IL-6 has recently been recognized as a myokine (1), which is released

Transcription

1 DIABETES-INSULIN-GLUCAGON-GASTROINTESTINAL Role of Adenosine 5 -Monophosphate-Activated Protein Kinase in Interleukin-6 Release from Isolated Mouse Skeletal Muscle Stephan Glund, Jonas T. Treebak, Yun Chau Long, Romain Barres, Benoit Viollet, Jorgen F. Wojtaszewski, and Juleen R. Zierath Department of Molecular Medicine and Surgery (S.G., Y.C.L., R.B., J.R.Z.), Section for Integrative Physiology, Karolinska Institutet, SE Stockholm, Sweden; Copenhagen Muscle Research Center (J.T.T., J.F.W.), Department of Exercise and Sports Sciences, University of Copenhagen, DK-2400 Copenhagen, Denmark; Institute Cochin (B.V.), Université Paris Descartes, Centre National de la Recherche Scientifique (Unité Mixte de Recherche 8104), Paris, France; and Institut National de la Santé et de la Recherche Médicale, Unité 567 (B.V.), Paris, France IL-6 is released from skeletal muscle during exercise and has consequently been implicated to mediate beneficial effects on whole-body metabolism. Using 5-aminoimidazole-4-carboxamide ribofuranoside (AICAR), a pharmacological activator of 5 -AMP-activated protein kinase (AMPK), we tested the hypothesis that AMPK modulates IL-6 release from isolated muscle. Skeletal muscle from AMPK 2 kinase-dead transgenic, AMPK 1 knockout (KO) and AMPK 3 KO mice and respective wild-type littermates was incubated in vitro, in the absence or presence of 2 mmol/liter AICAR. Skeletal muscle from wild-type mice was also incubated with the AMPK activator A Incubation of mouse glycolytic extensor digitorum longus and oxidative soleus muscle for 2 h was associated with profound IL-6 mrna production and protein release, which was suppressed by AICAR (P 0.001). Basal IL-6 release from soleus was increased between AMPK 2 kinase-dead and AMPK 1 KO and their respective wild-type littermates (P 0.05), suggesting AMPK participates in the regulation of IL-6 release from oxidative muscle. The effect of AICAR on muscle IL-6 release was similar between AMPK 2 KD, AMPK 1 KO, and AMPK 3 KO mice and their respective wild-type littermates (P 0.001), indicating AICAR-mediated suppression of IL-6 mrna expression and protein release is independent of AMPK function. However, IL-6 release from soleus, but not extensor digitorum longus, was reduced 45% by A Our results on basal and A mediated IL-6 release provide evidence for a role of AMPK in the regulation of IL-6 release from oxidative skeletal muscle. Furthermore, in addition to activating AMPK, AICAR suppresses IL-6 release by an unknown, AMPK-independent mechanism. (Endocrinology 150: , 2009) ISSN Print ISSN Online Printed in U.S.A. Copyright 2009 by The Endocrine Society doi: /en Received August 13, Accepted September 17, First Published Online September 25, 2008 IL-6 has recently been recognized as a myokine (1), which is released from skeletal muscle during prolonged exercise (2, 3). Various molecular mechanisms are implicated in the control of IL-6 production from skeletal muscle, including Ca 2 -dependent pathways mediated via calcineurin activation (4, 5), increased reactive oxygen species (6), autocrine stimulation by IL-6 (7, 8), and impaired glucose availability (9, 10). However, correlative studies implicate a relationship between 5 -AMP-activated protein kinase (AMPK) activity and IL-6 release (11 14). Exposure of isolated skeletal muscle or cultured myotubes to IL-6 directly increases AMPK phosphorylation (11, 15, 16) and glucose uptake and metabolism in a manner analogous to acute exercise. AMPK is a heterotrimeric complex, consisting of a catalytic -subunit and the regulatory - and -subunits, which functions as a fuel sensor to coordinate the balance between energy consuming and energy producing processes (reviewed in Ref. 17). In cultured human myotubes, small interfering RNA-mediated silencing of AMPK subunits prevents the effect of an AMP-analog, 5-aminoimidazole-4-carboxamide ribofuranoside (AICAR) to increase IL-6 mrna (8), directly implicating AMPK in IL-6 release. Exposure of mouse cardiac fibroblasts to AICAR or metformin increases IL-6 production (18), supporting the hy- Abbreviations: AICAR, 5-Aminoimidazole-4-carboxamide ribofuranoside; AMPK, 5 -AMP-activated protein kinase; EDL, extensor digitorum longus; F, forward; KD, kinase dead; KO, knockout; R, reverse. 600 endo.endojournals.org Endocrinology, February 2009, 150(2):

