Activation of c-Raf-1 Kinase Signal Transduction Pathway in α7 Integrin-deficient Mice*

Integrin α7-deficient mice develop a novel form of muscular dystrophy. Here we report that deficiency of α7 integrin causes an activation of the c-Raf-1/mitogen-activated protein (MAP) 2 kinase signal transduction pathway in muscle cells. The observed activation of c-Raf-1/MAP2 kinases is a specific effect, because the α7 integrin deficiency does not cause unspecific stress as determined by measurement of the Hsp72/73 level and activity of the JNK2 kinase. Because an increased level of activated FAK was found in muscle of α7 integrin-deficient mice, the activation of c-Raf-1 kinase is triggered most likely by an integrin-dependent pathway. In accordance with this, in the integrin α7-deficient mice, part of the integrin β1Dvariant in muscle is replaced by the β1A variant, which permits the FAK activation. A recent report describes that integrin activity can be down-modulated by the c-Raf-1/MAP2 kinase pathway. Specific activation of the c-Raf-1/MAP2 kinases by cell-permeable peptides in skeletal muscle of rabbits causes degeneration of muscle fibers. Therefore, we conclude that in α7integrin-deficient mice, the continuous activation of c-Raf-1 kinase causes a permanent reduction of integrin activity diminishing integrin-dependent cell-matrix interactions and thereby contributing to the development of the dystrophic phenotype.

Integrins are a family of heterodimeric transmembrane receptors consisting of an ␣and a ␤-subunit (1). Each subunit spans the plasmamembrane once, and in both cases the N termini are localized extracellularly binding the ligand, whereas the C termini face the cytoplasm. Integrins interact with proteins of focal adhesions that mediate the attachment of bundles of stress fibers (2). Moreover, numerous proteins involved in signal transduction are concentrated in these regions. Therefore integrins do not merely mediate cell attachment to the extracellular matrix but can also promote physiological processes such as migration and cell invasion. On the molecular level, integrin activation can modulate specific gene expression, cell proliferation, cell cycle progression, or prevention of apoptosis. Integrins can be considered as classic receptors without catalytic activity triggering a variety of signal transduction pathways upon ligand binding. In a simplified model, ligand occupation of integrins leads to the activation of the nonrecep-tor tyrosine focal adhesion kinase (FAK) 1 by its autophosphorylation at Tyr-397 (2,3). By this tyrosine phosphorylation a binding site (YpAEI motif) for the SH2 domain of proteintyrosine kinases of the c-Src family is generated. The subsequent Src-dependent phosphorylation at Tyr-576 and Tyr-577 increases the catalytic activity of FAK (4). The Src-dependent phosphorylation at Tyr-925 (YpENV motif) generates a binding site for the SH2 domain of the adapter protein Grb2. The adapter protein Grb2 consists of a middle SH2 domain known to bind to phosphotyrosine residues and N-and C-terminal SH3 domains (5). The SH3 domains of Grb2 are known to interact with SOS proteins (PXXP motif; Ref. 6), which modulate Ras activity, finally resulting in an activation of c-Raf-1 kinase.
Integrins mediate signals both from the extracellular matrix to the cytoplasm (outside-in signaling) and from the cytoplasm to the outside (inside-out signaling) (for a recent review see Ref. 7). Modulation of integrin activity is therefore of major biological significance for a variety of physiological processes like cell migration or differentiation. By the process of inside-out signaling, integrin affinity for its specific ligand is modulated in response to intracellular signals. In a recent report, the Ras/ Raf-initiated MAP2 kinase signal transduction pathway was described as a novel transcription-independent modulator of integrin activity (8). Activated MAP2 kinase down-regulates integrin activity without changing the phosphorylation state of the integrin ␤-subunit. Because ligand binding to certain integrins can result in a Ras/Raf-dependent activation of MAP2 kinase, this novel pathway most probably represents a negative feedback loop in integrin activity.
