Direct Cross-talk of Interleukin-6 and Insulin Signal Transduction via Insulin Receptor Substrate-1 in Skeletal Muscle Cells*

The exercise-induced interleukin (IL)-6 production and secretion within skeletal muscle fibers has raised the question of a putative tissue-specific function of IL-6 in the energy metabolism of the muscle during and after the exercise. In the present study, we followed the hypothesis that IL-6 signaling may directly interact with insulin receptor substrate (IRS)-1, a keystone in the insulin signaling cascade. We showed that IL-6 induces a rapid recruitment of IRS-1 to the IL-6 receptor complex in cultured skeletal muscle cells. Moreover, IL-6 induced a rapid and transient phosphorylation of Ser-318 of IRS-1 in muscle cells and in muscle tissue, but not in the liver of IL-6-treated mice, probably via the IL-6-induced co-recruitment of protein kinase C-δ. This Ser-318 phosphorylation improved insulin-stimulated Akt phosphorylation and glucose uptake in myotubes since transfection with an IRS-1/Glu-318 mutant simulating a permanent phospho-Ser-318 modification increased Akt phosphorylation and glucose uptake. Noteworthily, two inhibitory mechanisms of IL-6 on insulin action, phosphorylation of the inhibitory Ser-307 residue of IRS-1 and induction of SOCS-3 expression, were only found in liver but not in muscle of IL-6-treated mice. Thus, the data provided evidence for a possible molecular mechanism of the physiological metabolic effects of IL-6 in skeletal muscle, thereby exerting short term beneficial effects on insulin action.

Over the last two decades, the pleiotropic cytokine interleukin-6 (IL-6) 2 has been looked upon as a major component in the inflammatory network and the acute immune response (1). Since 1997, when the adipose tissue was recognized as a relevant IL-6 producing organ, accounting for 10 -35% of circulating IL-6 plasma levels in humans (2), this view was renewed and broadened. It has been established in numerous clinical trials and association studies that IL-6 plasma concentrations increased with weight gain and are associated with the development of insulin resistance (3)(4)(5)(6). Moreover, some reports provide evidence for elevated IL-6 levels as a risk factor for the manifestation of type 2 diabetes (3,4). Well in accordance with these clinical data, chronic IL-6 treatment of mice reduces insulin signal transduction in the liver and produces a mild state of insulin resistance in these animals (7,8). Thus, it is widely accepted that IL-6, similar to tumor necrosis factor-␣, induces insulin resistance in the liver and adipose tissue (9).
However, there is overwhelming evidence that besides immune competent cells and adipose tissue, the skeletal muscle is a further important source of IL-6 (10,11). Noteworthily, the IL-6 production in the muscle fibers is markedly increased by exercise, reaching interstitial IL-6 concentrations Ͼ1 ng/ml (i.e. 100-fold above physiological plasma concentrations) and, depending on the duration and intensity of the exercise, leading to elevated IL-6 plasma concentrations (12,13). Since exercise is one of the most important therapeutic interventions to cope with weight gain and insulin resistance (14), it is obvious to suggest that exercise-induced IL-6 expression may exert positive effects on insulin signal transduction in the skeletal muscle. In fact, moderate exercise has been shown to improve insulin sensitivity, glucose uptake, and glycogen synthesis in muscle (15)(16)(17). An increasing number of reports suggest that IL-6 could act as a myocyte-derived "exercise factor" that improves the energy supply of the muscle by increasing the output of glucose and fatty acids from the energy-storing organs liver and fat and by enhancing glucose uptake, fatty acid oxidation, and glycogen synthesis in muscle (18 -21). Apparently, this argues for some muscle-specific insulin-sensitizing effects of IL-6, but the underlying molecular mechanism is unclear.
