Tumor Necrosis Factor-related Weak Inducer of Apoptosis Augments Matrix Metalloproteinase 9 (MMP-9) Production in Skeletal Muscle through the Activation of Nuclear Factor-κB-inducing Kinase and p38 Mitogen-activated Protein Kinase

Destruction of skeletal muscle extracellular matrix is an important pathological consequence of many diseases involving muscle wasting. However, the underlying mechanisms leading to extracellular matrix breakdown in skeletal muscle tissues remain unknown. Using a microarray approach, we investigated the effect of tumor necrosis factor-related weak inducer of apoptosis (TWEAK), a recently identified muscle-wasting cytokine, on the expression of extracellular proteases in skeletal muscle. Among several other matrix metalloproteinases (MMPs), we found that the expression of MMP-9, a type IV collagenase, was drastically increased in myotubes in response to TWEAK. The level of MMP-9 was also higher in myofibers of TWEAK transgenic mice. TWEAK increased the activation of both classical and alternative nuclear factor-κB (NF-κB) signaling pathways. Inhibition of NF-κB activity blocked the TWEAK-induced production of MMP-9 in myotubes. TWEAK also increased the activation of AP-1, and its inhibition attenuated the TWEAK-induced MMP-9 production. Overexpression of a kinase-dead mutant of NF-κB-inducing kinase or IκB kinase-β but not IκB kinase-α significantly inhibited the TWEAK-induced activation of MMP-9 promoter. The activation of MMP-9 also involved upstream recruitment of TRAF2 and cIAP2 proteins. TWEAK increased the activity of ERK1/2, JNK1, and p38 MAPK. However, the inhibition of only p38 MAPK blocked the TWEAK-induced expression of MMP-9 in myotubes. Furthermore the loss of body and skeletal muscle weights, inflammation, fiber necrosis, and degradation of basement membrane around muscle fibers were significantly attenuated in Mmp9 knock-out mice on chronic administration of TWEAK protein. The study unveils a novel mechanism of skeletal muscle tissue destruction in pathological conditions.

Skeletal muscle atrophy/wasting has highly complex physiology that involves not only intracellular protein degradation but also several extracellular alterations including the breakdown of skeletal muscle basement membrane and intramuscular connective tissues (1,2). However, the identity of the molecules that cause extracellular matrix degradation in skeletal muscle during pathological conditions remains unknown. Matrix metalloproteinases (MMPs) 2 are a family of extracellular matrix-degrading enzymes that are collectively capable of degrading all the components of the extracellular matrix (3,4). These proteases are synthesized as secreted or transmembrane proenzymes and processed to an active form by the removal of an amino-terminal propeptide (3)(4)(5). MMP-2 (gelatinase A) and MMP-9 (gelatinase B) are the two major MMPs that can degrade several matrix proteins including type IV collagen, an integral component of the basement membrane (3)(4)(5). Excessive production of MMP-2 and MMP-9 has been found to be associated with tissue destruction in a number of disease states including rheumatoid arthritis, fibrotic lung disease, dilated cardiomyopathy, and cancer invasion and metastasis (6 -11). However, the role and the mechanisms regulating MMP-2 or MMP-9 expression during atrophying conditions have not been yet investigated.
Proinflammatory cytokines are considered as one of the most important mediators of skeletal muscle wasting in various chronic diseases (12). We have recently reported that tumor necrosis factor (TNF)-related weak inducer of apoptosis (TWEAK), a proinflammatory cytokine belonging to the TNF superfamily, is a potent skeletal muscle-wasting cytokine (13). Although the exact mechanisms of action of TWEAK in skeletal muscle remain unknown, TWEAK has been found to inhibit the differentiation of myoblasts into myotubes (14). TWEAK also augments the degradation of specific muscle proteins such as MyoD and myosin heavy chain fast type in cultured muscle cells (13,15). Furthermore TWEAK-induced muscle atrophy in vivo is accompanied with significant extracellular matrix abnormalities (13). In addition, we have also reported that the TWEAK-induced degradation of muscle proteins (e.g. myosin heavy chain fast type and MyoD) involves the activation of nuclear factor-B (NF-B) transcription factor (13,15).
NF-B is a major proinflammatory transcription factor that regulates the expression of a plethora of genes especially those involved in inflammatory and immune responses (16 -18). Recent genetic studies have provided strong evidence that constitutive activation of NF-B leads to skeletal muscle wasting (16,19). Conversely specific inhibition of NF-B prevents loss of skeletal muscle mass in response to catabolic stimuli such as tumor growth, denervation, and unloading (19,20). Depending on the type of stimuli, the activation of NF-B can occur via either a canonical or alternative pathway (18). The canonical NF-B signaling pathway involves the upstream activation of inhibitors of B (IB) kinase-␤ (IKK␤) and subsequent phosphorylation and degradation of IB proteins. On the other hand, activation of the alternative NF-B pathway requires the upstream activation of NF-B-inducing kinase (NIK) and IKK␣ and the proteolytic processing of p100 subunit into p52 (17,18,21). Although the activation of either the classical or alternative NF-B signaling pathway can lead to skeletal muscle wasting in response to specific stimuli (22,23), the upstream signaling mechanisms regulating NF-B activation and the genes that activated NF-B induces in skeletal muscle tissues remain largely unknown.
We hypothesize that one of the mechanisms by which TWEAK induces myopathy is through the increased production of specific extracellular proteases via NF-B-dependent mechanisms. To test our hypothesis, we performed a genome wide microarray analysis using RNA samples from control and TWEAK-treated C2C12 myotubes. Our results show that the expression and production of MMP-9 are drastically increased in TWEAK-treated cultured myotubes. Higher expression of MMP-9 was also observed in skeletal muscle tissues of TWEAK transgenic mice. Our study also shows that NIK-, IKK␤-, and protein kinase 38 (p38) MAPK-dependent activation of NF-B is responsible for the increased expression of MMP-9 in cultured muscle cells. Finally our experiments demonstrate that compared with control mice TWEAK-induced myopathy is significantly attenuated in Mmp9 knock-out mice.
