Physiol. Genomics Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 24: 97-104, 2006. First published August 23, 2005; doi:10.1152/physiolgenomics.00134.2005
1094-8341/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables and Figures
Right arrow All Versions of this Article:
24/2/97    most recent
00134.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Olaso, R.
Right arrow Articles by Melki, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Olaso, R.
Right arrow Articles by Melki, J.
Received 8 June 2005; accepted in final form 9 August 2005.
Physiological Genomics 24:97-104 (2006)
1094-8341/05 $8.00 © 2006 American Physiological Society

Activation of RNA metabolism-related genes in mouse but not human tissues deficient in SMN

Robert Olaso1, Vandana Joshi1, Julien Fernandez1, Natacha Roblot1, Sabrina Courageot1, Jean Paul Bonnefont2 and Judith Melki1

1 Molecular Neurogenetics Laboratory, Institut National de la Santé et de la Recherche Médicale (INSERM), E-223, University of Evry, Genopole, Evry
2 INSERM 393, Hôpital Necker-Enfants Malades, Paris, France


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 DISCLOSURES
 REFERENCES
 
Mutations of the survival of motor neuron gene (SMN1) are responsible for spinal muscular atrophies (SMA), a frequent recessive autosomal motor neuron disease. SMN is involved in various processes including RNA metabolism. However, the molecular pathway linking marked deficiency of SMN to SMA phenotype remains unclear. Homozygous deletion of murine Smn exon 7 directed to neurons or skeletal muscle causes severe motor axonal or myofiber degeneration, respectively. With the use of cDNA microarrays, expression profiles of 8,400 genes were analyzed in skeletal muscle and spinal cord of muscular and neuronal mutants, respectively, and compared with age-matched controls. A high proportion of genes (20 of 429, 5%) was involved in pre-mRNA splicing, ribosomal RNA processing, or RNA decay, and 18 of them were upregulated in mutant tissues. By analyzing other neuromuscular disorders, we showed that most of them (14 of 18) were specific to the SMN defect. Quantitative PCR analysis of these transcripts showed that gene activation was an early adaptive response to the lack but not reduced amount of full-length SMN in mouse mutant tissues. In human SMA tissues, activation of this program was not observed, which could be ascribed to the reduction but not the absence of full-length SMN.

survival motor neuron; gene expression profile; mouse models; human spinal muscular atrophy


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 DISCLOSURES
 REFERENCES
 
SPINAL MUSCULAR ATROPHIES (SMA) are characterized by degeneration of lower motor neurons leading to progressive limb and trunk paralysis associated with muscular atrophy. SMA is a frequent recessive autosomal disorder caused by mutations of the survival of motor neuron gene (SMN1, incidence 1 of 5,000–10,000 newborns; see Refs. 15 and 7 for review). SMN is a ubiquitously expressed protein of 294 amino acids with a molecular weight of 38 kDa. SMN forms a large multiprotein complex of ~1 MDa both in the cytoplasm and in the nucleus where it is concentrated in a structure called gems (for "gemini of coiled bodies"), most often associated with or identical to Cajal bodies (coiled bodies, Ref. 17). The interaction of SMN with components of this large complex is enhanced by SMN oligomerization. The identification of SMN-interacting proteins of known function strongly supports the view that SMN is involved and facilitates cytoplasmic assembly of small nuclear ribonucleoprotein particle (snRNP) into the spliceosome, a large RNA-protein complex that catalyzes the splicing reaction (26). In the nucleus, SMN appears to be directly involved in pre-mRNA splicing, transcription, and metabolism of ribosomal RNA (14, 22, 23). SMN has also been shown to interact directly with other proteins including Bcl-2 and p53, two proteins involved in apoptotic processes, although these interactions and the involvement of an apoptotic process in SMA remain to be clarified in vivo (6, 13, 27). Other direct or indirect partners of SMN include the far upstream element (FUSE)-binding protein, profilin II, a zinc finger protein called ZRP1, RNA helicase A, RNA polymerase II, RNA, and mSin3A (2, 9, 10, 20, 21, 25, 29). Therefore, SMN appears to be a multifunctional protein.

A mouse line carrying two loxP sequences flanking Smn exon 7 (SmnF7) has been established through homologous recombination. Cre-mediated deletion of Smn exon 7 (Smn{Delta}7) has been directed to either neurons or skeletal muscle ("neuronal" and "muscular" mutant, respectively; Refs. 4, 5, 8). Neuronal mutant mice develop a severe motor defect leading to complete paralysis and death at a mean age of 1 mo (8). Analysis of skeletal muscle revealed severe muscle denervation associated with a dramatic and progressive loss of motor axons contrasting with mild and late loss of motor neuron cell bodies. In addition, abnormal synaptic terminals filled with neurofilaments associated with defective axonal sprouting were observed at the neuromuscular junctions of mutant mice (5). SMN has been studied in other organisms. Reduced levels of SMN in zebrafish or point mutations of SMN in Drosophila cause defects in motor axon pathfinding or neuromuscular junctions, respectively (3, 19). Therefore, deficiency of SMN in various organisms has underlined an essential role of SMN for motor axon and neuromuscular junction development or maintenance. Surprisingly, homozygous deletion of Smn exon 7 directed to murine skeletal muscle only ("muscular mutant") led to a severe myopathic phenotype characterized by myofiber necrosis leading to muscle paralysis and death at a mean age of 1 mo (4). These data indicated that full-length SMN is essential for cell viability and showed that SMN{Delta}7 is not able to compensate for lack of full-length SMN.

