Euglycemic, hyperinsulinemic clamp tests have shown that adult ruminants are less insulin-sensitive than monogastric omnivores. The present study was carried out to elucidate possible cellular mechanisms contributing to this impaired insulin sensitivity of ruminants. Western blotting was used to measure glucose transporters 1 and 4 (GLUT1, GLUT4) in oxidative (musculus masseter and diaphragm) and glycolytic (musculus longissimus dorsi and semitendinosus) skeletal muscle in the crude membranes of pigs and cows. Muscles were characterized biochemically. To determine insulin-stimulated 3-O-d-[3H]-methylglucose (3-O-MG) uptake and GLUT4 translocation, porcine and bovine musculus semitendinosus strips were removed by open muscle biopsy and incubated without and with 0.1 or 20 mIU insulin/ml. GLUT4 translocation was analyzed using subcellular fractionation techniques to isolate partially purified plasma membranes and cytoplasmic vesicles and using Western blotting. GLUT4 protein contents were significantly higher in oxidative than in glycolytic muscles in pigs and cows. GLUT1 protein contents were significantly higher in glycolytic than in oxidative muscles in bovines but not in porcines. The 3-O-MG uptake into musculus semitendinosus was similar in both species. Maximum insulin-induced GLUT4 translocation into musculus semitendinosus plasma membrane was significantly lower in bovines than in porcines. These results indicate that GLUT1 is the predominant glucose transporter in bovine glycolytic muscles and that a reinforced insulin-independent glucose uptake via GLUT1 may compensate for the impaired insulin-stimulated GLUT4 translocation, resulting in a similar 3-O-MG uptake in bovine and porcine musculus semitendinosus. These findings may explain at least in part the impaired in vivo insulin sensitivity of adult ruminants compared with that of omnivorous monogastric animals.
- skeletal muscle
- glucose transporter 1
- glucose transporter 4
adult ruminants and monogastric omnivores differ to a great extent with regard to their glucose supply. Monogastric omnivores meet their glucose requirements mainly by enteral absorption, whereas ruminants, for the most part, use glucose produced by endogenous gluconeogensis, with propionate as the major precursor (52). Furthermore, because of these different nutrient strategies, adult ruminants seem to be less insulin-sensitive than monogastric omnivores (15, 36, 54).
Skeletal muscle is the main mammalian tissue of glucose utilization and therefore the major site of differences in insulin sensitivity (62). Myocyte glucose uptake is mediated by two specific 45–50 kDa transport glycoproteins (47). The insulin-independent glucose transporter 1 (GLUT1) is predominantly located in the muscle cell plasma membrane, and accounts for the basal glucose supply of the myocyte. The insulin-regulated glucose transporter 4 (GLUT4) recycles between the muscle cell plasma membrane (PM) and an intracellular tubulovesicular pool, where it is associated with cytoplasmic vesicles (CV). In hyperglycemic states, insulin is secreted from the endocrine pancreas and stimulates the myocyte glucose uptake by increasing the translocation of intracellular GLUT4 vesicles into the plasma membrane (35). Both transporters are highly conserved among mammalian species, especially with regard to their C-terminal domain, which is responsible for antigenic properties. Rat, pig, and bovine 38-amino-acid C-terminal regions of GLUT4 exhibit a single amino acid difference: asparagin508 in bovines is replaced by histidin in porcines and rats (2). There are no differences between rats (7), pigs (59), and bovines (8) in the corresponding GLUT1 sequences.
We assume that the impaired in vivo insulin sensitivity of adult ruminants compared with that of monogastric omnivores is due to a different distribution of glucose transporters in oxidative and glycolytic muscles or to an impaired translocation of GLUT4 from intracellular vesicles into the plasma membrane of muscle cells. The GLUT1 and GLUT4 distribution patterns of adult ruminants are still unknown. In calves and goat kids, Hocquette et al. (31) observed a distribution pattern inverse to that in rats (24, 28, 38, 45, 46, 53), whose oxidative muscles contained lower GLUT4 levels than glycolytic muscles. In humans suffering from insulin-independent diabetes mellitus (NIDDM), muscle cell GLUT4 contents are like those of healthy persons, but the ability of insulin to provoke GLUT4 translocation into the myocyte plasma membrane is interrupted (63). Such an impaired GLUT4 translocation could be a further reason for ruminant low insulin sensitivity.
To clarify our hypotheses, we determined the GLUT1 and GLUT4 contents of oxidative and glycolytic skeletal muscles of pigs and cows and analyzed the insulin-stimulated 3-O-methylglucose uptake and GLUT4 translocation into bovine and porcine musculus semitendinosus PM.
MATERIALS AND METHODS
The animal tests included in this study were reported to the government of the district Hannover, Lower Saxony, Germany according to the German law for the prevention of cruelty to animals (reference number: 509c-42502–01A68).
GLUT1 and GLUT4 in Porcine and Bovine Skeletal Muscles
Animals and tissue sampling.
