“w,- -

 

 

_. . «If, ,..
2.0 9F. . "7 er,
"mm? Cu.
5- . , 4
v .
v ‘. ‘

:2; :51 .
. 1., :12?» 1.
At .\ 1.“..4 3.. ‘ , .
r .1... : . .
. ...s
. 5rd

1.::‘
p... .1
1—3.“ .

,

Vaw‘
v.

a.
I‘lv .b,lx.

fir.

fivmvrn mm“?

3....

z.
IAF .
4.4.6.. ‘
:2!
f

1 z
, fl. .. .. i
.t... ‘r’. .u .
mam.» .gguq
.YnV $331.33?

.,a...

.
:1:

«.1 .. z. .

«M

E: .
. inf?!

l, l. ‘
miuwufl Hyman
:2 3g
.5. V: I. in“?! 9. .

.1.
91...”..21. 2:

II

.Wnfli‘l

 

 

THESIS

4 LIBRARY
200« 7’71 Michigar‘. State
5" “’0‘ University

 

 

 

This is to certify that the

dissertation entitled

Functional Analysis of Estrogen
Receptor-alpha mRNA Splicing Variants

presented by

Aliccia Berg Bollig

has been accepted towards fulfillment
of the requirements for

Ph.D. . Physiology
degree in

 

 

Mm profissor

il 11, 2
Date Apr 003

 

MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRCIDataDue.p65op.15

FUNCTIONAL ANALYSIS OF
ESTROGEN RECEPTOR-a mRNA SPLICING VARIAN TS

By

Aliccia Berg Bollig

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Physiology

2003

ABSTRACT
FUNCTIONAL ANALYSIS OF
ESTROGEN RECEPTOR-a mRNA SPLICING VARIANTS
By

Aliccia Berg Bollig

Analysis of mRN A prepared from a variety of estrogen responsive cells and
tissues has established that ERa mRNA is expressed as a mixture of transcripts. This
heterogeneity results largely from a pattern of alternative mRN A splicing that gives rise
to a family of correctly processed and exon—skipped ERa mRNAs. Although variant
forms of ERa derived by alternative splicing have been identified in all ERa expressing
tissues and cell lines examined and their expression levels are well documented in
published reports, little is understood regarding their function. We have reconstructed
ERor cDNAs representing the single exon-skipped variants ERAEZ through ERAE7 in
order to analyze their activities in a well defined cell transfection system.

The aims of this thesis are two-fold. First, it addresses the hypothesis that
certain ERor mRN A splicing variants retain many aspects of wt ERa biochemical
function; and secondly it addresses the hypothesis that EROL splicing variants alter gene
expression.

Biochemical analysis of ERa mRN A splicing variants has identified two exon-

skipped variants that share significant residual function compared to their full-length

counterpart, wt ERa. The variants with sequence deletions corresponding to exon 3
(ERAE3) or exon 5 (ERAES) are observed to display normal nuclear uptake and retain
the ability to interact with transcriptional co-regulators. Furthermore, the variant
ERAE3, like wt ERa, binds ligand. These properties confer on ERAE3 and ERAES the
ability to inhibit the activity of some ERa-responsive promoters through an interplay
with wt ERa signaling, and potentially to stimulate the transcription of other genes.
Results from this study suggest that potential targets for transcriptional activation by
these ERa splicing variants include the extracellular matrix protease, human
collagenase gene, and the human IGF-l gene. The promoters of these genes lack
consensus DNA-binding sites for ERa and appear to be regulated indirectly by ERor
isoforms through interactions with other transcription factors such as AP-l. That these
genes are implicated in tumorigenesis suggests that ERa splicing variants may have an
influence on cell growth and cancer progression.

The initial discovery of ERa mRNA splicing variants challenged any previous
understanding of ER function in gene transcription and cell signal transduction. That
understanding is further complicated by the present study which contributes to evidence
suggesting that ERa splicing variants may have a significant impact on the ability of the

breast and other tissues to respond to estrogens and other regulatory signals.

To my father

iv

ACKNOWLEDGMENTS

I am grateful to the members of my graduate committee—Drs. Susan Conrad,
Kathy Gallo, Sandra Haslam, Donald Jump, and Richard Miksicek—for their helpful
suggestions throughout the course of this work. I am especially grateful to Dr. Miksicek
who is both my respected mentor and friend.

Thank you Mary Morrison and Dr. David Ankrapp for your comradeship and
insights regarding this project. I sincerely enjoyed learning from and working with both

of you in the lab.

TABLE OF CONTENTS

Page
List of Figures ......................................................................................................... viii
Key to Abbreviations ................................................................................................ x
I. Introduction .......................................................................................................... 1
II. Literature Review ................................................................................................ 4
1. Molecular Mechanisms of Estrogen Receptor Action ............................. 4
1.1 Multiple genes ............................................................................ 5
1.2 Non-genomic actions of estrogen ............................................... 9
1.3 Structural and functional domains of ERa ................................ 12
1.4 Factors influencing ERtx activity: Coactivators ......................... 18
1.5 Factors influencing ERor activity: Corcpressors ........................ 33
1.6 Antiestrogens and partial agonists ............................................. 39
1.7 Regulation of ERa function by phosphorylation ....................... 43
2. Nonconcensus Estrogen Enhancer Elements .......................................... 47
3. Identification of ERa mRNA Splicing Variants ..................................... 51
4. Activity of ERa Splicing Variants .......................................................... 55
III. Biochemical Analysis of ERa mRNA Splicing Variants ................................. 59
1. Introduction ............................................................................................. 59
2. Results ................................................................................................... 61
2.1 Measurement of the DNA-binding activity of the ERa
mRN A splincing variants .................................................... 62
2.2 ERAE3, like wt ERa, binds ligand ............................................ 63
2.3 Subcellular localization of ERa splicing variants ..................... 64
2.4 Characterization of the transactivation function of
ERa splicing variants on the vitellogenin ERE ................... 65
2.5 Dimerization and co-regulator binding properties
of ERAE3 and ERAES ......................................................... 67
3. Discussion. ............................................................................................. 83

vi

IV. ERa Splicing Variants Lacking Exons 3 or 5 Display Promoter-Specific

Transcriptional Effects Through Nonconsensus ERES ................................ 92
1. Introduction ............................................................................................. 92
2. Results ................................................................................................... 95

2.1 Like wt ERa, ERAE3 activates the ovalbumin
gene promoter ...................................................................... 96

2.2 Both ERAE3 and ERAES activate the

Coll(-73)Luc reporter ........................................................... 98
2.3 ERAES activates the human [OF -1 promoter ........................... 101
4. Discussion. ............................................................................................ 118
V. Summary and Conclusions ............................................................................... 132
VI. Materials and Methods .................................................................................... 139
1. Expression Vectors ................................................................................ 139
2. Cell Culture, Transfection and CAT Assays ......................................... 139
3. Ligand Binding Analysis ....................................................................... 141
4. DNA Binding Assays ............................................................................. 141
5. Immunoblot Analysis ............................................................................. 142
6. In Vitro Protein-Protein Interaction Assays ........................................... 143
7. Immunohistochemical and Cytochemical Analysis ............................... 144
VII. List of References ........................................................................................... 146

vii

LIST OF FIGURES

Figure Page
1. ERa functional domains .................................................................................... 16
2. Model of nuclear receptor ligand-binding domain ....................................... i ...... 17
3. Diagram of SRC-le fimctional domains ............................................................ 29
4. Schematic of the SMRT functional domains ..................................................... 38
5. Depiction of ERa activation of consensus and nonconsensus

regulatory elements ............................................................................................ 50
6. Comparison of ERor mRNA Splicing variants and wt ERa structure ................ 54
7. ERa mRNA splicing variant immunoblot analysis ........................................... 69
8. ERE-binding by wt ERor and EROL splicing variants ......................................... 71
9. Ligand-binding capacity of wt ERoc and ERoc splicing variants ........................ 72
10. Binding of nitrile THC by wt ERor and ERAE3 ................................................. 73
11. Scatchard analysis of E2 binding by wt ERa and ERAE3 .................................. 74
12. Cellular localization of wt ERa and ERa splicing variants .............................. 76
13. Activation of an ERE by wt ERa and ERa splicing variants ............................ 77
14. Inhibition of wt ERa activity by ERAE3 and ERAES ....................................... 79
15. In vitro binding of ERa isoforms and SRC-le fragments ................................. 81
16. In vitro binding of ERa isoforms and SMRT .................................................... 82
17. Depiction of ERa C-terminus secondary structure ............................................ 86
18. Activation of the Ovalbumin promoter by ERa isoforms ................................. 106

viii

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Effect of c-jun overexpression on vaalb-CAT activated expression .............. 107

In vitro binding of c-jun and c-fos with wt ERa ................................................ 108
Activity of wt ERa and ERa splicing variants on pColl(-73)Luc ..................... 109
Effect of c-jun overexpression on pColl(-73)Luc activated expression ............. 110
Conditions of treatment for ERAE3 activation of pColl(-73)Luc ...................... 112
Activity of wt ERor and ERa splicing variants on p(AP1)3TK-CAT ................ 114
Activity of wt ERor and ERa splicing variants on the IGF-l promoter ............. 115
ERAES activation of IGF-I promoter fragments ............................................... 117
Transactivation Model 1: Coactivator recruitment ........................................... 127
Transactivation Model 2: Kinase activation ..................................................... 129
Transactivation Model 3: Corepressor neutralization ....................................... 131

ix

AD 1
AD2
AF 1
AF 2
AP-l
AR
CAT
CREB
CBP/p300
Coll
DBD
E2
EGF
ER
ERa
ERB
ERAEZ
ERAE3
ERAE4
ERAES
ERAE6
ERAE7
ERE
GR
GRIP]
GST
IGF-l
JN K
LBD
Luc
MAPK
Mab-l 7
mRN A
N-CoR
NLS
Ovalb
PMA
PR
PPAR

KEY TO ABBREVIATIONS

SRC-l activation domain-1

SRC-l activation domain-2

N-terminal transactivation function-l

C-terminal transactivation function-2

Activator Protein 1, representing homo- and heterodimers of jun and fos
Androgen Receptor

Chlorarnphenicol Acetyl Transferase, a reporter gene
Cyclic AMP response element binding protein

CREB binding protein, a family of 300 kDa transcription coactivators
-73 to +63 promoter fragment from the Collagenase gene
DNA—Binding Domain

17B-Estradiol

Epidermal Growth Factor

Estrogen Receptor

Estrogen Receptor-0t

Estrogen Receptor-[3

Exon 2-skipped variant of Estrogen Receptor-alpha
Exon 3-skipped variant of Estrogen Receptor-alpha
Exon 4-skipped variant of Estrogen Receptor-alpha
Exon 5-skipped variant of Estrogen Receptor-alpha
Exon 6-skipped variant of Estrogen Receptor-alpha
Exon 7-skipped variant of Estrogen Receptor-alpha
Estrogen Receptor Element

Glucocorticoid Receptor

Glucocorticoid Receptor Interacting Protein-1
Glutathione S-Transferase

Insulin-Like Growth Factor-1

c-Jun NHz-terminal Kinase

Ligand-Binding Domain

Luciferase, a reporter gene

Mitogen Activated Protein Kinase

ER—specific Monoclonal antibody

messenger Ribonucleic Acid

Nuclear Receptor Corepressor

Nuclear Localization Signal

-1342 to +7 promoter fragment from the chicken Ovalbumin gene
Phorbol 12-Myristate, 13-Acetate

Progesterone Receptor

Peroxisome Proliferator-Activated Receptor

RIP 1 40
RIP 1 60
SMRT
SRC- 1
Tam
TBP
TIFl
TIP 2
TR
VDR
wt ERoc

Retinoic Acid Receptor

Receptor Interacting Protein-140

Receptor Interacting Protein-160

Silencing Mediator of Retinoic Acid and Thyroid Hormone Receptor
Steroid Receptor Coactivator-1
4-hydroxytamoxifen

TATA Binding Protein

Transcriptional Intermediary Factor-l
Transcriptional Intermediary Factor-2

Thyroid Hormone Receptor

Vitamin D Receptor

595 amino acid wild-type Estrogen Receptor-alpha

xi

I. Introduction

Analysis of messenger RNA (mRN A) prepared from a variety of estrogen-
responsive cells and tissues has established that estrogen receptor-a (ERa) mRN A
is typically expressed as a mixture of transcripts. This heterogeneity results largely
from a pattern of alternative mRNA splicing that gives rise to a family of correctly
processed and exon-skipped variant ERa mRNAs (Gotteland et al., 1995; Hu et al.,
1996; Miksicek et al., 1993; Murphy et al., 1997). Although there is no consistent
ratio of relative expression, wild-type (wt) and variant ERa transcripts are
coexpressed in all ERa-positive cell lines and tissues examined (Castles et al., 1995;
Friend et al., 1997; Gotteland et al., 1995; Hu et al., 1996; Kuiper et al., 1996).

There has been extensive analysis at the RNA level of the pattern of
expression and abundance of ERor splicing variants, yet limited information is
available on their functional activity. ERor mRN A comprises sequences from eight
coding exons and is translated to yield a protein with discrete functional domains
(for review, see Tsai and O’Malley, 1994). Like other nuclear receptors, ERa is a
modular protein in that individual domains are capable of demonstrating
autonomous function within receptor mutants, as well as when they are introduced
into heterologous fusion proteins (Kumar et al., 1987; Tsai and O’Malley, 1994).
One of the major aims of this thesis is to examine the ftmction of ERa splicing

variants from the vantage point of what is known about the functional organization

of wt ERa. To test the hypothesis that the mRNA splicing variants retain wt ERor
activities, the function of six ERor splicing variants that arise by the deletion of one
of its internal exons (ERAEZ through ERAE7, where the deleted exon is indicated
numerically) was characterized. Biochemical analysis demonstrates that, the
omission of a particular exon results in the loss of ERa ftmction ascribed to that
exon. The deletion of an exon is also observed to disrupt activities attributed to
other exons and to bestow novel function on the receptor isoform. Most notably,
none of the exon-skipped variants effectively promotes gene expression from a
consensus estrogen response element (ERE).

The traditional view of estrogen action dictates that hormone binding by
nuclear localized ER elicits a conformational change in the receptor allowing it to
dimerize, recruit transcription coactivators, bind DNA at an ERE and activate target
genes (Tsai and O’Malley, 1994). More recently a novel pathway for regulation of
transcription by ERa has been described to involve cooperation of the receptor with
AP-l factors such as c-jun and c-fos (Gaub et al., 1990; Uht et al., 1997). An
important distinction of this nonclassical pathway for ERa action is that functional
domains within the receptor that are crucial for promoting gene expression from an
ERE appear to be dispensable for ERa activity on AP-l directed promoters (Gaub et
al., 1990; Uht et al., 1997). This suggests a potential role for ERoc splicing variants
in regulation of nonconsensus ERES. The second major aim of this thesis is to

identify a transcription regulatory role for ERa variants. More specifically, it

investigates whether ERa variants are transcriptionally active on promoters lacking
a consensus ERE which have previously been shown to be, or are predicted to be,
estrogen responsive.

This thesis comprises seven chapters beginning with a brief topic
introduction and a statement of its major aims and hypotheses. Chapter H provides a
review of relevant literature to summarize the discovery and analysis of ERa
splicing variants in estrogen responsive tissues and cell lines. Chapter II also
reviews the literature that defines our current understanding of gene regulation by
ERa. Chapter III details the research project which focuses on the biochemical
characteristics of ERa splicing variants compared to wt ERa. This project tests the
hypotheses that selected splicing variants retain certain aspects of wt ERa function
such as DNA and ligand binding, cellular localization, and protein-protein
interactions with steroid receptor co-regulators. The research on ERa variant
regulation of transcription is described in chapter IV. This research project tests the
hypothesis that ERa splicing variants positively regulate gene transcription and
identifies promoters that may be induced by specific ERa splicing variants. Chapter
V offers a summary and completes a context of relevance for the work described in
the chapters preceding it. The final chapters, VI and VII, provide an outline of

materials and methods and a list of references, respectively.

II. Literature Review

1. Molecular Mechanisms of Estrogen Receptor Action

Estrogen exerts many biological activities by influencing gene activity in
estrogen sensitive cells. This genomic response is imparted when estrogen binds a
ligand-activated transcription factor, the estrogen receptor (ER). Binding of
estrogen to the ER elicits a change in receptor conformation that allows the receptor
to bind to DNA and enhance gene expression from the promoters of regulated genes
(Kumar and Chambon, 1988; Tsai and O’Malley, 1994). ER-induced gene
expression supports cell growth and differentiation in many tissues. For example, it
has a role in the remodeling and development of the male and female reproductive
systems (Curtis Hewitt et al., 2000; Haslam, 1987; Hess et al., 1997), influences
brain and cardiovascular function (Kuroski de Bold, 1999; Toran-Allerand et al.,
1999), and it has a role in regulating bone growth and maintenance (Compston,
2001; Grumbach, 2000).

A nuclear receptor for estrogen was first characterized in rat uterus in 1966
when specific, high affinity binding of radiolabeled l7B-estradiol (E2) to cell
extracts was shown (Tofi and Gorski, 1966). Shortly thereafter, a general
description of a mechanism of steroid hormone action was proposed. Jensen and

colleagues described a two-step mechanism that is initiated when estrogen binds and

activates a cytoplasmic receptor (Jensen et al., 1968). The activation event is not
specifically defined; however, this model continues to explain that the activated
hormone-receptor complex is then translocated to the nucleus and binds chromatin
to cause a change in mRN A expression. Since this mechanism of steroid action was
first proposed, research has corrected and elaborated upon the details of this model
so that we now know that the ER is associated primarily with the cell nucleus in an
unliganded state (King and Greene, 1984; Welshons et al., 1984), and that estrogen
is an allosteric effector for the receptor. Ligand-bound ER forms homodimers
which bind to DNA at a specific nucleotide sequence called an estrogen response
enhancer element (ERE) for which the Xenopus vitellogenin A2 promoter ERE
sequence (5 'GGTCACA-GTGACC-3 ') serves as a paradigm (Klein-Hitpass et al.,
1988; Kumar and Chambon, 1988).

Continued research has taken the current understanding of ER-directed gene
expression beyond simply correcting the details of Jensen’s model. Investigators are
now confronted by a regulatory mechanism with unforeseen complexity and new

uncertainties.

1.1 Multiple genes

The transcriptional effects of estrogens are mediated by two closely related
receptor isoforms, ERa and ERB, each encoded by a separate gene. The human
ERa gene, which was first cloned by Chambon and colleagues in 1986 (Green et al.,

1986), is localized to the long arm of chromosome 6 (6q23) where it spans

approximately 450 Kb of genomic sequence (Gosden et al., 1986; Manasce et al.,
1993). The location of the more recently described human ERB gene has been
mapped to chromosome 14 (more specifically, l4q22-24) (Enmark et al., 1997;
Mosselman et al., 1996). Analysis of mRNA has determined that there is a wide
tissue distribution for both ERa and ERB, and mRNA splicing variants for each
isoforrn have been identified (Chu and Fuller, 1997; Enmark et al., 1997;
Mosselman et al., 1996; Murphy et al., 1997; Poola et al., 2002; Price et al., 2001;
Stabile et al., 2002). The variety of ERor splicing variants include mRNA variants
with a deletion of one exon, including exons 2, 3, 4, 5, and 7, as well as
combinations of exons (Miksicek et al., 1993; Murphy et al., 1997). Single exon-
skipped variants of ERB include mRNA isoforms missing exons 2, 3, 4, 5, and 6
(Chu and Fuller, 1997; Poola et al., 2002). Variants of ERB also include receptors
expressing an insert of 54 nucleotides within the ligand binding domain and another
ERB mRNA splicing variant that harbors both the exon 3 deletion and the 54 base
pair insert has also been reported (Price et al., 2001).

The ERa and ERB cDNAs share a great deal of identity, particularly in the
sequences corresponding to those regions of the receptor that bind ligand and DNA
(59 percent and 97 percent, respectively) (Enmark et al., 1997). The resulting
proteins share similarities in structure and in their function to regulate transcription.
Both ERa and ERB homodimerize in the presence of E2 (Ogawa et al., 1998). In

vitro and in viva binding assays suggest that ligand-bound ERa and ERB are

colocalized within the nucleus and form heterodimers (Matsuda et al., 2002; Ogawa
et al., 1998). The mouse homologs of ERa and ERB are shown to bind E2 with
similar affinity (although a slightly lower average dissociation constant is reported
for ERB) (Tremblay et al., 1997). Moreover, Ez-liganded ERa and ERB dimers
bind and activate a consensus ERE with similar effectiveness. This transcriptional
activity can be blocked by treatment with antagonists [CI 164, 384 and 4-
hydroxytamoxifen (tamoxifen) (Mosselman et al., 1996; Paech et al., 1997).
Tissues that are sensitive to estrogen generally express both isoforms of the
receptor; however, the mRNAs are often present at unequal ratios and some tissues
offer examples of ERa- and ERB— differential expression (Enmark et al., 1997,
Mosselman et al., 1996). Human ERa and ERB mRNA are coexpressed in many
tissues, including uterus, breast, lung, kidney, ovary and prostate. Examples of cells
that express only one of the receptor isoforms include human granulosa cells which
express only ERB mRNA and leydig cells of the human testis which only contain
ERa (Enmark et al., 1997). Comparing the ratio of expression in tissues that
coexpress the two isoforms Enmark et al. report that ERa is more abundant than
ERB in the human uterus and ERB predominates in the breast, however quantitative
analysis of ERoc and ERE mRNA in normal and breast cancer tissues suggests that
ERa is the dominant transcript in tumors and ERB expression is down regulated
during carcinogenesis (Enmark et al., 1997; Gustafsson and Warner, 2000; Iwao et

al., 2000). A study of relative protein levels in benign and malignant breast biopsies

by western blotting confirms that ERB is more abundant than ERa in the normal
human mammary gland samples. Although very little ERa was detected in the
normal mammary gland, it was strongly detected in invasive tumor samples
(Gustafsson and Warner, 2000).

Additional distinctions between ERa and ERB are revealed when fertility
and breast development are examined in mice with disrupted genes. ERa knockout
mice are infertile and, compared to wild-type mice expressing both ERa and ERB,
ERB knockout mice have decreased fertility and fewer and smaller litters (Krege et
al., 1998; Lubahn et al., 1993). In ERa knockout mice uterine weight is reduced
compared to wild-type littermates and the uteri of adult ERa knockout mice do not
undergo cyclic changes in response to treatment with ovarian steroid hormones
(Couse et al., 1995; Korach 1994; Lubahn et al., 1993). In contrast, the uteri of
adult ERB knockout mice resemble the wild-type phenotype (Krege et al., 1998;
Lubahn et al., 1993). Embryonic and fetal development of the mammary gland
occurs in the absence of either receptor, however, in ERor knockout mice
postpubertal ductal growth is absent. Like wild-type counterparts, ERB knockout
mice develop a normal ductal structure that fills the entire fat pad (Couse et al.,
2000; Gustafsson and Warner, 2000). Together with the relative abundance of
receptor isoforms observed in normal and malignant breast biopsies these findings
suggest that ERa and ERB each offer a unique contribution to estrogen

responsiveness in the breast. Further consideration and additional studies that

consider the relative ratio of expression of ERa and ERB may offer greater insight
into both normal breast development and the advancement of cancer.

Notable differences between ERa and ERE are also evident when their
capacity to effectively activate various promoters is compared. In agreement with
earlier studies, Paech and coworkers have shown that with E2 treatment ERa and
ERB were equally effective at activating an ERE-directed reporter gene construct in
HeLa cells (an ER-negative human cancer cell line of uterine origin). Treatment
with antiestrogens raloxifene, tamoxifen and ICI 164384 did not support this
expression. When their activities were compared on an AP-l response element
treatment with raloxifene, tamoxifen and ICI 164,384 induced gene expression from
an AP-l -directed reporter gene in HeLa cells cotransfected with ERB. E2 treatment
of ERB-expressing cells had no effect. Distinctively, E2 treatment induced
ERa-dependent gene expression while antiestrogen treatment had relatively little
effect on AP-l reporter gene activity in cells cotransfected with ERa ( Paech et al.,

1997)

1.2 Non-genomic actions of estrogen

While most of the biological effects of estrogens involve programmed
changes in gene expression that are mediated by receptors located within the
nucleus, rapid responses to estrogens have also been described that appear to
originate at the cell membrane and to be independent of gene transcription (for

review see, Coleman et al, 2001). These signaling events are considered to be non-

genomic and can be initiated by estrogen analogues that do not traverse the cell
membrane. BSA-conjugated B; does not enter the cell yet has been shown to bind
specifically at the cell membrane and within minutes (a time frame too rapid to
involve gene transcription) can cause an increase in intracellular signaling molecules
such as cAMP, calcium and inositol-l ,4,5-triphosphatc (1P3) in many cell types
(Aronica et al., 1994; Coleman et al., 2001; Fiorelli et al., 1996; Guo et al., 2002,
Improta-Brears et a1 ., 1999). The putative membrane receptor for estrogen has not
been fully characterized, however a study using ER-negative Chinese hamster ovary
(CHO) cells to transiently express ERa and ERB demonstrated that a small
percentage of BRO: and ERB were translocated to the membrane (Razandi et al.,
1999). Studies using membrane preparations from these ERa- and ERB-expressing
CHO cells indicate that membrane localized ERa and ERB are coupled with G-
proteins. Treatment with E2 stimulated Ca2+-dependent mitogen-activated protein
kinase (MAPK) activity and Gaq-dependent production of 1P3, both effects were
blocked by co-treatment with the ER antagonist ICI 182, 780 (Improta-Brears et. al.,
1999; Razandi et al., 1999).

Rapid stimulation of PKC activity by estrogen in primary cultures of female
rat growth plate chondrocytes is also attributed to a membrane localized receptor
and does not involve alteration of gene expression (Sylvia et al., 2000). Sylvia and
colleagues demonstrated that treatment of cultures with 17B-E2 produced a dose-
dependent increase in PKC activity which was mediated by G-protein-coupled

phospholipase C. PKC activity was unaltered by treatment with the ER agonist

10

diethylstilbesterol (DES) and co-treatment with [Cl 182, 780 did not block the effect
of 17B-E2, suggesting that Ez-dependent PKC activity in these cultures may not be
directed by either of the classical ERor or ERB receptors (Sylvia et al., 2000).
Increased release of the cell signaling molecule nitric oxide (NO) is
stimulated in vascular endothelial cells when Ez-liganded ERa activates
phosphatidylinositol-3-OH kinase (PI3-K). P13-K activates protein kinase B/Akt
(PKB/Akt) which subsequently activates endothelial nitric oxide synthase (eNOS)
(Hisamoto et al., 2001; Rameh and Cantley, 1999; Simonicini et al., 2000). ERa-
associated PI3-K activity is ligand-dependent and sensitive to antiestrogen
treatment. Co-transfection experiments using CHO cells determined that E;-
stirnulated PKB/Akt activity in these cultures was dependent on transfected ERor.
Notably, E2 treatment did not enhance PKB/Akt activity in cells transfected with
ERB (Hisamoto et al., 2001). Studies reported by Simonicini et al., suggest that the
activation of PI3-K by ERor is in part the effect of recruitment and binding of PI3-K
by ERor localized in the cytosol. ERor interaction with the PI3-K regulatory subunit
p850r is ligand-dependent and not supported by ICI 182, 780 binding. Consistent
with its inability to activate PI3-K, ERB did not complex with p850: (Simoncini et
a], 2000). Unlike ER-dependent PI3-K activation, both ERa and ERB mediate the
activation of MAPK in a number of cell types including the human breast tumor cell
line, MCF7, in the presence of E2 (Improta-Brears et al., 1999; Migliaccio et al.,

1996) However, although co-treatment with ICI 182, 780 or tamoxifen inhibits

ll

ERor-directed MAPK activation, ERE-directed activation is stimulated by
antiestrogen treatment (Wade et al., 2001 ). This is not an entirely unique effect as
the ERor-dependent increase in intracellular cAMP in breast cancer and uterine cell
lines was also found to be stimulated by both estrogens and antiestrogens (Aronica
et al., 1994). Further complicating our understanding of ER regulation is the
knowledge that CAMP acts as a second messenger which signals a PKA-dependent
binding of ERor and cyclin D1, an association which signals ligand-independent

activation of ERE—directed gene transcription (Lamb et al., 2000).

