evidence from transgenic mice that glucose transport is rate-limiting

Communication Vol. 268, No. 22, Issue of August 5 , pp. 16113-16115, 1993 THE JWRNAI. OF B1owur.u CHEMISTRY

0 1993 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

Evidence from Transgenic Mice That Glucose Transport Is Rate-limiting for Glycogen Deposition and Glycolysis in Skeletal Muscle*

(Received for publication, May 18, 1993, and in revised form, June 4, 1993)

Jian-Ming RenSI, Bess Adkins Marshallll 11, Eric A. GulveS, Jiaping GaoS, Daniel W. Johnsonn, John 0. Holloszyt, and Mike Mueckled** From the Departments of $Medicine and TlCell Biology and Physiology, Washington University Medical School, St. Louis, Missouri 63110

Aline of transgenic mice was constructed in which the human Glutl glucose transporter is overexpressed in skeletal muscle. Overexpression of Glutl protein was evident in epitrochlearis, extensor digitorurn longus (EDL), and quadriceps muscles, and resulted in 6.6-7.4- fold elevations in basal glucose transport activity as measured in isolated muscles in vitro. The elevated glucose transporter activity in the skeletal muscles of transgenic mice was associated with a 10-fold increase in glycogen concentration in EDL and quadriceps muscles that was not due to an increase in muscle glycogen synthase activity or a decrease in glycogen phosphorylase activity. The increased glucose transport activity also resulted in a 2-fold increase in muscle lactate concentration, with no increase in muscle glucose 6-phosphate. Despite a slight (10%) increase in muscle hexokinase activity, there was a 4-fold increase in total muscle free glucose in transgenic mice, indicating that hexokinase becomes rate-limiting for glucose uptake when the rate of glucose transport is very high. These results demonstrate that the muscle glycogen content can be dra- matically elevated by increasing the muscle Glutl protein level and that glucose transport is a rate-limiting step for muscle glucose disposal in normal, resting mice.

Skeletal muscle comprises 40% of the body mass and is the most important tissue for whole body glucose disposal in hu- mans (1, 2). Muscle cells express two glucose transporter iso- forms, Glutl and Glut4 (for reviews, see Refs. 3-5). Glut l re- sides constitutively in the plasma membrane and is believed to be responsible for basal glucose transport in muscle (4). We have constructed a line of transgenic mice in which human

Grants DK38495 (to M. M.) and DK18986 (to J. 0. H.), the Juvenile * This work was supported in part by National Institutes of Health

Diabetes Foundation International, and the Diabetes Research and Training Center at Washington University Medical School. The costs of publication of this article were defrayed in part by the payment of page

in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. charges. This article must therefore be hereby marked "advertisement"

5 Supported by National Institutes of Health Postdoctoral Training Grant AG00078.

11 Supported by National Institutes of Health Postdoctoral Training Grant DK07120.

** To whom correspondence and reprint requests should be addressed.

Glu t l is overexpressed in skeletal muscle.' The fasting and fed plasma glucose levels are lower, but insulin and glucagon levels are unaltered in the transgenic animals relative to control littermates. The transgenic mice also display a marked increase in the rate of glucose clearance following an oral glucose load. The present study was undertaken in an attempt to under- stand the cellular mechanisms by which alteration of muscle Glutl expression affects whole body glucose homeostasis. Our data show that increased glucose flux into muscle resulting from overexpression of Glutl increases the lactate and glycogen concentrations in muscle. These changes could not be attrib- uted to alterations in the activities of other enzymes involved in cellular glucose metabolism. These results indicate that glucose transport is a rate-limiting step for muscle glucose disposal in normal, fed mice.

EXPERIMENTAL PROCEDURES Materials-3-O-[3H1Methylglucose and [14Clmannitol were pur-

chased from Du Pont-New England Nuclear. '251-Labeled goat anti- rabbit IgG was from Amersham Corp. Glucose transporter (Glutl) antibody was raised against a synthesized peptide corresponding to the C-terminal sequence of human Glutl protein (7). Reagents for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGEY were of electrophoresis grade and were obtained from Bio-Rad. All other chemicals were purchased from Sigma.

