The Mechanisms of BPA Exposure and in the Developing Mammary Gland

The mechanisms of BPA exposure and in the developing mammary gland

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of

Philosophy in the Graduate School of The Ohio State University

Andrea R. Hindman

Graduate Program in Molecular, Cellular and Developmental Biology

The Ohio State University

2017

Dissertation Committee:

Craig J. Burd, Advisor

Helen M. Chamberlin

Ruth A. Keri (Case Western Reserve University)

Thomas Ludwig

Mark R. Parthun

Copyrighted by

Andrea R. Hindman

2017

Abstract

It is estimated that greater than 40,000 women in the United States will die from breast cancer this year and over 250,000 will be newly diagnosed. Greater than 70% of these cases are attributable to environmental factors, some of which are suspected to be a consequence of extensive human exposure to estrogenic compounds, collectively known as endocrine disrupting compounds (EDCs). Significant evidence from animal models suggests that chronic and/ or early-life exposure to known EDCs, such as bisphenol A

(BPA) and genistein (GEN) are linked to increased risk and mediators of epithelial transformation to impart later-life development of breast cancer. However, the consequences of vast human exposures are largely unknown, immeasurable and the molecular mechanisms are largely undefined. Herein, we have utilized two distinct models to address undefined facets of the EDC mechanism. We address molecular mechanisms of chronic exposures that are representative of observed human exposures and examine windows of differential in utero BPA exposure to narrow possible developmental mechanisms at time of exposure that may propagate EDC-mediated alterations to later-life. We emphasize the importance of understanding the transcriptional alterations and epigenetic landscape of adult mammary component cell types, given the primary focus of current work to be on embryonic manifestation of disease and susceptibility. Our analyses demonstrate sustained reprogramming of the estrogen response beyond cessation of chronic exposures. Further we attribute the well-

characterized BPA in utero phenotype to a period in mammary gland development corresponding to discrete tissue compartment interactions, corroborating the vital paracrine signaling that occurs at this in utero time of exposure. Our analyses are the first of its kind to map transcriptional and epigenetic alterations following in utero BPA exposure in the adult mammary gland. Taken together, our findings can inform analysis of human populations, determining the mechanisms of deregulation by in utero BPA exposures that contribute to later-life breast cancer risk and encourage alternative compounds. We believe while these mechanisms have been investigated specific to BPA, our findings can be extended to and provide a basis of concern for other known and suspected EDCs.

iii

Dedication

Dedicated, with love to my mother, Grace Ann,

whom at the end of this journey I always believed I

would have close to me celebrating this milestone

and beyond. Throughout my life, she gave me the

confidence and any and all resources she could to

ensure I had the best chance of accomplishing anything I set out to do, no matter how fleeting some of those aspirations were. Only recently have I realized how much she gave me and my sister. When I had the opportunity to ask her, in the hopes I could raise my children with similar courage and confidence that she cultivated in us, she simply said ‘I just wanted to make sure you knew you could do anything. You never know until you try.’ So I continue to try without you now and hope I can come close to the example you set forth in never settling on anything less than what you believed you could do and never taking no for an answer.

I also dedicate this work to my wonderfully supportive husband that I credit for my stamina towards completing my doctoral work. I could have never done it without your love and support and I look forward to our next adventure together, in happier times, welcoming our son, Evan, and pursuing the careers we both have imagined for ourselves.

I love you with all my heart and I could not have asked for a better partner in life (PIL).

Acknowledgments

I am grateful to have an army of support surrounding me. My Dad - for instilling in me hard work, dedication and respect for all unless they prove otherwise. Kara Barrett

– ever since that first phone call you dropped on me as I sat in my lonely apartment my first year of graduate school, you have been a constant source of empowerment. I have always counted on you to bounce off any experience I am faced with, not feel judged but rather understood and delivered with blunt-force honesty. Janelle Gabriel – through 3 years of being roommates, I am forever indebted to your friendship in science, brownies and reality television. There are too many more to thank but I will continue to hold you all close. Finally, my sister Emily Hickey and her kids - for their support whenever called upon. With love, thank you all so very much.

Importantly, my advisor Craig - who took a chance on me, bestowing the highest honor of being his first graduate student. He showed me the greatest measure of compassion at a time when it was most needed. His patience, attentiveness and ongoing commitment to supporting his trainees has instilled pride in me for having worked for him and given me the confidence and tools to face the next steps in my career. Thanks.

My amazing undergraduate mentees - Karen Wernke, Hannah Helber, Nanditha

Ravichandran and Claire Kovalchin who I admire and have contributed to a great amount of what is presented herein. Thank you all.

Finally, I am grateful to my committee members, my program and my fellow graduate students, particularly Lindsey Anstine who provided guidance for me. Gina

Sizemore - an invaluable resource in science and life. The labs of Ostrowski and Leone – they provided a vast amount of knowledge and resources among its members. Our dedicated lab technicians Alina Murphy and Ali Shapiro – both of whom, together with the Ostroiwski and Leone labs, I thank for their support and I credit with the progress I was able to make in the projects presented herein.

Vita

June 2006 Niagara Falls High School

Feb 2011 B.S. Biological Sciences, B.A. Chemistry, the State University of

New York at Buffalo

2012 to present Graduate Research Fellow, Molecular, Cellular and Developmental

Biology Program, Department of Molecular Genetics, The Ohio

State University

Publications

Hindman AR, Mo XM, Helber HL, Kovalchin CE, Ravichandran N, Murphy AR,

Kladney RD, Fagan AM, St. John PM and Craig J. Burd. Defining the window of female mammary gland susceptibility to in utero BPA exposure. Endocrinology. (ACCEPTED.

August 2017).

McAdams NM, Patterson AR and Gollnick P. Identification of a residue (Glu60) in

TRAP required for inducing efficient transcription termination at the trp attenuator independent of binding tryptophan and RNA. J. Bacteriol. 2017 Jan. doi:

10.1128/JB.00710-16.

vii

Patterson AR, Mo X, Shapiro A, Wernke KE, Archer TK, Burd CJ. Sustained reprogramming of the estrogen response following chronic exposure to endocrine disruptors. Mol Endocrinol. 2015 Jan 16:me20141237.

Bayfield OW, Chen CS, Patterson AR, Luan W, Smits C, Gollnick P, Antson AA. Trp

RNA-binding attenuation protein: modifying symmetry and stability of a circular oligomer. PLoS One. 2012;7(9):e44309. doi: 10.1371/journal.pone.0044309.

Fields of Study

Major Field: Molecular, Cellular and Developmental Biology

viii

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments...... v

Vita ...... vii

Publications ...... vii

Fields of Study ...... viii

Table of Contents ...... ix

List of Tables ...... xvi

List of Figures ...... xvii

Chapter 1: Introduction ...... 1

1.1 Breast cancer outlook and incidence...... 1

1.1.1 Statistics and factors of risk...... 1

1.1.2 Incidence influenced by industrialization and lifestyle...... 1

1.2 Estrogen action and its receptors...... 2

1.2.1 The sex hormone estrogen...... 2

1.2.2 Estrogen receptor alpha (α) is the key regulator of breast (mammary)

development...... 3 ix

1.2.3 Breast cancer the disease, treatment and recurrence...... 4

1.3 Endocrine disrupting compounds...... 6

1.3.1 Lessons learned in the vulnerability of hormonal signaling and breast cancer

risk: Diethylstilbestrol...... 6

1.3.2 The phenomenon of low-dose EDC exposures...... 7

1.3.3 The omnipresence of industrial EDC Bisphenol A: chronic human exposure. .. 8

1.3.4 BPA is postulated to contribute to breast cancer incidence, similar to DES...... 9

1.4 Mammary gland development...... 11

1.4.1 The mammary gland is a complex and heterogeneous glandular organ...... 11

1.4.2 The susceptibility of key mammary gland developmental stages to endocrine

disruption...... 14

1.4.3 The importance of epithelial-stromal interactions in mammary gland

development...... 15

1.5 Gaps in the field of endocrine disruption...... 16

1.5.1 Chronic EDC exposures to recapitulate human exposures...... 16

1.5.2 Elucidation of a poorly understood mechanism through examination of key

stages of gland development...... 17

1.5.3 The potential of component cell types of the adult mammary gland to reveal the

mechanism of in utero BPA missed by whole tissue analyses...... 18

1.5.4 Are the observed latent consequences of in utero exposure propagated by

epigenetic mechanisms? ...... 18

1.6 Research Purpose ...... 19

Chapter 2: Sustained Reprogramming of the Estrogen Response Following Chronic

Exposure to Endocrine Disruptors2 ...... 21

2.1 Abstract ...... 21

2.2 Introduction ...... 22

2.3 Materials and Methods ...... 24

2.4 Results ...... 28

2.4.1 Chronic EDC treatment reprograms global gene expression...... 28

2.4.2 Chronic EDC treatment reprograms E2-regulated gene expression...... 29

2.4.3 Deregulated ER response is independent of the activating ligand...... 32

2.4.4 Chronic EDC treatment causes differential ER recruitment...... 33

2.4.5 The EGR3 locus is epigenetically reprogrammed in EDC-treated cell lines. .. 34

2.4.6 Estrogen signaling defects persist after EDC removal...... 36

2.5 Discussion ...... 38

2.5.1 A model of chronic exposure...... 38

2.5.2 Chronic EDC exposure deregulates ER signaling in a compound- and gene-

specific manner...... 38

2.5.3 Chronic EDC deregulation persists in a compound- and gene-specific manner.

...... 40

2.5.4 Conclusions...... 41

Chapter 3: Varying susceptibility of the female mammary gland to in utero windows of

BPA exposure3 ...... 55

3.1 Abstract ...... 55

3.2 Introduction ...... 56

3.3 Materials and Methods ...... 59

3.4 Results ...... 67

3.4.1 In utero ERα expression in the developing mammary gland is strictly

mesenchymal...... 67

3.4.2 Discrete in utero windows of BPA treatment cause varying degrees of

mammary gland phenotypes in the developing gland...... 68

3.4.3 Varied in utero windows of BPA treatment change later-life ERα expression in

the tissue compartments of the mammary gland...... 71

3.4.4 In utero BPA exposure alters ductal morphology to varying degrees based upon

the window of treatment...... 72

3.5 Discussion ...... 73

3.5.1 The susceptibility of the mammary gland to BPA exposure is not uniform

throughout gestation...... 73

xii

3.5.2 The role of ERα in mediating the effects of in utero BPA exposure...... 75

3.5.3 The impact of stromal signaling on mammary gland defects...... 78

3.5.4 Conclusions: Different windows of varying mammary gland susceptibility to in

utero BPA exposure...... 80

Chapter 4: BPA in utero exposure alters later-life mammary development through deregulated transcriptional programming and extracellular matrix remodeling4 ...... 92

4.1 Abstract ...... 92

4.2 Introduction ...... 93

4.3 Materials and Methods ...... 97

4.3.1 Animals...... 97

4.3.2 Harvest, isolation and preparation of mouse mammary component cell types. 97

4.3.3 RNA-sequencing, validation, and analysis...... 98

4.3.4 Tissue preparation and analyses of mammary gland ECM components

following in utero exposure to BPA...... 99

4.3.5 Hydraulic permeability measurements performed with microfluidic dye...... 100

4.4 Results ...... 102

4.4.1 Component mammary cell types are effectively isolated following in utero

BPA exposure...... 102

4.4.2 Limited overlap of differentially regulated genes following in utero BPA

exposure between mammary component cell types...... 104

xiii

4.4.3 Pathway analysis reveals ECM organization and structure significantly

deregulated following in utero BPA exposure...... 111

4.4.4 In vitro collagen organization altered by in utero BPA exposed mammary

fibroblasts...... 112

4.4.5 ECM remodeling revealed through decreased hydraulic permeability as a result

of in utero BPA exposure...... 113

4.5 Discussion ...... 114

4.5.1 Tissue architecture and signaling dictate development...... 114

4.5.2 ECM remodeling directs postnatal mammary gland development and sets the

stage for tumorigenesis...... 116

4.5.3 Disruption of ECM homeostasis can be a pre-requisite or consequence of

tumorigenesis, imparting a mechanism for BPA propagation of later-life disease

susceptibility...... 117

4.5.4 Reduced hydraulic permeability indicates a more dense tissue composition. 118

4.5.5 Additional work to implicate BPA in mammary gland remodeling to promote

later-life susceptibility...... 119

Chapter 5: Concluding remarks, unpublished data and future directions...... 128

5.1 Summary, impact and innovation...... 128

5.2 Future directions...... 131

xiv

5.2.1 Utilizing in vitro methods to demonstrate altered molecular signaling by EDCs.

...... 131

5.2.2 Tissue-specific signaling in development and the requirement of ERα in the

mesenchyme to elicit later-life susceptibility...... 133

5.2.3 Epigenetic mechanism of BPA...... 134

5.3 Concluding remarks...... 138

References ...... 145

List of Tables

Table 2.1 cDNA primers utilized for transcriptional genes of interest.……...... ……..26

Table 2.2 ERE primers used for ER recruitment and DNA accessibility...……………..28

Table 3.1 Antibodies used for staining…………..………………………………………66

Table 3.2 Sholl analysis of epithelial branching density and complexity………….…....89

Table 4.1 - Gene signatures representative of purified adult mouse mammary gland cell types used to validate isolated cell types…………………………………………….…104

Table 4.2 - Luminal epithelial cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands…………………………………………106

Table 4.3 - Basal epithelial cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands…………………………………………………..107

Table 4.4 - Fibroblast cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands…………………………………………………...109

xvi

List of Figures

Figure 2.1 Chronic EDC treatment reprograms global gene expression………..…….…43

Figure 2.2 Chronic EDC treatment reprograms E2-regulated gene expression………....45

Figure 2.3 Deregulated ER response is independent of the activating ligand…...... …47

Figure 2.4 Chronic EDC treatment causes differential recruitment of ER………….…..48

Figure 2.5 Epigenetic reprogramming is seen at EGR3 response elements……….…….49

Figure 2.6 Changes in the E2 response persist after removal of the EDC………....……51

Figure 2.7 Histone protein expression……………..………………………………….....53

Figure 2.8 Chronic EDC treatment causes differential recruitment of ER……………....54

Figure 3.1 - ERα expression is strictly mesenchymal throughout perinatal murine mammary gland development……………...……………………………………………82

Figure 3.2 – BPA causes early-life defects in epithelial elongation within the mammary gland. …………………………………….……………………………………………...83

Figure 3.3 – Early-life epithelial dysfunction of the mammary gland by in utero BPA significantly correlates to the stromal compartment…………………………………….85

Figure 3.4 - In utero BPA changes later-life ERα expression of mammary tissue compartments. …………………………………………………...……………………...87

Figure 3.5 - In utero BPA significantly alters the density and complexity of the mammary epithelium………………………………………………………………………………..88

xvii

Figure 3.6. Negative staining controls for IHC……………...……………………....…..90

Figure 3.7 Representative data from Sholl analysis for each in utero treatment.

…………………………………………………………………………………………...91

Figure 4.1 – Discrete isolation of adult mammary component cell types following in utero BPA exposure……………………………………………..……………………...121

Figure 4.2 – Distinct transcriptional changes across adult component mammary cell types result in little overlap…………………………………………………………………...123

Figure 4.3 – Significant decrease in collagen density within the adult mammary tissue………………………………………………………………………………...….126

Figure 4.4 – Significant decrease in hydraulic permeability of an in vitro collagen matrix remodeled by in utero BPA-treated, adult mammary fibroblasts……….………...……127

Figure 5.1 – All significantly differentially methylated regions in the adult luminal epithelial cell population completely hypomethylated as a result of in utero BPA exposure……………………………………………………………………………..…140

Figure 5.2 – Differentially methylated regions of mammary gland component cell types suggest epigenetic reprogramming as a result of in utero BPA exposure……………………………………………………………………..…………141

Figure 5.3 – Differentially H3K27-trimethylated regions of adult mammary component epithelial cell types reveals a vastly hypermethylated luminal epithelial cell population following in utero BPA exposure and indicative of transcriptional repression………..143

xviii

Chapter 1: Introduction

1.1 Breast cancer outlook and incidence.

1.1.1 Statistics and factors of risk.

It is estimated that greater than 40,000 women in the United States will die from breast cancer this year and over 250,000 will be newly diagnosed [1]. Indeed as of 2017, breast cancer accounts for over 30% of cancers newly diagnosed in women and will be the second leading cause of death in women [1]. According to the National Cancer

Institute (NCI), the main risk factors for breast cancer development are gender, age and genetic alteration [2]. Importantly, genetic predispositions such as BRCA1 and 2 mutations, account only for 5-10% of breast cancers [3]. Thus, there could be many factors including lifestyle and environment that may influence breast cancer incidence.

Extensive epidemiological studies have further demonstrated the minor role genetics play in breast cancer susceptibility through studies of twin cohorts [4, 5].

1.1.2 Incidence influenced by industrialization and lifestyle.

Most staggering is the difference in documented incidences of breast cancer between industrialized countries compared to that of developing countries [6-9]. A 5-fold increase in the number of new cases of breast cancer among the United States and

Western Europe compared with Africa and Asia has been reported [10]. As a result, breast cancer has been deemed a disease that “afflicts the affluent,” or more Westernized

cultures [11, 12] with incidence rates reported by the World Health Organization to be highest in the United States (91/100,000) and lowest in countries located in East Asia

(18/100,000) (age-standardized) [13].1 While many factors within the realm of lifestyle and reproductive behaviors are accepted, the main underlying environmental causes remain to be identified and could be the result of a pervasive environmental mixture of hormonally active compounds.

1.2 Estrogen action and its receptors.

1.2.1 The sex hormone estrogen.

One undeniable contributor to breast cancer risk is that of life-long exposure to the endogenous steroid sex hormone estrogen [14-16]. Estrogen has many targets throughout the human body, including the brain, heart, liver, bone, and reproductive tract

[17]. Simply, the hypothalamus-pituitary-gonadal (HPG) axis controls estrogen synthesis through release of luteinizing hormone (LH). LH is targeted to theca cells within the female ovary, leading to the conversion of androgens to estrogen via the enzyme aromatase [18].

______

1 Important to note: While the disease itself has been shown to have greater incidence in industrialized countries, social disparities have provided context to greater rates of mortality among poorer populations within these countries; limited access to resources, diagnostic screening, prevention and treatment have sadly proven to be obstacles.

The HPG finely tunes the amount of circulating hormones and the typical levels of total endogenous estrogen in the premenopausal female body are in the range of 20 - 800 pg/mL (approximately 0.073 - 2.9 nM) and < 20 pg/mL in prepubertal females [19-21]. It has been clearly demonstrated that estrogen signaling is essential to the growth and maintenance of normal breast development and reproductive health. It is not surprising then, that approximately 70% of breast cancers are dependent upon estrogen signaling for growth and survival [22, 23]. Canonically, estrogen binds and activates the estrogen receptor (ER). Upon activation, ER trans-locates into the nucleus and is recruited to directly interact with DNA at estrogen response elements (EREs) or through binding other transcriptions factors like AP-1 [24] and transcriptional co-regulators (CoRs) [25-

27]. Activated ER recruits additional CoRs to further regulate proliferation and target gene expression. The conformational change in the receptor upon ligand binding facilitates the binding of these CoR proteins [28-30] and can dictate recruitment to specific EREs depending on the ligand or CoRs present.

1.2.2 Estrogen receptor alpha (α) is the key regulator of breast (mammary) development.

The ER is a member of the nuclear receptor super-family of transcription factors.

This family of transcription factors is defined by their ligand-dependent mechanisms

[31]. There are two subtypes, ERα and ERβ, which have distinct tissue and temporal expressions, and have been demonstrated by multiple genetic models to have non- redundant functions in development. ERα is the predominant regulator of mammary gland development compared to ERβ [27, 32]. In mice, the knock-out (KO) of ERα is not

lethal, however the mammary gland does not develop beyond the rudimentary morphogenesis of the epithelium [32], the reproductive tissues are unresponsive to hormone [32] and these females are infertile [33]. ERβ KO mice develop normally and maintain E2-responsiveness suggesting that ERα is the critical mediator of breast development and function. Further, several studies have concluded that ERα is critical in breast cancer progression due to an observed increase in the ERα/ERβ ratio [34, 35]. It is clear that ER has many sources of promiscuity in its activity, through ligand-binding and recruitment of CoRs, and even through different receptor subtypes. In fact, the binding pocket of ER has been observed to be accommodating to a large number of compounds, outside of its canonical ligand estrogen [36]. As a result, there are many opportunities for receptor activity to be deregulated due to estrogen-like compounds in the environment, deregulating breast development and later-life breast cancer susceptibility.

1.2.3 Breast cancer the disease, treatment and recurrence.

Breast cancer is the most common disease among Western women [1, 11, 37].

The factors regulating the breast including the hormone estrogen and its predominant receptor ERα, confer much complexity to the development and proliferation of this glandular organ and likewise the progression of disease. The breakdown of the disease pathogenesis is comprised of histological and molecular subtypes. Most simply, the breast consists of two tissue compartments: the epithelium and stroma. The histological subtypes of breast cancer are ductal carcinoma and lobular carcinoma, both structures of the epithelium. The most common molecular subtype is receptor positive, or hormone- responsive, and it encompasses approximately 70% of breast cancer cases. This subtype

expresses the estrogen and/ or progesterone (PR) receptors [22, 23]. The other molecular subtypes comprising the remaining approximately 30% of cases include those cancers mediated by human epidermal growth factor receptor (HER2; in some cases co-expressed with ER and/ or PR) and finally, receptor negative or triple negative breast cancer.

Receptor-positive cancers expressing ER and PR generally have a better prognosis; their responsiveness to hormone makes these cancers receptive to endocrine therapies targeting these receptors including anti-estrogens, aromatase inhibitors and selective estrogen receptor -modulators (SERMs) and -downregulators (SERDs). Since many of the proliferative effects of estrogens in breast cancer have been attributed to ERα, it is routinely utilized in pathological diagnosis for the purposes of treatment [38, 39].

Sadly, despite advances in earlier detection and treatments, approximately 20-

30% of all breast cancer patients will relapse with distant metastases [40] or develop endocrine-resistant disease [41, 42]. This probability of developing metastatic breast cancer appears to be stable despite advances and early detection [43], underscoring it as the second leading cause of death in women in the U.S. Further, breast cancer is unusual in that the manifestation of distant metastases is not only common, but continues to be a threat to patients beyond 10 years following the initial diagnosis [44]. A complete loss of

ER function is rare in endocrine-resistant disease and progression [45]. Concomitantly,

ER remains responsive to additional lines of endocrine therapies, suggesting that the receptor activity albeit altered, remains a key regulator of progression.

It is clear that a better understanding of the mechanisms dictating hormone- responsiveness and mediation of the ER in breast cancer are needed in order to identify

and therapeutically target disrupted pathways in metastatic disease. Given that over 70% of breast cancers are influenced by the environment and that the predominant receptor,

ERα, can be promiscuous in its ligand-binding activity, sets the stage for alternative mediators of ER activation to play a role in these poorly understood mechanisms and the implications for life-long susceptibility, disease and recurrence.

1.3 Endocrine disrupting compounds.

1.3.1 Lessons learned in the vulnerability of hormonal signaling and breast cancer risk:

Diethylstilbestrol.

Environmental compounds known to mimic and interfere with normal hormonal signaling have been classified as endocrine disrupting compounds (EDCs) [46-49]. This diverse class of compounds range from those industrially produced and more natural plant components, specifically termed phytoestrogens [50-52]. The ER has been shown to have affinity for and altered activity with many of these environmental contaminants [36,

53], thereby demonstrating their estrogenic potential. Life-long exposure to estrogen is a risk factor for breast cancer development [14-16]. The presence of these all-pervading compounds in our everyday environments has been postulated to contribute to our overall risk for the disease.

A tragic demonstration of the consequence of such exogenous estrogenic exposure began in the 1940’s with the administration of a known synthetic estrogen, diethylstilbestrol (DES) to pregnant women with the intent to prevent miscarriage.

Unfortunately, it was not until decades later that the administration of DES was

discontinued when daughters exposed in utero – “DES Daughters” – had a 40x increased incidence of clear cell adenocarcinoma of the vagina and cervix and a 2-fold increased incidence of breast cancer risk [54-56]. The risk for breast cancer did not manifest until after the age of 40 years old in these women. Interestingly, a significantly increased risk of breast cancer was also observed in the mothers’ administered the compound while pregnant [57, 58], asserting the range of hormonal consequences as a result of estrogenic exposures at vulnerable times in reproductive development and estrogen signaling. For these reasons, DES is classified not only as an EDC but as a known carcinogen [59]. The cautionary tale of in utero DES highlights the vulnerability of the in utero time point of exposure as well as the latency of later-life susceptibility. This latency in manifestation of disease presents one of the many unknown components of the endocrine disrupting mechanism and lends to the complexity of estrogen signaling relating to treatment and recurrence as well.

1.3.2 The phenomenon of low-dose EDC exposures.

Normal hormonal signaling operates at very discrete and finely tuned concentrations [19-21]. It can be imagined then, that the availability of alternative compounds capable of interacting with the endogenous receptors, such as EDCs, has the ability even at low concentrations to disrupt the normal hormonal signaling program.

Contrary to classical toxicological risk assessment where bulbous concentrations of a suspected human toxicant are tested to determine a safe dose, part of the complexity of endocrine disruption is the ability of EDCs to have affects at much lower than expected concentrations (as reviewed in [60]). In 2001, the Environmental Protection Agency

(EPA) charged the National Toxicology Program (NTP) to perform a review of low-dose literature and a selected explanation of their definition of low-dose is as follows: “a dose below the lowest dose at which a biological change (or damage) for a specific chemical has been measured in the past” and those doses that are measured in the human population (not through occupational exposures) [61].

An additional complexity of endocrine disruption is the dose-response curves associated with these compounds. Again, contrary to traditional toxicological assays that typically test candidate toxicants according to a linear dose-response with the expectation that the higher the dose, the more detrimental the biological effect; EDCs typically exhibit a non-monotonic dose response (NMDR) [60]. A NMDR can undulate between a positive and/ or negative slope and does not adhere to a consistently linear progression with increased dose. This is further confounding because if and when compounds are tested for a biological response, it cannot be assumed that at lower or higher doses the response can be extrapolated. Thus, many EDC studies must evaluate a range of concentrations in order to assess the biological implications of exposure [62]. Ultimately and as expected, while this phenomenon is attributed to the complexity of EDCs, it originates from the observation that endogenous hormones are responsive in this way.

There are many potential mechanisms that attempt to explain these phenomena; however, exactly how it occurs is poorly understood.

1.3.3 The omnipresence of industrial EDC Bisphenol A: chronic human exposure.

Synthetic estrogens, like DES, are no longer prescribed to pregnant women.

However, the increased use of an industrially produced EDC, bisphenol A (BPA), has

become a concern as it relates to breast cancer incidence. Worldwide it is estimated, by the Center for Disease Control and Prevention (CDC) National Biomonitoring Program that 5-6 billion pounds of BPA are produced every year and there is a percentage of growth in production observed every year [63] (reviewed in [61]). Specifically, BPA is a plasticizer that can be found in most consumer products, including but not limited to, plastic water and baby bottles, resins lining the inside of canned goods and other food packaging, dental sealants, medical equipment, thermal receipts and much more [48, 64].

As a consequence, BPA has been detected in the urine of greater than 95% of the U.S. population [65].

A diverse sampling of human biological tissues and fluids have reported levels of

BPA to be typically in the 1 – 10 nM range but as high as 80 nM [20], conferring extensive and continuous exposures. Most concerning are the levels of BPA found in human pregnancy-associated fluids including amniotic fluid [66-68] and demonstrations of this EDC to cross the placental barrier [69, 70]. Finally, toxicokinetic studies in rodents, non-human primates and humans have determined the biological half-life of

BPA to be approximately 6 hours [71-74], underscoring the constant exposure to this

EDC in our environments if such levels are consistently detectable in humans.

Most studies with the intent to evaluate the EDC effects carry out analyses following short and discrete times of exposure. However, it is clear that human exposures are chronic and occur over the entirety of an individual’s life.

