USE OF A TRANSGENIC MOUSE MODEL OF OVARIAN

HYPERSTIUMLUATION TO IDENTIFY THERAPEUTIC TARGETS AND

MECHANISMS IN HORMONE-INDUCED MAMMARY CANCER

by

ERIN LEE MILLIKEN

Submitted in partial fulfillment of the requirements

For the degree of Doctor of Philosophy

Thesis Advisor: Dr. Ruth A. Keri

Department of Pharmacology

CASE WESTERN RESERVE UNIVERSITY

August, 2005 CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

______

candidate for the ______degree *.

(signed)______(chair of the committee)

______

______

______

______

______

(date) ______

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

I dedicate this dissertation to Matt, whose patience and optimism I have drawn on repeatedly

over the past six years

and

to my grandmother, Virginia Gerrity

iii TABLE OF CONTENTS

Dedication iii List of Tables v List of Figures vi Acknowledgments viii Abstract x Chapter I Introduction 1 Chapter II Ovarian Hyperstimulation by Luteinizing 35 Hormone Leads to Mammary Gland Hyperplasia and Cancer Predisposition in Transgenic Mice Chapter III EB1089, a Vitamin D Receptor Agonist, Reduces 77 Proliferation and Decreases Tumor Growth Rate in a Mouse Model of Hormone-Induced Mammary Cancer Chapter IV Centrosome Amplification and p53 Signaling in a 98 Model of Hormone-Induced Mammary Cancer Chapter V Summary, Future Directions, and Conclusions 150 Reference List 177

iv LIST OF TABLES

Table II-1 Serum levels of estrogen, progesterone, and 60 prolactin are increased in LH-overexpressing animals Table IV-1 Genes associated with p53, 120 centrosomes/microtubules/polarity, and genomic instability Table IV-2 Genes upregulated in mammary glands of LH- 122 overexpressing relative to wild-type mice Table IV-3 Genes downregulated in mammary glands of LH- 129 overexpressing relative to wild-type mice

v LIST OF FIGURES

Figure I-1 Mammary gland development 29

Figure I-2 Chemical structures of 1, 25-dihydroxyvitamin D3 31 and EB1089 Figure I-3 Generation of LH-overexpressing mice 33 Figure II-1 Mammary gland development is accelerated in 61 LH-overexpressing mice Figure II-2 The mammary glands in adult, virgin LH- 63 overexpressing mice display a mid-pregnancy morphology Figure II-3 The mammary glands of adult, virgin LH- 65 overexpressing mice display an increase in epithelial cell proliferation, but no change in apoptosis Figure II-4 The mammary glands of adult virgin LH- 67 overexpressing mice display a mid-pregnancy phenotype at the molecular level Figure II-5 Mammary gland hyperplasia in LH- 69 overexpressing mice is dependent on ovarian input Figure II-6 LH-overexpressing mice are predisposed to 71 mammary cancer Figure II-7 DMBA-induced and spontaneous mammary 73 cancers in LH-overexpressing mice Figure II-8 Spontaneous mammary tumors from LH- 75 overexpressing largely lack expression of progesterone and estrogen receptors Figure III-1 Mammary gands of LH-overexpressing mice 92 demonstrate elevated expression of vitamin D

vi receptor Figure III-2 EB1089 decreases mammary epithelial cell 94 proliferation in LH-overexpressing mice Figure III-3 EB1089 displays anti-tumorigenic activity in a 96 subset of hormone-induced mammary tumors Figure IV-1 Pre-neoplastic mammary glands and mammary 139 tumors from LH-overexpressing mice display centrosome amplification Figure IV-2 A mutant form of p53 does not alter the latency of 141 hormone-induced mammary tumorigenesis Figure IV-3 mRNA levels of regulators of p53 stability in 143 mammary glands and tumors of LH- overexpressing mice Figure IV-4 p53 is activated in the mammary glands of wild- 145 type and LH-overexpressing mice by ionizing radiation Figure IV-5 Mammary epithelial cells of wild type and LH- 147 overexpressing animals demonstrate variable apoptotic response to ionizing radiation Figure V-1 Tamoxifen inhibits ductal branching and alveolar 171 bud formation in LH-overexpressing mice Figure V-2 Hormone-induced mammary gland hyperplasia 173 persists in the presence of EB1089 Figure V-3 EB1089 fails to inhibit the formation of hormone- 175 induced mammary gland hyperplasia

vii ACKNOWLEDGEMENTS

My graduate school experience has been influenced by a number of

people. First and foremost among these is my thesis advisor and mentor, Ruth

Keri. Ruth has given me endless support and encouragement, as well as much

needed sympathy throughout my time in her lab. In addition to molding me into a

scientific thinker, she has taught me, by example, the importance and benefit of

being open to the constructive criticism of others. I would also like to

acknowledge John Nilson, my scientific “grandfather”, whose confidence in me

has been unwavering and who has provided me with numerous opportunities to

experience the world of science outside the laboratory. I also appreciate the

guidance and input of the remaining members of my thesis committee – Amy

Wilson-Delfosse, Lloyd Culp, and Paul MacDonald.

Interacting with other students in a variety of settings has been the most

intellectually stimulating component of my graduate education. I didn’t really

understand the concept of “critical analysis” until I experienced prelim I

preparation with Helai Mohammad and Amelia Sutton. These sessions and the

many others that followed with many different students always challenged me

and taught me more than I learned in any class. The students in my own lab

have had a particularly important part in my education and life over the past

several years. Melissa Landis has been an example and a friend to me from my

first day in the Pharmacology Department. Her thoughtfulness and careful

consideration of scientific questions, as well as her patience and positive attitude

have amazed and inspired me. Jonathan Mosley, questioner of all ideas and

viii conventions, has, much to my benefit, afforded me many stimulating discussions

in and beyond the realm of science. My protracted tenure as a graduate student

has also allowed me to learn from “younger” students, like Jen Yori and Marjorie-

Montanez-Wiscovich, as they ask questions I’ve never thought to ask.

The work presented in this dissertation was made possible by

contributions from many other people. Kristen Lozada has tirelessly monitored all

of my mice and has made sure that I had all the reagents and equipment I

needed, even at a moments notice. Much of the early work on the CTP mice was

done by former Keri lab members Alireza Behrooz and Becky Ameduri. The VDR

work was done in conjunction with Xiaoxue Zhang in the Paul MacDonald lab.

The MRI experiments were made possible by Chris Flask and Jeff Duerk. Ira

Whitten contributed substantially to the assessment of centrosome amplification

in the CTP mammary glands. Melissa Landis performed the breeding and

palpating of the p53R172H/CTP mice. Finally, although their contributions may

not appear in any of the figures presented here, I would like to acknowledge

other members of the Keri lab: Lenka Yunk, my faithful undergraduate, and

Darcie Seachrist, who has provided not only expert technical advice, but a great

deal of encouragement and support.

In ending, I would like to acknowledge the incredible support I have

received from my family – Matt, Mom, Dad, Daniel, Phil, Evan, and all the

Gerritys, McCallions, Lees, Caseys, Fines, and Millikens. I am ever thankful that

their love does not depend on whether or not my experiments work and that,

even though they’re proud of me, this PhD is immaterial.

ix Use of a Transgenic Mouse Model of Ovarian Hyperstimulation to Identify

Therapeutic Targets and Mechanisms in Hormone-Induced Mammary

Cancer

Abstract

By

Erin L. Milliken

Epidemiological studies and clinical trials have revealed the critical impact that reproductive hormones have on breast cancer. The work described in this dissertation characterizes a transgenic mouse model, the LH-overexpressing mouse, which develops mammary gland hyperplasia and tumorigenesis in response to ovarian hyperstimulation, making it a unique model of hormone- induced mammary cancer. The hyperplasia in these mice is ovary-dependent and is due to a dramatic increase in proliferation of mammary epithelial cells.

Most of the spontaneous mammary tumors that form in this model are mammary intraepithelial neoplasias that lack expression of both estrogen and progesterone receptors.

The LH-overexpressing mice have been utilized to investigate potential therapeutic targets and identify mechanisms that contribute to hormone-mediated mammary gland pathology. Similar to observations made in human breast cancer, the pre-neoplastic mammary glands and tumors of LH-overexpressing mice demonstrate increased expression of vitamin D receptor, supporting the notion that this protein is a potential target for therapy. Treatment with EB1089, a vitamin D receptor agonist, results in decreased proliferation in the pre-neoplastic

x mammary glands of LH-overexpressing mice and reduces the growth rate of a

subset of established mammary tumors, providing evidence for both

chemopreventive and chemotherapeutic benefits of activating the vitamin D

receptor.

Analysis of pre-neoplastic mammary glands of LH-overexpressing mice

has revealed the presence of centrosome amplification, suggesting that genomic

instability may contribute to hormone-induced mammary tumorigenesis. Although

centrosome amplification and genomic instability in human cancers are often

associated with mutations in the p53 protein, the p53 sequence is intact in mammary tumors of LH-overexpressing mice. Furthermore, p53 is capable of being functionally activated in the mammary glands of transgenic mice in response to ionizing radiation, as evidenced by increased phosphorylation, upregulation of target genes, and induction of apoptosis. Along with the fact that introduction of a mutant form of p53 does not alter tumor latency, these data suggest that hormone-induced mammary tumors is not dependent on direct perturbation of the p53 signaling pathway. Future studies using the LH- overexpressing mouse should provide further insight into both potential therapeutic targets for treatment of hormone-induced mammary cancer and the mechanisms through which this disease progresses.

xi CHAPTER I

INTRODUCTION

Breast cancer is the most common cancer among American women, who exhibit a 1 in 8 probability of being diagnosed over the course of their lifetime (1).

Although mortality from breast cancer has steadily declined since the early

1990s, incidence has increased over the same time period and the United States will spend approximately $6.5 billion on the treatment of breast cancer this year

(2). Hence, incentive remains to further elucidate the mechanisms of breast cancer in order to facilitate prevention and improve treatment.

Risk Factors for Breast Cancer Development

A number of factors impact a woman’s risk of developing this disease, including age, family history, exposure to radiation, lifestyle behaviors and reproductive history. As is the case for many cancers, increased age is the weightiest prognostic indicator for breast cancer; women above the age of 60 are over seventy times more likely to be diagnosed with this disease over the next 10 years than women in their twenties (1). Genetic factors also play a role; the risk of diagnosis increases approximately two-fold if a female relative has previously been diagnosed (3). A significant portion of familial breast cancers (84%) have been linked to mutations in two genes: BRCA1 and BRCA2 (4). Women who carry mutations in either of these genes have a lifetime breast cancer risk of up to 87% (5,6); however, it is estimated that these cases account for only 5% of the

1 overall population burden of breast cancer (7). Lifestyle and behavior are other

contributors to breast cancer risk. Extensive data correlating body composition

and exercise with breast cancer risk has led the International Agency for

Research on Cancer to estimate that obesity and sedentary lifestyle account for

25% of breast cancers worldwide (8,9).

The timing and occurrence of reproductive events also have a significant

impact on a woman’s risk of developing breast cancer. Early menarche and late

menopause both impart increased risk (10-12), suggesting that extended

exposure to the hormonal milieu associated with normal menstrual cycles

contributes to this disease. This notion is supported by the fact that removal of

ovarian hormones by pre-menopausal bilateral oophorectomy conveys a

protective effect (13). Concordantly, pregnancy, which is associated with

increased levels of circulating hormones, causes a transient increase in breast

cancer risk compared to nulliparous women (14). However, the effects of pregnancy are multifarious; epidemiological studies consistently demonstrate that nulliparous women have a higher lifetime risk of developing breast cancer than their parous counterparts. Moreover, the timing of pregnancy also has influence:

the younger a woman is when she undergoes her first full-term pregnancy, the less likely she is to develop breast cancer over the course of her lifetime (15,16).

Several hypotheses have been proposed to explain parity-induced protection.

Full-term pregnancy has been suggested to decrease the population of cells predisposed to cancer by causing terminal differentiation of susceptible cells, by inducing selective loss of these cells through the apoptotic process of involution

2 (17), or by producing persistent changes in gene expression (18). Alternatively, persistent post-partum changes in mammogenic hormones have been documented in both humans and rodent models (19-21), supporting a global physiological explanation for this phenomenon.

The strong relationship between reproductive events and breast cancer occurrence has prompted extensive clinical investigation in an attempt to correlate breast cancer risk with altered serum levels of one or multiple reproductive hormones; however, the outcomes have not been entirely satisfying.

The most definitive and repeatable result has been the positive correlation between circulating levels of estradiol and breast cancer risk in postmenopausal women (22). Several groups have performed prospective studies on levels of steroid hormones such as estradiol, progesterone and testosterone in premenopausal women, but the vast majority have not found statistically significant associations with breast cancer risk (22-26). Despite efforts for normalization, this is likely due, at least in part, to variability in serum measurements of steroid hormones over the course of the menstrual cycle. It has not been any easier to identify a clear relationship between mammogenic peptide hormones and breast cancer. Serum prolactin levels have been found to be unrelated to breast cancer risk in one prospective study (27) and related in another (28).

It has become apparent that despite their classical characterization as endocrine growth factors, many hormones are able to act in a paracrine, autocrine or intracrine manner within a variety of tissues, including the breast.

3 Postmenopausal estrogen production in the skin, bone and brain is vitally

important for curtailing the detrimental effects of hormone loss on these tissues

(29). Adipose tissue is another extra-ovarian site of estrogen synthesis, which primarily occurs through the conversion of circulating androstenedione to estrone

by locally expressed aromatase. Adipose cells associated with normal breast

tissue have relatively low expression of aromatase; however, tumor-associated

factors cause a dramatic upregulation of this enzyme in stromal or interstitial cells

in more than 70% of breast carcinomas (30,31). Steroid sulfatase, which

produces estrone by hydrolysis of estrone sulfate, is also upregulated in 70-90%

of breast carcinomas (32,33). As a result, the concentration of estradiol in breast

cancer tissue from postmenopausal women is often considerably higher than that

in circulation (34). Peptide hormones can also be expressed within the breast.

Growth hormone is expressed in normal and neoplastic human breast (35).

Prolactin, which has been implicated in mammary cancer in multiple rodent

models (36-39), is expressed, along with its receptor, in the epithelial cells of the

vast majority of normal breast tissues and breast cancers that have been

assessed (40). This may explain why anti-prolactin therapies, such as

bromocriptine, which inhibit secretion from the pituitary, have had little to no

clinical success (41,42).

Breast Cancer Therapy by Hormonal Modulation

The role of hormones in the induction and progression of tumorigenesis

has been reinforced by the dramatically beneficial effects of anti-hormone

4 therapies. The most widely studied and utilized of these drugs limit estrogen signaling in the breast.

Selective Estrogen Receptor Modulators

Perhaps the most revolutionary advancement in breast cancer therapy

over the past thirty years has been the development of selective estrogen

receptor modulators (SERMs). These compounds provoke different estrogen

receptor (ER) responses depending on the target tissue, creating the potential to

maintain positive estrogenic effects in some tissues, while antagonizing estrogen

in the breast. One of the earliest and most widely used SERMs is tamoxifen,

which is an ER agonist in bone, liver and the cardiovascular system; an

antagonist in breast and brain; and displays mixed activity in the uterus. An

overview of 55 randomized clinical trials involving tamoxifen therapy, including a

total of 37,000 women, revealed substantial benefit with regard to recurrent

disease, contralateral breast cancer, mortality, and disease free survival. While

improved outcome was observed regardless of age; menopausal status; nodal

involvement; or the presence of metastatic disease, benefit was strongly related

to the presence of estrogen receptor in the primary tumor. Women with ER-

positive tumors treated with tamoxifen for 5 years demonstrated a 50% decrease

in recurrence and a 28% decrease in mortality, while their ER-negative

counterparts did not exhibit any evidence of benefit (43). The profound effect of

tamoxifen therapy prompted investigation into its potential as a chemopreventive

agent. The Breast Cancer Prevention Trial implemented by the National Surgical

5 Adjuvant Breast and Bowel Project administered tamoxifen or placebo to women

at high risk of developing breast cancer. Overall, tamoxifen resulted in a 49%

decrease in invasive breast cancer and a 50% decrease in noninvasive breast cancer; however, the benefit was highly selective for estrogen receptor positive

tumors. Women treated with tamoxifen developed 69% fewer ER-positive

tumors, while incidence of ER-negative tumors was not significantly altered (44).

Aromatase Inhibitors

Recently, aromatase inhibitors, which prevent the biosynthesis of estrogen from androstenedione, have provided an additional approach to anti-hormonal

therapy of postmenopausal women with breast cancer. Combined data from two clinical trials reveals that one such aromatase inhibitor, anastrozole, is as effective as, and sometimes more effective than, tamoxifen in the treatment of postmenopausal women with advanced breast carcinoma (45). Preliminary reports from the ATAC (Arimidex (anastrozole), Tamoxifen Alone or in

Combination) trial have also suggested that anastrozole treatment alone of early

breast cancer provides extended disease free survival and decreased risk of

contralateral disease compared to tamoxifen (46,47). Combination studies

indicate that administration of another aromatase inhibitor, letrozole, to women

following 2-5 years of tamoxifen treatment improves outcome (48,49).

Furthermore, although patients treated with aromatase inhibitors demonstrate

increased incidence of bone fracture and musculoskeletal disease, many side

effects commonly caused by tamoxifen, such as vaginal bleeding, venous

6 thrombosis, and increased risk of endometrial cancer are not observed with

aromatase inhibitors. These positive clinical outcomes have prompted a recent

recommendation from the American Society of Clinical Oncology for the inclusion

of an aromatase inhibitor in the adjuvant therapy paradigm for postmenopausal women with ER-positive breast cancers (50). It is of note to mention that

aromatase inhibitors are not recommended for premenopausal breast cancer

patients, as early studies indicated that these therapies are unable to adequately

suppress estrogen production from a fully functional ovary (51).

Mouse Models of Mammary Gland Development and Pathology

Advances in breast cancer treatment will be catalyzed by increased

understanding of the developmental processes of the breast. The morphology

and development of the mouse mammary gland is exceedingly similar to that of

the human breast (52); hence, substantial insight into the process of breast

development has been amassed by the ability to surgically and genetically

manipulate the laboratory mouse.

Growth and Development of the Mammary Gland

The mammary gland is primarily composed of three cells types: epithelial

cells, which make up the milk secreting alveoli and milk-carrying ducts;

myoepithelial cells, which line the basal border of the luminal epithelial cells and

contract to induce milk flow during lactation; and stromal cells, which provide

structural and molecular support for the mammary gland. Extensive

7 communication takes place between the different cellular compartments.

Development of the mammary gland is driven by steroid and peptide hormones,

which induce an assortment of local signaling molecules and second

messengers in order to orchestrate proper construction of a mammary gland

capable of producing sufficient milk during lactation (reviewed in (53)).

At birth, the mammary gland anlagen consists of a very rudimentary ductal

system in the vicinity of the nipple (Figure I-1). It remains in such a state until the

onset of puberty, at which time increases in reproductive hormones awaken the

gland and recommence development. The earliest morphological event of puberty is the appearance of terminal end buds (TEBs), specialized structures of highly proliferative epithelial cells at the ends of the existing ducts (Figure I-1).

These TEBs invade the fat pad of the mammary gland, giving rise to new luminal

and myoepithelial cells, and persist until the ducts reach the end of the fat pad,

approximately 3 weeks after the onset of puberty in the mouse. This process is

dependent on estrogen signaling as the estrogen receptor α knockout mice

(αERKO) exhibit severely stunted ductal elongation (54). Mammary gland

transplant studies reveal that both epithelial and stromal estrogen receptors are

required for appropriate ductal development in the presence of physiological

levels of estrogen (55), reinforcing the critical interplay between different cellular

compartments of the mammary gland. Recent studies have also identified a role,

although not a requirement, for progesterone in ductal elongation (56), which

appears to be mediated through stromal progesterone receptor (PR) (57). As the

8 ducts extend, secondary and tertiary branches arise, forming an arborized

network of epithelial cells that fills the mammary fat pad.

The disappearance of the terminal end buds signifies the end of pubertal

development and the completion of the adult mammary gland (Figure I-1);

however, the gland does not remain in a stagnant state. Successive rounds of

morphological and molecular changes occur regularly in response to the normal

hormonal oscillations of the estrous cycle. Fata et al. documented significant

morphological variation of mouse mammary glands (58). Some ductal trees

examined by whole mount exhibited limited secondary branching and were

devoid of alveolar structures, which are precursors to the milk forming units that

form during pregnancy; others displayed extensive higher order branching and numerous alveolar buds. The highest order morphology was exclusive to mice in diestrus, which is similar to the luteal phase of the human menstrual cycle and is characterized by high levels of progesterone. Glands of mice in diestrus also demonstrated the highest rates of proliferation and apoptosis (58).

Immunohistochemical analyses of human breast tissue have similarly revealed that proliferation and apoptosis of epithelial cells are more common during the luteal than the follicular phase of the menstrual cycle (59). Furthermore, cyclic changes in the expression of apoptosis related genes, such as Bcl-2 and Bax

(59); and extracellular matrix genes, such as laminin, fibronectin, and multiple collagens (60), have also been documented. These inherent variations in tissue cellularity and molecular characteristics suggest that the mammary gland

9 response to exogenous stimuli, including tumorigenic insults, may vary with stage of the estrous or menstrual cycle.

When pregnancy occurs, the alveolar buds of diestrus are maintained and differentiate into lobulo-alveolar structures, which, upon parturition, fulfill the objective of the mammary gland by completing differentiation into milk-forming units (Figure I-1). Analysis of the progesterone receptor knockout mouse (PRKO) mammary gland demonstrates that progesterone signaling through epithelial PR is essential for the tertiary branching and lobulo-alveolar formation that occurs at this time (57). Proliferation and functional differentiation of the lobulo-alveoli are also dependent on prolactin signaling through Jak2-Stat5a, as evidenced by knockout models of prolactin, prolactin receptor, and Stat5a, as well as the conditional Jak2 knockout, all of which demonstrate severe impairment in this capacity (61-64).

Upon weaning, milk stasis in the mammary gland initiates involution, which, through apoptosis and tissue remodeling, regenerates a tissue that is morphologically similar to that of a virgin animal (Figure I-1). Although morphological changes commence rapidly upon cessation of milk removal, the progression is reversible for up to 48 hours; it is at this time that massive apoptosis begins and the gland becomes committed to the process of involution.

The second phase of involution, which begins 3-5 days after pup removal in mice, includes degradation of basement membranes; alveolar collapse; and macrophage infiltration, culminating with the elimination of 50-80% of the mammary epithelial cells present in the lactating gland (reviewed in (65)). Gene

10 expression changes reflect the global changes that are occurring. Decreased

expression of milk protein genes during the early stages of involution are

paralleled by increases in apoptotic genes, such as Bax (66) and caspase-1 (67).

The final stages of remodeling are protease driven and are accompanied by

increased expression of metalloproteases and serine proteases like stromolysin-

1, stromolysin-3, gelatinase A, and urokinase-type plasminogen activator (68,69).

Vitamin D Receptor in Mammary Gland Development and Tumorigenesis

Although maintaining mineral homeostasis and sustaining skeletal integrity

are its classical roles, the vitamin D endocrine system has recently been shown to participate in a plethora of physiological functions including hair follicle cycling,

blood pressure regulation and mammary gland development (reviewed in (70)).

Vitamin D3, which is acquired through the diet or produced in the skin upon exposure to ultraviolet light, is converted to its most bioactive form, 1,25- dihydroxyvitamin D3 [1,25-(OH)2D3], by a series of hydroxylations. 1,25-(OH)2D3 transmits its signal by binding to the vitamin D receptor (VDR), a member of the nuclear receptor superfamily; this ligand-receptor interaction triggers VDR heterodimerization with the retinoid X receptor (RXR) resulting in a complex that modulates transcription of target genes.

The ability of vitamin D signaling to combat breast pathogenesis has been suggested based on numerous lines of evidence. Polymorphisms in at least one domain of the receptor have been associated with modified breast cancer risk, particularly for women with a family history of the disease (71). A distinct

11 polymorphism appears to impact risk of developing bone metatases (72). Several

epidemiological studies have found an inverse correlation between breast cancer

mortality and exposure to the ultraviolet sunlight that produces vitamin D3

(73,74). Furthermore, a Norwegian study found that women receiving treatment for their breast cancer during seasons of high vitamin D3 exposure displayed

improved prognosis (75). The impact of dietary vitamin D has been illustrated in

laboratory mice. When fed a high-fat, Western-style diet, these mice exhibit

hyperproliferation of pancreatic, prostate and mammary epithelial cells; however,

addition of calcium and vitamin D suppresses this phenomenon (76). Speculation

of the potential value of VDR as a chemotherapeutic target is supported by the

fact that high expression of VDR has been found in a large number of breast cancers (77,78).

The presence of vitamin D receptor in the mouse mammary gland is

spatially and developmentally regulated. During puberty, when the ducts are elongating to fill the fat pad, VDR is detected in nearly all luminal epithelial cells

but is conspicuously absent from the rapidly proliferating cap cells of the terminal

end bud (79). In the virgin adult gland, only a small percentage of ductal epithelial

cells maintain expression of VDR (79); however, in the event of pregnancy,

expression of VDR increases dramatically within mammary epithelial cells and

remains elevated throughout lactation and early involution (80). Evidence for a

potential function of vitamin D signaling during lactation is provided by an

experiment that demonstrated enhanced calcium uptake by explanted mammary

glands from pregnant mice in response to 1,25-(OH)2D3 (81). Interestingly, 1α-

12 hydroxylase, the primarily renal enzyme that catalyzes the final step in 1,25-

(OH)2D3 synthesis, is also expressed in the mammary gland and upregulated

during pregnancy and lactation (80), allowing for the possibility of ligand

synthesis as well as signaling at this site.

Generation of a series of genetically modified mice has led to the

identification of a role for VDR signaling in mammary gland development,

function and tumorigenesis. Mice deficient in vitamin D3 exhibit impaired casein

production in the mammary gland, further suggesting that this secosteroid is

important in lactation (81,82). Mice lacking the vitamin D receptor (VDRKO) display failure to thrive due to hypocalcemia and impaired skeletal formation (83); however, these defects can be overcome by administration of supplemental dietary calcium, allowing for study of the intrinsic role for VDR in mammary gland development. Mammary glands from peripubertal VDRKO female mice demonstrate enhanced ductal elongation and increased secondary branching compared to age-matched wild-types. VDR deficient mammary glands also exhibit enhanced responsiveness to estrogen, progesterone and lactogenic hormones both in vivo and ex vivo (79), supporting the notion that VDR antagonizes these mammary mitogenic hormones. During pregnancy, the absence VDR begets more extensive side branching and accelerated lobuloalveogenesis, resulting in larger alveoli and increased milk production.

Initiation of apoptosis during the first stage of involution is significantly diminished and glandular regression is delayed in VDRKO animals by approximately two days compared to wild-type mice (80). A role for VDR in resistance to mammary

13 tumorigenesis has been elucidated in the MMTV-neu model. Loss of a single copy of VDR is sufficient to shorten tumor latency and increase tumor incidence in these mice (84).

Numerous studies have also demonstrated the ability of 1,25-(OH)2D3 to inhibit growth of human breast cancer cells in culture (85-89). This anti- proliferative effect is due to G1 arrest and is correlated with a decrease in cyclin

D1 and increases in the cell cycle inhibitors p21 and p27 (88), with a concomitant

decrease in Cdk2 activity (90). Treatment of MCF-7 human breast cancer cells

with 1,25-(OH)2D3, as well as several related analogues, also induces

downregulation of estrogen receptor gene expression and protein levels (91,92),

prompting speculation that the growth inhibitory effects of 1,25-(OH)2D3 are

merely due to attenuation of estrogen signaling. However, treatment of several

estrogen receptor negative breast cancer cell lines (85,93,94), as well as anti-

estrogen resistant MCF-7 cells (95), with VDR agonists results in growth

inhibition, demonstrating the presence of estrogen-independent vitamin D effects.

In addition to inhibiting growth, treatment of MCF-7 cells with 1,25-(OH)2D3 induces trademarks of apoptosis, such as cell shrinkage, chromatin condensation and DNA fragmentation (96), as well as re-orientation of phosphatidylserine (97).

The calcemic effects of 1,25-(OH)2D3 in a physiological setting limit its

usefulness in in vivo experimental systems, as well as clinically. To address this

conundrum, a number of 1,25-(OH)2D3 analogues that maintain the ability to

inhibit mammary epithelial cell growth while exerting a diminished effect on

calcium homeostasis have been developed and characterized (98,99); one such

14 entity is the 20-normal analogue EB1089 (Figure I-2), which is at least one order

of magnitude more potent than the native hormone at inhibiting growth of MCF-7

cells and is significantly less hypercalcemic than 1,25-(OH)2D3 in mice (100).

Systemic administration of EB1089 prevented expansion of established MCF-7

xenografts in nude mice, even inducing complete regression of a subset of

tumors (101). Mice harboring MCF-7 xenografts have also been treated with

EB1089, as well as other vitamin D3 analogues, in combination with tamoxifen

(102), paclitaxel (103), and ionizing radiation (104); in each case, adjuvant therapy was more effective than either individual component at inhibiting growth or inducing tumor regression. In a study involving cardiac injection of the

metastastic breast cancer cell line MDA-MB-231, EB1089 reduced the number of

bone metastases and the bone tumor burden within each animal (105).

Treatment with EB1089 also inhibited growth of established nitrosomethyl urea

(NMU)-induced mammary tumors in rats (106); however, to date, the prevention

and treatment efficacies of EB1089 have not been assessed in any of the various

transgenic mouse models of mammary cancer.

Phase I clinical trials using EB1089 have been performed on patients with

advanced breast, colorectal, and pancreatic cancers (107,108). Both studies

found that the drug was well tolerated with adverse events limited to dose-

dependent effects on calcium metabolism. No effect on disease was observed. A

phase II trial involving patients with inoperable hepatocellular carcinoma found

similar drug tolerance and saw significant tumor response in a subset of patients

15 (109). Clinical trials assessing chemopreventive potential or the benefit of combination therapy have not yet been completed.

Mouse Models of Hormonal Contributions to Breast Cancer

As described above, hormones have a duplicitous effect on breast cancer risk. Prolonged exposure to cycling hormones between menarche and menopause results in increased risk (10-12) and the elevated hormone levels associated with pregnancy cause a transient increase in risk of being diagnosed with this disease (14). On the other hand, pregnancy imparts long-term protection from breast cancer (15,16). The impact of hormones of breast cancer risk is also observed in a number of rodent models of mammary cancer and research done using these models has provided some insight into hormonal modulation of mammary pathology.

The importance of estrogen in human breast cancer progression is substantiated by the successes of therapies that impede its activities (43-49).

The relationship between estrogen signaling and mammary cancer has also been corroborated in rodent models. Prolonged treatment with estrogens causes mammary tumor formation in rats (110). Although this phenomenon rarely occurs in mice in the absence of infection with the mouse mammary tumor virus (111), manipulation of estrogen has been shown to alter mammary tumor latency in the

MMTV-neu model. Treatment with 17β-estradiol leads to reduced tumor latency while treatment with tamoxifen causes a significant reduction in tumor incidence

16 (112). Tamoxifen treatment is also able to reduce incidence of tumor formation in transplanted p53 mammary epithelia (113).

The role of prolactin in mammary tumor formation has also been extensively probed. Transgenic mice that ectopically express high levels of prolactin from the liver develop mammary carcinomas (37). Another model of prolactin-induced mammary tumorigenesis mimics endogenous production of this peptide hormone by the mammary gland, resulting in formation of adenocarcinomas and adenosquamous neoplasms (38). Pituitary isografts, which result in elevated circulating levels of prolactin and progesterone, increase tumor formation in response to carcinogen treatment and are also capable of inducing tumors in the absence of carcinogen (114).

All of the models described above involve treatment with exogenous hormones and/or carcinogens or entail overexpression of hormones from an ectopic tissue. Recently, a unique model of hormone-mediated mammary cancer was developed that overexpresses luteinizing hormone from the gonadotropes of the pituitary; these mice are described in detail below and have been extensively utilized in the studies presented in this dissertation.

The protective effect of pregnancy on a woman’s lifetime risk of breast cancer has been well documented (15,16). Similarly, parous rats (19) and mice

(115) demonstrate refractoriness to carcinogen-induced mammary tumorigenesis, which can be mimicked by short term treatment with pregnancy levels of estrogen and progesterone (116) or by treatment with human chorionic gonadotropin (117). One of the leading hypotheses of pregnancy-mediated

17 protection from mammary tumorigenesis is that persistent changes in molecular pathways take place upon exposure to hormones. In this regard, increased and sustained expression of nuclear p53 was observed in the mammary epithelial cells of rats that had been exposed to protective levels of estrogen and progesterone (118), suggesting that increased activity of this tumor suppressor protein may contribute to parity induced protection of the mammary gland.

p53 p53: The Guardian of the Genome

p53 is a transcription factor that regulates expression of genes involved in a number of processes, including cell cycle arrest, DNA repair, and apoptosis

(reviewed in (119)). It is an example of a classic tumor suppressor gene because it does not appear to be essential for normal development, but loss of its function is associated with increased risk of cancer in a multitude of tissues in both humans and mouse models (120-122). It is the most commonly mutated gene in human cancer (123) and loss of its function is highly correlated with genomic instability, chromosomal aberrations, and aneuploidy (124), earning it the appellation, “Guardian of the Genome” (125).

To facilitate rapid response to activating events, p53 is primarily regulated at the level of protein stability. In normal cells, p53 is associated with Mdm2, an

E3 ubiquitin ligase that mediates degradation of the p53 protein (126). Disruption of this interaction is essential for activation of the p53 pathway, which occurs in response to a number of stress signals. Many forms of DNA damage, including

18 double and single strand breaks; base alkylation; depurination; and oxidative

damage, as well as changes in cell cycle induced by overexpression of

oncogenes such as myc and ras, trigger p53 activation (reviewed in (127,128)).

Activation of the p53 protein is marked by the addition of multiple post- translational modifications (129), which cause increased protein stability and

changes in cellular localization and interaction abilities, ultimately resulting in

transcriptional regulation of a number of target genes. p53 regulates transcription

by binding to consensus p53 response elements, which may be located

upstream of the transcriptional start site or in the first or second intron (127).

p53 activation can induce senescence, cell cycle arrest, or apoptosis. The

outcome depends on the genes that are affected. G1 arrest in response to p53 is

largely due to upregulation of p21, which inhibits the activity of the cyclin E-cdk2

complex (130). On the other hand, upregulation of 14-3-3σ leads to arrest of the

cell cycle in G2 by sequestering CDC25C, the phosphatase responsible for

activation of cyclin B-cdk1, in the cytoplasm (131). Induction of apoptosis in

response to p53 activation is due to changes in a number of genes including Bax,

Noxa, PIG8, and PUMA, resulting in caspase activation and release of

cytochrome c from the mitochondria (reviewed in (119)). The ultimate response

to p53 depends on cellular context, with intrinsic and extrinsic factors tempering

or augmenting p53 induced signals. Subtle differences in expression and/or

localization of p53-associated proteins may impact the sensitivity of p53 to

activation, the types of post-translational modifications that occur on the p53

protein, and the ability of p53 to bind to and regulate target genes.

19 p53 Mutations in Cancer

p53 is thought to be mutated in 50% of human cancers (123).

Furthermore, many patients with Li-Fraumeni cancer susceptibility syndrome

carry germline mutations of p53 (132). Unlike most tumor suppressor genes,

which are often truncated or deleted, the p53 mutations observed are most

commonly missense mutations that result in single amino acid substitutions

(133). The majority of documented mutations are in the DNA binding domain,

rendering the p53 protein unable to bind to its consensus response element.