GLUT1 and GLUT4 contents were estimated in oxidative and glycolytic muscles of six male castrated fattened pigs of the breed “Pic” and in the same muscles of six Holstein-Friesian cows. Average body weight of the pigs was 121 ± 6 kg. The pigs were housed in stable boxes and given a diet composed of barley (50%), wheat (28.5%), soybean meal (18%), soybean oil (1%), and mineral premix (2.5%). The cows were in the first or second lactation and weighed 580 ± 81 kg. They were tethered in stalls and fed to satisfy the nutrient requirements of maintenance and production of 30 kg milk per day. The diet consisted of maize (47%) and grass silage (22%), concentrate (15%), soybean meal (2%), molasses (7%), tapioca (6%), and mineral-vitamin premix (1%). Pigs and cows had free access to water throughout the entire experiment. The animals were slaughtered by stunning and exsanguination (bovines were slaughtered within 6 wk after parturition). Immediately after slaughter, tissue samples of ∼5 g were removed from the musculus masseter, diaphragm, musculus longissimus dorsi, and musculus semitendinosus. Muscle samples were washed in ice-cold 0.9% sodium chloride solution to remove blood cells. Then samples were cut free of visible connecting tissue, frozen in liquid nitrogen, and stored at −80°C until analysis of GLUT contents and enzyme activities.
Preparation of myocyte crude membranes.
Muscle GLUT1 and GLUT4 contents were determined by the Western blot procedure. This method has been shown to be appropriate for the analysis of the contents of both transporters in skeletal muscle of rats (45) and bovines (1). Muscle crude membranes were prepared according to the method of Gumà et al. (26). About 100 mg of muscle tissue were homogenized with an ultra-turrax for 2 × 5 s in 2.5 ml ice-cold homogenization buffer [20 mM HEPES, 250 mM sucrose, 5 mM sodium azide (NaN3), 1 μM leupeptin, 1 μM antipain, 2 μg/ml aprotinin, and 100 mM phenylmethylsulfonyl fluoride (PMSF) at pH 7.4]. Homogenate was centrifuged at 1,200 g and 4°C for 10 min, and the supernatant (SN1) was retained. The pellet (P1) was resuspended in 2.5 ml of homogenization buffer, rehomogenized, and centrifuged in the same way. The new pellet (P2) was discarded, and the supernatant (SN2) was pooled with SN1 and centrifuged in a Sorvall RC-5B high-speed centrifuge at 9,000 g and 4°C for 10 min to remove mitochondria. The pellet (P3) was discarded, and the supernatant (SN3) was centrifuged at 190,000 g and 4°C in a Beckmann Coulter Optima XL-100K ultracentrifuge. The resulting pellet P4 (crude membranes) was collected and resuspended in 500 μl of ice-cold homogenization buffer. Protein concentrations of the crude membrane preparations were determined using the protein assay according to Bradford (9) with BSA as a standard. Crude membrane proteins were precipitated with 10-mM TCA for 30 min at 4°C. Then crude membrane proteins were resuspended in Laemmli buffer containing 63 mM Tris·HCl, pH 6.8, 10% (vol/vol) glycerol, 2% (wt/vol) SDS, 5% (vol/vol) 2-mercaptoethanol, and 30 mM Bromphenol blue (42) to a final dilution of 30 μg/15 μl.
Electrophoresis and immunoblotting of membranes.
SDS-polyacrylamide gel electrophoresis was performed on crude membrane protein. Proteins were transferred to a Hybond nitrocellulose membrane in buffer consisting of 25 mM Tris, 192 mM glycine, and 10% methanol (vol/vol), pH 8.6. After transfer, the membranes were blocked with 10% (wt/vol) nonfat dry milk in Tris-buffered saline solution (20 mM Tris base, 137 mM sodium chloride [NaCl], 0.05% [wt/vol] Tween 20, pH 7.6) for 1 h at room temperature and then incubated overnight at 4°C with antibodies directed against 12-amino-acid carboxy terminus of rat GLUT4 (1:750), and rat whole GLUT1 protein (1:200) diluted in Tris-buffered saline solution, as described above, containing 5% (wt/vol) nonfat dry milk and 0.05% NaN3. The immune complex was detected using an ECL chemiluminescence system. The resulting autoradiograms were quantified by scanning densitometry (ScionImage, Scion, Frederick, MD). Resulting optical densities (OD) of the GLUT-specific bands were used to calculate the percent distribution of GLUT1 and GLUT4 among the four muscles of each animal analyzed here. The ODs of the GLUT-specific bands in the four muscles of each animal were summarized and equated with 100%. Afterwards the contents of GLUT1 and GLUT4 in all four muscles were calculated as percentages. This procedure made it possible to compare GLUT1 and GLUT4 distribution patterns between pigs and cows.
To test the specificity of the GLUT4 bands, we had the synthetic peptide of the carboxy terminal sequence of bovine GLUT4 synthesized by Biogenes GmbH (Berlin, Germany) and used this peptide sequence to compete with the GLUT4 antibody. Blots of rat, porcine, and bovine musculus masseter (GLUT4) were simultaneously incubated with the GLUT4 antibody, as well as with the corresponding synthetic peptide sequence. As expected, the bands corresponding to GLUT4 (Fig. 1A) were blunted in rats, pigs, and bovines by preabsorbtion of the primary antibody with the bovine peptide (Fig. 1B). Furthermore, to test the unspecific binding of the secondary antibody (horseradish peroxidase-labeled anti rabbit IgG at a final dilution of 1:10,000), some blots were incubated only with the secondary antibody. No binding signal was detected in the corresponding autoradiograms (Fig. 1C).
Metabolic classification of muscles.
The analyzed muscles were classified as oxidative or glycolytic according to isocitric dehydrogenase (ICDH; characteristic of oxidative metabolism) and lactate dehydrogenase (LDH, characteristic of glycolytic metabolism) activities. Specific activities (IU/mg) of both enzymes were measured spectrophotometrically as described previously (10, 21).
