Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The humoral and cellular changes occurring after strenuous muscular work resemble in some aspects the acute-phase response to trauma and inflammation. They consist of leukocyte mobilization and activation, proinflammatory (interleukin [IL]-1 and tumor necrosis factor [TNF]-), anti-inflammatory (IL-1ra, IL-10) and inflammation-responsive (IL-6) cytokine production, complement activation, increased release of arachidonic acid derived compounds, cellular infiltration, formation of free radicals, and increased blood coagulability (1, 2). IL-6 is a cytokine with pleiomorphic biologic effects expressed by mesenchymal, epithelial, and other cells that is upregulated in response to noxious stimuli, cytokines, and growth factors (3). A number of studies have consistently shown increased plasma levels of IL-6 in response to exercise, whereas results with TNF- and IL-1 have been more variable (4). In spite of the well-documented increase in circulating levels of IL-6 after exercise, the cellular source, as well as the stimulus for its release, remains elusive. Increased steady-state IL-6 mRNA levels in skeletal muscle homogenates after long distance running suggest that skeletal muscle tissue might be a source of circulating IL-6 released during exercise (4).

High-intensity muscular contraction is accompanied by intramuscular production of reactive oxygen species (ROS). The major sources of intracellular ROS generation during exercise are the mitochondrial electron transport chain, the cytosolic NADH-oxidase, and xanthine oxidase (XO) (9). Increasing evidence in the literature identifies ROS as general mediators of cellular responses (10). Exogenously applied H₂O₂ activates tyrosine and serine/threonine kinase cascades and exerts an inhibitory action on protein tyrosine phosphatase 1B (10, 11). Moreover, endogenously produced ROS participate in the signaling of growth factors and cytokines as they are generated following binding of these ligands to their cognate receptors (10). Based on the fact that human skeletal muscle cells have the inherent ability to express a variety of cytokines, including IL-6 (12), we hypothesized that ROS generated during exercise might regulate IL-6 production from skeletal myocytes, the most abundant cell type in skeletal muscle tissue. To test our hypothesis, we utilized an in vitro model of C2C12 murine skeletal myoblasts that are differentiated into myotubes when grown in low serum containing medium. We have shown that exposure of myotubes to ROS-producing agents results in an increase in IL-6 release, through a transcription-dependent mechanism that involves the activation of the redox-sensitive transcription factor nuclear factor B (NF-B).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents and Cell Culture

Tissue culture plates were obtained from Nalgen Nunc International (Rochester, NY). Dulbecco's modified Eagle's medium (DMEM), fetal calf serum (FCS), antibiotics, and LipofectAMINE were obtained from GIBCO BRL (Paisley, UK). The enzyme-linked immunosorbent assay (ELISA) kits for the murine IL-6 and the murine TNF- were purchased from R&D Systems Co. (Mineapolis, MN). The luciferase reporter gene assay was purchased from Boehringer Mannheim (Mannheim, Germany); pNF-B, pAP-1, and pTAL were obtained from Clontech Laboratories, Inc. (Palo Alto, CA); Bradford dye and nitrocellulose membranes were obtained from Biorad (Hercules, CA); enhanced chemiluminescence (ECL) Western blotting analysis system was purchased from Amersham Life Science (Buckinghamshire, UK); Taq polymerase was purchased from Promega (Madison, WI), and the DNA and RNA isolation kits were from Qiagen Inc. (Valencia, CA). The antibodies for IB- and phospho-IB- (p-IB-) were obtained from New England Biolabs (Beverly, MA). SB203580 was obtained from Calbiochem (San Diego, CA). All other reagents including horse serum, xanthine/xanthine-oxidase, pyrogallol, hydrogen peroxide, actinomycin D, cycloheximide, catalase, superoxide dismutase, N-acetyl-cysteine, and diethyldithiocarbamate were obtained from Sigma Chemical Co (St. Louis, MO).