Mice-Control (strain FVB/NV) and Mmp9 knock-out (strain FVB.Cg-Mmp9 tm1Tvu/J) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Transgenic mice expressing TWEAK under the control of muscle creatine kinase promoter have been described previously (13). All the experiments with animals were approved by the Institutional Animal Care and Use Committee of the University of Louisville. All the procedures were conducted in strict accordance with the Public Health Service Animal Welfare Policy.
Cell Culture-C2C12 and L6 myoblastic cell lines were obtained from the American Type Culture Collection (Manassas, VA). These cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum. C2C12 myoblasts were differentiated into myotubes by incubation in differentiation medium (DM; 2% horse serum in Dulbecco's modified Eagle's medium) for 96 h. Myotubes were maintained in DM, and medium was changed every 48 h. L6 myoblasts were differentiated into myotubes by incubation in Dulbecco's modified Eagle's medium containing 0.5% fetal bovine serum for 48 -72 h.
Short Interfering RNA (siRNA)-ON-TARGETplus siRNA duplexes for mouse TRAF2, IKK␣, IKK␤, NIK, and p38 and non-targeting control siRNA were obtained from Dharmacon RNA Technologies (Lafayette, CO). cIAP1 and cIAP2 siRNA were purchased from Santa Cruz Biotechnology. C2C12 myoblasts plated in either a 6-well (100,000 cells/well) or 24-well (20,000 cells/well) plate were transfected with siRNA duplexes using DharmaFect-3 transfection reagent following the manufacturer's instructions (Dharmacon RNA Technologies). The knockdown of specific proteins was confirmed by performing Western blotting.
Gene Expression Analysis Using Microarray Techniques-Total RNA was isolated from C2C12 myotubes after 18 h of TWEAK treatment using the Agilent total RNA isolation kit (Agilent Technologies, Palo Alto, CA). Each experiment was performed with a minimum of five replicates. The total RNA concentration was determined by a NanoDrop spectrophotometer, and RNA quality was determined by 18/28 S ribosomal peak intensity on an Agilent Bioanalyzer. For microarray expression profiling and real time PCR, RNA samples were used only if they showed little to no degradation. Custom cDNA slides were spotted with Oligator "Mouse Exonic Evidence-Based Oligonucleotide" mouse genome set with 38,467 cDNA probes (Illumina, Inc., San Diego), which allows simultaneous interrogation of 25,000 genes. A Q-Array2 robot (Genetix) was used for spotting. The array includes positive controls, doped sequences, and random sequence to ensure correct gene expression values were obtained from each array. A total of 250 ng of RNA was used to synthesize double-stranded cDNA using the low RNA input fluorescent linear application kit (Agilent). The microarray slides were scanned using a GSI Lumonics Sca-nArray 4200A Genepix scanner (Axon). The image intensities were analyzed using the ImaGene 5.6 software (Biodiscovery, Inc., El Segundo, CA). Expression analysis of microarray experiments was performed with GeneSpring 7.1 (Silicon Genetics, Palo Alto, CA) using the raw intensity data generated by the ImaGene software. Local background-subtracted total signal intensities were used as intensity measures, and the data were normalized using per spot and per chip locally weighted scatterplot smoothing normalization. The changes in transcript levels were analyzed utilizing a t test with Benjamini and Hochberg multiple testing correction.
Gelatin Zymography-The MMP-9 activity in C2C12 conditioned medium was determined by gelatin zymography as previously described (24). To determine MMP-9 activity in skeletal muscle tissues, muscle extracts were prepared in non-reducing lysis buffer (50 mM Tris-Cl (pH 8.0), 200 mM NaCl, 50 mM NaF, 0.3% Ipegal, and protease inhibitors). Equal amount of proteins (100 g/sample) were separated on 10% SDS-polyacrylamide gels containing 1 mg/ml gelatin B (Fisher Scientific) under nonreducing conditions. Gels were washed in 2.5% Triton X-100 for 1 h at room temperature followed by incubation in reaction buffer (50 mM Tris-HCl (pH 8.0), 50 mM NaCl, 5 mM CaCl 2 , and 0.02% sodium azide) for 36 -48 h at 37°C. To visualize gelatinolytic bands, gels were stained with Coomassie Blue at room temperature followed by extensive washing in destaining buffer (10% methanol and 10% acetic acid in distilled water). The gels were photographed for determination of gelatinolytic activity.
In Vitro Kinase Assay-The activation of JNK1 activity was determined by immunoprecipitation followed by in vitro kinase assays using GST-c-Jun-(1-79) proteins as a substrate as described previously (25).
Electrophoretic Mobility Shift Assay (EMSA)-The activation of NF-B and AP-1 transcription factors was measured by EMSA. A detailed procedure for the preparation of nuclear and cytoplasmic extracts and EMSA has been described previously (15).
Western Blotting-Western blotting was performed to measure the levels of different proteins using a standard protocol as described previously (15,26).
Transient Transfection and Reporter Gene Activity-C2C12 myoblasts plated in 24-well tissue culture plates were transfected with different plasmids using Effectene transfection reagent according to the manufacturer's protocol (Qiagen). Transfection efficiency was controlled by cotransfection of myoblasts with Renilla luciferase-encoding plasmid pRL-TK (Promega). When Ͼ90% confluent, the cells were differentiated by changing medium to DM and incubation for an additional 96 h. After appropriate treatments, specimens were processed for luciferase expression using Dual-Luciferase assay systems with reporter lysis buffer according to the manufacturer's instructions (Promega). Luciferase measurements were made using a luminometer (Berthold Detection Systems). Alkaline phosphatase activity in culture supernatants was measured by a standard assay using para-nitrophenyl phosphate as a substrate (24).
