THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1993 by The American Society for Biochemistry and Molecular Biology, Inc. Vof. ZSe, No. 2, Issue of January 15, pp. 1320-1325,1993

Printed in U.S.A.

Effect of Glucose Deprivation on GLUT 1 Expression in 3T3-Ll Adipocytes”

(Received for publication, July 23, 1992)

Harvey H. Kitzman, Jr.$, Robert J. McMahonS, Martin G. Williams, and Susan C. Frost8 From the Department of Bioefaemistfy and Mo~ecular Biology, University of Florida, Gainesvitle, Florida 32610

Elevated glucose transport rates during glucose dep- rivation are phenomena that have been observed in several different types of cells in culture. We show here that glucose transport rates in 3T3-Ll adipocytes increased by 10-fold within 18 h in response to glucose deprivation, confirming earlier work by Van Putten and Krans (Van Putten, J. P. M., and Krans, H. M. J. (1985) J. Biol. Chem. 260, 7996-8001). Mannose and 3-0-methylglucose (a nonmetabolizable glucose ana- log), but not fructose or galactose, blocked the increase in transport activity. Although the increase in trans- port was dependent on new protein synthesis, only a small and transient increase in GLUT 1 mRNA (less than 2-fold) was observed. In addition, the level of the normal isoform of GLUT 1 (46 kDa) did not increase. A lower molecular mass isoform (37 kDa) was observed but not until 15 h after glucose removal, the appear- ance of which was clearly not correlated with the increase in activity. Further, the extracellular glucose Concentration required to elicit accumulation of this form (p37) was 2 orders of magnitude less than that required for transport stimulation (6 pM uersus 600 pM glucose; p37 accumulation and transport activation, respectively). Interestingly, p37 was seen in the pres- ence of galactose, but not fructose, despite elevated transport activity with either sugar. The p37 isoform was slightly larger than N-glycosidase F-treated GLUT 1 (36 kDa), implying that this form is still glycosylated, albeit incompletely. It is not known if p37 is functional, but the time- and sugar-dependent appearance of the lower isoform suggests that p37 is not responsible for st~vation-induced transport but potentially represents an underglycosylated precursor of the normal, 46-kDa isoform of GLUT 1.

Glucose is transported by a family of integral membrane glycoproteins which share a high degree of sequence and probably structural similarity. These include GLUT 1, the constitutive transporter; GLUT 2, the high K, transporter; GLUT 3, the brain transporter; GLUT 4, the insulin-sensitive transporter; and GLUT 5, the intestinal transporter (for review, see Bell et al., 1990). Expression of GLUT 1 can be

* This work was supported by National institutes of Health Grants DK39135 (to S. C. F.) and DK45035 (to S. C. F.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement’’ in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

In partial fulfillment toward the Ph.D. degree at the University of Florida.

§ To whom correspondence should be addressed: Box 100245 JHMHC, Dept. of Biochemistry and Molecular Biology, University of Florida, Gainesville, FL 32610.

modulated by hormones (Calderhead et al., 1990; Tai et al., 1990; Tordjman et al., 1989; Kozka et al., 1991; Harrison et al., 1990; Asano et al., 1989), differentiation (Kaestner et al., 1989; Reed et al., 1990), cell cycle (Kitagawa et al., 1991; Cornelius et al., 1990; Hiraki et al., 1988), transformation (Godwin and Lieberman, 1991), and nutrients (Kletzien and Perdue, 1975; Van Putten and Krans, 1985; Haspel et al., 198G; Reed et al., 1990; Walker et al., 1988, 1989; Koivisto et al., 1991; Kozka et al., 1991; Pessin et al., 1982). Its response to glucose is particularly intriguing. In all cells tested to date, glucose deprivation increases the rate of glucose transport by a slow, protein synthesis-dependent process (Kletzien and Perdue, 1975; Van Putten and Krans, 1985). Beyond this, there does not appear to be a unifying mechanism by which this enhanced activity occurs. The early work by Kletzien and Perdue (1975) in chick embryo fibroblasts showed that in addition to protein synthesis, the increase in transport rates also required a transcriptional component. This may also be true of glia cells, in which a 4-6-fold increase in GLUT 1 mRNA and protein was noted in response to deprivation (Walker et al., 1988). However, in 3T3-C2 cells, RNA extracts from glucose-deprived cells actually show a reduced ability to make new protein, in uitro, compared with control cells, despite a 10-40-fold increase in the amount of the GLUT 1 in glucose-deprived cells, the total of which included both the normal isoform and a lower molecular weight form (Haspel et al., 1986). More recently, it has been suggested that glucose deprivation blocks degradation of the GLUT 1 transporter in these cells (Ortiz et al., 1992). In LG skeletal muscle cells, which contain both GLUT 1 and GLUT 4, an increase in the mRNA levels of the GLUT 1 transporter, but not GLUT 4, was observed with chronic glucose deprivation. However, the number of both transporters increased in the plasma mem- brane (Walker et al., 1989; Koivisto et al., 1991). In 3T3-Ll adipocytes, a %fold increase in total GLUT 1 protein was observed (the normal isoform, which did not change, plus a lower molecular weight isoform) (Reed et al., 1990). However, the relationship between the increase in transport activity and protein was not investigated. Kozka et al. (1991) have demonstrated a 2-3-fold increase in GLUT 1 content (not GLUT 4) on the cell surface of glucose-deprived 3T3-Ll adipocytes, but only the normal GLUT 1 isoform was identi- fied. This differs from chick embryo ~ b r o b l a s ~ , in which two GLUT 1 isoforms were present on the plasma membrane (Pessin et al., 1982). Thus it is clear that each type of cell has its own mechanisms for dealing with the stress of glucose deprivation which should be remembered in comparing such data.

