Endocrinology and Metabolism

Activation of AMPK is essential for AICAR-induced glucose uptake by skeletal muscle but not adipocytes

Hideyuki Sakoda, Takehide Ogihara, Motonobu Anai, Midori Fujishiro, Hiraku Ono, Yukiko Onishi, Hideki Katagiri, Miho Abe, Yasushi Fukushima, Nobuhiro Shojima, Kouichi Inukai, Masatoshi Kikuchi, Yoshitomo Oka, Tomoichiro Asano


5-Aminoimidazole-4-carboxamide ribonucleoside (AICAR) reportedly activates AMP-activated protein kinase (AMPK) and stimulates glucose uptake by skeletal muscle cells. In this study, we investigated the role of AMPK in AICAR-induced glucose uptake by 3T3-L1 adipocytes and rat soleus muscle cells by overexpressing wild-type and dominant negative forms of the AMPKα2 subunit by use of adenovirus-mediated gene transfer. Overexpression of the dominant negative mutant had no effect on AICAR-induced glucose transport in adipocytes, although AMPK activation was almost completely abolished. This suggests that AICAR-induced glucose uptake by 3T3-L1 adipocytes is independent of AMPK activation. By contrast, overexpression of the dominant negative AMPKα2 mutant in muscle markedly suppressed both AICAR-induced glucose uptake and AMPK activation, although insulin-induced uptake was unaffected. Overexpression of the wild-type AMPKα2 subunit significantly increased AMPK activity in muscle but did not enhance glucose uptake. Thus, although AMPK activation may not, by itself, be sufficient to increase glucose transport, it appears essential for AICAR-induced glucose uptake in muscle.

  • AMP-activated protein kinase
  • 5-aminoimidazole-4-carboxamide ribonucleoside
  • exercise

in skeletal muscle, both insulin and exercise stimulate glucose transport by inducing translocation of GLUT4 to the cell surface, although the respective transduction pathways differ somewhat (2, 9, 18, 29). In that regard, AMP-activated protein kinase (AMPK) was recently implicated in exercise-induced, insulin-independent glucose transport. This enzyme is a well known serine/threonine kinase that phosphorylates acetyl-CoA carboxylase (ACC), thereby reducing its activity (5,8, 20, 22, 31). Downregulation of ACC, in turn, suppresses the activity of malonyl-CoA, an inhibitor of carnitine palmitoyltransferase I (CPT I), resulting in activation of CPT I and fatty acid oxidation. However, the role of AMPK in exercise-induced glucose transport is still unclear; indeed, the AMPK substrate mediating glucose transport is not yet known.

AMPK exists as a heterotrimer composed of a catalytic α-subunit, two isoforms of which occur in skeletal muscle (α1 and α2), and two regulatory subunits (β and γ) (30). In response to elevation of the AMP/ATP ratio, perhaps as a result of exercise or hypoxia, Thr172 of the α-subunit is phosphorylated by AMPK kinase (AMPKK) (10). Some evidence suggests that only the α2-subunit is phosphorylated (7, 17, 30), but substitution of Thr172 with Ala (T172A) abolishes the kinase activity of the AMPK heterotrimer containing either isoform (3, 27).

5-Aminoimidazole-4-carboxamide ribonucleoside (AICAR) is an adenosine analog taken up by muscle and phosphorylated to form 5-aminoimidazole-4-carboxamide-1-β-d-ribofuranosyl-5′-monophosphate (ZMP), which stimulates AMPK activity (14, 28) and glucose transport (1, 10, 11, 16, 33) in skeletal muscle and has therefore been used to study exercise-induced, insulin-independent glucose uptake. Although the effect of AICAR on glucose uptake is presumed to be mediated largely by changes in AMPK activity, there is no evidence that AICAR does not also stimulate activation of other molecules as well. Furthermore, it was recently reported that, whereas AICAR- and hypoxia-induced glucose uptake are completely abolished in the skeletal muscle of transgenic mice overexpressing kinase-dead AMPK, exercise-induced glucose uptake is only partially inhibited (17). Clearly, much remains to be learned about the molecular mechanism underlying exercise- and AICAR-induced glucose uptake. In the present study, we confirmed that AICAR stimulates glucose uptake into both rat soleus muscle and 3T3-L1 adipocytes. In addition, by use of adenovirus-mediated gene transfer, wild-type AMPK α-subunit and a T172A dominant negative mutant were overexpressed to further investigate the role of AMPK in basal and AICAR-induced glucose transport.



