Research ArticleSpinal Muscular Atrophy

Systemic Delivery of scAAV9 Expressing SMN Prolongs Survival in a Model of Spinal Muscular Atrophy

See allHide authors and affiliations

Science Translational Medicine  09 Jun 2010:
Vol. 2, Issue 35, pp. 35ra42
DOI: 10.1126/scitranslmed.3000830


Spinal muscular atrophy is one of the most common genetic causes of death in childhood, and there is currently no effective treatment. The disease is caused by mutations in the survival motor neuron gene. Gene therapy aimed at restoring the protein encoded by this gene is a rational therapeutic approach to ameliorate the disease phenotype. We previously reported that intramuscular delivery of a lentiviral vector expressing survival motor neuron increased the life expectancy of transgenic mice with spinal muscular atrophy. The marginal efficacy of this therapeutic approach, however, prompted us to explore different strategies for gene therapy delivery to motor neurons to achieve a more clinically relevant effect. Here, we report that a single injection of self-complementary adeno-associated virus serotype 9 expressing green fluorescent protein or of a codon-optimized version of the survival motor neuron protein into the facial vein 1 day after birth in mice carrying a defective survival motor neuron gene led to widespread gene transfer. Furthermore, this gene therapy resulted in a substantial extension of life span in these animals. These data demonstrate a significant increase in survival in a mouse model of spinal muscular atrophy and provide evidence for effective therapy.


Spinal muscular atrophy (SMA) is an autosomal recessive neurodegenerative disorder characterized by progressive muscular weakness and hypotonia as a consequence of the loss of lower motor neurons. On the basis of the age of onset and the severity of the neuromuscular symptoms, four clinical phenotypes have been described (1). The most severe form, type 1 SMA, is a devastating childhood condition also known as Werdnig-Hoffmann disease. The disease begins by 6 months after birth and is fatal by 3 years of age (1). In 1995, the gene responsible for most cases of SMA, survival motor neuron (SMN), was identified at the chromosomal locus 5q13 (2). The human gene is duplicated with telomeric and centromeric copies, SMN1 and SMN2, respectively. The telomeric and centromeric copies of the SMN gene share >99.8% sequence homology. SMA is caused by mutations or deletion of the SMN1 gene, leading to depletion of SMN protein, because SMN2 fails to generate sufficient amounts of full-length protein. In particular, a C-to-T substitution in exon 7 affects its splicing, with subsequent exon 7 skipping (3), thereby generating a truncated, unstable form of SMN (SMNΔ7). SMN is a ubiquitously expressed gene, and it is unclear why the disease affects only the lower motor neurons. Mice have a single smn gene, the loss of which leads to early embryonic lethality (4). Transgenic mice (SMNΔ7) with the clinical features of SMA were generated by introducing copies of human SMN2 and SMNΔ7 into mice lacking endogenous smn (5) and have been extensively used for preclinical studies (69).

We have previously demonstrated that lentiviral vectors encoding SMN (10) could be successfully used to enhance SMN protein concentrations within motor neurons in the SMNΔ7 mouse model (5) with concomitant reduction in motor neuron death. However, the effect of this treatment on mouse survival was modest, with an increase in life expectancy of only up to 5 days compared to that of controls (10). It has been recently reported that motor neurons in mice require high concentration of SMN protein for normal function (11). We reasoned that further extension of survival by SMN replacement would be likely to require a gene-targeting strategy that increased expression of the SMN protein and resulted in more widespread distribution of the transgene.

The overall objective of the current study was to optimize SMN expression and enhance the distribution of the SMN protein in murine models of SMA. Self-complementary adeno-associated vector serotype 9 (scAAV9) mediates efficient and sustained transgene expression in spinal motor neurons after systemic administration in neonatal mice (12, 13). The advantage of using scAAV over traditional single-stranded DNA AAV is that it allows a faster expression of the transgene in vivo (14, 15).

Here, we have generated scAAV encoding a codon-optimized version of the SMN1 complementary DNA (cDNA) (16) and used it to enhance SMN protein expression more effectively than previously achieved in the murine spinal cord (10). We demonstrate that a single intravenous injection of scAAV9-coSMN (codon-optimized SMN) rescues the lethal phenotype and substantially improves survival in SMNΔ7 mice.


