Regulation of expression of the glucose transporter GLUT1 by glucose in 3T3-L1 adipocytes


Material Information

Regulation of expression of the glucose transporter GLUT1 by glucose in 3T3-L1 adipocytes
Physical Description:
xii, 165 leaves : ill. ; 29 cm.
McMahon, Robert Joseph, 1968-
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Monosaccharide Transport Proteins -- genetics   ( mesh )
Monosaccharide Transport Proteins -- isolation & purification   ( mesh )
Monosaccharide Transport Proteins -- metabolism   ( mesh )
Glucose -- physiology   ( mesh )
Gene Expression Regulation   ( mesh )
Glycosylation   ( mesh )
Adipocytes -- physiology   ( mesh )
Cricetulus   ( mesh )
Department of Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Biochemistry and Molecular Biology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1995.
Bibliography: leaves 158-164.
Statement of Responsibility:
by Robert Joseph McMahon.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002306499
oclc - 50189830
notis - ALQ9817
System ID:

This item is only available as the following downloads:

Full Text







This work is dedicated to my wife Laura and my daughter



I would like to acknowledge the people without whose

help and encouragement this project could not have been

completed. I would first like to acknowledge the members of

my committee, Dr. Charles Allen, Dr. Christopher West, Dr.

Michael Kilberg, and Dr. John Gander, who graciously agreed

to help partway through the completion of this thesis. A

debt of thanks goes to Dr. Susan Frost, whose ability to

simultaneously allow her students the freedom to investigate

problems in their own fashion and yet maintain their focus

constitute only two of the many admirable qualities I hope to

bring to my own lab.

I would also like to acknowledge the assistance given to

me on numerous occasions by Dr. Ron Laine and Marc Malandro.

My discussions with them have been invaluable in overcoming

obstacles that arise during any thesis. I would also like to

thank Maxine Fisher for always reminding me to cultivate

personal relationships between scientists which are essential

for a successful career.

I would also like to thank my parents, Joseph and

Florence, for the freedom to always pursue my ambitions with

their constant support. Lastly, I would like to thank my

wife Laura for her unending support and encouragement during

this period of immense personal and professional growth. Her

love and involvement have enriched during this experience

beyond measure.


ACKNOWLEDGMENTS ........................................... iii

LIST OF FIGURES ..........................................viii

ABSTRACT ....................................................x

OVERVIEW ....................................................1

GLUT Proteins and Glucose Transport....................... 1
GLUT Protein Family ....................................1
Glucose Transporter Structure ..........................4
Regulation of Glucose Transport...........................7
Regulation of Glucose Transport Activity by Glucose
Deprivation .........................................8
Regulation of Glucose Transporter Protein ..............8
3T3-L1 Adipocytes ........................................10

MATERIALS AND METHODS ...................................... 16

Materials ................................................16
Methods ..................................... ... ......... 17
Cell Culture ..........................................17
Antibody Production ...................................18
Peptide Purification of Anti-GTl Antibody ............. 21
Membrane Fraction Isolation ...........................23
Markwell Assay for Protein Determination .............. 26
Gel Electrophoresis ...................................27
Electrotransfer and Western Blotting .................. 27
Metabolic Radiolabeling of 3T3-L1 Adipocytes .......... 28
GLUT1 Immunoprecipitation .............................30
Endoglycosidase Digestions .........................32
Glycogen Isolation ....................................33
Glucose Assay .........................................34
Subfractionation of 3T3-L1 Adipocytes .................35
Cell Surface Biotinylation ............................37
Plasma Membrane Fragment Isolation ..................40


Introduction .............................................43
Characterization of Anti-GTl Antiserum ................ 44
Glucose Deprivation and GLUT1 Protein Expression ......48
Tryptic Digest of GLUT1 Transporter ................... 51
Effect of Tunicamycin Treatment on GLUT1 Protein ......54
Glycosidase Digestion of GLUT1 Glycoforms .............60

Specificity and Efficiency of GLUT1
Immunoprecipitation ...............................64


Synthesis and Glycosylation of GLUT1 in Control and
Glucose-deprived 3T3-L1 Adipocytes .................75
Regulation of GLUT1 Processing by Glucose ............. 78
Synthesis of GLUT1 Glycoforms During Glucose
Deprivation and Glucose Refeeding .................. 81
GLUT1 Half-Life in Control and Glucose-Deprived
Adipocytes ..........................................87

GLUT1 GLUCOSE TRANSPORTER ................................ 98

Introduction .............................................98
Time Course and Concentration Curve of Glucose
Assay ................. ................... ........102
Effect of Glucose-Deprivation on Glycogen Content
in 3T3-L1 Adipocytes .............................. 103
Glycogen Accumulation During Refeeding ............... 105
Effect of Fructose on Glycogen Depletion in
Glucose-Deprived Adipocytes .......................105
Effect of Insulin and 3-O-methylglucose on the
Glycogen Level in Glucose-Deprived 3T3-L1
Adipocytes Incubated with Fructose ................107
Prior Depletion of the Cellular Glycogen Pool
Results in a More Rapid Alteration of
Glycosylation in the Absence of Glucose ........... 111
Glycogen Content in the CHO Cell Line ................ 113
GLUT1 Glycosylation from Glucose-Deprived CHO
Cells ............................................. 118

ADIPOCYTES ..............................................126

Introduction ............................................126
Subfractionation of Control and Glucose-Deprived
3T3-L1 Adipocytes ................................. 131
Cell Surface Biotinylation of Control and Glucose-
deprived 3T3-L1 Adipocytes. ....................... 138

CONCLUSIONS AND FUTURE DIRECTIONS .........................147

Conclusions............................................. 147
Future Directions....................................... 154

LIST OF REFERENCES ........................................158

BIOGRAPHICAL SKETCH ....................................... 165


Figure page

1-1 Predicted Secondary Structure of Glucose Transporters. .4

2-1 Elution of GT1-Specific Antibodies. ................... 23

3-1 Characterization of Anti-GTl Antiserum. ...............45

3-2 Glucose Deprivation and GLUT1 Expression ............. 49

3-3 Tryptic Digest of p46 and p37. ........................52

3-4 Effect of Tunicamycin on Expression of GLUT1.......... 55

3-5 GLUT1 Expression Following Tunicamycin Washout. .......58

3-6 Glycosidase Digestion of GLUT1 Glycoforms. ............61

3-7 Immunoprecipitation of GLUT1 from Normal and Glucose-
Deprived 3T3-L1 Adipocytes ............................65

4-1 Biosynthesis of GLUT1 in Normal and Glucose-Deprived
Adipocytes ............................................76

4-2 Processing of GLUT1 in Normal and Glucose-Deprived
Adipocytes ...........................................79

4-3 Time Course of p37 GLUT1 Biosynthesis in Glucose-
Deprived Adipocytes ..................................82

4-4 Biiosynthesis of p37 GLUT1 During Glucose-Refeeding. ..85

4-5 Turnover of GLUT1 in Glucose-Deprived Adipocytes. .....88

4-6 Turnover of GLUT1 in Glucose-Refed Adipocytes. ........90

5-1 Time Course and Concentration Curve of Glucose Assay. 104


5-2 Glycogen Content and Glucose Deprivation. ............ 106

5-3 Glycogen Content and Glucose Refeeding. .............. 108

5-4 Glycogen Content in Fructose-Incubated Adipocytes. ... 109

5-5 Effect of Insulin and 3-O-methylglucose on Glycogen
Content ............................................112

5-6 Glycogen Depletion and GLUT1 Glycosylation. .......... 114

5-7 Glycogen Depletion and GLUT1 Glycosylation. .......... 116

5-8 GLUT1 Glycosylation and Glucose Deprivation in CHO
cells. ............................................... 119

5-9 GLUT1 Glycosylation and Glucose Deprivation in CHO
cells. ..................................... .........121

6-1 Subcellular Fractionation of 3T3-L1 Adipocytes ....... 133

6-2 Subcellular Fractionation of 3T3-L1 Adipocytes ....... 135

6-3 Cell Surface Biotinylation ...........................139

6-4 Plasma Membrane Fragment Isolation ................... 143

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy





Chairperson: Susan C. Frost
Major Department: Biochemistry and Molecular Biology

The regulation of GLUT1 expression by glucose was

analyzed in the 3T3-L1 adipocyte. Glucose deprivation of

3T3-L1 adipocytes for more than 12 hours resulted in the

accumulation of a second GLUT1 glycoform (p37). 48 hours of

glucose deprivation resulted in a 2-fold increase in total

GLUT1, a reflection of the accumulation of p37.

Metabolically labeled GLUT1 migrated as a wide band of

approximately 46 kDa. In contrast, metabolically labeled

GLUT1 protein in glucose-deprived adipocytes migrated at 37

kDa. Although at least 12 hours of glucose-deprivation were

required to affect GLUT1 glycosylation, glucose-deprived

cells quickly recovered the ability to correctly glycosylate

nascent GLUT1 upon the readdition of glucose.

The regulation of GLUT1 turnover by glucose was analyzed

by metabolic labeling and pulse-chase analysis. GLUT1 in

normal adipocytes exhibited a half-life of 14 hours. In

contrast, the turnover of the transporter in glucose-deprived

adipocytes was greater than 50 hours. The inhibition of

transporter turnover was readily reversed upon the re-

addition of glucose.

The glycogen content of 3T3-L1 adipocytes was assessed

as a potential carbohydrate source for protein glycosylation.

Normal 3T3-L1 adipocytes contained 0.537 0.097 pmol

glucose/106 adipocytes, which decreased in a first order

fashion upon glucose deprivation (0.0.052 pmol glucose/106

cells/hr, tl/2 = 6 hours). Fructose was unable to prevent the

depletion of glycogen in glucose-deprived adipocytes. In

glycogen-depleted cells, alterations in GLUT1 glycosylation

occurred more rapidly upon glucose deprivation than in

adipocytes with normal glycogen levels. Chinese Hamster

Ovary cells (CHO), which contain only 1% of the glycogen of

3T3-L1 adipocytes, exhibited the most rapid alteration in

GLUT1 glycosylation, occurring within 3 hours of glucose


The targeting of the alternatively glycosylated GLUT1

transporter was assessed by plasma membrane fragment

isolation, cell surface biotinylation, and subcellular

fractionation. All three techniques detected both GLUT1

glycoforms (p46 and p37) at the plasma membrane, indicating

that the targeting of the GLUT1 protein was not affected by

either the withdrawal of glucose or the alteration in N-

linked glycosylation.


GLUT Proteins and Glucose Transport

GLUT Protein Family

The maintenance of intracellular metabolism relies upon

the ability of the cell to incorporate solutes that cannot

readily diffuse across the plasma membrane. The uptake of

organic solutes, such as sugars and amino acids, is

accomplished by specific integral membrane proteins that

mediate the movement of molecules from the extracellular

environment to the cytosol. The transport of solutes across

the plasma membrane is accomplished by one of three

mechanisms (Voet and Voet, 1990). The first is passive

diffusion across the plasma membrane, a slow process due to

the low solubility of organic molecules in the hydrophobic

membrane interior. The second transport mechanism, termed

facilitated transport, is mediated by specific integral

membrane proteins that undergo conformational changes that

allow the solute access to the cytosol. The direction of

flux in this type of transport is determined solely by the

concentration gradient of the solute. Facilitated transport

describes many transport systems, including many salt, sugar,

and amino acid transporters. In some instances, solutes can

be transported against its concentration gradient if

transport is coupled to that of another solute that has a

concentration gradient of greater magnitude. The third

mechanism of transport is termed active transport. In this

process, the cleavage of high energy phosphoryl bonds

(usually in the form of ATP) is coupled to changes in the

conformation of the transport protein that result in the

translocation of the solute across the plasma membrane.

Since energy is expended in this process, solutes can be

transported against their concentration gradients. Examples

of this include the Na+, K+ ATPase and the Ca2+ ATPase.

Due to a large volume of scientific study, the transport

of glucose has served as a paradigm for the facilitated

transport of small, neutral organic solutes. The most well

understood glucose transport system exists in the mammalian

erythrocyte. The partial purification of a glucose

transporter protein was facilitated by its high concentration

in the erythrocyte plasma membrane. The purified transporter

from erythrocyte ghosts migrated as a wide band with an

average molecular weight of 55 kDa (Kasahara and Hinkle,

1977; Baldwin et al., 1979). Screening a cDNA expression

library with antiserum against the purified transporter

resulted in the isolation of a clone coding for a protein

capable of saturable, D-stereoisomer specific, glucose

transport (Mueckler et al., 1985; Birnbaum et al., 1986).

