Loss-of-function mutations in dysferlin cause muscular dystrophy, and dysferlin has been implicated in resealing membrane disruption in myofibers. Given the importance of membrane fusion in many aspects of muscle function, we studied the role of dysferlin in muscle growth. We found that dysferlin null myoblasts have a defect in myoblast–myotube fusion, resulting in smaller myotubes in culture. In vivo, dysferlin null muscle was found to have mislocalized nuclei and vacuolation. We found that myoblasts isolated from dysferlin null mice accumulate enlarged, lysosomal-associated membrane protein 2 (LAMP2)-positive lysosomes. Dysferlin null myoblasts accumulate transferrin-488, reflecting abnormal vesicular trafficking. Additionally, dysferlin null myoblasts display abnormal trafficking of the insulin-like growth factor (IGF) receptor, where the receptor is shuttled to LAMP2-positive lyososomes. We studied growth, in vivo, by infusing mice with the growth stimulant IGF1. Control IGF1-treated mice increased myofiber diameter by 30% as expected, whereas dysferlin null muscles had no response to IGF1, indicating a defect in myofiber growth. We also noted that dysferlin null fibroblasts also accumulate acidic vesicles, IGF receptor and transferrin, indicating that dysferlin is important for nonmuscle vesicular trafficking. These data implicate dysferlin in multiple membrane fusion events within the cell and suggest multiple pathways by which loss of dysferlin contributes to muscle disease.
Dysferlin gene mutations cause three forms of inherited human muscle disease that collectively are known as dysferlinopathies, limb–girdle muscular dystrophy type 2B (LGMD2B), Miyoshi myopathy (MM) and distal anterior compartment myopathy (DACM) (1,2). In LGMD2B, the proximal muscles of the limb and girdle are most weakened, whereas the more distal limb muscles are affected in MM. Dysferlinopathy patients have markedly elevated serum levels of creatine kinase, and frequently, an inflammatory infiltrate is seen in biopsies of skeletal muscle (3,4). The age of onset of muscle symptoms in the dysferlinopathies varies, but generally disease presents during the late second decade or third decade, and affected individuals may lose their ability to ambulate. To date, there is no cure for the dysferlinopathies and current therapies are ineffective.
Dysferlin is a 237 kDa protein that is highly expressed at the plasma membrane in mature skeletal muscle. Dysferlin contains at least six C2 domains and a transmembrane domain (1,5). C2 domains can bind phospholipids and proteins. The C2A domain of dysferlin binds phospholipids in the presence of calcium (6,7). A mutation in the C2A domain of dysferlin, V67D, causes reduced phospholipid binding and is known to cause muscular dystrophy, suggesting that calcium-sensitive phospholipid binding is important for muscle maintenance (6).
Mature muscle faces frequent and varied mechanical stress that can result in membrane damage. To accommodate this damage, adult muscle is highly regenerative. Trauma, intense physical activity and muscle disease can all lead to muscle membrane damage, necessitating rapid repair to prevent cell death. One mechanism of repair involves rapid resealing of the damaged membrane. In addition to membrane patching, membrane fusion events are also critical in muscle development and growth. Singly nucleated myoblasts fuse with one another creating de novo myotubes, as well as fuse to existing myotubes, facilitating muscle regeneration. In normal muscle, muscle repair is very efficient. Muscle damage is repaired by muscle regeneration, leaving no residual lasting deficit (reviewed in 8). However, in muscular dystrophy, a large group of heritable diseases, muscle degeneration exceeds muscle regeneration. This lack of efficient repair results in chronic remodeling with fatty replacement, fibrosis and progressive weakness (9,10).
Dysferlin is thought to mediate the fusion of vesicles during membrane repair. Dysferlin null myofibers have a significant delay in membrane resealing upon laser wounding in vitro (11). In vivo dysferlin-deficient muscle shows defects in regeneration after focal muscle injury induced by injection of the toxin notexin (12). Injured dysferlin null muscle displays increased immune infiltration and delayed functional recovery, implying dysferlin is also important for cellular recruitment and clearance in damaged muscle (12). Muscle histology from dysferlin null mice containing a muscle-specific dysferlin transgene to reintroduce dysferlin expression was indistinguishable from control mice at 8 months of age, a time frame when hallmarks of dystrophy are normally present in muscle (13). These data point towards a muscle intrinsic role for dysferlin in generating the dystrophic phenotype.
