COMPOSITIONS AND METHODS FOR MODULATING SMN ACTIVITY

CROSS-REFERENCE TO RELATED APPLICATIONS [001] This application claims the benefit of U.S. Provisional Application No. 61/040,964, filed March 31, 2008, and U.S. Provisional Application No. 61/093,931, filed September 3, 2008, which are herein incorporated by reference in their entireties.

FIELD OF THE INVENTION

[002] The present invention is related to the fields of genetics and neurobiology. In particular, the present invention describes a Drosophila model of human spinal muscular atrophy and genetic modifiers of the survival motor neuron (SMN) gene. The genetic modifiers provide novel therapeutic targets for treating spinal muscular atrophy.

BACKGROUND OF THE INVENTION

[003] Spinal muscular atrophy (SMA) is an autosomal recessive disorder which is the leading hereditary cause of infant death in humans. The disease is characterized by a progressive muscle weakness from proximal to distal with lower limbs more greatly affected than upper limbs. Three types of SMA have been described based on disease severity and age of onset. Type I affects approximately sixty percent of SMA patients with symptoms presenting within the first six months after birth. Death typically occurs within the first two years due to respiratory failure. Type II SMA has an onset between six months and eighteen months of age, and length of survival is dependent on the severity of respiratory impairment. Type III SMA patients, who have symptom onset between eighteen months and early childhood, usually do not have a decrease in life expectancy, although most are wheel chair bound at some point in their disease progression.

[004] SMA is characterized by loss of alpha-motor neurons in the anterior horn of the spinal cord, which is correlated with muscle paralysis and atrophy (Crawford and Pardo (1996) Neurobiol. Dis. 3: 97-110). Motor neuron degeneration is thought to be due to low levels of the survival motor neuron protein (Coovert et al. (1997) Hum. MoI. Genet. 6: 1205-1214; Lefebvre et al, (1997) Nat. Genet. 16: 265-269). Homozygous deletions

of the telomeric copy of the survival motor neuron (SMNl) gene located on chromosome 5q cause SMA (Lefebvre et al. (1995) Cell 80: 155-165). The majority of the population have a centromeric copy of the survival motor neuron (SMN2) gene, which can partially compensate for the loss of the SMNl gene product in SMA patients. [005] The essential difference between the SMNl and SMN2 genes is a C to T transition in exon 7 of the SMN2 gene, which causes this exon to be frequently omitted during alternative splicing (Lorson et al. (1999) Proc. Natl. Acad. Sci. USA 96: 6307-6311; Monani et al. (1999) Hum. MoI. Genet. 8: 1177-1183). The majority of transcripts produced from SMN2 lack exon 7 and result in the expression of a truncated protein product. The truncated protein does not oligomerize as well as full-length protein and is quickly degraded (Lorson et al. (1998) Nat. Genet. 19: 63-66; Le et al. (2000) Neurogenetics 3: 7-16). However, the ability of the SMN2 gene to generate low levels of full-length transcript and in turn full-length protein may explain why SMN2 copy number modulates the disease phenotype (Parsons et al. (1998) Am. J. Hum. Genet. 63: 1712- 1723).

[006] Currently, there are no effective therapeutics for SMA disease. In addition, the development of appropriate animal models to test potential treatments is difficult since the presence of the SMN2 gene is unique to humans. Most other organisms only have one copy of the SMN gene, disruption of which results in an embryonic lethal phenotype. Therefore, designing an animal model that resembles the SMA disease condition in humans and obtaining a viable organism in which to test candidate drug compounds or dissect the mechanism by which the SMN protein functions in motor neurons has been challenging. Thus, there is need in the art to identify other genes that interact with the SMN gene to understand the development of SMA disease and provide additional therapeutic targets.

SUMMARY OF THE INVENTION

[007] The present invention is based, in part, on the development of a Drosophila model system carrying a hypomorphic allele of the Smn gene, which more closely resembles the human SMA disease condition. Using such Drosophila model systems of human SMA, the inventors have identified novel genetic modifiers of Smn gene function. These novel

genetic modifiers provide therapeutic targets for the development of SMA treatments. Accordingly, the present invention provides a method of treating spinal muscular atrophy in a subject by manipulating the expression or activity of the identified genetic modifiers. [008] In one embodiment, the present invention includes a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that enhances bone morphogenetic protein signaling or fibroblast growth factor signaling. The agent can enhance the activity of one or more components in these signaling cascades, such as a BMP type II receptor, FGF receptor-2 or 3, regulatory SMAD activity, or MAP kinase activation. In some embodiments, the agent can be BMP, FGF, or other ligand of a BMP type II receptor or FGF receptor. [009] In another embodiment, the present invention provides a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that increases the expression or activity of a SMN agonist. SMN agonists can include Pumilio homolog 1, eIF-4E, MAPlB, Rhol, and plastin3. In some embodiments, the agent is an expression construct encoding a SMN agonist. In one embodiment, the SMN agonist is expressed using a muscle-specific promoter. In another embodiment, the SMN agonist is expressed using a neuron-specific promoter. In another embodiment, the agent is the SMN agonist itself. In still another embodiment, the agent is a compound that increases the expression or activity of a SMN agonist.

[0010] In still another embodiment, the present invention provides a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that decreases the expression or activity of a SMN antagonist. SMN antagonists can include Fmrl, Moesin, and slik. In some embodiments, the agent is an inhibitory RNA molecule or an antisense nucleic acid targeted to a sequence of a SMN antagonist. In other embodiments, the agent is a compound the decreases the expression or activity of a SMN antagonist.

[0011] The present invention also encompasses a method for modulating SMN biological function in a cell by manipulating the expression and/or activity of one or more of the genetic modifiers described herein. In one embodiment, the SMN biological function is enhanced in the cell. In another embodiment, the SMN biological function is reduced in the cell. The cell can be in vitro or in vivo.

[0012] The present invention also provides a method of screening candidate compounds for treatment of spinal muscular atrophy. In one embodiment, the method comprises exposing transgenic Drosophila to one or more candidate compounds, wherein the transgenic Drosophila comprise at least one transgene expressing an inhibitory RNA molecule against the Smn gene; comparing the phenotype of the exposed transgenic Drosophila to the phenotype of transgenic Drosophila not exposed to said one or more compounds; and selecting said one or more compounds that produce a change in phenotype, wherein the selected one or more compounds are therapeutic compounds for the treatment of spinal muscular atrophy. The change in phenotype can include a reduction in lethality, an increase in the number of synaptic boutons at the neuromuscular junction, and an increase in the number of pigmented pupae. In some embodiments, the inhibitory RNA molecule is an siRNA. The siRNA can be targeted to the full length Smn gene or a portion thereof. The transgene can be expressed ubiquitously or in a tissue- specific manner {e.g. muscle or neuronal).

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1. Specificity of the anti-SMN antibodies. (A-C) Wing discs from 3rd instar larvae overexpressing the UAS-FLAG-Smn transgenic rescue construct using the vestigalGAL driver were stained with antibodies against the FLAG peptide (green) (A) and SMN (red) (B). (C) Merge of (A) and (B) showing the overlapping expression of SMN and FLAG within the vestigal expression domain. (D) Wild-type and (E) vestigalGAL4 ^' , pWIZ[UAS-Smn-RNAi] ^N4 3rd instar wing discs were stained with antibodies against SMN (green). (F) Western blots of a serial dilution of S2 cell extracts (1: 20 μg, 2: 40 μg, 3: 60 μg, 4: 80 μg total protein) using the polyclonal (left) and monoclonal (right) antiserum against SMN recognize a single band of approximately 28kD in size.

[0014] Figure 2. SMN localizes to the post-synaptic region of the Drosophila NMJ. (A-D) SMN expression at the NMJ between muscle fibers 6 and 7. (A) Pre-synaptic anti- HRP staining (red), (B) post-synaptic anti-DLG staining (blue), (C) anti-SMN staining (green) and (D) a merge of (A-C). SMN expression co-localizes with DLG at the postsynaptic region of the NMJ. (E) SMN staining is also observed in muscle fibers and

discrete foci in nuclei (arrow). (F) Though no pre-synaptic SMN staining is observed, robust levels of SMN expression are seen in the larval brain. Scale bars in (D), (E), (F) represent 10 μm, 20 μm, and 50 μm, respectively.

[0015] Figure 3. SMN post-synaptic staining is abolished by muscle specific SMN knockdown. (A-F) The morphology of the NMJ between muscles 6 and 7 in the A2 segment was observed in different genetic backgrounds using antibodies against SMN (green) and the post-synaptic marker, Discs large (red). (A-C) Wild-type: anti-DLG (A), anti-SMN (B) and (C) merge of (A) and (B). (D-F) Transgenic animals containing how24BGAL4 and pWIZ[UAS-Smn-RNAi] ^N4 : anti-DLG (D), anti-SMN (E) and (F) merge of (D) and (E). In this background, SMN staining is reduced (E). [0016] Figure 4. Smn mutations cause lethality. (A) Schematic representation of the SMN protein and location of mutations corresponding to the Smn alleles used. The conserved Tudor domain and YG box are indicated. Insertion sites of the transposon- induced Smn ^m5960 and Smn ^{m u>9} alleles are denoted by triangles. Regions of the Smn transcript targeted by RNA interference (RNAi) are illustrated as lines under the SMN protein schematic. (B) For individuals of given phenotypes, the percentages of surviving individuals are shown and are normalized to wild-type. homozygotes die during late 2 ⁿ /early 3 ^r larval and pupal stages, though some escapers are detected. In contrast, 67% of the S homozygotes survive to adulthood. and trans-heterozygous combinations are also viable. In addition, a small deficiency uncovering the entire Smn transcript was generated ( ( ) ). We crossed all three Smn alleles to ( ) and found that both ^o /Of(3L)S and S heterozygotes die between the 2 ^nd and 3 ^rd instar larval stages, while -60% o /Of(3L are viable. Therefore, using lethality as a criterion, all three alleles behave as loss-of-function mutations with S ⁹ displaying the weakest phenotype of the three. No obvious maternal or paternal effect is observed for the different alleles, m: maternal contribution, p: paternal contribution. WT is wild- type (Canton-S). At least 100 individuals were examined for each genotype.

[0017] Figure 5. Drosophila Smn mutations elicit neuromuscular junction (NMJ) defects. (A-F) The morphology of the NMJ, as judged by bouton numbers, between

muscles 6 and 7 in the A2 segment was observed in different genetic backgrounds using the pre-synaptic (Synaptotagmin) and post-synaptic (Discs large) markers, shown in green and red, respectively. The following genotypes were examined: (A) wild-type ff)1109 / _/ S _o mn ff)1109 , ( s _n D)-. S _c mn 73Ao / _/ S _c mn K)1109. ^ Ofr- these combinations, Smn ⁷ °/Smn displayed the most robust NMJ defect. These defects are partially rescued by either (E) neuron-specific expression (elavGAL4) or (F) muscle-specific expression (how24BGAL4) of a UAS-FLAG-Smn transgene. (G) More complete rescue was achieved when this transgene was expressed using both drivers simultaneously. Bouton numbers were normalized to the ratio of the muscle area. Scale bars represent 20 μm. (H) Diagram of bouton numbers for genotypes from (A-F), normalized for muscle area. * P<0.05 was determined by the ANOVA multiple comparisons test. For each genotype at least 15 animals were examined. [0018] Figure 6. Pre-synaptic ghost bouton counts are elevated in Smn animals. The morphology of the NMJ between muscles 6 and 7 in the A2 segment was observed in different Smn backgrounds using the pre-synaptic (Synaptotagmin) and post-synaptic (Discs large) markers. Ghost bouton counts were determined by assessing the numbers of boutons that stained positive for Synaptotagmin and failed to stain for Discs large. All combinations examined displayed elevated numbers of pre-synaptic ghost boutons when compared to wild- type. [0019] Figure 7. Lethality strongly associates with loss of Smn function in muscle. Survival rates of animals expressing the N4, C24 and FL26B transgenic UAS-Smn-RNAi constructs under the control of the actinGAL4 (A), how24BGAL4 (B), and elavGAL4 (C) drivers were measured at the following developmental stages: embryo (day 0), 1 ^st instar larva (day 2), 3 ^rd instar larva (day 5), early pupa (day 7), late pupa (day 9), 2-day old adult (day 12). Each experiment was performed in triplicate. The empty pWIZ RNAi vector served as a control. The survival rates of animals were calculated and subtracted from control values. The N4, C24 and FL26B transgenic animals displayed graded viability among the drivers tested. Ubiquitous SMN knockdown (A) leads to pupal lethality. Muscle-specific SMN knockdown (B) leads to late pupal lethality only in animals harboring the stronger alleles (N4 and C24), whereas greater than 90% of FL26B individuals survive to adulthood. In contrast, reduction of SMN in neurons using N4 and

C24 (C) causes only very mild lethality (7%) when compared to control animals. (D) Western blots using an anti-SMN polyclonal antibody show reduction of SMN protein in 3 ^rd instar larvae for all three UAS-Smn-RNAi transgenic strains in combination with the ubiquitous actinGAL4 driver. The top panel shows a graded effect on SMN protein levels by the three constructs consistent with their effects on lethality. The bottom panel shows anti-α tubulin levels, which served as loading controls.

[0020] Figure 8. Muscle and neuron specific Smn RNAi knockdown causes NMJ defects. (A-I) Reduced SMN expression in the N4, C24 and FL26B UAS-Smn-RNAi transgenic constructs elicits graded effects on NMJ morphology using the ubiquitous actinGAL4 (A, D, G) as well as the tissue-specific how24BGAL4 (muscle) (B, E, H) and elavGAL4 (neuron) (C, F, I) drivers. Vector only (pWIZ) controls are shown (J, K, L). In these images the pre- and post-synaptic tissues are labeled with antibodies against Synaptotagmin (green) and Discs large (red), respectively. (M) Bouton counts for the NMJs from the genotypes shown in (A-L) were normalized for muscle area and subtracted from vector only controls. For each genotype at least 15 animals were examined. * P<0.01 and **P<0.05 was determined by the ANOVA multiple comparisons test. Scale bars represent 15 μm.

[0021] Figure 9. Schematic representation of the Smn modifier screen. Depicted are the crosses performed to identify enhancers and suppressors of 5Vn«-associated lethality. In the first stage of the screen, designed to identify Smn enhancers, tubulinGAL4 e/TM6B virgin females were mated to males from Exelixis collection strains. In this stage, the entire Exelixis collection, which affects approximately 50% of the Drosophila genome, was tested. In the Fl generation, mutations that resulted in synthetic lethality or reduced viability in trans with the Smn ^13Ao tubulinGAL4 e chromosome were defined as enhancers. In the second stage of the screen, males from Fl crosses that failed to show enhancement (P[EXeIiXiS]/+; Smn ^13Ao tubulinGAL4 e/TM6B) were mated to e/TMl, Me virgin females to identify mutations that suppressed th ¹ ° tubulinGAL4 el Smn ¹ °, e lethal phenotype. The F2 suppressor screen was performed with Exelixis mutations on first and second chromosomes as testing third chromosome mutations would require placing these mutations in cis with Smn. Additional assays were employed to eliminate false positives. Seventeen enhancers and ten suppressors met these criteria.

All 27 modifiers were subsequently examined for their ability to modify the Smn NMJ phenotype by GIuRIIA staining.