IRS-1 is a key player in insulin signaling in muscle, and its cross-talk with metabolic and mitogenic pathways is modulated in a very subtle and complex manner using the phosphorylation pattern of the Ser/Thrphosphorylation sites of IRS-1 as a regulatory mechanism for the interaction with its important downstream mediators, e.g. phosphatidylinositol-3 kinase (PI3K) (22)(23)(24)(25)(26)(27)(28). Since IL-6 regulates via the signal transducer and activator of transcription (STAT)-3-Janus kinase (JAK) pathway and the mitogen-activated protein kinase (MAPK) pathway several Ser/Thr-kinases shown to phosphorylate IRS-1 (29), we hypothesize that IL-6 could cross-talk with the insulin signaling cascade by phosphorylation of IRS-1. We showed in the present study that IL-6 induces the Ser-318 phosphorylation of IRS-1 in various muscle cell lines and in muscle tissue of mice but not in the liver, that IRS-1 coimmunoprecipitates with the IL-6 receptor (IL-6R) and protein kinase C (PKC)-␦ in an IL-6-dependent manner, and that Ser-318 phosphorylation may improve insulin action.
In Vivo IL-6 Treatment of Mice-Ten-week-old male C57Bl/6 mice were obtained from The Jackson Laboratory and studied after 2 weeks of acclimatization. They were maintained on a normal light/dark cycle and kept on a regular diet. For short term IL-6 effects, mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg of body weight) and xylazine (10 mg/kg of body weight), and a total of 2.5 g/kg of mouse recombinant IL-6 or 5 IU of insulin was injected into the inferior vena cava. Controls received a comparable amount of diluent. Tissues (liver, muscle) were removed at the indicated time points, and protein extracts were prepared as described (33). All procedures were approved by the local Animal Care and Use Committee.
Cell Culture and Transfection-Human primary skeletal muscle cells were grown from satellite cells obtained from percutaneous needle biopsies performed on the lateral portion of the quadriceps femoris (vastus lateris) muscle as recently described (34). When myoblasts reached 80 -90% confluence, the cells were fused for 5 days in ␣-MEM containing 5.5 mM glucose with 2% FCS and 2% antibiotic antimycotic solution (fusion medium). Stimulation of the myotubes with IL-6 was performed in fusion medium. C2C12 myoblasts were cultured in Dulbecco's modified Eagle's medium containing 25 mM glucose, 10% FCS, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. 1.0 ϫ 10 5 cells/well of a 6-well plate were transfected using the Ca 3 (PO 4 ) 2 -DNA-co-precipitation-method. Cells were lysed with 150 l of lysis buffer/well (50 mM Tris, pH 7.6, 150 mM NaCl, 1% Triton X-100, containing protease and phosphatase inhibitors). Total protein (400 g) was used for immunoprecipitation. L6 myoblasts were cultured in ␣-MEM containing 5.5 mM glucose, 10% FCS, 100 units/ml penicillin, and 100 g/ml streptomycin as described (35). For differentiation, 0.5 ϫ 10 5 cells/well of a 12-well plate or 2.0 ϫ 10 5 /well of a 6-well plate were seeded in differentiation medium (␣-MEM containing 5.5 mM glucose, 2% FCS, 100 units/ml penicillin and, 100 g/ml streptomycin). Myotubes were used 7-8 days after cell seeding. To obtain stable expression of IRS-1 WT, Ala-318, and Glu-318, L6 Glut4Myc myoblasts were transfected with Effectene (Qiagen, Hilden, Germany) according to the instructions of the supplier. 2.0 ϫ 10 5 cells/well in a 6-well plate received 0.8 g of IRS-1 expression vector and 0.2 g of pSVneo encoding neomycin resistance. Transfected cells were selected for their resistance to the antibiotic G418. G418-resistant clones were screened for IRS-1 expression and Ser-318 phosphorylation by Western blotting.
