Myocardial ischemia and increased heart work modulate the phosphorylation state of eukaryotic elongation factor-2.

Protein synthesis, in particular peptide chain elongation, is an energy-consuming biosynthetic process. AMP-activated protein kinase (AMPK) is a key regulatory enzyme involved in cellular energy homeostasis. Therefore, we tested the hypothesis that, as in liver, it could mediate the inhibition of protein synthesis by oxygen deprivation in heart by modulating the phosphorylation of eukaryotic elongation factor-2 (eEF2), which becomes inactive in its phosphorylated form. In anoxic cardiomyocytes, AMPK activation was associated with an inhibition of protein synthesis and an increase in phosphorylation of eEF2. Rapamycin, an inhibitor of the mammalian target of rapamycin (mTOR), did not mimic the effect of oxygen deprivation to inhibit protein synthesis in cardiomyocytes or lead to eEF2 phosphorylation in perfused hearts, suggesting that AMPK activation did not inhibit mTOR/p70 ribosomal protein S6 kinase (p70S6K) signaling. Human recombinant eEF2 kinase (eEF2K) was phosphorylated by AMPK in a time- and AMP-dependent fashion, and phosphorylation led to eEF2K activation, similar to that observed in extracts from ischemic hearts. In contrast, increasing the workload resulted in a dephosphorylation of eEF2, which was rapamycin-insensitive, thus excluding a role for mTOR in this effect. eEF2K activity was unchanged by increasing the workload, suggesting that the decrease in eEF2 phosphorylation could result from the activation of an eEF2 phosphatase.

Protein synthesis, in particular peptide chain elongation, is an energy-consuming biosynthetic process. AMPactivated protein kinase (AMPK) is a key regulatory enzyme involved in cellular energy homeostasis. Therefore, we tested the hypothesis that, as in liver, it could mediate the inhibition of protein synthesis by oxygen deprivation in heart by modulating the phosphorylation of eukaryotic elongation factor-2 (eEF2), which becomes inactive in its phosphorylated form. In anoxic cardiomyocytes, AMPK activation was associated with an inhibition of protein synthesis and an increase in phosphorylation of eEF2. Rapamycin, an inhibitor of the mammalian target of rapamycin (mTOR), did not mimic the effect of oxygen deprivation to inhibit protein synthesis in cardiomyocytes or lead to eEF2 phosphorylation in perfused hearts, suggesting that AMPK activation did not inhibit mTOR/p70 ribosomal protein S6 kinase (p70S6K) signaling. Human recombinant eEF2 kinase (eEF2K) was phosphorylated by AMPK in a timeand AMP-dependent fashion, and phosphorylation led to eEF2K activation, similar to that observed in extracts from ischemic hearts. In contrast, increasing the workload resulted in a dephosphorylation of eEF2, which was rapamycin-insensitive, thus excluding a role for mTOR in this effect. eEF2K activity was unchanged by increasing the workload, suggesting that the decrease in eEF2 phosphorylation could result from the activation of an eEF2 phosphatase.
Protein synthesis, in particular peptide chain elongation, is an energy-consuming biosynthetic process, accounting for a large proportion of the oxygen requirements of cells (1). Protein synthesis is regulated via the phosphorylation/dephosphorylation of translation factors and ribosomal proteins (2). The mammalian target of rapamycin (mTOR) 1 phosphorylates eu-karyotic initiation factor 4E-binding protein-1 (4E-BP1), thereby relieving its inhibitory action on eukaryotic initiation factor 4E (eIF-4E), which can then bind the mRNA cap and stimulate protein synthesis (3). The control of translation by mTOR is also exerted at elongation by regulating the phosphorylation of eukaryotic elongation factor-2 (eEF2) (4). The phosphorylation of eEF2 at Thr-56 by a specific calcium-and calmodulin-dependent eukaryotic eEF2 kinase (eEF2K) leads to its inactivation (5). The p70 ribosomal protein S6 kinase (p70S6K) lies downstream of mTOR and phosphorylates Ser-366 of eEF2K, causing inactivation (4). Therefore, mTOR activation can result in a stimulation of protein synthesis by decreasing eEF2 phosphorylation. Indeed, one of the mechanisms by which insulin stimulates protein synthesis in the heart involves mTOR activation, the effect being blocked by the mTOR inhibitor, rapamycin, and by wortmannin, which blocks phosphatidylinositol 3-kinase (PI 3-kinase) upstream of mTOR (6). The stimulation of protein synthesis by insulin also involves a mTOR-independent mechanism via the phosphorylation and inactivation of glycogen synthase kinase-3. Glycogen synthase kinase-3 phosphorylates and inactivates the guanine nucleotide exchange factor, eukaryotic initiation factor 2B (eIF2B) (7).