[0022] Figure 10. wit overexpression in neurons exacerbates Siwn-dependent NMJ defects. To further investigate the interaction between wit and Smn at the NMJ, the neuron-specific driver, elavGAL4, was used to express WIT in neurons. (A-F) The morphology of the NMJ, as judged by bouton numbers, between muscles 6 and 7 in the A2 segment was observed in different genetic backgrounds using the pre-synaptic (Synaptotagmin) and post-synaptic (Discs large) markers, shown in green and red, respectively. The following genotypes were examined: genotypes from (A-F and wild-type). Consistent with previous reports, neural induced expression of the UAS-wit2A transgene had no obvious effect on NMJ bouton number. A synergistic effect was observed upon the addition of a single Smn allele or ) to this background, leading to a reduction of NMJ bouton numbers. The phenotype was more severe in the background showed an approximate 50% reduction in bouton numbers whil reduced the bouton count by 20%. elavGAL4, S ^13A + (B) and elavGAL4, + (C) individuals display no significant reduction in NMJ bouton numbers compared to wild-type (G). Bouton counts were determined as above. Error bars are s.e.m.; * P<0.02 was determined by the ANOVA multiple comparisons test to wild-type and all controls, n was 15-20 animals for each genotype. Bouton numbers for each genotype were normalized to the ratio of muscle areas. Scale bars represent 20 μm.

[0023] Figure 11. Loss of mad function enhances Smn NMJ defects. (A-F) The morphology of the NMJ, as judged by bouton numbers, between muscles 6 and 7 in the A2 segment was observed in different genetic backgrounds using the pre-synaptic (Synaptotagmin) and post-synaptic (Discs large) markers, shown in green and red, respectively. The following genotypes were examined: (A) wild-type, ( ) , ( ) ) Bouton counts for genotypes in (A-F). Introduction of into either a background dominantly reduces the Smn-dependent NMJ bouton count. Error bars are

s.e.m.; *P<0.02 was determined by the ANOVA multiple comparisons test to wild-type and all controls, n was 15-20 animals for each genotype. Bouton numbers for each genotype were normalized to the ratio of muscle areas. Scale bars represent 20 μm. [0024] Figure 12. Smn knockdown reduces pMAD signals. (A-B) Wild-type wing discs from 3 ^r instar larvae were stained with antibodies against SMN (red) (A) and phosphorylated MAD (pMAD) (green) (B). (C-D) 3 ^rd instar wing discs of engrailedGAL4, pWIZ[UAS-Smn-RNAi] ^N4 animals are stained with antibodies against SMN (red) (C) and pMAD (green) (D). (E) Merge of (C) and (D). pMAD staining is reduced in the posterior region of the wing disc where SMN expression is decreased (yellow line). (F) A wing from an engrailedGAL4 ^' , pWIZ[UAS-Smn-RNAi] ^N4 transgenic adult exhibits defects in the posterior crossvein regions and the distal portions of wing veins L4 and L5 (arrow). Scale bars represent 40 μm.

[0025] Figure 13. pMAD staining of vestigalGAL4, UAS-Smn-RNAi transgenic animals. (A-B) 3rd instar wing discs of vestigalGAL4 , pWIZ[UAS-Smn-RNAi] ^N4 animals are stained with antibodies against SMN (red) (A) and pMAD (green) (B). pMAD staining is reduced in the dorsoventral boundary of the wing disc where SMN expression is decreased (see Figure 12 for wild-type control).

[0026] Figure 14. A dad null allele rescues Smn NMJ defects. (A-D) The morphology of the NMJ, as judged by bouton numbers, between muscles 6 and 7 in the A2 segment was observed in different genetic backgrounds using the pre-synaptic (Synaptotagmin) and post-synaptic (Discs large) markers, shown in green and red, respectively. The following genotypes were examined: (A) wild-type (B homozygotes and (D) S (E) Bouton counts for genotypes in (A-D). individuals display strongly reduced NMJ bouton numbers while dad ^271'68 homozygotes have a greater than two-fold of bouton numbers re ₁ la ₄ t-i-ve ₄ t.o ₄ t- _λ h-e ani •ma ils. r T-p,he d _λ ou 1b1le mutants behave like ⁷ homozygotes. Error bars are s.e.m.; n is 15-20 animals for wild-type and . n is 30 for * and P<0.002 by the ANOVA multiple comparisons test. Bouton numbers for each genotype were normalized to the ratio of muscle area. Scale bars represent 15 μm.

[0027] Figure 15. Modification of Smn lethality by mutations in the FGF signaling pathway. 5m«-induced lethality was scored at the early pupal stage and measured as a percentage relative to the viability of sibling controls. Genotypes of experimental backgrounds are shown. (A) Smn RNAi was driven under the control of the ubiquitous tubGAL4 driver in a breathless (btl) heterozygous background. (B) Effects of the simultaneous reduction of both Smn and btl using the tubGAL4 driver. (C) Smn RNAi was driven under the control of tubGAL4 in a branchless (bnl) or stumps heterozygous background.

[0028] Figure 16. Reduction in synaptic size resulting from Smn RNAi. (A) Control NMJ of muscle 4. Synaptic boutons were monitored with antibodies against DLG and HRP. (B) Reduced bouton counts were observed under conditions of Smn RNAi. (C) Rescue of synaptic size by overexpression of wild type htl. (D) Quantification of boutons/synapse of control, 24BGAL4, UAS-Smn-RNAi and 24BGAL4, UAS-Smn-RNAi, UAS htl larvae.

[0029] Figure 17. Muscle expression of a dominant negative form of htl results in reduced synapse size. (A) Control NMJ of muscle 4. (B) NMJ in animals expressing a dominant negative form of htl in muscles. Synaptic boutons were monitored with antibodies against DLG and HRP.

DETAILED DESCRIPTION

[0030] The present invention is based, in part, on the identification of novel genetic modifiers of SMN biological activity. Using a genetic screen of a Drosophila model of spinal muscular atrophy (SMA), the inventors discovered several genes, which were previously unknown to be associated with the Smn locus. For example, components of the bone morphogenetic protein signaling cascade and the fibroblast growth factor signaling cascade suppressed the deleterious effects of loss of SMN function on the neuromuscular junction. Genes involved in regulation of translation and modulation of the cytoskeleton were also found to modify the phenotype due to loss of SMN function. These genetic modifiers of Smn function provide therapeutic targets for treatments of SMA. Accordingly, the present invention provides novel approaches for treating SMA in a human subject.

[0031] In one embodiment, the present invention includes a method of treating SMA in a subject in need thereof comprising administering to the subject an agent that enhances bone morphogenetic protein (BMP) signaling. Preferably, the subject is a human subject. BMP signaling plays a role in osteogenesis, cell differentiation, anterior/posterior axis specification, growth, and homeostasis. BMPs are members of the TGF beta superfamily of ligands, which include Growth and differentiation factors (GDFs), Anti-mullerian hormone (AMH), Activin, Nodal and TGFβ ligands (e.g. TGFβl, TGFβ2, TGFβ3). BMP signaling is initiated by ligand binding to the Type II BMP receptor, a serine/threonine receptor kinase, which in turn causes the recruitment and phosphorylation of a BMP type I receptor. Upon phosphorylation, the BMP Type I receptor phosphorylates a receptor regulated SMAD (R-SMAD) protein (e.g. SMADl, SMAD5, or SMAD9), which then interacts with a co-SMAD protein, SMAD4. The R-SMAD/co-SMAD complex translocates to the nucleus where it activates BMP-responsive genes. In some embodiments, the agent for treating SMA enhances the activity of a BMP type II receptor. For instance, the agent may be an agonist of a BMP type II receptor, such as BMP2 or BMP7. The receptor agonist may be a natural ligand of the BMP type II receptor or a synthetic ligand or an antibody or fragment thereof that binds and activates the receptor. In certain embodiments, the agent enhances the activity of one or more R- SMAD proteins, such as SMADl, SMAD5, and SMAD9. This enhancement can occur directly by increasing the expression, phosphorylation, or interaction of the R-SMAD with the receptor complex. The enhancement can also occur indirectly by increasing or enhancing the activity of the BMP type II receptor or its interaction with the type I receptor. The agent for treating SMA can also include nucleic acids encoding a BMP type II receptor, a BMP type I receptor, or a R-SMAD that mediates BMP signaling (e.g. SMADl, SMAD5, SMAD9). Such genetic therapy approaches employing adenoviral vectors encoding a BMP type II receptor have been successful for treating idiopathic pulmonary arterial hypertension in rats (Reynolds et al. (2007) Am J Physiol Lung Cell MoI Physiol. Vol. 292: LI l 82-Ll 192).

[0032] There are also two inhibitory SMAD proteins (e.g. SMAD6 and SMAD7), which negatively regulate BMP/TGF beta signaling. SMAD7 competes for binding of the Type I receptor with the R-SMAD proteins and prevents their phosphorylation. SMAD6 binds

to SMAD4 and prevents its interaction with phosphorylated R-SMAD proteins. Thus, in certain embodiments, the agent for treating SMA inhibits the activity of SMAD6 or SMAD7 protein, which acts to enhance signaling through the BMP and/or TGF-beta cascade. For example, agents that can inhibit the activity or expression of SMAD6 and/or SMAD7 include antisense oligonucleotides and inhibitory RNA molecules, such as small interfering RNAs (siRNAs), small hairpin RNAs (shRNAs), and ribozymes, targeted to SMAD6 and/or SMAD7 sequences. In some embodiments, the agents for inhibiting SMAD6 and/or SMAD7 activity include antibodies or fragments thereof that bind to SMAD6 and/or SMAD7 and interfere with their ability to bind to SMAD4 and/or the Type I BMP receptor.

[0033] In another embodiment, the present invention encompasses a method of treating SMA in a subject in need thereof comprising administering to the subject an agent that enhances transforming growth factor-beta (TGF-beta) signaling. Preferably, the subject is a human subject. TGF-beta signaling is similar in mechanism and function to BMP signaling and is known to play a role in embryogenesis, cell differentiation, and apoptosis. TGF-beta ligands (TGFβl, TGFβ2, TGFβ3) bind to type II TGF-beta receptors and initiate a signaling cascade similar to the BMP signaling cascade, which results in the phosphorylation of R-SMADs (e.g. SMAD2 and SMAD3). SMAD2/SMAD4 or SMAD3/SMAD4 complexes translocate to the nucleus to effect activation of TGF-beta responsive genes. Thus, in certain embodiments, an agent for treating SMA includes type II TGF-beta receptor agonists, such as natural TGF-beta ligands, synthetic ligands, antibodies or fragments thereof that bind and activate the type II receptor, or small molecule TGF-beta mimetics (see, e.g., Glaser et al. (2002) MoI Cancer Ther., Vol. l(10):759-68). The agent can also be a nucleic acid encoding a type II TGF-beta receptor, a type I TGF-beta receptor, or a R-SMAD that mediates TGF-beta signaling (e.g. SMAD2 or SMAD3). The invention also contemplates small molecule compounds that enhance TGF-beta or BMP signaling as agents useful for treating SMA in a subject.

[0034] In another embodiment, the present invention provides a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that enhances fibroblast growth factor signaling. Preferably, the subject is a human

subject. Fibroblast growth factors (FGFs), are involved in angiogenesis, wound healing, embryonic development, and cell proliferation and differentiation in various tissues. FGF signaling has been implicated in neural development and degeneration and has been shown to operate through mechanisms that include alternative splicing in vertebrates (Irving, et al. (2002) Development, Vol. 129(23): 5389-98; Shirasaki, et al. (2006) Neuron, Vol. 50(6): 841-53; Umemori, et al. (2004) Cell, Vol. 118(2): 257-70). Moreover, in C. elegans and in mice, FGF signals have been shown to play a role in axon branching (Szebenyi, et al. (2001) J Neurosci, Vol. 21(11): 3932-41) and NMJ maintenance (Bulow, et al. (2004) Neuron, Vol. 42(3): 367-74). There are four human FGF receptor genes (FGF-I, FGF-2, FGF-3, and FGF-4) each of which can generate different splice variants to give rise to different subtypes of receptors. The FGF receptors are receptor tyrosine kinases that dimerize and autophosphorylate upon ligand binding. The tyrosine phosphorylated residues in the cytoplasmic tails of the dimerized FGF receptors recruit different signaling molecules to initiate various signaling cascades, including the phosopholipase C/protein kinase C cascade and the RAS/mitogen-activated protein (MAP) kinase cascade. In certain embodiments, the agent for treating SMA enhances the activity of FGFR-2 and/or FGFR-3. For example, the agent can be a FGF receptor agonist including, but not limited to, natural or synthetic ligands of FGF receptors, antibodies that bind and activate the FGF receptors and fragments thereof, nucleic acids encoding one or more of the human FGF receptor genes or splice variants thereof, and small molecule compounds that activate FGF receptors. [0035] In another embodiment, the agent enhances downstream components of the FGF receptor signaling cascade. For instance, in some embodiments, the agent enhances RAS activation. In other embodiments, the agent enhances MAP kinase activation. The agent can, in certain embodiments, act directly on the downstream components of FGF signaling. Thus, agonists of RAS and MAP kinase {e.g. nucleic acids encoding these genes) are also contemplated as agents useful for treating SMA according to the methods of the invention.

[0036] Glycogen synthase kinase 3 (GSK-3) is thought to negatively affect FGF signaling. Thus, inhibition of GSK-3 can result in enhanced FGF signaling. In some embodiments, GSK-3 inhibitors can be used as agents for treating SMA as described

herein. GSK-3 inhibitors suitable for use in the methods of the invention include, but are not limited to, Hymenialdisine, Flavopiridol, Kenpaullone, Alsterpaullone, Azakenpaullone, Indirubin-30-oxime, 6-Bromoindirubin-30-oxime (BIO), 6- Bromoindirubin-30-acetoxime, Aloisine A, Aloisine B, TDZD8, Compound 12, Pyrazolopyridine 18, Pyrazolopyridine 9, Pyrazolopyridine 34, CHIR98014, CHIR99021 (CT99021), Compound 1, SU9516, ARA014418, Staurosporine, Compound 5a, Compound 29, Compound 46, GF109203x (bisindolylmaleimide I), Ro318220 (bisindolylmaleimide IX), SB216763, SB415286, CGP60474, Compound 8b, TWSl 19, Compound IA, Compound 17, and those described, for example, in U.S. Application Publication No. 20080262205.

[0037] The present invention also includes a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that increases the expression or activity of a SMN agonist. As used herein, a "SMN agonist" is a gene or protein that positively regulates SMN function. A SMN agonist can also refer to a gene or protein that acts cooperatively or synergistically with SMN and/or a gene or protein that replaces or compensates for SMN function. Non-limiting examples of SMN agonists include Pumilio homolog 1, eIF-4E, MAPlB, Rhol, plastin3, type II BMP receptor, type II TGF-beta receptor, R-SMAD protein (e.g. SMADl, SMAD2, SMAD3, SMAD5, and SMAD9), FGF-2 or FGF-3 receptor, RAS, and MAP kinase. [0038] In some embodiments, the agent for increasing expression or activity of a SMN agonist is an expression construct encoding the gene for the SMN agonist, wherein the gene is overexpressed following administration of the expression construct. For purposes of this application, the terms "expression construct," "expression vector," and "vector," are used interchangeably to demonstrate the application of the invention in a general, illustrative sense, and are not intended to limit the invention. A "vector" or "construct" is a composition of matter which can be used to deliver a nucleic acid of interest to the interior of a cell. Numerous vectors are known in the art including, but not limited to, linear polynucleotides, polynucleotides associated with ionic or amphiphilic compounds, plasmids, and viruses. Thus, the term "vector" includes an autonomously replicating plasmid or a virus. Examples of viral vectors include, but are not limited to, adenoviral

vectors, adeno-associated virus vectors, retroviral vectors, and the like. An expression construct can be replicated in a living cell, or it can be made synthetically. [0039] In one embodiment, an expression construct for expressing a SMN agonist comprises a promoter "operably linked" to a polynucleotide encoding a gene for a SMN agonist (e.g. a polynucleotide encoding Pumilio homolog 1, eIF-4E, MAPlB, Rhol, plastin3, type II BMP receptor, type II TGF -beta receptor, R-SMAD protein, FGF-2 or FGF-3 receptor, RAS, and MAP kinase). The phrase "operably linked" or "under transcriptional control" as used herein means that the promoter is in the correct location and orientation in relation to a polynucleotide to control the initiation of transcription by RNA polymerase and expression of the polynucleotide. A "promoter" refers to a DNA sequence recognized by the synthetic machinery of the cell, or introduced synthetic machinery, required to initiate the specific transcription of a gene. The term promoter will be used here to refer to a group of transcriptional control modules that are clustered around the initiation site for RNA polymerase I, II, or III. In some embodiments, constitutive promoters, such as the human cytomegalovirus (CMV) immediate early gene promoter, the SV40 early promoter, the Rous sarcoma virus long terminal repeat, rat insulin promoter and glyceraldehyde-3 -phosphate dehydrogenase, can be used to obtain high-level expression of the polynucleotide sequence of interest. The use of other viral or mammalian cellular or bacterial phage promoters which are well-known in the art to achieve expression of a polynucleotide sequence of interest is contemplated as well, provided that the levels of expression are sufficient for a given purpose. [0040] In some embodiments, the expression constructs for expressing a SMN agonist comprise a tissue-specific promoter operably linked to the polynucleotide encoding the SMN agonist gene. In one embodiment, the tissue-specific promoter is a muscle-specific promoter. Muscle-specific promoters suitable for use in constructs expressing SMN agonists include, but are not limited to, the myosin light chain-2 promoter, the α-actin promoter, the troponin 1 promoter; the Na+/Ca2+ exchanger promoter, the dystrophin promoter, the α7 integrin promoter, and the muscle creatine kinase (MCK) promoter. One of ordinary skill in the art can select other appropriate promoters for directing expression of the gene of interest in muscle cells. In another embodiment, the tissue- specific promoter is a neuron-specific promoter. Neuron-specific promoters can include,

but are not limited to, Tαl α-tubulin promoter, GluR2 promoter, synapsin 1 promoter, neuron-specific enolase (NSE) promoter, neuronal nicotinic acetylcholine receptor beta 2-subunit promoter, calcium/calmodulin kinase II promoter, platelet-derived growth factor b-chain (PDGF) promoter, and MAPlB promoter. Other appropriate neuron- specific promoters are known to those of skill in the art.