Immunohistochemistry-Serial cross-sections (10 mm thick) of frozen tibial anterior (TA) muscle were cut at midbelly on a cryostat microtome, mounted on glass slides, and fixed in 4% paraformaldehyde. The sections were then blocked in 10% bovine serum albumin in phosphate-buffered saline for 1 h and incubated with rabbit polyclonal laminin antibody overnight at 4°C under humidified conditions. Alexa Fluor 488 fluorescent dye conjugated to an anti-rabbit secondary antibody (Invitrogen) was used for detection. Slides were visualized with a fluorescence microscope (Nikon), and images were captured using a Nikon DS Fi1 camera. Fiber cross-sectional area in lamininimmunostained muscle sections was calculated using Nikon NIS Elements BR 3.00 software.
Analysis of Apoptosis and Necrosis-Myonuclear apoptosis in muscle cryosection was determined using a MEBSTAIN Apoptosis Kit Direct following the manufacturer's instructions (MBL International, Woburn, MA). Necrotic fibers in soleus muscle sections were identified by immunostaining muscle FEBRUARY 13, 2009 • VOLUME 284 • NUMBER 7 cryosections with Cy3-labeled goat anti-mouse IgG (Invitrogen), a commonly used method to quantify fiber necrosis in muscle sections (27). The number of IgG-filled fibers in two different sections of each soleus muscle (two muscles from each mouse) was counted. Data are presented as average number of necrotic fibers in muscle sections.

TWEAK Stimulates MMP-9 Expression in Skeletal Muscles
Statistical Analysis-Results are expressed as mean Ϯ S.D. The Student's t test or analysis of variance was used to compare quantitative data populations with normal distributions and equal variance. A value of p Ͻ0.05 was considered statistically significant unless otherwise specified.

RESULTS
Using cultured C2C12 myotubes, we investigated the effects and mechanisms by which TWEAK induces the expression and production of MMP-9 in skeletal muscles. Furthermore by chronic administration of TWEAK in control and Mmp9 knock-out mice, we investigated the role of MMP-9 in TWEAK-induced body and skeletal muscle loss.
Microarray Analysis Revealed Increased Expression of MMP-9 in TWEAK-treated C2C12 Myotubes-To determine how TWEAK affects the expression of various MMP family genes in skeletal muscle, we performed a genome wide microarray analysis using RNA samples from control and TWEAKtreated C2C12 myotubes. C2C12 myoblasts were first differentiated into myotubes by incubation in differentiation medium for 96 h. The myotubes were then treated with recombinant mouse TWEAK protein (10 ng/ml) for 18 h. The total RNA from control and TWEAK-treated myotubes was isolated and subjected to microarray analysis for the expression of MMP family genes. Among 40 genes of the MMP family that were present on our microarray slides, the expression of 14 MMPrelated genes was significantly (p Ͻ 0.05) affected on treatment of myotubes with TWEAK (Table 1). MMP-9 and TIMP-1 were among the prominent MMP-related genes whose expression was significantly increased, whereas MMP-2 and TIMP-2 topped the list of MMP genes whose expression was significantly down-regulated in response to TWEAK-treatment (Table 1).
To validate the microarray data, we performed quantitative real time PCR for a few select MMPs using their specific primers and using the same RNA samples that were used in the microarray experiment. As shown in Fig. 1A, the level of MMP-9 mRNA was significantly increased in TWEAK-treated myotubes compared with untreated control myotubes. Consistent with the microarray results, we observed significant changes in the levels of MMP-2, TIMP-1, and TIMP-2 in TWEAK-treated myotubes (Fig. 1A).
TWEAK Increases MMP-9 Production in C2C12 Myotubes and in Skeletal Muscle Tissues in Vivo-Because MMP-9 is a major extracellular protease that can degrade several components of basement membrane and intramuscular connective tissues (4, 28), we studied how TWEAK regulates the expression and production of MMP-9 in cultured myotubes. We first evaluated whether the increased expression of MMP-9 in C2C12 myotubes was accompanied with the increased production of MMP-9 protein in cell culture supernatants. The production of MMP-9 in culture supernatants was measured by gelatin zymography and Western blotting techniques. As shown in Fig. 1, B and C, treatment of C2C12 myotubes with B, C2C12 myotubes were treated with increasing concentrations of TWEAK for 18 h, and the production of MMP-9 in culture supernatants was measured by gelatin zymography. Data presented here show that TWEAK augments the secretion of MMP-9 in culture supernatants. C, culture supernatants of TWEAK-treated C2C12 myotubes were subjected to Western blot analysis for MMP-9 proteins. A representative immunoblot presented here shows that TWEAK increases the MMP-9 protein levels in culture supernatants in a dosedependent manner. D, L6 myotubes were treated with different concentrations of TWEAK protein, and the production of MMP-9 in culture supernatants was measured by Western blotting. Data presented here show that TWEAK enhances the production of MMP-9 in L6 myotubes. E, diaphragm, gastrocnemius, soleus, and tibial anterior muscles isolated from 6-month-old control and TWEAK transgenic (Tg) mice were analyzed for the expression of MMP-9 by gelatin zymography (upper panel) or Western blotting (lower panel). Data presented here demonstrate that the protein level of MMP-9 is increased in skeletal muscle of TWEAK transgenic mice compared with littermate control mice. WB, Western blotting. TWEAK increased the production of MMP-9 in culture supernatants in a dose-dependent manner. Consistent with the microarray and quantitative real time PCR data, TWEAK treatment slightly reduced the MMP-2 production in myotubes (Fig.  1B). TWEAK was also found to increase the production of MMP-9 in culture supernatants of L6 myotubes suggesting that the effect of TWEAK on MMP-9 production was not specific to C2C12 myotubes and that other muscle cells respond similarly to TWEAK with respect to MMP-9 production (Fig. 1D).
We also studied the effect of TWEAK on MMP-9 production in skeletal muscle of TWEAK transgenic mice (13) by zymography and Western blotting methods. As shown in Fig. 1E, the level of MMP-9 was significantly higher in diaphragm, soleus, and tibial anterior muscle of TWEAK transgenic mice compared with littermate control mice. Collectively these data provide strong evidence that TWEAK augments the production of MMP-9 in skeletal muscle cells and tissues.