Codon optimization enhances SMN protein expression in SMA type 1 fibroblasts

Previous studies suggest that a high copy number of SMN2 can rescue the smn−/− mice because sufficient SMN protein is produced in motor neurons (17). We therefore hypothesized that there is a threshold amount of SMN that needs to be expressed in motor neurons to support their normal function. The marginal effect on survival of SMN2+/+, SMNΔ7+/+, smn−/− (defined as SMNΔ7) reported in our previous work (10) might be due to insufficient SMN replacement in motor neurons. It has been reported that codon optimization of genes enhances protein expression (18). Here, we have generated a coSMN construct (Fig. 1A). The SMN protein expressed from coSMN cDNA displays 100% amino acid homology with the human protein. In a first step, we constructed lentiviral vectors harboring the human wild-type SMN (LV-wtSMN) or codon-optimized SMN (LV-coSMN). We compared the ability of these constructs to express SMN in SMA type 1 GM03813 fibroblasts. We used fibroblasts from type 1 SMA patients because these cells produce little or no SMN protein (19). Transduction of fibroblast cells was achieved with multiplicity of infection (MOI) of 20. High SMN protein concentrations were detected in fibroblasts that had been incubated with both constructs for 7 days (Fig. 1B). Furthermore, immunoblotting with antibodies to SMN revealed that SMN protein concentrations were up to 10-fold higher in LV-coSMN–transduced cells than in untransduced, control fibroblasts (n = 4, P < 0.001) (Fig. 1C). This level of overexpression is about twice as high as that achieved on LV-wtSMN transduction (n = 4, P < 0.05) (Fig. 1C). On the basis of this finding, we carried out all our studies using the most efficient construct (coSMN).

Fig. 1

Codon optimization enhances gene expression in SMA fibroblasts. (A) Alignment of the human SMN and coSMN. (B and C) Immunocytochemistry (B) and Western blot analysis (C) of SMN expression in SMA fibroblasts transduced with either LV-coSMN or LV-wtSMN 7 days after transduction (n = 4). *P < 0.05; ***P < 0.001, ANOVA followed by Bonferroni post hoc test. (D and E) Immunocytochemistry (D) and Western blot analysis (E) of SMN expression in SMA type 1 fibroblasts transduced with scAAV9-coSMN 6 days after transduction.

Widespread delivery of therapeutic genes to the central nervous system (CNS) is a challenge in SMA therapy for many reasons. First, delivery of therapeutic genes in the CNS parenchyma is restricted by the blood-brain barrier (BBB). Second, because of the severe nature of SMA and the possible developmental deficits, SMN replacement would be desirable at early stages of the disease. We therefore needed to choose a vector system that crosses the BBB and that allows fast, robust, and long-lasting gene transfer in neonatal mice. scAAV9 fulfils all these criteria (12, 20) and mediates efficient gene transfer in mice after systemic delivery. We generated scAAV9 that expressed human coSMN (scAAV9-coSMN) for testing in in vivo efficacy studies. Before in vivo gene transfer in SMNΔ7 mice, we first established that cells, transduced with scAAV9-encoding coSMN gene, produced sufficient SMN protein. In vitro transduction of the SMA type 1 GM03813 fibroblasts with scAAV9-coSMN indeed resulted in the restoration of SMN protein concentrations in this cell line (Fig. 1, D and E). Together, these data indicate that scAAV9 expressing SMN restores gems (Gemini of coiled bodies) with high efficiency in SMA patient–derived fibroblasts.

scAAV9-coSMN gene delivery rescues the SMNΔ7 mouse model

To explore the efficacy of SMN gene replacement on motor neuron survival in the SMNΔ7 mouse model, we injected scAAV9-coSMN (n = 9) or scAAV9-GFP (green fluorescent protein) (n = 9) vectors systemically into the facial vein of postnatal day 1 (PN1) SMNΔ7 mice. An SMNΔ7-untreated group (n = 10) was included as control. We intentionally chose this route of delivery, because systemic delivery of scAAV9 resulted in very efficient transduction of the CNS in SMNΔ7 mice (Fig. 2), a well-established model of SMA. SMNΔ7 mice display progressive muscle weakness and die on average 2 weeks after birth (5). In particular, daily measurement of body weight revealed slowing of growth and body weight decrease from PN6 (Fig. 3, A and B) (5, 21, 22). Inspection of the phenotype of these neonatal mice revealed that scAAV9-coSMN–treated SMNΔ7 animals displayed continuous body weight gain and a phenotype comparable to SMN2+/+, SMNΔ7+/+, smn+/− (defined as carrier) littermates (Fig. 3, A and B), whereas scAAV9-GFP–injected and untreated SMNΔ7 mice became immobile and less active and showed signs of muscle atrophy (Fig. 3A). Furthermore, scAAV9-coSMN–treated mice became progressively able to successfully complete a negative geotaxis reflex task (Fig. 3C), in which the animals are placed facing down on a steep surface and the time spent to turn themselves 180° and start climbing is recorded (21). To be considered successful, the task had to be completed within a maximal time of 60 s. Although only up to 40% of scAAV9-coSMN SMNΔ7 mice were scored as fully completing this task, the remaining 60% could still grip the surface, whereas none of the control SMNΔ7 pups could grip the surface after PN8.