1.3 Structural and functional domains of ERor

To understand the mechanisms by which ERor transactivates a promoter it is
useful to break the receptor down and describe its individual functional domains.
Like other nuclear receptors, BRor is a modular protein that has discrete functional
domains capable of demonstrating autonomous activity when they are expressed in
receptor mutants as well as when they are introduced into heterologous fusion
proteins (for review, see Beato and Sanchez-Pacheco, 1996; Kumar et al., 1987).
ERor is translated from mRNA with eight coding exons (Ponglikitmongkol et al.,
1988). An N-terminal transactivation function (AF 1) encoded by exon 1 and a
portion of exon 2 is thought to promote gene expression by interacting with and
recruiting other transcription factors (e.g., nuclear receptor coactivators) to the
promoters of responsive genes (Endoh et al., 1999; Kobayashi et al., 2000; Metivier

et al., 2001; Metivier et al., 2000; Metzger et al., 1995; Ofiate et al., 1998; Sathya et

12

al., 2002; Tremblay et al., 1999; Watanabe et al., 2001; Webb et al., 1998). Also,
AF 1 directly binds, and may thereby stabilize, basal transcription factors TFIIB,
TATA-binding protein (TBP), and the TBP-associated factor TAFn30 (Ing et al.,
1992; Jacq et al., 1994; Sadovsky et al., 1995). Derived fiom exons 2 and 3 is a
centrally located zinc-finger motif (commonly referred to as the DNA-binding
domain or DBD) that is essential for sequence-specific DNA binding and
transcriptional activation through a consensus ERE (Kumar et al., 1987). A receptor
dimerization interface is also located within this domain, specifically between the
first two cysteines of the second zinc finger (Schwabe and Rhodes, 1991). Within
the region encoded by exon 4 are the nuclear localization signals (NLS) and a hinge
region that allows for receptor conformation flexibility (Tsai and O’Malley, 1994;
Ylikomi et al., 1992). A ligand binding domain (LBD) confers regulatory function
to the receptor and is encoded by the C-terminal exons 4 through 8 (Kumar et al.,
1986). The ERor LBD structure resembles the crystal structures of other nuclear
receptor LBDs (Brzozowski et al., 1997; Nolte et al., 1998; Tanenbaum et al.,
1998). An arrangement of receptor a-helices numbered H3 to H12 contribute to the
ERor LBD structure. The ligand-binding pocket is formed principally by H3, H6,
H8, and H12. In the unliganded state H3, H6 and H8 form the interior cavity and
H12 extends away from this domain. When E2 binds the cavity it makes direct
contact with critical interior residues and effects a conformational change that
causes H12 to fold and contact H3, H5 and H11. As depicted in Fig. 2, this

orientation of H12 bars the cavity entrance and locks the hormone in the binding

13

pocket (Brzozowski et al., 1997; Chen et al., 1996). Within the LBD are binding
sites for heat shock proteins such as Hsp90, Hsp40, Hsp56, and Hsp70 which form a
heterologomeric complex with ER. Heat shock proteins bind to the nascent ER and
in concert with accessory proteins (e. g. p23, hip, hop, and CyP40) assemble to form
a dynamic multicomponent complex which promotes proper protein folding and
functions to maintain the mature receptor in a conformation which facilitates
receptor interaction with hormones (Howard et al., 1990, Kimmins et al., 1999).
The region encoded by exons 4 through 8 also includes additional regulated
determinants for subunit dirnerization, and a well-characterized C-tenninal
transactivation function (AF 2) (Kumar and Chambon, 1988). In the ligand-bound
state, AF 2 adopts a conformation that supports the association of ERor with factors
that are known to cooperate with the receptor to promote gene transcription (e.g.,
coactivators or other transcription factors) (Nolte et al., 1998; Tanenbaum et al.,
1998). With this association, ERor may act to enhance gene expression by ushering
additional transactivators to ERE-containing promoters. When the receptor is
unliganded or binds an antagonist, AF 2 does not bind transcription promoting
factors and may instead interact with proteins (generally referred to as corepressors)
that repress transcription (Nagy et al., 1999). Unlike AF 2 activity, the N-terminal
AFl activity is considered to act independently of ligand binding (Tora et al., 1989);
i.e., it is constitutively active and its function is generally not influenced by the
binding of antagonists that are otherwise shown to hinder AF 2 activity (Berry et al.,

1990; Metivier et al., 2001). The two activation domains function independently

l4

when introduced into heterologous fusion proteins and may fimction independently
in the context of the full-length receptor on certain promoters in a cell type-specific
manner (Berry et al., 1990; Tora et al., 1989); yet it is clear that for many estrogen-
sensitive promoters and cell types full transcriptional activity is the sum total of the
activities of both transactivation domains (Berry et al., 1990; McInery et al., 1996;

Sathya et al., 2002; Tora et al., 1989).

15

Hinge Region

 

 

 

 

AF‘l _
1 38 116 184 263 302 549 595“
V
l I [080 NLS w\'
A B

Fig. 1. A representation of ERor functional domains. The ER functional domains
and A-F regions are arranged as they reside along the length of the receptor.

Relative positioning of the nuclear localization signal (NLS), DNA-binding domain
(DBD), ligand binding domain (LBD), AF 1 , and AF 2 domains are depicted.

l6

 

Fig. 2. Three dimensional model of a nuclear receptor ligand-binding domain

with and without bound estradiol. Left, in the unliganded state helix 12 (H12)
extends away from the opening of the ligand-binding cavity which is formed by
helices H3, H5, and H11. Right, when estradiol binds to the cavity it effects a
conformation change that causes H12 to fold and contact H3, H5, and H11. This
folded orientation of H12 bars the cavity entrance and locks the hormone (shown

in green) in the binding pocket.

From L. P. Taylor and H. Akil. NRR Graphics Library,
http://biocheml . basic—sci. georgetown. edu/nrr/glib. htmI

1.4 Factors influencing ERor activity: Coactivators

Binding of estradiol to ERor elicits a conformational change that promotes
receptor dirnerization and translocation to an ERE in the nucleus of ERor-expressing
cells. However, the expression of ER and the accessibility of an ERE in a target cell
are not the sole requirements for an efficient genomic response. The process of
hormone binding and the resulting change in conformation also promotes an
interaction with proteins that enhance or deter ERor activity. The profile of ERor
transcriptional activity has been recently modified to incorporate a role for two new
classes of proteins, coactivators and corepressors. Coactivators function to enhance
the activity of ligand-bound nuclear receptors, and corepressors associate with
nuclear receptors in the unliganded state to repress activity (Nagy et al., 1999).

The existence of a coactivator-type protein was first suspected when
investigators observed a squelching of ERor activity in cells overexpressing the
receptor (Meyer et al., 1989). Transcriptional interference suggested that the
receptors were competing for a limiting factor that was fundamental to receptor
activity. This notion was ultimately reinforced when proteins were identified that
were shown to directly bind to ERor and enhance ERE-directed reporter gene
expression (Cavailles et al., 1994; Ofiate et al., 1995; Voegel et al., 1996). The
nature of coactivator involvement in nuclear receptor-mediated transcription was
firrther elucidated when over expression of a dominant negative isoform of a single
coactivator was shown to effectively block ERor-enhanced reporter gene expression

(Ofiate et al., 1995). In 1994 Malcolm Parker and colleagues described two murine

l8

proteins of 160 and 140 kDa (referred to as receptor interacting proteins, RIP160
and RIP140, respectively) that associated directly with mouse ERa in the presence
of E2 (Cavailles et al., 1994). Later it was shown that these particular proteins
enhanced ERor-directed transcription in reporter gene expression experiments
(Cavailles et al., 1995; Voegel et al., 1996). Since these proteins were
characterized, the list of coactivators has grown significantly to include, among
others, a group of “pl 60” nuclear receptor coactivators (NCOAI , NCOA2, and
NCOA3), the CBP/p300 and P/CAF proteins, and the multirneric SWI/SNF and
TRAP/DRIP/ARC complexes (Freedman, 1999; Janknecht and Hunter, 1996;
McKenna et al., 1999; Muchardt and Yaniv, 1993; Naar et al., 1999; Rachez et al.,
1999). The p160 coactivators from several species have been cloned and
characterized by a number of laboratories, giving rise to a confusing nomenclature.
For consistency, NCOAI (also called SRC-l , RIP160, and TIFl) will be referred to
by its more common name, SRC-l , while NCOA2 (TIF2/GRIP1/SRC-2) and
NCOA3 (ACTR/AIBl/RACB/p/CIP) will be referred to as TIF2 and ACTR,
respectively. In general, coactivators enhance nuclear receptor transcriptional
activity in the presence of an agonist and do not associate with antagonist-bound
receptor. Coactivator frmction requires recruitment to the promoter by the receptor
or by other DNA-binding transcription factors. The growing list of coactivators and
similarities in their behavior suggest that they are functionally redundant proteins.
However, continued research hints that there are preferences among the various

nuclear receptors (NR5) for individual coactivators (Glass and Rosenfeld, 2000;

19

Ding et al., 1998; Kalkhoven et al., 1998; Voegel et al., 1996).

There are three classes of coactivators distinguished by the nature of their
intrinsic activities which appear to be critical for their role in nuclear receptor-
directed gene transcription (for review, see Glass and Rosenfeld, 2000). The first
class of coactivators includes mammalian homologues of the yeast SW12/SNF2
factor named BRGl and hBrm (Laurent et al., 1993; Khavari et al., 1993; Muchardt
and Yaniv 1993). SWI/SNF functions as a multiprotein coactivator complex and
has DNA-stimulated ATPase activity (Laurent et al., 1993; Owen-Hughes et al.,
1996). SWI/SNF complexes are recruited to promoter elements by interactions with
sequence specific DNA-binding proteins (including ER). With the hydrolysis of
ATP they effect a local remodeling of chromatin structure that facilitates binding of
additional transcription regulatory factors and correlates with enhanced gene
transcription (Chiba et al., 1994; Kingston and Narlikar, 1999; Laurent et al., 1993).
A second class of coactivators are known to possess a histone acetyltransferase
(HAT) functional domain. Examples of coactivators with HAT activity include
SRC-l , ACTR, P/CAF, and CBP/p300. As with the SWI/SNF complex,
translocation of these coactivators to DNA allows their intrinsic HAT activities
function, remodeling chromatin and facilitating the binding of additional
transcriptional regulatory proteins to specific DNA sequences (Freedman, 1999;
Ogryzko et al., 1996; Spencer et al., 1997; Wolffe, 1994; Yang et al., 1996c). The
third class of coactivators is characterized by the TRAP/DRIP/ARC complex.

TRAP/DRIP/ARC subunits assemble to form a multiprotein composite coactivator

20

that binds both nuclear receptors and other coactivators (Fondell et a1, 1999; Glass
and Rosenfeld, 2000; K0 et al., 2000; Nair et al., 1999; Rachez et al., 1999). It is
unclear how TRAP/DRIP/ARC functions to enhance NR—dependent transcription.
The capacity for TRAP/DRIP/ARC to interact with NRs and potentially recruit other
coactivators may in part define their role in NR-directed gene transcription. The
TRAP/DRIP/ARC complex does not have intrinsic HAT activity, however, it
associates with other coactivators including CBP/p300 which do have HAT activity
(Ko et al., 2000).

The p160 coactivators (SRC-l , TIF2, and ACTR (also referred to as AIBl,
RAC3, or p/CIP) share similarities in structure, activation function, and receptor
binding properties (Freedman, 1999) and have received much attention in the
literature. To date, the relationship between coactivator and ERor activity is best
understood within the context of SRC-l , which was the first nuclear receptor
accessory factor reported to efficiently stimulate the AF 2 activity of ERor (Ofiate et
a1 ., 1995).

In a study designed to assess direct interaction between coactivators and
nuclear receptors, researchers observed varying affinities for interaction between the
nine nuclear receptors and two coactivators examined (Ding et al., 1998).
Differences are especially notable in the in vitro preference for TIF2 binding by the
androgen receptor (AR) and progesterone receptor (PR) compared to the relatively
weak affinity these receptors exhibited for SRC-l (Ding et al., 1998). Comparing

TIF2 and SRC-l activities, results fi'om reporter gene experiments suggest that TIF2

21

enhances transactivation by AR, ER, and PR, but not by the glucocorticoid receptor
(GR), retinoic acid receptor (RAR), thyroid hormone receptor (TR) and vitamin D
receptor (V DR), whereas SRC-l associates with and stimulates the AF 2 activity of
all nuclear receptors (Ding et al., 1998; Nolte et al., 1998; Ofiate et al., 1995;
Voegel et al., 1996). In more comprehensive studies SRC-l activity was tested on
promoters responsive to transcription factors other than nuclear receptors.
Coexpression of SRC-l does not affect cyclic AMP response element-binding
protein (CREB) activity on a CREB-responsive promoter in reporter gene assays
unlike the CBP/p300 coactivator. Therefore, although SRC-l has been shown to
enhance transcription by nuclear receptors, it is not a general coactivator for all
inducible transcription factors (Freedman et al., 1999; Nestin et al., 2000; Oflate et
al, 1995). In contrast, CBP/p300 has been shown to bind both NRS and CREB,
allowing it to function as a signal integrator (Kamei etal., 1996)

Three human SRC-l isoforms have been cloned (Hayashi et al., 1997;
Kalkhoven et al., 1998). The isoforms are encoded by the same gene and have an
identical amino acid sequence up to residue 1385. The SRC-la C-terminus extends
forty-two amino acids longer than SRC-le and the last fourteen C-terminal amino
acids of SRC-le (1385-1399) are not conserved in SRC-la (Kalkhoven et al., 1998).
SRC-lq differs fiom SRC-la by deletion of one C-terminal glutamine residue
(amino acid 1386 of SRC-la) (Hayashi et al., 1997). It has been shown that SRC-
la, SRC-le and SRC—lq function similarly and all three share a capacity to bind

NRs and enhance NR-induced reporter gene transcription (Hayashi et al., 1997;

22

Kalkhoven et al., 1998; Ma et al., 1999; Ofiate et al., 1995). Two reports suggests
that SRC-le may be more effective than SRC-la and SRC-lq in coactivating NR
transcription (Hayashi et al., 1997; Kalkhoven et al., 1998).

Binding and transactivation studies suggest, and studies of co-crystal
structures strengthen the argument, that the conformational change induced by an
agonist is required for formation of an active nuclear receptor/SRC-l complex
(Darimont et al., 1998; Ding et al., 1998; Nagy et al., 1999; Nolte et al., 1998;
Ofiate et al., 1995). The necessity for SRC-l activity in vivo is exemplified in
studies of ovariectomized mice with a disrupted SRC-l gene. Xu and coworkers
measured a decrease in estrogen-induced uterine growth in ovariectomized SRC-l -
null mice compared to heterozygous mutants and wild-type counterparts (Xu et al.,
1998). These authors also report that the endocrine feedback control system and
mammary gland proliferation and differentiation in response to estrogen and
progesterone treatment are disrupted in homozygous SRC-l mutants (Xu et a1 .,
1998). It is important to stress that, although stunted, there was uterine and
mammary gland development in these SRC-l -mutant mice with hormone (estrogen
and progesterone) treatment and that both male and female homozygous mutants
were fertile (Xu et al., 1998). Results fiom these studies suggest that SRC-l is
important for steroid hormone-mediated gene transcription, since mice lacking SRC-
1 exhibit partial resistance to E2 treatment. Conversely, these results also
substantiate the notion that one coactivator may, to some extent, compensate for

another.

23

As noted above, coactivator-like accessory factors were first hypothesized
when receptor activity squelching was observed in transient transfection studies
(Meyer et al., 1989). SRC-la over expression is shown to reverse the squelching of
PR—directed gene expression by cotransfected ligand-bound ERor in some contexts,
in a dose-dependent manner (Lopez et al., 1999; Ofiate et al., 1995). This
observation is tempered by reports of other efforts to study the limiting influence of
SRC-l which indicate that over expression of SRC-l may itself attenuate reporter
gene expression (Lopez et al., 1999). Studies suggest that the degree of coactivator
influence on receptor activity is cell type-specific, and under certain transfection
conditions using a range of exogenous coactivator expression levels, the presence of
too little or too much coactivator may be equally ineffective at potentiating receptor
activity or alleviating receptor activity squelching (Cavailles et al., 1995; Kamei et
al., 1996; Lopez et al., 1999). These results led some researchers to investigate
whether or not SRC-l can titrate out another factor that in turn becomes limiting. In
separate transient transfection experiments it has been shown that p300/CBP over
expression can enhance cooperative SRC-l/ERa activity, and under different
conditions, it can alleviate SRC-l attenuation of ERor-directed reporter gene
expression (Kamei et al., 1996; Lopez et al., 1999; Smith et al., 1996).

When considering the mechanism by which SRC-l enhances ERa activity it
is necessary to reiterate that evidence from binding studies suggests that SRC-l acts
by interacting directly with ERa and is recruited by the receptor to the promoter

where it modulates gene expression. This puts the coactivator at the scene of

24

transcription initiation, but it does not fully explain how this complex enhances
receptor activity. Two modes of activity are thought to contribute to SRC-l activity.
First, SRC-l associates with other coactivators (e.g., CBP/p300) (Yao et al., 1996)
and it directly binds the basal transcription factors TFIIB and TBP (Takeshita et al.,
1996). It may in effect be recruiting other necessary factors to the ERor-responsive
promoter for efficient transcription and acting to stabilize the transcription initiation
complex. Secondly, SRC—l is shown to have HAT activity, a function that, as
described earlier, correlates with remodeling of the chromatin structure and
increased gene transcription (Spencer et al., 1997; Wolffe, 1994).

An (Jr-helical receptor-binding motif with the sequence LXXLL is conserved
among many of the coactivators and is sufficient for ligand-dependent nuclear
receptor binding (Heery et al., 1997). Site-directed mutagenesis of residues in this
motif reduce or abolish interaction with nuclear receptors (Heery et al, 1997;
Kalkhoven et al., 1998). Three of these sequences are localized to the central region
of all SRC-l isoforms and one additional LXXLL motif is present in the divergent
C-terminus of SRC-1a(Kalkhoven et al., 1998). Each of them contributes to
interaction with ERor, but mutational studies demonstrate that maintaining the
integrity of just the second of the three central motifs is sufficient for in vitro ERor
binding. This observation, and reports which suggest that there are different
affinities among a variety of receptor/coactivator complexes that interact via this
binding sequence, indicate that motif location and the context of the residues

surrounding this motif may influence receptor binding preference (Ding et al., 1998;

25

Kalkhoven et al., 1998; Voegel et al., 1996).

The binding of SRC-l can be mapped to a conserved C-terminal (Jr-helix in
the LBD of nuclear receptors (Danielian et al., 1992; Darimont et al., 1998; Nolte et
al., 1998; Shiau et al., 1998; Tanenbaum et al., 1998). When a nuclear receptor
binds its cognate ligand, receptor helix 12 folds to expose a surface to accommodate
SRC-l binding. Crystal structures of the LBDs of multiple nuclear receptors,
including ERor, indicate that helix 12 extends outwardly from the body of the
unliganded receptor. When the receptor binds estrogen, helix 12 is tightly packed
against the LBD and makes direct contact with the estrogen molecule (Darimont et
al., 1998; Nolte etal., 1998; Shiau et al., 1998; Tanenbaum et al., 1998). This
conformation brings a conserved glutamic acid and lysine residue to the receptor
surface in an orientation that allows them to form hydrogen bonds with the amino
and carboxy-terminal ends, respectively, of the contacting SRC-l LXXLL motif.
The length of the LXXLL helix is accommodated by a hydrophobic clefi which
forms between these specified glutamic acid and lysine residues (Darimont et al.,
1998; Nolte et al., 1998; Shiau et al., 1998). Additional contacts occur between
coactivators and helices 4 and 5 of the LBD. Coactivator binding is impaired by
mutations in this region, and binding of antagonists such as raloxifene to ERor do
not support the appropriate conformational change to expose the receptor SRC-l
docking site (Barettino et al., 1994; Brozozowski et al., 1997; Danielian et al., 1992;
Durand et al., 1994). Co-crystal structure analysis indicates that the two AF 2

domains of a nuclear receptor dimer contact separate LXXLL motifs of a single

26

SRC-l coactivator molecule (Nolte et al., 1998).

In addition to the well characterized C-terminal interaction, SRC-l also
associates with an N-terminal region of ERa. In most contexts, the synergistic
activity of AF 1 and AF 2 is required for maximal Ez-stimulated activity on an ERE;
i.e., the full-length receptor, with both AF 1 and AF 2 intact, activates ERE-directed
gene transcription more effectively in many cell types than mutated counterparts
lacking one of the activation functions (McInery et al., Metivier et al., 2001; 1996;
Tora et al., 1989). The first convincing data that SRC-l interacts with both AF 1 and
AF 2 came from experiments in which truncation mutants corresponding to the AF 1 -
containing N-terminus and AF2-containing C-terminus of ERor were expressed in
the presence of SRC-l (Kraus et al., 1995). When both receptor mutants were
simultaneously expressed, the effect on reporter gene activation was synergistic in
the presence of SRC-l compared to either receptor mutant expressed individually
alone or in combination with SRC-l (McInery et al., 1996). Considering that it has
been shown that SRC-l binds the ERor N-teminus in vitro (Metivier et al., 2001;
Ofiate et al., 1998), it is plausible that SRC-l and perhaps other coactivators, as
well, are simultaneously binding both the N- and C-terminus mutants and thereby
integrating the ligand-independent AF 1 fimction and Ez-induced AF 2 activity to
fully activate the reporter gene in these studies. This model has received
considerable experimental support and is believed to account for much of the cell-
type specificity that has been observed for AF] function (Benecke et al., 2000;

Kobayashi et al., 2000; Metivier et al., 2001).

27

As mentioned above, members of the p160 class of coactivators have an
acetyltransferase catalytic activity and multiple nuclear receptor interaction motifs.
They also have intrinsic activation domains that correlate with binding sites for
other coactivators (Koh et al., 2001; Ofiate et al., 1998; Sheppard et al., 2001). The
centrally located SRC-l activation domain 1 (ADl) binds the homologous CBP and
p300 factors (Sheppard et al., 2001). Each of these are themselves coactivators with
intrinsic HAT activity that interact with other activating proteins such as P/CAF
(Ogryzko et al., 1996; Yang et al., 1996c). A C-terminal activation domain (AD2)
within the p160 coactivators binds coactivator-associated arginine methyltransferase
1 (CARMl) (Koh et al., 2001). Recruitment of CARMI to the transcription
intitiation complex by a nuclear receptor coactivator, and its ability to synergistically
enhance steroid receptor-mediated transcription in the presence of coexpressed pl 60
coactivator, renders CARMl a secondary cofactor (Chen et al., 1999; Koh et al.,
2001). The same mutation that eliminated the ability of CARMl to methylate
histone H3 in vitro also reduced CARMl ’s ability to enhance transcriptional
activation in reporter gene experiments (Chen et al., 1999). It appears that
CARMl ’s ability to selectively methylate histone H3 coincides with enhancement of

steroid receptor-mediated transcription.

28

1| % i b;% , -ADz - 1 399 aa

570-780
989-1 240

 

 

     

 

 

 

 

 

 

 

781 -988

 

1 241 -1 399

Fig. 3. Diagram of SRC-le functional domains. The location of the nuclear
receptor (NR) binding domains are indicated by hatched boxes. The activation
function domains are represented by black boxes. The SRC- 1 e amino acid

coordinates are indicated under the diagram.

29

The knowledge that some coactivators possess different activities does little
to help our understanding of why there are so many factors that frmction as NR
coregulators. Many coactivators share the same functional domains (e.g., both SRC-
1 and ACTR have HAT activity) and furthermore it is unclear why multiple
coactivator activities are even necessary. These uncertainties will be addressed
when the mechanism by which coactivators potentiate NR-directed transcription is
better understood. In a review of coregulator function, Glass and Rosenfeld describe
three models to explain how multiple coactivators might have a role in regulating
NR transactivation (Glass and Rosenfeld, 2000). The first model proposes that
various coregulatory complexes are recruited in a sequential fashion such that one
coactivator may initially bind a NR and subsequently be exchanged for another. The
binding of each successive coactivator may present a different activity that is
required in a particular chronological order to effectively enhance transcription. A
second model speculates that various coactivators are simultaneously recruited by
NRs to a promoter. This model suggests that proteins from each of the functionally
distinct coactivator classes may act in concert to modulate gene transcription. The
final model considers that the same gene may be responsive to the assembly of a
variety of coactivator complexes. Multiple signal transduction events may
independently effect the expression of the same gene to varying degrees by causing
different coactivators, coordinated in either a simultaneous or sequential manner, to
be recruited to the target gene promoter.

A better understanding of how cofactors and NRs effect transcription may be

30

advanced with a greater appreciation that promoters are organized into nucleosomes
and that chromatin structure contributes to regulation of promoter activity.
Transactivation studies using the mouse mammary tumor virus (MMTV) promoter
emphasize this point. The MMTV promoter is packaged into a regular array of
nucleosomes which suppresses basal transcription when the MMTV promoter is
organized into replicating chromatin (Bresnick et al., 1992). GR binding to
elements in this nucleoprotein associated promoter induces a chromatin remodeling
event that allows secondary transcription factors to bind, thereby activating
transcription. When the MMTV promoter is transiently transfected into cells it
adopts a less organized structure that allows secondary transcription factors to bind
constitutively. Compared to the chromatin template, basal activity for the
transfected promoter is relatively higher and a smaller degree of induction by
liganded GR is observed (Archer et al., 1992; Fragoso et al., 1998).

Like the MMTV promoter, the nucleoprotein structure of the genomic p82
promoter has also been mapped (Sewack and Hansen, 1997). The p82 promoter
contains an ER binding enhancer site and is transcriptionally upregulated by
liganded ER. Using a chromatin irnmunoprecipitation (ChIP) assay Sewack et. a1.
demonstrate specifically that the acetylation of two histones within the p82
chromatin array is induced with estradiol treatment in ER expressing cells.
Moreover, they correlate this acetylation with DNAse 1 hypersensitivity and
promoter activation (Sewack et al., 2000). Rather than simply knowing that

acetylation and nuclease sensitivity is increased generally, this experimental method

31

allows researchers to address the question of which nucleoproteins are acetylated
and directs their attention to where any subsequent disruptions may occur within the
targeted promoter nucleosome structure. ChIP assays have also been used to
demonstrate that both the coactivator BRGl , and the basal transcription factor TBP,
interact with the p82 promoter chromatin in an ER—dependent manner (DiRenzo, et.
al., 2000; Sewack et al., 2000). The association of BRGl with the promoter
complex plays a significant role in the positive regulation of the p82 gene. From
studies with the p82 promoter we know that coactivation of ER by SRC-l requires
BRGl (Sewack et al., 2000). Results from studies investigating GR regulation of
the MMTV promoter offer further insight into BRGl function in NR transactivation.
Liganded-GR recruits BRGl to the MMTV promoter where the presence of both GR
and BRGl is required for recruitment of the secondary transcription factor, nuclear
factor 1 (NFl). BRGl subsequently directs the ATP-dependent dissociation of GR
and BRGl from the chromatin template (Fletcher et al., 2000). It is unclear whether
NFl remains associated with the MMTV promoter after the dissociation of GR and
BRGl , however, it is clear that the association of GR and BRGl with the chromatin
is a dynamic process. Corroborating this observation, McNally et al. report that
liganded-, fluorofore-labeled GR undergoes continuous exchange between being
bound to and dissociating from chromatinized MMTV DNA regulatory elements in

living cell cultures (McNally et al., 2000).

32

1.5 Factors influencing receptor activity: Compressors

In contrast to coactivators, corepressors are a class of proteins that interact
with unliganded nuclear receptors and repress transcription (for review, see
Horowitz et al., 1996). Two of the better characterized examples of corepressors are
the ubiquitously expressed, homologous proteins SMRT (silencing mediator of
retinoic acid and thyroid hormone receptor) and N-CoR (nuclear receptor
corepressor) (Chen et al., 1995; Horlein et al., 1995). Biochemical studies have
determined that SMRT and N-CoR have multiple conserved silencing domains at
their N-terminus and distinct C-terminal receptor interaction domains (Li et al.,
1997; Nagy et al., 1999; Perissi et al., 1999; Web et al., 2000). Their role in
regulating transcription is especially apparent when the mechanism for gene
silencing by the RAR and the TR is considered. Unlike ER, class H nuclear
receptors such as RAR and TR form heterodimers with RXR and localize to their
cognate DNA response elements as unliganded heterodimers. In this state they
strongly repress basal expression from promoters containing TR and RAR enhancer
elements (Baniahmad et al., 1992). This transcriptional silencing effect of class H
receptors is mediated by the interaction of unliganded receptor with corepressors
(Chen et al., 1995; Horlein et al., 1995). Basically, ligand binding causes
corepressor dissociation fiom the DNA-bound receptor complex, encourages
coactivator interaction, and ultimately triggers gene transcription (Chen et a1 ., 1995;
Horlein et al., 1995; Nagy et al., 1999; Perissi et al., 1999). The suggestion that the

SMRT and N-CoR proteins are transcriptional corepressors that facilitate repression

33

by unliganded TR and RAR is supported by protein-protein interactions and
transient transfection experiments with expression vectors for the GAL4 DBD—fused
SMRT and N-CoR proteins. SMRT and N-CoR complex with unliganded
RAR/RXR or TR/RXR heterodimers that are associated with their cognate DNA
enhancer elements (Chen et al., 1995; Horlein et al., 1995). A repression of basal
transcription from a GAL4-responsive reporter gene in cells over expressing GAL4
DBD-fused SMRT or N-CoR C-terminally truncated mutants demonstrates that
SMRT and N-CoR have multiple independent silencing function domains with
activity that can be conferred on heterologous fusion proteins (Chen et al., 1995;
Hdrlein et al., 1995). The silencing domains correspond with binding sites for
Sin3A/B proteins (Li et al., 1997). Sin3 acts as a scaffold protein that links SMRT
and N-CoR with basal transcription factors and histone deacetylases to form a
multisubunit repressor complex (Heinzel et al., 1997; Nagy et al., 1997; Wong and
Privalsky, 1998) Reducing the availability of Sin3 and its associated histone
deacetylase RPD3 by antibody blocking correlates with abolition of N-CoR-
dependent transcriptional repression (Heinzel et al., 1997). In a similar role to that
of histone acetylases, deacetylases may influence transcription by affecting DNA
structure. DNA wraps around histone proteins to form nucleosomes, and
nucleosomes give shape to chromatin (Wu and Grunstein, 2000). Acetylation
destabilizes nucleosomes by causing DNA to be less tightly associated with
histones. Enhanced transcriptional activation parallels increased histone acetylation

(Bartsch et al., 1996; Wolffe, 1994). On the other hand deacetylation stabilizes the

34

nucleosome structure and decreased transcriptional activity is generally associated
with increased deacetylation (Burke and Baniahmad, 2000; Chen and Townes,
2000)

The location and structure of nuclear receptor interaction domains found in
coactivators are conserved in SMRT and N-CoR. An I/LXXII motif, which
resembles the coactivator LXXLL motif, dictates the sites of nuclear receptor
interaction with corepressor (Nagy et al., 1999; Perissi et al., 1999; Webb et al.,
2000). The C-terminal half of SMRT contains two I/LXXII motifs, and the C-
terminus of N-CoR has three motifs that contribute to receptor binding (Nagy et al.,
1999; Perissi et al., 1999; Webb et al., 2000). Since the nuclear receptor interaction
domains conserved in SMRT and N-CoR resemble those described for coactivators,
researchers speculated that the domains for coactivator and corepressor binding to
nuclear receptors may overlap. The I/LXXII motif is believed to have an
amphipathic (rt-helical structure with a non-polar surface on one side that, like the
coactivator L)O(LL motif, could be a site for hydrophobic interaction with NRS
(Feng et al., 1998, Mak et al., 1999; Nolte et al., 1998). Studies found that point
mutations of specific amino acids at or near receptor residues critical for coactivator
binding located in receptor helices 4 and 5 also disrupt corepressor binding (F eng et
al., 1998; Nagy et al., 1999; Perissi et al., 1999). However, carboxy terminal
truncation of RAR or TR, which deletes helix 12 and results in a receptor mutant
that shows reduced coactivator binding, exhibits enhanced interaction with SMRT

(Li et al., 1997). Further truncation of the C-terminus that removed helices 11 and

35

12 of TR and RAR reduced SMRT interaction with the receptor implicating
sequences within helix 11 in corepressor binding. It appears that corepressors bind
nuclear receptors at multiple sites and, to some extent, share similar points of
contact with coactivators; yet a clear distinction arises when receptor helix 12 is
considered. To recapitulate, coactivator binding is severely impaired by mutations
in helix 12 (Collingwood et al., 1997), but corepressors still bind receptor helix 12
mutants (Li et al., 1997). Moreover, additional studies suggest that corepressors do
not efficiently dissociate from helix 12-mutant nuclear receptors (Chen et al., 1996).
Unlike class 11 nuclear receptors (such as RAR and TR), which form
unliganded RXR heterodimers that bind DNA, class I nuclear receptors (i.e., steroid
receptors, including ERor) only bind DNA as ligand-activated homodimers. In the
ligand-bound state, ERor assumes a conformation that promotes coactivator
interaction and ERor-directed gene transcription (Darimont et al., 1998; Ding et al.,
1998; Nagy et al., 1999; Nolte et al., 1998; Ofiate et al., 1995). Researchers have
also shown that ERor binds SMRT both in the absence or presence of ligand (Smith
et al., 1997), but it is unclear what consequence this may have on transcriptional
signaling. Recent reports suggest that one consequence of corepressor binding may
explain the mixed agonist/antagonist nature of tamoxifen, which under some
circumstances inhibits ERor activity and at other times acts as a stimulatory agent
(Katzenellenbogen et al., 1996; Smith et al., 1997). Whether tamoxifen acts as an
agonist or antagonist when bound to the receptor may be dictated by the relative

abundance of coactivators and corepressors in a particular cell. Using a reporter

36

gene transfection assay in yeast cells, Graham and coworkers demonstrated that
when the coactivator L7/SPA is coexpressed with tamoxifen-liganded ERor the
agonist activity of tamoxifen was enhanced (Graham et al., 2000). L7/SPA is
shown to interact with NRS and enhance partial agonist activity of antagonists bound
to PR, GR and ERor (Jackson et. al., 1997). Similar increased activation was
observed when TIF2 or SRC-l and ERor were coexpressed with an ER—responsive

reporter gene construct in HeLa cells (Webb et al., 1998).

