Muscle fibers attach to laminin in the basal lamina using two distinct mechanisms: the dystrophin glycoprotein complex and the α7β1 integrin. Defects in these linkage systems result in Duchenne muscular dystrophy (DMD), α2 laminin congenital muscular dystrophy, sarcoglycan-related muscular dystrophy, and α7 integrin congenital muscular dystrophy. Therefore, the molecular continuity between the extracellular matrix and cell cytoskeleton is essential for the structural and functional integrity of skeletal muscle. To test whether the α7β1 integrin can compensate for the absence of dystrophin, we expressed the rat α7 chain in mdx/utr−/− mice that lack both dystrophin and utrophin. These mice develop a severe muscular dystrophy highly akin to that in DMD, and they also die prematurely. Using the muscle creatine kinase promoter, expression of the α7BX2 integrin chain was increased 2.0–2.3-fold in mdx/utr−/− mice. Concomitant with the increase in the α7 chain, its heterodimeric partner, β1D, was also increased in the transgenic animals. Transgenic expression of the α7BX2 chain in the mdx/utr−/− mice extended their longevity by threefold, reduced kyphosis and the development of muscle disease, and maintained mobility and the structure of the neuromuscular junction. Thus, bolstering α7β1 integrin–mediated association of muscle cells with the extracellular matrix alleviates many of the symptoms of disease observed in mdx/utr−/− mice and compensates for the absence of the dystrophin- and utrophin-mediated linkage systems. This suggests that enhanced expression of the α7β1 integrin may provide a novel approach to treat DMD and other muscle diseases that arise due to defects in the dystrophin glycoprotein complex. A video that contrasts kyphosis, gait, joint contractures, and mobility in mdx/utr−/− and α7BX2-mdx/utr−/−mice can be accessed at http://www.jcb.org/cgi/content/full/152/6/1207.
The defective association of skeletal and cardiac muscle with their surrounding basal lamina underlies the pathologies associated with a variety of muscular dystrophies and cardiomyopathies (Matsumura and Campbell 1994; Hayashi et al. 1998; Lim and Campbell 1998). Duchenne muscular dystrophy (DMD) is a congenital X-linked myopathy that is caused by a lack of the dystrophin protein and affects approximately 1 in 3,300 males. Patients with DMD experience progressive muscle deterioration and debilitation that severely restricts mobility. Death due to cardiac and respiratory failure usually occurs in the second decade of life.
Mutations in the dystrophin gene result in a lack of dystrophin, a 427-kD protein localized at the cytoplasmic side of the plasma membrane of skeletal and cardiac muscle cells (Monaco et al. 1986; Matsumura and Campbell 1994; Campbell 1995). In association with dystroglycans, syntrophins, and sarcoglycans, dystrophin links the cell cytoskeleton to laminin in the extracellular matrix. In the absence of one or more components of the dystrophin linkage system, the association of fibers with the surrounding basal lamina is compromised, leading to the myopathy observed. Thus, the molecular continuity between the extracellular matrix and the cell cytoskeleton is essential for the structural and functional integrity of muscle.
The integrins are αβ heterodimeric receptors that bind extracellular matrix proteins and interact with the cell cytoskeleton (Hynes 1992). The α7β1 integrin is a laminin receptor on skeletal and cardiac muscle (von der Mark et al. 1991; Song et al. 1992) and serves as a transmembrane link between the basal lamina and muscle fibers. Multiple isoforms of the α7 and β1 chains are generated by developmentally regulated RNA splicing, resulting in a family of receptors with diverse structures and functions (for reviews see Hodges and Kaufman 1996; Burkin and Kaufman 1999).
The α7 integrin chain is encoded by a single autosomal gene on human chromosome 12q13 (Wang et al. 1995). Three alternative cytoplasmic domains (α7A, α7B, and α7C) and two extracellular domain variants (X1 and X2) of the protein have been identified (Collo et al. 1993; Song et al. 1993; Ziober et al. 1993). Four additional alternatively spliced isoforms of the extracellular domain have been predicted by nucleotide sequence analysis (Leung et al. 1998; Vignier et al. 1999).
The β1 chain cytoplasmic domain also undergoes developmentally regulated alternative splicing. β1A is the most common isoform of the β1 chain and is expressed in a wide variety of tissues including replicating myoblasts. The alternative β1D form is generated upon differentiation of myoblasts to myofibers (van der Flier et al. 1995; Zhidkova et al. 1995; Belkin et al. 1996, Belkin et al. 1997).
Mutations in the genes that encode the many components of the dystrophin glycoprotein complex cause a variety of muscular dystrophies. Mutations in the α7 gene also cause congenital myopathies (Hayashi et al. 1998). Thus, both the integrin- and dystrophin-mediated transmembrane linkage systems contribute to the functional integrity of skeletal muscle. Interestingly, there is an increase in the amount of α7 transcript and protein in DMD patients and mdx mice (the mouse model that has a mutation in its dystrophin gene) (Hodges et al. 1997). This led us to suggest that enhanced expression of the integrin may partially compensate for the absence of the dystrophin glycoprotein complex (Hodges et al. 1997; Burkin and Kaufman 1999).
