Contribution of Dysferlin Deficiency to Skeletal Muscle Pathology in Asymptomatic and Severe Dystroglycanopathy Models: Generation of a New Model for Fukuyama Congenital Muscular Dystrophy

Defects in dystroglycan glycosylation are associated with a group of muscular dystrophies, termed dystroglycanopathies, that include Fukuyama congenital muscular dystrophy (FCMD). It is widely believed that abnormal glycosylation of dystroglycan leads to disease-causing membrane fragility. We previously generated knock-in mice carrying a founder retrotransposal insertion in fukutin, the gene responsible for FCMD, but these mice did not develop muscular dystrophy, which hindered exploring therapeutic strategies. We hypothesized that dysferlin functions may contribute to muscle cell viability in the knock-in mice; however, pathological interactions between glycosylation abnormalities and dysferlin defects remain unexplored. To investigate contributions of dysferlin deficiency to the pathology of dystroglycanopathy, we have crossed dysferlin-deficient dysferlin sjl/sjl mice to the fukutin-knock-in fukutin Hp/− and Large-deficient Large myd/myd mice, which are phenotypically distinct models of dystroglycanopathy. The fukutin Hp/− mice do not show a dystrophic phenotype; however, (dysferlin sjl/sjl: fukutin Hp/−) mice showed a deteriorated phenotype compared with (dysferlin sjl/sjl: fukutin Hp/+) mice. These data indicate that the absence of functional dysferlin in the asymptomatic fukutin Hp/− mice triggers disease manifestation and aggravates the dystrophic phenotype. A series of pathological analyses using double mutant mice for Large and dysferlin indicate that the protective effects of dysferlin appear diminished when the dystrophic pathology is severe and also may depend on the amount of dysferlin proteins. Together, our results show that dysferlin exerts protective effects on the fukutin Hp/− FCMD mouse model, and the (dysferlin sjl/sjl: fukutin Hp/−) mice will be useful as a novel model for a recently proposed antisense oligonucleotide therapy for FCMD.


Introduction
Muscular dystrophies are a heterogeneous group of genetic disorders characterized by the progressive loss of muscle strength and integrity. Several lines of evidence have established that the structural linkage between the muscle extracellular matrix and the cytoskeleton is essential in preventing the progression of muscular dystrophy [1]. The dystrophin-glycoprotein complex (DGC) forms the structural linkage, and mutations in components of this complex cause several forms of muscular dystrophy, including Duchenne and limb-girdle muscular dystrophies (LGMDs) [2]. Within the DGC, aand b-dystroglycans (DG) act as a molecular bridge between the extracellular matrix and the cytoskeleton. a-DG is a highly glycosylated extracellular subunit that functions as a receptor for extracellular matrix proteins such as laminins. Omannosyl glycosylation and a novel phosphodiester-linked modification of O-mannose, termed post-phosphoryl modification, are necessary for a-DG to serve as a functional laminin receptor [3,4]. a-DG is anchored on the plasma membrane through non-covalent interaction with a transmembrane-type b-DG, which in turn binds to the dystrophin-actin cytoskeleton.
