Skeletal Muscle AMP-activated Protein Kinase Is Essential forthe Metabolic Response to Exercise in Vivo*□S

Received for publication, March 23, 2009, and in revised form, May 13, 2009 Published, JBC Papers in Press, June 12, 2009, DOI 10.1074/jbc.M109.021048

Robert S. Lee-Young‡1, Susan R. Griffee‡, Sara E. Lynes‡, Deanna P. Bracy‡, Julio E. Ayala‡§, Owen P. McGuinness‡§,and David H. Wasserman‡§

From the ‡Department of Molecular Physiology and Biophysics and the §Mouse Metabolic Phenotyping Center, VanderbiltUniversity School of Medicine, Nashville, Tennessee 37232

AMP-activatedproteinkinase (AMPK)hasbeenpostulated asa super-metabolic regulator, thought to exert numerous effectson skeletal muscle function, metabolism, and enzymatic signal-ing. Despite these assertions, little is known regarding the directrole(s) of AMPK in vivo, and results obtained in vitro or in situare conflicting. Using a chronically catheterized mouse model(carotid artery and jugular vein), we show that AMPK regulatesskeletal muscle metabolism in vivo at several levels, with theresult that a deficit in AMPK activity markedly impairs exercisetolerance. Compared with wild-type littermates at the same rel-ative exercise capacity, vascular glucose delivery and skeletalmuscle glucose uptake were impaired; skeletalmuscle ATP deg-radation was accelerated, and arterial lactate concentrationswere increased in mice expressing a kinase-dead AMPK�2 sub-unit (�2-KD) in skeletal muscle. Nitric-oxide synthase (NOS)activity was significantly impaired at rest and in response toexercise in �2-KD mice; expression of neuronal NOS (NOS�)was also reduced. Moreover, complex I and IV activities of theelectron transport chain were impaired 32 � 8 and 50 � 7%,respectively, in skeletal muscle of �2-KD mice (p < 0.05 versuswild type), indicative of impairedmitochondrial function. Thus,AMPK regulates neuronal NOS� expression, NOS activity, andmitochondrial function in skeletal muscle. In addition, theseresults clarify the role of AMPK in the control ofmuscle glucoseuptake during exercise. Collectively, these findings demonstratethat AMPK is central to substratemetabolism in vivo, which hasimportant implications for exercise tolerance in health and cer-tain disease states characterized by impaired AMPK activationin skeletal muscle.

Theubiquitously expressed serine/threonineAMP-activatedprotein kinase (AMPK)2 is an ��� heterotrimer postulated to

play a key role in the response to energetic stress (1, 2), becauseof its sensitivity to increased cellular AMP levels (3). Pharma-cological activation of AMPK (primarily via the AMP analogueZMP) increases catabolic processes such as GLUT4 transloca-tion (4, 5), glucose uptake (6, 7), long chain fatty acid (LCFA)uptake (8), and substrate oxidation (6). Concomitantly, phar-macological activation of AMPK inhibits anabolic processes,and in skeletal muscle genetic reduction of the catalyticAMPK�2 subunit eliminates these pharmacological effects(9–12). Thus, AMPK has been proposed to act as a metabolicmaster switch (2, 13, 14). Physiologically, exercise at intensitiessufficient to increase free cytosolic AMP (AMPfree) levels is apotent stimulus of AMPK, preferentially activatingAMPK�2 inskeletal muscle (15–17). The metabolic profile of skeletal mus-cle during moderate to high intensity exercise is remarkablysimilar to skeletal muscle in which AMPK has been pharmaco-logically activated (i.e. increases in catabolic processes). This isconsistent with the hypothesis that AMPK activation isrequired for themetabolic response to increased cellular stress.Given this, it is surprising that the direct role(s) of skeletal mus-cle AMPK during exercise under physiological in vivo condi-tions is unknown.A number of studies have tried to attribute causality to the

AMPK and metabolic responses to exercise using transgenicmodels. In mouse models in which AMPK�2 protein expres-sion and/or activity has been impaired, contractions performedin isolated skeletal muscle in vitro, ex vivo, or in situ have dem-onstrated that skeletal muscle glucose uptake (MGU) is normal(9, 10), partially impaired (11, 18), or ablated (19). Furthermore,ex vivo skeletal muscle LCFA uptake and oxidation in responseto contraction appears to be AMPK-independent (20, 21). Akey limitation of these studies is that the experimental modelswere not physiological. Under in vivo conditions, mice express-ing a kinase-dead (18) or inactive (22) AMPK�2 subunit in car-diac and skeletal muscle have impaired voluntary and maximalphysical activity, respectively, indicative of a physiological rolefor AMPK during exercise. In this context, obese non-diabeticand diabetic individuals have impaired skeletal muscle AMPKactivation during moderate intensity exercise (23) as well asduring the post-exercise period (24), yet the contribution of thisimpairment to the disease state is unclear. Thus, in vivo studies

* This work was supported, in whole or in part, by National Institutes of HealthGrants U24 DK-59637 and R01 DK-54902 (to D. H. W.).

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Tables S1 and S2 and Figs. S1–S5.

1 Supported by a mentor-based fellowship from the American Diabetes Asso-ciation. To whom correspondence should be addressed: 823 Light Hall,2200 Pierce Ave., Nashville, TN 37232. Fax: 615-322-7236; E-mail:robert.s.lee-young@vanderbilt.edu.

2 The abbreviations used are: AMPK, AMP-activated protein kinase; %QG, per-cent cardiac output to gastrocnemius muscle; �2-KD, AMPK �2 kinase-dead subunit; ETC, electron transport chain; eNOS, endothelial NOS; LCFA,long-chain fatty acid; MGU, muscle glucose uptake; NEFA, nonesterifiedfatty acid; NO, nitric oxide; NOS, nitric-oxide synthase; nNOS, neuronalNOS; nNOS�, muscle isoform of nNOS; OXPHOS, oxidative phosphoryl-ation; TEI, tissue extraction index; BisTris, 2-[bis(2-hydroxyethyl)-

amino]-2-(hydroxymethyl)propane-1,3-diol; ANOVA, analysis of vari-ance; Cr, creatine; PCr, phosphocreatine; SVL, superficial vastus latera-lis; 2-[14C]DG, 2-[14C]deoxyglucose; and 3H-R-BrP, [9,10-3H]-(R)-2-bro-mopalmitate; WT, wild type; ZMP, 5-aminoimidazole-4-carboxamideriboside monophosphate .

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 284, NO. 36, pp. 23925–23934, September 4, 2009© 2009 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

SEPTEMBER 4, 2009 • VOLUME 284 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 23925

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

are essential to define the role of AMPK in skeletal muscle dur-ing exercise.Physical exercise of a moderate intensity is an effective

adjunct treatment for chronic metabolic diseases such as obe-sity and type 2 diabetes (25). Given the importance of elucidat-ing themolecularmechanism(s) regulating skeletalmuscle sub-strate metabolism during exercise and the putative role ofAMPK as a critical mediator in this process, we tested thehypothesis that AMPK�2 is functionally linked to substratemetabolism in vivo.