Insulin-Stimulated Glucose Uptake in Porcine and Bovine Musculus Semitendinosus
Animals and tissue sampling.
Insulin-stimulated glucose uptake and GLUT4 translocation were assessed in porcine and bovine musculus semitendinosus explants by an in vitro ex vivo procedure. Six female fattened Deutsche Landrasse pigs with an average body weight of 80 kg were kept in groups of three animals in stable boxes with free access to water. The animals received a diet of barley (79%), soybean meal (15%), soybean oil (3%), and a mineral premix (3%). The energy content of the ratio was calculated to obtain a daily body weight gain of 750 g. Six nonpregnant, nonlactating 3.7 ± 1.6-year-old cows with body weights of 614 ± 132 kg were tethered in single boxes with free access to water and fed with hay (74.5%), wheat straw (24.5%), and mineral-vitamin premix (1%) according to their maintenance requirements. After an overnight fast, tissue samples (∼5 g) of musculus semitendinosus were removed from the animals by an open muscle biopsy procedure. Pigs were anesthetized with azaperone (2 mg/kg body wt im) and ketamine (10 mg/kg body wt im) combined with an epidural anesthesia (0.5 ml/10 cm neck-rump length 2% Procain). Cows received an epidural anesthesia (6 ml 2% lidocaine). All animals survived the surgical treatment without any adverse effects such as lameness or wound infection. Immediately after removal, muscle explants were transferred to equilibration buffer [Krebs-Henseleit-bicarbonate (KHB) buffer containing (in mM) 18 mannitol, 2 pyruvate, 5 HEPES, as well as 0.1% BSA, pH 7.4]. Equilibration and all further incubation steps were carried out in a shaking (110 pm) water bath at 35°C (30). To avoid oxygen deficiency of samples the equilibration buffer, as well as all other solutions used in this experiment, was gassed continuously for at least 30 min with Carbogen [95% O2 : 5% CO2] before they were used in the experiments. Equilibration, muscle strip preparation, and all incubation steps were performed under Carbogen atmosphere.
The rate of glucose transport into bovine and porcine musculus semitendinosus was measured using 3-O-d-[3H]methylglucose (3-O-MG) (58). After equilibration, one part of each muscle sample was dissected into small strips of about 20 mg. To minimize muscle damage, preparation was done with an iris scissors and an iris forceps using a stereo magnifier. Because of this technique, it was possible to mobilize small muscle fiber bundles. Muscle strips were preincubated in 2.5 ml/10 mg muscle wet tissue of Carbogen-gassed glucose-free KHB buffer containing 5 mM HEPES, 20 mM mannitol, 0.1% BSA, and, in case of insulin stimulation, 0.1 or 20 mIU insulin/ml. Muscle strips were incubated without insulin to permit measurement of the basal 3-O-MG uptake. The insulin concentration of 0.1 mIU/ml reflects a plasma insulin value resulting from an intravenous glucose load of between 5.6 (55) and 11.2 (50) mmol/l in pigs and of 2.8 mmol/l (49) in cows. In humans, the 3-O-MG uptake into skeletal muscle strips after incubation with 0.1 mIU insulin/ml also correlates positively with the whole body glucose utilization during the euglycemic hyperinsulinemic clamp test (62). We conclude that the insulin concentration of 0.1 mIU/ml in the incubation buffer was suitable to determine the insulin-dependent 3-O-MG uptake under physiological conditions. To estimate the maximum insulin-induced 3-O-MG uptake in porcine and bovine musculus semitendinosus, the muscle strips were stimulated with the supraphysiological insulin concentration of 20 mIU/ml. After 30 min of preincubation, muscles were incubated for a further 20 min in 2.5 ml/10 mg muscle wet tissue of Carbogen-gassed KHB buffer containing 3-O-methylglucose (10 mM), 3-O-MG, and [1-14C]mannitol (final specific radioactivities of 37 kBq/ml and 2.8 kBq/ml, respectively), 5 mM HEPES, 10 mM mannitol, 0.1% BSA, and, in case of insulin stimulation, 0.1 or 20 mIU insulin/ml. After incubation, muscles were removed, blotted dry on a Whatman filter, and digested in 0.5 ml of 1 M KOH at 70°C for 20–30 min. Muscle digest was transferred into scintillation vials, and the tubes were rinsed once with 0.5 ml of distilled water. Scintillation fluid was added to the vials, the mixture was acidified, and radioactivity was measured. The rate of glucose transport into the muscle strips was calculated in μmol 3-O-MG × ml intracellular water (H2Oi.c.)−1 × 20 min−1 according to Cartee and Bohn (13).
Insulin-Stimulated GLUT4 Translocation in Bovine and Porcine Musculus Semitendinosus
For the determination of the GLUT4 translocation, further portions of musculus semitendinosus samples were dissected into small strips of ∼100 mg, using the same equipment and technique as described above. This minimum weight was needed throughout this incubation study to receive sufficient material for muscle membrane fractionation. Muscle strips were incubated for 30 min in 2.5 ml/10 mg muscle wet tissue of Carbogen-gassed KHB buffer containing (in mM) 5 HEPES, 10 glucose, 10 mannitol, as well as 0.1% BSA, and in case of insulin stimulation, 0.1, or 20 mIU insulin/ml. After incubation, muscles were removed, blotted dry on a Whatman filter, and freeze-clamped in liquid nitrogen.
