Background
As the leading genetic cause of infant deaths, spinal muscular atrophy (SMA) is a devastating and incurable neuromuscular disorder [
1,
2]. SMA affects 1 in 6,000 to 10,000 births and results from deletions or mutations in the survival motor neuron 1 (
SMN1) gene [
1‐
3]. The primary pathological hallmark of SMA is the loss of lower motor neurons from the spinal cord and corresponding muscular atrophy with subsequent paralysis and in most severe cases, death [
1,
2].
The complete loss of the SMN protein is embryonic lethal [
4]. In humans however, a recent duplication event in chromosome 5 has given rise to the centromeric
SMN2 gene [
3]. While both
SMN1 and
SMN2 genes differ by only a few nucleotides, a critical C to T substitution lies within position 6 of
SMN2 exon 7 [
5,
6]. This silent mutation results in the aberrant splicing of exon 7, giving rise to the biologically unstable SMNΔ7 protein [
3,
5]. Although the
SMN2 gene produces predominantly the SMNΔ7 protein, a small amount of full-length SMN is still produced [
3]. Thus, the number of
SMN2 gene copies in SMA patients is a key modifier of disease severity [
3,
7,
8].
One of the major hurdles in SMA is to understand how the loss of a ubiquitously expressed protein leads to the specific loss of spinal cord motor neurons. Work from various research groups has identified distinct roles for SMN in neurodevelopment, neuromaintenance, RNA metabolism, at the neuromuscular junction (NMJ) and in skeletal muscle (reviewed in [
9]). As of yet however, none of these various functions of the SMN protein have been recognized as being solely responsible for SMA pathogenesis.
Work from our laboratory has shown that Smn depletion in cellular and mouse models results in altered expression and localization of a number of regulators of actin cytoskeletal dynamics [
10‐
12]. Indeed, analysis of spinal cords from SMA mice revealed a significant increase in active RhoA (RhoA-GTP) [
12], a major upstream regulator of the actin cytoskeleton [
13]. RhoA-GTP signaling in neuronal cells modulates various cellular functions such as growth, neurite formation, polarization, regeneration, branching, pathfinding, guidance and retraction (reviewed in [
14,
15]). Our previous work demonstrated that administration of the RhoA/Rho kinase (ROCK) inhibitor Y-27632 [
16] leads to a dramatic increase in survival in an intermediate mouse model of SMA [
12]. Recently Nölle
et al. demonstrated that knockdown of Smn in PC12 cells affects the phosphorylation state of downstream effectors of ROCK, supporting the value of the ROCK pathway as a therapeutic target for SMA pathogenesis [
17].
In the present work, we have treated SMA mice with fasudil, a ROCK inhibitor approved for US clinical trials. We show that fasudil dramatically improves the lifespan and increases muscle fiber size in Smn
2B/-
SMA mice. Furthermore, we report for the first time that ROCK inhibition restores normal expression of markers of skeletal muscle development in SMA mice. Our study highlights the beneficial effects of ROCK inhibition not only for SMA pathogenesis but also for any degenerative disease that has NMJ and skeletal muscle development defects. Importantly, as fasudil is currently being used in US Food and Drug Administration (FDA)-approved clinical trials for other disorders, re-purposing it is an exciting and feasible therapeutic approach for the treatment of SMA.
Methods
Animal models
The
Smn
2B/-
mice were established in our laboratory and maintained in our animal facility on a C57BL/6 × CD1 hybrid background. The 2B mutation consists of a substitution of three nucleotides in the exon splicing enhancer of exon 7 [
18,
19]. The
Smn knock-out allele was previously described by Schrank
et al. [
20] and
Smn
+/-
mice were obtained from The Jackson Laboratory (Bar Harbor, Maine, USA). All animal procedures were performed in accordance with institutional guidelines (Animal Care and Veterinary Services, University of Ottawa).
Fasudil administration
Fasudil (LC Laboratories; Woburn, Massachusetts, USA) was diluted in water and administered by a modified oral gavage procedure [
21] to
Smn
2B/-
and
Smn
2B/+
mice from post-natal day (P)3 to P21. The three fasudil dosage regimens were as follows: low dose (30 mg/kg once daily), medium dose (30 mg/kg twice daily), and high dose (30 mg/kg twice daily from P3 to P6; 50 mg/kg twice daily from P7 to P13; 75 mg/kg twice daily from P14 to P21). Vehicle-treated animals received water. Survival and weight were monitored daily.