Myoblasts derived from mouse skeletal muscle (C2C12) were cultured in DMEM containing 10% FCS supplemented with 10 U/ml penicillin and 100 µg/ml streptomycin (growth medium) at 37°C, in a humidified incubator with 5% CO₂. C2C12 were differentiated into myotubes as previously described (13). When myoblast cultures reached confluence, they were switched to DMEM containing 2% heat-inactivated horse serum supplemented with antibiotics (differentiation medium) for 48 h._.Murine aortic endothelial cells (MAEC) were cultured in DMEM containing 10% FCS supplemented with 10 U/ml penicillin and 100 µg/ml streptomycin.

IL-6 Measurements

C2C12 cells were cultured for 2 d in 24 multiwell clusters until they reached confluence, and were then allowed to differentiate into myotubes for 48 h. Myotubes were incubated with pyrogallol (250 µM), xanthine/xanthine-oxidase (330 µM/50 mU), hydrogen peroxide (500 µM) or TNF- (10 ng/ml) with or without pretreatment with diethyldithiocarbamate (a pharmacologic inhibitor of NF-B; 100 µM), actinomycin D (an inhibitor of transcription; 0.5 µM), cycloheximide (an inhibitor of protein synthesis; 30 µM), catalase (2,000 U/ml), superoxide dismutase (300 U/ml), SB203580 (2 µM), or N-acetylcysteine (NAC) (10 mM). After 24 h, supernatants were collected and centrifuged for 5 min at 6,000 rpm in a tabletop microcentrifuge (Hettich, Tuttlingen, Germany) to remove floating cells. Following centrifugation, pellets were discarded and supernatants used for ELISA in accordance to the manufacturer's instructions. C2C12 cell monolayers in the multiwell plates were lysed with 1N NaOH. Protein amounts per well were determined by the Bradford method and were used to normalize for the values obtained for cytokine release.

Reverse Transcriptase/Polymerase Chain Reaction

Competitor molecules were constructed as follows: a -actin plasmid with known sequence was used as DNA template. Hybrid primers that recognized -actin sequences and murine IL-6 cDNA sequences were synthesized (sense: TGACAACCACGGCCTTCT GGCCGGGACCTGAC, antisense: TTCTGCAAGTGCATCAT CGCGGCAATGCCAGGGT) and used to amplify a 400-base pair product as visualized when run on a 1.2% agarose gel. The final product (IL-6 competitor) was isolated and diluted to a concentration of 0.5 ng/ml. C2C12 were grown in monolayers and allowed to differentiate after reaching confluence. They were then treated with H₂O₂ 500 µM for 6 h and total RNA was isolated according to the manufacturer's instructions. mRNA was reversed transcribed and the resulting cDNA was amplified by competitive polymerase chain reaction (PCR). Each reaction contained a different amount of the competitor molecule (serial 5x dilutions) and cDNA amplified with a set of IL-6 primers recognizing native IL-6 sequences as well as sequences of the competitor (sense: TGA CAACCACGGCCTTC and antisense: TTCTGCAAGTGCAT CATCG). The conditions used were: 94° for 5 min, 94° for 1 min, 53° for 1 min, 72° for 1.5 min, repeated for a total of 29 cycles, and final extension was performed at 72° for 7 min. The products were run on 1.5% agarose gel and visualized under a UV lamp. The point where the two visualized bands (200 bp from the amplification of the native IL-6 cDNA and 400 bp from the competitor molecule amplification) were of equal intensity was considered as the isopoint of the reaction (where equal starting concentrations of IL-6 cDNA and competitor existed in the PCR mix).

Transfections

C2C12 myoblasts were plated in six-well plates at a density of 2 × 10⁴/cm² and allowed to reach 40-60% confluence. Cells were transfected with vector alone (pTAL) or a plasmid containing the luciferase coding sequence under the control of upstream NF-B sites (pNF-B) or AP-1 sites (pAP-1). Transfections were performed using LipofectAMINE at a DNA/lipid ratio of 2.5 µg plasmid DNA/9 µl lipid. When transfected cells reached confluence, growth medium was replaced by differentiation medium for 48 h. Myotubes were then exposed to ROS and after 24 h lysed and assayed for luciferase activity. To normalize for transfection efficiency, we cotransfected a plasmid coding for lacZ (under the control of an SV40 promoter) with either pTAL, pNF-B, or pAP-1. -Galactosidase activity was measured using standard methods. Transfections using a mutant form of IB- (S32A/ S36A) were performed in a similar fashion using the LipofectAMINE Plus reagent.

