Regulatory, Integrative and Comparative Physiology

Exercise promotes α7 integrin gene transcription and protection of skeletal muscle

Marni D. Boppart, Sonja E. Volker, Nicole Alexander, Dean J. Burkin, Stephen J. Kaufman


The α7β1 integrin is increased in skeletal muscle in response to injury-producing exercise, and transgenic overexpression of this integrin in mice protects against exercise-induced muscle damage. The present study investigates whether the increase in the α7β1 integrin observed in wild-type mice in response to exercise is due to transcriptional regulation and examines whether mobilization of the integrin at the myotendinous junction (MTJ) is a key determinant in its protection against damage. A single bout of downhill running exercise selectively increased transcription of the α7 integrin gene in 5-wk-old wild-type mice 3 h postexercise, and an increased α7 chain was detected in muscle sarcolemma adjacent to tendinous tissue immediately following exercise. The α7B, but not α7A isoform, was found concentrated and colocalized with tenascin-C in muscle fibers lining the MTJ. To further validate the importance of the integrin in the protection against muscle damage following exercise, muscle injury was quantified in α7−/− mice. Muscle damage was extensive in α7−/− mice in response to both a single and repeated bouts of exercise and was largely restricted to areas of high MTJ concentration and high mechanical force near the Achilles tendon. These results suggest that exercise-induced muscle injury selectively increases transcription of the α7 integrin gene and promotes a rapid change in the α7β integrin at the MTJ. These combined molecular and cellular alterations are likely responsible for integrin-mediated attenuation of exercise-induced muscle damage.

  • exercise
  • injury
  • repeated bout effect
  • tenascin-C

several classes of cell adhesion molecules provide mechanisms for attaching cells to each other and to their extracellular environment. These include integrins, cadherins, and members of the immunoglobulin and selectin families (23). Integrins are transmembrane heterodimers of noncovalently bound α- and β-subunits. Integrins bind to specific ligands in the extracellular matrix, form clusters, and recruit cytoskeletal and cytoplasmic proteins that stabilize the cell and provide a means for communication between the outside and inside of the cell. Eighteen α-subunits and 8 β-subunits have been characterized, and at least 24 heterodimers have been identified (46).

α7β1 is the predominant integrin in adult skeletal muscle. The α7 subunit determines the specificity of ligand binding to laminin in the basal lamina surrounding individual muscle fibers, whereas the β1 subunit participates in linkage with actin via several subsarcolemmal proteins, including α-actinin, talin, vinculin, paxillin, and tensin. Several other integrin α-subunits are expressed in skeletal muscle in association with β1, including α1, α3, α4, α5, α6, α9, α10, α11, and αv (15, 29, 30, 39, 46). Overlapping functions of α-subunits have made it difficult to interpret the role of individual integrins in muscle function; however, α4, α5, α6, α7, and αv are believed to be the major integrin subunits involved in muscle differentiation (for a review, see Ref. 33). Expression of the α5 and α6 integrins are decreased and α7 expression is increased during myotube formation (5, 7, 44). This coincides with developmentally regulated changes in matrix composition in the basement membrane as well as alternative RNA splicing that generates new α7 and β1 cytoplasmic domains (28, 45). Transfection of HEK-293 cells to express an α7-chain also results in the downregulation of α1, α3, α5, and α6 integrin (42). The functions of α9, α10, and α11 subunits in mature skeletal muscle remain unknown (29, 30, 39). In mice lacking the α7 integrin, β1 chains associate with the α5 subunit in a fibronectin-rich environment (36); however, muscle integrity is still compromised in these mice. Mutations in the human α7 gene also have been shown to cause congenital myopathies (20). Therefore, the α7β1 integrin appears to be critical for skeletal muscle development and function. The α7β1 integrin is present throughout the sarcolemma, and it is enriched at myotendinous (2) and neuromuscular (32) junctions. The localization of the α7β1 integrin at the myotendious junction is unique as α1, α3, α5, α6, and αv subunits are not concentrated in this region (2). Mice deficient in the α7 subunit develop a muscular dystrophy that is most apparent at the myotendinous junction (34, 36). Patients with Duchenne muscular dystrophy and dystrophic mdx mice exhibit increased amounts of α7β1 integrin (21) and transgenic overexpression of the α7 integrin chain maintains myotendinous and neuromuscular junction integrity, ameliorates disease, and extends longevity in mice with a severe form of muscular dystrophy (8). Direct muscle injury or transection of the soleus muscle results in regeneration that includes increased expression of the RNA splice variants that encode the alternative cytoplasmic (α7A and α7B) and extracellular (X1 and X2) α7 chain domains (24, 45). As a result, α7A and α7X1 isoforms are predominantly increased during the dynamic adhesion phase of muscle repair (24). These studies indicate that the α7β1 integrin responds to skeletal muscle injury and stabilizes muscle-tendon interactions in critical weight-bearing areas of adult skeletal muscle.