[0041] In other embodiments, the expression constructs comprise inducible promoters that can be activated to produce the gene product under certain conditions. Inducible promoters are known in the art, and include, but are not limited to tetracycline promoter, metallothionein HA promoter, heat shock promoter, steroid/thyroid hormone/retinoic acid response elements, the adenovirus late promoter, and the inducible mouse mammary tumor virus LTR.

[0042] In certain embodiments, an agent that increases the expression or activity of a SMN agonist can be the protein product of the SMN agonist itself (e.g. Pumilio homolog 1, eIF-4E, MAPlB, Rhol, plastin3, type II BMP receptor, type II TGF-beta receptor, R- SMAD protein, FGF-2 or FGF-3 receptor, RAS, and MAP kinase) or another protein or gene product that regulates the expression or activity of the SMN agonist, such as a transcription factor or other protein that acts upstream of the SMN agonist in a particular signaling cascade. In one embodiment, the agent is a small molecule compound that directly or indirectly increases the expression and/or activity of the SMN agonist. [0043] The present invention also encompasses a method of treating spinal muscular atrophy in a subject in need thereof comprising administering to the subject an agent that decreases the expression or activity of a SMN antagonist. As used herein, a "SMN antagonist" is a gene or protein that negatively regulates SMN function. A SMN antagonist can also refer to a gene or protein that acts to interfere or compete for binding with SMN target proteins. SMN antagonists include, but are not limited to, Fmrl, Moesin, slik, SMAD6, and SMAD7. In some embodiments, an agent that decreases the expression or activity of a SMN antagonist is a small molecule compound that directly or indirectly decreases the expression and/or activity of the SMN antagonist. In other embodiments, an agent that decreases the expression or activity of a SMN antagonist is an antibody or fragment thereof that binds to the SMN antagonist and prevents its interaction with other proteins and/or inhibits its activity.

[0044] In certain embodiments, an agent that decreases the expression or activity of a SMN antagonist is an antisense nucleic acid targeted to a sequence of the SMN antagonist. Suitable antisense nucleic acids can comprise ribonucleotides or deoxyribonucleotides and preferably, have at least one chemical modification. Such modifications include without limitation locked nucleic acids, peptide nucleic acids, sugar modifications, such as 2'-O-alkyl (e.g. 2'-O-methyl, 2'-O-methoxyethyl), 2'-fluoro, and 4' thio modifications, and backbone modifications, such as one or more phosphorothioate, morpholino, or phosphonocarboxylate linkages (see, for example, U.S. Patent Nos. 6,693,187 and 7,067,641, which are herein incorporated by reference in their entireties). Other modifications of antisense nucleic acids to enhance stability and improve efficacy, such as those described in U.S. Patent No. 6,838,283, which is herein incorporated by reference in its entirety, are known in the art and are suitable for use in the methods of the invention. Preferable antisense nucleic acids useful for inhibiting the expression and/or activity of a SMN antagonist are about 20 to about 200 nucleotides in length. Antisense nucleic acids can comprise a sequence that is at least partially complementary (e.g. at least about 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% complementary) to a gene sequence for a SMN antagonist or portion thereof. In one embodiment, the antisense nucleic acid comprises a sequence that is 100% complementary to a gene sequence for a SMN antagonist or portion thereof. The antisense nucleic acid can target either a coding or non-coding region of the SMN antagonist gene. In some embodiments, the antisense nucleic acid targets a mRNA transcript from the SMN antagonist gene.

[0045] In other embodiments, an agent that decreases the expression or activity of a SMN antagonist is an inhibitory RNA molecule targeted to a sequence of the SMN antagonist. The inhibitory RNA molecule may be a double-stranded, small interfering RNA (siRNA) or a short hairpin RNA molecule (shRNA) comprising a stem-loop structure or a ribozyme. The double-stranded regions of the inhibitory RNA molecule may comprise a sequence that is at least partially identical and partially complementary, e.g. about 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% identical and complementary, to a coding or non-coding region of a gene sequence for a SMN antagonist. In one embodiment, the double-stranded regions of the inhibitory RNA molecule may contain

100% identity and complementarity to the gene sequence for a SMN antagonist. In another embodiment, the inhibitory RNA molecule targets a mRNA transcript from the SMN antagonist gene.

[0046] The antisense nucleic acid or inhibitory RNA molecule targeted to a SMN antagonist can be encoded on an expression construct as described herein. In one embodiment, the antisense nucleic acid or inhibitory RNA molecule is under the control of a tissue-specific promoter. In a preferred embodiment, the tissue-specific promoter is a muscle-specific promoter. In another preferred embodiment, the tissue-specific promoter is a neuron-specific promoter.

[0047] Preferably, a pharmaceutically effective amount of an agent for treating SMA is administered to the subject (e.g. human subject). As used herein, the term "pharmaceutically effective amount" means an amount that improves one or more symptoms of SMA. Symptoms of SMA include, but are not limited, to muscle weakness, muscle atrophy, motor neuron loss, decreased life expectancy, poor muscle tone, decreased or absent deep tendon reflexes, twitching of leg, arm or tongue muscles, abnormal gait, or difficulty breathing. In some embodiments of the invention, at least one symptom of SMA is alleviated following administration of an agent that increases the expression or activity of a SMN agonist, an agent that decreases the expression or activity of a SMN antagonist, or an agent that enhances BMP or FGF signaling. In one embodiment, motor unit number estimation can be used to monitor symptoms of a SMA patient. Motor unit number estimation is a technique that allows the determination of the number of motor units present in a muscle by measuring electromyography signals. Preferably, an increase in the number of motor units and/or the size of the motor units is observed following administration of a pharmaceutically effective amount of an agent for treating SMA.

[0048] Formulation of an agent described herein for treatment purposes comprises combining pharmaceutically effective amounts of the agent of the invention with pharmaceutically acceptable diluents, preservatives, solubilizers, emulsifiers, adjuvants and/or carriers. Such compositions include diluents of various buffer content (e.g., Tris- HCl, acetate, phosphate), pH and ionic strength; additives such as detergents and solubilizing agents (e.g., Tween 80, Polysorbate 80), anti-oxidants (e.g., ascorbic acid,

sodium metabisulfite), preservatives (e.g., Thimerosol, benzyl alcohol) and bulking substances (e.g., lactose, mannitol); incorporation of the material into particulate preparations of polymeric compounds such as polylactic acid, polyglycolic acid, etc. or into liposomes. Protein agents of the invention may be produced as fusion proteins to modulate or extend the half- life of the protein. Such fusion proteins may include human serum albumin, transferrin, other serum proteins, etc. Such compositions may influence the physical state, stability, rate of in vivo release, and rate of in vivo clearance of the present compounds. See, e.g., Remington's Pharmaceutical Sciences, 18th Ed. (1990, Mack Publishing Co., Easton, Pa. 18042) pages 1435-1712. The compositions may be prepared in liquid form, or may be in dried powder, such as lyophilized form. Implantable sustained release formulations are also contemplated. Preferably, pharmaceutical compositions will be prepared in a form appropriate for the intended application and be essentially free of pyrogens, as well as other impurities that could be harmful to humans or animals.

[0049] Colloidal dispersion systems, such as macromolecule complexes, nanocapsules, microspheres, beads, and lipid-based systems including oil-in-water emulsions, micelles, mixed micelles, and liposomes, can be used as delivery vehicles for the therapeutic agents described herein, especially for nucleic acid-based therapeutic agents (e.g. expression vectors, antisense nucleic acids, and inhibitory RNA molecules). Commercially available fat emulsions that are especially suitable for delivering the nucleic acid agents of the invention to tissues, such as skeletal muscle tissue, include Intralipid®, Liposyn®, Liposyn® II, Liposyn® III, Nutrilipid, and other similar lipid emulsions. A preferred colloidal system for use as a delivery vehicle in vivo is a liposome (i.e., an artificial membrane vesicle). The preparation and use of such systems is well known in the art. Exemplary formulations are also disclosed in US 5,981,505; US 6,217,900; US 6,383,512; US 5,783,565; US 7,202,227; US 6,379,965; US 6,127,170; US 5,837,533; US 6,747,014; and WO 03/093449, which are herein incorporated by reference in their entireties.

[0050] Administration of the agents according to the methods of the present invention may be via any common route so long as the target tissue (e.g. skeletal muscle, motor neurons) is available via that route. This includes oral, nasal, or buccal. Alternatively,

administration may be by intradermal, subcutaneous, intramuscular, intraperitoneal, intrathecal, intraventricular, intraparenchymal, intraarterial or intravenous injection, or by direct injection into skeletal muscle tissue or motor neurons. The therapeutic agents described herein would normally be administered as pharmaceutically acceptable compositions, as described herein. The agents may also be administered parenterally or intraperitoneally. By way of illustration, solutions of the therapeutic agents as free base or pharmacologically acceptable salts can be prepared in water suitably mixed with a surfactant, such as hydroxypropylcellulose. Dispersions can also be prepared in glycerol, liquid polyethylene glycols, and mixtures thereof and in oils. Under ordinary conditions of storage and use, these preparations generally contain a preservative to prevent the growth of microorganisms.

[0051] The pharmaceutical forms suitable for injectable use include, for example, sterile aqueous solutions or dispersions and sterile powders for the extemporaneous preparation of sterile injectable solutions or dispersions. Generally, these preparations are sterile and fluid to the extent that easy injectability exists. Preparations should be stable under the conditions of manufacture, storage, and administration (depot delivery) and should be preserved against the contaminating action of microorganisms, such as bacteria and fungi. Appropriate solvents or dispersion media may contain, for example, water, ethanol, polyol (for example, glycerol, propylene glycol, and liquid polyethylene glycol, and the like), suitable mixtures thereof, and vegetable oils. The proper fluidity can be maintained, for example, by the use of a coating, such as lecithin, by the maintenance of the required particle size in the case of dispersion and by the use of surfactants. The prevention of the action of microorganisms can be brought about by various antibacterial an antifungal agents, for example, parabens, chlorobutanol, phenol, sorbic acid, thimerosal, and the like. In many cases, it will be preferable to include isotonic agents, for example, sugars or sodium chloride. Prolonged absorption of the injectable compositions can be brought about by the use in the compositions of agents delaying absorption, for example, aluminum monostearate and gelatin.

[0052] Sterile injectable solutions may be prepared by incorporating the therapeutic agents in an appropriate amount into a solvent along with any other ingredients (for example as enumerated above) as desired, followed by filtered sterilization. Generally,

dispersions are prepared by incorporating the various sterilized active ingredients into a sterile vehicle which contains the basic dispersion medium and the desired other ingredients, e.g., as enumerated above. In the case of sterile powders for the preparation of sterile injectable solutions, the preferred methods of preparation include vacuum- drying and freeze-drying techniques which yield a powder of the active ingredient(s) plus any additional desired ingredient from a previously sterile-filtered solution thereof. Upon formulation, solutions are preferably administered in a manner compatible with the dosage formulation and in such amount as is therapeutically effective. The formulations may easily be administered in a variety of dosage forms such as injectable solutions, drug release capsules and the like. For parenteral administration in an aqueous solution, for example, the solution generally is suitably buffered and the liquid diluent first rendered isotonic for example with sufficient saline or glucose. Such aqueous solutions may be used, for example, for intrathecal, intravenous, intramuscular, subcutaneous and intraperitoneal administration. Preferably, sterile aqueous media are employed as is known to those of skill in the art, particularly in light of the present disclosure. Some variation in dosage will necessarily occur depending on the stage of SMA {e.g. type I, type II, or type III) to be treated and individual characteristics of the subject to be treated {e.g. size, age, overall health, etc). The person responsible for administration will, in any event, determine the appropriate dose for the individual subject. Moreover, for human administration, preparations should meet sterility, pyrogenicity, general safety and purity standards as required by FDA Office of Biologies standards.

[0053] The present invention also includes a method for modulating SMN biological function in a cell. As used herein, "SMN biological function" refers to the function of native SMN protein in a cell, which includes, but is not limited to, the assembly of spliceosomal small nuclear ribonucleoproteins (snRNPs) and associated spliceosome functions, regulation of axonal morphology, survival of alpha motorneurons, and maintenance of the neuromuscular junction.

[0054] In one embodiment, the method for modulating SMN biological function in a cell comprises increasing the expression or activity in the cell of one or more genes selected from the group consisting of Mothers against Dpp {Mad), Daughters against Dpp {Dad), pumilio,futsch, MAPlB, heartless {htl), stumps, branchless {bnl), FGFR-2, Rhol, elF-

4E,fimbrin,plastin3, slik, and the genes listed in Table 1. In some embodiments, increasing the expression or activity of the one or more genes comprises delivering an expression construct to the cell, wherein the expression construct encodes the one or more genes and wherein the one or more genes are overexpressed following delivery of said expression construct to the cell. In certain embodiments, SMN biological function is enhanced following delivery of the expression construct. For instance, SMN biological function is enhanced in the cell following delivery of an expression construct encoding a gene selected from the group consisting of CGl 0776 (wit), human bone morphogenetic protein receptor Il (BMPRII), CG8127 (Eip75B), human peroxisome proliferator- activated receptor gamma (PPAR]), CGl 927, Mothers against Dpp (Mad), pumilio, elF- 4E,futsch, MAPlB, CG32134 (btl), heartless {hit), stumps, branchless (bnl), FGFR-2, FGFR-3, Rhol ,fimbrin, andplastin3.

[0055] In other embodiments, SMN biological function is reduced in the cell following delivery of an expression construct. For example, SMN biological function is reduced in the cell following delivery of an expression construct encoding a gene selected from the group consisting of Daughters against Dpp (Dad), CG8920, human TDRD7, CGl 3775, human RASD2, CG1697 (rho-4), human RHBDL3, CG6203 (Fmrl), moesin, radixin, human Fmrl, and slik.