Animals and Experimental Design-The construction of transgenic mice overexpressing the human Glutl glucose transporter was described previously.' All animals used for experiments were littermates resulting from the breeding of a single line of Glutl overexpressing mice (GlutlbP carrying a single copy of the transgenic locus with non-transgenic C57BU6 x SJL F, mice, Litters thus consisted of a -50:50 ratio of control and transgenic mice. Animals were housed in a room main- tained at 23 "C with a fixed 12-h lighffdark cycle and given free access to Purina chow and water ad libitum. All experiments were performed in the morning (10 a.m.). Fed mice were anesthetized with an injection of sodium pentobarbital (5 mg/100 g of body weight). Intact epitrochlearis and extensor digitorum longus muscles were excised and incubated in vitro for determination of glucose transport activity (8). For the measurements of Glutl protein, muscle metabolites, and enzyme activity, muscles were clamp frozen immediately after dissection. All experiments were conducted independently on at least two separate batches of littermates, except the glycogen synthase and glycogen phosphorylase measurements, which were performed on one batch of mice.

Muscle Incubation-Muscles were incubated for 30 min in 2 ml of oxygenated Krebs-Henseleit bicarbonate buffer ( M B (9)) containing 8 mM glucose and 32 mM mannitol. The flasks were shaken in a Dubnoff incubator at 35 "C. The gas phase in the flasks was 95% 065% CO,.

Determination of Glucose Zkansport Activity-After the initial incubation period, the muscles were rinsed in the absence of glucose for 10 min at 29 "C in 2 ml of oxygenated KHB containing 40 mM mannitol. Glucose transport activity was then measured using either the non- metabolizable glucose analogue 3-O-methyl-~-ghcose (3-MG) or 2-deoxy-o-glucose (2-DG) as described previously (8, 10). For measurement of 3-MG transport, the muscles were incubated in 1.5 ml of KHB containing 32 mM [14Clmannitol (3.9 pCilmmol) and 8 mM 3-U-[3H]methylglucose (300 pCilmmo1) for 7 min at 29 "C. For measurement of 2-DG transport, the muscles were incubated in 1.5 ml of KHB containing 39 mM P4C1mannitol (3.9 pCi/mmol) and 1 mM 2-deo~y-[l,2-~H]glucose (1.7 mCilmmo1) for 20 min at 29 "C. The gas phase in all flasks during both the rinse and incubation periods was 95% 02/5% CO,. The muscles were then processed, and the extracellular space and intracel-

Marshall, B. A., Ren, J.-M., Johnson, D. W., Gibbs, E. M., Lillquist, J. S., Soeller, W. C., Holloszy, J . O., and Mueckler, M. (1993) J. Biol. Chem., in press.

The abbreviations used are: PAGE, polyacrylamide gel electrophoresis; KHB, Krebs-Henseleit bicarbonate buffer; 3-MG, 3-O-methyl-~-glucose; 2-DG. 2-deoxy-~-glucose; EDL, extensor dlgitorum longus.

16113

16114 Muscle Glucose Metabolism in Dansgenic Mice lular 3-MG and 2-DG concentrations were determined as described previously (8). 3-MG transport is expressed as micromoles of 3-MG.ml of intracellular water-l.7 rnin". 2-DG transport is expressed as micromoles of 2-DG.ml of intracellular water-l.20 min-'.

Quantitation of GLUT1 Protein Concentration-The method for measurement of GLUTl immunoreactivity was similar to that previously described for measurement of Glut4 (10). Briefly, an aliquot of muscle homogenate containing 50 pg of protein was solubilized in Laemmli sample buffer, subjected to SDS-PAGE, and then electrophoretically transferred to a nitrocellulose filter. Glucose transporter protein was detected with the polyclonal antibody F350, which is specific for the GLUTl isoform, followed by 9 - l abe led goat anti-rabbit IgG.

Measurement of Muscle Metabolites-Frozen quadriceps and EDL muscles were weighed and homogenized in perchloric acid (11). An aliquot of the HC104 homogenate was used for measurement of glycogen with the amyloglucosidase method (12). The remaining homogenate was centrifuged a t 3,000 x g for 10 min a t 4 "C. The supernatant was neutralized with 2 N KOH containing 0.4 M imidazole and 0.4 M KC1 and centrifuged at 4 "C in a microcentrifuge. The neutralized supernatant was used for measurement of lactate, glucose, and glucose 6-phosphate (13).