3T3-Ll a d i p o c ~ s have a high demand for glucose, as a substrate in energy production, as well as in fatty acid and triglyceride synthesis. To facilitate glucose uptake, these cells contain two transporter types: GLUT 1 and GLUT 4. Under basal conditions, glucose transport is mediated by GLUT 1.

1320

Glucose Deprivation in 3T3-Ll Adipocytes 1321

Insulin stimulates the t ransl~at ion of GLUT 4 transporters from an intracellular storage vesicle in the trans-Golgi net- work to the plasma membrane (Blok et al., 19881, a process which occurs quickly and in a protein synthesis-independent fashion (Pettengell and Frost, 1989). This contrasts with the requirement for protein synthesis in starvation-induced activ- ity as discussed above. The starvation-induced requisite may result in the synthesis of new GLUT 1, an activator of GLUT 1 or its translocation, or a new stress GLUT 1 isoform. We were particularly intrigued by the appearance of a lower molecular weight form of GLUT 1 in 3T3-Ll adipocytes (Reed et ai., 1990) as a stress isoform. Our goal was to determine its role in the up-regulation of transport activity in 3T3-Ll adipocytes. First, we found only one species of GLUT 1 mRNA, which changed little over the course of starvation. Second, we observed little change in the normal GLUT 1 isoform (p46) during glucose deprivation. The lower molecular weight isoform (p37) became apparent between 15 and 18 h of glucose deprivation. Incubation with galactose, as in glucose deprivation, resulted in the appearance of p37 in the same fashion, whereas fructose treatment prevented the appearance of p37. Whether we glucose deprived cells or substituted fructose or galactose for glucose in the medium, we observed significantly elevated transport rates within 6 h. Thus the process of transport activation is uncoupled from the appear- ance of p37, suggesting that it is not a new "stress" GLUT 1 isoform, but rather represents an underglycosylated form of GLUT 1. Potential mechanisms for activating the transporter are discussed.

EXPERIMENTAL PROCEDURES

~ u t e r ~ ~ ~ ~ - D M E M ' (430-2100 EG), calf serum (38N73091, and fetal bovine serum (51K612) were obtained from GIBCO. Glucose- free DMEM was made in a manner comparable to commercially available DMEM (Dulbecco and Freeman, 1959). Fructose, galactose, mannose, 3-O-methylg~ucose, and ra~oimmunoassay grade bovine serum albumin (A-7888) were obtained from Sigma. ['H]Deoxyglu- cose and ["PIdATP and the enhanced chemiluminescence kit were obtained from Amersham Corp. The random primers DNA labeling system (8187SA) was obtained from Bethesda Research Laboratories. N-Glycosidase F was obtained from Boehringer Mannheim. All other reagents used were of the highest quality commercially available.

Cell Culture and Glucose Starvation-Cells were grown and differ- entiated following the procedure of Frost and Lane (1985). 35-mm plates (2.1 X 10' cells) were used for glucose transport activity experiments, 60-mm plates (5 x lo6 cells) were used for protein identification, and 10-cm plates (12 X lo6 cells) were used for RNA isolation. Cells were utilized between days 8 and 12 postdifferentia- tion. Cells were fed complete medium, 10% fetal bovine serum 24 h prior to treatment. For glucose starvation studies, the cells were incubated in glucose-free DMEM containing 10% dialyzed fetal bo- vine serum for the times indicated. In the experiments involving other sugars, glucose-free DMEM was supplemented with fructose, galac- tose, or mannose at 25 mM.

Glucose Transport-Glucose transport was assayed essentially as described by Frost and Lane (1985). Briefly, cells were washed 3 X 3.0 ml with Krebs-Ringer phosphate buffer (128 mM NaCl, 4.7 mM KC], 1.25 mM MgSO,, 1.25 mM CaCl,, 5.0 mM phosphate salts, pH 7.4) a t 37 "C. Plates were ~ui l ibrated with the buffer for 10 min. [3H]Deoxyglucose was then added to the plates a t a final concentra- tion of 0.2 mM (1 pCi/pmol) for 10 min. The reaction was terminated by washing the plates 3 X 3.0 ml with phosphate-buffered saline (120 mM NaCl, 2.7 mM KCl, 10 mM phosphate salts, pH 7.4). The cells were lysed in 0.1% SDS. An aliquot was counted for radioactivity.

RNA Lolation and Northern Analysis-RNA was isolated by the guanidinium thiocyanate procedure (Church and Gilbert, 1984) as described previously by Wong et al. (1991). Briefly, an aliquot of 20 pg of total RNA was loaded onto a 1% formaldehyde-agarose gel. The gels were run for 12-16 h at 40 V with constant buffer recirculation.

