Therapy development in spinal muscular atrophy

Transcription

1 n e u r o d e g e n e r a t i o n p e r s p e c t i v e Therapy development in spinal muscular atrophy Michael Sendtner Proximal spinal muscular atrophy (SMA) is the predominant form of motor neuron disease in children and young adults. In contrast to other neurodegenerative disorders, SMA is a genetically homozygous autosomal recessive disease that is caused by deficiency of the survival motor neuron (SMN) protein. This homogeneity should in principle facilitate therapy development. Previous therapy approaches have focused on upregulation of SMN expression from a second SMN (SMN2) gene that gives rise to low amounts of functional SMN protein. Drug development to target disease-specific mechanisms at cellular and physiological levels is in its early stages, as the pathophysiological processes that underlie the main disease symptoms are still not fully understood. Mouse models have helped to make conceptual progress in the disease mechanism, but their suitability in the search for therapeutic agents remains to be validated an issue that is ubiquitous to the translational therapeutic research of other neurodegenerative diseases. Human induced pluripotent stem cell technology for generation of large numbers of human motor neurons could help to fill this gap and advance the power of drug screening. In parallel, advances in oligonucleotide technologies for engineering SMN2 pre-mrna splicing are approaching their first clinical trials, whose success depends on improved technologies for drug delivery to motor neurons. If this obstacle can be overcome, this could boost therapy development, not only for SMA but also for other neurodegenerative disorders. The identification of underlying gene defects is a key for understanding the pathophysiology of neurodegenerative diseases and a vantage point for the development of therapies. Starting with the discovery of mutations in the SOD1 gene in familial amyotrophic lateral sclerosis (ALS) in 1993, it soon became clear that the same disease can also be caused by mutations in other genes (reviewed in ref. 1). Despite these advances, success in the development of therapies for ALS was marginal, partly owing to the heterogeneity of genetic conditions causing this disease 2. In ALS, more than 90% of cases seem sporadic, and the remaining 10% of cases that are familial fall into genetically and clinically different groups 1, hence suggesting that customized therapies tailored to disease subtypes might be needed to overcome the individual molecular pathologies. In contrast to ALS, proximal spinal muscular atrophy (SMA, OMIM ) is a classical autosomal recessive disorder that is the most common form of degenerative motor neuron disease in children and young adults 3. This makes Institute for Clinical Neurobiology, University of Wuerzburg, Wuerzburg, Germany. Correspondence should be addressed to M.S. (sendtner_m@klinik.uni-wuerzburg.de). Published online 25 June 2010; doi: /nn.2565 SMA interesting as a model case for other neurodegenerative diseases, in particular for development of concepts and tools for therapies. In SMA, the prevailing symptom is proximal skeletal muscle weakness stemming from degeneration of spinal and bulbar motor neurons. Although several forms of SMA are distinguished by clinical and electrophysiological parameters, more than 90% of these forms are caused by homozygous deletion or mutation of the SMN1 gene on human chromosome 5. The corresponding SMN protein is ubiquitously expressed and is localized to nuclear complexes known as Gemini of coiled bodies, or Gems, which are involved in small nuclear ribonucleoprotein processing and recycling (reviewed in ref. 4). In some types of neurons in particular, in motor neurons the SMN protein is also found at relatively high quantities in the cytoplasm of the cell bodies, in axons and in axon terminals 5 (Fig. 1). The SMN1 gene is part of an inverted segmental duplication on human chromosome 5, and, as a consequence, two functional copies of the SMN gene are found in humans; both are expressed 4. The SMN1 and SMN2 genes differ by five nucleotide exchanges 6, two of them within exons. In most transcripts derived from the SMN2 gene, a translationally silent cytosine to thymidine exchange at position 6 of exon 7 is responsible for the skipping of exon 7 during splicing. This mutation abolishes an exonic splice enhancer site (ESE) 7 and generates a new exonic splicing silencer domain 8 for the last coding exon of the SMN