37

$01 302 R101 RIDZ

 

 

 

 

 

1 134 475 743 1495 aa

Fig. 4. Schematic of the SMRT functional domains. The diagram shows where
functional domains are roughly positioned along the length of the 1495 amino acid
SMRT protein. The hatched boxes represent receptor interaction domains (RID)

and black boxes specify silencing domains (SD).

38

1.6 Antiestrogens and partial agonists

Estrogens are critical for the normal development and firnction of
physiological processes in a variety of tissues, however, estrogen is also implicated
in the promotion of pathological processes such as breast and uterine cancer
(Gustafsson and Warner, 2000; Lavinsky et al., 1998; Lawson et al., 1999; Speirs et
al., 1999). Awareness of estrogen related pathologies has made the ER a target for
endocrine therapy and spawned the development of selective estrogen receptor
modulators (SERMS) which bind ER and influence its activity. Initial research
focused on the ability of estrogen-like compounds to disrupt the proliferative effect
estrogen has on tissues such as the breast and uterus. Now SERMS are also
investigated to assess their potential to mimic the positive effects estrogen has on
physiological parameters such as bone growth and maintenance, and cardiovascular
and neurological fitness (Dechering et al., 2000). Ideally, SERMS would offer the
beneficial estrogen effects in some tissues while lacking or inhibiting the
undesirable estrogen effects in other tissues. Speculation that further investigation
of SERMS may disclose a cancer preventative agent has prompted even more
research (King et al., 2001).

The SERM classification catagorizes a group of compounds that range from
very effective estrogen receptor agonists to pure antiestrogens which interact with
both ERor and ERB to differentially regulate their function in control of cell growth
and signaling (Gustafsson and Warner, 2000; Katzenellenbogen et al., 2000).

Antiestrogens are competitive antagonists of ER activity. They occupy the ERor and

39

ERB LBD to the exclusion of agonists and inhibit the ER transactivating function.
Many ER—binding compounds have been synthesized and confirmed to have an
antagonistic nature. This array of antiestrogens effect inhibition of ER
transactivation to varying degrees that are shown to be cell type- and promoter-
specific (Berry et al., 1990; McDonnell et al., 1995: Yang et al., 1996a; Yang et al.,
1996b).

Antiestrogens are divided into two basic classes: pure antiestrogens and
partial, or mixed, antiestrogens. The related compounds ICI 164,384 and [CI
182,780 exemplify pure antiestrogens. In vivo studies indicate that [Cl 182,780
inhibits uterine and mammary cell growth and consistently blocks estrogen-induced,
ERor-directed transcription of an ERE reporter gene (Bowler et al., 1989; Nuttall et
al., 2000; Wakeling et al., 1991; Wijayaratrre et al., 1999).

The chemicals raloxifene and tamoxifen are examples of mixed
antiestrogens. As a relatively recently developed drug, raloxifene has received
comparatively less attention in the literature than other antiestrogens; however, a
few reports on raloxifene studies offer a great deal of insight into the character of
both pure and mixed antiestrogen activity. Raloxifene inhibits an estradiol growth
response in breast and uterine cells; conversely, it has a stimulating effect on bone
growth and remodeling (Draper et al., 1996; Huston, 1999). Administration of
raloxifene stimulates TOE-[33 expression in the femur of ovariectomized rats and
cotransfection studies indicate that raloxifene induces ERor-dependent activation of

the TGF-B3 promoter (Yang et al., 1996a; Yang et al., 1996b). Although we were

40

unable to confirm their results in HeLa cells, Yang et al. report that cotransfected
ERor strongly activates a reporter gene construct driven by the TGF-B3 promoter in
human osteosarcoma MG63 cells treated with raloxifene. The TGF-B3 promoter
lacks a consensus ERE and was not activated by E; treatment in these studies.
Moreover, in a reversal of roles, cotreatrnent with E2 antagonized raloxifene action
(Yang et al., 1996b).

A comparison of the crystal structures of the E2- and raloxifene-bound
ERor LBD indicates that the receptor undergoes a different conformational change
upon binding of structurally dissimilar ligands (Brzozowski et al., 1997). When
raloxifene settles into the LBD, a portion of the raloxifene molecule is not
accommodated by the cavity (it protrudes from the pocket) and prevents H12 from
folding into the body of the LBD as it does when estrogen binds (Brzozowski et al.,
1997, Ekena et al., 1997). These observations provide direct evidence that the
binding of structurally distinct compounds effects distinct changes in receptor
conformation. Further research suggests that this difference modulates the ability of
nuclear receptors to recruit co-regulators (Lavinsky et al., 1998; Smith et al., 1997).

Like raloxifene, tamoxifen is a mixed antiestrogen and acts as an agonist (to
varying degrees), or antagonist of ER-induced gene expression in a cell type- and
promoter-specific manner (Katzenellenbogen et al., 1996, Montano et al., 1998).
Tamoxifen antagonizes estrogen stimulated growth of breast cancer cells in culture
and is ofien used successfully in cancer treatment to cause grth cessation and

regression of ER-positive breast tumors (Furman et al., 1992; Lamy et al., 2002).

41

The duality of tamoxifen action is evident in the treatment of some breast cancer
patients for whom it is observed that tamoxifen antagonizes the proliferative effects
of estrogen in the breast, yet encourages the cell proliferation that introduces an
increased risk for carcinogenesis in the uterus (Juneja et al., 2002; Decensi, et al.,
2002; Shang and Brown, 2002).

Activation of the quinone reductase (QR) gene by tamoxifen-liganded ERor
and ERB provides another example of the agonist activity of tamoxifen. QR is an
enzyme that decreases the generation of hydroxyl radicals in a cell by reducing
quinones. This may be a factor in defining tamoxifen’s putative role as an anti-
cancer agent. Tamoxifen regulation of the QR promoter shows complete reverse
pharmacology such that ER directed QR expression is suppresed by E; in breast
cancer cells (Montano et al., 1998). From these examples it appears that the
promoter, tissue-type and receptor isoform all play a role in SERM signaling.

Tamoxifen binding supports dimer formation and receptor DNA binding
(Cheskis et al., 1997). As noted above, ER has two distinct transactivation domains.
The C-terminal AF 2 is regulated by conformation of the LBD. AF 1 at the N-
terminus is not directly regulated by ligand. Both contribute to the total activity of
the estrogen-bound full-length receptor on an ERE (Benecke et al., 2000; kabayashi
et al, 2000; Metivier et al., 2001; Tora et al., 1989). When tamoxifen binds to ERor,
in contrast to E2 binding, the resulting LBD conformation impairs AF 2 activity;
however, AF 1 activity remains intact (Berry etal., 1990; Brzozowski et al., 1997).

Therefore, with tamoxifen bound, a receptor may dimerize and translocate to a

42

promoter and effect transcription via a functional AF 1 domain. Notably, although
ERor and ERB share a great deal of homology, their sequence is most divergent at
the N-terminus (Enmark et al., 1997). Perhaps by affecting how a receptor interacts
with cofactors, any N-terminal structural difference between ERor and ERB may
distinguish the effect tamoxifen binding has on the receptors (Berry et al., 1990;
Lazennec et al., 2001). Furthermore, differential expression of receptor isoforms
may partially explain tissue specific responses to tamoxifen treatment (Gustafsson
and Warner, 2000; Shang and Brown, 2002). Evidence is building to suggest that
the relative amounts and profile of corepressor and coactivator expression may also
be a contributing factor. A decreased level of N-CoR expression correlates with the
observation of tamoxifen-resistant growth in both cultured human breast cancer cells
and in mouse breast tumors (Lavinsky et al., 1998). In transient transfection studies,
over expression of SRC-l is reported to enhance tamoxifen-stimulated transcription
in a dose—dependent manner (Smith et al, 1997). Conversely, over expression of
SMRT or N-CoR reduced tamoxifen-stimulated transcription (Lavinsky et al., 1998;
Smith et al., 1997). Consistent with the notion that the selective interaction and
relative levels of co-regulators may influence the activity of tamoxifen-liganded
receptor, in vitro binding studies have shown that both corepressors and coactivators

can bind tamoxifen-liganded receptor (Smith et al., 1997).

1.7 Regulation of ERor function by phosphorylation

Steroid hormone receptors, including ERa, are substrates for a variety of

43

kinases and phosphatases. The state of phosphorylation of specific receptor residues
affects receptor function (Arnold et al., 1997; Joel et al., 1998). Unliganded
receptors are basally phosphorylated and become hyperphosphorylated following
association with hormone. This increased phosphorylation observed with agonist
binding is not supported by the binding of receptor antagonists (Denner et al., 1990;
Denton et al., 1992; Orti et al., 1989). In vitro phosphorylation of purified ERor by
Src family tyrosine kinases (e.g., p60"’src and p56'°") on tyrosine 537 correlates with
maximum estrogen binding capacity and ERE association (Arnold et al., 1997,
Arnold et al., 1995b). In vitro dephosphorylation by protein-tyrosine phophataselB
(PTPlB), or a phenylalanine substitution of this residue significantly decreases ERor
honnone—binding capacity (approximately 90 percent) and also inhibits DNA
binding (Arnold et al., 1997, Arnold et al., 1995b). Results from studies designed to
investigate intracellular localization of phosphorylated ER in Ez-treated MCF-7 cells
(an ER—positive breast carcinoma cell line) demonstrate that ninety percent of the
MCF-7 ER population is nuclear localized and phosphorylated at tyrosine 537. The
smaller percentage of ER found in the MCF-7 cytosolic fraction is not
phosphorylated at tyrosine 537 (Arnold et al., 1995b).

Increased affinity of ERor for its ERE and enhanced activation of an ERE-
directed reporter gene construct by ERor is found to occur with phosphorylation of
serine 167 (Arnold et al., 1995c; Frddin and Gammeltoft, 1999; Joel et al., 1998).
Serine 167 phosphorylation is regulated by a mitogen-activated protein (MAP)

kinase signal transduction cascade. Serine 167 is directly phosphorylated by a p90

kDa ribosomal S6 kinase (RSK; also known as MAPK-activated protein kinase-1,
MAPKAP-Kl) (F rodin and Gammeltofi, 1999; Joel et al., 1998). RSK is activated
when it is phosphorylated by the MAPK, extracellular signal-regulated kinase
(ERK) (Blenis, 1993; Joel et al., 1998). RSK and ERK phosphorylate both cytosolic
and nuclear localized proteins (Frodin and Gammeltofi, 1999). Over expression of
either RSK or ERK enhances endogenous ER-mediated transcription of an ERE-
driven reporter gene construct in Ez-treated MCF-7 cells (Joel et al., 1998).

Serine 118 is located in the ER A/B region and is also phosphorylated by
ERK. Phosphorylation of serine 118 by ERK correlates with enhanced ERor gene
activation in cotransfection studies (Kato et al., 1995). Kato and coworkers
observed increased expression from an ERE-driven reporter gene construct in ER-
negative COS-1 cells when ERK was overexpressed with either ligand-bound wt ER
or a truncated receptor containing receptor regions A-C (including the AF 1 and
DBD domains). This suggests that phosphorylation of serine 118 modulates AF 1
activity. Furthermore, these authors report that coexpression of either Ras or its
substrate Raf (MAPK kinase kinase) enhanced activity of the AF 1 -containing
truncated mutant (Kato et al., 1995).

Some studies suggest that ER and growth factor receptor signal transduction
pathways may cooperate to transduce the cell response to growth factors such as
IGF-l , TGFor, and EGF (Bunone et al., 1996; Ignar-Trowbridge et al., 1996;
Newton et al., 1994). The ERK kinase pathway that converges on the ER is

activated by membrane-associated receptor tyrosine kinases, examples of which

45

include receptors for the growth factors EGF and IGF -1 (for review, see Pearson et
al., 2001). Receptor tyrosine kinases phosphorylate and activate Ras which in turn
activates Raf. MAPKK or MEK is subsequently phosphorylated and activated by
Raf. The substrate for MEK is ERK (Pearson et al., 2001; Frddin and Gammeltofi,
1999). TGFa similarly acts via a parallel MAP kinase cascade which through
phosphorylation of p38 MAPK induces phosphorylation of ERK and RSK (F rodin
and Gammeltofi, 1999). The Treatment of cells with IGF-l, EGF or TGFor
stimulates hyperphosphorylation of ERor and is reported to result in both enhanced
ERa-dependent transcription in cotransfection assays and cell growth (Bunone et

al., 1996; Ignar-Trowbridge et al., 1996; Kato et al., 1995; Newton et al., 1994).

46

2. Non-consensus Estrogen Enhancer Elements

Recent reports have identified an increasing number of genes with promoters
lacking an ERE that show ERor-enhanced gene expression. For these promoters,
transcriptional stimulation by ERor appears to be mediated by an indirect
mechanism, involving synergistic protein-protein interactions with known
transcription factors including AP-l and SP-l (Gaub et al., 1990; Krishnan et al.,
1994)

The expression levels of the RARor, cathepsin D and c-fos genes are
upregulated by Ez-liganded ERor. The ERor enhancer element maps to a GC-rich
Spl binding motif in the promoter region of these genes (Duan et al., 1998;
Krishnan et al., 1994; Sun et al., 1998). Porter and colleagues demonstrate that
ERor and Spl directly bind in solution (Porter et al., 1997). Using a gel mobility
shift assay, another report confirms that both ERor and Spl complex with a putative
Spl/ERor response element in the c—fos promoter (Duan et al., 1998). Collectively,
these studies demonstrate that the E2 responsiveness of the RARor, cathepsin D, and
c-fos promoters in transient transfection studies requires the presence of ERor, Spl ,
and the identified Spl binding motif (Duan et al., 1998; Krishnan et al., 1994; Sun
etaL,1998)

AP-l describes the ubiquitous jun/fos family of transcription factors whose
activity is crucial for the efficient expression of a wide variety of genes. As an

important downstream target for the MAPK signaling cascades, AP-l is a central

47

player in mediating the effects of serum and growth factors on cellular proliferation
(Boyle et al., 1991; Westwick et al., 1994). A variety of estrogen-responsive genes
have been described that lack a palindromic ERE, but instead contain one or more
consensus AP-l elements (5 '-TGAG/CTCA-3 '), with or without a degenerate ERE
or ERE half-site (5'-GGTCA-3' or 5’TGACC-3'). Examples of such genes include
ovalbumin, which is induced by E; in chicken oviduct cells (Tora et al., 1988), and
the rat insulin-like growth factor-1 (IGF-l) gene whose expression is stimulated by
E; in the uterus of ovariectomized-hypophysectomized rats and in cultured rat
osteoblast cells (Ernst and Rodan, 1991; Murphy and Ghahary, 1990). An AP-l
enhancer motif identified in the chicken IGF-l promoter is essential for E2 and
phorbol ester-stimulated gene transcription (U mayahara et al., 1994). Phorbol ester
serves to stimulate increased AP-l activity by directly activating PKC, which in turn
activates Ras to initiate a MAP kinase cascade that sequentially includes the
activation of MEKK (MAPK kinase kinase), c-jun N-terminal kinase kinase (JNKK)
and c-jun N-terminal kinase (JN K). JN K ultimately phosphorylates and thereby
activates AP-l (Boyle et al., 1991; Bokemeyer et al., 1996; Vuong et al., 2000;
Westwick et al., 1994).

Like the rat IGF-l gene, the chicken ovalbumin promoter does not have an
ERE and the ERor regulatory element maps to a critical AP-l site (Gaub et al., 1990;
Tora et al., 1988). That this effect is mediated by AP-l is supported by the
observation that the synergistic induction of ERor-dependent expression by E2 and

phorbol ester cotreatrnent is further enhanced by cotransfection with c-fos or c-jun

48

(Gaub et al., 1990).

Reporter gene cotransfection studies with expression vectors for AP-l
isoforms and ERor in HeLa cells indicate that a similar mechanism regulates the
human collagenase promoter (Webb et al., 1995). The minimal region of the
collagenase promoter reported to be responsive to tamoxifen-liganded wt ERor (and
to a variety of ERa mutants) harbors a critical AP-l element and lacks a consensus
ERE. The activity of ERor on the collagenase promoter was enhanced with AP-l (c-
jun or c-fos) over expression (Webb et al., 1995). Furthermore, evidence that ERor
regulation converges with AP-l-directed gene transcription is provided by results
from protein binding assays indicating that c-jun is able to bind to wt ERor in vitro

(Webb et al., 1995).

49

 

 
   
     
   
 

cell nucleus

 
   

GGTCAnnnTGACC
TGAGICTCA

TGAGICTCA

 

Fig. 5. A diagram depicting wt ERor transcriptional activation of consensus
and nonconsensus regulatory elements. The left side of the figure represents a
receptor dimer acting through a consensus ERE to transactivate gene expression.
The receptors in the lower middle and on the right are cast in a nonclassical
transactivating role on a consensus AP-l motif (TGAG/CTCA). In this role it is
unclear if ER directly binds AP-l factors (e.g., jun and fos) or indirectly interacts
with them through association with coactivators that bind AP-l factors.

50

3. Identification of ERor mRNA Splicing Variants

Numerous variant ERor cDNAs have been cloned and sequenced from a
variety of breast tumors and established tumor cell lines (Castles et al., 1995:
Gotteland et al., 1995; Miksicek et al., 1993; Murphy et al., 1997). The most
common variants harbor a precise deletion of one or more of the internal exons that
contribute to the structure of the mature ERor protein, suggesting that they arise as a
result of imprecise splicing of the primary ERor mRNA transcript. ERor cDNAs
with deletions corresponding to exons 2, 3, 4, 5, and 7 have been identified, along
with a large number of more complex variants (Castles et al., 1995: Gotteland et al.,
1995; Miksicek et al., 1993; Murphy et al., 1997). These basic variants are referred
to as ERAE2 through ERAE7, where the deleted exon is indicated numerically.
While there is no consistent ratio of relative expression, wt and variant ERor
transcripts are apparently coexpressed in all ERa-positive cell lines and tissues
(Castles et al., 1995; Gotteland et al., 1995; Hu et al., 1996). Quantitation of
individual variants shows that each variant generally represents a minority of ERor
mRN A; however, as a population, splicing variants can constitute 50 percent or
more of the total ERor mRNA (Castles et al., 1995; Erenburg et al., 1997; Gotteland
et al., 1995; Murphy et al., 1997). While most efforts to evaluate expression levels
of individual variants have focused on mRNA, Traish et al. report that of 29 human
breast tumors they analyzed, 45 percent contained ER protein with a defect in the

DNA-binding domain. Using gel mobility shift analysis and site-directed

51

monoclonal antibodies these authors also report that 3 of the 29 tumors contained
ER protein which had defects in both the N-terminal A/B region and the DNA-
binding domain (Traish et al., 1995).

Quantitation of ER variant transcripts has revealed that the population of
ERor transcripts is mixed and varies among the tissues and cell types examined;
furthermore, wt ERor is usually the major transcript expressed. However, some
studies indicate that the presence of a single splicing variant can sometimes exceed
wt ERa in amount. Quantitation of mRNA transcripts in the BT-20 breast cancer
cell line shows that ERAES is the major ERor isoform expressed in these cells
(Castles et al., 1993; Hall et al., 1990). ERAES comprises 68 percent of the ERor
mRNA population while wt ERor measures only 8 percent. Results from
immunoblot analysis agree with mRNA analysis and demonstrate that ERAES is the
predominant ER protein expressed in BT-ZO cells (Castles et al., 1995). The
receptor status in the BT-2O cancer cell line, which was considered to be ER-
negative by ligand-binding analysis, raises the notion that the ratio of variant and wt
ERor expression may affect hormone responsiveness and cancer progression. In this
context, analysis of wt ERor and ERAES mRN A in various stocks of the human
breast tumor cell line MCF-7 showed that an elevated ratio of ERAES compared to
wt ER mRNA correlated with the ability of some MCF-7 sublines to grow
independently of estrogenic stimulation (Klotz et al., 1995).

Studies also indicate that, while ERAEB tends to be underrepresented in

52

breast tumors and tumor cell lines, it appears to constitute 50 percent or more of
ERor mRNA in both stromal fibroblasts and epithelial cells isolated fi'om reduction
mammoplasty specimens (Erenburg et al., 1997). This leads to speculation that
higher relative abundance of ERAE3 may be a feature of normal breast tissue, where
it may slow cell growth relative to breast tumor cells. Examination of the growth of
MCF-7 cells stably transfected with ERAE3 supports this notion. Estrogen
stimulated the ability of the parental cell line to form colonies in soft agar. Reduced
colony formation was observed for ERAE3 expressing clones (Erenburg et a1 .,

1997)

53

DNA Hormone

 

 

 

 

 

 

 

 

 

 

wildtype N c 66.6kDa

AF‘I NLS AF2

I VII VIII
Exon Structure r l .11.": IV 'V .V| ' Zi III IIIIII
ensez n—i-c:_r§-1_::i;jj:j::j}-c 17.0 kDa
ERAE3 N———-" f ' ' ' ' c 62.3kDa

ERAE4 N——-.t‘ 1W0 54.1 kDa.

 

ERAES n—Icmc j;:-::::::r-c 41w»
ERAEG N MyIIIZG-C 53.0 kDa
ERAE7 N We ri-c 52.2 kDa

Fig. 6. Comparison of ERor mRNA splicing variants and wt ERor structure.
The locations of the various functional domains of ERa are depicted along with the
exon sequence from which they are derived. The variants are referred to by deleted
exon. The size and molecular weight of each of the variants are predicted from the
translational reading frame of the sequenced cDNA clones. Dashed lines indicate
regions of the major open reading frame of the full-length ERor protein that are
missing from each variant. The nuclear localization signal is circled. The regions
encompassing the DNA- and ligand-binding domains are marked by darkened and
hatched boxes, respectively. The AF 1 domain and AF 2 core domains are indicated

where they reside within the N- and C-termini of the wild-type receptor.

54

4. Activity of ERor Splicing Variants

While there has been extensive analysis of the pattern of expression and
abundance of ERor splicing variants, limited information has been available until
recently on their functional activity. Early reports indicated that ERAES can support
weak, cell type-dependent activity (Chaidarrrn and Alexander, 1998; Fuqua et al.,
1991; Rea and Parker, 1996). Also, when tested on an ERE, both ERAEB and
ERAES are dominant negative receptor forms in the presence of wt ERor (Ohlsson
et al., 1998; Wang and Miksicek, 1991). Taken together these two observations
suggested to us that splicing variants may play a role in regulating gene
transcription.

Fuqua and colleagues reported that ERAES (which contains the AF 1 domain,
but lacks AF 2 and the regulatory functions imparted by the LBD), is constitutively
active in promoting transcription from an ERE in a heterologous yeast reporter gene
assay (Fuqua et al., 1991). These authors also described that over expression of
ERAES in a stably transfected breast cancer cell line supported greater proliferation
compared with control cells, as well as imparting a tamoxifen-resistant phenotype
(Fuqua and Wolf, 1995). In the human osteosarcoma cell line U2-OS, coexpression
of ERAES significantly enhances ERE-directed reporter gene expression induced by
wt ERor (Chaidarun and Alexander, 1998). The presence of a constitutively active
receptor variant able to exert a mitogenic effect in breast tumor cells in the absence

of E2 or in the presence of tamoxifen was thought to be an appealing explanation for

55

aquired antiestrogen resistance observed in previously responsive breast tumors and
cell lines (Wiseman et al., 1993; Wolf and Jordan, 1994). However, this model is
now challenged by conflicting observations that ERAES and closely related,
genetically engineered ERor mutants do not efficiently induce transcription from an
ERE reporter when transiently transfected into the ER-negative HeLa or chicken
embryo fibroblast (CEF) cell lines, or promote proliferation in stably transfected
breast tumor cells (Kumar and Chambon, 1988; Rea and Parker, 1996). A
constitutive transcriptional activity has also been reported for the ERAE4 variant.
When transfected into the ER-negative Chinese hamster ovary CHO k1 cell line,
ERAE4 was observed to induce an ERE-containing reporter gene construct in a
hormone-independent manner (Pasqualini et al., 2001). Like reports that describe
ERAES to be a constitutively active variant, the significance of this finding is called
into question by results from another study which suggests that ERAE4 does not
bind DNA in vitro and ftuthermore does not transactivate an ERE in human
choriocarcinoma JEG3 cells (Koehorst et al., 1994).

Aside from the potential for ERor variants to activate a promoter, the mRN A
splicing variants ERAE3 and ERAES are reported to have an inhibitory effect on wt
ERor transactivating activity in cotransfection studies (Ohlsson et al., 1998; Wang
and Miksicek, 1991). A 70 percent inhibition of transcriptional activation by E;-
liganded wt ERor on an ERE-driven reporter gene was observed in HeLa cells when

ERAE3 and wt ERa were coexpressed (Wang and Miksicek, 1991). In the ER-

56

negative nontumorigenic human breast cell line HMT-3522S 1 , coexpression of an
equal amount of ERAES also significantly inhibited stimulation of an ERE reporter
construct by wt ERor (Ohlsson et al., 1998). Increasing the ratio of transfected
variant to wt ERor demonstrates that the repression of wt ERor by ERAE3 and
ERAES is dose-related and becomes nearly complete when the variants are present
in sufficient excess over the intact receptor (Ohlsson et al., 1998; Wang and
Miksicek, 1991). This observation has physiological significance in the case of cells
such as the BT-20 breast tumor cell line that predominantly express one of these
splicing variants. Additionally, stable over expression of ERAE3 in MCF-7 cells to
levels seen in normal mammary epithelial cells dramatically reduced the ERor-
dependent expression of endogenous p82 mRN A and anchorage independent growth
of MCF-7 cells (Erenburg et al., 1997).

ERAES retains both the DBD and AF 1 domains and, based on what we
know of the modular character of wt ERor, presumably also retains the functions
ascribed to these domains. With the DBD and AF 1 intact, ERAES may also retain
the ability to bind an ERE or interact with co-regulators, thereby possessing the
capacity to transactivate genes or disrupt wt ERor activity by competing for DNA or
coactivator binding. The fact that ERAE3 lacks an intact DBD must account, at
least in part, for its inability to direct gene induction through an ERE, even though it
retains fully functional AF 1 and AF 2 domains. Furthermore, ERAE3 inhibition of

wt ERor activity on an ERE-driven reporter gene may be a result of its ability to

57

prevent wt ERor from binding to an ERE (Miksicek et al., 1993; Wang and
Miksicek, 1991).

The limited observations described above hint at a functional role for some
of the ERor mRNA splicing variants. Considering the prevalence of exon-skipped
ERor variants it is clear that further study is warranted to more thoroughly evaluate

their activities, particularly regarding their capacity to regulate gene transcription.

58

III. Biochemical Analysis of ERor mRNA Splicing Variants

Published in the journal Molecular Endocrinology
(Bollig and MIkSicek, 14(5): 634-649, 2000)

1. Introduction

Numerous variant ERor cDNAs have been cloned and sequenced from breast
tumors and established tumor cell lines (Castles et al., 1995; Gotteland et al., 1995;
Miksicek et al., 1993; Murphy et al., 1997). The most common variants harbor a
precise deletion of one of the internal exons from the eight that contribute to the
structure of the mature ERor protein, suggesting that they arise as a result of
imprecise splicing of the primary ERor mRNA transcript.

Like other nuclear receptors, ERor is a modular protein in that individual
domains demonstrate autonomous function (Beato and Sanchez-Pacheco, 1996;
Kumar et al., 1987). Based on mutational studies, it can reasonably be assumed that
the exclusion of a particular exon will predictably result in a receptor variant lacking
the function ascribed to that exon. Additionally, it is probable that the loss of a
particular exon will result in unpredictable functional deficits or perhaps even
bestow a novel function on the variant receptor. This study examines the function of
wt ERor splicing variants from the vantage point of what is known about the
functional organization of wt ERor. Concurrently, the process of examining splicing

variants, like mutational studies, improves our understanding of wt ERor firnction.

