|Home | About | Journals | Submit | Contact Us | Français|
The dystrophin-glycoprotein complex maintains the integrity of skeletal muscle by associating laminin in the extracellular matrix with the actin cytoskeleton. Several human muscular dystrophies arise from defects in the components of this complex. The α7β1-integrin also binds laminin and links the extracellular matrix with the cytoskeleton. Enhancement of α7-integrin levels alleviates pathology in mdx/utrn−/− mice, a model of Duchenne muscular dystrophy, and thus the integrin may functionally compensate for the absence of dystrophin. To test whether increasing α7-integrin levels affects transcription and cellular functions, we generated α7-integrin-inducible C2C12 cells and transgenic mice that overexpress the integrin in skeletal muscle. C2C12 myoblasts with elevated levels of integrin exhibited increased adhesion to laminin, faster proliferation when serum was limited, resistance to staurosporine-induced apoptosis, and normal differentiation. Transgenic expression of eightfold more integrin in skeletal muscle did not result in notable toxic effects in vivo. Moreover, high levels of α7-integrin in both myoblasts and in skeletal muscle did not disrupt global gene expression profiles. Thus increasing integrin levels can compensate for defects in the extracellular matrix and cytoskeleton linkage caused by compromises in the dystrophin-glycoprotein complex without triggering apparent overt negative side effects. These results support the use of integrin enhancement as a therapy for muscular dystrophy.
CELLS CAN RECOGNIZE AND INTERACT with their environment through membrane receptors specific for different extracellular matrix proteins. The heterodimeric integrins are evolutionarily conserved receptors for matrix proteins found in all metazoans (37). In skeletal muscle, the association of the extracellular matrix with the cytoskeleton is essential for maintaining the integrity of the sarcolemma, sustaining appropriate myofiber architecture, and correctly transmitting the forces produced by muscle contractions (5, 42). The α7β1-integrin is part of a linkage system that maintains such transmembrane associations in skeletal muscle (10, 66). It binds to laminin in the basal lamina that surrounds muscle fibers. Within muscle fibers, the integrin associates with the subsarcolemmal cytoskeleton network through focal adhesion complexes (5). The α7β1-integrin is distributed along the sarcolemma at costameres, and it is also concentrated at neuromuscular and myotendinous junctions (2, 48). This localization suggests that the integrin has a pivotal role in maintaining these specialized structures and their functions in skeletal muscle. Congenital myopathies caused by mutations in the human integrin α7-gene (ITGA7) confirm the importance of the α7β1-integrin in maintaining normal skeletal muscle physiology (34). Secondary deficiencies in the integrin are also common in patients with other muscular dystrophies and myopathies of unknown etiology and likely contribute to the pathologies that develop (54, 55).
The dystrophin-glycoprotein complex also binds laminin and has similar roles in skeletal muscle as the integrin transmembrane linkage system (14). Mutations in the genes encoding components of the dystrophin complex result in different types of muscular dystrophies, including Duchenne muscular dystrophy (DMD) (18). The mdx mouse, a genotypic model of DMD, has been used extensively to study this disease and possible therapies. However, unlike DMD patients, the mdx mouse has a mild dystrophic phenotype and near-normal life span (8, 15). Utrophin, a mammalian homolog of dystrophin, is increased in DMD patients and mdx mice and has been hypothesized to compensate in part for the loss of dystrophin (44, 56, 71). The mdx/utrn−/− mice lacking dystrophin and utrophin exhibit progressive muscular dystrophy, markedly reduced mobility and life span, and severe abnormalities at the neuromuscular and myotendinous junctions (20, 30). Since the pathology of skeletal and cardiac muscle of mdx/utrn−/− mice more closely resembles that in DMD patients than the pathology of mdx mice, the doubly deficient mouse is a useful and appropriate model for developing and testing new therapies for muscular dystrophy (4, 12, 13, 27, 31).
In addition to increased utrophin expression, DMD patients and mdx mice also have more α7β1-integrin (36, 74). Other cytoskeleton-associated proteins in integrin focal adhesions, such as vinculin and talin, are also elevated in mdx mice (44). This increase in integrin and other components of adhesion complexes led us to suggest that the integrin and dystrophin complexes have overlapping roles in linking the extracellular matrix and cell cytoskeleton (36). Thus increasing the amount of integrin may functionally compensate for the absence of dystrophin in DMD patients. This hypothesis was tested and validated by showing that transgenic expression of α7 alleviates the development of severe muscular dystrophy in mdx/utrn−/− mice (12, 13). Furthermore, in the absence of both the integrin and dystrophin linkage systems (mdx/α7−/− or gmi mice), extreme severe muscular dystrophies develop (1, 32, 58). These results confirm that the α7β1-integrin and dystrophin complexes are functionally complementary and important in maintaining skeletal muscle integrity. The beneficial effects of increasing integrin in mdx/utrn−/− mice include increased longevity, maintaining the structure of the myotendinous and neuromuscular junctions, reduced apoptosis and kyphosis, enhanced regeneration, and hypertrophy (12, 13). Thus exploring the therapeutic potential of increasing α7β1-integrin in dystrophin deficient muscle remains promising.