2 Endocrinology, February 2009, 150(2): endo.endojournals.org 601 pothesis that AMPK activation can directly trigger IL-6 production. In contrast, in vitro AICAR exposure of human adipose tissue and human skeletal muscle cells, as well as intact rat soleus muscle, is associated with a decrease in IL-6 gene expression and secretion (19). Here we tested the hypothesis that AMPK is required for in vitro IL-6 production and release from oxidative and glycolytic mouse muscle at rest and in response to pharmacological activators of AMPK. Using AMPK 2 kinase dead (KD) mice (20, 21), AMPK 1 knockout (KO) mice (22), and AMPK 3 KO mice (23), we determined whether IL-6 production and release is dependent on intact AMPK signaling. Materials and Methods Animals The experiments were approved by the Regional Animal Ethical Committee (Stockholm, Sweden) and the Danish Animal Experimental Inspectorate. AMPK 2 KD mice (20, 21), AMPK 1 KO mice (22) backcrossed in seven generations on an SV129 background, AMPK 3 KO mice (23), and their respective wild-type littermates were studied. Mice were maintained on a 12-h light, 12-h dark cycle and received standard rodent chow. Muscle incubation procedure Extensor digitorum longus (EDL) and soleus muscles were removed from anesthetized mice [Avertin, 2,2,2-tribromo ethanol 99%, and tertiary amyl alcohol at ml per gram body weight ip for AMPK 3 KO mice and respective wild-type littermates or pentobarbital (6 mg per 100 g body weight) for AMPK 2 KD and AMPK 1 KO mice and respective wild-type littermates] and incubated for 30 min at 30 C in vials containing preoxygenated (95% O 2-5% CO 2 ) incubation buffer [Krebs-Henseleit-buffer containing 5 mmol/liter HEPES, 15 mmol/liter mannitol, and 5 mmol/liter glucose]. Muscles were then transferred to new vials containing 1 ml of fresh incubation buffer, either with or without AICAR (2 mmol/liter), ionomycin (1 mol/liter), or A (100 mol/liter), as indicated. AICAR is an adenosine analog that can be taken up into intact skeletal muscle, is phosphorylated to form 5-aminoimidazole-4-carboxamide ribonucleotide, and mimics the activating effects of AMP on AMPK (24). A is a direct activator of AMPK that mimics the effects of AMP (25). After 2 h of incubation at 30 C, muscles were removed, immediately frozen in liquid nitrogen, and stored at 80 C for further analysis. The incubation media were stored at 20 C until further analysis of IL-6 concentration. Skeletal muscle IL-6 release IL-6 release from skeletal muscle was measured as the concentration of IL-6 in the media at the end of the 2-h incubation period. IL-6 concentration was determined by ELISA (R&D Systems, Abingdon, UK). RNA purification and DNA synthesis Muscles were homogenized in 1 ml Trizol reagent (Sigma, St. Louis, MO), and RNA was isolated. Purified RNA was then treated with deoxyribonuclease I (DNA free kit; Ambion, Huntingdon, UK). Subsequently RNA was used as template for cdna synthesis [SuperScript first strand synthesis system (Invitrogen, Stockholm, Sweden) with random hexamer primers]. All procedures were performed according to the manufacturer s protocols. Quantitative real-time PCR Real-time PCR with the ABI PRISM 7000 sequence detector system and fluorescence-based SYBR-green technology (Applied Biosystems, Foster City, CA) was used to quantify cdna. PCR was performed in a final volume of 25 l. All samples were analyzed in duplicates and normalized against the expression of acidic ribosomal phosphoprotein P0 (36B4) (26), using relative quantification method. The forward (F) and reverse (R) primer sequences were as follows: for IL-6 (Gen- Bank NM ) CTGCAAGAGACTTC- CATCCAGTT (F) and AGGGAAGGCCGTG- GTTGT (R), for 36B4 (GenBank NM ) TGCAGCAGATCCGCATGT (F) and CTT- GCGCATCATGGTGTTCT (R). FIG. 1. Effect of AICAR and ionomycin on in vitro IL-6 release and mrna production. EDL and soleus muscle from wild-type mice was incubated in the absence (basal) or presence of either 1 mol/liter ionomycin (IM), 2 mmol/liter AICAR, or both, as indicated. IL-6 release into the media was determined after 2 h incubation (A and B). Data are expressed as picograms IL-6 per milliliter of media. For better comparison of the fiber-type dependency of IL-6 release, basal IL-6 release from EDL is indicated as a dashed line (B). IL-6 mrna concentration was determined in muscles after 2 h incubation (C and D). Data were standardized to 36B4 expression and are presented as arbitrary units (a.u.). Results are mean SEM for n 6 7 muscles/condition. ***, P for AICAR effect using one-way ANOVA. Statistics Data are expressed as mean SEM. Statistical differences were determined using either a one- or two-way ANOVA, with or without repeated measures as indicated in the figure legends, followed by Tukey s post hoc analysis, as appropriate. Student s t test was used to compare basal IL-6 release between fiber types. Levene s test of equality was performed to ensure equality of variances between the groups. When significance was observed, data were log transformed. In all cases, log transformation was sufficient to obtain equal variances among groups. P 0.05 was considered significant.