The abbreviation used is: DMEM, Dulbecco's modified Eagle's medium.

After washing, the RNA was then ele~trophoretically transferred from the gel to an uncharged nylon membrane (Genescreen, Du Pont-New England Nuclear) and cross-linked with uv light for 3.5 min.

Probe Synthesis and Hybridization-The cDNAs that were used to make the GLUT 1 and GLUT 4 probes were the generous gift of Dr. Maureen Charron (Albert Einstein College of Medicine). The insert was labeled by primer extension (Feinberg and Vogelstein, 1983). Specifically, 0.5 pg of DNA was denatured by boiling at 100 "C for 5 min. After briefly cooling (3 min), it was incubated with a 6-base primer, deoxynucleotides, one of which was labeled ( ~ 3 z P J ~ T P , dCTP, dTTP, dGTP), and DNA polymerase. Free nucleotides were removed by gel chromatography over Sephadex G-50. The eluted labeled DNA was used directly to probe the RNA blot generated above. Hybridization was performed for 12-16 h in a buffer containing 1% bovine serum albumin ~ra~oimmunoassay grade), 1 mM EDTA, 0.5 M Na2HP04, pH 7.2, and 1% SDS at 60 "C. The membranes were washed under high stringency conditions (3 X 10-min washes in 1 mM EDTA, 40 mM Na2HP04, pH 7.2, and 1% SDS at 65 "c) and exposed to Kodak X-Omat AR film for various lengths of time. Relative intensity of each band was quantified by video densitometry within the linear range of the film using the BioImage Visage 110 (Millipore Biolmage, Ann Arbor, MI).

Western Analysis of GLUT Protein-Cells were lysed in 2 ml of TES (20 mM Tris, pH 7.4, 1 mM EDTA, 255 mM sucrose containing 1 mM phenyimethylsulfonyl fluoride). The celis were homogenized using 10 strokes of a motor-driven pestle. The homogenate was then centrifuged at 1,000 X g in an sS34 rotor (Sorvall) for 30 min at 4 "C. The supernatant was collected and centrifuged at 212,000 X g in a 70.1 Ti rotor in a Beckman L8-70 ultracentrifuge for 60 min at 4 'C. The supernatant was discarded, and the pellet was washed once and recollected as above. The final pellet was resuspended in 500 p1 of TES, homogenized, and frozen at -20 "C until all samples were collected, The samples were thawed at room temperature and protein determined using the method of Markwell et al. (1978). Proteins were then separated on reducing, 7.5% polyacrylamide-SDS gels as de- scribed by Laemmli (1970). 10 pg of protein was diluted with sample dilution buffer containing 1 mM phenylmethylsulfonyl fluoride. Sam- ples, along with low molecular weight markers (Bio-Rad) were then loaded onto the gels, and the gels were run for 3.5 h at 150 V. The proteins were transferred to nitrocellulose following the method of Clancy and Czech (1990). Immunodetection was performed using enhanced chemiluminescence. Briefly, the nitr~ellulose was blocked for 1.5 h in 20 mM Tris and 137 mM NaCl (TBS) containing 0.1% Tween 20 and 10% nonfat dry milk at room temperature. The nitrocellulose was then incubated with either a polyclonal antibody against GLUT 1 (generously provided by Dr. E. Michael Gibbs, Pfizer) or a commercially available antibody against the carboxyl terminus from East Acres Biochemicals at a 1:1,00O,dilution in TBS/ Tween (containing 10% milk) for 1 h. After washing, the nitrocellu- lose was then incubated with secondary antibody linked to horse- radish peroxidase at a l : l ~ , O O O dilution in TBS/Tween (containing 10% milk) for 1 h. Finally the nitrocellulose was incubated with detection reagents for 1 min. The nitrocellulose was exposed to Kodak X-Omat AR film for 1-5 min. Video densitometry was performed to measure the relative intensity of the bands in the linear range of the film.

A similar procedure was utilized for GLUT 3 detection. As a positive control, membranes from rat cerebrum (30 pg) were run in parallel. Antisera against the carboxyl-terminal sequence of GLUT 3

(NIH) and used at a dilution of 1:800. (amino acids 474-493) was generously provided by Dr. Ian Simpson

N-Glycosidase F Treatment-Membranes (10 pg) were sonicated for 30 s in TES containing 1 mM phenylmethylsulfonyl fluoride with or without N-glycosidase F (1 unit) in a total volume of 50 pl and then incubated at 37 'C for 1 h. The samples were then diluted with sample dilution buffer and loaded onto a 7.5% polyacrylamide gel as described above.