As a next step in evaluating an integrin-mediated therapy for DMD, we have generated α7β1-integrin-inducible C2C12 myoblasts and myotubes using a tetracycline-inducible system and evaluated the effects of increasing α7 on cellular functions and differentiation. We have used Affymetrix microarrays to explore whether increasing the amount of integrin affects the transcription profile of skeletal muscle in vitro and in vivo. A threefold increase in α7 levels in C2C12 cells promoted their adhesion to laminin, stimulated proliferation, decreased apoptosis when challenged with staurosporine, and did not affect differentiation. Moreover, neither threefold more integrin in C2C12 cells, nor an eightfold increase of integrin in skeletal muscle in vivo, altered normal transcription. Likewise, no overt toxic effects were observed. Therefore enhancing α7-integrin expression may be useful as a therapy for muscular dystrophies characterized by compromised dystrophin complexes.
The cDNA encoding the rat integrin α7BX2 isoform was amplified from MCKα7BX2 plasmid (13) with pfu polymerase (Stratagene, La Jolla, CA) using the following conditions: 95°C for 4 min, followed by 35 cycles of 95°C for 1 min, 65°C for 30 s, 72°C for 1 min, and a final extension at 72°C for 10 min. The forward primer (5′-ATGAATTCTCCCATGGCCAGGATTCCGAG-3′) and reverse primer (5′-TATCTAGAGCGAATTGGGTACACTTACCTG-3′) were used. The amplified α7BX2 was cloned into pcDNA4/TO vector of the T-Rex system (Invitrogen, Carlsbad, CA), and the sequence of the final plasmid was verified.
Protocols for animal use were approved by the Institutional Animal Care and Use Committee, University of Illinois at Urbana-Champaign. The α7-integrin transgenic mice used in this study were previously described (7). Weights of 30 animals of each sex were measured weekly for up to 30 wk. Five-week-old female transgenic mice and their wild-type controls (SJ6/C57BL6) were used for microarray studies. Animals were euthanized by CO2 asphyxiation, and muscles were immediately dissected and snap frozen in liquid nitrogen.
C2C12 and L8E63 cells were cultured as previously described (9). Cells were transfected with linearized plasmids by using lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. Stably transfected cells were selected with 7.5 μg/ml blasticidin and 500 μg/ml zeocin and were analyzed by immunofluorescence. Clones of the Tetα7-C2C12 cells with the lowest basal level and the highest expression level of rat α7 inducible by 1 μg/ml tetracycline were used for further studies.
Mouse monoclonal antibody O26 was used at a concentration of 10 μg/ml to selectively detect rat α7-integrin by immunofluorescence (9). Polyclonal rabbit antibody against the α7B cytoplasmic domain (CDB347) was used to recognize both mouse and rat α7-integrin (67). Monoclonal antibody O5 and CDB347 were used in Western blots to detect rat and total integrin α7, respectively (67). Rabbit polyclonal antibody against the cytoplasmic domain of the β1D integrin chain was generously provided by Dr. W. K. Song (Department of Life Science, Kwangju Institute of Science and Technology, Kwangju, Korea). Monoclonal antibody against myosin heavy chain (MF20) was used to determine the fusion index. Rabbit anti-caspase-3 antibody (Cell Signaling Technology, Danvers, MA) was used to detect apoptotic myoblasts. Fluorophore and horseradish peroxidase-labeled secondary antibodies were purchased from Jackson ImmunoResearch (West Grove, PA).
Ten-micrometer sections of gastrocnemius muscle from 5-wk-old animals were frozen in liquid nitrogen-cooled isopentane (Sigma, St. Louis, MO), were cut by using a Leica CM1900 series cryostat (Nussloch, Germany), and were placed on microscope slides (Surgipath, Richmond, IL). Sections were fixed in 100% acetone at −20°C for 1 min, were rinsed in tap water for 10 min, and were stained with hematoxylin and eosin using standard histological procedures. Fiber cross-sectional areas were measured by using the advanced measurements component of Open-Lab software (Improvision, Lexington, MA). The cross-sectional areas of 1,000–1,200 fibers from gastrocnemius muscle of three 12-wk-old animals per genotype were measured, and the distributions were displayed using SigmaPlot (Systat Software, San Jose, CA).
To detect rat α7-integrin, live cells were incubated sequentially with 10 μg/ml O26 antibody and secondary antibody, each for 1 h at 37°C, with extensive washing in between, and were then fixed in ice-cold methanol for 5 min and rehydrated in PBS for 30 min. For myosin heavy chain staining, myotubes were fixed in ice-cold methanol for 5 min, rehydrated in PBS for 30 min, and incubated with MF20 antibody in 1% horse serum-PBS and secondary antibody, each for 1 h. For staining of active caspase-3, cells were fixed in ice-cold methanol for 5 min, washed in PBS, and incubated with a 1:100 dilution of primary antibody for 1 h, followed by secondary antibody for 1 h. Coverslips were mounted by using Vectashield (Vector Labs, Burlingame, CA). Tissue sections were fixed in −20°C acetone for 1 min, rehydrated in PBS for 10 min, and blocked in PBS containing 10% horse serum for 30 min. Primary and secondary antibodies in PBS containing 1% horse serum were applied sequentially, each for 1 h, followed by extensive washings. Immunofluorescent images were acquired using a Leica DMRXA2 microscope equipped with an AxioCam digital camera (Zeiss, Thornwood, NY) and analyzed with the use of OpenLab software. Immunofluorescence analyses of α7 were performed on cells used in all experiments.