Fukuyama congenital muscular dystrophy (FCMD: MIM 253800) is an autosomal recessive disorder characterized by severe muscular dystrophy, abnormal neuronal migration associated with mental retardation and, frequently, eye abnormalities [5]. We identified fukutin, the gene responsible for FCMD, and a 3-kb SINE-VNTR-Alu (SVA) retrotransposon insertion into the 39 UTR of fukutin as the founder mutation in FCMD [6]. This insertion causes abnormal splicing that leads to the production of non-functional fukutin protein [7]. The introduction of antisense oligonucleotides that target the splice acceptor and splicing enhancers prevented the pathogenic abnormal splicing by SVA in the cells of FCMD patients as well as model mice that carry the retrotransposal insertion [7]. Point mutations in fukutin have been reported in patients both inside and outside Japan, and recent studies have revealed a broad clinical spectrum for fukutindeficient muscular dystrophies [8]. In FCMD, a-DG is abnormally glycosylated, and its laminin-binding activity is decreased [3]. Several other forms of muscular dystrophy are caused by abnormal glycosylation of a-DG; collectively, these conditions are termed ''dystroglycanopathies''. More than 10 genes have been identified as causative genes in dystroglycanopathies [9][10][11][12][13][14], some of which encode products that possess enzyme activities involved in synthesizing O-mannosyl sugar chains on a-DG [15][16][17][18]. Fukutin, LARGE, and Fukutin-related protein (FKRP) participate in forming the post-phosphoryl moiety [4,19]. Overall, dystroglycanopathy gene products appear to be involved in Omannosyl chain synthesis and post-phosphoryl modification; mutations in these pathways commonly result in abnormal glycosylation of a-DG and reduced ligand-binding activity, disrupting the DG-mediated linkage between the extracellular matrix and the cytoskeleton [2]. Defects in DGC components or a-DG glycosylation disrupt the linkage between the extracellular matrix and the cytoskeleton, thus rendering the sarcolemma more susceptible to contractioninduced damage. This is thought to trigger an increase in intracellular Ca 2+ concentration, eventually leading to necrosis and myofiber degeneration. Myofibers possess an intrinsic mechanism for repair of damaged membranes, and dysferlin plays a pivotal role in the skeletal muscle membrane repair pathway. In humans, dysferlin deficiency leads to LGMD2B, Miyoshi myopathy or a distal myopathy with anterior tibial onset [20]. Dysferlindeficient mice show defective membrane repair and also develop muscular dystrophy [21]. Several proteins are known to interact with dysferlin [20], and it is expected that these proteins also participate in membrane repair. For example, mitsugumin 53 (MG53, also known as TRIM72) has been implicated in vesicle trafficking to the damage site during the membrane repair process [22].
We previously described a new FCMD mouse model that carries the retrotransposal insertion in the mouse fukutin ortholog [23]. These knock-in mice exhibit hypoglycosylated a-DG but do not develop muscular dystrophy. Therefore, these mice are not suitable for testing effectiveness of the antisense oligonucleotide therapy for FCMD. Although skeletal muscle-selective fukutin conditional knock-out mice, namely MCK-fukutin-cKO and Myf5-fukutin-cKO, show dystrophic phenotype [24], they are not applicable for the examination of the antisense oligonucleotide therapy because they do not possess the retrotransposal insertion. We previously reported that the small amount of normally glycosylated a-DG remaining in the skeletal muscle of the knock-in mice prevents muscular dystrophy [23]. However, it is not clear whether this residual glycosylation alone is sufficient to maintain skeletal muscle membrane integrity. We hypothesized that dysferlin functions compensate for presumed membrane fragility caused by a reduced interaction between a-DG and laminin. Furthermore, the exact contribution of dysferlin and dysferlin-interacting proteins to the pathology of dystroglycanopathy is not known. To investigate this question, we crossed dysferlin-deficient mice with two distinct dystroglycanopathy mouse models and analyzed the resultant phenotypes. In addition, if the double mutant mice carrying the retrotransposal insertion show worse dystrophic phenotype than those of dysferlin mutant mice, they can be the first model for the novel antisense oligonucleotide therapy for FCMD.

Animals
Dysferlin-deficient SJL/J mice, a strain with a large deletion in the Dysf gene [25], were purchased from Charles River Japan. The transgenic mouse carrying a neo cassette disruption of one fukutin allele (fukutin +/2 ) [26] and the transgenic knock-in homozygous mutant mouse carrying the retrotransposal insertion in the mouse fukutin ortholog (fukutin Hp/Hp ) have been described previously [23]. Genotyping for the Dysf mutant allele and the fukutin mutant allele was performed as described previously [23,25]. All animal procedures were approved by the Animal Care and Use Committee of Kobe University Graduate School of Medicine (P120202-R2) in accordance with guidelines of Ministry of Education, Culture, Sports, Science and Technology (MEXT) and Japan Society for the Promotion of Science (JSPS). The animals were housed in cages (2-4 mice per cage) with wood-chip bedding in an environmentally controlled room (25uC, 12 h lightdark cycle) and provided food and water ad libitum at the animal facility of Kobe University Graduate School of Medicine. Welltrained and skilled researchers and experimental technicians, who have knowledge of methods to prevent unnecessary excessive pain, handled the animals and carried out the experiments. Euthanization was done by cervical dislocation. At sacrifice, the muscles were harvested and snap-frozen in liquid nitrogen (for biochemistry) or in liquid-nitrogen-cooled isopentane (for immunofluorescence and histology). The number and ages of animals used in each experiment is indicated in Figure legends and graphs.

