Acute Regulation of Glut1 Function: the Role Of

Home , GLUT1, Stomatin

ACUTE REGULATION OF GLUT1 FUNCTION: THE ROLE OF

DETERGENT-RESISTANT MEMBRANE DOMAINS

DARRELL RUBIN

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Thesis Adviser: Dr. Faramarz Ismail-Beigi

Department of Pathology

CASE WESTERN RESERVE UNIVERSITY

August 2004

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the dissertation of

______

candidate for the Ph.D. degree *.

(signed)______(chair of the committee)

______

(date) ______

*We also certify that written approval has been obtained for any proprietary material contained therein. Dedication

To my family and all the people providing inspiration or motivation along the way.

Table of Contents

Table of Contents …………………………………………………………………. v

List of Tables and Figures ………………………………………………………... vii

Acknowledgements ………………………………………………………………... ix

List of Abbreviations ……………………………………………………………... x

Abstract ……………………………………………………………………………. xi

Chapter 1 Introduction

Members of the Glucose Transporter Family …………………….. 1

Clone 9 cells and the Regulation of Glucose Transport ………….. 6

Protocols for Subcellular Fractionation …………………………... 8

Glucose Transporter Compartmentalization ……………………… 16

The Activation of Glucose Transporters ………………………….. 20

Membrane Microdomains and Glucose Transport Regulation …… 31

Chapter 2 Distribution of Glut1 in Detergent-Resistant Membranes

(DRMs) and non-DRM domains: Effect of Treatment with

Azide

Introduction ……………………………………………………….. 52

Materials and Methods ……………………………………………. 55

Results …………………………………………………………….. 60

Discussion ………………………………………………………… 65

Chapter 3 Distribution of Glut1 in Clone 9 cells and to Plasma

Membrane DRM and non-DRM Domains: Effect of Azide

Introduction ……………………………………………………….. 79

Materials and Methods ……………………………………………. 82

Results …………………………………………………………….. 89

Discussion ………………………………………………………… 94

Chapter 4 The Effects of Methyl-β-Cyclodextrin on Glut1 and Glucose

Transport

Introduction ……………………………………………………….. 117

Materials and Methods ……………………………………………. 120

Results …………………………………………………………….. 123

Discussion ………………………………………………………… 125

Chapter 5 Summary and Future Studies

Summary ………………………………………………………….. 140

Future Studies ……………………………………………………... 141

Works Cited ……………………………………………………………………... 145

vi List of Figures and Tables

Table 1.1 The glucose transporter family ………………………………... 37

Figure 1.1 Schematics of SLCA2 glucose transporters …………………... 39

Figure 1.2 Dendrogram SLCA2 family glucose transporters ……………... 41

Figure 1.3 Protocols for subcellular fractionation ………………………… 43

Figure 2.1 Low density DRMs in Triton X-100 cell lysate ………………... 69

Figure 2.2 DRMs in cell lysate produced with other detergents …………... 71

Figure 2.3 Low density DRMs in Triton X-100 cell lysate: Azide effect … 73

Table 2.1 The azide effect on glucose transport and Glut1 in the DRMs ... 75

Figure 2.4 Low density DRMs in the plasma membrane: Azide effect …… 77

Figure 3.1 Electron microscopy of preparative gradient fractions …………. 101

Figure 3.2 Protein distribution in Clone 9 cell post-nuclear homogenate …... 103

Table 3.1 Protein distribution in Clone 9 cell post-nuclear homogenate …. 104

Figure 3.3 Glut1 colocalization with other proteins ………………………… 106

Table 3.2 Glut1 colocalization with other proteins ………………………... 107

Figure 3.4 Caveolin-1 in plasma membrane microdomains ………………... 109

Figure 3.5 Glut1 in plasma membrane microdomains: Azide effect ………. 111

Figure 3.6 Glut1 distribution in plasma membrane microdomains …………. 113

Figure 3.7 Glut1 content in plasma membrane microdomains ……………... 115

Table 3.3 Glut1 in plasma membrane microdomains: Azide effect ……… 116

vii Figure 4.1 mβcd blocks azide – induced glucose transport ………………… 130

Table 4.1 Glucose transport by inhibitors of respiration blocked by mβcd ... 131

Figure 4.2 mβcd-treated Clone 9 cells can recover the azide response ……... 133

Figure 4.3 mβcd blocks glucose transport stimulated by ETC inhibition …... 135

Figure 4.4 mβcd decreases Glut1 association with the DRM ………………. 137

Table 4.2 mβcd decreases Glut1 association with the DRM ………………. 139

viii

Acknowledgements

I thank Dr. Faramarz Ismail – Beigi for his patient guidance and his generous dedication to my education and progress. The other members of my committee Dr. George Dubyak,

Dr. Alan Levine, Dr. Sanjay Pimplikar, and Dr. Alan Tartakoff contributed their time, expertise, and guidance for which I am grateful. The department of Pathology, led first by Dr. Micheal Lamm and succeeded by Dr. George Perry, has been a great place for my training. I have had the pleasure of learning from two laboratories during the course of

my graduate education and both provided numerous experiences which have helped my

development as a scientist tremendously. I also thank my immediate family for their

unwavering support during this time in my career as a student. Of my friends, I

especially thank Michael Payne for his support. Finally, I acknowledge the staff and

director of the Medical Scientist Training Program, especially former program

coordinator Felicite Katz and director emeritus Dr. John Nilson, for their support.

List of Abbreviations

Homog homogenization

SDH succinic dehydrogenase

M-J McKeel and Jarret

DRM detergent – resistant membrane

PM plasma membrane

LDM low density membrane

HDM high density membrane

ETC electron transport chain

3H tritium

x ACUTE REGULATION OF GLUT1 FUNCTION: THE ROLE OF

DETERGENT-RESISTANT MEMBRANE DOMAINS

Abstract

DARRELL RUBIN

Identifying which processes and proteins control glucose transport could provide

important clues to understanding and treating a number of clinical entities including

diabetes and some cancers. Glucose transport across the plasma membrane occurs by

either sodium-dependent or independent glucose transporters. In order to study the

mechanisms which control acute changes in glucose transport by sodium-independent

glucose transporters, we use the non-transformed rat liver – derived Clone 9 cell line.

These cells respond to the acute inhibition of oxidative phosphorylation by azide with a

4-6 fold stimulation of glucose transport and a 1.8 fold increase in the amount of glucose transporter 1 (Glut1) in the plasma membrane. In Clone 9 cells under basal conditions,

~38 % of Glut1 in the post-nuclear lysate is localized to the detergent-resistant membrane

(DRM) microdomains. Acute exposure to azide decreased this figure by ~40 %.

In order to examine the effects of azide on Glut1 localization to the plasma membrane of the Clone 9 cell, we performed subcellular fractionation of the post-nuclear homogenates. Approximately 30 % of the Glut1 in the post-nuclear homogenate was recovered in the plasma membrane (PM) compartment and 50 % of this PM Glut1 localized to the DRM fraction. Acute inhibition of oxidative phosphorylation with azide

xi resulted in a 1.6-fold increase in the total abundance of Glut1 in the PM and was associated with a 2.9 fold increase in the abundance of Glut1 in the non-DRM fraction but no significant change in the content of Glut1 in the DRM fraction. We conclude that in the Clone 9 cell Glut1 localizes to detergent-resistant membrane microdomains in the plasma membrane. Moreover, in these cells the azide – induced increase in glucose transport is associated with an increase of Glut1 abundance in the non – DRM fraction of the plasma membrane and a decrease of Glut1 association with the DRM fraction of a membrane compartment other than the plasma membrane. These findings indicate that the distribution of Glut1 to the DRM and non-DRM domains of the cellular membrane compartments contribute, in part, to the mechanisms of glucose transport regulation in response to the acute inhibition of oxidative phosphorylation.

xii Chapter 1

Introduction

Abnormal responses to increased blood glucose are an important clinical signal suggesting the patient may be at risk for or already suffer from pathological processes leading to diabetes (Bjornholt et al., 2001). It has also been suggested that the prognosis of patients with certain tumors correlates with the expression of Glut1 (Kurokawa et al.,

2004; Vordermark et al., 2003). Identifying which processes and proteins control glucose transport may facilitate the development of improved therapies for managing the course of disease in these patients.

Members of the Glucose Transporter Family

There are many members of the facilitative glucose transporter family (gene name Solute

Carrier (SLCA) 2A) designated Glut1 – 14 including the proton/myoinositol co- transporter (HMIT) (Table 1.1) (Joost et al., 2001; Wood et al., 2003b). These integral membrane proteins have 12 transmembrane domains and a glycosylation site present in one extracellular loop. The most recent additions to the Glut family were identified based on the greater than 28% identity of their primary sequences with that of Glut1 and the presence of the so-called “sugar transporter signature” sequences (Figure 1.1) which are conserved among Glut1 – 5 (Joost et al., 2001). The family was subdivided into three classes based on sequence identity (Figure 1.2). The following short discussion regarding the Glut family summarizes the excellent reviews provided by G. Joost and B.

1 2

Thorens (Joost et al., 2001) as well as work from several other groups (Mueckler, 1990;

Mueckler, 1994).

Glut1, the founding member of the Class I subfamily, is expressed in many diverse cell types but most abundant in erythrocytes, vascular endothelium, and the basolateral membranes of intestinal and kidney epithelial cells (Table 1.1) (Mueckler,

1994; Zhang et al., 1999a). Expression is broader in fetal tissues and includes the placenta. The transporter has an intermediate affinity for glucose relative to other Class I members and is also believed to transport dehydroascorbic acid, a precursor of Vitamin C

(Brown et al., 2002; Vera et al., 1993; Vera et al., 1994). Its subcellular distribution is broad in non-confluent cultured cells (Rubin et al., 2004); and some have even reported it in the nucleus although there is no speculation as to what it may be doing there

(Pantaleon et al., 2001). The expression and function of Glut1 are regulated by a variety of stimuli and agents including hypoxia and inhibitors of oxidative phosphorylation

(Bashan et al., 1992; Behrooz et al., 1997; Hamrahian et al., 1999; Ismail-Beigi, 1993;

Mercado et al., 1989; Ouiddir et al., 1999; Shetty et al., 1993a; Shetty et al., 1992; Shi et al., 1995). Many believe that Glut1 provides much of the basal, non-insulin dependent glucose transport in tissues; however, further analysis of the recently discovered members of the Glut family may challenge this assertion (Mueckler, 1994; Takata, 1996).

Glut2 is expressed in hepatocytes, pancreatic beta cells, and absorptive epithelia.

Its low affinity and high - capacity transport activity allow intracellular glucose to closely reflect plasma glucose levels in those cells. These properties are believed to facilitate its role in the regulation of glucose homeostasis. Glut3 is expressed in neurons, muscle, hepatocytes, kidney, and placental vascular endothelial cells (Stuart et al., 2001). The 3

Glut3 gene is believed to have been duplicated resulting in the newly discovered gene designated Glut14 (Wu et al., 2002a). Glut4 is expressed in insulin-responsive tissues such as adipose, skeletal muscle, and cardiac muscle. The majority of Glut4 is retained in intracellular membrane compartments until cells are exposed to insulin, or some other stimuli, whereupon some of this intracellular pool translocates to the plasma membrane.

The founding member of the Class II glucose transporters was Glut5, a fructose transporter, which is expressed in intestine, kidney, muscle, and adipose (Kayano et al.,

1990). It was joined by Glut7, 9 and 11 in the mid 1990s (Joost et al., 2001). The predicted amino acid sequence of Glut7 is 55% identical to that of Glut5. When heterologously expressed in Xenopus oocytes, it transports glucose as well as fructose; however, in contrast to Class I Glut proteins, this transport is insensitive to cytochalasin

B and phloretin (Li et al., 2004). There have been no reports of a characterization of

Glut9 activity (Phay et al., 2000). Glut11 is ~40% identical to Glut5 and its overexpression increased glucose uptake which could be partially inhibited by fructose

(Doege et al., 2001; Wu et al., 2002b). Membranes of Glut11- overexpressing cells demonstrated cytochalasin B-inhibitable glucose transport when reconstituted into liposomes (Wu et al., 2002b).

Class III Glut proteins (Glut 6, 8, 10, and 12) are substantially different from both

Class I and II proteins as the glycosylation site in Class III proteins is in extracellular loop 9 instead of loop 1. Glut6 and 8 have intracellular retention motifs and remain intracellular even upon overexpression (Joost et al., 2001). Mutation of these intracellular retention motifs allowed surface expression and increased glucose transport.

Both transporters also demonstrated glucose inhibitable - cytochalasin B binding activity 4

(Doege et al., 2000a; Doege et al., 2000b; Lisinski et al., 2001; Murata et al., 2002).

Insulin-induced translocation of Glut8 to the plasma membrane was reported for murine blastocysts (Carayannopoulos et al., 2000) but this has not been reported for 3T3-L1 adipocytes which also express Glut 8 (Scheepers et al., 2001).

Heterologous expression of Glut10 in Xenopus oocytes demonstrated cytochalasin

B- and phloretin-inhibitable glucose uptake which could be competed with galactose but not fructose (Dawson et al., 2001). 3T3-L1 adipocytes express Glut10 but not Glut12; while the human SGBS adipocyte cell line expresses both (Wood et al., 2003a). Similar to Glut6 and 8, Glut12 has intracellular retention motifs; however, they are at both the N and C termini. In spite of these motifs, heterologous expression in Xenopus oocytes demonstrated cytochalasin B – inhibitable glucose uptake that could be partially inhibited by fructose (Rogers et al., 2003; Rogers et al., 2002). Glut12 is also expressed in the

MCF-7 human cell line.

Of all the newly identified Glut family members, HMIT is certainly the most conspicuous. Its predicted sequence has three intracellular retention signals and three glycosylation sites. It shares the highest sequence similarity with sequences from

Caenorhabditis elegans and Arabidopsis thaliana (Uldry et al., 2001). No hexose activity was demonstrated when a mutant HMIT lacking one or all the retention signals was expressed in Xenopus oocytes; however, there was cytochalasin B- and phloretin – inhibitable uptake of myo-inositol.

The identification of more Glut family members has helped investigators to identify and study the conserved residues which confer substrate binding and functional properties (Joost et al., 2001). There is extensive literature on the topic of Glut1 structure 5

and the examination of the primary sequence which has helped establish the role of

particular residues and conserved motifs in this protein (Hruz et al., 2001). Mutation

strategies, especially when used with affinity - labeling reagents selective for either the extracellular or intracellular substrate binding sites, have provided much of the correlative evidence with regard to Glut1 structure – function relationships.

Crystallographic or NMR data on the three-dimensional structure of glucose transporters have not appeared; however, this does not minimize the importance of the conclusions drawn from the aforementioned body of work or from proposed 3D models derived from studies of the kinetics of ligand binding (Cloherty et al., 2001; Hamill et al., 1999;

Sultzman et al., 1999) or algorithmic predictions from the primary sequence (Zeng et al.,

1996; Zuniga et al., 2001).

Clone 9 cells and the Regulation of Glucose Transport

Glut1 is believed to mediate glucose transport in human red blood cells, fibroblasts, and

Clone 9 cells, a non-transformed, rat liver-derived cell line. With regard to recent

discoveries (Joost et al., 2001; Wood et al., 2003b), there are no published data to exclude the expression of class 2 and 3 Glut family members in Clone 9 cells, e.g. Glut3, judging from preliminary evidence from this laboratory. As these data are neither well developed nor complete, this evidence will not be considered further in this dissertation.

Exposure of Clone 9 cells to sodium cyanide or sodium azide, both inhibitors of cytochrome a3 in the cytochrome c oxidase complex of the electron transport chain,

increases lactate production in a time - dependent manner (Mercado et al., 1989; Shi et al., 1995). This increase in anaerobic metabolism is driven by a biphasic increase of glucose transport. The acute phase of the response is observed within 1-2 hours and mediated entirely by post-translational mechanisms while the chronic phase (2 – 24 hours) involved an increase in Glut1 transcription, translation, and plasma membrane expression (Mercado et al., 1989; Shetty et al., 1993a; Shetty et al., 1992; Shetty et al.,

1993b). The acute phase increase of glucose transport was not accompanied by a corresponding fold increase in the abundance of Glut1 in the plasma membrane (Mercado et al., 1989). Further, the Vmax of transport increased while neither the Km for glucose nor

the energy of activation for transport changed (Hamrahian et al., 1999). These

characteristics suggest that both activation (“unmasking”) of Glut1 pre-existing in the

plasma membrane and translocation of intracellular transporters to the plasma membrane

contribute to the acute stimulation of glucose transport. While the concept of glucose

transporter activation has had detractors (Holman et al., 1990), possible mechanisms for 7

the activation of glucose transport have been studied for many years (Furtado et al., 2002;

Kandror, 2003).

Redox changes have not been excluded as a possible mechanism whereby Glut1 function may be regulated; however, several studies suggest they have little or no contribution to the regulation of glucose transport by azide. Thus, while it is known that

H2O2 stimulates glucose transport in addition to a host of other responses in Clone 9 cells

(Grune et al., 2001; Grune et al., 1995; Grune et al., 2002a; Grune et al., 2002b; Prasad et

al., 1999), studies using thiol-containing antioxidants suggest that changes in the redox

state of the cells may not be the mechanism whereby azide stimulates glucose transport

(Hamrahian et al., 1999). Preliminary studies from this laboratory with N-acetylcysteine

have produced a similar finding (data not shown). In Xenopus oocyte studies, mutants

of Glut1 in which each cysteine was mutated individually or together retain transport

function; however, cys-209 deserves additional consideration as it is in a membrane

proximal position and may be acylated (Hartel-Schenk et al., 1992). Acylation has been

known to facilitate protein-protein interactions, membrane microdomain association, and

protein activity in general (Brown et al., 2000; Charrin et al., 2002; Loisel et al., 1996). 8

Protocols for Subcellular Fractionation

The mainstream proposals for the mechanisms of increased glucose transport have been

tested in several mammalian cell types including primary rat adipocytes, mouse 3T3-L1

adipogenic cells, rat myogenic cells, and Clone 9 cells. For adipocytes and muscle cells,

it seems fairly clear that some stimuli which increase glucose transport cause glucose

transporters to translocate from intracellular compartments to the plasma membrane. In addition to this widely reported and accepted phenomenon, there is evidence for an increase in the intrinsic activity of transporters at the plasma membrane. Reports regarding this topic will be considered in some detail below. The foundation of all studies concerning the mechanisms of glucose transport regulation is the separation of cellular glucose transporters into compartments, typically an intracellular pool and a plasma membrane pool. Various techniques have been used for this separation including biochemical subcellular fractionation and microscopy.

Early reports of subcellular fractionation established the power of combining differential centrifugation with density-gradient centrifugation for the preparation of distinct membrane compartments from homogenates (Boone et al., 1969; Chauveau et al.,

1962; Davoren et al., 1963; Holter et al., 1953; McKeel et al., 1970; Rodbell, 1967 ;

Rothschild, 1961). These early studies identified compartments mainly by morphology, biochemical composition (including nucleic acid and phospholipids), and some enzyme activities e.g. adenylyl cyclase (Chauveau et al., 1962; Rothschild, 1961). The physical

parameters defined by some of the more seminal reports have assumed an almost

definition status for the compartments they describe. 9

In studies of the compartmentalization of adenyl cyclase activity, Davoren and

Sutherland wrote that the separation of nuclei from membranes could be performed

“efficiently by utilizing a dispersion procedure which extensively fragments the

membranes and yet does little damage to the nuclei” (Figure 1.3 a) (Davoren et al.,

1963). Layering this dispersed material, or homogenate, onto a 20% glycerol cushion

before centrifugation allowed the DNA to sediment while adenylyl cyclase activity

remained in the supernate. Further fractionation of this supernate by differential

centrifugation and continuous density-gradient centrifugation revealed one major peak of

adenyl cyclase activity through a range of density peaking at 1.18. The separation of

DNA, i.e. the nuclei, from other less dense materials and the definition of a parameter

(density) describing adenylyl cyclase activity in the homogenate were strong

demonstrations of the usefulness of density centrifugation in conjunction with differential

centrifugation for subcellular fractionation.

In 1967, Rodbell introduced a hypotonic lysis protocol for the preparation of

cellular “ghosts” from primary rat adipocytes for the stated purpose of recapitulating the

semi-permeable nature of the plasma membrane in vitro so as to facilitate the study of hormone effects on membrane transport processes (Rodbell, 1967). While these “ghosts” achieved the stated goal, they contained significant quantities of nuclei, mitochondria, and other membrane populations. In 1969, Boone and colleagues reported a protocol which improved the Rodbell technique for the purpose of obtaining a more homogenous plasma membrane fraction (Figure 1.3 b). In this protocol, HeLa cells were first homogenized in a hypotonic buffer instead of being passively lysed in hypotonic buffer.

The given rationale was that this step would prevent “morphologically indistinguishable 10

smooth vesicles derived from a number of organelles within the cell” from contaminating

the preparation of plasma membrane vesicles (Boone et al., 1969). Although it was a large assumption that one could distinguish plasma membrane vesicles from intracellular

membrane vesicles based on morphology, the procedures set by this study represent the

beginning of a gold standard for subcellular fractionation.

The Boone protocol isolated what were called “cell ghosts” by floating a low-

speed pellet of the homogenate through a sucrose-step gradient. The reported loss of 45

% of the surface material (the details of this determination are delineated below) at this

point suggested that any quantitative identification of compartmentalized proteins or activities would be a lower limit. The ghosts were sonicated and centrifuged at low- speed to remove particles that were assumed to be “either much larger or denser than the plasma membrane vesicles…” thereby preserving the “spherical vesicles 100 – 500 µm in

diameter…definitely identified as being derived from the microvillus-like projections of

the plasma membranes …”. The supernate from the low speed centrifugation was further

fractionated on a 20 – 50% continuous sucrose gradient producing three zones of

apparently arbitrary division. Analysis of these zones by electron microscopy found zone

one (which included the peak absorbance values at A280) to consist of “empty smooth

vesicles” with occasional ribosomes. Zone two consisted of smooth vesicles with

occasional ribosomes, larger smooth vesicles, and mitochondria. Zone 3 contained

mitochondria and “dilated vesicles of rough endoplasmic reticulum.” Based on the

relative homogeneity of the morphology of zone 1 contents (the average density of those

fractions was 1.137) and relative lack of other elements, it was determined to consist of primarily plasma membrane vesicles. In an attempt to explain the apparent presence of 11

the so-called plasma membrane vesicles in zones 2 and 3, the authors suggest that their

association with ribosomes resulted in their apparently greater density. The weakness of

the morphology argument withstanding, these findings are unmistakably the foundation

of current subcellular fractionation protocols and represent a standard as evidenced by the

citation record (McKeel et al., 1970; Simpson et al., 1983).

Curiously, the most important finding of this paper was not mentioned in the

discussion of the findings. Surface labeling with 125iodine-labeled horse immunoglobulin raised against whole HeLa cells allowed the enrichment and distribution of surface material to be tracked throughout the protocol. It was reported that the peak activity of

125I per unit protein was found at a density of 1.154 (35.1 %, 0.54 M sucrose) and

enriched 49 fold. This density value disagreed with the density of 1.137 (31.68 %, 0.49

M sucrose) for morphologically-identified plasma membrane fractions. Surprisingly, the

authors attempted to explain the discrepancy by noting the density identified by surface

labeling would probably be greater due to large mass of antibody that would be required

to give the high specific activity they observed at that density. In spite of these disparate

results regarding the density of surface material, the density of 1.14 was cited by McKeel

and Jarett (McKeel et al., 1969) in their report describing the biochemical fractionation of

primary rat adipocyte membranes and adenylyl cyclase activity and cited again by

Cushman and colleagues regarding the action of insulin on glucose transporter

translocation (Simpson et al., 1983).

The McKeel-Jarett (M-J) homogenization protocol further improved upon the use

of density gradients for membrane compartment fractionation by reducing the pre –

preparative gradient fractionation procedures (Figure 1.3 c) (McKeel et al., 1970). They 12

used adenylyl cyclase activity as a marker of the plasma membrane (an assertion that has

been challenged by other groups) (Buck et al., 1999; Davoren et al., 1963; Parkinson et

al., 2001; Rodbell, 1967; Schulze et al., 1998), and several other biochemical markers to

monitor the enrichment and distribution of other membrane compartments e.g. the nucleus (DNA); mitochondria (succinic dehydrogenase (SDH) activity); endoplasmic

reticulum (RNA and NADH-cytochome c reductase activity). Their protocol differed

significantly from the Boone protocol in that the material applied to the preparative linear sucrose gradient was the post-nuclear homogenate instead of a more processed material; however, even this level of processing should remove a substantial quantity of material

(as evidenced by the Boone protocol) resulting in data that may represent a lower limit in

terms of protein quantity or enzyme activity per fraction. After centrifugation through

the preparative linear sucrose gradient, the particulate matter observed at densities 1.14

and 1.18 were recovered and called the “plasma membrane” (citing the work of Boone et

al) and the “gradient mitochondria” (based on their own characterization) respectively.

Interestingly, they do not mention the discrepant 1.154 density of the plasma membrane fraction reported by Boone et al.

The biochemical studies of the fractions derived from this protocol demonstrate that the enrichment (or activity per unit protein) of markers in the plasma membrane fraction (PM) were consistent with the contemporary understanding of how these markers should be distributed although no 5′ nucleotidase activity was detected in any fraction.

Moreover, it was clear that the fractions overlapped to some degree with respect to the compartments they represented. DNA was low in the PM (0.58% of the total homogenate) and high in the nuclear fraction (85%) as expected. SDH specific activity 13

was approximately 90% lower in the plasma membrane fraction than in the gradient mitochondria and microsomal fractions. The specific activity of RNA and NADH

cytochrome c reductase in the PM were approximately 30% of that in the microsomal fraction. The specific activity of adenylyl cyclase in the absence of sodium fluoride was substantial in the PM while little to none was detected in either the microsomal or mitochondrial fractions. In the presence of sodium fluoride, adenylyl cyclase was observed in both mitochondrial and microsomal fractions but still substantially lower (70 and 92% lower respectively) than that observed in the plasma membrane fraction under the same conditions.

In spite of the reported reliance on the 1.14 density for isolation of the plasma membrane fraction, the biochemical evidence for the segregation of nuclear and mitochondrial compartments is very clear. The incomplete fractionation of some marker activities, e.g. NADH-cytochrome c reductase activity and adenylyl cyclase activity, suggested that either the compartments represented by these fractions have intrinsic activity or the compartment(s) which represent(s) the true location of these activities co- fractionates with other compartments. In spite of this unknown, the protocol was an improvement and represented an advance in the field.

Thirteen years later in 1983, Simpson and colleagues used a biochemical fractionation technique very similar to that described by McKeel and Jarret in order to examine the effects of insulin on glucose transporters in various membrane compartments

(Figure 1.3 d) (Simpson et al., 1983). Interestingly, they used a step gradient instead of a continuous gradient for the preparation of the plasma membrane and other fractions; however, they reported neither their rationale for this departure from the McKeel and 14

Jarret protocol nor studies of the equivalence of the two methods. In any case, they indicated that the revised method did not completely separate all subcellular compartments; however, they demonstrated substantial enrichment of various activities in certain fractions which were then labeled according to the presumed origin of that activity, i.e. adenylyl cyclase activity is attributed to the plasma membrane. The idea that adenylyl cyclase activity represented the plasma membrane was taken to a logical extreme when the 5′-nucleotidase activity in the mitochondrial fraction was corrected for what they called “plasma membrane contamination” based on the co-fractionating adenylyl cyclase activity. In spite of this zealous use of correction factors, the lack of detection of 5′-nucleotidase in the plasma membrane fraction of the same cell type by

McKeel and Jarret is not mentioned.

