% (95% C.I.: 83.61–92.06) Gys1 mRNA knockdown, respectively in quadriceps muscle (Figure 2B, Table S1). Corroborating the reduction in mRNA expression, compared to control ASO, ASO#2 administration decreased GYS1 protein, measured by Western blot, in key Pompe tissues of Gaa^−/− mice by 100%, 98%, and 70% in quadriceps, diaphragm, and heart, respectively. (Figure 2C, Table S1).

To evaluate the impact of Gys1 knockdown on excessive glycogen accumulation in the muscle of the Pompe mice, we measured glycogen content by a colorimetric biochemical assay in the muscle of the 20-week-old Gaa^−/− mice following 16 weeks of Gys1 ASO treatment. Compared to the control ASO, treatment with ASO#1 led to a 41.13% reduction in glycogen content in the quadriceps (322 µg/100 mg tissue vs. 501 µg/100 mg tissue) and a 15.12% reduction in the diaphragm (539 µg/100 mg tissue vs. 648 µg/100 mg tissue). In contrast, ASO#2 resulted in a more substantial decrease, with glycogen levels falling by 58.47% in the quadriceps (212 µg/100 mg tissue vs. 501 µg/100 mg tissue) and by 57.07% in the diaphragm (296 µg/100 mg tissue vs. 648 µg/100 mg tissue) (Figure 2D, Table S1). In the heart, we observed variable degree of reduction of GYS1 protein levels, on average the impact of both ASOs was more pronounced in quadriceps and diaphragm (Figure 2C, Table S1). The control ASO had no effect on Gys1 expression and glycogen content (Figure 2B–F).

In line with the biochemical measurements, glycogen content assessed in histological sections using PAS staining confirmed that Gys1 ASOs reduced aberrant glycogen accumulation in the quadriceps muscle of Gaa^−/− mice (Figure 2E). Quantitative analysis of PAS-stained images using Imaris revealed a median intensity of 119 (95% C.I.: 103–135) in 4.5-month-old Gaa^−/− mice treated with control ASO, 74 (95% C.I.: 70–78) in mice treated with ASO#1, and 65 (95% C.I.: 60–70) in mice treated with ASO#2 (Figure 2E and F). In view of the enhanced efficacy in the early-treatment setting, we selected ASO#2 to further investigate if Gys1 ASO treatment could revert the Pompe pathology in older Gaa^−/− mice.

2.3 ASO-mediated Gys1 reduction partially reverses glycogen accumulation in the skeletal muscle of older Gaa^−/− mice

The typical patient with late onset Pompe disease presents in the clinic is in their 30s–40s, when the disease has already progressed. To assess if therapeutic treatment with Gys1 ASO could reverse the glycogen accumulation and myopathy in aged model, we dosed Gaa^−/− mice with Gys1 ASO#2 starting at 3 months of age (Cohort 2) (Figure 3A). The Gys1 ASO#2 was administered by weekly subcutaneous injections at 25 mg/kg for 16 weeks (Figure 3A).

The rotarod motor performance assay showed that the ASO#2-mediated knockdown of Gys1 expression did not significantly slow down the deterioration of motor function in Pompe mice compared to the control ASO. However, minor improvements were observed in 5 out of 10 ASO#2-treated animals, whereas only 1 animal in the PBS group and none in the control ASO group showed improvement (Figure 3B). After completing the treatment, we sacrificed the animals at 6.5 months of age to study the quadriceps muscle, diaphragm, and heart tissues.

In the quadriceps muscle, we found an 83.34% (95% C.I.: 89.65–73.21) knockdown of mouse Gys1 mRNA and complete depletion of GYS1 protein compared to control ASO (Figure 3C and D, Table S1). Similarly, GYS1 protein was reduced by 80% in diaphragm and 70% in heart compared to mice treated with control ASO (Figure 3D, Table S1). Interestingly, the expression of Gys1 mRNA and GYS1 protein in quadriceps muscle of 6.5-month-old Pompe mice was not significantly different from that of the wild-type mice (Figure 3C and D). This may be because the key difference between Pompe and wild-type muscle GYS1 expression is not in total protein levels but in GYS1 phosphorylation, as reported by Ullman et al. (2024).³²

