INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

SKELETAL MUSCLE glucose uptake in vivo is determined by the glucose transport activity of the sarcolemma and the transsarcolemmal glucose gradient (TSGG). Considerable work has been done to study the role and regulation of glucose transport into muscle (47). Far less attention has been given to the role of the TSGG and the factors that maintain it. The TSGG is determined by the delivery of glucose to the outer surface of the sarcolemma and removal of glucose from the inner surface. Muscle glucose delivery is a function of the muscle blood flow and diffusion distance, and removal of glucose from the inner sarcolemma surface is determined by diffusion and glucose phosphorylation. Clearly, measurement of the TSGG would give valuable insight into the control of glucose uptake under various physiological and pathophysiological conditions. The glucose concentrations on the outer and inner surfaces of the sarcolemma, however, are impossible to obtain directly. Although total tissue glucose can be measured accurately in biopsied or excised muscle, it is impossible to translate these measurements into a number representing the TSGG. This is because it is very difficult to distinguish interstitial from intracellular glucose. Furthermore, even if one were able to make the distinction between interstitial and intracellular glucose directly, spatial and physical barriers within extracellular and intracellular compartments make the determination of glucose at the membrane surfaces impossible. Because the TSGG cannot be directly determined, important questions regarding control of glucose uptake remain unanswered.

The aim of these studies was to define the roles of glucose transport and the TSGG in the regulation of basal and insulin-stimulated glucose uptake in vivo. For these purposes, a novel method was used for simultaneously assessing indexes of the TSGG and muscle glucose metabolism. This method applies infusions of isotopic glucose analogs [3-O-[³H]methylglucose (3-O-[³H]MG), [U-¹⁴C]mannitol ([U-¹⁴C]MN), and 2-deoxy-[³H]glucose ([2-³H]DG)] and principles of glucose countertransport in a chronically catheterized, awake rat model. Changes in the TSGG can be obtained because the ratio of intracellular to extracellular 3-O-[³H]MG concentration at equilibrium is a functional consequence of the TSGG (6, 12, 41, 42). In addition, an index of muscle glucose metabolism is calculated from muscle phosphorylated [2-³H]DG ([2-³H]DGP) and plasma [2-³H]DG. These measurements can distinguish the roles of glucose transport and the TSGG in control of muscle glucose uptake.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animal maintenance and surgical procedures. Male Sprague-Dawley rats (Sasco, Omaha, NE) were individually housed at 23°C on a 0600-1800 light cycle and allowed free access to water and a predetermined weight of chow (65% carbohydrate, 11% fat, 24% protein). The rats were housed under these conditions for ~1 wk, by which time their weights had reached 250-300 g. During this period, the food conversion index (FCI; ratio of weight gained to food consumed) for each rat was calculated. Animals were anesthetized with a 50:5:1 mixture of ketamine (Avoco, Fort Dodge, IA), Rompun (Avoco), and acepromazine (Haver, Shawnee, KS). The left common carotid artery and the right jugular vein were catheterized (PE50, Clay Adams, Parsippany, NJ). Catheters were tunneled under the skin, exteriorized, secured at the back of the neck, filled (~60 µl) with a 3:1 mix of glycerol and heparin, and flame sealed. Immediately postsurgery, each animal received 40,000 U of penicillin G (Marsam, Cherry Hills, NJ) and 5 ml of sterile saline subcutaneously. All procedures were preapproved by the Vanderbilt University Animal Care Committee and followed the National Research Council Guide for the Care and Use of Laboratory Animals. In the postsurgery period, animal weights and food intake were monitored daily and only animals in which presurgery weight and FCI were restored were used for experiments. On the day of the experiment the catheters were aspirated, flushed with a saline-heparin solution, and connected to silcone rubber tubing for sampling.

