LKB1, an upstream AMPK kinase, regulates glucose and lipid metabolism in cultured liver and muscle cells

BBRC Biochemical and Biophysical Research Communications 351 (2006) 595–601 www.elsevier.com/locate/ybbrc

LKB1, an upstream AMPK kinase, regulates glucose and lipid metabolism in cultured liver and muscle cells Kenta Imai, Kouichi Inukai *, Yuichi Ikegami, Takuya Awata, Shigehiro Katayama Division of Endocrinology and Diabetes, Department of Medicine, Saitama Medical School, Morohongo 38, Moroyama, Iruma-gun, Saitama 350-0495, Japan Received 27 September 2006 Available online 18 October 2006

Abstract LKB1 is a 50 kDa serine/threonine kinase that phosphorylates and activates the catalytic subunit of AMPK at its T-loop residue Thr 172. We prepared adenoviruses expressing the constitutive active (wild-type) form (CA) or dominant negative (kinase inactive, D194A mutant) form (DN) of LKB1 and overexpressed these proteins in cultured myotubes (C2C12 cells) and rat hepatoma cells (FAO cells). When analyzed by immunoblotting with the antibody against Thr172-phosphorylated AMPK, the phosphorylation of AMPK was increased (2.5-fold) and decreased (0.4-fold) in cells expressing CA and DN LKB1, respectively, as compared with Lac-Z expressing control cells. Immunoprecipitation experiments, using isoform-specific antibody, revealed these alterations of AMPK phosphorylation to be attributable to altered phosphorylation of AMPK a2, but not a1 catalytic subunits, strongly suggesting the a2 catalytic subunit to be the major substrate for LKB1 in mammalian cells. In addition, adiponectin or AICAR-stimulated AMPK phosphorylation was inhibited by overexpression of DN LKB1, while phenformin-stimulated phosphorylation was unaffected. These results may explain the difference in AMPK activation mechanisms between AMP and phenformin, and also indicate that AMPK phosphorylation by LKB1 is involved in AMP-stimulated AMPK activation. As a downstream target for AMPK, AICAR-induced glucose uptake and ACCb phosphorylation were found to be significantly reduced in DN LKB1 expressing C2C12 cells. The expression of key enzymes for gluconeogenesis, glucose6-phosphatase and phosphoenolpyruvate carboxykinase, was also dependent on LKB1 activities in FAO cells. These results demonstrate that LKB1 is a crucial regulator of AMPK activation in muscle and liver cells and, therefore, that LKB1 activity is potentially of importance to our understanding of glucose and lipid metabolism. Ó 2006 Elsevier Inc. All rights reserved. Keywords: LKB1; AMPK; C2C12 myotubes; FAO cells; Gluconeogenesis

AMP-activated protein kinase (AMPK) has been implicated as a key regulator of physiological energy dynamics, including glucose transport, gluconeogenesis, and lipolysis [1–3]. One of the most studied events activating AMPK is muscle contraction. In this process, AMPK is activated following a rise in the intracellular AMP:ATP ratio. AMPK is a heterotrimeric complex comprised of a catalytic a subunit and regulatory b and c subunits. Once AMPK is activated by phosphorylation of Thr 172 in the T-loop of the a subunit, AMPK regulates a number of transcriptional factors and signal transduction proteins, such as acetyl CoA car-

*

Corresponding author. Fax: +81 492 76 1430. E-mail address: [email protected] (K. Inukai).

0006-291X/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2006.10.056

boxylase (ACC) [4], hepatic nuclear factor a (HNFa) [5], and p38 mitogen-activated protein kinase (MAPK) [6], and thereby restores intracellular ATP levels. Though many studies have focused on its physiological response to energy demands, AMPK can also be activated by a variety of signals, such as ischemia [7], hypoxia [8], and oxidative stress [9]. Adiponectin, an adipocyte-derived insulinsensitizing hormone, also reportedly activates AMPK [10]. Among chemical agents, 5-aminoimidazole-4-carboxamide riboside (AICAR), which is converted to an AMP mimetic within the cell, is an AMPK activator [11]. Metformin, which is widely used to lower blood glucose levels in Type 2 diabetic subjects, also activates AMPK [12], although the mechanism by which metformin activates AMPK remains unclear.