Increasing the workload of heart stimulates protein synthesis, leading to cardiac hypertrophy (8). The stretch of the ventricular wall could be the mechanical trigger for this effect. Therefore, protein synthesis plays an important role in the adaptation of cell size and myofibrillar content to mechanical stress. However, the signal transduction pathway leading to the stimulation of protein synthesis due to an increased workload is unknown. Like insulin, increasing the workload stimulates glycolysis via a wortmannin-sensitive pathway (9,10). Activation of PI 3-kinase and mTOR could therefore be implicated in the stimulation of protein synthesis by increasing the workload, which was one of the aspects studied in this report.
In contrast to insulin and increased workload, oxygen deprivation inhibits protein synthesis (11,12). In skeletal muscle, an inhibition of mTOR/p70S6K signaling following AMP-activated protein kinase (AMPK) activation has been proposed to explain the inhibition of protein synthesis by oxygen deprivation (13). AMPK acts as an energy and nutrient sensor in cells and is activated by an increased AMP:ATP ratio, as occurs in the absence of oxygen or in response to other cellular stresses (14,15). Once activated, AMPK stimulates ATP-producing pathways and inhibits energy-consuming processes (14 -16). In this * The work was supported by the Federal Program Interuniversity Poles of Attraction (Belgium), by the Directorate General Higher Education and Scientific Research, French Community of Belgium, by the Fund for Medical Scientific Research (Belgium), and by European Union contract QLG1-CT-2001-01488 (AMPDIAMET). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Both authors contributed equally to this work and should be considered as joint first authors.
Preparation, Incubation, and Treatment of Cardiomyocytes-Cardiomyocytes were prepared from male Wistar rats (9). Cells (about 20 -30 mg of wet weight/ml) were incubated at 37°C in 20-ml vials with constant agitation and treated as described in the legend for Fig. 1. Following incubation, extracts were prepared using a standard lysis buffer containing a mixture of protease and phosphatase inhibitors (18).
Perfusion Protocols of Isolated Hearts-Hearts from fed male Wistar rats were perfused by the Langendorff method (9) or in ''working conditions'' (19) at 37°C in a recirculating system with 100 ml of Krebs-Henseleit bicarbonate buffer containing 5 mM glucose and 2.5 mM CaCl 2 in equilibrium with a gas phase containing either 95% O 2 and 5% CO 2 (normoxic conditions) or 95% N 2 and 5% CO 2 (anoxic conditions). The perfusion protocols and preload and afterload of the working hearts are described in the legends for the figures. Hearts were freeze-clamped at the indicated times for the preparation of extracts for immunoblotting and enzyme assay. Samples of the frozen left ventricules were homogenized in 9 volumes of lysis buffer on ice, using an Ultra-Turrax homogenizer (18).
