INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

IN MAMMALS, SKELETAL MUSCLE is the main tissue responsible for insulin-stimulated glucose uptake, an essential process for maintenance of normal glucose homeostasis. Glucose entrance across sarcolemmal membranes is thought to be the limiting step in skeletal muscle glucose metabolism (26). Entry occurs through a facilitated diffusion process mediated by two isoforms of the glucose transporter protein, GLUT-1 (22) and GLUT-4 (reviewed in Ref. 16). GLUT-1 is mainly associated with the cell surface and may mediate basal glucose transport. In the basal (nonstimulated) state, the more abundant GLUT-4 is largely confined to an intracellular compartment (reviewed in Ref. 16). Glucose transport is increased by insulin, hypoxia, or muscle contractions and primarily results from GLUT-4 translocation from intracellular stores to the sarcolemmal compartment (11; reviewed in Ref. 16). Alternatively, glucose transport may also be altered by changes in the intrinsic activity of the surface GLUT-4 (10, and references therein).

In contrast to the extensive investigations carried out in mammalian species, very little information is available in avian species, which, like mammals, are endotherms showing a regulated glycemia despite extremely variable needs in tissue glucose metabolism and irregular food intake. This apparent disinterest may be related to the fact that birds such as chickens show a relative insensitivity to the hypoglycemic effect of insulin in vivo (28). More recent studies indicated some differences between avian and mammalian glucose transport in skeletal muscle. Indeed, although antibodies against mammalian GLUTs led to a positive immunodetection of GLUT-1 in various chicken tissues, including brain, skeletal muscle, and myotubes, no immunoreactive GLUT-4 was detected (6). These results suggested that either the chicken GLUT-4 transporter is very different from the mammalian one and hence is not recognized by the antibodies used or that this type of transporter is not expressed in chicken tissues. However, there is some in vitro evidence for insulin or insulin-like growth factor-I-mediated stimulation of glucose uptake in chicken myotubes (6). Although the stimulation is much lower in amplitude than that observed in isolated mouse soleus muscle (24) or rat perfused hindlimb muscles (14, 16), it suggests that an insulin-dependent glucose transporter may be expressed in avian skeletal muscle. To our knowledge, no experimental data have yet demonstrated the existence of such a transporter in avian skeletal muscle.

The aim of this work was, therefore, to reevaluate the characteristics and regulation of glucose transport in skeletal muscles of an avian species, the Muscovy duck (Cairina moschata). We report that duckling sarcolemmal vesicles exhibit a saturable component of D-glucose transport. By Western blot analysis, it is shown that skeletal muscle membranes contain proteins homologous to mammalian GLUT-1 and GLUT-4 transporters. Functionally, translocation of internal GLUT-4 stores is controlled by insulin and regulated in accordance to the metabolic needs of the tissue.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Chemicals. Human insulin (Actrapid) was from NOVO Nordisk, U-D-[¹⁴C]glucose (9.36 Gbq/mmol) and 1-(N)-L-[³H]glucose (518 GBq/mmol) were from NEN-DuPont de Nemours, France, and ¹²⁵I-labeled protein A was from Amersham, France. Other chemicals were purchased from Sigma (St. Quentin Fallavier, France).

Animals. Male Muscovy ducklings (Cairina moschata L, pedigree R31, Institut National de la Recherche Agronomique) were obtained from a commercial stockbreeder (Ets Grimaud). They were fed a commercial mash (Aliment Genthon Démarrage) ad libitum and had free access to water. Ducklings were kept in a constant photoperiod (8:16-h light-dark cycle). From the age of 1 wk, ducklings were caged in groups of six for a period of 4 wk at an ambient temperature (T_a) of 25°C [thermoneutral controls (TN)]. Animals were taken in a postabsorptive phase and killed by decapitation, and gastrocnemius muscle was sampled, frozen in liquid nitrogen, and kept at 70°C until used for membrane preparations. Another batch of TN ducklings, which were fasted for 2 h before the start of the surgery, was used in perfused muscle experiments in vitro. To enhance the metabolic needs of skeletal muscle, a third batch of ducklings was exposed to 4°C T_a from the age of 1 wk. This cold-acclimation (CA) schedule described by Barré et al. (1) leads to the development of increased muscle thermogenic capacities. At experiment, 5-wk-old ducklings weighed between 1.3 and 1.5 kg.

