Although a diminished ability of tissues and organisms to tolerate stress is a clinically important hallmark of normal aging, little is known regarding its biochemical basis. Our goal was to determine whether age-associated changes in AMP-activated protein kinase (AMPK), a key regulator of cellular metabolism during the stress response, might contribute to the poor stress tolerance of aged cardiac and skeletal muscle. Basal AMPK activity and the degree of activation of AMPK by AMP and by in vivo hypoxemia (arterial Po2 of 39 mmHg) were measured in cardiac and skeletal muscle (gastrocnemius) from 5- and 24-mo-old C57Bl/6 mice. In the heart, neither basal AMPK activity nor its allosteric activation by AMP was affected by age. However, after 10 min of hypoxemia, the activity of α2-AMPK, but not α1-AMPK, was significantly higher in the hearts from old than from young mice (P < 0.005), this difference being due to differences in phosphorylation of α2-AMPK. Significant activation of AMPK in the young hearts did not occur until 30 min of hypoxemia (P < 0.01), stress that was poorly tolerated by the old mice (mortality = 67%). In contrast, AMPK activity in gastrocnemius muscle was unaffected by age or hypoxemia. We conclude that the age-associated decline in hypoxic tolerance in cardiac and skeletal muscle is not caused by changes in basal AMPK activity or a blunted AMPK response to hypoxia. Activation of AMPK by in vivo hypoxia is slower and more modest than might be predicted from in vitro and ex vivo experiments.
one of the characteristics of biological aging is a diminished ability of organisms to tolerate various forms of stress (26, 29). An impaired stress tolerance is also observed in isolated tissues such as cardiac and skeletal muscle, where the sensitivity to ischemic or oxidative stress increases with age (1, 4, 5, 12, 14, 19). This is a clinically important issue in light of the increasingly older population of the developed world, although little is known regarding its biochemical basis.
During myocardial ischemia the availability of high-energy phosphates may be more limited in aged than in young hearts (6, 25, 28). This raised the possibility that metabolic signaling dictated by the intracellular energy charge may be different in the aged tissue. In isolated hearts, the ATP depletion and subsequent AMP accumulation induced by ischemia or anoxia activate the AMP-activated protein kinase (AMPK), a central component of the cellular stress response that shifts metabolism toward ATP restoration (7, 15, 21). Dually activated by AMP and by upstream phosphorylation, this kinase promotes fatty acid oxidation, glucose uptake, and glycogenolysis while it inhibits anabolic processes such as fatty acid synthesis (21, 24). While ischemia and anoxia cause robust activation of AMPK in the isolated heart and skeletal muscle, the extent to which this occurs in vivo, where an intact neuroendocrine system can strongly influence cellular energetics, has not been established.
Recent data indicate that the activity of AMPK or its yeast homologue, Snf1, may be altered with age (23, 32). Moreover, genetic mutations in the AMPK genes cause severe dysfunction of cardiac and skeletal muscles, suggesting that alterations in AMPK have clinical consequences and may potentially contribute to the decline in stress tolerance observed with aging (2, 3). However, predicting in what direction AMPK activity might change with age, and how these changes would affect stress tolerance, is difficult. In vitro studies in yeast and human fibroblasts suggest that activity of AMPK increases with age (23, 32). However, some characteristics of aging such as reduced insulin sensitivity would be more consistent with decreased AMPK activity or a diminished ability to activate it (33). The exact role played by AMPK in stress tolerance in vivo is largely unknown.
Therefore, the main goal of this study was to determine whether normal aging affects basal AMPK activity and its degree of activation by in vivo acute hypoxia in cardiac and skeletal muscle. In vivo hypoxia was chosen as the stimulus not only because of its clinical relevance but also because this information is conspicuously lacking. Additionally, we assessed whether aging affects the ability of AMP to allosterically activate AMPK. When effects of age or hypoxia on AMPK activity were observed, the phosphorylation status of AMPK was assessed. The heart and gastrocnemius muscle were chosen for study so that comparisons between a highly oxidative muscle (the heart) and a highly glycolytic muscle (gastrocnemius) could be made.
MATERIAL AND METHODS
Animals and tissue collection.