Histological and Immunofluorescence analysis
For H&E staining, cryosections (7 mm) were stained for 2 min in hematoxylin, 1 min in eosin, and then dehydrated with ethanol and xylenes. For Masson trichrome staining, sections were fixed with Bouin's solution (Sigma) for 1 hour at 60uC. The slides were incubated in solution A (5% trichloroacetic acid, 5% potassium dichromate) for 30 min, and then stained with Weigert's iron hematoxylin (Muto Chemical Co Ltd) for 15 min. After a rinse with 0.5% HCl in 70% ethanol and a subsequent rinse with warm water, the slides were incubated in solution B (0.5% phosphotungstic acid, 2.5% phosphomolybdic acid) for 1 min, and then stained with FUCHSIN-PONCEAU solution. The slides were washed with 1% acetic acid, incubated in 2.5% phosphomolybdic acid for 5 min, washed with 1% acetic acid, stained with aniline blue, washed with 1% acetic acid, dehydrated, and mounted.
For immunofluorescence analysis, sections were treated with cold ethanol/acetone (1:1) for 1 min, blocked with 5% goat serum in MOM Mouse Ig Blocking Reagent (Vector Laboratories) at room temperature for 1 h, and then incubated with primary antibodies diluted in MOM Diluent (Vector Laboratories) overnight at 4uC. The slides were washed with PBS and incubated with Alexa Fluor 488-conjugated or Alexa Fluor 555-conjugated secondary antibodies (Molecular Probes) at room temperature for 30 min. Permount (Fisher Scientific) and TISSU MOUNT (Shiraimatsu Kikai) were used for H&E staining and immunofluorescence, respectively. Sections were observed under fluorescence microscopy (Leica DMR, Leica Microsystems).
For quantitative evaluation of muscle pathology, the percentages of myofiber with centrally located nuclei were counted for at least 1,000 fibers for each genotype (n.4). For evaluation of the F4/80-positive and the collagen I-positive area, the immunofluorescence signal was quantitatively measured using Image J software. Statistical analysis was performed using values represent means with standard deviations, and p values ,0.05 were considered significant (Student's t-test and Mann-Whitney U test).

Generation of double mutant mice exhibiting both abnormal glycosylation of a-DG and dysferlin deficiency
To generate double mutant mice, we crossed dysferlin-deficient SJL/L mice (dysferlin sjl/sjl ) [25] with two distinct dystroglycanopathy models, fukutin-deficient or Large-deficient mice. Previously we reported a transgenic knock-in homozygous mutant mouse carrying the retrotransposal insertion in the mouse fukutin ortholog (fukutin Hp/Hp ) [23]. Compound heterozygous mice carrying the retrotransposal insertion and a neo cassette fukutin disruption (fukutin Hp/2 ) showed more abnormal glycosylation of a-DG than did mice homozygous for the insertion (fukutin Hp/Hp mice), although fukutin Hp/2 mice did show a detectable amount of residual a-DG glycosylation [23]. For the current study, we generated double mutant mice with the (dysferlin sjl/sjl : fukutin Hp/ 2 ) genotype (Fig. 1A). The other dystroglycanopathy model, Large-deficient Large myd mouse (Large myd/myd ) [27,28] show abnormal glycosylation with no detectable amount of properly glycosylated a-DG. The ligand binding activity of a-DG in Large myd/myd mice is greatly reduced compared with that in fukutin Hp/2 mice [23]. Breeding strategies, genotypes, and abbreviations for these double mutant mice and their controls are shown in Figure 1A and 1B.
