Distinct signaling pathways are activated in response to mechanical stress applied axially and transversely to skeletal muscle fibers.

In the diaphragm muscle we tested the hypothesis that MAP kinase signaling pathways are activated by mechanical stress and such signaling pathways are dependent on the direction in which mechanical stress is applied. Although equal magnitudes of mechanical stress were applied axially and transversely a greater level of activation of ERK1/2, p38, Raf-1, p90 RSK, Elk-1, and the DNA binding activity of AP-1 transcription factor was produced when the muscle was stretched transversely than when stretched axially. A significant up-regulation in protein tyrosine phosphorylation was observed in axially or transversely loaded diaphragm muscles and the activation of ERK1/2 was completely inhibited by genistein (protein-tyrosine kinase inhibitor). Pretreatment of muscles with wortmannin (phosphoinositide 3-kinase inhibitor), TMB-8 (antagonist of intracellular calcium release), GF109203X (PKC inhibitor), or PD98059 (MEK1/2 inhibitor) blocked the activation of ERK1/2 kinases in response to axial but not to transverse loading. On the other hand, pretreatment of muscles with protein kinase A inhibitors H-7 and KT5720 completely suppressed the activation of ERK1/2 in response to transverse loading only. Taken together with the alterations of MAP kinases and the findings of elevations of downstream transcription targets, our data are consistent with two distinct MAP kinase signal transduction pathways in response to mechanical stress.

In vivo, skeletal muscles function in a three-dimensional mechanical environment in which forces are generated along the muscle fibers and transmitted transversely across the lateral surfaces of the muscle cell, as well as at the ends of the muscle cell (1). Any elastic sheet of muscle that is mechanically loaded in the direction of the muscle fibers may have a lengthtension relationship that is altered from when the sheet is also mechanically loaded in the direction orthogonal to the muscle fibers (2). Force transmission in skeletal muscles varies greatly with muscle fiber architecture (3,4). Muscle fibers may be organized in parallel, in series, or in a complex multipinnate arrangement. Muscle fibers in the mouse diaphragm span the entire length of the muscle and they are arranged in parallel. However, in vivo, the diaphragm unlike most other skeletal muscles is mechanically loaded biaxially. That is, in vivo, the diaphragm is subjected to stresses in the direction of the muscle fibers and in the direction perpendicular to the fibers in the plane of the muscle sheet. Therefore, length tension in the direction of the muscle fibers and transverse to the fibers are important in mediating signal transduction.
The network of mitogen-activated protein (MAP) 1 kinases is a very complex signaling system and it acts in many different physiological processes, such as cellular proliferation, differentiation, inflammatory responses, apoptosis, and development. Recent evidence has shown that mechanical stresses or forces alter the growth and differentiation of a number of tissues and cells (5)(6)(7). The mechanism that relates the mechanical forces with the response of skeletal muscle cells has been investigated recently and several signaling molecules have been demonstrated to play significant roles in mechanical force-initiated signal transduction (8). Among them are the MAP kinases, which are activated by mechanical stresses. Activation of MAP kinases may link the effects of mechanical stress to the biochemical responses and gene expression (9 -11). The MAP kinases act via the regulation of several transcription factors initiating the expression of a variety of immediate and delayed response genes (12,13). In mammalian cells, three parallel pathways of MAP kinases have been described, i.e. extracellular signal related kinase (ERK1/2), protein kinase 38 (p38), and c-Jun-N-terminal kinases (JNKs). These kinases require dual phosphorylation on threonine and tyrosine residues by specific upstream protein kinase. JNK and p38 are often referred to as stress-activated protein kinases (SAPK1/JNK and SAPK2/ p38), and ERK1/2 are often described as the kinases involved in growth factor stimulation (14,15).
The physiologic significance of activation of MAP kinases by mechanical forces is not yet understood. In particular, it is not known whether in skeletal muscles mechanical forces activate specific transcription factors, e.g. activator protein-1 (AP-1). AP-1 is a redox-sensitive transcription factor that plays an important role in cell proliferation, differentiation, and response to various environmental stimuli (16,17). AP-1 consists of a homodimer and heterodimers of members of the Jun family (c-Jun, JunB, and JunD) and heterodimers of the Jun and Fos (c-Fos, FosB, Fra1, and Fra2) families (18). Activation of AP-1 leads to its binding to the DNA at consensus sites that in turn regulate the transcription of AP-1-dependent genes (18,19). Accumulating evidence suggests that the activity of AP-1 is regulated at least in part by the MAP kinase signaling pathways (20,21). The transcriptional activity of several components of AP-1 transcription factor such as c-Jun, activating transcription factor 2, c-Fos, and Elk-1 is increased upon their phosphorylation by ERK1/2 and JNKs (20).
Although, recently it has been established that mechanical stress activates the ERK1/2, JNKs, and p38 MAP kinases in skeletal muscle cells (22)(23)(24)(25)(26)(27)(28)(29)(30), the pathways that are involved in the activation of these kinases in response to mechanical stress remain unknown. Furthermore, most of the studies on the activation of MAP kinases were carried out in vitro using hind limb muscle cells, which are loaded in vivo along the direction of the long axis of the muscle fibers. In contrast, the diaphragm muscle is mechanically loaded in vivo with transdiaphragmatic pressure, and therefore, its muscles are subjected to mechanical forces not only in the direction of the muscle fibers but also in the direction transverse to the fibers (1). It is therefore important to investigate the effects of mechanical forces applied to the diaphragm muscle axially and transversely. In the present study we investigated the mechanism of activation of ERK1/2 in response to constant static stretch applied either in the axial direction of the muscle fibers or in the transverse direction of the fibers. Additionally, we also report the effects of mechanical stretch on the activity of AP-1 and Elk-1 transcription factors in muscle fibers of the diaphragm.