Reverse Transcription-PCR and Real-time Quantitative PCR Analysis-Total RNA of muscle and liver was isolated with PeqGOLD TriFast (Peqlab, Erlangen, Germany). Reverse transcription of total RNA (1 g) was performed using avian myeloblastosis virus reverse transcriptase with the first-strand cDNA synthesis kit for reverse transcription-PCR. Aliquots (2 l) of the reverse transcription reactions were then submitted in duplicate to online quantitative PCR with the LightCycler system with SYBR green using the FastStart DNA-MasterSYBR Green I as described (36). The following SOCS-3 primer pair was used: sense, 5Ј-gctggccaaagaaataacca-3Ј; antisense, 5Ј-agctcaccagcctcatctgt-3Ј, product of 224 bp. The PCR was performed in a volume of 20 l: 2 l of FastStart DNA-MasterSYBR Green I, 3 mM MgCl 2 , and primers according to a primer concentration of 1 M. The instrument settings were: Denaturing at 95°C for 10 min, 45ϫ denaturing at 95°C for 15 s, annealing at 66°C for 10 s, elongation for 9 s.
Small Interfering RNA (siRNA)-siRNA oligonucleotides targeting PKC-␦ were designed and synthesized and annealed at Dharmacon Research, Lafayette, CO. An unrelated siRNA targeting firefly luciferase was used as control in all experiments. Transfection was performed with CellPhect (Amersham Biosciences, Buckinghamshire, UK) with 100 pmol of siRNA according to the instructions of the manufacturer. Briefly, 1 ϫ 10 5 of C2C12 cells/well were seeded in 6-well plates and transfected in Dulbecco's modified Eagle's medium containing 25 mM glucose, 10% FCS without antibiotics. 24 h after the glycerol shock, cells were stimulated as indicated.
Statistical Analysis-Results presented are derived from at least three independent experiments. Means Ϯ S.E. were calculated, and groups of data were compared using Student's t test. Statistical significance was set at p Ͻ 0.05.

IL-6 Induces Ser-318 Phosphorylation of IRS-1 in Muscle and in Mus-
cle Cell Culture Models-Since previous studies showed that serine phosphorylation of IRS-1 modulates insulin signal transduction, we studied whether IL-6 treatment of mice (2.5 g/kg intravenously for 5 min) induces Ser phosphorylation of IRS-1 using phospho-site-specific antibodies against inhibitory serine residue 307 (23) and the Ser-318 phosphorylation previously reported from our group (24,32). We detected a rapid and significant increase in the phosphorylation of Ser-307 in liver but not in muscle of IL-6-treated mice (Fig. 1, A and C). Phosphorylation of Tyr-705 of STAT-3 demonstrated the activation of IL-6-dependent pathways. Insulin treatment resulted in a clear phosphorylation of Ser-307 in muscle (Fig. 1A) as reported previously (23). When we studied the phosphorylation of Ser-318 of IRS-1 in these animals, we found just the opposite; IL-6 induced a rapid (5 min) and significant phosphorylation of this serine residue in muscle but not in liver ( Fig. 1, B and C). Noteworthily, we observed a clear although less pronounced phosphorylation of STAT-3 after IL-6 treatment in muscle when compared with liver tissue, indicating that the muscle responds to IL-6.
To investigate the rapid and unexpected IL-6-induced phosphorylation of Ser-318 in skeletal muscle, we studied this effect in cultured skeletal muscle cells. Using three different cell models, we observed a rapid (2 min) and transient IL-6-induced phosphorylation of Ser-318 in human myotubes, murine C2C12 myoblasts, and rat L6 myotubes, which reached a maximum after 5-10 min of IL-6 stimulation, whereas the phosphorylation of STAT-3 proceeded (Fig. 2, A-C). Together the data show that IL-6 induces a very rapid phosphorylation of a distinct serine residue (i.e. Ser-318 in muscle) on IRS-1, an adaptor molecule previously described as an essential mediator in the signal transduction of insulin and IGF-1 (38). The following questions arise from this novel finding. First, how are the IL-6 and the insulin signaling pathways linked to each other; second, which kinase is responsible for the IL-6-stimulated Ser-318 phosphorylation, and third, what is the physiological function of this serine phosphorylation of IRS-1 in muscle ?