Enzyme Assay and in Vitro Phosphorylation-AMPK was purified (18) and assayed as described (16). Human GST⅐eEF2K (1.5 g) was incubated with AMPK (1 unit/ml) and 0.1 mM [␥-32 P]MgATP (specific radioactivity, 1000 cpm/pmol) in a final volume of 50 l at 30°C in the presence and absence of 0.2 mM AMP. At the indicated times, aliquots (5 l) were removed for SDS-PAGE and phosphorimaging for the measurement of 32 P incorporation. Heart extracts or human recombinant GST⅐eEF2K were assayed for eEF2K activity at 30°C in a reaction mixture containing 20 mM Hepes-Tris-Mes (pH 7), 3 mM MgCl 2 , 2 mM EDTA, 1 mM dithiothreitol, 2% (v/v) glycerol with 4.14, 1.625, or 0.23 mM CaCl 2 /5 mM EGTA to give free Ca 2ϩ concentrations of 1 M, 100 nM, or 10 nM, respectively, at pH 7, and 10 g/ml calmodulin, 70 pmol purified eEF2, 0.1 mM [␥-32 P]MgATP (specific radioactivity, 1000 cpm/ pmol) in a final volume of 50 l. After 5 min, the reactions were stopped by the addition of Laemmli sample buffer for SDS-PAGE in gels containing 10% (w/v) acrylamide. 32 P incorporation was calculated by phosphorimaging the dried gels. One unit of protein kinase activity corresponds to the formation of one nmol/min under the conditions of the various assays.
Immunoblotting-Proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. The percentage of acrylamide used for the gels depended on the protein under study. Acrylamide concentrations were 15% (w/v) for 4E-BP1, 12% (w/v) for eIF2␣ FIG. 1. AMPK activation leads to eEF2 phosphorylation in rat cardiomyocytes and in Langendorff-perfused rat hearts. After 10 min of equilibration, cardiomyocytes were incubated under anoxic conditions (95% N 2 , 5% CO 2 ) or treated with 1 M oligomycin. AMPK activity (A) and eEF2 phosphorylation (B) were measured after 10 min of anoxia or oligomycin treatment. In A, the rate of protein synthesis was measured by the incorporation of [ 14 C]leucine into trichloroacetic acid-precipitable material. AMPK activity and protein synthesis rates were significantly different between the three groups (p Ͻ 0.05, n ϭ 3). In B, extracts (50 g) were subjected to SDS-PAGE and immunoblotted (IB) with anti-phospho (Thr-56) eEF2 antibody (upper panel). The membrane was stripped and probed with anti-full-length antibody that does not distinguish the phosphorylated protein (lower panel). In C and D, after 20 min of equilibration, hearts were submitted to no-flow ischemia or anoxia by changing the gas phase (95% N 2 , 5% CO 2 ) for 10 min. AMPK activity (C) and eEF2 phosphorylation (D) were assessed as described above. In C, AMPK activities are the means Ϯ S.E. from at least five hearts. Results from three separate experiments are presented in D. *, a statistically significant difference with respect to the controls (p Ͻ 0.05). and ERK1/2, and 10% (w/v) for eEF2 and eIF2B. The membranes were probed with the relevant primary antibodies, and immunoreactive bands were detected with protein G-peroxidase by enhanced chemiluminescence.
Phosphorylation Site Identification by Mass Spectrometry-Phosphorylated bands corresponding to the GST⅐eEF2K (M r 119000) were cut from Coomassie Blue-stained gels, concentrated, and digested with 1 g of sequencing grade trypsin as described (20). Peptides were separated by reverse-phase narrow-bore HPLC at a flow rate of 200 l/min, and radioactive peaks were analyzed by nano-electrospray ionization tandem mass spectrometry (nano-ESI-MS/MS) in a LCQ Deca XP Plus ion-trap mass spectrometer (ThermoFinnigan, San Jose, CA) (20).
Metabolite Measurements and Other Methods-AMP and ATP (21), phosphocreatine (PCr), and creatine (Cr) (22) were measured in neutralized perchloric acid extracts as cited. Protein was estimated by the method of Bradford (23) using bovine serum albumin as a standard. Protein synthesis was measured as described (17).

RESULTS
Oxygen Deprivation Inhibits Protein Synthesis, Activates AMPK, and Increases eEF2 Phosphorylation in Rat Cardiomyocytes-Isolated cardiomyocytes were incubated for 10 min with 1 M oligomycin, an inhibitor of oxidative phosphorylation, or under anoxic conditions (N 2 replacing O 2 in the gas phase). Protein synthesis was inhibited 48% by anoxia and 22% by oligomycin treatment, and there was an inverse relationship between the rates of protein synthesis and AMPK activation (Fig. 1A). Incubation of cardiomyocytes under anoxic conditions increased the phosphorylation state of Thr-56 of eEF2, the effect of anoxia being more pronounced than that of oligomycin (Fig. 1B), in line with the extent of AMPK activation.