A biochemical assay demonstrated that the ASO#2-mediated knockdown of Gys1 expression was associated with a trend toward reduced glycogen accumulation in the quadriceps (32.66% reduction; 214 µg vs. 307 µg/100 mg tissue, not significant by one-way ANOVA, p < .02 by unpaired t test) and the diaphragm (32.33% reduction; 343 µg vs. 492 µg/100 mg tissue, not significant with one-way ANOVA, p < .05 with unpaired t test) of Gaa^−/− mice, compared to the control ASO. In contrast, glycogen accumulation in the heart was 660.5 µg/100 mg in ASO#2-treated mice versus 752.5 µg/100 mg in control ASO-treated mice (p = .88), indicating no statistically significant difference between the groups (Figure 3E, Table S1). The lower GYS1 protein knockdown efficiency in the heart (70%) compared to the quadriceps (98%) and diaphragm (85%) likely accounts for the reduced impact on glycogen reduction in cardiac tissue. Furthermore, quantitative analysis of histological sections of the quadriceps muscle using PAS staining confirmed an improvement of the aberrant glycogen accumulation following treatment with Gys1 ASO#2. The median staining intensity was 119 (95% C.I.: 113–125) in 6.5-month-old Gaa^−/− mice treated with control ASO, and 73 (95% C.I.: 69–77) in mice treated with ASO#2 (Figure 3F). To minimise technical variability in the PAS staining data, all samples from every cohort were processed simultaneously. Notably, a similar PAS staining mean intensity of 119 was observed in the quadriceps of Pompe mice treated with control ASO in both cohort 1 (4.5 months old, Figure 2E and F) and cohort 2 (6.5 months old, Figure 3E and F), despite differences in glycogen content: 501 µg/100 mg tissue in cohort 1 versus 307 µg/100 mg tissue in cohort 2. This discrepancy underscores the semi-quantitative nature of PAS staining.

2.4 ASO-mediated Gys1 knockdown in combination with enzyme replacement therapy further reverses Pompe pathology, preserving motor function in older Gaa^−/− mice

Substrate reduction therapy (SRT) using ASO to remove the source of new glycogen synthesised via GYS1 showed benefit for skeletal muscle, with variable effect on the heart in both 4.5- and 6.5-month-old Gaa^−/− mice. We reasoned that combining Gys1 ASO, which blocks the synthesis of new glycogen, with GAA enzyme replacement therapy (ERT) to simultaneously remove preexisting glycogen in the skeletal muscle and heart of older Pompe mice, could provide additive efficacy in muscle and especially the heart. Therefore, we dosed 4-month-old Gaa^−/− mice with Gys1 ASO#2 at 25 mg/kg by weekly subcutaneous injections for 6 weeks, followed by a 10-week-long co-administration phase where we combined ASO#2 and ERT, in the form of recombinant human alpha-acid glucosidase (rhGAA) enzyme (Cohort 3) (Figure 4A). rhGAA was dosed intravenously every 2 weeks. Alternatively, Gaa^−/− mice were dosed with either ASO#2 alone or ERT alone (Figure 4A). Three groups—wild-type, PBS, and control ASO-treated Gaa^−/− mice—were included in the study as controls, with the latter being used as the best comparator for the Gys1 ASO#2-treated Gaa^−/− mice.

The ASO#2+ERT combination treatment led to a steady increase in rotarod performance over the 4-month treatment period, with gains of 18%, 16%, 21%, and 39%, respectively compared to control ASO after 1, 2, 3, and 4 months of treatment. In contrast, ASO#2 alone increased rotarod performance during the first 3 months by 3%, 20%, and 25%, and 15% gain compared to control ASO. ERT alone was unable to prevent the functional decline in aging Pompe mice, resulting in a downward change in rotarod performance by –4%, –17%, –11%, and –17% after 1, 2, 3, and 4 months of treatment, respectively. The control ASO group in contrast showed significant declines of –7%, –35%, –57%, and –66% over the same time period (Figure 4B). The histogram in Figure 4C clearly highlights the effects of ASO#2 alone, ASO#2+ERT and ERT alone treatment on rotarod performance at the final time point.

At the end of the study, the 8-month-old mice were sacrificed for further biochemical and IHC analysis of the muscle tissue. Treatment with ASO#2, either alone or in combination with ERT, resulted in a 93%–99% reduction in mRNA expression and complete ablation of GYS1 protein in the quadriceps (Figure 4D and E). Similarly, minimal residual GYS1 protein, with 85–90% reduction was detected in the diaphragm and heart with the same treatment groups. As expected, treatment with ERT alone reduced GYS1 protein levels in the quadriceps and diaphragm of Gaa^−/− mice by 25% compared to control ASO treated mice. However, ERT had no effect on the heart, where GYS1 protein expression remained similar to control ASO treated Gaa^−/− mice (Figure 4E, Table S1).