Utrophin, a protein homologous to dystrophin, is also increased in DMD patients and mdx mice (Law et al. 1994; Pons et al. 1994). Utrophin associates with many of the same proteins as dystrophin and further increasing utrophin may, in part, also compensate for the absence of dystrophin (Tinsley et al. 1996).
Although DMD patients (Monaco et al. 1986) and mdx mice (Bulfield et al. 1984; Sicinski et al. 1989) both lack dystrophin, the pathology that develops in the mdx mouse is much less severe than that observed in humans. The differences in the extent of pathology may be due to several factors, including the enhanced expression and altered localization of utrophin (Law et al. 1994; Pons et al. 1994) and the α7 integrin chain (Hodges et al. 1997) in mdx mice. In addition, differences in utilization of skeletal muscles by humans compared with mice in captivity may also contribute to the decreased level of pathology seen in mdx mice. In contrast, mdx/utr−/− mice lack both dystrophin and utrophin and have a phenotype that is similar to that seen in DMD patients. These double mutant mice develop severe progressive muscular dystrophy and die prematurely between 4 and 20 wk of age (Deconinck et al. 1997b; Grady et al. 1997b).
To explore the hypothesis that enhanced expression of the α7β1 integrin may compensate for the absence of the dystrophin glycoprotein complex and reduce the development of severe muscle disease, transgenic mice were made that express the rat α7 chain. We report that mdx/utr−/− mice with enhanced expression of the α7BX2 chain isoform show greatly improved longevity and mobility compared with nontransgenic mdx/utr−/− mice. Transgenic mice maintained weight and had reduced spinal curvature (kyphosis) and joint contractures. Transgenic expression of the α7BX2 chain also reduced the degree of mononuclear cell infiltration and expression of fetal myosin heavy chain (fMyHC) in muscle fibers. Together these results show that enhanced expression of α7BX2β1D integrin significantly reduces the development of muscular dystrophy.
Materials and Methods
Muscle Creatine Kinase-α7BX2 Integrin Construct
The cDNA encoding the rat α7BX2 integrin isoform was cloned into pBK-RSV vector (Stratagene) downstream of the 3.3-kb mouse muscle creatine kinase (MCK) promoter and the mouse α7 integrin cell surface localization signal sequence using the restriction sites AatII and KpnI. The MCK promoter was provided by Dr. Stephen Hauschka (University of Washington, Seattle, WA). The construct was verified by DNA sequencing. Previous studies have shown that the MCK promoter is only active in heart and skeletal muscle (Jaynes et al. 1986; Johnson et al. 1989; Shield et al. 1996). The expression and functionality of the MCK-α7BX2 integrin construct were verified by transfecting C2C12 myoblasts (Burkin et al. 1998, Burkin et al. 2000).
Production of Transgenic mdx/utr−/− Mice
The MCK-α7BX2 construct was gel purified. mdx/utr+/− female mice were superovulated, mated to mdx/utr+/− male mice, and fertilized oocytes were collected. The MCK-α7BX2 construct was microinjected into male pronuclei, and injected oocytes were placed into pseudopregnant mice at the University of Illinois Transgenic Animal Facility. Resulting pups were weaned at 3 wk of age, and genomic DNA was isolated from 0.5-cm tail clips using a DNA isolation kit (Promega). Primers (MCK1, 5′-caagctgcacgcctgggtcc-3′; and AATII, 5′-ggcacccatgacgtccagattgaag-3′), used to amplify between the MCK promoter and the α7 integrin cDNA, produced a 455-bp fragment only in transgenic mice. Transgenic mdx/utr+/− male mice were bred with mdx/utr+/− female mice to produce transgenic α7BX2-mdx/utr−/− offspring. The mdx/utr+/− mice (Grady et al. 1997b) used in these experiments were provided by Dr. Joshua Sanes (Washington University, St. Louis, MO).
The mdx mutation was screened by the amplification-resistant mutation system described by Amalfitano and Chamberlain 1996. A new forward primer (Int22-306F, 5′-catagttattaatgcatagatattcag-3′) upstream of the mdx mutation site was used to yield a larger, 275-bp band. The status of the utrophin gene was analyzed by PCR using the primers 553, 554, and 22803 described previously by Grady et al. 1997a. C57BL6 × SJ6 mice were used as wild-type controls.