Assessment of insulin action on perfused muscle in vitro. The duckling limb surgical preparation described by Marmonier et al. (20) was used. Briefly, the left lower limb was surgically isolated under halothane anesthesia. The mass of perfused muscle was generally in the range 23-27 g. Perfusion was via the popliteal artery through polyethylene tubing (ID = 1.2 mm) connected to a nonrecirculating constant flow system thermostated at 37°C. An erythrocyte-free Krebs-Henseleit buffer was used as the perfusion medium, with a composition (in mM) of 120 NaCl, 4.8 KCl, 1.2 KH₂PO₄, 1.2 MgSO₄, 7 H₂O, 2.5 CaCl₂, and 24 NaHCO₃. In addition, the buffer contained 10 mM glucose and 2% BSA, pH 7.4, and was continuously equilibrated with carbogen (95% O₂-5% CO₂). Perfusions were conducted at 0.47 ml · min¹ · g¹ to ensure appropriate oxygenation of the muscle preparation (20). After an initial equilibration period of 45 min to reach a steady state, aliquots of arterial and venous perfusates were used for glucose determination after either saline or insulin addition to the arterial perfusate. The insulin clamp (500 or 5,000 µIU/ml) lasted for 15-20 min. For membrane preparations, gastrocnemius muscle was freeze-clamped in situ at the end of the perfusion with aluminum clamps cooled by liquid nitrogen. Muscle samples were kept frozen at 70°C until used for membrane preparation.

Comparative perfusion experiments were also performed using a perfused rat hindlimb preparation and the experimental setting described above. Hindlimbs of Wistar rats weighing 240-280 g were perfused at 32°C with the Krebs buffer described above at a flow rate of 6-7 ml/min.

Sarcolemmal membrane preparation. Sarcolemmal membrane vesicles were prepared according to the method of Grimditch et al. (8), with modifications. Skeletal muscle obtained after birds were killed, or after perfusion experiments in vitro, was minced in a sucrose-HEPES (SH) buffer containing 250 mM sucrose and 20 mM HEPES (pH 7.4) and homogenized with a polytron homogenizer (2 bursts of 8 s at setting 8). The mixture was then filtered through a piece of nylon mesh of 40-µm pore size to give a crude homogenate. A volume of KCl (3 M) and sodium pyrophosphate (250 mM) solution equal to 10% of the crude homogenate was added to solubilize contractile proteins. The mixture was then spun for 50 min at 184,000 g. The supernatant was discarded, and the pellet was resuspended in SH buffer using a manual glass-Teflon tissue homogenizer. DNase (30,000 U/20 ml homogenate) was added, and the mixture was incubated in a shaking water bath for 60 min at 30°C. After incubation, an equal volume of ice-cold SH buffer was added and the mixture was kept on ice for 5 min. The final suspension was then spun at 750 g for 8 min. The supernatant was spun at 184,000 g for 50 min, and the resultant pellet was resuspended in 45% sucrose solution using a manual glass-Teflon tissue homogenizer. This suspension of crude muscle membranes constituted the lower layer of a discontinuous sucrose gradient consisting of equal volumes of 45, 38, 32, 30, 27, and 12% sucrose solutions. Tubes were then spun for 16 h at 64,000 g. The sarcolemmal fraction, located between the 12 and 27% layers, was sampled, washed, and pelleted by centrifugation in 1 mM NaHCO₃ for 50 min at 184,000 g. The final pellet containing the purified sarcolemmal fraction was resuspended in appropriate buffers (see below), and the relevant biochemical assays were performed. All the above steps were carried out at 4°C and pH 7.4.