Mice and Tissue Preparation-Mice (SV129 strain) were housed and fed in stainless steel cages on a 12 h on and 12 h off lighting schedule. The animal facility is a virus-free facility. Experimental protocols have been approved by the Animal Protocol Review Committee of the Baylor College of Medicine Animal Program (animal welfare assurance number A-3823-01) and are assigned protocol number AN-1727. All procedures were conducted in strict accordance with the United States Public Health Service animal welfare policy. Mice weighing 21-24 g were anesthetized with an intravenous injection of pentobarbital (0.5-0.7 ml/kg). The diaphragmatic muscle was excised from each animal and immediately immersed into a muscle bath containing a modified Krebs-Ringers solution (in mM: 137 NaCl, 5 KCl, 1 NaH 2 PO 4 , 24 NaHCO 3 , 2 CaCl 2 , 1 MgSO 4 , pH 7.4) bubbled with 95% O 2 , 5% CO 2 . The solution was maintained at a temperature of 25°C throughout the muscle preparation and experimental phase of this study. We excised hemidiaphragms that included a portion of lower ribs and the central tendon. To obtain the length-tension relationship, four silk suture position markers (7-0 or 8-0 Surgilene) were sutured on the abdominal surface of the left costal hemidiaphragm. To minimize the boundary effects, all markers were placed in the central region of the muscle. The markers were placed in an approximate 1-mm 2 square configuration, two pairs of marker were aligned in the direction along the fibers and two pairs were aligned in the direction transverse to the muscle fibers.
Ex Vivo Mechanical Loading of the Diaphragm-We used a biaxial tissue testing apparatus to apply mechanical stress to the tissue along the muscle fiber direction as well as transverse to the fiber direction. The detailed description of this apparatus was described previously (1). Using the entire costal muscle of the left hemidiaphragm mechanical loading was applied by passively stretching the diaphragm in the fibers direction by applying passive tension of about 0.4 N/cm. This is equivalent to a passive mechanical stretch of about 50% from the unstressed state. Unstressed length or length of excised muscle is the shortest length of the muscle and it is equivalent to the in vivo length of the diaphragm at total lung capacity. Optimal length is about 125% of this unstressed length. Therefore, our stretch of 50% places the diaphragm at about 120% of optimal muscle length. In a different mechanical loading protocol we stretched the diaphragm muscle in the direction transverse to the muscle fibers by applying tension of 0.4 N/cm. This is equivalent to a mechanical stretch of about 20% from the unstressed state. In each of these two protocols the muscle was held in the stretched state for 15 min. The total time postmortem did not exceed 1 h.
Thickness Measurements and Data Acquisition-Unstressed muscle thickness measurements were obtained from the excised diaphragm and a digital image of the muscle surface was created as described previously (31). Briefly, the surface area of the excised tissue was determined by using Image Tool (version 2.0, ddsdx.uthscsa.edu). Excess water was gently removed from the surface of the tissue with a cotton-tipped swab, and the tissue sample was immediately weighed. Thickness was then computed as t ϭ m/Ad, where t is muscle thickness in cm, m is muscle mass in grams (g), A is the surface area of the muscle sample in cm 2 , and d is muscle density Ϸ1.06 g/cm 3 (31). Each loading cycle began at the unstressed length (defined as the muscle length that produced no passive tension). Stress was computed as applied tension divided by muscle thickness of the unstressed tissue (stress ϭ tension/ thickness), where stress is in N/cm 2 . Tension is computed by the measured force in grams and divided by the width of the tissue clamp and is measured in N/cm. In the direction of the muscle fibers clamp width was 10 mm, whereas muscle width in the transverse fiber direction was 4 mm. The force data for either axis was collected at a sample rate of 10 Hz using a data acquisition board (model Lab-PC-1200/AI, National Instruments) and LabVIEW software (version 4.0). The force data were stored in an ASCII file for post-analysis of the length-tension relationships.
Strain Calculations-Mechanical strains were calculated based on methods established in our previous work (32). Briefly, coordinates for any three markers are denoted x i and y i (i ϭ 1, 2, or 3), where three markers define a triangle in a plane. Subsequently, the four-sutured markers would define four triangles. For example, triangle A would be defined by markers 1, 2, and 3 whereas triangle B would be defined by markers 1, 2, and 4. The displacements of the markers from an unstressed state to higher stressed states of loading are denoted u i for along the fibers and v i for transverse to the fibers, for example, as follows.
This equation, with known values of the displacements and the coordinates of three markers substituted for u i , x i , and y i , provides a set of three equations for the three coefficients a 1 , a 2 , and a 3 . Similarly, the coordinates of the markers provide information for the following equation.
v i ϭ a 4 ϩ a 5 x i ϩ a 6 y i (Eq. 2) The values of the coefficients a 2 , a 3 etc. were used to find the partial derivatives that were substituted into the following equations defined to calculate strains.
x ϭ ␦u/␦x (Eq. 3) y ϭ ␦v/␦y (Eq. 4) xy ϭ ␦u/␦y ϩ ␦v/␦x (Eq. 5) ⑀ x denotes strain developed along the fibers, ⑀ y denotes strain developed transverse to the fibers, and ⑀ xy denotes shear strain. Strains were computed relative to unstressed length. Unstressed length was defined as the length of the markers in the absence of applied forces. The final strains computed as the average of the strains from the four triangles. For small strains, ⑀ x and ⑀ y are equivalent to the fractional change of muscle fiber length relative to the unstressed length in their respective direction. Our mechanical data are plotted using extension ratios (1.0 ϩ strain values). Assay of ERK1/2, p90 RSK, Elk-1, p38, PKC, and Raf-1-After muscles were stretched for 15 min, samples were washed with phosphatebuffered saline and homogenized in lysis buffer A (20 mM HEPES, pH 7.4, 2 mM EDTA, 250 mM NaCl, 0.3% Nonidet P-40, 2 g/ml leupeptin, 2 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 0.5 g/ml benzamidine, 1 mM dithiothreitol, and 1 mM sodium orthovanadate, 50 mM sodium fluoride, 10 mM ␤-glycerophosphate). The protein concentration of each sample was measured using the Bio-Rad protein assay reagent. An 80-g aliquot of protein was resolved on each lane on 10% SDS-PAGE, electrotransferred onto nitrocellulose membrane, and probed with the phosphospecific anti-p44/42 MAPK (Thr 202 /Tyr 204 ) or anti-p38 (Thr 180 /Tyr 182 ) MAP kinase or anti-Raf-1 (Ser 259 ), anti-Elk-1 (Ser 383 ), or p90 RSK (Ser 380 ) antibodies (Cell Signaling Technology, Inc.) raised in rabbits and detected by chemiluminescence (ECL, Amersham Biosciences). The bands obtained were quantitatively assessed with Personal Densitometer Scan version 1.30 using ImageQuant software version 3.3 (Amersham Biosciences). Total amount of these kinases in the cell extracts was determined by using antibodies, which bind to both phosphorylated and unphosphorylated proteins.