596

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

LKB1 is a 50 kDa serine/threonine kinase that was originally identified as the product of the gene mutated in the autosomal dominantly inherited Peutz–Jeghers syndrome [13]. Recently, LKB1 was demonstrated to phosphorylate and activate the catalytic a-subunit of AMPK at its T-loop residue Thr 172 in a cell-free system [14,15]. Thus, AMPK is regulated by upstream kinases of which the tumor suppressor, LKB1, was the first to be identified. AMPK is minimally activated in cells that lack or have decreased LKB1 expression [15,16]. LKB1 forms a heterotrimeric complex, with regulatory proteins termed STRAD and MO25, which are required for its activation and cytosolic targeting [17,18]. Thus, whether LKB1 actually functions as a central regulator of organismal metabolism remains uncertain, but it has become apparent in the last few years that LKB1 plays a crucial role in activating AMPK. In the present study, we prepared adenoviruses expressing a dominant negative (kinase inactive, D194A mutant) form of LKB1 and overexpressed this protein in cultured myotubes (C2C12 cells) and rat hepatoma cells (FAO cells). In these cells, AMPK phosphorylation was markedly decreased (0.4-fold), as compared with that in Lac-Z expressing control cells. Under these conditions, we investigated molecules downstream from the AMPK cascade and obtained convincing evidence that LKB1 influences a variety of AMPK-related functions, indicating LKB1 to play an important role in glucose and lipid metabolism in muscle and liver. Materials and methods Cell culture and chemicals. C2C12 or L6 myoblasts were maintained in DMEM with 4500 mg/L glucose or 1000 mg/L glucose, respectively, containing 10% fetal calf serum (Life Technologies, Inc.) at 37 °C in 5% CO2. After the myoblasts had reached subconfluence, differentiation was induced by treatment with DMEM containing 5% horse serum for 7 days, at which time formation of myotubes was maximal. FAO cells, derived from a rat hepatoma, were maintained in DMEM containing 10% fetal calf serum at 37 °C in 5% CO2 and used for experiments at 90% confluence. The 3T3-L1 fibroblasts were maintained in DMEM containing 10% donor calf serum at 37 °C in 10% CO2. Two days after the fibroblasts had reached confluence, differentiation was induced by treating the cells with DMEM containing 0.5 mM 3-isobutyl-1-methylxanthine, 4 lg/ml dexamethasone, and 10% fetal bovine serum for 48 h. Cells were refed with DMEM supplemented with 10% fetal bovine serum for the following 6 days. When >90% of the cells expressed the adipocyte phenotype, the cells were used for the experiments. After the media had been exchanged for KRB buffer (25 mM Hepes, 118 mM NaCl, 4.8 mM KCl, 1.3 mM CaCl2, 1.2 mM KH2PO4, 1.3 mM MgSO4, 5 mM NaHCO3, 5.5 mM glucose, and 0.07% BSA), the indicated concentrations of chemicals were added before cell lysate preparation. AICAR (Toronto Research Chemicals Inc., Canada), phenformin (Sigma, MO), and human recombinant adiponectin (R&D Systems Inc., MN) were purchased. Antibodies and Western blotting. Western blotting was performed as previously described [19]. Briefly, after incubation with the indicated chemicals, cell lysates from C2C12 myotubes or FAO hepatoma cells were washed with ice-cold PBS, lysed in ice-cold lysis buffer (50 mM Tris, 150 mM NaCl, 1 mM Na3VO4, 1 mM NaF, 1% NP40, 1 mM EDTA, 1 mM EGTA, and 2.5 mM sodium pyrophosphate), and then centrifuged at 14,000g for 10 min at 4 °C. Supernatants including tissue protein extracts were resolved on 10% SDS–polyacrylamide gel, followed by electrophoretic transfer to a nitrocellulose membrane. Membranes were

