Abstract
Spinal muscular atrophy (SMA) is a neuromuscular disease caused by loss of the SMN1 gene. Although lower motor neurons are a primary target, there is evidence that peripheral organ defects contribute to SMA. Current SMA gene therapy uses a single, high titre intravenous bolus of AAV9-SMN resulting in impressive, yet limited amelioration of the clinical phenotype. However, risks of this treatment include liver toxicity. Intrathecal administration is under clinical trial but was interrupted due to safety concerns in a concomitant animal study. As there is no direct comparison between the different delivery strategies while avoiding high dose toxicity, we injected SMA mice with low dose scAAV9-cba-SMN either intravenously (IV) for peripheral SMN restoration or intracerebroventricularly (ICV) for CNS-focused SMN restoration. Here, IV injections restored SMN in peripheral tissues but not CNS, while ICV injections mildly increased SMN in the periphery and the CNS. Consequently, only ICV treatment rescued motor neuron degeneration. Surprisingly, both treatments resulted in an impressive rescue of survival, weight, motor function, and peripheral phenotypes including liver and pancreas pathology. Our work highlights independent contributions of peripheral organs to SMA pathology and suggests that treatments should not be restricted to the motor neuron.
Introduction
Spinal muscular atrophy (SMA) is a devastating childhood neurodegenerative disorder characterized by the loss of lower motor neurons and skeletal muscle atrophy. Untreated and severely affected patients suffer from proximal and progressive muscle weakness, leading to complications such as scoliosis, respiratory failure, and early death 1. Recent therapeutic advances, notably in an AAV9-mediated gene therapy, are bringing hope to SMA patients, but toxicity concerns of the high viral dose required demand a closer look at the therapeutic potential of low dose treatment.
SMA is caused by deletion or mutation in the Survival of Motor Neuron 1 (SMN1) gene 2. The SMN protein product is essential and a complete SMN loss results in early embryonic lethality. SMA severity is thus mediated by a second gene, SMN2, which produces a limited amount of SMN protein. Recent advances in SMA therapeutics have changed the landscape of SMA disease prognosis as the use of therapeutics is becoming mainstream for SMA patients. There are currently three FDA-approved therapies available for SMA patients: nusinersen, risdiplam, and onasemnogene abeparvovec. While nusinersen and risdiplam act on the SMN2 gene to increase SMN protein production, onasemnogene abeparvovec uses a blood-brain-barrier penetrant adeno-associated virus serotype 9 (AAV9) vector carrying SMN cDNA to encode for the missing SMN protein. This gene therapy treatment is administered as a one-time intravenous (IV) injection, aiming to induce long-term expression of SMN protein in the periphery as well as the CNS 3,4. However, there is no bioavailability data in humans so far. It is also unknown how long expression lasts in different tissues because episomal viral DNA is diluted during cell division, and permanent expression will likely be restricted to post-mitotic cells such as neurons 5. Limited data from similar AAV gene therapies targeting mitotic cells has confirmed transgene expression and clinical benefit for up to 6 years after vector administration 6–8. So far, systemic delivery has been approved for therapy. However, CNS specific routes of administration are being explored for this therapy, as are various viral concentrations 9.
Though targeting of motor neurons is often the focus for SMA therapy, a wide range of non-neuronal tissues are affected in SMA patients and animal models. Muscle intrinsic defects 10–12, cardiac defects 13,14, fatty acid metabolism defects 15,16, glucose metabolism defects 17,18, immune organ defects 19–21 and gastrointestinal dysfunction 22,23 have been observed in SMA patients and characterized in mouse models of the disease. In fact, the most effective experimental gene therapy treatments in SMA pre-clinical models were delivered systemically with the use of ubiquitous promoters 4,24,25. However, safety concerns have arisen over the high doses of AAV required to transduce the CNS when administered intravenously. AAV vectors are known to have a higher affinity for transduction of the liver compared to other tissues 26, causing liver damage in some onasemnogene abeparvovec treated patients 4, while primates treated with a high dose of a similar construct in a pre-clinical trial demonstrated liver toxicity, liver failure, and neuronal degeneration 27. Alternative routes of administration of this drug are also being explored, though a trial evaluating cerebrospinal fluid delivery of onasemnogene abeparvovec was suspended due to concerns with neuronal toxicity in pre-clinical studies 9. Further, new data has shown that long-term overexpression of SMN has neurotoxic effects and leads to motor dysfunction over time in AAV9-SMN treated mice 28. With serious risks associated with high dosage via both routes of delivery, it is important to understand the benefits associated with an approach using low doses comparing CNS versus peripheral delivery. This will allow for the ideal route of delivery to be determined, as well as the development of therapies using lower doses and fewer safety risks.
