The glucose-lowering agent sodium tungstate increases the levels and translocation of GLUT4 in L6 myotubes through a mechanism associated with ERK1/2 and MEF2D
The aim of this study was to investigate the action of the glucose-lowering compound sodium tungstate on glucose transport in muscle myotubes and to unravel the molecular events underlying the effects observed.
We studied the effects of tungstate on 2-deoxy-d-glucose uptake, levels and translocation of the glucose transporters GLUT4 and GLUT1, and Glut4 (also known as Slc2a4) promoter activity. We also measured the modifications of individual components of the signalling pathways involved in the effects observed.
Tungstate increased 2-deoxy-d-glucose uptake in differentiated L6 myotubes through an increase in the total amount and translocation of GLUT4 transporter. The effects on glucose uptake were additive to those of insulin. Tungstate activated transcription of the Glut4 promoter, as shown by an increase in Glut4 mRNA, and by a promoter reporter assay. The assay of deletions of the Glut4 promoter indicated that the effect of tungstate is mediated by the myocyte enhancer factor 2 (MEF2)-binding domain. Accordingly, MEF2 levels and DNA binding activities were increased in response to the treatment. Tungstate-induced glucose uptake and GLUT4 transcriptional activation were dependent on the activation of extracellular signal-regulated kinases 1 and 2 (ERK1/2), while no changes were observed in the phosphorylation state of the β subunit of the insulin receptor, in the phosphatidylinositol 3-kinase pathway or in the activation of 5′AMP-activated protein kinase.
Tungstate activates glucose uptake in myotubes through a novel ERK1/2-dependent mechanism. This effect is exerted by an increase in the content and translocation of the GLUT4 transporter. This is the first report of a glucose-lowering compound activating Glut4 transcription through an ERK1/2-dependent increase in MEF2 levels.
KeywordsERK1/2 GLUT1 GLUT4 GLUT4 promoter L6 myotubes MEF2 family Sodium tungstate
5′AMP-activated protein kinase
electrophoretic mobility shift assay
extracellular signal-regulated kinases 1 and 2
myocyte enhancer factor 2
nuclear factor 1
The major metabolic pathways regulated by insulin include the stimulation of glucose transport in muscle. Glucose uptake is mediated in this tissue mainly by the insulin-regulated glucose transporter GLUT4. In the absence of insulin, this transporter slowly recycles between an intracellular storage compartment and the plasma membrane. In response to the hormone or to physical exercise, the amount of GLUT4 located at the plasma membrane increases, which accounts for the augmented transport capacity . Insulin-induced GLUT4 translocation is dependent on the activation of phosphatidylinositol 3-kinase (PI3K) [2, 3], while physical exercise enhances translocation by the activation of 5′AMP-activated protein kinase (AMPK) . A novel mechanism of glucose signalling through a PI3K-independent mechanism, but dependent on the activation of extracellular signal-regulated kinases 1 and 2 (ERK1/2) and phospholipase D, has also been described and may be relevant to GLUT translocation [2, 5].
A second regulatory mechanism of glucose transport is the control of GLUT4 (also known as SLC2A4) gene expression, which is subject to tissue, hormonal and metabolic regulation (as reviewed in ). Changes in GLUT4 gene expression levels are observed in physiological states of altered glucose homeostasis and in those of relative insulin deficiency . The GLUT4 gene is subjected to a complex system of tissue-specific and metabolic regulation. Two regulatory domains of the 5′-flanking DNA region are required to direct a complete programme of GLUT4 gene regulation [7, 8, 9, 10]. The region referred to as domain II or myocyte enhancer factor 2 (MEF2) domain binds the isoforms of the MEF2 family of transcription factors (MEF2A, MEF2B, MEF2C and MEF2D) [11, 12, 13]. A second region, named domain I, binds nuclear factor I (NF1)  and a transcriptional activator named GLUT4 enhanced factor [13, 14, 15].