Fig. 2

scAAV-mediated gene transfer in SMNΔ7 mice. (A) Colocalization of GFP and CGRP in lumbar spinal motor neurons. Scale bars, 100 μm (top panel), 20 μm (bottom panel). (B) Western blot analysis of SMN expression in different tissues 24 days after injection. (C) Immunohistochemistry showing extensive overexpression of SMN in lumbar motor neurons of SMNΔ7 mice 60 days after scAAV-coSMN administration. Scale bars, 20 μm.

Fig. 3

A single scAAV9-coSMN injection improves neonatal motor function and rescues fatalities in SMNΔ7 mouse model of SMA. (A) Pictures showing the same SMNΔ7 pups at different ages, as indicated. The pups were injected at PN1 with either scAAV9-coSMN or scAAV9-GFP. For reference, a gender-matched carrier littermate was included. (B) Body weight growth (PN0 to PN21) in SMNΔ7 injected with scAAV9-GFP or scAAV9-coSMN. An untreated group was included as control. *P < 0.05; ***P < 0.001, scAAV9-coSMN versus no virus mice; #P < 0.05, ###P < 0.001, scAAV9-coSMN versus carrier mice, two-way ANOVA with Bonferroni post hoc test. (C) Negative geotaxis test highlighting significant improvement in motor function of scAAV9-coSMN compared to controls. P < 0.0001, two-test for trend. (D) Kaplan-Meier cumulative survival curves in the different experimental groups. The crosses identify mice that were euthanized for humane reasons before they displayed any major motor neuron impairment. The number indicates the mouse ID. See text for details.

We saw a robust increase in survival of the SMNΔ7 mice after systemic delivery of scAAV9-coSMN (n = 9) (Fig. 3D). Survival analysis revealed that 80% of scAAV9-coSMN–treated animals showed an extension in the life span (69.1 ± 14.2.8, n = 8) (one mouse is still alive with a maximal survival of >190 days), whereas untreated and scAAV9-GFP–injected SMNΔ7 transgenic mice die on average at 13.9 ± 0.99 (n = 10) and 11.2 ± 1.32 (n = 9) days of age, respectively. Only one scAAV9-coSMN–treated SMNΔ7 mouse showed SMA-like symptoms with modest extension in survival and was humanely euthanized at PN16 (Fig. 3D). To explore the reason behind this early death, we carried out Western blot analysis to determine the amount of SMN protein in this mouse. This analysis revealed poor gene transfer in this particular animal (fig. S1), suggesting probable technical failure during intravenous administration of scAAV9-coSMN. Other mice did not die as a consequence of motor impairment because the motor function and behavior of these animals remained comparable to those of carrier littermates (Fig. 3D). In particular, mouse S74 (Fig. 3D) suddenly looked moribund at PN24 and was therefore culled. Macroscopic postmortem examination of its tissues revealed gut hemorrhage, whereas Western blot analysis of SMN protein demonstrated a significant increase in protein concentration in several tissues, achieving protein concentrations comparable to those of a carrier littermate (Fig. 2B). Mouse S45 (Fig. 3D) showed significant hyperactivity and was humanely euthanized at ~60 days of age after advice from the veterinary surgeon, in adherence with UK regulations. Mice S46, S59, S60, and S71 (Fig. 3D) displayed sudden respiratory distress and behavioral abnormality, with disinterest and reduced provoked response, and were therefore culled. Macroscopic postmortem examination of their tissues revealed heart enlargement. Histological analysis demonstrates pathology of the heart, characterized by patchy interstitial fibrosis of the ventricles, particularly affecting the inner part of the ventricular wall (fig. S2A). We saw increased fiber size variation (fig. S2A). There was no fiber necrosis. Small blood vessels within the myocardium and heart valves appeared unremarkable. Measurement of copy number revealed that these mice express two copies of SMN2.