Western blot analysis revealed pronounced GYS1 protein expression in quadriceps, diaphragm and heart of Gaa^−/− mice treated with PBS, approximately 20%–30% higher than in wild-type mice, while Gys1 mRNA levels showed more than a two-fold increase in quadriceps (Figure 4D and E).

Importantly, the systemic administration of Gys1 ASO#2, alone or in combination with ERT, was well tolerated in Pompe mice with no significant changes in body or organ weights, and no significant elevation in liver transaminases, blood urea nitrogen, or muscle creatine kinase in plasma compared to control groups (Figure S2).

Compared to control ASO, the biochemical glycogen assay revealed that ASO#2 alone did not result in a statistically significant reduction in quadriceps muscle glycogen content (177 µg vs 300.5 µg/100 mg tissue, p = .16). However, ASO#2 alone led to a significant 50.86% reduction in diaphragm glycogen levels (227 µg vs. 478 µg/100 mg tissue, p < .05) in Gaa^−/− mice (Figure 4F, Table S1). The ASO#2+ERT combination had an even greater effect, reducing glycogen levels by 54.65% in quadriceps muscle (131 µg vs. 300.5 µg/100 mg tissue) and by 83.13% in the diaphragm (84 µg vs. 478 µg/100 mg tissue) (Figure 4F, Table S1).

Compared to control ASO, ERT alone had some minor, but not significant effect on glycogen accumulation in Gaa^−/− mice, with quadriceps glycogen content at 217.5 µg vs. 300.5 µg/100 mg tissue (p = .48), and diaphragm glycogen at 256 µg vs. 478 µg/100 mg tissue (p = .07). ERT was more efficacious in the heart, leading to a 97.72% reduction in heart glycogen content compared to control ASO when dosed alone (213 µg vs. 1854 µg/100 mg tissue), and a 87.51% reduction when combined with Gys1 ASO#2 (456 µg vs. 1854 µg/100 mg tissue), and a 34.55% reduction when ASO#2 was used alone (1335 µg vs. 1854 µg/100 mg tissue) (Figure 4F, Table S1).

PAS staining of quadriceps muscle histological sections corroborated the biochemical assay results, showing a marked reduction in glycogen content, particularly in the ASO#2 group (median intensity: 63, 95% C.I.: 56– 71) and the ASO#2+ERT group (median intensity: 48, 95% C.I.: 43–53), compared to the control ASO group (median intensity: 122, 95% C.I.: 94–150) and the ERT alone group (median intensity: 104.5, 95% C.I.: 90–119) (Figure 4G and H). Thus, the motor function improvement correlated well with the reduction in muscle glycogen content.

2.5 Effect of Gys1 ASO and ERT on muscle regeneration in Gaa^−/− mice

Centrally located nuclei are a hallmark feature observed in certain muscle disorders, particularly those involving muscle regeneration and repair indicating newly regenerated myofibres following damage.^{33, 34} Approximately 600 quadriceps myofibres from 3–5 mice per group (1–2 images/mouse) were analysed for centralised nuclei. In the 8-month-old Gaa^−/− mice, we observed a significant increase in the percentage of myofibre with centralised nuclei in quadriceps muscle (1.5%, 95% CI: 0.3–3.3) compared with WT mice (1.1%, 95% CI: 21.4–60.9), confirming impaired muscle regeneration (Figure S3). Treatment with Gys1 ASO#2, either alone or in combination with ERT, or ERT led to alleviation these abnormalities. The percentage of myofibre with centralised nuclei in Gaa^−/− mice treated with Gys1 ASO#2 was 20.7% (95% CI: 5.2–36.3), while the combined Gys1 ASO#2 + ERT treatment reduced it to 11.9% (95% CI: 1.3–22.4). ERT alone also provided some improvement, with a reduction to 18.7% (95% CI: 9.0–28.5), compared to the control ASO group, which exhibited a high percentage of centralised nuclei at 48.9% (95% CI: 31.4–66.5) (Figure S3). These results suggest that Gys1 ASO#2, with or without ERT, may enhance muscle regeneration and alleviate some of the pathological features associated with Pompe disease.