Both adult (5 mo old) and old (24–26 mo old) C57Bl/6 mice were housed in the University of Wisconsin Medical School animal facility following National Institutes of Health guidelines for the use of laboratory animals. Only male mice were studied to eliminate the potentially confounding effects of cyclic variability in estrogen on the tissue of interest. Mice had ad libitum access to food and water until the time of death. On the day of death (animals were killed during the first half of their 12-h light cycle) each mouse was weighed and placed in an airtight container with 4% isoflurane in 100% oxygen. When a state of deep anesthesia was reached (complete loss of muscle tone), mice were aseptically intubated, placed on a 37°C heating pad, and then ventilated at 1.1 ml·min−1·g body wt−1 (Harvard Apparatus, Minivent) with 2% isoflurane in 100% oxygen. Some mice were maintained on isoflurane in 100% oxygen (hyperoxic groups), and others were switched to isoflurane in 7.9% O2 and nitrogen (hypoxia groups). Organs were collected after either 10 min of hyperoxia, 10 min of hypoxia, or 30 min of hypoxia. The heart and gastrocnemius muscle (in this order) were removed, freeze-clamped in liquid nitrogen, and stored at −80°C until assayed. In a separate group of adult mice, arterial blood gases [arterial Po2 (PaO2) and Pco2 (PaCO2)] were measured to ensure that hyperoxemia and hypoxemia were effectively achieved under these conditions (Table 1). Only adult mice were used for the blood gas experiments because age-associated changes in the murine lung that might affect blood gases during mechanical ventilation appear to be minimal (17, 22).
AMPK enzymatic activity.
For AMPK analysis, tissues were homogenized in buffer A (in mM: 30 HEPES, 2.5 EGTA, 3 EDTA, 20 KCl, 40 glycerolphosphate, 40 NaF, 4 NaPPi, 1 Na3VO4; and 0.1% Igepal CA-630, 32% glycerol, 2% protease inhibitor cocktail, Sigma P-8340) and centrifuged 10 min at 14,000 rpm in an Eppendorf 5415C microfuge, and the supernatant was mixed with an equal volume of buffer B (in mM: 30 HEPES, 2.5 EGTA, 3 EDTA, 70 KCl, 20 glycerolphosphate, 20 NaF, 4 NaPPi, 1 Na3VO4) and spun again. Two-hundred micrograms of protein from this second supernatant were immunoprecipitated with protein A/G agarose beads (Santa Cruz Biotech) and either the anti-α1-AMPK or anti-α2-AMPK antibody (Upstate) for 2 h at 4°C. The beads containing the immunoprecipitated AMPK complexes were washed, resuspended in reaction buffer (in mM: 40 HEPES, 80 NaCl, 5 MgCl2, 1 DTT) and incubated for 10 min at 37°C in a thermomixer in the presence of 0.2 mM SAMS peptide, 0.2 mM [32P]ATP (specific activity 1,500–4,000 cpm/pmol), and either 0 or 0.2 mM AMP for each sample. After incubation, the beads were quickly pelleted and an aliquot of each supernatant was spotted on P-81 phosphocellulose paper that was washed in 1% phosphoric acid and then in acetone and air-dried. The incorporated radioactivity was counted in a TriCarb 3000 beta scintillation counter. Day-to-day coefficient of variation (CV%) of the assay in an internal control (rat heart) was 18% (n = 5).
To investigate if changes in AMPK activity were due to changes in the phosphorylation state of the enzyme, we measured its activity in the presence of phosphatase PP2Cα, known to dephosphorylate and deactivate AMPK (10). For these experiments, tissues were homogenized and immunoprecipitated as before in HEPES buffer without phosphatase inhibitors, and the beads were resuspended in an assay mix containing 0.1 μU/μl PP2Cα (Upstate) and incubated at 30°C for 60 min. This dephosphorylation protocol was established by testing the extent to which different concentrations of PP2Cα, incubation times, and temperatures decreased AMPK activity relative to the nondephosphorylating conditions. We found that this protocol and concentration of PP2Cα caused the maximal decrease in AMPK activity achievable by dephosphorylation . The beads were washed and incubated in reaction mixture as before to measure the AMPK activity. A separate piece of the same tissue sample was homogenized and assayed under nondephosphorylating conditions for comparison.
To determine the effects of AMP and age on AMPK activity, two-way ANOVA was conducted for the data collected under hyperoxic conditions and for those collected after 10 min of hypoxemia. Likewise, to test for the effect of hypoxemia on AMPK activity within each age group, two-way ANOVA was conducted.
Arterial oxygenation and survival.
Arterial blood gases were measured during hyperoxic (100% inspired O2) and hypoxic (7.9% inspired O2) mechanical ventilation in a separate group of 5-mo-old mice. As shown in Table 1, we were able to reproducibly induce a moderate degree of arterial hypoxemia while controlling PaCO2 and pH. Based on published oxygen-hemoglobin dissociation curves, a PaO2 of 39 mmHg represents an arterial oxygen saturation of ∼75% (20). This degree of arterial hypoxemia was well tolerated for 10 min by both 5- and 24-mo-old mice (89% and 100% survival, respectively). However, whereas 100% of the young mice survived to 30 min of hypoxemia, only two of six old mice survived (33% survival), evidencing a decline in hypoxic tolerance with aging. This high mortality precluded us from collecting tissues and, therefore, biochemical data from the 30-min time point in the old mice.
