A Peroxovanadium Compound Stimulates Muscle Glucose Transport as Powerfully as Insulin and Contractions Combined

bpV(phen) was obtained from Calbiochem-Novabiochem. Pork insulin was purchased from Eli Lilly. 2-deoxy-d-[1,2-³H]glucose (2DG) was obtained from American Radiolabeled Chemicals, whereas donkey anti-rabbit [¹²⁵I]IgG and [U-¹⁴C]mannitol were from DuPont-New England Nuclear. Adenine-9-β-d-arabinofuranoside hydrate (ara-A) was purchased from ICN Biomedicals. The anti-phosphotyrosine, anti-PI 3-K, and anti-rat COOH-terminal IRS-1 antibodies were purchased from Upstate Biotechnology. Horseradish peroxidase-conjugated donkey anti-mouse IgG was purchased from Jackson ImmunoResearch Laboratories. Enhanced chemiluminescence reagents were from Amersham. The 2-N-4 (1-azi-2,2,2,-trifluoroethyl)-ben-zoyl-1,3-bis-(d-mannose-4-yloxy)-2-propylamine (ATB-[2-³H]BMPA) was obtained from Toronto Research Chemicals. The bicinchoninic acid protein assay kit was purchased from Pierce Chemical. All other chemicals were obtained from Sigma.

All experimental procedures were approved by the Washington University animal studies committee. Male Wistar rats (100–140 g) were obtained from Charles Rivers Laboratories and given rat chow and water ad libitum. After an overnight fast, animals were anesthetized with an intraperitoneal injection of pentobarbital sodium (5 mg/100 g body wt), and the epitrochlearis and soleus muscles were removed (29). The soleus muscles were split into thin strips before incubation (30).

Muscles were incubated in a shaking incubator at 35°C for 30 min in 2 ml of Krebs-Henseleit bicarbonate (KHB) buffer containing 0.1% radioimmunoassay-grade BSA, 8 mmol/l glucose, and 32 mmol/l mannitol. In some experiments, one of the following inhibitors (wortmannin, LY294002, sodium dantrolene, or ara-A) was added to the incubation medium. Flasks were gassed continuously with 95% O₂-5% CO₂. After the initial incubation, muscles were transferred to 2 ml KHB buffer with 0.1% BSA, 8 mmol/l glucose, and 32 mmol/l mannitol with or without insulin and/or bpV(phen), and then they were incubated for 60 min at 35°C. If an inhibitor was present during the initial incubation, it was also present during the incubation with insulin or bpV(phen).

After incubation in the presence or absence of 0.1 mmol/l bpV(phen) or 2 mU/ml insulin for 60 min, some of the muscles were stimulated to contract before measurement of glucose transport activity. A total of 10 tetanic contractions were produced by stimulating the muscle at 100 Hz for 10 s at a rate of 1 contraction/min for 10 min as described previously (31).

Muscles were transferred to 1.0 ml KHB containing 0.1% BSA, 4 mmol/l 2DG (1.5 μCi/ml), 36 mmol/l [U-¹⁴C]mannitol (0.3 μCi/ml), and the same additions present during the previous incubations. The flasks were incubated with shaking at 30°C for 20 min with a gas phase of 95% O₂-5% CO₂. The assay was terminated by blotting and then clamp-freezing the muscles. Extracellular space and intracellular 2DG concentration were determined as previously described (32,33).

The quantity of GLUT4 at the cell surface was assessed in epitrochlearis muscles using the ATB-[2-³H]BMPA exofacial photoaffinity labeling technique, as described previously (34).

Muscles were homogenized in perchloric acid, and glycogen content was assessed by the amylogucosidase method (35).

Epitrochlearis muscles were incubated for 30 min at 35°C in 2 ml KHB buffer containing 0.1% BSA, 8 mmol/l glucose, and 32 mmol/l mannitol, and then they were transferred to fresh medium of the same composition with or without insulin (50 mU/ml) or bpV(phen) (0.1 mmol/l) and incubated for 5 or 60 min at 35°C. After incubation, muscles were frozen and stored at −80°C. Preparation of muscle extracts, immunoprecipitation with anti-rat IRS-1 antibody, and measurement of IRS-1 tyrosine phosphorylation and IRS-1-bound PI 3-K by immunoblotting were performed as described previously (36). Phosphorylation of PKB at the Ser473 site was quantified as previously described (37).

Data are presented as the means ± SE. Multiple group comparisons were analyzed using ANOVA. A Newman-Keuls post hoc test was performed to assess differences when indicated by the ANOVA.