59

Following are results from experiments designed to assess receptor capacity to
translocate to the nucleus, bind to DNA, bind ligand, participate in protein

complexes, and promote gene transcription.

60

2. Results

Cytomegalovirus (CMV) promoter-driven ERor cDNA expression vectors
representing the exon-skipped variants ERAE2 through ERAE7 have been
constructed in order to enable their functional characterization in a well defined cell
transfection system. This assembled pool of ERor splicing variants also includes the
hypothetical receptor ERAE6, even though this variant is not readily identified in
vivo. Figure 6 diagrams the ERor mRN A splicing variants examined, showing the
positions of deleted exons and their consequences with respect to protein structure.
Deletion of exon 2, 5, 6, or 7 all cause a frame-shift mutation resulting in premature
termination of translation, thereby generating a diverse class of C-terminally
truncated receptor forms. Omission of either exon 3 or 4 does not disrupt the
mRN A reading flame, but produces a receptor protein with an internal deletion.

Transient expression in Cos7 cells demonstrates that each of these variants
translates to a stable protein able to accumulate to readily detectable levels within
transfected cells (Fig. 7). The C037 cell line is an ER-negative SV40 large T-
antigen transformed cell line derived from the simian kidney cell line CV-l.
Expression of SV40 large T-antigen allows for replication of our CMV-derived
expression vectors which contain the SV40 origin of replication (Gluzrnan 1981).
Based on immunoblot analysis with an N-terminal monoclonal antibody (Mab-17),
which recognizes an epitope within exon 1 common to all of the variants (Neff et

al., 1994), we observe that the mobility of the six variant proteins is consistent with

61

their predicted molecular weights. No immunoreactivity is observed in mock

transfected cells, confirming the specificity of the Mab—l 7 antibody.

2.1 Measurement of the DNA-binding activity of the ERor splicing variants
Efficient DNA binding by ERor requires the c00peration of several functional
elements within this protein, including the centrally located DBD and a ligand-
inducible subunit dirnerization motif located near the C-terminal end of the LBD
(Kumar and Chambon, 1988; Kumar et al., 1987). Since all of the ERor splicing
variants sustained deletions within various regions of this protein, it was of interest
to systematically assess the DNA-binding ability of each variant. For this purpose,
gel mobility shift assays were performed using extracts prepared from Ez-treated,
transiently transfected Cos7 cells. Extracts were incubated with a 32P-labeled
oligonucleotide (AGGTCACAGTGACCT) containing a consensus ERE from the
Xenopus vitellogenin A2 promoter. As expected, variants that harbor a deletion
within the DBD (ERAE2 and ERAE3) are completely unable to recognize the ERE
(Fig. 8, lanes 5—8). Less predictably, the loss of exons contributing to the LBD also
result in a strong defect in ERE recognition (Fig. 8, lanes 9-16). For ERAES and
ERAE7, however, this appears to be a quantitative defect in DNA binding. The
addition of the monoclonal antibody, Mab-17, to the binding reactions consistently
results in the recovery of weak DNA binding by ERAES (Fig. 8, lane 12).
Presumably, the role of the bivalent antibody is to stabilize the interaction of

receptor subunits withtheir palindromic binding site, mimicking the function of the

62

missing dirnerization motif present within the C-terminus of the LBD. These results
suggest the possible existence of cell-specific constituents that perform a similar
function in viva (e.g., co-regulators) and may account for the variable activity of
ERAES and related constructs in different cell types (Chaidarun and Alexander,
1998; Fuqua et al., 1991; Rea and Parker, 1996). We have also observed the
formation of a complex between ERAE7 and labeled ERE probe (Fig. 8, lane 16).
Overall, however, the relative weakness of DNA-binding observed in these studies
raises serious questions about the extent to which any of the variants, including
ERAES and ERAE7, are able to recognize and bind to a consensus ERE, in viva.
Furthermore, that ERAE7 binds an ERE in vitro has little transcriptional relevance
in light of the observation that ERAE7 is not translocated to the nucleus when

expressed in Cos7 cells (see below).

2.2 ERAE3, like wt ERor, binds ligand

To test the ability of the ERor mRNA splicing variants to bind hormone, we
performed a saturation binding assay on whole-cell extracts from Cos7 cells
transiently transfected with wt ERor or the ERor variants. Only wt ERor and ERAE3
were able to bind 3H-labeled E2, whereas all of the remaining variants demonstrated
no specific ligand binding (Fig. 9). The individual deletion of exons 2, 4, 5, 6, and 7
effectively eliminates all, or a significant portion, of the LBD (see Fig. 6), consistent
with their loss of hormone binding. These results were confirmed using an in viva

ligand-binding assay in which the binding of a fluorescent estrogen analog was

63

visualized in cells cultured on glass cover slips. Cos7 cells transiently expressing
the individual variants or wt receptor were treated with the fluorescent ligand, nitrile
tetrahydrochrysene (nitrile THC) (Miksicek et al., 1995). Only those cultures
transfected with wt ERor or ERAE3 were observed to stain with this ligand. In both
cases, staining was localized tightly within the nucleus. This suggests that among
the variants examined, wt ERor and ERAE3 exclusively bind ligand and in both
cases, the receptors are translocated normally to the nucleus of expressing cells (Fig.
10). The affinities of ERAE3 and wt ERa for 3H-labeled 132 were compared with
Scatchard analysis. The measured dissociation constants were 0.66 nM for wt ERor

and 0.79 nM for ERAE3 (Fig. 11).

2.3 Subcellular localization of ERor splicing variants

To more carefully assess the subcellular localization of ERor splicing
variants, including those that fail to bind ligand, Cos7 cells were transiently
transfected with expression vectors encoding wt ERor or individual variants. These
receptors were detected in transfected cells by indirect immunofluorescence staining
(using the Mab-17 monoclonal antibody) and confocal microscopy. Similar to wt
ERor, ERAE3 and ERAES localize to the nuclei of transfected cells, although ERAES
showed perinuclear as well as nuclear staining (Fig. 12). These results are
consistent with the fact that both ERAE3 and ERAES retain a NLS immediately

downstream of the DBD (Tsai and O’Malley, 1994; Ylikomi et al., 1992).

Subcellular localization studies have also been completed for the exon 2, 4, 6, and 7
deletion variants. Each of these proteins can be readily detected in transfected cells,
but they all possess dramatic defects in nuclear targeting (Fig. 12). Nuclear
targeting of wt ERor is governed, in large part, by a tripartite karyophilic signal
present within exon 4 (Ylikomi et al., 1992). Loss of this signal is therefore
consistent with the cytoplasmic pattern of distribution of mutants such as ERAE2
and ERAE4, both of which lack protein corresponding to exon 4 sequences.
Inappropriate presentation or folding of this signal must account for the defects in
nuclear localization seen with ERAE6 and ERAE7, since the NLS is retained in
these variants. Based on their subcellular distribution, we would predict that only
ERAE3 and ERAES, like wt ERor, would have the potential to exert nuclear effects,
such as modulating gene transcription. Moreover, the inability of the cytoplasmic
variants ERAE2, ERAE4, ERAE6, and ERAE7 to dimerize with wt ERor (see below)
predicts that their subcellular distribution will not be influenced by coexpression

with the nuclear isoforms (wt ERor, ERAE3, and ERAES).

2.4 Characterization of the transactivation function of ERor splicing variants
on the vitellogenin ERE

HeLa cell cotransfection experiments designed to assess the transcriptional
activity of individually expressed ERor splicing variants have failed to demonstrate

any significant ability of variant receptors to support gene activation through an

65

ERE, with the possible exception of the ERAES variant, which is reported to display
a low level of constitutive transcriptional activity on an ERE—driven reporter in
some, but not all, cell types examined (Chaidarun and Alexander, 1998; Fuqua et
al., 1991; Rea and Parker, 1996). In our hands none of the variants were effective
transcriptional activators of an ERE-containing promoter. ERAES repeatedly
showed only modest constitutive activity (5 percent of wt ERor induction) on an
ERE-directed reporter plasmid cotransfected into HeLa cells (see Fig. 13).

It is important to recognize that the tissues and cell lines that express these
variants also express wt ERor. It has previously been reported that the ERAE3
variant acts as a dominant negative mutant when it is coexpressed with wt ERor in
HeLa cells treated with B; (Wang and Miksicek, 1991). In the human breast
epithelial cell line HMT-352281, ERAES has also been shown to disrupt
transactivation by agonist-bound wt ERor of an ERE reporter gene (Ohlsson et al.,
1998). To clarify whether this is a frmction unique to these variants, we completed a
series of experiments to test whether the remaining exon-skipped ERor variants also
support transcriptional inhibitory effects. When examined in a HeLa cell
cotransfection assay in which the expression of pERE-TK-CAT was driven by E;-
bound wt ERor, a 5-fold molar excess of any of the splicing variants lacking exons 2,
4, 6, or 7 failed to inhibit the Ez—dependent induction of CAT gene expression by
intact receptor. In agreement with previously published results, the ERAE3 aned

ERAES variants both demonstrated a dominant inhibitory activity at all molar ratios

66

tested (Fig. 14). Coupled with the observation that equal amounts of plasmid DNA
support similar levels of variant expression (see Fig. 7), it appears that ERAE3 and

ERAES are approximately equivalent in their inhibitory activity in HeLa cells.

2.5 Dimerization and co-regulator binding properties of ERAE3 and ERAES

We next questioned whether a direct interaction between the variants and
factors responsible for ERE—directed transcription might explain the inhibitory
effects of ERAE3 and ERAES on wt ERa activity; specifically, we tested for a direct
interaction of ERAE3 and ERAES with wt ERor and SRC-le. The C-terminus of wt
ERor and fragments of the SRC-le protein were expressed as fusion proteins with
glutathione S-transferase (GST) and attached to glutathione-Sepharose beads.
Binding assays with GST fused to the C-terminus ofwt ERor (GST-AF2) and ”s-
methionine labeled in vitro translated receptor demonstrate that ligand-bound wt
ERor and ERAE3, but not ERAES, dimerize with the LBD of ERor in solution (Fig.
15A). In experiments with GST-SRC-le fragments, both ERAE3 and ERAES were
observed to bind to regions of the coactivator that also bind wt ERor (Fig. ISB).
ERAE3 and wt ERor bind the SRC-le fiagments comprising amino acids 579-780
and 989-1240 and do so only in the presence of E2. In contrast, binding of ERAES
to the 989-1240 amino acid fragment is constitutive (Fig. 15B).

It is previously reported that wt ERor binds corepressors (Smith et al., 1997).

In order to further evaluate the relationship between co-regulators and the splicing

67

variants we tested whether ERAE3 and ERAES also possess a corepressor binding
function. In vitro binding studies Show that wt ERor, ERAE3, and ERAES share a
capacity to bind the corepressor SMRT (Fig. 16). For wt ERa and ERAE3, binding
occurs with or without the addition of E2 and in the presence of tamoxifen. The

strongest SMRT association is observed in vehicle treated ERAE3 and ERAES

samples.

68

Fig. 7. Immunoblot analysis of wt ERor and ERor mRNA splicing variants.

Cos7 cells were cultured in phenol red-free DMEM supplemented with 10%
calf serum and transiently transfected with 10 pg of pCMV4-derived expression
vectors for the indicated ERor isoform. After 48 hours in culture, transfected cells
were resuspended in protease inhibitor—supplemented binding buffer and whole-cell
extracts were prepared by sonication. Samples containing 20 pg of soluble protein
were analyzed by SDS-PAGE and electrophoretically transferred to nitrocellulose
filters. Immunoblots were probed with the ERor-specific monoclonal antibody,
Mab-l7, obtained from a hybridoma culture supemate that was diluted with an equal
volume of PBS. Immunoreactive protein was visualized by chemiluminescence
using a horseradish peroxidase-conjugated secondary antibody. The figure is

representative of three independent transfection experiments.

69

Fig. 8. Gel mobility shift analysis of ERor mRNA splicing variant DNA-Binding
Activity. Cos7 cells were cultured in phenol red-free DMEM supplemented with
10% calf serum and transiently transfected with 10 pg of the pCMV4 expression
plasmids containing the indicated ERor cDNAs. Whole-cell extracts were prepared
by sonication in protease inhibitor-supplemented binding buffer. For each assay, 8
pl aliquots containing 30 pg of soluble protein were incubated with 6 frnol (40,000
cpm) of a 32P-labeled consensus ERE oligonucleotide. DNA-protein complexes
were resolved from fiee DNA probe on a nondenaturing 5% polyacrylamide gel.
Confirmation that the indicated band represents an authentic ERG/DNA complex is
provided by the ability of an ERor-specific monoclonal antibody (Mab-17) to
supershift this complex (compare lane 4 with lane 3). The figure is representative of
three independent transfection experiments. Position of the antibody-supershifted

complex is indicated by an arrow.

70

Ema; V”
.mmrwv

mmlm

name

V
eoeoegeeoaeooeve

Eu.

 

 

d M w h m m N. m mdcfidndwnh‘mdmdfla

9090a»
%o®¢¢:¢eeo%

71

1250-1
1000-

750-

fmollmg proteln
01
O
O

N
01
O

 

 

O

pCMV4 wt ER ERAE2 ERAE3 ERAE4 ERAES ERAEB ERAE7

Fig. 9. Ligand-binding capacity of ERor and ERor mRNA splicing variants.
Ligand binding capacity was assessed by measuring the specific association of 10
nM 3H-1713-estradiol with wt ERor and its splicing variants transiently expressed in
Cos7 cells. Cells were transiently transfected with 10 pg of the indicated pCMV4-
derived receptor expression vectors and cultured for 48 hours in phenol red-free
DMEM supplemented with 5% charcoal-treated calf serum. Cells were resuspended
in extraction buffer containing protease inhibitors and sonicated. Aliquots of whole-
cell extracts containing 200 pg of soluble protein were incubated overnight at 4°C
with 3H-labeled l7B-estradiol in the presence or absence of a ZOO-fold molar excess
of unlabeled E2. Free ligand was separated from bound ligand by treatment with
dextran-coated charcoal. The figure represents one of three independent

experiments.

72

 

pCMV4 wt ER or ERAE3

Fig. 10. Binding of a fluorescent estrogen analog (nitrile THC) by wt ERor and
ERAE3. Cos 7 cells transiently transfected with 10 pg of the wt EROt and ERAE3
expression vectors were plated on glass coverslips and cultured in phenol red-free
DMEM supplemented with 5% charcoal-treated calf serum. After 24 hours, cells
were stained in culture with 10’7 M nitrile THC. Cells were mounted and

photographed using epifluorescence microscopy. Magnification, 400x.

73

 

 

 

 

Specific Estradiol Binding
(% Maximal)

 

 

 

 

 

 

Eetredlol Concentratlon (nM)

 

Fig. l 1. Scatchard analysis of E2 binding by wt ERor and ERAE3. Cos7 cells
were cultured in phenol red-free DMEM supplemented with 5% charcoal-stripped
calf serum and transiently transfected with 10 pg of the indicated pCMV4-derived
receptor expression vectors. After 48 hours, cells were resuspended in extraction
buffer and sonicated. Aliquots of whole-cell extracts containing 200 pg of soluble
protein were incubated overnight at 4°C with various concentrations (0.1 nM - 10
nM) of 3'H-labeled 17B-estradiol in the presence or absence of a ZOO-fold molar
excess of unlabeled E2. Free ligand was separated from bound ligand by treatment
with dextran-coated charcoal. The measured dissociation constants were 0.66 nM
for wt ERor and 0.79 nM for ERAE3. The figure represents one of three independent

experiments.

74

Fig. 12. Localization of ERor and ERor splicing variants by confocal
microscopy. Cos7 cells transfected with 10 pg of the indicated pCMV4-derived
expression vector were plated on glass cover slips and cultured in phenol red-free
DMEM supplemented with 5% charcoal stripped calf serum. After 24 hours from
transfection, receptor isoforms were detected by indirect immunofluorescence
staining using an ERa-specific monoclonal antibody (Mab-l 7) and a rhodamine-
conjugated secondary antibody. Irnmunoreactivity was observed with confocal
microscopy. The upper left panel shows a representative field from control cells
transfected with empty expression vector (pCMV4) displaying minimal nonspecific
background. Bar, 10 pm.

75

 

 

 

250T r-
3‘ 200"
E Fvehiclej
‘5 ._ IE
< 150 -
l-
< a
0 100-- _,
g . .
'8
8. 50--
to

 

 

 

 

pCMV4 ERAE2 ERAE3 ERAE4 ERAES ERAEG ERAE7 wt ER

Fig. 13. Transcriptional activity of wt ERor and receptor variants on an ERE.
HeLa cells were plated at 2 x 10‘6 cells per 100 mm dish and cultured in phenol red-
free DMEM supplemented with 5% charcoal-treated calf serum. Cells were
cotransfected with 16 pg of reporter (pERE-TK-CAT) and 1 pg of the specified
ERor expression vector and treated as indicated with vehicle (0.001% ethanol) or 10
nM 17B-estradiol (E). CAT assays were normalized for equal amounts of protein.
Values are expressed relative to vehicle-treated empty expression vector, pCMV4.

The figure represents results from one of at least three transfection experiments.

77

Fig. 14. Effect of ERAE3 and ERAES on wt ERa transcription activity. HeLa
cells were plated (2 x 10'6 cells per 100 mm dish) and cultured in phenol red-free
DMEM supplemented with 5% charcoal-treated calf serum. Cells were
cotransfected with 16 pg of pERE-TK-CAT reporter gene, 1 pg of wt ERor
expression vector, and increasing amounts of expression vectors for ERAE3 or
ERAES (from O to 20 pg). The ratios of wt ERor to variant expression plasmid used
in each transfection are indicated. The total amount of DNA in each transfection
was held constant with the addition of empty expression vector, pCMV4. CAT
assays were normalized for equal amounts of protein. For each of the panels, values
are expressed relative to the maximal CAT expression induced by 17B-estradiol

treatment in the absence of coexpressed ERAE3 or ERAES.

78

% Maximal CAT Activity

 

100‘

40-

 

 
   
 

50..
410--

100?
80:-

60-!-

20:.

 

ERAE3

uvehlcle
leetradlel

 

 

 

 

1:25 1:10

wt ERor:Variant

79

1 :20

 

Fig. 15. In vitro binding of receptor variants with GST-fused wt ERa and
SRC-le fragments. A, In vitro translated 35S-methionine-labeled wt ERa, ERAE3
and ERAES were incubated with bacterial expressed GST or GST fusion protein
expressing the ERor LBD (GST-LBD) bound to glutathione-Sepharose beads in
NETN buffer supplemented with protease inhibitors. Binding was studied in the
absence or presence of 17B-est1adiol as indicated. Bound proteins were separated
on a discontinuous 10% polyacrylarnide SDS-PAGE gel and visualized by
autoradiography. The top panel represents 10% of the radiolabeled input. The
figure is representative of one of three independent experiments. B, In vitro
translated 35S-methionine-labeled wt ERa, ERAE3 and ERAES were incubated with
GST fusion proteins expressing SRC-le fragments which included amino acids 570-
780 (lanes 1 and 2), 781-988 (lanes 3 and 4), 989-1240 (lanes 5 and 6) and 1241-
1399 (lanes 7 and 8).

80

1/10 Input: ,3? g

6'
47117,};

‘ t m"'-“.'.'. rr.‘.'v Ma;.‘.rrrvut.’.n- v. - . . .7

GST-LBD Pull-down:
ERAE3 eases thRa
58m”? 5‘. * t

' l».
‘o
. e . '.t
“'1' '
.. , ,' v
‘ qr. . ,
.‘ ‘
P ’ ‘
- e . .,.»~v, . I,r.r . , . . __' . -1... _, .
.. ‘-,. . , '.‘.v .3. _,.-. , _~ .« ._. y .. . ‘ 4

B
GST-SRC1 Pull-down:

(570-780) (781-988) (989-1 240) (1241-1399)
Estredlol - g- - + - + - +

95"‘Ai/‘P
.' 1,1. 1 ..

m? “
» i.

thRor “ ' .-

m

V‘, 3rd: ”e ‘4. ,_- ,-.:‘..-.'-.‘.a. ,-.~_‘x..'¥
' ..'..’

_ f".-. . Luv-’3“ l . xv r. ,. .
«an m ‘ ‘ ‘ *' ' “ “ “W New. ?
ERAES ~. . ._ _ . »

- a.
\ mut- an». "Lafissfs‘vrniniwk‘u - ' -

81

1/10 input

 

 

N “
LIB—Tan x_E_Ta_m
GST SMRT
ERAE3 ERAES

v E Tam v E Tam s s,
GST SMRT a,

Fig. 16. In vitro binding of receptor variants with GST-fused SMRT. In vitro
translated 3SS-methionine-labeled wt ERor, ERAE3 and ERAES were incubated with
bacterial expressed GST or a GST fusion protein expressing SMRT (GST-SMRT)
bound to glutathione-Sepharose beads in NETN buffer supplemented with protease

inhibitors. Binding was studied in the absence (V) or presence of 17B-estradiol (E)

or tamoxifen (Tam) as indicated. Bound proteins were separated on a discontinuous

10% polyacrylamide SDS-PAGE gel and visualized by autoradiography. The figure

represents one of three independent experiments.

82

3. Discussion

Our efforts to functionally characterize exon-skipped ERor mRNA splicing
variants have identified two receptor isoforms that possess the ability to modulate
ERor activity on an ERE. Although their protein structure is significantly altered,
the ERAE3 and ERAES splicing variants retain many of the activities attributed to
the full-length receptor. Loss of exon 3 results in a receptor protein with an internal
deletion that lacks a major portion of the DBD and therefore prevents ERAE3 from
binding to a consensus ERE, as confirmed by gel mobility shift analysis. However,
ERAE3 retains the LBD and NLS, thereby allowing it to bind hormone with an
affinity similar to wt ERor and translocate to the nucleus.

The deletion of exon 5 causes a frame-shift mutation and results in a C-
tenninally truncated form of the receptor. Loss of the LBD predictably renders
ERAES unable to bind E2. Nonetheless, ERAES still retains the NLS, and
immunofluorescence analysis shows nuclear staining in Cos7 cells transfected with
this variant. The receptor isoforms resulting from deletion of exons 2, 5, 6, and 7 do
not bind ligand. Like ERAES, ERAE2, ERAE6, and ERAE7 are C-terminally
truncated receptor isoforms that are missing all (ERAE2) or most of the LBD
(ERAE6 and ERAE7) and many of the residues critical for ligand binding.

The loss of exon 4 causes an infrarne deletion within the receptor that leaves
the C-terminus and much of the LBD intact; however, this variant does not bind

ligand. Without exon 4 this variant is missing 112 central amino acids that contain

83

the NLS and encompass helices 1, 3 and 4 which comprise the proximal portion of
the LBD (Fig. 17). Furthermore, helix 3 makes a number of close contacts with ring
A of 17b-estradiol, including Glu353 which forms a direct hydrogen bond with the
phenolic hydroxyl group. Exon 4 therefore appears to be crucial for formation of a
functional estrogen-binding pocket. While it is unknown what impact the loss of
exon 4-encoded helices has on the conformation of the remaining portion of the
LBD (helices 5 through 12), it is likely that the overall structure of the LBD remnant
is severely distorted. The same probably holds true for ERAES, ERAE6, and ERAE7

truncation mutants.

84

Fig. 17. A depiction of the secondary structure of the ERor ligand binding
domain. The ERor ligand binding domain comprises helices 1 through 12. Labeled
arrows specify exon junctions. Ligand contacts are indicated in helices 3, 6, 8, and
11. Sites for receptor dirnerization and coactivator contacts are also shown (black

and hatched bars, respectively).

85

 

88: 5 252.0: J1 :1... .

 

 

 

« a
+5.... a 11.x u

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a a .
:1... m - 10.... m i h--.
96: «B .5620: a ..... v.33.
III... a
19.x m , zo=x q
s _ a:
1e=x m N 88: c: o:
I
12.x .3
NH |
:2... 2
In..." an
. d .625 3:808
98: NE .5520: I 2335.0: 2.1231
a 33.222 02.8."?

 

 

 

86

Rather than serving to stimulate transcription on a consensus ERE, results
from transient transfecion experiments in HeLa cells that combine either ERAE3 or
ERAES with wt ERor and an ERE-driven reporter gene indicate that these isoforms
actually function to inhibit transcriptional activation by wt ERa. The dominant
negative character of ERAE3 and ERAES suggests that, like wt ERor, these variants
are able to interact with at least one component of the ERE-directed transcription
complex in a manner that disrupts positive gene regulation by wt ERor. Based on
gel mobility shift assay analysis, it is unlikely that transcriptional interference by
these variants involves binding to an ERE to the exclusion of wt Ear. Our DNA
binding analysis indicates that ERAES can bind only weakly to DNA, and only when
the formation of this complex is stabilized by the addition of a bivalent antibody.
The role of the antibody in this case is presumably to substitute for the missing
dimerization interface and to tether the receptor subunits together in a form more
able to interact with DNA. DNA binding by ERAE7 similarly requires the addition
of antibody, but this binding is even less efficient than binding by ERAES. The
relevance of ERE binding by ERAE7 is further reduced with the consideration that
ERAE7 does not translocate to the nucleus. Interestingly, a correlation exists among
the ERa variants between their ability to translocate to the nucleus and their
transcriptional inhibitory effect on wt ERor activity in mammalian cells. As of yet,
no clear function has been established for the ERAE7 variant in mammalian cells,

despite an earlier report that ERAE7 is a dominant inhibitor of wt ERa function in

87

yeast (Fuqua etal., .1992). This is noteworthy since a number of quantitative studies
indicate that ERAE7 generally represents the most abundant of the ERor splicing
variants in breast tumors (summarized in Murphy et al, 1997).

It has previously been reported that ERAE3 can prevent wt ERor fiom
binding to DNA (Miksicek et al., 1993). That ERAE3 inhibits both DNA complex
formation and transactivation by wt ERa suggests that the potential targets of
interaction by ERAE3 may include contact with the receptor or competitive
interaction with coactivators or other receptor-associated factors. ERor function may
be disrupted when ERAE3, which lacks the DBD but retains the hormone-inducible
dirnerization domain, forms mixed dimers with wt ERor that are inefficient at
binding stably to DNA. We are able to show that, in the presence of E2, ERAE3 (but
not ERAES) can form a stable complex with the LBD of ERor fused to GST attached
to glutathione-Sepharose beads. This is consistent with a model for direct inhibition
of the DNA binding activity of the full-length receptor by ERAE3.

Experiments using fiagments from SRC-le fused to GST indicate that both
ERAE3 and ERAES can bind a nuclear receptor coactivator. Similar to the pattern
of wt ERor interaction with SRC-le, in vitro translated ERAE3 is able to associate in
an Ez-dependent manner with two regions of the steroid coactivator SRC-le (amino
acids 570-780 and 989-1240). This agrees with previous reports that describe three
conserved nuclear receptor-binding motifs (LXXLL) within the 570-780 amino acid

region and a distinct site for AF] interaction within the 989-1240 amino acid

88

fiagment (Kalkhoven et al, 1998; Webb et al., 1998). The major site for SRC-l
interaction with ERa corresponds to the AF 2 domain (Feng et al., 1998), a region
that is retained only in the ERAE3 variant. Isoforrns of SRC-l are potent enhancers
of agonist-bound ERor and are required for its full transcriptional activity in viva
(Ofiate et al., 1995; Xu et al., 1998). Transfection experiments in Ez-treated HeLa
cell cultures demonstrate that coexpression of mutants containing the C-terminus of
ERor can attenuate ERor-dependent gene expression and that this decreased activity
can be overcome with simultaneous over expression of the SRC-l-related
coactivator TIF2 (V oegel et al., 1996). These results suggest that coactivators are
limiting factors for which the receptors are competing and that ERAE3, like wt ERor,
is a target for SRC-l binding.

In a surprising result fi'om cotransfection studies using engineered mutants of
ERor, maximal expression of an ERE-containing reporter gene could be observed
when SRC-l was transfected simultaneously with separate N- and C-tenninal
fragments of ERor, containing the AF l/DBD and the LBD/AF 2 regions, respectively
(McInery et al., 1996). These results suggest that separate AF 1- and AF 2-containing
ERa polypeptides can interact in a transcriptionally productive manner, provided
they are brought together by SRC-l. Furthermore, they provide an initial indication
that SRC-l interacts separately and perhaps directly with both the AF 1 and AF 2
domains. More support for this notion is provided by our observations that ERAES

binds the SRC-le amino acid fragment 989-1240 in solution. This result suggests

89

the possibility that the inhibitory function of ERAES, which itself is relatively
inefficient at binding DNA or activating transcription through an ERE, most likely
results from competition with wt ERor for interaction with SRC-l. ERAE3, ERAES
and wt ERor are also found to interact with the corepressor SMRT. In agreement
with previously published results, we find that wt ERor binds the SMRT corepressor
in a ligand-independent manner (Smith et al., 1997). Results fiom in vitro binding
studies Show a similar degree of wt ERor and SMRT binding in the absence or
presence of E2 or tamoxifen. Although ERAE3 also binds SMRT in the presence of
tamoxifen and E2, increased binding was consistently observed in vehicle treated
samples. These results indicate that the interaction between ERAE3 and corepressor
may be enhanced in cells when ligand is unavailable. Thus far there is little reported
evidence to suggest what consequence corepressor binding by wt ERor or ERor
splicing variants may have on modulating transcription. However, it is conceivable
that the activity of some promoters may be affected when ERAE3 or ERAES bind
corepressor and thus make it less available to associate with other nuclear receptors.
The predicted effect would be a positive influence on corepressor-sensitive
promoters that are not directly activated by ERAE3 or ERAES. For example,
ERAE3 and ERAES are likely to interfere with basal repression of promoters by TR
or RAR, which is effected by receptor association with corepressors in the absence
of ligand (Chen et al., 1995; Horlein et al., 1995).

