Glucose transport is the rate limiting step to glucose metabolism in the fasted condition. Skeletal muscle is the main site of postprandial glucose uptake and metabolism in the body, and thus glucose uptake into skeletal muscle is essential for maintaining blood glucose homeostasis. Glucose transport has been shown to be depressed in muscle of diabetic and nondiabetic obese patients in vitro, and the decrease in transport is believed to be the cause of hyperglycemia.

Because glucose transport into muscle is a key element in maintaining glucose homeostasis, the question arises as to whether the level of expression of GLUT4 may be responsible for the depressed glucose transport that is seen in muscle of obese and diabetic patients. We have previously observed a 20% decrease in GLUT4 protein levels in skeletal muscle of obese and obese-Type 2 diabetic patients (³). However, when patients lost weight and regained insulin sensitivity, they did not have increased muscle GLUT4 protein (⁵). This suggests that depressed GLUT4 protein was not responsible for the insulin resistance. Other research in humans showed that GLUT4 levels in lean, obese, and Type 2 diabetes was not significantly altered (^2,24). Taken together these studies show that a decrease in GLUT4 protein is not responsible for insulin resistance.

Animal studies however, have shown the importance of GLUT4 in maintaining euglycemia in a mouse model of Type 2 diabetes (⁷). Introducing a GLUT4 transgene to cause overexpression of GLUT4, Gibbs’ group showed that upregulation of the GLUT4 protein restored glycemic control in diabetic mice (⁷). Additionally another group found that small changes in GLUT4 expression ultimately lead to a significant improvement in fasting blood glucose levels (¹³). Thus, although depressed GLUT4 protein is not responsible for insulin resistance, increased expression of GLUT4 protein could be a therapeutic target for the treatment of diabetes (⁴⁰).

THE EFFECT OF EXERCISE ON GLUT4 PROTEIN AND mRNA EXPRESSION IN SKELETAL MUSCLE

Exercise training has been shown to increase total GLUT4 protein in skeletal muscle in both rats and humans (^2,6,12,27,28). Exercise training in lean and obese Zucker rats increased GLUT4 protein by 1.7- and 2.3-fold, respectively, compared with GLUT4 protein levels in sedentary rats. These were rats run on a treadmill 5 d·wk⁻¹ for 18–30 wk and sacrificed 48 h after the last exercise session (⁶). Another study using male Sprague-Dawley rats measured GLUT4 protein content after 6 wk of voluntary running in exercise-wheel cages (²⁷). Rats were sacrificed 72 h after the last bout of exercise to avoid any residual effects of acute exercise on glucose metabolism (²⁷). GLUT4 protein content increased 60% in plantaris muscle. After one exercise session (two 3-h bouts separated by 45 min), rats had a 50% increase in GLUT4 protein expression in epitrochlearis 16 h after the session. After a second session, rats exhibited a twofold increase in GLUT4 protein (²⁵).

In studies of trained and sedentary men, the GLUT4 protein content of the trained group was roughly twofold greater than the sedentary group (¹⁰). Plasma insulin levels during an oral glucose tolerance test were also significantly lower in the trained men (¹⁰). A chronic exercise training study in humans was conducted with previously sedentary middle-aged men. The subjects for this study were enlisted to participate in a 14-wk exercise training regimen (¹²). GLUT4 protein concentration in skeletal muscle increased by 1.8-fold after 14 wk of exercise training (¹²). Short-term exercise training studies were also conducted in sedentary individuals. After 1 wk of cycle ergometer training, there was a 2.8-fold increase in muscle GLUT4 protein content in seven sedentary males (¹¹). In another study, eight healthy individuals who had not performed regular endurance exercise for at least 6 months before the experiment exercised for 7–10 d on a cycle ergometer (⁸). All subjects responded to the exercise with an increase in GLUT4. These studies suggest that increased muscle GLUT4 protein content may play a role in increasing insulin sensitivity with exercise training in patients of all ages, with or without diabetes (^2,10).

To determine whether exercise increases GLUT4 protein by increasing transcription of the gene (²⁰), trained and untrained rats were exercised 8 d on a treadmill then sacrificed at 30 min, 3 h, or 24 h after the last bout of exercise. Transcription of the GLUT4 gene, as detected by run-on analysis, was significantly increased at 3 h after exercise in both trained and untrained groups. The increase in GLUT4 gene transcription in the untrained rats was lower (1.4-fold) compared with the trained rats (1.8-fold) (²⁰). These data demonstrate that exercise transiently increases GLUT4 gene transcription. The level of GLUT4 mRNA increases but returns to baseline by 24 h, whereas GLUT4 protein increases daily in response to the transient increases in gene transcription and reaches a steady state after several days of training.