We used a low dose of scAAV9-cba-SMN paired with either IV or ICV delivery to generate two different patterns of viral transduction. IV delivery induced a strong SMN-expression in the periphery while omitting the spinal cord. ICV scAAV9-cba-SMN application resulted in a modest SMN increase. Strikingly, both applications led to an impressive and similar effect on survival, weight gain, and motor function, underlining the importance of peripheral SMN expression. However, while there was a similar rescue effect on peripheral phenotypes such as liver and pancreas defects, motor neurons were rescued in ICV injected mice only. Together, these results show the efficiency of SMN gene-replacement therapy with a low viral dose. Moreover, rescue mechanism varies depending on the delivery route and is different between a central and peripheral treatment strategy. Further studies are required that explore a combination of low dose delivery routes in an scAAV9-cba-SMN treatment regimen to fully exploit the different route-dependent benefits while avoiding toxic effects.
Results
Intracerebroventricular scAAV9-cba-SMN administration to Smn2B/- mice results in a mild increase in SMN in the CNS and the periphery while intravenous injection results in restoration of SMN in peripheral tissues only
Our objective was to compare the phenotypic rescue of SMN replacement focused on the periphery with a replacement strategy primarily targeting the CNS. Therefore, we used a low dose of a blood-brain-barrier penetrant scAAV9 viral vector either administered IV or ICV to Smn2B/- mice. This virus expresses SMN under the control of a chicken beta actin promoter like the commercially available SMA treatment onasemnogene abeparvovec (Fig. 1A). To evaluate the tissue tropism of each application route, SMN protein levels in treated and untreated Smn2B/+ and Smn2B/- mice were obtained by western blot to compare between the neuromuscular system compartment (spinal cord and muscle) and the liver (the major peripheral target of AAV9 in patients). Both application routes lead to increased SMN protein, demonstrating the functionality of the virus (Fig. 1B,F). However, there was a difference in the tissue specific increases between routes of injection. ICV treatment produced a substantial yet mild increase of SMN protein levels in liver and spinal cord, and a trend towards an increase in muscle (Fig. 1B-E), while IV treatment completely restored SMN protein to the muscle and liver, but not the spinal cord (Fig. 1F-I). Not surprisingly, these results indicate that systemic scAAV9-cba-SMN administration more efficiently restores peripheral SMN levels compared to a CNS-directed approach. Importantly, systemic scAAV9-cba-SMN application at this titre omitted spinal cord transduction (Fig. 1I), which could only be achieved by a CNS-directed approach. Therefore, subsequent benefits in IV treated animals are most likely not mediated by a rescue of motor neurons.
Both routes of scAAV9-cba-SMN delivery significantly ameliorate SMA-like pathophysiology in Smn2B/- mice with more pronounced effects after systemic administration
To compare the pathophysiological impact of peripheral SMN restoration (IV treatment) vs. partial CNS and peripheral SMN restoration (ICV treatment), we evaluated survival, body weight, and motor function. Mice were monitored every 2 days for survival and weight gain and were subjected to three motor function tests to assess muscle strength (Fig. 2). Both IV and ICV treated Smn2B/- mice showed a full rescue of survival within the observational period, while untreated Smn2B/- mice had a median survival of 23 days (Fig. 2A). Both treatments produced a partial rescue in weight gain with no differences between treatment groups (Fig. 2B). IV and ICV treatment also significantly improved motor function scores. Though not significant, both treatments trended towards improved righting times. This became apparent at P19, where ICV and IV treated mice were able to immediately right themselves while Smn2B/- mice generally were not (Fig. 2C). In the pen test, both IV and ICV treated Smn2B/- mice had significantly longer balancing times than untreated Smn2B/- mice from P19 to P25 (Fig. 2D). Interestingly, in the mesh grip test, times were improved in IV treated Smn2B/- mice but not ICV treated mice compared to untreated Smn2B/- mice (Fig. 2E), suggesting greater strength recovery in distal muscles of IV treated mice. This indicates that important aspects of the SMA-like pathophysiology are influenced regardless of the locus of SMN restoration. Interestingly, IV treatment had more comprehensive impact on the motor functions emphasizing the importance of peripheral SMN restoration.