As shown in Fig. 1A, bpV(phen) potently stimulates glucose transport activity in rat epitrochlearis. A maximal stimulatory effect was obtained at ∼0.1 mmol/l bpV(phen). The rate of 2DG transport increased for ∼60 min and then leveled off in muscles exposed to 0.1 mmol/l bpV(phen) (Fig. 1B).

We tested the hypothesis that the effects of low concentrations of insulin and bpV(phen) on glucose transport are synergistic. This hypothesis turned out to be wrong, because the actual rates of glucose transport induced by 5 μmol/l bpV(phen) together with 10 or 20 μU/ml insulin were essentially the same as the values calculated by adding the transport rates induced by insulin alone and bpV(phen) alone (data not shown).

The effects of 100 μmol/l bpV(phen) on glucose transport activity (Fig. 2A) and GLUT4 translocation to the cell surface (Fig. 2B) were greater than those of either a maximally effective insulin stimulus or a maximally effective exercise stimulus. As in previous studies (21,22), the maximal effects of insulin and exercise on glucose transport and GLUT4 translocation to the cell surface were additive (Fig. 2). There was no significant difference between the effect of bpV(phen) and the combined maximal effects of insulin and exercise (Fig. 2). When muscles were incubated with maximally effective concentrations of both insulin and bpV(phen), the increase in 2DG transport was not significantly different from that induced by bpV(phen) alone (Fig. 2). This finding is in keeping with the evidence that peroxyvanadate compounds stimulate glucose transport by increasing tyrosine phosphorylation of the IRSs, thus activating insulin signaling.

Muscle contractions stimulate muscle glucose transport by a mechanism that is independent of the insulin signaling pathway, and the maximal effects of contractions and insulin on glucose transport and GLUT4 translocation to the cell surface are additive. Therefore, because bpV(phen) appears to stimulate glucose transport via the insulin signaling pathway, we examined the possibility that the effects of bpV(phen) and muscle contractions are also additive. However, the combined effects of bpV(phen) and contractions on both 2DG transport and GLUT4 translocation to the cell surface were no greater than that of bpV(phen) alone (Fig. 2).

The increases in glucose transport induced by insulin and by contractions are similar in the epitrochlearis (Fig. 2A). In contrast, the responsiveness to insulin is much greater than the responsiveness to contractions in the soleus (Fig. 3). This phenomenon may be explained by the presence of a large pool of “insulin responsive GLUT4” and a small pool of “contraction responsive GLUT4” in the soleus (31). As a consequence, although the increase in glucose transport induced by bpV(phen) was greater than that induced by a maximally effective insulin stimulus, the relative difference in magnitude between insulin-stimulated and bpV(phen)-stimulated glucose transport was much smaller in the soleus than in epitrochlearis (Fig. 3). We therefore concentrated on the effect of bpV(phen) in the epitrochlearis in this study.

The findings that bpV(phen) stimulates glucose transport to the same extent as exercise plus insulin, and that the effect of contractions and bpV(phen) are not additive, raised the possibility that bpV(phen) stimulates both the contraction-activated and the insulin-activated pathways. There is evidence that AMP kinase (AMPK) plays a role in the stimulation of glucose transport by exercise (38–40). ara-A is taken up by skeletal muscle and converted to araATP (9-β-D arabinofuranoside ATP), a competitive inhibitor of AMPK (41), which partially inhibits contraction-stimulated glucose transport (42). We found that 2 mmol/l ara-A had no inhibitory effect on the stimulation of glucose transport by bpV(phen) or insulin, but it inhibited contraction-stimulated glucose transport by ∼60% (Fig. 4). There is also evidence that the increase in cytosolic Ca²⁺ during excitation-contraction coupling is involved in the stimulation of glucose transport by exercise (22,43,44). The stimulation of glucose transport by agents that release Ca²⁺ from the sarcoplasmic reticulum can be blocked with 5 μmol/l dantrolene (43). However, dantrolene had no inhibitory effect on the stimulation of glucose transport by bpV(phen) (3.97 ± 0.38 μmol · ml⁻¹ · 20 min⁻¹ with dantrolene vs. 4.08 ± 0.24 μmol · ml⁻¹ · 20 min⁻¹ without dantrolene [means ± SE for 6 muscles with and 12 muscles without dantrolene]).

