53;»
\ HMS

\l
,.
31$

 

Iv Iw-nofi-A'“ L. 'fl'FI

. V_,.,‘,,,.‘, _v

NS” :3?

33"

t

 

THEStS

(-‘J

\
_v
lb

 

uamv llljl’gff‘lll lllll llllll
Michigan State

University

 

 

 

This is to certify that the

dissertation entitled

Arachidonic Acid Regulation of Lipogenic Gene
Expression in Adipocytes and Hepatocytes

presented by

Michelle K. Mater

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Animal Science

 

 

294*.”- W

Major pigessor

Date AUQUSt 28, 1998

 

MSU is an Affirmative Action/Equal Opportunity Institution O~ 12771

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

”mm afizooa

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DateDuo.p65-p.15

 

 

ARACHIDONIC ACID REGULATION OF LIPOGENIC GENE EXPRESSION
IN ADIPOCYTES AND HEPATOCYTES

BY

Michelle Kay Mater

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Animal Science

1998

ABSTRACT

ARACHIDONIC ACID REGULATION OF LIPOGENIC GENE EXPRESSION IN
ADIPOCYTES AND HEPATOCYTES

BY

Michelle Kay Mater

Previous observations have shown that dietary
polyunsaturated fatty acids (PUFA) can inhibit the
expression of lipogenic genes in both liver and epididymal
fat. Transcriptional regulation of genes by PUFA have given
rise to three hypothetical mechanisms of action: a
prostanoid pathway, a PPAR pathway or a prostanoid/PPAR
independent pathway. Using the $14 gene as a model for
lipogenic gene expression, preliminary studies showed that
in cultured adipocytes, arachidonic acid (20:4n-6; AA)
inhibited both mRNA expression and transcriptional activity
in reporter studies more potently than eicosapentaenoic acid
(EPA). This result prompted the analysis of AA inhibition
in both adipocytes and hepatocytes. In adipocytes, AA
inhibition of $14 gene expression was reversed by the
cyclooxygenase inhibitor, flurbiprofen, implicating a
prostanoid pathway. Results also indicated PGE; inhibited
514 expression and Sl4CAT activity. The PGEz effect was

blocked by pertussis toxin, a Gi-protein inhibitor.

Prostaglandin (PG) agonist results, coupled with inhibitors
of specific signal transduction pathways, suggested PGBZ
action was mediated through a Cay'linked pathway.

In hepatocytes, AA and PG also inhibited $14
expression. However, the AA effect was not blocked by
flurbiprofen, indicating the PUFA control of gene expression
did not require cyclooxygenase. In liver, PG are produced
by Kupffer cells and act on hepatocytes. In adipocytes PG
are synthesized by both preadipocytes and adipocytes. PG-
agonist studies implicate involvement of EP3 receptors, but
failed to reveal the mediator of PGE2 action in hepatocytes.

Promoter deletion analysis indicated that both AA and
PGE2 targeted similar regions within the $14 promoter,
showing that AA can act directly on parenchymal cells or
through a paracrine mechanism to alter parenchymal cell gene
expression. Presumably, AA enters Kupffer cells, is
converted to PG and secreted to interact with EP3 receptors
on parenchymal cells. Interestingly, both PGEz and AA have
similar effects on gene in parenchymal cells.

In conclusion, these studies have shown that dietary n-
6 PUFA conversion to PG can provide another route for the
control of hepatic and adipogenic lipogenic gene expression.
I speculate this route plays a significant role in hepatic
gene expression under conditions stimulating Kupffer cell PG

synthesis, i.e. inflammation.

To my family:

Dad
Mom
Thinh
Pamela
Michael
Janet
Mindi
Terry
Morgan
Cassandra
Whitney
Marc
Michelle
Kayleigh
Ryan
Blade

and

my PHD (Pretty Happy Dog)

iv

 

AflflflflflflbEDGflflflMNTS

I would like to thank several people who have helped me
during my graduate career. I would first like to thank my
committee members: Drs. Bergen, Rozeboom, Romsos, Dewitt and
Jump. Although I started under the direction of Dr. Bergen
in the Department of Animal Science, Dr. Jump in the
Department of Physiology has invested much of his time and
research knowledge in my projects and in me. I greatly
appreciate his help and advise. I would also like to thank
Dr. Romsos who contributed both his time and space in his
lab. Thanks to Dr. Dewitt for his help and suggestions as
well. Thanks to Dr. Rozeboom for his help and assistance
with all the departmental requirements and his advice about
the pig business.

Others in the labs I have worked in have made my
graduate career easier and more enjoyable. Thanks
especially to Dr. Annette Thelen, my friend and lab-mate for
all her help, advise and ice-cream. Thanks also to Dr.
David Pan, Sharon DeBar, Dr. Bing Ren, Dr. Patty Weber, Dr.
Nathalie Trottier and Dr. Matt Doumit for their assistance
in the past few years. Scott Kramer and Josep Garcia-Sirera
have also been great friends and very supportive.

Finally, I would like to thank my family. Even though

they have often been confused about my research, they have

never failed to support my choices. My parents have been
especially helpful in providing free food and relaxation
whenever I needed it. Words can not expression my thanks to
my devoted friend, Blade, who not only listened but
understood my research better than any other person outside
of the lab. She has put up with a lot of stress because of
me and I can not imagine having done this without her
listening abilities and her jokes.

Thank-you everyone!

vi

IHUIEE CHPIQONTEHEBS

LIST OF TABLES ........................................... ix
LIST OF
FIGURES.. .................................................. x
LIST OF
ABBREVIATIONS ............................................ xii
INTRODUCTION ............................................... 1
CHAPTER 1
LITERATURE
REVIEW ..................................................... 3
Lipid Metabolism ...................................... 3
Lipid Synthesis ....................................... 5
Lipid Oxidation ....................................... 7
Dietary Fat ..................................... 10
Polyunsaturated Fatty Acids ................... ..11
Possible Mechanisms for PUFA Control of
Lipogenic Gene Expression ............... ..15
Adipocytes ..... . ..................................... 18
Adipogenic Regulators ........................... 20
Adipogenic Transcription Factors ..... . .......... 27
Lipogenic Gene Expression in Adipocytes ......... 36
Arachidonic Acid ..................................... 37
Prostaglandins ....................................... 45
Prostaglandin Receptors ......................... 45
Actions of Prostaglandins ....................... 49
Hepatocyte and Kupffer Cell Interactions. ..... ....... 53
Tumor Necrosis Factor ................... ... .55
TNF and P63 in Primary Hepatocyte Culture ....... 62
Calcium Mediators. ........ .................. . ...... 65
Regulation of FAS and $14 ..... . .................... ..70
PAS and $14 Promoter....... ..................... 74
Concluding Statements and Hypothesis ................. 78
CHAPTER 2
ADIPOCYTES.. .............................................. 80
Aims..... ............................................ 82
Introduction... ...................................... 83
Methods and Materials ................................ 86

vii

Results ............................................. 94

Discussion ......................................... 125
Conclusions ........................................ 129

CHAPTER 3

HEPATOCYTES.... ......................................... 132
Aims ............................................. ..134
Introduction.... ................................... 134
Methods and Materials .............................. 136
Results...... ...................................... 138
Discussion... ...................................... 158
Conclusions ........................................ 163

CHAPTER 4

FUTURE STUDIES .......................................... 165
Questions .......................................... 165
Introduction ....................................... 166
Methods and Materials .............................. 167
Results ............................................ 167
Discussion ......... . ............................. ..173
Conclusions and Future Directions .................. 177

APPENDIX A

ADIPOCYTE EXPERIMENTS - SOLUTIONS ....................... 181

APPENDIX B

HEPATOCYTE PREPARATION .................................. 185

REFERENCES .............................................. 191

viii

 

LIST OF TABLES

 

TABLE NUMBER TABLE NAME

PAGE

2.1 Fatty Acid Sensitivity of $14 Constructs.......109

2.2 Total Triglycerides in Adipocytes .............. 122

3.1 PGE2 Signal Transduction Inhibitors ...... . ..... 144

3.2 Prostaglandin Agonists Effects on $14 CAT
Activity in Adipocytes and Hepatocytes ......... 147

3.3 AA Metabolism Inhibitor Effects on $14 CAT
Activity.. ............................. . ....... 150

ix

 

LIST OF FIGURES

 

NWO
Hi4
unv

HFJ
b

wwwww'ww DON
QQOSU'IbUJN

FIGURE NUMBER FIGURE NAME PAGE
1.1 Fatty Acid Elongation and Desaturation ............... 6
1.2 Lipid Metabolism.- ................................... 10
1.3 Structure of AA and EPA ............................. 13
1.4 Adipocyte Transcription Factors ..................... 37
1.5 Arachidonic Acid Metabolism ......................... 39
1.6 Overall Prostaglandin Synthesis and Signaling ....... 53
1.7 Summary of TNFa Action .............................. 58
1.8 Calcium Signaling Pathways .......................... 69
1.9 The 814 Promoter .................................... 78
2.1 The Effect of Fatty Acids on Adipocyte mRNA Levels..95
2.2 Effect of Fatty Acids on Sl4CAT Activity in

Preadipocytes and Adipocytes ........................ 97
Dose Response of Fatty Acids in Adipocytes on

Sl4CAT Activity ..................................... 98
Continuous Fatty Acid Treatment on Sl4CAT

Activity in Adipocytes .............................. 99
Oil Red 0 Stained Adipocytes ....................... 102
$14 Plasmid Construct Maps ........ ° ................. 105
Fatty Acid Specificity ............................. 106
Gel Shift Analysis of TSE-l and TSE-2 .............. 111
Pioglitazone Treatment of Adipocytes ............... 113
Flurbiprofen Blocks the Arachidonic Acid

Suppression of Sl4CAT Activity.....................115
PGE2 Inhibition of Sl4CAT Activity and RNA in
Adipocytes ......................................... 117

Inhibition of Sl4CAT Activity by CAMP and A23187...118
Effects of PGE2 and PT, H7 or Verapamil on

Sl4CAT Activity .................................... 120
Fatty Acid Profiles in Adipocytes .................. 124
AA Suppression of Sl4CAT Activity is Not

Reversed by Flurbiprofen ........................... 139
PGE2 Effects on Hepatic Gene Expression .......... ..140
Inhibition of Sl4CAT Activity by PG ................ 141

Effects of PGE2 and Inhibitors on Sl4CAT Activity..143
PGE2 Receptor Agonist effects on Sl4CAT Activity...146

PG and Agonist Structures .......................... 148
Specificity of PGE2 on the $14 Promoter ............ 152
Deletion Analysis of the $14 Promoter with

AA and PGE2 ........................................ 154
Y Box and C-Region are Required for PUFA

Control of $14 ..................................... 156

3.10 Promoter Context Dictates Control by PUFA and PGE2.158
3.11 Hepatocyte and Kupffer Cell Interactions ........... 162

 

FIGURE NUMBER FIGURE NAME PAGE

4.1 IL-1a and TNFa Effects on Sl4CAT Activity .......... 168
4.2 Dose Response of TNFa .............................. 169
4.3 TNFa Effects on $14 mRNA ........................... 170
4.4 $14 Promoter Analysis on Sl4CAT Activity by TNFa.. 171
4.5 TNFa Effect on $14 Proximal Promoter Elements ...... 173

xi

 

AA
ACS
ADDl

AOX
CamK

C/EBPa
CYP

DEX
EGF
EPA
FAS
FGF
GH
GLUT4

HSL
IBMX
IGF-l
IP3
LAP
LIP
LPL
OA
Ob17
Obl771
PEPCK
PDGF
PG
PKA
PKC
PPAR
PPRE

PRR
PUFA
RXR
SCD
SRE
SREBP
TAl

TGF
TH

TRE
TRR

LIST OF ABBREVIATIONS

Arachidonic acid (20:4, n-6)

Acyl CoA synthetase

Adipocyte determination and differentiation
dependent factor 1 (also known as SREBP)
Acyl CoA oxidase

Calcium calmodulin kinase

CCAAT enhancer binding protein alpha

Cytochrome P450 (many subtypes:2E1, 4A2, 2C23,

etc.)
Dexamethasone
Epidermal growth factor
Eicosapentaenoic acid (20:5, n-3)
Fatty acid synthase
Fibroblast growth factor
Growth Hormone
Glucose transporter 4 (in adipose and muscle
tissue)
Hormone sensitive lipase
Isobutylmethylxanthine
Insulin-like growth factor 1
inositol triphosphate
Liver activating protein
Liver inhibitory protein
Lipoprotein lipase
Oleic acid
Adipogenic cell line from ob/ob mice
Subclone of Obl7 cells
Phosphoenolpyruvate carboxykinase
Platelet derived growth factor
Prostaglandin
Protein kinase A
Protein kinase C
Peroxisome proliferator activated receptor
Peroxisome proliferator activated receptor
response element
Pluripotent response region
Polyunsaturated fatty acids
Retinoid X receptor
Stearoyl CoA desaturase (1 or 2)
Sterol response element
SRE binding protein
Adipogenic cell line from mice embryo 10Tl/2
mesenchymal cells treated w/ S'azacytidine
Transforming growth factor
Thyroid hormone
Thyroid hormone response element
Thyroid hormone response region

xii

in”

 

VLDL Very low density lipoproteins
WAT White adipose tissue

3T3-L1 Adipogenic cell line from mice
3T3-F442A Adipogenic cell line from mice

xiii

 

INHHKMWUCTUIRU

Nearly 40% of the calories in a typical American diet
is derived from fat. Approximately half of this is from
animal fat (meat and dairy) while the other half is from
plants, in the form of oils. Research has shown that high
dietary fat intake is correlated with heart disease,
diabetes, several forms of cancer and obesity. These health
risks led to the recommendation of decreasing dietary intake
to a maximum of 30% of calories from fat. However,
recommendations have not changed the eating habits of the
majority of Americans. Nearly one third of American adults
are considered obese and perhaps more frightening, a quarter
of American children are overweight (Food Fats and Health,
1991).

Dietary fat and fat synthesized de novo contribute to
total body fat. The liver and adipose tissues are the main
organs involved in fat synthesis and storage. Fats,
endogenous or dietary, can follow different pathways in
these cells. In the liver, triglycerides can be converted

into phospholipids for cell membranes, packaged as VLDL and

 

secreted or oxidized. Adipose tissue is the principal site
for triglyceride storage but also releases fatty acids from
triglycerides when the body requires energy. Fatty acids
derived from the diet or adipose tissue can regulate gene
expression of many genes. For example, polyunsaturated
fatty acids can lower the expression of lipogenic genes,
those involved in synthesizing fatty acids (Jump et al.,
1996).

The focus of my research has been to study the
regulation of genes involved in lipid metabolism,
particularly how PUFA regulate lipogenic gene expression.

It is not well understood how PUFA regulate gene expression.
Using the 814 gene as a model of lipogenic gene expression,
my studies have shown that PUFA can regulate the same gene
in two different tissues, i.e. liver and adipose, through
different mechanisms. However, at the promoter level, there
appears to be a common target of action (Jump et al., 1993;
1996;.1997a).

This dissertation will describe my findings of how
dietary PUFA regulates lipogenic gene expression in cultured
3T3-L1 adipocytes and primary hepatocytes. These studies,
coupled with previous work from our lab (Ren et al., 1996;
1997) reveal the presence of multiple mechanisms in the

liver for PUFA regulation of lipogenic gene expression.

.-‘L...’

 

 

CHAPTER 1

C3UUPEER.1.

LITERATURE REVIEW

Our laboratory focuses on the study of lipid metabolism
and its regulation by hormones and dietary factors. This
literature review will describe lipid metabolism in liver
and white adipose tissue. I will also describe the effect
of dietary fat on these tissues. Finally I will discuss the
regulation of $14, a model for hormonal and dietary
regulation of lipogenic gene expression, by the
polyunsaturated fatty acid, arachidonate, in both adipocytes

and hepatocytes.

Lipid Metabolism

Lipids play an important role in energy storage, cell
membrane structures and the synthesis of important bioactive
compounds. Dietary lipids have also been shown to regulate
gene expression. ‘This section will discuss lipid
metabolism, synthesis, oxidation and the role of dietary
fat.

Fat in the body is derived either from the diet or

synthesized de novo. In humans, dietary fats or

 

 

triglycerides, are broken down into fatty acids as a result
of bile salts, pancreatic lipase and intestinal lipase in
the small intestine. The fatty acids are absorbed into the
intestinal villi and packaged into chylomicrons. The
chylomicrons are transported to the lymph system from the
villi and enter the blood stream (Stryer, 1988; Alberts et
al., 1994; Vance and Vance, 1985). Fatty acids released
from chylomicrons by lipoprotein lipase (LPL) are taken up
by the adipose, muscle or other tissues. The fatty acids
enter adipocytes and are re-esterified into triglycerides
for storage. The chylomicron remnants, containing
cholesterol and apo-proteins, are transported to the liver
and taken up by receptor mediated endocytosis (Cooper,
1997).

The liver packages triglycerides into very low density
lipoproteins (VLDL). Triglycerides in VLDL are synthesized
endogenously or are derived from the diet. Once released,
lipases act on VLDLs, releasing fatty acids. The
triglyceride level decreases as fatty acids are removed,
increasing the density of the particle. These remaining
components are either converted to low density lipoproteins
(LDL) or taken up by the liver. LDL contain cholesterol and
are considered the main transporters of this molecule.
Another transporter of triglycerides are high density

lipoproteins (HDL). These molecules contain a small amount

 

of triglycerides and mainly function to pick up cholesterol
in the plasma for esterification (Stryer, 1988; Alberts et

al., 1994).

Lipid Synthesis

VLDL from the liver contain triglycerides and
phospholipids which are derived from endogenous or dietary
fatty acids. Fatty acids can be synthesized in the liver
and the adipose tissue of humans. However, fatty acid
synthesis occurs predominantly in the liver and only when a
high carbohydrate diet is consumed (Jungermann and
Kietzmann, 1996; Hellerstein et al., 1996). The first step
in fatty acid synthesis requires acetyl CoA and acetyl CoA
carboxylase (ACC) to form malonyl CoA. Fatty acid synthase
(FAS), using acetyl CoA as a primary molecule, sequentially
adds malonyl CoA to form palmitate (16:0) (Stryer, 1988;
Volpe and Vagelos, 1976). Palmitate can be elongated and/or
desaturated to either stearate (18:0) or oleate (18:1).
Oleate can also be elongated but desaturation beyond n-9 is
not possible in mammals due to the lack of appropriate
desaturases. Linoleic and linolenic acid (18:2n-6 and
18:3n-3, respectively) are therefore required in our diets
(see Figure 1.1) (Vance and Vance, 1985). These fatty acids

are used in the production of triglycerides which are used

for the synthesis VLDLs, phosphoglycerides or sphingolipids

(Stryer, 1988; Alberts et al., 1994).

 

ENDOGENOUS DIET
16:0 16:1
1 1
18:0—> 18:1n-9_l 18:2n-6" 18:3n-6 18:3n-3
1 l 18:2n-9 l 1 1
20:0 20:1n-9 l 20:2n-6 20:3n-6 18:4n-3
l l 20:2n-9 1
22:0 22:1n-9 l 20:4n-6 20:4n-3

20:3n-9 l 1

22:4n-6 20:5n-3

“"* = desaturation .l = elongation

 

 

 

Figure 1.1. Fatty Acid Elongation and Desaturation. Fatty
acids can be elongated and desaturated in vivo. This figure
represents the pathways present in mammals. Thicker arrows
indicate more prominent pathways, horizontal arrows indicate
desaturation steps and vertical arrows indicate elongation
steps. This figure was adapted from a figure in Vance and
Vance, 1985.

Enzymes involved in fatty acid synthesis are regulated
at both transcriptional, translational and post-
translational levels. For example, ACC activity is induced
by citrate and suppressed by palmitoyl CoA. Phosphorylation

of ACC by protein kinase A (PKA) following treatment of

cells with epinephrine or glucagon will suppress ACC, while
insulin will induce ACC activity (Stryer, 1988).
Transcription of ACC is also regulated by epinephrine,
glucagon and insulin. Diet also affects these enzymes. For
example, high carbohydrate diets increase ACC and FAS
activity while polyunsaturated fatty acids (PUFA) decrease
activity of these enzymes (Kim and Freake, 1996; Jump et
al., 1996; Hellerstein et al., 1996). Most of the activity
level of FAS is due to protein and RNA levels and is
therefore controlled at the pretranslational level rather
than the post-translational level. The same is true for the
lipogenic gene, S14 (Clarke and Jump, 1993). A more
extensive discussion of the regulation of FAS and 814 will

be presented in a later section.

Lipid Oxidation

Most humans store enough energy in the form of fat to
last a month. When energy is required, these stored
triglycerides are degraded by lipases, releasing fatty acids
and one glycerol molecule. The glycerol can be used for
glucose synthesis but the fatty acids can not (hence sugar
can not be synthesized from fats). Fatty acid oxidation

occurs in the mitochondria of many tissues, including liver,
.heart, muscle and.brain. The major form of oxidation is D—

<xxidation where the chain length is reduced by two carbons

each cycle. The result is the release of acetyl CoA which
is further oxidized in the mitochondria to produce ATP and
C02 (Stryer, 1988; Alberts et al., 1994; Vance and Vance,
1985).

Fatty acids are released by lipases on the cell surface
and must be taken into the mitochondria for oxidation. Once
inside the cell, the fatty acid is converted to a CoA
thioester by acyl CoA synthetases at the outer mitochondrial
membrane. Once through the outer mitochondrial membrane,
the fatty acid thioesters are transported across inner the
mitochondrial membrane to the matrix by carnitine. This
translocation requires carnitine palmitoyltransferase I and
II (CPTI and CPTII), present on the inner and outer side of
the inner mitochondrial membrane, respectively. CPTII
transfers the thioester to carnitine and CPTI removes the
carnitine and the fatty acid thioester is released into the
intermembrane space where oxidation can occur. Thus the
fatty acid thioester is converted to an acylcarnitine and
back to a thioester before oxidation begins (Vance and
Vance, 1985).

Fatty acid oxidation is induced when energy is needed
by the body. Thus during times of hunger, fatty acid
oxidation is increased. The enzymes involved in fatty acid
oxidation, 3-hydroxyacyl CoA dehydrogenase and thiolase, are

inhibited by NADH and acetyl CoA which indicate that energy

is high. Glucagon and epinephrine increase their activities
and transcription while insulin will inhibit this (Stryer,
1988). Fatty acids are synthesized or oxidized upon changes
in body energy status. Thus, the enzymes involved in these
processes are regulated by the energy status in the cells.
A summary figure for lipid metabolism is given in Figure
1.2.

Fatty acids are also oxidized in peroxisomes.
Peroxisomes are organelles which have hydrogen peroxide-
producing flavin oxidases and catalases. These organelles

function to metabolize hydrogen peroxide, but also contain
several lipid oxidative enzymes. Peroxisomal B-oxidation is

increased by peroxisome proliferators such as clofibrate,
gemfibrozil, hypolipidemic agents and high fat diets.
Peroxisomes contain oxidases which oxidize fatty acids and

release heat in contrast to the energy released from

mitochondrial B-oxidation (Reddy and Mannaerts, 1994). One

peroxisomal B-oxidative enzyme, acyl CoA oxidase (AOX), is

regulated at the transcriptional level by peroxisome
proliferators. AOX mRNA levels increase in rats fed a diet
containing fish oil (contains long chain n-3 fatty acid) or
gemfibrozil, a hypolipidemic drug (Ren et al., 1997). Very
long chain fatty acids are almost exclusively oxidized by
peroxisomes. Prostaglandins and bile acids are also

oxidized by the peroxisome (Reddy and Mannaerts, 1994).

 

 

[Dietary fatTW

 

 

 

 

 

 

 

 

 

 

 

l LIVER
)I lipogenesis

Ehylomicrons F a I B-oxidation

l I TG synthesis

mbilization] I VLDL production
A I Repackaging of TG

norms: '
I lipogenesis 1

storage

 

 

I lipolysis

 

A

 

 

 

 

Figure 1.2. Lipid Metabolism. Dietary fat is taken in,
absorbed as chylomicrons and dumped into circulation. The
adipose depot can store TG from chylomicrons while the liver
can pick the remnants or the chylomicrons for repackaging
into VLDL. The liver can also synthesize fatty acids
(lipogenesis) for T6 formation and release.

Dietary’Eat

Dietary fat is required in the diets of humans and
animals. Fats are necessary for energy, steroid and
prostanoid production and cell membrane components. Dietary
fat also improves taste and palatability of food. However,
too much dietary fat can have negative side effects (Food
Fats and Health, 1991).
A Both amount and type of dietary fat are important when
considering health factors. Too much fat is linked to
obesity, diabetes, heart disease and cancer. Perhaps even
more importantly, the type of fat consumed is linked to

these diseases (Food Fats and Health, 1991). Dietary fat is

10

either saturated or unsaturated. Saturated fat, fatty acids
with no double bonds, has been linked to these diseases as
well as insulin resistance and hypertension. In contrast,
fats with more than one double bond, polyunsaturated n-3
fatty acids, are correlated with lower heart disease (Food
Fats and Health, 1991). Gene expression is also affected by
degree of saturation. Polyunsaturated n-3 fats lower VLDL
production from the liver and both n-6 and n-3 PUFA decrease
enzymes involved in fatty acid synthesis while saturated
fats do not (Jump et al., 1997a). This dissertation will
focus on the regulation of gene expression by n-6
polyunsaturated fats. A more extensive review of PUFA

follows.

Polyunsaturated Fatty Acids

Polyunsaturated fatty acids (PUFA) are fatty acids
containing two or more double bonds. Two long chain PUFA
discussed here are eicosapentaenoic acid 20:5n-3 (EPA) and
arachidonic acid 20:4n-6 (AA). AA and EPA are synthesized
by animals from linoleic 18:2n-6 and a-linolenic acid
18:3n-3, respectively. Animals can not synthesize these
precursors because they can not add double bonds past the
ninth carbon on a fatty acid. Linoleic and a-linolenic

acids are therefore essential in animal diets. Most human

11

K4 1" "7".“ in q

diets also include some AA and EPA. For example, fish (or
fish oil) contains n-3 PUFA (Food, Fats and Health, 1991).

Arachidonic acid contains twenty carbons with four
double bonds, an n—6 fatty acid. EPA also has twenty carbons
but has five double bonds and is an n-3 fatty acid (see
Figure 1.3). The difference being one more double bond in
EPA.

PUFAs suppress lipogenic enzyme activity by inhibiting
gene transcription. N-3 PUFA also inhibit triglyceride
synthesis but the mechanism is not known (Clarke and Jump,
1993). Early work with rats fed linoleic and linolenic acid
determined that diets containing these fatty acids could
lower G6PD and FAS activity (Clarke et al., 1976; Flick et
al., 1977; Hodge and Salati, 1997). Since that time,
transcription, RNA stability and enzyme activity of
lipogenic enzymes have been studied to determine how PUFAs
affect these processes. Most lipogenic genes have been
shown to be suppressed at the transcriptional level, rather
than RNA stability or direct inhibition of the enzyme. FAS,
SCD1, pyruvate dehydrogenase, ACC and 814 all have lower
‘expression in the liver as a result of dietary PUFA (Clarke
and Jump, 1994; 1993; Clarke and Abraham, 1992; Clarke et
al., 1990a). In contrast G6PD is suppressed at the post-
transcriptional level as a result of changes in RNA

stability (Hodge and Salati, 1997).

12

 

OOH

 

n-6 n-3
ARACHIDONIC ACID EICOSAPENTAENOIC
ACID

 

 

 

Figure 1.3. Structure of AA and EPA. Each fatty acid is
shown above. The arrows indicate the first carbon double-
bonded from the opposite end of where the numbering system
starts. Hence, AA contains a double bond

on the 5, 8, 11 and 14fl‘carbon.

Fish oil has been extensively studied because it has
been shown to lower serum triglycerides, VLDL secretion and
increase fatty acid oxidation in humans. A diet high in
fish, such as that of Eskimos, is highly correlated to lower
heart disease as well (Food, Fat and Health, 1991; Jump et
al., 1997a). In our lab, fish oil has been used in several
~rat feeding studies. In every instance, rats fed diets
containing fish (menhaden) oil, have lower expression of
$14, FAS and PK mRNAs. Serum triglyceride levels are also
lower compared to olive oil fed animals. In mice, mRNAs
encoding proteins involved in peroxisomal and microsomal

fatty acid oxidation, such as AOX and CYP4A (a P-450 gene),

13

are increased with this dietary regime as compared to the
olive oil fed animals (Ren et al., 1997). These effects on
hepatic gene expression are due to direct effects on
hepatocytes. There is no evidence to indicate that PUFA
regulation of hepatic gene expression requires extraheptic
metabolites or hormones. Adipose tissue also responds to
fish oil feeding leading to a suppression of $14 and FAS
mRNA. However, the fish oil effect on adipose tissue is
slower than in liver. Moreover, the mechanism for this
control is unknown (Jump et al., 1993).

Like EPA, AA suppresses lipogenic gene expression in
both hepatocytes and adipocytes. AA has previously been
shown to lower FAS expression in hepatocytes (Armstrong et

al., 1991; Jump et al., 1994), but not increase the
peroxisomal B-oxidative enzyme AOX (Jump et al., 1997a).

Interestingly, EPA increases in AOX and CYP4 mRNA in vivo
and primary hepatocytes. One proposed mechanism of how EPA
regulates gene expression is through peroxisome proliferator
activated receptors (PPAR), which bind DNA and are

transcription factors (these will be discussed later).
Using the PPARa knockout mouse, Ren et al. (1997) reported
that EPA required PPARa to induce AOX and CYP4 mRNA levels.

In this same report, Ren also showed that while 814 and FAS

were suppressed by EPA, their suppression did not require

PPARa (Ren et al., 1997). These results indicate that PPAR

14

are required for PUFA regulation of some genes (AOX, CYP4A)

but not others (514, FAS). Because AA failed to activate

PPARa (Ren et al., 1997), the AA regulation of hepatic gene

expression does not require PPARa. This thesis will
examine how AA suppresses lipogenic gene expression in
cultured adipocytes and primary hepatocytes. The following
section describes the possible mechanisms for PUFA control

of $14.

Possible Mbchanisms for PUFA Control of Lipogenic Gene
Expression

Efforts to determine how PUFA are suppressing 814 (and
FAS) have led to several possibilities. One possibility is
that peroxisome proliferator activated receptor (PPAR) is
the PUFA regulatory factor (Jump et al., 1996; Clarke and

Jump, 1996). PPAR are a family of transcription factors

including PPARa (predominant in liver), PPARy no

predominant in adipose tissue, 71 ubiquitous) and PPARB
(many tissues) which bind to DNA as heterodimers with
retinoid X receptor (RXR) at peroxisome proliferator
response elements (PPRE) and affect transcription (Braissant
et al., 1996). Activators include fatty acids, leading to

the hypothesis that PUFA are ligands for PPAR, and therefore

can regulate transcription via PPAR. However, PPARa, the

15

most abundant PPAR in liver, does not bind the $14 promoter
although PPARa can interfere with the T3 response by
sequestering RXR (Ren et al., 1996). AOX and CYP4A have
been shown to require PPARa to be regulated by PUFA but
this was not the case with 514 and FAS (Ren et al., 1997;
Ren et al., 1996). The lipogenic gene, stearoyl CoA
desaturase (SCDl), is also suppressed by PUFA but is induced
by the peroxisome proliferator clofibrate (Miller and
Ntambi, 1996). Opposing this theory that PUFA use PPAR to
regulate gene expression is the fact that saturated fats
bind and activate PPAR as well (Gottlicher, et al., 1992).
It is known that PPAR bind several arachidonic acid
metabolites including 15-deoxy-A3L“ PGJg, 8S-HETE, 88-HEPE,
PGI (Forman et al., 1997). In primary hepatocytes

arachidonic acid failed to activate PPARa. These studies

suggest that PPAR may not be involved in the AA control of

lipogenic gene expression (Ren et al., 1997).

A second possibility is that another PPAR such as y or

B/S is involved in the PUFA response (Jump et al., 1996;
'Clarke and Jump, 1996). The peroxisome proliferator WY-
14,643 inhibits $14 mRNA expression and Sl4CAT activity in
primary hepatocytes (Jump et al., 1995; Ren et al., 1996).

Perhaps this suppression is through a PPAR other than

PPARa.

16

A third possibility is that PUFA suppression is through
the production of prostanoids from PUFA (Jump et al., 1996;
Clarke and Jump, 1996). Both AA and EPA can be metabolized
to prostaglandins although much less EPA than AA is
metabolized in this way (Smith, 1989). However, prostanoid
inhibitors did not reverse the inhibition of FAS activity
seen with corn oil treatment (Flick et al., 1977). However,
it is not known if arachidonic acid is acting through a
prostanoid dependent pathway, and this question will be
addressed in this thesis. Arachidonic acid metabolites,
eicosanoids, are abundant and will be discussed in more
detail later in the literature review.

Other possible mechanisms also present potential
modifiers of transcription by PUFA. For example, alteration
of the redox state of the cell could change transcription.
PUFA may change the phosphorylation state of proteins such
as PKA or protein kinase C (PKC), which phosphorylate
transcription factors. PKC is known to be activated by
unsaturated fats, including monounsaturated fats. However,
monounsaturated fatty acids do not suppress lipogenic
'enzymes (Clarke and Jump, 1994). PUFA decrease lipogenesis
and SCDl, leading to a decline in 16:0 and 18:1. Perhaps
this reduces the fatty acid availability for cell membrane
components, changing membrane fluidity (Clarke and Jump,

1996; Clarke and Jump, 1993). However, studies in primary

17

 

hepatocytes do not support a generalized effect on plasma
membrane signaling (Liimatta et al., 1994).

The regulation of $14 by PUFA has been a focus of our
lab. Several hypotheses exist but definitive answers do
not. This dissertation will focus on how arachidonic acid

is suppressing $14 in adipocytes and hepatocytes.

Adipocytes

Because a large portion of this thesis deals with fatty
acid regulation of adipocyte gene suppression, I will now
present a brief overview differentiation and fatty acid
metabolism in adipocytes. Formation of adipocytes begins
with “determination” of stem cells destined to become either
myoblasts, chondroblasts or preadipocytes. Once
determination is complete, preadipocytes are destined to
become adipocytes. After preadipocytes become confluent, or
contact inhibited, and the proper hormonal signals are
present, differentiation begins. During the differentiation
process there are changes in the levels of 300 proteins,
nearly a third of which change in the first five hours
(Sadowski et al., 1992). The cells go on to accumulate
lipids until the adipocyte has multilocular or a unilocular
lipid droplet. In adult animals adipocytes can also be
formed by induction of replication and differentiation of

preadipocytes. Regardless of when the preadipocyte becomes

18

 

an adipocyte, once differentiated, it is an extremely
efficient lipogenic and lipolytic machine (Smas and Sul,
1995; Shillabeer and Len, 1994; Cornelius et al., 1994;
Ailhaud et al., 1991).

Adipocytes function to store energy in the form of
triglycerides in humans and animals. When energy is needed,
fatty acids are released into the circulation where they are
taken up by tissues and oxidized for energy. During times
when energy is plentiful, the adipocytes takes up
triglycerides for storage (Alberts et al., 1994). Adipose
tissue and cultured adipocytes also produce the hormone
leptin. Adipose production of leptin is important in
regulation of satiety and body weight maintenance. In mice,
leptin production is reduced by fasting and then increased
by refeeding. Mice with defective leptin (ob/ob) are
hyperphagic and extremely obese, even when pair fed to their
lean littermates (MacDougald et al., 1995; Coleman, 1978).

Differentiation involves a multitude of factors
including various inducers and inhibitors as well as
expression of specific transcription factors. Though most
of the research presented here is done using only fully
differentiated adipocytes, the many signaling compounds that
trigger adipocyte formation and the transcription factors
expressed in adipocytes also induce lipogenic gene

expression. Each of the following sections, Adipogenic

19

Regulators and Adipogenic Transcription Factors, will
describe the multitude of factors involved in the complexity

of adipocyte differentiation.

Adipogenic Regulators

Regulators of adipocyte differentiation include
hormones, fatty acids, peroxisome proliferators,
prostaglandins, growth factors and other assorted compounds.
Each of these are used in adipogenic cell culture to
increase the percentage of cells which become lipid-filled
adipocytes. Most adipogenic cell lines are cultured and
differentiated in the presence of serum but recently the use
of serum~free medium has necessitated the determination of
which regulators are required for adipocyte differentiation.

The importance of growth hormone (GH) in adipocyte
differentiation was discovered when attempting to remove it
from serum. Without GH, differentiation does not occur in
serum-free cultured preadipocytes (Ailhaud et al., 1992a;
Doglio et al., 1986). GH works through a cell surface
receptor linked to a Janus Kinase (JAK), a tyrosine kinase
which signals to a number of transduction pathways (Alberts
et al., 1994). Research showed that some adipogenic cell
lines require growth hormone for differentiation while
others allow IGF-l substitution. PKC activators can also

substitute for growth hormone to induce differentiation.

20

Another mediator of growth hormone action may be c-Fos
activation. GH may act by activation of c-Fos, PKC or IGF-
1, but all adipogenic cell lines require growth hormone or
IGF-l to develop (Smas and Sul, 1995; Ailhaud et al., 1991).

Thyroid hormone also induces adipocyte differentiation,
although it does not play a main role. For example, in Ob17
cells thyroid hormone is necessary for differentiation when
these cells are cultured in serum-free conditions. However,
cAMP can fulfill the thyroid hormone requirement (Cornelius
et al., 1994; Smas and Sul, 1995).

Glucocorticoids induce differentiation of
preadipocytes to adipocytes. Dexamethasone, a synthetic
glucocorticoid, is routinely used to differentiate the
adipogenic cell lines 3T3-L1 and 3T3-F442A, as well as other
cell lines (Smas and Sul, 1995; Chapman et al., 1985). One
possible mechanism of action by glucocorticoids is the
ability to block synthesis of prostaglandins. Indomethacin,
a potent inhibitor of cyclooxygenase, is used to
differentiate another adipocyte cell line, TAl (Smas and
Sul, 1995). Treatment with cortisone has proven to increase
the breakdown of AA into PGIz. The PGI; can increase cAMP
levels which will increase differentiation of preadipocytes
to adipocytes. This is supported by the fact that
prostaglandins or arachidonic acid can substitute for

glucocorticoids in some cell differentiation media (Ailhaud

21

et al., 1992). Glucocorticoids can also act by directly
increasing gene expression. Glucocorticoids bind DNA via
the glucocorticoid receptor and may increase expression of
genes involved in adipocyte differentiation.
Glucocorticoids also increase the expression of the
lipogenic genes FAS and 814 in adipocytes post
differentiation (Lepar and Jump, 1989; Jump and MacDougald,
1993).