Work from Bansal et al. (11) supports that dysferlin is required for normal muscle repair in the mature myofiber. This observation does not preclude other roles for dysferlin, such as a regulator of muscle growth and regeneration. Chiu et al. (12) found that dysferlin null myoblasts have defective cytokine secretion in cell culture, which correlates with a delay in immune infiltration upon notexin injury in dysferlin null mice. Additionally, human dysferlin null myoblast cultures have a decreased fusion potential, suggestive of a fusion defect in dysferlinopathy patients (14). These data provide evidence that dysferlin has a role in myoblasts, in addition to mature skeletal muscle, that may effect the capacity for muscle regeneration.
To understand the role of dysferlin in muscle repair, we studied the process of muscle growth in dysferlin null mice and in cultured cells. We utilized the naturally occurring A/J ETn dysferlinopathy mouse model that harbors a retroposon in intron 4 and expresses no dysferlin (15). A control line was generated for this dysferlin null allele by breeding the null allele into the 129SVJ background. We found that dysferlin null myoblasts are defective in myoblast–myotube fusion. In vivo, this was recapitulated as dysferlin null day 1 pups displayed vesicle accumulation and centrally placed nuclei. Dysferlin null myoblasts were shown to have an accumulation of vesicles by both electron and immunofluorescence microscopy. Infusion of insulin-like growth factor 1 (IGF1) failed to elicit muscle growth in dysferlin null muscle in vivo. Additionally, we found that dysferlin null fibroblasts display many of the same defects. These data show that dysferlin regulates vesicle trafficking and myoblast function, both of which are important for muscle regeneration upon damage.
Dysferlin null muscle has reduced fiber size
The dysferlin null allele arose through a retrotransposon insertion into the dysferlin locus in A/J mice, a commonly used strain at the Jackson Laboratory (15). Through breeding, this allele was introduced into the 129/SV emst/J line through successive generations to create the dysferlin null allele in the 129SVJ background (Dysf). A schematic is shown in Figure 1A. Skeletal muscle from 8-week-old 129/SV emst/J wild-type and dysferlin null mice was harvested, sectioned in cross-section and visualized by hematoxylin and eosin staining (Fig. 1B). Dysferlin null muscles have a reduction in the number of largest myofibers (arrowhead) and have an increased number of smaller-sized fibers (arrowhead) as shown in the histogram in Figure 1C. Furthermore, dysferlin null muscle fibers have reduced cross-sectional area (mean Dysf 658 µm2) compared with age-matched wild-type controls (mean WT 841 µm2; P = 0.001; n > 378 fibers per genotype) (Fig. 1D).
Neonatal dysferlin null muscle accumulates vesicles
We examined hind limb muscle from newborn, postnatal day 1 mice and found that dysferlin null muscle had evidence of centrally placed nuclei and fiber size variability (Fig. 2A and B) unlike wild type. Higher magnification images are shown on the right. When viewing the muscle longitudinally, the central nuclei were arranged in a row separated by a core-like zone that lacks sarcomeres and instead shows vesicular structures (Fig. 2B, arrow). These central cores were also visualized in some of the myofibers in the cross-sectional images (Fig. 2A, arrow). Additionally, in dysferlin null neonatal myofibers, the lysosomal marker lysosomal-associated membrane protein 2 (LAMP2) (red) stained vesicular structures between myofiber nuclei (blue) (Fig. 2C, white arrow). This is consistent with the vacuoles seen in Figure 2B being lysosomal in nature. Vesicles were rarely seen in wild-type muscle. At 8 weeks of age, these changes were less evident, suggesting that these postnatal changes may reflect a delay of maturation. This delay of maturation may be especially evident during this late stage of developing mouse muscle, a point where myofibers have been formed but ongoing myoblast fusion contributes significantly to muscle growth (16,17).
Dysferlin null myoblasts have reduced fusion potential
The convergence of many processes such as cell signaling, secretion, migration and protein synthesis contribute to myofiber enlargement. Additionally, large myofibers result from an increased number of myoblasts fusing into the fiber as well as hypertrophy of the individual fiber. Wild-type myoblasts were isolated and placed in low serum media to induce differentiation. Cell lysates were harvested at day 0, day 3 and day 6 of differentiation, with the myotube number increasing with each day of differentiation. Dysferlin expression is low in day 0 primary myoblast cultures, but steadily increases with differentiation state of the cultures (Fig. 3A). This was confirmed by immunofluorescence microscopy showing dysferlin expression in both myoblasts (arrowhead) and in mature myotubes (arrow) (Fig. 3B).