Determination of Muscle Enzyme Actiuities-For determination of glycogen synthase activity, muscle was homogenized a t -10 "C in 1 ml of 30 mM /3-glycerophosphate, pH 7.0, containing 60% glycerol, 50 nm NaF, and 5 mM EDTA. Glycogen synthase activity was measured in the direction of glycogen synthesis a t 30 "C by the method of Kornfeld and Brown (14). Glycogen synthase activity measured in the presence of 5 mM glucose 6-phosphate is called total activity, while activity measured in the absence of added glucose 6-phosphate is defined as synthase I activity. For determination of glycogen phosphorylase activity, muscle was homogenized at -10 "C in 0.2 ml of 100 mM Tris-HC1 (pH 7.5) containing 60% glycerol, 50 mM potassium fluoride, and 10 nm EDTA. Homogenates were then diluted with 0.8 ml of the same buffer without glycerol and homogenized further at 0 "C. Phosphorylase activity was measured in the direction of glycogen breakdown at 30 "C as described previously (15). Phosphorylase measured in the presence of 3 nm AMP is called total activity, whereas activity measured in the absence of added AMP is defined as phosphorylase a activity. For measurement of hexokinase activity, quadriceps muscle was homogenized in 175 nm KCl, 10 nm GSH, and 2 m EDTA (pH 7.4). The homogenate was centrifuged at 700 x g for 15 min, and the supernatant was assayed for hexokinase as described by Uyeda and Racker (16).

Statistical Analysis-Data in the text and figures are given as means S.E. For comparisons between two groups, the Student's t test was

performed. Differences were considered to be significant a t p < 0.05.

RESULTS AND DISCUSSION

Western blot analysis was performed to estimate the relative Glutl protein content in different skeletal muscles from mice carrying a single copy of the transgenic human Glutl locus and their nontransgenic littermates. As shown in Fig. 1, the amount of Glutl protein per mg of total muscle protein was greater in the transgenic mice for all of the muscle types examined. An accurate -fold increase could not be determined because of the

-36.5

Control Transgenic FIG. 1. Relative levels of Glutl protein in skeletal muscles of

normal and transgenic littermates. Fifty micrograms of total protein homogenate from each muscle were subjected to SDS-PAGE using a 10% resolving gel and then immunoblotted with a rabbit polyclonal antibody directed against the C terminus of Glut l as described under "Experimental Procedures." The mobilities of molecular mass stan- dards are indicated. The sharp bands at -38 kDa in each lane represent an abundant unidentified muscle protein that is detected under the conditions used for this Western blot. Quad., quadriceps; Epi., epitrochlearis.

very low level of expression of endogenous Glutl in the control mice. Muscle Glut4 protein levels are unaltered in the Glutl- overexpressing mice.'

Glucose transport activity was initially measured in isolated epitrochlearis muscle using the non-metabolizable glucose analogue, 3-MG. 3-MG transport activity measured in vitro was 0.69 2 0.23 pmol/ml/7 min in control mice and was 7-fold elevated in the muscles from transgenic littermates. The rate of 3-MG transport in skeletal muscle becomes nonlinear when the intracellular concentration of 3-MG attains -25% of the extracellular concentration (8). Because of this limitation, 2-DG (1 mM) uptake was also measured to estimate the very high glucose transport activity in muscles of the transgenic mice. 2-DG is phosphorylated by hexokinase and, therefore, trapped after entering cells. Because the product 2-DG phosphate is a very weak inhibitor of hexokinase, the rate of intracellular 2-DG accumulation under our conditions is linear for at least 30 min in the muscles of the transgenic mice (data not shown). The rate of 2-DG uptake was markedly elevated in the epitrochlearis and EDL muscles of the transgenic mice, being 7.4-fold higher in epitrochlearis and 6.6-fold higher in the EDL of the transgenic as compared with the control mice (Table I). These results demonstrate that the human Glutl protein is function- ally active in at least two different skeletal muscle types in the transgenic mice, confirming and extending our previous obser- vations. '

To determine the role of the glucose transport process in overall muscle glucose metabolism, we measured glycogen and lactate concentrations in clamp-frozen quadriceps and EDL muscles from fed mice. As shown in Fig. 2, the muscle glycogen concentration was 10-fold higher in both the quadriceps and the EDL muscles of the transgenic mice. These levels are much greater than that seen under normal physiologic conditions. For example, muscle glycogen concentrations in the transgenic mice are far greater than those measured in conditions of muscle glycogen supercompensation after recovery from glycogen- depleting exercise (11). The high muscle glycogen concentration in transgenic mice was not due to an activation of glycogen synthase, as the active (I) form of glycogen synthase in muscles of the transgenic mice was only 50% as high as in the control muscles (Table 11). In fact, total (I + D) glycogen synthase activity was also lower in the transgenic mice, perhaps in re- sponse to the excessive accumulation of glycogen. Note that the increased glycogen accumulation was similar in both transgenic muscle types, despite the greater level of expression of human Glutl in EDL uersus quadriceps (see Fig. 1). This ob- servation suggests that some step beyond transport becomes rate-limiting for glycogen synthesis when the glucose transporter level is increased to the level seen in the quadriceps of the transgenic animals.