3 602 Glund et al. Role of AMPK in Muscle IL-6 Release Endocrinology, February 2009, 150(2): Results Effect of ionomycin and AICAR on IL-6 production from EDL or soleus muscle IL-6 was readily detectable in the media after the 2-h incubation of either EDL (Fig. 1A) or soleus (Fig. 1B) muscle, indicating that even under basal (resting) conditions, mouse skeletal muscle produces IL-6. We observed a fiber-type dependence on IL-6 release. Basal IL-6 release was significantly higher in oxidative soleus vs. glycolytic EDL muscle ( vs pg/ml, respectively, P 0.001). Addition of the Ca 2 -ionophore, ionomycin (1 M) to the media was without effect on basal IL-6 release in either EDL (Fig. 1A) or soleus (Fig. 1B) muscle. We do not confirm earlier finding of increased IL-6 release from rat soleus muscle after ionomycin stimulation (27) in our mouse muscle preparation. Addition of AICAR to the incubation media of either EDL (Fig. 1A) or soleus (Fig 1B) led to a marked reduction (P 0.001) in IL-6 release. We next evaluated whether changes in IL-6 release are correlated with IL-6 mrna. Ionomycin treatment was without effect on basal IL-6 mrna expression in either EDL (Fig. 1C) or soleus (Fig 1D) muscle after the 2-h incubation. In parallel with the results for IL-6 release, AICAR exposure led to a similar reduction (P 0.001) in IL-6 mrna levels. These results provide evidence to suggest AICAR prevents IL-6 release by blocking IL-6 mrna production. FIG. 2. Skeletal muscle IL-6 release in relation to AMPK activity. Soleus muscle was incubated in basal media for 1, 2, or 3 h (A) or either basal or AICAR (2 mmol/liter)-containing media for 2 h (B). Incubation media were collected and IL-6 release into the media was determined. Data are expressed as picograms IL-6 per milliliter of media. At the end of the incubation period, phosphorylation of AMPK and its substrate ACC was determined by immunoblot analysis. Unstimulated soleus muscle served as control (CON). A representative image is shown (C). A quantification of AMPK phosphorylation is shown. Data are expressed as arbitrary units (D). Equal loading was ensured by performing an immunoblot analysis for AMPK protein (C). Results are mean SEM for n 4 (A) and n 5 6 (B D) muscles per condition. ***, P for effect of AICAR vs. basal. Skeletal muscle IL-6 release in relation to AMPK activity IL-6 was undetectable in the media after 1 h incubation of isolated soleus muscle. However, IL-6 was detected in the media after 2 3 h incubation (Fig. 2A). An autocrine mechanism likely contributes to basal IL-6 release under the applied conditions because the IL-6 media concentration reached levels above that achieved in serum under normal fed conditions from resting mice [ 50 pg/ml in basal media from soleus muscle after 2 h incubation vs. 10 pg/ml in serum from fed mice (Glund, S., and J. R. Zierath, unpublished)]. Incubation of soleus muscle in the presence of AICAR led to a marked reduction (P 0.001) in basal IL-6 release (Fig. 2B). To study the effect of the incubation protocol and AICAR on AMPK activity, we determined phosphorylation of AMPK and its downstream substrate Acetyl-CoAcarboxylase. Incubation of soleus muscle for 2 h under the basal condition was without effect on AMPK and ACC phosphorylation (Fig. 2, C and D). As expected, addition of AICAR (2 mmol/ liter) to the incubation media significantly increased phosphorylation of AMPK and ACC (Fig. 1, C and D). AMPK and skeletal muscle IL-6 release To determine whether the effect of AICAR on in vitro IL-6 release is dependent on AMPK, we studied skeletal muscle from three mouse models with impaired AMPK signaling. Basal IL-6 release from EDL muscle (Fig. 3A) was similar between AMPK 2 KD and wild-type mice. Interestingly, IL-6 release from soleus muscle (Fig. 3B) was greater in AMPK 2KDvs. wild-type mice (P 0.05). IL-6 protein release from soleus muscle was also greater in AMPK 2 KD muscle after AICAR stimulation ( vs for AMPK 2 KDvs. wild-type, respectively, P 0.05); however, the AICAR-mediated suppression of IL-6 release was similar between AMPK 2 KD and wild-type mice (80 vs. 70% and 80 vs. 75% reduction in IL-6 release for EDL and soleus muscle, respectively (P 0.001). This suggests AICAR-mediated inhibition of in vitro IL-6 release is independent of AMPK, whereas the basal IL-6 release appears to be partly AMPK dependent. We next determined whether changes in IL-6 release are correlated with IL-6 mrna. Basal IL-6 mrna measured in either EDL (Fig. 3C) or soleus (Fig. 3D) muscle was similar between AMPK 2 KD and wild-type mice. AICAR exposure reduced IL-6 mrna in EDL (Fig. 3C) and soleus (Fig. 3D) muscle, and this effect was similar between AMPK 2 KD and wild-type mice. AMPK 2 KD mice retain about 30 40% of basal AMPK 1 activity in skeletal muscle (28). Consequently, our results with AMPK 2 KD cannot exclude the potential contribution of the AMPK 1 isoform in the basal or AICAR-mediated IL-6 response. To

