|Home | About | Journals | Submit | Contact Us | Français|
Duchenne muscular dystrophy (DMD) and other types of muscular dystrophies are caused by the loss or alteration of different members of the dystrophin protein complex. Understanding the molecular mechanisms by which dystrophin-associated protein abnormalities contribute to the onset of muscular dystrophy may identify new therapeutic approaches to these human disorders. By examining gene expression alterations in mouse skeletal muscle lacking α-dystrobrevin (Dtna−/−), we identified a highly significant reduction of the cholesterol trafficking protein, Niemann-Pick C1 (NPC1). Mutations in NPC1 cause a progressive neurodegenerative, lysosomal storage disorder. Transgenic expression of NPC1 in skeletal muscle ameliorates muscular dystrophy in the Dtna−/− mouse (which has a relatively mild dystrophic phenotype) and in the mdx mouse, a model for DMD. These results identify a new compensatory gene for muscular dystrophy and reveal a potential new therapeutic target for DMD.
Duchenne muscular dystrophy (DMD) is a fatal, muscle-wasting disease, affecting ~1 in 3500 live male births. Patients with DMD display muscle weakness early in life, experience progressive loss of mobility due to severe muscle degeneration, and ultimately die in the early twenties, usually from respiratory or cardiac failure. DMD results from the loss of functional dystrophin, a large intracellular membrane protein required for the formation of the dystrophin-associated protein complex (DPC), a large complex of transmembrane glycoproteins (Ervasti et al., 1990 ), adaptor proteins, and signaling proteins (Hoffman et al., 1987 ). Despite significant advances in gene therapy aimed at replacing the defective dystrophin gene (Cox et al., 1993 ; Ragot et al., 1993 ; Wang et al., 2000 ; Gregorevic et al., 2006 ), DMD still remains without an effective treatment.
The DPC forms a physical link between the intracellular actin cytoskeleton and the extracellular matrix. Members of the DPC include the dystroglycans (which provide transmembrane linkage between dystrophin and laminin), γ-actin, the sarcoglycans, and the syntrophin–dystrobrevin scaffold for signaling proteins (Ervasti, 2007 ). Mutations in DPC proteins cause other forms of muscular dystrophies, such as the limb-girdle muscular dystrophies (Ozawa et al., 2005 ). Although the DPC complex has been extensively studied, the downstream molecular and cellular alterations that lead to muscle degeneration in DMD are largely unknown. A more complete understanding of these pathways may reveal new therapeutic targets that slow the progression of muscle degeneration.
In this study, we performed gene expression analyses on skeletal muscle from mice lacking α-dystrobrevin (Dtna−/−), which display a mild dystrophic phenotype compared with mdx mice (Grady et al., 1999 ), in order to identify new genes contributing to muscular dystrophy. The Dtna−/− mice have a highly significant reduction in Niemann-Pick C1 (Npc1) transcript and protein. Mutations in Npc1 are responsible for Niemann-Pick C1 (NPC1) disease, a progressive and ultimately fatal, autosomal recessive, neurodegenerative disease, affecting ~1 in 150,000 live births. The function of NPC1 is not well understood but seems to be involved in regulating intracellular cholesterol transport (Liscum et al., 1989 ). Cells lacking NPC1 exhibit an accumulation of LDL-derived unesterified cholesterol in the endosomal/lysosomal pathway, which may impair trafficking or other cellular functions (Sokol et al., 1988 ). NPC1 has been studied extensively in brain and liver, the tissues more severely affected by the disease. The role of NPC1 in skeletal muscle has not been examined, to our knowledge.
To determine whether the restoration of NPC1 improves the health of dystrophic muscle, we expressed Npc1 in skeletal muscle of both Dtna−/− and mdx mice. Here, we show that NPC1 expression in skeletal muscle significantly ameliorates the dystrophic phenotype in both of these mouse models of muscular dystrophy. These findings identify NPC1 as a booster gene for compromised muscle (Engvall and Wewer, 2003 ) and reveal a potential new therapeutic target for muscular dystrophies.