c-Jun Kinase Assay-The c-Jun kinase assay was performed by a method described earlier with minor modifications (33). After the application of mechanical stress to the muscles for 15 min at 25°C the muscles were washed with phosphate-buffered saline and cell extracts were prepared by grinding the muscle in lysis buffer containing 20 mM HEPES, pH 7.4, 2 mM EDTA, 250 mM NaCl, 0.3% Nonidet P-40, 2 g/ml leupeptin, 2 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 0.5 g/ml benzamidine, and 1 mM dithiothreitol. Cell extracts (600 -700 g of protein/sample) were immunoprecipitated with 0.5 g of anti-JNK1 antibody for 4 -6 h at 4°C. Immune complexes were collected by incubation with protein A-Sepharose beads for 1 h at 4°C. The beads were washed three times with lysis buffer followed by two additional washes with kinase buffer (20 mM HEPES, pH 7.4, 1 mM dithiothreitol, 50 mM NaCl). Kinase assays were performed for 15 min at 30°C with GST-Jun-(1-79) as a substrate (2 g/sample) in 20 mM HEPES, pH 7.4, 10 mM MgCl 2 , 1 M dithiothreitol, and 10 Ci of [␥-32 P]ATP. Reactions were stopped with the addition of 20 l of 2ϫ SDS sample buffer, boiled for 5 min, and subjected to 10% SDS-PAGE. GST-Jun-(1-79) was visualized by staining with Coomassie Brilliant Blue, and the dried gel was exposed to x-ray film to see the radioactive bands of GST-c-Jun-(1-79).
Electrophoretic Mobility Shift Assays-To determine AP-1 activation, electrophoretic mobility shift assays were carried as described previously (34) with some modification. The muscles, after being mechanically stretched, were weighed frozen using dry ice and ground using a manual grinder in low salt lysis buffer containing 10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2.0 g/ml leupeptin, 2.0 g/ml aprotinin, and 0.5 mg/ml benzamidine (18 l buffer/milligram of muscle tissue). The tissue extract in low salt lysis buffer was allowed to swell on ice for 10 min, vortexed vigorously for 10 s, followed by two freezethaw cycles. The homogenate was centrifuged for 10 s at 4°C, the supernatant was removed, and the nuclear pellet was resuspended (5 l/mg of original tissue weight) in ice-cold nuclear extraction buffer (20 mM HEPES, pH 7.9, 420 mM NaCl, 1 mM EDTA, 1 mM EGTA, 10 mM MgCl 2 , 20% glycerol, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2.0 g/ml leupeptin, 2.0 g/ml aprotinin, 0.5 mg/ml benzamidine) and incubated on ice for 30 min with intermittent vortexing. Samples were centrifuged for 5 min at 4°C, and the supernatant (nuclear extract) was either used immediately or stored at Ϫ70°C. The protein content was measured using the Bio-Rad protein assay reagent. Electrophoretic mobility shift assays were performed by incubating 20 g of nuclear extract with 16 fmol of the 32 P-end-labeled AP-1 consensus oligonucleotide 5Ј-CGCTTGATGAGTCAGCCGGAA-3Ј (underline indicates AP-1 binding sites) for 15 min at 37°C. The incubation mixture included 2.5 g of poly(dI⅐dC) in a binding buffer (25 mM HEPES, pH 7.9, 0.5 mM EDTA, 0.5 mM dithiothreitol, 1% Nonidet P-40, 5% glycerol, 50 mM NaCl). The DNA-protein complex thus formed was separated from free oligonucleotide on 7.5% native polyacrylamide gels using buffer containing 50 mM Tris, 200 mM glycine, pH 8.5, and 1 mM EDTA. The gel was then dried and the radioactive bands were visualized by exposing the gel to the x-ray film for a suitable amount of time. The bands obtained were quantitatively assessed with Personal Densitometer Scan version 1.30 using ImageQuant software version 3.3 (Amersham Biosciences). The specificity of the AP-1 band was confirmed by cold competition assays.
Statistical Analysis-All experiments were repeated at least three times unless otherwise indicated. Results are expressed as mean Ϯ S.D. Statistical analysis used Student's t test or analysis of variance to compare quantitative data populations with normal distribution and equal variance. A value of p Ͻ 0.05 was considered statistically significant unless specified.

Stress-Strain Relationship Along and
Transverse to the Muscle Fibers-A representative set of uniaxial length tension relationships of the left costal muscle sheet of the diaphragm in either the muscle fiber direction or transverse to the muscle fibers is shown in Fig. 1. These data demonstrate that tension along the muscle fibers and transverse to the fibers increase continuously with increasing mechanical stretch. However, the length tension of the diaphragm in the axial direction is shifted to the right compared with the length-tension curve in the transverse plane, that is, the normal diaphragm muscle has a greater extensibility in the direction of the muscle fibers than transverse to the fibers. The nonlinear behaviors in the two directions are somewhat different. The stress-strain relation in the muscle direction is only mildly nonlinear over the sizable range of strains that our data cover. The nonlinear stiffening in the transverse direction appears to be more pronounced. Using all normal diaphragms, we computed the average extension ratio, , at tension of ϳ0.4 N/cm along the fiber direction and found to be greater in the direction of the muscle fibers than transverse to the fibers (along fibers, ϭ 1.5; transverse to fibers, ϭ 1.2).