Antibodies and Reagents
For immunofluorescence analysis, the mouse monoclonal antibody O26 was used to detect the transgenic rat α7 chain. Polyclonal anti-α7CDA (345) and anti-α7CDB (347) were used in Western blots to detect the α7A and α7B cytoplasmic domains, respectively (Song et al. 1993; Martin et al. 1996). The amino acid sequences of the mouse and rat α7 chain cytoplasmic domains are identical, thus these antisera will detect both species with equal affinity. Peptides used to make these polyclonal antibodies were used as blocking controls. Rabbit polyclonal antibodies to the cytoplasmic domains of the β1A and β1D integrin chains were provided by Dr. W.K. Song (K-JIST, Kwang-ju, Korea; Kim et al. 1999). Dystrophin was detected using an antidystrophin monoclonal antibody (MANDRA1) (Sigma-Aldrich). Antiutrophin monoclonal antibody (NCL-DRP2) was purchased from Novacastra. The anti-fMyHC monoclonal antibody 47A was a gift from Dr. Peter Merrifield (University of Western Ontario, London, Ontario, Canada). The anti-creatine kinase monoclonal antibody was obtained from ADI Diagnostics. Acetylcholine receptor (AChR) clusters were detected with rhodamine-labeled α-bungarotoxin (Molecular Probes). FITC-labeled donkey anti–mouse and anti–rabbit antibodies were purchased from Jackson ImmunoResearch Laboratories.
Western Blot Analysis
Samples of muscle tissue were extracted in 200 mM octyl-β-d-glucopyranoside or 1% NP-40 in 50 mM Tris-HCl, pH 7.4, 2 mM PMSF, 1:200 dilution of Protease Cocktail Set III (Calbiochem), 1 mM CaCl2, and 1 mM MgCl2 at 4°C for 1 h. The supernatants were collected and protein concentrations were determined using Bradford assays. Equal amounts of extracted muscle proteins were separated on 8% polyacrylamide-SDS gels at 40 mA for 50 min, and the protein was transferred to nitrocellulose filters. Blocked filters were incubated with 1:500 dilutions of either polyclonal anti-α7CDA (345), anti-α7CDB (347), or anti-β1D antibody. HRP-linked anti–rabbit secondary antibody (Jackson ImmunoResearch Laboratories) was used to detect primary antibodies. Immunoreactive protein bands were detected using an ECL kit (Amersham Pharmacia Biotech). Specificity of the bands was determined using the blocking peptides used to raise the antibodies. Blots were reprobed with an anti-creatine kinase antibody. The intensities of the α7 bands were compared with creatine kinase and to total protein stained with Ponceau S using ImageQuant software. Comparable results were obtained by both methods.
Southern Blot Analysis
Mouse genomic DNA was isolated from whole blood or liver using a genomic DNA isolation kit (Promega). DNA was cleaved with EcoRI and KpnI at 3 U/μg of DNA for 16 h. DNA fragments were separated on 0.8% agarose gels and alkaline transferred to Hybond-XL nylon membranes (Amersham Pharmacia Biotech) (Sambrook et al. 1989). A 367-bp probe from the rat α7 3′-untranslated domain was isolated. The probe was directly labeled with HRP using a North2South nonradioactive kit (Pierce Chemical Co.). The hybridized blots were washed following the manufacturer's instructions. Probes were detected using an ECL substrate. Blots were exposed to x-ray film for 1–30 min.
Quadriceps muscles from 10-wk-old wild-type male mdx, mdx/utr−/−, and α7BX2-mdx/utr−/− mice were embedded in OCT compound (Tissue-Tek) and frozen in liquid nitrogen–cooled isopentane. Using a Leica CM1900 series cryostat, 10-μm sections were cut and placed on microscope slides coated with 1% gelatin, 0.05% chromium potassium sulfate. Sections were fixed in −20°C acetone for 1 min, rehydrated in 1× PBS for 10 min, and blocked in PBS containing 10% horse serum for 15 min. The rat α7 chain was detected using 5 μg/ml of purified O26 monoclonal antibody directly labeled with Alexa 488 (Molecular Probes). The anti-β1D antibody was used at a 1:100 dilution in 1% horse serum in PBS. The antidystrophin antibody was used at a 1:100 dilution, whereas antiutrophin and anti-fMyHC antibodies were diluted 1:2 in 1% horse serum in PBS. Rhodamine-labeled α-bungarotoxin was used at a 1:3,000 dilution to detect neuromuscular junctions (NMJs).
Endogeneous mouse immunoglobulin was blocked before the addition of monoclonal antibodies using 60 μg/ml goat anti–mouse monovalent Fabs (Jackson ImmunoResearch Laboratories) in 1% horse serum in PBS for 30 min at room temperature. Slides were then washed three times for 5 min in 1% horse serum in PBS. Primary antibodies were added for 1 h at room temperature. Slides were washed three times for 5 min in 1% horse serum in PBS. Primary antibodies were detected with a 1:100 dilution of FITC-labeled donkey anti–mouse or anti–rabbit antibody in 1% horse serum in PBS. Slides were mounted using Vectorshield mountant (Vector Laboratories). Localization of the antibody was observed with a ZEISS Photomicroscope III. Images were acquired with a Sony DXC9000 color video CCD camera using SiteCam software and a ZEISS Axiocam digital camera (see Fig. 10).