Considering all of the protocols discussed here, it is apparent that significant enrichment of various subcellular compartments became achievable by combining differential centrifugation with density-gradient centrifugation. While there have been many adaptations of these protocols, their application has remained essentially the same.

Modernity has introduced immunoblots to supplement biochemical activity assays; however, the classic protocols have remained the standard.

Of the significant criticisms of the technique, the fastidious requirement for the consistency of its application by individual investigators looms large. In order to compare the experimental results of different reports, one must have some reasonable estimation of the equivalence of the fractions upon which they report; however, it must also be considered that different cell types may have differences in the expression and compartmentalization of certain proteins and activities. Many factors, including 15 homogenizer clearance and application, affect the yields and co-purification of subcellular membrane compartments thus reinforcing the notion that these protocols bring about enrichment as opposed to purification and that compartments must be expected to overlap to some degree (Joost et al., 1988; Lee et al., 1999; McKeel et al.,

1970; Weber et al., 1988). While keeping the history and criticisms of this technique in mind, it is certainly true that, in spite of its shortcomings, it has been indispensable for the many major strides taken towards understanding the regulation of glucose transport. 16

Glucose Transporter Compartmentalization

The early studies of glucose transporter abundance in the plasma membrane relied

on non-immunologic methods of detection. The observation that cytochalasin B could

inhibit glucose transport and, indeed, bind to red blood cells in a glucose-inhibitable

fashion inspired the development of this reagent for the purposes of glucose transporter

detection (Czech, 1976; Lin et al., 1974). Wardzala et al used 3H-cytochalasin B to

demonstrate an insulin-induced increase in glucose-inhibitable cytochalasin B-binding sites (KD of 120 nM) in the plasma membrane of primary rat adipocytes isolated using

the method of McKeel and Jarret (M-J protocol) (Wardzala et al., 1978). This study was

advanced in 1980 when an analysis of glucose-inhibitable cytochalasin B binding sites

(CB binding) demonstrated not only more binding sites in the microsomes relative to the

plasma membrane but also an insulin-induced decrease in these sites (Cushman et al.,

1980; Suzuki et al., 1980). With these core findings, it was proposed that the mechanism

for insulin action involved the inducible translocation of transporters from an intracellular

pool to the plasma membrane; a proposal that came to be described as the “recruitment

hypothesis” (Cushman et al., 1980; Oka et al., 1984).

Czech and colleagues advanced the use of cytochalasin B as an affinity reagent

for the detection of glucose transporters when they reported its ability to covalently

crosslink to its binding proteins in human red blood cells upon irradiation with ultraviolet

light (Carter-Su et al., 1982). This property was exploited in an effort to more rigorously

demonstrate the recruitment of intracellular transporters to the plasma membrane of

primary rat adipocytes as described by Wardzala and Cushman (Oka et al., 1984). After

exposing intact adipocytes to 3H-CB and UV light, relatively crude fractions of the 17

plasma membrane were isolated using a truncated form of the McKeel and Jarret protocol

which suggests that the CB labeling observed was an upper limit of the binding that

would have been expected in a more refined plasma membrane preparation. The microsomes were prepared from the material remaining after the crude plasma membrane fraction was removed and fractionated into the so-called high and low density microsomes (HDM and LDM respectively) by differential centrifugation. The names of these compartments appear to be misnomers since their isolation appears to have little to

do with density (unless the size of all microsomes are equal).

Czech and colleagues reaffirmed the recruitment hypothesis and further found an

insulin-induced decrease in the same parameter in the microsomal fraction (it is not clear

from their writing whether this was observed in either the LDM fraction or the total

microsome preparation). They also advanced the field by using a poorly membrane-

permeable glucose analogue, ethylidene glucose, to prevent the CB labeling of surface

glucose transporters while allowing the labeling of intracellular transporters. This elegant

experiment definitively demonstrated that intracellular glucose transporters, represented

by the LDM fraction, were diminished upon insulin treatment of primary rat adipocytes

with insulin thereby supporting the notion of glucose transporter trafficking in response

to the hormone.

Another membrane-impermeant affinity reagent used in the study of glucose

transporters is the relatively hard to acquire tritiated ATB-BMPA (2-N-[4-(1’-

azitrifluoroethyl)benzoyl]-1,3-bis-(D- mannos-4-yloxy)-2- propylamine) developed by

Holman et al (Holman et al., 1988; Holman et al., 1990; Holman et al., 1986; Holman et

al., 1982; Parkar et al., 1985). Protocols combining labeling with this reagent, subcellular 18

fractionation, and/or immunoprecipitation of specific glucose transporters allowed

detailed studies of insulin’s effects on glucose transporters.

Yet another technique for detecting plasma membrane proteins is surface biotinylation (Busch et al., 1989; Hare et al., 1989) . Although the direct observation of

glucose transporter biotinylation is lacking, the technique has proved very useful for

several investigators (Barros et al., 2001; Shetty et al., 1993b). One caveat to the

application of this method is the possibility that differences observed between control and

treated cells potentially could be due to changes in the protein or lipid microenvironment

whereby the proteins which are biotinylated become more or less accessible.

Microscopic studies of plasma membrane proteins took advantage of firm cell

attachments to various surfaces to make plasma membrane lawns for immunofluorescent

labeling. Although several protocols for producing these lawns have been described, the

core technique involves growing cells on some substratum, such as poly-L-lysine - coated

plastic, and removing the apical, non-attached membrane compartments by either

sonication or other shearing techniques, hypotonic lysis, rapid freeze-thaw cycles, or

some combination of these (Koseoglu et al., 1999; Moore et al., 1987; Robinson et al.,

1992; van den Berghe et al., 1996).

The study of glucose transporter trafficking derived great benefit from the

contributions of several groups of innovative thinkers. The discovery that membrane

compartments could be separated by mass and density was probably the most important

observation for the development of the leading theories in the field. The combination of

differential and density-gradient centrifugation with affinity labeling allowed the

transporters to be more confidently divided into surface and intracellular pools leading to 19 the famous recruitment hypothesis for the mechanism of insulin-stimulated glucose transport. Advances and discoveries in the field of membrane compartmentation, as well as in the field of microscopy, promises to further refine the picture of glucose transporter disposition especially as regards the nature of its regulation at the plasma membrane. 20

The Activation of Glucose Transporters

As the study of glucose transport regulation progressed, it became apparent that under

certain conditions, changes in glucose transport did not completely match changes in the

abundance of glucose transporters in the plasma membrane fraction. These observations

led to a controversial proposal that in addition to, or in some cases instead of, transporters

trafficking between the plasma membrane and an intracellular compartment to facilitate

changes in glucose transport, there was regulation of transporter intrinsic function. It is

worthwhile to mention that since Glut6 – 14 and HMIT were only recently discovered

they could not have been considered at the time these studies were published; however,

their expression and potential role should be investigated as the proper antibodies or other

detection reagents become widely available (Kandror, 2003; Lisinski et al., 2001; Wilson

et al., 1995; Wood et al., 2003a).

Studies regarding the regulation of intrinsic transporter function have

encompassed many different model systems including whole animals, primary rat

adipocytes, the mouse 3T3-L1 adipogenic cell line, and others. It is the purpose of this

section to review the literature regarding this type of regulation and present the most

current understanding of its relevance.

One of the first reports to demonstrate the above mentioned discrepancy

concerned the insulin response of primary rat adipocytes (Karnieli et al., 1981).

Cushman and colleagues reported a 1.5 minute lag between the t1/2 of the increase in CB binding sites in the plasma membrane fraction (t1/2 = 4 minutes) and that of the increase in

glucose transport (t1/2 = 2.5 minutes). These data implied a step, in addition to

translocation, was required to make glucose transporters fully functional at the plasma 21

membrane. These authors alluded to the possibility that their report may have included some methodological error; however, their methods were the state of the art at that time.

In a later report, they tested the contribution of an intracellular pool of glucose transporters to insulin-stimulated glucose transport by mixing plasma membrane vesicles with an excess of so-called low density microsomes, LDMs (Joost et al., 1988). The contribution of these microsomes to glucose transport activity was low and thus it was concluded that the discrepancy between the increase in transporter abundance and transport activity would remain unexplained even with LDM-contamination of the plasma membrane fraction. An earlier report contradicted the finding of a relative low activity in the microsomal pool of transporters; however, the preparation was somewhat different in that the glucose transporters were reconstituted into liposomes before activity was measured (Suzuki et al., 1980). As the field developed, the criticism of LDM- contamination of the plasma membrane fraction became still less of a concern due to the development of more and better techniques for the study of surface glucose transporters.

Supporting the concept of intrinsic glucose transporter regulation, it was demonstrated in the mid 1980’s that stimulating the β - adrenergic receptors of primary rat adipocytes with isoproterenol in the presence of adenosine deaminase partially inhibited insulin-stimulated glucose transport without affecting the translocation of CB binding sites to the plasma membrane fraction (Green, 1983a; Green, 1983b; Joost et al.,

1986; Kuroda et al., 1987; Smith et al., 1984). These studies suggested that extracellular adenosine contributed to insulin stimulation glucose transport but not transporter transloction. In fact, adenosine deaminase has also been shown to decrease the affinity labeling of surface glucose transporters with ATB-BMPA (Vannucci et al., 1992), which 22 is believed to interact preferentially with active transporters (this concept will be discussed in more detail below) (Barnes et al., 2002; Harrison et al., 1992), thereby suggesting its ability to regulate transporter intrinsic activity.

In 1987 the concept of glucose transporter activation was demonstrated in an animal model when Kahn et al reported an anomalous 3 fold increase of the insulin response of rat epididymal adipocytes isolated from streptozotocin-diabetic rats treated with insulin 7 – 8 days (Kahn et al., 1987). While the in vivo insulin exposure did increase the expression of intracellular transporters resulting in an ~30% higher recruitment to the plasma membrane in response to insulin, this change did not account for the 3-fold increase in glucose transport with “little change” in the Km. Oddly, these authors used a one hundred fold higher dose of insulin for the studies of cytochalasin B- binding in subcellular fractions as compared to the dose used for glucose transport studies. This difference may suggest a further discrepancy between glucose transporter abundance and glucose transport as the difference between the two might have been still greater if the CB-binding assays were performed on fractions from adipocytes stimulated with a lower dose of insulin.

Many other studies using different techniques as well as different cell types have confirmed the initial observation of Cushman and colleagues (Karnieli et al., 1981) regarding the discrepancy between the stimulation of glucose transporter translocation to the plasma membrane and the stimulation of glucose transport (Clancy et al., 1991; Clark et al., 1991; Diamond et al., 1993; Fisher et al., 1996; Gibbs et al., 1988; Harrison et al.,

1991; Harrison et al., 1990; Harrison et al., 1992; Joost et al., 1988; Kanai et al., 1993;

Somwar et al., 2001; Zhang et al., 1998). In spite of this, there are some studies, even 23

from the Cushman lab, that do not agree with these findings (Holman et al., 1990). In

1990, antibodies directed against Glut1 and Glut4 had become available and were used in

combination with the tritiated ATB-BMPA affinity reagent in order to observe clearly the

disposition of the two transporters in the plasma membrane of primary rat adipocytes in

the presence and absence of insulin. As the reagent was impermeable to the plasma

membrane, surface transporters could be labeled and subsequently immunoprecipitated

thereby allowing the detection of insulin-induced changes at the plasma membrane with

much less trepidation regarding contamination with intracellular membranes (Holman et

al., 1990). The results were no doubt startling at the time. It was reported that after insulin exposure total ATB-BMPA labeling was increased approximately 6 fold over basal whereas glucose transport was increased ~20 fold; however, the labeling of immunoprecipitated Glut1 and Glut4 was increased ~5 and 20 fold respectively (Holman et al., 1990). This led to the assertion that translocation of both Glut4 and Glut1 could

account for the insulin-induced increase in glucose transport completely thereby

indicating that insulin caused no activation of transporter intrinsic function. It is worth

mentioning that these studies were performed in the presence of added adenosine, while

those of Karnieli et al were not, a condition which could have been confounding for

reasons mentioned previously and discussed below.

Almost one year after this dissenting study suggesting that the activation of

transporter intrinsic function did not occur, Cushman and colleagues published another

study involving primary rat adipocytes and the 3H-ATB-BMPA reagent reversing their

position (these experiments were performed without added adenosine) (Clark et al.,

1991). It was reported that ATB-BMPA surface labeling did increase with insulin - 24

stimulation; however, in agreement with previous studies (Karnieli et al., 1981), neither

the time course nor the fold change of surface labeling corresponded with that of glucose transport. They reported a ~1 minute lag between the t1/2 of the increase in glucose transporters in the plasma membrane and the increase glucose transport which approximated the 1.5 minute difference reported in the earlier study. These authors go on to speculate as to the mechanisms of activation and suggest that a proportion of basal

transporters with low intrinsic activity could be activated by covalent or structural

processes or by changes in their environment. This last suggestion indicated the early

interest of these investigators in the concept of transporter regulation by microenvironment (Joost et al., 1987).

In 1992, the interpretation of data generated using the ATB-BMPA reagents was refined by a study of 3T3-L1 adipocytes which suggested the ATB-BMPA reagent preferentially labeled active glucose transporters (Harrison et al., 1992). Like primary rat adipocytes exposed to cycloheximide for 90 minutes (Baly et al., 1987; Matthaei et al.,

1988), 3T3-L1 adipocytes exposed to anisomycin or cycloheximide for 5 hours increased glucose transport substantially (7 and 5 fold respectively). While surface labeling with

3H-ATB-BMPA demonstrated ~3.5-fold and ~1.8-fold increases in Glut1 and Glut4 labeling respectively, immunoblot of these proteins suggested there was no change in the plasma membrane after exposure to these inhibitors of protein synthesis. It seemed from these data that in this context there was not only an activation of glucose transporter intrinsic function but also preferential labeling of active transporters with 3H-ATB-

BMPA. Another glucose transporter affinity-labeling reagent [125I]IAPS-forskolin (3-

[125I]iodo-4-azidophenethylamido-7-O-succinyldeacetylforskolin) shows similar 25

properties as ATB – BMPA (Clancy et al., 1991). The propensity of these reagents to

preferentially label transporters in a transport – stimulated context has proved useful in

other studies of the activation of glucose transporter function (Barnes et al., 2002).

The work reviewed thus far was from studies of primary or transformed adipocytes. There are many other studies which examine the regulation of glucose transporter intrinsic function in cell types such as muscle and erythrocytes (Fisher et al.,

1996). Klip and colleagues applied many techniques developed in the study of adipocytes to muscle cells, in particular the rat L6 myogenic cell line (Bashan et al.,

1993; Bashan et al., 1992; Bilan et al., 1992a; Bilan et al., 1992b; Bilan et al., 1991; Klip

et al., 1990; Mitsumoto et al., 1991; Mitsumoto et al., 1992a; Mitsumoto et al., 1992b;

Ramlal et al., 1988; Somwar et al., 2001; Taha et al., 1997). Their findings confirmed

those from adipocyte studies (Furtado et al., 2002; Somwar et al., 2001); however, in the

L6 myogenic cell line, the interpretation of these results is complicated by the expression

of Glut3 protein and the apparent regulation of its subcellular localization and activity by

insulin and IGF-1 (Bilan et al., 1992a; Ramlal et al., 1988; Wilson et al., 1995).

In any case, Klip and colleagues found that the lag between increased transport

and the appearance of Glut4 at the plasma membrane also occurred in L6 myotubes

(Somwar et al., 2001). Additionally, they showed that inhibition of p38 MAPK α and β

with the small molecule inhibitor SB203580 (4-(fluorophenyl)-2-(4- methylsulfonyl-

phenyl)-5-(4-pyridyl) imidazole) partially blocked insulin-stimulated glucose transport

(IC50 ~1 µM) in both L6 myotubes and 3T3-L1 adipocytes, but not the translocation of either Glut1 or Glut4 to the plasma membrane (Somwar et al., 2001; Somwar et al.,

2002). For the latter finding, they employed microscopy of plasma membrane sheets as 26

well as an enzyme – linked immmunosorbent assay (ELISA) technique to detect surface- labeled Glut4-Myc in whole cells. Since the ELISA analysis was of intact cells, it clearly demonstrated the increase of Glut4 surface expression while controlling for the confounding factor of Glut4 vesicles that could be docked, but not fused, with the plasma membrane (Lange et al., 1990a; Lange et al., 1990b; Vannucci et al., 1992). Thus, the translocation of Glut4 and the activation of glucose transport were dissociated in this model and a role for p38 in the mechanism of transporter activation was suggested

(Somwar et al., 2001; Somwar et al., 2002; Tsakiridis et al., 1996). p38 had been implicated in the regulation of glucose transport activity by a variety of other studies using p38 – stimulating agents such as inhibitors of protein synthesis, hyperosmotic shock, and AICAR (5-amino 4-imidazole carboxamide riboside 5’-triphosphate) – stimulated AMP – activated kinase (AMPK) activity (Abbud et al., 2000; Barnes et al.,

2002; Barros et al., 2001; Barros et al., 1997b; Clancy et al., 1991; Gould et al., 1995;

Harrison et al., 1992; Xi et al., 2001).

In erythrocytes, the regulation of glucose transporter intrinsic activity has been relatively easier to observe given the general lack of intracellular membrane compartments (Diamond et al., 1993; Wood et al., 1969; Zhang et al., 1998). In a very detailed study, Carruthers and colleagues demonstrated what they called transporter

“derepression” in pigeon erythrocytes which have mitochondria and nuclei but no other detectable intracellular membrane structures (Diamond et al., 1993). This report showed an increase in glucose transport after “poisoning” the erythrocytes with a mitochondrial respiration uncoupler and demonstrated no increase in Glut1 protein in the cell based on a competition ELISA. 27

In human erythrocytes, it is believed that there are no intracellular membranes in

the basal state (Turner et al., 2003) and as such stimulation of glucose transport activity

could be due to a change in the intrinsic activity of Glut1 which is highly expressed in

these cells. It has been reported that human erythrocytes express both Glut5 and Glut2 in

addition to Glut1 although this has yet to be widely confirmed or cited (Concha et al.,

1997). Cytochalasin B binding sites have been tested rigorously in these cells and shown

to have 3 binding classes; 1) glucose – sensitive binding sites representing the glucose

transporter; 2) cytochalasin E – sensitive binding sites and; 3) glucose and cytochalasin E

insensitive binding sites (Jung et al., 1977). In the presence of glucose, cytochalasin E,

which neither binds nor inhibits Glut1, not only increases the glucose-inhibitable

cytochalasin B binding sites (class I) but also increases glucose transport reportedly by

conversion of class III binding sites into Class I binding sites (Lachaal et al., 1990; Zhang

et al., 1998). It was suggested that this stimulation of Glut1 detection and function

reflected an “unmasking” of previously inactive (or less active) Glut1 molecules possibly representing changes in protein-protein interactions or the microenvironment of the

transporter.

Another cell type that has demonstrated great utility in studies of the regulation of

glucose transporter intrinsic activity has been Clone 9 cells, a non-transformed rat liver-

derived cell line described earlier. Acute exposure of these cells to sodium cyanide or

sodium azide, both inhibitors of cytochrome a3 in the cytochrome c oxidase complex of the electron transport chain (ETC), increased glucose transport entirely by post- translational mechanisms (Mercado et al., 1989; Shetty et al., 1993a; Shetty et al., 1992;

Shetty et al., 1993b; Shi et al., 1995). It was proposed that the stimulation of glucose 28

transport represented, in part, an increase in the intrinsic activity of the transporter for

several reasons: 1) the fold increase in glucose transport under these conditions was not accompanied by a corresponding fold increase in the abundance of Glut1 in the plasma membrane (Koseoglu et al., 1999), 2) there was no change in either the Km of uptake or decrease in the energy of activation for transport (Hamrahian et al., 1999; Mercado et al.,

1989).

In 1995, Shi et al published that exposing Clone 9 cells to azide for 1 hour

significantly increased cytochalasin B (CB) binding in cellular membranes (Shi et al.,

1995). The Kd of binding after azide-exposure was not different from control (~0.26

µM); however, the total binding capacity of the membranes increased ~4 fold to ~7

pmol/mg protein. Similar to human erythrocytes (Jung et al., 1977; Zhang et al., 1998),

Clone 9 cells have three CB binding sites: 1) glucose - sensitive, 2) cytochalasin E -

sensitive, and 3) sites insensitive to both reagents. Exposing cell membranes to both

glucose and cytochalasin E increased the number of glucose-inhibitable binding sites at

the expense of the third class of binding sites. Interestingly, incubating human

erythrocyte derived Glut1-containing liposomes with cytosol from Clone 9 cells exposed

to azide decreased their total cytochalasin B binding capacity but not if the cytosol was

heated first or if Glut1-C terminal peptide (amino acids 427-492) was included. This

suggested the existence of a protein inhibitory factor whose association with the

membrane fraction, possibly the C-terminus of Glut1, was interrupted by azide exposure.

They identified 35S-labeled 28 and 70 kDa proteins (which could be not labeled with 32P), as binding partners of Glut1 in cytosol from Clone 9 cells using GST-fusions of the Glut1

C-terminus. Although providing some evidence for the notion that protein-protein 29

interactions are important for the regulation of Glut1 intrinsic activity, the identity of

these proteins has not been reported.

In other studies, Koseoglu et al and Barros et al, reported that there was no

increase of Glut1 in plasma membrane sheets of Clone 9 cells treated with 5 mM azide or

exposed to hyperosmolarity despite increased glucose transport thereby adding to the

body of evidence suggesting regulation of glucose transporter intrinsic function (Barros

et al., 2001; Koseoglu et al., 1999). In 3T3 L1 adipocytes hyperosmotic shock increased

plasma membrane Glut4 significantly (1.35-fold); however, when compared to the ~ 5

fold increase in glucose transport, the suggestion of an increase in transporter intrinsic

activity remains (Barros et al., 2001). Additionally, this increase in transport was

insensitive to wortmannin and SB203580 suggesting that in this context neither PI3K nor

p38 respectively were involved in the mechanism of activation. (Barros et al., 1997a;

Barros et al., 2001).

Exposure of either mouse 3T3-L1 adipocytes or rat Clone 9 cells to inhibitors of protein synthesis (anisomycin, cycloheximide, or puromycin) also stimulated glucose transport without a corresponding increase in transporters at the plasma membrane

(Barros et al., 1997b; Clancy et al., 1991; Harrison et al., 1992). Both anisomycin and puromycin activate the Jun-N terminal kinase, JNK, and p38 MAPK. In 3T3-L1 adipocytes, inhibitor studies with the p38 inhibitory SB203580 implicated p38 α and β in

the mechanism of the response of glucose transport to anisomycin (Barros et al., 1997b;

Gould et al., 1995). Additionally, SB203580 does block 5-aminoimidazole-4-

carboxamide ribonucleoside (AICAR) – induced activation of glucose transport in Clone

9 cells suggesting that in this context p38 may also have a role in this mechanism of 30

action (Abbud et al., 2000; Barnes et al., 2002; Xi et al., 2001). SB203580 does not work

to block the azide-induced increase in glucose transport observed in Clone 9 cells

suggesting that the operative mechanism of activation is independent of p38 α and β

(Barros et al., 2001). These findings suggest that the activation of glucose transporter

intrinsic function may involve cell – type and pathway – specific regulatory features.

In Clone 9 cells, broad spectrum interventions have been shown to inhibit the

azide effect, e.g. calcium chelation, while other more specific agents such as low

concentration wortmannin or SB203580 do not (Barros et al., 2001; Barros et al., 1995;

Davies et al., 2000). The effects of Ca2+ chelation could support a role for the complex

interaction between the lipid microenvironment and glucose transporters (Carruthers et

al., 1988) as evidenced by the inhibition of basal and stimulated glucose transport by

calcium chelation (Quintanilla et al., 2000) and the effects of calcium on membrane fluidity (Livingstone et al., 1980; Schootemeijer et al., 1994). This concept was also supported by studies with 3T3-L1 fibroblasts which showed that increasing the sphingomyelin: cholesterol ratio modestly increased basal glucose transport while inhibiting insulin – stimulated glucose transport without affecting Glut1 transporter detection in a so – called crude plasma membrane fraction (Al-Makdissy et al., 2003). 31

Membrane Microdomains and Glucose Transport Regulation

The interest in the effect of membrane microenvironment on glucose transporter function

was evident even before the recruitment hypothesis was proposed to explain the effects of

insulin on glucose transport. Studies of human erythrocyte glucose transporters

reconstituted into model membranes indicated the important and complex interactions

between lipid composition (especially cholesterol), membrane fluidity, and glucose

transporter activity (Carruthers et al., 1984; Carruthers et al., 1988; Connolly et al., 1985;

Stubbs et al., 1984; Yuli et al., 1981).

In studies of glucose transporter translocation (Oka et al., 1984), Czech and

colleagues write the following:

“A basic unproved tenet of the recruitment hypothesis has been that the low density membrane fraction containing high levels of glucose transporter in controls cells indeed resides in an intracellular compartment rather than on the cell surface as a domain of the plasma membrane… it is possible that low density membranes containing glucose transporters are part of the cell surface and are sheared away from microdomains [emphasis added] or invaginations of the plasma membrane during homogenization. Insulin might act to activate these transporters by a mechanism that involves structural changes in these membrane domains such that they are not sheared by homogenization… transporter activation by insulin might involve lateral movement of transporters away from these membrane microdomains...”

Although some the points in this statement have proven untenable, the idea that

membrane microdomains influence glucose transporter recruitment and regulation has

been discussed often in recent years although not without controversy (Chamberlain et

al., 2002; Chiang et al., 2001; Gustavsson et al., 1996; Kumar et al., 2004; Parpal et al.,

2001; Rubin et al., 2003).

The concepts surrounding membrane microdomains have been studied for many years by examining lipid interactions in model membrane systems (London, 2002;

London et al., 2000). These studies indicated that mixed phase lipid systems in cells 32

could be generated from the association of sphingolipids and facilitated by cholesterol.

In vitro arrangements of certain ratios of these lipids in bilaers could assume a liquid ordered (Lo) phase which was intermediate to, but could co-exist with, liquid – disordered phases of the bilayer. Cholesterol is considered to be one of the primary factors that allowed the formation of Lo phases with the sphingolipids found in cell membranes.

Over the past 15 years or so, detergent resistant lipid domains in cells (Yu et al.,

1973) have received significant attention from the scientific community as the manifestation of the lipid interactions long observed in model membrane systems (Brown et al., 2000; Brown et al., 1992; Hooper, 1999; Simons et al., 1997; Skibbens et al.,

1989). These detergent resistant membranes (DRMs) have been described as “lateral organizations of membrane lipids” biochemically recognized by their insolubility in low concentrations of non-ionic detergents at low temperatures (4oC) and the low-density imparted by their saturated lipid and cholesterol content. Simons et al termed these

DRMs “lipid rafts” (Simons et al., 1997) and proposed the following commentary regarding their nature: 1) native rafts exist only in living cells, 2) DRMs are non-native, i.e. they are aggregated rafts, and 3) raft composition will vary with extraction conditions.

