If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Palmitate, but Not Unsaturated Fatty Acids, Induces the Expression of Interleukin-6 in Human Myotubes through Proteasome-dependent Activation of Nuclear Factor-κB*
To whom correspondence should be addressed: Dept. of Internal Med., Div. of Endocrinology, Metabolism, and Pathobiochemistry, University of Tübingen, Otfried-Müller-Strasse 10, D-72076 Tübingen, Germany. Tel.: 49-7071-29-87599; Fax: 49-7071-29-5974;
* This work was supported by the German Research Council (KFO 114) and by Grant Schl 239-7 from the German Research Foundation (to E. D. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Circulating interleukin-6 (IL-6), insulin, and free fatty acid (FFA) concentrations are associated with impaired insulin action in obese and type 2 diabetic individuals. However, a causal relationship between elevated plasma FFAs and IL-6 has not been shown. Because skeletal muscle represents a major target of impaired insulin action, we studied whether FFAs may affect IL-6 expression in human myotubes. We demonstrate that specifically saturated FFAs, e.g. palmitate (0.25 mm), induce IL-6 mRNA expression and protein secretion by a proteasome-dependent mechanism that leads to a rapid and chronic activation of nuclear factor-κB. Insulin, high glucose concentrations, or unsaturated FFAs did not activate IL-6 expression. In fact, the unsaturated FFA linoleate inhibited palmitate-induced IL-6 production. Because inhibition of palmitate metabolism by the acyl-CoA synthetase inhibitor triacsin C did not abolish IL-6 expression, it appears that the palmitate molecule per se exerts the observed effects. Furthermore, we show that in human myotubes, IL-6 activates the phosphorylation of signal transducer and activator of transcription 3 in concentrations similar to hepatocytes. However, no inhibitory effect of IL-6 on insulin action, determined as phosphatidylinositol 3-kinase association with insulin receptor substrate-1, Akt phosphorylation, and glycogen synthesis, was detected. We conclude that IL-6 expression may be modulated by the composition of circulating FFA, e.g. by diet, and that skeletal muscle cells could be target cells for IL-6.
). IL-6 is produced by many different cells but in vivo particularly by stimulated monocytes/macrophages, fibroblasts, adipocytes, and vascular endothelial cells (
). However, recent observations suggest that IL-6 production is also modulated by various conditions in apparently healthy individuals. Pedersen and co-workers (
) have demonstrated that exercise can markedly increase plasma IL-6. The marked exercise-induced IL-6 protein expression was shown to occur exclusively within skeletal muscle fibers (
Several studies have shown that adipose tissue produces and secretes IL-6, and that 10–35% of the body's basal circulating IL-6 is derived from adipose tissue (
). In individuals with obesity and overt type 2 diabetes, IL-6 plasma levels are 2–3-fold higher than in control subjects and are associated with reduced insulin action, i.e. insulin resistance (
). The IL-6 production in fat and skeletal muscle tissue suggests hitherto unknown effects of IL-6 besides its role in the inflammatory network. The liver is a likely target of IL-6 produced by skeletal muscle and adipose tissue. IL-6 has been shown to increase blood glucose through elevated hepatic glucose output (
). Chronic exposure to IL-6 causes impairment of early insulin receptor signaling in the livers of mice, resulting in reduced whole-body insulin sensitivity (
). In both human hepatocarcinoma cell line HepG2 and primary hepatocytes, IL-6 inhibits insulin receptor signal transduction and downstream insulin action, specifically glycogen synthesis (
). However, it is unclear whether IL-6 could also activate IL-6 signaling cascades in skeletal muscle cells. It has been argued that the levels of IL-6 receptors expressed by human myotubes are not sufficient to mediate IL-6 signaling or that subsequently the sensitivity of myotubes to respond to IL-6 is markedly low compared with other cells, e.g. hepatocytes (
The aim of the present study was to investigate the effects of glucose, insulin, and circulating free fatty acids on IL-6 gene expression and protein secretion in skeletal muscle cells. These metabolic factors are increased in obese, insulin-resistant individuals, and skeletal muscle is the tissue mainly responsible for peripheral insulin resistance. Furthermore, we studied the expression of IL-6 receptor components gp130 and IL-6R and investigated the sensitivity of human myotubes to IL-6 by examining the concentration-dependent effect of IL-6 on signal transducer and activator of transcription 3 (STAT-3) tyrosine phosphorylation. We demonstrate that specifically saturated FFAs, e.g. palmitate, induce IL-6 mRNA expression and protein secretion by a proteasome-dependent mechanism, leading to inhibitor of κB-α (IκB-α) degradation and nuclear factor-κB (NF-κB) activation. Insulin, high glucose concentrations, or unsaturated FFAs did not activate IL-6 expression. In fact, the unsaturated FFA linoleate inhibited palmitate-induced IL-6 production. Furthermore, we show that in human myotubes, IL-6 activates the Janus kinase-STAT pathway in concentrations similar to hepatocytes; however, no inhibitory effect of IL-6 on insulin action, determined as phosphatidylinositol 3-kinase association with insulin receptor substrate-1 (IRS-1), Akt phosphorylation, and glycogen synthesis was detected.