Membrane fractionation and characterization.
The insulin-induced GLUT4 translocation was assessed by measuring musculus semitendinosus PM and CV GLUT4 contents after stimulation with the different insulin concentrations. To isolate both membrane compartments, muscle crude membranes were prepared as described above and then fractionated on a continuous sucrose gradient [10–60% (wt/vol) sucrose in 20 mM HEPES, 5 mM NaN3, 1 μM leupeptin, 1 μM antipain, 2 μg/ml aprotinin, 100 mM PMSF, and 1% BSA, pH 7.4] at 95,000 g and 4°C for 14 h. To remove BSA, 1-ml fractions of the gradient were diluted 1:10 in homogenization buffer and centrifuged at 190,000 g and 4°C for 90 min. The resulting pellets were resuspended in 300 μl of homogenization buffer, and protein concentrations of the resulting solutions were determined again by the method of Bradford (9). To identify musculus semitendinosus plasma membrane, specific activities of the sarcolemmal marker enzyme 5′-nucleotidase (5′-NT) were measured spectrophotometrically (20). Finally, PM and CV proteins were precipitated and resuspended in Laemmli buffer to a final protein concentration of 20 μg/15 μl.
Electrophoresis and immunoblotting of membranes.
SDS-polyacrylamide gel electrophoresis and Western blotting were performed as described above on PM and CV protein of each insulin incubation step. In addition to GLUT4 detection, protein contents of the α1 subunit of the Na+/K+-ATPase were determined by Western blotting (primary antibody was diluted 1:250) to verify the identification of PM and CV on the basis of specific 5′-NT activities. Western blotting results were determined and quantified as described above. For the estimation of the insulin-induced GLUT4 translocation into the PM, the percent increase in PM GLUT4 was estimated in relation to the basic values (incubation without insulin). For the comparison of PM GLUT4 amounts of both species, the percentage of PM GLUT4 of each insulin incubation step was calculated according to the formula: For the evaluation of the membrane fractionation procedure, bovine musculus semitendinosus strips were incubated without and with 20 mIU insulin/ml. Then sucrose gradient fractions 4, 5, 7, and 8 were analyzed for 5′-NT activities and their Na+/K+-ATPase and GLUT4 protein contents. The 5′NT-activities of all 10 gradient fractions of musculus semitendinosus membrane preparation were determined in porcines.
Most commonly used chemicals, porcine insulin, and the protease inhibitors antipain, leupeptin, and aprotinin were from Sigma-Aldrich (Saint Louis, MO). All chemicals were of the highest purity grade available. Western blotting reagents and materials (Hybond nitrocellulose membrane, Hyperfilm, ECL reagents, and secondary horseradish-peroxidase-linked anti-rabbit and anti-mouse IgG-antibodies) were from Amersham Biosciences Europe GmbH (Freiburg, Germany). The 3-O-MG was from NEN Life Science Products (Boston, MA). d-[1-14C]mannitol was from PerkinElmer Life Sciences (Boston, MA). Bradford reagent was from Bio-Rad (Hercules, CA). Polyclonal antibodies against the 12 C-terminal amino acid residues of rat GLUT4 and rat GLUT1 raised in rabbit were from Biotrend Chemicals GmbH (Cologne, Germany). Monoclonal antibody against α1-subunit of mouse Na+/K+-ATPase was from Alexis Biochemicals GmbH (Grünberg, Germany). Azaperone (Stresnil) was obtained from Janssen Animal Health (Neuss, Germany); ketamine (Ursotamin) was obtained from Serumwerk Bernburg (Bernburg Germany); procain (Isocain 2%) was obtained from Selectavet (Holzolling, Germany); and lidocaine (Lidocain 2%), from Vétoquinol/Chassot (Ravensburg, Germany).
Data are expressed as means ± SE. Statistical analysis was carried out using SigmaStat 3.0 (SPSS). Normal distribution of data was confirmed by the Kolmogoroff-Smirnoff test. A one-way ANOVA for repeated measures was performed to evaluate significant differences in the percentages of GLUT1 and GLUT4 in the four muscles of bovines and porcines analyzed here. When ANOVA detected significant differences of main effects, means were compared with a subsequent Tukey's test. The same statistical procedure was used to analyze the 5′-NT activity in sucrose gradient fractions to identify PM and CV. Data from the glucose uptake studies, the GLUT4 translocation, and the LDH and ICDH activities were analyzed by a two-way ANOVA for repeated measures (factors: species × insulin concentration and species × muscle, respectively) and a subsequent Tukey's test. Statistical analyses were also carried out to evaluate the muscle fractionation procedure using a two-way ANOVA for repeated measures (factors: sucrose gradient fraction × insulin concentration) and a Tukey's test. Correlations between GLUT contents and LDH and ICDH activities were calculated using the linear regression model. When data were not distributed normally an ANOVA on ranks was carried out. In all cases, the significance was set at P = 0.05.