Assessment of the purity of sarcolemmal fraction. The quality of the sarcolemmal fraction was determined using classic marker enzymes. The purification level was estimated using the activity of the K⁺-stimulated p-nitrophenylphosphatase (KpNPPase) according to the method of Grimditch et al. (8). The contamination by sarcoplasmic reticulum and mitochondria was assessed by the activities of the NADPH cytochrome c reductase (30) and the succinate dehydrogenase (25), respectively. Protein content was assessed by the bicinchoninic acid method using BSA as standard.

Glucose transport determination. The uptake of radiolabeled glucose into sarcolemmal vesicles was determined using a filtration method under zero-trans conditions. All measurements were performed at 20°C with freshly prepared membrane vesicles.

The transport medium contained (in mM) 10 Tris (pH 7.4), 100 NaCl, 4 KCl, 1 MgCl₂, and varying concentrations of unlabeled substrates plus a fixed concentration of labeled substrates (D-[¹⁴C]glucose, 10 µCi/ml; and L-[³H]glucose, 40 µCi/ml) and varying amounts of sucrose up to 100 mM to adjust osmolarity. The reaction was started by adding 10 µl of membrane suspension to 40 µl of transport medium. After a given incubation time, the preparation was quenched by adding 1 ml ice-cold stop solution (transport medium without radioisotope but containing 0.2 mM phloretin to inhibit further mediated D-glucose transport). Membrane vesicles were collected on nitrocellulose filters (0.45 µm, Millipore, Saint-Quentin en Yvelines, France) and washed with 5 ml ice-cold stop solution. Vesicle-associated radioactivity was determined by liquid scintillation. The specific carrier-mediated uptake of D-glucose was determined by subtracting the initial rate of L-glucose uptake from that of D-glucose. An incubation time of 5 s was chosen to estimate the initial rate of uptake and applies to all the results presented in Tables 1-3 and Figs. 1-7, except for the time course experiment. Maximal transport rate (V_max) is extrapolated from sarcolemmal vesicles values to 1 g wet wt muscle using yield of activity of KpNPPase, the marker enzyme of plasma membrane.

Western blot analysis. Proteins contained in sarcolemmal membrane (30 µg protein) and crude membrane (40 µg protein) fractions were separated by 12% SDS-PAGE according to the method of Laemmli (18). Separated proteins were electrophoretically transferred to a nitrocellulose membrane (Biorad, Ivry-sur-Seyne, France), blocked with BSA, and incubated overnight at 4°C (4) with polyclonal antibodies against mammalian GLUT-1, GLUT-4, or GLUT-5 transporters (East Acres, Southbridge, MA). The antibody-transporter complexes were revealed by two methods. Qualitativedetection of the complexes was realized using a second antibody conjugated to alkaline phosphatase activity, which was detected by colorimetric analysis. Semiquantitative determination of GLUT-4-like protein was performed by autoradiography after incubation of the antibody-transporter complex with ¹²⁵I-labeled protein A (2 µCi/ml). Transfer membranes were washed and then exposed to Kodak XAR-5 films for 48 h at 80°C. Autoradiographic bands were analyzed by densitometry using a Biocom Phoretix Software System. Autoradiographic exposures were within the linear response capability of the film. Quantification was relative to a "GLUT-4 standard" used for the whole experiment and always deposited in parallel to duckling muscle samples. "GLUT-4 standard," defined as an arbitrary unit, was the immunological signal obtained from 10 µg of a unique crude membrane fraction of rat skeletal muscle.

Comparison between groups of animals was made after correction for changes in enrichments and yields of membrane fractions due to insulin perfusion or cold exposure. This correction also decreased variability of values within the different series of animals. GLUT-4 content of each sarcolemmal fraction was calculated per unit of KpNPPase activity present in this fraction. Next, the result was extrapolated to the KpNPPase activity present in 1 g wet wt muscle and expressed as units of GLUT-4 per gram wet weight.