When induced to differentiate, myoblasts in culture undergo a predictable change in cell morphology; several days into differentiation, myoblasts retract flattened lateral cell extensions and become elongated with terminal extensions at each end. Dysferlin null myoblasts, visualized by anti-desmin staining, displayed similar morphologic changes as seen in wild-type myoblasts (Fig. 3C). The ability to elongate and form small microtubes that express desmin is consistent with the ability to undergo the same degree of myoblast and myotube differentiation. The elongated cell shape change of the prefusion myoblast coupled with the ability to form binucleate, desmin-positive microtubes indicates that dysferlin is not required for these stages. Upon further culture, where large myotubes should form, we found that dysferlin null myoblasts did not form the largest myotubes. This was quantified by counting the number of nuclei per desmin-positive myotube. Dysferlin null cultures had reduced numbers of myotubes with greater than four nuclei per myotube, 9%, compared with wild-type cultures containing 29% (P = 0.0001) (Fig. 3D). Qualitatively, this is shown in Figure 3E, where dysferlin null cultures had reduction in multinucleated myotubes. Data were collected from four independent cultures per genotype.
Dysferlin null myoblasts accumulate dense core granules near the plasma membrane
Because the dysferlin null cells have a defect in myoblast–myotube fusion, we examined the ultrastructure of dysferlin null and wild-type myoblasts by electron microscopy (Fig. 4A, n > 25 cells per genotype). The electron micrographs revealed a significant increase in large vacuolar structures with the appearance of lysosomes, as well as an accumulation of dense core granules (Fig. 4A and B). Dense core granules have been defined as vesicles originating from the trans-Golgi network that are secretory in nature and contain cargo such as hormones and lysosomal proteins (reviewed in 18). These dense core granules were 2.5 times larger than wild-type vesicles and accumulated near the plasma membrane in dysferlin null myoblasts (Fig. 4C and D) (WT, n = 36 cells and 70 vesicles; Dysf n = 25 cell and 135 vesicles). The accumulation of vesicles near the plasma membrane is consistent with a model in which dysferlin facilitates fusion of vesicles to the plasma membrane. Reduced fusion of vesicles with the plasma membrane may then lead to an accumulation of vesicles near the membrane in addition to shuttling of vesicles to alternative pathways. Double-walled autophagic vesicles were rarely seen in either wild-type or dysferlin null myoblasts approximately 0–1 per cell or 1 per every 50 vesicles.
Dysferlin null myoblasts accumulate acidic vesicles
To investigate the specific nature of the vesicle accumulation, we analyzed the lysosomal marker LAMP2. LAMP2 labels late-stage vesicles in the lysosomal pathway (19). Myoblasts were labeled with anti-LAMP2 (red) and DAPI (blue) and visualized by immunofluorescence microscopy (Fig. 5A). The white boxes highlight the area magnified to the right in Figure 5A. Dysferlin null myoblasts contained an increased number of LAMP2-positive lysosomes compared with wild-type control myoblasts (mean values 44.73 and 16.85, respectively, P = 0.04) (Fig. 5B). Furthermore, the dysferlin null lysosomes were four times larger than the lysosomes in wild-type myoblasts (n > 200 lysosomes per genotype; P = 0.02) (Fig. 5C). Interestingly, in the dysferlin null myoblasts, the LAMP2-positive vesicles aggregated near the nucleus and were not seen in the cell extensions to the same extent as wild-type cells, pointing towards a trafficking defect in dysferlin null cells.
Microtubule-associated protein1 light chain 3 (LC3) is a marker found on double-walled autophagic vesicles before fusion with lysosomes. Upon activation of the autophagic pathway, LC3 is converted to a cleaved form (20). By immunoblot, dysferlin null myoblasts contained levels of LC3 similar to wild-type myoblasts (Supplementary Material, Fig. S1). Thus, dysferlin cells have an increased number of enlarged, acidic lysosomal vesicles without obvious activation of a known component of the autophagic pathway.