Because p53 normally binds to DNA as a tetramer, the mutated protein also has

the ability to exert a dominant negative effect on any wild-type protein that is

present; however, harboring a mutated form of p53 is not equivalent to losing the protein altogether. Cultured cells expressing mutant protein grow more quickly than cells lacking p53 altogether (134) and microarray experiments demonstrate that mutant p53 proteins acquire novel transcriptional activities while retaining a very small proportion of wild-type activity (134,135). One explanation for these

“gain of function” behaviors is the ability of mutant p53 proteins to interact with and inhibit p63 and p73, members of the p53 family that are unable to interact with wild-type p53 (136-139).

p53 in Breast Cancer

The fact that many patients with Li-Fraumeni syndrome that contain

germline mutations in p53 display an increase risk of breast cancer established a

role for p53 in the prevention of this disease (140); however, although 50% of

20 human cancers harbor mutations in p53, a comprehensive meta-analysis showed

that only 20% of breast cancers express mutant forms of this protein (141). p53

mutations are more commonly found in high grade than low grade ductal

carcinomas in situ, suggesting that acquisition of mutant p53 may not be a

common initiating event in breast tumorigenesis (142). Given the robust tumor

suppressor capabilities of the p53 signaling pathway, aspiring breast tumors

must develop a way to subvert p53 function without necessarily selecting for its

mutation. This could be accomplished by alterations in pathway members either

upstream or downstream of p53. A few examples of such alterations have been

reported in the literature. ATM, a kinase responsible for activating p53 in

response to DNA damage, is mutated or downregulated in a subset of breast

tumors (143). Another upstream activator of p53, Chk2, exhibits decreased

expression in some breast cancers (144) and is also mutated in selected patients

with Li-Fraumeni syndrome (145). p53 function may be decreased in the small portion of breast tumors that exhibit amplification of Mdm2 (146) or MdmX (147), both of which prevent binding of p53 to DNA. Alterations have also been found in

p53 target genes. Epigenetic silencing of 14-3-3σ, an enforcer of p53 growth

arrest, has been observed and mutations in PIG8, a mediator of p53-dependent

apoptosis, have been documented in early onset breast cancer (148,149). A

number of these observations are correlative and should be explored on a

functional level; furthermore, it is likely that additional mechanisms of bypassing

p53-mediated protection exist. Utilization of mouse models of tumorigenesis will

facilitate identification and functional characterization of these mechanisms.

21 p53 Mouse Models

Mice with targeted disruption of the p53 gene are viable and

developmentally normal (120,121); however, both knockout models that have

been generated display tumor formation with high penetrance and virtually all of

the mice lacking p53 are dead by 10 months of age. The most commonly

observed tumors in these mice are lymphomas and sarcomas; the aggressive nature of these tumors likely precludes tumor formation in other tissues that may be susceptible. To circumvent this barrier, p53 null mammary epithelia were transplanted into the fat pads of mammary glands of wild-type mice, resulting in

the formation mammary tumors with high penetrance (150). Transplanted p53

null epithelia also demonstrate increased tumor formation in response to

hormone stimulation with pituitary isografts alone or in combination with the

carcinogen DMBA (151). These tumors were all adenocarcinomas and

frequently aneuploid (151). Loss of p53 also accelerates mammary

tumorigenesis in mice that express mutant forms of BRCA1 (152,153) or

overexpress Wnt-1 in the mammary gland (154). Wnt-1 mammary tumors lacking

p53 exhibited increased incidence of aneuploidy and chromosomal amplifications

and deletions (154).

The role of p53 in mammary tumorigenesis has also been investigated

using transgenic mice that overexpress mutant forms of p53 selectively in the

mammary gland. A commonly observed missense mutation in human breast

cancer results in the substitution of histidine for arginine 175 (arginine 172 in

mouse); like many other p53 mutants, this protein is unable to bind DNA, but

22 enhances the tumorigenic potential of cells lacking p53, revealing complex “gain of function/loss of function” behavior (155). Transgenic mice were developed that overexpress mouse p53R172H under the direction of the whey acidic protein

(WAP) promoter (156). Although these mice develop few spontaneous mammary tumors, tumor formation induced by treatment with the carcinogen DMBA is significantly accelerated in WAP-p53R172H mice compared to wild-type controls

(157). Mammary tumor latencies were also decreased when the R172H transgene was co-expressed with erbB2/neu or a mutant form of IGF-1 in the mammary gland (156,158), establishing a cooperative role for p53 in these models. One common feature of these models is the increased rate of aneuploidy and genomic instability in tumors that develop in the presence of the mutant form of p53 (156-158). The presence of p53 mutations in human cancers is also correlated with increased levels of genomic instability (124).

Genomic Instability and Centrosome Amplification

Genomic instability, which describes the increased likelihood that a cell will acquire heritable genomic alterations, has long been recognized as a characteristic of tumors. Defects in DNA mismatch repair (MMR) or nucleotide excision repair (NER) can lead to microsatellite instability, a type of genomic instability which results in sequence mutations. Genomic instability may also become manifest at the level of the chromosome, resulting in inversions; rearrangements; deletions of partial chromosomes; or gains or losses of entire chromosomes (reviewed in (159)). A cell that contains an abnormal assemblage

23 of chromosomes is described as aneuploid, while chromosomal instability (CIN) refers to an increased rate of chromosomal alteration.

Genomic instability has also been correlated with the presence of centrosome amplification. Centrosomes are the major microtubule organizing structures of vertebrate cells. In a normal cell undergoing division, one centrosome forms each pole of the spindle and plays an important role in chromosome segregation. Centrosome amplification describes a condition in which 1) a cell contains supernumerary (≥3) centrosomes, 2) centrosomes are increased in size, and/or 3) centrosomes contain an increased number of centrioles. Centrosomal aberrations have been observed in a diverse array of human tumors, including those of the breast (160-163), pancreas (164), and prostate (165), as well as numerous cell lines derived from tumors (166). Many speculate that centrosome amplification causes multipolar spindle formation, which results in missegregation of chromosomes during anaphase and creates a source of chromosomal instability. In fact, multiple studies have determined a positive correlation between centrosome amplification and both aneuploidy and chromosomal instability (163,167). Evidence that centrosome amplification is an early event in tumorigenesis (rather than a byproduct) is provided by the fact that it has been identified in in situ carcinomas of the breast, cervix, and prostate

(163,168) as well as in pre-neoplastic cells in a rat model of mammary carcinogenesis (169). In vitro experiments have also indicated that centrosome amplification can precede and drive genomic instability as the human

24 papillomavirus (HPV)-16 E7 protein induces centrosome abnormalities prior to the appearance of genomic abnormalities (170).

The cause of centrosome amplification is unknown, although alterations in several genes have been associated with changes in centrosomes. Embryonic fibroblasts from p53 null mice demonstrate centrosome amplification along with a high degree of aneuploidy. Centrosome amplification was also correlated with mutant p53 status in squamous cell carcinomas of the head and neck (171,172); however, centrosome amplification has been documented in breast tumors that contain wild-type p53 (163). Abnormal centrosome number has also been observed in fibroblasts expressing mutant forms of BRCA1 (173) or BRCA2 (174) as well as in the mammary tumors of mice expressing HER2/neu (175) or a conditional mutant for BRCA1 (176). Aurora-A, a serine/threonine kinase that is upregulated in many human cancers (177-180), can cause centrosome amplification when overexpressed in cultured human breast cells (181). The mechanism(s) of centrosome amplification must be further elucidated to determine whether the processes that contribute to this phenomenon may be potential targets for therapeutics.

A Transgenic Mouse Model of Ovarian Hyperstimulation

The vital roles of hormones in the growth and development, as well as the pathology of the breast have been presented in this chapter. The distinct responsiveness of the mammary gland to hormones obliges recognition of endocrine contributions to signaling pathways and cellular processes that may

25 proceed hormone-independently in other tissues. In this regard, a unique model of hormone-induced mammary tumorigenesis was recently developed and has been utilized extensively in the studies presented in this dissertation.

Transgenic mice were created that overexpress the β-subunit of bovine luteinizing hormone (bLHβ; see Figure I-3); expression of this transgene is specifically targeted to the gonadotropes of the anterior pituitary by the bovine gonadotropin α-subunit promoter. The half-life of this hormone was increased 2-3 fold by addition of the C-terminal peptide (CTP) of the human chorionic gonadotropin β subunit (hCGβ) to the C-terminus of bLHβ (182). While male transgenics display normal levels of LH and maintain fertility, female counterparts have dramatically increased levels of luteinizing hormone as early as two weeks of age leading to an array of reproductive and physiological abnormalities.

As a result of LH-overexpression, female transgenic mice undergo precocious puberty and display increased circulating levels of estradiol (182), testosterone (183), and progesterone (182,184). These mice are infertile due to multiple reproductive defects. The chronic nature of LH exposure leads to anovulation. Although ovulation and pregnancy can be achieved by treatment with PMSG/hCG, embryos are resorbed at midgestation due to maternal hormone imbalance. A lack of uterine receptivity was also documented using transplanted embryos (185).

Ovaries from LH-overexpressing females become enlarged, cystic and hemorrhagic with the onset of puberty (185). These ovaries demonstrate accelerated folliculogenesis (184) and premature depletion of primordial and

26 primary follicles (186). Transgenic females maintained in the CF-1 strain develop

granulosa cell tumors by 5 months of age (182,183); however, F1 hybrid mice

created by breeding one generation into CD-1, C57BL/6 or SJL strains fail to develop such tumors despite similar hormonal profiles. These hybrid strains instead display a cystic phenotype with extensive luteinization (183). The frequency of granulosa cell tumor appearance in a backcross experiment suggests that three unlinked, recessive genes may be responsible for strain- dependent appearance of this pathology (183). Ovulatory surges simulated by cyclic hCG injection prevented formation of granulosa cell tumors in CF-1 transgenic mice and led to a luteoma phenotype indistinguishable from F1 hybrid

ovarian pathology (187).

Ovarian hyperstimulation caused by elevated LH leads to secondary

pathologies in other tissues including the adrenal gland, the pituitary and the

mammary gland. LH-overexpressing females display enlarged adrenal glands

and increased circulating levels of corticosterone which decline upon

ovariectomy (188). Interestingly, unlike wild-type counterparts, adrenal glands

from transgenic females express LH receptor and, when cultured, responded to

an LH agonist (188). The pituitaries of LH-overexpressing females are enlarged

by 16 weeks of age due to increased proliferation of Pit-1-positive cells

(lactotropes, somatotropes and thyrotropes). This results in increased circulating

levels of prolactin and growth hormone, but not thyroid stimulating hormone.

Functioning pituitary adenomas are fully penetrant by 10 months of age and foci

of all three Pit-1 positive cell lineages have been observed (189).

27 The mammary gland is also a well established target of ovarian hormones.

The effect of LH-induced ovarian hyperstimulation on mammary gland

development and tumorigenesis in this transgenic mouse model is the focus of

this dissertation.

Statement of Purpose

Hormones play a critical role in the initiation and progression of breast

cancer. Improvements in chemoprevention and chemotherapy will require further

characterization of mechanisms of hormone-induced mammary cancer as well as

the identification of potential therapeutic agents. The studies presented in this

dissertation describe a unique mouse model of hormone-induced mammary tumorigenesis (Chapter II). This model is utilized to investigate the therapeutic potential of a vitamin D receptor agonist (Chapter III) and identify p53-

independent centrosome amplification as a trait of hormone-induced mammary

tumors (Chapter IV).

28 Figure I-1

Mouse mammary gland development. At birth (pre-puberty), the mouse mammary gland consists of a rudimentary epithelial ductal structure (arrow) in the vicinity of the nipple. During puberty, terminal end buds (an example is marked by an asterisk) drive extension of the ducts through the fat pad. The adult mammary gland consists of an arborized ductal structure that fills the fat pad. In the event of pregnancy, the epithelial content of the mammary gland increases dramatically with secondary and tertiary branching of the ducts and the formation of lobulo-alveolar structures. During the functional state of lactation, the gland is dominated by milk-engorged lobulo-alveoli. These structures and a large number of the epithelial cells regress via apoptosis during involution. At the completion of involution the morphology of the mammary gland is similar to that of a virgin adult. (LN=lymph node)

29 LN

pre-puberty

LN *

puberty

LN

adult

LN LN

involution pregnancy

LN

lactation

30 Figure I-2

Chemical structures of 1, 25-dihydroxyvitamin D3 and EB1089. EB1089 (B) is an analog of 1, 25-dihydroxyvitamin D3 (A). Differences between the two structures are highlighted in red. EB1089 contains two double bonds and three additional carbons on its sidechain. Although both compounds bind to and activate the vitamin D receptor, EB1089 is significantly less hypercalcemic in animals than 1, 25-dihydroxyvitamin D3.

31 A. OH

HO OH

1, 25-dihydroxyvitamin D3

B. OH

HO OH

EB1089

32 Figure I-3

Generation of LH-overexpressing mice. The bovine gonadotropin subunit-α promoter was used to target expression of the cDNA for the β-subunit of bovine luteinizing hormone (LHβ) to the gonadotropes of the pituitary. The C-terminal peptide (CTP) of human chorionic gonadotropin-β fused to the 3’ end of the LHβ cDNA resulted in a 2-3-fold increase in the half-life of the LH heterodimer relative to heterodimer containing native LHβ (182). The construct also contains the polyadenylation signal from Simian Virus 40 and the first intron of the LHβ gene

(light green).

33 34 CHAPTER II

OVARIAN HYPERSTIMULATION BY LUTEINIZING HORMONE LEADS TO

MAMMARY GLAND HYPERPLASIA AND CANCER PREDISPOSITION IN

TRANSGENIC MICE∗

Introduction

Development of the mammary gland is a hormonally regulated process.

Estrogen, progesterone, and prolactin mediate progression of the rudimentary

ductal system present at birth to an extensive network following puberty and

finally into a differentiated milk producing organ during pregnancy and lactation

(53). These same hormones also regulate the occurrence and timing of reproductive events that influence a woman’s risk of developing breast cancer.

An increase in risk is observed for individuals who exhibit early age at menarche, late age of menopause, late age at first full term pregnancy or nulliparity (191-

193). Conversely, late age at menarche, surgically-induced menopause, and young age at first full term pregnancy appear to have protective effects

(11,15,194). Parity also confers protection against mammary cancer in rodents and this effect can be mimicked by treatment with human chorionic gonadotropin

(hCG) or combination therapy with estrogen and progesterone (195). However, the mechanisms that contribute to this phenomenon are unknown (19,196).

Reproductive hormones are also implicated in both initiation and progression of mammary tumors. Treatment with tamoxifen, which acts as an estrogen

∗ The data presented in this chapter has, in part, been published in (190).

35 antagonist in the breast, significantly reduces cancer recurrence and mortality in

patients with estrogen receptor positive breast tumors (43). In addition, tamoxifen

treatment can decrease the incidence of breast cancer in women at high risk of

developing the disease (44).

Deciphering the complex signals that hormones convey to the mammary

gland will lead to an increased understanding of the events surrounding tumor

formation and growth which will ultimately spawn novel therapeutic strategies.

Clearly, this will require development of appropriate model systems. While tissue

culture paradigms have utility for pathway dissection, it is not possible to

reproduce the complex interactions that occur between different hormones and the multiple cell types of the mammary gland in vivo. In addition, although human studies are most relevant, they are not conducive to experimental manipulation or control. Hence, the development of animal models is most advantageous, allowing for hormonal alterations within a physiological background. The ability to manipulate the murine genome has made the mouse a powerful tool for establishing the roles of various genes in mammary gland development and carcinogenesis. Furthermore, several previously generated transgenic mouse models develop mammary tumors that are morphologically similar to human

breast cancers (52).

To date, most mouse models of mammary cancer involve infection with

the mouse mammary tumor virus (MMTV) or employ the MMTV or whey acidic

protein (WAP) promoters to target expression of oncogenes or proto-oncogenes

to the mammary gland (197-200). While useful, these models do not address the

36 role of hormones in mammary tumor formation and progression. Other approaches, such as ovariectomy, hypophysectomy, pituitary isograft, and hormone injections have been used to determine the hormonal dependence of mammary tumor cells injected into mouse mammary glands (201-203). In addition, increased prolactin and progesterone levels induced by pituitary isografts enhance the ability of the carcinogen N-methyl-N-nitrosurea (MNU) to cause mammary tumors in mice (114). While each of these studies provides evidence for the role of hormones in tumorigenesis, they usually involve surgical manipulation, use of immortalized or transformed cell lines, treatment with super- physiological doses of hormone or addition of carcinogens, all of which make translation to human carcinogenic processes difficult. More physiological evidence that prolactin may be a key player in tumorigenesis is provided by the appearance of mammary tumors in a transgenic mouse model that expresses rat prolactin under control of the metallothionein promoter; however, this approach does not address the potentially cooperative roles of estrogen and progesterone and detaches prolactin from its conventional regulation and normal site of synthesis. Furthermore, although tumors were reported after 11 months of age, definitive tumor latency was not assessed (37).

While the breast is not typically considered an LH-responsive tissue, receptors for luteinizing hormone (LH)/hCG have been found in human breast cancers and breast cancer cell lines. In addition, hCG can inhibit proliferation of some breast cancer cell lines in culture (204,205). While these data are intriguing, regression of the mammary gland upon ovariectomy indicates that the

37 predominant effect of LH in vivo is indirect through stimulation of estrogen

production by the ovary (206). Furthermore, female estrogen receptor knockout

(αERKO) mice display significant defects in mammary gland development despite elevated levels of LH (207,208). Increased breast cancer risk associated with high levels of estrogen suggests that LH activity may actually play a role in the pathology of this disease. Indeed, in recent clinical trials, ablation of ovarian estrogen through chronic treatment of breast cancer patients with gonadotropin releasing hormone (GnRH) analogs alone and in combination with tamoxifen has led to tumor regression and increased survival (209,210,210).

Transgenic mice (LH-overexpressing) that overexpress luteinizing

hormone from the gonadotropes of the anterior pituitary have been previously

generated (182). While male mice maintain normal levels of LH, females display

levels of LH that are five to ten fold higher than control animals (182).

Consequently, estrogen (182), progesterone (185), and prolactin (188) are

elevated in female transgenic animals. These mice exhibit precocious puberty

(184), anovulation (185), infertility (182), and develop various ovarian pathologies in a strain dependent manner (183). Because estrogen, progesterone, and prolactin play imperative roles in mammary gland development and function, we investigated the impact of the LH-mediated alteration in the hormonal milieu on the mouse mammary gland. Herein we report that mammary glands of LH- overexpressing mice undergo precocious development and maintain an apparent

pregnancy-like state of hyperplasia that is dependent on ovarian hormones.

Furthermore, these mice develop spontaneous mammary tumors and display

38 accelerated carcinogen-induced mammary tumorigenesis when compared to wild-type controls. We propose that the LH-overexpressing mouse may be a useful model for exploring the hormonal regulation of pathways involved in transformation of the mammary epithelium in vivo.

Materials and Methods

Animals

Mice harboring a transgene that confers overexpression of luteinizing hormone have been described previously ((188); see Figure I-3). These mice have been maintained in the CF-1 genetic background. Studies described herein utilized either young mice from the CF-1 strain or mice from mixed genetic backgrounds containing either CF-1 and C3H; CF-1 and FVB; or CF-1 and

C57BL/6 components. No obvious strain-specific differences were observed in the morphology of the female transgenic mouse mammary glands. In addition, the ovaries of all young mice (less than 5 months of age) had numerous follicular cysts while all mice older than 5 months exhibited luteomas of pregnancy as previously described (182). Mice were housed in micro-isolator units with a 12 h light/dark cycle and given food and water ad libitum. Transgene genotyping was performed using PCR as previously described (188). All mouse studies were approved by the Institutional Animal Care and Use Committee at Case Western

Reserve University.

39 Hormone Measurements

Blood was collected from randomly cycling animals by cardiac puncture following asphyxiation with CO2. Samples were obtained at various times of day. After clotting, sera were prepared by centrifugation and collection of supernatant. Sera were stored at –20C until the time of assay. 17β-estradiol and progesterone levels were measured using RIA kits from Pantex (Santa Monica, CA) that have previously been validated for use with mouse sera (185,211). Limits of detection are 10 pg/ml and 0.2 ng/ml, respectively. Each sample was assayed in duplicate.

Prolactin serum levels were measured at the National Pituitary Hormone Center

(Harbor-UCLA Medical Center Torrance, CA, USA).

Morphological Analyses of Mammary Tissue

Whole mounts of female mammary glands were generated using inguinal

(#4) glands that had been immersed in Kahle’s fixative for 2-4 hours followed by overnight immersion in Carmine Alum stain (2% carmine, 5% aluminum potassium sulfate in water) at 4 C. Stained glands were dehydrated by graded ethanol washes, cleared with xylene and mounted on glass slides with Permount

(212).

For histological sections, thoracic (#2/3) or inguinal (#4) glands were removed and fixed in Kahle’s fixative or 4% paraformaldehyde overnight. Tissue was embedded in paraffin and 5 μm sections cut. Sections were deparaffinized in xylene, rehydrated with graded concentrations of ethanol and rinsed in phosphate buffered saline (PBS). Sections not being used for immunohistochemical analysis were stained with hematoxylin and eosin.

40 Immunohistochemistry

Antigen retrieval was performed by boiling samples in 10 mM citrate

buffer (213) for 10 minutes. Samples undergoing an immunoperoxidase reaction

were treated for 20 minutes with 3% hydrogen peroxide in methanol in order to

inactivate endogenous peroxidase activity. All incubations were carried out at

room temperature unless otherwise indicated.

For detection of progesterone receptor, samples were incubated with

goat serum for 10 minutes before treatment with an anti-PR antibody (DAKO

Corporation, #A0098, 1:100 dilution) for 1 hour. The rabbit IgG Vectastain Elite

ABC Kit (Vector Laboratories) was used for detection of the progesterone

receptor antibody. Concentrations of reagents recommended by the

manufacturer were used. After a PBS wash, samples were incubated for 30

minutes with the biotinylated goat anti-rabbit antibody (1:200 dilution), washed

again with PBS, then treated with the Vectastain ABC reagent for 30 minutes.

The secondary antibody was detected by incubation with 3,3’-diaminobenzidine

tetrachloride solution (Sigma, cat #D4168) for 2-5 minutes. Sections were

counterstained with methyl green, dehydrated, cleared, and subsequently

mounted with Permount. Immunohistochemistry for estrogen receptor was

performed at Baylor College of Medicine. Results shown are representative of

staining performed on at least five different animals with tumors.

For assessment of cells undergoing DNA synthesis, animals received an intraperitoneal injection of BrdU in dH2O (0.1 mg/gram body weight, Sigma) two hours before being killed. After deparaffinization, rehydration, and antigen

41 retrieval, samples were treated with 10% goat serum for 10 minutes. Following a

one-hour incubation with mouse anti-BrdU antibody (Becton Dickinson, 347580,

1:150 dilution) the samples were washed with PBS and treated with FITC-

conjugated goat anti-mouse antibody (Jackson ImmunoResearch, cat #115-095-

003, 1:300 dilution) for 1 hour (214). Samples were mounted with Vectashield

mounting media with propidium iodide diluted 1:4 in Vectashield mounting media

(Vector Laboratories). The number of BrdU-positive epithelial cells was counted in 6-10 fields from each animal (n=3 for both groups).

Apoptotic cells were detected using the TdT-FragEL DNA Fragmentation

Detection Kit (Oncogene Research Products). Samples were treated in

accordance with manufacturer’s instructions. In short, samples were

deparaffinized and hydrated, then permeabilized with proteinase K (2 mg/ml).

After inactivation of endogenous peroxidase activity and equilibration,

biotinylated dNTPs were applied. After incubation for 1.5 hours at 37 C the

reaction was terminated and samples were treated with a streptavidin-HRP

conjugate. Signal was detected by incubation with 3,3’-diaminobenzidine

tetrachloride for 2-10 minutes. Sections were counterstained with methyl green,

dehydrated, cleared, and subsequently mounted with Permount. The number of

apoptotic epithelial cells was counted in 10-12 fields per sample (n=3 for both

groups).

Gene Expression Profiling

Gene expression profiling was performed using Affymetrix Murine 11K

GeneChips which contain probes for 11,000 genes and expressed sequence

42 tags. Samples were prepared in concordance with recommended protocols from

Affymetrix. In short, total RNA was collected from mammary glands using TRIzol

(Invitrogen) and mRNA was subsequently isolated with an Oligotex mRNA Midi

Kit (Qiagen). Following generation of double stranded cDNA with the Superscript

Double-Stranded cDNA Synthesis Kit (Invitrogen), a High Yield RNA Transcript

Labeling Kit (Affymetrix) was employed to create biotinylated cRNA. After purification with the RNeasy Mini Kit (Qiagen), samples were hybridized to the chip and scanned with an argon-ion laser. Data were analyzed using Affymetrix software, including Microarray Suite, MicroDB, and Data Mining Tool.

Ovariectomy

Transgenic female mice were either ovariectomized or received a sham

surgery under avertin anesthesia at 4 months of age. The sham surgery involved

a flank incision, exposure of the ovary, reinsertion of the ovary, and placement of

a wound clip. Following a 21-day recovery period, mice were killed and thoracic

(#2/3) mammary glands removed and prepared for histological examination as

described above.

Carcinogen Treatments

Transgenic and wild-type female mice were given 7,12

dimethylbenz(a)nthracene (DMBA) (1 mg/100μl corn oil) or corn oil by oral

gavage at 5 weeks of age. A second dose of either DMBA or vehicle was given

1 week later. Mice were then palpated weekly for detection of developing

mammary tumors. Tumor latency was measured using Kaplan-Meier analysis.

Mice were killed upon the appearance of tumors or if the animals appeared ill.

43 Any mice that were obviously sick, but did not have overt mammary tumors, were censored from the study. These animals usually had either hydronephrosis (188) or pituitary adenomas (189) that were secondary to ovarian hyperstimulation caused by LH-overexpression.

Analysis of Milk Protein Gene Expression

Mammary glands were collected from wild-type female mice at 9, 12, 14, 16, and

18 days post-coitus or adult virgin transgenic and wild-type mice. Total RNA was isolated using Trizol (Gibco) according to manufacturer instructions. Northern blots containing 37.8 μg total RNA for each sample were probed with [32P] end- labeled, oligodeoxynucleotide probes complementary to the murine β-casein (5'

GTC TCT CTT GCA AGA GCA AGG GCC 3') or WAP (5' CAA CGC ATG GTA

CCG GTG TCA 3') genes (215). Blots were hybridized in Quikhyb solution

(Stratagene) for 2 h at 50 C followed by washing and exposure to x-ray film. To control for epithelial composition, blots were stripped and re-probed with a

[32P]dATP-labeled, randomly primed-probe to murine cytokeratin 8. Cytokeratin

8 template for probe production was generated by RT-PCR amplification of mammary gland RNA with the following primers: (forward, 5’

GTGCCCAGTACGAGGACATT 3’) and (reverse,

5’ GTGAGTCCCCCATAGGATGA 3’). Blots were hybridized in Quikhyb solution at 50 C for 30 minutes, washed, and exposed to x-ray film.

44 Results

Mammary gland development is accelerated in LH-overexpressing mice.

LH-overexpressing mice have elevated LH as early as post-natal day 14

(184), and, as a result, undergo precocious puberty compared to their wild-type

counterparts. Using age at vaginal opening as an indicator, LH-overexpressing

mice enter puberty on post-natal days 21-22, at least 5 days earlier than wild-

type littermates (183,184). To assess whether early ovarian activity in transgenic

mice impacts mammary gland morphogenesis, we examined glands from

transgenic and wild-type mice at 3 and 5 weeks of age. By 3 weeks of age,

accelerated development of the mammary gland was observed in LH- overexpressing mice compared to age-matched wild-type controls (Figure II-1A-

B). In transgenic animals, primary ducts were elongated, nearly reaching the centrally located lymph node, and numerous alveolar buds were present. In contrast, wild-type mice displayed age-appropriate pre-pubertal ductal development without significant accumulation of alveolar buds. The accelerated growth of the transgenic mammary gland was more apparent at 5 weeks of age

(Figure II-1C-D). While the ductal network in wild-type animals had progressed through only half of the mammary gland fat pad, the entire fat pad of the transgenic mouse was filled with ducts that displayed abundant alveoli and a loss of terminal end buds. Although all animals were virgins, the morphological pattern observed in the transgenic females was remarkably similar to that observed at mid to late pregnancy in normal mice (53).

45 The mammary glands in adult, virgin LH-overexpressing mice display a mid- pregnancy phenotype.

In addition to precocious maturation, mammary glands from adult transgenic mice had extensive epithelial hyperplasia. As shown in Figure II-2, histological examination of mammary glands from 3-month-old transgenic mice revealed considerable alveolar development with accumulation of lipid droplets.

Moreover, distended ducts were often observed with deposition of material within these ducts. Conversely, the mammary glands from age-matched wild-type littermates displayed a limited accumulation of epithelial cells. The accumulation of epithelial cells in transgenic animals was due to a clear increase in proliferation. A 12-fold increase in the number of S-phase epithelial cells was measured by incorporation of BrdU (p<0.01; Figure II-3A-E). Further supporting an elevation in proliferative rate, expression profiling of mammary glands of LH- overexpressing mice revealed a 2.2-fold increase in Ki-67 mRNA levels compared to their wild-type counterparts (average of data from two gene expression microarrays). In contrast to the increase in proliferative rate, a compensatory change in the number of apoptotic cells was not observed (Figure

II-3F).

The histological presentation of the mammary glands from adult, virgin transgenic mice supports the notion that these glands have undergone changes that normally accompany pregnancy. To further explore this possibility, we examined the expression pattern of molecular markers of pregnancy-induced differentiation within mammary glands of transgenic and wild-type mice. WAP, β-

46 casein, and WDNM1 are milk proteins whose corresponding mRNAs accumulate at specific stages of pregnancy (215). The earliest of these genes to be detected

in pregnant mammary glands is WDNM1, followed sequentially by β-casein and

WAP (215). To determine if expression of the milk protein genes is activated in

mammary glands of adult LH-overexpressing mice, we compared expression

levels in virgin transgenic and wild-type mice to those observed in glands

collected from wild-type mice at various stages of pregnancy (Figure II-4).

Expression of WAP and β-casein were nearly undetectable in mammary tissue collected from virgin, wild-type mice. In contrast, expression of these genes in the mammary glands of LH-overexpressing mice was very similar to that observed at day 14 of pregnancy in wild-type animals. A similar pattern was observed for WDNM1 (data not shown). Coupled with the morphological data

described above, these gene expression data support the hypothesis that

mammary glands in virgin transgenic animals with excessive LH have a

phenotype that simulates mid-pregnancy.

Mammary gland hyperplasia in LH-overexpressing mice is dependent upon ovarian input.

Serum levels of LH have previously been reported to be 5-10 fold elevated

in LH-overexpressing mice compared to wild-type littermates (184). In response

to high LH, increases in serum estrogen, progesterone, testosterone, and

prolactin were also observed in young animals (184,185,188). Because estrogen,

progesterone, and prolactin are known to impact mammary gland development,

47 we measured serum levels of these hormones in mice at several different ages to

identify the hormonal profile that supports the development of mammary

hyperplasia and tumors in transgenic mice (Table 1). Significant increases in estrogen were detected in LH-overexpressing mice as early as 5 weeks of age, consistent with the extensive hyperplasia observed at this time. By 10 weeks of age, progesterone levels were also dramatically increased. In addition, serum prolactin levels were elevated; however, this was only observed later in life and may be due to chronic elevation in circulating estrogens, which increase lactotrope secretion of prolactin (216). Although prolactin levels in 19-week-old transgenics were highly variable, ranging from 67 to 762 ng/ml, all animals examined at this age displayed significantly higher serum concentrations than age matched controls (5-7 ng/ml).

While the mammary gland is not normally considered a target of LH

action, the hyperplasia observed in LH-overexpressing mice could be due to

direct stimulation of the gland by LH. Supporting this notion are studies by

Russo and colleagues that revealed a protective effect of hCG on chemically-

induced mammary cancers in rats (217,218). Since hCG binds the same receptor

as LH (219), it is possible that this effect is mediated directly at the mammary

gland. In addition, LH receptors have recently been identified in mammary

epithelium (204,205,220). Thus, we sought to determine whether the mammary

gland hyperplasia observed in LH-overexpressing transgenic mice was due

solely to a direct action of LH on the mammary gland, or required ovarian

hyperstimulation caused by excess LH. We used an ovariectomy (OVX)

48 paradigm to assess the relative importance of the ovary to mammary hyperplasia

in LH-overexpressing mice. Although this approach eliminates ovarian factors,

LH levels would actually increase due to the loss of steroid negative feedback on

transgene expression (221). Transgenic mice were OVX or sham-operated and

mammary glands collected 21 days following surgery. As shown in figure 5,

mammary gland hyperplasia was persistent in sham-operated control mice. In

contrast, loss of ovarian input caused complete regression of the gland. These

data indicate that the ovary is an obligatory intermediate of LH action on the

mammary gland in LH-overexpressing mice.

LH-overexpressing mice are predisposed to mammary cancer.

Parity has a dual effect on breast cancer risk. Although pregnancy is

associated with long-term protection against breast cancer, a short-term

increased risk has been observed immediately following pregnancy (14,222).

With this in mind, we speculated that the LH-overexpressing mice might be

predisposed to development of mammary cancer. In addition, we observed

occasional spontaneous cancers in older transgenic mice, but not in age-

matched wild-type littermates. To directly assess whether the LH-overexpressing

mice were particularly susceptible to developing mammary cancers, we treated

virgin transgenic and wild-type controls with the mammary carcinogen,

7,12-dimethylbenz(a)anthracene (DMBA). Female mice were treated with DMBA

or corn oil at five and six weeks of age and tumor development was assessed by

weekly palpation (Figure II-6). Tumor phenotype was evaluated using the criteria

49 recommended by the Annapolis Pathology Panel (223). Transgenic mice that

received DMBA developed invasive mammary carcinomas with squamous

metaplasia (Figure II-7A) at a mean latency of 13.5 (± 3.8) weeks post-treatment.

All transgenic mice developed tumors by 20 weeks following the second DMBA

treatment. In contrast, wild-type mice treated with DMBA developed mammary

tumors at a much slower rate, with only 20% penetrance by 56 weeks following

carcinogen treatment. Lung metastases were also observed in most mice harboring DMBA-induced tumors. More importantly, transgenic animals treated with vehicle stochastically developed mammary tumors with 50% penetrance by

41 weeks following corn oil administration, while control wild-type animals failed to develop tumors throughout the entire course of the experiment. Most of the spontaneous tumors identified in LH-overexpressing mice were mammary intraepithelial neoplasias (MINs), exhibiting multiple layers of epithelial cells and atypical nuclear cytology while remaining within the confines of the basement membrane; both low and high-grade MINs were observed (Figure II-7B). Lesions were multifocal and histological examination indicates that they originated in the terminal ductal lobular unit (TDLU) of the mammary gland. In addition to the palpable tumors that were observed in older mice, non-palpable low grade MINs were identified in untreated, transgenic mice as young as 20 weeks of age

(Figure II-7C). The potential of these lesions to progress to malignancy was realized in a number of tumor-bearing animals that demonstrated invasive acinar or solid carcinomas with occasional lung metastases (Figure II-7D). Additionally, two animals developed spontaneous nipple adenomas. From these data, we

50 conclude that LH-overexpressing mice are predisposed to mammary

carcinogenesis when compared to their wild-type littermates.

Serum levels of estrogen, progesterone, and prolactin were measured in

tumor bearing transgenic mice and compared to age-matched wild-type control

animals (Table 1). LH-overexpressing mice with mammary tumors display

elevated serum concentrations of all of these hormones relative to wild-type

controls. Furthermore, serum levels of estrogen in tumor bearing mice are

significantly higher than concentrations observed in tumor-free transgenic

animals at any other age (p<0.01, Mann-Whitney U-test). Conversely, while

tumor bearing mice demonstrate higher concentrations of progesterone than

young transgenic animals, there is not a significant change relative to 20 week

old LH-overexpressing mice. However, of note is the extensive variation

observed in progesterone levels of tumor bearing animals, which ranged from

110 to 3000 ng/ml while age matched wild-type animals demonstrated

concentrations from 5 to 42 ng/ml. The most dramatic difference observed was in

prolactin levels of tumor bearing animals, which ranged from 1,056 to 16,340

ng/ml, whereas wild-type animals at the same age displayed concentrations of 23 to 41 ng/ml.