A more recent study using transgenic mice confirmed the increases in GLUT4 gene transcription (¹⁵). Using transgenic mice with a construct containing the chloramphenicol-acyltransferase (CAT) reporter gene driven by varying lengths of the human GLUT4 promoter, our lab demonstrated that exercise increased the expression of CAT, demonstrating that the GLUT4 promoter has “exercise responsive elements” (ERE).

REGULATION OF THE GLUT4 PROMOTER

The glucose transporter, GLUT4, demonstrates tissue-specific regulation and is regulated both hormonally and metabolically (¹⁹). Understanding how and where the GLUT4 promoter is regulated will help in developing strategies in treating insulin resistance (⁴⁰). To determine the DNA elements of the promoter that are necessary for regulation of the GLUT4 gene, studies were conducted using varying lengths of the human GLUT4 (hG4) promoter attached to the CAT reporter gene (²²). Transgenic mice with 2.4 kilobases of the 5′-flanking region of the hG4 gene had tissue-specific expression of CAT reporter that was the same as endogenous mouse GLUT4 mRNA. Smaller promoter constructs were generated at lengths of 1975, 1639, 1154, 730, and 412 bp of the hG4 fused to CAT, which allowed identification of regions important for normal GLUT4 regulation and expression in skeletal muscle (²²). Both the −412 bp and −730 bp promoters demonstrated a lack of tissue specificity, having expression in brain and liver, sites in which GLUT4 is not endogenously expressed. Interestingly, the −412 bp and −730 bp promoters also did not exhibit the characteristic decrease in CAT in the diabetic state as did the longer constructs. Likewise, insulin treatment, which ameliorated the diabetic state in the −1154 bp to −1975 bp constructs, did not change CAT levels in the −412 bp and −730 bp constructs. Upon inspection of the sequence between −412 bp and −895 bp, a consensus myocyte enhancer factor 2 (MEF2) binding site was located from −473 to −464 bp. To further study the importance of MEF2, six founder lines were developed where part of the MEF2 consensus sequence was mutated to determine whether the MEF2 site effected GLUT4 expression. No significant gene expression was found in GLUT4 expressing tissues when the MEF2 site was mutated. In mobility shift assays, Olson’s lab has also shown that in vitro translated MEF2A and MEF2C have the ability to carry out specific binding to the GLUT4 MEF2-binding sequence (³¹). These studies demonstrated that the shortest promoter length that still resulted in normal GLUT4 expression was 895 bp and that the MEF2 sequence was necessary.

Ezaki et al. (⁴) have also done extensive work to identify the regulatory regions of the GLUT4 promoter (^4,32). Using varying length constructs with the GLUT4 promoter and luciferase reporter gene, specific regions were identified as necessary for GLUT4 expression. Initially, the −1000 to −442 region was identified as necessary for normal GLUT4 expression. Further studies showed that the region from −522 to −420 contains an E-box and an MEF2 binding site. Mutations of the MEF2 binding site resulted in a near complete loss of reporter gene expression.

Work with transcription factors MEF2 and the myogenic basic-helix-loop-helix (bHLH) MyoD suggests that they function together to regulate the phenotype of skeletal muscle (^16,29). MEF2 is a transcription factor with many functions that include embryogenic development in the brain, heart, and skeletal muscle (¹⁸). There are four isoforms of the MEF2 protein: A, B, C, and D. The two that appear to be active in skeletal muscle are A and C (³¹). MEF2 contains a known nuclear localization sequence encompassing amino acids 472–507 in the primary sequence (³⁸). This indicates MEF2 can be activated to translocate to the nucleus. More recent work indicates that MEF2 co-localizes to the nucleus with histone deacetylase 4 (HDAC4) (¹). Some members of the HDAC family are MEF2 inhibitors; HDAC4 is one of these.