CNS but not peripheral scAAV9-cba-SMN delivery partially rescues spinal cord motor neuron degeneration in Smn2B/- mice
To understand the mechanisms which underly the pathophysiological changes, we next analyzed the pathological features by histology. Motor neuron number was investigated to determine the impact of IV or ICV scAAV9-cba-SMN treatment on motor neuron degeneration (Fig. 3). ICV treatment partially protected against motor neuron degeneration, while there was no difference in motor neuron number between IV treated Smn2B/- mice and untreated Smn2B/- mice (Fig. 3A-E). Additionally, we measured NfL plasma levels, which is a common outcome measure of neurodegeneration in the CNS and a candidate biomarker of neurodegeneration, disease state and therapeutic efficacy in SMA 29. As expected, blood plasma NfL levels were elevated in Smn2B/- mice, reflecting the motor neuron degeneration (Fig. 3F). Importantly, ICV treatment resulted in a reduced elevation in plasma NfL (Fig. 3F). In contrast, there was no protection afforded by IV treatment as evidenced by the relatively high blood plasma NfL levels, which is in line with a lack of SMN restoration in the spinal cord. The latter result is suggestive that IV-mediated beneficial effects on pathophysiology occur despite ongoing motor neuron degeneration.
CNS-directed scAAV9-cba-SMN injection in Smn2B/- mice rescues neuromuscular junction pathology better than IV treatment
Next, we analysed the impact of motor neuron degeneration and/or increased muscle intrinsic SMN levels on muscle function. While there was no change in fibre size in the TA muscle by any of the scAAV9-cba-SMN injections (Supplementary Fig. 1), our in-depth analyses of the NMJ pathology revealed significant differences. (Fig. 4). A variety of NMJ defects are present in Smn2B/- mice including abnormal endplate morphology, neurofilament accumulation, and denervation 30. Neurofilament accumulation and NMJ denervation were quantified to determine the degree of NMJ pathology after both routes of scAAV9-cba-SMN injection. Both IV and ICV treatments partially rescued neurofilament accumulation in the TVA of P19 Smn2B/- mice, with a higher efficacy in ICV-injected animals (Fig. 4A-D,I). This difference became more apparent when evaluating endplate occupancy, where only ICV injections resulted in a partial rescue (Fig. 4E-H, J), which is in line with an ICV mediated rescue of motor neurons and no rescue in IV injected animals. However, the mild effect of IV-delivered scAAV9-cba-SMN on neurofilament accumulation (Fig. 4I) points towards a muscle-intrinsic mechanism since there is no rescue in motor neuron numbers but an increased muscle SMN expression in this group (Fig. 3F and Fig. 1H).
IV and ICV scAAV9-cba-SMN treatment partially rescue peripheral organ defects in Smn2B/- mice
We examined peripheral tissues to evaluate if scAAV9-cba-SMN injections affect pathological aspects that may be independent of the neuromuscular disease mechanisms. Several metabolic defects have been observed in SMA patients and mouse models of SMA. Fatty acid metabolism defects can be observed in Smn2B/- mice through lipid accumulation in the liver 15, while glucose metabolism defects are observed through abnormal glucose homeostasis and pancreatic defects 17. Both IV and ICV scAAV9-cba-SMN administration prevented hepatic microvesicular steatosis in Smn2B/- mice (Fig. 5A-D), similar to what was previously observed with peripheral SMN restoration in this mouse model 31. Of note, the gastrointestinal tract of IV and ICVC treated mice also appeared to be healthy, unlike in untreated Smn2B/- mice where it shows signs of low motility and malfunction (data not shown). Moreover, Smn2B/- mice pancreata show a higher percentage of glucagon-producing alpha cells compared to insulin-producing beta cells and both treatments resulted in a partial rescue in the percentage of alpha cells within the islets compared to Smn2B/- mice (Fig. 5F-J). Accordingly, IV as well as ICV scAAV9-cba-SMN injections restored blood glucose levels to normal in Smn2B/- mice (Fig. 5E). In summary, ICV and IV injections resulted in the same degree of peripheral rescue, which is in line with the overall increase of SMN in the periphery in response to both application routes.