GLUT1 and GLUT4 Contents of Porcine and Bovine Oxidative and Glycolytic Muscles
Muscle metabolic properties were characterized by the measurement of specific ICDH and LDH activities (see Table 1). In porcines and in bovines, specific activities of the oxidative enzyme ICDH were significantly (P < 0.05) higher in musculus masseter and diaphragm than in the longissimus dorsi and semitendinosus muscles. ICDH activity was significantly (P < 0.05) higher in musculus masseter than in diaphragm in bovines, but not in porcines. ICDH activity of the musculus masseter was significantly (P < 0.05) higher in cows than in pigs, but in the diaphragm, ICDH activity was significantly (P < 0.05) higher in pigs. There were no differences in glycolytic muscles with regard to specific ICDH activities. Specific LDH activities were significantly (P < 0.05) higher in musculus longissimus dorsi and semitendinosus than in musculus masseter and diaphragm of both species. No significant differences were found in either muscle group in porcines or bovines. The specific LDH activities of masseter and longissimus dorsi were significantly (P < 0.05) higher in pigs than in cows. ICDH and LDH activities correlated negatively in both species (porcines: r = −0.839; bovines: r = −0.768; P < 0.05; n = 16).
Representative autoradiograms of the GLUT1 and GLUT4 contents of the four examined muscles from pigs and from cows are shown in Fig. 2A and Fig. 3A, respectively. In porcines, GLUT1-specific bands were similar in all four muscles. GLUT4-specific bands were strong in the oxidative muscles masseter and diaphragm but weak in the glycolytic longissimus dorsi and semitendinosus. In cows, GLUT1-specific bands were strong in the glycolytic muscles longissimus dorsi and semitendinosus, and weak in the oxidative muscles masseter and diaphragm, but this was not the case in porcines. Strongest GLUT4 bands were detected in bovine musculus masseter, followed by diaphragm and both glycolytic muscles. According to these autoradiograms, porcines showed no significantly different GLUT1 contents among all four analyzed muscles (Fig. 2B). GLUT4 contents in pigs were significantly (P < 0.05) higher in masseter and diaphragm than in longissimus dorsi and semitendinosus. In cows (Fig. 3B), there were significantly (P < 0.05) higher GLUT1 contents in musculus longissimus dorsi and musculus semitendinosus than in musculus masseter and diaphragm. The distribution of GLUT4 in bovine muscles was similar to that of porcines. Significantly (P < 0.05), the highest levels were detected in masseter, followed by diaphragm and longissimus dorsi and semitendinosus.
GLUT4 contents correlated positively (P < 0.05) with ICDH activity (pigs: r = 0.801; cows: r = 0.844, n = 16), and negatively (P < 0.05) with LDH activity (pigs: r = −0.80; cows: r = −0.972, n = 16) in both species. Because of the different GLUT1 distribution in porcines and bovines, only cows showed a negative (P < 0.05) correlation between muscle GLUT1 concentration and specific ICDH activity (pigs: r = −0.129; cows: r = −0.759; n = 16) and a positive (P < 0.05) correlation between GLUT1 and LDH activity (pigs: r = 0.101; cows: r = 0.898; n = 16).
Insulin-Stimulated 3-O-MG Uptake in Porcine and Bovine Musculus Semitendinosus
The rates of basal and insulin-stimulated 3-O-MG uptake in porcine and bovine musculus semitendinosus are shown in Fig. 4. In pigs, the basal 3-O-MG uptake was 1.63 ± 0.17 μmol·g−1·20 min−1. Stimulation of muscle strips with the physiological insulin concentration of 0.1 mIU/ml did not result in a significant change in 3-O-MG uptake. Stimulation with the supraphysiological insulin concentration of 20 mIU/ml caused a significant (P < 0.05), 1.5-fold increase in 3-O-MG uptake. In bovines, the basal 3-O-MG uptake was 1.31 ± 0.15 μmol·g−1·20 min−1. After stimulation with the physiological concentration of 0.1 mIU insulin/ml, there was a 1.5-fold increase in 3-O-MG uptake, which was significant (P < 0.05). Maximum stimulation with 20 mIU insulin/ml did not further augment the 3-O-MG uptake significantly in cows. There were no significant differences between the species in 3-O-MG uptake.
Insulin-Stimulated GLUT4 Translocation in Porcine and Bovine Musculus Semitendinosus
Membrane fractionation and characterization.
Musculus semitendinosus PM and CV were identified by determination of specific 5′-NT activities in the sucrose gradient fractions. Figure 5 shows the distribution of 5′-NT activities, Na+/K+-ATPase, and GLUT4 contents in the sucrose gradient fractions 4, 5, 7, and 8 of bovine musculus semitendinosus strips after stimulation without and with 20 mIU insulin/ml. Without and with insulin, 5′NT-activities and Na+/K+-ATPase contents were significantly (P < 0.05) higher in fractions 4 and 5 than in fractions 7 and 8. Stimulation of muscle strips with 20 mIU insulin/ml lead to significantly (P < 0.05) higher GLUT4 contents in sucrose gradient fraction 4 than in fraction 7. Additionally, GLUT4 contents in fraction 4 of insulin-stimulated muscle strips were higher (P < 0.05) than those in fraction 4 of muscle strips incubated without insulin. GLUT4 contents of sucrose gradient fraction 8 were partly below the detection limit.