gene. The resulting SMN protein lacks the C-terminal 16 amino acid residues, which are replaced by four amino acids encoded by exon 8 sequences 9. The corresponding protein is less stable, and the altered C terminus of the SMN protein cannot self-associate anymore, making it less active 4,10. As a consequence, the SMN2 gene cannot fully compensate for deficiency of the SMN1 gene. However, some transcripts from the SMN2 gene that include the exon 7 encoded domain do undergo correct alternative splicing, so that 10% to 30% of functional full-length SMN proteins are produced from the SMN2 gene. This genetic situation implies several directions for therapy development: (i) strategies to increase full-length SMN protein production from the SMN2 gene, either by stimulating promoter activity or by enhancing exon 7 inclusion, (ii) gene therapy to supply extra copies of a functional SMN gene or the SMN cdna in motor neurons, and (iii) strategies focusing on compensating the deficits that are caused by SMN deficiency leading to motor neuron dysfunction and the disease symptoms. The last of these directions requires deeper insight into the mechanisms by which SMN protein deficiency leads to muscle weakness. These three therapeutic avenues depend on the availability of several tools. For one, optimized cell culture systems are needed for studying the molecular pathophysiology and to be used in drug screening. For another, animal models can help us understand why motor neurons are the predominantly affected cell type, and whether and how other cell types contribute to disease progression. Animal nature neuroscience VOLUME 13 NUMBER 7 JULY

2 Figure 1 Targets for therapy in SMA. Schematic diagram of the SMN1 and SMN2 genes on human chromosome (chr.) 5. In SMA, both copies of the SMN1 gene are deleted or mutated. The SMN2 gene is also expressed, but most of the resulting gene products give rise to a truncated SMN protein lacking the regions encoded by exon 7. This is caused by a C-to-T transition at position 6 of exon 7, leading to disruption of a splice enhancer site and generation of a new splice silencer site. Targets for therapy are marked as red circles. Increase of SMN2 promoter activity gives rise to enhanced production of truncated SMN2Δ7 mrna, but also to enhanced production of SMN2 full-length mrna and SMN protein. Restoration of splicing and inclusion of exon 7 by means of antisense oligonucleotides forms a second target for therapy development. Bottom, SMN protein is normally found in both the nucleus and the cytoplasm of spinal motor neurons (right); the deficit in SMN expression (left) depletes SMN immunoreactivity in both regions. Reproduced from ref. 12. SMN2 C-to-T transition Centromere models of disease can also serve as the proof of therapeutic strategies. Finally, comparative studies are needed to mitigate conditions that prevent successful transfer of technology from the bench to the bedside and to increase the predictive power of animal- or cell-based models for human trials. Mouse and cell culture models for therapy development Given that spinal muscular atrophy is caused by homozygous deletion of the SMN1 gene, gene knockout in model organisms appeared as the first step toward modeling the disease. However, the evolutionary duplication of the SMN gene in humans is not found in mice and flies, and homologous recombination of the Smn locus in mice leads to complete depletion of the SMN protein, causing early embryonic lethality 11. However, slowly progressive degeneration of motor neurons is recapitulated well in Smn +/ mice 12, which have a ~50% reduction in Smn protein, particularly in the cytoplasm (Fig. 1). Transgenic expression of two copies of the human SMN2 gene on a mouse Smn / background results in a gene dosage dependent phenotype resembling severe forms of SMA 13. These mice normally die shortly after birth and cannot be used to test the efficacy of drug candidates on postnatal motor behavior. Isolated motor neurons from Smn / SMN2 tg mice do not show any abnormalities in cell survival, but they manifest defects in axonal growth and growth cone morphology 14. Similarly, morpholinomediated knockdown of Smn in zebrafish 15 does not compromise the survival of motor neurons but severely retards axonal extension and disturbs pathfinding, and Smn deficiency causes abnormalities in neurotransmission at neuromuscular endplates 16 (Fig. 2). Thus, disturbed neuromuscular endplate development and function, and corresponding abnormalities in synaptic transmission 16,17, may