The study of glucose transport then underwent another

expansion when it was discovered that the transport of

glucose was not accomplished by the same protein in every

cell type. Screening cDNA libraries from other tissues at

low stringency led to the cloning of additional glucose

transporters, which in total are now known as the GLUT

protein family (GLUcose Transporter), a group of structurally

and functionally related integral membrane glycoproteins that

mediate the tissue-specific uptake of glucose (reviewed by

Gould and Bell, 1990). The expression of GLUT1 is virtually

ubiquitous; in most tissues, it is thought to be responsible

for a basal, constitutive transport activity. GLUT2 is

expressed predominantly in liver and is characterized by a

relatively high affinity constant; its role has been ascribed

to equilibrating glucose between the hepatocyte and blood

plasma according to the extracellular glucose concentration.

GLUT3 is expressed primarily in fetal muscle and brain

tissue. GLUT4 is expressed in insulin-sensitive tissues such

as heart, muscle, and adipose. The localization of GLUT4 can

be altered upon the binding of insulin, where the transporter

is translocated from an intracellular compartment to the

plasma membrane. GLUT5 is a recently cloned transporter that

is characterized by a high affinity for fructose; its role in

glucose transport and utilization has not yet been

elucidated. GLUT 6, a recently cloned gene, is a non-

expressed pseudogene. GLUT7, a component of the glucose-6-

phosphatase system, is localized in the endoplasmic reticulum

of the hepatocyte where its role is vital in gluconeogenesis.

The complex arrangement of transporter family, exhibiting

tissue-specific expression, developmental regulation, and

hormonal regulation, lends itself well to being capable of

delivering glucose to different tissues with varying

requirements for carbohydrate.

Glucose Transporter Structure

Although GLUT proteins exhibit a high degree of tissue-

specific expression, analysis of the primary sequences of the

cloned transporters reveals a high degree of homology. The

GLUT family exhibits approximately 68% overall homology; the

highest degree occurs in the membrane spanning regions, and

the lowest conservation occurs in the amino and carboxy

terminal ends and the large intracellular loop between

membrane spanning regions 6 and 7 (Carruthers, 1990). The

predicted topology of the transporter is depicted in Figure

1. The disposition of the transporter in the membrane was

based upon hydrophobicity and hydrophilicity measurements

which predict putative membrane spanning domains (Mueckler et

al., 1985). This model predicts that the GLUT proteins

exhibit 12 membrane spanning regions, and places both the

amino and carboxy terminal ends facing the cytoplasm

(Mueckler et al., 1985). Infrared spectroscopy and circular

dichroism have revealed a high degree of a-helical nature in

the transporter protein, which likely represents the membrane

spanning regions (Chin et al., 1986; Alvarez et al., 1987).

This topological model has received a significant amount of

experimental support (Shanahan et al., 1984; Cairns et al.,

Membrane 1 3 4 5 7 9 10 11s


Figure 1-1 Predicted Secondary Structure of GLUT1 The
predicted topology of the GLUT1 transporter, based on
hydropathy values (Mueckler et al., 1985). The amino- and
carboxy-terminal ends of the transporter are designated by
[N] and [C], respectively. The intracellular- and
extracellular-facing sides of the transporter are designated
by [IN] and [OUT], respectively. The boxes represent
membrane spanning regions and the branched chain represents
oligosaccharide attached to the extracellular loop between
membrane spanning regions 1 and 2.

1984; Cairns et al., 1987; Davies et al., 1987). Utilization

of peptide-specific antibodies directed against amino or

carboxyl terminal ends of GLUT1, in conjunction with tryptic

digestion, allowed the analysis of the predicted

intracellular and extracellular faces of the transporter.

Using an antiserum that recognized GLUT1 in its native form

and a preparation of erythrocyte membranes of known

orientation, it was determined that the carboxy terminus

resided intracellularly because the transporter was

recognized only in vesicles which were predominantly "inside

out". Furthermore, tryptic digests of the transporter

resulted in two large fragments of only slightly different

molecular weights, supporting the hypothesis that the major

cleavage site in the protein is near the middle of the large

intracellular loop between membrane spanning regions 6 and 7.

Antiserum specific for the amino terminus of the transporter

recognized a large diffuse band, suggesting that this

fragment contained the oligosaccharide, confirming that the

putative N-linked glycosylation site in the first

extracellular domain was utilized. Labeling the erythrocyte

transporter with the competitive ligand cytochalasin B before

proteolysis demonstrated that the binding site for this

ligand resided in the carboxy terminal half of the protein

(Holman and Rees, 1987).

Apart from the amino acid sequence, a second aspect that

deserves review is the glycosylation of the glucose

transporters. All glucose transporters are glycosylated at

an asparagine residue located between the first and second

membrane spanning region (Carruthers, 1990). Although the

complete oligosaccharide sequence is not known, some

characteristics have been elucidated. All members of the

glucose transporter family can be digested with N-glycanase F

or endoglycosidase F to release the oligosaccharide,

indicating N-linked glycosylation (Haspel et al., 1986;

Kitzman et al., 1994). The oligosaccharide is usually a

large structure that accounts for approximately 10 kDa of

migration by SDS-PAGE, or 20% of its total relative molecular

weight. Additionally, GLUT proteins migrate as diffuse bands

when analyzed by SDS-PAGE, but whether this heterogeneity

results from differences in the oligosaccharide structure

(known as microheterogeneity), or from different charges

associated with the terminal sugar chains, is not known.

Regulation of Glucose Transport

The regulation of glucose transport has received much

attention due to the central role the transporter plays in

the utilization of glucose. Glucose transport is regulated by

many factors, including hormonal influence and cellular

stress. The influence of hormones on the transport process

has been extensively reviewed elsewhere (Czech et al., 1992;

Czech, 1995) and will not be discussed in this work. Rather,

I will highlight the role of cellular stress, particularly

nutrient deprivation, on the regulation of glucose

transporter expression.

Regulation of Glucose Transport Activity by Glucose

Glucose deprivation, a specific type of nutrient

deprivation, augments glucose uptake in a variety of cell

types, including glial cells (Walker et al., 1989), avian

fibroblasts (Martineau et al., 1972; Kletzein and Purdue

1975; Pessin et al., 1982; Yamada et al., 1983), mammalian

fibroblasts (Germinario et al., 1982; Haspel et al., 1986,

Ortiz and Haspel, 1993), muscle cells (Walker et al., 1989;

Koivisto et al., 1991; Maher and Harrison, 1991), and

adipocytes (van Putten and Krans, 1985; Reed et al., 1990;

Tordjman et al., 1990). Several characteristics of this

enhanced uptake are common among these cell types. Glucose

uptake enhancement results from an increase in the Vm, of the

transport process with only a small change in the affinity

(Km) (Kletzien and Purdue, 1975; van Putten and Krans, 1975).

Transport activation also contains a protein synthesis-

dependent component, since exposure to cycloheximide

attenuates this response (Tordjman et al., 1990; Kitzman et

al., 1994). Metabolism or metabolites of glucose are not

required since incubation with 3-0-methyl glucose and 2-

deoxyglucose (2-deoxyglucose is only phosphorylated) can

prevent transport activation (Kitzman et al., 1994).

Regulation of Glucose Transporter Protein

Glucose deprivation also regulates glucose transport by

regulating the GLUT1 transporter protein level. Although not

as intensely studied as the regulation of transport activity,

progress has been made in elucidating the regulation of GLUT1

protein by glucose. The regulation of transporter expression

can be categorized into two classes; the first, where the

level of the transporter changes in response to glucose

deprivation, and the second, where a second transporter form

accumulates. Glucose deprivation results in an increase in

GLUT1 protein levels in some cell types, including the

fibroblast cell line 3T3-C2 and the kidney cell line NRK

(Haspel et al., 1986; Haspel et al., 1991). This effect,

however, is not conserved in all cell types; in 3T3-L1

adipocytes, there is no change in the level of normal GLUT1

transporter (Reed et al., 1990; Kitzman et al., 1994). With

extended periods of glucose deprivation, an additional

protein recognized by anti-GLUT1 antiserum accumulates. This

phenomenon is observed in several other cell types, including

the 3T3-L1 fibroblasts, normal rat kidney cells, and a Wilm's

tumor cell line (Haspel et al., 1986, Haspel et al., 1990;

Shawver et al., 1987; Haspel et al., 1986, Haspel et al.,

1990; Kitzman et al., 1994). This second protein migrates as

analyzed by SDS-PAGE approximately 10 kDa lower than the

normal GLUT1 protein. The glycosylation state of this

protein was investigated by treatment with endoglycosidase F

or N-Glycanase F, which remove all N-linked oligosaccharides.

The migration of this second protein can be altered slightly

upon this treatment, indicating that like the normal GLUT1

protein, it is modified with N-linked oligosaccharide (Haspel

et al., 1986; Kitzman et al., 1994). Furthermore, the

migration of the normal GLUT1 protein and the low molecular

weight protein were found to be identical upon digestion with

N-Glycanase F, indicating that resultant core proteins were

of identical molecular weight. Most investigators have now

accepted the assumption that the second protein is indeed a

different form of the GLUT1 transporter (Haspel et al., 1986;

Haspel et al., 1990; Kitzman, et al., 1993).

3T3-L1 Adipocytes

White adipose tissue plays an important role in the

storage of triglyceride resulting from excess blood glucose.

With this important role in the disposal of glucose, the

transport system for glucose is highly regulated. Many

studies involving the transport of glucose utilize adipocytes

isolated from fat pads, but these cells exhibit several

disadvantages, including the propensity to dedifferentiate

once isolated, and the inability to culture these cells for

extended periods of time. For these reasons, the work

described herein utilized the 3T3-L1 adipocyte, an adipocyte

model cell line, to study glucose transporter regulation.

In 1974, Green and Kehinde isolated several subclones of

mouse fibroblasts 3T3 cells that exhibited the propensity to

accumulate large lipid droplets once confluent. This

phenomenon could be prevented or reversed by passing and

reculturing the cells at low density. Extensive subcloning

of these cells resulted in the establishment of the 3T3-L1

(Lipid accumulating subclone 1) cell line in which

essentially all cells accumulate fat droplets (Green and

Meuth, 1974).

After observation of the onset of lipogenesis during the

differentiation of 3T3-L1 cells, it was established that the

3T3-L1 adipocytes express the proteins necessary to carry out

the metabolism characteristic of true adipose tissue.

Lipoprotein lipase (Spooner et al., 1979), fatty acid

synthase (Student et al., 1980; Wise et al., 1980),

acquisition of hormonal sensitivity, including to insulin

(Rubin et al., 1977; Resh et al., 1982), enzymes involved in

phosphatidylcholine and phosphoethanolamine biosynthesis

(Coleman et al., 1978), stearoyl CoA desaturase and fatty

acid binding protein (Cook et al., 1988) are all expressed

specifically during the onset of differentiation of 3T3-L1

cells. The large body of information regarding the

metabolism of the 3T3-L1 adipocyte has established this cell

line as a valid model for both the differentiation and

physiology of true adipose tissue.

3T3-L1 adipocytes express two members of the glucose

transporter family; GLUT1, the basal transporter, and GLUT4,

the insulin-sensitive glucose transporter. It has been

estimated by cell surface labeling techniques in normal and

insulin-treated adipocytes that each cell expresses

approximately 1 x 106 and 0.3 x 106 GLUT1 and GLUT4

transporter molecules, respectively (Calderhead et al.,

1990). The 3T3-L1 cell line has been used extensively in the

study of glucose transport regulation, including regulation

by insulin (reviewed by Czech, 1995), cholera toxin and

dibutyryl cyclic AMP (Clancy and Czech, 1990), cadmium, and
tumor necrosis factor a (Stephens et al., 1992).

In particular, the 3T3-L1 adipocyte has been used to

investigate the effects of glucose deprivation on glucose

transport activity and glucose transporter expression. The

withdrawal of glucose from the culture medium of 3T3-L1

adipocytes results in a 10-fold increase in the rate of

glucose transport, consistent with the responses of other

cell types to glucose deprivation (van Putten and Krans,

1985, Kitzman et al., 1994). Furthermore, the transport

activation results from an increase in the V, of the

transport system, with little change in the affinity of the

transporter for glucose (Km)((van Putten and Krans, 1985).

Other groups have since confirmed this result (Tordjman et

al., 1990; Kitzman et al., 1994). An increase in the maximal

velocity of transport without a change in the Km suggests

that there is either a greater number of transporter

molecules at the cell surface available for transport, or

that the activity of the transporter has been increased,

possibly by allosteric or covalent modification. Transport

activation, beginning at 3 hours post glucose withdrawal and

reaching a maximum at 24 hours, is protein synthesis-

dependent as assessed by cycloheximide sensitivity (Tordjman

et al., 1985, Kitzman et al., 1994). Additionally,

cycloheximide attenuated transport activation at any time

during glucose deprivation (Kitzman et al., 1994). Up-

regulation of glucose transport by glucose deprivation was

concentration-dependent, with a transport activation

threshold of approximately 0.5 mM. Additionally, transport

activation was completely reversible upon the re-addition of

glucose, indicating that down-regulation of transport was

also glucose-dependent (Kitzman et al., 1994). The

dependence of transport activation upon glucose or one of the

metabolites of glucose was tested by analyzing the up-

regulation of transport in the presence of several other

sugars; some of which are not transported by glucose

transporters, some of which are only ligands of the

transporter, and some which are transported but not

metabolized (Haspel et al., 1990; Kitzman et al., 1994).