Statistical analyses. Values are presented as means ± SE. Statistical significance of observed variations was assessed by ANOVA followed by post hoc tests and was recognized at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Enzymatic characterization of sarcolemmal fractions. Activities of marker enzymes in homogenate and membrane fractions are shown in Table 1. The KpNPPase activity, the enzyme commonly used as a marker of plasma membranes, was considerably increased (18-fold) in the sarcolemmal fraction compared with the original homogenate and 4.4-fold by comparison with the crude membrane fraction, indicating a high enrichment in plasma membranes.

Specific activities of NADPH cytochrome c reductase and succinate dehydrogenase were increased by 1.6- and 0.6-fold, respectively, in the sarcolemmal fraction, indicating a low contamination by sarcoplasmic reticulum and mitochondrial membranes.

Existence of a facilitated D-glucose transport. The time course of D-glucose uptake by sarcolemmal vesicles for an extravesicular glucose concentration of 5 mM is shown in Fig. 1. The curve is characteristic of a diffusional process leading to a diffusion equilibrium. D-Glucose uptake was linearly related to time during the first 10 s of incubation. The choice of an incubation time of 5 s for all subsequent analysis was convenient for a precise determination of the initial rate of uptake.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Time course of glucose transport by sarcolemmal vesicles. Incubation medium contained (final concentrations): D-glucose + D-[14C]glucose (5 mM + 10 µCi/ml), NaCl (100 mM), KCl (4 mM), MgCl2 (1 mM), sucrose (95 mM), Tris (10 mM); pH 7.4. Incubation was performed at 20°C. Other conditions were as described in EXPERIMENTAL PROCEDURES. Each point is mean of triplicate determinations.

To determine the kinetic parameters of D-glucose transport by sarcolemmal vesicles isolated from duckling gastrocnemius muscle, the initial rate of D-glucose uptake was examined as a function of extravesicular D-glucose concentration. L-Glucose uptake was determined at similar extravesicular concentrations to correct D-glucose uptake for simple diffusion and filter retention. Figure 2 shows that the initial rate of specific D-glucose uptake was a saturable function of the extravesicular D-glucose concentration and followed typical Michaelis-Menten kinetics. Kinetic parameters of transport were determined by nonlinear regression analysis with a Graph-Pad-Prism package. The apparent transport constant (K_m) was 16.7 mM, and the V_max was 1.2 nmol · s¹ · mg protein¹. An extrapolation to the whole muscle would give a maximal transport capacity of 10.5 nmol · s¹ · g muscle¹. The demonstration of a saturable component of D-glucose transport in sarcolemmal vesicles indicated the existence of sarcolemmal D-glucose transporters.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Kinetics of D-glucose transport by sarcolemmal vesicles. Incubation conditions were those described in Fig. 1, except that variable glucose concentrations were used. Differing concentrations were osmotically equilibrated by sucrose. Incubation time was 5 s. Data are means ± SE obtained from 5 ducklings, each determination being performed in triplicate. Carrier-mediated D-glucose () was obtained by subtracting simple diffusion as determined with L-glucose uptake () from total D-glucose uptake ().

Identification of D-glucose transporters in duckling skeletal muscles. By using specific antibodies against mammalian GLUT transporters, we investigated the presence of the three hexose transporters described in mammalian skeletal muscle: the ubiquitous GLUT-1, the insulin-dependent GLUT-4, and the fructose transporter GLUT-5. On account of the existence of intracellular stores of GLUT-4 in mammalian muscles, immunodetection was performed on crude membrane fractions. Positive and negative controls were performed in parallel.

Figure 3 shows that polyclonal antibodies directed against rat GLUT-1 recognized a protein of similar size (~45 kDa) in Western blots of duckling crude membranes. No other immunoreactive band was visible on the gel. A band of similar size was also obtained with rat erythrocyte ghost membranes used as a positive control, whereas no signal was obtained with duckling intestinal microvillar membranes, a structure devoid of GLUT-1 transporters in mammals. Comparison of the GLUT-1 immunoreactive proteins between rats and ducklings suggests similar molecular weight and degree of glycosylation in the two species.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Western blot identification of GLUT-1 homologous protein in duckling skeletal muscle. A: rat ghost membranes (50 µg protein). B: crude membranes from duckling gastrocnemius muscle (40 µg protein). C: duckling intestinal brush-border membranes (50 µg protein). Proteins were separated by 12% SDS-PAGE and transferred to a nitrocellulose membrane. GLUT-1 homologous protein was detected by a polyclonal antibody against rat GLUT-1 transporter (dilution 1/500). Blot was developed by use of a second antibody conjugated to alkaline phosphatase activity. Scanning of nitrocellulose membranes after immunoblotting was performed with an Apple "Onescanner."