Activation of ERK1/2 and p90 RSK by Mechanical Stretch-Because the diaphragm muscle in the transverse direction of the muscle fibers is less compliant as compared when loaded axially in the fiber direction ( Fig. 1), we investigated the effect of mechanical stress applied axially as well as transversely on the activation of MAP kinases. Our preliminary experiments showed that mechanical stress of the diaphragm either axially or transversely activates the MAP kinases as early as 5 min at the onset of mechanical stress with maximum activation occurring at 15 min of mechanical loading (data not shown). Therefore, in all MAP kinase assays we stretched the muscles for 15 min and then the tissue was immediately lysed in lysis buffer. As assayed by phosphospecific immunoblot analysis, stretching of the muscles either axially or transversely resulted in the induction of ERK1/2 phosphorylation ( Fig. 2A). However, the level of phosphorylation of ERK1/2 proteins in response to the applied load appears to be dependent on the direction along which the mechanical load was applied. Phosphorylation level of ERK1/2 was significantly greater in response to mechanical stretch applied in the transverse direction of the muscle fibers than that applied in the direction of the muscle fibers. Muscle length is expressed as a fraction of muscle unstressed length (length at which no passive force is applied). The uniaxial data along the direction of the muscle fibers is shown with E and q. The open circles indicate the stretching of the muscle tissue along the fibers, and the closed circles indicate the unloading in the direction along muscle fibers. Uniaxial data for loading transverse to the direction of the muscle fibers are shown with छ and ϩ. The diamonds indicate the stretching curve of the muscle transverse to fibers, and crosses indicate the unloading curve. The data demonstrate that extensibility and compliance is much smaller when stresses are applied in the transverse direction of the muscle fibers than when applied in the axial direction of the fibers.
Because ERK1/2 phosphorylate cytosolic protein p90 RSK (35,36) we also studied the effect of mechanical stress on the phosphorylation level of p90 RSK protein. As shown in Fig. 2B (upper panel), mechanical stress of muscles led to an elevated level of phosphorylation of p90 RSK with a significantly higher level of phosphorylation occurring in those muscles stretched in the direction transverse to the muscle fibers. The increased phosphorylation level of ERK1/2 and p90 RSK in response to mechanical stress was not attributed to an alteration in the level of these proteins in muscle tissues, because the total amount of these proteins before and after the application of mechanical stress remained essentially the same (Fig. 2, A and B, lower panels). Because mechanical stress was applied equally in the direction of the muscle fibers and transverse to the fibers (refer to "Experimental Procedures") the difference in the level of activation of ERK1/2 and p90 RSK was not because of the difference in the magnitude of mechanical stress between stress applied in the fiber direction and stress applied in the transverse direction to the muscle fibers. These data thus clearly show that mechanical stresses activate the ERK1/2 and p90 RSK to different levels depending on the direction in which mechanical stresses are applied.
Effect of Mechanical Stretch on the Activity of JNK1 and p38 MAP Kinase-Because the level of phosphorylation of ERK1/2 was dependent on whether mechanical stretch was applied to the diaphragm muscle axially or transversely, we also studied the effect of mechanical stretch applied in either direction on the activity of JNK1 and p38 MAP kinase. The JNK1 activity was assayed by phosphorylation of the GST-c-Jun protein. Interestingly, the mechanical loading of the muscles led to a drastic increase in the JNK1 activity, however, in contrast to ERK1/2, there was no difference in the level of activation of JNK1 between the axially stretched and those transversely stretched diaphragm muscles (Fig. 3A). The activity of p38 kinase was also increased in response to both axial and transverse stress with significantly higher levels of activation in those transversely loaded diaphragms (Fig. 3B). The effects of mechanical loading on MAP kinase activation were specific, as the activity of a cell cycle kinase cyclin-dependent kinase-2 (cdk2) was not altered (Fig.  3C). Mechanically stretching the diaphragm muscle either axially or transversely did not affect the total cellular level of JNK1, p38, or cdk2 (data not shown).
PD98059 Inhibits the Phosphorylation of ERK1/2 and p90 RSK in Response to Axial Loading but Not to Transverse Loading-P98059 is a specific inhibitor of the MAP kinase kinase (MEK1/ 2), a kinase that phosphorylates ERK1/2 in the classical ERK1/2 signaling pathway (37). We investigated the effect of PD98059 on the phosphorylation of ERK1/2. Published reports have shown that 25-50 M PD98059 can completely inhibit the activity of MEK1/2 kinases in all cells including skeletal muscles (37,38). We found that pretreatment of the diaphragm for 30 min with 50 M PD98059 completely inhibited the phosphorylation of ERK1/2 in axially loaded muscles (Fig. 4A). On the other hand, there was no effect of PD98059 on the level of phosphorylation of ERK1/2 kinases in the transversely loaded muscles. These results show that MEK1/2 are involved in the activation of ERK1/2 in response to axial stress, whereas different mechanisms of signaling are involved in the activation of ERK1/2 in response to transverse stress. PD98059 did not have any significant effect on the activation of JNK and p38 MAP kinases (data not shown). The effect of PD98059 on the phosphorylation level of p90 RSK in the diaphragm was also investigated. Interestingly, pretreatment of the diaphragm with PD98059 completely inhibited the phosphorylation of p90 RSK in those muscles that were axially loaded, whereas the p90 RSK phosphorylation in response to transverse loading remained unchanged (Fig. 4B). These results suggest that p90 RSK is only phosphorylated by an ERK1/2-dependent mechanism in response to axial mechanical stress. It further confirms that MEK1/2 are not involved in the activation of The diaphragm muscles were exposed to different passive mechanical loadings for 15 min followed by their lysis in lysis buffer. Equal amounts of proteins were subjected to immunoblot analysis with phosphorylated ERK1/2 and p90 RSK antibodies. The total amount of the proteins was determined by using normal antibodies against ERK1/2 and p90 RSK proteins. A representative blot and the quantitation of multiple experiments shown here shows that loading of muscles along or transverse to the direction of fiber orientation activates ERK1/2 and p90 RSK (top panels) and an equal amount of ERK1/2 and p90 RSK protein was present in each lane (bottom panels). ERK1/2 in response to transverse loading of the diaphragm.
Activation of Raf-1 in Response to Axial or Transverse Stress-Raf-1, a serine/threonine protein kinase is an important component of the mitogen-and mechanical-induced signal transduction response (4,14). To further delineate the pathways involved in the activation of MAP kinases in response to mechanical loading, the effect of mechanical stress on the activation of Raf-1 kinase was also studied. Although not statistically significant an equal amount of stress applied axially and transversely seems to activate the Raf-1 kinase with somewhat higher level of activation (not statistically significant) in transversely loaded muscle (Fig. 5).