scAAV9-mediated efficient gene transfer in neonatal SMNΔ7 mice

To maximize gene transfer, we injected 10 μl of high-titer scAAV vector solutions, corresponding to ~1011 vector genome, systemically into the facial vein of SMNΔ7 mice at PN1. scAAV9-GFP–injected mice (n = 9) were killed when they reached the clinical end point (11.2 ± 1.32 days of age). Gene transfer was not specific to the CNS, because GFP expression was detectable in other organs; the strongest expression was in the heart and liver (fig. S2B). Extensive reporter gene GFP expression was observed in the spinal cord (fig. S2B). Moreover, when staining for calcitonin gene–related peptide (CGRP) to identify motor neurons more directly, we were able to assess the gene transfer efficiency in spinal motor neurons (Fig. 2A). Consistent with previous reports (12, 20), CGRP labeling of spinal cord sections from scAAV9-GFP–injected mice revealed scAAV9-mediated efficient gene transfer to spinal motor neurons. Cell counts showed that 66.5, 45, and 55% of motor neurons were transduced in lumbar, thoracic, and cervical spinal cord, respectively. To monitor transgene expression in scAAV9-coSMN–treated mice, we used antibodies to SMN to perform immunohistology and Western blot analysis. At 24 days after injection of scAAV9-coSMN into the facial vein, overexpression of SMN protein was seen in the lumbar spinal cord, muscle, and liver (Fig. 2B). The gene transfer efficiency of scAAV9-coSMN was confirmed by immunofluorescence in spinal motor neurons (Fig. 2C). All these findings thus indicate that we achieved a high gene transfer efficacy through systemic delivery of the scAAV9 vector in the SMNΔ7 mouse model.

scAAV9-encoding coSMN improves the phenotype of the SMNΔ7 mouse model

Rescued SMNΔ7 mice were weaned at 21 days of age. We continued to monitor their growth and health into adulthood (Fig. 4, A and B). Although the rescued mice were slightly smaller than their carrier littermates, their body weights continued to increase normally (Fig. 4B). The scAAV9-coSMN–treated mice had a normal appetite, groomed their fur, and were as mobile and spontaneously active as wild-type mice. Their skin (paws, mouth) had a normal color. However, we observed a slight unsteadiness of gait and occasional hindlimb splay reflex impairment. To objectively quantify the neuromuscular performance of treated SMNΔ7 mice, we used the CatWalk gait analysis system and rotarod testing after weaning the animals (Fig. 5). Motor function, assessed by the rotarod task, showed no significant motor impairment in the rescued mice compared to carrier littermates (Fig. 5A). We observed some variability in scAAV9-coSMN–treated mice on the rotarod, which we attribute to their hyperactivity and preference to jump from the apparatus rather than to loss of motor function (Fig. 5A). Gait analysis confirmed lack of major motor performance deficits in rescued mice (Fig. 5, B to D). However, the gait analysis highlighted a postural difference, namely, a generally reduced interpaw distance specifically between the hind paws (fig. S3). Together, these data show that the motor function of scAAV9-coSMN–injected mice closely resembles the performance of carrier littermates.

Fig. 4

Rescued SMNΔ7 mice show a phenotype comparable with their carrier littermates. (A) Picture showing a rescued mouse at PN56 with his gender-matched carrier littermate. Note the short tail of SMNΔ7-treated mouse with scAAV9-coSMN. (B) Body weight growth in rescued and their smn+/− littermates. *P < 0.05; **P < 0.01; ***P < 0.001, two-way ANOVA followed by Bonferroni post hoc test.

Fig. 5

Motor function assessment in rescued SMNΔ7 mice. (A) Best performance in an accelerated rotarod test highlights a higher variability in rescued mice in comparison with their littermates. *P < 0.05, two-way ANOVA followed by Bonferroni post hoc test. (B) Representative gait traces analyzed by the CatWalk software. (C and D) Stride length measurement shows a progressive increase in both front limb (C) and hindlimb (D) for both rescued and carrier mice.