Discussion
Here, we evaluated the beneficial effects of a low dose scAAV9-cba-SMN injection in Smn2B/- mice, either ICV or IV, to compare a CNS-centred SMN restoration to a peripheral-centred SMN restoration. ICV injection produced a mild increase in SMN levels within the CNS as well as in peripheral tissues of these mice, although SMN was not fully restored to wild type levels. However, the mild increase in SMN was sufficient to produce a partial rescue of neuronal degeneration and a robust rescue of the NMJ pathology, resulting in a significant amelioration of survival, weight and motor function. IV injection of the scAAV9-cba-SMN completely omitted transduction of the CNS, but fully restored SMN levels in peripheral tissues. This was surprising because of the blood-brain-barrier penetrant properties of the AAV serotype. However, this enabled us to evaluate the contribution of a peripheral SMN reduction to the overall SMA-like pathology. The significant rescue observed in weight, survival, motor function, and metabolic defects in IV-treated animals was therefore likely a result of increased SMN in peripheral tissues. In accordance with a lack of central SMN restoration, there was no rescue of motor neuron degeneration. The modest rescue of NMJ neurofilament accumulation in IV-injected mice was possibly due to muscle-intrinsic SMN restoration. Both application routes showed the same degree of rescue of liver and pancreatic phenotypes. Overall, these results, summarized in Figure 6, emphasize the importance of SMN restoration to the peripheral organs and suggests further investigation into the motor neuron independent mechanisms contributing to weight, survival, motor function, and metabolic defects in SMA.
The impact of the loss of SMN protein on non-neuronal tissues is yet to be described in detail, but a variety of tissues are known to be affected in SMA patients as well as pre-clinical models of the disease 32. With the increasing use of SMN replacement therapy, it is important to understand the independent contributions of peripheral tissues to disease as well as the SMN protein requirements for these tissues. Our results agree with other published data that demonstrate the importance of treating the peripheral organs to achieve a full rescue. Knockdown of SMN protein in motor neurons was shown to cause an SMA-like phenotype in mice that was milder than that produced by ubiquitous knockdown 33, while pre-clinical nusinersen trials demonstrated 25-fold greater survival in mice receiving systemic treatment compared to CNS-restricted treatment 34.
A recent pre-clinical study demonstrated that AAV9-SMN gene therapy in SMNΔ7 mice restricted to neurons does not rescue the SMA phenotype in SMA mice, while ubiquitous expression does 35. Interestingly, this latter report demonstrated a mild increase in SMN within the spinal cord after IV injection of 4.5 × 1010 vg/mouse, a dose similar to our study. However, it is important to note that a different promoter, the phosphoglycerate kinase gene promoter, was used and that experiments on SMA mice using the cba promoter typically use a dose of at least 1×1011 vg/mouse to achieve transduction of the CNS 25,36. Our results emphasize the importance of targeting the peripheral organs when treating SMA and demonstrate the independent contributions of these tissues to SMA disease. This is also emphasized by clinical data, where patients treated with the CNS-specific treatment nusinersen can often remain severely disabled despite improvements in motor function, requiring ventilatory or nutritional support 37,38.
We were surprised to observe a slight rescue of NMJ neurofilament accumulation in our IV-injected mice, despite no restoration of SMN protein to the spinal cord in this group. This finding hinted towards possible muscle-intrinsic mechanisms of NMJ pathology, as IV injection fully restored SMN protein to muscle. However, studies have demonstrated a rescue of survival in SMA mice after muscle-specific expression of SMN, but no effect on NMJ pathology or synaptic function 39,40. NMJ pathology is instead likely related to SMN protein levels in motor neurons, and a consequence of motor neuron dysfunction 41. Interestingly however, motor neuron specific restoration of SMN in SMNΔ7 mice partially rescued endplate size and denervation, but not neurofilament accumulation 42. This agrees with our results that neurofilament accumulation may not be fully determined by the motor neuron, but also by other cells. Though neurofilament accumulation in neurodegenerative diseases is thought to be caused by hyper-phosphorylation of the protein within neurons 43, it is possible that terminal Schwann cells may have an effect on neurofilament phosphorylation and axonal transport 44. One explanation for the rescue of neurofilament accumulation in IV-injected mice may be that the scAAV9-cba-SMN was capable of transducing terminal Schwann cells to impact the degree of neurofilament accumulation in NMJs.