4 Endocrinology, February 2009, 150(2): endo.endojournals.org 603 Effect of A mediated AMPK activation on muscle IL-6 release EDL and soleus muscle were incubated in the presence of a specific AMPK activator, A Incubation of soleus skeletal muscle in the presence of A reduced basal IL-6 release (P 0.001), although the effect was less potent compared with AICAR [ 45% reduction in the presence of A (Fig. 5) vs % reduction in the presence of AICAR (Figs. 1 4)]. In contrast to AICAR, A was without effect on IL-6 release from EDL muscle. FIG. 3. Effect of AICAR on in vitro skeletal muscle IL-6 release and mrna production in AMPK 2 KD mice. EDL and soleus muscle from wild-type or AMPK 2 KD mice were incubated in the absence (basal) or presence of 2 mmol/liter AICAR. Muscle IL-6 release into the media was determined after 2 h incubation (A and B). Data are expressed as picograms IL-6 per milliliter of media. IL-6 mrna concentration was determined in the incubated muscles (C and D). Results were standardized to 36B4 mrna expression and are presented as arbitrary units (a.u.). Results are mean SEM for n 6 7 muscles/condition. ***, P for AICAR effect;, P 0.05 for genotype effect. Statistical analyses were performed using two-way ANOVA. determine whether basal or AICAR-mediated IL-6 release is dependent on the AMPK 1 isoform, IL-6 release from AMPK 1 KO mice was determined under basal and AICAR-stimulated conditions. Basal IL-6 release from EDL (Fig. 4A) was comparable between AMPK 1 KO, and wild-type mice. Interestingly, basal IL-6 release from soleus muscle (Fig. 4B) lacking AMPK 1 was significantly greater compared with wild-type muscle, providing further evidence for a role for AMPK in regulating basal IL-6 release. AICAR exposure blocked basal IL-6 release to a similar extent in EDL and soleus muscle from AMPK 1 KO mice and respective wild-type littermates (P 0.001). We also determined whether the AMPK 3 subunit plays a role in skeletal muscle IL-6 release using AMPK 3 KO mice. The AMPK 3 subunit is predominantly expressed in skeletal muscle, with greatest expression observed in muscle enriched in fast-twitch glycolytic fibers (29). Basal IL-6 release from either EDL (Fig. 4C) or soleus (Fig. 4D) was comparable between AMPK 3 KO and wild-type mice. Similar to the response observed in AMPK 2 KDand AMPK 1 KO mice, AICAR exposure blocked IL-6 release from EDL (Fig. 4C) and soleus (Fig. 4D) muscle from AMPK 3 KO mice (P 0.001). Discussion IL-6 is a myokine that is released from skeletal muscle in response to physical exercise (1). Skeletal muscle contraction (30), as well as pharmacological agents that increase AMPK activity (8, 18, 19), have been suggested to promote (8, 18) or suppress (19) IL-6 production. Thus, AMPK activation has been implicated in the mechanism governing IL-6 production and release (8, 13). Here we determined the role of AMPK in basal, AICAR, and A mediated IL-6 release from either fast-twitch glycolytic (EDL) or slow-twitch oxidative (soleus) mouse skeletal muscle in vitro. Using animal models in which AMPK subunits have been either deleted or inactivated, we provide evidence implicating AMPK in the regulation of basal IL-6 release from isolated oxidative skeletal muscle. We also observe the AMPK activating agent AICAR suppresses IL-6 release from glycolytic and oxidative skeletal muscle and this effect is retained in mouse models with impaired AMPK signaling. The direct AMPK activator A prevents IL-6 release only from oxidative soleus muscle. Our data implicate AMPK in the control of IL-6 release in resting muscle. Moreover, we show AMPK-independent pathways are likely to contribute to IL-6 release under AICAR-stimulated conditions. Several lines of evidence underscore the cellular mechanism(s) underlying IL-6 release from skeletal muscle (4, 5, 7 10, 15, 27). In healthy humans, a 3-h infusion of recombinant IL-6 increased IL-6 mrna levels 120-fold (7), providing evidence for a role for autocrine regulation of skeletal muscle IL-6 production. In cultured cells, the autocrine effect of IL-6 is dependent on intracellular calcium (8). Here we confirm that isolated EDL or soleus muscle produces and releases IL-6, as evident by the increased media concentration of IL-6 and skeletal muscle mrna expression after the 2-h incubation period. Basal IL-6 release was greater in oxidative soleus, compared with glycolytic EDL muscle, consistent with previous evidence from rat skeletal muscle (31). Of relevance, insulin sensitivity is increased in oxidative, compared with glycolytic muscle (32). IL-6 can regulate satellite cell-mediated skeletal muscle hypertrophy (33) and directly stimulate differentiation of human skeletal muscle cells (15). Thus, it is tempting to speculate that higher basal IL-6 release from ox-