Functional analysis of ERor exon-skipped splicing variants shows that, while

90

several of the isoforms are functionally incapacitated by their deletions, two of the
variants clearly retain activities assigned to the full-length receptor. The ERAE3 and
ERAES variants represent stable receptor isoforms that, like wt ERor, localize
efficiently to the nucleus where they may interact with proteins known to associate
with the transcription apparatus. However, when acting through a consensus ERE,
these variants completely lack (ERAE3) or show only weak (ERAES) transcriptional
stimulatory activity, consistent with their poor DNA binding ability. On the
contrary, both variants serve to inhibit the ability of coexpressed wt ERor to promote
transcription of ERE-containing genes. At the same time, the ability of ERAE3 (and
presumably also ERAES) to interact productively with nuclear receptor coactivators
or corepressors gives these ERor splicing variants the potential to stimulate or

otherwise modulate gene expression.

91

IV. ERor Splicing Variants Lacking Exons 3 or 5 Display Promoter-

Specific Transcriptional Effects Through Nononsensus ERES

1. Introduction

A wide variety of variant ERor mRNAs are recognized in the literature
(Dowsett et al., 1997; Fuqua et al., 1993; Murphy et al., 1997; van Dijk etal.,
1997). Those most commonly observed harbor a precise deletion of one of the
internal exons that contribute to the structure of the full-length ERor protein.
Despite their discovery nearly a decade ago (Fuqua et al., 1991; Wang and
Miksicek, 1991), relatively little is known about the functional role of ERa splicing
variants. The most prominent activity described for two of the splicing variants
(ERAE3 and ERAES) is to interfere with the transcriptional activity of wt ERor on
ERE-containing genes, where they behave as dominant-negative mutants (see Fig.
14). The inhibitory activities of ERAE3 and ERAES are likely to result in part from
their ability to sequester nuclear receptor coactivators or other limiting transcription
factors into a non-productive complex. In addition, ERAE3 may block binding of wt
ERor to DNA by forming functionally inactive, mixed dimers (see Fig. 15A, Wang
and Miksicek, 1991).

The in-frame deletion of exon III harbored by ERAE3 causes partial loss of
its DNA-binding domain DBD, but leaves the LBD intact. Predictably, ERAE3 does

not bind an ERE; yet its ability to bind and potentially respond to hormones is

92

maintained. Conversely, the LBD is severely truncated in ERAES resulting in the
complete loss of hormone binding. Although ERAES retains the DBD, its DNA-
binding activity is impaired as a consequence of the loss of helix 11 within the LBD,
which contains critical contact sites for subunit dirnerization (Brzozowski et al.,
1997 ; Kumar and Chambon, 1988). While the ERAES splicing variant is reported to
display modest transcriptional activity in certain cell- and promoter-specific contexts
(Chaidarun etal., 1998; Fuqua et al., 1991; ), its relatively poor binding to the
palindromic ERE suggests that it is unlikely to be a significant effector of
transcription through a consensus ERE. Similarly, the remaining exon-skipped ERor
splicing variants (ERAE2, ERAE3, ERAE4, ERAE6, and ERAE7) appear to
completely lack transcriptional stimulatory activity when tested in transient
transfection assays with an ERE-containing reporter plasmid (Fig. 13). In the
studies detailed below, ERAE3 and ERAES splicing variants receive significant
attention for the reason that these receptor isoforms retain some of the functional
characteristics of wt ERor. Most notably, despite their mutations, they localize to the
nucleus, associate with the steroid receptor coactivator SRC-le, and ultimately are
observed to modulate gene transcription (see Figs. 12, 14, 15, and below).

In recent years, a novel pathway for regulation of transcription by estrogen
receptors (both ERor and ERB) has been described that involves cooperation of the
receptors with the AP-l transcription factors, c-jun and c-fos (Gaub et al., 1990; Uht
et al., 1997). This has been broadened to include additional genes for which

estrogen regulation was mapped to binding sites for the Spl (Duan et al., 1998) and

93

ATF2 transcription factors (Sabbah et al., 1999), which bind to GC-rich and CRE
(cyclic AMP response element)-like motifs, respectively. An important feature of
these nonclassical pathways for estrogen action is that functional domains within the
receptor that are crucial for transcriptional activation through a consensus ERE are
in some cases dispensable for ERor activity on AP-l, Spl , or CRE-regulated
promoters. Mutational analysis revealed that the DBD was not always required for
ERor-dependent expression of these genes (Duan et al., 1998; Gaub et al., 1990;
Sabbah et al., 1999). Clearly, the mechanisms of ERor transcriptional activity and
DNA targeting are complicated by these reports. This led us to consider the
possibility that the nuclear ERor splicing variants (specifically ERAE3 and ERAES)
may function as transcriptional regulators through noncanonical hormone response
elements as opposed to consensus ERES. The following results describe four
promoters lacking an ERE that are efficiently stimulated by either ERAE3 or
ERAES. In some instances the activity is shared by wt ERor, while in other cases
gene activation is limited to the ERAE3 or the ERAES variants alone. The
promoters tested in these studies attracted our interest because they are reported to
be ERor regulated and/or contain an AP-l motif. They include the chicken
ovalbumin, human collagenase and human IGF-l promoters, and a synthetic

promoter consisting of three consensus AP-l elements.

94

2. Results

In agreement with previously published reports (Wang and Miksicek, 1991;
Rea and Parker, 1996), Fig. 13 shows that ERor splicing variants do not effectively
promote gene expression from a palindromic ERE . However, this understanding
does not preclude the possibility that transcriptional effects of ERa variants may be
directed at promoters with nonclassical hormone response elements. To further
investigate the role of ERor splicing variants in the regulation of gene expression, we
used a cell culture transfection system to assay for receptor activity on a variety of
reporter gene constructs that may be targets for regulation by ERor even though they
lack a consensus ERE. For preliminary studies, HeLa cells were transfected with
each of these reporters together with wt ERor or variant ERor expression plasmids.
HeLa cells express minimal endogenous ERor and yet possess a transcriptional
milieu that enables reporter plasmids to respond efficiently to transfected receptor.
Paired cultures were treated with 10 nM E2 in combination with 20 nM PMA, or
with vehicle alone. This was done to ensure that an activating ligand was present, if
necessary, and that AP-l factors were maximally stimulated in the event that

promoter activation required stimulation of PKC.

95

2.1 Like wt ERa, ERAE3 activates the ovalbumin gene promoter

Although there is no consensus ERE in the ovalbumin promoter, it has been
shown that this gene is regulated by wt ERor. Mapping of the ERa regulatory
element places it at a critical AP-l site near the transcription start site (Schweers et
al., 1990; Tora et al., 1988). We performed cotransfection experiments in HeLa
cells using vectors expressing wt ERor or the exon-skipped variants ERAE2 through
ERAE7 and a CAT reporter gene construct, vaalb-CAT driven by a fragment of
the ovalbumin promoter (-1342 to +7 relative to transcription start site) described to
encompass much of the regulatory sequence of this gene (Schweers et al., 1990;
Tora et al., 1988). Results from these experiments indicate that both wt ERor and
ERAE3 support inducible gene expression from the ovalbumin promoter (Fig. 18).
The remaining single exon-skipped variants are transcriptionally inactive on this
reporter construct (Fig. 18, inset). For wt ERa, this corroborates previously
published reports (Gaub et al., 1990; Tora et al., 1988). Maximal activity was
measured in cultures treated with both phorbol 12-myristate, 13-acetate (PMA, a
phorbol ester) and E2, where a 16-fold induction was observed. Like wt ERor,
ERAE3 reproducibly induced this reporter, despite its lack of an intact DBD. While
the induction by ERAE3 (averaging 9-fold) is less than that supported by wt ERa,
the activity of ERAE3 equaled and occasionally exceeded that of the intact receptor
in several individual experiments, confirming that this variant can be a potent

inducer of transcription (Fig. 18). In both cases cotreatrnent with PMA and E2 is

96

highly synergistic as E2 treatment alone has no significant effect and PMA treatment
alone supports only modest induction for wt ERor (2.5-fold relative to vehicle
control). Tamoxifen treatment of wt ERor- or ERAE3 -transfected cultures, either
alone or together with phorbol ester, also had no significant effect on vaalb-CAT
expression. This contrasts with the stimulatory activity of tamoxifen observed on
other AP-l containing estrogen-responsive reporter genes (Webb et al., 1995). In
control cells transfected with an empty CMV expression vector, treatment with
PMA yielded negligible reporter gene activity. This suggests that, in the absence of
ERor, activation of endogenous AP-l alone is not an effective inducer of
transcription from the ovalbumin promoter in these cells. To confirm that wt ERor
and ERAE3 cooperate with AP-l factors to regulate transactivation of the ovalbumin
promoter, we measured vaalb-CAT expression in HeLa cells cotransfected with
both a receptor isoform and a c-jun expression vector. Transcriptional activity of wt
ERor and ERAE3 supported by PMA and E2 cotreatrnent was enhanced by c-jun over
expression. While the presence of endogenous AP-l tended to obscure the
synergism between c-jun and wt ERor in this system, the combined effects of these
transcription factors were slightly more than additive. Enhanced activation was also
observed when c-jun was coexpressed with ERAE3 (Fig. 19). Exogenous c-jun
alone elicited only a modest response to PMA and E2 treatment. These
observations, combined with the dual requirement for activation of both AP-l and

ERa, strongly suggest that these factors are acting cooperatively on the ovalbumin

97

promoter. Further consideration of the mechanism of activation by wt ERor and
ERAE3 excludes a requirement for direct binding of receptor to the ovalbumin
promoter. ERAE3 is devoid of DNA-binding activity, arguing that activation of
ovalbumin by these receptors must be mediated indirectly through specific protein-
protein interactions. Figure 20 demonstrates that ERor interacts with c-jun and c-fos
in solution and suggests that the receptor may be recruited to the ovalbumin
promoter by interacting with AP-l factors. Binding assays with GST-fused ERor
and 35S-labeled c-jun and c-fos (translated individually or in combination in vitro)
demonstrate that, independent of E2 treatment, ERor binds c-jun and c-fos under all
treatment conditions tested. Although ERa variants were not specifically included
in these binding studies, Webb and coworkers have reported an interaction between
c-jun and the isolated N-terminus of ERot, a region conserved in both ERAE3 and
ERAE5 (Webb et al., 1995). We therefore infer that ERAE3 and ERAE5, like their

intact counterpart, are also able to interact with AP-l homo- and heterodimers.

2.2 Both ERAE3 and ERAE5 activate the Coll(-73)Luc reporter

The promoter of the human collagenase gene has received considerable
attention as a target for nonclassical regulation by both ERor and ERB. Collagenase
is a member of a family of matrix metalloproteases, which are regulated similarly

and demonstrate enhanced expression in response to injury and during inflammatory

reactions (Angel et al., 1986; Herrlich et al., 1986; Whitharn et al., 1986).

98

Collagenase gene expression is elevated in certain tumor cells and may contribute to
advancement of metastasis (Liotta 1986). In cultured fibroblasts and skin cells
treated with carcinogens and tumor promoting agents, induction of collagenase is
mediated by phorbol ester treatment (Angel et al., 1986; Angel et al., 1985). A
short region of the collagenase promoter (-73 to +63 relative to the transcription start
site) which harbors an AP-l element is reported to direct estrogen- or tamoxifen-
regulated gene expression of a luciferase reporter gene in a variety of cell lines
(Paech et al., 1997; Uht et al., 1997 ; Webb et al., 1995). In contrast to published
reports, we find that wt ERor behaves no differently on the pColl(-73)Luc reporter
than the empty pCMV4 control vector (Fig. 21). In both cases, a relatively modest
(approximately 6-fold) increase is observed that depends on activation of
endogenous AP-l by PMA treatment. This same promoter is also activated by
ERAE3 and to a lesser extent by ERAE5 in response to combined treatment with
estrogen and PMA (Fig. 21). Like wt ERor, the remaining splicing variants (ERAE2,
ERAE4, ERAE6, and ERAE7) were without significant effect. Substantiating the
necessity for PMA treatment and the role of AP-l in receptor regulation of the
collagenase promoter, coexpression of c-jun is observed to enhance ERAE3- and
ERAES- induced pColl(-73)Luc reporter gene expression (Fig. 22).

Additional analysis indicates that maximal activation of pColl(-73)Luc by
ERAE3 requires cotreatrnent with both E2 and PMA (Fig. 23). Phorbol ester alone
(or in combination with ER ligands) supports a modest induction of the collagenase

reporter that is independent of coexpressed receptor. Regulation pColl(-73)Luc by

99

ERAE5 similarly required activation of endogenous AP-l by PMA, but was
independent of E2 treatment, consistent with the inability of ERAE5 to bind ligand.
The partial agonist tamoxifen was without effect on induction of this reporter by
ERAE3, either alone or in the presence of PMA. Treatment with E2 or tamoxifen
alone or together with PMA had no effect on the activity of pColl(-73)Luc
cotransfected with wt ERor (or pCMV 4) compared to PMA alone (Fig. 23, inset).
The human collagenase promoter therefore demonstrates a difference in behavior
compared to the vaalb-CAT reporter which was not induced by ERAE5, but was
activated by both ERAE3 and wt ERor in cells cotreated with E2 and PMA.

Due to the stimulatory behavior of ERAE3 and ERAE5 on the ovalbumin and
collagenase promoters, both of which contain AP-l sites in their native promoter
contexts, it was of interest to examine the effects of the various ERor splicing
variants on a reporter plasmid containing an AP-l site in the context of a promoter
not otherwise regulated by ERor or estrogens. For this purpose, we used p(AP-1)3-
TK-CAT, which contains three consensus AP-l sites in tandem upstream of a
minimal thymidine kinase promoter from Herpes simplex virus. Somewhat
unexpectedly, ERAE5 (but not wt ERa, ERAE3, or any of the remaining splicing
variants) was able to induce this promoter approximately 7-fold in HeLa cells (Fig.
24) although this difference failed to reach statistical significance (P S 0.1). Like
the pColl(-73)Luc reporter described above, ERAE5 activity on the p(AP-1)3TK-

CAT reporter required PMA treatment and did not depend on the presence of E2

100

(Fig. 24, inset). It is clear that by some reports ERa is an estrogen-dependent
inducer of transcription through a consensus AP-l element (Gaub et al., 1990; Webb
et al., 1999). These results demonstrate that the activity of ERor on AP-l containing
promoters is a complicated and conditional effect. The data presented here agree
well with a previous report demonstrating that a C-terminally truncated receptor
lacking the LBD is more active than E2-liganded wt ERor on an analogous AP-lfl" K
reporter construct (Uht et al., 1997). Also, it is evident that the context of an AP-l
site within the promoter is critical for conferring its ability to respond, and the

degree of that response, to wt ERor or any of the ERor splicing variants.

2.3 ERAE5 activates the human IGF-l promoter

Insulin-like grth factor I (IGF-l) is expressed in and effects cellular
growth in many tissues. In general, growth hormone acts as the major regulatory
factor for the IGF-l gene; however, several other factors (including thyroid
hormone, epidermal grth factor, parathyroid hormone and estrogen) function to
modulate IGF -1 expression (Ernst et al., 1989; McCarthy et al., 1989; Murphy et al.,
1990; Stewart et al., 1996; Wolf et al., 1989). Induction of IGF -1 is stimulated by
E2 in the uterus of ovariectomized / hypophysectomized rats (Murphy et al., 1990;
Murphy et al., 1988) and in cultured rat osteoblast cells (Ernst et al., 1989). In
transient cotransfection studies a defined region of the chicken IGF-l promoter
directed reporter gene induction by liganded ER (Umayahara et al., 1994). Like the

ovalbumin and collagenase promoters mentioned above, the IGF-l promoter lacks a

101

consensus ERE sequence (Kim et al., 1991; Hall et al., 1992). Further complicating
this picture are independent data showing that in primary fetal rat osteoblasts E2 can
block the increase in IGF-l synthesis caused by parathyroid hormone (PTH) or
prostaglandin E2 (PGE2) (McCarthy et al., 1997). It has been shown that this
hormonal regulation of lGF-l expression by PTH and PGE2 is mediated by a binding
site for CCAAT/enhancer-binding protein-delta (C/EBP-fi) located between
positions +202 and +209 within exon 1 of this gene (Umayahara et al., 1999). For
the chicken promoter, a positive ERor regulatory element was mapped to an AP-l
motif essential for both E2 and phorbol ester-stimulated gene transcription
(Umayahara etal., 1994). Although there is a great deal of homology among IGF -1
promoters of various species, this AP-l site is not conserved in mammals (Hall et
al., 1992; Kim et al., 1991; Umayahara et al., 1994). The human IGF-l promoter is
thus different from the promoter examples described above in that it lacks any
consensus binding sites for either AP-l or Spl that might represent likely targets for
regulation by ERor. However, Nagaoka and colleagues report that PKC activation
by PMA treatment increases the rate of IGF-I transcription in the human
macrophage-like cell line U937 (Nagaoka et al., 1990). This suggests that AP-l
factors or another PKC-responsive transcription factor positively regulate the human
[GP -1 gene. Furthermore, they may be functioning at a nonconsensus AP-l or
ATF2 binding site.

When cotransfected into HeLa cells, treatment with PMA induced

expression of a luciferase reporter plasmid containing 1.63 kb of sequence upstream

102

from the human IGF-l transcription start site (pIGF1-1630-Luc) (Fig. 25). This
same reporter was unresponsive to wt ERor transfection. IGF-l promoter activity is
strongly and constitutively induced by ERAE5, greater than 30-fold compared to
cells transfected with control plasmid. PMA treatment of ERAE5 transfected cells
has no additional stimulatory effect on ERAE5 induction of this reporter gene. In
fact, PMA treatment slightly decreases ERAE5 activity (Fig 25). This argues against
the involvement of PKC or PKC-activated transcription factors such as AP-l in the
mechanism of ERAE5 induction of human IGF-l expression. In an attempt to
determine where within the IGF-l promoter the ERAE5 regulation is directed,
reporter gene transfection assays were performed with a series of 5'-deletion mutants
of the human IGF-l promoter. ERAE5 induced expression from reporter gene
constructs containing -1630, -926, -592, and -235 bp (the number reflects
nucleotides relative to the transcription start site) of the human IGF -1 promoter (Fig.
26). However, induction of the shortest length of promoter tested (the -235 bp
fragment) was significantly less compared to the others. Unexpectedly, ERAE5 was
most active on the -592 bp fragment. These results indicate that repressor elements
may reside in the human IGF -1 promoter between -l630 and -592 and that a highly
responsive ERAE5 enhancer element is located between nucleotides -592 and -235.
It should be noted, however, that overall activity was progressively lost with each
truncation of the IGF -1 promoter (Fig. 26, inset). Although its induction by ERAE5

was less than that observed for other promoter lengths, the -1630 bp promoter

103

supported maximal reporter gene expression . Efforts to transfer ERAE5-
responsiveness onto a heterologous promoter were unsuccessful, suggesting that
activation of IGF-1 expression by ERAE5 may require complex interactions between

transcription factor binding sites throughout the IGF-l promoter.

104

Fig. 18. Activation of the ovalbumin gene promoter by wt ERor and ERAE3.
HeLa cells were plated (2 x 10" cells per 100 mm dish) and cultured in phenol red-
free DMEM supplemented with 5% charcoal-treated calf serum. Cells were
cotransfected with 16 pg of reporter (vaalb-CAT) and 1 pg of the specified ERor
expression vector. Cultures were treated as indicated: vehicle control (0.001%
ethanol and 0.002% DMSO); 20 nM phorbol 12-myristate, 13-acetate (PMA); 10
nM WIS-estradiol (E); E + PMA; and 100 nM tamoxifen (T); or T + PMA. CAT
assays were normalized for equal amounts of protein. Values are expressed relative
to vehicle-treated empty expression vector, pCMV4. Error bars represent the SEM
of three independent transfection experiments (*, P < 0.001). Inset, a transient
transfection reporter gene assay was used to test the activity of each of the single

exon-skipped variants on vaalb-CAT.

105

 

 

 

 

Relative CAT Activity

20-

15-

10-

 

 

D
0

d
O

Ivehicle '7
’ DE+PMA

1

 

Relative CAT Activity

 

 

 

 

 

0 J a
poll“ [M32 IME: [M24 ENE! 3M3. INLET M In

 

 

 

.vehicle
IPMA
EJE
IE+PMA
UT
T-«i-PMA

 

 

 

 

P < 0.001

 

 

 

 

 

 

 

106

 

 

 

4o-- —

30 .. lvehicle P<0-05
DE-i-PMA I I.
2... I l

10-- * I
0._£.~_£4.j+‘— L|

 

Relative CAT Activity

 

 

 

 

 

ERAE3 "’ " + + - "
thRa _ _ — _ + 1"
cJun - + " + " +

Fig. 19. Effect of c-jun over expression on wt ERor and ERAE3 activation of the
ovalbumin promoter. HeLa cells were plated (2 x 10* cells per 100 mm dish) and
cultured in phenol red-fiee DMEM supplemented with 5% charcoal-treated calf
serum. Cells were cotransfected with 16 pg of reporter (pERE-TK-CAT), 1 pg of
the specified ERor expression vector and 2 pg of c-jun expression vector where
indicated. Cells were treated with vehicle or 10 nM l7B-estradiol and 20 nM
phorbol lZ-myristate, 13-acetate (E + PMA). CAT assays were normalized for
equal amounts of protein. Values are expressed relative to vehicle-treated empty
expression vector, pCMV4. Error bars represent the SEM of four independent

experiments (*, P < 0.05; ns, not significant).

107

1110 input: GST-Elia Pull-down:

.- + II + ambl

W “a? Al I'?
. , ,1
.3 "are; ‘

s . “.qemnv- n—vb-

  

edun

   

. '. .-‘s<-;..s;.ss:‘s. “Zia
. Q1; c-fos

W'- e a .. I
3 4

firms-hits
.I" ' 2:1:
I' . '.

c40n
-i- c-foe

   

.—
.
. ‘4'7I“.“".‘§_‘F. .,

1 2
GST GST-ER

 

 

Fig. 20. Binding of c-jun and c-fos with GST-fused wt ERa. Right panel, in
vitro translated 35S-methionine-labeled c-jun and c-fos were incubated
independently and in combination (c-jun + c-fos) with bacterial expressed GST
(lanes 1 and 2) or GST-ERor (a GST fusion protein expressing wt ERor, lanes 3 and
4). GST proteins were bound to glutathione-Sepharose beads in NETN buffer
supplemented with protease inhibitors. Binding was studied in the absence or
presence of 17B-estradiol as indicated. Bound proteins were separated on a
discontinuous 10% polyacrylarrride SDS-PAGE gel and visualized by
autoradiography. Left panel shows 10% of the radiolabeled input. The figure

represents results from one of three independent experiments.

108

P5 0.1

 

 

 

 

 

 

80 T T
g. 60... ~-
.2
‘6
< Psoas
i '
a 40... l ,
'8 I vehicle
3 r: E-i-PMA
m Pseos
.2 l
E
g 20--

 

 

pCMV4 ERAE2 ERAE3 ERAE4 ERAES ERAEB ERAE7 wt ERG

Fig. 21. Activity of ERa splicing variants on a cotransfected human
collagenase reporter plasmid. HeLa cells were plated at 1 x 10‘6 cells per 60 mm
dish and transiently transfected with 10 pg of the pColl(-73)Luc reporter gene and
0.5 pg of receptor expression vector. Cells were cultured in phenol red-lice
DMEM supplemented with 5% charcoal-heated calf serum and treated with vehicle
or 10 nM WIS-estradiol and 20 nM phorbol 12-myristate, l3-acetate (E + PMA).
Luciferase activities were normalized for protein and results are eXpressed relative
to vehicle-treated empty expression vector (pCMV4) controls. Error bars represent
the SEM of three independent transfection experiments.

109

 

 

 

 

 

r P 5' 0'01 lIIDCMWI—P-é—Oi-I (poiiivnlféil.|
chm

300 [ f I
3. P g 0.1
E 250 .. l'_ 'l
E .T.
g 200 1-
9
£3 nvehicle
8 15° " o E+PMA
._l
'22, 100 «-
E
Q)
CE 50 ..

o . I

 

 

 

pCMV4 c-jun ERAE3 ERAEa ERAE5 ERAE5 thFia thRa
+ c-jun + c-jun + c-jun

Fig. 22. Effect of c-jun coexpression on ERAE3 and ERAE5 transactivation of
the pColl(-73)luc reporter gene. HeLa cells were plated (l x 10'6 cells per 60 mm
dish) and transiently transfected with 10 pg of reporter gene, 0.5 pg of receptor
expression vector, and 2 pg c-jun expression vector where indicated. Cells were
cultured in phenol red-free DMEM supplemented with 5% charcoal-treated calf
serum and treated with vehicle or 10 nM 17B-estradiol and 20 nM phorbol 12-
myristate, 13-acetate (E + PMA). Luciferase activities were normalized for protein
and results are expressed relative to vehicle-treated empty expression vector
(pCMV4) controls. Error bars represent the SEM of three independent transfection

experiments.

110

Fig. 23. Induction of the pColl(-73)Luc reporter gene by ERAE3 requires 1713-
estradiol and PMA cotreatment. HeLa cells were plated at 1 x 10‘6 cells per 60
mm dish and transiently transfected with 10 pg of reporter gene and 0.5 pg of
receptor expression vector. Cells were cultured in phenol red-free DMEM
supplemented with 5% charcoal-treated calf serum and treated as indicated: vehicle
control; 20 nM phorbol 12-myristate, 13-acetate (PMA); 10 nM 17B-estradiol (E); E
+ PMA; and 100 nM tamoxifen (T); or T + PMA. CAT assays were normalized for
equal amounts of protein. Values are expressed relative to vehicle-treated empty
expression vector, pCMV4. Luciferase activities were normalized for protein and
results are expressed relative to vehicle-treated empty expression vector (pCMV4)
controls. Error bars represent the SEM of three independent experiments. The
indicated significance C“, P 50.1; **, P _< 0. 05) is relative to corresponding vehicle
control. Inset, a representative transient transfection assay was used to determine
the transcriptional activity of wt ERa on the pColl(-73)Luc reporter gene in the

presence of tamoxifen.

lll

 

 

 

 

 

MEI-'1

pCMV4

 

25 V

2.3.3 823.2... 2:23.

 

 

 

 

 

 

100 1-

 

 

 

 

 

 

 

 

T+PMA

 

uE-i-PMA
IT

Ivehlcle

DE
BPMA

 

0.-

.1 u .r
0 0 0 0
8 6 4 2

>=>=o< 03.2.03 5...»..me .x.

 

 

Fig. 24. Transcriptional activity of ERor splicing variants on the p(APl)3TK-
CAT reporter plasmid. HeLa cells were plated at 1 x 10“5 cells per 60 mm dish
and transiently transfected with 0.5 pg of pCMV4 vectors expressing the indicated
receptor isoforms along with 10 pg of the p(AP1)3TK-CAT reporter gene. Cells
were cultured in phenol red-free DMEM supplemented with 5% charcoal-treated
calf serum and treated with vehicle or 10 nM 17B-estradiol and 20 nM phorbol 12-
myristate, 13-acetate (E + PMA). CAT assays were normalized for equal amounts
of protein. Values are expressed relative to vehicle-treated empty expression vector,
pCMV4. Error bars represent the SEM of three independent experiments. Inset, a
transient transfection assay was used to determine ligand requirements for ERAE5

activation of the p(AP1)3TK-CAT reporter gene.

113

 

 

Relative CAT Activity

 

 

 

   
  

 

33513:: I vehlcle

 

525152 IT+PMA

 

 

 

 

 

I vehicle
a E+PMA

 

flflfl

pCMV4 ERAE2 ERAE3 ERAE4 ERAE5 ERAE6 ERAE7 wt ERa

 

114

 

 

P5031
iooTI '
P3035

 

\l
01
l

   

a vehicle
I E+PMA

 

%Maxlmum Luciferase Activity
N U!
01 O

 

 

pCMV4 ERAE2 ERAE3 EHAE4 ERAES ERAEG ERAE7 wt ERa

Fig. 25. Transcriptional activity of wt ERa and wt ERG. splicing variants on
the human lGF-l promoter. HeLa cells were plated at l x 10"5 cells per 60 mm
dish and transiently transfected with 0.5 pg of pCMV4 vectors expressing the
indicated receptor isoforms along with 10 pg of the IGF -1 promoter driven reporter
gene pIGF1-1630-Luc. Cells were cultured in phenol red-free DMEM
supplemented with 5% charcoal-treated calf serum and treated with vehicle or 10
nM WIS-estradiol and 20 nM phorbol 12-myristate, 13-acetate (E + PMA).
Luciferase activities were normalized for protein and results are expressed relative
to vehicle-treated empty expression vector (pCMV 4) controls. Error bars represent
the SEM of three independent experiments.

115

Fig. 26. ERAE5 induces the expression of a reporter gene driven by a series of
truncated IGF—l promoters. HeLa cells were plated at 1 x 10‘6 cells per 60 mm
dish and transiently transfected with 0.5 pg of the pCMV4 derived ERAE5
expression vector and 10 pg of a luciferase reporter gene construct driven by various
lengths of the IGF-l promoter. Cells were cultured in phenol red-free DMEM
supplemented with 5% charcoal-treated calf serum. The length of [GP -1 promoter
used in each luciferase reporter construct is specified by the number of base pairs
upstream from the IGF-l transcription start site: -1630, -926, -592 and -235 bp.
Luciferase activities were normalized for protein and results are expressed relative
to empty expression vector (pCMV4) controls for each reporter construct. Error
bars represent the SEM of at least three independent experiments. The inset shows
absolute promoter activity in control cells cotransfected with the empty pCMV4
expression vector compared to the enhanced activity supported by ERAE5

cotransfection from one representative experiment.

116

 

 

 

 

ERAE5

 

 

[3592
I235

I1630
I926

 

pCMV4

 

 

 

 

 

 

 

 

mgmzcwgwomwzn o W. M. R — I _:  ,, m

 

 

- n p p . p

. . u u u .1 .
4 0 6 2 8 4

2 2 1 1

c2520:- Eon.