Mice lacking dystrophin (mdx) and mice heterozygous for the Niemann Pick type C1 mutant allele were obtained from The Jackson Laboratory (Bar Harbor, ME). The NPC1 mutant allele contains the insertion of a retroposon, resulting in a premature truncation, deleting 11 of the 13 transmembrane domains. Breeding pairs were bred to produce Npc1-null (Npc1−/−), heterozygous (Npc1+/−), and wild-type (Npc1+/+) offspring. Genotyping was performed according to The Jackson Laboratory protocols (jaxmice.jax.org). Dtna−/− mice were a generous gift of Joshua Sanes (Harvard University, Boston, MA). All animal experiments were performed with the approval of the Institutional Animal Care and Use Committee at the University of Washington (Seattle, WA).
Quadriceps muscles were dissected from 6-wk-old male Dtna−/− and their wild-type, sex-matched littermates. Total RNA was isolated from each quadriceps muscle by using TRIzol Reagent (Invitrogen, Carlsbad, CA) and the RNeasy Mini kit (QIAGEN, Valencia, CA). For each sample, 20 μg of total RNA was processed for gene expression analysis, using Affymetrix Murine Genome U74Av2, U74Bv2, and U74Cv2 arrays, following Affymetrix protocols (Affymetrix, Santa Clara, CA; www.affymetrix.com). RNA from eight experimental and eight control mice was analyzed individually on the U74v2 arrays. Double-stranded cDNA was synthesized from total RNA (SuperScript Choice system; Invitrogen). Single-stranded, biotin-labeled cRNA (BioArray HighYield RNA Transcript Labeling kit [T7]; Enzo Diagnostics, New York, NY) was synthesized by in vitro transcription, and fragmented. The resulting fragmented, biotin-labeled cRNA (15 μg) was hybridized to the Affymetrix arrays, which were labeled with the GeneChip Fluidics Station 400 (Affymetrix) and scanned with the GeneChip Scanner 3000 (Affymetrix) at the Center for Array Technologies at the University of Washington. Initial data analysis was performed with Affymetrix Microarray Suite 5.0.
Real-Time quantitative RT-PCR was performed using TaqMan chemistry and the ABI 7000 sequence detection system (Applied Biosystems, Foster City, CA). Quadriceps muscles were dissected from 6-wk-old male Dtna−/− mice and their wild-type, sex-matched littermates. Real-time quantitative RT-PCR was performed on each sample using 50 ng of RNA, One-Step RT-PCR Master Mix reagents (Applied Biosystems), and Npc1 primers (5′-AATGCGGTCTCCTTGGTCAA-3′ and 5′-GCTCTCGTTATATGGCTGCAGAA-3′; Integrated DNA Technologies, Coralville, IA) and probe (5′ 6-FAM d(CACAGAAATGCCACAGCTCATCACCAA) BHQ-1 3′, Biosearch Technologies,), or control 18S primers and probe (4310893E; Applied Biosystems). The relative expression of Npc1 mRNA was normalized to 18S RNA in the same sample. Each sample was run in duplicate.
A 16-amino acid peptide (KAKRHTTYERYRGTER), corresponding to the cytosolic carboxy-terminal domain of murine NPC1 (GenBank accession no. BC052437, residues 1256–1271) was synthesized (Macromolecular Resources, Colorado State University, Fort Collins, CO), conjugated to keyhole limpet hemocyanin (Pierce Chemical, Rockford, IL), and injected into New Zealand White rabbits (Covance Research Products, Princeton, JN). Antibodies were affinity purified using peptide bound to UltraLink Iodoacetyl Gel columns (Pierce Chemical).