Experimental procedures. Food was taken away from rats ~5 h before the beginning of a study. At t = 100 min an infusion of saline (n = 6) or insulin (Novo Nordisk, Princeton, NJ) at a rate of 0.6 (n = 5), 1.0 (n = 7), or 4.0 (n = 7) mU · kg¹ · min¹ was begun. Also at t = 100 min, primed infusions of [U-¹⁴C]MN (3.5 µCi primer and 60 nCi/min infusion) and 3-O-[³H]MG (25 µCi primer and 150 nCi/min infusion) were begun. 3-O-[³H]MG is transported in the cell but is not further metabolized, whereas [U-¹⁴C]MN is restricted to the extracellular space and thus serves as a marker of the extracellular space. In another protocol, rats were studied for a longer duration (340 min) to assess muscle isotope equilibration (i.e., achievement of a steady state). Beginning at t = 300 min, 3-O-[³H]MG and [U-¹⁴C]MN were primed and infused as described above. At t = 0 min a constant-rate infusion of [2-³H]DG (900 nCi/min) was started and continued to the end of the study. [2-³H]DG is transported into the cell and phosphorylated.

Processing of blood and muscle samples. Plasma glucose concentrations were measured by the glucose oxidase method using an automated glucose analyzer (Beckman Instruments, Fullerton, CA), and immunoreactive insulin was measured using a double antibody method (40). Total plasma radioactivity ([U-¹⁴C]MN, 3-O-[³H]MG, [2-³H]DG) was determined after deproteinization with barium hydroxide [Ba(OH)₂, 0.3 N] and zinc sulfate (ZnSO₄, 0.3 N) and centrifugation. Radioactivity was determined in 10 ml of Ecolite+ scintillation fluid (ICN, Irvine, CA) by dual-labeled liquid scintillation counting (Beckman LS 5000TD, Beckman Instruments). To distinguish plasma 3-O-[³H]MG and [2-³H]DG radioactivity, plasma samples were treated with Ba(OH)₂ and ZnSO₄, incubated with a solution containing (as final concentrations) 2.5 mg/ml yeast hexokinase (21 U/mg solid; Sigma, St. Louis, MO), 100 mM KCl, 40 mM tris(hydroxymethyl)aminomethane-Cl, 20 mM MgCl₂, and 4 mM EDTA (pH 8.1), incubated at room temperature for 30 min, and then retreated with Ba(OH)₂ and ZnSO₄. Tests in our laboratory have shown that the yeast hexokinase phosphorylates >99% of [2-³H]DG to [2-³H]DGP, and >98% of the [2-³H]DGP is removed by the final Ba(OH)₂ and ZnSO₄ treatment. Muscle samples were homogenized in 0.5% perchloric acid (PCA), centrifuged, and neutralized with 10 N KOH. One aliquot of homogenate was counted without further treatment, as described for plasma samples, to yield total muscle counts ([U-¹⁴C]MN, 3-O-[³H]MG, [2-³H]DG, [2-³H]DGP, and [2-³H]DGP in glycogen). A second aliquot of homogenate was treated with Ba(OH)₂ and ZnSO₄ to remove free [2-³H]DGP and [2-³H]DGP incorporated into glycogen and then counted to yield [U-¹⁴C]MN, 3-O-[³H]MG, and [2-³H]DG radioactivity. A third aliquot of homogenate was incubated with yeast hexokinase (as described for plasma samples), treated with Ba(OH)₂ and ZnSO₄ to remove [2-³H]DGP, and then counted to give [U-¹⁴C]MN and 3-O-[³H]MG radioactivity. Because [U-¹⁴C]MN is unaffected by analytical methods [i.e., hexokinase treatment and Ba(OH)₂ and ZnSO₄], radioactivity in the ¹⁴C counting window in treated samples was normalized to its radioactivity in untreated samples to provide an internal control for each plasma and muscle sample. Tissue glucose was measured after deproteinization with PCA by an enzymatic method (33).

Calculations. The distribution of mannitol between tissue and extracellular spaces was used to calculate the fraction of extracellular to total water space in biopsies by the equation

(1)

where F_e is the fraction of the tissue water that is extracellular and the subscripts t and e refer to tissue and extracellular compartments. [U-¹⁴C]MN_e concentrations were assumed to equal plasma [U-¹⁴C]MN, since mannitol is not extracted by muscle. Intracellular and extracellular water spaces were used to calculate intracellular substrate concentrations.