EXPERIMENTAL PROCEDURES

Animal Maintenance—All procedures were approved by theVanderbilt University Animal Care and Use Committee. Maleand female C57BL/6J mice expressing a kinase-dead AMPK�2subunit (�2-KD) in cardiac and skeletal muscle (18) and wild-type (WT) littermate mice were studied. Twenty one days afterbirth, littermates were separated by gender, maintained inmicroisolator cages, fed a standard chow diet (5.5% fat byweight; 5001 Laboratory Rodent Diet, Purina), and had accessto water ad libitum. All mice were studied at 16 weeks of age.Exercise Stress Test—Peak oxygen consumption (VO2peak)

was assessed using an exercise stress test protocol. Two daysprior to the exercise stress test, all mice were acclimatized totreadmill running by performing 10min of exercise at a speed of10 m�min�1 (0% incline). To determine VO2peak, mice wereplaced in an enclosed single lane treadmill connected to Oxy-max oxygen (O2) and carbon dioxide (CO2) sensors (ColumbusInstruments, Columbus, OH). Following a 30-min basal period,mice commenced running at 10m�min�1 on a 0% incline. Run-ning speed was increased by 4 m�min�1 every 3 min until micereached exhaustion, defined as the time point whereby miceremained at the back of the treadmill on a shock grid for �5 s.O2 consumption and CO2 production were assessed at 30-sintervals throughout the basal and exercise periods. Basal val-ues are representative of the final 10 min of the basal period.Prior to the VO2peak test, body weight was measured, and bodycomposition was assessed using an mq10 NMR analyzer(Bruker Optics, The Woodlands, TX). Given that changes inwhole body VO2 during exercise closely reflect changes occur-ring within exercising muscle (26), all oxygen consumptionmeasurements were expressed per kg of lean body mass(kgLBM).Metabolic Experiments—Following the exercise stress test,

surgical procedures were performed as described previously(27) to catheterize the left common carotid artery and rightjugular vein for sampling and infusions, respectively. The cath-eters were exteriorized, sealed with stainless steel plugs, andkept patentwith saline containing 200 units�ml�1 heparin and 5mg�ml�1 ampicillin. Mice were housed individually post-sur-gery, and body weight was recorded daily. Five days followingsurgery, all mice performed a 10-min bout of exercise at theirpre-determined experimental running speed (see below).Experiments were performed 2 days later.Approximately 1 h prior to the experiment, Micro-Re-

nathane tubing was connected to the exteriorized catheters,and all mice were placed in the enclosed treadmill to acclimateto the environment. At t � 0 min, a base-line arterial blood

sample was taken for the measurement of arterial glucose,plasma insulin, plasma nonesterified fatty acids (NEFA),plasma lactate, and hematocrit. Mice then remained sedentaryor performed a single bout of exercise. Sedentary mice wereallowed to move freely in the stationary treadmill for 30 min.Mice that exercised were divided into three groups as follows:1) �2-KD mice performed a maximum of 30 min of treadmillexercise at 70% of their maximum running speed; 2) WT miceran at the same absolute running speed as �2-KD mice; 3) WTmice ran at the same relative intensity as �2-KDmice. Runningtime was matched between groups.In all mice, a bolus containing 13�Ci of 2-[14C]deoxyglucose

(2-[14C]DG) and 26 �Ci of [9,10-3H]-(R)-2-bromopalmitate(3H-R-BrP) was injected into the jugular vein at t � 5 min toprovide an index of tissue-specific glucose and LCFA uptakeand clearance, respectively. At t� 7, 10, 15, and 20min, arterialblood was sampled to determine blood glucose, plasma NEFA,plasma lactate, and plasma 2-[14C]DG and 3H-R-BrP. Hemato-crit was measured at t � 20 min, and at t � 30 min or exhaus-tion, arterial blood was taken for the measurement of bloodglucose, plasma insulin, plasma NEFA, plasma lactate, plasma2-[14C]DG, and 3H-R-BrP. Following the final arterial bloodsample, 50 �l of yellow DYE-TRAK� microspheres (15 �m;Triton Technology Inc., San Diego) were injected into thecarotid artery, followed by a small flush of saline, to assessthe percentage of cardiac output to gastrocnemius (%QG) andthe left and right kidney. Mice were then anesthetized with anarterial infusion of sodium pentobarbital (3 mg). The soleus,right gastrocnemius, superficial vastus lateralis (SVL), heart,and brain were rapidly excised, frozen in liquid nitrogen, andstored at �70 °C. The left gastrocnemius and left and right kid-ney were placed into 15-ml polypropylene tubes and stored at4 °C prior to microsphere analysis.Echocardiography—Transthoracic echocardiograms were

performed as described previously (28). Mice were acclimatedto the procedure over 3 days. Immediately following treadmillexercise, two-dimensional targeted M-mode echocardio-graphic images were obtained at the level of the papillary mus-cles from the parasternal short axis view and recorded at aspeed of 150 cm/s for the measurement of heart rate. Echocar-diograms were completed within 72 � 13 s after exercise. Leftventricular wall thickness, end diastolicmeasurements, and leftventricular end systolic dimensions were determined asdescribed previously (28) and are the average of three to fiveconsecutive selected sinus beats using the leading edge tech-nique. Heart rate was determined from the cardiac cyclesrecorded on the M-mode tracing.Plasma and Tissue Radioactivity—Plasma 2-[14C]DG radio-

activity was assessed by liquid scintillation counting followingdeproteinization with 0.3 N Ba(OH)2 and 0.3 N ZnSO4 asdescribed previously (29). Plasma 3H-R-BrP radioactivity wasdetermined directly from the plasma via liquid scintillationcounting. Tissue 2-[14C]DG and 3H-R-BrP were determinedusing amodifiedmethod of Folch et al. (30). Chloroform:meth-anol (2:1)was added to a portion of tissue that had been crushedin liquid nitrogen using a mortar and pestle, homogenized onice, and stored at 4 °C for 60min. KCl (0.1 M) was then added tothe homogenate, and samples were centrifuged at 3500 � g for

Physiological Role of AMPK in Skeletal Muscle

23926 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 36 • SEPTEMBER 4, 2009

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

15 min. The upper aqueous phase (containing 2-[14C]DG) wasused to determine 2-[14C]DG -P as described previously (29). Aportion of the lower lipid phase (containing 3H-R-BrP) wasused to determine tissue 3H-R-BrP content (31).Plasma Hormones and Metabolites—Immunoreactive

plasma insulinwas assayedwith a double antibodymethod (32),and plasma NEFA were measured spectrophotometricallyusing an enzymatic colorimetric assay (NEFA C kit, WakoChemicals Inc.). Plasma lactate was determined enzymatically(33), with lithium L-lactate (Sigma) used as the standard. Arte-rial glucose levels were determined directly from �5 �l of arte-rial blood samples using anACCU-CHEK�Advantagemonitor(Roche Diagnostics).Muscle Metabolites—For muscle glycogen determination,

2 M HCl was added to a portion (�10 mg) of crushed tissuesamples, which were then incubated at 100 °C for 2 h and neu-tralized with 0.667 M NaOH. Glucose units were determinedusing an enzymatic fluorometric method (33). Muscle lactate,PCr, Cr, andATPwere analyzed from�20mg of crushed tissueusing enzymatic fluorometric techniques (33). ADPfree andAMPfree were calculated as described previously (34).Microsphere Isolation—Tissues were digested overnight in

1 M KOH at 60 °C. Following sonication with Triton X-100,microspheres were suspended in ethanol containing 0.2% (v/v)HCl, followed by ethanol. The microsphere:ethanol solutionwas evaporated at room temperature, and 200 �l of N,N-di-methylformamide (Sigma) was added to elute the fluorescentdye from the microspheres. The absorbance of the N,N-dim-ethylformamide solution was determined at 450 nm.Immunoblotting—Muscle samples were homogenized in

lysis buffer (50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM

EGTA, 10% glycerol, 1% Triton X-100, 1 mM dithiothreitol, 1mM phenylmethylsulfonyl fluoride, 10 �g/ml trypsin inhibitor,5 �l/ml protease inhibitor mixture, 50 mM NaF, and 5 mM

sodium pyrophosphate). Samples were centrifuged at 10,000 �g, and protein content in the supernatant was determined usingthe Bradford method. Protein expression of AMPK�1 and -�2,acetyl-CoA carboxylase-�, neuronal (n) nitric-oxide synthase(NOS), and endothelial (e) NOS was determined from 75 �g ofwhole cell lysate. Inducible NOS was immunoprecipitatedusing 200 �g of protein in conjunction with immobilizedRecomb protein A beads (Pierce) and an anti-inducible NOSmouse monoclonal antibody (BD Biosciences). Proteins wereseparated using NuPAGE 4–12% BisTris gels (Invitrogen) andtransferred to polyvinylidene difluoridemembranes. Blots wereprobed with anti-AMPK�1 rabbit monoclonal antibody (1:500;Abcam, Cambridge, MA), anti-AMPK�2 goat polyclonal anti-body (1:100; Santa Cruz Biotechnology), anti-nNOS mousemonoclonal antibody (1:500; BD Biosciences), anti-eNOS rab-bit polyclonal antibody (1:100; Abcam,MA), and anti-inducibleNOS mouse monoclonal antibody (1:100; BD Biosciences).Antibody binding was detected with either IRDyeTM 800-con-jugated anti-rabbit IgG (1:10,000), IRDyeTM 700-conjugatedanti-mouse IgG (1:10,000), or IRDyeTM 800-conjugated anti-goat IgG secondary antibodies (Rockland Immunochemicals,Inc., Gilbertsville, PA). Acetyl-CoA carboxylase-� proteinexpression was detected using IRDyeTM 800-labeled streptavi-din (1:5,000; Rockland).

