Xenobiotic-dependent regulation of gene expression in human placental cell lines

MISSING IMAGE

Material Information

Title:
Xenobiotic-dependent regulation of gene expression in human placental cell lines
Physical Description:
x, 115 leaves : ill. ; 29 cm.
Language:
English
Creator:
Zhang, Liyan, 1963-
Publication Date:

Subjects

Subjects / Keywords:
Gene Expression Regulation   ( mesh )
Placenta -- drug effects   ( mesh )
Trophoblasts   ( mesh )
Cell Line   ( mesh )
Xenobiotics -- pharmacology   ( mesh )
Benzopyrenes -- pharmacology   ( mesh )
Tetrachlorodibenzodioxin -- analogs & derivatives   ( mesh )
Tetrachlorodibenzodioxin -- pharmacology   ( mesh )
Transforming Growth Factor alpha -- drug effects   ( mesh )
Transforming Growth Factor beta -- drug effects   ( mesh )
Epidermal Growth Factor -- drug effects   ( mesh )
Plasminogen Activator Inhibitor 2 -- drug effects   ( mesh )
Proto-Oncogene Proteins c-myc -- drug effects   ( mesh )
Receptors, Aryl Hydrocarbon   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 103-114).
Statement of Responsibility:
by Liyan Zhang.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 49016241
ocm49016241
System ID:
AA00011226:00001

Full Text











XENOBIOTIC-DEPENDENT REGULATION OF GENE EXPRESSION IN HUMAN
PLACENTAL CELL LINES













BY


LIYAN ZHANG














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY


UNIVERSITY OF FLORIDA

1996



























This dissertation is dedicated to Yukun, and also to my parents, brothers and sisters.














ACKNOWLEDGEMENTS


I would like to sincerely thank my mentor, Dr. Kathleen Shiverick, for her

guidance and patience throughout the course of the completion of this work. What I have

gained in knowledge and capability under her direction, I will carry with me into a

hopefully long-lasting career in science.

I also wish to respectfully thank the members of my dissertation committee, Dr.

Stephen Baker, Dr. Thomas Rowe, Dr. Jeffrey Harrison, Dr. Margaret James and Dr.

Rosalia Simmen, for their advice and support. I wish to also thank the past and present

members of Dr. Shiverick's laboratory, including Terry Medrano, Grantly Charles, Paul

Saunders, Dr. Scott Masten, Dr. Mary Vaccarello, Dr. Yara Smit, and Dr. Phyllis Conliffe

for their help and for providing a pleasant working atmosphere. I would also like to

express my appreciation to my fellow graduate students and all the faculty members in the

Department of Pharmacology for their help and friendship. My sincere thanks also go to

the administrative staff of the Department of Pharmacology, especially Judy Adams,

Barbara Reichert, Cookie Mundorff, Donna Desmond, and Patsy Messinger.

In addition, I would like to acknowledge the following people for providing

materials or instruments essential to the completion of this research: Dr. Thomas Rowe, Dr.

Paul Klein, Dr. William Greenlee, Dr. Oliver Hankinson, Dr. Christopher Bradfield, and

Dr. Stephen Safe.

Finally, I thank my husband, Yukun, for standing by me as I followed my dream,

and my parents and parents-in-law for their love and support.















TABLE OF CONTENTS


ACKNOWLEDGEMENTS............................................... ...................... iii

KEY TO ABBREVIATIONS.................... ................. vii

ABSTRACT ................... ................................................... ix

CHAPTER 1: INTRODUCTION...................... ............ ..................... 1

Human Placenta: A Primary Target Organ of CYP1A1 Inducers ........................ 1
The Objective of This Study................................................................2
Cultured Placental Cells: An In Vitro Model for the Study of Fetoplacental
T oxicity ................................... ................................................ 3
General Characterization of the Mechanisms of Action of TCDD and BaP..............4
Ah Receptor..................................................................... ... 4
Metabolic Activation ................................... ............ ........ 5
EGF Receptor: A Possible Marker of Placental Toxicity.................................... 5
TCDD-responsive Growth Control Genes......................................... 7
Trophoblast Growth Control Factors... ................................ ......10
TGF-a ..................................... .....................................10
TG F-pl ......................................... ....... ... ....... ..... ........... 1
c-Myc.............................................. ..... ........................12
PAI-2 .................................................................. 13

CHAPTER 2: MATERIALS AND METHODS ......................... .......................15

Materials ........................................................ ...........................15
Chemicals and Bioreagents..................................................................15
Recombinant cDNA Clones............................................................ 16
Antibodies and ELISA Kits.............................................................16
Methods.....................................................................................17
Cell Culture and Chemical Treatment.................................................. 17
EGF Binding Assay.............................................. ...................17
Western Immunoblot Analysis............................................ ............. 18
General Procedure ..................................................................18
CYP1A1 Protein......................................................................18
EGF Receptor Protein ................................................................18
c-M yc Protein ......................................................................... 19
RNA Isolation and Northern Blot Analysis........................................... 19
Nuclear Run-off Assay .................................................................20
mRNA Stabilization Assay............................................................... 21
ELISA for Secreted Proteins.............................................................21
TGF- 13 Assay.....................................................................21
hCG Assay ........................................................................22
Cell Proliferation Assay ..................................................................22


iv









MTT Assay ................................... ............................22
[3H]Thymidine Incorporation Assay ...............................................23
In Vitro Invasion Assay ..................................................................23
Protein Assay ........................................................ .......... ......24
Data Analysis ..........................................................................24

CHAPTER 3: EVIDENCE THAT THE HUMAN PLACENTAL CELL LINES
BEWO AND JEG-3 HAVE A FUNCTIONAL AH RECEPTOR SYSTEM............25
Introduction ...........................................................................25
Results .................................................................................. 26
Expression of Ah Receptor and Amt mRNA in BeWo and JEG-3 Cells .......26
Superinduction of CYPIAl mRNA and Structure-Activity Specificity .........27
Induction of CYPIAl Protein.....................................................27
D iscussion............................ .................... ....................28

CHAPTER 4: EFFECTS OF TCDD AND BAP ON EGF RECEPTOR
EXPRESSION...............................................................................34
Introduction ...............................................................................34
Results.................................................... ..............................35
Effects of TCDD and BaP on Specific Binding of 125I-EGF to BeWo
and JEG-3 Cells ................................................................35
Effects of TCDD and BaP on EGF Receptor Protein............................36
Relationship between CYPIAl Induction and EGF Receptor Changes.........37
Effects of Actinomycin D and Cychloheximide on BaP-Mediated
Changes in EGF Receptors .................................................38
Effects of TCDD and BaP on Steady State EGF Receptor mRNA
Levels.............................................. ............................ 38
Discussion........................... .........................................39

CHAPTER 5: EFFECTS OF TCDD AND BAP ON TGF-a, TGF-j1, C-MYC
AND PAI-2 GENE EXPRESSION.........................................................56
Introduction ............................................................... ................56
R results .......................................................................................57
Effects of TCDD and BaP on the Steady State mRNA Levels for TGF-
a, TGF-pl, PAI-2 and c-myc ..................................................57
Effects of TCDD and BaP on the Rate of TGF-pl and c-myc Gene
Transcription.................................................................58
Effects of BaP on the Stability of TGF-1p and c-myc mRNA....................59
Effects of TCDD and BaP on TGF-pl and c-Myc Protein Levels ............. 60
D iscussion................................................ ..............................60

CHAPTER 6: EVALUATION OF THE EFFECTS OF TCDD AND BAP ON
CELLULAR GROWTH AND ENDOCRINE RESPONSES OF BEWO AND
JE G -3 C EL L S................................................................................. 77
Introduction ............................................................................77
R esults............................................. ...................................... 77
Effects of TCDD and BaP on Cell Proliferation as Measured by the
MTT Conversion Assay .....................................................77
Effects of TCDD and BaP on JEG-3 Cell Proliferation as Measured by
[3H]Thymidine Incorporation and Cell Number Changes....................78
Effects of TCDD and BaP on BeWo Cell Proliferation as Measured by
[3H]Thymidine Incorporation.................................................79









Effects of TCDD and BaP on Secretion of the Hormone hCG .................79
Differential Effects of TCDD and BaP on JEG-3 Cell Invasiveness.............80
D iscussion.................................. ........ ............................... 80

CHAPTER 7: CONCLUSIONS AND FUTURE DIRECTIONS...........................95

LIST OF REFERENCES ........................... ..... ................... 103

BIOGRAPHICAL SKETCH ........................................ .................... 115












KEY TO ABBREVIATIONS


AD actinomycin D

Ah aryl hydrocarbon

AhR aryl hydrocarbon receptor

Arnt aryl hydrocarbon receptor nuclear translocator

a-NF a-naphthoflavone

BaP benzo(a)pyrene

BPDE benzo(a)pyrene-7,8-diol-9,10-epoxide

BSA bovine serum albumin

c-myc cellular myc gene or ribonucleic acid

c-Myc c-myc oncoprotein

cDNA complementary deoxyribonucleic acid

CHX cycloheximide

CYPIA1 cytochrome P 450 1A1

DMSO dimethyl sulfoxide

DNA deoxyribonucleic acid

DRE/XRE dioxin responsive element/xenobiotic responsive element

DTT dithiothreitol

ECM extracellular matrix

EGF epidermal growth factor

EGFR epidermal growth factor receptor

ELISA enzyme-linked immunosorbent assay

FBS fetal bovine serum

HCB 2,2',4,4',5,5'-hexachlorobiphenyl

hCG human chorionic gonadotropin










hr hours

IGF-II insulin-like growth factor I

IGFBP insulin-like growth factor binding protein

kb kilobases

kd dissociation constant

kDa kilodaltons

MMP matrix metalloproteinase

MNF 3'-methoxy-4'-nitroflavone

mRNA messenger ribonucleic acid

PAI plasminogen activator inhibitor

PBS phosphate-buffered saline

PCB polychlorinated biphenyl

PCDF polychlorinated dibenzofurans

PDGF platelet-derived growth factor

PMSF phenyl-methyl sulfonyl fluoride

RNA ribonucleic acid

SE standard error

SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis

TCB 3,3',4,4'-tetrachlorobiphenyl

TCDD 2,3,7,8-tetrachlorodibenzo-p-dioxin

TGF transforming growth factor

TIMP tissue inhibitor of metalloproteinases

t-PA tissue plasminogen activator

u-PA urokinase plasminogen activator














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


XENOBIOTIC-DEPENDENT REGULATION OF GENE EXPRESSION IN HUMAN
PLACENTAL CELL LINES

By

Liyan Zhang

December 1996



Chairman: Kathleen T. Shiverick
Major Department: Pharmacology and Therapeutics


This study evaluated the usefulness of human placental trophoblastic

choriocarcinoma cell lines BeWo and JEG-3 to study the effects of the prototype

environmental chemicals on the expression of important trophoblast growth regulatory

genes. The environmental chemicals studied are the aryl hydrocarbon (Ah) receptor

agonists, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and benzo(a)pyrene (BaP). The

placental genes under study include the growth factors, transforming growth factor (TGF)-

a, TGF-pl and the epidermal growth factor (EGF) receptor, the proteinase inhibitor,

plasminogen activator inhibitor (PAI)-2, and the protooncogene, c-myc. This study first

demonstrated that both cell lines possess a functional Ah receptor system as evidenced by

CYP1Al induction following TCDD and BaP exposure. Secondly, induction of CYP1Al

does not appear to be directly linked with loss of EGF receptor. BaP decreased EGF

receptor binding and protein in both cell lines, without affecting EGF receptor mRNA

level. TCDD, however, was not found to alter the EGF receptor expression in either cell

line. Thirdly, significant dysregulation of TGF-PI and c-myc gene expression was









observed with BaP but not TCDD in JEG-3 cells. BeWo cells primarily showed effects of

both chemicals on hCG secretion, BaP on TGF-31 and TCDD on TGF-a expression.

Finally, BaP has been shown to inhibit choriocarcinoma cell proliferation and invasion,

whereas TCDD increased cell invasion without affecting cell proliferation.

In summary, BaP-mediated changes in EGF receptor, TGF-31 and c-myc

expression were correlated with altered trophoblast proliferation and invasiveness. These

altered processes may underlie mechanisms by which xenobiotics such as those found in

cigarette smoke disrupt placental function and lead to fetal growth retardation. Data also

indicate that TCDD and BaP produce placental toxicity through different mechanisms. In

addition, this study supports the feasibility of using the BeWo and JEG-3 cell lines to

investigate biomarkers and mechanisms of placental toxicity.














CHAPTER 1
INTRODUCTION


Human Placenta: A Primary Target Organ of CYPIAl Inducers

Maternal cigarette smoking during pregnancy has been associated with increased

incidence of spontaneous abortions, congenital malformations and fetal growth retardation

(Pirani, 1978; Naeye, 1980; Sachs, 1989; Alderman et al., 1994; Gabriel, et al., 1994;

Handler et al., 1994). Changes in placental epidermal growth factor (EGF) receptors

(Wang et al., 1988; Gabriel, et al., 1994), amino acid uptake (Rowell, 1981) and endocrine

function (Mochizuki et al., 1984) have been reported in association with maternal cigarette

smoking. Cigarette smoke contains a number of biologically active compounds (Hoffmann

et al., 1978), but it is not clear which of those compounds are responsible for fetoplacental

toxicity. It has been reported that maternal cigarette smoking induced placental cytochrome

P450 1Al (CYPIAl) activity (Sesardic et al., 1990), produced smoking-related covalent

DNA adducts in human placenta (Everson et al., 1986), and increased sister chromatid

exchange frequency in cytotrophoblasts (Shulman et al., 1991). In this regard, a major

polycyclic carcinogen present in the cigarette smoke is benzo(a)pyrene (BaP), an inducer of

CYP1A1 and a DNA-damaging agent (Hoffmann et al., 1978; Kaiserman and Rickert,

1992). Study of primary human placental cells found that exposure to BaP directly resulted

in an alteration in amino acid uptake (Guyda, 1991) and a loss of EGF receptor binding and

autophosphorylation (Guyda et al., 1990). The selective alterations in EGF, but not

insulin, receptors in the placentas of women who smoked have been linked with fetal

growth retardation in these pregnancies (Gabriel et al., 1994). Thus, smoking-related









fetotoxicity may involve direct effects of cigarette smoke on the placenta which may be

mediated by BaP and related compounds.

As observed with infants of cigarette smoking mothers, low birth weights were also

observed in infants born to mothers who consumed polychlorinated biphenyl

(PCB)/dibenzofuran (PCDF)-contaminated rice oil or fish during pregnancy (Sunahara et

al., 1987; Lindstrom et al., 1995). Further study has found that decreased birth weights in

infants following in utero exposure to PCBs/PCDFs were associated with decreased

functional EGF receptors in the placentas (Sunahara et al., 1987). CYP1Al activity was

also found to be markedly induced in PCB/PCDF-exposed placentas (Wong et al., 1986).

These studies clearly indicate that human placental CYP1A1 activity is highly inducible by

environmental xenobiotic exposure and the human placenta is a primary target organ of

PCBs/PCDFs, a large group of halogenated hydrocarbons represented by its prototype

2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Furthermore, animal studies have shown

that gestational exposure to TCDD or PCB mixtures can result in fetotoxicity and

teratogenicity without overt toxic effects on the mother (McNulty, 1985; Golub et al.,

1991; Rier et al., 1993), suggesting that the potential mechanisms of fetotoxicity of the

CYP1A1 inducers may involve direct effects of these chemicals on the placenta.


The Objective of This Study


This research was undertaken to further investigate molecular mechanisms by

which the prototype CYPIA1 inducers, TCDD and BaP, have disruptive effects on

placental function. It is well known that the induction of CYPIAl is via the cytosolic aryl

hydrocarbon (Ah) receptor, which is a ligand-dependent transcription factor. Substantial

evidence indicates that TCDD and related Ah receptor ligands act as endocrine disruptors

and growth modulators by persistently altering the expression of a battery of genes.

However, little is known regarding the Ah receptor regulated genes in human placental

cells. Our hypothesis is that TCDD and BaP alter placental function by disrupting local









autocrine and paracrine networks which are recognized to be important for the regulation of

trophoblast proliferation, differentiation and invasiveness. The objective of this project

was to identify the cellular and molecular processes altered by TCDD and BaP in human

placental cells in an effort to develop an in vitro model for the study of placental toxicity of

environmental chemicals.



Cultured Placental Cells: An In Vitro Model for the Study of Fetoplacental Toxicity


Study of the human fetoplacental toxicity of environmental chemicals has been

difficult because of a lack of sufficient quantities of early pregnancy tissue, as well as the

fact that placental trophoblasts in primary culture do not actively proliferate (Lewis et al.,

1994). It is essential, therefore, to explore the use of alternate models of proliferative

human trophoblasts.

The human trophoblastic cell lines BeWo and JEG-3, derived from gestational

choriocarcinoma have been shown to retain many characteristics of normal trophoblasts and

display a number of the functional features of differentiated syncytiotrophoblasts (Pattillo et

al., 1968; Kohler and Bridson, 1971; Hochberg et al., 1991). Both cell lines have been

used as a model for in vitro studies of regulation of trophoblastic gene expression and

placental endocrine function (Kato and Braunstein, 1991; Chiang and Main, 1994; Matsuo

and Strauss III, 1994), although their validity as a placental model has not been

established. Our interest is in the feasibility of using cultured placental cell lines as a

placental system to investigate the potential toxicity of environmental chemicals. We have

chosen BeWo and JEG-3 cell lines as a model system to investigate the events which occur

during Ah receptor-mediated CYP1A1 induction in human placental trophoblasts.

The next four sections of this Chapter present a brief review of the current

understanding of the mechanisms of action of TCDD and BaP, Ah-responsive genes and

trophoblast growth control networks.