Ischemia and Anoxia Increase eEF2 Phosphorylation in Langendorff-perfused Rat Hearts-Similarly, in Langendorff-perfused rat hearts, no-flow ischemia and anoxia induced AMPK activation (Fig. 1C) and an increase in eEF2 phosphorylation (Fig. 1D). Phosphorylation of eEF2 was detectable within 2 min and was maximal between 5 and 10 min of no-flow ischemia and correlated with AMPK activation (not shown). The phosphorylation states of initiation factors 4E-BP1 and eIF2␣, initiation factors that are also known to be regulated by phosphorylation, were unchanged under anoxia or ischemia for 10 min (not shown).
Rapamycin Has No Effect on the Rate of Protein Synthesis and eEF2 Phosphorylation in Rat Cardiomyocytes and Langendorff-perfused Rat Hearts-In isolated cardiomyocytes, rapamycin inhibited mTOR as it blocked Thr-389 phosphorylation of p70S6K induced by a mixture of insulin (0.1 M), leucine (1 mM), and glutamine (10 mM) (Fig. 2A). The phosphorylation of Thr-389 of p70S6K by mTOR is required for its activation (24). Rapamycin did not mimic the effect of oxygen deprivation to inhibit protein synthesis in cardiomyocytes (Fig. 2C), suggesting that this inhibition did not result from a reduction in mTOR/p70S6K signaling. Moreover, the level of eEF2 phosphorylation in ischemic-perfused hearts was unaffected by rapamycin, although there was a small increase in eEF2 phosphorylation in the controls (Fig. 2B). The failure of rapamycin to inhibit basal protein synthesis, despite a slight increase in eEF2 phosphorylation, indicates that the mTOR/p70S6K pathway was not active and could not be implicated in the control of eEF2 phosphorylation under these conditions.
AMPK Phosphorylates and Activates eEF2K-The upstream kinase that phosphorylates Thr-56 of eEF2 is a highly specific calcium-and-calmodulin-dependent protein kinase called eEF2K (5). When measured in the presence of calcium, eEF2K was activated 3-fold in extracts prepared from isolated Langendorff-perfused rat hearts subjected to 10 min of no-flow ischemia (Fig. 3A). However, in the absence of calcium, some eEF2K activity was detectable, as observed in crude extracts from certain cells (25). Recombinant human GST⅐eEF2K was phosphorylated by AMPK in a time-and AMP-dependent fashion to a stoichiometry of 0.7 mol of phosphate incorporated per mol of enzyme (Fig. 3B). Phosphorylation resulted in a 2-2.5-fold increase in eEF2K activity that was restricted to the total activity measured in the presence of calcium (Fig. 3C). In an attempt to identify phosphorylation sites, GST⅐eEF2K was maximally phosphorylated with AMPK and [␥-32 P]MgATP. Following SDS-PAGE, bands corresponding to eEF2K were digested with trypsin, and the resulting peptides were separated by reverse-phase HPLC. A radioactive peak was detected, and its labeling was increased by AMP (not shown). Analysis of the labeled peak by nano-ESI-MS/MS allowed the identification of Ser-78 as one phosphorylation site for AMPK in the fragmentation spectrum of the phosphopeptide 69 YSSSGSPAN p SFHFK 82 , where "p" represents the phosphorylated residue (not shown).
