Toll-like receptors (TLRs) comprise a critical sentinel that monitors body compartments for the presence of pathogens. Skeletal muscle expresses TLRs and responds to pathogen-associated molecular patterns (PAMPs), such as lipopolysaccharide (LPS), by mounting an innate immune response. In the present study, we used C2C12 myocytes as a model system for skeletal muscle during infection. C2C12 cells responded to LPS in a time frame and with a pattern of gene expression that faithfully mimicked the response of skeletal muscle to LPS in vivo. LPS from a variety of Escherichia coli serotypes stimulated IL-6 synthesis. C2C12 cells expressed TLR1–7, but not TLR8 or TLR9, mRNA by RT-PCR. A synthetic tripalmitoylated cysteine-, serine-, and lysine-containing peptide (Pam) and LPS from Porphyromonas gingivalis, two TLR2 ligands, also stimulated IL-6 expression. LPS and Pam stimulated luciferase activity driven from NF-κB and IL-6 promoter-containing plasmids, and this response was blunted when the NF-κB binding site was mutated. LPS- and Pam-stimulated IL-6 expression was inhibited by the proteasome inhibitor MG-132 and the IκB kinase-2 (IKK2) inhibitor 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide (TPCA-1). Pam-stimulated NF-κB and IL-6 promoter activities were disrupted by a dominant-negative form of TLR2, but not TLR4. Local injection of LPS or Pam into the gastrocnemius muscle stimulated IL-6 mRNA expression in the injected, but not the contralateral, muscle. The LPS- but not Pam-stimulated expression of IL-6 mRNA was blunted in skeletal muscle of mice carrying an inactivating mutation in TLR4. The data suggest that skeletal muscle and muscle cells recognize pathogen-associated molecules with specific TLRs to initiate an IL-6 transcriptional response.
- skeletal muscle
- muscle cells
- immune response
skeletal muscle has the potential to be exposed to multiple pathogen-associated molecules. During infection, this includes bacterial cell wall components (9, 32), bacterial DNA (12), and viral RNA (24). These molecules represent pathogen-associated molecular patterns (PAMPs) that are recognized by the host's defense system (31). The afferent limb of the immune system comprises a series of pattern recognition receptors with homology to the Drosophila Toll protein. The Toll-like receptors (TLRs) specifically recognize PAMPs (2, 33), which stimulate the efferent limb of the immune system to secrete cytokines and activate macrophages (44, 45). Skeletal muscle also releases nitric oxide (NO) to kill invading microorganisms, but excess NO can also impair contractility and contribute to muscle wasting (39).
Two of the most studied PAMPs are acetylated and palmitoylated lipoproteins and peptides, which bind to TLR2, and LPS (endotoxin) from Escherichia coli, which binds to TLR4. A tripalmitoylated peptide containing cysteine, serine, and lysine (Pam) induces a robust cytokine response in vivo and in vitro by binding to TLR2, and Pam and LPS are widely used to mimic the systemic and molecular effects of gram-positive and -negative bacteria, respectively (10, 42).
Recently, we demonstrated that intraperitoneal injection of LPS stimulates the expression of multiple cytokines in rat and mouse skeletal muscle (25). LPS specifically utilizes TLR4 to increase cytokines and signaling molecules, such as the inhibitor of NF-κB (IκB) in muscle (16). LPS fails to induce the mRNA for these proteins in C3H/HeJ mice, which harbor a mutation in the LPS receptor (TLR4), suggesting that TLR4 is necessary for the LPS response in vivo.
Because skeletal muscle is composed of many cell types, we have examined the specific effect of LPS on the inflammatory response in muscle cells in culture (8, 13, 14). LPS stimulated IL-6 and NO synthase-2 (NOS2) expression in C2C12 myocytes and primary cultures of human myocytes (13). LPS also stimulated downstream signaling components of TLR4, including phosphorylation of IL receptor-associated kinase (IRAK-1), degradation of IκB, activation of JNK, and stimulation of an NF-κB reporter plasmid (14, 16). TLR4 expression is necessary for LPS-stimulated NF-κB activity in C2C12 cells, because a dominant-negative form of the receptor with a mutation in the Toll IL-1 receptor (TIR) domain inhibits luciferase activity driven off an NF-κB reporter plasmid (16).
Because skeletal muscle may be exposed to a variety of PAMPs from gram-positive and -negative bacteria, as well as viruses, we have examined the expression of TLR1–9 mRNA in C2C12 cells and skeletal muscle. In addition, the functional behavior of the receptors was compared by challenging TLR2 with Pam, TLR3 with polyinosinic-polycytidylic acid [poly(I:C)], TLR4 with LPS, TLR7 with imidiquimod (R837), and TLR9 with bacterial DNA to stimulate IL-6 synthesis. We found that TLR2 and TLR4, but not TLR3 and TLR9, ligands potently stimulate IL-6 protein and mRNA expression. The response to TLR2 and TLR4 ligands is transcriptionally mediated, because it can be blocked by the transcriptional inhibitor 5,6-dichloro-β-d-ribofuranosylbenzimidazole (DRB). The LPS- and Pam-induced increase in IL-6 was also blocked by 2-[(aminocarbonyl)amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide [TPCA-1, an IκB kinase (IKK2) inhibitor], MG-132 (a proteasome inhibitor), and dexamethasone (an anti-inflammatory). In addition, TLR2 ligands stimulate NF-κB and IL-6 promoter activity, and this response was specifically blocked by a dominant-negative form of TLR2, but not TLR4.