Antibodies
The primary antibodies used were as follows: mouse anti-actin (1:800; Fitzgerald; Acton, Massachusetts, USA), mouse anti-Smn (1:5000; BD Transduction Laboratories; Mississauga, Ontario, Canada), rabbit anti-phosphorylated cofilin (1:250; Chemicon; Billerica, Massachusetts, USA), rabbit anti-cofilin (1:500; Millipore; Billerica, Massachusetts, USA), rabbit anti-phosphorylated cofilin 2 (1:500; Cell Signaling; Danvers, Massachusetts, USA), rabbit anti-cofilin 2 (1:500; Millipore), mouse anti-myogenin (1:250; BD Transduction Laboratories), rabbit anti-HB9 (1:50; Abcam; Cambridge, Massachusetts, USA), mouse anti-2H3 (1:250; Developmental Studies Hybridoma Bank; Iowa City, Iowa, USA) and mouse anti-SV2 (1:250; Developmental Studies Hybridoma Bank). The secondary antibodies used were as follows: horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:5000; Bio-Rad; Mississauga, Ontario, Canada), HRP-conjugated goat anti-rabbit IgG (1:5000; Bio-Rad), DyLight goat anti-mouse (1:250; Jackson ImmunoResearch; West Grove, Pennsylvania, USA), goat anti-rabbit biotin-SP-conjugated (1:200; Dako; Burlington, Ontario, Canada)), streptavidin-Cy3-conjugated (1:600; Jackson ImmunoResearch) and Alexa Fluor 680 goat anti-mouse (1:5000; Molecular Probes; Burlington, Ontario, Canada). The α-bungarotoxin (BTX) conjugated to tetramethylrhodamine isothiocyanate was from Molecular Probes (5 μg/mL).
Immunoblot analysis
Equal amounts of spinal cord and tibialis anterior (TA) muscle tissue extracts were separated by electrophoresis on 10% SDS-polyacrylamide gels and blotted onto nitrocellulose membranes (Amersham; Baie d'Urfe, Quebec, Canada). The membranes were blocked in 5% non-fat milk in TBST (10 mM Tris-HCl pH 8.0, 150 mM NaCl, and 0.1% Tween 20 (Sigma; Oakville, Ontario, Canada)). Membranes were incubated overnight at 4°C with primary antibody, followed by a one-hour incubation with the secondary antibody. All washes were performed with TBST. Signals were visualized using the ECL or the ECL plus detection kit (Amersham). Exposure times were chosen based on the saturation of the highest amounts of protein.
Hematoxylin and eosin staining
Spinal cord (L1-L2 lumbar regions) and TA muscle sections (5 μm) were deparaffinized in xylene and fixed in 100% ethanol. Following a rinse in water, samples were stained in hematoxylin (Fisher; El Paso, Texas, USA) for three minutes, rinsed in water, dipped 40 times in a solution of 0.02% HCl in 70% ethanol and rinsed in water again. The sections were stained in a 1% eosin solution (BDH; Billerica, Massachusetts, USA) for one minute, dehydrated in ethanol, cleared in xylene and mounted with Permount (Fisher). Images were taken with a Zeiss Axioplan2 microscope, with a 20× objective.
Quantitative assays were performed on three mice for each genotype and five sections per mouse. Motor neurons were identified by their shape and size (> 10 μm in diameter) in the same designated area of the ventral horn region of the spinal cord. Every fifth section was analyzed and the subsequent totals were multiplied by five to give an estimate of total motor neuron number. Only motor neurons with visible nuclei were counted so as to prevent double-counting. For TA quantitative assays, the area of muscle fiber within designated regions of the TA muscle sections was measured using the Zeiss AxioVision software.
Immunohistochemistry
For immunohistochemistry, spinal cord sections were first deparaffinized in xylene (3 × 10 minutes), fixed in 100% ethanol (2 × 10 minutes), rehydrated in 95% and 75% ethanol (5 seconds each) and placed for 5 minutes in 1 M Tris-HCl pH 7.5. Sections were then placed in boiling sodium citrate antigen retrieval buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0) for 20 minutes in the microwave. The sections were rinsed for 10 minutes under running cold tap water and incubated for 2 hours at room temperature (RT) in blocking solution (TBLS (10% NaN3), 20% goat serum, 0.3% Triton X-100). This was followed by an overnight incubation at 4°C with the primary antibody. Subsequently, sections were incubated for 1 hour at RT with the biotinylated rabbit antibody followed by a 1 hour incubation at RT with streptavidin-Cy3. All washes were done with PBS. Hoechst (1:1000) was added to the last PBS wash followed by the slides being mounted in fluorescent mounting medium (Dako). Images were taken with a Zeiss confocal microscope, with a 20× objective, equipped with filters suitable for Cy3/Hoechst fluorescence.