AMPK and NOS Activity Assays—AMPK�2 and -�1 weresequentially immunoprecipitated using 200 �g of protein, 2 �gof a rabbit AMPK�2 polyclonal antibody (Abcam), 2 �l of arabbit AMPK�1 monoclonal antibody (Abcam), and immobi-lized Recomb protein A beads (Pierce). AMPK activity in theimmune complexes was measured for 24 min at 30 °C (withinthe pre-determined linear range) in the presence of 200 �M

AMP and calculated as picomoles of phosphate incorporatedinto the SAMS peptide (100 �M; GenWay Biotech) per min permg of protein subjected to immunoprecipitation.NOS activity was measured on gastrocnemius and SVLmus-

cle. Samples were homogenized in lysis buffer, and 5 �l of sam-ple (�70 �g of protein) was added to pre-heated assay buffer(1.15mMNADPH, 4�MBH4, 100 nM calmodulin, 0.7mMCaCl,0.63 �M FAD, 3 �M L-[3H]arginine). The assay was performedfor 7 min at 37 °C (within the linear range), and NOS activitywas measured with or without the NOS inhibitor N�-nitro-L-arginine methyl ester (1 mM). NOS activity is the differencebetween samples incubated with or without N�-nitro-L-argi-nine methyl ester and was calculated as picomoles ofL-[3H]arginine converted to picomoles of L-[3H]citrulline permin per mg of protein.OXPHOSActivity Assays—Post-600� g supernatants of gas-

trocnemius muscle were prepared as described previously (35).Briefly, frozen samples were homogenized in 120 mM KCl, 20mM HEPES (pH 7.4), 2 mM MgCl, 1 mM EGTA, and 5 mg/mlbovine serum albumin and centrifuged twice at 600 � g for 10min at 4 °C. The second supernatant was stored in 2 �g/�laliquots at �70 °C. All assays were performed at 30 °C in a finalvolume of 1 ml using a SpectraMax Plus384 spectrophotometer(Molecular Devices). Prior to measurement of complex I, I �III, II, and II� III activity, sampleswere diluted 1:1 in hypotonicmedia (final concentration of 25 mM potassium phosphate (pH7.2), 5 mM MgCl) and freeze-thawed three times.Complex I activity (NADH:ubiquinone oxidoreductase; EC

1.6.5.3) was measured by following the decrease in absorbancedue to the oxidation of NADH at 340 nm, with 425 nm as thereference wavelength (� � 6.81mM�1�cm�1) (35). The reactionwas initiated by adding 30 �g of protein to the assay buffer (25mM potassium phosphate (pH 7.2), 5 mMMgCl, 2 mMKCN, 2.5mg/ml bovine serum albumin (fraction V), 130 �M NADH, 65�M decylubiquinone, 2 �g/ml antimycin A) and monitored for5 min. Rotenone (2 �g/ml) was added, and the reaction wasmonitored for 3 min. Complex I activity is the differencebetween total enzymatic rates and rates obtained in the pres-ence of rotenone. Complex I � III (NADH-cytochrome c oxi-doreductase) activity was determined as described previously(36) with minor modifications. The reaction was initiated byadding 30 �g of protein to the assay buffer (50 mM potassiumphosphate (pH 7.2), 80�M cytochrome c (bovine heart), 130�M

NADH, 2 mM KCN, 5 mM MgCl). The increase in absorbancedue to the reduction of ferricytochrome c (� � 19 mM�1�cm�1)was monitored for 3 min at 550 nm with 580 nm as the refer-ence wavelength. Rotenone (2 �g/ml) was added, and the reac-tion wasmonitored for a further 3min. Complex I� III activityis the rotenone-sensitive rate.Complex II activity (succinate:ubiquinone oxidoreductase;

EC 1.3.5.1) was measured by following the reduction of 2,6-

Physiological Role of AMPK in Skeletal Muscle

SEPTEMBER 4, 2009 • VOLUME 284 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 23927

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

dichlorophenolindophenol at 600 nmwith 750 nm as the refer-ence wavelength (� � 19.1 mM�1�cm�1) (35). Samples (30 �g)were incubated in 25 mM potassium phosphate (pH 7.2), 5 mM

MgCl, and 20 mM succinate (pH 7.2) for 10 min at 30 °C. Anti-mycin A (2 �g/ml), rotenone (2 �g/ml), KCN (2 mM), and 2,6-dichlorophenolindophenol (50 �M) were added, and the reac-tion was monitored for 3 min. Decylubiquinone (65 �M) wasadded, and the reaction was monitored for a further 3 min. Forthe measurement of complex II � III activity (succinate-cyto-chrome c oxidoreductase), 30�g of proteinwas added to 25mM

potassium phosphate (pH 7.2), 2 mM KCN, 20 mM succinate(pH 7.2), 2 �g/�l rotenone and incubated at 30 °C for 10 min.Ferricytochrome c was added (37.5 �M), and the increase inabsorbance due to the reduction of ferricytochrome c wasmeasured for 3 min at 550 nm with 580 nm as the referencewavelength.Complex IV activity (cytochrome c oxidase; EC 1.9.3.1) was

measured by following the decrease in absorbance at 550 nmdue to the oxidation of ferrocytochrome c, with 580 nm as thereference wavelength (� � 19.1 mM�1�cm�1) (35). Samples (10�g) were added to 20mM potassium phosphate, 15 �M ferrocy-tochrome c, and 450�Mn-dodecyl�-D-maltoside, and the reac-tionwasmonitored for 30 s. Complex IV activity was calculatedfrom the initial rate. Ferrocytochrome cwas prepared by adding5 �M dithiothreitol to 200 �M ferricytochrome c. After 20 min,the 550 nm/565 nm ratio was determined, and ferricytochromec was considered reduced if the ratio was between 10 and 20.Citrate synthase wasmeasured on 10 �g of sample as describedby Barrientos (37).Calculations—The tissue-specific clearance of 2-[14C]DG

and 3H-R-BrP (Kg andKf, respectively) and themetabolic indexfor glucose and LCFA (Rg and Rf) were calculated as describedpreviously (38). Kg and Kf are used as concentration-indepen-dent indices of muscle glucose and LCFA uptake, respectively.Rg and Rf are concentration-dependent indices of muscle glu-cose and LCFA uptake, respectively.Percent cardiac output was calculated from fluorescent

intensity as described previously (39) and is expressed as per-cent cardiac output to the tissue (%QT), where %QT � (fT/fRef)�(tissueaverage/tissuemouse). fT and ƒRef are the fluorescentintensity of the tissue and reference sample, respectively. Ade-quacy of microsphere mixing was assumed if %Q to the left andright kidneywaswithin 10%.Of the 43mice infusedwithmicro-spheres, 34 met the inclusion criteria for analysis.The amount of 2-[14C]DG-P present in the gastrocnemius

muscle as well as the amount of microspheres trapped withinthe gastrocnemius muscle were used to determine the glucosetissue extraction index (TEI). The glucose TEI was calculatedby expressing the percentage of 2-[14C]DG-P (expressed rela-tive to the amount infused) relative to the percentage of micro-spheres (expressed relative to the amount infused). For theechocardiography experiments, an index linearly related to car-diac outputwas calculated as heart rate� (diastolic left ventric-ular internal dimension3 � systolic left ventricular internaldimension3) (28).Statistical Analyses—Data are means � S.E. Statistical anal-

ysis was performed using a Student’s t test, one-way analysis ofvariance (ANOVA), one-way repeated measures ANOVA, or

two-way repeated measures ANOVA where appropriate withthe statistical software package SigmaStat. If the ANOVA wassignificant (p � 0.05), specific differences were located usingFisher’s least significant difference test.