Cells were washed once and harvested in PBS containing 2 mM PMSF. Gastrocnemius and soleus muscles were powdered in liquid nitrogen before extraction. Proteins were extracted and quantified as previously described (12, 13). Equal amounts of proteins were separated on 8% polyacrylamide-SDS gels and were transferred to nitrocellulose membranes. Blocked membranes were incubated with the respective primary antibodies in 2% nonfat milk for 1 h. Horseradish peroxidase-conjugated secondary antibodies were used to detect bound primary antibodies. Immunore-active protein bands were detected using an enhanced chemiluminescence kit (Amersham, Arlington Heights, IL). Bands were quantified by using ImageQuant software. Data were obtained from three independent experiments.
Cell adhesion was measured as previously described (19). Briefly, Costar 24-well polystyrene plates (Sigma) were coated with substrates in 0.1% BSA in PBS for 2 h at room temperature. The wells were then washed three times with PBS and were blocked with 1% BSA in PBS for 2 h. DMEM (250 μl) with 0.1% BSA was added to the wells and was preincubated for 1 h at 37°C. Cells were washed once with PBS, detached by 0.2 g/l EDTA in PBS, pelleted, and resuspended in DMEM containing 1% BSA. 5 × 104 cells were added to each well and incubated at 37°C for 30 min. Nonadherent cells were removed by gentle rinsing with PBS, and the attached cells were fixed with 4% paraformaldehyde-1% glutaraldehyde in PBS and stained with 1% toluidine blue. After extensive rinses in distilled water, the stained cells were counted using a Nikon inverted microscope equipped with an ocular micrometer. For each well, at least 10 randomly selected fields were measured. Data from three experiments in triplicates are presented.
Tetα7-C2C12 cells were grown for 36 h with or without tetracycline, then detached and seeded in growth medium containing different concentrations of fetal bovine serum (FBS). Cell numbers in triplicate samples were measured daily. Doubling times were calculated from linear regression analysis of cell numbers versus time. Data were derived from three independent experiments.
Tetα7-C2C12 cells were grown for 36 h with or without tetracycline and were starved for 48 h in serum free growth medium to synchronize them at G1/G0. Growth medium containing 20% FBS with or without tetracycline was then added. Cells were collected and fixed in suspension with 70% ethanol 0, 8, 24, and 32 h later. Cells were treated with RNase and stained with propidium iodide. Flow cytometry was carried out using a Coulter XL-MCL flow cytometer (Beckman Coulter, Fullerton, CA). The percentages of cells in each phase of the cell cycle were determined by using ModFit software (Verity Software, Topsham, ME). The data from three experiments in triplicates are presented.
Tetα7-C2C12 cells were cultured on laminin-coated glass coverslips for 36 h with or without tetracycline. Staurosporine (0, 1, or 2 μM; Cell Signaling Technology) was added for 3 h to induce apoptosis. After immunostaining, the numbers of active caspase-3-positive cells and the total numbers of cells were counted in 100 randomly selected ×40 fields. Apoptosis assays were done in triplicate in three experiments.
RNAs were extracted from cells and gastrocnemius and soleus complex muscles. Triplicate cultures of Tetα7-C2C12 myoblasts were grown with or without tetracycline for 36 h and were extracted using TRIzol (Invitrogen). RNA pellets were resuspended in diethyl pyrocarbonate-treated water and subjected to a cleanup protocol using an RNeasy minikit (Qiagen, Valencia, CA). The RNA was prepared from 18 mice of each genotype. Triplicate samples for each genotype were made by pooling the RNA from 6 mice. RNA quality was examined using a bioanalyzer (Agilent Technologies, Santa Clara, CA), denaturing gel analysis, and OD260/OD280 ratios. Total RNA was further processed for use on Affymetrix mouse 430 2.0 microarray slides (Affymetrix, Santa Clara, CA) at the University of Illinois Keck Center for Comparative and Functional Genomics. Affymetrix CEL files were analyzed by using GeneSifter (VizX Labs, Seattle, WA), and the probe level analysis algorithm GC-RMA (robust multichip average with adjustment for GC content of probes) was applied. These data have been deposited in the National Center for Biotechnology Information Gene Expression omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) and are accessible through GEO series accession numbers GSE 8312 and GSE 8313. Microarray statistical analysis was performed using the t-test with Benjamini and Hochberg [false discovery rate (FDR)] multiple-testing correction for two group comparisons. Except where indicated, probe sets were defined as differentially expressed if the Benjamini and Hochberg (FDR) corrected P was <0.05 and the absolute value of the average fold difference was ≥2. The correlation coefficient of arrays between groups was calculated using bioconductor and R software packages (28, 69).