DNA chip technology is a powerful tool allowing the analysis of changes in expression of several thousand genes in parallel and the characterization of complex transcriptional activities. To determine downstream gene expression changes resulting from the Smn gene defect, gene expression profiles were analyzed in skeletal muscle and spinal cord of muscular and neuronal mutants, respectively, and compared with age-matched controls. Comparison of gene expression profiles revealed an overexpression of RNA metabolism-related genes in both skeletal muscle and spinal cord of mutant mice but not in human SMA tissues.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 DISCLOSURES
 REFERENCES
 
Animals.
Muscular (HSA-Cre, Smn{Delta}7/F7) or neuronal (NSE-Cre, SmnF7/F7) mutant mice were generated as previously described (4, 8). Mice were maintained on C57BL/6J genetic background. Smn+/F7 animals from the same litter were used as controls. Transgenic mice expressing G93A mutation within the superoxide dismutase 1, cytosolic (SOD1) gene (11) and mdx mice were purchased from Jackson Laboratory and Charles River Laboratories (Arbresle, France), respectively. Animals were genotyped by PCR amplification of DNA extracted from tail biopsies (8, 11). All animal procedures were performed with approval from the Veterinary Offices of the Essonne Department and from the Research Department of the Ministry of National Education, Research and Technology (agreement A91-228-2 and 3429).

Human tissues.
Spinal cord (n = 3) or skeletal muscle (n = 3) from fetuses predicted for having SMA was collected. Spinal cord (n = 5) or skeletal muscle (n = 7) from fetuses of the same age and affected by a defect unrelated to neuromuscular disorders was used as controls. Prenatal diagnosis of these diseases was performed by medical geneticists without any connection to this study. Samples were collected after abortion of fetuses predicted for having these diseases and after the informed and systematic consent of parents. Skeletal muscle of human SMA patients (n = 6) or subjects affected by diseases unrelated to SMA (n = 5) was collected in the frame of muscle biopsies after the informed consent of patients or parents. SMA patients or fetuses carried homozygous deletion of SMN1 exon 7 (data not shown). All tissue samples were available from the Research Tissues Bank, a Center of Biological Resources located at the Pitie-Salpetriere Hospital. Samples were obtained from surgical operation (surgery scrap) or from autopsy materials from fetuses (with systematic consent of parents) following the laws and the rules. Tissues have been harvested by an individual who had no connection to the study itself.

RNA extraction, cDNA labeling, and array hybridization.
Tissues were dissected and stored at –80°C. Total RNA was extracted using Trizol reagent (Invitrogen) according to the manufacturer's instructions and purified using RNeasy columns (Qiagen). RNA concentration and integrity were assessed by absorbance at 260 and 280 nm and gel electrophoresis. Microarray experiments were performed on mouse samples. cDNA was synthesized from 20 µg of RNA and labeled with either cyanine 3 (Cy3) or cyanine 5 (Cy5; Perkin Elmer) using a direct labeling cDNA synthesis kit (Agilent). Cy5-labeled mutant cDNA was then mixed with Cy3-labeled control cDNA, and the mixture was hybridized to mouse microarrays containing ~8,400 cDNA probes (G4104A, Agilent) for 17 h at 65°C. The slides were then washed and dried, and images were read using an Agilent DNA microarray scanner (G2565AA). The microarray data are available at the EMBL-European Bioinformatics Institute (EBI) ArrayExpress web site (http://www.ebi.ac.uk/arrayexpress/query/entry) under accession number E-MEXP-131.

Data analysis.
For the analysis of microarray experiments, ratio data were extracted from scanned microarray images using Agilent G2567AA Feature Extraction software (v7.5) and processed using Luminator software (Rosetta). A population of control spots was present on each slide and was used to calculate population statistics of spot intensities and background region, using a 99% level of confidence. Normalization was conducted using the software. To assess the significance of the log ratio (level of intensity in Cy5-labeled sample relative to Cy3-labeled sample), and therefore the confidence in the gene's differential expression, the log ratio error and P value were calculated for all features based on an error model. The significance of the feature intensity was calculated using two-sided Student's t-test. A detailed description of algorithms that were utilized can be found at http://www.chem.agilent.com/scripts/literaturePDF.asp?iWHID=37629.