10-μm cryosections from the quadriceps muscles of 5-, 8-, and 10-wk-old wild-type, mdx, mdx/utr−/−, and transgenic mdx/utr−/− mice were placed on uncoated slides and stained with hematoxylin and eosin. The occurrence of central nuclei was scored in ≥1,000 fibers in two mice from each line. The extent of mononuclear cell infiltration was estimated as the percentage of fields (40× objective) in which two or more clusters (>10 cells) of mononuclear cells were detected. At least 80 fields in four sections from each of the duplicate mice were scored.
Sternomastoid muscles from 5-wk-old animals were fixed in 2% glutaraldehyde and 2.5% paraformaldehyde in 0.2 M Sorenson's phosphate buffer, embedded in LX-112 epoxy (Ladd Research Industries), sectioned at 0.1 μm using a Reichert Ultracut E Ultramicrotome, stained with uranyl acetate and lead citrate, and viewed with a Hitachi H-600 transmission electron microscope at a 20,000× magnification.
X-Ray and Magnetic Resonance Imaging
Spinal curvature (kyphosis) in 10-wk-old mdx, mdx/utr, and transgenic α7BX2-mdx/utr−/− mice was visualized by x-ray imaging using a Siemens Heliodent 70 x-ray machine (model D3104). X-rays were taken at 70 kVp and 7 mA. At least two mice of each genotype were analyzed.
Magnetic resonance imaging (MRI) of 10-wk-old wild-type, mdx, mdx/utr−/−, and α7BX2-mdx/utr−/− mice was used to visualize soft tissues. Mice were imaged at 1-mm thickness using a 4.7 T/31 cm Surrey Medical Imaging Spectrophotometer. At least two mice of each genotype were analyzed.
Survival data from 84 mdx/utr−/− mice and 43 transgenic α7BX2-mdx/utr−/− mice were analyzed using the Kaplan-Meier method (Kaplan and Meier 1958). Survival curves were generated for both populations and the data were compared using log-rank (Peto et al. 1977) and Wilcoxon (Conover 1980) statistical tests.
Online Supplemental Material
A video that contrasts kyphosis, gait, joint contractures, and mobility in mdx/utr−/− and α7BX2-mdx/utr−/− mice can be viewed at http://www.jcb.org/cgi/content/full/152/6/1207.
Production of Transgenic Mice Expressing the Rat α7BX2 Chain
To test the hypothesis that the α7β1 integrin linkage system can alleviate severe muscle disease, transgenic mice were produced that express the rat α7 chain. DNA encoding the rat α7 integrin α7BX2 isoform, under the transcriptional control of the mouse MCK promoter, was cloned and shown to have biological activity in vitro (Burkin et al. 1998). The 3.3-kb MCK promoter limits transcription to differentiated skeletal and cardiac muscle, confining the effects of overexpression to these tissues (Donoviel et al. 1996). The 7.1-kb construct, MCK-α7BX2, was used to express the rat integrin in mdx/utr−/− mice. Due to the mortality of the double knockout mice, the rat transgene was initially introduced into a heterozygous (mdx/utr+/−) background, and these animals were then bred to produce double knockout transgenic offspring. The ratio of offspring followed expected Mendelian genetics, indicating that the transgenic expression of the rat α7 integrin did not have an obvious effect on embryonic development. Two independent lines of transgenic mice were produced: α7BX2#2 and α7BX2#9. In this report, the data from the α7BX2#2 line are presented, but similar results were obtained with both.
The presence of the rat α7 transgene was detected by both PCR and Southern blot analyses. Using MCKI and AATII primers, a 455-bp product was amplified only in transgenic mice (Fig. 1 A). Southern blot analysis produced a strong 7.1-kb band only in transgenic mice. This is the expected size of the EcoRI and KpnI digested MCK-α7BX2 construct (Fig. 1 B). A weak 14.2-kb band was also detected by Southern blot analysis, suggesting that a portion of the constructs had lost one of these restriction sites.
The status of the utrophin gene was analyzed by PCR using the primers 553, 554, and 22803 described previously (Grady et al. 1997a). The 640-bp product is amplified when the wild-type utrophin allele is present, whereas the 450-bp product is amplified when the utrophin mutant allele is present (Fig. 1 C).
The status of the dystrophin gene was determined by the amplification-resistant mutation detection system (Amalfitano and Chamberlain 1996). Using the mdx-specific primer set, a 275-bp mutant allele is detected, whereas in separate reactions the wild-type–specific primer set detects a 275-bp wild-type allele. Fig. 1 D shows three different genotypes at the dystrophin locus. Mouse 2 is wild-type at the dystrophin locus, mouse 3 is heterozygous (mdx/+), and mouse 4 is mdx.