Protein Tyrosine Phosphorylation Is Required for the Activation of ERK1/2 in Response to Mechanical Loading-Protein tyrosine phosphorylation by both receptor and nonreceptor proteintyrosine kinases is known to trigger downstream signaling events in response to various stimuli (39). We next investigated whether mechanical loading of diaphragm muscle can alter the protein tyrosine phosphorylation in diaphragm muscle. Immunoblotting with phosphotyrosine monoclonal antibody showed that a significant increase in phosphorylation of approximately 67-kDa protein was observed in response to both axial and transverse loading of diaphragm muscles (Fig 6A). The phosphorylation level was much higher in transversely loaded diaphragm muscle as compared with those loaded axially. Furthermore, an additional protein of approximately 85 kDa seems to be phosphorylated in response to transverse loading only (Fig. 6A).
To investigate whether the protein tyrosine phosphorylation is required for the activation of ERK1/2 in response to mechanical loading, the diaphragm muscles were preincubated for 45 min with 50 M genistein (a protein-tyrosine kinase inhibitor) followed by mechanical stretching for 15 min. As shown in Fig.  6B, pretreatment of diaphragm muscles with genistein appears to completely inhibit the activation of ERK1/2 in response to both axial and transverse loading. These data strongly suggest that protein-tyrosine kinases are involved in the activation of ERK1/2 in response to both axial and transverse mechanical stress.
Phosphoinositide 3-Kinase (PI3K) Is Involved in Activation of ERK1/2 in Response to Axial Loading-Using a specific and potent inhibitor of PI3K, wortmannin, we investigated the role of PI3K in the activation of ERK1/2 in response to mechanical loading of the diaphragm. Pretreatment of muscles with 400 nM wortmannin resulted in a significant inhibition in the phosphorylation of ERK1/2, whereas wortmannin did not have any effect on the phosphorylation level of ERK1/2 in response to transverse loading (Fig. 7).

Role of Intracellular Ca 2ϩ , Calmodulin-dependent Protein Kinase II (CaMKII), and Stretch-activated (SA) Channels in Activation of ERK1/2 in Response to Mechanical Loading-
Ca 2ϩ ions mediate a large number of cellular responses by binding to specific intracellular proteins, which may be considered Ca 2ϩ receptors (40,41). The Ca 2ϩ concentration in the cells in response to mechanical stress is increased by either release of Ca 2ϩ ions from intracellular stores and/or by Ca 2ϩ influx from the extracellular environment mainly through SA channels (7,42). Pretreatment of diaphragm muscle with TMB-8 (an intracellular Ca 2ϩ antagonist) suppressed the activation of ERK1/2 in response to axial loading but not to transverse loading (Fig. 8A). The CaMKII appears not to be involved in the activation of ERK1/2 in response to either axial or transverse stress, as the inhibition of CaMKII activity by KN-93 (a specific inhibitor of CaMKII) did not have any effect on the activation of ERK1/2 (Fig. 8, right panel). Furthermore, pretreatment of diaphragm muscle with 20 M gadolinium III chloride (an SA channel blocker) or removing the extracellular Ca 2ϩ from the physiologic Krebs-Ringer solution did not affect the activation of ERK1/2 in response to either axial or transverse mechanical stress (data not shown). This indicates that either SA channels or extracellular Ca 2ϩ influx is not required for the stress-induced activation of ERK1/2 in diaphragm muscles.

FIG. 3. Effect of mechanical loading of diaphragm muscles on the activity of JNK1 (A) and p38 MAP kinase (B), and cyclin-dependent kinase 2 (cdk2) (C) activity.
Diaphragm muscles were stretched for 15 min in the direction along or transverse to the fiber orientation. Muscles were then immediately lysed in the lysis buffer. The JNK1 and cdk2 activity were determined by immunoprecipitating equal amounts of proteins with anti-JNK1 and anti-cdk2 antibodies, respectively, followed by in vitro kinase assays using GST-c-Jun-(1-79) for JNK1 or histone H1 for cdk2 as substrates. The activity of p38 MAP kinase was determined by immunoblot analysis with phosphospecific p38 MAP kinase antibodies. The result presented here shows that both longitudal and transverse loading of the diaphragm muscle activate JNK1 and p38 MAP kinase, whereas no effect was observed on the activity of cdk2.

Protein Kinase C Is Required for the Activation of ERK1/2 in Response to Axial Whereas Protein Kinase A (PKA) Is Needed for Activation of ERK1/2 in Response to Transverse Loading-
One of the early steps in cell signaling is the activation of various isoforms of PKC and PKA (43)(44)(45). Because intracellular Ca 2ϩ mobilization was required for the activation of ERK1/2 in response to axial loading, we investigated whether the protein kinase C or protein kinase A have any role in mechanotransduction in skeletal muscle. Interestingly, H-7, an inhibitor of protein kinase A completely blocked the activation of ERK1/2 in response to transverse loading while there was only a marginal inhibition in response to axial loading (Fig. 9A). Because H-7 is also known to inhibit the activity of PKC at higher concentrations, we used KT5720, a very specific inhib-itor of PKA having no effects on PKC to study the activation of ERK1/2, and found that KT5720 only inhibited the activation of ERK1/2 in response to transverse loading but not to axial loading, suggesting that PKA is required for the activation of ERK1/2 in response to transverse stress (Fig. 9B). GF109203X, a specific inhibitor of PKC, on the other hand inhibited the activation of ERK1/2 only in response to axial loading, and there was no effect on the activation of ERK1/2 in response to transverse loading (Fig. 9C). These data thus suggest that protein kinase C activation is required for the activation of ERK1/2 in response to axial stress, whereas PKA is needed in transverse stress.
Activation of AP-1 and Elk-1 Transcription Factors by Mechanical Stretch-Our data demonstrate a greater level of ac- FIG. 4. Effect of PD98509 on the activation of ERK1/2 (A) and p90 RSK (B). The diaphragm muscles were treated for 30 min in 50 M PD98509 followed by stretching of the muscles either axially or transversely for 15 min. The muscles were lysed in lysis buffer and equal amounts of proteins were subjected to immunoblot analysis using phospho-ERK1/2 and phospho-p90 RSK antibodies. The representative blot and the quantitation of multiple experiments using the densitometer presented here show that PD98059 inhibited the activation of ERK1/2 and p90 RSK in response to axial stress but not in response to transverse stress. tivation of MAP kinases in response to transverse stretch than axial stretch (Figs. 2 and 3). We wondered if such activation would lead to the activation of transcription factors that are the phosphorylation targets of MAP kinases. Using the electrophoretic mobility shift assay, we investigated the effect of mechanical stretch of diaphragm muscle sheets in either the fiber or transverse fiber direction on the activation of the AP-1 transcription factor. As shown in Fig. 10 an increased DNA binding activity of AP-1 was observed as a result of mechanical stress. Furthermore, the level of activation of AP-1 was greater in the transversely loaded muscles as compared with the axially loaded muscles.