Incubation of glucose-deprived 3T3-L1 adipocytes with

fructose, galactose, or maltose, none of which are

internalized by the glucose transporter, although maltose (a

disaccharide of glucose) is bound by the transporter, had no

effect on the up-regulation of transport activity. 3-0-

methylglucose, which is transported by the glucose

transporter but not metabolized, and 2-deoxyglucose, which is

only phosphorylated, completely abolish transport activation

(Kitzman et al., 1994). These observations suggest that

neither the binding of the transporter nor the metabolism of

glucose is required for prevention of transport activation;

rather, they suggest that a molecule that is either glucose

itself or a molecule closely resembling glucose is the

regulatory factor controlling transport activity.

Another effect of glucose deprivation in 3T3-L1

adipocytes is the appearance of a second protein recognized

by anti-GLUT1 antiserum that migrates at a relative molecular

weight of 37 kDa, in contrast to the normal GLUT1 transporter

which migrates at a relative molecular weight of 46 kDa (Reed

and Vang, 1990; Kitzman et al., 1994). The migration of

these two proteins, when treated with N-Glycanase F, is

identical (36 kDa) (Haspel et al., 1986; Kitzman et al.,

1994). This led several groups to conclude that the p37

protein is a second GLUT1 transporter which contains a

different oligosaccharide than the normal transporter. Like

transport activation, the appearance of p37 was prevented by

exposure to cycloheximide, even when added 12 hours after the

withdrawal of glucose, indicating that the appearance of p37

required new protein synthesis (Kitzman et al., 1994). The

threshold glucose concentration required to elicit the

accumulation of p37 was 0.005 mM. The difference in glucose

concentration (approximately 100-fold) for the activation of

glucose transport and the appearance of p37 strongly suggests

that p37 is not involved in, or responsible for, the

activation of glucose uptake (Kitzman et al., 1994). In

agreement with this conclusion, incubation of glucose-

deprived cells with fructose prevents the appearance of p37

while allowing the activation of glucose transport (Haspel et

al., 1986, Kitzman et al., 1994). Aside from the

determination of the glycosylation state of p37 and the

sensitivity of its appearance to cycloheximide, little else

is known about the regulation of the glucose transporter

protein by glucose. It should be noted that there is one

report where glucose deprivation resulted in an increase in

the level of the normal transporter protein, but the

discrepancy between this and other reports may lie in the

fact that equal numbers of cell equivalents were not analyzed

(Tordjman et al., 1990). The hypothesis that p37 is a second

GLUT1 protein synthesized in glucose-deprived 3T3-L1

adipocytes evokes several important questions regarding its

genesis, processing, turnover, and targeting. The goal of

this work is to describe how the expression of GLUT1 is

altered in the absence of glucose. This includes how the

transporter is processed to its final form, how it is

degraded, and to determine whether its altered glycosylation

state affects its targeting to the plasma membrane.



Dulbecco's Modified Eagle's Medium (D-MEM), Modified

Earle' Medium, a modification (a-MEM), and D-MEM base powder

was obtained from Life Technologies, Inc. Calf serum (Cat.

No. 1100-90) and fetal bovine serum (FBS) (Cat. No. 1020-75)

were obtained from Intergen. Glucose-free FBS was prepared

by dialyzing against phosphate buffered saline, pH 7.4, (PBS)

for 48 hours at 4C, with a molecular weight cutoff of 13,000

(dialysis tubing obtained from SpectraPor). L-methionine, L-

cysteine, L-proline, D-glucose, D-fructose, D-xylose, 3-0-

methylglucose, methylisobutylxanthine, dexamethasone, protein

A Sepharose, Keyhole Limpet Hemocyanin, Sephadex G-100,

Freund's Complete and Incomplete Adjuvant, polyoxyethylene 9

lauryl ether (C12E9), octanoic acid, bovine serum albumin,

poly-L-lysine (M.W. > 150,000), and anti-rabbit IgG

conjugated horseradish peroxidase were obtained from Sigma.

Insulin was a generous gift of Dr. Ronald Chance of Eli Lilly

Corp. Tran35S-Label (1100 Ci/mmol) was obtained from ICN.

Aquacide was obtained from Calbiochem. Streptavidin-agarose,

m-maleimidobenzoyl-N-hydroxysulfosuccinimide ester (sulfo-

MBS), the Sulfolink matrix, and sulfosuccinimidyl-6-

(biotinamido) hexanoate (NHS-LC-Biotin) were obtained from

Pierce. All other chemicals were obtained through either

Fisher or Sigma.


Cell Culture

Cell culture and differentiation of 3T3-L1 fibroblasts

were performed as previously described (Frost and Lane,

1985). Cells were seeded in plastic tissue culture dishes

and cultured in D-MEM containing 10% calf serum (CS) for

seven days, refeeding every other day until confluent. After

seven days, the cells were incubated in D-MEM containing 10%

fetal bovine serum (FBS), 1 pg/ml insulin, 0.5 mM

methylisobutylxanthine, and 0.25 mM dexamethasone (Kuri-

Haruch and Green, 1978; Rubin et al., 1978; Reed and Lane,

1980). Two days later, the cells were incubated in D-MEM/10%

FBS containing 1 pM insulin only. Thereafter, the cells were

fed with D-MEM/10% FBS alone. Cells were used for

experiments between days 8 and 12 post-differentiation. To

prepare glucose-deprived 3T3-L1 adipocytes, the normal

culture medium was aspirated and replaced with glucose-free

D-MEM containing 10% dialyzed FBS for the indicated periods

of time.

CHO-K1 cells, obtained from American Type Cell Culture,

Rockville, MD, and LEC1 CHO cells, obtained as a kind gift of

Dr. Pamela Stanley, (Albert Einstein College of Medicine, NY,

NY) were seeded onto plastic tissue culture dishes (100 cm2

for most experiments) and grown to confluence in a-MEM medium

containing 10% FBS. Cells were used between days 1 and 5

post-confluence. For glucose deprivation of the CHO-K1 cells

and the LEC1 CHO cells, the a-MEM medium was aspirated and

replaced with glucose-free DMEM containing 10% dialyzed FBS.

This medium was supplemented with 0.1 mM proline in order to

satisfy the auxotrophic requirement of the CHO cell line .
Incubation of all cells with other sugars (i.e., xylose,

fructose, 3-0-methylglucose) was performed by adding the

appropriate amount of crystalline sugar in glucose-free DMEM

containing 10% dialyzed FBS to obtain a final concentration

of 25 mM.

Antibody Production

The detection of GLUT1 transporter protein was

accomplished through the generation of a polyclonal antiserum

against a peptide corresponding to the carboxy terminus of

the amino acid sequence. This region was chosen because of

its charged nature and its low degree of identity compared to

the same region in other members of the GLUT family.

Although use of this crude antiserum is adequate for the

detection of GLUT1 in Western blotting applications, it was

necessary to purify the GLUT1-specific antibodies for use in

immunoprecipitation experiments. The following sections

describe the generation of the anti-GLUT1 antiserum and the

purification of the peptide-specific antibodies.

Some small peptides are poorly immunogenic; therefore,

in order to ensure antigenicity, they are often coupled to

larger proteins, termed immune carriers, (Harlow and Lane,

1988). Covalent conjugation of small proteins or peptides to

these immune carriers results in an antigen large enough to

elicit an immune response. Antiserum prepared in this manner

contain antibodies against the immune carrier and the

antigen. The proteins typcially chosen as immune carriers

are BSA, ovalbumin, or keyhole limpet hemocyanin. Careful

choice of the immune carrier can result in antiserum which

recognizes only the peptide because the immune carrier is

absent from the experimental system. Antisera prepared

against synthetic peptides have several advantages over

antisera generated against whole proteins (Harlow and Lane,

1988). First, the antibody can be generated without knowing

the entire sequence of the protein or purifying the protein.

Second, the antiserum can be generated in such a way that it

will be specific for specific regions of the protein of

interest. Finally, peptide-specific antibodies have a higher

potential of being conformation specific.

For the preparation of an antiserum against GLUT1, a

peptide (CEELFHPLGADSQV) that corresponds to amino acid

residues 480-492 of the GLUT1 sequence was chosen as the

antigenic region. This peptide, designated GT1, was

synthesized by the Protein Chemistry Core facility at the

University of Florida and verified by amino acid analysis.

An additional cysteine was added to the amino terminus of the

peptide in order to facilitate covalent coupling to the

immune carrier. A keyhole limpet hemocyanin-GTl conjugate

was generated by using a thiol-specific cross-linking

reagent, sulfo-m-maleimidobenzoyl-N-hydroxysuccinimide ester

(sulfo-MBS). Keyhole limpet hemocyanin (KLH) (60 mg) was

dissolved in 10 ml 10 mM potassium phosphate, pH 7.3 and

dialyzed at 4C against the same for 48 hours. The

dialysate was concentrated to approximately 4 ml by covering

the dialysis bag in a liberal amount of Aquacide and wrapping

in foil (Calbiochem). The concentrated dialysate was divided

into microcentrifuge tubes and centrifuged in microcentrifuge

tube for 10 minutes to remove any protein aggregates. The

concentration of protein in the supernatant was estimated by

measuring the absorbance at 280 nm using the milligram

extinction coefficient for KLH of 1.6. KLH (6 mg) from a 12

mg/ml solution in 10 mM potassium phosphate, pH 7.3 was

reacted at room temperature with 0.6 mg sulfo-MBS for 45 min.

The KLH-MBS conjugate was separated from free sulfo-MBS by

passing the reaction mixture over a 10 ml G-100 column pre-

equilibrated in 50 mM potassium phosphate, pH 6.0. The KLH-

MBS conjugate is excluded in the void volume, and the free

sulfo-MBS is retarded in the column. Ten fractions (1 ml

each) were collected and measured at 280 nm. Peak fractions,

containing KLH-MBS were pooled and reacted with the GT1

peptide (6 mg) for 3 hours at room temperature with end-over-

end rotation. The pH of the reaction mixture was adjusted

with HC1 in order to produce the maximum amount of

precipitate. The reaction mixture was then allowed to stand

upright overnight at 4C to collect precipitate, which

represents the KLH-peptide conjugate.

A polyclonal antiserum to this conjugate was then

generated by injection of the KLH-GT1 conjugate (200 yg)

emulsified in Freund's Complete Adjuvant into the popliteal

lymph node of a New Zealand white rabbit (Sigel et al.,

1983). The rabbit was boosted by intradermal injection with

an equal amount of the GT1-KLH conjugate emulsified in

Freund's Incomplete Adjuvant 28 days later, and then again

two weeks later. Serum (designated as anti-GT1) was

collected every week by laceration of the medial ear vein.

Additional boost injections were administered approximately

every 6 weeks. This crude antiserum was divided into 1 ml

aliquots and stored at -200C.

Antiserum against a carboxy-terminal peptide of the

insulin-responsive glucose transporter GLUT4 (CSTELEYLGPDEND)

and an amino-terminal peptide of the glucose-regulated

protein GRP78 (EEEDLLEDVGTVC) were generated in the same

manner. This antiserum were used unpurified for Western


Peptide Purification of Anti-GT1 Antibody

Although the crude antiserum generated against the KLH-

GT1 conjugate recognized the peptide, a fraction of the whole

serum containing the total IgG was purified using a

modification of a previously described method (Dankert et

al., 1985). The eventual purpose of this technique was to

produce an antibody fraction that could be used for the

purification of peptide specific antibodies.

First, the pH of approximately 15 ml whole rabbit serum

was adjusted to 5 with 3M acetic acid. While vigorously

stirring, 0.75 ml octanoic acid (1 part in 20) was added

dropwise; the solution was then stirred an additional 30 min

at room temperature. The precipitate was collected at 41,000

x g for 30 min at 4C and the pellet discarded. The

supernatant was then removed and its volume determined. An

equal volume of ammonium sulfate (saturated solution in

water) was added and stirred overnight at 4C. The

precipitate was then collected by centrifugation at 41,000 x

g for 30 min at 4C. The pellet, consisting of primarily

immunoglobulin, was dissolved in PBS and the concentration of

this solution was estimated by measuring its absorbance at

280 nm (using a milligram extinction coefficient of 1.4).