Because PAMPs stimulate a systemic and a local release of cytokines, we also tested whether direct intramuscular injection of Pam and LPS stimulated the local expression of IL-6 within skeletal muscle. Both ligands stimulated IL-6 expression in the injected, but not the contralateral, muscle. In addition, LPS-stimulated IL-6 expression was normal in wild-type mice, whereas the IL-6 response to LPS, but not Pam, was markedly attenuated in C3H/HeJ mice, which carry a TLR4 mutation. Our data suggest that skeletal muscle and muscle cells respond to specific TLR ligands to initiate a robust innate immune response.
MATERIALS AND METHODS
Experimental protocol for C3H/HeJ mice.
C3H/HeSnJ and C3H/HeJ mice (Jackson Laboratories, Bar Harbor, ME) were housed in a controlled environment and provided water and rodent chow ad libitum for 2 wk before they were used in experiments. Mice (8–9 wk old, 25.0 ± 0.3 g body wt) were randomly assigned to one of six groups that received intramuscular injections: C3H/SnJ mice injected with saline (Wt/Sal), C3H/HeJ mice injected with saline (HeJ/Sal), C3H/SnJ mice injected with LPS (Wt/LPS), C3H/HeJ mice injected with LPS (HeJ/LPS), C3H/SnJ mice injected with Pam (Wt/Pam), and C3H/HeJ mice injected with Pam (HeJ/Pam). In the experiment depicted in Figs. 10 and 11, mice were injected intramuscularly with Pam (InvivoGen, San Diego, CA; 100 μg/muscle) or LPS (Sigma-Aldrich, St. Louis, MO; 50 μg/muscle). The doses of LPS and Pam were based on a preliminary dose-response study and are similar to those used by other investigators (37). Some mice in each group were injected with the same volume (50 μl) of sterile 0.9% saline in the hindlimb muscles. After 4 h, the mice were anesthetized with a mixture of ketamine (Fort Dodge Animal Health, Fort Dodge, IA) and xylazine (Bayer, Shawnee Mission, KS) at 90 and 9 mg/kg, respectively. This time point was chosen because it corresponds to the peak of IL-6 mRNA expression in mice and rats injected intraperitoneally with LPS. Blood was collected from the inferior vena cava in heparinized syringes. Hindlimb skeletal muscle (gastrocnemius and plantaris) from both legs was dissected from each animal, wrapped in aluminum foil, and frozen in liquid nitrogen. Mice were killed by cardiac excision and subsequent exsanguination. Tissues were later powdered under liquid nitrogen using a mortar and pestle and stored at −70°C. All experiments were approved by the Institutional Animal Care and Use Committee at the Pennsylvania State University College of Medicine and adhere to the National Institutes of Health guidelines for the use of experimental animals.
The C2C12 mouse myoblast cell line (American Type Culture Collection, Manassas, VA) was used for all studies. Cells were grown in 100-mm petri dishes (Greiner Bio-One, Frickenhausen, Germany) and cultured in minimal essential medium containing 10% fetal bovine serum, penicillin (100 U/ml), streptomycin (100 μg/ml), and amphotericin B (25 μg/ml; all from Mediatech, Herndon, VA). Cells were grown to confluence and switched to fresh serum-containing medium before addition of LPS, cytokines, or other agents. C2C12 cells were used as mixed cultures of myoblasts and myotubes. In the experiment depicted in Fig. 1B, the cells were switched to medium containing 10% bovine calf serum and allowed to differentiate into myotubes for >2 wk to confirm the expression of TLR2 and TLR4 in cultures comprising >90% myotubes. E. coli LPS (026:B6; Sigma Aldrich) and ultra pure E. coli LPS 011:B4 and K12, poly(I:C), and an unmethylated oligodeoxynucleotide (ODN 2216; 5′-ggGGGACGATCGTCgggggg-3′) and its negative control (5′-ggGGGAGCATGCTGcggggg-3′; InvivoGen) were added to the cells to mimic bacterial and viral components, which have the potential to stimulate an immune response and bind to TLRs. A variety of compounds were purchased from Calbiochem (La Jolla, CA), including MG-132, TPCA-1, and DRB, and used to characterize the response of myocytes to LPS and Pam.
Transient transfection assays.