The ␣ 7 ␤ 1 integrin, a receptor for the basement membrane glycoprotein laminin, is predominantly expressed in muscle (9,10). Mice lacking the ␣ 7 integrin subunit develop a novel form of muscular dystrophy (11).
Here we describe an activation of the c-Raf-1/MAP2 kinase signal transduction pathway in muscle tissue derived from ␣ 7 integrin-deficient mice. Specific activation of this pathway in rabbit skeletal muscle is involved in degeneration of muscle fibers. Therefore, permanent activation of this signaling cascade might contribute to the manifestation of the dystrophic phenotype by down-modulation of integrin activity.

EXPERIMENTAL PROCEDURES
Animals and Antibodies-Tissues were obtained from mice (11,12) at several ages ranging from 30 days to 1.5 years. Chinchilla bastard rabbits were used for the injection experiments. Primary antibodies used were rabbit sera, goat sera, or mouse monoclonal antibodies against c-Raf-1 kinase, MAP2 kinase, JNK2, Hsp72/73, FAK, phosphotyrosine, and activated MAP2 kinase, all from Santa Cruz Biotechnology. The integrin ␤ 1 -specific monoclonal antibody was purchased from * This work was supported by a grant from the Mü nchener Medizinische Wochenschrift (to E. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Life Technologies, Inc. The integrin ␤ 1 -specific antiserum was kindly provided by Dr. Staffan-Johannson, and the ␤ 1D -specific antiserum was provided by Dr. Engvall. The monoclonal antibodies Q19/10 and HBV25-19 specific for the PreS2 domain of the hepatitis B virus surface antigen were kindly provided by Dr. Gerlich and Dr. Mimms.
For Western blot analysis, proteins were separated by SDS-polyacrylamide gel electrophoresis (13) and transferred onto polyvinylidene difluoride membranes (Millipore, 0.45 m). Membranes were washed for 30 min with PBS/Tween and preblocked for 15 min in PBS/Tween (0,05%) containing 10% nonfat dry milk powder (blocking solution). Incubation with the first antibody diluted in blocking solution was performed for 2 h at room temperature. After extensive washing with PBS/Tween, the bound antibody was detected using anti-mouse peroxidase-conjugated or anti-rabbit peroxidase-conjugated antibodies (Amersham Pharmacia Biotech), respectively. Following 45 min of washing, the blots were stained using the ECL reagent (Amersham Pharmacia Biotech).
Immunoprecipitations-Lysates were prepared as described above. To the lysate containing 500 g of total protein, 20 l of a protein A/G-Sepharose suspension (Santa Cruz Biotechnology) were added and incubated at 4°C under permanent shaking for 15 min. After centrifugation the supernatant was incubated with a specific polyclonal antiserum at 4°C for 90 min under permanent shaking. Then 25 l of the protein A/G-Sepharose suspension were added, and the incubation was continued for 45 min. The beads were washed twice in 500 mM LiCl, 100 mM Tris-HCl, pH 7.4, followed by a final wash in 10 mM Tris-HCl, pH 7.4. The washed precipitates were resuspended in adequate buffer.
Immunocomplex Assays-For each assay, immunoprecipitated kinases were resuspended in 36 l of kinase buffer and incubated at 30°C for 15 min. The reaction was stopped by addition of SDS sample buffer. An aliquot was loaded on a 10% SDS gel and analyzed by autoradiography. In the case of c-Raf-1, the kinase buffer contained 25 mM Tris-HCl, pH 7.5, 25 mM ␤-glycerophosphate, 10 mM MgCl 2 , 10 mM ATP, 1 mM dithiothreitol, in the presence of 5 Ci of [␥-32 P]ATP and 1 g of MEK (Santa Cruz Biotechnology) per assay. In the case of MAP2, the kinase buffer contained 25 mM Hepes, pH 7.5, 10 mM MgCl 2 , 1 mM dithiothreitol, 50 M ATP, 5 Ci of [␥-32 P]ATP, and 2.5 g of basic myeloglycoprotein (Santa Cruz Biotechnology) as a substrate per assay. In the case of JNK2, the kinase buffer contained 20 mM Hepes, pH 7.6, 20 mM MgCl 2 , 10 mM ␤-glycerophosphate, 2 mM dithiothreitol, 50 M ATP, 5 Ci of [␥-32 P]ATP, and 1 g of recombinant glutathione Stransferase-Jun (Santa Cruz Biotechnology).