TWEAK Activates Both Classical and Alternative NF-B Signaling Pathways in C2C12 Myotubes-Several published reports suggest that NF-B is one of the most important transcription factors; its activation leads to increased expression of MMP-9 in response to divergent stimuli (14). We studied the effect of TWEAK on the activation of NF-B in myotubes by EMSA. As shown in Fig. 2A, treatment of C2C12 myotubes with TWEAK increased the activation of NF-B in a time-depend-ent fashion. We also investigated the effect of TWEAK on the phosphorylation and degradation of IB␣ protein, an important step in the activation of the classical NF-B signaling pathway (18). As shown in Fig. 2B, treatment of myotubes with TWEAK significantly increased the phosphorylation (upper panel) and degradation (middle panel) of IB␣ protein. Furthermore the phosphorylation of the p65 (RelA) subunit of NF-B also significantly increased in response to TWEAK treatment (Fig. 2B) indicating that TWEAK activates the classical NF-B signaling pathway in myotubes.
We next investigated the effect of TWEAK on the levels of p100 and p52 proteins in C2C12 myotubes. As shown in Fig. 2C, treatment of myotubes with TWEAK reduced the level of p100 protein with a concomitant increase in the level of p52 protein suggesting that TWEAK also activates the alternative NF-B signaling pathway in myotubes. We also studied whether TWEAK increases the transcriptional activity of NF-B in myotubes. C2C12 myoblasts were transiently transfected with pNF-B-SEAP (NF-B reporter plasmid) and differentiated into myotubes by incubation in DM for 96 h. Myotubes were then treated with TWEAK for 18 h, and the levels of secreted alkaline phosphatase (SEAP) were measured by a standard assay using para-nitrophenyl phosphate as substrate. Treatment of myotubes with TWEAK significantly increased the transcriptional activity of NF-B in myotubes (Fig. 2D). Similar to NF-B, we also measured the transcriptional activation of AP-1, a transcription factor that cooperates with NF-B to induce gene expression (17). As shown in Fig. 2D, TWEAK also increased the transcriptional activity of AP-1 in myotubes.
NF-B and AP-1 Are Involved in the TWEAK-induced Production of MMP-9 in Myotubes-We next investigated the role of NF-B and AP-1 transcription factors in TWEAK-induced production of MMP-9 from myotubes. The activation of NF-B was blocked by overexpression of IB␣⌬N protein (a dominant negative inhibitor of NF-B) using adenoviral vector. The role of AP-1 in TWEAK-induced MMP-9 production was studied using NDGA, a pharmacological inhibitor of AP-1 (29). As shown in Fig. 3A, the TWEAK-induced production of MMP-9 in culture supernatants was completely blocked by overexpression of IB␣⌬N protein. Similarly pretreatment of myotubes with NDGA also significantly inhibited the production of MMP-9 in culture supernatants on treatment with TWEAK (Fig. 3B).  FEBRUARY 13, 2009 • VOLUME 284 • NUMBER 7

JOURNAL OF BIOLOGICAL CHEMISTRY 4443
We also measured the effect of specific inhibition of NF-B or AP-1 on the activation of MMP-9 promoter. C2C12 myoblasts were transfected with either pcDNA3-FLAG-IB␣⌬N or pcDNA-TAM67 (a dominant negative inhibitor of AP-1) plasmid (24) along with a reporter plasmid expressing luciferase under the control of mouse MMP-9 promoter (in a 1:10 ratio). The myoblasts were differentiated into myotubes, and following TWEAK treatment for 18 h, the luciferase activity in cell extracts was measured. As shown in Fig. 3C, overexpression of either IB␣⌬N or TAM67 significantly blocked the TWEAKinduced activation of MMP-9 promoter in myotubes. We also studied the role of NF-B and AP-1 in TWEAK-induced expression of MMP-9 in myotubes using a reporter gene assay in which the conserved AP-1, NF-B, or SP-1 binding sites in the human MMP-9 gene promoter were mutated (30). TWEAK-induced activation of human MMP-9 promoter was significantly blocked by point mutations in either AP-1 or NF-B but not SP-1 binding site (Fig. 3D).
NIK and IKK␤ but Not IKK␣ Are Involved in TWEAK-induced Activation of MMP-9 Promoter-The activation of NF-B in response to different stimuli involves upstream activation of a number of signaling proteins; the most important are NIK, IKK␣, and IKK␤. These kinases are central to determining which NF-B pathway (classical or alternative) is activated in response to specific stimuli, the composition of activated NF-B complex, and the set of genes NF-B targets upon its activation (18,21). Using their dominant negative inhibitors, we investigated the role of NIK, IKK␣, and IKK␤ in TWEAK-induced activation of MMP-9 promoter. As shown in Fig. 4A, overexpression of a dominant negative mutant of NIK (DN-NIK) or IKK␤ (DN-IKK␤) significantly blocked the TWEAK-induced activation of MMP-9 promoter in C2C12 myotubes. On the other hand, overexpression of a dominant negative mutant of IKK␣ only marginally (not significantly) affected the activation of MMP-9 promoter in response to TWEAK (Fig. 4A). We also investigated the contribution of IKK␣, IKK␤, and NIK in the activation of NF-B on treatment with TWEAK. Consistent with the MMP-9 promoter activation, overexpression of a dominant negative mutant of either NIK or IKK␤ but not a dominant negative mutant of IKK␣ significantly blocked the TWEAK-induced activation of NF-B in myotubes (Fig. 4B). A few studies have indicated that Rho GTPases, especially Rac1 and Cdc42, interact with TWEAK receptor (31,32). However, we found that overexpression of a Rho GDP dissociation inhibitor protein that blocks the activation of RhoA, Rac1, and Cdc42 GTPases (33) did not affect either the activation of MMP-9 promoter or NF-B in myotubes (Fig. 4, A and B).