be responsible for muscle weakness, the leading symptom in humans with SMA. Strategies to increase SMN protein in motor neurons Transgenic expression of two copies of a cdna encoding the human SMNΔ7 gene product in Smn / SMN2 tg mice prolongs survival to about 3 weeks after birth 18. By contrast, mice with more than six copies of b 2a 1 1 2a 2b C-to-U mutation Responsible for exon skipping 1 2a 2b % skipping exon 7 10% with exon 7 New stop codon in D7 mrna Unstable SMN2Δ7 mrna and protein Smn-deficient motor neuron SMN1 Stop codon Stable SMN mrna and protein Control motor neuron Telomere 8 1 2a 2b Human chr. 5 SMA region Pre-mRNA mrna Protein Neuron the human SMN2 gene on an Smn / background appear healthy 13. Interestingly, a similar phenotypic rescue is seen in humans with homozygous SMN1 deletion but extra copies of SMN2 (ref. 19). Thus, increased SMN2 copy number can restore motor neuron function in SMA, and the increased expression of SMN from the SMN2 gene is considered a major avenue to therapy, as SMN2 gene dosage is correlated with disease severity. SMN expression from its endogenous promoter is highest during embryonic development, and expression decreases after birth, so strategies maintaining high embryonic expression into later stages seem promising for SMA therapy. High-throughput screening to identify drugs that increase expression of SMN from the SMN2 gene has been performed, mostly in fibroblasts or other non-neuronal cells derived from subjects with SMA. Such screens 20 resulted in the identification of several drug candidates, such as nonselective histone deacetylase (HDAC) inhibitors valproic acid, trichostatin A and sodium butyrate that are known to upregulate transcription of 2% of all genes 21, including SMN 22,23. Unfortunately, the relative effects of these drug candidates has lagged behind expectations, both in mouse models and in humans. The median difference in survival of trichostatin A treated Smn / SMN2 tg SMN Δ7tg mice was 3 days 24 ; sodium butyrate prolonged survival by 6 days in another mouse model of SMA 22, and valproic acid showed some effects on motor neuron survival and function in a model of mild SMA 25. Clinical studies with valproic acid showed only small effects in patients 26. New drug candidates have been identified that seem to be more potent in this context 27,28. A major drawback of using drugs that do not act specifically on the SMN2 promoter lies in unwanted side effects in other cell types and toxic effects on motor neurons 29. This points to problems that could contribute to these disappointing results in clinical trials and highlights the need for integration of additional symptom-related parameters such as restoration of axon growth, synapse formation and neural excitability in cell-based screening programs. These issues are also important for 796 VOLUME 13 NUMBER 7 JULY 2010 nature neuroscience

3 Figure 2 Axonal defects in Smn-deficient motor neurons. (a) Smn / SMN2 tg motor neurons show defects in formation of presynaptic structures, as shown by the lack of accumulation of voltage-gated calcium channels (Cav2.2) in the tip of axonal growth cones and lack of colocalization with other proteins of the active zone, such as piccolo (green). Reproduced from ref. 17. (b) Diminished neuromuscular endplate currents (EPC) in tibialis anterior muscle of postnatal Smn-deficient (SMA) mice. The deficit in neurotransmission is caused by a deficit in release of synaptic vesicles. CL, control; *P < Reproduced from ref. 16. a Smn +/+ SMN2 Smn / SMN2 Cav2.2 Piccolo b CL 5 na 2 ms SMA other neurodegenerative diseases, and combined efforts to identify molecules that prevent axon and synapse degeneration could also be of relevance for a much broader spectrum of disorders of the nervous system. Strategies to increase SMN protein from the SMN2 gene Besides the strategies aiming at upregulating SMN2 promoter activity for higher expression of the SMN protein, efforts have been made to modulate pre-mrna splicing to get higher rates of inclusion of exon 7 into the processed SMN mrna from the SMN2 gene. Strategies for modulating pre-mrna splicing have been developed for other neuromuscular disorders, including Duchenne muscular dystrophy (DMD) and myotonic dystrophy 30. These strategies are normally based on antisense oligonucleotide (AON) technologies not only to knock down specific transcripts in antiviral therapy 31 but also to modulate mrna splicing (for example, in DMD 32 ). Some of these approaches go back more than 10 years and in the cancer