incubated for 1 h at RT with the appropriate primary antibody. Commercial antibodies against phospho-AMPK (Thr 172) (Cell Signaling Technology, CA), the a1 and a2 subunits of AMPK (Santa Cruz Biotechnology, Inc., CA), phospho-Akt (Ser 473) (Cell Signaling Technology, CA), mouse LKB1 (Santa Cruz Biotechnology, Inc., CA), and acyl-CoA carboxylase b (Upstate Cell Signaling, NY) were purchased. After blotting with the indicated secondary antibody, detection was performed using an ECL chemiluminescent kit (Amersham Pharmacia Biotech, UK) according to the manufacturer’s instructions. Quantitations were performed using a Molecular Imager (Bio-Rad Lab, CA). Immunoprecipitation was performed as previously described [20], using the anti-a1 and a2 subunits of AMPK antibody. Immunoprecipitates were then boiled in Laemmli sample buffer containing 100 mM DTT, subjected to SDS–PAGE, and finally to Western blotting using the anti-phospho AMPK antibody. RNA preparaion and real time PCR. Total RNA from FAO cells was isolated with Isogen (Nippon Gene, Japan). cDNA was synthesized from the purified RNA using a reverse transcriptase kit (Amersham Pharmacia Biotech, UK) according to the manufacturer’s instructions. For quantitative analysis of glucose 6 phosphatase (G6Pase) and phosphoenolpyruvate carboxykinase (PEPCK), we conducted real-time PCR using an ABI PRISM Model 7000 (Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. The primer sets and probes for rat G6Pase (Assay ID; Rn00565347_m1) and rat PEPCK (Rn01529009_g1) were purchased. Wild-type and dominant negative forms of the LKB1 cDNA construct. A full length mouse LKB1 cDNA was isolated from mouse hepatic RNA by reverse transcriptase PCR. PCR oligonucleotide sequences used for PCR were as follows: coding strand, 5 0 -AGGATGGACGTGGCGGACCCC GAGCCG-3 0 ; non-coding strand, 5 0 -CAGTCACTGCTTGCAGG CCGAGAGCCG-3 0 . The cDNA was subcloned into TA vectors, pCRII (Invitrogen Life Technologies), sequenced to confirm their identities and observed to have no unexpected mutations. The single amino acid substitution of Alanine for Asparagine-194 of LKB1 was performed as previously described [21]. Briefly, the A to C point mutation was introduced at nucleotide 581 of the wild-type LKB1 cDNA by PCR, using the synthesized oligonucleotide 5 0 -CTCAAGATCTCGGCCCTCGGTGTTGCC GAGGCC-3 0 , in which A at nucleotide 581 had been replaced by C. This point mutation resulted in substitution of alanine (GCC) for asparagine (GAC) at residue 194 of the encoded LKB1. Generation of recombinant adenoviruses and transfection. Recombinant adenovirus containing full-length murine LKB1 cDNA was prepared by homologous recombination of expression cosmid cassettes containing the corresponding cDNAs and the parental adenovirus genome, as described previously [22]. Lac Z adenovirus was used as a control. The amplified adenoviruses were purified and concentrated using cesium chloride ultracentrifugation. The resultant viruses were then dialyzed into phosphatebuffered saline. For adenovirus mediated transfection, cultured cells were incubated with DMEM containing the adenoviruses for 2 h at 37 °C, and the growth medium was then added. Experiments were performed 3 days after transfection. Assay of glucose uptake. C2C12 myotubes were incubated for 30 min in KRB buffer, with or without 2 mM AICAR. Then, the cells were incubated in glucose-free Krebs–Ringer phosphate (KRP) medium containing 2-deoxy-D-[1,2-3H(N)]glucose (2-DG) (2.25 mCi/ml). 2-DG uptake was measured during the last 20 min of incubation. AICAR was present throughout the wash and the glucose uptake incubation. Glucose transport in L6 myotubes or 3T3-L1 adipocytes with or without 107 M human insulin (Novo Nordisk) was assayed as described previously [23]. After the cells had been lysed with 0.1% Triton, radioactivity was measured using liquid scintillation counting.

Results Wild-type LKB1 is reported to behave as a constitutive active form in terms of its kinase activity [24]. We prepared adenoviruses expressing the wild-type or the dominant negative (kinase inactive, D194A mutant) form of LKB1,

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

occur in the absence of LKB1 activity, suggesting that LKB1 plays a crucial role in AICAR-mediated AMPK activation. To further assess which ligands are affected by LKB1, we stimulated C2C12 myotubes with phenformin, adiponectin, and AMP in the absence or presence of LKB1 activation (Fig. 2). All of these ligands increased AMPK phosphorylation (2.8-, 1.7-, and 2.1-fold, respectively) in control C2C12 cells (Fig. 2, lanes 1 and 2). Similar enhancements in AMPK phosphorylations following treatment with each ligand were observed in CA expressing cells (Fig. 2, lanes 3 and 4). On the other hand, when DN expressing cells were treated with adiponectin or AMP, no increase in AMPK phosphorylation was observed, while only phenformin treatment markedly increased AMPK activation (Fig. 2, lanes 5 and 6). These results may explain the different mechanisms of AMPK activation and also indicate that AMPK phosphorylation by LKB1 is involved in AMP-stimulated AMPK activation. To investigate which isoform of AMPK LKB1 is preferentially phosphorylated in vivo, we performed immunoprecipitation experiments using isoform-specific antibody. When immunoprecipitates of the a2 isoform antibody were blotted with antiphospho-Thr172 of AMPK, there was a marked increase in phosphorylation in CA expressing C2C12 cells, while phosphorylation was unchanged in those with the a1 isoform (Fig. 3, lanes 1 and 3), revealing AMPK phosphorylation alteration in CA expressing cells to be attributable to altered phosphorylation of the AMPK

designated CA and DN, respectively, and overexpressed these proteins in cultured myotubes (C2C12 cells) or rat hepatoma cells (FAO cells). After adenovirus gene transfer, a large amount of LKB1 or mutated protein was observed by Western blotting using anti-LKB1 antibody (Fig. 1, upper panel). These cultured cells expressed only a small amount of endogenous LKB1, as compared with recombinant LKB1 protein expressed via adenovirus gene transfection. The amounts of a1 or a2 AMPK subunits did not change significantly with LKB1 induction. The phosphorylation of Akt-Ser473, which is involved in the insulin signaling cascade, was not significantly altered. When analyzed by immunoblotting with the antibody against Thr172-phosphorylated AMPK, the phosphorylation of AMPK in C2C12 myotubes was increased (2.5-fold) and significantly decreased (0.4-fold) in cells expressing CA and DN, respectively, as compared with Lac-Z expressing control cells (Fig. 1A, lanes 1, 3 and 5). The AMPK phosphorylations in FAO cells tended to be similar to those in C2C12 myotubes (Fig. 1B). DN overexpression did not lead to complete loss of basal activity, which explains the presence of upstream kinases other than LKB1. Interestingly, the AICAR treatment did not significantly increase AMPK-phosphorylation in DN overexpressing cells, while significant enhancement of AMPK phophorylation was observed in control and CA overexpressing cells treated with AICAR (Fig. 1, lanes 2, 4 and 6). Thus, we speculate that AICAR-induced AMPK phosphorylation does not