To further confirm the role of IKK␣, IKK␤, and NIK in TWEAK-induced MMP-9 production in myotubes, we studied the effect of siRNA-mediated knockdown of NIK, IKK␣, and IKK␤ on MMP-9 production in response to TWEAK. C2C12 myoblasts were transfected with either non-targeting control siRNA or mouse NIK, IKK␣, or IKK␤ siRNA duplexes; differentiated into myotubes; and treated with soluble TWEAK protein. Consistent with the data using dominant negative mutants, siRNA-mediated knockdown of NIK (Fig. 4C) and

. Role of NF-B and AP-1 in TWEAK-induced MMP-9 production.
A, C2C12 myoblasts were transduced with Ad.Control or Ad.FLAG-IB␣⌬N adenovirus (multiplicity of infection, 200) and differentiated into myotubes by incubation in differentiation medium. The myotubes were then treated with TWEAK (100 ng/ml) for 18 h, and the production of MMP-9 in culture supernatants was measured by gelatin zymography. Data presented here show that overexpression of IB␣⌬N protein blocks the TWEAK-induced production of MMP-9 in myotubes. B, C2C12 myotubes were pretreated with NDGA (20 M) for 2 h followed by treatment with TWEAK (100 ng/ml) for 18 h. The production of MMP-9 in culture supernatants was studied by gelatin zymography. Data presented here show that NDGA inhibits the production of MMP-9 in culture supernatants. C, C2C12 myoblasts were transfected with vector alone, pcDNA-FLAGIB␣⌬N, or pcDNA3-TAM67 plasmid along with mouse MMP-9 reporter plasmid in a 1:10 ratio. The myoblasts were differentiated into myotubes and treated with TWEAK (100 ng/ml) for 18 h, and the luciferase activity in cell lysates was measured using a commercially available kit. Data presented here show that the activation of MMP-9 promoter in response to TWEAK was significantly blocked by overexpression of either IB␣⌬N or TAM67 proteins. *, p Ͻ 0.05, values significantly different from TWEAK-treated myotubes transfected with vector alone. D, C2C12 myoblasts were transfected with either wild-type human MMP-9 promoter-luciferase reporter construct (Ϫ670/ϩ34 wt) or with the promoter in which either AP-1 (Ϫ670/ϩ34 AP-1m), NF-B (Ϫ670/ϩ34 NF-B1m), or SP-1 (Ϫ670/ϩ34 SP-1m) binding site was mutated. The transfected myoblasts were differentiated into myotubes by incubation in DM for 96 h. The myotubes were then treated with TWEAK for 18 h, and the luciferase activity in cell extracts was measured. Data are presented as means Ϯ S.D. Data presented here show that mutation in either AP-1 or NF-B site inhibits the MMP-9 promoter-dependent expression of luciferase gene upon treatment with TWEAK. #, p Ͻ 0.05, values significantly different from TWEAK-treated myotubes transfected with wild-type (Ϫ670/ϩ34) MMP-9 promoter construct. ZG, zymography. IKK␤ (Fig. 4E) but not IKK␣ (Fig. 4D) reduced the TWEAKinduced production of MMP-9 in myotubes. Collectively these data suggest that TWEAK induces MMP-9 production in C2C12 myotubes through the activation of NIK and IKK␤.
Involvement of TRAF2 and cIAP2 in TWEAK-induced MMP-9 Production in Myotubes-TRAFs constitute a growing family of cytoplasmic adapter proteins that interact with a large number of signaling receptors/proteins (34). Published reports suggest that the cytoplasmic domain of Fn14, a receptor for TWEAK, contains binding sites for TRAFs (35).
Accumulating evidence also suggests that cIAP1 and cIAP2 proteins are involved in various signal transduction pathways, including NF-B activation in response to TNF␣ (36). To further understand the signaling mechanisms by which TWEAK induces MMP-9 production in myotubes, we investigated the role of TRAF2, cIAP1, and cIAP2 proteins by using an siRNA technique. C2C12 myoblasts were transfected with control, TRAF2, cIAP1, or cIAP2 siRNA duplexes and differentiated into myotubes by incubation in differentiation medium. The cells were treated with TWEAK, and the production of MMP-9 in culture supernatants was measured by Western blotting. Knockdown of TRAF2 (Fig. 5A) and cIAP2 (Fig. 5C) significantly inhibited the TWEAK-induced production of MMP-9 in myotubes. On the other hand, siRNA-mediated knockdown of cIAP1 did not affect the production of MMP-9 in response to TWEAK.
TWEAK Activates ERK1/2, JNK1, and p38 MAPK in Myotubes-MAPK pathways are among the best studied signaling pathways that are activated in response to various stimuli and modulate the activity of many transcription factors and the expression of a wide spectrum of genes (37). In mammalian cells, three parallel mitogen-activated protein kinase pathways have been described, i.e. ERK1/2, p38, and JNKs (37). Available literature suggests that depending upon the stimuli the activation of ERK1/2, JNK1, and/or p38 kinase can lead to the production of MMP-9 (14). We investigated whether TWEAK can activate MAPK in myotubes. C2C12 myotubes were treated with soluble TWEAK protein for different time intervals ranging from 0 to 120 min. The activation of ERK1/2 and p38 MAPK was determined by immunoblotting with antibodies that recognize their respective phosphorylated proteins. The activation C-E, C2C12 myoblasts were transfected with control, NIK, IKK␣, or IKK␤ siRNA duplexes and differentiated into myotubes by incubation in differentiation medium for 72 h. The cells were treated with TWEAK (100 ng/ml), and production of MMP-9 protein in culture supernatants and levels of NIK, IKK␣, or IKK␤ proteins in cell extracts were measured by Western blotting. Data presented here show that the siRNA-mediated knockdown of NIK and IKK␤ but not IKK␣ inhibits the TWEAK-induced production of MMP-9 in myotubes. FIGURE 5. Role of TRAF2, cIAP1, and cIAP2 proteins in TWEAK-induced MMP-9 production in myotubes. C2C12 myoblasts were transfected with control, TRAF2, cIAP1, or cIAP2 siRNA and differentiated into myotubes. The cells were then treated with TWEAK protein (100 ng/ml) for 18 h, and the production of MMP-9 in culture supernatants and the levels of TRAF2, cIAP1, or cIAP2 proteins in myotubes were measured by Western blotting using their specific antibodies. Data presented here show that the siRNA-mediated knockdown of TRAF2 (A) or cIAP2 (C) but not cIAP1 (B) inhibits the TWEAKinduced MMP-9 production in myotubes. FEBRUARY 13, 2009 • VOLUME 284 • NUMBER 7 of JNK1 was measured by immunoprecipitation and in vitro kinase assay. As shown in Fig. 6, TWEAK increased the activation of all three MAPKs (i.e. ERK1/2, JNK1, and p38) in myotubes.