5 604 Glund et al. Role of AMPK in Muscle IL-6 Release Endocrinology, February 2009, 150(2): FIG. 4. Effect of AICAR on in vitro skeletal muscle IL-6 release and mrna production in AMPK 1 KO mice and AMPK 3 KO mice. EDL (A and C) and soleus (B and D) muscle from either AMPK 1 KO mice or AMPK 3 KO mice and their respective wild-type littermates were incubated in the absence (basal) or presence of 2 mmol/liter AICAR. IL-6 release into the media was determined after the 2-h incubation. Data are expressed as picograms IL-6 per milliliter of media. Results are mean SEM for n 7 8 (AMPK 1 KO) and n 3 4 (AMPK 3 KO) muscles per condition. ***, P for AICAR effect;, P 0.05 for genotype effect. Statistical analyses were performed using two-way ANOVA with repeated measures. idative skeletal muscle might also enhance insulin sensitivity, which is a characteristic feature of the oxidative fiber type. However, as is evident from the time-course experiment, at the 2-h time point, the muscle is markedly stimulated under the basal in vitro condition to produce IL-6. Thus, AICAR may specifically inhibit IL-6 production induced by this putative stimulus. Whereas the exact stimulus for the basal IL-6 release is unknown, FIG. 5. Effect of A on in vitro skeletal muscle IL-6 release. EDL and soleus muscle was incubated in the presence or absence of the AMPK activator A , and IL-6 release into the media was determined. Data are expressed as picograms IL-6 per milliliter of media. Results are mean SEM for n 5 muscles per condition. ***, P for effect of A vs. basal. reactive oxygen species, local hypoxia, or mechanical stress during the muscle dissection period may increase IL-6 production. Furthermore, IL-6 can possibly increase its own production via an autocrine mechanism (7). The effect of ionomycin stimulation on IL-6 protein release was modest compared with previous work in rat soleus muscle in vitro (27). Despite a tendency (33%; P N.S.) for an increase in IL-6 protein release from EDL and soleus muscle, ionomycin stimulation was without effect on IL-6 production. The reasons for the lack of ionomycin stimulation might be manifold and may possibly arise from species differences (mouse vs. rat muscle) as well as methodological aspects. Importantly, rat soleus muscle contains a greater portion of oxidative muscle fibers compared with mouse soleus muscle (34). Because IL-6 release from rat oxidative muscle was greater, the difference in fiber-type composition between mouse and rat skeletal muscle may account for the smaller sensitivity to ionomycin. In the previous study of rat soleus muscle, basal IL-6 release reached 20 pg/ml after 1 h, whereas in the present study of mouse muscle, IL-6 concentrations were undetectable at this time point. The discrepancy between these studies may be related to the longer (2 h) muscle incubation period used in the present study (27). Because the basal IL-6 levels were relatively high, we speculate that a Ca 2 -mediated autocrine regulation by IL-6 may directly contribute to the increase in basal IL-6 production (30), thereby masking an effect of ionomycin. Basal IL-6 release from isolated soleus muscle increases in an exponential-like fashion over time. We speculate that autocrine stimulation by IL-6 contributes to in vitro muscle IL-6 release, a process that is also implicated in skeletal muscle IL-6 production during later stages of exercise (30). Exercise is associated with increased skeletal muscle IL-6 release as well as increased AMPK activity. Paradoxically, we and others (19) provide evidence that AICAR exposure of incubated isolated skeletal muscle results in diminished IL-6 release and mrna production (19), despite pronounced activation of AMPK. Interestingly, AICAR stimulation resulted in an increase in IL-6 mrna production in C2C12 cells, which could be prevented by knockdown of AMPK 1 and - 2 isoforms using small interfering RNA (8). In mouse cardiac fibroblasts, exposure to either AICAR or metformin increases IL-6 protein release (18), providing evidence for a role of AMPK in IL-6 production. In contrast, AICAR blocks IL-6 protein release from human skeletal muscle cells (19) as well as human adipose tissue or adipocytes in vitro (12, 14). Similarly, we show AICAR stimulation prevents IL-6 release and mrna production from isolated mouse skeletal muscle in vitro. The discrepancy between

6 Endocrinology, February 2009, 150(2): endo.endojournals.org 605 the effects of AICAR on IL-6 release in some cultured cell systems vs. isolated tissues is unapparent but may be related to the cell type or differentiation state. Furthermore, AMPK is a heterotrimeric molecule with up to 12 possible combinations of subunit isoforms, and the tissue expression or cellular localization of a particular AMPK heterotrimer might be essential to elicit specific downstream responses. To investigate whether basal IL-6 release as well as the AICAR suppression of IL-6 release and mrna production depends on functional AMPK, we used transgenic and knockout mouse models with impaired AMPK signaling in skeletal muscle. AMPK-dependent metabolic and mitogenic responses in skeletal muscle from AMPK 2 KD and AMPK 3 KO mice are partly suppressed or completely abolished (35). We provide evidence implicating AMPK in the regulation of basal IL-6 release from oxidative skeletal muscle, which was elevated in AMPK 2 KD and AMPK 1 KO mice compared with their respective wild-type littermates. In vitro exposure of human skeletal myotubes to IL-6 alters gene transcription and promotes myotube fusion and formation (15); thus, the modest increase in IL-6 release may contribute to changes in the transcriptional profile of