Tissues were dissected and quick frozen in liquid nitrogen. Membrane preparations and immunoblotting were performed as described previously (Garver et al., 2000 ) with the following modifications. Tissues were homogenized in buffer A [25 mM 2-(N-morpholino)ethanesulfonic acid, pH 6.5, 150 mM NaCl, and protease inhibitors [Sigma-Aldrich, St. Louis]) and centrifuged for 30 min at 100,000 × g, generating a cytosolic fraction (supernatant) and membrane-enriched fraction (pellet). Pellets were resuspended in buffer A plus 1% Triton X-100 and centrifuged at 2000 × g for 10 min. Protein content was determined using the bicinchoninic acid protein assay (Pierce Chemical). Gels (4–15% or 4–20% Tris-HCl gradient; Bio-Rad, Hercules, CA) were loaded with 5–25 μg of protein and run and transferred to nitrocellulose membrane (Millipore, Billerica, MA). Primary antibody was detected using horseradish peroxidase-conjugated secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA) and SuperSignal West Femto Maximum Sensitivity Substrate (Pierce Chemical) using a charge-coupled device camera (Alpha Innotech, San Leandro, CA).
Single muscle fiber preparations were adapted from Percival et al. (2007) . The fibers were incubated for 45 min in Alexa Fluor 488-conjugated donkey anti-rabbit secondary antibody (Invitrogen) diluted in blocking buffer (0.05% saponin and 10% goat serum in phosphate-buffered saline) and washed. Labeled fibers were separated on glass slides in ProLong Gold antifade mounting media, containing 4,6-diamidino-2-phenylindole (DAPI) (Invitrogen) for visualization of nuclei. Confocal microscopy images were obtained using an LSM 510 META (Carl Zeiss, Thornwood, NY) at the W. M. Keck Center for Advanced Studies in Neural Signaling (University of Washington). Stacks of 40 serial optical sections measuring 0.44 μm in thickness were obtained at 0.44-μm intervals through the muscle fibers, using a Plan-Neofluar 40×/1.3 oil objective at 2× zoom. The sections were merged into a single image using the “maximum” projection method.
Teased muscle fibers were incubated with block solution (0.2% bovine serum albumin, 0.2% fish skin gelatin [Sigma-Aldrich] in phosphate-buffered saline) for 20 min. The fibers were stained with block solution containing 1 mg/ml filipin (Sigma-Aldrich) according to the method of Carozzi et al. (2000) . Fibers were imaged using an Axioskop 2 microscope (Carl Zeiss).
The cDNA encoding full-length Npc1 was obtained from American Type Culture Collection (9890203; Manassas, VA). On sequence comparison of this clone (GenBank accession no. BC052437) with that of mouse Npc1 sequence (GenBank accession no. AF003348), multiple expressed sequence tags, and the mouse genome, we observed several sequence discrepancies. We considered the consensus sequence generated from all of the sequences to be correct. Therefore, by PCR, we made a single base mutation, a3963g, resulting in an amino acid alteration, K1273E, to match the ATCC sequence to that of the consensus. The Npc1 cDNA was cloned into the NotI site of pBSX-HSAvpA (Crawford et al., 2000 ), a generous gift from Dr. Jeffrey Chamberlain (University of Washington). The linearized construct was injected into the pronuclei of C57BL/6 x C3H embryos (Transgenic Resources Program, Department of Comparative Medicine, University of Washington), and the resulting progeny were genotyped by PCR (5′-GATGAA GCAGACAGTATTCAGC-3′ and 5′-CAGTTCGGCTCGCGGAGCAC-3′). Four founder lines carrying the Npc1 transgene were identified [Tg(Npc1)] and bred onto Dtna−/− [Dtna−/−-Tg(Npc1)] and mdx [mdx-Tg(Npc1)] backgrounds.
Tibialis anterior (TA), soleus, and diaphragm muscles were dissected, embedded in O.C.T. (Tissue-Tek) and quick-frozen in liquid nitrogen-cooled isopentane. Cross sections (8 μm) taken from the midbelly of the muscles were fixed in 95% ethanol and stained with Gill's no. 3 hematoxylin and eosin Y (H&E). Microscope (Axioskop 2; Carl Zeiss) images of the entire muscle were used to assess the presence of centrally located nuclei. A minimum of five mice per genotype was analyzed. Dtna−/−-Tg(Npc1) were examined at 3 mo of age and mdx-Tg(Npc1) at 8 wk of age. ImageJ software (National Institutes of Health, Bethesda, MD) was used to measure Feret's diameter (Briguet et al., 2004 ) of muscle fibers. Measurements were taken from an entire cross section of soleus muscle from each of three mice per genotype examined. Coefficients of variation (percentages) of the fiber diameters were calculated as SD of the muscle fiber size/mean fiber size × 100.