We also investigated the effect of mechanical stress on the activation of Elk-1. As demonstrated in Fig. 11, Elk-1 is strongly activated when the diaphragm muscle sheet was stretched in the transverse fiber direction than when stretched in the axial direction. These results thus show that a greater level of activation of MAP kinases in the transversely loaded diaphragms is linked to activation of additional downstream targets of this signal transduction pathway. DISCUSSION We show that the diaphragm muscle responses to mechanical stress by activating the ERK1/2, JNKs, and p38 MAP kinases. Furthermore, the mechanical stress activates the Raf and the p90 ribosomal S6 kinase (p90 RSK) as well as the transcription factors AP-1 and Elk-1. We also show that there are two distinct signaling pathways activated in response to mechanical stress applied axially and transversely to muscle fibers of the diaphragm. In particular, our data show that PI3K, PKC, and MEK1/2 are activated when the mechanical stress of the diaphragm is applied axially but not activated when stress is applied transversely. On the other hand, cyclic AMP-dependent PKA is activated only in response to transverse mechanical stress. This is despite the fact that we applied the same magnitude of mechanical stress axially and transversely to the muscle fibers.
Mechanism of Activation of ERK1/2 Signaling Pathways by Axial and Transverse Loading of Diaphragm Muscle-The diaphragm is mechanically loaded in vivo with transdiaphragmatic pressure, and therefore, its muscles are subjected to mechanical forces not only in the direction of the muscle fibers but also in the direction transverse to the fibers. Therefore, it is important to study mechanical signal transduction not only when mechanical stress is applied along the muscle fibers but also when mechanical stress is applied in the transverse direction to the fibers. Our experiments were conducted such that mechanical stress was applied in either the direction of the muscle fibers or transverse to the fibers. In our experimental protocols axial stress applied to the muscle was essentially the same as that applied transversely, and data in Fig. 1 show that this resulted in mechanical stretch that is about twice as much in the axial direction than in the transverse fiber direction. The stress-strain relationships shown in Fig. 1 clearly demonstrate that muscle compliance is small and extensibility is less in the transverse direction to the muscle fibers compared with those in the direction of the muscle fibers. That is when mechanical stress was applied equally in either direction the resistance to mechanical stretch is greater in the transverse direction of the muscle fibers than in the direction of the fibers. The greater FIG. 5. Effect of mechanical loading of diaphragm muscles on the activity of Raf-1. Muscles were mechanically loaded in either the axial or transverse direction of the diaphragm muscle fibers and the muscles were immediately lysed in a lysis buffer at the end of a 15-min constant mechanical stretch. The cell extracts were analyzed by immunoblotting with phospho-Raf-1 antibodies. The representative blot and the quantitation from three experiments show that the Raf-1 activity is increased by either axial or transverse loading with somewhat slightly higher levels of Raf-1 activation in transversely loaded muscles as compared with axially loaded muscles.
FIG. 6. Involvement of protein-tyrosine kinases in mechanical stretch-induced activation of ERK1/2. A, diaphragm muscles were mechanically loaded for 15 min in either the axial or transverse direction of the muscle fibers followed by its lysis in lysis buffer. The cell extracts were analyzed by immunoblotting with phosphotyrosine (Tyr(P)-100) monoclonal antibodies. The representative blot shows that the protein tyrosine phosphorylation is increased upon mechanical loading. B, diaphragm muscles were treated with 50 M genistein for 45 min followed by loading of muscles along or transverse to the direction of the muscle fiber for 15 min. A representative blot and quantitation from two independent experiments show that genistein inhibits the phosphorylation of ERK1/2 in response to either axial or transverse mechanical loading. activation of ERK1/2, p38, and Raf-1 kinases, as well as the transcription factors, AP-1 and Elk-1, in response to transverse mechanical stress are consistent with greater transverse muscle stiffness compared with axial stiffness.
Earlier reports have shown that mechanical tension alone can regulate cell growth and metabolism and the intracellular signaling proteins may link changes in the tension to this cellular adaptation (46). Recently it has been shown that mechanical stress stimulates the signaling pathways of ERKs, JNKs, and p38 MAP kinases in a variety of cell lines including skeletal muscle cells (8,9,22,23,29). In all these studies, however, the investigators applied axial stress to myocytes isolated from hind limb muscles. In agreement with these studies we found that axial mechanical stress activates ERK1/2, JNK1, and p38 MAP kinase. Additionally, we have shown that mechanical stress activates p90 RSK and Elk-1, the downstream targets of ERK1/2.