An antibody against the AMPKα subunit was prepared by immunizing rabbits with a GST-rat AMPKα2 subunit fusion protein. The antibodies raised against AMPKα were then affinity purified, as previously described (24).

Cell culture.

3T3-L1 fibroblasts were maintained in DMEM supplemented with 10% donor calf serum (Life Technologies) under a 10% CO2atmosphere at 37°C. Two days after the fibroblasts reached confluence, differentiation was induced by incubating the cells for 48 h in DMEM supplemented with 10% fetal bovine serum, 0.5 mmol/l IBMX, and 4 mg/ml dexamethasone. Thereafter they were maintained in DMEM supplemented with 10% fetal bovine serum for an additional 4–10 days; once >90% of cells expressed the adipocyte phenotype, the cells were used for experimentation.

Generation of recombinant adenovirus.

cDNAs encoding rat AMPKα1 and -α2 subunits were produced by PCR using a rat embryo cDNA library. A cDNA encoding the T172A AMPKα2 point mutant was produced by PCR with mutated primers, as reported previously (27). All AMPK constructs were designated to contain a c-myc tag at the NH2 terminus. Recombinant adenoviruses used to express wild-type and T172A α1- and α2-subunits were constructed by homologous recombination of the expression cosmid cassettes containing the corresponding cDNAs and the parental virus genome, as described previously (13). The amplified adenoviruses were purified and concentrated using cesium chloride ultracentrifugation. The resultant viruses were then dialyzed into phosphate-buffered saline containing 10% glycerol.

Gene transfer to 3T3-L1 adipocytes and rat skeletal muscle.

For gene transfer into 3T3-L1 adipocytes, cells were incubated for 6 h at 37°C in DMEM containing recombinant adenovirus, after which the virus-containing medium was replaced with normal growth medium. Experiments were performed 2 days later.

For transfer into skeletal muscle, 4-wk-old male Sprague-Dawley rats (Tokyo Experimental Animals, Tokyo, Japan) were anesthetized with pentobarbital sodium (60 mg/kg body wt ip), after which the fur was shaved from the lateral portion of both hindlimbs, a 5-mm incision was made in the skin, and 100 μl of the adenoviral vector (1.3 × 1010 pfu/ml) were injected intramuscularly. Each animal received the wild-type or T172A AMPKα2 gene in the right or left leg and the green fluorescent protein (GFP) gene in the opposite leg (6). Adenovirus encoding GFP served as a control. Five days after adenovirus injection, food was withdrawn for 12 h, the rats were anesthetized with pentobarbital sodium (60 mg/kg body wt ip), and the soleus muscles were dissected out for experimentation.

The study protocol was approved by the Institutional Review Board of the Institute for Adult Disease, Asahi Life Foundation. Animal care was in accordance with the policies of the University of Tokyo at all times.

AMPK assay.