Regulation of eEF2 Phosphorylation by Anoxia under Working Conditions-Since eEF2K is a Ca 2ϩ -dependent protein kinase, basal eEF2 phosphorylation could be different in a model of isolated working heart, in which a higher calcium transient has been observed. Therefore, isolated working hearts were perfused with an afterload of 100 cm of H 2 O and a preload of 20 cm of H 2 O (physiological conditions) under normoxia and were then submitted to anoxia. Heart function was monitored continuously with a pressure transducer, indicating a drop in aortic pressure within the first minutes of anoxia (Fig. 4A). After 5 min of anoxia, AMPK was maximally activated about 4-fold (Fig. 4B), and at 10 min of anoxia, the AMP/ATP ratio was increased 2-3-fold (Fig. 4C), whereas the PCr/Cr ratio was reduced by about 30% (Fig. 4D). An increase in eEF2 phosphorylation was observed after 5 and 10 min of anoxia (Fig. 4E) as in unloaded hearts (Fig. 1D). . The membrane was stripped and probed with anti-full-length antibody that does not distinguish the phosphorylated protein (lower panel). In B, eEF2 phosphorylation was assessed in Langendorff-perfused hearts submitted to no-flow ischemia for 10 min in the presence or absence of rapamycin 250 nM. In C, protein synthesis was measured in cardiomyocytes as described in the legend for Fig. 1. eEF2 Is Dephosphorylated during Increased Heart Work-In perfused working hearts, we also studied the effects of increased heart work on eEF2 phosphorylation, as this condition is known to increase protein synthesis (8). Heart work was increased by raising the afterload from 60 (low load) to 120 cm of H 2 O (high load). In contrast to anoxia, an acute increase in pressure load resulted in a decrease in eEF2 phosphorylation, which occurred within 1 min and persisted over 15 min (Fig.  5A). This could explain the increase in protein synthesis in response to an increased workload (8), as eEF2 phosphorylation inhibits protein synthesis. The effect of increasing the workload to dephosphorylate eEF2 was unaffected by wortmannin and rapamycin (Fig. 6A), which are inhibitors of PI 3-kinase and mTOR respectively. Surprisingly, Thr-389 phosphorylation of p70S6K was also decreased under high load conditions, suggesting that mTOR was inhibited (Fig. 6B). This contrasts with the situation in rat cardiomyocytes treated with insulin, in which p70S6K is activated and protein synthesis is stimulated, both effects being blocked by rapamycin (6) (see also Fig. 2A). No significant decrease in eEF2K activity was observed after 15 min of high load (Fig. 5B), suggesting that the decrease in eEF2 phosphorylation might be due to the activation of an eEF2 phosphatase. Taken together, these results show a decrease in eEF2 phosphorylation in response to in-creasing the workload, for which mTOR/p70S6K does not seem to play a role.
Increasing the workload resulted in a transient increase in 4E-BP1 phosphorylation, which was rapamycin-insensitive (Fig. 6C). The anti-phospho antibody used in these experiments recognizes Thr-36 and probably cross-reacts with Thr-45 (the surrounding sequences are almost identical). However, there was no change in eIF2␣ phosphorylation state (not shown). In a previous study, we showed that the stimulation of glycolysis by increasing the workload, like the metabolic effects of insulin, were wortmannin-sensitive and did not require AMPK (10). The MEK/ERK and PI 3-kinase pathways are both stimulated in response to insulin and implicated in the phosphorylation and inactivation of eEF2K. Therefore, we checked two downstream targets of these two pathways (ERK and eIF2B respectively) to exclude their implication in the dephosphorylation of eEF2 observed during the increased workload. The phosphorylation state of eIF2B tended to increase slightly on increasing the workload, and no changes in the phosphorylation states of ERK1/2 were detected (not shown). DISCUSSION In this report, we show that AMPK activation correlates with an increase in eEF2 phosphorylation in anoxic cardiomyocytes FIG. 3. Effect of AMPK on eEF2K activity. In A, eEF2K was assayed in extracts (15 g) from Langendorff-perfused hearts as described