IL-6 Induces IRS-1 Association with IL-6 Receptor Complex-Our results showing an unexpectedly rapid, IL-6-induced Ser phosphorylation of IRS-1 indicated a possible interaction of proximal IL-6 signaling molecules and IRS-1. To test this hypothesis, we transfected C2C12 cells with IRS-1 and performed co-immunoprecipitation studies using antibodies both against IL-6 receptor subunit gp130, which is not unique for IL-6 signal transduction, and against the specific IL-6 receptor subunit IL-6R. IL-6 signal transduction as assessed by STAT-3 phosphorylation FIGURE 1. IL-6 treatment induces different Ser phosphorylation of IRS-1 in muscle and liver. Ten-week-old male C57Bl/6 mice were treated with recombinant IL-6 (2.5 g/kg intravenously (i.v.)) or 5 IU of insulin (Ins) for 5 min, muscle and liver tissue was obtained, and protein extracts were prepared as described under "Experimental Procedures." Proteins (150 g) were separated by 5% SDS-PAGE, and Ser phosphorylation of IRS-1 was studied using anti-phospho-Ser-307 (A, p-Ser-307) and anti-phospho-Ser-318 (B, p-Ser-318) antibodies. Phosphorylation of STAT-3 was monitored to show activation of the IL-6 signaling pathway. Membranes were reprobed for protein amount. con, control; p-Tyr-705, phospho-Tyr-705. C, densitometric quantification of the relative IL-6-induced phosphorylation of Ser-307 and Ser-318 in muscle and liver extracts related to IRS-1 protein. The values of untreated mice were set as 1 (n ϭ 5; mean Ϯ S.E.; * , p Ͻ 0.05 versus untreated mice). rel., relative. was similar in wild-type and IRS-1-transfected cells (Fig. 3A). In nontransfected cells, we detected a faint signal for endogenous IRS-1 in the gp130 immunoprecipitate that increased after short term IL-6 stimulation (Fig. 3B). In IRS-1-transfected cells, the basal and IL-6-stimulated IRS-1 association was much more pronounced (Fig. 3B). By immunoprecipitation of IL-6R, the amount of co-precipitated IRS-1 is also markedly enhanced in IRS-1-transfected cells after 10 min of IL-6 stimulation (Fig. 3C). The results indicate that ligand-induced IL-6 receptor complex stimulation rapidly recruits IRS-1.
Co-recruited PKC-␦ Phosphorylates Ser-318-The serine 318 residue of IRS-1 was previously identified by in vitro phosphorylation of IRS-1-GST fusion proteins and subsequent mass spectrometry-based sequencing as a PKC-dependent phosphorylation site (24,39). Further studies using this in vitro approach revealed that all three classes of PKC isoforms, classical, novel and atypical PKCs, can phosphorylate Ser-318 (40). Since an important role for PKC-␦ in IL-6 signaling has been demonstrated (41), we hypothesized that this PKC isoform could be responsible for IL-6-induced Ser-318 phosphorylation of IRS-1. To elucidate the possible participation of PKC-␦, we co-transfected active or kinasenegative PKC-␦ and IRS-1 in C2C12 cells. Overexpression of active PKC-␦ caused a very strong increase in Ser-318 phosphorylation in unstimulated cells, with no further obvious increase after IL-6 stimulation (Fig. 4A). In control cells, IL-6 treatment enhances Ser-318 phosphorylation as expected, whereas overexpression of the kinase-inactive PKC-␦ caused no change in Ser-318 phosphorylation (Fig. 4A). The data suggest that this serine residue can be phosphorylated by PKC-␦ in skeletal muscle cells. To strengthen our hypothesis, we investigated the effect of IL-6 on Ser-318 phosphorylation in C2C12 cells after knockdown of PKC-␦ using a siRNA silencing approach (Fig. 4B). The IL-6induced phosphorylation of STAT-3 remained unaffected in siRNA oligonucleotides targeting PKC-␦ (si-PKC-␦)-transfected cells, but phosphorylation of Ser-318 was blocked completely (Fig. 4B). Thus, we conclude that PKC-␦ mediates the effect of IL-6 on Ser-318 phosphorylation.