RESULTS

Exercise Capacity and Oxygen Consumption in Vivo AreImpaired in �2-KD mice during an Exercise Stress Test—At 16weeks of age no significant differences were observed between�2-KD mice and WT littermates with respect to body weight(24 � 2 versus 25 � 1 g forWT and �2-KD, respectively), mus-clemass (77� 1 versus 78� 2% bodyweight), or fatmass (8.5�1.4 versus 9.4 � 0.4% body weight). Basal VO2 was similarbetween genotypes (78� 5 versus 79� 6ml�kgLBM�1�min�1 forWT and �2-KD, respectively) as was the respiratory exchangeratio (0.77 � 0.02 versus 0.77 � 0.01). During an exercise stresstest,�2-KDmice displayedmarked exercise intolerance as seenby impairments in maximum running speed (38 � 1 versus21 � 1 m�min�1 for WT and �2-KD, respectively; p � 0.001)and running time (23 � 1 versus 10 � 1 min; p � 0.001). VO2during the stress test increased at a similar rate in WT and�2-KDmice (Fig. 1A); however, VO2peak was reduced in�2-KDmice (142� 2 versus 113� 4ml�kgLBM�1�min�1; p� 0.001). Asa result, �2-KDmice were exercising at a greater percentage ofVO2peak compared with WT mice at any given absolute workrate (supplemental Table S1). Respiratory exchange ratio wassimilar between genotypes at exhaustion (0.89 � 0.01 versus0.90 � 0.03). At a VO2 of �90 ml�kgLBM�1�min�1, VCO2increased disproportionately compared with VO2 in WT mice(Fig. 1B), reflecting a change in either substrates utilized oracidosis. This effect was not apparent in �2-KDmice (Fig. 1C).Acute Exercise Experiment, Controlling for Relative and

Absolute Exercise Intensity—To examine the role of AMPK�2in the regulation of skeletal muscle metabolic flux in vivo,�2-KD mice performed a single bout of treadmill exercise at70% of their maximum running speed (�2-KD70%). Because ofthe difference in maximum running speed between the geno-types, WT mice that exercised at the same absolute speed as�2-KD70% did so at �45% of their maximum running speed(WT45%; supplemental Table S2). To best equate results to�2-KD70%, a second group ofWTmice was exercised at 70% oftheir maximum running speed (WT70%; supplemental TableS2). As demonstrated in the results that follow, controlling forabsolute and relative exercise intensity is essential for interpre-tation of the physiological and metabolic responses to exercisein vivo.AMPK� Protein Expression and AMPK Activity Is Impaired

in Skeletal Muscle of �2-KD Mice—Similar to other musclegroups (11), expression of the �2-KD subunit in gastrocnemiusmuscle was increased relative to native AMPK�2 (98 � 8%higher in �2-KD compared with WT, p � 0.01; supplementalFig. S1A). A concomitant decrease in AMPK�1 expression wasobserved in the gastrocnemius of �2-KDmice (51 � 11% lowerin �2-KD compared with WT, p � 0.02; supplemental Fig.S1A). Similar findings for AMPK�2 and �1 expression wereobserved in SVL muscle (data not shown). In gastrocnemiusmuscle of WT mice, AMPK�2 (supplemental Fig. S1B) andAMPK�1 activities (supplemental Fig. S1C) increased in an

Physiological Role of AMPK in Skeletal Muscle

23928 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 36 • SEPTEMBER 4, 2009

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

intensity-dependent manner. AMPK�2 and -�1 activities werebarely detectable in the gastrocnemius of �2-KD mice undersedentary conditions and did not change in response to exer-cise. Gastrocnemius acetyl-CoA carboxylase-� Ser221 phos-

phorylation was similar between genotypes at rest andincreased to a similar extent in all groups in response to exercise(supplemental Fig. S1D).SkeletalMuscle ATPConcentrationsDecrease in�2-KDMice

during Exercise in Vivo—In response to exercise, no significantchanges in ATP were observed in the gastrocnemius of WT45%

or WT70% (Table 1). In contrast, exercise significantlydecreased gastrocnemius ATP levels in �2-KD70%. Lactate andcreatine (Cr) significantly increased, whereas phosphocreatine(PCr), PCr:(PCr � Cr), and glycogen significantly decreasedduring exercise in all groups (Table 1). In �2-KD70% andWT70%, ADPfree, AMPfree, and AMPfree:ATP all increased inresponse to exercise (Table 1). The similar increase in AMPfreeand AMPfree:ATP observed between �2-KD70% and WT70%

shows that, by this criteria, cellular stress was equally elevatedin these groups compared with WT45%. This finding empha-sizes the need to exercise mice at the same relative work inten-sity to obtain comparable energetic responses in vivo.Arterial Metabolites and Hormones Are Altered in �2-KD

Mice at Rest and during Steady State Exercise in Vivo—Anincrease in exercise intensity resulted in significantly lowerarterial glucose levels in WT mice (Fig. 2A). Compared withWT70%, arterial glucose levels during exercisewere significantlygreater in �2-KD70%. Arterial NEFAs (Fig. 2B) and insulin (Fig.2C) decreased to similar concentrations in all groups duringexercise. Although no differences in basal insulin levels wereobserved between individual groups, basal insulin levels weregreater in�2-KD70% comparedwith the combined average of allWT mice (98 � 6 versus 71 � 7 pM, p � 0.05). Arterial lactateincreased over time in all exercise groups (Fig. 2D), and a sig-nificant group effect was observed with �2-KD70% � WT70% �WT45% (p � 0.01).Indices of Glucose Uptake, but Not LCFA Uptake, Are

Impaired in Skeletal Muscle of �2-KD Mice during Exercise inVivo—In WT mice, an increase in exercise intensity increasedthe plasma disappearance of 2-[14C]DG at 7 and 10 min (Fig.3A). Gastrocnemius Kg (Fig. 3B) and Rg (Fig. 3C) also increasedinWT70% compared withWT45%. At the same relative exerciseintensity, the disappearance of plasma 2-[14C]DG was attenu-ated at 7 min in �2-KD70% mice when compared with WT70%.In �2-KD70% mice, gastrocnemius Kg was impaired by �60%when compared withWT70% mice (Fig. 3B). Gastrocnemius Rg

FIGURE 1. Oxygen consumption is impaired in 16-week-old chow-fedC57BL/6J mice expressing a kinase-dead form of AMP-activated proteinkinase �2 (�2-KD) in cardiac and skeletal muscle. Compared with WT lit-termates, the increase in oxygen consumption (VO2) during an exercisestress test is attenuated in �2-KD mice (A). B and C, carbon dioxide production(VCO2) during an exercise stress test was plotted against VO2 for WT and�2-KD mice, respectively. Note the change of slope of VCO2 versus VO2 in WTmice (denoted by the arrow) that is not present in �2-KD mice. Data aremean � S.E. for n � 8 –9. kgLBM indicates kilograms of lean body mass.

TABLE 1Measured and calculated metabolites (normalized to total creatine levels) and glycogen at rest and immediately following exercise ingastrocnemius muscle of 16-week-old chow-fed C57BL/6J mice expressing WT or kinase-dead form of AMP-activated protein kinase�2 (�2-KD) in cardiac and skeletal muscleData are mean � S.E. for n � 5–7 per group.

MetaboliteSedentary Exercise

WT �2-KD WT45% WT70% �2-KD70%

ATP (�mol�100 g�1) 29.1 � 2.2 30.5 � 1.6 25.7 � 1.1 25.4 � 1.2 19.7 � 2.7aLactate (�mol�100 g�1) 12.3 � 2.2 12.3 � 1.7 57.7 � 14.1a 54.0 � 12.6a 78.4 � 15.5aPCr (�mol�100 g�1) 62.4 � 3.1 57.8 � 3.3 28.5 � 8.3a 29.5 � 8.5a 16.1 � 5.2aCr (�mol�100 g�1) 37.6 � 3.1 42.2 � 3.3 71.5 � 8.3a 70.5 � 8.5a 83.9 � 5.2aPCr:(PCr � Cr) 0.60 � 0.03 0.58 � 0.03 0.28 � 0.08a 0.29 � 0.08a 0.12 � 0.03aADPfree (nmol�100 g�1) 154 � 19 178 � 22 255 � 57a 372 � 90a 451 � 117aAMPfree (nmol�100 g�1) 0.9 � 0.2 1.2 � 0.2 3.2 � 1.3 7.8 � 3.4a 7.5 � 2.3aAMPfree:ATP 0.05 � 0.01 0.03 � 0.00 0.13 � 0.06 0.33 � 0.15a 0.42 � 0.13aGlycogen (�mol�100 g�1) 901 � 174 630 � 101 390 � 106a 345 � 62a 162 � 34a

a p � 0.05 versus corresponding basal.