Insulin is one of the most common hormones used to

 

induce differentiation of fat cells. Insulin is used at
high concentrations and is probably acting through the IGF—l
receptor (Cornelius et al., 1994). However, insulin may
also play a role by increasing the glucose transporters
GLUT—l and GLUT-4, which increase expression of other
lipogenic genes (Ailhaud et al., 1992). In adipocytes,
insulin binds to its receptor, causing phosphorylation of
the receptor and other proteins including insulin receptor
substrate (IRS). IRS subsequently a¢tivates Ras which
phosphorylates Raf which phosphorylates MAP kinase. Through
this pathway, insulin increases glucose storage and
lipogenesis in adipocytes. If the activation of Ras by IRS-
1 is inhibited, 3T3-L1 cells will not differentiate,
indicating the requirement of the insulin signaling pathway
for differentiation (Porras and Santos, 1996; Lamphere et

al., 1994; Yenush and White, 1997).

22

Fatty acids can also increase adipocyte
differentiation. In Ob17 and Ob1771 cells, arachidonic acid
(AA) is an inducer of differentiation, acting through an
increase in cAMP and diacylglycerol (Smas and Sul, 1995;
Gaillard et al., 1989). However, in other cell lines, AA
is an inhibitor of differentiation (Smas and Sul, 1995).
Long term treatment of Ob1771 cells (an adipogenic cell
line) with palmitate led to differentiation. Furthermore, a
nonmetabolizeable form of palmitate could repeat this
finding (Ailhaud et al., 1995). Shillabeer and Lau (1994)
reported that the saturated fats were better inducers of
adipocyte formation than polyunsaturated fatty acids in rats
fed different diets. Grimaldi et al. (1992) used palmitate
to induce adipogenesis as well as lipogenic genes. These
data indicate that fatty acids, especially saturated fats,
in the diet can increase fat cell numbers and
differentiation.

Some peroxisome proliferators have recently been shown
to induce differentiation of adipocytes. Peroxisome
proliferators include a variety of chemical compounds
including plasticizers, thiazolidinediones, fatty acids,
steroids and xenobiotics. Thiazolidinediones are
particularly active on preadipocytes. Treatment of
preadipocytes with these compounds can quicken the

transition into lipid-filled adipocytes (Chawla and Lazar,

23

1994; Chawla et al., 1994; Tontonoz et al., 1994b; 1994c).
The actions of these compounds will be discussed further in
the next section.

Prostaglandins both inhibit and induce adipocyte
differentiation depending on the cell line and the
prostaglandin tested. The three prostaglandins produced in

adipose tissue and adipocytes (in vitro) are PGE2, PGI; and
small amounts of PGng(Apbert et al., 1996). These PGs are

synthesized from arachidonic acid by cyclooxygenase. PGI;
(or prostacyclin) is reported to act via an increase in cAMP

and is thought to be an inducer of differentiation in Ob17

and Ob1771 cells. PGE2 and PGng do not act via an increase

in cAMP but PGFyz is known to induce differentiation through
the stimulation of PKC in Ob17 cells (Ailhaud et al.,1991;
Smas and Sul, 1995). However, in 3T3-L1 cells, PGFfl1.iS an
inhibitor of differentiation, as is PGE2. As mentioned
above, the cyclooxygenase inhibitor indomethacin, induces
differentiation in the adipogenic cell line TAl. It is
unknown if blocking prostaglandin synthesis is the reason
that indomethacin induces differentiation in these cells
(Smas and Sul, 1995; Shillabeer et al., 1996). Recently
Lehmann et al., (1997) found that indomethacin itself binds
PPARy. Perhaps it is this pathway that indomethacin is

using to induce differentiation of adipocytes.

24

Prostaglandins may bind PPARy, which is known to enhance

differentiation. One prostaglandin, lS-deoxy-A””“
prostaglandin J2, has been identified as a ligand for
PPARyZ, although this ligand has never been detected in vivo
(Forman et al., 1995; Kliewer et al., 1995). While
prostaglandins can play a role in differentiation, it is
unknown exactly how they are involved in the process.

Growth factors generally inhibit differentiation of
adipocytes, with the exception of IGF-1. Epidermal growth
factor (EGF), platelet derived growth factor (PDGF),
transforming growth factor (TGF) and fibroblast growth
factor (FGF) all inhibit fat development either in rats or
in adipogenic cell lines (Smas and Sul, 1995; Cornelius et
al., 1994). Growth factors activate mitosis (cell
division), not differentiation. These mitogenic factors
will not cause differentiation even once the cells arev
confluent. Overexpression of c-myc, a mitogenic signaling
molecule, also prevents differentiation (Cornelius et al.,
1994).

IGF-l is the only growth factor required for
differentiation. It can also substitute for GH in some cell
lines requiring GH in serum-free differentiation media.
IGF-l acts at the cell surface through its receptor which is
linked to a tyrosine kinase signaling cascade including the

GTP-activated ras. This is further substantiated by the

25

fact that cells transfected with H-ras can be differentiated
without GH or IGF-l (Smas and Sul, 1995; Cornelius et al.,

1994).

Other compounds are important in blocking adipocyte
differentiation. Tumor necrosis factor a (TNFa) and other
cytokines are inhibitors of differentiation. Even in
combination with normal differentiation inducers, TNFa will

not allow differentiation and actually depresses the
expression of several lipogenic genes in adipocytes (Smas

and Sul, 1995; Hauner et al., 1995; Ninomiya-Tsuji et al.,
1993). TNFa also interferes with the insulin receptor
substrate IRS-1 and may be a cause for insulin resistance.
By blocking phosphorylation of IRS-1, TNFa blocks
lipogenesis and glucose utilization initiated by insulin

(Hotamisgil et al., 1996b; Yenush and White, 1997). TNFa
has also been shown to reduce C/EBP and PPARy expression in

treated adipocytes. C/EBP and PPARy, which will be

discussed in the next section, are inducers of adipocyte
differentiation. Each is involved in maintaining the
adipogenic state of fat cells (Zhang et al., 1996; Stephens
and Pekala, 1992). Retinoic acid also inhibits
differentiation, causing de-differentiation or cell death.
Retinoic acud is thought to suppress the expression of the

adipogenic transcription factor C/EBP (Schwarz et al., 1997;

26

Cornelius et al., 1994). Other differentiation inhibitors
include phorbol esters, dimethylsulfoxide (DMSO) and
endothelin-l (Cornelius et al., 1994). [TNF will be further
discussed later.]

In summary, adipocyte differentiation has many
regulators. Over all, there appears to be a requirement for
an increased cAMP levels, a glucocorticoid stimulation and
activation of the insulin/IGF-l tyrosine kinase pathway in
order to induce differentiation (Cornelius et al., 1994;
Smas and Sul, 1995; Ailhaud et al., 1991). As will be
discussed below, most of these differentiation regulators

also affect the expression of $14.

Adipogenic Transcription Factors
The adipocyte transcription factors involved in

adipocyte differentiation have been described. They include
PPARy, the C/EBP family and SREBP/ADDl. These function to

induce transcription of other genes involved in
differentiation and lipogenesis. Each of these have

putative binding sites in the $14 promoter as well.
Peroxisome proliferator activated receptors (PPARy) are
a part of the nuclear hormone receptor family which includes

PPARyl PPARyZ, PPARa and PPARB (NUCl, FAAR, PPARS are all

homologs to PPARB). PPARyl and 2 differ only in the 5'

27

untranslated region. PPARyZ mRNA is slightly longer at 2.1
kb vs. 1.8 kb for PPARyl in rodents. The most important

PPAR in adipocyte differentiation is PPARyZ. PPARyZ is

specifically expressed in high levels in adipocytes and at
much lower levels in other tissues (Tontonoz et al., 1994b;

Braissant et al., 1996; Schoonjans et al., 1996b; 1996c).
PPARyZ can be also be phosphorylated as a result of insulin
or MAP kinase activation, perhaps indicating some crosstalk
between insulin and PPARy in adipocyte differentiation

(Zhang et al., 1996; Adams et al., 1997; Camp and Tafuri,

1997; Hu et al., 1996a; Reginato et al., 1998).

Endogenous PPARyZ is expressed in 3T3-L1 and 3T3-F442A
cells within 1-2 days post confluence of preadipocytes
(Tontonoz et al., 1994b) but PPARy expression alone is often
enough to induce adipocyte formation. Overexpression of
PPARyZ can cause differentiation in fibroblasts (Tontonoz et

al., 1994c). Overexpression of FAAR in conjunction with

fatty acids will convert muscle cells into lipid containing
cells (Teboul et al., 1995; Hu et al., 1995). The PPARyZ

ligand, pioglitazone, used with PPARyZ expression is an even
stronger inducer of differentiation in non-adipogenic cells
(Brun et al., 1996). Clearly, PPARy plays a leading role in

adipocyte differentiation.

28

 

Activators of PPARs include long chain fatty acids and
their metabolites, plasticizers and hypolipidemic drugs.
Although considered orphan receptors only a few years ago,
several ligands have now been determined for each of the
PPARs. PPARyZ ligands include 15-deoxy-A12,14 prostaglandin
J2 (15-PGJ2) and the thiazolidinediones (Kliewer et al.,

1995; Forman et al., 1995; Lehmann et al., 1995).

Leukotriene B4, 8(S) HETE and carbaprostacyclin bind PPARa

and carbaprostacyclin also binds PPARB. It is interesting
to note that all of the naturally occurring ligands are
derivatives of arachidonic acid (Mandrup and Lane, 1997;
Brun et al., 1996; Spiegelman and Flier, 1996; Yu et al.,
1995).

Indomethacin and ibuprofen have also been shown to be
PPARy ligands. This further substantiates the finding that
indomethacin activates differentiation in TAl adipocytes,
perhaps by acting through PPARy rather than by blocking
prostaglandin production. Ibuprofen was shown to increase
activity of PPARa as well. The authors concluded that

adipocyte differentiation and peroxisome proliferation may
be initiated by common pathways (Lehmann et al., 1997).
Because PPARs are activated by fatty acids, it has been

suggested that they are the mediators of PUFA regulation of

genes. Fatty acids and peroxisome proliferators increase B-

29

oxidation of fatty acids. For example, both the peroxisome

proliferator, WY-14,643 and eicosatetraynoic acid (20:4,n-6
triple bonds), activate PPARa and the peroxisomal enzyme,
acyl CoA oxidase (AOX) (Keller et al., 1993). AOX contains
an PPRE and binds PPARa (Varanasi et al., 1996). Another

gene regulated by fatty acids and peroxisome proliferators
is apolipoprotein A-II. The authors here suggest that fatty
acids and peroxisome proliferators are both acting to
modulate gene transcription at through the PPRE (Vu-Dac et
al., 1995). These findings suggest that the two regulators,
fatty acids and peroxisome proliferators, may use the same
pathway to influence gene transcription (Keller et al.,
1993; Kliewer et al., 1997).

Some of the thiazolidinediones are now being used to
increase insulin sensitivity in non-insulin dependent
diabetes (NIDDM) patients (Brun et al., 1996). Pioglitazone
and other thiazolidinediones can lower plasma insulin,
triglycerides and glucose in NIDDM humans and animals. In
the obese Zucker (fa/fa) rat, pioglitazone increased the use
of glucose and insulin sensitivity. However, the treatment
also increased adipocyte differentiation and expression of

specific lipogenic genes: fatty acid synthase (FAS),
phosphoenolpyruvate carboxykinase (PEPCK), C/EBPa, GLUT4

and decreases leptin expression. The increase in adipocytes

is not a desirable side effect in the treatment of NIDDM

30

patients, who already tend to be obese. However, humans in
clinical trials with the thiazolidinedione troglitazone have

not yet shown weight gain (Hallakou et al., 1997).
PPAPw2 is known to activate several lipogenic genes

including the adipocyte gene (aP2), malic enzyme (ME),

lipoprotein lipase (LPL) and PEPCK as well as some B-

oxidative genes (AOX) and P450 cytOchrome genes (P450 4A6)
(Hunter et al., 1996; Castelein et al., 1994). Like the
other PPARs, PPARyZ binds to DNA at its peroxisome
proliferator response element (PPRE), a DR-l, as a
heterodimer with retinoid X receptor (RXR). Usually the
PPRE is located far upstream of the transcription start site
and acts as an enhancer of transcription (Bernlohr et al.,
1985; Amri et al., 1991; Tontonoz et al., 1994a; Graves et
al., 1992; Spiegelman and Flier, 1996).

Several genes contain PPRE sites. The aP2 gene is

often used as an early marker of adipocyte differentiation.
This gene has two binding sites for PPARyZ, its key

regulator and inducer (Tontonoz et al., 1994a). LPL was
known to be activated in adipocytes and liver by treatment
with peroxisome proliferators. Later a PPRE was found in
the LPL promoter by gel shift assays (Schoonjans et al.,
1996a). Similarly, the acyl-coenzyme A synthetase gene was
found to contain a PPRE and be activated by peroxisome

proliferators (Schoonjans et al., 1995). Kallen and Lazar

31

(1996) and De Vos et al. (1996) reported that leptin was
down regulated by thiazolidinediones. These findings
indicate that PPAR can both increase and decrease
transcription of target genes.

Another adipogenic transcription factor is the CCAAT
enhancer binding protein (C/EBP) family. C/EBPa, B and 5

are leucine zipper transcription factors each involved in
differentiation. This transcription factor tends to bind
near the proximal promoter in target genes. The members of
this family have basic helix-loop-helix binding domains and

homo- or heterodimerize with other family members through a

leucine zipper region. Both B and 6 are expressed early,

then a is produced during the last part of differentiation.

Inhibition of C/EBP expression in preadipocytes but not in
adipocytes may play a role in regulation when the cells are
allowed to differentiate (Brun et al., 1996; Cornelius et
al., 1994; Spiegelman and Flier, 1996; Mandrup and Lane,
1997; Jiang et al., 1998).

Like PPARy, overexpression of C/EBP can induce
differentiation of nonadipogenic cells (Wu et al., 1996;
Freytag et al., 1994) or preadipocytes (Lin and Lane, 1994;
Christy et al., 1991). Furthermore, antisense C/EBP
expression in 3T3-L1 cells prevented expression of aPZ,
GLUT4 and stearoyl CoA desaturase 1 (SCDl) as well as

inhibited triglyceride accumulation (Lin and Lane, 1992).

32

In C/EBPa knockout mice, the adipocytes do not contain as

much lipid and have lower lipogenic gene expression than
wildtype mice (Cornelius et al., 1994; Mandrup and Lane,
1997; Darlington et al., 1995). These findings indicate
that the C/EBP family plays a key role in adipocyte
differentiation.

Many variants of the C/EBP family exist and have been
described. C/EBPB is also known as liver activator protein

(LAP) and its alternative splice product is liver inhibitory

protein (LIP). LIP, which does not have a trans-activation

domain, can block activation caused by LAP. C/EBPa also
has two isoforms, a 30kDa protein (p30C/EBPa) and a 42kDA

protein (p42C/EBPa). These are different products of a

single gene. The p42 but not p30 inhibits mitosis in 3T3-L1

adipocytes but both activate gene transcription of aP2
(Cornelius et al., 1994). C/EBPB and 6 appear to be
expressed first and transiently in preadipocytes treated

with the hormonal inducers; then C/EBPa expression occurs
and growth is arrested, initiating differentiation. C/EBPB

expression requires cAMP while 8 requires glucocorticoids,

two of the requirements for differentiation (Mandrup and

Lane, 1997).
C/EBPa interacts with other gene promoters as well.

Miller et al. (1996b) and Hwang et al. (1996) determined

33

that the leptin gene contains a C/EBPa binding site, and

mutation of this site removed activation by this
transcription factor. GLUT4 also contains a C/EBP binding
site, which allows for transactivation of this gene
(Kaestner et al., 1990). Both aP2 and SCD have promoter
binding sites for C/EBP as well (Christy et al., 1989;
Herrera et al., 1989).

Even more confusing is the interaction between the two

transcription factors, C/EBP and PPARyZ. C/EBPB and 8
induce expression of C/EBPa, which can also up—regulate
itself. C/EBPB induces PPARy which, in turn, can induce

C/EPBa. Both C/EBPa and PPARy induce the expression of two

lipogenic genes already mentioned: aP2 and PEPCK (Tontonoz
et al., 1994b; 1995; Cheneval et al., 1991). The leptin
gene promoter also contains sites for each of these
transcription factors but the factors themselves seem to
antagonize one another, perhaps explaining why
thiazolidinediones depress leptin expression (Hollenberg et
al., 1997; Kallen and Lazar, 1996). The interactions
between these transcription factors make the process of
adipocyte differentiation and gene expression extremely
complex.

A third transcription factor involved in adipocyte

differentiation is adipocyte differentiation and

34

 

determination factor 1 (ADDl), also know as sterol response
element binding protein 1 (SREBP-1). This protein also
contains a basic helix-loop-helix domain like the C/EBP
family but binds to “E-boxes” or CANNTG sequences as a
dimer. This transcription factor was originally found to
play a role in cholesterol regulation. However, ADDl has
been shown to play a role in adipocyte differentiation (Kim
and Spiegelman, 1996). Specifically, overexpression of ADDl
under nonadipogenic conditions will induce the expression of

the lipogenic genes LPL and FAS. These authors also showed
that ADDl expression will increase the activity of PPARy.

These results indicate that ADDl may play a role in

differentiation itself and/or through another of the
transcription factors, PPARy. Another indication of the

role SREBP plays in adipocytes is its binding to the
lipogenic gene, FAS. Both Bennet et al. (1995) and Tontonoz
et al. (1993) showed that the FAS promoter binds SREBP, and
with another transcription factor Spl, could increase
activity in a FAS promoter-luciferase reporter system. 314
also contains an E box in its carbohydrate response region
(-1457 to -1428 bp). This E box binds ADDl and
transcription is activated in NIH 3T3 cells transfected with
a 514 reporter vector (Kim et al., 1995).

The three major transcriptional regulators of adipocyte

differentiation are PPARy, C/EBP and ADDl. All can increase

35

 

differentiation of preadipocytes into adipocytes directly or
indirectly as well as influence the expression or activity
of each other and adipogenic genes. Spiegelman and Flier

(1996) describe these interactions as shown in Figure 1.4.

Lipogenic Gene Expression in Adipocytes

Once differentiation has begun in adipocytes, several
lipogenic and lipolytic genes are expressed. An early
marker of differentiation is lipoprotein lipase (LPL).
Other markers include aP2, adipsin, glycerol-B-phosphate
dehydrogenase (GPD) and PEPCK. Some of the genes with
increased expression include acetyl CoA carboxylase (ACC),
malic enzyme (ME), fatty acid synthase (FAS), ATP-citrate

lyase, $14, GLUT4, hormone sensitive lipase (HSL), stearoyl-
CoA desaturase, angiotensinogen, insulin and B adrenergic

receptors and glucocorticoid receptors (Butterwith, 1994;
Mackall et al., 1976; Cornelius et al., 1994). All of these
genes appear after differentiation of adipocytes (Lepar and
Jump, 1989; Cornelius et al., 1994).

In general, lipogenic gene expression increases with
insulin, glucocorticoid or carbohydrate treatment and
decreases with polyunsaturated fatty acid treatment and
agents which cause an increase in cAMP in both liver and
adipose tissue. For example, PUFA, but not saturated fat

diets, lower expression of liver stearoyl CoA desaturase

36

(SCD) in both lean and obese rats (Jones et al., 1996) and
in mice and hepatocytes (Landschulz et al., 1994). A
discussion of the control fatty acid synthase and $14 in
liver and adipose tissue is included at the end of this

chapter.

 

 

¢—— PPARy/RXR _DD1

“\CI/QQ

C/EBPBor6
O O [Adipocyte Differentiation]

l

Lipogenic
Gene
Expression

 

 

 

 

 

 

Fatty Acid
Metabolism

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 1.4. Adipocyte Transcription Factors. Initially
C/EBPB and/or 6 is induced by hormonal changes which give
rise to expression of PPARy,.ADD1 and C/EBPa. C/EBPa may

allow for the continued presence of PPARy, both of which are
thought to be important for adipocyte differentiation
(Spiegelman and Flier, 1996).

Arachidonic Acid
While both EPA and AA are polyunsaturated fats and
suppress lipogenic gene expression, AA also plays a

significant role as a precursor for eicosanoids (all AA

37

metabolites are termed eicosanoids). Very little AA is free
within the cytosoll Most AA in the body is present in cell
membrane phospholipids in the sn-2 position. With the
proper signals, phospholipase Ag cleaves the ester linkage,
and AA is released into the cell. It can then be
metabolized by one of the three major enzyme groups:
epoxygenases, lipoxygenases and cylcooxygenases (Irvine,
1982; Makita et al., 1996; Clark et al.,l995; Camandola et
al., 1996). Figure 1.5 describes the metabolism of AA.

The metabolites produced by epoxygenases include
hydroxyeicosatetraenoic acids (HETEs) and
epoxyeicosatrienoic acids (EETs). These compounds are
produced from a cytochrome P-450 pathway which functions to
monooxygenate arachidonic acid. Some of the epoxy products
are involved in vasodilation/contraction and regulation of
ion flux. For example, 14,15-EET has been shown to cause an
increase of calcium output from liver cells while 5,6-EET
increases intracellular calcium in liver. Inhibitors of
epoxygenases have also been reported to block the calcium
release induced by arachidonic acid. However, the
incredible lack of stability of these compounds makes
studying their effects difficult (Makita et al., 1996;
Fitzpatrick and Murphy, 1989; Graber et al., 1997; Oliw,

'1994).

38

 

Arachidonic Acid

Epoxygen:E§/////' Cyclooxygena;;\\\\\Lipoxygenase

Epoxyeicosatrienoic PGHz Hydroperoxyeicosa-
acids 1 tetranoic acids
PGDz, PGE2, PGI; |_4:HETES
EETs PGan, TxAz diHETES
Leukotrienes
HETEs Prostaglandins
Thromboxanes

 

 

Figure 1.5. Arachidonic Acid Metabolism. Arachidonic acid
can be metabolized in three separate pathways: epoxygenase,
cyclooxygenase and lipoxygenase to result in several
metabolites listed here.

The cytochrome P4505 (CYP) are microsome associated
enzymes that oxidize hydrophobic compounds. There are
several families of CYP and some, e.g. 2C, 2E and 4A,
monooxygenate fatty acids. For example, CYP4A oxidizes m-l
mono and polyunsaturated fatty acids in liver microsomes
(Fitzpatrick and Murphy, 1989; Oliw, 1994). Peroxisome
proliferators increase the activity of some of these enzymes
in the liver and some are regulated by PPAR (Demoz et al.,
1994). Some metabolites produced in the liver include 20-
HETE, 19—HETE, 18-HETE, 17-HETE and 16-HETE by CYP4A.
CYP2E1 produces 19R-HETE, 19S-HETE and 18R-HETE. Of these,
20-HETE has been found in urine and 19-HETE has been proven

to have biological activity. Some epoxides can be used in

39

 

phospholipid synthesis while others are metabolized into

epoxy prostaglandins by PGH synthase in the liver (Oliw,

1994). Unlike liver, I found no evidence for epoxygenase
products in the adipocyte or in adipose tissue.

The lipoxygenase pathway produces leukotrienes, HETEs
and HPETEs. Lipoxygenase is stimulated by calcium in
activated neutrophils, monocytes and macrophages. The
leukotrienes produced include LTAq, LTBq, LTC4iand LTD4
depending on the cell type involved. Leukotrienes act
mostly to induce chemotaxis, chemokinesis and aggregation of
cells. LTB. in particular, is proinflammatory. The
involvement of leukotrienes and their connection to asthma
is a focus of much research (Christie et al., 1993). In
another study, dietary fish oil depressed cyclooxygenase
product formation but increased lipoxygenase product levels.
The significance of more lipoxygenase products produced in
these rats is not known (Hwang et a1, 1988).

In the rat liver, liver fatty acid binding protein (L-
FABP) binds lipoxygenase metabolites. Although L-FABP binds
fatty acids like AA and CA, PGE1, PGE2 and leukotriene B4;
15-HPETE, 5-HETE and 15-HETE all bind with greater affinity
(Raza et al., 1989). L-FABP binding of these metabolites
may reduce their rate of further oxidation. It may also
function to carry a specific fatty acid, such as arachidonic

acid, to the nucleus to regulate transcription (Ek et al.,

40

 

1997). In liver, FABP has also been shown to bind PGA and
PGJ} with greater affinity than arachidonic acid (Khan and
Sorof, 1990). In the adipocyte, the fatty acid binding
protein is similar to L-FABP: adipocyte lipid binding
protein (ALBP or aP2). ALBP has a higher affinity for
oleoyl CoA than L-FABP in the liver. However, both bind
long chain fatty acids very well and play a role in fatty
acid metabolism (Gossett et al., 1996).

Prostaglandins and thromboxanes are produced from the
cyclooxygenase metabolism of arachidonic acid. EPA can also
be converted to prostaglandins (3-series), however, AA is a
much preferred substrate of the cyclooxygenase enzyme even
with equal release of the two fatty acids. EPA can also act
to suppress AA metabolism in platelets (Belury et al., 1989;
Hwang et al., 1988). One reason for this suppression is due
to the fact that EPA is a poor substrate for the
constitutively produced cyclooxygenase in most tissues,
which much prefers arachidonic acid (Laneuville et al.,
1995). AA is also more mitogenic than EPA. In studies with
Swiss 3T3 cells, AA was shown to activate the expression of
c-fos and Egr-l, both early growth gene products, whereas
EPA did not. This study also determined that the mitogenic
response of AA was due to PGE2 production followed by PKC
activation (Sellmayer et al., 1996). Further discussion of

the cyclooxygenase products of AA will follow this section.

41

 

Arachidonic acid can induce other signaling pathways
without being metabolized. Wolf et al. (1986) showed that
in permeabolized islet cells, AA caused an immediate release
of calcium similar to levels induced by IP3. In a similar
experiment with neutrophils, calcium was also released upon
treatment with arachidonic acid (Beaumier et al., 1987).
Blobe et al. (1995) hypothesized that arachidonic acid g
activates PKC if calcium was present. Using rat liver

microsomes, Chan and Turk (1987) showed that arachidonic

 

acid could cause an intracellular increase in calcium which
could not be blocked by lipoxygenase or cyclooxygenase
inhibitors. These data indicate that arachidonic acid
itself may be involved in cell signaling, possibly via a
calcium mediated process.

Other examples of direct action of AA on gene
expression include GLUT4 and SCDl. AA suppresses GLUT4 in
adipocytes treated with 50uM AA for forty-eight hours
(Tebbey et al., 1994). In adipocytes treated with
cyclooxygenase, epoxygenase or lipoxygenase inhibitors, AA
still caused the RNA stability of the SCDl gene to be
depressed while oleate and stearate had no effect on
expression (Sessler et al., 1996).

Arachidonic acid is also involved in the inflammatory
response. AA can block transcription of inflammatory genes

such as IL—8, E-selectin and ICAM-1, however, some AA

42

metabolites are pro-inflammatory increasing cytokine and
coagulant levels (Stuhlmeier et al., 1997). In humans who
voluntarily consumed a lower n-6:n-3 fatty acid diet, the
cytokines involved in inflammation, IL-1 and TNFa, are
decreased (Camandola et al., 1996).

Arachidonic acid, cytokines and peroxisome

proliferators have been shown to act by increasing the

activity of NF-KB. This transcription factor binds to and

 

activates transcription of several genes involved in
inflammation. Gel shift binding of NF-KB increased with

AA, was reversed by AA + indomethacin and could not be
induced by EPA. For example, when the promonocytic cell

line U937 is treated with arachidonic acid, the
transcription factor NF-KB is stimulated. Further

experiments indicated that PGE2(a metabolite of AA, the

synthesis of which can be blocked by indomethacin) could

also increase NF-KB binding (Camandola et al., 1996).

Leung et al. (1996) reported NF-KB is activated in rat liver

upon treatment with the peroxisome proliferator

iciprofibrate.
NF-KB must move from the cytoplasm to the nucleus to
induce transcription of its target genes. In the cytoplasm,

NF-KB is bound by IKB. Phosphorylation of IKB cause the

release of NF—KB, which in turn increases transcription of

the target genes after nuclear translocation (see Figure
1.5). TNFa causes phosphorylation of 1x8, for example
(Alberts et al., 1994). AA can inhibit the phosphorylation
of IKB and hence, AA inhibits transcription of some

inflammatory genes, suppressing the immune system

(Stuhlmeier et al., 1997). Glucocorticoids also inhibit NF-
KB activation by increasing IKB, acting as an immune

suppressant (McKay and Cidlowski, 1998). At the same time,

most AA metabolites, such as PG, are proinflammatory,
possibly by increasing NF-KB action (Stuhlmeier et al.,

1997). These results imply that AA action can depend on its
metabolism and on which metabolic pathway is most active in
the cell.

The production of eicosanoids from arachidonic acid (n-
6) is suppressed by EPA, an n-3 fatty acid (Kinsella et al.,
1990). Overproduction of eicosanoids can lead to immune
suppression and high levels of inflammation, as discussed
above. Hence a diet containing the correct ratio of n-6:n-3
is very important for body homeostasis. Although some n-6
fatty acid is required for eicosanoid synthesis, n-3 long
chain PUFAs can substitute for membrane phospholipid needs.
N-3 fatty acids have also been shown to decrease eicosanoid
production and inhibit conversion of linoleate to
arachidonate (Hwang et al., 1988). Releasing too much

arachidonic acid from cell membranes may lead to

44

 

overproduction of eicosanoids. This suggests that
consumption of more n-3 fatty acids may help off set this
possibility (Kinsella et al., 1990). However, it is not
well understood how dietary fat levels relate to membrane
phospholipid. Currently, an equal balance of n—6:n-3 fatty
acids in the diet is recommended (Kinsella et al., 1990;

Food Fats and Health, 1991).

Prostaglandins

As indicated above, one possible pathway of PUFA
suppression of $14 is through a prostanoid regulatory
mechanism. Prostanoids are produced in both the liver and
adipose tissue from fatty acids. The most common precursor
is arachidonic acid, and therefore, the 2-series prostanoids
most prevalent (Coleman et al., 1994). These prostaglandins
have diverse effects in their target tissues. The actions
of prostaglandins in adipose tissue and liver are focused on

after a general description of prostaglandin receptors.

Prostaglandin Receptors
Before discussing prostaglandin receptors, below is a
list of definitions which are helpful for the remainder of

this discussion.

G-Proteins = are trimeric membrane proteins activated by a
ligand binding to its receptor. Upon activation the alpha

 

45

 

subunit can further activate adenylate cyclase (G3) or
phospholipase C (Gmm) or activation of phosphodiesterase
Uh) or other pathways. Prostaglandin receptors are linked
to G-proteins.

Gs - activates adenylate cyclase, increasing cAMP and
subsequent activation of protein kinase A.

Gi - activates a phosphodiesterase, decreasing cAMP and
inhibition of protein kinase A.

Gmb - activates phospholipase C, causing IP3 and
diacylglycerol (DAG) formation. IP3 increases

intracellular calcium while DAG with the calcium
activates protein kinase C.

Pertussis toxin = acts by ADP-ribosylating the alpha subunit
of G proteins. This disconnects the G protein from its
receptor. It has been used to study prostaglandin signaling
pathways. Reversal of PG action by PT indicates a Giior Go
pathway (inhibition of cAMP or activation of phospholipase
C).

Prostaglandins act on cell surface G-protein linked

receptors. PGE2 has four receptors:EP1, EP2, EP3 and EP4
while PGan binds to the FP receptor. EP2 and EP4 receptors

activate a stprotein and increase cAMP. EPl is linked to a
pertussis toxin-insensitive G protein which activates
phospholipase C, PKC and intracellular calcium release. The
EP3 receptor is both sensitive or insensitive to inhibition
by pertussis toxin and depending on the subtype, linked to
at least three separate pathways. There are four splice
variants of the EP3 receptor in the cow: a, b, c and d; each
of which have different actions. The A subtype of EP3
inhibits adenylate cyclase, B and C induce cAMP and D can

activate Gq, Gi, G3 to increase turnover phosphoinositides

46

 

3»

"pm

5‘

O A

it

'fl

I \
g:

If

Pd

Rib

CE

‘b

and calcium levels. EP3-d is similar to the EPl receptor
(Coleman et al., 1994; Asboth et al., 1996).

EP receptors have a variety of actions in many
different cell types depending on which subtype is present.
EP3 subtypes are also present in mice, humans and rabbits
and like the cow, each subtype has different actions. In
the kidney, EP3 receptors can be inhibited by pertussis
toxin to block the decrease of cAMP formation. The EP3
receptor linked to Gq is insensitive to pertussis toxin in
bovine adrenal glands. EP3 receptors are also present in
adipocytes where they function to inhibit lipolysis by
blocking the epinephrine increase of cAMP (Coleman et al.,
1994; Negishi et al., 1995). Bone cells use EPl and EP4
receptors to increase calcium and cAMP by PGE2, respectively
(Suda et al., 1996). Myometrial cells contain all four EP
receptors but EPl is sensitive to pertussis toxin and EP3 is
not, though both increase intracellular calcium (Asboth et
al., 1996). In summary, PGE2 increases cAMP, decreases cAMP
or activates phospholipase C to increase intracellular
calcium depending on the receptor present on the target

cell.
The other prostaglandins, PGI; and PGng, also have

cell surface receptors. PGI; binds the IP receptor, which
is found in vascular smooth muscle tissue, nerve tissue,

thymus, platelets and lung. IP receptors are linked to both

47

 

adenylate cyclase activation via Gsiand intracellular
calcium increase (Coleman et al., 1994). In WAT, PGI; was
shown to activate G, and increase cAMP coupled to release of

calcium but also may act through a PPAR system (Aubert et

al., 1996). PGI; has also been shown to act through PPARa

(Hertz et al., 1996) . PGan binds FP receptors which act by

increasing intracellular calcium. The Gq protein is thought

to be involved (insensitive to pertussis toxin here) and

 

—-'al‘

causes a concomitant increase in IP3*with calcium (Coleman
et al., 1994). In preadipocytes, PGFfixinhibits
differentiation. This inhibition was reversible when cells

were also treated with a calcium calmodulin-dependent

protein kinase blocker, KN-62 (Miller et al., 1996a). Lepak

and Serrero (1995) reported that PGFy: could induce

transcription of TGFa which prevented differentiation.

These examples demonstrate the diversity and complexity of
prostaglandin action.

Liver is known to have PG receptors, although how many
types are present is not known. Hashimoto et al. (1997)
reported that EP3 receptors were present in hepatocytes
although in mice liver EP3 receptors were not detected by
Sugimoto et al. (1992). It is also likely that because PGE2
has been shown to increase or decrease cAMP levels in both

adipocytes and hepatocytes that these cells contain EP2 or

48

EP4 receptors (Watanabe et al., 1986). Down regulation of
the PGE2 receptor involved in activation of cAMP has also

been shown in rat liver (Robertson et al., 1980).

Actions of Prostaglandins
Primarily PGE2 and PGI; are produced in adipose tissue
of humans and rats. PGE2 and PGI; act to control blood flow

and lipolysis in adipose tissue (Vassaux et al., 1992). In

 

cell culture, adipocytes also produce prostaglandins, l
although at a much lower level than preadipocytes (Hyman et
al., 1982; Vassaux et al., 1992; Richelsen et al., 1992).

The PGE series is usually referred to as
“antilipolytic” in adipose tissue. PGE2 acts primarily to
decrease cAMP, countering the glucagon effect (Christ and
Nugteren, 1970; Castan et al., 1994; Watanabe et al., 1986;
Richelsen, 1987). Antilipolytic effects by PGE2 in adipose
tissue is thought to be through EP3 receptors (Negishi et
al., 1995; Strong et al., 1992).

.Acting opposite of PGE2, PGIz increases adenylate
cyclase activity and is considered lipolytic (Aubert et al.,
1996). Insulin can inhibit both PGE2 and PGI; release in
adipose tissue, perhaps relating blood pressure to insulin
levels (Chatzipanteli et al., 1996). PGE2 and PGI; also
appear to counteract the action of one another in rat

adipose tissue when both adipocytes and endothelial cells

49

are present (Chatzipanteli et al., 1992; Parker et al.,
1989). PGIZ, while considered lipolytic in adipose tissue,
can be adipogenic and will substitute for arachidonic acid
or cAMP in the differentiation media of Ob1771 cells (Negrel
et al., 1989).

In the adipocyte, prostaglandins can act through
multiple pathways. Some studies suggest PGE2 action on
adipocytes is anti-lipolytic and acts via blocking an
increase in cAMP (Vassaux et al., 1992; Christ and Nugteren,
1970; Chatzipanteli et al., 1992). Others suggest that PGE2
acts through protein kinase C via a receptor coupled to
phospholipase C in Swiss 3T3 fibroblasts (Danesch et al.,
1996). Long and Pekala (1996a and 1996b) showed that GLUT4
expression in adipocytes was depressed by 20:4,n-6 and PGE2
and further showed that cAMP was produced in adipocytes upon
treatment with PGE2. These studies indicate that adipocytes
have the capacity to respond to prostaglandins by all three
mechanisms discussed above: increasing cAMP, decreasing cAMP
or altering intracellular calcium levels.

Prostaglandins are also produced in the liver by
Kupffer cells upon injury, sepsis or other stimulus.
Released PG act in a paracrine fashion on the surrounding
hepatocytes. After liver injury, PGE2 can increase

hepatocyte proliferation through the EP3 receptor (Hashimoto

et al., 1997) . PGE2 and PGan can also induce DNA synthesis

50

in cultured hepatocytes in conjunction with hepatocyte
growth factor (Adachi et al., 1995; Refnes et al., 1995).
Skouteris and Kaser (1991) measured DNA synthesis in
hepatocytes treated with PGE2 and PGan (which are induced
by EGF). Not only did indomethacin (cyclooxygenase
inhibitor) block the EGF effect, but PG treatment increased
DNA synthesis, and this increase could be reversed with the
calcium ion channel blocker, verapamil. This work implies
that arachidonic acid was metabolized as a result of the
growth factor treatment to produce the PG to cause the
increase in DNA synthesis in hepatocytes. Later, this
indomethacin inhibition was repeated in TGF treated
hepatocytes. Again, prostaglandins seem to be actively
involved in liver regeneration by increasing DNA synthesis
(Skouteris and McMenamin, 1992).