EXPERIMENTAL PROCEDURES
Materials—Cell culture media and supplements were from Invitrogen. Oligonucleotides were synthesized by Invitrogen. Klenow enzyme, poly(deoxyinosinic-deoxycytidylic acid), reagents for cDNA synthesis, and the Light Cycler system were from Roche (Mannheim, Germany); protein G Sepharose CL-4B was from Amersham Biosciences; phosphatase inhibitors, fatty acids, essentially fatty acid-free bovine serum albumin (BSA), MG 132, triacsin, myriocin, cycloserine, and human recombinant IL-6 were from Sigma; monoclonal anti-IκB-α antibody was from Alexis (Grünberg, Germany); antibodies against phospho-IκB-α serine 32, phospho-STAT-3 Tyr-705, STAT-3, and phospho-Akt serine 473 were from Cell Signaling (Frankfurt, Germany); antibodies against p85 subunit of phosphatidylinositol 3-kinase and IRS-1 were from Upstate Biotechnology (Lake Placid, NY); antibodies against NF-κB p65, p50, and IκB-β were from Santa Cruz Biotechnology (Santa Cruz, CA); the monoclonal anti-IRS-1 antibody used for immunoprecipitation experiments was made by Nanotools (Teningen, Germany); anti-human IL-6 antibody was from R&D Systems (Minneapolis, MN);[α-32P]dATP was from Hartmann (Braunschweig, Germany), d-[U-14C]glucose (250–360 mCi/mmol) was from Perkin Elmer.
Cell Culture—Primary skeletal muscle cells were grown from satellite cells obtained from percutaneous needle biopsies performed on the lateral portion of quadriceps femoris (vastus lateris) muscle as described recently (
). The donors were normal-weight, healthy Caucasian subjects. All experiments were performed on the first pass of subcultured cells that were plated at ∼5 × 104 cells in 60-cm2 dishes in a 1:1 mixture of α-minimum Eagle's medium and Ham's F-12 supplemented with 20% fetal bovine serum, 1% chicken embryo extract, and 0.2% antibiotic antimycotic solution (growth medium) as described (
). When myoblasts reached 80–90% confluence, the cells were fused for 4 days in α-minimum Eagle's medium containing 5.5 mm glucose with 2% fetal bovine serum and 0.2% antibiotic antimycotic solution (fusion medium). On day 5, cells were stimulated in fusion medium containing FFAs as indicated or with 30 mm glucose for 48 h. HepG2 hepatoma cells were cultured in α-minimum Eagle's medium containing nonessential amino acids supplemented with 2 mm glutamine, 1 mm pyruvate, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 10% FCS according to Staiger et al. (
Reverse Transcription-PCR and Real-time Quantitative PCR Analysis—Total RNA was isolated with the RNeasy kit (Qiagen). Reverse transcription of total RNA (1 μg) was performed using avian myeloblastosis virus reverse transcriptase with a first-strand cDNA synthesis kit for reverse transcription-PCR. Aliquots (2 μl) of the reverse transcription reactions were then submitted in duplicate to online quantitative PCR with the Light Cycler system with SYBR green using the FastStart DNA-MasterSYBR Green I. Initial real-time amplifications were examined by agarose gel electrophoresis, followed by ethidium bromide staining to verify that the primer pairs amplified a single product of the predicted size. Subsequent aliquots of the PCR reaction where checked by melting curve analysis as provided by the Light Cycler system. The following primer pairs were used: IL-6: sense, CCAGCTATGAACTCCTTCTC; antisense, GCTTGTTCCTCACATCTCTC; product of 425 bp; gp130: sense, ACAGAACAGCATCCAGTGTC; antisense, AATCTGGCTCCAAGTTGAGG; product of 569 bp; IL-6R: sense, GACAATGCCACTGTTCACTG; antisense, GCTAACTGGCAGGAGAACTT; product of 377 bp. The PCR was performed in a volume of 20 μl: 2 μl of FastStart DNA-MasterSYBR Green I, 3 mm MgCl2, and primers according to a primer concentration of 1 μm. The instrument settings were: denaturing at 95 °C for 10 min; 45× denaturing at 95 °C for 15 s, annealing at 63 °C for 10 s, and elongation for 17s for IL-6; 45× denaturing at 95 °C for 15 s, annealing at 70 °C for 10 s, and elongation for 23 s for gp130; and 45× denaturing at 95 °C for 15 s, annealing at 68 °C for 10 s, and elongation for 15 s for IL-6R. Quantification was performed by online monitoring for identification of the exact time point at which the logarithmic linear phase was distinguishable from the background. Serially diluted samples obtained by PCR with the above-mentioned primers from human myotubes were used as external standards in each run. The cycle numbers of the logarithmic linear phase were plotted against the logarithm of the concentration of the template DNA, and the concentration of cDNA in the different samples was calculated with the Light Cycler software (version 5.32).