To determine the reliability of microarray data, dye swap experiments were performed using liver and heart samples of wild-type mouse. cDNA from mouse liver and heart was synthesized and labeled with Cy3 and Cy5, respectively, and vice versa and hybridized following the protocol described above. The relative ratio of each transcript was expressed as log10 [heart (Cy5): Liver (Cy3)] and compared with that of –log10 [liver (Cy5): heart (Cy3)]. A total of 3,384 genes were selected based on P value < 0.02. A high correlation coefficient (98.5%) was observed between the two experiments (Supplemental Material S1; available at the Physiological Genomics web site).1

To determine the most reliable criteria for two-dimensional cluster and further molecular investigation, correlation between microarray data and real-time quantitative PCR amplification of reverse transcripts was evaluated. For the correlation analysis, the same RNA samples used for microarray experiments were reverse transcribed, and real-time PCR amplification of transcripts was performed. Aldolase was used as the internal control, since no significant difference was observed between control and mutant samples (skeletal muscle or spinal cord; data not shown). A total of 45 microarray data corresponding to 23 genes were examined. Two-fold expression changes in mutant-to-control ratio associated with high signal homogeneity (P value < 0.02) resulted in 82% correlation between microarray and real-time PCR amplification data. All inconsistent data (8 of 45) were coming from negative mutant-to-control ratio found in microarray data (data not shown; {chi}2, P = 0.003). To reduce the proportion of false negatives, more stringent criteria were applied. Microarray data were selected when positive fold change was greater than or equal to 2 and negative change was less than or equal to –3. These criteria resulted in 89% correlation between microarray and real-time PCR amplification data. Data were visualized and analyzed using Luminator software (Rosetta). Two-dimensional clusters were generated using the agglomerative hierarchical clustering algorithm (average link). To update the slide annotation of expressed sequence tags (ESTs; which represented 40% of cDNA loaded on the microarray slide), a search for sequence homology was carried out from gene accession number. Software written with python 2.3 (BioPython module) was elaborated to compare nucleotide sequence with those of the National Center for Biotechnology Information (NCBI) database, using blastN (ftp://ftp.ncbi.nlm.nih.gov/blast/executables/release/2.2.8/netblast-2.2.8-ia32-linux.tar.gz). Sequences with homology >80% were then validated manually, leading to 18% of unknown sequences. The same procedure was applied to the accession number of known genes loaded onto slides and selected in our study. Among 238 known genes, names were updated for 11 sequences.

Real-time quantitative PCR amplification analysis.
cDNA was synthesized from total RNA using an RT reaction. Briefly, 5 µg of total RNA were mixed to oligo(dT) (50 µM), dATP, dGTP, dCTP, and dTTP (1 mM each). After 5 min of incubation at 65°C, 20 µl of 5x first-strand buffer (Invitrogen) and DTT (10 mM), 5 µl of RNase out (Invitrogen), and 3 µl of Superscript II (Invitrogen) were added to the reaction. Synthesis was achieved by incubating the mix at 42°C for 50 min. Then, the reaction was inactivated by heating at 70°C for 15 min. The cDNAs were stored at –20°C before being used as template for quantitative PCR. Oligonucleotides were designed using primer express software (Applied Biosystems). Primer sequences of mouse and human genes are shown in Supplemental Material S2. Real-time quantitative PCR was carried out in 96-well plates, using the SYBR Green master mix (Applied Biosystems) according to manufacturer's instructions and processed on an ABI prism 7000 (Applied Biosystems). The thermal conditions used were 2 min at 50°C and 10 min at 95°C, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. For each set of primers, optimal primer concentration was first determined, and then a standard curve using increased dilution of cDNA was established in control tissues. Each reaction was performed at least in duplicate to evaluate the relative level of gene expression in tissue samples. Transcript levels were normalized to that of aldolase, since no difference in aldolase transcript level was observed between control and SMN-deficient tissues (3 mice in each group). ß-Actin was selected as internal control, since no difference was observed between human control and SMA tissues. All values are means ± SE. Student's t-test was used in statistical analyses.

Immunoblotting experiments.
Skeletal muscle of 1-mo-old control (SmnF7/+) and muscular mutant (HSA-Cre, SmnF7/{Delta}7) mice were rapidly frozen and crushed in liquid nitrogen using a mortar and pestle. Pulverized tissue samples were transferred into 5 vol of a buffer containing 25 mM sodium phosphate (pH 7.2), 5 mM EDTA, and 1% SDS supplemented with protease inhibitor cocktail (Sigma, St Louis, MO) and boiled for 5 min. Protein sample (50 µg) was mixed with an equal volume of 2x Laemmli buffer, electrophoresed, and then transferred. For FHL2 and p21 immunodetection, anti-FHL2 (1/1,000; a gift from Dr. Roland Schüle) and anti-p21 (1/1,000; BD Biosciences, San Jose, CA) monoclonal antibodies were used. For actin immunodetection, anti-actin monoclonal antibodies were used (1:10,000; Chemicon International, Temecula, CA). After washes in PBS and 0.05% Tween 20, membranes were incubated with horseradish peroxidase-conjugated secondary antibody (1/5,000 to 1/10,000; Jackson Laboratories, West Grove, PA), and the immune complexes were revealed using chemiluminescent detection reagents (Pierce, Rockford, IL).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 DISCLOSURES
 REFERENCES
 