 

ll7

 

3. Discussion

The traditional view of estrogen action depicts ERa as the central player in
target gene selection by virtue of its ability to bind to DNA containing a close match
to the ERE consensus sequence (Tremblay et al., 1997). It has been necessary to
modify this view to accommodate the newly discovered ERB, which binds to the
same palindromic ERE sequence (Chien et al., 1994). Recent evidence also
indicates that many estrogen-regulated genes may contain noncanonical hormone
response elements organized around AP-l or SP1 sites. Among these targets are
genes such as IGF-l, cyclin D1, c-fos, c-myc, RARa, collagenase, and cathepsin D
that play important roles in controlling cell grth and tissue remodeling (Duan et
al., 1998; Gaub et al., 1990; Krishnan et al., 1994; Pilihowska et al., 1997; Sabbah
et al., 1999; Sun et al., 1998). It appears that the structural requirements for ERa to
regulate many noncanonical hormone response elements are less stringent than
transcriptional stimulation through a consensus ERE (Duan et al., 1998; Gaub et al.,
1990; Sabbah et al., 1999). For this reason, we have examined the behavior of ERa
splicing variants on several promoter constructs containing noncanonical regulatory
elements. In agreement with previously published reports, our studies indicate that
wt ERa stimulates ovalbumin promoter activity. The actvity of wt ERa and
engineered ERa mutants on this promoter is mapped to a critical AP-l motif which
when mutated causes the loss of ERG responsiveness (Gaub et al., 1990; Tora et al.,

1988). Maximal activation of the ovalbumin promoter by wt ERG requires estradiol

”8

and phorbol ester co-treatment. This induction by wt ERa is enhanced with c-jun
cotransfection. The dual requirement for estradiol and a phorbol ester treatment and
the enhanced activity observed with c-jun over expression concurs with promoter
mutational studies to suggest that ERa regulation of the ovalbumin promoter is
directed by an AP-l motif identified in the upstream region of the promoter (Gaub et
al., 1990; Tora et al., 1988). Furthermore, these results also indicate that this
regulation involves a cooperative interaction between EROL and AP-l factors. A
significant estradiol- and phorbol ester-dependent induction of ovalbumin activity is
also supported by ERAE3, despite its lack of an intact DNA-binding domain. While
it is evident that the ERAE3 splicing variant is not as active as wt ERa on this
promoter, the capacity for ERAE3 to induce the ovalbumin promoter implies that
direct DNA binding by the receptor is not required for this regulation. Indeed,
induction of this promoter by wt ERa was clearly shown not to require the direct
binding of the full-length receptor to the DNA (Paech, et al., 1997).

In contrast to the situation with vaalb-CAT, the activity of wt ER): on the
remaining three noncanonical response element-containing promoters examined
here was either non-existent or insignificant compared to that of the ERAE3 or
ERAE5 splicing variants. Specifically, both ERAE3 and ERAE5 effectively
enhanced expression of the pColl(-73)Luc reporter, and ERAE5 alone strongly
activated pIGF-Luc and p(AP-l)3TK-CAT. The behavior of these reporters in

response to the various receptor isoforms was not readily predicted from their

119

known biochemical properties and significant differences are evident for each of the
reporters tested.

Similar to ERa-directed regulation of the ovalbumin promoter, activity of
ERa isoforms on the collagenase promoter can be mapped to an AP-l motif.
Induction of pColl(-73)Luc reporter construct, which contains a short region of
collagenase promoter including the AP-l motif, by ERAE3 required both estradiol
and phorbol ester treatment and ERAE3 induction was enhanced with c-jun over
expression. Considering the facts that ERAE3 lacks the DBD, that ERAE3
cooperates with c-jun to maximally activate this promoter, and that c-jun binds the
ERa N-terminus (Webb et al., 1995), these results strongly suggest that ERAE3 is
recruited to the AP-l motif through protein-protein interactions to activate the
collagenase promoter. Whether transcriptional stimulation of reporters containing
. noncanonical hormone response elements (such as vaalb-CAT or pColl(-73)Luc)
involves a direct interaction between ERa or its isoforms and the AP-l components
jun and fos, or an indirect interaction mediated through a bridging factor such as
CBP/p300 or SRC-l needs to be further clarified. Additionally, the identities of the
downstream targets for activation by phorbol ester, while presumed to involve the
PKC pathway, remain to be identified. Potential targets to consider include AP-l,
nuclear receptor co-regulators, or the ERa isoforms themselves.

Activation of the collagenase promoter construct by ERAE5 in HeLa cells

required PMA treatment and was enhanced by c-jun over expression suggesting that

120

ERAE5 cooperates with AP-l factors to regulate this promoter. Induction of the
p(AP-1)3TK-CAT reporter gene by ERAE5 also required phorbol ester treatment. In
contrast, ERAE5 induction of the IGF-l promoter was not enhanced by phorbol ester
treatment. In fact, ERAE5 activity on the IGF-l promoter was slightly attenuated
with PMA treatment. This suggests that activation of AP-l , presumably by
stimulation of PKC and phosphorylation of AP-l factors (e.g., c-jun and c-fos),
plays different roles in the stimulation of these various reporter constructs. These
data also suggest that the presence of a functional AF 2 domain is detrimental for
activation of some (but not all) noncanonical hormone response elements by ERa
since ERAE5, but not wt ERa or ERAE3 stimulated the p(AP-1)3TK-CAT and
pIGF(-1630)Luc reporter plasmids. ERAE5 activity on the pColl(-73)Luc, p(AP-
1)3TK—CAT, and pIGF(-1630)Luc reporters showed no estrogen responsiveness,
consistent with the inability of this variant to bind ligand. In contrast to the reporter
plasmids described above, neither wt EROL, ERAE3, nor ERAE5 stimulated reporter
gene expression from a variety of control promoters representing the HSV thymidine
kinase or c-Met genes, or the early region of Simian Virus 40.

From our studies, we know that wt ERa, ERAE3 and ERAE5 do not
transactivate the same promoters. Compared to the activities observed for ERAE3,
ERAE5 activity is complicated by results that show that efficient induction of the
collagenase and AP-l reporters by this splicing variant requires treatment with

phorbol ester, while stimulation of the IGF -1 promoter does not. The disparities

121

observed for the behavior of these promoters in response to the receptor isoforms
indicate that the context of the AP-l motif (or other noncanonical ERES) can exert a
significant influence on how these promoters respond to ERon and its variants.
Although our cotransfection studies serve to identify promoters activated by
ERa mRNA splicing variants without directly addressing how they do so, published
studies suggest a variety of potential mechanisms that could account for our
observations. According to the first of these models, regulation of some
noncanonical promoters by wt ERa and ERoc splicing variants may involve a direct
protein-protein interaction with other upstream factors such as c-jun and c-fos that in
turn act through their cognate DNA-response elements (summarized in Fig. 27). As
described above, this appears to be the case for ERAE3 regulation of the collagenase
and ovalbumin promoters. Stimulation of transcription by ERAE3 appears to
require dual activation by both E; and phorbol ester. Evidence suggests that
ERAE3-directed transcription of the collagenase and ovalbumin promoters involves
both receptor interaction with AP-l factors and ligand-induced recruitment of
transcriptional co-regulators. The role of ERAE3 and ERAE5 in this activation may
be to recruit coactivator proteins through their intrinsic activation domains. Two
activation functions have been described in wt ERa, AF] is localized to the amino-
terminus and is expressed in both ERAE3 and ERAE5. A second estrogen-inducible
activation function (AF2) resides within the carboxy-terminal ligand-binding

domain and is therefore present in ERAE3, but not in ERAE5. Studies performed

122

using wt ERa demonstrate that both of these activation fimctions correspond to
binding sites for various co-regulatory proteins including CBP/p300 or p160
coactivators such as SRC-l (Endoh et al., 1999; Feng et al,. 1998; Kamei et al.,
1996; Kobayashi et al., 2000; Metivier et al., 2001; Metivier et al., 2000; Tremblay
et al., 1999; Watanabe et al., 2001; Webb et al., 1998). Indeed, we have previously
demonstrated that both ERAE3 and ERAE5, like wt ERa, can interact with the
nuclear receptor coactivator SRC-l. For ERAE3, this interaction involves at least
two SRC-l interaction sites and is dramatically stimulated by ligand. Only the
constitutive amino-terminal site (AF 1) is present in ERAE5 (supporting ligand-
independent interaction with SRC-l ).

Cotransfection and binding studies suggest that ERAE3 and ERAE5 may
cooperate with AP-l factors to activate transcription of noncanonical promoters
containing an AP-l motif. A study showing increased induction of the pColl(-
73)Luc reporter by an ERa-DBD mutant (which resembles ERAE3) when it is
expressed as a VP16 chimera provides evidence that ERAE3 is translocated to an
AP-l motif to stimulate transcription (Webb et al., 1995). Yet there is no direct
evidence that ERAE5 is physically recruited to the promoters that it regulates.
Moreover, the IGF-l promoter does not contain an AP-l motif. Clearly other
mechanisms of activation must be considered. If a receptor isoform is not located to
the site of transcription initiation perhaps it effects transcription by modulating the
activity or expression of other transcription factors (see Fig. 28). Previous reports

have indicated that ERa is capable of inducing the mitogen-activated protein kinase

123

(MAPK) cytoplasmic signalling cascade by a non-genomic mechanism that
stimulates the expression and activation of transcription factors such as c-fos that
ultimately influence cell growth (Migliaccio et al., 1996; van der Burg et al., 1991).
Alternatively, receptor isoforms that are not necessarily chromatin-associated may
still regulate transcription indirectly by binding to a repressor protein that if it were
readily available might otherwise act to silence the induced genes (see Fig. 29).
Results from in vitro binding studies show that wt ERa, ERAE3 and ERAE5 can all
bind the corepressor SMRT. For the promoters we tested, however, titration of
SMRT is unlikely to be the mechanism for transcriptional activation since SMRT
interaction with ERa appears to be constitutive and does not adequately account for
the promoter specificity of the various ER isoforms. Although SMRT does not
appear to function in this role, there may be other unidentified inhibitory factors that
may.

While our transfection studies do not directly address the impact that ERAE3
and ERAE5 have on the transcription of endogenous genes in normal tissues or in
tumors, these results suggest that they nonetheless have the potential to make
distinctive and possibly unique contributions to patterns of gene regulation. As an
extracellular matrix protease, collagenase (along with other metalloproteases) is
likely to play an important role in the tissue remodeling and stromal invasion that
occurs in metastatic disease. The IGF-l gene is similarly expressed in a variety of
tissues, including the liver, bone, uterus and breast where it contributes to the

normal growth and differentiation of these structures (Murphy et al., 1990; Murphy

124

et al., 1988; Pilihowska et al., 1997; Stewart et al., 1996). In the breast, some
uncertainty remains regarding the cell type of origin of IGF-I and with respect to its
precise role in breast tumor growth. It is likely that the mammary epithelium is
exposed to IGF-l produced locally by the breast stroma as well as from systemic
sources such as the liver. In addition, some breast tumors produce IGF-I and other
growth factors of their own (Pilihowska et al., 1997). Expression of IGF-I appears
to be deregulated in some tumors, contributing an autocrine component to the tumor
cell growth. The ability of ERAE5 to stimulate transcription of the IGF-l gene is
therefore of potential clinical significance. Interestingly, we observe that ERAE5,
but not wt ERa promotes expression of a cotransfected IGF-l reporter plasmid,
predicting that IGF-l production by breast tumors that overexpress ERAE5 will be
insensitive to tamoxifen treatment. Figures 27 through 29 summarize three models
depicting what we believe are the most likely mechanisms for transcriptional
regulation by the ERa splicing variants ERAE3 and ERAE5. It should be
emphasized that these mechanisms may coexist in estrogen receptor expressing

tissues, since they are not mutually exclusive.

125

Fig. 27. Transactivation Model 1: Coactivator recruitment. A model depicting
a mechanism of transactivation whereby promoter specificity and kinase dependence
are governed by the selection of coactivators that each ER isoform recruit. J and F
refer to jun and fos, respectively, while TF is meant to refer to an undefined

transcription factor.

126

 

>1-..

gone. in 032220.. 133232.:

23.5.5:
3.0.1.2130... a

3.9.5.2932. 00328"

 

 

31:53.0: 02:on

127

Fig. 28. Transactivation Model 2: Kinase Activation. Model two demonstrates
that promoter specificity and phorbol ester dependence may be governed by the
identity of protein kinases with which each ER isoform can interact. J and F refer to
jun and fos, respectively, while TF is meant to refer to an undefined transcription
factor. PK signifies protein kinases such as JNK, MAPK, c-src, and casein kinase II

(CKII).

128

 

UOIWMIOV 999"” 334i lapOW

Fig. 29. Transactivation Model 3: Corepressor Neutralization. The third model
suggests a mechanism for transcriptional regulation whereby promoter specificity
and kinase dependence are governed by the spectrum of corepressors which

associate with each ER isoform. Abbreviations are as defined in Figs. 27 and 28.

I30

2.002 aw” 00308000.. 2053:5203

 

V. Summary and Conclusions

The ERa gene gives rise to a markedly heterogeneous population of RNA
transcripts that includes exon-skipped, as well as correctly spliced ERa mRN As.
(Castles et al., 1995; Gotteland et al., 1995; Murphy et al., 1997). For this
dissertation project, the biochemical properties of ERa variants lacking a single
internal coding exon were characterized and compared to wt ERa function. These
studies have shown that, while all of the variants can be expressed to give rise to
stable proteins, ERAE2, ERAE4, ERAE6, and ERAE7 represent functionally
impaired receptor isoforms that fail to localize to the nucleus and have no
demonstrable effects (either stimulatory or inhibitory) on gene expression. Whether
these variants fulfill some as of yet undefined role within the cytoplasmic
compartment is currently unknown. It should be stressed, however, that none of
these cytoplasmic ERa variants can bind ligand and therefore cannot be involved in
mediating any of the biological effects of estrogens. The ERAE3 and ERAE5
variants, in contrast, represent nuclear isoforms that retain at least some aspects of
receptor function. For ERAE3, this includes the ability to bind ligand, to form
homologous or mixed dimers with itself or wt ERa, respectively, and to interact
with at least some of the same co-regulatory proteins (specifically, SRC-le and
SMRT) that modulate the transactivating fiinction of nuclear receptor family

members. Although it is unable to bind ligand or dimerize, the ERAE5 variant may

I32

bind weakly to DNA, but more importantly retains the ability to interact with the
nuclear receptor coactivator SRC-le and the nuclear corepressor SMRT. These
properties enable both ERAE3 and ERAE5 to potentially exert complex effects on
gene expression. In stark contrast with wt ERa, ERAE3 and ERAE5 do not
effectively promote gene transcription from an ERE. The observation that ERAE3
and ERAE5 inhibit wt ERor activity on a palindromic ERE provided an early
indication that these variants have the capacity to modulate gene transcription.
More compelling evidence of their transcriptional activity that is seen in transfection
experiments involving reporter gene constructs containing nonconsensus regulatory
elements. Our observations provide the first clear evidence that two ERor splicing
variants, namely ERAE3 and ERAE5, like wt ERa can exert significant positive
effects on gene transcription on their own and raise the intriguing possibility that
each one of these receptor isoforms may target a unique, but overlapping subset of
genes for transcriptional regulation.

A novel pathway for ER function has been described that involves
cooperation of the ER with AP-l (i.e., jun and fos family members) and perhaps
with other transcription factors as well. An important distinction of this nonclassical
pathway for estrogen action is that functional domains within ERa that are crucial
for action through palindromic estrogen response elements (most notably the DBD)
appear to be dispensable for activation of transcription by ERa through AP-l

elements. We have confirmed that two ERa splicing variants (ERAE3 and ERAE5)

133

can stimulate transcription of several promoters that contain AP-l or other
transcription factor binding sites, rather than classical ERa binding motifs. These
include promoters for the human collagenase and IGF-1 genes.

AP-l and its isoforms represent a family of nearly ubiquitous transcription
factors whose activity is crucial in mediating the effects of serum and growth factors
on cellular proliferation (Boyle et al., 1991; Westwick et al., 1994). Some of the
target genes involved in the AP-l-directed pathway for estrogen action include
significant players in the transformation process (e.g., growth factors and matrix
proteases). Only rarely do human breast tumors display a simple pattern of variant
receptor expression or contain more variant than wt ERa mRN A. Even so, it is
conceivable that the modest increases in expression of selected ERa splicing
variants relative to wt ERor that are observed in some tumors and cell lines might
tend to direct ERa transcriptional activity away from ERE-containing genes towards
genes involved in cell grth and transformation (driven by AP-l response
elements).

Several studies using quantitative PCR analysis have concluded that certain
variants (notably ERAE5 and ERAE7) may be elevated relative to wt ERa in breast
tumors generally, or in subsets of tumors with defined metastatic potential or steroid
receptor status (Daffada et al., 1995; Leygue et al., 1996). One such study has
suggested that elevated expression of ERAE5 may be indicative of the ER-lPR+
tumor phenotype and may account for some cases of tamoxifen resistance (Daffada

et al., 1995). It has also been proposed that changes in the expression of ERAE3

I34

relative to wt ERoc may interfere with ERor function and modulate the growth
stimulatory effects of estrogens. A comparison of ERAE3 mRN A expression in
normal and tumorigenic breast specimens suggests that ERAE3 expression is
reduced in breast tumors and studies show that stable over expression of ERAE3 can
slow estrogen-dependent MCF-7 cell grth (Erenburg et al., 1997). Interplay
between ERa splicing variants and ERB (and it’s variants) probably also exists.
Defining ER function is complicated when we consider that ERa, ERB and
each of their variants are all often coexpressed in cells. Estrogen signaling is a
phenomenon that involves the combined contribution of many factors including
coregulatory proteins, enhancer binding sites and all of the various ER isoforms.
ERa, ERB and variants interact with and probably compete for binding with
corepressors and coactivators in vivo, thereby limiting the availability of specific
cofactors in certain cells to the detriment or benefit of gene transcription. Similarly,
ERa, ERB and the ER variants, through protein-protein interactions, may compete
for positioning on enhancer elements. The coexpression of ERB and ERa splicing
variants may also affect in vivo dimer formation. Since ERa heterodimerizes with
both ERAE3 and ERB, it is likely that receptor homology allows for ERB and
ERAE3 dimer formation as well. Such a dimer may direct promoter activities
differently than ERB homodimers or ERa/ERB heterodimers. For example, a
particular dimer may bind a specific promoter for which another dimer has little or

no affinity; and these dimer isoforms may compete to modulate the activity of a

135

single promoter.

Efforts described herein to establish a transcriptional role for ERa mRNA
splicing variants have identified three genes with promoters that are regulated by
one or both of the isoforms, ERAE3 and ERAE5. Expression of a reporter gene
driven by the chicken ovalbumin promoter is induced by ERAE5 in cells treated
with PMA, and by ERAE3 and wt ERa in response to E2 and PMA cotreatment.
Estrogen regulation of the ovalbumin promoter has been mapped to a critical AP-l
enhancer motif (Gaub et al., 1990). Over expression of c-jun enhanced wt EROL,
ERAE3 and ERAE5 activities on the ovalbumin promoter. The human collagenase
promoter also harbors an AP-1 element described to direct E2 regulated gene
expression (Uht et al., 1997). A short region of the collagenase promoter (~73 to
+63 relative to the transcription start site) containing an AP-l motif is activated by
ERAE5 and Ez-liganded ERAE3 in the presence of PMA. In contrast to the report
from Uht and colleagues, in our hands wt ERa demonstrates no activity on this
promoter. Further analysis indicates that ERAE3 activation of the collagenase
promoter requires PMA and E2 cotreatment. These results provide additional
evidence that, like wt ERa, ERAE3 and ERAE5 can activate gene expression.
Interestingly, these results also indicate that although estrogen regulation for both
the collagenase and the ovalbumin promoter has been mapped to an AP-l motif,
these promoters are not activated with the equal effectiveness by the same receptors.

This distinction suggests that flanking sequences surrounding the AP-l motif may

136

confer a difference on the mechanism of regulation, as well as the structural
requirements for ERa-induced activity. The differences seen in regulation of
various AP-l-containing promoters by structurally distinct ERa isoforms is reflected
in the ability of other NRs to transactivate AP- 1 -directed promoters. In
cotransfection studies reported by Barnberger et al. (1996), unliganded PR was also
observed to act through an AP-l motif to stimulate reporter gene expression
(Bamberger et al., 1996). In contrast, GR inhibits basal activity of a reporter gene
construct driven by a region of collagenase promoter containing the AP-l response
element (Webb et al., 1995).

Another promoter included in our studies is the human IGF-l promoter.
Reporter gene expression fi‘om the human IGF -1 promoter is modestly induced by
PMA treatment. This promoter does not respond to wt ERa or ERAE3 and is
constitutively induced by ERAE5. It is unclear where ERAE5 regulation is directed
in the sequence of the IGF-l promoter, since it lacks an ERE or consensus AP-l
element (Kim et al., 1991). However, 5’ deletion analysis suggests that an ERAE5
regulatory element may be localized in the region between nucleotides -532 and -
235. Efforts to transfer ERAE5 responsiveness onto a heterologous promoter were
unsuccessful, showing that this effect requires support from a core promoter
element.

With the characterization of the ERB isoform and the initial discovery of
ERa mRN A splicing variants, the results described above call into question the

appealing, but simplistic notion that all of the biologically important regulatory

137

effects of ERa are mediated by a single 595 amino acid form of this receptor that
was originally described by Chambon and colleagues in 1986 (Green et al., 1986).
Rather, there is growing evidence that ERa splicing variants (most notably ERAE3
and ERAE5) may have a significant impact on the ability of the breast and other
tissues to respond to estrogens and other regulatory signals. This regulation appears
to occur through complex genomic and possibly non-genomic pathways that were
not fully envisioned when Jensen and colleagues first proposed their two-step

mechanism for estrogen action.

I38

VI. Materials and Methods

1. Expression Vectors

Plasmids for ERa mRNA splicing variant cDNAs were generated as
derivatives of pCMV4 (Andersson et al., 1989) and pcDNA3.1 (Invitrogen, San
Diego, CA), which support high levels of receptor expression in HeLa and C087 cell
lines (Miksicek et al., 1995). Plasmids expressing ERAE4, ERAE5, and ERAE6
were generated using synthetic oligonucleotides to construct the variant splice
junctions within an otherwise wt ERa cDNA expression plasmid. The remaining
plasmids were constructed with the use of flanking restriction sites to shuttle cloned
cDNAs (Wang and Miksicek, 1991) into the appropriate expression vectors. Mouse
c-jun and c-fos cDNA, cloned into the pCMV2 expression vector, were provided by

L. McCabe (Michigan State University, East Lansing, Mich.)

2. Cell Culture, Transfection, and CAT Assays

C037 and HeLa cells were grown in phenol red-free DMEM supplemented
with 10% calf-serum, SmM HEPES (pH 7.4), 2 mM glutamine, penicillin (50 U/ml),
and strptomycin (50 ug/ml). Cells were transfected by the CaPO4 method, as

previously described (Jordan et al., 1996). HeLa cells (1 x 106 cells per 60-mm

dish) were transfected with 0.5 ug of the indicated ERa expression plasmid, 2 pg of

139

the c-jun expression plasmid and 10 pg of reporter plasmid. The reporter plasmids
used for these experiments included pERE-TK-CAT (Klock et al., 1987), vaalb-
CAT (Schweers et al., 1990), p(AP-1)3TK-CAT kindly provided by L. McCabe
(Dept. of Physiology, MSU), pColl(-73)Luc received from P.J. Kushner (Webb et
al., 1995), and pIGF-Luc constructs (-926, -592, -235) obtained from P. Rotwein
(Kim et al., 1991). Calf thymus DNA (10 pg) was added as carrier. After overnight
incubation with DNA, culture medium was replaced with 5% charcoal-treated
serum-supplemented DMEM. Cell cultures were treated with the indicated
hormones (20 nM PMA, 10 nM 17B-estradiol, 100 nM tamoxifen) or vehicle
(ethanol + DMSO, 1% each). After 24-hour incubation, cells were harvested and
CAT assays were performed as previously described (Gorman et al., 1982) using
100 pg protein. Quantitation of CAT activities was performed by phosphorimage
analysis of thin layer chromatographs (ImageQuaNT, Molecular Dynamics, Inc.,
Sunnyvale, CA). For experiments involving biochemical or cytochemical analysis
of ERa variants, Cos7 cells were similarly transfected with 10 pg of the indicated
expression plasmid and 10 pg of calf thymus carrier DNA. After overnight
exposure to DNA, cells were cultured for 48 hours in 10% calf serum-supplemented
DMEM. All experiments involving extracts from transfected cells were normalized
with respect to protein, as measured using the method of Lowry et al. (Lowry et al.,
1951). Two-way AN OVA and comparison with Student’s t test were used to assess
statistical differences between groups. The level of statistical significance is

indicated in each figure.

140

3. E2 Binding Analysis

Ligand-binding assays were performed as previously described (Neff et al.,
1994). Whole-cell extracts were prepared from transfected Cos7 cells that were
resuspended and sonicated in extraction buffer (20mm HEPES, pH 7.4, 20%
glycerol, 0.4 M KCL, ImM MgClz) supplemented immediately before use with
protease inhibitors (0.05 mg/ml each of chymostatin, trypsin inhibitor, antipain,
leupeptin, aprotinin, and pepstatin). Aliquots containing 200 pg of protein were
incubated overnight at 4° C with various concentrations (0.1 nM - 10 nM) of 3H-
labeled E; (N EN Life Science Products, Boston, Mass.) in the presence or absence
of a ZOO-fold molar excess of unlabeled E2. Free ligand was separated fiom bound
ligand by treatment with dextran-coated charcoal. For determination of equilibrium
binding constants, these data were plotted according to the method of Scatchard

(Scatchard, 1949).

4. DNA Binding Assays

DNA binding assays were performed as previously described (N eff et al.,
1994). Aliquots containing 30 pg of protein from extracts prepared as above from
transfected Cos7 cells were preincubated for 15 minutes at room temperature in 10

pl binding buffer [10 mM HEPES (pH 7.4), 1 mM dithiothreitol, and 20% glycerol]

141

containing 1 pg poly (dI-dC), with or without 1 pl of added human ER—specific
monoclonal antibody (Mab-17), generated as described (N eff et al., 1994).
Approximately 6 frnol (40,000 cpm) ofa 32P-labeled double-stranded ERE
oligonucleotide (Wang and Miksicek, 1991) were added to the samples and
incubated for 30 minutes at room temperature, followed by an additional S-minute
incubation at 4° C. Samples were then loaded on a preelecrophoresed nondenaturing
5% polyacrylarnide gel that was run in 0.5 x Tris-Borate-EDTA at 275 V for 2

hours. The gel was dried and exposed for autoradiography.

5. Immunoblot Analysis

Discontinuous 12% SDS-PAGE was carried out as previously described
(Laemmli, 1970). After electrophoresis of 30 pg of whole-cell protein from extracts
of transfected Cos7 cells, proteins were elecrophoretically transferred to
nitrocellulose filters with 3 Trans Blot apparatus (Bio-Rad Labotatories, Inc.
Richmond, CA) using the procedure previously detailed (Erickson et al., 1982).
Immunblots were probed with the ER-specific monoclonal antibody, Mab—17,
obtained from a hybridoma culture supemate that was diluted with an equal volume
of PBS (Neff et al., 1994). Irnmunoreactive protein was visualized by enhanced
chemiluminescence using a horseradish peroxidase-conjugated goat antimouse IgG,
following manufacturer’s instructions (Amersham Pharmacia Biotech, Arlington

Heights, IL)

142

6. In Vitro Protein-Protein Interaction Assays

Variant and wt ERa receptor protein was translated in the presence of 358-
methionine using TNT Coupled Reticulocyte System (Promega Corp., Madison,
WI). GST-fusion proteins were expressed in the pGEX system (Pharmacia Biotech,
Uppsala, Sweden) (Cavailles et al., 1994; Kalkhoven et al., 1998). The GST-SMRT
fusion vector was provided by R. M. Evans . The GST-AF 2 and GST-SRC-le
fusion vectors were provided by and M. G. Parker. Overnight cultures of
transformed bacteria were diluted 1:20 and cultured for 2 hours before protein
expression was induced with the addition of isopropyl B-D-thiogalactoside (IPTG,
0.2 mM final concentration). Bacteria were collected by centrifugation 2 hours
following IPTG induction, and pellets were resuspended in 400 pl of extraction
buffer supplemented with protease inhibitors. Cells were sonicated briefly, and the
resulting lysates were centrifuged for 20 minutes at 20,000 rpm, 4° C. Protein
concentrations were determined (Lowry et al., 1951) and extracts were diluted to 2
pg/ pl in extraction buffer and stored at -70° C until binding assays were performed.

Before use in protein interaction assays, 25 pl of glutathione-Sepharose 4B
beads (Pharmacia Biotech) were washed three times in 100 pl NETN [0.5% Nonidet
P-40, 1 mM EDTA, 20 mM Tris (pH 8.0), 100 mM NaCl] and suspended in 100 pl
NETN, 0.5% powdered milk. Washed beads were incubated with 40 pg of GST-
fusion protein for 2 hours, rotating at room temperature. Beads complexed with

GST-fusion proteins were washed three times with 100 pl NETN, 4° C. For protein

143

binding assays, 5 pl of in vitro translated protein was added to washed complexed
beads resuspended in 100 pl NETN supplemented with protease inhibitors (as
above) with and without 2.5 pM E2. After a 2-hour incubation during which the
samples were rotated at room temperature, the beads were pelleted and washed four
times with 100 pl NETN, 4° C. Bound proteins were separated on a discontinuous
10% polyacrylamide SDS-PAGE gel (Laemmli, 1970). The gels were dried and

exposed for autoradiography.