Blood (50–100 μl) was collected from the saphenous vein and allowed to clot for 20 min. Serum was obtained after centrifugation of the blood at 5000 × g for 8 min at room temperature. We used the CK Liquid-UV test (StanBio Life Sciences, Elkhart, IN) and followed the manufacturer's protocol for the determination of serum creatine kinase activity. A minimum of five mice per genotype was analyzed.
For statistical analyses of microarray data, two-tailed Student's paired t tests were performed on the signal intensities of all probe sets to identify significant differences between the control and experimental groups. A value of p < 0.005 was considered significant. A two-tailed Student's paired t test was performed on the quantitative RT-PCR data to determine the significance of the difference in Npc1 expression between the control and experimental groups. We performed one-way analysis of variance tests with Tukey's post test for determination of statistical significance of both central nucleation and creatine kinase activity, using GraphPad Prism version 4.00 (GraphPad Software, San Diego, CA).
To identify genes with altered expression in a mouse model of muscular dystrophy, we compared gene expression levels from muscle of Dtna−/− mice to littermate controls. The Dtna−/− mouse was chosen for these studies because its mild dystrophy does not evoke the aggressive immune response that predominates gene expression analysis studies in mdx muscle (Porter et al., 2002 ). The goal of these studies was to identify genes with expression altered before the onset of muscle degeneration. To improve the probability of identifying early gene expression changes, we used relatively young mice (6 wk) and selected a muscle (quadriceps) that is only mildly affected (based on histological analysis). We analyzed large numbers of muscle samples independently so that even small, but statistically significant changes would be revealed. Using this approach, we identified >200 transcripts that were differentially regulated in Dtna−/− muscle (Supplemental Tables 1 and 2). Because of its high level of statistical significance (p = 3.0e−7) and its involvement in a known neurodegenerative disease, our attention was directed toward the gene encoding Npc1. The array data indicated that the expression level of Npc1 transcript is reduced by ~50% in the Dtna−/− muscle (Supplemental Table 1). We used real-time quantitative RT-PCR to confirm the 50% reduction (p < 0.05) of Npc1 transcript in Dtna−/− quadriceps muscle (Figure 1).
Because reduction in mRNA does not necessarily lead to a reduction in protein levels we compared NPC1 protein levels in wild-type and Dtna−/− muscle (Figure 1). Like the Npc1 transcript levels, NPC1 protein levels are also markedly decreased in Dtna−/− muscle, compared with wild-type samples. This was particularly evident in the membrane-enriched samples as expected for the transmembrane NPC1 (Figure 1).
We used immunofluorescence to determine the location of NPC1 in skeletal muscle (Figure 2). Labeling of individual muscle fibers using our NPC1 antibody showed a specific punctate pattern not present in the NPC1 knockout muscle. This labeling colocalized with lysosomal membrane protein (LAMP1), a lysosomal marker (Figure 2).
Having established that the levels of NPC1 were reduced in the Dtna−/− mouse model of muscular dystrophy, we hypothesized that restoration of skeletal muscle NPC1 levels could ameliorate the dystrophic phenotype. We tested this hypothesis by generating transgenic mice expressing Npc1 specifically in skeletal muscle. We obtained four lines of transgenic NPC1 mice: Tg(Npc1)19, Tg(Npc1)55, Tg(Npc1)56, and Tg(Npc1)58. Two of these transgenic lines, Tg(Npc1)19 and Tg(Npc1)58, showed high expression of NPC1 in skeletal muscle (Figure 3). The transgenic mice showed no adverse effects of high levels on NPC1 in muscle. Expression of the NPC1 transgene results in a qualitative increase in the number of visible puncta but no increase in diffuse or sarcolemma labeling (Supplemental Figure 1). Each of the four Tg(Npc1) lines was crossed onto the Dtna−/− and onto the mdx backgrounds.