Activation of a high proportion of RNA metabolism-related genes in mutant mice.
Gene expression profiles were determined in mouse tissues. The gastrocnemius muscle of control (SmnF7/+) and muscular mutants (HSA-Cre, SmnF7{Delta}7) and the spinal cord of control (SmnF7/+) and neuronal mutants (NSE-Cre, SmnF7/F7) were analyzed. Twenty- or thirty-day-old mutant and control mice were used for this study. These ages were chosen based on the phenotypic characteristics of both mutants. At 20 days of age, neuronal or muscular mutants did not display any motor defect despite a Cre recombinase activity in >80% of motor neurons or muscle fibers, respectively (4, 5, 8). At 30 days of age, both mutants display motor defects associated with marked histological changes in targeted tissues (4, 5, 8). Microarray analysis was performed on at least three animals of each group. To exclude physiological noise, only genes found to share regulation in at least two of three mutant mice (or 3 of 4) at each age (20 or 30 days old) and in each tissue (skeletal muscle or spinal cord) were retained.

On the basis of these criteria, a set of 429 genes that displayed expression changes in either mutant skeletal muscle or spinal cord was selected (Supplemental Materials S3 and S4). Among them, 255 genes were significantly regulated in skeletal muscle of 30-day-old mutant mice and 100 genes in spinal cord of 30-day-old neuronal mutants. The number of genes expressed in response to the SMN defect increased from 20 to 30 days of age in both muscular (3.5-fold, 70 to 255 genes) and neuronal mutants (1.5-fold, 66 to 100 genes), which indicated progression of molecular defects during the disease course (Supplemental Material S4).

To characterize the molecular pathways involved in response to the SMN defect, the regulated genes were subdivided into groups according to the Gene Ontology (GO) classification system (http://www.informatics.jax.org/userdocs/go_help.shtml). For a given functional class, the proportion of genes regulated in mutant skeletal muscle was compared with that of mutant spinal cord (Supplemental Material S5). No significant difference in the pattern of functional classes was found during disease progression in either skeletal muscle or spinal cord (data not shown). For this reason, classification was applied from genes regulated at either 20 or 30 days of age (Supplemental Material S5). The proportion of genes differentially expressed in neuronal or muscular mutants was similar in several functional classes including cytoskeleton, metabolism, and transcription. However, markedly different proportions of genes involved in inflammation, muscle contraction, or RNA metabolism were observed between mutant skeletal muscle and spinal cord (Supplemental Material S5). The proportion of genes involved in inflammation was higher in mutant spinal cord (14%) than in mutant skeletal muscle (2%). The proportion of genes involved in muscle contraction was markedly lower in mutant spinal cord (1%) than in mutant skeletal muscle (5%), a feature consistent with the expression pattern of these genes restricted to skeletal muscle (Supplemental Materials S4 and S5).

Interestingly, a high proportion of genes (5%, 20 of 429) were involved in pre-mRNA splicing, ribosomal RNA processing, or RNA decay (Table 1 and Supplemental Material S4). Eighteen of them were upregulated and two downregulated (Supplemental Material S4). The proportion of genes involved in RNA metabolism was higher in mutant skeletal muscle (7%) than in mutant spinal cord (2%, Supplemental Material S5). The low proportion of motor neurons with respect to the other cell types in the spinal cord could lead to underestimation of the number of overexpressed genes using a fold change greater than or equal to two. In agreement with this, 10 of 16 genes that were overexpressed in skeletal muscle were also found to be upregulated in spinal cord when using less-stringent criteria (fold change ≥1.3, Fig. 1). Real-time PCR analysis of transcripts encoding galectin-3 (Lgals3), embryonic lethal abnormal vision (ELAV)-like protein 1 (Elavl1), Ewing sarcoma homolog (Ewsr1), and pre-mRNA branch site protein p14 (SAP14) showed an expression of these genes in both spinal cord and skeletal muscle of wild-type mice and confirmed their overexpression in mutant mice, in agreement with the microarray data (Fig. 1 and Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. List of genes involved in RNA metabolism and regulated in Smn mutant mice

 


View larger version (30K):
[in this window]
[in a new window]
  		Fig. 1. Classification of overexpressed RNA metabolism-related genes in Smn mutant mice. A: according to the expression changes of the genes in amyotrophic lateral sclerosis (ALS) and Smn mutant mice, genes were classified as nonspecific (left) or specific to survivor motor neuron (SMN) defect (right). Expression changes are given as means of mutant-to-control ratios (ordinate, %) from microarray data. B: expression of RNA metabolism-related genes in Smn mutant mice as determined by quantitative PCR analysis. Transcript levels of 4 RNA metabolism-related genes were determined in skeletal muscle and spinal cord of 30-day-old control (n = 2–4, open bars) and Smn mutant mice (n = 2–4, solid bars). The expression level is given as the ratio of the target gene to aldolase (arbitrary units). *P < 0.05.