1.3.4 BPA is postulated to contribute to breast cancer incidence, similar to DES.

Epidemiological studies have demonstrated that industrialized countries like the

US, which heavily produce and rely on the utility of BPA, have increased incidence of breast cancer [6-9]. Notably, studies of migrants from low- (Asian) to high-risk (U.S.) populations had increased breast cancer risk in succeeding generations [75, 76]. While these studies implicate lifestyle changes, presumably increased exposures to industrialized consumer products, one cannot discount other factors these studies neglect to account for. Finally, the prolific use of BPA industrially, only dawned between the

1970s and 80s and thus the consequence of human exposure has not been fully realized as with DES, which took decades for the increased risk to manifest. Thus, while the effects of human BPA exposure are not well-defined, in vitro and in vivo models have done well to fill in the gaps.

As an industrial estrogen and like DES, the EDC BPA is structurally capable of interacting with ER and causing its own unique transcriptional programs [36, 53]. BPA has demonstrated alterations to mammary gland development and morphology in rodent models and in the rhesus monkey [77-79]; alterations to structures and signaling which make the organ more susceptible to later-life breast cancer risk. However, like most

EDCs these effects are dose- and time-of-exposure- dependent and in the case of BPA have been shown to have a NMDR as well. Generally, these alterations manifest as a prolonged period of nascent structures that make the gland more susceptible to carcinogens [80] or alternatively, increased differentiation, volume and proliferation of the adult epithelium [77].

The propensity for disease manifestation as a result of exposure to BPA has been well-characterized in mouse models [77, 78, 80-87], demonstrating increased cancer susceptibility similar to that seen following treatment with DES [88-90]. In these mouse studies, while BPA alone has not been enough to invoke breast cancer development, exposures have been shown to increase susceptibility through the development of ductal hyperplasias [91], and result in a higher incidence of tumors in models using predisposed transgenic alterations (NCR nu/nu, MMTV-erbB2) [92] or through a second assault by a known carcinogen (such as 7,12-dimethylbenz[a]anthracene]) [93]. Similar and more direct consequences have been achieved in rats, where perinatal BPA exposure was sufficient to induce malignancies and lesions [94].

The mechanisms underlying BPA disruption of normal breast (mammary) development and estrogen signaling have yet to be elucidated, though the phenotypic consequences are clear. Taken together with demonstrated human exposures, the necessity to define how environmental components, presumably BPA as a major player, contribute to human risk for breast cancer. The adverse effects upon incidence and recurrence of disease as well as the efficacy of current therapies are at stake.

1.4 Mammary gland development.

1.4.1 The mammary gland is a complex and heterogeneous glandular organ.

Breast cancer is a complex disease and much of its complexity derives from normal breast development, component cell types and morphology. Much of what is known about breast development is deduced from studies in mice. During

embryogenesis, the mouse fetal mammary gland goes through several developmental stages that are independent of hormone activity [95-98]. Specifically, ERα knockout mice have normal mammary glands at birth demonstrating that the receptor is only required after birth for mammary development [99, 100]. The initiating event occurs approximately at embryonic day (E) 10 [101], where along the mammary or milk lines that run in a rostro-caudal direction on the ventral surface of the body, primordial epithelial buds form [102]. In the days following, E12-13, the epithelial bud increases in size and invaginates into the underlying dermis. The cells in close vicinity to these growing epithelial buds condense into what becomes the primary mesenchyme [102], serving as the only ERα-expressing compartment of this developing embryonic gland

[86, 103] (also, see Chapter 3, Figure 3.4.1). Importantly, in mice, ERα expression has been reported in the epithelium only in post-natal development [104]. This dense mesenchymal sheath consists of fibroblast cells. The bud remains relatively unchanged and completely surrounded by ERα-expressing mammary mesenchyme between E13-15.

Finally, at approximately E16, the epithelial bud experiences a burst of proliferation. As a result, the epithelial sprout elongates beyond the primary mammary mesenchyme and towards the fat pad, consisting of adipocytes [101]. Rudimentary epithelial branching is observed beginning at approximately E16 through birth, resulting in a small ductal tree, emanating from the nipple.

For the purposes of work presented herein, in utero development and exposures will be emphasized. However, briefly, post-natal development involves the filling-out of the mammary fat pad, first beginning with the invasion of nascent and highly

proliferative terminal end buds (TEBs) as in mice or terminal duct lobular units (TDLUs) in humans. This is mediated by estrogen secreted by the ovaries. These structures eventually further differentiate and give rise to increased arborization of the mammary epithelium to fill the entire mammary fat pad (as reviewed in [105]. This branching process is regulated by a number of signaling factors, including transforming growth factor beta (TGFβ) and local cues as orchestrated by gland tissue architecture [106]. The branching process coincides with the onset of the estrous cycling. The gland remains relatively static until pregnancy when rapid proliferation and remodeling occur, followed by hormonal regulation of lactation and gland remodeling, known as involution.

Importantly, the additional complexities of the mammary gland are present within the diverse populations of component cell types. The epithelial ducts have an interior lumen lined with luminal epithelial cells, 30% of which typically express ERα [107]. This luminal layer is then surrounded by ER-negative basal epithelial cells, approximately

20% of which are comprised of mammary stem cells (MaSCs) [108]. This layer is in contact with the basement membrane of the epithelial duct. The mammary fat pad is dominated by fat-filled adipocytes and this sea of fat is bestrewn with stromal fibroblasts.

However, fibroblasts are more often closely associated with the ductal basement membrane and are intricately involved with the extracellular matrix (ECM) mediation of epithelial development and architecture [109, 110]. Indeed, while the epithelium is typically the site of tumorigenesis and disease, the density of the stroma/ fibroblasts has been identified as a major risk factor for the development of breast cancer as well [111,

112]. Additionally component cell types of the mammary gland include immune, lymphatic and vascular cells.

In consideration of the BPA-mediated mechanism of incidence and disease and understanding the propagation of effects incurred from in utero exposures to adult-life, it is clear that discrete alterations to these component cell types may play a role.

1.4.2 The susceptibility of key mammary gland developmental stages to endocrine disruption.

The intricacies of the developing mammary gland present three distinct stages that are highly susceptible to EDC exposures, including: (1) embryonic/ fetal, (2) pubertal and

(3) pregnancy [113, 114].The mammary gland is unique in that much of its dynamic and functional development occurs post-natally as orchestrated by hormonal signaling. In contrast, embryonic breast development relies heavily on signaling intricacies and cross- talk between tissue compartments: the epithelium and the mesenchyme [101, 115].

The susceptibility of the embryonic window of exposure was well-demonstrated by the administration of DES to pregnant women and their “DES Daughters” exhibiting increased later-life incidence of breast cancer. Interference of EDCs during this time have the potential to disrupt or reprogram critical tissue compartment interactions affecting later-life development and signaling necessary for mammary gland functionality. The pubertal window of exposure presents many factors of susceptibility, both in the onset of critical hormone signaling, namely through estrogen and progesterone, and in the highly proliferative and dynamic structures present during this time. Specifically, in mice these are the terminal end buds (TEBs) and in humans the terminal duct lobular units (TDLUs)

[77]. These structures have a higher propensity to be vulnerable to carcinogens in that they are undifferentiated and highly proliferative. The final stage of gland maturation occurs in pregnancy and is further susceptible through its reliance on hormonal signaling and tissue microenvironment to remodel the gland for lactation, through prolactin and growth hormones. Pregnancy also presents another time in mammary development where proliferation and differentiation of the epithelium is high and thus vulnerable to endocrine disruption [113].

Taking into account the developmental timing of EDC exposures can provide insights into later-life susceptibilities observed and ultimately ways to address and prioritize prevention.

1.4.3 The importance of epithelial-stromal interactions in mammary gland development.

Tissue architecture and microenvironment are critical for organ function and morphogenesis. Eloquent in vitro explant studies within the mammary gland have demonstrated that tissue compartment interactions guide early-life gland development and are not orchestrated by systemic hormonal signaling [115]. Interestingly, in the embryonic mammary gland, there are subsets of the mammary mesenchyme which have different potentials to influence epithelial development. Recalling that there is a dense

ERα-expressing mammary mesenchyme which completely surrounds the developing mammary epithelial bud through E15, versus a more distal ‘secondary’ mesenchyme making up the precursor fat pad. The former was shown to induce hyperplastic growth in the epithelium of adult mice when grafted into their fat pads [116]. Such experiments further demonstrated that mesenchyme dictates developmental specificity. Principally,

when mammary epithelium was recombined with mammary mesenchyme, a normal epithelial tree resulted. Conversely, when mammary epithelium was recombined within a salivary mesenchyme, a salivary-like epithelial tree was generated (as reviewed in [110,

115, 117]. It follows then, that the dynamics of the adult epithelium (regardless of glandular origin) are retained as well, when recombined and under the influence of embryonic mammary mesenchyme [118]. Finally, it has been further shown that disruption of the remodeling and structural activities of the surrounding mesenchyme, mainly through the ECM, correlates to epithelial breakdown, an epithelial-mesenchymal transition (EMT) and tumorigenesis [119, 120].

The potential of BPA to modify tissue compartments has been shown at perinatal time points in mammary development through decreased collagen deposition (Trichrome staining) [85] and the transcriptional up-regulation of a gene (Tgm2) that encodes a protein involved in collagen cross-linking [86]. Both of these phenotypes are associated with ECM and breast density. However, little has been shown to demonstrate these mesenchymal/ ECM alterations by BPA in adult life, following early-life exposures, highlighting a gap in the field.

1.5 Gaps in the field of endocrine disruption.

1.5.1 Chronic EDC exposures to recapitulate human exposures.

The phenotypes and breast cancer incidence are well-characterized in the rodent mammary gland; however the mechanisms by which BPA elicits these programs are not.

Tissue bio-monitoring of BPA indicates that this EDC is pervasive in our environment. A

shortcoming of the field our work has aimed to address was that many analyses, in an effort to evaluate human risk, focus on short-term administered exposures resulting in the immediate activation of the ER and acutely-driven transcriptional programs. This unfortunately does not recapitulate the demonstrated chronic exposures of the human population. We have used an in vtiro chronic treatment model to address this gap and characterize sustained molecular effects as a result of chronic EDC exposures.

1.5.2 Elucidation of a poorly understood mechanism through examination of key stages of gland development.

Further, in utero BPA exposures are a concern and come to the forefront, firstly due to the demonstrated levels of this EDC found in pregnancy-associated fluids. The similar phenotypes and propensity for cancer susceptibility and progression between the known carcinogen DES (only administered during in utero time points) and EDC, BPA exacerbates this concern. The distinct tissue compartments (epithelium and stroma) of the developing in utero mammary gland present discrete intrinsic mechanisms which may be alternatively mediated by BPA to confer later-life changes affecting the interactions of these tissues. It is further unclear if the mammary gland has different susceptibilities to

EDC exposure across these in utero developmental stages. Some investigation into the in utero developmental windows linked to DES-induced increased risk found that the reproductive tract tumors found in ‘DES daughters’ were specifically associated with first trimester exposures [54-56]. Though, more in-depth analysis of DES exposure windows and breast cancer has not been performed, our work sought to narrow the mechanistic possibilities of in utero BPA by examining discrete windows of exposure

modeled after the key stages of development at time of exposure. Taken together with clues from the elucidated modulators of these discrete tissue compartment interactions in normal mammary development, it becomes possible to attribute more specific signaling changes as influenced by in utero exposure to BPA.

1.5.3 The potential of component cell types of the adult mammary gland to reveal the mechanism of in utero BPA missed by whole tissue analyses.

The consequences of BPA exposure are not known in the human population, but it was clear from DES exposures that there was latency, a matter of several decades, between exposure and increased incidences of cancers. We have not reached a point, since the onset of prolific production and use of BPA, which would manifest the anticipated consequences of breast cancer risk. Given this latency, and the focus of the field on embryonic developmental alterations, our work has sought to elucidate a mechanism by which in utero deregulation is propagated to later-life. Thus, in this body of work we administer in utero BPA exposures and our in vivo analyses have focused on alterations at the adult time point. The complexity of the mammary gland through changes in the adult mammary component cell types has been prioritized in our analyses to this end. To date, this complexity has been largely overlooked or more vaguely assigned to tissue compartments which can fail to tease out the nuances of changes achievable by in utero exposures in each distinct cell type.

1.5.4 Are the observed latent consequences of in utero exposure propagated by epigenetic mechanisms?

There has been evidence implicating epigenetic alterations and reprogramming in the mechanism of in utero BPA. This has included the utilization of the Agouti gene in the viable yellow agouti (Avy) mouse [121] and maternal administration of BPA that was shown to induce DNA hypomethylation. This hypomethylation was manifested through a change in coat color of offspring, normally brown agouti mice to yellow [122]. This unmethylated state is associated with increased risk of disease and demonstrated the epigenetic modifying potential of in utero BPA. More recently, in utero exposure to both

BPA and DES were shown to increase both histone 3 lysine 27 trimethylation

(H3K27me3) and its enzyme, enhancer of zeste 2 (EZH2), indicative of gene repression

[87]. In our examination of adult alterations to the mammary gland, we aim to attribute discrete epigenetic changes and reprogramming to the component cell types that may account for the propagation of later-life disease susceptibility observed.

1.6 Research Purpose

To date, many epidemiological and tissue biomonitoring studies of the human population have revealed widespread and sustained exposures to EDC compounds in our environment. The phenotype and later-life consequence of in utero EDC exposures, including that of BPA, have been well-described. Further, the onset of incidence to later- life disease is delayed from a vulnerable in utero time of exposure until an adult time point in development, as demonstrated by DES exposures. We postulate this latency suggests propagation of effects and a mechanism either through discrete changes to signaling between the two tissue compartments at time of exposure, or within component

cell types, as mediated by BPA. The exact mechanism of EDC-mediated development directing later-life disease is not known. Herein and presented in Chapter 2, we first sought to mimic life-time exposure to EDCs in our environment through an in vitro chronic treatment model to explore global transcriptional effects resulting from long-term exposures and determine sustained changes in signaling and recruitment of ER. Given the dependency on epithelial-stromal tissue interactions throughout normal embryonic mammary gland development, in Chapter 3 we proposed that discrete windows of in utero BPA exposure at key stages of tissue interactions at time of exposure may narrow the developmental mechanism of BPA at this time point. In acknowledgment of the plasticity and heterogeneity of the developing mammary gland through its component cell types, we investigate cell type-specific changes in transcription to reveal a mechanism that relates BPA in utero exposures to known risk factors for breast cancer development in Chapter 4. Finally, this body of work concludes by further utilizing component mammary gland cell types to determine a mode of propagation of effects, through epigenetic mechanisms as part of future directions. These briefly include examination of DNA methylation and enrichment of a known repressive mark of transcription, histone 3 lysine 27 trimethylation (H3K27me3).

Chapter 2: Sustained Reprogramming of the Estrogen Response Following Chronic

Exposure to Endocrine Disruptors2

2.1 Abstract

The pervasive nature of estrogenic industrial and dietary compounds is a growing health concern linked to cancer, obesity, and neurological disorders. Prior analyses of endocrine disruptor action have focused primarily on the short-term consequences of exposure. However, these studies are unlikely to reflect the consequences of constant exposures common to industrialized countries. Here, we examined the global effects of long-term endocrine disruption on gene transcription and estrogen signaling. Estrogen- dependent breast cancer cell lines were chronically treated with physiologically relevant levels of bisphenol A or genistein for more than 70 passages. Microarray analysis demonstrated global reprogramming of the transcriptome when compared to a similarly cultured control cell line. Estrogen responsive targets showed diminished expression in both the presence and absence of estrogen. Estrogen receptor recruitment, H3K4 monomethylation, and DNase accessibility were reduced at nearby response elements.

______

2Andrea R. Patterson, Xiaokui Molly Mo, Ali Shapiro, Karen E. Wernke, Trevor K. Archer, and Craig. J. Burd. Sustained reprogramming of the estrogen response following chronic exposure to endocrine disruptors. Mol Endocrinol. 2015 Jan.

Based upon these observations, we investigated the potential of long-term EDC exposure to initiate persistent transcriptional reprogramming. Culture of chronically exposed cell lines in the absence of the endocrine disruptors did not reverse many of the signaling defects that accumulated during treatment. Taken together, these data demonstrate that chronic exposure to endocrine disrupting compounds can permanently alter physiological hormone signaling.

2.2 Introduction

Estrogen signaling is essential for normal development and physiology, but when deregulated, promotes pathological conditions such as cancer. Canonically, estrogens bind the estrogen receptor (ERα or ERβ), inducing dimerization, DNA binding, and the subsequent recruitment of co-regulatory molecules responsible for target gene expression.

Interaction of the ER with DNA can be either direct (through estrogen responsive elements/EREs) or indirect (via interaction with other transcription factors such as AP-

1[24]), serving to regulate defined gene ontologies.

Endocrine disruptors (EDCs) are industrial and dietary compounds that alter normal estrogen signaling and have been implicated in cancer, infertility, obesity, diabetes and neurological disorders [123, 124]. EDCs, like bisphenol A (BPA) and genistein (GEN), bind the ER directly and stimulate compound-specific transcriptional programs that are disparate from that of the estrogen-bound receptor [53, 125]. The ability of EDCs to stimulate aberrant ER signaling is a growing health concern. BPA is pervasive in industrialized countries where it is used in the manufacturing of plastics,

dental sealants and food packaging. As a result, BPA is detectable in the serum of >95% of Americans [126]. Serum BPA levels as high as 80 nM have been reported, but typical levels remain in the 1-10 nM range [48, 65, 127]. Other estrogenic compounds, like

GEN, are abundant in our diet. While serum GEN levels typically range from 10-60 nM

[128], individuals on a soy-rich diet can exhibit levels up to 140 nM [128, 129]. The ER- binding affinities of GEN and BPA are ~10-10000-fold less than 17-beta estradiol (E2)

[130, 131], but EDC levels within the range of human exposure are nonetheless capable of altering ER action [132, 133].

A specific fear surrounding EDCs has grown from the well-established link between life-long estrogen exposure and breast cancer [14, 134]. As EDCs are ubiquitous in industrialized countries, prolonged exposure to these compounds may increase breast cancer risk. To address this concern, prior studies have focused on defining the immediate effects of EDCs on ER signaling. However, these studies fail to address the persistent nature of real-life EDC exposures that may initiate distinct phenotypes compared to acute exposures [135-137]. For example, while GEN acutely promotes estrogen-dependent growth, chronic exposure of human breast cancer cells (MCF7) inhibited estrogen-dependent cellular proliferation [138]. In the setting of acute exposures, each EDC distinctly reprograms ER signaling. Therefore, the long-term consequences of EDC exposure are also likely to be compound-specific.

Here, we investigated the impact of chronic BPA and GEN exposure on estrogen- dependent breast cancer models. To faithfully mimic human exposure, these EDCs were administered for over 1 year in the presence of physiological hormones. Our work reveals

that chronic EDC exposure initiates compound-specific transcriptional reprogramming.

Specifically, ER target genes showed decreased expression and reduced receptor recruitment to nearby response elements. We show that decreases in ER binding are attributed to epigenetic changes (i.e. the loss of H3K4me1 and DNA accessibility at enhancer elements). These alterations in E2-induced ER activity are persistent, remaining long after the removal of EDCs from the culture. Together, these data demonstrate the potential of chronic EDC exposure to permanently modify the normal estrogen signaling axis, suggesting that prolonged exposures result in lifetime health consequences.

2.3 Materials and Methods

2.3.1 Cell culture.

MCF7 and T47D cells lines were acquired from ATCC and maintained in

Dulbecco’s Modified Eagle Medium (DMEM) supplemented with phenol red (Gibco® by Life Technologies) 10% (v/v) fetal bovine serum (FBS), 1% (v/v) penicillin- streptomycin solution and 2mM ʟ-glutamine. Chronic EDC-treated cells were cultured in the presence of either no EDC (MCF7-F, T47D-F), 50 nM bisphenol A (BPA; Sigma

MKBF3852V; MCF7-B, T47D-B) or 50 nM genistein (GEN; Sigma G6649-5MG;

MCF7-G, T47D-G) for > 70 passages. Fresh medium was added to the cultures three times per week.

2.3.2 Western immunoblotting.

Chronically treated MCF7 cells and low passage MCF7 cells were plated at equal density and allowed to adhere overnight. The following day cells were harvested and

lysed in RIPA buffer (150 mM NaCL, 1% IGEPAL, 0.5% deoxycholate, 0.1% SDS in 50 mM Tris pH 8.0) with added protease inhibitors. Total protein was quantified (Bio-Rad

Bradford protein assay) and subjected to SDS-PAGE. Gels were transferred to PVDF and probed with antibodies specific for ERα (GeneTex, #EPR4097) and β-actin (Cell

Signaling, #3700). Immunoblots were imaged on a LI-COR Odyssey.

2.3.3 RNA isolation, RT-PCR and real-time PCR.

Cells were initially seeded and then cultured for 72 hours in phenol red free

DMEM supplemented with 5% charcoal dextran treated FBS (CDT). Cells were then stimulated with ethanol (EtOH) vehicle, 10 nM E2, 100 nM BPA or 100 nM GEN for 3 hours and RNA was harvested using RiboZolTM (Amresco). Higher doses of BPA and

GEN were used in the RT-PCR experiments as compared to the long term exposure to maximize the ER response for these short time point experiments. RNA was reverse transcribed using the Improm II RT System (Promega). The resulting cDNA was analyzed using (real-time) primers specific for the indicated genes (Table 2.1).

Table 2.1 cDNA primers utilized for transcriptional genes of interest.

A minimum of 3 biological replicates, measured in triplicate, were performed for each target. Data was normalized to the vehicle treated MCF7-F cell line and is depicted as the mean and standard error. Statistical significance was determined using a two-tailed student’s t-test.

2.3.4 cDNA microarray and real-time PCR validation.

Cells were cultured and treated as described for RT-PCR analysis. RNA was isolated using the Total RNA Isolation Nucleospin® RNA II kit (Macherey-Nagel).

Microarray analysis was performed by the Microarray and Next Generation Sequencing

Shared Resource at The Ohio State University Comprehensive Cancer Center using

GeneChip® Human Transcriptome Arrays (vs. 2.0, Affymetrix). For gene profiling microarray, we normalized the probe cell intensity (CEL) with the RMA method using

Expression Console Software (Affymetrix). Analysis of variance was performed in order to identify differentially expressed genes. Analysis was done using SAS 9.3 software

(SAS, Inc, Cary, NC). Microarray intensity files can be accessed through GEO

(http://www.ncbi.nlm.nih.gov/geo/) with the accession number GSE59345.

2.3.5 Chromatin immunoprecipitation (ChIP).

ChIP was performed as previously described [139]. Briefly, cells were cultured in

DMEM with 5% CDT for 72 hours followed by stimulation with vehicle control or 10 nM E2 for 45 minutes. Cells were cross-linked, sonicated and immunoprecipitated with antibodies specific for ERα (GeneTex-EPR4097), IgG control (Cell Signaling) or

H3K4me1 (Abcam, ab8895). Enriched DNA was quantified using the Qubit® dsDNA

HS assay kit (Life Technologies). Real-time PCR was performed using 400pg DNA/well with primers amplifying estrogen response elements for the target genes of interest (Table

2.2). Data shown is representative of at least 3 independent biological replicates.

Table 2.2 ERE primers used for ER recruitment and DNA accessibility.

2.3.6 DNase sensitivity assays.

Chronically treated MCF7 cells grown in DMEM/CDT for 72 hours were harvested and DNase assays were performed as previously described [140]. Data shown is representative of at least 3 independent biological replicates.

2.4 Results

2.4.1 Chronic EDC treatment reprograms global gene expression.

Exposure to environmental estrogens is both constant and ubiquitous within industrialized countries. To generate an in vitro model of chronic EDC exposure, we supplemented the media of MCF7 cells with 50 nM BPA or GEN for greater than 70

passages (> 1 yr; Figure 2.1A). Two model cell lines, termed MCF7-B (BPA) and

MCF7-G (GEN), were generated in addition to a control cell line, MCF7-F (FBS), grown in parallel without EDC addition. The concentrations of BPA and GEN chosen for these studies were within the documented ranges detected in human serum [65, 128, 129].

Using microarray analysis, we examined whether chronic EDC exposure could induce transcriptional reprogramming. Cells were cultured in the absence of hormone for

72 hours and then stimulated for 3 hours with 10 nM E2 or ethanol vehicle (EtOH; Figure

2.1A). Microarray-based transcriptome analysis revealed genes that were statistically different in the chronically treated cell lines compared to the MCF7-F cells (Figure 2.1B).

The heatmap depicts the top 100 genes that demonstrated the highest fold changes

(minimum > 2-fold, p<0.05) between MFC7-F (EtOH) vs. –B (EtOH) and MCF7-F vs. –

G (EtOH) (200 total genes). Although many deregulated targets (up- or down-regulated ≥

2-fold) in the MCF7-B and -G lines were distinct (Figure 2.1B), chronic EDC treatment was generally associated with increases in transcription (Figure 2.1C). There were 115 genes showing higher expression in the MCF7-G cell line compared to only 21 genes showing lower expression in the MCF7-G cell line when compared to the MCF7-F control (p < 0.05, > 2-fold). The MCF7-B cell line showed a similar ratio (110/25) of genes showing higher expression when compared to the control.

2.4.2 Chronic EDC treatment reprograms E2-regulated gene expression.

We observed many established ER targets deregulated within the microarray dataset. Therefore, we reanalyzed our microarray data, focusing on the top 150 ER target genes in the MCF7-F cell line that changed ≥ 1.5-fold in response to E2. Hierarchical

clustering revealed that while a large portion of E2-responsive genes remained similarly regulated, a unique signature representing specific deregulated genes was observed in the treated cell lines (MCF7-B and -G) (Figure 2.1D). Moreover, many E2-responsive genes showed altered expression in both the estrogen-depleted (CDT) and -stimulated state.

However, EDC-dependent transcriptional reprogramming was not universal.

Approximately 25% of E2-responsive genes were similarly regulated in all three cell lines plus or minus E2 (Figure 2.1E). In contrast to the unbiased clustering of genes performed in figure 1B and C, we found that most E2-responsive genes showed decreased expression in cell lines exposed to chronic BPA or GEN treatment (Figure

2.1E). Specifically, there were 127 genes induced greater than 1.5-fold by E2 in the

MCF7-F cell line that was not induced to that extent in the MCF7-B and -G cell lines. In contrast, only 8 and 24 genes were more highly induced by E2 in the MCF7-B or -G cell lines, respectively. In relation to E2-mediated gene repression, the chronically treated cell lines had a larger number of repressed genes than the MCF7-F control (Figure 2.1E, right diagram). We only observed 21 genes repressed by E2 treatment in the MCF7-F cell line, while 36 and 68 genes were repressed (p<0.05, > 1.5-fold) in the MCF7-B and –G lines, respectively. Taken together, there was a much larger percent of E2-regulated genes repressed in the chronically treated cell lines (MCF7-B/G) than the MCF-7 control (p <

0.0001, chi-square test). These data show that chronic EDC treatment causes the deregulation of E2 responses, frequently resulting in reduced target gene expression.

To further characterize the estrogen response in cell lines chronically treated with

EDCs, we performed a more detailed analysis of known E2-responsive genes (Figure

2.2A-C). These genes were selected based on a combination of either strong E2-mediated induction at 3 hours or the presence of a strong ERα ChIP peaks proximal to the transcriptional start sites. Each cell line was cultured for 72 hours in the absence of steroid hormones (CDT) and then stimulated with 10 nM E2 or vehicle for 3 hours

(Figure 2.1A). As seen in our microarray analyses, a subset of estrogen-responsive genes responded similarly in all cell lines (e.g. PS2, PGR; Figure 2.2A). Additionally, EGR3 was significantly (p < 0.05) repressed in both the E2-stimulated MCF7-B and -G cell lines compared to control MCF7-F (Figure 2.2B). Finally, a third group of gene targets demonstrated an EDC-specific E2-response (e.g. MSMB, MYC, GREB, HDAC11; Figure

2.2C). For example, MSMB expression increased in the MCF7-B cell line yet had normal expression in MCF7-G cells compared to the control (Figure 2.2C). Interestingly, many targets with altered expression prior to E2-stimulation (EtOH) also had expression that differed from the MCF7-F control in the E2 treated samples (Figure 2.2B, C). For example, the EGR3 gene in the MCF7-G cell line had much lower basal activity and E2- induced gene expression, but the fold enrichment after E2 treatment was similar to the

MCF7-F cell line. These data suggest that many of these targets can still respond to ER activation, but the overall transcriptional program is abnormal. Changes in gene expression in the absence of ER activation suggest that the mechanism driving deregulation is associated with alterations to the genomic loci.