RNA extracted from cells or skeletal muscle was treated with RNase-free DNase I (Fisher Scientific, Pittsburgh, PA) for 10 min at room temperature to remove genomic DNA. cDNA was synthesized by using RETROscript kits (Ambion, Austin, TX) with random decamers using 100 ng of cDNA as templates. GAPDH was used as an internal control. Primer sequences for each gene are listed in Table 1. PCR reactions were performed by using the following conditions: start at 94°C for 2 min, cycle at 94°C for 30 s, 55°C to 61°C (depending on the primer used) for 30 s, 72°C for 45 s, and a final extension at 72°C for 10 min. Cycle numbers were determined to be within the exponential amplification range for each gene. PCR products were electrophoresed in 1.5% agarose gels. Intensities of amplicon bands were quantified using ImageQuant software and were normalized to GAPDH control bands from the same samples. To provide independent confirmation, RNAs for RT-PCR and microarray analyses were purified from separate animals.
All averaged data are presented as means ± SE. Comparisons between two groups were performed by unpaired t-test. One-way ANOVA followed by Tukey’s posthoc analysis was used if more than two groups were compared. Statistics were computed by using SYSTAT software (Systat Software, San Jose, CA). Differences were considered significant at P < 0.05.
Alleviation of severe muscular dystrophy in mdx/utrn−/− mice by α7β1-integrin demonstrates the overlapping functions between the integrin and the dystrophin glycoprotein complexes (12, 13). To further explore developing a therapy for muscular dystrophy on the basis of enhancing α7-integrin levels, we have examined the functional and transcriptional consequences of increasing α7β1-integrin in muscle cells. Mouse C2C12 cells were used to derive α7-integrin tetracycline-inducible clones of myogenic cells that express the rat α7-integrin. RT-PCR confirmed expression of the rat α7-integrin transgene after tetracycline induction, whereas noninduced cells did not express rat α7 RNA (Fig. 1A). Western blot analysis of myoblasts and myotubes induced with tetracycline detected the rat integrin, whereas no transgene-derived protein was observed in noninduced cells (Fig. 1, B and C). A threefold increase in total integrin was detected in myoblasts and a 1.5-fold increase was present in myotubes (Fig. 1B). To be functionally active, the integrin must be targeted to the cell membrane. Immunofluorescence staining of live cells with anti-rat α7 monoclonal antibody (O26) demonstrated that induced rat α7-integrin was localized to the cell membrane in both myoblasts and myotubes (Fig. 1C). Thus the induced integrin localized to the cell membrane to form α7β1-integrin focal adhesions.
To test whether rat α7-integrin expressed in C2C12 cells is functional, we determined whether the induced α7-integrin promotes adhesion to different extracellular matrix proteins. Induced and noninduced Tetα7-C2C12 cells were detached and plated in complete medium on BSA-, laminin-, or fibronectin-coated 24-well plates. The numbers of cells that adhered after 30 min were determined (Fig. 2A). Increased integrin promoted cell adhesion on laminin and decreased adhesion on fibronectin (Fig. 2B). Virtually no cells adhered to BSA-coated wells, which confirms that adhesion was substrate specific.
Integrins can function with receptors for growth factors such as EGF and hepatocyte growth factor to regulate cell proliferation (16, 59). We therefore tested whether induced Tetα7-C2C12 cells, with approximately threefold more integrin, have a different proliferation rate. Induced or noninduced Tetα7-C2C12 cells were seeded at the same density in growth medium containing different concentrations of FBS, with or without tetracycline. Wild-type C2C12 cells were used to exclude the possibility that tetracycline itself affects cell proliferation. No differences in doubling times were observed in induced or noninduced cells grown in medium containing 20% or 10% FBS. However, lower concentrations of FBS (0.5%, 1%, 2%, and 5%) slowed cell proliferation of noninduced cells but did so to a lesser extent in cells with elevated integrin levels (Fig. 3A). These results suggest that α7β1-integrin may sensitize myoblasts to growth factors to activate cell proliferation.
To further test whether additional α7β1-integrin renders myoblasts more sensitive to growth factors, sparsely seeded myoblasts were grown for 48 h in serum-free medium, with or without tetracycline, to synchronize the cells at G1/G0 without inducing differentiation. The medium was then changed to contain 20% FBS, with or without tetracycline, and cell cycle progression was determined by flow cytometry. Tetα7-C2C12 cells induced with tetracycline progressed into G2/M faster than noninduced cells (Fig. 3B). This is the first report to show that α7β1-integrin regulates myoblast proliferation and cell cycle progression, and it supports previous findings that myoblasts proliferate more rapidly when grown on laminin (26).
The α7β1-integrin is expressed on replicating myoblasts, quiescent satellite cells, and mature skeletal muscle fibers, and it is involved in myoblast differentiation and maintaining muscle integrity (10). Myoblast fusion and expression of myosin heavy chain were used to monitor whether increased α7β1 affected cell differentiation. C2C12 and Tetα7-C2C12 cells were grown to confluence, at which point differentiation medium with or without tetracycline was added. The percentage of myosin heavy chain-positive cells and the cell fusion index indicate no significant differences due to increased integrin (Fig. 4, A and B).
