AMP-activated Protein Kinase Suppresses Protein Synthesis in Rat Skeletal Muscle through Down-regulated Mammalian Target of Rapamycin (mTOR) Signaling*

AMP-activated protein kinase (AMPK) is viewed as an energy sensor that acts to modulate glucose uptake and fatty acid oxidation in skeletal muscle. Given that protein synthesis is a high energy-consuming process, it may be transiently depressed during cellular energy stress. Thus, the intent of this investigation was to examine whether AMPK activation modulates the translational control of protein synthesis in skeletal muscle. -ri-bonucleoside were The activity unchanged compared with whereas 2 AMPK was significantly increased resulted in a reduction in protein synthesis to 45% of the control value. This depression was associated with decreased activation of protein kinases in the mammalian target of rapamycin (mTOR) signal transduction pathway as evi-denced by reduced phosphorylation of protein kinase B on Ser 473 , mTOR on Ser 2448 , ribosomal protein S6 on Thr 389 , and eukaryotic initiation factor eIF4E-bind-ing protein on Thr . A reduction in eIF4E associated with eIF4G to 10% of the control value In contrast, eIF2B activity remained unchanged in response to AICAR treatment and therefore would not appear to contribute to the depression in protein synthesis. m M (cid:1) -glycero-phosphate, 0.1 m M phenylmethylsulfonyl fluoride, 1 m M benzamidine, 1 m M dithiothreitol (DTT), and 0.5 m M sodium vanadate. The remaining homogenate was centrifuged at 10,000 (cid:4) g for 10 min at 4 °C. The resulting supernatant was combined with an equal volume of SDS sample buffer and then subjected to protein immunoblot analysis as described previously (18). Samples were analyzed for the phosphorylation status of 4E-BP1 on Thr 37 , S6K1 on Thr 389 , the (cid:2) -subunit of eIF2 (eIF2 (cid:2) ) on Ser 51 , PKB on Ser 473 , and mTOR on Ser 2448 by Western blot analysis using phosphorylation site-specific antibodies. With the excep- tion of the anti-phospho-eIF2 (cid:2) (Ser 51 ) antibody, which was obtained from BioSource International, Hopkinton, MA, the anti-phosphospecific antibodies were obtained from Cell Signaling Technology, Beverly, MA. Total PKB and mTOR were measured by Western blot analysis using antibodies that recognize both the phosphorylated and unphosphoryl-ated proteins. No change in PKB or mTOR content was observed under any of the experimental conditions. For quantitation of the amount of eIF4G present in the eIF4G (cid:1) eIF4E complex, eIF4E was immunoprecipi- tated from 10,000 (cid:4) g supernatants using a monoclonal antibody. Samples were subjected to immunoblot analysis using a polyclonal antibody to eIF4G to assess the association of eIF4G with eIF4E (18). Results were normalized to the amount of eIF4E in the immunoprecipitates.

the energy status of the cell and mediating subsequent metabolic events. AMPK has been referred to as an energy-sensing/ signaling protein within the cell that responds to changes in the ratio of ATP/AMP as well as phosphocreatine/creatine (1,2). Changes in the cellular energy state activate AMPK through various mechanisms involving allosteric regulation of AMPK, activation by an upstream AMPK kinase, and diminished activity of phosphatases (3). AMPK activation increases glucose uptake and fatty acid oxidation in muscle (4) as well as up-regulates expression of various metabolic genes (e.g. the glucose transporter, GLUT4, uncoupling protein-3, and cytochrome c) (5)(6)(7). Consequently AMPK serves as a sensor/modulator of intermediary metabolism by directing cellular events to increase energy availability and sustain high energy phosphate levels.
Research using in vitro systems has shown that AMPK can be activated under artificial conditions such as treatment with high fructose or 2-deoxyglucose, heat shock, and inhibitors of oxidative phosphorylation (3). Pharmacological use of 5-aminoimidazole-4carboxamide 1-␤-D-ribonucleoside (AICAR) has been commonly utilized to directly activate AMPK without altering cellular concentrations of ATP, ADP, and AMP (8). Additionally, starvation and endurance exercise result in increased activity of AMPK in skeletal muscle (9 -11). Exercise alters the adenine nucleotide ratios and serves as a physiological context for AMPK activation. Recently specific catalytic isoforms of AMPK (␣ 1 and ␣ 2 ) have been shown to be differentially regulated by exercise intensity with ␣ 2 AMPK exhibiting greater metabolic sensitivity compared with the ␣ 1 isoform (10,12).