skeletal muscle from AMPK 2 KD mice (21). Previous evidence suggests that IL-6 activates AMPK (11, 15, 16). Mice lacking IL-6 display diminished AMPK phosphorylation in skeletal muscle and adipose tissue at rest and after exercise (11). Resting serum IL-6 levels in wild-type mice are less than 10 pg/ml (Glund, S., and J. R. Zierath, unpublished data), whereas the media concentration of IL-6 in the present study was markedly increased. This raises the possibility that autocrine stimulation of IL-6 production (7) has contributed to the increased basal IL-6 release as observed in the present study. Moreover, basal IL-6 release was increased in soleus, but not EDL, from transgenic and knockout mice with altered AMPK signaling. Consequently, our results may indicate a feedback mechanism in which IL-6 blunts its own release from oxidative skeletal muscle through an AMPK-dependent pathway. In contrast to results obtained for AMPK 2 KD and AMPK 1 KO mice, basal IL-6 release from soleus of AMPK 3 KO mice is unaltered. This may be explained by the finding that AMPK subunit expression in this KO mouse model is unaltered (29), despite the lack of AMPK 3. Furthermore, whereas AMPK subunits are expressed in slow-twitch oxidative skeletal muscle, AMPK 3 mrna and protein are undetectable (29). Our results therefore provide evidence to suggest the AMPK isoforms are necessary for the regulation of basal IL-6 release from oxidative muscle, whereas the 3 isoform is dispensable. AICAR suppression of IL-6 release was similar in skeletal muscle from AMPK 2 KD and AMPK 1 KO mice, compared with the respective wild-type littermates. AMPK 2 KD mice retain 10 20% of the total AMPK 2 activity and 30 40% of total AMPK 1 activity in skeletal muscle (28). Moreover, AMPK 1 KO mice display a 1.6-fold increase in AMPK 2 expression (22). Because there is some residual AMPK activity in all of the mouse models used in this study, our results cannot completely exclude the involvement of AMPK in the AICAR-mediated suppression of IL-6 release. Exercise is associated with increased activity of heterotrimers in human skeletal muscle (36), and exercise-mediated AMPK 2 activity correlates with IL-6 release (13). Taken together, our findings implicate the AMPK 1 isoform in the regulation of basal IL-6 release from muscle. To extend our findings with AICAR, we also used the direct AMPK activator A as an alternative means to modulate AMPK activity. Exposure to A was associated with a suppression of IL-6 release from soleus but not EDL muscle. Furthermore, the A mediated suppression of IL-6 release from soleus muscle was less pronounced compared with the effect of AICAR. This was surprising, given that both AICAR and A are potent AMPK activators. Importantly, IL-6 release from oxidative muscle was increased in mouse models with impaired AMPK signaling, and activation of AMPK with A reduced IL-6 release from oxidative soleus muscle. The effect of AICAR to suppress IL-6 release was greater compared with A Moreover, AICAR, but not A , suppressed IL-6 release from glycolytic EDL muscle. Taken together, our data implicate AMPK in the regulation of IL-6 release from oxidative muscle tissue. AMPK-independent effects are likely to contribute to the effects of AICAR on IL-6 release. In conclusion, using transgenic and knockout mouse models to perturb AMPK signaling, we provide evidence that AMPKdependent pathways regulate IL-6 release from isolated oxidative skeletal muscle. Moreover, we provide evidence that in vitro exposure of mouse skeletal muscle to AICAR reduces IL-6 mrna expression and protein release, at least in part via AMPKindependent pathways. The local production and release of IL-6 may provide an additional level of metabolic and mitogenic regulation in skeletal muscle. Acknowledgments We thank Dr. Morris J. Birnbaum (Howard Hughes Medical Institute and The Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA) for the kind donation of the AMPK KD mice. We also thank Dr. Kei Sakamoto (MRC Protein Phosphorylation Unit, College of Life Sciences, University of Dundee, Dundee, Scotland) for kindly providing the A compound. We thank Dr. Anna Krook for valuable comments during the preparation of this manuscript. Address all correspondence and requests for reprints to: Juleen R. Zierath, Department of Molecular Medicine and Surgery, Section for Integrative Physiology, Karolinska Institutet, von Eulers väg 4, 4th Floor, S Stockholm, Sweden. juleen.zierath@ki.se. This work was supported by grants from the Swedish Research Council, the Swedish Diabetes Association, the Foundation for Scientific Studies of Diabetology, the Swedish Centre for Sports Research, the Danish Medical Research Council, the Novo Nordisk Foundation, the Danish Diabetes Association, the Copenhagen Muscle Research Centre, the Strategic Research Foundation (INGVAR II), the Commission of the European Communities (Contracts LSHM-CT EXGENESIS and LSHM-CT EUGENE2). Current work address for S.G.: Boehringer Ingelheim Pharma GmbH & Co. KG, Metabolic Diseases Research, Biberach an der Riss, Germany. Disclosure Statement: The authors have nothing to disclose.

