Prior AICAR Stimulation Increases Insulin Sensitivity in Mouse Skeletal Muscle in an AMPK-Dependent Manner

2042 Diabetes Volume 64, June 2015

Rasmus Kjøbsted,1,2 Jonas T. Treebak,1,2 Joachim Fentz,1 Louise Lantier,3,4,5 Benoit Viollet,3,4,5 Jesper B. Birk,1 Peter Schjerling,6 Marie Björnholm,7 Juleen R. Zierath,2,7 and Jørgen F.P. Wojtaszewski1

Diabetes 2015;64:2042–2055 | DOI: 10.2337/db14-1402

An acute bout of exercise increases glucose uptake in rodents (1–6) and may persist for up to 48 h after exercise, skeletal muscle by an insulin-independent mechanism. depending on carbohydrate availability (7–9). Improved mus- In the period after exercise, insulin sensitivity to in- cle insulin sensitivity postexercise is mediated by one or creased glucose uptake is enhanced. The molecular several local contraction-induced mechanisms (10) involving mechanisms underpinning this phenomenon are poorly both enhanced transport and intracellular processing of glu- understood but appear to involve an increased cell cose. This period is characterized by increased GLUT4 protein surface abundance of GLUT4. While increased proximal abundance at the plasma membraneandenhancedglycogen insulin signaling does not seem to mediate this effect, synthase activity (11,12). These changes occur independent elevated phosphorylation of TBC1D4, a downstream target of both insulin (Akt) and exercise (AMPK) signal- of global protein synthesis (13), including both total GLUT4 ing, appears to play a role. The main purpose of this and glycogen synthase protein content (4,11), and are inde- study was to determine whether AMPK activation pendent of changes in proximal insulin signaling, including increases skeletal muscle insulin sensitivity. We found Akt activation (3,4,13–17). that prior AICAR stimulation of wild-type mouse muscle AMPK is a heterotrimeric complex consisting of catalytic increases insulin sensitivity to stimulate glucose uptake. (a1/a2) and regulatory subunits (b1/b2andg1/g2/g3). Of However, this was not observed in mice with reduced or the 12 heterotrimeric combinations, only 3 and 5 combina- SIGNAL TRANSDUCTION ablated AMPK activity in skeletal muscle. Furthermore, tions have been found in the skeletal muscle of human and prior AICAR stimulation enhanced insulin-stimulated mouse, respectively (18,19). AMPK is activated in response 649 711 phosphorylation of TBC1D4 at Thr and Ser in to various stimuli that increase cellular energy stress (e.g., wild-type muscle only. These phosphorylation events metformin, hypoxia, hyperosmolarity, muscle contraction, were positively correlated with glucose uptake. Our and exercise) (20). With energy stress, intracellular concen- results provide evidence to support that AMPK activa- trations of AMP and ADP accumulate. This activates AMPK tion is sufﬁcient to increase skeletal muscle insulin sen- allosterically and decreases the ability of upstream phospha- sitivity. Moreover, TBC1D4 phosphorylation may tases to dephosphorylate Thr172, which further increases facilitate the effect of prior AMPK activation to enhance glucose uptake in response to insulin. AMPK phosphorylation and activity (21). Like exercise, AICAR increases AMPK activity in skeletal muscle (22), which partly mimics the metabolic changes observed dur- The effect of insulin on skeletal muscle glucose uptake is ing muscle contraction (23). increased in the period after a single bout of exercise. This TBC1D4 is involved in insulin-stimulated glucose phenomenon is observed in muscle from both humans and transport in skeletal muscle (24) and is regulated via

1Section of Molecular Physiology, August Krogh Centre, Department of Nutrition, 7Integrative Physiology, Department of Molecular Medicine and Surgery, Karolinska Exercise and Sports, University of Copenhagen, Copenhagen, Denmark Institutet, Stockholm, Sweden 2 The Novo Nordisk Foundation Center for Basic Metabolic Research, Section of Corresponding author: Jørgen F.P. Wojtaszewski, [email protected]. Integrative Physiology, University of Copenhagen, Copenhagen, Denmark Received 11 September 2014 and accepted 20 December 2014. 3INSERM, U1016, Institut Cochin, Paris, France 4CNRS, UMR8104, Paris, France R.K. and J.T.T. share first authorship. 5Université Paris Descartes, Sorbonne Paris Cité, Paris, France © 2015 by the American Diabetes Association. Readers may use this article as 6Institute of Sports Medicine, Department of Orthopedic Surgery, Bispebjerg long as the work is properly cited, the use is educational and not for profit, and Hospital and Center for Healthy Aging, Faculty of Health Sciences, University of the work is not altered. Copenhagen, Copenhagen, Denmark See accompanying article, p. 1901. diabetes.diabetesjournals.org Kjøbsted and Associates 2043 phosphorylation at multiple sites by Akt (25), thereby and extensor digitorum longus (EDL) muscles were increasing translocation of GLUT4 to the plasma mem- dissected and suspended in incubation chambers (Multi brane. AMPK also targets TBC1D4; however, this does not Wire Myograph System; DMT, Aarhus, Denmark) con- seem to directly affect glucose uptake (26). As insulin taining Krebs-Ringer buffer (KRB) (117 mmol/L NaCl, (Akt) and exercise/AICAR (AMPK) signaling pathways 4.7 mmol/L KCl, 2.5 mmol/L CaCl2, 1.2 mmol/L converge on TBC1D4, this may explain how exercise KH2PO4, 1.2 mmol/L MgSO4, 0.5 mmol/L NaHCO3, modulates insulin action to regulate glucose transport in pH 7.4) supplemented with 0.1% BSA, 8 mmol/L man- skeletal muscle. Supporting this concept, TBC1D4 phos- nitol, and 2 mmol/L pyruvate. During the entire incuba- phorylation is elevated in skeletal muscle several hours tion period, the buffer was oxygenated with 95% O2 and after an acute bout of exercise in both rodents and 5% CO2, and maintained at 30°C. After 10 min of pre- humans, concomitant with increased insulin sensitivity incubation, muscles were incubated for 50 min in the to stimulate glucose uptake in the postexercise period absence or presence of 1 mmol/L AICAR (Toronto Re- (15,16,27–30). search Chemicals, Toronto, Ontario, Canada) in 100% Prior AICAR stimulation increases skeletal muscle human serum from healthy overnight-fasted men. The insulin sensitivity (13). However, because AICAR exerts use of serum is necessary to elicit an effect of AICAR on multiple AMPK-independent effects (31), the direct re- muscle insulin sensitivity (13). Soleus and EDL muscles lationship between AMPK and muscle insulin sensitivity were allowed to recover in the absence of AICAR in has not been established. Thus, the primary purpose of modified KRB supplemented with 5 mmol/L glucose, 5 the current study was to determine whether AMPK di- mmol/L mannitol, and 0.1% BSA for 4 h (soleus muscle) rectly regulates skeletal muscle insulin sensitivity on or 6 h (EDL muscle). During recovery, the medium was glucose uptake. We established an ex vivo protocol using replaced once every hour to maintain an adequate glu- mouse muscle to study insulin sensitivity after prior cose concentration. Subsequently, paired muscles from AICAR stimulation and tested the hypothesis that each animal were incubated for 30 min in KRB in the AMPK is necessary for the effect of AICAR to enhance absence or presence of a submaximal concentration insulin sensitivity. Furthermore, we evaluated TBC1D4 (100 mU/mL) of insulin (Actrapid; Novo Nordisk, Bagsvaerd, phosphorylation status because this protein is a conver- Denmark). The uptake of 2-deoxyglucose was measured gence point for insulin- and exercise-mediated signaling during the last 10 min of the 30-min period by adding events. 1 mmol/L [3H]2-deoxyglucose (0.056 MBq/mL) and 7 mmol/L [14C]mannitol (0.0167 MBq/mL) to the incu- RESEARCH DESIGN AND METHODS bation medium. After incubation, muscles were harvested, fi Animals/Humans washed in ice-cold KRB, quickly dried on lter paper, and All experiments were approved by the Danish Animal frozen in liquid nitrogen. Experimental Inspectorate and the regional animal ethics committee of Northern Stockholm and complied with Muscle Processing the European Union Convention for the Protection of Muscles were homogenized in 400 mLofice-cold Vertebrate Animals Used for Scientific Purposes (Council buffer (10% glycerol, 20 mmol/L sodium pyrophos- of Europe 123, Strasbourg, France, 1985). Except for phate, 1% NP-40, 2 mmol/L phenylmethylsulfonyl fl the wild-type (WT) mice (C57BL/6J; Taconic, Ejby, uoride [PMSF], 150 mmol/L sodium chloride, 50 Denmark) used in Figs. 1, 3E, and 8, the animals used mmol/L HEPES, 20 mmol/L b-glycerophosphate, 10 fi mmol/L sodium fluoride, 1 mmol/L EDTA, 1 mmol/L in this study were muscle-speci c kinase-dead a2-AMPK fi EGTA, 10 mg/mL aprotinin, 3 mmol/L benzamidine, (AMPK KD) (32), muscle-speci c a2- and a1-AMPK double-knockout (AMPK mdKO) (33), and g -AMPK KO 10 mg/mL leupeptin, and 2 mmol/L sodium orthova- 3 3 mice (34) with corresponding WT littermates used as nadate, pH 7.5) for 2 30 s at 30 Hz using steel beads controls. All mice in this study were female (mean weight and a TissueLyzer II (QIAGEN, Hilden, Germany). 24.3 6 0.2 g) and were maintained on a 12:12 light-dark Homogenates were rotated end over end for 1 h before cycle (6:00 A.M. to 6:00 P.M.) with unlimited access to centrifugation at 16,000g for 20 min. The supernatant standard rodent chow and water. Serum was obtained (lysate) was collected, frozen in liquid nitrogen, and 2 from healthy young men in accordance with a protocol stored at 80°C for later analyses. approved by the Ethics Committee of Copenhagen (protocol #H-3–2012–140) and complied with the ethical Glucose Uptake Measurements 3 guidelines of the Declaration of Helsinki II. Informed con- Glucose uptake was assessed by the accumulation of [ H] 14 sent was obtained from all participating subjects before 2-deoxyglucose in muscle with the use of [ C]mannitol they entered the study. (PerkinElmer, Waltham, MA) as an extracellular marker. Radioactivity was measured in 250 mL of lysate by liquid Muscle Incubations scintillation counting (Ultima Gold and Tri-Carb 2910 TR; Fed animals were anesthetized by intraperitoneal injec- PerkinElmer) and was related to the specific activity of the tion of pentobarbital (10 mg/100 g body wt) before soleus incubation buffer. 2044 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015