The rescued mice displayed a distinctive phenotype starting from PN18. Their tails were shorter and thicker than those of the carrier littermates. A similar feature was also observed in a previous mouse model (Smn−/− with more than two copies of SMN2) of moderate SMA (23) or in more severe mouse models (Smn−/− with two copies of SMN2) in which the life span was prolonged with either a genetic (11) or a pharmacological approach (24). Furthermore, all the animals at around PN40 displayed ear inflammation starting from the edge and then spreading inward through the tissue of the pinna. The ear skin became necrotic by PN60 (Fig. 4A). Distal necrosis has also been recently reported in two children affected by severe SMA (25) and in SMNΔ7 mice whose life span was extended by pharmacological and nutritional care (24). Finally, moderate swelling and inflammation of the eyelid rim was observed in seven of eight rescued mice. Curiously, this phenomenon seemed to be restricted to the eye opposite to the vein that was originally injected. These observations were not reported in scAAV9-coSMN– or scAAV9-GFP–injected carrier mice, which suggests that the tail and ear phenotypes neither are caused by SMN overexpression nor are an immune response to AAV9 but are an intrinsic part of the phenotype of the SMNΔ7 mice with extended survival.


Our gene therapy strategy can overcome two potential major bottlenecks in SMA therapy: (i) generation of a vector capable of mediating the high SMN expression necessary for normal motor neuron function and (ii) identification of a vector system that can efficiently restore SMN protein widely in the CNS and other vital organs. Together, the most important findings of our study are as follows. (i) Codon optimization of the SMN cDNA allows enhancement of protein expression. When using the scAAV9 vector system encoding the coSMN sequence, we achieved substantial restoration of SMN protein expression in the type 1 SMA fibroblasts. These vectors can thus be successfully used for transfer of therapeutic genes to motor neurons for treatment of motor neuron degeneration. (ii) We report here that systemic delivery of scAAV9-coSMN can reverse the phenotype and induce a large increase in the life expectancy of a clinically relevant rodent model of type 1 SMA. This approach achieves a substantial therapeutic effect on survival in the SMA mouse model. In a previous study (10), gene transfer of SMN with a rabies G–pseudotyped lentiviral vector, which is retrogradely transported from muscle to spinal motor neurons, had only a marginal therapeutic effect on the course of SMA disease. To explain the minimal effect on survival in this previous report, we propose the following: (i) Because the approach described previously exclusively targets motor neuron populations, SMN expression may need to be restored not only in motor neurons, but also in other cell types and organs for effective treatment; and (ii) the amount of SMN restored in motor neurons was not sufficient to establish normal function and survival of motor neurons. Our current gene therapy strategy, based on systemic delivery of scAAV9 associated with codon optimization of SMN, can overcome these two therapeutic bottlenecks.

SMA is currently incurable mainly because of the severity of the effects on the neuromuscular system and a lack of methods to efficiently deliver therapeutically attractive molecules to the CNS. We believe that our findings are noteworthy because SMNΔ7 transgenic mice develop a severe phenotype only a few days after birth, first by a decrease in their body weight, and then by motor neuron degeneration at 9 days of age. As the disease progresses, they also develop proximal muscle weakness and atrophy, resulting in end-stage paralysis and death at ~13.27 days of age (5). The SMNΔ7 model is widely used for therapeutic testing (69). So far, only a few gene therapy investigations have been reported in this field (10, 26, 27). Our treatment with scAAV9-coSMN gene transfer robustly reversed the phenotype and increased survival in 80% of treated SMNΔ7 mice. Treatment was effective in rescuing the mice from clinical signs such as muscle weakness and atrophy, immobility, respiratory distress, weight loss, and paralysis. Only one of nine scAAV9-coSMN–treated animals displayed signs of disease and was humanely euthanized by 16 days of age. This animal was ineffectively treated because it exhibited low concentration of SMN protein and RNA, as revealed by Western blotting and quantitative polymerase chain reaction (qPCR) analysis. The most likely explanation for this outcome is a technical failure during intravenous delivery of the therapeutic vector in this PN1 pup.

This study highlights the considerable therapeutic potential of scAAV9 for treating human SMA. Furthermore, the noninvasive mode of vector delivery used in this study and the high efficacy achieved provide a realistic and strong rationale for advancing vector-based SMN replacement approaches toward clinical trials in human patients. However, translating our strategy to human clinical trial will require further preclinical studies in rodents and nonhuman primates. These preclinical studies will provide useful information about dosing, toxicology, and the time window when the treatment could be applied in babies with SMA.