General Characterization of the Mechanisms of Action of TCDD and BaP


Ah Receptor


It is generally believed that the biochemical and toxic effects of TCDD and related

halogenated aromatics are mediated by the Ah receptor, a cytosolic protein and a ligand-

activated transcription factor (Whitlock Jr., 1993; Hankinson 1995; Ferandez-Salguero et

al., 1996). The Ah receptor is distinct from the steroid receptor family in two notable

ways. First, there is no known endogenous physiological ligand for the Ah receptor;

second, the Ah receptor is a basic-helix-loop-helix protein, whereas the steroid receptors

are zinc-finger proteins (Whitlock Jr., 1993; Dolwick et al., 1993; Hankinson 1995). The

known ligands for the Ah receptor are foreign planar aromatic compounds, including

polycyclic aromatic hydrocarbons represented by BaP, as well as halogenated aromatic

compounds whose prototype is TCDD. A strong correlation exists between the binding

affinity of various ligands to the Ah receptor, the potency of these chemicals as a CYP1A1

inducer, and their ability to produce various toxic responses (Safe, 1990). The current

view of the mechanism by which the Ah receptor regulates gene transcription is that the

binding of specific ligands to this receptor induces a biochemical and conformational

change which converts the protein into an active DNA binding form (Whitlock Jr., 1993;

Hankinson 1995).

Northern blot analysis indicates that the human Ah receptor mRNA is expressed in

all tissues examined, which include liver, lung, heart, kidney, brain, skeletal muscle and

placenta (Dolwick et al., 1993). It is worth noting that placental tissue is one of the richest

known sources of Ah receptor in humans (Manchester et al., 1987; Dolwick et al., 1993).

Importantly, the high levels of Ah receptor in human placenta may account for the great

sensitivity of this tissue to receptor agonists, such as those found in cigarette smoke and

PCB-contaminated rice oil (Wong et al., 1986; Lucier et al 1988). A recent study

demonstrates that the extracts of urban air and vehicle exhaust particulates contain









significant amounts of Ah receptor binding activity, the majority of which is attributed to

unidentified polycyclic aromatic hydrocarbons (Mason, 1994). In light of the high level of

Ah receptor expression, placenta is regarded as a very sensitive tissue for investigation of

the potential human risk of exposure to environmental polycyclic aromatic pollutants.


Metabolic Activation


Although both BaP and TCDD bind to the Ah receptor and induce CYP1A1

activity, only BaP is metabolized by CYPIA1 into a series of reactive metabolites. The

latter can cause teratogenicity (Legraverend et al., 1984), immunotoxicity (Kong et al.,

1994) and hematotoxicity (Holladay and Smith, 1994; Zhu et al., 1995), as well as DNA

damage resulting in mutagenicity and carcinogenicity (Levin et al, 1978). In contrast,

TCDD is resistant to metabolism and all, or at least most, of the TCDD-induced biologic

effects are mediated by TCDD-Ah receptor complex via regulation of a battery of structural

genes (Greenlee et al., 1990; Huff et al., 1994). In this regard, BaP may be more toxic

than TCDD because the toxic effects of BaP can be amplified and exacerbated by its

reactive metabolites. It is postulated that the carcinogenicity of TCDD may result from

alterations in the DNA-damaging potential of some endogenous compounds, as well as

disregulation of cellular differentiation and/or division by an initial interaction of TCDD

with the Ah receptor (Huff et al., 1994). Therefore, our comparative study of BaP- and

TCDD-mediated effects on related endpoints in the same cell line will further our

knowledge of the importance of initial interaction of ligands with the Ah receptor and

metabolic activation in the mechanisms of action of BaP- and TCDD-type carcinogens.


EGF Receptor: A Possible Marker of Placental Toxicity


The EGF receptor is a well-characterized 170 kDa single polypeptide

transmembrane glycoprotein which is detectable in a wide variety of tissues in vivo and cell

lines in culture (Adamson, 1990). Human placenta shows a high level of expression of









EGF receptor which is localized in the proliferative cytotrophoblasts in very early placenta,

and subsequently in mitotically inactive differentiated syncytiotrophoblasts as gestation

advances (Ladines-Llave et al., 1991). There is an increase in the number and binding

capacity of EGF receptors in the trophoblast with advancing normal gestation (Carson et

al., 1983; Lai and Guyda, 1984; Adamson, 1990). EGF has been shown to stimulate

trophoblast proliferation and endocrine function (Lai and Guyda, 1984; Morrish et al.,

1987; Maruo et al., 1992). Thus, evidence strongly supports a physiological role for the

EGF receptor system in normal fetoplacental growth and development throughout

pregnancy.

Study of placentas from cigarette smokers and nonsmokers found that EGF-

stimulated receptor kinase activity was markedly decreased in placental membrane proteins

from smokers (Lucier et al., 1987; Wang et al., 1988). In this regard, BaP is a potent

polycyclic carcinogen that is present in the particulate phase of cigarette smoke (Hoffmann

et al., 1978; Kaiserman and Rickert, 1992). Our study of human placental cells in primary

culture found that exposure to BaP directly resulted in a dose-dependent selective loss of

EGF receptor binding activity and autophosphorylation, which was greatest in cells from

first trimester placentas (Guyda et al., 1990). It is significant that the original observation

of selective alterations in EGF, but not insulin, receptors in the placentas of women who

smoked was recently confirmed by Gabriel et al. (1994), and further shown to be linked

with intrauterine growth retardation in these pregnancies. A separate study of women who

were exposed to polychlorinated biphenyl-contaminated rice oil found that birth weights

were decreased in infants following in utero exposure in association with decreased

placental EGF receptor tyrosine kinase activity (Lucier et al., 1987; Sunahara et al., 1987).

These and other reports (Fujita et al., 1991; Fondacci et al., 1994) provide substantial

evidence that EGF receptors are altered in placental membranes from women whose fetuses

show intrauterine growth retardation. Thus alterations in EGF receptor might provide a

good biomarker of effect for fetoplacental toxic xenobiotics. In comparison with previous









studies in normal trophoblasts, the present study of EGF receptors in choriocarcinoma cells

provides information on similarities and differences between normal and transformed

trophoblast cells in response to xenobiotics, as well determines the validity of the

choriocarcinoma cells as a model system for placental toxicity studies.



TCDD-responsive Growth Control Genes


At present, the best characterized response to TCDD is the transcriptional regulation

of the CYP1Al gene. It is well known that transcriptional activation involves the binding

of the liganded Ah receptor to several DNA recognition sites known as xenobiotic-

responsive elements (XREs) or dioxin-responsive elements (DREs) in upstream enhancer

regulatory regions of the CYP1Al gene (Greenlee et al., 1990; Whitlock Jr., 1993;

Hankinson, 1995). The receptor-XRE interaction increases the accessibility of the

downstream promoter and hence activates CYP1A1 transcription. Thus, the binding of the

liganded Ah receptor to the DNA recognition sites in upstream enhancers is a pivotal event

in the mechanism of TCDD action. Theoretically, any gene that contains 5'-regulatory

regions highly homologous to the XREs of the CYP1A1 gene may be regulated by the

TCDD-Ah receptor complex. Dependent upon tissue or gene specific factors, the

expression of XRE-containing genes can potentially be induced or suppressed by the

liganded Ah receptor complex. Increasing evidence indicates that TCDD alters the

expression of a number of genes important in cell growth and differentiation, including the

EGF receptor, transforming growth factor (TGF)-a, EGF, TGF-Ps, tumor necrosis factor-

a (TNF-a), interleukin- 1j (IL-lI ), plasminogen activator inhibitor 2 (PAI-2), urokinase-

type plasminogen activator (u-PA), tissue plasminogen activator (t-PA), cathepsin D (an

aspartyl protease), and the estrogen receptor (Hudson et al., 1985; Gierthy et al., 1987;

Abbott and Birnbaum, 1989, 1990a and 1990b; Astroff et al., 1990; Choi et al., 1991; Lin

et al., 1991; Safe et al., 1991; Sutter et al., 1991; Gaido et al., 1992; Gaido and Maness,

1994; Sewall and Lucier, 1995; Vogel and Abel, 1995).









One of the hallmarks of the effects of TCDD on growth control genes is significant

tissue- and species-specificity. TCDD decreases EGF receptor binding activity in human

keratinocytes and mouse liver without affecting the amount of mRNA for the EGF receptor

(Hudson et al., 1985; Lin et al., 1991). In contrast, EGF receptor mRNA was reported to

be decreased by TCDD exposure in rat uterus and liver in correlation with the loss of EGF

binding (Astroff et al., 1990; Sewall et al., 1995). TCDD has also been shown to reduce

the number of EGF receptors in hepatic plasma membrane of rat, guinea pig, mouse and

hamster, with different degrees of sensitivity and potency (Madhukar et al., 1984). It

warrants note, however, that TCDD has also been reported to stimulate EGF receptor

expression and proliferation in the mouse embryonic palate and ureter epithelial cells

(Abbott and Birnbaum, 1989 and 1990a).

Studies with both normal and transformed human keratinocytes have shown that

TCDD increased the steady-state mRNA levels for TGF-a, IL-1 and PAI-2 (Choi et al.,

1991; Sutter et al., 1991; Gaido et al., 1992; Gaido and Maness, 1994), and suppressed

the mRNA expression for TGF-P2 (Gaido et al., 1992; Gaido and Maness, 1994),

whereas TGF-a and PAI-2 mRNAs were not altered by TCDD in rat liver (Vanden Heuvel

et al., 1994). Moreover, one study demonstrated that the induction of TGF-a expression

in transformed SCC-12F keratinocytes by TCDD occurs post-transcriptionally by increased

mRNA stabilization, while TGF-p2 expression is reduced due to a decrease in the rate of

TGF-32 gene transcription (Gaido et al., 1992). There are alterations in protein levels of

TGF-a, TGF-p2, IL-11 and PAI-2 which correlate with altered mRNA levels following

TCDD exposure in nontransformed human keratinocytes (Gaido and Maness, 1994). u-PA

protein levels were induced in these cells, however, the mRNA level for u-PA was not

altered following TCDD treatment (Gaido and Maness, 1994).

A recent study with the human breast cancer cell line MCF-7 has shown that the

mRNA levels of TGF-a, TGF-P3, IL-I and TNF-a were increased by TCDD, whereas

the TGF-pl and TGF-P2 mRNAs were unchanged (Vogel and Abel, 1995). This study









further reported that the enhanced secretion of TGF-P was accompanied by an inhibition of

cell growth by TCDD (Vogel and Abel, 1995). Furthermore, TCDD has been reported to

reduce the expression of TGF-a, EGF and TGF-P 1 in mouse palatal epithelial and

mesenchymal cells (Abbott and Birnbaum, 1990b). However, it is important to emphasize

the following: 1) TGF-a mRNA is not affected in mouse (Lin et al., 1991) and rat (Vaden

Heuvel et al., 1994) liver, and increased in human keratinocytes (Choi et al., 1991; Gaido

et al., 1992; Gaido and Maness, 1994) and breast cancer cells (Vogel and Abel, 1995) by

TCDD; 2) TGF-l 1 mRNA is not changed by TCDD in either human keratinocytes (Gaido

et al., 1992) or breast cancer cells (Vogel and Abel, 1995).

Studies in MCF-7 mammary tumor cells indicate that TCDD reduces estrogen-

induced secretion of t-PA and cathepsin D (Gierthy et al., 1987; Safe et al., 1991). A

further study showed that the expression of hepatic and uterine estrogen receptor protein

was decreased by TCDD in mice (Devito et al., 1992). It has also been observed that

TCDD caused a decrease in nuclear estrogen receptor levels in MCF-7 and wild-type Ah-

responsive Hepa lclc7 cells, but was inactive in Ah non-responsive mutant Hepa lclc7

cells (Safe et al., 1991). In this regard, the human estrogen receptor gene has been shown

to contain a DNA sequence that binds activated mouse and human Ah receptors, providing

a possible mechanism for transcriptional regulation of the estrogen receptor by TCDD

(White and Gasiewiz, 1993).

Altogether, patterns of growth control genes may be differentially altered by TCDD

in separate species, tissues and/or cells. Modulation of gene expression by TCDD results

in differential changes in either mRNA or protein level for specific genes and can occur by

multiple mechanisms. These mechanisms may include transcriptional control,

posttranscriptional modulation of mRNA, translational control and posttranslational

modulation of protein. Our effort to identify the important TCDD-responsive growth

regulatory genes in placental cells serves to add more knowledge to how growth control









genes respond to TCDD differently in various tissues, and also increases our understanding

of tissue-specific responsiveness of TCDD.



Trophoblast Growth Control Factors


Human trophoblast proliferation, differentiation and invasiveness is regulated by

local autocrine/paracrine networks. The major networks involve 1) the growth factors,

TGF-Ps, TGF-t/EGF and their common receptor, the EGF receptor (Frolik et al., 1983;

Filla et al., 1993; Amemiya et al., 1994), 2) the cytokine, IL-I (Kauma et al., 1990), 3)

the protooncogenes, e. g. c-myc and c-sis (Goustin et al., 1985), and 4) the proteases and

their inhibitors such as u-PA and its inhibitor PAI-2 (Feinberg et al., 1989; Hofmann et al.,

1994).


TGF-a


TGF-a is widely expressed in a variety of normal adult cells, embryos and fetuses

(Lee et al., 1995). The mature 50 amino acid form of human TGF-a is synthesized as an

internal part of a transmembrane 160 amino acid precursor (Derynck et al., 1984; Wong et

al., 1989; Anklesaria et al., 1990; Lee et al., 1995). Not only the mature secreted form but

also the membrane-anchored precursor binds to the EGF receptor, leading to signal

transduction and cellular response (Derynck et al., 1984; Lee et al., 1995). Notably,

TGF-a is found in maternal decidual cells, villous and extravillous cytotrophoblasts, and

villous syncytiotrophoblasts throughout pregnancy (Filla et al., 1993; Lysiak et al., 1993;

Horowitz et al., 1993), which is colocalized with its receptor, the EGF receptor (Ladines-

Llave et al., 1991; Filla et al., 1993). The proliferation of first trimester human trophoblast

cells in culture is stimulated by exogenous TGF-a (Lysiak et al., 1993 and 1994) In a

study with choriocarcinoma cell lines JEG-3 and Jar, Lewintre et al. (1994) found that

TGF-a elevates the levels of mRNA, protein and catalytic activity of 17p-hydroxysteroid









dehydrogenase type 1 that catalyzes the reversible interconversion of estrone and estradiol.

This study suggests that TGF-a may play a role in estrogen production in the human

placenta.

In light of the above evidence, TGF-a appears to have an autocrine,

paracrine/juxtacrine role in trophoblast growth and function. Together with evidence that

TGF-a is differentially regulated by TCDD in human keratinocytes (Choi et al., 1991;

Gaido et al., 1992, Gaido and Maness, 1994) and breast cancer cells (Vogel and Abel,

1995), mouse embryonic palatal cells (Abbott and Bimbaum, 1990b) and liver (Lin et al,

1991), and rat liver (Vanden Heuvel et al., 1994), it is of common interest to both the

developmental biologist and the toxicologist as to whether the expression of the TGF-a

gene in human placental cells is altered by xenobiotics.


TGF-pl


One of the initial sources for purification of TGF-P to homogeneity is the human

placenta (Frolik et al., 1983). All tissues found at the maternal-fetal interface, including

first-trimester decidua, placenta, and placental membrane, contain TGF-p and express

TGF-p1 mRNA (Kauma et al., 1990). In this regard, TGF-P immunoreactive protein is

localized in the cytoplasm of villous syncytiotrophoblast and extravillous trophoblast cells

throughout gestation and TGF- 31 mRNA is expressed in both syncytiotrophoblasts and

cytotrophoblasts (Lysiak et al., 1995), suggesting that trophoblast cells themselves can

regulate their own invasiveness in an autocrine manner. TGF-P derived from decidua and

trophoblasts is the prime mediator in the control of invasion by first trimester trophoblasts

(Lala and Graham, 1990; Graham and Lala, 1991). A mechanism by which TGF-pl acts

to control invasion is through reduction in collagenase type IV activity which parallels

increased expression of tissue inhibitor of metalloproteinases in trophoblasts (Lala and

Graham, 1990; Graham et al., 1994). A second mechanism by which TGF-pl blocks









trophoblast invasion is by inhibition of proliferation and enhanced differentiation of human

trophoblast cells into noninvasive syncytiotrophoblasts (Graham et al., 1992).

In view of the major regulatory roles of TGF-l1 in trophoblast proliferation,

differentiation and invasiveness, it is meaningful to explore whether xenobiotics disrupt

TGF-l1 expression and subsequently alter placental autocrine/paracrine growth control

networks.


c-Mvc


c-Myc is a well known nuclear oncoprotein, a sequence-specific transcription

factor, which regulates a variety of genes important in normal cellular proliferation and

differentiation processes (Vastrik et al., 1994). During human placental development, the

pattern of c-Myc expression is closely linked to the highly proliferative and invasive

phenotype of cytotrophoblasts (Goustin et al., 1985; Maruo and Mochizuki, 1987). The

nonproliferative differentiated syncytiotrophoblasts have not been found to display

detectable levels of c-Myc expression (Ohlsson, 1989). Cultured first trimester

trophoblasts respond to PDGF with elevated levels of c-myc mRNA and protein

expression, accompanied by an activation of DNA synthesis (Goustin et al., 1985). Thus,

studies indicate that c-Myc plays a role in normal trophoblast proliferation and placental

development.

Not surprisingly, c-myc mRNA is constitutively highly expressed in

choriocarcinoma cells (Nachtigal et al., 1992; Arbiser et al., 1993). Its expression can be

suppressed by the chemotherapeutic drug methotrexate (MTX) in both BeWo and JEG-3

cells (Nachtigal et al., 1992; Arbiser et al., 1993). In this regard, MTX has been shown to

be able to induce these choriocarcinoma cells to change their usual cytotrophoblastic

phenotype and acquire morphologic and functional characteristics typical of intermediate

trophoblasts (Taylor et al., 1991; Nachtigal et al., 1992). Altogether, these results reveal

an inverse relationship between c-myc mRNA expression and trophoblast differentiation.









Further studies of the effects of TCDD and BaP on c-myc gene expression and

choriocarcinoma cell proliferation may increase our understanding of interactions between

protooncogene expression and trophoblast growth.