After first serum-starving the cells for 3 h in serum-free DMEM, and then preincubating them for 1 h in Krebs-Ringer-HEPES buffer, 3T3-L1 adipocytes were stimulated by incubating them for 60 min in Krebs-Ringer- HEPES buffer containing 2 mmol/l AICAR. Alternatively, intact soleus muscles were incubated for 60 min in 2 ml of Krebs-Henseleit bicarbonate (KHB) buffer supplemented with 8 mmol/l glucose, 32 mmol/l mannitol, and 0.1% bovine serum albumin in 20-ml flasks in a shaking water bath at 35°C. Thereafter, the muscles were incubated for 60 min in KHB buffer, with or without 2 mmol/l AICAR, during which the flasks were gassed continuously with 95% O2-5% CO2. In both preparations, AICAR stimulation was stopped by freezing with liquid nitrogen. The 3T3-L1 adipocytes and muscles were then lysed in 10 vol/wt of buffer A [50 mmol/l Tris · HCl (pH 7.5), 50 mmol/l NaF, 5 mmol/l sodium pyrophosphate, 1 mmol/l EDTA, 1 mmol/l dithiothreitol (DTT), 0.1 mmol/l phenylmethylsulfonyl fluoride, and 10% glycerol] containing 1% Triton X-100; the insoluble material was removed by centrifugation, and the supernatants were collected. Aliquots of supernatant containing equal amounts of protein were incubated with anti-myc tag or anti-AMPKα subunit antibody. The resultant immune complexes were precipitated with protein A Sepharose (Pharmacia Biotech), after which they were washed twice with buffer A and then twice withbuffer B [50 mmol/l HEPES (pH 7.5), 1 mmol/l EDTA, 1 mmol/l DTT, and 10% glycerol]. The AMPK activity in the immunoprecipitates was assessed as a function of phosphorylation of the SAMS peptide (5, 27, 31). Assay reagents {40 mmol/l HEPES (pH 7.0), 200 μmol/l SAMS peptide, 200 μmol/l AMP, 80 mmol/l NaCl, 0.8 mmol/l EDTA, 0.8 mmol/l DTT, 8% glycerol, and 200 μmol/l [γ-32P]ATP} were added directly to the immunoprecipitate, and the mixture was incubated for 15 min at 30°C with shaking. Aliquots were then removed and spotted onto circle filters (P81 Whatman), which were then washed three times with 1% H3PO4 and once with acetone, air dried, and counted using a Molecular Imager (Bio-Rad).

Immunoprecipitation and immunoblotting.

The supernatants from adipocyte and muscle lysates, prepared as described above, were immunoprecipitated with anti-myc or anti-AMPKα subunit antibody. The immunoprecipitates were then boiled in Laemmli sample buffer containing 100 mmol/l DTT. SDS-PAGE and Western blotting were performed as described previously (23), with anti-AMPKα subunit antibody as a probe.

Assay of glucose transport.

Rat soleus muscles were isolated and incubated for 30 min in KHB buffer, with or without AICAR (2 mmol/l) or human insulin (2 mU/ml, Novolin R; Novo Nordisk). The muscles were then rinsed for 10 min at 29°C in 2 ml KHB buffer containing 40 mmol/l mannitol and 0.1% BSA and then incubated for 20 min at 29°C in 1.5 ml of KHB buffer containing 8 mmol/l 2-deoxy-d-[1,2-3H(N)]glucose (2-DG) (2.25 mCi/ml), 32 mmol/l [14C]mannitol (0.3 mCi/ml), 2 mmol/l sodium pyruvate, and 0.1% BSA. AICAR or insulin was present throughout the wash and the glucose uptake incubation. After the incubation, muscles were rapidly blotted, weighed, and solubilized in 1 ml of Soluene 350 (Packard). Radioactivity in the resultant samples was counted using a liquid scintillation counter. 2-DG uptake rates were corrected for extracellular trapping with mannitol counts (19).

Glucose transport in 3T3-L1 adipocytes was performed as described previously (23).


Characterization of anti-AMPKα antibody and overexpression of wild-type and T172A AMPKα2 subunit in 3T3-L1 adipocytes.