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

Statistical analyses. Statistical comparisons were made between the basal group and each insulin-infused group with Student's t-test using Statview (Abacus, Berkeley, CA) and a Macintosh computer. Differences were considered statistically significant at P < 0.05. Data are expressed as means ± SE. Statistical differences are presented in the legends to Figs. 1-6 and Tables 1 and 2.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Arterial plasma insulin, arterial plasma glucose, glucose infusion rates, and arterial plasma 3-O-[³H]MG. Plasma insulin concentrations (Table 1) ~1.2-, 1.8-, and 4.9-fold above values obtained during saline infusion were achieved with 0.6, 1.0, and 4.0 mU · kg¹ · min¹ insulin infusion rates, respectively. Arterial insulin concentrations at the two higher insulin doses were significantly greater than values obtained during saline infusion alone. There was no significant difference in plasma glucose among any of the groups (Table 1). The glucose infusion rate required to maintain euglycemia was increased in an insulin dose-dependent manner (Table 1).

1

1

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Arterial plasma 3-O-[3H]methylglucose (3-O-[3H]MG) concentrations were determined at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data represent means ± SE. 3-O-[3H]MG concentrations were significantly decreased by all insulin infusion rates in relation to saline (P < 0.05); dpm, disintegrations/min.

Skeletal muscle glucose concentrations. Skeletal muscle glucose concentration, which is the sum of tissue glucose inside and outside the cell, fell insignificantly compared with basal at the low insulin dose (0.6 mU · kg¹ · min¹) in the soleus. Soleus glucose concentration was equal to basal at the two higher insulin doses (Table 2). In contrast, glucose concentration fell by 35-45% in muscles that have a high composition of fast-twitch fibers (vastus lateralis and gastrocnemius). Because the change in muscle glucose will reflect the balance between glucose utilization and glucose delivery, these data suggest that the muscles comprised of fast-twitch fibers have a greater deficit in muscle glucose delivery compared with their ability to use glucose.

Skeletal muscle extracellular water space. The fraction of H₂O that was extracellular in the soleus was 0.33 ± 0.02, 0.35 ± 0.07, 0.41 ± 0.02, and 0.38 ± 0.02 at insulin infusions of 0.0, 0.6, 1.0, and 4.0 mU · kg¹ · min¹, respectively. The fraction of tissue H₂O that is extracellular was increased significantly above basal at the 1.0 mU · kg¹ · min¹ insulin infusion rate (P < 0.05).

Skeletal muscle S_i/S_o. S_i/S_o in the soleus was 0.45 ± 0.10 after a 140-min infusion of isotopes alone. Extending the infusion of [U-¹⁴C]MN and 3-O-[³H]MG by 200 min did not lead to a further increase in S_i/S_o (0.53 ± 0.03), indicating that 3-O-[³H]MG had reached equilibrium in the soleus. A clear steady-state S_i/S_o was not apparent in other muscles, and ratios are not presented. The S_i/S_o response to increasing insulin in the soleus is shown in Fig. 2. S_i/S_o was increased above values seen with saline infusion at the highest insulin dose (P < 0.005).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Ratio of intracellular to extracellular 3-O-[3H]MG concentration (Si/So) was calculated in soleus at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data represent means ± SE. Si/So was increased at highest insulin dose in soleus (P < 0.005).