We then studied whether IL-6 stimulates the association of both IRS-1 and PKC-␦ with the IL-6R. As shown in Fig. 4C, the minor co-precipitation of endogenous PKC-␦ with IL-6R in wild-type cells is increased by IL-6 treatment, and this increase is similar in IRS-1-transfected cells. Transfection of C2C12 cells with PKC-␦ or both PKC-␦ and IRS-1 led to a comparable stimulation; the amount of endogenous as well as co-transfected PKC-␦ was increased in the IL-6R immunopre-   Fig. 3A were immunoprecipitated with anti-IL-6 R antibodies (IP) and immunoblotted for PKC-␦ (IB). D, C2C12 cells were co-transfected with IRS-1 and PKC-␦ as indicated and stimulated with 20 ng/ml mouse recombinant IL-6 for 10 min. Control cells were transfected with empty expression vector pcDNA3 (con). Co-immunoprecipitation was performed with anti-IL-6R antibodies, and IRS-1 and PKC-␦ were detected in the immunocomplexes by immunoblotting. The arrow indicates the blots that were obtained after shorter exposure time. E, cell extracts shown in D were monitored for IRS-1 and PKC-␦ expression and phosphorylation of STAT-3. F, cell extracts as shown in B were immunoprecipitated with anti-IL-6R antibodies, and IRS-1 was detected in the immunocomplexes by immunoblotting. Representative immunoblots of four independent experiments are shown. MARCH 17, 2006 • VOLUME 281 • NUMBER 11

JOURNAL OF BIOLOGICAL CHEMISTRY 7063
cipitates in an IL-6-dependent manner (Fig. 4, C and D). The amount of co-precipitated IRS-1 was not enhanced by transfection of PKC-␦ (Fig.  4D) or influenced by transfection of kinase-negative PKC-␦ (data not shown). Of note, IL-6-induced phosphorylation of STAT-3 was comparable in control cells and after transfection of IRS-1 and PKC-␦ (Fig. 4, A and E) or IRS-1 and kinase-negative PKC-␦ (Fig. 4A). Moreover, when PKC-␦ expression is down-regulated in siRNA-transfected C2C12 cells (Fig. 4B), the IL-6-induced recruitment of IRS-1 in the IL-6R complex is not affected (Fig. 4F). Thus, IL-6 induced the association of IRS-1 and IL-6R independent of PKC-␦. The data also suggest that IRS-1 and PKC-␦ were in close proximity, thus enabling PKC-␦ to phosphorylate Ser-318 of IRS-1.

IL-6 Treatment Does Not Influence Tyrosine Phosphorylation of IRS-1-
To study the biological function of the IL-6-stimulated recruitment of IRS-1 to the IL-6R complex and the subsequent Ser-318 phosphorylation of IRS-1, we investigated the tyrosine phosphorylation of IRS-1 in IRS-1-transfected C2C12 cells after stimulation with IL-6 for 10 min or insulin for 5 min. Although IL-6 treatment of the IRS-1-transfected cells did not induce the tyrosine phosphorylation of total IRS-1 protein, stimulation with insulin led to the expected increase of this phosphorylation (Fig. 5A). Although a huge amount of IRS-1 was co-immunoprecipitated with the IL-6R, no tyrosine phosphorylation of this IRS-1 fraction was detectable in both IL-6-and insulin-stimulated cells (Fig. 5A). Similarly, we observed no tyrosine phosphorylation of co-precipitated IRS-1 in control cells (Fig. 5A). To further investigate the role of Ser-318, we transfected C2C12 cells with IRS-1/Ala-318 to exclude phosphorylation at this site and IRS-1/Glu-318 to mimic permanent phosphorylation. We observed no effect of Ala-318 or Glu-318 on the tyrosine phosphorylation of total IRS-1 after stimulation with IL-6, insulin, or both (Fig.  5B). Preincubation with IL-6 neither reduced nor increased the amount of tyrosine-phosphorylated IRS-1 (Fig. 5B). These data indicate first that treatment of C2C12 cells with IL-6 did not inhibit insulin signal transduction at the level of IRS-1 tyrosine phosphorylation. Secondly, a potential regulation of insulin action by IL-6-induced phosphorylation of Ser-318 takes place downstream of IRS-1. Fig. 2C, IL-6 induced phosphorylation of Ser-318 in L6 myotubes with a similar time course when compared with C2C12 cells (Fig. 2B). To confirm the results on IRS-1/IL-6R complex formation and tyrosine phosphorylation of IRS-1 obtained in C2C12 cells, we studied the effects of IL-6 in L6 myotubes after stimulation with IL-6 or insulin (Fig. 6A). Again, we observed no effect of IL-6 on tyrosine phosphorylation of total IRS-1, whereas insulin induced this tyrosine phosphorylation (Fig. 6B). IL-6 increased the amount of coimmunoprecipitated IRS-1 in the IL-6R complex, but again, no tyrosine phosphorylation of this IRS-1 fraction was detected (Fig. 6B).