[0056] In another embodiment, the method for modulating SMN biological function in a cell comprises decreasing the expression or activity in the cell of one or more genes selected from the group consisting of Mothers against Dpp (Mad), Daughters against Dpp (Dad), pumilio, futsch, MAPlB, heartless (htl), stumps, branchless (bnl), FGFR-2, Rhol , eIF-4E,fιmbrin,plastin3, slik, and the genes listed in Table 1. In certain embodiments, decreasing the expression or activity of the one or more genes comprises delivering an expression construct to the cell, wherein the expression construct encodes a nucleic acid that attenuates the expression or activity of the one or more genes and wherein the expression of the one or more genes is reduced following delivery of the expression construct to the cell. In some embodiments, SMN biological function is enhanced in the cell following delivery of the expression construct. For instance, SMN biological function is enhanced in the cell following delivery of an expression construct encoding a nucleic acid that attenuates the expression or activity of a gene selected from

the group consisting of Daughters against Dpp (Dad), CG8920, human TDRD7, CGl 3775, human RASD2, CGl 697 (rho-4), human RHBDLS, CG6203 (Fmrl), moesin, radixin, human Fmrl, and slik. In other embodiments, SMN biological function is reduced in the cell following delivery of the expression construct. For instance, SMN biological function is reduced in the cell following delivery of an expression construct encoding a nucleic acid that attenuates the expression or activity of a gene selected from the group consisting of CGl 077 '6 (wit), human bone morphogenetic protein receptor II (BMPRII), CG8127 (Eip75B), human peroxisome proliferator-activated receptor gamma (PPAR)), CGl 927, Mothers against Dpp (Mad),pumilio, eIF-4E,futsch, MAPlB, CG32134 (btl), heartless (htl), stumps, branchless (bnl), FGFR-2, FGFR-S, Rhol, fimbrin, and plastin3. The nucleic acid that attenuates the expression or activity of one or more genes can be antisense nucleic acids or inhibitory RNA molecules (e.g. ribozymes, siRNAs, shRNAs) as described herein. The nucleic acid targets a sequence from at least one or more of the genes listed above.

[0057] In another embodiment, the present invention provides a method for modulating SMN biological function in a cell comprising contacting the cell with at least one compound, wherein the at least one compound modulates expression or activity of a gene selected from the group consisting of Mothers against Dpp (Mad), Daughters against Dpp (Dad), pumilio,futsch, MAPlB, heartless (htl), stumps, branchless (bnl), FGFR-2, Rhol, eIF -4E, fimbrin, plastin3, slik, and the genes listed in Table 1. In some embodiments, the at least one compound increases the expression or activity of the one or more genes. In certain embodiments, SMN biological function is enhanced in the cell following contact with the at least one compound, wherein the at least one compound increases the expression or activity of a gene selected from the group consisting of CGl 0776 (wit), human bone morphogenetic protein receptor II (BMPRII), CG8127 (Eip75B), human peroxisome proliferator-activated receptor gamma (PPAR]), CGl 927, Mothers against Dpp (Mad),pumilio, eIF-4E,futsch, MAPlB, CG32134 (btl), heartless (htl), stumps, branchless (bnl), FGFR-2, FGFR-3, Rhol , fimbrin, andplastin3. [0058] In other embodiments, SMN biological function is reduced in the cell following contact with the at least one compound, wherein the at least one compound increases the expression or activity of a gene selected from the group consisting of Daughters against

Dpp (Dad), CG8920, human TDRD7, CGl 3775, human RASD2, CGl 697 (rho-4), human RHBDL3, CG6203 (Fmrl), moesin, radixin, human Fmrl, and slik. [0059] In yet another embodiment, the at least one compound decreases the expression or activity of the one or more genes. In certain embodiments, SMN biological function is enhanced in the cell following contact with the at least one compound, wherein the at least one compound decreases the expression or activity of a gene selected from the group consisting of Daughters against Dpp (Dad), CG8920, human TDRD7, CGl 3775, human RASD2, CG1697 (rho-4), human RHBDL3, CG6203 (Fmrl), moesin, radixin, human Fmrl , and slik. In other embodiments, SMN biological function is reduced in the cell following contact with the at least one compound, wherein the at least one compound decreases the expression or activity of a gene selected from the group consisting of CGl 0776 (wit), human bone morphogenetic protein receptor II (BMPRII), CG8127 (Eip75B), human peroxisome proliferator-activated receptor gamma (PPAR]), CGl 927, Mothers against Dpp (Mad),pumilio, eIF-4E,futsch, MAPlB, CG32134 (btϊ), heartless (htl), stumps, branchless (bnl), FGFR-2, FGFR-3, Rhol ,fimbrin, andplastin3. [0060] SMN biological function can be modulated according to the methods of the invention in various cell types including, but not limited to, C elegans cells, mammalian cells (e.g. human cells), insect cells, and zebrafish cells. Non-limiting examples of insect cells are, Spodoptera frugiperda (Sf) cells, e.g. Sf9, Sf21, Trichoplusia ni cells, e.g. High Five cells, and Drosophila S2 cells. Examples of mammalian cells include COS cells, baby hamster kidney cells, mouse L cells, LNCaP cells, Chinese hamster ovary (CHO) cells, human embryonic kidney (HEK) cells, Vero cells, CVl cells, HeLa cells, MDCK cells, Hep-2 cells, and muscle cell lines (e.g. C2C12 cells). Human cells can include fibroblasts from SMA patients, human skeletal muscle cells, human spinal motor neurons, embryonic stem cell (ES)-derived alpha motor neurons from SMA animal models, or induced pluripotent stem cell (IPS)-derived motor neurons from SMA patients or SMA animal models. The cell may be in vitro or in vivo. Methods of delivering expression constructs and nucleic acids to cells are known in the art and can include, for example, calcium phosphate co-precipitation, electroporation, microinjection, DEAE- dextran, lipofection, transfection employing polyamine transfection reagents, cell sonication, gene bombardment using high velocity microprojectiles, and receptor-

mediated transfection. Homo logs and orthologs of the genes listed in Table 1 and described herein are also included in the present invention.

[0061] The present invention also provides a hypomorphic Smn Drosophila strain as an improved model of SMA disease. The Drosophila genome harbors a single copy of the Smn gene, which encodes a highly conserved homologue of SMN. The Smn loss of function allele, Smn ¹ °, results in recessive larval lethality and, importantly, neuromuscular junction abnormalities [15,18,23]. The inventors employed an RNA interference (RNAi) strategy to create a series of loss of function {e.g. reduced function) Smn alleles, whose pheno types mimic the dosage dependent nature of human SMA pathology. Thus, these hypomorphic Smn strains more closely mimic the human disease state and can be used to screen candidate compounds for novel treatments for SMA. [0062] Thus, the present invention provides a method of screening candidate compounds for treatment of SMA. In one embodiment, the method comprises exposing transgenic Drosophila to one or more candidate compounds, wherein the transgenic Drosophila comprise at least one transgene expressing an inhibitory RNA molecule against the Smn gene; comparing the phenotype of the exposed transgenic Drosophila to the phenotype of transgenic Drosophila not exposed to said one or more compounds; and selecting said one or more compounds that produce a change in phenotype, wherein the selected one or more compounds are therapeutic compounds for the treatment of spinal muscular atrophy. In some embodiments, the change in phenotype is a reduction in lethality. In other embodiments, the change in phenotype is an increase in the number of synaptic boutons at the neuromuscular junction. For instance, the number of synaptic boutons can be measured by staining muscle tissue from the exposed Drosophila with fluorescent pre- and post-synaptic markers to visualize pre-synaptic terminals and post-synaptic junctions. Suitable markers for visualizing pre-synaptic terminals include, but are not limited to, synaptotagmin (SYT), horseradish peroxidase, NC82-bruchpilot-active zone marker, and FasII-peri-active zone, while suitable markers for visualizing post-synaptic sites include Discs Large (DLG), glutamate receptors (GIuRIIA, GIuRIIC), Dystroglycan, Dystrophin, and Coracle. In still other embodiments, the change in phenotype is an increase in the number of pigmented pupae. The appearance of pigmentation can be used as an indicator

of the stage of development a pupa has reached and therefore may be used to determine a shift in the lethal phase of a population of pupae (see Example 6). [0063] Exposing the transgenic Drosophila to one or more candidate compounds can comprise exposing the Drosophila to an aerosolized from of the one or more candidate compounds or can comprise feeding the one or more candidate compounds to the transgenic Drosophila. By way of example, a general feeding protocol comprises exposing 1st instar larvae to a filter paper soaked with a solution of the one or more candidate compounds of various concentrations for six hours, transferring the exposed animals to normal food for 18 hours, and repeating the procedure for three subsequent days until they reach the late 3rd larval instar stage. The one or more candidate compounds can include proteins, peptides, polypeptides, polynucleotides, oligonucleotides, RNA molecules {e.g. siRNA, shRNA), or small molecules. [0064] In certain embodiments, the transgenic Drosophila comprise at least one transgene expressing an siRNA targeted to the Drosophila Smn gene. In one embodiment, the siRNA is directed to the full length Smn gene. In another embodiment, the siRNA is directed to the amino terminal portion of the Smn gene. The amino terminal portion may include the entire Tudor domain. In still another embodiment, the siRNA is directed to the carboxy terminal portion of the Smn gene. The carboxy terminal portion preferably does not include the Tudor domain. Preferably, the expression level of the Smn gene is significantly reduced in the transgenic Drosophila carrying the transgene. Exemplary siRNA constructs for targeting the Smn gene or portions thereof are described in Example 1 and Figure 4A. The transgene can be expressed ubiquitously by driving the transgene with ubiquitous drivers, such as the tubulin and actin drivers. In some embodiments, the transgene is expressed in muscles using a muscle-specific driver (e.g. how24BGAL4, 24B-GaU, cl 79-Gal4, M12-Gal4, and MHC-GaU). In other embodiments, the transgene is expressed in neurons using a neuronal driver (e.g. elavGAL4, 1407-GaU, OKo-GaU, OK309-GaU).

[0065] This invention is further illustrated by the following additional examples that should not be construed as limiting. The contents of all references, patents and published patent applications cited throughout this application, as well as the Figures, are incorporated herein by reference in their entirety.

EXAMPLES

Example 1. Mutations in Smn compromise viability and cause neuromuscular junction defects

[0066] Elegant biochemical studies established the importance of the SMN protein in a ubiquitous, multimeric complex involved in the assembly of spliceosomal small nuclear ribonucleoproteins (snRNPs) [9,10,11,12,13,14]. Despite its seemingly fundamental and indispensable role in cellular metabolism, reduction of SMN leads to a specific neurodegenerative profile associated with SMA disease [1,15,16,17,18]. Though several recent studies indicate that SMN influences motor neuron axonal morphology [19,20], it remains unclear whether SMN has a specific neuromuscular junction (NMJ) function, and whether the functional requirement for SMN activity is increased at the NMJ than elsewhere in the organism.

[0067] The dichotomy between the ubiquitous housekeeping function of Smn and the very specific neuromuscular SMA phenotype raises the question whether Smn functions differently at the neuromuscular junction (NMJ) than in other tissue types. Specifically, whether SMN has a differential expression pattern in neurons and muscle and whether SMN concentrates to any particular cellular compartments at the NMJ remain open questions.

[0068] To determine in which tissue(s) SMN is expressed in Drosophila, antibodies against full-length Drosophila SMN were raised and the expression pattern for SMN was monitored, particularly at the NMJ. In Western blots performed on lysates derived from S2 cells, 3 ^rd instar larvae, and wild-type adult heads, the antibody recognizes a single ~28kD band [18], corresponding to the predicted molecular weight of Drosophila SMN (Figure 1 and data not shown). Moreover, when a FLAG-tagged Smn transgenic construct (UAS-FLAG-Smn) was expressed under the control of the vestigialGAL4 driver, SMN and FLAG staining overlapped at the dorsal-ventral (DV) boundary of 3 ^rd instar larval wing discs. In addition, vestigialGAL4-dirQctQd expression of an inducible RNAi allele of Smn (see Example 6) abolished the SMN staining pattern along the DV boundary of the larval wing disc (Figure 1). Together, these results indicate the specificity of the antibody raised against SMN.

[0069] Using this antibody, SMN expression at the NMJ was probed and SMN antigens were found to be clearly concentrated at the post-synaptic regions in the muscle, co- localizing with the post-synaptic marker Discs Large (DLG) (Figure 2A-D) [29]. Under these conditions, antigens in the pre-synaptic region of the motor neuron terminal were not detected (as defined by horseradish peroxidase (HRP) staining) at the NMJ (Figure 2A-D). SMN staining was also observed within muscle fibers and at discrete foci in muscle nuclei (Figure 2C and E), which presumably reflect SMN localization in Cajal bodies (gems) as demonstrated for mammalian cells [9], and in Drosophila ovarian nurse cells and oocytes [30]. This post-synaptic NMJ expression pattern of SMN is abolished by muscle-specific Smn RNAi knockdown, again demonstrating the specificity of the anti-SMN antibodies (Figure 3). Consistent with its general role in snRNP assembly, SMN was detected in all tissues examined, including muscle (Figure 2A-D) and neurons (Figure 2F). However, at the Drosophila NMJ, SMN is concentrated at the post-synaptic regions in the muscle.

Mutations in Smn compromise viability

[0070] Previous studies determined that loss of Smn function results in larval lethality [15,18]. Two additional Smn alleles found within the Exelixis collection (Harvard Medical School), and were examined [25,26]. Sequence analysis of both strains indicates each allele harbors a transposon insertion within the Smn coding region (at amino acids 193 for and Kl 36 for ) that is predicted to introduce a premature stop codon. (Figure 4A). Unlike th ^o allele [15,18], which is 100% lethal in homo- and hemizygous (Smn ⁷ °/Df(Smn)) backgrounds (Figure 4B), the Smn allele produces a small percentage of escapers (3.3%) when mutant larvae are isolated and cultured at low density. On the other hand, Smn allele is semi-viable (67.7%) (Figure 4B), indicating that the and Smn ^m96 ° alleles are not null mutations as previously suggested [18]. By examining the viability of various Smn allelic combinations (Figure 4B), was determined to be weakly hypomorphic as it retains some degree of viability in all cases tested, while appears to act as a strong loss-of-function allele since it fails to complement both and a small deficiency that uncovers Smn, Figure 4B). The null allele of Smn,

Df(3L)SW , was generated by imprecise excision of the _J P{EPgy2}EY14384. 1,626 excision events were isolated and 17 failed to complement Smn ^73Ao , including Df(3L)Smn ^X7 . Subsequent sequence analysis of Df(3L)Smn ^X7 determined the excision event removed almost the entire SMN transcript without affecting nearby loci (93 bp upstream of the transcription start site through all but the final 44 bp of the 3' UTR). Ubiquitous (tubulinGAL4, actinGAL4) expression of UAS-FLAG-Smn rescued Smn ⁵ lethality, demonstrating the lethality was associated with a loss of Smn activity. This is consistent with earlier studies showing ectopic SMN expressed under the control of a ubiquitous driver {tubulinGAL4) rescued Smn ^13Ao lethality [15].

Loss of Smn causes neuromuscular junction defects

[0071] SMA patients experience motor neuron degeneration and muscle atrophy [1,4]. Consistent with this, previous work has shown that a loss of Smn function results in defects at the Drosophila NMJ [15]. To confirm and extend these results, the NMJ phenotype observed in various Smn genetic backgrounds was examined by quantitatively assessing the morphology of the NMJ through examination of synaptic bouton numbers between muscles 6 and 7 of the 3 ^rd instar larval NMJs. These boutons were visualized by using antibodies against the Synaptotagmin (SYT) (pre-synaptic) and DLG (postsynaptic) proteins, respectively (Figure 5A-G). The following Smn genotypes, which were capable of reaching the 3 ^r instar larval stages {Smn ^{1 0} I Smn , and therefore amenable to dissection, were examined.