There has been much controversy regarding the observation and isolation of lipid rafts and it is clear from this controversy that raft quality differs depending on cell type and the detergent quality and quantity used to extract them (Garavito et al., 2001; Munro, 2003;

Röper et al., 2000; Schuck et al., 2003).

In 1988, the concept of functionally clustered sphingolipids was proposed by

Simons and van Meer as a sorting mechanism for lipids and proteins in the MDCK polarized epithelial cell model (Simons et al., 1988). The association of proteins with 33

these functional clusters of lipids, or DRMs, appeared to occur during their transit through the trans-Golgi network which was believed to be the location of the apical and basolateral sorting processes in this cell model (Skibbens et al., 1989). In 1992, Brown and colleagues published their seminal report on the isolation and characterization of low density, DRMs from Madine-Darby canine kidney cells (MDCK) (Brown et al., 1992).

This study demonstrated that these DRMs were enriched in sphingolipids and cholesterol as would have been expected from the in vitro studies discussed earlier (London, 2002;

London et al., 2000). These membrane domains were soon reported to exist in both polarized and non-polarized cells and proposed to serve as both sorting domains in the

Golgi and as focal points for protein aggregation at the plasma membrane, a process integral to signal transduction from some receptors (Arcaro et al., 2001; Arcaro et al.,

2000; Chiang et al., 2001; Scherer et al., 1994; Simons et al., 1997; Wilson et al., 2000).

Many determinants have been proposed for the association of proteins with

DRMs. The transmembrane domains of some proteins have an intrinsic affinity for this lipid environment, e.g. the hemeagglutinin transmembrane domain (Scheiffele et al.,

1997). Other membrane proteins, e.g. caveolin 1, bind lipids like cholesterol and this may drive their interaction with the DRMs (in the case of caveolin, the DRMs are a specialized population called caveolae) (Fielding et al., 2000; Murata et al., 1995;

Schlegel et al., 2001; Simons et al., 1997; Smart et al., 1994). Glycophosphatidylinositol- anchored proteins were the first proteins to be studied intensively with regard to their

DRM association (Brown et al., 1992; Harder et al., 1998; Kenworthy et al., 2000; Varma et al., 1998; Zacharias et al., 2002). In response to certain stimuli, some proteins, e.g. certain Src family tyrosine kinases, are covalently modified with lipids which facilitates 34

their translocation from the cytosol to membrane microdomains (Dunphy et al., 2001;

Melkonian et al., 1999; Webb et al., 2000)). Finally, protein-protein interactions are an important way to facilitate signaling from or through DRMs as evidenced by the inducible-association of insulin receptor signaling intermediates with the DRMs (Chiang

et al., 2001).

The removal or sequestration of cholesterol with various agents, such as the β-

cyclodextrins has been useful for examining the role of the cholesterol-rich DRM

domains in many processes including vesicle trafficking, signal transduction, and solute transport (Chamberlain et al., 2002; Hoessli et al., 2000; Martens et al., 2001). The effect

of cholesterol manipulation on cells in vitro has been characterized carefully and thoroughly (Irie et al., 1982; Kilsdonk et al., 1995; Yancey et al., 1996). The β-

cyclodextrins seemed to be relatively specific for cholesterol since, in certain cells, it was

reported that relatively short exposures to the reagent extracted little of the highly

abundant membrane phospholipids (Kilsdonk et al., 1995). Whole cell toxicity was also

noted to be low under these conditions; however, in other conditions, toxicity correlated

with the cholesterol extraction efficacy of the β-cyclodextrin employed. There is some

evidence that the morphology and function of certain organelles can be profoundly

affected by these agents (Hansen et al., 2000).

Studies using cholesterol manipulating reagents have implicated membrane

cholesterol as an integral part of cellular glucose metabolism. Although there is some

controversy, it seems that the insulin receptor and elements of its effector pathways

localize to the DRMs and require their integrity for proper activity (Chiang et al., 2001;

Gustavsson et al., 1999; Parpal et al., 2001; Vainio et al., 2002). A few studies indicate 35

that Glut4 itself associates with the DRMs (Gustavsson et al., 1996; Karlsson et al., 2002;

Scherer et al., 1994). An association between Glut1 and DRMs has been found in several

different cell types including HeLa cell – fibroblast hybrids, Clone 9 cells, human

erythrocytes, 3T3-L1 fibroblasts, and 3T3-L1 adipocytes (Kumar et al., 2004; Rubin et

al., 2003; Rubin et al., 2004; Sakyo et al., 2002). These consistent reports suggest the

DRMs have some role in the regulation of glucose transport. In 3T3 – L1 fibroblasts, an

increased sphingomyelin: cholesterol ratio was found to increase basal glucose transport but to inhibit insulin – stimulated glucose transport. While these studies support a role for membrane microdomains in the regulation of glucose transport, one published study suggested that the role of cholesterol in the regulation of Glut1 function is minimal

(Samuel et al., 2001).

The putative inhibitory association between Glut1 and stomatin, an integral

membrane protein known to reside in DRMs (Salzer et al., 2001; Snyers et al., 1999a),

suggested that DRMs may influence glucose transport regulation in Clone 9 cells (Rubin

et al., 2003; Zhang et al., 2001; Zhang et al., 1999b). Although the function of stomatin

(also known as human erythrocyte band 7.2b) is unknown, it may be involved in the

regulation of membrane structure, sodium channel activity, and membrane signaling in

both mammalian and nematode systems (Desneves et al., 1996; Hiebl-Dirschmied et al.,

1991; Rajaram et al., 1999; Salzer et al., 1993; Sedensky et al., 2001; Snyers et al., 1997;

Snyers et al., 1998; Snyers et al., 1999a; Snyers et al., 1999b; Stewart, 1997). Given the

association between Glut1 and stomatin as well as the potentially inhibitory lipid

composition of the DRMs (Carruthers et al., 1984; Gidwani et al., 2001), it has been

hypothesized that DRMs are important for the regulation of glucose transport (Rubin et 36

al., 2003). This hypothesis has been investigated by several other groups (Barnes et al.,

2003; Kumar et al., 2004). The results of our studies, detailed herein and presented elsewhere (Rubin et al., 2003; Rubin et al., 2004), report the association of Glut1 with the

DRMs of Clone 9 cells and the effects of the inhibition of oxidative phosphorylation on this association.

Table 1.1. The glucose transporter family and their tissue distribution1 (Joost et al.,

2001; Mueckler, 1994; Wood et al., 2003b; Zhang et al., 1999a)

Class Isoform1 Expression

I Glut1 Erythrocytes, brain vascular endothelium, astrocytes, intestinal mucosa, kidney

I Glut2 Liver, pancreatic islets, kidney, intestinal mucosa

I Glut3 Neurons, muscle, liver, kidney, placental endothelium (Stuart et al., 2001)

I Glut4 Adipose, muscle, heart, brain

II Glut5 Small intestine, kidney, heart, muscle, adipose, erythrocytes (?),testes, spermatozoa (Concha et al., 1997; Shepherd et al., 1992)

III Glut6 Brain, spleen, leukocytes, kidney (Lisinski et al., 2001)

II Glut7 Small intestine, colon, testis, prostate

III Glut8 Adipose, blastocysts, brain, heart, skeletal muscle, testes (Carayannopoulos et al., 2000; Doege et al., 2000b; Scheepers et al., 2001)

II Glut9 Brain, heart, pancreas, peripheral blood leukocytes (Doege et al., 2000a)

III Glut10 Heart, adipose, adipose vascular cells, liver, lung, skeletal muscle, brain, pancreas, placenta (Wood et al., 2003a)

II Glut11 Heart, skeletal muscle, kidney, small intestine, peripheral blood, and other tissues (McVie-Wylie et al., 2001; Wu et al., 2002b)

III Glut12 Heart, skeletal muscle, adipose, adipose vascular cells, small intestine, brain, chondrocytes, mammary gland (Richardson et al., 2003; Rogers et al., 2003; Rogers et al., 2002; Wood et al., 2003a)

III HMIT2 Heart, brain, adipose, kidney (Uldry et al., 2001)

I Glut143 Testis (Wu et al., 2002a)

1These is some variation in reports of message expression depending on the method used for message detection, i.e. Northern blot or PCR. 2 HMIT – proton/myoinositol co-transporter (HMIT). 3 Glut14 is believed to be a duplication of Glut3.

Figure 1.1. Schematics of SLCA2 family Class I and III glucose transporters. These

images represent the 12 – transmembrane domain model of the facilitated glucose transporter (Mueckler et al., 1985) and highlight residues conserved in all known family members in closed circles and those conserved in a particular class indicated in open

boxes (adapted from Joost, et al., 2001). Not shown here is one major difference between

Class I and Class II which is the absence of the tryptophan (W) on the cytoplasmic side

of helix 10 which is known to be involved in the binding of cytochalasin B to Glut1

(Joost et al., 2001). 39

Figure 1.2. A dendrogram depicting the relationships between all classes of glucose transporters in the SLCA2 family. The length of each branch depicts the extent of differences between members (adapted from Joost, et al., 2001). Glut14 is not shown; however, is believed to be a duplication of Glut3 (Wu et al., 2002a).

Figure 1.3. Flow charts of protocols for the isolation of membrane compartments.

a) This protocol describes the characterization of adenyl cyclase activity found in the membranes of pigeon erythrocytes (adapted from Davoren and Sutherland). This

protocol was used to demonstrate that adenyl cyclase activity associated with a cell

membrane compartment other than the nucleus (Davoren et al., 1963). 43

Pestle High pressure Homogenization Homogenization in 155 mM in 155 mM NaCl

600 x g Layered onto 600 x g 15 20% glycerol 15 minutes minutes cushion pellet

Majority of Majority of adenyl cyclase adenyl cyclase activity but not activity separated from DNA, mitochondria h 600 x g 600 x g 15 minutes 15 minutes supernate pellet

78 k x g pellet

Majority of Minority of adenyl cyclase adenyl cyclase activity activity Pestle homogenization into 0.9M or 1.5M sucrose and isopynic gradient formation

90 k x g 4 hrs

Assay of 105 k x g pellet of fractions demonstrate a broad peak of adenyl cyclase activitycentered at density 1.18 44

Figure 1.3. Flow charts of protocols for the isolation of membrane compartments.

b) This protocol describes the characterization of a plasma membrane fraction derived from HeLa cell homogenates (adapted from Boone et al 1969). The results of this protocol were cited by other protocols that were highly influential in the study of the glucose transporter biology (Boone et al., 1969; McKeel et al., 1970; Simpson et al.,

1983).

3.5 billion cells Hypotonic 3 – 5 strokes washed with Lysis (10 mM Tris, in 50 ml Dounce 3 krpm 15 minutes 10 mM Tris, 1 mM Mg2+); homogenizer; 1 mM Mg2+ TM buffer adjusted to 250 147 mM NaCl mM sucrose

Sonicated 5 seconds

pellet supernate 3 krpm 15 min

Resuspended pellet supernate to 45% sucrose in TM pH 8.6 then overlaid with 40%, 35%, and 8.5% Loaded onto sucrose. 20 – 50% gradient 20 krpm 10 min

20 krpm 16 hrs 35 – 40 % 40 – 45% pellet interface interface Fractions diluted to 10% sucrose in TM pH 8.6 And pelleted at Diluted to 10% Unbroken nuclei 40 krpm 45 min sucrose then cells and 3.5 krpm for “debris” 20 minutes

ghosts

Zone 1 Zone 2 Zone 3

Figure 1.3 b continued

Zone 1 Zone 2 Zone 3

• 30 – 33% sucrose • Distinguishing OD peak • Average density 1.137 • Empty smooth vesicles of similar sizes •Occasionally with ribosomes

•33 – 38% sucrose •smooth vesicles occasionally with ribosomes •Larger smooth vesicles •Rough vesicles

•38 – 43% sucrose •Mitochondria •Endoplasmic reticulum 47

Figure 1.3. Flow charts of protocols for the isolation of membrane compartments. c) This protocol describes the characterization of a plasma membrane fraction derived from the homogenate of primary adipose cell homogenates (adapted from McKeel et al

1970). This protocol represents a standard for membrane fractionation in the field of the glucose transporter biology (McKeel et al., 1970).

Cells isolated from Fine mesh Washed in Shift to 4oC; collagenase- digested filtered and cells Homog buffer: 10 strokes at 18 krpm epididymal fat pads collected as 1 k x g 10 mM Tris-HCl 35 ml Teflon pestle of 12 – 16 rats in modified 30 sec pellet 1 mM EDTA homogenizer Kreb’s buffer with 10mg/ml 250 mM sucrose albumin pH 7.4 20oC

16 k x g 15 minutes 1 k x g 10 minutes

Discard pellet; supernate Pellet Supernate

13 k x g 5 minutes Washed in homog buffer; Centrifuged 1 k x g 10 min Centrifuged at Discard supernate; 160 k x g for 45 min resuspend pellet in homog buffer and Pellet Supernate repeat to obtain P6; resuspend P6. Discard supernate; resuspend pellet in Centrifuged 16 k x g Resuspended in homog buffer and 20 min; the supernate 600 x g 5 minutes homog buffer w/o centrifuge 160 k x g discarded and the pellet EDTA (buffer II); for 45 mintues called nuclei resuspended and layered onto linear sucrose or Discard pellet; Ficoll gradients centrifuge supernate Discard supernate; 13 k x g 5 min resuspend pellet in buffer II; called Centrifuged at Microsomes 24 k x g for 90 min Resuspend pellet in buffer II; call Crude mitochondria Supernate

Figure 1.3 c continued

Supernate

Centrifuged 16 k x g 20 min; the supernate discarded and the pellet resuspended and layered onto linear sucrose or Ficoll gradients

Centrifuged at 24 k x g for 90 min

Isolated particulate bands at densities 1.14 and 1.18

Band 1.14 Band 1.18

Resuspended in Resuspended in homog buffer homog buffer

Centrifuged at Centrifuged at 16 k x g for 15 min 16 k x g for 10 min

Resuspended in Resuspended in homog buffer; called homog buffer; called Gradient Mitochondria Plasma membrane 50

Figure 1.3. Flow charts of protocols for the isolation of membrane compartments. d) This protocol is a variation of that described by McKeel et al, 1970 and was used to characterize the disposition of glucose transporters in the plasma membrane primary adipose cell homogenates in response to insulin (Simpson et al., 1983). Interestingly, they do not use a preparative linear sucrose gradient for the isolation of the plasma membrane fraction instead opting for a step gradient; the rationale for this choice was not indicated in the text of their report.

80 million cells 7 nM Insulin at Washed x 2 in 10 strokes washed x 5 with 37oC w shaking Homog buffer: 55 ml Teflon pestle 10 mM Krebs- 20 mM Tris-HCl homogenizer Ringer bicarbonate 1 mM EDTA 30 mM Hepes 7.4 255 mM sucrose 10 mg/ml BSA pH 7.4 c Shift to 4 C 20oC

16 k x g 15 minutes

pellet supernate

Washed x 2 in Centrifuged at Homog buffer 48 k x g for 20 min

pellet supernate Resuspended in Homog buffer layered onto 1.12 M sucrose cushion; centrifuged 101 k x g for 70 minutes Washed once and Centrifuged resuspended in at 212 k x g 70 min, homog buffer; resuspended and called high density called low density pellet interface microsomes microsomes

Washed once and Resuspended to resuspended in 50 ml, centrifuged homog buffer; at 48 k x g 45 min, called mitochondria, resuspended and nuclei, and debris called plasma membrane

Chapter 2

Distribution of Glut1 in Detergent-Resistant Membranes (DRMs) and non-DRM domains: Effect of Treatment with Azide

(Parts of this chapter have been published. Rubin D and Ismail-Beigi F. Am J Physiol

Cell Physiol 285: C377-383, 2003)

Introduction

Glut1, a widely expressed member of the facilitative glucose transporter family, is an

intrinsic membrane glycoprotein with 12 putative transmembrane domains (Joost et al.,

2001). Its expression and function are regulated by a variety of stimuli and agents

including serum, growth factors, transformation, hypoxia, and inhibitors of oxidative

phosphorylation, with the latter two being associated with stimulation of 5’-AMP-

activated protein kinase (AMPK) (Abbud et al., 2000; Barnes et al., 2002; Behrooz et al.,

1998; Hamrahian et al., 1999; Hayashi et al., 2000; Ismail-Beigi, 1993; Mercado et al.,

1989; Shetty et al., 1993a; Shetty et al., 1992; Shetty et al., 1993b; Xi et al., 2001). The

enhancement of Clone 9 cell glucose transport in response to inhibition of oxidative

phosphorylation by azide is biphasic, with the acute response (1-2 h) being mediated

entirely by post-translational mechanisms (Hamrahian et al., 1999; Shetty et al., 1992;

Shetty et al., 1993b). This response does not involve a substantial increase in the

abundance of Glut1 in the plasma membrane, and the increase in the Vmax of transport

52 53

occurs without a change in the affinity of the transporter for glucose or a decrease in the

energy of activation of transport (Hamrahian et al., 1999; Mercado et al., 1989). These

characteristics suggest that the acute stimulation of glucose transport involves activation

(“unmasking”) of Glut1 transporters pre-existing in the plasma membrane.

Our laboratory has suggested that the integral membrane protein stomatin

(Stewart, 1997) may be an inhibitory binding partner of Glut1 in Clone 9 cells (Zhang et

al., 2001; Zhang et al., 1999b). Although its functions are not clear, stomatin appears to

play a role in the regulation of cell membrane structure and function (Desneves et al.,

1996; Hiebl-Dirschmied et al., 1991; Rajaram et al., 1999; Salzer et al., 1993; Salzer et

al., 2001; Sedensky et al., 2001; Snyers et al., 1997; Snyers et al., 1998; Snyers et al.,

1999a; Snyers et al., 1999b). The presence of stomatin in detergent – resistant membrane

domains (Salzer et al., 2001; Samuel et al., 2001) and its interaction with Glut1 suggested

that DRMs may play an important role in control of Glut1 function.

DRMs represent a revision of the fluid-mosaic model of cell membranes. It has

become clear that the heterogeneous mixture of lipids in cellular membranes can form

domains of closely packed lipids that contain specific proteins and high levels of

sphingolipids and cholesterol (Brown et al., 2000; London et al., 2000; Simons et al.,

1997). DRMs are biochemically defined by their lipid content, low density, and their

resistance to solubilization in buffers containing non-ionic detergents at low temperatures

(4 oC); hence their designation as low-density, detergent resistant, membrane (DRM) domains.

The reported localization of stomatin in DRMs of human RBCs (Salzer et al.,

2001; Samuel et al., 2001) and the suggested correlation between stomatin expression and 54 decreased glucose transport (Zhang et al., 2001; Zhang et al., 1999b) prompted us to pose the following set of questions: 1) Is Glut1 also, in part, concentrated in DRMs in Clone 9 cells?, 2) Is stomatin localized to the DRM fraction in cells other than the human RBC?, and 3) Does exposure of cells to inhibitors of oxidative phosphorylation (such as azide) result in a redistribution of stomatin and/or Glut1 between the DRM and non-DRM domains? These and other questions are relevant, especially in light of a recent report indicating that Glut1, but not Glut3, is mostly localized in the DRM fraction in HeLa cells (Sakyo et al., 2002). Results of our studies demonstrate that a fraction of cell Glut1 and the bulk of stomatin reside in DRMs under basal conditions, and that stimulation of glucose transport in response to azide is associated with a partial redistribution of Glut1, but not of stomatin or caveolin-1, out of the DRM domains. 55

Materials and Methods

Materials Cell culture reagents were purchased from Invitrogen (Grand Island, NY), plasticware through Fisher Scientific, and all other chemicals from Sigma Aldrich or

Fisher Scientific, unless otherwise noted. Fugene 6 was purchased from Roche

(Indianapolis, IN). Sulfo-NHS-SS-biotin reagent and the Dc Protein Assay kit were

purchased from BioRad (Hercules, CA). 4 ml ultracentrifugation tubes were purchased

from Kendro. Antibodies were purchased from the following companies: mouse anti-

caveolin-1 (Transduction Laboratories), rabbit anti-Glut1 (Chemicon, Temecula, CA),

mouse anti-human transferrin receptor-1 (hTfR1) (Chemicon, Temecula, CA),

horseradish peroxidase-conjugated anti-rabbit and anti-mouse IgG secondary antibodies

(Sigma Aldrich). Mouse anti-human stomatin monoclonal antibody was a gift from Dr.

Rainer Prohaska (University of Vienna). Western blots were developed using the

enhanced chemiluminescence kit from Santa Cruz Biotechnology (Santa Cruz, CA).

Nitrocelluose membrane was purchased from Schliecher & Schuell (Keene, NH).

Cell culture Clone 9 cells were maintained in DMEM supplemented with 10% bovine

calf serum (v/v) and penicillin-streptomycin. Cells were transfected with pcDNA3-

human stomatin plasmid (Zhang et al., 2001) or the empty vector according to the

manufacturer’s protocol to enable the detection of stomatin by the available antibody.

Stable transfectants were selected with 0.5 mg/ml Geneticin (G418). Surviving clones

were harvested using cloning cylinders and maintained in DMEM containing 0.2 mg/ml

G418. A clone expressing the human stomatin was employed in these studies. The rate

of glucose transport in the chosen clone under basal conditions and following exposure to 56

5mM azide was equivalent to non-transfected Clone 9 cells. Culture medium was

replaced with fresh medium 16-24 hours prior to initiation of experiments. 3T3-L1

fibroblasts were maintained as described (Prasad et al., 1999).

Fractionation by sucrose density step-gradient centrifugation Four 10 cm plates of

confluent Clone 9 cells or 3T3-L1 fibroblasts were washed 3 times with ice-cold PBS and

then scraped into 0.5 ml of ice-cold PBS. All subsequent steps were performed at 4 oC.

Cell pellets were lysed in 410 µl Buffer A (0.5% Triton X-100, 50 mM Hepes, 150 mM

NaCl, 5 mM EDTA, 5 mM EGTA, 20 mM sodium pyrophosphate, 1 mM sodium

orthovanadate, 20 mM sodium fluoride, and 0.1 µM PMSF; pH 7.4). 385 µl of the post- nuclear lysate was loaded into the bottom of a 4-ml ultracentrifuge tube for the Sorvall

TST 60.4 swinging bucket rotor, adjusted to 40% sucrose with 60% sucrose in Buffer B

(20 mM Tris, 150 mM NaCl, 1 mM EDTA, 0.1 µM PMSF; pH 7.4) to a total volume of 1

ml and then overlaid with 2 ml 35% sucrose and 1 ml 5% sucrose in Buffer B. Samples

were centrifuged at an average of 150 k x g for 17 hours at 4 oC. 500 µl fractions were collected from the top of the tube. 30 µl aliquots of each sample were subjected to SDS-

PAGE.

In separate gels, samples of whole cell post-nuclear lysates from control and azide-treated cells were also used for western blotting. The gels were transferred to nitrocellulose membranes, blocked with 5% (w/v) powdered milk, and then incubated overnight with the appropriate primary antibodies. After three washes with TBST

(0.05% Tween-20 in Tris-buffered saline), membranes were incubated with the 57

appropriate horseradish peroxidase-conjugated secondary antibody, washed with TBST,

and developed by enhanced chemiluminescence.

In each experiment, samples from fractions 2 - 8 from control or azide-treated

cells were prepared for western blotting on the same membrane. Moreover, in each

experiment and for each antigen, the densities of the appropriate bands in fractions 2 and

3 were added and divided by the total of densities in all the fractions. The Bio-Rad Gel

Doc 1000 system was used to determine band density.

One step separation of soluble and insoluble material Post-nuclear cell lysates were

prepared as described above using either Triton X-100, Igepal (NP-40), or CHAPS (all at

0.5% in buffer A). Lysates were centrifuged at an average of 128 k x g for 1 hour at 4 oC in 500 µl tubes in a Beckman tabletop ultracentrifuge. Supernates were removed and pellets were resuspended in an equal volume of fresh buffer. Equal volumes of the supernate and resuspended pellet were analyzed.

Detergent-free isolation of DRMs This method was adapted from a previously described protocol (Song et al., 1996). Cells were lysed with 0.5 M sodium carbonate (pH 11) in

Buffer A without detergent. Lysates were homogenized with 20 strokes of a Teflon

Potter-Elvehjem homogenizer. Finally, aliquots of the homogenate were sonicated with a probe-type sonciator (550 Dismembranator, Fisher Scientific) using four 2-second

sessions while on ice. All subsequent steps were performed at 4 oC. This material was processed as described for isolation and analysis of the DRM.

Isolation of plasma membrane DRMs and their fractionation The method of cell surface

biotinylation was adapted from a previously described protocol (Shetty et al., 1993b).

Confluent cells in 10 cm plates were rinsed 3 times with cold PBS then incubated with 3

ml cold biotinylation buffer (120 mM NaCl, 30 mM NaHCO3, 5 mM KCl, 0.1 mg/ml sulfo-NHS-SS-Biotin; pH 8.5) for 30 minutes while rocking at 4 oC. After 3 washes with

5 ml stop buffer (140 mM NaCl, 20 mM Tris, 5 mM KCl; pH 7.5), cells were scraped into 0.5 ml stop buffer. All subsequent steps were performed at 4 oC. Cell pellets were lysed in 400 µl of Buffer A. The post-nuclear lysate was fractionated in the sucrose- density step gradient as described above. 300 µl of fractions 2, 3, 6, 7, and 8 were incubated with streptavidin-agarose beads overnight at 4 oC on a rotisserie shaker; fractions 1, 4, and 5 were excluded since only trace amounts of Glut1 immunoreactivity were present in these samples. The supernatant of each sample was saved and the beads were washed three times with Tris-buffered saline containing 1 mM EDTA (TBS-EDTA;

pH 7.4), and then eluted with 2x Laemmli buffer. 40% of the eluate, called the pulldown

(pd), and 10% of the supernate (sup) was fractionated by SDS-PAGE and analyzed by

immunoblot and enhanced chemiluminescence. Films were scanned and the density of

the bands, measured using the Bio-Rad Gel Doc 1000 system, were used to calculate the

content of Glut1 in each sample. The fraction of cell Glut1 present in plasma membrane

DRMs was calculated by dividing the sum of Glut1 content in the pulldown of fractions 2

plus 3 by the total amount of Glut1 calculated as the sum of Glut1 content in the

pulldown and supernate of all fractions. The ratio of Glut1 in plasma membrane DRMs

to total cell Glut1 in control and azide-treated cells was corrected by subtracting the ratio

of the same from the mock-biotinylated cells prepared in parallel with each experiment. 59

Preparation of red blood cell (RBC) ghosts RBC ghosts were prepared from freshly

drawn blood as previously described (Zhang et al., 1999b). The ghosts were resuspended

in Buffer A containing 1% Triton X-100 and then subjected to the sucrose density step

gradient and the subsequent procedures described above.

Glucose Transport Cytochalasin B-inhibitable glucose transport was measured using 3H-

3-O-methylglucose, as previously described (Hamrahian et al., 1999; Mercado et al.,

1989).

Statistical Analysis All results are expressed as mean ± SEM. Two-tailed t-test was used

and a P value of < 0.05 was considered significant.