The ammonium sulfate was then removed by dialysis against

three changes of PBS over 48 hours at 4C with a molecular

weight cutoff of 13,000 daltons.

The purification of peptide specific antibodies from the

total antibody fraction provides an advantage of reducing

non-specific protein--antibody interactions that could

complicate the interpretation of data gathered by

immunoprecipitation. Peptide-specific antibodies were

purified from the total IgG fraction by affinity

chromatography over a column of GT1 peptide immobilized on an

inert substrate. The GT1 peptide (2 mg) was covalently

coupled to a Sulfolink matrix through the additional cysteine

residue on the GT1 peptide following the manufacturer's

instructions. Sulfolink matrix consists of 6% cross-linked

beaded agarose with a terminal iodoacetyl group that confers

reactivity with sulfhydryl groups under mild (neutral pH)

conditions. An additional advantage of the Sulfolink matrix

is that the linkage between the peptide and the matrix is

irreversible; other matrices utilizing sulfhydryl chemistry

can be reversed under reducing conditions. Anti-GTl

antibodies were purified using the following method. The

total IgG fraction (15 mg each batch) purified from whole

serum was incubated with the peptide column and rotated end-

over-end at room temperature for 2 h. Unbound IgG was washed

from the column with PBS, monitoring the eluate by absorbance

at 280 nm until a baseline was reached. Specifically bound

IgG was then eluted with 0.1 M glycine, pH 3.0, by collecting

1 ml fractions. The eluate was immediately neutralized with

0.1 ml 1M Tris-Base. Peak fractions, as assessed by

absorbance at 280 nm, were pooled and dialyzed against PBS

for 12 h at 4C. The final IgG was stored at -20C as a 0.5

mg/ml solution in PBS.

Membrane Fraction Isolation

To facilitate the detection of integral membrane proteins

such as the glucose transporter, it is often desirable to

00 0.4

0 0.2

S 0.1

1 2 3 4 5 6 7 8 9 10

Fraction Number

Figure 2-1 Elution Profile of Peptide-Purified anti-GT1
Antibody IgG from a total antibody fraction was incubated
with the anti-GT1 column for 2 hours at room temperature.
Unbound antibody was washed from the column with PBS, and
specifically bound antibody was eluted with 0.1 M glycine, pH
3.0. The absorbance of 1 ml fractions collected during the
elution was then measured at 280 nm. This profile is
representative of at least ten elution profiles performed on
this column.

partially purify a cellular fraction which is enriched in

membrane protein. This affords the advantage of preparing

relatively concentrated preparations of the membrane protein

of interest, and furthermore, reduces non-specific

interactions in procedures such as Western blotting and

immunoprecipitation. To this end, GLUT1 was routinely

partially purified by preparing a cell lysate fraction which

was depleted of cytosolic and nuclear proteins. To prepare

this fraction, the culture medium was aspirated and the cell

monolayer rinsed in phosphate-buffered saline (PBS) at 40C.

Cells were then scraped into 5 ml TES buffer [20 mM Tris-HCl,

pH 7.4, 1 mM EDTA, 250 mM sucrose, and protease inhibitors

(20 pg/ml each of aprotonin, leupeptin, pepstatin, TPCK,

TLCK, and 1 mM PMSF)] and homogenized by 20 strokes in a 10

ml Potter-Elvejhem homogenizing flask with a teflon pestle.

The homogenate was centrifuged at 1300 x g for 5 min at 4C

to remove nuclei and unbroken cells. The supernatant was

centrifuged at 212,000 x g for 1 hour at 4C to pellet the

remaining particulate fraction. The supernatant, containing

cytosolic proteins, was discarded and the final membrane

pellet resuspended in TES buffer (usually 0.3 0.5 ml) by

homogenization. The resuspended membrane fraction was

quickly frozen by submersion in liquid nitrogen and then

stored until use at -20oC.

Markwell Assay for Protein Determination

Several assays have been developed for the determination

of protein concentration in biological samples. All protein

determinations, however, suffer the limitation that certain

compounds, such as detergents or urea, can interfere with the

accuracy of the assay. For the determination of protein in

samples containing membranes, the sample must be extracted

with detergent in order to obtain a true reflection of the

protein content. The most popular method, a procedure

developed by Lowry, has come into wide use because of its

ease and sensitivity (Lowry et al., 1951). The Lowry protein

assay, however, is interfered by the presence of detergents,

especially sodium dodecyl sulfate (SDS) (Peterson, 1979).

Protein determinations in this work were most commonly

performed on membrane fractions, necessitating the routine

use of the Markwell modification of the Lowry assay for

protein determination. The Markwell assay is a modification

of the Lowry protein determination which is compatible with

certain detergents, including SDS.

To generate a standard curve, samples of purified bovine

serum albumin (BSA) (0-100 pg) were diluted to a total of 0.1

ml with water. Aliquots of the unknown samples (usually 5 -

25 pl) were also diluted to 0.1 ml water. If the unknown

protein samples were in buffer, an equal volume of that

buffer was also added to the protein standards. Assay

solution (1 ml of [100 parts 2.0% Na2CO3, 0.4% NaOH, 0.16%

Na', K+ tartrate, 1.0% SDS to 1 part 4% CuSO4]), was added,

mixed well, and incubated at room temperature for 10 minutes.

Folin Phenol (0.1 ml, IN) was then added to each sample and

mixed well immediately after addition. The samples were

incubated for 45 min at room temperature for color

development. The absorbance of the samples was then

determined at 650 nm. The protein concentration of the

unknown samples was then determined from the standard curve.

Gel Electrophoresis

Gel electrophoresis was performed essentially as

originally described (Laemmli et al., 1970). Protein samples

were diluted with an equal volume of sample dilution buffer

consisting of 4% SDS, 6M urea, 10% B-mercaptoethanol, 0.15

mg/ml bromophenol blue, 40% glycerol and 20 mM Tris-base, pH

6.6. Gels were run at approximately 45 V overnight at room


Electrotransfer and Western Blotting

The detection of proteins by the Western blotting method

requires that the protein sample be transferred to a

substrate, usually nitrocellulose. The transfer buffer and

conditions closely follow the original procedure outlined by

Towbin et al., (1979). The electrophoresis gel was removed

from the gel apparatus and soaked in transfer buffer for 30

min at room temperature prior to transfer. Resolved proteins

were then transferred to nitrocellulose at 250 mA for 2.5

hours at 4'C in 25 mM Tris-base (unadjusted pH), 192 mM

glycine, 20% methanol. The blot was stained briefly in amido

black (0.2% in 40% methanol, 10% acetic acid) and destined

(40% methanol, 10% acetic acid). The blot was blocked in

Tris-buffered saline/0.1% Tween-20 (TBS-T) containing 5% non-

fat dry milk (NFDM) for 30 min at room temperature. The blot

was agitated constantly with an orbital shaker throughout the

procedure. The blot was then incubated in the same solution

containing a 1:500 dilution of the anti-GT1 antiserum for 1

hour at room temperature. After extensively washing the blot

in TBS-T, the blot was incubated in 1:50,000 dilution of goat

anti-rabbit antibody coupled to horseradish peroxidase in

TBS-T/5% NFDM for 1 hour at room temperature. The blot was

then extensively washed in TBS-T before before being

visualized with Amersham's Enhanced Chemiluminescence

detection reagents following manufacturer's directions.

Light emissions from the blot were captured on Hyperfilm MP

(Amersham). Band intensity was quantitated in the linear

range of the film on a Visage Bioscan video densitometer.

Metabolic Radiolabelinq of 3T3-L1 Adipocytes

In order to characterize individual proteins for their

turnover and processing, they need to be metabolically

labeled for a defined period of time. Usually, this is

accomplished with the use of radioactively labeled amino

acids, such as 35S labeled methionine and/or cysteine. In

order to determine the time required for the degradation of

the labeled protein, the label is then "chased" by incubating

the cells in a large excess of unlabeled amino acids in order

to dilute the specific activity of the intracellular amino

acid pools (Harlow and Lane, 1988). Immunoprecipitation and

recovery of the protein of interest during this period allows

the analysis of the radioactivity present in the initially

labeled protein pool and its disappearance due to degradation

as a function of time. The same strategy can be used for

very short labeling periods where the protein is labeled, and

then the processing or glycosylation of the protein can then

be followed by observing shifts in relative molecular weight

as assessed by gel electrophoresis. This section describes a

metabolic labeling procedure optimized for the labeling and

immunoprecipitation of GLUT1 expressed at endogenous levels

in 3T3-L1 adipocytes for subsequent gel electrophoresis and


Cells were incubated in 8 ml methionine- and cysteine-

free D-MEM without added serum for 1 hour to deplete the

intracellular pools of methionine and cysteine. The

depletion medium was aspirated and the cells incubated in 2

ml methionine- and cysteine-free DMEM containing 200 pCi/ml

Tran35S-Label for the specific times indicated. To initiate

the chase periods, the labeling medium was aspirated and

replaced with 8 ml complete DMEM/10% FBS containing 2 mM

methionine and 2 mM cysteine. These conditions were

established to generate an adequate signal for fluorography

while minimizing the time of methionine and cysteine

depletion based on previous reports of possible alterations

in protein turnover with prolonged amino acid depletion

(Haspel et al., 1985).

GLUT1 Immunoprecipitation

Immunoprecipitation is a procedure by which a particular

protein can be isolated from a crude mixture of proteins by

incubation and binding of specific antibodies. Once this

immunocomplex is formed, it is captured on a substrate,

commonly protein A Sepharose, so that it can be removed from

solution. Proteins associated in a non-specific manner can

then be removed by washing the protein A Sepharose-

immunocomplex. Finally, the immunoprecipitated protein can

be released by destroying the interaction between the

antibody and the protein A, usually with strong denaturing

solutions such as urea, low pH, SDS, or a combination of

these conditions (Harlow and Lane, 1988).

GLUT1 was immunoprecipitated from the total membrane

fraction in the following manner. A total membrane fraction

was obtained by first washing the monolayer in cold PBS. The

cells were then scraped into TES buffer and homogenized by 20

strokes in a 10 ml Potter-Elvejhem flask with a teflon

pestle. The homogenate was then centrifuged at 212,000 x g

for 1 hour at 40C to collect a total membrane pellet. The

final membrane pellet was then homogenized with 20 strokes of

a 2 ml Potter-Elvehjem homogenizing flask with a teflon

pestle in 1 ml extraction buffer [PBS containing 1 mM EDTA,

2% C12E9, 0.1% SDS and protease inhibitor cocktail as above].

Any insoluble material was then removed by centrifugation in

a microcentrifuge for 5 min at 4C. The supernatant was then

recovered, frozen in liquid nitrogen, and stored at -20"C.

In order to eliminate non-specific interactions between

membrane proteins and immunoglobulin, thawed extracts

(containing between 1.5 2 mg membrane protein) were

incubated with an unrelated non-immune antiserum (5 pl) and

collected with a 50% suspension of Protein A Sepharose (25

pl) for 1 h at 4C. Samples were centrifuged briefly in a

microcentrifuge to pellet the protein A Sepharose pellet and

the pre-cleared supernatants transferred to new

microcentrifuge tubes. Extracts were adjusted to equal

protein concentration in 1 ml with additional extraction

buffer and incubated with the anti-GT1 antibody (5 pg) for 3

hours at 4C with end-over-end rotation. The 50% Protein A

Sepharose suspension in PBS was then added (25 pl) for 2

hours to collect the immunoprecipitates. The Protein A

Sepharose was then washed 3 times with extraction buffer,

followed by 4 times for 10 min with extraction buffer

containing 1M NaCl. Immunoprecipitates were released by

incubation in 0.1 ml sample dilution buffer containing 6 M

urea and 10% BME for 15 min at 37C. The supernatants were

then loaded onto an 8% SDS-PAGE gel and run overnight at 50

V. For fluorography, the gels were fixed in 10% TCA/40% MeOH

for 30 min, soaked in water for 30 min, and then soaked in 1M

sodium salicylate for 1 h before drying at 60C under vacuum.

Dried gels were exposed to Amersham Hyperfilm typically for 4


Endoqlycosidase Digestions

Endoglycosidases refer to a class of enzymes that

catalyze the cleavage of oligosaccharide chains at specific

sugar residues. These enzymes are often useful for the

characterization of oligosaccharides on glycoproteins. For

the analysis of the oligosaccharide on GLUT1, two specific

endoglycosidases were chosen. The first, N-Glycanase F,

cleaves between the two GlcNAc residues adjacent to the

asparagine residue on the protein. This enzyme can be used

to determine if glycoproteins are post-translationally

modified with an N-linked oligosaccharide. This enzyme will

not cleave oligosaccharides attached to serine or threonine

resides (O-linked). The second endoglycosidase,

endoglycosidsase H, also cleaves between the two GlcNAc

residues but is specific for the high mannose type of N-

linked oligosaccharide. This enzyme is useful for not only

determining if a protein contains an N-linked

oligosaccharide, but can be used to distinguish among high

mannose type and complex type oligosaccharides. Like N-

Glycanase F, this enzyme will not cleave oligosaccharides

attached in an O-linked manner.