[0072] The most severe reduction in NMJ bouton numbers was observed in a genetic background (Figures 5A-D and F). The semi-viable mutation displayed a moderate reduction in NMJ bouton numbers, consistent with its weakly hypomorphic nature (Figure 5C). Surprisingly, the strong loss of function mutation, though homozygous lethal, failed to exhibit a detectable change in NMJ bouton numbers in an background. However, an increase in pre-synaptic ghost bouton numbers [33,34] (where pre-synaptic SYT was not accompanied with postsynaptic DLG) was observed in these individuals (Figure 6), indicating that the Smn ^m596 ° allele does, indeed, disrupt NMJ morphology. The NMJ phenotype associated with

individuals was rescued partially by neuronal or muscle-directed expression of a UAS-FLAG-Smn transgene (Figure 5E-G), suggesting that SMN expression in either tissue is sufficient to restore, at least partially, NMJ morphology.

Loss of Smn function in muscles causes lethality

[0073] Though it is clear that global reduction of SMN function elicits a larval lethal phenotype (Figure 4B), the relative requirement of SMN in muscle versus neuron remains unresolved. To address this question directly, inducible Smn RNAi strains (N4, C24 and FL26B) were created, which can be expressed using tissue-specific GAL4 drivers. To generate the Smn RNAi constructs, three different portions of the Smn cDNA were cloned into the pWIZ vector: the entire cDNA (FL constructs), the amino-terminal portion up to and including the entire Tudor domain (N constructs) and the carboxy- terminal portion after, but not including, the Tudor domain (C constructs) (Figure 4A). These constructs were then introduced into w ¹¹¹⁸ embryos by germ-line transformation according to standard procedures. Multiple independent insertions were obtained for each construct, including the and transgenic strains that were used for this series of experiments. SMN expression was reduced in neuronal and muscle lineages using the pan-neuronal elavGAL4 [35] and pan-muscle how24BGAL4 drivers, respectively, to express the Smn RNAi constructs.

[0074] Reduction of SMN in either tissue causes lethality, however, loss of SMN expression in the muscle results in an earlier onset of lethality, which we consider to be a more severe phenotype (Figure 7B-C). In the strongest Smn RNAi allele, N4, muscle- specific SMN reduction results in 70% mortality (Figure 7B), while neuronal specific reduction results in 7% mortality (Figure 7C). As RNAi is less efficient in neurons, a GAL4-driven dicer construct was added to increase the efficacy of SMN reduction under these conditions [38]; this resulted in no obvious enhancement of lethality in all Smn RNAi and elavGAL4 backgrounds (data not shown). The GAL4 repressor GAL80 was expressed in neurons using the pan neuronal n-syb driver [39] to overcome the potential leakiness o/the how24BGAL4 driver. Since the lethality observed for muscle specific reduction of SMN more closely resembles ubiquitous SMN reduction (compare Figure

7A and B), the requirement of SMN in the muscle (using how24BGAL4) appears to be more important for viability than its requirement in the neurons.

Muscle and neuronal expression is required for normal NMJ morphology [0075] Similar to the tissue-dependent lethality experiments above, we sought to assess the impact SMN activity has on NMJ morphology using our UAS-Smn-RNAi strains, which can be expressed using tissue-specific GAL4 drivers. [0076] SMN expression was selectively reduced in neuron and muscle tissues by crossing the UAS-Smn-RNAi alleles to the elavGAL4 and how24BGAL4 drivers as they provide the earliest tissue specific expression and most robust lethal effect (Figure 7 and data not shown). Visualized by SYT (pre-synaptic) and DLG (post-synaptic) staining, NMJs of Smn RNAi animals containing either a muscle- or neuron-specific GAL4 driver revealed a reduction in the number of synaptic boutons compared to vector alone controls (Figure 8A-M). In the N4 strain, both neuron and muscle specific attenuation of SMN cause approximately 50% reduction in bouton numbers (Figure 8B, C, K-M), a reduction comparable to what is observed in Smn ^13Ao /Smn ^muω larvae (Figure 5D, H). Therefore, NMJ morphology is dependent upon both pre- and post-synaptic SMN activity. [0077] Previous studies demonstrated that mutations in Smn cause a decrease in staining for the post-synaptic neurotransmitter receptor subunit, GIuRIIA [15]. To corroborate these results and to extend the characterization of the tissue-specific requirement of SMN at the NMJ, the GIuRIIA [40,41,42] expression pattern was examined in the UAS-Smn- RNAi backgrounds. A consistent and significant quantitative reduction in synaptic GIuRIIA levels was observed when Smn expression was decreased using either neuron- (elavGAL4) or muscle-specific (mhcGAL4) drivers. GAL4-only controls had no significant effect on GIuRIIA staining intensity. Consistent with the trend observed for the severity of the lethal phenotype, the strongest Smn RNAi alleles caused the greatest reduction in GIuRIIA expression levels, suggesting that GIuRIIA levels are sensitive to the dose of functional SMN protein and thus, would be a useful phenotypic metric in which to validate potential modifiers of the Smn NMJ phenotype. [0078] These data indicate that normal NMJ morphology requires SMN activity in both muscle and neurons. However, it appears that loss of SMN activity in the muscle causes a

more severe lethal phenotype (Figure 7B), a conclusion that is consistent with the finding that the SMN protein is concentrated in the post-synaptic regions in muscle (Figure 2A- D).

Example 2. Identification of genetic modifiers oiSmn

[0079] To gain insights into the genetic circuitry capable of modulating SMN activity in vivo, a genetic approach was employed to screen for genes that affect 5Vn«-dependent processes using the Exelixis collection of transposon- induced mutations [25,26]. The benefits of using the collection in a genetic screen have been previously described [24]. Notably, the collection covers approximately 50% of the genome and harbors both gain- as well as loss-of- function mutations when exposed to GAL4 due to the presence of UAS sequences within the insertional transposons [25,26]. While the molecular coordinates of each insertion site is known, gene assignments are sometimes ambiguous, as the modifying transposon may have inserted between two genes. [0080] The screen was carried out in two stages to identify both enhancers and suppressors of 5m«-associated lethality (Figure 9). The strong correlation observed between the degree of lethality and NMJ phenotypes using the Smn RNAi lines suggested the use of lethality as a screening parameter would be successful in identifying components of the SMN genetic network that might also affect the NMJ. Both phases of the screen utilized the Smn ⁷ ° allele, which gives a robust NMJ defect, and importantly, contains a point mutation in the YG box (Figure 4A), which is the location of a documented human SMNl mutation [3].

[0081] The first stage was an Fl screen designed to identify insertions that produced synthetic lethality or semi-lethality in an Smn heterozygous background, which will hereafter be referred to as enhancers. The screen performed combined elements from standard Fl and F2 screens. This "combination screen" was identical to a standard F2 screen with the exception that the crosses were designed to identify synthetic lethal interactions with Smn in the Fl . In this screen th allele was utilized in cis with a ubiquitously expressed tubulinGAL4 driver (Lee and Luo, 1999). Initially, tubulinGAL4 e/TM6B virgin females were crossed to the entire Exelixis mutant collection to identify insertions that elicit Fl synthetic lethality or reduced viability. From

these results, a strain to be a candidate enhancer was defined as one that displayed a viability of less than 30%. Using this criterion, the entire Exelixis collection was screened and 17 insertions that result in + lethality were identified (see Table 1 below). [0082] In the second stage of the screen, the ability of mutations to suppress Smn- dependent larval lethality was tested. F 1 males from strains that failed to elicit synthetic lethality were crossed to e/TM 1 , Me virgins to test for their ability to suppress homozygous Smn ° larval lethality. In the F2 screen, candidate suppressors were identified by the presence of individuals bearing the marker ebony (e), which is visible in both pupae and adults. 7170 strains (as Smn ⁷ ° is located on the third chromosome, third chromosome insertions were excluded) were screened and ten suppressors of homozygous Smn ° lethality were identified (see Table 1 below). [0083] Listed in Table 1 are the insertions that enhance (top) or suppress (bottom) 5m« ^73Ao -dependent lethality. Due to the site of transposon insertion, unambiguous gene assignments were not possible in all instances (shaded). Strains whose designations begin with "d" or "f" contain GAL4 responsive elements (UAS), whereas strains beginning with "c" or "e" are not GAL4-inducible. Gene assignments were determined using FlyBase. Human homologs were determined using NCBI BLAST, NCBI UniGene (NCBI) or ENSEMBL genome browser. Annotated functions were determined based on FlyBase, NCBI Entrez Gene and SMART. Modification of the NMJ morphology between muscles 6 and 7 in the A2 segment was assayed in the elavGAL4 pWIZ[UAS- Smn-RNAi] background in trans with all identified modifiers using the pre-synaptic (Horseradish peroxidase) and post-synaptic (GIuRIIA) markers. In the three cases that did not show significant phenotypic alteration, additional allele was also used. The degrees of change observed in GIuRIIA staining were categorized as follows: +++, strong; ++, moderate; +, weak; N.E., No Effects.

[0084] To correlate modifier activity with the NMJ, we investigated whether all of the Smn modifiers (10 suppressors and 17 enhancers) could disrupt Smn RNAi-dependent NMJ defects, using synaptic GIuRIIA staining as an assay to quantify the degree to which the Smn phenotype was modified by the interacting mutation. For this assay, the C24 Smn RNAi line was employed because it displays intermediate phenotypic strength. In all but two cases, the combination of the modifier insertion mutation induced a statistically significant change in the C24 GIuRIIA phenotype (Table 1 and data not shown). Amongst the validated modifier insertions, the degree of enhancement or suppression varied depending on the locus; control crosses demonstrated that there were no significant 5Vn«-independent changes in GIuRIIA localization for the tested insertion lines. Three lines (fO4448, d09801 & d00698) failed to modify C24 GIuRIIA staining and were retested using a weaker Smn RNAi strain, N13 (Strain S04448 and d09801 enhanced, whereas d00698 showed no interaction (data not shown), highlighting the importance of the NMJ phenotype as a secondary screening tool (Table I)). Thus, the majority of the modifiers of the Smn ^13Ao lethal phenotype were confirmed by a second, independent assay. All but one of these insertions modified the Smn NMJ phenotype, validating the efficacy of the screen and suggesting that the screen is an effective tool in the identification of candidate genes that may be relevant to the SMA disease state. [0085] Though none of the unambiguously identified modifier genes have an obvious role in snRNP assembly, the canonical role for SMN; we did recover genes (wishful thinking, fmrl and cutup) that have been shown previously to function at the NMJ [27,28,63,64]. Moreover, the majority of the remaining genes, which had no previously known NMJ function, also modified Smn NMJ phenotypes. Thus our genetic approach was efficient in identifying genes related to Smn NMJ function. This suggests that a similar approach utilizing a hypomorphic Smn allele (e.g. UAS-Smn-RNAϊ) that more closely approximates the dosage dependent nature of the human disease condition may identify additional members of the Smn genetic circuitry (see Example 6). [0086] An analysis of the interacting loci according to molecular functions reveals an assortment of functional categories including cytoskeleton interaction proteins (moe and ctp), transcription factors (net) and metabolic enzymes (CGl 7323 and CG10561). Identified interactors also include members of several signal transduction pathways (e.g.

BMP (wit), FGF (btϊ) and Nuclear Hormone Receptor (Eip75E)), raising the possibility that these evolutionarily conserved signaling pathways integrate with SMN or targets of SMN function(s).

Example 3. SMN activity affects BMP signaling

[0087] This example describes experiments designed to further examine the relationship between wishful thinking (wit) and Smn in greater detail, wit was of particular interest because it has been previously implicated in NMJ function [27,28] and thus could serve as a paradigm for validating the ability of the screen to identify bonaβde Smn genetic modifiers.

[0088] wit encodes a type II bone morphogenetic protein (BMP) receptor that functions as a retrograde signaling component in neurons [27,28]. wit loss-of- function mutations cause NMJ defects, whereas wit gain-of-function causes no obvious NMJ morphological changes. As the wit allele identified as an Smn enhancer, wit ^d02492 , is associated with a GAL4-responsive transposon, it seemed likely that it represented a gain-of-function mutation. Consistent with this notion, an independent UAS-wit transgene [27,28] behaved in a similar fashion to wit ^d02492 under the conditions used in the screen (data not shown). In addition, increased expression of WIT was detected in wit ^d02492 animals containing tissue-specific GAL4 drivers (data not shown).

[0089] Over-expression of WIT in neurons using the neuron-specific elavGAL4 driver in either an Smn ¹ ° or an Smn heterozygous background resulted in reduced NMJ bouton numbers relative to elavGAL4 and UAS-wit; elavGAL4 controls (Figure 10). This result suggests that the Smn-dependent NMJ phenotype is sensitive to elevated WIT levels.

[0090] Given the involvement of wit at the NMJ and its interaction with Smn, we hypothesized that an Smn heterozygous background leads to an increase in sensitivity to the dosage of BMP during NMJ development. Thus, under conditions of elevated levels of WIT in Smn heterozygotes, it is possible that normal BMP signaling at the NMJ is altered, perhaps due to titration of the BMP ligand, thereby resulting in NMJ defects. If this hypothesis is correct, mutations of the BMP components downstream of wit should also enhance the Smn NMJ phenotype. Therefore, we tested whether Mothers against dpp

(Mad) and Smn interact at the NMJ. Mad encodes the Drosophila homolog of R-Smad, a downstream effector of the pathway [34,43,44]. Pathway activation leads to phosphorylation of MAD (pMAD), and its subsequent translocation to the nucleus where it regulates gene expression [34,43,44]. To examine the consequences of Smn/Mad interaction at the NMJ, the hypomorphic Mad allele [34] was used in combination with multiple Smn alleles to monitor the phenotypic effects at the NMJ. The moderate reduction in number of NMJ boutons caused by the hypomorphic M allele (Figure 1 ID and G) is clearly exacerbated by mutations in Smn (Figure 1 IE-G). These results suggest that perturbations in BMP signaling are able to modify Smn-dependent phenotypes at the larval NMJ.

[0091] To further validate the link between SMN and the BMP signaling pathway the effect of reduced SMN levels on pMAD expression was examined. Though Mad is required for retrograde signaling in neurons at the NMJ [34,45], a lack of detectable pMAD staining at the NMJ precluded the use of the NMJ as a means to assess whether SMN can affect its expression. Instead, the pMAD expression pattern adjacent to the anterior-posterior compartment boundary of 3 ^rd instar larval wing discs [46] (Figure 12) was examined using engrailedGAL4 and vestigalGAL4 directed expression of the N4 RNAi transgene (Figure 12 and Figure 13 respectively). Regions in which SMN levels are reduced display attenuated pMAD staining (Figure 12C-E). Moreover, adult wing abnormalities occur in regions of reduced SMN expression, including thicker wing veins and shorter posterior cross-veins (Figure 12F). These phenotypes are similar to phenotypes elicited by mutations in other BMP pathway components such as thickveins (tkv) and glass bottom boat (gbb) [45,47,48]. Thus, BMP signaling in the wing appears to be affected by loss of SMN activity through the regulation of activated Mad, corroborating the link between Smn and the BMP signaling pathway. [0092] Next, the relationship between the BMP pathway antagonist, Daughters against dpp (Dad), and Smn was examined. Dad encodes the Drosophila homolog of mammalian anti-Smad and acts as a Mad antagonist [44,49,50]. Since Dad mutants exhibit presynaptic overgrowth [49], we tested whether the null mutation could rescue the Smn NMJ phenotype. Consistent with previous reports [49], 3 ^rd instar larvae homozygous for display more dispersed SYT expression at the NMJ than control larvae

(Figure 14C). However, in contrast to previous studies, the total bouton number, as determined by DLG post-synaptic staining, was found to be only slightly reduced. Importantly, the NMJ phenotype was suppressed by the introduction of Dad ^271'68 (Figure 14B, D, E), providing genetic evidence that a third element of the BMP pathway interacts with Smn. It appears that elevating BMP activity through a complete loss of Dad function suppresses the effects of Smn mutations on the NMJ (Figure 14D, E). Thus, pharmacological reagents that increase BMP signaling may ameliorate Smn- associated NMJ defects, thereby identifying a set of targets of potential therapeutic value. [0093] SMN may act in the muscle to influence retrograde BMP signaling through the WIT receptor, for example by regulating the activity of the WIT ligand (GBB). It is also possible that SMN functions cell-autonomously in the neurons to affect the activity of MAD or its antagonist, DAD. As the BMP signaling pathway has been implicated in other neurodegenerative diseases, including Duchenne Dystrophy and Marfan Syndrome [65], it is probable that BMP signaling also plays a role in the pathology of SMA in humans.