Figure 4 shows that polyclonal antibodies directed against rat GLUT-4 transporter recognized a protein of similar size (~45 kDa) in crude membranes prepared from rat (positive control) and duckling skeletal muscle. No signal was obtained for rat erythrocyte ghost membranes chosen as a negative control. These results confirmed that there was no cross-reactivity between the antibodies used for the determination of GLUT-1 and GLUT-4.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Western blot identification of GLUT-4 homologous protein in duckling skeletal muscle. A: standard proteins: phosphorylase B (107 kDa), albumin (85 kDa), and ovalbumin (47 kDa). B: crude membranes of rat skeletal muscle (20 µg protein). C: crude membranes of duckling gastrocnemius muscle (40 µg protein). D: rat ghost membranes (50 µg protein). Conditions of immunodetection were those described in Fig. 3. A polyclonal antibody directed against rat GLUT-4 transporter (dilution 1/500) was used.

Figure 5 shows that polyclonal antibodies directed against human GLUT-5 transporter recognized an homologous protein in duckling intestinal microvillar membrane, which was selected because this structure contains a large amount of GLUT-5 transporters in mammals (data not shown). By contrast, no signal was detectable in sarcolemmal membranes from duckling gastrocnemius muscle nor in rat erythrocyte ghost membranes (negative control).

View larger version (6K): [in this window] [in a new window]   Fig. 5.   Western blot analysis of GLUT-5 homologous protein in duckling skeletal muscle and intestinal brush-border membranes. A: duckling sarcolemmal membranes (40 µg protein). B: duckling intestinal brush-border membranes (50 µg protein). C: rat ghost membranes (50 µg protein). Conditions of immunodetection were those described in Fig. 3. A polyclonal antibody directed against human GLUT-5 transporter (dilution 1/500) was used.

Quantification of GLUT-4 and GLUT-1 homologous transporters in crude and sarcolemmal membrane fractions of duckling skeletal muscle. Specific quantification of GLUT-4 homologous protein was performed using ¹²⁵I-labeled protein A and expressed as units per gram of muscle, as defined in EXPERIMENTAL PROCEDURES. Crude membranes isolated from duckling gastrocnemius muscle contained 93.8 ± 25.5 U/g muscle of GLUT-4 homologous protein, whereas sarcolemmal membranes contained only 18.4 ± 4.2 U/g muscle. The arbitrary unit was defined as the amount of GLUT-4 contained in 10 µg of crude membrane protein obtained from rat skeletal muscle. These results, therefore, suggest that when ducklings were killed in a postabsorptive state, the majority of muscle GLUT-4 homologous protein was located in an intracellular compartment, whereas only a minority (~20%) was present at the cell surface on sarcolemmal membranes. This situation resembles that observed in mammalian skeletal muscle in the absence of insulin or exercise stimulation.

The protein A method did not allow us to quantify muscle GLUT-1 homologous protein, because the signal was much too faint. This observation may be linked to the low level of GLUT-1 homologous protein expressed in duckling muscle, as noted in mammals.

Response of glucose transport and transporters to insulin in duckling skeletal muscle. To obtain further experimental evidence for the existence of a GLUT-4-like glucose transporter in duckling skeletal muscle, we investigated the characteristics of glucose uptake by a perfused muscle preparation in vitro and analyzed the transporter distribution between internal and sarcolemmal membranes. Table 2 shows the effects of insulin on glucose uptake of perfused duckling muscles and a comparison with perfused rat hindlimb. There was a dose-dependent increase in glucose uptake in both groups of animals. Perfusion of muscles with a supramaximal concentration of insulin (5,000 µU/ml) increased glucose uptake more than 5-fold in ducklings and 13-fold in rats. The net increment in uptake (stimulated minus basal) was 15.6 µmol · g¹ · h¹ in ducklings, lower than that found in rats (24.5 µmol · g¹ · h¹). A marked difference in the response to insulin (17.6 vs. 9.6 µmol · g¹ · h¹) was also observed between duckling and rat muscles at the lower concentration of insulin (500 µU/ml).