Taken together, our data indicates that blocking the synthesis of new glycogen while concomitantly enhancing its clearance is beneficial in treating both the skeletal muscle and cardiac abnormalities in Pompe disease (Figure 6).

2.6 ASO-mediated Gys1 knockdown with or without ERT, but not ERT alone, ameliorates the autophagic markers in the skeletal muscle of Gaa^−/− mice

Defective autophagy is a major factor contributing to the muscle pathology in Pompe disease.⁷ Consistent with the highly integrated functions of lysosomes and autophagosomes, it has been shown that lysosomal dysfunction in Pompe disease leads to a block in autophagic flux and accumulation of autophagic debris, particularly in muscle tissue.³⁵ It is important to note that the Gaa^−/− mouse model faithfully recapitulates the autophagic defects described in tissues from Pompe patients.³⁰

Incomplete autophagic flux can be assessed by measuring the aberrant accumulation of LC3 and p62 proteins. Additionally, immunostaining of histological sections for lysosomal-associated membrane protein 1 (LAMP1) provides a method to detect enlarged lysosomes.^{36, 37} We investigated the efficacy of Gys1 ASO, alone or in combination with ERT (Cohort 3, Figure 4A), in ameliorating the defective autophagy in Gaa^−/− mice. Western blot analysis confirmed accumulation of p62 and LC3-II proteins in quadriceps muscle and heart of 8-month-old Gaa^−/− mice treated with PBS (Figure 5A). ERT alone alleviated the autophagic defect in the heart but not significantly in the skeletal muscle, as indicated by the reduced levels of LC3-II and LAMP1 proteins in cardiac tissue with no major effect in quadriceps muscle (Figure 5A and B), highlighting the tissue-specific limitation of ERT. On the other hand, while Gys1 ASO#2 contributed to a lesser extent, combination Gys1 ASO#2 with ERT resulted an additive benefit, significantly enhancing the reduction of the autophagic LC3, LAMP1 markers in both skeletal and cardiac muscles (Figure 5A and B). In addition, immunostaining of histological sections confirmed that ASO#2+ERT administration reduced LAMP1 protein levels in quadriceps muscle of Gaa^−/− mice (Figure 5B), suggesting that the combination therapy could rescue the autophagy pathway dysfunction in Pompe muscle after 4 months of treatment. The effect was more pronounced in the ASO#2+ERT group than in the Gys1 ASO#2-only group (Figure 5B). Our data indicates that the combination of Gys1 ASO and ERT might provide therapeutic benefit by targeting various aspects of Pompe pathogenesis, namely glycogen accumulation and autophagy defects (Figure 6).

2.7 Systemic Gys1 knockdown does not mitigate Pompe pathology in the central nervous system

Pompe disease is a complex multi-systemic condition with growing evidence of central nervous system (CNS) involvement.³⁸ Indeed, we found remarkable abnormalities in the spinal cord of Gaa^−/− mice at 6.5 months of age (Figure S4). The pathologic changes were visible in histological sections of the spinal cord, where Nissl staining (toluidine blue) revealed the presence of swollen cell bodies, activated glial cells, and central nucleated cells. Indeed, we observed a statistically significant increase in number of neurons with eccentric nuclei in the spinal cord of Gaa^−/− mice compared to WT controls (p < .0001, Fisher's Exact Tests, 30–40 neurons per group; Figure S3, data not shown). Subcutaneous injection of either control ASO or Gys1 ASO#2 did not ameliorate the CNS phenotype of Gaa^−/− mice (Figure S4, data not shown). These results are consistent with the inability of systemically dosed unconjugated ASOs to cross the blood-brain barrier.^39-41

4.1 Identification of antisense oligonucleotides (ASOs) targeting mouse Gys1

All ASOs used in the studies described here were 16 nucleotides in length with a phosphorothioate (PS) backbone and three 2′-constrained ethyl (cEt)-modified nucleotides at both ends (3-10-3 gapmer configuration). A well-characterised control ASO which does not hybridise to any mouse mRNA sequence was included in the experiments. The oligonucleotides were synthesised and purified as previously described.⁵⁹ ASOs targeting murine Gys1 were designed and initially screened for activity in B16-F10 cells via electroporation. The most active ASOs were subsequently evaluated in wild-type mice to identify ASOs that were safe and active in vivo. A dose–response confirmation experiment for the two lead ASOs, ASO#1 and ASO#2, was performed in wild-type mice by subcutaneous injection twice a week for 5.5 weeks at 12.5, 25, 50, and 100 mg/kg total weekly doses (shown in Figure S1A). All ASOs were formulated in phosphate buffered saline (PBS) and were administered to mice by subcutaneous injection. The mouse ASO screening studies were conducted under protocols approved by the Institutional Animal Care and Use Committees at Ionis Pharmaceuticals.