Immunofluorescence-Cryosections of soleus muscle were obtained and processed as described previously (11). Briefly, after blocking the samples with 5% normal goat serum in PBS/Tween for 1 h, incubation with the c-Raf-1 kinase-specific rabbit derived antiserum diluted was performed in 2% normal goat serum in PBS/Tween for 1 h. The slides were washed and incubated with the Cy3-conjugated secondary antibody (Dianova) diluted in 2% normal goat serum in PBS/Tween for 45 min. After final washing, sections were mounted in Vectrashield (Vector Lab, Burlingame, CA) and analyzed on an DMR fluorescence microscope (Leica).
Injection Experiments-Protein was injected into the erector trunci muscle in a concentration of 2.5 mol (100 l) in PBS in a 1:1 dilution with green ink (Sigma). To minimize individual effects in these experiments, the PreS1PreS2 protein was injected into the left muscle, and the mutant was injected into the right muscle. After 24 h, the equal amount of protein was injected again. The rabbits were sacrificed 24 h later, and the muscle specimen was prepared as described above.
Histochemical Analysis-Hematoxylin and eosin staining on cryosections was performed as described elsewhere (11).

In ␣ 7 Integrin-deficient Mice c-Raf-1 and MAP2 Kinases Are
Activated-Integrin ␣ 7 deficiency causes the development of a new form of muscular dystrophy in mice (11). Because integrins are involved in major signal transduction pathways, we investigated whether loss of the ␣7 subunit has an influence on the activity of c-Raf-1 kinase. The activity of c-Raf-1 kinase was determined by immunocomplex assays using recombinant MEK as substrate in lysates derived from skeletal muscle of hind limb and diaphragm of ␣ 7 -deficient and wild type control mice. The assays show a significant activation of the c-Raf-1 kinase in the case of ␣ 7 integrin-deficient mice as compared with wild type controls in both tissues (hind limb and diaphragm) (Fig. 1A).
In vivo, the activation of c-Raf-1 kinase can be transduced via MEK and results in subsequent activation of MAP2 kinase. Therefore, the activity of MAP2 kinase in lysates of hind limb and diaphragm was determined by immunocomplex assays using basic myeloglycoprotein (MBP) as substrate. Deficiency of ␣ 7 integrin induces a strong activation of MAP2 kinase in both tissues as compared with the wild type ( Fig. 1B).
To exclude the possibility that the observed effect was due to an increased expression level of these kinases, the total amounts of c-Raf-1 kinase (Fig. 1C) and MAP2 kinase ( Fig. 1D) were determined by Western blot analysis of hind limb-derived lysates. The comparison of integrin ␣ 7 -deficient mice and controls shows equal amounts of both kinases. Moreover, it was shown by Western blotting that the precipitates that were subjected to immunocomplex assays contained comparable amounts of MAP2 kinase (Fig. 1E) or c-Raf-1 kinase (data not shown). This set of experiments confirms that the observed increase of c-Raf-1/MAP2 kinase activity in integrin ␣ 7 -deficient mice is indeed due to an activation of the c-Raf-1/MAP2 kinase signal transduction pathway.