In recent years, the effect of mechanical axial stress on cell growth and regulation of expression of several genes has been investigated. However, there is little published work directed toward understanding the biochemical signaling pathways initiated by different mechanical stresses. The ERK pathways involve the sequential phosphorylation and activation of serine kinase Raf, MAP kinase (MEK1/2), and two ERK isoforms (ERK1/ERK2). The ERK1/2 phosphorylate and activate cytosolic substrates such as p90 RSK and nuclear transcription factors such as c-Myc, Elk-1, and c-Fos (14,18). The use of specific inhibitors has been a powerful tool in delineation of the downstream MAP signaling in response to numerous stimuli. It has been reported that PD98059 is a highly specific inhibitor of MEK1/2 with very little inhibition reported on other members of MAP kinase family (37,38). Our results using PD98059 suggest the involvement of MEK1/2 in the activation of ERK1/2 only in response to axial stress. Interestingly, PD98059 did not have any effect on the phosphorylation level of ERK1/2 in response to transverse stress indicating that different kinases might be involved in the phosphorylation of ERK1/2 during transverse loading of the diaphragm muscle. This is further supported by a higher level of activation of ERK1/2 in response to transverse stress as compared with axial stress. Although, ERK1/2 have been proposed to be the direct phosphorylation target of the MEK1/2 in the classical signal transduction, this pathway is not a universal phenomenon. Signal transduction in cells involves cross-talk, interconnection between signaling pathways, and shared substrates among multiple enzymes (47,48). The increased phosphorylation of ERK1/2 in response to either form of mechanical stress also led to an increase in the activation of p90 RSK. Similar to ERK1/2 treatment with PD98059 completely inhibited the activation of p90 RSK in response to axial stress, whereas no effect was observed in the case of transverse loading. These results suggest that application of mechanical stress triggers the activation of p90 RSK via the ERK1/2 mechanism. PD98059 did not affect the activity of JNK1 and p38 MAP kinase (data not shown), suggesting that ERK1/2 are the only MAP kinases that are phosphorylated by MEK1/2 in response to axial mechanical stress and the activity of JNKs and p38 MAP kinase is regulated by different pathways. Indeed, the cell signaling pathways that activate JNK1 and p38 MAP kinases in response to various stimuli have been well established (14), however, the pathways that lead to the activation of JNKs and p38 MAP kinase in response to mechanical loading remain unknown.
Despite a great deal of research in the last decade, the mechanotransduction remains poorly understood. In the absence of specific cell surface receptors, it remained unclear how mechanical forces are converted into intracellular signals coupled with nuclear gene expression. Some investigators have proposed that transduction of mechanical signals to biological signals occur via the extracellular matrix/integrin/cytoskeletal axis (7,49). Mechanical stress causes deformation of the sarcolemma, which may directly or indirectly cause conformational changes in protein (and subsequent activation of them) that are anchored to the inner surface of cell membranes, or in transmembrane proteins. Several effectors enzyme such as protein-tyrosine kinases, phospholipases, and protein kinase C isoenzymes, ion channels, such as the SA channels, are examples of sarcolemmal proteins that might be affected by mechanical stresses (7).
To understand the upstream events that lead to the activation of MAP kinases in response to axial and transverse mechanical loading, we investigated the role of several effector molecules and second messengers. Interestingly, there was a significant increase in the protein tyrosine phosphorylation of an approximately 67-kDa protein upon the application of mechanical stress to the diaphragm muscle. The phosphorylation level, however, was much higher in transversely loaded diaphragm muscle as compared with axially loaded muscle. Furthermore, there was an additional tyrosine phosphorylation of an approximately 85-kDa protein in transversely loaded diaphragm muscle (Fig. 6A). The protein tyrosine phosphorylation seems to be required for the activation of ERK1/2 in response to both axial and transverse loading as the inhibition of proteintyrosine kinases activity by genistein abrogated the mechanical stress-induced phosphorylation of ERK1/2 in diaphragm muscle (Fig. 6B). What role protein-tyrosine kinase plays in the activation of ERK1/2 in response to mechanical stress remains a subject of much speculation. Some recent reports, however, suggested that focal adhesion kinase, a nonreceptor tyrosine kinase, localized at focal adhesion is involved in mechanosensing in fibroblast and endothelial cells (50,51). Several other reports also suggested that members of Src family of proteintyrosine kinases such as c-Src constitute a part of the mechanotransduction in endothelial cells in response to shear stress (Ref. 52, and references therein). Although, the nature of pro- FIG. 7. Effect of wortmannin on the activation of ERK1/2 in response to mechanical loading. Diaphragm muscles were treated with 400 nM wortmannin for 30 min followed by loading of muscles along or transverse to the direction of muscle fiber for 15 min. The tissue extracts were made and equal amounts of protein were subjected to immunoblot analysis using phospho-ERK1/2 antibodies. A representative blot and the quantitative data from three independent experiments presented here clearly show that wortmannin inhibits the ERK1/2 phosphorylation when muscles are loaded in the direction of the muscle fibers but not in the transverse direction of the fibers.
tein-tyrosine kinase is not known at present, our data suggest that protein-tyrosine kinase are involved in stress-induced activation of ERK1/2 in the diaphragm muscle.
Another striking observation of the present investigation is the involvement of PKC in response to axial stress whereas PKA was involved in response to transverse stress. The activation of ERK1/2 of the diaphragm muscle in response to applied axial but not transverse mechanical stress requires the intracellular mobilization of Ca 2ϩ ions as antagonist of intra-cellular Ca 2ϩ TMB-8 blocked the activation of ERK1/2 (Fig. 8). ERK1/2 activation in response to either axial or transverse stress does not involve SA channels or the extracellular Ca 2ϩ influx (data not shown). Furthermore, the inhibition of PKC activity also led to the inhibition of ERK1/2 activation in response to axial loading only (Fig. 9C), suggesting that Ca 2ϩ ions plays an important role in mediating the cellular response in axial loading of skeletal muscle fibers. In addition to increasing the activity of PKC, elevation of free Ca 2ϩ level in the cells also results in its binding and activation of calmodulin (CaM), which consequently modulates the activity of protein kinases and phosphatases (40,41). One such enzyme Ca 2ϩ /calmodulindependent protein kinase II (CaMKII) that has broad substrate specificity and mediates various biological responses (53) is, however, not involved because the inhibition of CaMKII with KT-93 did not alter the activation of ERK1/2 in response to mechanical stress (Fig. 8, right panel). for 45 min followed by mechanical loading of muscles in either the fiber or transverse to the fiber direction for 15 min. Consequently we conducted immunoblotting with phospho-ERK1/2 antibodies. Representative data from two independent experiments suggest that PKA inhibitors H-7 and KT5720 inhibit the activation of ERK1/2 in response to transverse stress, whereas PKC inhibitor GF109203X inhibits the activation of ERK1/2 in response to axial stress.
FIG. 10. Effect of mechanical loading of diaphragm muscle on the DNA binding activity of AP-1. Diaphragm muscles stretched along the muscle fibers or transverse to the fibers were subjected to extraction of nuclear proteins. Equal amounts of nuclear proteins were allowed to bind with 32 P-labeled oligonucleotide containing AP-1 consensus sites. The DNA-protein complex was analyzed using the electrophoretic mobility shift assay as described under "Experimental Procedures." The loading of muscles along the muscle fibers (lane 3) as well as transverse to the fibers (lane 5) increased the DNA binding activity of AP-1. Furthermore, there is a higher level of activity when muscle was loaded transversely than when muscle was loaded axially. The specificity of the bands was confirmed by using an excess of unlabeled oligonucleotide in the DNA-protein complex of transversely loaded muscles.