Effect of IL-6 on IRS-1 Recruitment and Tyrosine Phosphorylation in L6 Myotubes-As demonstrated in
Effect of IL-6-induced Ser-318 Phosphorylation on Insulin Signal Transduction-Several reports on the pathophysiological function of serine phosphorylation of IRS-1 pointed to a desensitizing action of specific serine residues on insulin signal transduction, e.g. among others, the intensively studied Ser-307 phosphorylation (22). However, there is increasing evidence for a more subtle regulation of insulin action by serine phosphorylation of IRS-1 with some serine sites also positively modulating insulin signal transduction (28,42). Very recently, we could demonstrate that the phosphorylation of Ser-318 is not per se inhibitory but could even improve insulin signal transduction in the early phase of insulin action (32). Here, we support our data using L6GLUT4 myotubes, which in contrast to C2C12 cells exhibit insulinsensitive glucose uptake (35), stably transfected with IRS-1 WT, IRS-1/ Ala-318, or IRS-1/Glu-318. Mimicking a permanent Ser-318 phosphorylation in the Glu-318-transfected cells led to an enhanced increase in phosphorylation of Akt after 5 min of insulin stimulation when compared with IRS-1 WT and Ala-318-transfected cells (Fig. 7A). This improvement of insulin action is further reflected by the increase in insulin-stimulated glucose uptake in the Glu-318 myotubes, determined as incorporated 2-   our group (32). These results indicate that IL-6-induced phosphorylation of Ser-318 could enhance insulin-stimulated glucose uptake in muscle cells. Of note, treatment of IRS-1 WT myotubes with IL-6 caused a slight but significant increase in glucose uptake in the absence of insulin (1.14 Ϯ 0.05-fold; p Ͻ 0.007), which was not observed in IRS-1/Ala-318-transfected cells (0.98 Ϯ 0.07-fold; data not shown).

IL-6 Treatment of Mice Induces a Rapid and Marked Induction of SOCS-3 Expression in Liver but Not in Skeletal
Muscle-Our data did not support any inhibitory action of IL-6 on insulin signal transduction in skeletal muscle, as has been reported, especially for liver and hepatocytes. Since the rapid induction of SOCS-3 expression by IL-6 has been demonstrated as one important mechanism for the IL-6-mediated reduction of insulin action in hepatocytes (43), we hypothesized a different regulation of SOCS-3 expression in muscle and liver. Therefore, we measured SOCS-3 mRNA expression in vivo in muscle and liver tissue of mice treated for 30 and 60 min with 2.5 g/kg of IL-6. IL-6 plasma levels in these animals increased up to 63.3 Ϯ 6.3 and 60.1 Ϯ 9.4 pg/ml after 30 and 60 min of IL-6 intraperitoneal application when compared with 7.2 Ϯ 0.3 pg/ml in control mice. In the liver, we observed a strong, ϳ10-fold induction of SOCS-3 mRNA expression, both after 30 min and after 60 min of IL-6 treatment, whereas the increase in SOCS-3 levels in muscle was not significant (Fig. 8).

DISCUSSION
The exercise-induced IL-6 production and secretion within skeletal muscle fibers have raised the question of a putative role of IL-6 in the energy supply and energy storage of the muscle during and after exercise. This function has to be considered as a short term, physiological, and probably local effect of IL-6 within minutes or probably a few hours. In the present study, we followed the hypothesis that IL-6 may influence muscle energy metabolism by direct interaction of the IL-6 receptor complex with important molecules of the insulin signaling cascade in muscle.