RESULTS

Van Putten and Krans (1985) have shown that glucose deprivation increases glucose transport activity in nearly lin- ear fashion over time in 3T3-Ll adipocytes. After an initial lag phase following incubation with glucose-free DMEM, we were able to confirm the increase in activity, although the rates steadied at about 24 h (Fig. 1). In the absence of glucose, other sugars had differentiai effects on this process. Both

1322 Glucose Deprivation in 3T3-Ll Adipocytes

Y 2.0 F 4

. . I V”r+Man I /*

0 6 12 18 24 30 36 42 48

HOURS OF TREATMENT

FIG. 1. Effect of chronic glucose deprivation on glucose transport in 3T3-LI adipocytes. Adipocytes in 35-mm plates were incubated in the absence of glucose or in the presence of fructose, galactose, mannose, or 3-0-methylglucose (MeClc) each a t 25 mM. Deoxyglucose transport was assayed for 10 min at the indicated times, as described under “Experimental Procedures.” The rate data are reported in nmol/106 cells/min f S.E. Duplicate assays were per- formed for each time point, and a complete set for each sugar was repeated a t least twice.

mannose and 3-0-methylglucose suppressed the increase in transport, whereas fructose and galactose treatment had no effect on the up-regulation. Recall that fructose and galactose compete poorly with glucose for transport (Ki values = 50.7 and 73.5 mM, respectively; Van Putten and Krans (1985)) and presumably enter the cell via unique transporters. In contrast, both mannose and 3-O-methylglucose, a nonmetabolizable analog of glucose, compete well for glucose uptake (Ki values = 7.6 and 2.2 mM, respectively; Van Putten and Krans (1985)) and thus enter the cell through the same transporter. These data imply that the “receptor” which elicits regulation of the transporter is stereospecific (apparently with the same spec- ificity of the transporter, itself) recognizing glucose and not one of its metabolites.

3T3-Ll adipocytes contain two transporter isoforms: GLUT 1, the constitutive transporter, and GLUT 4, the insulin-sensitive transporter (Kaestner et al., 1989). There is no immunoreactive GLUT 3 protein in either control or glucose-deprived cells (data not shown). Thus, either GLUT 1 or GLUT 4 or a combination of them might be responsible for the elevated transport rates. Earlier work demonstrated that insulin-stimulated glucose transport activity decreases over time in response to glucose deprivation (Van Putten and Krans, 1985). In fact, GLUT 4 mRNA declined to near undetectable levels (t112 = 6 h) in response to glucose depri- vation in 3T3-Ll adipocytes (Fig. 2). Although this finding is unique among the cell types examined, it suggests that an increase in GLUT 4 is not responsible for the deprivation- induced increase in transport rates. In contrast, GLUT 1 mRNA changed very little with this same treatment, reaching its highest level a t about 2 h after the onset of glucose deprivation (1.8-fold) and then returning to near control levels (Fig. 2). In addition, actinomycin D (4 p ~ ) , added when glucose was removed, had only a small effect on the increase in activity (after 24 h of treatment, 1.55 ? 0.05 uersus 1.19 & 0.02 nmol/106 cells/min, glucose-deprived uersus actinomy- cin-treated glucose-deprived cells), suggesting that the depri- vation-induced increase in transport activity had little de- pendence on transcriptional activation. Although our GLUT 1 data are similar to those observed in L6 muscle cells (Walker et al., 1989) they clearly differ from those obtained in primary cultures of rat glial cells, in which a 6-fold increase in GLUT 1 mRNA was seen in response to glucose deprivation (Walker et al., 1988). Interestingly, whereas GLUT 4 mRNA declined in 3T3-Ll adipocytes, GLUT 4 mRNA in L6 skeletal muscle cells remained constant during glucose deprivation (Koivisto

A hours 0 2 4 6 12 2 4 36 48

C F S F S F S F S F S F S F S

GLUT 1 8

G L U T 4 I)WYmul)MYa+O. u’,M E 2.0

6 e@ e”. GLUT 1 mRNA

W N

E 1.5! ‘@’@\ /@ : A-A GLUT 4 mRNA

2

5 ’‘O . ‘AA\

v) 0.5

A

W

e- @

n

W 0 A-A”---

0.0 , ALL 0 6 12 18 24 30 36 42 48

HOURS OF GLUCOSE DEPRIVATION

FIG. 2. Northern analysis of GLUT 1 and GLUT 4 mRNA during glucose deprivation. 10-cm plates of adipocytes were in- cubated in the absence or presence of 25 mM glucose, and total RNA was isolated at the indicated times. Panel A, Northern analysis, using cDNA probes for GLUT 1 and GLUT 4, was performed as described under “Experimental Procedures.” C, control; F, 25 mM glucose; S, glucose-deprived. Panel R, the GLUT bands were quantitated by video densitometry and normalized to actin mRNA levels which did not change over the course of glucose deprivation. This experiment is representative of two independent experiments.

et al., 1991). Thus, the different transporters appear to be regulated in a cell-specific manner under the same metabolic condition.