Protein expression from the rat α7 chain transgene was determined by immunofluorescence analysis of cryosections using the rat-specific α7 monoclonal antibody O26 (Fig. 2). The rat α7 chain was only detected by immunofluorescence in the muscle of transgenic mice (Fig. 2). Immunofluorescence also showed the absence of dystrophin in muscle fibers and the absence of utrophin at NMJs in both transgenic and nontransgenic mdx/utr−/− mice (Fig. 2). The fluorescent specks seen in Fig. 2 upon staining mdx/utr−/− muscle with mouse antidystrophin, antiutrophin, and anti-α7 integrin antibodies are visible in the absence of primary antibody and are due to residual staining with secondary anti–mouse antibody.
The alternative spliced form of the β1 integrin chain, β1D, is expressed in differentiated skeletal and cardiac muscle (van der Flier et al. 1995; Zhidkova et al. 1995; Belkin et al. 1996). Compared with the β1A, β1D may form stronger linkages between the cell cytoskeleton and extracellular matrix (Belkin et al. 1997). Immunofluorescence analysis showed β1A levels were elevated in fibers of mdx/utr−/− mice compared with wild-type and mdx animals. This is indicative of muscle that is not fully differentiated. In contrast, α7BX2-mdx/utr−/− mice had normal levels of β1A integrin. Immunofluorescence and Western blot analysis showed that mdx and mdx/utr−/− mice have more cell surface β1D chain than wild-type mice. This increase in β1D coincided with an increase in endogenous α7 chain in nontransgenic mdx and mdx/utr−/− mice as well as total α7 in α7BX2-mdx/utr−/− mice. The α7BX2-mdx/utr−/− mice also had an additional 1.5-fold more β1D compared with mdx/utr−/− mice (Fig. 3 and Fig. 4 C). Thus, an increase in the α7BX2β1D integrin is promoted by increased expression of the α7 transgene.
As reported previously, mdx mice express approximately twofold more α7 integrin mRNA than wild-type controls (Hodges et al. 1997). No further increase in α7 protein was detected in the mdx/utr−/− animals. The amount of α7BX2 protein in the α7BX2-mdx/utr−/− mouse hindlimb detected by Western blots was 2.0–2.3-fold greater than the endogenous α7BX2 chain in mdx/utr−/− mice (Fig. 4A and Fig. B). As expected, the levels of α7AX2 were equivalent in the transgenic and nontransgenic mice (Fig. 4A and Fig. B).
α7BX2-mdx/utr−/− Mice Exhibit Increased Longevity and Mobility
Longevity was significantly extended in the α7BX2-mdx/utr−/− transgenic mice (Fig. 5). Kaplan-Meier survival analysis (Kaplan and Meier 1958) of 84 nontransgenic and 43 transgenic mdx/utr−/− mice demonstrated that the observed differences in survival of these populations were statistically significant (P < 0.001). Log-rank (Peto et al. 1977) and Wilcoxon rank-sum tests (Conover 1980) showed that the difference in survival emerged soon after birth and was maintained throughout the observed lifetime of the animals. The mdx/utr−/− mice used in these experiments developed severe muscular dystrophy and 50% died before 12 wk of age. The median age at death of the transgenic mdx/utr−/− mice was 38 wk, a threefold increase over that observed in nontransgenic mdx/utr−/− littermates. These findings were consistent in both male and female mice. The oldest α7BX2-mdx/utr−/− mouse was killed at 64 wk of age.
Compared with mdx mice that exhibit minimal pathology, mdx/utr−/− mice do not maintain weight. Instead, these mice undergo a crisis period that results in weight loss and premature death at 4–20 wk of age (Deconinck et al. 1997b; Grady et al. 1997b). In contrast, α7BX2-mdx/utr−/− transgenic mice did not show sudden weight loss. Animal weight stabilized between 20 and 25 g (Fig. 6). No significant differences were found in the weights of mdx mice compared with α7BX2-mdx mice 3–30 wk of age. Thus, extra α7BX2 chain itself does not promote weight gain.
By 8 wk of age, mdx/utr−/− mice exhibited limited mobility and a waddling gait. In contrast, α7BX2-mdx/utr−/− littermates had highly improved mobility compared with mdx mice. A video that contrasts kyphosis, gait, joint contractures, and mobility in mdx/utr−/− and α7BX2-mdx/utr−/− transgenic mice is accessible at http://www.jcb.org/cgi/content/full/152/6/1207.