Both PGE; and PGan have been shown to block the effect
of glucagon treatment of hepatocytes and be reversed by
pertussis toxin (Melien et al., 1988). Glycogen breakdown
is induced by PGE2 in the liver. In Kupffer cell culture,
glucagon induced PGE2, PGDz and PGan synthesis. These PG
then inhibited the glucagon induced breakdown of glycogen.
Furthermore, pertussis toxin could block this PGEzieffect in
hepatocytes. These results indicate that PGE2 is acting
through a pertussis toxin-sensitive Gy-linked receptor which

acts by decreasing cAMP levels (Garrity et al., 1983; 1989;

51

Hespeling et al., 1995a; 1995b; Okumura et al., 1988). In
contrast, others have found that pertussis toxin could not
reverse the PGEQ effects on glycogen breakdown but was
likely acting through a calcium mediated process (Kanemaki
et al., 1993; Mine et al., 1990). In other experiments, EGF
stimulation of hepatocytes increased a G; protein to
activate phospholipase C resulting in an increase in
calcium. The calcium release was blocked when the rats were
pretreated with pertussis toxin before hepatocyte isolation
(Yang et al., 1991; 1993). Clearly G proteins which
activate calcium release exist on hepatocyte membranes, but
which EP receptor PGE2 is acting through is not clear (see

Figure 1.6).

52

 

 

 

[Arachidonic Acidl

 

'\

i

 

 

 

 

 

'\
/

 

 

 

 

Increase Dec ease .
I cAMP J CAMP [IP3/Calc1um]

 

 

 

 

 

 

 

 

Figure 1.6. Overall Prostaglandin Synthesis and Signaling.
Prostaglandins are synthesized from arachidonic acid and act
on cell surface receptors to activate G proteins. This
diagram is a general overview of the synthesis and signal
pathways PG can act through. Notice that PGE2 has 4
receptors and that EP3 has subtypes that work through all
three pathways.

aqpatocyto and Kupffer Coll Interactions

The nonparenchymal liver cells include Kupffer cells,
pit cells, fat cells and endothelial cells. Kupffer cells
were discovered by von Kupffer in 1876. He called them
“sternzellen” or star cells due to their shape. Although
the star shaped cells and their associating cells later
proved to be the above four separate types of cells, all are

located in the sinusoidal walls of the liver. Of these cell

53

types, Kupffer cells are the most important in their
interactions with hepatocytes (Billiar and Curran, 1992).

Kupffer cells are macrophages and vary in location,
shape and size. Internally, Kupffer cells contain many
lysosomes, vacuoles and phagosomes, typical of the average
macrophage. They can be mobile and associate with pit cells
and lymphocytes following inflammation. Kupffer cells
function to fight infections and in detoxification. They
are also in close contact with hepatocytes. Kupffer cell-
hepatocyte interactions include receptor mediated, gap
junction mediated, factor mediated as well as cell-to-cell
interactions. The most important of these is the receptor
mediated communication (Billiar and Curran, 1992).

The most common interaction between hepatocytes and
Kupffer cells is receptor mediated communication. For
example, stimulation of Kupffer cells causes a release of
cytokines which interact with nearby hepatocytes via a
receptor mediated process. Only the Kupffer cells secrete
these compounds, not the hepatocytes. For example, one

effect of Kupffer cells cytokines on hepatocytes is the
inhibition of cytochrome P-450 enzymes. TNFa,

prostaglandins, IL-6 and IL-1 are released from Kupffer
cells during acute phase response to bacteria, viruses,
immune complexes, colloidal compounds and other substances._

These molecules go on to cause a variety of actions on all

54

 

the liver cell types. Growth factors are also secreted by
Kupffer cells and induce hepatocytes to proliferate via a
receptor mediated system in response to liver injury
(Billiar and Curran, 1992; Mion et al., 1995). Some of
these cytokines will be discussed further in the following

sections.

Tumor Necrosis Factor (TNF)

Tumor necrosis factor (TNF) is produced by Kupffer
cells in the liver. A portion of this thesis project
involved using TNF to determine its effect on lipogenesis in
hepatocytes. A brief review of TNF and its action follows.
Most of this information was taken from an excellent review
by Tracey, 1997.

TNF was simultaneously discovered in 1975 by two
separate laboratories as an agent causing cancer cell death
and a cause of wasting in chronic infectious diseases.
After isolation and cloning, TNF was found to be a 17kDa
protein having two forms: a and B. Although originally
thought to be an endogenous anti-cancer protein, TNF is not
Aspecific to cancer cells but can target nearly all types of
cells. TNF can cause all the symptoms of septic shock.
Furthermore, antibodies to TNF can prevent these side

effects of septic shock in baboons, even when bacteria is

55

 

present in their blood. This cytokine has a variety actions
in many different physiological states (Tracey, 1997).

TNF is produced mainly by macrophages, which upon
stimulus by endotoxin or lipopolysaccharide (LPS), will
release the TNF. Other stimuli include enterotoxins,
nitrites, calcium ionophores, irradiation, oxygen radicals,
IL-l, phorbol esters, viruses and TNF itself, just to name a
few. After release, the macrophages will not secrete
further TNF until a 3-7 day recovery period has passed.

Expression of TNF is increased through transcriptional

regulation at a Y-box and an NF-KB enhancer sequence.

Removal of the KB sequence in the TNF promoter will
eliminate the LPS stimulated TNF release. In immunocytes,
NF-KB is translocated to the nucleus to allow for

transcription of proinflammatory genes (as well as TNF
itself). Other transcriptional regulators of TNF include
AP-2, CREB and AP-l. RNA stability and translational
modifications also play a role in controlling TNF levels.
Synthesis of TNF can depend on leukotrienes, cAMP,
prostaglandins, tyrosine kinases and calcium levels (Tracey,
1997).

TNF binds cell surface receptors and activates several
secondary pathways and mediators (Figure 1.7). Some of
these include leukotrienes, prostaglandins, ceramides,

interleukins, nitric oxide, cortisol, insulin, glucagon,

56

 

epinephrine, tyrosine kinases and calcium depending on the
target cell. These responses to TNF depend on the cell
types but include the following physiological effects:
shock, lactic acidosis, hypoxia, fever, anorexia, release of
hormones to regulate glucose, diarrhea, sodium uptake in
muscle cells, depressed motility of the gastrointestinal
tract and system inflammation and many others. For example
in the vascular system, TNF causes production of nitric
oxide (NO) causing vasodilation of blood vessels and
leakage. NO levels are generated at toxic levels, causing
myocardial and endothelial cell death, although the TNF
levels can be toxic itself to these cells (Tracey, 1997;
Fournier et al., 1997). In contrast, PGE2 has been shown to
inhibit NO in cultured hepatocytes (Harbrecht et al., 1996).
TNF is also involved in diabetes and obesity. TNF is
lethal to islet cells and is often produced at higher levels
in diabetic patients. In a septic shock, TNF depresses
appetite and induces thermogenesis hypothalamically as well
as causes insulin resistance. In obese animals, appetite is
not suppressed but TNF levels are increased from adipocytes.
Although TNF levels are high in these obese animals, they do
not show the typical wasting seen in subjects with
sufficient appetite control systems when treated with TNF

(Tracey, 1997).

57

 

TNFa

Shock, acidosis,
fever, anorexia

hormones, etc.

Other \
Transcriptional
egulator Prostaglandins,
Calcium
Interleukins,

Glucagon, NO,
Cortisol, Insulin

 

 

 

 

 

 

Figure 1.7. Summary of TNF Action. This diagram is a
simplified summary of how TNF may act to produce in vivo

responses. NF-KB is bound by I-KB in the cytosol. Upon TNF
binding to its receptor, I-KB is phosphorylated and this

releases NF-KB which can then translocate to the nucleus and
activate transcription of several genes. Other
transcriptional regulators are also likely playing a role in
TNF stimulation of gene transcription.

Adipose tissue significantly contributes to TNFa
levels. In obese mice models and in obese humans, TNFa
levels are greater than in lean or normal controls. High
levels of TNFa. correlate extremely well with
hyperinsulinemia, an indicator of insulin resistance in
euglycemic states. It is thought that TNFa interferes with

the normal insulin signaling pathway by blocking the

58

tyrosine kinase activity of the insulin receptor which in
turn causes insulin resistance (Hotamisligil et al., 1995;
1996b). In contrast to adipocytes, in preadipocytes TNFa
can increase glucose transport and upregulate the GLUTl
transporter (Cornelius et al., 1990). In adipocytes, TNFa
downregulates GLUT4 and aP2 (Stephens and Pekala, 1992).
TNFa is both depressing differentiation and causing insulin
resistance in adipocytes.

The actions of TNFa released from Kupffer cells and
acting on the hepatocyte appear somewhat contradictory to
the in vivo and adipocyte data. In rats treated with TNFa,
liver lipogenesis and serum triglycerides increase and LPL
activity decreases (Grunfeld et al., 1990; Feingold et al.,
1990). Using mice resistant to LPS (they could not produce
TNF when treated with LPS), Adi et al. (1992) demonstrated
that TNF was required for the lipogenic effects of LPS. The
wild type mice had increased lipogenesis as measured by
fatty acid synthesis in the liver while the mutant strain
had lower fatty acid synthesis and much less TNF and IL-1
message. However, in hepatocyte cell culture, TNFa did not
increase lipogenesis in experiments reported by Brass and

Vetter (1994). TNFa also increases lipolysis in

adipocytes, which in vivo, causes a release of fatty acids

into circulation. The liver then re-esterifies and releases

59

 

these in the form of VLDLs which may account for the higher
triglyceride levels seen with TNFa treatment (Feingold et
al., 1992; Grunfeld et al., 1990).

The search for the TNFa mediator involved in the
induction of lipogenesis has resulted in several
possibilities: IL-1, IL-6, sphingomyelin, interferon a and
prostaglandins. Grunfeld et al. (1990) ruled out
prostaglandins as a TNFa mediator but suggested that IL-6
could be involved. Brass and Vetter (1994) reported that
IL-6 and a prostaglandin E agonist both increased hepatocyte
lipogenesis but TNFa did not. Wang et a1. (1995) reported

that TNF does not increase cAMP in hepatocytes but does
increase cAMP levels in Kupffer cells, possibly as a result
of higher PGE2 levels. These findings make it difficult to
determine what compounds are released from Kupffer cells and
their cumulative result on lipogenesis as they collectively
target hepatocytes.

One TNF signaling pathway is through the sphingomyelin
(SM) cycle. TNFa causes hydrolysis of sphingomyelin,
-generating ceramide. The ceramide, a signaling molecule,
can mimic TNF action in some cell lines. Both lead to
apoptosis in leukemia cells, for example. This action is
blocked by the use of PKC activators, suggesting that the

action of ceramide (or TNF via ceramide) may involve the

60

inhibition of PKC. The PKC inhibition hypothesis is further
substantiated by the activation of protein phosphatase by

ceramide although ceramide can activate certain PKC isoforms
and MAPK. Ceramide can also increase NF-KB activity and

PGE2 secretion (Hannun, 1997; Hannun and Obeid, 1995). Also
in leukemia cells, Jayadev et al. (1994) reported that TNFa

increased AA release which in turn increased sphingomyelin
metabolism and generation of ceramide. Both TNF and
ceramide have been shown to lower GLUT4 expression in
adipocytes. This report indicates this signaling pathway
exists in adipocytes in addition to the usual inflammatory
cells (Long and Pekala, 1996a; 1996b).

As mentioned above, another factor indicated in TNF
action is IL-1. IL-1 is a proinflammatory cytokine which
binds a separate and distinct cell surface receptor from
TNF. Both can cause similar effects, however. Several.
inducers of IL-1 such as LPS, phorbol esters, radiation and
viruses also induce TNF. Kupffer cells also produce both of
these cytokines, although monocytes produce much more IL-l
than Kupffer cells. IL—l binds a receptor which causes
activation of a MAP kinase also activated by LPS. Blocking
this MAP kinase phosphorylation can stop production of TNF
and IL-1. There is some evidence that this MAP kinase
phosphorylation involves a Janus kinase (JNK) or Stat

pathway. IL—l can also increase prostaglandin synthesis

61

 

(Tocci and Schmidt, 1997). IL-la can also activate nuclear

translocation of NF-xB, like TNF does (Brasier et al.,

1990). Others have reported that IL-1, acting through the
sphingomyelin/ceramide pathway, regulates the oxidative gene
CYP2CII in hepatocytes (Nikolova-Karakashian et al., 1997).
It is unknown how IL—l, TNF and PG interact to influence Sl4
expression in the hepatocyte. One part of this thesis will
be to look at the effect of these three players in the

suppression of $14.

TNF and P65 in Primary Hepatocyte Culture

Isolation of hepatocytes from adult rats using the
collagenase perfusion technique is selective for hepatocytes
only. This implies that the primary culture contains only
hepatocytes. However, it is possible that small amounts of
Kupffer or other nonparenchymal cells are carried through
into the culture. Moreover, it is Kupffer cells, not
hepatocytes, that produce most, if not all, the
prostaglandins. PGD2 is the most predominant prostaglandin
produced by Kupffer cells but PGE2 is thought to be
important as well. Hepatocytes, although they are not
synthesizing the prostaglandins, have a great capacity for
metabolizing prostanoids and leukotrienes (Billiar and

Curran, 1992; Billiar et al., 1990).

62

 

During sepsis, Kupffer-hepatocyte interaction becomes
very important in glucose and fatty acid metabolism. Sepsis
increases output of interleukins (IL), prostanoids and TNFa
from the Kupffer cells. IL-1 is linked to an increase in
gluconeogenesis while TNFa increases glucose metabolism and
insulin levels. PGDQ increases glycogenlysis in
hepatocytes, an effect that can be blocked with indomethacin
(an inhibitor of cyclooxygenase). Lipolysis and
gluconeogenesis are also stimulated in both the liver and
adipose tissue upon sepsis via the nervous system and
subsequent elevation of adrenaline, cortisol and glucagon.
Sepsis patients then become hyperglycemic, even with
adequate levels of insulin present followed by hypoglycemia.
Patients also become hypertriglyceridemic as triglycerides
are released from the adipose tissue from TNFa action. In
the liver, fatty acid oxidation and release of free fatty
acids is increased. In some experiments, TNFa and IL—1
have lowered LPL activity, which may contribute to the
higher levels of triglycerides in the blood. If no
epinephrine or cortisol is present, TNFa has also been
shown to lower triglyceride production in hepatocytes
(Billiar and Curran, 1992; Rodriguez de Turco and Spitzer,

1990).

63

 

Although both TNFa and prostaglandins are released
from Kupffer cells, they can have different effects. In
contrast to TNFa, PGDz and PGE2 were shown to decrease VLDL

production in hepatocytes by a calcium or cAMP mediated
pathway (Bjornsson et al., 1992). PGE2 has also been
reported to increase glycogenolysis (i.e. increase glucose
output) by a calcium mediated pathway (Mine et al., 1990),
but inhibit glucagon induced PEPCK expression by the same
reported mechanisms (Valera et al., 1993; Puschel and
Christ, 1994).

There has been some research with PGE2 and its effect
in hepatocytes. “It is thought that PGE2 inhibits PEPCK
expression by decreasing cAMP levels through an EP3 receptor
in hepatocytes. This receptor, linked to a G; protein, is
sensitive to pertussis toxin inhibition. Pertussis toxin
has been used to reverse the PGEzteffect on PEPCK activity
(Puschel and Christ, 1994). These authors also reported
that PGE2 increased the decay of PEPCK mRNA. However, if
the PGE2 was added two hours after the addition of glucagon,
there was still some inhibition of PEPCK activity. This
inhibition was not reversible by pertussis toxin, suggesting
another pathway than inhibition of cAMP by PGE2. Others
have reported that an increase in calcium causes inhibition
of PEPCK in hepatocytes. By decreasing extracellular

calcium with EGTA, Valera et al. (1993) reversed the

64

 

decrease in PEPCK message. However, these authors did not
indicate reversal of PGE2 action with any agent. Another
example in cultured hepatocytes, is the report by Mine et
al. (1990) in which glucose output was increased by PGE2.
These authors could find absolutely no increase in cAMP but
did measure an increase in intracellular calcium and IP3
production. These findings suggest that PGE2 is most likely
involved in activating more than one signaling pathway in

the hepatocyte when regulating glucose output.

Calcium Mediators

As discussed above, many of the mentioned compounds use
calcium as a signaling pathway. Arachidonic acid and
prostaglandins both can increase intracellular calcium
levels in many different cell types. Vasopressin is often
used to activate IP3 to increase calcium. This section will
briefly discuss calcium cell signaling and some examples of
cell systems in which calcium levels were measured in
response to various treatments.

The change in intracellular calcium levels is commonly
used as a signaling pathway in cells. Calcium release is
one pathway used by Gq-proteins, for example. After binding
of a ligand to a cell surface receptor, the G-protein is
activated, causing phospholipase C (PLC) to be activated.

PLC cleaves phosphatidylinositol bisphosphate to give rise

65

 

to diacylglycerol (DAG) and inositol triphosphate (IP3).
IP3 is coupled to a calcium channels on the endoplasmic
reticulum which releases calcium into the cytoplasm. DAG
can activate PKC, which is calcium dependent. PKC has a

variety of actions including activation of MAP kinases and
subsequently NF-KB. Calcium ionophores like A23187 can

mimic the IF: effect in many cell types while phorbol esters
like TPA mimic the DAG effects. Another protein activated
by calcium is calmodulin which can regulate
calcium/calmodulin-dependent protein kinases (CaM Kinases)
or calcium ion pumps (Alberts et al., 1994).

Calcium levels are also affected by activation of ion
channels on the cell surface. The hetero-trimeric G;
protein can directly act upon a calcium ion channel in the
plasma membrane to allow calcium influx when the receptor is
activated. (The G1 protein also acts to decrease cAMP, as
discussed in the prostaglandin receptor section.) Calcium
ion channels can also be “voltage gated”, which respond to
membrane polarity. In muscle cells for example, calcium ion
channels on the cell surface are voltage gated but ion
‘Channels on the sarcoplasmic reticulum are not (Alberts et
al., 1994). Figure 1.8 summarizes calcium signaling.

Calcium is a major player in many of the different
signaling pathways. Several examples of calcium release in

hepatocytes or adipocytes in response to the above mentioned

66

 

compounds can be found. Vasopressin, acting through IP3,
has been used in hepatocytes to determine the effects of
intracellular calcium release on gene expression (Duddy et
al., 1989). Several have reported (Hughes et al., 1987;
Butta et al.,1993; and Fernando and Barritt, 1994; Fernando
et al., 1997; Mellgren et al., 1997) that vasopressin
stimulates calcium release through a pertussis toxin
sensitive GTP protein in hepatocytes. In another example,
Kass et al. (1990) reported that vasopressin caused a
calcium influx through ion channels in hepatocytes as a
secondary

response after the initial intracellular calcium release.
Sphingosine has also been shown to activate calcium release
from intracellular stores to activate glycogen phosphorylase
in hepatocytes (Im et al., 1997). Berven and Barrit (1994;
Berven et al., 1995; 1994) localized a G-protein sensitive
to pertussis toxin that upon activation, increased

calcium release into hepatocytes. In WAT, Izawa et al.
(1994) used adrenal corticotropin hormone to show that this
hormone increased arachidonic acid release which increased
calcium in the cells. Furthermore, they could block this
calcium release by pertussis toxin. Calcium is also
released in human platelet cells treated with arachidonic
acid, even when metabolism of AA is blocked by lipoxygenase

or cyclooxygenase inhibitors (Tohmatsu et al., 1989). These

67

d ...-L.

examples demonstrate just a few of the wide variety of
compounds using calcium as a second messenger in their
signal transduction pathways.

Fatty acids can also change calcium levels. In
hepatocytes, palmitoyl CoA has been shown to cause a release
of intracellular calcium (Fulceri et al., 1993). Gomez—
Munoz et al. (1991) reported that palmitate and linolenate
increased glycogen phosphorylase activity through a calcium
mediated pathway. In rat adipocytes, palmitate increases
glucose uptake. In experiments using a intracellular
calcium chelator (quinZ-AM), the palmitate was unable to
stimulate the glucose uptake indicating the importance of
calcium in the signaling pathway (Thode et al., 1989).

$14 is suppressed by an increase in intracellular
calcium. Sudo and Mariash (1996) showed that hepatocytes
treated with the calcium ionophore A23187, had suppressed
814 mRNA. Vasopressin treatment did not result in any
change in $14. Blocking the calcium release from the
endoplasmic reticulum with thapsigargin prevented the
increase in $14 expression normally seen with glucose
treatment. These results suggest that the endoplasmic
reticulum calcium pool is necessary for the glucose
induction of 814 but the IP3 induced calcium release (via

vasopressin) is not.

68

 

G protein PLC

 

 

 

 

Figure 1.8. Calcium Signaling Pathways. Binding of a
ligand to the receptor on the target cell activates a G-
protein. G-proteins can activate PLC which generates DAG
and IP3. IP3 can activate calcium release from
intracellular stores. PKC is activated by DAG and calcium.
One target of PKC activation is a MAP kinase which
phosphorylates I-kB to cause release of NF-kB. NF-kB, once
released, can translocated to the nucleus and activate gene
transcription of targeted genes. Calcium can also activate
CaMK through calmodulin. Other possible effects may also
result as calcium levels change in the cell.

Calcium seems to be involved in insulin-induced
lipogenesis in adipocytes as well as regulate some lipolytic
and lipogenic enzymes via other agents such as parathyroid
hormone and vitamin D (Ritz et al., 1980; Avasthy et al.,

1988). Very few examples exist for direct measurement of

lipogenic gene expression regulation by calcium per se. As

discussed previously, PGan acts through a calcium mediated

69

 

pathway to inhibit adipocyte differentiation, and therefore,
lipogenic genes requiring the differentiated state (like
FAS) would not increase (Sessler and Ntambi, 1998).

These examples represent only a few of the wide variety
of mediators that can cause an increase in calcium in the
cells. Control of each of these and how they interact is
not always clear. While the focus of this thesis is not on
calcium signaling, the regulation of lipogenic gene
expression by AA and PGE2 might involve calcium mediated

mechanisms and will be addressed in Chapters 2 and 3.

Regulation of FAS and 814

Both FAS and 314 are expressed in the adipocyte and
hepatocyte and have been used as models to study lipogenic
gene expression. Neither is detectable in the preadipocyte
but mRNA levels for FAS dramatically increase in 3T3-L1
cells upon differentiation while 814 expression increases
only after adipocytes are treated with glucocorticoids
(Lepar and Jump, 1989). Glucagon or cAMP lowers and insulin
raises FAS and $14 mRNA levels in these cells (Moustaid et
a1, 1993; Paulauskis and Sul, 1988; Jump and MacDougald,
1993; Sul et al., 1993; Jump et al., 1990a).
Triiodothyronine (T3) has proven to increase transcription
of these genes, leading to increases in mRNA. T3 also

increases mRNA stability of FAS in 3T3-L1 adipocytes

7O

(Moustaid and Sul, 1991). 814 mRNA expression in 3T3-L1
adipocytes is not induced by T3 treatment, but
interestingly, S14 mRNA is increased in epididymal fat after
rats are treated with T3. Fat treatment also changes 514
expression in adipose tissue. WAT 814 message levels in
rats fed fish oil are approximately 50% of the levels in
rats on high carbohydrate diets even after only a few hours
(Jump and Oppenheimer, 1985; Lepar and Jump, 1989; Jump et
al., 1993).

In the rat liver or hepatocyte, 814 and FAS regulation
is again nearly identical. Thyroid hormone increases mRNA
expression of both genes, as does insulin or carbohydrates.
Starvation and polyunsaturated fatty acids (PUFA) inhibit
transcription. Rats fed a diet of menhaden oil (contains n-
3 PUFA), have depressed $14 and FAS levels within a few
hours as well as lower serum triglycerides (Jump et al.,
1996; 1993; Blake and Clarke, 1990). After starvation,
refeeding induces $14 and FAS expression in the liver about
25-fold but in adipose tissue only FAS is induced at these
levels after refeeding (Kim and Freake, 1996).

The greatest differences between FAS and 814 are the
sizes of their genes and promoters. FAS is a large gene and
protein. The FAS gene includes over 18,000 base pairs
containing forty-two introns. The 250kDa protein functions

as a homodimer with each monomer comprising seven separate

71

enzyme functions in three domains. Sequence analysis
revealed that its promoter contains transcriptional
regulatory sequences similar to the estrogen response
element, glucocorticoid response element, thyroid response
element and the progesterone response element. Eight
sequences of the cAMP transcriptional regulatory element
(CRE) are also present (Amy et al., 1990; 1992).

The 514 protein has been extensively studied in our
laboratory as a model for lipogenic gene regulation. $14 is
a small protein (M=17,000) with mRNA lengths of 1.3 and
1.47kb (the 2 sizes differ only in the 3’ untranslated
region). This gene is expressed in lipogenic tissues, and
like FAS, is extremely responsive to nutritional status of
the cell or animal. Carbohydrates (or eating a normal chow
diet in rats) induce transcription while starvation and
polyunsaturated fatty acids suppress transcription. Thyroid
hormone greatly induces expression in rat liver or
hepatocytes (Jump, 1989; Jump et al., 1990; 1984; 1993).

High expression of 814 only occurs in the liver,
lactating mammary tissue and adipose tissue. The $14
promoter and gene is also nearly homologous between humans
and rats (Ota et al., 1997). The protein has been localized
to the nucleus by immunohistochemical methods (Kinlaw et
al., 1993). Recently experiments using_antisense Sl4

expression in hepatocytes showed that $14 is involved in the

72

induction of ATP citrate lyase, FAS, ME activity, PEPCK and
PK (Brown et al., 1997; Kinlaw et al., 1995). Cunningham et
al. (1997) reported that 814 is a homodimer which co-
precipitates with an unknown hepatic protein. 814 has also
been shown to be expressed in mammary gland (Jump and
Oppenheimer, 1985). In breast cancer cells lipogenesis is
increased and overexpression of FAS implies a poor
prognosis. 814 has also been found to be increased in human
breast cancer (Moncur et al., 1998). These results suggest
that 514 may be a transcriptional regulator of lipogenic
genes in all three of the tissues where it is expressed.

In summary, 814 and FAS are both lipogenic genes with
similar regulation. Each has hormonal controls: T3,
insulin, glucagon and glucocorticoids. Both are controlled
by diet: fasting, feeding, carbohydrates and dietary fats.
Both also have tissue specific regulation, being expressed
in the liver and adipocyte. The regulation of these genes
has led to great interest in the promoters of each gene.
Because 814 has been very amenable for research purposes and
the promoter is well characterized, 814 has been used as a
model to study lipogenic gene expression in our laboratory.
The following section will focus on several aspects of the

S14 promoter.

73

FAS and 814 Promoter

As discussed in the above section, FAS is regulated
much like $14. Soncini et al. (1995) used a transgenic
mouse expressing the reporter gene chloramphenicol acetyl
transferase (CAT) under the control of the FAS promoter to -
2.1kb to study regulation of FAS by various compounds. CAT
was expressed in the correct tissues (liver, WAT and mammary
gland) and was induced by glucocorticoids and suppressed by
cAMP as expected. However, diets containing fish oil
greatly suppressed the endogenous FAS but only diminished
CAT activity about 50%. This observation was repeated in
hepatocytes upon transfection of the reporter vector (-2195
to -60. Several sites were located upon promoter analysis
including putative Spl, T3 and two GRE binding domains. CAT
activity increased when hepatocytes were treated with
insulin or T3 treatment and dexamethasone compared to
control treated cells but not with insulin or T3 treatment
alone. It was unexpected that T3 and insulin would not
increase CAT activity since both of these agents increase in
vivo levels of FAS mRNA (Clarke and Jump, 1993). Others
have mapped the insulin sensitive region in the FAS promoter
to -67 to -52bp (Moustaid et al., 1993; 1994). These
results indicate that the FAS promoter is not limited to -
2500 base pairs. Work with the FAS promoter in CAT reporter

systems often gives disparate responses. This lab and

74

others have also attempted to determine the PUFA response
region in FAS but efforts to date have not yet been
successful. These responses with FAS have prompted our use
of 814 as a model to study lipogenic gene regulation.

Unlike FAS, the promoter region of $14 is well defined
and has been more amenable to experimentation. Several
regions in the $14 promoter which bind specific factors have
been mapped. The thyroid hormone response region (TRR) at -
2.7 kb was first detected by DNase I treatment of rat liver
nuclei (Jump, 1989). Two other important regions, the
pluripotent response region (PRR) from -1.6 to -1.4 kb and
the proximal promoter region (+19 to -260) were also mapped
as important regulatory regions (Jump, 1989; Jump et al.,
1990; 1988).

The PRR (—l.6 to -1.4) has been implicated as the
carbohydrate and insulin responsive region. This region is
also responsive to glucocorticoids and retinoic acid in
adipocytes although this gene is not responsive to these
compounds in hepatocytes (unpublished observations by A.
Thelen and D. Jump). Towle et al., 1996, reported that -
1476 to -1422 region was required for the glucose response
of 514 in hepatocytes. These authors also reported that two
other factors, upstream stimulatory factor (USF) and an
unknown factor were involved in the carbohydrate response.

USF binds E-boxes, two of which are present in the $14 PRR.

75

Although mutation of the E-boxes did not modify the glucose
response, altering the spacing between them did change the
responsiveness. These authors concluded that while USF may
not be the carbohydrate response factor, it may be necessary
for binding or modification by another factor to allow for
the carbohydrate response of 814. However, recent
experiments in the our laboratory have indicated that
removal of the $14 PRR (-l600 to -1400bp) does not diminish
the glucose response in hepatocytes (unpublished data).

The proximal promoter region spans the region from +19
to -290bp with +1 being the transcription start site. This
area contains a TATA box (-27 to -21), an NF-l binding site
(—63 to ~48) and two regions designated the Y box and C
regions (see Figure 1.9). The Y box contains an inverted
CCAAT box (-83 to -104) which binds C/EBP, NF-l and NF-Y.
Recent experiments have determined that T3 induction of 514
requires NF-Y binding to the Y box, implicating an
interaction between the TRR and the proximal promoter for
transcriptional regulation of 814 (Jump et al., 1997b).

NF-Y, or CCAAT box binding factor (CBF), is a trimeric
protein that binds DNA. Two of the subunits (B and C) bind
to DNA in a “handshake” motif, similar to that seen in
histones H2a and H2b. NF—Y functions to recruit histone
acetylases (e.g. GCNS, CBP and P/CAF) to the DNA (Currie,

1998). Histone transferases transfer the acetate from

76

acetyl CoA to the amino group of lysine. Histones within
nucleosomes are acetylated at lysine residues at their N-
terminal end leading to changes in DNA/histone interaction.
Such modifications are thought to be involved in local
remodeling of chromatin to enable additional_transcription
factors to bind DNA. The 314 Y box and entire proximal
promoter region is within a DNase I hypersensitive region, a
structure indicative of local chromatin remodeling (Jump et
al., 1997b).

The proximal promoter region has been implicated as the
PUFA response region. Analysis using hepatocytes
transfected with various promoter deletion $14 CAT
constructs treated with EPA showed that the region between -
80 and -220 is sensitive to this PUFA (Jump et al., 1993).
814 mRNA in primary hepatocytes is suppressed by several
PUFAs including linoleic acid (18:2n-6), a and 7 linolenic
acids (18:3n-6 and 18:3n-3), arachidonic acid (20:4n-6) and
EPA (20:5n—3) (Jump et al., 1994; Ren et al., 1997).

It is not known how PUFAs control the 814 gene
transcription. However, the Y box is involved in
transcriptional regulation of 314 (Jump et al., 1997b) and
the Y box is located in the PUFA-RR, an element expected to
participate in the PUFA control. This thesis will test the

hypothesis that NF-Y, per se, is the target for PUFA action.

77

 

Thyroid Hormone Insulin Glucose
RXR PPARyZ
NF-Y NF-l PIC

/\\ l J l
W

-2.8 -2.5 -1.6 -1.4 -.3 -.1 —.05 -.03kb

 

TRR PRR C-region Y Box/NF-l TATAA

 

 

 

Figure 1.9. The $14 Promoter. The $14 promoter and its
major binding domains are shown above. TRR= thyroid hormone
response region and PRR= pluripotent response region. The
different factors which bind 814 are listed with an arrow to
its appropriate binding location.

Concluding Statements and Hypothesis

Regulation of lipid metabolism has proven to be
extremely complex. Dietary, hormonal and tissue type can
affect the transcription of the genes involved. A main goal
of our laboratory is to determine how PUFAs regulate
lipogenic gene transcription. We have used the 814 gene as
a model to define the molecular basis of PUFA regulation of
gene transcription. Previous research in our laboratory has
demonstrated that PUFA suppress Sl4 gene transcription in
both adipose tissue and liver and target the proximal
promoter region (-220 to -80bp) of this gene (Jump et al.,

1993). Furthermore, Ren et al. (1996; 1997) showed that EPA

78

suppression of 814 was not through a PPAR mediated pathway
in vivo. Very recently, a Y-box has been found in the PUFA-
RR of the $14 promoter (Jump et al., 1997).

These results lead to the following hypothesis: PUFA
suppression of lipogenic gene expression in adipocytes is
similar to that seen liver. To address this hypothesis the
following questions are asked: 1) how does AA regulate
lipogenic gene expression in adipocytes? 2) Is the adipocyte
pathway of PUFA regulation of lipogenic gene expression also
utilized by hepatocytes? 3) Is the Y box the target for PUFA

regulation of 814 gene transcription?

79

CHAPTER 2

CHAPTER 2

ADIPOCYTES

Nearly a third of American adults are considered obese.
Obesity, an imbalance in metabolism, is a result of
hyperplasia and hypertrophy of adipocytes as a result of
consuming too many calories. Fat in the diet contributes to
this caloric load. The typical American diet contains 37%

of the calories as fat although FDA recommendations suggest
consuming under 30%. Obesity is correlated to serious
health implications including heart disease, hypertension,
insulin resistance and diabetes. Controlling dietary fat
intake may help manage these health risks (Food Fats and
Health, 1991).

Dietary fat includes saturated, monounsaturated and
polyunsaturated fatty acids (PUFA) . A diet high in
saturated fat is linked to heart disease, cancer and
diabetes. In contrast, polyunsaturated fats, especially n-3

fatty acids, are thought to lower the risk of heart disease
by lowerin t ' l 'd ' ' '
9 serum r1g ycer1 es (Zemel, 1995, Rocch1n1,

19
95 ; Food, Fats and Health, 1991) . A major focus of

80

 

research is how these different fats cause or prevent health

risks.
Much of the research on dietary fat has focused on

rodent models. Dietary saturated fat has been shown to
increase fat cell mass in rats at a higher rate than PUFA

(Shillabeer and Lau, 1994). In rats, dietary

polyunsaturated fats lower hepatic de novo lipogenesis. In

the liver, n-3 polyunsaturated fats decrease the synthesis

of fatty acids and triglycerides and lower enzyme activity
fatty acid synthase and

and message for malic enzyme,
1990a;

gliicose-6-phosphate dehydrogenase (Clarke et a1,

Clarke and Jump, ' 1994) .
Unlike the liver, the role of PUFA in the adipocyte is

not: 433 well understood. Previously, our laboratory reported
that rats fed a diet of n-3 fatty acids had lower $14 and

FAS expression in both the liver and epididymal fat tissue

Several possible PUFA regulatory

(Jump et al., 1993).
A PPAR dependent mechanism was

me chanisms were proposed .
However,

Studied in the liver as the $14 PUFA regulator.

itrleEB pathway was found not to be operative in 514
Understanding how

suppression (Ren et al., 1996; 1997).
PO PAS regulate lipogenic enzymes, lipogenesis and
t -

rlglyceride synthesis in the adipocyte is the focus of this

:2
hapter. These findings may be important both

81

 

physiologically and pharmacologically in the understanding
and treatment of obesity or diseases related to obesity.
This chapter examines the effect of polyunsaturated
fatty acids (PUFA) on lipogenic gene expression in cultured
3T3-L1 adipocytes. Preliminary work indicated that
arachidonic acid (20:4,n-6; AA) and eicosapentaenoic acid
(20:5,n-3; EPA) suppressed mRNAs encoding fatty acid

synthase (FAS) and $14, but had no effect on B-actin. A

clonal cell line containing a stably transfected S14CAT
fusion gene was established to evaluate the transcriptional
corrtrol of $14 by PUFA. Given the high potency of 20:4,n—6
on $14 mRNA expression and CAT activity and previous reports
thaat: dietary n-3 fatty acids lowered $14 expression in both
liver and epididymal fat in rats, the following hypothesis
was developed: “Arachidonic acid suppression of adipocytes
is similar to the PUFA suppression of $14 in the liver”.
Several aims were developed to address how arachidonic acid

SL1E>IDlressed $14 expression in adipocytes.

JUDHS

Ain'- 1: Where is the adipocyte PUFA response region in the

I‘dlll‘ :2:

$14 promoter?

Do peroxisomal proliferator activated receptors

(PPARs) bind the $14 pluripotent response region

(PRR)?

82

 

Are PPARs involved in the PUFA control of $14 in

Aim 3:
adipocytes?
Aim 4 Does arachidonic acid regulate Sl4 gene expression
through a prostanoid pathway?
Aim 5 Are PUFAs altering lipid metabolism in adipocytes?

INHTKHNWCTIIHI
Polyunsaturated fatty acids (PUFA), particularly the

highly unsaturated n-3 fatty acids, when fed to rodents
irflaibit hepatic de novo lipogenesis, triglyceride synthesis

anti secretion and induce peroxisomal and microsomal fatty
1994; Jump et al., 1996;

acid oxidation (Clarke and Jump,
1980, Thomassen et al., 1982,

1990; Rambjor et
1981;

1994; Neat et al.,
1988; Aarsland et al.,
Toussant et al.,

1993;

Flatmark et al.,

al. , 1996; Rustan et al.,
Clarke et al., 1990a, 1990b; Blake and Clarke, 1990).