In order to verify that decreased estrogen-targeted gene expression was not due to

EDC-induced changes in ERα expression, protein was harvested from chronically treated cells and subjected to SDS-PAGE followed by immunoblotting. ERα was highly

expressed across all three cell lines comparable to low passage MCF7 cells (Figure

2.2D). Changes in estrogen-dependent signaling could also be influenced by stochastic events occurring during the long-term culture of each MCF7 derivative. Therefore, we compared the transcriptional response of early passage MCF7 (LP) cells to the MCF7-F line. We looked at both E2-dependent and –independent targets that were differentially expressed in the chronically treated cell lines. At these targets, MCF7-F cells behave similarly to low passage MCF7 cells (Figure 2.7).

To corroborate these findings, we generated chronically treated cultures of another estrogen-dependent breast cancer cell line, T47D. Similar to the MCF7-B and –G lines, chronic EDC-treated T47D cultures showed diminished activation of EGR3 and

HDAC11 (Figure 2.2E). We also observed a significant (p < 0.05) decrease in PS2 (also known as TFF1) expression in the T47D-B cell line. At these targets, transcriptional reprogramming in response to chronic EDC administration is conserved across the MCF7 and T47D cell lines.

2.4.3 Deregulated ER response is independent of the activating ligand.

The ability of chronic EDC exposure to alter canonical estrogen-dependent transcriptional programs suggested that acute responses to non-canonical ligands might also be affected in these lines. As these cells may have adapted to exposure of EDCs, a ligand-specific activation of the ER may be observed. To test this possibility, each MCF7 cell line was cultured in CDT for 72 hours and then stimulated with either 100 nM BPA or 100 nM GEN (Figure 2.3). Examination of classic estrogen-responsive genes revealed a statistically significant (p < 0.005) decrease in PS2 expression specific to the MCF7-B

line when treated with GEN (Figure 2.3A). This difference was particularly interesting as

E2-stimulated PS2 expression was nearly equivalent in the MCF7-F, -B, and –G cultures

(Figure 2.2A). While this difference may be a ligand-specific defect in ER signaling, it also is more likely indicative of a subtle deregulation only observed under conditions of minimal ER activation as is the case with low affinity EDCs. This defect in signaling may have been masked in figure 2 by the 10 nM E2 dose designed to maximize ER activation.

Along those lines, the MCF7-B cell line showed reduced transcription of

HDAC11 (Figure 2.3B) similar to MCF7-G under conditions of EDC activation. In this model, HDAC11 may be better grouped with EGR3 in that it has reduced sensitivity to estrogenic stimulation in both EDC treated lines. When we looked at EGR3, the target showing the greatest reduction in gene expression (Figure 2.2B), it had a similar inability to respond to the non-canonical signaling of BPA or GEN (Figure 2.3 C). Finally, much of the altered gene expression in the chronically treated cells at ER-regulated genes

(Figure 2.2C) showed identical deregulation in the presence of EDC ligands (Figure

2.3D, and data not shown). These data demonstrate that defects in ER-target gene activation are ligand-independent and that the chronically treated cells did not adapt to the EDC exposure by generation mechanisms to respond to EDCs instead of canonical ligands.

2.4.4 Chronic EDC treatment causes differential ER recruitment.

To identify the molecular mechanisms driving deregulation of ER activity following chronic EDC exposure, we performed ChIP assays in the established MCF7 cell lines. Each culture was stimulated with 10 nM E2 for 45 minutes and then harvested

for analysis. Real-time PCR primers were designed to assess EREs proximal to a set of well-characterized E2-responsive genes. In all three MCF7 lines (-F, -B and -G), the ER was strongly recruited to an ERE near the PS2 gene (Figure 2.4A). This finding was not surprising given that PS2 was equally expressed in the E2-stimulated control and EDC- exposed cell lines (Figure 2.2A). In general, alterations in gene expression observed in the MCF7-B and –G lines were paralleled by changes in ER recruitment (Figure 2.4B,

C). For example, reduced expression of MYC, EGR3 and HDAC11 in the MCF7-G line was accompanied by decreases in gene-specific ERE recruitment (Compare Figure 2.2B,

C, E to Figure 2.4B, C). Although this correlation between ER binding activity and gene expression was consistent across most targets, it was not universal. The MCF7-B line demonstrated loss of ER recruitment at the MYC locus, but no observable decrease in transcription. However, the MCF7-B cell line generally showed reduced ER recruitment at all analyzed sites, but that affect was most pronounced at the repressed targets (Figure

2.4, compare C to A).

2.4.5 The EGR3 locus is epigenetically reprogrammed in EDC-treated cell lines.

Many of the EREs that showed reduced ER recruitment were located in gene enhancers. To examine whether enhancer activity was altered in response to chronic EDC exposure, we assessed the histone 3 lysine 4 monomethylation (H3K4me1) status of relevant EREs. H3K4me1 is a histone mark enriched at active enhancer elements and has been implicated in cell-type specific transcriptional responses [141]. We postulated that a decrease in H3K4me1 might be responsible for the reduced ability of ER to locate and bind specific EREs in cell lines chronically treated with BPA or GEN. Therefore, we

examined the status of H3K4me1 prior to estrogen stimulation. In the absence of hormone, MCF7-B and -G cells showed equal or greater levels of H3K4me1 at the PS2

ERE than those observed in the control line (MCF7-F). Specifically, H3K4me1 at this site was higher in the MCF7-G line than any other culture (Figure 2.5A, left), paralleling the enhanced recruitment of ER to the PS2 promoter observed in these cells (Figure

2.4A). The MSMB gene which showed elevated expression in the MCF7-B line, demonstrated no increase in ER binding (Figure 2.4B) and no significant change in

H3K4me1 (Figure 2.4.5A, right). MSMB did demonstrate a very high basal expression in the absence of hormone (see Figure 2.2C), suggesting that deregulation of this locus may not be ER dependent. However, at the two EGR3 response elements we observed a decrease in H3K4me1 that was restricted to MCF7 cells chronically treated with EDCs

(Figure 2.5A, see bottom panels). Again, these results paralleled the change in ER recruitment observed at the EGR promoter of the MCF7-B and -G cell lines (Figure 2.4C, left). We further investigated the repressive histone marks H3K9me3 and H3K27me3 at these genomic loci, but did not see change in these modifications that would suggest active repression of the target genes (data not shown). These data show that chronic EDC exposure alters the chromatin landscape leading to defects in the recruitment of activated

ER to specific target gene promoters. We thus probed the microarray data for transcriptional changes in histone modifying enzymes and discovered only one transcript

(MLL1/KMT2A) that was significantly altered > 1.5-fold in the chronically treated cell lines. MLL1 was downregulated (1.53-fold MCF7-B, 1.35-fold MCF7-G, p < 0.05) and is a methyltransferase associated with the H3K4me2 and H3K4Me3 marks. In order to

ensure that global changes in H3K4 methylation were not influencing the ChIP signatures, we performed western blots demonstrating no changes in total histone H3,

H3K4me1, and H3K4me3 for each cell line (Figure 2.8).

DNA accessibility is known to control nuclear receptor recruitment to target genes [142, 143]. Since EREs in the chronically treated cell lines showed a loss of

H3K4me1 (Figure 2.5A, see EGR3 EREs), we investigated if the chromatin architecture at these sites was concomitantly altered using DNase hypersensitivity assays. Here, DNA was subjected to a limited DNase I digestion and then analyzed by real-time PCR to determine the relative resistance of key EREs to enzymatic cleavage. The PS2 and MSMB genes, both of which showed efficient binding by ER, showed either no change or a slight increase in accessibility (Figure 2.5B, upper panels). In the case of PS2, the MCF7-G cell line demonstrated a statistically significant decrease that corresponds to increased ER recruitment (Figure 2.3A) and H3K4me1 (Figure 2.4A), but not to transcriptional output

(Figure 2.1). In contrast, EGR3 EREs were more resistant to DNase digestion in the

GEN- and BPA-treated cell lines (Figure 2.5B, bottom panels). Therefore, chronic exposure to EDCs results in reduced accessibility of nearby enhancer elements and decreased recruitment of receptor to these EREs.

2.4.6 Estrogen signaling defects persist after EDC removal.

Our data show that chronic EDC treatment causes gene-specific reprogramming of the estrogen response. Since chromatin alterations play a role in this process, the mechanism by which long-term EDC treatment alters hormone signaling may be

epigenetic. By definition, epigenetic events are heritable changes to the chromatin structure that are sustained through multiple cell divisions.

To test whether alterations in estrogen signaling caused by chronic EDC exposure were epigenetic, we removed GEN and BPA from our MCF7-B and –G lines for several passages (Figure 2.6A). First, we confirmed that EDC removal did not alter ERα expression levels in the MCF7 cell lines. Immunoblot demonstrated that ERα protein levels were unchanged after removal of the EDCs (compare Figures 2.2D and 2.6B).

Next, we assessed E2-mediated gene expression in these same cell lines. As expected, genes that were initially unaffected by chronic BPA or GEN treatment (Figure 2.2A), remained unaffected in MCF7-B and -G lines cultured in the absence of EDC for 2 or 9 passages (Figure 2.6C). However, the deregulation of EGR3 (Figure 2.2B and 2.3C) was sustained in the MCF7-B and -G cell lines 2-9 passages after EDC removal (Figure

2.6D). Interestingly, two genes, MSMB and HDAC11, that previously showed EDC- specific effects (Figures 2.2C), demonstrated enhanced deregulation in both EDC-treated cell lines (Figure 2.6E) after several passages without treatment. Such sustained alterations at the estrogen responsive genes were typical, but not universal. For example,

MYC, a target deregulated solely in the MCF7-G cell line (Figure 2.2C), showed similar induction across all cell lines after removal of the EDCs (data not shown). Thus, while most of the genes deregulated by the estrogen response persisted, some had the ability to return to normal expression states after continued culture in the absence of EDCs. The observed changes in ER recruitment and DNA accessibility at nearby enhancer elements

were also maintained in the chronically treated cell lines following removal of the EDCs

(Figure 2.9).

Together, these data demonstrate that chronic EDC treatment can permanently alter the ability of estrogen to activate ER-target genes and further implicates the potential of these compounds to epigenetically reprogram hormone-responsive breast cancer cells.

2.5 Discussion

2.5.1 A model of chronic exposure.

Detectable levels of EDCs in human sera and urine indicate that people in industrialized countries are constantly exposed to environmental hormones. However, prior work in the field has focused on acute exposures in an effort to demonstrate the ability of these compounds to activate ER signaling. Herein, we describe sustained changes to the estrogen signaling program following chronic EDC exposure that cannot be represented in acute models. Utilizing a cell culture system to mimic chronic BPA and

GEN exposure, we show reduced expression of many E2-responsive genes under both basal and E2-stimulated conditions (Figure 2.1D, 2.2B, C). While the acute EDC activation of ER stimulates expression (Figure 2.3, MCF7-F), chronic exposure led to reduced expression of many ER-responsive genes (Figure 2.1E, 2, 3, MCF7-B, -G).

2.5.2 Chronic EDC exposure deregulates ER signaling in a compound- and gene-specific manner.

It is well established that EDCs can bind the ER, regulate gene expression, and drive estrogen-dependent proliferation [53, 125, 132, 133, 144-148], presumably dependent upon the conformation of the ligand-bound ER and its ability to recruit distinct subsets of transcriptional co-regulators [149, 150]. Following chronic EDC exposure, we found that canonical ER signaling is also deregulated in a compound- and gene-specific manner. For instance, MSMB expression was induced dramatically in MCF7-B cells, but did not increase in the similarly treated MCF7-G line (Figure 2.2C). Likewise, MYC expression was decreased in MCF7-G cells alone (Figure 2.2C). Other genes were similarly deregulated in both the BPA and GEN treated cells (EGR3, HDAC11). Gene target specificity may be dependent upon the receptor confirmation induced by each ligand, thus modulating co-regulator recruitment. Also, BPA and GEN are known to induce different estrogenic activities. For example, BPA has been shown to have a better ability to induce the ER tethering activity at AP-1 response elements when compared to

GEN [125]. GEN has been shown to bind ERβ 20-30 fold stronger than ERα, however in vitro assays suggest that GEN only has weak preference for ERβ in gene transactivation

[151]. Finally, BPA and GEN are known to act through alternate pathways other than ER that may contribute to unique gene signatures [152, 153].

Interestingly, while many ER targets were deregulated, the receptor was still able to at least partially modulate gene expression. In fact, in some cases the alterations in gene expression are mostly attributed to basal activity and the relative fold induction after

ER activation is roughly equivalent. Nonetheless, the transcriptional output at these genes in the chronically treated cells is substantially altered even in the presence of E2. A

potential caveat of these experiments is that CDT media retains a small amount of estrogenic compounds and basal activity is not truly estrogen-free. Thus, gene expression reductions in the EtOH treated samples could be indicative of a loss in the residual estrogenic activity of CDT media. Moreover, ER could retain the ability to reorganize the gene to induce full activity, but at some targets is not being effectively recruited. There is also the possibility that other factors synergize with ER at these loci. Transcription factors tend to bind in hotspots where an enhancer may recruit multiple transcription factors [139]. The changes we see in ER recruitment and enhancer chromatin architecture may also be affecting the recruitment of other transcription factors.

2.5.3 Chronic EDC deregulation persists in a compound- and gene-specific manner.

Many changes in gene expression and chromatin architecture persisted even after removal of the EDCs (Figure 2.6, Supplemental figure 2.8). These data show that chronic

EDC exposure may promote persistent, compound- and gene-specific defects. However, while some changes in transcription appeared permanent, MYC did not display evidence of epigenetic reprogramming and normal expression patterns were restored after removal of the EDC. Clearly, the complexity of ER signaling extends to chronic EDC action and the transcriptional changes observed may be dependent upon the diverse set of molecular and estrogenic targets of these compounds. In particular, ligand specific recruitment of co-regulators may play a critical role in the gene specificity of epigenetic modulation.

Co-regulators often possess chromatin modifying enzymatic activities [154]. Thus, the proposed role of co-regulators in ligand-specific ER signaling is particularly relevant to the epigenetic modifications in cells chronically exposed to EDCs. The recruitment of

chromatin modifying complexes to specific ER targets may drive the gene-specific epigenetic reprogramming observed in our cell lines as a result of chronic treatment.

Alternatively, EDCs are known to selectively modulate the expression of chromatin readers, writers, and other modifying enzymes [87]. It is possible that this process also influences the epigenetic landscape of cells chronically exposed to BPA or GEN.

However, the deregulation of E2-signaling described herein is target-specific, suggesting that general alterations in the expression of chromatin modifying enzymes is not solely responsible for the ERE reprogramming caused by chronic EDC exposures. Instead, our data put forth a model wherein differential co-regulator recruitment results in transient changes in gene transcription. Upon chronic exposure, the repetitive activity of these cofactors induces epigenetic reprogramming, leading to long-term consequences on E2- signaling even after subsequent removal of the EDC.

2.5.4 Conclusions.

Our work demonstrates that chronic EDC exposures can permanently influence hormonal signaling. As such, the prevalence and persistence of compounds like BPA and

GEN in the environment of industrialized countries may have unrealized and underappreciated health consequences. Unfortunately, epidemiological studies investigating the impact of EDCs on human health are often conflicting, making it difficult to assess whether a specific compound has beneficial or detrimental consequences. Overcoming the complexities of ER signaling through the study of EDCs and their mechanisms of hormonal disruption will begin to elucidate suspected and observed consequences of our chronic exposure.

2.6 Acknowledgements

We thank Dr. C. E. Burd, X. Guan and K. LaPak for critical reading and advice on the manuscript. The research reported in this publication was supported by the

National Institute of Environmental Health Sciences of the National Institutes of Health under Award Number R00ES019918. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of

Health.

Figure 2.1 Chronic EDC treatment reprograms global gene expression. A.

Continued.

Figure 2.1 continued.

A model of chronic EDC exposure was generated by passaging cells >70 passages under normal FBS conditions (MCF7-F) with additional 50 nM BPA (MCF7-B) or GEN

(MCF7-G). B. Microarray analyses were performed on the three generated cell lines and two-dimensional hierarchical clustering applied to significant, differentially expressed genes (≥ 2-fold). C. Genes whose expression was higher (> MCF7-F) or lower (< MCF7-

F) by ≥ 2-fold in the MCF7-B and MCF7-G lines as compared to MCF7-F cells are represented by Venn diagram. D. Two-dimensional hierarchical clustering of the three cells lines was performed using E2-responsive genes significantly regulated (≥ 1.5-fold) in the MCF7-F cell line between EtOH and E2. E. A Venn diagram depicts the overlap of genes with increased or decreased expression following E2 treatment for 3 h.

Figure 2.2 Chronic EDC treatment reprograms E2-regulated gene expression. A-C.

Continued.

Figure 2.2 continued.

Control, MCF7-F (white columns) and chronically treated, MCF7-B (gray columns) and

MCF7-G (black columns) cells were induced with 10 nM E2 for 3 hours and relative mRNA was gene specific as E2-responsive targets either showed no change (A), decreased expression in both chronically treated cell lines (B), or altered expression that was EDC- specific (C). D. An measured by qRT-PCR at genes known to rapidly respond to E2. An altered E2 response was immublot of ERα protein expression specific (C). D.

An immublot of ERα protein expression was performed on the three cell lines to determine the effect of chronic EDC exposures on receptor expression. Low-passage

MCF7 cells is presented for comparison. β-actin is depicted as a loading control. E. We performed a replicate experiment of chronic EDC exposure in another E2-responsive breast cancer cell line, T47D. The E2 response showed results similar to the MCF7 cell lines at multiple ER target genes. RT-PCR data are expressed relative to the mean of

RPL13a mRNA, and are presented as a mean ± SE; n=5. * represents p ≤ 0.05. Fold activation over basal (EtOH) for each cell line is represented above the E2 treated column.

Figure 2.3 Deregulated ER response is independent of the activating ligand. A-D.

MCF7-F (white columns), MCF7-B (gray columns) and MCF7-G (black columns) cells were induced with 100 nM BPA or 100 nM GEN, respectively, for 3 hours and relative mRNA was measured by qRT-PCR for expression of PS2 (A), HDAC11 (B), EGR3 (C), and MSMB (D). Data are expressed relative to the mean of RPL13a mRNA, and are presented as a mean ± SE; n=4. * represents p ≤ 0.05. Fold activation over basal (EtOH) for each cell line is represented above the E2 treated column.

Figure 2.4 Chronic EDC treatment causes differential recruitment of ER. ChIP assays were performed to determine the level of receptor recruitment to EREs following activation with 10 nM E2 A-C. We examined genes unaltered (A), EDC-specifically altered (B) and mutually down-regulated by chronic EDC treatment (C) as depicted in

Figure 2. The response elements’ relative position to the transcriptional start site (TSS) is depicted under each graph. Enriched ChIP DNA was measured by qPCR and the data depicted is the mean ± SE; n=3. * represents p ≤ 0.05.

Figure 2.5 Epigenetic reprogramming is seen at EGR3 response elements. A. ChIP assays were performed on the MCF7-F, -B, and -G cells lines prior to ER activation with antibodies specific for the enhancer mark H3K4me1.

Continued.

Figure 2.5 continued.

ChIP enriched DNA was analyzed by qPCR with primers specific for response elements near the PS2, MSMB and EGR3 genes. Data depicted as the mean ± SE; n=3. B. DNase accessibility assays were performed under identical conditions to the ChIP assays.

Relative DNase resistance was measured by qPCR at the response elements associated with the PS2, MSMB and EGR3 genes. Data represents the DNA resistant to digestion relative to MCF7-F cells and is depicted as the mean ± SE; n=3. * represents p ≤ 0.05

Figure 2.6 Changes in the E2 response persist after removal of the EDC. A. Diagram depicting the experimental model to test sustained E2 signaling defects. MCF7-F, -B and

-G cells were all placed under normal culturing conditions in the absence of BPA and

GEN for a minimum of 2 passages followed by analysis of E2 signaling. B. Following either 2 or 9 passages off chronic EDC treatment, ER protein expression was analyzed as described in Figure 2.2D.

Continued. 51

Figure 2.6 continued.

C-E. The response of each cell line to 10 nM E2 was measured by qRT-PCR following 2 and 9 passages off chronic EDC treatment at genes shown to be unaffected (C), mutually repressed (D), and EDC-specifically deregulated (E) in Figure 2. Data are expressed relative to the mean of RPL13a mRNA, and are presented as a mean ± SE; n=3. * represents p ≤ 0.05. Fold activation over basal (EtOH) for each cell line is represented above the E2 treated column.

Figure 2.7 Histone protein expression. An immublot of histone protein expression was performed on the three cell lines to determine whether chronic EDC exposures affect histone modifications indicative of active transcription (H3K4me3) and enhancer activity

(H3K4me1) and total histone 3 protein.

Figure 2.8 Chronic EDC treatment causes differential recruitment of ER. ChIP assays were performed to determine the level of receptor recruitment to additional EREs following activation with 10 nM E2. Enriched ChIP DNA was measured by qPCR and the data depicted is the mean ± SE; n=3. * represents p ≤ 0.05.

Chapter 3: Varying susceptibility of the female mammary gland to in utero windows

of BPA exposure3

3.1 Abstract

In utero exposure to the endocrine disrupting compound bisphenol A (BPA) is known to disrupt mammary gland development and increase tumor susceptibility in rodents. It is unclear if there are different periods of in utero development that may be more susceptible to BPA exposure. Herein, pregnant CD-1 mice were exposed to BPA at different times during gestation that correspond to specific milestones of in utero mammary gland development. Mammary glands of early-life and adult female mice, exposed in utero to BPA, were morphologically and molecularly (ERα and Ki67) evaluated for developmental abnormalities. We found that BPA treatments occurring prior to mammary bud invasion into the mesenchyme (embryonic day (E), E12.5) incompletely resulted in the measured phenotypes of mammary gland defects.

______

3Andrea R. Hindman, Xiaokui Molly Mo, Hannah L. Helber, Claire E. Kovalchin, Nanditha Ravichandran, Alina R. Murphy,

Raleigh D. Kladney, Abigail M. Fagan, Pamela M. St. John and Craig J. Burd. Varying susceptibility of the female mammary gland to in utero windows of BPA exposure. (ACCEPTED, AUGUST 2017)

Exposing mice up to the point in which the epithelium extends into the precursor fat pad

(E16.5) resulted in a nearly complete BPA phenotype while exposures during epithelial extension (E15.5-18.5) resulted in a partial phenotype. Furthermore, the relative differences in phenotypes between exposures highlights significant correlations between early-life molecular changes (ERα and Ki67) in the stroma and the epithelial elongation defects in mammary development. These data further implicate BPA action in the stroma as a critical mediator of epithelial phenotypes.

3.2 Introduction

Estrogen signaling is critical for breast organogenesis and development.

Exogenous estrogenic exposures during in utero development can have life-long consequences. The first evidence of risk associated with in utero estrogenic exposures was from the use of synthetic estrogen, diethylstilbestrol (DES), prescribed to women beginning in the 1940’s. The use of DES was discontinued when the daughters exposed in utero - “DES daughters” - had a 40x increased risk of clear cell adenocarcinoma of the vagina and cervix [54-56]. Later investigation of these women revealed a 2-fold increased risk of breast cancer over the age of 40 years old [55]. While synthetic estrogens are no longer prescribed to pregnant women, rodent studies have demonstrated that in utero exposure to environmental estrogens, such as bisphenol A (BPA), can also contribute to mammary gland tumor formation [77, 82, 93] . BPA, like DES, has been shown to aberrantly bind and activate the estrogen receptor (ER) [53, 155].

BPA is an industrial plasticizer that is ubiquitous in our environments, having been detected in greater than 95% of the U.S. population, including human amniotic fluids [20, 65-70, 156]. Although the effects of in utero BPA exposure in humans have yet to be well-defined, the detriment of this exposure has been well-characterized in murine models [77, 78, 80-87], demonstrating increased cancer susceptibility similar to that seen in DES treated rodents [88-90]. Specifically, in utero BPA-treated early-life mammary glands have previously been shown to have an increased number of terminal end buds (TEBs) when normalized to reduced epithelial area[80], reduced number of terminal duct ends (TDEs), decreased epithelial elongation through the fat pad and less stromal proliferation [77, 83, 157]. Furthermore, in utero BPA exposure increases susceptibility to cancer through the development of ductal hyperplasias [91], and results in a higher incidence of tumors in models using predisposed transgenic alterations (NCR nu/nu, MMTV-erbB2) [92] or through a second assault by a carcinogen (such as 7,12- dimethylbenz[a]anthracene]) [93]. Similar phenotypes have been reported in rats [82, 84,

158, 159], where perinatal BPA exposure induced malignancies and lesions - without additional exposure to a carcinogen [94] - and in the rhesus monkey through altered development [79].

During embryogenesis, the fetal mammary gland goes through several developmental stages that are independent of estrogen activity [95-98]. ERα knockout mice have normal mammary glands at birth demonstrating that the receptor is only required after birth for mammary development [99, 100]. However, it is unclear if perinatal receptor activity is essential for driving BPA-induced pathologies. In utero

mammary gland development begins between E10.5 and 11.5 [101] when the primordial mammary epithelial buds form along the milk line. At this time, the mammary mesenchyme condenses underneath the developing epithelial bud [102] and serves as the only ERα-expressing compartment of the developing gland [86, 103]. Mouse ERα expression has been reported in the epithelium only in post-natal development [104].

Once these developing epithelial buds sink beneath the epidermis starting at E13.5, they remain completely surrounded by this ERα-positive mesenchyme until approximately

E16.5 [102], when a boost in proliferation causes the bud to sprout out of this mesenchyme, towards the fat pad and into primary rudimental branching [101]. It is unclear if the mammary gland has different susceptibilities to EDC exposure across these in utero developmental stages. However, reproductive tract tumors found in DES daughters were specifically associated with first trimester exposures [54-56]. Analysis of

DES exposure windows and breast cancer has not been performed, but as the interaction of the perinatal ERα-positive mesenchyme and the ERα-negative epithelial bud is developmentally modulated, the impact of estrogenic compounds on the developing gland may have very discrete windows of susceptibility.

Here, we examine different windows of in utero exposure to BPA in order to narrow the scope of developmental mechanisms responsible for the previously observed later-life epithelial defects and cancer susceptibility. While previous studies have extensively demonstrated other developmental time points of sensitivity, including peripubertal and pregnancy [92, 113, 114, 158, 160, 161], we have focused our efforts on the susceptibility of in utero exposures. BPA has been shown to cross the placental barrier

and has been detected in human amniotic fluid [20, 66-70, 156], thus we treated pregnant

CD-1 mice to a low dose (25 μg/ kg • bw/ day) of BPA during different periods of gestation. We then analyzed several previously reported phenotypes of these exposures in in utero-treated female offspring [77, 79, 81, 83-85, 93] to determine whether different time points of BPA exposure resulted specific mammary gland defects. We found that treating mice after the ERα-negative mammary epithelial bud had completely sunk into the condensed ERα-positive mesenchyme (E12.5) was sufficient to cause previously reported phenotypes. Furthermore, we revealed a significant correlation linking decreased stromal proliferation and increased ERα expression to delayed epithelial growth in early development suggesting that these defects may in fact be directed by mesenchymal dysfunction. Thus, our findings support a role for BPA to disrupt tissue compartment interactions of the developing mammary gland, driving later-life epithelial defects and cancer susceptibility that have previously been observed.

3.3 Materials and Methods

3.3.1 Animals.

Animal experiments were performed in compliance with protocols approved by The Ohio State University Institutional Animal Care and Use Committee

(IACUC, Protocol #2013A00000030) and in accordance with the accepted standard of humane animal care. Only female CD-1 mice were used in this study. Mice were maintained in polysulfone cages and fed a diet containing minimal levels of phytoestrogen (Harlan2019X). Sexually mature female CD-1 mice (8 weeks of age or

older) were mated and identification of a vaginal plug was taken to be embryonic day (E)

0.5. Pregnant mice were intraperitoneally (IP) injected daily with sesame oil vehicle control or 25μg/ kg • bw BPA (in sesame oil) between E8.5 and 18.5. This dose has been previously described, leading to alterations in the mammary gland and uterus of exposed mice [83, 87]. Mice from the same litter were divided amongst the different endpoints

(4.7 weeks, 14 weeks, 20 weeks) to minimize littermates being used for the same analysis. In sample sets using littermates, the litters were compared against the remaining samples to determine if the litter was statistically different. In cases when a litter was statistically different, the average of the littermates was used as one data point.