Cell adhesion regulates proliferation, migration, and differentiation, and it also supports cell survival (63, 75). Integrin-mediated adhesion regulates apoptosis in a variety of cells, including skeletal muscle (43, 68). In laminin-2-deficient congenital muscular dystrophy, the amount of α7β1-integrin is secondarily reduced, leading to a loss of muscle fiber adhesion to the extracellular matrix, apoptosis, and muscle wasting (74). Likewise, α7-transgenic mdx/utrn−/− mice with elevated levels of integrin also have fewer apoptotic nuclei than mdx/utrn−/− mice as measured by terminal deoxynucleotidyl transferase dUTP-mediated nick-end labeling staining (12). We therefore used the Tetα7-C2C12 cells to examine whether integrin can prevent apoptosis in myoblasts in vitro. Tetα7-C2C12 myoblasts were grown in tetracycline to express excess integrin and were then treated with staurosporine. Staurosporine-induced apoptosis in C2C12 cells can be identified by staining for active caspase-3 (35, 50). Apoptotic cells also shrink and can be identified morphologically (Fig. 5A, a, c, and e). Immunofluorescence staining revealed less caspase-3 activity in tetracycline-induced cells compared with noninduced and control
C2C12 cells (Fig. 5A, b, d, and f). The enhanced expression of integrin in myoblasts reduced apoptosis by approximately twofold in cells grown on laminin (Fig. 5B). Thus it appears that enhanced expression of α7β1-integrin can protect myoblasts from apoptosis.
Transgenic mice expressing the integrin α7BX2 subunit, under control of the muscle creatine kinase promoter, were generated as described previously (7). Immunoblot analysis of protein extracts of the gastrocnemius and soleus muscles from α7-transgenic and wild-type control mice revealed an eightfold increase in α7BX2 protein in the transgenic muscle (Fig. 6A). Immunofluorescence staining with the rat-specific O26 monoclonal antibody and a polyclonal antibody against the α7B cytoplasmic domain demonstrated a high level of expression of the rat integrin α7-transgene (Fig. 6B). The overwhelming majority of the integrin was localized to the sarcolemma; however, some cytoplasmic localization was also detected (Fig. 6B) and is likely due to an excess of α7-chain. Hematoxylin and eosin staining revealed no discernable differences between the skeletal muscles of the transgenic mice and their wild-type controls (Fig. 6B). Similarly, no significant differences in body weight (Fig. 6C) or average fiber cross-sectional areas were observed (Fig. 6D), although there was a 12% increase in the percentage of smaller fibers in the transgenic mice. The eightfold more α7-chain expressed in the transgenic skeletal muscle did not produce any overt negative changes under normal physiological conditions.
Integrin α7β1 can interact with signaling molecules and transcription factors such as FHL2 and FHL3 in skeletal muscles (60, 76). Therefore, for integrin overexpression to be an effective therapy, it would be best if it did not disrupt normal cell functions and gene expression. As a global index of that, transcription profiles of noninduced and induced Tetα7-C2C12 cells and of wild-type mice and integrin α7-transgenic mice were determined by using Affymetrix microarray analysis.
Comparison of the data from the noninduced and induced Tetα7-C2C12 cells revealed relatively few differences in their overall transcriptional profiles (Fig. 7A). The correlation coefficient of the induced and noninduced myoblast arrays was 0.996, which demonstrated the high similarity in their transcription profiles. Seven probes representing six genes indicated greater than twofold changes with Benjamini and Hochberg adjusted P values< 0.05 (Table 2). These six genes are involved in cell metabolism (Cth, Ugt1a2, Ank2, and pigt), regulation of apoptosis (Trib3 and Angptl4), and endoplasmic reticulum (ER) overloading responses (Trib3). Among the six genes, only Trib3 had two separate probe sets, and both revealed similar changes in expression levels. This was confirmed by semiquantitative RT-PCR (Fig. 7B). A full list of probe sets that revealed changes in expression of ~1.5-fold, with adjusted P values< 0.05, is in Supplemental Table 1. (The online version of this article contains supplemental data.)
Trib3, the mammalian homolog 3 of Drosophila tribbles, was first identified as an inhibitor of the Akt pathway, and it is highly expressed in liver (22). More recent reports demonstrated that Trib3 expression can be induced by stress-related signals downstream of CHOP (53, 77). Interestingly, CHOP was also increased 1.6-fold in our array analysis. These increases in expression of Trib3, CHOP, and several other genes involved in cellular stress responses may represent a generalized response to the overexpression of membrane receptors and adapt the myoblast to increased α7-chain. However, Trib3 is not normally expressed in adult skeletal muscle, and its expression is dramatically decreased upon differentiation of C2C12 cells (38). Thus this increase of Trib3 transcripts in myoblasts is not likely significant to the response of muscle fibers to increases in α7β1-integrin.