C2C12 cells were plated at 50% confluence in 24-well plates. Cells were switched to serum-free medium and transfected with a pNFκB-Luc reporter vector (BD Biosciences, Palo Alto, CA), the full-length IL-6 promoter, or a vector in which the NF-κB binding site was mutated (provided by Dr. O. Eickelberg, Yale University, New Haven, CT). All cells were also transfected with a pRL-SV40 control vector expressing luciferase from Renilla reniformis (Promega, Madison, WI). Plasmids were added as preformed complexes with Lipofectamine 2000 at a lipid-to-DNA ratio of 5:1. After 2 h, cells were allowed to recover in serum-containing medium for 16 h. Cell extracts were isolated at various times after the addition of LPS in reporter lysis buffer (Promega) and frozen until assay. In some transfection studies, the cells also received a third plasmid expressing a dominant-negative form of TLR2 or TLR4 (pZERO-mTLR2 or pZERO-mTLR4; InvivoGen) in which the TIR domain was deleted from the mouse TLR2 and TLR4 genes. Luciferase reporter activity was measured with firefly and Renilla luciferase assay reagents (dual luciferase kit, Promega) on a luminometer set for an integration time of 9 s (Turner Biosystems, Sunnyvale, CA). Renilla luciferase activity was used to normalize for transfection efficiency.
RNA isolation and RNase protection assay.
Total RNA, DNA, and protein were extracted from C2C12 cells or tissues in a mixture of phenol and guanidine thiocyanate (TriReagent, Molecular Research Center, Cincinnati, OH) using the manufacturer's protocol. RNA was separated from protein and DNA by the addition of bromochloropropane and precipitation in isopropanol. After a 75% ethanol wash and resuspension in formamide, RNA samples were quantified by spectrophotometry; 10 μg of RNA were used for each assay. Riboprobes were synthesized from a custom multiprobe mouse template set containing a probe for IL-6, suppressor of cytokine signaling (SOCS3), NOS2, and L32 mRNA detection (Pharminigen, San Diego, CA). The labeled riboprobe was hybridized with RNA overnight using an RNase protection assay (RPA) kit according to the manufacturer's protocol (Pharminigen). Protected RNAs were separated using a 5% acrylamide gel (19:1 acrylamide-bisacrylamide). Gels were transferred to blotting paper and dried under vacuum on a gel dryer. Dried gels were exposed to a PhosphorImager screen (Molecular Dynamics, Sunnyvale, CA), and the resulting data were quantified using ImageQuant software and normalized to the mouse ribosomal protein L32 mRNA signal in each lane.
Northern blot analysis was performed to quantify TLR2 and TLR4. Total RNA (25 μg) was electrophoresed under denaturing conditions on a 1.1% agarose gel containing 6% formaldehyde. RNA was transferred to Nytran Supercharge membranes (Schleicher & Schuell, Keene, NH) using a TurboBlotter (Schleicher & Schuell). After they were baked, the blots were hybridized at 42°C in ULTRAhyb (Ambion, Austin, TX). Oligonucleotides for TLR2 (5′-GCTGCTGTGAGTCCGGAGGGAATAGAGGTGAAAGACCTGG-3′), and TLR4 (5′-CCGGCTCTTGTGGAAGCCTTCCTGGATGATGTTGGCAGC-3′) were labeled by Tet (Promega) tailing with [32P]dATP (Amersham, Arlington Heights, IL). For normalization of RNA loading, a rat 18S oligonucleotide was labeled in the same manner. Oligonucleotides were synthesized by the Macromolecular Core Facility at the Pennsylvania State College of Medicine using a Perceptive Biosystems Expedite 8909 nucleic acid synthesizer. All membranes were washed twice in 2× saline-sodium citrate-0.1% SDS at 42°C for 5 min and once in 0.2× saline-sodium citrate-0.1% SDS at 42°C for 15 min. Membranes were exposed to a PhosphorImager screen, and the resultant data were calculated as described above for the RPA.
Total RNA was isolated from C2C12 cells and mouse skeletal muscle and spleen. The RNA was treated with DNase I to remove contaminating genomic DNA. All RNA was reverse transcribed with the Superscript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA). PCR was conducted using HotStarTaq DNA polymerase (Qiagen, Valencia, CA) at 95°C for 15 min followed by cycles of 30 s each at 94°C, 55°C, and 72°C and termination at 72°C. For the C2C12 data presented in Fig. 8A, a total of 35 PCR cycles were conducted with primers for TLR1–7. For analysis of TLR8 and TLR9 in Fig. 8B, 35 PCR cycles were conducted, whereas TLR2, TLR4, and hypoxanthine phosphoribosyltransferase (HPRT) underwent 30, 30, and 25 PCR cycles, respectively. All cycle levels were nonsaturating. The primer sequences for the mouse TLR RT-PCR were identical to those previously reported by Caramalho et al. (7). PCR products were run out on a 1.2% agarose gel containing ethidium bromide (Fisher Scientific, Pittsburg, PA), and the image was captured with a Gene Genius Bio Imaging System (Syngene, Frederick, MD). TLR2 and TLR4 mRNA levels were normalized to HPRT mRNA in the same sample. Each reaction was conducted with RNA that had been incubated with or without RT to eliminate the possibility of a false-positive signal generated from genomic DNA.