Physiological Role of AMPK in Skeletal Muscle

SEPTEMBER 4, 2009 • VOLUME 284 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 23929

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

during exercise was also impaired �35% in �2-KD70% micewhen compared with WT70% (Fig. 3C).An increase in exercise intensity tended to increase Kg in the

soleus ofWTmice (p � 0.07 forWT70% versusWT45%; supple-mental Fig. S2A), whereasKg inWT70%was significantly greaterthanWT45% in SVL (supplemental Fig. S2B). These results par-alleled findings observed for Rg in soleus (supplemental Fig.S2C) and SVL (supplemental Fig. S2D). As with the gastrocne-

mius, Kg in soleus and SVL was impaired �30% in �2-KD70%mice when compared withWT70%; however, soleus and SVL Rgwas similar between �2-KD70% and WT70%.

Taken together, these findings show that glucose concentra-tion-dependent (Rg) and -independent (Kg) indices ofMGU areimpaired in �2-KDmice during exercise in vivo compared withWT mice exercising at the same relative intensity. The findingthat MGUwas greater inWT70% compared withWT45% showsfor the first time that the 2-[14C]DGmethod (38) can be used todetermine the effect of different exercise intensities onmultiplemuscle groups in vivo.Indices of LCFA clearance (Kf) and uptake (Rf) are shown in

supplemental Fig. S3. Kf increased to similar rates during exer-cise in soleus, gastrocnemius, and SVL of �2-KD70% andWT70%. In WT45% Kf responses were generally reduced. Rf sig-nificantly increased in soleus, gastrocnemius, and SVL of�2-KD70%. In WT70%, Rf significantly increased in soleus andgastrocnemius, whereas Rf was elevated in gastrocnemius ofWT45%. Given that Kf and Rf increased normally in response toexercise in �2-KD mice, it can be concluded that AMPK�2 isnot essential for skeletal muscle LCFA uptake during exercisein vivo. This is in agreementwith previous studies performed exvivo (20, 21).Percent Cardiac Output to Skeletal Muscle Is Altered in

�2-KD Mice at Rest and during Exercise in Vivo—Under basalconditions %QG was �2.5-fold greater in �2-KD mice com-pared with WT mice (Fig. 3D). Exercise increased %QG inWT45% (�4.5-fold) andWT70% (�4-fold). Exercise did not alter%QG in �2-KD70%. The glucose TEI did not differ between�2-KD and WT mice at rest (Fig. 3E). Exercise increased theglucose TEI to a similar extent in �2-KD70% and WT70%, dem-onstrating that the impairment in MGU seen in the gastrocne-mius of �2-KD70% compared with WT70% during exercise was

likely due to reduced substratedelivery (i.e. %QG). The TEI did notincrease in WT45%, demonstratingthat in WT mice the extraction ofglucose by skeletal muscle is ac-celerated as exercise intensityincreases.Cardiac Fuel Uptake and Func-

tion during Exercise in Vivo Are NotImpaired in �2-KD Mice—CardiacKg was similar between genotypes atrest, and exercise did not signifi-cantly increase Kg in any group(supplemental Fig. S4A). Cardiac Rgwas also similar between genotypesat rest, and exercise significantlyincreased cardiac Rg in �2-KD70%and WT70% (supplemental Fig.S4B). Cardiac Rg did not increaseduring exercise in WT45% and wassignificantly less than cardiac Rg in�2-KD70% and WT70%. No signifi-cant differences were observed withrespect to Kf or Rf in cardiac musclebetween any of the three groups

FIGURE 2. Arterial blood glucose (A), plasma NEFAs (B), plasma insulin (C),and plasma lactate (D) at rest and during exercise in 16-week-old chow-fed C57BL/6J mice expressing a WT or kinase-dead form of AMP-acti-vated protein kinase �2 (�2-KD) in cardiac and skeletal muscle. Follow-ing a 1-h fast, chronically catheterized mice performed a maximum of 30 minof running on a motorized treadmill, and arterial blood was sampled at timesshown. Data are mean � S.E. for n � 7–9 per group. †, p � 0.05 versus corre-sponding basal value; **, p � 0.01 versus �2-KD70%; f, main effect for time, p �0.05; § main effect for group, p � 0.01.

FIGURE 3. Plasma 2-[14C]DG clearance (A), gastrocnemius glucose clearance (Kg; B), glucose uptake(Rg; C), percent cardiac output to gastrocnemius (%QG; D), and the glucose TEI (E) in 16-week-old chow-fed C57BL/6J mice expressing a WT or kinase-dead (KD) form of AMP-activated protein kinase �2 incardiac and skeletal muscle. Following a 1-h fast, mice were given a bolus injection of 13 �Ci of 2-[14C]DG, andRg and Kg values were calculated as described (see “Experimental Procedures”). Following a final blood sample,fluorescent microspheres were injected into the carotid artery to determine %QG (see “Experimental Proce-dures”). Data are mean � S.E. for n � 8 –9 per groups except for D and E where n � 5–7 per group. *, p � 0.05versus corresponding WT; **, p � 0.05 versus WT70%; ***, p � 0.05 versus �2-KD70%; †, p � 0.05 versus corre-sponding sedentary group.

Physiological Role of AMPK in Skeletal Muscle

23930 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 36 • SEPTEMBER 4, 2009

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

(data not shown). Heart rate (683 � 12 versus 669 � 9 versus650 � 5 beats�min�1 for WT45%, WT70%, and �2-KD70%,respectively) and cardiac output (14 � 1 versus 14 � 1 versus15 � 1 ml�min�1 for WT45%, WT70%, and �2-KD70%, respec-tively) were similar between groups in response to exercise.Thus, a kinase-dead AMPK�2 subunit in cardiac muscle doesnot adversely affect substrate uptake or cardiac function inresponse to exercise.NOS Expression and Activity Are Reduced in Skeletal Muscle

of �2-KD Mice—In skeletal muscle, AMPK has been shown tointeract with the endothelial and neuronal isozymes of NOS(eNOS and nNOS, respectively) (8, 40). In gastrocnemius of�2-KDmice, expression of the skeletalmuscle isoformof nNOS(nNOS�) was reduced �35% compared with WT mice (Fig.4A). Gastrocnemius eNOS expression was similar betweengenotypes (0.05 � 0.01 versus 0.04 � 0.00 arbitrary units forWT and �2-KD, respectively), whereas expression of inducibleNOS was also similar (0.32 � 0.05 versus 0.37 � 0.02 arbitraryunits forWT and�2-KD, respectively). nNOS� expression wasalso impaired in SVL of �2-KDmice (0.77 � 0.09 versus 0.50 �0.05 arbitrary units forWT and �2-KD, respectively, p � 0.02).In the basal state, total NOS activity was �35% lower in gas-

trocnemius of �2-KD mice (Fig. 4B). Exercise increased NOSactivity in WT70% but not in either WT45% or �2-KD70%. Basal

NOS activity was also impaired in SVL muscle of �2-KD mice(supplemental Fig. S5); however, exercise did not alter SVLNOS activity in any group, a finding that may be related to lessrecruitment of thismuscle (i.e. attenuatedRg andKg when com-paredwith gastrocnemiusmuscle). Thus, AMPK is required forfull expression of nNOS�, as well as NOS activity at rest and inresponse to exercise. The observation that NOS activityincreased in gastrocnemius of WT70% but not WT45% showsthat NOS activity is sensitive to exercise intensity.Activities of Specific Electron Transport Chain (ETC) Com-

plexes Are Reduced in Skeletal Muscle of �2-KD Mice—Thefinding that exercise capacity, VO2peak, andATP generation areimpaired, although changes in arterial lactate levels are accel-erated in �2-KD mice during exercise despite normal extrac-tion of glucose in skeletal muscle, led us to hypothesize thatmitochondrial function is impaired in these mice. Support forthis hypothesis comes from the finding that a reduction innNOS� protein expression, such as seen in the presentstudy, is associated with impaired activity of enzymesinvolved in skeletal muscle OXPHOS (41, 42). As shown inTable 2, complex I and complex IV activities of the ETC weresignificantly impaired in sedentary �2-KD mice when com-pared with WT mice, whereas no changes were observed forcomplex I � III, II, or II � III activities. Identical findings wereobserved if complex activities were normalized to citrate syn-thase levels, which did not differ between genotypes (50 � 7versus 52 � 11 �mol�min�1�mg�1 for WT and �2-KD, respec-tively). Thus, the impairment in skeletal muscle ETC com-plexes in �2-KD mice was not because of a nonspecific reduc-tion in mitochondrial content, a finding that is in agreementwith previous observations demonstrating no alteration inmitochondrial density, DNA, and other markers of mitochon-drial content and biogenesis in gastrocnemius muscle ofuntrained �2-KD mice (43).