1988;
PUFA

effect on these metabolic pathways is controlled by the
transcription of specific genes involved in these pathways.

£231? eexample, fish oil rapidly inhibit the transcription of

genes encoding the fatty acid synthase (FAS) and the 814
while inducing expression of genes encoding acyl

E31:<3"t3€31n,
c:
(DJAX (oxidase (AOX) and cytochrome P450 4A2 (Cyp4A2), enzymes

1L
r1"‘Dlved in peroxisomal and microsomal fatty acid oxidation,

1:
espectively (Ren et al., 1996; 1997).
The molecular basis for PUFA-mediated control of

83

 

hepatic gene expression involves at least two distinct
pathways. One pathway requires the peroxisome proliferator
activated receptor (PPAR), a member of the steroid-thyroid
supergene family. In liver, PPARa is the principal PPAR
subtype accounting for the fatty acid control of AOX and
CYP4A2 (Ren et al., 1997). PPARa is activated by
peroxisome proliferators, including fatty acids. The second
pathway is independent of PPAR: and is involved in the
PUFA-mediated suppression of 814, FAS and L-pyruvate kinase
(Ren et al., 1996; 1997; Liimatta et al., 1994). Specific
fatty acid regulated transcription factors have not yet been
identified for the PUFA-mediated control of FAS, 814 or L-
PK. While these genes are subject to complex control by
insulin, T3 and glucose (Clarke and Jump, 1994, Jump et al.,
1996; Liimatta et al., 1994), the cis-regulatory targets for
PUFA control of S14 and L-PK do not converge with the i
Principal targets for endocrine or carbohydrate control.
InStead, the PUFA-regulatory elements converge with targets
that serve an ancillary role in hormone/nutrient control of
-gene transcription (Jump et al., 1993; Liimatta et al.,
1994).
While the liver serves as a major target for fatty acid
J:‘egullated gene expression, recent studies indicate that
white adipose tissue might also be a target for fatty acid

c
<>1~*‘tricol of gene expression (Jump et al., 1993; Ailhaud et

84

 

 

al., 1995; 1996). Fatty acids promote differentiation of
preadipocytes to adipocytes, a process that involves PPARyZ.

PPAR'y2 is activated by thiazolidinediones, a class of

insulin sensitizing drugs, as well as fatty acids and

prostanoids (Tontonoz et al., 1994a; 1994b, 1994c;

Spiegelman and Flier, 1996; Kliewer et al., 1995; Kliewer et

al., 1995; 1997; Forman et al., 1995; 1997). Prostanoids,
products of arachidonic acid metabolism, have also been

reported to promote adipocyte differentiation in culture and

to affect gene expression (Ailhaud et al., 1992; Kliewer et

al., 1995; 1997; Forman et al., 1995).

Since feeding rats for 5 days on diets containing fish

oil leads to a 50% suppression of mRNAms and mRNAs“ in

epididymal fat (Jump et al., 1993), I was interested in

determining whether the mechanism of PUFA-mediated

suPpression of lipogenic gene expression in adipocytes is
Similar to that found in liver. Accordingly, I examined the
effects of specific mono- and polyunsaturated fatty acids on
li53<>genic gene expression in 3T3-L1 cells.

- L1 cells differentiate in vitro from preadipocytes
(fibroblasts) to adipocytes and this differentiation is

accompanied by the induction of lipogenic genes as well as

1:€3'::‘319tors that bind lipogenic and lipolytic hormones

( -
Al lhaud et al., 1992) . My results show that while specific

9
DEA inhibit $14 and FAS gene expression in cultured

85

 

adipocytes, the mechanism for control involves a prostanoid
pathway. This mechanism of control differs from the one

previously described in liver (Jump et al., 1993).

METHODS AND MATERIALS

CELL CULTURE 3T3-L1 cells were grown to 2 days post
confluence in Dulbecco's Modified Eagles Medium (D-MEM)
supplemented with 10% calf serum, penicillin (100 units/l)

and streptomycin (0.1 mg/l). Differentiation was initiated

with D—MEM containing 10% fetal bovine serum (FBS), 1 uM

insulin, 1 uM dexamethasone (DEX) and 250 uM isobutylmethyl

xanthine (IBMX) for 48 hours. After initiating
differentiation, media was replaced with D-MEM supplemented

with 10% FBS and insulin and changed every 2—3 days. All
Cells were maintained at 37°C and 5% COz/95% 02 (See Appendix

A for further cell culture solutions.)
Several 3T3-L1 cell lines containing stably transfected

sl4CAT fusion genes have been previously described
'(baéi<:t30ugald and Jump, 1992). The monoclonal C11 cell line
contains Sl4CAT124 (CAT: chloramphenicol acetyl
t“Talusferasm reporter gene and the $14 promoter to extending
from -4315 to +19 bp fused upstream from CAT. The
monoclonal cell line was derived from the original

p .
Qll’clonal line by differential dilution. Cell lines were

86

   

k

 

selected based on ability to differentiate and express

detectable CAT activity. Cell line #11 was chosen (hence

C11 cells) and all experiments were done using this cell

The cells transfected with TKCAT208, MamNeoCAT or

line.
TKCAT208 contains

RSVCAT are pools of G418-resistant cells.

the region extending from -1.6 to -1.4 kb upstream from the

$14 transcription start site fused upstream from the

thymidine kinase (TK) promoter. The -1.6/-1.4 kb region

contains the glucocorticoid and adipocyte-specific elements
required for $14 expression in adipocytes (MacDougald and

Jump, 1992). MamNeoCAT (Clontech) contains the
glucocorticoid inducible MMTV promoter while RSVCAT (from S.

Conrad, Michigan State University) contains the Rous Sarcoma

Virus promoter. With the exception of MamNeoCAT, these

stably transfected cell lines were prepared by co-
transfection with SV2Neo and colonies were selected and

maintained in 0.4mg/ml G418 (Geneticin, Life Technologies)

unt i l confluent .
Other monoclonal cell lines were derived using the

calcium phosphate transfection method then selected by

differential dilution under G418 (geneticin) pressure.

Cells were plated at 600,000 cells per 100mm plate in 10ml
growth media the day before transfection. On day of

t
ransfection, 30119 of CsCl prepared vector and Bug of CsCl

D
h~§Dared selection vector (SV2Neo) were mixed with 0.5ml 2X-

87

 

BSS and 0.5ml CaClz. This mixture was incubated at room
temperature for 20 minutes then added drop-wise to plate of
cells. The plate was swirled and replaced into incubator at
37 C at 3% C02. Media was removed the next day and replaced
with growth media and placed in normal incubator conditions
(10% C02). 'When cells reached 70-80% confluence, they were
rinsed with PBS, trypsinized and split 1:5 in media
containing 400pg/ml geneticin. Media was changed as
necessary until colonies appeared (2 weeks or more) and
cells were further cloned with 96 well plates for monoclonal
cell line establishment or pooled for polyclonal cell line
establishment (Chen and Okayama, 1987) .
From storage, all cell lines were initially plated with
O. 4mg/ml G418 and maintained until confluence in 0.2mg/ml
G418 . After initiating differentiation, cells were
maintained in the absence of G418. Cells were treated with
fatty acids at concentrations indicated in figures and
a1"Ways at a 5:1 ratio with fatty acid-free bovine serum
albumin. Other treatment concentrations are indicated in
the figure legends. Following treatments, cells were
assayed for CAT activity and protein content as previously

described (Jump et al., 1993). CAT Units: 1"C-acetylated
c
113~Oramphenicol CPM/lOO pg protein/hour.
RNA ANALYSIS Total RNA was isolated from cells using

R z:
N STAT-60 (Tel-Test B, Friendswood, Texas) or Gibco's

88

 

Trizol. RNA (20 pg/lane) was electrophoretically separated
in 1% agarose-formaldehyde gels (Jump et al., 1994).
Northern blots were prepared and probed with radiolabeled
cDNA (Jump et al., 1994) for $14 (Sl4ExoPEII6), FAS (Pas-l;
from H.S. Sul, Univ. Calif.-Berkeley) and B-actin (L. Kedes,
Stanford, Palo Alto, CA).

RNA was electrophoresed in a 1% agarose, 1X MOPS, 1.7%
formaldehyde gel (10-20pg/lane). After running the gel,

gels were rinsed in water for 15 minutes and transferred to
nitrocellulose membrane using the turbo-blotter system with
10X SSC as transfer buffer (Schleicher and Schuell, Keene,
NH). After transfer blots were marked with pencil for
lanes, allowed to dry then baked for 2 hours in a vacuum

oven (nylon blots were UV linked). Blots were prehybridized
for at least 2 hours in prehybridization buffer at 429C then

hybridized overnight with desired cDNA probe. Probes were
synthesized using Gibco's Random Primers DNA Labeling System

and cleaned with spin columns from 3' Prime 5' Prime. Blots

were washed in .1X SSC, .1% SDS at 55-60%:. Blots were then

exposed to film for 24-48 hours at -80°C before development

or exposed to the phosphoimager screen for 1-24 hours. This
protocol is a combination of Fourney et al., 1988, and Jump

et al., 1984.

89

 

GEL SHIFT.ANALYSIS Gel shift analysis involves four
steps: labeling of the oligonucleotide,
transcription/translation of the receptor, binding of
receptor(s) to oligonucleotide and gel preparation.

Oligonucleotide Labeling: Oligonucleotides were

annealed by heating an equal amount of both Oligonucleotides
in 1X TEN to 859C for five minutes and cooling to room
temperature slowly. Oligos were end labeled by incubating

the following mix for 30 minutes at 30W3: Zul polynucleotide
kinase buffer, lul 100mM DTT, annealed oligos (100-200ng),
1p.l polynucleotide kinase, Sul 32-P.yATP and water to 20111.

The reaction was stopped with Sul 200mM EDTA. One

microliter was counted for specific activity determination.

Oligos were cleaned by a spin column or with TCA
precipitation and recounted for cpm/lu”

Transcription/translation of receptors was done using
Promega's Transcription/translation kit according to the
manufacturer's protocol. One or two micrograms of plasmid

was required for this reaction. Receptors were stored at -
I80°C and kept on ice at all times after synthesis.

The binding reaction was set up after receptors were
made and oligos labeled. Each reaction contained 4.5ul

dialysis buffer, lpl DTT, 0.51.11 100mM MgClz and lpl dI:dC,

ZlLl of receptor or receptor mix and 5000cpm labeled oligo

90

(usually lul). As a control, one tube contained no

receptor (only oligo) and another contained 2ul

unprogrammed cell lysate instead of receptor. The reaction

incubated at room temperature for 20 minutes. Before
loading, Sui of stop buffer was added to each tube.

Gel: 8% polyacrylamide gels were used for gel shift
analysis. These were prepared or the Biorad precast gels
were used. The precast gels were less likely to separate

bands as clearly as the poured gels. In a .25X TBE buffer,
the gels were run at 350 volts for about 1 hour at 4°C. The

precast gels were run at half this speed for the same amount
of time. After electrophoresis, gels were dried for 1—2
hours (the precast gels take at least 2 hours) and exposed
to film overnight.

TRIGLYCERIDE AND EATTY ACID ANALYSIS Total

triglycerides were analyzed using a Sigma Triglyceride Assay
Kit and Oil Red O staining. For triglyceride extraction,
cells were rinsed with PBS and then scraped into lml
methanol in a corex tube and 2ml of chloroform was added to
each tube. The sample was homogenized and lml of .15M
acetic acid was added. The tube was vortexed and
centrifuged for 10 minutes at 1000rpm. The lower layer was
removed and aspirated under nitrogen. The remaining

triglyceride was resuspended in 95% ethanol and analyzed

91

using the Sigma's triglyceride kit according to their
protocol.

Oil Red O stains triglycerides. After staining as
described previously (Mater, 1994), lml of 100% ethanol was
added to the plate to remove the stain. The absorbance of
each sample was determined at 540nm on a Beckman
Spectrophotometer.

For fatty acid analysis, cells were rinsed with
phosphate buffered saline then scraped into microtubes in
methanol. Cell extracts were frozen until assayed. Fatty
acid analysis of adipocyte fatty acids involved conversion
of fatty acids to methyl esters of total cell lipids by
direct trans-esterification using boron trichloride/methanol
(14% w/v, Sigma). After methylation, extracts were

extracted twice with hexane then dried under nitrogen. The
fatty acids were then resuspended in 100ul hexane. The

composition of the fatty acid methyl ester was determined by
a capillary gas liquid chromatography using a Hewlett-
Packard 5890 gas chromatograph fitted with a 50 m x 0.025 mm
(id) CP-Sil 88 capillary column (Chrompack, Middleburry, The
Netherlands) and a flame ionization detector. A temperature
gradient program from 150°C to 250°C at 1°C/min was used.
Injection port and detector temperatures were 240W3. The
fatty acid methyl esters were identified by comparing their

retention times versus those of authentic standards.

92

ISOLATION OF ADIPOCYTE NUCLEI Nuclei were isolated
from both preadipocytes and adipocytes with this protocol.
Media was removed from 100mm plates of cells and rinsed with
PBS. Three milliliters of Solution A was added to the
plates and swirled. Cells were allowed to sit and swell for
5 minutes at room temperature and then the solution was
removed. Three milliliters of Buffer A were added to the
cells, cells were scraped and placed into a sterile
homogenizing flask. After a brief homogenization, the cells
were layered over 5ml of Buffer B and centrifuged at 3000 x
g for 10 minutes. The solution was decanted, cells washed
in 3ml of Buffer C and again centrifuged at 1000 x g for
five minutes. The solution was decanted and nuclei were
resuspended in 200ul of Buffer D. After removing Sul for
absorption reading at 260nm, the remaining was stored at —
80°C.

Nuclear proteins were obtained from isolated nuclei
with this procedure. All procedures are carried out on ice.
Nuclei were resuspended in Buffer D (from Nuclei isolation
procedure) at 100 Aum units per milliliter. The nuclei were
placed in a polyallomer tube (65V-Dupont vertical rotor,
13.5m1 tubes) and nuclear lysis buffer was added to 7ml.
1.0ml 4M (NH4)2804 was added and adjusted to 12ml with
nuclear lysis buffer containing 0.5M (NH4)ZSO4. After

capping and mixing, the tube sat on ice for one hour. The

93

chromatin was then sedimented for 80 minutes at 40,000rpm.
The volume of the supernatant was determined and placed into
a new tube. Solid (NH4)ZSO,; was added at .3mg/ml to tube and
mixed on ice overnight. The extract was again spun for 30
minutes at 45,000 rpm. The supernatant was removed and the

pellet carefully resuspended in dialysis buffer and allowed
to shake overnight at 4°C. The extract was then dialyzed

against 100 volumes of dialysis buffer overnight with one

buffer change after 4 hours. Nuclear proteins were

recovered into sterile tube, labeled and stored at ~80%2

after 2ul was removed for protein concentration.

IREERHUTS
PRELIMINARI RESULTS

PUEA Suppress 814 and EAS Geno Expression in L1
Adipocytes. Treatment of primary hepatocytes with 20:4, n-6
or 20:5, n-3 lead to a suppression of $14 and FAS mRNA with
an ED“,<:100 uM (Jump et al., 1993). To determine if PUFA
inhibited the expression of these mRNAs in cultured 3T3-L1
adipocytes, cells were treated with vehicle, 18:1, n-9;
20:4, n-6; or 20:5, n-3 for 48 hrs (Figure 2.1). 18:1,n-9
had no significant effect on mRNAs encoding 814, FAS or B—

actin. In contrast, both 20:4, n-6 and 20:5, n-3 suppressed

94

Sl4 mRNA by 85 and 70%, respectively. FAS mRNA was
suppressed by 70% following 20:4 treatment and ~40%
following 20:5 treatment. Actin mRNA was unaffected by
these treatments. These results indicated that the
previously reported effects of dietary PUFA on adipocyte FAS
and $14 gene expression may be due to direct effects of PUFA
on fat cells. This effect is fat type specific and can not
be attributed to a generalized fatty acid effect since 18:1
did not affect any mRNA examined. Moreover, the lack of a
PUFA effect on actin mRNA suggested that the inhibition of

$14 and FAS was gene specific.

 

 

 

 

 

 

 

 

0.2 #-

    

  

 

 

mRNA in Adipocytes
12
‘ ‘ EE - =
0.8 .. £5 E: g
g E; E— : UAbum'n
2 0.6 -- g E _——_=*‘ BOA
m 04‘ 22% 33% g 2:: IAA

 

 

 

 

 

 

0- —

 

 

 

(D
.L
b

fi-Actin

 

 

 

Figure 2.1. The effect of fatty acids on adipocyte mRNA
levels. RNA was isolated from fully differentiated

adipocytes after 48 hours of treatment with SOuM albumin and

250uM fatty acid. Samples were in triplicate and standard
deviations of the mean are shown. This is a representative
graph of at least 3 experiments.

95

PUFA regulates 814 at the trenscriptionel level. In an
effort to establish the mechanism for control, a stable cell
line containing the Sl4CAT fusion gene was used. This
Sl4CAT fusion gene contains the cis-regulatory elements
required for the adipocyte-specific glucocorticoid-mediated
activation of transcription of this gene. Accordingly, basal
and DEX-mediated induction of CAT activity was examined in
preadipocytes and adipocytes receiving vehicle, 18:1, 20:4
and 20:5 (Figure 2.2).

In preadipocytes, CAT activity was expressed at low
levels and was not induced by DEX or affected by any fatty
acid treatment (Fig. 2.2A). In contrast, DEX induced Sl4CAT
~18-fold in adipocytes (Fig. 2.2B). While 18:1 treatment
had no effect, 20:4,n-6 and 20:5,n-3 both inhibited CAT
activity by >70%. The effect of the fatty acid treatment on
CAT activity by both basal (no DEX treatment) and induced
(DEX treatment) was comparable (Fig. 2.2b) indicating that
20:4,n-6 and 20:5,n-3 acted on the basal expression and not

on DEX-mediated transactivation.

96

 

 

 

 

CATActivity in Pnadipocybs B CATAcflvityin Adipocytes
1200 12500
1d!) 10”
g an g
c c 7500
E °°° E
5 400 5 m
200 2500
0 O
Alum 0A M 6% M GA AA BA

 

 

 

 

Figure 2.2. Effect of fatty acids on Sl4CAT Activity in
Preadipocytes and Adipocytes. Cells were treated with 250uM
fatty acid and 50pM albumin for 48 hours and CAT activity
was measured. Solid bars are also treated with luM
dexamethasone. Means of triplicate samples with standard
deviations are shown. These graphs are one representative of
several experiments are shown. A: Preadipocytes B:
Adipocytes

Dose Response Analysis. In primary hepatocytes, both
20:4,n-6 and 20:5, n-3 are equipotent inhibitors of Sl4CAT
activity (Jump et al., 1993). To determine if adipocytes
responded to PUFA like hepatocytes, L1 adipocytes containing
the stably integrated Sl4CAT gene were treated with fatty
acids ranging from 50 to 1000uM (Figure 2.3). While
treatment of cells with 18:1,n-9 up to SOOpM had no
significant effect on CAT activity; lmM 18:1,n-9 reduced CAT
activity by ~60%. Both 20:4, n—6 and 20:5, n-3 inhibited
Sl4CAT activity. However, EDW for 20:4,n—6 was 6-fold
lower than 20:5,n—3. In contrast to liver (Jump et al.,

1993), 20:4, n-6 is a more potent inhibitor of $14 gene

expression than 20:5,n-3 in adipocytes.

97

 

Dose Response of Fatty Acids In Adipocytes

160

+AA
120 1’ I j +8!

 

 

RdflhaCATUMb

 

 

 

 

 

0 50 25) 5M) 1mm

 

 

pMofFA

 

Figure 2.3. Dose Response of Fatty Acids in Adipocytes on
Sl4CAT activity. Fully differentiated adipocytes were

treated with varying doses of fatty acids and lpM
dexamethasone for 48 hours and CAT activity was measured.
Values were normalized to the albumin control. Standard
deviations are shown.

Continuous fatty acid treatment analysis. Saturated
and polyunsaturated fatty acids have been reported to
stimulate adipocyte differentiation in OB1771 cells (Ailhaud
et al., 1995; 1996; 1992). To determine whether PUFA
affected adipocyte differentiation, specific fatty acids

were added to the medium after removal of the

differentiation medium and maintained in the medium for 8
days at SOuM (Figure 2.4). While 18:1 had no effect on DEX-
induced Sl4CAT expression, 18:2,n-6; 18:3,n-6 and 20:5,n-3

suppressed CAT activity by <30%. In contrast, cells treated

98

with 20:4,n-6 showed a >90% suppression of both basal

(Figure 2.4) and DEX-induced Sl4CAT activity (not shown).

 

Continuously Treated Adipocytes

 

 

 

 

 

 

Alb 0A LA yLN AA EPA

 

 

 

Figure 2.4. Continuous Fatty Acid Treatment on Sl4CAT
Activity in Adipocytes. C11 cells were treated with SOuM

fatty acid and lOuM albumin for 8 days and CAT activity was
determined. Alb = albumin, OA = oleic acid (18:1n-9), LA =

linoleic acid (18:2n-6), aLN = a-linolenic acid (18:3n-6),
AA = arachidonic acid (20:4n-6) and EPA = eicosapentaenoic
acid (20:5n-3). This is a representative of at least 2
separate experiments and each bar represents an average of
three samples with standard deviation shown.

Fatty Acid Treatment has little effect on lipid
accumulation in adipocytes. Cells treated with various
fatty acids were stained with the lipid specific stain, Oil
Red 0. As shown in Figure 2.5, cellular lipid accumulated
fairly equally among all treatments. This figure shows the

adipocytes before, during and after differentiation. Even

99

with the various fatty acid treatments, the cells looked

similar microscopically and after Oil Red 0 staining.

100

Figure 2.5. Oil Red 0 Stained Preadipocytes and Adipocytes.
Adipocytes were fixed and stained with Oil Red O. A:
Confluent preadipocytes before differentiation mix was
added, B: Adipocytes not differentiated but kept in culture
for 12 days with the usual media present in all treatments,

C: Adipocytes differentiated and treated with SOuM AA for 5
days post-differentiation media, D: Adipocytes
differentiated and treated with SOuM EPA for 10 days
post-differentiation media. There were no differences
visually detectable before or after staining in all the
treatments, regardless of fatty acid present.

101

     

            

«View. A. . .. ... Aw. .Waiwfi...m

  

: s. . . .V
1 . . . . \ VA . x
.. u .. .5 u .
. ,; .
, ... .

._ . . . ._
... . , . G» ..
3.. sun.

 

 

102

AIMIl RESULTS
Aim 1: Where is the PUFA response region in the $14
promoter?

Several stable cell lines were established from
transfection of reporter vectors containing different $14
promoter constructs to determine the PUFA response
element(s). The different constructs are shown in Figure
2.6. As controls, cell lines containing the MMTV (MAMNeo
plasmid) and thymidine kinase (TK) promoters with the CAT
reporter gene were also established. In some cases,
monoclonal cell lines were used (selection based on
adipocyte phenotype and CAT activity levels).

PUEA.Effects are specific to the 814 gene. To show the
specificity of PUFA action on $14 expression and that PUFA
did not have generalized effects on glucocorticoid
activation of gene transcription, the effect of PUFA on the
expression of the glucocorticoid inducible MMTV promoter was
examined (Figure 2.7). DEX induced CAT activity ~8-fold and
PUFA did not affect either the basal or induced level (not
shown) of CAT activity. There were also no fatty acid
effects in RSVCAT adipocytes on CAT activity, again showing
fatty acid specificity to different promoters (Fig 2.7).
Similar results were obtained in cells containing stably
integrated TKCATPAN (Fig 2.7). This plasmid contains the

$14 glucocorticoid response region fused upstream from the

103

thymidine kinase basal promoter (TKCATPAN, Methods and
Materials). These results argue against the $14
glucocorticoid response region (between -1069 and -1588 bp)
as the cis-regulatory target for PUFA-mediated suppression
of $14 gene expression in adipocytes. Based on previous
reports of the functional elements controlling 814 gene
expression in liver and preadipocytes (Jump et al., 1993;
MacDougald and Jump, 1992), these findings implicated the
$14 proximal promoter (i.e. -290 to -1 bp) as a likely
target for PUFA action. However, additional studies will be
required to localize the cis-regulatory target for PUFA

action.

104

 

-4315 -2897 -1601 -220 -120 -80 ____¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CAI'
l // | // l I l l
Sl4CAT124 ’9, ’0,
‘—————>
I CAI'
Sl4CAT170
L n CAT
Sl4CAT171 U
(mutation at -138)
Sl4CAT176 I .AT
(mutation at r1
-104/-111) lJ
Sl4CAT196
(block mutation I CAT
—80 to -120)
Sl4CAT198
(block mutation I CAT
of -80 to -220)
Sl4CAT127 _,
L CAT
Sl4CAT121 _____.
(block mutation I I CAT
of -87 to -1381) I l—
‘—————s
Sl4CAT131 l J 1 CAT
(block mutation of Li, I
-290 to -1386)
‘TKCAT202 [TK Element CAT
—————9
TKCAT206 I 314 lTK Element CAT
_(contains 814 region 1
of -1069 to —1386)
7......
TKCAT208 | 514 ITK Element CAT

 

 

(contains 514 region I
of -1069 to -1588)

 

 

 

 

Figure 2.6. $14 Plasmid Construct Maps. Each plasmid used
in the adipocyte transfections are shown above. The $14
promoter region included in each plasmid is described and
shown. All contain CAT as the reporter gene.

105

 

Fatty Acid Specificity

 

 

I WT
UTKCATPAN
D RSVCAT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 2.7.. Fatty Acid Specificity. The fully
differentiated cell lines shown above were treated with the

usual fatty acid (250pM) and albumin (SOuM) for 48 hours and
CAT activity was determined. The RSVCAT data was divided by
100 for making this graph. Each of these shown is a
representative of 3 separate experiments. Each bar
represents a triplicate sample with standard deviation error
bars.

Localization of PUFA-RR was unsuccessful in adipocytes.
Several different $14 promoter constructs were tested in
order to determine which region was sensitive to arachidonic
acid suppression. Some of the cell lines had previously
been established in the our laboratory and monoclonal cell
lines were created from these cell lines. Other cell lines
were established as described in Materials and Methods. As
shown in Figure 2.7, the pluripotent response region (-1069

to -1588) of the 814 promoter was not sensitive to fatty

acids in TKCATPAN adipocytes. Therefore, other cell lines

106

containing constructs with more proximal regions of the $14
promoter were established and tested for this project.

Table 2.1 summarizes the PUFA localization results in
adipocytes. Each cell line was treated with AA and compared
to OA treatment. The data confirmed that the PUFA response
region was located between -220 and -80 bp of the $14
promoter in cell lines containing Sl4CAT131 and 5 of 6
clones of Sl46CAT198. Sl4CAT127 was not sensitive to AA and
should have been with 1600 bp of the $14 promoter present.
This cell line was developed many years ago, and efforts to
develop a monoclonal cell line were unsuccessful as was
creating a new cell line with this plasmid. One clone of
the Sl4CAT198 cell lines was also sensitive to AA. This was
unexpected since it is missing the -220 to -80bp region. It
is unclear why this one clonal cell line would be sensitive
to AA while 5 clones were not, although the site of
integration by the plasmid into the genome of the cell could
affect sensitivity.

The RSV constructs were all established as polyclonal
cell lines. CAT activity was measured but results were
Variable and not identical to the $14 plasmids. For
example, CAT activity in cells containing RSVCAT139 should
have been sensitive to only AA, not both AA and OA.

Even with these exceptions, overall, the results

confirmed that AA is suppressing 814 through the promoter

107

region between -220 and ~80bp as expected from previous
reports of the PUFA response region (Jump et al., 1993).
Efforts to localize the PUFA RR more precisely in adipocytes
were not successful. Chapter 3 will re-examine this issue

in transfected primary hepatocytes.

108

Table 2.1.

FA

Sensitivity of $14 Constructs in Adipocytes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CGNSTRUCT PROMOTER REGION SUPPRESSED BY AA
Sl4CAT124 -4315 to +19 Yes
(C11)
Sl4CATl71 -2897 to +19 Yes (3 clones)
w/mutation at -138
Sl4CAT196 —2897 to +19 with a block No (3 clones)
mutation of -120/-80
Sl4CAT198 -2897 to +19 with a block Yes (1 clone)
mutation of —220/-80 No (5 clones)
Sl4CAT127 -1601 to +19 No
Sl4CAT121 -1588 to +1381 ND*
and -87 to -8
Sl4CAT131 -1588 to -1386 Yes
and -290 to -8
TKCAT208 -1588 to -1069 No
(TKCATPAN) with TKgpromoter
TKCAT206 -1386 to -1069 No**
with TK promoter
RSVCAT136 TK element + RSV No
RSVCAT137 -1400 to -1200 kb and No
-220 to -120 + RSV
RSVCAT138 1400 to -1200 and No
-120 to -80 + RSV
RSVCAT139 1400 to -1200 and Yes (both OA &
—220 to -80 + RSV AA)
RSVCAT14O 1400 to -1200 and RSV No
RSVCAT141 1400 to -1200 + TK element No
+ RSV
RSVCAT148 -1309 to -1069 + RSV No
Table 2.1. Fatty Acid Sensitivity of $14 Constructs. This

tables lists the plasmids stably transfected into adipocytes

and their respective promoter regions.

All cells were

treated with 250uM AA and SOuM albumin or albumin alone for
48 hours and CAT activity was determined.
was tested at least 3 times and a summary of the results are

given.

and this is indicated on the table.

*CAT counts at or below background
**CAT counts barely above background

109

Each cell line

Some cell lines had very little or no CAT activity

 

 

AIMIZ RESULTS
Aim.2: Do peroxisomal proliferator activated receptors
(PPARs) bind the 814 pluripotent response region (PRR)?
PPARyZ binds the TSE-l Region but TSE-z. Using gel
shift analysis, the tissue specific elements (TSE-l and 2)
were tested for PPAR binding. TSE-l and TSE-2 are regions

in the $14 promoter which bind transcription factors.

Binding to TSE-l is more active after differentiation, while
binding to TSE-2 is active during the preadipocyte stage

(MacDougald and Jump, 1992). As shown in Figure 2.8, only
PPARyZ and not PPARa and FAAR bound to TSE-l. HNF-4 also
did not bind this region (not shown). The AOX-PPRE was used

as a positive control for PPAR binding.

110

TSE-l TSE-Z

l 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

W

 

    

 

 

 

 

 

Figure 2.8. Gel Shift Analysis of TSE-l and TSE-Z. TSE-l
and TSE-2 were shifted with translated receptors. Lanes 1
and 10 contain the labeled TSE-l or TSE-2 oligonucleotides
only. Lanes 2 and 9 include TSE-l or TSE-2 with the control
translation cocktail. Lanes 3 and 11 contain Zug RAR, Lanes
4 and 12 contain Zpg RXR, Lanes 5 and 13 contain Zug PPARyZ,
Lanes 6 and 14 contain lug each of RXR and RAR, Lanes 7 and
15 contain lug each of RAR and PPARyZ and Lanes 8 and 16
contain lug each of RXR and PPARyZ. The arrow indicates the
site of RXR/PPARyZ heterodimer shifting TSE-l but not TSE-2.
This experiment was repeated at least 3 times and a
representative example is shown above. Other receptors

tested included FAAR, PPARa and HNF4, none of which bound
to TSE-l or TSE-2.

111

.AEM 3 RESULTS
Aim 3: Are PPARs involved in the PUFA control of 814 in
adipocytes?
PPARyZ and PUFA do not interact in adipocytes. The

region of PPARyZ binding (Aim 2) did not correlate with the
proposed PUFA target (Aim 1). Because the promoter targets
for these two factors did not bind the same regions (PPARyZ

bound a region between -1550 and -1530bp while the PUFA-RE
is between -220 and -80bp), it is unlikely that PUFA are

acting through a PPAR pathway. To further substantiate this
finding, adipocytes were treated with the PPARy2 activator,

pioglitazone. Pioglitazone, a thiazolidinedione, is known
to increase differentiation. Treatment of adipocytes during
the 48 hour differentiation period with pioglitazone did
indeed increase lipid accumulation as seen microscopically.
CAT activity ten days after differentiation was enhanced in
cells treated during differentiation with pioglitazone (not
shown). This was probably a result of better
differentiation. In general the fatter the cells and the

'more fat cells present, the higher the CAT activity
(observation). In contrast, activation of PPARyZ by

pioglitazone in the last 48 hours of culture in fully
differentiated cells had no effect on $14 CAT activity (Fig.

2.9).

112

 

Pioglitazone Dose Response In Adipocytes

 

............
............

.....

CATUMN

 

 

    

..............................
.............
.............

Mb ubjm “L0 50 W16-
mkanmmume

 

 

 

Figure 2.9. Pioglitazone Treatment of Adipocytes. Fully
differentiated adipocytes were treated for 48 hours with the
indicated dose of pioglitazone and CAT activity was
determined. The bars are means of three samples with
standard deviations shown. Several other experiments using

1-10uM pioglitazone duplicated these results.

In conclusion of Aims 2—3, PUFA and PPARyZ target

different sites in the $14 promoter. Pioglitazone had no

effect on CAT activity (Figure 2.9) while PUFA had a
suppressive effect (preliminary results). PPARyZ plays a

role in adipocyte differentiation and may also be involved
in $14 expression early in adipocyte differentiation.
However, in these studies, at the time when 514 shows PUFA
sensitivity, it does not show sensitivity to pioglitazone as
measured by CAT activity. These results do not indicate any

interaction between PUFA.and PPARmZ in regulation $14 gene

transcription in fully differentiated adipocytes.

113

AIM 4 RESULTS
.Aim.4: Does arachidonic acid regulate Sl4 gene expression
through a prostanoid pathway?

.Arachidonic Acid Inhibits Sl4CAT Expression Through a
Prostanoid Pathway. The goal has been to determine the
mechanism of PUFA regulation of adipocyte lipogenic gene
expression. The dose response studies have already shown
that in contrast to liver, 20:4,n-6 is a 6-fold more potent
inhibitor of lipogenic gene expression than 20:5, n-3. When
compared to 20:4,n-6, 20:5,n-3 is a poor substrate for the
synthesis of prostaglandins by cyclooxygenase 1 and 2
(Laneuville et al., 1995). Thus, the differential potency
of 20:4,n-6 and 20:5,n-3 suggest that 20:4,n-6 may be
converted to prostanoids which, in turn, induce changes in
adipocyte lipogenic gene expression. This is in keeping
with others who have established that adipocytes convert
20:4,n-6 to the prostaglandins, PGE2, PGF} and PGI; (Hyman et
al., 1982; Shillabeer et al., 1996; Smas and Sul, 1995;
Shillabeer and Lau, 1994).

To determine if 20:4, n-6 requires metabolism to
prostanoids, we used the cyclooxygenase inhibitor
flurbiprofen. Cells were also treated with the
nordihydroguaiaretic acid (NDGA) and clotrimazole,
inhibitors of lipoxygenase and monooxygenase activity,

respectively.

114

While flurbiprofen blocked the 20:4,n—6 inhibitory
effect on CAT activity, NDGA and clotrimazole had no effect
(Figure 2.10). Thus, 20:4,n-6—mediated inhibition of 814
gene expression requires cyclooxygenase and implicated a
role for prostanoids in regulating adipocyte lipogenic gene

expression.

 

Reversal of AA Suppression of S14

 

CATUMB

     

 

 

 

 

 

Alb AA AA+CI AA+F| AA+N AA+T

 

 

 

Figure 2.10. Flurbiprofen blocks the Arachidonic Acid
Suppression of Sl4CAT Activity. Fully differentiated
adipocytes were treated with the indicated compounds and

250uM arachidonic acid (except for albumin (SOuM) control)
for 48 hours and CAT activity was determined. This
experiment is a representative of at least three experiments
and each bar represents a mean of 3 samples with standard

deviation error bars. Cl= clotrimazole (lOuM), Fl=
flurbiprofen (100uM), N=NDGA (SOuM) and Tr= Triacsin C
(louM).

PGE2 suppresses both Sl4CAT activity and $14 mRNA. To

further examine this prostanoid dependent pathway of $14

gene suppression, the effect of specific prostanoids on

115

adipocyte lipogenic gene expression was evaluated by

treating L1 adipocytes with PGE2 or PGan at 1011M (Figure

2.11). Both PGE2 and PGng (not shown) inhibited CAT
activity. Dose response analyses show that PGE2 inhibits

Sl4CAT expression with an ED“,of ~5uM, a concentration well

below the EDso of ~50pM for 20:4,n-6. Analysis of 814 mRNA
following PGE2 treatment showed a similar decline (Fig
2.11). FAS mRNA showed a like results with PGE2 treatment
(not shown). Based on these results, I concluded that
20:4,n-6 is converted to prostaglandin in adipocytes and
that PGE2 and PGan inhibit adipocyte lipogenic gene
expression.

Signal Transduction Pathway for Plz Control of 814

Gene Transcription. PGE2 and PGF2a regulate cell function

through G-protein linked plasma membrane receptors (Smith,
1989; Asboth et al., 1996; Hamon et al., 1993; Nagai et al.,
1996; Uehara et al., 1994; Danesch et al., 1994). Depending
on the G-protein linkage, PGE2 can increase or decrease

cellular cAMP levels or elevate IP3 and Ca+2 levels.

116

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PGE2 Effects in Adipocytes
5
IRMK
cunIAmmw
i
DMSO PGE;
Figure 2.11. PGE2 Inhibition of Sl4CAT Activity and RNA in i

adipocytes. Fully differentiated adipocytes were treated

with DMSO or IOuM prostaglandin for 48 hours and CAT
activity was determined or $14 RNA was measured. CAT Units
were divided by 1000 for presentation. The CAT data is
representative of several experiments. RNA results are
representative of 2 separate experiments. The bars are
means of triplicate samples and include standard deviations.

PGan effects on CAT activity are identical but are not
shown.