Although transcriptional regulation did not appear to play a role in transport activation in glucose-deprived 3T3-Ll adipocytes, translational control has been suggested (Van Putten and Krans, 1985). To determine if an increase in transporter protein was responsible for the increase in activ- ity, we measured the relative amount of immunoreactive GLUT 1 over glucose deprivation. As shown in Fig. 3A, we were able to identify two isoforms of GLUT 1 protein. The normal isoform (which in our hands migrated as a 46-kDa species, ~ 4 6 ~ ) changed little over the course of glucose depri- vation. A lower molecular weight GLUT 1 isoform, which migrated as a 37-kDa species (p37) appeared between 15 and 18 h after glucose withdrawal and continued to accumulate through 48 h. Thus, the total transporter pool (p46 + p37) increased by about 2-fold within 48 h, but the increase in transport activity was not correlated with either the appear- ance of p37 or the increase in the total pool. In addition (Fig. 4, A and B ) , the concentration of extracellular glucose at which a 50% increase in transport activity was observed was 2 orders of magnitude greater than that which elicited the appearance of p37 (Kso = 500 p M uersus 5 pM for transport activation and p37 accumulation, respectively). Interestingly, when the cells were incubated with either fructose or man- nose, p37 did not appear (Fig. 5, A and B). Recall that mannose but not fructose blocked the increase in transport activity in response to glucose deprivation. Surprisingly, the appearance and accumulation of p37 in galactose-treated cells were similar to those in glucose-starved cells (Fig. 5B, inset). This contrasts with the finding in 3T3-C2 fibroblasts, the

*Mobilities of integral membrane proteins, such as the glucose transporter, on SDS-polyacrylamide gel electrophoresis can vary with gel system, percentage of acrylamide, and sample preparation (Rizzolo et al., 1976).

Glucose Deprivation in 3T3-LI Adipocytes 1323

A

t imdh) 0 6 1 2 1 5 1 8 2 4 3 6 4 8

&. -37kDa

HOURS OF GLUCOSE DEPRIVATION

FIG. 3. Western ana lys i s o f GLUT 1 protein during glucose deprivation. Total membranes were isolated from cells incubated in the presence or absence of 25 mM glucose for the times indicated. Panel A, GLUT 1 was identified by immunoblotting as described under "Experimental Procedures." Since glucose deprivation had no effect on total protein levels (H. H. Kitzman, Jr. and S. C. Frost, unpublished observation), samples were loaded by constant protein. Panel B, bands were quantitated by video densitometry. The intensity of both the 46-kDa (p46) and 37-kDa (p37) bands is reported with respect to time zero. Shown is an experiment representative of a t least five independent experiments.

- p46

- p37

B

W $ W Ly

M

a".GT

60 50 40 30 a 20 \

10 0 0 .001 .01 0.1 1.0 13

GLUCOSE CONCENTRATION (mM)

FIG. 4. Effect of glucose concentration on transport act ivi ty a n d p37 accumulation. Cells were incubated for 12 (transport determination) or 24 h (p37 analysis) with refeeding every 2 h a t the appropriate glucose concentration. Panel A, total membranes were prepared and GLUT 1 identified by immunoblotting as described under "Experimental Procedures." Panel B, the p46 and p37 bands were quantitated by video densitometry. The p37 bands are reported relative to zero glucose. The p46 bands did not change relative to glucose concentration (densitometric analysis not shown). Transport activity is shown as a percent of that observed a t zero glucose (0.60 & 0.024 nmol/106 cells/min, n = 4). The data represent an average of two independent experiments. GT, glucose transport activity.

control cell line for 3T3-Ll cells, in which galactose as well as fructose blocked the appearance of p37. This difference could be explained if galactose is not transported or metabo- lized by 3T3-Ll adipocytes. However, even if transported, galactose would supply only UDP-Gal and UDP-Glc, which would be insufficient for dolichol-linked oligosaccharide bio- synthesis (Zubay, 1988). By comparison, both fructose and

A time (h) 0 6 1215 18243648

0 5 .

O O L L " " "

0 6 17 18 ? 4 30 36 S2 L8

POLRS C r TQFhTUEF.'

FIG. 5. Effect of var ious sugars on the accumulation of p37 during glucose deprivation. Total membranes were isolated from cells incubated in the absence of glucose but in the presence of galactose, fructose, or mannose a t 25 mM. Panel A, GLUT 1 was identified by immunoblotting as described under "Experimental Pro- cedures." Panel B, the p46 and p37 bands were quantitated by video densitometry. The intensity of p46 and p37 (inset) is relative to time zero for each sugar. The dotted line represents the p46 and p37 data (glucose deprivation) from Fig. 3B for comparison. Shown is an experiment which is representative of three independent experiments.

glucose + + - - N-gF - + - +

FIG. 6. N-Glycosidase F treatment of isolated membranes. Membranes (10 pg) isolated from control (+glucose, time 0 ) and glucose-starved (-glucose, time 48 h) cells were treated with or without 1 unit of N-glycosidase F ( N - g F ) as described under "Ex- perimental Procedures." Immunoblot analysis revealed three distinct forms: the normal isoform (p46), the glucose-deprived form (p37), and the aglyco form (p36). This experiment is representative of at least five independent experiments.

mannose would lead to the synthesis of UDP-Glc, UDP- GlcNAc, and GDP-mannose, which would provide appropri- ate substrates for core oligosaccharide synthesis.

To examine the extent of glycosylation of p37, we treated isolated membranes with N-glycosidase F to remove aspara- gine-linked oligosaccharide. Fig. 6 demonstrates the appear- ance of a unique immunoreactive form after glycosidase treat- ment in both control and glucose-deprived cells. This aglyco form migrated a t about 36 kDa (p36), slightly below p37. With this small change, it is important to mention that heat- inactivated glycosidase F did not alter the mobility of p37 (data not shown). These observations together with the sugar- dependent appearance of p37 (Figs. 3-5) suggest that p37 represents a form which is incompletely glycosylated.