Activation of p38 MAPK Contributes to the TWEAK-induced MMP-9 Production and the Activation of NF-B in Myotubes-
We next investigated whether the inhibition of any of the three MAPKs affects the TWEAK-induced expression and production of MMP-9 in C2C12 myotubes. Pretreatment of myotubes with SB203580 (a pharmacological inhibitor of p38 MAPK) but not PD98059 (a pharmacological inhibitor of ERK1/2) or SP600125 (a pharmacological inhibitor of JNK) significantly blocked the TWEAK-induced increase in MMP-9 mRNA levels (Fig. 7A). Similarly analysis of culture supernatants by gelatin zymography revealed that the inhibition of p38 MAPK but not ERK1/2 or JNK1 was effective in blocking the TWEAK-induced production of MMP-9 in myotubes (Fig. 7B). The role of p38 MAPK in TWEAK-induced MMP-9 production was also studied using an siRNA approach. As shown in Fig. 7C, siRNA-mediated knockdown of p38 protein significantly inhibited the TWEAK-induced production of MMP-9 in culture supernatants.
Published reports also suggest that in response to specific stimuli and cell type the activation of p38 MAPK can lead to the activation of NF-B (24,38). However, the role of p38 MAPK in TWEAK-induced NF-B activation remains unknown. We tested the possibility that p38 MAPK is involved in TWEAK-induced activation of NF-B and hence the expression of MMP-9 in myotubes. C2C12 myoblasts were transiently transfected with a dominant negative mutant of MKK3 and MKK6 (DN-MKK3/MKK6), the upstream activator of p38 MAPK (37), along with NF-B reporter plasmid (1:10 ratio) followed by their differentiation into myotubes and treatment with TWEAK. Interestingly overexpression of DN-MKK3/6 proteins significantly blocked the TWEAK-induced transcriptional activation of NF-B (Fig. 7D).

TWEAK-induced Myopathy Is Attenuated in Mmp9 Knock-out
Mice-Because TWEAK induces the expression of MMP-9 in myotubes and in skeletal muscle of mice in vivo, we investigated whether TWEAK-induced muscle loss can be prevented in Mmp9 knock-out mice. We used a protocol of TWEAK treatment of mice similar to that described earlier with some modification (13). Three-week-old control and Mmp9 knock-out mice FIGURE 6. TWEAK activates ERK1/2, JNK1, and p38 MAPK in myotubes. C2C12 myotubes were treated with 100 ng/ml TWEAK protein for the indicated time intervals, and the activation of ERK1/2, JNK1, and p38 MAPK was studied using the protocols described under "Experimental Procedure." Data presented here show that TWEAK activates ERK1/2 (A), JNK1 (B), and p38 MAPK (C) in myotubes. Treatment with TWEAK did not alter the total level of ERK1/2, JNK1, or p38 MAPK in myotubes. IP, immunoprecipitation. FIGURE 7. Role of MAPK in TWEAK-induced MMP-9 production in myotubes. C2C12 myotubes were preincubated for 2 h with the indicated concentrations of PD98059 (ERK1/2 inhibitor), SP600125 (JNK1 inhibitor), or SB203580 (p38 kinase inhibitor) followed by treatment with TWEAK (100 ng/ml) for 18 h. A, the transcript level of MMP-9 measured by quantitative real time PCR showed that pretreatment with SB203580 but not PD98059 or SP600125 inhibits the TWEAK-induced expression of MMP-9 in myotubes. B, analysis of culture supernatants by zymography reveled that SB203580 inhibits the TWEAK-induced production of MMP-9 in myotubes. C, C2C12 myoblasts were transfected with control or p38 MAPK siRNA and incubated in differentiation medium for 72 h followed by treatment with TWEAK (100 ng/ml) for 18 h. The production of MMP-9 protein in culture supernatants and the levels of p38 protein in cell extracts were measured by Western blotting. Data presented here demonstrate that siRNA-mediated knockdown of p38 protein inhibits the TWEAK-induced MMP-9 production in myotubes. D, C2C12 myoblasts were transfected with vector alone or dominant negative mutants of MKK3 and MKK6 (equal amounts) along with pNF-B-SEAP plasmid (1:10 ratio). The myoblasts were differentiated into myotubes and treated with TWEAK (100 ng/ml) for 18 h, and SEAP activity in culture supernatants was measured. Data (mean Ϯ S.D.) presented here show that DN-MKK3/6 significantly inhibits the TWEAK-induced MMP-9 production in myotubes. *, p Ͻ 0.05, values significantly different from TWEAK-treated myotubes transfected with vector alone.