To determine if changes in intracellular ca+2 or cAMP
was involved in prostaglandin-mediated suppression of $14
gene expression, L1 adipocytes containing the stably
integrated Sl4CAT fusion gene were first treated with A23187
(a calcium ionophore) or 8-CTP-cAMP and isobutylmethyl
xanthine (IBMX) to elevate intracellular calcium or cAMP
respectively (Figure 2.12). Treatment of cells with 8-CTP-
cAMP plus IBMX or the A23187 inhibited CAT activity by 350%.
These studies demonstrated that alterations in intracellular
cAMP or Ca'+2 markedly suppress Sl4 gene transcription in

fully differentiated adipocytes (Figure 2.12).

117

 

Inhibition of S14 CAT Activity

 

 

 

 

 

 

 

 

Veh ’ MR I A23187

 

 

 

Figure 2.12. Inhibition of 814 CAT Activity by cAMP and
A23187. Fully differentiated adipocytes were treated with

SOuM 8-CPT-cAMP + SOpM IBMX or luM A23187 for 48 hours and
CAT activity was determined. This graph is 5 experiments
combined and normalized to vehicle. Percent errors are
shown for each mean.

Depending on the G-protein linkage, PGE2 can activate
protein kinase A, protein kinase C or calcium-regulated
mechanisms. To determine which pathway affects $14 gene
expression, PGE2-treated adipocytes were treated with H7 and
staurosporin, inhibitors of A and C-kinases, or pertussis

toxin an inhibitor of G; /G°-linked processes (Figure 2.13) .

(h/Gd-linked processes promote a decrease in intracellular
.cAMP or an activation of phospholipase CA and elevation in

intracellular Ca+2 through release of inositol 1,4,5-
phosphate [IP3] (Uehara et al., 1994; Danesch et al., 1994).
A rise in cAMP is associated with inhibition of Sl4CAT
expression, therefore, pertussis toxin will provide a means

to evaluate Ca+2 -regu1ated processes. While PGE; treatment

118

of cells inhibits CAT by ~60%, co-treatment with H7 or
staurosporin at doses sufficient to inhibit both A and C-
kinases did not block the PGEzleffect (Fig. 2.13).

Treatment of cells with the an calcium channel blocker,
verapamil, also failed to block the PGEzieffect (Fig. 2.13).
Only pertussis toxin blocked the PGE2<effect on CAT activity
(Figure 2.13). Pertussis toxin also partially reversed the
PGE2 inhibition of 314 mRNA (not shown). Treatment of cells
with pertussis toxin, H7 or verapamil in the absence of PGE2
had no effect on Sl4CAT expression (not shown). Taken
together, these studies suggest that PGEQ operates through a
Ch/Gdrlinked signaling system which may involve changes in

intracellular calcium.

119

 

Reversal of PGE2 Effects

 

3

CAT Units
E

 

 

 

PGE; - + +PT +H7 +V

 

 

 

Figure 2.13. Effects of PGE2 and PT, H7 or Verapamil on
Sl4CAT Activity. Fully differentiated adipocytes were

treated without (--) or with (+) louM PGE2 alone or with
lOOng/ml pertussis toxin (PT), 10uM H7 or 10uM Verapamil (V)

for 48 hours before CAT activity was determined. This
experiment is a representative of at least 3 separate
experiments and each bar is a mean of 3 samples with
standard deviation error bars indicated.
AIM 5 RESULTS

Aim 5: : Are PUFAs altering lipid metabolism in adipocytes?

Effect of PUEA on Adipocyte Differentiation and Lipid
Accumulation. Saturated and polyunsaturated fatty acids
have been reported to stimulate adipocyte differentiation in
081771 cells (Ailhaud et al., 1992; 1995; 1996). To
determine if PUFA affected adipocyte differentiation,
specific fatty acids were added to the medium after removal
of the differentiation medium and maintained in the medium

for 10 days at SOuM. Continuous treatment of cells with

SOuM fatty acid had no obvious effect on differentiation as

120

assessed by triglyceride accumulation (Table 2.2), oil red
staining (Fig 2.5) or total fatty acids (Figure 2.14).

Fatty acid analysis of cells treated with the specific fatty
acids showed an expected accumulation of the specific fatty
acid in the triglyceride/phospholipid pool. For example,
20% of the fatty acids within the phospholipid/triglyceride
pool was 20:4,n-6 in cells receiving continuous 20:4,n-6
treatment. In contrast, cells treated with either no fat or
18:1,n-9 show less than 3% 20:4,n-6 (Table 2.2).

A corresponding decline in $14 and FAS mRNAs was also
found following continuous fatty acid treatment like that
seen in Figure 2.1 (not shown). The decline in FAS and 814
gene expression would suggest that de novo lipogenesis was
inhibited in these cells. However, the triglyceride levels
and the level of 16:0 were unaffected by these treatments.
Presumably, these fatty acids accumulate from the fetal calf
serum supplemented in the culture medium. Thus, in the
presence of a continuous supply of exogenous PUFA,
inhibition of de novo lipogenesis may have little impact on
triglyceride accumulation in adipocytes.

Another observation is that in cells treated with
18:2n-6 or 18:3n-6 (Figure 2.14) show no change in 20:4n—6
levels. This indicates a lack of subsequent desaturation

and elongation of the exogenously supplied 18-carbon PUFA.

121

Thus, the exogenous FA enter either the TG/phospholipid pool

for storage or are not further remodeled.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 2.2. Total Triglycerides in Adipocytes
Treatment I rcs Inn: 18:1 10:2 18:3 I 20:4 I 20:5
Total T6: I I 34:5 37il| 3311 403:4 [ 35:1:8 I 39:7

Patty Percent of total fatty acid
Acid
16:0 59.8 55.3 46.7 44.4 46.3 45.3 46.2
18:0 10.4 9.7 8.6 8.8 8.3 10.5 8.0
18:1n-9 20.9 24.5 32.9 19.1 17.9 18.2 19.7
18:2n-6 1.9 2.0 2.1 16.0 4.1 1.8 1.8
18:3n-6 0.7 1.0 0.5 0.9 11.0 0.6 0.4
18:3n-3 0.4 0.8 1.5 1.6 4.0 0.4 0.5
20:4n-6 3.3 3.5 3.0 3.8 4.2 20.0 2.7
20:5n-3 0.7 1.1 2.9 3.4 2.8 1.2 19.0
22:6n-3 1.8 2.2 1.7 2.0 1.4 2.0 1.6
Table 2.2. Total Triglycerides in Adipocytes. Total

triglycerides were determined as described in Materials and
Methods. FCS:fetal calf serum, Alb:albumin. There was no
statistical differences between treatments for total TG.

The lower part of the table lists the percent of total fatty

acids after continuous treatment with 50uM of the indicated
fatty acid throughout differentiation.

122

Figure 2.14. Fatty Acid Profiles in Adipocytes.
Immediately following removal of the differentiation mix,

cells were treated with the fatty acids at SOuM continuously
until harvest for fatty acid analysis (10 days). All fatty .

acid treatments contained lOuM albumin in DMEM containing
PBS. The fatty acid treatments are as acids which were
analyzed are shown on the graph and are all n-6 fatty acids
except 20:5 and 22:6 which are n-3. Each bar represents

7 samples and standard error are shown. The Y-axis shows the
percent of all the fatty acids in sample.

123

@HNN now You

 

 

gigwm

gosgm
o... .8822... a
28.235
swoon
c.5320

 

 

92. Now «How one tow

 

858.3 s 8&2“. an! ace“.

«3.0..

cWW mes M0110 Named

124

DISCUSSION

Treatment of adipocytes with fatty acids have profound
effects on differentiation and gene expression (Ailhaud et
al., 1992; 1995; 1996; Tontonoz et al., 1994c; Forman et
al., 1995; Long and Pekala, 1996; Vassaux et al., 1992).

The principal mechanisms for the effects of fatty acids on
adipocyte gene expression were thought to involve either
prostanoid or PPAR-mediated pathways. Our studies show that
the dominant negative suppression of lipogenic gene
expression by PUFA in 3T3-L1 cell line is mediated by a
prostanoid pathway and not through a PPAR-mediated pathway.
PUFA suppress $14 and FAS mRNA levels and Sl4CAT expression
in cultured 3T3-L1 adipocytes and this effect occurs only in
fully differentiated adipocytes. The suppression of $14
gene transcription is 6-fold more sensitive to 20:4,n-6 than
20:5,n-3 and can be blocked by the cyclooxygenase inhibitor,
flurbiprofen. These results are consistent with a
prostanoid-mediated pathway for the control of lipogenic
gene expression. Both PGE2 and PGan inhibit lipogenic gene
expression and this effect is blocked with pertussis toxin
implicating an Gi/Go-coupled signaling pathway.

Cyclooxygenases convert 20:4,n-6 to prostanoids and
prostanoids elicit a wide range of biological effects

through plasma membrane receptors coupled to G-proteins.

125

Prostaglandins are produced locally and act in a paracrine
fashion (Smith, 1989). PGE2, PGIziand PGE} are the most
predominant prostaglandins produced in 3T3-L1 cells,
although more prostaglandins are produced in preadipocytes
than adipocytes (Hyman et al., 1982; Long and Pekala, 1996;
Vassaux et al., 1992; Richelsen et al., 1992; Christ and
Nugteren, 1970). Long and Pekala showed that GLUT4
expression was depressed by 20:4,n-6 and PGE2 and further
showed that cAMP was produced in adipocytes upon treatment
with PGE2. While activation of protein kinase A by cAMP has
a strong suppressive effect on Sl4CAT expression in L1
adipocytes (Fig. 2.13), treatment of adipocytes with the
phorbol ester, PMA, an activator of PKC (not shown), had no
effect on CAT activity. Treatment of cells with inhibitors
of both protein kinase A and C, i.e. H7 and staurosporin,
failed to block the PGE2-mediated inhibition of CAT
activity. Therefore, it is unlikely that the PGE2-mediated
inhibition of CAT activity is due to an increase in cAMP
and/or subsequent activation of protein kinase A or C. In
contrast, treatment of cells with pertussis toxin, which
blocks Gi/Go-linked pathways, was the only agent found to
reverse the PGE2 inhibition on Sl4CAT activity (Figure
2.14). Gh/Gd-linked signaling pathways inhibit Gg-linked
activation of adenylate cyclase or promote an elevation in

inositol 1:4.5-phosphate (IP3) and intracellular Ca“:2

126

(Richelsen et al., 1992). As a decrease in cAMP would be
expected to induce Sl4CAT expression, the involvement of the

Ch/Gd—linked pathway implicates activation of phospholipase
CB: release of 1P5 and elevation of intracellular free Ca”K

Treatment with the calcium ionophore A23187 also suppresses
Sl4CAT expression (Figure 2.13). The control of $14 gene
transcription by calcium will be addressed in future
research.

Some studies suggest PGEziaction on adipocytes is anti-
lipolytic and acts via blocking an increase in cAMP (Vassaux
et al., 1992; Christ and Nugteren, 1970; Chatzipanteli et
al., 1992). Others suggest that PGEzacts through protein

kinase C via a receptor coupled to phospholipase C in Swiss
3T3 fibroblasts (Danesch et al., 1994). PGFa,treatment of

3T3-L1 preadipocytes leads to an increase in intracellular
calcium, activation of calcium/calmodulin dependent protein
kinase (CaM kinase) and inhibition of differentiation
(Miller et al., 1996). Ntambi and Takova (1996) reported
that A23187 inhibited differentiation of 3T3-Ll cells but
had no effect when used two days after differentiation.
These confusing reports make it difficult to determine how
prostaglandins act on only differentiated adipocytes. In
contrast to the Long and Pekalas’ (1996) studies, my results
suggest that the inhibition of Sl4CAT expression is not due

to an increase in cAMP but through a separate pathway,

127

 

perhaps calcium linked. Further research is required to
definitively show that PGE2 suppresses Sl4 gene
transcription via a calcium linked pathway.

Another possibility is that prostaglandins are working

through a nuclear receptor pathway. Prostaglandins are
reported to bind PPARyZ, a nuclear receptor (Spiegelman and
Flier, 1996; Kliewer et al., 1997; 1995; Forman et al.,
1997) . While PPARyZ plays an important role in
adipogenesis, which includes the induction of lipogenic
enzymes, these results show that pioglitazone-activated

PPARmZ has no effect on Sl4CAT expression in fully

differentiated adipocytes. Though PPARyZ did bind to the

$14 promoter, this data does not indicate that PGEzior
arachidonic acid are acting through a PPAR mediated pathway
to suppress $14 in fully differentiated adipocytes.

Finally, my results would appear to conflict with
reports by others documenting that rats fed vegetable oils
do no show significant changes in adipose tissue lipogenesis
(Clarke et al., 1976; 1977). One explanation is that the
.dose of AA used here is higher than in vivo levels. Our

results do agree with these in vivo studies by showing that
linoleic (18:2n-6) and a-linolenic (18:3n-3) and y—linolenic
(18:3n-6) acids, found in vegetable oils, were ineffective

inhibitors of Sl4CAT activity (Figure 2.4). Interestingly,

128

epididymal fat mRNAs encoding 814 and FAS were suppressed
50% after 5 days on a diet containing 10% fish oil (Jump et
al., 1993). EPA, a component of fish oil, effectively
suppressed $14 and FAS mRNAs (Figure 2.1) and CAT activity
(Figure 2.2) in adipocytes. When compared to 18-carbon
PUFA, the highly unsaturated fatty acids found in fish oil
were more potent suppressers of lipogenic gene expression
(Clarke and Jump, 1994; Jump et al., 1996; 1994). Thus any
conflict with earlier reports can be explained by the use of
fish oil vs. vegetable oil for in vivo studies and the dose

of AA or EPA used in adipocyte cell culture.

CKNRIBUSHKNNS

Aim 1 of this chapter was to determine the PUFA
response region in the $14 promoter. Previously, this
laboratory reported the PUFA response region (PUFA-RR) was
located in the proximal promoter region )-220/-80 bp). The
work presented here suggests that in adipocytes the same
PUFA-RR is targeted by 20:4,n-6/PGE2. However, I was unable
to define further the specific PUFA target in adipocytes.
'This aim will be addressed again in the next chapter.

The second and third aims assessed the involvement of

PPAR as the mediator of PUFA action on. Using gel shift
analysis, PPARyZ was found to bind the $14 promoter in the

carbohydrate/glucocorticoid region, far upstream of the

129

proximal promoter. However, pioglitazone, a PPARyZ

activator, had no effect on $14 gene expression in fully

differentiated adipocytes. Therefore, I concluded that
although PPARyZ does bind the 814 promoter, PPARyZ does not

mediate the PUFA control of this gene.

Aim 4 determined how arachidonic acid suppressed $14
gene expression. These results show that $14 is suppressed
by AA in cultured 3T3-L1 adipocytes, and this effect is
reversed with flurbiprofen, a cyclooxygenase inhibitor.
Furthermore, the PGEzsuppression of $14 was reversed by‘
pertussis toxin, linking the suppression to a Gi/Go
signaling cascade and not a PPAR mediated pathway.

Aim 5 was originally intended to address the visible
media changes in fatty acid treated cells. However,
analysis of total triglycerides and staining of treated
cells showed no changes, regardless of treatment or
treatment time in adipocytes. Fatty acid analysis of
continuously treated adipocytes indicated that whatever fat
was fed to the cells, the cells contained (“you are what you
eat” hypothesis). There was no indication that triglyceride
levels changed even when lipogenic gene expression was
suppressed by >80% by the fatty acid treatment. These
observations indicate that while a specific fatty acid can
alter lipogenic gene expression, adipocytes still accumulate

fat from the diet.

130

In summary, my results show that lipogenic gene
expression is suppressed by PUFA in cultured 3T3-L1
adipocytes. The results presented here are consistent with
a requirement for AA conversion to prostanoids. The
prostanoids may activate a Gy“% linked signaling cascade
that leads to the inhibition of $14 gene transcription.
This mechanism for control is different for PUFA-mediated

suppression of lipogenic gene expression in the liver.

 

131

CHAPTER 3

(HEAEHHMR 3

HEPATOCYTES

Like the adipocyte, the liver is a major site for lipid
metabolism. Results in Chapter 2 indicated that arachidonic
acid suppressed $14 gene transcription through a prostanoid
pathway in adipocytes. How PUFA regulate 814 in the liver
is unknown, although a PPAR-dependent pathway has been ruled
out (Ren et al., 1996; 1997). Flick et al. (1977) reported
that treatment of rats on an n-6 fatty acid diet with
indomethacin failed to block the n-6 PUFA suppression of
fatty acid synthase (FAS) activity, indicating a prostanoid
independent pathway of PUFA suppression. Therefore, the
hypothesis for Chapter 3 is “PUFA regulation of hepatic
lipogenic gene expression is through a PPAR and prostanoid
independent pathway”.

As in the adipocyte, PUFA suppress hepatic $14 gene
expression. In the rat, the liver rapidly responds to a
diet containing n-3 PUFA by suppressing $14 and FAS
expression (Jump et al., 1993). Several hepatic genes are
regulated by PUFA including FAS, malic enzyme (ME),

glucokinase (GK), citrate lyase, acetyl CoA carboxylase

132

(ACC), pyruvate kinase, A? desaturase, S14 (Clarke and Jump,

1994; Clarke and Abraham 1992; Clarke et al., 1990a), AOX
and CYP4A (Ren et al., 1996; 1997). In rats fed a fish oil
diet for five days, the hepatic expression FAS, 814, ME,
pyruvate kinase and glucokinase are all depressed compared
to olive oil fed rats (Jump et al., 1994). In contrast, AOX F
and CYP4A mRNA expression in the liver increases in mice fed 5

n-3 PUFA compared to chow fed mice (Ren et al., 1997).

 

As discussed in Chapter 1, $14 gene transcription is _
controlled by a variety of factors in the liver, namely
.insulin, thyroid hormone and dietary components. Because
PUFA control of adipocyte Sl4 gene expression was found to
require cyclooxygenase, Aim 1 will address this requirement
in hepatocytes. Subsequently, Aim 2 will address the
prostanoid control of hepatic $14 expression.

As discussed in Chapter 2, results suggested that the
PUFA-RR was located between -220 and -80bp in the $14
promoter. Results in adipocytes were not conclusive with
further efforts to locate a more specific PUFA-RR. Aim 3
will use transfected hepatocytes to further narrow the
promoter region targeted by AA and other suppressive

compounds.

133

Aim.1: Does arachidonic acid regulation of $14 gene
expression in hepatocytes require cyclooxygenase?

Aim 2: Are PG regulating 814 in hepatocytes and if so, how?
Aim 3: Where is the cis-regulatory element required for AA

and PGE; regulation of S14 gene transcription?

INTRODUCTION

The effect of dietary n-6 polyunsaturated fatty acids
(PUFA) on hepatic de novo lipogenesis was first reported by
Allman and Gibson (1965) while studying the effects of
essential fatty acid deficiency on hepatic lipid metabolism.
Those studies showed that removal of n-6 PUFA from the diet
led to a rise in de novo lipogenesis. In subsequent
studies, Flick, et al. (1977) reported that administration
of indomethacin in vivo failed to block the n-6 PUFA
suppression of fatty acid synthase (FAS) activity. N-6 PUFA
are precursors of prostaglandins and indomethacin inhibits
prostaglandin synthesis. These observations led
investigators to suggest the n-6 PUFA regulation of hepatic

lipogenesis did not involve prostaglandins.

134

However, recent studies on the effects of n-6 PUFA on

lipogenic gene expression in cultured adipocytes indicated
20:4,n-6 suppressed (EDu,~»50 uM) mRNAs encoding FAS and

the 814 protein. The 20:4,n-6 effect on $14 CAT activity

could be blocked by flurbiprofen, a non-specific

cyclooxygenase inhibitor (Chapter 2). PGE2 and PGE;Cl also

suppressed Sl4 mRNA (ED“K10uM). Using a cell line

containing a stably transfected Sl4CAT fusion gene as a
monitor of transcriptional effects, PGEziwas found to
suppress Sl4CAT activity through a pertussis-sensitive
Ch/Gd-linked regulatory pathway. Thus, the 20:4, n-6
control of lipogenic gene expression in 3T3-L1 adipocytes
required cyclooxygenase.

Based on these findings and the fact that primary
hepatocytes have prostanoid receptors and are responsive to
prostanoids, I reexamined the requirement of cyclooxygenase
in the n-6 PUFA regulation of hepatic lipogenic gene
expression. Here, results show that the 20:4, n-6
suppressed lipogenic gene expression in the hepatic
,parenchymal cell does not involved metabolism through a
cyclooxygenase. However, PGEQ suppresses mRNAs encoding
FAS, 814 and L-pyruvate kinase (L-PK). Using the Sl4CAT as
a monitor for transcriptional control, these results show
that the cis-regulatory targets for PGEziand 20:4, n-6 map

to the PUFA-response regions within 514 proximal promoter (-

135

220 to -80bp). These studies indicated that prostanoids also
affect hepatic lipogenic gene expression and that PUFA and
prostanoids might utilize a common pathway at the genomic

level to control 814 gene transcription.

IUNTERUJUUS AmmilflETHIEHB

EEPATOCYTE PREPARATION AND CULTURE A complete protocol
for hepatocyte preparation is given in Appendix B. Briefly,
male Sprague—Dawley rats were starved overnight then
anesthetized using Nembutol (interperitoneal injection) or
methoxyflurane (inhalation). After shaving and washing, the
skin and muscle layers were cut back to reveal the hepatic
vein. The vein was catheterized and the liver perfused with
a wash buffer then a digestion buffer containing either
collagenase or Liberasem (Boehringer Mannheim). The liver
was removed and hepatocytes were separated through a Percoll
cushion. Hepatocytes were plated at 3 million cells per
60mm plate for CAT assays or 10 million cells per 100mM
plate for RNA. Cells were plated in Williams E media
Containing 10% fetal bovine serum, 10mM lactate, 200nM

insulin and 10nM dexamethasone. Transfection (4-6 hours

later) with reporter plasmids (2ug/plate), thyroid hormone

receptor expression vector (lug/plate) and Lipofectinm

136

(6.6pl/pg DNA) was performed in the same media as plating
media but without serum. The next morning media was changed
Williams E media containing 25mM glucose, lpM insulin and

10nM dexamethasone unless otherwise noted. Media was
changed after 24 hours; all experiments were terminated
after 48 hours.

RNA ISOLATION AND NORTHERN BLOTTING Total RNA was
isolated using Triazol" (Gibco) according to the
manufacturer's protocol. RNA was electrophoresed,
transferred to nitrocellulose membranes and probed with
radiolabelled cDNA as described in Chapter 2 (Materials and
Methods). Radiolabelled probes used here include those
listed in Chapter 2 as well as AOX (T. Osumi, Japan),
cyclooxygenase I and II (D. Dewitt, Michigan State
University) and CYP4A2 (A. Thelen, Michigan State
University).

CAT ASSATS Following treatments, cells were assayed
for CAT activity and protein content as previously described

(Jump et al., 1993). CAT Units: 1"C-acetylated

chloramphenicol CPM/100 pg protein/hour.

137

RIEHIUTS

AIMII RESULTS

Aim 1: Does arachidonic acid regulation of $14 gene
expression in hepatocytes require cyclooxygenase?

814 suppression by arachidonic acid is not reversible
by flurbiprofen. Based on the observation in adipocytes 814
CAT activity suppression by AA could be reversed by the
cyclooxygenase inhibitor, flurbiprofen, reversal of Sl4CAT
activity in hepatocytes by flurbiprofen was also tested.

Primary hepatocytes were transfected with Sl4CAT124 and
treated with 50pM Albumin or 250pM AA with or without lOOpM

flurbiprofen for 48 hours. As expected, AA inhibited CAT
activity approximately 60% but flurbiprofen could not
reverse this inhibition. The lipoxygenase inhibitor, NDGA,
was also unable to reverse the AA suppression of $14 CAT
activity (data not shown). These results indicate that AA
suppression of $14 gene expression does not require
metabolism through the cyclooxygenase or lipoxygenase

pathway (Figure 3.1).

138

 

Flurbiprofen and AA in Hepatocytes

 

 

1.2
1.0 <-
g 0.0 ..
a
5 00
=3 '°‘

 

 

 

 

.0
N

 

 

 

 

 

.0
o

 

 

No Flurbiprofen 100pM Flurbiproien

 

 

 

Figure 3.1. AA Suppression of Sl4CAT Activity is Not
Reversed by Flurbiprofen. Hepatocytes transfected with

Sl4CAT124 (-4315/+19) were treated with 250pM OA or AA alone
or with 100pM flurbiprofen for 48 hours. CAT activity was

determined and all data normalized to the OA treatment. The
bars represent 3 pooled experiments and standard error is
included. No differences existed between treatments with or
without flurbiprofen.
AIMIZ RESULTS

Aim.2: Are PG regulating $14 in hepatocytes and if so, how?

Pug suppresses 814 message and CAT activity. PG
suppressed $14 mRNA and CAT activity in adipocytes (Chapter
2). Therefore, similar experiments were set up with

hepatocytes to test suppression of $14 mRNA and CAT

activity. Figure 3.2 shows that 314 message in hepatocytes
is suppressed 66% by 10pM PGEziafter a 48 hour treatment.

Other mRNAs were also measured in response to this

treatment. FAS mRNA levels were suppressed ~60% and PK

139

~35%. AOX and CYP4A2 were not effected by the treatment.
These results indicate that like PUFA, PGEzican depress mRNA
levels for the lipogenic genes FAS, PK and 814. However,
unlike EPA which induces AOX and CYP through PPAR (Ren et
al., 1996; 1997), the oxidative genes AOX and CYP are not
effected by PGE2 treatment in hepatocytes. Interestingly,
like PGE2,.AA also has no consistent effect on AOX and

CYP4A2 mRNA in hepatocytes (data not shown).

 

 

 

 

 

 

 

mRNAs in Hepatocytes
25“ nvume
use:
2 3AA
£3
5 I
15
z
E w
0.5
o

 

 

 

 

 

 

 

 

 

 

 

 

CYRMU!

s14 ' FAS

 

 

 

Figure 3.2. PGE2 Effects on Hepatic Gene Expression.

Hepatocytes were treated with DMSO or 10pM PGE2 for 48
hours. Total RNA was isolated, transferred and probed for
$14, FAS, PK, AOX and CYP4A2. This graph represents pooled
data from 2 separate experiments of triplicate plates each.
The data was normalized to the control (DMSO). Standard
error of the pooled PGE2 data is indicated.

140

As described in Chapter 2, $14 CAT activity in
adipocytes was inhibited by PG treatment. To examine the
effects of prostanoids on hepatic gene transcription,
hepatocytes were treated with various prostaglandins to
determine if CAT activity was also suppressed in these cells
transfected with $124 (-4315/+19 bp). The results are shown
in Figure 3.3. Both PGE2 and PGan. inhibit Sl4 CAT activity
by about 54% and 43% respectively in transfected
hepatocytes. PGI; is very unstable and was inhibitory only
when freshly prepared (data not shown). These results

indicate that PG can suppress $14 transcription in

hepatocytes.

 

CAT Activity in Hepatocytes

 

 

Pmumuoflkmma'

.0
N
on

 

 

   

0.00 -

 

 

 

PGE2 PGan

 

 

 

Figure 3.3. Inhibition of $14 CAT activity by PG. Primary
hepatocytes transfected with $124 and treated with the

indicated PG at 10pM for 48 hours. CAT activity was
determined and data from several experiments were pooled and
normalized to the DMSO control. Standard error is shown.

141

Pug suppression of CAT activity was not affected by
inhibitors of signal transduction pathways. In adipocytes,
the PG suppression of 514 CAT activity was reversed by
pertussis toxin. Several inhibitors were also used here in
attempts to block the PGE2 inhibition of $14 CAT activity in
hepatocytes (inhibitors used in both cell types are
summarized in Table 3.1). The results of the hepatocyte
studies are shown in Figure 3.4. The PGE2 suppression was
reversed by pertussis
toxin treatment in adipocytes, however, in hepatocytes this
was not the case. As shown in Figure 3.4, the PGE2
suppression was not reversed by pertussis toxin. Pertussis
toxin blocks 6;, a G-protein which inhibits cAMP formation.
Pertussis toxin was used at 100ng/ml and the cells treated
with the PGE2 concurrently. Cells were also pretreated with
pertussis toxin from 15 minutes to 12 hours. However,
pretreatment of hepatocytes with pertussis toxin also had no
effect and could not consistently block the inhibition by

PGE2 or AA (not shown).
H7, a PKA/PKC inhibitor at 10pM, also had no effect.

Other PKA or PKC inhibitors were unable to reverse the PCB;

effect as well. These are shown in Table 3.1. Verapamil,

an ad calcium ion channel blocker, also was unable to

reverse the inhibition by PGE2. Verapamil was used at 10pM'

and had no effect. Chelating extracellular calcium with

142

EGTA was toxic to the cells (not shown). Each of these
compounds used were tested alone and used at the highest
nontoxic dose, measured both by protein levels and CAT

activity.

 

Reversal of PGE2 Effect in Hepatocytes

 

750 ..

CAT Units

250 i

 

 

 

 

0
Fee, " + +PT +H7 +v

 

 

 

 

Figure 3.4. Effects of PGE2 and Inhibitors on $14 CAT
activity. Hepatocytes were transfected with Sl4CAT124 and
treated for 48 hours without (--) or with (+) 10pM PGE2 and
also plus 100ng/ml pertussis toxin (PT), 10pM H7 or 10pM
verapamil. CAT activity was determined. This is a
representative graph of at least three separate experiments
and includes the standard error of each triplicate on each

bar. CAT units = cpm/hour/lOOpg of protein.

143

Table 3.1.

PGE2 SIGNAL TRANSDUCTION INHIBITORS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DRUG(SOLVENT) ACTION ADIPOCYTE EEPATOCYTE
RESULT RESULT
H7 (water) inhibits PKA no effect at no effect at
' and PKC at 10“” 10pM
conc. used
H89 inhibits PKA no effect at no effect at
(water) specifically 10pM 10pM
Staurosporin inhibits PKA no effect at no effect at
(DMSO) and PKC 10nM 10nM
KN-62 inhibits Caz" inhibitory NA
(DMSO) calmodulin but slight
kinase II rev. of PGE2
at 25pM
Verapamil blocks «1 Ca some reversal slight to no
(DMSO) ion channels Of PGE reversal Of
inhibition at PGE inhib.
10pM at 10pM
BAPTA-AM blocks no effect at no effect at
(DMSO) intracellular 10pM 20pM
Ca release
Thapsigargin releases toxic at NA
(DMSO) intracell. Ca, 170nM
and inhibits
Ca ATPase
TMB-8 (water) Ca antagonist NA no effect at
and PKC inhib. 15pM
-blocks Ca
release from
ER
Pertussis blocks Gi/q/o reverses PGE2 no
Toxin proteins effect at 25 consistent
(glycerol) to effect even
100ng/ml with pretrt.
PD980059 blocks MEK, or NA no reversal
(DMSO) p44 (a MAPK) of PGE effect
up to 50pM w/I
pretreatment
Table 3.1. PGE2 Signal Transduction Inhibitors. The listed

agents were used at the stated concentrations with 10pM PGE2
in either adipocytes or hepatocytes. CAT activity was
determined and results are described in the table. All
experiments were done at least twice. NA= not applied to
that cell system.

144

PGE2 acts through an EP3 cell surface receptor. In
adipocytes, all three PG receptor agonists were effective in
suppressing $14 CAT activity (not shown). In order to
determine which prostaglandin receptor was active in
hepatocytes, agonists to each of the PGE2.neceptors were
used to treat cells. As shown in Figure 3.5, only
sulprostone, an EP3 receptor agonist, was able to mimic the
suppression shown by PGEi. This indicates that EP3
receptors are present and active in these hepatocytes while
EPl and EP2 are not active in suppressing Sl4 CAT activity.
In contrast, the adipocytes appeared to be sensitive to all
the agonists, indicating each signal transduction pathway
(cAMP increase, cAMP decrease and IP3/calcium levels) was
operative in this cell type. A summary of the prostaglandin
receptor agonist actions and responses of both cell types
are also shown in Table 3.2. The structure of the different

prostaglandins and PG agonists are shown in Figure 3.6.

145

 

EP Receptor Effect in Hepatocytes

 

.0

FdHCATAdMW'

 

 

 

DMSO EP1 EP2 EP3

 

 

 

Figure 3.5. PGE2 Receptor Agonist Effects on $14 CAT
Activity. Hepatocytes, transfected with Sl4CAT124, were

treated with DMSO or 10pM EP1, EP2 or EP3 agonists. EP1=17-
phenyl trinor PGE2, EP2=11-deoxy 16,16 dimethyl PGE2,
EP3=sulprostone. The CAT data was pooled from several
experiments and normalized to the control (DMSO). Standard
errors of the pooled data is included on the graph.

146

 

Table 3.2. PROSTAGLANDIN AGONISTS EFFECTS ON 814 CAT
ACTIVITY IN ADIPOCYTES AND HEPATOCYTES

 

 

 

 

 

 

 

 

 

 

 

 

DRUG(SOLVENT) ACTION ADIPOCYTE HEPATOCYTE
RESULT RESULT

PGE2 (DMSO) prostaglandin inhibitory inhibitory
at 10pM at 10pM

pGan (DMSO) prostaglandin inhibitory inhibitory
at 10pM at 10pM

PGI; (DMSO) prostaglandin inhibitory inhibitory
at 10pM or at 10pM or
unstable unstable

17-phenyl EP1 receptor inhibitory no effect at

trinor PGE2 agonist at 10pM 10pM

(DMSO)

ll-deoxy EP2 receptor inhibitory no effect at

16,16 dm PGE2 agonist at 10“” 10pM

(DMSO)

Sulprostone EP3 receptor inhibitory inhibitory

(DMSO) agonist at 10pM at 10pM

17-phenyl FP receptor inhibitory inhibitory

trinor PGan agonist at at 10pM

(DMSO) 10pM

 

 

 

Table 3.2. Prostaglandin Agonists Effects on $14 CAT
Activity in Adipocytes and Hepatocytes. A summary of the
different prostaglandins used in both cell types are shown
here. Cells were treated with 10pM PG and CAT activity was
determined. Each of the above PG were used in at least
three different experiments in each cell type.

147

Prostaglandins

Prostaglandin 82
(P532)

Prostaglandin I;
(PGI2)

11-deoxy-16,16~dimethyl PGE;
(EP1.Agonist)

17-pheny1 trinor PGE2
(EP2 Agonist)

Figure 3.6.

above.
Chemical

PG and Agonist Structures.

Prostaglandin F2.
(PGFZe)

OH

”on / .

OH

16,16-dimethyl PGE2

Sulprostone
(EP3 agonist)

0 h

o l 0
w 2 ’~\;’:-c”3
/ // \> O
m o—
E.

O"

17-phenyl trinor PGI-‘2.
(FP Agonist)

 

The chemical
structure of each prostaglandin and agonist is depicted

These structure models were copied from Cayman
(Ann Arbor, Michigan) product literature.

148

 

 

 

 

 

 

.AA Suppression of 814 CAT activity could not be
reversed in hepatocytes. No tested inhibitor could reverse
the PG inhibition of $14 CAT activity in hepatocytes. Also,
although both AA and PG inhibited $14 expression and CAT
activity, AA suppression is not dependent upon PG production
in hepatocytes (Figure 3.1). With these facts, I tested
some other agents with the intention of trying to reverse
the AA inhibition of CAT activity. These are shown in Table
3.3. In most cases, the highest non-toxic dose was used.

In all cases, this table represents CAT activity in
hepatocytes or adipocytes transfected with Sl4CAT124, which
contains the entire 814 promoter (-4300bp).

As shown in Table 3.2, only flurbiprofen reversed the
inhibition of Sl4CAT activity and only in adipocytes.
Although the efficacy of the drugs were not tested, each was
used at the highest nontoxic dose and at levels comparable
to reports in the literature. Perhaps arachidonic acid
itself rather than a specific metabolite can regulate $14
gene expression in hepatocytes. Further research will be
required to determine how both AA and PG are acting to

Suppress $14 gene transcription in hepatocytes.

149

 

 

 

 

 

 

 

 

 

 

 

Table 3.3. AA METABOLISM INHIBITOR EFFECTS ON 314 CAT
ACTIVITY
DRUG(SOLVENT) ACTION ADIPOCYTE EEPATOCYTE
RESULT RESULT
Flurbiprofen blocks COX I 100pM: no effect at
(DMSO) and II reverses AA 100pM
inhibition
of CAT
activity
NS-398 (DMSO) blocks COXII toxic NA
Clotrimazole inhibits slight can be toxic
(DMSO) P450 group 2 reversal of at 15pM but
epoxygenases AA removes FA
inhibition effects
at 50pM
Aminobenzotriaz suicide NA no effect at
ole inhibitor of 15pM, may
(DMSO) P450 mono- increase CAT
oxygenases activity
NDGA inhibitor of no effect at no effect at
(DMSO) lipoxygenase 100pM 50pM
s but also
COX, EPOXYs .
Triacsin inhibitor of no effect at no effect at
(DMSO) FA -CoA 10pM 10pM
formation,
PPAR
activator
Vitamin E anti-oxidant NA no effect up
(DMSO) to 100pM
b- stops NA no effect up
Mercaptoethanol peroxide to 100pM
(water) formation
from FA
Nembutol induces NA no effect up
(pentobarbital) CYP2A1 & 2B1 to 100pM
Miconazole blocks P450 inhibitory NA
(DMSO) monooxygenas to CAT
es activity at

 

 

 

lSpM

 

 

Table 3.3. AA Metabolism Inhibitor Effects On 814 Cat

Activity.

The listed agents were used to treat adipocytes

or hepatocytes also treated with 250pM AA at the

concentrations listed.
result is listed the table.

CAT activity was measured and the
Most experiments were done at

least twice. NA = not applied to that cell system.

150

 

AIMIS RESULTS

Aim.3: Where is the cis-regulatory element required for AA
and PGE2
regulation of $14 gene transcription?