Activation of c-Raf-1/MAP2 Kinase Signal Transduction Pathway in ␣ 7 Integrin-deficient Mice Is a Specific Process-Activation of c-Raf-1/MAP2 kinases can be mediated by specific pathways that are induced by specific receptor/ligand interactions. Nevertheless, an activation by unspecific cellular stress factors has also been described (14). Therefore, it was investigated whether integrin ␣ 7 deficiency gives rise to unspecific cellular stress. Two parameters were chosen for analysis. The level of Hsp72/73, which is known to be elevated by cellular stress, was determined by Western blotting in lysates derived from hind limb or diaphragm, respectively. The comparison of wild type controls and ␣ 7 integrin-deficient mice revealed no difference in Hsp72/73 levels ( Fig. 2A). In addition, the activity of JNK2 (also designated SAPK for stress-activated protein kinase) was determined by immunocomplex assays. Liver tissue derived from transgenic mice overproducing the LHBs protein of hepatitis B virus (12) served as the positive control. The determination of the JNK2 kinase activity in both tissues (Fig.  2B) confirmed that the loss of ␣ 7 integrin does not cause unspecific cellular stress.
Activation of c-Raf-1 Kinase Is a Muscle Cell-specific Process-Skeletal muscle tissue consists of several different cell types. Therefore we wanted to investigate whether the observed activation of c-Raf-1 kinase takes place in muscle cells. Activation of c-Raf-1 kinase is associated with a translocation of the activated kinase to the membrane (Fig. 3A). Cryosections of the soleus muscle were stained with a c-Raf-1 kinase-specific antiserum and analyzed by indirect immunofluorescence. In the case of sections derived from wild type mice, a weak homogenous staining of the muscle fibers was observed characteristic for the inactive state of c-Raf-1 kinase. In contrast, the staining of sections derived from ␣ 7 -deficient mice showed an intensive staining of the subsarcolemmal region, reflecting the translocation of activated c-Raf-1 kinase to the plasma membrane (Fig. 3B). These results indicate that the observed activation of c-Raf-1 kinase occurs within muscle cells.
In the Integrin ␣ 7 -deficient Mice Part of the Integrin ␤ 1D Variant in Muscle Is Replaced by the ␤ 1A Variant-The loss of one integrin ␣ subunit could influence the overall expression pattern of integrin subunits. To determine whether the loss of the integrin ␣ 7 subunit affects the amount of its heterodimeric binding partner ␤ 1 , Western blot analysis using ␤ 1 integrinspecific antisera was performed. As shown in Fig. 4A, the One representative assay using 100-day-old mice is shown. B, activity of MAP2 kinase was determined by immunocomplex assays using MBP and [␥-32 P]ATP as substrates. In the case of hind limb lysates (lanes 1 and 2) derived from integrin ␣ 7 -deficient mice (lane 1), about 2.2-fold activation was measured, and in the case of the diaphragm lysates (lanes 3 and 4) derived from the integrin ␣ 7 -deficient mice, about 3.3-fold activation of the MAP2-kinase was determined as compared with the respective wild type controls (lanes 2 and 4). The given factors are the mean values of three independent experiments. One representative assay using 100-day-old mice is shown. C and D, Western blot analysis of lysates derived from hind limb of an integrin ␣ 7 knock out mouse (lane 1) and an adequate wild type control (lane 2) using a c-Raf-1 kinase-specific antiserum (C) or a MAP2 kinase-specific antiserum (D) shows that neither the level of the c-Raf-1 kinase (C) nor the level of MAP2 kinase (D) is affected by ␣ 7 integrin deficiency. E, hind limb-derived lysates of an integrin ␣ 7 knock out mouse (lane 1) and adequate wild type control (lane 2) were immunoprecipitated using the MAP2-specific monoclonal antibody as described above for panel B. The immunoprecipitates were analyzed by Western blotting using the MAP2-specific goat-derived antiserum. The blot shows that comparable amounts of MAP2 were precipitated.  1 and 2) and diaphragm (lanes 3 and 4) of integrin ␣ 7 knock out mice (lanes 1 and 3) and adequate wild type controls (lanes 2 and 4) using a Hsp72/73-specific antiserum shows that the Hsp72/73 level is not affected by ␣ 7 integrin deficiency. B, activity of JNK2 kinase was determined by immunocomplex assays using recombinant glutathione S-transferase-Jun and [␥-32 P]ATP as substrates. No significant difference in JNK2 kinase activity was measured in lysates derived from hind limb and diaphragm of integrin ␣ 7 -deficient mice (Ϫ/Ϫ) and wild type controls (ϩ/ϩ) from 100-day-old mice. As positive control, lysates derived from livers of LHBs transgenic mice were compared with their respective wild type controls.