A single bout of predominantly eccentric or muscle lengthening exercise results in skeletal muscle damage and soreness (37). However, a second bout of similar exercise results in minimal muscle damage, and this protection from injury may persist for up to 6 mo (38). The molecular mechanisms underlying this protective adaptation, or “repeated bout” effect, have not been fully elucidated. Neural, mechanical, and cellular (increased sarcomere length, inflammatory responses) adaptations have been proposed to be responsible for the repeated bout effect (35). These include changes in motor unit activation between repeated bouts, increased muscle stiffness, and increased sarcomere length and inflammatory responses, respectively. One mechanism for increasing resistance to mechanical reinforcement is enhanced localization and/or expression of cytoskeletal proteins, such as desmin, following eccentric exercise (3). However, conflicting evidence suggests that desmin is not responsible for the repeated bout effect, since desmin null mice exhibit less disruption of myofibrils (41). Other cytoskeletal proteins or molecules associated with the cytoskeleton, such as integrins, may be responsible for the adaptations that occur in response to a single bout of eccentric exercise.

We recently demonstrated that the amount of α7β1 integrin increases in response to downhill running exercise and that transgenic overexpression of this integrin prevents exercise-induced muscle damage in mice (6). The purpose of this study was to illuminate the molecular and cellular mechanisms that underlie integrin-mediated protection against exercise. We demonstrate that the increase in α7 integrin chain previously observed in exercised wild-type mice is due to the selective increase in transcription of the α7 integrin gene. In addition, immediate alterations in α7 integrin localization at the myotendinous junction in wild-type mice were visualized by immunofluoresence and quantified following exercise. The resulting increased adhesion at sites of high-force transmission may provide the stabilization necessary to prevent muscle fiber injury during exercise. Finally, α7−/− mice were used to examine the role of the α7β1 integrin in the injury response to exercise and the repeated bout effect.


Animals and downhill running exercise.

Protocols for animal use were approved by the Institutional Animal Care and Use Committee of the University of Illinois at Urbana-Champaign. α7−/− Mice (C57BL6-α7βgal strain) were produced in the University of Nevada Transgenic Center as described previously (16). All experiments were conducted at approximately the same time of day. Five-wk-old female wild-type and α7−/− mice remained at rest (basal), completed a single downhill running exercise (−20°, 17 m/min, 30 min), or completed two bouts of downhill running exercise separated by 1 wk. Speed on the treadmill was gradually increased from 10 to 17 m/min during the 21-min warm-up period. Mice completing exercise were euthanized via cervical dislocation immediately, 3 h, 24 h, and 1 wk postexercise. Mice subjected to a repeat bout of exercise were euthanized 3 h (RNA expression data) or 24 h (muscle injury data) following the second bout of exercise. To account for changes in RNA during development, 6-wk-old female wild-type mice that remained at rest were used as controls for the mice examined 1 wk postexercise (RNA expression data only). The gastrocnemius-soleus complexes were rapidly dissected and either frozen in liquid nitrogen prior to RNA extraction or frozen in liquid nitrogen-cooled isopentane for immunohistochemistry studies (n = 3–5 animals per experimental group).

Semiquantitative PCR.

