Activation of AMP-activated protein kinase (AMPK) by exercise and metformin is beneficial for the treatment of type 2 diabetes. We recently found that, in cultured cells, the LKB1 tumor suppressor protein kinase activates AMPK in response to the metformin analog phenformin and the AMP mimetic drug 5-aminoimidazole-4-carboxamide-1-β-d-ribofuranoside (AICAR). We have also reported that LKB1 activates 11 other AMPK-related kinases. The activity of LKB1 or the AMPK-related kinases has not previously been studied in a tissue with physiological relevance to diabetes. In this study, we have investigated whether contraction, phenformin, and AICAR influence LKB1 and AMPK-related kinase activity in rat skeletal muscle. Contraction in situ, induced via sciatic nerve stimulation, significantly increased AMPKα2 activity and phosphorylation in multiple muscle fiber types without affecting LKB1 activity. Treatment of isolated skeletal muscle with phenformin or AICAR stimulated the phosphorylation and activation of AMPKα1 and AMPKα2 without altering LKB1 activity. Contraction, phenformin, or AICAR did not significantly increase activities or expression of the AMPK-related kinases QSK, QIK, MARK2/3, and MARK4 in skeletal muscle. The results of this study suggest that muscle contraction, phenformin, or AICAR activates AMPK by a mechanism that does not involve direct activation of LKB1. They also suggest that the effects of excercise, phenformin, and AICAR on metabolic processes in muscle may be mediated through activation of AMPK rather than activation of LKB1 or the AMPK-related kinases.
- AMP-activated protein kinase
amp-activated protein kinase (AMPK) is a key regulator of cellular pathways that consume and generate cellular energy (4, 10). AMPK is activated by the elevation in cellular 5′-AMP that accompanies a fall in the ATP/ADP ratio due to the reaction catalyzed by adenylate kinase (9). One of the best studied physiological processes that activates AMPK is exercise in skeletal muscle, where AMPK stimulates uptake of glucose and the oxidation of fatty acids (27). AMPK is also activated by metformin, the most widely utilized drug for reducing blood glucose levels in type 2 diabetic patients (32). The mechanism by which metformin, or its closely related analog phenformin, activates AMPK is unknown but is not thought to involve changes in intracellular levels of AMP or the ATP/ADP ratio (7, 12).
The activation of AMPK by both energy depletion (i.e., during exercise) and phenformin requires phosphorylation of the catalytic subunit of AMPK at its T-loop residue (Thr172, in both the α1 and α2 catalytic subunit isoforms of AMPK) by an upstream kinase(s) (9). Recent work performed in Saccharomyces cerevisiae (14, 18, 25) and in mammalian cells (11, 23, 28) has demonstrated that a protein kinase termed LKB1 is the primary kinase that mediates the T-loop phosphorylation of AMPK. LKB1 is a 50-kDa serine/threonine kinase that is the product of the gene mutated in the autosomal dominantly inherited cancer disorder termed Peutz-Jeghers syndrome (3).
We have recently demonstrated that 11 other poorly studied protein kinases that belong to the AMPK subfamily (NUAK1, NUAK2, BRSK1, BRSK2, QSK, SIK, QIK, MARK1, MARK2, MARK3, MARK4), are activated in vitro following phosphorylation of their T-loop threonine residues by LKB1 (16). We have also developed assays to measure the activities of the endogenous AMPK-related kinases and found that the activities of these enzymes are markedly reduced in LKB1-deficient fibroblasts or HeLa cells (16). These results imply that LKB1 functions as a master “upstream” protein kinase, regulating the activities of AMPK and the AMPK-related kinases. LKB1 and AMPK-related kinase activities have thus far been studied only in cultured cell lines and not in tissues of relevance to diabetes. Because the beneficial effects of exercise and metformin for type 2 diabetes are thought to be at least partly mediated through activation of AMPK, we sought to determine whether LKB1 and AMPK-related kinases are also regulated by muscle contraction and/or phenformin in rat skeletal muscles.
RESEARCH DESIGN AND METHODS
Protease-inhibitor cocktail tablets were obtained from Roche (no. 1697498; Lewes, Sussex, UK), phenformin from Sigma (Poole, UK) and 5-aminoimidazole-4-carboxamide-1-β-d-ribofuranoside (AICAR) from Toronto Research Chemicals (North York, ON, Canada). [γ-32P]ATP, protein G-Sepharose, and enhanced chemiluminescence reagent were purchased from Amersham Biosciences (Little Chalfont, UK). All peptides were synthesized by Dr. Graham Bloomberg at the University of Bristol.