Our previous data and those of others suggest that muscle cell membrane fragility due to loss of DG or its functional glycosylation triggers disease manifestation [24,29]. However, we have not observed evidence indicating membrane fragility in fukutin Hp/2 skeletal muscle [23]. To investigate whether membrane fragility is associated mechanistically with the deteriorated phenotype of the (dysferlin sjl/sjl : fukutin Hp/2 ) mice, we analyzed the population of albumin-positive muscle fibers. Intracellular albumin staining often is used as an indicator of muscle fiber damage or increased membrane permeability [30]. Immunofluorescence analysis suggested that the albumin-positive myofibers were almost absent in both (dysferlin sjl/+ : fukutin Hp/+ ) and (dysferlin sjl/+ : fukutin Hp/2 ) and only sparsely observed in (dysferlin sjl/sjl : fukutin Hp/+ ) skeletal muscles, whereas they appeared increased in (dysferlin sjl/sjl : fukutin Hp/2 ) skeletal muscle (Fig. 5A). Quantification of albumin-positive fibers also confirmed significant deterioration of the myofiber membrane fragility in the (dysferlin sjl/sjl : fukutin Hp/2 ) mice (Fig. 5B). These data suggest that skeletal muscle fibers in (dysferlin sjl/+ : fukutin Hp/2 ) mice have latent membrane fragility, which is protected partially by dysferlin functions, and membrane fragility caused by synergy of reduced a-DG glycosylation and dysferlin-deficiency underlies the deteriorated phenotype of the (dysferlin sjl/sjl : fukutin Hp/2 ) mice.
Characterization of muscular dystrophic changes in (dysferlin sjl/sjl : Large myd/myd ) mice We subsequently analyzed the histopathology of (dysferlin sjl/sjl : Large myd/myd ) mice. Large myd/myd mice show severe muscular dystrophic phenotypes such as infiltration of connective and fat tissues and marked variation in fiber size [28]. Almost all a-DG is hypoglycosylated in Large myd/myd mice [23]. We confirmed that the pathology of (dysferlin sjl/+ : Large myd/myd ) mice was more severe than that in (dysferlin sjl/sjl : Large myd/+ ) mice (Fig. 6). To examine whether the dysferlin functions have protective roles in Large myd/myd skeletal muscle, we compared the pathology in (dysferlin sjl/+ : Large myd/myd ) and (dysferlin sjl/sjl : Large myd/myd ) mice. The (dysferlin sjl/+ : Large myd/myd ) mice showed necrotic and centrally nucleated fibers, indicating frequent cycles of muscle degeneration and regeneration (Fig. 6C). In addition, some animals showed signs of advanced muscular dystrophic changes such as variations in fiber size and connective tissue infiltration (Fig. 6D). The (dysferlin sjl/sjl : Large myd/myd ) mice exhibited severe pathology, including marked variation in fiber size and large areas with infiltration ( Fig. 6E and F). We evaluated these pathologies quantitatively by measuring the areas of macrophage or connective tissue infiltration and the population of albumin-positive muscle fibers (Fig. 6I, J, and K). Both the macrophage-infiltrated area and the population of albumin-positive muscle fibers tended to be larger in (dysferlin sjl/sjl : Large myd/myd ) than in (dysferlin sjl/+ : Large myd/myd ); however, we did not observe statistically significant differences between the two groups. Furthermore, quantification of collagen I immunofluorescence showed no significant difference in connective tissue infiltration between (dysferlin sjl/sjl : Large myd/ myd ) and (dysferlin sjl/+ : Large myd/myd ) skeletal muscles. These results suggest that dysferlin function produces limited protective effects against the progression of severe muscular dystrophy in Large myd/myd mice. Interestingly, however, when compared with the (dysferlin +/+ : Large myd/myd ) mice, the (dysferlin sjl/sjl : Large myd/ myd ) mice showed significant increases in F4/80, collagen I and intracellular albumin staining (Fig. 6I, J, and K). The amount of dysferlin protein in total lysates from (dysferlin sjl/sjl : Large myd/myd ) and (dysferlin sjl/+ : Large myd/myd ) skeletal muscles was estimated to be ,20% and ,60% of that from (dysferlin +/+ : Large myd/myd ) muscle, respectively (Fig. 6L). These results suggest that the dramatic reduction in the amount/activity of dysferlin protein may be associated with a worse phenotype in the (dysferlin sjl/sjl : Large myd/myd ) mice. Overall, our results suggest that the protective effects of dysferlin on dystroglycanopathy phenotype appear to be diminished when the dystrophic pathology is severe and progressive and also may depend on the amount of dysferlin proteins.