Gastrocnemius-soleus complexes were crushed into a fine powder using a mortar and pestle. RNA was extracted using Trizol (Invitrogen), and RNA concentrations were determined spectrophotometrically at 260 nm. RNA was treated with DNase to eliminate potential genomic DNA contamination, and reverse transcription and PCR were performed with 1.0 μg using a RETROscript kit (Ambion). cDNAs were amplified by PCR in a DNA Thermal Cycler (MJ Research). PCR was done on triplicate samples prepared from each animal in each group. Each reaction contained 500 ng cDNA and primers for total α7 integrin; primers to distinguish the α7A and α7B isoforms; primers to distinguish α7X1 and α7X2; primers to distinguish total β1, β1A, and β1D; or primers for α4, α5, and α6 integrin subunits (Table 1). Cycle parameters for the amplifications of transcripts are listed in Table 2. Cycle times reported for each primer pair were within the linear increase in PCR product. PCR products were separated and visualized in 1.5% agarose gels with ethidium bromide. Band intensities were quantitated using ImageQuant software and normalized to GAPDH. GAPDH levels were not significantly altered by downhill running.

View this table:
Table 1.

Primer sequences for integrin subunits and α7-integrin splice variants

View this table:
Table 2.

Cycle parameters for cDNA amplification

Immunohistochemistry of α7 integrin at the myotendinous junction.

Localization of the α7A and α7B integrin isoforms was examined at myotendinous junctions in the distal end of gastrocnemius-soleus complexes by immunofluorescence microscopy. Sections 8-μm thick were fixed in acetone and blocked with PBS containing 5% BSA. Endogenous mouse immunoglobulin was blocked with 70 μg/ml goat anti-mouse monovalent Fab fragments (Jackson ImmunoResearch). Sections were reacted with rabbit polyclonal antibodies to the α7A cytoplasmic domain (α7CDA; 1:500) or the α7B cytoplasmic domain (α7CDB; 1:500) (45), and subsequently with a rat monoclonal antibody that recognizes tenascin-C (1:100; Sigma). FITC-labeled donkey anti-rabbit (1:100; Jackson ImmunoResearch) and rhodamine-labeled donkey anti-rat (1:100; Jackson ImmunoResearch) were used to detect integrin and tenascin, respectively. Slides were mounted using Vectashield containing DAPI (Vector Laboratories) and examined with a Leica DMRXA2 microscope (using the ×20 and ×40 objectives and Excitation480/Emission527 filters). Images were acquired using an AxioCam digital camera (Zeiss) and OpenLab software. To quantify the linear immunolocalization of α7A and α7B at the perimeters of myotendinous junctions, the lengths of integrin immunofluorescence at tenascin-positive myotendinous junctions were measured in three sections from each of three to four animals per group using the advanced measurements component of OpenLab. Total myotendinous junction areas in wild-type and α7−/− mice also were measured using OpenLab.

Assessment of myofiber damage.

Wild-type and α7−/− mice were injected with Evans blue dye (0.5 mg/ml, 0.05 ml/10 g body wt) 90 min prior to a single bout of exercise or a repeat bout of exercise 1 wk following the first exercise bout. For mice completing a repeat bout of exercise, Evans blue dye was injected 90 min prior to the second bout. Twenty-four hours following a single or repeat bout of exercise, gastrocnemius-soleus complexes were dissected and frozen as described above. Frozen 8-μm-thick sections were cut (3 sections per sample, separated by 100 μm) from both the proximal and distal ends of the muscle and a total of 50 fields were observed per animal by fluorescence microscopy (×40 objective, Excitation570/Emission640 filters). Mean numbers of Evans blue positive fibers in 50 fields are given.

Statistical analysis.

All averaged data are presented as the means ± SE. Comparisons between groups were performed by one-way ANOVA, followed by Tukey's post hoc analysis (Sigma Stat). An unpaired t-test was used to analyze α7 integrin perimeter at the myotendinous junction in the basal state and immediately postexercise. Differences were considered significant at P < 0.05.


Injury-producing exercise increases total α7 integrin mRNA.

Total α7 RNA increased 5.4-fold in wild-type mice by 3 h postexercise compared with animals that were not exercised (Fig. 1), and the amount of α7 RNA decreased to base level by 24 h postexercise. In contrast to α7 RNA, the amount of α4 integrin RNA decreased ∼57% 3 h postexercise and remained decreased up to 1 wk postexercise. No significant changes in expression of α5 integrin RNA were detected at any time following exercise. No changes in the levels of α6 integrin RNA were detected 3 and 24 h postexercise; however, α6 expression was decreased 84% 1 wk later.

Fig. 1.