Membrane protein (20 pg) from control or glucose-

deprived adipocytes and 10 pg from LEC1 CHO cells were

denatured in 50 mM BME/0.5% SDS in a final volume of 50 Mp

for 20 min at 37C. Then, for N-glycosidase F digestions,

the samples were brought to 150 mM Tris-HC1, pH 8.0, 1% NP-

40, protease inhibitors as above, 1.25 U N-glycosidase F and

then incubated at 37C for 2 hours. For endoglycosidase H

digestions, the denatured protein sample was brought to 75 mM

sodium citrate, pH 5.5, protease inhibitors as above, and 2.5

mU endoglycosidase H and incubated at 37C for 2 hours. An

equal volume of 2X sample dilution buffer was added and

loaded onto an 8% SDS-PAGE gel for separation.

Glycoqen Isolation

Glycogen, a polymerized form of glucose, is stored by

most cells (Voet and Voet, 1990). Glycogen breakdown

provides a source of fuel for energy production by glycolysis

as well as other pathways. The determination of glycogen in

this work is not made directly; rather the glycogen is

hydrolyzed with acid to produce free glucose which can then

be assayed directly by the method described below. The

results of this assay are therefore represented as the amount

of glucose contained in glycogen. The procedure used was

based on a previously described method for the isolation of

glycogen from liver (Pfleiderer 1963). After rinsing in

phosphate-buffered saline, pH 7.2 (PBS) at 40C, cells were

scraped into a 15 ml polypropylene tube in 1 ml PBS. The

cells were sonicated for 10 sec on power 1.5 at 50% duty

cycle on a Branson Sonifier 450. KOH (2.0 ml, 30% w/v) was

added and the mixture mixed vigorously. The sample was then

incubated in a boiling water bath for 15 min. The sample was

removed and cooled briefly before the addition of 3.5 ml 95%

EtOH. After mixing, the samples were heated in a boiling

water bath for 3 min and then cooled to room temp. The

precipitate was collected in a clinical centrifuge at 1300 x

g for 5 min at room temperature. The clear supernatant was

removed by aspiration and the precipitate washed in 1 ml 95%

EtOH by resuspension. The suspension was transferred to a

microcentrifuge tube and recollected at 13,300 x g for 5 min.

The supernatant was aspirated and the precipitate stored at

-200C. When all samples were collected, the precipitate was

dissolved in 0.2 ml 2N H2S04 and incubated in a boiling water

bath for 2 hours. The samples were cooled before the

addition of 0.15 ml 2N NaOH and 0.65 ml H20. The hydrolysates

were used for glucose determination as described below.

Glucose Assay

The assay chosen for the determination of glucose

released from glycogen is based upon the following reaction

mechanism. Glucose is first phosphorylated by hexokinase in

the presence of ATP. Glucose-6-phosphate is then oxidized in

the presence of NAD+ by glucose-6-phosphate dehydrogenase,

forming 6-phosphogluconate. This results in the production

of NADH in direct proportion to the initial concentration of

glucose. To perform the assay, 0.1 ml glycogen hydrolysate

was diluted with 0.4 ml H2SO4 / NaOH/ H20 mixture (2/1.5/6.5).

Samples of this solution (0.1 ml) were assayed in triplicate.

Samples, as well as the assay buffer, were pre-warmed to

37C. Assay reagent (1.0 ml) was mixed with 0.1 ml sample

and incubated at 370C for 30 min. The absorbance of the

samples was then measured at 340 nm.

Subfractionation of 3T3-L1 Adipocytes

Subfractionation is a procedure for separating and

purifying the different compartments of the cell. This

usually involves breaking the cell open while attempting to

maintain the integrity of the intracellular organelles. When

these individual organelles are then subjected to centrifugal

force, these organelles separate from each other on the basis

of both their size and density. In this manner, the

localization of proteins can be determined. One disadvantage

of this technique, is that the difference in density and size

between different cell organelles is relatively small. This

allows incomplete separation of these organelles resulting in

cross-contamination; that is, that a part of the golgi

fraction may be present in the plasma membrane, and so forth.

The following section describes a protocol used to separate

fractions enriched in plasma membrane, high density

microsomes, and low density microsomes. The high density

membrane fraction is usually enriched in proteins contained

in the endoplasmic reticulum and the golgi apparatus. The

low density membrane fraction is enriched in proteins

characteristic of endosomes, lysosomes, and the trans-golgi

network. Although all subfractionation procedures exhibit

some cross-contamination, this procedure will still provide

useful information in conjunction with other procedures to

analyze protein targeting.

Plasma membrane, high density, and low density membrane

fractions were isolated with the following subfractionation

procedure developed by Fisher and Frost (in preparation).

3T3-L1 adipocytes cultured in 10 cm tissue culture plates (5

plates per condition) were washed in PBS at 40C before being

scraped into 20 ml TES buffer. The cells were passed by a

ball bearing in a homogenizing block with a 0.0025 inch

clearance ten times at 40C. The homogenate was centrifuged

at 17,369 x g for 15 min at 4C [A]. The supernatant from

[A] was centrifuged at 48,254 x g for 30 min at 40C [B].

The pellet from [A] was resuspended into 25 ml TES buffer and

layered over 11.5 ml sucrose cushion of TES containing 1.12M

sucrose. This step gradient was centrifuged at 95,133 x g

for 65 min at 4C in a swinging bucket rotor [C]. The

interface from the step gradient, representing the plasma

membrane enriched fraction, was centrifuged at 48,254 x g for

30 min at 40C [D]. The pellet from [D] was resuspended into

TES buffer and centrifuged again at 48,254 x g for 30 min at

4C. This washed plasma membrane pellet was resuspended into

TES buffer, frozen in liquid nitrogen and stored at -200C.

The pellet resulting from the [B], representing the high

density membrane fraction, was resuspended into TES buffer

and centrifuged at 48,254 x g for 30 min at 40C. This washed

pellet was resuspended into 0.3 ml TES, and stored in a

manner similar to the plasma membrane fraction. The

supernatant from [B] was centrifuged at 212,000 x g for 75

min at 40C. The pellet, representing the low density

membrane fraction, was resuspended in TES buffer and

centrifuged at 212,000 x g for 75 min at 40C. This washed

pellet was resuspended into 0.3 ml TES, frozen in liquid

nitrogen, and stored at -200C. For Western blot analysis,

samples were diluted into sample dilution buffer containing 6

M urea and 10% B-mercaptoethanol before separation by SDS-

PAGE and transfer to nitrocellulose as described above.

Cell Surface Biotinylation

The modification of proteins with membrane impermeable

chemicals has recently proven to be an effective tool in the

study of membrane proteins, including the vitronectin

receptor (Nesbitt and Horton, 1992), Thy-i and lymphocyte

surface marker proteins (Arni et al., 1992), surface proteins

of 3T3-L1 fibroblasts (Hare and Taylor, 1991), surface

proteins of 3T3-L1 adipocytes (Hare and Taylor, 1992),

insulin receptors (Levy-Toledano et al., 1993), and tumor

necrosis factor receptor (Hsu and Chao, 1993). Several

membrane-impermeant modifiers are now commercially available,

with different functional groups to confer different chemical

reactivities. Many of these modifiers are conjugated to

biotin, a small molecule that has a tight affinity for

avidin. Biotin has many advantages as a molecular tag; it is

small and in many cases does not abolish biological activity

of ligands or enzymes, it can be conjugated to many different

types of compounds and so is quite versatile in its uses, and

finally, it can be readily detected by binding of avidin or

streptavidin. The binding between avidin and biotin is the

tightest binding described, with a dissociation constant of

approximately 10-15. Avidin and streptavidin can be detected

by several methods; one method utilizes radioactively labeled

avidin or streptavidin, a second utilizes fluorescently

labeled avidin or streptavidin that can be visualized by

illuminating with fluorescent light, a third method utilizes

enzyme-avidin or streptavidin conjugates which can be

detected by incubation with the proper substrate which can

lead to the production of color or light.

In this work, a membrane impermeant biotin probe which

exhibits reactivity towards free amine groups was utilized.

Once the cell surface was derivitized with the biotin probe,

the membrane fraction was solubilized and biotinylated

proteins recovered with an immobilized form of streptavidin.

Streptavidin was chosen over avidin because of avidin's

extremely high isoelectric point (>10) which can lead to non-

specific interactions. The isoelectric point of streptavidin

is approximately 5, and exhibits much lower non-specific

reactivity in immunoaffinity assays. Once the biotinylated

proteins are recovered, unbiotinylated proteins are removed

and the specifically bound proteins analyzed by SDS-PAGE and

Western blotting.

Specifically, 3T3-L1 adipocytes were rinsed in PBS at

40C and maintained at 4*C. When the cells were equilibrated

to 40C, the PBS was removed and the cells washed in PBS, pH

8.5 at 40C. Cells were then incubated in 2.0 ml of the same

solution in the presence or absence of 0.5 mg/ml sulfo-NHC-

LC-biotin for 1 hour. The solution was then aspirated and

the cells washed in TES to quench any remaining biotin probe.

A membrane fraction was then prepared and solubilized in the

extraction buffer used for immunoprecipitation. After the

removal of insoluble material by centrifugation, the

clarified supernatant was incubated in the presence of 50 pl

streptavidin-agarose for 6 hours at 40C with end-over-end

rotation. The complexes were then washed 5 times in

extraction buffer to remove non-specific interactions.

Specifically bound proteins were released by incubating the

complexes in sample dilution buffer containing 6 M urea and

10% B-mercaptoethanol at 370C for 30 min. The released

proteins were then analyzed by SDS-PAGE and a modified

Western blotting technique.

For the detection of the biotinylated proteins, the gel

was transferred to nitrocellulose as described before.

Following transfer, the blot was briefly stained in amido

black (0.2% w/v in 40% methanol, 10% acetic acid) and

destined. Non-specific binding sites were blocked by

incubating the blot in 5% non-fat dry milk (NFDM) in Tris-

buffered saline containing 0.1% Tween-20 (TBS-T) for 1 hour

at room temperature. The blot was then washed in TBS-T

briefly before a streptavidin-horseradish peroxidase

(streptavidin-HRP) conjugate was added in 0.5% NFDM/TBS-T at

a final concentration of 2 x 10-5 mg/ml for 1 hour at room

temperature. The blot was then extensively washed at room

temperature with TBS-T to remove the unbound and non-

specifically bound streptavidin-HRP. The blot was then

incubated in the ECL Enhanced Chemiluminescent Western

blotting detection reagents according to the manufacturer's

instructions. The developed blot was then exposed to film

for a period of 10 sec 10 min.

Plasma Membrane Fragment Isolation

Another technique which has recently become popular for

the analysis of processes occurring at the plasma membrane is

the isolation of plasma membrane fragments. In this

technique, cells that are plated onto glass coverslips are

adhered to the surface by incubating the cells in poly-L-

lysine of high molecular weight (<100,000 M.W.). The

mechanism of adherence by poly-L-lysine is not exactly known,

but it is thought that the highly positive nature of the

poly-L-lysine polymer acts to bind the negatively charged

plasma membrane and the negatively charged glass surface in

an ionic interaction. Gentle sonication is then used to

remove parts of the cell that are not adhered to the surface,

and results in large fragments of the plasma membrane left on

the glass coverslip. This procedure has been used

extensively for the characterization of the assembly of

clathrin-coated pits which occurs at the inside of the plasma

membrane surface (Lin et al., 1992; Lin et al., 1994) and the

purification of caveolin (Chang et al., 1991). Recently, it

has also been modified for use in the 3T3-L1 adipocytes for

analyzing the translocation of the GLUT1 glucose transporter

to the plasma membrane in response to insulin treatment

(Robinson and James, 1992). This procedure, therefore, can

be used to generate pure plasma membrane fragments which are

free from contamination of endoplasmic reticulum and golgi

membrane fractions.