Example 4. FGF signaling and Smn interaction at the neuromuscular junction

[0094] The FGF pathway is a hierarchical signaling cascade in which receptor tyrosine kinase activation leads to the regulation of target gene expression. In Drosophila, this pathway has been demonstrated to control the development of branching morphogenesis of the tracheal system [67] as well as the establishment of the mesoderm and its derivate, the musculature [68-75]. Pathway activation is mediated by the two known Drosophila FGF receptor orthologs, breathless (btl) and heartless (htl). btl, which functions in the tracheal system, is activated by its ligand branchless (bnl) [76], whereas htl, which functions in the mesoderm and muscles, is activated either by thisbe (ths) [77] or pyramus (pyr) [74]. Both receptors act through Sos-Grb2 to activate Ras/Raf/MAP kinase signaling. Additional regulation of Ras/Raf/MAP kinase signaling occurs through stumps [75], which regulates the phosphatase corkscrew (csw) [78]. In turn, Csw negatively regulates sprouty (sty), itself a negative regulator of Raf, thereby leading to MAPK activation [79].

[0095] The genetic screen (see Example 2) identified breathless (btl) as a modifier of 5m«-dependent lethality. As btl encodes one of the two known Drosophila FGF receptors, its recovery suggested a connection between the FGF pathway and Smn. Subsequent experiments demonstrated that the partial lethality associated with the ubiquitous expression of RNAi directed against SMN using the UAS-Smn-RNAi ^FL26B (tubGAL4: :FL26B) strain was rendered fully lethal by ubiquitous expression of constitutively active btl (Figure 15A). Moreover, this lethality was significantly enhanced by btl (Figure 15B), the allele identified in the screen. Approximately 30% of tubGAL4::FL26B animals die at an early pupal stage; this percentage is increased to 50% if these animals also carry the btl mutation. Subsequent experiments demonstrated btl altered an Smn RNAi-induced NMJ phenotype (data not shown). Moreover, loss of function mutations of stumps, an effector of FGF signals, or branchless (bnl), a ligand for the btl receptor, enhance tubGAL4::FL26B lethality, further demonstrating that SMN activity can be modulated by FGF signals (Figure 15C).

[0096] The link between FGF signaling and SMN was reinforced by the finding that ectopic expression of the second Drosophila FGF receptor, heartless (htl), rescues the NMJ phenotype associated with N4, the strongest of the Smn RNAi alleles (Figure 16). In addition, the expression of a dominant negative htl transgene using the how24B muscle driver elicits a strong NMJ phenotype (Figure 17), further implicating the FGF pathway in the development and/or maintenance of the Drosophila NMJ.

[0097] To further examine the epistatic relationship between FGF signaling and SMN at the 3rd larval instar NMJ, a set of extant FGF pathway mutations is used. The modulation of FGF activity in the NMJ using mutations in the pathway establishes possible FGF related NMJ phenotypes. Further analyses simultaneously modulating the activities of both loci (gain- and loss-of- function combinations) establishes the point of intersection between SMN and the FGF pathway in the NMJ. To avoid the embryonic lethality caused by classical FGF pathway mutations, the GAL4/UAS system [80] is used in combination with a temperature sensitive GAL80 (GAL80ts) construct (as necessary) [81] to alter the levels of FGF components in a tissue-specific fashion. Shifting the GAL80ts strain from a non-permissive to a permissive temperature activates GAL80ts, thereby antagonizing GAL4 function in a precise temporal manner. This allows for the determination of

specific spatial and temporal requirements of the genes under investigation. Thus, a set of mesodermal/muscle (how24BGAL4, MHCGAL4 and M12GAL4) and nervous system (elavGAL4, OK6GAL4 and OK309GAL4) drivers that express in an increasingly restricted set of cells within each tissue are examined.

[0098] The FGF pathway components whose expression levels are altered include the receptors {htl, btϊ), the ligands (bnl, thb) and a subset of the downstream effectors {stumps, csw, and sty). To reduce expression levels, the extant UAS-RNAi and UAS- dominant negative transgenic strains currently available is used. The consequences of overexpression of these genes and constitutive activation of both receptors in manner analogous to the experiments described above is examined.

[0099] Given the demonstrated effect of loss of htl function in the muscle on the larval NMJ (see Figure 17), expression levels of other pathway members is altered to determine whether the observed htl NMJ defects can be phenocopied. The muscle-specific drivers described above is used to direct expression of RNAi constructs to reduce levels of both receptors and selected cytoplasmic effectors {stumps and csw). In addition, a dominant negative btl allele and a full-length sty (an inhibitor) transgenic strain is used to further assess loss of FGF pathway function in muscle. Since loss of FGF signaling in the muscle leads to reduction of the NMJ and overexpression of htl in the muscle rescues Smn- dependent NMJ defects, it is likely that increasing this signaling may lead to a corresponding increase in the NMJ. Therefore, the consequences of FGF pathway activation is assessed by expressing a dominant negative version of the inhibitor, sty, which would be expected to lead to pathway activation and potentially cause synaptic overgrowth.

[00100] In another series of experiments, the role of FGF signaling in modulating SMN- dependent muscle loss is examined. In mouse models, FGF 6 has been shown to play an important role in the regeneration of experimentally damaged skeletal muscles [82]. FGF activity in muscle functions to maintain normal and adapt to new physiological homeostasis states, and in vitro, autocrine FGF signaling promotes survival of vascular smooth muscle cells (SMC) by a Ras-MAPK coupled pathway [83]. Different components of the FGF-Ras-MAPK pathway is used to determine whether a similar mechanism is active in Drosophila muscle survival under conditions in which SMN

function is reduced using RNAi or different allelic combinations This will provide a genetically tractable system of muscle deterioration/degeneration that is used to model muscle atrophy observed in SMA patients.

[00101] Overexpression of the FGF receptor HtI rescues most of the muscle defects associated with muscle-specific reduction of Smn (see Figure 16). It will be determined whether the canonical FGF-Ras-MAPK pathway is involved in ameliorating this muscle phenotype by directing expression of dominant negative and activated Ras (Ras^ ⁷ and , respectively) using muscle-specific GAL4 drivers and assaying for effects on the Smn muscle phenotype. As Ras is also a downstream effector of EGFR signaling, the possibility that Ras activation in muscles may result through activation of EGFR signaling is controlled by testing whether activated EGFR and dominant negative EGFR can also affect the Smn-dependent muscle deterioration/degeneration. MAPK activity is monitored using an antibody directed against phosphorylated MAP kinase (phospho ERK) in muscles in which Smn function has been decreased. Activated MAPK influences apoptosis through its transcriptional regulation of the pro-apoptotic gene, hid. Interestingly, loss-of-function for hid rescues Smn RNAi-induced lethality (data not shown).

Example 5. Functional modifiers of SMN associated with the cytoskeleton and protein translation.

[00102] The canonical cellular function of SMN is to participate as a member of the Gemin complex, which is involved in RNA metabolism and splicing. The ubiquitous, "house keeping" function of SMN is an apparent contradiction considering the rather specific neuromuscular phenotype that accompanies loss of SMN function in SMA. Remarkably, this phenotype seems to be conserved across species as the loss of SMN function in Drosophila also affects NMJ morphology. A series of studies raise the possibility the SMN may actually have a general neuronal or, potentially, a specific NMJ function in addition to its role in the Gemin complex (reviewed in [I]). For example, microtubule associated SMN containing granules have been observed in rat spinal

motorneurons [84]. These granules, which contain RNA and proteins (RNPs), display bidirectional movement and are speculated to reflect a specific neuronal role for SMN. Given that SMN is known to bind RNA and RNPs, the possibility that SMN may participate in a complex(es) that is distinct from the Gemin complex and is used in the neurons to transport mRNA [85, 86] and ribonucleoprotein particles [9, 87, 88] has been raised. Moreover, Bassel and Kelic [89] suggest that SMN may also participate in the localized translational apparatus that is thought to be important in neurites. Interestingly, using the screen of the Exelixis collection, we have identified SMN modifiers whose function has been associated with the cytoskeleton (moesin (moe) and Fimbrin (Fim)) and the translational machinery ((pumilio (pum), eIF-4E, Fmrl) .

Pumilio and eIF-4E

[00103] The Drosophila Pumilio (Pum) protein is an RNA binding protein that acts as a translational regulator during embryonic patterning and germ-line development [90]. Recent findings demonstrate that Pum also plays an important role at the NMJ [91, 92]. In neurons, it appears to function in the homeostatic control of excitability via down- regulation of paralytic (para) [93], a voltage-gated sodium channel. In addition it may more generally modulate local protein synthesis in neurons via translational repression of eIF-4E [91, 93]. Therefore, it is significant that both pumilio and eIF-4E were isolated as strong Smn modifiers.

[00104] Though pum was identified as a suppressor of Smn, the type of pum allele identified does not allow a priori determination whether the isolated mutations reflect gain or loss of pum function. Given that overexpression of full-length Pum results in an expansion of the NMJ [91], it is reasonable that a gain-of- function pum mutation was responsible for suppression of the Smn phenotype. To verify this potential interaction, a UAS-pum (full length) construct is used in combination with ubiquitous GAL4 drivers to determine whether increasing Pum expression levels in several backgrounds of reduced Smn function results in suppression of the Smn NMJ phenotype. In parallel, the impact of the strong hypomorphic pum ^ET1 and pum ^ET9 heteroallelic combination on the NMJ phenotypes in different Smn genetic backgrounds is assessed. In addition to investigating the epistatic relationships between Smn and pum, the tissue-specific requirements of the

two loci for the observed interactions is determined by using muscle and neuronal GAL4 drivers to vary the levels of either Pum (via UAS-pum) or SMN (via Smn RNAi and UAS- Smn-FLAG) and assaying the effect these combinations have on the NMJ. [00105] eIF-4E was also identified as a suppressor of Smn. As the eIF-4E message is one of the major targets of Pum, its expression and localization serves as an indicator of Pum activity. eIF-4E controls translation initiation by binding to the 5'-m7Gppp cap- structure of mRNA, thereby affecting the recruitment of mRNA to the ribosome. Schuster and colleagues have shown that post-synaptic aggregates of eIF-4E can be visualized in the larval NMJ [94]. Alteration of these aggregates in the absence of SMN would suggest that an interaction of SMN with the local translation machinery might underlie its role in neuromuscular development and function. These experiments investigate the connection between SMN, Pum and eIF-4E and may allow for the prediction of a specific role for SMN in motor neurons that may serve to solve the dichotomy between the ubiquitous housekeeping function of SMN and the very specific neuromuscular SMA phenotype.

Fragile X mental Retardation 1 (Fmrl) and its target futsch

[00106] The Fmrl (fragile X mental retardation 1) locus in humans encodes the fragile X mental retardation protein (FMRP), a protein that acts as a shuttle within cells by carrying molecules of messenger RNA (mRNA). FMRP is normally expressed in high abundance in many tissues including brain and has been implicated in localized translation [89]. The Drosophila FMRl ortholog {Fmrl) was identified as an SMN modifier (see Example 2 and Table 1) that affects NMJ phenotype. Reduced or increased Fmrl activity in Drosophila results in enlarged synaptic terminals and fewer and larger synaptic boutons, respectively, defects that are accompanied by altered neurotransmission. These phenotypes mimic those observed in mutants of futsch, which encodes the Drosophila ortholog of the microtubule-associated protein, MaplB. Fmrl was shown to act as a translational repressor of futsch regulating microtubule-dependent synaptic growth and function [95]. Fmrl is thought to be a part of the motoneuron translational apparatus and may potentially link Smn activity to this machinery. [00107] Fmrl was identified as an enhancer of loss of Smn function phenotypes, however, a priori determination of whether this allele is a loss- or gain-of- function

mutation was not possible. In a first set of experiments, the nature of the interaction between these two loci is established using an array of loss- and gain-of function alleles to assay for changes in NMJ morphology as measured by the number of boutons per synapse. As loss of Fmrl results in an expansion of the NMJ and overexpression leads to reduced synaptic size [95], it is likely that the observed interaction was due to overexpression of Fmrl . Therefore, a UAS-Fmrl (full length) construct is used in combination with ubiquitous GAL4 drivers to determine whether increasing Fmrl function in several backgrounds of reduced Smn function results in enhancement of the Smn NMJ phenotype. The impact of a strong hypomorphic Fmrl allele on the NMJ phenotypes in different hypomorphic Smn genetic backgrounds is also assessed. To extend these studies, the tissue-specific requirements of the two loci for the observed interactions is determined in a similar manner as that described for pum by varying the levels of either Fmrl (via UAS-Fmrl) or Smn (via Smn RNAi and UAS-Smn-FLAG) at the NMJ. Immunohistochemistry and Western blot analyses are used to assess the effects of the various mutant combinations on the expression and localization of SMN and Fmrl as well as the expression and localization of Futsch, the MAPlB ortholog, which was also isolated as a suppressor of Smn. The futsch message is one of the major Fmrl targets [95] and thus serves to mark its activity. Fmrl associates with futsch mRNA, and in Fmrl null mutants, Futsch levels in the nervous system are increased [95]. Neuronal directed overexpression of Fmrl results in a corresponding reduction in Futsch expression, indicating that Fmrl negatively regulates/«tec/z. Consequently, Fmrl levels inversely regulate the level of Futsch in the nervous system.

[00108] Recent evidence suggests a strong association between the SMN complex and FMRP in cultured hypothalamus slices [96]. Thus, taken together our genetic data combined with the finding that SMN and FMRP are present in the same protein complex provide strong evidence of a functional interaction between these two genes. It is a plausible hypothesis that SMN, through its association with this complex, is also regulating Futsch. Futsch is required for microtubule assembly and plays an important role in NMJ morphology in Drosophila [97]. Thus, if reduced levels of SMN directly regulate Futsch levels and are responsible for the NMJ phenotype, overexpression of futsch in motorneurons (elavGAL4::UAS-futsch) can rescue the NMJ phenotype caused

by loss of Smn. This interaction is validated by reducing Futsch levels using the amorphic futsch ⁿ⁵⁸ allele in a genetic background with reduced levels of SMN.

Links between SMN and the cvtoskeleton

[00109] Moesin (Moe) encodes an ERM family protein and localizes to the apical region of the plasma membrane where it physically interacts with the actin cytoskeleton to maintain epithelial integrity [98]. Apical actin decreases in a loss-of- function Moe mutant background, whereas constitutively-active Moe mutations (Moe ^{τ55 D} ) upregulate cortical F-actin levels in epithelial cells [99]. Reinforcing the notion that filamentous actin organization relies on Moe function, overexpression of Rhol GTPase, a regulator of Moe activity, phenocopies the Moe loss-of- function phenotype in the wing imaginal disc [99]. At the NMJ, loss of Moe elicits a synaptic overgrowth phenotype [100], indicating that Moe also modulates the localization of actin filaments at the growth cone. [00110] Decreased Moe activity suppresses Smn-dependent lethality and NMJ defects (see Example 2 and Table 1). To extend this finding, the NMJ morphology derived from loss- (Moe ^G0323 ) and gain-(Moe ^T559D ) of-function Moe alleles [99] in backgrounds with reduced Smn function is examined. To determine whether the ability of Moe to suppress the Smn phenotypes is tissue specific, a GAL4-responsive RNAi allele (UAS-Moe-RNAϊ) to reduce Moe transcript levels selectively in neurons and muscles is used. The cellular localization of endogenous Moe in relation to SMN at the NMJ is monitored by immunohistochemical analysis using anti-Moe [99] and anti-SMN antibodies. The observed suppression of Moe loss-of- function on loss of SMN function phenotypes may be associated with effects on actin filament organization at the NMJ. To test this hypothesis, the NMJ cytoskeleton is examined (using antibodies that recognize actin filaments) in a hypomorphic Smn background that contains either gain or loss of function Moe (Moe ^G0323 and Moe ^T559D ) mutations.