PGE2 inhibition is specific to the 814 promoter. To
examine the effects of PGEzion parenchymal cell gene
transcription, primary hepatocytes were transiently
transfected with either an Sl4CAT reporter gene ($170) or
RSVCAT. 5170 contains the $14 promoter extending from +19
bp to -2.8 kb. This promoter contains a proximal promoter
region required for proper initiation of gene transcription
(MacDougald and Jump, 1991) and two enhancers. One enhancer
located between -1.6 and 1.4 kb is a target for insulin and
glucose induction of $14 gene transcription. A second
enhancer located between -2.8 and -2.5 kb contains 3 thyroid
hormone response elements (TRE) which are targets for
thyroid hormone receptors (TR) binding in association with
the retinoid X receptor (RXR). Cells transfected with $170
and treated with T3, insulin and glucose express high levels
of CAT activity. RSVCAT contains the RSV promoter/enhancer
fused to CAT and cells transfected with this plasmid

constitutively express high levels of CAT activity. PGE2

(10 pM) suppressed Sl70 CAT activity by ~50% (Figure 3.5).

151

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Promoter Specificity by PGE2
140
120 0 .maso
area
3 100 .
so ..
b
so ..
g T
e 40 ._ l
20 ..
o .
s14CAT124 RSVCAT

 

 

 

Figure 3.7 Specificity of PGE2 on the $14 Promoter.
Hepatocytes were transfected with the above plasmids and

treated without or with 10pM PGE2 for 48 hours. CAT
activity was determined and normalized to the control. This
is a representative graph of at least 3 separate
experiments.

Arachidonic acid and PGE2 target the same regions of
the 814 promoter. Although both PGE2 and AA both inhibit
Sl4 CAT activity, AA does not require conversion to PG to be
inhibitory to this gene. This experiment was designed to
determine if PGEziand AA targeted the same site in the $14
promoter in hepatocytes. Several 814 promoter constructs
were transfected into primary hepatocytes which were treated
with either arachidonic acid or PGE2. .As shown in Figure
3.8, arachidonic acid progressively shows less inhibition in

CAT activity as the promoter is truncated from -2897 to -80

152

base pairs. The greatest inhibition (78%) is with $170, and
no inhibition is detectable with $158, which only includes -
80 base pairs and the TRR. About 50% inhibition is seen
with 814155, approximately 2/3 of that seen with 8170.

PGE2, is less inhibitory even in 8170 than arachidonic acid
(50% vs. 78%). There is approximately 20% inhibition with
both 8155 and $16 and none with $158 by PGE2. These results
indicate that 1)arachidonic acid and PGEziare targeting at
least 2 sites to suppress 514: a region between -220 and -
2897, the region between -220 and -80, and 2) PGE2 is
somewhat less inhibitory than arachidonic acid. While there
is overlap of the targets between these two compounds, there
are also some differences in specificity and level of
inhibition. PGE2 is metabolized very quickly by hepatocytes
and may explain why inhibition was not always as strong as

AA.

153

 

 

 

 

[I PGE:

  

 

 

IE

 

 

 

 

 

[20:4,n-6 3000
100
400
1000
S124J-—i- . 13.1
50 100

CAT Activity
% Inhibition

 

 

 

 

Figure 3.8. Deletion Analysis of the $14 Promoter with AA
and PGE2. The plasmids were transfected into hepatocytes

with TRBl and then treated with 250pM arachidonic acid or

10pM PGEQ and the corresponding controls (albumin + OA and
DMSO respectively) for 48 hours. The data from at least 3
separate experiments was pooled. The results are presented
as percent of control with standard errors. RCU = relative
CAT Units.

The I box and the C-region are necessary for 814
suppression by AA and PGE2. Results from the above
experiments indicated that the region between -220 and -80
is responsible for at least 2/3 of the inhibition of $14
gene transcription by PGE2 and AA. Accordingly, several
other plasmids were used to more specifically determine
which regions were required for the inhibition by AA and

PGE2. This region contains a Y-box, which binds NF-Y and is

154

 

 

required for the T3 induction of $14 gene transcription
(Jump et al., 1997b). To determine whether the Y-box or
other elements within the -220/-80bp PUFA-RR, a series of
plasmids were constructed (Figure 3.9). Each plasmid
contains the RSV-TATA box (-60/+20 bp) and the $14 TRR (-
2.8/-2.5 bp). Various components of the S14 proximal
promoter and other promoters are inserted between the TRR
and TATA-box.

Only $14 promoter constructs (R131 and R119) containing
the 514 Y box and the C region (upstream of the Y box)
showed inhibition by AA and PGE2. These results indicate
that regions surrounding the Y-box (-99 to -104) are
important for regulation by both of these compounds. R157
and R117, containing either the albumin Y box or 814 Y box
alone respectively, was only slightly sensitive to AA. R132
contains a Y box from the TK gene which is flanked by an Spl
site on either side of the Y box. This plasmid (RSV132) was
not sensitive to arachidonic acid. By comparing RSV119 and
RSV159, which contain the Y box and either the upstream or
downstream flanking region respectively, the AA suppression
is similar to R131 with only R119, which contains the Y box
and the upstream C region. Plasmids containing a Y box from
another gene (R157) or the Y-box competitor C/EBP (R158)
were not suppressed by AA. These results clearly point out

that the 514 Y box and the C region alone are not effective

155

but that both are required for the full effect.
Furthermore, suppression is specific to the 814 Y box (due

to its flanking regions) and not to other Y boxes.

 

-—I-Rla "5060‘
--I—-R158 100
-—-RI16 1000
-—.—-Rl17 2000
-—_RI$ 1000
_—- ”19 2°00
- smI ICAT IR'32 5000
—R'31 25‘”

I
I l I I

 

 

 

 

 

Um ICMT 0 20 - 4) m
Isuzu 'c/mm CATActiVity

%InhibitionbyAA

 

 

 

 

Figure 3.9. Y Box and C-Region are Required for PUFA
Control of $14. The above $14 promoter constructs were

treated with either 250pM OA or AA. CAT activity was
determined and normalized to OA. This graph is combined
data from three experiments and includes the percent error.
RCU = relative cat units

PUEAs and PGE2 regulation of 814 gene transcription is
dictated.by promoter context. Based on above results with
AA, I wanted to determine if PGEzinhibition of $14 gene
transcription also required the Y box and its flanking
regions. The n-3 fatty acid, EPA was tested as well to

determine if it too required this region for inhibition of

156

$14 CAT activity. Hepatocytes were transfected with the RSV
plasmids R131 and R132 and treated with 10pM PGE2«or 250pM

AA or EPA. Figure 3.10 illustrates the effect of these
compounds on CAT activity transfected with R131 and R132.
R131 contains the $14 promoter region between -80 and -220
(includes the S14 Y box and flanking regions) while R132
contains the TK Y box flanked by Spl sites. Only R131,
which contains the 814 Y box and the C-region, is suppressed
by AA, EPA and PGE2. This result again stresses the
importance of the Y box and its requirement for the flanking

regions for suppression of this gene.

157

 

-220 -80

:3 R131 TRR - CAT
- R132 IT'RRI-r sus CAT |

120

 

 

 

 

100 v

80 ..

 

D RSV131
I RSV132

00 w

 

 

 

40%

waMdCuMd

20 4.

 

 

 

 

 

 

 

 

 

 

EPA PGE2

 

 

 

Figure 3.10. Promoter Context Dictates Control by PUFA and
PGE2. Hepatocytes transfected with RSV131 and RSV132 were

treated with 250pM EPA or AA or 10pM PGE2 for 48 hours and
CAT activity was determined. Results were normalized to the
control (albumin or DMSO) and percent of control was plotted
on the graph. This graph depicts at least three combined
experiments for each treatment and the standard error.

DISCUSSION

The analysis of prostanoid regulation of hepatic
lipogenic gene expression was prompted by studies showing
that 20:4,n-6 suppression of lipogenic gene expression in
3T3-L1 adipocytes required cyclooxygenase (Chapter 2).
These studies suggested that 20:4, n—6 was converted to

prostanoids in adipocytes. Further results indicated that

158

PGE2 and PGFzcl suppressed mRNA encoding FAS and 814 through

a pertussis-toxin sensitive Gi/Go regulatory pathway in
adipocytes. In the this chapter, results show that the
20:4, n-6 mediated inhibition of hepatic lipogenic gene
expression does not require cyclooxygenase activity (Figure
3.1). Moreover, Northern analysis failed to detect mRNAs
encoding PGHSl or 2 in hepatic parenchymal cells (not
shown). Thus, n—6 PUFA-mediated inhibition of lipogenic

gene expression in hepatic parenchymal cells does not

 

require a prostaglandin intermediate. This finding does not
exclude the possibility that some other 20:4,n-6 metabolite
might be generated in parenchymal cells and activate a
signaling pathway. For example, cytochrome P450 mediated
fatty acid metabolism can potentially generate PUFA
metabolites that activate signaling cascades that affect
gene transcription (Capdevila et a1, 1990). Studies are
currently in progress to evaluate the role these pathways
play in dietary PUFA regulation of hepatic lipogenic gene
expression.

While dietary PUFA does not require cyclooxygenase for
its regulatory effects on hepatic gene expression, a number

of reports have indicated that 20:4,n-6 cyclooxygenase

 

products from non-parenchymal cells can act in a paracrine
fashion on parenchymal cells. For example, prostaglandins

produced by Kupffer cells in response to injury, sepsis or

159

other stimuli act in a paracrine fashion on the surrounding
hepatocytes to alter specific functions (Billiar and Curran,
1992). PGE2 increase hepatocyte proliferation through the
EP3 receptor (Hashimoto et al., 1997) and induce glycogen
breakdown (Garrity.et al., 1987; 1989; Hespeling et al.,
1995a; 1995b) . Glucagon induces PGE2, P602 and PGde
synthesis and release from Kupffer cells. Interestingly,
these prostaglandins counter glucagon-mediated
glycogenolysis. Thus the effect of PGEQion hepatic
metabolism in vivo is likely to be influenced by the
physiological status of the animal.

Our studies show that PGEziacts on primary cultures of
hepatic parenchymal cells to regulate mRNAs encoding genes
involved in lipogenesis. PGE2 suppresses mRNAs encoding
proteins involved in lipid synthesis (FAS, 814 and LPK) but
had no consistent effect on mRNAs encoding proteins involved
in non-mitochondrial fatty acid oxidation (AOX and CYP4A2).
Others have reported that PGE2«effects on glycogen
metabolism were linked to Gi-linked EP3 receptor that
decreased cAMP levels (Garrity et al., 1987; 1989; Hespeling
et al., 1995a; 1995b). I did not detect changes in hepatic
CAMP levels in response to PGE2 treatment (not shown).
Depending on the subtype, EP3 receptors can augment or

decrease intracellular CAMP as well as activate a PLC to

change intracellular IP3 and calcium. PGFn.is thought to

160

regulate intracellular IP3 and calcium levels. Based on
this reasoning, the PGE2 and PGFZG suppression of hepatic

lipogenic gene expression is consistent with an activation
PLC. This finding is consistent with our earlier results
with adipocytes (Chapter 2).

At the molecular level, both 20:4,n-6 and PGE2 suppress
Sl4 mRNA by inhibiting gene transcription (Figure 3.11).
Efforts to map the cis-regulatory targets for PGEziaction
showed that the 20:4,n-6 and PGE2 regulatory mechanism share
similar elements. In contrast to the inhibition of $14 gene
expression by peroxisome proliferator activated receptors
(Ren et al., 1996), neither 20:4, n-6 or PGE2 inhibited $14
gene expression through the thyroid hormone response region.
Thus, PGEzidoes not have generalized effects on thyroid
hormone regulation of this gene. Instead, the $14 proximal
promoter (-220 to -80 bp) was the principal target for both
PGEziand 20:4,n—6 suggesting that these two regulatory
pathways converge on common elements within the 814 promoter
to control its transcription. One key transcription factor
regulating 814 within the PUFA/PGE2 response region (-220 to
480 bp) is N-FY. NF-Y binds a Y-box at -104/-99 bp. It is a
heterotrimeric transcription factor that is critical for the
functioning of the 2 upstream enhancers (Jump et al.,
1997b). Any mutation or substitution of this element

essentially abrogates $14 gene transcription. Thus, factors

161

controlling NF-Y action impact on the transcriptional

capacity of the $14 gene.

 

20:4,n-6

/

      
   
   
   

Parenchymal

Cell
Kupffer
Cell ‘mRNA S14
\ * mRNA FAS
/ * mRNA L-PK

 

 

 

Figure 3.11. Hepatocyte and Kupffer Cell Interactions.
This picture demonstrates the actions of PGEziand AA in the
liver. Kupffer cells synthesize PGE2 from AA. Both AA and
PGE2 can act independently on hepatocytes to change gene
expression.

Here, I show that the promoter context dictates
sensitivity of NF-Y to PUFA/PGEQ control. NF-Y binds many
promoters (Wright et al., 1995). The thymidine kinase
promoter region at -115/-35 bp (Wagner et al., 1981), was
used here as it binds NF-Y in the context of Spl, to
evaluate its role in PUFA/PGEziaction. A reporter gene
containing the TK NF-Y elements (R132) was insensitive to
20:4,n-6, 20:5,ne3 and PGE2 (Figure 3.10). However,

substituting the $14 PUFA-RR for the TK element conferred

162

both PUFA and PGE2 control to the gene. The difference
between these two elements is that in the TK promoter, NF-Y
is flanked by Spl binding sites. No SP1 binding sites are
found within the PUFA-RR. Thus, factors other than NF-Y
that are present in the $14 PUFA-RR are critical for
PUFA/PGE2 control of $14 gene transcription. Studies are in

progress to identify these other factors.

CKNRNUUSICWEB

In summary, these studies have shown that 20:4,n-6 acts
directly on hepatocytes to suppress lipogenic gene
transcription. There is no requirement for cyclooxygenase
activity for this control mechanism (Figure 3.1). However,
specific prostanoids can act on the liver through EP3 and FP
receptors to activate a G-protein linked signaling cascade
that probably involves PLC, IP3 and calcium. In vivo, these
prostanoids arise from non-parenchymal cells and act in a
paracrine fashion on parenchymal cells to affect
carbohydrate and lipid metabolism. Taken together with our
'previous studies on PPARs (Ren et al., 1997), fatty acids
can regulate lipogenic gene transcription through 3 distinct
pathways: one is PPAR-dependent, another is prostanoid-
dependent and a third pathway is PPAR and prostanoid-

independent. Given the paucity of PPAR in the human liver

163

(Gonzalez et al., 1997), this latter pathway is probably the
operative pathway involved in the dietary PUFA suppression
of hepatic lipogenic gene expression in both rodents and

human under normal physiological conditions.

 

 

164

CHAPTER 4

CfiHUNEER 4

FUTURE EXPERIMENTS

PGE2 was found to regulate lipogenic gene expression in I

both adipocytes (Chapter 2) and hepatocytes (Chapter 3). AA

 

is converted to PG in adipocytes but not in hepatic
parenchymal cells (hepatocytes) . Thus, the PGE2 must be
derived from nonparenchymal cells to affect parenchymal cell
gene expression. This finding led to interest in other non-

parenchymal cell factors and how they might affect hepatic
gene expression. Accordingly, I tested the TNFa and IL-la

in primary hepatocytes for suppression of $14 gene
transcription.
This chapter will describe some preliminary experiments

to answer the following questions:

21. Do the Kupffer cell products IL—la and TNFa affect 814

expression in hepatocytes?

2. If so, where is the target in the 814 promoter for the

effective compounds?

165

INTRODUCTION

Because prostaglandins were inhibitory to CAT activity
in hepatocytes, this raised questions about the role other,
Kupffer cell products played in $14 gene expression. During

sepsis or injury, Kupffer cells release a variety of
interleukins, TNFa, prostaglandins, growth factors and
nitric oxide, all of which can affect both glucose and lipid
metabolism. The animal becomes hypertriglyceridemic and
hyperglycemic then quickly hypoglycemic. Hepatic fatty acid
synthesis and oxidation increase and lipolysis increases in

adipocytes as a result of TNFa (Billiar and Curran, 1992).

I was interested in determining how these products might
change 514 expression in hepatocytes.

Sl4 promoter analysis indicated that the Y box and its
flanking regions were necessary for inhibition by both AA
and PGE2 (Chapter 3). The second question will address the
issue of what other compounds may target this same region in
the $14 promoter. A similar target may be indicative of a
similar co-factor involved in the suppression of $14 gene

transcription.

166

NDUNERIAJHBJENDllflTTHOTHB

HEPATOCYTE PREPARATION AND CULTURE Hepatocytes were
prepared as described in Chapter 3 and Appendix B.
RNA ISOLATION’AND NORTHERN BLOTTING Total RNA was

isolated, blotted and probed as described in Chapter 2

(Materials and Methods).
CAT ASSATS Following treatments, cells were assayed

for CAT activity and protein content as previously described

(Jump et al., 1993). CAT Units: 14C-acetylated

chloramphenicol CPM/lOO pg protein/hour.

lflBSUEflEB

TNFa, but not IL-ld, suppressed $14 gene transcription.
Due to the inhibition of $14 by the Kupffer cell product,
PGE2, other Kupffer cell products were tested. TNFa,knu;
not IL-ld, suppressed CAT activity in primary hepatocytes
transfected with Sl4CAT124 (Figure 4.1). Because only TNFa
showed an effect, a dose response curve was determined in
hepatocytes treated with 0.5-10ng/ml of TNFa and CAT
activity was measured. This result (Figure 4.2) determined

that the EDMI= 1ng/ml TNFa in primary hepatocytes as

167

measured by CAT activity. However, RNA analysis indicated
that lOng/ml TNFa was required for 50% suppression of S14
mRNA (Figure 4.3). IL-la had no effect on $14 mRNA levels
(not shown). Other preliminary experiments indicate the
mRNAs encoding PK and ApoC3 are also suppressed by TNFa at

10ng/ml (data not shown). Further effects of TNFa and its

in vivo activator, LPS, on gene expression are currently

underway in our laboratory.

 

 

 

 

 

   

 

 

Kupffer Cell Products in Hepatocytes
12000
0000 ..
:8
C
D
l-
5 4000.
o .
Vehicle lL-1a TNFa

 

 

 

Figure 4.1. IL-la and TNFa Effects on $14 CAT Activity.
Primary hepatocytes transfected with Sl4CAT124 were treated

with vehicle, 1ng/ml Il-la or TNFa for 48 hours and CAT
activity was determined. This is a representative graph of
several experiments.

 

168

 

TNFa Dose Response

 

 

 

 

 

 

0.00 t 1 +
0 0.01 0.05 0.1 1

ng/ml TNFa

 

 

 

Figure 4.2. Dose Response of TNFa. Primary hepatocytes
transfected with Sl4CAT124 were treated with TNFa at dose
from 0 to lng/ml for 48 hours and CAT activity was
determined. The data was normalized to the 0 dose value and
percent error was determined. Each point represents the

mean of 3 plates.

169

 

TNFa Effects on 814 mRNA

120%

 

100% .-

Pmmmuoflkmbb
§

 

 

   

 

 

Vehicle TNFa

 

 

Figure 4.3. TNFa Effects on $14 mRNA. Primary hepatocytes
were treated for 48 hours with vehicle or long/ml TNFa and

then total RNA was harvested and probed for $14 mRNA. This

graph represents data pooled from 2 separate experiments

each done with duplicate samples and percent error is shown.
TNFa targets the Y box. Because the target in the $14

promoter for both arachidonic acid and PGE2 both involve the

proximal promoter, I was curious about whether TNFa also

targeted this region. TNFa was used to treat hepatocytes
transfected with several different Sl4 promoter deletion
_constructs. These results are shown in Figure 4.4. TNFa
showed good inhibition (70%) with $170 and 60% inhibition of
both 5155 and $156 and some effect on $158 (18%). There is
also some inhibition of TK222, implying that the TRR and/or
the TK element may be targeted by this compound. These

results are complex but seem to indicate that the same

170

region targeted by AA and PGE2, also is targeted by TNFa in
the suppression of 514. However, further research will be
necessary to determine if TNFa also targets other regions

such as the TRR or the TK element.

 

 

Sl4 Promoter Analysis by TNFa

 

 

 

 

 

_m moo

 

IrRR ITK ICATI TK224

 

 

II'RR I—eoICATI S158
[TRR ['120 [CAT] S156
ERRI'ZZO [CAT] S155

[TRR I-4300 ICATI S124

Easmid Identification] 096 56% 100%

Percent Inhibition of CAT Activity

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.4. $14 Promoter Analysis on CAT Activity by TNFa.
Hepatocytes were transfected with the above plasmids and

treated for 48 hours with 1ng/ml TNFa. CAT activity was
determined and normalized to the control treatment. This

graph represents 3 separate experiments of combined data and
percent error.

The RSV constructs used in Chapter 3 determined that
PGEziand AA both required the Y box and the C-region and was

promoter context specific. The same plasmids were used here
to determine if TNFa also required the Y box and flanking

regions. The results are shown in Figure 4.5. Unlike AA

171

and PGE2, TNFa required only the Y box and not the flanking

regions to give the full suppressive effect. R131, which
contains the Y box and both the upstream and downstream
regions, was suppressed 50% with TNF treatment as was R117,
containing only the Y box. However, R116 also showed
suppression by TNF (40%). Perhaps TNFa is acting through
the Y box without requiring the flanking regions to aid in
the inhibition. Both this result and Chapter 3 indicate
that the Y box appears to play an extremely important role
in the suppression of $14 transcription by a variety of

factors.

172

 

he

 

 

 

 

In.“
R131

 

J-

1 l

C] TATA I own 0 50 100
NF] B NFY Albumin

% of Control
I s14 m I cIEBP Albumin

 

 

 

Figure 4.5. TNFa Effect on 514 Proximal Promoter Elements.
Primary hepatocytes transfected with the above plasmids were

treated for 48 hours with lug/ml TNFa. CAT activity was
determined and normalized to the vehicle and presented as %
of control. The above graph is pooled data from at least 3
separate experiments with percent error given. CAT activity
was very low in cells transfected with R116 and R158.

DICUSSION

Question 1 asked what other Kupffer cell products
Change 814 expression. Of the cytokines tested, TNFa was

the most inhibitory. This cytokine is also released in the

liver as a result of sepsis or LPS stimulation. The actions
of TNFa released from Kupffer cells and acting on the

hepatocyte appear somewhat contradictory to the in vivo and

173

 

adipocyte data in the literature. In rats treated with
TNFa, liver lipogenesis and serum triglycerides increase
and LPL activity decreases (Grunfeld et al., 1990; Feingold
et al., 1990). However, in hepatocyte cell culture, TNFa

did not increase lipogenesis in experiments reported by
Brass and Vetter (1994). I did not measure lipogenesis

directly, rather, results shown in Figures 4.2 and 4.3
indicate that TNFa is inhibiting the lipogenic gene model
(814) mRNA expression in primary hepatocytes. My treatments
are also chronic treatments (48 hours) compared to the above
in vivo results, which are acute treatments. Thus, the
reports that TNFa does not increase lipogenesis in
hepatocytes and my results that $14 transcription is not

increased by TNFa are similar.

The search for what mediates TNFa induction of
lipogenesis has resulted in several reported possibilities:
IL—1, IL-6, interferons and prostaglandins. Grunfeld et al.
(1990) ruled out prostaglandins as a TNFa mediator but
suggested that IL—6 could be involved. Brass and Vetter
4(1994) reported that IL-6 and a prostaglandin E agonist both
increased hepatocyte lipogenesis. There also appears to be
some interaction between the compounds released during

sepsis.

174

Another mediator of TNFa is IL—1. IL-1 is a

proinflammatory cytokine which binds a separate and distinct
cell surface receptor from TNF. Both can cause similar
effects, however. Several inducers of IL-1 such as LPS,
phorbol esters, radiation and viruses also induce TNF.
Kupffer cells also produce both of these cytokines, although
monocytes produce much more IL-1. IL-1 binds a receptor
which causes activation of a MAP kinase also activated by
LPS. Blocking this MAP kinase phosphorylation can stop
production of TNF and IL-1. There is some evidence that
this MAP kinase phosphorylation involves a Janus kinase
(JNK) or Stat pathway. IL-l can also increase prostaglandin
synthesis (Tocci and Schmidt, 1997). However, my results

with IL-1 on 814 transcription do not indicate that this

compound is mimicking TNFa or that IL-la is a mediator of
TNFa or PGE2.

TNFa has a cell surface receptor belonging to the
cytokine family. Although no attempts were made to reverse
the TNFa suppression of S14 CAT activity, it is likely that

this compound acts through its receptor to activate a
JAK/STAT phosphorylation cascade, perhaps activating some

mediator require for gene suppression. One such possible

mediator is NF-KB, another is sphingomyelin (see literature

review). It is not known if NF-KB binds the 814 promoter or

175

if ceramide is suppressive to 814. Further research will be

necessary to determine how TNF suppresses 814.
These results show that AA, PGE2, and TNFa have at

least two separate targets each in the $14 promoter, one of
which the Y box. I have focused on the Y-box region and
determined that this suppression requires the surrounding
regions and is promoter context specific. The Y-box alone
does not result in more than 25% suppression of CAT activity
by any compound tested but adding the flanking regions

results in at least 2/3 of the total suppression seen with
the full length promoter when using AA, PGEzior TNFa. The

other portion in the $14 promoter necessary for the
suppression by these agents remains to be determined.
During sepsis or liver injury, Kupffer cells release

many different compounds. I did not test these compounds in
any combinations except PGEzand TNFa. Separately they

inhibited CAT activity about 50% and together inhibition was
75% (data not shown). This result indicates that they are
additive and probably operating through distinct regulatory
mechanisms. Other combinations of cytokines and
prostaglandins have not been attempted with this system.
There is an abundance of literature indicating several
complex relationships between Kupffer cell products, and
much more research will have to done before scientists can

hope to understand these interactions.

176

CONCLUSIONS AND FUTURE DIRECTIONS

ADIPOCYTES: Although results show that in adipocytes,
AA acts through prostaglandins which involve a G-protein
signaling mechanism to suppress 814, what that G-protein
system is regulating is still unclear. While calcium seems
to be involved in the signaling, that has not been
definitively proven here. Perhaps the inhibitors tried were
not effective or perhaps there are other agents that would
be better utilized for blocking intracellular calcium
release. Intracellular calcium changes were not measured
and experiments set up to do so may be necessary to verify
calcium involvement.

I also did not test reversal of EPA suppression in
adipocytes by the cyclooxygenase inhibitor. EPA can be
metabolized to prostanoids, although this pathway is more
prevalent with AA. It would be interesting to know if
flurbiprofen could reverse 814 suppression in EPA treated
adipocytes.

HEPATOCYTES: In the hepatocyte experiments, no
inhibitor reversed the PGEzior AA inhibition of 814 gene
transcription. Again, the inhibitors may not have been
effective. Pertussis toxin has reversed PGEzieffects in
hepatocytes in other work (see literature review) but could

not in my system. Perhaps the dose or time of treatment I

177

 

used was not correct, though several doses and times were
attempted. However, if PGE2 is acting through a different
pathway than those reported in the literature for $14
suppression, pertussis toxin may not be an effective agent
and other signaling pathway inhibitors should be used.
While PG were inhibitory to $14 in hepatocytes, further
research will have to be done to determine how they are
acting to cause suppress of transcription.

Arachidonic acid also acts to suppress Sl4
transcription and does so independently of PG and PPARs.
Yet the many inhibitors of arachidonic acid metabolism,
oxidation and blockers of signaling pathways could never
reverse this inhibition in hepatocytes. This seems to imply
that the fatty acid is directly changing transcription, but
how? Is AA changing binding of a co-factor, and if so, what
is this co-factor? Is AA inhibiting the binding of a known
transcription factor that normally binds and if so, which
one? Perhaps it is an unknown transcription factor that has
reduced binding when PUFA are present. Considering how
complex transcriptional control has become as more research
is done, it is likely that there is more than one factor
involved. Again, further research will have to be completed
to answer these questions about PUFA control of gene

transcription.

178

r ....
‘I

 

 

Although preliminary, it is very interesting that the
cytokine TNFa, which is released from Kupffer cells in the
liver, also suppresses $14 gene transcription. LPS or
sepsis is known to increase lipogenesis, yet TNFa depressed

Sl4 gene transcription in hepatocytes. While this may be
due to differences in treatment times and cell types, it is
a very curious observation. For example, would diet change
this effect or influence it? Is the sepsis induced
hypertriglyceridemia in animals or humans harmful and could
it be changed by drugs or diet in patients?

The promoter targeting results demonstrate importance
of the Y box in controlling the transcription of 514. While
the target in the $14 promoter has been narrowed to the Y

box and its flanking region, the co-factor involved in the
inhibition is still unknown. Do AA, PGE2 and TNFa directly

interact with the DNA binding at Y box or is there a common
factor they interact with? Determination of this common
factor, if indeed one exists, may prove to answer the above
questions about how AA and PGEziare suppressing this gene.
Current efforts are underway in our laboratory to address
this question.

In conclusion, this dissertation has shown that in
adipocytes, AA suppression is through a prostanoid pathway

linked to a G-protein signaling system. I have also

demonstrated that AA, PGE2 and TNFa inhibit 814 gene

179

 

 

transcription via the Y-box and its flanking regions in
hepatocytes, and that AA acts in a prostanoid and PPAR

independent pathway.

180

 

 

APPENDIX A

AIWHQMDIXIIL

Solutions for growth of 3T3-L1 cells

1.

10.

11.

DMEM: Gibco (Gaithersburg, MD), 1 package DMEM or
13.37g, 3.7g NaHC03, 4.89 HEPES (20mM final
concentration), dissolve, adjust to pH 7.0, volume to 1
liter, filter sterilize in hood, store at 4 C.

Calf Serum: Gibco, Heat inactivated for 45 minutes at
56 C.

Fetal Bovine Serum: Atlanta Biologicals (Atlanta, GA)

Penicillin/Streptomycin Solution: Sigma Chemicals (St.
Louis, MO), lOOKU/ml of penicillin and 10mg/ml
streptomycin.

Insulin: Gibco, lmM insulin in 0.2M HCl—0.1M NaCl,
filter sterilized, stored at 4 C.

IBMX: Sigma Chemicals, 100mM IBMX in 0.1N NaOH, filter
sterilized, stored at 4 C.

Dexamethasone: Sigma, 10mM dexamethasone in DMSO,
filter sterilized, stored at 4 C.

Albumin: Boehringer Mannheim, 6.8 g of fatty acid free
bovine serum albumin dissolved in 50 ml of DMEM, filter
sterilize (.45micron filter), store at 4 C.

Phosphate Buffer Saline (PBS): see Molecular Cloning
(Maniatis vol. 3.)

10X Trypsin EDTA: Sigma, diluted to 1X in PBS.

Geneticin: Gibco, 100mg/ml geneticin in water, filter
sterilized, stored at -20 C.

Solutions for transfection of 3T3-L1 cells

1.

0.25 M CaClz : 3.675 g CaClzin 100 ml glass distilled
water

181

 

 

2X-BSS: 50mM BES, pH 6.95, 280 mM NaCl, 1.5mM Nafifixh
For 100ml solution: add BES 1.0669 and 1.64 9 NaCl to
50 ml glass distilled water and adjust pH to 6.95 with
HCl then fill to volume of 90ml. Make NagHPO4
(0.2129/100ml water) and add 10ml to BES and NaCl
solution to get 1.5mM N83HPO4.

Solutions for RNA and Northern blots

 

1.

1% Large gel (250ml): Dissolve 2.5 9 agarose, 25 ml 10X
MOPS and 245 ml water by boiling then cool to 60 C. In
the hood, add 7.5ml formaldehyde while stirring and
immediately pour then allow to

gel.

Running buffer: 1X MOPS

10X MOPS: 0.2M MOPS, 50mM sodium acetate, 10mM EDTA, pH
7.0, filter sterilize and/or autoclave.

RNA Sample Preparation: Dissolve 10-20 RNA in 5pl TE-8

and add 25pl electrophoresis sample buffer (made
fresh). Mix and incubate at 65 C for 15 minutes then

add 1pl 1mg/m1 EtBr (this is a 1:10 dilution of the
usual EtBr). Mix and quick spin and load sample.

Electrophoresis Sample Buffer: 0.75ml deionized
formamide, 0.15ml 10X MOPS, 0.24ml formaldehyde, 0.1ml
water, 0.1ml glycerol, 0.08ml 10% (w/v) bromophenol
blue.

10X SSC: see Molecular Cloning (Maniatis) vol 3.
10% SDS: see Molecular Cloning (Maniatis) vol 3.

Prehyb/hybridization solution: 50% deionized formamide,
0.5X Denhardts, 5X SSC, 0.1% SDS, 50mM sodium phosphate
buffer pH 6.5, 0.1mg/ml salmon sperm DNA (boiled before
addition), 0.25mg/ml tRNA (boiled before addition) and
water to volume.

cDNA Probe was made according to Gibco's protocol,
cleaned with spin columns from 3 Prime 5 Prime
(Boulder, CO) and boiled for 10 minutes. The probes
were added to hybridization solution at 2 million
cpm/ml and allowed to hybridize overnight.

182

Solutions for Gel Shifts

1.

TEN (Tris-EDTA-NaCl): see Molecular Cloning (Maniatis
vol. 3).

Polynucleotide kinase buffer:500mM Tris-Cl, pH 7.5;
100mM MgClz; lmM spermidine and lmM EDTA. Store at -20
C.

100mM DTT:Store at -20 C.
100mM MgClz: Store at -20 C.

Poly dI:dC: resuspended in TE-8 and add NaCl to 0.1M at
a final concentration of 2mg/ml. Heat to 90 C for 10

minutes and cool slowly to room temperature. Store at
-20 C.

Nuclear protein dialysis buffer: 25mM Tris-Cl, pH 7.5;
40mM KCl, 0.1mM EDTA and 10% glycerol.

8% acrylamide gel:16ml 30% acrylamide-bisacrylamide;
1.5ml 10X TBE, pH 8.3; 0.3ml 20% NP-40 and 41.9ml GD-

water. Mix and degas. Add 300pl 20% ammonium

persulfate (made fresh) and 60pl TEMED. Mix gently and
pour into casting stand. Let the gels polymerize for 2
hours and store at 4 C wrapped in plastic wrap.

TBE (Tris-Borate-EDTA): see Molecular Cloning (Maniatis
vol. 3).

Solutions of nuclei isolation & nuclear protein preparation

1.

2.

Solution A: 20mM Tris-Cl, pH 7.5; 2mM MgClz.
Buffer A: 10mM Tris-Cl, pH 7.5; 5mM MgClz; 320mM
sucrose;

0.2%nonidet P-40; lmM DTT*; lmM PMSF*.

Buffer B: 10mM Tris-Cl, pH 7.5; 5mM MgClz; 880mM
sucrose; 0.2%nonidet P-40; lmM DTT*; lmM PMSF*.

Buffer C: 10mM Tris-Cl, pH 7.5; 5mM MgClz; 250mM
sucrose; 0.2%nonidet P-40; lmM DTT*; lmM PMSF*.

Buffer D: 50mM HEPES, pH 7.5; 5mM MgClz; 60pM EDTA; 40%
glycerol; lmM DTT*; lmM PMSF*.

183

Nuclear lysis buffer: 10mM HEPES, pH 7.6; 100mM KCl;
3mM MgClz; 0.1mM EDTA; 10% glycerol; lmM DTT*; lmM
PMSF*.

4 M (NH4);SO4 pH 7.9

Nuclear protein dialysis buffer: 25mM Tris—Cl, pH 7.5;
40mM KCl, 0.1mM EDTA and 10% glycerol; lmM DTT*; lmM
PMSF*.

* Autoclave all solutions for 20 minutes then add DTT
and PMSF at time of use.

184

APPENDIX B

AEHHNMDIEIIB

Hepatocyte Preparation Protocol

Required items

 

koooqmmwai-I

10.
11.
12.
13.
14.

15.
16.
17.

18.
19.

operating tools

gauze pads

500ml beaker w/ stir bar (autoclaved)
500ml beaker w/ funnel (autoclaved)
silastic tubing

hemacytometer

sutures (in ethanol)

elastic bands

bubblers (autoclaved)

Hepatocyte filter UV/ethanol sterilized
catheter

syringes/needles

heparin

Methoxyflurane

water bath at 42°C

peristaltic-pump

Oz/COz gas

Betadinem

Bottle top filters

Prepared Items

 

IT IS IMPERATIVE THAT THE BUFFERS ARE PERFECT AND THE pH IS
CORRECT FOR GOOD HEPATOCYTE PREPS. ABSOLUTELY NO SOAP IS
TOLERATED, SO‘NASH THE GLASSNARE CAREFULLY.

Perfusion Buffer I (PB-I)

For preparing 1 Liter of 10X Buffer

 

1.42 M NaCl 82.9 g
.067 M KCl 5.0 g
0J.M HEPES 23.8 g

Dissolve in ddH20, pH to 7.4
Filter

Preparing 400 ml of 1X Buffer

Filter the following into sterile bottle
360 ml ddH20

40 ml 10X PB-I

185

Perfusion Buffer II (PB—II)

Preparing 2 Liters of PB-II

 

66.7 mM NaCl 7.8 9 NaCl or 33.4 ml 4 M NaCl
6.7mMKCl 1.0 gKCl or 4.6 ml 3MKC1
100 mM HEPES 47.66 g HEPES

4.8 mM CaC1202H20 1.41 g CaC1202H20

Dissolve in ddeO, pH to 7.6
Filter sterilize in 400 ml aliquots into sterile
bottles

Liberase-Albumin Solution

Stock Liberase comes in 70mg aliquots. Add 10ml PB-II
to bottle, shake gently to dissolve, let sit on ice for 10
minutes then aliquot into 10 glass vials (1 ml each). Store

at ~20W3. On the day of hepatocyte preparation, remove 1
vial and let it thaw on ice for at least 10 minutes.
Previously albumin was used in PB-II but this is optional.
Liberase works without the albumin. Once the PB-II has
gassed 30 minutes, pour PB—II into this beaker. Before
filtering, add the lml of Liberase. If albumin is used, the
Liberase should not be added until most of the albumin is
dissolved to prevent loss of activity of the Liberase. The
Liberase should always be on ice or left frozen until it is
used.