3.3.2 In utero exposure to BPA and tissue preparation.

To evaluate perinatal ERα expression in the developing mammary gland

(Figure 1), treated pregnant mice were euthanized at E13.5, 15.5, 16.5, 17.5, 18.5 and

19.5 and developing fetuses were harvested and genotyped for sex to omit males, fixed for 72 hours in 10% neutral buffered formalin, transferred to 70% ethanol embedded in paraffin and cut in 6μM sections. Sections were collected by taking four consecutive

6μM sections, skipping 30μM of tissue and repeating the four 6μM sections in order to encounter one of the developing placodes. Spacing out consecutive cuts was employed in order to provide favorability in locating mammary placodes and later epithelial rudiments. As a result, images are representative obtained for each perinatal time point evaluated. The regions of the Y chromosome used for sex determination were zinc finger protein, Y-linked (ZFY) [F: CACAGAAAGATGAACTTCAGAAAGA, R:

TTCAACTAAGCTACATTAAGTGACC] and sex-determining region (SRY)

[F:ATGTCAGCTGTTAGTAAGTAGGTAAG,

R:CTACACAGAGAGAAATACCCAAA].

The in utero dosing schemes utilized to examine discrete windows of exposure, and administered by IP injection as described, correspond to: BPA1 (E8.5-12.5): the period during milk line formation but prior to bud invagination into the mesenchyme;

BPA2 (E8.5-16.5): the period from milk line formation through bud invagination, but prior to epithelial invasion into the fat pad; BPA3 (E15.5-18.5): the period where epithelial extension begins toward the fat pad until it has begun branching; and BPA4

(E8.5-18.5) the duration of exposure that has resulted in previously reported phenotypes

(Figure 2A).

For the purpose of morphological analysis of susceptibility, the fourth inguinal mammary gland on each side of the animal at 4.5 (early-life, n ≥ 7), 14 and 20 weeks

(adult/ later-life, n ≥ 9 and 3, respectively) was harvested for each treatment group. For all animals at least 8 weeks of age, the estrus cycle was monitored by vaginal cytology

[162-165] and glands were harvested during the estrous stage. Stage of estrus was confirmed through Hematoxylin & Eosin (H&E) analysis of dissected uterus, ovaries and vagina from each animal [163, 165]. For each mouse, one gland was whole-mounted and carmine-aluminum stained, (using the National Toxicology Program Animal Studies

Protocol; section XIII, appendix 6, section E) [166] for morphological analyses; the other gland was fixed in 10% neutral buffered formalin for 24 hours, transferred to 70% ethanol, paraffin embedded and cut in 4μM sections for immunohistochemical analysis.

Thus, data presented herein are from matched mammary glands for whole mount morphological analyses and immunohistochemistry (IHC).

3.3.3 Amniotic fluid sample preparation and HPLC with fluorescence detection.

Beginning at E8.5, pregnant CD-1 female mice were IP injected as described, until E15.5. At E15.5, mice were IP injected with a final dose of BPA or oil vehicle and harvested 1 hour later for fetal amniotic fluid. Fluid from all embryonic sacs from 1 single female mouse was pooled for each treatment replicate. The liquid from amniotic fluid samples (on average approximately 600 l) was transferred to microcentrifuge tubes and lyophilized. A modification to an existing method for liquid-liquid extraction of BPA was developed [167]. Aliquots of 1000 l of chloroform:methanol (50:50) were added to each pellet and the samples were mixed on an Eppendorf Thermomixer at 1400 rpm for

24 hours followed by centrifugation at 10.2k rpm for 3 minutes. The supernatant was extracted and dried under vacuum and 500 l of acetonitrile was added to each tube.

Each sample was filtered using a 0.2 m Corning nylon syringe filter prior to HPLC injections. Chromatography was performed on a Waters HPLC instrument (WAT 600E controller, 717 plus autosampler) with a fluorescence detector (WAT 2475) that has a linear. A Waters Spherisorb ODS2 analytical column (4.6 x 250 mm, 5 m particle size) with a Supelguard LC-18 guard column (2 cm x 4 cm, 5 m) was used for analysis. An acetonitrile:H2O mixture was used as the mobile phase with a steep gradient beginning at

50% and ending at 100% acetonitrile over a period of 6 minutes at a flow of 1 ml/min

[168]. Each 10 l HPLC injection was excited with 225 nm light and fluorescence was collected at 305 nm. Chromatograms showing the fluorescence emission as a function of 62

time were collected for samples that originated from mouse amniotic fluid and for control samples consisting of BPA in acetonitrile (1 M and 1 nM concentrations), acetonitrile, and water. The retention times of BPA in the chromatograms from control samples were compared to those from the mouse samples to verify the presence of (untransformed)

BPA and to estimate the amount present in the amniotic fluid. The limit of detection of control solutions of BPA is sub-picomolar in concentration, however, in order to estimate the amount of BPA present in the samples, we relied on the linearity of the BPA fluorescence intensity as a function of concentration. The calibration curves generated showed that BPA from control solutions showed a fluorescence intensity that was linear with concentration to picomolar quantities. The average concentration of BPA in the amniotic fluid from pregnant mice on E16.5 one hour after IP injection was 0.13 nM

(data not shown).

3.3.4 Mammary whole mount morphological and immunohistochemistry analyses.

For morphological analysis of early-life (4.5 weeks of age) carmine-aluminum stained mammary whole mounts, epithelial elongation was measured in millimeters +/- from a line on the leading edge of the lymph node, to the most distal point of epithelial branching. Number of terminal duct ends (TDEs) was counted as the most distal ends of epithelial ducts and branches. Number of terminal end buds (TEBs) was counted on these same mammary whole mounts, with qualifying TEBs meeting a threshold of 0.15 mm in width. Epithelial area (%) was determined at this early-life time point, again on these same carmine-aluminum stained mammary whole mounts using color thresholding and selection in ImageJ [169]. Percent epithelial area was used to normalize the number of

TEBs determined for each treatment and is represented as a ratio. All measurements were performed on digital images obtained using a dissection microscope (Zeiss Stemi SV11

Apt stereomicroscope) at 0.6x magnification, equipped with a camera (AxioCam 506 color) and Zeiss ZEN software and confirmed through blind analysis (Figure 2, n≥ 7).

For morphological analysis of adult (14 and 20 weeks of age) carmine-aluminum stained mammary whole mounts, epithelial area (%) was also determined using color thresholding and selection in ImageJ [169] (Figure 5). Digital images were obtained similarly on a dissection microscope and with one field of view from each biological replicate being uniformly selected within the leading edge of the mammary gland in front of the lymph node (14 weeks, n ≥ 9; for 20 weeks, n ≥ 3). To further examine epithelial branching density and complexity, Sholl analysis was performed on the entire epithelial tree, as previously described [170] (Table 1, n ≥ 9). Images were obtained on a dissection microscope at 0.6x magnification. Briefly, the entire epithelial tree of each biological replicate was processed in ImageJ to optimize the contrast of the tree from the surrounding mammary fat pad, background was subtracted, and noise was removed; the grayscale image was thresholded and skeletonized to discern pixels of interest. Sholl analysis was performed using the plug-in for ImageJ. Since this analysis was performed on 14 weeks old whole mounted glands, instead of using the point of epithelial tree origin, a centralized point behind the lymph node was chosen across all glands analyzed.

Choosing such a point was previously reported to not affect Sholl results. Sholl profiles representative of treatment means are depicted as the heap map of concentric rings across the gland (Supplemental 2A), an inset depicting the skeletonized epithelium

(Supplemental 2B), a linear plot of number of intersections (N) versus the distance from the centralized point chosen behind the lymph node (Supplemental 2C) and a semi-log plot of epithelial branching complexity (number of intersections normalized to total mammary epithelial area, N/mm2) versus the distance from the centralized point chosen behind the lymph node (Supplemental 2D).

For IHC analysis of both early-life and adult glands, all sections were stained using a Bond Rx autostainer (Leica), using automated software (Bond Rx version 4.0) to run optimized protocols. Briefly, slides were baked at 65⁰C for 15 minutes, and automated software performed dewaxing, rehydration, antigen retrieval, blocking, primary antibody incubation, post primary antibody incubation, detection (DAB or RED), and counterstaining using Bond reagents (Leica). Samples were then removed from the machine, dehydrated through ethanols and xylenes, mounted, and cover-slipped.

Antibodies for the following markers were diluted in antibody diluent (Leica): rabbit antibodies αSMA (1:5000, Abcam), ERα (1:2000, E115 Genetex), and Ki67 (1:200, Abcam); and rat antibody: cytokeratin 8 (1:2000, TROMA-I-C, DSHB, University of Iowa) (Table

3.1).

Table 3.1 Antibodies used for staining.

IHC dual-stained for both early-life and adult glands was analyzed by first taking four field-of-view images, per biological replicate at 20x magnification using the Vectra

Automated Multispectral imaging system. At 4.5 weeks old mammary tissue, TDEs and ducts were imaged and for 14 weeks old mammary tissue, alveolar buds and ducts were imaged, resulting in 4 images per structure in each biological replicate. Tissue compartment segmentation and quantification of positivity was performed manually using InForm Advanced Image Analysis Software by Perkin Elmer (version 2.0.2)

(Figure 3.3, 4). Specifically, the InForm Tissue Finder software was used to determine tissue compartment positivity and this process consisted of the following: (1) manual tissue compartment segmentation was performed, using smooth muscle actin (SMA)-

positive staining as a designation of the epithelial compartment, (2) positive cells were identified by the object-based segmentation functionality, and (3) scoring utilizing 2-bin positivity to identify component DAB signal as positive for the staining being performed

(ERα or Ki67) and a threshold to differentiate between DAB+ and DAB- (H&E only) cell nuclei. Scoring was done separately for each tissue compartment, thresholds for positivity were determined for each compartment and applied across all images and the results were positivity (%) values normalized to the area of the total tissue by the InForm software.

3.3.5 Statistical analysis.

Data were analyzed by analysis of variance (ANOVA) followed by post-hoc comparisons. The multiplicity was adjusted by using Holms-Bonferroni method to control the overall type I error rate at 0.05 [171]. Pearson correlation and linear regression methods were used to evaluate the association between percent of ERα or

Ki67 positive cells and epithelial elongation (Figure 3.3D). SAS 9.4 software was used for data analysis (SAS, Inc; Cary, NC).

3.4 Results

3.4.1 In utero ERα expression in the developing mammary gland is strictly mesenchymal.

Many of the defects associated with in utero BPA exposures are observed in the adult epithelium; the misappropriated activation of ERα is postulated to be critical for the observed pathologies associated with increased cancer susceptibility. However, ERα has been reported to be expressed strictly in the mesenchymal cells surrounding the mammary epithelial bud in in utero mouse development [102, 172]. Thus, we first

examined the expression of ERα in the developing perinatal mammary gland in the presence and absence of BPA (Figure 3.1). Following in utero BPA treatment, ERα was examined by IHC in the developing mammary gland at E13.5, 15.5, 16.5, 17.5, 18.5, and

19.5. Observed expression at all in utero time points revealed that ERα was strictly expressed in the mammary mesenchyme surrounding the ERα-negative developing mammary epithelial bud (Figure 3.1, see oil 20x). At the E13.5 and 16.5 time points, virtually all peripheral epithelial cells are in contact with ER positive mesenchyme.

Notably, as the ERα-negative epithelial rudiments extend toward the mammary fat pad and past the primary ERα-positive mesenchyme (after approximately E15.5), more ER negative stromal cells are in contact with the peripheral epithelial cells (Figure 3.1, oil

E18.5, 20x; oil 18.5, 10x). In utero exposure to BPA showed similar expression of ERα to oil treated controls. (Figure 3.1 BPA, bottom images).

3.4.2 Discrete in utero windows of BPA treatment cause varying degrees of mammary gland phenotypes in the developing gland.

To date, previous studies have focused on exposures spanning all of in utero mammary gland development. In order to assess whether the various stages of in utero mammary development have different susceptibilities to BPA exposure, we utilized four specific dosing schemes structured around mammary gland in utero developmental milestones, to assess whether a window of exposure would best account for the well- characterized later-life epithelial defects (Figure 3.2A). Windows of BPA exposure were designed to include the period during milk line formation but prior to bud invagination into the mesenchyme (BPA1: E8.5-12.5); the period from milk line formation through

bud invagination (BPA2: E8.5-16.5); the period where epithelial extension begins toward the fat pad until it has begun branching (BPA3: E15.5-18.5); and the duration of exposure that resulted in previously reported phenotypes (BPA4: E8.5 -18.5). Pregnant mice were

IP-injected daily with oil or 25 μg/ kg• bw BPA for the indicated time periods between

E8.5 and E18.5 (Figure 3.2A). Using carmine-stained mammary whole mounts of in utero-treated 4.5 weeks old mice, we found that the previously reported delayed epithelial growth phenotype was dependent on the in utero window of BPA treatment received.

Specifically, epithelial elongation from the leading end of the mammary gland lymph node (dashed line) to the most distal epithelial branching (solid line) demonstrated a clear defect in mice treated after E12.5 (Figure 3.2B). Mice treated prior to E12.5 (BPA1) had similar elongation to control, while the other in utero windows of treatment, BPA2 (p <

0.001) and -3 (p < 0.05) were found to be significantly less elongated compared to control and similarly to mice treated throughout all of in utero mammary development

(BPA4, p < 0.001) (Figure 3.2B, n ≥ 7). Early-life glands showed that only BPA4 (p <

0.005) treatments demonstrated a statistically significant reduction in number of TDEs

(Figure 3.2C, TDEs indicated by white arrows and including black arrows; graph Figure

3.2D). Though following the trend of decreased number of TDEs, BPA2 and BPA3 did not reach significance. Further, recapitulation of phenotypes observed in previous studies included an observed increase in the ratio of TEBs (Figure 3.2C, TEBS indicated by black arrows; graph Figure 3.2E) to total epithelial area in treatment periods beyond

E12.5 (Figure 3.2D) with BPA4 reaching significance (p < 0.005), while BPA2 and

BPA3 demonstrated a phenotypic trend that did not reach statistical significance.

Using mammary gland tissue sections from the matched inguinal glands of the 4.5 weeks old mice, we examined ERα expression and proliferation (Ki67) in both the epithelial tissue compartment and the stroma surrounding terminal ducts ends (TDEs, these included TEBs) following varied in utero BPA treatment. IHC staining of smooth muscle actin (SMA) was used as a marker to differentiate the epithelial tissue compartment from the surrounding stroma. Positivity for ERα and Ki67 was determined within the epithelium (inset, green) and stroma (inset, pink) for each treatment (Figure

3.3A, B; n ≥ 7). ERα-positivity in the stromal compartment surrounding TDEs was found to be significantly increased in the BPA2 (p < 0.05), -3 (p < 0.05), and -4 (p < 0.001) treatments compared to control, and BPA1 was not (Figure 3.3A, STROMA).

Conversely, Ki67- positivity in the stromal compartment surrounding TDEs was found to be significantly decreased in the BPA2 (p < 0.001), with a trend towards decreased Ki67 positivity in BPA3 and -4 treatments compared to control not reaching significance

(Figure 3.3B, STROMA). Ductal Ki67 positivity was also quantified in both tissue compartments with no significant changes observed across treatments (data not shown).

Molecular changes, both ERα and Ki67, in TDE epithelium were only significant in

BPA4 (Figure 3.3A, B, EPITHELIUM).

Notably, early-life dysfunction of mammary gland epithelial elongation (Figure

3.2B) correlated with the molecular changes in the stromal compartment (Figure 3.3C).

Cross comparison of all mammary glands, regardless of the exposure, revealed that epithelial elongation significantly correlated inversely with the ERα-positivity in the stroma surrounding the TDEs ERα expression (Figure 3.3C, left graph, R = -0.54, p <

0.0001) and directly with proliferation (Figure 3.3C right graph, R = 0.44, p < 0.001).

This correlation whole holds true even if the BPA4 treatment representing all exposures is removed (data not shown). Epithelial elongation also significantly correlated to TDE epithelium ERα- (Figure 3.3D, left graph, R = -0.32, p < 0.05) and Ki67-positivity

(Figure 3.4D, right graph, R = 0.43, p < 0.05), but only when BPA4 was considered along with all other treatment groups.

3.4.3 Varied in utero windows of BPA treatment change later-life ERα expression in the tissue compartments of the mammary gland.

To evaluate whether the early-life defects (Figures 2 and 3) propagated to later- life mammary development, we looked at morphological and molecular changes in adult

14 weeks old mice following our different windows of in utero exposure. Examining dual-stained tissue sections taken from in utero-treated 14 weeks old mice, we found that each treatment window had different ERα-positivity profiles, particularly for the epithelial-alveolar structures and surrounding stroma (Figure 3.4, n ≥ 4). Consistent with the pubescent developing mammary gland, BPA1 mice showed unchanged ERα- positivity in both the stromal and epithelial compartments of the adult mammary gland when compared to control. Comparable to BPA1, BPA3 mice also showed no changes in when compared to control in either tissue compartment. However, ERα-positivity was significantly decreased in the BPA4 treatment compared to control in both tissue compartments (p < 0.05). While BPA2 did not reach statistical significance in stromal

ERα positivity compared to control, it was significantly decreased in the epithelial

compartment (Figure 3.4). Ki67-positivity was not significantly different than control in any treatment or tissue compartment in the adult mammary gland (data not shown).

3.4.4 In utero BPA exposure alters ductal morphology to varying degrees based upon the window of treatment.

To evaluate the ability of our model to recapitulate previous reports that in utero

BPA exposure increases epithelial volume and alters ductal morphologies [77, 83], carmine-stained mammary whole mounts of adult 14 (Figure 3.5A, n ≥ 9) and 20 (Figure

3.5B, n ≥ 3) week-old mice were examined. Similar to the trends seen for ERα staining at

14 weeks of age (Figure 3.4), exposures BPA2 and -4 showed statistically significant increases in epithelial area (Figure 3.5A, p < 0.05), while no change was observed in the

BPA1 and -3 treatments. This phenotype was consistent through 20 weeks (Figure 3.5B, p < 0.05). Interestingly, BPA1 and -3 treatments did show a trend towards increased epithelial area at this later time point of 20 weeks, but it did not reach statistical significance (p = 0.051 and 0.082, respectively). Further investigation of alterations to the mammary epithelium was performed through use of Sholl analysis to measure branching density and complexity of 14 week old mice [170]. The BPA2 (p < 0.05) and -4 (p <

0.0001) treatments demonstrated significantly more intersections than control mice

(Table 3.2, Sum N column). The BPA1 group was indistinguishable from the control mice. The BPA3 treatment was comparable to the trend set by BPA2 and -4 mice, but did not reach statistical significance compared to control (p = 0.13). Further, the branching density (N/mm2) which normalizes the number of intersections to the total area of the mammary gland epithelium revealed a significant difference compared to control, again

in both BPA2 (p < 0.005) and -4 (p < 0.001) treatments (Table 3.2). Branching complexity, or the calculated Sholl regression coefficient (ƙ), is described by the degree branching continues in distal regions of the epithelial tree [170]. The calculated value for the coefficient, ƙ, was significantly lower in BPA2 (p < 0.05) and -4 (p < 0.0001) compared to control which indicated that the epithelial branching is more consistent/ complex throughout the entire epithelium [170]; BPA3 followed this trend without reaching statistical significance (Table 1, p = 0.053). Taken together, these data revealed that the majority of in utero BPA phenotypes throughout development can be recapitulated and propagated to adult life, solely with exposures post E12.5.

3.5 Discussion

3.5.1 The susceptibility of the mammary gland to BPA exposure is not uniform throughout gestation.

In utero EDC exposures initiate detrimental effects upon hormone-dependent organogenesis that increase the risk of developing cancer. Studies of DES daughters demonstrated that exposures during the first trimester resulted in the greatest risk for developing vaginal cancers [55, 173, 174]. However, no similar analysis has been performed to identify a specific window of susceptibility for breast cancer. While BPA has been shown increase terminal end buds and cancer risk in rodents similar to the in utero effects of DES [88-90], the developmental mechanism of this disruption and propagation to later-life end points remains unclear. We thus tested various windows of in utero exposure to BPA based on the stages of murine mammary gland development

[102, 175, 176]. Utilizing this model consisting of four BPA dosing schemes, we revealed that developmental defects in the mammary gland were observed when the mice were treated after the epithelial bud had invaded the ERα-positive mesenchyme (post

E12.5). BPA1 (E8.5-12.5) elicited little to no evidence of any previously reported defects. Consistent with previously reported phenotypes, BPA2 and -4 (post E12.5 through E18.5) were found to exhibit significant defects. At our early-life time point of

4.5 weeks old, BPA exposure resulted in delayed epithelial elongation, and a trends toward both reduced number of TDEs and increased ratios of TEBs normalized to the total epithelial area (Figure 3.2B and C). While most BPA treatments, BPA1-3, recapitulated the previous finding that proliferation within the epithelial compartment of the TEB, our BPA4 exposure, in contrast, had a significant decrease in proliferation

(Ki67-positivity) compared to control. However, this discrepancy could be due to the fact that this measurement was done in a mixture of TDEs and TEBs, since we did not qualify

TEBs specifically with a size threshold. Overall, the majority of significant early- and later-life phenotypes were observed in windows of exposure including E12.5 to E16.5

(BPA2-4).

Interestingly, the developmental window E12.5 to 16.5 correlates to a time when the potential for the most direct and complete influence of the ERα-positive mesenchyme upon the ERα-epithelium is observed (Figure 3.1, E13.5 and 16.5). BPA3 only exhibited a partial phenotype, not reaching significance in any of the later-life molecular and morphological phenotypes observed. The BPA3 treatment corresponds to an in utero time

in mammary development when the elongating ERα-negative epithelial rudiments are no longer completely surrounded by the ERα-positive mesenchyme.

A potential caveat of our model is that the most significant treatments, BPA2 and

-4 also happened to span the longest period of in utero development (8 and 10 days compared to 4 and 3 days of BPA1 and -3, respectively). However, it is important to recognize that BPA3, while only spanning 3 embryonic days of exposure, did result in a significant but diminished phenotype while the BPA1, spanning 4 days, had no observable changes to the mammary gland. Thus, the duration of exposure is not the principle driver of the observed defects. Furthermore, due to the rate of clearance of BPA from the animal, exposures to BPA persist after the treatment window. For example, while the BPA2 treatment group received a final BPA dose on E16.5, previously published work has demonstrated that BPA administered by 10mg/kg IP injections (40x higher than the dosage used herein) only becomes undetectable 18-24 hours after injection [177]). Thus, the exact exposure period extends beyond the treatment windows and cannot be discretely demarcated. Nonetheless, the different windows in our model produced varying degrees of mammary gland phenotypes and treatments prior to E12.5 produced defects that were not significant for most end points measures.

3.5.2 The role of ERα in mediating the effects of in utero BPA exposure.

The effects of BPA have long been suspected to elicit deregulated mammary development and later-life susceptibilities through its ability to bind and differentially activate the estrogen receptors [53, 131, 155, 178-180]. Evidence suggests that ERα activation in the stroma may be the driver of BPA-induced defects [85, 86, 104]. More

recently, BPA-induced defects in mice were shown to be blocked by treatment with the estrogen antagonist fulvestrant, suggesting that BPA may act through nuclear ERs in the stroma [181]. Together with our work herein, supports classical findings that the stroma plays a pivotal role in later epithelial morphogenesis [176, 182-184], and extends the impact of the stroma to be involved in the mechanism of disruption by BPA to affect later-life susceptibilities. Thus, the BPA-induced mesenchyme may reprogram the epithelium through paracrine signaling following inappropriate ERα activation. As a result, one would predict that paracrine signaling would have the most influence at the point in mammary development when the epithelial bud is completely surrounded by

ERα-positive mesenchyme. In fact, our BPA2 and -4 treatments displayed the most significant phenotype from BPA exposure. This period represents the time beginning at

E12.5, when the developing epithelial placode has extended into the underlying dermis and ERα-positive mesenchyme [102, 183]. At E16.5, the epithelial bud sprouts out of the mesenchyme and towards the fat pad (Figure 3.1, diagrammed in Figure 3.2A). While

ERα-positive stromal cells are observed around the extending epithelial bud at this later time point, most expression remains near the nipple sheath (Figure 3.1, E18.5 10x). Thus,

ERα-dependent paracrine signaling after bud extension may be affecting a smaller percentage of the developing epithelium. This possibility is highlighted by mice treated after the epithelial bud had begun to elongate towards the fat pad (BPA3) which demonstrated a significant epithelial phenotype in early development (Figure 3.2B and

3.3A), but failed to show significant defects in the adult gland (Figure 3.4, 3.5 and Table

1). This finding suggests that either the influence of BPA-mediated paracrine signaling

during this time is weaker, or that some phenotypes are induced through ER-independent mechanisms.

Previously, it has been shown in mice that only 4% of stromal cells are weakly

ERα-positive immediately in postnatal development, and continues to increase in intensity and number of positive cells through puberty and adulthood [172]. ERα- positivity within the control-treated early-life mammary gland recapitulates this finding

(Figure 3.3A, STROMA). Interestingly, BPA causes an increase in the ERα expressing cells of the TDEs at 4.5 weeks (BPA4) compared to significantly reduced epithelial expression at 14 weeks as observed in BPA2 (p < 0.05) and -4 (p < 0.05). The importance of ERα signaling in the stroma of the developing mammary gland has until recently remained unclear. A recent study has shown that BPA can directly alter mammary gland development in an ex vivo model system that eliminates the effects BPA may have on the hypothalamic-pituitary-gonadal axis. These data support the hypothesis that BPA acts directly on the ERα positive stroma to influence mammary gland development.

The demonstrated ability of BPA to interact with other hormone receptors, including ERβ and a membrane-bound receptor (GPR30 or GPER), that act on mammary gland development cannot be dismissed. Consequently, both receptors, like ERα, initiate hormone-responsive activity and signaling effects [155, 178, 179, 185, 186]. However, it is clear from knockout models of these receptors that ERα predominantly facilitates later- life mammary gland development [95-98, 185]. Furthermore, mice displaying the most mammary gland defects (BPA2) were exposed during a period when ERβ expression is not detected in the developing epithelial bud (E18) [86]. Taken together, our exposure

model supports the idea that BPA specifically acts on in utero mesenchymal ER positive cells to reprogram signaling that imparts later-life defects observed; emphasizing ERα activation must be critical for BPA disruption of mammary gland development.

The impact of ERα signaling in our murine model, however can be convoluted by the potential differences in mammary gland in utero expression of the receptor compared in mice and humans tissue compartments. While mice lack epithelial ERα expression until after birth, humans do express the receptor beginning after week 30 of gestation.

Furthermore, no human study has ever demonstrated in utero mesenchymal ERα expression. Few studies have looked prior to week 18 of gestation in the human mammary gland for ERα expression [187-189], and no study has looked prior to week

12. The period of mouse development that is most susceptible to in utero BPA exposure in our mouse model (E12.5-16.5) corresponds to weeks 7-14 of human gestation, well before any ERα epithelial expression in humans [172]. This period represents the end of the first trimester of human gestation, and is also the window in which susceptibility to vaginal cancers and particularly to the vaginal mis-differentiation defects, was found to be the highest [55, 190]. While no study has correlated breast cancer susceptibility to the timing of exposure, one study did correlate these vaginal defects to breast cancer risk [55,

56]. These combined data suggest that human susceptibility to cancer in the mammary gland occurs prior to epithelial ERα expression in the third trimester.

3.5.3 The impact of stromal signaling on mammary gland defects.

Our model of discrete in utero windows of BPA exposure further supports the long-established importance of stromal-epithelial interactions in mammary gland

morphogenesis [176, 182, 183, 191]. Estrogenic compounds can act directly on mammary gland tissue to effect epithelial development despite the lack of ER in the epithelium [181]. Furthermore, the fetal transcriptomes of each tissue compartment were specifically and differentially altered by BPA exposure and highlighted the potential of

BPA to disrupt the interactive partnership between the epithelium and stroma in the developing breast [86]. However, it is unclear whether these changes to the transcriptome are both necessary and sufficient to drive adult phenotypes and susceptibility to cancer.