Muscular dystrophy is characterized, in part, by increased numbers of regenerating muscle fibers (identified by central nucleation) and damaged muscle membranes. The latter leads to elevated levels of serum creatine kinase, which provides a body-wide assessment of myofiber sarcolemmal integrity. We assessed the effect of NPC1 expression on muscular dystrophy using these two parameters.
We determined the percentage of centrally-nucleated muscle fibers from H&E-stained cross sections of a predominantly slow-twitch (soleus), fast-twitch (TA), and respiratory (diaphragm) muscle of each genotype (Figure 4). In both the soleus and diaphragm muscles, transgenic expression of NPC1 restored central nuclei counts to wild-type levels. High expression of transgenic NPC1 significantly reduced the percentage of centrally nucleated myofibers by 83% [Dtna−/−-Tg(NPC1)19] and 76% [Dtna−/−-Tg(NPC1)58] relative to Dtna−/−. We also observed smaller, but still statistically significant, reductions (40 and 34%) in the two lines expressing NPC1 at lower levels [Dtna−/−-Tg(NPC1)55 and Dtna−/−- Tg(NPC1)56, respectively] (data not shown). In diaphragm, central nuclei counts were significantly reduced in the two high NPC1-expressing transgenic lines [Dtna−/−-Tg(NPC1)19 (89%) and Dtna−/−-Tg(NPC1)58 (93%)], but the two transgenic lines expressing NPC1 at low levels did not show significant decreases (data not shown). In the TA muscle, the percentage of myofibers with central nuclei was not significantly changed in all four transgenic lines compared with Dtna−/− mice (Figure 4; data not shown).
Dtna−/− transgenic mice expressing high levels of NPC1 also exhibited marked reductions in serum creatine kinase activity. As shown in Figure 4, transgenic expression of NPC1 in Dtna−/− mice restored creatine kinase levels to wild-type levels in both the Dtna−/−-Tg(NPC1)19 and Dtna−/−-Tg(NPC1)58 lines. The reduced number of central nuclei and reduced serum creatine kinase levels demonstrate that the dystrophic phenotype in the Dtna−/− mouse is ameliorated by transgenic skeletal muscle expression of NPC1. This amelioration is dose dependent and the degree of amelioration varies among the muscle fiber type.
To determine whether high expression of NPC1 alters the phenotype in a model of severe muscular dystrophy, we used the mdx mouse, a mouse studied extensively as a model of DMD (Bulfield et al., 1984 ). All four transgenic NPC1 lines were bred onto the mdx background (Figure 5). The extent of muscular dystrophy was assessed in 8-wk-old wild-type, mdx, and mdx-Tg(Npc1) mice by quantifying the percentage of centrally nucleated myofibers, determining serum creatine kinase activity and by measuring myofiber size variability (Figure 5).
In muscles of the soleus, TA, and diaphragm, high expression of NPC1 significantly reduced the percentage of centrally nucleated fibers relative to muscles from mdx mice (Figure 5). Transgenic expression of NPC1 in mdx mice also significantly reduced serum creatine kinase levels (Figure 5). Both of the transgenic lines expressing NPC1 at high levels [mdx-Tg(Npc1)19 and mdx-Tg(Npc1)58] showed sharply decreased creatine kinase levels (by 54 and 75%, respectively) compared with mdx. In fact, creatine kinase levels in mdx-Tg(Npc1)58 mice were not significantly different from wild-type levels.
An additional characteristic of dystrophic muscle is marked variability in muscle fiber diameter. We determined the variability of fiber diameter in the soleus muscle of the mdx-Tg(Npc1)58 mouse because it seemed to be the most improved based on the degree of improvement in central nucleation and creatine kinase levels. Indeed, we found that although fiber sizes varied significantly in the mdx soleus compared with wild type, transgenic expression of NPC1 significantly reduced the size variability (Figures 5). Together, these data demonstrate that high expression of NPC1 in mdx skeletal muscle can significantly improve the dystrophic phenotype in this mouse model of Duchenne muscular dystrophy.