To estimate the effects of insulin on glucose transport kinetics, sarcolemmal vesicles were isolated from insulin-stimulated or unstimulated perfused duckling muscles and the presence of GLUT-4 homologous protein was assessed by immunodetection. Because KpNPPase is associated solely with plasma membranes, the yield of enzyme activity of each preparation was used for expressing and comparing other parameters measured in sarcolemmal vesicles. This adjusted the data for differences in sarcolemmal protein enrichments, which were slightly higher, but not significantly, in fractions prepared from insulin-perfused muscle than those prepared from muscle perfused without insulin. Similar increases in protein yield in crude and sarcolemmal fractions after insulin were also observed by others (9). Figure 6 illustrates the results obtained in a typical experiment. Stimulation with insulin induced a marked increase in the initial rate of D-glucose transport by sarcolemmal vesicles. This was associated with a concomitant increase in the sarcolemmal content of GLUT-4 homologous protein as determined by Western blots. By contrast, there was no increase in the total amount of muscle GLUT-4 homologous protein after stimulation with insulin. These results were reproduced in all perfusion experiments, and mean values are presented in Table 3. Kinetic constants of carrier-mediated and nonmediated D-glucose uptake of sarcolemmal vesicles are also shown. The nonspecific glucose uptake was not affected by insulin stimulation, as indicated by the constancy of the diffusional constant of the transport. The K_m of the transporter for the substrate was also unchanged. By contrast, the V_max of specific D-glucose transport was increased 2.7-fold by insulin. Similarly, insulin induced a 2.4-fold concomitant increase in the content of GLUT-4 homologous protein in the sarcolemmal fraction. The total content of GLUT-4 homologous protein was, however, not altered by insulin, suggesting that insulin markedly affected the cellular distribution of the protein, in favor of sarcolemmal membranes.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effect of insulin on D-glucose uptake by sarcolemmal vesicles and cellular distribution of D-glucose transporters in duckling gastrocnemius muscle. Results of a typical experiment realized with sarcolemmal vesicles or membrane fractions obtained from in vitro perfused leg muscles with or without insulin stimulation. A: mean of triplicate assays of carrier-mediated D-glucose uptake measured in sarcolemmal vesicles as a function of sugar concentration. B: corresponding Western blots of muscle crude membrane proteins (CM; 40 µg protein) and sarcolemmal membranes (SM; 30 µg protein). Quantification was performed with protein A.

Muscle glucose transport and distribution of GLUT-4 homologous protein in CA ducklings. CA ducklings are known to exhibit increased capacity for skeletal muscle thermogenesis and, therefore, represent a convenient model to investigate the influence of a prolonged increased energy expenditure of the cells on glucose transport. Figure 7 shows the results of a typical experiment comparing the characteristics of sarcolemmal vesicle glucose transport and the corresponding cellular distribution of GLUT-4 homologous protein in a TN and a CA duckling. Thus data were corrected as described in the case of insulin-perfused muscle preparations. Such as those obtained from insulin-perfused muscles, data were corrected to take into account the fact that sarcolemmal fractions prepared from CA ducklings presented a higher protein enrichment than those of controls. Cold-acclimation resulted in a higher initial transport rate of D-glucose by isolated sarcolemmal vesicles, indicating an increased V_max of specific D-glucose transport. The rise in V_max was related to a concomitant increase in the proportion of cellular GLUT-4 homologous proteins associated with the sarcolemmal fraction.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Effect of cold acclimation on D-glucose uptake by sarcolemmal vesicles and cellular distribution of GLUT-4 homologous protein in duckling skeletal muscle. Results of a typical experiment realized with sarcolemmal vesicles and membrane fractions obtained from gastrocnemius muscle of a thermoneutral control (TN) and a cold-acclimated (CA) duckling. A: means of triplicate assays of carrier-mediated D-glucose uptake measured in sarcolemmal vesicles as a function of glucose concentration. B: distribution of GLUT-4 homologous protein between internal pool and sarcolemmal fraction. GLUT-4 homologous protein was quantified with protein A on Western blots of muscle crude membrane or sarcolemmal fractions. Internal pool was determined by difference between crude and sarcolemmal fractions.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The present study provides a substantial body of evidence supporting the existence of a glucose transporter in duckling skeletal muscle with molecular and functional properties resembling those of the mammalian GLUT-4 transporter.