Injury-producing exercise increases α7 RNA expression. Wild-type mice remained at rest (basal) or were run downhill (20° decline) at 17 m/min for 30 min (n = 3–5 per group). Gastrocnemius-soleus complexes were dissected 3 h postexercise (3PE), 24 h postexercise (24PE), or 1 wk postexercise (1wkPE). Top: bands from single samples are shown; however, triplicate samples from each mouse in each group were quantified. Bottom: RT-PCR analysis of α4, α5, α6, and α7 integrin demonstrate that only α7 integrin is increased postexercise, whereas α4 and α6 integrin gene expression are decreased. All values are normalized to GAPDH and represent means ± SE. *P < 0.05 vs. corresponding basal; #P < 0.05 vs. corresponding 24PE; ^P < 0.05 vs. all corresponding groups.

To further define the alterations in α7 RNA expression detected 3 h postexercise, we examined the changes in the transcripts encoding the α7A and α7B and the α7X1 and α7X2 isoforms 3 h, 24 h, and 1 wk postexercise. A 4.9-fold increase in α7A RNA and a 2.6-fold increase in α7B transcripts was detected 3 h postexercise (Fig. 2). α7X1 RNA was increased ∼13-fold 3 h postexercise, but remained only slightly elevated at 24 h and 1 wk (Fig. 2). A 3.5-fold increase in α7X2 RNA was also detected 3 h postexercise. Thus, the increase in total α7 seen 3 h postexercise appears to be reflected in both the cytoplasmic and extracellular domain isoforms. Two isoforms of the β1 integrin (β1A and β1D) heterodimerize with α7 chains. Total β1 RNA was not significantly changed at any time following exercise, nor were changes noted in the splice variants encoding β1A and β1D (data not shown). To determine whether a second bout of exercise promotes further increases in total α7 integrin RNA, RNA levels were examined 3 h following a second bout of exercise (1 wk following the first bout) and compared with 6-wk-old mice that were not exercised. No increase in total α7 RNA was found in mice following a second bout of exercise (data not shown).

Fig. 2.

Injury-producing exercise increases all α7 integrin splice-variant RNAs. Wild-type mice remained at rest (basal) or were run downhill (20° decline) at 17 m/min for 30 min (n = 3–6 per group). Gastrocnemius-soleus complexes were dissected 3PE, 24PE, or 1-wk PE. RT-PCR analysis of α7A, α7B, α7X1, and α7X2 integrin demonstrate that all α7 integrin isoforms are increased postexercise. *P < 0.05 vs. corresponding basal; **P < 0.01 vs. corresponding basal.

α7B Integrin is concentrated at the myotendinous junction following exercise.

α7 Integrin is localized at the sarcolemma (44), myotendinous junctions (2), and neuromuscular junctions (32) in adult skeletal muscle. Immunolocalization of α7A and α7B with tenascin-C demonstrates prominent expression of the α7 integrin between tendons and individual muscle fibers (Fig. 3A). When anti-α7 and anti-tenascin primary antibodies were omitted from the protocol, no staining was seen in wild-type mice and no α7 immunofluorescence was evident in α7−/− mice, confirming the specificity of these antibodies. Staining of α7A at the perimeter of the myotendinous junction was increased almost twofold immediately postexercise (Fig. 3B). An immediate increase in α7B localization at the perimeter of the myotendinous junction was also observed. Rapid mobilization of the α7 integrin to sites of high mechanical force transmission in the myoteninous junction, or mechanically induced conformational changes that increase accessibility of the integrin extracellular and cytoplasmic domains at myotendinous junctions to antibody may be responsible for increased detection of α7 integrin localization following exercise (4, 45). Interestingly, the areas of the myotendinous junctions defined by tenascin-C staining were 91% greater in the basal state of α7−/− mice compared with wild-type mice (wild-type = 292,745 ± 65,469 μm2, α7−/− = 558,655 ± 65,469 μm2; P < 0.05), and exercise did not alter the myotendinous junction area in either group (data not shown). Thus, the α7 integrin appears to be important to the normal architecture and cohesion of these junctions, and it may be rapidly mobilized or activated in response to exercise.

Fig. 3.