Figure 6 shows the 5′-NT activity of all 10 gradient fractions of 6 pigs. The highest 5′-NT activity per microgram membrane protein was detected in fraction 4; the lowest activity combined with a sufficient protein content was found in fraction 7. According to these results, fraction 4 was identified as PM and fraction 7 as CV. The 5′-NT activities of bovine and porcine musculus semitendinosus PM and CV are summarized in Table 2. These activities were ∼7-fold (pigs) and 4-fold (cows) higher in the plasma-membrane fraction than in the intracellular membranes. Enrichment was not affected by insulin treatment either in porcines or in bovines. To verify the membrane characterization by 5′-NT activities, the membrane Na+/K+-ATPase contents were analyzed by Western blotting. Representative autoradiograms of Na+/K+-ATPase concentrations of porcine and bovine PM and CV are shown in Fig. 7.
Figure 7 also presents typical Western blots of GLUT4 in PM and CV without and after insulin stimulation. In porcines (Fig. 7A), incubation of musculus semitendinosus strips with 20 mIU insulin/ml resulted in a 32% increase in GLUT4 in PM and a 37% decrease in GLUT4 in CV compared with the basal situation (without insulin). Muscle strip incubation with 0.1 mIU insulin/ml did not significantly alter GLUT4 in PM or in CV. In bovine musculus semitendinosus (Fig. 7B), GLUT4 in PM was 23% higher in the muscle strips that had been stimulated with 20 mIU insulin/ml, whereas GLUT4 in CV was 23% lower than under basal conditions. Stimulation of muscle strips with 0.1 mIU insulin/ml in bovines also had no effect on GLUT4 in PM or in CV. Figure 8 gives the percentage of GLUT4 in porcine and bovine musculus semitendinosus plasma membranes. In pigs, 61.00 ± 4.17% of the GLUT4 in the combined PM and CV content were located in the plasma membrane when muscle strips had been incubated without insulin. Stimulation of musculus semitendinosus strips with 20 mIU insulin/ml increased the percentage of GLUT4 in PM significantly (P < 0.05) up to 77.24 ± 1.54%, whereas incubation with the physiological insulin concentration of 0.1 mIU/ml had no significant effect. In bovines, insulin-free incubation resulted in 53.22 ± 2.32% GLUT4 in PM. Stimulation with 0.1 mIU insulin/ml did not significantly increase the GLUT4 content of PM. GLUT4 in PM increased significantly (P < 0.05) to 64.82 ± 2.80% only when bovine musculus semitendinosus strips had been incubated in 20 mIU insulin/ml. Supraphysiological insulin stimulation resulted in a significantly (P < 0.05) higher percentage of GLUT4 in the PM of musculus semitendinosus in porcines than in bovines.
The objective of this study was to elucidate possible cellular mechanisms contributing to the impaired in vivo insulin sensitivity of adult ruminants compared with monogastric omnivores by determining the GLUT1 and GLUT4 contents of oxidative and glycolytic muscles of cows and pigs and insulin-stimulated 3-O-MG uptake and the GLUT4 translocation into the myocyte plasma membrane of bovine and porcine musculus semitendinosus. GLUT1 levels of glycolytic muscles were significantly higher than those of oxidative muscles in bovines, but not in porcines. On the other hand, insulin-induced maximum GLUT4 translocation into the bovine musculus semitendinosus PM was significantly lower than in porcines.
Distribution of GLUT1 and GLUT4 in skeletal muscle.
In rats, GLUT1 and GLUT4 are expressed in different amounts in red oxidative and white glycolytic muscle (24, 28, 38, 45, 46, 53). In the present study, musculus masseter and diaphragm represent oxidative muscles, and musculus longissimus dorsi and semitendinosus glycolytic muscles. Muscles were metabolically characterized using oxidative (ICDH) and glycolytic (LDH) marker enzymes. Several studies in rodents have demonstrated a closer link of muscle GLUT4 contents to the oxidative capacity than to the fiber-type composition (for review see Ref. 16). LDH has been used as a marker enzyme of glycolytic muscles, and ICDH as a marker of oxidative muscle metabolism in goats and calves (31). In our experiments in pigs and in cows, ICDH activities were significantly higher in musculus masseter and diaphragm than in musculus longissimus dorsi and semitendinosus, while LDH activities showed the inverse distribution (Table 1). These data are in agreement with findings in calves (31) and, at least with regard to LDH activities, in fattened pigs (56). In addition, the high LDH activities in bovine musculus longissimus dorsi and semitendinosus correspond to the high amounts of α-white fibers of these muscles (40). A higher glycolytic metabolism of porcine skeletal muscle than in bovines is indicated by the higher ICDH activity in bovine than in porcine musculus masseter and by the higher LDH activity in porcine musculus masseter and musculus longissimus dorsi than in bovines.