The concept of AMPK acting as an energy sensor suggests that cellular processes that utilize ATP, and are not vital to short term survival, are potential control points for regulation by the protein kinase (13). Thus, a hierarchy may exist for ensuring sufficient energy availability during an energetic stress and the anabolic process of protein synthesis may be diminished to support that dominant function. The acute control of global rates of protein synthesis is predominantly executed at the level of translational initiation with the modulation of various eukaryotic initiation factors (eIFs) (14).
The protein kinase referred to as the mammalian target of rapamycin (mTOR), which serves as a convergence point for signaling by growth factors and amino acids to the mRNA binding step of translation initiation is involved in modulation of the phosphorylation of the binding protein for the eukaryotic initiation factor 4E, i.e. 4E-BP1. It also acts to control the phosphorylation status of the 70-kDa ribosomal protein S6 kinase (S6K1).
Modulation of these translation initiation events allows for more immediate control of protein synthesis and is responsive to changes associated with acute metabolic or nutritional alterations. Therefore, the present investigation examined whether or not activation of AMPK by treatment with AICAR would depress translational initiation. We hypothesized that during an apparent cellular energy stress induced by AICAR, increased AMPK activity would diminish translation initiation and attenuate protein synthesis.

EXPERIMENTAL PROCEDURES
Animals-Animal facilities and the experimental protocol were reviewed and approved by the Institutional Animal Care and Use Committee of The Pennsylvania State University College of Medicine. Male Sprague-Dawley rats (ϳ175 g) were kept on a 12-h light:dark cycle with food (Harlan-Teklad Rodent Chow, Madison, WI) and water provided freely.
AICAR Injections-Rats were injected subcutaneously with AICAR (1 mg/g of body weight) in sterile 0.9% NaCl, or controls were given an equivalent volume of 0.9% NaCl (n ϭ 6 -10 animals per group). A flooding dose (1.0 ml/100 g body of weight) of L-[2,3,4,5,6-3 H]phenylalanine (150 mmol/liter) was injected via the tail vein 50 min after the subcutaneous injections for the measurement of rates of synthesis of total mixed muscle protein (15). Rats were sacrificed by decapitation 1 h after receiving the subcutaneous injection. Previous research has demonstrated that AMPK activity peaks between 1 and 2 h following AICAR injection (16). The gastrocnemius muscles were rapidly dissected and frozen in liquid nitrogen with the total time elapsed for freezing tissue being less than 60 s.
Measurement of Protein Synthesis-The fractional rate of synthesis (K s ) was estimated from the rate of incorporation of radioactive phenylalanine into total mixed muscle protein using the specific radioactivity of serum phenylalanine as representative of the precursor pool (17). The actual time for incorporation of the radiolabeled phenylalanine into protein was taken as the time elapsed from injection until freezing of muscle in liquid nitrogen.
Analysis of Protein Kinase B (PKB)/mTOR Signaling to eIFs-Gastrocnemius muscles were weighed and homogenized in 7 volumes of buffer containing 20 mM HEPES (pH 7.4), 100 mM potassium chloride, 0.2 mM EDTA, 2 mM EGTA, 50 mM sodium fluoride, 50 mM ␤-glycerophosphate, 0.1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 1 mM dithiothreitol (DTT), and 0.5 mM sodium vanadate. The remaining homogenate was centrifuged at 10,000 ϫ g for 10 min at 4°C. The resulting supernatant was combined with an equal volume of SDS sample buffer and then subjected to protein immunoblot analysis as described previously (18). Samples were analyzed for the phosphorylation status of 4E-BP1 on Thr 37 , S6K1 on Thr 389 , the ␣-subunit of eIF2 (eIF2␣) on Ser 51 , PKB on Ser 473 , and mTOR on Ser 2448 by Western blot analysis using phosphorylation site-specific antibodies. With the exception of the anti-phospho-eIF2␣(Ser 51 ) antibody, which was obtained from BioSource International, Hopkinton, MA, the anti-phosphospecific antibodies were obtained from Cell Signaling Technology, Beverly, MA. Total PKB and mTOR were measured by Western blot analysis using antibodies that recognize both the phosphorylated and unphosphorylated proteins. No change in PKB or mTOR content was observed under any of the experimental conditions. For quantitation of the amount of eIF4G present in the eIF4G⅐eIF4E complex, eIF4E was immunoprecipitated from 10,000 ϫ g supernatants using a monoclonal antibody. Samples were subjected to immunoblot analysis using a polyclonal antibody to eIF4G to assess the association of eIF4G with eIF4E (18). Results were normalized to the amount of eIF4E in the immunoprecipitates.