DISCUSSION

This study supports for the first time in vivo the hypothesisthat AMPK is a critical mediator of the metabolic response toexercise.We demonstrate that AMPK regulates skeletalmusclemetabolism in vivo at multiple levels, with the overall resultbeing that a defect in AMPK�2 subunit activity in skeletal mus-cle grossly impairs exercise tolerance. Without a functionallyactive AMPK�2 subunit, glucose uptake during exercise in vivois impaired in different skeletal muscle groups of �2-KD micecompared with WT littermate mice exercising at the same rel-ative intensity. This may be due in part to impaired substrate

FIGURE 4. NOS protein expression (A) and NOS activity (B) in gastrocnemiusmuscle of 16-week-old chow-fed C57BL/6J mice expressing a WT or kinase-dead (KD) form of AMP-activated protein kinase �2 in cardiac and skeletalmuscle. A, SDS-PAGE was performed on 75 �g of whole cell lysate from gastro-cnemius muscle. B, NOS activity at rest and in response to exercise was deter-mined in gastrocnemius muscle as described (see “Experimental Procedures”).Data are mean � S.E. for n � 5 per group. *, p � 0.05 versus corresponding WT; **,p � 0.05 versus WT70%; †, p � 0.05 versus corresponding sedentary group.

TABLE 2Electron transport chain complex activities in gastrocnemius muscleof 16-week-old chow-fed C57BL/6J mice expressing a WT orkinase-dead form of AMP-activated protein kinase �2 (�2-KD) incardiac and skeletal muscleData are mean � S.E. for n � 5–6 per group. Activities are expressed asnmol�min�1�mg�1.

Complex Enzyme WT �2-KD

I NADH:ubiquinone oxidoreductase 74.9 � 7.4 51.0 � 5.9aI � III NADH:cytochrome c oxidoreductase 11.1 � 3.7 6.6 � 2.6II Succinate:ubiquinone oxidoreductase 10.7 � 2.1 9.5 � 2.7II � III Succinate:cytochrome c oxidoreductase 13.4 � 3.0 8.0 � 1.7IV Cytochrome c oxidoreductase 144.7 � 15.2 72.3 � 9.4a

a p � 0.05 versus corresponding WT.

Physiological Role of AMPK in Skeletal Muscle

SEPTEMBER 4, 2009 • VOLUME 284 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 23931

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

delivery to exercising muscle (estimated via %QG) as the glu-cose TEI was not different between the two genotypes. SpecificETC complex activities in skeletal muscle were also impaired in�2-KD mice. Taken together with the findings of decreasedskeletal muscle ATP concentrations, greater arterial lactateaccumulation, and reductions in VO2peak duringmaximal exer-cise in�2-KDmice, our findings suggest that the exercise intol-erance in �2-KDmice is the result of impaired energy-produc-ing oxidative pathways (see Fig. 5).The novel finding that complex I and complex IV activities of

the ETC were impaired in the gastrocnemius of �2-KD micereveals new insight regarding the role of AMPK in skeletalmus-cle, and it provides a mechanism that could account for or con-tribute to the exercise intolerance observed in the �2-KDmouse. Complex I and complex IV represent the proximal anddistal ETC complexes, respectively, and thus play an integralrole in OXPHOS and the generation of ATP. A deficiency incomplex I activity will lead to excess levels of NADH and a lackof NAD�, resulting in impairedKrebs cycle function and elevatedblood lactate (44), the latter being observed in�2-KDmice duringexercise in this study. A deficiency in complex IV activity wouldimpair theprotongradient required for subsequentATPsynthesis(45), explaining the accelerated net ATP degradation observed inskeletal muscle of �2-KD mice during exercise in vivo. Impor-tantly, the changes in complex I and complex IV activities in�2-KD mice occurred despite similar levels of citrate synthaseactivity when comparedwithWTmice. This agreeswith previousfindings showing thatmitochondrialdensity,mitochondrialDNA,cytochrome c protein expression, �-aminolevulinate synthasemRNA expression, and peroxisome proliferator-activated recep-tor � coactivator-1� mRNA expression are similar in gastrocne-

miusmuscle of untrained�2-KD andWT mice (43). Thus, a functionallyinactive AMPK�2 subunit is suffi-cient to impair mitochondrial func-tion,without adversely alteringmark-ers of muscle mitochondrial content.Although OXPHOS capacity was

impaired in skeletal muscle of�2-KD mice, it is unclear whetherthe �2-KD subunit per se wasdirectly responsible for this phe-nomenon. A novel finding withimportant implications was thatnNOS� protein expression wasimpaired in skeletal muscle of�2-KD mice. This finding is sup-ported by the close associationbetween AMPK�2 and nNOS�(40). A decrease in nNOS� proteinexpression has been associated withimpairments in OXPHOS. Indeed,in skeletal muscle of patients withamyotrophic lateral sclerosis,reduced nNOS� expression ishighly associated with impairedETC complex activities (42). Simi-larly, in skeletalmuscle of nnos��/�

mice, ETC complex activities are reduced (41). Thus, theimpairments in OXPHOS within skeletal muscle of �2-KDmice may be due to a direct impairment of AMPK or indirecteffects mediated by reductions in nNOS� protein expression.The reduced nNOS� expression may have also caused an

impairment in muscle blood flow, as %QG did not increase inresponse to exercise in �2-KD mice, whereas an �4-foldincrease was observed inWT70% andWT45%. It has been shownthat vasodilation in response to mild exercise is significantlyimpaired in animal models where nNOS� is partially impairedor ablated in skeletal muscle (46). Likewise, Lau et al. (47) haveshown that �50% of contraction-induced arteriolar dilation invitro is dependent on nNOS�. Conversely, restoring nNOS� atthe sarcolemma of skeletal muscle significantly improves theexercise-induced increase in skeletalmuscle perfusion (48). It iswell known that contracting muscle releases nitric oxide (NO)(49, 50). Given that NOS activity in gastrocnemius of �2-KDmice was impaired in response to intense exercise, it is a plau-sible hypothesis that NO efflux from �2-KD mice was alsoimpaired. NO is a potent stimulator of vasodilation (51), and assuch impaired NOS activity during exercise may have also sup-pressed arteriolar relaxation in �2-KD mice. Aside fromnNOS�, it has been shown that the gastrocnemius of �2-KDmice contains significantly fewer capillaries compared withWT mice (52). Given that exercise normally causes a redistri-bution of blood flow toward contractingmuscle (53), fewer cap-illaries in the gastrocnemius of �2-KD mice might have alsoresulted in less blood flow to this tissue during exercise.We found for the first time that suppressed activation of

AMPK�2 in skeletal muscle during exercise in vivo was associ-ated with �60 and �35% reductions in concentration-inde-

FIGURE 5. Proposed model describing the role of skeletal muscle AMPK�2 during exercise in vivo. Ourresults show that skeletal MGU during exercise is dependent on AMPK�2 activation, as mice expressing �2-KDhave impaired MGU when compared with WT mice at the same relative exercise intensity. The impaired MGUin �2-KD mice is at least partially because of reduced vasodilation, which arises from an inability of AMPK toactivate NOS and thus stimulate NO production. The impairment in AMPK�2 activation and/or reductions inthe skeletal muscle isoform of neuronal NOS (nNOS�) also attenuate mitochondrial function. This reducesmitochondrial ATP generation and diverts glucose toward anaerobic ATP generation, resulting in elevatedplasma lactate levels. The whole body phenotype of these impairments is a reduction in exercise tolerance.G-6-P, glucose 6-phosphate.