Western Blotting

C2C12 cells were cultured in six-well plates, and were pretreated and lysed in lysis buffer (1% NP40, 50 mM NaCl, 0.1% sodium dodecyl sulfate, 50 mM NaF, 1 mM Na₃VO₄, 50 mM Tris-HCl, 0.1 mM EGTA, 0.5% deoxycholic acid, 1 mM EDTA, aprotinin 10 µg/ml, leupeptin 10 µg/ml, and 1 mM PMSF). Cell lysates were rocked for 30 min at 4°C followed by a brief centrifugation at 14,000 rpm. Sample aliquots (50 µg/lane) were electrophoresed on 10% SDS-polyacrylamide gels and transferred to a nitrocellulose membrane at 20 V overnight at 4°C in a buffer containing 25 mM Tris and 700 mM glycine. Membranes were subsequently incubated for 2 h at room temperature with 5% dry milk in buffer containing 0.1% (vol/vol) Tween 20 in Tris-buffered saline (TTBS). The following day, membranes were incubated with the primary antibody in TTBS, containing 1% milk for 2 h at room temperature, then washed three times with TTBS for 20 min each time. Finally, membranes were incubated for 1 h with horseradish peroxidase-conjugated secondary antibody and washed again twice with TTBS and once with TBS. Immunoreactive protein bands were visualized using the ECL system.

Data Analysis and Statistics

Data are presented as means ± standard error of the mean (SEM) of the indicated number of observations. Cytokine values are expressed as pmol/mg protein or as percent of the control value. Statistical comparisons between groups were performed using one-way analysis of variance followed by a post hoc test, or a Student's t test, as appropriate. Differences among means were considered significant when P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of ROS on IL-6 Production from Skeletal Myoblasts and Myotubes

Differentiation of C2C12 myoblasts to myotubes was confirmed by light microscopy, as differentiated muscle cells appeared multinucleated, elongated, and fusing into tubes. In addition, using a Myo D-specific antibody, we determined an increase in the expression of this myogenic HLH-transcription factor in myotubes (data not shown).

Myotubes exhibited higher basal secretion of IL-6 compared with myoblasts (0.17 ± 0.01 versus 0.056 ± 0.001 pg/mg protein). To investigate the effects of exposure to ROS on IL-6 production, C2C12 cells were exposed to pyrogallol (PYR) and IL-6 was measured in the supernatant after 24 h. Unlike myoblasts that did not respond to ROS-generating agents, skeletal myotubes incubated with PYR exhibited a 12.2-fold increase in IL-6 release (Figure 1A). Another potential cellular source of IL-6 in the skeletal muscle tissue in vivo is the endothelial cell. To test if endothelial cells can release IL-6 in response to ROS, MAEC were exposed to the indicated concentrations of PYR or X/XO for 24 h. ROS-treated MAEC cells did not exhibit an increase in IL-6 production (Figure 1B). Similarly, SV-40 transformed mouse EC derived from lymph node stroma, a murine microvascular endothelial cell line, failed to increase IL-6 release when exposed to ROS-producing agents (data not shown).

View larger version (9K): [in this window] [in a new window]   View larger version (14K): [in this window] [in a new window]   Figure 1.   Reactive oxygen species stimulate IL-6 production in myotubes. (A) C2C12 nondifferentiated skeletal myoblasts or C2C12-differentiated skeletal myotubes were exposed to the ROS-producing reagent, pyrogallol (PYR, 250 µM), for 24 h. IL-6 production in the supernatants was determined by ELISA. Values are means ± SEM; n = 9-12, *P < 0.05 from control. (B) Murine aortic endothelial cells were exposed to the indicated concentrations of pyrogallol (PYR) or xanthine/xanthine-oxidase (X/XO) for 24 h. IL-6 production in the supernatants was determined by ELISA. Values are means ± SEM; n = 3.