Immunoblot analysis of lysates from 3T3-L1 adipocytes overexpressing myc-tagged wild-type α1- or α2-subunit or the T172A α1- or α2-mutant, by use of anti-myc tag antibody as a probe, yielded bands at ∼64 kDa, which corresponds to the overexpressed α1- and α2-proteins; no band was observed in control cells overexpressing GFP, however (Fig.1, A and B,top). The antibody raised against the full-length α2-subunit also recognized the α1-subunit, although with only about one-half of the efficiency (Fig. 1 A, bottom). Thus the weak bands obtained from GFP-expressing cells were considered to be endogenously expressed α1- and α2-subunits (Fig. 1,A and B, bottom). The intensity of the bands corresponding to the overexpressed AMPKα1 or -α2 subunit were approximately fivefold stronger than those corresponding to the endogenous subunits (Fig. 1, A and B,bottom).

Fig. 1.

Overexpression of AMP-activated protein kinase (AMPK) α-subunit in 3T3-L1 adipocytes. A: lysates from 3T3-L1 adipocytes overexpressing green fluorescent protein (GFP), wild-type AMPKα1 (WT-α1), or wild-type AMPKα2 (WT-α2) were subjected to SDS-PAGE and Western blotting using anti-myc (top) or anti-AMPKα (bottom) antibody as a probe. B:top: lysates from cells overexpressing WT-α1 or -α2 or an AMPKα1 or -α2 T172A point mutant (T172Aα1 or -α2) were analyzed as in A, using anti-myc antibody as a probe. Bottom: cells were solubilized, immunoprecipitated (IP), and immunoblotted (IB) with anti-AMPKα subunit antibody.

The effect of AMPK overexpression on AICAR-induced AMPK activity in 3T3-L1 adipocytes.

Stimulating 3T3-L1 adipocytes overexpressing wild-type AMPKα1 or -α2 with 2 mmol/l AICAR increased the kinase activity present in the anti-myc antibody immunoprecipitates by ∼50 or 80% , respectively (Fig. 2, A andC). The basal kinase activity in immunoprecipitates from cells expressing the T172A α1- or α2-mutant was much lower; in fact, it was not different from control, and AICAR elicited no increase in kinase activity.

Fig. 2.

Effect of AMPK overexpression on 5-aminoimidazole-4-carboxamide ribonucleoside (AICAR)-induced AMPK activity in 3T3-L1 adipocytes. Cells were incubated with or without 2 mmol/l AICAR and then solubilized and immunoprecipitated with anti-myc (A, C) or anti-AMPKα (B, D) antibody. AMPK activity in the immunocomplexes was measured as a function of phosphorylation of SAMS peptide. Bars depict means ± SE of 3 independent experiments. A and C: * P < 0.05, ** P < 0.001 vs. WT-α1 or -α2 AICAR (−); #P < 0.0005 vs. WT-α1 or -α2 AICAR (+). Band D: * P < 0.005, ** P < 0.001 vs. GFP AICAR (−); #P< 0.001 vs. GFP AICAR (+).

Basal and AICAR-induced AMPK activity present in anti-AMPKα antibody immunoprecipitates from cells overexpressing wild-type AMPKα1 or -α2 were very similar to those from control cells overexpressing GFP (Fig. 2, B and D). By contrast, the basal and stimulated AMPK activities in cells overexpressing the T172A α1- or α2-mutant were both significantly lower than control: basal AMPK activity was reduced by ∼70%, and AICAR elicited no increase in that activity (Fig. 2, Band D).

Glucose transport in 3T3-L1 adipocytes overexpressing AMPK.

AICAR stimulation increased 2-DG uptake approximately twofold in 3T3-L1 adipocytes expressing GFP, confirming earlier observations (25). Overexpression of neither the wild-type nor the T172A form of AMPKα1 or -α2 had any effect on basal or AICAR-induced glucose uptake (Fig. 3), which was somewhat surprising because AMPK activity was clearly diminished in cells expressing a T172A mutant (Fig. 2, Band D).

Fig. 3.