[00111] To increase understanding of the genetic relationship between Moe and Smn, a molecular genetic approach is employed to assess the ability of Moe and two of its regulators, Rhol and Sterile20-like kinase (slik) [99, 101], to affect Smn expression in the Drosophila wing imaginal disc where the relationship between Moe and these regulators has been examined. Moe, Rhol and slik transcript levels are reduced in the wing

epithelium by introducing GAL4-responsive RNAi alleles of each gene [102] into backgrounds carrying either the engrailedGAL4 or vestigialGAL4 drivers. SMN expression is monitored using antibodies generated against Drosophila SMN to determine whether mutations in Moe, Rhol and/or slik affect SMN expression. Once the relationships between Smn, Moe and its regulators are established in the wing epithelium, the results are corroborated in the larval NMJ. Given that loss of Moe function suppresses the NMJ phenotypes caused by loss of Smn, alterations in Rhol or slik activity are examined to determine an effect on the Smn«-dependent NMJ phenotype. As GAL4 directed expression of UAS-Rhol phenocopies Moe loss-of-function phenotypes in the wing disc, Rhol overexpression is also likely to suppress the Smn-dependent phenotype. Whether neuronal or muscle overexpression of Rhol can rescue NMJ phenotypes caused by reduction of Smn is tested by employing muscle (how24BGAL4 and MHCGAL4) and neuronal (elavGAL4) specific drivers in a reduced Smn background. In addition, as Slik phosphorylates Moe and activates Moe function, hypomorphic slik mutations may also suppress the NMJ abnormalities associated with Smn.

Interaction between Fimbrin and Smn at the NMJ

[00112] Plastin 3 is an actin filament bundling protein that has been implicated in the maintenance of cell shape and cell polarity [103]. A recent study identified a subset of human female subjects, who despite having lost both copies of SMNl , do not display severe signs of SMA as they express increased levels of Plastin 3 [104]. This result suggests that Plastin 3 can act as a suppressor of SMA [104]. Consistent with this finding, axonal defects caused by the lack of SMN in culture neurons and in zebrafish motoneurons were rescued by increasing Plastin 3 levels [104]. Importantly, this suppression caused by Plastin 3 upregulation is associated with increased levels of F- actin [103]. Although Plastin 3 is found in a complex with SMN, they have not been shown to directly bind to one another [104]. Thus, the direct/physical links between SMN and Plastin 3 remain unknown.

[00113] We identified a loss-of-function mutation of the Drosophila ortholog of Plastin 3, Fimbrin (Fim), as an enhancer of 5m«-linked lethality. To further explore the relationship between Smn and Fim, full length and mutated (inactive) forms of Fim are

cloned into FLAG-HA-tagged GAL4 responsive vectors to allow tissue specific expression of these transgenes. Neuronal and muscle specific drivers (see above) are used to ectopically express full length Fim in Smn mutant backgrounds to determine whether Fim functions in a tissue specific manner to alleviate the Smn NMJ defects. Morphological changes of the NMJ are monitored using pre- and post-synaptic markers in addition to actin filament markers. As a negative control, a mutated form of Fim lacking the phosphorylation site (Ser5) essential for F-actin bundling [105] is expressed. As Moe also functions as an actin binding protein, Moe, Rhol or slik mutations [99, 101] are introduced into a Fim background to determine their combined effects on the structure of the Drosophila NMJ. These studies allow further characterization of the genetic relationship between Fim and Smn.

Example 6. Constructing RNAi-based hypomorphic Smn alleles [00114] Since the clinical severity of SMA correlates with the amount of SMN expression, we sought to better model the disease by generating a set of Smn alleles with varying degrees of SMN activity using RNAi. A GAL4-inducible vector was used to produce three different double-stranded RNAi transgenic constructs targeted against the full-length SMN protein (FL) as well as the amino -terminal (N) (the entire 5' portion of the protein up to and including the Tudor domain) and carboxy-terminal (C) (the 3 ' portion of the protein after, but not including, the Tudor domain) SMN regions (Figure 4A).

[00115] Ten independent transgenic strains for each type of construct (C, N and FL) were generated and examined for their effects on lethality when SMN activity was reduced or eliminated using either tubulinGAL4 or actinGAL4, two ubiquitous GAL4 drivers. It was difficult to differentiate between the lethal phases of many strains in the tubulinGAL4 background, presumably due to its higher levels of expression. Instead, we were able to use the timing of lethality in the presence of actinGAL4 to choose three lines (fUAS-Smn-RNAif ⁴ (N4), [UAS-Smn-RNAif ²⁴ (C24) and [UAS-Smn-RNAif ^L26B (FL26B)) that define a set of alleles representing the broadest range of detectable lethality for further analysis (Figure 7A).

[00116] Of all strains generated, N4 displayed the most severe phenotype, causing mortality at the early pupal stage. C24 was less severe and resulted in lethality at a later pupal stage than N4, while FL26B was semi- viable and was therefore the weakest allele of the three (Figure 7A). Under the control of the tubulinGAL4 driver, N4 caused a similar phenotype to those observed for the Smn ¹ ° and Smn ⁵ mutations, suggesting that N4 is a strong hypomorphic Smn allele (data not shown). The efficiency of RNAi in the N4 and C24 strains precluded us from testing whether ectopic SMN expression could rescue the RNAi-induced lethality. However, we do note that the fully penetrant pupal lethality induced by the expression of tubulinGAL4-directed FL26B is completely rescued by the addition of the UAS-FLAG-Smn construct to this genetic background (data not shown).

[00117] Consistent with these results, examination of protein derived from 3rd instar larvae from the above strains in the presence of the actinGAL4 driver revealed significant reductions in SMN expression levels (Figure 7D), further suggesting the observed lethality is the direct result of SMN protein attenuation. Although the three strains did not display apparent differences in the degree of reduction of SMN under these conditions, the genetic results with respect to viability and subsequent experiments investigating NMJ morphology (see Example 1) strongly suggest these RNAi-induced Smn strains result in varying degrees of SMN activity and therefore, alleles of different strengths. Importantly, these hypomorphic Smn alleles provide a Drosophila model system that more closely mimics the human spinal muscular atrophy disease condition.

A pilot screen for dominant modifiers of Smn RNAi-induced pupal lethality [00118] Given the previous success in screening for genetic modifiers of Smn activity (see Example 2), a genetic screen was designed to take advantage of the Smn RNAi strains' mimicry of the dosage sensitive nature of SMA seen in humans to identify additional members of the Smn genetic network. The wide range of lethal phases displayed by these strains suggested that an appropriate combination of driver and RNAi strain could be selected that would render the modification assay significantly more sensitive than that used in the previous screen (see Example 2). The FL26B strain, in combination with the tubGAL4 driver (tubGAL4: :FL26B), results in a fully penetrant,

pupal lethality that can be used to assess qualitatively and quantitatively potential dominant genetic interactions.

[00119] Qualitatively, the presence of adults or the absence of pupae can be used as indicators of either suppression or enhancement, respectively. Control crosses of tubGAL4::FL26B to a deficiency for Smn resulted in the expected enhancement of lethality as very few pupae appear in comparison to sibling controls (see Table 2 below). In contrast, crossing tubGAL4::FL26B to a UAS-Smn-FLAG construct resulted in suppression of lethality as all pupae reached adulthood. The quantitative nature of this assay results from the non-uniform lethal phase displayed by tubGAL4::FL26B pupae. During the course of normal pupal development, which lasts 120 hours at 25°C, the pupa begins to become pigmented at approximately 96 hours. Thus, the appearance of pigmentation can be used as an indicator of the stage of development a pupa has reached and therefore may be used to determine a shift in the lethal phase of a population of pupae. In control crosses to the isogenic background of the Exelixis collection, approximately 40% of dead \ubGAL4 : :FL26B pupae reach the stage at which they become pigmented while the remaining 60% die beforehand (see Table 2). In test crosses, most of the previously identified modifiers exhibited a shift in the relative levels of the two populations in the expected manner (see Table 2 for two examples: a suppressor, CG3136 and an enhancer, CG 1927), thereby demonstrating the ability to detect modification by measuring the shift in lethal phase between populations.

Table 2. Modification of Smn RNAi elicited pupal lethality.

[00120] With these experiments serving as proof of principle, we conducted a pilot screen of the Exelixis collection of transposon insertions using the hypomorphic Smn allele (tubGAL4::FL26B) in a manner similar to that described in Example 2 for the original screen. Screening 3200 lines (20% of the collection) yielded 400 lines to be retested. Of these 400, 150 (4.7%) passed the retest. This rate of recovery is similar to those observed for other screens of the Exelixis collection and suggests a screen of the entire collection should yield approximately 750 modifiers. To assess the biological content recovered by the pilot screen, we grouped those insertions that could be assigned to a single gene into several functional categories, including BMP signaling and RNA processing. No genes associated with the known function of SMN, RNA processing, were identified as modifiers in our previous screen. In contrast, 18 of the 150 (12%) modifiers recovered in the pilot screen fall into that category, demonstrating a significant increase in the sensitivity of our RNAi-based assay. In addition, the overlap of functional categories recovered in both screens {e.g. BMP signaling) confirms the RNAi-based screen is capable of identifying processes demonstrated to affect the NMJ in an Smn- dependent fashion. We isolated several molecules that fall in the same genetic circuitry as molecules identified in our first screen. This includes/«tec/z, which is a target of Fmrl, one of the modifiers identified in the first screen (see Table 1). We also isolated new modifiers, pumilio and eIF-4E, which have been previously demonstrated to play a bona fide role in determining the structure and function of the NMJ [91, 93]. Thus, the RNAi- based hypomorphic strains provide a sensitive model system that more closely resembles the Smn gene dosage found in spinal muscular atrophy. Such a system is useful for identifying genetic modifiers that can be novel therapeutic targets as well as for testing candidate compounds for therapeutic efficacy of the spinal muscular atrophy disease test. [00121] It is understood that the disclosed invention is not limited to the particular methodology, protocols and materials described as these can vary. It is also understood

that the terminology used herein is for the purposes of describing particular embodiments only and is not intended to limit the scope of the present invention which will be limited only by the appended claims.

[00122] Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.

REFERENCES

1. Monani UR (2005) Spinal muscular atrophy: a deficiency in a ubiquitous protein; a motor neuron-specific disease. Neuron 48: 885-896.

2. Frugier T, Nicole S, Cifuentes-Diaz C, Melki J (2002) The molecular bases of spinal muscular atrophy. Curr Opin Genet Dev 12: 294-298.

3. Wirth B (2000) An update of the mutation spectrum of the survival motor neuron gene

(SMNl) in autosomal recessive spinal muscular atrophy (SMA). Hum Mutat 15:

228-237.

4. Crawford TO, Pardo CA (1996) The neurobiology of childhood spinal muscular atrophy. Neurobiol Dis 3: 97-110.

5. Lefebvre S, Burglen L, Reboullet S, Clermont O, Burlet P, et al. (1995) Identification and characterization of a spinal muscular atrophy-determining gene. Cell 80: 155- 165.

6. Lefebvre S, Burlet P, Liu Q, Bertrandy S, Clermont O, et al. (1997) Correlation between severity and SMN protein level in spinal muscular atrophy. Nat Genet 16: 265-269.

7. Lorson CL, Hahnen E, Androphy EJ, Wirth B (1999) A single nucleotide in the SMN gene regulates splicing and is responsible for spinal muscular atrophy. Proc Natl Acad Sci U S A 96: 6307-6311.

8. Monani UR, Lorson CL, Parsons DW, Prior TW, Androphy EJ, et al. (1999) A single nucleotide difference that alters splicing patterns distinguishes the SMA gene SMNl from the copy gene SMN2. Hum MoI Genet 8: 1177-1183.

9. Liu Q, Dreyfuss G (1996) A novel nuclear structure containing the survival of motor neurons protein. Embo J 15: 3555-3565.

10. Massenet S, Pellizzoni L, Paushkin S, Mattaj IW, Dreyfuss G (2002) The SMN complex is associated with snRNPs throughout their cytoplasmic assembly pathway. MoI Cell Biol 22: 6533-6541.

11. Wan L, Battle DJ, Yong J, Gubitz AK, KoIb SJ, et al. (2005) The survival of motor neurons protein determines the capacity for snRNP assembly: biochemical deficiency in spinal muscular atrophy. MoI Cell Biol 25: 5543-5551.

12. Paushkin S, Gubitz AK, Massenet S, Dreyfuss G (2002) The SMN complex, an assemblyosome of ribonucleoproteins. Curr Opin Cell Biol 14: 305-312.

13. Eggert C, Chari A, Laggerbauer B, Fischer U (2006) Spinal muscular atrophy: the

RNP connection. Trends MoI Med 12: 113-121.

14. Meister G, Eggert C, Fischer U (2002) SMN-mediated assembly of RNPs: a complex story. Trends Cell Biol 12: 472-478.

15. Chan YB, Miguel- Aliaga I, Franks C, Thomas N, Trulzsch B, et al. (2003)

Neuromuscular defects in a Drosophila survival motor neuron gene mutant. Hum MoI Genet 12: 1367-1376.

16. McWhorter ML, Monani UR, Burghes AH, Beattie CE (2003) Knockdown of the survival motor neuron (Smn) protein in zebrafish causes defects in motor axon outgrowth and pathfinding. J Cell Biol 162: 919-931.

17. Murray LM, Comley LH, Thomson D, Parkinson N, Talbot K, et al. (2008) Selective vulnerability of motor neurons and dissociation of pre- and post-synaptic

pathology at the neuromuscular junction in mouse models of spinal muscular atrophy. Hum MoI Genet 17: 949-962.

18. Rajendra TK, Gonsalvez GB, Walker MP, Shpargel KB, SaIz HK, et al. (2007) A

Drosophila melanogaster model of spinal muscular atrophy reveals a function for SMN in striated muscle. J Cell Biol 176: 831-841.

19. Oprea GE, Krober S, McWhorter ML, Rossoll W, Muller S, et al. (2008) Plastin 3 is a protective modifier of autosomal recessive spinal muscular atrophy. Science 320:

524-527.

20. Rossoll W, Jablonka S, Andreassi C, Kroning AK, Karle K, et al. (2003) 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: 801-812.

21. McAndrew PE, Parsons DW, Simard LR, Rochette C, Ray PN, et al. (1997)

Identification of proximal spinal muscular atrophy carriers and patients by analysis of SMNT and SMNC gene copy number. Am J Hum Genet 60: 1411- 1422.

22. Feldkotter M, Schwarzer V, Wirth R, Wienker TF, Wirth B (2002) Quantitative analyses of SMNl and SMN2 based on real-time lightCycler PCR: fast and highly reliable carrier testing and prediction of severity of spinal muscular atrophy. Am J Hum Genet 70: 358-368.

23. Miguel- Aliaga I, Chan YB, Davies KE, van den Heuvel M (2000) Disruption of SMN function by ectopic expression of the human SMN gene in Drosophila. FEBS Lett 486: 99-102.

24. Kankel MW, Hurlbut GD, Upadhyay G, Yajnik V, Yedvobnick B, et al. (2007)

Investigating the genetic circuitry of mastermind in Drosophila, a notch signal effector. Genetics 177: 2493-2505.

25. Parks AL, Cook KR, Belvin M, Dompe NA, Fawcett R, et al. (2004) Systematic generation of high-resolution deletion coverage of the Drosophila melanogaster genome. Nat Genet 36: 288-292.

26. Thibault ST, Singer MA, Miyazaki WY, Milash B, Dompe NA, et al. (2004) A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac. Nat Genet 36: 283-287.