Figure 1—Glucose uptake (A) and phospho-AMPK (pAMPK) Thr172 (B) in soleus and EDL muscles in response to acute AICAR stimulation (50 min, 0.5–4 mmol/L) (n =7–8). †P < 0.05 vs. control (0 mmol/L) within muscle type. C: Glucose uptake in soleus and EDL muscles after 4 and 6 h of recovery from prior AICAR treatment (50 min, 1 mmol/L), respectively (n = 8). D: Glucose uptake in EDL muscle incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (n = 24). †P < 0.05 vs. basal control; *P < 0.05 vs. basal value within group; #P < 0.05 vs. response to insulin in control (interaction: insulin 3 AICAR). E: Glucose uptake in soleus muscle incubated with or without insulin (100 mU/mL) 4 h after prior AICAR stimulation (50 min, 1 mmol/L) (n = 8). *P < 0.001, main effect of insulin. Data were analyzed by one-way ANOVA (A and B), paired t test (C), or two-way repeated-measures ANOVA (D and E). Data are expressed as the means 6 SEM. F: Representative Western blot image. AU, arbitrary units; hr, hour; SOL, soleus.

SDS-PAGE and Western Blot Analyses All membranes were stripped with buffer (100 mmol/L Total protein abundance in muscle lysates was determined 2-mecaptoethanol, 2% SDS, 62.5 mmol/L Tris-HCl, pH by the bicinchoninic acid method (ThermoFisher Scientific, 6.7) and reprobed with new primary antibodies for the de- Waltham, MA). Muscle lysates were prepared in Laemmli tection of other phosphorylation sites on identical proteins buffer and heated for 10 min at 96°C. Equal amounts of or the corresponding total proteins. The stripping proce- protein were separated by SDS-PAGE on 5% or 7% self-cast dure was verified by reincubating membranes with second- gels and transferred to polyvinylidene fluoride membranes ary antibodies for the detection of primary antibodies that using semidry blotting. Membranes were blocked for 5–10 were possibly still bound. min in 2% skim milk or 3% BSA and probed with primary and secondary antibodies. Proteins with bound antibody Antibodies were visualized with chemiluminescence (Millipore) using The following antibodies were from Cell Signaling Tech- a digital imaging system (ChemiDoc MP System; Bio-Rad). nology (Danvers, MA): anti–phospho-AMPK-Thr172 (catalog diabetes.diabetesjournals.org Kjøbsted and Associates 2045