Values are means ± SE. Unless otherwise noted, each experimental condition was tested in triplicate and each experiment was repeated twice. Data were analyzed by analysis of variance followed by Student-Newman-Keuls test. Statistical significance was set at P < 0.05. For the study in which LPS and Pam were injected intramuscularly, the number of mice per group was five for HeJ/Sal, HeJ/LPS, SnJ/Sal, and SnJ/LPS and nine for HeJ/Pam and SnJ/Pam.
TLR2 and TLR4 ligands dose and time dependently stimulate IL-6 protein and mRNA expression.
To test whether TLR2 and TLR4 ligands increase IL-6 expression in myoctyes, we first compared the ability of different LPS serotypes to stimulate IL-6 synthesis in C2C12 cells. Ultrapure LPS from E. coli 011:B4 and E. coli K12 stimulated IL-6 synthesis equivalently, with an ED50 of 0.1 ng/ml. Both ultrapure LPS preparations were 100-fold more potent than a crude preparation of LPS from the E. coli 026:B6 strain (Fig. 1A). C2C12 cells expressed TLR2 and TLR4 mRNA in myoblasts and myotubes, suggesting that C2C12 cells express these TLRs independent of their state of differentiation (Fig. 1B). The TLR2 and TLR4 receptors were functional, as evidenced by the stimulation of IL-6 synthesis by two TLR2 ligands (LPS from P. gingivalis and Pam) and E. coli LPS (Fig. 1C). In contrast, a double-stranded RNA mimetic and TLR3 ligand [poly(I:C)] failed to increase IL-6 synthesis (Fig. 1D). In general, the TLR2 ligand Pam was more potent than the TLR4 ligand LPS at stimulating IL-6 synthesis. LPS and Pam also increased IL-6 mRNA expression dose dependently (Fig. 2), and both ligands stimulated IL-6 mRNA and IL-6 protein with a similar potency (Figs. 1D and 2B).
We next determined whether TLR2 and TLR4 ligands increase IL-6 expression with a similar time course. Although TLR2 and TLR4 share many intracellular adapter and signaling molecules, one might anticipate that the two receptors would display some diversity in signal transduction or in the time course in which they stimulate IL-6 synthesis. A significant 30- to 40-fold increase in IL-6 mRNA was observed 2 h after stimulation with Pam or LPS. IL-6 mRNA remained elevated above control levels up to 24 h after exposure (Fig. 3, A and B). Pam and LPS stimulated IL-6 protein synthesis and secretion within 2 h. IL-6 protein levels reached a plateau between 2 and 4 h (Fig. 3C). Because PAMPs also increase other proinflammatory cytokines in vivo, we examined TNF expression in the C2C12 cells in the presence of the two TLR ligands. LPS and Pam increased TNF-α mRNA within 1–2 h (Fig. 3D). Although LPS and Pam increased TNF-α mRNA, the level of TNF-α protein in conditioned medium and cell extracts from C2C12 cells was below the limit of detection of our assay (15 pg/ml).
TLR2 and TLR4 ligands stimulate NF-κB and IL-6 promoter activity.
To determine whether TLR2 and TLR4 ligands stimulate IL-6 transcription and promoter activity, we pretreated C2C12 myocytes with the transcriptional inhibitor DRB and subsequently exposed the cells to LPS or Pam. LPS and Pam stimulated IL-6 protein expression over a 4-h time course, and the response to both ligands was completely blocked by pretreatment of the cells with DRB (Fig. 4A). DRB also completely blocked LPS- and Pam-stimulated IL-6 mRNA expression and the corresponding increases in LPS- and Pam-induced NOS2 and SOCS3 mRNA, as evidenced by RPA (Fig. 4B). Both ligands stimulated firefly luciferase activity when it was driven by an NF-κB reporter plasmid or a plasmid containing the full-length IL-6 promoter (Fig. 5A). Pam stimulated both reporters with an identical dose-response curve, and the transcriptional responses exhibited a dose-response pattern comparable to those seen for the induction of IL-6 mRNA and protein (cf. Figs. 1D, 2B, and 5B). LPS-stimulated IL-6 promoter activity was largely dependent on the presence of the IL-6 promoter's NF-κB binding site. Mutation of the NF-κB site reduced LPS-induced promoter activity by 60%.
TLR signaling is restricted to TLR2, TLR4, and TLR7 in C2C12 cells.
To determine whether other TLR ligands could stimulate IL-6 synthesis, we tested whether an oligonucleotide containing CpG dinucleotides in the context of the core sequence GACGTT would stimulate IL-6 synthesis. This sequence is an optimal sequence for stimulating mouse TLR9 (3) and mimics bacterial DNA in some cell types. Although LPS and Pam stimulated IL-6 synthesis, the bacterial DNA mimetic and its negative control oligonucleotide did not stimulate IL-6 synthesis (Fig. 6A). In addition, the bacterial DNA mimetic did not synergize or antagonize the LPS-induced increase in IL-6. To test whether LPS and Pam signal through a glucocorticoid-sensitive signal transduction pathway, we also examined whether the synthetic glucocorticoid dexamethasone could inhibit IL-6 synthesis. Dexamethasone inhibited LPS- and Pam-stimulated IL-6 synthesis (Fig. 6B).