We next addressed the question of whether allosteric activation of glycogen synthase by glucose-6-P was necessary for the

TABLE I 2-Deoxyglucose uptake in isolated muscles of control and

transgenic mice Epitrochlearis and EDL muscles were incubated a t 35 "C for 30 min

in KHB containing 8 nm glucose and 32 nm mannitol, in the absence of insulin. Muscles were then incubated at 29 "C for 10 min in glucose-free medium containing 40 mM mannitol. Glucose transport activity was assessed by measuring the intracellular accumulation of 2-deoxyglucose at 39 "C (see "Experimental Procedures"). Results are expressed as micromoles of 2-deoxyglucose taken up per milliliter of intracellular water in 20 min. Values are means = S.E. for 5-6 muscles.

Control Transgenic

Epitrochlearis' 0.79 f 0.12 5.86 2 0.74" EDL 0.60 f 0.02 3.95 f 0.26"

"p < 0.05, control versus transgenic.

Muscle Glucose Metabolism i n Transgenic Mice

Control Tran~penlc Control Trsnsoenic

FIG. 2. Muscle glycogen concentration. Quadriceps and EDL muscles from six control and six transgenic mice were excised, weighed, and homogenized in 3 M HClO,. An aliquot of the homogenate was assayed for glycogen with amyloglucosidase as described under “Exper- imental Procedures.” Values are expressed as means 2 S.E. in pmoVg of wet muscle. **, p < 0.001 (unpaired t test).

TABLE I1 Muscle glycogen synthase and phosphorylase activities i n control and

transgenic mice Quadriceps and EDL muscles were excised, weighed, and homoge-

nized in 100 mM Tris-HC1 buffer. Glycogen synthase activity was measured in quadriceps muscle in the direction of glycogen synthesis at 30 “C (see “Experimental Procedures”). Glycogen synthase activity measured in the presence of 5 mM glucose 6-phosphate is termed total activity (I + D), while activity measured in the absence of added glucose 6-phosphate is termed active form (I). Glycogen phosphorylase was measured in EDL muscle in the direction of glycogen breakdown at 30 “C (see “Experimental Procedures”). Phosphorylase activity measured in the presence of 3 mM AMP is termed total activity (a + b ) , whereas activity measured in the absence of added AMP is termed active form ( a ) . The ratio of the active form over the total activity approximates enzyme activity under physiological conditions. Values are means f S.E. of 6 muscles and are expressed as pmoVg/min.

Control Transgenic

Glycogen synthase I 0.70 2 0.11 I + D

0.33 f 0.04” 2.88 5 0.44

VI + D (%I 2.28 2 0.29”

24.4 f 1.5 14.4 2 0.6”

a 6.8 2 0.7 7.1 0.6 b a l a + b (9%)

25.0 2 2.3 22.8 2 3.1 27.4 2 1.9 32.9 f 3.0

Glycogen phosphorylase

“ p < 0.05, control versus transgenic.

TABLE I11 Muscle metabolite concentrations in control and transgenic mice

Quadriceps muscle was excised, weighed, and homogenized in per- chlorate. The homogenate was centrifuged at 3,000 x g for 10 min at 4 “C. The supernatant was neutralized with 2 N KOH containing 0.4 M imidazole and 0.4 M KC1 and centrifuged at 4 “C in a microcentrifuge. The neutralized supernatant was assayed for lactate, glucose, and glucose 6-phosphate (see “Experimental Procedures”). Values are means S.E. for 9-13 muscles, and expressed as pmoVg wet muscle.

Control Transgenic

Glucose 6-phosphate Lactate Glucose

1.3 2 0.3 1.4 f 0.3 6.2 2 0.9 14.5 2 1.4”

1.14 f 0.25 4.88 f 0.55“

“ p < 0.001, control versus transgenic.

increased glycogen deposition. As shown in Table 111, the level of Glc-6-P in quadriceps muscle was not significantly different between the transgenic mice and controls. These data indicate that the normal basal glycogen synthase activity was sufficient to permit a massive accumulation of muscle glycogen in the transgenic mice.

To evaluate the possibility that the higher muscle glycogen content observed in transgenic mice might be due to a reduction in glycogen phosphorylase activity, phosphorylase activity was measured in the EDL muscle. Neither the proportion of the active phosphorylase a form nor the total activity (a + b ) was altered in transgenic mice (Table 11). These results indicate that the increased muscle glycogen accumulation in the transgenic mice was entirely a consequence of the increase in glucose transport activity.