Discussion
Here we have characterized the contribution of dysferlindeficiency to the pathology of dystroglycanopathy using double mutant mice for dysferlin and a-DG glycosylation. To date, several dystroglycanopathy model mice have been established. Large myd mice [28] and knock-in mice carrying the FKRP P448L mutation [32] show no detectable amounts of functionally glycosylated a-DG, no laminin binding activity, and progressive muscular dystrophy. On the other hand, other dystroglycanopathy mouse models do not show a muscular dystrophy phenotype [23]. We previously reported that a small amount of intact a-DG in fukutin Hp/2 mice is sufficient to maintain muscle cell integrity, thus preventing muscular dystrophy [23]. These results and others suggest that the presence of functionally glycosylated a-DG can decrease disease severity [33,34]. In the present study, however, we showed that although (dysferlin sjl/+ : fukutin Hp/2 ) mice did not exhibit a muscular dystrophy phenotype, (dysferlin sjl/sjl : fukutin Hp/2 ) mice developed a more exacerbated phenotype than did the dysferlin single-mutant (dysferlin sjl/sjl : fukutin Hp/+ ) mice. It has been widely accepted that a-DG glycosylation plays an important role in preventing diseasecausing membrane fragility by maintaining a tight association between the basement membrane and the muscle cell membrane, and its defects produce muscle membrane that is susceptible to damage [24,29]. The synergically exacerbated phenotype of the (dysferlin sjl/sjl : fukutin Hp/2 ) mice suggests latent membrane fragility in fukutin-deficient fukutin Hp/2 skeletal muscle. Indeed, the increased number of intracellular albumin-positive fibers in the (dysferlin sjl/sjl : fukutin Hp/2 ) mice also supports this hypothesis. It is assumed in the fukutin Hp/2 myofiber that interaction between the basement membrane and the cell membrane may be weakened, and therefore disease-causative membrane damage could occur during  It is known that dysferlin plays a role in membrane repair pathway and several proteins are known to interact with dysferlin, suggesting that dysferlin forms a protein complex during the membrane repair process. MG53 has been shown to interact with dysferlin and participate in membrane repair, and genetic disruption of MG53 in mice results in muscular dystrophy [22]. Caveolin-3 is known to interact with dysferlin and MG53 [31,35]. In the present study, however, we did not observe compensatory upregulation of these proteins in fukutin Hp/2 mice, suggesting that dysferlin functions other than membrane repair may play protective roles in the fukutin Hp/2 mice. Recently, accumulating evidence has suggested new dysferlin roles other than membrane repair, such as T-tubule formation, maintenance, and stabilizing stress-induced Ca 2+ signaling [36,37]. In addition, it has been reported that dysferlin deficiency leads to increased expression of complement factors and that complement-mediated muscle injury is associated with the pathogenesis of dysferlin-deficient muscular dystrophy [38]. Therefore, it is possible that such impairments independently or synergically contribute to the pathology of the double mutant mice.