Array analysis of gastrocnemius and soleus muscles from α7-transgenic mice was done to determine the effect of increasing integrin in vivo. Twenty-nine probe sets representing 20 genes reported more than twofold increases in transcription, and three probe sets identified more than twofold decreases in expression of three genes (Table 3). A 36-fold increase in α7-transcripts was found in the transgenic mice. Genes whose expression significantly changed ~1.5-fold are listed in Supplemental Table 2. Similar to the results from the myoblast array analysis, relatively few genes were found to be differentially expressed in the integrin transgenic mice (Fig. 7A). The 0.999 correlation coefficient of the wild-type and α7-integrin transgenic mice array analysis indicates the high similarity of their transcription profiles. Significant changes of several genes noted in the array data (Teme58, pttg1, hspa5, and armet) were confirmed by semiquantitative RT-PCR (Fig. 7B).
Gene ontology analysis revealed that many of the changes in gene expression are related to cell metabolism, including protein synthesis, posttranslational processing, and protein transport. These genes include ribosomal protein S4 (1110033J19Rik), branched-chain amino acid aminotransferase (Bcat2), stromal cell-derived factor 2-like protein 1 (SDF2L1), thyroid hormone receptor interactor 11 (Trip11 or TRIP230), transmembrane protein 58 (Tmem58, similar to rat Scotin), progressive myoclonic epilepsy, type 2 gene-α (Epm2a), disulfide isomerase-associated 6 (Pdia6), coatomer protein complex subunit-α (Copa), and heat shock 70-kDa protein 5 (Hspa5, Grip, or Bip).
Other groups of genes that were differentially transcribed are related to cell division, G protein signaling, and the interferon-inducible p200 family of proteins. Tacc2, Prc1, and pttg1 (securin) are regulators of cell division and showed increased expression. Their relation to the integrin is unclear, but an increase in satellite cell proliferation was noted in α7mdx/utrn−/− mice (12). RGS5 and xpr1 are involved in G protein signaling (3, 17), and their expression was increased in the integrin transgenic mice. Five probes representing three proteins in the interferon-inducible p200 family were significantly increased in the integrin transgenic mice (Table 3). Members of this family have specific functions in skeletal and cardiac muscles (46). Ifi204 promotes myoblast differentiation and is increased following MyoD-dependent differentiation of muscle precursor cells (46). Therefore, the increased expression of these genes may enhance skeletal muscle differentiation in the α7-integrin transgenic mice.
Several expressed sequence tags are also differentially expressed. Among these, increased expression of Armet, an extracellular protein, may represent a modification of the extracellular environment by muscle fibers with enhanced integrin.
Neither β1-integrin gene expression nor the mRNAs encoding other focal adhesion components increased commensurate with the increases in α7-chain in both myogenic cells and skeletal muscle. Immunofluorescence staining also did not show an increase in β1D in the α7-transgenic mice (Fig. 6B). The localization of α7-chain inside the transgenic muscle fibers (Fig. 6B, arrows) and the decreased expression of integrin α4 (ITGA4) transcripts seen in the microarray data indicates the β1-chain may be limiting in the α7-transgenic mice. Normal expression of β1-integrin in α7-knockout mice also suggests that the regulation of transcription of α7 and β1 is independent in skeletal muscle (25, 58), and it explains the lack of changes in β1 mRNA evidenced in our arrays. Since no alterations in transcription of the genes encoding the components of the dystrophin glycoprotein complex were observed in both array analyses, it is likely that the enhanced structural linkage and signaling provided by increased integrin accounts for the alleviation of dystrophic pathology in α7-transgenic mdx/utrn−/−mice (12, 13).
Array analysis of ADAM12 transgenic mice, in which an increase in ADAM12 partially rescued mdx mice, reported an increase of α7-protein but not mRNA in skeletal muscle (41, 51). In both our cell and tissue arrays, no significant change in transcription of any of the ADAM genes was detected, which suggests that α7β1-integrin may function downstream or in concert with ADAM12 in promoting muscle regeneration and preventing muscular dystrophy. This conclusion is also supported by the inability of overexpressing ADAM12 to alleviate the pathology of laminin-2-deficient mice in which integrin α7β1 is secondarily reduced (33).
The array analyses reported in the present study show that few perturbations of transcription result from increasing α7β1-integrin in myoblasts or skeletal muscle and encourage the development of an integrin-mediated therapy for muscular dystrophy.
The α7β1-integrin is a laminin receptor on muscle progenitor cells and mature skeletal muscle fibers. Developmentally regulated expression and alternative splicing of the α7-integrin chain are important for appropriate myogenic cell proliferation, adhesion, migration, and differentiation (19, 78). Alleviation of muscular dystrophy by transgenic overexpression of the integrin in mdx/utrn−/− mice revealed the potential for using it as a therapy for muscular dystrophy (13). This led us to examine the biological effects of enhancing integrin levels in myoblasts and skeletal muscle fibers. Using tetracycline-inducible integrin-expressing myoblasts, we have demonstrated that increasing the amount of the α7β1-integrin improves myoblast adhesion and proliferation and imparts resistance to staurosporine-induced apoptosis but does not interfere with myogenic differentiation. These effects were achieved without markedly altering gene expression.