PAI-2

Mammalian embryonic development and growth require implantation of the

blastocyst into the uterus. During hemochorial placentation, characteristic of humans and

rodents, embryonic trophoblast cells invade through the uterine epithelium and deep into

the maternal stroma. Invasion of trophoblasts requires cell surface-associated proteolysis

that is potent but properly controlled (Lala and Graham, 1990). Of interest, PAI-2, the

primary inhibitor of u-PA, has been suggested to play a role in the controlled invasion of

the maternal decidua by trophoblasts during human implantation (Feinberg et al., 1989;

Hofmann et al., 1994). In particular, PAI-2 has been immunohistochemically colocalized

with u-PA and PAI-1 in cytotrophoblasts, intermediate trophoblasts and

syncytiotrophoblasts at the maternal-fetal interface in early human implantation sites

(Hofmann et al., 1994), with PAI-2 being the predominant PAI in villous

syncytiotrophoblasts in term human implantation sites (Feinberg et al., 1989). Plasma

PAI-2 levels are undetectable in nonpregnant women, but increase progressively in normal

pregnancy, and then decrease dramatically soon after delivery (Kruithof et al., 1987), likely

reflecting the completion of the placental function. Decreased plasma levels of PAI-2 have

been found to be associated with intrauterine growth retardation, suggesting an impaired

placental and fetal growth (Boer et al., 1988; Estelles et al., 1991; Lindoff and Astedt,

1994). Thus, evidence has shown that PAI-2 is an important regulatory protein of normal

implantation and a possible marker of placental growth.

Interestingly, characterization of the promoter 5'-regulatory region of PAI-2 gene

(Kruithof and Cousin, 1988) has revealed a XRE core sequence (5'-TNGCGTG-3') that

provides the recognition motif for the Ah receptor (Sutter et al., 1991). In this regard,






14


TCDD exposure has been shown to stimulate the expression of PAI-2 mRNA in human

keratinocytes, primary hepatocytes, monocytic cells, and hepatoma and breast cancer cells

(Sutter et al., 1991; Gaido and Maness, 1994; Gohl et al, 1996; Dohr et al, 1995).

Therefore, it is of great interest to determine whether the expression of PAI-2, a possible

placental function marker, is regulated by TCDD or BaP in human placental cells.














CHAPTER 2
MATERIALS AND METHODS

Materials


Chemicals and Bioreagents

BaP, cycloheximide and actinomycin D were obtained from the Sigma Chemical

Co. (St. Louis, MO), and TCDD from Midwest Research Institute (Kansas City, MO)

through the National Cancer Institute Chemical Carcinogen Reference Repository. a-

naphthoflavone (a-NF) was from Eastman Kodak Co. (Rochester, NY), 3,3',4,4'-

tetrachlorobiphenyl (TCB) from RFR, Corp. (Hope, RI), and 2,2',4,4',5,5'-

hexachlorobiphenyl (HCB) from Ultra Scientific (North Kingstom, RI). 3'-methoxy-4'-

nitroflavone (MNF) was kindly provided by Dr. Stephen Safe (Texas A&M University,

College Station, TX). Recombinant human EGF was purchased from Genzyme

(Cambridge, MA). 125I-protein A, [- 32P]dCTP and [a-32P]UTP were obtained from

ICN Pharmaceuticals Inc (Costa Mesa, CA), and 125I-EGF and [3H-methyl] thymidine

from Amersham Life Sciences (Arlington Heights, IL). The prime-It II random primer

labeling kit and NucTrap probe purification columns were obtained from Stratagene (La

Jolla, CA). CellTiter 96M non-radioactive cell proliferation assay kit was obtained from

Promega (Madison, WI), and the Fisher Diagnostics LeukoStat stain kit was purchased

from Fisher Scientific (Pittsburgh, PA). The FastTrack mRNA isolation kit was from

Invitrogen Corporation (San Diego, CA). Oligo(dT) cellulose, cell culture media and

antibiotics were from Gibco/BRL (Grand Island, NY), and fetal bovine serum from

Hyclone Laboratories (Logan, UT). All other chemicals were reagent or molecular biology

grade and were obtained from standard commercial sources.









Recombinant cDNA Clones

Plasmids containing DNA for EGF receptor (pE7), TGF-a (phTGFl-10-3350),

TGF-l1 (phTGFB-2), c-myc (pG1-5'-c-myc), CYP1A1 (phPl-450-3') and P-actin

(HHCI89; 65129) were obtained from the American Type Culture Collection (ATCC)

(Rockville, MD). The plasmids containing cDNA for Ah receptor (phuAhR, Dolwick et

al., 1993), Art (pBM5/Neo-MI-l, Hoffman et al., 1991) and PAI-2 (clone 18, Sutter et

al., 1991) were kindly provided by Dr. Christopher Bradfield (Northwestern University,

Chicago, IL), Dr. Oliver Hankinson (University of California at Los Angeles, Los

Angeles, CA) and Dr. William Greenlee (University of Massachusetts, Worcester, MA),

respectively. The probes used were a 2.7 kb Smal fragment for Ah receptor, a 2.6 kb

BamHI fragment for Arnt, a 1.0 kb EcoRI fragment for CYP1A1, a 2.4 kb Clal fragment

for EGF receptor, a 3.3 kb EcoRI fragment for TGF-a, a 2.1 kb EcoRI fragment for TGF-

p1, a 2.1 kb BstZI fragment for PAI-2, a 1.6 kb Sad fragment for c-myc and a 1.1 kb
EcoRI fragment for P-actin. All the probes are human cDNAs, except the c-myc probe

which is human genomic DNA.


Antibodies and ELISA Kits

The polyclonal sheep anti-human EGF receptor and rabbit anti-human c-myc

antisera were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY), and the

polyclonal goat anti-rat CYP1A1 antiserum from Gentest (Woburn, MA). The rabbit

biotinylated anti-sheep IgG was from Vector Laboratories Inc. (Burlingame, CA), while

horseradish peroxidase labeled goat anti-rabbit IgG and rabbit anti-goat IgG were from
Bio-Rad Laboratories (Hercules, CA). The Micro-Elisa total Peta-hCG test kit was

purchased from Leinco Technologies Inc. (St. Louis, MO), and the TGF-p1 ELISA

system from Promega (Madison, WI).












Cell Cultures and Chemical Treatment

Human placental choriocarcinoma cell lines BeWo and JEG-3 were obtained from

ATCC. BeWo cells were cultured in Ham's F-12 medium supplemented with 15% FBS,

and JEG-3 cells in Eagle's minimum essential medium supplemented with 10% FBS,

respectively, in a humidified atmosphere containing 5% CO2 at 370C. All complete media
were supplemented with penicillin (100 ig/ml), streptomycin (100 g.g/ml) and

amphotericin B (2.5 gg/ml). Cells were grown to confluency and media changed every 2

to 3 days. Confluent cells were subcultured after trypsinization. Unless otherwise

indicated, all experiments were initiated when cells were at approximately 40-60%

confluence. Cells were cultured in the absence or presence of various chemicals, added

either in DMSO or ethanol with final concentration of the vehicle being 0.1%, for the

various times as indicated. The appropriate vehicle was added to control cultures. For

RNA and protein analysis, cells were cultured in duplicate or triplicate for each experiment

point. For proliferation assays, cells were cultured in triplicate or quadruplicate.

EGF Binding Assay

Cells were washed three times with PBS to remove BaP or TCDD and then

incubated with 100 pM 125I-EGF in the presence or absence of unlabeled EGF for 90 min

at room temperature or 5 hr at 40C. After careful rinsing to remove unbound radioactivity,

the cells were solubilized in 0.5 N NaOH and the total binding of 125I-EGF was

determined by gamma counting. Specific binding was expressed as the difference between

radioactivity bound in the absence (total binding) and presence (nonspecific binding) of

excess unlabeled EGF (100 nM). For Scatchard analysis, cells were incubated with

increasing concentrations of 1251-EGF (1.25 to 200 pM) for 1.5 hr at room temperature.

Nonspecific binding of ligand was measured by adding excess unlabeled EGF to cultures









for each concentration of 125I-EGF. Each point on the Scatchard plot represents specific

binding of 1251-EGF.

Western Immunoblot Analysis


General procedure. Cells were rinsed two to three times, collected by scraping with

a rubber policeman, and lysed in 1 ml PBS using three freeze-thaw cycles. The total cell

membrane fraction was obtained by centrifugation at 12,000g for 10 min at 40C and

resuspended in PBS. Alternatively, cells were scraped into lysis buffer containing 50 mM

Tris-HC1, pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCI, 1 mM

ethylene glycol-bis(P-aminoethyl ether) N,N,N',N'-tetraacetic acid (EGTA), sodium

fluoride (NaF), sodium orthovanadate (Na3VO4) and phenyl-methyl sulfonyl fluoride
(PMSF), and lgg/ml aprotinin, leupeptin and pepstatin. The lysate was transferred to a

microcentrifuge tube using a syringe fitted with a 21 gauge needle, incubated on ice for 30-

60 min, and centrifuged at 12,000g for 15 min at 40C. The supernatant liquid is the total
cell lysate. Samples of membrane or total cell lysate protein (100 gig) were then separated

by 7.5% or 10% SDS-PAGE and transferred electrophoretically to nitrocellulose filters

using 25 mM Tris, 192 mM glycine buffer at pH 8.2, with or without 20% methanol

according to the procedure of Towbin et al. (1979).

CYPIAl protein. The blotted nitrocellulose sheet was blocked for 20 min in 3%

gelatin in 20 mM Tris, containing 500 mM NaCI, pH 7.5, and then incubated with goat

anti-rat CYP1Al (1:500 dilution) or preimmune goat serum for 2 hr, followed by

peroxidase conjugated anti-goat IgG for 1 hr as described previously (Wang et al., 1988).

The immunoreactive protein was visualized by incubation with 3-amino-9-ethylcarbazole in

the presence of 0.015% hydrogen peroxide, and quantitated by optical density scanning

using a Microtek ScanMaker II scanner and NIH image program.

EGF receptor protein. The blotted nitrocellulose filters were washed in 100 mM

Tris containing 0.1% (v/v) Tween 20 and 0.9% NaCI, pH 7.5 (TTBS) for 30 min, and









then incubated sequentially with polyclonal anti-human EGF receptor antiserum (diluted to

1 gg/ml in TTBS) or preimmune sheep serum for 60 min, followed by biotinylated anti-

sheep IgG for 60 min, and Vectastain ABC reagent for 30 min. Immunoreactive bands

were visualized and quantitated as described above. Alternatively, quantitation of

immunoreactive EGF receptor protein was carried out by a modification of the method of

Gargosky et al. (1992). In brief, nitrocellulose filters were blocked with 1% (w/v) BSA in

TTBS for 18 hr at 40C, then incubated sequentially with sheep anti-EGF receptor antiserum

or preimmune sheep serum, rabbit anti-sheep IgG, and 125I-protein A. The
immunoreactive bands were quantitated by scanning cpm using a Petascope 603 blot

analyzer, after which filters were exposed to X-Omat film at -800C for 12 to 18 hr for

autoradiography.

c-Myc protein. The blotted nitrocellulose sheet was blocked for 30 min in 5% fat-

free dried milk in PBS, and then incubated with rabbit anti-human c-myc protein (0.5
gg/ml) or preimmune rabbit serum overnight at 40C, followed by peroxidase conjugated

goat anti-rabbit IgG for 1.5 hr at room temperature. The blots were washed thoroughly

with deionized water after removing the primary and secondary antibodies, respectively.

The immunoreactive protein was visualized by enhanced chemiluminescence detection

system (Amersham Life Sciences, Arlington Heights, IL) and Kodak X-Omat film,

according to the manufacturer's instructions, and the band quantitated by densitometry as

described above.


RNA Isolation and Northern Blot Analysis

Total cellular RNA was isolated from cultured cells by acid guanidinium thiocyanate

phenol-chloroform extraction according to Xie and Rothblum (1991). Poly (A)+ RNA was

prepared by the method of Celano et al (1993). Alternatively, poly (A)+ RNA was

extracted directly from the cells using FastTrack mRNA isolation kits, according to the

manufacturer's instructions (Invitrogen). For Northern blotting, 40 gg of total cellular









RNA or 10 gg of poly (A)+ RNA was denatured, fractionated in 1% agarose formaldehyde

gel and transferred to nitrocellulose or nylon membranes. The probes were labeled with
[a-32P]dCTP using a random primer DNA labeling kit. Prehybridization was carried out

in 50% formamide containing 5 X Denhardt's solution, 4 X SSC (20 X SCC is 3 M NaCI,

0.3 M sodium citrate, pH 7.0), 0.1% (w/v) SDS, 40 mM sodium phosphate, and 0.25

mg/ml yeast RNA at 420C overnight. The hybridization was performed at 420C for 20 to

40 hr in the same buffer but containing 1 X Denhardt's solution and the 32P-probes as

indicated. The filter was washed twice in 2 X SSC/0.1% SDS at room temperature for 30

min, then twice in 0.2 X SSC/0.1% SDS at 420C for 30 min, and once at 0.1 X SSC/0.1%

SDS at 650C for 15 min. Alternatively, hybridization of RNA with the indicated DNA

probes was carried out 2 to 4 hr at 680C in ExpressHyb hybridization solution as instructed

(CLONTECH Laboratories, Inc., Palo Alto, CA). The presence of specific RNA was

detected by autoradiography, after which, filters were stripped and reprobed with each of

the indicated DNA probes. Hybridization signals were quantitated by densitometry with
the each message standardized to the p-actin transcript.


Nuclear Run-off Assay

The nuclear run-off transcription assay was performed as described by Greenberg
and Ziff (1984) with slight modifications. Briefly, 100 itl nuclei (2 X 107) collected by

Nonidet P-40 lysis were added to 100 p.1 2 X reaction buffer (10 mM Tri-HCI, pH 8, 5

mM MgC12, 300 mM KC1, 5 mM dithiothreitol, 1 mM ATP, CTP and GTP), and 100 pCi
of [a-32P]UTP followed by incubation at 300C for 45 min. After degradation of the DNA

by 10 U/ml of RNase-free DNase I (Boehringer Mannheim), the nascent transcripts were

isolated by acid guanidinium thiocyanate phenol-chloroform extraction, and hybridized
with nylon blots with 2 gIg of specific c-myc, TGF-p1, CYP1A1 and P-actin cDNA

immobilized in each dot. The hybridization was processed in 2 ml of 500 mM sodium

phosphate buffer, pH 7.2, 7% SDS for 48 h at 650C. After an initial wash in 250 mM









sodium phosphate buffer containing 1% SDS, the blots were washed twice in 100 mM

sodium phosphate buffer containing 1% SDS, 15 min each at 650C. The specific

hybridization signal was detected by autoradiography and quantitated by densitometry.


mRNA Stabilization Assay

Cells were treated with 10 ViM BaP or 0.1% DMSO for 24 hr, and actinomycin D at

5 lig/ml was added to the media. At the indicated times after addition of actinomycin D,

total RNA was isolated and processed for Northern analysis, and the hybridized signal was

quantitated by densitometry as described above.


ELISA for Secreted Proteins

Conditioned media were clarified by passing through a 0.45-im filter, and protein

inhibitors were added to a final concentration of 2 j.g/ml pepstatin A, 1 jg/ml leupeptin, 2

jig/ml aprotinin and 1 mM PMSF. The conditioned media were stored at -700C or -200C

until use.

TGF-l 1 assay. The specific detection of biologically active TGF-p 1 in the media

was carried out with a TGF-I1 ELISA system (Promega), according to the manufacturer's

direction. Briefly, the latent form TGF-Pl in the media was activated by adding 1 ll of 1

N HCI to 50 il sample medium. Maxi-Sorp (high protein affinity) 96 well ELISA plates

were coated with monoclonal anti-TGF-1p overnight at 40C, blocked and incubated with

the acid-activated sample media or TGF-I 1 standard for 1.5 hr at room temperature. After

washing for 5 times with wash buffer, the plates were incubated with polyclonal anti-TGF-

1p for 2 hr at room temperature, and washed as described above, followed by incubation

with a species-specific horse radish peroxidase conjugated antibody for 2 hr at room

temperature. The unbound conjugate was removed by washing, and the specifically bound
TGF-P 1 was detected using a chromogenic substrate. The extinction at 450nm was then









recorded using an ELISA plate reader, which is directly proportional to the amount of

biologically active TGF-P I in the test sample.

hCG assay, Assay of hCG in the media was carried out with a Micro-Elisa total

peta-hCG test kit (Leinco Technologies, Inc.). Briefly, the test sample was allowed to

react simultaneously with the coated and conjugated antibodies, resulting in the hCG

molecule being sandwiched between the solid phase and enzyme-linked antibodies. After a

30 min incubation at room temperature, the sample well was washed thoroughly to remove

free enzyme-labeled antibody. An enzyme substrate-chromogen was added to the well and

incubated for 15 min at room temperature resulting in the development of a blue color. The

addition of 1.0 N H2S04 converted the color to yellow. The extinction at 450nm was then

recorded, which is directly proportional to the concentration of hCG in the sample.


Cell Proliferation Assay


MTT assay. Cell proliferation was determined using the tetrazolium (3-[4,5-

dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) MTT dye assay according to the

CellTiter 96T non-radioactive cell proliferation assay procedure (Promega). This assay is

based on the cellular conversion of the tetrazolium salt into a formazan product that is

quantitated using an ELISA plate reader. Direct comparisons between 3H-thymidine

incorporation and the MMT assay have shown less than a 5% difference between assays

for determination of growth-factor concentrations. Briefly, cells were subcultured for 20

hr in 96-well tissue culture plate at a density of 5.0 X 103/well for BeWo and 2.5 X

103/well for JEG-3, and exposed to various compounds for 48 hr in the presence or

absence of FBS, followed by a 4 hr incubation with MTT dye. After solubilization of the

formazan product, the extinction at 595nm was recorded with 690nm as reference

wavelength using an ELISA plate reader. Relative cell proliferation was determined by

comparing the extinction with that of control cells.