Results

Glut1 association with the DRM in the post-nuclear lysate of Clone 9 cells

Confluent Clone 9 cells stably transfected to express the human form of stomatin were

lysed at 4 oC in buffer containing 0.5% Triton X-100 and the post-nuclear lysate (PNL) was fractionated by ultracentrifugation from the bottom of a sucrose step gradient

(Figure 1a). In keeping with previous reports that the transferrin receptor is present in

the non-DRM fraction and is solubilized under the above conditions (Harder et al., 1998),

the receptor was found in the high-density fractions (numbers 6, 7, and 8). On the other

hand, caveolin-1, which is known to be present in caveolae and is not solubilized under the above conditions, was concentrated in the low-density DRM fractions (numbers 2 and

3) (Scherer et al., 1994). Similar to reports in human RBC ghosts (Salzer et al., 2001;

Samuel et al., 2001), we found that the bulk of stomatin is present in the DRM fractions; in repeated experiments ~70% of stomatin was concentrated in the DRMs. In contrast,

Glut1 in the PNL was distributed in both the DRM and non-DRM fractions, with the fraction of Glut1 localized to the DRMs averaging 38 ± 6% (mean ± SE, n = 8). A similar distribution of Glut1 was observed employing non-transfected Clone 9 cells, or buffers containing 1% Triton X-100 (data not shown).

We next determined whether the above-noted presence of Glut1 in the DRM fraction is unique to Clone 9 cells. Employing 3T3-L1 fibroblasts and human RBCs, which are known to express Glut1 near-exclusively (Tordjman et al., 1989; Zhang et al.,

1998; Ziehm et al., 1993), we found Glut1 to be present in DRM fractions of both cells

(Fig. 1b for fibroblasts and Fig. 1c for RBCs). The fraction of cell Glut1 localized to the 61

DRM fraction of 3T3-L1 fibroblast post-nuclear lysate was 29 ± 1%, and for human

RBCs was 20 ± 3%. In 3T3-L1 fibroblasts, the bulk of caveolin-1 (>95%) was localized

in fractions #2 and #3. Similar to the findings in Clone 9 cells, a fraction of Glut1 co-

localized with stomatin in DRMs of RBC ghosts; this is presumably the case in 3T3-L1

fibroblasts although an antibody that recognizes the murine form of stomatin was not

available.

The presence of Glut1 in detergent-resistant membrane domains of Clone 9 cells

was additionally determined after preparation of cell lysates with buffers containing 0.5%

CHAPS or 0.5% NP-40 (Figure 2). As expected, caveolin-1 remained in the detergent-

insoluble fraction, while the transferrin receptor was solubilized. Glut1 was present in

both fractions employing either of the two additional detergents. However, the fraction

of Glut1 remaining in the detergent-resistant fraction (pellet) varied with the detergent

used; in two independent experiments employing the above protocol, the percentage of

Glut1 in the insoluble fraction averaged 31, 70 and 39% using Triton X-100, CHAPS,

and NP-40, respectively. Despite the variability, it is clear that a significant fraction of

cell Glut1 in the post-nuclear lysate is present in the DRM fraction.

In our attempt to verify the presence of Glut1 in DRMs, we employed a

fractionation procedure that does not employ detergents (Song et al., 1996). In this

method, cells are homogenized in 0.5 M Na2CO3 pH 11 (which will also remove peripheral membrane proteins) followed by fragmentation of membranes by sonication prior to the sucrose-density gradient centrifugation. Employing this procedure and by varying the extent of sonication in several different experiments, we could not obtain ideal conditions that would lead to the expected partitioning of caveolin-1 and the 62

transferrin receptor. We observed that while the amount of Glut1 in the light fractions

decreased with increased sonication, caveolin-1 became increasingly associated with

high-density fractions (data not shown). We conclude that the carbonate/sonication

procedure cannot be applied in a quantitatively valid manner in these cells. Nevertheless,

it appears that a percentage of cellular Glut1 may be present in the low-density fraction.

The Effect of Azide on Glut1 association with the DRM in the post-nuclear lysate of

Clone 9 cells

The observation that a fraction of Glut1 and the bulk of stomatin reside in DRMs, and

considering that stomatin is a Glut1-associated protein that appears to exert an inhibitory

effect on Glut1 function (Zhang et al., 2001), led us to test the hypothesis that treatment

of cells with azide and the resulting stimulation of glucose transport is associated with a

redistribution of Glut1 and/or stomatin from the DRM fraction. In keeping with previous

results (Hamrahian et al., 1999; Shetty et al., 1992; Shetty et al., 1993b), exposure of

Clone 9 cells to 5 mM azide for 90 minutes resulted in 4.5 ± 0.8-fold stimulation of

glucose transport (P < 0.05). Upon exposure to azide for 90 minutes, we found that the abundance of Glut1 in the low-density fraction decreased, while the distribution of stomatin and caveolin-1 remained unchanged (Figure 3a). In repeated experiments, the content of Glut1 in DRMs of Clone 9 cells decreased by ~40% (from 38% to 22% in control and azide-treated cells, respectively), while the fractions of stomatin and caveolin-1 in cell fractions 2 and 3 (DRMs) remained unchanged (Figure 3b).

Importantly, the sum-total of Glut1 content in all the gradient fractions per unit protein was equal in control and azide-treated cells (P = not significant). Moreover, and in 63

keeping with previous results (Shetty et al., 1992; Shetty et al., 1993b), the content of

Glut1 in whole-cell lysates was equal in control cells and in Clone 9 cells exposed to 5

mM azide for 90 minutes. The distribution of transferrin receptor was unaltered by

treatment of cells with azide (data not shown).

To examine the time course of action of azide on Glut1 redistribution and glucose

transport, we performed experiments following 30 minutes of exposure to azide. In three

experiments performed after 30 minutes of exposure to 5 mM azide, the content of Glut1

in the DRM fraction decreased by 25 ± 2% (mean ± SE, P < 0.05), while the distribution of stomatin remained unaltered; the rate of glucose transport increased 2.5 ± 0.5-fold in

cells treated with azide for 30 minutes (P < 0.05) (Table 1).

The effect of azide on the abundance of Glut1 in the DRM fraction was also determined in 3T3-L1 fibroblasts. Similar to the finding in Clone 9 cells, the content of

Glut1 in the low-density fractions decreased by ~55% from 29 ± 1% to 13 ± 5% in control cells and in cells treated with 5 mM azide for 2 hours, respectively (n = 4; P

<0.05). In parallel experiments, the rate of glucose transport in fibroblasts increased 2.8

± 0.5-fold in cells exposed to 5 mM azide for two hours (p < 0.05).

The above results in Clone 9 cells and 3T3-L1 fibroblasts led us to conclude that treatment of cells with azide, and the resulting stimulation of Glut1-mediated glucose transport, is associated with a net decrease in the content of Glut1 in DRMs. It is hence possible that the movement of Glut1 away from DRMs with their relatively low membrane fluidity (Gidwani et al., 2001) and the predicted reduction in the association of

Glut1 with stomatin might mediate, in part, the activation (“unmasking”) of Glut1. This

possibility, along with the fact that DRM microdomains are present in both the plasma 64

membrane as well as intracellular membranes (Gkantiragas et al., 2001), prompted

studies on the distribution of Glut1 in plasma membrane DRMs (PM-DRMs) in control

and stimulated cells. In these experiments, cells were first surface-biotinylated (Shetty et

al., 1993b) and following lysis in Triton X-100 and sucrose-density centrifugation,

samples were incubated with streptavidin-agarose beads. This protocol yields an estimate

of the content of Glut1 in the PM-DRM of control and stimulated cells, although the

value may represent an underestimate since the biotinylation reaction and isolation procedure may be incomplete. Nevertheless, in three independent experiments, after 90 minutes of exposure to 5 mM azide, the percentage of PNL cell Glut1 present in the plasma membrane-DRM fraction decreased by 52 ± 12% (P < 0.05) from 5.9% to 2.8%

in control and azide-treated cells, respectively (Figure 4a and 4b). 65

Discussion

We demonstrate the presence of Glut1 in the DRM domain in three different cell types.

The results of our studies are consistent with the reported localization of Glut1 in DRMs

in Hela cells and with the presence of Glut4 in DRMs of the plasma membrane of

adipocytes (Gustavsson et al., 1996; Karlsson et al., 2002; Ros-Baro et al., 2001; Sakyo et

al., 2002; Scherer et al., 1994). Furthermore, we show the redistribution of Glut1 out of this domain upon treatment with azide. These findings raise a number of intriguing questions, including, 1) What is the mechanism by which a fraction of cell Glut1 is incorporated and retained in DRMs?, 2) What sequence of events lead to movement of

Glut1 out of these domains?, 3) Is the movement selective to Glut1?, and 4) To what extent, if any, does the re-distribution of Glut1 account for the stimulation of glucose transport? Concerning the latter point, the co-localization of Glut1 and stomatin in the

DRMs, and our previous finding of the inhibitory effect of stomatin on Glut1 function

(Zhang et al., 2001), raises the possibility that the fraction of Glut1 in this membrane microdomain may be inactive (“masked”).

Mechanisms proposed for localization of proteins to DRMs include the affinity of the protein’s transmembrane domains for the sphingolipid- and cholesterol-rich membrane microdomains, anchoring of proteins by glycophosphatidylinositol linkage,

protein-protein interactions, or co- or post-translational acylation (palmitoylation)

(Brown et al., 2000). While the mechanism by which Glut1 is associated with the

DRMs is not known, we speculate that palmitoylation may be a potential mechanism.

We propose this for several reasons, including a) palmitoylation is reversible and can

occur in the time scale of the observed stimulation of glucose transport in response to 66

azide (Baker et al., 1997; Loisel et al., 1996), b) Glut1 contains an ideally positioned

cysteine (Cys 207) that can potentially serve as a substrate for palmitoylation (Charrin et

al., 2002; de Vetten et al., 1988; Hartel-Schenk et al., 1992), and c) Glut1 has been

reported to be palmitoylated (Pouliot et al., 1995). Furthermore, protein acyltransferase

activity has been localized to the DRM fraction in some cell types, thereby implicating its

compartmentalization as a factor in the recruitment of other proteins to this domain

(Dunphy et al., 2001).

At present, the signals and sequence of events leading to the movement of Glut1

out of DRMs in response to treatment with azide are not known. Likewise, the role of the

redistribution of Glut1 in the observed enhancement of glucose transport following

inhibition of oxidative phosphorylation remains unresolved. Nevertheless, based on

previous observations that stomatin acts as a Glut1-binding protein and appears to have

an inhibitory effect on Glut1 function (Zhang et al., 2001; Zhang et al., 1999b), it is

tempting to speculate that the movement of Glut1, but not of stomatin, out of the DRMs

in response to azide may contribute to “activation” of Glut1 intrinsic function. At the

present time, the magnitude of this contribution is not known especially considering the

relative change in the content of Glut1 in DRMs compared to the larger increment in the rate of glucose transport. Further studies are necessary to explore the underlying mechanisms and to determine the role of Glut1 redistribution in the stimulation of

glucose transport in response to the inhibition of oxidative phosphorylation. 67

Acknowledgements

We thank Bonnie Gorzelle, Donna Horvath, and Bridget Roth for their able technical assistance. We also thank Dr. Rainer Prohaska for his generous gift of the anti-human stomatin antibody. This study was supported in part by NIH grant, DK45945. 68

Figure 2.1. Distribution of Glut1, stomatin, transferrin receptor, and caveolin-1 in

detergent-soluble and -insoluble fractions. (A). Stably-transfected Clone 9 cells

expressing human stomatin were lysed in buffer containing 0.5% Triton X-100, and fractionated in a sucrose density gradient. Fractions #1 (top of the gradient) and #9

(pellet) showed no reaction to any of the antibodies employed and are not shown. (B).

3T3-L1 fibroblasts were fractionated as above. (C). Human red blood cell ghosts were

lysed in buffer containing 1% Triton X-100 prior to fractionation. Similar results were

obtained in at least four independent experiments performed on each of the cell types.

A Fraction 2 3 4 5 6 7 8 121 Anti-hTfR1 86 69 52 Anti-Glut1 40 Anti-Stomatin

28 Anti-Caveolin-1 22

B Fraction 2 3 4 5 6 7 8

86 Anti-Glut1 69 52 28 Anti-Caveolin 1 22

C Fraction 2 3 4 5 6 7 8

86 69 Anti-Glut1 52 40 Anti-Stomatin 70

Figure 2.2. Distribution of Glut1, stomatin, transferrin receptor, and caveolin-1 in detergent-soluble and -insoluble fractions employing different non-ionic detergents.

Transfected Clone 9 cells were lysed in buffer containing Triton X-100, CHAPS, and

NP-40 (all at 0.5%), and then subjected to the one-step fractionation procedure detailed under Material and Methods. Similar results were obtained in two independent experiments.

Triton X-100 CHAPS NP-40 Fraction W S P W S P W S P 121 Anti-hTfR1 86 69 Anti-Glut1 52 40 Anti-Stomatin

28 Anti-Caveolin-1 22

W=Whole; S = Soluble; P = Pellet 72

Figure 2.3. Effect of azide on the distribution of Glut1, stomatin, transferrin

receptor, and caveolin-1 in detergent-soluble and -insoluble fractions. (A).

Transfected Clone 9 cells were treated with diluent or 5 mM azide for 90 min prior to

lysis in Triton X-100 and fractionation. The resulting blots were incubated with the specific antibodies shown. (B). In control or azide-treated cells, the densities of bands

representing cell Glut1, stomatin, and caveolin-1 present in fractions 2 plus 3 were

expressed as a ratio to the total of densities of the appropriate bands in all fractions. The

experiment was repeated five times and results averaged (mean ± SE). *denotes P <0.05. 73

A Fraction 2 3 4 5 6 7 8 2 3 4 5 6 7 8 121 86 Anti-hTfR1 69 52 Anti-Glut1

40 Anti-Stomatin 28 22 Anti-Caveolin-1 Control 90 min 5 mM azide 74

1.2

0.8

0.6

0.4 Fraction in the DRMs

0.2

0 Glut1 Stomatin Caveolin-1

Table 2.1. Effect of exposure of Clone 9 cells to azide on stimulation of glucose transport and the % decrease of Glut1 in DRMs

Duration of Exposure Fold Stimulation of Glucose Transport % Decrease of Glut1 in DRM 30 minutes 2.5 ± 0.5* 25 ± 2%

90 minutes 4.5 ± 0.8* 38 ± 11%

Clone 9 cells were exposed to 5mM azide for 30 and 90 minutes. The effect of azide on glucose transport and on the % decrease of Glut1 in the DRM fractions was determined. Mean ± S.E.; n =

3-4. *P < 0.05. 76

Figure 2.4. Effect of azide on the distribution of Glut1 in plasma membrane DRMs.

(A). Transfected Clone 9 cells were treated with diluent or 5 mM azide for 90 min prior

to cell surface biotinylation, lysis in Triton X-100, sucrose-density centrifugation, and

streptavidin-agarose pull down. Cells not reacted with the biotinylation reagent were also

examined. “pd” represents 40% of the material pulled down by the streptavidin-agarose

beads; “sup” represents 10% of the material remaining in the supernate. (B). The

percentage of total cell Glut1 present in fractions 2 plus 3 in control and azide-treated

cells (and corrected for non-specific “pulldown”) was determined in parallel. The

fractions of the sample used in the “pd” and “sup” as well as the densitometric readings

of Glut1 in all fractions were employed in the calculation of the percentage of cell Glut1 present in the plasma membrane DRMs. The experiment was repeated three times and results averaged (mean ± SE). *denotes P <0.05. 77

A Fraction 2 3 pd post pd post

Control

Azide

No Biotinylation

Loaded 40% pulldown and 10% post-pulldown supernate. 78

0.070 Control 0.060 90 min 5 mM Azide

0.050

0.040 * 0.030

0.020

0.010 Fraction of total Glut1 PM-DRM Glut1 total of Fraction 0.000

Chapter 3

Distribution of Glut1 in Clone 9 cells and to Plasma Membrane DRM

and non-DRM Domains: Effect of Azide

(Parts of this chapter have been submitted for publication and presented elsewhere.

Rubin D and Ismail-Beigi F. The FASEB Journal 18: A697, 2004)

Introduction

Glut1, a widely expressed member of the facilitative glucose transporter family, is an

intrinsic membrane glycoprotein with 12 putative transmembrane domains (Joost et al.,

2001). The transporter is thought to mediate much of the basal, non-insulin-dependent transport of glucose (Ismail-Beigi, 1993; Mueckler, 1994). Glut1 expression and function is regulated by a variety of stimuli and agents including serum, growth factors, transformation, hypoxia, and inhibitors of oxidative phosphorylation, with the latter two being associated with stimulation of 5’-AMP-activated protein kinase (AMPK) (Abbud et al., 2000; Barnes et al., 2003; Barnes et al., 2002; Bashan et al., 1992; Behrooz et al.,

1997; Behrooz et al., 1998; Hamrahian et al., 1999; Hayashi et al., 2000; Ismail-Beigi,

1993; Mercado et al., 1989; Ouiddir et al., 1999; Shetty et al., 1993a; Shetty et al., 1992;

Shi et al., 1995; Xi et al., 2001). The enhancement of Glut1-mediated glucose transport

in response to inhibition of oxidative phosphorylation is biphasic, with the acute response

(1-2 h) being mediated entirely by post-translational mechanisms and resulting in a 4- to

10-fold increase in the rate of glucose transport. The chronic response at 24 hours is

mediated at multiple levels including transcriptional, post-transcriptional, and post-

79 80

translational mechanisms and results in a further 2 to 3 fold increase in glucose transport

above that achieved during the acute phase(Shetty et al., 1992; Shetty et al., 1993b).

The acute stimulation of glucose transport in Clone 9 cells in response to the

inhibition of oxidative phosphorylation after 1 and 2 hours exposure to azide was

previously determined to involve 1.5- and 1.8-fold increases, respectively, in the

abundance of Glut1 in the plasma membrane (Shetty et al., 1993b). The stimulation of

glucose transport is associated with an increase in the Vmax of transport without a change

in the affinity of the transporter for glucose, or a decrease in the energy of activation for

glucose transport (Hamrahian et al., 1999; Mercado et al., 1989). These characteristics

suggest that the acute stimulation of glucose transport predominantly involved activation

or “unmasking” of glucose transporters pre-existing in the plasma membrane with

translocation having a relatively minor role.

Results of our recent study employing 0.5% Triton X-100 at 4oC suggested that

~38 % of the Glut1 in post-nuclear lysates of Clone 9 cells is concentrated in low density,

detergent-resistant membrane microdomains (DRMs); this was also true to a similar

extent in 3T3-L1 fibroblasts and human red blood cells (Rubin et al., 2003). DRMs, also

known as lipid rafts, are biochemically defined by their high content of cholesterol and sphingolipids, low density, and resistance to solubilization in buffers containing non- ionic detergents such as Triton X-100 at 4 oC. DRMs are found in the plasma membrane

(PM) as well as in intracellular membranes, and are enriched in specific proteins (Brown et al., 2000; London et al., 2000; Simons et al., 1997). Some integral membrane proteins including caveolin-1, hemagglutinin, the insulin receptor, and Glut4 have been shown to associate with DRMs (Gustavsson et al., 1999; Saltiel et al., 2003; Scheiffele et al., 1997; 81

Scherer et al., 1994; Skibbens et al., 1989; Vainio et al., 2002). There is on-going controversy as to the exact relationship between the insulin receptor, Glut4, and DRMs as well as the effect of insulin on this relationship (Bickel, 2002; Chamberlain et al., 2002;

Karlsson et al., 2002; Scherer et al., 1994; Smith et al., 1998; Souto et al., 2003).

Our previous work also demonstrated that in post-nuclear lysates of Clone 9 cells exposed to 5 mM azide for 90 minutes, the percentage of Glut1 in the DRM fraction was decreased by 40 %. The association of Glut1 with DRMs, and the azide-induced decrease in that association, raised the possibility that the abundance of Glut1 in the non-

DRM fraction of the plasma membrane may have an important role in the control of glucose transport. These considerations prompted us to study the subcellular distribution of Glut1 in some detail in control and azide-treated cells by posing the following set of questions: 1) What is the subcellular distribution of Glut1 in the Clone 9 cells, 2) What fraction of Glut1 in the PM localizes to the DRMs and non-DRM domains under basal conditions and, 3) Does exposure to azide affect the distribution of Glut1 between DRMs and non-DRM domains of the plasma membrane? We have employed cell fractionation and confocal imaging techniques to address these questions. Our results demonstrate that i) approximately 30% of Glut1 in the post-nuclear homogenate (PNH) of Clone 9 cells under basal conditions resides in the plasma membrane fraction; ii) approximately 50 % of Glut1 in the plasma membrane resides each in the DRMs and non-DRM domains under basal conditions; and iii) exposure to azide is associated with a ~2.9 fold increase in the abundance of Glut1 in the non-DRM domain of the plasma membrane fraction without a significant change in the abundance of Glut1 in the DRMs of the PM. Some of the above findings have been presented recently (Rubin et al., 2004). 82

Materials and Methods

Materials Cell culture reagents were purchased from Invitrogen (Carlsbad, CA),

plasticware and glassware through Fisher Scientific, chemicals and equipment for

electron microscopy were purchased from Electron Microscopy Sciences (Fort

Washington, PA), and unless otherwise noted, all other chemicals were from Sigma

Aldrich or Fisher Scientific. 4 ml ultracentrifugation tubes fitting the Sorvall TST 60.4

swinging bucket rotor, 36 ml ultracentrifugation tubes fitting the Sorvall AH629

swinging bucket rotor, 0.5 ml ultracentrifugation tubes fitting the Beckman TLA 100.1

rotor were purchased from Kendro.

Antibodies were purchased from the following companies: mouse monoclonal

anti-caveolin-1 (BD Transduction Laboratories), mouse monoclonal anti-trans Golgi

Network 38 (Oncogene Research Products, San Diego, CA), rabbit anti-Glut 1

(Chemicon, Temecula, CA), mouse monoclonal anti-Na+/K+-ATPase alpha 1 (Sigma

Aldrich, St. Louis, MO), mouse monoclonal anti-human transferrin receptor 1 (Alpha

Diagnostics International, Inc, San Antonio, TX), rabbit anti-calnexin (Santa Cruz

Biotechnology, Inc, Santa Cruz, CA), rabbit anti-caveolin 1 (Santa Cruz Biotechnology,

Inc), horseradish peroxidase-conjugated anti-rabbit and anti-mouse IgG secondary antibodies (Sigma Aldrich), highly cross-absorbed Alexa Fluor 633-conjugated goat anti- rabbit IgG (Molecular Probes, Eugene, OR), highly cross-absorbed Alexa Fluor 488-

conjugated goat anti-mouse IgG (Molecular Probes). Mouse ascites containing anti-rat

protein disulfide isomerase (PDI) monoclonal antibody was a gift from Dr. Michael

Lamm (Case Western Reserve University, Cleveland, OH). Purified mouse

immunoglobulin containing anti-Na+/K+-ATPase alpha 1 was a gift from Dr. George 83

Dubyak (Case Western Reserve University, Cleveland, OH). The anti-fade mounting

media Vectashield with Dapi and Fluoromount G were purchased from Vector

Laboratories, Inc (Burlingame, CA) and Fisher Scientific respectively. Nitrocellulose

membrane was purchased from Schliecher & Schuell (Keene, NH). Western blots were

developed using the enhanced chemiluminescence kit from Santa Cruz Biotechnology

(Santa Cruz, CA).

o Cell culture Clone 9 cells were maintained at 37 C and 8 % CO2 in low glucose DMEM supplemented with 10 % bovine calf serum (v/v) and penicillin-streptomycin. Culture medium was replaced with fresh medium containing 10 % bovine calf serum (v/v) 16-24 hours prior to initiation of experiments. All experiments in control and azide-treated cells were performed in parallel.

Glucose Transport Cytochalasin B-inhibitable glucose transport was measured using 3H-

3-O-methylglucose, as previously described (Hamrahian et al., 1999; Mercado et al.,

1989; Rubin et al., 2003).

Isolation and fractionation of the plasma membrane This method was adapted from

McKeel and Jarret and others (McKeel et al., 1970; Shetty et al., 1993b; Simpson et al.,

1983). Confluent cells in 45 to 50 10-cm dishes were washed twice then scraped twice

with 0.5 ml ice cold PBS. All subsequent steps were performed at 4 oC. Cell pellets were homogenized using 10 gentle strokes of a 15 ml Dounce homogenizer in approximately 10 ml of homogenization buffer (0.25 M sucrose in Buffer A (20 mM Tris 84

pH 7.4, 1 mM EDTA, and 0.1 µM PMSF)). The homogenate was centrifuged 20 minutes

at 27 k x g (15 krpm) in the Sorvall SS-34 fixed-angle rotor. The supernate contained

approximately 2 % of the Glut1 content and was discarded. The pellet was resuspended

in approximately 4 ml in homogenization buffer and homogenized with 5 gentle strokes

of a Potter-Elvehjem Teflon pestle homogenizer. The homogenate was diluted to 15 ml

with homogenization buffer and centrifuged 5 minutes at 470 x g and the supernate

saved. This new pellet containing nuclei, unbroken cells, and some membranes was

washed using 4 – 5 ml of homogenization buffer, centrifuged, and the supernate pooled

with the previous supernate; this fraction was labeled as the post-nuclear homogenate,

PNH, and was used in subsequent experiments. A preparative sucrose step gradient was

formed by layering 18 ml of the PNH onto an 18 ml cushion of 1.12 M sucrose in Buffer

A in 36 ml tubes fitting the Sorvall AH-629 swinging bucket rotor. Samples were

centrifuged at ~ 100 k x g (24 krpm) for 1 hour at 4 oC. The interface between the two phases of the gradient (the plasma membrane-enriched fraction), the sucrose cushion, and the gradient pellet were removed and saved. The supernate above the interface of the gradient had an average of less than 10 µg protein with only minor amounts of Glut1 (n =

3), and was discarded. The gradient pellet was resuspended to ~ 1 ml in Buffer A. The

interface was diluted to 15 ml with buffer A and centrifuged at 27 k x g for 36 minutes in

the Sorvall SS-34; this pellet was resuspended to approximately 100 – 200 µl with Buffer

A, and labeled as the plasma membrane fraction. Approximately 75% of the sucrose

cushion of the preparative sucrose step gradient was diluted to 36 ml with Buffer A and

centrifuged at ~ 100 k x g for 1 hour at 4 oC in the Sorvall AH-629. This pellet was

resuspended in 100 – 200 µl of buffer A and labeled as the sucrose cushion. 85

In the “float” experiments, 100 µg of protein from samples of the plasma membrane fraction, sucrose cushion, and gradient pellet were centrifuged in the Beckman tabletop ultracentrifuge TLA 100.1 fixed-angle rotor at an average of 128 k x g (60 krpm)

for 30 minutes at 4 oC (there was no anti Glut1 immunoreactivity detected in the supernates). The pellets were solubilized in buffer B (50 mM Hepes, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA, 20 mM sodium pyrophosphate, 1 mM sodium orthovanadate,

20 mM sodium fluoride, and 0.1 µM PMSF; pH 7.4) with the indicated concentrations of

Triton X-100, then transferred to 4 ml centrifugation tubes for the Sorvall TST 60.4

swinging bucket rotor, and brought to 385 µl with either buffer A or buffer B containing

no detergent (there was no difference in the results obtained with either). The samples

were adjusted to 40 % sucrose using 65% sucrose in buffer C (20 mM Tris, 150 mM

NaCl, 1 mM EDTA, 0.1 µM PMSF; pH 7.4) to a total volume of 1 ml and overlaid with 2

ml 35 % sucrose and 1 ml 5 % sucrose in buffer C. Samples were centrifuged at an

average of 150 k x g (38 krpm) overnight at 4 oC. 500 µl fractions from each “float” tube

were collected from the top. 30 µl or 60 µl of fractions 2 - 8 were solubilized with 4 x

Laemmli buffer, fractionated by SDS-PAGE and analyzed by immunoblot.