We used gene expression profiling of Dtna−/− mouse muscle to identify downstream cellular and/or molecular alterations that result from the loss of α-dystrobrevin, and contribute to the onset of the muscular dystrophy of Dtna−/− mice. Dtna−/− muscle displays a milder dystrophic phenotype than mdx muscle (Grady et al., 1999 ). Expression profiling studies of mdx muscle have yielded large changes associated with immune response and regeneration in the dystrophic muscle (Porter et al., 2002 ; Porter et al., 2004 ). To identify alterations in proteins and/or signaling pathways that lead to degeneration, we sought to minimize the influence of responses associated with inflammation and muscle repair by examining the less severely affected Dtna−/− quadriceps muscle at 6 wk of age. We analyzed large numbers of samples independently so statistical significance of small changes could be assessed. Selection of important changes was based on statistical significance, rather than an arbitrary magnitude change.
We identified >200 differentially expressed transcripts in the Dtna−/− quadriceps muscle. Several of the genes with increased expression, such as insulin-like growth factor-II, are likely involved in muscle regeneration, similar to trends observed in publicly available data sets (Gene Expression Omnibus database; http://www.ncbi.nlm.nih.gov/geo/) from expression analyses of other dystrophy models (Bakay et al., 2002 ; Haslett et al., 2002 ; Porter et al., 2002 ; Tseng et al., 2002 ). In this study however, we focused on Niemann Pick type C1 (Npc1), which showed a highly significant (p = 3.02e−7) ~50% decrease in transcript levels in Dtna−/− muscle. Npc1 had previously not been associated with the muscular dystrophies. However, a search using Gene Expression Omnibus database revealed that Npc1 transcript levels are decreased in muscles from mdx and dysferlin-null mice (Tseng et al., 2002 ; Wenzel et al., 2005 ).
NPC1 is a multispan membrane protein, residing primarily in late endosomes/lysosomes (Higgins et al., 1999 ; Neufeld et al., 1999 ; Garver et al., 2000 ) as well as the trans-Golgi network (Higgins et al., 1999 ) and caveolin-1 containing vesicles (Higgins et al., 1999 ; Garver et al., 2000 ). The absence of NPC1 results in intracellular accumulations of large amounts of unesterified cholesterol and glycosphingolipids in late endosomes/lysosomes. In skeletal muscle, we found that NPC1 is localized primarily in lysosomes. Although cholesterol has been shown to accumulate in every tissue of Npc1−/− mice, including skeletal muscle (Xie et al., 1999 ), the accumulation could not be observed in skeletal muscle fibers stained with filipin (Supplemental Figure 2). To our knowledge, neither the function of NPC1 in skeletal muscle, nor the effects of NPC1 deficiency in skeletal muscle has been examined.
To determine whether the reduction of NPC1 contributes to the muscular dystrophy of the Dtna−/− mouse, we generated Dtna−/− mice expressing transgenic Npc1 under the control of the human skeletal α-actin (HSA) promoter (Brennan and Hardeman, 1993 ). Many of the dystrophic characteristics normally seen in Dtna−/− muscle are ameliorated in Dtna−/−-Tg(Npc1) mice. Serum creatine kinase levels and the percentage of centrally nucleated fibers in Dtna−/− soleus and diaphragm muscles were restored to near normal levels by transgenic expression of NPC1. Perhaps more importantly, muscle-specific expression of NPC1 dramatically improved the phenotype of the more severely dystrophic mdx mice. Thus, our results suggest a new avenue for treatment of Duchenne muscular dystrophy in humans.