Because TLR ligands may contain trace contaminants that have the potential to stimulate IL-6 synthesis, we also pretreated LPS and Pam with polymyxin B, a specific inhibitor of LPS activity. The LPS-induced increase in IL-6 synthesis was completely blocked by as little as 0.5 μg/ml of polymyxin B. In contrast, up to a 16-fold higher concentration of polymyxin B did not inhibit Pam-stimulated IL-6 synthesis (Fig. 6C). Polymyxin B also selectively inhibited LPS- but not Pam-induced IL-6 promoter activity in skeletal myocytes (Fig. 6D). Polymyxin B did not alter basal IL-6 promoter activity, suggesting that the reagents used for transfecting the myocytes were not contaminated with LPS.
Muscle and muscle cells express TLR1–7, but not TLR8 and TLR9.
Because single-stranded RNA viruses may also infect skeletal muscle, we tested whether a TLR7 ligand could stimulate an IL-6 response in C2C12 cells. The immune modulator R837, a potent TLR7 ligand, stimulated IL-6 protein and mRNA expression only minimally. More importantly, R837 enhanced the ability of LPS and Pam to stimulate IL-6 protein and mRNA expression (Fig. 7).
To obtain a comprehensive picture of the repertoire of TLRs expressed by C2C12 cells, we reverse transcribed total RNA from C2C12 cells and probed for individual TLRs by PCR. C2C12 cells were found to express the mRNA for TLR1–7, and the PCR signal was dependent on the presence of RT, in all cases ruling out potential contamination by genomic DNA (Fig. 8A). C2C12 cells did not express TLR8 or TLR9 mRNA after up to 35 PCR cycles, and these TLR mRNAs were also not detected in RNA isolated from mouse skeletal muscle. In contrast, TLR8 and TLR9 mRNA were readily detected in RNA isolated from the spleen when used as a positive control (Fig. 8B). TLR2, but not TLR4, mRNA levels were elevated in C2C12 cells treated with LPS or Pam, as determined by semiquantitative RT-PCR (Fig. 8C), in agreement with our previous finding, as determined by Northern blotting (16).
LPS and Pam activate IL-6 through an IKK- and proteasome-dependent pathway.
To test whether signaling pathways that activate NF-κB are involved in LPS- and Pam-induced IL-6 synthesis, we used proteasomal and IKK2 inhibitors to block LPS- and Pam-induced IL-6 expression. TLR2 and TLR4 ligands activate the 20S proteasome to degrade IκB. This facilitates translocation of NF-κB to the nucleus and the activation of NF-κB-dependent genes. Two key steps in this process are the phosphorylation of IκB by an IκB kinase (IKK2) and degradation of IκB by the proteasome. We examined whether the proteasomal inhibitor MG-132 and the IKK2 inhibitor TPCA-1 could block IL-6 synthesis. LPS and Pam stimulated IL-6 synthesis 20- to 30-fold (Fig. 9A). MG-132 completely inhibited LPS- and Pam-stimulated IL-6 synthesis. LPS-, Pam-, and TNF-stimulated IL-6 synthesis was also inhibited by the IKK2 inhibitor TPCA-1 (Fig. 9B).
A dominant-negative TLR2, but not TLR4, construct blocks Pam-induced NF-κB and IL-6 promoter activity.
To better define the receptor that Pam utilizes in skeletal myocytes to stimulate NF-κB and IL-6 promoter activities, we transfected C2C12 cells with dominant-negative forms of TLR2 and TLR4. These receptor constructs have a deletion in their TIR domain that allows for ligand binding but disrupts their intracellular interaction with adapter proteins such as myeloid differentiation marker 88. Pam-stimulated NF-κB and IL-6 promoter activity in the C2C12 cells was not altered by cotransfection with dominant-negative TLR4 (Fig. 10, A and B). In contrast, cotransfection of myocytes with dominant-negative TLR2 completely inhibited Pam-induced NF-κB and IL-6 promoter activity (Fig. 10, C and D).
Intramuscular injections of LPS and Pam directly stimulate IL-6 expression in skeletal muscle.
In the final study, we determined whether an intramuscular injection of Pam or LPS would stimulate IL-6 synthesis in skeletal muscle, because the systemic immune response to PAMPs also stimulates the synthesis of cytokines, such as IL-1 and TNF-α, which may secondarily affect IL-6 expression. This experimental paradigm would help determine whether TLR2 and TLR4 are functional not only in C2C12 myoblasts but also in bona fide skeletal muscle in vivo. Therefore, we examined IL-6 expression in gastrocnemius muscle after an intramuscular injection of LPS or Pam. LPS and Pam increased IL-6 mRNA expression fivefold in the gastrocnemius 4 h after injection compared with injection of an equal volume of saline (Fig. 11). IL-6 mRNA expression was restricted to the injected muscle, with no IL-6 expression detectable in the contralateral muscle. Injection of saline alone also slightly increased IL-6 mRNA expression, but this response was always less than that observed in muscle injected with TLR ligands.