27. Aberle H, Haghighi AP, Fetter RD, McCabe BD, Magalhaes TR, et al. (2002) wishful thinking encodes a BMP type II receptor that regulates synaptic growth in Drosophila. Neuron 33: 545-558.

28. Marques G, Bao H, Haerry TE, Shimell MJ, Duchek P, et al. (2002) The Drosophila

BMP type II receptor Wishful Thinking regulates neuromuscular synapse morphology and function. Neuron 33: 529-543.

29. Zito K, Fetter RD, Goodman CS, Isacoff EY (1997) Synaptic clustering of Fascilin II and Shaker: essential targeting sequences and role of Dig. Neuron 19: 1007-1016.

30. Liu JL, Murphy C, Buszczak M, Clatterbuck S, Goodman R, et al. (2006) The

Drosophila melanogaster Cajal body. J Cell Biol 172: 875-884.

31. Littleton JT, Bellen HJ, Perin MS (1993) Expression of synaptotagmin in Drosophila reveals transport and localization of synaptic vesicles to the synapse. Development 118: 1077-1088.

32. Littleton JT, Stern M, Schulze K, Perin M, Bellen HJ (1993) Mutational analysis of

Drosophila synaptotagmin demonstrates its essential role in Ca(2+)-activated neurotransmitter release. Cell 74: 1125-1134.

33. Ataman B, Ashley J, Gorczyca D, Gorczyca M, Mathew D, et al. (2006) Nuclear trafficking of Drosophila Frizzled-2 during synapse development requires the PDZ protein dGRIP. Proc Natl Acad Sci U S A 103: 7841-7846.

34. Eaton BA, Davis GW (2005) LIM Kinase 1 controls synaptic stability downstream of the type II BMP receptor. Neuron 47: 695-708.

35. Robinow S, White K (1988) The locus elav of Drosophila melanogaster is expressed in neurons at all developmental stages. Dev Biol 126: 294-303.

36. Brand AH, Perrimon N (1993) Targeted gene expression as a means of altering cell fates and generating dominant pheno types. Development 118: 401-415.

37. Michelson AM (1994) Muscle pattern diversification in Drosophila is determined by the autonomous function of homeotic genes in the embryonic mesoderm. Development 120: 755-768.

38. Dietzl G, Chen D, Schnorrer F, Su KC, Barinova Y, et al. (2007) A genome-wide transgenic RNAi library for conditional gene inactivation in Drosophila. Nature 448: 151-156.

39. DiAntonio A, Burgess RW, Chin AC, Deitcher DL, Scheller RH, et al. (1993)

Identification and characterization of Drosophila genes for synaptic vesicle proteins. J Neurosci 13: 4924-4935.

40. DiAntonio A, Petersen SA, Heckmann M, Goodman CS (1999) Glutamate receptor expression regulates quantal size and quantal content at the Drosophila neuromuscular junction. J Neurosci 19: 3023-3032.

41. Elias GM, Nicoll RA (2007) Synaptic trafficking of glutamate receptors by MAGUK scaffolding proteins. Trends Cell Biol 17: 343-352.

42. Lahey T, Gorczyca M, Jia XX, Budnik V (1994) The Drosophila tumor suppressor gene dig is required for normal synaptic bouton structure. Neuron 13: 823-835.

43. Inoue H, Imamura T, Ishidou Y, Takase M, Udagawa Y, et al. (1998) Interplay of signal mediators of decapentaplegic (Dpp): molecular characterization of mothers against dpp, Medea, and daughters against dpp. MoI Biol Cell 9: 2145-2156.

44. Patterson GI, Padgett RW (2000) TGF beta-related pathways. Roles in

Caenorhabditis elegans development. Trends Genet 16: 27-33.

45. McCabe BD, Marques G, Haghighi AP, Fetter RD, Crotty ML, et al. (2003) The

BMP homolog Gbb provides a retrograde signal that regulates synaptic growth at the Drosophila neuromuscular junction. Neuron 39: 241-254.

46. Tanimoto H, Itoh S, ten Dijke P, Tabata T (2000) Hedgehog creates a gradient of

DPP activity in Drosophila wing imaginal discs. MoI Cell 5: 59-71.

47. Haerry TE, Khalsa O, O'Connor MB, Wharton KA (1998) Synergistic signaling by two BMP ligands through the SAX and TKV receptors controls wing growth and patterning in Drosophila. Development 125: 3977-3987.

48. Khalsa O, Yoon JW, Torres-Schumann S, Wharton KA (1998) TGF-beta/BMP superfamily members, Gbb-60A and Dpp, cooperate to provide pattern

information and establish cell identity in the Drosophila wing. Development 125:

2723-2734.

49. Sweeney ST, Davis GW (2002) Unrestricted synaptic growth in spinster-a late endosomal protein implicated in TGF-beta-mediated synaptic growth regulation. Neuron 36: 403-416.

50. Tsuneizumi K, Nakayama T, Kamoshida Y, Kornberg TB, Christian JL, et al. (1997)

Daughters against dpp modulates dpp organizing activity in Drosophila wing development. Nature 389: 627-631.

51. Battle DJ, Kasim M, Yong J, Lotti F, Lau CK, et al. (2006) The SMN complex: an assembly machine for RNPs. Cold Spring Harb Symp Quant Biol 71: 313-320.

52. Zhang Z, Lotti F, Dittmar K, Younis I, Wan L, et al. (2008) SMN deficiency causes tissue-specific perturbations in the repertoire of snRNAs and widespread defects in splicing. Cell 133: 585-600.

53. Gabanella F, Butchbach ME, Saieva L, Carissimi C, Burghes AH, et al. (2007)

Ribonucleoprotein assembly defects correlate with spinal muscular atrophy severity and preferentially affect a subset of spliceosomal snRNPs. PLoS ONE 2: e921.

54. Kariya S, Park GH, Maeno-Hikichi Y, Leykekhman O, Lutz C, et al. (2008) Reduced

SMN protein impairs maturation of the neuromuscular junctions in mouse models of spinal muscular atrophy. Hum MoI Genet.

55. Lee S, Sayin A, Grice S, Burdett H, Baban D, et al. (2008) Genome -wide expression analysis of a spinal muscular atrophy model: towards discovery of new drug targets. PLoS ONE 3: el404.

56. Gavrilina TO, McGovern VL, Workman E, Crawford TO, Gogliotti RG, et al. (2008)

Neuronal SMN expression corrects spinal muscular atrophy in severe SMA mice while muscle-specific SMN expression has no phenotypic effect. Hum MoI Genet 17: 1063-1075.

57. Hsieh-Li HM, Chang JG, Jong YJ, Wu MH, Wang NM, et al. (2000) A mouse model for spinal muscular atrophy. Nat Genet 24: 66-70.

58. Jablonka S, Karle K, Sandner B, Andreassi C, von Au K, et al. (2006) Distinct and overlapping alterations in motor and sensory neurons in a mouse model of spinal muscular atrophy. Hum MoI Genet 15: 511-518.

59. Monani UR, Sendtner M, Coovert DD, Parsons DW, Andreassi C, et al. (2000) The human centromeric survival motor neuron gene (SMN2) rescues embryonic lethality in Smn(-/-) mice and results in a mouse with spinal muscular atrophy. Hum MoI Genet 9: 333-339.

60. Avila AM, Burnett BG, Taye AA, Gabanella F, Knight MA, et al. (2007) Trichostatin

A increases SMN expression and survival in a mouse model of spinal muscular atrophy. J Clin Invest 117: 659-671.

61. Hua Y, Vickers TA, Okunola HL, Bennett CF, Krainer AR (2008) Antisense masking of an hnRNP A1/A2 intronic splicing silencer corrects SMN2 splicing in transgenic mice. Am J Hum Genet 82: 834-848.

62. Simon V, Ho DD, Abdool Karim Q (2006) HIV/AIDS epidemiology, pathogenesis, prevention, and treatment. Lancet 368: 489-504.

63. Pan L, Zhang YQ, Woodruff E, Broadie K (2004) The Drosophila fragile X gene negatively regulates neuronal elaboration and synaptic differentiation. Curr Biol 14: 1863-1870.

64. Sachdev P, Menon S, Kastner DB, Chuang JZ, Yeh TY, et al. (2007) G protein beta gamma subunit interaction with the dynein light-chain component Tctex-1 regulates neurite outgrowth. Embo J 26: 2621-2632.

65. Cohn RD, van Erp C, Habashi JP, Soleimani AA, Klein EC, et al. (2007) Angiotensin

II type 1 receptor blockade attenuates TGF -beta-induced failure of muscle regeneration in multiple myopathic states. Nat Med 13: 204-210.

66. Johnson KG, Tenney AP, Ghose A, Duckworth AM, Higashi ME, et al. (2006) The

HSPGs Syndecan and Dallylike bind the receptor phosphatase LAR and exert distinct effects on synaptic development. Neuron 49: 517-531.

67. Ghabrial, A., et al. (2003) Branching morphogenesis of the Drosophila tracheal system. Annu Rev Cell Dev Biol. 19: 623-47.

68. Shishido, E., et al. (1997) Requirements of DFRl /Heartless, a mesoderm-specific

Drosophila FGF receptor, for the formation of heart, visceral and somatic muscles, and ensheathing of longitudinal axon tracts in CNS. Development 124(11): 2119-28.

69. Beiman, M., B.Z. Shilo, and T. VoIk (1996) Heartless, a Drosophila FGF receptor homolog, is essential for cell migration and establishment of several mesodermal lineages. Genes Dev. 10(23): 2993-3002.

70. Gisselbrecht, S., et al. (1996) heartless encodes a fibroblast growth factor receptor

(DFR1/DFGF-R2) involved in the directional migration of early mesodermal cells in the Drosophila embryo. Genes Dev. 10(23): 3003-17.

71. Michelson, A.M., et al. (1998) Dual functions of the heartless fibroblast growth factor receptor in development of the Drosophila embryonic mesoderm. Dev Genet. 22(3): 212-29.

72. Schulz, RA. and K. Gajewski (1999) Ventral neuroblasts and the heartless FGF receptor are required for muscle founder cell specification in Drosophila. Oncogene 18(48): 6818-23.

73. Shishido, E., et al. (1993) Two FGF-receptor homologues of Drosophila: one is expressed in mesodermal primordium in early embryos. Development 117(2): 751-61.

74. Stathopoulos, A., et al. (2004) pyramus and thisbe: FGF genes that pattern the mesoderm of Drosophila embryos. Genes Dev. 18(6): 687-99.

75. Vincent, S., et al. (1998) The Drosophila protein Dof is specifically required for FGF signaling. MoI Cell. 2(4): 515-25.

76 . Sutherland, D., C. Samakovlis, and M.A. Krasnow (1996) branchless encodes a

Drosophila FGF homolog that controls tracheal cell migration and the pattern of branching. Cell 87(6): 1091-101.

77. Kerr, D.A., et al. (2003) Human embryonic germ cell derivatives facilitate motor recovery of rats with diffuse motor neuron injury. J Neurosci. 23(12): 5131-40.

78. Petit, V., et al. (2004) Downstream-of-FGFR is a fibroblast growth factor-specific scaffolding protein and recruits Corkscrew upon receptor activation. MoI Cell Biol. 24(9): 3769-81.

79. Jarvis, L.A., et al. (2006) Sprouty proteins are in vivo targets of Corkscrew/SHP-2 tyrosine phosphatases. Development 133(6): 1133-42.

80. Brand, A.H. and N. Perrimon (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118(2): 401-15.

81. Lee, T. and L. Luo (1999) Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morphogenesis. Neuron 22(3): 451-61.

82. Floss, T., H.H. Arnold, and T. Braun (1997) A role for FGF-6 in skeletal muscle regeneration. Genes Dev. 11(16): 2040-51.

83. Miyamoto, T., et al. (1998) Autocrine FGF signaling is required for vascular smooth muscle cell survival in vitro. J Cell Physiol. 177(1): 58-67.

84. Zhang, H.L., et al. (2003) Active transport of the survival motor neuron protein and the role of exon-7 in cytoplasmic localization. J Neurosci. 23(16): 6627-37.

85. Bertrandy, S., et al. (1999) The RNA-binding properties of SMN: deletion analysis of the zebrafish orthologue defines domains conserved in evolution. Hum MoI Genet. 8(5): 775-82.

86. Lorson, CL. and EJ. Androphy (1998) The domain encoded by exon 2 of the survival motor neuron protein mediates nucleic acid binding. Hum MoI Genet. 7(8): 1269-75.

87. Jones, K.W., et al. (2001) Direct interaction of the spinal muscular atrophy disease protein SMN with the small nucleolar RNA-associated protein fibrillarin. J Biol Chem. 276(42): 38645-51.

88. Mourelatos, Z., et al. (2001) SMN interacts with a novel family of hnRNP and spliceosomal proteins. Embo J. 20(19): 5443-52.

89. Bassell, GJ. and S. Ke lie (2004) Binding proteins for mRNA localization and local translation, and their dysfunction in genetic neurological disease. Curr Opin Neurobiol. 14(5): 574-81.

90. Asaoka-Taguchi, M., et al. (1999) Maternal Pumilio acts together with Nanos in germline development in Drosophila embryos. Nat Cell Biol. 1(7): 431-7.

91. Menon, K.P., et al. (2004) The translational repressor Pumilio regulates presynaptic morphology and controls postsynaptic accumulation of translation factor eIF-4E. Neuron 44(4): 663-76.

92. Chen, G., et al. (2008) Identification of synaptic targets of Drosophila pumilio. PLoS

Comput Biol. 4(2): el000026.

93. Muraro, N.I., et al. (2008) Pumilio binds para mRNA and requires Nanos and Brat to regulate sodium current in Drosophila motoneurons. J Neurosci. 28(9): 2099-109.

94. Sigrist, S.J., et al. (2000) Postsynaptic translation affects the efficacy and morphology of neuromuscular junctions. Nature 405(6790): 1062-5.

95. Zhang, Y.Q., et al. (2001) Drosophila fragile X-related gene regulates the MAPlB homolog Futsch to control synaptic structure and function. Cell 107(5): 591-603.

96. Piazzon, N., et al. (2008) In vitro and in cellulo evidences for association of the survival of motor neuron complex with the fragile X mental retardation protein. J Biol Chem. 283(9): 5598-610.

97. Roos, J., et al. (2000) Drosophila Futsch regulates synaptic microtubule organization and is necessary for synaptic growth. Neuron 26(2): 371-82.

98. Bretscher, A., K. Edwards, and R.G. Fehon (2002) ERM proteins and merlin: integrators at the cell cortex. Nat Rev MoI Cell Biol. 3(8): 586-99.

99. Speck, O., et al. (2003) Moesin functions antagonistically to the Rho pathway to maintain epithelial integrity. Nature 421(6918): 83-7.

100. Seabrooke, S. and B.A. Stewart (2008) Moesin helps to restrain synaptic growth at the Drosophila neuromuscular junction. Dev Neurobiol. 68(3): 379-91.

101. Hughes, S. C. and R.G. Fehon (2006) Phosphorylation and activity of the tumor suppressor Merlin and the ERM protein Moesin are coordinately regulated by the Slik kinase. J Cell Biol. 175(2): 305-13.

102. Dietzl, G., et al. (2007) A genome -wide transgenic RNAi library for conditional gene inactivation in Drosophila. Nature 448(7150): 151-6.

103. Matsudaira, P. (1994) The fimbrin and alpha-actinin footprint on actin. J Cell Biol.

126(2): 285-7.

104. Oprea, G.E., et al. (2008) Plastin 3 is a protective modifier of autosomal recessive spinal muscular atrophy. Science 320(5875): 524-7.

105. Janji, B., et al. (2006) Phosphorylation on Ser5 increases the F-actin-binding activity of L-plastin and promotes its targeting to sites of actin assembly in cells. J Cell Sci. 119(Pt 9): 1947-60.