field have already progressed to phase 3 clinical studies (summarized in ref. 30). The strategy of using the AON technology for exon skipping is also approaching first clinical trials in DMD for restoring the reading frame of dystrophin transcripts, with the ultimate goal of converting a severe disease phenotype into a mild one. A first clinical trial in four people with DMD in which a single dose of AON in a single muscle was injected led to dystrophin restoration for periods of several months 32. Conceptually speaking, this strategy is hampered by the difficulty of AON delivery to target tissue(s) for example, deep into the heart, which is also affected in DMD. To overcome this delivery issue, other strategies have been developed that transduce an antisense gene with viral vectors. Such constructs are normally based on U7 (ref. 33) or related small nuclear ribonucleoprotein genes to target the oligonucleotides to spliceosomes. In these constructs, sequences that hybridize to the 5 splice site are replaced by the AON antisense sequence (reviewed in refs. 30,34). Suitable viral vectors include those based on adeno-associated virus (AAV), which seem most promising for human gene therapy because constructs for exon skipping are small and can easily fit into AAV vectors. These viruses also show impressive long-term expression in mouse models 33. However, the situation could be more problematic in humans, in whom these viruses seem more immunogenic than in mice 35, thus precluding repeated delivery and limiting long-term expression. Unfortunately, verbatim transfer of AON application from cancer and DMD therapy to SMA is not possible, considering the fundamental differences in the disease mechanism. The situation in cancer and Duchenne muscular dystrophy therapy is such that gene expression or exon splicing is suppressed by the antisense oligonucleotides. However, the molecular situation is different in SMA, where a strategy to enhance exon inclusion must be applied an interesting if somewhat daunting challenge for cell biologists. In contrast to therapeutic strategies aiming at increasing SMN2 promoter activity, the oligonucleotide-based strategy is site-specific for exon 7 of the SMN2 gene and thus more specific than the more pleiotropic approach of upregulating CL SMA SMN2 promoter activity. Despite the conceptual challenges, however, proof of principle for this therapeutic avenue has been demonstrated in a mouse model of the severe form of SMA, in which a bifunctional U7 small nuclear RNA was expressed as a transgene 36. Whereas average survival in control mice was 5 days after birth, more than 50% of mice with a transgenic U7 exon splice enhancer construct were still alive at 100 days after birth, some mice surviving even beyond 300 days. This finding marks a point of reference for therapy development in SMA. The path to this success was paved by systematic characterization of all splice enhancer and silencer sites in exon 6, exon 7 and neighboring introns that influence exon 7 inclusion and thus could be used for therapy 37. Inhibitory regions in both exon 7 and the last intron between exon 7 and 8 (ref. 38) were identified, as well as a region within exon 7 that binds the splicing activator. On the basis of these data, AONs were developed and tested for their efficacy in increasing exon 7 inclusion in mrna transcripts from the SMN2 gene in fibroblasts from SMA patients and in mouse models 37,39. The treatment increased SMN protein levels, but in vivo effects in mouse models were observed mainly in liver and skeletal muscle, not in the spinal cord 37,39. This result was as expected because the AONs used in this study do not penetrate the blood-brain-barrier. The challenge here is the development of suitable application techniques for delivery of these substances to motor neurons in vivo. Moreover, the AONs or compound molecules that are designed to increase exon 7 inclusion from the SMN2 gene need to be tested in motor neurons to identify potential cell type specific conditions that influence splicing of the SMN pre-mrna. Even more important will be determining whether this approach alters disease-specific defects such as the disturbed presynaptic differentiation and synaptic vesicle release that seems to be responsible for motor neuron dysfunction and muscle weakness in SMA. Viral gene transfer of SMN and splicing modifiers to neurons Lentivirus-based gene transfer has been used for RNA interference (RNAi) as a therapeutic strategy in a mouse model of familial ALS 40. Such application has increased life expectancy in the animal model by 80%, highlighting the clinical potential of these treatment methods. Similar lentivirus-based approaches have been used in SMA models 41, but the therapeutic effects were much smaller EPC (na) * nature neuroscience VOLUME 13 NUMBER 7 JULY