Spontaneous mammary tumors from LH-overexpressing largely lack expression

of progesterone and estrogen receptors.

Growth of mammary tumors can be classified as either hormone dependent or hormone independent. These assessments are based on response

51 to hormone treatment (111). In humans, approximately one-third of mammary

tumors are hormone dependent while the remaining two-thirds are independent.

Conversely, almost all mammary tumors that develop in mice are hormone

independent (111). One indication of the hormonal requirements for growth is the

presence of receptors for estrogen or progesterone. Analysis of human breast

tumors has revealed a significant correlation between expression of estrogen and

progesterone receptors (224,225). Furthermore, the presence of estrogen

receptor in human breast tumors usually indicates that the tumor will at least be

initially responsive to chemical ablation of ovarian hormones using treatments such as tamoxifen and/or GnRH analogs (226,227). Hence, determination of the receptor status can be important for predicting hormone dependency of tumors.

Immunohistochemistry for progesterone and estrogen receptors was performed on spontaneous mammary tumors from LH-overexpressing mice. While normal epithelial cells adjacent to tumors express both receptors, most tumor cells did not express detectable levels of either PR or ER (Figure II-8). This is consistent with the co-expression pattern for these two receptors often observed in humans

(224,225). Small patches of ER-positive and PR-positve cells were observed in

some tumors (<1% cells in a given tumor; data not shown).

Discussion

We have shown that persistent overexpression of LH from the pituitary of

transgenic mice leads to precocious mammary gland development and ovary-

dependent mammary hyperplasia. Hyperplasia is due to an increase in epithelial

52 cell proliferation by an order of magnitude compared to wild-type controls. In view of the fact that early puberty is a risk factor for breast cancer in humans

(191,228), it will be important to determine whether precocious puberty in LH- overexpressing mice significantly contributes to their acquisition of mammary tumors. Mammary glands of adult LH-overexpressing mice display morphological similarity to those of mice at mid-pregnancy and also express the milk protein genes β-casein, WAP, and WDNM1. The pregnancy phenotype is intriguing given the importance that pregnancy-related mammary gland changes play in the determination of breast cancer risk. Parity is known to impart a protective effect in humans as well as rodents (15,229,230). This phenomenon can be mimicked in rodents by treating with hCG, a placental hormone which binds to the LH receptor, or a combination of estrogen and progesterone; both paradigms render animals resistant to the tumorigenic effects of carcinogens (117,195). These observations have spurred the proposition of hCG treatment for women who plan to delay their first pregnancy (231). The susceptibility of LH-overexpressing mice to spontaneous tumor formation indicates that persistent exposure to this hormone may have detrimental effects. Therefore, the timing and duration of such treatment must be taken under serious consideration. An additional concern stems from the report that a transient increase in breast cancer detection is observed in women who have recently given birth; this is most likely due to the hormone rich environment stimulated by pregnancy that would favor growth of previously transformed cells (14). Accordingly, manipulation of the hormonal

53 milieu of a nulliparous woman may result in a short-term increase in breast

cancer risk.

The molecular mechanisms underlying parity-mediated protection from

breast cancer are unknown although several hypotheses have been suggested.

One possibility is that terminal differentiation of the mammary gland brought

about by pregnancy or hormone treatment results in the removal of cells that are

particularly susceptible to transformation (196). Alternatively, it has been

postulated that parous animals display an altered hormonal milieu that modifies

the biochemical profile of the mammary epithelia (19). Although the rat has been

the animal model of choice for investigating the impact of hormonal manipulation,

recent work has confirmed that parous mice demonstrate a resistance to

carcinogen-induced tumorigenesis, providing evidence that the two rodent

models are comparable (115). In contrast to the protection bestowed on

mammary glands by pregnancy or estrogen/progesterone treatment, the

pregnancy-like state of LH-overexpressing mice renders the mammary glands of

these mice vulnerable to cancer formation rather than imparting protection. This

discrepancy may stem from variations in the degree of hormone elevation or they

may be due to the chronic nature of hormone exposure in the LH-overexpressing mice compared to the transient increase in hormone levels that occurs during pregnancy. It is also important to note that, although the mammary glands of LH- overexpressing mice mimic pregnancy morphologically, they may not have achieved parity. Perhaps the alterations observed in these mice are more similar

to the changes caused by perphenazine (195), a dopamine antagonist that

54 induces mammary gland differentiation but does not offer protection from

carcinogens, than those induced by pregnancy or estrogen/progesterone

treatment. Interestingly, the hormone treatment paradigm that renders rats refractory to carcinogen-induced tumors triggers a 9-fold increase in estrogen

and only a 4-fold increase in progesterone (195). Alternatively, perphenazine

leads to increases in circulating prolactin and progesterone concentrations, but

does not significantly alter estrogen levels (195). Although serum measurements

cannot be directly compared, LH-overexpressing mice have only a modest

increase (< 3-fold) in estrogen levels compared to wild-type mice of the same

age, while progesterone levels are more than 20-fold higher in tumor free

transgenic animals and prolactin is nearly 40-fold higher in 20-week-old

transgenic mice compared to age-matched wild-type controls. However, while the

hormonal milieu of LH-overexpressing mice increases susceptibility to the

carcinogen DMBA, treatment with perphenazine does not appear to significantly

increase the rate of tumor formation in MNU treated rats. Also, the increased

mammary tumor susceptibility of LH-overexpressing mice may be due to the fact

that they never reach the fully differentiated state of lactation. However, while lactation appears to enhance the protective effect of pregnancy in humans (232-

235), rats that have undergone full term pregnancy but not lactation demonstrate resistance to carcinogen-induced tumorigenesis suggesting that lactation is not

required for hormone-mediated protection in rodents (236). Finally, unlike wild-

type mice that have achieved parity or received transient hormone treatment, the

mammary glands of LH-overexpressing mice never undergo an involution-like

55 event. The absence of tissue remodeling that is normally associated with this

process may contribute to the tumor susceptibility of these animals. Further

studies will be required to define the precise hormonal conditions that predispose

LH-overexpressing mice to cancer and, more importantly, to dissect the molecular mechanisms that govern both hormone-induced mammary tumorigenesis as well as protection from carcinogenic insults.

LH-overexpressing mice represent an autochthonous mammary tumor model in which the mice are susceptible to tumor induction by the carcinogen

DMBA and also develop spontaneous mammary tumors with a mean latency of

41 weeks. To date, in vivo studies of the hormonal components of breast cancer

have often depended on external manipulations such as subcutaneous hormone

pellets or pituitary isografts (114,195). Persistent treatment with exogenous

estrogen in this fashion leads to the formation of mammary tumors in rats;

however, this effect is rarely observed in mice (reviewed in (111)). In contrast,

continual administration of medroxyprogesterone acetate causes mammary

adenocarcinomas in BALB/c virgin female mice (237). This effect of progesterone

appears to be strain specific. Lastly, transgenic mice expressing prolactin under

the control of the metallothionein promoter have been constructed, but these

mice express the transgene primarily in the liver, a tissue that does not normally

secrete prolactin (37). In contrast to these approaches, LH-overexpressing mice

provide a model for studying the pathology that occurs as a result of chronic

hyperstimulation of the intact pituitary-gonadal axis. Given the dramatic increase

in serum prolactin levels observed in LH-overexpressing mice with mammary

56 tumors (Table 1), it is conceivable that this hormone plays a significant role in the

tumorigenic process. However, several lines of evidence suggest that prolactin

may require additional inputs to cause tumor formation. In particular, the

prolactin-overexpressing mice described above develop mammary tumors, but

with an extended latency compared to that observed in the LH-overexpressing mice (37). Furthermore, although pituitary isografts, which cause an increase in prolactin levels and a subsequent increase in progesterone production by the ovary, enhance susceptibility of the mouse mammary gland to carcinogen-induced tumorigenesis, they rarely lead to the formation of spontaneous mammary tumors (114,238). One possibility is that prolactin functions to promote growth of transformed cells, leading to the formation of a discernable tumor. In support of this notion, many mice with mammary tumors were also found to contain functioning pituitary prolactinomas (John Nilson, personal communication). Extensive studies utilizing dopamine agonists, such as bromocriptine (216), or crosses with prolactin deficient mice (61) will be required to assess the specific role that prolactin plays in tumorigenesis in this model.

Spontaneous mammary tumors of LH-overexpressing mice do not display

detectable expression of progesterone receptor and rarely express estrogen

receptor. This is consistent with the observation that most mouse mammary

tumors lack expression of steroid hormone receptors (111) and suggests that the

mammary tumors of LH-overexpressing mice are hormone-independent.

Consistent with this notion, experiments performed on a limited number of tumor-

57 bearing mice indicate that mammary tumor growth is sustained upon removal of

the ovary (data not shown).

Similarities observed between mammary glands of transgenic and

pregnant mice, including overall morphology and expression of milk proteins,

prompted us to assess the molecular resemblance of these tissues on a global

scale. Gene expression profiling revealed many genes that behave similarly in

transgenic and pregnant glands when compared to virgin wild-type glands (data

not shown). Presumably, these genes are regulated, directly or indirectly, by the

altered hormonal milieu present in both transgenic and pregnant animals. On the

other hand, there are also a number of genes that are differentially expressed

between these two samples. Differentially expressed genes will be of significant

interest given the fact that the LH-overexpressing mice display an increased susceptibility to carcinogen-induced mammary cancer, while pregnancy actually imparts protection (239). Expression profiling also indicates that the dramatic

increase in the proliferative index of the mammary glands of LH-overexpressing

mice (Fig II-3, data not shown) can be accounted for, at least in part, by changes

in several cell cycle regulatory genes. Messenger RNA levels of cyclins A, B1

and B2, as well as cyclin dependent kinase-1 (cdk1) are increased in mammary

glands from LH-overexpressing animals compared to wild-type age-matched

controls, while expression of an inhibitor of cell cycle progression, p18, is

decreased (data not shown). Further assessment of the significance of these

changes is currently underway.

58 In summary, we have reported that the mammary glands of mice

overexpressing LH undergo precocious development and acquire spontaneous

tumors in the absence of forced proto-oncogene overexpression. Molecular

profiling of the mammary glands of these mice throughout development using

global approaches such as gene chip expression analyses should provide further

insight into the mechanisms of hormone-mediated hyperplasia and tumorigenesis. These mice may also be of use for testing prospective hormonally-relevant anti-cancer drugs. Revealing signaling pathways that are

important in hormone-induced tumorigenesis may also lead to the identification of

potential therapeutic targets and strategies for treatment of hormone-induced

human breast cancers.

59 Table II-1

Serum levels of estrogen, progesterone, and prolactin are increased in LH- overexpressing animals.

17β-estradiol progesterone prolactin

(pg/ml) (ng/ml) (ng/ml) Age (wks) x ± s.d. n x ± s.d. n x ± s.d. n 5 WT 10.0 ± 8.2 4 18.6 ± 19.7 4 11.5 ± 4.7 4 CTP 75.3 ± 40.5* 3 31.9 ± 26.4** 3 31.3 ± 40.0 4 10 WT 44.7 ± 32.0 9 16.5 ± 17.5 9 21.4 ± 33.4 5 CTP 71.5 ± 24.5 6 433.4 ± 385.5** 6 56.2 ± 39.2 5 20 WT 58.0 ± 18.6 5 15.9 ± 5.6 5 6.33 ± 1.2 3 CTP 148.0 ± 74.1** 8 316.0 ± 163.1** 8 237.8 ± 260.1* 6 41 WT 29.0 ± 7.3 4 17.5 ± 17.2 4 31.5 ± 9.8 4 ≥41 (w/ ** 4326.3 ± CTP 280.6 ± 177.8**# 8 603.1 ± 986.8 8 8 tumors) 5160.4** Age matched samples were compared using the Mann-Whitney U-test. * indicates P<0.05 when compared to age-matched wild-type control ** indicates P<0.01 when compared to age-matched wild-type control

60 Figure II-1

Mammary gland development is accelerated in LH-overexpressing mice.

Whole mounts of inguinal mammary glands (#4) were prepared from wild-type (A and C) or LH-overexpressing (B and D) mice that were either 3 (A and B) or 5 (C and D) weeks of age. “L” indicates the mammary gland associated lymph node.

61 wild type LH A B

3 wk L

L

C D

L L 5 wk

62 Figure II-2

The mammary glands in adult, virgin LH-overexpressing mice display a mid-pregnancy morphology. Histological sections of thoracic mammary glands

(#2/3) from 4-month-old wild-type (WT) and LH-overexpressing (LH) mice (100X magnification). Sections were stained with hematoxylin and eosin. Asterisks indicate collections of epithelial cells.

63 * * * WTWT * LHLH

64 Figure II-3

The mammary glands of adult, virgin LH-overexpressing mice display an increase in epithelial cell proliferation, but no change in apoptosis. 12-week old wild-type (A and C) and transgenic (B and D) animals were injected with

BrdU two hours before being killed, and cells in S-phase of the cell cycle were detected with an anti-BrdU antibody followed by a FITC-conjugated secondary antibody (green, A and B). Sections were mounted with media containing propidium iodide (red, C and D), allowing for visualization of all nuclei (propidium iodide staining did not facilitate reliable assessment of apoptosis). (E)

Quantification of S-phase epithelial cells was done by calculating the percentage of BrdU-positive cells in 6-10 fields from each animal (n=3 for both groups). The number of BrdU positive cells was increased in the LH-overexpressing mice compared to virgin wild-type mice (P<0.01, Mann-Whitney Mean-Rank U-test).

(F) Apoptotic cells were detected using a FragEL DNA Fragmentation Assay.

The percentage of apoptotic cells was calculated for 10-12 fields per animal (n=3 for both groups). Values represent the mean + SD. All pictures are magnified

200X.

65 wild type LH A B BrdU

C D PI

2.5 E 12.0 F 10.0 2.0 8.0 1.5 6.0 1.0 4.0 2.0 0.5

0.0 cells % apoptotic 0.0 WT LH WT LH % BrdU positive cells

66 Figure II-4

The mammary glands in adult, virgin LH-overexpressing mice display a mid-pregnancy phenotype at the molecular level. Northern blots were performed with total RNA isolated from the mammary glands of virgin 4-month- old wild-type (V) and transgenic (LH) littermates. For comparison, RNA was also collected from wild-type animals at days 9, 12, 14, 16, and 18 of pregnancy and day 1 of lactation (L). End-labeled oligonucleotide probes against WAP and β- casein were used simultaneously. To control for epithelial cell content, the blot was stripped and re-analyzed with a cytokeratin 8-specific probe.

67 wild type day of pregnancy VLV 912141618LLH H

β-Casein

WAP

Cyto- keratin 8

68 Figure II-5

Mammary gland hyperplasia in LH-overexpressing mice is dependent on ovarian input. Four-month old transgenic mice were either ovariectomized

(OVX) or sham operated (intact). Three weeks following the surgery, thoracic

(#2/3) glands were collected, immersed in Kahle’s fixative, sectioned, and stained with hematoxylin and eosin (n=3 for each group; magnification = 100X).

69 Intact

OVX

70 Figure II-6

LH-overexpressing mice are predisposed to mammary cancer. Transgenic and wild-type female littermates were treated with either DMBA or corn oil at 5 and 6 weeks of age (indicated by gray line). Mammary tumor development was assessed by weekly palpation. The following numbers of animals were used: wild-type, DMBA treated = 15 (black circles); transgenic (Tg), DMBA treated = 13

(black squares); wild-type, vehicle = 18 (open circles); transgenic, vehicle treated

= 15 (open squares). Censored events (removal of an animal from the experiment because of death or sickness) are marked with an “X”. Spontaneous tumor latency in transgenic mice was found to be 41 weeks using Kaplan-Meier survival analysis.

71 100

80

60 x x 40 Tg-DMBA WT-DMBA xx 20 Tg-oil % Tumor Free WT-oil x Censored 0 0102030405060 Age (weeks)

72 Figure II-7

DMBA-induced and spontaneous mammary cancers in LH-overexpressing

mice. Hematoxylin and eosin stained sections of mammary tumors from LH- overexpressing mice (magnification = 100X). (A) Invasive mammary carcinoma with squamous metaplasia of transgenic mouse treated with DMBA. (B)

Spontaneous infiltrating mammary adenocarcinoma from an LH-overexpressing mouse treated with corn oil. (C) A non-palpable, in situ carcinoma identified in a

23-week-old LH-overexpressing mouse. Arrow indicates adjacent normal mammary epithelial tissue. (D) A lung metastasis identified in a transgenic female with a spontaneous mammary adenocarcinoma. The “L” indicates normal lung tissue and the “T” is indicative of the mammary cancer metastasis. All tissues were fixed with Kahle’s fixative.

73 AA BB

CC DD LL

TT

74 Figure II-8

Spontaneous mammary tumors from LH-overexpressing mice largely lack

expression of progesterone and estrogen receptors. Surrounding normal mammary epithelial cells (arrowhead) express progesterone receptor (A) or estrogen receptor (B), but the adjacent spontaneous adenocarcinoma does not

(asterisk in both panels). Sections are counterstained with methyl green

(magnification = 100X). Dark staining within the tumor tissue is non-nuclear remnants of the methyl green counterstain.

75 A.

*

B.

*

76 CHAPTER III

EB1089, A VITAMIN D RECEPTOR AGONIST, REDUCES PROLIFERATION

AND DECREASES TUMOR GROWTH RATE IN A MOUSE MODEL OF

HORMONE-INDUCED MAMMARY CANCER∗

Introduction

Breast cancer is the most common cancer in women in the United States, who exhibit a 1 in 8 chance of being diagnosed with this disease over their lifetime (1). Although currently available treatment paradigms, which include hormone modulators such as tamoxifen, are often effective, many cancers are refractory to therapy and others recur following initial responsiveness. The identification of novel therapeutic targets and subsequent development of effective modulators of these targets will assist in reducing the incidence and mortality of this disease.

The vitamin D endocrine system represents one emerging target of therapeutic interest in breast tissue. Vitamin D3 is acquired through the diet or

produced in the skin upon exposure to ultraviolet light. 1, 25-dihydroxyvitamin D3

[1,25-(OH)2D3], the bioactive metabolite of vitamin D3, binds to the vitamin D

receptor (VDR), a member of the nuclear receptor superfamily, and regulates

transcription of select target genes (reviewed in (70)). Several lines of

epidemiological evidence suggest that activation of VDR may impart protection

against breast cancer progression. Risk of fatal breast cancer is inversely

∗ The data presented in this chapter has been published in (240).

77 proportional to the intensity of local sunlight (74) and post-treatment disease

prognosis is improved if chemotherapy is administered during seasons of high

levels of vitamin D3 exposure (75). Furthermore, high dietary intake of vitamin D

has been correlated with decreased breast cancer risk in women (241,242).

Immunohistochemical analysis has revealed that the vitamin D receptor is

elevated in human breast cancers compared to normal breast tissue (78). VDR is

present in more than 75% of breast tumors and is associated with a diverse set of molecular phenotypes, evoking the possibility that VDR may be a broad spectrum therapeutic target (243).

Studies of the endogenous role of vitamin D receptor in the mammary

gland have been performed using genetically engineered mice. Mice deficient in

VDR have revealed that it is able to limit hormone-driven mammary gland

development. Compared to age-matched wild-type littermates, peripubertal

female VDR-/- mice exhibit enhanced ductual elongation and increased

secondary branching (79), events ascribed to estrogen and progesterone activity,

respectively (53). Furthermore, mammary glands from mice lacking VDR display

enhanced responsiveness to exogenous estrogen, progesterone, and lactogenic

hormones both in vivo and ex vivo (79). The importance of VDR in resistance to

tumorigenesis has been verified by a recent report that loss of a single copy of

VDR is sufficient to shorten tumor latency and increase tumor incidence in the

MMTV-neu mouse model of mammary cancer (84).

In addition to mouse models, numerous studies have demonstrated the

ability of 1,25-(OH)2D3 to inhibit growth of human breast cancer cells in culture

78 (85-88). This antiproliferative effect is due to G1 arrest and is correlated with a

decrease in cyclin D1, increases in the cell cycle inhibitors p21 and p27 (88), and a decrease in Cdk2 activity (90). In MCF-7 cells, 1,25-(OH)2D3 treatment is also

associated with downregulation of estrogen receptor (ER) (91,92); however, VDR

agonists also deter growth of several estrogen receptor negative breast cancer

cell lines, demonstrating the existence of estrogen receptor-independent

mechanisms of action (85,93,94). In addition to growth inhibition, cells exposed

to 1,25-(OH)2D3 also display trademarks of apoptosis, such as cell shrinkage,

chromatin condensation and DNA fragmentation (96).

Initially, the hypercalcemic effects of 1,25-(OH)2D3 limited in vivo scrutiny of the therapeutic potential of VDR activation, but the development and characterization of several 1,25-(OH)2D3 analogs that maintain the ability to

inhibit mammary epithelial cell growth while exerting a diminished effect on

calcium homeostasis has facilitated investigation in experimental animal models

(98,99). One such compound, EB1089, inhibits growth of MCF-7 cells with a

potency at least one order of magnitude greater than that exhibited by the native

hormone. Most importantly, this agent is significantly less hypercalcemic than

1,25-(OH)2D3 in mice (100). Systemic administration of EB1089 prevented

expansion of established MCF-7 xenografts in nude mice, even inducing

regression of a subset of tumors (101). Treatment with EB1089 also inhibited

growth of established nitrosomethyl urea (NMU)-induced mammary tumors in

rats (106); however, to date, prevention and treatment efficacies of 1,25-(OH)2D3 analogs have not been assessed in a spontaneous model of mammary cancer.

79 In this regard, transgenic mice that overexpress luteinizing hormone (LH)

provide a unique model of hormone-induced mammary cancer (182,190).

Elevated LH causes ovarian hyperstimulation and leads to increased levels of

several mammogenic hormones, including estradiol, progesterone and prolactin.

Exposure to this altered hormonal milieu results in mammary gland hyperplasia

and the formation of spontaneous mammary tumors (190), making the LH-

overexpressing mouse one of the few experimental models that reflects the

critical contribution of reproductive hormones to human breast cancer. This

contribution has been long acknowledged and is evidenced by numerous

epidemiological studies that correlate the timing and occurrence of several

hormonally regulated events, such as menarche, pregnancy and menopause,

with risk of breast cancer diagnosis (10,13-16). Furthermore, women treated

with tamoxifen, a selective estrogen receptor modulator that antagonizes ER in

the breast, display a significantly decreased risk of developing invasive breast

cancer (244). Hence, identifying approaches to combat the pathological effects of

hormones on the mammary gland, both preventatively and therapeutically,

should contribute to reductions in disease incidence and mortality and models

such as the LH-overexpressing mouse should facilitate this process.

The studies presented herein provide evidence that treatment with a

vitamin D3 analog, EB1089, has the ability to limit hormone-induced proliferation of endogenous mammary epithelial cells and reduce the growth rate of a subset of spontaneous mammary tumors in vivo.

80 Materials and Methods

Animals

Mice were housed in microisolator plus units with a 12 hour light/dark

cycle and given food and water ad libitum. During treatment with EB1089 or

vehicle, all mice were maintained on a 0.1% low-calcium diet (BioServe,

Frenchtown, NJ) to avoid the potential development of hypercalcemia. All mouse studies were approved by the Institutional Animal Care and Use Committee at

Case Western Reserve University.

Gene Expression Profiling

Affymetrix U74Av2 gene chips were probed with biotinylated cRNA

generated from mammary glands of wild-type and transgenic mice at the

specified ages. Double stranded cDNA was reverse transcribed from 10 μg of total RNA using the Superscript Double Stranded cDNA Synthesis kit (Invitrogen,

Carlsbad, CA; cat# 11917-010) and an oligo-dT primer coupled to the T7 RNA polymerase promoter. Subsequently, biotinylated cRNA was synthesized using the Enzo BioArray High Yield RNA Transcript Labeling Kit (Affymetrix, Santa

Clara, CA; cat# 900182). Data was analyzed using Affmetrix MicroArray Suite 5.0

(MAS 5.0). VDR was identified as part of a list of genes that was at least two-fold changed in at least one transgenic sample relative to the appropriate wild-type sample.

Immunohistochemistry

Thoracic mammary glands were removed and fixed in 4%

paraformaldehyde. Five micron sections were deparaffinized in xylene,

81 rehydrated with graded concentrations of ethanol and rinsed in phosphate buffered saline (PBS). Unless otherwise noted, all incubations were performed at room temperature and washes done in PBS. The primary antibodies utilized were the 9A7 rat anti-chicken VDR antibody (Affinity Bioreagents, Golden, CO; cat# MA1-710) and mouse anti-BrdU (Becton Dickinson Immunocytometry

Systems, San Jose, CA, cat# 347580). Vector ABC kits containing biotinylated anti-rat and anti-mouse secondary antibodies were used (Vector Laboratories,

Burlingame, CA; cat# PK-4004, PK-4002) and diamino-benzidine (DAB) was employed for detection. Nuclei were counterstained with hematoxylin QS (Vector

Laboratories; cat# H-3404).

For VDR, antigen retrieval was performed in 2N HCl for 30 minutes at

37°C. Non-specific binding was blocked by pre-incubation in 1.5% normal rabbit serum in PBS for 30 minutes. The antibody for VDR was used at a concentration of 5 μg/ml with 1.5% normal goat serum in PBS and incubated overnight at 4°C.

After washing, biotinylated anti-rat secondary antibody (1:200) was incubated for

1 hour at 37°C. The percentage of VDR positive cells was quantitated from 10-20 individual sections per animal (n=5 WT, 6 LH) and data analyzed with a Student’s t-test.

Sections to be assayed for BrdU incorporation were first boiled for 10 minutes in 10 mM citric acid, pH 6.0 to expose antigenic sites. After one hour incubation with the primary antibody (1:150) and subsequent washes, sections were exposed to secondary antibody (1:300) for 1 hour. The percentage of BrdU positive nuclei was determined in 10-20 sections (1100-1200 total cells) from

82 each animal and data was analyzed by a one-tailed Student’s t-test (n=4 WT, 4

LH). Epithelial cells comprising both ducts and terminal end buds were included in the analysis.

EB1089 treatment of mice during development

EB1089 was obtained from Leo Pharmaceutical Products (Ballerup,

Denmark). LH-overexpressing female mice heterozygous for the VDR null allele

were generated in a mixed genetic background of CF1, FVB and C57BL/6J.

Female mice were injected subcutaneously with 50 μl of sesame oil or with 50 μl

of freshly prepared EB1089 (0.027 μg per animal) diluted in sesame oil once per

day from 3 to 5 weeks of age. Mice were weighed daily to monitor for any global

toxicity. Weight gains of mice in vehicle and EB1089 treated groups were

indistinguishable. Twenty four hours following the last injection, the mice were

injected intraperitoneally with 0.1 mg of 5-bromo-2-deoxyuridine (BrdU; Sigma,

cat# B5002) per gram body weight. Two hours later, the mice were asphyxiated

with carbon dioxide and their mammary glands processed for

immunohistochemistry.

EB1089 treatment of tumor bearing mice

Transgenic mice were palpated weekly. Magnetic resonance imaging

(MRI) was performed upon detection of a mammary tumor and again 2 weeks

later in order to determine a basal growth rate for each tumor. Treatment was

then initiated. Mice were given intraperitoneal injections of vehicle or 0.05 μg

EB1089 in PBS every 48 hours. After 14 days of treatment, tumors were

reevaluated using MRI. Percent change in growth rate (GR) was calculated as

83 follows: % change GR = ((GRpost - GRb)/GRb)*100. GRb (basal growth rate) =

change in tumor volume over 14 days prior to treatment initiation. GRpost (post- treatment growth rate) = change in tumor volume from Day 0 to Day 14 of treatment. Hence, a percent change in growth rate between -100 and 0 reflects a decreased tumor growth rate, while a percent change in growth rate below -100 indicates tumor regression.

Magnetic Resonance Imaging

Mice were anesthetized under 2.0% isoflurane gas in oxygen for 5 min and transferred to a clinical 1.5T MRI scanner (Siemens Sonata). A nosecone was used to deliver 1-2% isoflurane in oxygen to the animals during the MRI scans. The sedated animals were placed into a cylindrical, phased-array mouse coil (ID = 33mm) developed in-house to maximize the signal-to-noise ratio for the high resolution murine images. To visualize potentially small mammary tumors, the mice were scanned with a high resolution (270um x 270um), T1-weighted spin echo acquisition (TR/TE = 700/13ms, slice thickness = 1mm, number of averages = 6). The slices were acquired with zero gap in between to ensure an accurate tumor volume measurement. The T1-weighted acquisition was selected to provide good contrast between the normally bright, lipid-rich mammary glands and the darker tumors. The area of the tumors was measured in each image

(slice) by manually tracing along the edge of the tumor. The total tumor volume was calculated by multiplying the slice-by-slice tumor areas by the slice thickness

(1 mm) and summing the individual slice volumes.

84 Results

Mammary Glands of LH-overexpressing Mice Demonstrate Elevated Expression

of Vitamin D Receptor.

In order to identify the molecular profile associated with the development

of hormone-induced mammary gland hyperplasia and tumorigenesis, Affymetrix

gene expression profiling was performed on the mammary glands of LH-

overexpressing mice and wild-type littermates at 5, 8, 12, and 19 weeks of age, as well as on spontaneous mammary tumors and age-matched controls.

Included among the genes that are differentially expressed between wild-type

and transgenic mammary glands is the vitamin D receptor (Figure III-1A). While

the signal associated with the VDR transcript is below the limit of detection in most of the wild-type samples, mammary glands from 8, 12, and 19-week old LH- overexpressing mice display robust expression, with relative signal intensities between 6- and 17-fold higher than those of age-matched controls. Increased expression of VDR is maintained in tumors, with tumors having two-fold more

VDR mRNA compared to age-matched wild-type mice.

To determine whether the observed expression change also occurs at the

protein level, we assessed VDR expression immunohistochemically. As

expected, nuclear VDR staining is present in a relatively small subset of

mammary epithelial cells in adult virgin wild-type mice (79) (Figure III-1B);

however, the hyperplastic mammary glands of adult LH-overexpressing mice

display a 2.3-fold higher proportion of VDR-positive epithelial cells (Figure III-1C

85 and E; p<0.01). Furthermore, VDR is detected in nearly every cell of the hormone-induced mammary tumors (Figure III-1D).

EB1089 Decreases Mammary Epithelial Cell Proliferation in LH-overexpressing

Mice.

The presence of high levels of vitamin D receptor in the mammary glands of LH-overexpressing mice led us to speculate that the growth inhibitory potential of this receptor could be exploited to limit the pathological effects of the altered hormonal milieu. In order to determine whether activation of VDR has the ability to inhibit the proliferative effects of the elevated mammogenic hormones in these mice, we implemented a two week treatment of peripubertal transgenic mice with

EB1089, a 1,25-(OH)2D3 analog with less in vivo hypercalcemic activity than the native hormone. BrdU incorporation was used to assess epithelial cell proliferation. Exposure to EB1089 reduced the number of proliferating mammary epithelial cells by thirty-five percent compared to vehicle treated controls (Figure

III-2; p<0.05).

EB1089 Displays Anti-tumorigenic Activity in a Subset of Hormone-Induced

Mammary Tumors.

Evidence for EB1089 efficacy in the hyperproliferative endogenous mammary gland prompted us to investigate the influence of this compound on established hormone-induced mammary tumors of LH-overexpressing mice.

Mice were palpated weekly for the presence of tumors. Upon tumor detection,

86 mice were subjected to magnetic resonance imaging (MRI); the resulting images

were used to determine tumor volume (see Materials and Methods). As inter-

mouse variability in tumor growth rate is observed in this model, each tumor was

used as its own control. The change in tumor volume over the two weeks

preceding initiation of treatment established the basal growth rate of each tumor.

Subsequent to the second measurement of the tumor, treatment with vehicle or

EB1089 was begun. After two weeks of treatment, a third MR image was

generated and changes in tumor volume were calculated. The tumor growth rate

actually increased in all but one of the vehicle treated animals (Figure III-3A). On

the other hand, half of the mice that were treated with EB1089 displayed a

decreased rate of tumor growth. Although variability in tumor response precluded

statistical significance, it is remarkable that EB1089 actually caused regression of

two tumors to less than half of their pre-treatment volume. Representative MR

images of a mammary tumor that regressed in response to EB1089 are shown in

Figures III-3B (pre-treatment) and 3C (post-treatment).

Discussion

Extensive epidemiological and in vitro evidence supports a role for vitamin D signaling in limiting growth of mammary epithelial cells. Although

estrogen receptor is not required for VDR-mediated growth inhibition of breast

cancer cell lines (85,93,94), mammary glands from mice deficient in VDR

demonstrate an enhanced proliferative response to estrogen combined with other

mammogenic hormones (79). These data suggest that at least one facet of

87 vitamin D anti-proliferative action involves opposition to proliferative hormonal stimuli. With this in mind, we attempted to harness the growth inhibitory qualities

of VDR activation in order to hinder hormone-induced mammary gland pathology

within a transgenic mouse model.

Importantly, vitamin D receptor is strongly expressed in the mammary glands and tumors of LH-overexpressing mice. Although expression of this protein is well documented in human breast cancers (77,78,245), to date, interrogation of mouse models of this disease has been limited to the MMTV-neu mouse (84). The LH-overexpressing mouse represents a unique model of mammary tumorigenesis with a molecular profile and morphology distinct from that of MMTV-neu (data not shown), hence providing a valuable outlet for probing the potential of VDR activation in tumors not associated with amplification of

HER2/neu.

The VDR promoter is activated in response to estradiol (246) and VDR

mRNA levels increase in human breast cancer cell lines that have been exposed

to progestins (247), suggesting that the altered hormonal milieu of the LH-

overexpressing mice may contribute to the observed changes in vitamin D

receptor expression. However, while pre-neoplastic mammary epithelial cells in

these mice express steroid hormone receptors and exhibit ovarian dependency,

tumors largely lack expression of both progesterone and estrogen receptors

((190) and Milliken and Keri, unpublished observations), revealing that VDR

overexpression is not dependent on ER or PR. The p38 and JNK MAPK

88 pathways have also been shown to activate VDR (248), indicating that a broad array of signaling pathways may impinge on the regulation of this protein.

The study reported herein shows, for the first time, that treatment with a

VDR agonist can have a significant impact on the proliferative index of the intact mammary gland in the presence of elevated mammogenic hormones. The anti- proliferative effect of vitamin D and several related analogs has been extensively observed in the controlled environment of cultured human breast cancer cells

(85-88) and has also been documented in xenograft models (101); however, verification of in vivo efficacy in targeting an endogenous tissue should stimulate enthusiasm for the potential of VDR agonists to act as chemopreventive agents.

The value of chemoprevention is illustrated by the Breast Cancer Prevention

Trial, in which tamoxifen significantly decreased the incidence of both invasive and non-invasive breast cancers in high risk patients (249). While promising, it is important to note that tamoxifen did not alter the incidence of estrogen receptor negative breast cancers. The identification of chemopreventive agents that impact a broad spectrum of cancer phenotypes is, as yet, an unresolved challenge. In this regard, the vitamin D receptor provides a promising target and

1,25-(OH)2D3 analogs such as EB1089 may facilitate this effort.

Data regarding the chemotherapeutic potential of 1,25-(OH)2D3 analogs in breast cancer is limited. A Phase I clinical trial reports low toxicity, but no tumor response to EB1089 in patients with advanced disease (107); however, studies done with earlier stage patients have not yet been published. In animal models, a report that EB1089 can reduce the volume of carcinogen-induced mammary

89 tumors in rats is significant and intriguing (106), but it is important to determine the effectiveness of such compounds in a diverse assemblage of models, including those that approximate a more natural disease progression. The LH- overexpressing model of hormone-induced mammary cancer reflects the important role that hormones play in breast cancer, thus providing an informative outlet for discovery of chemopreventive and chemotherapeutic agents. In the

current study, half of tumors exposed to EB1089 decreased their growth rate,

suggesting that activation of VDR signaling can counteract the intrinsic

proliferative drive of at least a subset of tumors. The reason for variability in

tumor response is unknown, although it is not a function of tumor size or basal

growth rate (data not shown). Furthermore, VDR expression is not a

distinguishing factor, as the protein was detected by immunohistochemistry in

both responding and non-responding tumors (data not shown). Morphological

and molecular variability among mammary tumors of the LH-overexpressing

mouse has been observed and may be partially due to the presence of genomic

instability early in tumorigenic progression (Milliken et al., in preparation). The

occurrence of both EB1089 responsive and non-responsive tumors in LH-

overexpressing mice may make it a useful model for identification of indicators

that predict the outcome of treatment with VDR agonists.