Endocytic recycling is delayed in dysferlin myoblasts
Work by others has shown that dysferlin is important for vesicle fusion at the plasma membrane during myofiber repair (11). We showed that vesicle accumulation is not confined to myofibers but is also seen in myoblasts (Fig. 4). We hypothesize that dysferlin null myoblasts may have defects in endocytic recycling, due to delayed fusion of vesicles with the plasma membrane. To evaluate this, myoblasts were incubated with transferrin. Transferrin recycling employs both endocytosis and exocytosis. After transferrin engages the receptor on the cell surface, both transferrin and its receptor are rapidly internalized by endocytosis and then recycled back to the plasma membrane without degradation in the lysosome (21). This allows the receptor to be reused in later rounds of endocytosis. Dysferlin null and wild-type myoblasts were incubated for 60 min with transferrin conjugated to the fluorescent fluorophore Alexa-488. Dysferlin null myoblasts accumulate transferrin-488 (green) near the nucleus (DAPI, blue), presumably in the endocytic recycling compartment (Fig. 6A). This accumulation of transferrin-488 near the nucleus occurred in 80% of dysferlin null myoblasts, whereas only 20% of wild-type myoblasts displayed this phenotype (Fig. 6B). These data were gathered from at least four cultures isolated from independent animals.
Next, we evaluated the rate of endocytic recycling of the transferrin receptor. Primary myoblasts isolated from wild-type and dysferlin null mice were cultured and incubated with transferrin-488. After 1 h, cultures were chased with unlabeled transferrin for 0, 20, 40 or 60 min. Then, cells were analyzed using flow cytometry for Alexa-488 fluorescence. Dysferlin null cultures contained a higher percentage of Alexa-488 fluorescent-positive cells than the wild-type control at all time points. To determine the rate of endocytic recycling, the fluorescence for each time point was normalized to the initial florescence at time 0 and the single exponential decay curve was plotted. The best-fit curve was plotted for dysferlin null cultures (red dotted line) and wild-type (green dotted line) following the methods described in Girones and Davis (22) (Fig. 6C). The endocytic recycling rate, k, of the dysferlin null culture was calculated as 0.001, whereas that of wild type was 0.036 (P = 0.01) (22). These data demonstrate that dysferlin null myoblasts have a defect in endocytic recycling of the transferrin receptor and underscore dysferlin's function in vesicle trafficking and endocytic recycling.
Dysferlin null myoblasts accumulate IGFR1 aggregates that colocalize with lysosomes
Knowing that dysferlin regulates recycling of the transferrin receptor and that dysferlin null muscle has defective muscle maturation and repair, we hypothesized that there may be a defect in the recycling of the IGF receptor, since IGF is known to stimulate myofiber growth (23,24). To study this process, we isolated primary myoblasts from wild-type and dysferlin null myoblasts and serum starved the cultures for 1 h to allow the IGF1 receptor (IGFR1) to shuttle to the plasma membrane. Cultures were incubated with IGF for 60 min, followed by fixation and staining with anti-IGFR1 and anti-LAMP2 antibodies (Fig. 7). Confocal microscopy showed that dysferlin null myoblasts contained large aggregates of IGFR1 not seen in wild-type cells. These aggregates colocalized with LAMP2, a lysosomal marker, consistent with the IGF receptor being shuttled to the degradative pathway. These aggregates generally are found near the nucleus and not in the cell extensions similar to the LAMP2 and transferrin-488 staining described early. This suggests that dysferlin participates in the recycling of a variety of vesicle pathways, which carry distinct cargo, including the IGF1 receptor.
Dysferlin null skeletal muscle is unresponsive to IGF1 in vivo
We hypothesized that the mis-trafficking of the IGF1 receptor to the lysosome may alter the ability of muscle to respond to the IGF1 ligand, decreasing the hypertrophic response. To determine whether dysferlin null muscle is defective in muscle growth in vivo, wild-type and dysferlin null mice were treated with 2 mg/kg IGF1 delivered via an osmotic pump for 28 days using a previously published protocol that has shown to increase the size of myofibers in vivo, although not increasing muscle weight (25,26). It is possible that larger amounts of IGF1 or longer exposure to IGF1 may result in increased muscle mass. As expected, after 4 weeks of IGF1 treatment, the cross-sectional area of myofibers from wild-type mice quadriceps increased by >30% from 1576 to 2043 µm2 (P < 0.001), whereas dysferlin null mouse myofibers were unresponsive to IGF1 exposure with no change from 1334 to 1360 μm2 (P = 0.71) (Fig. 8A and B, n > 200 fibers per genotype). We conclude that dysferlin null skeletal muscle does not respond to IGF1 in vivo consistent with a defect in myoblast–myotube growth.