We observed a rescue in liver steatosis, with no apparent difference between the degree of rescue across the two methods of delivery. In Smn2B/- mice, liver steatosis and fatty acid metabolism defects lead to several functional defects including reduced protein production, impaired hemostasis, and reduced insulin like growth factor 1 (IGF1) levels 31. Rescue of liver function therefore likely allowed for overall better health and survival, and a possible restoration of IGF1 levels may also have contributed to improved growth and weight gain, although this needs to be experimentally verified. We also observed a rescue in blood glucose levels and a partial rescue in pancreatic defects. Smn2B/- mice demonstrate an imbalance between alpha and beta cells in the pancreatic islets, as well as glucose intolerance, hyperglucagonemia, and elevated glucose at P19 15,17. Rescue of these defects likely improved survival, as these metabolic changes are associated with higher mortality and sudden cardiac death 45. Further, the metabolic rescue observed in IV-treated Smn2B/- mice provides important evidence that the metabolic defects in SMA are motor neuron independent.
Unfortunately, systemic (IV) gene therapy presents significant risks for SMA patients due to liver toxicity and is limited in real world application due to the required high viral doses in older patients. Intrathecal application has been explored to allow for a lower dose to be used, reducing the risk of liver toxicity as well as the overall cost of the drug, but significant toxicity concerns have also been flagged for this method of delivery 28(p9). This treatment method also may prevent transduction of the periphery, limiting the effectiveness of the therapy. Intravenous treatment is therefore likely worth the risk, as it is becoming more and more apparent that targeting the peripheral organs is essential for complete treatment of SMA. An alternative approach could suggest pairing a low dose systemic administration with a low dose intrathecal administration, in order to target the periphery while allowing for a less potent dose to be delivered to motor neurons and avoiding the potential toxicity associated with overexpression of SMN.
As treatments extend the lives of SMA patients and continue to change the landscape and natural history of the disease, it will be important to focus on the quality of life of SMA patients. Current treatments are by no means a cure for SMA. Treated patients may begin to develop new symptoms and face obstacles that have not been faced before by untreated patients. The data herein highlight the importance of adopting a whole-body approach to SMA treatment, focusing on each specific tissue and its independent SMN requirements. As treated patients continue to age, it is likely that peripheral symptoms will become more apparent and have a larger impact on a patient’s health. Our results display that focusing on treating the peripheral organs will not only improve patients’ quality of life but also likely impact survival and overall health.
Materials and Methods
Animals
Smn2B/- (C57BL/6J background) mice were developed by our laboratory 46,47 and housed at the University of Ottawa Animal Care Facility. Smn2B/- mice were produced by breeding Smn+/- mice (C57BL/6J) to Smn2B/2B mice (C57BL/6J). The Smn2B/- mice are a model of SMA and asymptomatic heterozygous Smn2B/+ mice are used as controls in these experiments. Animals were cared for according to the Canadian Council on Animal Care.
scAAV9-cba-SMN treatment
The self-complementary AAV9-cba-SMN vector was produced as previously described and was titred by real time qPCR 31. The scAAV9-cba-SMN expresses human SMN under the control of a chicken beta actin (cba) promoter and was produced at a titre of 9.62 × 1013 viral genomes (vg)/mL. Mice at post-natal day 1 (P1) were administered scAAV9-cba-SMN through facial vein injection (5×1010 vg per pup administered in a volume of 20 μ L; four litters of 8-12 mice) or ICV injection (5×1010 vg per pup in a volume of 3 μ L; three litters of 6-12 mice). Two litters from each treatment were monitored for motor function, weight, and survival and sacrificed at P60. Mice were tattooed by Animal Care Facility staff around P4-6 to allow for a specific animal’s growth and motor function to be tracked over time. Mice were weighed every 2 days. Other groups were sacrificed at P19 for collection of blood and various tissues.