4 than those of transgenic expression of extra copies of the SMN2 gene 13 or the U7 exon splice enhancer construct 36. These approaches have not yet entered the clinic, but there are reasons for optimism. Although the initial AAV-based vectors did not show efficacy in mouse models similar to that of lentiviral vectors for RNAi treatment of familial ALS 42, new vectors have been developed that allow more efficient gene transfer into motor neurons, and these vectors have recently shown high efficacy for treatment of SMA mouse models when the SMN gene is transduced 43,44. When these mice are treated with such viral vectors, they survive at least as long as mice treated by conventional transgenic expression of a U7 small nuclear RNA for splicing modification of the SMN2 gene product 36. Further progress has been made by the demonstration that these AAV vectors transduce genes to spinal motor neurons in nonhuman primates after intramuscular injection 45 and by the recent report that these viruses are relatively selective for motor neurons after systemic injection 46. However, it is still not clear, considering potential problems of AAV with immunogenicity in humans 35, whether these tools will show similar efficacy in patients. Translation from mouse models to therapy: requirement for both new tools and validation In cell culture and mouse models of ALS, neurotrophic factors such as ciliary neurotrophic factor (CNTF), or other growth and differentiation factors, such as vascular endothelial factor (VEGF), can improve motor neuron survival, maintenance and function of neuromuscular endplates 47,48. Nevertheless, previous clinical trials with these factors failed, or efficacy has not been proven yet (reviewed in ref. 47). Two main barriers to translating the effects in mouse models to humans became apparent: (i) poor pharmacokinetics and bioavailability, so these factors did not reach motor neurons in sufficient quantities, and (ii) unwanted side effects, as neurotrophic factors that normally act in close cell-to-cell contact were administered systemically and bound to receptors on cells other than motor neurons for example, CNTF receptors on liver cells thus causing fever and cachexia 47. Similar problems were observed with a broad range of other therapeutic approaches that were tested in superoxide dismutase (SOD) G93A mice and other mouse models expressing mutant forms of transgenic SOD1. None of these treatments in the mouse models have thus far translated to positive effects in the clinic. However, the overexpression of additional transgenes of mutant SOD1 does not reflect the genetic situation in familial ALS in which only one allele of SOD1 is expressed. Artificial protein aggregation and other defects that are caused by the expression of mutant SOD from multiple copies could be responsible for this lack of transferability, at least in part. Altogether, these problems point to differences between mouse and human that negatively influence the translation of therapeutic effects from mouse models to clinically effective treatment. These issues from virus-based therapeutic approached for sporadic ALS also apply to current efforts in SMA, and probably apply equally to other neurodegenerative disorders as well. The efficacy of drug candidates that increase promoter activity of SMN2 could differ between mouse and human cells, and their pharmacokinetics might also differ between mouse and human. Furthermore, if drug candidates that lack specificity for specific promoters detrimentally modulate expression of other genes in mouse and human motor neurons, this could cause side effects that are not predictable from studies with mouse models. Potential confounding factors from bench to bedside also apply to studies with AON approaches. Previous studies with AONs in mouse models of DMD have shown relatively good success, but the small size of the animal makes it easier to target multiple muscle groups by injection than is possible in humans 30. In this respect, new technologies to deliver AONs with high specificity and efficacy to targets that is, muscle in DMD or motor neuron in SMA or ALS are necessary. Among these techniques, new developments in the field of viral gene transfer could