#2531), anti–phospho-acetyl-CoA carboxylase (ACC) Ser79 were subsequently subcloned into a p3xflag-cmv-9–10 vec- (catalog #3661), anti-Akt2 (D6G4) (catalog #3063), tor using NotIandKpnI cloning sites before amplification anti–phospho-Akt-Thr308 (catalog #9275), anti–phospho- in Escherichia coli TOP10 cells (Invitrogen). Plasmid DNA Akt-Ser473 (catalog #9271), anti–phospho-TBC1D1-Thr590 was extracted using an endotoxin-free Plasmid Mega Kit (catalog #6927), anti–phospho-TBC1D4-Ser318 (catalog #8619), (QIAGEN) and was diluted in isotonic saline solution to anti–phospho-TBC1D4-Ser588 (#8730), and anti–phospho- a final concentration of 2 mg/mL. DNA (50 mg) was injected TBC1D4-Thr642 (catalog #8881). Anti–DYKDDDDK-Tag into the tibialis anterior muscle 2 h after treatment (FLAG-Tag) (catalog #F1804; Sigma-Aldrich), anti–phospho- with hyaluronidase (Sigma-Aldrich) (one injection of 30 TBC1D1-Ser237 (catalog #2061452; Millipore), anti-TBC1D1 units/muscle, 1 unit/mL), and gene electrotransfer was as previously described (35), anti-AS160 (TBC1D4) (cata- performed as previously described (24). Seven days after log #07–741; Millipore), anti–phospho-TBC1D4-Ser711 as gene electrotransfer, phosphorylation of TBC1D4 Thr649 previously described (26), and anti–AMPK-a2 (catalog and Ser711 was assessed in the tibialis anterior muscle of #SC-19131; Santa Cruz Biotechnology). The antibodies anesthetized (8 mg pentobarbital/100 g body wt) animals used for AMPK activity measurements were anti–AMPK- in response to retro-orbital injection of either saline so- g3, anti–AMPK-a1, and anti–AMPK-a2, all of which were lution or insulin (10 units/kg). Ten minutes after injec- provided by Professor D.G. Hardie (University of Dundee, tion, the tibialis anterior muscle was removed, quickly Dundee, Scotland, U.K.). frozen in liquid nitrogen, and stored at 280°C for subsequent analysis. AMPK Activity Assay Five different AMPK trimer complexes have been detected Statistics in mouse skeletal muscle: a2b2g3, a2b1g1, a2b2g1, Statistical analyses were performed using SigmaPlot a1b1g1, and a1b2g1 (19). a2b2g3-AMPK activity was version 11.0 (SYSTAT, Erkrath, Germany) and SPSS measured on g3-AMPK immunoprecipitates (IPs) from version 20 (IBM) software. SPSS version 20 was used for 300 mg of muscle lysate using AMPK-g3 antibody, three-way ANOVA with repeated measures, while all – G-protein coupled agarose beads (Millipore) and IP buffer other analyses were performed using SigmaPlot version fl (50 mmol/L NaCl, 1% Triton X-100, 50 mmol/L sodium uo- 11.0. Data are presented as the mean 6 SEM. One-, ride, 5 mmol/L sodium-pyrophosphate, 20 mmol/L Tris- two-, or three-way ANOVAs with or without repeated base, pH 7.5, 500 mmol/L PMSF, 2 mmol/L dithiothreitol, measures was used to assess statistical differences, 4 mg/mL leupeptin, 50 mg/mL soybean trypsin inhibitor, 6 where appropriate. When a three-way interaction oc- mmol/L benzamidine, and 250 mmol/L sucrose). Samples curred (P , 0.05; genotype 3 AICAR 3 insulin), were treated as previously described (19,36). In short, a two-way ANOVA with repeated measures was used after overnight end-over-end rotation at 4°C, IPs were on each genotype (WT and KD or WT and mdKO) in centrifuged for 1 min at 2,000g and washed once in IP order to determine the site of interaction between 3 buffer, once in 6 assay buffer (240 mmol/L HEPES, 480 AICAR and insulin (P , 0.05; AICAR 3 insulin). Any 3 mmol/L NaCl, pH 7.0), and twice in 3 assay buffer (1:1). main effects of genotype are included in the figure The activity assay was performed for 30 min at 30°C in legends. For post hoc testing, a Student-Newman-Keuls a total volume of 30 mL of kinase mix (40 mmol/L HEPES, test was used. Correlation analyses were performed by 80 mmol/L NaCl, 833 mmol/L dithiothreitol, 200 mmol/L determination of Pearson product moment correlation AMP, 100 mmol/L AMARA peptide, 5 mmol/L MgCl2, coefficient. Differences were considered statistically sig- 200 mmol/L ATP, and 2 mCi of [g-33P]-ATP; PerkinElmer). nificant at P , 0.05. The reaction was terminated by adding 10 mLof1%phosphoric acid. Twenty microliters of the reaction mix were RESULTS spotted on P81 filter paper. Filter papers were subsequently 3 Prior AICAR Stimulation Increases Insulin Sensitivity in washed 4 15 min in 1% phosphoric acid. 33P radioactiv- EDL Muscle but Not in Soleus Muscle ity was analyzed on dried filter paper using a Storm 850 Acute AICAR stimulation increased glucose uptake and PhosphorImager (Molecular Dynamics). The combined ac- AMPK phosphorylation in both soleus and EDL muscles tivity of a2b1g1anda2b2g1 was measured on superna- (Fig. 1A, B,andF). We then determined the time point tants from the g3-AMPK IPs using the AMPK-a2 antibody at which glucose uptake had reversed to basal levels in for a second IP, and the combined activity of a1b1g1and order to evaluate the effect of AICAR on insulin sensi- a b g a 1 2 1 was measured on supernatants from the 2-AMPK tivity. In WT soleus and EDL muscle, glucose uptake IPs using a1-AMPK antibody for a third IP. reversed to basal levels after 4 and 6 h of recovery In Vivo Gene Electrotransfer from AICAR stimulation, respectively (Fig. 1C). Prior TBC1D4 WT and TBC1D4 T649A and S711A DNA mutant AICAR treatment increased the effect of a submaximal constructs, containing T-to-A and S-to-A point mutations, insulin concentration (100 mU/mL) to stimulate glucose respectively, were commercially and individually synthesized uptake in the EDL muscle, but not in the soleus muscle of from the gene encoding mouse TBC1D4 (GeneArt; Life WT mice (Fig. 1D and E). Based on these results, we chose Technologies, Darmstadt, Germany). All three constructs to use only EDL muscle for subsequent experiments. 2046 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015

Prior AICAR Stimulation Increases Muscle Insulin AICAR treatment increased ACC phosphorylation to Sensitivity in an AMPK-Dependent Manner a greater extent in muscle from WT littermates than in In order to clarify whether the effect of AICAR on insulin muscle from both transgenic models (although the increase sensitivity is dependent on AMPK, we took advantage of was significant only in WT mice from the mdKO model), the AMPK KD and AMPK mdKO mouse models in which indicating a maintained effect of prior AICAR stimulation AMPK activity is decreased or ablated in skeletal muscle. on AMPK in muscle cells. Therefore, we measured AMPK Prior AICAR stimulation increased insulin sensitivity in activity in WT EDL muscle that had been previously stim- isolated EDL muscle from WT littermates but failed to ulated with AICAR. The combined activity of a1b1g1and increase insulin sensitivity in both transgenic models (Fig. a1b2g1was;1.4-fold higher compared with unstimulated 2A and B). The incremental increase in insulin-stimulated control muscle (P = 0.037), while a2b2g3activitywas glucose uptake (glucose uptake with insulin minus basal ;2.3-fold higher (P , 0.001) (Fig. 3E). In contrast, the glucose uptake) was significantly higher after prior AICAR combined activity of AMPK trimer complexes a2b2g1and stimulation in WT littermates only (Fig. 2C and D). a2b1g1 was unchanged by prior AICAR treatment. This AMPK Activity and Signaling indicates a persistent effect of prior AICAR stimulation on fi As AICAR acutely increases phosphorylation of AMPK and speci c AMPK trimer activity in particular a2b2g3 activity. the downstream target ACC, we investigated whether this Prior AICAR Stimulation Increases Muscle Insulin effect was maintained into recovery. Phosphorylation of Sensitivity in an AMPK-g3–Dependent Manner AMPK and ACC was increased in EDL muscle previously A persistent increase in AMPK a2b2g3 activity after stimulated with AICAR independent of genotype (Fig. 3A– AICAR stimulation prompted us to test the hypothesis D, H,andI). We assume that the observed increase in ACC that the effect of AICAR to enhance muscle insulin sensi- phosphorylation in muscle from both transgenic mouse tivity is mediated through the AMPK a2b2g3 trimer com- models after prior AICAR treatment corresponds to plex. Indeed, prior AICAR stimulation failed to increase AMPK-independent effects of AICAR on ACC phosphoryla- muscle insulin sensitivity in whole-body g3-AMPK KO tion or AMPK activation in nonmuscle cells. However, prior mice (Fig. 3F and G). For unknown reasons, prior AICAR