The interactions between the respective tissue compartments are vital in the development of the breast; disruption can promote tumorigenesis [183, 191-196]. Thus, determining the influence in utero BPA exposure has upon these vital processes and receptor signaling is paramount in determining the mechanism by which BPA mediates later-life breast cancer susceptibility. In early development of the mammary gland following in utero exposure to BPA, we found a significantly increased number of ERα- positive stromal cells surrounding the TDEs (Figure 3.3A, STROMA). As ERα-positive stromal cells are reported to increase during puberty [172, 197], this observation may suggest that the mammary glands are undergoing precocious puberty. Furthermore, stromal ERα staining has been previously shown to be confined to undifferentiated mesenchymal cells rather than adipocytes or fibroblasts [197], suggesting an increase in mesenchymal stem cells (MSCs) which are known to promote tumor growth and metastasis. Taken together, BPA may stimulate epithelial development indirectly through stromal ERα-positive progenitors [80, 83, 91, 193, 198].

3.5.4 Conclusions: Different windows of varying mammary gland susceptibility to in utero BPA exposure.

Our murine in utero model showed significant defects in the developing mammary epithelium (Figures 3.2-5) despite their lack of ERα expression at the time of

BPA exposure (Figure 3.1). These significant developmental defects are associated with in utero treatments that occur when the mammary epithelial bud was completely surrounded by the ERα-expressing mesenchyme (E12.5-16.5, BPA2-4) (Figure 3.1 and

3.2A). Our findings reinforce the idea that the timing of exposure even within the perinatal period can have a significant influence on later-life cancer susceptibility. This idea tragically proceeds from the treatment of pregnant women with both DES in the

1940’s and thalidomide in the 1950’s. Both drugs, elicited developmental defects in offspring specific to the in utero exposure received [55, 199]. In the case of DES, the comprehensive follow-up report emphasized that women exposed in utero and at least 40 years of age, had a 2-fold increased incidence of breast cancer [54-56]. The risk was more pronounced in women with vaginal epithelial changes previously shown to be associated with first trimester exposures. In humans, the initial mammary bud invades the parenchyma during the 8th week of gestation correlating to E12.5 of murine development

[172]. It was evident in our investigation that this period elicited the most detrimental epithelial defects both in early and adult mammary development. Together, these data implicate reprogrammed ERα activation in the fetal stroma as a mediator of later-life mammary gland defects following exposure to estrogenic compounds in our

environment. Our work clarifies the need to demonstrate the requirement of ERα expression in the stroma during in utero exposure to BPA.

3.6 Acknowledgments

We would like to thank Raleigh D. Kladney for his expertise in IHC staining. We would like to thank Jason Bice for his reliable support in tissue preparation for IHC. We would like to thank James Dowdle, PhD, María Cecilia Cuitiῆa DVM, PhD and

Christopher Koivisto in methods support for estrus staging and fetal tissue preparation, respectively. Finally, we would like to thank Corinne Haines, Rebecca Hennessey, Kyle

LaPak and Xiangnan Guan for their critical reading of this manuscript.

Figure 3.1 - ERα expression is strictly mesenchymal throughout perinatal murine mammary gland development. Immunohistochemistry of representative time points of fetal ERα expression at embryonic days (E) 13.5, 16.5 and 18.5 of in utero treated mice with oil control and BPA. Arrows indicate exterior of the fetus and origin of invaginating epithelial cells from the epidermis, nipple sheath. At no embryonic time point was ERα expressed in the mammary epithelium.

Figure 3.2 – BPA causes early-life defects in epithelial elongation within the mammary gland. A. Schematic time line of varied in utero treatment model with oil control or windows of in utero treatment with 25 μg/ kg• bw/ day BPA, beginning embryonic day (E) 8.5 and varied through 18.5. BPA1: E8.5-12.5; BPA2: E8.5-16.5;

BPA3: E15.5-18.5; and BPA4: E8.5-18.5. Dissection light microscope images representative of mammary glands from 4.5 weeks old mice were harvested and whole mounts stained with carmine aluminum. B. Epithelial elongation indicated by direction of growth +/- from the leading edge of the lymph node (dashed line) to the most distal rudiment of epithelial tree (solid line). Graph represents comparison of epithelial elongation in millimeters, between treatments; each dot represents separate biological replicates, n ≥ 7.

Continued. 83

Figure 3.2 continued.

C. The number of terminal duct ends (TDEs) is indicated by white and black arrowheads and the number of terminal end buds (TEBs), meeting a size threshold of ≥ 150 pixels in width, are indicated by black arrowheads. D. The number of TDEs, including TEBs, was quantified between the various BPA treatments. E. The number of TEBs normalized to total epithelial area is plotted. Each dot represents a separate biological replicate, n ≥ 7.

Open circle symbols indicate the carmine-stained mammary gland image representative of each treatment mean in (B) and (C). * = p < 0.05.

Figure 3.3 – Early-life epithelial dysfunction of the mammary gland by in utero BPA significantly correlates to the stromal compartment. Representative immunohistochemistry (IHC) of an oil in utero-treated mammary gland from 4.5 weeks old mice, dual-stained with smooth muscle actin (SMA) and A. ERα or B. Ki67 (images,

20x magnification); inset is the same oil-treated to show tissue compartment segmentation by InForm Advanced Image Analysis Software, used to differentiate the stromal (red) and epithelial (green) tissue compartments. The in utero treatment windows with BPA are derived similarly as indicated. Graphs present A. ERα or B. Ki67 positivity

(%) in the stroma (left), and epithelium (right) of the terminal duct ends (TDEs), including terminal end buds. Each dot represents separate biological replicates, n ≥ 7 and open circle symbols indicate representative images of the mean shown for each treatment.

Continued.

Figure 3.3 continued.

Correlation of epithelial elongation across treatments (Figure 3.2B) to C. the stromal compartment surrounding the TDEs ERα- (left) and Ki67 (right) positivity and D. the epithelial compartment of TDEs ER- (left) and Ki67-positivity (right). * = p < 0.05, ** = p < 0.0001. Dotted lines indicate 95% confidence interval.

Figure 3.4 - In utero BPA changes later-life ERα expression of mammary tissue compartments. Representative IHC of oil and varied with BPA in utero-treated mammary glands from adult 14 weeks old mice, dual-stained (as in Figure 3.3) with

SMA and ERα focused on alveolar structures (images, 20x). Graphs present ERα positivity (%) quantified in each tissue compartment of the alveolar structures of the epithelium (left) and the surrounding stroma (right). Each dot represents separate biological replicates, n ≥ 4 and open circle symbols indicate representative images shown for each treatment. * = p < 0.05.

Figure 3.5 - In utero BPA significantly alters the density and complexity of the mammary epithelium. Epithelium quantitation by color thresholding in ImageJ as pixels in epithelium out of total pixels in field of view (%).Representative dissection light microscope images of carmine-stained mammary gland whole mounts from adult mice,

A. 14 weeks old and B. 20 weeks old, treated with oil and for varied windows in utero with BPA (as described in Figure 2). Graphs present area of epithelium (%) as quantified in ImageJ. Each dot represents separate biological replicates, A. n ≥ 9 and (B) n ≥ 3 and open circle symbols indicate the mammary gland representing that treatment as shown. *

= p < 0.05.

Table 3.2 - Sholl analysis of epithelial branching density and complexity. Number of mice per treatment group is indicated in parentheses (n ≥ 8). Values are means ± SEM;

Sum N: total intersections in mammary epithelial area (MEA); log (N/mm2): intersections/ MEA less the area occupied by the lymph node or branching density; ƙ:

Sholl regression coefficient or branching complexity, the slope of log (N/mm2); CV: coefficient of variation. * p < 0.05, and ** p < 0.0001.

Figure 3.6. Negative staining controls for IHC. Immunohistochemistry was performed without primary antibodies on 4.5 weeks old mammary gland tissue for A. ERα and B.

Ki67 (Table 1), applying DAB for colorimetric detection, optimized time determined for each antibody. C. Staining for ERα performed on ER KO adult mammary and uterine tissue demonstrating antibody specificity. ERα KO animals were generated by crossing

ERfloxed animals [200] to Sox2-Cre (Jackson Laboratories).

Figure 3.7. Representative data from Sholl analysis for each in utero treatment. All data shown are representative of the means presented in Table 1 and were performed on the entire 14 weeks old epithelial tree of the mammary gland, using a centralized point behind the lymph node to begin the radii for Sholl analysis. A. Heat maps of concentric rings across the mammary gland epithelial tree. B. An inset of branching contained in the leading edge of the epithelial tree. C. Linear plot of the number of intersections (N) versus the distance from the centralized point chosen behind the lymph node. D. Semi- log plot of epithelial branching complexity (number of intersections normalized to total mammary epithelial area, N/mm2) versus the distance from the centralized point chosen behind the lymph node. The nested table indicates the ending radii and the number of times the Sholl concentric rings intersect the epithelial tree (N) of the representative treatment replicate.

Chapter 4: BPA in utero exposure alters later-life mammary development through

deregulated transcriptional programming and extracellular matrix remodeling4

4.1 Abstract

Widespread exposures to the endocrine disrupting compound (EDC) bisphenol A

(BPA) have been demonstrated in human populations through diverse epidemiological studies. Further, there is a mounting body of experimental evidence in rodent and non- human primate models that clearly demonstrate alterations to mammary gland development, function and disease susceptibility as result of in utero BPA exposures.

While these phenotypes are well-characterized, the molecular mechanisms are unknown.

The heterogeneity of the mammary gland through its discrete tissue compartments, and even more so through the component cell types, presents complexity in the determination of a mechanism for BPA disruption of development to impart later-life risk of disease and cell type-specific alterations have not been examined. Thus, we have isolated component mammary gland cell types (luminal epithelia, basal epithelial/ mammary stem cells and fibroblasts) in order to tease out discrete alterations that may mediate the consequences of in utero BPA exposure.

______

4Andrea R. Hindman, XingYan Kuang, Kyle Voytovich, Claire E. Kovalchin, Alex Avendano, Clarissa Wormsbaecher, Jason Bice,

Hannah L. Helber, Alina R. Murphy, Ali C. Shapiro, Jonathan Song and Craig. J. Burd. (UNPUBLISHED). 92

Utilizing RNA-sequencing methods we identified significantly deregulated genes for each component cell type in response to in utero BPA exposure and performed pathway analysis. We found that extracellular matrix organization (ECM) was a major pathway deregulated across all cell types. Surprisingly, we found either a significant increase in

ECM component macromolecules. Normal homeostasis of the ECM was demonstrated to be disrupted through the measurement of hydraulic permeability through an in vitro collagen matrix, as remodeled by in utero BPA-treated mammary fibroblasts. This provides a mechanism by which the effects of in utero BPA exposure is propagated to later-life, specifically attributed to the fibroblast cell population.

4.2 Introduction

Bisphenol A (BPA) is an endocrine disrupting compound (EDC) ubiquitous in our environment due to its prolific use in the manufacture of plastics and resins [48, 61, 63,

64]. As a result, detectable and biologically relevant levels of BPA have been found in numerous human biological tissues and fluids, and of most concern, human amniotic fluid [20, 68, 126]. Importantly, in utero and other early-life exposures in rodents and primates have demonstrated alterations to mammary gland morphogenesis and function, and increased susceptibility to disease [77-80, 114]. Specifically, perinatal BPA exposure in mice has been shown to induce intraductal hyperplasias [91] and cause in situ carcinoma in rats [94]. Further, an increased risk of tumorigenesis in models of breast cancer in mice was observed following exposure to carcinogen [93] and similarly in rats

[82]. It is clear BPA targets the mammary gland and that early-life exposures are

propagated to later-life in the form of increased susceptibility to disease. Taken together with observations that more developed and industrialized countries, like the U.S. have a higher incidence of breast cancer risk than their less developed counterparts [6-9] adds urgency to understand BPA’s impact on the human population.

The consequences of early-life exposure to BPA are not known in humans.

However, the derivation of risk associated with BPA stems from prescribed exposures to an EDC and known carcinogen, diethylstilbestrol (DES). Following DES administration to pregnant women, their offspring, or “DES daughters” were found to have increased incidence of vaginal and cervical cancers and a 2-fold increased incidence of breast cancer [54-56]. The risk for breast cancer did not manifest until after the age of 40 years old in these women. Since the widespread use and manufacture of BPA beginning in the

1950’s, it is anticipated that we have not reached a point where we would realize the consequence of this compounds’ omnipresence in our environment.

To date, the morphological phenotypes and observed susceptibilities resulting from early-life exposure to BPA are well-characterized in rodent models; however the molecular mechanisms remain largely undefined. Further, much of the weight-of- evidence for the purposes of elucidating the BPA exposure mechanism have focused on the context of embryonic, close to the time of or during in utero exposures, and early-life

[86, 87, 181]. While it is plausible that alterations at these early times in development may mechanistically account for observed later-life susceptibilities, alterations propagated and the mechanism of conserving those changes to adult-life may also prove to be just as informative for determining the BPA in utero mechanism.

Another layer of impact relating to the complexity of BPA action in the mammary gland stems from the glands’ intricacies of both tissue-dependent interactions, between the epithelia and stroma, as well as component cell types. The epithelial ducts have an interior lumen lined with luminal epithelial cells, 30% of which typically express ERα

[107]. This luminal layer is then surrounded by ER-negative basal epithelial cells, approximately 20% of which are comprised of mammary stem cells (MaSCs) [108, 201].

This basal layer is in contact with the basement membrane of the epithelial duct and on the exterior, closely surrounded by fibroblasts which are critical for extracellular matrix

(ECM)-mediation of epithelial development and architecture [109, 110]. The mammary fat pad is dominated by adipocytes. Additionally component cell types of the mammary gland include immune, lymphatic and vascular cells.

Recent work has sought to define tissue-specific alterations to impart a mechanism resulting from in utero BPA exposure by utilizing laser capture microdissection and examining transcriptome changes at embryonic day (E) 19.5 [86].

They demonstrated transcriptional effects to be more pronounced in the epithelial compartment, specifically, changes in genes related to ECM organization and adipogenesis. Further, they concurrently evaluated these changes in an ERα knock-out mouse model, thereby demonstrating the necessity of the receptor in these observed transcriptional effects. However, given the cellular heterogeneity of the mammary gland and demonstrated latency of later-life susceptibility as a result of BPA in utero exposure, we postulated that cell type-specific nuances may be lost in simply evaluating dissected, whole tissue compartments. Through our examination of discrete adult mammary gland

cell types, we expect to elucidate changes neglected by previous studies performed on whole tissues.

Thus, herein, we have treated pregnant CD-1 mice in utero with BPA to evaluate transcriptional changes at an adult time point of mammary gland development (14 weeks of age) at a dose, similarly administered and previously shown to elicit gland morphological defects associated with susceptibility [87]. This dose is also well below the Environmental Protection Agency’s calculated reference dose of 50µg/ kg • bw daily

[48, 202]. In order to account for heterogeneity, we employed a mechanical and enzymatic method to isolate mammary component cell types followed by fluorescence- activated cell sorting (FACs) to discern component epithelial cell populations [203]. We found little overlap of deregulated transcript abundances between component cell types; however similar pathways were deregulated relating to ECM organization as well as other mediators of tissue architecture and cellular protein production. To this end, macromolecule components of the ECM (collagens) were observed to be significantly increased as a result of in utero BPA exposure. The functionality of these transcriptional and collagen deposition changes were demonstrated in reduced hydraulic permeability of an in vitro collagen matrix as remodeled by BPA-treated mammary fibroblasts. These findings support a model that more specifically imparts the mechanism of BPA-mediated later-life risk to in utero exposed mammary fibroblasts and corroborates previous findings that ECM remodeling and tissue stiffness play a critical role as a risk factor in breast cancer development.

4.3 Materials and Methods

4.3.1 Animals.

Animal experiments were performed in compliance with protocols approved by

The Ohio State University Institutional Animal Care and Use Committee (IACUC,

Protocol #2013A00000030) and in accordance with the accepted standard of humane animal care. Only female CD-1 mice were used in this study. Mice were maintained in polysulfone cages and fed a diet containing minimal levels of phytoestrogen

(Harlan2019X). Sexually mature female CD-1 mice (8 weeks of age or older) were mated and identification of a vaginal plug was taken to be embryonic day (E) 0.5. Pregnant mice were intraperitoneally (IP) injected daily with sesame oil vehicle control or 25μg/ kg • bw

BPA (in sesame oil) between embryonic days (E) 9.5 and 18.5. This dose has been previously described, leading to alterations in the mammary gland and uterus of exposed mice [83, 87]. All experimental treatments and end points are representative of ≥3 litters to mitigate litter effects.

4.3.2 Harvest, isolation and preparation of mouse mammary component cell types.

Following in utero exposure, litters were culled for females only, weaned, and grown up to 12-14 weeks of age when harvested for mammary glands, taking care to exclude the lymph nodes. Mechanical and enzymatic isolation of the component mouse mammary cell types for each in utero treatment, oil and BPA, were achieved through the use of Smalley 2010 protocol [203]. Briefly, between 12-20 female CD-1 mice, 12-14 weeks old, were sacrificed and the 4th and 5th inguinal mammary glands excised and mechanically minced to produce a fine, semi-liquid slurry. This slurry was further

digested in a collagenase/ trypsin mix on a platform shaker at 37°C. The centrifuged

(350x g) pellet contains epithelial organoids, stromal and red blood cells and fibroblasts.

The pellet is processed through a series of digestive and dissociative steps, including a plating of cells onto a 150mm cell culture dish to separate epithelial organoids from fibroblasts. The resulting epithelial cell suspensions for each treatment were stained and separated by flow cytometric sorting (FACSAria II cell sorter, BD Sciences). IgG controls were used for all fluorescent antibodies (FITC and PE-conjugated, BD Sciences) as described [203]. The number of cells in the test sample for each treatment was adjusted to 106/ mL. First, DAPI was used to remove dead cells, Cd45 to remove leukocytes and finally Cd24 and Cd49f to separate the component epithelial cell populations: Cd24High/

Cd49fLow (luminal epithelial) and Cd24Low/ Cd49fHigh (basal epithelial cell/ stem cell).

Component mammary epithelial cell populations were sorted into 2% FBS and 1% BSA- coated tubes to prevent sample loss. Plated fibroblasts were maintained at 37°C/ 5% CO2/

5% O2 and grown to 80% confluency. All three cell types were washed in 1x phosphate buffered saline (PBS) and DNA/ RNA harvested using the ZR-Duet DNA/RNA mini prep kit (Zymo Research).

4.3.3 RNA-sequencing, validation, and analysis.

Libraries were prepared from cell type-specific RNA and next-generation sequencing was performed utilizing Illumina HiSeq 2500 (The Ohio State University,

Genomics Shared Resource). RNA-seq was performed in duplicate, from two separately treated and isolated biological replicates. Sequencing reads were deduplicated using

FastUniq [204]

(http://sourceforge.net/projects/fastuniq/) and aligned to the mouse MM10 genome with

TopHat splice read-mapper [205]. Transcriptional gene counts were generated with

Cufflinks [206, 207] and then normalized accounting for ‘batch effects’ between replicates in Combat [208] (http://biosun1.harvard.edu/complab/batch/). Significant differential gene expression was determined using NoiSeq [209]

(https://www.bioconductor.org/packages/release/bioc/html/NOISeq.html). ToppGene

Suite [210] (http://toppgene.cchmc.org) was utilized for gene list functional enrichment and pathway analysis. Gene lists for component mammary cell types were identified, meeting the following parameters: p-value ≤ 0.05 and a fold change ≥ 0.4 or ≤ -0.04 in gene expression between oil- and BPA in utero-treatments.

4.3.4 Tissue preparation and analyses of mammary gland ECM components following in utero exposure to BPA.

To evaluate ECM component content of adult mammary tissue following in utero

BPA exposure, female mice were treated and culled as described (Figure 4.1A); the fourth inguinal mammary gland on each side of the animal at 14 weeks of age (adult/ later-life, n ≥ 11) was harvested for each treatment. All animals were staged in the estrous cycle prior to harvest as determined by vaginal cytology [162-165] and glands were harvested during the stage of estrus. Stage of estrus was confirmed through Hematoxylin

& Eosin (H&E) analysis of dissected uterus, ovaries and vagina from each animal [163,

165]. For each mouse, one gland was whole-mounted and carmine-aluminum stained,

(using the National Toxicology Program Animal Studies Protocol; section XIII, appendix

6, section E) [166] for potential morphological analyses and confirmation of in utero

BPA phenotype; the other gland was fixed in 10% neutral buffered formalin for 24 hours, transferred to 70% ethanol, paraffin embedded and cut consecutively, in 4μM or 12μM sections for histological analyses of the ECM major macromolecular component, collagen (Picro-sirius red). For staining methods, slides were baked, deparaffinized and hydrated to water. A Picro-sirius red (PSR) staining kit was used to image collagen and connective tissue content (in red) (ab150681, Abcam).

Analyses for all histological staining was performed by taking two, field-of-view images per biological replicate (n ≥ 10), one selected from the leading edge of the mammary gland (in front of the lymph node), and one on the opposite side of the lymph node (LN) (Figure 4.3A-C). Images were acquired at 4x magnification using a light microscope (Nikon Eclipse 50i microscope), equipped with a camera (Axiocam506 color,

Zeiss) and Zeiss Zen Pro software. Images were randomized prior to analysis in ImageJ

[169], in order to blind the histological analysis and quantitation of ECM component content. The area of the LN was removed from the image prior to thresholding. PSR was simply thresholded, in color, to quantify the total red stained area within the tissue. The intensity of this red-stained area was also quantified in ImageJ.

4.3.5 Hydraulic permeability measurements performed with microfluidic dye.

Measurement of hydraulic permeability as a result of treated fibroblasts reorganization of experimentally available collagen I was performed as previously described [211]. Briefly, soft lithography methodologies were employed to fabricate straight channel polydimethylsiloxane (PDMS) microfluidic devices (L: 5mm, W:

500µM, T: 1mm) [212]. Neutralized (NaOH) acidic rat tail type I collagen (Corning)

100

solutions were prepared at 6mg/mL in media for the following separate experimental conditions: (1) containing oil vehicle-treated fibroblasts, (2) containing in utero BPA- treated fibroblasts (1800 cells/µL) and (3) no added fibroblasts for acellular conditions.

PDMS microfluidic devices were pre-coated with 100µg/mL fibronectin prior to collagen mixture injection, as previously determined to be optimum [213, 214]. Polymerization of the collagen mixtures, with and without fibroblasts, was permitted prior to the addition of fresh media. Fresh media was replaced daily over 72 hours to permit the re-organization of the provided collagen by the treated fibroblasts. At this time, a height-based (1.8-

2.2cm) hydrostatic pressure difference was applied between channel ports of the PDMS microfluidic device, promoting the flow of a rhodamine fluorescent fluid (used to track the flow rate) through the to the semi-porous collagen matrix. Hydraulic permeability was calculated using Darcy’s law (Equation 1) [215]:

Where µ is fluid viscosity of the cell culture medium (approximated with water),

ν is the average fluid velocity, ΔL is the length of the PDMS microfluidic device channel and ΔP is the pressure difference across the channel due to differences in the fluid height of the two ports on either side of the device, given by (Equation 2):

Where ρ is the density of the flow medium, ɡ is the acceleration due to gravity

(9.81m/s2) and հ is the fluidic height difference between device ports.

101

4.4 Results

4.4.1 Component mammary cell types are effectively isolated following in utero BPA exposure.

Mammary glands of 12-14 weeks old female mice, exposed in utero to BPA

(25µg/ kg • bw/ day) by IP injection E9.5 - 18.5 (Figure 4.1A, treatment schematic), were harvested and component cell types isolated by fluorescence activated cell sorting

(FACS) [203]. Representative images depict morphological defects of carmine aluminum-stained, 14 weeks old mammary gland whole mounts treated in utero with sesame oil vehicle control or BPA (in sesame oil) (Figure 4.1A, images). Insets depict greater detail of differential epithelial branching and volume at this adult time point of mammary gland development. These defects have been characterized more thoroughly as previously described (see Chapter 3, Figure 3.1), following our in utero BPA treatment model.

The component mammary cell types that were isolated include fibroblasts, luminal epithelial and basal epithelial cells and are depicted in Figure 4.1B-D. Fibroblasts were first separated by plating out on a culture dish and thus the flow cytometric sorting depicted by the two color fluorescence plot only represents the separation of the luminal and basal epithelial cell populations (Figure 4.1B). The separation between the two epithelial cell populations was discrete. The luminal epithelial cell population segregated by Cd24High/ Cd49fLow and the basal epithelial cell population, which contains approximately 10-20% mammary stem cells (MaSCs), segregated by Cd24Low/ Cd49fHigh

(Figure 4.1B). Overall, we found no difference in the population sizes of component

102

epithelial cell types isolated and ultimately recovered and counted by FACS (Figure

4.1C).

RNA-sequencing (RNA-seq) was performed on these 14 weeks old mouse mammary component cell types following in utero exposure to BPA (n = 2). The heatmap depicts signature genes [86, 216] that are representative of the cell types being isolated, luminal epithelial (Oil- and BPA-L), basal epithelial/ MaSC (-B) and fibroblasts

(-F) (Figure 4.1D and Table 4.1). We observed that the purity of our cell types was independent of the in utero treatment received. For example, the first panels of the heatmap (left) depicting the luminal epithelial cell type of in utero oil and BPA highly express the luminal epithelial gene signature (in blue) compared to the other cell types.

This finding provided further confirmation that there was limited cross-contamination between our three cell populations of interest.

103

Table 4.1 Gene signatures representative of purified adult mouse mammary gland cell types used to validate isolation (Figure 4.1D). [217, 218]

4.4.2 Limited overlap of differentially regulated genes following in utero BPA exposure between mammary component cell types.

RNA-seq analysis was used to identify significantly deregulated genes for each mammary cell population (Table 4.2, 3, 4) isolated from 14 weeks old mice following in utero BPA exposure compared to oil vehicle (Figure 4.2A-C, p ≤ 0.05 and absolute fold change ≥ 0.4). Again, these heatmaps are representative of two separately isolated biological replicates, indicated by -1 and -2 for each treatment. Much fewer genes were significantly deregulated in the luminal epithelial cell population (36 genes) compared to basal epithelial/ MaSCs and fibroblast populations (160 and 134 genes, respectively).

Interestingly, 139 genes of the total 160 significantly differentially expressed genes 104

MaSC population (almost 90%, indicated by blue), compared to vehicle oil-treated

(Figure 4.2B). Conversely, significant and differentially regulated genes of the luminal epithelial and fibroblast cell populations were more evenly distributed between those genes having higher versus lower expression levels compared to control; both cell populations having approximately 40% of significantly deregulated genes with lower expression than control (Figure 4.2A, C).

Table 4.2 Luminal epithelial cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands (Figure 4.2A, fold change ≥ 40%, p ≤ 0.05;

105

36 total genes represented in heatmap). Gene symbol, fold change and p-value as indicated.

106

Table 4.3 Basal epithelial cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands (Figure 4.2B, fold change ≥ 40%, p ≤ 0.05; 160 genes total represented in the heatmap). Gene symbol, fold change and p-value as indicated.

Continued. 107

Table 4.3 continued.

108

Table 4.4 Fibroblast cell significant gene list identified through RNA-seq analysis isolated from adult mammary glands (Figure 4.2C, fold change ≥ 40%, p ≤ 0.05; 135 genes total represented in the heatmap). Gene symbol, fold change and p-value as indicated.

Continued. 109

Table 4.4 continued.

110

Overlap of significant (p ≤0.05) and differential (absolute fold change ≥ 0.4) gene expression between cell populations was limited (Figure 4.2D). The limited number of genes that did overlap across all cell populations was mostly related to ribosomal binding and processes as well as ECM components such as collagens. The basal epithelial/ MaSC population had hepatocyte growth factor receptor/ Met proto-oncogene (Met) and translationally controlled tumor protein-1 (Tpt1) in common with fibroblasts. Met has been extensively implicated in breast development [219-222] and has been more recently shown to be dependent on ERα expression in the perinatal mammary gland [86]. Tpt1 has been shown to be involved in malignant cancer progression [223] and indicative of poor prognosis in breast cancer [224]. Both the luminal epithelial and fibroblast cell populations had several genes related to ECM components, including collagen type I

(Col1a1) that was shared between both. Further, the top 5 most significantly deregulated genes in the fibroblast population consisted of all collagen type genes and all of these were more highly expressed as a result of in utero BPA exposure.