We also evaluated the possibility that bpV(phen) stimulates glucose transport by increasing Ca²⁺ influx from the medium. The absence of Ca²⁺ in the medium had no effect on the stimulation of glucose transport by bpV(phen) (3.93 ± 0.44 μmol · ml⁻¹ · 20 min⁻¹ for bpV(phen) without Ca²⁺ vs. 4.07 ± 0.35 μmol · ml⁻¹ · 20 min⁻¹ for bpV(phen) with Ca²⁺ [means ± SE for 4 muscles per group]). If bpV(phen) induced an increase in cytosolic Ca²⁺, this would, by resulting in phosphorylase activation, bring about a decrease in glycogen. However, muscles incubated with bpV(phen) showed increases in glycogen similar to those induced by insulin [glycogen (in μmol/g wet wt): basal 11.4 ± 0.8; insulin 15.9 ± 1.2; bpV(phen) 15.1 ± 1.0]. Taken together, these results provide evidence that bpV(phen) does not activate the pathway by which contractions stimulate glucose transport.

A number of studies have shown that the PI 3-K inhibitor wortmannin, at a concentration of 1 μmol/l, completely blocks the stimulation of glucose transport by insulin in muscle without affecting the increase in glucose transport induced by contractions (21,25,45). As in these previous studies, 1 μmol/l wortmannin prevented the stimulation of glucose transport by insulin, but it had no effect on contraction-stimulated glucose transport (Fig. 4). Wortmannin and another PI 3-K inhibitor, LY294002, completely blocked the stimulation of glucose transport by bpV(phen) (Fig. 4). Wortmannin also blocked bpV(phen)-induced GLUT4 translocation (Fig. 2B). These findings provide strong evidence that the stimulation of glucose transport by bpV(phen) is mediated entirely via the insulin-signaling pathway.

The finding that bpV(phen) induces a greater increase in glucose transport than insulin, despite, apparently, stimulating glucose transport entirely via the insulin-signaling pathway, led us to examine the possibility that bpV(phen) is more potent than insulin as an activator of this pathway. As shown in Fig. 5, treatment of muscles with bpV(phen) for 5 min resulted in a minimal increase in IRS-1 tyrosine phosphorylation. However, after 60 min of incubation with bpV(phen), IRS-1 tyrosine phosphorylation was increased ∼11-fold. In contrast, the early ∼5-fold increase in tyrosine phosphorylation of IRS-1 by insulin had partially reversed by 60 min, at which time IRS-1 tyrosine phosphorylation in the bpV(phen)-treated muscles was ∼3.5-fold greater than in the insulin-treated muscles. A similar pattern was seen with IRS-1 binding of PI 3-K (Fig. 6). After 60 min of incubation, the amount of PI 3-K bound to IRS-1 was ∼2.7-fold greater in muscles treated with bpV(phen) than in the insulin-treated muscles. The increase in PKB phosphorylation on serine 473 after 60 min of treatment was also greater in response to bpV(phen) than in response to insulin (Fig. 7).

Insulin induces a graded increase in muscle glucose transport which, in our epitrochlearis muscle preparation, varies over an approximately fivefold range when insulin concentration is increased from 20 μU/ml to 1 mU/ml (46). Similarly, the increase in muscle glucose transport induced by bpV(phen) increases with increasing concentration before leveling off at a bpV(phen) concentration of ∼100 μmol/l. However, as our results show, bpV(phen) can stimulate glucose transport to about twice as great an extent as a maximally effective insulin concentration in epitrochlearis muscle. The greater stimulation of glucose transport by bpV(phen) is mediated by translocation of more GLUT4 to the cell surface. Glucose transport can also be stimulated by exercise in muscle, and the maximal effects of contractions and insulin on glucose transport are additive (22). Our finding that the maximal effect of bpV(phen) is similar to that of insulin and contractions together raised the possibility that the signal generated by bpV(phen) causes translocation to the plasma membrane of all of those GLUT4 vesicles that normally respond to contractions or to insulin. In this context, it seemed likely that bpV(phen) might stimulate both the insulin- and the contraction-activated pathways.

There is evidence that the increase in cytosolic Ca²⁺ that triggers muscle contraction is involved in mediating the contractile activity-induced increase in glucose transport (22). It has been reported that vanadate may induce an increase in cytosolic Ca²⁺ concentration in various cell types, suggesting that bpV(phen) might activate the contraction-stimulated pathway (13). However, our results show that the contraction-activated pathway is not involved, and that the greater effect of bpV(phen) on glucose transport is mediated entirely via the IRS-PI 3-K signaling pathway. The findings that support this conclusion are that 1) preventing Ca²⁺ release from the sarcoplasmic reticulum with dantrolene and extracellular Ca²⁺ entry by using Ca²⁺-free medium has no effect on bpV(phen)-stimulated glucose transport; 2) bpV(phen) induces greater IRS-1 phosphorylation and PI 3-K binding than insulin; 3) ara-A, which partially inhibits contraction-stimulated transport, does not inhibit bpV(phen)-stimulated glucose transport; and, most importantly, 4) the stimulation of glucose transport by bpV(phen) is completely blocked by the PI 3-K inhibitors wortmannin and LY294002. It is well documented that wortmannin does not inhibit contraction-stimulated glucose transport (21,25,45) (Fig. 4).