Neuromuscular junction immunohistochemistry
Transversus abdominis (TVA) and TA muscle sections were labeled by immunohistochemistry to allow quantification of neuromuscular innervation as described previously [
12,
22]. Briefly, TVA muscles were immediately dissected from recently sacrificed mice and fixed in 4% paraformaldehyde (Electron Microscopy Science; Hatfield, Pennsylvania, USA) in PBS for 15 minutes. Post-synaptic acetylcholine receptors were labeled with αBTX for 10 minutes. Muscles were then permeabilized in 2% TritonX for 30 minutes and blocked in 4% bovine serum albumin/1% TritonX in PBS for 30 minutes before incubation overnight in primary antibodies and visualized with DyLight-conjugated secondary antibodies. Whole TA muscles were dissected and fixed in 4% paraformaldehyde. Following the removal of connective tissue, the TA muscles were incubated with αBTX Alexa Fluor 555 conjugate for 20 minutes at RT. Whole TVA muscle and a thin filet of TA muscle were mounted in Dako fluorescent mounting media. Images were taken with a Zeiss confocal microscope equipped with filters suitable for fluorescein isothiocyanate (FITC)/Cy3/fluorescence. Categorization of pre- and postsynaptic morphologies was performed as previously reported [
23].
Pen test
Balance and strength were assessed using the pen test as described [
24]. Mice were placed on a suspended pen at different time-points (P12, 14, 17 and 21). The latency to fall from the pen was measured with a plateau of 30 seconds. At each time-point, individual mice were assessed three consecutive times.
Statistical methods
All statistical analyses were performed using the GraphPad Prism software. For the Kaplan-Meier survival analysis, the log-rank test was used and survival curves were considered significantly different at P < 0.05. When appropriate, the Student's two-tail t test for paired variables and one-way ANOVA were used to test for differences between samples and data were considered significantly different at P < 0.05.
Discussion
Previous work has implicated the RhoA/ROCK pathway in SMA pathogenesis [
10,
12,
17]. In the present study, we demonstrate that targeting the ROCK pathway with the inhibitor fasudil significantly increases the lifespan of the
Smn
2B/-
SMA mice. The increased survival is independent of Smn expression, weight gain, pen test performance and pre-synaptic NMJ phenotype. We find, however, that fasudil benefits post-synaptic pathology and muscle development. Importantly, the results obtained from other fasudil clinical trials are proof-of-principle of its feasibility and availability as a therapeutic approach for the treatment of SMA. Future SMA clinical endeavors should therefore consider assessing the beneficial potential of ROCK inhibitors.
Smn protein levels remained significantly low in both fasudil-treated spinal cord and muscle samples of SMA mice. These findings are important when considering therapeutic avenues for SMA. There are presently many strategies being developed to increase the expression of SMN, such as gene therapy, modulation of transcription and splicing of
SMN2, and the use of various histone deacetylase (HDAC) inhibitors (reviewed in [
34‐
36]). Although these therapeutic approaches show promising results, they remain in pre-clinical stages and may not be as efficient if administered to mid- to late-symptomatic patients [
37]. It is therefore crucial to understand the pathological molecular pathways that are affected upon SMN loss and how these can be modulated to attenuate their degenerative effects. Along with other research groups, we have shown that the RhoA/ROCK pathway is indeed perturbed in SMA cellular and animal models and that its targeting leads to a significant beneficial outcome [
10,
12,
17].
We had previously identified the upregulation of RhoA-GTP in the spinal cords of
Smn
2B/-
mice [
12]. The misregulated RhoA/ROCK pathway in the spinal cord was, therefore, the primary target of our Fasudil therapeutic strategy [
12]. Interestingly, we have observed that fasudil does not prevent the motor neuron loss that occurs in the
Smn
2B/-
mice. In fact, the most apparent effects of fasudil appear to be the restoration of normal skeletal muscle growth and development, as well as increased post-synaptic EP area. A number of recent reports suggest that the SMN protein may have a muscle-intrinsic role that influences SMA pathology ([
28‐
30] and JGB, unpublished data). Active RhoA has previously been shown to positively regulate the expression of myogenin [
38,
39]. Furthermore, work performed in avian and murine myoblasts shows that inhibition of ROCK promotes exit from the cell cycle and subsequent terminal differentiation [
40]. Indeed, myoblasts treated with the ROCK inhibitor Y-27632 display increased differentiation, cell fusion and myotube formation [
40]. Fasudil's inhibition of the RhoA/ROCK pathway most likely restores the normal skeletal muscle developmental program of
Smn
2B/-
mice via modulation of myoblast differentiation and fusion, as well as myogenin expression. The fasudil-dependent increase in myofiber size could lead to the subsequent increase in EP size. Indeed, a positive correlation has previously been established between myofiber size and motor EP size [
41]. Furthermore, various reports suggest that post-synaptic differentiation and formation is initially muscle-dependent and motor axon-independent [
42,
43]. Our study, therefore, highlights two important points. Firstly, therapeutic strategies that improve skeletal muscle and EP growth should be considered when developing therapies for SMA. Secondly, ROCK inhibition may have positive outcomes in other pre-clinical disease models characterized by muscle atrophy and NMJ pathology.