Western Blotting—Cells were lysed with 600 μl of lysis buffer/10-cm dish (50 mm Tris, pH 7.6, 150 mm NaCl, and 1% Triton X-100, containing protease and phosphatase inhibitors). Cytosolic extracts of myotubes or immunoprecipitated proteins were separated by sodium dodecyl sulfate polyacrylamide (7.5%) gel electrophoresis. Proteins were transferred to nitrocellulose by semi-dry electroblotting (transfer buffer: 48 mm Tris, 39 mm glycine, 0.0375% sodium dodecyl sulfate, and 20% (v/v) methanol). Then nitrocellulose membranes were blocked with NET buffer (150 mm NaCl, 50 mm Tris/HCl, pH 7.4, 5 mm EDTA, 0.05% Triton X-100, and 0.25% gelatin) and incubated with the first antibody (diluted 1:1000 in NET) overnight at 4 °C. After washing with NET buffer, the membranes were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG for 1 h at room temperature. Visualization of immunocomplexes was performed by enhanced chemiluminescence as described previously (
Enyzme-linked Immunosorbent Assay for IL-6 and Determination of DNA Concentrations—Supernatants of myotubes were collected at the indicated time points, and IL-6 protein concentration was measured by the Quantikine human IL-6 immunoassay (R&D Systems, Minneapolis, MN). DNA in total cell extracts was measured by fluorometry with bisbenzimidazol (
). Synthetic oligonucleotides containing a high-affinity binding site for NF-κB GTTAGTTGAGGGGACTTTCCCAGGC were end-labeled with [α-32P]dATP (3000 Ci/mm) and Klenow enzyme and were incubated with up to 10 μg of nuclear protein in 20 μl of 22 mm Hepes-KOH, pH 7.9, 70 mm KCl, 2.2 mm dithiothreitol, and 10% glycerol on ice for 20 min. Poly(deoxyinosinic-deoxycytidylic acid) (0.05 mg/ml) was added as unspecific competitor. The samples were run on a 5% non-denaturating polyacrylamide gel in a buffer containing 25 mm Tris-HCl, pH 8.0, 190 mm glycine, and 1 mm EDTA. Gels were dried and analyzed by autoradiography.
Glycogen Synthesis—Glycogen synthesis was assayed as described by Krutzfeldt et al. (
) with modifications. Fused myotubes in 6-well plates were pretreated for 90 min with IL-6 (20 ng/ml) in serum-free α-minimum Eagle's medium containing 5.5 mm glucose. Cells were then incubated with 10 or 100 nm insulin for 30 min before addition of d-glucose/d-[U-14C]glucose (final concentration, 10 mm; 0.3 μCi/well). After 60 min at 37 °C, the supernatants were aspirated, and the cells were washed three times with ice-cold PBS and lysed in 300 μl of 30% KOH. Aliquots (10 μl/well) were removed for determination of protein content. The extracts were heated for 30 min at 95 °C, and glycogen (final concentration, 2 mg/ml) was added as a carrier. Glycogen was precipitated with 550 μl of 95% ethanol and collected by centrifugation for 10 min at 10,000 × g. The glycogen pellet was resuspended in 500 μl of water and counted by liquid scintillation counting. Glycogen synthesis was expressed as cpm/well/μg of protein.
Statistical Analysis—Results presented are derived from at least three independent experiments. Means ± S.E. were calculated, and groups of data were compared using Student's t test. Statistical significance was set at p < 0.05.
RESULTS
Activation of IL-6 mRNA Expression by Saturated Fatty Acids in Human Myotubes—We studied whether metabolic parameters, such as insulin or high glucose concentrations, affect the expression of IL-6 in human myotubes. High glucose concentrations (30 mm) showed no effect on IL-6 mRNA expression after a 48-h incubation period, either alone or in combination with 10 or 100 nm insulin, and 100 nm insulin alone reduced IL-6 mRNA expression levels by 30% (Fig. 1A). Because we observed recently that 0.5 mm saturated fatty acids up-regulate glutamine:fructose-6-phosphate aminotransferase expression in human myotubes after 20 h (
), we used this experimental condition to evaluate the effect of different FFAs on IL-6 expression. As shown in Fig. 1B, stimulation of myotubes with 0.5 mm palmitate or stearate for 20 h resulted in a significant increase in IL-6 mRNA expression. The effect of unsaturated FFA palmitoleate, oleate, and linoleate was similar to the effect observed in control cells treated with BSA (fatty acid free) alone. Incubation with FFAs had no effect on cell number or DNA content (data not shown). The effect of palmitate was studied in more detail because: (1) the plasma concentration of FFA is <0.7 mm in the postabsorptive state and might rise to >1 m m after a fatty meal or during fasting; and (2) palmitate is the most abundant saturated FFA in plasma (palmitate, 20–35%; stearate, ∼10%). When we examined the effect of increasing concentrations of palmitate on IL-6 mRNA expression, we found no effect with 0.125 mm palmitate but a maximal stimulation with 0.25 mm palmitate (data not shown). Therefore, we used this concentration for additional studies. Because TNF-α is a known strong inducer of IL-6, we compared the effect of palmitate and TNF-α on IL-6 expression in human myotubes. As shown in Fig. 1C, 0.25 mm palmitate induced IL-6 mRNA to a similar extent as TNF-α. We also observed a small effect of FFA-free BSA, which was used as a carrier for FFA (1.4-fold). In additional studies, we corrected all FFA effects to the FFA-free BSA control.
Fig. 1Regulation of IL-6 mRNA expression in human myotubes. Human myotubes were stimulated with 30 mm glucose (HG) or insulin for 48 h (A), with different fatty acids for 20 h as indicated (B), or with 5 nm TNF-α for 4 and 8 h (C). IL-6 mRNA expression was measured by real-time PCR. Results are expressed as mean ± S.E. of at least four separate experiments; rel., relative. A, mRNA expression of untreated cells was set as 1. #, p < 0.05 versus control. B, mRNA expression of BSA cells was set as 1. *, p < 0.05 versus BSA. C, mRNA expression of TNF-α control cells was set as 1. *, p < 0.05 versus control cells (con).