Our results showed, rather unexpectedly, that the double-mutant (dysferlin sjl/sjl : Large myd/myd ) mice did not exhibit significant deterioration of muscle pathology compared with the single-mutant (dysferlin sjl/+ : Large myd/myd ) mice. These data suggest that the protective effects of dysferlin in Large myd/myd mice were slightly or much reduced compared with those in fukutin Hp/2 mice. Since Large myd/myd mice showed severe and rapid progressive pathology while fukutin Hp/2 mice were asymptomatic, our data suggest that the protective effect of dysferlin may be less when disease pathology is advanced and/or severe. It has been reported that a double mutant of dysferlin and dystrophin produced a more exacerbated phenotype than did either single mutant [39]. In our colony, Large myd/myd mice show much more severe and rapid progressive pathology than do dystrophin-deficient mdx mice, supporting our hypothesis of a limited protective effect of dysferlin in dystrophic pathology. Interestingly, the (dysferlin sjl/sjl : Large myd/myd ) mice, however, showed a significantly worse phenotype that did the (dysferlin +/+ : Large myd/myd ) mice. In addition, there is a tendency toward a worse phenotype in the order of dysferlin amount, i.e. (dysferlin +/+ : Large myd/myd ), (dysferlin sjl/+ : Large myd/myd ), and (dysferlin sjl/sjl : Large myd/myd ). These data support the possibility that the protective effect of dysferlin is present even in the severe dystrophic Large myd/myd mice. We conclude that dysferlin has the potential to protect muscular dystrophy progression; however, its effect may depend on disease severity and the amount/activity of dysferlin proteins.
Recently, we showed that the retrotransposal insertion in the 39-UTR region of fukutin causes abnormal mRNA splicing, which is induced by a strong splice acceptor site in SVA and a rare alternative donor site in the last exon, to produce an aberrantly spliced fukutin protein [7]. The introduction of antisense oligonucleotides that target the splice acceptor, the predicted exonic splicing enhancer, and the intronic splicing enhancer prevented the pathogenic exon trapping by SVA in the cells of FCMD patients as well as model mice (fukutin Hp/Hp and fukutin Hp/2 ) [7]. This therapeutic strategy can potentially be applied to almost all FCMD patients in Japan, and can therefore be the first radical clinical treatment for dystroglycanopathies. However, there was no animal model to test the effectiveness of the antisense oligonucleotide therapy. Since fukutin Hp/2 mice do not exhibit any signs of muscular dystrophy [23], they are not a great model for examining therapeutic effects of this strategy. Skeletal muscle-selective fukutin cKO mice, MCK-fukutin-cKO and Myf5-fukutin-cKO, showed dystrophic pathology [24], but they do not possess the retrotransposal insertion, and thus they are not applicable for testing the antisense oligonucleotide therapy. Our present study demonstrates more severe dystrophic phenotype of (dysferlin sjl/sjl : fukutin Hp/2 ) mice compared with (dysferlin sjl/sjl : fukutin Hp/+ ) mice. Since the (dysferlin sjl/sjl : fukutin Hp/2 ) mice possess the retrotransposal insertion and show dystrophic phenotype, they will be used as the first model for evaluation of the antisense oligonucleotide therapy for FCMD. There is a possibility that the absence of dysferlin could add hurdles on how to interpret the results of the antisense oligonucleotide treatments; however, our quantitative assessments established in this study could overcome this issue. For example, macrophage infiltration (Fig. 4B), connective tissue infiltration (Fig. 4D), and membrane fragility in quadriceps muscles (Fig. 5B) were significantly increased only in the (dysferlin sjl/sjl : fukutin Hp/2 ) mice. These parameters in the (dysferlin sjl/sjl : fukutin Hp/+ ) mice were not changed compared with those in the (dysferlin sjl/+ : fukutin Hp/+ ) and the (dysferlin sjl/+ : fukutin Hp/2 ) mice, and therefore can be used for quantitative evaluation for therapeutic effects of the antisense oligonucleotide treatments. We hope that generation of this novel FCMD model and establishment of the quantitative evaluation for disease severity will accelerate the future translational researches to overcome FCMD.