Physiological Role of AMPK in Skeletal Muscle

23932 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 36 • SEPTEMBER 4, 2009

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

pendent and -dependent indices of MGU, respectively. Theseimpairments became apparent when work intensity was nor-malized to the same relative work rate inWT and �2-KDmice.As mentioned above, a novel observation in �2-KD mice wasthat %QG did not increase in response to exercise. As a result,the impairment in MGU in �2-KDmice could at least partiallybe ascribed to a reduction in vascular glucose delivery to thecontracting muscle. In line with this theory, we demonstratedthat the glucose TEI in gastrocnemius of �2-KDmice was sim-ilar to WT mice exercising at the same relative intensity, sug-gesting that the muscle had adequate capacity to extract glu-cose from the blood.A key aspect of this studywas utilizingWTmice exercising at

the same relative and absolute levels as �2-KD mice. To date,we are unaware of any study that has controlled for this variablewhen utilizing exercise as a means to amplify metabolic signalsin rodents. Dzamko et al. (21) recently reported that VO2 andwhole body substrate oxidation (assessed via indirect calorim-etry) was similar between �2-KD and WT mice exercising atthe same absolute running speed. This agrees with findingsfrom this present study; however, the findings of Dzamko et al.(21) do not account for the difference in relative exercise inten-sity. Our exercise stress test demonstrated that at the sameabsolute running speed �2-KD mice were exercising at agreater percentage of VO2peak. This is evidenced by the elevatedcellular stress in skeletal muscle of �2-KD mice (i.e. AMPfreeand AMPfree:ATP) compared with WT mice at the same abso-lute running speed, observations also observed inAMPK�2�/�

and WT mice exercising at the same absolute speed (54). Fur-thermore, in the present study Kg and Rg were elevated in SVLof �2-KD70% compared with WT45% demonstrating that thismuscle group, comprised primarily of fast glycolytic fibers, wasrecruited to a greater extent in �2-KD mice. Based on theseobservations, we propose that the relative exercise intensityshould be compared in rodent exercise studies, as is generallydone in human exercise studies. Performing an exercise stresstest in rodents provides valuable data pertaining to maximumrunning speed and running time and facilitates interpretationof subsequent studies examining physiological responses to anacute bout of exercise in vivo.

Aside fromexamining the role ofAMPK�2 in skeletalmuscleduring exercise, we also addressed the role of AMPK�2 in car-diacmuscle. Expression of the�2-KD transgene is driven by themuscle creatine kinase promoter (18), which is present in car-diac and skeletal muscle (55). The muscle creatine kinase pro-moter activity, and thus expression of the �2-KD transgene, ismuch lower in cardiac muscle (18). Nevertheless, cardiac func-tion of�2-KDmice has been shown to be significantly impairedin metabolically challenged states such as ischemia (56), and ithas been suggested that the exercise intolerance of �2-KDmice may be due to impairments in cardiac function asopposed to skeletal muscle defects (21, 57). In this studychanges in cardiac glucose and LCFA uptake, heart rate, andcardiac output in �2-KDmice were similar toWTmice exer-cising at the same relative intensity. Thus, a functionallyinactive AMPK�2 subunit in cardiac muscle does not appearto impair substrate uptake or cardiac function during phys-iological exercise conditions.

In conclusion, we show for the first time that exercise per-formed in vivo by�2-KDmice elicits a phenotype characterizedby impaired VO2peak, exercise intolerance, enhanced ATP deg-radation in skeletal muscle, and lactic acidosis. At the samerelative exercise intensity, MGU is impaired in �2-KD micecompared withWTmice. This is not because of attenuation inthe fractional extraction of glucose by skeletal muscle but likelyto impaired vascular glucose delivery to skeletal muscle. Wealso show that AMPK regulates nNOS� protein expression andNOS activity, as well asmitochondrial function in skeletal mus-cle. Based on existing literature (46), it is likely that the effects ofimpaired AMPK activation on vascular and mitochondrialfunction are to an extentmediated by changes in nNOS�. Thus,our findings demonstrate novel roles for AMPK in skeletalmuscle and provide new insight into the role of AMPK duringphysiological exercise. Our findings have implications forchronic metabolic disease states such as obesity and type 2 dia-betes, which are characterized by suppressed skeletal muscleAMPK�2 activity during exercise.

Acknowledgments—We thank Prof. Morris Birnbaum (University ofPennsylvania) for kindly supplying the �2-KDmice used for breeding.We thankDrs. ZhiZhangWang and Jeffrey Rottman of the VanderbiltMouseMetabolic Phenotyping Center, Cardiovascular Pathophysiol-ogy Core, for performing the echocardiography. We also thank Asso-ciate Professor Rodney Snow (Deakin University, Victoria, Australia)for helpful discussions regarding the metabolite analyses.

REFERENCES1. Jorgensen, S. B., and Rose, A. J. (2008) Front. Biosci. 13, 5589–56042. Hardie, D. G. (2007) Nat. Rev. Mol. Cell Biol. 8, 774–7853. Corton, J. M., Gillespie, J. G., and Hardie, D. G. (1994) Curr. Biol. 4,

315–3244. Kurth-Kraczek, E. J., Hirshman,M. F., Goodyear, L. J., andWinder,W.W.

(1999) Diabetes 48, 1667–16715. Koistinen, H. A., Galuska, D., Chibalin, A. V., Yang, J., Zierath, J. R., Hol-

man, G. D., andWallberg-Henriksson, H. (2003)Diabetes 52, 1066–10726. Merrill, G. F., Kurth, E. J., Hardie, D. G., andWinder, W.W. (1997) Am. J.

Physiol. 273, E1107–E11127. Bergeron, R., Russell, R. R., 3rd, Young, L. H., Ren, J.M.,Marcucci,M., Lee,

A., and Shulman, G. I. (1999) Am. J. Physiol. 276, E938–E9448. Shearer, J., Fueger, P. T., Vorndick, B., Bracy,D. P., Rottman, J. N., Clanton,

J. A., and Wasserman, D. H. (2004) Diabetes 53, 1429–14359. Jørgensen, S. B., Viollet, B., Andreelli, F., Frøsig, C., Birk, J. B., Schjerling, P.,

Vaulont, S., Richter, E. A., and Wojtaszewski, J. F. (2004) J. Biol. Chem.279, 1070–1079

10. Fujii, N.,Hirshman,M. F., Kane, E.M.,Ho, R. C., Peter, L. E., Seifert,M.M.,and Goodyear, L. J. (2005) J. Biol. Chem. 280, 39033–39041

11. Lefort, N., St-Amand, E., Morasse, S., Cote, C. H., and Marette, A. (2008)Am. J. Physiol. Endocrinol. Metab. 295, E1447–E1454

12. Jørgensen, S. B., Treebak, J. T., Viollet, B., Schjerling, P., Vaulont, S.,Wojtaszewski, J. F., and Richter, E. A. (2007) Am. J. Physiol. Endocrinol.Metab. 292, E331–E339

13. Kemp, B. E., Stapleton, D., Campbell, D. J., Chen, Z. P., Murthy, S.,Walter,M., Gupta, A., Adams, J. J., Katsis, F., Van Denderen, B., Jennings, I. G.,Iseli, T., Michell, B. J., and Witters, L. A. (2003) Biochem. Soc. Trans. 31,162–168

14. Winder, W. W., and Hardie, D. G. (1999) Am. J. Physiol. 277, E1–E1015. Chen, Z. P., Stephens, T. J., Murthy, S., Canny, B. J., Hargreaves, M., Wit-

ters, L. A., Kemp, B. E., and McConell, G. K. (2003) Diabetes 52,2205–2212

16. Musi, N., Hayashi, T., Fujii, N., Hirshman,M. F.,Witters, L. A., andGood-

Physiological Role of AMPK in Skeletal Muscle

SEPTEMBER 4, 2009 • VOLUME 284 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 23933

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

year, L. J. (2001) Am. J. Physiol. Endocrinol. Metab. 280, E677–E68417. Wojtaszewski, J. F., Nielsen, P., Hansen, B. F., Richter, E. A., and Kiens, B.

(2000) J. Physiol. 528, 221–22618. Mu, J., Brozinick, J. T., Jr., Valladares, O., Bucan, M., and Birnbaum, M. J.