Motor function tests
Three motor function tests were performed. Righting reflex test (Treat-NMD SOP D_M.2.2.002, treat-nmd.org) evaluated overall body strength, pen test (Treat-NMD SOP SMA_M.2.1.001) assessed motor balance and coordination, and mesh grip test (Treat-NMD SOP SMA_M.2.1.002) evaluated limb strength. Motor function tests were performed as reported previously 48. Briefly, righting reflex test was performed from P7 to P19. In this test, the mouse is placed on its back on a flat surface and the time to right itself is measured (up to a maximum of 30 sec). Pen test involves placing the mouse on a pen and recording the length of time they are able to balance. This test was performed from P19 to P25. Mesh grip test was performed from P13 to P25. This test measures the strength of a mouse’s limbs by timing their latency to fall when suspended from a mesh. Tests were performed every 2 days. Of note, tests were performed by two different evaluators using the lab’s controlled and established protocol.
Blood collection and plasma analysis
Mice were euthanized at P19 by decapitation after anaesthesia in a CO2 chamber. Upon decapitation, blood was collected using Microcuvette CB 300 K2E tubes (Sarstedt, Newton, NC) coated with K2 EDTA. Samples were spun at 2 000 g using 5424R centrifuge (Eppendorf, Hamburg, Germany) for 5 min at room temperature. Forty-five L of the plasma supernatant was then collected in a 1.5 mL Eppendorf tube and stored at -80° C. Samples were thawed on ice and stored at 4° C the day before the assay was to be performed. Samples were analyzed using the Simoa NF-Light ® assay (Quanterix, Billerica, MA) on a Simoa HD-1 analyzer to determine the concentration of neurofilament light chain (NfL) protein. Blood glucose was measured upon blood collection using a Freestyle Precision Neo meter with Freestyle Precision Blood Glucose Test Strips (Abbott, Chicago, IL). About 1 μ L of blood was extracted from the collection tube and applied to the test strip to measure blood glucose concentration.
Tissue processing and staining
After euthanasia at P19, liver and tibialis anterior (TA) muscles were fixed in 1:10 dilution buffered formalin (Thermo Fisher Scientific, Waltham, MA) for 48 h at 4°C and then transferred to 70% ethanol at 4°C until processing. Pancreata were fixed in 4% paraformaldehyde (PFA) for 48 h at 4°C and then transferred to 70% ethanol at 4°C until processing. Lumbar spinal cord was fixed in 4% PFA overnight at 4°C then prepared for cryosectioning as previously described 48. The abdominal musculature was dissected and fixed in 4% PFA and the transversus abdominis (TVA) was separated from the abdominal musculature as per 49.
TA, liver, and pancreas samples were processed at the University of Ottawa Department of Pathology and Laboratory Medicine and embedded in wax using a LOGOS microwave hybrid tissue processor (Milestone Medical, Kalamazoo, MI). Paraffin block tissues were cut with a microtome at 3-4 µm thickness. Sections of TA and liver were stained with hematoxylin and eosin (H&E) using an XL CV5030 autostainer (Leica, Wertzler, Germany). Samples stained with H&E were scanned with a MIRAX MIDI digital slide scanner (Zeiss, Oberkochen, Germany). Images were acquired using Panoramic Viewer 1.15.4 (3DHISTECH, Budapest, Hungary) at different magnifications. Muscle fibre size was quantified using ImageJ (version 1.53). Approximately 100-200 fibers were counted in different areas of the muscle section to ensure appropriate coverage. The area of each fibre was measured to calculate a mean fibre size for each animal. Sections of pancreas were deparaffinized in 3 washes of xylene substitute Histo-Clear (National Diagnostics, Atlanta, GA) for 5 min each followed by 2 washes of a 50/50 mixture of absolute ethanol and Histo-Clear for 5 min each. Slides were gradually rehydrated in 100% (v/v), 95% (v/v), 70% (v/v), 50% (v/v), and 0% (v/v) ethanol. Slides were incubated in 0.5% Triton-X-100 (Millipore Sigma, Burlington, MA) in PBS for 5 min, washed 3× with PBS, then blocked in 20% goat serum, 0.3% Triton-X-100 in TBS for 2 h. Slides were incubated with primary antibodies for insulin and glucagon (Supplementary Table 1) in 2% goat serum, 0.3% Triton-X-100 in TBS overnight at 4°C and then washed 3× with PBS. Slides were incubated with secondary antibodies (Supplementary Table 1) in 2% goat serum, 0.3% Triton-X-100 for 1 h and then washed 3× with PBS. DAPI (1:1 000) was added to the last PBS wash, followed by the slides being mounted in Dako Fluorescent Mounting Medium (Agilent, Santa Clara, CA). Images were taken with an Axio Imager M2 microscope (Zeiss), with a 20x objective, equipped with filters suitable for DAPI/ fluorescence.