Time-dependent Increase in Cellular IL-6 mRNA Levels and IL-6 Protein Production in Palmitate-treated Myotubes—Next, we studied the time response of palmitate-induced IL-6 mRNA and protein expression. After 1 h of incubation with 0.25 mm palmitate, an increased IL-6 mRNA level was detected that was not further enhanced during the next 7 h of palmitate stimulation (1.5–2.0-fold; Fig. 2A). The increase in IL-6 mRNA levels peaked after 24 h (5-fold; Fig. 2A) and remained elevated after 48 h. To assess whether the palmitate-induced increase in IL-6 mRNA expression was translated into an enhanced IL-6 protein production and subsequent secretion, IL-6 protein content in cellular supernatants was determined by enzyme-linked immunosorbent assay. As shown in Fig. 2B, palmitate induced a significant increase in IL-6 protein production as early as after 1 h, which increased with time. After 24 h, 3.0 ± 0.3 ng/μg of DNA were produced and secreted into the supernatant, and after 48 h, 5.7 ± 1.4 ng/μg of DNA were accumulated. The data indicate that the rapid increase in IL-6 mRNA is readily translated into IL-6 protein, and that the increases in IL-6 protein production remain high after 24 and 48 h, as suggested by the high cellular IL-6 mRNA levels at these time points (Fig. 2A).
Fig. 2Time-course of palmitate-induced IL-6 mRNA and protein expression. Human myotubes were stimulated with 0.25 mm palmitate for the indicated time points. Results are expressed as mean ± S.E. of at least four separate experiments; rel., relative. A, palmitate-induced expression of IL-6 mRNA was measured by real-time PCR. IL-6 mRNA levels of cells treated with BSA for identical times were set as 1. *, p < 0.05 versus BSA. B, induction of IL-6 protein accumulation by palmitate. IL-6 protein concentrations in the supernatant of palmitate-treated cells were measured by enzyme-linked immunosorbent assay and corrected to the DNA content and expressed as ng/μg of DNA. *, p < 0.05 versus BSA.
Induction by Palmitate Is Proteasome-dependent and Does Not Require Metabolism to Acyl-CoA—Because the transcription factor NF-κB has been shown to be involved in the activation of IL-6 gene expression in various tissues and cells (
), we tested the possibility of NF-κB mediating the palmitate-induced IL-6 expression in human myotubes. The activation of NF-κB requires proteasome activity (
). To elucidate whether proteasome activity or oxidative stress is induced by palmitate and mediates the up-regulation of IL-6 mRNA expression, inhibition experiments were performed. Activation of myotubes with 0.25 mm palmitate in the presence of 0.5 mm antioxidant α-lipoic acid did not prevent IL-6 mRNA induction (Fig. 3A). Similarly, preincubation with 0.1 mm α-lipoic acid for 48 h, which was shown to prevent oxidative stress-mediated activation of NF-κB (
), failed to inhibit IL-6 mRNA expression (Fig. 3A). However, the presence of the proteasome inhibitor MG 132 (10 μm) completely blocked palmitate-induced IL-6 expression (Fig. 3, A and B). Thus, activation of NF-κB by a proteasome-dependent pathway could be involved in IL-6 up-regulation by palmitate.
Fig. 3Effect of inhibitors of fatty acid metabolism and of proteasome inhibition on palmitate-induced IL-6 expression. Human myotubes were stimulated with 0.25 mm palmitate for 20 h with or without inhibitors. IL-6 mRNA expression was measured by real-time PCR; IL-6 protein concentration in the supernatant was measured by enzyme-linked immunosorbent assay. Results are expressed as mean ± S.E. of at least four separate experiments; rel., relative. IL-6 mRNA expression and IL-6 protein concentration were shown as fold-induction of values of BSA-treated cells. A, myotubes were preincubated for 30 min with 10 μm proteasome inhibitor MG 132 or 0.5 mm α-lipoic acid and then stimulated with palmitate. Preincubation with 0.1 mm α-lipoic acid was performed for 48 h before palmitate stimulation. *, p < 0.05 versus BSA; #, p < 0.05 versus palmitate without inhibitor. B, myotubes were preincubated for 30 min with increasing concentrations of MG 132 before stimulation withpalmitate for 20 h. *, p < 0.05 versus BSA; #, p < 0.05 versus without MG 132. C, myotubes were preincubated with 5 μm triacsin C for 30 min before treatment with palmitate and/or linoleate for 20 h. *, p < 0.05 versus BSA; #, p < 0.05 versus palmitate; §, p < 0.05 palmitate with triacsin C versus palmitate. D, myotubes were preincubated with 10 μm myriocin or 1 mm cycloserine for 30 min before treatment with palmitate for 20 h. *, p < 0.05 versus BSA. E, IL-6 protein concentration in the supernatant of myotubes treated for 20 h with FFA with or without inhibitors as indicated. *, p < 0.05 versus BSA; #, p < 0.05 versus palmitate.