were given chronic intraperitoneal injection of either phosphate-buffered saline alone or soluble TWEAK protein (100 ng/g of body weight) biweekly for 2 weeks (i.e. a total of four injections). Twenty-four hours after the final TWEAK injection, the mice were sacrificed, and body and skeletal muscle weight of each mouse was measured. As shown in Fig. 8A, the gain in body weight of TWEAK-treated Mmp9 knock-out mice was significantly higher than that of corresponding TWEAKtreated wild-type mice (control 1.45 Ϯ 0.2 g versus Mmp9 knock-out 2.4 Ϯ 0.3 gm). In addition, TA muscle weight was also significantly higher in TWEAK-treated Mmp9 knock-out mice compared to wild-type mice treated with TWEAK (Fig.  8B). Chronic administration of TWEAK in wild-type mice led to the degradation of basement membrane as evident by diffused immunostaining with laminin antibody (Fig. 8C). However, the degradation of basement membrane was not evident in Mmp9 knock-out mice treated with TWEAK (Fig. 8C). We also performed hematoxylin and eosin staining of a TA muscle cryosection of control and TWEAK-treated control and Mmp9 knock-out mice. As shown in Fig. 8D, chronic administration of TWEAK led to fiber population of variable diameters and appearance of dark stained nuclei of inflammatory cells between adjacent fibers, a typical feature during myopathy. However, TA muscle from TWEAK-treated Mmp9 knock-out mice showed fibers of uniform diameter and no clustering of inflammatory cells between adjacent fibers (Fig. 8D). Furthermore the average fiber cross-sectional area in laminin-stained cryosections of TA muscle of TWEAK-treated Mmp9 knockout mice was significantly higher than that of TWEAK-treated control mice (Fig. 8E). Treatment with TWEAK did not induce myonuclear apoptosis in either control or Mmp9 knock-out mice (negative data not shown). Furthermore there was also no evidence of fiber necrosis in TA, diaphragm, gastrocnemius, or quadriceps muscle of either TWEAK-treated control or Mmp9 knock-out mice. However, analysis of soleus muscle revealed the presence of necrotic fiber after treatment of mice with TWEAK (Fig. 8F). Interestingly the number of necrotic fibers was significantly reduced in soleus muscle of TWEAK-treated Mmp9 knock-out mice compared with control mice (Fig. 8G). Collectively these data suggest that TWEAK-induced production of MMP-9 contributes to myopathy in vivo.

DISCUSSION
MMPs represent one of the most important extracellular proteases that can degrade several extracellular matrix proteins in skeletal muscle tissues (4,28). However, the expression of different MMPs and their potential role in structural and functional deterioration of skeletal muscle during pathological conditions remain largely unknown. Our experiments in this study demonstrate that TWEAK drastically augments the expression of MMP-9 in cultured myotubes (Fig. 1). Interestingly although the expression of MMP-9 was increased by severalfold, the level of TIMP-1, a physiological inhibitor that binds to MMP-9 in a stoichiometric ratio (3)(4)(5), was only marginally increased (Fig.  1A). Furthermore the increased levels of MMP-9 in skeletal muscle tissue of transgenic mice expressing TWEAK in skeletal muscle (Fig. 1E) and upon in vivo administration of TWEAK in normal mice (data not shown) suggest that one of the potential mechanisms by which TWEAK induces myopathy is through the increased production of MMP-9 (and possibly some other MMPs) in the skeletal muscle microenvironment. Indeed our results in this study demonstrate that TWEAK-induced myopathy is significantly attenuated in Mmp9 knock-out mice (Fig. 8).
The transcription of MMP-9 gene is under the control of a 2.2-kb upstream regulatory sequence harboring binding sites for AP-1, NF-B, SP-1, and PEA3/Ets, which are conserved in both human and mouse (9,39,40). Accumulating evidence suggests that the activation of NF-B is one of the most important signaling events required for the inducible expression of MMP-9 in response to different stimuli (14). Our results in this study demonstrate that TWEAK augments the DNA binding and transcriptional activities of NF-B in cultured myotubes (Fig. 2, A and D). Furthermore our results also suggest that the activation of NF-B and to some extent AP-1 is the seminal event in TWEAK-induced production of MMP-9 in skeletal muscle. Specific inhibition of NF-B activity or point mutations in either NF-B or AP-1 binding sites effectively blocked the TWEAK-induced transcriptional activation of MMP-9 promoter and the production of MMP-9 in cultured myotubes (Fig. 3).
TWEAK is a relatively newly identified cytokine that functions through binding to Fn14 receptor (41). Although TWEAK has been shown to induce various biological responses including inflammation, angiogenesis, osteoclastogenesis, skeletal muscle wasting, and apoptosis (41), the TWEAK-induced intracellular signaling pathways remain poorly understood. Our results in this study suggest that TWEAK can cause the activation of both classical and alterative NF-B signaling pathways in myotubes as evidenced by the increased phosphorylation of IB␣ and p65 proteins (Fig. 2B) and proteolytic processing of p100 protein into p52 (Fig. 2C), respectively. Our experiments also indicate that TWEAK induces the production of MMP-9 through the activation of IKK␤, a key kinase of the classical NF-B signaling pathway (Fig. 4, A and E). Interestingly we observed that the overexpression of a dominant negative inhibitor (Fig. 4A) of NF-B-inducing kinase (DN-NIK) or siRNA-mediated knockdown of NIK (Fig. 4C) blocked the TWEAK-induced activation of MMP-9 in myotubes. NIK is an important kinase of the NF-B signaling pathway that is activated in response to a wide variety of stimuli and generally leads to the activation of IKK␣ (18,21). A recent report also suggests that functional NIK is required for TWEAK-induced processing of p100 into p52 protein (42). However, our data in this study suggest that NIK causes MMP-9 expression in myotubes using a pathway that may not involve the participation of IKK␣. This contention is supported by our results demonstrating that the overexpression of either DN-NIK or DN-IKK␤ but not DN-IKK␣ significantly blocked the TWEAK-induced activation of MMP-9 promoter (Fig. 4A) and NF-B in myotubes (Fig. 4B). Similarly siRNA-mediated knockdown of NIK (Fig. 4C) and IKK␤ (Fig. 4E) but not IKK␣ (Fig. 4D) blocked the TWEAKinduced production of MMP-9 in myotubes. Although the exact mechanisms by which NIK mediates TWEAK-induced MMP-9 expression in myotubes remain unclear, the IKK␣-independent function of NIK in MMP-9 expression might not be very startling because NIK has been implicated in regulation of multiple signaling events. Besides IKK␣, NIK has also been shown to phosphorylate and activate IKK␤ under specific conditions (43,44). NIK also phosphorylates the p65 transactivation domain (45)(46)(47) and activates MAPK signaling pathways (48,49). Indeed our results demonstrate that TWEAK augments the phosphorylation of p65 protein at Ser-536 residue (Fig. 2B) and activates MAPK (Fig. 5) in myotubes. There is also a report indicating that, like TWEAK, the expression of MMP-9 in response to osteopontin occurs through the activation of NIK (50).