Measurement of eIF2B Activity-The guanine nucleotide exchange activity of eIF2B in skeletal muscle was measured by the exchange of [ 3 H]GDP bound to eIF2 for nonradioactively labeled GDP as described previously (19).
Statistical Analysis-Data are presented as means Ϯ S.E. Results were compared using a two-tailed, two-sample (equal variance) Student's t test to assess differences between treatment groups. Statistical significance was set at an ␣ level of p Ͻ 0.05.

RESULTS
The experimental model utilized in the studies reported herein is based on the use of the chemical AICAR to artificially activate AMPK. AICAR is internalized by the cell and subsequently phosphorylated to form an AMP analog, termed ZMP, that acts as a metabolic activator of both AMPK and AMPK kinase without altering the adenine nucleotide ratios within the cell (8). To assess the effectiveness of AICAR treatment in the present study, the activity of the ␣ 1 and ␣ 2 isoforms of AMPK were measured following immunoprecipitation of the kinase from skeletal muscle of rats administered either AICAR or vehicle alone 1 h before analysis. As shown in Fig. 1, activation of the AMPK ␣ 1 isoform was unchanged in AICARtreated rats compared with rats administered vehicle alone (control rats). In contrast, ␣ 2 AMPK activity was significantly elevated (p Ͻ 0.001) in AICAR-treated animals compared with controls.
To determine whether or not activation of ␣ 2 AMPK in response to AICAR treatment was associated with a change in protein synthesis, in vivo rates of the synthetic process were measured using the flooding dose technique (15). The results show that in AICAR-treated rats, protein synthesis in skeletal muscle was depressed to 55% of the value (p Ͻ 0.02) observed in control animals (Fig. 2). To examine potential mechanisms regulating the reduction in protein synthesis associated with increased ␣ 2 AMPK activity, several key regulatory steps in translation initiation were investigated.
The regulation of eIF2 is an important event during translation initiation in maintaining global rates of protein synthesis. The first step in translation initiation is the binding of methionyl-tRNA i to eIF2⅐GTP to form a ternary complex that subsequently binds to the 40 S ribosomal subunit. Formation of the ternary complex can be modulated by phosphorylation of eIF2 on the ␣-subunit by converting it into a competitive inhibitor of eIF2B activity (21). In the present study the phosphorylation state of eIF2␣ was examined by protein immunoblot analysis using an anti-phosphopeptide antibody that only recognizes eIF2␣ when it is phosphorylated on Ser 51 . The results show that the relative phosphorylation of eIF2␣ was reduced to 80% of the control value (p Ͻ 0.05) in AICAR-treated rats. Because the activity of eIF2B can be modulated by the phosphorylation state of eIF2␣, the guanine nucleotide exchange activity of eIF2B was measured in extracts of muscle from AICAR-treated and control animals. However, the activity of eIF2B was not significantly different between the two groups (0.063 Ϯ 0.009 and 0.056 Ϯ 0.008 pmol of GDP exchanged/min, respectively).  1. The effects of AICAR on AMPK ␣ 1 and ␣ 2 activity. Muscle homogenates were centrifuged at 14,000 ϫ g for 20 min at 4°C. Supernatants (150 g) were subjected to immunoprecipitation with specific antibodies to the ␣ 1 or ␣ 2 catalytic subunits of AMPK and BioMag goat anti-rabbit beads, and AMPK activity was measured in the immunoprecipitate as described under "Experimental Procedures" (n ϭ 6 per group). † †, p Ͻ 0.001 versus control group.