Figure 2—Glucoseuptake(A and B)orD glucose uptake (insulin minus basal) (C and D) in EDL muscle from either AMPK KD mice (A and C)orAMPKmdKOmice(B and D) and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (KD n =9–10, mdKO n = 12). Data are expressed as the means 6 SEM. A:AICAR3 insulin 3 genotype interaction (P < 0.001). *P < 0.05 vs. basal value within genotype; #P < 0.001 vs. response to insulin in WT control mice (interaction AICAR 3 insulin). B:AICAR3 insulin 3 genotype interaction (P < 0.05). *P < 0.001 vs. basal value within genotype; #P < 0.001 vs. response to insulin in WT control (interaction AICAR 3 Insulin). C: Data are extracted from the raw data given in A. †P < 0.001 vs. control within genotype. D: Data are extracted from the raw data given in B. †P < 0.05 vs. control within genotype. hr, hour. diabetes.diabetesjournals.org Kjøbsted and Associates 2047

Figure 3—Phospho-AMPK (pAMPK) Thr172 (A and C) and phospho-ACC (pACC) Ser212 (B and D) in EDL muscle from either AMPK KD mice (A and B) or AMPK mdKO mice (C and D) and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (KD n =9–10, mdKO n = 12). AMPK trimer specific activity in EDL muscle from C57BL/6J mice incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (N =4–6) (E). Glucose uptake (F)orD insulin-stimulated glucose uptake (insulin minus basal) (G) in EDL muscle from g3-AMPK KO mice and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (n =6–8). Data are expressed as the means 6 SEM. A: †Main effect of AICAR (P < 0.05). Main effect of genotype (P < 0.001). B: †Main effect of AICAR (P < 0.01). Main effect of genotype (P < 0.001). C: AICAR 3 insulin 3 genotype interaction (P < 0.001). †P < 0.001 vs. control within genotype; #P < 0.001 vs. response to AICAR in WT basal (interaction AICAR 3 insulin). D: †P < 0.01 vs. control value within genotype; ‡P < 0.001 vs. response in 2048 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015 treatment still affected basal glucose uptake in WT lit- transgenic models. Importantly, the effect of prior AICAR termates in this particular experiment, suggesting that treatment was site specific, as insulin-stimulated phos- the acute effect of AICAR on glucose uptake was not fully phorylation of TBC1D4 Ser324 and Ser595 was unaffected reversed. (Fig. 6E–J) Akt Signaling Glucose Uptake Correlates With TBC1D4 Site-Specific Prior AICAR stimulation potentially enhances muscle Phosphorylation Levels insulin sensitivity to stimulate glucose uptake by regulat- To investigate whether AICAR/AMPK increases muscle ing proximal insulin-signaling proteins. To investigate insulin sensitivity through TBC1D4, we performed a cor- this, we measured the phosphorylation of Akt Thr308 and relation analysis between D values (insulin minus basal) Ser473. Insulin did not further increase the phosphoryla- on muscle glucose uptake and TBC1D4 phosphorylation. tion of Thr308 and Ser473 in muscle previously stimulated We found that glucose uptake and phosphorylation of 711 with AICAR compared with control muscle (Fig. 4A–F). TBC1D4 Ser was positively correlated in WT littermates from both AMPK mouse models (P , 0.001 and TBC1D1 Signaling P , 0.01; Fig. 7A and B, respectively). Correlating data for TBC1D1 is a closely related paralog of TBC1D4 that is glucose uptake and phosphorylation of TBC1D4 Thr649 regulated by both AMPK and Akt, and regulates glucose revealed a more scattered pattern that was positively cor- transport (37–40). As AMPK increases the phosphoryla- , 231 related in WT littermates from the AMPK KD strain (P tion of TBC1D1 Ser in response to contraction and 0.01; Fig. 7C), but was not correlated in WT littermates AICAR, and as Akt increases the phosphorylation of 590 from the AMPK mdKO strain (P = 0.18; Fig. 7D). In Thr in response to insulin, we investigated whether addition, the phosphorylation levels of TBC1D4 Thr649 changes in TBC1D1 phosphorylation occurred in parallel and Ser711 were positively and strongly correlated in with the increase in muscle insulin sensitivity after prior , 231 WT littermates from both AMPK mouse models (P AICAR stimulation. Phosphorylation of TBC1D1 Ser 0.001; Fig. 7E and F). was markedly increased in WT muscle previously stimu- 649 711 lated with AICAR. Prior AICAR stimulation also modestly Phosphorylation Levels of TBC1D4 Thr and Ser increased the phosphorylation of TBC1D1 Ser231 in mus- May Be Causally Linked cle from AMPK KD and mdKO mice (Fig. 5A, B, E, and F). AMPK has been shown to regulate the phosphorylation of TBC1D4 Ser711, and in muscle overexpressing a 4P mutant Furthermore, insulin increased the phosphorylation of 711 590 of TBC1D4 (in which Ser is not mutated) the phosphor- TBC1D1 Thr in both mouse models independent of 711 genotype (Fig. 5C–F). However, in AMPK mdKO mice ylation of Ser is severely blunted (26). In order to investigate whether changes in phosphorylation level of TBC1D4 and WT littermates, insulin-stimulated phosphorylation 711 649 of TBC1D1 Thr590 in prior AICAR-stimulated muscle Ser affect TBC1D4 Thr phosphorylation and vice was decreased compared with control muscle (Fig. 5D). versa, TBC1D4-WT, TBC1D4-S711A, and TBC1D4-T649A constructs were expressed in mouse tibialis anterior mus- TBC1D4 Signaling cle by gene electrotransfer. Insulin increased the phos- TBC1D4 (like TBC1D1) has been identified as a substrate phorylation of TBC1D4 Thr649 in muscle expressing of both AMPK and Akt in skeletal muscle (25,26), and TBC1D4-WT or TBC1D4-S711A, but Thr649 phosphoryla- the phosphorylation of TBC1D4 is critical for insulin- tion levels were significantly blunted in the latter (Fig. 8A stimulated glucose uptake (24,41). In addition, TBC1D4 is and C). Insulin increased the phosphorylation of TBC1D4 phosphorylated at multiple sites in the postexercise pe- Ser711 in muscle expressing TBC1D4-WT, and this re- riod in parallel with enhanced muscle insulin sensitivity sponse was completely ablated in muscle expressing (15,16,27–30). This indicates that the regulation of mus- TBC1D4-T649A (Fig. 8B and C). Our results suggest that cle insulin sensitivity is linked to TBC1D4 phosphoryla- the phosphorylation levels of TBC1D4 Thr649 and Ser711 tion. We found an increased effect of insulin on TBC1D4 are mutually dependent on each other. Thr649 and Ser711 phosphorylation in muscle previously stimulated with AICAR compared with control muscle DISCUSSION (Fig. 6A–D, I, and J). Furthermore, this effect was depen- Several lines of evidence imply that AMPK activation dent on AMPK, because no difference in insulin-mediated regulates skeletal muscle insulin sensitivity. In C2C12 phosphorylation was observed between control and prior myotubes, AICAR stimulation or hyperosmotic stress AICAR-stimulated muscle from either of the two AMPK increases insulin sensitivity, which is inhibited by the