[3H1Thymidine incorporation assay. Cells were cultured at 5.0 X 104 cells/well in

24-well plates for 20 hr and treated with various concentration of TCDD or BaP for 5 days,
or 10 nM TCDD and 10 gM BaP for various times as indicated. Medium was changed

every 24 to 48 hr. Cells were transferred to serum-free medium 20 hr before the addition

of tritiated thymidine and pulsed with 1 jiCi/ml [3H]thymidine for the last 3 hr. After

trypsinization, cells were harvested onto glass fiber filter strips with a cell harvester, and

incorporated radioactivity was determined by liquid scintillation counting. Replicate

cultures were harvested for cell number counting. The cell number was counted using a

hemocytometer.


In Vitro Invasion Assay

The Matrigel invasion assay was performed using a two compartment Boyden

Chamber (Terranova et al., 1986). Briefly, subconfluent cell cultures were incubated with

BaP or TCDD for 48 hr. Cells were collected and resuspended in complete media to a

density of 2.0 X 105 cells/ml after trypsinization. Twenty-eight microliters of the cell

suspension, in the presence of the respective chemicals, were added to the lower

compartment of the Boyden chamber. The lower compartment was overlaid with a
matrigel-coated porous (8 pm diameter pores) polyvinyl-pyrrolidone-free polycarbonate

membrane (Nuclepore, Pleasanton, CA), and the upper compartment fastened on. The

chamber was inverted and incubated in a humidified atmosphere containing 5% C02 at

370C for 2 to 3 hr, allowing the cells to attach to the membrane. The chamber was next
placed upright and 50 1tl media without chemicals was added to the upper compartment.

After incubation for 18 hr, the membrane was removed and scraped free of uninvaded cells

on the lower side of the membrane. The cells that invaded through the membrane were

stained with LeukoStat stain set and counted under a microscope.






24


Protein Assay


Protein content of membrane suspensions and total cell lysates was determined by

the method of Bradford (1976) using BSA as a standard.


Data Analysis


All experiments were performed using duplicate to quadruplicate cultures at each

concentration of EGF, BaP, TCDD and/or time point. A one-factor ANOVA was

employed to assess dose effects followed with Fisher's protected least significant

difference (PLSD) test when a significant dose effect (p < 0.05) was detected in the

ANOVA. A two-factor ANOVA was employed to analyze the BaP dose effect, the EGF

treatment effect and the potential interactive effect on cell proliferation. An unpaired

Student's t-test was also used to analyze the data. All statistical analyses were performed

with the Macintosh StatView512+TM or MicrosoftExcel program.














CHAPTER 3
EVIDENCE THAT THE HUMAN PLACENTAL CELL LINES BEWO AND JEG-3
HAVE A FUNCTIONAL AH RECEPTOR SYSTEM



Introduction

Evidence indicates that TCDD-induced toxicities are initiated by activation of the Ah

receptor system. The Ah receptor is a ligand-activated transcription factor that functions in

partnership with the Ah receptor nuclear translocator (Arnt) protein. Upon binding of a

ligand, the Ah receptor undergoes a series of biochemical changes which include

dissociation of hsp90 proteins, translocation to the nucleus, and dimerization with the Arnt

protein. The heterodimer of the liganded-Ah receptor and Arnt proteins binds to upstream

enhancer region specific DNA sequences termed XREs/DREs, resulting in the activation of

a battery of genes, including drug metabolizing enzymes and growth regulatory factors,

which ultimately lead to the observed biologic response (Whitlock Jr., 1993; Hankinson,

1995; Safe, 1995). The response requires both the Ah receptor and the Arnt proteins.

TCDD fails to activate CYPIA1 transcription in Ah receptor-defective cells and in Arnt-

defective cells (Whitlock Jr., 1993; Hankinson, 1995). Therefore, potential success in the

exploration of TCDD-regulated genes is highly dependent on the choice of an appropriate

cell model which not only expresses both Ah receptor and Art gene products, but also

possesses a known inducible positive control gene. To date, the best characterized

response to TCDD and related Ah receptor ligands is the transcriptional activation of the

CYPIAI gene. Figure 3-1 summarizes the current understanding of the molecular

mechanism of induction of the CYP1Al gene expression.

An earlier study with JEG-3 cells observed that aryl hydrocarbon hydroxylase, an

enzyme activity associated with CYPIA1, was induced by TCDD and a benzothiazole









derivative 2-(4'-chlorophenyl)benzothiazone; however, no detectable immunoreactive

CYP1A1 protein was observed by Western analysis in the JEG-3 cells following exposure

to these chemicals (Kirenlampi et al., 1989). Thus this study was not conclusive as to

whether the CYPIAl gene in JEG-3 cells is inducible by Ah receptor agonists. Since this

report, no study has established that the induction of aryl hydrocarbon hydroxylase is truly

through the induction of CYPIAI gene expression in JEG-3 cells, nor is any information

available regarding the Ah receptor system in BeWo cells. It is, therefore, necessary to

determine whether the BeWo and JEG-3 cells have a functional Ah receptor system. The

results presented in this Chapter provide direct evidence that the BeWo and JEG-3 cells

express Ah receptor and Art mRNA and are able to form a functional Ah receptor complex

following Ah receptor agonist stimulation.



Results


Expression of Ah Receptor and Amt mRNA in BeWo and JEG-3 Cells


We initially examined the Ah receptor, Arnt and CYP1A1 steady state mRNA levels

in control, TCDD and BaP-treated cells. Northern blot analysis revealed the presence in

both BeWo and JEG-3 cells of a single Ah receptor transcript of 6.6 kb and three Arnt

transcripts of 4.2, 2.6 and 1.8 kb (Figure 3-2), as has been reported in human placental

tissues as well as in human liver, HepG2 and mouse Hepa-l cell lines (Hoffman et al.,

1991; Dolwick et al., 1993). As shown in Figure 3-2, both Ah receptor and Art mRNA

transcripts were constitutively expressed. BeWo cells had a lower steady state level of Ah

receptor mRNA than JEG-3 cells, while the Amt mRNA was abundant and comparable in

both cell lines. TCDD or BaP treatment at 48 hr had little effect on the steady state mRNA

levels for Ah receptor and Art.

Unlike the Ah receptor and Arnt mRNA, CYPIAI mRNA transcript was

undetectable in either BeWo or JEG-3 control cells by Northern blot analysis. A 3.0 kb









CYPlAl mRNA, however, was highly induced by TCDD and BaP (Figure 3-2),

indicating the ability of the BeWo and JEG-3 cells to form a functional Ah receptor

complex following exposure to the Ah receptor ligands, as previously reported in human

keratinocytes and mouse hepatoma cells (Gaido et al., 1992; Israel et al., 1985).


Superinduction of CYP1A1 mRNA and Structure-Activity Specificity

Treatment of JEG-3 cells with 10 gg/ml cycloheximide (CHX), a protein synthesis

inhibitor, for 24 h either in the presence or absence of Ah receptor ligands resulted in

substantial increases in CYP1Al mRNA (Figure 3-3). The induction of CYPlA1 mRNA

in cells treated with TCDD or BaP plus CHX was much greater than that in cells treated

with TCDD, BaP or CHX alone. Thus, a superinduction was observed in choriocarcinoma

cells treated simultaneously with the Ah receptor ligands and the protein synthesis inhibitor,

which is a well-known property of Ah receptor-mediated induction of CYP1AI mRNA in

hepatoma cells (Whitlock Jr., 1993; Hankinson, 1995; Safe, 1995).

Moreover, treatment with the PCB congener 3,3',4,4'-tetrachlorobiphenyl (TCB),

a weak Ah receptor ligand, resulted in a weak induction of CYP1A1 mRNA, while

2,2',4,4',5,5'-hexachlorobiphenyl (HCB), a non-Ah receptor ligand, was unable to induce

any CYP1AI mRNA expression in JEG-3 cells (Figure 3-3). This congener specificity is

consistent with previous observations on structure-activity relationships for binding to the

Ah receptor (Safe, 1990 and 1992). The observed superinduction and sterospecificity

support the involvement of the Ah receptor.


Induction of CYP1Al Protein

Western immunoblot analysis showed that a 55 kDa immunoreactive CYP1A1

protein was induced by TCDD and BaP in a concentration dependent manner in both BeWo

and JEG-3 cells (Figure 3-4). In addition, figure 3-4 indicates that induction of CYPIAl is

a sensitive marker for exposure to as low as 0.1 nM TCDD or 1 JiM BaP, and was









maximal at 10 nM TCDD or 10 pM BaP in both cell lines. Insofar as 100 tlg of cell protein

was applied to each lane, the intensities of the respective lanes indicate that more

immunoreactive CYP1Al protein was present in JEG-3 cells, compared with BeWo cells,

which is correlated with the higher level of CYPIAl mRNA induction in the JEG-3 cells

(Figure 3-2).





The present study for the first time demonstrated that the human placental

trophoblastic choriocarcinoma cell lines BeWo and JEG-3 expressed both Ah receptor and

Arnt mRNA, and CYP1A1 mRNA was highly inducible by the two prototype Ah receptor

ligands, TCDD and BaP (Figure 3-1 and 2). The induced CYP1A1 mRNA is expressed as

CYP1Al protein was induced by TCDD and BaP (Figure 3-4). These data contrast with

the previous report that no immunoreactive CYPIA1 protein was detected by Western

analysis in JEG-3 cells following TCDD treatment (Kirenlampi et al., 1989). One possible

reason is that the antibody (Mab 1-7-1, mouse anti-rat liver P450) previously used may not

recognize the specific epitopes of the CYP1Al proteins in JEG-3 cells. In addition, the

present results indicate that CYP1Al mRNA and protein are induced by TCDD and BaP to

a greater extent in JEG-3 cells than in BeWo cells and are correlated with the higher level

expression of the Ah receptor mRNA in JEG-3 cell line.

The present finding that CYPIA1 mRNA was superinduced by simultaneous

treatment of TCDD or BaP with CHX is in agreement with the previous studies in human

keratinocytes (Gaido et al., 1992), human breast cancer cells MCF-7 and MDA-MB-231

(Arellano et al., 1993), Hepa lclc7 mouse hepatoma cells (Israel et al., 1985), and quail

aortic smooth muscle cells (Ou and Ramos, 1995). The mechanism of the superinduction

is not fully understood. It has been shown that the CHX-mediated superinduction of

CYP1AI required a functional Ah receptor complex (Israel et al., 1985) and involved

multiple DNA-binding factors that interact with the XREs/DREs (Saatcioglu et al., 1990).









It is now generally agreed that CHX inhibits the synthesis of a labile dominant repressor

which competes with the liganded Ah receptor complex and superinduction ensues

(Whitlock Jr., 1993; Hankinson, 1995).

The finding that the CYP1Al mRNA in JEG-3 cells was slightly induced by the

PCB congener TCB is consistent with previous reports that TCB is 100-times less potent

than TCDD in binding to the Ah receptor and inducing CYP1A1 expression. In contrast,

the observation that HCB congener does not induce CYP1A1 mRNA is consistent with

previous reports that HCB does not bind to the Ah receptor or induce CYP1A1 expression

(Safe, 1990 and 1992). The sterospecificity further supports the role of the Ah receptor in

the induction of CYP1A1 mRNA.

In conclusion, the placental cell lines BeWo and JEG-3 possess a functional Ah

receptor system, and respond to TCDD and BaP directly with an induction of CYPIAl

mRNA and protein expression. The induction of the CYP1A1 gene expression exhibits

structure-activity specificities in response to PCB congeners with different binding

affinities to the Ah receptor, and superinduction occurs following simultaneous treatment

with TCDD or BaP plus CHX. These characteristics are consistent with well-documented

responses of other cells to the Ah receptor ligands (Israel et al., 1985; Safe, 1990; Sutter et

al., 1991; Gaido et al., 1992; Arellano et al., 1993; Whitlock Jr., 1993; Hankinson, 1995;

Ou and Ramos, 1995). Thus, data indicate that the BeWo and JEG-3 cells are suitable for

the investigation of the events which occur during Ah receptor-mediated CYP1A1 induction

in human placental trophoblastic cells.












0 09










34~ 0 Q,
.,,M
1R.1















00
mc 0
*g| i:









0 Z,
m
bo
|1|P














0
u a 0
B c, k^u


siJ^









eo 00.
0 '" 0
s1l ll
8) P S'
So S

$^.^
11*1
ilit!
Bo ^^
i111.!8'














BeWo JEG-3




SE
E e S
bj a a a a .coe a a:
S a H aa 8s ^ ^ a


AhR b )
(6.6 kb) 4s


Arnt
(4.2 kb)


(2.6 kb)
(1.8 kb)


U-. A A


CYP1A1
(3.0 kb)

I ,


4 A A AAAA&'Aa


P-actin
(2.0 kb)


Figure 3-2. Northern blot analysis of Ah receptor, Arnt and CYP1Al mRNA. Cells were
incubated with or without TCDD and BaP for 48 hr. Poly(A)+RNA, 10 rg, was
denatured, blotted, and hybridized with 32p-labeled cDNA probes as described in Materials
and Methods.


~'-- -------- I


i -Ak-


1


;I;i 1 I- -1


























F3-actin
(2.0 kb) -

B)

"
Se I

CYPIA1
(3.0 kb)
P-actin

(2.0 kb)





Figure 3-3. Northern analysis of CYP1A1 mRNA. A) Superinduction of CYP1A1
mRNA in the presence of CHX; B) Induction of CYP1Al mRNA by two PCB congeners,
TCB, a weak Ah receptor agonist, and HCB, a non-Ah receptor agonist. Cells were
treated with 10 nM TCDD, 10 1M BaP, 10 pg/ml CHX, TCDD plus CHX, and BaP plus
CHX for 24 hr, or 10 M TCB and HCB for 48 hr. Total RNA, 40 rg, was denatured,
fractionated, transferred, and hybridized sequentially with CYP1Al and p-actin probes.
fractionated, transferred, and hybridized sequentially with CYP IAlI and P-actin probes.





















BeWo JEG-3

TCDD 0 0.1 1 10 100 0 0.1 1 10 100
(nM)
CYP1AI
(55 kDa)




BeWo JEG-3
BaP
) 0 1 10 50 0 1 10 50

CYPIAI
(55 kDa) -"









Figure 3-4. Effects of TCDD and BaP on immunoreactive CYPIAl proteins. The BeWo
and JEG-3 cells were harvested following incubation with TCDD or BaP for 48 hr. Total
cell protein, 100 ug, was separated by 10% PAGE, transferred to nitrocellulose, and
immunostained with anti-CYPIAl antibody.














CHAPTER 4
EFFECTS OF TCDD AND BAP ON EGF RECEPTOR EXPRESSION



Introduction

Human placenta shows a high level of expression of EGF receptor which is
colocalized with its ligands EGF and TGF-at in trophoblasts throughout gestation (Ladines-

Llave et al., 1991; Filla et al., 1993), suggesting a physiological role for the EGF receptor

system in normal fetoplacental growth and development. Substantial evidence indicates

that EGF receptors are altered in placental membranes from women whose fetuses show

intrauterine growth retardation (Lucier et al., 1987; Sunahara et al., 1987; Wang et al.,

1988; Fujita et al., 1991; Fondacci et al., 1994; Gabriel et al., 1994). In this regard,

maternal cigarette smoking or exposure to PCBIPCDF mixtures has been associated with

decreased EGF receptor tyrosine kinase activity in the placenta, which has been further

shown to be linked with intrauterine growth retardation in these pregnancies (Lucier et al.,

1987; Gabriel et al., 1994). It has been proposed that alterations in EGF receptor may be a

good placental biomarker of exposure to toxic halogenated aromatic compounds (Lucier et

al., 1987).

Several known inducers of CYPIAl, including TCDD and BaP, have been shown

to downmodulate EGF receptor expression in various species, tissues and cell lines

(Ivanovic and Weinstein, 1982; Kirenlampi et al., 1983; Madhukar et al., 1984; Hudson et

al., 1985; Astroff et al., 1990; Guyda et al., 1990; Lin et al., 1991; Sewall et al., 1993 and

1995). The mechanism of chemical-mediated EGF receptor downmodulation is not

currently understood, neither is the mechanistic link between CYP1Al induction, EGF

receptor downmodulation, and chemical-mediated toxicity. The study presented in this









Chapter evaluated changes in EGF receptor number (Bmax), binding affinity (Kd),

internalization, and protein and mRNA content in the Ah-responsive choriocarcinoma cells

BeWo and JEG-3 following TCDD and BaP treatment. A goal is to evaluate whether there

is a dose-relationship between EGF receptor expression and CYP1Al induction.


Results



Effects of TCDD and BaP on Specific Binding of 125I-EGF to BeWo and JEG-3 cells

Scatchard analysis showed a single class of binding site with a Kd of 0.043 nM for

BeWo and 0.048 nM for JEG-3 cells (Figure 4-1), which is similar to the high affinity

binding site previously reported in human placental cells (Guyda et al., 1990) and human

placental membranes (Lucier et al., 1987). Cultures treated with 10 nM TCDD for 48 hr

did not show any changes in the Bmax and Kd values in either cell line. In contrast,
cultures treated with 10 giM BaP for 48 hr exhibited 34% and 42% decreases in the Bmax in

BeWo and JEG-3 cells, respectively, with little change in Kd values (Figure 4-1).

The next experiment evaluated whether the BaP-induced loss of EGF binding was

mediated by internalization of cell surface receptors. Internalization of the EGF receptor in

choriocarcinoma cells does not occur at 40C as evidenced by the fact that more than 90% of

the 125I-EGF bound to both BeWo and JEG-3 cells was dissociated by an acid wash

procedure which removes only surface bound ligand. 125I-EGF-specific binding to BaP-

treated cells was reduced to the same extent when binding was measured at 40C or room

temperature (Figure 4-2), while TCDD had no effect on the binding under either condition.

These data indicate, therefore, that the BaP-related decrease in EGF binding is due to an

alteration in available high affinity cell surface binding sites rather than internalization of the

ligand.