AMPK Down-regulates PKB/mTOR Signaling 23978
To further examine potential mechanisms involved in the depression of protein synthesis associated with ␣ 2 AMPK activation, the effect of AICAR treatment on eIF4G association with eIF4E was examined. As shown in Fig. 3A, the amount of eIF4G associated with eIF4E was decreased to ϳ10% of the control value (p Ͻ 0.01) in muscle from AICAR-treated rats. One mechanism for regulating the binding of eIF4G to eIF4E involves phosphorylation of 4E-BP1, which releases eIF4E from the inactive 4E-BP1⅐eIF4E complex and allows it to bind to eIF4G. In muscle from AICAR-treated rats, phosphorylation of 4E-BP1 on Thr 37 , a priming event for phosphorylation of the protein on additional residues that promote its dissociation from eIF4E, was significantly reduced (p Ͻ 0.01) (Fig. 3B). Phosphorylation of Thr 37 on 4E-BP1 is mediated by a protein kinase referred to as mTOR (22). Another downstream target of mTOR is the 70-kDa S6K1. Similar to its effect on 4E-BP1 phosphorylation, AICAR treatment reduced phosphorylation of Thr 389 on S6K1 to 5% of the value (p Ͻ 0.01) observed in control rats (Fig. 3C). Together the changes in 4E-BP1 and S6K1 phosphorylation suggest that the activity of mTOR was repressed in skeletal muscle of AICAR-treated rats.
The stimulation of the protein kinase activity of mTOR by growth factors such as insulin or insulin-like growth factor-I is mediated in part by phosphorylation of protein kinase B on Ser 473 , which results in its activation (23). PKB subsequently phosphorylates a residue (Ser 2448 ) on mTOR that is present in a domain that normally acts to repress mTOR protein kinase activity (24). To examine whether or not activation of AMPK might result in changes in PKB activity and thereby alter phosphorylation of mTOR, protein immunoblot analysis was performed using antibodies specific for mTOR phosphorylated on Ser 2448 and PKB phosphorylated on Ser 473 . As shown in Fig.  4, A and B, the relative phosphorylation of both mTOR on Ser 2448 and PKB on Ser 473 in AICAR-treated rats was proportionately decreased to ϳ40% of the control value (p Ͻ 0.05). DISCUSSION AMPK is recognized as having a well established role in the regulation of energy production and nutrient flux within skeletal muscle during periods of energetic stress. The present study provides the first in vivo evidence that AMPK activation directly affects translational initiation and protein synthesis in skeletal muscle and that these responses are mediated through the mTOR signaling pathway. The results also further establish AMPK as a unique energy sensor that not only modulates glucose and fatty acid metabolism but also appears to regulate, FIG. 3. eIF4E⅐eIF4G association and phosphorylation state of Thr 37 on 4E-BP1 and Thr 389 on S6K1 following AICAR treatment in skeletal muscle. A, eIF4E was immunoprecipitated from 10,000 ϫ g supernatants using a monoclonal antibody. Samples were subjected to immunoblot analysis using a polyclonal antibody to eIF4G and a monoclonal antibody to eIF4E. The results are expressed as a ratio of eIF4G to eIF4E (n ϭ 6 per group). B, phosphorylation of 4E-BP1 on Thr 37 was assessed using an anti-phospho-4E-BP1 antibody (n ϭ 10 per group). C, phosphorylation of S6K1 on Thr 389 was determined using an antiphospho-S6K1 antibody (n ϭ 10 per group). †, p Ͻ 0.01 versus control group. CON, control.

FIG. 4. Phosphorylation state of Ser 473 on PKB and Ser 2448 on mTOR in response to AICAR treatment in skeletal muscle.
A, phosphorylation of PKB on Ser 473 was determined using an anti-phospho-PKB antibody (n ϭ 10 per group). B, phosphorylation of mTOR on Ser 2448 was assessed using an anti-phospho-mTOR antibody (n ϭ 6 per group). *, p Ͻ 0.05 versus control group. CON, control.
FIG. 2. Fractional rate of skeletal muscle protein synthesis in response to AICAR. Protein synthesis was estimated from the rate of incorporation of radioactive phenylalanine into total mixed muscle protein using the specific radioactivity of serum phenylalanine as representative of the precursor pool as described under "Experimental Procedures" (n ϭ 6 per group). *, p Ͻ 0.05 versus control group.