Tungstate is an oral glucose-lowering [16, 17, 18, 19, 20, 21, 22] and anti-obesity  agent. This compound has a low toxicity profile in animals and humans, and is currently undergoing Phase II clinical trials. Tungstate normalises carbohydrate metabolism in liver [16, 18, 19], stimulates insulin secretion and regenerates pancreatic beta cells in neonatally streptozotocin-induced diabetic rats . In streptozotocin-induced diabetic rats, tungstate also increases GLUT4 production and translocation in diaphragm . Here, we dissect the molecular mechanisms that drive the tungstate-induced increase in glucose transport in muscle cells. We found that tungstate increases glucose transport in L6 differentiated myotubes through an ERK1/2-dependent mechanism involving an increase in the production and translocation of GLUT4. This double effect on glucose transport emphasises the glucose-lowering potential of this compound.
Sodium tungstate, PD98059 and horseradish peroxidase-conjugated secondary antibodies were from Sigma (St Louis, MO, USA). 5-Aminoimidazole-4-carboxamide-1-β-d-ribofuranoside (AICAR) was from Toronto Research Chemicals (North York, ON, Canada). Tissue culture media and supplements were from Sigma and Invitrogen (Carlsbad, CA, USA). FCS was from Cultek (Madrid, Spain). ERK1/2 and phosphotyrosine (4G10) antibodies were from Upstate (Waltham, MA, USA). Phospho-p44/42 mitogen-activated protein kinase E10 (Thr202/Tyr204), AMPKα2 and phospho-AMPKα2 (Thr172) antibodies were from Cell Signaling (Beverly, MA, USA). Phospho-protein kinase B/Akt (Ser473) antibody was from Nanotools (Teningen, Germany). GLUT1, NF1, MEF2, MEF2C and MEF2D antibodies were from Santa Cruz Biotechnologies (Santa Cruz, CA, USA).
The L6.C11 rat skeletal muscle myoblast line (ECACC No.92102119) was grown in DMEM containing 10% (vol./vol.) FCS. Cells were differentiated by culturing for 5 to 6 days in DMEM containing 2% (vol./vol.) FCS (>50% fusion into multinucleated myotubes). Prior to each experiment, myotubes were incubated in FCS-free medium for 18 to 24 h. For long-term incubations, a tungstate concentration of 100 μmol/l was chosen, while higher concentrations (500 μmol/l to 1 mmol/l) were used for short-term experiments.
Differentiated myotubes were incubated in FCS-free medium for 18 to 24 h. Tungstate and insulin treatments were performed in serum-free medium. Triplicate measurements of 2-deoxy-d-[1-3H]glucose (2-DG) uptake were taken after 10 min of incubation as described .
Membrane fractions from myotubes were prepared as described [26, 27]. 5′-Nucleotidase and cytochrome c reductase activities were assayed as marker enzymes for plasma membranes and low-density microsomes, respectively .
GLUT1 and GLUT4 protein analyses
L6 myotube membrane fraction proteins were electrophoresed in 10% (wt/vol.) SDS-polyacrylamide gels and processed for western blot with anti-GLUT1 or anti-GLUT4 antibodies (Biogenesis, Poole, UK). Immunoreactive bands were visualised by chemiluminescence and quantified with NIH Image Software .
cDNA was generated with a kit (First-strand cDNA Synthesis; GE Health Care, Uppsala, Sweden) using total RNA from the myotubes . Real-time PCR was performed using SYBR Green I dye and the primers: Glut4 5′-GTCATCAACGCCCCACAGAA-3′ and 5′-GCAAGGACAGTGGACGCTCTCTTTC-3′; MyoD (also known as Myod1) 5′-AGGGAAGGGAAGAGCAGAAG-3′ and 5′-TACACCTGTTACACCCGAGAT-3′; and β-actin 5′-GGCCAACCGTGAAAAGATG-3′ and 5′-GGATCTTCATGAGGTAGTCTGTC-3′. Amplifications were carried out using a Stratagene Mx3000P system (Cedar Creek, TX, USA). In each run, standard curves were generated for a primer set by serial dilution of plasmid DNA encoding the relevant cDNA. Melting curves were generated after each run and the fidelity of RT-PCR was also confirmed by electrophoresis (not shown). Glut4 and MyoD mRNA levels were calculated using the real-time PCR standard curve method.
Protein phosphorylation analysis
L6 myotubes were incubated in FCS-free medium for 18 to 24 h and then treated with tungstate (1 mmol/l) or insulin (100 nmol/l) in serum-free medium. After treatment, plates were processed as described in . Proteins were separated by SDS-PAGE and immunoblotted with selected antibodies.