Fibroblasts from dysferlin null muscle accumulate acidic vesicles and transferrin similar to myoblasts
Dysferlin is expressed highly in muscle and in myotubes in culture, but is also found at lower abundance in nearly all other cell types (27). By electron microscopy, the dysferlin null fibroblasts accumulated more vesicles than wild-type fibroblasts similar to what was seen for dysferlin null myoblasts (Fig. 9A). Additionally, the cells contained an increased number of acidic vesicles labeled with LAMP2 (Fig. 9B). Similar to the dysferlin null myoblasts, dysferlin null fibroblasts accumulated IGF receptor aggregates (Fig. 9C) and accumulated transferrin-488 near the nucleus after a 1 h pulse with transferrin-488 (Fig. 9D). These data show that dysferlin-deficient fibroblasts display many of the same defects as dysferlin-deficient myoblasts and support that such cells may be a viable venue in which to study dysferlin function.
Mature muscle is highly regenerative. The large size of individual muscle cells, coupled with the forces of contraction, makes muscle tissue prone to damage such that it is estimated that 1–2% of myonuclei turn over weekly (28). Dysferlin is a membrane-associated protein that facilitates the resealing of normal muscle membrane (11). The absence of dysferlin leads to delayed resealing and a loss of the calcium sensitivity of resealing in muscle (11). These cell-based observations are consistent with the in vitro data showing that the first C2 domain of dysferlin, C2A, binds to negatively charged phospholipids (6). In muscular dystrophy, ongoing and concomitant cycles of muscle injury occur. The identification of dysferlin as a genetic cause of muscular dystrophy is consistent with its role in mediating the resealing of the sarcolemma in injured myofibers (1,29). However, the loss of dysferlin may alter any number of membrane fusion events in muscle, including cell-based regeneration. The data shown here support additional roles for dysferlin beyond the myofiber. Dysferlin is expressed in myoblasts, and dysferlin null myoblasts show accumulation of intracellular vesicles reminiscent of what is described for muscle. It is possible that under normal conditions, dysferlin acts to suppress vesicle formation. We favor the model where dysferlin promotes vesicle fusion because the analogous synaptotagmins promote vesicle fusion (30). The calcium-sensitive phospholipid-binding properties of synaptotagmin and dysferlin C2 domains are similar. The ferlin family member, otoferlin, also promotes calcium-sensitive membrane fusion (31). However, it is possible that ferlin family members promote some types of fusion whereas inhibiting others. These data are consistent for a broad role for dysferlin in regulating vesicle fusion and multiple intracellular events that rely on normal vesicle fusion. Interestingly, the expression of dysferlin is not upregulated in response to damage like the homolog, myoferlin (32), perhaps because higher levels of dysferlin in mature muscle are sufficient to mediate responses to injury.
The dysfunction of dysferlin null myoblasts may be tightly coupled to developmental delay. Neonatal dysferlin null muscle displays many features reminiscent of immature muscle with centrally placed nuclei and lack of full sarcomere formation. The presence of lysosomes and vesicles in neonatal muscle may arise from a growth defect, or alternatively may contribute to the growth defect. A broad defect in vesicle fusion is likely to impair many properties, including those for normal differentiation. This correlates with previously published data showing that loss of dysferlin decreases levels of myogenin, a critical transcription factor required for proper muscle development (14). Additionally, dysferlin null muscle has defects in muscle regeneration upon injury (12). Whether dysferlin is mediating muscle formation from the myoblast, myotubes, or both, during muscle growth and or repair is still not fully resolved.
Dysferlin null muscle did not hypertrophy in response to continuous in vivo stimulation with IGF1, whereas control muscle cross-sectional area increased by 30%. IGF1 induces skeletal muscle growth and repair in normal and atrophied skeletal muscle, whereas lack of IGF1 causes muscle atrophy and reduced muscle regeneration (33–35). The lack of response to IGF1 by dysferlin null muscle is not likely to be absolute. For example, higher doses of IGF may be able to overcome this defect by recruiting a sufficient number of receptors to recycle so that adequate signaling is achieved for muscle growth.