16115

The primary fates of glucose entering muscle are storage as glycogen, conversion to lactate, or complete oxidation via the Krebs cycle. I n vivo glucose clamp experiments show that the majority of glucose taken up by muscle in the resting state is converted to glycogen (2). Our data demonstrate that a large increase in basal glucose transport in muscle is associated with a massive accumulation of glycogen. This finding indicates that in normal nontransgenic mice, glucose transport is a rate-limiting step for glycogen synthesis. The finding that the lactate concentration was more than 2-fold higher in quadriceps muscle of the transgenic mice as compared with the controls (Table I) suggests that the increased glucose flux into muscle associated with the elevation of Glutl protein also increased the rate of glycolysis. This interpretation is consistent with the finding that plasma lactate was also elevated in the transgenic mice.’ Thus, it appears that muscle glucose transport activity limits both the rate of glycogen synthesis and the rate of glycolysis in muscles of fed, resting mice.

There is virtually no free glucose in normal skeletal muscle because muscle hexokinase activity appears to be sufficiently high to keep pace with the rate of glucose transport under physiologic conditions (6). The free glucose measured in whole muscle homogenates generally represents the glucose in the extracellular space, which is in near equilibrium with plasma glucose. To determine whether hexokinase becomes rate-limiting for muscle glucose disposal when glucose influx is very high, we measured the free glucose concentration in quadriceps muscle. As shown in Table 111, the whole muscle free glucose concentration was 4-fold higher in transgenic mice as compared with controls. The high free glucose concentration observed in transgenic mice cannot be explained by a higher glucose concentration in the extracellular space, as the plasma glucose concentration was lower in transgenic mice than in the controls,’ or by decreased hexokinase activity, as the hexokinase activity measured in whole muscle homogenate was slightly higher in the transgenic mice than the controls (1.90 2 0.06 versus 1.65 2 0.06 pmol/midg, wet weight, p < 0.05).

Our data demonstrate that an increase in glucose transport activity due to overexpression of Glutl in skeletal muscle results in a large increase in muscle glycogen content and elevation of the muscle lactate concentration. These results indicate that in resting, fed mice glucose transport activity limits the rates of glycogen synthesis and glycolysis in skeletal muscle.

Acknowledgments-We thank May Chen, Xiang-Jing Wang, Connie Skillington, and Guofeng Zhou for their excellent technical assistance.

REFERENCES 1

2,

3

4. 5. 6. 7.

8.

9.

10.

11.

12. 13.

14. 15.

16.

Baron, A. D., Brechtel, G., Wallace, P., and Edelman, S. V. (1988) Am. J. Physiol. 255, E769-E774

Shulman, G. I., Rothman, D. L., Jue, T., Stein, P., DeFronzo, R. A., and Shul- man, R. G. (1990) N . Engl. J. Med. 322, 223-228

Burant, C. F., Sivitz, W. I., Fukumoto, H., Kayano, T., Nagamatsu, S., Seino, S., Pessin, J. E., and Bell, G. I. (1991) Recent Prog. Horm. Res. 47, 349-387

Mueckler, M. (1990) Diabetes 39,6-11

Ziel, F. H., Venkatesan, N., and Davidson, M. B. 11988) Diabetes 37, 885-890 Devaskar, S. U., and Mueckler, M. M. (1992) Pediatr. Res. 31, 1-13

Haney, P. M., Slot, J. W., Piper, R. C., James, D. E., and Mueckler, M. (1991) J.

Young, D. A., Uhl, J. J., Cartee, G. D., and Holloszy, J . 0.11986) J. Biol. Chem.

Krebs, H. A., and Henseleit, K. (1932) Hoppe-Seylerk 2. Physiol. Chem. 210,

Gulve, E. A., Henriksen, E. J., Rodnick, K. J. , Youn, J. H., and Holloszy, J. 0.

Cartee, G. D., Young, D. A., Sleeper, M. D., Zierath, J., Wallberg-Henriksson,

Passoneau, J. V., and Lauderdale, V. R. (1974) Anal. Biochem. 60, 405412 Lowry, 0. H., and Passoneau, J . V. (1972) A Flexible System of Enzymatic

Kornfeld, R., and Brown, D. H. (1962) J. Biol. Chem. 237, 1772-1777 Young, D. A., Wallberg-Henriksson, H., Crenshaw, J., Chen, M., and Holloszy,

Uyeda, K., and Racker, E. (1965) J. Biol. Chem. 240, 46824688

Cell Biol. 114, 689-699

261,16049-16053

33-66

11993) Am. J. Physiol. 264, E319-E327

H., and Holloszy, J. 0 . (1989) Am. J. Physiol. 256, E494-E499

Analysis, Academic Press, New York

J. 0 . (1985)Am. J. Physiol. 248, C406-C409

evidence from transgenic mice that glucose transport is rate-limiting

Documents