WT (interaction genotype 3 AICAR). E: †††Effect of AICAR within group (P < 0.001). F: AICAR 3 insulin 3 genotype interaction (P < 0.05). *P < 0.001 vs. basal values within genotype; #P < 0.05 vs. response to insulin in WT control (interaction AICAR 3 insulin); †P < 0.001 control vs. AICAR. G: Data are extracted from the raw data given in F. †P < 0.001 vs. control within genotype. H and I: Representative Western blot images from AMPK KD and mdKO studies, respectively. AU, arbitrary units; hr, hour; IB, immunoblotting. diabetes.diabetesjournals.org Kjøbsted and Associates 2049

Figure 4—Phospho-Akt (pAkt) Thr308/Akt2 protein (A and B) and pAkt Ser473/Akt2 protein (C and D) in EDL muscle from either AMPK KD mice (A and C) or AMPK mdKO mice (B and D) and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (KD n =9–10, mdKO n = 12). Data are expressed as the means 6 SEM. A: *Effect of insulin (P < 0.001). B: *Effect of insulin (P < 0.001). †Main effect of AICAR (P = 0.048). Main effect of genotype (P < 0.001). C:*P < 0.001 vs. basal; #P < 0.05 vs. response to insulin in control (interaction insulin 3 AICAR). Main effect of genotype (P =0.033).D: *Effect of insulin (P < 0.001); §P < 0.001 vs. response to insulin in WT (interaction insulin 3 genotype); ‡genotype 3 AICAR interaction (P =0.034).E and F: Repre- sentative Western blot images from AMPK KD and mdKO mouse studies, respectively. AU, arbitrary units; IB, immunoblotting.

unspeciﬁc AMPK inhibitor compound C (42). Similarly, AICAR as a modulator of enzymes such as glycogen insulin sensitivity is increased in myotubes transfected phosphorylase, glucokinase, and phosphofructokinase with a constitutive active form of AMPKa, which is also (31). However, because AICAR did not increase insulin suppressed by compound C (43). Furthermore, AICAR sensitivity in muscle from AMPK KD or mdKO mice, any fails to increase insulin action in cells transfected with possible AMPK-independent effect of AICAR does not a dominant-negative form of AMPKa (43). Collectively, seem to account for changes in glucose uptake in re- our data and those obtained in cell culture systems sponse to insulin. (42,43) suggest that AMPK plays an important role in The improved insulin-stimulated glucose uptake in mediating AICAR-induced increases in skeletal muscle in- muscle previously stimulated with AICAR occurred insulin sensitivity to stimulate glucose transport. dependently of changes in proximal insulin signaling (Akt AICARistakenupbythecell,whereitactsasanAMP phosphorylation). This is consistent with earlier ﬁndings mimetic, thus potentially affecting multiple proteins showing that prior AICAR treatment does not increase regulatedbyAMP.Withinrecentyears,anincreased either Akt phosphorylation or phosphoinositide-3 kinase number of AMPK-independent effects of AICAR have activity in rat skeletal muscle (13). Similar observations been described together with reports identifying have been made in both human and rodent skeletal muscle 2050 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015

Figure 5—Phospho-TBC1D1 (pTBC1D1) Ser231/TBC1D1 protein (A and B) and pTBC1D1 Thr590/TBC1D1 protein (C and D) in EDL muscle from either AMPK KD mice (A and C) or AMPK mdKO mice (B and D) and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (KD n =9–10, mdKO n = 12). Data are expressed as the means 6 SEM. A: †P < 0.001 vs. control value within genotype; ‡genotype 3 AICAR interaction (P < 0.01). B: †P < 0.001 vs. control value within genotype; ‡genotype 3 AICAR interaction (P = 0.001); #insulin 3 AICAR interaction (P =0.041).C: *Effect of insulin (P < 0.001). Main effect of genotype (P < 0.001). D:*P < 0.001 vs. basal; #P < 0.001 vs. response to insulin in control (interaction insulin 3 AICAR). E and F: Representative Western blot images from AMPK KD and mdKO studies, respectively. AU, arbitrary units; IB, immunoblotting.

after an acute bout of exercise (3,4,13,16,17). Based on this, increased the phosphorylation of TBC1D1 Ser231 in both the mechanism responsible for the AMPK-dependent in- basal and insulin-stimulated muscle, and insulin increased crease in muscle insulin sensitivity likely involves signal the phosphorylation of TBC1D1 Thr590 in all groups, our transduction downstream of Akt, implicating a role for results suggest that the phosphorylation of TBC1D1 TBC1D1 or TBC1D4. Ser231 and Thr590 is not sufﬁcient for regulating muscle We evaluated the phosphorylation status of key sites insulin sensitivity in response to prior AICAR treatment. on TBC1D1 previously shown to increase in response to This is supported by ﬁndings of identical TBC1D1 Ser231 AICAR, muscle contraction, exercise, or insulin (16,37,44,45). phosphorylation, and similar increases in insulin-stimulated Phosphorylation of TBC1D1 Ser231 was markedly in- PAS-TBC1D1 and Ser590 phosphorylation in previously creased in muscle from WT mice, and only modestly in- rested or exercised muscle from humans and rats creased in muscle from AMPK KD and mdKO mice 6 h (15,16,29). after AICAR treatment. Conversely, phosphorylation of In addition to TBC1D1, we also analyzed the phos- TBC1D1 Thr590 was increased in response to insulin in- phorylation status of TBC1D4 at multiple sites because it dependent of genotype. Given that prior AICAR stimulation has been suggested to play a prominent role in regulating diabetes.diabetesjournals.org Kjøbsted and Associates 2051

Figure 6—Phospho-TBC1D4 (pTBC1D4) Thr649/TBC1D4 protein (A and B), pTBC1D4 Ser711/TBC1D4 protein (C and D), pTBC1D4 Ser324/ TBC1D4 protein (E and F), and pTBC1D4 Ser595/TBC1D4 protein (G and H) in EDL muscle from either AMPK KD mice (A, C, E, and G)or AMPK mdKO mice (B, D, F, and H) and corresponding WT littermates incubated with or without insulin (100 mU/mL) 6 h after prior AICAR treatment (50 min, 1 mmol/L) (KD n =8–10, mdKO n = 12). Data are expressed as the mean 6 SEM. A: AICAR 3 insulin 3 genotype interaction (P < 0.05). *P < 0.05 vs. basal value within genotype; #P < 0.05 vs. response to insulin in WT control (interaction AICAR 3 insulin). B: AICAR 3 insulin 3 genotype interaction (P < 0.001). *P < 0.05 vs. basal value within genotype; #P < 0.001 vs. response to insulin in WT control (interaction AICAR 3 insulin); †effect of AICAR within basal (P < 0.001). C: AICAR 3 insulin 3 genotype interaction (P < 0.01). *P < 0.001 vs. basal value within genotype; #P < 0.001 vs. response to insulin in WT control (interaction AICAR 3 insulin); †effect of AICAR within basal (P < 0.05). D: AICAR 3 insulin 3 genotype interaction (P < 0.05). *P < 0.05 vs. basal value within genotype; #P < 0.001 vs. response to insulin in WT control (interaction AICAR 3 insulin); †borderline effect of AICAR within basal (P = 0.05). E: *Effect of insulin (P < 0.001). Main effect of genotype (P < 0.01). F and G: *Effect of insulin (P < 0.001). H: *Effect of insulin (P < 0.001); §P < 0.05 vs. response to insulin in WT (interaction insulin 3 genotype). I and J: Representative Western blot images from AMPK KD and mdKO studies, respectively. AU, arbitrary units; IB, immunoblotting. 2052 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015