The probable explanation for the finding that insulin is unable to maximally activate the IRS-PI 3-K signaling pathway is that tyrosine phosphorylation of the IRSs by IR tyrosine kinase is countered by dephosphorylation by PTPs. As a consequence, the maximal steady-state level of IRS phosphorylation induced by insulin represents a balance between the actions of IR tyrosine kinase and PTPs. In this context, it is of interest that the PTPs, leukocyte antigen-related PTP, and PTP-1B act as negative regulators of insulin signaling and may play a role in some forms of insulin resistance (47,48). It has, therefore, been suggested that PTPs are potential targets for the treatment of type 2 diabetes (47,48). In contrast to the effect of insulin, the IRS tyrosine phosphorylation mediated by the inhibition of PTPs by bpV(phen) is unopposed, and although it occurs more slowly, IRS-1 tyrosine phosphorylation attains a level ∼3.5-fold greater than the steady-state level of IRS-1 tyrosine phosphorylation induced by a maximally effective insulin concentration. These results provide support for the suggestion that inhibitors of PTPs, such as vanadate and peroxovanadates, might be useful in the treatment of insulin resistance (13), if it can be shown that they do not have toxic side effects at doses that are effective in lowering blood glucose.

The finding that the maximal effects of insulin and exercise on GLUT4 translocation and glucose transport are additive has raised the possibility that there may be two pools of GLUT4-containing vesicles in muscle, and that those in one pool are translocated in response to insulin, whereas the others respond to the signal generated by exercise (1,27). The present results suggest that all of those GLUT4 vesicles in muscle that are normally translocated to the plasma membrane in response to the combined effects of insulin and exercise can also be translocated by the more powerful activation of the IRS-PI 3-K signaling pathway by bpV(phen). The evidence for this is that the increase in GLUT4 at the cell surface induced by bpV(phen) is as great as that induced by the combined maximal effects of insulin and contractions, and the response to contractions plus bpV(phen) is no greater than that to bpV(phen) alone. Viewed in this context, our results argue against the concept that the GLUT4 vesicles that are normally translocated by exercise are unable to respond to the signal mediated by the insulin-signaling pathway. The alternative possibility that the additional GLUT4 recruited by bpV(phen) could come from a large pool of “insulin-responsive” GLUT4 vesicles, some of which are not recruited by a maximally effective insulin concentration, seems less likely in view of our finding that the effects of bpV(phen) and contractions together are not greater than that of bpV(phen) alone.

How the larger IRS-PI 3-K pathway signal generated by bpV(phen) brings a greater number of GLUT4s to the cell surface is part of the larger question of how insulin induces a graded response, because the greater effect of bpV(phen) can be thought of as an extension of the insulin dose-response curve. Not enough is currently known regarding how insulin mediates GLUT4 translocation to warrant detailed speculation regarding the mechanism underlying the graded response. However, the existence of GLUT4 vesicles that do not respond to a maximally effective insulin concentration but are instead translocated in response to the more powerful activation of the IRS-PI 3-K signaling pathway by bpV(phen) suggests that the signal generated by insulin is weaker in some regions of the cell than in others. This hypothesis also provides an explanation for the graded response to insulin. It seems possible in this context that the GLUT4 vesicles that normally respond to exercise but not to insulin are located in regions in which the insulin signal is too attenuated to cause translocation.

In summary, the present results show that a maximally effective concentration (100 μmol/l) of bpV(phen) can stimulate GLUT4 translocation and glucose transport in muscle to as great an extent as maximally effective insulin and contraction stimuli in combination. This effect of bpV(phen) is mediated via the IRS-PI 3-K signaling pathway, which is activated more powerfully by bpV(phen) than by insulin.

This work was supported by National Institutes of Health Grant DK18986 and Clinical Nutrition Research Unit Grant DK56341. L.A.N. was supported initially by an American Diabetes Association Mentor-Based Postdoctoral Fellowship and subsequently by Institutional National Research Service Award AG00078.

We thank Tim Meyer, Helen Host, and May Chen for excellent technical assistance and Victoria Reckamp for expert assistance in preparation of the manuscript.