While studying the necessity of a functional MEF2 binding site, the results indicated that MEF2 was necessary but not sufficient for normal GLUT4 expression (³¹). In a follow-up to their previous work, Olson’s lab identified a 30-bp regulatory element they named Domain I, located upstream of the MEF2 binding site at −742 to −712 bp upstream from the initiation site (²³). They cloned the protein that binds to Domain I and named it GLUT4 enhancer factor (GEF) (²³). GEF was cloned and found to consist of an 1100-bp gene that encodes a 28-kDa peptide. GEF binds specifically to Domain I and an unlabeled wildtype oligonucleotide successfully competes for GEF. The cDNA sequence also contains a nuclear localization sequence, indicating that it is localized to the nucleus. Deletion or mutation of the Domain I binding site results in a decrease in gene expression or inhibits complex formation, respectively (²³). Therefore, both MEF2 and GEF protein binding are necessary for normal GLUT4 mRNA expression (²³).

Transgenic mice with deletions of varying portions of the human GLUT4 promoter resulted in the identification of an ERE within the promoter region. In this study, several lines of mice expressing the CAT reporter gene, driven by varying lengths of the human GLUT4 promoter, were exercised to determine the effects on GLUT4 gene transcription (¹⁵). CAT response was significant when the transgene was driven by at least −895 bp of the promoter (¹⁵). When the −730 bp promoter was tested, the exercise response was lost. These results were similar to those of Olson and Pessin (²²) with STZ diabetes and suggest that GEF is required for the exercise response.

Other studies in transgenic mice also indicate that there are exercise response elements between bases −1000 and −442 in the mouse GLUT4 promoter that regulate gene expression (^32,33). Together, these exercise studies suggest that MEF2 and GEF are key transcription factors mediating the exercise response of GLUT4. The important elements of the human GLUT4 promoter are shown in Figure 1.

SIGNALS REGULATING GLUT4 EXPRESSION

Exercise increases GLUT4 expression and transcription factors MEF2 and GEF are likely involved. How the signal is relayed from the contracting myocyte to regulatory pathways of gene transcription is of great interest. Our hypothesis is that changes in the energy state of the cell or fluctuations in Ca²⁺ levels is/are responsible for regulation of GLUT4 gene transcription during exercise.

Muscle contraction increases intracellular Ca²⁺ concentrations, and this results in the activation of calcium-dependant signaling pathways. The calcium, calmodulin-dependent phosphatase, calcineurin, promotes gene expression in slow oxidative fibers (¹⁷), and results from studies in the Williams laboratory suggest that MEF2 mediates the effects of calcineurin gene regulation (³⁵). In a study using mice with an MEF2-dependent reporter gene, forced expression of calcineurin or nerve stimulation upregulated the MEF2-dependant reporter gene (³⁵). Further study of the relationship between calcineurin and MEF2 showed that calcineurin dephosphorylates MEF2, increasing the transcriptional potency of MEF2 (³⁶). Because MEF2 is required for the exercise response of GLUT4, calcineurin would seem to be a viable candidate as a signaling molecule.

Another candidate signaling protein to mediate the calcium effect during exercise is calcium-calmodulin activated kinase (CAM kinase). Ojuka et al. (²¹) reported that caffeine treatment of L6 myocytes caused a fourfold increase in cytosolic Ca²⁺ and an increase in GLUT4 protein and mRNA. When treated with dantrolene, a Ca²⁺ channel blocker, in combination with caffeine the Ca²⁺ response was blunted. Removal of dantrolene from the medium resulted in a fourfold increase in cytosolic Ca²⁺. Caffeine treatment induced a nearly twofold increase in MEF2A and MEF2D protein levels. KN93, a CAMK inhibitor, prevented the caffeine-induced increases in MEF2A, MEF2D, and GLUT4 proteins in myocytes. These results suggest that Ca²⁺ may regulate GLUT4 expression during exercise by increasing the activity of CAMK, which acts to increase MEF2 and GLUT4 transcription.

The factors regulating GLUT4 expression have also focused on changes in metabolite concentrations and changes in the energy charge of the cell. Skeletal muscle contraction results in an increase in the AMP/ATP ratio and in a decrease in the creatine phosphate/creatine levels. To study the role of energy charge on GLUT4 expression, Yaspelkis et al. (³⁷) exposed gastrocnemius muscle to low frequency electrical stimulation from 14 to 28 d. The energy charge of the muscle was decreased and expression of GLUT4 was increased. ATP, and creatine phosphate were inversely related to GLUT4 protein concentration (¹⁴).

Another method to change energy charge is to lower creatine phosphate. When rats were fed β-guanidinopropionic acid (β -GPA), a creatine analog, creatine and phosphocreatine were depleted in skeletal muscle, whereas GLUT4 and mitochondrial oxidative enzymes were increased (²⁶). These changes in high-energy phosphate concentrations and increased enzyme activity are similar to the changes seen during muscle contraction (²⁶).