 
With SMN having a role in RNA metabolism in vitro, we determined whether the expression changes of RNA metabolism-related genes were specific to SMN defect. To this end, a gene expression profile was determined in spinal cord of transgenic mice overexpressing the SOD1-G93A gene, a mouse model of amyotrophic lateral sclerosis (ALS) (11). Fourteen of eighteen RNA metabolism-related genes that were upregulated in the Smn mutants did not display any changes in ALS, indicating that they were specific to the SMN defect (Figs. 1 and 2). With the use of real-time PCR analysis, transcript levels of a subset of these genes were evaluated in spinal cord of affected ALS transgenic mice (n = 3, 4 mo old) and confirmed the microarray data (Fig. 2). These results allowed classifying the genes involved in RNA metabolism into SMN-specific and nonspecific groups (Fig. 1).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 2. Expression analysis of RNA metabolism-related genes in mouse models of other neuromuscular disorders. By using real-time RT-PCR analysis, transcript levels were quantified in skeletal muscle of mdx (n = 3, gray bars) and control mice (n = 3, open bars) and in spinal cord of ALS (n = 3, solid bars) and control litter mice (n = 3, open bars). Expression level is given as the ratio of the target gene to aldolase of the same litter (arbitrary units). *P < 0.05.

 
Within the SMN-specific group of genes, eight were overexpressed in mutant skeletal muscle but not spinal cord. To know whether these genes could be activated in response to myofiber degeneration in a nonspecific manner, quantitative analysis of transcripts encoding Elavl1, Ewsr1, and SAP14 was performed in skeletal muscle of mdx mice (n = 3, 4 mo old), a mouse model of Duchenne muscular dystrophy (Fig. 2). Transcript levels of these genes were similar in mdx and control mice. In contrast, Lgals3, which was upregulated in both Smn mutant and ALS transgenic mice, was also upregulated in mdx, indicating a nonspecific response to tissue damage (Fig. 2). Therefore, a total of 14 overexpressed genes involved in RNA metabolism were specific to the SMN defect (Fig. 1). They encode proteins having a role in pre-mRNA splicing, ribosomal RNA processing, or RNA decay (Table 1).

Expression of the RNA metabolism-related genes was studied during the disease course of mutant skeletal muscle. At 9 days of age, the full-length SMN transcripts remain present, although at a low level, in muscular mutants (4), and transcript levels of RNA metabolism-related genes showed no significant difference compared with wild-type skeletal muscle (Table 2). At 20 or 30 days of age, the full-length SMN transcripts are no longer detected (4). Dramatically increased levels of RNA metabolism-related gene transcripts were observed showing a tight correlation between lack of full-length SMN transcripts and upregulation of RNA metabolism-related genes (Table 2). These data indicate that activation of the RNA metabolism-related genes was an early adaptive event in response to the lack of SMN, a feature in agreement with a role of SMN in RNA metabolism in vivo.


View this table:
[in this window]
[in a new window]
  		Table 2. Expression changes of RNA metabolism-related genes during mouse disease course

 
Lack of activation of RNA metabolism-related genes in human SMA tissues.
To determine whether activation of RNA metabolism-related genes found in Smn mutant mice occurred in human SMA disease, real-time quantitative PCR amplification analysis of orthologous transcripts was performed in human SMA tissue samples including skeletal muscle and spinal cord. Skeletal muscle (n = 3) and spinal cord samples (n = 3) collected from 10- to 19-wk-old fetuses predicted for having SMA linked to SMN1 mutations were analyzed. Skeletal muscle (n = 7) and spinal cord samples (n = 5) from fetuses having a defect unrelated to neuromuscular disorders were used as controls. ß-Actin was chosen as internal control, since no difference of expression level was observed between control and SMA samples (data not shown). Seven orthologous genes involved in the RNA metabolism were not activated in SMA compared with control tissues (Fig. 3 and Table 1). To determine whether activation of these genes might occur at a later stage of the disease course, skeletal muscles were collected from affected patients (n = 6) or controls (n = 5) and analyzed. Quantitative analysis of transcript levels of the same genes did not reveal any change in SMA compared with control tissues (Fig. 3 and Table 1). In addition, the Brunol 3 gene, which was recently shown to be overexpressed in muscle cells from a single SMA patient, was found to have an expression unchanged or even downregulated in human skeletal muscle or spinal cord tissues at either fetal or postnatal stages (1) (Fig. 3). These data indicate that activation of several RNA metabolism-related genes did not occur in human SMA.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 3. Expression analysis of RNA metabolism-related genes in human SMA tissues. Real-time RT-PCR amplification analysis of RNA metabolism-related genes was performed in adult skeletal muscle (A), fetal skeletal muscle (B), and fetal spinal cord (C). In each group, SMA tissues (solid columns) were analyzed and compared with control (open columns). The no. of samples tested for each gene is as follows: human adult skeletal muscle control, n = 5; SMA adult skeletal muscle, n = 6; fetal control skeletal muscle, n = 7; fetal SMA skeletal muscle, n = 3; fetal control spinal cord, n = 5; fetal SMA spinal cord, n = 3. Expression level is shown as the ratio of the target gene to ß-actin (arbitrary units). *P < 0.05. nd, Not detectable.