Lumbar SC was prepared for choline acetyltransferase (ChAT) staining of motor neurons as previously desribed 48. The number of ChAT positive motor neurons per ventral horn was recorded for ten different sections per animal, each separated by at least 100 μ m to avoid counting the same motor neuron twice. TVA muscles were dissected and stained for neurofilament and synaptic vesicle protein to visualize neuromuscular junctions (NMJs) as before with slight alterations 49. Briefly, the abdominal musculature was dissected and fixed in 4% PFA and the TVA was separated from the abdominal musculature under a dissection microscope. The TVA was incubated with a tetramethylrhodamine (TRITC) conjugated bungarotoxin for 30 min at room temperature. The tissue was then incubated overnight at 4°C with primary antibodies for neurofilament and synaptic vesicle protein 2 (Supplementary Table 1). Tissues were then incubated in secondary antibodies (Supplementary Table 1). Tissues were mounted with Dako Fluorescent Mounting Medium (Agilent) and imaged using Axio Imager M2 microscope (Zeiss) with Z-stack feature at 40x magnification. At least 40 NMJs were counted for each animal. Each NMJ was quantified as either normal or displaying neurofilament accumulation, and endplates were noted as either occupied or unoccupied.
Western blot
Tissue processing and immunoblotting was performed as previously described with slight alterations 19. After euthanasia at P19, thoracic spinal cord, liver, and TA muscle were dissected, and flash frozen in Microvette CB 300 Z tubes (Sarstedt) in liquid nitrogen. Protein was extracted from frozen tissue by homogenization of tissue with RIPA lysis buffer and PMSF (Cell Signalling, Danvers, MA). Protein concentrations of samples were determined using Bradford Assay (Bio-Rad, Hercules, CA). Twenty μ g of protein was loaded onto a 12% acrylamide gel and subject to sodium dodecyl sulfate polyacrylamide gel electrophoresis. Proteins were transferred to a PVDF membrane (Immobilon-P or Immobilon-FL, Millipore, Burlington, MA) and blocked for 1 h at room temperature in 5% (w/v) milk powder in TBS-T or Odyssey blocking buffer (Li-Cor, Lincoln, NE). See Supplementary Table 1 for primary and secondary antibodies. Signals were detected with Odyssey CLx (Li-Cor) or using Pierce ECL Western Blotting Substrate (Thermo Scientific) and SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Scientific). Densitometry of western blot bands was performed using ImageJ (v. 1.53). Raw values were normalized by geometric mean and used for subsequent housekeeping normalization (α-tubulin values) from the same blot.
Statistical analysis
Data are presented as the mean standard error of the mean. One-way ANOVA with Tukey’s post-test or Two-way ANOVA with Bonferroni post-test were used to compare multiple means. Statistical tests were performed using GraphPad Prism 5. Significance was indicated by * for P≤0.05, ** for P≤0.01, and *** for P≤0.001. Images were blinded prior to quantification.
Funding
RK was supported by Muscular Dystrophy Association (USA) (grant number 575466); Muscular Dystrophy Canada; and Canadian Institutes of Health Research (CIHR) (grant number PJT-156379). AR was supported by a CNMD STAR Award from the University of Ottawa Brain and
Mind Institute. MOD was supported by Frederick Banting and Charles Best CIHR Doctoral Research Award.
Competing Interests
RK received honoraria and travel accommodations from Roche as an invited speaker at their global and national board meetings in 2019. RK and the Ottawa Hospital Research Institute have a licensing agreement with Biogen for the Smn2B/- mouse model. MOD received honoraria and travel accommodations from Biogen for speaking engagements at the SMA Summit 2018 held in Montreal, Canada and SMA Academy 2019 held in Toronto, Canada. These COI are outside the scope of this study. All other authors have no competing interests to declare.
Author contributions
AR and RK designed research
AR, MOD, AB, RY, ST, and DRT performed experiments
BLS and VTC provided material support
AR, MOD, and NH analyzed the data
AR, NH, and RK wrote the manuscript with input from all authors
RK designed the study
Acknowledgements
We thank Dr. Ronald Booth and Adrienne Rowan from the EORLA at The Ottawa Hospital for assistance with some of the NfL assays.