One possibility is that these mammary tumors possess differing capacities

for transcriptional modulation of vitamin D receptor target genes in response to

agonist. Hampered response may be due to limiting levels of cooperative

signaling components or the presence of an inhibitory factor. In either case, basal

90 expression of VDR target genes may gauge pathway functionality. In this regard,

variability in mRNA levels of the VDR target genes osteopontin and 25-

hydroxyvitamin D-24 hydroxylase has been observed in mammary tumors of LH-

overexpressing mice despite equivalent levels of VDR (data not shown). To

determine whether relatively low levels of these genes predict refractoriness to

EB1089, correlation of pre-treatment mRNA expression, obtained via biopsy, with

treatment response should be performed in future trials.

The efficacy of EB1089 in this study, in combination with previously

published evidence, suggests that vitamin D receptor agonists warrant attention

as chemopreventive and chemotherapeutic agents. While VDR activation has

demonstrated a beneficial effect on its own, it is likely that the observed anti-

tumorigenic consequences would be magnified if it were coupled with additional approaches. This scheme has been validated in mice harboring MCF-7 xenografts; EB1089 treatment in combination with tamoxifen (102), paclitaxel

(103), or ionizing radiation (104) was more effective than the individual components at inhibiting growth or inducing tumor regression. It should be of value to extend this combinatorial paradigm to transgenic mouse models as well as other in vivo models of cancer to assist determination of optimal treatment paradigms utilizing vitamin D3 analogs prior to introduction into clinical

therapeutic programs.

91 Figure III-1

Mammary glands of LH-overexpressing mice demonstrate elevated expression of vitamin D receptor. (A) Affymetrix gene expression analysis reveals increased VDR mRNA in mammary glands of LH-overexpressing compared to age-matched wild-type mice at 8, 12, and 19 weeks of age as well as in tumors. Data was analyzed using Affymetrix MAS 5.0 (gene detection: hatched bars=absent, solid bars=present; signal intensities normalized to 5-wk wild-type; asterisk indicates sample significantly increased compared to age- matched wild-type control). Immunohistochemistry for VDR was performed on mammary glands of adult wild-type (B), and LH-overexpressing mice (C), as well as a mammary tumor (marked by asterisk) from an LH-overexpressing mouse

(D). (E) The percentage of epithelial cells expressing VDR is increased 2.3-fold in the non-tumor bearing mammary glands of adult LH-overexpressing mice relative to age-matched wild-type mice (*p<0.01). Examples of positively staining nuclei are denoted by arrows in (B). Images are magnified 200X.

92 A VitaminVitamin DD Receptor 60 wild type 50 * 40 LH 30 * 20 * * * 10

normalized signal 0 5 wks 8 wks 12 wks 19 wks ≥ 41 wks tumors

B C

DD 60 E *

40 ** 20

% VDR positive cells 0 WT LH

93 Figure III-2

EB1089 decreases mammary epithelial cell proliferation in LH-

overexpressing mice. EB1089 or vehicle was injected subcutaneously into LH- expressing female mice for two weeks after weaning. At 5 weeks of age, the mammary glands from vehicle (A) and EB1089 (B) treated animals were assessed for proliferation using BrdU labeling analysis (400X magnification). (C)

EB1089 treated animals exhibited 35% fewer BrdU positive mammary epithelial cells than vehicle treated controls (*p<0.05).

94 95 B A C

% BrdU positive cells 20 10 30 Vehicle EB1089 Vehicle EB1089 * Figure III-3

EB1089 displays anti-tumorigenic activity in a subset of hormone-induced mammary tumors. Tumor bearing animals were treated with vehicle or EB1089 and tumor mass was evaluated using magnetic resonance imaging (MRI). (A)

Percent change in growth rate of tumors upon treatment with vehicle or EB1089

(see Materials and Methods). Treatment with EB1089 for 14 days resulted in decreased rate of tumor growth in half of mice; two of these tumors regressed more than 50% (indicated by asterisks). MR images of a tumor-bearing mouse on Day 0 (B) and Day 14 (C) of EB1089 treatment (tumor is indicated by the white arrow; Ros (rostral) and Cau (caudal) indicate the orientation of the mouse,

L=liver).

96 A 200 vehicle 150 EB1089 100

50

0

-50

-100

% change in growth rate growth in % change -150 * -200 *

B Day 0 Ros C Day 14 Ros

L L

Cau Cau

97 CHAPTER IV

CENTROSOME AMPLIFICATION AND P53 SIGNALING IN A MODEL OF

HORMONE-INDUCED MAMMARY CANCER

Introduction

Aneuploidy and genomic instability are common features of human cancers. Although cell cycle checkpoints normally prevent propagation of cells with altered chromosomal content, cells undergoing transformation often acquire the ability to bypass such controls and select for chromosomal populations that favor unregulated growth. One hypothesized cause of genomic instability is centrosome amplification, which is regularly observed in tumors. Centrosomes are the microtubule-organizing centers of animal cells and play important roles in establishing cell polarity, as well as establishing the mitotic spindle and ensuring proper distribution of chromosomes during cell division. Centrosomal aberrations have been observed in a diverse array of human tumors, including those of the breast (160-163), pancreas (164), and prostate (165), as well as numerous cell lines derived from tumors (166). Many speculate that centrosome amplification leads to genomic instability by causing multipolar spindle formation, which results in missegregation of chromosomes during anaphase and creates a source of chromosomal instability. In fact, centrosome amplification has been correlated with aneuploidy and chromosomal instability in breast cancer (163) and other cancers (164-166); however controversy exists regarding whether centrosome amplification is a cause or result of tumorigenesis. Evidence that centrosome

98 amplification is an early event in tumorigenesis (rather than a byproduct) is

provided by the fact that it has been identified in in situ carcinomas of the breast,

cervix, and prostate (163,168) as well as in pre-neoplastic cells in a rat model of

mammary carcinogenesis (169). In vitro experiments have also indicated that

centrosome amplification can precede and drive genomic instability as the

human papillomavirus (HPV)-16 E7 protein induces centrosome abnormalities

prior to the appearance of genomic abnormalities (170). At the least, centrosome

amplification is a sign that the cell is failing to enact appropriate checkpoint

controls.

The cause(s) of centrosome amplification in human cancers are unknown;

however, much speculation has surrounded the role of p53 in this process. Early

passage embryonic fibroblasts from p53-/- mice develop centrosome amplification

(250). Work in culture indicates that this is largely due to deregulation of the

centrosome duplication cycle (172); p53 null MEFs initiate centrosome duplication earlier in the cell cycle than their wild-type counterparts and are able to reduplicate their centrosomes even when DNA synthesis is inhibited (251).

This relationship has also been observed in human cancers: centrosome amplification in squamous cell carcinomas of the head and neck has been correlated with mutant p53 status (171,172); however, the link between p53 and centrosome amplification in breast cancer is less clear.

It is clear that aberrations in p53 signaling increase breast cancer risk.

Patients with Li-Fraumeni cancer susceptibility syndrome that carry germline mutations in p53 are at increased risk of developing breast cancer (140).

99 Similarly, transplantation of p53 null mouse mammary epithelial cells into a

cleared fat pad results in spontaneous tumor formation (151). However, the

frequency of somatic p53 mutation in breast cancer appears to be relatively low.

Overall, 50% of human cancers are thought to harbor p53 mutations, but

comprehensive meta-analysis showed that only 20% of breast cancers express mutant p53 (141); hence, the breast demonstrates a predisposition to employ

methods of tumor progression that are independent of p53 mutation status. This

trend also holds true for centrosome amplification, which is observed in 80% of

breast tumors, but occurs independently of p53 mutation (163).

The robust ability of the p53 pathway to prevent propagation of damaged

and potentially tumorigenic cells has caused many to hypothesize that in the

absence of p53 mutations other components of the p53 signaling pathway must

be altered in order to facilitate tumor formation. In this regard, amplification and overexpression of Hdm2 and HdmX, well characterized negative regulators of

p53 activity, have been documented in breast cancer (146,147,252), but even

combined, these instances account for only a very small proportion of the cases that occur independent of p53 mutations. Hence, there is room for significant expansion of knowledge regarding mechanisms of centrosome amplification and genomic instability in breast cancer. Members of the p53 pathway, both upstream and downstream, are good starting points for such analyses, but it is equally possible that modifications that are distantly associated with p53, or completely independent, play significant roles in these processes.

100 The ability to study mechanisms of disease development and progression is limited in humans, creating the need for animal models in order to facilitate examination and manipulation of biological systems. Transgenic mice overexpressing luteinizing hormone (LH) demonstrate spontaneous mammary tumor formation in response to elevated levels of mammogenic hormones (190), providing a unique model that reflects the well-established role of reproductive hormones in breast cancer risk and progression. Exposure to hormones during pregnancy creates a transient increase in risk of being diagnosed with breast cancer (14); however, women who have undergone a full-term pregnancy, especially before the age of twenty, have a much lower risk of developing this disease over the course of their lifetime (15,16). Pregnancy-mediated protection is also observed in mice and rats and can be mimicked in rodents by treatment with estrogen and progesterone (117,195). A role for p53 in pregnancy-mediated protection has been suggested based on the observation that increased levels and activity of p53 occur in mouse and rat mammary glands after treatment with estrogen and progesterone (118); however, further exposure to hormones is able to overcome pregnancy-induced protection (253), suggesting that hormones can overwhelm the protective effects of p53. This may explain why the breast, which is highly responsive to hormones, demonstrates a proclivity for the development of tumors independent of p53 mutation.

The following study reports the occurrence of centrosome amplification in tumors and pre-neoplastic mammary tissue of LH-overexpressing mice. Tumor development and centrosome amplification occur in the absence of p53

101 mutations, and use of a bitransgenic mouse model suggests that tumor

development is independent of change in p53 status. Phosphorylation of p53 in

response to ionizing radiation indicates that the p53 signaling pathway is intact

upstream of p53, and subsequent induction of p21 and Bax suggests that p53

becomes transcriptionally active; thus, the LH-overexpressing mice should

provide a useful model to investigate hormone-mediated mechanisms that allow

for the development of tumors independent of mutations in p53.

Materials and Methods

Animals

Mice were housed in microisolator plus units with a 12 hour light/dark

cycle and given food and water ad libitum. All mouse studies were approved by

the Institutional Animal Care and Use Committee at Case Western Reserve

University. Mice were asphyxiated with carbon dioxide. Tissues used for immunohistochemistry were fixed in 4% paraformaldehyde and subsequently

embedded in paraffin. Irradiated mice were exposed to 5 Gy whole body

irradiation administered using a 137Cs irradiatior and killed 2 or 6 hours later.

When indicated, stage of estrous cycle was determined by vaginal smear. WAP-

p53R172H mice were generously provided by Jeff Rosen.

Centrosome Staining

Mammary glands or tumors were removed and minced into 0.5-1mm

pieces with a scalpel. Cells were dissociated in collagenase type III (Worthington;

561 units collagenase/ml 1X PBS) with gentle agitation for approximately 30

102 minutes. Cells were pelleted and washed three times with PBS, resuspended in

PBS, and spun onto slides using a cytospin centrifuge. Cells were

fixed/permeabilized with ice cold methanol for 2 minutes. After blocking in 1% normal goat serum/1% Triton in PBS, cells were incubated overnight with anti- gamma tubulin antibody (Sigma T6557; 1:500). Cells were then incubated with fluorescein-conjugated goat anti-mouse secondary antibody (Jackson

ImmunoResearch, 115-095-003; 1:500) for one hour at room temperature. Slides were mounted with Vectashield Mounting Media containing DAPI counterstain

(Vector Laboratories, H-1200; diluted 1:4 in Vectashield mounting media, H-

1000). 150-1000 cells per animals were quantified. Results are reported as mean

± standard deviation.

p53 Sequence Analysis

Total RNA from mammary tumors (n=8) or wild-type mammary glands

(n=4) was DNase-treated and reverse transcribed with reverse primers for p53.

cDNA for p53 was amplified using Platinum High Fidelity Taq Polymerase in two

overlapping pieces using the following primer pairs: 5’-

ATCCTGGCTGTAGGTAGCGA-3’ with 5’-CTCCGTCATGTGCTGTGACT-3’ for

the 5’ end and 5’-AACTATGGCTTCCACCTGG-3’ with 5’-

CCCCACTTTCTTGACCATTG-3’ for the 3’ end. Both fragments of p53 were gel

purified using a GFX Purifiication Kit (Amersham, 27-9602-01) and sequenced by

Davis Sequencing (Davis, CA).

103 Real-Time RT-PCR

Reverse transcription using random primers was carried out on DNase-

treated total RNA. TaqMan Gene Expression Assays were purchased from

Applied Biosystems (Foster City, CA) for Mdm2 (Mm00487656_m1),

Mdm4/MdmX (Mm00484944_m1), CDKI1A/p21 (Mm00432448_m1), and TBP

(Mm00446973_m1). PCR was carried out using the ABI PRISM 7900HT

Sequence Detection System at CASE Comprehensive Cancer Center’s Gene

Expression Array Core Facility. Quantitation relative to TBP was performed using

ABI Sequence Detection Software. Samples were normalized to a 16-week-old

wild-type sample (Mdm2, MdmX) or a non-irradiated wild-type sample (p21).

Data was analyzed using a two-tailed student’s t-test.

Western Blotting

For analysis of pSer15-p53, 150 μg protein was run on an 8% SDS-

polyacrylamide Laemmli gel and transferred to PVDF membrane. The membrane

was blocked in 5% milk in TBST. Washes were performed in TBST. Primary

antibody (Cell Signaling, 9284S) was used at a dilution of 1:50 in 5% BSA and

was incubated overnight at 4°C. Incubation with peroxidase conjugated goat anti- rabbit secondary antibody (Santa Cruz, sc-2054; 1:5000) for 1 hour at room

temperature was used for detection. β-tubulin (Oncogene Research Products,

CP07; 1:5000) was used as a loading control and was detected using

peroxidase-conjugated goat anti-mouse secondary antibody (Santa Cruz, sc-

2055; 1:3000).

104 Northern Blotting

For analysis of Bax mRNA, 20 μg total RNA was run on a 1% agarose gel

containing formaldehyde and transferred to Nytran membrane. The membrane

was probed with a 32P-labelled hBax cDNA probe that was synthesized using the

DECAprime II kit (Ambion, #1455). Hybridization was carried out in QuikHyb

(Stratagene) according to manufacturer’s instructions.

Measurement of Apoptosis

Apoptosis was assessed using the ApopTag Red In Situ Apoptosis

Detection Kit (Chemicon, S7165). 5 μm sections were stained according to the manufacturer’s instructions and were counterstained and mounted with

Vectashield mounting media containing DAPI diluted in Vectashield mounting media (1:4). The percentage of TUNEL positive cells was determined by counting at least 1000 cells per animal. Differences in treatment groups were analyzed using a student’s t-test.

Gene Expression Profiling

Total RNA was isolated from thoracic mammary glands of nine wild-type and

nine LH-overexpressing mice at 16 weeks of age. RNA from three animals was

pooled, generating a total of three wild-type and three transgenic samples.

Double-stranded cDNA was made from the pooled RNA samples using the

Superscript Double Stranded cDNA Synthesis kit (Invitrogen, Carlsbad, CA; cat#

11917-010) and an oligo-dT primer coupled to the T7 RNA polymerase promoter.

Biotin-labeled cRNA was generated with the Enzo BioArray High Yield RNA

Transcript Labeling Kit (Affymetrix, Santa Clara, CA; cat# 900182) and purified

105 using the Qiagen RNEasy kit (cat# 74104). Samples were hybridized to

Affymetrix U74Av2 GeneChips® and data was analyzed using Affymetrix

MicroArray Suite 5.0 (MAS 5.0). Comparisons were made between each wild- type and each transgenic sample (total of 9 comparisons). In order to warrant further analysis, genes were required to meet the following criteria: 1) gene must be called “present” by Affymetrix in at least one sample, 2) gene must be called changed (increase, decrease, moderate increase, moderate decrease) in all comparisons (each wild-type compared to each transgenic sample), and 3) gene must have a p-value of <0.05 when analyzed using a Student’s t-test. Data mining was carried out in part using GeneSpring software (Silicon Genetics,

Redwood, CA).

Results

Pre-neoplastic Mammary Glands and Mammary Tumors from LH-overexpressing

Mice Display Centrosome Amplification

Centrosome amplification has been documented in early stage breast cancer (163), and is hypothesized to contribute to tumorigenesis by inducing

genomic instability. In order to determine whether centrosome amplification is a

feature of tumors in a mouse model of hormone-induced mammary cancer, we

dissociated mammary tumors from LH-overexpressing mice and mammary tissue

from age-matched wild-type mice and stained the resulting cells for γ-tubulin, a

centrosomal protein. All cells from wild-type mammary glands contained only one

or two centrosomes, reflecting normal variation in cell cycle stage; however,

106 increased centrosome numbers were observed in 1.6 ± 0.1% of cells from mammary tumors (Figure IV-1). These cells contained as few as three centrosomes, but, in many cases, ten or more centrosomes were visible in a single cell. These cells were not visually distinguishable from cells containing normal centrosome number; specifically, the nuclei displayed normal morphology and were not suggestive of an apoptotic state. It has been postulated that centrosome amplification is a result of tumor progression rather than a driving force of tumorigenesis (254). Although the presence of amplified centrosomes in these cells suggests deregulation of centrosome replication and cell cycle checkpoints within the environment of an established tumor, we wanted to

determine if this phenomenon occurs in early stages of hormone-induced

mammary cancer in this model. To investigate this question, we utilized mammary tissue from wild-type and transgenic mice that were 16 weeks of age.

The earliest age at which we have observed histological evidence of mammary tumors in this model is 19 weeks of age (personal observation); hence, analysis of tissue from 16 week old animals should provide insight into processes that are taking place prior to the formation of an overt tumor. Quantification of centrosome number in pre-neoplastic mammary glands showed 1.9 ± 0.9% of cells contained increased centrosome number (Figure IV-1), indicating that the altered hormonal milieu of the LH-overexpressing mice results in a mammary gland that is permissive to accumulation of cells with centrosome abnormalities and that these irregularities precede tumor formation.

107 Mammary Tumors of LH-overexpressing Mice Develop Independent of Mutations in p53

p53 mutations are found in only 20% of human breast cancers (141); however, they are present in 50% of human cancers (123) and p53 function is tightly coupled to maintenance of genomic stability and has been correlated with centrosome amplification in squamous cell cancers of the head and neck (171).

Hence, we analyzed cDNA from mammary tumors to determine whether they contained mutations in the coding region of p53. The entire coding region of p53

was amplified by RT-PCR and subjected to sequence analysis. No evidence of

sequence mutations were found when the p53 coding sequence from tumor

cDNA (n=8) was compared to that of wild-type littermates (n=4) (data not shown).

These data indicate that wild-type p53 mRNA is being expressed in the

mammary tumors of LH-overexpressing mice.

A Mutant Form of p53 Does Not Alter the Latency of Hormone-Induced

Mammary Tumorigenesis

The cooperative role of p53 in other mouse models of mammary

tumorigenesis has been revealed by introduction of a mutant form of p53 using a

transgenic approach (156). The p53R172H mutation in mice is analogous to the

human R175H mutation, one of the most common p53 mutations found in human

cancer (140). Although the function of this mutant is somewhat complex, it is

known to dimerize with wild-type p53 and prevents binding to DNA (255). To

determine whether introduction of this mutant to the mammary glands of the LH-

108 overexpressing mice has an impact on hormone-induced mammary

tumorigenesis, we bred the LH-overexpressing mice with transgenic mice

expressing this mutant under the mammary-selective, hormone-responsive whey

acidic protein (WAP) promoter. WAP-p53R172H mice do not develop tumors on

their own, but expression of this transgene accelerates mammary tumor formation in mice that also overexpress neu (156) or a mutant form of IGF-1

(158), as well as mice treated with the carcinogen DMBA (157). Tumor formation

was monitored in LH-overexpressing and LH-overexpressing/p53R172H mice by

weekly palpation. Kaplan-Meier analysis revealed no change in the tumor latency

of bitransgenic mice compared to LH-overexpressing mice despite strong

expression of the p53 transgene in the mammary glands of bi-transgenic mice

(Figure IV-2). This suggests that inactivation of p53 is not a rate-limiting step of

hormone-induced mammary tumorigenesis in the LH-overexpressing mouse.

mRNA Levels of Regulators of p53 Stability in Mammary Glands and Tumors of

LH-overexpressing Mice

Some tumors, including those of the breast, that contain wild-type p53 demonstrate increased expression of either Mdm2 (146,252) or MdmX (147), providing a mechanism for p53 inactivation that is unrelated to the integrity of its sequence. Although little is known about the regulation of MdmX, estradiol induction of Mdm2 expression in MCF-7 human breast cancer cells has been documented (256). We hypothesized that expression of one or both of these genes may be elevated in the mammary glands of LH-overexpressing mice,

109 resulting in decreased p53 activity. Analysis was perfomed using relative

quantitative RT-PCR. As seen in Figure IV-3, levels of Mdm2 mRNA are not

significantly different between wild-type and transgenic samples at 16 weeks of

age or in tumors compared to age-matched wild-type controls. A 1.5-fold

increase in MdmX mRNA was observed in the mammary glands of LH-

overexpressing mice relative to wild-type; however, tumors contained levels of

MdmX that were significantly lower than age-matched wild-type (43% decreased)

as well as 16 week old transgenic mammary glands (40% decreased). An age

dependent effect was also seen in wild-type animals: older animals demonstrated

a 1.6-fold increase in MdmX when compared to 16 week old mice. Altogether,

these data indicate that modestly elevated expression of MdmX, but not Mdm2,

may be a potential mechanism of p53 inactivation in pre-neoplastic mammary

glands of LH-overexpressing mice.

p53 is Activated by Ionizing Radiation in the Mammary Glands of Wild-Type and

LH-overexpressing Mice

To more directly assess the functional capacity of p53 in the mammary

glands of LH-overexpressing mice, we subjected mice to a challenge known to

activate the p53 signaling pathway. Ionizing radiation causes double-stranded

DNA breaks, initiating a cellular response that should result in activation of p53 followed by cell cycle arrest, DNA repair, and/or apoptosis (257). Aberrations at any level of the p53 pathway can impede these protective responses; however, the absence of p53 activation would suggest that a defect has occurred upstream

110 of p53 in the mammary glands of these mice. Therefore, we assessed activation

of p53 by way of phosphorylation of serine 18 in response to ionizing radiation.

Western blot analysis of mammary glands collected 2 hours after exposure to IR

revealed significant serine 18 phosphorylation of p53 in wild-type and transgenic

mammary glands, as well as in tumors (Figure IV-4A). This suggests that the

portion of the p53 signaling pathway between dsDNA break detection and

phosphorylation of serine 18 on p53 is intact.

Although stabilization and phosphorylation of p53 suggest that p53 is

transcriptionally active, it is possible that altered levels of coregulators could alter

the capacity of p53 to modulate transcription of its target genes. To indirectly assess the transcriptional activity of p53, we measured induction of two well- established p53 target genes: p21 and Bax. p21 mRNA levels were measured by relative quantitative RT-PCR. Strong induction of p21 was observed in both wild- type (12.5-fold) and transgenic (7.1-fold) mammary gland 6 hours after exposure to IR (Figure IV-3B). Induction of Bax was also observed in irradiated mice, and was independent of genotype (Figure IV-3C). Increased phosphorylation of p53 along with the upregulation of p21 and Bax strongly suggest that the p53 signaling pathway is functionally active in response to double-strand DNA breaks.

111 Mammary Epithelial Cells of Wild-type and LH-overexpressing Animals

Demonstrate Variable Apoptotic Response to Ionizing Radiation

While the presence of cells with amplified centrosomes in the mammary

glands of LH-overexpressing mice does not establish centrosome abnormalities

as the cause of the tumors that eventually form, it does suggest that mammary

cells in this model persist when they most likely should be induced to die. This

behavior, evasion of apoptosis, is a hallmark of cancer (258). p53 is necessary

for inducing apoptosis in cells that have acquired double-strand DNA breaks, as well as other genomic aberrations (reviewed in (257)). Loss of p53 function results in increased cell survival upon exposure to radiation, which may result in tumor formation (259). Several pieces of data have accumulated suggesting that

progesterone can inhibit apoptosis of mammary epithelial cells, both in vitro

(260,261) and in vivo (262-264); hence, we hypothesized that the elevated levels

of progesterone observed in the LH-overexpressing mice may provide the

mammary epithelial cells with an increased survival signal, allowing them to

evade apoptosis even in the presence of genomic aberrations.

To test this hypothesis, we measured apoptosis in response to ionizing

radiation. Overall, no difference in the induction of apoptosis was detected

between the mammary glands of wild-type and LH-overexpressing mice (Figure

IV-5). Extensive variability was observed in the induction of apoptosis in the

mammary glands of both genotypes, despite the fact that the lymph node housed

within the mammary gland consistently demonstrated high levels of apoptosis

upon exposure to IR (Figure IV-5). Although stage of the estrous cycle at the time

112 of radiation and death was determined by vaginal smear for wild-type mice (LH-

overexpressing mice are anovulatory), apoptotic response did not appear to

correlate with stage of estrous cycle or serum levels of progesterone (Figure IV-

5, data not shown). Hence, the altered hormone profile of LH-overexpressing

mice does not impart protection from radiation-induced apoptosis and a hormonal

basis for variability in apoptosis induction has not been identified for either

genotype.

Gene Expression Profiling Reveals Differences between Mammary Glands of

Wild-Type and LH-overexpressing Mice

Although they have not yet developed tumors, the mammary glands of 16-

week-old LH-overexpressing mice are hyperplasic and have centrosome

amplification (Figure IV-1), suggesting the altered hormonal milieu of these

animals creates an environment that is conducive to tumorigenesis. Analysis of

this gland should be beneficial for identifying early events in mammary tumor progression. In this regard, gene expression profiling was performed on mammary glands from 16 week old wild-type and LH-overexpressing mice using the Affymetrix U74Av2 platform, which probes approximately 12,000 genes and

ESTs. 622 genes were identified to be differentially expressed between wild-type and transgenic mice based on the filtering criteria described in the Materials and

Methods. Twenty of these genes were found to be associated with p53 signaling; centrosomes, polarity, and/or microtubules; or genomic stability (Table IV-1). The remaining changed genes are listed in Table IV-2 (upregulated in transgenic

113 mammary glands) and Table IV-3 (downregulated in transgenic mammary

glands). Verification and functional analysis of genes that are differentially

expressed in the mammary glands of LH-overexpressing mice, particularly those

in Table IV-1, may reveal mechanisms of hormone-induced centrosome

amplification and mammary tumorigenesis.

Discussion

Human tumors frequently exhibit aneuploidy and genomic instability,

which seem not only to contribute to tumor formation, but also to the acquisition of invasive and metastatic phenotypes (265). Amplification of centrosomes offers one potential mechanism by which cells develop genomic instability. Cells with excess centrosome number often form multipolar spindles during mitosis, increasing the risk of imbalanced chromosome segregation. Normal, non-dividing cells contain only one centrosome, the duplication of which is coupled to DNA replication during S-phase; however, alterations in many proteins, including p53, have been shown to uncouple centrosome and DNA replication, resulting in excess centrosome number (172,251).

Some controversy exists about whether aneuploidy, genomic instability,

and centrosome amplification are causes, or merely results, of tumorigenesis.

Support for the notion that centrosome amplification is an early event in

tumorigenesis is provided by the fact that it has been observed, along with mitotic

spindle defects and chromosome instability, in pre-invasive human breast

carcinomas in situ (168). Amplified centrosomes also occur in pre-neoplastic

114 cells in the mammary glands of rats treated with methylnitrosourea (169) and, as

shown in the current study, in the pre-neoplastic cells of the mammary glands of

LH-overexpressing mice (Figure IV-1B). If the mammary tumors that eventually

develop in the mammary glands of these mice begin from one of these cells

containing an increased number of centrosomes, one might expect virtually all of

the tumor cells to contain centrosome amplification; however, mammary tumors

in this model exhibit a degree of centrosome amplification similar to that of pre-

neoplastic tissue (Figure IV-1D). This does not necessarily rule out a tumor

precursor with more than two centrosomes; evidence has accrued for

centrosome coalescence, the merging of multiple centrosomes to form a single

functional microtubule organizing center (266,267). In fact, it has been

hypothesized that, in order to maintain mitotic fidelity, once a cell has acquired a

favorable genome, it must functionally inactivate excess centrosomes (267).

Although the mere presence of amplified centrosomes in the tumors and pre- neoplastic mammary glands of LH-overexpressing mice does not prove that this phenomenon causes hormone-induced mammary tumors, it does suggest that the environment created by elevated hormones is permissive to the accumulation of cells with centrosome aberrations. If centrosome amplification is a driving force behind the formation of these tumors the mechanism would likely be through the abnormal mitotic spindle formation and subsequent aneuploidy; thus, it would be of interest to determine the karyotypes of a number of mammary tumors from LH- overexpressing mice.

115 p53 is the most commonly mutated gene in human cancers (123) and

deletion of p53 leads to centrosome amplification in mouse embryonic fibroblasts

(250). Mutations in p53 have also been correlated with centrosome amplification

in squamous cell cancers of the head and neck in humans (171); however, p53 is

only mutated in approximately 20% of human breast cancers (141) and

centrosome amplification commonly occurs in the absence of mutation of p53 in

these tumors (163,171). In this regard, the mammary tumors of LH-

overexpressing mice seem to be like the majority of breast cancers – devoid of

p53 mutations while exhibiting centrosome amplification. Indeed, abrogation of p53 does not appear to be a rate-limiting step in mammary tumor formation in the

LH-overexpressing mice, as introduction of the R172H DNA binding mutant, which is equivalent to one of the most common p53 mutation in human cancer

(R175H) (268), does not alter tumor latency (Figure IV-2).

An intact p53 coding sequence does not ensure functional activity of the

p53 signaling pathway. Several reports have described the existence of cancers

that contain alterations in proteins directly associated with p53. Mdm2, the

ubiquitin ligase that chronically mediates degradation of p53, is amplified or

upregulated in a small subset of breast cancers (146,252) and its overexpression

causes mammary tumors in transgenic mice (269). There is a correlation

between Mdm2 overexpression and expression of estrogen receptor in human

breast tumors (270,271) and estradiol induces expression of Mdm2 in MCF-7

human breast cancer cells (256). As the LH-overexpressing mice have elevated

levels of estradiol (190), it would not be surprising if Mdm2 was upregulated in

116 the mammary glands of these mice; however, no changes in mRNA levels were observed in pre-neoplastic glands or tumors (Figure IV-3). MdmX, which also inhibits p53 activation (272), has recently been found to be amplified or overexpressed in a subset of human breast tumors (147). Unlike Mdm2, MdmX is moderately upregulated in pre-neoplastic mammary glands of LH-overexpressing mice (Figure IV-3). Little is known about what regulates MdmX expression, but elevated levels in the mammary glands of LH-overexpressing mice suggest that hormones may play a role. The loss of upregulation of MdmX observed in mammary tumors of this same model is consistent with regulation by one of the steroid hormones, as these tumors do not retain expression of ER or PR (Figure

II-8, (190)). Unfortunately, lack of a specific antibody to MdmX has precluded determination of whether changes observed at the level of the mRNA are reflected at the level of the MdmX protein.

There are several proteins upstream of p53 that, if altered, may inhibit or attenuate p53 functional activity. For example, loss of ATM (273), Chk2 (274),

BRCA1 (275), or BARD1 (275) results in significantly decreased activation and/or stabilization of p53 in response to ionizing radiation. Hence, efficient activation of p53 in response to ionizing radiation suggests that signaling components upstream of p53 are intact. Activation of p53 upon exposure to ionizing radiation requires phosphorylation of serine 18 (serine 15 in humans) by ATM (276,277).

The extensive accumulation of p53 phosphorylated on serine 18 upon exposure to ionizing radiation (Figure IV-4A) indicates that the signaling pathway upstream of p53 that detects double-stranded DNA breaks is intact in pre-neoplastic

117 mammary glands and tumors of LH-overexpressing mice; however, this does not necessarily mean that p53 is transcriptionally active. Although not a direct test of p53 transcriptional activity, induced expression of p21 and Bax, both archetypal p53-target genes, in the mammary glands of irradiated wild-type and LH- overexpressing mice strongly suggests functional activity of p53. Of course, this does not exclude the possibility that other p53-target genes, which may be distinctly regulated, are not differentially induced or suppressed.

The lack of a difference between radiation-induced apoptosis in the mammary glands of wild-type and LH-overexpressing mice implies that the portion of the p53 pathway that is liable for responding to double-strand breaks is not defective in the pre-neoplastic mammary glands of transgenic mice.

Progesterone has been reported to prevent apoptotic processes associated with post-lactational involution in the mammary gland; however, this does not rule out the possibility that there may be other hormone-induced modifications in p53 signaling that contribute to the survival and propagation of cells containing centrosome amplification or other genomic abnormalities.

Overall, the data presented in this chapter show that, despite the fact that p53 can be functionally activated, the mammary glands of LH-overexpressing mice develop centrosome amplification and eventually form tumors. It is likely that hormone-induced changes in the mammary glands of these mice allow some cells to bypass the normal protective mechanisms normally initiated by p53 or other caretaker pathways. The fact that p53 can be activated suggests that implementation of the p53 protection program is being attenuated downstream of

118 p53. Further analyses of this model of hormone-induced mammary tumorigenesis should provide insight into mechanisms employed by the breast to allow for survival of transformed cells independent of p53 mutation. Expanded knowledge in this area should assist in development of chemopreventives, chemotherapeutics and diagnostic indicators for breast cancer.

119 Table IV-1. Genes associated with p53*, centrosomes/microtubules/polarity§, and genomic instability¥

Affymetrix Genbank Fold Gene Namea Accession Accession Changeb Ref Mid1 interacting protein (MIG12)§ 95135_at AI844396 0.33 (278) Sucrose non-fermenting protein (SNF-1) related kinase§ 97429_at AW048113 0.36 (279) Sestrin 1* 95731_at AI843106 0.48 (280) Tubulin, alpha 1§ 100342_i_at M28729 0.5 (281) Kinesin 2§ 97984_i_at AA709672 0.59 (282) Hairy/enhancer of split-related with YRPW motif 1* 95671_at AJ243895 0.63 (283) Nucleobindin§ 94839_at M96823 0.63 (284) Yippee-like 3§ 96615_at AI840137 0.67 (285) (286,28 F-box and WD-40 domain protein 2§ 100056_at X54352 0.71 7) Pituitary tumor transforming 1*¥ 101027_s_at AF069051 1.7 (288) ASF1 anti-silencing function 1 homolog A¥ 98914_at AI853634 2 (289) Crumbs homolog 3§ 104584_f_at AI845823 2.06 (290) Pelota homolog¥ 161280_r_at AV292400 2.09 (291) Clusterin* 161294_f_at AV003873 2.12 (292) Kinesin family member 20A§ 161856_f_at AV059766 2.36 (282) Growth arrest and DNA damage-inducibe gene, gamma¥ 101979_at AF055638 2.38 (293) Excision repair 2¥ 99397_at U97572 2.42 (294) Zinc-binding protein 89* 99502_at U80078 3.42 (295) Regulator of G-protein signaling 16* 161609_at AV349152 4.56 (296) (297,29 A kinase (PRKA) anchor protein 9§ 93464_at AI561567 5.38 8)

aData were analyzed using Affymetrix MicroArray Suite 5.0. In order to be considered changed, genes must fulfill all of the following criteria: 1) be called

“present” by the Affymetrix algorithm in at least one sample, 2) be called changed

(increase, decrease, moderate increase, moderate decrease) by the Affymetrix algorithm in all comparisons (each wild-type compared to each transgenic sample), and 3) have a p-value of <0.05 when analyzed using a Student’s t-test

120 (wild type vs. transgenic). bFold change represents expression in transgenic mammary glands relative to wild type and is derived from the Affymetrix signal log ratio (SLR).