A

597

B

Fig. 1. Effect of CA or DN LKB1 overexpression on AICAR-induced AMPK phosphorylation in C2C12 myotubes (A) and FAO cells (B). Cells were transfected with adenoviruses expressing Lac Z (lanes 1 and 2), CA LKB1 (lanes 3 and 4), and DN LKB1 (lanes 5 and 6). After the cells had been incubated in KRB buffer with or without 2 mM AICAR for 30 min, they were lysed in ice-cold lysis buffer and centrifuged at 14,000g for 10 min at 4 °C. Supernatants including tissue protein extracts were resolved on 10% SDS–polyacrylamide gel, followed by electrophoretic transfer to a nitrocellulose membrane. Membranes were incubated for 1 h at RT with the antibody against phospho-AMPK (Thr 172), mouse LKB1, the a1 and a2 subunits of AMPK and phospho-Akt (Ser 473). After blotting with the indicated secondary antibody, detection was performed using an ECL chemiluminescent kit. Representative data from four experiments are presented. Each column shows the means ± SE obtained from these experiments. *Significant difference (P < 0.05) relative to AMPK phosphorylation of control cells in the absence of AICAR, **significant difference (P < 0.05) relative to AMPK phosphorylation of CA expressing cells in the absence of AICAR, as determined by Student’s t test. N.S. means; not significant.

598

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

Fig. 2. Effect of CA or DN LKB1 overexpression on phenformin (A), adiponectin (B), and AMP (C)-induced AMPK phosphorylation in C2C12 myotubes. C2C12 myotubes were transfected with adenoviruses expressing Lac Z (lanes 1 and 2), CA LKB1 (lanes 3 and 4), and DN LKB1 (lanes 5 and 6). After the cells had been incubated in KRB buffer with or without 2 mM phenformin, 30 lg/ml adiponectin, and 0.2 mM AMP for 30 min, they were lysed in ice-cold lysis buffer and centrifuged at 14,000g for 10 min at 4 °C. Supernatants including tissue protein extracts were resolved on 10% SDS– polyacrylamide gel, followed by electrophoretic transfer to a nitrocellulose membrane. Membranes were incubated for 1 h at RT with the antibody against phospho-AMPK (Thr 172). After blotting with the indicated secondary antibody, detection was performed using an ECL chemiluminescent kit. Representative data from four experiments are presented. Each column shows the means ± SE obtained from these experiments. *Significant difference (P < 0.05) relative to AMPK phosphorylation of control cells in the absence of each ligand, **significant difference (P < 0.05) relative to AMPK phosphorylation of DN expressing cells in the absence of each ligand, as determined by Student’s t test. N.S. means; not significant.

Fig. 3. Effect of CA or DN LKB1 overexpression on AICAR-induced phosphorylation of a1 (A) or a2 (B) AMPK isoform. C2C12 myotubes were transfected with adenoviruses expressing Lac Z (lanes 1 and 2), CA LKB1 (lanes 3 and 4,) and DN LKB1 (lanes 5 and 6). After the cells had been incubated in KRB buffer with or without 2 mM AICAR for 30 min, they were lysed in ice-cold lysis buffer and centrifuged at 14,000g for 10 min at 4 °C. Immunoprecipitation was performed using the anti-a1 (A) or a2 (B) subunit of the AMPK antibody. Immunoprecipitates were then boiled in Laemmli sample buffer containing 100 mM DTT, then subjected to SDS–PAGE and finally to Western blotting using anti-phospho AMPK antibody. Representative data from four experiments are presented. Each column shows the means ± SE obtained from these experiments. *Significant difference (P < 0.05) relative to AMPK phosphorylation of control cells in the absence of AICAR, **significant difference (P < 0.05) relative to AMPK phosphorylation of CA expressing cells in the absence of AICAR, as determined by Student’s t test. N.S. means; not significant.

a2, but not the a1, catalytic subunit. These results strongly suggest the a2, rather than the a1, catalytic subunit to be involved in LKB1 induced AMPK activation. Given that LKB1 controls the AMPK cascade, induction of CA or DN LKB1 is expected to influence functions downstream from the AMPK cascade. As a downstream target for AMPK, we first evaluated ACCb phosphorylation and the regulation of key gluconeogenesis enzymes in FAO cells. As shown in Fig. 4A, ACCb phosphorylation essentially paralleled that of AMPK, suggesting that LKB1 controls ACCb phosphorylation

by altering AMPK activity. G6Pase expression was markedly suppressed in response to enhanced AMPK activity, i.e., by induction of CA or by AICAR treatments, while being unaffected by DN induction (Fig. 4B). These results indicate that the basal state of AMPK activity corresponds to maximal expression of G6Pase and that suppression of AMPK activity below the basal level does not lead to increased G6Pase expression. On the other hand, PEPCK expression in the basal state of AMPK activity was inhibited, while suppression of AMPK activity by DN LKB1 induction led to