under "Materials and Methods" in the presence or absence of Ca 2ϩ at pH 7. The values are the means Ϯ S.E. from at least three different hearts. In B, GST⅐human eEF2K (1.5 g) was incubated with AMPK (1 unit/ml) and [␥-32 P]MgATP (specific radioactivity 1000 cpm/pmol) in a final volume of 50 l at 30°C in the presence and absence of 0.2 mM AMP. At the indicated times, aliquots (5 l) were removed for SDS-PAGE and phosphorimaging for measurement of 32 P incorporation. In C, GST⅐human eEF2K (1.5 g) was incubated with 0.2 mM AMP and 1 mM MgATP in a final volume of 25 l for 20 min at 30°C in the presence or absence of AMPK (1 unit/ml). Aliquots (2.5 l) were assayed for eEF2K activity for 5 min in reaction mixtures at the indicated Ca 2ϩ concentrations at pH 7 as described under "Materials and Methods." The results are the means of three separate determinations. *, a statistically significant difference with respect to the controls (p Ͻ 0.05). and ischemic heart. The increase in eEF2 phosphorylation during ischemia occurred before any change in the phosphorylation state of 4E-BP1 and was not mimicked by rapamycin. This suggests that the inhibition of mTOR signaling is not involved in the AMPK-induced inhibition of basal protein synthesis, confirming previous observations in hepatocytes (17) and cardiomyocytes (26). In hepatocytes incubated with amino acids, p70S6K becomes activated, and this effect was antagonized by AMPK activation (27,28). Also, AMPK activation by 5-amino-4-imidazolecarboxamide riboside in rat skeletal muscle in vivo inhibited both the mTOR pathway and protein synthesis (13). Therefore, AMPK activation inhibits mTOR signaling and protein synthesis, but only if the pathway is first switched on, e.g. by amino acids or insulin.
In extracts from anoxic hepatocytes (17) or ischemic-perfused hearts (Fig. 3A), eEF2K was activated, suggesting that AMPK might phosphorylate and activate eEF2K. In some cells, treatments that lead to a rise in cAMP activate eEF2K by increasing the calcium-independent activity (29), whereas in other systems, the total activity measured in the presence of calcium is increased (30). The situation is further complicated by the proposed existence of isozymes of eEF2K (31), which could exhibit different kinetic properties. Here we show that in ischemic hearts, AMPK activation leads to an increase in eEF2K activity, which is mainly reflected in an increase in the total, calcium-dependent activity of the extracts (Fig. 3A). However, we cannot exclude that an inhibition of eEF2 phosphatase by AMPK could also contribute toward an increase in overall phosphorylation of eEF2. In vitro, AMPK phosphorylated a FIG. 4. Effect of anoxia on AMP/ATP, PCr/Cr ratio, AMPK activity, and eEF2 phosphorylation in isolated working hearts. Hearts were perfused at 100 cm of H 2 O afterload and 20 cm of H 2 O preload. After 20 min of equilibration, hearts were submitted to anoxia by changing the gas phase (95% N 2 , 5% CO 2 ) for 10 min. In A, the aortic pressure recording from a representative experiment is shown. Hearts were freeze-clamped at the indicated times for measurement of AMPK activity (B) and for the determination of the intracellular AMP/ATP (C) and PCr/Cr (D) concentration ratios. The values are the means Ϯ S.E. for at least five hearts. *, a statistically significant difference with respect to the controls (p Ͻ 0.05). In E, Thr-56 phosphorylation of of eEF2 was assessed in extracts (50 g of protein) as described in the legend for Fig. 1. IB, immunoblot. human recombinant GST⅐eEF2K (Fig. 3B), which led to eEF2K activation (Fig. 3C). Moreover, eEF2K activation was restricted to the calcium-dependent activity (Fig. 3C), similar to the activation of eEF2K seen in ischemic hearts (Fig. 3A). When the AMPK-phosphorylated human GST⅐eEF2K was digested with trypsin and analyzed by HPLC and mass spectrometry, Ser-78 was identified as a phosphorylation site for AMPK in a peak whose labeling was increased in the presence of AMP. This serine residue is conserved in eEF2K from human, mouse, and rat. However, the HPLC profile indicated the existence of other sites that could account for the change in activity.