4.4.3 Pathway analysis reveals ECM organization and structure significantly deregulated following in utero BPA exposure.

Pathway analysis conducted on the significantly deregulated gene lists of each component mammary cell type overall revealed extracellular matrix (ECM) components, organization and structure as a result of in utero BPA exposure (Table 4.5). While the basal epithelial/ MaSC population had limited deregulated pathways related to ECM

(Table 4.5, Cellular components: focal adhesion and gap junctions), most deregulated pathways were related to ribosomal binding, processing and translation. This major

111

finding in the basal epithelial/ MaSC population, and more subtly in the other two cell populations, is interesting considering approximately 30% of the total protein mass of a multicellular animal consists of collagens and other component macromolecules needed to structure the ECM [225]. Despite the limited genes found to be significantly deregulated in the luminal epithelial cell population (36 genes total; versus 160 and 134 genes in basal/ MaSC and fibroblasts, respectively), there were many molecular functions and biological processes in common with the fibroblast cell population (Table 4.5). These included platlet-derived growth factor (PDGF) and generally growth factor binding and signaling, collagen binding and organization, and mammary neoplasms (ToppGene,

Disease: luminal epithelia-specific, p = 2.03E-04; ToppGene, Disease: fibroblast-specific, p = 1.04E-03) and epithelial morphogenesis (ToppGene, Disease: fibroblast-specific).

4.4.4 In vitro collagen organization altered by in utero BPA exposed mammary fibroblasts.

The coordinated deregulation of ECM organization by each cell type led us to first examine some of the ECM macromolecule content within the adult mammary tissue through histological methods. Utilizing 14 weeks of age mammary gland tissue, treated in utero with BPA or oil vehicle control, histological staining for collagen (PSR) was performed (Figure 4.3). Consistent with our RNA-seq transcriptional findings, we found that PSR staining exhibited a significant increase in collagen area (red-staining) as well as the intensity of that red stain throughout the adult mammary gland tissue as visualized by PSR (red, Figure 4.3 A).

112

4.4.5 ECM remodeling revealed through decreased hydraulic permeability as a result of in utero BPA exposure.

Transcriptional changes indicating ECM deregulation, particularly the observation in the fibroblast cell population that many collagen genes were more highly expressed as a result of in utero BPA, led us to examine a more functional way of assessing alterations to the ECM. Hydraulic permeability relates to the interstitial flow of water and solutes through a biological medium, which are critical processes related to physiology, inflammation and cancer progression [215]. The ECM architecture is comprised of macromolecules including collagen fibers and GAGs [215, 225, 226], through which this interstitial fluids flows.

To determine the effect of in utero BPA treatment upon hydraulic permeability, we utilized a single channel microfluidic assay to assess the remodeling capacity of treated fibroblasts. The channel was injected with a 3-D collagen matrix which was seeded with control and in utero BPA-treated fibroblasts isolated as described (Figure 1) from adult mammary glands; fibroblasts were permitted to remodel the matrix over 72 hours. A microfluidic tracer dye was used to determine the rate of fluid flow through this semi-porous, remodeled matrix (Figure 4.4A). Velocity measurements were determined to be significantly decreased as a result of in utero BPA-treated fibroblast remodeling compared to oil control (Figure 4.4B, C). The reduction in permeability as a result of in utero BPA exposure was 2.5-fold compared to oil and 1.5-fold compared to the acellular control. These results indicate that in utero BPA exposure promotes tissue-specific remodeling mediated through the fibroblasts, changing the permeability. Following that

113

the same amount of collagen was applied to each channel suggests that the remodeling we infer from the transcriptional changes can related to both collagen deposition and arrangement of fibers.

4.5 Discussion

Our work herein sought to impart a mechanism of in utero BPA exposure by which discrete transcriptional changes to the component cell types present in the adult mammary gland may play a role in propagating the consequences of exposure to account for well-described phenotypes and later-life susceptibility. Our isolation method resulted in distinct mammary cell populations, including fibroblasts, luminal epithelial cells and basal epithelial/ MaSCs. Further, the component cell type transcriptional deregulation as a result of in utero BPA exposure was distinct, in that the significantly deregulated genes identified, demonstrated little overlap.

4.5.1 Tissue architecture and signaling dictate development.

The concept that stroma directs epithelial development is well characterized and accepted [182]. Principally, when mammary epithelium was recombined with mammary mesenchyme, a normal epithelial tree resulted. Conversely, when mammary epithelium was recombined within a salivary mesenchyme, a salivary-like epithelial tree was generated (as reviewed in [110]) [115, 117]. As demonstrated in the embryonic mammary gland, there are subsets of the mammary mesenchyme which have different potentials to influence epithelial development. Recall in mouse embryonic mammary development that there is a dense ERα-expressing mesenchyme which completely surrounds the

114

developing epithelial bud, versus a more distal ‘secondary’ mesenchyme making up the precursor fat pad. The former was shown to induce hyperplastic growth in the epithelium of adult mice when grafted into their fat pads [116]. Further, up-regulation genes involved in the specific mediation of epithelial development were found to be exclusive to the subset of the surrounding mesenchyme as opposed to the more distal, non-specific pre-cursor fat pad mesenchyme [118, 227]. In fact, the ability of this fat pad precursor mesenchyme to direct epithelial development has been shown to be non-specific and unable to direct morphogenesis. Further, the dynamics of the adult epithelium (regardless of glandular origin) are retained as well, when recombined and under the influence of embryonic mammary mesenchyme [118].

A major component of the stromal compartment is that of the ECM which not only serves as a physical and mechanical support for tissue organization but is also involved in many biochemical and biophysical signaling pathways to regulate cell behavior [110, 228]. Deregulation in key pathways such as ECM organization and protein production among component cell types suggest that BPA may predispose the developing mammary gland via altered ECM signaling and organization incurred at time of exposure. We propose this as a mechanism by which BPA orchestrates gland development following in utero exposure and propagates later-life risk for cancer pathologies.

115

4.5.2 ECM remodeling directs postnatal mammary gland development and sets the stage for tumorigenesis.

Mammary gland dynamic and functional development occurs post-natally as orchestrated by hormonal and paracrine signaling. The ECM has been shown to be critical in these processes of postnatal development, closely associated with normal epithelial morphogenesis [38, 229] and an active participant in tumor progression [230-

232]. Further, it has been shown that disruption of the remodeling and structural activities of the surrounding mesenchyme, mainly through the ECM, correlates to epithelial breakdown, epithelial-mesenchymal transition (EMT) and tumorigenesis [119, 120].

Our cell type-specific approach identified many genes within our fibroblast cell population involved with ECM remodeling, including TGFβ, Wnt, β-catenin and growth factor signaling (as reviewed in [110]) [219] to affect epithelial development. These are direct signaling pathways that impact EMC remodeling as it may relate to BPA in utero exposure that may be missed in whole tissue compartment analyses. Additionally, fibroblasts have been shown to be involved in much of the synthesis of ECM components and enzymatic mediators of tissue organization and as it relates to cancer development

[233, 234]. Thus BPA may alternatively mediate this role of the fibroblast cell population at time of exposure.

116

4.5.3 Disruption of ECM homeostasis can be a pre-requisite or consequence of tumorigenesis, imparting a mechanism for BPA propagation of later-life disease susceptibility.

We found that a major ECM macromolecular component, collagen, was significantly increased in our adult mammary tissue as a result of in utero BPA exposure.

This finding corroborated our transcriptional and pathway findings that collagens and

ECM components were significantly increased or deregulated following in utero BPA exposure. Specifically, the fibroblast significant gene list revealed that the top 5 significantly deregulated genes were all collagens, all showing an increase compared to control. Indeed, fibroblasts are responsible for the bulk of transcription and secretion of interstitial collagen [235]. It has been previously shown that upon examination of perinatal glands (E18.5) evaluated for area of connective tissue (Masson’s trichrome)

[85], a significant increase in collagen localization was observed immediately surrounding the epithelial compartment at this perinatal time point, consistent with previously described functions of the more immediate mammary mesenchyme surrounding the epithelium and the precursor fat pad mesenchyme. This depicts the dynamics of ECM remodeling in the developing mammary gland and how it can evolve from early- to later-life. We postulate our work can clarify the mechanism of propagated exposure to later-life through the characterization of changes in the adult mammary gland cell types since it is in the adult that in utero exposures to these EDCs have been shown to manifest risk for cancer development.

117

Our finding that at an adult time point of mammary development, more collagen resulting from in utero BPA exposures could mean an alteration to ECM remodeling and increased deposition of collagen and other major ECM macromolecules. Indeed, disruption of collagen homeostasis has been demonstrated to be a prerequisite for tumor invasion and progression [236] and the observed increase in collagen deposition, demonstrated by PSR staining, supports a change in collagen content thereby disrupting the normal architecture and permeability of the mammary tissue [237, 238]. This could be consistent with rodent studies that demonstrated increased susceptibility to later-life disease through in utero BPA exposures [239]. While in utero exposure is usually not enough, additional exposures to carcinogens or changes in our environment may elicit the consequence of increased risk.

4.5.4 Reduced hydraulic permeability indicates a more dense tissue composition.

Our finding that in utero BPA-treated, adult mammary fibroblasts remodel an in vitro collagen matrix to be less permeable to fluid flow indicates a promotion of ECM density and stiffening. This confers the ability of BPA to contribute to one of the greatest risk factors for breast cancer – breast density. Mammographic breast density has demonstrated a 4-fold increased incidence of developing breast cancer [111, 112, 240,

241]. Further, matrix and stromal stiffening have been associated with the promotion of tumor initiation, progression and invasion in breast cancer [242, 243].

Evident in our analysis, the amount of collagen deposition is a likely contributor to the mechanism of in utero BPA exposure. In our in vitro hydraulic permeability assay, the same amount of collagen was provided in each microfluidic flow channel. Thus, it is

118

also possible that the arrangement and organization of collagen fibers may contribute to differential permeability. Thus, it is necessary to further demonstrate the functionality of this phenotype to impart increased stiffness of the ECM matrix, in order to faithfully impart this mechanism to BPA in utero exposure.

4.5.5 Additional work to implicate BPA in mammary gland remodeling to promote later- life susceptibility.

In order to fully characterize the mechanism by which in utero BPA exposure may influence breast stiffening that can confer a mechanism of later-life susceptibility, more must be done to characterize the arrangement and possibly the enzymatic mediation of remodeling. We could evaluate protease-dependent degradation of collagen in order to account for the decreased collagen we observed. Typically this enzymatic degradation is catalyzed by lysyl oxidases (LOX) [244]. These enzymes typically function in concert with metalloproteinases (MMPs). Thus, evaluation of these mediators of ECM remodeling could the take the form of the following: (1) identifying them within our gene lists, significant or not, and determining whether our RNA-seq analyses missed them through application of cut-off parameters; (2) measure relative expression through RT-

PCR analysis within our component cell types; or (3) manipulate their contribution to matrix remodeling through the use of shRNAs or viral knock-down of these genes in our mammary fibroblasts within a 3-D in vitro or microfluidic device model. Given that ECM remodeling is a prerequisite for tumor initiation, progression and invasion, markers for epithelial-mesenchymal transition (EMT) could also be evaluated. While, it would be best to begin with the RNA-seq gene lists of the mammary component cell types, other

119

EMT candidates could include E-cadherin, Snail, β-catenin, Wnt and TGFβ. The latter three had been identified as deregulated pathways within our RNA-seq analysis, particularly between the luminal epithelial cell population and fibroblasts.

The increased amount of collagen present in our adult, in utero BPA-treated mammary tissue, taken together with decreased hydraulic permeability implicates changes to the collagen network organization, through deposition specifically. However, other interesting parameters of ECM organization could be further demonstrated in fiber thickness, angles and length of collagen and other ECM macromolecular components.

Such parameters can be evaluated through the use of polarized light microscopy. This can be performed on the already stained PSR mammary tissue. The PSR stain has birefringent properties when polarized light is applied that can provide information about the type of collagen fibers present in the tissue and their thickness. To obtain additional information about the collagen arrangement and properties, second harmonic generation (SHG) microscopy can be utilized [245].SHG utilizes the wavelength of light, ~950nm, to specifically image collagen I fibers, which is the predominant collagen type present in tissues. This imaging can provide information about collagen I fiber length, width and angle [246].

4.6 Acknowledgements

We would like to thank the Michael C. Ostrowski lab, specifically Gina M.

Sizemore, Subhasree Balakrishnan and Anisha M. Hammer for their help in refining our mouse mammary cell type harvest and isolation technique and our exploration of

120

extracellular matrix methods to corroborate our RNA-seq findings, respectively. The project described was supported by Award Number Grant UL1TR001070 from the

National Center for Advancing Translational Sciences. The content is solely the responsibility of the authors and does not necessarily represent the official views of the

National Center for Advancing Translational Sciences or the National Institutes of

Health. We would like to acknowledge the following Core Shared Resource facilities at

The Ohio State University: the Analytical Cytometry and Cell Sorting (ACCSSR; Bryan

McElwain and Katrina Miller), Genomics (GSR; Pearlly Yan, Ph.D.), Biostatistics (BSR) and Microscopy (MSR; Sara Cole, Ph.D. and Brian Kemmenoe).

121

Figure 4.1 – Discrete isolation of adult mammary component cell types following in utero BPA exposure. A. Schematic timeline of in utero BPA treatment used to investigate cell type-specific changes to component mammary gland cell types.

Representative carmine aluminum-stained 14 weeks old mammary glands to show altered morphogenesis of the epithelium.

Continued.

122

Figure 4.1 continued.

B. Fluorescence activated cell sorting (FACs) dot plot of single-cell epithelial populations, isolated from mice treated A, luminal epithelial, and basal epithelial/ MaSC.

*Fibroblasts are isolated prior to sorting. C. Distribution of the epithelial cell populations recovered from isolation and FACs sorting. D. Enrichment of cell type-specific populations was validated by RNA-seq as compared to the average transcript abundance for genes across samples. Expression is as indicated and the distribution of the Z-score.

High expression indicated in blue and low expression in yellow. Top quadrant of genes are specific for luminal epithelial cells (all blue), middle quadrant specific for basal epithelial/ MaSC and bottom quadrant specific for fibroblasts.

123

Figure 4.2 – Distinct transcriptional changes across adult component mammary cell types results in little overlap. A-C. Heatmaps depict significantly deregulated genes within each adult, mammary component cell type, luminal epithelial, basal epithelial/

MaSC and fibroblasts, respectively. Genes with p ≤ 0.,05 and a fold change of 0.4 are represented in the heatmap. D. Vann diagram depicting overlap of genes between the component cell types.

124

Table 4.5 – Pathway analysis of transcriptional changes reveals the extracellular matrix as a highly deregulated pathway among mammary component cell types following in utero BPA exposure. Analysis performed in ToppGene with the gene lists identified in Figure 4.2, 3 and 4. Parentheses preceding pathways indicated represent the order of significance. The final number in parentheses in each category per cell type is indicative of the total number of significantly deregulated pathways and processes listed in ToppGene. Numbers in the brackets indicate the degree of significance of that pathway, the p-value.

125

Figure 4.3 – Significant increase in collagen density within the adult mammary tissue. Component macromolecules of the ECM are histologically evaluated. Images are at 4x magnification. Amount of ECM macromolecule present if quantified using ImageJ thresholding and represented as a percent within each in utero treatment. Picro-sirius red

(PSR) to image collagen and connective tissue (red). * p ≤ 0.05.

126

Figure 4.4 – Significant decrease in hydraulic permeability of an in vitro collagen matrix remodeled by in utero BPA-treated, adult mammary fibroblasts. A.

Microfluidic, single channel device is injected with a solution containing collagen I, treated mammary fibroblasts within cell media (indicated by the dotted box). Pressure is applied to the channel in order to promote fluid flow through the channel, and hydraulic permeability (κ) is calculated. B. Fluorescent dye used to track the progress of flow through the microfluidic flow cell and through the deposited collagen matrix indicated in

A for each condition. Yellow arrows indicate the change in length measured from the initial bulk of dye detected in the left panel, the single dotted line. C. Hydraulic permeability represented in each condition with either oil- or BPA- in utero treated adult mammary fibroblasts, or containing no fibroblasts (acell). * p ≤ 0.05. 127

Chapter 5: Concluding remarks, unpublished data and future directions.

5.1 Summary, impact and innovation

The in utero phenotype of EDC exposure is well-characterized in animal models; however the mechanisms and consequences of this exposure to incidence of disease are less understood. Unlike the deliberate administration of what would become known as a carcinogen and EDC, DES, our constant and chronic exposure to BPA is much less realized or calculated. The pervasive manufacture and appreciable detection of it within many human biological tissues and fluids and its demonstrated ability to interfere with hormone signaling, BPA is suspected to be a likely contribution to incidences of hormone-dependent diseases, such as breast cancer. Given the complexity of hormone- dependent tissues, their signaling and tissue compartment interactions that direct development over a lifetime, the urgency of understanding the mechanisms of EDC exposure is upon us.

Our body of work herein, sought to address some of the mechanistic gaps that remain in the field of endocrine disruption as they relate to BPA, but we anticipate will translate to a better understanding of general EDC modes of action. In Chapter 2, utilizing an in vitro model in human, hormone-dependent breast cancer cell lines, we addressed the lack of mechanistic understanding related to chronic human exposures that occur over a lifetime to these EDCs (BPA and GEN). While many studies before had

128

emphasized the lack of mechanistic knowledge, their approach seemingly fell short as a result of acute and short-term exposures. We felt it was important to model human lifetime exposure, in order to evaluate transcriptional alterations, receptor recruitment and

DNA accessibility, by treating cells long-term, over a year, in culture with these EDCs.

Strikingly, our work demonstrated that not only were estrogen targets altered as a result of chronic exposures, but given time in the absence of exposure, some of these targets demonstrated sustained reprogramming. This work demonstrated the potential of EDCs to propagate effects of chronic exposure beyond a time of cessation, and that changes to signaling can be sustained. Our evaluation of histone modifications, coupled with sustained signaling changes adds further support to the potential of EDC compounds to perpetuate their effects over multiple generations, in this case the passaging of cells, implicating an epigenetic mechanism.

In Chapter 3, we took into consideration the developmental intricacies of tissue compartment interactions and demonstrated differences in compartment signaling at in utero time of exposure in order to narrow a possible mechanism for BPA. Based on the limited analysis of exposure windows for DES administration that associated first trimester exposures to reproductive tract tumors, we postulated that BPA may also exhibit mechanisms of action related to key developmental stages at time of exposure.

Utilizing four treatment schemes, we were able to demonstrate that varying windows of in utero exposures were not equal in their ability to elicit well-described phenotypes, and the phenotype exhibited by our full embryonic BPA exposure. The most susceptible window to in utero BPA exposure appeared to occur at a time when in mammary gland

129

development, the developing epithelial bud is completely surrounded by the primary and immediately adjacent mesenchyme. This primary mammary mesenchyme, known in mice to exhibit the sole expression of ERα, also has up-regulation of critical gene expression necessary for epithelial morphogenesis and function. Our work corroborated the necessity of tissue compartment interactions at this critical time in mammary gland development and provided a mode of action for BPA and its potential to interfere with these vital processes.

Finally in Chapter 4, in acknowledgement that the component cell types add an additional layer of complexity to the potential mechanism of BPA disruption of mammary gland development and disease, we optimized a component cell type isolation protocol [203]. While we know from developmental studies that the interactions between the two tissue compartments are critical for epithelial morphogenesis and function, we postulated that examining discrete alterations to component cell types may better inform the mechanism of BPA disruption and elaborate on molecular changes that could potentially be missed examining whole tissue compartment. Further, to this end, we sought also to examine the adult time point as a way of deciphering the molecular changes present in the mammary gland at a time when the consequence of in utero exposure would manifest. This reasoning is derived from the demonstrated latency of

DES exposure in ’DES daughters’, and the evidence in numerous animal models, of increased susceptibility of the adult gland to tumorigenesis, following in utero and perinatal exposures and provoked with exposure to carcinogen. Our finding that transcriptional alterations to the component adult mammary cell types converged on

130

deregulated ECM signaling and organization supports a mechanism by which BPA alters early-life tissue compartment interactions through altered remodeling at time of in utero exposure and that could be sufficient to propagate susceptibility to later-life. Our most significant finding was the functional demonstration of this potential remodeling, mediated by treated fibroblasts to alter an in vitro collagen matrix to reduce hydraulic permeability. Given the risk associated with breast density and its implications for breast cancer, this provides more evidence that BPA is capable of contributing to breast cancer incidence in the human population.

5.2 Future directions.

5.2.1 Utilizing in vitro methods to demonstrate altered molecular signaling by EDCs.

Many in vitro studies have been valuable in their demonstration of molecular mechanisms of EDC action including receptor affinities and preferences [130, 131, 247-

249], alternative signaling cascades [24, 178] and differential transcriptional regulation

[53, 155, 249]. Further, and in the case of our work (Chapter 2), the EDCs have potential and have been demonstrated to drive epigenetic reprogramming. In these ways, the actions of EDCs are dependent on their interaction with the receptors responsible for mediating endogenous hormone action. This is logical not only based on the immense weight-of-evidence of EDCs but also based on what is known about endogenous hormone activity. These include their potencies at discretely regulated, low-dose concentrations within the body, and the biological context of action dependent on tissue- specific criteria including, the number of receptors present, binding affinity and the

131

availability of CoRs [250]. For these same reasons their activities can be unpredictable

[247].

One specific proposed mechanism by which EDCs elicit alternative programs is through recruitment of an array of CoRs upon receptor binding [25, 26, 251] and these

CoRs have been shown to be both specific to the target tissue and implicated in pathologies and have activating and repressive activities [252]. For example, the coactivator proteins SRC-3 and SRC-1 (SRC proto-oncogene, non-receptor tyrosine kinase) are frequently over-expressed and/ or amplified in breast cancer and have also been implicated in recurrent disease (as reviewed in [253]) [254]. In general, more than 400 of these CoR proteins have been identified which is representative of the diversity in responses that could be elicited upon EDC binding of a hormone receptor [255]. We know ER is promiscuous in that it can bind an array of compounds, due to its generous ligand-binding pocket. Ligand binding promotes a conformational change within the receptor that dictates which CoRs are recruited to affect downstream signaling and biological processes. This is specifically facilitated by a highly conserved, helix 12 located in the activation function-2 (AF-2) domain of the receptor [256, 257]. Depending on the orientation of this helix, determines which CoRs will bind. Further, the recruitment of CoRs has consequences for endocrine-resistant breast cancer as well [258]. These possibilities of differential CoR recruitment could be examined through utilization of stable isotope labeling of amino acids in cell culture (SILAC) followed by mass spectrometry (MS) [259, 260]. In human breast cancer cell lines, separate lines treated with a DMSO-control, a biologically relevant dose of BPA and a relevant dose of

132

estrogen as a positive control could be compared in vitro. In the presence of isotope under each treatment condition, CoR proteins recruited due to activating ligand (DMSO, BPA or estrogen) can be distinguished and detected by MS.

5.2.2 Tissue-specific signaling in development and the requirement of ERα in the mesenchyme to elicit later-life susceptibility.

In identifying a more discrete window of susceptibility, possibly best accounting for epithelial morphologies that contribute to risk, the scope of developmental paracrine signaling altered by in utero BPA exposure can be narrowed. Since many of the molecular mediators of this developmental stage, and specific to each tissue compartment, have been identified [102, 175] our work could be taken a step further by manipulating them. This could be accomplished through knock-down of these mediators at this discrete time of in utero exposure and then assessing the later-life susceptibility to carcinogenic exposure.

While we correlated our findings to ERα expression in the embryonic mesenchyme, we did not demonstrate its requirement for the phenotype as a result of in utero BPA exposure. Recent work has demonstrated this functional difference utilizing an ER antagonist [261]. They were able to show that while BPA-mediated effects upon epithelial morphology were effectively blocked by the antagonist, estrogen was not [181].

An interpretation of this would be that BPA is acting through stromal ER (see Chapter 3,

Figure 3.4.1) [102, 172] while estrogen can endogenously act through alternative pathways. Finally, our lab has developed a mouse model that has specific mesenchymal knock-out of ERα as initiated through the tenascin C (Tnc) mesenchymal-specific

133

promoter [200]. Thus, utilizing this model we can further demonstrate the requirement of

ERα in the stroma at time of in utero BPA exposure.

5.2.3 Epigenetic mechanism of BPA.

Sustained reprogramming of the estrogen response as a result of chronic BPA exposure demonstrated the potential of BPA to initiate epigenetic alterations (Chapter

2). Further, animal models have shown demonstrated epigenetic programs elicited by

BPA [84, 87, 262]. Specifically, one study utilized hormone-responsive breast cancer cell lines and in utero- treated mice to examine the effect of both DES and BPA on the polycomb repressive group 2 (PRC2) catalytic subunit Enhancer of zeste homolog 2

(Ezh2) expression and function [87]. They found that as a result of exposure, both DES and BPA increase the expression of Ezh2 as well as its repressive mark, histone 3 lysine

27 trimethylation (H3K27me3). Interestingly, DNA methylation status and Ezh2 have been implicated in breast cancer risk and prognosis [263-265]. Taken together, our ongoing work has been to utilize our cell type-specific approach to investigate transcriptional (Chapter 3) and epigenetic alterations (Chapter 4) following in utero

BPA exposure in the adult mammary gland. This is in contrast to the studies mentioned herein, as they examined more pubertal/ early-life time point of mammary gland development compared to our intended examination of the adult gland (12-14 weeks of age).

Preliminarily, through whole genome bisulfate-sequencing of component mammary cell types and exon enrichment (Agilent SureSelect) there was significantly more differentially methylated regions (DMRs) identified in our luminal epithelial cell

134

population following in utero BPA treatment compared to both the basal epithelial/

MaSC and fibroblast populations (Figure 5.1). In addition, while the latter populations demonstrated even distributions of DMRs between hyper- and hypo-methylation, the luminal population was almost entirely hypomethylated. Given that cancer occurs in epithelium, this finding of hypomethylation in the luminal epithelial cell population following in utero BPA exposure, specifically the population associated with hormone receptor expression, is significant. Indeed, genomic hypomethylation has been linked to metastatic disease compared to primary tumors [266]. This particularly has implications for the incidences of recurrence. In addition, DNA hypomethylation has been significantly correlated to breast carcinoma [267] and rendering prognostic value to this state of methylation.

Upon further examination of hypomethylation of enriched genomic regions within our adult, in utero treated mammary cell populations we observed an expansion of DMRs in promoter regions within our luminal epithelial cell population compared to fibroblasts

(see purple, Figure 5.2A). Thus, we performed motif analysis on luminal epithelial DMRs within promoters to identify transcription factors recruited to these hypomethylated promoters and pathway analysis revealed that they converged on Wnt signaling.

Interestingly, our pathway analysis performed on our RNA-seq of adult, in utero BPA- treated component cell populations (Chapter 3, Table 4.4.1) also identified Wnt signaling to be significantly deregulated in our fibroblast cell populations. Wnt signaling is a critical regulator of embryonic mammary development as well [175, 268]. However, despite this common pathway between our transcriptional and epigenetic analyses, we

135

found little overlap between hypomethylated luminal promoters and genes found to be transcriptionally up-regulated 2-fold (Figure 5.2B). However, the lack of overlap may be due to the low coverage achieved in our exon-enriched methylation sequencing and improvements to this are in progress.

Finally, we have optimized and are now implementing the use of an ultra-low input native chromatin immunoprecipitation (ULI-N-ChIP) method [269] in order to investigate cell type-specific enrichment of specific histone modifications in the adult mammary gland following in utero BPA exposure. Since our cell type isolation methods only yield a limited number of cells to work with, we have optimized this protocol utilizing only 1 million cells per ChIP. Given the evidence for increased H3K27me3 following in utero exposures to both DES and BPA [87] we started with this histone modification in our cell type-specific assessment. While we are currently in the analysis phase of our first biological replicate we have validated the generated libraries, following

H3K27me3-ChIP, of our component epithelial cell types. By real-time PCR we analyzed targets known to have H3K27me3 in luminal (Foxa1) and basal (Wnt7b) epithelial cell populations (Figure 5.2C). We have begun to examine the epithelial cell populations following enrichment with H3K27me3 and have preliminarily found that as of result of in utero BPA exposure, the adult luminal epithelial cell population is vastly hypermethylated compared to the basal epithelial cell population which has a similar distribution between the two states (hyper- vs. hypo-methylation) (Figure 5.3A). This distribution of H3K27me3 is further demonstrated by the overlap between the two epithelial cell populations with many more genes hypermethylated in the luminal

136

compared to basal (Figure 5.3B). Finally, we show that our method has preliminarily yielded treatment specific results and that both epithelial cell populations have so far been observed to exhibit similar enrichment following exposure (Figure 5.3). Future directions include evaluating the H3K27me3 status of the fibroblast cell population as well as performing pathway analyses on regions differentially enriched for the transcriptional mark of repression.