In contrast to axial loading, the activation of Ca 2ϩ ions or PKC is not required for the activation of ERK1/2 in response to transverse loading. Instead, cyclic AMP-dependent PKA is needed for the activation of ERK1/2 in response to transverse loading. PKA mediates several of the effects of the elevated levels of cAMP in the cells and has been reported to both activate and inhibit the ERK1/2 signaling pathway either directly or through the activation of Rap1 (45). Although both PKA and PKC can modulate the activity of ERK1/2 pathways by phosphorylation of Raf kinases, PKA preferentially targets B-Raf, whereas PKC alters the activity of the Raf-1 isoform (7,45,54,57).
PI3K, a lipid kinase complex, is another important regula-tory protein that is involved in the initiation of different signaling pathways and regulating the major functions of the cell (55,56). The precise role of PI3K in MAP kinase signaling pathways is not very clear. Both PI3K-dependent and PI3Kindependent activation of MAP kinases has been reported (57-61). Ikeda et al. (62) have shown that the activation of ERK1/2 in endothelial cells in response to cyclic strain is dependent on the PI3K and p21 ras . Similar to these findings our results with wortmannin suggest that the activation of ERK1/2 pathways in response to axial stress of the diaphragm involves PI3K (Fig. 7). However, the same is not true in the case of transverse loading. We did not get any inhibition in the activation of ERK1/2 in response to transverse loading by wortmannin even when it was used at higher concentrations (up to 1 M, not shown). How PI3K regulates the activation of ERK/12 in response to axial mechanical stress remains speculative. The main effectors of PI3K include the mitogen transducing signaling proteins (protein kinase C, phosphoinoside-dependent kinases, and small G-proteins) that are activated either via their interaction with the lipid products of PI3K such as phosphatidylinositol 3,4,5-P 3 or PI3K-dependent phosphorylation of proteins. Activation of PI3K activates phosphoinositide-dependent kinase-1, which can directly or along with phosphatidylinositol 3,4,5-P 3 modulate the activity of various isoform of PKC (7,44,57). There are also reports that suggest that lipid products of PI3K can cause the release of Ca 2ϩ stores and hence activate the Ca 2ϩ -dependent signaling pathways (42,44). Activation of AP-1 Transcription Factor by Mechanical Stress-The present study for the first time demonstrates that mechanical stress can activate the DNA binding activity of the AP-1 transcription factor in skeletal muscles. Although the role of AP-1 in mechanical stress response of skeletal muscles is still recondite, the activity of AP-1 has been shown to be very important for various cellular responses (18) and may have a role in mechanotransduction through modulation of the late load-response genes through an as yet undefined transcription cascade mechanism. Indeed, increased mRNA levels of several FIG. 11. Effect of mechanical loading of diaphragm muscles on the activity of Elk-1. Diaphragm muscles were subjected to mechanical stress applied either axially in the muscle fiber direction or transverse to the fibers. The muscles were immediately lysed in a lysis buffer. The activity of Elk-1 was determined by immunoblot analysis with phospho-specific Elk-1 antibodies. The data shown here is representation of three experiments carried out independently.
FIG. 12. Possible mechanism of activation of ERK1/2 and AP-1 in skeletal muscles. Schematic representation of the potential mechanism of activation of ERK1/2 and AP-1 transcription factor in response to mechanically loading of skeletal muscles either axially or transversely as supported by the data. Despite activation of ERK1/2 by either form of mechanical stress, the MEK1/2, PI3K, and PKC are involved in the activation of ERK1/2 in response to axial mechanical stress whereas PKA is involved only in response to applied transverse stress. genes including insulin-like growth factor I (63), fibroblast growth factor (64), myogenic regulatory growth factor (65), c-Jun (66), and smpx (67) have been reported in response to mechanical stretch in skeletal muscle cells. Vandenburgh et al. (68) have shown that the activity of cyclooxygenase increases within 4 h of initiating mechanical stretch in skeletal muscle cells. Similarly, the expression of the nitric-oxide synthase (NOS) gene is augmented following mechanical stimulation of rat skeletal muscle cells (69). The promoter region in either genes of COX and NOS has been reported to contain the consensus binding sites for the AP-1 transcription factor (70 -72). The activation of AP-1 by mechanical stress thus suggests that one mechanism by which mechanical loading could lead to the increased expression of either COX or NOS is through the activation of AP-1 transcription factor.
The exact mechanism of activation of AP-1 in response to mechanical stress is not understood, however, we observed a significantly higher level of AP-1 activity in transversely loaded muscles as compared with those loaded axially. AP-1 is a collective term referring to dimeric transcription factors composed of Jun, Fos, Elk-1, or activating transcription factor subunits that bind to a common DNA site, the AP-1-binding site (18). The transcriptional activity of several AP-1 subunits is increased upon their phosphorylation by MAP kinases, especially ERK1/2 and JNK1 (20,73). We also found that the phosphorylation level of Elk-1, a physiological substrate of ERK1/2, is increased by application of axial or transverse loading (Fig. 11). It is therefore possible that the increased activation of AP-1 in response to mechanical stress is a direct result of the activation of MAP kinases. Based on our findings in this report we provide here a schematic model of the activation of MAP kinases and AP-1 transcription factor (Fig. 12). The schematic show clearly that there are two distinct signaling pathways in the diaphragm as function of the direction of applied mechanical stress. It is very possible that those distinct pathways exist in all skeletal muscles that are mechanically loaded in vivo with axial and transverse stress and those include the diaphragm, abdominal wall, and intercostal muscles.
In summary, our study provides the first evidence that mechanical signaling is directional in the skeletal muscle cells of the diaphragm and also demonstrates an up-regulation in the activity of specific transcription factors in response to mechanical stress. Because most of the intracellular signaling pathways lead to an altered gene expression, our results support the hypothesis that gene expressions can be differentially regulated depending on the direction in which mechanical forces are applied.