Note added in proof: During the evaluation and revision of this paper, a similar result has been published (29).

Materials and Methods

Viral production

coSMN (generated by Geneart AG) or human SMN cDNA was cloned between the Bam HI and the Xho I sites of a lentiviral vector genome. Self-inactivating lentiviral vector stocks pseudotyped with the vesicular stomatitis virus glycoprotein envelope protein were prepared by transient calcium phosphate transfection of the human embryonic kidney 293T cell line as described (28). Viral titers were estimated by quantifying the p24 capsid protein by enzyme-linked immunosorbent assay (28).

To generate scAAV9 for in vivo procedures, we used an scAAV plasmid coding for GFP under the control of the cytomegalovirus promoter. The coSMN sequence was cloned between the Age I and the Sbf I restriction sites. High-titer scAAV9 vectors were then prepared by Virapur with a three-plasmid transient cotransfection system involving a plasmid encoding the Rep2Cap9 sequence (pAAV2/9), a helper plasmid (pHelper, Stratagene), and the vector genomes (scAAV-coSMN or scAAV-GFP).

Cell culture

GM03813 fibroblasts derived from an SMA type 1 patient (19) (Coriell Cell Repository) were used to demonstrate gene transfer efficiency from different viral vectors. In particular, lentiviral vectors were used at an MOI of 20, whereas for scAAV9, 2 × 1010 vector genomes were used to transduce 5 × 104 cells. Seven days after transduction, the cells were either fixed with 4% paraformaldehyde or lysed for protein extraction.


All the procedures involving animals were performed according to the UK Home Office regulations. SMNΔ7 mice (5) were purchased from The Jackson Laboratory (stock 005025) and maintained in a controlled facility in a 12-hour dark-light photocycle with free access to food and water. Carrier animals were used for breeding, and the offspring were genotyped immediately after birth by PCR amplification of the transgenes according to the protocols provided by The Jackson Laboratory. SMNΔ7 copy number was verified as described (5).

SMNΔ7 mice were injected in the facial vein under isoflurane anesthesia with 1011 vector genome of either scAAV9-GFP (n = 9) or scAAV9-coSMN (n = 9). The mice were then allowed to recover, rolled in the sawdust from their original cage, and immediately returned to their cage. No episodes of litter exclusion were observed. To check the motor function of the animals, we performed a battery of behavioral tests (21) on a daily basis. In particular, the negative geotaxis reflex was used as described (21). Starting from PN21, the motor function of rescued SMNΔ7 mice was assessed in comparison with that of their carrier littermates every 7 days. To evaluate motor coordination and balance, we placed the mice on a rotarod treadmill and the rotation speed was smoothly increased from 4 to 40 rpm over 5 min. Each animal was assayed twice per session, recording the best performance. Furthermore, the gait was analyzed by computerized footprint recording. Briefly, the footage was recorded by a camera placed under a glass surface on which the mice were allowed to walk freely. CatWalk 7.1 software (Noldus) was then used to calculate several gait parameters.

Western blotting

Proteins were extracted from either cells or tissues (5) as described, electrophoretically separated onto a 10% SDS-polyacrylamide gel, and transferred onto a polyvinylidene difluoride membrane (Millipore). The following primary antibodies were used: mouse antibody to α-tubulin (1:2500 to 10,000; Calbiochem), mouse antibody to SMN (1:2500; BD Transduction), and rabbit antibody to GFP (1:2500; Clontech) followed by horseradish peroxidase–conjugated goat antibody to mouse (1:10,000; Bio-Rad) or to rabbit (1:4000; DakoCytomation).