Figure 7—A and B: Pearson correlations between the D insulin value (insulin minus basal) on glucose uptake and phospho-TBC1D4 (pTBC1D4) Ser711 in WT littermates from KD and mdKO mice, respectively. C and D: Pearson correlations between D insulin value (insulin minus basal) on glucose uptake and pTBC1D4 Thr649 in WT littermates from KD and mdKO mice, respectively. E and F: Pearson correlations between D insulin value (insulin minus basal) on pTBC1D4 Thr649 and Ser711 in WT littermates from KD and mdKO mice, respectively. Sample size is n =10–12. R2 and signiﬁcance level are indicated in the respective panel. To visualize any bias due to grouping effect (control and prior AICAR treatment), we also provide Pearson correlations (dashed lines) based on data from the individual groups. Control, open symbols; prior AICAR treatment, closed symbols. AU, arbitrary units; hr, hour.

both insulin-stimulated glucose uptake and insulin action action on Ser711 phosphorylation in WT mouse muscle in skeletal muscle after exercise (24,27–29). Recent stud- but not in AMPK KD or mdKO mouse muscle previously ies (26,44), using site-speciﬁc antibodies, suggest that stimulated with AICAR are consistent with this notion. only the phosphorylation of TBC1D4 Ser711 is increased In contrast with TBC1D4 Ser711, the phosphorylation in mouse skeletal muscle in response to exercise, AICAR, of Thr649 seems to be important for insulin-stimulated or ex vivo muscle contraction. Because AICAR-mediated glucose uptake in mouse EDL muscle (26,41). However, phosphorylation of TBC1D4 Ser711 is dependent on this site is not regulated by acute AICAR treatment AMPK (26), the AMPK-dependent increase in insulin sen- (26,46). Thus, the potentiated effect of insulin on sitivity after AICAR treatment may be mediated through TBC1D4 Ser711 phosphorylation by prior AICAR treat- changes in TBC1D4 Ser711 phosphorylation during acute ment appears to mediate an enhanced AMPK-dependent AICAR stimulation. Our data showing increased insulin phosphorylation of Thr649, which may facilitate the diabetes.diabetesjournals.org Kjøbsted and Associates 2053