7. Immunohistochemical and Cytochemical Analysis

Indirect immunofluorescence analysis was performed as previously
described (Neff et al., 1994) using Cos7 cells that were plated and transfected on
glass cover slips. On the second day after transfection, cells were washed three
times wih Tris-buffered saline (TBS), fixed for 3 minutes in cold 95% methanol,
rehydrated by three washes with TBS, and incubated 30 minutes at 37° C with
primary antibody (Mab-l 7 hybridoma supemate used at a l :10 dilution in TBS).
Bound antibody was detected by staining with a rhodamine-conj ugated affinity-
purified goat anti-mouse IgG (Roche Molecular Biochemicals, Indianapolis, IN)
diluted 1:2000 in TBS, and incubating for 30 minutes at 37° C in the presence of
0.02 pg/ml of 4’, 6-diamidine-2-phenylindole dihydrochloride. Confocal images

were recorded using the Odyssey system (Noran Instruments, Middleton, WI) on an

144

Optiphot 2 Nikon (Melville, NY) microscope. Fluorescent ligand staining of
transfected Cos7 cells was performed as described (Miksicek et al., 1995) on live,
whole-cell mounts treated in DMEM with 10—7 M nitrile THC. For these studies,
cells were visualized using a Nikon UFX microscope equipped with a 100 watt
mercury lamp for fluorescence excitation, and a 40 x 0.7 numerical aperture Plan

objective.

145

VII. List of References

Andersson S, Davis DN, Dahlback H, Jomvall H, Russell DW (1989) Cloning,
structure, and expression of the mitochondrial cytochrome P450 sterol 26-
hydroxylase, a bile acid biosynthetic enzyme. J Biol Chem, 264, 8222-8229.

Angel P, Poting A, Mallick U, Rahmsdorf HJ Schorpp M, Herrlich P (1986)
Induction of metallothionein and other mRNA species by carcinogens and

tumor promoters in primary human skin fibroblasts. Mol Cell Biol, 6, 1760-
1766

Angel P, Rahmsdorf HJ, Pfiting A, Lficke-Huhle C, Herrlich P (1985) 12-0-
tetradecanoylphorbol-l 3-acetate (TPA)-induced gene sequences in human
primary diploid fibroblasts and their expression in SV40-transformed
fibroblasts. J Cell Biochem, 29, 351-360

Arnold SF, Melamed M, Vororjeikina DP, Notides AC, Sasson S (1997) Estradiol-
binding mechanism and binding capacity of the human estrogen receptor is
regulated by tyrosine phosphorylation. Mol Endocrinol, 11, 48-53.

Arnold SF, Vorojeikina DP, Notides AC (1995) Phosphorylation of tyrosine 537 on
the human estrogen receptor is required for binding to an estrogen response
element. J Biol Chem, 270, 30205-30212.

Arnold SF, Oboum JD, Jaffe HJ, Notides AC (1995) Phosphorylation of the human
estrogen receptor by mitogen-activated protein kinase and casein kinase H:
consequence on DNA binding. J Steroid Biochem Mol Biol, 55, 163-172.

Aronica SM, Kraus WL, Katzenellenbogen BS (1994) Estrogen action via the
cAMP signaling pathway: stimulation of adenylate cyclase and CAMP-
regulated gene transcription. Proc Natl Acad Sci USA, 91, 8517-8521.

Bamberger AM, Bamberger CM, Gellersen B, Schulte HM (1996) Modulation of
AP-l activity by the human progesterone receptor in endometrial
adenocarcinoma cells. Proc Natl Acad Sci USA, 93, 6169-6174.

Baniahmad A, Kohne AC, Renkawitz R (1992) A transferable silencing domain is

present in the thyroid hormone receptor, in the v-erbA oncogene product and
in the retinoic acid receptor. Cell, 11, 1015-1023.

146

Barettino D, Vivanco Ruiz MM, Stunnenberg HG 1994 Characterization of the

ligand- dependent transactivation domain of thyroid hormone receptor.
EMBO J, 13, 3039-3049.

Bartsch J, Truss M, Bode J, Beato M (1996) Moderate increase in histone
acetylation activates the mouse mammary tumor virus promoter and
remodels its nucleosome structure. Proc Natl Acad Sci USA, 93, 10741-
10746.

Beato M, Sanchez-Pacheco A (1996) Interaction of steroid hormone receptors with
the transcription initiation complex. Endocr Rev, 17, 587-609.

Benecke A, Chambon P, Gronemeyer H (2000) Synergy between estrogen receptor
alpha activation functions AF 1 and AF 2 mediated by transcription
intermediary factor TIF2. EMBO Rep, 1, 151-157.

Berry M, Metzger D, Chambon P (1990) Role of the two activating domains of the
oestrogen receptor in the cell-type and promoter-context dependent agonistic
activity of the antioestrogen 4-hydroxytamoxifen. EMBO J, 9, 2811-2818.

Blenis J (1993) Signal transduction via the MAP kinases: proceed at your own risk.
Proc Natl Acad Sci USA, 90, 5889-5892.

Bokemeyer D, Sorokin A, Dunn MJ (1996) Multiple intracellular MAP kinse
signaling cascades. Kidney Int, 49, 1187-1198.

Bowler J, Lilley TJ, Pittam JD, Wakeling AE (1989) Novel steroidal pure
antiestrogens. Steroids, 54, 71-99.

Boyle WJ, Smeal T, Defize LHK, Angel P, Woodgett JR, Karin M, Hunter T (1991)
Activation of protein kinase C decreases phosphorylation of cJun at sites that
negatively regulate its DNA-binding activity. Cell, 64, 573-584.

Bresnick EH, Rories C, Hager GL (1992) Evidence that nucleosomes on the mouse
mammary tumor virus promoter adopt specific translational positions.
Nucleic Acids Res, 20, 865-870.

Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, Ohman L,

Greene GL, Gustafsson J-A, Carlquist M (1997) Molecular basis of agonism
and antagonism in the oestrogen receptor. Nature, 389, 753-758.

147

Bunone G, Briand PA, Miksicek RJ, Picard D (1996) Activation of the unliganded
estrogen receptor by EGF involves the MAP kinase pathway and direct
phosphorylation. EMBO J, 15, 2174-2183.

Burke LJ, Baniahmad A (2000) Co-repressors 2000. FASEB J, 14, 1876-1888.

Castles CG, F uqua SAW, Klotz DM, Hill SM (1993) Expression of a constitutively
active extrogen receptor variant in the estrogen receptor-negative BT-20
human breast cancer cell line. Cancer Res, 53, 5934-5939.

Castles CG, Klotz DM, Fuqua SAW, Hill SM (1995) Coexpression of wild-type and
variant oestrogen receptor mRNAs in a panel of human breast cancer cell
lines. Br J Cancer, 71, 974-980.

Cavailles V, Dauvois S, Danielian PS, Parker MG (1994) Interaction of proteins

with transcriptionally active estrogen receptors. Proc Natl Acad Sci USA,
91, 10009-10013.

Cavailles V, Dauvois S, L’Horset J, Lopez G, Hoare S, Kushner PJ, Parker MG
(1995) Nuclear factor RIPl40 modulates transcriptional activation by the
estrogen receptor. EMBO J, 14, 3741-3751.

Chaidarun SS, Alexander J M (1998) A tumor-specific truncated estrogen receptor
splice variant enhances estrogen-stimulated gene expression. Mol
Endocrinol, 12, 1355-1366.

Chen D, Ma H, Hong H, Koh SS, Huang S-M, Schurter BT, Aswad DW, Stallcup
MR (1999) Regulation of transcription by a protein methyltransferase.
Science, 284, 2174-2177.

Chen H, Lin RJ, Schlitz L, Chakravarti D, Nash A, Nagy L, Privalsky ML, Nakatani
Y, Evans RM (1997) Nuclear receptor coactivator ACTR is a novel histone

acetyltransferase and forms a multirneric activation complex with P/CAF a
CBP/p300. Cell, 90, 569-580.

Chen JD, Evans RM (1995) A transcriptional co-repressor that interacts with nuclear
hormone receptors. Nature, 377, 454-457.

Chen JD, Umesono K, Evans RM (1996) SMRT isoforms mediate repression and

anti-repression of nuclear receptor heterodimers. Proc Natl Acad Sci USA,
93, 7567-7571.

148

Chen WY, Townes TM (2000) Molecular mechanism for silencing virally
transduced genes involves histone deacetylation and chromatin
condensation. Proc Natl Acad Sci USA, 97, 377-382.

Cheskis BJ, Karathanasis S, Lyttle CR (1997) Estrogen receptor ligands modulate its
interaction with DNA. J Biol Chem, 272, 11384-11391.

Chien CH, Wang FF, Hamilton TC (1994) Transcriptional activation of c-myc
proto-oncogene by estrogen in human ovarian cancer cells. Mol cell
Endocrinol, 99,1 1-19.

Cho H, Katzenellenbogen BS (1993) Synergistic activation of estrogen receptor-
mediated transcription by estradiol and protein kinase activators.
Mol Endocrinol, 7, 441-452.

Chu S and Fuller PJ (1997) Identification of a splice variant of the rat estrogen
receptor 0 gene. Mol Cell Endocrinol, 132, 195-199.

Coleman KM, Smith CL (2001) Intracellular signaling pathways: nongenomic
actions of estrogens and ligand-independent activation of estrogen receptors.
Front Bioscience, 6, 1379-1391.

Collingwood TN, Butler A, Tone Y, Clifton-Bligh RJ Parker MG, Chatterjee VK
(1997) Thyroid horrnone-mediated enhancement of heterodirner formation
between thyroid hormone receptor beta and retinoid X receptor. J Biol
Chem, 272, 13060-13065.

Compston JE (2001) Sex steroids and bone. Physiol Rev, 81, 419-447.

Couse JF, Curtis SW, Washburn TF, Lindzey J, Golding TS, Lubahn DB, Smithies
O, Korach KS (1995) Analysis of transcription and estrogen insensitivity in

the female mouse afier targeted disruption of the estrogen receptor gene. Mol
Endocrinol, 9, 1441-1454.

Couse JF, Hewitt SC, Korach KS (2000) Receptor null mice reveal contrasting roles

for estrogen receptor a and b in reproductive tissues. J Steroid Biochem Mol
Biol, 74, 287-296.

Curtis Hewitt S, Couse J F, Korach KS (2000) Estrogen receptor transcription and

transactivation: estrogen receptor knockout mice: what their phenotypes
reveal about mechanisms of estrogen action. Breast Cancer Res, 2, 345-352.

149

Daffada AAI, Johnston SRD, Smith IE, Detre S, King N, Dowsett M (1995) Exon 5
deletion variant estrogen receptor messenger RNA expression in relation to
tamoxifen resistance and progesterone receptor/p32 status in human breast
cancer. Caner Res, 55, 288-293.

Danielian PS, White U, Lees JA, Parker MH (1992) Identification of a conserved
region required for hormone dependent transcriptional acivation by steroid
hormone receptors. EMBO J, 11, 1025-1030.

Darimont BD, Wagner RL, Apriletti J W, Stallcup MR, Kushner PJ, Baxter JD,
Fletterick RJ, Yamarnoto KR (1998) Structure and specificity of nuclear
receptor-coactivator interactions. Gene Dev, 12, 3343-3356.

Dechering K, Boersma C, Mosselman S (2000) Estrogen receptors alpha and beta:
two receptors of a kind? Current Med. Chem, 7, 561-576.

Denner LA, Schrader WT, O’Malley BW, Weigel NL (1990) Hormonal regulation
and identification of chicken progesterone receptor phosphorylation sites. J
Biol Chem, 265, 16548-16555.

Denton RR, Koszewski NJ, Notides AC (1992) Estrogen receptor phosphorylation.

Hormonal dependence and consequence on DNA binding. J Biol Chem,
267, 7263-7268.

Decensi A, Fontana V, Bruno S, Gustavino C, Gatteschi B, Costa A (1996) Effect of
tamoxifen on endometrial proliferation. J Clin Oncol, 14, 434-40

Dhandaparri KM, Brann DW (2002) Protective effects of estrogen and selective
estrogen receptor modulators in the brain. Biol Repro, 67, 1379-1385.

Ding XF, Anderson CM, Ma H, Hong H, Uht RM, Kushner PJ, Stallcup MR (1998)
Nuclear receptor-binding sites of coactivators glucocorticoid receptor
interacting protein 1 (SRC-l ): multiple motifs with different binding
specificities. Mol Endocrinol, 12, 302-313.

DiRenzo J, Shang Y, Phelan M, Sif S, Myers M, Kingston R, Brown M (2000)
BRG-l is recruited to estrogen-responsive promoters and cooperates with

factors involved in histone acetylation. Mol Cell Biol, 20, 7541-7549.

Dowsett M, Daffada A, Chan CM, Johnston SR (1997) Oestrogen receptor mutants
and variants in breast cancer. Eur J Cancer, 33, 1177-1183.

150

Draper MW, Flowers DE, Huster WJ, Neild JA, Harper KD, Arnaud C (1996) A
controlled trial of raloxifene (LY139481) HCL: impact on bone turnover and
lipid profile in healthy postmenopausal women. J Bone Miner Res, 11, 835-
842.

Duan R, Porter W, Safe S (1998) Estrogen-induced c-fos protooncogene expression
in MCF-7 human breast cancer cells: role of estrogen receptor Spl complex
formation. Endocrinol, 139, 1981-1990.

Durand BM, Saunders M, Gaudon C, Roy B, Losson R, Chambon P (1994)
Activation function 2 (AF-2) of retinoic acid receptor and 9-cis retinoic acid
receptor: presence of a conserved autonomous constitutive activating domain
and influence of the nature of the response element on AF -2 activity.

EMBO J, 13, 5370-5382.

Ekena K, Weis KE, Katzenellenbogen JA, Katzenellenbogen BS (1997) Different
residues of the human estrogen receptor are involved in te recognition of
structurally diverse estrogens and antiestrogens. J Biol Chem, 272,
5069-5075.

Endoh H, Maruyama K, Masuhiro Y, Kobayashi Y, Goto M, Tai H, Yanagisawa J,
Metzger K, Hashimoto S, Kato S (1999) Purification and identification of
p68 RNA helicase acting as a transcriptional coactivator specific for the
activation function 1 of human estrogen receptor alpha. Mol Cell Biol, 19,
5363-5372.

Enmark E, Pelto-Huikko M, Grandien K, Lagercrantz S, Lagercrantz J, Fried G,
Nordenskjéld M, Gustafsson J -A (1997) Human estrogen receptor B-gene
structure, chromosomal localization, and expression pattern. J Clinical
Endocrinol, 82, 4258-4265.

Ernst M, Rodan GA (1991) Estradiol regulation of insulin-like growth factor-1 gene
expression in osteoblastic cells: evidence for transcriptional control. Mol
Endocrinol, 5, 1081-1089.

Ernst M, Heath JK, Rodan GA (1989) Estradiol effects on proliferation, messenger
ribonucleic acid for collagen and insulin-like growth factor-1, and
parathyroid hormone-stimulated adenylate cyclase activity in osteoblastic
cells from calvariae and long bones. Endocrinology, 125, 825-833

151

Erenburg I, Schachter B, Mira y Lopez R, Ossowski I (1997) Loss of an estrogen
receptor isoform (ERaAE3) in breast cancer and the consequences of its

reexpression: interference with estrogen-stimulated properties of malignant
transformation. Mol Endocrinol, 11, 2004-2015.

Erickson PF, Minier LN, Lasher RS (1982) Quantitative electorphoretic transfer of
polypeptides from SDS polyacrylamide gels to nitrocellulose sheets: a
method for their re-use in irnmunoautoradiographic detection of antigens. J
Immunol Methods, 51, 241-249.

Ernst M, Rodan GA (1991) Estradiol regulatioin of insulin-like growth factor-1 gene

expression in osteoblastic cells: evidence for transcriptional control. Mol
Endocrinol, 5, 1081-1089.

Feng W, Ribiero RCJ, Wagner RL, Nguyen H, Apriletti J W, Fletterick RJ, Baxter
JD, Kushner PJ, West BL (1998) Hormone-dependent coactivator binding to
a hydrophobic clefi on nuclear receptors. Science, 280, 1747-1749.

Fiorelli G, Sori F, Frediani U, Franceschelli F, Tanini A, Tosti-Guerra, Benvenuti S,
Gennari L, Becherini L, Brandi ML (1996) Membrane binding sites and non-
genomic effects of estrogen in cultured human pre-osteoclastic cells. J
Steroid Biochem Mol Biol, 59, 233-240.

Fletcher TM, Xiao N, Mautino G, Baumann CT, Wolford R, Warren BS, Hager GL
(2002) ATP-dependent mobilization of the glucocorticoid receptor durring
chromatin remodeling. Mol Cell Biol, 22, 3255-3263.

F ondell JD, Guerrnan M, Malik S, Roeder RG (1999) Thyroid hormone receptor-
associated proteins and general positive cofactors mediate thyroid hormone
receptor function in the absence of the TATA box-binding protein-associated
factors of TFHD. Proc Natl Acad Sci, 96, 1959-1964.

Fragoso G, John S, Toberts MS, Hager GL (1995) Nucleosome positioning on the
MMTV LTR results fiom the frequency-biased occupancy of multiple
frames. Genes Dev, 9, 1933-1947.

Freedman LP (1999) Increasing the complexity of coactivation in nuclear receptor
signaling. Cell, 97, 5-8.

Friend KE, Resnick EM, Ang LW, Shupnik MA (1997) Specific modulation of
estrogen receptor mRN A isoforms in rat pituitary throughout the estrous

cycle and in reponse to steroid hormones. Mol Cell Endocrinol, 13,
147-155.

152

Fr6din M, Gammeltofi S (1999) Role and regulation of 90kDa ribosomal S6 kinase
(RSK) in signal transduction. Mol Cell Endocrinol, 151, 65-77.

Fuqua SAW, Charnness GC, McGuire WL (1993) Estrogen receptor mutations in
breast cancer. J Cell Biochem, 51, 135-139.

Fuqua SAW, Fitzgerald SD, Allred DC, Elledge RM, Nawaz Z, McDonnel DP,
O’Malley BW, Green GL, McGuire WL (1992) Inhibition of estrogen
receptor action by a naturally occurring variant in human breast tumors.
Cancer Res, 52, 483-486.

Fuqua SAW, Fitzgerald SD, Charnness GC, Tandon AK, McDonnell DP, Nawaz Z,
O’Malley BW, McGuire WL (1991) Variant human breast tumor estrogen
receptor with constititive transcriptional activity. Cancer Res, 51, 105-109.

F uqua SAW, Wolf DM (1995) Molecular aspects of estrogen receptor variants in
breast cancer. Breast Cancer Res Treat, 35, 233-241.

Furrnan E, Rushkin E, Margalit R, Bendel P, Degani H (1992) Tamoxifen induced
changes in MCF7 human breast cancer: in vitro and in vivo studies using

nuclear magnetic resonance spectroscopy and imaging. J Steroid Biochem
Mol Biol, 43, 189-195.

Gaub MP, Bellard M, Scheuer I, Chambon P, Sassone-Corsi P (1990) Activation of
the ovalbumin gene by the estrogen receptor involves the Fos-Jun complex.
Cell, 63, 1267-1276.

Glass CK, Rosenfeld MG (2000) The coregulator exchange in transcriptional
functions of nuclear receptors. Genes Dev, 14, 121-141.

Gluzrnan Y (1981) SV40-transformed simian cells support the replication of early
SV40 mutants. Cell, 23, 175-182.

Gotteland M, Desauty G, Delarue J C, Liu L, May E (1995) Human estrogen receptor
messenger RNA in both normal and tumor breast tissues. Mol Cell
Endocrinol, 112, 1-13.

Gorman CM, Moffat LF, Howard BH (1982) Recombinant genomes which express

chloramphenicol acetyltransferase in mammalian cells. Mol Cell Biol, 2,
1044-105 1 .

153

Gosden JR, Middleton PG, Rout D (1986) Localisation of the human oestrogen
receptor gene to chromosome 6q24-q27 by insitu hybridisation. Cytogenet
Cell Genet, 43, 218-220.

Graham JD, Bain DL, Richer JK, Jackson TA, Tung L, Horowitz KB (2000)
Nuclear receptor conformation, coregulators, and tamoxifen-resistant breast
cancer. Steroids, 65, 579-584.

Green S, Walter P, Kumar V, Krust A, Bomet JM, Argos P, Chambon P (1986)
Human oestrogen receptor cDNA: sequence, expression and homology to v-
erbA. Nature, 320,134-139.

Grumbach MM (2000) Estrogen, bone, grth and sex: a sea change in
conventional wisdom. J Pediatr Endocrinol Metab, 13, Suppl 6, 1439-1455.

Guo Z, Krucken J, Benten WP, Wunderlich F (2002) Estradiol-induced nongenomic
calcium signaling regulates genotropic signaling in macrophages. J Biol
Chem, 277, 7044-7050.

Gustafsson J -A, Warner M (2000) Estrogen receptor [3 in the breast: role in estrogen
responsiveness and development of breast cancer. J Steroid Biochem Mol
Biol, 74, 245-248.

Hall LJ, Kajimoto Y, Bichell D, Kim S-W, James PL, Counts D, Nixon LJ, Tobin
G, Rotwein P (1992) Functional analysis of the rat insulin-like growth factor

I gene and identification of an IGF—I gene promoter. DNA Cell Biol, 11,
301 -3 13.

Hall RF, Lee CSL, Alexander IE, Shine J, ClarkCL, Sutherland R (Steroid hormone
receptor gene expression in human breast cancer cells: inverse relationship

between oestrogen and glucocoticoid receptor messenger RNA levels. Int J
Cancer, 46, 1081-1087.

Haslam SZ (1987) Role of sex steroid hormones in normal mammary gland
function. Neville MC, Daniel CW (eds.) The Mammary Gland. Plenum,
New York, 499-533.

Hayashi Y, Ohmori S, Ito T, Seo H (1997) A splicing variant of steroid receptor
coactivator-1 (SRC-le): the major isoform of SRC-l to mediate thyroid
hormone action. Biochem Biophys Res Corn, 236, 83-87.

Heery Dm, Kalkhoven E, Hoare S, Parder MG (1997) A signature motif in
coactivators mediates binding to nuclear receptors. Nature, 387, 733-736.

154

Heinzel T, Lavinsky RM, Mullen T-M, Stiderstrtim M, Laherty CD, Torchia J, Yang
W-M, Brard G, Ngo SD, Davie JR, Seto E, Eisenman RN, Rose DW, Glass
CK, Rosenfeld MG (1997) A complex containing N-CoR, mSin3 and
histone deacetylase mediates transcriptional repression. Nature, 387, 43-48.

Herrlich P, Angel P, Rahmsdorf HJ, Mallick U, Pfiting A, Hieber L, Lficke-Huhle
C, Schorpp M (1986) The mammalian genetic stress response. Adv Enzyme
Reg, 25, 485-504.

Hess RA, Bunick D, Lee KH, Bahr J, Taylor JA, Korach KS, Lubahn DB (1997) A
role for oestrogens in the male reproductive system. Nature, 390, 447-448.

Hisamoto K, Ohmichi M, Kurachi H, Hayakawa J, Kanda Y, Nishio Y, Adachi K,
Tasaka K, Miyoshi E, Fujiwara N, Taniguchi N, Murata Y (2001) Estrogen
induces the Akt-dependent activation of endothelial nitric-oxide synthase in
vascular endothelial cells. J Biol Chem, 276, 3459-3467.

Horlein AJ, Naar AM, Heinzel , Torchia J, Gloss B, Kurokawa R, Ryan A, Kamel
Y, Soderstrém M, Glass CK, Rosenfeld MG (1995) Ligand-independent
repression by the thyroid hormone receptor mediated by a nuclear receptor
co-repressor. Nature, 377, 397-404.

Horowitz KB, Jackson TA, Bain DL, Richer JK, Takimoto GS, Tung L (1996)
Nuclear receptor coactivators and corepressors. Mol Endocrinol, 10,
1167-1177.

Howard KJ, Holley SJ, Yamarnoto KR, Distelhorst CW (1990) Mapping the hsp90
binding region of the glucocorticoid receptor. J Biol Chem, 265,
11928-11935.

Hu C, Hyder SM, Needleman DS, Baker VV (1996) Expression of estrogen receptor
variants in normal and neoplastic human uterus. Mol Cell Endocrinol, 118,
173-179.

Huston L (1999) Raloxifene reduces breast cancer risk. Lancet, 353, 43.
Ignar—Trowbridge DM, Pimentel M, Parker MG, McLachlan JA, Korach KS (1996)
Peptide growth factor cross-talk with the extrogen receptor requires the NB

domain and occurs independently of protein kinase C or estradiol.
Endocrinol, 137, 1735-1744.

155

Improta-Brears T, Whorton AR, Codazzi F, York JD, Meyer T, McDonnell DP
(1999) Estrogen-induced activation of mitogen-activated protein kinase
requires mobilization of intracellular calcium. Proc Natl Acad Sci USA, 96,
4686-4691.

Ing NH, Beekman HM, Tsai SY, Tsai MJ, O’Malley BW (1992) Members of the
steroid homone recepttor superfamily interact with TFHB (S300-II). J Biol
Chem, 267, 17617-17623.

Iwao K, Miyoshi Y, Egawa C, Ikeda N, Noguchi S (2000) Quantitative analysis of
estrogen receptor-B mRNA and its variants in human breast cancers. Int J
Cancer, 88, 733-736.

Jackson TA, Richer JK, Bain DL, Takimoto GS, Tung L, Horowitz KB (1997) The
partial agonist activity of antagonist-occupied steroid receptors is controlled
by a novel hinge domain-binding coactivator L7/SPA and the corepressors
N-CoR or SMRT. Mol Endocrinol, 11, 693-705.

Jacq X, Brou C, Lutz Y, Davidson 1, Chambon P, Tom L (1994) Human Tafn30 is
present in a distinct TFHD complex and is required for transcriptional
activation by the estrogen receptor. Cell, 79, 107-117.

J anknecht R, Hunter T (1996) A growing coactivator network. Nature, 383, 22-23.

Jensen EV, Suzuki T, Kwashima T, Stumpf WE, Jungblut PW, DeSombre ER
(1968) A 2 step mechanism for interaction of estradiol with rat uterus. Proc
Natl Acad Sci USA, 59, 632-638.

J uneja M, Jose R, Kekre AN, Viswanathan F, Seshadri (2002) Tarnoxifen-induced
endometrial changes in postmenopausal women with breast carcinoma. Int J
Gynaecol Obstet, 76, 279-84.

Joel PB, Smith J, Sturgill TW, Fisher TL, Blennis J, Lanningan DA (1998) p90rsk1
regulates estrogen receptor-mediated transcription through phosphorylation
of Set-167. Mol Cell Biol, 18, 1978-1984.

Jordan M, Schallhom A, Wurm FM (1996) Transfecting mammalian cells:
optimization of critical parameters affecting calcium-phosphate precipitate
formation. Nucleic Acids Res, 24, 596-601.

Kalkhoven E, Valentine JE, Heery DM, Parker MG (1998) Isoforms of steroid

receptor co-activator 1 differ in their ability to potentiate transcription by the
oestrogen receptor. EMBO J, 17, 232-243.

156

 

Kamei Y, Xu L, Heinzel T, Torchia J, Kurokawa R, Gloss B, Lin S-C, Heyrnan RA,
Rose DW, Glass CK, Rosenfeld MG (1996) A CBP integrator complex
mediatees transcriptional activation and AP-l inhibition by nuclear
receptors. Cell, 85, 403-414.

Kato S, Endoh H, Masuhiro Y, Kitarnoto T, Uchiyama S, Sasaki H, Masushige S,
Gotoh Y, Nishida E, Kawashima H, Metzger D, Chambon P (1995)
Activation of the estrogen receptor through phosphorylation by mitogen-
activated protein kinase. Science, 270, 1491-1494.

Katzenellenbogen J A, O’Malley BW, Katzenellenbogen BS (1996) Tripartite steroid
hormone receptor pharmacology: interaction with multiple effector sites as a
basis for the cell- and promoter-specific action of these hormones. Mol
Endocrinol, l0, 1 19-13 1 .

Katzenellenbogen BS, Choi I, Delage-Mourroux R, Ediger TR, Martini PGV,
Montano M, Sun J, Weis K, Katzenellenbogen JA (2000) Molecular
mechanisms of estrogen action: selective ligands and receptor pharmacology.
J Steroid Biochem Mol Biol, 74, 279-285.

Kaufrnann M (1997) A review of endocrine Options for the treatment of advanced
breast cancer. Oncology, 54 suppl, 2-5.

Khavari PA, Peterson CL, Tamkun JW, Mendel DB, Crabtree GR (1993) BRGI
contains a conserved domain of the SWIZ/SNFZ family necessary for normal
mitotic growth and transcription. Nature, 366, 170-174.

Kim SW, Lajara R, Rotwein P (1991) Structure and function of a human insulin-like
growth factor-1 gene promoter. Mol Endocrinol, 5, 1964-1972.

Kimmins S and MacRae TH (1999) Maturation of steroid receptors: an example of
functional cooperation among molecular chaperones and their associated
proteins. Cell Stress Chaperones, 5, 76-86.

King MC, Wieand S, Hale K, Lee M, Walsh T, Owens K, Tait J, Ford L, Dunn BK,
Costantino J, Wickerharn L, Wolmark N, Fisher B (2001) Tamoxifen and
breast cancer incidence among women with inherited mutations in BRCA1
and BRCA2: national surgical adjuvant breast and bowel project (N SABP-
Pl) Breast Cancer Prevention Trial. JAMA, 286, 2251-2256.

King WJ, Greene GL (1984) Monoclonal antibodies localize oestrogen receptor in
the nuclei of target cells. Nature, 307, 745-747.

157

Kingston RE, Narlikar GJ (1999) ATP-dependent remodeling and acetylation as
regulators of chromatin fluidity. Genes Dev, 13, 2339-2352.

Klein-Hitpass L, Ryffel GU, Heitlinger E, Cato AC (1988) A 13 bp palindrome is a
functional estrogen responsive element and interacts specifically with
estrogen receptor. Nucleic Acids Res, 16, 647-663.

Klock G, Strahle U, Schfitz G (1987) Oestrogen and glucocorticoid responsive
elements are closely related but distinct. Nature, 329, 734-736.

Klotz DM, Castles CG, Fuqua SAW, Spriggs LL, Hill SM (1995) Differential
expression of wild-type and variant ERmRNAs by stocks of MCF-7 breast
cancer cells may account for differences in estrogen responsiveness.
Biochem Biophys Res Corn, 210, 609-615.