Intriguingly, the dramatic increase in skeletal muscle myofiber size of fasudil-treated
Smn
2B/-
mice is not accompanied by changes in weight or strength, when compared to vehicle-treated
Smn
2B/-
mice. Previous studies have reported this phenomenon, providing a variety of potential explanations. In cases of sarcoplasmic hypertrophy, the non-contractile myofiber components expand while muscular strength remains unchanged [
44]. Further, the characterization of a post-natal myogenin knockout mouse model revealed normal skeletal muscle size albeit with a 30% weight loss compared to control littermates [
45]. The authors suggest that this phenotype is caused by a slower growth rate and perturbed energy homeostasis [
45]. Finally, Rehfeldt
et al. showed that mice homozygous for the
Compact myostatin mutation (
C/C) display muscular hyperplasia and increased muscle weight but with a reduction in overall body weight [
46]. The authors also identify a reduction in the number of capillaries per muscle in the
C/C mice, subsequently impacting oxidative metabolism [
46]. Interestingly, recent work in the severe SMA mouse model demonstrated a significant decrease in the capillary bed density within skeletal muscle [
47]. Thus, the findings mentioned above highlight the fact that an increase in muscle size and or weight does not necessarily positively correlate with an increase in body weight. Regardless, the restoration of myofiber growth and skeletal muscle development by fasudil, in the absence of weight gain, appears to be sufficient to provide therapeutic benefits to the
Smn
2B/-
mice.
In recent years, it has been postulated that SMA may be a die-back neuropathy, where the motor axons initially reach the EP but subsequently retract as disease progresses [
48‐
50]. This hypothesis suggests that synapses are selectively vulnerable in SMA, with synapse loss preceding cell body degeneration. In addition, it has been suggested that neurons undergo compartmental degeneration, where the soma, axons and synapses of neurons possess specific and compartmentalized mechanisms of degeneration [
51‐
53]. It therefore follows that therapeutics which target distal compartments of the cell, such as the synapse or axon, can be protective to the cell body. In our study, we show that while fasudil administration has little impact upon the initial loss of motor neurons, it dramatically increases myofiber and EP size in SMA mice. We therefore suggest that this improvement in post-synaptic parameters stabilizes the synaptic connections and subsequently protects the remaining motor neurons. Consistent with this observation, the surviving synapses constitute NMJs that will eventually develop and mature normally. Given the tight correlation between EP maturation and neuromuscular activity (reviewed in [
54]), fasudil may indirectly improve NMJ transmission, subsequently ameliorating motor EP maturation. Alternatively, considering the crucial role of the actin cytoskeleton in the redistribution of acetylcholine receptors (AChRs) during post-synaptic remodeling [
55,
56], fasudil's modulation of actin dynamics could directly restore normal AChR clustering. Clearly, the understanding and identification of fasudil's influence on NMJ maturation in SMA mice requires further investigation. Nevertheless, our work highlights the applicability of the compartmental degeneration hypothesis to SMA pathogenesis and the potential of therapies aimed at preventing synaptic degeneration.
ROCK has evolved as an important therapeutic target in various models of cardiovascular disease, spinal cord injury and glaucoma (reviewed in [
57‐
59]). Furthermore, the ROCK inhibitor fasudil, which has been approved in US clinical trials, has shown beneficial effects in patients with vasospastic angina [
60], stable effort angina [
61], general heart failure [
62] and pulmonary hypertension [
63]. It has now become evident that the pathogenic misregulation of the RhoA/ROCK pathway in various Smn-depleted cellular and animal models can also be modulated by the ROCK inhibitors Y-27632 and fasudil, leading to significant positive outcomes [
10,
12,
17].
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
MB, LMM and RK conceived and designed the experiments. MB, LMM and CLA performed the experiments. MB and LMM analyzed the data. JGB performed the initial characterization of misregulated myogenin expression in SMA skeletal muscle. MB and LMM wrote the paper. All authors read and approved the final manuscript.