We have provided here for the first time evidence that IL-6 induces a rapid, occurring within 10 min, recruitment of IRS-1, one keystone in insulin signal transduction, to the IL-6 receptor complex. We detected endogenous IRS-1 as well as co-expressed IRS-1 by immunoprecipitation using antibodies against both IL-6 receptor subunits, gp130 and the IL-6R, in IL-6-stimulated C2C12 cells. The sequential regulation of this rapid complex formation is yet unknown. Theoretically, IRS-1 could directly associate with the phosphorylated tyrosine residues of the receptor subunit gp130 via the phosphotyrosine binding or Src homology (SH) 2 domains. Since the exact peptide sequence that is required for the binding of IRS-1 to the tyrosine-phosphorylated insulin receptor (NPXY motif) is not present in gp130, it could be speculated that SH2 domains-containing adaptor molecules that are tyrosine-phosphorylated by IL-6 mediate the ligand-induced binding of IRS-1 to the IL-6 receptor complex.
Another major finding of the present study was that IL-6 modifies IRS-1 by the rapid and transient phosphorylation of Ser-318, which was demonstrated in the muscle of IL-6-treated mice and in various skeletal muscle cells. Noteworthily, this modification was not observed in the liver of these mice, although the activation of IL-6 signaling pathways by the IL-6 treatment, as demonstrated by the phosphorylation of STAT-3, was much more pronounced in liver when compared with muscle.
We have provided several lines of evidence that PKC-␦ could be the kinase responsible for the IL-6-induced Ser-318 phosphorylation. First, PKC-␦ was recruited to the IRS-1-containing IL-6 receptor complex by IL-6 stimulation, which is well in line with a previous report demonstrating the association of this PKC with gp130 via STAT-3, thereby regulating IL-6 signaling (41). Second, overexpression of PKC-␦ resulted in a strong phosphorylation of Ser-318 also in unstimulated cells, whereas overexpression of a kinase-negative PKC-␦ was uneffective, and third, knock-down of PKC-␦ prevented the IL-6-induced Ser-318 phosphorylation. Moreover, PKC-␦ was able to phosphorylate this serine residue using an in vitro phosphorylation assay (data not shown).  Ten-week-old male C57Bl/6 mice were treated with recombinant IL-6 (2.5 g/kg intraperitoneally (i.p.)) for 30 and 60 min. Liver and muscle tissue was obtained, and RNA was prepared as described under "Experimental Procedures." SOCS-3 expression was determined by reverse transcription-PCR. Shown are the relative (rel.) expression units (mean Ϯ S.E.; n ϭ 5). Asterisks denote significant differences in IL-6-treated versus untreated mice.
Of note, another PKC isoform, namely atypical PKC-, has been demonstrated to mediate the insulin-dependent Ser-318 phosphorylation in baby hamster kidney cells (24) and in skeletal muscle cells (32). Thus, one serine residue could be phosphorylated by different PKC isoforms depending on the stimulus.
The functional relevance of the Ser-318 phosphorylation was demonstrated in L6GLUT4 myotubes expressing IRS-1 WT, IRS-1/Ala-318, the non-phosphorylated form, or IRS-1/Glu-318, which simulates a permanent phospho-Ser-318 modification. The permanent Ser-318 "phosphorylation" of IRS-1 enhances insulin-stimulated Akt phosphorylation and glucose uptake after short term insulin stimulation of the myotubes, as shown in the present study and in a very recent study (32), whereas mutation of Ser-318 to alanine reduced Akt phosphorylation in the early phase of insulin stimulation. Thus, the phosphorylation of this serine residue could initially improve early insulin action. Of note, the IL-6-induced Ser-318 phosphorylation is a very rapid and transient modulation and was no longer visible after 30 min of IL-6 stimulation. The first published participation of Ser-318 in the reduction of insulin signal transduction (24) is only visible after long term insulin treatment with high insulin doses, representing an insulin-resistant state but not the situation during physical activity when insulin concentrations are low. Further studies indicate that besides a positive modulation of insulin action in the early phase, the insulin-induced phosphorylation of Ser-318 is more likely to be a signal for the physiological attenuation of insulin action. Following this line, in the IRS-1/Glu-318 myotubes treated for a long term with insulin, the glucose uptake is attenuated similar to the wild-type cells (32).