3T3-L1 cells were grown on 22 mm round glass coverslips

(Fisher) and differentiated as described earlier. Once fully

differentiated, cells were incubated in the presence or

absence of glucose for the indicated period of time. The

cells were then washed 3 times in PBS at 40C. The glass

coverslip was removed and placed into a clean 35 mm tissue

culture before being incubated in the same solution

containing 0.5 mg/ml poly-L-lysine (M.W. > 100,000) for 1

minute. The monolayer was then washed three times at 40C

with 3.0 ml 1/3X sonication buffer [1X sonication buffer; 70

mM KC1, 30 mM HEPES, pH 7.5, 5 mM MgC12, 3 mM EGTA]. This

solution was aspirated and replaced with 1 ml sonication

buffer at 4"C. The monolayer was then immediately placed

under a 1/2-inch tapped flat horn at a distance of 5 mm and

sonicated for approximately 2 sec at power 1.5 on a Heat

Systems Sonicator. The glass coverslip was then washed three

times at 40C in PBS. All last traces of PBS were carefully


removed from the coverslip and tissue culture dish. Sample

dilution buffer (50 pl) was added directly to the coverslip

and the glass scraped with a rubber scraper. The sample

dilution buffer was then recovered and loaded onto a mini

SDS-PAGE gel for separation. The presence of GLUT1 and GRP78

was detected by Western blotting as described above.



Glucose deprivation results in the appearance of a lower

molecular weight protein recognized by anti-GLUTI antiserum

in several cell lines, including 3T3-C2 fibroblasts (Haspel

et al., 1991), 3T3-L1 fibroblasts (Reed et al., 1990), 3T3-L1

adipocytes (Reed et al., 1990; Tordjman et al., 1990; Kitzman

et al., 1994), and normal rat kidney cells (Haspel et al.,

1990). The appearance of this low molecular weight protein

can be prevented by incubation with fructose, suggesting that

sugar metabolism is required to prevent its appearance

(Haspel et al., 1986; Kitzman et al., 1994). In normal rat

kidney cells and 3T3-C2 cells, the migration of this protein

was not altered upon digestion by endoglycosidase F and

therefore was interpreted to be unglycosylated (Haspel et

al., 1986; Haspel et al., 1990). In 3T3-L1 adipocytes,

however, the lower molecular weight protein is glycosylated,

but with a smaller oligosaccharide (Kitzman et al., 1994).

Careful examination of this protein and its relationship to

the normal GLUT1 protein in 3T3-L1 adipocytes has not been

reported. This section describes the characterization of the

normal GLUT1 protein as well as the lower molecular weight

protein that appears in glucose-deprived 3T3-L1 adipocytes.

These analyses will be performed with a combination of

Western blotting, glycosylation inhibitors, tryptic digests,

and glycosidase digestions. The goal of this section is to

test the hypothesis that the lower molecular weight protein

detected by Western blot in glucose-deprived cells is a

second form of the GLUT1 transporter that arises not from

alterations in the core protein, but through alterations in



Characterization of Anti-GT1 Antiserum

In order to confirm that the anti-GTl antiserum

specifically recognized the GLUT1 carboxy-terminal peptide,

dot blots and Western blots were performed. In Figure 3-1 A,

varying amounts of the GT1 peptide (CEELFHPLGADSQV) and GT4

peptide (CSTELEYLGPDEND) were blotted onto dry

nitrocellulose, dried, and detected with the anti-GT1

antiserum. This result indicates that the anti-GT1 antiserum

at a dilution of 1:1000 can detect as little as 0.1 ng GT1

peptide. The antiserum, however, did not recognize the

carboxy-terminal GLUT4 peptide when blotted in a similar

manner, demonstrating that the reaction was specific for the

GLUT1 peptide. Note that the GLUT1 peptide and the GLUT4

peptide share little identity. The observation that the

anti-GLUT1 antiserum does not recognize the GLUT4 peptide

E-) E-1
I 0
o I
0 -H i-H
0 0)
E0 H

0 0r
0C 0

'0 *4m 0 e

0 ,QE-

H 00 *) mc
o 00

* 0 0 0

0 >-1 0 p
a W0

0 t-),4'

4)n m
0p U -H

00 3
- 0 C Ul
-4 m 0

U 04a '

R0.4 -
-4 00 V
r m0. '









a L




S i z

& -

reduces the possibility that this antiserum will recognize

the other transporter normally expressed in the 3T3-L1


The anti-GTl antiserum was then tested in a traditional

Western blot to confirm that the correct protein from 3T3-L1

adipocytes would be recognized. Equal amounts (30 pg) of

membrane protein from normal and glucose-deprived (48 hours)

cells were resolved by SDS-PAGE and transferred to

nitrocellulose. GLUT1 was then detected by Western blot.

The anti-GTl antiserum recognizes a wide band of average

molecular weight of 46 kDa, consistent with the recognition

of a similar band with other anti-GLUT1 antiserum (Haspel et

al., 1985; Haspel et al., 1986; Kitzman et al., 1994) (Figure

3-1 B). It should be noted that the migration of

glycoproteins in SDS-PAGE can deviate from its true molecular

weight according to the size of the oligosaccharide and the

percentage of acrylamide in the resolving gel (Leach et al.,

1980). It is therefore assumed that, despite slightly

different migration, the protein recognized by this antiserum

and other antisera are recognizing the same protein. In

glucose-deprived cells, the lower molecular weight protein,

which migrates at 37 kDa (p37), is also recognized by the

anti-GTl antiserum. The pre-immune serum (PIS) does not

detect either of these bands and shows no other cross-

reactivity to other membrane proteins from 3T3-L1 adipocytes.

Glucose Deprivation and GLUT1 Protein Expression

The effect of glucose deprivation on the expression of

GLUT1 protein was analyzed by Western blotting. 3T3-L1

adipocytes were incubated in glucose-free medium for a total

of 48 hours, and a membrane fraction was prepared at the

times indicated. Membrane proteins (25 pg/lane) were then

resolved by SDS-PAGE and transferred to nitrocellulose. The

anti-GT1 antiserum was then used to probe the protein blot in

order to visualize both the relative amount and molecular

weight of the GLUT1 transporter protein. The normal

glycoform of the GLUT1 transporter from 3T3-L1 adipocytes is

observed at time 0; it migrates with an average molecular

weight of approximately 46 kDa (p46) and exhibits the wide

heterogeneous glycosylation expected (Figure 3-2 A). For the

first 15 hours of glucose deprivation, both the level and the

migration of this protein is relatively unchanged. After 15

hours, however, the appearance of the second immunoreactive

protein at 37 kDa (p37) is observed. After 48 hours of

glucose-deprivation, the total level of immunoreactive

protein is increased by 2-fold. This increase is solely

attributable to the accumulation of the 37 kDa protein

(Figure 3-2 B), since the normal GLUT1 protein does not

change significantly during this period. Also in this

figure, the effect of cycloheximide, a protein synthesis

inhibitor, on the accumulation of p37 was analyzed. The

Western blot indicates that cycloheximide, even when added

Figure 3-2 Effect of Cycloheximide on GLUT1 Expression
During Glucose Deprivation 3T3-L1 adipocytes were incubated
in glucose-free medium for up to 48 hours and membrane
fractions prepared at the times indicated. Panel A Membrane
proteins probed for GLUT1 by Western blot with anti-GTl
antiserum. The right panel shows cells incubated with
cycloheximide (20 pM) added 12 hours after the withdrawal of
glucose. These samples were then analyzed in an identical
manner. Panel B Densitometry of GLUT1 protein from panel A.
These results are representative of at least 4 independent

A Glucose Deprivation, hours
0 3 6 9 15 24 36 48 15 24 36 48

Is I -p37

-cycloheximide +cycloheximide
at 12 hours

B o -l

12 24 36 48
Glucose Deprivation, hours

twelve hours after the onset of glucose deprivation,

prevented the appearance of p37 (Figure 3-2 A, right panel).

It has been generally accepted that p37 is a second GLUT1

protein (Haspel et al., 1986; Haspel et al., 1990; Tordjman

et al., 1991; Kitzman et al., 1994), although a careful

examination of both proteins has not been reported. The

following sections will describe experiments suggesting that

p37 is another form of the GLUT1 transporter only synthesized

in glucose-deprived cells.

Tryptic Digest of GLUT1 Transporter

To determine whether p46 and p37 arise from the same

protein, tryptic digests were performed with the expectation

that peptides generated by proteolysis would be identical if

both p37 and p46 derived from the same protein. It should be

noted that although many tryptic peptides can be generated,

only those retaining the carboxy-terminal region against

which the antiserum was generated could be recognized in the

Western blot. Membrane proteins from normal or glucose-

deprived (48 hours) cells were first isolated. These

membrane proteins were then incubated in the presence of

trypsin (10 Mg) for the times indicated, when the reaction

was stopped with a 10-fold excess concentration of soybean

trypsin inhibitor. These samples were then resolved by 10%

SDS-PAGE and transferred to nitrocellulose. GLUT1 proteins

were then detected with the anti-GTl antiserum in a Western

blot. GLUT1 peptides generated by partial trypsin treatment

Figure 3-3 Tryptic Digest of p46 and p37 3T3-L1 adipocytes
were incubated in the presence (+Glucose) or absence (-
Glucose) of glucose for 48 hours before a total membrane
fraction was prepared. These membrane proteins were
incubated trypsin (10 pg) for the times indicated before the
reaction was stopped with excess trypsin inhibitor. The
digested membrane proteins were then probed for GLUT1 by
Western blot with anti-GTl antiserum. Molecular weight
standards are indicated. This result is representative of at
least two independent experiments.

Incubation time, min 0 0.5 1 2 3 4 5 6 10

45- Aa.I.. ---






- p46
- p37

migrated at relative molecular weights of 25 kDa, 21 kDa, and

17 kDa (Figure 3-3 A and 3-3 B). These peptides,

corresponding to the carboxy-terminal half of the GLUT1

transporter which do not contain the glycosylation site, are

identical in control and glucose-deprived cells. No

additional peptides were observed from glucose-deprived

cells, as would be expected if that protein were of different

length or sequence. Together these data suggest that the

core protein of the normal GLUT1 protein and p37 protein is

very similar.

Effect of Tunicamycin Treatment on GLUT1 Protein

Tryptic digest experiments suggested that the protein

from which GLUT1 and p37 arise is very similar; therefore,

the possibility that the difference in relative molecular

weight between p46 and p37 was a difference in glycosylation

was tested. To prevent N-linked glycosylation of GLUT1, 3T3-

L1 adipocytes were treated with various concentrations (0-10

pg/ml) of the N-linked glycosylation inhibitor tunicamycin

dissolved in 20 mM potassium hydroxide for 24 hours.

Membrane proteins were then isolated and probed for GLUT1 by

Western blot using the anti-GLUT1 antiserum. The normal form

of the transporter, p46, (lane KOH, or 20 mM potassium

hydroxide alone) is drastically different from the

transporter from tunicamycin-treated cells, which migrates at

a relative molecular weight of 36 kDa (p36), in agreement

with the predicted molecular weight from the amino acid

I 0) H

0 4J WH )
o 0 (a -

- iO-4 0
0 d H '4

0 U .0 -I
0 0 0-H
U 0 .oC
0 0O-H 4(
0 0
r e 4 J

HU -
H *O H


00 ) 0

WM & 3E-1
0 4) WO t.

*-4 -H 0 4
am S 0

o E

0 Q)4 r-.4
4H -0 0

Cl H

0 -I m O0

00 *OC
ZH 0 d

O .- r.l

E-I 01 >0
44 O4.4
0 0 0
-0 *.0
d0 U 0 0

*1o c4 r
('44-4 to0
C V 44r. -

4-4-O -'o
*H 0A 0-1 0
r V -H0 (D



0" 1.



sequence (Figure 3-4, + glucose). The next species, p37,

present in the glucose-deprived cells, migrates slightly but

distinctly above the core protein with a difference in

molecular weight of approximately 1-2 kDa (Figure 3-4 -

glucose). The GLUT1 protein resulting from tunicamycin

treatment is identical whether it is derived from normal or

glucose-deprived adipocytes, supporting the conclusion that

the core protein in glucose-deprived adipocytes is identical

in size, and that differences in molecular weight between p46

and p37 arise from differences in glycosylation.

The production of unglycosylated GLUT1 protein provided

an opportunity to follow GLUT1 processing if the effect of

tunicamycin could be removed. A previous report demonstrated

that the processing of the insulin receptor in 3T3-L1

adipocytes could be followed after tunicamycin washout

(Ronnet et al., 1988). To this end, adipocytes were exposed

to tunicamycin (1 pg/ml) for 24 hours in either the presence

or absence of glucose. Following this treatment, the medium

was changed into medium without tunicamycin for 24 hours.