ROS-induced IL-6 release by C2C12 myotubes was concentration-dependent with maximal IL-6 release being observed at 330 µM/50 mU and 600 µM for X/XO and H₂O₂, respectively (Figure 2). Pretreatment of myotubes with superoxide dismutase (SOD, 300 U/ml) or catalase (CAT, 2,000 U/ml) blocked the ROS-induced IL-6 release (Figure 3). Although the highest concentrations of ROS-producing agents did exert toxic effects on myotubes as assessed by a decrease in cell number and protein content (data not shown), upregulation of IL-6 production was also observed under conditions where toxicity by ROS could not be documented by the methods used.

View larger version (11K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 2.   Concentration dependence of ROS-induced IL-6 production. C2C12 skeletal myotubes were exposed to the indicated concentration of hydrogen peroxide (H2O2; A) or xanthine/xanthine-oxidase (X/XO; B) for 24 h. IL-6 production in the supernatants was determined by ELISA. Values are means ± SEM; n = 3, *P < 0.05 from control.

View larger version (12K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 3.   Superoxide dismutase and catalase inhibit ROS-stimulated IL-6 release. C2C12 skeletal myotubes were pretreated with either (A) superoxide dismutase (SOD, 300 U; striped bars) or (B) catalase (CAT, 2,000 U; striped bars) for 15 min, and then exposed to the ROS-producing agent shown (pyrogallol [PYR], 250 µM; xanthine/xanthine-oxidase [X/XO], 330 µM/50 mU; or H2O2 500 µM) for 24 h. Solid bars indicate vehicle in both A and B. IL-6 production in the supernatants was determined by ELISA. Values are means ± SEM; n = 3, *P < 0.05 from control.

To determine the importance of endogenously produced ROS in the increase of IL-6 production, C2C12 were exposed to TNF- with or without pretreatment with an anti-oxidant. Cells exposed to N-acetyl-cysteine did not exhibit increased IL-6 production in response to TNF- stimulation (Figure 4).

View larger version (9K): [in this window] [in a new window]   Figure 4.   Inhibition of endogenous ROS blocks TNF--induced IL-6 secretion. C2C12 skeletal myotubes were pretreated with vehicle or N-acetyl-cysteine (10 mM) for 30 min and then exposed to TNF- (10 ng/ml). IL-6 was determined in the cell culture supernatants after 24 h. Values are means ± SEM; n = 6, *P < 0.05 from control.

ROS-Induced IL-6 Release Is Transcription-Dependent

To investigate the mechanism of ROS-induced IL-6 release, we determined steady-state mRNA levels in myotubes after exposure to H₂O₂ (Figure 5A); data from these experiments showed a 5-fold increase in IL-6 mRNA following H₂O₂ treatment. Moreover, pretreatment of skeletal myotubes with either actinomycin D (ACT D 0.5 µM) or cycloheximide (CHX 30 µM) for 2 h before incubation with PYR or X/XO resulted in complete inhibition of the ROS-stimulated IL-6 release (Figure 5B), suggesting that ROS stimulate IL-6 production via a transcription-dependent mechanism.

View larger version (50K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 5.   ROS-induced IL-6 secretion is transcription-dependent. (A) Lanes 1-8 show the products from the amplification of 5 µg cDNA in the presence of competitor from either control cells or cells treated with 500 µM H2O2 for 6 h. The concentration of the competitor in the first reaction (lane 1) was 0.5 ng and subsequent lanes had a 5× dilution of the competitor (native IL-6 band is present at 200 bp whereas the 400-bp band corresponds to the competitor). The lane in which bands of equal intensity were observed (lane 4 for the CTL and lane 3 for the H2O2) is considered the isopoint of the PCR reaction where equal starting amounts of IL-6 cDNA and IL-6 competitor were present in the PCR reaction mix (4 pg for the control and 20 pg for the H2O2). The picture shown is representative of an experiment performed two times with similar results. (B) C2C12 myotubes were pretreated with actinomycin D (ACT D, 0.5 µM) or cycloheximide (CHX, 30 µM) for 2 h, followed by exposure to pyrogallol (250 µM) or H2O2 (500 µM) for 24 h. Values are means ± SEM; n = 3, *P < 0.05 from control.