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

599

Fig. 4. Effect of CA or DN LKB1 overexpression on AICAR-induced phosphorylation of ACCb (A), G6Pase (B), and PEPCK expression (C) in FAO cells. Fao cells were transfected with adenoviruses expressing Lac Z (lanes 1 and 2), CA LKB1 (lanes 3 and 4), and DN LKB1 (lanes 5 and 6). (A) After the cells had been incubated in KRB buffer with or without 2 mM AICAR for 30 min, cell lysates were resolved on 10% SDS–polyacrylamide gel, followed by electrophoretic transfer to a nitrocellulose membrane. Membranes were incubated for 1 h at RT with the antibody against phospho-ACCb. After blotting with the indicated secondary antibody, detection was performed using an ECL chemiluminescent kit. (B,C) cDNA was synthesized from the purified total RNA from Fao cells using reverse transcriptase. For quantitative analysis of glucose 6 phosphatase (G6Pase) (B) and phosphoenol carboxykinase (PEPCK) (C), we conducted real-time PCR using an ABI PRISM Model 7000 (Applied Biosystems, Foster City, CA). Representative data from four experiments are presented. Each column shows the means ± SE obtained from these experiments. *Significant difference (P < 0.05) relative to AMPK phosphorylation of control cells in the absence of AICAR, **significant difference (P < 0.05) relative to AMPK phosphorylation of CA expressing cells in the absence of AICAR, as determined by Student’s t test. N.S. means; not significant.

Fig. 5. Effect of CA or DN LKB1 overexpression on AICAR-induced glucose uptake in C2C12 myotubes (A), insulin dependent glucose uptake in L6 myotubes (B), and 3T3-L1 adipocytes (C). Cells were transfected with adenoviruses expressing Lac Z (lanes 1 and 2), CA LKB1 (lanes 3 and 4), and DN LKB1 (lanes 5 and 6). (A) After C2C12 myotubes had been incubated for 30 min in KRB buffer with or without 2 mM AICAR, they were incubated in glucose-free Krebs–Ringer phosphate (KRP) medium containing 2-deoxy-D-[1,2-3H(N)]glucose (2-DG) (2.25 mCi/ml). 2-DG uptake was measured during the last 20 min of incubation. (B,C) glucose Transport in L6 myotubes or 3T3-L1 adipocytes with or without 107 human insulin (Novo Nordisk) was performed as described previously [23]. After the cells had been lysed with 0.1% Triton, radioactivity was measured using liquid scintillation counting. Representative data from four experiments are presented. Each column shows the means ± SE obtained from these experiments. *Significant difference (P < 0.05) relative to AMPK phosphorylation of control cells in the absence of AICAR. N.S. means; not significant.

increased PEPCK expression (Fig. 4C). These subtle modulations by LKB1 may explain why only G6Pase is constantly activated in the basal state of AMPK activity, while PEPCK is activated only when AMPK activity is inhibited, such as in a state of starvation. We further examined the effect of LKB1 on the glucose uptake activity in C2C12 myotubes. Activation of AMPK was reported to increase glucose uptake in an insulin independent manner in skeletal muscle or myotube-derived cultured cells [6,11]. As shown in Fig. 5A, AICAR treatment or the induction of CA resulted in a slight, but statistically significant, increase in 2-DG uptake as compared with the control cells, while 2-DG uptake was suppressed in DN

expressing cells. These results suggest that LKB1 controls glucose uptake by altering AMPK activity. Then, we examined whether altering LKB1 activity would have effects on the insulin-stimulated glucose uptake in L6 myotubes and 3T3-L1 adipocytes. When CA LKB1 was overexpressed, slight increases in 2-DG uptake (1.7- and 1.5-fold, respectively) were observed in the absence of insulin (Fig. 5B and C). AICAR treatments also enhanced 2-DG uptake (data not shown), as previously reported [25], indicating that AMPK activation enhances glucose transport activity in insulin responsive cells. Nevertheless, there were no significant differences in insulin-stimulated glucose uptake among Lac Z, CA, and DN overexpressing cells. These