 
To determine whether the discrepancies between human and mouse transcriptional programs involved other pathways, some genes unrelated to RNA metabolism were investigated. They included the cyclin-dependent kinase inhibitor 1A (p21), four and a half LIM domains 2 (FHL2), and stathmin-like protein 4 (STMN4), which were deregulated in mutant mice (Supplemental Material S4 and Fig. 4). They encode proteins involved in various molecular pathways including cell cycle withdrawal (12), cell differentiation (18), or microtubule dynamics (24). Expression changes of these genes were confirmed by the analysis of either transcripts or proteins or both in mouse mutant tissues (Fig. 4). In human SMA tissues, quantitative PCR analysis of orthologous transcripts revealed similar changes (Fig. 4).



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 4. Expression analysis of biomarkers in mouse and human tissues. A: real-time RT-PCR amplification analysis of selected genes was performed on spinal cord and skeletal muscle of control and mutant mice (at least 3 animals in each group). Quantitative analysis of orthologous transcripts was performed in spinal cord and skeletal muscle of human control fetuses (n = 4 and n = 5, respectively) and SMA fetuses (n = 3 in each group). Expression level is given as the ratio of the target gene to aldolase (in mouse) or to ß-actin (in human, arbitrary units). Expression level of each gene was analyzed in control (open bars) and mutant tissues (solid bars). Values are means ± SE. *P < 0.05. B: immunoblotting experiments of proteins extracted from skeletal muscle of 30-day-old control (lanes 1–3) and Smn muscular mutant mice (lanes 4–6). Note the higher amount of FHL2 and p21 in mutant tissues compared with controls and actin expression. Controls and mutants of lanes 1 and 4, lanes 2 and 5, and lanes 3 and 6 belong to the same litter.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 DISCLOSURES
 REFERENCES
 
This study was able to define the pattern of transcriptional changes resulting from the SMN defect only. Indeed, skeletal muscle was collected from the muscular mutant in which the Smn exon 7 deletion was directed to skeletal muscle and not to neurons (4), and from spinal cord of the neuronal mutant carrying the same mutation in neurons but not in skeletal muscle (8). Therefore, our approach avoided collecting transcriptional adaptations of skeletal muscle to motor neuron degeneration or adaptations of spinal cord to skeletal muscle necrosis.

Of special interest was the observation of an activation of genes involved in RNA metabolism. They represent a high proportion of regulated genes in Smn mutant mice (5%). Excluding those found in ALS or mdx mice, which represent nonspecific responses, 14 genes were upregulated in response to SMN defect. Quantitative RT-PCR analysis confirmed the microarray data and showed that activation of this pathway was an early event during disease course. In vitro experiments or cell culture systems have shown that SMN plays an essential role in RNA metabolism (20, 21, 23, 26). In Smn mutant mice, the overexpression of genes involved in this process and/or interacting with SMN (such as Ewing sarcoma homolog, 28) might represent an adaptive response of RNA processing machinery to the lack of a component involved in the same process. However, activation of this pathway is not sufficient to prevent myofiber or neuron degeneration. These results provide, therefore, the first indirect evidence for a role of SMN in RNA metabolism in vivo.

Importantly, no overexpression of orthologous RNA metabolism-related genes specific to SMN defect was found in skeletal muscle or spinal cord of human SMA samples at either fetal or postnatal stages. How do we explain theses discrepancies between human and mouse tissues mutated for SMN? In skeletal muscle of 9-day-old mutant mice, marked reduction but not complete absence of full-length SMN transcripts is observed without any significant changes of the RNA metabolism-related genes. At later stages (20 or 30 days of age), Cre-mediated deletion of Smn exon 7 is completed, and full-length Smn transcripts are no longer detected. Then, an overexpression of RNA metabolism-related genes is observed. These data strongly suggest that this transcriptional program is activated in response to the lack but not marked reduction of full-length SMN in vivo. In human SMA, SMN2, a highly homologous copy of the SMN1 gene, remains present, encoding a low level of full-length SMN transcripts that are translated into a functional protein (16). We can hypothesize that full-length SMN transcripts and proteins, at a low level, are likely sufficient for the RNA processing, while its complete lack would be deleterious, leading to an activation of the RNA processing machinery that could be regarded as a compensatory mechanism. Anderson et al. (1) have shown an upregulation of Brunol 3, an RNA-binding protein, in mutant mice or myoblast culture from an SMA patient. The analysis of Brunol 3 transcript level in either fetal or postnatal skeletal muscle or spinal cord from human SMA patients was not able to reproduce these results. Altogether, these results provide strong evidence for an involvement of the RNA processing in mouse but not human tissues mutated for SMN, which raises the question of whether this pathway is defective in human SMA.

Although these results showed important discrepancies in the transcriptional program of RNA metabolism-related genes between human and mouse tissues mutated for SMN, other pathways involving p21, FHL2, and STMN4 genes (12, 18, 24) displayed similar expression changes in both murine and human SMA. The role that these genes may have in SMA pathogenesis will require in-depth analysis of the gene products in SMA. The availability of biomarkers specific to SMA should contribute to better evaluation of natural human SMA course and should be very helpful for accurate estimation of the potential benefit of therapeutics in SMA.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
DISCLOSURES
 REFERENCES
 
This work was supported by INSERM, AFM, Families of SMA (US), the Amyotrophic Lateral Sclerosis Association (US), the Conseil Régional d'Ile de France, and the Foundation Bettencourt Schueller.