Role of p38 and NF-B in ROS-Stimulated IL-6 Release

Evidence in the literature suggests that activation of p38 plays an important role in IL-6 from myocytes (14). To study the involvement of p38 in the ROS-induced IL-6 release, C2C12 were pretreated with the pharmacologic inhibitor of p38 kinase SB203580 before the H₂O₂ exposure (Figure 6). SB203580 blocked the H₂O₂-induced increase in IL-6 production, suggesting that ROS-induced IL-6 release depends on p38 activation.

View larger version (9K): [in this window] [in a new window]   Figure 6.   Inhibition of p38 blocks the H2O2-induced IL-6 production. C2C12 skeletal myotubes were pretreated with vehicle (dimethyl sulfoxide) or SB203580 (2 µM) for 1 h and then exposed to H2O2 (500 µM) for 24 h. Supernatants were then collected and IL-6 measured by ELISA. Values are means ± SEM; n = 3, *P < 0.05 from control.

To test the involvement of NF-B in the ROS-induced IL-6 release, C2C12 myotubes were pretreated with H₂O₂ and IB- phosphorylation and degradation was determined by Western blotting (Figure 7A). H₂O₂ induced a time-dependent phosphorylation of IB- and promoted its degradation, suggesting that this pathway may contribute to the ROS-induced IL-6 production. To demonstrate that NF-B activation by ROS in skeletal myotubes leads to an increase in gene expression, we transfected C2C12 with a plasmid expressing the luciferase reporter gene under the control of six B sites (Figure 7B). Treatment of the transiently transfected C2C12 with PYR and X/XO induced a 4.2- and a 2.3-fold increase in NF-B-dependent luciferase activity. Using an AP-1-driven reporter gene, we observed a 2.4- and a 3.4-fold increase in luciferase activity in response to PYR and X/XO. Furthermore, pretreatment of myotubes with diethyldithiocarbamic acid (DETC, 100 µM) for 2 h, resulted in partial inhibition of PYR-stimulated IL-6 release and complete inhibition of H₂O₂-induced IL-6 increase (Figure 8A). Moreover, transient transfection with a form of IB- that does not get phosphorylated by IKK and thus behaves as a transdominant repressor of IB- blocked the H₂O₂-stimulated increase in IL-6 production (Figure 8B), suggesting that ROS-induced IL-6 release is NF-B-mediated.

View larger version (45K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 7.   ROS activate NF-B in C2C12 cells. (A) Cells were serum starved overnight and exposed to hydrogen peroxide (H2O2, 500 µM) for the indicated time. C2C12 were then lysed, processed by SDS-PAGE, transferred on nitrocellulose membranes, and blotted with an IB- or a phospho-specific IB- antibody. Equal protein loading was confirmed using a -actin antibody. Blots shown are representative of three independent experiments. (B) C2C12 skeletal myoblasts were transiently transfected and differentiated into myotubes. Myotubes were then incubated with either pyrogallol (PYR, 250 µM) or xanthine/xanthine-oxidase (330 µM/50 mU) for 24 h. Cells were lysed and luciferase/-gal activities were measured. RLU values are calculated as the difference in the luciferase/-gal ratio exhibited by pNF-B or pAP-1 minus pTAL exposed to vehicle, (control), PYR, or X/XO. Values are the means ± SEM; n = 4, *P < 0.05 from control.

View larger version (11K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   Figure 8.   NF-B inhibition blocks ROS-stimulated IL-6 release. (A) Cells were pretreated with the inhibitor of NF-B diethyldithiocarbamate (DETC, 100 µM) for 2 h and then incubated with pyrogallol (PYR, 250 µM) or hydrogen peroxide (H2O2, 300 µM) for 24 h. Values are means ± SEM; n = 3, *P < 0.05 from control. (B) C2C12 skeletal myoblasts were transiently transfected with a control vector or a mutant non-degrading form of IB- that acts as a dominant repressor. After 48 h cells were incubated with H2O2 (500 µM) and cell culture supernatants were collected after an additional 24 h for IL-6 determination. Values are means ± SEM; n = 9, *P < 0.05 from control.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Plasma levels of IL-6 have consistently been reported to rise after strenuous exercise (2, 6, 7, 15). In line with this observation, we recently reported that resistive breathing, a form of strenuous exercise for the respiratory muscles, results in elevated plasma levels of IL-6 (16). Production of IL-6 in contracting human skeletal leg muscles (as assessed by the arterial-venous IL-6 concentration difference of the exercising leg) can account for the exercise-induced increase in plasma IL-6 (17). Furthermore, Jonsdottir and coworkers reported that electrically induced muscle contractions induce IL-6 mRNA expression in rat skeletal muscles, and Ostrowski and colleagues observed increased IL-6 mRNA expression in five out of eight leg muscle biopsy homogenates of marathon runners after the race (4, 18). Together, these results suggest that contracting muscles are the tissues producing IL-6 during intense exercise. Although previous studies have excluded blood mononuclear cells and peripheral blood leukocytes as the source of IL-6 in exercise models, the cellular source of IL-6 production is currently unknown (3, 19, 20).