600

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

results suggest that the mechanism of insulin-stimulated glucose uptake is not involved in AMPK activation. Discussion AMPK is a heterotrimeric complex comprised of an a catalytic subunit and regulatory b and c subunits. There are two isoforms of the a subunit, a1 and a2. In response to energy stress, AMP directly binds to tandem repeats of the crystathionine-b synthase (CBS) domains in AMPK c subunits, causing a conformational change that exposes the activation loop in the a subunit, allowing it to be phosphorylated by LKB1 [26]. A previous report suggested that binding of AMP to AMPK is likely to be the principal regulatory mechanism stimulating phosphorylation of AMPK by LKB1 [27]. In this study, AMPK activation by AMP or treatment with AICAR, which is converted into an AMP mimetic within the cell, was abolished in DN overexpressing cells. Our observations strongly support the idea that intact LKB1 activity is required for AMPK activation by AMP. In contrast, no phenformin-induced AMPK phosphorylation was observed. The mechanism by which metformin, or its closely related analog phenformin, activates AMPK has not yet been established, but is not thought to involve changes in intracellular levels of AMP [28,29]. Our results may account for the difference in AMPK activation mechanisms between AMP and phenformin. The effect of LKB1 overexpression on adiponectin-induced AMPK activation was also investigated. Based on our results, adiponectin is likely to be involved in LKB1-induced AMPK activation, though the mechanism remains unclear. Further study is anticipated to clarify this issue, including the possibility of LKB1 being located in the adiponectin signaling pathway. Recently, besides LKB1, Ca(2+)/calmodulin-dependent protein kinase kinase b (CaMKKbeta) was also observed to activate AMPK [30]. Moreover, thrombin was reported to activate AMPK in endothelial cells via a pathway involving CaMKKbeta [31]. In that earlier study, downregulation of LKB1 abolished phosphorylation of AMPK with AICAR, which is compatible with our data, while having no effect on thrombin-induced AMPK activation. Thus, we suspect that the mechanism by which CaMKK activates AMPK is likely to be different from that of LKB1. In our study, overexpression of constitutive active and dominant negative forms of LKB1 resulted in increased and decreased phosphorylation of Thr172, in a isoform of AMPK, respectively. As AMPK activity is thought to parallel AMPK phosphorylation, our results indicate that LKB1 controls AMPK activity as an upstream signaling molecule. According to a previous report [24], LKB1 forms a heterotrimeric complex, with regulatory proteins termed STRAD and MO25, and activation of AMPKK by LKB1 requiring STRAD and MO25. Thus, it is intriguing that the modulation of LKB1 activity on its own affected AMPK activity and also the AMPK-related cascade. As

a downstream target for AMPK, ACCb is phosphorylated by AMPK, which leads to the inhibition of fatty acid synthesis [4]. HNF4a is a downstream target of AMPK and one of the effects of AMPK activation is reduced expression of HNF4a target genes, such as those for enzymes involved in gluconeogenesis, e.g., G6Pase or PEPCK [5]. AICAR-stimulated activation of p38 MAPK is mediated by AMPK, and the p38 MAPK cascade is downstream from AMPK in the signaling pathway of AICAR-stimulated glucose transport [6]. Using myotubes and hepatoma cell lines, our present study provides convincing evidence that LKB1 plays a major role in regulating all of these downstream signals of AMPK. Although there are two isoforms of AMPKa, a1 and a2, AICAR treatment or moderate-intensity endurance muscle contraction preferentially stimulates only the a2 isoform [32]. In addition, AICAR-stimulated glucose transport was abolished in skeletal muscles from whole body a2 KO mice, but not in muscles from whole body a1 KO mice [33]. These previous observations strongly suggest the AMPK activation, which follows a rise in the cellular AMP:ATP ratio, to be mediated only by the a2, rather than the a1, isoform. In the present study, neither treatment with AICAR nor induction of CA LKB1 enhanced phosphorylation of the AMPK a1 isoform. However, a previous study found that LKB1 activates AMPK complexes containing either AMPK a1 or a2 in cell-free assays [27], which contradicts our data. This discrepancy may be attributable to the experimental conditions, i.e., intra-versus extracellular. In conclusion, the induction of constitutive active or dominant negative LKB1 in C2C12 myotubes and the FAO hepatoma cell line resulted in a marked increase or decrease in AMPK phosphorylation, respectively. Glucose uptake and ACCb phosphorylation paralleled the AMPK phosphorylation alteration in response to LKB1 induction. Furthermore, G6Pase expressions were inhibited by CA LKB1 induction while PEPCK expressions were increased by DN LKB1 induction in FAO cells. These results indicate that LKB1 is a crucial regulator of AMPK activation in muscle and liver cells and, therefore, that LKB1 activity is of potential importance to our understanding of glucose and lipid metabolism. References [1] D.G. Hardie, D. Carling, The AMP-activated protein kinase: fuel gauge of mammalian cells, Eur. J. Biochem. 246 (1997) 259–273. [2] D.G. Hardie, D. Carling, M. Carlson, The AMP-activated/SNF1 protein kinase subfamily: metabolic sensors of the eukaryotic cell? Annu. Rev. Biochem. 67 (1998) 821–855. [3] D.G. Hardie, J.W. Scott, D.A. Pan, E.R. Hudson, Management of cellular energy by the AMP-activated protein kinase system, FEBS Lett. 546 (2003) 113–120. [4] W.W. Winder, D.G. Hardie, Inactivation of acetyl-CoA carboxylase and activation of AMP-activated protein kinase in muscle during exercise, Am. J. Physiol. Endocrinol. Metab. 270 (1996) E299–E304. [5] Y.H. Hong, U.S. Varanasi, W. Yang, T. Leff, AMP-activated protein kinase regulates HNF4alpha transcriptional activity by inhibiting