   DISCLOSURES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 DISCLOSURES
REFERENCES
 
There is no conflict of interest.


   ACKNOWLEDGMENTS
 
We are very grateful to Mark Lathrop for making available the Agilent scanner, Dr. Virginie Rouiller-Fabre for assistance with human control biopsies, Dr. Roland Schüle for the gift of anti-FHL2 antibody, and A. Munnich and R. Friedman's departments and the biobank of the Association Française contre les Myopathies (AFM) for providing human biopsies.


   FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: J. Melki, Molecular Neurogenetics Laboratory, INSERM, Univ. of Evry, E0223, GENOPOLE, 2 rue Gaston Crémieux, CP5724, 91057 Evry, France (e-mail: j.melki{at}genopole.inserm.fr)

10.1152/physiolgenomics.00134.2005.

1 The Supplemental Material for this article is available online at http://physiolgenomics.physiology.org/cgi/content/full/00134.2005/DC1. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 DISCLOSURES
 REFERENCES
 

  1. Anderson KN, Baban D, Oliver PL, Potter A, and Davies KE. Expression profiling in spinal muscular atrophy reveals an RNA binding protein deficit. Neuromuscul Disord 14: 711–722, 2004.[Medline]
  2. Campbell L, Hunter KM, Mohaghegh P, Tinsley JM, Brasch MA, and Davies KE. Direct interaction of Smn with dp103, a putative RNA helicase: a role for Smn in transcription regulation? Hum Mol Genet 9: 1093–1100, 2000.[Abstract/Free Full Text]
  3. Chan YB, Miguel-Aliaga I, Franks C, Thomas N, Trulzsch B, Sattelle DB, Davies KE, and Van Den Heuvel M. Neuromuscular defects in a Drosophila survival motor neuron gene mutant. Hum Mol Genet 12: 1367–1376, 2000.
  4. Cifuentes-Diaz C, Frugier T, Tiziano FD, Lacene E, Roblot N, Joshi V, Moreau MH, and Melki J. Deletion of murine SMN exon 7 directed to skeletal muscle leads to severe muscular dystrophy. J Cell Biol 152: 1107–1114, 2001.[Abstract/Free Full Text]
  5. Cifuentes-Diaz C, Nicole S, Velasco ME, Borra-Cebrian C, Panozzo C, Frugier T, Millet G, Roblot N, Joshi V, and Melki J. Neurofilament accumulation at the motor endplate and lack of axonal sprouting in a spinal muscular atrophy mouse model. Hum Mol Genet 11: 1439–1447, 2002.[Abstract/Free Full Text]
  6. Coovert DD, Le TT, Morris GE, Man NT, Kralewski M, Sendtner M, and Burghes AH. Does the survival motor neuron protein (SMN) interact with Bcl-2? J Med Genet 37: 536–539, 2000.[Free Full Text]
  7. Frugier T, Nicole S, Cifuentes-Diaz C, and Melki J. The molecular bases of spinal muscular atrophy. Curr Opin Genet Dev 12: 294–298, 2002.[CrossRef][ISI][Medline]
  8. Frugier T, Tiziano FD, Cifuentes-Diaz C, Miniou P, Roblot N, Dierich A, Le Meur M, and Melki J. Nuclear targeting defect of SMN lacking the C-terminus in a mouse model of spinal muscular atrophy. Hum Mol Genet 9: 849–858, 2000.[Abstract/Free Full Text]
  9. Gangwani L, Mikrut M, Theroux S, Sharma M, and Davis RJ. Spinal muscular atrophy disrupts the interaction of ZPR1 with the SMN protein. Nat Cell Biol 3: 376–383, 2001.[CrossRef][ISI][Medline]
  10. Giesemann T, Rathke-Hartlieb S, Rothkegel M, Bartsch JW, Buchmeier S, Jockusch BM, and Jockusch H. A role for polyproline motifs in the spinal muscular atrophy protein SMN. Profilins bind to and colocalize with smn in nuclear gems. J Biol Chem 274: 37908–37914, 1999.[Abstract/Free Full Text]
  11. Gurney ME, Pu H, Chiu AY, Dal Canto MC, Polchow CY, Alexander DD, Caliendo J, Hentati A, Kwon YW, Deng HX, Chen WC, Zhai P, Sufit RL, and Siddique T. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 264: 1772–1775, 1994.[Abstract/Free Full Text]
  12. Halevy O, Novitch BG, Spicer DB, Skapek SX, Rhee J, Hannon GJ, Beach D, and Lassar AB. Correlation of terminal cell cycle arrest of skeletal muscle with induction of p21 by MyoD. Science 267: 1018–1021, 1995.[Abstract/Free Full Text]
  13. Iwahashi H, Eguchi Y, Yasuhara N, Hanafusa T, Matsuzawa Y, and Tsujimoto Y. Synergistic anti-apoptotic activity between Bcl-2 and SMN implicated in spinal muscular atrophy. Nature 390: 413–417, 1997.[CrossRef][Medline]
  14. Jones KW, Gorzynski K, Hales CM, Fischer U, Badbanchi F, Terns RM, and Terns MP. Direct interaction of the spinal muscular atrophy disease protein SMN with the small nucleolar RNA-associated protein fibrillarin. J Biol Chem 276: 38645–38651, 2001.[Abstract/Free Full Text]