The 5′-AMP-activated protein kinase (AMPK) has been on the center stage of muscle metabolism research for the past decade. AMPK is a heterotrimeric enzyme composed of the catalytic α subunit and the noncatalytic β and γ subunits (^30,34). Two α subunits have been identified (α₁ and α₂); expression and activity varies depending on the tissue (³⁰). The subunit weights are: α, 63 kDa; β, 40 kDa; and γ, 38 kDa (³⁰). This allows for different combinations of heterotrimeric AMPK enzymes. The α isoenzymes have different activity as well. In liver, the α₁ subunit accounts for 94% of the activity (³⁰). Strong homology exists between the catalytic cores of α₁ and α₂, but outside the core the identity is weaker (³⁰). The relative concentrations and activities of α₁ and α₂ vary, with α₁ being more abundant than α₂ in skeletal muscle, but α₂ having the higher activity (³⁰).

The change in the energy balance of the cell is the primary cause for AMPK activation (³⁴). Shifts in cell energy equilibrium regulate demand for glucose uptake. When the ATP to AMP ratio is decreased (i.e., the energy charge is low), the cell detects this, and there is an increase in demand for ATP production. The increase in AMP activates AMPK both directly and indirectly. AMP allosterically activates AMPK and it also activates AMPK kinase (³⁴), which then phosphorylates and activates AMPK. AMPK has been called the fuel gauge of the cell due to the apparent ability to sense low energy levels and activate or inhibit the appropriate molecules to reestablish the proper energy balance.

To study the effect of AMPK activation on long-term gene expression, rats were injected daily for 5 d with 5-aminoimidazole-4-carboxamide-1-β-_D-ribofuranoside (AICAR), an activator of AMPK. Previously, we and others have found that AICAR treated mice and rats have increased AMPK activity (⁹) and GLUT4 mRNA (^1,2,39). The 5-d treatment resulted in significant increases in GLUT4 protein in both gastrocnemius and epitrochlearis by 60 and 100%, respectively (⁹). In another 5-d study, AICAR treatment (3 h·d⁻¹) of L6 myotubes resulted in a 75% increase in GLUT4 content. These same cells also had a twofold increase in MEF2A and D protein levels (²¹). Treatment of L6 cells with AraA, an AMPK inhibitor, prevents an AICAR-induced increase in GLUT4 (²¹). These findings in isolated myocytes suggest the involvement of AMPK and MEF2 in GLUT4 regulation. However, results with AICAR and ARA must be considered with caution because these compounds are not specific for only AMPK.

GLUT4 gene transcription studies in transgenic mice showed that mRNA was increased with AICAR treatment in white and red quadriceps by 70 and 50%, respectively (³⁹). This in vivo treatment of transgenic mice with AICAR resulted in a significant increase in relative GLUT4 mRNA levels in the −895 bp but not the −730 bp promoter. These results are very similar to those with exercise and STZ diabetes and they suggest the involvement of MEF2 and GEF.

SUMMARY AND CONCLUSIONS

Transcription of the GLUT4 gene is transiently activated after an acute bout of exercise and GLUT4 protein can be increased as much as two- to threefold after a few days of repeated exercise bouts. Studies of the GLUT4 promoter have identified two sets of DNA sequences that are important for metabolic regulation and also for increased transcription of the gene in response to exercise. These DNA elements have been shown to bind the transcription factors myocyte enhancer factor 2 (MEF2) and GLUT4 enhancer factor (GEF). The mechanisms that activate these proteins remain important areas of research.

Signals that link muscle contraction to the activation of transcription factors (MEF2, GEF) involved in increased expression of GLUT4 during exercise is another area needing further research. Two signals that show promise are changes in the energy charge (acting through AMP activated kinase [AMPK]) and changes in intracellular calcium (acting through calcineurin [a calcium-calmodulin activated phosphatase] and calcium-calmodulin activated kinase [CAMK]). There is good evidence that both increased AMPK activity and increased CAMK activity cause increased transcription of the GLUT4 gene. It remains to be demonstrated that exercise is acting through one or both of these mechanisms. Figure 2 shows our hypothesis as to the mechanisms that regulate GLUT4 transcription in response to exercise.

This work was supported by a grant from the National Institutes of Health (R01 DK38416).