4.2 Characterisation of the Gaa^−/− animal model of Pompe disease

The Gaa^−/− Pompe mouse model was a gift of Dr. Nina Raben.⁶⁰ The mouse model is a constitutive homozygote knockout of the murine acid alpha-glucosidase (Gaa) gene. One-month-old Gaa^−/− mice, which show only the initial signs of tissue glycogen accumulation without severe myopathy, were used in prevention studies, whereas 3-month-old Gaa^−/− mice, which show marked glycogen accumulation and motor function impairment, were used in reversal studies. All the intervention studies performed in either the Gaa^−/− mouse model or wild-type C57BL/6 control mice were conducted in accordance with the Institutional Animal Care and Use Committee at the University of California, Irvine (UCI), under protocol IACUC AUP-19-075, and were consistent with Federal guidelines. The mice were housed at UCI animal facility that maintained a constant temperature (22°C) and controlled light and dark cycles (12:12). The overall condition of the mice was monitored daily throughout the experimental process to assess general health and mitigate any pain and suffering.

4.3 Administration of ASOs to Gaa^−/− mice

Gaa^−/− or wild-type control mice were dosed with ASOs once a week at 25 mg/kg via subcutaneous injection. For the early-treatment study (Cohort 1) described in Figure 2, 1-month-old Gaa^−/− mice were randomly divided into groups for treatment with different Gys1 ASOs, control ASO, or vehicle (PBS), n = 12–17 mice per group. To study therapeutic effects of Gys1 ASO in mice with established pathology (Cohort 2) described in Figure 3, 3-month-old Gaa^−/− mice were randomly divided into three groups for treatment with Gys1 ASO#2 (n = 15), control ASO (n = 10), or vehicle (PBS) (n = 10). The mice received a total of 16 weekly ASO doses (16 weeks duration) and were sacrificed 48 h after the last ASO dose. Body weight changes, and the weight of liver, kidney, and spleen were recorded to exclude potential toxicity of the ASOs in the context of the long dosing paradigm in diseased Gaa^−/− mice.

4.4 Adjunct treatment using Gys1 ASO in combination with ERT in Gaa^−/− mice

Four-month-old Gaa^−/− mice were dosed with Gys1 ASO#2 or control ASO by subcutaneous injection once a week at 25 mg/kg for 6 weeks. Starting from week 7 the mice additionally received 20 mg/kg recombinant human GAA (rhGAA) (Genzyme Corporation, Cambridge, MA; Cat.: NDC 58468-0160-1) via intravenous injection once every 2 weeks. The antihistaminic agent diphenhydramine hydrochloride was administered to the mice via intraperitoneal injection 15 min before receiving the human enzyme, starting from the second rhGAA administration. The weekly subcutaneous ASO administration continued throughout the ERT phase for a total of 16 doses. The mice were euthanised at 8 months of age, after having been treated with Gys1 ASO for 4 months, the last two of which overlapped with the ERT; n = 5 per treatment group.

4.5 Accelerating rotarod performance test

Muscle function was evaluated using a rotarod apparatus (Rotamex-5, Columbus Instruments, Columbus, OH). Briefly, mice were placed on a rotarod whose speed was set to progressively increase from 4 to 40 rotations per minute (rpm) in 300 s. The latency to fall from the rotating rod was recorded as a measure of muscle strength. Assessments were done on three consecutive days, three to five trials each day with at least 2-min rest intervals between the trials. The mean latency time to fall off the rotarod from three independent trials from the last two days was used for analysis.

4.6 Biochemical analysis of serum from mice for toxicity studies

Blood samples were collected from mice via cardiac puncture under deep anesthesia using isoflurane inhalation. Immediately after collection, approximately 300–500 µL of whole blood were placed into serum separator tubes (Greiner Bio-One Minicollect tubes; Cat.: 450472). Samples were allowed to clot for at least 20 min at room temperature and then centrifuged at 3000 rpm for 15 min. Serum was transferred to fresh tubes and stored at –80°C before being used for clinical biochemistry evaluation at IDEXX Laboratories (IDEXX Laboratories Inc., Maine, USA). Parameters measured include alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, blood urea nitrogen, creatine kinase.