Numerous studies on rats have shown higher GLUT1 and higher GLUT4 levels in oxidative muscles than in glycolytic muscles (24, 28, 38, 45, 46, 53). These findings correspond to those of our study only for the GLUT4 contents of skeletal muscles in both species. In our investigations, oxidative muscles contained ∼2.4 times more GLUT4 than glycolytic muscles in porcines, and 3.0 times more in bovines. Findings from earlier studies on the GLUT4 contents in ruminants are inconsistent. Hocquette and colleagues (31) detected significantly higher GLUT4 contents in musculus semitendinosus than in musculus masseter and diaphragm of calves and adult goats. On the other hand, the same group identified higher GLUT4 mRNA levels in oxidative muscles than in glycolytic muscles (32). Our results agree with those of Abe et al. (3) for GLUT4 protein and mRNA levels in cattle skeletal muscles. Little is known about the GLUT1 distribution pattern in skeletal muscles of pigs and cows. Because it is generally assumed that GLUT1 accounts for basal muscle glucose demands, the GLUT1 distribution patterns of the four analyzed muscles of both species are partly unexpected. Unlike the findings in rats (24, 38, 45), no different GLUT1 levels in oxidative and glycolytic muscles were seen in pigs in our studies, although porcines, like rats, are typical monogastric omnivores, according to their nutrient strategies. The finding of different GLUT1 contents in oxidative and glycolytic muscles of rats and pigs may be due to the different muscles investigated. Even more surprising are the significantly higher GLUT1 contents in bovine glycolytic muscles than in oxidative muscles. As a consequence, the GLUT1 contents of the myocytes and LDH activities correlated positively in cows but not in pigs. We do not know whether lactation could have caused these differences between cows and pigs in our experiments. An insulin deficiency with regard to glucose utilization has been reported in lactating goats (17), which is probably caused by a 20 to 30% decrease in muscle GLUT4 protein contents (5). GLUT1 gene expression is high in the mammary gland of cows during lactation, and GLUT4 is not detectable in the udder of lactating cows (61). GLUT1 expression in epithelial cells of mammary glands of rats and mice is also increased during lactation (12, 44, 48). As early as 1990, Burnol et al. (11) concluded that only GLUT1 is present in rat mammary gland, while GLUT4 is absent. A higher GLUT1 and a lower GLUT4 level in glycolytic muscles of lactating cows could be an effective mechanism for maintaining a basic glucose supply, which prevents insulin-sensitive excessive glucose transport into these muscle cells, thus facilitating a preferential insulin-independent glucose supply for the mammary gland. It would be worthwhile to test the hypothesis of such an interaction between lactation and glycolytic muscle in further experiments on cows.
Data from the present study strongly indicate that GLUT1 is the predominant glucose transporter of bovine glycolytic skeletal muscle. Consequently, we may expect a less insulin-dependent 3-O-MG uptake in bovine glycolytic muscles than in porcines.
Insulin stimulated 3-O-MG uptake into skeletal muscles.
In cows, 3-O-MG uptake into musculus semitendinosus increased significantly when muscle strips had been stimulated with 0.1 mIU insulin/ml, but supraphysiological insulin stimulation with 20 mIU/ml had no further effects. The maximum insulin-induced increase (1.8-fold) in the uptake of 3-O-MG into bovine musculus semitendinosus was in the range of that previously described for 2-deoxyglucose uptake into musculus rectus abdominis of calves (1.8-fold) (31). 3-O-MG uptake into porcine musculus semitendinosus strips was enhanced only after incubation with the high concentration of 20 mIU insulin/ml. In previous studies the insulin-induced 3-O-MG uptake in rat musculus epitrochlearis increased between 1.6-fold and 3.1-fold when muscles had been incubated with 0.1 mIU insulin/ml for 20 min, whereas supramaximal insulin stimulation (20 mIU/ml) yielded an increase between 2.8- and 3.8-fold (13, 27). In human musculus quadriceps femoris, there was a 1.8-fold increase in 3-O-MG uptake after 1 h of incubation with 0.1 mIU insulin/ml (30). The relatively small increments of 3-O-MG uptake in cows and pigs compared with humans and rats could possibly be due to the high glycolytic activity of musculus semitendinosus compared with musculus quadriceps femoris (humans) and musculus epitrochlearis (rats). However, this explanation does not sufficiently account for the small enhancement of porcine 3-O-MG uptake induced by 0.1 mIU insulin/ml. This may partially be due to a relatively high basal 3-O-MG uptake. The reasons for this phenomenon are still unclear. In addition to insulin, GLUT4 translocation, and therefore glucose uptake into myocytes, is stimulated by hypoxia (14) and exercise (18, 22). In the latter case, GLUT4 vesicles seem to be recruited from a different transporter pool, and the translocation is mediated by a different signal cascade than the one activated by insulin (25). It may be that the high basal 3-O-MG uptake in porcines results from a greater contribution to the insulin-independent signaling pathway in this species. Other reasons like hypoxia or membrane damage during muscle strip preparation are unlikely because samples of cows and pigs were treated in the same way and the high basal 3-O-MG uptake in pigs corresponds to a high GLUT4 content in musculus semitendinosus PM after incubation without insulin. The similar basal and insulin-stimulated 3-O-MG uptake into musculus semitendinosus in pigs and cows could have been due to a similar insulin-regulated GLUT4 activation in pigs and cows or to a higher insulin-independent glucose transport by GLUT1 in cows.
Insulin-stimulated GLUT4 translocation into the myocyte PM.
For the determination of the insulin-stimulated GLUT4 translocation, PMs and CVs of porcine and bovine musculus semitendinosus were isolated by sucrose gradient fractionation. Both membrane fractions were identified by measuring the activity of the plasma membrane marker enzyme 5′-NT. The activity of this enzyme was four times higher in the PM of bovine and seven times higher in the PM of porcine musculus semitendinosus than in the corresponding CMs. These enrichments were in the range of previous works on humans (37) and rodents (41). This difference was confirmed by measuring protein contents of the α1 subunit of the Na+/K+-ATPase in PMs and CVs. This enzyme subunit is not influenced by insulin stimulation in skeletal muscles, and therefore it is a valid marker of the PM in these muscles (4, 33, 45). As described in humans (26, 57) and rats (43, 51), Na+/K+-ATPase-specific bands were much stronger in the PM than in CVs in porcines and bovines (Figs. 5 and 7). The quality of the membrane fractionation was confirmed by the finding that insulin induced a translocation of GLUT4 vesicles from the sucrose gradient fraction 7 (CV) to the fraction 4 (PM), whereas the distribution of 5′-NT activities and Na+/K+-ATPase contents were not influenced by insulin stimulation (Fig. 5).