Immunoblot films derived from the various fractions were scanned and the

intensity of the bands was measured using the Bio-Rad Gel Doc 1000 system. The

abundance of each antigen in the various cell fractions was calculated utilizing the

intensity of the band, the amount of protein loaded in the lane, and the total yield of

protein in the fraction. In the “float” preparations, Glut1 content in the DRM and non-

DRM fractions of the PM in azide-treated cells were calculated relative to control. The

fractional distribution of Glut1 between the DRM and non-DRM fractions was calculated 86

by dividing the sum of Glut1 intensity in fractions 2 plus 3 by the total Glut1 intensity in

all fractions.

To estimate azide-induced changes in Glut1 content, the ratio of Glut1 intensity in

azide-treated over control cells was calculated from analysis of 3 µg samples of the

plasma membrane or from the total intensity across “float” gradient fractions 2 – 8 on the

same immunoblots of the plasma membrane, cushion, and gradient pellet . The Glut1

content of PM in control cells was set to 1.0, and the Glut1 content of PM in azide-treated

cells was calculated relative to control. The abundance of Glut1 in the DRM and non-

DRM fractions in the PM of control and azide-treated cells was then calculated by

multiplying the fractional distribution by the total intensity of Glut1 in the PM.

Electron Microscopy All steps were performed at room temperature unless otherwise

indicated. The plasma membrane, sucrose cushion, and gradient pellet fractions of the

preparative sucrose step gradient were fixed with 2% glutaraldehyde in 0.1 M sodium cacodylate buffer for 2 hours then at 4 oC overnight. Each sample was washed three

times with mild agitation for 15 minutes with 0.1 M sodium cacodylate buffer then post-

fixed with 1% OsO4 in 0.1 M cacodylate buffer for 1 hour and washed three times for 15

minutes each with double-distilled water. Samples were incubated with 1 % uranyl

acetate in double-distilled water for 1 hour for en-bloc staining then washed three times

for 15 minutes with double-distilled water. Samples were dehydrated by incubation

through several changes of ethanol in water at progressively increasing concentration (50

% for 10 min, 75 % overnight, 95 % for 10 min, then incubated with 100 % three times

for 15 min). This was followed by three changes in 100 % propylene oxide (PO) for 15 87

min per change. The samples were infiltrated with resin by incubating for one hour with

three changes of an increasing ratio of resin to PO each hour (1 : 2, 1 : 1, and 2 : 1).

Finally, samples were incubated with 100 % resin overnight then embedded in resin at 60

oC overnight before sectioning, mounting, and viewing.

Immunocytochemistry These methods were modified from procedures described previously (Shetty et al., 1993b). Clone 9 cells were grown in four-well chamber slides

(Fisher Scientific) then fixed by washing with -20 oC methanol, incubating with the same for 3 minutes at -20 oC, and washing with -20 oC acetone. The pattern of staining was no different if these steps were performed on the bench or in the -20 oC cold room.

Paraformaldehyde fixation (0.5 – 4 %) did not change the pattern of staining observed after methanol/acetone fixation; specifically the apparent proportion of Glut1 at the cell surface was not different using either method. After air drying, the cells were permeabilized with 0.2 % Triton X-100 for 10 minutes at 4 oC, washed with cold PBS, then blocked for 1 hour at room temperature with 10 % fetal bovine serum (FBS) in PBS supplemented with 0.01 % Thimerosal (referred to as FBS-T). Primary antibodies were diluted to an appropriate concentration in FBS-T and incubated at 4 oC overnight. The cells were washed by gentle agitation in five separate one liter water baths at room temperature. Secondary antibodies were diluted to an appropriate concentration in FBS-

T and incubated at room temperature for 30 minutes. After washing, the chambers and gasket were removed from the slide and approximately 30 µl of a 1:1 mixture of

Vectashield and Fluoromount G was applied before a 50 x 22 mm coverslip was placed

and sealed with quick-drying clear fingernail polish. 88

Confocal Microscopy and Analysis Images were acquired with a Zeiss LSM 510 inverted laser-scanning confocal fluorescence microscope in the Case Western Reserve University

Ireland Comprehensive Cancer Center confocal microscopy facility using both X 40 numerical aperture 1.3 and X 63 numerical aperture oil planapochromat objectives.

Confocal images of Alexa fluor 633 fluorescence were acquired using either 543-nm or

633-nm excitation light from a HeNe laser and a 650-nm long pass filter. Images of

Alexa fluor 488 flourescence were acquired using 488-nm excitation light from an argon laser and a 500-550 band pass barrier filter. Images were analyzed with the Metamorph image analysis program (Universal Imaging). Fluorescence intensity was recorded from regions of interest covering an entire cell; intensity values were exported to Microsoft

Excel for analysis.

Statistical Analysis All results are expressed as mean ± standard error of the mean

(SEM). Data was analyzed for statistical significance using the two-tailed, unpaired

Student’s t-test or by chi-squared analysis, as appropriate.

Results

Cellular Distribution of Glut1 in Clone 9 cells

To gain a better understanding of mechanisms underlying the stimulation of glucose

transport in response to inhibition of oxidative phosphorylation, we determined the

cellular distribution of Glut1 in Clone 9 cells using subcellular fractionation and

immunocytochemistry techniques. Biochemical fractions of confluent Clone 9 cells were

prepared as defined in Materials and Methods using a preparative sucrose step-gradient

which yielded a plasma membrane-enriched fraction, the sucrose cushion, and a gradient

pellet fraction. Each fraction was analyzed by electron microscopy and was found to

have distinct morphological characteristics such as the distribution of vesicle size, as well

as common elements including vesicles that resembled mitochondria. The sucrose

cushion fraction contained a relatively more homogenous population of vesicles. It was

also noted that the mitochondria-like vesicles were least abundant in the plasma

membrane fraction and most abundant in the gradient pellet (Figure 1).

Each fraction was analyzed by immunoblot for the Na+/K+-ATPase α1 subunit,

protein disulfide isomerase (PDI), calnexin, caveolin-1, and Glut1 (Figure 2). In repeated experiments, and in agreement with previous reports regarding the efficacy of the fractionation procedure we have adapted (McKeel et al., 1970; Simpson et al., 1983), both the Na+/K+-ATPase α1 subunit and Glut1 were enriched ~15 fold in the PM

compared to the PNH. Both protein disulfide isomerase (PDI) and calnexin are residents of the endoplasmic reticulum and are thought to be involved in protein folding (Ellgaard et al., 2003; Kaetzel et al., 1987; Oliver et al., 1996). While the plasma membrane- enriched fraction also contained PDI and calnexin, the fractional amount of PDI in the 90

PM was less than that observed for Na+/K+-ATPase α1 subunit, caveolin-1, and Glut1

(Table 1). The fractionation pattern of PDI agrees with the notion that endoplasmic

reticulum membranes may connect with the plasma membrane, as shown in certain cell

types (Desjardins, 2003; Gagnon et al., 2002; Garin et al., 2001).

Results of several experiments aimed at estimating the relative abundance of Glut1 and

the other antigens in the three fractions of the preparative sucrose step gradient are

summarized in Table 1. On average ~28 ± 5 % of the Glut1 in the PNH was found in the

plasma membrane fraction, while the majority (52 ± 7 %) was found in the gradient pellet

and considerably less in the sucrose cushion (10 ± 2 %). The distribution of Na+/K+-

ATPase α1 subunit and caveolin-1 were similar to that of Glut1. In contrast, the bulk of

PDI was found in the gradient pellet with only ~10 % being present in the PM. These analyses suggested that much of the Glut1 present in the post-nuclear homogenate (PNH) may be derived from regions of the cell other than the plasma membrane. Consistent with this premise are that Glut1 is present in all three fractions, and that the bulk of Glut1 is found in the gradient pellet which also contains the bulk of a known endoplasmic reticulum antigen.

To further determine the distribution of Glut1 in Clone 9 cells, its localization and colocalization with the proteins mentioned above as well as with the Trans Golgi

Network 38 (TGN38) protein was determined. Immunocytochemistry by confocal microscopy was performed and the results analyzed with the Metamorph image analysis program (Figure 3). In preliminary experiments, various fixation techniques including

0.5 – 4 % paraformaldehyde were employed. We found that the images were superior with use of methanol/acetone fixation, and importantly that the ratio of Glut1 in the 91 plasma membrane to total cell Glut1 immunofluorescence did not appear to be higher using the other fixation techniques. It is evident in the images shown in Figure 3 that a significant amount of Glut1 is present in various parts of the cell other than the PM. In keeping with the previous immunofluorescence studies of Glut1in these cells (Koseoglu et al., 1999; Shetty et al., 1993b), we found a significant amount of Glut1 in the peri- nuclear region. Although the fine detail of Glut1 immunofluorescence in Clone 9 cells was obscured by the thresholding techniques inherent in the analysis program, an upper limit of the colocalization of Glut1 immunofluorescence with that of other antigens was determined (Table 2). Analysis of the colocalization of each antigen with Glut1 demonstrates a significant degree (> 95%) of association. On the other hand, the reverse analysis, i.e. colocalization of Glut1 with each antigen, demonstrated that Glut1 associated the most with Na+/K+-ATPase α1, reflecting that the two integral membrane proteins may share a common cellular distribution. Taken together, the above results derived from subcellular fractionation and immunofluorescence support the conclusion that substantial amounts of Glut1 is present in various intracellular pools in Clone 9 cells.

Distribution of Glut1 in DRM and non-DRM fractions of the plasma membrane of control

Clone 9 cells

Given the higher content of cholesterol and sphingolipids in the PM compared to intracellular membranes (Dawidowicz, 1987; Ohvo-Rekila et al., 2002), we first determined the concentration of Triton X-100 required for a valid estimation of DRMs and non-DRM domains in the PM by assessing the distribution of caveolin-1 in samples of plasma membrane fractions solubilized at various detergent: protein (mass: mass) 92

ratios (Figure 4). It is generally accepted that the bulk of caveolin-1 in cell membranes

is localized in DRMs (Scherer et al., 1994). Caveolin-1 was present only in the low density fractions (fractions 2 and 3) at a detergent: protein ratio of 4.16, and was increasingly distributed to the higher density fractions as that ratio was increased; in those experiments, the amount of protein and the volume of extraction was constant.

Based on these results, a detergent: protein ratio of 4.16 was employed in all subsequent experiments on the plasma membrane-enriched fraction.

The distribution of Glut1 in DRM and non-DRM domains of the PM in control cells is shown in Figure 5 and is summarized in Figure 6. The average Glut1 distribution to the DRMs and non-DRM domains were 50 ± 7% and 46 ± 7%, respectively (mean ± SE, n = 9). It should be noted that the 50% distribution of Glut1 in the DRM fraction of PM is higher than the ~38% value previously reported in the post- nuclear lysates of these cells (Rubin et al., 2003).

Effect of azide on the distribution of Glut1 in the DRM and non-DRM fractions of the

plasma membrane

Upon exposure to 5 mM azide for 90 minutes, there was a 4.6 ± 0.3-fold stimulation of

cytochalasin B-inhibitable glucose transport in accord with previous reports (Hamrahian

et al., 1999; Mercado et al., 1989; Rubin et al., 2003; Shetty et al., 1992; Shetty et al.,

1993b). At the same time, we found that the fractional distribution of plasma membrane

Glut1 in the DRM domain was somewhat decreased and increased in the non-DRM

domain relative to that in control cells; however, neither change was statistically

significant (Figure 6). The abundance of Glut1 in the plasma membrane fraction of

azide-treated cells increased 1.6 ± 0.2-fold (mean ± SE, n = 9) relative to control, but this

change also did not reach statistical significance. However, treatment with azide did

result in a significant 2.9 ± 0.8-fold increase in the content of Glut1 in the non-DRM

domain and a non-significant 1.1 ± 0.1-fold increase in the DRMs of the PM (Figure 7 and Table 3). To identify the cellular source of Glut1 “translocating” to the non-DRM domain of the PM, we ascertained whether the Glut1 content of the sucrose cushion and gradient pellet fractions (and their DRM and non-DRM domains) was changed in azide- treated cells. In three independent experiments, no statistically significant decrease in either of the fractions or their DRM and non-DRM domains was observed (data not shown). 94

Discussion

Results of previous studies employing cell fractionation and immunofluorescence had suggested that a significant fraction of Glut1 in Clone 9 cells may be localized in sites other than the plasma membrane (Rubin et al., 2003; Shetty et al., 1993b). This issue is of importance because the fractional distribution of Glut1 between plasma membrane and intracellular sites would place an upper limit on the extent by which translocation of

Glut1 to the PM can account for any short-term stimulation of glucose transport.

Employing a more detailed subcellular fractionation procedure and using confocal microscopy, we estimate that ~30% of Glut1 in Clone 9 cell post-nuclear homogenate is localized to the plasma membrane. The co-localization of Glut1 with the Na+/K+-ATPase

α1 subunit by confocal microscopy correlated well with the distribution of these integral membrane proteins in the preparative sucrose gradient. Results also showed that Glut1 was co-localized the least with trans-Golgi network 38 protein. Glut1 significantly co- localized with caveolin-1 in the microscopy and fractionation experiments. Of note, the previous result demonstrating that ~38% of Glut1 in post-nuclear lysates co-fractionated with caveolin-1 (Rubin et al., 2003) and the present result of ~22% co-localization by

confocal microscopy (Table 2) raises the possibility that a fraction of Glut1 resides in

DRMs that are not caveoli. This premise needs to be examined by immunoelectron microscopy or immunoabsorption of caveolin-1-containing vesicles.

In contrast to the ~38% localization of Glut1 in DRMs in post-nuclear cell lysates,

~50% of Glut1 in the plasma membrane is localized to the DRM domains (Table 3).

This result is based on separation of DRM and non-DRM domains of plasma membrane

according to the distribution of caveolin-1, a known DRM-resident protein (Scherer et al., 95

1994), at varying detergent: protein ratios. The DRM microdomains have been implicated in the regulation of some signaling proteins including the receptor IgE

(Wilson et al., 2000), the insulin receptor (Baumann et al., 2000; Vainio et al., 2002), and

CD9 (Claas et al., 2001), as well as specific cellular functions including solute transport

(e.g. potassium channels (Martens et al., 2001)). Hence, the distribution of Glut1

between the DRM and non-DRM domains of the plasma membrane may be of

physiological importance and may play a regulatory role in the control of Glut1-mediated

glucose transport (Rubin et al., 2003).

Exposure of Clone 9 cells to 5 mM azide for 90 min induced a significant ~2.9-

fold increase in the content of Glut1 in the non-DRM domain of the plasma membrane

without a change in the abundance of Glut1 in the DRMs. This finding strongly suggests

that Glut1 is translocated from one or more intracellular compartments to the non-DRM

domain of the PM, and that the net increase in the abundance of Glut1 in this domain of

the PM may play an important role in the stimulation of glucose transport in response to

inhibition of oxidative phosphorylation. The higher content of Glut1 in the plasma

membrane-enriched fraction (1.6-fold) or in the non-DRM domain of the PM (2.9-fold) is

less than the ~4.7-fold increase in the rate of glucose transport measured following 90

minutes of exposure to 5 mM azide. Taken together, the results suggest that while

translocation of Glut1 to the non-DRM domain of the PM may account for some of the

stimulation of glucose uptake, “activation” steps may be necessary for the full transport

response.

In a previous work we reported that treatment with 5 mM azide for 90 min was

associated with an ~40 % decrease in the abundance of Glut1 in the DRM domains of 96

post-nuclear cell lysates prepared with 0.5% Triton X-100 at 4 oC and a ~55% decrease in

Glut1 association with the DRM in the plasma membrane as isolated by the surface biotinylation technique; using surface biotinylation the content of Glut1 in DRMs decreased from 5.9 % of Glut1 in the post-nuclear lysates to 2.8 % (Rubin et al., 2003).

Given that in the current study we did not observe a net decrease in the Glut1 content of

DRMs in the plasma membrane fraction, we examined the DRM and non-DRM domains of the sucrose cushion and gradient pellet of the post-nuclear homogenate to identify the source of the decrease in Glut1 in DRMs and to determine the source of Glut1 that translocated to the non-DRM domain of the PM. We failed to identify either of the sources in our analysis of the other fractions of the PNH from control and azide-treated cells, perhaps due to the small fraction of cell Glut1 in the DRMs of the plasma membrane or to experimental variability which would preclude a measurement of small changes. An alternative source of Glut1 could be the nuclear pellet which contains some membranes in addition to nuclei and unbroken cells; this fraction was not examined in our studies. Finally, a caveat to the interpretation of our previous findings is the inability to directly biotinylate Glut1. Hence, the isolation of Glut1-containing DRMs is dependent on other DRM proteins which become biotinylated. Any change in the localization of these proteins in the DRMs of azide-treated cells would result in a decrease in the ability of surface biotinylation to isolate Glut1-containing DRMs. Further studies are required to identify the source of the increase in Glut1 abundance in the non-

DRM domain of the plasma membrane.

It has recently been reported that treatment of Clone 9 cells with 10 mM methyl-

β-cyclodextrin (mβcd), an agent that extracts cellular cholesterol and disrupt DRMs, 97 stimulated basal glucose transport to the same extent as 1 hour exposure to 5 mM azide, with no additive effect upon the combination of the two (Barnes et al., 2003). These results support our hypothesis that Glut1 localization to the DRM may be inhibitory

(perhaps due to stomatin), and that Glut1 molecules in the non-DRM domain are in an active functional state (Rubin et al., 2003). This hypothesis is made plausible by the potentially inhibitory lipid microenvironment of DRMs for transport proteins due to their low fluidity (Gidwani et al., 2001), and the presence of the DRM-resident protein stomatin, a potentially inhibitory Glut1-interacting protein (Barnes et al., 2003; Zhang et al., 2001). Dissociation of Glut1 from DRMs or its translocation from intracellular pools to the non-DRM domains of the PM could facilitate the presentation of active functional transporters to the cell surface. The results of our current study are consistent with this model. In contrast to the above, the results of a recent study on the stimulation of glucose transport (Glut1) in 3T3-L1 adipocytes in response to 12 hours of total glucose deprivation (in the presence of fructose) showed that the stimulation was associated with an increase in the percentage of Glut1 in the DRM domain of the plasma membrane

(Kumar et al., 2004) and suggested that Glut1 transporters associated with the DRM are functionally active. It is not known whether the discrepancy reflects the different cell models employed, the extreme condition of glucose deprivation with alterations in the glycosylation of Glut1, or changes in other potential modifications of newly synthesized

Glut1.

It is becoming increasingly clear that the distribution of Glut1 (and other proteins) in the DRM and non-DRM domains of the plasma membrane may be of great physiological importance. Protein-protein interactions (Baumann et al., 2000; Chiang et 98

al., 2001), acylation (Melkonian et al., 1999), possibly lipid-binding properties (Murata et

al., 1995), and elements of the transmembrane domain (Scheiffele et al., 1997) have been proposed as mechanisms that underlie the association of proteins with DRMs. Further studies are necessary to determine whether Glut1 present in the DRM versus non-DRM domains of the PM exhibit differential function with respect to glucose transport. 99

Acknowledgements

We thank Fumie (Jill) Nishiyama, Sriram Kasturi, Donna Horvath and Bonnie Gorzelle

for their able technical assistance. We also thank Midori Hitomi (Case Western Reserve

University, Department of Neuroscience electron microscopy facility) for her capable

assistance with electron microscopy; and Pamela Davis for the use of the Metamorph

Analysis software suite. This study was supported in part by NIH grant R01 DK61994-

01A1 to F. Ismail-Beigi. Studies employing confocal microscopy were supported by

R01 NS39469 to A.-L. Nieminen.

100

Figure 3.1. Electron microscopy of the three the fractions of the preparative sucrose density step gradient. Fractions were prepared as described in the Materials and

Methods and analyzed by electron microscopy. The scale bar represents 1 µm. 101

Cushion Gradient Pellet Plasma membrane 102

Figure 3.2. Distribution of various proteins in post-nuclear homogenate and its sub fractions. Immunoblots were prepare as described in the Materials and Methods. The amount of protein present in each lane is shown above each lane and the average yield of protein in each fraction is shown below. 9.3 PNH 30µg 0.3 Cushion 3µg Figure 2 3.5 Pellet 3µg 103 0.5 PM 3µg Anti-Na A Anti-Glut1 Anti-Caveolin 1 Anti-Protein DisulfideIsomerase Anti-Calnexin verage yield(mg) + /K/ + A TPase alpha1 104

Table 3.1. Fractional content of various proteins in fractions of the preparative sucrose step gradient.

Fraction Glut1 PDI Calnexin Na+/K+ ATPase α 1 Caveolin 1

Sucrose cushion 0.10 ± 0.02 0.07 ± 0.04 0.08 ± 0.01 0.15 ± 0.01 0.19 ± 0.09

Gradient pellet 0.53 ± 0.07 0.82 ± 0.10 0.46 ± 0.19 0.43 ± 0.15 0.43 ± 0.17

Plasma membrane 0.28 ± 0.05 0.10 ± 0.06 0.45 ± 0.19 0.42 ± 0.14 0.38 ± 0.09

The average protein content of fractions was as follows: cushion, 0.3 mg; pellet, 3.5 mg; and plasma membrane, 0.5 mg. The content of each antigen in each fraction was calculated using the relative densitometric intensity of the antigen per unit protein in the fraction and the protein content of the fraction. The fractional content of each antigen in each fraction was calculated as its content over the total (sucrose cushion + gradient pellet + PM). There were 3 experiments performed for each antigen except Glut1 for which there was 8 experiments.

105

Figure 3.3. Colocalization of Glut1 with markers of intracellular organelles and

other antigens in Clone 9 cells. Clone 9 cells were fixed using the methanol/acetone

method, permeabilized with 0.2% Triton X-100 for 10 minutes at 4 oC, washed, blocked and incubated with the indicated primary antibody at 4 oC overnight while rocking. The cells were washed and incubated with the appropriate secondary antibody at room temperature for 30 minutes. After washing, the coverslip was applied with mounting medium, sealed, and imaged using a Zeiss LSM 510 inverted laser-scanning confocal fluorescence microscope as indicated in the Materials and Methods. 106

Glut 1 Antigen Overlay

PDI a b c

TGN 38 d e f

Sodium Potassium ATPase alpha 1 g h i

Caveolin 1 j k l 107

Table 3.2. Colocalization of Glut1 with markers of intracellular organelles and other antigens in Clone 9 cells by confocal microscopy.

Average Average Percent Percent Antigen Integrated Integrated Colocalization Colocalization Intensity of Intensity of of Antigen of Glut1 Antigen Glut1 with Glut1 with antigen (x 106) (x 106)

Protein Disulfide Isomerase 5.7 ± 1.5 13 ± 3.3 99 ± 0.27 54 ± 1.8

Trans Golgi Network 38 0.89 ± 0.12 15 ± 1.8 99 ± 0.20 12 ± 0.99

Na+/K+-ATPase α 1 8.7 ± 1.2 15 ± 3.3 95 ± 1.17 81 ± 4.3

Caveolin 1 0.93 ± 0.21 9.5 ± 0.88 98 ± 0.34 22 ± 4.0

The number of cells analyzed from at least 2 experiments for each antigen is as follows:

Protein Disulfide Isomerase, 15; Trans-Golgi Network 38, 17; Na+/K+-ATPase α1, 17; and

Caveolin-1, 7. 108

Figure 3.4. The distribution of caveolin-1 in DRM and non-DRM fractions of

plasma membrane-enriched fraction as a function of the concentration of Triton X-

100. 100 µg of plasma membrane derived from control Clone 9 cells was solubilized in a

constant volume of lysis buffer with varying concentrations of Triton X-100 at 4oC. The solubilized material was fractionated and analyzed for caveolin-1 as detailed in Materials and Methods. The detergent: protein ratio is expressed as a mass ratio. Similar results were obtained in two additional experiments. 109

detergent: protein ratio HMW 1 2 3 4 5 6 7 8

28 4.16 22

28 6.25 22 Anti-Caveolin1

8.33 28 22

12.49 28 22 110

Figure 3.5. The distribution of plasma membrane Glut1 between DRM and non-

DRM fractions in control Clone 9 cells. 100 µg of plasma membrane derived from control Clone 9 cells was solubilized, fractionated, and analyzed for Glut1 immunoreactivity as described in Materials and Methods. This experiment was repeated several times with similar results. 111

Control

Bottom Top of Gradient of Gradient 8 7 6 5 4 3 2

69 Anti-Glut1 52 112

Figure 3.6. Fractional distribution of Glut1 in the DRM and non-DRM domains of the plasma membrane. 100 µg of plasma membrane derived from control and azide- treated Clone 9 cells was solubilized, fractionated, and analyzed for Glut1 immunoreactivity. The fractional distribution of Glut1 in the DRM and non-DRM domains was calculated as described in Materials and Methods. This experiment was repeated 9 times and the results were averaged; mean ± SE. 113

0.700

0.600 Control 0.500 Azide 0.400 0.300 0.200 0.100 0.000 Fraction of Total Glut1 Intensity Glut1 Total of Fraction DRM Non -DRM 114

Figure 3.7. The effect of azide on the content of Glut1 in the DRM and non-DRM fractions of the plasma membrane. 100 µg of plasma membrane derived from Clone 9 cells treated with diluent or for 90 minutes with 5 mM azide were analyzed for Glut1 immunoreactivity as described in Materials and Methods. This experiment was repeated

9 times and the results are summarized in Table 3.

115

detergent: Control 5 mM Azide 90 min protein ratio HMW 8 7 6 5 4 3 2 2 3 4 5 6 7 8

69 4.16 52 116

Table 3.3. Effect of azide on the content of Glut1 in the DRM and non-DRM domains of the plasma membrane.

DRM Non-DRM

Control 0.50 ± 0.07 0.46 ± 0.07

Azide 0.58 ± 0.12 1.03 ± 0.23*

Azide/Control 1.12 ± 0.11 2.87 ± 0.84**

In each experiment, the content of Glut1 was calculated utilizing the relative abundance of Glut1 in the PM of control (set at 1.0) and azide-treated cells, and the fractional distribution of Glut1 in the

DRM and non-DRM domains; experiments on control and azide-treated cells were performed in parallel (mean ± SE; n = 9). * denotes p < 0.05 by two-tailed, unpaired Student t-test. ** denotes p <

0.05 by Chi-squared test.