Downhill exercise increases α7A localization and α7B colocalization with tenascin-C at the myotendinous junction (MTJ). Wild-type and α7−/− mice remained at rest (basal) or were run downhill (20° decline) at 17 m/min for 30 min (n = 3–4 per group). Gastrocnemius-soleus complexes were dissected immediately postexercise (IPE) or 24PE. α7 Integrin is labeled with FITC-conjugated (green) secondary antibody, and tenascin-C is labeled with rhodamine-conjugated (red) secondary antibody. Secondary antibody-only controls were blank (not shown). A, left: presence of α7A and α7B at the MTJ. T, tendon; SkMs, skeletal muscle. Arrows depict fibers coexpressing tenascin-C and α7B in wild-type mice or fibers solely expressing tenascin-C in α7−/− mice. 24PE: higher magnification (×40) of fibers coexpressing integrin and tenascin-C. B: length of perimeters at the MTJ defined by α7A and α7B. C: number of cells coexpressing α7B and tenascin-C at the MTJ. Values are means ± SE. *P < 0.05 vs. basal; #P < 0.05 vs. IPE.

While examining the presence of the α7 integrin at the myotendinous junction, we noted that α7B and tenascin-C were colocalized in fibers lining the junction (Figs. 3, A and C). Fibers expressing α7B and tenascin-C were predominantly found in the soleus muscle near the Achilles tendon that attaches to the soleus and gastocnemius muscles. α7B immunofluorescence was consistently higher in the soleus compared with the gastrocnemius in both the basal and exercised muscle. These may represent sites of initiation of muscle fiber necrosis since intracellular expression of tenascin-C is only observed in mice with compromised muscle fibers, such as fibers deficient in dystrophin (19). The number of fibers in which high concentrations of α7B and coexpression of intracellular tenascin-C were detected was minimal in the basal state, increased twofold immediately postexercise, and increased 3.8-fold by 24 h postexercise (Fig. 3C). This response was specific to α7B since α7A did not colocalize with tenascin-C at these sites in the myotendinous junction. In α7−/− mice, fibers entirely positive for intracellular tenascin-C were frequently observed at the myotendinous junction. A further increase in the number of tenascin-C positive fibers and an increase in their distance from the junction were observed in α7−/− mice 24 h postexercise (Fig. 3A).

Injury is pronounced following single and repeat bouts of exercise in α7−/− mice.

Increased expression of cytoskeletal proteins, such as desmin and integrin, following injury-producing exercise or muscle contraction suggests a role for these molecules in stabilizing myofibers and rendering the cells less susceptible to damage following further exercise (6). Since the expression of α7 integrin RNA and protein are increased following exercise (6, 25), and excess α7β1 integrin prevents muscle damage (6), we hypothesized that the integrin may also prevent damage following a second bout of damaging exercise. Evans blue dye uptake was used to quantify injury. Compared with basal levels, more muscle injury was detected after a single bout of exercise in α7−/− mice than in wild-type animals (Fig. 4). Injury was evident in the proximal ends of gastrocnemius-soleus complexes; however, damage was most extreme at the distal ends of these complexes where the prominent junctions at the Achilles tendon are more susceptible to damage (27). Muscle damage at these sites was most severe in the α7−/− mice, and this confirms the role of the α7β1 integrin in protecting muscle from exercise-induced injury. At the distal sites, where injury is most prominent, a second bout of exercise further exacerbated the amount of injury in the α7−/− mice. This was not seen at the proximal ends of these muscle complexes of α7−/− mice, where the absence of integrin did not appear to promote significantly more injury. Thus, the protective effects afforded by the α7β1 integrin seen in wild-type mice appear to be site specific and most prominent at sites most susceptible to mechanical-induced damage. The repeat bout effect seen in proximal muscle was not dependent on the presence of integrin. Again, and in contrast, the integrin has a more profound effect at distal sites where it appears to be essential for the repeated bout effect, i.e., exercise-induced protection. Thus, both integrin-dependent and -independent mechanisms may underlie exercise-induced protection of muscle at different sites.

Fig. 4.

Increased injury at the distal gastrocnemius muscle in α7−/− mice following single and repeated bouts of exercise. Wild-type and α7−/− mice were injected with Evans blue dye and 90 min later remained at rest [basal (B)] or were run downhill (20° decline) at 17 m/min for 30 min (n = 5 per group) 1× (1Ex) or 2× separated by 1 wk (2Ex). Gastrocnemius-soleus complexes were dissected 24 h following the last bout of exercise. Values are means ± SE. *P < 0.05 vs. wild-type basal; **P < 0.05 vs. α7−/− basal.