In situ muscle contraction.
Male Sprague-Dawley rats were obtained from Harlan (Bicester, Oxon, UK). The studies performed were approved by the University of Dundee Ethics Committee and performed under a UK Home Office project license. Rats weighing ∼150 g were anesthetized with pentobarbital sodium (50 mg/kg body wt ip), the sciatic nerves to both legs were surgically exposed, and electrodes were attached. One leg was subjected to electrical stimulation for 2.5, 5, 10, or 20 min (train rate, 1/s; train duration, 500 ms; pulse rate, 100 Hz; duration, 0.1 ms at 2–5 V), and the other leg served as sham-operated control, as described previously (19). Immediately after nerve stimulation, rats were killed by cervical dislocation, and hindlimb muscles were rapidly removed and then frozen in liquid nitrogen.
Rats weighing 50–70 g were killed by cervical dislocation, and extensor digitorum longus (EDL) muscles were rapidly removed. Isolated EDL muscles were incubated as previously described (13, 19). Briefly, muscles were incubated in Krebs-Ringer bicarbonate buffer (KRB), pH 7.4, containing 2 mmol/l pyruvate for 60 min in the absence or presence of 2 mmol/l AICAR or 10 mmol/l phenformin. The buffers were continuously gassed with 95% O2-5% CO2 and maintained at 37°C. At the end of the incubation period, muscles were quickly frozen in liquid nitrogen.
The specific AMPKα1 and AMPKα2 antibodies (11) and the phosphospecific antibodies recognizing AMPK phosphorylated on the T-loop were generated as described previously (24). The LKB1 antibody used for immunoblotting was raised in sheep against the mouse LKB1 protein (21), and that for immunoprecipitation was raised in sheep against the human LKB1 protein (2). The AMPK-related kinase antibodies employed in this study, including those recognizing QIK, QSK, and MARK4, were described previously (16). The antibody recognizing both MARK2 and MARK3 isoforms (16) was from Upstate Biotechnology (anti-c-TAK no. 05–680, Buckingham, UK).
Preparation of muscle lysates.
Muscles were pulverized and homogenized in ice-cold lysis buffer containing 50 mmol/l Tris·HCl, pH 7.5, 1 mmol/l EGTA, 1 mmol/l EDTA, 1% Triton X-100, 1 mmol/l sodium orthovanadate, 50 mmol/l sodium fluoride, 5 mmol/l sodium pyrophosphate, 0.27 mol/l sucrose, 0.1% 2-mercaptoethanol, and “complete” protease inhibitor cocktail (1 tablet per 50 ml). Homogenates were centrifuged at 13,000 g for 10 min at 4°C, supernatants were removed, and aliquots were snap frozen in liquid nitrogen.
LKB1−/− knockout and wild-type LKB1+/+ immortalized mouse embryonic fibroblasts were generated and cultured as described previously (11). Control HeLa cells that do not express LKB1 and HeLa cells expressing either wild-type or catalytically inactive LKB1[D194A] were prepared and cultured as described before (20). Cells cultured to near confluence on 10-cm diameter dishes in medium containing 10% (vol/vol) serum were lysed in 1 ml of lysis buffer described above.
Muscle or cell lysates (25–50 μg) or immunoprecipitated protein (from 0.5–1 mg of lysate) was heated in SDS sample buffer and subjected to SDS-PAGE and electrotransfer to nitrocellulose membranes. Membranes were then blocked in 50 mmol/l Tris·HCl, pH 7.5, 0.15 mol/l NaCl, and 0.1% Tween (TBST) containing 10% skimmed milk and probed for 16 h at 4°C in TBST, 5% skimmed milk, and 1 μg/ml of the indicated antibodies. Detection of proteins was performed using horseradish peroxidase-conjugated secondary antibodies and the enhanced chemiluminescence reagent.
Immunoprecipitation and assay of LKB1 employing LKBtide substrate.