Cardiac α1-AMPK activity.
Figure 1 summarizes the effects of AMP, age, and hypoxemia on cardiac α1-AMPK activity. Age had no effect on AMPK activity when tissues were collected during hyperoxemia or after 10 min of hypoxemia. Contrary to in vitro studies, 10 min of hypoxemia did not increase cardiac α1-AMPK activity significantly in either age group (7, 21). However, after 30 min of hypoxemia, cardiac α1-AMPK activity was significantly elevated relative to the hyperoxic condition in the young mice, although the degree of activation was modest (1.7-fold). Under all conditions, 200 μM AMP significantly increased cardiac α1-AMPK activity to the same extent in both age groups.
Cardiac α2-AMPK activity.
As in the case of α1-AMPK, cardiac α2-AMPK activity was not affected by age when tissues were collected during hyperoxic conditions (Fig. 2). However, when tissues were collected after 10 min of hypoxemia, AMPK activity was higher in old than in young mice (P < 0.005). It was not until after 30 min of hypoxemia that cardiac α2-AMPK was significantly elevated in the young mice (P < 0.01), as was the case with α1-AMPK. Again, under all conditions 200 μM AMP significantly increased cardiac α2-AMPK activity to the same extent in both age groups.
Dephosphorylation of cardiac α2-AMPK.
To determine if increases in cardiac α2-AMPK activity with hypoxemia and age were mediated by an increased degree of phosphorylation of AMPK, tissue homogenates were assayed in the absence and in the presence of phosphatase PP2Cα (Figs. 3 and 4) (10). Figure 3 shows a group of experiments in young mice in which activation of α2-AMPK after 30 min of hypoxemia was 2.7-fold higher than in hyperoxemic hearts (P < 0.05). When the same samples were assayed in the presence of the phosphatase PP2Cα, this difference was completely eliminated. This indicates that the hypoxia-induced differences in AMPK activity were due to differences in the level of AMPK phosphorylation.
Similarly, as shown in Fig. 4, after 10 min of hypoxemia, old hearts had α2-AMPK activities higher than young hearts (P < 0.05). When the assay was performed in the presence of PP2Cα, this difference was eliminated. These experiments indicated that both the age and the hypoxia-induced activation of AMPK that we observed in vivo occur via phosphorylation of AMPK.
Skeletal muscle (gastrocnemius) α1- and α2-AMPK activities.
In skeletal muscle from young mice, neither 10 nor 30 min of hypoxemia caused α1- or α2-AMPK activity to increase above the hyperoxemic value (Figs. 5 and 6). Similarly, in the old mice, hypoxemia failed to increase α1- or α2-AMPK activity. As was the case in the heart, the amount of allosteric activation of either AMPK isoform by 200 μM AMP was not affected by age during any of the experimental conditions.
Several pieces of indirect evidence suggest that AMPK, a serine/threonine kinase that is a key regulator of cellular metabolism, is affected by normal aging (23, 32). Our purpose was to determine, in mammalian cardiac and skeletal muscle, whether aging affects three aspects of AMPK activity, namely basal activity, responsiveness to in vivo hypoxemia, and allosteric activation by AMP.
Effects of age on cardiac AMPK activity.
Recent data in cultured human fibroblasts and yeast indicate that enzymatic activity of AMPK, and its yeast homologue Snf1, increases with age (23, 32). Here, we present data comparing basal AMPK activity in young and old mammalian hearts. In contrast to those reports in vitro, we found that age did not affect the basal activity of either the α1- or the α2-isoform of AMPK in the heart. It is worth noting that our surgical setup and freeze-clamp method for tissue harvest were intended to circumvent potential artifactual AMPK activation related to the respiratory depression that accompanies barbituric anesthesia (13). However, we cannot rule out the possibility that the intubation and ventilation procedures might themselves have caused some activation of AMPK. Therefore, we use the term “basal” to describe the degree of AMPK activity in the well-oxygenated tissues we studied. Our results suggest that the age-associated increases in AMPK activity reported in vitro are either cell-type specific or an effect only observed in cell culture, where aging is defined in terms of the ability of cells to divide, not the passage of time.
The ability of the heart to tolerate ischemia and hypoxia decreases with advanced age via unknown mechanisms (1, 4–6, 8, 14). Because activation of AMPK is thought to be an important compensatory response to ischemia, we hypothesized that with advanced age, the hypoxic activation of AMPK may be blunted, explaining, at least in part, the poor hypoxic tolerance of the aged heart (21). We found that whereas the hypoxic activation of the α1-isoform followed a similar time course at both ages, activation of the α2-isoform was slightly accelerated in the aged heart. These data indicate that a lack of AMPK activation does not play a role in the poor hypoxic tolerance of the old heart. Rather, the accelerated activation of the α2-isoform of AMPK in the aged heart could be regarded as an adaptive or compensatory response to the more severe energetic deterioration that others have reported (6, 25, 28, 31).