Because the increase in transport activity during glucose deprivation is dependent on new protein synthesis (Van Put- ten and Krans, 1985), we wanted to determine if the appear- ance of p37 exhibited a similar requirement. To distinguish

1324 Glucose Deprivation in 3T3-Ll Adipocytes

- cyclo + cyclo at 12h

hours Q 3 6 9 15243648 15243648

FIG. 7. Effect of cycloheximide on the accumulation of p37. Cells were glucose deprived for the indicated times. Cycloheximide (cyclo, 20 PM) was added to some cells at 12 h after the removal of glucose and maintained through 48 h. This experiment is represent- ative of at least three independent experiments.

between the deprivation-induced increase in transport activ- ity (Fig. 1) and the appearance of p37 (Fig. 3), cycloheximide was added 12 h after removing glucose from the medium. Cycloheximide stabilizes the transport activity at the level seen a t 12 h.2 In addition, low concentrations of cycloheximide (20 p ~ ) were utilized to inhibit protein synthesis without inhibiting lysosomal proteases, which is observed a t higher concentrations (Yamada et al., 1983). As shown in Fig. 7, i t is clear that cycloheximide prevented the appearance and ac- cumulation of p37 in glucose-deprived cells with little effect on p46.

DISCUSSION

It has been known for several years that the increase in glucose transport activity in glucose-deprived cells is caused by a change in the Vmax of transport, not a change in the K,,, (Kletzien and Perdue, 1975). An increase in Vmax can be explained by an increase in the number of transporters a t the cell surface, an increase in the intrinsic activity of the trans- porters, or a combination of both. An increase in the number of transporters could result from new synthesis of the trans- porter or inhibition of the degradation of existing transport- ers. Either, or a combination of both, would lead to an increase in the amount of transporter protein. However, we have observed little change in the level of the normal GLUT 1 isoform despite a 10-fold increase in transport activity. Thus in 3T3-Ll adipocytes, protein level does not correlate with transport activity. In contrast to our data, Haspel et al. (1986) and Ortiz et al. (1992) showed that the normal GLUT 1 isoform increased by 5-fold in 3T3-C2 fibroblasts which cor- related with a 5-fold increase in transport activity. Interest- ingly, the increase in the normal isoform of GLUT 1 was not blocked by cycloheximide (Haspel et al., 1986) despite the established requirement of new protein synthesis for elevated transport activity (Kletzien and Perdue, 1975; Van Putten and Krans, 1985; Fung et al., 1986).

Alternatively, GLUT 1 could exist in intracellular storage vesicles (Zorzano et al., 1989) and translocate to the plasma membrane in a manner analogous to the recruitment theory as postulated for insulin-stimulated glucose transport (Cush- man and Wardzala, 1980; Suzuki and Kono, 1980). Indeed, Kozka et al. (1991) have recently demonstrated a 2-3-fold increase in GLUT 1 content on the cell surface of glucose- deprived 3T3-Ll adipocytes, although this by itself could not account for the 10-fold change in transport rates.

Another alternative to the above hypotheses is the synthesis of a new glucose transporter in response to glucose depriva- tion, a stress isoform, that differs from the other members of the transporter family. Indeed, a second immunoreactive transporter (p37) has been observed by us and others (Haspel et al., 1986; Reed et al., 1990). This second species could emerge via several mechanisms. I t could result from the synthesis of a smaller protein as a result of a novel message

‘ H. H. Kitzman, Jr., and S. C. Frost, unpublished data.

or translational error. A novel message seems unlikely since the Northern blots revealed only a single species of GLUT 1 mRNA. In addition, upon N-glycosidase F treatment the p37 appeared to migrate with the normal form of naked protein (36 kDa). Thus it seems unlikely that gross translational errors have occurred as a result of glucose deprivation. On the other hand, p37 requires new synthesis. Thus, this form could represent an oligosaccharide-deficient or aberrantly glycosy- lated precursor of p46. If so, this may prove useful in the study of the biosynthesis of the transporter. Alternatively, it could result from the degradation of normal GLUT 1 in an attempt by the cells to scavenge sugar residues in the face of glucose deficiency. Whether a precursor or degradation prod- uct of p46, the functional status of p37 has not been investi- gated. Although our evidence suggests that p37 is not an aglyco form, the literature is equivocal regarding the activity of the deglycosylated form of the transporter. Feugeas et ai. (1990) showed that removal of the oligosaccharide side chain by enzyme treatment inhibited transport activity. However, mutation of the asparagine residue only reduced its affinity for glucose (Asano et al., 1991). Regardless, a role of p37 in transport activation is unlikely because its kinetic appearance is not correlated with enhanced transport activity (by either time or glucose concentration), and it fails to accumulate in the presence of fructose despite elevated transport activity.