The BaP-mediated decrease in EGF binding was time-dependent. The inhibition of

EGF specific binding was observed within 6 hr following BaP treatment, which is similar

to observations in human keratinocytes (Hudson, et al., 1985), mouse embryo fibroblasts

(Ivanovic and Weinstein, 1982) and hepatoma cells (Karenlampi et al., 1983). The

inhibitory effect persisted at a constant level for at least 96 hr in BeWo cells, whereas the

EGF specific binding progressively decreased in JEG-3 cells until 96 hr, the last time point

examined (Figure 4-3). The sustained inhibitory effect of BaP on EGF binding is

comparable to that previously reported with TCDD in human keratinocyte SCC-12F cells

(Hudson et al., 1985).

Effects of TCDD and BaP on EGF Receptor Protein


The EGF receptor was detected by Western immunoblot as a 170 kDa band in the

total cell membrane fraction of both BeWo and JEG-3 cells (Figure 4-4), which was

comparable to the receptor species detected in human A-431 cells and rat liver microsomes

(data not shown). The level of immunoreactive EGF receptor protein was significantly

decreased following exposure to BaP in a concentration-dependent manner. In this regard,

EGF receptor protein was decreased from control by 20, 41 and 52% in BeWo cells, and

by 23, 58 and 53% in JEG-3 cells, following exposure to 1, 10 and 50 JIM of BaP for 48

hr, respectively. These data provide further evidence that the reduced 125I-EGF cell

surface binding was due to the loss of total cellular EGF receptor protein. In contrast,

exposure to TCDD for 48 hr did not alter the level of total cell EGF receptor

immunoreactive protein in either cell line (Figure 4-4). The level of immunoreactive EGF

receptor protein was 97, 106, 99 and 110% of control in BeWo cells, and 82, 85, 95 and

92% of control in JEG-3 cells, following exposure to 0.1, 1, 10 and 100 nM of TCDD for

48 hr, respectively.

In SCC-12F human keratinocyte cells the kinetics for the BaP- and TCDD-mediated

decrease of EGF binding were found to be different (Hudson et al., 1985), such that









inhibition of EGF binding by BaP was maximal by 24 hr, whereas TCDD treatment for 72

hr was required to produce maximal inhibition. For this reason, we treated BeWo and

JEG-3 cells with TCDD for longer time period and then determined the EGF receptor

protein levels. No significant change in the EGF receptor protein level was observed

following TCDD treatment for 24 to 96 hr in either cell line (Figure 4-5).

Relationship between CYP1A1 Induction and EGF Receptor Changes

A concentration-related decrease in the level of total cell EGF receptor

immunoreactive protein was observed in association with the induction of CYPIA1 protein

in both cell lines following BaP treatment (Figure 4-6). In contrast, exposure to TCDD for

48 hr caused no alteration in the level of EGF receptor protein in either cell line, despite the

marked induction of the CYP1A1 protein. To determine whether the induction of CYPIAl

is required for BaP to decrease the EGF receptor protein content, we then studied the
effects of the Ah receptor antagonists a-naphthoflavone (a-NF) and 3'-methoxy-4'-

nitroflavone (MNF) on the BaP-mediated loss of EGF receptor protein. Although
cotreatment of a-NF (1 and 10 gM) and TCDD partially antagonized the TCDD induced

CYPIAI protein expression in a concentration-related manner, cotreatment of JEG-3 cells
with 10 tM of a-NF and 10 [lM of BaP did not in any way inhibit the induction of

CYP1A1 by BaP (Figure 4-7). An unexpected finding was that a-NF alone at 10 jiM

shared partial agonist activity with some induction of CYPIA1, as well as partial loss of

EGF receptor protein. A similar observation has been reported with mouse Hepa-1

(Karenlampi et al., 1983) and embryo fibroblast cells (Ivanovic and Weinstein, 1982) in

which a-NF alone markedly decreased EGF binding. Thus, the partial agonist activity of

a-NF in choriocarcinoma cells makes it difficult to make a conclusion about the role of

CYPIA1 induction in BaP-mediated loss of EGF receptor.

A second Ah receptor antagonist, MNF, was also evaluated. MNF alone at 1 and
10 ItM did not induce CYP1A1 mRNA in JEG-3 cells (Figure 4-8), and TCDD-induced








CYPIAl mRNA expression was completely abolished by cotreatment with either 1 or 10

gIM MNF. Thus MNF acted as a pure antagonist of TCDD action. However, MNF even

at 10 iM was unable to antagonize the induction of CYPIAI by 10 gM BaP, when added

either 2 hr earlier or at the same time as BaP. In addition, BaP produced a comparable

decrease in EGF receptor protein level in the absence or presence of MNF. Therefore, the

results were again not definitive on whether there was a mechanistic link between CYP1A1

induction and EGF receptor loss in cells treated with BaP.


Effects of Actinomvcin D and Cycloheximide on BaP-Mediated Changes in EGF Receptors

We next investigated whether BaP-mediated downmodulation of EGF receptor

protein in JEG-3 cells requires the synthesis of other mRNAs and proteins, using the RNA

synthesis inhibitor actinomycin D (AD) and the protein synthesis inhibitor cycloheximide

(CHX). The presence of AD or CHX alone had no significant effect on EGF receptor

protein levels after 16 to 24 hr exposure (Figure 4-9). When JEG-3 cells were treated

simultaneously with 10 ;tM BaP and 5 gg/ml AD or 10 jig/ml CHX, both inhibitors

completely blocked BaP-induced downmodulation of EGF receptor protein level. These

data provide evidence that the BaP-mediated loss of EGF receptor was dependent on de

novo mRNA and protein synthesis. In addition, induction of CYP1Al protein was

blocked by treatment with AD and CHX, and this was in association with the block of EGF

receptor downmodulation in JEG-3 cells cotreated with BaP and AD or CHX. Thus data

suggest that CYPlA1 activity or metabolism of BaP may be essential in the BaP-induced

loss of EGF receptor protein.


Effects of TCDD and BaP on Steady State EGF Receptor mRNA Levels

The effect of TCDD and BaP on the steady state level of EGF receptor mRNA was

analyzed by Northern blot techniques using a pE7 EGF receptor cDNA probe. BeWo and

JEG-3 cells both showed two transcripts of 10 and 5.6 kb (Figure 4-10). Quantitation of









these bands further demonstrated that the steady state level of EGF receptor mRNA was not

significantly changed by BaP or TCDD treatment in either cell line (Figure 4-11). These

data suggest that the BaP-mediated decrease in EGF receptor protein does not involve

changes in the steady state level of mRNA.



Discussion

Both BeWo and JEG-3 cells were found to express EGF receptor mRNA

transcripts of 10 and 5.6 kb and express immunoreactive EGF receptor protein of 170 kDa,

which is in agreement with values previously reported in human placenta and placental cell

cultures (Wang et al., 1988; Guyda et al., 1990; Fujita et al., 1991; Gabriel et al., 1995).

BaP treatment of both BeWo and JEG-3 cells resulted in a concentration-related decrease in

the binding of 125I-EGF, in agreement with earlier reports in primary cultures of early

gestation human placental cells exposed to BaP (Guyda et al., 1990). Moreover, Scatchard

analysis indicates a loss of high affinity EGF binding sites in choriocarcinoma cells

following treatment with BaP. In this regard, dose-dependent loss of EGF binding has

been previously reported with BaP in cultured human keratinocytes (Hudson et al., 1985),

mouse embryo fibroblasts (Ivanovic and Weinstein, 1982) and hepatoma cell lines

(Kiirenlampi et al., 1983). In the choriocarcinoma cells, our data indicate that the reduced

binding of 125I-EGF to whole cells is not due to altered internalization of cell surface

receptors because EGF binding was still significantly decreased at 40 when internalization

is minimal. In addition, Western analysis confirmed that the decrease in binding of 1251-

EGF was associated with a loss of the EGF receptor protein following BaP treatment.

These results further support our previous finding that the smoking-related defect in

placental EGF receptor autophosphorylation appeared to be due to the loss of EGF receptor

protein (Wang et al., 1988).

Although CYP1Al inducers were found to decrease EGF binding in various cells

more than a decade ago (Ivanovic and Weinstein, 1982; Kirenlampi et al., 1983; Madhukar









et al., 1984; Hudson et al., 1985), the mechanistic link between CYPIAl induction and

EGF receptor downmodulation still remains unclear. Structure-activity relationship and

study of congenic Ah-responsive and -nonresponsive mouse strains have indicated a role of

the Ah receptor in mediating EGF receptor downmodulation (Ivanovic and Weinstein,

1982; Lin et al., 1991; Safe, 1995). In mouse fibroblasts, Ivanovic and Weinstein (1982)

found that exposure of BaP led to a time- and concentration-dependent decrease of EGF

binding, whereas the highly reactive electrophilic metabolite of BaP, BaP-7,8-diol-9,10-

epoxide (BPDE), did not significantly alter EGF binding. A comparison of a series of 16

polycyclic compounds further showed a correlation between the capacity to inhibit EGF

binding and the apparent affinity of the same compound for the Ah receptor, as well as their

ability to induce the CYP1A1 system (Ivanovic and Weinstein, 1982). In this regard, Lin

et al (1991) reported that the Ah locus mediates the effects of TCDD on the hepatic EGF

receptor in C57BUL6J mice.

The present study, however, found that TCDD resulted in a dose-dependent

CYP1A1 induction without altering EGF receptor protein or binding, providing evidence

that there is no causal relationship between CYP1AI induction and EGF receptor

downmodulation in choricarcinoma cells. These results clearly demonstrate that occupancy

of the Ah receptor and/or interaction of the liganded-Ah receptor complex with the XRE

DNA sequences per se does not affect EGF binding or EGF receptor protein level in

choriocarcinoma cells, since both TCDD and BaP bind to the Ah receptor and induce

CYPlAI mRNA and protein, but only BaP alters EGF binding and EGF receptor protein

level (Figure 4-1 to 9). On the other hand, these data imply that the loss of EGF binding

and protein may be a consequence of metabolism of BaP to reactive metabolites, which

does not occur with TCDD treatment due to its resistance to metabolism. In this regard, the

most toxic BaP metabolite, BPDE, has been shown to effectively block EGF binding in the

mouse hepatoma c4 mutant cell line which has a defective Ah receptor, as well as in the

parent Hepa-1 cell cultures, whereas BaP itself decreased EGF binding only in Hepa-l









parent cells (Karenlampi et al., 1983). In these experiments, TCDD was not found to

affect the EGF binding in Hepa- cells. In addition, it is noteworthy that aryl hydrocarbon

hydroxylase, an enzyme activity associated with CYPIA1, is highly inducible in JEG-3

cells by TCDD, providing evidence that JEG-3 cells have the capacity to metabolize BaP to

reactive intermediates (Karenlampi et al., 1989).

Studies with inhibitors of RNA and protein synthesis showed that block of

CYPIAl induction by AD or CHX also prevented the BaP-induced loss of EGF receptor

protein, evidence which further supports that metabolism of BaP may be essential in the

decrease of EGF receptor protein. However, these studies cannot exclude the possibility

that another repressor protein may be involved in the downmodulation of the EGF receptor,

particularly because neither AD nor CHX is a specific inhibitor of CYPIAl mRNA or

protein synthesis. In addition, data indicate that BaP-mediated loss of EGF receptors does

not involve changes in the steady state mRNA level (Figure 4-11), suggesting that

alterations in EGF receptor synthesis, protein processing and half life or modulation of

autocrine networks by BaP are likely involved in EGF receptor downmodulation.

The Ah receptor activation pathway is a multistep process which involves the

binding of agonist to the receptor, recruitment of the partner factor Arnt, and interaction of

the liganded receptor-Arnt heterodimer with the target DNA XRE sequences. Disruption of

any of these steps can block the Ah receptor-mediated induction of CYP1A1. The

mechanism of antagonistic action of a-NF and MNF on Ah receptor function is not fully

understood. Using Arnt-deficient cells and chimeric receptor techniques, Wilhelmsson et

al. (1994) demonstrated that Ah receptor sequences rather than Arnt mediate the

antagonistic effects of a-NF. At the same time, these authors found that the c-NF-

occupied Ah receptor was able to recruit Arnt, resulting in XRE binding activity. In
agreement with the experimental data, this study has shown that a-NF has dual functions

with antagonism of TCDD induction of CYP1A1 when co-administered with TCDD, as

well as action as a partial agonist in inducing CYP1A1 expression when administered









alone. The observed lack of antagonistic effect of a-NF on BaP induction of CYP1Al is

possibly due to their similar potency in stimulating Ah receptor function in JEG-3 cells.

MNF is a newly characterized pure Ah receptor antagonist in MCF-7 human breast

cancer cells (Lu et al., 1995), in which cotreatment with 0.01 to 10 IIM MNF plus 1 nM

TCDD was reported to cause a concentration-dependent inhibition of TCDD-induced

formation of the nuclear Ah receptor complex, CYPIAI mRNA level and ethoxyresorufin

O-deethylase activity. In the present study, coexposure of JEG-3 cells to MNF and TCDD

completely blocked TCDD induction of CYP1Al protein (Figure 4-8). No apparent

antagonistic effect of MNF on BaP induction of CYPIAl was observed, however, when

JEG-3 cells were treated simultaneously with BaP and MNF. The absence of an

antagonistic effect of MNF on BaP induction of CYP1AI cannot be explained by their

relative potencies in stimulating the Ah receptor function, particularly since MNF alone did

not show any partial agonistic effect (Figure 4-8).

There is some evidence for multiple Ah receptor proteins. In addition to the well-

characterized 8S Ah receptor, an intracellular 4S polycyclic hydrocarbon-binding protein

has been reported to be involved in the induction of CYP1A1 in mouse hepatoma cells

(Sterling et al., 1994). In this regard, evidence has shown that the Ah receptor complex

exists as two distinct forms in MCF-7 human breast cancer cells (Lu et al., 1995) and in

guinea pig hepatic cytosol following TCDD treatment (Swanson et al., 1993; Bank et al.,

1995). Merchant et al. (1992) found that BaP bound to both the 4S binding protein and the

Ah receptor, although the 4S binding protein was not required for the induction of

CYP1A1. However, Manchester et al. (1987) found that only a specific Ah receptor

binding peak rather than binding in 4 to 5S region was observed in sucrose density gradient

profiles when human placental cytosols were incubated with [3H]BaP. Moreover, our data

in Chapter 3 (Figure 3-2) has shown that the higher level expression of the Ah receptor

mRNA is directly correlated with a higher level induction of the CYP1A1 mRNA in JEG

cells, compared with that in BeWo cells following BaP exposure (Figure 4-1).









Collectively, evidence strongly supports a mechanism in which BaP induces CYP1Al via

the Ah receptor. However, our finding that the pure Ah receptor antagonist MNF

completely inhibited CYP1Al induction by TCDD, but not by BaP, raises questions as to

whether TCDD and BaP may induce CYPIAl through different subtypes of the Ah

receptor or through different forms of the DNA-binding Ah receptor complex.

The steady state level of mRNA for EGF receptors was not found to be

significantly altered by either BaP or TCDD exposure in the BeWo and JEG-3

choriocarcinoma cell lines. In contrast, Fujita et al (1991) reported that the EGF receptor

and its mRNA levels were decreased in placentas from pregnancies with intrauterine

growth-retardation and diabetes mellitus. TCDD was associated with a reduction in both

EGF binding and EGF receptor mRNA steady state levels in rat uterus (Astroff et al.,

1990). However, exposure of human keratinocytes in vitro to TCDD and livers from

TCDD-treated mice showed a reduction in maximal EGF binding without a change in the

amount of mRNA for the EGF receptor (Lin et al., 1991). The observation that TCDD

induces expression of TGF-ac in keratinocytes (Choi et al., 1991) led to the hypothesis that

increased expression of this peptide growth factor may lead to internalization of the EGF

receptor and activation of cell responses. In recent studies, however, TGF-a mRNA was

not found to be increased in rat liver by TCDD (Vanden Heuvel et al., 1994), whereas EGF

receptor mRNA was reported to be decreased in correlation with the loss of EGF binding

(Sewall et al., 1995). Finally, it warrants note that TCDD has also been reported to

stimulate EGF receptor expression and proliferation in the embryonic palate and ureter

epithelial cells, which may be unique to early development (Abbott and Birnbaum, 1989

and 1990a). Thus, data suggest that the effect of BaP or TCDD on EGF receptor mRNA

levels may be a species and/or tissue-specific response.

In conclusion, there is no evidence of a direct causal relationship between CYP1Al

induction and EGF receptor downmodulation in choriocarcinoma cells. Metabolism of BaP






44


by induced CYP1A1 to reactive metabolites or synthesis of a repressor protein may be

responsible for the observed loss of EGF receptor in BaP-treated cells.


































B (fmole/mg)


Cell Treatment Bmax [fmole/mg, (%)] Kd (nM)

BeWo
Control 27.7 (100) 0.043
TCDD, 10 nM 32.7(118) 0.067
BaP, 10 gM 18.3(66) 0.059
JEG-3
Control 77.6 (100) 0.048
TCDD, 10 nM 95.1 (123) 0.063
BaP, 10 pM 45.1(58) 0.100



Figure 4-1. Scatchard plot analysis of 1251-EGF binding to control and treated
cells. Cells were treated with 0.1% DMSO (control), 10 nM TCDD or 10 gM BaP
for 48 hr prior to incubation with increasing concentrations of 125I-EGF (1.25 to
200 pM) at room temperature for 90 min in the presence or absence of excess
unlabeled EGF. Specific binding was determined as described under Methods and
expressed as fmole 125I-EGF bound per mg total cell protein. Bmax (fmole/mg)
and Kd (nM) values for 125I-EGF binding to BeWo and JEG-3 cells were
determined by Scatchard analysis of equilibrium binding data.

















control
TCDD 10 nM


60- BaPlOM T
T T

S 40 -
p0


20




240 40 240 40
BeWo JEG-3


Figure 4-2. Effects of TCDD and BaP on specific binding of 125I-EGF to BeWo
and JEG-3 cells. The cells were treated with 0.1% DMSO (control), 10 nM TCDD
or 10 gM BaP for 48 hr prior to incubatuion at room temperature (240C) for 90 min
or at 40C for 5 hr with 100 pM of 125I-EGF in the presence or absence of unlabeled
100 nM EGF. The specific binding was normalized with respect to protein in each
sample and expressed as cpm per ig protein of total cell lysate. Values are the mean
SE of triplicate cultures. Differs from control at p < 0.05 by t test.