Percoll

Preparing the Percoll Stock
90 ml Percoll

10 ml sterile 10X PBS

1 ml 1 M HEPES, pH 7.4

 

Store at 4°C.
Williams E Media

Preparing 1 Liter

1 package Williams E media powder

2.2 g sodium bicarbonate

5.5 g HEPES

12 ml penicillin/streptomycin antibiotic

1. 5ml 5 N NaOH

one of the following:
1. 43 m1 lactate (for 10mM lactate)
4. 5 9 glucose (for 25mM glucose

 

186

Plating Media - 10mM lactate in Williams E Media + 10% FBS

Transfection Media - 10mM lactate in Williams E Media +
200nM insulin + 10nM dexamethasone

Treatment media - 10mM lactate or 25mM glucose in Williams E
+ 1pM insulin + 10nM dexamethasone

Protocol

 

The day before

1. Starve a cage of rats. Use the “do not feed” tape on
the cage so ULAR doesn’t feed them. Starve at least
one more rat than what is planned on being used.

2. Turn on the blower on the hood. Place a filter (1 per
rat) in the glass dish with 70% ethanol covering it.
UV light should be ON.

3. Make sure all the proper equipment is autoclaved.

4. Prepare at least 500ml of 10% PBS/Williams E 10mM
Lactate media per rat. Williams E 10mM media will also
be needed for transfection as well as treatment media.

5. Make Perfusion buffers I and II are ready.

The day of preparation

1. Turn UV light to regular light in the hood. Turn on
and light gas.

2. Spray off PB-I and PB-II bottles w/ ethanol and place a
sterile bubbler in each bottle. Place weight around
bottle neck. Wrap top w/ sterile tin foil. Place in

the waterbath set at 42°C.
3. Gas each w/ 95%02/5%C02 for 30 minutes. The tubing
with clamp should be used for PB-II.
Get 2 buckets of ice. Get the rat cage from ULAR.
Weigh out Albumin (lg/100ml PB-II) into glass beaker
with the stir bar. Place it back in the hood on stir
plate. (THIS IS NOT REQUIRED WITH THE LIBERASE
PROTOCOL)
~6. Prepare 0.5ml heparin in syringe. Prepare orange cap
tube w/ methoxyflurane in dryer hose on surgery table.
Prepare surgery table and utensils.
Rinse tubing on pump with 70% ethanol and drain.
By now the gassing should be complete, clamp off PB-II
tubing, remove bottle and pour PB-II into beaker (w/
the albumin if using this) in the hood. Allow the
albumin to dissolve, then add the Liberase aliquot
(lml). Once this is complete, sterile filter back into
the same bottle. Place the weight back over the top,
rewrap w/ tinfoil and put back into the water bath.

U105

(Dd

187

10.

11.

12.
13.

14.

15.

16.

17.

18.

19.

Wash the pump tubing carefully with 70% ethanol. DO
NOT TOUCH WITH HANDS AFTER THE TUBING IS WASHED.
Carefully place the tubing into the buffers. The
longest tubing goes into PB-II bottle. Turn on pump
and pull up PB-II to the joint, then pump PB-I through
tubing. PB-I should be coming out the other end with
no bubbles.

Weigh the rat. Place him in the sock. Wrap him up
with the towel bending his tail up toward his head.
This will prevent him from going backwards.

Place his head into the tubing about to his eyes.
Gently but firmly hold his head in the tube until he is
unconscious. Carefully remove towel and sock. Fasten
legs w/ rubber bands to surgery table (another person
to assist with this step is handy). Keep his head
partially in the tube throughout procedure. Monitor
his breathing and adjust tube to keep him alive but
still unconscious.

Inject ~500pl heparin into the tail vein.

Shave belly from groin to sternum. (The better shaved
he is the easier it is to cut the skin.)

Wash thoroughly with Betadine solution. Rinse with 70%
ethanol. Be generous with each.

Make sure the rat is breathing and everything is still
ready before cutting.

Make a small incision in the middle lower portion of
his belly. Carefully detach skin from muscle layer
below. Cut toward his sides from the middle and angle
slightly superior as you cut transversely. The cut
should resemble a “V”. Once you have loosened the skin
sufficiently, use the large hemostats to pull back the
skin. Further loosen and cut skin until it is pulled
back from the muscle layer all the way around the rib
cage.

Rinse muscle layer with ethanol. Be careful! The
ethanol is cold to the rat and he takes a deep breath.
A deep breath means more anesthetic, which can kill
him. Adjust tube until his breathing returns to
normal.

Again, make a medial incision at the base of his belly.
Don't go too deep, preventing any slicing of internal
organs! Make a cut toward each side and up to his
front legs, angling upward from the middle point (cut
both sides). Use the small hemostats to hold back the
muscle layer. It should slide over the ribs and stay
there. Make sure the cut is low enough (toward the
table) so blood can run out later.

Push back the organs with a piece of gauze. The
hepatic vein should be apparent. Adjust organs and
gauze so the vein is easy to access.

188

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.
30.

‘31.

32.

33.

Using the small tweezers, place a piece of thread on
the rat so one end is near the vein. Carefully make a
hole with the tweezers in your left hand and pull the
thread under the vein. Make a LOOSE knot.

Make sure the tubing is ready and not twisted. Lay it
close to the vein, usually on the gauze covering the
organs.

Cannulate the vein. Tighten the knot. Turn pump on to
1 and gently attach tubing to catheter. Turn up pump
to 4, simultaneously clipping the main vein going down
his backbone.

The liver should perfuse immediately. The muscle or
skin layer may have to be cut further to allow good
drainage. Using the spatula, carefully make sure all
the lobes are perfusing. DON'T BUMP THE CATHETER!
After PB-I is mostly through the liver, turn pivot so
PB-II is perfused through. Perfusion speeds and
amounts depends on the rat and many other things.
Watch the liver, don't let it get too mushy.

4 100mM petri dishes will be needed. Place 3 in the
hood and one next to the rat’s head. Once perfusion is
finished, remove liver, place it in the petri dish and
carry it to the hood. Take the large forceps to the
hood as well.

Add ice cold plating media to each dish and rinse the
liver three times in successive dishes.

Use a comb to break up the liver in the 4m‘dish. Hold
the liver with the forceps and comb to until the liver
is demolished.

Place filter over beaker with the funnel. Carefully
place the combed liver solution in the filter and allow
it to filter through. There will be some debris that
will not go through. Rinse dish w/ a small amount of
media. Rinse filter.

Place contents of beaker into a 50ml orange cap tube.
Spin for 10 minutes at 50 g. Keep the tube on ice at
all times from this point on.

Make 4 tubes of Percoll Stock-Media solution. (10mls
of each in each tube = 20mls per tube). (Start
cleaning up the mess.)

Dump off supernatant from tube and resuspend pellet in
10-15ml media. Once the pellet is COMPLETELY
resuspended, fill tube to 40mls with media. Place
20mls in each of 2 Percoll-Media tubes on top of the
Percoll-media mix. Do NOT mix the layers. Spin 10
minutes at 270 g.

Remove supernatant and interface, saving the pellet.
Repeat step 31.

Remove supernatant and resuspend pellet in 25ml media.
Centrifuge 5 minutes at 50 g.

189

34.
35.

36.

37.

Dump supernatant and resuspend pellet in 30mls.

Count the cells (# of plates = average # of cells per
grid if 30mls was used to resuspend the pellet).
Plate cells (3 million/60mm plate or 10million/100mm
plate)

Transfect about 4 hours later.

Transfection Protocol

1.

UTDUJN

Prepare transfections in blue capped tube = lml lactate

media + 2pg reporter + 1pg receptor + 6.6pl lipofectin
per plate.

Let preparation sit 20-30 minutes after mixing gently.
Rinse plates with 2ml PBS.

Add 2ml lactate media.

dd lml transfection mix.

190

REFERENCES

'IREETEUQMOES

Aarsland, A.M., M. Lundquist, B. Borreston and R.K. Berg.

1990. On the effect of peroxisomal B-oxidation and
carnitine palmitoyltransferase activity by eicosapentaenoic
acid in liver and hearts from rats. Lipids 25:546-548.

Adachi, T., S. Nakashima, S. Saji, T. Nakamura and Y.
Nozawa. 1997. Possible involvement of pertussis toxin-
sensitive G protein in hepatocyte growth factor-induced
signal transduction in cultured rat hepatocytes: pertussis
toxin treatment inhibits activation of phospholipid
signaling, calcium oscillation, and mitogen-activated
protein kinase. Hepatology 26:295-300.

Adachi, T., S. Nakashima, S. Saji, T. Nakamura and Y.
Nozawa. 1995. Roles of prostaglandin production and
mitogen-activated protein kinase activation in hepatocyte
growth factor-mediated rat hepatocyte proliferation.
Hepatology 21:1668-1674.

Adams, M., M.J. Reginato, D. Shao, M.A. Lazar and V.K.
Chatterjee. 1997. Transcriptional activation by peroxisome

proliferator-activated receptor 7 is inhibited by
phOSphorylation at a consensus mitogen-activated protein
kinase site. Journal of Biological Chemistry 272:5128-5132.

Adi, S., A.S. Pollock, J.K. Shinenaga, A.H. Moser, K.R.
Feingold and C. Grunfeld. 1992. .Role for monokines in the
metabolic effects of endotoxin. Journal of Clinical
Investigations 89:1603-1609.

Ailhaud, G., E. Amri, S. Bardon, S. Barcellini-Couget, B.
Bertrand, R. M. Catalioto, C. Dani, A. Doglio, C. Forest, D.
Gaillard, P. Grimaldi, A. Ibrahimi, R. Negrel, C. Vannier
and G. Vassaux. 1991. Growth and differentiation of
regional adipose tissue:molecular and hormonal mechanisms.
International Journal of Obesity 15:87-90.

Ailhaud, G., E.-Z. Amri and P. Grimaldi. 1995. Fatty acids
and adipose cell differentiation. Prostaglandins
Leukotrienes and Essential Fatty Acids 52:113-115.

Ailhaud, G., E.-Z. Amri and P. Grimaldi. 1996. Fatty acids

and expression of lipid-related genes in adipose cells.
Proceedings of the Nutrition Society 55:151-154.

191

Ailhaud, G., P. Grimaldi and R. Negrel. 1992. Cellular and
molecular aspects of adipose tissue development. Annual
Review of Nutrition 12:207-233.

Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts and J.D.
Watson. 1994. Molecular Biology of the Cell, 3“’ed.,
Garland Publishing, Inc., New York, New York.

Amri, E.-Z, G. Ailhaud and P. Grimaldi. 1991. Regulation
of adipose cell differentiation. II. Kinetics of induction
of the aP2 gene by fatty acids and modulation by
dexamethasone. Journal of Lipid Research 32:1457-1463.

Alberts, B., D. Bray, J. Lewis, M. Raff, K. Roberts and J.D.
Watson. 1994. Molecular Biology of the Cell, 31:" ed.,
Garland Publishing, Inc., New York.

Allmann, D.W. and D.M. Gibson. 1965. Fatty acid synthesis
during early linoleic acid deficiency in the mouse. Journal
of Lipid Research 6:51-62.

Amri, E.-Z., G. Ailhaud and P.A. Grimaldi. 1994. Fatty
acids as signal transducing molecules:involvement in the

differentiation of preadipose to adipose cells. Journal of
Lipid Research 35:930-937.

Amri, E.-Z., B. Bertrand, G. Ailhaud and P. Grimaldi. 1991.
Regulation of adipose cell differentiation. I. Fatty acids
are inducers of the aP2 gene expression. Journal of Lipid
Research 32:1449-1456.

Amy, C.M., B. Williams-Ahlf, J. Naggert and S. Smith. 1990.
Molecular cloning of the mammalian fatty acid synthase gene
and identification of the promoter region. Biochemical
Journal 271:675-679.

Amy, C.M., B. Williams-Ahlf, J. Naggert and S. Smith. 1992.
Intron-exon organization of the gene for the multifunctional
animal fatty acid synthase. Proceedings of the National
_Academy of Sciences 89:1105-1108.

Armstrong, M.K., W.L. Blake and S.D. Clarke. 1991.
Arachidonic acid suppression of fatty acid synthase gene
expression in cultured rat hepatocytes. Biochemical and
Biophysical Research Communications 177:1056-1061.

Asboth, G., S. Phaneuf, G.N. Europe-Finner, M. Toth and A.L.

Bernal. 1996. Prostaglandin Eziactivates phospholipase C
and elevates intracellular calcium in cultured myometrial

192

v4 .1.i-lac.ulrrf‘1

 

 

cells:involvement of EP1 and EP3 receptor subtypes.
Endocrinology 137:2572-2579.

Aubert, J., G. Ailhaud and R. Negrel. 1996. Evidence for a
novel regulatory pathway activated by (carba)prostacyclin in
preadipose and adipose cells. FEBS Letters 397:117-121.

Avasthy, N., J.Y. Jeremy and P. Dandona. 1988. The role of
calcium in mediating phorbol ester- and insulin-stimulated
adipocyte lipogenesis. Diabetes Research 9:91-95.

Beaumier, L., N. Faucher and P.H. Naccache. 1987.
Arachidonic acid-induced release of calcium in permeablized
human neutrophils. FEBS Letters 221:289-292.

Belury, M.A., K.E. Patrick, M. Locniskar and S.M. Fischer.
1989. Eicosapentaenoic and arachidonic acid:comparison of
metabolism and activity in murine epidermal cells. Lipids
24:423-429.

Bennet, M.K., J.M. Lopez, H.B. Sanchez and T.F. Osborne.
1995. Sterol Regulation of fatty acid synthase promoter.
Journal of Biological Chemistry 270:25578-25583.

Bernlohr, D.A., M.A. Bolanowski, T.J. Kelly and M.D. Lane.
1985. Evidence for an increase in transcription of specific
mRNAs during differentiation of 3T3-Ll preadipocytes.
Journal of Biological Chemistry 260:5563-5567.

Berven, L.A. and G.J. Barritt. 1994. -A role for a
pertussis toxin-sensitive trimeric G-protein in store-
operated Cah'inflow in hepatocytes. FEBS Letters 346:235-
240.

Berven, L.A., M.F. Crouch, F. Katsis, B.E. Kemp, L.M.
Harland and G.J. Barritt. 1995. Evidence that the
pertussis toxin-sensitive trimeric GTP-binding protein Ga;
is required for agonist-and store-activated Cay'inflow in
hepatocytes. Journal of Biological Chemistry 270:25893-
25897.

Berven, L.A., B.P. Hughes and G.J. Barritt. 1994. A slowly
ADP-ribosylated pertussis-toxin-sensitive GTP-binding
regulatory protein is required for vasopressin-stimulated
Ca*'inflow in hepatocytes. Biochemical Journal 299:399-
407.

Billiar, T.R. and R.D. Curran, eds. Hepatocyte and Kupffer
Cell Interactions. Boca Raton, Florida: CRC Press, 1992.

 

 

193

 

Billiar, T.R., T.W. Lysz, R.D. Curran, B.G. Bentz, G.W.
Machiedo and R.L. Simmons. 1990. Hepatocyte modulation of
Kupffer cell prostaglandin.Eb production in vitro. Journal
of Leukocyte Biology 47:304-311.

Bjornsson, O.G., J.D. Sparks, C.E. Sparks and G.F. Gibbons.
1992. Prostaglandins suppress VLDL secretion in primary rat
hepatocyte cultures:relationships to hepatic calcium
metabolism. Journal of Lipid Research 33:1017-1027.

Blake, W.L. and S.D. Clarke. 1990. Suppression of rat
hepatic fatty acid synthase and $14 gene transcription by
dietary polyunsaturated fat. Journal of Nutrition 120:1727-
1729.

Blobe, G.C., W.A. Khan and Y.A. Hannun. 1995. Protein
Kinase C: Cellular target of the second messenger

arachidonic acid? Prostaglandins Leukotrienes and Essential
Fatty Acids 52:129-135.

Bradford, M.M. 1976. A rapid and sensitive method for the
quantification of microgram quantities of protein utilizing
the principle of protein-dye binding. Analytical
Biochemistry 72:248-254.

Braissant, O., R. Foufelle, C. Scotto, M. Dauca and W.
Wahli. 1996. Differential expression of peroxisome
proliferator activated receptors (PPARs): tissue
distribution of PPAR-a, -B and -y in the adult rat.
Endocrinology 137:354-366.

Brasier, A.R., D. Ron, J.E. Tate and J.F. Habener. 1990. A
family of constitutive C/EBP-like DNA binding proteins

attenuate the IL-ld induced, NFKB mediated trans-activation
of the angiotensinogen gene acute-phase response element.
EMBO Journal 9:3933-3944.

Brass, E.P. and W.H. Vetter. 1994. Interleukin-6, but not

tumor necrosis factor-a, increases lipogenesis in rat
-hepatocyte primary cultures. Biochemical Journal 301:193-
197.

Brown, S.B., M. Maloney and W.B. Kinlaw. 1997. “Spot 14”
protein functions at the pretranslational level in the
regulation of hepatic metabolism by thyroid hormone and
glucose. Journal of Biological Chemistry 272:2163-2166.

Brun, R.P., J.B. Kim, E. Hu, 8. Altiok, B.M. Spiegelman.
1996a. Adipocyte differentiation: a transcriptional

194

 

regulatory cascade. Current Opinion in Cell Biology 8:826-
832.

Brun, R.P., P. Tontonoz, B.M. Forman, R. Ellis, J. Chen,
R.M. Evans and B.M. Spiegelman. 1996b. Differential
activation of adipogenesis by multiple PPAR isoforms. Genes
and Development 10:974-984.

Butta, N., E. Urcelay, C. Gonzalez-Manchon, R. Parrilla and

M.S. Ayuso. 1993. Pertussis toxin inhibition of ar-
adrenergic or vasopressin-induced Cay'fluxes in rat liver.
Journal of Biological Chemistry 268:6081-6089.

Butterwith, S.C. 1994. Molecular events in adipocyte
development. Pharmacological Therapeutics 61:399-411.

Camandola, S., G. Leonarduzzi, T. Musso, L. Varesio, R.
Carini, A. Scavazza, E. Chiarpotto, P.A. Baeuerle and G.

Poli. 1996. Nuclear factor KB is activated by arachidonic
acid but not by eicosapentaenoic acid. Biochemical and
Biophysical Research Communications 229:643-647.

Camp, H.S. and S.R. Tafuri. 1997. Regulation of peroxisome
proliferator-activated receptor 9 activity by mitogen-
activated protein kinase. Journal of Biological Chemistry
272:10811-10816.

Castan, I., P. Valet, N. Quideau, T. Viosin, L. Ambid, M.
LaBurthe, M. Lafontan and C. Carpene. 1994. Antilipolytic

effects of a2-adrenergic agonists, neuropeptide Y,
adenosine, and PGE1 in mammal adipocytes. American Journal
of Physiology 266:R1141-R1147.

Castelien, H., T. Gulick, R.E. Declercq, G.P. Mannaerts,
D.A. Moore and M.L. Baes. 1994. The peroxisome proliferator
activated receptor regulates malic enzyme gene expression.
Journal of Biological Chemistry 269:26754-26758.

Chan, K.-M. and J. Turk. 1987. Mechanism of arachidonic
acid-induced Ca2+ mobilization from rat liver microsomes.
Biochimica et Biophysica Acta 928:186-193.

Chapman, A.B., D.M. Knight and G.M. Ringold. 1985.
Glucocorticoid regulation of adipocyte
differentiation:hormonal triggering of the developmental
program and induction of a differentiation-dependent gene.
Journal of Cell Biology 101:1227-1235. '

195

Chatzipanteli, K., C. Head, J. Megerman and L. Axelrod.
1996. The relationship between plasma insulin level,
prostaglandin production by adipose tissue, and blood
pressure in normal rats and rats with diabetes mellitus and
diabetic ketoacidosis. Metabolism 45:691-698.

Chatzipanteli, K., S. Rudolph and L. Axelrod. 1992.
Coordinate control of lipolysis by prostaglandin E2 and
prostacyclin in rat adipose tissue. Diabetes 41:927-935.

Chawla, A. and M.A. Lazar. 1994. Peroxisome proliferator
and retinoid signaling pathways co-regulate preadipocyte
phenotype and survival. Proceedings of the National Academy
of Sciences 91:1786-1790.

Chawla, A., E.J. Schwarz, D.D. Dimaculangan and M.A. Lazar.

1994. Peroxisome proliferator-activated receptor (PPAR) 1:
adipose-predominant expression and induction early in
adipocyte differentiation. Endocrinology 135:798-800

Cheneval, D., R.J. Christy, D. Geiman, P. Cornelius and M.D.
Lane. 1991. Cell-free transcription directed by the 422
adipose P2 gene promoter: Activation by the CCAAT/enhancer
binding protein. Proceedings of the National Academy of
Sciences 88:8465-8469.

Christ, E.J. and D.H. Nugteren. 1970. The biosynthesis and
possible function of prostaglandins in adipose tissue.
Biochimica et Biophysica Acta 218:296-307.

Christie, P.E., J.P. Arm and T.H. Lee. 1993. Leukotrienes.
Mast Cell in Health and Disease, Marcell Dekker, Inc., New
York, NY, 62:393-407.

Christy, R.J., K.H. Kaestner, D.E. Geiman and M.D. Lane.
1991. CCAAT/enhancer binding protein gene promoter: Binding
of nuclear factors during differentiation of 3T3-L1
preadipocytes. Proceedings of the National Academy of
Sciences 88:2593-2597.

Christy, R.J., V.W. Yang, J.M. Ntambi, D.E. Geiman, W.H.
Landschulz, A.D. Friedman, Y. Nakabeppu, T.J. Kelly and M.D.
Lane. 1989. Differentiation-induced gene expression in
3T3-L1 preadipocytes: CCAAT/enhancer binding protein
interacts with and activates the promoters of two adipocyte-
specific genes. Genes and Development 3:1323-1335.

Clark, J.D., A.R. Schievella, E.A. Nalefski, L.-L. Lin.

1995. Cytosolic phospholipase A2. Journal of Lipid
Mediators and Cell Signaling 12:83-117.

196

Clarke, S.D. and S. Abraham. 1992. Gene
expression:nutrient control of pre- and posttranscriptional
events. FASEB 6:3146-3152.

FAS, 514, review, PUFA

Clarke, S.D., M.K. Armstrong and D.B. Jump. 1990a. Dietary
polyunsaturated fats uniquely suppress rat liver fatty acid
synthase and $14 mRNA content. Journal of Nutrition
120:225-231.

Clarke, S.D., M.K. Armstrong and D.B. Jump. 1990b.
Nutritional control of rat liver fatty acid synthase and $14
mRNA abundance. Journal of Nutrition 120:218-224.

Clarke, S.D. and D.B. Jump. 1993. Regulation of gene
transcription by polyunsaturated fatty acids. Progress in
Lipid Research 32:139-149.

Clarke, S.D. and D.B. Jump. 1993. Regulation of hepatic
gene expression by dietary fats:a unique role for
polyunsaturated fatty acids. Nutrition and Gene Expression,
CRC Press, pp. 227-246.

Clarke, S.D. and D.B. Jump. 1994. Dietary polyunsaturated
fatty acid regulation of gene transcription. Annual Reviews
of Nutrition 14:83-98.

Clarke, S.D. and D.B. Jump. 1996. Polyunsaturated fatty
acid regulation of hepatic gene transcription. Lipids
31(Suppl):S7-Sll.

Liver, PUFA, PPAR, review

Clarke, S.D., D.R. Romsos and G.A. Leveille. 1976.
Specific inhibition of hepatic fatty acid synthesis exerted
by dietary linoleate and linolenate in essential fatty acid
adequate rats. Lipids 11:485-490.

Clarke, S.D., D.R. Romsos and G.A. Leveille. 1977.
,Differential effects of dietary methyl esters of long-chain
saturated and polyunsaturated fatty acids on rat liver and
adipose tissue lipogenesis. Journal of Nutrition 107:1170-
1180.

Coleman, D.L. 1978. Obese and diabetes: two mutant genes
causing diabetes-obesity syndromes in mice. Diabetologia
14:141-148.

Coleman, R.A., B.C. Reed, J.C. Mackall, A.K. Student, M.D.
Lane and R.M. Bell. 1978. Selective changes in microsomal

197

 

enzymes of triacylglycerol phosphatidylcholine, and
phosphatidylethanolamine biosynthesis during differentiation
of 3T3-L1 preadipocytes. Journal of Biological Chemistry
253:7256-7261.

Coleman, R.A., W.L. Smith and S. Narumiya. 1994.
Classification of prostanoid receptors: properties,
distribution, and structure of the receptors and their
subtypes. Pharmacological Reviews 46:205-229.

Cooper, A.D. 1997. Hepatic uptake of chylomicron remnants.
Journal of Lipid Research 38:2173-2192.

Cornelius, P., O.A. MacDougald and M.D. Lane. 1994.
Regulation of adipocyte development. Annual Reviews of
Nutrition 14:99-129.

 

Cornelius, P., M. Marlowe, M.D. Lee and P.H. Pekala. 1990.

The growth factor-like effects of tumor necrosis a.
Journal of Biochemistry 265:20506-20516.

Cunningham, B.A., M. Maloney and W.B. Kinlaw. 1997. Spot
14 protein-protein interactions: evidence for both homo- and
heterodimer formation in vivo. Endocrinology: 138:5184-
5188.

Currie, R.A. 1998. NF-Y is associated with the histone
acetyltransferases GCN5 and P/CAF. Journal of Biological
Chemistry 273:1430-1434.

Danesch, U., P.C. Weber and A. Sellmayer. 1994.
Arachidonic acid increases c-fos and Egr-1 mRNA in 3T3
fibroblasts by formation of prostaglandin Eziand activation

of protein kinase C. Journal of Biological Chemistry
269:27528-27263.

Danesch, U., P.C. Weber and A. Sellmayer. 1996.
Differential effects of n-6 and n-3 polyunsaturatd fatty
acids on cell growth and early gene expression in Swiss 3T3
.fibroblasts. Journal of Cellular Physiology 168:618-624.

Darlington, G.J., N. Wang and R.W. Hanson. 1995. C/EBPa:a
critical regulator of genes governing integrative metabolic
processes. Current Opinion in Genetics and Development
5:565-570.

Demoz, A., H. Vaagenes, N. Aarsaether, E. Hvattum, J.
Skorve, M. Gottlicher, J.R. Lillehaug, G.G. Gibson, J.-A.
Gustafsson, S. Hood and R.K. Berge. 1994. Coordinate
induction of hepatic fatty acyl-CoA oxidase and P4504A1 in

198

rat after activation of the peroxisome proliferator-
activated receptor (PPAR) by sulphur-substituted fatty acid
analogues. Xenobiotica 24:943-956

De Vos, P., A.-M. Lefebvre, S.G. Miller, M. Guerro-Millo, K.
Wong, R. Saladin, L.G. Hamann, B. Staels, M.R. Briggs and J.
Auwerx. 1996. Thiazolidinediones repress ob gene
expression in rodents via activation of peroxisome

proliferator activated receptor 7. Journal of Clinical
Investigation 98:1004-1009.

Doglio, A., C. Dani, P. Grimaldi and G. Ailhaud. 1986.
Growth hormone regulation of the expression of
differentiation-dependent genes in preadipocyte Ob1771
cells. Biochemical Journal 238:123-129.

Duddy, S.K., C.E.N. Kass and S. Orrenius. 1989. Ca”-
mobilizing hormones stimulate Cafl'efflux from hepatocytes.
Journal of Biological Chemistry 264:20863-20866.

Ek, B.A., D.P. Cistola, J.A. Hamilton, T.L. Kaduce and A.A.
Spector. 1997. .Fatty acid binding proteins reduce 15-
lipoxygenase-induced oxygenation of linoleic acid and
arachidonic acid. Biochmica et Biophysica Acta 1346:75-85.

Feingold, K.R., M. Soued, M.K. Serio, A. Adi, A.H. Moser and
C. Grunfeld. 1990. The effect of diet on tumor necrosis

factor stimulation of hepatic lipogenesis. Metabolism
39:623-632.

Feingold, K.R., W. Doerrler, C.A. Dinarello, W. Fiers and C.
Grunfeld. 1992. Stimulation of lipolysis in cultured fat
cells by tumor necrosis factor, interleukin-1, and the
interferons is blocked by inhibition_of prostaglandin
synthesis. Endocrinology 130:10-16.

Fernando, K.C. and G.J. Barritt. 1994. Evidence from
studies with hepatocyte suspensions that store-operated Ca2+
inflow requires a pertussis toxin-sensitive trimeric G-
.protein. Biochemical Journal 303:351-356.

Fernando, K.C., R.B. Gregory, F. Katsis, B.E. Kemp and G.J.
Barritt. 1997. Evidence that a low-molecular-mass GTP-
binding protein is required for store-activated Cay'inflow
in hepatocytes. Biochemical Journal 328:463-471.

Fitzpatrick, E.A. and R.C. Murphy. 1989. Cytochrome P-450
metabolism of arachidonic acid:formation and biological
actions of “epoxygenase”-derived eicosanoids.
Pharmacological Reviews 40:229-241.

199

Flatmark, T., A. Nilsson, J. Kvannes, T.S. Eikhorn, M.H.
Fukami, H. Kryvi and E.N. Christiansen. 1988. On the

mechanism of induction of enzymes systems for peroxisomal B-
oxidation of fatty acids in rat liver by diets rich in
partially hydrogenated fish oil. Biochimica et Biophysica
Acta 962:122-130.

Flick, P.K., J. Chen and P.R. Vagelos. 1977. Effect of
dietary linoleate on synthesis and degradation of fatty acid

synthetase from rat liver. Journal of Biological Chemistry
252:4242-4249.

Task Force Report. Food Fats and Health. Council for
Agricultural Sciences and Technology, Ames, Iowa: 1991

Forman, B.M., J. Chen and R.M. Evans. 1997. Hypolipidemic
drugs, polyunsaturated fatty acids, and eicosanoids are

ligands for peroxisome proliferator-activated receptors a

and 7. Proceedings of the National Academy of Sciences
94:4312-4317.

Forman, B.M., P. Tontonoz, J. Chen, R.P. Brun, B.M.
Spiegelman and R.M. Evans. 1995. lS-Deoxy-A 12,14-
Prostaglandin J2 is a ligand for the adipocyte determination

factor PPARy. Cell 83:803-812.

Fourney, R.M., J. Miyakoshi, R.S. Day III and M.C. Paterson.
Northern blotting: Efficient RNA staining and transfer.
Focus 10:5-6.

Fournier, T., V. Fadok and P.M. Henson. 1997. Tumor
necrosis factor-a inversely regulates prostaglandin.Eb and

prostaglandin.Eb production in murine macrophages. Journal
of Biological Chemistry 272:31065-31072.

Freytag, S.O., D.L. Paielli and J.D. Gilbert. 1994.

Ectopic expression of the CCAAT/enhancer -binding protein a
promotes the adipogenic program in a variety of fibroblastic
'cell lines. Genes and Development 8:1654-1663.

Fulceri, R., A. Gamberucci, G. Bellomo, R. Giunti and A.
Benedett. 1993. CoA and fatty acyl-CoA derivatives
mobilize calcium from a liver reticular pool. Biochemical
Journal 295:663-669.

Gaillard, D., R. Negrel, M. Lagarde and G. Ailhaud. 1989.
Requirement and role of arachidonic acid in the

200

 

differentiation of pre-adipose cells. Biochemical Journal
257:389-397.

Garrity, M.J., B.P. Brass and R.P. Robertson. 1987. PGE
binding in isolated hepatocytes: regulation during fasting.
Advances in Prostaglandin, Thromboxane, and Leukotriene
Research 17:686-690.

Garrity, M.J., M.M. Reed and E.P. Brass. 1989. Coupling of
hepatic prostaglandin receptors to adenylate cyclase through
a pertussis toxin sensitive guanine nucleotide regulatory
protein. Journal of Pharmacology and Experimental
Therapeutics 248:979-983.

Garrity, M.J., K.R. Westcott, T.L Eggerman, N.H. Andersen,
D.R. Stormm and R.P. Robertson. 1983. Interrelationships
between PGE; and PGI; binding and stimulation of adenylate
cyclase. American Journal of Physiology 244:E367-E372.

Gomez-Munoz, A., P. hales and D.N. Brindley. 1991.
Unsaturated fatty acids activate glycogen phosphorylase in
cultured rat hepatocytes. Biochemical Journal 276:209-215.

Gonzalez, E.J. 1997. The role of peroxisome proliferator
activated receptor a in peroxisome proliferation,
physiological homeostasis, and chemical carcinogenesis.
Advances in Experimental Medicine and Biology 422:109-125.

Gossett, R.E., A.A. Frolov, J.B. Roths, W.D. Behnke, A.B.
Kier and F. Schroeder. 1996. Acyl-CoA binding proteins:
Multiplicity and Function. Lipids 31:895-918.

Gottlicher, M., E. Widmark, Q. Li and J.-A. Gustafsson.
1992. Fatty acids activate a Chimera of the clofibric acid-
activated receptor and the glucocorticoid receptor.
Proceedings of the National Academy of Sciences 89:4653-
4657.

Graber, M.N., A. Alfonso and D.L. Gill. 1997. Recovery of
.Ca“'pools and growth in Ca2+ pool-depleted cells is
mediated by specific epoxyeicosatrienoic acids derived from
arachidonic acid. Journal of Biological Chemistry
272:29546-29553.

Graves, R.A., P. Tontonoz, S.R. Ross and B.M. Spiegelman.
1991. Identification of a potent adipocyte-specific
enhancer:involvement of an NF-l-like factor. Genes and
Development 5:428-437.

201

Graves, R.A., P. Tontonoz and B.M. Spiegelman. 1992.
Analysis of tissue-specific enhancer:ARF6 regulates
adipogenic gene expression. Molecular and Cellular Biology
12:1202-1208.

Grimaldi, P.A., S.M. Knobel, R.R. Whitesell and N.A.
Abumrad. 1992. Induction of aP2 gene expression by
nonmetabolized long-chain fatty acids. Proceedings of the
National Academy of Sciences 89:10930-10934.

Grunfeld, C., S. Adi, M. Soued, A. Moser, W. Feirs and K.R.
Feingold. 1990. Search for mediators of the lipogenic
effects of tumor necrosis factor: potential role of
interleukin 6. Cancer Research 50:4233-4238.

Hallakou, S., L. Doare, F. Foufelle, M. Kergoat, M. Guerre-
Millo, M.F. Berthault, I. Dugail, J. Morin, J. Auwerx and P.
Ferre. 1997. Pioglitazone induces in vivo adipocyte
differentiation in the obese Zucker fa/fa rat. Diabetes
46:1393-1399.

Hamon, M., J. Berthaut, A. Wisner, F. Dray, B. Descomps and
H.T. Dao. 1993. Modulation of human myometrial PGE2
receptor by GTP characterization of receptor subtype.
Prostaglandins 46:251-268.

Hannun, Y.A. 1997. Sphingolipid second messengers: tumor
suppressor lipids. Proceedings of Experimental Medicine and
Biology 400A:305-312.

Hannun, Y.A and L.M. Obeid. 1995. Ceramide: an -
intracellular signal for apoptosis. Trends in Biological
Sciences 20:73-77.

Harbrecht, B.G., Y.M. Kim, B.M. Wirant, R.A. Shapiro and
T.R. Billiar. 1996. PGE2land LTBq inhibit cytokine-
stimulated nitric oxide synthase type 2 expression in
isolated rat hepatocytes. Prostaglandins 52:103-116.

_Harris, P.K. and R.F. Kletzien. 1994. Localization of a
pioglitazone response element in the adipocyte fatty acid-
binding protein gene. Molecular Pharmacology 45:439-445.

Hashimoto, N., T. Watanabe, Y. Ikeda, H. Yamada, S.
Taniguchi, H. Mitsui and K. Kurokawa. 1997. Prostaglandins
induce proliferation of rat hepatocytes through a
prostaglandin E2 receptor EP3 subtype. American Journal of
Physiology 272:6597-604.

202

 

Hauner, H., Th. Petruschke, M. Russ, K. Rohrig and J. Eckel.

1995. Effects of tumour necrosis factor alpha (TNFa) on
glucose transport and lipid metabolism of newly-
differentiated human fat cells in cell culture.
Diabetologia 38:764-771.

Herrara, R., H.-S. Ro, G.S. Robinson, K.G. Xanthopoulos and
B.M. Spiegelman. 1989. A direct role for C/EBP and the AP-
1-binding site in gene expression linked to adipocyte
differentiation. Molecular and Cellular Biology 9:5331-
5339.

Hertz, R., I. Berman, D. Keppler and J. Bar-Tana. 1996.
Activation of gene transcription by prostacyclin analogues
is mediated by the peroxisome-proliferators activated
receptor (PPAR). European Journal Biochemistry 235:242-247.

Hespeling, U., K. Jungemann and G.P. Puschel. 1995a.
Feedback-inhibition of glucagon-stimulated glycogenolysis in
hepatocyte/Kupffer cell cocultures by glucagon-elicited
prostaglandin production in Kupffer cells. Hepatology
22:1577-1583.

Hespeling, U., G.P. Puschel, K. Jungemann O. Gotz and J.
Zwirner. 1995b. Stimulation of glycogen phosphorylase in
rat hepatocytes via prostanoid release from Kupffer cells by
recombinant rat anaphylatoxin C5a but not by native human
C5a in hepatocyte/Kupffer cell co-cultures. FEBS Letters
372:108-112.

Hollenberg, A.N., V.S. Susulic, J.P. Madura, B. Zhang, D.E.
Moller, P. Tontonoz, P. Sarraf, B.M. Spiegelman and B.B.
Lowell. 1997. Functional antagonism between CCAAT/Enhancer
binding protein-a and peroxisome proliferator-activated

receptor-y on the leptin promoter. Journal of Biological
Chemistry 272:5283-5290. -

Hotamisligil, G.S, P. Arner, J.F. Caro, R.L. Atkinson and
B.M. Spiegelman. 1995. Increased adipose tissue expression
‘of tumor necrosis factor-a in human obesity and insulin
resistance. Journal of Clinical Investigations 95:2409-
2415.

Hotamisligil, 0.3., p. Peraldi, A. Budavari, R. Ellis, M.F.
White and B.M. Spiegelman. 1996b. IRS-1 mediated
inhibition of insulin receptor tyrosine kinase activity in

TNF-a and obesity-induced insulin resistance. Science
271:665-668.

203

 

Hu, E., J.B. Kim, P. Sarraf and B.M. Spiegelman. 1996.
Inhibition of adipogenesis through MAP kinase-mediated

phosphorylation of PPARy. Science 274:2100-2103.

Hu, E., P. Tontonoz and B.M. Spiegelman. 1995.
Transdifferentiation of myoblasts by the adipogenic

transcription factors PPARy and C/EBPa. Proceedings of the
National Academy of Sciences 92:9856-9860.