The present study examines the effect of injury-producing exercise on the expression and localization of α7β1 integrin in skeletal muscle. A single bout of downhill running exercise selectively increased transcription of the α7 integrin gene shortly after the initiation of exercise and more α7A and α7B protein was detected at myotendinous junctions immediately following exercise. We found α7B, but not α7A, in cells at myotendinous junctions that also express tenascin-C in response to injury. Muscle injury following downhill running exercise was more pronounced in α7−/− mice compared with wild-type mice, especially in areas in tendinous tissue that are exposed to high forces. Increased muscle damage in exercised α7−/− mice and the lack of a protective effect following a second bout of eccentric exercise suggest the α7β1 integrin may attenuate development of injury and mediate the repeated bout effect.

Expression of α7β1 integrin transcripts and protein are increased and may partially compensate for the loss of dystrophin in Duchenne patients and in mdx mice (21) as well as in mdx/utrn−/− mice that develop a severe form of muscular dystrophy (8). α7 Integrin RNA also is increased in response to ablation injury in the soleus (25). The increased amounts of α7 integrin RNA and protein (6) in response to a single bout of exercise and in disease confirm that the α7 integrin is important to the structural and functional integrity of skeletal muscle.

A rapid and transient fivefold increase in total α7 integrin RNA was observed 3 h postexercise and likely underlies the increase in α7 integrin protein 24 h postexercise previously reported (6). The increase in α7 transcription is specific; α5 integrin remained unchanged, and α4 and α6 integrin RNAs decreased following exercise. A dramatic reduction of α6 on the cell surface has also been detected after transfection of HEK-209 cells to express α7 (42).

Microarray analysis of mouse plantaris muscle following voluntary wheel running shows an approximate twofold increase in α7 integrin RNA (12). Seven days of aortic constriction created a chronic state of pressure overload on the ventricle in the heart and promoted increases in α1 and α5 integrin transcripts and α7 and β1D protein in the myocardium (1). Additional studies have shown that components of focal adhesion complexes, including focal adhesion kinase (FAK) and paxillin, increase in response to stretch and functional overload (17). These studies and our previous report showing prevention of injury in mice overexpressing the α7 chain indicate that the α7β1 integrin is a mechanosensitive molecule that is increased in muscle and promotes stability in response to exercise, mechanical loading, and/or injury. Whereas integrins containing α1, α3, or αv chains are not localized to the myotendinous junction, they may not be responsive to injury-producing exercise in skeletal muscle.

Developmentally regulated alternative splicing of α7 transcripts generate alternative cytoplasmic (α7A and α7B) and extracellular (X1 and X2) domains (45). The α7B is expressed in myoblasts, adult muscle fibers, and other cell types, whereas α7A is expressed upon terminal muscle differentiation and is restricted to adult skeletal muscle (13, 48). Both X1 and X2 splice variants are expressed in myoblasts and myofibers. Upon myoblast differentiation there is an increase in X1 compared with X2, but between birth and 90 days the ratio of X2/X1 increases in skeletal muscle (22, 48). In this study, transient increases in the RNAs encoding all isoforms of α7 integrin chain were observed following exercise. Whereas little to no α7X1 was detected in the basal state, the significant amount of α7X1 found 3 h postexercise represents a major shift in RNA splicing. The α7X1 and α7A isoforms are also preferentially increased during regeneration following muscle ablation injury in the soleus (25). These increases in α7 RNAs may establish new links between integrin and laminin at specific sites of muscle damage and promote muscle repair and/or regeneration. This is highly consistent with the increased integrity of the myotendinous junction in dystrophic mdx/utrn−/− mice fortified by transgenic overexpression of α7 (9). This leads us to suggest that exercise that enhances α7 integrin expression may improve muscle integrity and function in human muscular dystrophies characterized by the absence of dystrophin or by secondary reductions in the α7β1 integrin (40).

Engagement of the α7β1 integrin with laminin and/or antibodies activates and promotes conformational changes that result in access to antibody and its association with the cytoskelton (4, 31, 45, 47). The increased amount of α7 integrin detected at the myotendinous junction immediately postexercise likely results from such changes in integrin association with laminin and/or mobilization of laminin or integrin at these sites of high-force transmission. The result may be increased resistance to mechanical force.