Palmitate has been linked to the activation of several signaling pathways, e.g. by its metabolism to diacylyglycerol and subsequent activation of protein kinase C (
). All of these events require activation of palmitate to palmitoyl-CoA by acyl-CoA synthetase. Therefore, we studied whether this conversion is required for the effect on IL-6 expression. In the presence of triacsin C, an inhibitor of acyl-CoA synthetase (
), no reduction of IL-6 mRNA expression was detected (Fig. 3C); in contrast, the effect of palmitate on IL-6 mRNA expression was almost doubled (8.3 ± 0.9 versus 4.7 ± 0.3). Moreover, we observed that in the presence of equal amounts of linoleate and palmitate, the palmitate-induced up-regulation of IL-6 was prevented (Fig. 3C). Furthermore, our observation that myriocin and cycloserine, both inhibitors of ceramide biosynthesis, had no effect on palmitate-induced IL-6 expression further supports the view that activation of palmitate to palmitoyl-CoA is not necessary for the observed palmitate effect (Fig. 3D).
The inhibitory action of MG 132 and linoleate was also found on palmitate-induced IL-6 protein production (Fig. 3E). In accordance with the data obtained on the mRNA level, α-lipoic acid had no effect on enhanced IL-6 protein production (Fig. 3E). In the presence of the acyl-CoA synthetase inhibitor triacsin C, IL-6 protein induction was not prevented (Fig. 3E). Thus, unmetabolized, nonesterified palmitate is the responsible signaling molecule leading to IL-6 mRNA and protein expression.
Palmitate Activates NF-κB—Because inhibition of proteasome activity with MG 132 blocked IL-6 mRNA induction completely and because activation of the transcription factor NF-κB requires proteasome-dependent degradation of IκB-α, we investigated the hypothesis that stimulation of myotubes with palmitate resulted in activation of NF-κB. Using electrophoretic mobility shift assays, we found a clear induction of NF-κB DNA binding activity in nuclear extracts obtained from palmitate-treated cells (Fig. 4A, lane 3) compared with almost no detectable DNA-binding activity to the NF-κB consensus sequence in control extracts or nuclear extracts obtained from high glucose-stimulated myotubes (Fig. 4A, lanes 1 and 2). A weak NF-κB DNA binding was observed in linoleate-treated cells (Fig. 4A, lane 4); however, this was in the same range as the NF-κB DNA-binding activity found in BSA-treated cells (Fig. 4B, lane 2). In the presence of the proteasome inhibitor MG 132, the palmitate-stimulated NF-κB activation was prevented and was comparable with BSA treatment alone (Fig. 4B, lane 4 versus lane 2). The specificity of the NF-κB DNA binding was demonstrated by competition with 10- and 30-fold excess of unlabeled oligonucleotide (Fig. 4B, lanes 5 and 6). The activation of NF-κB was detected after 30 min (data not shown) and 1 h of palmitate treatment, and the DNA-binding activity appeared to be even higher after 4 and 20 h of palmitate stimulation (Fig. 4C, lanes 2–4). To identify NF-κB subunits contributing to DNA-binding activities, supershift experiments were performed. We observed p65 to be the major component of the NF-κB complex (Fig. 4C, lanes 5–7), and anti-p50 antibodies also attenuated the NF-κB binding, which was clearly detectable in nuclear extracts obtained after 20 h of palmitate treatment (Fig. 4C, lanes 8–10).
Fig. 4Palmitate-induced activation of NF-κB. Nuclear proteins were incubated with 40,000 cpm of the 32P-labeled oligonucleotide containing a consensus binding site for NF-κB. Specific binding is marked by the arrow. Free DNA probe is not shown. A, electrophoretic mobility shift assays were performed with nuclear extracts of myotubes either stimulated with 0.5 mm FFAs or with 30 mm glucose for 48 h. Unspecific binding is marked by the bracket. con, control cells. B, electrophoretic mobility shift assays were performed with nuclear extracts of myotubes stimulated with 0.5 mm palmitate for 20 h without (lane 3) or after preincubation with MG 132 (lane 4). Lanes 5 and 6 show competition experiments using 10- or 30-fold molar excess of unlabeled NF-κB consensus binding site and nuclear proteins of myotubes stimulated with 0.5 mm palmitate for 20 h. Unspecific binding is marked by the bracket. con, control cells. C, nuclear extracts of myotubes stimulated with 0.5 mm palmitate for 1, 4, or 20 h were preincubated with 2 μg of anti-NF-κB antibodies specific for p65 (lanes 5–7) or p50 (lanes 8–10). NF-κB binding is marked by the lower bracket; supershifted complexes are marked by the upper bracket. Unspecific binding is not shown. D and E, Western blots with cytosolic extracts of myotubes stimulated with 0.5 mm palmitate for the indicated time points. D, equal amounts of cytosolic extracts were separated on a 7.5% SDS-PAGE, and IκB-α and IκB-β were identified with a specific antibody. Columns show the densitometric quantification of the IκB-α protein. The mean value of BSA-treated cells is defined as 1. Each value is expressed as mean ± S.E. of n = 4; rel., relative. #, p < 0.01 versus BSA. E, equal amounts of cytosolic extracts were separated on a 7.5% SDS-PAGE, and phosphorylation of IκB-α on serine 32 was identified with a specific antibody. One representative immunoblot is shown with the reprobe of the same blot using monoclonal anti-IκB-α antibodies.