FIG. 3. Activated c-Raf-1 kinase within muscle cells of ␣ 7 integrin-deficient mice. Immunofluorescence analysis (400-fold magnification) of ethanol-fixed cryosections of soleus muscle derived from a 100-day-old wild type control mouse (A) and an ␣ 7 integrin-deficient mouse (B) using a c-Raf-1 kinase-specific rabbit-derived antiserum shows the translocation of c-Raf-1 kinase to the subsarcolemmal compartment in integrin ␣ 7 -deficient mice reflecting the muscle cell-specific activation of this kinase. deficiency of the ␣ 7 subunit does not influence the total amount of ␤ 1 integrin.
In mice, two different variants of the integrin ␤ 1 subunit have been described: ␤ 1A and ␤ 1D . The integrin ␤ 1D variant is strictly muscle-specific, whereas the ␤ 1A variant is ubiquitously expressed. Western blot analysis using an antiserum specific for the integrin ␤ 1D variant revealed that integrin ␣ 7 deficiency is accompanied with a strong reduction of the ␤ 1D level in muscle (Fig. 4B). Because the total amount of ␤ 1 integrins remains unchanged, it can be concluded that a decrease in the ␤ 1D level can only be adjusted by an increase in ␤ 1A . Therefore, in the integrin ␣ 7 -deficient mice part of the ␤ 1D variant in muscle is replaced by the ␤ 1A variant.
Tyrosine Phosphorylation of FAK Is Increased in ␣ 7 Integrindeficient Mice-Activation of c-Raf-1 kinase can be initiated by a broad variety of different receptors and signal transduction pathways. In the case of an integrin-dependent activation, the initial step in the cascade is the tyrosine phosphorylation of FAK at Tyr-397 and Tyr-925 (2), finally resulting in the activation of c-Raf-1 kinase.
To test whether the tyrosine phosphorylation of FAK is affected in ␣ 7 integrin-deficient mice, FAK was immunoprecipitated from lysates derived from the hind limb of wild type and ␣ 7 integrin-deficient mice. The precipitates were analyzed by Western blotting using a phosphotyrosine-specific serum. The For this purpose, the HBV-derived PreS1/PreS2 protein was used. This protein was chosen because it possesses two properties: (i) Because of an amphipatic ␣ helix at the C terminus of the PreS2 domain, the protein is cell-permeable. 2 (ii) The PreS2 domain was shown to trigger the activation of the c-Raf-1/ MAP2 signal transduction cascade (for an overview see Ref. 16). A protein with mutations in the amphipatic ␣ helix abolishing cell permeability and the activator function (17) 2 served as the negative control in this study. These mutations do not affect the reactivity with the antisera Q19/10 and HBV25-19, which were used to detect these proteins.
For the injection experiments, the protein solutions were mixed with an equal volume of green ink to label the injection channel. Two injections were performed over a time period of 48 h, and the tissues were prepared as described under "Experimental Procedures." To confirm that the PreS1/PreS2 protein indeed penetrates the muscle cells, immunostaining of cryosections with the PreS2-specific antisera was performed. The microscopic analysis (Fig. 6) shows the injection channel labeled by the green ink (Fig. 6, A and C). In the case of the control, only the intercellular space was labeled, whereas the cells remained unstained (Fig. 6D). In the case of the PreS1/PreS2 protein injection, muscle cells adjacent to the injection channel were stained, indicating that the PreS1/PreS2 protein has entered the cells (Fig. 6B).