121 Table IV-2. Genes upregulated in mammary glands of LH-overexpressing relative to wild-type mice

Affymetrix GenBank Fold Gene namea Accession Accession Changeb CD9 antigen 95661_at L08115 1.28 Ribosomal protein 10 98342_at M93980 1.35 High mobility group box 3 98038_at AF022465 1.36 Actin, gamma, cytoplasmic 1 96573_at M21495 1.38 G protein pathway suppressor 1 96943_at AW125234 1.45 Isoleucine-tRNA synthetase 93752_at AI848393 1.45 Ribosomal protein L7 97696_r_at M29015 1.45 Cathepsin H 94834_at U06119 1.46 SET and MYND domain containing 5 102375_at AI843738 1.46 Transducer of ErbB-2.1 99532_at D78382 1.46 Casein alpha 96030_at M36780 1.47 Cytokine inducible SH2-containing protein 100022_at D89613 1.47 RIKEN cDNA 2210008M09 gene 160702_at AA791885 1.47 Valyl-tRNA synthetase 2 97894_at AF109905 1.49 Methionine adenosyltransferase II, beta 94469_at AW120950 1.5 CUB domain containing protein 1 104198_at AW211760 1.51 Aldehyde dehydrogenase 18 family, member A1 95738_at AW124889 1.52 DNA segment, Chr 7, Wayne State University 128, expressed 103862_r_at AA388099 1.52 Ribosomal protein S26 98564_f_at U67770 1.52 SEC61, gamma subunit 92636_f_at U11027 1.52 RAB11 family interacting protein 5 (class I) 95618_at AI843884 1.53 Ribosomal protein S28 98085_f_at U11248 1.54 RIKEN cDNA 1810015C04 gene 95518_at AW122893 1.54 RIKEN cDNA 3110001N18 gene 93987_f_at AW121568 1.54 RIKEN cDNA 2310008M10 gene 160569_at AI848235 1.55 Platelet-activating factor acetylhydrolase, isoform 1b, alpha1 subunit 100576_at U57746 1.56 cDNA sequence BC005537 94830_at AI854300 1.57 Nucleolar protein family A, member 2 97824_at AW121031 1.58 CEA-related cell adhesion molecule 2 102804_at X67279 1.6 Sh3 domain YSC-like 1 103813_at D85926 1.6 Claudin 7 99561_f_at AF087825 1.62 Glycoprotein 49 A leukocyte immunoglobulin-like receptor, subfamily B, member 4 92217_s_at U05265 1.62 Villin 2 100084_at X60671 1.62 Expressed sequence C77080 161696_f_at AV232292 1.63 Interferon-stimulated protein 103432_at AW122677 1.63 Vesicle-associated membrane protein 8 100345_f_at W65964 1.63 Chloride intracellular channel 1 95654_at AF109905 1.64 RIKEN cDNA 3110006P09 gene 97463_g_at AI847584 1.64 Sec61 alpha 1 subunit 97882_at AB032902 1.64 Interleukin 18 binding protein 92689_at AB019505 1.66

122 Affymetrix GenBank Fold Gene namea Accession Accession Changeb Spermatogenesis associated 13 100958_at AI647003 1.66 Glutamate-cysteine ligase , modifier subunit 160335_at U95053 1.67 Serine protease inhibitor, Kunitz type 1 97206_at AW230369 1.67 Extracellular proteinase inhibitor 103051_at X93037 1.68 PQ loop repeat containing 1 160801_at AW061073 1.68 Dehydrogenase/reductase (SDR family) member 8 102370_at AA822174 1.69 Solute carrier family 39 (metal ion transporter), member 6 103459_at AW124544 1.69 Protein phosphatase 1, catalytic subunit, beta isoform 100088_at M27073 1.7 SRY-box containing gene 10 102856_at AF047389 1.7 WW domain binding protein 5 100522_s_at U92454 1.7 Ring finger protein 149 98915_at AI849082 1.71 Signal sequence receptor, beta 101061_at AI845293 1.71 X-box binding protein 1 94821_at AW123880 1.71 Zinc finger, DHHC domain containing 9 104741_at AI843023 1.72 Claudin 3 94493_at AF095905 1.74 RIKEN cDNA 2810417H13 gene 100116_at AI122538 1.74 Complement component 3 93497_at K02782 1.75 Mitogen activated protein kinase kinase kinase 1 103020_s_at AI317205 1.75 Expressed sequence C79248 94689_at C79248 1.76 Solute carrier family 9 (sodium/hydrogen exchanger), isoform 3 regulator 1 97243_at U74079 1.76 DnaJ (Hsp40) homolog, subfamily C, member 3 102414_i_at U28423 1.78 Keratinocyte associated protein 2 97269_f_at AI848699 1.78 PRKC, apoptosis, WT1, regulator 93439_f_at AA260005 1.78 Ring finger protein 5 101421_at AF030001 1.78 Transmembrane emp24 protein transport domain containing 7 160183_f_at AI846109 1.78 hematopoietic-specific early-response A1-d 93869_s_at U23781 1.79 hematopoietic-specific early-response A1-b 102914_s_at U23778 1.8 RIKEN cDNA 1810009M01 gene 97885_at AB031386 1.8 RIKEN cDNA 3732413I11 gene 96104_at AI047107 1.8 Tenascin C 101993_at X56304 1.8 Asparaginyl-tRNA synthetase 95070_at AW125874 1.81 Frequently rearranged in advanced T-cell lymphomas 2 103699_i_at AI646638 1.81 N-myc downstream regulated gene 1 96596_at U52073 1.81 Pleckstrin homology-like domain, family A, member 1 160829_at U44088 1.81 Ribonuclease T2 96653_at AI851762 1.81 Syntaxin 3 100499_at D29797 1.81 Transcribed locus 96530_at AI503926 1.81 CD14 antigen 98088_at X13333 1.82 Riboflavin kinase 101050_at AF031381 1.82 RIKEN cDNA E230022H04 gene 103082_at AI847507 1.82 RIKEN cDNA 1600029D21 gene 97413_at AI121305 1.83

123 Affymetrix GenBank Fold Gene namea Accession Accession Changeb Testis expressed gene 261 101517_at X81058 1.83 Casein beta 99130_at X04490 1.84 TPA regulated locus 160549_at M23568 1.85 RIKEN cDNA 1110008P14 gene 96709_at AI839839 1.86 RIKEN cDNA 1300010A20 gene 104398_at AI846222 1.86 Phosphofructokinase, liver, B-type 92637_at J03928 1.87 Nuclear factor of kappa light chain gene enhancer in B-cells inhibitor, alpha 104149_at AI642048 1.88 Solute carrier family 16 (monocarboxylic acid transporters), member 1 101588_at AF058055 1.89 Adipose differentiation related protein 98589_at M93275 1.9 Casein gamma 100463_at D10215 1.91 Cathepsin S 98543_at AJ223208 1.91 Myc induced nuclear antigen 102340_at AA833077 1.91 Spermidine/spermine N1-acetyl transferase 1 96657_at L10244 1.91 TEA domain family member 2 96940_at Y10026 1.91 Arginase type II 98473_at AF032466 1.92 Adenosine deaminase 98632_at M10319 1.94 G protein-coupled receptor 172B 96201_at AW123978 1.94 Lymphocyte cytosolic protein 1 94278_at D37837 1.94 Casein kappa 99065_at M10114 1.95 Elongation factor RNA polymerase II 2 103892_r_at AI197161 1.96 RIKEN cDNA 1110001C20 gene 95458_s_at AW121960 1.96 Sarcolemma associated protein 98553_at AW124175 1.96 cDNA sequence BC037006 104643_at AI850846 1.98 Phospholipase D3 100607_at AF026124 1.98 Suppressor of variegation 4-20 homolog 2 (Drosophila) 102686_at AI605493 1.98 Lysozyme 100611_at M21050 1.99 Solute carrier family 41, member 1 103547_at AI837116 1.99 ASF1 anti-silencing function 1 homolog A 98914_at AI853634 2 Chloride channel 3 94464_at AF029347 2.01 Vitamin D receptor 99964_at 2.01 Protease, serine, 8 100909_at AA760364 2.02 Solute carrier family 5 (sodium/glucose cotransporter), member 1 94357_at AA591002 2.02 CD24a antigen 100600_at M58661 2.04 DNA segment, Chr 3, University of California at Los Angeles 1 95708_at AI843466 2.04 ROD1 regulator of differentiation 1 93829_at AW107884 2.04 Eukaryotic translation initiation factor 5B 98141_at AA647048 2.05 Wingless-related MMTV integration site 5B 103513_at M89799 2.06 Antigen identified by monoclonal antibody Ki 67 99457_at X82786 2.08 CCR4-NOT transcription complex, subunit 3 102018_at AI854879 2.08 Potassium intermediate/small conductance calcium-activated channel, subfamily N, member 4 102198_at AF042487 2.09

124 Affymetrix GenBank Fold Gene namea Accession Accession Changeb Nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 3 97975_at D85612 2.11 RIKEN cDNA 5031439G07 gene 104284_at AI846126 2.14 Nuclear factor, erythroid derived 2, like 3 99527_at AB013852 2.16 Semaphorin 4D 160836_at U69535 2.16 Beta galactoside alpha 2,6 sialyltransferase 1 94432_at AI117157 2.17 Keratin complex 1, acidic, gene 19 102121_f_at AU040563 2.17 MARCKS-like protein 97203_at X61399 2.17 Solute carrier family 11 (proton-coupled divalent metal ion transporters), member 2 104451_at AI852578 2.17 N-myc downstream regulated-like 160464_s_at U60593 2.21 cDNA sequence BC005662 160897_at AW060889 2.22 NMDA receptor-regulated gene 1 93246_at AW260482 2.22 Forkhead box M1 98306_g_at Y11245 2.23 Peroxisomal biogenesis factor 11a 103660_at AF093669 2.25 RAB6 98927_at AI851048 2.25 CD320 antigen 102343_at AF110520 2.26 Disabled homolog 2 98045_s_at U18869 2.28 Insulin-like growth factor binding protein, acid labile subunit 97987_at U66900 2.28 Lipopolysaccharide binding protein 96123_at X99347 2.28 RIKEN cDNA 6030432P03 gene 99866_at AI642417 2.28 IQ motif containing GTPase activating protein 1 100561_at AW209098 2.3 cDNA sequence BC026744 92398_at AI848723 2.31 Peroxiredoxin 4 93495_at U96746 2.35 RIKEN cDNA 0610011I04 gene 96605_at AI787183 2.35 Kinesin family member 20A 161856_f_at AV059766 2.36 Cytosolic acyl-CoA thioesterase 1 103581_at Y14004 2.41 Laminin, beta 3 92759_at U43298 2.41 ATPase, H+ transporting, lysosomal V0 subunit a isoform 2 102273_at M31226 2.42 Excision repair cross-complementing rodent repair deficiency, complementation group 2 99397_at U97572 2.42 Glycerol kinase 97525_at U48403 2.42 Ubiquitin D 92715_at AL078630 2.45 Glucuronidase, beta 97538_at M19279 2.48 Integrin alpha 6 95511_at X69902 2.48 UDP-glucose ceramide glucosyltransferase 96623_at AI853172 2.48 Cytochrome b-245, alpha polypeptide 97013_f_at AW046124 2.49 CASP8 and FADD-like apoptosis regulator 103217_at Y14041 2.51 CCR4 carbon catabolite repression 4-like 101787_f_at X16672 2.52 Protein tyrosine phosphatase, receptor type, V 92662_g_at U36488 2.52 Granzyme B 102877_at M12302 2.53 Tumor protein D52 160249_at U44426 2.53 RIKEN cDNA 5430432M24 gene 103343_at AI845815 2.55 Semaphorin 4B 95387_f_at AA266467 2.55 Aminoadipate-semialdehyde synthase 103389_at AJ224761 2.64

125 Affymetrix GenBank Fold Gene namea Accession Accession Changeb Neutrophil cytosolic factor 2 102326_at AB002664 2.67 Riken cDNA 4930413B13 gene 161227_r_at AV260411 2.68 N-acetylneuraminic acid synthase 104147_at AW122052 2.74 Nuclear factor I/X 100307_at AA002843 2.74 RIKEN cDNA 6330505N24 gene 104207_at AI430272 2.78 Solute carrier family 39 (zinc transporter), member 4 97320_at AI842734 2.8 DNA segment, Chr 6, Wayne State University 176, expressed 95541_at AW125506 2.83 Proviral integration site 1 99384_at M13945 2.83 Nephronectin 103721_at AA592182 2.84 Prostaglandin E receptor 4 (subtype EP4) 103362_at D13458 2.85 Phospholipid scramblase 1 102839_at D78354 2.86 Gap junction membrane channel protein beta 2 98423_at M81445 2.88 WW domain containing transcription regulator 1 95627_at AW046038 2.99 CD5 antigen-like 93445_at AF011428 3.02 FK506 binding protein 11 97964_at AW122851 3.02 ATP-binding cassette, sub-family G (WHITE), member 2 93626_at AF103875 3.09 Ets homologous factor 102243_at AF035527 3.1 Carbonic anhydrase 2 92642_at M25944 3.2 Guanine nucleotide binding protein, alpha 13 100514_at M63660 3.21 Jun-B oncogene 102362_i_at U20735 3.23 RIKEN cDNA 0610040J01 gene 95937_at AI429010 3.24 Transcription factor 20 100947_at AI847906 3.25 Ubiquitin associated protein 2-like 95964_at C77404 3.26 Stratagene 937302 cDNA 102348_at AI551087 3.29 Acyl-CoA synthetase long-chain family member 4 104017_at AB033887 3.31 Defensin beta 1 100882_at AF003525 3.37 Zinc finger protein 148 99502_at U80078 3.42 SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily c, member 1 102062_at U85614 3.43 Desmoglein 2 104480_at AI152659 3.48 Nucleobindin 2 102197_at AJ222586 3.54 Eukaryotic translation initiation factor 4E binding protein 2 94353_at U75530 3.57 RIKEN cDNA 5930418K15 gene 102870_at AW125272 3.58 Carboxyl ester lipase 99939_at U37386 3.64 Interleukin 6 receptor, alpha 104268_at X51975 3.71 T-cell immunoglobulin and mucin domain containing 2 103794_i_at AA986114 3.72 Polymeric immunoglobulin receptor 99926_at AB001489 3.74 DEAH (Asp-Glu-Ala-His) box polypeptide 36 95944_at AV299153 3.76 Glycine decarboxylase 95603_at AW123955 3.77 Solute carrier family 7 (cationic amino acid transporter, y+ system), member 5 104221_at AB017189 3.81 Histone H2A.1 94805_f_at M33988 3.83

126 Affymetrix GenBank Fold Gene namea Accession Accession Changeb cDNA sequence BC004044 100949_at AI461767 3.87 HLA-B-associated transcript 3 93912_at AW047616 3.89 Peptidoglycan recognition protein 1 104099_at AF076482 3.89 Tumor necrosis factor receptor superfamily, member 11a 101632_at AF019046 3.97 DNA segment, Chr X, ERATO Doi 242, expressed 97142_at C80153 4.02 Transcribed locus 104477_at AW047643 4.02 Glycosylation dependent cell adhesion molecule 1 98063_at M93428 4.09 Endogenous provirus Imposon1 envelope gene 101313_r_at U95783 4.16 RIKEN cDNA 1810057B09 gene 98958_at AA759910 4.18 Phosphatase and tensin homolog 160614_at U92437 4.27 Expressed sequence AI451896 96513_at AA794350 4.36 RE1-silencing transcription factor 103720_at AI449034 4.73 U2AF homology motif (UHM) kinase 1 160778_at AI846236 4.9 Coactosin-like 1 95466_at AI837006 5.12 DEAD (Asp-Glu-Ala-Asp) box polypeptide 6 93965_r_at AF038995 5.29 Gap junction membrane channel protein beta 6 94391_at AW123650 5.4 Quiescin Q6 96602_g_at AW045751 5.82 Serum amyloid A 3 102712_at X03505 6.1 Killer cell lectin-like receptor subfamily B member 1B 94744_at M77677 6.49 SKI interacting protein 104158_at AW046671 6.99 Fatty acid binding protein 3, muscle and heart 94214_at X14961 7.07 Protein kinase, lysine deficient 1 161270_i_at AV319920 7.99 Inactive X specific transcripts 99126_at L04961 8.17 Casein delta 98814_at V00740 8.55 RIKEN cDNA 1300017J02 gene 103407_at AA895838 8.57 Sulfotransferase family 1D, member 1 160537_at AF026073 8.77 Cathepsin E 104696_at AJ009840 9.58 Placentae and embryos oncofetal gene 101368_at M32484 10.53 Opioid growth factor receptor-like 1 94693_at C79135 18.07

aData were analyzed using Affymetrix MicroArray Suite 5.0. In order to be considered changed, genes must fulfill all of the following criteria: 1) be called

“present” by the Affymetrix algorithm in at least one sample, 2) be called changed

(increase, decrease, moderate increase, moderate decrease) by the Affymetrix algorithm in all comparisons (each wild-type compared to each transgenic sample), and 3) have a p-value of <0.05 when analyzed using a Student’s t-test

127 (wild type vs. transgenic). bFold change represents expression in transgenic mammary glands relative to wild type and is derived from the Affymetrix signal log ratio (SLR).

128 Table IV-3. Genes downregulated in mammary glands of LH-overexpressing relative to wild-type mice

Affymetrix GenBank Fold Gene Namea Accession Accession Changeb Inter-alpha trypsin inhibitor, heavy chain 2 104519_at X70392 0.06 Fibrinogen, gamma polypeptide 93096_at AA986050 0.08 Tumor-associated calcium signal transducer 2 160651_at AI563854 0.11 Carboxylesterase 3 101539_f_at AW226939 0.13 Protocadherin alpha 11 160610_at D86916 0.13 Cbp/p300-interacting transactivator with Glu/Asp-rich carboxy-terminal domain 1 160705_at U65091 0.15 Potassium channel, subfamily K, member 2 104652_at AI849601 0.17 Prominin 1 93390_g_at AF039663 0.17 Glutathione S-transferase, mu 2 93009_at J04696 0.19 Amylase 1, salivary 101058_at J00356 0.2 Carboxypeptidase E 99643_f_at X61232 0.2 ELOVL family member 6, elongation of long chain fatty acids 94418_at AI839004 0.21 Ectonucleotide pyrophosphatase/phosphodiesterase 2 97317_at AW122933 0.22 Glutamyl aminopeptidase 102373_at M29961 0.22 ST6 (alpha-N-acetyl-neuraminyl-2,3-beta-galactosyl- 1,3)-N-acetylgalactosaminide alpha-2,6- sialyltransferase 5 92403_at AB030836 0.22 Insulin-like growth factor binding protein 3 95082_at AI842277 0.23 Interferon, alpha-inducible protein 27 92718_at AI158810 0.23 Melanocortin 2 receptor accessory protein 104325_at AI461631 0.24 Retinol binding protein 4, plasma 96047_at U63146 0.24 Solute carrier family 1 (neutral amino acid transporter), member 5 92582_at L42115 0.24 UDP-Gal:betaGlcNAc beta 1,3-galactosyltransferase, polypeptide 2 92341_at AF029791 0.24 Angiotensinogen 101887_at AF045887 0.25 Prostaglandin E receptor 3 (subtype EP3) 96588_at D10204 0.25 Resistin 102366_at AA718169 0.25 Neuronatin 97520_s_at X83569 0.26 ADP-ribosyltransferase 3 98924_at Y08027 0.27 Adrenergic receptor, beta 3 161900_f_at AV373835 0.27 Cytochrome P450, family 2, subfamily e, polypeptide 1 93996_at X01026 0.27 Heat shock 27kDa protein 8 160139_at AI848798 0.27 Solute carrier family 2 (facilitated glucose transporter), member 4 102314_at M23383 0.27 RIKEN cDNA 6530401D17 gene 95523_at AI839718 0.28 Thyroid stimulating hormone receptor 98328_at U02602 0.28 Phosphofructokinase, platelet 97834_g_at AI853802 0.29 Serine (or cysteine) proteinase inhibitor, clade A, member 3C 102707_f_at X61597 0.29 Aldehyde dehydrogenase family 1, subfamily A7 94778_at U96401 0.3

129 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb Cell death-inducing DFFA-like effector c 102016_at M61737 0.3 Cyclin-dependent kinase inhibitor 2B 101900_at AF059567 0.3 Endomucin 93885_g_at AB034693 0.3 Hydroxysteroid 11-beta dehydrogenase 1 97867_at X83202 0.3 Orosomucoid 1 100436_at M27008 0.3 Superoxide dismutase 3, extracellular 94902_at U38261 0.3 Adiponectin receptor 2 104605_at AW047554 0.31 Chemokine (C-X-C motif) ligand 12 162234_f_at AV139913 0.31 Manic fringe homolog 100508_at AF015769 0.31 Parathyroid hormone receptor 1 98482_at X78936 0.31 Sorbin and SH3 domain containing 1 160320_at U58883 0.31 Synuclein, gamma 104280_at AF017255 0.31 Tryptophan hydroxylase 1 99972_at J04758 0.31 Vanin 1 104165_at AJ132098 0.31 ATPase, Na+/K+ transporting, alpha 2 polypeptide 99481_at AI839697 0.32 N-myc downstream regulated gene 2 96088_at AB033921 0.32 Plasma membrane associated protein, S3-12 102763_at AF064748 0.32 Pleckstrin homology-like domain, family A, member 3 98056_at AI846214 0.32 Protein tyrosine phosphatase, receptor type, D 93485_at AI844911 0.32 Fibroblast growth factor receptor 2 93090_at M23362 0.33 FXYD domain-containing ion transport regulator 1 93040_at AF091390 0.33 Glutathione transferase zeta 1 160350_at AW060750 0.33 Malic enzyme, supernatant 101082_at J02652 0.33 Melanocortin 2 receptor 92429_at D31952 0.33 Microsomal glutathione S-transferase 3 96258_at AI843448 0.33 Prion protein 100606_at M18070 0.33 RIKEN cDNA 9130213B05 gene 92971_at AW125849 0.33 Solute carrier family 19 (sodium/hydrogen exchanger), member 1 94419_at L23755 0.33 Sulfotransferase family 1A, phenol-preferring, member 1 103087_at L02331 0.33 Amphiregulin 99915_at L41352 0.34 Beta-3-adrenergic receptor 92536_at X72862 0.34 Epoxide hydrolase 2, cytoplasmic 93051_at Z37107 0.34 Frizzled homolog 4 95771_i_at U43317 0.34 Membrane metallo endopeptidase 92331_at M81591 0.34 Mesenchyme homeobox 2 99937_at Z16406 0.34 Peroxisome proliferator activated receptor gamma 97926_s_at U10374 0.34 RIKEN cDNA 2310016A09 gene 96122_at AW049373 0.34 Hephaestin 104194_at AF082567 0.35 Phytanoyl-CoA hydroxylase 96608_at AF023463 0.35 Retinoic acid receptor responder (tazarotene induced) 2 97835_at AI842828 0.35 SPARC related modular calcium binding 1 96712_at AI848508 0.35 A kinase (PRKA) anchor protein (gravin) 12 95022_at AB020886 0.36 Cysteine-rich protein 1 (intestinal) 94061_at M13018 0.36 Glutathione S-transferase, theta 1 95019_at X98055 0.36

130 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb Ras interacting protein 1 104146_at AI853551 0.36 RIKEN cDNA B430320C24 gene 97711_at AI606967 0.36 Glucan (1,4-alpha-), branching enzyme 1 96803_at AW210370 0.37 Glutamate-ammonia ligase 94852_at U09114 0.37 Hormone-sensitive lipase 103083_at U69543 0.37 Peripheral myelin protein 102395_at Z38110 0.37 RIKEN cDNA 1190002H23 gene 160359_at AI854358 0.37 RIKEN cDNA 5730469M10 gene 96634_at AI850090 0.37 Aldehyde oxidase 1 104011_at AB017482 0.38 cDNA sequence AB023957 96132_at AB023957 0.38 Growth hormone receptor 99107_at M31680 0.38 Interleukin 1 receptor-like 1 98500_at D13695 0.38 PFTAIRE protein kinase 1 93421_at AF033655 0.38 Protein tyrosine phosphatase, receptor type, M 92309_i_at X58287 0.38 RIKEN cDNA 2310010J17 gene 161018_at AI661590 0.38 RIKEN cDNA A430096B05 gene 97826_at AI465965 0.38 Thymoma viral proto-oncogene 2 160558_at U22445 0.38 Amine oxidase, copper containing 3 102327_at AF078705 0.39 Latent transforming growth factor beta binding protein 4 97347_at AA838868 0.39 Neuropilin 1 95016_at D50086 0.39 Procollagen, type XVIII, alpha 1 101881_g_at L22545 0.39 Receptor (calcitonin) activity modifying protein 2 99444_at AJ250490 0.39 ST3 beta-galactoside alpha-2,3-sialyltransferase 6 102208_at AI153959 0.39 Zinc finger protein 275 104002_at AI153693 0.39 Leptin 98443_at AI882416 0.4 Sarcoglycan, epsilon 101861_at AF031919 0.4 3-ketoacyl-CoA thiolase B 99571_at AW012588 0.41 Cyclin-dependent kinase inhibitor 2C 160638_at U19596 0.41 Dermatopontin 96742_at AA717826 0.41 Endoglin 100134_at X77952 0.41 Indolethylamine N-methyltransferase 97402_at M88694 0.41 Interferon activated gene 203 93321_at AF022371 0.41 Leukotriene C4 synthase 92401_at U27195 0.41 Multiple PDZ domain protein 93887_at AI854351 0.41 Phosphoenolpyruvate carboxykinase 1, cytosolic 160481_at AF009605 0.41 Protein phosphatase 2 (formerly 2A), regulatory subunit A (PR 65), beta isoform 103912_at AW046563 0.41 RIKEN cDNA 1100001G20 gene 101912_at AI019679 0.41 Sarcospan 102378_at U02487 0.41 Transmembrane protein 43 98633_at AI854863 0.41 Twist gene homolog 1 98028_at M63649 0.41 Aquaporin 7 101859_at AB010100 0.42 Calcium and integrin binding family member 2 99536_at AB016080 0.42 Carnitine acetyltransferase 103646_at X85983 0.42 Melanoma cell adhesion molecule 160458_at AI853261 0.42 RIKEN cDNA 4921515A04 gene 104494_at AI642098 0.42

131 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb RIKEN cDNA 9630050M13 gene 160963_at AI551141 0.42 Sulfide quinone reductase-like (yeast) 94515_at AW208628 0.42 Superoxide dismutase 1, soluble 100538_at M35725 0.42 Tensin like C1 domain-containing phosphatase 96825_at AI854794 0.42 BCL2/adenovirus E1B 19kDa-interacting protein 1, NIP3 93836_at AF041054 0.43 Cerebellar degeneration-related 2 93094_at U88588 0.43 Haptoglobin 96092_at M96827 0.43 Integral membrane protein 2A 93511_at L38971 0.43 Lectin, galactose binding, soluble 1 99669_at X15986 0.43 Pyruvate carboxylase 93308_s_at M97957 0.43 RIKEN cDNA 4631408O11 gene 104445_at AW046694 0.43 Septin 4 94079_at X61452 0.43 Solute carrier family 25 (mitochondrial carrier, citrate transporter), member 1 162358_i_at AV218217 0.43 Acyl-Coenzyme A oxidase 1, palmitoyl 101515_at AF006688 0.44 Annexin A2 100569_at M14044 0.44 Brain acyl-CoA hydrolase 100539_at AI841279 0.44 Expressed sequence AI413331 93472_at AA796989 0.44 Inhibitor of DNA binding 4 96144_at AJ001972 0.44 Insulin-like growth factor 1 95546_g_at X04480 0.44 Nidogen 2 93563_s_at AB017202 0.44 RIKEN cDNA 1100001H23 gene 98033_at AA710132 0.44 Serum deprivation response 160373_i_at AI839175 0.44 Activating transcription factor 5 103006_at AB012276 0.45 Adenylate kinase 1 96801_at AJ010108 0.45 Cyclin-dependent kinase inhibitor 1C 95471_at U22399 0.45 Procollagen, type V, alpha 2 92567_at L02918 0.45 Procollagen, type VI, alpha 3 101110_at AF064749 0.45 SRY-box containing gene 18 104408_s_at L35032 0.45 Von Willebrand factor homolog 103499_at AI843063 0.45 Adenylate cyclase 4 95308_at AA397054 0.46 Branched chain ketoacid dehydrogenase E1, beta polypeptide 102302_at L16992 0.46 Guanine nucleotide binding protein, alpha inhibiting 1 104412_at AI153412 0.46 Inhibitor of DNA binding 1 100050_at M31885 0.46 Mast cell protease 6 93622_at M57626 0.46 Methylcrotonoyl-Coenzyme A carboxylase 1 (alpha) 94940_at AW123316 0.46 Phosphorylase kinase alpha 2 104003_at AA822296 0.46 Potassium channel tetramerisation domain containing 17 101856_at AI836771 0.46 Protease, serine, 11 (Igf binding) 96920_at AW125478 0.46 Protocadherin gamma subfamily C, 3 94449_at AI854522 0.46 RAS, dexamethasone-induced 1 99032_at AF009246 0.46 Riken cDNA 6430556B02 gene 161436_s_at AV345565 0.46 Thrombomodulin 104601_at X14432 0.46 Zinc finger homeobox 1a 99052_at D76432 0.46

132 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb Flavin containing monooxygenase 1 101991_at D16215 0.47 Nicotinamide N-methyltransferase 101473_at U86108 0.47 Progressive ankylosis 100948_at AW049351 0.47 Ral GEF with PH domain and SH3 binding motif 1 103556_at AI840158 0.47 RNA binding motif, single stranded interacting protein 2 92517_at AB026583 0.47 Stearoyl-Coenzyme A desaturase 2 95758_at M26270 0.47 Aquaporin 1 93330_at L02914 0.48 ATP-binding cassette, sub-family C (CFTR/MRP), member 9 97172_s_at D86037 0.48 Cysteine dioxygenase 1, cytosolic 96346_at AI854020 0.48 Cytochrome P450, family 4, subfamily b, polypeptide 1 103353_f_at D50834 0.48 Diacylglycerol O-acyltransferase 1 104371_at AF078752 0.48 RIKEN cDNA 1700037H04 gene 97210_at AW048446 0.48 Secretogranin III 162237_f_at AV328553 0.48 Sterol carrier protein 2, liver 93278_at M91458 0.48 Zinc finger, CW-type with coiled-coil domain 2 161009_at AW123037 0.48 DNA segment, Chr 17, ERATO Doi 288, expressed 96785_at AF110520 0.49 Hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase 7 160104_at AA824102 0.49 Isovaleryl coenzyme A dehydrogenase 104153_at AW047743 0.49 Laminin, alpha 4 104587_at U69176 0.49 Monoglyceride lipase 97511_at AI846600 0.49 Phosphatidic acid phosphatase 2a 98508_s_at D84376 0.49 Phosphoglucomutase 2 104313_at AI842432 0.49 RIKEN cDNA 1110035L05 gene 95052_at AI839150 0.49 RIKEN cDNA 2410003P15 gene 103090_at AI838742 0.49 S100 calcium binding protein A6 (calcyclin) 92770_at X66449 0.49 Serine (or cysteine) proteinase inhibitor, clade A, member 3N 104374_at M64086 0.49 UDP-N-acetyl-alpha-D-galactosamine:polypeptide N- acetylgalactosaminyltransferase 2 97553_at AI841884 0.49 AE binding protein 1 100412_g_at AF053943 0.5 CD1d2 antigen 101897_g_at M63697 0.5 Enoyl Coenzyme A hydratase, short chain, 1, mitochondrial 95426_at AW048512 0.5 Glycogen synthase 3, brain 98496_at U53218 0.5 Myeloblastosis oncogene 92644_s_at M12848 0.5 NAD(P) dependent steroid dehydrogenase-like 98631_g_at AW106745 0.5 Procollagen, type VIII, alpha 1 100308_at X66976 0.5 Quaking 160726_at U44940 0.5 RIKEN cDNA 2610019F03 gene 100902_at AI846549 0.5 S100 calcium binding protein A10 (calpactin) 92539_at M16465 0.5 Thyroid hormone receptor alpha 92348_at AI841022 0.5 Uridine-cytidine kinase 1 94381_at L31783 0.5 CCAAT/enhancer binding protein (C/EBP), alpha 98447_at M62362 0.51 Complement component 1, q subcomponent, receptor 93454_at AF081789 0.51

133 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb 1" Elastin 92836_at AA919594 0.51 Glyoxalase 1 93269_at AI848952 0.51 Low density lipoprotein receptor 160832_at Z19521 0.51 Patatin-like phospholipase domain containing 2 96348_at AW121217 0.51 Phospholipid transfer protein 100927_at U28960 0.51 RIKEN cDNA 1810044O22 gene 103619_at AI850017 0.51 RIKEN cDNA 2610001E17 gene 160298_at AW122012 0.51 RIKEN cDNA 3110043O21 gene 100464_at AI840585 0.51 RIKEN cDNA 4930422J18 gene 161817_f_at AV376312 0.51 CD302 antigen 160387_at AI853900 0.52 Low density lipoprotein receptor-related protein 1 101073_at X67469 0.52 Paraoxonase 3 93940_at L76193 0.52 Actin-binding LIM protein 1 103574_at AI841606 0.53 ADP-ribosylation factor-like 2 binding protein 98084_at AI849834 0.53 CCR4 carbon catabolite repression 4-like 99535_at AW047630 0.53 Coronin, actin binding protein, 2B 97365_at AI848032 0.53 Dipeptidase 1 103644_at D13139 0.53 FMS-like tyrosine kinase 1 98452_at D88689 0.53 Forkhead box P1 104415_at AA833293 0.53 Glycerol-3-phosphate acyltransferase, mitochondrial 101867_at U11680 0.53 Lysosomal-associated protein transmembrane 4B 100571_at AW123934 0.53 Major histocompatibility complex, class I-related 101433_at AF010452 0.53 Oxysterol binding protein-like 11 94439_at AW124363 0.53 Reticulocalbin 1 160896_at D13003 0.53 RIKEN cDNA 8430420C20 gene 103916_at AI850713 0.53 Thrombospondin 2 94930_at L07803 0.53 Tyrosine kinase receptor 1 99936_at X80764 0.53 UDP-Gal:betaGlcNAc beta 1,4-galactosyltransferase, polypeptide 6 102936_at AW125314 0.53 Adipsin 99671_at X04673 0.54 Catalase 160479_at M29394 0.54 Cathepsin F 97336_at AJ131851 0.54 cDNA sequence BC013529 103580_at AI845588 0.54 Nidogen 1 100120_at L17324 0.54 Procollagen, type VI, alpha 1 95493_at X66405 0.54 RIKEN cDNA A630051L19 gene 104661_at AF031164 0.54 RIKEN cDNA B430218L07 gene 93467_at AI852734 0.54 Synaptopodin 94991_at AW046661 0.54 Vascular endothelial growth factor A 103520_at M95200 0.54 Zinc finger and BTB domain containing 20 94780_at AI987985 0.54 Apoptosis-associated tyrosine kinase 100994_at AF011908 0.55 Dual specificity phosphatase 1 104598_at X61940 0.55 Integrin alpha 161497_f_at AV093331 0.55 Membrane bound C2 domain containing protein 96767_at AF098633 0.55 Peroxisomal membrane protein 2 104098_at L28835 0.55