To examine whether LPS stimulates IL-6 mRNA expression in skeletal muscle through the TLR4 receptor, we injected LPS and Pam intramuscularly into mice that harbor a mutation in TLR4 (C3H/HeJ mice). Pam, but not LPS, stimulated IL-6 mRNA expression in muscle from C3H/HeJ mice (Fig. 12). In comparison, LPS and Pam strongly stimulated IL-6 mRNA expression in wild-type C3H/SnJ mice.
The direct effects of pathogen-associated molecules on immune cells have been well characterized. Yet much less is known about the interaction of PAMPs with force-generating cells, such as skeletal myocytes, which constitute 40–50% of total body mass. In the present study, we examined the response of C2C12 myocytes to two PAMPs: LPS and Pam. These two ligands, but not double-stranded RNA or bacterial DNA, were able to stimulate an innate immune response in C2C12 cells, as evidenced by the synthesis of IL-6. This response suggests that skeletal myocytes have functional TLR2 and TLR4 but may lack other TLRs. In support of this conclusion, TLR2 and TLR4 mRNAs were detected in C2C12 myoblasts as well as differentiated myotubes, and this finding is consistent with the expression of TLR2 and TLR4 in rat and mouse skeletal muscle (25). In addition, C2C12 cells were found to express TLR1–7, but not TLR8 and TLR9, by RT-PCR, suggesting that myocytes have the potential to respond to many TLR-specific ligands.
The presence of functional TLR2 and TLR4 in skeletal muscles was confirmed by the intramuscular injection of Pam and LPS. Both ligands increased IL-6 mRNA expression in vivo. Muscle injury induced by the injection of saline also increased IL-6, and this was consistent with eccentric exercise resulting in muscle injury, JNK activation, and IL-6 synthesis (4). In our experience, injection of TLR ligands such as LPS and Pam always gave a more robust increase in IL-6 mRNA expression than saline injection. Thus PAMPs initiate an innate immune response that is more pronounced than the response to simple muscle injury alone. Furthermore, this response is unlikely to be mediated by a secondary systemic release of cytokines, because the contralateral control muscle did not demonstrate an IL-6 response of a similar magnitude.
The LPS-induced increase in intramuscular IL-6 mRNA was specific to TLR4, because it occurred in wild-type mice, but not in C3H/HeJ mice that carry a mutation in the receptor's TIR domain. By comparison, intramuscular injection of the TLR2 ligand Pam increased IL-6 mRNA equivalently in wild-type and C3H/HeJ mice, suggesting that Pam employs a separate receptor (presumably TLR2) to mount an immune response in skeletal muscle. We confirmed the role of TLR2 in the response to Pam in C2C12 cells transfected with dominant-negative versions of TLR2 and TLR4. Pam increased NF-κB and IL-6 promoter activity in C2C12 cells, and this was completely blocked in cells cotransfected with dominant-negative TLR2, but not dominant-negative TLR4.
We utilized IL-6 as a marker cytokine for the innate immune response of myocytes and skeletal muscle, but LPS has been demonstrated to stimulate the expression of a variety of other pro- and anti-inflammatory molecules as well (15). Although LPS and Pam stimulated TNF-α mRNA expression in this study, we previously determined that this response is very short-lived compared with IL-6 mRNA (15). The short-term expression of TNF-α mRNA may explain why we have been unable to detect TNF-α protein in C2C12-conditioned media or cell extracts. Recent studies provide multiple potential explanations for the transient expression of TNF-α in C2C12 cells. Motoyama et al. (30) found that an endogenous inhibitor of NF-κB (IκB-ζ), when overexpressed in RAW264.7 cells, inhibits TNF-α synthesis but enhances IL-6 expression. We have found that LPS and Pam induce a sustained expression of IκB-ζ that lasts for ≥60 h (unpublished observation), and we hypothesize that this would limit TNF-α mRNA while simultaneously enhancing IL-6 expression in myocytes. TNF-α mRNA is also subject to translational regulation by RNA binding proteins. T cell intracytoplasmic antigen (TIA-1) is a translational silencer that binds to AU-rich elements in TNF-α mRNA to prevent its translation. In mouse macrophages, TIA-1 selectively inhibits TNF-α, but not IL-6, mRNA translation (27, 36). It is possible that high levels of TIA-1 in C2C12 cells may inhibit TNF-α translation and, therefore, limit our ability to detect the protein. Activation of the cells by immune modulators in addition to LPS may be necessary to overcome the negative influence of IκB-ζ and TIA-1 on TNF-α expression.