In addition to activation by AMPK, a fall in intracellular pH, as occurs during ischemia (32), could participate in increasing the phosphorylation state of eEF2, as eEF2K activity rises as the pH drops below 7 (33). The increase in eEF2 phosphorylation is therefore likely to be due to a combination of phosphorylation/activation of eEF2K by AMPK and stimulation of eEF2K activity as a result of the fall in pH. In cortical neurons, oxidative stress induced by H 2 O 2 treatment has been shown to inhibit protein synthesis by a process involving the phosphorylation of eEF2, which it was proposed could have been sec-ondary to a rise in intracellular calcium (34). However, it is noteworthy that AMPK is activated in CCL13 cells treated with H 2 O 2 (20).
Calcium levels are constantly varying in cardiac myocytes, whereas they are at basal levels in most other cells. Therefore, a mechanism that is independent of calcium would be important to ensure increased phosphorylation of eEF2, since a decreased calcium transient is observed within a few minutes in the absence of oxygen, the effect being more pronounced when the intracellular pH falls (35,36). In contrast, increasing the workload increases the calcium transient of cardiomyocytes. However, perfusion under working conditions did not result in eEF2 phosphorylation, and the effect of anoxia on eEF2 phosphorylation was still apparent. This suggests that an additional mechanism maintains eEF2 in its dephosphorylated form in normoxic conditions, thereby counteracting the effect of calcium on eEF2K activity and maintaining protein synthesis.
In contrast to the situation in hypoxia, an acute increase in workload resulted in a decrease in eEF2 phosphorylation. This could be involved in the stimulation of protein synthesis in response to an increased workload (8) and reinforces the hy-FIG. 6. Effects of wortmannin and rapamycin on the dephosphorylation of eEF2 during acute increase in afterload. Hearts were submitted to an increase in afterload (from 60 to 120 cm of H 2 O) after 15 min of equilibration and freeze-clamped at 1 and 15 min under high load conditions. Wortmannin or rapamycin was added to the perfusion medium after 5 min of equilibration at final concentrations of 100 and 250 nM, respectively. Me 2 SO was the control vehicle. In A and C, eEF2 and 4E-BP1 (Thr-36 and Thr-45) phosphorylation were assessed in extracts (50 g) as described in the legend for Fig. 1. IB, immunoblot. In B, phosphorylation of Thr-389 in p70S6K was measured at 1 and 15 min under high load conditions. pothesis that AMPK is not required for the acute adaptation of the myocardium to changes in pressure load (10). The decrease in eEF2 phosphorylation by increasing the workload was wortmannin-and rapamycin-insensitive, and no changes in the phosphorylation states of ERK1/2 were detected (not shown). Also, the phosphorylation state of eIF2B, which lies downstream of glycogen synthase kinase-3 in the insulin-signaling pathway, did not decrease on increasing the workload, which contrasts with the effect of insulin in decreasing eIF2B phosphorylation (7,37). Furthermore, there was no change in eEF2K activity on increasing the workload (Fig. 5B). Therefore, we suggest that the decrease in eEF2 phosphorylation state could result from the activation of an eEF2 phosphatase. The protein phosphatase that specifically dephosphorylates eEF2 was proposed to be protein-serine/threonine phosphatase 2A. Indeed, protein-serine/threonine phosphatase 2A was suggested to be involved in the dephosphorylation of eEF2 in response to agonists such as angiotensin II, endothelin-I, and phenylephrine, which have been shown to mimic the hypertrophic response to an increased hemodynamic load (38). It was shown that protein-serine/threonine phosphatase 2A and the phosphorylation and inactivation of eEF2K by p90 rsk /p70S6K1 probably contribute toward the dephosphorylation of eEF2 and thus to an increase in protein synthesis in cardiac myocytes under these conditions (39).
In conclusion, during ischemia AMPK not only participates in the stimulation of glycolysis but is also involved in the inhibition of protein synthesis via eEF2K activation and the subsequent phosphorylation and inactivation of eEF2. In contrast, glycolysis and protein synthesis (at least for eEF2 phosphorylation) during increased workload are controlled by two separate pathways, namely PI 3-kinase activation for the stimulation of glycolysis and activation of a phosphatase for eEF2 (Fig. 7).