  15. Lefebvre S, Burglen L, Reboullet S, Clermont O, Burlet P, Viollet L, Benichou B, Cruaud C, Millasseau P, Zeviani M, Le Paslier D, Frézal J, Cohen D, Weissenbach J, Munnich A, and Melki J. Identification and characterization of a spinal muscular atrophy-determining gene. Cell 80: 155–165, 1995.[CrossRef][ISI][Medline]
  16. Lefebvre S, Burlet P, Liu Q, Bertrandy S, Clermont O, Munnich A, Dreyfuss G, and Melki J. Correlation between severity and SMN protein level in spinal muscular atrophy. Nat Genet 16: 265–269, 1997.[CrossRef][ISI][Medline]
  17. Liu Q and Dreyfuss G. A novel nuclear structure containing the survival of motor neurons protein. EMBO J 15: 3555–3565, 1996.[ISI][Medline]
  18. Martin B, Schneider R, Janetzky S, Waibler Z, Pandur P, Kuhl M, Behrens J, von der Mark K, Starzinski-Powitz A, and Wixler V. The LIM-only protein FHL2 interacts with ß-catenin and promotes differentiation of mouse myoblasts. J Cell Biol 159: 113–122, 2002.[Abstract/Free Full Text]
  19. McWhorter ML, Monani UR, Burghes AH, and Beattie CE. Knockdown of the survival motor neuron (Smn) protein in zebrafish causes defects in motor axon outgrowth and pathfinding. J Cell Biol 162: 919–931, 2003.[Abstract/Free Full Text]
  20. Mourelatos Z, Abel L, Yong J, Kataoka N, and Dreyfuss G. SMN interacts with a novel family of hnRNP and spliceosomal proteins. EMBO J 20: 5443–5452, 2001.[CrossRef][ISI][Medline]
  21. Pellizzoni L, Baccon J, Charroux B, and Dreyfuss G. The survival of motor neurons (SMN) protein interacts with the snoRNP proteins fibrillarin and GAR1. Curr Biol 11: 1079–1088, 2001.[CrossRef][ISI][Medline]
  22. Pellizzoni L, Charroux B, Rappsilber J, Mann M, and Dreyfuss G. A functional interaction between the survival motor neuron complex and RNA polymerase II. J Cell Biol 152: 75–85, 2001.[Abstract/Free Full Text]
  23. Pellizzoni L, Kataoka N, Charroux B, and Dreyfuss G. A novel function for SMN, the spinal muscular atrophy disease gene product, in pre-mRNA splicing. Cell 95: 615–624, 1998.[CrossRef][ISI][Medline]
  24. Ravelli RB, Gigant B, Curmi PA, Jourdain I, Lachkar S, Sobel A, and Knossow M. Insight into tubulin regulation from a complex with colchicine and a stathmin-like domain. Nature 428: 198–202, 2004.[CrossRef][Medline]
  25. Williams BY, Hamilton SL, and Sarkar HK. The survival motor neuron protein interacts with the transactivator FUSE binding protein from human fetal brain. FEBS Lett 470: 207–210, 2000.[CrossRef][ISI][Medline]
  26. Yong J, Wan L, and Dreyfuss G. Why do cells need an assembly machine for RNA-protein complexes? Trends Cell Biol 14: 226–232, 2004.[CrossRef][ISI][Medline]
  27. Young PJ, Day PM, Zhou J, Androphy EJ, Morris GE, and Lorson CL. A direct interaction between the survival motor neuron protein and p53 and its relationship to spinal muscular atrophy. J Biol Chem 277: 2852–2859, 2002.[Abstract/Free Full Text]
  28. Young PJ, Francis JW, Lince D, Coon K, Androphy EJ, and Lorson CL. The Ewing's sarcoma protein interacts with the Tudor domain of the survival motor neuron protein. Brain Res Mol Brain Res 119: 37–49, 2003.[Medline]
  29. Zou J, Barahmand-pour F, Blackburn ML, Matsui Y, Chansky HA, and Yang L. Survival motor neuron (SMN) protein interacts with transcription corepressor mSin3A. J Biol Chem 279: 14922–14928, 2004.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Am. J. Pathol.Home page
J. Vitte, C. Fassier, F. D. Tiziano, C. Dalard, S. Soave, N. Roblot, C. Brahe, P. Saugier-Veber, J. P. Bonnefont, and J. Melki
Refined Characterization of the Expression and Stability of the SMN Gene Products
Am. J. Pathol., October 1, 2007; 171(4): 1269 - 1280.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Tables and Figures
Right arrow All Versions of this Article:
24/2/97    most recent
00134.2005v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Olaso, R.
Right arrow Articles by Melki, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Olaso, R.
Right arrow Articles by Melki, J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2006 by the American Physiological Society.