Real-time QPCR

RNA was extracted from animal tissues with the SV Total RNA Isolation System (Promega) according to the manufacturer’s protocol. RNA (0.5 to 1 μg) was then retrotranscribed to cDNA with random hexamers by means of SuperScript III First-Strand Synthesis System (Invitrogen). The samples were diluted to 12.5 ng/μl and used as a template for real-time QPCR. To detect the presence of coSMN, we used the following primers at a concentration of 300 nM: forward, 5′-TCTGCGAAGTGGCCAACAA-3′; reverse, 5′-CACTTGGCTCTCGTTCTCGTTT-3′. β-Actin messenger RNA was used as the housekeeping transcript, and it was detected with the forward (5′-CAATGAGCGCGGTTCCGATACC-3′) and the reverse (5′-TCAACGTCACACTTCATGATGGA-3′) primers at a concentration of 300 nM. The reaction was performed with Brilliant II SYBR Green QPCR master mix (Stratagene) in a MX3000P Real-Time PCR System (Stratagene). Briefly, after an initial denaturation step (10 min, 95°C), 40 cycles were performed at 95°C for 30 s followed by 60°C for 1 min followed by a ramp increase in temperature up to 95°C to generate dissociation curves. The relative expression was calculated with the ΔΔCt method (ABI PRISM 7700 Sequence Detection System protocol, Applied Biosystems).

Immunocytochemistry and immunohistochemistry

Fibroblasts were stained with mouse antibody to SMN (1:1000) followed by Cy3-conjugated goat secondary antibody to mouse (1:200; The Jackson Laboratory) and nuclear counterstaining with Hoechst 33342 (10 μg/ml; Sigma). Animals were terminally anesthetized with pentobarbital and then transcardially perfused with phosphate-buffered saline (PBS) supplemented with 2 IU of heparin (Sigma) followed by 4% paraformaldehyde in PBS. Relevant tissues were then postfixed 24 to 48 hours in paraformaldehyde. Spinal cords were then cryoprotected in 30% sucrose and mounted in optimal cutting temperature compound, whereas hearts were dehydrated and paraffin-embedded. Spinal cord sections of 20-μm thickness were prepared on a sliding cryostat microtome (Leica) and collected onto gelatin-coated microscope slides. Immunohistochemistry was then performed with mouse antibody to SMN (1:1000; Transduction Laboratories) and rabbit antibody to CGRP (1:3000; Sigma-Aldrich) followed by Cy3-conjugated goat antibody to mouse (1:200; The Jackson Laboratory) and fluorescein isothiocyanate–conjugated goat antibody to rabbit (1:200; The Jackson Laboratory). The percentage of transduced motor neurons was determined by counting the number of cells showing CGRP-GFP double staining. Heart sections of 5 μm were prepared on a microtome (Leica) and collected onto precoated microscope slides for hematoxylin and eosin staining.

Statistical analysis

Statistical analysis was performed with GraphPad Prism 5. Statistical significance was assessed with one- or two-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test as stated in the figure legends.

Supplementary Material

Fig. S1. Lack of gene transfer in the SMNΔ7 scAAV9-coSMN–injected mouse that died at PN16.

Fig. S2. Histological analysis of the heart in SMNΔ7 and carrier mice.

Fig. S3. Gait analysis highlighted a postural difference between rescued and SMN+/− mice.

Video S1. Video recording of scAAV9-coSMN SMNΔ7 mouse (the smaller) and his carrier littermate at PN21 freely exploring a clean cage.

Video S2. Video recording of a carrier littermate at PN21.

Video S3. Video recording of a scAAV9-coSMN–injected SMNΔ7 mouse at PN21.


  • * These authors contributed equally to this work.

  • Citation: C. F. Valori, K. Ning, M. Wyles, R. J. Mead, A. J. Grierson, P. J. Shaw, M. Azzouz, Systemic delivery of scAAV9 expressing SMN prolongs survival in a model of spinal muscular atrophy. Sci. Transl. Med. 2, 35ra42 (2010).

References and Notes

  1. Acknowledgments: We thank K. Mitrophanous for his help with codon optimization of SMN. Lentiviral scAAV-GFP and pAAV2/9 plasmids were gifts from N. Deglon, J. Gray, and J. Wilson, respectively. We also thank J. Haylor who allowed us to use his stereomicroscope for the intravenous injection and S. Wharton for helpful advice on heart pathology. Funding: Jennifer Trust for SMA, SMA Trust, Sheffield Hospitals Charitable Trust, Wellcome Trust (P.J.S.), and FightSMA (M.W.). Author contributions: C.F.V. and K.N. designed, performed, and analyzed the in vitro and in vivo experiments; M.W. was involved in lentiviral production and titration; R.J.M., A.J.G., and P.J.S. dealt with the UK Home Office regulation and edited the paper; M.A. designed the experiments; M.A. and C.F.V. wrote and edited the paper. Competing interests: The authors declare that they have no competing interests.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article