Thus, we must consider activation of preexisting transport- ers to explain the increase in transport rates in glucose- deprived 3T3-Ll adipocytes. There are two aspects that make this activation event unique with respect to other studies which support activation of the GLUT 1 system (Gould et al., 1989: Harrison et al., 1991a, 1991b, 1992; Clancy et ai., 1991; Gibbs et al., 1988). First is its dependence on protein synthe- sis. Thus the synthesis of a new protein is required for the increase in glucose transport activity. The turnover of this protein must be fairly slow because if cycloheximide is added at 12 h to prevent further synthesis, the elevated activity is maintained for at least an additional 12 h in the absence of added glucose. The fact that transcription is not required suggests that the message for this protein is present before the initiation of starvation. Interestingly, Clancy et al. (1991) showed that cycloheximide, a t high concentrations, increased glucose transport activity in fed 3T3-Ll adipocytes, which may suggest the existence of an endogenous inhibitor of GLUT 1, although a t low concentrations of the inhibitor (which are specific for protein synthesis), we have been unable to show elevated transport activity. The second aspect of starvation-induced activation which is unique is the require- ment for glucose (and not one of its metabolites), suggesting that this protein has the same specificity as the transporter, itself. Although Marshall and his colleagues have shown glu- cose-dependent desensitization of insulin-stimulated glucose transport (for review, see Marshall et al. (1991)), glucose metabolism is required for this phenomenon (Garvey et al., 1987). We are aware of only two “known” intracellular pro- teins which bind glucose; hexokinase (as a substrate) and glycogen phosphorylase (as an allosteric effector). Earlier work has shown that hexokinase activity is not affected by starvation (Shiu et al., 1977). Phosphorylase a, although in- hibited by glucose (for review, see Johnson (1992)), is found primarily in muscle and liver. Thus, it seems unlikely that either of these would be involved in translational control of the GLUT 1 transport system in the wide variety of cells that respond to glucose deprivation. Other “unidentified” glucose- binding proteins have been characterized (Lee and Lipmann, 1977, 1978). These proteins (presumably oligomers of a pro- tein with a monomeric molecular mass of 18 kDa) were found

Glucose Deprivation in 3T3-Ll Adipocytes 1325

in loose association with the plasma membranes of trans- formed cells. Upon purification, these proteins increased the rate of glucose transport by a process which did not require de novo transcription and were additive to serum stimulation of transport. In addition, the specificity of binding showed a spectrum similar to that of the glucose transporter. Whether these proteins are responsible for transporter activation in 3T3-U adipocytes is yet to be determined.

Acknowledgments-We acknowledge with pleasure the assistance of Dr. Harry Nick and his colleagues in the measurement of trans- porter message levels.

REFERENCES Asano, T., Shihasaki, Y., Ohno, S., Taira, H., Lin, J.-L., Kasuga, M., Kanazawa,

Y., Akanuma, Y., Takaku, F., and Oka, Y. (1989) J. Biol. Chem. 264,3416-

Asano, T., Katagiri, H., Takata, K., Lin, J.-L., Ishihara, H., Inukai, K., Tsukuda, 3420

K., Kikuchi, M., Hirano, H., Yazaki, Y., and Oka, Y. (1991) J. Biol. Chem. 266,24632-24636

Bell, G. I., Kayano, T., Buse, J. B., Burant, C. F., Takeda, J., Lin, D., Fukomoto, H., and Seino, S. (1990) Diabetes Care 13,198-208

Blok, J., Gibbs, E. M., Lienhard, G. E., Slot, J. W., and Geuze, H. J. (1988) J. Cell Biol. 106 , 69-76

Calderhead, D. M., Kitagawa, K., Tanner, L. I., Holman, G. D., and Lienhard, G. E. (1990) J. Biol. Chem. 265,13800-13808

Church, G. M., and Gilbert, W. (1984) Proc. Natl. Acad. Sci. U. S. A . 81,1991- 1995

Clancy, B. M., and Czech, M. P. (1990) J. Biol. Chem. 265,12434-12443 Clancy, B. M., Harrison, S. A,, Buxton, J. M., and Czech, M. P. (1991) J. Biol.

Cornelius, P., Marlowe, M., Lee, M. D., and Pekala, P. H. (1990) J. Biol. Chem.

Cushman, S. W., and Wardzala, L. J. (1980) J. Biol. Chem. 255,4758-4762 Dulbecco, R., and Freeman, G. (1959) Virology 8,396-397 Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biochem. 132,6-13 Feugeas, J.-P., Neel, D., Pavia, A. A,, Laham, A., Goussault, Y., and Derappe,

Frost, S. C., and Lane, M. D. (1985) J. Biol. Chem. 260,2646-2652 Fung, K. P., Choy, Y. M., Chan, T. W., Lam, W. P., and Lee, C. Y. (1986)

Garvey. W. T.. Olefskv. J. M.. Matthaei. S.. and Marshall. S. (1987) J. Biol.

Chem. 266,10122-10130

265,20506-20516

C. (1990) Biochim. Biophys. Acta 1030,60-64

Bcochem. Bzophys. Res. Commun. 134,1231-1237

Ch&. 262 , '189-197 '

6685

. , . . ,

Gibbs, E. M., Lienhard, G. E., and Gould, G. W. (1988) Biochemistry 27,6681-

Godwin, A. K., and Lieberman, M. W. (1991) Mol. Carcinogenesis 4,275-285 Gould, G. W., Lienhard, G. E., Tanner, L. I., and Gibbs, E. M. (1989) Arch.