Hughes, B.P., J.N. Crofts, A.M. Auld, L.C. Read and G.J.
Barrit. 1987. Evidence that a pertussis-toxin-sensitive
substrate is involved in the stimulation by epidermal growth
factor and vasopressin of plasma-membrane Caa'inflow in
hepatocytes. Biochemical Journal 248:911-918.

Hunter, J., A. Kassam, C.J. Winrow, R.A. Rachubinski and
J.P. Capone. 1996. Crosstalk between the thyroid hormone
and peroxisome proliferator-activated receptors in
regulation peroxisome proliferator-responsive genes.
Molecular and Cellular Endocrinology 116:213-221.

Hwang, D.H., M. Boudreau, P. Chanmugam. 1988. Dietary
linolenic acid and longer chain n-3 fatty acids: comparison
of effects on arachidonic acid metabolism in rats. Journal
of Nutrition 118:427-437.

Hyman, B.T., L.L. Stoll and A.A. Spector. 1982.
Prostaglandin production by 3T3-L1 cells in culture.
Biochimica et Biophysica Acta 713:375-385.

Im, D.S., T. Fujioka, T. Katada, Y. Kondo, M. Ui and F.
Okajima. 1997. Characterization of Sphingosine 1-
phosphate-induced actions and its signaling pathways in rat
hepatocytes. American Journal of Physiology 35:91091-1099.

Irvine, R.F. 1982. How is the level of free arachidonic
acid controlled in mammalian cells? Biochemical Journal
204:3-16.

Review, AA

Izawa, T., T. Mochizuki, T. Komabayashi, K. Suda and M.
Tsuboi. 1994. Increase in cytosolic free CaB'in
corticotropin-stimulated white adipocytes. American Journal
of Physiology 266:E418-E426.

Jayadev, S., C.M. Linardic and Y.A. Hannun. 1994.
Identification of arachidonic acid as a mediator of
sphingomyelin hydrolysis in response to tumor necrosis

factor a. Journal of Biological Chemistry 269:5757-5763.

204

 

Jiang, M.S., Q. Tang, J. McLenithan, D. Geiman, W.
Shillinglaw, W.J. Henzel and M.D. Lane. 1998. Derepression

of the C/EBPa gene during adipogenesis: identification of

AP-Za as a repressor. Proceedings of the National Academy
of Sciences 95:3467-3471.

Jones, B.H., J.H. Kim, M.B. Zemel, R.P. Woychik, E.J.
Michaud, W.O. Wilkison and N. Moustaid. 1996. Upregulation
of adipocyte metabolism by agouti protein: possible
paracrine actions in yellow mouse obesity. American Journal
of Physiology 270:E192-El96.

Jump, D.B. 1989. Rapid induction of rat liver $14 gene
transcription by thyroid hormone. Journal of Biological
Chemistry 264:4689-4703.

Jump, D.B., M. Badin and A. Thelen. 1997b. The CCAAT box
binding factor, NF-Y, is required for thyroid hormone
regulation of rat liver Sl4 gene transcription. Journal of
Biological Chemistry 272:27778-27786.

Jump, D.B., A. Bell and V. Santiago. 1990. Thyroid hormone
and dietary carbohydrate interact to regulate rat liver Sl4
gene transcription and chromatin structure. Journal of
Biological Chemistry 265:3474-3478.

Jump, D.B., S.D. Clarke, 0. MacDougald and A. Thelen. 1993.
Polyunsaturated fatty acids inhibit Sl4 gene transcription
in rat liver and cultured hepatocytes. Proceedings of the
National Academy of Sciences 90:8454-8458.

Jump, D.B., S.D. Clarke, A. Thelen and M. Liimatta. 1994.
Coordinate regulation of glycolytic and lipogenic gene
expression by polyunsaturated fatty acids. Journal of Lipid
Research 35:1076-1084.

Jump, D.B., S.D. Clarke, A. Thelen, M. Liimatta, B. Ren and
M. Badin. 1997a. Dietary fat, genes and human health.
.Dietary Fat and Cancer, Plenum Press, NY, pp. 167-176.

Jump, D.B., S.D. Clarke, A. Thelen, M. Liimatta, B. Ren and
M. Badin. 1996. Dietary polyunsaturated fatty acid
regulation of gene transcription. Progress in Lipid
Research 35:227-241.

Jump, D.B. and O.A. MacDougald. 1993. Hormonal regulation

of gene expression in cultured adipocytes. Journal of
Animal Science 71(Suppl):56-64.

205

 

Jump, D.B., P. Narayan, H. Towle and J.H. Oppenheimer.
1984. Rapid effects of triiodothyronine on hepatic gene
expression. Journal of Biological Chemistry 259:2789-2797.

Jump, D.B. and J.H. Oppenheimer. 1985. High basal
expression and 3,5,3-triiodothyronine regulation of
messenger ribonucleic acid 814 in lipogenic tissues.
Endocrinology 117:2259-2266.

Jump, D.B., B. Ren, S. Clarke and A. Thelen. 1995. Effects
of fatty acids on hepatic gene expression. Prostaglandins
Leukotrienes and Essential Fatty Acids 52:107-111.

Jump, D.B., A. Veit, V. Santiago, G. Lepar and L.
Herberholz. 1988. Transcriptional activation of the rat
liver Sl4 gene during postnatal development. Journal of
Biological Chemistry 263:7254-7260.

Jungermann, K. and T. Kietzmann. 1996. Zonation of
parenchymal and nonparenchymal metabolism in liver. Annual
Reviews of Nutrition 16:179-203.

Kaestner, K.H., R.J. Christy and M.D. Lane. 1990. Mouse
insulin-responsive glucose transporter gene:
Characterization of the gene and trans-activation by the
CCAAT/enhancer binding protein. Proceedings of the National
Academy of Sciences 87:251-255.

Kallen, C.B. and M.A. Lazar. 1996. Antidiabetic
thiazolidinediones inhibit leptin (ob) gene expression in
3T3-L1 adipocytes. Proceedings of the National Academy of
Sciences 93:5793-5796.

Kanemaki, T., H. Kitade, M. Kaibori, K. Sakitani, Y.
Hiramatsu, Y. Kamiyama, S. Ito and T. Okumura. 1998.
Interleukin lbeta and interleukin 6, but not tumor necrosis
factor alpha, inhibit insulin-stimulated glycogen synthesis
in rat hepatocytes. Hepatology 27:1296-1303.

Kanemaki, T., H. Kitade, Y. Hiramatsu, Y. Kamiyama and T.
'Okumura. 1993. Stimulation of glycogen degradation by
prostaglandin E2 in primary cultured rat hepatocytes.
Prostaglandins 45:459-474.

Kass, G.E.N., J. Llopis, S.C. Chow, S.K. Duddy and S.
Orrenius. 1990. Receptor-operated calcium influx in rat
hepatocytes. Journal of Biological Chemistry 265:17486-
17492.

206

 

Keller, H., C. Dreyer, J. Medin, A. Mahfoudi, K. Ozato and
W. Wahli. 1993. Fatty acids and retinoids control lipid
metabolism through activation of peroxisome proliferator-
activated receptor-retinoid X receptor heterodimers.
Proceedings of the National Academy of Sciences 90:2160-
2164.

Khan, S.H. and S. Sorof. 1990. Preferential binding of
growth inhibitory prostaglandins by the target protein of a
carcinogen. Proceedings of the National Academy of Sciences
87:9401-9405.

Kim, J.B. and B.M. Spiegelman. 1996. ADDl/SREBPl promotes
adipocyte differentiation and gene expression linked to
fatty acid metabolism. Genes and Development 10:1096-1107.

Kim, J.B., G.D. Spotts, Y.-D. Halvorsen, H.-M. Shih, T.
Ellenberger, H.C. Towle and B.M. Spiegelman. 1995. Dual
DNA binding specificity of ADDl/SREBPl controlled by a
single amino acid in the basic helix-loop-helix domain.
Molecular and Cellular Biology 15:2582-2588.

Kim, T.-S. and H.C. Freake. 1996. High carbohydrate diet
and starvation regulate lipogenic mRNA in rats in a tissue-
specific manner. Journal of Nutrition 126:611-617.

Kinlaw, W.B., P. Tron and L.A. Witters. 1993.
Endocrinology 133:645-650.

Kinlaw, W.B., J.L. Church, J. Harmon and C.N. Mariash.
1995. Direct evidence for the role of “Spot 14” protein in
the regulation of lipid synthesis. Journal of Biological
Chemistry 270:16615-16618. -

Kinsella, J.E., K.S. Broughton and J.W. Whelan. 1990.
Dietary unsaturated fatty acids: interactions and possible
needs in relation to eicosanoid synthesis. Journal of
Nutritional Biochemistry 1:123-141.

‘Kliewer, S.A., J.M. Lenhar, T.M. Willson, I. Patel, D.C.
Morris and J.M. Lehmann. 1995. A prostaglandin J2
metabolite binds peroxisome proliferator activated receptor

1 and promotes adipocyte differentiation. Cell 83:813-819.

Kliewer, S.A., S.S. Sundsett, S.A. Jones, P.J. Brown, G.B.
Wisely, C.S. Koble, P. Devchand, W. Wahli, T.M. Willson,
J.M. Lenhard and J.M. Lehmann. 1997. Fatty acids and
eicosanoids regulate gene expression through direct
interactions with peroxisome proliferator-activated

207

 

receptors a and 7. Proceedings of the National Academy of
Sciences 94:4318-4323.

Lamphere, L., C.L. Carpenter, Z.F. Sheng, R.G. Kallen and
G.E. Lienhard. 1994. Activation of PI 3 kinase in 3T3-L1
adipocytes by association with insulin receptor substrate-1.
American Journal of Physiology 266:E486-E494.

Landshulz, K.T., D.B. Jump, O.A. MacDougald and M.D. Lane.
1994. Transcriptional control of the stearoyl-CoA
desaturase-1 gene by polyunsaturated fatty acids.

Biochemical and Biophysical Research Communications 200:763-
768. -

Lane, M.D., F.-T. Lin, O.A. MacDougald and M. Vasseur-
Cognet. 1996. Control of adipocyte differentiation by

CCAAT/enhancer binding protein a (C/EBPa). International
Journal of Obesity 20:891-896.

Laneuville, O., D.K. Breur, N. Xu, Z.H. Huang, D.A. Gage,
J.T. Watson, M. Lagarde, D.L. Dewitt and W.L. Smith. 1995.
Fatty acid substrate specificities of human prostaglandin-
endoperoxide H sythase-l and -2. Journal of Biological
Chemistry 270:19330-19336.

Lehmann, J.M., J.M. Lenhard, B.B. Oliver, G.M. Ringold and
S.A. Kliewer. 1997. Peroxisome proliferator-activated

receptors a and y are activated by indomethacin and other
non-steroidal anti-inflammatory drugs. Journal of
Biological Chemistry 272:3406-3410.

Lehmann, J.M., L.B. Moore, T.A. Smith-Oliver, W.O. Wilkison,
T.M. Willson and S.A. Kliewer. 1995. An Antidiabetic
thiazolidinedione is a high affinity ligand for peroxisome

proliferator-activated receptor 1 (PPARy). Journal of
Biological Chemistry 270:12953-12956.

Lepak, N.M. and G. Serrero. 1995. Prostaglandin F2a

stimulates transforming growth factor-a expression in
adipocyte precursors. Endocrinology 136:3222-32229.

Lepar, G.J. and D.B. Jump. 1989. Hormonal regulation of
the 814 gene in 3T3-F442A cells. Molecular Endocrinology
3:1207-1214.

Lepar, G.J. and D.B. Jump. 1992. Retinoic acid and
dexamethasone interact to regulate Sl4 gene transcription in
3T3-F442A adipocytes. Molecular and Cellular Endocrinology
84:65-72.

208

Leung, L.K. and H.P. Glauert. 1996. Reduction of the

concentrations of prostaglandins E2 and F20, and
thromboxane B2 in cultured rat hepatocytes treated with the
peroxisome proliferator ciprofibrate. Toxicology Letters
85:143-149.

Liimatta, M., H.C. Towle, S. Clarke and D.B. Jump. 1994.
Dietary polyunsaturated fatty acids interfere with the
insulin/glucose activation of L-type pyruvate kinase gene
transcription. Molecular Endocrinology 8:1147-1153.

Lin, F.-T. and M.D. Lane. 1992. Antisense CCAAT/enhancer-
binding protein RNA suppresses coordinated gene expression
and triglyceride accumulation during differentiation of 3T3-
Ll preadipocytes. Genes and Development 6:533-544.

Lin, F.-T. and M.D. Lane. 1994. CCAAT/enhancer binding

protein a is sufficient to initiate the 3T3-L1 adipocyte
differentiation program. Proceedings of the National
Academy of Sciences 91:8757-8761.

Long, S. and P.H Pekala. 1996a. Lipid mediators of insulin
resistance:ceramide signaling down-regulates GLUT4 gene
transcription in 3T3-L1 adipocytes. Biochemical Journal
319:179-184.

Long, S. and P.H Pekala. 1996b. Regulation of GLUT4 gene
expression by arachidonic acid. Journal of Biological
Chemistry 271:1138-1144.

MacDougald, O.A., C.-S. Hwang, H. Fan and M.D. Lane. 1995.
Regulated expression of the obese gene product (leptin) in
white adipose tissue and 3T3-L1 adipocytes. Proceedings of
the National Academy of Sciences 92:9034-9037.

MacDougald, O.A. and D.B. Jump. 1992. Localization of an
adipocyte-specific retinoic acid response domain controlling
Sl4 gene transcription. Biochemical and Biophysical
.Research Communications 188:470-476.

Mackall, J.C., A.K. Student, S.E. Polakis and M.D. Lane.
1976. Induction of lipogenesis during differentiation in a
preadipoctye cell line. Journal of Biological Chemistry
251:6462-6464.

Makita, K., J.R. Falck and J.H. Capdevila. 1996.
Cytochrome P450, the arachidonic acid cascade, and
hypertension:new vistas for an old enzyme system. FASEB
Journal 10: 1456-1463.

209

 

Mandrup, S. and M.D. Lane. 1997. Regulating adipogenesis.
Journal of Biological Chemistry 272:5367-5370.

McKay, L.I. and J.A. Cidlowski. 1998. Cross-talk between

nuclear factor-KB and the steroid hormone receptors:
mechanisms of mutual antagonism. Molecular Endocrinology
12:45-56.

Melien, O., R. Winsnes, M. Refsnes, I.P. Gladhaug and T.
Christoffersen. 1988. Pertussis toxin abolishes the

inhibitory effects of prostaglandins E1, E2, I2 and F20. on
hormone-induced CAMP accumulation in cultured hepatocytes.
European Journal of Biochemistry 172:293-297.

Mellgren, G., T. Bruland, A.P. Doskeland, T. Flatmark, O.K.
Vintermyr and 8.0. Doskeland. 1997. Synergistic
antiproliferative actions of cyclic adenosine 3',5'-
monophosphate, interleukin-18, and activators of

Cachalmodulin-dependent protein kinase in primary
hepatocytes. Endocrinology 138:4373-4383.

Miller, C.W., D.A. Casimir and J.M. Ntambi. 1996a. The
mechanism of inhibition of 3T3-L1 preadipocyte '

differentiation by prostaglandin Fun Endocrinology
137:5641-5650.

Miller, C.W. and J.M. Ntambi. 1996. Peroxisome
proliferators induce mouse liver stearoyl CoA desaturase 1

gene expression. Proceedings of the National Academy of
Sciences 93:9443-9448.

Miller, S.G., P.D. Vos, M. Guerro-Millo, K. Wong, T.
Hermann, B. Staels, M.R. Briggs and J. Auwerx. 1996b. The
adipocyte specific transcription factor C/EBPa modulates ob

gene expression. Proceedings of the National Academy of
Sciences 93:5507-5511.

Mine, T., I. Kojima and E. Ogata. 1990. Mechanism of
prostaglandin E2-induced glucose production in rat
hepatocytes. Endocrinology 126:2831-2836.

Mion, F., R. Jasuja and D.E. Johnston. 1995. The
contribution of hepatocytes to prostaglandin synthesis in
rat liver. Prostaglandins Leukotrienes and Essential Fatty
Acids 53:109-115.

Moncur, J.T., J.P. Park, V.A. Memoli, T.K. Mohandas and W.B.
Kinlaw. 1998. The “Spot 14" gene resides on the telomeric

210

 

end of the 11q13 amplicon and is expressed in lipogenic
breast cancers: implications for control of tumor
metabolism. Proceedings of the National Academy of Sciences
95:6989-6944.

Moustaid, N., R.S. Beyer and H.S. Sul. 1994.
Identification of an insulin response element in the fatty

acid synthase promoter. Journal of Biological Chemistry
269:5629-5634.

Moustaid, N., K. Sakamoto, S. Clarke, R.S. Beyer and H.S.
Sul. 1993. Regulation of fatty acid synthase gene
transcription. Sequences that confer a positive insulin
effect and differentiation-dependent expression in 3T3-L1
preadipocytes are present in the 332 bp promoter.
Biochemical Journal 292:767-772.

Moustaid, N. and H.S. Sul. 1991. Regulation of expression
of the fatty acid synthase gene in 3T3-L1 cells by
differentiation and triiodothyronine. Journal of Biological
Chemistry 266:18550-18554.

Nagai, M., K. TuChiya and H. Kojima. 1996. Prostaglandin
E2 increases the calcium concentration in rat brown
adipocytes and their consumption of oxygen. Prostaglandins
51:377-386.

Neat, C.E., M.S. Thomassen and H. Osmunden. 1980.

Induction of peroxisomal B-oxidation of rat liver by high
fat diets. Biochemical Journal 186:369-371.

Negishi, M., Y. Sugimoto and A. Ichikawa. 1995.
Prostaglandin receptors. Journal of Lipid Mediators and
Cell Signaling 12:379-391.

Negrel, R., D. Gaillard and G. Ailhaud. 1989. Prostacylin
as a potent effector of adipose-cell differentiation.
Biochemical Journal 257:399-405.

,Nikolova-Karakashian, M., E.T. Morgan, C. Alexander, D.C.
Liotta and A.H. Merrill, Jr. 1997. Bimodal regulation of

ceramidase by interleukin-18. Journal of Biological
Chemistry 272:18718-18724.

Ninomiya-Tsuji, J. P.M. Torti and G.M. Ringold. 1993.
Tumor necrosis factor-induced c-myc expression in the
absence of mitogenesis is associated with inhibition of
adipocyte differentiation. Proceedings of the National
Academy of Sciences 90:9611-9615.

211

 

Ntambi, J.M. and T. Takova. 1996. Role of Cay'in the
early stages of murine adipocyte differentiation as
evidenced by calcium mobilizing agents.

Oliw, E.H. 1994. Oxygenation of polyunsaturated fatty
acids by cytochrome P450 monooxygenases. Progress in Lipid
Research 33:329-354. '

Ota, Y., A. Mariash, J.L. Wagner and C.N. Mariash. 1997.
Cloning, expression and regulation of human $14 gene.
Molecular and Cellular Endocrinology 126:75-81.

Parker, J., J. Lane and L. Axelrod. 1989. Cooperation of
adipocytes and endothelial cells required for catecholamine
stimulation of PGI2 production by rat adipose tissue.
Diabetes 38:1123-1132.

Paulauskis, J.D. and H.S. Sul. 1988. Cloning and
expression of mouse fatty acid synthase and other specific
mRNAs. Development and hormonal regulation in 3T3-L1 cells.
Journal of Biological Chemistry 263:7049-7057.

Porras, A. and E. Santos. 1996. The insulin/RAS pathway of
adipocytic differentiation of 3t3 L1 cells:dissociation
between Raf-1 kinase and the MAPK/RSK cascade.

International Journal of Obesity 20(Suppl 3):S43-551.

Puschel, G.P. and B. Christ. 1994. Inhibition by PGE2iof
glucagon-induced increase in phosphoenolpyruvate
carboxykinase mRNA and acceleration of mRNA degradation in
cultured rat hepatocytes. FEBS Letters 351:353-356.

Raza, H., J.R. Pongubala and S. Sorof. 1989. Specific high
affinity binding of lipoxygenase metabolites of arachidonic
acid by liver fatty acid binding protein. Biochemical and
Biophysical Research Communications 161:448-455.

Reddy, J.K. and G.P. Mannaerts. 1994. Peroxisomal lipid
metabolism. Annual Reviews of Nutrition 14:343-370.

.Refsnes, M. O.F. Dajani, D. Sandnes, G.H. Thoresen, J.
Rottingen, J.-G. Iversen and T. Christoffersen. 1995. On
the mechanisms of the growth-promoting effect of
prostaglandins in hepatocytes:The relationship between
stimulation of DNA synthesis and signaling mediated by
adenylyl cyclase and phosphoinositide-specific phospholipase
C. Journal of Cellular Physiology 164:465-473.

Reginato, M.J., S.L. Krakow, S.T. Bailey and M.A. Lazar.
1998. Prostaglandins promote and block adipogenesis through

212

 

opposing effects on peroxisome proliferator-activated
receptor 7. Journal of Biological Chemistry 272:1855-1858.

Ren, B., A.P. Thelen, J.M. Peters, E.J. Gonzalez and D.B.
Jump. 1997. Polyunsaturated fatty acid suppression of
hepatic fatty acid synthase and 814 gene expression does not

require peroxisome proliferator-activated receptor a.
Journal of Biological Chemistry 272:26827-26832.

Ren, B., A. Thelen and D.B. Jump. 1996. Peroxisome

proliferator-activated receptor a inhibits hepatic $14 gene
transcription. Journal of Biological Chemistry 271:17167-
17173.

Richelsen, B. 1987. Factors regulation the production of
prostaglandin E2 and prostacyclin (prostaglandin I2) in rat
and human adipocytes. Biochemical Journal 247:389-394.

Richelsen, B., J.D. Borglum and 8.8. Sorensen. 1992.
Biosynthetic capacity and regulatory aspects of
prostaglandin E2 formation in adipocytes. Molecular and
Cellular Endocrinology 85:73-81.

Ritz, E., C.C. Heuck and R. Boland. 1980. Phosphate,
calcium and lipid metabolism. Advances in Experimental and
Medical Biology 128:197-208.

Robertson, R.P., K.R. Westcott, D.R. Storm and M.G. Rice.
1980. Down regulation in vivo of PGE receptors and
adenylate cyclase stimulation. American Journal of
Physiology 239:e75-E85.

Rocchini, A.P. 1995. Insulin Resistance, Obesity and
Hypertension. Journal of Nutrition 125:17188-17243.

Rodriguez de Turco, E.B. and J.A. Spitzer. 1990.

Eicosanoid production in nonparenchymal liver cells isolated
from rats infused with E. coli endotoxin. Journal of
Leukocyte Biology 48:488-494.

Sadowski, H.B., T.T. Wheeler and D.A. Young. 1992. Gene
expression during 3T3-L1 adipocyte differentiation. Journal
of Biological Chemistry 266:4722-4731.

Schoonjans, K., J. Peinado-Onsurbe, A. Lefebvre, R.A.
Heyman, M. Briggs, S. Deeb, B. Staels and J. Auwerx. 1996a.

PPARa and PPARy activators direct a distinct tissue-specific
transcriptional response via a PPRE in the lipoprotein
lipase gene. EMBO Journal 15:5336-5348.

213

 

Schoonjans, K., B. Staels and J. Auwerx. 1996b. The
peroxisome proliferator activated receptors (PPARs) and
their effects on lipid metabolism and adipocyte
differentiation. Biochimica et Biophysica Acta 1302:93-109.

Schoonjans, K., B. Staels and J. Auwerx. 1996c. Role of
peroxisome proliferator-activated receptor (PPAR) in
mediating the effects of fibrates and fatty acids on gene
expression. Journal of Lipid Research 37:907-925.

Schoonjans, K., M. Watanabe, H. Suzuki, A. Mahfoudi, G.
Krey, W. Wahli, P. Grimaldi, B. Staels, T. Yamamoto and J.
Auwerx. 1995. Induction of the acyl-coenzyme A synthetase
gene by fibrates and fatty acids is mediated by a peroxisome
proliferator response element in the C promoter. Journal of
Biological Chemistry 270:19269-19276.

Schwarz, E.J., M.J. Reginato, D. Shao, S.L. Krakow and M.A.
Lazar. 1997. Retinoic acid blocks adipogenesis by

inhibiting C/EBPB-mediated transcription. Molecular and
Cellular Biology 17:1552-1561.

Sellmayer, A., U. Danesch and P.C. Weber. 1996. Effects of
different polyunsaturated fatty acids on growth-related
early gene expression and cell growth. Lipids 31:837-840.

Serrero, G., N.M. Lepak and S.P. Goodrich. 1992. Paracrine
regulation of adipose differentiation by arachidonate

metabolites: prostaglandin F2a inhibits early and late

markers of differentiation in the adipogenic cell line 1246.
Endocrinology 131:2545-2551.

Sessler, A.M., N. Kaur, J.P. Palta and J.M. Ntambi. 1996.
Regulation of stearoyl-CoA desaturase 1 mRNA stability by
polyunsaturated fatty acids in 3T3-L1 adipocytes. Journal
of Biological Chemistry 271:29854-29858.

Sessler, A.M. and J.M. Ntambi. 1998. Polyunsaturated fatty
~acid regulation of gene expression. Journal of Nutrition
128:923-926.

Shillabeer, G., Z.-H. Li, G. Hatch, V. Kumar and D.C.W. Lau.
1996. A novel method for studying preadipocyte
differentiation in vitro. International Journal of Obesity
20(Suppl3):S77-883.

Shillabeer, G. and D.C.W. Lau. 1994. Regulation of new fat

cell formation in rats:the role of dietary fats. Journal of
Lipid Research 35:592-600.

214

Skouteris, G.G. and M.R. Kaser. 1991. Prostaglandin E2iand

Fad mediate the increase in c-myc expression induced by EGF
in primary rat hepatocyte cultures. Biochemical and
Biophysical Research Communications 178:1240-1246.

Skouteris, G.G. and M. McMenamin. 1992. Transforming

growth factor-a—induced DNA synthesis and c-myc expression
in primary rat hepatocyte cultures is modulated by
indomethacin. Biochemical Journal 281:729-733.

Smas, C.M. and H.S. Sul. 1995. Control of adipocyte
differentiation. Biochemical Journal 309:697-710.

Smith, W.L. 1989. The eicosanoids and their biochemical
mechanisms of action. Biochemical Journal 259:315-324.

Smith, W.L. 1992. Prostanoid biosynthesis and mechanisms
of action. American Journal of Physiology 263:F181-191.

Smith, W.L., R.M. Garavito and D.L. Dewitt. 1996.
Prostaglandin endoperoxide H synthases (cyclooxygenases)-1
and -2. Journal of Biological Chemistry 271:33157-33160.

Smith, L.K., K.M. Rice and C.W. Garner. 1996. The insulin-
induced down-regulation of IRS-1 in 3T3-L1 adipocytes is
mediated by a calcium-dependent thiol protease. Molecular
and Cellular Endocrinology 122:81-92.

Soncini, M., S.-F. Yet, Y. Moon, J.-Y. Chun and H. S. Sul.
1995. Hormonal and nutritional control of the fatty acid
synthase promoter in transgenic mice. Journal of Biological
Chemistry 270:30339-30343.

Spiegelman, B.M. and J.S. Flier. 1996. Adipogenesis and
Obesity:Rounding out the big picture. Cell 87:377-389.

Stephens, J.M., R.F. Morrison and P.F. Pilch. 1996. The
expression and regulation on STATs during 3T3-ll adipocytes
.differentiation. Journal of Biological Chemistry 271:10441-
10444.

Stephens, J.M. and P.H. Pekala. 1992. Transcriptional
repression of the C/EPB-a and GLUT4 genes in 3T3-L1

adipocytes by tumor necrosis factor-a. Journal of
Biological Chemistry 267:13580-13584.

Strong, P., R.A. Coleman and P.P.A. Humphrey. 1992.
Prostanoid-induced inhibition of lipolysis in rat isolated

215

adipocytes:probable involvement of EP3 receptors.
Prostaglandins 43:559-566.

Stuhlmeier, K.M., and J.J. Kao and E.H. Bach. 1997.
Arachidonic acid influences proinflammatory gene induction

by stabilizing the inhibitor-KBa/nuclear factor-KB (NF-KB)
complex thus suppressing the nuclear translocation of NF-KB.

Stryer, L, editor.- 1988. Biochemistry, 31“! ed., W.H.
Freeman and Company, New York.

Sudo, Y. and C.N. Mariash. 1996. Lowering glucose depletes
a thapsigargin-sensitive calcium pool and inhibits
transcription of the $14 gene. Endocrinology 137:4677-4684.

Sudo, Y. and C.N. Mariash. 1994. Two glucose-signaling
pathways in $14 gene transcription in primary hepatocytes: a
common role of protein phosphorylation. Endocrinology
134:2532-2540.

Sugimoto, Y., T. Namba, A. Honda, Y. Hayashi, M. Negishi, A.
Ichikawa and S. Narumiya. 1992. Cloning and expression of
a cDNA for mouse prostaglandin E receptor EP3 subtype.
Journal of Biological Chemistry 267:6463-6466.

Sul, H.S., N. Moustaid, K. Sakamoto, C.M. Smas, N. Gekakis
and A. Jerkins. 1993a. Nutritional and hormonal regulation
of genes encoding enzymes involved in fat synthesis.
Nutrition and Gene Expression, CRC Press, pp. 207-226.

Sul, H.S., C.M. Smas and N. Moustaid. 1993b. Positive and
negative regulators of adipocyte differentiation. Journal
of Nutritional Biochemistry 4:554-562.

Tebbey, P.W. and T.M. Buttke. 1992. Arachidonic acid
regulates unsaturated fatty acid synthesis in lymphocytes by
inhibiting stearoyl-CoA desaturase gene expression.
Biochimica et Biophysica Acta 1171:27-34.

Tebbey, P.W., K. M. McGowan, J.M. Stephens, T.M. Buttke and
P.H. Pekala. 1994. Arachidonic acid down-regulates the
insulin-dependent glucose transporter gene (GLUT4) in 3T3-L1
adipocytes by inhibiting transcription and enhancing mRNA
turnover. Journal of Biological Chemistry 269:639-644.

Teboul, L., D. Gaillard, L. Staccini, H. Inaders, E.-Z. Amri
and P.A. Grimaldi. 1995. Thiazolidinediones and fatty
acids convert myogenic cells into adipose-like cells.
Journal of Biological Chemistry 270:28183-28187.

216

Thode, J., H.A. Perhadsingh, J.H. Ladenson, R. Hardy and
J.M. McDonald. 1989. Palmitic acid stimulates glucose
incorporation in the adipocyte by a mechanism likely
involving intracellular calcium. Journal of Lipid Research
30:1299-1305.

Tocci, M.J. and J.A. Schmidt. 1997. Interleukin-1:
Structure and function. Cytokines in Health and Disease,
Marcell Dekker, Inc., New York, NY, pp. 1-23.

Tontonoz, P., R.A. Graves, A.I. Budavari, H. Erdjument-
Bromage, M. Lui, E. Hu, P. Tempst and B.M. Spiegelman.
1994a. Adipocyte-specific transcription factor ARF6 is a
heterodimeric complex of two nuclear hormone receptors,

PPARy and RXRa. Nucleic Acids Research 22:5628-5634.

Tontonoz, P., E. Hu, J. Devine, E.G. Beale and B.M.

Spiegelman. 1995. PPARyZ regulates adipose expression of
the phosphoenolpyruvate carboxykinase gene. Molecular and
Cellular Biology 15:351-357.

Tontonoz, P., E..Hu, R.A. Graves, A.I. Budavari and B.M.

Spiegelman. 1994b. mPPARyZ:tissue-specific regulator of an
adipocyte enhancer. Genes and Development 8:1224-1234.

Tontonoz, P., E. Hu and B.M. Spiegelman. 1994c.

Stimulation of adipogenesis in fibroblasts by PPARmZ, a
lipid-activated transcription factor. Cell 79:1147-1156.

Tontonoz, P., J.B. Kim, R.A. Graves and B.M. Spiegelman.
1993. ADD1:a novel helix-loop-helix transcription factor
associated with adipocyte determination and differentiation.
Molecular and Cellular Biology 13:4753-4759.

Toussant, M.J., M.D. Wilson and S.D. Clarke. 1981.
Coordinate suppression of liver acetyl CoA carboxylase and
fatty acid synthase by polyunsaturated fat. Journal of
Nutrition 11:146-153.

Towle, H.C., E.N. Kaytor and H.-M. Shih. 1996. Metabolic
regulation of hepatic gene expression. Biochemical Society
Transactions 24:364-368.

Tracey, K.J. 1997. Tumor Necrosis Factor. Cytokines in
Health and Disease, Marcell Dekker, Inc., New York, NY, pp.
223-236.

Uehara, T., S. Hoshino, M. 01, Y. Tokumitsu and Y. Nomura.
1994. Possible involvement of phosphatidylinositol-specific

217

phospholipase C related to pertussis toxin-sensitive GTP-
binding proteins during adipocytes differentiation of 3T3-L1
fibroblast:negative regulation of protein kinase C.
Biochimica et Biophysica Acta 1124:302-310.

Valera, A., G. Solanes and F. Bosch. 1993. Calcium-
mobilizing effectors inhibit P-enolypyruvate carboxykinase
gene expression in cultured rat hepatocytes. FEBS 333:319-
324.

Vance, D.E. and J.E. Vance, eds. 1985. Biochemistry of
Lipids and Membranes, The Benjamin/Cummings Publishing
Company, Inc.

Varanasi, U., R. Chu, Q. Huang, R. Castellon, A.V. Yeldandi
and J.K. Reddy. 1996. Identification of a peroxisome
proliferator-responsive element upstream of the human
peroxisomal fatty acyl coenzyme A oxidase gene. Journal of
Biological Chemistry 271:2147-2155.

Vassaux, G., D. gaillard, C. Darimont, G. Ailhaud and R.
Negrel. 1992. Differential response of preadipocytes and
adipocytes to prostacyclin and prostaglandin
E2:physiological implications. Endocrinology 131:2393-2398.

Volpe, J.J. and P.R. Vagelos. 1976. Mechanism and
regulation of biosynthesis of saturated fatty acids.
Physiological Reviews 56:339-417.

Vu-Dac, N., K. Schoonjans, V. Kosykh, J. Dallongeville, J.-
C. Fruchart, B. Staels and J. Auwerx. 1995. Fibrates
increase human apolipoprotein A-II expression through
activation of the peroxisome proliferator-activated
receptor. Journal of Clinical Investigations 96:741-750.

Wagner, M.J., J.A. Sharp and W.C. Summers. 1981.
Nucleotide sequence of the thymidine kinase gene of herpes

simplex virus type 1. Proceedings of the National Academy
of Sciences 78:1441-1445.

Wang, H., M.S. Goligorski and C.C. Malbon. 1997. Temporal
activation of Ca2+ calmodulin-sensitive protein kinase type
II is obligate for adipogenesis. Journal of Biological
Chemistry 272:1817-1821.

Wang, J.-C., P.E. Stromstedt, R.M. O’Brien and D.K. Granner.
1996. Hepatic nuclear factor 3 is an accessory factor
required for the stimulation of phosphoenolpyruvate
carboxykinase gene transcription by glucocorticoids.
Molecular Endocrinology 10:794-800.

218

Wang, P., S.M. Tait, Z.F. Ba and I.H. Chaudry. 1995. Tumor

necrosis factor-a administration increases Kupffer cell
cyclic adenosine monophosphate levels. Shock 4:351-355.

Watanabe, T., K. Umegaki and W.L. Smith. 1986. Association
of a solubilized prostaglandin E2 receptor from renal
medulla with a pertussis toxin-reactive guanine nucleotide
regulatory protein. Journal of Biological Chemistry
261:13430-13439.

Wolf, B.A., J. Turk, W.R. Sherman and M.L. McDaniel. 1986.
Intracellular Ca2+ mobilization by arachidonic acid.
Journal of Biological Chemistry 261:3501-3511.

Wright, K.L, T.L. Moore, B.J. Vilen, A.M. Brown and J.P.Y.
Ting. 1995. Major histocompatibility complex class II-
associated invariant chain gene expression is up-regulated
by cooperative interactions of Spl and NF-Y. Journal of
Biological Chemistry 270:20978-20986.

Wu, Z., N.L.R. Bucher and S.R. Farmer. 1996. Induction of

peroxisome proliferator-activated receptor 1 during the
conversion of 3T3 fibroblasts into adipocytes is mediated by

C/EBPB, C/EBP6 and glucocorticoids. Molecular and Cellular
Biology 16:4128-4136.

Yang, L., G. Baffy, S.G. Rhee, D.R. Manning, C.A. Hansen and
J.R. Williamson. 1991. Pertussis toxin-sensitive G;
protein involvement in epidermal growth factor-induced

activation of phospholipase C-y in rat hepatocytes. Journal
of Biological Chemistry 266:22451-22458.

Yang, L., A.M. Camoratto, G. Baffy, S. Raj, D.R. Manning and
J.R. Williamson. 1993. Epidermal growth factor-mediated
signaling of Gi-protein to activation of phospholipases in
rat-cultured hepatocytes. Journal of Biological Chemistry
268:3739-3746.

Yenush, L. and M.F. White. 1997. The IRS-signaling system
during insulin and cytokine action. BioEssays 19:491-500.

Yu, K., W. Bayona, C.B. Kallen, H.P. Harding, C.P. Ravera,
G. McMahon, M. Brown and M.A. Lazar. 1995. Differential
activation of peroxisome proliferator-activated receptors by
eicosanoids. Journal of Biological Chemistry 270:23975-
23983.

219

 

Zemel, M.B. 1995. Insulin resistance, obesity and
hypertension: An overview. Journal of Nutrition 125:17158-

1717S.

Zhang, B., J. Berger, G. Zhou, A. Elbrecht, S. Biswas, S.
White-Carrington, D. Szalkowski and D.E. Moller. 1996.
Insulin- and mitogen-activated protein kinase-mediated
phosphorylation and activation of peroxisome and

proliferator-activated receptor 7. Journal of Biological
Chemistry 271:31771-31774.

 

220

"illiillilllill“