Incubation of musculus semitendinosus strips with 20 mIU insulin/ml led to a 1.23-fold increase in PM GLUT4 content in bovines and to a 1.32-fold increase in porcines. These enhancements are somewhat lower than those reported for rat hind limb muscle (1.5- to 2.5-fold) (18, 29) and for human musculus vastus lateralis (1.6-fold) (14, 26) and, especially in porcines, are probably due to the high basal PM GLUT4 levels. In porcines, these high PM GLUT4 contents correspond well to the high 3-O-MG uptake after incubation without insulin, and they therefore may indicate an insulin-independent GLUT4 recruitment in porcines. The maximum GLUT4 translocation was lower in bovines than in porcines (Fig. 8). As influences of sex, pregnancy, and lactation of the animals used in this experiment can be excluded, this impaired GLUT4 translocation in bovine glycolytic skeletal muscles may be a species-specific characteristic.
Two remarkable results have to be discussed concerning the relationship between glucose transport and GLUT4 translocation. On the one hand, insulin-induced GLUT4 translocation and 3-O-MG uptake developed in parallel, but, as has been reported in humans and rodents (for a review, see Ref. 19), the insulin-stimulated GLUT4 translocation slightly underestimated the insulin-activated 3-O-MG uptake in bovines as well as in porcines (with 20 mIU insulin/ml; in porcines there was a 1.5-fold increase in 3-O-MG uptake and a 1.32-fold increase in GLUT4 in PM; in bovines, there was a 1.8-fold increase in 3-O-MG uptake and a 1.23-fold increase in GLUT4 in PM). It appears unlikely that this difference results from the lower weight of muscle strips used for the insulin stimulated 3-O-MG uptake compared with the strips analyzed for the GLUT4 translocation. Samples were incubated in all incubation studies in a volume of 2.5 ml per 10 mg of muscle wet weight; that is, identical incubation buffer to muscle strip weight ratios were kept throughout both incubation studies. The maximum insulin-induced 3-O-MG uptake was similar in bovines and porcines, although the maximum insulin-stimulated GLUT4 translocation was lower in cows than in pigs. A possible explanation for these findings is that the activity of insulin not only increased the GLUT4 translocation into myocyte PM but also stimulated the intrinsic activity of GLUT4, that is, the glucose-transporting capacity of GLUT4 (18, 23, 39). This would account for the quantitative difference between glucose uptake and GLUT4 translocation in both species. An explanation for the similar 3-O-MG uptake into bovine and porcine musculus semitendinosus, although GLUT4 translocation was lower in cows than in pigs, is that the bovine insulin-independent glucose transport via GLUT1 may compensate for the smaller insulin-induced GLUT4 translocation. Because of the high GLUT1 levels in bovine glycolytic muscles, this seems likely.
It has been speculated that in ruminants the insulin-independent glucose utilization is of greater importance than in monogastric omnivores, and it may compensate at least in part for the impaired in vivo insulin sensitivity in adult ruminants (6, 15, 17, 36). Details of the cellular mechanisms that contribute to this impaired insulin sensitivity, as well as those of the higher compensatory insulin-independent glucose uptake in ruminants, are still unknown. Taking all of our results into account, it appears that the lower glucose utilization of adult ruminants, demonstrated by euglycemic, hyperinsulinemic clamp tests (36), is at least partially due to a smaller effect of insulin on GLUT4 translocation. The substantial importance of the insulin-independent glucose metabolism in ruminants may be an adaptation to the nutritional peculiarity of these animals. In ruminants, glucose availability from food is low, since glucose is fermented in the forestomach. The bulk source of glucose that the ruminants need originates from gluconeogenic processes (52). As a consequence, plasma glucose levels, their fluctuations, and glucose-induced pancreatic insulin secretion are lower (34) than in monogastric animals (60). The predominance of GLUT1, and thus an insulin-independent glucose uptake into glycolytic muscles observed in the present study, may contribute to this peculiarity. No doubt, insulin sensitivity in lactating and nonlactating ruminants and an increased GLUT1 content and an impaired GLUT4 translocation are attractive and highly interesting questions and a challenge in future studies. It would be of interest to understand the reasons for the impaired ability of insulin to stimulate GLUT4 translocation into ruminant myocyte PMs, particularly, with regard to the density of insulin receptors in the PM and to the postinsulin receptor signal transduction.
This work was supported by the Deutsche Forschungsgemeinschaft (DFG) grant SA-160/17–1.
The authors wish to thank Prof. Dr. K.-H. Waldmann, Clinic for Pigs, Small Ruminants, Forensic Medicine and Ambulatory Service, and Prof. Dr. J. Rehage, at the Clinic for Cattle (Foundation University of Veterinary Medicine Hannover, Hannover, Germany) and Dr. J. Voigt, Department of Physiological Nutrition “Oskar Kellner”, Research Institute for the Biology of Farm Animals, Dummerstorf, Germany, for providing access to the experimental animals. We are also grateful to Dr. J. McAlister-Hermann for the careful English proof reading.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society