In the next set of experiments it was analyzed whether the internalized PreS1/PreS2 protein truely activates the c-Raf-1/ MAP2 signal transduction cascade under these conditions. To test this, 10 cryosections covering the injection channel were FIG. 4. Selective reduction of the integrin ␤ 1D variant in muscle in ␣ 7 integrin-deficient mice. A, Western blot analysis of lysates derived from hind limb and diaphragm of ␣ 7 knock out mice and the adequate wild type control using a ␤ 1 -specific polyclonal antiserum. The blot shows that the total amount of ␤ 1 integrins is not affected by ␣ 7 integrin deficiency. B, Western blot analysis of lysates derived from hind limb (lanes 1 and 2) and diaphragm (lanes 3 and 4) of 100-day-old ␣ 7 knock out mice (lanes 2 and 4) and the adequate wild type control (lanes 1 and 3) using a ␤ 1D -specific antiserum. The blot shows that in the case of ␣ 7 deficiency, the amount of the ␤ 1D isoform is reduced as compared with the wild type control. pooled, adjusted to equal protein concentration, and analyzed by Western blotting using an antiserum specific for activated MAP kinases. The Western blot shows a strong increase in the amount of activated MAP2 kinase in the case of the lysates derived from the PreS1/PreS2 protein-injected tissue as compared with the control (Fig. 6E). This demonstrates that the PreS1/PreS2 protein significantly activates the c-Raf-1/ MAP2 signal transduction pathway under these experimental conditions. (Fig. 6) indicates that the chosen experimental system is suitable for studying the consequences of an activation of the c-Raf-1/MAP2 signal transduction cascade for the integrity of muscle tissue in vivo.

Activation of the c-Raf-1/MAP2 Signal Transduction Pathway Can Cause Degeneration of Muscle Fibers-This set of experiments
In accordance with the experiments described above (Fig. 4), it was analyzed whether the activation of the c-Raf-1/MAP2 signal transduction cascade affects the level of ␤ 1 integrins in total. Therefore Western blotting of the lysates described above was performed using an integrin ␤ 1 -specific antiserum. The Western blot (Fig. 7A) demonstrates that the injection of the PreS1/PreS2 protein does not influence the total amount of ␤ 1 integrins as compared with the control.
The phenotype of the muscle tissue was investigated by histochemical analysis of the cryosections. The Fig. 7 (B and C) are composed of overlapping micrographs and provide an overview at a 100-fold magnification of the tissue close to the injection channel as marked by the green ink. In both samples mild fiber necrosis can be detected because of the injection event. The overview of the control injected tissue (Fig. 7C) as well as the more detailed analysis (200-fold magnification, Fig.  7F) show all normal features of intact muscle fibers. In contrast to this, the PreS1/PreS2 protein injected tissue (Fig. 7B, 200fold magnification; Fig. 7, D and E) shows severe changes of muscle tissue histology. An increased variability of muscle fiber diameter and many fibers with centrally localized nuclei could be observed. These are typical characteristics of muscle fiber regeneration. Moreover, basophilic degenerated and necrotic fibers as well as infiltration of phagocyting cells could be detected. This histological analysis of the PreS1/PreS2 protein injected tissue reveals ongoing processes that are typical for muscle fiber degeneration and regeneration resembling early stages of muscular dystrophies. These data demonstrate for the first time that the activation of the c-Raf-1/MAP2 signal transduction pathway can cause the typical symptoms of a muscular dystrophy. DISCUSSION Integrin ␣ 7 deficiency was shown to cause a novel form of muscular dystrophy in mice (11). In accordance, patients lacking the integrin ␣ 7 subunit develop a myopathy (18). Integrin ␣ 7 -deficient mice are viable and fertile, indicating that myogenesis in principle is not affected by loss of ␣ 7 integrin. This gives rise to the question of which molecular trigger causes the destruction of the organized muscle architecture, resulting in the dystrophic phenotype.