Western blots of 5′AMP-activated protein kinase
L6 myotubes were incubated in FCS-free medium for 24 h and then treated in the absence or presence of insulin (100 nmol/l), sodium tungstate (1 mmol/l) or AICAR (0.5 mmol/l). Plates were flash-frozen in liquid nitrogen and scraped with lysis buffer: 50 mmol/l Tris–HCl (pH 7.4), 150 mmol/l NaCl, 1% (vol./vol.) Triton X-100, 10% (vol./vol.) glycerol, 1 mmol/l EDTA, 1 mmol/l EGTA, 10 mmol/l NaF, 10 mmol/l sodium pyrophosphate, 1 mmol/l sodium orthovanadate, 1 mmol/l dithiothreitol, 4 μg/l leupeptin and 0.5 mmol/l phenylmethylsulfonyl fluoride. The lysates were centrifuged for 30 min at 11,000×g and 4°C. Samples were separated by SDS-PAGE and immunoblotted with anti-AMPKα2 or anti phospho-AMPKα2 (Thr172) antibodies.
Design of reporter gene constructs
The pGL3-ratGLUT4 construct was made by cloning a genomic DNA fragment encompassing positions −2212 to + 164 of the rat Glut4 gene from clone −2212/+147G4CAT (a gift from A. Zorzano ), into pGL3-Basic vector (Promega, Madison, WI, USA). pGL3-ratGLUT4Δ1 was obtained by deletion of a SmaI–BstXI fragment of pGL3-ratGLUT4, encompassing positions −502 to −243. pGL3-ratGLUT4Δ2 was generated by removing a SmaI–SacI fragment, which encompasses positions −1042 to −502. The sequence of all constructs was verified by automated DNA sequencing.
DNA transfection assays
Cells were used at 80–90% confluence. Transfection was performed using LipofectAMINE 2000 (Invitrogen). The DNA mixture comprised the various GLUT4-luciferase reporters and the reference plasmid pRL-TK (ratio 1:5). The pRL-TK plasmid codes for Renilla luciferase under the control of the tk promoter as an internal control. Following removal of the transfection reagent, fresh medium was added to the cultures. In the case of cultures set to differentiate into myotubes, differentiation medium was applied 24 h later. Next, where indicated, the cells were incubated for 24 h with insulin (100 nmol/l) or tungstate (0.1–1 mmol/l), and luciferase activity was determined using the dual luciferase method (Promega). Results were standardised for Renilla luciferase activity. All transfections were performed at least three times, in triplicate, using at least two preparations of plasmid DNA. To allow comparison of the expression patterns between myoblasts and myotubes, the data are expressed as relative changes in luciferase activity: luciferase activity of control myoblasts was normalised to a value of 100% and the remaining activities were referred to this value.
Preparation of nuclear extracts
Nuclear extracts from L6 cells were prepared as described . Aliquots of the extracts (40 μg) were resolved by 10% (wt/vol.) SDS-PAGE and analysed by immunoblotting using anti-NF1, anti-MEF2, anti-MEF2C or anti-MEF2D antibodies.
Electrophoretic mobility shift assays
Electrophoretic mobility shift assays (EMSAs) were performed using a double-stranded oligonucleotide probe that contains the functional MEF2 binding site present in the Glut4 promoter: forward, 5′-GATCGCTCTAAAAATAACCCTGTCG-3′; reverse, 5′-CGACAGGGTTATTTTTAGAGCGATC-3′; the recognition sequence for MEF2 is italicised . The double-stranded oligonucleotide was end-labelled using T4 polynucleotide kinase and γ-[32P]ATP (GE Healthcare). Binding reactions of the probes (200 fmol) were performed with 10 μg protein from nuclear extracts, at 4°C for 45 min, in 20 μl of the binding buffer consisting of 20 mmol/l HEPES, pH 7.9, 2.5 mmol/l MgCl2, 10% (vol./vol.) glycerol, 1 mmol/l DTT, 0.1 mg/ml bovine serum albumin and 30 ng/μl of polydeoxyinosinic-deoxycytidylic acid. Competitive assays were conducted under the same conditions, with the addition of 100-fold molar excess of unlabelled oligonucleotides. For supershift studies, extracts were pre-incubated on ice for 30 min with 1 μg of MEF2A, MEF2C or MEF2D antibody before addition of the labelled probe. The DNA–protein complexes were electrophoresed on 5% (wt/vol.) non-denaturing polyacrylamide gels at 4°C in 45 mmol/l Tris, 45 mmol/l borate pH 8 and 1 mmol/l EDTA buffer.