Harrison, S. A., Buxton, J. M., Clancy, B. M., and Czech, M. P. (1990) J. Biol. Biochem. Biophys. 268,264-275

Chem. 265 , 20106-20116

Harrison. S. A,. Buxton. J. M.. Clancv. B. M.. and Czech. M. P. (1991a) J. B i d . ~ Chem. 266, i9438-19449

u. s. A. a8.7829-7843 Harrison, S. A., Buxton, J. M., and Czech, M. P. (1991b) Proc. Natl. Acad. Sei.

Jancy, B. M., Pessino, A., and Czech, M. P. (1992) J. Biol. Chem. 267,3783-3788

M. (1986) J. Biol. Chem. 261,6778-6789

13655-136fi2

Haspel, H. C., Wilk, E. W., Birnbaum, M. J., Cushman, S. W., and Rosen, 0.

Hiraki, Y., Rosen, 0. M., and Birnbaum, M. J. (1988) J. Bid . Chem. 263 ,

Johnson, L. N. (1992) FASEB J. 6,2274-2282 Kaestner, K. H., Christy, R. J., McLenithan, J. C., Braiterman, L. J., Cornelius,

P., Pekala, P. H., and Lane, M. D. (1989) Proc. Natl. Acad. Sci. U. S. A . 8 6 ,

Kitagawa, T., Msasumi, A,, and Akamatsu, Y. (1991) J. Biol. Chem. 2 6 6 , 3150-3154

Kletzien, R. F., and Perdue, J. F. (1975) J. Biol. Chem. 250,593-600 18066-18071

Koivisto, U. M., Martinez-Valdez, H., Bilan, P. J., Burdett, E., Ramlal, T., and

Ky:%!. J., Clark, A. E., and Holman, G. D. (1991) J. Biol. Chem. 266,11726-

-. . . -. . -

Klip, A. (1991) J. Biol. Chem. 2 6 6 , 2615-2621

Laemmli, U. K. (1970) Nature 227,680-685 Lee, S. G., and Lipmann, F. (1977) Proc. Natl. Acad. Sci. U. S. A . 7 4 , 163-167 Lee, S. G., and Lipmann, F. (1978) Proc. Natl. Acad. Sci. U. S. A. 75 , 5427-

11 131

5431

Bzochem. 87, 206-210 Markwell, M. A. K., Haas, S. M., Bieber, L. L., and Tolbert, N. E. (1978) Anal.

Mm!$all, S., Garvey, W. T., and Traxinger, R. R. (1991) FASEB J. 5 , 3031-

Ortiz, P. A,, Honkanen, R. A,, Klingman, D. E., and Haspel, H. C. (1992)

Pessin, J. E.,%llotson, L. G., Yamada, K., Gitomer, W., Carter-Su, C., Mora, Biochemist 3 1,5386-5393

R., Isselbacher, K. J., and Czech, M. P. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 2286-2290

Pettengell, K., and Frost, S. C. (1989) Biochem. Biophys. Res. Commun. 161 , fi.?.l-fi.?Q

3UJb

Reed, B. C., Shade, D., Alperovich, F., and Vang, M. (1990) Arch. Biochem.

R' IZZO f 0, L. J., LeMaire, M., Reynolds, J. A., andTanford, C. (1976) Biochemistry

Shiu, R. P., Pouyssegur, J., and Pastan, I. (1977) Proc. Natl. Acad. Sci. U. S. A.

"- "_ Bio hys. 279 , 261-274

15,3433-3440

74 2Rdn-RAdA ~ -," _- I-..

Suzuki, K., and Kono, T. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,2542-2545 Tai, P.-K. K., Liao, J.-F., Chen, E. H., Dietz, J., Schwartz, J., and Carter-Su,

Tordjman, K. M., Weingan , K A , James, D. E., and Mueckler, M. M. (1989)

Van Putten, J. P. M., and Krans, H. M. J. (1985) J. Biol. Chem. 2 6 0 , 7996-

Walker, P. S., Donovan, J. A,, Van Ness, B. G., Fellows, R. E., and Pessin, J.

Walker, P. S . , Ramlal, T., Donovan, J. A., Doering, T. P., Sandra, A., Klip, A,,

Wong, W. T., Frost, S. C., and Nick, H. S. (1991) Biochem. J. 275,313-319 Yamada, K., Tillotson, L. G., and Isselhacher, K. J. (1983) J. Biol. Chem. 258 ,

Zorzano A. Wilkinson, W Kotliar, N Thiodis G., Wadzinski B. E., Ruoho,

Zubay, G. (1988) Biochemistry, 2nd ed., p. 648, Macmillan, New York

C. (1990) J. Biol. Chem. 265,21828-21834

Proc. Natl. Acad. Scr. U. 8. A: 86, 7761-7765

8001

E. (1988) J. B i d . Chem. 2 6 3 , 15594-15601

and Pessm, J. E. (1989) J. Bzol. Chem. 264,6587-6595

9786-9792

A. E.,'and Pilch, P. F. (1689) J. BioL'khem. 2 6 4 , 12358-123133