134 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb RIKEN cDNA 2310004I03 gene 96211_at AI846896 0.55 Adenosine A1 receptor 92374_at AW120691 0.56 Angiopoietin 2 92210_at AF004326 0.56 AU RNA binding protein/enoyl-coenzyme A hydratase 96650_at AI837724 0.56 Eukaryotic translation initiation factor 4E binding protein 1 100636_at U28656 0.56 Expressed sequence AI464131 104034_at AI846672 0.56 FK506 binding protein 9 93731_at AF090334 0.56 Gelsolin 93750_at J04953 0.56 Microfibrillar-associated protein 2 101095_at L23769 0.56 Reelin 96591_at U24703 0.56 Semaphorin 4A 94063_at X85991 0.56 Succinate dehydrogenase complex, subunit A, flavoprotein (Fp) 94080_at AI835718 0.56 7-dehydrocholesterol reductase 98989_at AF057368 0.57 Calpain 2 101040_at D38117 0.57 Fatty acid binding protein 5, epidermal 160544_at AJ223066 0.57 Interleukin 13 receptor, alpha 1 103723_at AA608387 0.57 Laminin B1 subunit 1 101948_at X05212 0.57 Lymphoblastomic leukemia 100468_g_at X57687 0.57 Myeloid-associated differentiation marker 96285_at AJ001616 0.57 Nuclear factor I/X 101930_at Y07688 0.57 Protein kinase C, delta binding protein 97496_f_at AW048944 0.57 RAB34, member of RAS oncogene family 160317_at AI835712 0.57 UDP glucuronosyltransferase 1 family, polypeptide A6 99580_s_at U16818 0.57 Coronin, actin binding protein 1C 98107_at AW123801 0.58 Glycogenin 1 162262_f_at AV357306 0.58 Hemochromatosis 104014_at Y12650 0.58 Homeo box A5 103086_at Y00208 0.58 Melanoma antigen, family D, 1 96703_at AB029448 0.58 Paraoxonase 2 104378_at L48514 0.58 RIKEN cDNA 1190005I06 gene 93143_at AI844196 0.58 Stromal interaction molecule 1 100952_at U47323 0.58 Dynamin 1 103031_g_at L31397 0.59 Rab6 interacting protein 1 104108_at AJ245569 0.59 RIKEN cDNA 2210419D22 gene 93877_at AI834777 0.59 Sphingomyelin phosphodiesterase 1, acid lysosomal 100099_at Z14132 0.59 Aldehyde dehydrogenase 2, mitochondrial 96057_at AI647493 0.6 Branched chain aminotransferase 2, mitochondrial 100443_at AF031467 0.6 CXXC finger 5 95701_at AW124069 0.6 Glutathione S-transferase, mu 1 93543_f_at J03952 0.6 Keratin complex 2, basic, gene 7 97920_at AA755126 0.6 Protein tyrosine phosphatase 4a3 160862_at AF035645 0.6 RIKEN cDNA 2310001H13 gene 96124_at AI835632 0.6 Secreted acidic cysteine rich glycoprotein 97160_at X04017 0.6 Amyloid beta (A4) precursor-like protein 2 93498_s_at M97216 0.61 DNA segment, Chr 10, ERATO Doi 214, expressed 94526_at AI848453 0.61

135 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb Enoyl coenzyme A hydratase 1, peroxisomal 93754_at AF030343 0.61 Gap junction membrane channel protein alpha 1 100064_f_at M63801 0.61 Inositol polyphosphate-5-phosphatase A 97858_at AW049190 0.61 Peptidase 4 101042_f_at U51014 0.61 SH3-binding domain glutamic acid-rich protein like 93806_at AI848671 0.61 Tenascin XB 102916_s_at AB010266 0.61 Vascular endothelial growth factor B 103001_at U43836 0.61 Adiponectin, C1Q and collagen domain containing 99104_at U49915 0.62 Benzodiazepine receptor, peripheral 93042_at D21207 0.62 Diaphanous homolog 1 160976_at AA222943 0.62 FK506 binding protein 10 99082_at L07063 0.62 Lymphocyte antigen 6 complex, locus C 93077_s_at D86232 0.62 Matrilin 2 98475_at U69262 0.62 P450 (cytochrome) oxidoreductase 99019_at D17571 0.62 Protein kinase, cAMP dependent, catalytic, alpha 103559_at M12303 0.62 RIKEN cDNA E430034L04 gene 101106_at AI853331 0.62 3-hydroxyisobutyrate dehydrogenase 97279_at AI837615 0.63 Adenylate cyclase 6 102321_at M93422 0.63 Amyloid beta (A4) precursor protein-binding, family B, member 1 interacting protein 102710_at AF020313 0.63 Bernardinelli-Seip congenital lipodystrophy 2 homolog 93080_at AF069954 0.63 Electron transferring flavoprotein, beta polypeptide 96947_at AW046273 0.63 Ephrin B2 160857_at U30244 0.63 Fasciculation and elongation protein zeta 2 101934_at AI851119 0.63 Growth arrest specific 6 99067_at X59846 0.63 Hairy/enhancer-of-split related with YRPW motif 1 95671_at AJ243895 0.63 Protein phosphatase 2, regulatory subunit B (B56), alpha isoform 93826_at AI956230 0.63 RAS p21 protein activator 4 160965_at AA163960 0.63 RIKEN cDNA E030024M05 gene 99366_at AI553536 0.63 SAR1a gene homolog 1 100635_at L20294 0.63 Selenium binding protein 1 100596_at M32032 0.63 Dehydrogenase/reductase (SDR family) member 7 95620_at AW120882 0.64 DNA segment, Chr 12, ERATO Doi 647, expressed 93775_at AI841894 0.64 DNA segment, Chr 15, ERATO Doi 366, expressed 94561_at AI836140 0.64 DNA segment, Chr 15, ERATO Doi 366, expressed 94561_at AI836140 0.64 Exostoses (multiple)-like 3 95591_at AI843335 0.64 Guanosine diphosphate (GDP) dissociation inhibitor 1 97313_at U07950 0.64 RAS p21 protein activator 3 93319_at U20238 0.64 RIKEN cDNA 1110038D17 gene 95119_at AA866888 0.64 RIKEN cDNA 2210402C18 gene 103560_at AW124401 0.64 Glyceronephosphate O-acyltransferase 94562_at AI843968 0.65 Inositol 1,4,5-trisphosphate 3-kinase B 104395_at AW121826 0.65 Lipin 1 98892_at AI846934 0.65 Riken cDNA 2410087O15 gene 161889_f_at AV102160 0.65 RIKEN cDNA 5730472N09 gene 160661_at AI840615 0.65 Transgelin 2 160162_at AI852545 0.65

136 Affymetrix GenBank Fold Gene Namea Accession Accession Changeb 3-monooxygenase/tryptophan 5-monooxygenase activation protein, gamma polypeptide 95716_at AW125041 0.66 Carnitine deficiency-associated gene expressed in ventricle 1 102215_at Y10495 0.66 F-box protein 45 93217_at AW049642 0.66 Guanine nucleotide binding protein, beta 5 100122_at L34290 0.66 Integrin beta 1 binding protein 1 100990_g_at AJ001373 0.66 Netrin 1 97977_at AA645293 0.66 Peroxisome biogenesis factor 19 95074_at AW125309 0.66 Progesterone receptor membrane component 1 101585_at AF042491 0.66 RIKEN cDNA 2610318G18 gene 96688_at AI845463 0.66 Sparc/osteonectin, cwcv and kazal-like domains proteoglycan 2 104375_at AI844853 0.66 RIKEN cDNA 2310042M24 gene 96649_at AW046552 0.67 Tetraspanin 6 92555_at AF053454 0.67 Annexin A11 102815_at U65986 0.69 Choline phosphotransferase 1 93994_at AW212131 0.69 Deltex 2 homolog 96818_at AA762522 0.69 Moesin 160308_at AI839417 0.69 RIKEN cDNA B430110G05 gene 160392_at AW060358 0.69 AHNAK nucleoprotein 160255_at AA657044 0.7 RIKEN cDNA 1500003O03 gene 96607_at AW124902 0.7 Ring finger protein 141 101939_at AI849344 0.7 S100 calcium binding protein A13 100959_at X99921 0.7 Fatty acid synthase 98575_at X13135 0.73 Aldehyde dehydrogenase family 7, member A1 97450_s_at AA986258 0.75 Succinate dehydrogenase complex, subunit B, iron sulfur (Ip) 95053_s_at AA674669 0.75 RIKEN cDNA 1810011K17 gene 104052_at AI840921 0.78 Quininoid dihydropteridine reductase 96948_at AI845337 0.83

aData were analyzed using Affymetrix MicroArray Suite 5.0. In order to be considered changed, genes must fulfill all of the following criteria: 1) be called

“present” by the Affymetrix algorithm in at least one sample, 2) be called changed

(increase, decrease, moderate increase, moderate decrease) by the Affymetrix algorithm in all comparisons (each wild-type compared to each transgenic sample), and 3) have a p-value of <0.05 when analyzed using a Student’s t-test

(wild type vs. transgenic). bFold change represents expression in transgenic

137 mammary glands relative to wild type and is derived from the Affymetrix signal log ratio (SLR).

138 Figure IV-1

Pre-neoplastic mammary glands and mammary tumors from LH- overexpressing mice display centrosome amplification. Pre-neoplastic mammary glands and mammary tumors were collected from LH-overexpressing mice and control mammary tissue was collected from age-matched wild-type mice. Tissue was dissociated and centrosomes were detected using an antibody for γ-tubulin (green). Nuclei are counterstained with DAPI (blue). While all cells from wild-type mammary tissue (A) demonstrated a normal complement of centrosomes (1 or 2 per cell), many cells from pre-neoplastic mammary glands

(B) and tumors (C) exhibited a significant increase in centrosome number. (D)

Quantification, expressed as mean ± SD, indicated that 1.9± 0.9% of pre- neoplastic cells and 1.6± 0.1% of tumor cells contained >2 centrosomes.

139 ACB

D 3 2.5

2

1.5

1

0.5

0

% cells with >2 centrosomes % cells 16 wk 16 wk >40 wk LH WT LH WT tumors n=6 n=5 n=5 n=5

140 Figure IV-2

A mutant form of p53 does not alter the latency of hormone-induced mammary tumorigenesis. LH-overexpressing mice were crossed with mice expressing p53R172H in the mammary gland under the control of the WAP promoter. Tumors were detected by weekly palpation. (A) Bi-transgenic animals

(n=12) developed mammary tumors with a latency that was not significantly different from LH-overexpressing mice (n=16). (B) Transgene expression in the mammary glands of bi-transgenic mice (p53/LH) was verified by northern blot analysis.

141 A.

Tumor Latency

100 X X X

80

60 Censored X WT X 40 LH p53/LH XX 20 Percent Tumor Free X 0 0 5 10 15 20 25 30 35 40 45 50 55 age (weeks)

B. p53 Expression: WT p53 LH p53/LH

p53

142 Figure IV-3 mRNA levels of regulators of p53 stability in mammary glands and tumors of LH-overexpressing mice. To determine whether p53 activity may be diminished by increases in Mdm2 and/or MdmX, both of which negatively regulate p53, RT-PCR was carried out on RNA from pre-neoplastic mammary glands, tumors, and appropriate age-matched wild-type controls. Mdm2 levels were not significantly changed between any of the experimental groups. MdmX mRNA levels were 1.5-fold increased in pre-neoplastic mammary gland from a

16 week old animal relative to age-matched wild-type controls. Tumors from LH- overexpressing mice, however, exhibited a 43% decrease in MdmX mRNA compared to age-matched wild-type mammary tissue. *indicates p<0.05 relative to age-matched wild-type control. §indicates p<0.05 relative to sample from 16 week old animal of the same genotype. n=3 for all groups

143 3

2.5 § 16 wk WT 2 * 16 wk LH 1.5 * § >40 wk WT 1 LH tumors

0.5

0 Fold to TBP Change Relative Mdm2 MdmX

144 Figure IV-4 p53 is activated in the mammary glands of wild-type and LH- overexpressing mice by ionizing radiation. To determine whether the signaling pathway upstream of p53 was intact in the mammary glands, wild-type and LH-overexpressing mice were exposed to ionizing radiation and mammary tissue was collected 2 or 6 hours later. (A) Western blot analysis shows that p53 is phosphorylated at serine 18 in mammary glands from irradiated wild-type and pre-neoplastic transgenic mice, as well as in tumors from LH-overexpressing mice, relative to non-irradiated controls (2 hours post-IR). (B) p21 mRNA levels are upregulated in the mammary glands of wild-type and pre-neoplastic transgenic mice 6 hours after exposure to radiation. mRNA levels were measured using ABI TaqMan RT-PCR. Values are expressed relative to TBP and are normalized to a wild-type, non-irradiated sample. *indicated p<0.05 relative to genotype matched non-irradiated control. (C) Northern blot analysis shows upregulation of Bax mRNA in the mammary glands of wild-type and LH- overexpressing mice 2 hours after exposure to radiation. The ethidium bromide stained gel is shown as a loading control.

145 WT LH LH tumors IR: -+-+-+-+-+-++ A. pSer15-p53

β-tubulin

B. 15 * 10 *

5

FoldRelative Change to TBP 0 WT WT LH LH IR: - + +-

WT LH C. IR: - - + + -+- +

Bax

28S

18S

146 Figure IV-5

Mammary epithelial cells of wild-type and LH-overexpressing animals

demonstrate variable apoptotic response to ionizing radiation. Adult wild-

type and transgenic animals were exposed to ionizing radiation and killed 6 hours later. Apoptosis was detected by TUNEL assay. TUNEL-positive cells are red; nuclei are counterstained with DAPI (blue). Autofluorescence of mammary tissue and red blood cells appears in green. Non-irradiated wild-type (WT) and transgenic (LH) mice have low levels of apoptosis in mammary epithelial cells

(MEC) (A, G). Mammary epithelial cells of both genotypes exhibit a highly variable rate of apoptosis upon exposure to ionizing radiation. Wild-type mammary glands with high (C) and low (E) induction of apoptosis upon exposure to ionizing radiation display similar induction of apoptosis in the lymph node (D,

F). Mammary glands from LH-overexpressing mice with high (H) and low (I) induction of apoptosis upon exposure to radiation (lymph node images not shown). Quantification of apoptosis of mammary epithelial cells is shown in (J).

There is no difference in apoptosis induction between wild-type and LH- overexpressing animals (data analyzed using student’s t-test). Apoptotic response in wild-type mice does not correlate with stage of the estrous cycle

(red=estrus, purple=metestrus, blue=diestrus)

147 WT-MEC WT-lymph node AB non IR

CD

IR

EF

IR

148 LH-non-IR LH-IR G H

I

J 14 12

10 estrus 8 metestrus ditestrus 6 4

2 % TUNEL positive 0 WT LHWT LH IR: --++ n= 9 313 7

149 CHAPTER V

SUMMARY, FUTURE DIRECTIONS, AND CONCLUSIONS

SUMMARY

The goal of my thesis project has been to elucidate mechanisms of

hormone-induced mammary tumorigenesis. Contributions in this area should eventually lead to improved approaches for prevention and treatment of breast cancer, decreasing the social and individual burden of this disease. Most of the work presented in Chapters III and IV of this dissertation were made possible by the initial characterization of the LH-overexpressing mouse as a unique model in which to study hormone-induced mammary tumorigenesis (Chapter II, (190)).

Despite extensive, long-standing evidence that reproductive hormones make significant contributions to breast cancer, at the time of its publication this work

represented only the second example of a transgenic mouse model of hormone- induced mammary cancer. The earlier model overexpressed supra-physiological levels of prolactin from the liver, resulting in mammary tumors with an extended latency (37). The fact that LH-overexpressing mice develop spontaneous

mammary tumors reinforces the powerful impact that hormones can have on the

mammary gland, even in the absence of exposure to a carcinogen or forced

expression of an oncogene. This model will facilitate future exploration of

treatments for hormone-induced breast cancer and elucidation of the

mechanisms that drive this disease, both in the Keri laboratory and in other

150 laboratories with related research endeavors. Examples of both of these

applications can be found in this dissertation.

The studies presented in Chapter III (240) illustrate the utility of the LH-

overexpressing mouse model for testing chemopreventive and chemotherapeutic

agents. Although several lines of epidemiological and experimental evidence

have implicated the vitamin D signaling pathway in breast cancer progression,

the work presented in Chapter III of this thesis is the first time that the efficacy of

a VDR agonist has been investigated in a mouse model of spontaneous

mammary cancer. The ability of the VDR agonist EB1089 to reduce the

proliferative potential of the mammary glands of LH-overexpressing mice

supports the notion that such agonists have potential as chemopreventive agents. Moreover, the chemotherapeutic promise of VDR agonists is promoted

by the positive response observed in a subset of hormone-induced mammary

tumors treated with EB1089. These data support the further investigation of the

VDR signaling pathway in breast cancer. Importantly, the results generated by

these studies also validate the LH-overexpressing mouse as a worthy model for

investigation of other potential therapeutics.

Although testing potential therapeutic agents is an important aspect of

fulfilling the goal to decrease the incidence and mortality of breast cancer, it is

also important to elucidate mechanisms of this disease in order to facilitate the development of rational, innovative approaches for prevention and treatment. In this regard, the data presented in Chapter IV provide insight into the hormone- induced molecular and cellular changes that take place in the mammary gland

151 and correlate with increased susceptibility to cancer development. The pre- neoplastic mammary glands and mammary tumors of LH-overexpressing mice demonstrate centrosome amplification, suggesting that genomic instability may be a characteristic of hormone-induced mammary cancer. Of note, centrosome amplification in the mammary glands of LH-overexpressing animals occurs in the presence of wild-type p53. The majority of human breast cancers also contain wild-type p53 in conjunction with centrosome amplification (141), suggesting that mechanisms of tumor development in this mouse model may be similar to those involved in the progression of human disease. Utilization of the LH- overexpressing mice will assist in determining the p53-independent changes that allow for centrosome amplification and the hormonal contributions to these changes. Identification of such alterations is essential, because, although hormonal contributions to breast cancer are well established, the means by which hormones impact risk are unknown; enhanced understanding in this area would facilitate targeted disruption of harmful hormone activity.

The work presented in Chapters II, III, and IV has increased understanding of the hormonal component of mammary cancer, largely through utilization of a transgenic mouse model of this disease; however, the results of these studies have generated a new series of questions to be addressed by further research. In this regard, the following sections of this chapter will present data generated by preliminary experiments and propose future experiments that can be done to augment knowledge of the hormonal basis of breast cancer.

152 Materials and Methods

Animals

Mice were housed in microisolator plus units with a 12 hour light/dark cycle and

given food and water ad libitum. Wild-type CF-1 mice were obtained from

Charles River Laboratories (Wilmington, MA). Unless otherwise specified, during

treatment with EB1089 or vehicle, all mice were maintained on a 0.1% low-

calcium diet (BioServe, Frenchtown, NJ) to avoid the potential development of

hypercalcemia. All mouse studies were approved by the Institutional Animal Care

and Use Committee at Case Western Reserve University.

Tamoxifen treatment of mice during development

Peripubertal LH-overexpressing and wild-type female mice were given

daily subcutaneous injections of tamoxifen (150 μg in sterile PBS; Sigma T5648)

or vehicle. Treatment was initiated at 20 days of age and ended at 5 weeks of

age, at which time the animal was killed and mammary glands collected.

EB1089 treatment of mice during development

EB1089 was obtained from Leo Pharmaceutical Products (Ballerup,

Denmark). Peripubertal LH-overexpressing and wild-type female mice were given intraperitoneal injections of EB1089 (25 ng in safflower oil) or vehicle every 48 hours. Treatment was initiated at 20 days of age and ended at 5 weeks of age, at

which time the animal was killed and mammary glands collected.

EB1089 treatment of adult mice

Beginning at 10 weeks of age, LH-overexpressing and wild-type mice

were given intraperitoneal injections of EB1089 (50 ng in safflower oil) or vehicle

153 every 48 hours for three weeks. Mammary glands were assessed by whole

mount.

PRELIMINARY RESULTS AND FUTURE STUDIES

Prospects for Chemoprevention Studies

Tamoxifen and Hormone-Induced Mammary Gland Pathology

Estrogen is one of the primary hormones responsible for mammary gland

development and has been strongly implicated in breast cancer incidence and

progression. The primary role for estrogen in mammary gland development

appears to be in ductal elongation, as mice lacking the estrogen receptor display

severely stunted ductal development (54). In humans, the contributions of

estrogens to breast cancer are evidenced by the correlation between increased circulating levels of estradiol and breast cancer risk in post-menopausal women

(22), as well as the effectiveness of therapies that interfere with estrogen

signaling. One example of the latter is tamoxifen, a selective estrogen receptor modulator (SERM) that is an estrogen receptor agonist in bone, liver and the cardiovascular system; an antagonist in the breast and brain; and displays mixed activity in the uterus (299). Use of tamoxifen in the clinic has proved extremely beneficial for women with ER-positive breast cancer. Women with ER-positive tumors treated with tamoxifen for 5 years demonstrated a 50% decrease in disease recurrence and a 28% decrease in mortality (43). Moreover, prophylactic treatment of women at high risk of developing breast cancer results in an approximately 50% decrease in risk of both invasive and non-invasive ER-

154 positive breast tumors (44). Given the elevated levels of 17β-estradiol in LH-

overexpressing mice (Table II-1), it was of interest to determine the effects of

tamoxifen on the mammary pathology of this mouse model of hormone-induced

mammary cancer. Serum measurements indicate that 17β-estradiol levels are

elevated as early as 5 weeks of age (Table II-1), a fact that is illustrated by the

increased ductal elongation observed in peri-pubertal mice (Figure II-1). To

ascertain the role of estrogen on early mammary gland development, wild-type

and LH-overexpressing animals were treated with tamoxifen from 20 days of age

to 5 weeks of age. Wild-type mice treated with tamoxifen displayed decreased

ductal elongation (Figure V-1A), consistent with the known role of estrogen in the

mammary gland. In contrast, the ductal elongation of transgenic mice treated

with tamoxifen was not substantially different from that of vehicle treated animals;

however, qualitative assessment suggests that less branching and formation of

fewer alveolar buds occurs in the mammary glands of LH-overexpressing mice in

the presence of tamoxifen (Figure V-1B). This was somewhat surprising, as

secondary branching and the formation of alveolar buds are functions

traditionally assigned to progesterone and prolactin (57,61). This phenotype may

reflect the loss of transcriptional upregulation of progesterone receptor by

estrogen (300-302), which could be determined by immunohistochemistry for PR

in these mammary glands.

The lack of estrogen receptor expression in established mammary tumors

of LH-overexpressing mice (Figure II-8) makes it unlikely that they would respond to tamoxifen treatment; however, there is evidence that, unlike in humans,

155 tamoxifen can inhibit the formation of ER-negative mammary tumors in mice

(113). This suggests that hormone-responsive tumors in humans and hormone-

independent tumors in mice develop by similar mechanisms, supporting the use

of such mouse models for studying early events in mammary tumorigenesis. To determine whether tamoxifen can inhibit mammary tumor development in LH- overexpressing mice, slow-release pellets containing tamoxifen or vehicle could be inserted subcutaneously at 20 days of age (before development of hyperplasia) or at 8 weeks of age (after hyperplasia is established) and mammary tumor formation would be monitored by weekly palpation. If tamoxifen treatment results in failure to develop mammary tumors or extends tumor latency, this would suggest that signaling by estrogen is important for tumor formation in this model. Although this may be due to direct effects on the mammary gland, it is also possible that tamoxifen treatment will reduce circulating levels of prolactin or alter other endocrine signaling pathways. Given the developmental phenotype observed in LH-overexpressing mice treated with tamoxifen (Figure V-1B), it is also possible that tamoxifen treatment will inhibit the formation of specific, tumor- susceptible cell types in the mammary gland. This may or may not be due to alterations in the progesterone receptor. To investigate these possibilities, serum prolactin levels should be measured and morphological analysis as well as assessment of PR expression should be performed on mammary glands that have been subjected to long-term tamoxifen treatment.

156 The Role of Vitamin D Receptor in Mammary Cancer

The protective role of vitamin D signaling in breast cancer has been

suggested by several lines of epidemiological evidence as well as numerous in vitro studies. These data suggest that activation of the vitamin D receptor through environmental, dietary, or pharmacological means may provide protection against breast cancer development or inhibit its progression. Until recently, however, the efficacy of VDR agonists had not been investigated in a spontaneous in vivo model of mammary cancer. Data presented in this dissertation revealed that the vitamin D receptor is upregulated in the hyperplastic mammary glands and tumors of LH-overexpressing animals.

Upregulation of VDR is also observed in human breast cancers (78), suggesting

that it may be a useful therapeutic target. The notion that VDR may be an

effective target for chemoprevention is supported by the fact that treatment with

the VDR agonist EB1089 reduced the proliferative rate of mammary epithelial

cells in the LH-overexpressing mouse by 35% (Figure III-2). This effect suggests

that exposure to EB1089 may reduce the risk of mammary tumor development in

this model, and possibly in humans. As LH-overexpressing animals demonstrate

hyperplasia of the mammary gland prior to tumorigenesis, we assessed whether

the growth inhibitory effect of EB1089 is sufficient to reverse or inhibit this

hyperplasia.

Mammary gland hyperplasia in LH-overexpressing mice is dependent on

ovarian factors (Figure II-5; (190)). VDR has been shown to inhibit the effects of

ovarian and lactogenic hormones on the mammary gland (79); thus, we

157 hypothesized that activation of the vitamin D receptor with EB1089 may interfere

with the proliferative effects of the ovarian and lactogenic hormones, resulting in

reversal of mammary gland hyperplasia. To test this hypothesis, LH-

overexpressing animals were aged to 10 weeks of age, by which time they

develop extensive mammary gland hyperplasia, and then treated with EB1089

for three weeks. Although this treatment paradigm was shown to induce

hypercalcemia in LH-overexpressing animals in a parallel experiment (data not

shown), verifying the activity of EB1089, whole mount analysis revealed

mammary gland hyperplasia in these mice was not reversed (Figure V-2).

Mammary gland morphology was also unchanged in wild-type mice treated with

EB1089 compared to those treated with vehicle (Figure V-2). These data indicate

that pharmacological activation of VDR is unable to reverse the effects of

mammogenic hormones on the mammary gland, whether they are present at

normal or elevated levels. One possible explanation is that, while VDR activation

is able to reduce proliferation, it is not sufficient to reverse previously established

hyperplasia.

To determine whether exposure to EB1089 is sufficient to prevent hormone-induced mammary gland hyperplasia, wild-type and LH-overexpressing mice were treated with EB1089 or vehicle beginning at 20 days of age. The initiation of this study precedes the commencement of post-natal mammary gland development and the appearance of mammary gland hyperplasia in LH- overexpressing mice. Mice were killed at 5 weeks of age and mammary gland development was assessed by whole mount (Figure V-3). Analysis of the extent

158 of ductal elongation and branching, activities ascribed to estrogen and progesterone, respectively, did not reveal a significant difference between animals treated with EB1089 and those treated with vehicle. This indicates that

EB1089 is not sufficient to inhibit hormone-mediated peri-natal mammary gland development, in the presence of either normal physiological or super- physiological levels of hormones. Failure of EB1089 to reverse or inhibit hyperplasia is somewhat surprising because studies using the VDRKO mouse showed that VDR limits mammary gland hyperplasia induced by estrogen, progesterone, and lactogenic hormones (79). One possible explanation for this apparent discrepancy is that ligand is not rate-limiting in the pathway by which

VDR inhibits hormone-induced hyperplasia. Even with the observed increase in

VDR in the mammary glands of LH-overexpressing mice (Figure III-1), limiting levels of receptor and/or coregulators may prevent maximal anti-hormone activity. While it is clear that EB1089 is sufficient to reduce mammary gland proliferation (Figure III-2), the modest reduction in cell growth (35%) is not adequate to inhibit or reverse hyperplasia to an extent that is evident by gross morphological analysis.

Despite the inability of short-term EB1089 to inhibit or reverse mammary gland hyperplasia in LH-overexpressing mice, the decreased proliferative rate observed in young transgenic animals suggests that activation of VDR may be chemopreventive. To directly test this hypothesis, LH-overexpressing animals could be treated with EB1089 over a long period of time to determine if activation of VDR impacts tumor incidence or latency. The most efficient form of drug

159 delivery over an extended period of time is through use of slow release pellets

that can be implanted under the skin of the animal. Pellets could be implanted

pre-pubertally or in adult LH-overexpressing mice with fully developed mammary

glands. Initially, EB1089-releasing pellets would be maintained throughout the

remainder of the life of the animal; however, if protective effects were observed, it

would be interesting to determine whether exposure during a specific period of

time, such as puberty, is sufficient to impart protection. The results of these

studies could have important implications for human breast cancer prevention.

Protection conveyed by VDR agonists would minimally reinforce the need for

optimization of vitamin D3 exposure and would provide support for clinical trials in

which a VDR agonist is administered to individuals at high risk for developing breast cancer. Chemoprevention could additionally be assessed in combination with tamoxifen.

Although EB1089 is not sufficient to inhibit or reverse mammary gland

hyperplasia in the LH-overexpressing mice, it was able to inhibit growth of a

subset of the spontaneous tumors that develop in these glands (Figure III-3).

Most likely this effect is not due to modulation of mammogenic hormone

signaling, because these tumors are largely devoid of steroid hormone receptors

and continue to grow in the absence of the ovary (Figure II-8, data not shown).

The factors that determine responsiveness to EB1089 are unknown; however,

the LH-overexpressing mouse may provide a useful tool for identifying response

predictors. This would require large scale characterization of pre-treatment

tumors and subsequent correlation with EB1089 response. Small portions of

160 tumors could be removed prior to the onset of treatment and analyzed using

high-throughput genomic or proteomic techniques. The mouse containing the

remaining portion of the tumor would be treated with EB1089 and tumor growth rate would be monitored. It is possible that all tumors that decrease their growth rate in response to EB1089 do so by a common mechanism; in this case, data generated by high-throughput approaches could be pared down by identification of genes or proteins that are differentially expressed or modified in all responsive tumors. However, it is also quite possible that multiple mechanisms of EB1089 response are employed by mammary tumors of LH-overexpressing mice.

Multiple mechanisms or predictors would be better identified by hierarchical clustering, which would group together samples with similar patterns of expression, perhaps reflecting similar mechanisms of responsiveness or resistance.

Utilization of different mouse models of mammary cancer would also

advance understanding of mechanisms of EB1089-mediated tumor inhibition

and/or regression. The well characterized MMTV-neu model demonstrates

decreased tumor latency and increased tumor incidence in a heterozygous

VDRKO background, demonstrating a role for VDR in neu-induced tumorigenesis

(84); however, the efficacy of VDR agonists such as EB1089 has not been tested

in this model. While tumors of both LH-overexpressing and MMTV-neu mice are

ER-negative, mammary tumors induced in transgenic mice that express prolactin

in the mammary gland under the control of the hormone-nonresponsive neu-

related lipoprotein promoter (NRL-prolactin) can be either ER-positive or ER-

161 negative (38). Testing EB1089 in this model may allow for dissection of mechanisms that are both hormone dependent and independent. Use of a model with ER-positive tumors would also allow for assessment of EB1089 treatment in conjunction with tamoxifen, perhaps providing insight into the potential of such combination therapies in human trials. Molecular comparison of tumors from

MMTV-neu or NRL-prolactin mice that are responsive and/or non-responsive to

EB1089 to those from LH-overexpressing mice would provide further insight into the mechanism(s) of VDR-mediated anti-tumorigenic signaling.

Prospects for Mechanistic Studies

Centrosome Amplification in Hormone-Induced Mammary Cancer

The studies reported in Chapter IV of this thesis show the existence of centrosome amplification in pre-neoplastic mammary glands and tumors of LH- overexpressing mice (Figure IV-1); however, it is not known whether centrosome amplification is actually driving mammary tumorigenesis in this model. The mechanism by which centrosome amplification is thought to cause tumorigenesis is through generation of aneuploidy as a consequence of multi-polar spindle formation. Thus, existence of aneuploidy in the mammary tumor cells of LH- overexpressing mice would be consistent with the involvement of centrosome amplification in tumor initiation. One method for determining genomic abnormalities is comparative genomic hybridization (CGH), which involves labeling of tumor DNA and normal reference DNA with two different fluorescent dyes, followed by co-hybridization to normal reference DNA. Differences in

162 hybridization reveal locations of copy number changes within the genome. This

approach provides benefits over more traditional approaches, such as spectral

karyotyping (SKY), because it does not require the use of metaphase

chromosomes, which can be difficult to obtain from tissue. The availability of

genome arrays for CGH is also increasing. This platform carries the promise of

increased resolution as well as relative technical simplicity.

If mammary tumors from LH-overexpressing mice are aneuploid, the next

step would be to determine whether they exhibit chromosomal instability (CIN).

While aneuploidy describes the state of having a non-diploid karyotype, CIN

refers to the active rate of change in chromosome population. CIN has been correlated with centrosome amplification in pre-invasive human carcinoma in situ

(168). Determination of CIN involves analysis of the chromosomal content of

many cells within a tumor. This can be done on a large scale using SKY if sufficient metaphase spreads are obtainable, but it can also be performed by

FISH (fluorescent in situ hybridization) using a smaller number of probes. CIN is calculated as the percent of cells with non-modal signals for a particular chromosome or chromosome region (163) and indicates that the karyotype of the cell is in a constant state of flux. This is an important tumor characteristic as it is

thought to facilitate selection of genotypes that have increased growth capacity

and invasive abilities (303).

As centrosome amplification is observed in a relatively small number of

cells in the mammary glands of LH-overexpressing mice, determination of

cellular characteristics that typify the presence of amplified centrosomes would

163 require isolation of cells with more than two centrosomes. Fluorescence activated

cell sorting (FACS) analysis could be used to separate cells with normal centrosome content from those containing three or more centrosomes. This could be done by labeling cells with a fluorescent-conjugated antibody for γ-

tubulin, a component of pericentriolar material. Cells could also be labeled with a

luminal epithelial cell marker, such as cytokeratin 19, to ensure comparison of

like cell types. Once populations of both cell types were obtained, gene expression profiling could be performed to identify differences in the transcriptome that correlate with the presence of centrosome amplification. There

have been several genes that have been linked to centrosome amplification. For example, Aurora A (181), cyclin E (304), and mutant forms of BRCA1 (173) and

BRCA2 (174) have all been shown to induce centrosome amplification; therefore, it is possible that we would observe alterations in the expression these genes, but it is also likely that we will identify additional genes involved in centrosome amplification including some genes that are regulated by mammogenic hormones.

The ability to isolate cells with centrosome amplification would also allow more direct assessment of the tumorigenic potential of these cells. These cells could be injected into the cleared fat pad of a syngeneic or nude mouse to determine whether they will generate a tumor. Cells with normal centrosome number could be transplanted into the contralateral fat pad of the same mouse for an internal control. Transplants could be introduced into either wild-type or

LH-overexpressing hosts. Tumor outgrowth in a wild-type host would suggest

164 that the cells had achieved commitment to tumorigenesis and no longer needed

the support of the altered hormonal milieu of the LH-overexpressing mouse.

Alternatively, hormonal input may still be necessary. Regardless, if centrosome

amplification is a sign of progression to tumorigenesis, cells with amplified

centrosomes should form tumors more quickly than cells with normal

centrosomes, which may not possess tumorigenic ability at all.

p53 in Hormone-Induced Mammary Cancer

MdmX: a Potential Negative Regulator of p53 Signaling in Early Stages of

Hormone-Induced Mammary Cancer

Although the sequence of p53 was found to be intact in the mammary

tumors of LH-overexpressing mice, RT-PCR analysis of MdmX mRNA showed

that it is 1.5-fold upregulated in the pre-neoplastic mammary glands of these

mice (Figure IV-3), providing a potential mechanism by which p53 activity may be compromised. The observed increase in MdmX gene expression is modest and

RT-PCR does not identify the cells in which this increase is taking place. In situ

hybridization may reveal whether the observed increase in MdmX mRNA is

occurring in the epithelial cells of the mammary gland. Inhibition of p53 activity

would require that the MdmX protein is correspondingly upregulated; thus,

western blot or immunohistochemistry should be performed to determine whether

excess MdmX protein is available for sequestering p53. Completion of these

studies hinges on the availability of a specific antibody, which we have not

identified at this time.