The exact molecular mechanism for the insulin-sensitizing effect of the phosphorylation of Ser-318 on glucose uptake in skeletal muscle cells remains open. We detected no effect of IL-6 on the tyrosine phosphorylation of the IRS-1 fraction recruited to the IL-6R complex or of IL-6 or Ser-318 on insulin-stimulated tyrosine phosphorylation of total IRS-1. These data suggest that under some circumstances, serine/threonine phosphorylation could regulate the function of IRS-1 without direct effects on its tyrosine phosphorylation. It could be speculated that IRS-1 might serve as a multidocking site protein in receptor complexes beside the insulin receptor.
The next downstream effector of the insulin signaling cascade is PI3K. An IL-6-dependent PI3K activation was described in other cell culture models (44). We also observed an IL-6-induced association of the p85 subunit to the IL-6 receptor complex in C2C12 cells. 3 This could provide evidence for the hypothesis that the delicate molecular balance of the regulatory p85 and the catalytic subunit p110 of PI3K is regulated by IL-6, leading to a sensitization of the PI3K-Akt pathway for insulin since the p85 subunits exhibits a negative effect on insulin signaling, probably due to their greater abundance when compared with the p110 subunit (45). However, we found no evidence for a regulation of the p85 recruitment to both IL-6R complex and IRS-1 by phosphorylation of Ser-318. 3 Thus, the role of PI3K in IL-6-enhanced insulin action in skeletal muscle cells remains open.
Our data may help explain the discrepancy of how IL-6 may act in an apparently opposite way on insulin signal transduction in different tissues and tissue-specific cell types. From previous reports, one could assume different molecular mechanisms by which IL-6 could reduce insulin signal transduction; one could include the mammalian target of rapamycin (mTOR) or Jun-activated kinase-mediated phosphorylation of Ser-307 of IRS-1, which in turn reduces insulin signaling (46). Another mechanism includes IL-6-induced expression of SOCS-3, which in turn attenuates insulin signaling in hepatocytes (43). In the present study, we have shown that none of these mechanisms is operating in skeletal muscle, supporting the observation that IL-6 is not reducing insulin action in skeletal muscle cells (36). Moreover, our data have provided evidence for an IL-6-induced stimulatory effect on insulin action via muscle-specific phosphorylation of Ser-318 of IRS-1. Together with the recently reported muscle-specific IL-6-induced Ser-473 phosphorylation of Akt, which in turn enhances insulin-stimulated glycogen synthesis in human myotubes (47), our data suggest a molecular mechanism concerning how insulin signaling is modulated differently in different tissues.
The putative metabolic function of the contraction-induced IL-6 production described here could be a local function within skeletal muscle during or after exercise. Such a role for IL-6 in skeletal muscle energy balance has been hypothesized by several groups, including the improvement of fatty acid oxidation (48,49), enhancement of glucose disposal rate (19), activation of AMP-activated kinase (50), and more in general, involvement in exercise endurance (51). Noteworthily, this physiological role of the exercise-induced IL-6 in muscle is not in contrast to several reports demonstrating a role for pathophysiological elevated IL-6 plasma levels in the development of insulin resistance (3)(4)(5)(6).
First, systemic levels of IL-6 are chronically increased by obesity or inflammation, whereas exercise-induced IL-6 production is transient; second, exercise predominantly increases local concentrations of IL-6 in muscle; and third, IL-6 appears to exert tissue-specific effects.
In conclusion, IL-6 is able to induce a direct interaction of the IL-6 receptor complex with IRS-1 in skeletal muscle cells, thereby improving insulin action. This possibly physiological and local function of exerciseinduced IL-6 has to be discriminated from the pathophysiological function of chronically elevated circulating levels of IL-6 that are able to induce insulin resistance.