Enhanced Expression of the α7BX2 Chain Stabilizes Regeneration in mdx/utr−/− Mice
Nuclei are normally localized along the periphery of myofibers, whereas in regenerating muscle nuclei are centrally located (DiMario et al. 1991). Regeneration is also accompanied by a transient reversion to expression of fetal isoforms of myosin heavy chain (fMyHC) (Matsuda et al. 1983; Saad et al. 1987). Hindlimb sections from 5-, 8-, and 10-wk-old wild-type, mdx, mdx/utr−/−, and α7BX2-mdx/utr−/− mice were stained with hematoxylin and eosin to determine the extent of mononuclear infiltration and centrally located nuclei (Fig. 7 and Table). Immunofluorescence of fMyHC was also determined. Degeneration and regeneration that are characteristic of muscle disease occur earlier in mdx/utr−/− animals compared with mdx mice (Fig. 7 and Table). These results are consistent with the earlier onset of necrosis and cell infiltration reported previously in these animals (Deconinck et al. 1997b; Grady et al. 1997b). The occurrence of central nuclei in α7BX2-mdx/utr−/− mice was similar to that in mdx/utr−/− mice, indicating that enhanced expression of the integrin does not prevent early degeneration and regeneration. Likewise, fMyHC expression was most extensive at 5 wk in the mdx/utr−/− and α7BX2-mdx/utr−/− mice. In contrast, mdx mice exhibited very little fMyHC at 5 wk. At 8 wk, fMyHC was elevated in mdx mice, and at 10 wk it was reduced, indicating that a cycle of degeneration and regeneration was followed by stabilization. The shift from the β1A to β1D chain supports this conclusion. At all ages examined, the extent of fMyHC expression in the α7BX2-mdx/utr−/− animals was intermediate between that found in the mdx and mdx/utr−/− animals. In the 8- and 10-wk-old transgenic mdx/utr−/− mice, fMyHC expression approached that in mdx mice (Fig. 7). This decreased expression of fMyHC in α7BX2-mdx/utr−/− mice paralleled the greater integrity of tissue seen in the 8- and 10-wk-old transgenic animals compared with the mdx/utr (−/−) mice. The extensive mononuclear cell infiltration observed in the mdx/utr−/− mice (89–97% positive fields) was also partially reduced in the α7BX2-mdx/utr−/− animals (Fig. 7). The percentage of fields with cell infiltration was reduced approximately 17 and 19% in 5- and 8-wk-old transgenic versus nontransgenic mdx/utr−/− mice, respectively. This represents a minimal estimate since the areas of cell infiltration were considerably larger in the nontransgenic animals (Fig. 7). Thus, enhanced expression of the α7β1 integrin does not alter the initial degenerative cycle, but once regeneration has taken place, the additional integrin appears to stabilize muscle integrity, reducing muscle pathology.
Kyphosis and Joint Contractures Are Alleviated in α7BX2-mdx/utr−/− Mice
Severe curvature of the spine (kyphosis) in DMD patients and mdx/utr−/− mice is due to a failure of the muscles that would normally support the spinal column (Oda et al. 1993). X-ray images showed that both kyphosis and rib cage compression were markedly reduced in α7BX2-mdx/utr−/− mice compared with mdx/utr−/− littermates (Fig. 8). This was confirmed by whole body MRI which visualized not only the tissues surrounding the spinal column, but bundles of muscle fibers, the heart, lung, and other soft tissues. The reduction in kyphosis promoted by the enhanced expression of integrin in the α7BX2-mdx/utr−/− animals is likely a major factor in their survival. Kyphosis results in the diaphragm being pushed forward, compromising lung capacity and diaphragm function, and thereby contributing to cardiopulmonary failure. A partial reduction of kyphosis may therefore have dramatic effects on survival.
A hallmark of diseased musculature is the failure to extend limb muscles, resulting in joint contractures. Hindlimb joint contractures are conspicuous in mdx/utr−/− mice but are markedly reduced in the α7BX2-mdx/utr−/− mice (Fig. 9). The reduction in hindlimb joint contractures allows the mice to have greatly improved mobility.
Structural Changes in the NMJs of α7BX2-mdx/utr−/− Mice
The NMJs in utr−/− mice exhibit a significant reduction in the number of synaptic folds and density of AChRs (Deconinck et al. 1997a; Grady et al. 1997a). This is exacerbated in mdx/utr−/− mice which show even greater reduction in postsynaptic folding and AChR density (Deconinck et al. 1997b; Grady et al. 1997b). The postsynaptic plate of the NMJ in the mdx/utr−/− mice appears en face as discrete boutons rather than as a continuous folded structure (Grady et al. 1997b; Rafael et al. 2000).
Since the α7β1 integrin is normally found at NMJs (Martin et al. 1996) and participates in the clustering of AChRs in C2C12 cells (Burkin et al. 1998, Burkin et al. 2000), we compared the structure of NMJs from 5-wk-old wild-type, mdx/utr−/−, and α7BX2-mdx/utr−/− mice (Fig. 10). Longitudinal sections of hindlimb muscles were stained with rhodamine-labeled α-bungarotoxin, and images of en face sections of the postsynaptic membrane were analyzed. Fluorescence staining of the NMJs of mdx/utr−/− mice appeared less intense compared with wild-type mice and showed discrete boutons. In contrast, most NMJs from α7BX2-mdx/utr−/− mice appeared more continuous. The postsynaptic membrane in longitudinal sections prepared from sternomastoid muscles was examined by EM. The normal folded morphology of the postsynaptic membrane is markedly reduced in the mdx/utr−/− animals, as reported previously (Deconinck et al. 1997b; Grady et al. 1997b). As seen in Fig. 10, postsynaptic folding was partially restored in the α7BX2-mdx/utr−/− mice. Thus, enhanced levels of the α7β1 integrin help maintain the normal structure of the NMJ.