4.7 Gene expression analysis using quantitative real-time PCR

Tissue samples were isolated from mice and snap frozen in liquid nitrogen. Frozen tissues were homogenised in TRIzol reagent (Thermo Scientific, Waltham, MA; Cat.: 15596026), and total RNA was extracted according to the manufacturer's instructions. Real-time qPCR analysis was performed using the following reagents: mouse Gys1 (Taqman™ assay ID: Mm01962575_s1), mouse Gapdh (Taqman™ assay ID: Mm 99999915_g1, housekeeping gene). In the initial assessment of ASO activity in wild-type mice (Figure S1A) real-time qPCR analysis was performed using the following primer-probe sets: mouse Gys1 (forward primer: 5′-TGATGAAGAGAGCCATCTTTGC-3′; reverse primer: 5′-AGGAGTCGTCCAGCATGTTGT-3′; probe: 5′-ACTCAGCGGCAGTCTTTCCCACCA-3′), mouse cyclophilin A (forward primer: 5′-TCGCCGCTTGCTGCA-3′; reverse primer: 5′-ATCGGCCGTGATGTCGA-3′; probe: 5′-CCATGGTCAACCCCACCGTGTTC-3′). Reactions were conducted in triplicate on a QuantStudio7 real-time PCR system (Thermo Fisher Scientific, Waltham, MA), following the manufacturer's instructions.

4.8 Biochemical measurement of tissue glycogen content

Mouse tissues were collected, snap frozen in liquid nitrogen upon collection, and stored at –80°C. Approximately 10 mg of frozen quadriceps muscle were rinsed in 300 µL of ice-cold PBS, then homogenised in 200 µL of double-distilled water using a Wheaton dounce tissue grinder (DWK Life Sciences, Rockwood, TN) as follows: 10–20 passes on loose pestle followed by 10–20 passes on tight pestle, keeping the tissue grinder on ice. The tissue lysates were then heated at 100°C for 10 min and then centrifuged at 15 000 × g for 10 min at 4°C before glycogen measurement. Glycogen concentration was measured using a colorimetric assay (Abcam, Ab65620), according to the manufacturer's instructions. The resulting signals were analysed using the BioTek Gen5 software (Agilent Technologies, La Jolla, CA), and normalised to the mount of the tissue sample used.

4.9 Western blot analysis of GYS1 protein and autophagy markers

Approximately 30 mg of frozen tissue was homogenised in 500 µL of RIPA buffer (Sigma-Aldrich, St. Louis, MO; Cat.: R0278) supplemented with a proteinase inhibitor cocktail (Sigma-Aldrich, St. Louis, MO; Cat.: P8340) using a handheld electric homogeniser (Qsonica, Newtown, CT; Cat.: XL-2000). The tissue homogenate was then agitated on a Benchmark Roto-Mini Plus (Benchmark Scientific, Sayreville, NJ) for 2 h at 4°C. Lastly, the tissue lysate was centrifuged at 20 000 × g at 4°C for 30 min, and the supernatant was collected for analysis. The sample protein concentration was measured using the Micro BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA; Cat.: 23235). 20 µg of total protein lysate was resolved by electrophoresis on Bis-Tris 4–12% NuPAGE polyacrylamide gels using the Novex Mini Cell system (Invitrogen/Thermo Fisher Scientific, Waltham, MA). The proteins were then transferred to a PVDF membrane, which was later blocked with 1% bovine serum albumin (BSA) in Tris-buffered saline containing 0.1% Tween 20 (TBS-T). The primary antibodies used in our studies were: anti-GYS1 (Abcam, Waltham, MA; Cat.: ab 40810; 1:5 000 dilution), anti-LC3B-I/II (Abcam, Waltham, MA; Cat.: ab192890; 1:5000 dilution), anti-p62 (Abcam, Waltham, MA; Cat.: ab56416; 1:20 000 dilution), anti-LAMP1 (Abcam, Waltham, MA; Cat.: ab27170; 1:2000 dilution), anti-GAPDH (Abcam, Waltham, MA; Cat.: ab18160; 1:10 000 dilution; served as loading control), anti-β-ACTIN (Abcam, Waltham, MA; Cat.: ab8227; 1:5000 dilution; served as alternative loading control). Horseradish peroxidase (HRP)-linked goat anti-mouse or goat anti-rabbit secondary antibodies were used (1:5000 dilution). On average, each experiment included n = 5 samples per group, with three technical replicates per sample. Densitometric quantification was performed using ImageJ software (National Institutes of Health, Bethesda, MD).