The key to NF-κB regulation is the IκB proteins, which retain NF-κB in the cytoplasm. Phosphorylation of IκB-α by IκB kinases triggers its polyubiquitinylation and degradation, thereby releasing NF-κB, which translocates to the nucleus (
). Therefore, we investigated the molecular mechanism of palmitate-induced NF-κB activation by studying the amount of IκB-α and its serine 32 phosphorylation in palmitate-treated myotubes. Immunoblots using anti-IκB-α antibodies revealed a reduction of IκB-α protein as early as after 1 h of palmitate stimulation, and the level of IκB-α protein remained reduced after 4 and 20 h of palmitate treatment, well in line with the kinetics of activated DNA-binding activity of NF-κB (Fig. 4D). In contrast, IκB-β protein levels were unchanged (Fig. 4D). In immunoblots with anti-phospho-serine 32 antibodies, a large increase in phosphorylation of IκB-α on serine 32 was detected after 30 min of palmitate treatment, whereas later the phosphorylation decreased to basal levels (Fig. 4E). Thus, a rapid, palmitate-induced phosphorylation of IκB-α is the signal for the proteasome-dependent degradation of this protein, thereby leading to translocation to the nucleus and binding to consensus sequences and subsequently to IL-6 mRNA expression.
IL-6 Signaling in Human Myotubes—Because we could clearly demonstrate that human myotubes express and secrete relevant amounts of IL-6, we studied the putative auto/paracrine effect of IL-6. IL-6 signaling requires the expression of receptors gp130 and the IL-6 receptor (IL-6R), acting as a heterodimer in the binding and signal transduction of IL-6 (
). Using real-time PCR, we found that both receptors were expressed in human myotubes, and that the expression was not affected by palmitate or linoleate (Fig. 5, A and B). Compared with expression levels in HepG2 cells, the content of gp130 mRNA in myotubes was 38.8 ± 8% of that found in HepG2 cells, whereas IL-6R mRNA levels were 6.3 ± 4%. Because an early event after binding of IL-6 to the gp130/IL-6R dimer is activation of Janus kinases and subsequent rapid phosphorylation of STAT (
), we studied the effect of increasing concentrations of human recombinant IL-6 on STAT-3 phosphorylation in human myotubes. Immunoblots with anti-phospho-STAT-3 antibodies clearly demonstrated IL-6-dependent phosphorylation of STAT-3, with a comparable effect of stimulation with 3–60 ng/ml of IL-6 for 20 min (Fig. 5C). The concentrations used in this experiment are similar to the amount of IL-6 protein secreted by the myotubes (up to 30 ng/ml of IL-6 were found in palmitate-treated cells). Thus, IL-6 signaling pathways are conserved in human myotubes, and the cells are sensitive to IL-6-mediated activation.
Fig. 5Action of IL-6 on myotubes. mRNA expression of gp130 (A) and IL-6R (B) in human myotubes. Cells were stimulated with fatty acids for 20 h as indicated, and mRNA expression of gp130 (A) and IL-6R (B) was measured by real-time PCR. Results are expressed as mean ± S.E. of at least four separate experiments; rel., relative. mRNA expression in BSA-treated cells was set as 1. C, human myotubes were stimulated with human recombinant IL-6 for 20 min with the indicated concentrations. Equal amounts of cytosolic extracts were separated on a 7.5% SDS-PAGE, and STAT-3 protein and phosphorylation of STAT-3 on tyrosine 705 were identified with specific antibodies. A representative immunoblot with the phospho-site-specific antibody and reprobe of the same blot with polyclonal anti-STAT-3 antibody is shown in the upper part of the figure. con, control cells. Columns show densitometric quantification of the phosphorylated STAT-3 related to the STAT-3 protein. The mean value of control cells is defined as 1. Each value is expressed as mean ± S.E. of n = 4. *, p < 0.05 versus control. D, supernatant of myotubes stimulated with 0.25 mm palmitate for 20 h (palm-preconditioned medium) or control supernatant (control medium) were preincubated with neutralizing anti-IL-6 antibodies in a final concentration of 5 μg/ml for 10 min (+ anti-IL-6). Then, myotubes were treated with the different supernatants as indicated. Equal amounts of cytosolic extracts were separated on a 7.5% SDS-PAGE, and STAT-3 protein and phosphorylation of STAT-3 on tyrosine 705 were identified with specific antibodies. A representative immunoblot with the phospho-site-specific antibody and reprobe of the same blot with polyclonal anti-STAT-3 antibody is shown in the upper part of the figure. Columns show densitometric quantification of the phosphorylated STAT-3 related to the STAT-3 protein. The mean value of control cells is defined as 1. Each value is expressed as mean ± S.E. of n = 4. #, p < 0.05 anti-IL-6 versus palm-preconditioned medium.
Next, we studied whether the palmitate-induced expression of IL-6 could account for IL-6-mediated signals in the myotubes. Supernatants of palmitate-treated myotubes were used to induce IL-6 signaling in these cells. Incubation of the myotubes with these supernatants for 20 min activated phosphorylation of STAT-3, whereas supernatant of untreated control cells had no effect (Fig. 5D). Moreover, this activation is IL-6-dependent, because addition of neutralizing IL-6 antibodies clearly reduced STAT-3 phosphorylation (Fig. 5D). Thus, palmitate-induced IL-6 production mediates activation of IL-6 signaling pathways in myotubes.