Tenascin-C binds to integrins (α8β1, αvβ3, αvβ6, α9β1) (10, 11, 14) and other extracellular matrix proteins, and thereby promotes adhesion and anti-adhesion processes (11, 14). Tenascin-C is made in fibroblasts and chondrocytes in the extracellular matrix of the myotendinous and osteotendinous junctions, the superficial layer of the articular cartilage (26), and all musculoskeletal regions that transmit high forces. Tenascin-C is also localized at sites enriched in inflammatory cells in skeletal muscle of patients with Duchenne and Becker muscular dystrophy and myositis (19) and in degenerating/regenerating fibers of mdx mice (43). Tenascin-C was abundant in the cytoplasm of fibers surrounding the myotendinous junction in α7−/− mice, particularly in areas enriched in mononuclear cells (Fig. 3). Using tenascin-C as a marker, the areas of myotendinous junctions were found to be increased in α7−/− mice in the preexercise state. Thus the α7β1 integrin is essential to the normal architecture and function of the myotendionous junction, and this is disrupted in muscular dystrophy and in α7−/− mice (36). Concentration and increases in α7β1 at these sites may protect against exercise-induced muscle damage.

A rapid increase in tenascin-C mRNA arises in response to mechanical loading of the chicken anterior latissimus dorsi muscle, and this increase is independent of macrophage infiltration (18). We also observe an increase in tenascin-C immunofluorescence in wild-type muscle at the myotendinous junction immediately following muscle lengthening exercise. The integrin has been proposed to be a mechanosensor involved in the initiation of tenascin-C expression (10). Both α7 integrin and tenascin-C proteins are expressed in specific cells at the myotendinous junction and both are believed to be involved in muscle regeneration (9, 24). Whether the α7β1 integrin directly affects tenascin-C expression or vice versa remains to be determined.

Different theories have been proposed to explain the mechanism underlying the repeated bout effect, that is, the decrease in muscle damage that occurs following a subsequent bout of accustomed eccentric exercise. Mechanical adaptations within the muscle (involving increased passive and dynamic muscle stiffness), cellular adaptations, changes in inflammatory responses, and modifications to excitation-contraction coupling may be involved, but evidence of a defining factor is lacking (35). We previously demonstrated that the amount of α7 chain is increased 24 h postexercise and remained elevated 1 wk later (6). In this study, α7 RNA was increased following a single bout of exercise, and no further increase in α7 expression was detected following a second bout of exercise 1 wk later. Thus, once the level of α7 integrin is enhanced, no further synthesis appears necessary following a second insult. The increase in muscle damage in α7−/− mice detected in this study, particularly at the distal ends of fibers, confirms that α7β1 integrin protects against injury induced by eccentric exercise and that an increase in the amount and localization of this integrin may underlie the repeated bout effect at specific sites. Future studies are needed to examine the sustainability of α7β1 expression at the myotendinous junction following exercise and the molecular mechanism by which this integrin protects against muscle damage.

Perspectives and Significance

We have shown that injury-producing exercise promotes selective expression of the α7 integrin gene. This results in the enhanced synthesis of α7 chain and underlies integrin-mediated protection against exercise-induced damage. Mechanical strain, rather than factors associated with injury, likely provides the stimulus for this increased integrin expression and localization considering the rapid time course by which these events occur in the muscle. The contribution of this adhesion factor to the myotendinous junction may provide the stabilization necessary during the relatively long time course required for complete skeletal muscle regeneration and repair. The persistence of enhanced adhesion, particularly at sites of prominent junctions most susceptible to damage, may optimally protect against degradation associated with subsequent muscle lengthening and strain.


This work was supported by National Institute on Aging Grant RO1-AG-014632 and a grant from the Muscular Dystrophy Association (to S. J. Kaufman) and National Institutes of Health Division of Research Resources Grants P20-RR-018751 and P20-RR-15581 (to D. J. Birkin).


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 26a.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35.
  37. 36.
  38. 37.
  39. 38.
  40. 39.
  41. 40.
  42. 41.
  43. 42.
  44. 43.
  45. 44.
  46. 45.
  47. 46.
  48. 47.
  49. 48.
  50. 49.
View Abstract