Our results demonstrate for the first time that enhanced expression of the α7β1 integrin can alleviate the development of muscular dystrophy and significantly extend longevity. Mice lacking both dystrophin and utrophin were used in this study because in the absence of both proteins direct substitution of dystrophin with utrophin is precluded. This results in the development of severe muscular dystrophy and premature death, symptoms that closely resemble those seen in DMD (Deconinck et al. 1997b; Grady et al. 1997b).
The α7BX2-mdx/utr−/− mice reported here have ∼2.0–2.3-fold more α7BX2 chain than their nontransgenic littermates. The β1D chain, partner to α7, is also increased in the α7BX2 transgenic mice. The increased levels of α7β1 integrin led to a threefold extension in median survival time, markedly improved mobility, and reduced kyphosis and joint contractures in the transgenic mdx/utr−/− mice. Kaplan-Meier survival analysis of the transgenic and nontransgenic mdx/utr−/− mice shows that the extension of longevity due to expression of the transgene is statistically significant and is evident early and throughout the life of the animals.
The survival times of the mdx/utr−/− mice in these experiments differ slightly from those reported previously. The original reported longevity of the mdx/utr−/− mice bred from the mdx/utr+/− line used to produce the animals in our experiments was 4–14 wk (Grady et al. 1997b). More recently, a life span of 4–20 wk has been reported (Grady et al. 1999) and occasional longer living mice have been noted (Sanes, J.R., personal communication). We too have noted some “outliers”: 6 of 84 mdx/utr−/− mice survived beyond 22 wk, with the oldest mouse dying at 36 wk of age. The transgenic and nontransgenic mice with extended life spans were reevaluated for expression of dystrophin and utrophin by PCR and immunofluorescence and were again found deficient in both. Nevertheless, α7BX2-mdx/utr−/− mice are clearly distinct in longevity, mobility, and histology from nontransgenic littermates. The median life span of the α7BX2-mdx/utr−/− mice was 38 wk, whereas the median life span for those not receiving the transgene was 12 wk.
Although the mechanism by which enhanced expression of the α7β1 integrin alleviates the development of the dystrophic phenotype is not currently understood, multiple effects that result from additional α and β integrin chains are possible.
Expression of the β1D chain is restricted to differentiated skeletal and cardiac muscle (van der Flier et al. 1995; Zhidkova et al. 1995; Belkin et al. 1996, Belkin et al. 1997). In contrast, the β1A chain is present in a wide variety of cell types including myogenic precursor cells. The β1D cytoplasmic domain may act to arrest the progression of myoblast proliferation and alter subcellular localization and affinity of α7β1 for its ligand and the association of the α7β1 with the cell cytoskeleton (Belkin et al. 1997). Therefore, increased β1D expression in α7BX2 transgenic mice may increase the interaction between the extracellular matrix, sarcolemma, and the cell cytoskeleton, stabilizing muscle integrity. Moreover, β1A, characteristic of nonmuscle cells and undifferentiated muscle, is increased in mdx/utr−/− and decreased in the transgenic mdx/utr−/− animals. The shift from β1A and increased β1D reflects fewer mononuclear cell infiltrates and increased stability of muscle fibers in the rescued mice.
The α7BX2 integrin chain is normally concentrated at neuromuscular and myotendinous junctions (Martin et al. 1996) and at intrafascicular junctions (Duxson, M., personal communication). In patients with DMD and in mdx and mdx/utr−/− mice, endogenous expression of the α7 integrin protein is increased, and the α7BX2 isoform is also found extrajunctionally (Hodges et al. 1997). This increase in expression and redistribution of α7β1 integrin in dystrophic mice is also seen with utrophin that is normally confined to NMJs (Matsumura et al. 1992). Immunolocalization of integrin encoded by the rat α7 transgene detected with anti–rat α7 antibodies shows that the rat α7 protein is also distributed more globally in the α7BX2-mdx/utr−/− animals. Enhanced expression of the integrin therefore contributes to the mechanical integration and stability between muscle fibers and at their junctional sites. Other possible mechanisms may also underlie how the α7β1 integrin rescues mdx/utr−/− mice.
Whereas the MCK promoter drives transcription in skeletal and cardiac muscle (Donoviel et al. 1996), enhanced expression of the α7β1 integrin in the heart may also contribute to the rescue of these animals. However, expression of utrophin in skeletal muscle, but not cardiac muscle, of mdx/utr−/− mice increased survival and reduced pathology (Rafael et al. 1998). These observations suggest that the loss of skeletal muscle integrity is the major factor in the development muscle pathology in mdx/utr−/− mice.