Skeletal muscle [G]_om, [G]_im, and TSGG. [G]_om, [G]_im, and TSGG were calculated under two extreme theoretical conditions, which are defined in Calculations. In the first condition, [G]_im is contained in such a small fraction of the intracellular H₂O that, regardless of whether [G]_im is relatively high or low, it contributes negligibly to the tissue glucose mass (the symbol  represents terms calculated in this condition). In the second condition, the opposite is assumed. Glucose was assumed to be distributed evenly throughout intracellular H₂O (designated with the symbol ). Within the bounds of these extremes lay the true values of [G]_om, [G]_im, and TSGG. [G]_om was virtually unchanged with increasing insulin. [G]_om decreased gradually with increasing insulin concentration (Fig. 3). [G]_im was not significantly different from zero under basal conditions and at insulin doses of 0.6 and 1.0 mU · kg¹ · min¹ but rose to 2.4 ± 0.6 at an insulin dose of 4.0 mU · kg¹ · min¹ (P < 0.05; Fig. 3). [G]_im was also not significantly different from zero in the basal state and at insulin infusion rates of 0.6 and 1.0 mU · kg¹ · min¹. At the higher insulin dose, however, it rose significantly, reaching 0.9 ± 0.2 mM at an insulin infusion of 4.0 mU · kg¹ · min¹ (P < 0.005).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Outer sarcolemmal glucose concentrations ([G]om and [G]om) (top) and inner sarcolemmal glucose concentrations ([G]im and [G]im) (bottom) were calculated in soleus at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data represent range bounded by means ± SE for [G]om and [G]om and [G]im and [G]im. [G]om decreased gradually with increasing insulin concentration. [G]im and [G]im were not significantly different from a glucose concentration of 0 at basal insulin and at insulin doses of 0.6 and 1.0 mU · kg1 · min1 but rose significantly at the insulin dose of 4.0 mU · kg1 · min1 (P < 0.05-0.005).

1

1

1

1

1

1

1

1

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Transsarcolemmal glucose gradients (TSGG and TSGG) were calculated in soleus at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data represent range bounded by means ± SE for TSGG and TSGG. TSGG was significantly reduced compared with basal at 4.0 mU · kg1 · min1 insulin infusion (P < 0.02). TSGG was reduced compared with basal at insulin infusions of 1.0 and 4.0 mU · kg1 · min1 (P < 0.05-0.01). TSGG and TSGG were not significantly different from 0 at highest insulin dose.

Skeletal muscle R_g. The R_g response in the soleus is shown in Fig. 5. R_g was 13.3 ± 3.0, 36.3 ± 9.2, 31.8 ± 4.2, and 35.1 ± 2.1 µmol · 100 g¹ · min¹ at insulin infusions of 0.0, 0.6, 1.0, and 4.0 mU · kg¹ · min¹, respectively. The increment in soleus R_g at the 0.6 mU · kg¹ · min¹ insulin infusion occurred without a decrease in the TSGG. Comparison of Figs. 4 and 5 shows that the TSGG narrowed at the higher insulin infusion rates even after R_g had reached a maximum. These data indicate that glucose transport is not maximal and that those factors responsible for sustaining the glucose gradient (glucose delivery to muscle and/or glucose phosphorylation in muscle) have become rate limiting. The increase in [G]_im at the highest insulin dose suggests that glucose phosphorylation is one of those factors.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Glucose metabolic index (Rg) in soleus was determined at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data represent means ± SE. An insulin infusion increased Rg significantly above values seen with saline infusion alone (P < 0.05-0.001).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

A great number of studies have assessed the means by which insulin stimulates glucose transport, and much has been learned. Far fewer studies have assessed the means by which the downhill glucose gradient from outside to inside the muscle cell is maintained in the face of the marked increase in glucose transport that occurs during insulin stimulation. The results of this study show that a small physiological increase in insulin can stimulate muscle R_g by approximately threefold without significantly affecting the TSGG in vivo. This suggests that mechanisms of glucose delivery to soleus and glucose phosphorylation within the soleus keep pace with the increase in glucose transport at this concentration of insulin. If the increase in insulin-stimulated glucose transport is uncompensated for by parallel increases in muscle glucose delivery and/or intracellular glucose phosphorylation, the TSGG will be reduced. This is, in fact, what occurred at higher insulin concentrations. Stimulation of glucose transport exceeded the stimulation of mechanisms that maintain the TSGG, and the TSGG fell from 5 mM in the basal state to <1 mM at an insulin infusion rate of 4.0 mU · kg¹ · min¹. The fact that R_g is sustained despite a progressive fall in TSGG indicates that glucose transport continues to increase even after R_g has achieved a maximum. The TSGG at the high insulin infusion rate was not significantly different from zero (i.e., no gradient existed), suggesting that TSGG, and not glucose transport, had become rate limiting for muscle glucose uptake.