We intend to evaluate additional marks of methylation, including one implicated in a recent study performed on both fetal and pubertal rats following in utero BPA exposure [84]. As a result of in utero BPA exposure, they found an increase in the activational histone 3 lysine 4 trimethylation (H3K4me3). Though this was limited to a single promoter, we believe the significance of methylation status can be resolved by examining cell types individually, that would otherwise be missed as whole tissue or glands. Again, we are interested in the adult rendering of methylation status to impart an epigenetic mechanism for BPA in utero exposure.

With our implementation of adult mammary component cell type analyses following in utero BPA exposure, we hope to resolve discrete transcriptional and epigenetic mechanisms. Through which, we intend to identify potential biomarkers of

EDC exposure as well as molecular mediators that are indicative or contribute to risk for breast cancer. While our work is specific to BPA, we believe the molecular mechanisms can be extrapolated to other EDC compounds prevalent in our environment.

137

5.3 Concluding remarks.

During a time where consumer opinion and outrage drive business models and decisions, it is no wonder BPA has been voluntarily phased out of use in the manufacture of certain products, most notably baby bottles in 2008 [270, 271]. This inclined the U.S. Food and

Drug Administration (USFDA) to amend its regulations for use of BPA in these related products as of 2013.5 However, what the consumer does not know/ cannot know will not hurt them? Despite labels declaring products to be ‘BPA free,’ many of them alternatively contain similarly formulated compounds that have been demonstrated to activate and interfere with the same or worse consequences to physiological processes and pathologies attributed to known EDCs (as reviewed in [272]) [155]. Thus, our work herein and the work of others are critical in order to emphasize the detriment to human health and disease these unregulated and hormonally active compounds in our environments pose.

An additional obstacle facing scientists and especially that of regulatory agencies assessing human risk to EDC exposures is the problem of mixtures. While current work herein and otherwise has done well to characterize biological effects and susceptibility following exposure to single EDC compounds, the fact remains, our environment presents a pervasive mixture of suspected EDCs that affect human health and disease

[273].

______

5Though this action taken by the USFDA was mitigated in the reasoning it provided, citing provisions set by manufacturers to eliminate BPA from their production (based on consumer demands), rather than concerns over safety. Indeed, FDA maintains its stance on BPA being safe for use in food packaging

138

To this end it is postulated that references doses set for single compounds maygrossly underestimate human risk since they fail to account for other environmentally available compounds likely present [273, 274].

139

Figure 5.1 – All significantly differentially methylated regions (DMRs) in the adult luminal epithelial cell population completely hypomethylated as a result of in utero

BPA exposure. Glands from adult mice, 14 weeks old, treated in utero with BPA or oil were harvested and cell types isolated as described (see Chapter 4, Figure 4.1A). DNA was isolated (Zymo Research DuetTM DNA/RNA miniprep kit), libraries made and bisulfite converted (Agilent SureSelect Mouse Methyl-Seq with exon enrichment).

Distribution of differentially methylated regions (DMRs: hypo- and hyper-methylated) by indicated cell type.

140

Figure 5.2 – Differentially methylated regions (DMRs) of adult mammary gland component cell types suggest epigenetic reprogramming as a result of in utero BPA exposure. Glands from adult mice, 14 weeks old, treated in utero with BPA or oil were harvested and cell types isolated as described (see Chapter 4, Figure 4.4.1A). DNA was isolated (Zymo Research DuetTM DNA/RNA Miniprep kit). A. Libraries made and bisulfite converted (Agilent SureSelect Mouse Methyl-Seq with exon enrichment).

Continued.

141

Figure 5.2 continued.

Distribution of differentially methylated regions (DMRs) found to be hypomethylated as a result of in utero BPA exposure in the fibroblast and luminal epithelial cell populations.

Motif analysis, identifying transcription factors enriched in these DMRs performed on the expanded (compare purple) promoter elements found to be hypomethylated in the luminal epithelial cell population revealed Wnt signaling as a key deregulated pathway.

B. Overlap of hypomethylated DMRs with those genes found to be 2-fold up-regulated in our transcriptional analyses of the luminal epithelial cell population. C. Ultra-low input chromatin immunoprecipitation (ULI-N-ChIP) of H3K27me3 was performed on enriched

(by FACs, see Chapter 4, Figure 4.1B) luminal and basal epithelial/ MaSC populations

(with non-specific IgG). Immunoprecipitated DNA was analyzed by real time PCR for targets known to have H3K27me3 in luminal (Foxa1) and basal (Wnt7b) epithelial populations. RPS 25 served as a negative control.

142

Figure 5.3 – Differentially H3K27trimethylated regions (DMRs) of adult mammary gland component cell types suggest epigenetic reprogramming as a result of in utero

BPA exposure. Glands from adult mice, 14 weeks old, treated in utero with BPA or oil were harvested and cell types isolated as described (see Chapter 4, Figure 4.1A). Adult mammary gland cell types, luminal epithelial, basal epithelial and fibroblast were washed in PBS and flash frozen in 1 million cell aliquots as determined by FACs for the epithelial cell populations and hemocytomer for the fibroblast cell population. Ultra low input, native chromatin immunoprecipitation (ULI-N-ChIP) [269] assays were performed on each cell population with results for the adult mammary gland epithelial cell populations treated in utero preliminarily represented here.

Continued.

143

Figure 5.3 continued.

A. Enrichment of histone 3 lysine 27 trimethylation (H3K27me3; Diagenode antibody) was performed by ULI-N-ChIP and libraries made and sequenced. Distribution of differentially H3K27-trimethylated regions (DMRs), hyper- or hypo-methylated with the repressive chromatin mark in each epithelial cell population as a result of in utero BPA exposure. B. Overlap of DMRs between adult mammary gland epithelial cell populations. C. Example of BPA-treatment specific enrichment of H3K27me3 between epithelial cell types.

144

References

1. Siegel, R.L., K.D. Miller, and A. Jemal, Cancer Statistics, 2017. CA Cancer J Clin, 2017. 67(1): p. 7-30. 2. Institute, N.C. Breast cancer risk in American women. 2012 Sept. 24, 2012 [cited 2017 June 22]; Available from: https://www.cancer.gov/types/breast/risk-fact- sheet. 3. Bussard, K.M. and G.H. Smith, The mammary gland microenvironment directs progenitor cell fate in vivo. Int J Cell Biol, 2011. 2011: p. 451676. 4. Lundqvist, E., et al., Co-twin control and cohort analyses of body mass index and height in relation to breast, prostate, ovarian, corpus uteri, colon and rectal cancer among Swedish and Finnish twins. Int J Cancer, 2007. 121(4): p. 810-8. 5. Lichtenstein, P., et al., Environmental and heritable factors in the causation of cancer--analyses of cohorts of twins from Sweden, Denmark, and Finland. N Engl J Med, 2000. 343(2): p. 78-85. 6. Brown, S.B., D.S. Morrison, and T.G. Cooke, Increasing incidence of breast cancer: distinguishing between the effects of birth cohort and a national breast screening programme. Breast Cancer Res Treat, 2009. 116(3): p. 603-7. 7. Chia, K.S., et al., Profound changes in breast cancer incidence may reflect changes into a Westernized lifestyle: a comparative population-based study in Singapore and Sweden. Int J Cancer, 2005. 113(2): p. 302-6. 8. Swerdlow, A.J., et al., Trends in cancer incidence and mortality in Scotland: description and possible explanations. Br J Cancer, 1998. 77 Suppl 3: p. 1-54. 9. dos Santos Silva, I. and A.J. Swerdlow, Recent trends in incidence of and mortality from breast, ovarian and endometrial cancers in England and Wales and their relation to changing fertility and oral contraceptive use. Br J Cancer, 1995. 72(2): p. 485-92. 10. M Morrow, H.B.H., JR Harris Malignant Tumors of the Breast. 10 ed. DeVita, Hellman, and Rosenberg’s Cancer: Principles and Practice of Oncology, ed. T.L. VT Jr. DeVita, SA Rosenberg 2015, Philadelphia PA: Wolters Kluwer Health. 11. Krieger, N., Is breast cancer a disease of affluence, poverty, or both? The case of African American women. Am J Public Health, 2002. 92(4): p. 611-3. 12. Jemal, A., et al., Global cancer statistics. CA Cancer J Clin, 2011. 61(2): p. 69- 90. 13. Orgnaization, W.H., Cancer incidence, mortality, and prevalence worldwide. 2000. 14. Chlebowski, R.T., Breast Cancer after Use of Estrogen plus Progestin in Postmenopausal Women. The New England journal of medicine., 2009. 360(6): p. 573.

145

15. Clemons, M. and P. Goss, Estrogen and the risk of breast cancer. N Engl J Med, 2001. 344(4): p. 276-85. 16. Spink, B.C., et al., Long-term estrogen exposure promotes carcinogen bioactivation, induces persistent changes in gene expression, and enhances the tumorigenicity of MCF-7 human breast cancer cells. Toxicol Appl Pharmacol, 2009. 240(3): p. 355-66. 17. Gustafsson, J.A., What pharmacologists can learn from recent advances in estrogen signalling. Trends Pharmacol Sci, 2003. 24(9): p. 479-85. 18. Burger, H.G., Androgen production in women. Fertil Steril, 2002. 77 Suppl 4: p. S3-5. 19. MW Elmlinger, W.K., MB Ranke Reference ranges for serum concentrations of lutropin (LH), follitropin (FSH), estradiol (E2), prolactin, progesterone, sex hormone-binding globulin (SHBG), dehydroepiandrosterone sulfate (DHEAS), cortisol and ferritin in neonates, children and young adults. Clin Chem Lab Med Vol. 40. 2002. 1151-1160. 20. Vandenberg, L.N., et al., Urinary, circulating, and tissue biomonitoring studies indicate widespread exposure to bisphenol A. Cien Saude Colet, 2012. 17(2): p. 407-34. 21. KA, J., Handbook of endocrinology ed. G.G. HM Kaplan. Vol. 1. 1996: CRC Press. 1-42. 22. Lim, E., O. Metzger-Filho, and E.P. Winer, The natural history of hormone receptor-positive breast cancer. Oncology (Williston Park), 2012. 26(8): p. 688- 94, 696. 23. Glass, A.G., et al., Breast cancer incidence, 1980-2006: combined roles of menopausal hormone therapy, screening mammography, and estrogen receptor status. J Natl Cancer Inst, 2007. 99(15): p. 1152-61. 24. Dahlman-Wright, K., et al., Interplay between AP-1 and estrogen receptor alpha in regulating gene expression and proliferation networks in breast cancer cells. Carcinogenesis, 2012. 33(9): p. 1684-91. 25. McKenna, N.J. and B.W. O'Malley, Combinatorial control of gene expression by nuclear receptors and coregulators. Cell, 2002. 108(4): p. 465-74. 26. Baek, S.H. and M.G. Rosenfeld, Nuclear receptor coregulators: their modification codes and regulatory mechanism by translocation. Biochem Biophys Res Commun, 2004. 319(3): p. 707-14. 27. Shughrue, P.J., et al., Comparative distribution of estrogen receptor-alpha (ER- alpha) and beta (ER-beta) mRNA in the rat pituitary, gonad, and reproductive tract. Steroids, 1998. 63(10): p. 498-504. 28. Tora, L., et al., The human estrogen receptor has two independent nonacidic transcriptional activation functions. Cell, 1989. 59(3): p. 477-87. 29. McKenna, N.J., R.B. Lanz, and B.W. O'Malley, Nuclear receptor coregulators: cellular and molecular biology. Endocr Rev, 1999. 20(3): p. 321-44. 30. Heery, D.M., et al., A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature, 1997. 387(6634): p. 733-6. 146

31. Mangelsdorf, D.J., et al., The nuclear receptor superfamily: the second decade. Cell, 1995. 83(6): p. 835-9. 32. Hewitt, S.C., J.C. Harrell, and K.S. Korach, Lessons in estrogen biology from knockout and transgenic animals. Annu Rev Physiol, 2005. 67: p. 285-308. 33. Couse, J.F., et al., Tissue distribution and quantitative analysis of estrogen receptor-alpha (ERalpha) and estrogen receptor-beta (ERbeta) messenger ribonucleic acid in the wild-type and ERalpha-knockout mouse. Endocrinology, 1997. 138(11): p. 4613-21. 34. Osborne, C.K., et al., Role of the estrogen receptor coactivator AIB1 (SRC-3) and HER-2/neu in tamoxifen resistance in breast cancer. J Natl Cancer Inst, 2003. 95(5): p. 353-61. 35. Sarwar, N., et al., Phosphorylation of ERalpha at serine 118 in primary breast cancer and in tamoxifen-resistant tumours is indicative of a complex role for ERalpha phosphorylation in breast cancer progression. Endocr Relat Cancer, 2006. 13(3): p. 851-61. 36. Brzozowski, A.M., et al., Molecular basis of agonism and antagonism in the oestrogen receptor. Nature, 1997. 389(6652): p. 753-8. 37. Weigelt, B., J.L. Peterse, and L.J. van 't Veer, Breast cancer metastasis: markers and models. Nat Rev Cancer, 2005. 5(8): p. 591-602. 38. Beral, V., Breast cancer and hormone-replacement therapy in the Million Women Study. Lancet, 2003. 362(9382): p. 419-27. 39. Rossouw, J.E., et al., Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women's Health Initiative randomized controlled trial. Jama, 2002. 288(3): p. 321-33. 40. (EBCTCG), E.B.C.T.C.G., Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet, 2005. 365(9472): p. 1687-717. 41. Jeselsohn, R., et al., The Evolving Role of the Estrogen Receptor Mutations in Endocrine Therapy-Resistant Breast Cancer. Curr Oncol Rep, 2017. 19(5): p. 35. 42. Williams, N. and L.N. Harris, The renaissance of endocrine therapy in breast cancer. Curr Opin Obstet Gynecol, 2014. 26(1): p. 41-7. 43. Welch, H.G., D.H. Gorski, and P.C. Albertsen, Trends in Metastatic Breast and Prostate Cancer--Lessons in Cancer Dynamics. N Engl J Med, 2015. 373(18): p. 1685-7. 44. S Hellman, J.R.H., Diseases of the Breast ed. M.L. JR Harris, M Morrow & CK Osborne. 2000: Lippincott Williams & Wilkins, Philadelphia. 45. Hoefnagel, L.D., et al., Prognostic value of estrogen receptor alpha and progesterone receptor conversion in distant breast cancer metastases. Cancer, 2012. 118(20): p. 4929-35. 46. Tomooka, Y. and H.A. Bern, Growth of mouse mammary glands after neonatal sex hormone treatment. J Natl Cancer Inst, 1982. 69(6): p. 1347-52. 47. Fenton, S.E., et al., Persistent abnormalities in the rat mammary gland following gestational and lactational exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxicol Sci, 2002. 67(1): p. 63-74. 147

48. Vandenberg, L.N., et al., Bisphenol-A and the great divide: a review of controversies in the field of endocrine disruption. Endocr Rev, 2009. 30(1): p. 75- 95. 49. Zoeller, R.T., et al., Endocrine-disrupting chemicals and public health protection: a statement of principles from The Endocrine Society. Endocrinology, 2012. 153(9): p. 4097-110. 50. Oostenbrink, B.C., et al., Simulations of the estrogen receptor ligand-binding domain: affinity of natural ligands and xenoestrogens. J Med Chem, 2000. 43(24): p. 4594-605. 51. Bolger, R., et al., Rapid screening of environmental chemicals for estrogen receptor binding capacity. Environ Health Perspect, 1998. 106(9): p. 551-7. 52. Jordan, V.C., et al., Structure-activity relationships of estrogens. Environ Health Perspect, 1985. 61: p. 97-110. 53. Hong, E.J., et al., Identification of estrogen-regulated genes by microarray analysis of the uterus of immature rats exposed to endocrine disrupting chemicals. Reprod Biol Endocrinol, 2006. 4: p. 49. 54. Hatch, E.E., et al., Cancer risk in women exposed to diethylstilbestrol in utero. Jama, 1998. 280(7): p. 630-4. 55. Hoover, R.N., et al., Adverse health outcomes in women exposed in utero to diethylstilbestrol. N Engl J Med, 2011. 365(14): p. 1304-14. 56. Palmer, J.R., et al., Prenatal diethylstilbestrol exposure and risk of breast cancer. Cancer Epidemiol Biomarkers Prev, 2006. 15(8): p. 1509-14. 57. Colton, T., et al., Breast cancer in mothers prescribed diethylstilbestrol in pregnancy. Further follow-up. Jama, 1993. 269(16): p. 2096-100. 58. Greenberg, E.R., et al., Breast cancer in mothers given diethylstilbestrol in pregnancy. N Engl J Med, 1984. 311(22): p. 1393-8. 59. (NTP), N.T.P., ed. Report on Carcinogens 12 ed. 60. Vandenberg, L.N., et al., Hormones and endocrine-disrupting chemicals: low- dose effects and nonmonotonic dose responses. Endocr Rev, 2012. 33(3): p. 378- 455. 61. Welshons, W.V., S.C. Nagel, and F.S. vom Saal, Large effects from small exposures. III. Endocrine mechanisms mediating effects of bisphenol A at levels of human exposure. Endocrinology, 2006. 147(6 Suppl): p. S56-69. 62. Vandenberg, L.N., Non-monotonic dose responses in studies of endocrine disrupting chemicals: bisphenol a as a case study. Dose Response, 2014. 12(2): p. 259-76. 63. Burridge, E., Bisphenol A: product profile. 2003. 64. Talsness, C.E., et al., Components of plastic: experimental studies in animals and relevance for human health. Philos Trans R Soc Lond B Biol Sci, 2009. 364(1526): p. 2079-96. 65. Calafat, A.M., et al., Exposure of the U.S. population to bisphenol A and 4- tertiary-octylphenol: 2003-2004. Environ Health Perspect, 2008. 116(1): p. 39-44. 148

66. Engel, S.M., et al., Xenobiotic phenols in early pregnancy amniotic fluid. Reprod Toxicol, 2006. 21(1): p. 110-2. 67. Gerona, R.R., et al., Bisphenol-A (BPA), BPA glucuronide, and BPA sulfate in midgestation umbilical cord serum in a northern and central California population. Environ Sci Technol, 2013. 47(21): p. 12477-85. 68. Chen, M., et al., Determination of bisphenol-A levels in human amniotic fluid samples by liquid chromatography coupled with mass spectrometry. J Sep Sci, 2011. 34(14): p. 1648-55. 69. Balakrishnan, B., et al., Transfer of bisphenol A across the human placenta. Am J Obstet Gynecol, 2010. 202(4): p. 393.e1-7. 70. Nishikawa, M., et al., Placental transfer of conjugated bisphenol A and subsequent reactivation in the rat fetus. Environ Health Perspect, 2010. 118(9): p. 1196-203. 71. Doerge, D.R., et al., Pharmacokinetics of bisphenol A in neonatal and adult CD-1 mice: inter-species comparisons with Sprague-Dawley rats and rhesus monkeys. Toxicol Lett, 2011. 207(3): p. 298-305. 72. Doerge, D.R., et al., Pharmacokinetics of bisphenol A in neonatal and adult Sprague-Dawley rats. Toxicol Appl Pharmacol, 2010. 247(2): p. 158-65. 73. Volkel, W., et al., Metabolism and kinetics of bisphenol a in humans at low doses following oral administration. Chem Res Toxicol, 2002. 15(10): p. 1281-7. 74. Volkel, W., N. Bittner, and W. Dekant, Quantitation of bisphenol A and bisphenol A glucuronide in biological samples by high performance liquid chromatography- tandem mass spectrometry. Drug Metab Dispos, 2005. 33(11): p. 1748-57. 75. Bray, F., P. McCarron, and D.M. Parkin, The changing global patterns of female breast cancer incidence and mortality. Breast Cancer Res, 2004. 6(6): p. 229-39. 76. Ziegler, R.G., et al., Migration patterns and breast cancer risk in Asian-American women. J Natl Cancer Inst, 1993. 85(22): p. 1819-27. 77. Macon, M.B. and S.E. Fenton, Endocrine disruptors and the breast: early life effects and later life disease. J Mammary Gland Biol Neoplasia, 2013. 18(1): p. 43-61. 78. Soto, A.M., et al., Does cancer start in the womb? altered mammary gland development and predisposition to breast cancer due to in utero exposure to endocrine disruptors. J Mammary Gland Biol Neoplasia, 2013. 18(2): p. 199-208. 79. Tharp, A.P., et al., Bisphenol A alters the development of the rhesus monkey mammary gland. Proc Natl Acad Sci U S A, 2012. 109(21): p. 8190-5. 80. Munoz-de-Toro, M., et al., Perinatal exposure to bisphenol-A alters peripubertal mammary gland development in mice. Endocrinology, 2005. 146(9): p. 4138-47. 81. Fenton, S.E., C. Reed, and R.R. Newbold, Perinatal environmental exposures affect mammary development, function, and cancer risk in adulthood. Annu Rev Pharmacol Toxicol, 2012. 52: p. 455-79. 82. Durando, M., et al., Prenatal bisphenol A exposure induces preneoplastic lesions in the mammary gland in Wistar rats. Environ Health Perspect, 2007. 115(1): p. 80-6.

149

83. Markey, C.M., et al., In utero exposure to bisphenol A alters the development and tissue organization of the mouse mammary gland. Biol Reprod, 2001. 65(4): p. 1215-23. 84. Dhimolea, E., et al., Prenatal exposure to BPA alters the epigenome of the rat mammary gland and increases the propensity to neoplastic development. PLoS One, 2014. 9(7): p. e99800. 85. Vandenberg, L.N., et al., Exposure to environmentally relevant doses of the xenoestrogen bisphenol-A alters development of the fetal mouse mammary gland. Endocrinology, 2007. 148(1): p. 116-27. 86. Wadia, P.R., et al., Low-dose BPA exposure alters the mesenchymal and epithelial transcriptomes of the mouse fetal mammary gland. PLoS One, 2013. 8(5): p. e63902. 87. Doherty, L.F., et al., In utero exposure to diethylstilbestrol (DES) or bisphenol-A (BPA) increases EZH2 expression in the mammary gland: an epigenetic mechanism linking endocrine disruptors to breast cancer. Horm Cancer, 2010. 1(3): p. 146-55. 88. Ninomiya, K., et al., Effects of neonatally administered diethylstilbestrol on induction of mammary carcinomas induced by 7, 12-dimethylbenz(a)anthracene in female rats. Toxicol Pathol, 2007. 35(6): p. 813-8. 89. Kawaguchi, H., et al., Effects of fetal exposure to diethylstilbestrol on mammary tumorigenesis in rats. J Vet Med Sci, 2009. 71(12): p. 1599-608. 90. Yoshikawa, T., et al., Effects of neonatally administered low-dose diethylstilbestrol on the induction of mammary carcinomas and dysplasias induced by 7,12-dimethylbenz [a] anthracene in female rats. In Vivo, 2008. 22(2): p. 207-13. 91. Vandenberg, L.N., et al., Perinatal exposure to the xenoestrogen bisphenol-A induces mammary intraductal hyperplasias in adult CD-1 mice. Reprod Toxicol, 2008. 26(3-4): p. 210-9. 92. Jenkins, S., et al., Chronic oral exposure to bisphenol A results in a nonmonotonic dose response in mammary carcinogenesis and metastasis in MMTV-erbB2 mice. Environ Health Perspect, 2011. 119(11): p. 1604-9. 93. Weber Lozada, K. and R.A. Keri, Bisphenol A increases mammary cancer risk in two distinct mouse models of breast cancer. Biol Reprod, 2011. 85(3): p. 490-7. 94. Acevedo, N., et al., Perinatally administered bisphenol a as a potential mammary gland carcinogen in rats. Environ Health Perspect, 2013. 121(9): p. 1040-6. 95. Korach, K.S., et al., Estrogen receptor gene disruption: molecular characterization and experimental and clinical phenotypes. Recent Prog Horm Res, 1996. 51: p. 159-86; discussion 186-8. 96. Krege, J.H., et al., Generation and reproductive phenotypes of mice lacking estrogen receptor beta. Proc Natl Acad Sci U S A, 1998. 95(26): p. 15677-82. 97. Couse, J.F. and K.S. Korach, Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev, 1999. 20(3): p. 358-417.

150

98. Mallepell, S., et al., Paracrine signaling through the epithelial estrogen receptor alpha is required for proliferation and morphogenesis in the mammary gland. Proc Natl Acad Sci U S A, 2006. 103(7): p. 2196-201. 99. Hiremath, M., et al., Parathyroid hormone-related protein activates Wnt signaling to specify the embryonic mammary mesenchyme. Development, 2012. 139(22): p. 4239-49. 100. Narbaitz, R., W.E. Stumpf, and M. Sar, Estrogen receptors in mammary gland primordia of fetal mouse. Anat Embryol (Berl), 1980. 158(2): p. 161-6. 101. Robinson, G.W., A.B. Karpf, and K. Kratochwil, Regulation of mammary gland development by tissue interaction. J Mammary Gland Biol Neoplasia, 1999. 4(1): p. 9-19. 102. Robinson, G.W., Cooperation of signalling pathways in embryonic mammary gland development. Nat Rev Genet, 2007. 8(12): p. 963-72. 103. Lemmen, J.G., et al., Expression of estrogen receptor alpha and beta during mouse embryogenesis. Mech Dev, 1999. 81(1-2): p. 163-7. 104. Saji, S., et al., Estrogen receptors alpha and beta in the rodent mammary gland. Proc Natl Acad Sci U S A, 2000. 97(1): p. 337-42. 105. Inman, J.L., et al., Mammary gland development: cell fate specification, stem cells and the microenvironment. Development, 2015. 142(6): p. 1028-42. 106. Nelson, C.M., et al., Tissue geometry determines sites of mammary branching morphogenesis in organotypic cultures. Science, 2006. 314(5797): p. 298-300. 107. Ricketts, D., et al., Estrogen and progesterone receptors in the normal female breast. Cancer Res, 1991. 51(7): p. 1817-22. 108. Shackleton, M., et al., Generation of a functional mammary gland from a single stem cell. Nature, 2006. 439(7072): p. 84-8. 109. Wiseman, B.S. and Z. Werb, Stromal effects on mammary gland development and breast cancer. Science, 2002. 296(5570): p. 1046-9. 110. Nelson, C.M. and M.J. Bissell, Of extracellular matrix, scaffolds, and signaling: tissue architecture regulates development, homeostasis, and cancer. Annu Rev Cell Dev Biol, 2006. 22: p. 287-309. 111. Provenzano, P.P., et al., Collagen density promotes mammary tumor initiation and progression. BMC Med, 2008. 6: p. 11. 112. Slanetz, P.J., P.E. Freer, and R.L. Birdwell, Breast-density legislation--practical considerations. N Engl J Med, 2015. 372(7): p. 593-5. 113. Brisken, C. and B. O'Malley, Hormone action in the mammary gland. Cold Spring Harb Perspect Biol, 2010. 2(12): p. a003178. 114. Fenton, S.E., Endocrine-disrupting compounds and mammary gland development: early exposure and later life consequences. Endocrinology, 2006. 147(6 Suppl): p. S18-24. 115. Kratochwil, K., Organ specificity in mesenchymal induction demonstrated in the embryonic development of the mammary gland of the mouse. Dev Biol, 1969. 20(1): p. 46-71.