The role of the α7β1 integrin in the formation of the postsynaptic membrane (Burkin et al. 1998, Burkin et al. 2000) suggests that increased integrin expression may enhance the development and stability of the NMJs. Dystrophin and utrophin are also concentrated at the postsynaptic membrane and mdx, utr−/−, and mdx/utr−/− mice show progressive alterations of the ultrastructure of these sites (Deconinck et al. 1997b; Grady et al. 1997b). Whereas wild-type and utr−/− mice have NMJ endplates that are highly folded and continuous, mdx and mdx/utr−/− mice show discontinuous NMJs that are described as discrete “boutons” (Grady et al. 1997a,Grady et al. 1997b; Rafael et al. 2000). Whereas both mdx and utr−/− mice show a reduction in the number of synaptic folds compared with wild-type mice, mdx/utr−/− mice show even fewer synaptic folds (Deconinck et al. 1997b; Grady et al. 1997b). Transgenic expression of the α7BX2 chain appears to maintain the normal structure of the postsynaptic membrane in mdx/utr−/− mice.
In the absence of dystrophin, there is an increase in total muscle calcium (Bertorini et al. 1982) and an elevation of intracellular calcium ([Ca2+]i) in isolated dystrophic myofibers (Turner et al. 1988). These increases have been attributed to leaky calcium channels in dystrophic muscle compared with normal muscle. The [Ca2+]i increase may activate Ca2+-dependent proteolysis and increase muscle degeneration (Denetclaw et al. 1994). [Ca2+]i levels are also regulated by signaling through the α7β1 integrin (Kwon et al. 2000), suggesting that this integrin may contribute to the maintenance of calcium levels in myofibers. If so, the transgenic expression of the α7BX2 chain may regulate the activity of calcium channels, stabilizing [Ca2+]i levels in mdx/utr−/− myofibers and reducing Ca2+-dependent proteolysis and muscle degeneration.
Enhanced expression of the α7 integrin may contribute to additional changes in the expression of other proteins, both within the cell and in the extracellular matrix. For example, matrix stability or modeling may potentiate both mechanical and signal transduction capacities of muscle (Colognato et al. 1999). This dual role for the integrin is consistent with analyses of α7−/− mice. The myotendinous junctions of fast fibers are compromised in α7-deficient mice (Mayer et al. 1997). These myofibers also exhibit a partial shift from β1D to β1A integrin and activation of the c-Raf-1/mitogen-activated protein kinase 2 signaling pathway. These changes cause a reduction of integrin-dependent association of fibers and the basal lamina, contributing to the dystrophy that develops in these mice (Saher and Hildt 1999). As shown here, increased α7 chain leads to increased β1D.
A broad phenotype is seen in children with congenital muscular dystrophies that arise from mutations in the α7 gene (Hayashi et al. 1998). These patients exhibit congenital myopathy, delayed motor milestones, and severe impairment of mobility. These phenotypes are consistent with a role for α7β1 integrin in the formation and stability of the postsynaptic membrane, myotendinous junctions, and overall stability of muscle integrity.
Since enhanced expression of the α7β1 integrin can alleviate many of the symptoms of severe muscular dystrophy in mdx/utr−/− mice, it appears that the integrin- and dystrophin-mediated linkage systems between myofibers and the extracellular matrix are in many ways functionally complementary mechanisms. As such, the enhanced expression of the α7β1 integrin may be a novel approach to alleviate DMD and treat α7 integrin congenital muscular dystrophies. Moreover, increasing integrin levels may prove effective in reducing the development of other muscular dystrophies and cardiomyopathies that arise from compromised expression of other components of the dystrophin glycoprotein complex.
This paper is dedicated to the memory of Dr. Kiichi Arahata, a distinguished scientist who taught us much about muscle and its diseases.
We gratefully acknowledge the generous gift of the mdx/utr+/− mice provided by Dr. Joshua Sanes (Washington University) and the MCK promoter and advice kindly provided by Drs. Stephen Hauschka and Jean Buskin (University of Washington). We thank Dr. Jonathan Henry (University of Illinois) for his helpful discussions of the X-ray and MRI analyses and Mr. Eric Chaney for his skillful assistance with the EM. Mice were generated at the University of Illinois Transgenic Animal Facility under the direction of Dr. Matt Wheeler, with the expert technical assistance of Ms. Melissa Hentges. MRI data were obtained at the University of Illinois Medical Resonance Imaging Center under the direction of Dr. Paul Lauterbur, with the technical assistance of Ms. Sherrie Frydenger.
Supported by the Muscular Dystrophy Association and the National Institutes of Health.
The online version of this article contains supplemental material.
Abbreviations used in this paper: AChR, acetylcholine receptor; DMD, Duchenne muscular dystrophy; fMyHC, fetal myosin heavy chain; MCK, muscle creatine kinase, MRI, magnetic resonance imaging; NMJ, neuromuscular junction.