In this study, we show that three different LPS serotypes stimulate IL-6 synthesis and that other TLR ligands, including Pam and LPS from P. gingivalis (two TLR2 ligands), also stimulate IL-6 synthesis in myoctyes. In general, the TLR2 ligand Pam was slightly more potent than the TLR4 ligand LPS at stimulating an immune response, including the synthesis of IL-6 protein and mRNA. LPS and Pam stimulated IL-6 protein and mRNA expression equivalently over time when a maximally stimulating dose of each ligand was compared. The rapid expression of IL-6 in response to LPS was similar to that observed in other cell types, although C2C12 cells appear to be more sensitive than macrophages to TLR2 ligands (41). TNF-α mRNA expression was also increased in response to TLR2 and TLR4 ligands, as were NOS2 and SOCS3 mRNA. It is likely that all these responses are due to changes in the transcription of the corresponding genes. We found that a transcriptional inhibitor completely blocked LPS- and Pam-induced IL-6, SOCS3, and NOS2 mRNA expression. LPS and Pam also increased IL-6 promoter activity. In total, these findings strongly suggest that TLR ligands increase transcription of the endogenous IL-6 gene in skeletal muscle.
Other TLR ligands were essentially without effect on IL-6 synthesis in C2C12 myocytes. Poly(I:C), a double-stranded RNA mimetic and TLR3 ligand, did not stimulate IL-6 synthesis, although it did increase NOS2 mRNA expression, albeit at a very high dose (25 μg/ml; unpublished observation). This is consistent with our ability to detect TLR3 mRNA in C2C12 cells, but this receptor is not coupled to pathways that stimulate IL-6 synthesis as effectively as TLR2 and TLR4 in this cell type. Similarly, an oligonucleotide containing unmethylated DNA in the context of CpG dinucleotides also failed to increase IL-6 synthesis, suggesting that C2C12 myocytes do not express functional TLR9. Indeed, when we examined C2C12 cells, we found that they expressed TLR1–7 but lacked TLR8 and TLR9, and this pattern was also seen in mouse skeletal muscle. Some TLRs are known to be expressed intracellularly, and the addition of naked DNA or RNA to the cells in the absence of a lipid carrier may have diminished our chances of observing a response to PAMPs composed of nucleic acids.
We have also determined that R837, a potent antiviral molecule that activates TLR7, has a limited ability to stimulate IL-6 synthesis on its own but has a synergistic effect when added in the presence of LPS or Pam. The enhanced expression of IL-6 in response to LPS and R837 is consistent with R837 acting as an immunomodulator that increases interferon-regulated genes and that these genes enhance the response to LPS (5). The above-mentioned finding suggested that C2C12 cells express TLR7, and this was also confirmed by RT-PCR.
IL-6 expression in skeletal muscle and muscle cells is potently attenuated by the synthetic glucocorticoid dexamethasone (14, 15, 25). In this study, we demonstrated that dexamethasone inhibited LPS- and Pam-induced IL-6 synthesis in C2C12 cells. Inhibition of both TLR ligands with dexamethasone is consistent with TLR2 and TLR4 sharing a common downstream signaling pathway. Because NF-κB is often transrepressed by the glucocorticoid receptor (40, 50) and the IL-6 promoter is known to have an NF-κB-responsive element, we tested whether inhibition of upstream regulators of NF-κB signaling could block IL-6 synthesis. We previously showed that the proteasomal inhibitor MG-132 could prevent the LPS-stimulated degradation of IκB, although MG-132 can positively regulate IL-6 expression over time (14).
We have shown that MG-132 inhibits LPS- and Pam-induced IL-6 synthesis in cells exposed to MG-132 for a short period of time. In addition, an inhibitor of IKK2, the kinase that phosphorylates IκB and triggers its subsequent degradation by the proteasome, also inhibited LPS- and Pam-induced IL-6 synthesis. This agrees with the finding that IKK2-null mice fail to make IL-6 in response to IL-1 (46). Thus phosphorylation of downstream substrates of IKK2, including IκB-α and NF-κB p65, may be necessary for complete transcriptional activation of the IL-6 promoter (22). Indeed, when C2C12 myoctyes were transfected with an IL-6 reporter plasmid in which the NF-κB binding site was mutated, LPS-stimulated IL-6 promoter activity was reduced compared with the wild-type promoter. These findings are in agreement with a similar study examining IL-1β-induced IL-6 promoter activity (28).
Additional signaling pathways may also stimulate IL-6 synthesis (11). For example, inhibitors of the p38 and JNK pathway block LPS-induced IL-6 protein and mRNA expression, whereas an MEK-ERK inhibitor is without effect, suggesting that PAMPs exhibit signaling specificity (14). The specificity of these pathways originates from MEK-ERK kinase (MEKK3), a potent inducer of NF-κB and MAPK. IL-6 induction by IL-1 and LPS requires MEKK3, because mouse embryonic fibroblasts that lack MEKK3 fail to make IL-6 in response to these two ligands, although they continue to respond to bacterial DNA (19). In MEKK3-deficient cells and cells that overexpress a dominant-negative form of MEKK3, LPS-stimulated JNK and p38 phosphorylation and NF-κB-mediated activation potential are severely reduced in luciferase assays (19). Saccani et al. (38) demonstrated that activation of the p38 kinase plays an important role in marking a subset of inflammatory genes for subsequent NF-κB recruitment. The IL-6 promoter is marked in this fashion, because LPS-induced phosphorylation of histone H3 occurs on the IL-6 promoter, as evidenced by chromatin immunoprecipitation assays. The ability of p38 and JNK kinase inhibitors to block LPS-stimulated IL-6 synthesis in C2C12 myocytes may therefore relate to these inhibitors keeping the promoter inaccessible to NF-κB.