Chapter 4

The Effects of Methyl-β-Cyclodextrin on Glut1 and Glucose Transport

Introduction

The presence of glucose transporters in low density, detergent-resistant, membrane

microdomains (DRMs) has gained much attention over the past decade. Glut4 and Glut1,

but not Glut3, have been associated with these domains (Sakyo et al., 2002). Glut1

distribution to the DRMs has been reported for several different cell types including

HeLa cell - fibroblast hybrid cells, Clone 9 cells, human erythrocytes, 3T3-L1 fibroblasts,

and 3T3-L1 adipocytes (Kumar et al., 2004; Rubin et al., 2003; Rubin et al., 2004;

Scherer et al., 1994). Evidence of a functional consequence of transporter association

with the DRMs has also accumulated. It has been suggested, so far without dissent, that

these domains may be involved in vesicle trafficking, e.g. of glucose transporter -

containing vesicles (Chamberlain et al., 2002; Chiang et al., 2001; Ros-Baro et al., 2001;

Simons et al., 1988); however, the idea of glucose transporter regulation by translocation

between DRMs and non-DRM fractions is controversial. Some reports indicate that

movement from the bulk membrane into the DRM domain facilitates increased glucose

transport (Gustavsson et al., 1996; Kumar et al., 2004); while other reports suggest that the DRM domains may be inhibitory and translocation away from them facilitates increased glucose transport (Barnes et al., 2003; Rubin et al., 2003). In contrast to these studies, at least one report has been published that suggested the influence of cholesterol - rich lipid microdomains on the regulation of glucose transport is minimal as observed in

117 118

studies of human erythrocytes (Samuel et al., 2001). All of these findings were based on

studies of changes is glucose transporter association with the DRM in response to some stimulus or changes in cell glucose transport after exposure to a reagent expected to

disrupt the DRMs by perturbing membrane lipid content or arrangement.

The interaction between cholesterol and sphingolipids is believed to be an

important factor for the existence of lipid microdomains in cellular membranes (London,

2002; London et al., 2000). There are several reagents that are widely believed to perturb

these microdomains; these include lipid sequestering agents (e.g. filipin) and lipid

disrupting agents (e.g. the β - cyclodextrins, sphingomyelinase, and cholesterol oxidase)

(Allan et al., 1988; Al-Makdissy et al., 2003; Gustavsson et al., 1999; Liu et al., 2004;

Robinson et al., 1980; Röper et al., 2000; Smart et al., 1994; Yeagle, 1985). Of these

reagents, only the β-cyclodextrins will be considered further by the studies presented

here. This family of compounds seems to be relatively specific for cholesterol, since the

extraction of more abundant membrane lipids is reportedly low in certain cell types (Irie

et al., 1982; Kilsdonk et al., 1995; Yancey et al., 1996). Cellular toxicity also seems to be low given proper dosage and duration of exposure; however, there may be adverse effects for certain organelles even under these conditions (Hansen et al., 2000).

Previous work has shown that in post-nuclear lysates of Clone 9 cells, Glut1 association with the detergent-resistant membrane domain is decreased by azide exposure thus suggesting that these domains may be important for the regulation of glucose transport (Rubin et al., 2003). In order to further explore this hypothesis, the effect of cholesterol extraction on the increase in glucose transport induced by azide and various other inhibitors of oxidative phosphorylation, as well as on the localization of Glut1 to 119

the DRM in post-nuclear lysates, was examined. The following results are presented: i)

pre-treatment with methyl-β-cyclodextrin (mβcd) reversibly inhibits the ability of azide

to stimulate glucose transport, ii) pre-treatment with mβcd also inhibits the stimulation of glucose transport with other inhibitors of oxidative phosphorylation, iii) mβcd decreases glucose transporter association with the DRM. 120

Materials and Methods

Materials Cell culture reagents were purchased from Invitrogen (Carlsbad, CA),

plasticware and glassware through Fisher Scientific, and unless otherwise noted, all other

chemicals were from Sigma Aldrich or Fisher Scientific. 4 ml ultracentrifugation tubes

fitting the Sorvall TST 60.4 swinging bucket rotor were purchased from Kendro. 3-O-

methyl-D- glucose (3-OMG) was purchased from Amersham (Chicago, IL).

Antibodies were purchased from the following companies: mouse monoclonal anti-

caveolin-1 (BD Transduction Laboratories), rabbit anti-Glut 1 (Chemicon, Temecula,

CA), horseradish peroxidase-conjugated anti-rabbit and anti-mouse IgG secondary

antibodies (Sigma Aldrich). Nitrocellulose membrane was purchased from Schliecher &

Schuell (Keene, NH). Western blots were developed using the enhanced

chemiluminescence kit from Santa Cruz Biotechnology (Santa Cruz, CA).

Cell culture Clone 9 cells were maintained in DMEM supplemented with 10% bovine calf serum (v/v) and penicillin-streptomycin. Culture medium was replaced with fresh medium 16-24 hours prior to initiation of experiments.

Cell Treatment Cells were pre-treated with either diluent or 10 mM mβcd in serum – free

DMEM for 30 minutes. Then, cells were washed with Hanks buffered salt solution and

incubated in media supplemented with either diluent or an inhibitor of oxidative

phosphorylation at the indicated concentration: 5 mM azide (cytochrome c oxidase

inhibitor), 1 µM rotenone (complex I inhibitor), 2.5 µM carbonyl cyanide m- chlorophenyl-hydrazone (CCCP; uncoupler of mitochondrial respiration). Glucose 121

transport was previously known to be stimulated after 90 minutes exposure to each of the

above inhibitors at the concentrations given (Hamrahian et al., 1999).

Glucose Transport The glucose transport assay began when the treatment media was

removed and the uptake media applied (Hamrahian et al., 1999; Mercado et al., 1989).

Uptake media consists of 1 µCi/ml 3H-3-O-methylglucose (3-OMG) in serum-free

DMEM, and either 50 µM cytochalasin B (CB) in DMSO or an equal volume of diluent.

After one minute, the uptake media was removed and the cells were immediately washed

with stop solution (100 mM MgCl2 and 0.1 mM phloretin in deionized water). The difference between 3-OMG uptake in the presence of CB and diluent is equal to CB – inhibitable 3-OMG accumulation.

Fractionation by sucrose density step-gradient centrifugation Four 10 cm plates of confluent Clone 9 cells were washed 3 times with ice-cold PBS and scraped into 0.5 ml of ice-cold PBS. All subsequent steps were performed at 4 oC. Cell pellets were lysed in

410 µl Buffer A (0.5% Triton X-100, 50 mM Hepes, 150 mM NaCl, 5 mM EDTA, 5 mM

EGTA, 20 mM sodium pyrophosphate, 1 mM sodium orthovanadate, 20 mM sodium fluoride, and 0.1 µM PMSF; pH 7.4). 385 µl of the post-nuclear lysate was loaded into

the bottom of a 4-ml ultracentrifuge tube for the Sorvall TST 60.4 swinging bucket rotor,

adjusted to 40% sucrose with 60% sucrose in Buffer B (20 mM Tris, 150 mM NaCl, 1

mM EDTA, 0.1 µM PMSF; pH 7.4) to a total volume of 1 ml and then overlaid with 2 ml

35% sucrose and 1 ml 5% sucrose in Buffer B. Samples were centrifuged at an average 122

of 150 k x g for 17 hours at 4 oC. 500 µl fractions were collected from the top of the tube.

Aliquots of 30 µl from fractions 2 – 8 were further fractionated by SDS-PAGE, transferred to nitrocellulose membranes, blocked with 5% (w/v) powdered milk, and then incubated overnight with the appropriate primary antibodies. After three washes with

TBST (0.05% Tween-20 in Tris-buffered saline), membranes were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody, washed with TBST, and developed by enhanced chemiluminescence. The density of the appropriate bands was determined using the Bio-Rad Gel Doc 1000 system. The densities of these bands in fractions 2 and 3 were added and divided by the total density of the same bands in all the fractions.

Statistical Analysis All results are expressed as mean ± SE. Either chi squared analysis or the two-tailed, unpaired Student t-test was used and a P value of < 0.05 was considered significant. 123

Results

To examine the consequences of exposure to methyl-β-cyclodextrin on glucose transport stimulation by the inhibition of oxidative phosphorylation, Clone 9 cells were pre-treated with 10 mM methyl-β-cyclodextrin for 30 minutes, and then exposed to one of several inhibitors of the mitochondrial electron transport chain as described in Materials and

Methods. Pre-treating Clone 9 cells with mβcd resulted in a statistically non-significant

1.3 ± 0.12 - fold (mean ± SE, n = 16) (Figure 1) increase in glucose transport while exposure to 5 mM azide for 90 minutes increased glucose transport 3.3 ± 0.29 – fold

(mean ± SE, n = 8). Pre-treatment with mβcd inhibited the azide response to a statistically non-signficant 1.5 ± 0.11 – fold (mean ± SE, n = 10) increase; this inhibition was also observed at 5 mM mβcd (data not shown). Treatment with rotenone and carbonyl cyanide m-chlorophenyl-hydrazone (CCCP), also inhibitors of oxidative phosphorylation, significantly stimulated glucose transport, as expected (Table 1) and this effect was inhibited by pre-treatment with mβcd (Figure 2). The effect of mβcd was not permanent as cells recovered the azide response after removal of the reagent and incubation overnight in serum – free medium (Figure 3). Cells treated only with mβcd and allowed to recover had glucose transport no different than basal (0.95 ± 0.04 – fold

(mean ± SE, n = 6) change) while azide increased transport 3.1 ± 0.11 – fold (mean ± SE, n = 6).

Because the effect of mβcd could have indicated some decrease in the viability of a portion of cells, membrane integrity was assessed by Trypan Blue exclusion. There was no difference between cells treated with either diluent or mβcd at the exposure level 124

used for the experiments reported here although there was an increase in staining at and beyond one hour of exposure (data not shown).

As methyl-β-cyclodextrin is believed to disrupt cholesterol – rich membrane microdomains, its effect on Glut1 distribution to the DRMs and non-DRM domains of post-nuclear lysates was examined (Figure 4A). Exposure to 10 mM mβcd for 30 minutes appeared to decrease Glut1 association with the DRM from 48 ± 12 % to 21 ± 4

% (mean ± SE, n = 3) and increase Glut1 association with the non – DRM domain from

45 ± 13 % to 76 ± 5 % (mean ± SE, n = 3) (Figure 4B). Neither of these changes was

statistically different; however, the trend of the data set seems consistent (Table 2).

Oddly, in one experiment, the distribution of caveolin – 1 was unaffected (data not shown). 125

Discussion

Since the association between Glut1 and the detergent – resistant membrane domain in

post-nuclear lysates was altered by exposure to azide (Rubin et al., 2003), the effect of methyl-β-cyclodextrin (mβcd) on the stimulation of glucose transport by azide and other inhibitors of the electron transport chain (ETC) was tested. While it did not alter basal transport, mβcd inhibited the stimulation of glucose transport by three different inhibitors of oxidative phosphorylation; each of which inhibits the ETC by an independent mechanism (see Materials and Methods). The ability of mβcd to inhibit glucose transport response to each of these inhibitors suggests either that each has some membrane cholesterol – dependent mechanism for the stimulation of glucose transport in common or there was a general effect of mβcd that resulted in resistance to all the agents.

Two findings argue against the latter interpretation of these findings. First, basal transport does not seem to be significantly affected by mβcd alone. Second, membrane integrity, as measured by Trypan blue exclusion, was not grossly changed after exposure to the conditions employed in these studies. In addition to these findings, the ability of these cells to recover their ability to respond to azide after a sufficient period of time indicates that the cells can correct whatever changes were caused by exposure to mβcd even in the absence of serum.

It has recently been reported that treatment of Clone 9 cells with 10 mM mβcd for

60 minutes stimulated basal glucose transport to the same extent as a 60 minute exposure to 5 mM azide although the combination of the two was neither additive nor inhibitory

(Barnes et al., 2003). These authors did not report observing any signs of toxicity with their treatment protocol; in contrast to this report, our laboratory has found that treatment 126 with 10 mM mβcd longer than 30 minutes has deleterious effects on Clone 9 cell membrane integrity as indicated by increased permeability to Trypan blue as well as alterations in cell morphology, i.e. rounding and an increase in highly refractive cells

(data not shown).

In any case, Barnes and colleagues were able to demonstrate an association between the mβcd – induced increase of glucose transport and a decrease in the association of Glut1 with the DRM fraction of post-nuclear lysates. While the results presented in this chapter were not statistically consistent, they indicate a decrease in

Glut1 association with the DRM; however, this decrease was not associated with an increase in glucose transport. The reasons for the discrepancy between the findings of our lab and that of Barnes and colleagues are unclear. The different treatment conditions employed for the two studies could be contributory. Alternatively, it may be that there are real differences between the Clone 9 cells employed; however, such a difference has not been noted in the literature or the American Type Culture Collection on – line database.

Considering these studies (Barnes et al., 2003; Rubin et al., 2003) as well as those present here, we suggest that DRMs may have a permissive role in the mechanism of azide-induced stimulation of glucose transport. The inability to stimulate glucose transport after exposure to mβcd (or other conditions where an increase in the sphingolipids: cholesterol ratio is expected (Al-Makdissy et al., 2003)) suggests that the integrity of the DRM-enriched lipids are required for the stimulation of glucose transport.

Further, the azide – induced decrease in Glut1 association with the DRM in cellular membranes (Rubin et al., 2003) could reflect an activation, or disinhibition, of less active 127

Glut1 molecules in a couple of different ways. First, the DRM itself could be a

potentially inhibitory element due to its lipid composition and low fluidity (Carruthers et

al., 1984; Carruthers et al., 1988; Gidwani et al., 2001). Second, there could be an

inhibitory Glut1-interacting protein co-resident in the DRM, e.g. stomatin (Zhang et al.,

2001). Dissociation of Glut1 from either of these elements could allow the presentation

of a more functional transporter at the non-DRM domain of the plasma membrane in response to inhibitors of oxidative phosphorylation or other conditions that stimulate glucose transport. To wit, there is at least one paper which suggests that Glut1 association with the DRMs may increase glucose transport. This study showed an increase in the association of Glut1 with the DRM in response to extreme glucose deprivation (in the presence of fructose) a condition which is also known to increase glucose transport (Kumar et al., 2004). There is no clear explanation as to why these studies would yield results that contradict the interpretation of the work presented and discussed here.

As detergent-resistant membrane microdomains may have an important role in the regulation of glucose transport, more studies are required to understand the mechanisms whereby they contribute to transporter regulation. Investigating the mechanisms of the redistribution of glucose transporters between DRM and non-DRM domains in response to certain stimuli may provide interesting and important clues toward understanding the acute regulatory events which control transporter function. 128

Acknowledgements

We thank Donna Horvath and Bonnie Gorzelle for their able technical assistance. This study was supported in part by NIH grant R01 DK45945 to F. Ismail-Beigi. 129

Figure 4.1. Methyl-β-cyclodextrin inhibits azide – induced glucose transport. Cell

treatments and glucose transport were performed as described in the Materials and

Methods. * denotes p < 0.05 by Chi-squared test. 130

4.0 * 3.5

3.0 Serum-free DMEM wash

2.5 10mM mbCD, wash

2.0 wash, 5mM Azide

Fold Effect 1.5 10mM mbCD, wash, 1.0 5mM Azide

0.5

0.0

131

Table 4.1. The fold change of glucose transport after exposure of Clone 9 cells to

various inhibitors of oxidative phosphorylation is inhibited by pre-treatment with

methyl-β-cyclodextrin.

Inhibitor No pre-tx mβcd

None1 1.0 ± 0.16 1.3 ± 0.12

Azide 1 3.3 ± 0.29* 1.5 ± 0.11

Rotenone 3.2 ± 0.30* 2.1 ± 0.25

CCCP 3.4 ± 0.31* 1.9 ± 0.13

1 These results also reported in Figure 1. * denotes p < 0.05 relative to one by Chi- squared test. The number of transport assays for each condition is as follows: no inhibitor, 6; mβcd treatment (tx) alone, 16; azide, 8; mβcd pre-tx then azide, 10; rotenone,

6; mβcd pre-tx then rotenone, 6; CCCP, 3; mβcd pre-tx then CCCP, 3. 132

Figure 4.2. Methyl-β-cyclodextrin inhibits the stimulation of glucose transport by the inhibition of oxidative phosphorylation at ETC sites other than cytochrome c oxidase. Cell treatments and glucose transport were performed as described in the

Materials and Methods. The number of transport assays for each condition is detailed in

Table 1. * denotes p < 0.05 by Chi-squared test. 133

4.0 * * * 3.5

3.0

2.5 Inhibitor 2.0 mbcd, Inhibitor

Fold Effect 1.5

1.0

0.5

0.0 Azide Rotenone CCCP

134

Figure 4.3. Clone 9 cells treated with methyl-β-cyclodextrin can recover the azide- response. Cell treatments and glucose transport were performed as described in the

Materials and Methods. * denotes p < 0.05 by Chi-squared test.

135

4.5 * 4.0

3.5 * 3.0 Azide 2.5 mbcd, recovery 2.0 mbcd, recovery, azide Fold Effect 1.5

1.0

0.5

0.0

136

Figure 4.4. Methyl-β-cyclodextrin decreases Glut1 localization to the DRM. A)

Cells were treated and prepared as described in Materials and Methods. 30 µl samples of

sucrose gradient fractions 2 – 8 were subjected to SDS – PAGE and transferred to

nitrocellulose whereupon immunoblots using anti-Glut1 antibody were performed and

analyzed as described. B) A graphical summary of the data regarding the fractional

association of Glut1 with the DRM (Figure 4A) and non-DRM domains gathered from three independent experiments as described. The average percentages of Glut1 association with either the DRM or non –DRM domains were not statistically different between control and mβcd treated cells as determined by the two-tailed, unpaired Student

t-test. 137

A Control 30 min 10mM mbCD 2 3 4 5 6 7 8 2 3 4 5 6 7 8 Fraction

Anti-Glut1 138

0.90 0.80 0.70

0.60

0.50 Control 0.40 Azide 0.30 Fraction of Glut1 of Fraction 0.20 0.10 0.00 DRM Non - DRM

139

Table 2. Methyl-β-cyclodextrin decreases the percentage of Glut1 associated with the DRM domain.

Experiment Control mβcd

1 39 26

2 29 27

3 77 10

Mean ± SE 48 ± 12 21 ± 4

The average percentages of Glut1 association with the DRM between control and mβcd were not statistically significant by the two-tailed, unpaired

Student t-test.

Chapter 5

Summary and Future studies

Summary

This work has demonstrated an association between Glut1 and the detergent – resistant membrane (DRM) domains present in Clone 9 cells. This association may facilitate the acute regulation of Glut1 function as it was shown that exposing these cells to azide significantly decreased Glut1 association with the DRM by 40 % in post – nuclear lysate while increasing glucose transport by 3 – 4 fold. In the plasma membrane compartment of Clone 9 cells, azide did not decrease the association of Glut1 with the DRM; however, the abundance of Glut1 in the non – DRM fraction was increased significantly. Finally, methyl-β-cyclodextrin could inhibit the increase in glucose transport stimulated not only

by azide but also rotenone and CCCP each of which inhibits oxidative phosphorylation

by different mechanisms. Taken together, these findings suggest that the association

between Glut1 and the DRMs is a part of the mechanism that manages the acute control

of glucose transport in Clone 9 cells. Further studies will be required to substantiate the

findings and interpretation of this work. Toward that end, the following passages

describe studies that may support the findings here and provide more clues toward

understanding the mechanisms of the azide – induced stimulation of glucose transport.

140 141

Future Studies

Verification of Glut1 Association with Membrane Microdomains

The verification of Glut1 association with plasma membrane microdomains by another

independent measure is a prudent choice for future studies. Immunoelectron microscopy

of plasma membrane sheets has been used by several groups, along with varying degrees of statistical analysis, as evidence of discrete microdomains at the cell surface (Prior et al., 2003a; Prior et al., 2003b; Wilson et al., 2000). The clustering of Glut1 can be assessed by the distribution of anti – Glut1 antibody as revealed with a colloidal gold

labeled secondary antibody.

Further Analysis of Azide – Induced Changes in Glut1 Compartmentalization

The membrane compartment from which there is an azide-induced decrease of Glut1 in

the DRMs of post-nuclear lysates has yet to be identified. As suggested in the preceding

text, our inability to observe that compartment in our studies of the post-nuclear

homogenate may be because it is associated with the discarded nuclear fraction. Studies

to approach the identity of this compartment should include plasma membrane

preparations isolated from total cell membranes or whole homogenate so as not exclude

any membrane populations. In addition to studies of azide, the effect of methyl-β- cyclodextrin on Glut1 compartmentalization may also be informative in this context. In the event the compartment of the azide – induced decrease of Glut1 in the DRM remains unobserved, a more classical approach to the fractionation (McKeel et al., 1970; Simpson et al., 1983) could be employed in order to assess changes that may be masked by our adaptation of these methods. 142

An assessment of glucose transporter activity in the DRM and non-DRM

compartments of the plasma membrane is conspicuously absent from the studies

presented here. Given that others in the laboratory have been unable to reliably

recapitulate glucose transport in vitro, a surrogate measure of transporter activation can

be obtained using those membrane-impermeable, glucose transporter labeling reagents

that have been suggested to preferentially label the more active transporters e.g. ATB-

BMPA or IAPS-forskolin. Plasma membranes fractions could be prepared from cells

labeled with tritiated ATB-BMPA and the distribution of radioactivity assessed by a

number of techniques. In the first instance, autoradiograms of electrophoresis gels, either

dried or transferred to nitrocellulose, may provide an image of the labeled transporters.

Transfer to nitrocellulose has the advantage that one could compare the fractional

distribution of radiolabeled transporters to that of total transporters as detected by

immunoblot. A second technique would be to use scintillation to measure the

radioactivity in slices of gels through which samples from labeled cells have been

electrophoresed.

Investigation of the Mechanism Whereby Glut1 associates with the DRM

While we, and others, have established and confirmed the association of Glut1 with low

density detergent-resistant membrane domains in the plasma membrane of Clone 9 cells,

the mechanism of this association remains a mystery (Kumar et al., 2004; Rubin et al.,

2003; Sakyo et al., 2002) although there are several interesting possibilities. Given that

the palmitoylation of the transporter has been reported (Pouliot et al., 1995) and that Cys-

209 of Glut1 is membrane proximal and thus a good candidate for acylation (Charrin et 143

al., 2002; de Vetten et al., 1988; Hartel-Schenk et al., 1992), the role of this residue in

Glut1 association with the DRMs should be examined. There are several techniques

which may be helpful. First, metabolic labeling with tritiated fatty acids and

immunoprecipitation could be used to assess acylation of Glut1 (Melkonian et al., 1999).

Second, conservative and non-conservative mutations of Cys-209 would allow a direct

assessment of the role of this residue for association with the DRMs given proper folding

and trafficking of the mutant protein. If the mutation is applied in the context of the

extracellular loop – Myc tagged Glut1, the Myc tag would allow the assessment of

surface expression of the mutant transporter by several techniques including microscopy,

flow cytometry, and ELISA. Finally, labeling of reactive cysteine residues with biotin –

cysteine (Humphries et al., 2002) could indicate the accessibility of those residues in wildtype Glut1 both in the basal condition and in the presence of azide.

Discovery of Glut1 – interacting proteins

As the increase of Glut1 abundance in the non-DRM domains of the plasma membrane fraction does not completely account for the azide-induced increase in glucose transport

we have observed (Rubin et al., 2003), other mechanisms may be contributing to this

phenomenon. The discovery of additional Glut1 – interacting proteins (Lachaal et al.,

1990; Oliver et al., 1996; Zhang et al., 1999b) could provide important clues to these

mechanisms. Glut1 – associated proteins could be thoroughly evaluated and assessed by

combining subcellular fractionation, magnetic separation – affinity isolation (Howell et

al., 1989; Pasquali et al., 1999), gel electrophoresis, and mass spectrometry. A rigorous

analysis of protein – protein interactions would involve 2D – gel analysis of 144

immunoprecipitates obtained under a variety of preparation conditions including different

detergent types and concentrations, high and low salt concentrations, and the presence or

absence of divalent cations. It is very clear that in the case of membrane proteins

changing these parameters can drastically affect not only the results of the

immunoprecipitation but also of the electrophoresis (Charrin et al., 2002; Pasquali et al.,

1997; Santoni et al., 2000).

Works Cited

Abbud, W., Habinowski, S., Zhang, J. Z., Kendrew, J., Elkairi, F. S., Kemp, B. E., Witters, L. A., and Ismail-Beigi, F. (2000). Stimulation of AMP-activated protein kinase (AMPK) is associated with enhancement of Glut1-mediated glucose transport. Arch Biochem Biophys 380, 347-352.

Allan, D., and Walklin, C. M. (1988). Endovesiculation of human erythrocytes exposed to sphingomyelinase C: a possible explanation for the enzyme-resistant pool of sphingomyelin. Biochim Biophys Acta 938, 403-410.

Al-Makdissy, N., Younsi, M., Pierre, S., Ziegler, O., and Donner, M. (2003). Sphingomyelin/cholesterol ratio: an important determinant of glucose transport mediated by GLUT-1 in 3T3-L1 preadipocytes. Cell Signal 15, 1019-1030.

Arcaro, A., Gregoire, C., Bakker, T. R., Baldi, L., Jordan, M., Goffin, L., Boucheron, N., Wurm, F., van der Merwe, P. A., Malissen, B., and Luescher, I. F. (2001). CD8beta endows CD8 with efficient coreceptor function by coupling T cell receptor/CD3 to raft- associated CD8/p56(lck) complexes. J Exp Med 194, 1485-1495.

Arcaro, A., Gregoire, C., Boucheron, N., Stotz, S., Palmer, E., Malissen, B., and Luescher, I. F. (2000). Essential role of CD8 palmitoylation in CD8 coreceptor function. J Immunol 165, 2068-2076.

Baker, L., and Storm, D. (1997). Dynamic palmitoylation of neuromodulin (GAP-43) in cultured rat cerebellar neurons and mouse N1E-115 cells. Neurosci Lett 234, 156-160.

Baly, D. L., and Horuk, R. (1987). Dissociation of insulin-stimulated glucose transport from the translocation of glucose carriers in rat adipose cells. J Biol Chem 262, 21-24.

Barnes, K., Ingram, J. C., Bennett, M. D., Stewart, G. W., and Baldwin, S. A. (2003). Methyl-beta-cyclodextrin stimulates glucose uptake in Clone 9 cells: a possible role for lipid rafts. Biochem J Pt.

Barnes, K., Ingram, J. C., Porras, O. H., Barros, L. F., Hudson, E. R., Fryer, L. G. D., Foufelle, F., Carling, D., Hardie, D. G., and Baldwin, S. A. (2002). Activation of GLUT1 by metabolic and osmotic stress: potential involvement of AMP-activated protein kinase (AMPK). J Cell Science 115, 2433-2442.

Barros, L. F., Baldwin, S. A., and Griffiths, M. (1997a). Rapid activation of GLUT1 by osmotic stress. Biochem Soc Trans 25, 485S.

Barros, L. F., Barnes, K., Ingram, J. C., Castro, J., Porras, O. H., and Baldwin, S. A. (2001). Hyperosmotic shock induces both activation and translocation of glucose transporters in mammalian cells. Pflugers Arch 442, 614-621.

145 146

Barros, L. F., Marchant, R. B., and Baldwin, S. A. (1995). Dissection of stress-activated glucose transport from insulin-induced glucose transport in mammalian cells using wortmannin and ML-9. Biochem J 309, 731-736.

Barros, L. F., Young, M., Saklatvala, J., and Baldwin, S. A. (1997b). Evidence of two mechanisms for the activation of the glucose transporter GLUT1 by anisomycin: p38(MAP kinase) activation and protein synthesis inhibition in mammalian cells. J Physiol 504, 517-525.

Bashan, N., Burdett, E., Guma, A., Sargeant, R., Tumiati, L., Liu, Z., and Klip, A. (1993). Mechanisms of adaptation of glucose transporters to changes in the oxidative chain of muscle and fat cells. Am J Physiol 264, C430-440.

Bashan, N., Burdett, E., Hundal, H. S., and Klip, A. (1992). Regulation of glucose transport and GLUT1 glucose transporter expression by O2 in muscle cells in culture. Am J Physiol 262, C682-690.