Muscle lysate protein (0.3–0.5 mg) or 1.0 mg of mouse embryonic fibroblasts or HeLa cell lysate protein was incubated at 4°C for 1 h on a shaking platform with 5 μl of protein G-Sepharose covalently conjugated to 5 μg of human LKB1 antibody (2). The immunoprecipitates were washed twice with 1 ml of lysis buffer containing 0.5 mol/l NaCl and twice with 1 ml of buffer A (50 mmol/l Tris·HCl, pH 7.5, 0.1 mmol/l EGTA, and 0.1% 2-mercaptoethanol). Phosphotransferase activity toward the LKBtide peptide [SNLYHQGKFLQTFCGSPLYRRR residues 241–260 of human NUAK2 with 3 additional Arg residues added to the COOH terminus to enable binding to p81 paper (16)] was then measured in a total assay volume of 50 μl consisting of 50 mmol/l Tris·HCl, pH 7.5, 0.1 mmol/l EGTA, 0.1% 2-mercaptoethanol, 10 mmol/l magnesium acetate, 0.1 mmol/l [γ-32P]ATP (∼200 cpm/pmol), and 200 μmol/l LKB1tide peptide. The assays were carried out at 30°C with continuous shaking to keep the immunoprecipitates in suspension and were terminated after 20 min by applying 40 μl of the reaction mixture onto p81 membranes. These were washed in phosphoric acid, and the incorporated radioactivity was measured by scintillation counting. One unit of activity was defined as that which catalyzed the incorporation of 1 nmol of 32P into the substrate per minute.
Immunoprecipitation and assay of AMPK and AMPK-related kinase.
Muscle lysate protein (0.05–1 mg) was incubated at 4°C for 1 h on a shaking platform with 5 μg of the corresponding antibody, which had been previously conjugated to 5 μl of protein G-Sepharose. The immunoprecipitates were washed twice with lysis buffer containing 0.5 mol/l NaCl and twice with buffer A. Phosphotransferase activity toward the AMARA peptide was then measured in a total assay volume of 50 μl as described above.
Data are expressed as means ± SE. Statistical analysis was undertaken using a paired Student's t-test and one-way analysis of variance. When analysis of variance revealed significant differences, further analysis was performed using Tukey's post hoc test for multiple comparisons. Differences between groups were considered statistically significant when P < 0.05.
LKB1 is not activated in response to muscle contraction in rat gastrocnemius muscle.
Hindlimb muscle contractions (in situ contraction) were induced via electrical stimulation of the sciatic nerve in anesthetized rats. We initially measured AMPKα1 (Fig. 1A, left), AMPKα2 (Fig. 1A, right), and phosphorylation of AMPKα1/α2 at Thr172, the T-loop residue (Fig. 1B), in response to contraction in rat gastrocnemius muscle (mixture of red and white gastrocnemius muscles). Contraction increased AMPKα2 activity at all time points measured, while AMPKα1 activity remained unchanged. Previous studies have also found that the AMPKα1 is not activated in contracting muscle under conditions that AMPKα2 is stimulated (8, 31). Total AMPKα1/α2 Thr172 phosphorylation, assessed using a phosphospecific antibody, was elevated in response to muscle contraction, confirming that contraction induced the phosphorylation of AMPK at Thr172 by upstream kinase(s).
LKB1 catalytic activity was measured in rat gastrocnemius muscle after its immunoprecipitation, employing the LKBtide peptide described in research design and methods. Significant basal LKB1 activity was detected in muscle, but its activity was not altered by contraction at any time point (Fig. 1C). To demonstrate that the LKB1 activity assay employed was specific, LKB1 was immunoprecipitated from either LKB1−/− knockout or LKB1+/+ mouse embryonic fibroblasts (11) or from HeLa cells that do not express LKB1 as well as HeLa cell lines stably expressing wild-type or catalytically inactive LKB1 (20) (Fig. 1D). These control studies demonstrate that the LKB1 activity measured was vastly higher in the cell lines expressing LKB1 compared with the LKB1-deficient cells.
LKB1 is not activated in response to muscle contraction in different muscle fiber types. Skeletal muscles are composed of a heterogeneous population of fiber types that vary according to their contractile and metabolic properties. We therefore measured LKB1 activity in different muscle fiber types in response to 5 min of in situ contraction where we saw the highest AMPKα2 activity over the time course examined in mixed (red and white) gastrocnemius muscle (Fig. 1A). For this purpose, sciatic nerve stimulation was performed to induce muscle contraction in multiple hindlimb muscles that encompass a varying range in fiber type composition (19).