TRAFs are some of the most important adaptor proteins that are involved in receptor-mediated activation of a number of cell signaling pathways. Published reports suggest that the cytoplasmic domain of Fn14 receptor contains a putative TRAF binding sequence, and genetic ablation of TRAF2 and TRAF5 inhibits the TWEAK-induced activation of NF-B in cultured mouse embryonic fibroblasts (35,42). Recent evidence also suggests that cIAP1 and cIAP2 proteins through their interaction with TRAF2 protein mediate the TNF␣-induced activation of NF-B (36, 51). Vince et al. (52) showed that interaction of Fn14 with TWEAK causes the recruitment of TRAF2-cIAP1 complex, and lysosomal degradation of this complex is required for TWEAK-induced activation of non-canonical NF-B. Consistent with published reports, we found that siRNA-mediated knockdown of TRAF2 inhibits the TWEAK-induced production of MMP-9 in myotubes (Fig. 5A). Interestingly we also found that cIAP2 (Fig. 5C) but not cIAP1 (Fig. 5B) was required for the production of MMP-9 in response to TWEAK in myotubes. Although the exact mechanisms by which TRAF2 and cIAP2 mediate TWEAK-induced expression of MMP-9 remain to be elucidated, these experiments suggest that, like TNF-␣, cIAP2 might also be one of the important components of the TWEAK-induced signaling pathways.
A plethora of literature exists suggesting that MAPKs are involved in the expression and production of MMP-9 in different cell types (14). However, the role of individual MAPKs (i.e. ERK1/2, JNK, and p38 MAPK) in the expression of MMP-9 seems to be dependent on the cell type and the stimulus (Ref. 24 and references therein). Our results show that although TWEAK can activate all three MAPKs in myotubes (Fig. 6) the activation of only p38 MAPK was required for the TWEAKinduced production of MMP-9 in myotubes (Fig. 7, A, B, and C). Furthermore our data suggesting that overexpression of dominant negative mutants of MKK3 and MKK6 (DN-MKK3/6), the upstream activators of p38 MAPK (37), contributes to the TWEAK-induced activation of NF-B provide an important link between the activation of p38 MAPK and NF-B and hence MMP-9 expression in myotubes (Fig. 7D). Although p38 MAPK is known to regulate a diverse set of cellular responses, it is important to recognize that p38 MAPK might be playing a pivotal role in regulation of skeletal muscle metabolism especially in response to proinflammatory molecules. Recent reports suggest that several other muscle-wasting stimuli such as TNF␣ and lipopolysaccharide induce the expression of atrogin-1, a muscle-specific E3 ubiquitin ligase, through the activation of p38 MAPK (53,54). Furthermore similar to TWEAK, TNF␣ induces the expression of MMP-9 in skeletal muscle cells through the activation of p38 MAPK (24). Collectively these evidences indicate that, besides NF-B, p38 MAPK might serve as an important molecular target to blocking the expression of various muscle-wasting molecules in skeletal muscle.
An important observation of our study was that the TWEAK-induced loss of body weight was significantly attenuated in Mmp9 knock-out mice compared to control mice (Fig.  8A). Furthermore TA muscle weight (Fig. 8B) and average muscle fiber cross-sectional area (Fig. 8E) were significantly higher in TWEAK-treated Mmp9 knock-out mice compared to control mice suggesting that the increased production of MMP-9 contributes to the TWEAK-induced muscle atrophy in vivo. Furthermore the reduced numbers of inflammatory cells (Fig.  8D) and necrotic fibers (Fig. 8, F and G) in skeletal muscle tissues of TWEAK-treated Mmp9 knock-out mice indicate that the increased levels of MMP-9 in muscle tissues contribute to inflammation and fiber necrosis in skeletal muscle. This role of MMP-9 in skeletal muscle inflammation/necrosis is not unusual. This is because MMP-9 is a well known mediator of the inflammatory response, apoptotic and necrotic cell death, and tissue degradation in several disease states (6 -11). Although collagen IV present in skeletal muscle basement membrane is the most important degradation target of MMP-9, to some extent MMP-9 can also hydrolyze collagens I, III, and V in skeletal muscle tissues either directly or by enhancing the activation of other MMPs (14). Furthermore MMP-9 can degrade laminin, fibronectin, entactin, and elastin (14,(55)(56)(57), which are all components of skeletal muscle extracellular matrix. Recent studies have also shown that ␤-dystroglycan, an important component of the dystrophin-glycoprotein complex on muscle membrane, is also one of the most important proteolytic targets of MMP-9 (58,59). Our results demonstrating reduced degradation of basement membrane around muscle fibers in TWEAK-treated Mmp9 knock-out mice compared to control mice (Fig. 8C) suggest that MMP-9 is involved in the degradation of the components of extracellular matrix in skeletal muscle that might lead to increased inflammation and fiber necrosis.
In summary, we provide the first evidence that TWEAK can induce MMP-9 production in skeletal muscle and that the increased production of MMP-9 contributes to the loss of skeletal muscle mass in vivo. We have also delineated the signaling mechanisms by which TWEAK might be inducing MMP-9 expression in muscle cells (Fig. 9). Although the cause and effect relationship between the production of MMP-9 and skeletal muscle wasting has not yet been established in various muscular disorders, it is possible that similar mechanisms might be involved in skeletal muscle tissue destruction in other conditions where expression of MMP-9 is increased. Indeed the ongoing studies in our laboratory suggest that constitutive overexpression of MMP-9 in skeletal muscle causes severe myopathy in vivo in mice. 3 In future studies, it will also be important to identify the specific substrates that MMP-9 degrades in the skeletal muscle microenvironment and whether available molecular and pharmacological inhibitors of MMP-9 can prevent myopathy in different disease states.