The alleviation of severe muscular dystrophy in mice by enhanced expression of the α7β1-integrin results from the functional enhancement of muscle fiber adhesion, strengthening of myotendinous and neuromuscular junctions expansion of regenerative capacity, and enhanced signaling, leading to decreased apoptosis and fiber survival (12, 13). As shown here, transgenic overexpression of the α7-chain in wild-type mice did not alter body weight or skeletal muscle histology, nor did it dramatically change the skeletal muscle global gene expression. Thus, increasing α7-integrin is beneficial to muscle and has only minimal adverse effects.
In myoblasts, tetracycline induction resulted in threefold overexpression of α7-integrin, whereas only a 1.5-fold increase was observed in myotubes. The smaller fold increase observed in myotubes is likely due to increased synthesis of endogenous mouse α7-integrin during C2C12 differentiation (67, 79). Furthermore, consistent with previous reports that α7-integrin increases non-muscle-cell adhesion to laminin (19, 23, 72), we confirmed that enhanced expression of integrin α7β1 in C2C12 myoblasts also increases their adhesion to laminin; however, it correspondingly decreases adhesion to fibronectin. Similar effects were reported when α7-integrin was overexpressed in Chinese hamster ovary cells, suggesting that there may be a competition between α7 and other integrin α-subunits for the common β1-chain (72). In support of this hypothesis, a decrease of α4 integrin transcripts was also detected in the α7-integrin transgenic mice. These data suggest a potential competition between different integrin α-chains for their common β1-subunit as a mechanism of regulating myoblast affinity for different matrix proteins.
The increased capacity of cells with enhanced levels of α7β1-integrin to proliferate in low serum concentrations suggests that the integrin may render satellite cells more responsive to limiting concentrations of growth factors in vivo. These quiescent muscle stem cells that are activated by low levels of growth factors may contribute to an increase in the number of precursor cells and enhanced regeneration and repair such as seen in α7-integrin transgenic mdx/utrn−/− mice (12). Similarly, the α7β1-integrin renders myofibers highly sensitive to respond to low levels of agrin, and this promotes clustering of acetylcholine receptors, an early stage in the development of neuromuscular junctions (9, 11).
Increased apoptosis of both myofibers and myoblasts in dystrophic muscle has been reported (49, 65, 73), and caspase-3 activation has been shown to participate in this process. We observed an approximate 50% reduction in the number of caspase-3 activated cells following staurosporine treatment of cells with enhanced expression of α7β1, which suggests that the integrin can function as an antiapoptotic signal in myogenic cells. α7-Integrin transgenic mdx/utrn−/−mice likewise have fewer apoptotic cells than their nontransgenic counterparts (12). In accordance with this hypothesis, increase in integrin does protect against exercise-induced muscle damage (7). Thus, enhancing α7-integrin expression can spare skeletal muscle from injury or apoptosis without compromising normal physiological functions.
Several “complementary” genes have been shown to rescue or alleviate the pathology of dystrophic muscle when their expression levels are increased (24). However, transcriptional changes induced by overexpression of these genes in myoblasts and skeletal muscle have yet to be fully examined. Evaluation of the global changes in transcription resulting from complementary genes in transgenic mice is essential for determining potential side effects of these gene therapies. Here we utilized array profiling to evaluate the overexpression of α7-integrin as a therapy for muscular dystrophy. Neither the threefold increase in α7-protein in myoblasts nor the eightfold increase in mature skeletal muscle significantly altered their respective transcriptional profiles. In both cases, relatively few genes had greater than twofold changes in expression, and the majority of these changes in transcription are likely related to mild ER stress posed by overexpression of the α7-chain. Others are related to cell division, G protein signaling, and the interferon-inducible p200 family of proteins. Among them is RGS5, a marker for pericytes (6), and these cells are myogenic in humans and may be used in stem cell therapy for muscular dystrophy (21). The increased expression of RGS5 in α7-integrin transgenic mice may result in an increase in the number of pericytes and the higher regenerative capacity and reduced pathology seen in α7mdx/utrn−/−mice (12, 13). Similarly, the interferon-inducible p200 family of proteins such as p204 may also promote muscle differentiation in α7-transgenic mice.
Several possibilities may explain the lack of more dramatic changes in the transcription profiles of α7-integrin-overexpressing myoblasts and skeletal muscle. First, α7β1-integrin primarily functions as a receptor, mediating the transsarco-lemma linkage of muscle fibers to the extracellular matrix. Its relatively short cytoplasmic tail suggests it does not have enzymatic activity, and thus it has less potential for initiating signals compared with other receptors (67). Therefore, although enhanced expression of the integrin improves structural linkages through the sarcolemma, it does not necessarily promote major changes in transcription. Accordingly, alleviation of pathology in dystrophic mice by enhancing integrin levels is likely to be largely due to the restoration of the transsarco-lemma linkages and specialized skeletal muscle structures such as the myotendinous and neuromuscular junctions (12, 13).