Effect of IL-6 on Insulin Signaling in Myotubes—Increases in circulating IL-6 plasma levels have been associated with insulin resistance in humans (
), and IL-6 has been demonstrated to reduce insulin-stimulated signaling in human hepatocytes, e.g. by decreased tyrosine phosphorylation of IRS-1, by decreased association of the p85 subunit of phosphatidylinositol 3-kinase with IRS-1, by subsequently impaired phosphorylation of Akt on serine 473, and by reduced glycogen synthesis (
). IRS-1-associated phosphatidylinositol 3-kinase is an early and crucial event in the insulin signaling cascade in skeletal muscle. Therefore, we investigated, in human myotubes, the effect of IL-6 on insulin-stimulated p85/IRS-1 complex formation. Insulin concentrations of 1 and 10 nm are sufficient to induce insulin receptor tyrosine kinase activity in human myotubes (
). Stimulation of myotubes with 1 and 10 nm insulin activated the association of p85 with IRS-1, as shown by co-immunoprecipitation and immunoblotting of p85 and IRS-1 (Fig. 6A). Preincubation with 20 ng/ml of IL-6 for 90 min, which had been shown to exhibit a maximum reduction of insulin-stimulated p85/IRS-1 complex formation in hepatocytes (
), had no inhibitory effect on p85/IRS-1 association in myotubes. Accordingly, insulin-stimulated phosphorylation of serine 473 of Akt was not attenuated in the myotubes after IL-6 preincubation, whereas IL-6 alone induced a weak phosphorylation of Akt (Fig. 6B). Similarly, insulin-induced glycogen synthesis was not reduced by IL-6 (2.23 ± 0.17-fold after stimulation with 100 nm insulin and preincubation with IL-6 compared with 2.31 ± 0.13-fold with insulin alone; Fig. 6C). When myotubes were treated with 10 nm insulin, we observed a stimulatory effect of IL-6 on glycogen synthesis (2.28 ± 0.18 and 1.49 ± 0.06, respectively; Fig. 6C). Because both cell types, hepatocytes and myotubes, were sensitive to similar insulin concentrations and the IL-6 concentrations necessary for STAT-3 phosphorylation were comparable, we conclude that, in human myotubes, IL-6 is not a mediator of cellular insulin resistance.
Fig. 6Effect of IL-6 pretreatment on insulin signaling pathways. Human myotubes were stimulated with insulin as indicated after preincubation with IL-6 (20 ng/ml) for 90 min. A, insulin-stimulated association of the p85 subunit of phosphatidylinositol 3-kinase with IRS-1 was detected by coprecipitation using monoclonal anti-IRS-1 antibodies. Representative Western blots are shown in the upper part of the figure. Columns show densitometric quantification of coprecipitated p85 related to IRS-1 protein. The mean value of untreated cells is defined as 1. Each value is expressed as mean ± S.E. of n = 7; ns, nonsignificant. B, insulin-stimulated phosphorylation of Akt on serine 473. Equal amounts of cytosolic extracts were separated on a 7.5% SDS-PAGE, and Akt protein and phosphorylation of Akt on serine 473 were identified with specific antibodies. A representative immunoblot with the phospho-site-specific antibody and reprobe of the same blot is shown in the upper part of the figure. Columns show densitometric quantification of the phosphorylated Akt related to Akt protein. The mean value of untreated cells is defined as 1. Each value is expressed as mean ± S.E. of n = 4; ns, nonsignificant. C, insulin-stimulated glycogen synthesis. Glycogen synthesis (cpm/μg protein) of untreated cells is defined as 1. Each value is expressed as mean ± S.E. of n = 6; ns, nonsignificant.
), have been shown to inhibit insulin action in skeletal muscle. In line with these reports, pretreatment of our myotubes with 0.5 mm palmitate for 20 h attenuated the glycogen synthesis stimulated with 100 nm insulin to 80% (1.68 ± 0.08 compared with 2.09 ± 0.14, respectively; data not shown). However, from our results it seems unlikely that IL-6 is the mediator of fatty acid-induced insulin resistance in skeletal muscle cells.
DISCUSSION
In the present study, we demonstrate that saturated FFAs, in contrast to unsaturated FFAs, high glucose concentrations, or insulin, activate gene expression and protein production of IL-6 in human myotubes. The induction occurs rapidly within 1 h and is maintained during the next 48 h. We also found a small but significant effect of FFA-free BSA alone on IL-6 gene expression. We suggest that contamination with low endotoxin amounts in the BSA preparation accounted for this induction of IL-6 because NF-κB activity, known to be readily induced by endotoxins (
), is slightly enhanced in BSA-treated myotubes. The novel finding that saturated FFAs induce IL-6 expression is striking because the current view is that IL-6 expression is regulated by mediators of the pro- and anti-inflammatory network. In our study, we focused on palmitate, because a maximum effect was achieved with 0.25 mm palmitate, a plasma concentration that is also present in the postprandial state and during fasting. Although the presence of unsaturated fatty acids alone did not enhance IL-6 mRNA levels, coincubation of myotubes with equimolar concentrations of linoleate and palmitate completely prevented the stimulatory action of the saturated FFAs. A protective effect of unsaturated FFAs on palmitate-mediated lipotoxicity has also been described recently for pancreatic β-cell apoptosis (
). Unsaturated FFAs promote channeling of palmitate into triglyceride pools away from pathways leading to apoptosis. Thus, stimulation of palmitate metabolism to triglycerides would also be a possible explanation for the inhibitory effect of linoleate in our experiments.
Further support for this hypothesis is provided by our results demonstrating that the free, non-metabolized palmitate is responsible for the activation of IL-6 gene transcription and not palmitoyl-CoA or other converted metabolites. The bacterial compound triacsin C, which had been shown to inhibit acyl-CoA synthetase in micromolar concentrations (
), did not prevent IL-6 induction by palmitate but, in fact, enhanced IL-6 mRNA expression significantly compared with the effect of palmitate alone. These data imply that accumulation of free palmitate because of blockade of its activation to palmitoyl-CoA results in higher IL-6 expression levels, whereas triggering the metabolism of palmitate in the presence of the unsaturated fatty acid linoleate reduces the concentration of free palmitate to ineffective levels.