El
r'1




















100

0 Ohr
S3i 6hr
75- 24hr
o U 48 hr
&] 72 hr
S0 96hr







0_
50 -








BeWo JEG-3



Figure 4-3. Time-course of BaP effects on 125I-EGF binding to whole cells. The
BeWo and JGE-3 cells were treated with 10 pM BaP for 6, 24, 48, 72 and 96 hr
prior to incubation at room temperature for 90 min with 205 pM 125I-EGF in the
presence or absence of unlabeled 328 nM EGF. The specific binding was
normalized with respect to protein in each sample and expressed as fmole 125I-EGF
bound per mg protein of total cell lysate. Values are the mean SE of duplicate
cultures. The points without the standard error bars indicate that the individual SEs
are too small to be shown. Differs from time 0 point at p <0.05 by t test.













---- BeWo

150 ........ ....... JEG-3



100



50


I I
0.1 1

TCDD (nM)


1 I
10 100


BaP (M)


Figure 4-4. Effects of TCDD and BaP on immunoreactive EGF receptor protein in
BeWo and JEG-3 cells. The cells were harvested following incubation with TCDD
or BaP for 48 hr. Cell protein, 100 Mg, was electrophoresed, transferred, and
immunostained with sheep anti-EGF receptor. The immunoreactive proteins were
then quantitated, with the average value of the controls being arbitrarily set as
100%. Values are the mean SE of three separate experiments. *Differs from
control at p < 0.05 by Fisher PLSD and t test.
























--C-- BeWo
S150- JEG-3
8 ................ JEG -3






50-



0
0 24 48 72 96

Time (hr)

Figure 4-5. Time-course of the effect of TCDD on EGF receptor protein
levels. Cells were treated with 10 nM TCDD for various times as indicated.
Total cell protein, 100 gg, was separated on 7.5% PAGE, transferred to
nitrocellulose, and immunostained with anti-EGF receptor. The
immunoreactive band of 170 kDa was quantitated by densitometry
scanning, with the average density of the controls being set as 100 %.
Values are the mean SE of three separate experiments.


LI'U
























EGFR
(170 kDa)


a -

_ea


I~,ji~4


CYP1A1
(55 kDa)








Figure 4-6. Comparison of the effects of TCDD and BaP on EGF receptor and CYPIAI.
Cells were treated with TCDD or BaP for 48 hr. Total cell protein, 100 Ag, was subjected
to PAGE, transferred, and immunostained with anti-EGF receptor or anti-CYPlAl
antibody as described in Materials and Methods. Shown are a representative immunoblot
of JEG-3 cells.


















TCDD(nM)


10 10 10


BaP (M) 10 -


10 10


a-NF (M) 1 10 1 10 1
CYPIA1
(55 kDa) :


EGIFR
(170 WDe) IH Wl r^


I


Figure 4-7. Effects of the Ah receptor antagonist a-NF on BaP-mediated changes in EGF
receptor expression and CYP1A induction in JEG-3 cells. The cells were treated with the
respective chemicals at various concentrations for 48 hr. Total cell protein, 100 ug, was
separated by PAGE and transferred in duplicate blots. One blot was immunostained with
anti-CYPIAl, and the other with anti-EGF receptor.


m I














10 10 10 -
S 10 10 10


MNF (M) 1


10 1 10 1 10


EGFR U U I
(170 kDS) inl l Bl


Figure 4-8. Effects of the Ah receptor antagonist MNF on BaP-mediated changes in EGF
receptor protein and CYPlAl induction in JEG-3 cells. The cells were treated with the
respective chemicals at various concentrations for 24 hr. Total cell protein, 100 gg, was
separated by PAGE and transferred in duplicate blots. One blot was immunostained with
anti-CYPlAl, and the other with anti-EGF receptor.


TCDD (nM)
BaP (M)


CYP1A1 I
(55 kDa) I ""


-40 OW


...A[












TCDD (nM)

BaP (pM)


AD (pg/ml) -

CHX (Ig/ml) -


5 5 5 -

- 10 10 10


EGFRiHr rr
(170OkDa) 1 111 1 p p


CYP1AI
(55 kDa)


U U


F-.
[-1-


Figure 4-9. Effects of AD and CHX on BaP-mediated changes in EGF receptor protein
and CYP1A1 induction in JEG-3 cells. The cells were treated with the respective chemicals
at various concentrations for 16 to 24 hr. Total cell protein, 100 ig, was separated by
PAGE and transferred in duplicate blots. A) One blot was immunostained with anti-
CYP1A1, and the other with anti-EGF receptor. B) Quantitation of the intensity of the 170
kDa EGF receptor band by densitometry, with the control values being set as 100%.
Results are the mean SE of three separate experiments. p < 0.05 as compared with
control; A p < 0.05 as compared with BaP alone by t test.


10 -


10 10


10 10


T T i




IlU flf .l. .


'm Noe

















BeWo JEG-3



0 0


EGFR
(10kb) O

(5.6 kb) *SS6SS


.-actin
(2.0 kb)






Figure 4-10. Northern blot analysis of EGF receptor mRNA. Cells were incubated with
or without TCDD and BaP for 48 hr. Poly(A)+RNA, 10 pg, was denatured, blotted, and
hybridized with 32P-labeled cDNA probes as described in Materials and Methods. Shown
are representative autoradiograms of the Northern blot.




















0 -.-0-- BeWo

S.......o........ JEG-3

0 1
0 0.1 1 10 100
< TCDD (nM)
S 2-
S -0-- BeWo T
1.5-......... JEG-3 T

S........ ..... JEG-3



I I T
1 .




0.5-


0
0 1 10 100
BaP (gM)

Figure 4-11. Steady state level of EGF receptor mRNA in BeWo and JEG-3 cells
following culture in the absence or presence of TCDD and BaP for 48 hr.
Poly(A)+RNA, 10 gg (TCDD) or total RNA, 40 gg (BaP), was denatured, blotted,
hybridized with the 32P-labeled EGF receptor cDNA probe, and subsequently
rehybridized with the 32P-labeled 3-actin cDNA as shown in Figure 4-10. Shown
are the results from quantitation of the 10 and 5.6 kb transcripts of EGF receptor
mRNA, with the ratio of EGF receptor message to 3-actin message in the control
cells being set as 1. Values are the mean SE of three (BaP) or five (TCDD)
separate experiments.














CHAPTER 5
EFFECTS OF TCDD AND BAP ON TGF-a, TGF-pI, C-MYC AND PAI-2 GENE
EXPRESSION


Introduction

Increasing evidence indicates that TCDD and related Ah receptor ligands act as

endocrine disruptors and growth modulators by persistently altering the expression of

important growth control genes through sustained activation of the Ah receptor system (as

reviewed by Huff et al., 1994; Safe, 1995b; Birnbaum, 1995; DeVito and Bimbaum,

1995). TCDD has been shown to alter the expression of a number of genes important in
cell growth and differentiation, including transforming growth factor (TGF)-a, TGF-P,

and plasminogen activator inhibitor 2 (PAI-2) (Abbott and Birnbaum, 1990b; Choi et al.,

1991; Sutter et al., 1991; Gaido et al., 1992; Gaido and Maness, 1994; Vogel and Abel,

1995). One of the hallmarks of the effects of TCDD on growth control genes is significant

tissue- and species-specificity, as stated earlier in Chapter 1. Little is known regarding the

nature of Ah receptor regulated genes in human placental cells.

As discussed in the Introduction Chapter, TGF-c, TGF-pl, c-myc and PAI-2 are

all expressed in human placental trophoblast cells and are major partners in the autocrine

and paracrine networks which control trophoblast proliferation, differentiation and

invasiveness (Goustin et al., 1985; Maruo and Mochizuki, 1987; Feinberg et al., 1989;

Ohlsson, 1989; Kauma et al., 1990; Lala and Graham, 1990; Graham et al., 1992; Filla et

al., 1993; Horowitz et al., 1993; Lysiak et al., 1993, 1994 and 1995; Hofmann et al.,

1994). This study was undertaken to investigate whether TCDD and BaP alter placental

function by disrupting the local autocrine and paracrine networks of these important








trophoblast growth control genes. The objective of the study was to identify potential

biomarkers of placental toxicity for exposure to environmental chemicals such as TCDD

and BaP. The results indicate that the steady state level of TGF-a mRNA was increased by

TCDD in BeWo cells, whereas BaP modulated TGF-P1 and c-myc mRNA and protein

expression in JEG-3 cells. Evidence further suggests that developmental windows may

exist in placental and trophoblast growth for altered responses to environmental chemicals.



Results



Effects of TCDD and BaP on the Steady State mRNA Levels for TGF-a. TGF-P 1. PAI-2
and c-myc


Our initial experiments examined the effects of TCDD and BaP on the steady state
mRNA levels for TGF-a, TGF-p 1, PAI-2 and c-myc in both BeWo and JEG-3 cells

(Figure 5-1). Expression of CYPlA1, a gene under direct transcriptional control by the Ah
receptor ligands, was used as a positive control. CYP1A1, TGF-fl, PAI-2 and c-myc

mRNAs were expressed in choriocarcinoma cells as 3.0, 2.5, 2.3 and 2.4 kb transcripts,

respectively, values which are in agreement with previous reports in other tissues

(Ohlsson, 1989; Sutter et al., 1991; Greenberg and Ziff, 1984; Gaido et al., 1992). In
contrast, the TGF-a mRNA in BeWo and JEG-3 cells was detected as a single 2.3 kb

transcript, which is an alternate of the 4.5 kb transcript reported to be expressed by most

cells (Lee et al., 1995). CYPIAI mRNA expression was induced by TCDD at 0.1, 1, 10

and 100 nM in a concentration-dependent manner in both cell lines. Comparable induction

of the CYP1A1 mRNA was observed following treatment with 10 pM BaP.

Exposure to TCDD for 48 hr resulted in a significant concentration-dependent
increase in the steady state level of TGF-a mRNA in BeWo cells, but not in JEG-3 cells.

In contrast, the steady state mRNA levels for TGF- 1, PAI-2 and c-myc were not altered

by TCDD in either cell line (Figure 5-2). Conversely, exposure to 10 pM BaP for 48 hr









significantly increased the steady state TGF-11 mRNA level in both cell lines, and

decreased the steady state c-myc mRNA level in JEG-3 cells, but not in BeWo cells. The

steady state levels for TGF-a and PAI-2 mRNA were not altered by BaP in either cell line.

The effects of BaP on the steady state TGF-pl and c-myc mRNA levels were further found

to be concentration-related (Figure 5-3). The steady state TGF-31 mRNA levels were

increased 1.5-, 3.0- and 2.9-fold over control, while c-myc mRNA levels were decreased
by 38, 61 and 67%, respectively, following exposure of JEG cells to 1, 10 and 20 gM BaP

for 48 hr.

We next examined the time course for changes in mRNA expression of TGF-pl

and c-myc. Exposure to 10 nM TCDD for 6 to 120 hr (5 days) had no effect on the steady

state levels of TGF-31 and c-myc mRNA. In contrast, BaP exposure caused a time-

dependent change in the steady state TGF-pl and c-myc mRNA level, and the time

dependence was different from that for the induction of CYPIA1 (Figure 5-4). A marked

increase in CYP1Al mRNA level was detected at 6 hr, the earliest time point examined,
following exposure to 10 nM TCDD and 10 gM BaP, and levels progressively increased

until 72 hr and remained elevated for 5 days, the last time point examined. Significant

increases in TGF-I31 mRNA level, however, were not detectable until 24 hr after addition

of BaP. By 72 hr, TGF-31 mRNA level in BaP-treated cells was increased 4-fold over

control and remained significantly elevated for 5 days. In contrast, c-myc mRNA level was

not changed at 6 and 24 hr, but was significantly decreased by 50% at 72 hr, and further
depressed by 60% at 5 days in cultures treated with 10 PM BaP.


Effects of TCDD and BaP on the Rate of TGF-Dll and c-myc Gene Transcription


Nuclear run-off analysis was performed to determine whether the observed changes
in the steady state mRNA levels of TGF-p1 and c-myc occurred at the level of mRNA

transcription (Figure 5-5). CYP1Al was used as a positive control for transcriptional

activation by TCDD and BaP. Transcription of CYP1A1 was activated at both 6 and 24 hr









following exposure to 10 nM TCDD and 10 gtM BaP. Consistent with previous reports

(Israel et al., 1985), CHX alone resulted in little increase in the rate of CYPIAI

transcription. However, cotreatment with CHX plus TCDD caused a 3-fold increase in the

rate of CYPIAI transcription compared to TCDD alone. Cotreatment with CHX plus BaP

also caused a 2-fold increase in CYP1AI gene transcription compared to BaP alone. The
rate of TGF-P 1 transcription was unaltered at 6 hr, but was significantly induced 4-fold in

BaP-treated cells at 24 hr. The rate of c-myc transcription was also unaltered at 6 hr, but

appeared to be increased at 24 hr following BaP treatment. No significant effect of TCDD

treatment was observed on the rate of TGF-P 1 and c-myc transcription at either 6 or 24 hr

(Figure 5-5). In addition, CHX alone or in combination with TCDD or BaP did not affect
the rate of TGF-P 1 and c-myc transcription at 6 hr.


Effects of BaP on the Stability of TGF-pl and c-myc mRNA


Studies were further performed to evaluate the effects of BaP treatment on the
stability of TGF-P 1 and c-myc mRNA. The decay rate of c-myc mRNA in the presence of

actinomycin D was found to be significantly increased in BaP-treated cells (Figure 5-6).

The half life of the c-myc mRNA was shortened from 30 min in control cells to 18 min in

BaP-treated cells. Thus results indicate that the decrease in the steady state c-myc mRNA

level following BaP treatment could be accounted for by the decrease in stability of the c-

myc mRNA. In contrast, no alteration in TGF-p1 mRNA stability was observed. The half

life of the TGF-pI mRNA was approximately 10 hr in both control and BaP-treated cells.

The finding that the stability of TGF-P31 mRNA was not increased following BaP exposure

further supports the previous finding that the increase in the steady state TGF-1I mRNA

level is due to the increase in the rate of TGF-P 1 transcription.









Effects of TCDD and BaP on TGF-l1 and c-Myc Protein Levels

TGF-lI was undetectable in the media from control or treated JEG-3 cells using an

ELISA assay which detects biologically active TGF-Pl. Activation of the conditioned

media (CM) by low pH, however, converts latent TGF-p31 to the bioactive form such that a

TGF-3I concentration of 1280 165 pg/ml was measured in the CM of control cells. Data

in Figure 5-7 show that BaP treatment for 48 hr increased TGF-pl secretion by JEG-3

cells in a concentration-related manner. Due to the high variability between assays,

however, the increase was not statistically significant. The secretion of TGF-Pl also

appeared to be increased slightly following exposure to TCDD for 48 hr.

Western blot analysis using an antibody to human c-myc protein identified a minor

and a major immunoreactive band with an apparent molecular mass of approximately 67

(Myc-1) and 64 (Myc-2) kDa (Figure 5-8), respectively. These observed values for Myc-1

and -2 in choriocarcinoma cells are in agreement with previous reports in other tissues

(Munger et al., 1992; Packham and Cleveland, 1995). Quantitation of the major 64 kDa

band revealed a concentration-dependent decrease in Myc level following BaP, but not

TCDD, treatment. Myc levels were significantly decreased by 38, 46, 54 and 50% at 1, 5,
10 and 20 ViM BaP, respectively, after treatment for 48 hr. Quantitation of the minor 67

kDa band showed similar results (data not shown).



Discussion

TCDD modulation of growth control genes shows significant tissue and species

specificity (Safe, 1995b; Birbaum, 1995). Our studies indicate that TCDD increased TGF-
a mRNA level in BeWo but not in JEG-3 human placental cells (Figure 5-1 and 2). The

differential response to TCDD observed in BeWo cells compared with JEG-3 cells may

reflect developmental differences in the state of differentiation of these two cell lines, the

latter being more invasive into cultured reepithelialized endometrial fragments (Griimmer et









al., 1994). In this regard, TGF-a mRNA level has been shown to be increased by TCDD

in human primary and SCC-12F keratinocyte and breast cancer MCF-7 cells (Gaido et al.,

1992; Choi et al., 1991; Vogel and Abel, 1995), but not in mouse and rat liver (Lin et al.,

1991; Vanden Heuvel et al., 1994). Consistent with the earlier observations in human

keratinocyte SCC-12F and breast cancer MCF-7 cell lines (Gaido et al., 1992; Vogel and

Abel, 1995), the present study has shown that the steady state mRNA level or transcription
rate of TGF-P l was not altered by TCDD in either BeWo or JEG-3 cell line (Figure 5-1, 2,

4 and 5). It warrants note, however, that TCDD has been shown to reduce the expression
of TGF-a and TGF-31 in mouse embryonic palate epithelial and mesenchymal cells

(Abbott and Bimbaum, 1990), which may be unique to early development.

In choriocarcinoma cells, the steady state c-myc mRNA level was found to be

unchanged following TCDD exposure, which is in agreement with an earlier report in

mouse Hepa-1 hepatoma cells (Puga et al., 1992). TCDD has been shown to decrease c-

Myc DNA binding activity by modulating its state of phosphorylation in guinea pig adipose

tissue (Enan and Matsumura, 1994 and 1995). It has been further reported that the

expression of PAI-2 mRNA was increased by TCDD in human SCC-12F keratinocyte,

primary hepatocyte, monocytic U937, hepatoma HepG2 and breast cancer MDA-MB 231

cells (Sutter et al., 1991; Dohr et al., 1995; Gohl et al., 1996). However, Vanden Heuvel

et al. (1994) found that the PAI-2 mRNA level was not altered by TCDD in rat liver, which

is in agreement with our present finding in choriocarcinoma cells.