The production of the unglycosylated transporter p36 is

confirmed after exposure to tunicamycin for 24 hours (Figure

3-5). However, after the medium is changed into medium

without tunicamycin, no reduction in the level of the

unglycosylated transporter is observed. The inclusion of 25

mM N-acetylglucosamine was added in order to compete with the

action tunicamycin for the glycosylation process. The

results from this experiment indicate that even with the

Figure 3-5 GLUT1 Expression Following Tunicamycin Washout
Normal or glucose-deprived (24 hours) 3T3-L1 adipocytes were
incubated in the presence of tunicamycin (1 pg/ml) for 24
hours. The tunicamycin-containing medium was then removed
and replaced with either medium containing tunicamycin (1
pg/ml), normal medium, or glucose-free medium containing 25
mM N-acetylglucosamine, for 24 hours. Membrane proteins were
then probed for GLUT1 by Western blot with anti-GTl

2 C
+ J <

4+ 3I

45- ago"
I a 0

- p46

inclusion of N-acetylglucosamine, the production of

unglycosylated transporter could not be prevented. Together,

these data indicate that although tunicamycin resulted in the

generation of unglycosylated GLUT1, the effect of tunicamycin

could not be removed and therefore elimated tunicamycin as a

useful tool in the analysis of GLUT1 processing. The

inability to remove tunicamycin in these experiments is not

consistent with the previous report on the processing of the

insulin receptor, but it may lie in the processing of a

multi-subunit protein such as the insulin receptor is

fundamentally different from a monomeric glycoprotein like


Glycosidase Digestion of GLUT1 Glycoforms

In order to characterize the type of oligosaccharide

present on p46 and p37 GLUT1, membrane proteins from normal

or glucose-deprived cells were digested with either

endoglycosidase H, which cleaves high mannose

oligosaccharides, or N-Glycosidase F, which cleaves all N-

linked oligosaccharides. For comparison of these glycoforms,

membrane proteins from the LEC1 CHO cell line were also

tested. The LEC1 CHO cell line is a glycosylation mutant

that is deficient in N-acetylglucosamine transferase I, the

first committed step towards the production of hybrid and

complex type glycoproteins (Stanley and Chaney, 1985). As a

Em 0

of a-

o e4)

10 (

w -1 44 to

S0 4- -d

4 d- Q501

mOU .

+1 .4 -> 4
Sd ai

0 0 M
0 -1 r.

> 40 i m

0 -0

-a H I )

ma ) a0

m5 (a 0 4

m 0 41 (0
0 4a) 0E )
oVl- -
*r 1 o




+ I


I +

+ I


I +

+ I




result, glycoproteins in this cell line are terminated at the

high mannose stage, and remain endoglycosidase H sensitive.

Figure 3-6 shows a Western blot of these membrane proteins

probed for GLUT1 using the anti-GLUTI antiserum. The

migration of the normal GLUT1 glycoform present in 3T3-L1

adipocytes was not altered by treatment with endoglycosidase

H, indicating that this GLUT1 glycoform is not a high mannose

type glycoprotein. The migration of p46 was altered,

however, by digestion with N-Glycosidase F, confirming

earlier observations that GLUT1 is an N-linked glycoprotein

(Kitzman, et al., 1993). The molecular weight of the N-

Glycanase F digested GLUT1 protein is 36 kDa, in agreement

with the molecular weight predicted by amino acid sequence

and the molecular weight of GLUT1 from tunicamycin-treated

adipocytes. In glucose-deprived adipocytes, neither the

normal nor the low molecular weight GLUT1 glycoform was

sensitive to digestion by endoglycosidase H, indicating that

neither is a high mannose glycoprotein. Once again, the

migration of the normal and low molecular weight GLUT1 was

resolved to the core 36 kDa protein upon digestion with N-

Glycanase F. This data confirms earlier observations that

the difference between the normal and low molecular weight

GLUT1 glycoforms is due to differences in glycosylation and

not changes in the primary sequence of the protein. To

confirm that the endoglycosidase H in this assay was active,

membrane proteins from the LEC1 CHO cell line were also

tested. The GLUT1 glycoform present in these cells (p38),

which was predicted to be a high mannose glycoprotein, was

sensitive to digestion with endoglycosidase H and was

resolved to a 36 kDa core protein. Together, these data

indicate that the low molecular weight GLUT1 present in

glucose-deprived 3T3-L1 adipocytes is glycosylated, but with

a relatively small oligosaccharide (approx. 1 kDa). While

this oligosaccharide is N-linked, it is not a high mannose

structure as might be expected for a normal processing

intermediate prematurely terminated due to the lack of

glucose. This oligosaccharide on the low molecular weight

GLUT1 from glucose-deprived 3T3-L1 adipocytes likely

represents an alternative oligosaccharide generated in the

absence of glucose.

Specificity and Efficiency of GLUT1 Immunoprecipitation

The procedure for immunoprecipitation of GLUT1 from 3T3-

L1 adipocytes was assessed using metabolically labeled

control or glucose-deprived adipocytes. A wide band

migrating at 46 kDa (p46) was immunoprecipitated from

membrane extracts of control cells using purified anti-GT1

antibody (-GT1, + glucose) (Figure 3-7). In contrast, a 37

kDa protein was immunoprecipitated from cells deprived of

glucose for 36 hours (-GT1, glucose). The specificity of

the immunoprecipitation procedure was verified by

immunoprecipitating GLUT1 in the presence of 1 pg of the GT1

peptide. Inclusion of this peptide completely blocked the

immunoprecipitation of p46 GLUT1 in control cells, as well as


S r. o0
U Wa) rO
8 BS

(d DU aO

4O ,0 40-
F a) ao

r-i O ,-o
m fu 4)
0H 1 l 0 H

*.al g'g

\o ., I + v.
-,1 4
4.3 44

0- 4 :3 r r -4

) V *r i' 0) 0

0 4) a.r -43 H.

A4 U

o -. M H
-1 4 W2 Eo

e "., U +, 0)

+) a)o 0 .

ia 0 i o I d

4 a) no

P4 u Z-rq > I + a

'C +



0' +




r- (O

p37 GLUT1 in glucose-deprived cells (+GT1, glucose). In a

similar experiment, a GLUT4 carboxy-terminal peptide

(CSTELEYLGPDEND) did not block the immunoprecipitation of

GLUT1 (+GT4, glucose). GLUT1 protein of either molecular

weight was not observed in membrane extracts

immunoprecipitated with Protein A Sepharose alone, with

preimmune serum, or with 5 pg IgG from an unrelated antisera

(-Ab, PIS, IRR, glucose). GLUT1 protein was also not

observed when extracts were immunoprecipitated with an equal

amount of the unbound antibody fraction (flow through) from

peptide purification of the same antiserum, indicating that

the purification step removed the majority of the GLUT1

specific IgG (FT, glucose).


This chapter described experiments designed to

demonstrate that the second immunoreactive protein that

appears during glucose deprivation of 3T3-L1 adipocytes is a

second glycoform of GLUT1. Tryptic digest of membrane

proteins from normal and glucose-deprived cells indicates

that peptides detected with the anti-GTl antiserum migrated

in an identical fashion. I interpret these data to suggest

that there were no gross alterations in the length of the

GLUT1 core protein. The expression of a different GLUT1

protein, therefore, could not account for the difference in

molecular weight between p46 and p37.

The effect of tunicamycin on the relative molecular

weight of the transporter supports this conclusion. These

experiments indicate that the core protein generated in

tunicamycin-treated cells, regardless of the presence or

absence of glucose, migrated at the same molecular weight (36

kDa). This type of experiment analyzes the GLUT1 protein

before it has an opportunity to be glycosylated. Enzymatic

deglycosylation of p37 and p46, where the oligosaccharide is

removed after glycosylation, also results in the same protein

migrating at 36 kDa, consistent with the tunicamycin

experiment. Additional information about the type of

oligosaccharide on p37 was obtained. Although p37 was

sensitive to digestion with N-Glycosidase F, indicating that

it was an N-linked glycoprotein, p37 was not sensitive to

endoglycosidase H, indicating that this oligosaccharide was

not of the high mannose class. In order to confirm that the

endoglycosidase H reaction was working correctly, digestion

of GLUT1 derived from the LEC1 CHO cell line was performed

and verified that the GLUT1 from this glycosylation mutant

was a high mannose type glycoprotein, despite its similar

migration to p37. Earlier studies utilizing glucose-deprived

CHO cells indicated that an alternative glycosylation pathway

exists, where the cell synthesized a truncated

oligosaccharide core, GlcNAc2MansGlc3, that was subsequently

added to proteins synthesized in the endoplasmic reticulum

(Rearick, et al., 1981). This alternative oligosaccharide

was shown to be endoglycosidase H resistant. The

endoglycosidase H insensitivity and size of the

oligosaccharide on p37 are consistent with this type of post-

translational modification. The biosynthesis and processing

of the transporter in the absence of glucose will be

addressed in more detail in Chapter 4.

Together, these data provide evidence that p37 is an

alternatively glycosylated form of GLUT1. The protein

component is likely unaltered in the absence of glucose.

Furthermore, the difference in relative molecular weight

between p46 GLUT1 and p37 GLUT1 resides in a difference in

the size of the oligosaccharide. p37 GLUT1 is glycosylated

in glucose-deprived adipocytes, although with an

oligosaccharide drastically reduced in size. This

oligosaccharide is likely not a processing intermediate

prematurely terminated in the absence of glucose, due to the

observation that the small oligosaccharide is approximately

the same size as a high mannose core but is endoglycosidase H

insensitive. The next chapter will describe experiments

designed to elucidate biosynthesis, processing, and turnover

of these GLUT1 glycoforms in normal and glucose-deprived 3T3-

L1 adipocytes.



The biosynthesis of integral membrane glycoproteins

involves a complex journey through multiple membrane

compartments containing specifically localized enzymes.

During the translation of polypeptides through the

endoplasmic reticulum, a preformed oligosaccharide is

transferred by the enzyme oligosaccharyltransferase to

asparagine residues on the protein when they occur in the

recognition sequence Asn-X-Ser/Thr (Voet and Voet, 1990).

The transfer of oligosaccharides is regulated by a variety of

factors, including cell type, conformation of the protein

during translation, nutritional status, and distance from the

recognition site to the inside face of the endoplasmic

reticulum (Kornfeld and Kornfeld, 1980).

The pre-assembled oligosaccharide core is synthesized on

both the cytsolic and lumenal faces of the endoplasmic

reticulum. Sugar residues are added sequentially to a lipid

carrier molecule, dolichol pyrophosphate, generating a

growing oligosaccharide core. The first seven sugars, two N-

acetylglucosamine residues and five mannose residues, are

donated from their respective GDP nucleotide derivatives

generating the oligosaccharide dolichol-PP-GlcNAc2Man5. The

second seven sugars, 4 mannose residues and 3 glucose

molecules, are added from their respective dolichol phosphate

derivatives. The completed core oligosaccharide, dolichol-

PP-GlcNAc2MangGlc3, is transferred cotranslationally to the

growing polypeptide chain in the endoplasmic reticulum, and

is identical regardless of the glycoprotein in question.

Cleavage and processing of this oligosaccaride, however,

varies widely depending upon the type of cell, the individual

glycoprotein, and the final target of the glycoprotein.

Processing of the oligosaccharide begins immediately in

the endoplasmic reticulum upon the attachment of the core

oligosaccharide. Two specific glucosidases (glucosidase I

and glucosidase II) remove the first and second two glucose

residues, respectively, generating Asn-GlcNAc2Mang. Recently,

reglucosylation of the core oligosaccharide in the

endoplasmic reticulum has been described, although the exact

purpose of this enzymatic reaction has not yet been clearly

elucidated (Kornfeld R. reference). When the glycoprotein

has moved from the endoplasmic reticulum into the cis golgi,

mannosidases begin the sequential removal of four of the nine

mannose residues, eventually resulting in Asn-GlcNAc2Mans.

For glycoproteins destined to be of the high mannose class,

additional mannose units are added onto this oligosaccharide.

For complex type glycoproteins, however, a variety of other

sugars are then added onto the oligosaccharide, often

creating large and highly branched structures. These

modifications occur in the medial golgi, trans golgi, and the

trans golgi network. Some of the later sugar modifications

catatalyzed are the addition of N-acetylglucosamine, sialic

acid, fucose and nueraminic acid. A third type of

glycoprotein, the hybrid type, exhibits characteristics of

both high mannose and complex oligosaccharides, and often

contains a bisecting N-acetylglucosamine residue (Kornfeld

and Kornfeld, 1980). Sulfation, acetylation, and

phosphorylation of sugars are other types of oligosaccharide

modifications that are possible. For lysosomal proteins,

specific mannose sugars are phosphorylated, and this serves

to act as the appropriate recognition mechanism for targeting

to the lysosomes (Kaplan et al., 1977; Goldberg and Kornfeld,


The glucose-dependent regulation of glucose transport

has been described in many cell types (for review see Klip et

al., 1994). In each of these cell types, a common result of

glucose deprivation is an increase in the maximal velocity

(Vnx) of glucose transport. However, the mechanisms) (i.e.,

transcriptional, post-transcriptional, and post-

translational) responsible for this stimulation vary

according to cell type. Several investigators have

postulated that one component of the transport upregulation

might be an alteration of transporter turnover, particularly

GLUT1 (Yamada et al., 1983; Germinario et al., 1982; Walker

et al., 1989; Shawver et al., 1987; Haspel et al., 1991). In

particular, if the degradation of the glucose transporter

were inhibited in glucose-deprived cells while the synthesis

of the transporter remained constant, the end result would be

an increased number of GLUT1 molecules available for glucose

transport (increasing the Vmx). Few studies, however, have

investigated the processing and turnover of GLUT1 directly

because of the technical difficulties involved (Haspel et

al., 1985). The primary goal of this chapter is to define

the processing pathway of the GLUT1 transporter and determine

the kinetics of GLUT1 turnover in both the presence and

absence of glucose.