We also gained insight into the molecular mechanism of palmitate-induced gene transcription. Palmitate triggers the rapid degradation of IκB-α within 1 h by stimulation of its phosphorylation on serine 32 after 30 min. This led to translocation of NF-κB into the nucleus and activation of NF-κB DNA binding within 1 h. Because IL-6 gene expression could be up-regulated by NF-κB activation that was mediated by direct binding to a binding site in the 5′-flanking region of the IL-6 gene (
), early induction of IL-6 expression was found after 1 h of palmitate treatment. In addition, we observed a chronic activation of NF-κB, because induction of DNA binding is also visible after 4 and 20 h of incubation with palmitate. This is in line with the chronic activation of IL-6 gene expression and could at least partially explain the finding that the cytoplasmic amount of IκB-α is also reduced after 4 and 20 h. The exact molecular mechanism of this sustained down-regulation of IκB-α remains to be determined, because phosphorylation of IκB-α was only observed after 30 min. Proteasome activity is clearly necessary for the chronic activation of NF-κB and subsequent IL-6 mRNA and protein expression found after 20 h of palmitate treatment, because the presence of proteasome inhibitor MG 132 completely blocked all of these events. A lipid-induced degradation of IκB-α was also described in healthy volunteers (
). In these studies, the activation of protein kinase C isoforms, e.g. PKC θ, by increased diacylglycerol levels was implicated as the responsible mechanism for NF-κB activation. Our finding that palmitate metabolism is not necessary to achieve the up-regulation of gene expression does not support the possibility that diacylglycerol is the signal transducer for palmitate-enhanced IL-6 production. The mechanism of how free palmitate triggers IκB-α phosphorylation and proteasome-mediated protein degradation remains open. Our finding that NF-κB subunit p50 is also part of the DNA binding complex suggests the involvement of other pathways leading to chronic NF-κB activation (
To address the question of whether IL-6 produced by skeletal muscle cells has auto/paracrine function, we investigated whether IL-6 signaling is transmitted in human myotubes. We clearly demonstrated that myotubes are sensitive to IL-6 in a similar range as hepatocytes or HepG2 hepatoma cells (
), as shown by dose-dependent STAT-3 phosphorylation with 3 ng/ml of IL-6, resulting in a near maximum STAT-3 phosphorylation, although the expression level of IL-6R was low compared with HepG2 cells. This is, to our knowledge, the first report on the IL-6 signaling capacities of human skeletal muscle cells. Thus, myotubes may be added to the list of IL-6 target cells.
Following this line, we investigated the effect of IL-6 on insulin signaling. IL-6 has been shown to impair insulin signaling in hepatocytes and HepG2 cells (
). However, using similar experimental conditions as in the studies performed with HepG2 cells and hepatocytes, we observed no inhibitory effect of IL-6 on insulin-stimulated phosphatidylinositol 3-kinase p85 subunit/IRS-1 complex formation and on insulin-stimulated serine 473 phosphorylation of Akt. Activation of phosphatidylinositol 3-kinase and Akt are key events in insulin signaling in skeletal muscle that are necessary for insulin stimulation of both glucose transport and glycogen synthesis. Accordingly, insulin-induced glycogen synthesis was not reduced by IL-6 in the myotubes. Thus, we conclude that IL-6 is unlikely to be a mediator of cellular insulin resistance in skeletal muscle cells. A recent study shows that a 5-day infusion of IL-6 failed to reduce insulin receptor signaling in the skeletal muscle of mice (
Our results promote interest in the possible (patho)physiological role of elevated muscle IL-6 production after exposure to high saturated fatty acid levels. It has been hypothesized that one function of muscle-derived IL-6 is to down-regulate TNF-α, thus acting as an anti-inflammatory agent and as an enhancer of insulin sensitivity by counter-regulating TNF-α activity (
). During prolonged exercise when skeletal muscle is confronted with increasing concentrations of fatty acids originating from other tissues, energy utilization should be favored in the muscle, while other organs act as energy suppliers. Thus, we hypothesize that IL-6 released from the muscle triggers energy supply through the lipolytic properties of IL-6 on adipose tissue and the glucoregulatory effect of IL-6 leading to hepatic glucose production.
In conclusion, our data provide clear evidence that skeletal muscle cells produce considerable amounts of IL-6 in response to increasing saturated FFA concentrations. Specifically, a positive association between palmitate and plasma IL-6 levels, which was independent of adiposity, was found in 55 subjects of the Tuebingen Family Study for obesity and type 2 diabetes.
). Thus, there is evidence that up-regulated IL-6 gene expression by saturated fatty acids could also be present in humans. The role of this potential increased IL-6 production of the skeletal muscle remains to be determined.
Acknowledgments
We thank D. Burt and N. Stefan for critical discussion.
66523-1&id=gr1.jpg){kind=link}
66523-1&id=gr2.jpg){kind=link}
66523-1&id=gr3.jpg){kind=link}
66523-1&id=gr4.jpg){kind=link}
66523-1&id=gr5.jpg){kind=link}
66523-1&id=gr6.jpg){kind=link}