AICAR-stimulated 2-deoxy-d-[1,2-3 H(N)]- glucose (2-DG) uptake by 3T3-L1 adipocytes. Cells overexpressing AMPKα1 (A) or -α2 (B) were incubated with or without 2 mmol/l AICAR for 1 h. 2-DG uptake was then assayed as described inmaterials and methods. Bars depict means ± SE of 3 independent experiments.

Overexpression of the wild-type α2-subunit or the T172A mutant in soleus muscle.

SDS-PAGE and immunoblotting were used to determine the level of adenovirus-mediated overexpression of wild-type AMPKα2 or the T172A α2-mutant in rat soleus muscle. AMPKα2 was detected as a band at a slightly greater molecular weight than the endogenous subunit (Fig.4 A). The levels of expression of the overexpressed wild-type and T172A forms were comparable to the level of endogenous AMPKα. Moreover, compared with the levels seen in control cells, overexpression of either wild-type α2- or the T172A α2-mutant resulted in a 50% reduction in the level of endogenous AMPKα subunit detected (Fig. 4 A). The glycogen levels in rats overexpressing GFP, wild-type, or T172A AMPKα2 were 75.8 ± 8.5, 70.2 ± 12.5, and 73.2 ± 13.1 μmol/g, respectively, and did not differ significantly.

Fig. 4.

Effect of overexpression of wild-type AMPKα2 (WT-α2) and the T172A point mutant (T172A-α2) on AMPK activity in rat soleus muscle. A: samples in left 3 lanes were immunoprecipitated with anti-myc antibody; those in theright 3 lanes were immunoprecipitated with anti-AMPKα subunit antibody. All were immunoblotted with anti-AMPKα subunit antibody. B: isolated muscle was incubated for 60 min in KHB buffer, with or without 2 mmol/l AICAR. AMPK activity in the immune complexes was measured as a function of phosphorylation of SAMS peptide. Bars depict means ± SE of 3 independent experiments; * P < 0.05 vs. GFP AICAR (−), ** P < 0.05 vs. GFP AICAR (+).

The effect of AMPK overexpression on AICAR-induced AMPK activity.

Overexpression of the wild-type α2-subunit in rat soleus muscle increased basal and AICAR-induced AMPK activity by 30 and 40%, respectively, compared with overexpression of GFP (Fig. 4 B). Stimulation with 2 mmol/l AICAR increased by ∼50% the level of activity present in AMPK immunoprecipitates from muscle overexpressing either GFP or wild-type AMPKα2. Overexpression of the T172A α2-mutant had no significant effect on basal AMPK activity but completely blocked the AICAR-induced AMPK activation. Thus the T172A α2-mutant functioned as a dominant negative form in both 3T3-L1 cells and rat soleus muscle.

Glucose uptake by rat soleus muscle overexpressing AMPK.

Uptake of 2-DG into isolated soleus muscle was increased twofold by AICAR stimulation (Fig.5 A). Overexpression of wild-type α2 had no significant effect on basal or AICAR-induced 2-DG uptake, whereas overexpression of the T172A α2-mutant slightly but significantly suppressed the basal 2-DG uptake and abolished the AICAR-induced increase in uptake. Insulin-stimulated 2-DG uptake, by contrast, was unaffected by overexpression of the T172A α2-mutant (Fig. 5 B).

Fig. 5.

Glucose uptake by soleus muscle overexpressing wild-type AMPKα2 subunit or a T172A point mutant. Isolated muscles were incubated for 60 min in KHB buffer, with or without 2 mmol/l AICAR (A) or 2 mU/ml human insulin (B). 2-DG uptake was assayed as described in materials and methods. Bars depict means ± SE of 3 independent experiments; * P < 0.05 vs. GFP AICAR (−), ** P < 0.01 vs. GFP AICAR (+).


Skeletal muscle takes up glucose in response to insulin stimulation and contraction (e.g., during exercise). The insulin-induced response is mediated via a pathway in which the insulin receptor (IR), insulin receptor substrates (IRS), and phosphatidylinositol (PI) 3-kinase activation are essential components (4, 12, 15). In contrast, exercise-induced glucose uptake by muscle is insulin independent and does not require PI 3-kinase activation (21). Instead, it appears that AMPK is involved in exercise-induced glucose uptake, because AICAR, an activator of AMPK, stimulates glucose uptake in a manner similar to muscle contraction.