In 3T3-L1 adipocytes, glucose deprivation does not

result in a significant change in the level of GLUT1 protein

(as measured by Western blotting) over the period in which

transport induction is observed (Kitzman et al., 1994).

Beyond this time, we and others have reported the

accumulation of a lower molecular weight GLUT1 species which

results in a 2-fold increase in the total GLUT1 mass by 48

hours of glucose deprivation. (Reed et al., 1990; Kitzman et

al., 1994; Haspel et al., 1991). The role of this second

GLUT1 protein in transport stimulation has been investigated

and found not to be responsible for the activation of

transport during glucose deprivation (Kitzman et al., 1994).

However, the origin and characteristics of this lower

molecular weight GLUT1 and its relationship to the normal

GLUT1 glycoform were unknown. Several scenarios are

possible. One possibility is that the low molecular weight

GLUT1 glycoform is a normal intermediate in GLUT1 processing

that accumulates during glucose deprivation. In this case,

the transporter in glucose-deprived adipocytes would be

processed in the same manner as in normal cells, but due to

the inactivation of a processing enzyme or the depletion of a

critical precursor, a premature GLUT1 glycoform becomes the

terminal glycoform. This hypothesis assumes that the core

oligosaccharide produced in glucose-deprived cells is the

same as in normal cells. Alternatively, the low molecular

weight GLUT1 glycoform could arise from the glycosylation of

the core protein with a different oligosaccharide that is

produced only in the absence of glucose. In this scenario,

the oligosaccharide is aberrant before being added to the

protein, and further processing is not observed because the

aberrant oligosaccharide is not suitable substrate for the

processing enzymes found in the golgi apparatus. A third,

less likely, possibility is that the low molecular weight

GLUT1 transporter is a degradation product, that arises in

glucose-deprived cells due to the action of proteases on the

normal transporter, reducing its apparent molecular weight to

37 kDa. The latter is not likely due to the identical

molecular weight that the transporter exhibits either when

incubated in the presence of tunicamycin, when treated with

N-Glycanase F to remove the oligosaccharide, or compared to

the expected molecular weight of the core transporter protein

given by amino acid sequence. To address these issues,

existing methods of GLUT1 immunoprecipitation were modified

to improve recovery and reduce non-specific interactions.

This has allowed a clear model of the processing and turnover

of the metabolically labeled transporter to be obtained in

both normal and glucose-deprived adipocytes.


Synthesis and Glycosylation of GLUT1 in Control and Glucose-
deprived 3T3-L1 Adipocytes

To determine if glucose plays a role in regulating the

synthetic rate of GLUT1, control and glucose-deprived

adipocytes were methionine and cysteine depleted before being

metabolically labeled for 10 to 60 min. GLUT1 was then

immuoprecipitated from solubilized membrane extracts and

analyzed by SDS-PAGE and fluorography. Figure 4-1 A shows

the fluorograph of the GLUT1 immunoprecipitates. Two GLUT1

species were synthesized in control cells during the first 10

min of labeling; a 36 kDa protein, and a glycosylated

intermediate migrating at 42 kDa. The 36 kDa protein is

likely the core protein based on the predicted molecular

weight of the core GLUT1 protein and tunicamycin experiments

(see Chapter 3). When normal adipocytes were labeled for an

additional 10 min, the mature GLUT1 protein (46 kDa) was

observed along with p36 and p42. No further GLUT1 proteins

were observed out to 60 minutes of labeling. When glucose-

deprived cells were labeled for 10 min, only the 37 kDa GLUT1

protein was observed. The 36 kDa core GLUT1 protein is

likely present but obscured due to the width of the p37

Figure 4-1 Biosynthesis of GLUT1 in Normal and Glucose-
Deprived 3T3-L1 Adipocytes Panel A Cells, cultured with or
without glucose for 36 hours, were metabolically labeled with
200 pCi/ml Tran35S-Label for the times indicated as described
in Chapter 2. GLUT1 was then immunoprecipitated from
membrane extracts and analyzed by SDS-PAGE and fluorography.
Panel B Densitometry of total GLUT1 during labeling. Total
GLUT1 intensity was obtained in all experiments by summing
the appropriate bands after densitometry of each species
separately (i.e. p36+p42+p46). Inset, radioactivity in
membrane fraction during labeling relative to 10 min label.
This result is representative of at least two independent

Glucose + + + + -

45- -p46
p- p42
S9P36 P37

10 20 30 60
Labeling Time, min



ioo- 1 6
S100 6

S0 10 20 30 40 50 60

" 0 Control

0 10 20 30 40 50 60

Labeling Time, minControl
- 50- Glucose-deprived


0 10 20 30 40 50 60

Labeling Time, min

protein. Lighter exposures provided no further

discrimination between p37 and any other GLUT1 protein. When

glucose-deprived adipocytes were labeled for 20 min or more,

no additional GLUT1 species were observed. The relative

synthetic rate of GLUT1 in normal and glucose-deprived

adipocytes was measured by adding the densitometric intensity

of all GLUT1 bands at each time point. Although the

intensity of total GLUT1 protein in glucose-deprived cells

was greater than that of control cells at every time point

(Figure 4-1 B), this effect was not specific to GLUT1 when

taking into account an equal increase in radioactivity of the

membrane fraction from glucose-deprived cells (Figure 4-1 B,

inset). This phenomenon, therefore, could represent an

increase in the specific activity of the methionine and

cysteine pools in glucose-deprived cells. Taken together,

these data demonstrate that the synthetic rate of GLUT1 was

unaffected by glucose availability despite the differences in


Regulation of GLUT1 Processing by Glucose

To examine the processing pathway of GLUT1 in control

and glucose-deprived adipocytes, normal or glucose-deprived

adipocytes were metabolically labeled for 10 min and then

chased for a total of 60 min. Cells were harvested for

immunoprecipitation during the indicated times during the

chase period. GLUT1 was then immunoprecipitated from these

membrane extracts and analyzed by SDS-PAGE and fluorography.

Figure 4-2 Processing of GLUT1 in Normal and Glucose-
Deprived 3T3-L1 Adipocytes Cells incubated with (+ glucose)
or without (- glucose) glucose for 36 hours were labeled with
500 pCi/ml Tran35S-Label for 10 min and then chased for a
total of 60 min. GLUT1 was immunoprecipitated from membrane
extracts of each and analyzed by SDS-PAGE and fluorography.
Panel A Control adipocytes chased in the presence of
glucose. Panel B Glucose-deprived adipocytes chased in the
absence of glucose. Panel C Glucose-deprived cells chased
in the presence of glucose. Shown is the result of a single

A Glucose + + + + +

45- i J '. p46
.II8 -p42

B Glucose -

45- I


C Glucose + + + +

45 -

m .4- -p37
Chase Time, min 0 10 20 30 60

In control cells, the 36 kDa and 42 kDa GLUT1 precursors

synthesized during the pulse disappeared concomitant with the

emergence of the mature GLUT1 at 46 kDa after 20 min of chase

(Figure 4-2 A). In contrast, p37 synthesized in glucose-

deprived cells was unchanged in either molecular weight or

intensity when chased in the absence of glucose (Figure 4-2

B). Although this data indicated that p37 GLUT1 was not

processed in glucose-deprived cells, it did not rule out the

possibility that p37 could be processed if glucose were

present. This hypothesis was tested by labeling glucose-

deprived (36 hours) adipocytes for 10 min and then chasing

for 60 minutes in the presence of 25 mM glucose. It was

found that like the experiment that contained no glucose

during the chase, p37 GLUT1 remained unchanged (Figure 4-2

C). These results demonstrate that although 3T3-L1

adipocytes have the capacity to glycosylate GLUT1 in the

absence of glucose, the transporter cannot be further

processed due to either the structure of the oligosaccharide

or the location of GLUT1 along the processing pathway.

Synthesis of GLUT1 Glycoforms During Glucose Deprivation and
Glucose Refeedinq

To determine the time required for glucose deprivation

to affect GLUT1 glycosylation, normal adipocytes were placed

into glucose-free medium and labeled for 1 hour at specific

times during the next 36 hours of glucose deprivation. GLUT1

was then immunoprecipitated from solubilized membrane


0l .4 .3
.. -o
I "


+ 0

+ ^ N ) X

+ co 8
I o ~5

C) 0]

(0 ID
0:*M a 6*


extracts and analyzed by SDS-PAGE and fluorography. This

type of experiment provides a "snapshot" of what GLUT1

glycoform(s) are being synthesized during glucose-

deprivation. Despite glucose withdrawal, the normal p46

GLUT1 glycoform was the only GLUT1 protein synthesized during

the first 12 hours of glucose deprivation (Figure 4-3 A).

Beyond 12 hours of glucose deprivation, however, the

synthesis of GLUT1 shifted to the p37 glycoform.

Densitometry of the GLUT1 bands indicated that the total

amount (addition of all bands) of GLUT1 in glucose-deprived

cells remained nearly constant (Figure 4-3 B). To examine

the reversibility of this effect, cells glucose-deprived for

36 hours were placed into medium containing 25 mM glucose and

labeled for 1 hour during various points during the

refeeding. As expected, p37 was the only form synthesized in

the glucose-deprived cells. Within 6 h of refeeding,

however, the normal p46 GLUT1 glycoform was the only GLUT1

protein synthesized (Figure 4-3 C). Furthermore, the

readdition of glucose resulted in a 3.5 fold increase in

GLUT1 (Figure 4-3 D) despite only a 40% increase in total

membrane radioactivity (data not shown). To examine this

rapid recovery in more detail, we performed the same

experiment on glucose-deprived (36 hours) cells and placed

them into glucose containing medium. Cells were then labeled

every hour for the first 6 hours of refeeding. Surprisingly,

within 1 h of glucose readdition, only the normal p46



U W4)


o (Cgo
o o'-

3 0 4 4
H .0 4 1-

a-4 C
,4i V

0 a

a*H 44 >

O 00


0 Q)

0 0 il
o 0.a
4 -I C) a)

0 -If p -P

oa) e

m 'a C

+ .44 IL

o + Co
+ a


glycoform was synthesized; no metabolically labeled p37 was

observed (Figure 4-4).

GLUT1 Half-Life in Control and Glucose-Deprived Adipocytes

To determine the glucose-dependent regulation of GLUT1

turnover in 3T3-L1 adipocytes, control and glucose-deprived

cells were metabolically labeled for 1 hour and then chased

in either the presence or absence of glucose. Cells were

harvested at various times during the chase period for

immunoprecipitation of GLUT1. GLUT1 immunoprecipitates were

then analyzed by SDS-PAGE and fluorography. The normal GLUT1

glycoform (p46) in control adipocytes exhibited a half-life

of approximately 14 h (Figure 4-5 A, B). When normal

adipocytes were labeled for 1 hour and then chased in the

absence of glucose, the turnover of GLUT1 was similar for

approximately the first 12 hours to that of cells chased in

the presence of glucose. Thereafter, the turnover of GLUT1

was significantly inhibited, exhibiting a turnover time of

greater than 50 hours. Likewise, the aberrant GLUT1

glycoform (p37) synthesized in glucose-deprived cells was not

significantly degraded when chased in the absence of glucose

(Figure 4-6 A, B). Additionally, a second GLUT1 glycoform,

although not present during the initial labeling period (time

= 0) appeared during the chase period. The turnover of this

GLUT1 glycoform, which exhibited a relative migration of

Figure 4-5 Turnover GLUT1 in Glucose-Deprived 3T3-L1
Adipocytes Control adipocytes were labeled with 200 pCi/ml
Tran35S-Label for 1 hour and then chased in the presence (+)
or absence (-) of glucose for a total of 48 hours. GLUT1 was
then immunoprecipitated from membrane extracts and analyzed
by SDS-PAGE and fluorography. Panel A Fluorograph of GLUT1
immunoprecipitates. Panel B Densitometry of total GLUT1
immunoprecipitates from panel A. Results shown are
representative of at least 4 independent experiments.

Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EUOR0LQ6E_HG6FKV INGEST_TIME 2012-09-13T15:05:28Z PACKAGE AA00011803_00001