In a recent report it was described that activated Ras/c-Raf-1 pathway is able to down-modulate the activity of ␤1 or ␤3 integrins via MAP2 kinase in a transcription-independent manner (8). Modulation of integrin activity is of major significance for a variety of biological processes as cell migration, which can be considered as an alternation between a high affinity (attachment) and low affinity (detachment) state. Because integrins are known to activate MAP2 kinase by the Ras/Raf mediated cascade, this pathway most probably represents a negative feedback loop controlling integrin activity.
In this study we demonstrate that ␣ 7 integrin deficiency results in a permanent activation of the c-Raf-1/MAP2 kinase signal transduction pathway in muscle cells. In principle, the observed activation of this signaling cascade could be because of an unspecific secondary effect. However, arguing against an unspecific effect are two observations. First, an activation because of unspecific cellular stress was excluded because both tested markers indicating cellular stress, increased activity of JNK2 and increased amount of Hsp72/73, were not affected by ␣ 7 integrin deficiency. Second, the activation of c-Raf-1 and MAP2 kinases in ␣ 7 integrin-deficient mice is not an age-dependent process, because it was observed at different age stages (between 30 days and 1.5 years; data not shown). The dystrophic phenotype becomes evident in the diaphragm at an age of 2 months (11). Therefore these data indicate that the activation of c-Raf-1/MAP2 kinase signal transduction pathway by the ␣ 7 integrin deficiency can be observed before the phenotype of muscular dystrophy becomes evident. Therefore, it can be excluded that the activation of this signaling cascade is a secondary effect caused by the degenerating/regenerating processes of the muscular dystrophy.
In the integrin ␣ 7 -deficient mice an increased phosphorylation of FAK can be observed, suggesting that the activation of ␤ 1 integrins is involved in the activation of c-Raf-1 kinase. In a recent report it was demonstrated that the integrin ␣ 7 ␤ 1 negatively regulates the function of the integrin ␣ 5 ␤ 1 (20). Therefore, the lack of the ␣ 7 subunit could result in an activation of the integrin ␣ 5 ␤ 1 , resulting in an activation of FAK.
Moreover in the integrin ␣ 7 -deficient mice a significant change in the composition of the ␤ 1 integrin variants occurs. The integrin ␤ 1D variant is the only integrin ␤ 1 variant that is expressed in the sarcolemma of wild type muscle. In the integrin ␣ 7 -deficient mice part of the integrin ␤ 1D variant is replaced by the integrin ␤ 1A variant, resulting in an unaltered amount of ␤ 1 integrins. These two variants only differ in their cytoplasmic domains, also affecting their function (21)(22)(23)(24). The integrin ␤ 1D variant mediates a tighter binding to the matrix, and its activity cannot be regulated (21,23). Therefore, the reduction of the ␤ 1D variant could contribute directly to a diminished tissue stability. In contrast, the integrin ␤ 1A variant is well known to be involved in dynamic functions, for example the activation of FAK (21). An interaction of integrin ␤ 1D and FAK has not been found in vivo. That might be due to the fact that the in vivo binding motif for FAK only exists in the integrin ␤ 1A variant (15). Under the conditions of a strong overproduction, however, a ␤ 1D -dependent activation of FAK in vitro was reported (23). Therefore, in the integrin ␣ 7 -deficient mice the change in the integrin ␤ 1 variant composition in combination with the loss of the integrin ␣ 7 negative regulatory effect can be causative for the observed activation of FAK and subsequently of the c-Raf-1/MAP2 kinase signal transduction pathway.