165 Little is known about the regulation of MdmX levels; in fact, there are no published reports describing the MdmX promoter. Upregulation in the mammary glands of the LH-overexpressing mouse suggests a role for mammogenic hormones. To determine the hormone responsiveness of the MdmX promoter, it could be cloned and linked to a reporter gene to facilitate analysis in cell culture.

Based on the hormonal profile of the LH-overexpressing mice, estrogen, progesterone, and prolactin would be the first potential regulators to be tested.

Responsiveness of the MdmX promoter to these hormones could be assessed in

MCF-7, T47D, and HC11 cells respectively. Interestingly, sequence analysis of a portion of the MdmX promoter reveals several potential Sp1 and Ap1 binding sites, as well as a potential estrogen receptor binding site.

p53 Target Genes

Two p53 target genes, p21 and Bax, were upregulated in response to

ionizing radiation in the mammary glands of both wild-type and LH-

overexpressing mice (Figure IV-4). This suggests that p53 is functionally

activated in the pre-neoplastic mammary glands of LH-overexpressing mice.

Although the induction of p21 in irradiated wild-type mammary glands (12.5-fold)

was not statistically different from the induction in irradiated transgenic glands

(7.1-fold), there is a trend of lower induction in transgenic glands. Addition of

more animals to this study would clarify whether or not a difference in p21 induction in response to ionizing radiation exists. p21 is a cyclin-dependent kinase inhibitor that is responsible for inducing cell cycle arrest in response to

166 p53 activation. In addition to verifying its transcriptional upregulation in the mammary glands of LH-overexpressing mice, it would be of interest to determine whether cell cycle arrest is occurring. Estrogen treatment of MCF-7 breast cancer cells alters the ability of p21 to associate with, and thus inhibit, the cyclin E-cdk2 complex without changing expression levels of p21 (305,306); hence, even with equal levels of p21 protein, the altered hormonal milieu of the LH-overexpressing animals may prevent p21 mediated cell cycle arrest. Failure to undergo cell cycle arrest when necessary could allow for propagation of cells at high risk of undergoing transformation.

Although both p21 and Bax appear to be upregulated in transgenic mammary glands upon exposure to radiation, it is possible that p53 regulation of other target genes is compromised, attenuating p53 function and eventually allowing for centrosome amplification and tumor formation. Extensive work has been done to identify both direct and indirect p53 target genes (307-314); thus it would be possible to perform large-scale analysis of a diverse array of p53 target genes to generate a more complete comparison of changes in p53 target genes that occur upon irradiation of wild-type and transgenic mammary glands. There are several platforms that would facilitate such a comparison. Large scale analysis could be performed using Affymetrix GeneChips, which currently probe up to 39,000 murine transcripts and variants. A more directed approach could be designed using TaqMan Low Density Arrays from Applied Biosystems, which accommodate scrutiny of up to 128 genes.

167 Cell Survival in the Mammary Gland

While IR-induced induction of apoptosis was variable in the mammary glands of both wild-type and LH-overexpressing mice (Figure IV-5), the nature of the variability observed in wild-type mice is particularly intriguing. Mammary glands of transgenic mice display a range of apoptosis induction, but wild-type mice appear to exhibit a bimodal response to ionizing radiation; the mammary glands either have extensive induction of apoptosis or no induction at all. There are at least two explanations for the variation observed in wild-type mice. First, the mammary epithelial cells in the non-apoptotic glands may be preferentially activating a cell cycle arrest program. It is well documented that p53 can activate either apoptosis, through target genes such as Bax (315); Noxa (316); and Puma

(317), or cell cycle arrest through p21 (130) or 14-3-3σ (131). To determine whether this is indeed the case, analysis of proliferation using a marker such as phospho-histone H3 (318) could be performed. A second possibility is that the mammary epithelial cells in the non-apoptotic glands have implemented a program by which they can bypass p53-mediated apoptosis and cell cycle arrest.

This trait would be very conducive to carcinogenesis, as cells with damaged DNA would be permitted to proliferate. In either case, identifying the biological characteristics that determine the response exhibited by the mammary gland upon exposure to ionizing radiation would be informative from a carcinogenesis perspective and may eventually provide the groundwork for optimization of chemotherapy.

168 In order to ascertain the factors which determine the response of the mammary gland to ionizing radiation, large-scale experimental approaches

should be utilized. Insight could be gained from identifying basal differences

between mammary glands (before they are exposed to ionizing radiation) as well

as differences in response to IR. To achieve this, one or two mammary glands

(the mouse has ten) could be surgically removed just before the mouse is

subjected to radiation, allowing each mouse to act as its own control. A small

amount of blood could also be collected at this time for baseline measurements

of serum hormone levels. Mice would be killed 6 hours after exposure to radiation

(or sham irradiation) and the remaining mammary glands as well as cardiac

blood collected at that time. Collected mammary tissue could be subjected to

several analyses. TUNEL staining of post-IR tissue would indicate whether or not

the mammary glands are susceptible to IR-induced apoptosis. RNA and protein

could be isolated from both pre- and post-IR tissue. RNA could be used to

perform large scale gene expression profiling, while protein could be subjected to

two-dimensional PAGE analysis to identify changes in expression and/or post-

translational modification taking place at the protein level. Differences identified

through both of these approaches should help to identify systemic, local and

intrinsic factors that influence the response of the mammary gland to an insult

like ionizing radiation and generate considerable insight into mechanisms that

regulate mammary cancer risk.

169 CONCLUSIONS

The body of work presented in this dissertation describes the characterization of a transgenic mouse model, the LH-overexpressing mouse, which develops mammary gland hyperplasia and tumors in response to ovarian hyperstimulation.

This is one of the few mouse models that reflects the essential contribution of hormones to breast cancer risk and progression. The studies reported here illustrate the utility of this model for investigating potential therapeutic targets and identifying mechanisms of hormone-induced mammary cancer. Continued research on the mammary gland pathology of the LH-overexpressing mouse should advance understanding of the hormonal component of breast cancer and eventually spur improvements in chemoprevention and chemotherapy.

170 Figure V-1

Tamoxifen inhibits ductal branching and alveolar bud formation in LH-

overexpressing mice. Wild-type and LH-overexpressing mice were treated with

tamoxifen from 20 days of age to 5 weeks of age and mammary glands were

collected for whole mount analysis. (A) Wild-type mice treated with tamoxifen

(n=3) exhibit decreased ductal elongation compared to vehicle treated controls

(n=2). The leading edges of ductal growth are indicated by the closed arrows. (B)

LH-overexpressing mice treated with tamoxifen (n=3) exhibit similar ductal

extension, but decreased secondary branching and alveolar bud formation than vehicle treated control mice (n=3). Open arrows point to examples of alveolar buds. LN=lymph node

171 A. Wild Type LN

LN vehicle

5X 10X

LN LN tamoxifen

5X 10X

B. LH-overexpressing

LN LN vehicle

5X 10X LN

LN tamoxifen 5X 10X

172 Figure V-2

Hormone-induced mammary gland hyperplasia persists in the presence of

EB1089. Beginning at 10 weeks of age, wild-type and LH-overexpressing mice

were treated with EB1089 for three weeks and then mammary gland morphology

was assessed by whole mount. (A) EB1089 does not induce any significant morphological change in wild-type mammary glands. (B) Mammary gland

hyperplasia in LH-overexpressing is unchanged in the presence of EB1089.

Number of animals in each treatment group is indicated beside picture.

LN=lymph node

173 A. Wild Type

vehicle LN n=5

EB1089 LN n=4

B. LH-overexpressing

vehicle LN n=3

EB1089 LN n=4

174 Figure V-3

EB1089 fails to inhibit the formation of hormone-induced mammary gland hyperplasia. Wild-type and LH-overexpressing mice were treated with EB1089 for 2 weeks beginning at 20 days of age. Mammary gland development was assessed by whole mount. Mammary gland development was not altered by

EB1089 in wild-type (A) or LH-overexpressing mice (B). Number of animals in each treatment group is indicated beside picture. LN=lymph node

175 A. Wild Type

vehicle LN n=5

EB1089 LN n=6

B. LH-overexpressing

vehicle LN n=5

EB1089 LN n=6

176 Reference List

(1) Ries LAG, Eisner MP, Kosary CL et al. SEER Cancer Statistics Review, 1975-2002 (http://seer.cancer.gov/csr/1975_2002/) based on November 2004 SEER data submission, posted to the SEER web site 2005. National Cancer Institute Bethesda, MD. 2005.

(2) Cancer Progress Report - 2003 Update, National Cancer Institute, NIH, DHHS, Bethesda, MD, February 2004, http://progressreport.cancer.gov/. National Cancer Institute, NIH, DHHS. 2005.

(3) Pharoah PD, Day NE, Duffy S et al. Family history and the risk of breast cancer: a systematic review and meta-analysis. Int J Cancer. 1997;71(5):800-809.

(4) Ford D, Easton DF, Stratton M et al. Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. The Breast Cancer Linkage Consortium. Am J Hum Genet. 1998;62(3):676-689.

(5) Easton DF, Ford D, Bishop DT. Breast and ovarian cancer incidence in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Am J Hum Genet. 1995;56(1):265-271.

(6) Ford D, Easton DF, Bishop DT et al. Risks of cancer in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Lancet. 1994;343(8899):692- 695.

(7) Ford D, Easton DF, Peto J. Estimates of the gene frequency of BRCA1 and its contribution to breast and ovarian cancer incidence. Am J Hum Genet. 1995;57(6):1457-1462.

(8) Vainio H, Kaaks R, Bianchini F. Weight control and physical activity in cancer prevention: international evaluation of the evidence. Eur J Cancer Prev. 2002;11 Suppl 2:S94-100.

(9) McTiernan A. Behavioral risk factors in breast cancer: can risk be modified? Oncologist. 2003;8(4):326-334.

(10) Hsieh CC, Trichopoulos D, Katsouyanni K, Yuasa S. Age at menarche, age at menopause, height and obesity as risk factors for breast cancer: associations and interactions in an international case-control study. Int J Cancer. 1990;46(5):796-800.

(11) Kvale G, Heuch I. Menstrual factors and breast cancer risk. Cancer. 1988;62(8):1625-1631.

177 (12) Brinton LA, Schairer C, Hoover RN, Fraumeni JF, Jr. Menstrual factors and risk of breast cancer. Cancer Invest. 1988;6(3):245-254.

(13) Irwin KL, Lee NC, Peterson HB et al. Hysterectomy, tubal sterilization, and the risk of breast cancer. Am J Epidemiol. 1988;127(6):1192-1201.

(14) Lambe M, Hsieh C, Trichopoulos D et al. Transient increase in the risk of breast cancer after giving birth. N Engl J Med. 1994;331(1):5-9.

(15) MacMahon B, Purde M, Cramer D, Hint E. Association of breast cancer risk with age at first and subsequent births: a study in the population of the Estonian Republic. J Natl Cancer Inst. 1982;69(5):1035-1038.

(16) Rao DN, Ganesh B, Desai PB. Role of reproductive factors in breast cancer in a low-risk area: a case-control study. Br J Cancer. 1994;70(1):129-132.

(17) Chodosh LA, D'Cruz CM, Gardner HP et al. Mammary gland development, reproductive history, and breast cancer risk. Cancer Res. 1999;59(7 Suppl):1765-1771s.

(18) Ginger MR, Gonzalez-Rimbau MF, Gay JP, Rosen JM. Persistent changes in gene expression induced by estrogen and progesterone in the rat mammary gland. Mol Endocrinol. 2001;15(11):1993-2009.

(19) Thordarson G, Jin E, Guzman RC et al. Refractoriness to mammary tumorigenesis in parous rats: is it caused by persistent changes in the hormonal environment or permanent biochemical alterations in the mammary epithelia? Carcinogenesis. 1995;16(11):2847-2853.

(20) Musey VC, Collins DC, Musey PI et al. Long-term effect of a first pregnancy on the secretion of prolactin. N Engl J Med. 1987;316(5):229- 234.

(21) Ingram DM, Nottage EM, Roberts AN. Prolactin and breast cancer risk. Med J Aust. 1990;153(8):469-473.

(22) Key TJ, Verkasalo PK. Endogenous hormones and the aetiology of breast cancer. Breast Cancer Res. 1999;1(1):18-21.

(23) Thomas HV, Key TJ, Allen DS et al. A prospective study of endogenous serum hormone concentrations and breast cancer risk in premenopausal women on the island of Guernsey. Br J Cancer. 1997;75(7):1075-1079.

(24) Rosenberg CR, Pasternack BS, Shore RE et al. Premenopausal estradiol levels and the risk of breast cancer: a new method of controlling for day of the menstrual cycle. Am J Epidemiol. 1994;140(6):518-525.

178 (25) Wysowski DK, Comstock GW, Helsing KJ, Lau HL. Sex hormone levels in serum in relation to the development of breast cancer. Am J Epidemiol. 1987;125(5):791-799.

(26) Helzlsouer KJ, Alberg AJ, Bush TL et al. A prospective study of endogenous hormones and breast cancer. Cancer Detect Prev. 1994;18(2):79-85.

(27) Wang DY, De Stavola BL, Bulbrook RD et al. Relationship of blood prolactin levels and the risk of subsequent breast cancer. Int J Epidemiol. 1992;21(2):214-221.

(28) Hankinson SE, Willett WC, Michaud DS et al. Plasma prolactin levels and subsequent risk of breast cancer in postmenopausal women. J Natl Cancer Inst. 1999;91(7):629-634.

(29) Nelson LR, Bulun SE. Estrogen production and action. J Am Acad Dermatol. 2001;45(3 Suppl):S116-S124.

(30) Clyne CD, Kovacic A, Speed CJ et al. Regulation of aromatase expression by the nuclear receptor LRH-1 in adipose tissue. Mol Cell Endocrinol. 2004;215(1-2):39-44.

(31) Miller WR, Anderson TJ, Jack WJ. Relationship between tumour aromatase activity, tumour characteristics and response to therapy. J Steroid Biochem Mol Biol. 1990;37(6):1055-1059.

(32) Saeki T, Takashima S, Sasaki H et al. Localization of Estrone Sulfatase in Human Breast Carcinomas. Breast Cancer. 1999;6(4):331-337.

(33) Suzuki T, Nakata T, Miki Y et al. Estrogen sulfotransferase and steroid sulfatase in human breast carcinoma. Cancer Res. 2003;63(11):2762- 2770.

(34) Blankenstein MA, van d, V, Maitimu-Smeele I et al. Intratumoral levels of estrogens in breast cancer. J Steroid Biochem Mol Biol. 1999;69(1-6):293- 297.

(35) Mol JA, Henzen-Logmans SC, Hageman P et al. Expression of the gene encoding growth hormone in the human mammary gland. J Clin Endocrinol Metab. 1995;80(10):3094-3096.

(36) Welsch CW, Jenkins TW, Meites J. Increased incidence of mammary tumors in the female rat grafted with multiple pituitaries. Cancer Res. 1970;30(4):1024-1029.

(37) Wennbo H, Gebre-Medhin M, Gritli-Linde A et al. Activation of the prolactin receptor but not the growth hormone receptor is important for

179 induction of mammary tumors in transgenic mice. J Clin Invest. 1997;100(11):2744-2751.

(38) Rose-Hellekant TA, Arendt LM, Schroeder MD et al. Prolactin induces ERalpha-positive and ERalpha-negative mammary cancer in transgenic mice. Oncogene. 2003;22(30):4664-4674.

(39) Vomachka AJ, Pratt SL, Lockefeer JA, Horseman ND. Prolactin gene- disruption arrests mammary gland development and retards T-antigen- induced tumor growth. Oncogene. 2000;19(8):1077-1084.

(40) Reynolds C, Montone KT, Powell CM et al. Expression of prolactin and its receptor in human breast carcinoma. Endocrinology. 1997;138(12):5555- 5560.

(41) Clinical trial of 2-Br- -ergocryptine (CB154) in advanced breast cancer. Eur J Cancer. 1972;8(2):155-156.

(42) Fritze D, Queisser W, Schmid H et al. Prospective randomized trial concerning hyper- and normoprolactinemia and the use of bromoergocryptine in patients with metastatic breast cancer. Onkologie. 1986;9(6):305-312.

(43) Tamoxifen for early breast cancer: an overview of the randomised trials. Early Breast Cancer Trialists' Collaborative Group. Lancet. 1998;351(9114):1451-1467.

(44) Fisher B, Costantino JP, Wickerham DL et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Studys. J Natl Cancer Inst. 1998;90(18):1371-1388.

(45) Bonneterre J, Buzdar A, Nabholtz JM et al. Anastrozole is superior to tamoxifen as first-line therapy in hormone receptor positive advanced breast carcinoma. Cancer. 2001;92(9):2247-2258.

(46) Baum M, Budzar AU, Cuzick J et al. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the ATAC randomised trial. Lancet. 2002;359(9324):2131-2139.

(47) Baum M, Buzdar A, Cuzick J et al. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early-stage breast cancer: results of the ATAC (Arimidex, Tamoxifen Alone or in Combination) trial efficacy and safety update analyses. Cancer. 2003;98(9):1802-1810.

180 (48) Coombes RC, Hall E, Gibson LJ et al. A randomized trial of exemestane after two to three years of tamoxifen therapy in postmenopausal women with primary breast cancer. N Engl J Med. 2004;350(11):1081-1092.

(49) Goss PE, Ingle JN, Martino S et al. A randomized trial of letrozole in postmenopausal women after five years of tamoxifen therapy for early- stage breast cancer. N Engl J Med. 2003;349(19):1793-1802.

(50) Winer EP, Hudis C, Burstein HJ et al. American Society of Clinical Oncology technology assessment on the use of aromatase inhibitors as adjuvant therapy for postmenopausal women with hormone receptor- positive breast cancer: status report 2004. J Clin Oncol. 2005;23(3):619- 629.

(51) Dowsett M, Stein RC, Coombes RC. Aromatization inhibition alone or in combination with GnRH agonists for the treatment of premenopausal breast cancer patients. J Steroid Biochem Mol Biol. 1992;43(1-3):155-159.

(52) Cardiff RD, Wellings SR. The comparative pathology of human and mouse mammary glands. J Mammary Gland Biol Neoplasia. 1999;4(1):105-122.

(53) Hennighausen L, Robinson GW. Think globally, act locally: the making of a mouse mammary gland. Genes Dev. 1998;12(4):449-455.

(54) Bocchinfuso WP, Korach KS. Mammary gland development and tumorigenesis in estrogen receptor knockout mice. J Mammary Gland Biol Neoplasia. 1997;2:323-334.

(55) Mueller SO, Clark JA, Myers PH, Korach KS. Mammary Gland Development in Adult Mice Requires Epithelial and Stromal Estrogen Receptor alpha. Endocrinology. 2002;143(6):2357-2365.

(56) Shi HY, Lydon JP, Zhang M. Hormonal defect in maspin heterozygous mice reveals a role of progesterone in pubertal ductal development. Mol Endocrinol. 2004;18(9):2196-2207.

(57) Humphreys RC, Lydon JP, O'Malley BW, Rosen JM. Use of PRKO Mice to Study the Role of Progesterone in Mammary Gland Development. J Mammary Gland Biol Neoplasia. 1997;2(4):343-354.

(58) Fata JE, Chaudhary V, Khokha R. Cellular turnover in the mammary gland is correlated with systemic levels of progesterone and not 17beta-estradiol during the estrous cycle. Biol Reprod. 2001;65(3):680-688.

(59) Feuerhake F, Sigg W, Hofter EA et al. Cell proliferation, apoptosis, and expression of Bcl-2 and Bax in non-lactating human breast epithelium in relation to the menstrual cycle and reproductive history. Breast Cancer Res Treat. 2003;77(1):37-48.

181 (60) Ferguson JE, Schor AM, Howell A, Ferguson MW. Changes in the extracellular matrix of the normal human breast during the menstrual cycle. Cell Tissue Res. 1992;268(1):167-177.

(61) Horseman ND, Zhao W, Montecino-Rodriguez E et al. Defective mammopoiesis, but normal hematopoiesis, in mice with a targeted disruption of the prolactin gene. EMBO J. 1997;16(23):6926-6935.

(62) Ormandy CJ, Camus A, Barra J et al. Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse. Genes Dev. 1997;11:167-177.

(63) Liu X, Robinson GW, Wagner KU et al. Stat5a is mandatory for adult mammary gland development and lactogenesis. Genes Dev. 1997;11:179- 186.

(64) Wagner KU, Krempler A, Triplett AA et al. Impaired alveologenesis and maintenance of secretory mammary epithelial cells in Jak2 conditional knockout mice. Mol Cell Biol. 2004;24(12):5510-5520.

(65) Marti A, Lazar H, Ritter P, Jaggi R. Transcription factor activities and gene expression during mouse mammary gland involution. J Mammary Gland Biol Neoplasia. 1999;4(2):145-152.

(66) Li M, Liu X, Robinson G et al. Mammary-derived signals activate programmed cell death during the first stage of mammary gland involution. Proc Natl Acad Sci U S A. 1997;94(7):3425-3430.

(67) Boudreau N, Sympson CJ, Werb Z, Bissell MJ. Suppression of ICE and apoptosis in mammary epithelial cells by extracellular matrix. Science. 1995;267(5199):891-893.

(68) Talhouk RS, Bissell MJ, Werb Z. Coordinated expression of extracellular matrix-degrading proteinases and their inhibitors regulates mammary epithelial function during involution. J Cell Biol. 1992;118(5):1271-1282.

(69) Lund LR, Romer J, Thomasset N et al. Two distinct phases of apoptosis in mammary gland involution: proteinase-independent and -dependent pathways. Development. 1996;122(1):181-193.

(70) Sutton AL, MacDonald PN. Vitamin D: more than a "bone-a-fide" hormone. Mol Endocrinol. 2003;17(5):777-791.

(71) Sillanpaa P, Hirvonen A, Kataja V et al. Vitamin D receptor gene polymorphism as an important modifier of positive family history related breast cancer risk. Pharmacogenetics. 2004;14(4):239-245.

182 (72) Schondorf T, Eisberg C, Wassmer G et al. Association of the vitamin D receptor genotype with bone metastases in breast cancer patients. Oncology. 2003;64(2):154-159.

(73) Garland CF, Garland FC, Gorham ED. Calcium and vitamin D. Their potential roles in colon and breast cancer prevention. Ann N Y Acad Sci. 1999;889:107-119.

(74) Garland FC, Garland CF, Gorham ED, Young JF. Geographic variation in breast cancer mortality in the United States: a hypothesis involving exposure to solar radiation. Prev Med. 1990;19(6):614-622.

(75) Robsahm TE, Tretli S, Dahlback A, Moan J. Vitamin D3 from sunlight may improve the prognosis of breast-, colon- and prostate cancer (Norway). Cancer Causes Control. 2004;15(2):149-158.

(76) Xue L, Lipkin M, Newmark H, Wang J. Influence of dietary calcium and vitamin D on diet-induced epithelial cell hyperproliferation in mice. J Natl Cancer Inst. 1999;91(2):176-181.

(77) Friedrich M, Villena-Heinsen C, Tilgen W et al. Vitamin D receptor (VDR) expression is not a prognostic factor in breast cancer. Anticancer Res. 2002;22(3):1919-1924.

(78) Friedrich M, Axt-Fliedner R, Villena-Heinsen C et al. Analysis of vitamin D- receptor (VDR) and retinoid X-receptor alpha in breast cancer. Histochem J. 2002;34(1-2):35-40.

(79) Zinser G, Packman K, Welsh J. Vitamin D(3) receptor ablation alters mammary gland morphogenesis. Development. 2002;129(13):3067-3076.

(80) Zinser GM, Welsh J. Accelerated mammary gland development during pregnancy and delayed postlactational involution in vitamin D3 receptor null mice. Mol Endocrinol. 2004;18(9):2208-2223.

(81) Mezzetti G, Monti MG, Casolo LP et al. 1,25-Dihydroxycholecalciferol- dependent calcium uptake by mouse mammary gland in culture. Endocrinology. 1988;122(2):389-394.

(82) Bhattacharjee M, Wientroub S, Vonderhaar BK. Milk protein synthesis by mammary glands of vitamin D-deficient mice. Endocrinology. 1987;121(3):865-874.

(83) Li YC, Pirro AE, Amling M et al. Targeted ablation of the vitamin D receptor: an animal model of vitamin D-dependent rickets type II with alopecia. Proc Natl Acad Sci U S A. 1997;94(18):9831-9835.

183 (84) Zinser GM, Welsh J. Vitamin D receptor status alters mammary gland morphology and tumorigenesis in MMTV-neu mice. Carcinogenesis. 2004.

(85) Chouvet C, Vicard E, Devonec M, Saez S. 1,25-Dihydroxyvitamin D3 inhibitory effect on the growth of two human breast cancer cell lines (MCF- 7, BT-20). J Steroid Biochem. 1986;24(1):373-376.

(86) Eisman JA, Sutherland RL, McMenemy ML et al. Effects of 1,25- dihydroxyvitamin D3 on cell-cycle kinetics of T 47D human breast cancer cells. J Cell Physiol. 1989;138(3):611-616.

(87) Eisman JA, Koga M, Sutherland RL et al. 1,25-Dihydroxyvitamin D3 and the regulation of human cancer cell replication. Proc Soc Exp Biol Med. 1989;191(3):221-226.

(88) Verlinden L, Verstuyf A, Convents R et al. Action of 1,25(OH)2D3 on the cell cycle genes, cyclin D1, p21 and p27 in MCF-7 cells. Mol Cell Endocrinol. 1998;142(1-2):57-65.

(89) Brenner RV, Shabahang M, Schumaker LM et al. The antiproliferative effect of vitamin D analogs on MCF-7 human breast cancer cells. Cancer Lett. 1995;92(1):77-82.

(90) Wu G, Fan RS, Li W et al. Modulation of cell cycle control by vitamin D3 and its analogue, EB1089, in human breast cancer cells. Oncogene. 1997;15(13):1555-1563.

(91) Swami S, Krishnan AV, Feldman D. 1alpha,25-Dihydroxyvitamin D3 down- regulates estrogen receptor abundance and suppresses estrogen actions in MCF-7 human breast cancer cells. Clin Cancer Res. 2000;6(8):3371- 3379.

(92) Stoica A, Saceda M, Fakhro A et al. Regulation of estrogen receptor-alpha gene expression by 1, 25-dihydroxyvitamin D in MCF-7 cells. J Cell Biochem. 1999;75(4):640-651.

(93) Flanagan L, Packman K, Juba B et al. Efficacy of Vitamin D compounds to modulate estrogen receptor negative breast cancer growth and invasion. J Steroid Biochem Mol Biol. 2003;84(2-3):181-192.

(94) Love-Schimenti CD, Gibson DF, Ratnam AV, Bikle DD. Antiestrogen potentiation of antiproliferative effects of vitamin D3 analogues in breast cancer cells. Cancer Res. 1996;56(12):2789-2794.

(95) Nolan E, Donepudi M, VanWeelden K et al. Dissociation of vitamin D3 and anti-estrogen mediated growth regulation in MCF-7 breast cancer cells. Mol Cell Biochem. 1998;188(1-2):13-20.

184 (96) Simboli-Campbell M, Narvaez CJ, Tenniswood M, Welsh J. 1,25- Dihydroxyvitamin D3 induces morphological and biochemical markers of apoptosis in MCF-7 breast cancer cells. J Steroid Biochem Mol Biol. 1996;58(4):367-376.

(97) Narvaez CJ, Zinser G, Welsh J. Functions of 1alpha,25-dihydroxyvitamin D(3) in mammary gland: from normal development to breast cancer. Steroids. 2001;66(3-5):301-308.

(98) Maenpaa PH, Vaisanen S, Jaaskelainen T et al. Vitamin D(3) analogs (MC 1288, KH 1060, EB 1089, GS 1558, and CB 1093): studies on their mechanism of action. Steroids. 2001;66(3-5):223-225.

(99) Bouillon R, Okamura WH, Norman AW. Structure-function relationships in the vitamin D endocrine system. Endocr Rev. 1995;16(2):200-257.

(100) Colston KW, Mackay AG, James SY et al. EB1089: a new vitamin D analogue that inhibits the growth of breast cancer cells in vivo and in vitro. Biochem Pharmacol. 1992;44(12):2273-2280.

(101) VanWeelden K, Flanagan L, Binderup L et al. Apoptotic regression of MCF-7 xenografts in nude mice treated with the vitamin D3 analog, EB1089. Endocrinology. 1998;139(4):2102-2110.

(102) Vink-van Wijngaarden T, Pols HA, Buurman CJ et al. Inhibition of breast cancer cell growth by combined treatment with vitamin D3 analogues and tamoxifen. Cancer Res. 1994;54(21):5711-5717.

(103) Koshizuka K, Koike M, Asou H et al. Combined effect of vitamin D3 analogs and paclitaxel on the growth of MCF-7 breast cancer cells in vivo. Breast Cancer Res Treat. 1999;53(2):113-120.

(104) Sundaram S, Sea A, Feldman S et al. The combination of a potent vitamin D3 analog, EB 1089, with ionizing radiation reduces tumor growth and induces apoptosis of MCF-7 breast tumor xenografts in nude mice. Clin Cancer Res. 2003;9(6):2350-2356.

(105) El Abdaimi K, Dion N, Papavasiliou V et al. The vitamin D analogue EB 1089 prevents skeletal metastasis and prolongs survival time in nude mice transplanted with human breast cancer cells. Cancer Res. 2000;60(16):4412-4418.

(106) Colston KW, Pirianov G, Bramm E et al. Effects of Seocalcitol (EB1089) on nitrosomethyl urea-induced rat mammary tumors. Breast Cancer Res Treat. 2003;80(3):303-311.

185 (107) Gulliford T, English J, Colston KW et al. A phase I study of the vitamin D analogue EB 1089 in patients with advanced breast and colorectal cancer. Br J Cancer. 1998;78(1):6-13.

(108) Evans TR, Colston KW, Lofts FJ et al. A phase II trial of the vitamin D analogue Seocalcitol (EB1089) in patients with inoperable pancreatic cancer. Br J Cancer. 2002;86(5):680-685.

(109) Dalhoff K, Dancey J, Astrup L et al. A phase II study of the vitamin D analogue Seocalcitol in patients with inoperable hepatocellular carcinoma. Br J Cancer. 2003;89(2):252-257.

(110) Geschickter CF. Mammary carcinoma in the rat with metastasis induced by estrogen. Science. 1939;89:35-37.

(111) Nandi S, Guzman RC, Yang J. Hormones and mammary carcinogenesis in mice, rats, and humans: a unifying hypothesis. Proc Natl Acad Sci U S A. 1995;92(9):3650-3657.

(112) Yang X, Edgerton SM, Kosanke SD et al. Hormonal and dietary modulation of mammary carcinogenesis in mouse mammary tumor virus- c-erbB-2 transgenic mice. Cancer Res. 2003;63(10):2425-2433.

(113) Medina D, Kittrell FS, Hill J et al. Tamoxifen inhibition of estrogen receptor-alpha-negative mouse mammary tumorigenesis. Cancer Res. 2005;65(8):3493-3496.

(114) Swanson SM, Guzman RC, Christov K et al. Pituitary-isografted mice are highly susceptible to MNU-induced mammary carcinogenesis irrespective of the level of alveolar differentiation. Carcinogenesis. 1994;15(7):1341- 1346.

(115) Medina D, Smith GH. Chemical carcinogen-induced tumorigenesis in parous, involuted mouse mammary glands. J Natl Cancer Inst. 1999;91(11):967-969.

(116) Grubbs CJ, Farnell DR, Hill DL, McDonough KC. Chemoprevention of N- nitroso-N-methylurea-induced mammary cancers by pretreatment with 17 beta-estradiol and progesterone. J Natl Cancer Inst. 1985;74(4):927-931.

(117) Russo IH, Koszalka M, Russo J. Human chorionic gonadotropin and rat mammary cancer prevention. J Natl Cancer Inst. 1990;82(15):1286-1289.

(118) Sivaraman L, Conneely OM, Medina D, O'Malley BW. p53 is a potential mediator of pregnancy and hormone-induced resistance to mammary carcinogenesis. Proc Natl Acad Sci U S A. 2001;98(22):12379-12384.

186 (119) Vousden KH, Lu X. Live or let die: the cell's response to p53. Nat Rev Cancer. 2002;2(8):594-604.

(120) Donehower LA, Harvey M, Slagle BL et al. Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature. 1992;356(6366):215-221.

(121) Jacks T, Remington L, Williams BO et al. Tumor spectrum analysis in p53- mutant mice. Curr Biol. 1994;4(1):1-7.

(122) Hollstein M, Sidransky D, Vogelstein B, Harris CC. p53 mutations in human cancers. Science. 1991;253(5015):49-53.

(123) Hollstein M, Rice K, Greenblatt MS et al. Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic Acids Res. 1994;22(17):3551-3555.

(124) Overholtzer M, Rao PH, Favis R et al. The presence of p53 mutations in human osteosarcomas correlates with high levels of genomic instability. Proc Natl Acad Sci U S A. 2003;100(20):11547-11552.

(125) Lane DP. Cancer. p53, guardian of the genome. Nature. 1992;358(6381):15-16.

(126) Mayo LD, Donner DB. The PTEN, Mdm2, p53 tumor suppressor- oncoprotein network. Trends Biochem Sci. 2002;27(9):462-467.

(127) Harris SL, Levine AJ. The p53 pathway: positive and negative feedback loops. Oncogene. 2005;24(17):2899-2908.

(128) Giaccia AJ, Kastan MB. The complexity of p53 modulation: emerging patterns from divergent signals. Genes Dev. 1998;12(19):2973-2983.

(129) Brooks CL, Gu W. Ubiquitination, phosphorylation and acetylation: the molecular basis for p53 regulation. Curr Opin Cell Biol. 2003;15(2):164- 171.

(130) El Deiry WS, Tokino T, Velculescu VE et al. WAF1, a potential mediator of p53 tumor suppression. Cell. 1993;75(4):817-825.

(131) Hermeking H, Lengauer C, Polyak K et al. 14-3-3 sigma is a p53-regulated inhibitor of G2/M progression. Mol Cell. 1997;1(1):3-11.

(132) Malkin D. p53 and the Li-Fraumeni syndrome. Cancer Genet Cytogenet. 1993;66(2):83-92.

187 (133) Sigal A, Rotter V. Oncogenic mutations of the p53 tumor suppressor: the demons of the guardian of the genome. Cancer Res. 2000;60(24):6788- 6793.

(134) Scian MJ, Stagliano KE, Ellis MA et al. Modulation of gene expression by tumor-derived p53 mutants. Cancer Res. 2004;64(20):7447-7454.

(135) O'Farrell TJ, Ghosh P, Dobashi N et al. Comparison of the effect of mutant and wild-type p53 on global gene expression. Cancer Res. 2004;64(22):8199-8207.

(136) Strano S, Fontemaggi G, Costanzo A et al. Physical interaction with human tumor-derived p53 mutants inhibits p63 activities. J Biol Chem. 2002;277(21):18817-18826.

(137) Strano S, Munarriz E, Rossi M et al. Physical and functional interaction between p53 mutants and different isoforms of p73. J Biol Chem. 2000;275(38):29503-29512.

(138) Di Como CJ, Gaiddon C, Prives C. p73 function is inhibited by tumor- derived p53 mutants in mammalian cells. Mol Cell Biol. 1999;19(2):1438- 1449.

(139) Gaiddon C, Lokshin M, Ahn J et al. A subset of tumor-derived mutant forms of p53 down-regulate p63 and p73 through a direct interaction with the p53 core domain. Mol Cell Biol. 2001;21(5):1874-1887.

(140) Malkin D, Li FP, Strong LC et al. Germ line p53 mutations in a familial syndrome of breast cancer, sarcomas, and other neoplasms. Science. 1990;250(4985):1233-1238.

(141) Pharoah PD, Day NE, Caldas C. Somatic mutations in the p53 gene and prognosis in breast cancer: a meta-analysis. Br J Cancer. 1999;80(12):1968-1973.