Regardless of the large variability in TCDD responsiveness, CYPIA1 mRNA is

uniformly induced by TCDD in human keratinocytes, hepatocytes, HepG2 and breast

cancer cells, mouse and rat liver (Lin et al., 1991; Sutter et al., 1991; Gaido et al., 1992;

Choi et al., 1991; Vanden Heuvel et al., 1994; Dohr et al., 1995; Vogel and Abel, 1995;

Gohl et al., 1996), as well as in choriocarcinoma cells as shown by this study (Figure 5-1

and 4). Our data, therefore, provide further support for action by TCDD through the








classic Ah receptor mechanism with tissue- and gene-specific differences in

responsiveness.

The mechanisms responsible for the tissue- and gene-specific differences in TCDD

responsiveness, however, are poorly understood. Gradin et al. (1993) found that the

nonresponsiveness of normal human fibroblasts to TCDD was due to the presence of a

constitutive XRE-binding factor. In a study of the relationships between DNA sequence,

receptor binding, and TCDD responsiveness, Lusska et al. (1993) demonstrated that the Ah

receptor-DNA binding event per se was not sufficient to confer TCDD responsiveness

upon a linked gene. Based upon a comparison of the DNA sequences of the Ah receptor

binding sites, the authors suggested a "functional consensus" recognition sequence, which

is more extended in length than the "core"-binding sequence 5'-TNGCGTG-3'. These

authors further proposed that the liganded Ah receptor can function in either positive or

negative fashion, depending on the regulatory context. Thus, the action of some cell- or

gene-specific factors might account for some of the gene-to-gene differences in TCDD

responsiveness.

BaP is a transplacental carcinogen (Bulay and Wattenberg, 1970), and its mutagenic

effects are well characterized (Levin et al., 1978). However, the biochemical events

associated with BaP exposure other than covalent DNA binding are poorly understood.

This study demonstrated for the first time that exposure of JEG-3 cells to BaP caused a

persistent increase in TGF-p 1 mRNA and a sustained depression in c-myc mRNA and

protein levels, along with concomitant induction of CYP1AI mRNA (Figure 5-3, 4 and 8).

Conversely, BaP was reported to increase the steady state c-myc mRNA level in rat aortic

smooth muscle cells (Sadhu et al., 1993). In addition, our data in Chapter 3 and 4 have

demonstrated that EGF receptor protein level was significantly decreased, accompanying

induction of CYPIAl protein following BaP exposure. Thus, growing evidence indicates

that, like TCDD, BaP also leads to species-, tissue- or gene-specific modulation of growth

control gene expression.









Our data indicate that BaP-induced changes in the steady state TGF-pl and c-myc

mRNA level may not be a primary response to the interaction of the liganded Ah receptor

complex with specific xenobiotic responsive elements (XREs) in the promoter region.
First, there is a temporal lag in the effect of BaP on TGF-pl1 and c-myc gene expression

relative to CYPlAI induction (Figure 5-4 and 5). Second, TCDD has no effect on these

two genes in JEG-3 cells, despite the CYP1Al induction (Figure 5-1, 2, 4 and 5).

Transcriptional activation of CYPIAI is a well-characterized primary response to the

binding of the ligand-bound Ah receptor complex to upstream XREs of the CYPlA1

transcription start site (Hankinson, 1995) and was observed at 6 hr, the earliest time point

examined, in JEG-3 cells following treatment with both TCDD and BaP (Figure 5-5).

However, the rate of TGF-P 1 transcription was not altered by BaP at 6 hr, nor by TCDD,

in contrast to the early increase in CYPIA1 transcription. Therefore, although the BaP-
mediated increase in the steady state TGF-p31 mRNA level involves direct transcriptional

regulation of the TGF-p 1 gene, this may not necessarily be regulated directly by the

liganded Ah receptor complex, but more likely represents a secondary response. In this

regard, MacLeod et al (1995) recently demonstrated that BPDE modification of GC-box

sequences in the promoter region of the hamster adenosine phosphoribosyl transferase gene

caused a substantial increase in the apparent affinity for the transcription factor Spl

(MacLeod et al., 1995). Moreover, this study has shown that Spl bound to the BPDE-

modified non-GC-box DNA fragment with relatively high affinity. This type of evidence

suggests that BPDE-DNA adduct sites can interact with Spl, which may selectively affect

transcription of specific genes.

TGF-P has been reported to downmodulate c-myc mRNA expression in the mouse

BALB/MK keratinocyte cell line, secondary cultures of human keratinocytes, and the

human MOSER colon carcinoma cell line (Pietenpol et al., 1990; Munger et al., 1992;
Mulder et al., 1988). The block in c-myc expression by TGF-pl1 has further been shown

through inhibition of transcriptional initiation (Pietenpol et al., 1990). In the present study,









the time course of upmodulation of TGF-3 1 by BaP preceded the downmodulation of c-

myc expression (Figure 5-4); however, it remains to be determined whether there is any

causal relationship for the inverse changes between these two factors. In JEG-3 cells, the

observed decrease in the steady state c-myc mRNA level following BaP treatment appears

to occur posttranscriptionally, based upon evidence from the mRNA stabilization assay

which showed decreased stability of c-myc mRNA following BaP treatment (Figure 5-6).

Altogether, the present study demonstrates that regulation of gene expression by BaP

results in differential changes in mRNA levels for specific genes and can occur by multiple

mechanisms, including transcriptional and posttranscrriptional regulation, as previously

described for TCDD (Gaido et al., 1992; Hankinson, 1995).
TGF-a has been shown to stimulate trophoblast proliferation (Lysiak et al., 1993)

and 17f-hydroxysteroid dehydrogenase type 1 activity which catalyzes the reversible

interconversion of estrone and estradiol in placental cells (Lewintre et al., 1994). Thus,
our finding that TCDD increased the TGF-a mRNA level in BeWo but not in JEG-3 cells

implicates that TCDD may interfere with normal human trophoblast proliferation and

endocrine function at a certain stage of placental development.
In this study, changes in c-Myc protein and TGF-pl levels were observed which

correlated with the alterations in c-myc and TGF-f1 mRNA levels following exposure of

JEG-3 cells to BaP, which may be relevant to retardation of fetal growth. These two genes

have been reported to be regulated in concert with changes that affect placental cell

proliferation, differentiation and invasiveness, with TGF-31 being anti-proliferative and

anti-invasive, and c-myc being linked with proliferative and invasive trophoblast activities

(Goustin et al., 1985; Ohisson, 1989; Schmid et al., 1989; Lala and Graham, 1990;

Graham and Lala, 1991; Graham et al., 1992; Cross et al., 1994). Studies with human
trophoblasts have shown that TGF-p 1 upregulates tissue inhibitor of metalloproteinases

and extracellular matrix proteins such as oncofetal fibronectin, as well as downregulates the

activity of u-PA and collagenase type IV (Lala and Graham, 1990; Graham and Lala, 1991;








Graham et al., 1994; Guller et al., 1995). Interestingly, c-Myc has been found to repress

collagen gene expression (Packham and Cleveland, 1995). Therefore, BaP-mediated

upmodulation of TGF-lI and downmodulation of c-myc may lead to accumulation of

extracellular matrix such as collagen in human placenta. Indeed, placentas from women

who smoke cigarettes have been shown to exhibit thickening of the basement membrane

and increased collagen content of the villous stroma (Asmussen, 1980).

Upmodulation of TGF-PI may also disrupt placental endocrine function. TGF-pI

has been found to inhibit both basal and EGF-stimulated human chorionic gonadotropin

and placental lactogen secretion by primary cultures of cytotrophoblasts (Morrish et al.,

1991) and mouse growth hormone releasing factor secretion by placenta (Yamaguchi et al.,

1994). In fact, serum levels of human placental lactogen were found to be lower in heavy

smokers than those in nonsmokers (Mochizuki et al., 1984). In mice c-myc mutant

embryos are small and retarded in development compared with their littermates (Davis et

al., 1993), providing direct evidence that c-Myc is necessary for normal embryonic

development. Therefore, the observed alterations in TGF-01 and c-myc gene expression

may underlie mechanisms by which xenobiotics such as those found in cigarette smoke

cause fetal intrauterine growth retardation.

In summary, the present study has demonstrated that 1) TCDD increased TGF-a

mRNA expression in BeWo but not in JEG-3 cells; 2) TCDD had no effect on the steady
state mRNA levels for TGF-pl, PAI-2 and c-myc in either cell line; 3) BaP increased TGF-

1l mRNA and protein expression at the level of gene transcription, while c-myc mRNA

and protein levels were decreased via a posttranscriptional destabilization of the mRNA in
JEG-3 cells; 4) BaP also caused a slight increase in TGF-PI mRNA level in BeWo cells; 5)

BaP had no effect on TGF-a and PAI-2 mRNA level in either cell line. The significance of

these changes is that a specific temporal expression of TGF-a, TGF- I and c-myc is

important for the control of trophoblast proliferation, differentiation and invasiveness. The
disruption of coordinated TGF-a, TGF-Ip and c-myc gene expression may directly






66


interfere with normal placental development, which subsequently may lead to altered fetal

growth. In addition, these data imply that different mechanisms may be involved in the

placental toxicity of TCDD and BaP.










A) TCDD (nM) BaP (iM)
0 0.1 1 10 100 10
CYP1A1
(3.0 kb)


TGF-a
(2.3 kb) 9 9 O0


TGF-P1
(2.5 kb)


PAI-2
(2.3 kb) MO_


c-myc
(2.4 kb)

I I
NowIglH


p-actin
(2.0 kb)


Figure 5-1. Northern blot analysis following treatment of cells with DMSO (control),
TCDD or BaP for 48 hr. A) BeWo; B) JEG-3. Poly(A)+ RNA, 10 pg, was separated on
1.0% formaldehyde-agarose gel, transferred to nylon membrane, and probed with 32p_
labeled cDNAs as indicated.


WWzAAJ












B) TCDD (nM) BaP (pM)
0 0.1 1 10 100 10
CYPIA1
(3.0 kb)


TGF-a
(2.3 kb) WW P


TGF-Pl
(2.5 kb)


PAI-2
(2.3 kb)


c-my2
(2.4 kb)


P-actin
(2.0 kb)


Figure 5-1 continued


I ri










control
TCDD 0.1 nM
TCDD 1 nM
TCDD 10 nM
TCDD 100 nM
BaP 10 uM

TT


control
TCDD O.1 nM*
TCDD 10 nM T
TCDD 10nM
TCDD 100 nM
BaP 10 tM


IT


TGF-PIl


100-
T
50-

0-
CYP


150

100- i
50-
0


c-myc PAI-2


Figure 5-2. Effects of TCDD on the steady state mRNA levels of TGF-a, TGF-pl, c-myc
and PAI-2. Each hybridization band as shown in Figure 5-1 was quantitated by
densitometry and normalized to p-actin message, with the control values being set as 1.
Values are the mean SE of three experiments. The inserts shown are for CYPIAI
induction as a positive control. *Differs from control by Fisher PLSD or t test.


BeWo


JEG-3 []


|]


TGF-a


- .





















S-TGF-pl **
0 3- T T
0 c-myc








Sand -myc following BaP treatment. JEG-3 cells were treated with 0.1 DMSO or
T-






10 and 20 gM BaP in 0.1% DMSO for 48 hr and poly(A)+ RNA was isolated for
Northern analysis. The blots were sequentially hybridized with the 32P-labeled TGF-pl,
c-myc and p-actin probes, and quantitated by densitometry, with the ratio of each
message to p-actin message in the control cells being set as i. Values are the mean SE
of four separate experiments. The points without the standard error bars indicate that the
individual SEs are too small to be shown. *p <0.05 as compared with control by PLSD
and t test.












A) Time (h)



CYP1AI
(3.0 kb)


6 24 72 120
C T B C T B C T B C T B

A& h& ~k U


TGF-(1 k)l l
(2.5 kb)


c-myc
(2.4 kb)




P-actin
(2.0 kb)


Figure 5-4. Time-dependent changes in the steady state mRNA levels of TGF-pl and c-
myc. JEG-3 cells were treated with 0.1% DMSO (C), 10 nM TCDD (T), or 10 PM BaP
(B). Poly(A)+ RNA was extracted for Northern analysis at 6, 24, 72, and 120 hr after
treatment. The blots were hybridized with 32P-labeled probes as indicated. A)
Autoradiograms of the Northern blot; B) Quantitation of the TGF-pl and c-myc mRNA,
with the ratio of each message to p-actin message in the control cells being set as 1. Values
are the mean SE of three experiments. The points without the standard error bars indicate
that the individual SEs are too small to be shown. *p < 0.05 as compared with control by t
test.






















BaP, TGF- 1
.0.O- BaP, c-myc
------- TCDD,TGF-PI
----A---- TCDD, c-myc


24 48 72 96 120

Time (hr)


Figure 5-4 continued


B) 7-



6-



5-



4-



i 3-



2-



1-


I

















C TCDD BaP
c-myc

TGF-P 1

CYP1A1 J

P-actin


Treatment TGF-P13 c-myc CYPIAI
6h

TCDD, 10nM 1.01 0.12 0.99 0.31 2.89 2.2.03
BaP, 10 lM 1.03 0.19 1.07 0.33 2.29 1.52
CHX, 10 lg/ml 1.25 0.37 1.02 0.21 0.99 0.20

TCDD+CHX 1.06 0.20 0.97+ 0.21 7.96+ 1.6*

BaP + CHX 1.14 0.47 0.86 0.16 4.45 1.42*

24 h

TCDD, 10nM 1.000.15 0.73 0.16 18.83 17.45
BaP, 10 IM 4.14 1.23* 3.99 1.57 12.80 2.49*


Figure 5-5. Effects of TCDD and BaP on the rate of TGF-pl and c-myc transcription by
nuclear run-off assay. Cells were treated with 10 nM TCDD, 10 pM BaP or 0.1% DMSO (C)
in the presence or absence of 10 Rg/ml CHX. Nuclei were isolated at 6 and 24 hr and nuclear
run-off assay was performed. The [a-32p] UTP labeled nascent transcripts were hybridized
with nylon blots with 2 gg of specific c-myc, TGF-pl, CYPIA1 and B-actin cDNA in each
dot. A) A representative autoradiogram of the dot blots following 24 hr treatment; B)
Hybridization signal was quantitated by densitometry and normalized to p-actin. Results are
expressed as the relative level of gene transcription in treated cells versus control cultures.
Values are the mean SE of two (TCDD) to three (BaP) experiments. *p < 0.05 as compared
with control by t test.














30







I., Q5 control

120-_ BaP 10M

0 25 50 75 100
c-myc mRNA (% of initial)


30- ----





6-


~12 control
1 BaP IOgM

0 25 50 75 100
TGF-P1 mRNA (% of initial)

Figure 5-6. Effects of BaP on the stability of c-myc and TGF-p1 mRNA in the
presence of the RNA synthesis inhibitor AD. JEG-3 cells were pretreated with
0.1% DMSO or 10 pIM BaP for 24 hr, and AD was added directly to the media to a
final concentration of 5 gg/ml. Total RNA was extracted for Northern analysis at
the different time points after the addition of AD as indicated. The hybridization
signal was quantitated, with the ratio of each message to p-actin message at 0 time
point being set as 100%. *p <0.05 as compared with control by t test.
point being set as 100%. *p < 0.05 as compared with control by t test.












JVU -

400-


300-


200-
T...... ....
too ...... ...i
100 ... .. .... .....

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


0 1 5

BaP (iM)


10 20


0 1 10 100
TCDD (nM)
Figure 5-7. Effects of BaP and TCDD on the secretion of TGF-p1 by JEG-3 cells.
Cells were treated with BaP or TCDD for 48 hr, and the conditioned medium was
then collected and acid-activated for assay of TGF-01 levels. Values are the mean
SE of seven experiments.
















BaP (pM) TCDD (nM)

0 1 5 10 20 1 10 100


TCDD(nM)
I 10


BaP (tM)
Figure 5-8. Effects of TCDD and BaP on c-Myc protein expression in JEG-3 cells.
Cells were treated with the indicated concentrations of BaP or TCDD for 48 hr. Total
cell lysate (100 pg) was separated on 10% PAGE, transferred to nitrocellulose and
immunostained with anti-human c-myc. The bands were visualized by ECL detection
(A), and the intensity of the 64 kDa Myc-2 band were quantitated by densitometry (B).
Similar results were obtained with quantitation of the minor 67 kDa band (data not
shown). Values are the mean SE of seven experiments. p < 0.05 as compared with
control by Fisher PLSD.














CHAPTER 6
EVALUATION OF THE EFFECTS OF TCDD AND BAP ON CELLULAR GROWTH
AND ENDOCRINE RESPONSES OF BEWO AND JEG-3 CELLS



Introduction

The main objective of this project is to identify the target genes that represent

primary biologic responses to the environmental chemicals TCDD and BaP. The challenge

in this study is to use the knowledge of the alterations induced in gene expression by

TCDD and BaP to further test the hypothesis regarding likely changes in the more complex

biologic endpoints of cell proliferation, migration and endocrine function. This study has
demonstrated that 1) TCDD increases TGF-a mRNA expression in BeWo cells, 2) BaP

decreases EGF receptor levels in both BeWo and JEG-3 cells, 3) BaP increases TGF-pl

mRNA expression in both cell lines, 4) BaP decreases c-myc mRNA and protein

expression in JEG-3 cells. The next question was whether these changes in trophoblast

growth control networks alter characteristic trophoblastic cell functions of proliferation,

hormone secretion and invasiveness.



Results


Effects of TCDD and BaP on Cell Proliferation as Measured by the MTT Conversion


The MTT conversion assay was used to determine whether TCDD and BaP

treatment was associated with cytotoxic effects on cell division. Data in Figure 6-1 show
that exposure to 0.1-100 nM TCDD or 1 nM-50 pIM BaP had no significant effect on cell

proliferation up to 48 hr in BeWo and JEG-3 cells cultured in the presence of serum. In









contrast, under serum-free conditions, BeWo and JEG-3 cell proliferation was inhibited

30-50% with TCDD at 10 and 100 nM and BaP at 10 and 50 jM. Thus TCDD and BaP

exposure for 48 hr at higher concentrations adversely affected cell viability only under

serum-free conditions.