K. Imai et al. / Biochemical and Biophysical Research Communications 351 (2006) 595–601

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

dimer formation and decreasing protein stability, J. Biol. Chem. 278 (2003) 27495–27501. X. Xi, J. Han, J.Z. Zhang, Stimulation of glucose transport by AMPactivated protein kinase via activation of p38 mitogen-activated protein kinase, J. Biol. Chem. 276 (2001) 41029–41034. S.J. Baron, J. Li, R.R. Russell, D. Neumann, E.J. Miller, R. Tuerk, T. Wallimann, R.L. Hurley, L.A. Witter, L.H. Young, Dual mechanism regulating AMPK kinase action in the ischemic heart, Circ. Res. 96 (2005) 337–345. M.H. Zou, X.Y. Hou, C.M. Chi, S. Kirkpatick, F. Liu, M.H. Goldman, R.A. Cohen, Activation of 5 0 -AMP-activated kinase is mediated through c-Src and phosphoinositide 3-kinase activity during hypoxia- reoxygenation of bovine aortic endothelial cells: Role of peroxynitrite, J. Biol. Chem. 278 (2003) 34003–34010. S.L. Choi, S.J. Kim, K.T. Lee, J. Kim, J. Mu, M.J. Birnbaum, S.S. Kim, J. Ha, The regulation of AMP activated protein kinase by H2O2, Biochem. Biophys. Res. Commun. 287 (2001) 92–97. T. Yamauchi, J. Kamon, Y. Minokoshi, Y. Ito, H. Waki, S. Uchida, S. Yamashita, M. Noda, S. Kita, K. Ueki, K. Eto, Y. Akanuma, P. Froguel, F. Foufelle, P. Ferre, D. Carling, S. Kimura, R. Nagai, B.B. Kahn, T. Kadowaki, Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase, Nat. Med. 8 (2002) 1288–1295. G.F. Merrill, E.J. Kurth, D.G. Hardie, W.W. Winder, AICA riboside increases AMP-activated protein kinase, fatty acid oxidation, and glucose uptake in rat muscle, Am. J. Physiol. 273 (1997) E1107– E1112. G. Zhou, R. Myers, Y. Li, Y. Chen, X. Shen, J. Fenyk-Melody, M. Wu, J. Ventre, T. Doebber, N. Fujii, N. Musi, M.F. Hirshman, L.J. Goodyear, D.E. Moller, Role of AMP-activated protein kinase in mechanism of metformin action, J. Clin. Invest. 108 (2001) 1167– 1174. D.E. Jenne, H. Reimann, J. Nezu, W. Friedel, S. Loff, R. Jeschke, O. Muller, W. Back, M. Zimmer, Peutz–Jeghers syndrome is caused by mutations in a novel serine threonine kinase, Nat. Genet. 18 (1998) 38–43. A. Woods, S.R. Johnstone, K. Dickerson, F.C. Leiper, L.G. Fryer, D. Neumann, U. Schlattner, T. Wallimann, M. Carison, D. Carling, LKB1 is the upstream kinase in the AMP-activated protein kinase cascade, Curr. Biol. 13 (2003) 2004–2008. S.A. Hawley, J. Boudeau, J.L. Reid, K.J. Mustard, L. Udd, T.P. Makela, D.R. Alessi, D.G. Hardie, Complexes between the LKB1 tumor suppressor, STRADalpha/beta and MO25alpha/beta are upstream kinases in the AMP-activated protein kinase cascade, J. Biol. 2 (2003) 28. J. Boudeau, M. Deak, M.A. Lawlor, N.A. Morrice, D.R. Alessi, Heat shock protein 90 and Cdc37 interact with LKB1 and regulate its stability, Biochem. J. 370 (2003) 849–857. A.F. Baas, J. Boudeau, G.P. Sapkota, L. Smit, R. Medema, N.A. Morrice, D.R. Alessi, H.C. Clevers, Activation of the tumor suppressor kinase LKB1 by the STE20-like pseudokinase STRAD, EMBO J. 22 (2003) 3062–3072. J. Boudeau, J.W. Scott, N. Resta, M. Deak, A. Kieloch, D. Komander, D.G. Hardie, A.R. Prescott, D.M. Aalten, D.R. Alessi, Analysis of the LKB1–STRAD–MO25 complex, J. Cell Sci. 117 (2004) 6365–6375. K. Inukai, A.M. Shewan, W.S. Pascow, S. Katayama, D.E. James, Y. Oka, Carboxy terminus of glucose transporter 3 contains an apical membrane targeting domain, Mol. Endocrinol. 18 (2004) 339–349. K. Inukai, M. Funaki, M. Nawano, H. Katagiri, T. Ogihara, M. Anai, Y. Onishi, H. Sakoda, H. Ono, Y. Fukushima, M. Kikuchi, Y. Oka, T. Asano, The N-terminal 34 residues of the 55 kDa regulatory subunits of phosphoinositide 3-kinase interact with tubulin, Biochem. J. 346 (2000) 483–489.