have a prominent role. Human ips cell derived motor neurons as a first step for drug development in SMA Since the first report on reprogramming of mouse fibroblasts into socalled induced pluripotent stem cells (ips cells) by defined factors in 2006, human embryonic stem cell lines have been successfully derived from individuals with ALS 49 and SMA 50. These cells can be differentiated to motor neuron like cells, and these ips cell derived human motor neurons offer several advantages, as they seem more suitable to coping with factors that are unique to the affected motor neurons in SMA. It will be interesting to use ips derived human motor neurons to look for drugs for SMA because this could avoid potential differences between mouse and human motor neurons that affect the translatability of these approaches to clinical trials. These cells also exactly mirror the genetic situation in human SMA, and there is no need for artificial transgenic compensation of SMN gene defects. Thus, both AON approaches and strategies for upregulating SMN2 expression can be tested in a context of human cells in vitro. These cells can be used to add weight and predictability to current efforts with isolated mouse motor neurons to investigate the mechanism of therapeutic strategies that will normalize the disease process. For example, one could apply protocols to differentiate presynaptic axonal terminals in ips cells culturing the cells with synapse-specific forms of laminin 17 and other inducers of presynaptic differentiation which could in principle introduce additional functional parameters such as presynaptic differentiation and capacity for vesicle release to specifically model particular aspects of the disease (Fig. 2). This approach might reveal drug candidates that could prove effective in counteracting disease-specific abnormalities such as presynaptic impairment or axonal morphology defects. Defects in presynaptic structures that are necessary for synaptic vesicle release are prominent in isolated motor neurons from Smn / SMN2 tg mice (Fig. 2), a model for severe SMA 17, and they have also been observed in milder SMA mouse models in vivo 16. Nevertheless, the exact molecular mechanisms that lead to muscle weakness in SMA are not fully understood and are still under debate 4, so further studies on the exact mechanism of pathology are necessary. In these studies, mouse models for SMA and ips-derived motor neurons will complement one another as we bring in vitro data to the organismic level. Future perspectives and conclusions The main challenges toward successful therapies in patients with SMA include a better understanding of the disease pathophysiology and efficient delivery of therapeutic agents to the correct target tissue. Although the SMA field can learn from other areas of translational research, these tools and ideas must be adapted specifically to the SMA context, where the disease mechanism and conceptual obstacles may be different. It remains to be seen whether improved technologies for drug administration and further development of viral vectors that allow highly efficient and specific gene transfer to motor neurons can help to overcome the obstacles to deliver these drugs in such a way that they reach motor neurons in sufficient quantities and to limit unwanted side effects. Progress in this area is a general issue applicable for the treatment of other neurodegenerative disorders where the blood-brain barrier is a major consideration. Setbacks such as the recently reported potential problems of AAV with immunogenicity in humans 35 are expected, but researchers should not be discouraged and 798 VOLUME 13 NUMBER 7 JULY 2010 nature neuroscience

5 abandon their endeavors. As these problems become apparent, overcoming such obstacles will require concerted and continuing efforts. Moreover, better knowledge of the cellular dysfunction that is directly responsible for the disease phenotype could help to pave the way toward therapies that compensate for particular deficits in the disease conditions. Such efforts could benefit from mutually closer interaction between different fields of drug development research, such as those focused on SMA and ALS, and perhaps even from closer interaction with more distant fields such as those focused on Alzheimer s disease and other neurodegenerative disorders where synapse dysfunction and loss is a pathophysiological hallmark that correlates with disease symptoms. Efforts to prevent synaptic pruning, loss of presynaptic structures for synaptic vesicle