Figure 2A demonstrates the effect that contraction has on AMPKα2 activity and T-loop phosphorylation in several rat hindlimb muscles. Interestingly, the level of AMPKα2 activation varied in the different fibers. AMPKα2 activity did not change in soleus (predominantly slow-twitch muscle). In contrast, AMPKα2 activity was increased less than twofold in red gastrocnemius (a mixture of slow- and fast-twitch fibers), approximately sixfold in white gastrocnemius (predominantly fast-twitch muscle), and approximately tenfold in EDL (predominantly fast-twitch muscle). AMPKα2 protein expression was lower in soleus, consistent with its lower basal specific activity. Figure 2B illustrates the effect of contraction on LKB1 activity and protein expression in the same muscles used for the AMPKα2 assay. Protein expression and basal activity of LKB1 were similar across all muscle types. Muscle contraction did not significantly change LKB1 activity or protein expression.
Expression and activity of AMPK-related kinases in rat skeletal muscle.
We recently generated antibodies that can be used to immunoblot and immunoprecipitate the members of the AMPK-related kinases in cultured cells (16). In skeletal muscle, we were able to detect expression and activity of QSK, QIK, and MARK4 and the combined activity of MARK2/MARK3 measured using an antibody that specifically recognizes both of these MARK isoforms. For the other seven AMPK-related kinases (MARK1, NUAK1, NUAK2, SIK, MELK, BRSK1, BRSK2), although the antibody recognizes both endogenous and recombinant proteins expressed in various cell lines (16) and/or brain and testis extracts from rats (Sakamoto K, unpublished data), the expression and activity of these kinases were not detectable following the immunoprecipitation of 1 mg of rat skeletal muscle lysate. Figure 3A shows the expression and activities of QSK, QIK, MARK2/3, and MARK4 in noncontracted/basal and contracted white gastrocnemius muscle. Compared with the 600% stimulation of AMPKα2 under these conditions (Fig. 2A), muscle contraction did not cause significant increases in the AMPK-related kinases, although there was a tendency for a slight (10–25%) increase in QSK and QIK activity in contracted muscles compared with sham-operated muscles. Because of this, we also assayed QSK and QIK in EDL muscle where contraction stimulates AMPKα2 activity 10-fold (Fig. 2A). We found that there was also no significant elevation of QSK or QIK under these conditions (Fig. 3B).
AICAR and phenformin do not activate LKB1 or AMPK-related kinases in isolated rat EDL muscle in vitro.
To further study the effect of nucleoside monophosphates and phenformin on LKB1 activity in skeletal muscle, we incubated isolated rat EDL muscle in the presence or absence of 10 mmol/l phenformin or 2 mmol/l AICAR, the latter being converted by adenosine kinase to AICAR monophosphate (ZMP), a cellular mimetic of AMP (6). Phenformin rather than metformin was employed, as this analog of metformin activates AMPK much more rapidly than metformin (Hardie DG, unpublished results). AMPKα1 and AMPKα2 activity and T-loop phosphorylation were stimulated significantly by AICAR or phenformin (Fig. 4A). AICAR increased both AMPKα1 and AMPKα2, with a smaller effect on AMPKα1, whereas phenformin activated both AMPK isoforms approximately fourfold. In contrast, the activities of LKB1 (Fig. 4B) or QSK, QIK, MARK2/3, or MARK4 (Fig. 4C) were not significantly affected by AICAR or phenformin.
Recent work has provided strong evidence that LKB1 is the primary protein kinase acting upstream of AMPK. Biochemical studies demonstrated that LKB1 complexed to two regulatory subunits (STRADα/β and MO25α/β) can efficiently phosphorylate AMPK at its T-loop threonine residue, resulting in its activation. Moreover, in LKB1-deficient cells [mouse embryonic fibroblasts and HeLa cells (11)] or cells treated with heat shock protein 90 inhibitors that destabilize the LKB1 protein (28), AMPK is not activated by either phenformin or AICAR. The major finding of this study was that, in skeletal muscle, three treatments (contraction, phenformin, and AICAR) that induce activation of AMPK failed to activate LKB1 or a group of AMPK-related kinases that are downstream of LKB1 in cultured cells (16). These results strongly indicate that contraction, phenformin, and AICAR are not activating AMPK by directly stimulating LKB1 activity. Two previous studies havereported that an AMPK-activating activity measured in whole cell extracts was modestly stimulated (1.5-fold) by exercise in human skeletal muscle (5) or metformin in H4IIE cells (12). Our findings indicate that these effects are unlikely to result from a direct activation of LKB1. It is possible that a factor induced by exercise or metformin, which is present in total cell extracts, enhanced the rate at which LKB1 activated AMPK in these studies. Our results do not rule out the possibility that there is another activator of AMPK in skeletal muscle, distinct from LKB1, that phosphorylates AMPK on Thr172 in response to exercise, phenformin, or AICAR. To resolve this question, it will be necessary to analyze AMPK activation in skeletal muscle of an LKB1 conditional knockout mouse, because the complete LKB1 knockout is an embryonic lethal (30).