Second, recent reports of changes in signaling induced by exercise in α7-integrin transgenic wild-type mice provide evidence that the integrin is a mechanosensor in skeletal muscle (7). In α7-transgenic mice, our array analysis also reveals that the genes involved in G protein signaling (RGS5 and xpr) are changed in accordance with the enhanced α7-integrin. Since integrin and G protein signaling regulate one another (64) and both are involved in initiating mechanical signaling (45), the coordinated expression of these genes in skeletal muscle may regulate mechanosignal transduction and adaptation of fibers to contraction and stretch. Hence, much larger changes in transcriptional profiles may result from activation of α7β1-integrin by mechanical loading that is known to promote signaling. Alternatively, excess integrin may promote structural integrity or amplifications of transient signaling that do not result in altered gene expression. Further array analysis of integrin α7-transgenic mice following exercise is needed to determine whether activation of the integrin by mechanical stimuli regulates gene expression.
Third, although the α7-transgenic mice have a 36-fold increase in α7-RNA and an eightfold increase in α7-protein, no major changes were detected in either mRNA or protein levels of its interacting β1-subunit. Thus the amount of α7β1-integrin does not regulate β1-transcription. Likewise, there is no decrease in β1-chain in α7-null mice (36). Similarly, we did not observe changes in mRNA levels of other components of focal adhesion complexes in our array analysis. Limited amounts of β1-chain and/or other components of focal adhesion complexes (focal adhesion kinase, integrin-linked kinase, vinculin, and talin) may restrict signaling by the integrin and result in a lack of transcriptional regulation. Immunofluorescence localization of the α7-chain in transgenic mice revealed some staining within fibers, and this likely reflects limiting amounts of the β1-subunit. Unable to dimerize with the β1-subunit, excess α7-chain remained inside the fibers and likely triggered a mild ER-stress response as revealed in the array analyses. It will be interesting to determine whether enhancing the expression of both the α7- and β1-subunits will result in further regulation of signaling, altered transcription, or a greater alleviation of dystrophic pathology.
It should be noted that transcription is not uniform in all nuclei in muscle fibers; nuclei in proximity to junctional sites transcribe genes encoding proteins specific to those sites (47, 52, 61, 62). Therefore, it is possible that enhanced integrin expression triggers locally restricted transcriptional changes in nuclei near myotendinous and/or neuromuscular junctions, sites where the integrin is highly enriched. In our analysis, such changes may have been obscured by the relatively large amount of total RNA from other portions of the muscle fibers.
Although enhancing integrin levels in wild-type skeletal muscle did not dramatically alter transcription, in the environment of dystrophic muscle, that may not be the case. Dystrophic and healthy skeletal muscle differ with regard to the composition of their extracellular matrix, oxidative condition, sarcoplasmic Ca2+ concentration, protease activity, and other metabolic processes (14, 29, 70), and all of these differences may regulate α7-integrin activation and/or downstream transcription. In fact, an increase in α7-RNA and protein, as well as talin and vinculin, are seen in Duchenne patients and in mdx mice (36, 44). Therefore, expression profiling of α7-transgenic mdx/utrn−/− is needed to test whether increased α7β1-integrin affects transcription in dystrophic muscle and how those altered genes may contribute to alleviating pathology.
Lastly, the α7β1-integrin was recently reported to have tumor suppressor activity in several tissues and mutations in the α7-gene related to tumor formation (57). This interesting finding is consistent with earlier results in which developmentally defective myogenic cells were shown to be transformed and tumorigenic (40) and did not express the H36 antigen (39) [later shown to be α7-integrin (66)], and with the recent finding of a high rate of tumor formation in α7-null mice (DJ Burkin, unpublished observations). Thus, enhancing the levels of α7-integrin may also be useful as a therapeutic approach to cancer with minimal transcription effects.
In summary, the effects of increasing integrin levels in both muscle precursor cells and in skeletal muscle fibers are encouraging with respect to developing integrin enhancement as a therapy for muscular dystrophies. Whereas the integrin can promote myoblast adhesion, proliferation, and resistance to apoptosis, increasing integrin in precursor cells would yield increased numbers and more effective cells to repair dystrophic fibers. In addition, muscle fibers derived from such precursors would have more integrin to strengthen the linkage between the extracellular matrix and cytoskeleton and to diminish apoptosis. The increased proliferative capacity of such myogenic cells may also contribute to a more effective regenerative capacity and counter the depletion of stem cell populations that underlies progressive muscle wasting such as that seen in Duchenne patients. Moreover, increasing integrin levels in myoblasts and skeletal muscle does not promote major changes in gene expression, and therefore an integrin-based therapy is not likely to impart serious negative side effects.
Anti-β1-integrin monoclonal antibody was generously provided by Woo Keun Song (Department of Life Science, Kwangju Institute of Science and Technology, Kwangju, Korea). We thank Eric Chaney and James Mulligan for technical assistance. We also thank Marni D. Boppart, Suzanne E. Berry, Greg Q. Wallace, and Praveen B. Gurpur for helpful discussions.
The present study was supported by grants from the National Institutes of Health (NIH; R01-AG014632) and Muscular Dystrophy Association (to S. J. Kaufman) and NIH Grants P20RR018751 and P20RR15581 (to D. J. Burkin).