The decrease in TSGG is due, at least in part, to a limitation in glucose phosphorylation. Even our most conservative estimate showed an increase in [G]_im at the highest insulin concentration, suggesting that glucose phosphorylation is insufficient to maintain TSGG. This finding is consistent with the conclusions of several other studies that used different approaches (11, 15, 31, 65). It is less clear whether glucose availability to the outer surface of the muscle is also a limitation. [G]_om does not fall when it is calculated assuming [G]_im occupies a negligible volume, but falls to one-third of the basal value when it is calculated assuming that [G]_im exists throughout the entire intracellular space. It is probable that a deficit in glucose availability is more important in fast-twitch muscles, since the vastus lateralis and gastrocnemius both exhibit significant decreases in muscle glucose concentration in response to insulin infusions. The question of whether insulin may play a role in increasing the delivery of glucose to skeletal muscle is controversial. Although some investigators have reported an insulin-induced increase in limb blood flow (for review, see Ref. 1) and capillary perfusion (50), others have not observed these effects (e.g., 27, 44). This issue is further complicated by the results of one study in which insulin increased muscle blood flow but the areas of the muscle that received the increased flow were distinct from the areas in which the largest increases in glucose uptake occurred (49). Whether or not insulin has hemodynamic effects, the fall in total muscle glucose in type II muscles is evidence that the rate of glucose entry into the muscle as a whole must be less than the rate of glucose metabolism.

Three markedly different modeling approaches have been conducted using data from rats (15, 64) or humans (4). Each of these shows a greater insulin stimulation of 3-O-MG transport into the cell compared with transport out of the cell. This is exactly what is predicted from the increase in S_i/S_o in the present study. These models all give results consistent with the present study in that they predict a greater antagonism of 3-O-MG transport out of the cell in the presence of hyperinsulinemia in vivo. The principle of glucose countertransport has been used to estimate intracellular glucose in studies conducted in diverse model systems (6, 8, 12, 16, 38, 41, 42, 52). The linking of the transmembrane glucose distribution to 3-O-MG countertransport assumes that 1) glucose and 3-O-MG share the same transport system, 2) the reaction between carrier and sugar is rapid compared with carrier mobility, 3) the relative affinity of each sugar for the transport proteins is the same on the extracellular and intracellular sides of the plasma membrane, and 4) carrier mobility is independent of whether or not the transporter is bound to either sugar. These assumptions have been discussed in detail by Foley et al. (12). The validity of the first two assumptions has been repeatedly demonstrated (6, 7, 17, 18, 52, 59, 62). The third assumption is supported by the demonstration that the kinetic parameters for 3-O-MG are equal for entry and exit into adipocytes (56, 61), which use the same glucose transport proteins as muscle. Morever, this symmetry is independent of the method used to assess transport kinetics and whether insulin is present or absent (56). The fourth assumption is consistent with the lack of any marked asymmetry of the transport system in adipocytes, even when intracellular glucose and consequently the bound state of the inner aspect of the glucose transporter are less than the extracellular (18, 56).