Herein, we show that dysferlin regulates receptor trafficking as dysferlin null myoblasts had a reduced receptor recycling rate and mislocalized the IGFR1, resulting in aggregates within the myoblast. We speculate that the defect in vesicle trafficking in the dysferlin null myoblast may alter any number of receptors known to undergo recycling, and this likely alters many muscle and nonmuscle processes. Obtaining and studying primary myoblasts in culture, especially from dysferlinopathy patients, is difficult. Obtaining muscle from patients with muscle disease is complicated since these patients have reduced muscle mass. Moreover, many patients are diagnosed by genetic testing, eliminating the clinical need for muscle biopsy. We found that many defects found in the dysferlin null myoblast are also present in the dysferlin null fibroblast. We show that similar to dysferlin null myoblasts, dysferlin null fibroblasts had an increased number of acidic LAMP2 vesicles, aggregated IGFR1 and retained fluorescently labeled transferrin near the nucleus. These assays could be adapted to high-throughput screens to identify agents that reverse these processes. It is yet to be determined whether dysferlin null skin fibroblasts contain the same defects as muscle-derived fibroblasts, but if this phenomenon holds true, this finding will improve the ability to study dysferlinopathy patient mutations and facilitate sharing of resources.
MATERIALS AND METHODS
Generation of the Dysferlin 129 mouse line
The A/J ETn line from Jackson Laboratory stock 000646 was crossed to Jackson Laboratory 129/SV emst/J mouse stock 002065 to generate the dysferlin null allele in the 129 background. F1 offspring were backcrossed for six more generations while genotyping each cross for the presence of the retroposon insertion into the dysferlin gene (15) to create a dysferlin null line that is 98.4% 129/SV emst/J. This model is referred to throughout as Dysf.
Muscle fiber size analysis
The quadriceps muscles from 8-week-old wild-type and dysferlin null mice were dissected and frozen in liquid nitrogen-cooled isopentane. Sections from the mid-belly of the muscle were stained with hematoxylin and eosin. Using ImageJ particle analysis, the area of individual fibers was determined from over 350 fibers from at least two fields from three different animals per genotype. Mean fiber size and fiber variation were determined using Prism (Graphpad, La Jolla, CA, USA) (32). Statistics were done using an unpaired t-test.
Postnatal day 1 muscle analysis
The hind limbs from postnatal day 1 wild-type and dysferlin null mice were dissected and fixed in formalin. Hind limbs were paraffin-embedded, sectioned and stained with hematoxylin and eosin. Using a Zeiss Axiophot microscope and Ivision software (Carl Zeiss), ×40 and ×100 images were taken. At least three animals of each genotype were analyzed. For immunofluorescence microscopy, day 1 hind limb muscles were frozen in liquid nitrogen-cooled isopentane, sectioned and then fixed with 50% acetone–50% methanol for 2 min.
Primary myoblast cultures
Primary myoblasts and fibroblasts were isolated from day 0–1 wild-type and dysferlin null pups as described in Doherty et al. (36). Growth media was made using Ham's F-10 supplemented with 20% fetal bovine serum, 1% penicillin/streptomycin/actinomycin and 2.5 ng/ml of fibroblast growth factor (Fisher, PR-G5071). Differentiation media contained DMEM, 2% horse serum and 1% PSA. Cells were grown at 7% carbon dioxide. All reagents are from Invitrogen.
Immunostaining and microscopy
Isolated myoblasts and fibroblasts were grown on coverslips, fixed in PFA and stained. Anti-LAMP2 (Abcam, ab13524) was used at 1:600, anti-desmin (Sigma, DE-V-10) at 1:300, anti-dysferlin Hamlet at 1:200 (NCL), anti-IGFR1 at 1:200 (Santa Cruz, SC-713) and anti-γ sarcoglycan (37) at 1:200. Donkey anti-rat cy3 and goat anti-rabbit cy3 were used at 1:2500 (Jackson Immunologicals, West Grove, PA, USA), and goat anti-mouse and goat anti-rabbit 594 were used at 1:2500 (Molecular Probes). Blocking and antibody incubations were done in 1x PBS containing 5–10% fetal bovine serum. Coverslips were mounted using Vectashield with DAPI. Images were captured using either a Zeiss Axiophot microscope and Ivision software (Carl Zeiss) or a Leica SP2 scanning laser confocal microscope and LCS Leica Confocal Software. ImageJ was used to subtract background from each image and for analysis of particle size and particle number. Statistics were done using an unpaired t-test.