To test if myocytes produce IL-6 and to identify possible stimuli for its release that are relevant to exercise, we used the murine C2C12 skeletal muscle cell line. Similarly to what has been reported for cultured human myocytes (13), both C2C12 myotubes and myoblasts produce IL-6 under basal conditions. Interestingly, exposure of myotubes to ROS, which are known to accumulate intramuscularly during intense exercise, resulted in increased IL-6 production in a concentration-dependent-manner that was abolished by pretreatment with the antioxidant enzymes SOD and CAT. The ROS-stimulated IL-6 release was restricted to differentiated muscle cells, whereas undifferentiated myocytes and endothelial cells did not respond to ROS-generating agents. The increase in IL-6 production in response to ROS-generating agents is largely cell-specific. Normal human bronchial epithelial cells and the epithelial cell line HS-24 produce IL-6 when exposed to superoxide anions (O₂), but not H₂O₂, whereas human lung fibroblasts (WI-38-40) respond to neither agent (21). Moreover, ROS generated by increasing pO₂ stimulate IL-6 production in alveolar epithelial cells and ROS produced during hypoxia increase IL-6 from human endothelial cells (22, 23). The differential effect of PYR on IL-6 release from myoblasts versus myotubes might reflect differences in the expression of critical signaling components required for ROS-stimulated IL-6 expression that are upregulated during differentiation of myoblasts into myotubes. Alternatively, myoblasts might possess increased anti-oxidant defense mechanisms.

To establish a role for endogenously produced ROS in the IL-6 release, C2C12 cells were exposed to TNF-, a cytokine known to trigger the release of H₂O₂, in skeletal myocytes (24). Stimulation of cells with TNF- caused an increase in IL-6 production that was abolished by pretreatment of cells with the antioxidant N-acetyl-cysteine. These data suggest that similarly to exogenously supplied ROS, endogenously produced ROS following ligand binding trigger IL-6 production.

Regulation of IL-6 production has been shown to occur both at the transcriptional and post-transcriptional level (25). In alveolar type II epithelial cells X/XO- and H₂O₂-stimulated IL-6 production is not inhibited by inhibition of RNA polymerase by ACT D, but requires new protein synthesis (22). In human bronchial epithelial cells ROS-stimulated IL-6 production is preceded by increases in IL-6 steady-state mRNA levels (21). To examine the mechanism responsible for the increase in IL-6 production in skeletal myotubes following exposure to ROS-generating agents, we determined IL-6 steady-state mRNA levels. Exposure of cells to H₂O₂ caused an increase in IL-6 mRNA levels. Moreoever, pretreatment of C2C12 cells with a pharmacologic inhibitor of transcription (ACT D) or a pharmacologic inhibitor of protein synthesis (CHX) abolished the PYR- and X/XO-induced increase in IL-6 release, suggesting that in skeletal myotubes ROS induce IL-6 production through activation of the transcriptional machinery.

ROS activate several signaling pathways in mammalian cells (11). One family of signaling kinases that contribute to cytokine gene expression is the mitogen-activated protein kinase family. In cardiac myocytes, activation of the p38 pathway mediates the increase in IL-6 production through activation of NF-B (14). As ROS trigger activation of p38 (11), we tested the involvement of this kinase in the ROS-induced IL-6 release. C2C12 cells pretreated with the inhibitor of p38 SB203580 failed to exhibit increased production of IL-6 in response to H₂O₂ exposure, providing evidence that activation of p38 is crucial for the ROS-induced IL-6 release.