Calculation of [G]_om, [G]_im, and TSGG requires knowledge of the interstitial [G] and the K_m of glucose transport. Boundaries for interstitial [G] were achieved by using two extreme theoretical conditions. The first condition is one in which [G]_im is confined to such a small portion of the intracellular H₂O that it contributes negligibly to the total tissue glucose measured and all the glucose is confined to the extracellular space (Eq. 7). The opposite case is one in which glucose is dispersed uniformly in intracellular H₂O so that [G]_im is the glucose concentration of the entire intracellular H₂O. The values for [G]_om that are obtained with these approaches give a representative interstitial glucose concentration for these conditions. At some cells, however, this value may be too high and in others it may be too low. Regardless, the relationship of [G]_om to [G]_im will be the same. It is also difficult to precisely know K_m, since a range of 2 to 5 mM for the glucose transporter has been reported (47). In the calculations described above a K_m of 4 mM was used, since it is within the range measured for glucose transport by GLUT-4 in vitro and it is comparable to the venous glucose at which the K_m occurs in vivo (10, 63, 67). The responses of [G]_im, [G]_om, and TSGG to insulin are similar regardless of the K_m used to calculate them. It is also apparent that the value used for K_m has a much greater influence on our calculated values when S_i/S_o is relatively small, as in the basal state (as shown in Fig. 6). At an S_i/S_o of 1 (which is not different from the value in soleus at the highest insulin infusion rate), K_m does not influence this calculation.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Influence of Michaelis-Menten constant (Km) on  values of [G]om (top), [G]im (middle), and TSGG (bottom) calculated in soleus at insulin infusion rates of 0.0 (n = 6), 0.6 (n = 5), 1.0 (n = 7), and 4.0 (n = 7) mU · kg1 · min1. Data are mean values. These data show that the same types of responses are evident regardless of Km used in calculations.

The measurement of intracellular 3-O-MG concentration could also conceivably be increased if its intracellular volume of distribution were expanded. One could speculate that insulin could increase the volume of 3-O-MG distribution by either 1) accelerating exchange with a pool that equilibrates slowly under basal conditions or 2) increasing permeability of an intracellular physical barrier. The first postulate is unlikely to explain the increase in intramuscular 3-O-MG, since 3-O-MG is already in equilibrium with the contiguous cell water. It is difficult to envision the second, since insulin has not been shown to increase the permeability of intracellular organelles to glucose. Earlier studies have demonstrated that insulin stimulation does not affect the intracellular volume of 3-O-MG distribution in adipocytes (18, 59) or the tissue distribution volume of 3-O-MG in skeletal muscle (42). Studies in isolated cells (38) and tissues (6, 42) clearly show that the insulin-induced increase in 3-O-[³H]MG inside vs. outside the cell is secondary to effects of insulin on transmembrane glucose distribution and are not due to a primary increase in the volume of the intracellular glucose compartment. The reliance of S_i/S_o on transmembrane glucose distribution is exemplified by the demonstration that, if glucose is removed entirely from the incubation medium, the rate constants for 3-O-MG transport in and out are equivalent regardless of the insulin concentration. Conversely, rat skeletal muscle [3-O-MG] is reduced by hyperglycemia even in the presence of high serum insulin concentrations (42), as would be predicted by Eq. 6.

It is impossible to get an accurate value for intracellular glucose from direct glucose measurements in excised and biopsied tissue for several reasons. Estimates of intracellular glucose concentration require that interstitial glucose in tissue samples is known. Blood glucose concentrations have been used in place of interstitial glucose (66). This leads to spurious numbers, since estimates of skeletal muscle interstitial glucose concentration made using microdialysis show that it is well below arterial or venous plasma concentrations (34). Because 3-O-MG is not metabolized, its plasma concentration equals its interstitial concentration and can be used to calculate intracellular [3-O-MG] from measurements in tissue samples. Even if the intracellular glucose concentration could be determined, it may not be the most sensitive determinant of the glucose diffusion gradient in the cell, since physical barriers and spatial glucose gradients may compartmentalize glucose within the cell. In this regard, there is evidence that the glucose transporters (9, 13, 19, 36, 37, 39) and hexokinases (2, 26, 32, 58) are localized to specific regions within the skeletal muscle cell. Glucose concentration gradients may exist within the intracellular space so that concentrations are diminished from regions of high glucose transport to regions of high glucose phosphorylating activity. Because intracellular 3-O-[³H]MG is not metabolized, it can be measured and, at equilibrium, is homogenous within the contiguous intracellular space.