We next examined whether the BaP-mediated loss of EGF receptors altered EGF-

stimulated cell proliferation. EGF at concentrations of 100 (16 nM) and 200 ng/ml

significantly stimulated cell proliferation 1.5 to 2-fold under serum-free conditions in both

cell lines (Figure 6-2). When cells were exposed to EGF (100 ng/ml) and BaP together for

48 hr, the EGF stimulation of cell proliferation was still observed at 10 JM BaP, but was

significantly decreased from control at 50 gM BaP in both cell lines. A two factor

ANOVA, however, indicates that there was no significant interactive effect between BaP

and EGF on BeWo cell proliferation. In contrast, there was a significant interactive effect

of BaP and EGF on JEG-3 cell proliferation (p < 0.05). In this regard, a higher

concentration of EGF (200 ng/ml) protected JEG-3 cells, but not BeWo cells, from the loss

of EGF stimulation of cell proliferation in the presence of 50 gM BaP. This protective

effect of the higher dose of EGF in JEG-3 cells is similar to the protective effect of serum

on maintaining cell viability in the presence of 50 iM BaP (Figure 6-1).


Effects of TCDD and BaP on JEG-3 Cell Proliferation as Measured by flHlThymidine
Incorporation and Cell Number Changes

We next examined whether longer exposure to BaP and TCDD would alter JEG-3

cell proliferation in the presence of serum using the [3H]thymidine incorporation assay. A

time-dependent inhibition of cell proliferation was observed following BaP exposure over a

7 day period (Figure 6-3). The [3H] incorporation was significantly decreased by 42, 86
and 81% at 3, 5 and 7 days, respectively, following 10 PM BaP treatment in the presence

of serum. Data in Figure 6-4 show that the inhibitory effect was also concentration-

dependent following exposure to BaP for 5 days. [3H] Thymidine incorporation was









significantly inhibited by 55, 67, 73, and 76% at 1, 5, 10 and 50 iM BaP, respectively.

Similar results were observed in parallel experiments in which cell number was quantitated.

In contrast, no alterations in the [3H]thymidine incorporation and cell number was

observed following exposure to 1, 10 and 100 nM TCDD for 5 days or to 10 nM TCDD

for 1 to 7 days (Figure 6-3 and 4).


Effects of TCDD and BaP on BeWo Cell Proliferation as Measured by L3H1Thymidine
Incorporation

We also examined whether longer exposure of TCDD and BaP altered BeWo cell

proliferation. No alterations in [3H]thymidine incorporation was observed following

exposure to 10 nM TCDD for 3, 5 and 7 days or to 1, 10 and 100 nM TCDD for 5 days

(Figure 6-5). In contrast, exposure to 10 giM BaP significantly decreased BeWo cell

proliferation by 29, 49 and 46% at 3, 5 and 7 days, respectively. The inhibition was also

concentration-dependent. [3H]Thymidine incorporation was significantly decreased by 30,

46, 51 and 54% following exposure for 5 days to 1, 5, 10 and 20 JIM BaP, respectively.


Effects of TCDD and BaP on Secretion of the Hormone hCG

The next experiment evaluated whether TCDD and BaP exposure was associated

with altered trophoblast endocrine function as measured by secretion of the peptide

hormone hCG. Data in Figure 6-6 show that TCDD and BaP treatment for 48 hr

significantly inhibited basal hCG secretion by BeWo cells, but not by JEG-3 cells.

Hormone secretion by BeWo cells was reduced by 29, 40 and 30% at 1, 10 and 100 nM

TCDD after 48 hr treatment, respectively. The hCG concentration in the BeWo cell media
was significantly reduced by 41, 56 and 64% at 1, 10 and 50 tlM BaP after 48 hr

exposure, respectively (Figure 6-6). In JEG-3 cells, however, the level of basal hCG

secretion following TCDD or BaP treatment remained at a high level, as in control cells.









EGF is known to be a stimulant of hCG secretion and we next evaluated whether

the BaP-mediated loss of EGF receptors was correlated with an alteration in EGF-

stimulated hCG secretion. BeWo cells were first incubated with BaP for 48 hr, and then

treated with EGF for another 24 hr in serum-free medium. As shown in Figure 6-7, EGF

stimulated hCG secretion 3-fold in untreated control BeWo cells; the response, however,

was significantly decreased by 29 and 43% in BeWo cells pretreated with BaP at 1 and 10

gM, respectively. In contrast, in JEG-3 cells, BaP pretreatment did not alter the

stimulation of hCG secretion by EGF, with EGF at 100 ng/ml producing an appoximate 2-

fold increase in all three groups of JEG-3 cells. Thus differential effects of BaP were

observed on hCG secretion by BeWo and JEG-3 cells.


Differential Effects of TCDD and BaP on JEG-3 Cell Invasiveness


A final experiment evaluated JEG-3 cell invasiveness using the Boyden Chamber in

which cells were plated on Matrigel-coated membranes. Figure 6-8 shows that JEG-3 cell

invasion through Matrigel coated filters was significantly inhibited by BaP treatment,

whereas TCDD pretreatment significantly stimulated invasion. The number of invasive

cells was reduced by 84, 92, and 86% at 1, 10 and 20 g.M BaP after 48 hr pretreatment,

respectively (Figure 6-9). In contrast TCDD pretreatment for 48 hr increased the number

of invasive cells 1.4- and 2.7-fold at 10 and 100 nM, respectively.



Discussion

Trophoblast proliferation has been shown to be negatively regulated by TGF-3lI

and positively correlated with the level of c-Myc and EGF receptors (Gross et al., 1994;

Lala and Graham, 1990; Ohlsson, 1989). In the present study, the observed inhibition of

JEG-3 cell proliferation and invasion by BaP may well be mediated by the upmodulation of

TGF-l and the downmodulation of c-Myc and EGF receptor following BaP treatment.

Our findings may also have direct relevance to earlier observations that cytotrophoblast cell








proliferation was decreased in placentas from smokers (Sachs, 1989; Genbacev et al.,

1995). In this study, TCDD was not found to affect EGF receptor, TGF-pl, PAI-2 and c-

myc gene expression. Although the steady state TGF-a mRNA level was increased by

TCDD in BeWo cells, our effort to quantitate TGF-a protein in conditioned medium or

total cell lysate by Western analysis was not successful. Data have shown that

choriocarcinoma cell proliferation was not altered following TCDD exposure. Most

importantly, it was a major unexpected finding that JEG-3 cell invasiveness was

significantly increased by TCDD exposure, suggesting that some factors which control
trophoblastic cell invasion were altered by TCDD other than EGF receptor, TGF-a, TGF-

pl, PAI-2 and c-myc. In this regard, human keratinocytes (Gaido and Maness, 1994)

exposed to TCDD were found to have increased expression of the proteinase u-PA, which

can bind to the u-PA receptor in human trophoblast cells and lead to local proteolysis (Lala

and Graham, 1990; Strickland and Richards, 1992).

Trophoblast invasion through the uterine epithelium is promoted by proteinases that

degrade the extracellular matrix (ECM), and, conversely, limited by proteinase inhibitors

that broadly or specifically inhibit the proteinase activities (Lala and Graham, 1990;

Strickland and Richards, 1992; Cross et al., 1994). In addition, evidence indicates that the

trophoblast alters its migratory phenotype during placental development, such that

invasiveness decreases as the cytotrophoblast differentiates. The terminally-differentiated

syncytiotrophoblast is nonproliferative and noninvasive (Lala and Graham, 1990). Thus

our finding that TCDD increased JEG-3 cell invasiveness suggest that the cells may be less

differentiated following TCDD exposure.

Trophoblast proliferation and invasion are coordinatedly regulated by local

autocrine and paracrine growth control networks. Both proliferation and invasion are
downregulated by TGF-I31 that is secreted by the decidua and trophoblast (Graham and

Lala, 1991; Graham et al., 1992). However, recent data from Lala's laboratory provide

evidence that human trophoblast proliferation and invasion can be independently regulated








by locally produced growth factors. For example, insulin-like growth factor (IGF)-II,

produced by the cytotrophoblast, stimulates invasion and migration without affecting

proliferation (Lala and Lysiak, 1994; Lysiak et al., 1994a; Irving and Lala, 1995), whereas

EGF and TGF-a, secreted by the trophoblast and the decidua, do not influence invasion

but upregulate proliferation (Lysiak et al., 1993 and 1994b). These findings may partially

explain our observation that TCDD upregulated JEG-3 cell invasion without affecting cell

proliferation.

The mitogenic response of choriocarcinoma cells to EGF can be observed under

serum-free conditions, which is in agreement with the reported mitogenic action of EGF or

TGF-oa on cytotrophoblasts (Filla et al., 1993; Maruo et al., 1992). In choriocarcinoma

cells, we found that BaP inhibited both basal and EGF-stimulated cell proliferation under

serum-free conditions. The inhibitory effect of BaP on EGF-stimulated proliferation in

BeWo cells appears not to be specific since the relative percentage decrease in both groups

was similar and no significant interaction of BaP and EGF was observed by two-way

ANOVA. On the other hand, there was a significant interaction between BaP and EGF on

JEG-3 cell proliferation, suggesting that different mechanisms may be involved in the

regulation of BeWo and JEG-3 cell proliferation. TCDD has been shown to inhibit MCF-7

human breast cancer cell proliferation (Biegel and Safe, 1994) but to stimulate EGF

receptor expression and proliferation in the mouse embryonic palate and ureter epithelial

cells (Abbott and Birnbaum, 1989 and 1990a). In BeWo and JEG-3 cells, we found that

TCDD did not affect either EGF binding or cell proliferation.

An earlier clinical study found in pregnant women that serum levels of human

placental lactogen were lower in cigarette smokers than in nonsmokers (Mochizuki et al.,

1984). The present study found that the loss of EGF receptors in choriocarcinoma cells is

correlated with decreased basal and EGF-stimulated hCG secretion in BeWo cells, but not

in JEG-3 cells. The basal hCG secretion was also found to be decreased by TCDD in

BeWo cells, again not in JEG-3 cells. In this regard, recent studies with human









extravillous trophoblasts in explant culture have observed significant gestation-specific

differences in the ability of EGF to stimulate hCG production during early pregnancy

(Maruo et al., 1992; Genbacev et al., 1994). Our earlier study (Guyda et al., 1990) found

that BaP exposure inhibited EGF binding in human placental cell cultures from first

trimester, but not term placentas. Thus evidence indicates that developmental windows

exist in placental and trophoblast development for altered responses to EGF and chemical

toxicity. The differential effect of TCDD and BaP on hCG secretion observed in JEG-3

compared with BeWo cells may reflect developmental differences in the state of

differentiation of these two cell lines, with the former being more invasive and proliferative

in lower serum conditions (Grummer et al., 1994; Figure 6-3 and 5). EGF induces

differentiation of cytotrophoblasts to form syncytiotrophoblasts and to increase hCG

secretion in term placenta (Morrish et al., 1987), with hCG being able to act as an autocrine

regulator of further differentiation of cytotrophoblasts (Shi et al., 1993) by enhancing gap

junctional communication between trophoblasts (Cronier et al., 1994). Thus, our finding

that TCDD and BaP inhibited hCG secretion by BeWo cells suggests that differentiation in

this cell line may be adversely affected by both chemicals. In addition, different

mechanisms may be involved in the regulation of hCG secretion in these two cell lines, as

has been recently reported for JAR human choriocarcinoma cells (Licht et al., 1994).

In summary, BaP treatment resulted in a persistent inhibition of cell proliferation

and invasiveness, which may result from disregulation of the balance between positive and

negative regulators of trophoblast growth. TCDD exposure did not affect cell proliferation

but markedly increased JEG-3 cell invasiveness, suggesting that TCDD may directly shift

the balance between expression of proteinases and inhibitors of the proteinases which
control cell invasiveness, rather than through alterations in the expression of TGF-a, TGF-

p31, c-myc and EGF receptor. In addition, data suggest that placental endocrine function

may be affected by TCDD- and BaP-like environmental chemicals during selective periods

of placental and trophoblast development.


















14U

120-
SBeW 100 BeWo


r.00

80 ""' 'L \.
-...,,*-^ 60 -
I 50 40-
-- -- with serum ---- with serum
S .. ..... without serum 20 -. without serum
0 1 1 I 0 1 i
S 0 .001 .01 .1 1 10 100 0 .1 1 10 100
150 140-
S JEG-3 120- JEG-
T -r ,
cels wel 10p...ly,... ..fo0-120.. .. t a
"3 *100 a ^


) 60-
50- 1 40-
-- with serum with serum
.."""...- without serum ........ ........ without serum
0- i || 0
0 .001 .01 .1 1 10 100 0 .1 1 10 100
BaP (jiM) TCDD (rM)

Figure 6-1. Effects of TCDD and BaP on cell proliferation in the presence or absence of
FBS. BeWo and JEG-3 cells were cultured at densities of 5 X 103 and 2.5 X 103
cells/well, respectively, for 20 hr and treated with varying concentrations of TCDD or BaP
for another 48 hr in the presence or absence of FBS, respectively. Cell proliferation was
determined by the nonisotopic MTT assay. Proliferation of the control cells was set as
100%. Values are the mean SE of triplicate cultures from two separate experiments. The
points without the standard error bars indicate that the individual SEs are too small to be
shown. *p < 0.05 as compared to controls by Fisher PLSD and t test.













El control
m BaP 10 RM BeWo
200- BaP 50 tM
A A
150- T T
T T


100-
150 -



50


S 400 0 100 200
I-
a. E control JEG-3
l BaP 10 9tM
300-
SH BaP 50 M

200-



100 -



0-
0 100 200

EGF (ng/ml)

Figure 6-2. Effects of BaP on EGF-stimulated cell proliferation. BeWo and JEG-3 cells
were cultured at densities of 5 X 103 and 2.5 X 103 cells/well, respectively, for 20 hr and
treated with BaP and/or EGF for another 48 hr in the absence of FBS. Values are the mean
+ SE of triplicate cultures from two separate experiments. Differs from control cells (0
EGF, 0 BaP), "p < 0.05; differs from control (0 BaP) in each group, *p < 0.05 by Fisher
PLSD and t test.






86



100
-0--- control
S*............... TCDD
--0---- BaP


100



S ..........TCDD










10- ---- BaP
1-
S 1 I I

0 2 4 6 8
100

-0-- control
Figure 6-3. Tit e s of TCDD
S0. O i t -- BaP f F [

S .0""
2 1
S I/





0.1.
0 2 4 6 8
Time (d)

Figure 6-3. Time-dependent effects of TCDD and BaP on JEG-3 cell proliferation. Cells
were subcultured at 5.0 X 104 cells/well in 24-well plates for 20 hr, and treated with 10 nM
TCDD, 10 iM BaP or 0.1% DMSO in the presence of FBS. [3H]thymidine incorporation
and cell number were determined as described in Methods. Cpm values are the mean SE
of six replicate cultures; cell number values are the mean SE of six determinations from
duplicate cultures. The points without the standard error bars indicate that the individual
SEs are too small to be shown. *p < 0.05 as compared with control by t test.














150


8........O ........ BaP
0 BaP


I I



50-
I ... ...*


0 .

0 1 10 100
125
---D- TCDD
0. BaP
-100 T


75- \

...
50- -0\




0
25



0Il|
0 1 10 100
TCDD (nM)/BaP (W M)


Figure 6-4. Concentration-dependent inhibition of JEG-3 cell proliferation by BaP in the
presence of serum. Cells were treated with BaP or TCDD for 5 days, and the
incorporation of tritiated thymidine and cell number were determined as described in
Methods, with mean cpm and cell number values of control cells being set as 100%.
Values are the mean SE of three separate experiments. *p < 0.05 as compared with
control by Fisher PLSD.











A)

40


30-
o ,



20 .... .

S 10 control
.....o........ TCDD, 10 nM
--0- .... BaP, 10 RM
0
0 2 4 6 8
Time (d)
125

B)
100 .


2 75-



50-
25-- TCDD
0 .......-- aP
0- -----------

0 1 10 100
TCDD (nM)/BaP (gM)

Figure 6-5. Time-course and concentration-dependence of TCDD and BaP effects on
BeWo cell proliferation in the presence of serum. A) Time-course; B)
Concentration-dependence of cell proliferation following exposure to BaP or TCDD for 5
days. Cells were subcultured at 5.0 X 104 cells/well in 24-well plates, treated with TCDD
or BaP as indicated, and the incorporation of tritiated thymidine was determined as
described in Methods. Values are the mean SE of three separate experiments. The points
without the SE bars indicate that the individual SEs are too small to be shown. *p < 0.05
as compared with control by Fisher PLSD or t test.



















100




S 50-
8 -0-- JEG-3
....... ....... BeW o











0 --- --- ---I
0 1 10 100
TCDD (nM)
150



ST
100




50. a- ,

-0- JEG-3
........0........ BeW o
S BeWo

0 1 10 100
BaP (p M)

Figure 6-6. Effects of TCDD and BaP on hCG secretion. Cells were exposed to TCDD or
BaP for 48 hr, and the conditioned media were collected for assay of hCG levels. Results
are expressed as the mean SE of three (BaP) or four (TCDD) experiments. *p < 0.05 as
compared with controls by Fisher PLSD and t test.






















E3
0l


Control
EGF
EGF+BaP 1 gM
EGF+BaP 10 iM


BeWo JEG-3


Figure 6-7. Differential effects of BaP-pretreatment on EGF-stimulated hCG secretion.
Cells were treated with or without BaP for the first 24 hr in serum-containing medium and
then 24 hr in serum-free medium. After being washed three times with Hanks' solution,
the cells were exposed to serum free-medium with or without EGF (100 ng/ml, 17 nM) for
24 hr. hCG secretion by control BeWo (123.1 9.2 mIU/ml/24h) and JEG-3 (682.5
67.7mlU/ml/24h) cells was set as 100%. Values are the mean SE of six determinations
from triplicate cultures. *p < 0.05 as compared with untreated (0 EGF, 0 BaP) controls;
"p < 0.05 as compared with EGF alone by Fisher PLSD and t test.