Baumann, C. A., Ribon, V., Kanzaki, M., Thurmond, D. C., Mora, S., Shigematsu, S., Bickel, P. E., Pessin, J. E., and Saltiel, A. R. (2000). CAP defines a second signalling pathway required for insulin-stimulated glucose transport. Nature 407, 202-207.

Behrooz, A., and Ismail-Beigi, F. (1997). Dual control of glut1 glucose transporter gene expression by hypoxia and by inhibition of oxidative phosphorylation. J Biol Chem 272, 5555-5562.

Behrooz, A., and Ismail-Beigi, F. (1998). Induction of GLUT1 mRNA in response to azide and inhibition of protein synthesis. Mol Cell Biochem 187, 33-40.

Bickel, P. E. (2002). Lipid rafts and insulin signaling. Am J Physiol Endocrinol Metab 282, E1-E10.

Bilan, P. J., Mitsumoto, Y., Maher, F., Simpson, I. A., and Klip, A. (1992a). Detection of the GLUT3 facilitative glucose transporter in rat L6 muscle cells: regulation by cellular differentiation, insulin and insulin-like growth factor-I. Biochem Biophys Res Commun 186, 1129-1137.

Bilan, P. J., Mitsumoto, Y., Ramlal, T., and Klip, A. (1992b). Acute and long-term effects of insulin-like growth factor I on glucose transporters in muscle cells. Translocation and biosynthesis. FEBS Lett 298, 285-290.

Bilan, P. J., Ramlal, T., and Klip, A. (1991). IGF-I mediated recruitment of glucose transporters from intracellular membranes to plasma membranes in L6 muscle cells. Adv Exp Med Biol 293, 273-288.

Bjornholt, J. V., Erikssen, G., Liestol, K., Jervell, J., Erikssen, J., and Thaulow, E. (2001). Prediction of Type 2 diabetes in healthy middle-aged men with special emphasis on glucose homeostasis. Results from 22.5 years' follow-up. Diabet Med 18, 261-267. 147

Boone, C. W., Ford, L. E., Bond, H. E., Stuart, D. C., and Lorenz, D. (1969). Isolation of plasma membrane fragments from HeLa cells. J Cell Biol 41, 378-392.

Brown, D. A., and London, E. (2000). Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J Biol Chem 275, 17221-17224.

Brown, D. A., and Rose, J. K. (1992). Sorting of GPI-anchored proteins to glycolipid- enriched membrane subdomains during transport to the apical cell surface. Cell 68, 533- 544.

Brown, S., Georgatos, M., Reifel, C., Song, J. H., Shin, S. H., and Hong, M. (2002). Recycling processes of cellular ascorbate generate oxidative stress in pancreatic tissues in in vitro system. Endocrine 18, 91-96.

Buck, J., Sinclair, M. L., Schapal, L., Cann, M. J., and Levin, L. R. (1999). Cytosolic adenylyl cyclase defines a unique signaling molecule in mammals. Proc Natl Acad Sci U S A 96, 79-84.

Busch, G., Hoder, D., Reutter, W., and Tauber, R. (1989). Selective isolation of individual cell surface proteins from tissue culture cells by a cleavable biotin label. Eur J Cell Biol 50, 257-262.

Carayannopoulos, M. O., Chi, M. M., Cui, Y., Pingsterhaus, J. M., McKnight, R. A., Mueckler, M., Devaskar, S. U., and Moley, K. H. (2000). GLUT8 is a glucose transporter responsible for insulin-stimulated glucose uptake in the blastocyst. Proc Natl Acad Sci U S A 97, 7313-7318.

Carruthers, A., and Melchior, D. L. (1984). Human erythrocyte hexose transporter activity is governed by bilayer lipid composition in reconstituted vesicles. Biochemistry 23, 6901-6911.

Carruthers, A., and Melchior, D. L. (1988). Effects of lipid environment on membrane transport: the human erythrocyte sugar transport protein/lipid bilayer system. Annu Rev Physiol 50, 257-271.

Carter-Su, C., Pessin, J. E., Mora, R., Gitomer, W., and Czech, M. P. (1982). Photoaffinity labeling of the human erythrocyte D-glucose transporter. J Biol Chem 257, 5419-5425.

Chamberlain, L. H., and Gould, G. W. (2002). The vesicle- and target-SNARE proteins that mediate Glut4 vesicle fusion are localized in detergent-insoluble lipid rafts present on distinct intracellular membranes. J Biol Chem 277, 49750-49754.

Charrin, S., Manie, S., Oualid, M., Billard, M., Boucheix, C., and Rubinstein, E. (2002). Differential stability of tetraspanin/tetraspanin interactions: role of palmitoylation. FEBS Lett 516, 139-144. 148

Chauveau, J., Moule, Y., Rouiller, C., and Schneebeli, J. (1962). Isolation of smooth vesicles and free ribosomes from rat liver microsomes. J Cell Biol 12, 17-29.

Chiang, S. H., Baumann, C. A., Kanzaki, M., Thurmond, D. C., Watson, R. T., Neudauer, C. L., Macara, I. G., Pessin, J. E., and Saltiel, A. R. (2001). Insulin-stimulated GLUT4 translocation requires the CAP-dependent activation of TC10. Nature 410, 944-948.

Claas, C., Stipp, C. S., and Hemler, M. E. (2001). Evaluation of prototype transmembrane 4 superfamily protein complexes and their relation to lipid rafts. J Biol Chem 276, 7974- 7984.

Clancy, B. M., Harrison, S. A., Buxton, J. M., and Czech, M. P. (1991). Protein synthesis inhibitors activate glucose transport without increasing plasma membrane glucose transporters in 3T3-L1 adipocytes. J Biol Chem 266, 10122-10130.

Clark, A. E., Holman, G. D., and Kozka, I. J. (1991). Determination of the rates of appearance and loss of glucose transporters at the cell surface of rat adipose cells. Biochem J 278, 235-241.

Cloherty, E. K., Levine, K. B., and Carruthers, A. (2001). The red blood cell glucose transporter presents multiple, nucleotide- sensitive sugar exit sites. Biochemistry 40, 15549-15561.

Concha, II, Velasquez, F. V., Martinez, J. M., Angulo, C., Droppelmann, A., Reyes, A. M., Slebe, J. C., Vera, J. C., and Golde, D. W. (1997). Human erythrocytes express GLUT5 and transport fructose. Blood 89, 4190-4195.

Connolly, T. J., Carruthers, A., and Melchior, D. L. (1985). Effect of bilayer cholesterol content on reconstituted human erythrocyte sugar transporter activity. J Biol Chem 260, 2617-2620.

Cushman, S. W., and Wardzala, L. J. (1980). Potential mechanism of insulin action on glucose transport in the isolated rat adipose cell. Apparent translocation of intracellular transport systems to the plasma membrane. J Biol Chem 255, 4758-4762.

Czech, M. P. (1976). Characterization of (3H)cytochalasin B binding to the fat cell plasma membrane. J Biol Chem 251, 2905-2910.

Davies, S. P., Reddy, H., Caivano, M., and Cohen, P. (2000). Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351, 95-105.

Davoren, P. R., and Sutherland, E. W. (1963). The Cellular Location of Adenyl Cyclase in the Pigeon Erythrocyte. J Biol Chem 238, 3016-3023.

Dawidowicz, E. A. (1987). Dynamics of membrane lipid metabolism and turnover. Annu Rev Biochem 56, 43-61. 149

Dawson, P. A., Mychaleckyj, J. C., Fossey, S. C., Mihic, S. J., Craddock, A. L., and Bowden, D. W. (2001). Sequence and functional analysis of glut10: a glucose transporter in the type 2 diabetes-linked region of chromosome 20q12-13.1. Mol Genet Metab 74, 186-199. de Vetten, M., and P, A. (1988). The Rh polypeptide is a major fatty acid-acylated erythrocyte membrane protein. J Biol Chem 263, 18193-18196.

Desjardins, M. (2003). ER-mediated phagocytosis: a new membrane for new functions. Nat Rev Immunol 3, 280-291.

Desneves, J., Berman, A., Dynon, K., La Greca, N., Foley, M., and Tilley, L. (1996). Human erythrocyte band 7.2b is preferentially labeled by a photoreactive phospholipid. Biochem Biophys Res Commun 224, 108-114.

Diamond, D. L., and Carruthers, A. (1993). Metabolic control of sugar transport by derepression of cell surface glucose transporters. An insulin-independent recruitment- independent mechanism of regulation. J Biol Chem 268, 6437-6444.

Doege, H., Bocianski, A., Joost, H. G., and Schurmann, A. (2000a). Activity and genomic organization of human glucose transporter 9 (GLUT9), a novel member of the family of sugar-transport facilitators predominantly expressed in brain and leucocytes. Biochem J 350 Pt 3, 771-776.

Doege, H., Bocianski, A., Scheepers, A., Axer, H., Eckel, J., Joost, H. G., and Schurmann, A. (2001). Characterization of human glucose transporter (GLUT) 11 (encoded by SLC2A11), a novel sugar-transport facilitator specifically expressed in heart and skeletal muscle. Biochem J 359, 443-449.

Doege, H., Schurmann, A., Bahrenberg, G., Brauers, A., and Joost, H. G. (2000b). GLUT8, a novel member of the sugar transport facilitator family with glucose transport activity. J Biol Chem 275, 16275-16280.

Dunphy, J. T., Greentree, W. K., and Linder, M. E. (2001). Enrichment of G-protein Palmitoyltransferase Activity in Low Density Membranes. IN VITRO RECONSTITUTION OF Galpha i TO THESE DOMAINS REQUIRES PALMITOYLTRANSFERASE ACTIVITY. J Biol Chem 276, 43300-43304.

Ellgaard, L., and Helenius, A. (2003). Quality control in the endoplasmic reticulum. Nat Rev Mol Cell Biol 4, 181-191.

Fielding, C. J., and Fielding, P. E. (2000). Cholesterol and caveolae: structural and functional relationships. Biochim Biophys Acta 1529, 210-222.

Fisher, M. D., and Frost, S. C. (1996). Translocation of GLUT1 does not account for elevated glucose transport in glucose-deprived 3T3-L1 adipocytes. J Biol Chem 271, 11806-11809. 150

Furtado, L. M., Somwar, R., Sweeney, G., Niu, W., and Klip, A. (2002). Activation of the glucose transporter GLUT4 by insulin. Biochem Cell Biol 80, 569-578.

Gagnon, E., Duclos, S., Rondeau, C., Chevet, E., Cameron, P. H., Steele-Mortimer, O., Paiement, J., Bergeron, J. J., and Desjardins, M. (2002). Endoplasmic reticulum-mediated phagocytosis is a mechanism of entry into macrophages. Cell 110, 119-131.

Garavito, R. M., and Ferguson-Miller, S. (2001). Detergents as tools in membrane biochemistry. J Biol Chem 276, 32403-32406.

Garin, J., Diez, R., Kieffer, S., Dermine, J. F., Duclos, S., Gagnon, E., Sadoul, R., Rondeau, C., and Desjardins, M. (2001). The phagosome proteome: insight into phagosome functions. J Cell Biol 152, 165-180.

Gibbs, E. M., Lienhard, G. E., and Gould, G. W. (1988). Insulin-induced translocation of glucose transporters to the plasma membrane precedes full stimulation of hexose transport. Biochemistry 27, 6681-6685.

Gidwani, A., Holowka, D., and Baird, B. (2001). Fluorescence anisotropy measurements of lipid order in plasma membranes and lipid rafts from RBL-2H3 mast cells. Biochemistry 40, 12422-12429.

Gkantiragas, I., Brügger, B., Stüven, E., Kaloyanova, D., Li, X.-Y., Löhr, K., Lottspeich, F., Wieland, F. T., and Helms, J. B. (2001). Sphingomyelin-enriched Microdomains at the Golgi Complex. Mol Biol Cell 12, 1819-1833.

Gould, G. W., Cuenda, A., Thomson, F. J., and Cohen, P. (1995). The activation of distinct mitogen-activated protein kinase cascades is required for the stimulation of 2- deoxyglucose uptake by interleukin-1 and insulin-like growth factor-1 in KB cells. Biochem J 311 ( Pt 3), 735-738.

Green, A. (1983a). Catecholamines inhibit insulin-stimulated glucose transport in adipocytes, in the presence of adenosine deaminase. FEBS Lett 152, 261-264.

Green, A. (1983b). Glucagon inhibition of insulin-stimulated 2-deoxyglucose uptake by rat adipocytes in the presence of adenosine deaminase. Biochem J 212, 189-195.

Grune, T., Klotz, L. O., Gieche, J., Rudeck, M., and Sies, H. (2001). Protein oxidation and proteolysis by the nonradical oxidants singlet oxygen or peroxynitrite. Free Radic Biol Med 30, 1243-1253.

Grune, T., Reinheckel, T., Joshi, M., and Davies, K. J. (1995). Proteolysis in cultured liver epithelial cells during oxidative stress. Role of the multicatalytic proteinase complex, proteasome. J Biol Chem 270, 2344-2351.

Grune, T., Reinheckel, T., Li, R., North, J. A., and Davies, K. J. (2002a). Proteasome- dependent turnover of protein disulfide isomerase in oxidatively stressed cells. Arch Biochem Biophys 397, 407-413. 151

Grune, T., Reinheckel, T., North, J. A., Li, R., Bescos, P. B., Shringarpure, R., and Davies, K. J. (2002b). Ezrin turnover and cell shape changes catalyzed by proteasome in oxidatively stressed cells. Faseb J 16, 1602-1610.

Gustavsson, J., Parpal, S., Karlsson, M., Ramsing, C., Thorn, H., Borg, M., Lindroth, M., Peterson, K. H., Magnusson, K. E., and Stralfors, P. (1999). Localization of the insulin receptor in caveolae of adipocyte plasma membrane. Faseb J 13, 1961-1971.

Gustavsson, J., Parpal, S., and Stralfors, P. (1996). Insulin-stimulated glucose uptake involves the transition of glucose transporters to a caveolae-rich fraction within the plasma membrane: implications for type II diabetes. Mol Med 2, 367-372.

Hamill, S., Cloherty, E. K., and Carruthers, A. (1999). The human erythrocyte sugar transporter presents two sugar import sites. Biochemistry 38, 16974-16983.

Hamrahian, A. H., Zhang, J. Z., Elkhairi, F. S., Prasad, R., and Ismail-Beigi, F. (1999). Activation of Glut1 glucose transporter in response to inhibition of oxidative phosphorylation. Arch Biochem Biophys 368, 375-379.

Hansen, G. H., Niels-Christiansen, L. L., Thorsen, E., Immerdal, L., and Danielsen, E. M. (2000). Cholesterol depletion of enterocytes. Effect on the Golgi complex and apical membrane trafficking. J Biol Chem 275, 5136-5142.

Harder, T., Scheiffele, P., Verkade, P., and Simons, K. (1998). Lipid domain structure of the plasma membrane revealed by patching of membrane components. J Cell Biol 141, 929-942.

Hare, J. F., and Lee, E. (1989). Metabolic behavior of cell surface biotinylated proteins. Biochemistry 28, 574-580.

Harrison, S. A., Buxton, J. M., Clancy, B. M., and Czech, M. P. (1991). Evidence that erythroid-type glucose transporter intrinsic activity is modulated by cadmium treatment of mouse 3T3-L1 cells. J Biol Chem 266, 19438-19449.

Harrison, S. A., Buxton, J. M., Helgerson, A. L., MacDonald, R. G., Chlapowski, F. J., Carruthers, A., and Czech, M. P. (1990). Insulin action on activity and cell surface disposition of human HepG2 glucose transporters expressed in Chinese hamster ovary cells. J Biol Chem 265, 5793-5801.

Harrison, S. A., Clancy, B. M., Pessino, A., and Czech, M. P. (1992). Activation of cell surface glucose transporters measured by photoaffinity labeling of insulin-sensitive 3T3- L1 adipocytes. J Biol Chem 267, 3783-3788.

Hartel-Schenk, S., and Agre, P. (1992). Mammalian red cell membrane Rh polypeptides are selectively palmitoylated subunits of a macromolecular complex. J Biol Chem 267, 5569-5574. 152

Hayashi, T., Hirshman, M. F., Fujii, N., Habinowski, S. A., Witters, L. A., and Goodyear, L. J. (2000). Metabolic stress and altered glucose transport: activation of AMP-activated protein kinase as a unifying coupling mechanism. Diabetes 49, 527-531.

Hiebl-Dirschmied, C. M., Adolf, G. R., and Prohaska, R. (1991). Isolation and partial characterization of the human erythrocyte band 7 integral membrane protein. Biochim Biophys Acta 1065, 195-202.

Hoessli, D. C., Ilangumaran, S., Soltermann, A., Robinson, P. J., Borisch, B., and Nasir Ud, D. (2000). Signaling through sphingolipid microdomains of the plasma membrane: the concept of signaling platform. Glycoconj J 17, 191-197.

Holman, G. D., Karim, A. R., and Karim, B. (1988). Photolabeling of erythrocyte and adipocyte hexose transporters using a benzophenone derivative of bis(D-mannose). Biochim Biophys Acta 946, 75-84.

Holman, G. D., Kozka, I. J., Clark, A. E., Flower, C. J., Saltis, J., Habberfield, A. D., Simpson, I. A., and Cushman, S. W. (1990). Cell surface labeling of glucose transporter isoform GLUT4 by bis-mannose photolabel. Correlation with stimulation of glucose transport in rat adipose cells by insulin and phorbol ester. J Biol Chem 265, 18172- 18179.

Holman, G. D., Parkar, B. A., and Midgley, P. J. (1986). Exofacial photoaffinity labelling of the human erythrocyte sugar transporter. Biochim Biophys Acta 855, 115-126.

Holman, G. D., and Rees, W. D. (1982). Side-specific analogues for the rat adipocyte sugar transport system. Biochim Biophys Acta 685, 78-86.

Holter, H., Ottesen, M., and Weber, R. (1953). Separation of cytoplasmic particles by centrifugation in a density-gradient. Experientia 9, 346-348.

Hooper, N. M. (1999). Detergent-insoluble glycosphingolipid/cholesterol-rich membrane domains, lipid rafts and caveolae (review). Mol Membr Biol 16, 145-156.

Howell, K. E., Schmid, R., Ugelstad, J., and Gruenberg, J. (1989). Immunoisolation using magnetic solid supports: subcellular fractionation for cell-free functional studies. Methods Cell Biol 31, 265-292.

Hruz, P. W., and Mueckler, M. M. (2001). Structural analysis of the GLUT1 facilitative glucose transporter (review). Mol Membr Biol 18, 183-193.

Humphries, K. M., Juliano, C., and Taylor, S. S. (2002). Regulation of cAMP-dependent protein kinase activity by glutathionylation. J Biol Chem 277, 43505-43511.

Irie, T., Otagiri, M., Sunada, M., Uekama, K., Ohtani, Y., Yamada, Y., and Sugiyama, Y. (1982). Cyclodextrin-induced hemolysis and shape changes of human erythrocytes in vitro. J Pharmacobiodyn 5, 741-744. 153

Ismail-Beigi, F. (1993). Metabolic Regulation of Glucose Transport. J Membrane Biol 135, 1-10.

Joost, H. G., and Thorens, B. (2001). The extended GLUT-family of sugar/polyol transport facilitators: nomenclature, sequence characteristics, and potential function of its novel members (review). Mol Membr Biol 18, 247-256.

Joost, H. G., Weber, T. M., and Cushman, S. W. (1988). Qualitative and quantitative comparison of glucose transport activity and glucose transporter concentration in plasma membranes from basal and insulin-stimulated rat adipose cells. Biochem J 249, 155-161.

Joost, H. G., Weber, T. M., Cushman, S. W., and Simpson, I. A. (1986). Insulin- stimulated glucose transport in rat adipose cells. Modulation of transporter intrinsic activity by isoproterenol and adenosine. J Biol Chem 261, 10033-10036.

Joost, H. G., Weber, T. M., Cushman, S. W., and Simpson, I. A. (1987). Activity and phosphorylation state of glucose transporters in plasma membranes from insulin-, isoproterenol-, and phorbol ester-treated rat adipose cells. J Biol Chem 262, 11261- 11267.

Jung, C. Y., and Rampal, A. L. (1977). Cytochalasin B binding sites and glucose transport carrier in human erythrocyte ghosts. J Biol Chem 252, 5456-5463.

Kaetzel, C. S., Rao, C. K., and Lamm, M. E. (1987). Protein disulphide-isomerase from human placenta and rat liver. Purification and immunological characterization with monoclonal antibodies. Biochem J 241, 39-47.

Kahn, B. B., and Cushman, S. W. (1987). Mechanism for markedly hyperresponsive insulin-stimulated glucose transport activity in adipose cells from insulin-treated streptozotocin diabetic rats. Evidence for increased glucose transporter intrinsic activity. J Biol Chem 262, 5118-5124.

Kanai, F., Nishioka, Y., Hayashi, H., Kamohara, S., Todaka, M., and Ebina, Y. (1993). Direct demonstration of insulin-induced GLUT4 translocation to the surface of intact cells by insertion of a c-myc epitope into an exofacial GLUT4 domain. J Biol Chem 268, 14523-14526.

Kandror, K. V. (2003). A long search for Glut4 activation. Sci STKE 2003, PE5.

Karlsson, M., Thorn, H., Parpal, S., Stralfors, P., and Gustavsson, J. (2002). Insulin induces translocation of glucose transporter GLUT4 to plasma membrane caveolae in adipocytes. Faseb J 16, 249-251.

Karnieli, E., Zarnowski, M. J., Hissin, P. J., Simpson, I. A., Salans, L. B., and Cushman, S. W. (1981). Insulin-stimulated translocation of glucose transport systems in the isolated rat adipose cell. Time course, reversal, insulin concentration dependency, and relationship to glucose transport activity. J Biol Chem 256, 4772-4777. 154

Kayano, T., Burant, C. F., Fukumoto, H., Gould, G. W., Fan, Y. S., Eddy, R. L., Byers, M. G., Shows, T. B., Seino, S., and Bell, G. I. (1990). Human facilitative glucose transporters. Isolation, functional characterization, and gene localization of cDNAs encoding an isoform (GLUT5) expressed in small intestine, kidney, muscle, and adipose tissue and an unusual glucose transporter pseudogene-like sequence (GLUT6). J Biol Chem 265, 13276-13282.

Kenworthy, A. K., Petranova, N., and Edidin, M. (2000). High-resolution FRET microscopy of cholera toxin B-subunit and GPI- anchored proteins in cell plasma membranes. Mol Biol Cell 11, 1645-1655.

Kilsdonk, E. P., Yancey, P. G., Stoudt, G. W., Bangerter, F. W., Johnson, W. J., Phillips, M. C., and Rothblat, G. H. (1995). Cellular cholesterol efflux mediated by cyclodextrins. J Biol Chem 270, 17250-17256.

Klip, A., Ramlal, T., Bilan, P. J., Cartee, G. D., Gulve, E. A., and Holloszy, J. O. (1990). Recruitment of GLUT-4 glucose transporters by insulin in diabetic rat skeletal muscle. Biochem Biophys Res Commun 172, 728-736.

Koseoglu, M. H., and Beigi, F. I. (1999). Mechanism of stimulation of glucose transport in response to inhibition of oxidative phosphorylation: analysis with myc-tagged Glut1. Mol Cell Biochem 194, 109-116.

Kumar, A., Xiao, Y. P., Laipis, P. J., Fletcher, B. S., and Frost, S. C. (2004). Glucose deprivation enhances targeting of GLUT1 to lipid rafts in 3T3-L1 adipocytes. Am J Physiol Endocrinol Metab 286, E568-576.

Kuroda, M., Honnor, R. C., Cushman, S. W., Londos, C., and Simpson, I. A. (1987). Regulation of insulin-stimulated glucose transport in the isolated rat adipocyte. cAMP- independent effects of lipolytic and antilipolytic agents. J Biol Chem 262, 245-253.

Kurokawa, T., Yoshida, Y., Kawahara, K., Tsuchida, T., Okazawa, H., Fujibayashi, Y., Yonekura, Y., and Kotsuji, F. (2004). Expression of GLUT-1 glucose transfer, cellular proliferation activity and grade of tumor correlate with [F-18]-fluorodeoxyglucose uptake by positron emission tomography in epithelial tumors of the ovary. Int J Cancer 109, 926- 932.

Lachaal, M., Berenski, C. J., Kim, J., and Jung, C. Y. (1990). An ATP-modulated specific association of glyceraldehyde-3-phosphate dehydrogenase with human erythrocyte glucose transporter. J Biol Chem 265, 15449-15454.

Lange, K., and Brandt, U. (1990a). Insulin-responsive glucose transporters are concentrated in a cell surface-derived membrane fraction of 3T3-L1 adipocytes. FEBS Lett 261, 459-463.

Lange, K., and Brandt, U. (1990b). Restricted localization of the adipocyte/muscle glucose transporter species to a cell surface-derived vesicle fraction of 3T3-L1 155

adipocytes. Inhibited lateral mobility of integral plasma membrane proteins in newly inserted membrane areas of differentiated 3T3-L1 cells. FEBS Lett 276, 39-41.

Lee, W., Ryu, J. W., Souto, R. P., Pilch, P. F., and Jung, C. Y. (1999). Separation and partial characterization of three distinct intracellular GLUT4 compartments in rat adipocytes. Subcellular fractionation without homogenization. J Biol Chem 274, 37755- 37762.

Li, Q., Manolescu, A., Ritzel, M., Yao, S., Slugoski, M., Young, J. D., Chen, X. Z., and Cheeseman, C. I. (2004). Cloning and Functional Characterization of the Human GLUT7 Isoform (SLC2A7) from the Small Intestine. Am J Physiol Gastrointest Liver Physiol.

Lin, S., and Spudich, J. A. (1974). Biochemical studies on the mode of action of cytochalasin B. Cytochalasin B binding to red cell membrane in relation to glucose transport. J Biol Chem 249, 5778-5783.

Lisinski, I., Schurmann, A., Joost, H. G., Cushman, S. W., and Al-Hasani, H. (2001). Targeting of GLUT6 (formerly GLUT9) and GLUT8 in rat adipose cells. Biochem J 358, 517-522.

Liu, P., Leffler, B. J., Weeks, L. K., Chen, G., Bouchard, C. M., Strawbridge, A. B., and Elmendorf, J. S. (2004). Sphingomyelinase activates GLUT4 translocation via a cholesterol-dependent mechanism. Am J Physiol Cell Physiol 286, C317-329.

Livingstone, C. J., and Schachter, D. (1980). Calcium modulates the lipid dynamics of rat hepatocyte plasma membranes by direct and indirect mechanisms. Biochemistry 19, 4823-4827.

Loisel, T., Adam, L., Hebert, T., and Bouvier, M. (1996). Agonist stimulation increases the turnover rate of beta 2AR-bound palmitate and promotes receptor depalmitoylation. Biochemistry 35, 15923-15932.

London, E. (2002). Insights into lipid raft structure and formation from experiments in model membranes. Curr Opin Struct Biol 12, 480.

London, E., and Brown, D. A. (2000). Insolubility of lipids in triton X-100: physical origin and relationship to sphingolipid/cholesterol membrane domains (rafts). Biochim Biophys Acta 1508, 182-195.