7 606 Glund et al. Role of AMPK in Muscle IL-6 Release Endocrinology, February 2009, 150(2): References 1. Febbraio MA, Pedersen BK 2005 Contraction-induced myokine production and release: is skeletal muscle an endocrine organ? Exerc Sport Sci Rev 33: Ostrowski K, Rohde T, Zacho M, Asp S, Pedersen BK 1998 Evidence that interleukin-6 is produced in human skeletal muscle during prolonged running. J Physiol 508: Pedersen BK, Fischer CP 2007 Beneficial health effects of exercise the role of IL-6 as a myokine. Trends Pharmacol Sci 28: Banzet S, Koulmann N, Sanchez H, Serrurier B, Peinnequin A, Alonso A, Bigard X 2007 Contraction-induced interleukin-6 transcription in rat slowtype muscle is partly dependent on calcineurin activation. J Cell Physiol 210: Keller C, Hellsten Y, Steensberg A, Pedersen BK 2006 Differential regulation of IL-6 and TNF- via calcineurin in human skeletal muscle cells. Cytokine 36: Kosmidou I, Vassilakopoulos T, Xagorari A, Zakynthinos S, Papapetropoulos A, Roussos C 2002 Production of interleukin-6 by skeletal myotubes: role of reactive oxygen species. Am J Respir Cell Mol Biol 26: Keller P, Keller C, Carey AL, Jauffred S, Fischer CP, Steensberg A, Pedersen BK 2003 Interleukin-6 production by contracting human skeletal muscle: autocrine regulation by IL-6. Biochem Biophys Res Commun 310: Weigert C, Duefer M, Simon P, Debre E, Runge H, Brodbeck K, Haering HU, Schleicher ED 2007 Upregulation of IL-6 mrna by IL-6 in skeletal muscle cells: role of IL-6 mrna stabilization and Ca2 -dependent mechanisms. Am J Physiol Cell Physiol 293:C1139 C Chan MH, McGee SL, Watt MJ, Hargreaves M, Febbraio MA 2004 Altering dietary nutrient intake that reduces glycogen content leads to phosphorylation of nuclear p38 MAP kinase in human skeletal muscle: association with IL-6 gene transcription during contraction. FASEB J 18: Steensberg A, Febbraio MA, Osada T, Schjerling P, van Hall G, Saltin B, Pedersen BK 2001 Interleukin-6 production in contracting human skeletal muscle is influenced by pre-exercise muscle glycogen content. J Physiol 537: Kelly M, Keller C, Avilucea PR, Keller P, Luo Z, Xiang X, Giralt M, Hidalgo J, Saha AK, Pedersen BK, Ruderman NB 2004 AMPK activity is diminished in tissues of IL-6 knockout mice: the effect of exercise. Biochem Biophys Res Commun 320: Lihn AS, Jessen N, Pedersen SB, Lund S, Richelsen B 2004 AICAR stimulates adiponectin and inhibits cytokines in adipose tissue. Biochem Biophys Res Commun 316: MacDonald C, Wojtaszewski JF, Pedersen BK, Kiens B, Richter EA 2003 Interleukin-6 release from human skeletal muscle during exercise: relation to AMPK activity. J Appl Physiol 95: Sell H, Dietze-Schroeder D, Eckardt K, Eckel J 2006 Cytokine secretion by human adipocytes is differentially regulated by adiponectin, AICAR, and troglitazone. Biochem Biophys Res Commun 343: Al-Khalili L, Bouzakri K, Glund S, Lönnqvist F, Koistinen HA, Krook A 2006 Signaling specificity of interleukin-6 action on glucose and lipid metabolism in skeletal muscle. Mol Endocrinol 20: Glund S, Deshmukh A, Long YC, Moller T, Koistinen HA, Caidahl K, Zierath JR, Krook A 2007 Interleukin-6 directly increases glucose metabolism in resting human skeletal muscle. Diabetes 56: Hardie DG, Hawley SA, Scott JW 2006 AMP-activated protein kinase development of the energy sensor concept. J Physiol 574: Du JH, Xu N, Song Y, Xu M, Lu ZZ, Han C, Zhang YY 2005 AICAR stimulates IL-6 production via p38 MAPK in cardiac fibroblasts in adult mice: a possible role for AMPK. Biochem Biophys Res Commun 337: Lihn AS, Pedersen SB, Lund S, Richelsen B 2008 The anti-diabetic AMPK activator AICAR reduces IL-6 and