Glucose transport and transporters in duckling skeletal muscle. Present results have shown that the glucose transport in sarcolemmal vesicles isolated from duckling gastrocnemius muscle is substrate dependent, as described by Michaelis-Menten. Such kinetics indicate the existence in ducklings, as in mammals, of a facilitative uptake of glucose, which may be mediated by sarcolemmal membrane-associated glucose transporters. The K_m value was 17 mM in duckling sarcolemmal membranes. With the use of similar experimental conditions, respective K_m values of 18 and 17 mM were reported for sarcolemmal vesicles isolated from the red and the white regions of the rat gastrocnemius muscle (7).

Antibodies directed against rat GLUT-1 and GLUT-4 and human GLUT-5 have been used to characterize the hexose transporter isoforms present in duckling muscle membranes. Control Western blots showed the specificity of these antibodies in recognizing the isoforms of the transporter. Membrane fraction immunostaining indicated the presence of GLUT-1 and GLUT-4 homologous proteins in skeletal muscle of duckling. Anti-rat GLUT-1 antibodies have already been shown to recognize a homologous protein expressed by chick embryo fibroblasts (32), as well as chicken myotubes or various chicken tissues, including skeletal muscle (6). Contrary to the present results in ducklings, no immunostaining was detected in chicken myotubes or chicken tissues using antibodies against rat GLUT-4 (6). Variations in the antibodies used, rather than species, may account for this difference. In the present study, the positive immunostaining of duckling skeletal muscle membranes with anti-rat GLUT antibodies together with the fact that the immunoreactive proteins are of similar molecular weight to those found in mammals, forms an effective argument for the expression of GLUT-1 and GLUT-4 homologous proteins in duckling skeletal muscle. The structure of these proteins may, therefore, be relatively conserved between both classes. By contrast, the fructose transporter GLUT-5, present in mammalian skeletal muscle (13, 33), was not detected in duckling sarcolemmal membranes despite the fact that these antibodies recognized a GLUT-5 homologous protein in duckling intestinal microvillar membranes.

Insulin stimulation of glucose transport in duckling skeletal muscle and its relationship with GLUT-4 homologous protein in sarcolemmal membranes. Using an in vitro perfused muscle preparation, we have shown that duckling skeletal muscle glucose uptake was rapidly (15-20 min) and dose dependently stimulated by insulin (5.3-fold above basal at supraphysiological doses of insulin). Rapid stimulation of glucose uptake by insulin or insulin-like growth factor-I has also been reported in chicken myotubes in vitro (6). The increase in glucose uptake reported in the present study was, however, lower than that observed with a similar technique in a rat hindlimb preparation (Table 2). Although the mechanism responsible for such relative insulin resistance in duckling muscles remains to be determined, it could contribute to the relative insensitivity of birds to the hypoglycemic effect of insulin in vivo (28).

The mechanisms involved in the insulin-stimulated glucose uptake of duckling perfused muscles were investigated. Insulin perfusion induced a 2.7-fold increase in glucose transport by sarcolemmal membrane vesicles. This acute effect resulted from an increase in V_max with no change in K_m and is, therefore, comparable to that observed in mammalian skeletal muscle (23, 31). As in mammals (33), the insulin-induced increase in glucose transport was lower when measured with membrane vesicles than with perfused muscles. This is presumably due to the documented muscle fiber-specific effect of insulin (17). Indeed, the perfused limb preparation contains heterogeneous muscles, whereas only the gastrocnemius muscle was used for sarcolemmal vesicle isolation.