AMPK is an αβγ heterotrimer, and recent work shows that the γ-subunit contains two nucleotide-binding sites that are responsible for allosteric regulation by AMP and ATP (22). The binding of AMP to AMPK not only stimulates activity but also enhances two- to threefold the rate at which AMPK can be phosphorylated and activated by the LKB-STRAD-MO25 complex in vitro (11). Our finding that LKB1 is not activated by contraction or AICAR is consistent with the hypothesis that AMP or ZMP bind to AMPK, making it a better substrate for LKB1. This is reminiscent of the mechanism of activation of the PKB/Akt protein kinase by the upstream kinase phosphoinositide-dependent protein kinase-1 (PDK-1) (1). Insulin does not directly activate PDK-1, but instead leads to the generation of the second messenger phosphatidylinositol 3,4,5-triphosphate, which interacts with PKB/Akt and converts it into a better substrate for PDK-1 (1).
Phenformin stimulates AMPK without affecting the AMP or the ADP/ATP ratio in cells (7, 12) and does not influence the phosphorylation of the heterotrimeric AMPK complex by LKB1 in vitro (11). It is therefore currently unclear how phenformin activates AMPK. Possibilities include phenformin inhibiting an AMPK-specific phosphatase that does not target the AMPK-related kinases, although the related drug metformin did not affect dephosphorylation of AMPK by protein phosphatase 2C in vitro (12). Alternatively, phenformin may somehow generate a metabolite or other molecule that, inside the cell, binds to AMPK and promotes its activation by the LKB1 complex. Our findings in skeletal muscle are consistent with work in COS-7 (28) and fibroblast (16) cells, where LKB1 was not activated by phenformin or AICAR. It should be noted that we have employed the metformin analog phenformin in our studies because it activates AMPK more rapidly than metformin. We would have expected similar results had we employed metformin, as metformin is structurally closely related to phenformin. Further work is required, however, to formally establish whether metformin and phenformin activate AMPK by the same mechanism.
Although considerable evidence supports the conclusion that AMPK at least partly mediates glucose uptake in muscle following exercise and contraction (reviewed in Refs. 4 and 10), recent studies indicate that AMPK-independent pathways may also contribute to stimulation of glucose uptake (15, 17, 29). The AMPK-related kinases are currently poorly studied and are therefore possible candidates for mediating the AMPK-independent effects of glucose uptake following muscle contraction. Using previously characterized antibodies (16) that specifically recognize all 11 AMPK-related kinases that are regulated by LKB1, we were able to detect significant expression and activity of QSK, QIK, MARK2/3, and MARK4 in skeletal muscle. We failed to detect significant increases in the activities of these enzymes in response to contraction or to phenformin or AICAR. This is consistent with the finding that the AMP-related kinases measured in mouse embryo fibroblasts or HeLa cells expressing LKB1 were not stimulated by phenformin or AICAR (16). Taken together, these findings indicate that the effects of exercise, phenformin, and AICAR on metabolic processes in the muscle are mediated through activation of AMPK rather than AMPK-related kinases. Further studies are required to define the roles of the AMPK-related kinases in skeletal muscle and other tissues.
We thank the Association for International Cancer Research, Diabetes UK, the Medical Research Council and the pharmaceutical companies supporting the Division of Signal Transduction Therapy Unit at Dundee (AstraZeneca, Boehringer-Ingelheim, GlaxoSmithKline, Merck, Merck KgaA, and Pfizer) for financial support. O. Göransson is supported by a Wenner-Gren Foundation fellowship and the Foundation for Diabetes Research, Uppsala. D. G. Hardie is supported by a Programme Grant from the Wellcome Trust.
We thank Greg Stewart for affinity purification of AMPK antibodies; Alfonso Mora, Jose Lizcano, and Jérôme Boudeau for advice; Michael Hirshman and Laurie Goodyear for guidance in setting up the isolated muscle incubation system. We also thank Tomi Makela and Lina Udd (Helsinki University) for providing us with LKB1−/− knockout cells.
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 by American Physiological Society