AMPK, through TBC1D4 Ser711, regulates insulin action to stimulate glucose uptake. Previously, it has been shown (47) that discrepancies between Akt and TBC1D4 phosphorylation exist, indicating that only a small fraction of the insulin signal is necessary for mediating full glucose uptake in response to insulin. This is also observed in the current study where phosphorylation of Akt Ser308 and Thr473 does not match either phosphorylation of TBC1D4 Thr649 and Ser711 or glucose uptake in prior AICAR-stimulated WT muscle. However, in cases of normal (and perhaps increased) insulin sensitivity there seems to be a good correlation between plasma membrane GLUT4 and TBC1D4 phosphorylation (47). This indicates that glucose uptake and phosphorylation of TBC1D4 are associated, as also indicated by the correlations in the current study. Besides a change in TBC1D4 phosphorylation, it has been demonstrated that AICAR enhances insulin action in muscle cells by decreasing membrane cholesterol levels in an AMPK-dependent manner (48). This seems plausible since AMPK has been shown to decrease the activity of 3-hydroxy-3-methylglutaryl CoA reductase, the rate-limiting enzyme in cholesterol synthesis (49). We did not measure membrane cholesterol levels, and therefore we cannot rule out that muscle insulin sensitivity after prior AICAR stimulation was affected by a change in membrane cholesterol content. The enhanced insulin-stimulated glucose uptake after AICAR treatment seems to be dependent on a persistent increase in muscle g3-AMPK activity, as evidenced by AMPK activity measurements and insulin-stimulated glucose uptake in g3-AMPK KO mice. Furthermore, prior AICAR stimulation failed to increase insulin sensitivity in mouse soleus muscle in which the a2b2g3 complex Figure 8—Phosphorylation level of TBC1D4 Thr649 (A) and Ser711 represents ,2% of all AMPK trimer complexes (19). In (B) in the tibialis anterior muscle in response to retro-orbital injec- both human vastus lateralis muscle (18) and mouse EDL tion of saline solution or insulin (10 units/kg, 10 min) 7 days after muscle (19), the AMPK a2b2g3 trimer complex accounts muscle gene electrotransfer of TBC1D4-WT, TBC1D4-T649A, fi and TBC1D4-S711A. The top bands in the phospho-TBC1D4 for one- fth of all AMPK complexes (36). Of interest, (pTBC1D4) Ser711 blot are unspecific and do not represent actual AMPK-g3 protein level is markedly decreased in skeletal TBC1D4 protein. Total flag-tagged TBC1D4 protein indicates the muscle from trained humans (50), although insulin sen- n – expression level of the three constructs ( =46). Data are sitivity in general is increased. Conversely, enhanced expressed as the mean 6 SEM. A and B: *Effect of insulin within group (P < 0.05); #P < 0.05 vs. WT. C: Representative Western blot muscle insulin sensitivity after an acute bout of exercise image. AU, arbitrary units; N.D, not detected. seems to be lost in the trained state (51). Collectively, these results suggest that prior AICAR stimulation mimics the effect of exercise to enhance skeletal muscle insulin sensitivity, possibly through an AMPK-g3–dependent increased effect of insulin on glucose uptake. Such a re- mechanism. lationship is supported by the correlative relationship be- In conclusion, prior AICAR stimulation is sufficient to tween TBC1D4 Thr649 and Ser711 phosphorylation, and in enhance muscle insulin sensitivity. We provide evidence particular by the strong relationship between Ser711 phos- that this effect is likely mediated through AMPK signaling, phorylation and muscle glucose uptake. In addition, the as AICAR failed to increase insulin sensitivity in skeletal expression of the S711A TBC1D4 mutant decreased the muscle in which AMPK activity was blunted. Although total phosphorylation of TBC1D4 Thr649,whereas we observed no change in proximal insulin-signaling the T649A mutant severely decreased the phosphorylation events, the enhanced insulin-stimulated glucose uptake of Ser711 both in the presence or absence of insulin. observed after prior AICAR stimulation was associated This clearly indicates interdependence between the and positively correlated with increased TBC1D4 Thr649 two sites and supports a possible mechanism by which and Ser711 phosphorylation. This supports the idea that 2054 AMPK and Insulin Sensitivity in Skeletal Muscle Diabetes Volume 64, June 2015 prior activation of AMPK primes a pool of TBC1D4 to muscle: interactions at the level of phosphatidylinositol 3-kinase, Akt, and AS160. potentiate a subsequent effect of insulin to increase Diabetes 2007;56:2093–2102 GLUT4 translocation to the cell surface and enhance 7. Cartee GD, Young DA, Sleeper MD, Zierath J, Wallberg-Henriksson H, glucose uptake. At present, we have not succeeded in Holloszy JO. Prolonged increase in insulin-stimulated glucose transport in muscle establishing a mouse model for studying insulin sensi- after exercise. Am J Physiol 1989;256:E494–E499 tivity after prior muscle contraction. Therefore, future 8. Mikines KJ, Sonne B, Farrell PA, Tronier B, Galbo H. Effect of physical studies have to determine whether AMPK is also impor- exercise on sensitivity and responsiveness to insulin in humans. Am J Physiol 1988;254:E248–E259 tant for the enhanced insulin action after this interven- 9. Gulve EA, Cartee GD, Zierath JR, Corpus VM, Holloszy JO. Reversal of tion. Because TBC1D4 signaling by insulin is potentiated enhanced muscle glucose transport after exercise: roles of insulin and glucose. after exercise in both human and rat skeletal muscle Am J Physiol 1990;259:E685–E691 (15,16,29,30,52), our hypothesis is that the exercise- 10. Richter EA, Garetto LP, Goodman MN, Ruderman NB. Enhanced muscle induced increase in insulin sensitivity is also regulated glucose metabolism after exercise: modulation by local factors. Am J Physiol via an AMPK-TBC1D4 signaling axis. 1984;246:E476–E482 11. Hansen PA, Nolte LA, Chen MM, Holloszy JO. Increased GLUT-4 trans- Acknowledgments. The authors thank Ann-Marie Petterson, Integrative location mediates enhanced insulin sensitivity of muscle glucose transport after – Physiology Group, Department of Molecular Medicine and Surgery, Karolinska exercise. J Appl Physiol (1985) 1998;85:1218 1222 Institutet, Stockholm, Sweden, for her assistance. The authors also thank D.G. 12. Bogardus C, Thuillez P, Ravussin E, Vasquez B, Narimiga M, Azhar S. Effect Hardie (Division of Molecular Physiology, College of Life Sciences, University of of muscle glycogen depletion on in vivo insulin action in man. J Clin Invest 1983; Dundee, Scotland, U.K.) and L.J. Goodyear (Joslin Diabetes Center and Harvard 72:1605–1610 Medical School, Boston, MA) for the donation of antibodies. 13. Fisher JS, Gao J, Han DH, Holloszy JO, Nolte LA. Activation of AMP kinase Funding. This work was carried out as a part of the research programs enhances sensitivity of muscle glucose transport to insulin. Am J Physiol En- “Physical Activity and Nutrition for Improvement of Health” funded by the Uni- docrinol Metab 2002;282:E18–E23 versity of Copenhagen Excellence Program for Interdisciplinary Research and the 14. Bonen A, Tan MH, Watson-Wright WM. Effects of exercise on insulin binding UNIK project Food, Fitness & Pharma for Health and Disease (see www and glucose metabolism in muscle. Can J Physiol Pharmacol 1984;62:1500–1504 .foodfitnesspharma.ku.dk), which was supported by the Danish Ministry of 15. Funai K, Schweitzer GG, Sharma N, Kanzaki M, Cartee GD. Increased AS160 Science, Technology and Innovation, and by the Novo Nordisk Foundation Center phosphorylation, but not TBC1D1 phosphorylation, with increased postexercise for Basic Metabolic Research. The Novo Nordisk Foundation Center for Basic insulin sensitivity in rat skeletal muscle. Am J Physiol Endocrinol Metab 2009; Metabolic Research is an independent Research Center at the University of 297:E242–E251 Copenhagen that is partially funded by an unrestricted donation from the Novo 16. Funai K, Schweitzer GG, Castorena CM, Kanzaki M, Cartee GD. In vivo Nordisk Foundation (www.metabol.ku.dk). This study was also funded by the exercise followed by in vitro contraction additively elevates subsequent insulin- Danish Council for Independent Research Medical Sciences, the Novo Nordisk stimulated glucose transport by rat skeletal muscle. Am J Physiol Endocrinol Foundation, and the Lundbeck Foundation. J.T.T. was supported by a postdoctoral Metab 2010;298:E999–E1010 fellowship from the Danish Agency for Science, Technology and Innovation. 17. Hamada T, Arias EB, Cartee GD. Increased submaximal insulin-stimulated Duality of Interest. No potential conflicts of interest relevant to this article glucose uptake in mouse skeletal muscle after treadmill exercise. J Appl Physiol were reported. (1985) 2006;101:1368–1376 Author Contributions. R.K. helped to conceive of and design the re- 18. Wojtaszewski JF, Birk JB, Frøsig C, Holten M, Pilegaard H, Dela F. 5’AMP search, perform the experiments and analysis, and draft the manuscript. J.T.T. activated protein kinase expression in human skeletal muscle: effects of strength and J.F.P.W. helped to conceive of and design the research and draft the training and type 2 diabetes. J Physiol 2005;564:563–573 manuscript. J.F. helped to perform the experiments. L.L., B.V., J.B.B., P.S., 19. Treebak JT, Birk JB, Hansen BF, Olsen GS, Wojtaszewski JFP. A-769662 and M.B. helped to perform the analysis. All authors interpreted the results, edited and revised the manuscript, and read and approved the final version of activates AMPK beta1-containing complexes but induces glucose uptake through the manuscript. J.F.P.W. is the guarantor of this work and, as such, had full a PI3-kinase-dependent pathway in mouse skeletal muscle. Am J Physiol Cell – access to all the data in the study and takes responsibility for the integrity of the Physiol 2009;297:C1041 C1052 data and the accuracy of the data analysis. 20. Kahn BB, Alquier T, Carling D, Hardie DG. AMP-activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism. References Cell Metab 2005;1:15–25 1. Richter EA, Garetto LP, Goodman MN, Ruderman NB. Muscle glucose 21. Hardie DG, Ashford MLJ. AMPK: regulating energy balance at the cellular metabolism following exercise in the rat: increased sensitivity to insulin. J Clin and whole body levels. Physiology (Bethesda) 2014;29:99–107 Invest 1982;69:785–793 22. Merrill GF, Kurth EJ, Hardie DG, Winder WW. AICA riboside increases AMP- 2. Richter EA, Mikines KJ, Galbo H, Kiens B. Effect of exercise on insulin action activated protein kinase, fatty acid oxidation, and glucose uptake in rat muscle. in human skeletal muscle. J Appl Physiol (1985) 1989;66:876–885 Am J Physiol 1997;273:E1107–E1112 3. Wojtaszewski JF, Hansen BF, Kiens B, Richter EA. Insulin signaling in hu- 23. Miyamoto L, Egawa T, Oshima R, et al. AICAR stimulation metabolome man skeletal muscle: time course and effect of exercise. Diabetes 1997;46: widely mimics electrical contraction in isolated rat epitrochlearis muscle. Am J 1775–1781 Physiol Cell Physiol 2013;305:C1214–C1222 4. Wojtaszewski JF, Hansen BF, Gade, et al. Insulin signaling and insulin 24. Kramer HF, Witczak CA, Taylor EB, Fujii N, Hirshman MF, Goodyear LJ. sensitivity after exercise in human skeletal muscle. Diabetes 2000;49:325–331 AS160 regulates insulin- and contraction-stimulated glucose uptake in mouse 5. Cartee GD, Holloszy JO. Exercise increases susceptibility of muscle glucose skeletal muscle. J Biol Chem 2006;281:31478–31485 transport to activation by various stimuli. Am J Physiol 1990;258:E390–E393 25. Kramer HF, Witczak CA, Fujii N, et al. Distinct signals regulate AS160 6. Frøsig C, Rose AJ, Treebak JT, Kiens B, Richter EA, Wojtaszewski JFP. phosphorylation in response to insulin, AICAR, and contraction in mouse skeletal Effects of endurance exercise training on insulin signaling in human skeletal muscle. Diabetes 2006;55:2067–2076 diabetes.diabetesjournals.org Kjøbsted and Associates 2055