(2001)Mol. Cell 7, 1085–109419. Sakamoto, K., McCarthy, A., Smith, D., Green, K. A., GrahameHardie, D.,

Ashworth, A., and Alessi, D. R. (2005) EMBO J. 24, 1810–182020. Miura, S., Kai, Y., Kamei, Y., Bruce, C. R., Kubota, N., Febbraio, M. A.,

Kadowaki, T., and Ezaki, O. (2009)Am. J. Physiol. Endocrinol. Metab. 296,E47–E55

21. Dzamko, N., Schertzer, J. D., Ryall, J. G., Steel, R., Macaulay, S. L., Wee, S.,Chen, Z. P., Michell, B. J., Oakhill, J. S.,Watt, M. J., Jørgensen, S. B., Lynch,G. S., Kemp, B. E., and Steinberg, G. R. (2008) J. Physiol. 586, 5819–5831

22. Fujii, N., Seifert, M. M., Kane, E. M., Peter, L. E., Ho, R. C., Winstead, S.,Hirshman, M. F., and Goodyear, L. J. (2007) Diabetes Res. Clin. Pract. 77,S92-S98

23. Sriwijitkamol, A., Coletta, D. K., Wajcberg, E., Balbontin, G. B., Reyna,S. M., Barrientes, J., Eagan, P. A., Jenkinson, C. P., Cersosimo, E., De-Fronzo, R. A., Sakamoto, K., and Musi, N. (2007) Diabetes 56, 836–848

24. De Filippis, E., Alvarez, G., Berria, R., Cusi, K., Everman, S., Meyer, C., andMandarino, L. J. (2008)Am. J. Physiol. Endocrinol.Metab. 294,E607–E614

25. Sigal, R. J., Kenny, G. P., Wasserman, D. H., Castaneda-Sceppa, C., andWhite, R. D. (2006) Diabetes Care 29, 1433–1438

26. Knight, D. R., Poole, D. C., Schaffartzik, W., Guy, H. J., Prediletto, R.,Hogan, M. C., and Wagner, P. D. (1992) J. Appl. Physiol. 73, 1114–1121

27. Ayala, J. E., Bracy, D. P., McGuinness, O. P., andWasserman, D. H. (2006)Diabetes 55, 390–397

28. Rottman, J. N., Ni, G., Khoo, M., Wang, Z., Zhang, W., Anderson, M. E.,and Madu, E. C. (2003) J. Am. Soc. Echocardiogr. 16, 1150–1157

29. Ayala, J. E., Bracy, D. P., Julien, B. M., Rottman, J. N., Fueger, P. T., andWasserman, D. H. (2007) Diabetes 56, 1025–1033

30. Folch, J., Lees, M., and Sloane Stanley, G. H. (1957) J. Biol. Chem. 226,497–509

31. Shearer, J., Coenen, K. R., Pencek, R. R., Swift, L. L.,Wasserman, D.H., andRottman, J. N. (2008) Lipids 43, 703–711

32. Morgan, C. R., and Lazarow, A. (1965) Diabetes 14, 669–67133. Lowry, O. H., and Passonneau, J. V. (1972)A Flexible System of Enzymatic

Analysis, Academic Press, New York34. Lee-Young, R. S., Palmer, M. J., Linden, K. C., LePlastrier, K., Canny, B. J.,

Hargreaves, M., Wadley, G. D., Kemp, B. E., and McConell, G. K. (2006)Am. J. Physiol. Endocrinol. Metab. 291, E566–E573

35. Birch-Machin, M. A., and Turnbull, D. M. (2001) Methods Cell Biol. 65,97–117

36. Kwong, L. K., and Sohal, R. S. (2000) Arch. Biochem. Biophys. 373, 16–2237. Barrientos, A. (2002)Methods 26, 307–316

38. Kraegen, E. W., James, D. E., Jenkins, A. B., and Chisholm, D. J. (1985)Am. J. Physiol. 248, E353–E362

39. Maxwell, A. J., Schauble, E., Bernstein, D., and Cooke, J. P. (1998) Circu-lation 98, 369–374

40. Stephens, T. J., Chen, Z. P., Canny, B. J., Michell, B. J., Kemp, B. E., andMcConell, G. K. (2002) Am. J. Physiol. Endocrinol. Metab. 282,E688–E694

41. Schild, L., Jaroscakova, I., Lendeckel, U., Wolf, G., and Keilhoff, G. (2006)FASEB J. 20, 145–147

42. Soraru, G., Vergani, L., Fedrizzi, L., D’Ascenzo, C., Polo, A., Bernazzi, B.,and Angelini, C. (2007) Neuropathol. Appl. Neurobiol. 33, 204–211

43. Zong, H., Ren, J.M., Young, L. H., Pypaert,M.,Mu, J., Birnbaum,M. J., andShulman, G. I. (2002) Proc. Natl. Acad. Sci. U.S.A. 99, 15983–15987

44. Munnich, A., and Rustin, P. (2001) Am. J. Med. Genet. 106, 4–1745. Yoshikawa, S., Muramoto, K., Shinzawa-Itoh, K., Aoyama, H., Tsukihara,

T., Shimokata, K., Katayama, Y., and Shimada, H. (2006)Biochim. Biophys.Acta 1757, 1110–1116

46. Kobayashi, Y. M., Rader, E. P., Crawford, R. W., Iyengar, N. K., Thedens,D. R., Faulkner, J. A., Parikh, S. V.,Weiss, R.M., Chamberlain, J. S., Moore,S. A., and Campbell, K. P. (2008) Nature 456, 511–515

47. Lau, K. S., Grange, R. W., Isotani, E., Sarelius, I. H., Kamm, K. E., Huang,P. L., and Stull, J. T. (2000) Physiol. Genomics 2, 21–27

48. Lai, Y., Thomas, G. D., Yue, Y., Yang, H. T., Li, D., Long, C., Judge, L.,Bostick, B., Chamberlain, J. S., Terjung, R. L., and Duan, D. (2009) J. Clin.Invest. 119, 624–635

49. Hirschfield, W., Moody, M. R., O’Brien, W. E., Gregg, A. R., Bryan, R. M.,Jr., and Reid,M. B. (2000)Am. J. Physiol. Regul. Integr. Comp. Physiol. 278,R95–R100

50. Balon, T. W., and Nadler, J. L. (1994) J. Appl. Physiol. 77, 2519–252151. Vallance, P., Collier, J., and Moncada, S. (1989) Lancet 2, 997–100052. Zwetsloot, K.A.,Westerkamp, L.M.,Holmes, B. F., andGavin, T. P. (2008)

J. Physiol. 586, 6021–603553. Ross, R. M., Wadley, G. D., Clark, M. G., Rattigan, S., andMcConell, G. K.

(2007) Diabetes 56, 2885–289254. Klein, D. K., Pilegaard, H., Treebak, J. T., Jensen, T. E., Viollet, B., Schjer-

ling, P., and Wojtaszewski, J. F. (2007) Am. J. Physiol. Endocrinol. Metab.293, E1242–E1249

55. Bruning, J. C., Michael, M. D., Winnay, J. N., Hayashi, T., Horsch, D.,Accili, D., Goodyear, L. J., and Kahn, C. R. (1998)Mol. Cell 2, 559–569

56. Russell, R. R., 3rd, Li, J., Coven, D. L., Pypaert,M., Zechner, C., Palmeri,M.,Giordano, F. J., Mu, J., Birnbaum, M. J., and Young, L. H. (2004) J. Clin.Invest. 114, 495–503

57. Mu, J., Barton, E. R., and Birnbaum, M. J. (2003) Biochem. Soc. Trans. 31,236–241

Physiological Role of AMPK in Skeletal Muscle

23934 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 284 • NUMBER 36 • SEPTEMBER 4, 2009

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Ayala, Owen P. McGuinness and David H. WassermanRobert S. Lee-Young, Susan R. Griffee, Sara E. Lynes, Deanna P. Bracy, Julio E.

in VivoResponse to Exercise Skeletal Muscle AMP-activated Protein Kinase Is Essential for the Metabolic

doi: 10.1074/jbc.M109.021048 originally published online June 12, 20092009, 284:23925-23934.J. Biol. Chem. 

  10.1074/jbc.M109.021048Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

  http://www.jbc.org/content/suppl/2009/06/12/M109.021048.DC1

  http://www.jbc.org/content/284/36/23925.full.html#ref-list-1

This article cites 56 references, 16 of which can be accessed free at

by guest on August 17, 2020

http://ww

w.jbc.org/

Dow

nloaded from