R_g increases in soleus without an increase in S_i/S_o and [G]_im at an insulin dose of 0.6 mU · kg¹ · min¹, suggesting that hexokinase can accommodate the increase in insulin-stimulated glucose transport at this insulin concentration. It is also possible that insulin increases the hexokinase activity. This is consistent with the increase in the rate constant for glucose phosphorylation that was predicted by compartmental analysis (54) and studies using ¹⁸F-2-deoxyglucose (28). The basis for this change may relate to an increase in the fraction of hexokinase II that is bound to mitochondria and thus in its more active form (3). One study showed, however, that there is no change in the fraction of bound hexokinase II in human muscle during hyperinsulinemia (25). Although it is also possible that more hexokinase II is synthesized during the hyperinsulinemic clamp (35, 48), the time needed for this to occur in skeletal muscle is >6 h. An increased ability to phosphorylate glucose may also result from the reduction in glucose 6-phosphate that has been shown to occur in rat skeletal muscle at about the same insulin concentration as was obtained with the 0.6 mU · kg¹ · min¹ infusion (53). At higher insulin concentrations, glucose 6-phosphate begins to return to basal concentrations (53). In the present study, this correlated to the increase in [G]_im.

Basal and insulin-stimulated R_g are higher in oxidative than nonoxidative muscle (30). Hexokinase II (46, 58) and GLUT-4 (23, 29) are both more abundant, and oxidative fibers are better perfused than nonoxidative fibers of the rat. In the present study, insulin was shown to have different effects on the tissue glucose concentration of slow- and fast-twitch muscles. Soleus glucose concentration fell insignificantly and only transiently with increasing insulin concentrations. In contrast, vastus lateralis and gastrocnemius glucose concentrations both fell significantly (Table 2). Results obtained in human skeletal muscle, which contains both fast- and slow-twitch muscles, support the findings of the present study by showing that glucose concentration either falls slightly or stays the same during a hyperinsulinemic, euglycemic clamp (24, 51, 55, 57). The rat model is advantageous in that certain muscles are homogeneous for fiber type, permitting fiber type-specific effects to be assessed. Just as in the present study, the glucose concentration in a rat muscle that is predominantly fast twitch (rectus abdominus) was shown to decrease in the presence of hyperinsulinemia (66). The greater decrease in muscle glucose content in fast-twitch muscle suggests that glucose availability is a more serious limitation in these tissues and is more likely to compromise insulin-stimulated glucose uptake. This deficit is probably due to the lower blood flow and greater diffusion distances in these tissues.

Skeletal muscle comprises ~50% of total body mass and exhibits the greatest increases in glucose uptake in response to insulin and exercise. A knowledge of the regulation of skeletal muscle glucose uptake therefore is a prerequisite to understanding normal and pathophysiological whole body glucose uptake. These studies provide insight into the control of glucose uptake by showing that 1) increases in R_g can occur in response to insulin levels in the physiological range without decreasing TSGG, suggesting that muscle glucose delivery and glucose phosphorylation are adequate to accommodate the increased glucose transport flux; 2) insulin at high physiological or supraphysiological levels leads to a decrease in TSGG, which suggests that the increase in transport activity has exceeded glucose delivery to the muscle or glucose phosphorylation within the muscle; 3) maximum R_g is sustained even though TSGG continues to fall, indicating that glucose transport still has the capacity to increase and is not rate limiting for insulin-stimulated glucose uptake; and 4) in contrast to the soleus, which exhibits only a transient fall in muscle glucose, muscle consisting of fast-twitch fibers that are poorly perfused actually exhibits a 35-45% fall in tissue glucose, suggesting that glucose delivery is a major limitation in sustaining the TSGG during insulin stimulation in these tissues. In conclusion, control of glucose uptake is distributed between glucose transport and factors that determine the TSGG. The stimulation of glucose transport that occurs with increasing insulin concentrations places more importance on the factors that maintain glucose delivery to the muscle membrane and glucose phosphorylation inside the muscle.