Creatine kinase activity assay
Creatine kinase activity in L6 myotube cultures was assayed in accordance with a previously described method . Cells were lysed with 1% (vol./vol.) Triton X-100 in PBS. After centrifugation (16,000×g), creatine kinase activity of the supernatant fractions was measured using a creatine kinase NAC kit (Chemelex, Barcelona, Spain).
Results are expressed as means ± SEM for the number of experiments indicated. The statistical significance of variations was evaluated using one-way ANOVA. When a significant effect was found, post hoc comparisons of the means were done using the t adjusted Tukey test. A p value <0.05 was considered significant.
Tungstate stimulates glucose uptake
Tungstate treatment increases the total amount of GLUT4 and transcription of the gene
To determine the actions of tungstate on the transcription of Glut4, we assayed the effects of this compound on the Glut4 promoter. For this purpose, a pGL3-ratGLUT4 luciferase construct encompassing positions −2212 to +164 of the rat Glut4 gene that contain the main regulatory elements of the promoter  was transfected in L6 cells. Differentiated and non-differentiated cells were pre-incubated for 24 h with FCS-free medium and further incubated (24 h) with insulin (100 nmol/l) or tungstate (100 μmol/l and 1 mmol/l). The luciferase activity in myotubes was approximately twice that in myoblasts (Fig. 2c). Insulin treatment decreased the activity of the Glut4 promoter in myoblasts and myotubes, as previously described in adipocytes . In contrast, tungstate increased luciferase activity up to 1.6-fold at 100 μmol/l and 1 mmol/l. In non-differentiated cells, the compound still had a significant, but lower effect (Fig. 2c).
We next examined the role of domains I and II of the Glut4 promoter  on tungstate-stimulated Glut4 transcription. For this purpose, we used two constructs: pGL3-ratGLUT4Δ1, which includes the MEF2 binding site in the rat promoter [11, 31], and pGL3-ratGLUT4Δ2, which includes the putative NF1 binding site . L6 cells were transfected and luciferase activity was assayed in myotubes after the addition of insulin (100 nmol/l) or tungstate (1 mmol/l).
In contrast to the pGL3-ratGLUT4Δ2 construct, basal transcription levels from pGL3-ratGLUT4Δ1 were similar to those with the undeleted construct. The luciferase activity of myotubes transfected with pGL3-ratGLUT4Δ1 was not modified by treatment with tungstate. In contrast, insulin-treated cells still showed a significant reduction in transcription (Fig. 2d). These data indicate that the MEF2 binding site is essential for the promotion of Glut4 transcription induced by tungstate. To further this observation, we first studied whether tungstate treatment increased the amount of individual components of this family of transcription factors. A comparative analysis showed that MEF2A and MEF2C nuclear levels were moderately increased in tungstate-treated myotubes (∼1.25-fold) compared with either control or insulin-treated cells (∼0.85-fold). Strikingly, the nuclear content of MEF2D in tungstate-treated cells was increased more than twofold (Fig. 3a). Second, we analysed by means of an EMSA with a double-stranded oligonucleotide corresponding to the rat GLUT4 MEF2 binding domain whether MEF2 binding activity was modified by tungstate (Fig. 3c). MEF2 binding activity was significantly increased in tungstate-treated myotubes (∼1.65-fold) compared with control cells. A supershift analysis for distinct MEF2 isoforms using anti-MEF2 A, C and D antibodies showed changes in band mobility for MEF2D in the tungstate-treated cells (Electronic supplementary material [ESM] Fig. 1).