AMPK Down-regulates PKB/mTOR Signaling 23979 in part, the anabolic functions of mRNA translation and mixed muscle protein synthesis. Our data reveal that ␣ 2 AMPK activation by AICAR treatment results in significant suppression of the synthesis of total mixed muscle protein. Various metabolic states involving acute and chronic energetic imbalances, such as fasting, exercise, anorexia, and cachexia, have all been shown to depress protein synthesis (25)(26)(27)(28)(29). Identifying a molecular signal that initiates changes in mRNA translation and protein synthesis under these conditions has remained elusive. The concept of AMPK acting as an energy sensor is reinforced by the present investigation by establishing increased AMPK activity as a regulator of skeletal muscle protein synthesis under conditions of an energetic stress.
The dramatic suppression of skeletal muscle protein synthesis observed in the present study was mediated by alterations in several key steps in translation initiation. Phosphorylation of Thr 37 on 4E-BP1 and Thr 389 on S6K1 was significantly reduced with AICAR, indicative of a decrease in global rates of protein synthesis as well as the synthesis of specific proteins involved with the translational apparatus (mRNAs containing a terminal oligopyrimidine tract adjacent to the m 7 GTP cap, i.e. TOP mRNA), respectively. The hypophosphorylated 4E-BP1 would be expected to sequester eIF4E and prevent its association with eIF4G (30). Indeed the profound decrease in eIF4E associated with eIF4G observed in AICAR-treated rats corroborates this idea. Because the binding of eIF4E⅐mRNA complex to the ribosome requires association of eIF4E with eIF4G, the decreased binding of eIF4G to eIF4E observed in AICARtreated rats would contribute to the depression in protein synthesis noted.
The role of eIF2 in regulating mRNA translation with AMPK activation appears less clear. A reduction in eIF2B activity would be suggestive of a decrease in global rates of protein synthesis, but eIF2B activity remained unchanged with AICAR treatment. Moreover, the small reduction in eIF2␣ phosphorylation is somewhat paradoxical given that such a change should be associated with enhanced ternary complex formation, Met-tRNA i binding to ribosomes, and protein synthesis. The possibility exists that the diminished eIF2␣ phosphorylation was part of a response that represented an attempt by the cell to maintain a basal level of protein synthesis to prevent extensive catabolism. Nonetheless the function of eIF2 may not be crucial to regulating translation initiation under these conditions. This study represents the first investigation to demonstrate alterations in mTOR phosphorylation and accompanying changes in 4E-BP1 and S6K1 using an in vivo model. Our results suggest that AMPK may signal through PKB to downregulate the activity of mTOR and its downstream effectors. The exact mechanism(s) by which AMPK modulates PKB/ mTOR phosphorylation is unknown; however, in addition to Ser 473 , phosphorylation of Ser 308 on PKB was also reduced 2 suggesting kinases upstream of PKB (e.g. PDK1 or PDK2) may be targets for AMPK. Moreover, the possible involvement of changes in phosphatase activity (e.g. PP1 or PP2A) cannot be ruled out.
A recent investigation has proposed that mTOR serves as an energy sensor by monitoring changes in ATP concentrations (31). However, significant decreases in ATP concentrations in vivo are difficult to demonstrate under physiological conditions. Additionally, AMPK activation appears to be more sensitive to alterations in the ratio of AMP/ATP and phosphocreatine/creatine than to absolute changes in ATP concentrations. Therefore, AMPK may act as a molecular signal to control mRNA translation depending on the cellular adenine nucleotide ratios.
In conclusion, this research identifies a new cellular function for AMPK by its ability to modulate skeletal muscle protein synthesis and the phosphorylation state of translation initiation factors upon activation. Given the high energy consumption associated with protein synthesis in the cell, this anabolic function may be suppressed while cellular energy is either conserved or partitioned to maintain ATP concentrations. Furthermore, gaining a mechanistic appreciation for the regulation of skeletal muscle protein synthesis during an acute or chronic energy stress may provide nutritional and/or therapeutic strategies for the treatment of metabolic diseases. Ultimately AMPK should be viewed with an integrated approach by understanding its role in cellular function and how these events can potentially influence whole body metabolism of carbohydrate, fat, and protein.