Similar to enzymes such as phosphofructose kinase-1 or glycogen phosphorylase b, AMPK is regulated by AMP, its major allosteric activator during acute energetic stress (15). The fact that aging diminishes the ability of AMP to allosterically activate PFK raises the possibility that this might also occur with AMPK (30). However, no difference in the degree of activation of AMPK by 200 μM AMP was found between young and old mice. In fact, the degree of AMPK activation by AMP was remarkably similar across all of the conditions we studied.
Acute hypoxia-induced activation of cardiac AMPK.
In isolated hearts, acute ischemia or hypoxia causes a robust increase in AMPK activity (7, 21). For example, Beauloye et al. (7) showed that 10 min of zero-flow ischemia caused a 10-fold increase of total AMPK activity. This large activation of AMPK in the ex vivo heart contrasts with the relatively slow (requiring >10 min) and modest activation we observed in vivo. The main reason for this difference is undoubtedly that the degree of tissue ischemia and energetic disruption is less severe in our model of arterial hypoxemia than it is during zero-flow ischemia. Our intervention would be expected to decrease myocardial oxygen delivery to ∼75% of normal compared with the nearly complete absence of oxygen that occurs during zero-flow ischemia. Although the degree of hypoxic stress imposed by our protocol might therefore appear mild, it should be remembered that when applied to the 24-mo-old mice for 30 min, it caused death in 67% of the mice.
Although the hypoxemia-induced activation of AMPK we observed occurred in both the isoforms, increases in the α2-isoform were consistently greater than in the α1-isoform in both age groups. This isoform divergence has been observed in another physiological study and suggests a differential regulation of AMPK isoforms in vivo (9). In that study, the activation of cardiac AMPK observed during exercise is also consistent with the degree of hypoxic activation of AMPK seen in the present study. Those authors found that during high-intensity exercise (80% of maximal O2 consumption), total cardiac AMPK activity is increased twofold (9). Together, these data highlight the need to study AMPK activity in vivo to understand its physiological role.
Next, we set out to determine whether the increase in AMPK activation observed by age and hypoxia was caused by increased phosphorylation of AMPK. After dephosphorylation of the enzyme with phosphatase PP2Cα, the differences in activity observed with age or hypoxia were no longer present. Thus, consistent with most other scenarios in which acute activation of AMPK occurs, the increases in AMPK activity observed with age or hypoxemia are due to increased phosphorylation of AMPK.
Skeletal muscle AMPK activity.
The muscle weakness and frailty that occur with aging are associated with profound changes in muscle metabolism, most notably with the impairment of mitochondrial function that may compromise ATP production (11). Because mutations in AMPK genes that lead to changes in its activity can cause severe myopathy, one of our goals was to determine whether aging affects skeletal muscle AMPK activity (2, 3). Despite the reported changes in muscle metabolism associated with age, here we show that they do not include changes in basal AMPK activity. This lack of effect of age on skeletal muscle AMPK is somewhat surprising in light of the literature showing that aging causes metabolic changes and ATP depletion that might be expected to activate AMPK (11). However, our data support the possibility that while the aged skeletal muscle may be depleted in high-energy phosphates, the factor that most directly regulates AMPK activity, the AMP-to-ATP ratio, remains normal (27). It is also possible that different skeletal muscles present different degrees of metabolic deterioration with aging.
While skeletal muscle AMPK can be dramatically activated in vitro and ex vivo by hypoxia and ischemia, the extent to which AMPK is activated in muscle during acute hypoxemia in vivo has not previously been determined (16). Here we found that even 30 min of hypoxemia did not cause activation of AMPK in young mice. While this may seem surprising, previous studies have demonstrated that even more severe hypoxemia (PaO2 of <30 mmHg) does not cause ATP depletion or ADP accumulation in resting skeletal muscle in dogs (18). Therefore, our finding is not only consistent with these previous studies but confirms the lack of an energetic disruption in skeletal muscle during hypoxemia on a key downstream energy sensor, AMPK.
In summary, here, for the first time, we compared AMPK activity in cardiac and skeletal muscles from adult and senescent animals. Neither basal activity of AMPK nor allosteric activation by AMP was affected by aging. However, activation of AMPK during an acute hypoxic stress in vivo was higher in the heart of old animals. The effect of in vivo arterial hypoxemia on cardiac and skeletal muscle AMPK activity was remarkably modest compared with what occurs in vitro in the isolated heart or skeletal muscle. The present results support the more general view that age is associated with increased sensitivity to stress, at least in the heart. Future studies should focus on determining not only the cause of this increased sensitivity, but to the extent that it represents a maladaptation, strategies to reverse or prevent it.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society