151

116. Sakakura, T., Y. Sakagami, and Y. Nishizuka, Persistence of responsiveness of adult mouse mammary gland to induction by embryonic mesenchyme. Dev Biol, 1979. 72(2): p. 201-10. 117. Sakakura, T., Y. Nishizuka, and C.J. Dawe, Mesenchyme-dependent morphogenesis and epithelium-specific cytodifferentiation in mouse mammary gland. Science, 1976. 194(4272): p. 1439-41. 118. Cunha, G.R., et al., Mammary phenotypic expression induced in epidermal cells by embryonic mammary mesenchyme. Acta Anat (Basel), 1995. 152(3): p. 195- 204. 119. Lochter, A., et al., Matrix metalloproteinase stromelysin-1 triggers a cascade of molecular alterations that leads to stable epithelial-to-mesenchymal conversion and a premalignant phenotype in mammary epithelial cells. J Cell Biol, 1997. 139(7): p. 1861-72. 120. Sternlicht, M.D., M.J. Bissell, and Z. Werb, The matrix metalloproteinase stromelysin-1 acts as a natural mammary tumor promoter. Oncogene, 2000. 19(8): p. 1102-13. 121. Duhl, D.M., et al., Neomorphic agouti mutations in obese yellow mice. Nat Genet, 1994. 8(1): p. 59-65. 122. Dolinoy, D.C., D. Huang, and R.L. Jirtle, Maternal nutrient supplementation counteracts bisphenol A-induced DNA hypomethylation in early development. Proc Natl Acad Sci U S A, 2007. 104(32): p. 13056-61. 123. Palanza, P., et al., Prenatal exposure to endocrine disrupting chemicals: effects on behavioral development. Neurosci Biobehav Rev, 1999. 23(7): p. 1011-27. 124. Schug, T.T., et al., Endocrine disrupting chemicals and disease susceptibility. J Steroid Biochem Mol Biol, 2011. 127(3-5): p. 204-15. 125. Li, Y., et al., Endocrine-Disrupting Chemicals (EDCs): In Vitro Mechanism of Estrogenic Activation and Differential Effects on ER Target Genes. Environ Health Perspect, 2013. 121(4): p. 459-66. 126. AM Calafat, X.Y., L Wong, JA Reidy, and LL Needham, Exposure of the U.S. Population to Bisphenol A and 4-tertiary-Octylphenol: 2003–2004. Environmental Health Perspectives, 2008. 116(1): p. 39-44. 127. Vandenberg, L.N., et al., Human exposure to bisphenol A (BPA). Reprod Toxicol, 2007. 24(2): p. 139-77. 128. Grace, P.B., et al., Phytoestrogen concentrations in serum and spot urine as biomarkers for dietary phytoestrogen intake and their relation to breast cancer risk in European prospective investigation of cancer and nutrition-norfolk. Cancer Epidemiol Biomarkers Prev, 2004. 13(5): p. 698-708. 129. Allred, C.D., et al., Soy diets containing varying amounts of genistein stimulate growth of estrogen-dependent (MCF-7) tumors in a dose-dependent manner. Cancer Res, 2001. 61(13): p. 5045-50. 130. Kuiper, G.G., et al., Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology, 1998. 139(10): p. 4252-63.

152

131. Gertz, J., et al., Genistein and bisphenol A exposure cause estrogen receptor 1 to bind thousands of sites in a cell type-specific manner. Genome Res, 2012. 22(11): p. 2153-62. 132. Ricupito, A., et al., Effect of bisphenol A with or without enzyme treatment on the proliferation and viability of MCF-7 cells. Environ Int, 2009. 35(1): p. 21-6. 133. Han, D., H. Tachibana, and K. Yamada, Inhibition of environmental estrogen- induced proliferation of human breast carcinoma MCF-7 cells by flavonoids. In Vitro Cell Dev Biol Anim, 2001. 37(5): p. 275-82. 134. Clemons, M.G.P., Estrogen and the Risk of Breast Cancer. New England Journal of Medicine, 2001. 344(4). 135. Clarke, S.D., et al., Differential effects of acute and chronic estrogen treatment on thermogenic and metabolic pathways in ovariectomized sheep. Endocrinology, 2013. 154(1): p. 184-92. 136. Liu, F., et al., Effects of chronic and acute oestrogen replacement therapy in aged animals after experimental stroke. J Neuroendocrinol, 2012. 24(2): p. 319-30. 137. Kuba, T., et al., Acute and chronic estradiol replacements differentially alter corticosterone and COX-mediated responses to an inflammatory stimulus in female rats. Ethn Dis, 2010. 20(1 Suppl 1): p. S1-50-4. 138. Wang, J., S. Jenkins, and C.A. Lamartiniere, Cell proliferation and apoptosis in rat mammary glands following combinational exposure to bisphenol A and genistein. BMC Cancer, 2014. 14(1): p. 379. 139. Burd, C.J., et al., Analysis of chromatin dynamics during glucocorticoid receptor activation. Mol Cell Biol, 2012. 32(10): p. 1805-17. 140. Burch, J.B. and H. Weintraub, Temporal order of chromatin structural changes associated with activation of the major chicken vitellogenin gene. Cell, 1983. 33(1): p. 65-76. 141. Hon, G., W. Wang, and B. Ren. Discovery and annotation of functional chromatin signatures in the human genome. PLoS Comput Biol 2009 Nov [cited 5 11]; 2009/11/18:[e1000566]. 142. Burd, C.J. and T.K. Archer, Chromatin architecture defines the glucocorticoid response. Mol Cell Endocrinol, 2013. 380(1-2): p. 25-31. 143. He, H.H., et al., Differential DNase I hypersensitivity reveals factor-dependent chromatin dynamics. Genome Res, 2012. 22(6): p. 1015-25. 144. Krishnan, A.V., et al., Bisphenol-A: an estrogenic substance is released from polycarbonate flasks during autoclaving. Endocrinology, 1993. 132(6): p. 2279- 86. 145. Maggiolini, M., et al., Estrogen receptor alpha mediates the proliferative but not the cytotoxic dose-dependent effects of two major phytoestrogens on human breast cancer cells. Mol Pharmacol, 2001. 60(3): p. 595-602. 146. Korach, K.S., et al., Estrogen receptor-binding activity of polychlorinated hydroxybiphenyls: conformationally restricted structural probes. Mol Pharmacol, 1988. 33(1): p. 120-6. 147. Soto, A.M., et al., p-Nonyl-phenol: an estrogenic xenobiotic released from "modified" polystyrene. Environ Health Perspect, 1991. 92: p. 167-73. 153

148. Liu, C., et al., Exposure to bisphenol A disrupts meiotic progression during spermatogenesis in adult rats through estrogen-like activity. Cell Death Dis, 2013. 4: p. e676. 149. Barkhem, T., et al., Differential response of estrogen receptor alpha and estrogen receptor beta to partial estrogen agonists/antagonists. Mol Pharmacol, 1998. 54(1): p. 105-12. 150. vom Saal, F.S., et al., Chapel Hill bisphenol A expert panel consensus statement: integration of mechanisms, effects in animals and potential to impact human health at current levels of exposure. Reprod Toxicol, 2007. 24(2): p. 131-8. 151. Henley, D.V. and K.S. Korach, Physiological effects and mechanisms of action of endocrine disrupting chemicals that alter estrogen signaling. Hormones (Athens), 2010. 9(3): p. 191-205. 152. Banerjee, S., et al., Multi-targeted therapy of cancer by genistein. Cancer Lett, 2008. 269(2): p. 226-42. 153. Rezg, R., et al., Bisphenol A and human chronic diseases: current evidences, possible mechanisms, and future perspectives. Environ Int, 2014. 64: p. 83-90. 154. Kato S, Y.A.F.R., Nuclear receptor coregulators merge transcriptional coregulation with epigenetic regulation. Trends in biochemical sciences, 2011. 36(5): p. 272-81. 155. Li, Y., et al., Differential estrogenic actions of endocrine-disrupting chemicals bisphenol A, bisphenol AF, and zearalenone through estrogen receptor alpha and beta in vitro. Environ Health Perspect, 2012. 120(7): p. 1029-35. 156. Ikezuki, Y., et al., Determination of bisphenol A concentrations in human biological fluids reveals significant early prenatal exposure. Hum Reprod, 2002. 17(11): p. 2839-41. 157. Mori, T., et al., Long-term effects of neonatal steroid exposure on mammary gland development and tumorigenesis in mice. J Natl Cancer Inst, 1976. 57(5): p. 1057-62. 158. Jenkins, S., et al., Oral exposure to bisphenol a increases dimethylbenzanthracene-induced mammary cancer in rats. Environ Health Perspect, 2009. 117(6): p. 910-5. 159. Mandrup, K., et al., Low-dose effects of bisphenol A on mammary gland development in rats. Andrology, 2016. 160. Lamartiniere, C.A., et al., Exposure to the Endocrine Disruptor Bisphenol A Alters Susceptibility for Mammary Cancer. Horm Mol Biol Clin Investig, 2011. 5(2): p. 45-52. 161. Betancourt, A.M., et al., In utero exposure to bisphenol A shifts the window of susceptibility for mammary carcinogenesis in the rat. Environ Health Perspect, 2010. 118(11): p. 1614-9. 162. Westwood, F.R., The female rat reproductive cycle: a practical histological guide to staging. Toxicol Pathol, 2008. 36(3): p. 375-84. 163. Byers, S.L., et al., Mouse estrous cycle identification tool and images. PLoS One, 2012. 7(4): p. e35538.

154

164. Cora, M.C., L. Kooistra, and G. Travlos, Vaginal Cytology of the Laboratory Rat and Mouse: Review and Criteria for the Staging of the Estrous Cycle Using Stained Vaginal Smears. Toxicol Pathol, 2015. 43(6): p. 776-93. 165. Caligioni, C.S., Assessing reproductive status/stages in mice. Curr Protoc Neurosci, 2009. Appendix 4: p. Appendix 4I. 166. Program, N.T., Specifications for the conduct of studies to evaluate the reproductive and developmental toxicitiy of chemical, biological and physical agents in laboratory animals for the National Toxicology Program (NTP), N.T. Program, Editor. 2011. 167. Oberlies, N.H., et al., Microbial-mediated release of bisphenol A from polycarbonate vessels. Lett Appl Microbiol, 2008. 46(2): p. 271-5. 168. Yi, B., C. Kim, and M. Yang, Biological monitoring of bisphenol A with HLPC/FLD and LC/MS/MS assays. J Chromatogr B Analyt Technol Biomed Life Sci, 2010. 878(27): p. 2606-10. 169. Abramoff, M.D., Magalhaes, P.J., Ram, S.J., Image processing with Image J, in Biophotonics International. 2004. p. 36-42. 170. Stanko, J.P., M.R. Easterling, and S.E. Fenton, Application of Sholl analysis to quantify changes in growth and development in rat mammary gland whole mounts. Reprod Toxicol, 2015. 54: p. 129-35. 171. Hsu, J., Multiple comparisons: Theory and methods. 1996 ed, ed. G.S. Practitioner. 1996: Chapman and Hall/ CRC. 296. 172. Hovey, R.C., J.F. Trott, and B.K. Vonderhaar, Establishing a framework for the functional mammary gland: from endocrinology to morphology. J Mammary Gland Biol Neoplasia, 2002. 7(1): p. 17-38. 173. Adam, E., et al., Vaginal and cervical cancers and other abnormalities associated with exposure in utero to diethylstilbestrol and related synthetic hormones. Cancer Res, 1977. 37(4): p. 1249-51. 174. Herbst, A.L., H. Ulfelder, and D.C. Poskanzer, Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young women. N Engl J Med, 1971. 284(15): p. 878-81. 175. Hens, J.R. and J.J. Wysolmerski, Key stages of mammary gland development: molecular mechanisms involved in the formation of the embryonic mammary gland. Breast Cancer Res, 2005. 7(5): p. 220-4. 176. Sakakura, T., Mammary embryogenesis. The mammary gland: Development, regulation, and function, ed. M.C. Neville, and Daniel, C.W. 1987, New York: Plenum Press. 177. Pottenger, L.H., et al., The relative bioavailability and metabolism of bisphenol A in rats is dependent upon the route of administration. Toxicol Sci, 2000. 54(1): p. 3-18. 178. Prossnitz, E.R. and M. Barton, The G-protein-coupled estrogen receptor GPER in health and disease. Nat Rev Endocrinol, 2011. 7(12): p. 715-26. 179. Matsushima, A., et al., Bisphenol AF is a full agonist for the estrogen receptor ERalpha but a highly specific antagonist for ERbeta. Environ Health Perspect, 2010. 118(9): p. 1267-72. 155

180. Blair, R.M., et al., The estrogen receptor relative binding affinities of 188 natural and xenochemicals: structural diversity of ligands. Toxicol Sci, 2000. 54(1): p. 138-53. 181. Speroni, L., et al., New insights into fetal mammary gland morphogenesis: differential effects of natural and environmental estrogens. Sci Rep, 2017. 7: p. 40806. 182. Sakakura, T., New aspects of stroma-parenchyma relations in mammary gland differentiation. Int Rev Cytol, 1991. 125: p. 165-202. 183. Parmar, H. and G.R. Cunha, Epithelial-stromal interactions in the mouse and human mammary gland in vivo. Endocr Relat Cancer, 2004. 11(3): p. 437-58. 184. Maffini, M.V., et al., Stromal regulation of neoplastic development: age- dependent normalization of neoplastic mammary cells by mammary stroma. Am J Pathol, 2005. 167(5): p. 1405-10. 185. Kurosawa, T., et al., The activity of bisphenol A depends on both the estrogen receptor subtype and the cell type. Endocr J, 2002. 49(4): p. 465-71. 186. Prossnitz, E.R. and H.J. Hathaway, What have we learned about GPER function in physiology and disease from knockout mice? J Steroid Biochem Mol Biol, 2015. 153: p. 114-26. 187. Keeling, J.W., et al., Oestrogen receptor alpha in female fetal, infant, and child mammary tissue. J Pathol, 2000. 191(4): p. 449-51. 188. Naccarato, A.G., et al., Bio-morphological events in the development of the human female mammary gland from fetal age to puberty. Virchows Arch, 2000. 436(5): p. 431-8. 189. Friedrichs, N., et al., Immunohistochemical expression patterns of AP2alpha and AP2gamma in the developing fetal human breast. Histopathology, 2007. 51(6): p. 814-23. 190. Noller, K.L., et al., Clear-cell adenocarcinoma of the cervix after maternal treatment with synthetic estrogens. Mayo Clin Proc, 1972. 47(9): p. 629-30. 191. Shekhar, M.P., et al., Breast stroma plays a dominant regulatory role in breast epithelial growth and differentiation: implications for tumor development and progression. Cancer Res, 2001. 61(4): p. 1320-6. 192. Donjacour, A.A. and G.R. Cunha, Stromal regulation of epithelial function. Cancer Treat Res, 1991. 53: p. 335-64. 193. Cunha, G.R., et al., Stromal-epithelial interactions in adult organs. Cell Differ, 1985. 17(3): p. 137-48. 194. Farber, E., The multistep nature of cancer development. Cancer Res, 1984. 44(10): p. 4217-23. 195. van den Hooff, A., Stromal involvement in malignant growth. Adv Cancer Res, 1988. 50: p. 159-96. 196. Barcellos-Hoff, M.H., It takes a tissue to make a tumor: epigenetics, cancer and the microenvironment. J Mammary Gland Biol Neoplasia, 2001. 6(2): p. 213-21. 197. Haslam, S.Z. and K.A. Nummy, The ontogeny and cellular distribution of estrogen receptors in normal mouse mammary gland. J Steroid Biochem Mol Biol, 1992. 42(6): p. 589-95. 156

198. Shyamala, G. and A. Ferenczy, Mammary fat pad may be a potential site for initiation of estrogen action in normal mouse mammary glands. Endocrinology, 1984. 115(3): p. 1078-81. 199. Kim, J.H. and A.R. Scialli, Thalidomide: the tragedy of birth defects and the effective treatment of disease. Toxicol Sci, 2011. 122(1): p. 1-6. 200. Feng, Y., et al., Estrogen receptor-alpha expression in the mammary epithelium is required for ductal and alveolar morphogenesis in mice. Proc Natl Acad Sci U S A, 2007. 104(37): p. 14718-23. 201. Visvader, J.E. and J. Stingl, Mammary stem cells and the differentiation hierarchy: current status and perspectives. Genes Dev, 2014. 28(11): p. 1143-58. 202. (EPA), E.P.A. Integrated Risk Information System (IRIS): Bisphenol A. (CASRN 80–05–7). 1988; Available from: http://www.epa.gov/iris/subst/0356.htm. 203. Smalley, M., Isolation, Culture and Analysis of Mouse Mammary Epithelial Cells, in Mouse Cell Culture: Methods and Protocols, A.W.D. Tosh, Editor. 2010, Humana Press: United Kingdom. p. 139-170. 204. Xu, H., et al., FastUniq: a fast de novo duplicates removal tool for paired short reads. PLoS One, 2012. 7(12): p. e52249. 205. Trapnell, C., L. Pachter, and S.L. Salzberg, TopHat: discovering splice junctions with RNA-Seq. Bioinformatics, 2009. 25(9): p. 1105-11. 206. Trapnell, C., et al., Differential gene and transcript expression analysis of RNA- seq experiments with TopHat and Cufflinks. Nat Protoc, 2012. 7(3): p. 562-78. 207. Trapnell, C., et al., Differential analysis of gene regulation at transcript resolution with RNA-seq. Nat Biotechnol, 2013. 31(1): p. 46-53. 208. Johnson, W.E., C. Li, and A. Rabinovic, Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics, 2007. 8(1): p. 118- 27. 209. Tarazona, S., et al., Differential expression in RNA-seq: a matter of depth. Genome Res, 2011. 21(12): p. 2213-23. 210. Chen, J., et al., ToppGene Suite for gene list enrichment analysis and candidate gene prioritization. Nucleic Acids Res, 2009. 37(Web Server issue): p. W305-11. 211. Hammer, A.M., et al., Stromal PDGFR-alpha Activation Enhances Matrix Stiffness, Impedes Mammary Ductal Development, and Accelerates Tumor Growth. Neoplasia, 2017. 19(6): p. 496-508. 212. Qin, D., Y. Xia, and G.M. Whitesides, Soft lithography for micro- and nanoscale patterning. Nat Protoc, 2010. 5(3): p. 491-502. 213. Sung, K.E., et al., Control of 3-dimensional collagen matrix polymerization for reproducible human mammary fibroblast cell culture in microfluidic devices. Biomaterials, 2009. 30(27): p. 4833-41. 214. Bischel, L.L., D.J. Beebe, and K.E. Sung, Microfluidic model of ductal carcinoma in situ with 3D, organotypic structure. BMC Cancer, 2015. 15: p. 12. 215. Wiig, H. and M.A. Swartz, Interstitial fluid and lymph formation and transport: physiological regulation and roles in inflammation and cancer. Physiol Rev, 2012. 92(3): p. 1005-60. 157

216. Lim, E., et al., Transcriptome analyses of mouse and human mammary cell subpopulations reveal multiple conserved genes and pathways. Breast Cancer Res, 2010. 12(2): p. R21. 217. Kendrick, H., et al., Transcriptome analysis of mammary epithelial subpopulations identifies novel determinants of lineage commitment and cell fate. BMC Genomics, 2008. 9: p. 591. 218. Guo, Q., et al., Physiologically activated mammary fibroblasts promote postpartum mammary cancer. JCI Insight, 2017. 2(6): p. e89206. 219. Simian, M., et al., The interplay of matrix metalloproteinases, morphogens and growth factors is necessary for branching of mammary epithelial cells. Development, 2001. 128(16): p. 3117-31. 220. Montesano, R., et al., Identification of a fibroblast-derived epithelial morphogen as hepatocyte growth factor. Cell, 1991. 67(5): p. 901-8. 221. Yant, J., et al., In vivo effects of hepatocyte growth factor/scatter factor on mouse mammary gland development. Exp Cell Res, 1998. 241(2): p. 476-81. 222. Niranjan, B., et al., HGF/SF: a potent cytokine for mammary growth, morphogenesis and development. Development, 1995. 121(9): p. 2897-908. 223. Telerman, A. and R. Amson, The molecular programme of tumour reversion: the steps beyond malignant transformation. Nat Rev Cancer, 2009. 9(3): p. 206-16. 224. Amson, R., et al., Reciprocal repression between P53 and TCTP. Nat Med, 2011. 18(1): p. 91-9. 225. Frantz, C., K.M. Stewart, and V.M. Weaver, The extracellular matrix at a glance. J Cell Sci, 2010. 123(Pt 24): p. 4195-200. 226. Swartz, M.A. and M.E. Fleury, Interstitial flow and its effects in soft tissues. Annu Rev Biomed Eng, 2007. 9: p. 229-56. 227. Sakakura, T., Y. Sakagami, and Y. Nishizuka, Dual origin of mesenchymal tissues participating in mouse mammary gland embryogenesis. Dev Biol, 1982. 91(1): p. 202-7. 228. Cox, T.R. and J.T. Erler, Remodeling and homeostasis of the extracellular matrix: implications for fibrotic diseases and cancer. Dis Model Mech, 2011. 4(2): p. 165-78. 229. Hay, E.D., Extracellular matrix alters epithelial differentiation. Curr Opin Cell Biol, 1993. 5(6): p. 1029-35. 230. Polyak, K. and R. Kalluri, The role of the microenvironment in mammary gland development and cancer. Cold Spring Harb Perspect Biol, 2010. 2(11): p. a003244. 231. Schedin, P. and P.J. Keely, Mammary gland ECM remodeling, stiffness, and mechanosignaling in normal development and tumor progression. Cold Spring Harb Perspect Biol, 2011. 3(1): p. a003228. 232. Muschler, J. and C.H. Streuli, Cell-matrix interactions in mammary gland development and breast cancer. Cold Spring Harb Perspect Biol, 2010. 2(10): p. a003202. 233. Tomasek, J.J., et al., Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat Rev Mol Cell Biol, 2002. 3(5): p. 349-63. 158

234. Kalluri, R. and M. Zeisberg, Fibroblasts in cancer. Nat Rev Cancer, 2006. 6(5): p. 392-401. 235. De Wever, O., et al., Stromal myofibroblasts are drivers of invasive cancer growth. Int J Cancer, 2008. 123(10): p. 2229-38. 236. Friedl, P. and K. Wolf, Plasticity of cell migration: a multiscale tuning model. J Cell Biol, 2010. 188(1): p. 11-9. 237. Levental, K.R., et al., Matrix crosslinking forces tumor progression by enhancing integrin signaling. Cell, 2009. 139(5): p. 891-906. 238. Arnold, S.A., et al., Lack of host SPARC enhances vascular function and tumor spread in an orthotopic murine model of pancreatic carcinoma. Dis Model Mech, 2010. 3(1-2): p. 57-72. 239. Keri, R.A., et al., An evaluation of evidence for the carcinogenic activity of bisphenol A. Reprod Toxicol, 2007. 24(2): p. 240-52. 240. Boyd, N.F., et al., Mammographic density and the risk and detection of breast cancer. N Engl J Med, 2007. 356(3): p. 227-36. 241. Yaghjyan, L., et al., Mammographic breast density and subsequent risk of breast cancer in postmenopausal women according to tumor characteristics. J Natl Cancer Inst, 2011. 103(15): p. 1179-89. 242. Paszek, M.J., et al., Tensional homeostasis and the malignant phenotype. Cancer Cell, 2005. 8(3): p. 241-54. 243. Ronnov-Jessen, L. and M.J. Bissell, Breast cancer by proxy: can the microenvironment be both the cause and consequence? Trends Mol Med, 2009. 15(1): p. 5-13. 244. Xiao, Q. and G. Ge, Lysyl oxidase, extracellular matrix remodeling and cancer metastasis. Cancer Microenviron, 2012. 5(3): p. 261-73. 245. Provenzano, P.P., et al., Collagen reorganization at the tumor-stromal interface facilitates local invasion. BMC Med, 2006. 4(1): p. 38. 246. Bredfeldt, J.S., et al., Computational segmentation of collagen fibers from second- harmonic generation images of breast cancer. J Biomed Opt, 2014. 19(1): p. 16007. 247. Strunck, E., et al., Relative binding affinity does not predict biological response to xenoestrogens in rat endometrial adenocarcinoma cells. J Steroid Biochem Mol Biol, 2000. 74(3): p. 73-81. 248. Lee, G.S., et al., Effect of genistein as a selective estrogen receptor beta agonist on the expression of Calbindin-D9k in the uterus of immature rats. Toxicol Sci, 2004. 82(2): p. 451-7. 249. Gaido, K.W., et al., Evaluation of chemicals with endocrine modulating activity in a yeast-based steroid hormone receptor gene transcription assay. Toxicol Appl Pharmacol, 1997. 143(1): p. 205-12. 250. Vandenberg, L.N., et al., Regulatory decisions on endocrine disrupting chemicals should be based on the principles of endocrinology. Reprod Toxicol, 2013. 38: p. 1-15.

159

251. Aoyagi, S. and T.K. Archer, Dynamics of coactivator recruitment and chromatin modifications during nuclear receptor mediated transcription. Mol Cell Endocrinol, 2008. 280(1-2): p. 1-5. 252. Lonard, D.M. and B.W. O'Malley, Nuclear receptor coregulators: modulators of pathology and therapeutic targets. Nat Rev Endocrinol, 2012. 8(10): p. 598-604. 253. Dasgupta, S., D.M. Lonard, and B.W. O'Malley, Nuclear receptor coactivators: master regulators of human health and disease. Annu Rev Med, 2014. 65: p. 279- 92. 254. McBryan, J., et al., Metastatic progression with resistance to aromatase inhibitors is driven by the steroid receptor coactivator SRC-1. Cancer Res, 2012. 72(2): p. 548-59. 255. Routledge, E.J., et al., Differential effects of xenoestrogens on coactivator recruitment by estrogen receptor (ER) alpha and ERbeta. J Biol Chem, 2000. 275(46): p. 35986-93. 256. Arao, Y., et al., Estrogen receptor alpha AF-2 mutation results in antagonist reversal and reveals tissue selective function of estrogen receptor modulators. Proc Natl Acad Sci U S A, 2011. 108(36): p. 14986-91. 257. Danielian, P.S., et al., Identification of a conserved region required for hormone dependent transcriptional activation by steroid hormone receptors. Embo j, 1992. 11(3): p. 1025-33. 258. Graham, J.D., et al., Nuclear receptor conformation, coregulators, and tamoxifen- resistant breast cancer. Steroids, 2000. 65(10-11): p. 579-84. 259. Mann, M., Functional and quantitative proteomics using SILAC. Nat Rev Mol Cell Biol, 2006. 7(12): p. 952-8. 260. Ong, S.E., et al., Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol Cell Proteomics, 2002. 1(5): p. 376-86. 261. Osborne, C.K., A. Wakeling, and R.I. Nicholson, Fulvestrant: an oestrogen receptor antagonist with a novel mechanism of action. Br J Cancer, 2004. 90 Suppl 1: p. S2-6. 262. Kundakovic, M. and F.A. Champagne, Epigenetic perspective on the developmental effects of bisphenol A. Brain Behav Immun, 2011. 25(6): p. 1084- 93. 263. Widschwendter, M. and P.A. Jones, DNA methylation and breast carcinogenesis. Oncogene, 2002. 21(35): p. 5462-82. 264. Bachmann, I.M., et al., EZH2 expression is associated with high proliferation rate and aggressive tumor subgroups in cutaneous melanoma and cancers of the endometrium, prostate, and breast. J Clin Oncol, 2006. 24(2): p. 268-73. 265. Ding, L., et al., Identification of EZH2 as a molecular marker for a precancerous state in morphologically normal breast tissues. Cancer Res, 2006. 66(8): p. 4095- 9. 266. Ehrlich, M., DNA hypomethylation in cancer cells. Epigenomics, 2009. 1(2): p. 239-59.

160

267. Soares, J., et al., Global DNA hypomethylation in breast carcinoma: correlation with prognostic factors and tumor progression. Cancer, 1999. 85(1): p. 112-8. 268. Alexander, C.M., et al., Wnt signaling in mammary glands: plastic cell fates and combinatorial signaling. Cold Spring Harb Perspect Biol, 2012. 4(10). 269. Brind'Amour, J., et al., An ultra-low-input native ChIP-seq protocol for genome- wide profiling of rare cell populations. Nat Commun, 2015. 6: p. 6033. 270. Hitti, M., Stores to pull bisphenol A baby bottles. 2008: WebMD. 271. Bisphenol A. Available from: http://www.bisphenol-a.org. 272. Rochester, J.R. and A.L. Bolden, Bisphenol S and F: A Systematic Review and Comparison of the Hormonal Activity of Bisphenol A Substitutes. Environ Health Perspect, 2015. 123(7): p. 643-50. 273. Le Magueresse-Battistoni, B., et al., Endocrine disrupting chemicals in mixture and obesity, diabetes and related metabolic disorders. World J Biol Chem, 2017. 8(2): p. 108-119. 274. Silva, E., N. Rajapakse, and A. Kortenkamp, Something from "nothing"--eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ Sci Technol, 2002. 36(8): p. 1751-6.

161