The mouse IL-6 promoter contains a number of cis-acting elements, including consensus sequences for the redox-sensitive transcription factors NF-B and AP-1. The NF-B and AP-1 elements are located within the first 300 bp of the 5'-flanking region that is highly conserved across species (> 80%) (26). NF-B is involved in cytokine expression and can also be activated by oxidant and physical stress (27, 28). Increased IL-6 gene expression in response to a variety of stimuli including lipopolysaccharide, TNF-, IL-1 and ionizing radiation correlates strongly with activation of NF-B (29). Mutations in the B site of the human IL-6 promoter abolishes lipopolysaccharide-induced promoter activity, suggesting that this single regulatory element is crucial for LPS inducibility (29).

In most cell types, NF-B is maintained in an inactive form in the cytoplasm through association with IBs (33). Phosphorylation of the inhibitory subunit is followed by ubiquitination and degradation of IB by the 26S proteasome; the NF-B dimer then translocates to the nucleus, where it binds to its cognate DNA element and activates transcription (28). To test the involvement of NF-B in the ROS-stimulated increase in IL-6 release myotubes were exposed to H₂O₂ and the phosphorylation status of IB- was examined. In agreement with previously published data, exposure of myotubes to H₂O₂ induced a time-dependent phosphorylation and degradation of IB-, suggesting that NF-B becomes activated in response to oxidative stress in skeletal muscle (13, 24). To confirm that ROS-induced phosphorylation and degradation of IB- is functionally important, we used a heterologous expression system. Exposure of myotubes transiently transfected with a plasmid expressing the luciferase reporter gene under the control of B elements exhibited increased luciferase activity when exposed to ROS-generating agents. Direct evidence for the involvement of NF-B in the ROS-stimulated increase in IL-6 production is offered by experiments utilizing pharmacologic and gene-transfer approaches. Pretreament of cells with DETC, a pharmacologic inhibitor of NF-B, resulted in blockade of ROS-induced IL-6 production, and expression of a mutant form of IB- abolished the H₂O₂-induced IL-6 release.

Exposure of myotubes transfected with the luciferase gene under the control of AP-1 upstream sites also led to an increase in luciferase activity when cells were incubated with ROS-generating agents. AP-1 cis-acting elements are known to contribute to platelet-derived growth factor BB and transforming growth factor 1-induced IL-6 expression in rat osteoblasts and human lung fibroblasts, respectively (34, 35). It should be noted that exposure of cells to both agents is accompanied by an increase in intracellular ROS generation (10). Although both DETC and the IB- mutant abolished the H₂O₂-induced IL-6 production, we cannot rule out a role for AP-1 in the ROS-induced IL-6 production, as IL-6 expression might require AP-1/NF-B costimulation of the promoter region.

The importance of increased IL-6 production within the contracting skeletal muscles is unclear. Acutely, IL-6 might have a hormone-like glucoregulatory role, signaling that contracting muscle glycogen stores are reaching critically low levels and stimulating hepatic glucose output to maintain glucose homeostasis (36, 37). Moreover, prolonged elevation of circulating IL-6 has been implicated in the induction of delayed onset muscle soreness (2). This is in concert with its suggested role as a locally produced soluble mediator pointing to the site of injury, which in the case of intense exercise is the skeletal muscle (15, 38). However, IL-6 might also have beneficial effects, by inducing muscle proteolysis, initiating removal of injured myocytes, and allowing for muscle fiber regeneration to begin (39). Furthermore, IL-6 is a very potent stimulant of the hypothalamic-pituitary axis, leading to increases in the circulating levels of ACTH (40). The ACTH response would stimulate glucocorticoid secretion, which would suppress the late onset inflammatory response within the muscle. Therefore, overproduction of IL-6 by skeletal muscle may also be a protective mechanism, occurring in response to excess physical stress.

In conclusion, we have shown that ROS stimulate IL-6 release from skeletal myotubes, therefore identifying a novel cellular source which could be responsible for the excess amounts of this inflammation-responsive cytokine produced in strenuous-exercise models. Our results also indicate that the ROS-stimulated increase in IL-6 release is transcription-dependent and involves p38 and NF-B activation.