release, or postsynaptic alterations that contribute to loss of synaptic function are examples of lines of research that could make a difference to a broad range of neurodegenerative disorders. These combined efforts could broaden the chance for success, and thus the journey that started from the identification of the underlying genetic defect in SMA could finally lead to arrival at successful therapy in patients. Acknowledgments I thank R. Blum and R. Götz for critical reading and many helpful comments. Work in my laboratory on spinal muscular atrophy was supported by the SMA Foundation, the Hermann und Lilly Schilling Stiftung im Stifterverband der Deutschen Industrie and the Deutsche Forschungsgemeinschaft, grant SFB 581, B1. COMPETING FINANCIAL INTERESTS The author declares no competing financial interests. Published online at Reprints and permissions information is available online at reprintsandpermissions/. 1. Valdmanis, P.N., Daoud, H., Dion, P.A. & Rouleau, G.A. Recent advances in the genetics of amyotrophic lateral sclerosis. Curr. Neurol. Neurosci. Rep. 9, (2009). 2. Simpson, C.L. & Al-Chalabi, A. Amyotrophic lateral sclerosis as a complex genetic disease. Biochim. Biophys. Acta 1762, (2006). 3. Crawford, T.O. & Pardo, C.A. The neurobiology of childhood spinal muscular atrophy. Neurobiol. Dis. 3, (1996). 4. Burghes, A.H. & Beattie, C.E. Spinal muscular atrophy: why do low levels of survival motor neuron protein make motor neurons sick? Nat. Rev. Neurosci. 10, (2009). 5. Rossoll, W. et al. Specific interaction of Smn, the spinal muscular atrophy determining gene product, with hnrnp-r and gry-rbp/hnrnp-q: a role for Smn in RNA processing in motor axons? Hum. Mol. Genet. 11, (2002). 6. Wirth, B. An update of the mutation spectrum of the survival motor neuron gene (SMN1) in autosomal recessive spinal muscular atrophy (SMA). Hum. Mutat. 15, (2000). 7. Cartegni, L. & Krainer, A.R. Disruption of an SF2/ASF-dependent exonic splicing enhancer in SMN2 causes spinal muscular atrophy in the absence of SMN1. Nat. Genet. 30, (2002). 8. Kashima, T. & Manley, J.L. A negative element in SMN2 exon 7 inhibits splicing in spinal muscular atrophy. Nat. Genet. 34, (2003). 9. Mattis, V.B. et al. Novel aminoglycosides increase SMN levels in spinal muscular atrophy fibroblasts. Hum. Genet. 120, (2006). 10. Cho, S. & Dreyfuss, G. A degron created by SMN2 exon 7 skipping is a principal contributor to spinal muscular atrophy severity. Genes Dev. 24, (2010). 11. Schrank, B. et al. Inactivation of the survival motor neuron gene, a candidate gene for human spinal muscular atrophy, leads to massive cell death in early mouse embryos. Proc. Natl. Acad. Sci. USA 94, (1997). 12. Jablonka, S., Schrank, B., Kralewski, M., Rossoll, W. & Sendtner, M. Reduced survival motor neuron (Smn) gene dose in mice leads to motor neuron degeneration: an animal model for spinal muscular atrophy type III. Hum. Mol. Genet. 9, (2000). 13. Monani, U.R. et al. The human centromeric survival motor neuron gene (SMN2) rescues embryonic lethality in Smn / mice and results in a mouse with spinal muscular atrophy. Hum. Mol. Genet. 9, (2000). 14. Rossoll, W. et al. Smn, the spinal muscular atrophy-determining gene product, modulates axon growth and localization of beta-actin mrna in growth cones of motoneurons. J. Cell Biol. 163, (2003). 15. McWhorter, M.L., Monani, U.R., Burghes, A.H. & Beattie, C.E. Knockdown of the survival motor neuron (Smn) protein in zebrafish causes defects in motor axon outgrowth and pathfinding. J. Cell Biol. 162, (2003). 16. Kong, L. et al. Impaired synaptic vesicle release and immaturity of neuromuscular junctions in spinal muscular atrophy mice. J. Neurosci. 29, (2009). 17. Jablonka, S., Beck, M., Lechner, B.D., Mayer, C. & Sendtner, M. Defective Ca 2+ channel clustering in axon terminals disturbs excitability in motoneurons in spinal muscular atrophy. J. Cell Biol. 179, (2007). 18. Le, T.T. et al. SMNDelta7, the major product of the centromeric survival motor neuron (SMN2) gene, extends survival in mice with spinal muscular atrophy and associates with full-length SMN. Hum. Mol. Genet. 14, (2005). 19. Jedrzejowska, M. et al. Unaffected patients with a homozygous absence of the SMN1 gene. Eur. J. Hum. Genet. 16, (2008). 20. Jarecki, J. et al. Diverse small-molecule modulators of SMN expression found by high-throughput compound screening: early leads towards a therapeutic for spinal muscular atrophy. Hum. Mol. Genet. 14, (2005). 21. Marks, P.A., Richon, V.M., Miller, T. & Kelly, W.K. Histone deacetylase inhibitors. Adv. Cancer Res. 91, (2004). 