Since the transcription of the gene encoding creatine kinase (Ckm) is upregulated by MEF2 , we tested the effects of tungstate on total creatine kinase activity in L6-treated myotubes as additional proof of the action of this compound on this family of transcription factors. Creatine kinase activity was increased in cells treated either with insulin (20 or 100 nmol/l) or tungstate (0.1 mmol/l). The combination of both effectors produced an additive effect (Fig. 3d).
Taken together, these results indicate that tungstate treatment increased the total amount of GLUT4 in L6 myotubes through an upregulation of Glut4 transcription driven by MEF2 transcription factors.
Tungstate treatment increases GLUT4 translocation to the plasma membrane
Tungstate-induced GLUT4 translocation is independent of PI3K and AMPK activation
To gain insight into the targets of tungstate action involved in its actions in muscle cells, we studied how the key components of the insulin and AMPK transduction pathways are modulated by this compound. We first analysed the involvement of the insulin receptor in the tungstate-mediated increase in GLUT4 translocation. After incubation (5, 15 and 60 min) of myotubes with tungstate, no significant changes were observed in the phosphorylation of the insulin receptor β subunit. In contrast, insulin induced a rapid phosphorylation of its receptor (not shown).
We next assayed whether tungstate-stimulated GLUT4 translocation was dependent on the activation of AMPK . We analysed the key phosphorylation site for AMPK activation, Thr172 from α subunit. Neither tungstate nor insulin treatment modified the phosphorylation of this kinase, in contrast to the AMPK activator, AICAR (Fig. 5c).
Tungstate-stimulated GLUT4 synthesis and translocation is dependent on stimulation of the ERK1/2 signalling pathway
Since our data indicate that tungstate-induced activation of ERK1/2 is independent of PI3K or AMPK activation, we examined whether it was dependent on a transduction pathway variant that is stimulated by glucose [5, 36]. This pathway ultimately involves the ERK1/2-dependent activation of phospholipase D. To address this question, we measured 2-DG uptake in the presence of n-butanol, an inhibitor of phospholipase D . AICAR was used as positive control of 2-DG uptake. The positive effect on 2-DG uptake exerted by tungstate or AICAR (50 μmol/l) was completely blocked by pre-incubation with 1% (vol./vol.) n-butanol (Fig. 7b).
The effects of tungstate, a well-characterised glucose-lowering compound, have been studied in several key tissues involved in the maintenance of glucose homeostasis, namely liver, pancreas and adipose tissue [16, 17, 18, 19, 23, 24]. In contrast, little is known about the effects of this compound in muscle. A previous report showed that tungstate treatment normalises both the intracellular distribution of GLUT4 and its levels in diaphragms of diabetic animals, thereby suggesting that it increases glucose uptake in muscle . The present study addresses this question at the molecular level. First, our results demonstrate that tungstate treatment increases 2-DG transport in myotubes, through a rise in the total amount of GLUT4 transporter (long-term effect) and its translocation to the plasma membrane (short-term effect). Moreover, the short-term effects of tungstate and insulin on 2-DG uptake were additive, thereby indicating distinct but convergent signalling pathways. This finding may be of relevance for treatment purposes, as by acting through an alternative mechanism, this compound may circumvent the insulin resistance that is characteristic of type 2 diabetic patients. Accordingly, the tungstate-induced increase in 2-DG uptake did not involve activation of the insulin receptor and members of its amplification cascade. In addition, we show that this increase is also independent of the phosphorylation of AMPK. These observations indicate the existence of a PI3K- and AMPK-independent pathway to increase capacity for glucose uptake in muscle cells upon treatment with tungstate. We show that the tungstate-induced effects on GLUT4 (increased translocation and production) are dependent on the activation of ERK1/2. In this context, several authors have associated the activation of ERK1/2 with increased glucose uptake in adipocytes and muscle cells when incubated with elevated concentrations of glucose [2, 5]. They propose that high glucose concentrations increase GLUT4 translocation through the activation of AMPK and initiation of a signalling cascade involving ERK1/2 phosphorylation, and through the activation of phospholipase D as well as atypical protein kinase Cs. Our results suggest that tungstate partly stimulates this pathway, but without the participation of AMPK. This is supported by the observation that an inhibitor of phospholipase D blocks short-term tungstate-mediated 2-DG uptake. Furthermore, tungstate-stimulated glycogen deposition in hepatocytes occurs through an ERK1/2-dependent mechanism . This observation indicates that the activation of these kinases is a key step underlying the glucose-lowering actions of this compound.