601

[21] K. Inukai, T. Asano, H. Katagiri, M. Anai, M. Funaki, H. Ishihara, K. Tsukuda, M. Kikuchi, Y. Yazaki, Y. Oka, Replacement of both tryptophan residues at 388 and 412 completely abolished cytochalasin B photolabeling of the GLUT1 glucose transporter, Biochem. J. 302 (1994) 355–361. [22] S. Miyake, M. Makimura, Y. Kanegae, S. Harada, Y. Sato, K. Takamori, C. Tokuda, I. Saito, Efficient generation of recombinant adenoviruses using adenovirus DNA-terminal protein complex and a cosmid bearing the full-length virus genome, Proc. Natl. Acad. Sci. USA 93 (1996) 1320–1324. [23] H. Sakoda, T. Ogihara, M. Anai, M. Fujishiro, H. Ono, Y. Onishi, H. Katagiri, M. Abe, Y. Fukushima, N. Shojima, K. Inukai, M. Kikuchi, Y. Oka, T. Asano, Activation of AMPK is essential for AICAR-induced glucose uptake by skeletal muscle but not adipocytes, Am. J. Physiol. Endocrinol. Metab. 282 (2002) E1239–E1244. [24] J.M. Lizcano, O. Goransson, R. Toth, M. Deak, N.A. Morrice, J. Boudeau, S.A. Hawley, L. Udd, T.P. Makela, D.G. Hardie, D.R. Alessi, LKB1 is a master kinase that activates 13 kinases of the AMPK subfamily, including MARK/PAR-1, EMBO J. 23 (2004) 833–843. [25] S. Yamaguchi, H. Katahira, S. Ozawa, Y. Nakamichi, T. Tanaka, T. Shimoyama, K. Takahashi, K. Yoshimoto, M. Ohara, S. Nagamatsu, H. Ishida, Activators of AMP-activated protein kinase enhance GLUT4 translocation and its glucose transport activity in 3T3-L1 adipocytes, AM. J. Physiol. Endocrinol. Metab. 289 (2005) E643– E649. [26] B.B. Kahn, T. Alquier, D. Carling, D.G. Hardie, AMP-activated protein kinase: ancient energy gauge provides clues to modern understanding of metabolism, Cell Metab. 1 (2005) 15. [27] S.A. Hawley, J. Boudeau, J.L. Reid, K.J. Mustard, L. Udd, T.P. Makela, D.R. Alessi, D.G. Hardie, Complexes between the LKB1 tumor suppressor, STRADalpha/beta and MO25alpha/beta are upstream kinases in the AMP-activated protein kinase cascade, J. Biol. 2 (2003) 28. [28] L.G. Fryer, A. Parbu-Patel, D. Carling, The anti-diabetic drugs rosiglitazone and metformin stimulate AMP-activated protein kinase through distinct signaling pathways, J. Biol. Chem. 277 (2002) 25226– 25232. [29] S.A. Hawley, A.E. Gadalla, G.S. Olsen, D.G. Hardie, The antidiabetic drug metformin activates the AMP-activated protein kinase cascade via an adenine nucleotide-independent mechanism, Diabetes 51 (2002) 2420–2425. [30] A. Woods, K. Dickerson, R. Heath, S.P. Hong, M. Momcilovic, S.R. Johnstone, M. Carison, D. Carling, Ca2+/dependent protein kinase kinase-beta acts upstream of AMP-activated protein kinase in mammalian cells, Cell Metab. 2 (2005) 21–33. [31] N. Stahmann, A. Woods, D. Carling, R. Heller, Thrombin activates AMP-activated protein kinase in endothelial cells via a pathway involving Ca2+/dependent protein kinase kinase beta, Mol. Cell Biol. 26 (2006) 5933–5944. [32] M. Nakano, T. Hamada, T. Hayashi, S. Yonemitsu, L. Miyamoto, T. Toyoda, S. Tanaka, H. Matsuzaki, K. Ebihara, Y. Ogawa, K. Hosoda, G. Inoue, Y. Yoshimasa, A. Otaka, T. Fushiki, K. Nakao, a2 isoform-specific activation of 5 0 adenosine monophosphate-activated protein kinase by 5-aminoimidazole-4-carboxamide-1-b-D-ribonucleoside at a physiological level activates glucose transport and increase glucose transporter 4 in mouse skeletal muscle, Metabolism 55 (2006) 300–308. [33] S.B. Jorgensen, B. Viollet, F. Andreelli, C. Frosig, J.B. Birk, P. Schjerling, S. Vaulont, E.A. Richter, J.F. Wojtaszewski, Knockout of the alpha2 but not alpha1 5 0 -AMP-activated protein kinase isoform abolishes 5-aminomidazole-4-carboxamide-1-beta-ribofuranoside but not contraction-induced glucose uptake in skeletal muscle, J. Biol. Chem. 279 (2004) 1070–1079.