26. Treebak JT, Taylor EB, Witczak CA, et al. Identification of a novel phos- 40. An D, Toyoda T, Taylor EB, et al. TBC1D1 regulates insulin- and contraction- phorylation site on TBC1D4 regulated by AMP-activated protein kinase in skeletal induced glucose transport in mouse skeletal muscle. Diabetes 2010;59:1358– muscle. Am J Physiol Cell Physiol 2010;298:C377–C385 1365 27. Arias EB, Kim J, Funai K, Cartee GD. Prior exercise increases phosphory- 41. Chen S, Wasserman DH, MacKintosh C, Sakamoto K. Mice with AS160/ lation of Akt substrate of 160 kDa (AS160) in rat skeletal muscle. Am J Physiol TBC1D4-Thr649Ala knockin mutation are glucose intolerant with reduced insulin Endocrinol Metab 2007;292:E1191–E1200 sensitivity and altered GLUT4 trafficking. Cell Metab 2011;13:68–79 28. Treebak JT, Frøsig C, Pehmøller C, et al. Potential role of TBC1D4 in en- 42. Smith JL, Patil PB, Fisher JS. AICAR and hyperosmotic stress increase hanced post-exercise insulin action in human skeletal muscle. Diabetologia insulin-stimulated glucose transport. J Appl Physiol (1985) 2005;99:877–883 2009;52:891–900 43. Ju J-S, Gitcho MA, Casmaer CA, et al. Potentiation of insulin-stimulated 29. Pehmøller C, Brandt N, Birk JB, et al. Exercise alleviates lipid-induced in- glucose transport by the AMP-activated protein kinase. Am J Physiol Cell Physiol sulin resistance in human skeletal muscle-signaling interaction at the level of 2007;292:C564–C572 TBC1 domain family member 4. Diabetes 2012;61:2743–2752 44. Treebak JT, Pehmøller C, Kristensen JM, et al. Acute exercise and physi- 30. Schweitzer GG, Arias EB, Cartee GD. Sustained postexercise increases in ological insulin induce distinct phosphorylation signatures on TBC1D1 and AS160 Thr642 and Ser588 phosphorylation in skeletal muscle without sustained TBC1D4 proteins in human skeletal muscle. J Physiol 2014;592:351–375 increases in kinase phosphorylation. J Appl Physiol (1985) 2012;113:1852–1861 45. Taylor EB, An D, Kramer HF, et al. Discovery of TBC1D1 as an insulin-, 31. Daignan-Fornier B, Pinson B. 5-Aminoimidazole-4-carboxamide-1-beta-D- AICAR-, and contraction-stimulated signaling nexus in mouse skeletal muscle. J ribofuranosyl 59-monophosphate (AICAR), a highly conserved purine intermediate Biol Chem 2008;283:9787–9796 with multiple effects. Metabolites 2012;2:292–302 46. Ducommun S, Wang HY, Sakamoto K, MacKintosh C, Chen S. Thr649Ala- 32. Mu J, Brozinick JT Jr, Valladares O, Bucan M, Birnbaum MJ. A role for AMP- AS160 knock-in mutation does not impair contraction/AICAR-induced glucose activated protein kinase in contraction- and hypoxia-regulated glucose transport transport in mouse muscle. Am J Physiol Endocrinol Metab 2012;302:E1036– in skeletal muscle. Mol Cell 2001;7:1085–1094 E1043 33. Lantier L, Fentz J, Mounier R, et al. AMPK controls exercise endurance, 47. Hoehn KL, Hohnen-Behrens C, Cederberg A, et al. IRS1-independent defects mitochondrial oxidative capacity, and skeletal muscle integrity. FASEB J 2014;28: define major nodes of insulin resistance. Cell Metab 2008;7:421–433 3211–3224 48. Habegger KM, Hoffman NJ, Ridenour CM, Brozinick JT, Elmendorf JS. AMPK 34. Barnes BR, Marklund S, Steiler TL, et al. The 59-AMP-activated protein enhances insulin-stimulated GLUT4 regulation via lowering membrane choles- kinase gamma3 isoform has a key role in carbohydrate and lipid metabolism in terol. Endocrinology 2012;153:2130–2141 glycolytic skeletal muscle. J Biol Chem 2004;279:38441–38447 49. Carling D, Clarke PR, Zammit VA, Hardie DG. Purification and character- 35. Chen S, Murphy J, Toth R, Campbell DG, Morrice NA, Mackintosh C. ization of the AMP-activated protein kinase. Copurification of acetyl-CoA car- Complementary regulation of TBC1D1 and AS160 by growth factors, insulin and boxylase kinase and 3-hydroxy-3-methylglutaryl-CoA reductase kinase activities. AMPK activators. Biochem J 2008;409:449–459 Eur J Biochem 1989;186:129–136 36. Birk JB, Wojtaszewski JFP. Predominant alpha2/beta2/gamma3 AMPK 50. Frøsig C, Jørgensen SB, Hardie DG, Richter EA, Wojtaszewski JFP. 59-AMP- activation during exercise in human skeletal muscle. J Physiol 2006;577:1021– activated protein kinase activity and protein expression are regulated by en- 1032 durance training in human skeletal muscle. Am J Physiol Endocrinol Metab 2004; 37. Pehmøller C, Treebak JT, Birk JB, et al. Genetic disruption of AMPK sig- 286:E411–E417 naling abolishes both contraction- and insulin-stimulated TBC1D1 phosphoryla- 51. Mikines KJ, Sonne B, Tronier B, Galbo H. Effects of acute exercise and tion and 14-3-3 binding in mouse skeletal muscle. Am J Physiol Endocrinol detraining on insulin action in trained men. J Appl Physiol (1985) 1989;66:704– Metab 2009;297:E665–E675 711 38. Vichaiwong K, Purohit S, An D, et al. Contraction regulates site-specific 52. Castorena CM, Arias EB, Sharma N, Cartee GD. Postexercise improvement phosphorylation of TBC1D1 in skeletal muscle. Biochem J 2010;431:311–320 in insulin-stimulated glucose uptake occurs concomitant with greater AS160 39. Frøsig C, Jensen TE, Jeppesen J, et al. AMPK and insulin action— phosphorylation in muscle from normal and insulin-resistant rats. Diabetes 2014; responses to ageing and high fat diet. PLoS One 2013;8:e62338 63:2297–2308