22. Chang, J.G. et al. Treatment of spinal muscular atrophy by sodium butyrate. Proc. Natl. Acad. Sci. USA 98, (2001). 23. Sumner, C.J. et al. Valproic acid increases SMN levels in spinal muscular atrophy patient cells. Ann. Neurol. 54, (2003). 24. Avila, A.M. et al. Trichostatin A increases SMN expression and survival in a mouse model of spinal muscular atrophy. J. Clin. Invest. 117, (2007). 25. Tsai, L.K., Tsai, M.S., Lin, T.B., Hwu, W.L. & Li, H. Establishing a standardized therapeutic testing protocol for spinal muscular atrophy. Neurobiol. Dis. 24, (2006). 26. Swoboda, K.J. et al. Phase II open label study of valproic acid in spinal muscular atrophy. PLoS ONE 4, e5268 (2009). 27. Garbes, L. et al. LBH589 induces up to 10-fold SMN protein levels by several independent mechanisms and is effective even in cells from SMA patients nonresponsive to valproate. Hum. Mol. Genet. 18, (2009). 28. Butchbach, M.E. et al. Effects of 2,4-diaminoquinazoline derivatives on SMN expression and phenotype in a mouse model for spinal muscular atrophy. Hum. Mol. Genet. 19, (2010). 29. Rak, K. et al. Valproic acid blocks excitability in SMA type I mouse motor neurons. Neurobiol. Dis. 36, (2009). 30. Artsma-Rus, A. & van Ommen, G.J. Progress in therapeutic antisense applications for neuromuscular disorders. Eur. J. Hum. Genet. 18, (2010). 31. Pan, W.H. & Clawson, G.A. Antisense applications for biological control. J. Cell. Biochem. 98, (2006). 32. van Deutekom, J.C. et al. Local dystrophin restoration with antisense oligonucleotide PRO051. N. Engl. J. Med. 357, (2007). 33. Goyenvalle, A. et al. Rescue of dystrophic muscle through U7 small nuclear RNAmediated exon skipping. Science 306, (2004). 34. Khoo, B. & Krainer, A.R. Splicing therapeutics in SMN2 and APOB. Curr. Opin. Mol. Ther. 11, (2009). 35. Zaiss, A.K. & Muruve, D.A. Immunity to adeno-associated virus vectors in animals and humans: a continued challenge. Gene Ther. 15, (2008). 36. Meyer, K. et al. Rescue of a severe mouse model for spinal muscular atrophy by U7 small nuclear RNA-mediated splicing modulation. Hum. Mol. Genet. 18, (2009). 37. Hua, Y., Vickers, T.A., Okunola, H.L., Bennett, C.F. & Krainer, A.R. Antisense masking of an hnrnp A1/A2 intronic splicing silencer corrects SMN2 splicing in transgenic mice. Am. J. Hum. Genet. 82, (2008). 38. Singh, N.N., Singh, R.N. & Androphy, E.J. Modulating role of RNA structure in alternative splicing of a critical exon in the spinal muscular atrophy genes. Nucleic Acids Res. 35, (2007). 39. Hua, Y., Vickers, T.A., Baker, B.F., Bennett, C.F. & Krainer, A.R. Enhancement of SMN2 exon 7 inclusion by antisense oligonucleotides targeting the exon. PLoS Biol. 5, e73 (2007). 40. Ralph, G.S. et al. Silencing mutant SOD1 using RNAi protects against neurodegeneration and extends survival in an ALS model. Nat. Med. 11, (2005). 41. Azzouz, M. et al. Lentivector-mediated SMN replacement in a mouse model of spinal muscular atrophy. J. Clin. Invest. 114, (2004). 42. Towne, C., Raoul, C., Schneider, B.L. & Aebischer, P. Systemic AAV6 delivery mediating RNA interference against SOD1: neuromuscular transduction does not alter disease progression in fals mice. Mol. Ther. 16, (2008). 43. Foust, K.D. et al. Rescue of the spinal muscular atrophy phenotype in a mouse model by early postnatal delivery of SMN. Nat. Biotechnol. 28, (2010). 44. Passini, M.A. et al. CNS-targeted gene therapy improves survival and motor function in a mouse model of spinal muscular atrophy. J. Clin. Invest. 120, (2010). 45. Towne, C., Schneider, B.L., Kieran, D., Redmond, D.E. Jr. & Aebischer, P. Efficient transduction of non-human primate motor neurons after intramuscular delivery of recombinant AAV serotype 6. Gene Ther. 17, (2010). 46. Foust, K.D. et al. Intravascular AAV9 preferentially targets neonatal neurons and adult astrocytes. Nat. Biotechnol. 27, (2009). 47. Thoenen, H. & Sendtner, M. Neurotrophins: from enthusiastic expectations through sobering experiences to rational therapeutic approaches. Nat. Neurosci. 5 (suppl.), (2002). 48. Storkebaum, E., Lambrechts, D. & Carmeliet, P. VEGF: once regarded as a specific angiogenic factor, now implicated in neuroprotection. Bioessays 26, (2004). 49. Dimos, J.T. et al. Induced pluripotent stem cells generated from patients with ALS can be differentiated into motor neurons. Science 321, (2008). 50. Ebert, A.D. et al. Induced pluripotent stem cells from a spinal muscular atrophy patient. Nature 457, (2009). nature neuroscience VOLUME 13 NUMBER 7 JULY