4.10 Periodic acid-Schiff histological analysis

Periodic acid-Schiff (PAS) staining was performed to detect glycogen in histological tissue sections. Briefly, the quadriceps muscle was dissected and fixed in a 4% neutral buffered formalin solution at room temperature for 20 h. The fixed tissue was then rinsed in PBS and transferred to a 70% ethanol solution for storage. Paraffin embedding, sectioning (5 µm), and PAS staining were performed by Histoserv (Germantown, MD).⁶¹ Images of the histological slices were captured using an Aperio AT2 scanner (Leica Biosystems, Germany).

PAS Staining Signal Intensity Quantification: The fuchsia-coloured intensity within muscle fibres was measured using AI Microscopy Image Analysis Software (Imaris 10.2) in three steps: (1) Image conversion: 20× images from the Leica microscope were converted to.ims files using the Imaris file converter, then to greyscale for easier processing. The voxel size was adjusted according to the microscope manual. (2) Image segmentation: Machine learning segmentation was applied to distinguish foreground and background, followed by manual review to ensure accurate cell segmentation. (3) Data exporting and the median intensity was used for group analysis. For each treatment and control group, five images of quadriceps muscle from 4–5 animals, with at least 200 cells per image were analysed. To ensure accuracy, 50% of the images were reanalysed by a different individual. To quantify the number of myofibres with centralised nuclei, a minimum of 600 myofibres per animal were manually evaluated in the PAS-stained sections of quadriceps muscle (n = 4–5 mice per each group). The data is presented as the percentage of myofibres with a centralised nucleus.

4.11 Immunohistochemistry (IHC) studies

For lysosomal analysis, the expression of lysosomal associated membrane protein 1 (LAMP1) was assessed using IHC in tissue sections of quadriceps muscle embedded in frozen section medium (Fisher Scientific, Hampton, NH; Cat.: 22-046-511), as described previously.⁶² Briefly, 8 µm-thick cryosections (10 sections per each animal, 5 animals per treatment group) were fixed in 4% paraformaldehyde (PFA), blocked in donkey serum, and stained with an anti-LAMP1 primary antibody (Abcam, Waltham, MA; Cat.: ab24170; 1:500 dilution) at 4°C overnight. The sections were then washed three times with PBS and incubated with fluorescein-conjugated secondary antibody (Thermo Fisher Scientific, Waltham, MA; Cat.: A10042; 1:500 dilution) for 1 h at room temperature, then washed and mounted with DAPI-containing mounting media (Vectashield, Vector Laboratories, Newark, CA; Cat.: H-1200-10). The slides were imaged on a Zeiss LSM 900 confocal microscope (Carl Zeiss Microscopy, White Plains, NY). For the quantification of the LAMP1 staining, the percentage of LAMP1-positive myofibres was calculated and compared across treatment groups.

Nissl (toluidine blue) staining was used to study the morphology and pathology of neural tissue. The spinal cord was harvested from mice, fixed in 4% paraformaldehyde (PFA) for 24 h, and then incubated in a solution containing 30% sucrose in PBS overnight, embedded in optimal cutting temperature compound (OCT compound), and cryo-sectioned. Subsequently, the 10-micron cryosections were stained with 1% toluidine blue containing 1% borate solution as previously described.^{63, 64} Briefly, the slices were placed in Toluidine blue solution for 30 s and rinsed in triple distilled water. Stained slices were examined using light microscopy.

4.12 Statistical analysis

Statistical analysis was performed using GraphPad Prism software (GraphPad Software, Boston, MA), or RStudio software (RStudio, Boston, MA) in the case of Figure 1B. Unpaired t-test, one-way ANOVA, or two-way ANOVA were used, as specified in the respective figure legends. One-way ANOVA was used only when comparing between genotypes, among age groups, or among treatment groups). Contingency table analysis with Fisher exact test was used to analyse the neurons with eccentric nuclei in the cervical region of the spinal cord. p-values lower than .05 were considered statistically significant; *p < .05, **p < .01, ***p < .005, and ****p < .0001, ns = not significant. Error bars indicate the standard error of the mean (SEM).