One of the key actions of tungstate in myotubes is the stimulation of Glut4 gene transcription. Tungstate-induced Glut4 promoter activation requires the presence of the domain containing the MEF2 binding site. In addition, tungstate-treated myotubes increased the levels and DNA-binding activities of MEF2A, MEF2C and MEF2D. This increase was confirmed after analysis of creatine kinase activity in tungstate-treated myotubes (in muscle, the Ckm promoter contains a canonical MEF2 binding site) [33, 34]. This effect could contribute to the improvement in cardiac performance of tungstate-treated streptozotocin-induced diabetic rats . Taken together, these results indicate that the key point in the tungstate-stimulated transcription of Glut4 is the induced production of members of the MEF2 family. It has been described that Glut4 transcription is significantly upregulated through an increase in the production or activity of MEF2 transcription factors . In fact, an increase in MEF2A and MEF2D levels contributes to the adaptation of skeletal muscle to exercise  while, in diabetic animals, MEF2 DNA-binding activity and/or concentration is substantially reduced and correlates with a decrease in Glut4 transcription [12, 38].
In contrast to the effects of tungstate, insulin treatment moderately decreased the GLUT4 reporter gene activity in myotubes. This effect has not been previously described for muscle cells, but is consistent with what has been reported for adipocytes  and can be explained by the fourfold increase in NF1 levels induced by the hormone.
It is important to stress that the stimulatory actions of tungstate on Glut4 transcription were dependent on the activation of ERK1/2 through an increase in the content of MEF2, which again places these kinases at the centre of the mechanism of action of this compound.
In conclusion, our results show that tungstate treatment increases the potential for glucose transport in muscle cells by two mechanisms, both of which are ERK1/2-dependent. The first involves an increase in the translocation of GLUT4 and the second a MEF2-dependent increase in transcription of the transporter gene. Compounds like tungstate (currently in phase II clinical trials), which exert their action through the improvement of glucose uptake in skeletal muscle via an increase in MEF2 levels and GLUT4 translocation, are excellent candidates for restoring the impaired muscle glucose metabolism that is characteristic of diabetes mellitus, and through this reducing hyperglycaemia.
This work was supported by grants ERDF1FD97-0812 (Ministerio de Ciencia y Tecnología, Spain) and PAI-2003-04 (Plan Andaluz de Investigación), to R. Salto; and by grants SAF2004-06962, SAF2007-64722 (Ministerio de Educación y Ciencia, Spain), 2005-SGR-00570 (Generalitat de Catalunya), PI042402 (Instituto de Salud Carlos III [ISCIII], Spain), and a grant from Fundación Marcelino Botín to J. J. Guinovart. N. Sevillano is the recipient of a predoctoral fellowship from Fundación Ramón Areces. We thank T. Yates for assistance in preparing the English manuscript. CIBER de Diabetes y Enfermedades Metabólicas is an initiative of the ISCIII.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
- 2.Sajan MP, Bandyopadhyay G, Kanoh Y et al (2002) Sorbitol activates atypical protein kinase C and GLUT4 glucose transporter translocation/glucose transport through proline-rich tyrosine kinase-2, the extracellular signal-regulated kinase pathway and phospholipase D. Biochem J 362:665–674PubMedCrossRefGoogle Scholar
- 5.Bandyopadhyay G, Sajan MP, Kanoh Y et al (2001) Glucose activates protein kinase C-zeta /lambda through proline-rich tyrosine kinase-2, extracellular signal-regulated kinase, and phospholipase D: a novel mechanism for activating glucose transporter translocation. J Biol Chem 276:35537–35545PubMedCrossRefGoogle Scholar
- 20.Palanivel R, Sundravel S, Ravichandran P, Govindasamy S (1998) Insulin-like effect of ammonium para-tungstate on glycoprotein metabolism in streptozotocin diabetic rats. Med Sci Res 26:541–544Google Scholar
- 28.Rasband WS, Bright DS (1995) NIH Image: a public domain image processing program for the Macintosh. Microbeam Anal Soc J 4:137–149Google Scholar