(142) Done SJ, Eskandarian S, Bull S et al. p53 missense mutations in microdissected high-grade ductal carcinoma in situ of the breast. J Natl Cancer Inst. 2001;93(9):700-704.

(143) Angele S, Treilleux I, Taniere P et al. Abnormal expression of the ATM and TP53 genes in sporadic breast carcinomas. Clin Cancer Res. 2000;6(9):3536-3544.

(144) Sullivan A, Yuille M, Repellin C et al. Concomitant inactivation of p53 and Chk2 in breast cancer. Oncogene. 2002;21(9):1316-1324.

188 (145) Allinen M, Huusko P, Mantyniemi S et al. Mutation analysis of the CHK2 gene in families with hereditary breast cancer. Br J Cancer. 2001;85(2):209-212.

(146) McCann AH, Kirley A, Carney DN et al. Amplification of the MDM2 gene in human breast cancer and its association with MDM2 and p53 protein status. Br J Cancer. 1995;71(5):981-985.

(147) Danovi D, Meulmeester E, Pasini D et al. Amplification of Mdmx (or Mdm4) directly contributes to tumor formation by inhibiting p53 tumor suppressor activity. Mol Cell Biol. 2004;24(13):5835-5843.

(148) Umbricht CB, Evron E, Gabrielson E et al. Hypermethylation of 14-3-3 sigma (stratifin) is an early event in breast cancer. Oncogene. 2001;20(26):3348-3353.

(149) Gentile M, Ahnstrom M, Schon F, Wingren S. Candidate tumour suppressor genes at 11q23-q24 in breast cancer: evidence of alterations in PIG8, a gene involved in p53-induced apoptosis. Oncogene. 2001;20(53):7753-7760.

(150) Kuperwasser C, Hurlbut GD, Kittrell FS et al. Development of spontaneous mammary tumors in BALB/c p53 heterozygous mice. A model for Li-Fraumeni syndrome. Am J Pathol. 2000;157(6):2151-2159.

(151) Jerry DJ, Kittrell FS, Kuperwasser C et al. A mammary-specific model demonstrates the role of the p53 tumor suppressor gene in tumor development. Oncogene. 2000;19(8):1052-1058.

(152) Xu X, Wagner KU, Larson D et al. Conditional mutation of Brca1 in mammary epithelial cells results in blunted ductal morphogenesis and tumour formation. Nat Genet. 1999;22(1):37-43.

(153) Cressman VL, Backlund DC, Hicks EM et al. Mammary tumor formation in p53- and BRCA1-deficient mice. Cell Growth Differ. 1999;10(1):1-10.

(154) Donehower LA, Godley LA, Aldaz CM et al. Deficiency of p53 accelerates mammary tumorigenesis in Wnt-1 transgenic mice and promotes chromosomal instability. Genes Dev. 1995;9(7):882-895.

(155) Dittmer D, Pati S, Zambetti G et al. Gain of function mutations in p53. Nat Genet. 1993;4(1):42-46.

(156) Li B, Rosen JM, McMenamin-Balano J et al. neu/ERBB2 cooperates with p53-172H during mammary tumorigenesis in transgenic mice. Mol Cell Biol. 1997;17(6):3155-3163.

189 (157) Li B, Murphy KL, Laucirica R et al. A transgenic mouse model for mammary carcinogenesis. Oncogene. 1998;16(8):997-1007.

(158) Hadsell DL, Murphy KL, Bonnette SG et al. Cooperative interaction between mutant p53 and des(1-3)IGF-I accelerates mammary tumorigenesis. Oncogene. 2000;19(7):889-898.

(159) Draviam VM, Xie S, Sorger PK. Chromosome segregation and genomic stability. Curr Opin Genet Dev. 2004;14(2):120-125.

(160) Lingle WL, Lutz WH, Ingle JN et al. Centrosome hypertrophy in human breast tumors: implications for genomic stability and cell polarity. Proc Natl Acad Sci U S A. 1998;95(6):2950-2955.

(161) Lingle WL, Salisbury JL. Altered centrosome structure is associated with abnormal mitoses in human breast tumors. Am J Pathol. 1999;155(6):1941-1951.

(162) D'Assoro AB, Barrett SL, Folk C et al. Amplified centrosomes in breast cancer: a potential indicator of tumor aggressiveness. Breast Cancer Res Treat. 2002;75(1):25-34.

(163) Lingle WL, Barrett SL, Negron VC et al. Centrosome amplification drives chromosomal instability in breast tumor development. Proc Natl Acad Sci U S A. 2002;99(4):1978-1983.

(164) Sato N, Mizumoto K, Nakamura M et al. Centrosome abnormalities in pancreatic ductal carcinoma. Clin Cancer Res. 1999;5(5):963-970.

(165) Pihan GA, Purohit A, Wallace J et al. Centrosome defects can account for cellular and genetic changes that characterize prostate cancer progression. Cancer Res. 2001;61(5):2212-2219.

(166) Pihan GA, Purohit A, Wallace J et al. Centrosome defects and genetic instability in malignant tumors. Cancer Res. 1998;58(17):3974-3985.

(167) Ghadimi BM, Sackett DL, Difilippantonio MJ et al. Centrosome amplification and instability occurs exclusively in aneuploid, but not in diploid colorectal cancer cell lines, and correlates with numerical chromosomal aberrations. Genes Chromosomes Cancer. 2000;27(2):183- 190.

(168) Pihan GA, Wallace J, Zhou Y, Doxsey SJ. Centrosome abnormalities and chromosome instability occur together in pre-invasive carcinomas. Cancer Res. 2003;63(6):1398-1404.

190 (169) Goepfert TM, Adigun YE, Zhong L et al. Centrosome amplification and overexpression of aurora A are early events in rat mammary carcinogenesis. Cancer Res. 2002;62(14):4115-4122.

(170) Duensing S, Duensing A, Crum CP, Munger K. Human papillomavirus type 16 E7 oncoprotein-induced abnormal centrosome synthesis is an early event in the evolving malignant phenotype. Cancer Res. 2001;61(6):2356-2360.

(171) Carroll PE, Okuda M, Horn HF et al. Centrosome hyperamplification in human cancer: chromosome instability induced by p53 mutation and/or Mdm2 overexpression. Oncogene. 1999;18(11):1935-1944.

(172) Tarapore P, Fukasawa K. Loss of p53 and centrosome hyperamplification. Oncogene. 2002;21(40):6234-6240.

(173) Xu X, Weaver Z, Linke SP et al. Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic instability in BRCA1 exon 11 isoform-deficient cells. Mol Cell. 1999;3(3):389-395.

(174) Tutt A, Gabriel A, Bertwistle D et al. Absence of Brca2 causes genome instability by chromosome breakage and loss associated with centrosome amplification. Curr Biol. 1999;9(19):1107-1110.

(175) Montagna C, Andrechek ER, Padilla-Nash H et al. Centrosome abnormalities, recurring deletions of chromosome 4, and genomic amplification of HER2/neu define mouse mammary gland adenocarcinomas induced by mutant HER2/neu. Oncogene. 2002;21(6):890-898.

(176) Weaver Z, Montagna C, Xu X et al. Mammary tumors in mice conditionally mutant for Brca1 exhibit gross genomic instability and centrosome amplification yet display a recurring distribution of genomic imbalances that is similar to human breast cancer. Oncogene. 2002;21(33):5097- 5107.

(177) Tanaka T, Kimura M, Matsunaga K et al. Centrosomal kinase AIK1 is overexpressed in invasive ductal carcinoma of the breast. Cancer Res. 1999;59(9):2041-2044.

(178) Takahashi T, Futamura M, Yoshimi N et al. Centrosomal kinases, HsAIRK1 and HsAIRK3, are overexpressed in primary colorectal cancers. Jpn J Cancer Res. 2000;91(10):1007-1014.

(179) Gritsko TM, Coppola D, Paciga JE et al. Activation and overexpression of centrosome kinase BTAK/Aurora-A in human ovarian cancer. Clin Cancer Res. 2003;9(4):1420-1426.

191 (180) Li D, Zhu J, Firozi PF et al. Overexpression of oncogenic STK15/BTAK/Aurora A kinase in human pancreatic cancer. Clin Cancer Res. 2003;9(3):991-997.

(181) Zhou H, Kuang J, Zhong L et al. Tumour amplified kinase STK15/BTAK induces centrosome amplification, aneuploidy and transformation. Nat Genet. 1998;20(2):189-193.

(182) Risma KA, Clay CM, Nett TM et al. Targeted overexpression of luteinizing hormone in transgenic mice leads to infertility, polycystic ovaries, and ovarian tumors. Proc Natl Acad Sci U S A. 1995;92(5):1322-1326.

(183) Keri RA, Lozada KL, Abdul-Karim FW et al. Luteinizing hormone induction of ovarian tumors: oligogenic differences between mouse strains dictates tumor disposition. Proc Natl Acad Sci U S A. 2000;97(1):383-387.

(184) Risma KA, Hirshfield AN, Nilson JH. Elevated luteinizing hormone in prepubertal transgenic mice causes hyperandrogenemia, precocious puberty, and substantial ovarian pathology. Endocrinology. 1997;138(8):3540-3547.

(185) Mann RJ, Keri RA, Nilson JH. Transgenic mice with chronically elevated luteinizing hormone are infertile due to anovulation, defects in uterine receptivity, and midgestation pregnancy failure. Endocrinology. 1999;140(6):2592-2601.

(186) Flaws JA, Abbud R, Mann RJ et al. Chronically elevated luteinizing hormone depletes primordial follicles in the mouse ovary. Biol Reprod. 1997;57(5):1233-1237.

(187) Owens GE, Keri RA, Nilson JH. Ovulatory surges of human CG prevent hormone-induced granulosa cell tumor formation leading to the identification of tumor-associated changes in the transcriptome. Mol Endocrinol. 2002;16(6):1230-1242.

(188) Kero J, Poutanen M, Zhang FP et al. Elevated luteinizing hormone induces expression of its receptor and promotes steroidogenesis in the adrenal cortex. J Clin Invest. 2000;105(5):633-641.

(189) Mohammad HP, Abbud RA, Parlow AF et al. Targeted overexpression of luteinizing hormone causes ovary-dependent functional adenomas restricted to cells of the Pit-1 lineage. Endocrinology. 2003;144(10):4626- 4636.

(190) Milliken EL, Ameduri RK, Landis MD et al. Ovarian hyperstimulation by LH leads to mammary gland hyperplasia and cancer predisposition in transgenic mice. Endocrinology. 2002;143(9):3671-3680.

192 (191) Armstrong K, Eisen A, Weber B. Assessing the risk of breast cancer. N Engl J Med. 2000;342(8):564-571.

(192) Kvale G. Reproductive factors in breast cancer epidemiology. Acta Oncol. 1992;31(2):187-194.

(193) Stoll BA, Vatten LJ, Kvinnsland S. Does early physical maturity influence breast cancer risk? Acta Oncol. 1994;33(2):171-176.

(194) Kreiger N, Sloan M, Cotterchio M, Kirsh V. The risk of breast cancer following reproductive surgery. Eur J Cancer. 1999;35(1):97-101.

(195) Guzman RC, Yang J, Rajkumar L et al. Hormonal prevention of breast cancer: mimicking the protective effect of pregnancy. Proc Natl Acad Sci U S A. 1999;96(5):2520-2525.

(196) Russo J, Russo IH. Differentiation and breast cancer. Medicina (B Aires). 1997;57 Suppl 2:81-91.

(197) Callahan R, Smith GH. MMTV-induced mammary tumorigenesis: gene discovery, progression to malignancy and cellular pathways. Oncogene. 2000;19(8):992-1001.

(198) Tennant RW, Rao GN, Russfield A et al. Chemical effects in transgenic mice bearing oncogenes expressed in mammary tissue. Carcinogenesis. 1993;14(1):29-35.

(199) Hundley JE, Koester SK, Troyer DA et al. Differential regulation of cell cycle characteristics and apoptosis in MMTV-myc and MMTV-ras mouse mammary tumors. Cancer Res. 1997;57(4):600-603.

(200) Wang TC, Cardiff RD, Zukerberg L et al. Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature. 1994;369(6482):669-671.

(201) Lin TP, Guzman RC, Osborn RC et al. Role of endocrine, autocrine, and paracrine interactions in the development of mammary hyperplasia in Wnt- 1 transgenic mice. Cancer Res. 1992;52(16):4413-4419.

(202) Matsuzawa A, Yamamoto T. Lack of growth of a pregnancy-dependent mouse mammary tumor (TPDMT-4) in the absence of pituitary hormones. J Natl Cancer Inst. 1977;58(4):1087-1091.

(203) Matsuzawa A, Yamamoto T. Inhibited growth in vivo of a mouse pregnancy-dependent mammary tumor (TPDMT-4) by an antiestrogen, 2alpha, 3alpha-epithio-5alpha-androstan-17beta-ol (10275-S). Cancer Res. 1976;36(5):1598-1606.

193 (204) Lojun S, Bao S, Lei ZM, Rao CV. Presence of functional luteinizing hormone/chorionic gonadotropin (hCG) receptors in human breast cell lines: implications supporting the premise that hCG protects women against breast cancer. Biol Reprod. 1997;57(5):1202-1210.

(205) Meduri G, Charnaux N, Loosfelt H et al. Luteinizing hormone/human chorionic gonadotropin receptors in breast cancer. Cancer Res. 1997;57(5):857-864.

(206) Raafat AM, Hofseth LJ, Li S et al. A mouse model to study the effects of hormone replacement therapy on normal mammary gland during menopause: enhanced proliferative response to estrogen in late postmenopausal mice. Endocrinology. 1999;140(6):2570-2580.

(207) Bocchinfuso WP, Lindzey JK, Hewitt SC et al. Induction of mammary gland development in estrogen receptor-alpha knockout mice. Endocrinology. 2000;141(8):2982-2994.

(208) Couse JF, Korach KS. Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev. 1999;20(3):358-417.

(209) Klijn JG, Beex LV, Mauriac L et al. Combined treatment with buserelin and tamoxifen in premenopausal metastatic breast cancer: a randomized study. J Natl Cancer Inst. 2000;92(11):903-911.

(210) Klijn JG, de Jong FH, Lamberts SW, Blankenstein MA. LHRH-agonist treatment in clinical and experimental human breast cancer. J Steroid Biochem. 1985;23(5B):867-873.

(211) Keri RA, Andersen B, Kennedy GC et al. Estradiol inhibits transcription of the human glycoprotein hormone alpha-subunit gene despite the absence of a high affinity binding site for estrogen receptor. Mol Endocrinol. 1991;5(5):725-733.

(212) Rasmussen SB, Young LJT Smith GH. Preparing Mammary Gland Whole Mounts from Mice. Ip MM, Asch BB eds. Methods in Mammary Gland Biology and Breast Cancer Research. 75-85. 2000. New York, Kluwer Academic/Plenum Publishers.

(213) Katoh AK, Stemmler N, Specht S, D'Amico F. Immunoperoxidase staining for estrogen and progesterone receptors in archival formalin fixed, paraffin embedded breast carcinomas after microwave antigen retrieval. Biotech Histochem. 1997;72(6):291-298.

(214) Sivaraman L, Hilsenbeck SG, Zhong L et al. Early exposure of the rat mammary gland to estrogen and progesterone blocks co-localization of estrogen receptor expression and proliferation. J Endocrinol. 2001;171(1):75-83.

194 (215) Robinson GW, McKnight RA, Smith GH, Hennighausen L. Mammary epithelial cells undergo secretory differentiation in cycling virgins but require pregnancy for the establishment of terminal differentiation. Development. 1995;121(7):2079-2090.

(216) Ascoli M, Segaloff DL. Adenohypophyseal Hormones and Their Hypothalamic Releasing Factors. Hardman JG, Limbard LE eds. Goodman & Gilman's The Pharmacological Basis of Therapeutics. 1363- 1382. 1996. New York, McGraw-Hill.

(217) Russo IH, Koszalka M, Russo J. Effect of human chorionic gonadotropin on mammary gland differentiation and carcinogenesis. Carcinogenesis. 1990;11(10):1849-1855.

(218) Srivastava P, Russo J, Russo IH. Chorionic gonadotropin inhibits rat mammary carcinogenesis through activation of programmed cell death. Carcinogenesis. 1997;18(9):1799-1808.

(219) Segaloff DL, Sprengel R, Nikolics K, Ascoli M. Structure of the lutropin/choriogonadotropin receptor. Recent Prog Horm Res. 1990;46:261-301.

(220) Tao YX, Lei ZM, Rao CV. The presence of luteinizing hormone/human chorionic gonadotropin receptors in lactating rat mammary glands. Life Sci. 1997;60(15):1297-1303.

(221) Abbud RA, Ameduri RK, Rao JS et al. Chronic hypersecretion of luteinizing hormone in transgenic mice selectively alters responsiveness of the alpha-subunit gene to gonadotropin-releasing hormone and estrogens. Mol Endocrinol. 1999;13(9):1449-1459.

(222) Hsieh C, Pavia M, Lambe M et al. Dual effect of parity on breast cancer risk. Eur J Cancer. 1994;30A(7):969-973.

(223) Cardiff RD, Anver MR, Gusterson BA et al. The mammary pathology of genetically engineered mice: the consensus report and recommendations from the Annapolis meeting. Oncogene. 2000;19(8):968-988.

(224) Ormandy CJ, Hall RE, Manning DL et al. Coexpression and cross- regulation of the prolactin receptor and sex steroid hormone receptors in breast cancer. J Clin Endocrinol Metab. 1997;82(11):3692-3699.

(225) De Placido S, Gallo C, Perrone F et al. Prolactin receptor does not correlate with oestrogen and progesterone receptors in primary breast cancer and lacks prognostic significance. Ten year results of the Naples adjuvant (GUN) study. Br J Cancer. 1990;62(4):643-646.

195 (226) Rutqvist LE, Cedermark B, Fornander T et al. The relationship between hormone receptor content and the effect of adjuvant tamoxifen in operable breast cancer. J Clin Oncol. 1989;7(10):1474-1484.

(227) Kaufmann M, Jonat W, Kleeberg U et al. Goserelin, a depot gonadotrophin-releasing hormone agonist in the treatment of premenopausal patients with metastatic breast cancer. German Zoladex Trial Group. J Clin Oncol. 1989;7(8):1113-1119.

(228) Apter D, Reinila M, Vihko R. Some endocrine characteristics of early menarche, a risk factor for breast cancer, are preserved into adulthood. Int J Cancer. 1989;44(5):783-787.

(229) Grubbs CJ, Juliana MM, Hill DL, Whitaker LM. Suppression by pregnancy of chemically induced preneoplastic cells of the rat mammary gland. Anticancer Res. 1986;6(6):1395-1400.

(230) Grubbs CJ, Hill DL, McDonough KC, Peckham JC. N-nitroso-N- methylurea-induced mammary carcinogenesis: effect of pregnancy on preneoplastic cells. J Natl Cancer Inst. 1983;71(3):625-628.

(231) Rao CV. Does full-term pregnancy at a young age protect women against breast cancer through hCG? Obstet Gynecol. 2000;96(5 Pt 1):783-786.

(232) Newcomb PA. Lactation and breast cancer risk. J Mammary Gland Biol Neoplasia. 1997;2(3):311-318.

(233) Zheng T, Holford TR, Mayne ST et al. Lactation and breast cancer risk: a case-control study in Connecticut. Br J Cancer. 2001;84(11):1472-1476.

(234) Zheng T, Duan L, Liu Y et al. Lactation reduces breast cancer risk in Shandong Province, China. Am J Epidemiol. 2000;152(12):1129-1135.

(235) Purwanto H, Sadjimin T, Dwiprahasto I. Lactation and the risk of breast cancer. Gan To Kagaku Ryoho. 2000;27 Suppl 2:474-481.

(236) Sinha DK, Pazik JE, Dao TL. Prevention of mammary carcinogenesis in rats by pregnancy: effect of full-term and interrupted pregnancy. Br J Cancer. 1988;57(4):390-394.

(237) Lanari C, Molinolo AA, Pasqualini CD. Induction of mammary adenocarcinomas by medroxyprogesterone acetate in BALB/c female mice. Cancer Lett. 1986;33(2):215-223.

(238) Huseby RA, Soares MJ, Talamantes F. Ectopic pituitary grafts in mice: hormone levels, effects on fertility, and the development of adenomyosis uteri, prolactinomas, and mammary carcinomas. Endocrinology. 1985;116(4):1440-1448.

196 (239) Sinha DK, Pazik JE, Dao TL. Progression of rat mammary development with age and its relationship to carcinogenesis by a chemical carcinogen. Int J Cancer. 1983;31(3):321-327.

(240) Milliken EL, Zhang X, Flask C et al. EB1089, a Vitamin D Receptor Agonist, Reduces Proliferation and Decreases Tumor Growth Rate in a Mouse Model of Hormone-Induced Mammary Cancer. Cancer Lett. 2005.

(241) John EM, Schwartz GG, Dreon DM, Koo J. Vitamin D and breast cancer risk: the NHANES I Epidemiologic follow-up study, 1971-1975 to 1992. National Health and Nutrition Examination Survey. Cancer Epidemiol Biomarkers Prev. 1999;8(5):399-406.

(242) Shin MH, Holmes MD, Hankinson SE et al. Intake of dairy products, calcium, and vitamin d and risk of breast cancer. J Natl Cancer Inst. 2002;94(17):1301-1311.

(243) Berger U, Wilson P, McClelland RA et al. Immunocytochemical detection of 1,25-dihydroxyvitamin D3 receptor in breast cancer. Cancer Res. 1987;47(24 Pt 1):6793-6799.

(244) Fisher B, Costantino JP, Wickerham DL et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst. 1998;90(18):1371-1388.

(245) Friedrich M, Rafi L, Tilgen W et al. Expression of 1,25-dihydroxy vitamin D3 receptor in breast carcinoma. J Histochem Cytochem. 1998;46(11):1335-1337.

(246) Byrne IM, Flanagan L, Tenniswood MP, Welsh J. Identification of a hormone-responsive promoter immediately upstream of exon 1c in the human vitamin D receptor gene. Endocrinology. 2000;141(8):2829-2836.

(247) Vienonen A, Miettinen S, Manninen T et al. Regulation of nuclear receptor and cofactor expression in breast cancer cell lines. Eur J Endocrinol. 2003;148(4):469-479.

(248) Qi X, Pramanik R, Wang J et al. The p38 and JNK pathways cooperate to trans-activate vitamin D receptor via c-Jun/AP-1 and sensitize human breast cancer cells to vitamin D(3)-induced growth inhibition. J Biol Chem. 2002;277(29):25884-25892.

(249) Fisher B, Costantino JP, Wickerham DL et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst. 1998;90(18):1371-1388.

(250) Fukasawa K, Choi T, Kuriyama R et al. Abnormal centrosome amplification in the absence of p53. Science. 1996;271(5256):1744-1747.

197 (251) Tarapore P, Horn HF, Tokuyama Y, Fukasawa K. Direct regulation of the centrosome duplication cycle by the p53-p21Waf1/Cip1 pathway. Oncogene. 2001;20(25):3173-3184.

(252) Al Kuraya K, Schraml P, Torhorst J et al. Prognostic relevance of gene amplifications and coamplifications in breast cancer. Cancer Res. 2004;64(23):8534-8540.

(253) Thordarson G, Van Horn K, Guzman RC et al. Parous rats regain high susceptibility to chemically induced mammary cancer after treatment with various mammotropic hormones. Carcinogenesis. 2001;22(7):1027-1033.

(254) Nigg EA. Centrosome aberrations: cause or consequence of cancer progression? Nat Rev Cancer. 2002;2(11):815-825.

(255) Miller CW, Chumakov A, Said J et al. Mutant p53 proteins have diverse intracellular abilities to oligomerize and activate transcription. Oncogene. 1993;8(7):1815-1824.

(256) Okumura N, Saji S, Eguchi H et al. Distinct promoter usage of mdm2 gene in human breast cancer. Oncol Rep. 2002;9(3):557-563.

(257) Fei P, El Deiry WS. P53 and radiation responses. Oncogene. 2003;22(37):5774-5783.

(258) Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100(1):57- 70.

(259) Backlund MG, Trasti SL, Backlund DC et al. Impact of ionizing radiation and genetic background on mammary tumorigenesis in p53-deficient mice. Cancer Res. 2001;61(17):6577-6582.

(260) Vares G, Ory K, Lectard B et al. Progesterone prevents radiation-induced apoptosis in breast cancer cells. Oncogene. 2004;23(26):4603-4613.

(261) Ory K, Lebeau J, Levalois C et al. Apoptosis inhibition mediated by medroxyprogesterone acetate treatment of breast cancer cell lines. Breast Cancer Res Treat. 2001;68(3):187-198.

(262) Feng Z, Marti A, Jehn B et al. Glucocorticoid and progesterone inhibit involution and programmed cell death in the mouse mammary gland. J Cell Biol. 1995;131(4):1095-1103.

(263) Capuco AV, Li M, Long E et al. Concurrent pregnancy retards mammary involution: effects on apoptosis and proliferation of the mammary epithelium after forced weaning of mice. Biol Reprod. 2002;66(5):1471- 1476.

198 (264) Berg MN, Dharmarajan AM, Waddell BJ. Glucocorticoids and progesterone prevent apoptosis in the lactating rat mammary gland. Endocrinology. 2002;143(1):222-227.

(265) Cheng KC, Loeb LA. Genomic instability and tumor progression: mechanistic considerations. Adv Cancer Res. 1993;60:121-156.

(266) Sharp GA, Osborn M, Weber K. Ultrastructure of multiple microtubule initiation sites in mouse neuroblastoma cells. J Cell Sci. 1981;47:1-24.

(267) Brinkley BR. Managing the centrosome numbers game: from chaos to stability in cancer cell division. Trends Cell Biol. 2001;11(1):18-21.

(268) Hainaut P, Soussi T, Shomer B et al. Database of p53 gene somatic mutations in human tumors and cell lines: updated compilation and future prospects. Nucleic Acids Res. 1997;25(1):151-157.

(269) Lundgren K, Montes de Oca LR, McNeill YB et al. Targeted expression of MDM2 uncouples S phase from mitosis and inhibits mammary gland development independent of p53. Genes Dev. 1997;11(6):714-725.

(270) Hori M, Shimazaki J, Inagawa S et al. Overexpression of MDM2 oncoprotein correlates with possession of estrogen receptor alpha and lack of MDM2 mRNA splice variants in human breast cancer. Breast Cancer Res Treat. 2002;71(1):77-83.

(271) Sheikh MS, Shao ZM, Hussain A, Fontana JA. The p53-binding protein MDM2 gene is differentially expressed in human breast carcinoma. Cancer Res. 1993;53(14):3226-3228.

(272) Ghosh M, Huang K, Berberich SJ. Overexpression of Mdm2 and MdmX fusion proteins alters p53 mediated transactivation, ubiquitination, and degradation. Biochemistry. 2003;42(8):2291-2299.

(273) Delia D, Fontanella E, Ferrario C et al. DNA damage-induced cell-cycle phase regulation of p53 and p21waf1 in normal and ATM-defective cells. Oncogene. 2003;22(49):7866-7869.

(274) Takai H, Naka K, Okada Y et al. Chk2-deficient mice exhibit radioresistance and defective p53-mediated transcription. EMBO J. 2002;21(19):5195-5205.

(275) Fabbro M, Savage K, Hobson K et al. BRCA1-BARD1 complexes are required for p53Ser-15 phosphorylation and a G1/S arrest following ionizing radiation-induced DNA damage. J Biol Chem. 2004;279(30):31251-31258.

199 (276) Chao C, Saito S, Anderson CW et al. Phosphorylation of murine p53 at ser-18 regulates the p53 responses to DNA damage. Proc Natl Acad Sci U S A. 2000;97(22):11936-11941.

(277) Dumaz N, Meek DW. Serine15 phosphorylation stimulates p53 transactivation but does not directly influence interaction with HDM2. EMBO J. 1999;18(24):7002-7010.

(278) Berti C, Fontanella B, Ferrentino R, Meroni G. Mig12, a novel Opitz syndrome gene product partner, is expressed in the embryonic ventral midline and co-operates with Mid1 to bundle and stabilize microtubules. BMC Cell Biol. 2004;5(1):9.

(279) Farras R, Ferrando A, Jasik J et al. SKP1-SnRK protein kinase interactions mediate proteasomal binding of a plant SCF ubiquitin ligase. EMBO J. 2001;20(11):2742-2756.

(280) Budanov AV, Sablina AA, Feinstein E et al. Regeneration of peroxiredoxins by p53-regulated sestrins, homologs of bacterial AhpD. Science. 2004;304(5670):596-600.

(281) Liang P, MacRae TH. Molecular chaperones and the cytoskeleton. J Cell Sci. 1997;110 ( Pt 13):1431-1440.

(282) Cassimeris L, Skibbens RV. Regulated assembly of the mitotic spindle: a perspective from two ends. Curr Issues Mol Biol. 2003;5(3):99-112.

(283) Huang Q, Raya A, DeJesus P et al. Identification of p53 regulators by genome-wide functional analysis. Proc Natl Acad Sci U S A. 2004;101(10):3456-3461.

(284) Fialka I, Pasquali C, Kurzbauer R et al. Loss of epithelial polarity is accompanied by differential association of proteins with intracellular membranes. Electrophoresis. 1999;20(2):331-343.

(285) Hosono K, Sasaki T, Minoshima S, Shimizu N. Identification and characterization of a novel gene family YPEL in a wide spectrum of eukaryotic species. Gene. 2004;340(1):31-43.

(286) Freed E, Lacey KR, Huie P et al. Components of an SCF ubiquitin ligase localize to the centrosome and regulate the centrosome duplication cycle. Genes Dev. 1999;13(17):2242-2257.

(287) Miura M, Hatakeyama S, Hattori K, Nakayama K. Structure and expression of the gene encoding mouse F-box protein, Fwd2. Genomics. 1999;62(1):50-58.

200 (288) Yu R, Heaney AP, Lu W et al. Pituitary tumor transforming gene causes aneuploidy and p53-dependent and p53-independent apoptosis. J Biol Chem. 2000;275(47):36502-36505.

(289) Tyler JK, Adams CR, Chen SR et al. The RCAF complex mediates chromatin assembly during DNA replication and repair. Nature. 1999;402(6761):555-560.

(290) Roh MH, Fan S, Liu CJ, Margolis B. The Crumbs3-Pals1 complex participates in the establishment of polarity in mammalian epithelial cells. J Cell Sci. 2003;116(Pt 14):2895-2906.

(291) Adham IM, Sallam MA, Steding G et al. Disruption of the pelota gene causes early embryonic lethality and defects in cell cycle progression. Mol Cell Biol. 2003;23(4):1470-1476.

(292) Criswell T, Klokov D, Beman M et al. Repression of IR-inducible clusterin expression by the p53 tumor suppressor protein. Cancer Biol Ther. 2003;2(4):372-380.

(293) Vairapandi M, Balliet AG, Hoffman B, Liebermann DA. GADD45b and GADD45g are cdc2/cyclinB1 kinase inhibitors with a role in S and G2/M cell cycle checkpoints induced by genotoxic stress. J Cell Physiol. 2002;192(3):327-338.

(294) Sandoval MT, Zurita M. Increased UV light sensitivity in transgenic Drosophila expressing the antisense XPD homolog. Antisense Nucleic Acid Drug Dev. 2001;11(2):125-128.

(295) Bai L, Merchant JL. ZBP-89 promotes growth arrest through stabilization of p53. Mol Cell Biol. 2001;21(14):4670-4683.

(296) Buckbinder L, Velasco-Miguel S, Chen Y et al. The p53 tumor suppressor targets a novel regulator of G protein signaling. Proc Natl Acad Sci U S A. 1997;94(15):7868-7872.

(297) Keryer G, Di Fiore B, Celati C et al. Part of Ran is associated with AKAP450 at the centrosome: involvement in microtubule-organizing activity. Mol Biol Cell. 2003;14(10):4260-4271.

(298) Keryer G, Witczak O, Delouvee A et al. Dissociating the centrosomal matrix protein AKAP450 from centrioles impairs centriole duplication and cell cycle progression. Mol Biol Cell. 2003;14(6):2436-2446.

(299) Smith CL, O'Malley BW. Coregulator function: a key to understanding tissue specificity of selective receptor modulators. Endocr Rev. 2004;25(1):45-71.

201 (300) Kastner P, Krust A, Turcotte B et al. Two distinct estrogen-regulated promoters generate transcripts encoding the two functionally different human progesterone receptor forms A and B. EMBO J. 1990;9(5):1603- 1614.

(301) Savouret JF, Bailly A, Misrahi M et al. Characterization of the hormone responsive element involved in the regulation of the progesterone receptor gene. EMBO J. 1991;10(7):1875-1883.

(302) Kraus WL, Montano MM, Katzenellenbogen BS. Cloning of the rat progesterone receptor gene 5'-region and identification of two functionally distinct promoters. Mol Endocrinol. 1993;7(12):1603-1616.

(303) Cahill DP, Kinzler KW, Vogelstein B, Lengauer C. Genetic instability and darwinian selection in tumours. Trends Cell Biol. 1999;9(12):M57-M60.

(304) Mussman JG, Horn HF, Carroll PE et al. Synergistic induction of centrosome hyperamplification by loss of p53 and cyclin E overexpression. Oncogene. 2000;19(13):1635-1646.

(305) Planas-Silva MD, Weinberg RA. Estrogen-dependent cyclin E-cdk2 activation through p21 redistribution. Mol Cell Biol. 1997;17(7):4059-4069.

(306) Prall OW, Sarcevic B, Musgrove EA et al. Estrogen-induced activation of Cdk4 and Cdk2 during G1-S phase progression is accompanied by increased cyclin D1 expression and decreased cyclin-dependent kinase inhibitor association with cyclin E-Cdk2. J Biol Chem. 1997;272(16):10882-10894.

(307) Liu S, Mirza A, Wang L. Generation of p53 target database via integration of microarray and global p53 DNA-binding site analysis. Methods Mol Biol. 2004;281:33-54.

(308) Mirza A, Wu Q, Wang L et al. Global transcriptional program of p53 target genes during the process of apoptosis and cell cycle progression. Oncogene. 2003;22(23):3645-3654.

(309) Kannan K, Amariglio N, Rechavi G et al. DNA microarrays identification of primary and secondary target genes regulated by p53. Oncogene. 2001;20(18):2225-2234.

(310) Anazawa Y, Arakawa H, Nakagawa H, Nakamura Y. Identification of STAG1 as a key mediator of a p53-dependent apoptotic pathway. Oncogene. 2004;23(46):7621-7627.

(311) Burns TF, El Deiry WS. Microarray analysis of p53 target gene expression patterns in the spleen and thymus in response to ionizing radiation. Cancer Biol Ther. 2003;2(4):431-443.

202 (312) Li CQ, Robles AI, Hanigan CL et al. Apoptotic signaling pathways induced by nitric oxide in human lymphoblastoid cells expressing wild-type or mutant p53. Cancer Res. 2004;64(9):3022-3029.

(313) Nakamura Y. Isolation of p53-target genes and their functional analysis. Cancer Sci. 2004;95(1):7-11.

(314) Yoon KA, Nakamura Y, Arakawa H. Identification of ALDH4 as a p53- inducible gene and its protective role in cellular stresses. J Hum Genet. 2004;49(3):134-140.

(315) Miyashita T, Reed JC. Tumor suppressor p53 is a direct transcriptional activator of the human bax gene. Cell. 1995;80(2):293-299.

(316) Oda E, Ohki R, Murasawa H et al. Noxa, a BH3-only member of the Bcl-2 family and candidate mediator of p53-induced apoptosis. Science. 2000;288(5468):1053-1058.

(317) Nakano K, Vousden KH. PUMA, a novel proapoptotic gene, is induced by p53. Mol Cell. 2001;7(3):683-694.

(318) Ajiro K, Yoda K, Utsumi K, Nishikawa Y. Alteration of cell cycle-dependent histone phosphorylations by okadaic acid. Induction of mitosis-specific H3 phosphorylation and chromatin condensation in mammalian interphase cells. J Biol Chem. 1996;271(22):13197-13201.

203