The insulin-induced increase in glucose transport by sarcolemmal vesicles was associated with a 2.4-fold increase in the amount of the GLUT-4 homologous protein in the sarcolemma, yet no change occurred in the total amount of GLUT-4 homologous proteins in the crude membrane fraction. These results indicate that acute hyperinsulinemia induces a change in the cellular distribution of the protein from intracellular stores to plasma membrane, rather than affecting the translation of the protein. The quantitative increase in plasma membrane GLUT-4 was only half the increase in insulin-stimulated glucose transport, but similar results have been reported in mammals, with GLUT-4 increases in the plasma membrane ranging from 50 to 200%, whereas the increases in glucose transport in whole muscle are larger (200-800%; Ref. 10 and references therein). Part of the difference could also be due to the fact that GLUT-4 distribution analysis was restricted to gastrocnemius muscle. For instance, when analyzed in the same skeletal muscle, the quantitative increase in plasma membrane GLUT-4 (2.4×) induced by insulin was very similar to the quantitative increase in sarcolemmal vesicle glucose transport (2.7×).

Although we were unable to quantify the GLUT-1 homologous proteins in the present study, any contribution of this protein to the insulin-mediated glucose uptake is unlikely given the constancy of the GLUT-1 content reported in experiments with chicken myotubes when glucose uptake was stimulated (6). Similar observations have been made in mammals with respect to the action of insulin on skeletal muscle and adipose tissue (7, 33). In ducklings, the increase in the sarcolemmal content of GLUT-4 homologous protein may, therefore, result, as in mammals (12), from an insulin-mediated translocation of the protein to the sarcolemma from intracellular stores. The notion that the protein detected in duckling skeletal muscle is a GLUT-4-like protein similar to that described in mammals is reinforced by the fact that it shows an insulin-regulated cellular distribution, and its sarcolemmal location is associated with glucose transport. The mammalian model describing the action of insulin on glucose uptake is, therefore, likely to extend to avian species.

Stimulation of glucose transport by the energy expenditure of skeletal muscle cells. Because GLUT-4 in mammals is regulated not only by insulin, but also by metabolic needs or muscle activity (21, 23), we investigated the response of GLUT-4 homologous protein to a situation in which muscle energy expenditure is markedly increased. Indeed, long-term cold-exposed ducklings develop enhanced skeletal muscle thermogenic capacity to maintain homeothermy (1, 5). Present results show that cold-acclimation was associated with an increase in glucose transport by sarcolemmal vesicles (increased V_max). The rise in skeletal muscle glucose transport in CA ducklings differs from the situation in mammals, where no change in skeletal muscle glucose transport was found in CA animals (2, 27, 29). Furthermore, it was recently shown that rat GLUT-4 gene expression was decreased after 6 days in the cold (19). This difference may be related to the fact that skeletal muscle is the main thermogenic site in CA ducklings, contrary to rats, where brown adipose tissue is the specialized heat-producing tissue (5). In mammals, a rise in glucose transport is generally explained by a higher number of sarcolemmal transporters or by a stimulation of their intrinsic activity (3, 12, 15). Present results suggest that in ducklings, the cold-induced increase in glucose transport may be achieved, at least in part, by a cellular redistribution of the existing proteins in favor of the active sarcolemmal sites.

In conclusion, we have presented several experimental results, supporting, for the first time, the existence in avian skeletal muscle of a protein with molecular and functional homologies with the mammalian glucose transporter GLUT-4. Its cellular distribution is apparently a key modulator of cellular glucose uptake and is controlled by insulin and the energy expenditure of the cell.

Perspectives

The field of integrative physiology would benefit from studies on the role of the adaptive response of glucose transport on nonshivering thermogenesis and locomotive function in CA ducklings.