Kobayashi Y, Kitamoto T, Masuhiro Y, Watanabe M, Kase T, Metzger D,
Yanagisawa J, Kato S (2000) p3 00 mediates functional synergism between
AF-l and AF —2 of estrogen receptor alpha and beta by interacting directly
with the N-terrninal A/B domains. J Biol Chem, 275, 15645-15651.

Koehorst SGA, Cox JJ, Donker GH, Lopes da Silva S, Burbach JPH, Thijssen JHH,
Blankenstein MA (1994) Functional analysis of an alternatively spliced
estrogen receptor lacking exon 4 isolated from MCF-7 breast cancer cells
and meningioma tissue. Mol Cell Endocrinol, 101, 237-245.

Koh SS, Chen D, Lee YH, Stallcup MR (2001) Synergistic enhancement of nuclear
receptor function by p160 coactivators and two coactivators with protein
methylatransferase activities. J Biol Chem, 276, 1089-1098.

Korach KS (1994) Insights from the study of animals lacking functional estrogen
receptor. Science, 266, 1524-1527.

Kraus WL, McInerney EM, Katzenellenbogen BS (1995) Ligand-dependent,
transcriptionally productive association of the amino- and caboxyl-terminal

regions of a steroid hormone nuclear receptor. Proc Natl Acad Sci USA, 92,
12314-12318.

Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, Sar M, Korach
KS, Gustafsson J -A, Smithies O (1998) Generation and reproductive

phenotypes of mice lacking estrogen receptor [3. Proc Natl Acad Sci USA,
95, 15677-15682.

158

Krishnan V, Wang X, Safe S (1994) Estrogen receptor-Spl complexes mediate
estrogen-induced cathepsin D gene expression in MCF-7 human breast
cancer cells. J Biol Chem, 269, 15912-15917.

Kumar V, Chambon P (1988) The estrogen receptor binds tightly to its responsive
element as a ligand-induced homodirner. Cell, 55, 145-156.

Kumar V, Green S, Stack G, Berry M, J in JR, Chambon P (1987) Functional
domains of the human estrogen receptor. Cell, 51, 941-951.

Kumar V, Guren S, Staub A, Chambon P (1986) Localization of the oestradiol-
binding and putative DNA binding domain of the human oestrogen receptor.
EMBO J, 5, 2231-2236.

Kuiper GGJ M, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson J—A (1996)
Cloning of a novel estrogen receptor expressed in rat prostate and ovary.
Proc Natl Acad Sci USA, 93, 5926-5930.

Kuroski de Bold ML (1999) Estrogen natriuretic peptides and the renin-angiotensin
system. Cardiovasc Res, 41, 524-531.

Laemmli UK (1970) Cleavage of structural proteins during assembly of the head of
bacteriophage T4. Nature, 227, 680-685.

Lamb J, Ladha MH, McMahon C, Sutherland RL, Ewen ME (2000) Regulation of
the functional interaction between cyclin D1 and the estrogen receptor. Mol
Cell Biol, 20, 8667-8675.

Lamy PJ, Pujol P, Thezenas S, Kramar A, Rouanet P, Guilleux F, Grenier J (2002)
Progesterone receptor quantification as a strong prognostic determinant in
postmenopausal breast cancer women under tamoxifen therapy. Breast
Cancer Res Treat, 76, 65-71

Laurent BC, Treich I, Carlson M (1993) The yeast SNF2/SW12 protein has DNA-
stimulated ATPase activity required for transcriptional activation. Genes
Dev, 7, 583-591.

Lavinsky RM, Jepsen K, Heinzel T, Torchia J, Mullen TM, Schiff R, Del-Rio AL,
Ricote M, Ngo S, Gemsch J, Hilsenbeck SG, Osborne DK, Glass Ck,
Rosenfeld MG (1998) Diverse signaling pathways modulate nuclear receptor
recruitment of N-CoR complexes. Proc Natl Acad Sci USA, 95, 2920-2925.

159

Lawson J S, Field AS, Champion S, Tran D, Ishikura H, Trichopoulos D (1999) Low
oestrogen receptor alpha expression in normal breast tissue underlies low

breast cancer incidence in Japan. Lancet, 354, 1787-1788.

Lazennec G, Bresson D, Lucas A, Chauveau C, Vignon F (2001) BR beta inhibits
proliferation and invasion of breast cancer cells. Endocrinology, 142,
4120-4130.

Leygue ER, Watson PH, Murphy LC (1996) Estrogen receptor variants in normal
human mammary tissue. J Natl Cancer Inst, 88, 284-290.

Li H, Leo C, Schroen DJ, Chen JD (1997) Characterization of receptor interaction

and transcriptional repression by the corepressor SMRT. Mol Endocrinol,
11, 2025-2037.

Liotta LA 1986 Tumor invasion and metastases — role of the extracellular matrix:
rhoads memorial award lecture. Cancer Res, 46,1-7.

L0pez GN, Webb P, Shinsako JH, Baxter JD, Greene GL, Kushner PJ (1999)
Titration by estrogen receptor activation function-2 of targets that are
downstream from coactivators. Mol Endocrinol, 13, 897-909.

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with
the folin phenol reagent. J Biol Chem, 193, 265-275.

Lubahn DB, Moyer J S, Golding TS, Couse JF, Korach KS, Smithies O (1993)
Alteration of reproductive function but nor prenatal sexual development after
insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad

Sci USA, 90, 1 1162-1 1166.

Ma H, Hong H, Huang SM, Irvine RA, Webb P, Kushner PJ, Coetzee GA, Stallcup
MR (1999) Multiple signal input and output domains of the 160-kilodalton
nuclear receptor coactivator proteins. Mol Cell Biol, 19, 6164-6173.

Mak HY, Hoare S, Henttu PMA, Parker MG (1999) Molecular determinants of the
estrogen receptor-coactivator interface. Mol Cell Biol, 19, 3895-3903.

Matsuda K, Ochiai I, Nishi M, Kawata M (2002) Colocalization and ligand-
dependent discrete distribution of the estrogen receptor (ER) alpha and ER
beta. Mol Endocrinol, 10, 2215-2230.

160

McCarthy TL, Centrella M, Canalis E (1989) Parathyroid hormone enhances the
transcript and polypeptide levels of insulin-like growth factor I in osteoblast-
enriched cultures from fetal rat bone. Endocrinology, 124, 1247-1253

McCarthy TL, J i C, Shu H, Casinghino S, Crothers K, Rotwein P, Centrella M
(1997) l7B-estradiol potently suppresses CAMP-induced insulin-like growth

factor-I gene activation in primary rat osteoblast cultures. J Biol Chem, 272,
18132-18139

McDonnell DP, Clemm DL, Hermann T, Goldman ME, Pike J W (1995) Analysis of
estrogen receptor function in vitro reveals three distinct classes of
antiestrogens. Mol Endocrinol, 9, 659-669.

Mclnemey EM, Tsai MJ, O’Malley BW, Katzenellenbogen BS (1996) Analysis of
estrogen receptor transcriptional enhancement by a nuclear hormone receptor
coactivator. Proc Natl Acad Sci USA, 93, 10069-10073.

McKenna NJ, Lanz RB, O’Malley BW (1999) Nuclear receptor co-regulators:
cellular and molecular biology. Endocrinol Rev, 20, 321-344.

McNally JG, Muller WG, Walker D, Wolford R, Hager GL (2000) The
glucocorticoid receptor: rapid exchange with regulatory sites in living cells.
Science, 287, 1262-1268.

Menasce LP, White GRM, Harrison CJ, Boyle JM (1993) Localization of the
estrogen receptor locus (ESR) to chromosome 6q25.1 by FISH and a simple
post-fish banding technique. Genomics, 17, 263-265.

Metivier R, Penot G, F louriot G, Pakdel F (2001) Synergism between ER-alpha
transactivation function 1 (AF -1) and AF -2 mediated by steroid receptor
coactivator protein-1: requirement for the AF -1 alpha-helical core and for a

direct interaction between the N- and C-terminal domains. Mol Endocrinol,
15, 1953-1970.

Metivier R, Petit F G, Valotaire Y, Padkel F (2000) Function of N-terminal
transactivation domain of the estrogen receptor requires a potential alpha-

helical structure and is negatively regulated by the A domain. Mol
Endocrinol, 14, 1849-1871.

Metzger K, Ali S, Bonnert JM, Chambon P (1995) Characterization of the amino-

terrninal transcriptional activation function of the human estrogen receptor in
animal and yeast cells. J Biol Chem, 270, 9535-9542.

161

Meyer ME, Gronmeyer H, Turcotte B, Bocquel MT, Tasset D, Chambon P (1989)
Steroid hormone receptors compete for factors that mediate their enhancer
function. Cell, 57, 433-442.

Migliaccio A, di Domenico M, Castoria G, de Falco A, Bontempo P, Nola E,
Auricchio F (1996) Tyrosine kinase/p21ras/MAP-kinase pathway activation
by estradiol-receptor complex in MCF-7 cells. EMBO J, 15, 1292-1300.

Miksicek RJ, Carlson KE, Hwang K-J, Katzenellenbogen JA (1995) Studies using
fluorescent tetrahydrochrysene estrogens for in situ visualization of the
estrogen receptor in living cells. Mol Endocrinol, 9, 592-604.

Miksicek RJ, Lei Y, Wang Y (1993) Exon skipping gives rise to alternatively
spliced forms of the estrogen receptor in breast tumor cells. Breast Cancer
Res Treat, 26, 163-174.

Montano MM, Jaiswal AK, Katzenellenbogen BS (1998) Transcriptional regulation
of the human quinone reductase gene by antiestrogen-liganded estrogen
receptor-alpha and estrogen receptor-beta. J Biol Chem, 273, 25443-25449.

Mosselman S, Polman J, Dijkema R (1996) ERB: identification and characterization
of a novel human estrogen receptor. F EBS Lett, 392, 49-53.

Muchardt C, Yaniv M (1993) A human homologue of Saccharomyces cerevisiae
SNF2/ SW12 and Drosophila brrn genes potentiates transcriptional activation
by the glucocorticoid receptor. EMBO J, 12, 4279-4290.

Murphy LJ, Friesen HG (1988) Differential effects of estrogen and growth hormone
on uterine and hepatic insulin-like growth factor I gene expression in the
ovariectomized hypophysectomized rat. Endocrinology, 122, 325-332

Murphy LJ, Ghahary A (1990) Uterine insulin-like growth factor-1: regulation of
expression and its role in estrogen-induced uterine proliferation.
Endocr Rev, 3, 443-453.

Murphy LC, Leygue E, Dotzlaw H, Douglas D, Coutts A, Watson PH (1997)
Oestrogen receptor variants and mutations in human breast cancer. Ann
Med, 29, 221-234.

Naar AM, Beaurang PA, Zhou S, Abraham S, Solomon W, Tijan R (1999)

Composite co-activator ARC mediates chromatin directed transcriptional
activation. Nature, 398, 828-832

162

 

Nagaoka I, Trapnell BC, Crystal RG (1990) Regulation of insulin-like growth factor
I gene expression in the human macrophage-like .cell line U937. J Clinical
Invest, 85, 448-455

Nagy L, Kao H-Y, Chakravarti D, Lin RJ, Hassig CA, Ayer DE, Schreiber SL,
Evans RM (1997) Nuclear receptor repression mediated by a complex
containing SMRT, mSin3A, and histone deacetylase. Cell, 89, 373-380.

Nagy L, Kao H-Y, Love JD, Li C, Banayo E, Gooch JT, Krishna V, Chatterjee K,
Evans RM, Schwabe J WR (1999) Mechanism of corepressor binding and
release from nuclear hormone receptors. Genes Dev, 13, 3209-3216.

Neff S, Sadowski C, Miksicek RJ (1994) Mutational analysis of cysteine residues
within the hormone-binding domain of the human estrogen receptor
identifies mutants that are defective in both DNA-binding and subcellular
distribution. Mol Endocrinol, 8, 1215-1223.

Newton CJ, Buric R, Trapp T, Brockmeier S, Pagotto U, Stalla GK (1994) The
unliganded estrogen receptor (ER) transduces growth factor signals. J
Steroid Biochem Mol Biol, 48, 481-486.

Nolte RT, Wisely GB, Westin S, Cobb JE, Lambert MH, Kuokawa R, Rosenfeld
MG, Willson TM, Glass CK, Milburn MV (1998) Ligand binding and
coactivator assembly of the peroxisome proliferator-activated receptor-
garnma. Nature, 395, 137-143.

Nuttall ME, Stroup GB, Fisher PW, Nadeau DP, Gowen M, Suva LJ (2000) Distinct
mechanisms of action of selective estrogen receptor modulators in breast and
osteoblastic cells. Am J Physiol Cell Physiol, 279, 1550-1557.

Ogawa S, Inoue S, Watanabe T, Hiroi H, Orimo A, Hosoi T, Ouchi Y, Murarnatsu
M (1998) The complete primary structure of human estrogen receptor B

(hERB) and its heterodirnerization with ERa in vivo and in vitro. Biochem
Biophys Res Corn, 243, 122-126.

Ogryzko VV, Schiltz RL, Russanova V, Howard BH, Nakatani Y (1996) The
transcriptional coactivators p300 and CBP are histone acetyltransferases.
Cell, 87, 953-959.

Ohlsson H, Lykkesfeldt AE, Madsen MW, Briand P (1998) The estrogen receptor

variant lacking exon 5 has dominant negative activity in the human breast
epithelial cell line HMT-352281. Cancer Res, 58, 4264-4268.

163

Ofiate SA, Boonyaratanakomkit V, Spencer TE, Tsai SY, Tsai M-Y, Edwards DP,
O’Malley BW (1998) The steroid receptor coactivator-l contains multiple
receptor interacting and activation domains that cooperatively enhance the
activation function 1 (AF 1) and AF 2 domains of steroid receptors. J Biol

Chem, 273,121010-12108.

Ofiate SA, Tsai SY, Tsai M-J, O’Malley BW (1995) Sequence and characterization
of a coactivator for the steroid hormone receptor superfamily. Science, 270,
1354-1357.

Owen -Hughes T, Utley RT, Cote J, Peterson CL, Workman JL (1996) Persistent
site-specific remodeling of a nucleosome array by transient action of the
SWI/SNF complex. Science, 273, 513-516.

Paech K, Webb P, Kuiper GGJM, Nilsson S, Gustafsson J-A, Kushner PJ, Scarrlan
TS (1997) Differential ligand activation of estrogen receptors ERa and ERB
at APl sites. Science, 277, 1508-1510.

Pasqualini C, Guivarc’h D, Barnier J-V, Guibert B, Vincent J -D, Vernier P (2001)
Differential subcellular distribution and transcriptional activity of 2E3, 2E4,

and 2E3-4 isoforms of the rat estrogen receptor-0t. Mol Endocrinol, 15,
894-908.

Pearson G, Robinson F, Gibson TB, Xu B-E, Karandikar M, Berman K, Cobb MH
(2001) Mitogen-activated protein (MAP) kinase pathways: regulation and
physiological functions. Endocrine Rev, 22, 153-183.

Perissi V, Staszewski LM, Mclnemey EM, Kurokawa R, Krones A, Rose DW,
Lambert MH, Milburn MV, Glass CK, Rosenfeld MG (1999) Molecular

determinants of nuclear receptor-corepressor interaction. Genes Dev, 13,
3198-3208.

Pilihowska M, Kimua N, Fujiwara H, Nagura H (1997) Immunohistochemical study
of TGF-alpha, TGF-beta, EGF R, and IGF -1 expression in human breast
carcinoma. Mod Pathol, 10, 9699-9675

Ponglikitrnongkol M, Green S, Chambon P (1988) Genomic organization of the
human estrogen receptor gene. EMBO J, 7, 3358-3388.

Poola I, Abraham J, Baldwin K (2002) Identification of ten exon deleted ER beta
mRNAs in human ovary, breast, uterus and bone tissues: alternate splicing

pattern of estrogen receptor beta mRN A is distinct from that of estrogen
receptor alpha. FEBS Lett, 516, 133-138.

164

Porter W, Saville B, Hoivik D, Safe S (1997) Functional synergy between the
transcription factor Spl and the estrogen receptor. Mol Endocrinol, 11,
1569-1580.

Price RH, Butler CA, Webb P, Uht R, Kushner P, Handa RJ (2001) A splice variant
of estrogen receptor [3 missing exon 3 displays altered subnuclear

localization and capacity for transcriptional activation. Endocrinol, 142,
203 9-2049.

Rachez C, Lemon BD, Suldan Z, Bromleigh V, Gamble M, Naar AM, Erdjurnent-
Bromage H, Tempst P, Freedman LP (1999) Ligand-dependent transcription

activation by nuclear receptors requires the DRIP complex. Nature, 398,
824-828.

Rameh LE and Cantley LC (1999) The role of phosphinositide 3-kinase lipid
products in cell function. J Biol Chem, 274, 8347-8350.

Razandi M, Pedram A, Greene GL, Levin ER (1999) Cell membrane and nuclear
estrogen receptors (ERs) originate from a single transcript: studies of

ERa and ERB expressed in chinese hamster ovary cells. Mol Endocrinol, 13,
307-319.

Rea D, Parker MG (1996) Effects of an exon 5 variant of the estrogen receptor in
MCF-7 breast cancer cells. Cancer Res, 56, 1556-1563.

Robyr D, Wolffe AP, Walter W (2000) Nuclear hormone receptor co-regulators in
action: diversity for shared tasks. Mol Endocrinol, 14, 329—347.

Sabbah M, Courilleau D, Mester J, Redeuilh G (1999) Estrogen induction of the
cyclin D1 promoter: involvement of a CAMP response-like element. Proc
Natl Acad Sci USA, 96, 11217-11222

Sadovsky Y, Webb P, Lopez G, Baxter JD, Fitzpatrick PM, Giang-Ginsberg E,
Cavailles V, Parker G, Kushner PJ (1995) Transcriptional activators differ in
their responses to over expression of TATA-box-binding protein. Mol Cell
Biol, 15, 1554-1563.

Sanders M, McKnight GS (1988) Positive and negative regulatory elements control
the steroid-responsive ovalbumin promoter. Biochem, 27, 6550-6557

Scatchard G (1949) The attractions of proteins for small molecules and ions. Ann
NY Acad Sci, 51, 660-672.

165

 

Schwabe J WR, Rhodes D (1991) Beyond zinc fingers: steroid hormone receptors
have a novel structural motif for DNA recognition. Trends Biochem Sci,
16, 291-296.

Schweers LA, Frank DE, Weigel NL, Sanders MM (1990) The steroid-dependent
regulatory element in the ovalbumin gene does not function as a typical
steroid-response element. J Biol Chem, 265, 7590-7595.

Sewack GF, Ellis TW, Hansen U (2001) Binding of TATA binding protein to a
naturally positioned nucleosome is facilitated by histone acetylation. Mol
Cell Biol, 21,1404-1415.

Sewack GF and Hansen U (1997) Nucleosome positioning and transcrition-
associated chromatin alterations on the huan estrogen-responsive pS2
promoter. J Biol Chem, 272, 31118-31129.

Shang Y and Brown M (2002) Molecular determinants for the tissue specificity of
SERMS. Science, 295, 2465-2468.

Sheppard HM, Harries JC, Hussain S, Bevan C, Heery DM (2001) Analysis of the
steroid receptor coactivator 1(SRC1)-CREB binding protein interaction
interface and its importance for the function of SRCl. Mol Cell Biol, 21,
39-50.

Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ, Agard DA, Greene GL
(1998) The structural basis of estrogen receptor/coactivator recognition and
the antagonism of this interaction by tamoxifen. Cell, 95, 927-937.

Speirs V, Malone C, Walton DS, Kerin MJ, Atkin SL (1999) Increased expression
of estrogen receptor beta mRN A in tamoxifen-resistant breast cancer
patients. Cancer Res, 59, 5421-5424.

Sylvia VL, Boyan BD, Dean DD, Schwartz Z (2000) The membrane effects of 170-
estradiol on chondrocyte phenotypic expression are mediated by activation of
protein kinase C through phospholipase C and G-proteins. J Steroid
Biochem Mol Biol, 73, 211-224.

Simoncini T, Hafezi-Moghadam A, Brazil DP, Ley K, Chin WW, Liao JK (2000)

Interaction of oestrogen receptor with the regulatory subunit of
phosphatidylinositol-3-OH kinase. Nature, 407, 53 8-541.

166

Sluyser M 1995 Mutations in the estrogen receptor gene. Human Mut, 6, 97-103.

Smigel K (1998) Breast cancer prevention trial shows major benefit, some risk. J
Natl Cancer Inst, 90, 647-648.

Smith CL, Nawaz Z, O’Malley BW (1997) Coactivator and corepressor regulation
of the agonist/antagonist activity of the mixed antiestrogen, 4-
hydroxytarnoxifen. Mol Endocrinol, 11, 657-666.

Smith CL, Ofiate SA, Tsai M-J, O’Malley BW (1996) CREB binding protein acts
synergistically with steroid receptor coactivator-l to enhance steroid
receptor-dependent transcription. Proc Natl Acad. Sci USA, 93, 8884-8888.

Spencer TE, Jenster G, Burcin MM, Allis CD, Zhou J, Mizzen CA, McKenna NJ,
Oflate SA, Tsai SY, Tsai M-J, O’Malley BW (1997) Steroid receptor
coactivator-1 is a histone acetyltransferase. Nature, 389, 194-198.

Srabile LP, Davis AL, Gubish CT, Hopkins TM, Lukerich JD, Christie N,
Finkelstein S, Siegfried JM (2002) Human non-small cell lung tumors and
cells derived from normal lung express both estrogen receptor alpha and beta
and show biological responses to estrogen. Cancer Res, 62, 2141-2150.

Stewart CEH, Rotwein P (1996) Growth, differentiation, and survival: multiple
physiological function for insuling-like grth factors. Physiological Rev,
76, 1005-10026

Sun G, Porter W, Safe S (1998) Estrogen-induced retinoic acid receptor alpha 1

gene expression role of estrogen receptor-Spl complex. Mol Endocrinol, 12,
882-890.

Takeshita A, Yen PM, Misiti S, Cardona GR, Liu Y, Chin WW (1996) Molecular
cloning and properties of a full-length putative thyroid horone receptor
coactivator. Endocrinol, 137, 3594-3597.

Tanenbaum DM, Wang Y, Williams SP, Sigler PB (1998) Crystallographic
comparison of the estrogen and progesterone receptor’s ligand binding
domains. Proc Natl Acad Sci USA, 95, 5998-6003.

Tof’t D, Gorski J (1966) A receptor molecule for estrogens: isolation from the rat
uterus and preliminary characterization. Proc Natl Acad Sci USA, 53,
1574-1581.

167

Tora L, Gaub MP, Mader S, Dierich A, Bellard M, Chambon P (1988) Cell-specific
activity of a GGTCA half-palindromic oestrogen-responsive element in the
chicken ovalbumin gene promoter. EMBO J, 7, 3771-3778.

Tora L, White J, Brou C, Tasset D, Webster N, Scheer E, Chambon P (1989) The
human estrogen receptor has two independent nonacidic transcriptional
activation functions. Cell, 59, 477-487.

Toran-Allerand CD, Singh M, Setalo Jr. G (1999) Novel mechanisms of estrogen
action in the brain: new players in an old story. Front Neuroendocrin, 20,
97-121.

Traish AM, Al-Fadhli S, Klinge C, Kounine M, Quick TC (1995) Identification of
structurally altered estrogen receptors in human breast cancer by site-directed
monoclonal antibodies. Steroids, 60, 467-474.

Tremblay A, Tramblay GB, Labrie F, Giguere V (1999) Ligand-independent
recruitment of SRC-l to estrogen receptor beta through phosphorylation of
activation function AF -1. Mol Cell, 3, 513-519.

Tremblay GB, Tremblay A, Copeland NG, Gilbert DJ, Jenkins NA, Labrie F,
Giguere V (1997) Cloning, chromosomal localization, and firnctional
analysis of the murine estrogen receptor beta. Mol Endocrinol, 11, 353-365.

Truss M, Beato M (1993) Steroid hormone receptors: interaction with
deoxyribonucleic acid and transcription factors. Endocrine Rev, 14, 459-479

Tsai M-J, O’Malley BW (1994) Molecular mechanisms of action of steroid/thyroid
receptor superfamily members. Annu Rev Biochem, 63, 451-486.

Uht RM, Anderson CM, Webb P, Kushner PJ (1997) Transcriptional activities of
estrogen and gluccocorticoid receptors are functionally integrated at the
AP-l element. Endocrinology, 138, 2900-2908.

Umayahara , Kawamori R, Watada H, Irnano E, Iwarna N, Morishima T, Yarnasaki
Y, Kajimoto Y, Karnada T (1994) Estrogen regulation of the insulin-like
growth factor 1 gene transcription involves an AP-l enhancer. J Biol Chem,
269, 16433-16442.

Umayahara Y, Billiard J, Ji C, Centrella M, McCarthy TL, Rotwein P (1999)
CCAAT/enhancer-binding protein 8 is a critical regulator of insulin-like
grth factor-I gene transcription in osteoblasts. J Biol Chem, 274,
10609-10617.

168

 

van der Burg B, de Groot RP, Isbrficker L, Kruijer W, de Laat SW (1991) Oestrogen
directly stimulates growth factor signal transduction pathways in human
breast cancer cells. J Steroid Biochem Mol Biol, 40, 215-221.

van Dijk MA, Floore AN, Kloppenborg KI, van’t Veer LJ (1997) A functional assay
in yeast for the human estrogen receptor displays wild-type and variant
estrogen receptor messenger RNAs present in breast carcinoma. Cancer Res,
57, 3478-3485.

Vuong H, Patterson T, Shapiro P, Kalvaklanu DV, Wu R, Ma WY, Dong Z,
Kleeberger SR, Reddy SP (2000) Phorbol ester-induced expression of airway
squamous cell differentiation marker, SPRRl B, is regulated by protein
kinase Cd/Ras/MEKK17MKK1-dependent/AP-l signal transduction
pathway. J Biol Chem, 275, 32250-32259.

Voegel JJ, Heine MJ S, Zechel C, Chambon P, Gronemeyer H (1996) TIF2, a 160
kDa transcriptional mediator for the ligand-dependent activation function
AF -2 nuclear receptors. EMBO J, 15, 3667-3675.

Wade CB, Robinson S, Shapiro RA, Dorsa DM (2001) Estrogen Receptor ERor and
ERB exhibit unique pharmacologic properties when coupled to activation of
the mitogen-activated protein kinase pathway. Endocrinol, 142, 2336-2342.

Wakeling AE, Dukes M, Bowler J (1991) A potent specific pure antiestrogen with
cinical potential. Cancer Res, 51, 3867-3873.

Wang Y, Miksicek RJ (1991) Identification of a dominant negative form of the
human estrogen receptor. Mol Endocrinol, 5, 1707-1715.

Watanabe M, Yanagisawa J, Kitagawa H, Takeyarna K, Ogawa S, Arao Y, Suzawa
M, Kobayashi Y, Yano T, Yoshikawa H Masuhiro Y, Kato S (2001) A
subfamily of RNA-binding DEAD-box proteins acts as an estrogen receptor
alpha coactivator through the N-terrninal activation domain (AF -1) with an
RNA coactivator, SRA. EMBO J, 20, 1341-1352.

Welshons WV, Lieberman ME, Gorski J (1984) Nuclear localization of unoccupied
oestrogen receptors. Nature, 307, 747-749.

Webb P, Anderson CM, Valentine C, Nguyen P, Marimuthu A, West BL, Baxter
JD, Kushner PJ (2000) The nuclear receptor corepressor (N -CoR) contains
three isoleucine motifs (I/LXXII) that serve as receptor interaction domains
(IDs). Mol Endocrinol, 14, 1976-1985.

169

 

Webb P, Lopez GN, Uht RM, Kushner PJ (1995) Tamoxifen activation of the
estrogen receptor/AP-l pathway: potential origin for the cell-specific
estrogen-like effects of antiestrogens. Mol Endocrinol, 9, 443-456.

Webb P, Nguyen P, Shinsako J, Anderson C, F eng W, Nguyen MP, Chen D, Huang
S-M, Subramanian S, McKinemey E, Katzenellenbogen BS, Stallcup MR,
Kushner PJ (1998) Estrogen receptor activation fimction 1 works by binding
p160 coactivator proteins. Mol Endocrinol, 12, 1605-1618.

Westin S, Rosenfeld MG, Glass CK (2000) Nuclear receptor coactivators. Adv
Pharrnacol, 47, 89-112.

Westwick JK, Weitzel C, Minden A, Karin M, Brenner DA (1994) Tmnor necrosis
factor a stimulates AP-l activity through prolonged activation of the c-Jun
kinase. J Biol Chem, 269, 2639-26401.

Whitham SE, Murphy G, Angel P, Rahmsdorf H-J, Smith BJ, Lyons A, Harris TJR,
Reynolds JJ, Hierrlich P, Docherty AJP (1986) Comparison of human
stromelysin and collagenase by cloning and sequence analysis. Biochem J,
240, 913-916.

Wijayaratne AL, Nagel SC, Paige LA, Christensen DJ, Norris JD, Fowlkes DM,
McDonnell DP (1999) Comparative analyses of mechanistic differences
among antiestrogens. Endocrinol, 140, 5828-5840.

Wiseman LR, Johnson MD, Wakeling AE, Lykkesfeldt AE, May FEB, Westley BR
(1993) Type I lGF receptor and acquired tamoxifen resistance in oestrogen-
responsive human breast cancer cells. Eur J Cancer, 29A, 2256-2264.

Wolf KM, Jordan VC (1994) Characterization of tamoxifen stimulated MCF-7
tumor variants grown in athymic mice. Breast Cancer Res Treat, 31,
117-127.

Wolf M, Ingbar SH, Moses AC (1989) Thyroid hormone and growth hormone
interact to regulate insulin-like growth factor-I messenger ribonucleic acid
and circulating levels in the rat. Endocrinology, 125, 2905-2914

Wolffe AP (1994) Transcription: in tune with the histones. Cell, 77, 13-16.
Wong C-W, Privalsky ML (1998) Transcriptional repression by the SMRT-mSin3

corepressor: multiple interactions, multiple mechanisms, and a potential role
for TFIIB. Mol Cell Biol, 18, 5500-5510.

170