As introduced above, activated MAP2 kinase can down-modulate integrin activity (8). That can reduce cell-matrix interactions that are crucial for muscle integrity. To investigate whether activated MAP2 kinase can trigger the destruction of muscle architecture, we established a novel experimental approach: we activated the c-Raf-1/MAP2 kinase signal transduction pathway in muscle fibers by the means of an exogenous protein and investigated whether this activation causes degeneration of muscle tissue.
For this purpose, we used the HBV-derived cell-permeable FIG. 7. Degeneration of muscle fibers by activation of the c-Raf-1/ MAP2 signal transduction pathway. A, 10 cryosections covering the injection channel were pooled, adjusted to equal protein concentration, and analyzed by Western blotting using an antiserum specific for the ␤ 1 integrin subunit The total amount of the ␤ 1 integrin subunit was not affected by the injection of PreS1/PreS2 protein (lane 2) as compared with the control-injected sample (lane 1). B and C, HE stained cryosections of PreS1/PreS2 protein injected tissue (B) and control injected tissue (C). The panels were composed of three overlapping micrographs at a 100-fold magnification. D-F provide a more detailed analysis at 200-fold magnification of the areas labeled in panels B and C. In the case of the PreS1/PreS2 injected tissue (D and E), severe changes of muscle histology can be observed. An increased variability of muscle fiber diameter (arrow w), fibers with centrally localized nuclei (arrow x), basophilic fibers (arrow y), and infiltration of phagocyting cells (arrow z) can be observed. In the case of the control injected tissue (F) only the cells surrounding the injection channel were affected. PreS1/PreS2 protein. The PreS2 domain triggers an activation of the c-Raf-1/MAP2-kinase signal transduction pathway. These experiments were controlled using a mutant (17) 2 that displays no activator function and lacks the cell permeability. The reactivity with the antisera used for the detection of these HBV-specific proteins is not affected by this mutation.
In one set of experiments we determined the reliability of this novel experimental approach. Immunofluorescence microscopy analysis reveals that the PreS1/PreS2 protein indeed enters the muscle cells surrounding the injection channel, whereas the mutant was only found in the intercellular space.
Western blot analysis using an activated MAP-specific antiserum confirmed that the internalized PreS1/PreS2 protein triggers the activation of MAP2 kinase in this experimental system, whereas in the case of the control injected animals no activation could be observed, underlining the specificity of this experimental system. This set of experiments demonstrates that the cell-permeable PreS1/PreS2 protein is suitable to trigger specifically an exogenous activation of the c-Raf-1/MAP2 signal transduction cascade. This novel experimental system provides several advantages. It combines an easy handling with short latency periods. The PreS2 domain triggers highly specifically the activation of the c-Raf-1/MAP2 kinase cascade which has been investigated in detail (16). Furthermore, the existence of the mutant provides the proper control. Because of the spreading capacity of this cell-permeable protein, many cells distant from the injection channel are affected by internalization of the effector protein. This experimental approach provides a model to analyze the direct biological responses of an effector protein in the context of intact tissue.
The physiological effects of the activation of the c-Raf-1/ MAP2 signal transduction cascade on the architecture of the muscle tissue were investigated by histochemical analysis. Only in the case of the PreS1/PreS2 protein injection, severe changes of muscle tissue histology could be observed. Typical characteristics of muscle fiber degeneration like basophilic degenerated and necrotic fibers accompanied with infiltration of phagocyting cells could be observed. An increased variability of muscle fiber diameter and many fibers with centrally localized nuclei indicated that also muscle cell regeneration occurs. These data demonstrate that the activation of the c-Raf-1/ MAP2 kinase signal transduction cascade per se is sufficient to induce muscle degeneration and regeneration resembling early stages of muscular dystrophies.
In light of these data we conclude that integrin ␣ 7 deficiency causes a permanent c-Raf-1/MAP2 kinase activation. The activation of MAP2 kinase triggers muscle cell degeneration and regeneration. Therefore, in the integrin ␣ 7 -deficient mice the continuous activation of MAP2 kinase contributes to the development of the dystrophic phenotype.