AMPK, which exists as a heterotrimer composed of α-, β-, and γ-subunits, is activated by increases in the AMP/ATP ratio. The α-subunit, two isoforms of which (α1 and α2) have been identified, possesses the catalytic activity. Expression of the α1 subunit appears to be ubiquitous, whereas the α2-subunit appears to be expressed only in heart, liver, and skeletal muscle (26). Although skeletal muscle contains both isoforms, it may be that only the α2-subunit is phosphorylated, and thus activated, by AMPKK during exercise (7, 30). In this study, we constructed recombinant adenoviruses to express wild-type AMPKα1 or -α2 or a T172A point mutant, which, consistent with earlier findings, lacked kinase activity and was therefore considered to function as a dominant negative mutant.

AICAR increased AMPK activity twofold in 3T3-L1 adipocytes overexpressing GFP (control cells). Although expression of wild-type AMPKα1 or -α2 was increased fivefold in overexpressing cells, basal and AICAR-induced AMPK activities were similar to those in the control cells. This is very likely because the α-subunit is capable of catalytic activity only when complexed with the β- and γ-subunits (3, 27). That the availability of the regulatory subunits was a key determinant of AMPK activity means that one cannot determine whether changes in AMPK activity increase glucose uptake by 3T3-L1 adipocytes by overexpressing the α-subunit alone. On the other hand, whereas AICAR-induced AMPK activation was abolished by overexpression of the T172A α1- or α2-mutant, AICAR-induced 2-DG uptake was unaffected, making it very likely that AICAR stimulation increases glucose transport in 3T3-L1 adipocytes via a mechanism independent of AMPK activity.

Consistent with earlier findings (3, 17, 32), overexpression of AMPKα2 in skeletal muscle was associated with significantly diminished levels of the endogenous α-subunit. This likely reflects the fact that the α-subunit is unstable unless complexed with the β- and γ-subunits. Although the endogenous α-subunit was downregulated, overexpression of AMPKα2 significantly increased both basal and AICAR-stimulated AMPK activities. Still, despite the increased AMPK activity, 2-DG uptake was not enhanced in muscle overexpressing the wild-type subunit; apparently, AMPK activation alone is not sufficient to induce increases in glucose transport. Overexpression of a molecule upstream of AMPK, such as AMPKK, which has yet to be isolated, would enable one to address this issue more conclusively. In contrast, when AICAR-induced AMPK activation was abolished by overexpression of the T172A mutant, AICAR-induced 2-DG uptake was also markedly inhibited, which is consistent with the findings obtained using kinase-dead AMPKα2 transgenic mice (17). Thus, although AMPK activation may not, by itself, be sufficient for increased glucose transport, it does appear to be a necessary component of the pathway leading from AICAR stimulation to glucose uptake.

In summary, we found that AMPK activation is essential for AICAR-induced glucose transport in skeletal muscle, but not in 3T3-L1 adipocytes. In other words, AICAR appears to stimulate one or more molecules other than AMPK in adipocytes, which likewise leads to increased glucose transport. Our findings also show that, even in rat muscle, AICAR-induced glucose transport is dependent on actions taking place in addition to AMPK activation. Further investigation of the events occurring downstream of AMPK activation, and/or in other transduction pathways, will be required before a complete understanding of the mechanism of AICAR- and exercise-induced glucose transport is achieved.


  • Address for reprint requests and other correspondence: T. Asano, Dept. of Internal Medicine, Graduate School of Medicine, Univ. of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan (E-mail: asano-tky{at}umin.ac.jp).

  • 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.

  • First published February 19, 2002;10.1152/ajpendo.00455.2001


View Abstract