The physiological role of IL-6 in skeletal muscle is controversial. In cultured myotubes, IL-6 markedly increases the expression of the E3α-II ubiquitin ligase with a concomitant increase in the addition of ubiquitin to cellular proteins. Because E3α is highly expressed in skeletal muscle during cancer cachexia, IL-6 is thought to activate the ubiquitin-proteasome pathway to proteolytically cleave structural and regulatory proteins in skeletal muscle (1, 23). Infusion of IL-6 directly into skeletal muscle in vivo also caused muscle wasting and was associated with a 17% decrease in myofibrillar protein (18). IL-6 also caused muscle wasting when administered chronically for 1 wk (20) or when overexpressed in transgenic mice (47). In at least two cachexia models, IL-6 antibodies blunt the cachectic effect of IL-6 on skeletal muscle (17, 47).
In contrast to the negative effect of IL-6 on muscle protein in cachexia, IL-6 has anti-inflammatory effects when released by skeletal muscle during exercise (34). IL-6 is synthesized rapidly by exercising skeletal muscle, and provision of naturally produced IL-6 (via exercise) or infusion of IL-6 before the administration of endotoxin blunts TNF-α synthesis in humans (43). Exercise also blunts the overexpression of TNF-α mRNA in the mouse soleus, although IL-6 does not appear to be involved in this response, because exercise suppresses TNF-α mRNA, even in IL-6-deficient mice.
IL-6 is hypothesized to regulate glucose uptake and fatty acid oxidation in skeletal muscle, and, indeed, van Hall et al. (48) found that IL-6 stimulates lipolysis and fat oxidation in humans. IL-6 increased fatty acid oxidation in rat skeletal muscle and L6 myotubes (6, 35). Free fatty acids such as palmitate also increase IL-6 secretion in C2C12 cells and decrease the expression of the glucose transporter GLUT-4, consistent with the preferential use of fatty acids as a cellular fuel (21).
Palmitate and PAMPs increase IL-6 synthesis via a proteasome-dependent mechanism. Palmitate induces the degradation of IκB-α, increases the binding of NF-κB to DNA, and increases IL-6 mRNA in human myocytes. The above-mentioned responses are blocked by MG-132, as they are for PAMPs (49). The mechanism by which palmitate activates the NF-κB pathway is unknown, but kinases upstream of IκB-α may be activated, because IκB-α phosphorylation is one of the earliest end points that is increased by palmitate and TLR ligands (49). It is not known whether palmitate itself activates TLRs, but there is a strict requirement for the binding of a tripalmitoylated moiety to TLR2 (29). In addition, saturated fatty acids preferentially enhance NF-κB activity and cyclooxygenase-2 expression in RAW 264.7 cells (26). Future studies are needed to determine whether dominant-negative forms of TLRs or their downstream adapter proteins prevent palmitate-induced signaling in muscle cells.
In conclusion, our data provide clear evidence that mouse skeletal muscle and muscle cells express functional TLR2, TLR4, and TLR7, but not TLR8 and TLR9. TLR ligands potently stimulate IL-6 protein and mRNA expression in myocytes, and it is clear that the response to Pam requires TLR2, whereas the response to LPS requires TLR4. Additional accessory proteins may help mediate the effects of Pam and LPS in skeletal muscle. Because TLR2 is often found to dimerize with other TLRs, it is possible that bacterial lipopeptides may interact, for example, with TLR6. We can conclude that TLR2 is necessary for C2C12 cells to respond to Pam, but we do not know whether it is sufficient. Similarly, LPS stimulates IL-6 expression in skeletal muscle from wild-type mice but not as well in HeJ mice, which have a mutation in TLR4. It is therefore likely that TLR4 plays a major role in transducing the LPS signal but that other LPS-binding proteins may be necessary to stimulate an immune response in muscle.
TLR ligands potently stimulate IL-6 protein and mRNA expression in myocytes, and the response to both ligands is transcriptionally mediated, because it was blocked by DRB. LPS and Pam stimulate the NF-κB signaling pathway, and this pathway activates the IL-6 promoter. An IKK2 inhibitor and a proteasome inhibitor interrupt NF-κB signaling and IL-6 expression. TLR2 mediates the response to tripalmitoylated peptides, because Pam-induced NF-κB and IL-6 promoter activities were abrogated in C2C12 cells overexpressing a dominant-negative form of TLR2, but not TLR4. In contrast, the local expression of TLR4 in skeletal muscle appears to be critical to the ability of LPS to mount an immune response in this tissue, because the IL-6 response to LPS, but not Pam, was markedly attenuated in C3H/HeJ mice, which carry a TLR4 mutation. The potential role of TLRs in mediating the response to endogenous ligands and exercise remains to be determined.
This work was supported in part by National Institute of General Medical Sciences Grant GM-38032.
We thank Danuta Huber, Anne Pruznak, and Dr. David Yeagley for excellent technical assistance.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 the American Physiological Society