Martens, J. R., Sakamoto, N., Sullivan, S. A., Grobaski, T. D., and Tamkun, M. M. (2001). Isoform-specific localization of voltage-gated K+ channels to distinct lipid raft populations. Targeting of Kv1.5 to caveolae. J Biol Chem 276, 8409-8414.

Matthaei, S., Olefsky, J. M., and Karnieli, E. (1988). Cycloheximide decreases glucose transporters in rat adipocyte plasma membranes without affecting insulin-stimulated glucose transport. Biochem J 251, 491-497. 156

McKeel, D. W., and Jarett, L. (1969). Preparation and Characterization of Fat Cell Plasma Membrane. Fed Proc 28, 879.

McKeel, D. W., and Jarett, L. (1970). Preparation and characterization of a plasma membrane fraction from isolated fat cells. J Cell Biol 44, 417-432.

McVie-Wylie, A. J., Lamson, D. R., and Chen, Y. T. (2001). Molecular cloning of a novel member of the GLUT family of transporters, SLC2a10 (GLUT10), localized on chromosome 20q13.1: a candidate gene for NIDDM susceptibility. Genomics 72, 113- 117.

Melkonian, K. A., Ostermeyer, A. G., Chen, J. Z., Roth, M. G., and Brown, D. A. (1999). Role of lipid modifications in targeting proteins to detergent- resistant membrane rafts. Many raft proteins are acylated, while few are prenylated. J Biol Chem 274, 3910-3917.

Mercado, C. L., Loeb, J. N., and Ismail-Beigi, F. (1989). Enhanced glucose transport in response to inhibition of respiration in Clone 9 cells. Am J Physiol 257, C19-28.

Mitsumoto, Y., Burdett, E., Grant, A., and Klip, A. (1991). Differential expression of the GLUT1 and GLUT4 glucose transporters during differentiation of L6 muscle cells. Biochem Biophys Res Commun 175, 652-659.

Mitsumoto, Y., Downey, G. P., and Klip, A. (1992a). Stimulation of glucose transport in L6 muscle cells by long-term intermittent stretch-relaxation. FEBS Lett 301, 94-98.

Mitsumoto, Y., and Klip, A. (1992b). Development regulation of the subcellular distribution and glycosylation of GLUT1 and GLUT4 glucose transporters during myogenesis of L6 muscle cells. J Biol Chem 267, 4957-4962.

Moore, M. S., Mahaffey, D. T., Brodsky, F. M., and Anderson, R. G. (1987). Assembly of clathrin-coated pits onto purified plasma membranes. Science 236, 558-563.

Mueckler, M. (1990). Family of glucose-transporter genes. Implications for glucose homeostasis and diabetes. Diabetes 39, 6-11.

Mueckler, M. (1994). Facilitative glucose transporters. Eur J Biochem 219, 713-725.

Mueckler, M., Caruso, C., Baldwin, S. A., Panico, M., Blench, I., Morris, H. R., Allard, W. J., Lienhard, G. E., and Lodish, H. F. (1985). Sequence and structure of a human glucose transporter. Science 229, 941-945.

Munro, S. (2003). Lipid rafts: elusive or illusive? Cell 115, 377-388.

Murata, H., Hruz, P. W., and Mueckler, M. (2002). CONCISE COMMUNICATION Indinavir inhibits the glucose transporter isoform Glut4 at physiologic concentrations. Aids 16, 859-863. 157

Murata, M., Peranen, J., Schreiner, R., Wieland, F., Kurzchalia, T. V., and Simons, K. (1995). VIP21/caveolin is a cholesterol-binding protein. Proc Natl Acad Sci U S A 92, 10339-10343.

Ohvo-Rekila, H., Ramstedt, B., Leppimaki, P., and Slotte, J. P. (2002). Cholesterol interactions with phospholipids in membranes. Prog Lipid Res 41, 66-97.

Oka, Y., and Czech, M. P. (1984). Photoaffinity labeling of insulin-sensitive hexose transporters in intact rat adipocytes. Direct evidence that latent transporters become exposed to the extracellular space in response to insulin. J Biol Chem 259, 8125-8133.

Oliver, J. D., Hresko, R. C., Mueckler, M., and High, S. (1996). The glut 1 glucose transporter interacts with calnexin and calreticulin. J Biol Chem 271, 13691-13696.

Ouiddir, A., Planes, C., Fernandes, I., VanHesse, A., and Clerici, C. (1999). Hypoxia upregulates activity and expression of the glucose transporter GLUT1 in alveolar epithelial cells. Am J Respir Cell Mol Biol 21, 710-718.

Pantaleon, M., Ryan, J. P., Gil, M., and Kaye, P. L. (2001). An unusual subcellular localization of GLUT1 and link with metabolism in oocytes and preimplantation mouse embryos. Biol Reprod 64, 1247-1254.

Parkar, B. A., Midgley, P. J., and Holman, G. D. (1985). Interaction of hydrophobic bis (D-mannose) derivatives with adipocyte and erythrocyte sugar transport systems. Biochim Biophys Acta 814, 103-110.

Parkinson, N. A., and Bolsover, S. R. (2001). A nuclear location for Ca2+-activated adenylyl cyclases I and III in neurones. Brain Res Mol Brain Res 91, 43-49.

Parpal, S., Karlsson, M., Thorn, H., and Stralfors, P. (2001). Cholesterol depletion disrupts caveolae and insulin receptor signaling for metabolic control via insulin receptor substrate-1, but not for mitogen-activated protein kinase control. J Biol Chem 276, 9670- 9678.

Pasquali, C., Fialka, I., and Huber, L. A. (1997). Preparative two-dimensional gel electrophoresis of membrane proteins. Electrophoresis 18, 2573-2581.

Pasquali, C., Fialka, I., and Huber, L. A. (1999). Subcellular fractionation, electromigration analysis and mapping of organelles. J Chromatogr B Biomed Sci Appl 722, 89-102.

Phay, J. E., Hussain, H. B., and Moley, J. F. (2000). Cloning and expression analysis of a novel member of the facilitative glucose transporter family, SLC2A9 (GLUT9). Genomics 66, 217-220.

Pouliot, J. F., and Beliveau, R. (1995). Palmitoylation of the glucose transporter in blood- brain barrier capillaries. Biochim Biophys Acta 1234, 191-196. 158

Prasad, R. K., and Ismail-Beigi, F. (1999). Mechanism of stimulation of glucose transport by H2O2: role of phospholipase C. Arch Biochem Biophys 362, 113-122.

Prior, I. A., Muncke, C., Parton, R. G., and Hancock, J. F. (2003a). Direct visualization of Ras proteins in spatially distinct cell surface microdomains. J Cell Biol 160, 165-170.

Prior, I. A., Parton, R. G., and Hancock, J. F. (2003b). Observing cell surface signaling domains using electron microscopy. Sci STKE 2003, PL9.

Quintanilla, R. A., Porras, O. H., Castro, J., and Barros, L. F. (2000). Cytosolic [Ca(2+)] modulates basal GLUT1 activity and plays a permissive role in its activation by metabolic stress and insulin in rat epithelial cells. Cell Calcium 28, 97-106.

Rajaram, S., Spangler, T. L., Sedensky, M. M., and Morgan, P. G. (1999). A stomatin and a degenerin interact to control anesthetic sensitivity in Caenorhabditis elegans. Genetics 153, 1673-1682.

Ramlal, T., Sarabia, V., Bilan, P. J., and Klip, A. (1988). Insulin-mediated translocation of glucose transporters from intracellular membranes to plasma membranes: sole mechanism of stimulation of glucose transport in L6 muscle cells. Biochem Biophys Res Commun 157, 1329-1335.

Richardson, S., Neama, G., Phillips, T., Bell, S., Carter, S. D., Moley, K. H., Moley, J. F., Vannucci, S. J., and Mobasheri, A. (2003). Molecular characterization and partial cDNA cloning of facilitative glucose transporters expressed in human articular chondrocytes; stimulation of 2-deoxyglucose uptake by IGF-I and elevated MMP-2 secretion by glucose deprivation. Osteoarthritis Cartilage 11, 92-101.

Robinson, J. M., and Karnovsky, M. J. (1980). Evaluation of the polyene antibiotic filipin as a cytochemical probe for membrane cholesterol. J Histochem Cytochem 28, 161-168.

Robinson, L. J., Pang, S., Harris, D. S., Heuser, J., and James, D. E. (1992). Translocation of the glucose transporter (GLUT4) to the cell surface in permeabilized 3T3-L1 adipocytes: effects of ATP insulin, and GTP gamma S and localization of GLUT4 to clathrin lattices. J Cell Biol 117, 1181-1196.

Rodbell, M. (1967). Metabolism of isolated fat cells. V. Preparation of "ghosts" and their properties; adenyl cyclase and other enzymes. J Biol Chem 242, 5744-5750.

Rogers, S., Chandler, J. D., Clarke, A. L., Petrou, S., and Best, J. D. (2003). Glucose transporter GLUT12-functional characterization in Xenopus laevis oocytes. Biochem Biophys Res Commun 308, 422-426.

Rogers, S., Macheda, M. L., Docherty, S. E., Carty, M. D., Henderson, M. A., Soeller, W. C., Gibbs, E. M., James, D. E., and Best, J. D. (2002). Identification of a novel glucose transporter-like protein-GLUT-12. Am J Physiol Endocrinol Metab 282, E733-738. 159

Röper, K., Corbeil, D., and Huttner, W. B. (2000). Retention of prominin in microvilli reveals distinct cholesterol-based lipid microdomains in the apical plasma membrane. Nature Cell Biology 2, 582-592.

Ros-Baro, A., Lopez-Iglesias, C., Peiro, S., Bellido, D., Palacin, M., Zorzano, A., and Camps, M. (2001). Lipid rafts are required for GLUT4 internalization in adipose cells. Proc Natl Acad Sci USA 98, 12050-12055.

Rothschild, J. A. (1961). Sub-fractionation of Rat Liver Microsomes. Fed Proc 20, 145.

Rubin, D., and Ismail-Beigi, F. (2003). Distribution of Glut1 in detergent-resistant membranes (DRMs) and non-DRM domains: effect of treatment with azide. Am J Physiol Cell Physiol 285, C377-383.

Rubin, D., and Ismail-Beigi, F. (2004). Glut1, Detergent-Resistant Membranes (DRMs), and the Acute Stimulation of Glucose Transport: Effects at the Plasma Membrane. FASEB Journal 18, A697.

Sakyo, T., and Kitagawa, T. (2002). Differential localization of glucose transporter isoforms in non- polarized mammalian cells: distribution of GLUT1 but not GLUT3 to detergent-resistant membrane domains. Biochim Biophys Acta 1567, 165-175.

Saltiel, A. R., and Pessin, J. E. (2003). Insulin signaling in microdomains of the plasma membrane. Traffic 4, 711-716.

Salzer, U., Ahorn, H., and Prohaska, R. (1993). Identification of the phosphorylation site on human erythrocyte band 7 integral membrane protein: implications for a monotopic protein structure. Biochim Biophys Acta 1151, 149-152.

Salzer, U., and Prohaska, R. (2001). Stomatin, flotillin-1, and flotillin-2 are major integral proteins of erythrocyte lipid rafts. Blood 97, 1141-1143.

Samuel, B. U., Mohandas, N., Harrison, T., McManus, H., Rosse, W., Reid, M., and Haldar, K. (2001). The Role of Cholesterol and Glycosylphosphatidylinositol-anchored Proteins of Erythrocyte Rafts in Regulating Raft Protein Content and Malarial Infection. J Biol Chem 276, 29319-29329.

Santoni, V., Kieffer, S., Desclaux, D., Masson, F., and Rabilloud, T. (2000). Membrane proteomics: use of additive main effects with multiplicative interaction model to classify plasma membrane proteins according to their solubility and electrophoretic properties. Electrophoresis 21, 3329-3344.

Scheepers, A., Doege, H., Joost, H. G., and Schurmann, A. (2001). Mouse GLUT8: Genomic Organization and Regulation of Expression in 3T3- L1 Adipocytes by Glucose. Biochem Biophys Res Commun 288, 969-974.

160

Scheiffele, P., Roth, M. G., and Simons, K. (1997). Interaction of influenza virus haemagglutinin with sphingolipid- cholesterol membrane domains via its transmembrane domain. Embo J 16, 5501-5508.

Scherer, P., Lisanti, M., Baldini, G., Sargiacomo, M., Mastick, C., and Lodish, H. (1994). Induction of caveolin during adipogenesis and association of GLUT4 with caveolin-rich vesicles. J Cell Biol 127, 1233-1243.

Schlegel, A., and Lisanti, M. P. (2001). The caveolin triad: caveolae biogenesis, cholesterol trafficking, and signal transduction. Cytokine Growth Factor Rev 12, 41-51.

Schootemeijer, A., Van Beekhuizen, A. E., Tertoolen, L. G., De Laat, S. W., and Akkerman, J. W. (1994). Cytosolic calcium ions regulate lipid mobility in the plasma membrane of the human megakaryoblastic cell line MEG-01. Eur J Biochem 224, 423- 430.

Schuck, S., Honsho, M., Ekroos, K., Shevchenko, A., and Simons, K. (2003). Resistance of cell membranes to different detergents. Proc Natl Acad Sci U S A 100, 5795-5800.

Schulze, W., and Buchwalow, I. B. (1998). Adenylyl cyclase in the heart: an enzymocytochemical and immunocytochemical approach. Microsc Res Tech 40, 473- 478.

Sedensky, M. M., Siefker, J. M., and Morgan, P. G. (2001). Model organisms: new insights into ion channel and transporter function. Stomatin homologues interact in Caenorhabditis elegans. Am J Physiol Cell Physiol 280, C1340-1348.

Shepherd, P. R., Gibbs, E. M., Wesslau, C., Gould, G. W., and Kahn, B. B. (1992). Human small intestine facilitative fructose/glucose transporter (GLUT5) is also present in insulin-responsive tissues and brain. Investigation of biochemical characteristics and translocation. Diabetes 41, 1360-1365.

Shetty, M., Ismail-Beigi, N., Loeb, J. N., and Ismail-Beigi, F. (1993a). Induction of GLUT1 mRNA in response to inhibition of oxidative phosphorylation. Am J Physiol 265, C1224-1229.

Shetty, M., Loeb, J. N., and Ismail-Beigi, F. (1992). Enhancement of glucose transport in response to inhibition of oxidative metabolism: pre- and posttranslational mechanisms. Am J Physiol 262, C527-532.

Shetty, M., Loeb, J. N., Vikstrom, K., and Ismail-Beigi, F. (1993b). Rapid activation of GLUT-1 glucose transporter following inhibition of oxidative phosphorylation in clone 9 cells. J Biol Chem 268, 17225-17232.

Shi, Y., Liu, H., Vanderburg, G., Samuel, S. J., Ismail-Beigi, F., and Jung, C. Y. (1995). Modulation of GLUT1 intrinsic activity in clone 9 cells by inhibition of oxidative phosphorylation. J Biol Chem 270, 21772-21778. 161

Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387, 569- 572.

Simons, K., and van Meer, G. (1988). Lipid sorting in epithelial cells. Biochemistry 27, 6197-6202.

Simpson, I. A., Yver, D. R., Hissin, P. J., Wardzala, L. J., Karnieli, E., Salans, L. B., and Cushman, S. W. (1983). Insulin-stimulated translocation of glucose transporters in the isolated rat adipose cells: characterization of subcellular fractions. Biochim Biophys Acta 763, 393-407.

Skibbens, J. E., Roth, M. G., and Matlin, K. S. (1989). Differential extractability of influenza virus hemagglutinin during intracellular transport in polarized epithelial cells and nonpolar fibroblasts. J Cell Biol 108, 821-832.

Smart, E., Ying, Y., Conrad, P., and Anderson, R. (1994). Caveolin moves from caveolae to the Golgi apparatus in response to cholesterol oxidation. J Cell Biol 127, 1185-1197.

Smith, R. M., Harada, S., Smith, J. A., Zhang, S., and Jarett, L. (1998). Insulin-induced protein tyrosine phosphorylation cascade and signalling molecules are localized in a caveolin-enriched cell membrane domain. Cell Signal 10, 355-362.

Smith, U., Kuroda, M., and Simpson, I. A. (1984). Counter-regulation of insulin- stimulated glucose transport by catecholamines in the isolated rat adipose cell. J Biol Chem 259, 8758-8763.

Snyers, L., Thines-Sempoux, D., and Prohaska, R. (1997). Colocalization of stomatin (band 7.2b) and actin microfilaments in UAC epithelial cells. Eur J Cell Biol 73, 281- 285.

Snyers, L., Umlauf, E., and Prohaska, R. (1998). Oligomeric nature of the integral membrane protein stomatin. J Biol Chem 273, 17221-17226.

Snyers, L., Umlauf, E., and Prohaska, R. (1999a). Association of stomatin with lipid- protein complexes in the plasma membrane and the endocytic compartment. Eur J Cell Biol 78, 802-812.

Snyers, L., Umlauf, E., and Prohaska, R. (1999b). Cysteine 29 is the major palmitoylation site on stomatin. FEBS Lett 449, 101-104.

Somwar, R., Kim, D. Y., Sweeney, G., Huang, C., Niu, W., Lador, C., Ramlal, T., and Klip, A. (2001). GLUT4 translocation precedes the stimulation of glucose uptake by insulin in muscle cells: potential activation of GLUT4 via p38 mitogen- activated protein kinase. Biochem J 359, 639-649.

Somwar, R., Koterski, S., Sweeney, G., Sciotti, R., Djuric, S., Berg, C., Trevillyan, J., Scherer, P. E., Rondinone, C. M., and Klip, A. (2002). A dominant-negative p38 MAPK mutant and novel selective inhibitors of p38 MAPK reduce insulin-stimulated glucose 162

uptake in 3T3-L1 adipocytes without affecting GLUT4 translocation. J Biol Chem 277, 50386-50395.

Song, K. S., Li, S., Okamoto, T., Quilliam, L. A., Sargiacomo, M., and Lisanti, M. P. (1996). Co-purification and direct interaction of Ras with caveolin, an integral membrane protein of caveolae microdomains. Detergent-free purification of caveolae microdomains. J Biol Chem 271, 9690-9697.

Souto, R. P., Vallega, G., Wharton, J., Vinten, J., Tranum-Jensen, J., and Pilch, P. F. (2003). Immunopurification and characterization of rat adipocyte caveolae suggest their dissociation from insulin signaling. J Biol Chem 278, 18321-18329.

Stewart, G. W. (1997). Stomatin. Int J Biochem Cell Biol 29, 271-274.

Stuart, C. A., Wen, G., Williamson, M. E., Jiang, J., Gilkison, C. R., Blackwell, S. J., Nagamani, M., and Ferrando, A. A. (2001). Altered GLUT1 and GLUT3 gene expression and subcellular redistribution of GLUT4: protein in muscle from patients with acanthosis nigricans and severe insulin resistance. Metabolism 50, 771-777.

Stubbs, C. D., and Smith, A. D. (1984). The modification of mammalian membrane polyunsaturated fatty acid composition in relation to membrane fluidity and function. Biochim Biophys Acta 779, 89-137.

Sultzman, L. A., and Carruthers, A. (1999). Stop-flow analysis of cooperative interactions between GLUT1 sugar import and export sites. Biochemistry 38, 6640-6650.

Suzuki, K., and Kono, T. (1980). Evidence that insulin causes translocation of glucose transport activity to the plasma membrane from an intracellular storage site. Proc Natl Acad Sci U S A 77, 2542-2545.

Taha, C., Tsakiridis, T., McCall, A., and Klip, A. (1997). Glucose transporter expression in L6 muscle cells: regulation through insulin- and stress-activated pathways. Am J Physiol 273, E68-76.

Takata, K. (1996). Glucose transporters in the transepithelial transport of glucose. J Electron Microsc (Tokyo) 45, 275-284.

Tordjman, K. M., Leingang, K. A., James, D. E., and Mueckler, M. M. (1989). Differential regulation of two distinct glucose transporter species expressed in 3T3-L1 adipocytes: effect of chronic insulin and tolbutamide treatment. Proc Natl Acad Sci U S A 86, 7761-7765.

Tsakiridis, T., Taha, C., Grinstein, S., and Klip, A. (1996). Insulin activates a p21- activated kinase in muscle cells via phosphatidylinositol 3-kinase. J Biol Chem 271, 19664-19667. 163

Turner, E. J., Jarvis, H. G., Chetty, M. C., Landon, G., Rowley, P. S., Ho, M. M., and Stewart, G. W. (2003). ATP-dependent vesiculation in red cell membranes from different hereditary stomatocytosis variants. Br J Haematol 120, 894-902.

Uldry, M., Ibberson, M., Horisberger, J. D., Chatton, J. Y., Riederer, B. M., and Thorens, B. (2001). Identification of a mammalian H(+)-myo-inositol symporter expressed predominantly in the brain. Embo J 20, 4467-4477.

Vainio, S., Heino, S., Mansson, J. E., Fredman, P., Kuismanen, E., Vaarala, O., and Ikonen, E. (2002). Dynamic association of human insulin receptor with lipid rafts in cells lacking caveolae. EMBO Rep 3, 95-100. van den Berghe, N., Barros, L. F., van Mackelenbergh, M. G., and Krans, H. M. (1996). Clostridium botulinum C3 exoenzyme stimulates GLUT4-mediated glucose transport, but not glycogen synthesis, in 3T3-L1 adipocytes--a potential role of rho? Biochem Biophys Res Commun 229, 430-439.

Vannucci, S. J., Nishimura, H., Satoh, S., Cushman, S. W., Holman, G. D., and Simpson, I. A. (1992). Cell surface accessibility of GLUT4 glucose transporters in insulin- stimulated rat adipose cells. Modulation by isoprenaline and adenosine. Biochem J 288 ( Pt 1), 325-330.

Varma, R., and Mayor, S. (1998). GPI-anchored proteins are organized in submicron domains at the cell surface. Nature 394, 798-801.

Vera, J. C., Rivas, C. I., Fischbarg, J., and Golde, D. W. (1993). Mammalian facilitative hexose transporters mediate the transport of dehydroascorbic acid. Nature 364, 79-82.

Vera, J. C., Rivas, C. I., Zhang, R. H., Farber, C. M., and Golde, D. W. (1994). Human HL-60 myeloid leukemia cells transport dehydroascorbic acid via the glucose transporters and accumulate reduced ascorbic acid. Blood 84, 1628-1634.

Vordermark, D., and Brown, J. M. (2003). Endogenous markers of tumor hypoxia predictors of clinical radiation resistance? Strahlenther Onkol 179, 801-811.

Wardzala, L. J., Cushman, S. W., and Salans, L. B. (1978). Mechanism of insulin action on glucose transport in the isolated rat adipose cell. Enhancement of the number of functional transport systems. J Biol Chem 253, 8002-8005.

Webb, Y., Hermida-Matsumoto, L., and Resh, M. D. (2000). Inhibition of protein palmitoylation, raft localization, and T cell signaling by 2-bromopalmitate and polyunsaturated fatty acids. J Biol Chem 275, 261-270.

Weber, T. M., Joost, H. G., Simpson, I. A., and Cushman, S. W. (1988). Methods for Assessment of Glucose Transport Activity and the Number of Glucose Transporters in Isolated Rat Adipose Cells and Membrane Fractions. In Insulin Receptors, Part B: Clinical Assessment, Biological Responses, and Comparison to the IGF-1 Receptor, C. R. Kahn, and L. C. Harrison, eds. (New York, Alan R. Liss, Inc), pp. 171-181. 164

Wilson, B. S., Pfeiffer, J. R., and Oliver, J. M. (2000). Observing FcepsilonRI signaling from the inside of the mast cell membrane. J Cell Biol 149, 1131-1142.

Wilson, C. M., Mitsumoto, Y., Maher, F., and Klip, A. (1995). Regulation of cell surface GLUT1, GLUT3, and GLUT4 by insulin and IGF-I in L6 myotubes. FEBS Lett 368, 19- 22.

Wood, I. S., Hunter, L., and Trayhurn, P. (2003a). Expression of Class III facilitative glucose transporter genes (GLUT-10 and GLUT-12) in mouse and human adipose tissues. Biochem Biophys Res Commun 308, 43-49.

Wood, I. S., and Trayhurn, P. (2003b). Glucose transporters (GLUT and SGLT): expanded families of sugar transport proteins. Br J Nutr 89, 3-9.

Wood, R. E., and Morgan, H. E. (1969). Regulation of sugar transport in avian erythrocytes. J Biol Chem 244, 1451-1460.

Wu, X., and Freeze, H. H. (2002a). GLUT14, a duplicon of GLUT3, is specifically expressed in testis as alternative splice forms. Genomics 80, 553-557.

Wu, X., Li, W., Sharma, V., Godzik, A., and Freeze, H. H. (2002b). Cloning and characterization of glucose transporter 11, a novel sugar transporter that is alternatively spliced in various tissues. Mol Genet Metab 76, 37-45.

Xi, X., Han, J., and Zhang, J. Z. (2001). Stimulation of Glucose Transport by AMP- activated Protein Kinase via Activation of p38 Mitogen-activated Protein Kinase. J Biol Chem 276, 41029-41034.

Yancey, P. G., Rodrigueza, W. V., Kilsdonk, E. P., Stoudt, G. W., Johnson, W. J., Phillips, M. C., and Rothblat, G. H. (1996). Cellular cholesterol efflux mediated by cyclodextrins. Demonstration Of kinetic pools and mechanism of efflux. J Biol Chem 271, 16026-16034.

Yeagle, P. L. (1985). Cholesterol and the cell membrane. Biochim Biophys Acta 822, 267-287.

Yu, J., Fischman, D. A., and Steck, T. L. (1973). Selective solubilization of proteins and phospholipids from red blood cell membranes by nonionic detergents. J Supramol Struct 1, 233-248.

Yuli, I., Wilbrandt, W., and Shinitzky, M. (1981). Glucose transport through cell membranes of modified lipid fluidity. Biochemistry 20, 4250-4256.

Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002). Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 296, 913-916. 165

Zeng, H., Parthasarathy, R., Rampal, A. L., and Jung, C. Y. (1996). Proposed structure of putative glucose channel in GLUT1 facilitative glucose transporter. Biophys J 70, 14-21.

Zhang, J. Z., Abbud, W., Prohaska, R., and Ismail-Beigi, F. (2001). Overexpression of stomatin depresses GLUT-1 glucose transporter activity. Am J Physiol Cell Physiol 280, C1277-1283.

Zhang, J. Z., Behrooz, A., and Ismail-Beigi, F. (1999a). Regulation of glucose transport by hypoxia. Am J Kidney Dis 34, 189-202.

Zhang, J. Z., Hayashi, H., Ebina, Y., Prohaska, R., and Ismail-Beigi, F. (1999b). Association of stomatin (band 7.2b) with Glut1 glucose transporter. Arch Biochem Biophys 372, 173-178.

Zhang, J. Z., and Ismail-Beigi, F. (1998). Activation of Glut1 glucose transporter in human erythrocytes. Arch Biochem Biophys 356, 86-92.

Ziehm, D., Schurmann, A., Weiland, M., and Joost, H. G. (1993). Biphasic alteration of glucose transport in 3T3-L1 cells during differentiation to the adipocyte-like phenotype. Horm Metab Res 25, 71-76.

Zuniga, F. A., Shi, G., Haller, J. F., Rubashkin, A., Flynn, D. R., Iserovich, P., and Fischbarg, J. (2001). A three-dimensional model of the human facilitative glucose transporter Glut1. J Biol Chem 24, 24.