The loss or reduction of NPC1 by itself does not cause muscular dystrophy. We have examined several muscle types (TA, soleus, quadriceps, sternomastoid, and diaphragm) from 6-wk-old Npc1−/−, as well as 6- and 16-wk-old Npc+/− mice, and we have found no evidence of dystrophy in these mice (data not shown). However, we report here that transgenic expression of Npc1 in mdx skeletal muscle ameliorates the dystrophic phenotype. Such an apparent paradox has also been observed in the case of neuronal nitric-oxide synthase (nNOS). Neither nNOS-null mice (Chao et al., 1998 ), nor α-syntrophin-null mice (Kameya et al., 1999 ; Adams et al., 2000 ), which have reduced nNOS levels, are dystrophic. Yet, mdx mice, which also have reduced levels of nNOS (Brenman et al., 1995 ), are dystrophic, and transgenic expression of nNOS ameliorates the dystrophy in mdx mice (Wehling et al., 2001 ). A “two-hit” hypothesis has been suggested as a possible explanation for the discrepancy between the effect of nNOS reduction in the nNOS-null mouse and mdx mouse (Rando, 2001 ). According to this hypothesis, defects of the DPC are likely to have more than one biochemical consequence. Individually, either consequence may result in cell damage but alone is not enough to cause cell death; however, together they result in the severe necrosis observed in dystrophic muscle. In the case of nNOS, the reduction of nNOS causes ischemia in muscle as a result of the loss of protection to contraction-induced vasoconstriction (“first hit”), but DPC defects increase the vulnerability (“second hit”) of the muscle to ischemia, causing injury to the muscle. Although the loss of NPC1 alone may not be sufficient to cause muscle fiber degeneration, it may in the presence of DPC defects contribute to the pathophysiology of dystrophic muscle.
The mechanism by which NPC1 dysfunction causes neuronal degeneration in Niemann-Pick disease is not fully understood. However, molecular abnormalities in NPC1-null cells suggest possible links to known causes of muscle degeneration. One particularly intriguing connection involves the caveolins. Caveolin-3, the muscle-specific form, binds directly to the dystrophin complex members, β-dystroglycan (Sotgia et al., 2000 ) and nNOS (Garcia-Cardena et al., 1997 ; Venema et al., 1997 ) at the sarcolemma (Song et al., 1996 ), and it is required for the correct targeting of the dystrophin complex to cholesterol-sphingolipid rafts/caveolae (Galbiati et al., 2000 ).
A link between caveolin-3 regulation and the muscular dystrophies is well established. Muscles from Duchenne muscular dystrophy patients and mdx mouse muscles have elevated caveolin-3 levels (Vaghy et al., 1998 ; Repetto et al., 1999 ). Furthermore, mutations in caveolin-3 are the genetic basis for limb-girdle muscular dystrophy 1C (LGMD 1C) (Minetti et al., 1998 ). Finally, a Duchenne-like muscular dystrophy results from transgenic overexpression of caveolin-3 in mouse muscle (Galbiati et al., 2000 ).
Interestingly, elevated caveolin levels are also associated with NPC1 deficiency (Garver et al., 1999 ; Garver et al., 2002 ). In Npc1-heterozygous fibroblasts, caveolin levels are increased in plasma membrane caveolae (Garver et al., 2002 ). Whether the link between caveolin and NPC1 has a role in ameliorating the dystrophic phenotype in skeletal muscle requires further study.
We have identified a significant reduction of NPC1, a cholesterol and sphingolipid trafficking protein, in α-dystrobrevin-null skeletal muscle. Furthermore, we show that transgenic expression of Npc1 in skeletal muscle ameliorates the dystrophic phenotype of both Dtna−/− and mdx mice, two models of muscular dystrophy. Because cholesterol is known to affect plasma membrane rigidity and lipid-protein interactions, and because caveolae are involved in intracellular signaling, alterations in the cholesterol content of sarcolemmae or caveolae could adversely affect the structural integrity of the sarcolemmae, the ability of the membranes to repair themselves (Hernández-Deviez et al., 2008 ), and/or the localization/function of caveolar signaling proteins. The involvement of NPC1 offers new therapeutic targets for muscular dystrophies resulting from abnormalities of the dystrophin complex.
We thank Josh Sanes and Mark Grady for providing the Dtna−/− mice and Jeff Chamberlain for the HSA expression vector. This research was supported by National Institutes of Health grants P01 NS046788 and R01 NS33145 (to S.C.F.).
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E08-08-0811) on October 22, 2008.