Proteins transferred to membranes were immunoblotted with anti-dysferlin Hamlet (1:2000, NCL) or anti-tubulin (1:2000, Sigma). Secondary antibodies, goat anti-mouse HRP or goat anti-rabbit HRP (Jackson ImmunoResearch), were used at a dilution of 1:5000. Blocking and antibody incubations were done in 5% milk and rinsed with TBS-T. ECL-Plus chemiluminescence (Amersham-Pharmacia) and Kodak Biomax MS film were used for detection.
Cell morphology and fusion
Primary myoblasts were plated on NaOH-treated glass coverslips in a six-well plate. After 24 h, growth media was replaced with differentiation media. Cultures were fixed with PFA after 72 h of differentiation for morphology experiments—48 h for fusion experiments. Cultures were stained with anti-desmin and visualized using a Zeiss Axiophot microscope and Ivision software. Five random images were taken, each with equivalent number of nuclei from four cultures isolated from individual animals for each genotype. The number of nuclei per desmin-positive cell was counted and binned. An unpaired t-test was performed using Prism Graphpad.
Primary myoblasts and fibroblasts were plated on NaOH-treated glass coverslips. After 24 h, cells were rinsed with PBS and fixed in 2.5% glutaraldahyde. Cultures were treated as in Demonbreun et al. (25). Samples were photographed on a Tecnai electron microscope. Vesicle size was calculated using ImageJ. Statistics were done using an unpaired t-test.
Transferrin internalization and recycling assays
Primary myoblasts and fibroblasts were isolated as described earlier. Cells were plated on NaOH-treated glass coverslips in six-well plates. Experiments were conducted as described in Doherty et al. (36). Briefly, cells were incubated with Alexa-488-conjugated transferrin for 60 min for transferrin accumulation analysis. To assess transferrin recycling, cells were then chased with unlabeled transferrin for 0, 20, 40 or 60 min. Cells were fixed in 4% parafomaldehyde, mounted for visualization or FACS-sorted. Images were captured using a Zeiss Axiophot microscope and Ivision software (Carl Zeiss). FACS data were analyzed using FLOWJO software (Tree Star, Inc., Ashland, OR, USA). If a FACS sorted culture had fewer than 100 cells, it was excluded from analysis. The transferrin receptor recycling exponential decay constant was calculated as in Girones and Davis (22), using Prism (Graphpad). Data were acquired from six independent primary myoblast cultures per genotype. A t-test was done based on the k-constant derived from the best-fit curve of six individual cultures. Curves that did not converge were excluded.
IGFR1 localization assay
Primary myoblasts and fibroblasts were isolated and grown as described earlier. Cells were plated on NaOH-treated glass coverslips in six-well plates. The assay was performed as in Doherty et al. (36). Briefly, cells were serum starved and then stimulated with IGF1 for 0 and 60 min. Cultures were rinsed and fixed with PFA and stained for IGFR1 and LAMP2. Cells were fixed in Vectashield with DAPI. Images were taken on the Leica SP2 scanning laser confocal microscope and LCS Leica Confocal Software. ImageJ was used to subtract background from each image.
IGF1 treatment of mice
Dysferlin null mice were generated as described earlier. For 28 days, 2mg/kg of Long™ R3 IGF1 (SAFC Biosciences, Lenexa, KA, USA) or PBS were administered via Alzet pumps in wild-type and dysferlin age-matched mice (Fischer Scientific, model 1002) as described in references (25,26). Quadriceps muscles were harvested, sectioned and immunostained with an anti-λ-sarcoglycan antibody as described in Demonbreun et al. (25). Fiber cross-sectional area from at least three IGF1- and two PBS-treated animals per genotype and treatment was calculated using ImageJ. An unpaired t-test was performed using Prism (Graphpad).
Conflict of Interest statement. None declared.
Supported by NIH NS47726 and the Muscular Dystrophy Association.