IL-8 in human adipose tissue and skeletal muscle cells. Mol Cell Endocrinol 292: Mu J, Brozinick Jr JT, Valladares O, Bucan M, Birnbaum MJ 2001 A role for AMP-activated protein kinase in contraction- and hypoxia-regulated glucose transport in skeletal muscle. Mol Cell 7: Mu J, Barton ER, Birnbaum MJ 2003 Selective suppression of AMP-activated protein kinase in skeletal muscle: update on lazy mice. Biochem Soc Trans 31: Jorgensen SB, Viollet B, Andreelli F, Frosig C, Birk JB, Schjerling P, Vaulont S, Richter EA, Wojtaszewski JF 2004 Knockout of the 2 but not 15 -AMPactivated protein kinase isoform abolishes 5-aminoimidazole-4-carboxamide ribofuranosidebut not contraction-induced glucose uptake in skeletal muscle. J Biol Chem 279: Barnes BR, Marklund S, Steiler TL, Walter M, Hjalm G, Amarger V, Mahlapuu M, Leng Y, Johansson C, Galuska D, Lindgren K, Abrink M, Stapleton D, Zierath JR, Andersson L 2004 The 5 -AMP-activated protein kinase 3 isoform has a key role in carbohydrate and lipid metabolism in glycolytic skeletal muscle. J Biol Chem 279: Corton JM, Gillespie JG, Hawley SA, Hardie DG Aminoimidazole- 4-carboxamide ribonucleoside. A specific method for activating AMP-activated protein kinase in intact cells? Eur J Biochem 229: Goransson O, McBride A, Hawley SA, Ross FA, Shpiro N, Foretz M, Viollet B, Hardie DG, Sakamoto K 2007 Mechanism of action of A , a valuable tool for activation of AMP-activated protein kinase. J Biol Chem 282: Akamine R, Yamamoto T, Watanabe M, Yamazaki N, Kataoka M, Ishikawa M, Ooie T, Baba Y, Shinohara Y 2007 Usefulness of the 5 region of the cdna encoding acidic ribosomal phosphoprotein P0 conserved among rats, mice, and humans as a standard probe for gene expression analysis in different tissues and animal species. J Biochem Biophys Methods 70: Holmes AG, Watt MJ, Carey AL, Febbraio MA 2004 Ionomycin, but not physiologic doses of epinephrine, stimulates skeletal muscle interleukin-6 mrna expression and protein release. Metabolism 53: Jensen TE, Rose AJ, Jorgensen SB, Brandt N, Schjerling P, Wojtaszewski JF, Richter EA 2007 Possible CaMKK-dependent regulation of AMPK phosphorylation and glucose uptake at the onset of mild tetanic skeletal muscle contraction. Am J Physiol Endocrinol Metab 292:E1308 E Mahlapuu M, Johansson C, Lindgren K, Hjalm G, Barnes BR, Krook A, Zierath JR, Andersson L, Marklund S 2004 Expression profiling of the -subunit isoforms of AMP-activated protein kinase suggests a major role for 3in white skeletal muscle. Am J Physiol Endocrinol Metab 286:E194 E Fischer CP 2006 Interleukin-6 in acute exercise and training: what is the biological relevance? Exerc Immunol Rev 12: Banzet S, Koulmann N, Simler N, Birot O, Sanchez H, Chapot R, Peinnequin A, Bigard X 2005 Fibre-type specificity of interleukin-6 gene transcription during muscle contraction in rat: association with calcineurin activity. J Physiol 566: James DE, Jenkins AB, Kraegen EW 1985 Heterogeneity of insulin action in individual muscles in vivo: euglycemic clamp studies in rats. Am J Physiol 248:E567 E Serrano AL, Baeza-Raja B, Perdiguero E, Jardi M, Munoz-Canoves P 2008 Interleukin-6 is an essential regulator of satellite cell-mediated skeletal muscle hypertrophy. Cell Metab 7: Gregorevic P, Meznarich NA, Blankinship MJ, Crawford RW, Chamberlain JS 2008 Fluorophore-labeled myosin-specific antibodies simplify muscle-fiber phenotyping. Muscle Nerve 37: Long YC, Zierath JR 2006 AMP-activated protein kinase signaling in metabolic regulation. J Clin Invest 116: Birk JB, Wojtaszewski JF 2006 Predominant 2/ 2/ 3 AMPK activation during exercise in human skeletal muscle. J Physiol 577: