Inter-Individual Variation in Response to Estrogen and Implications for Breast Cancer Risk

University of Massachusetts Amherst ScholarWorks@UMass Amherst

Doctoral Dissertations Dissertations and Theses

5-8-2020

INTER-INDIVIDUAL VARIATION IN RESPONSE TO ESTROGEN AND IMPLICATIONS FOR BREAST CANCER RISK

Amye L. Black University of Massachusetts Amherst

Follow this and additional works at: https://scholarworks.umass.edu/dissertations_2

Part of the Cancer Biology Commons, and the Developmental Biology Commons

Recommended Citation Black, Amye L., "INTER-INDIVIDUAL VARIATION IN RESPONSE TO ESTROGEN AND IMPLICATIONS FOR BREAST CANCER RISK" (2020). Doctoral Dissertations. 1976. https://doi.org/10.7275/gxe7-4m14 https://scholarworks.umass.edu/dissertations_2/1976

This Open Access Dissertation is brought to you for free and open access by the Dissertations and Theses at ScholarWorks@UMass Amherst. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].

INTER-INDIVIDUAL VARIATION IN RESPONSE TO ESTROGEN AND IMPLICATIONS FOR BREAST CANCER RISK

A Dissertation Presented

AMYE L. BLACK

Submitted to the Graduate School of the University of Massachusetts Amherst in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

May 2020

Molecular and Cellular Biology

INTER-INDIVIDUAL VARIATION IN RESPONSE TO ESTROGEN AND IMPLICATIONS FOR BREAST CANCER RISK

A Dissertation Presented

AMYE L. BLACK

Approved as to style and content by:

______D. Joseph Jerry, Chair

______Sallie S. Schneider, Member

______Kathleen Arcaro, Member

______Thomas Zoeller, Member

______Scott Garman, Director Molecular and Cellular Biology DEDICATION

To my family, for their never-ending support.

ACKNOWLEDGMENTS

I would like to thank my advisor, D. Joseph Jerry, for welcoming me into the lab and giving me this great opportunity to work with primary human tissue and cell models.

He has always encouraged me to push myself, and in doing so has helped me develop as a scientist. I am also grateful for the support of Karen Dunphy, whose advice has been instrumental to my growth as a scientist.

In addition, I need to thank all the Jerry lab members, past and present, who have helped me along the way including Amy Roberts, Libby Daniele, Aliza Majewski, Prabin

Dhangada Majhi, Aman Sharma, Ceren Goral, Mary Hagen, and Shakirah Ssebyala.

Through lab meeting discussions and informal conversations, you helped me develop project ideas and push forward with this work. I am especially thankful for the support of

Amy Roberts, who has helped me out in the lab more times than I can count.

I am also thankful for the guidance of my committee members Sallie Schneider,

Kathleen Arcaro, and Thomas Zoeller throughout this journey. Their support and helpful suggestions at our meetings were crucial to finishing this work. I especially thank Sallie

Schneider and her lab members Stephanie Morin, Kelly Gregory, and Jennifer Ser-

Dolansky for their collaboration over the years. I also need to thank Michele Markstein for her support during my first lab rotation and beyond.

Lastly, I would like to thank my family and friends. I would not have been able to finish this work without your support.

ABSTRACT

INTER-INDIVIDUAL VARIATION IN RESPONSE TO ESTROGEN AND IMPLICATIONS FOR BREAST CANCER RISK

MAY 2020

AMYE L. BLACK, B.S., UNIVERSITY OF CONNECTICUT

Ph.D., UNIVERSITY OF MASSACHUSETTS AMHERST

Directed by: D. Joseph Jerry

Breast cancer is the second leading cause of cancer related death among women and the most prevalent cancer type in women worldwide. Many risk factors for breast cancer are related to estrogen exposure, including early menarche, late menopause, nulliparity, hormone replacement therapy, and high serum estrogen levels. These factors all involve prolonged or high levels of estrogen exposure. However, estrogen exposure during early pregnancy lowers risk by up to 50% and high dose estrogen treatment is an effective antitumor treatment in postmenopausal women. The mechanisms behind estrogen’s paradoxical role in both contributing to and reducing breast cancer risk are currently unknown. We hypothesized that a subset of women may be more sensitive to estrogen exposure, and the following experiments were designed to test this hypothesis and determine if sensitivity to estrogen is due to estrogen receptor levels or downstream regulation of estrogen signaling using 3 human breast tissue or cell models.

Our first model used to characterize responses to estrogen treatment was the human breast ex-vivo explant model. Donor human breast explants were treated with control or estrogen for 4 days and then transcriptional and functional responses were compared between individuals. Up to 80 or 100-fold variation in estrogen receptor alpha (ERα) and estrogen receptor beta (ERβ) mRNA levels were observed in explants. Responses to

estrogen also varied among individuals, although estrogen receptor target gene expression

(AREG, PGR, and TGFβ2), progesterone receptor protein (PR) expression, proliferation, and irradiation induced apoptotic responses were not significantly correlated with estrogen receptor levels. These results indicated that there are variable responses to estrogen exposure among individuals but that these responses do not appear connected to estrogen receptor expression levels. Our second model, conditionally immortalized primary human mammary epithelial cells (ciHMEC), also demonstrated variable responses in luciferase reporter assays with estrogen and xenoestrogen treatment, though the responses were more consistent than those observed in the human breast explant model. This was likely due to transfection of saturating levels of estrogen receptor.

Lastly, we examined estrogen-induced responses in 4 inducible ERα HMEC cell lines. While all 4 lines were able to activate luciferase reporter assays and expressed ERα protein, only 3 out of 4 demonstrated a proliferation response to estrogen. Expression of estrogen receptor target genes (AREG and PGR) however, was not regulated by estrogen treatment. Conditioned growth media assays demonstrated that this proliferative response was not due to secretion of factors into the media but is instead driven by intracellular factors. As others have implicated the pioneer factors FOXA1 and GATA3 in restoring estrogen-induced responses, we examined mRNA expression of these factors in our cells and found no correlation with estrogen-induced proliferation. These results imply that while FOXA1 and GATA3 are not sufficient to induce biological responses to estrogen.

Collectively, our work demonstrated individual variation in estrogen receptor expression and estrogen and xenoestrogen-induced responses. Sensitivity to estrogen treatment was not driven by estrogen receptor expression levels. Further, our results suggest intracellular

vii

factors, such as estrogen receptor coregulators, are important for modulation of estrogen- induced biological responses.

viii

TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ...... v

ABSTRACT...... vi

LIST OF TABLES ...... xii

LIST OF FIGURES ...... xiii

CHAPTER

1. INTRODUCTION ...... 1

Estrogen Exposure and Breast Cancer Risk...... 1 Familial Breast Cancer Risk ...... 2 Estrogen Sensitivity in Rodents ...... 3 Estrogen Receptors Alpha and Beta ...... 5 Estrogen Receptor Signaling...... 7 Pioneer Factors, Coregulators, and Estrogen Receptor Signaling ...... 9 Estrogen Receptor Levels in Normal Breast Epithelium ...... 11 Human Models for Estrogenic Responses in the Normal Breast ...... 12 Hypothesis and Objectives ...... 16 2. INTER-INDIVIDUAL VARIARION IN ESTROGEN-INDUCED RESPONSES IN

HUMAN BREAST EXPLANT CULTURES ...... 24

Introduction/Rationale ...... 24 Materials and Methods ...... 27 Sample Collection ...... 27 Microarray Analysis...... 27 Explant Culture ...... 28 RT-qPCR...... 29 Hematoxylin and Eosin Staining ...... 29 Immunohistochemistry ...... 30 Immunofluorescence ...... 30 Terminal Deoxynucleotide Transferase dUTP Nick-End Labeling (TUNEL) ...... 31 Results ...... 32 Variation in Estrogen Receptor Expression ...... 32 Human Breast Explant Model ...... 33 Optimization for RT-qPCR in explants ...... 34 Transcriptional Responses to Estrogen ...... 36 Functional Responses to Estrogen ...... 38

ERα and ERβ in human breast explants...... 40 Discussion ...... 40 3. COMPARISON OF ESTROGENIC RESPONSES IN PRIMARY HMECs...... 69

Introduction/Rationale ...... 69 Materials and Methods ...... 74 Cell culture ...... 74 Conditional immortalization ...... 74 RT-qPCR...... 75 Mammosphere Extreme Limiting Dilution Assay ...... 75 ERE-Luciferase Assays ...... 76 Cell Treatments ...... 77 Immunofluorescence ...... 77 Results ...... 78 Characterization of HMECs ...... 78 Measuring Estrogenic Responses in HMECs ...... 79 BP3 and PP are Estrogen Receptor Agonists in HMECs ...... 81 Limited Variation in Responses to E2, BP3, and PP in ciHMECs ...... 82 Feeder Cell Contamination in ciHMEC cultures ...... 84 Discussion ...... 85 4. ESTROGEN-INDUCED RESPONSES IN INDUCIBLE ERα HMECs ...... 101

Introduction/Rationale ...... 101 Materials and Methods ...... 107 pIND-ESR1 Construct Design ...... 107 Cell Culture ...... 108 Generation of inducible ERα HMEC lines ...... 109 Cell Treatments ...... 109 ERE-Luciferase Assays ...... 110 Western Blot for ERα ...... 110 Alamar Blue Proliferation Assay ...... 111 Conditioned Growth Media ...... 112 RT-qPCR...... 112 Results ...... 113 pIND-ESR1 Construct is Functional in HMECs ...... 113 Generation of iERα HMEC lines ...... 113 Luciferase Reporter Assays in iERα HMECs ...... 114 ERα Protein Expression in iERα HMECs ...... 116 Estrogen-Induced Proliferation and ERα Target Gene Responses ...... 118 E2-Induced Proliferation in iERα HMECs is Due to Cell Intrinsic Factors ...... 119 FOXA1 and GATA3 Expression is Not Sufficient for ERα Induced Responses ...... 122 Discussion ...... 123 5. SUMMARY AND FUTURE ...... 168

REFERENCES ...... 176

LIST OF TABLES

Table Page

1.1: ERα expression in the human breast ...... 18

2.1: Donor information for first set of human breast explants ...... 51

2.2: List of analyses performed on first set of explant donor tissues ...... 52

2.3: RT-qPCR primer sequences...... 53

2.4: Antibody information for IHC ...... 53

2.5: E2-induced gene response does not correlate with estrogen receptor expression ...... 54

2.6: ER mRNA levels do not correlate with functional responses to E2 ...... 54

3.1: Donor information for primary human mammary epithelial cells ...... 90

3.2: RT-qPCR primers ...... 90

3.3: Keratin expression profiles of HMEC lines ...... 91

4.1: Antibody information for Western Blot Analysis...... 128

4.2: RT-qPCR primer sequences...... 128

4.3: Summary of HMEC responses ...... 129

xii

LIST OF FIGURES

Figure Page

1.1: ERα and ERβ domains and homology ...... 19

1.2: ERα and ERβ agonists ...... 19

1.3: Estrogen receptors mediate the actions of estrogen ...... 20

1.4: Pathways for estrogen receptor signaling ...... 21

1.5: Looping model of chromatin interaction in estrogen receptor signaling ...... 22

1.6: Coregulators involved in mammary gland development ...... 22

1.7: Overview of human breast epithelium ...... 23

2.1: ESR1 and ESR2 expression diversity among women...... 55

2.2: ERβ protein expression decreases with age...... 55

2.3: Overview of human breast explant model ...... 56

2.4: Explant tissue architecture at T=0 and T=7 ...... 57

2.5: Heterogenous expression of ERα protein ...... 58

2.6: Optimization of RT-qPCR with KRT18...... 58

2.7: Comparison of ESR1 expression between donors and during culture ...... 59

2.8: ESR1 and ESR2 expression in nulliparous and parous explants ...... 60

2.9: E2 induces R-loop formation in human breast explants ...... 61

2.10: AREG, PGR, and TGFβ2 expression in nulliparous and parous explants ...... 62

2.11: E2-induced target gene responses do not correlate with ER mRNA expression ...... 63

2.12: PR regulation by E2 in human breast explants ...... 64

2.13: E2-induced proliferation in human breast explants ...... 65

xiii

2.14: Irradiation induced apoptosis in human breast explants ...... 66

2.15: γH2AX foci formation in human breast explants ...... 67

2.16: Effect of estrogen receptor specific agonists PPT and ERB041 on AREG, PGR, and TGFβ2 target gene response...... 68

3.1: Overview of Conditionally Immortalized HMEC model ...... 92

3.2: ciHMEC PCR Characterization ...... 93

3.3: ciHMEC Lines Can Form Primary Mammospheres...... 94

3.4: Comparison of Lipofectamine 2000 and Fugene 6 reagents ...... 95

3.5: Increasing amounts of ESR1 DNA in ERE-luciferase assays...... 96

3.6: Benzophenone-3 and propylparaben dose response curves in HMECs ...... 97

3.7: Limited variation in response to E2, BP3, and PP in HMECs...... 98

3.8: FOXA1 and GATA3 expression in ciHMECs ...... 99

3.9: NIH 3T3 contamination in ciHMEC lines ...... 100

4.1: Inducible ESR1 expression in pINDUCER14 backbone ...... 130

4.2: GFP expression in 293T cells transfected with pIND-ESR1 ...... 131

4.3: Transient transfection of 76N Tert cells with pIND-ESR1 ...... 132

4.4: FACS of 76N Tert cells after lentiviral infection with pIND-ESR1 ...... 133

4.5: Post FACS check of 76N Tert-ESR1 ...... 134

4.6: FACS of MCF10A cells after lentiviral infection with pIND-ESR1 ...... 135

4.7: Post FACS check of MCF10A-ESR1 ...... 136

4.8: FACS of HME-CC cells after lentiviral infection with pIND-ESR1 ...... 137

4.9: Post FACS check of HME-CC-ESR1 ...... 138

4.10: FACS of ME16C2 cells after lentiviral infection with pIND-ESR1 ...... 139

4.11: Post FACS check of ME16C2-ESR1 ...... 140

xiv

4.12: Doxycycline dose response in 76N Tert-ESR1 ...... 141

4.13: Doxycycline dose response in MCF10A-ESR1 ...... 142

4.14: Doxycycline dose response in HME-CC-ESR1 ...... 143

4.15: Doxycycline dose response in ME16C2-ESR1 ...... 144

4.16: Luciferase Reporter Assays with 76N Tert-ESR1 and MCF10A-ESR1 ...... 145

4.17: Luciferase Reporter Assays with HME-CC-ESR1 and ME16C2-ESR1 ...... 146

4.18: Fold change in ERE-Luc responses in 76N Tert-ESR1 ...... 147

4.19: Fold change in ERE-Luc responses in MCF10A-ESR1 ...... 148

4.20: Fold change in ERE-Luc responses in HME-CC-ESR1 ...... 149

4.21: Fold change in ERE-Luc responses in ME16C2-ESR1 ...... 150

4.22: ERα antibody optimization for Western blot analysis ...... 151

4.23: β actin expression during ERα antibody optimization for Western blot ...... 152

4.24: ERα protein expression in iERα HMEC lines compared to MCF7 ...... 153

4.25: Quantification of ERα protein expression in 76N Tert-ESR1 ...... 154

4.26: Quantification of ERα protein expression in MCF10A-ESR1 ...... 155

4.27: Quantification of ERα protein expression in HME-CC-ESR1 ...... 156

4.28: Quantification of ERα protein expression in ME16C2-ESR1 ...... 157

4.29: E2-induced proliferation in 76N Tert-ESR1 and MCF10A-ESR1 HMECs ... 158

4.30: E2-induced proliferation in HME-CC-ESR1 and ME16C2-ESR1 HMECs ...... 159

4.31: AREG expression in iERα HMEC lines...... 160

4.32: Overview of conditioned media assay ...... 161

4.33: Conditioned Growth Media Assay in 76N Tert ...... 162

4.34: Expansion of Conditioned Growth Media Assay in 76N Tert...... 163

4.35: Conditioned Growth Media Assay in MCF10A ...... 164

4.36: Conditioned Growth Media Assay in T47D with 76N Tert-ESR1 media ..... 165

4.37: Conditioned Growth Media Assay in T47D with MCF10A-ESR1 media..... 166

4.38: FOXA1 and GATA3 expression in HMEC lines ...... 167

xvi

CHAPTER 1

INTRODUCTION

Estrogen Exposure and Breast Cancer Risk

Breast cancer is the most prevalent cancer type among women worldwide and the second leading cause of cancer related death in women (Siegel et al., 2019). Known risk factors for breast cancer include age, genetics, obesity, increased breast density, early onset of menarche, late menopause, nulliparity, and late full-term pregnancy (Travis and Key,

2003; Samavat and Kurzer, 2015; Brooks et al., 2018). Many of these risk factors are related to endogenous estrogen levels. Chronic lifetime exposure to estrogen, through early menarche (RR = 1.3), late menopause (RR = 1.2-2.0), hormone replacement therapy with estrogen and progesterone (RR = 1.2), or having the highest quartile of serum estrogen levels (RR = 1.8-5.0) all increase breast cancer risk in women (Brinton et al., 1983, 1988;

Trichopoulos et al., 1972, 1983; Clemons and Goss, 2001; Rossouw et al., 2002; Eliassen et al., 2006; Dall and Britt, 2017). These reproductive risk factor suggest that longer lifetime exposure to estrogen increases breast cancer risk. Accordingly, treatment with selective estrogen receptor modulators (SERMs) to inhibit estrogen activity in the breast has been shown to lower breast cancer incidence (Cuzick et al., 2013). Estrogen is also known to contribute to breast tumor initiation and growth (Russo and Russo, 2006; Tian et al., 2018).

Strikingly, while a full-term pregnancy after the age of 35 increases breast cancer risk, a full-term pregnancy before the age of 20 decreases breast cancer risk by up to 50%

(MacMahon et al., 1970; Trichopoulos et al., 1983; Rosner et al., 1994; Lambe et al., 1996).

The effectiveness of parity in reducing breast cancer risk after an early pregnancy is

mimicked in mice with estrogen and progesterone treatment (Thordarson et al., 1995;

Sivaraman, 1998; Guzman et al., 1999; Rajkumar et al., 2007; Dunphy et al., 2008), further illustrating that estrogen can reduce breast cancer risk. High dose estrogen treatment is an effective antitumor therapy in postmenopausal women with breast cancer, possibly due to increased sensitivity to estrogen after the decrease estrogen levels during menopause

(Lonning et al., 2001; Coelingh-Bennink et al., 2017). Treatment of breast tumors with

SERMs may also sensitize tumors to estrogen-induced apoptosis (Yao et al., 2000; Song et al., 2001). This evidence suggests that high levels of estrogen exposure, depending on context, have antagonistic effects on breast cancer risk.

Familial Breast Cancer Risk

Familial breast cancer risk, or genetic susceptibility to breast cancer, is another major breast cancer risk factor. Genetic susceptibility to breast cancer varies widely among individuals. Twin studies suggest that hereditary factors account for approximately 27-

30% of overall breast cancer risk (Peto and Mack, 2000; Southey et al., 2013; Mucci et al.,

2016). High penetrance BRCA1, BRCA2, and TP53 mutations account for 16-20% of inherited risk, but are rare in the general population (Peto et al., 1999; Anglian Breast

Cancer Study Group, 2000; Thompson and Easton, 2004; Antoniou and Easton, 2006;

Lalloo and Evans, 2012). Moderate penetrance mutations (CHEK2, ATM, BRIP1, and

PALB2) make up approximately 3-5% of inherited breast cancer risk (Peto et al., 1999;

Anglian Breast Cancer Study Group, 2000; Thompson and Easton, 2004; Rahman et al.,

2007). These mutations are more common than BRCA1, BRCA2, or TP53 mutations but still occur in less than 1% of the population (Adank et al., 2011; Renwick et al., 2006; Seal

et al., 2006; Rahman et al., 2007). Genome wide association studies (GWAS) have also identified low penetrance mutations (CASP8, FGFR2, TOX3, MAP3K1, and LSP1) and single nucleotide polymorphisms (Easton et al., 2007; Ahmed et al., 2009; Kuchenbaecker et al., 2014; Michailidou et al., 2017) which are estimated to account for 8-15% of inherited breast cancer risk. However, the mechanisms of these sites identified through GWAS have are still unclear. These known inherited risk factors account for approximately 30-40% of inherited breast cancer, but the remaining 60-70% are due to currently unknown inheritable factors.

Estrogen Sensitivity in Rodents

Previous work has identified rodent strains with increased sensitivity or resistance to mammary tumor development in response to carcinogen or estrogen treatment. These rodent strains do not contain any previously known breast cancer risk genes, such as

BRCA1 or BRCA2. In DMBA induced models of mammary tumors, the Wistar-Kyoto rat strain is resistant to tumor development while the Wistar-Furth strain is uniquely susceptible (Gould, 1986; Lan et al., 2001). For estrogen-induced mammary tumor models, the ACI strain is uniquely susceptible to estrogen-induced mammary tumors while the

Copenhagen and Brown Norway strains are resistant (Shull et al., 1997; Spady et al., 1998;

Dennison et al., 2015; Shull et al., 2018). Treatment of susceptible ACI rats with tamoxifen reduces the development of estrogen-induced mammary tumors (Li et al., 2002; Singh et al., 2011), demonstrating the involvement of estrogen receptor signaling in mammary tumor development in the ACI strain. These studies have also identified quantitative trait loci (QTL), regions of the chromosomes linked to mammary tumor susceptibility and

resistance (Lan et al., 2001; Haag et al., 2003; Gould et al., 2004; Schaffer et al., 2006).

Several of these QTLs have been mapped to human orthologs previously identified in

GWAS for breast cancer risk loci (Schaffer et al., 2013; Colletti et al., 2014). Additionally, deletion of the mouse ortholog to a human breast cancer susceptibility locus, 8q24, reduces mammary tumor incidence in rats (Homer-Bouthiette et al., 2018). Taken together, these results illustrate the potential for estrogen to contribute to breast cancer development in certain genetic backgrounds, suggesting unique sensitivity to estrogen exposure.

In mice, BALB/c and C57BL/6 strains are also known to differ in susceptibility to estrogen-induced mammary tumor formation but also in other estrogen-induced responses in the mammary gland, including proliferation and apoptosis (Montero Girard et al., 2007;

Aupperlee et al., 2009). As estrogen is known to promote breast cancer risk in humans, we propose that some women may be uniquely susceptible to this exposure based on their genetic backgrounds, like the rodent strains mentioned previously. Further, some women may also be more sensitive to estrogen exposure, which may be linked to increased breast cancer risk. Increased breast density is a known risk factor for breast cancer which is positively correlated with higher serum estrogen levels (Brooks et al., 2018) and may indicate estrogen sensitivity. Specific SNPs in the ESR1 gene are associated with increased breast cancer risk, resistance to anti-estrogen treatment, and poor overall prognosis

(Schubert et al., 1999; Conway et al., 2005; Herynk et al., 2007; Abbasi et al., 2012).

Included in this list of SNPs with increased breast cancer risk are 908A/G (K303R) and

1608T/A (Y537N) ESR1 polymorphisms, which produce estrogen receptor protein that is hypersensitive to estrogen treatment or constitutively active (Conway et al., 2007;

Jeselsohn et al., 2014). Based on this data, if a subset of women is uniquely sensitive to

estrogen exposure because of their genetic background, it may be connected to increased breast cancer risk.

Estrogen Receptors Alpha and Beta

Biological responses to estrogen exposure are mediated through the two estrogen receptors: estrogen receptor alpha (ERα) and estrogen receptor beta (ERβ) (Heldring et al.,

2007; Nilsson et al., 2001). ERα and ERβ are ligand activated transcription factors in the steroid nuclear receptor family encoded by ESR1 and ESR2 respectively (Heldring et al.,

2007). ESR1 is located on chromosome 6q24-27 and ESR2 is on chromosome 14q22-24.

Both ERα and ERβ contain similar protein domains: an activation function (AF-1) domain,

DNA binding domain, a ligand binding domain, and an AF-2 domain (Figure 1.1) (Nilsson et al., 2001; Leitman et al., 2010). The AF-1 and AF-2 domains are both known to be involved in gene regulation (Kumar and Thompson, 1999) and binding of estrogen receptor coregulators (Webb et al., 1998; Tremblay et al., 1999; Benecke et al., 2000; Maggi, 2011).

The DNA binding domain of ERα and ERβ is required for the binding of these receptors to DNA, at estrogen response elements (EREs), to initiate estrogen-responsive gene transcription (Gruber et al., 2002; Heldring et al., 2007; Yasar et al., 2017). In mice with a mutated ERα DNA binding domain, mammary gland development is impaired, and uterine and mammary gland phenotypes resemble those observed in the full ERα knockout mice (Lubahn et al., 1993; Bocchinfuso and Korach, 1997; Ahlbory-Dieker et al., 2009), suggesting that DNA binding is essential to ERα function. The DNA binding domains of

ERα and ERβ are 97% homologous, indicating that they likely bind to many of the same regions of the DNA to control transcription. However, ERβ has lower affinity for EREs

than ERα and lower transactivation of ERE-reporters in luciferase assays with estrogen as a ligand (Hall and McDonnell, 1999; Yi et al., 2002).

The ligand binding domain is crucial for estrogen receptor function. 17β-estradiol

(E2) is the most prevalent estrogen in women (Gruber et al., 2002; Mukherjee et al., 2005) and acts as a ligand for both ERα and ERβ equally. Ligand binding to estrogen receptors initiates a conformational change, which allows for receptor dimers to form (Kumar et al.,

2011; Vrtačnik et al., 2014). These receptors can form homodimers or heterodimers. The ligand binding domain of ERα and ERβ are 55% homologous (Witkowska et al., 1997;

Gruber et al., 2004; Leitman et al., 2010), suggesting that it is possible for these receptors to be activated by different ligands. Accordingly, several estrogen receptor specific agonists exist (Figure 1.2). Propyl pyrazole triol (PPT) is an ERα specific agonist, with

410-fold selectivity for ERα over ERβ (Stauffer et al., 2000). Multiple ERβ have been developed, including ethenyl-2-(3-fluoro-4-hydroxyphenyl)-benzoxazolol (ERB041) and diarylpropionitrile (DPN); ERB041 has 200-fold selectivity for ERβ over ERα and DPN has 170-fold selectivity for ERβ (Harris et al., 2003; Meyers et al., 2001).

One class of endocrine disrupting chemicals, xenoestrogens, can also act as ligands for the estrogen receptors. Endocrine disrupting chemicals can disrupt normal hormone signaling by mimicking or antagonizing the effects of endogenous hormones and also by disrupting the synthesis and metabolism of endogenous hormones and their receptors

(Sonnenschein and Soto, 1998; Fernandez and Russo, 2010). Xenoestrogens have estrogenic activity, suggesting the potential to activate estrogen signaling in the breast

(Singleton and Khan, 2003). They are common in food, drinking water, and personal care products such as lotions, cosmetics, and sunscreens (Mortensen et al., 2014; Soni et al.,

2001). Xenoestrogens, therefore, have the potential to disrupt estrogen receptor signaling by acting as ligands for ERα and/or ERβ and could affect the balance between the protective and risk promoting effects of estrogen exposure on breast cancer risk.

Estrogen Receptor Signaling

Classical estrogen receptor signaling involves direct binding of ligand-bound estrogen receptor dimers to EREs, which then recruit coactivators or corepressors to control target gene transcription (Figure 1.3). Other pathways for estrogen signaling have been identified, including tethered, non-genomic, and ligand-independent pathways (Figure

1.4) (Heldring et al., 2007). The tethered pathway involves ligand-bound estrogen receptor dimer binding to other transcription factors bound to DNA, instead of EREs (Gaub et al.,

1990; Saville et al., 2000). Non-genomic estrogen receptor signaling involves either membrane bound estrogen receptors or other membrane bound receptors which in turn activate the estrogen receptors (Heldring et al., 2007; Levin, 2009). Membrane bound estrogen receptors are not sufficient to rescue mammary gland development in ERα knockout mice (Pedram et al., 2009), however, and thus does not likely play a significant biological role. Ligand-independent pathways involve estrogen receptor phosphorylation by kinases activated by growth factor signaling and are hypothesized to be involved in hormone-independent growth of breast tumors (Kato et al., 1995; Coutts and Murphy,

1998; Shim et al., 2000). More recent work has highlighted the increased complexity of classical estrogen receptor signaling, where estrogen receptors can control the expression of genes without EREs or other transcription factor binding sites through long range

chromatin looping, which brings distant genes into close proximity to where estrogen receptors are bound (Figure 1.5) (Fullwood et al., 2009).

Data from estrogen receptor knockout mice suggest that ERα function is crucial to mammary gland development, while ERβ is not. ERα knockout mice (αERKO) have rudimentary ductal mammary gland development similar to newborn mice (Bocchinfuso and Korach, 1997). Conversely, ERβ knockout mice (βERKO) mammary glands are indistinguishable from wild type mice (Krege et al., 1998). Therefore, the function of ERβ in the breast has long been questioned and is still partially unanswered. It is known that

ERβ, including its multiple isoforms, can form heterodimers with ERα (Moore et al., 1998).

The question, therefore, was whether ERβ could act as a modulator of ERα activity. Early work demonstrated that ERβ was able to act as a dominant inhibitor of ERα activity (Hall and McDonnell, 1999). Later studies confirmed the ability of ERβ to repress estrogen mediated proliferation in breast cancer cell lines (Williams et al., 2008), ERα induced activation of target genes pS2 and PGR, promoter binding, and recruitment of cofactors

(Matthews et al., 2006). ERβ is thought to block estrogen-induced proliferation through activation of MAPK and PI3K signaling (Cotrim et al., 2013) or in combination with p53

(Bado et al., 2017, 2018). However, while this work was done with the main ERβ isoform

(ERβ-1), other work has demonstrated a truncated ERβ isoform (ERβ-2, or ERβcx) is also capable of repressing ERα activity and coactivator recruitment while not influencing ERβ-

1 function (Ogawa et al., 1998). Accordingly, ERβ loss during breast cancer progression is associated with poorer prognosis (Omoto et al., 2001; Sugiura et al., 2007), resistance to tamoxifen treatment (Hopp, 2004), and increased risk of advanced tumor progression

(Roger et al., 2001; Shaaban et al., 2003; Skliris et al., 2003). Contradictory results are

obtained in other studies, however, which show poor prognosis with ERβ expression

(Novelli et al., 2008; Shaaban et al., 2008) even in ERα positive breast tumors (Saji et al.,

2002). For these reasons the exact function of ERβ is still unclear, especially regarding breast cancer development. It is accepted, though, that the role of ERα can potentially be modulated by ERβ and the ratios of ERα to ERβ may be important for determining the effects of estrogen treatment within a cell.

Pioneer Factors, Coregulators, and Estrogen Receptor Signaling

Pioneer factors and coregulators are essential determinants of estrogen receptor signaling and mammary gland development. Pioneer factors are transcription factors which can bind to specific sites in the chromatin and recruit other transcription factors or histone modifying enzymes to enhance or restrict transcription (Zaret and Carroll, 2011).

Forkhead box A1 (FOXA1) and GATA binding protein 3 (GATA3) are considered estrogen receptor pioneer factors, and FOXA1, GATA3, and ERα interaction is thought to maintain luminal epithelial cell phenotype in the mammary gland. FOXA1 regulates expression of ERα, terminal end bud formation, and ductal outgrowth during puberty

(Bernardo et al., 2010). GATA3 is also involved in terminal end bud formation as well as luminal cell differentiation and loss of GATA3 results in expansion of luminal progenitors

(Kouros-Mehr et al., 2006; Asselin-Labat et al., 2007). GATA3 and ERα have a cross- regulatory feedback loop and GATA3 is crucial for estrogen-induced growth in breast cancer cells (Eeckhoute et al., 2007). Attempts to generate ER negative breast cell lines expressing exogenous ESR1 have shown that FOXA1 and GATA3, along with ESR1, are

necessary for recovery of estrogen-induced phenotypes, including proliferation (Kong et al., 2011).

Estrogen receptor coregulators interact with FOXA1 and GATA3 pioneer factors as well as ERα and ERβ to modify estrogen receptor signaling. Many of these coregulators are important for different developmental stages of mammary gland development (Figure

1.6). Coactivators typically facilitate the recruitment of transcriptional machinery to the receptors while corepressors repress this recruitment, both acting through binding with the

AF-1 and AF-2 domains of the estrogen receptors (Heery et al., 1997; Nilsson et al., 2001;

Heldring et al., 2007). A total of over 200 estrogen receptor coregulators are known, most identified through interactions with ERα. Fewer coregulators have been examined for ERβ.

Known coactivators for the estrogen receptors include SRC1-3, BRCA1 and BRCA2,

CITED-1, Ccdn1, NCOA1-3, TIF2, PELP1, CBP, p300, CARM1, RIP140, p/CIP, and

PRMT1. Known co-repressors include NCOR, SMRT, NRIP1, MTA1, NFIB, and YBX1

(Kurebayashi et al., 2000; Maggi, 2011; Merrell et al., 2011; Jiang et al., 2013; Vrtačnik et al., 2014; Campbell et al. 2018).

Though affinity for estrogen is equal for both receptors, there is differential recruitment of coregulators by ERα and ERβ, and this can be affected by phosphorylation of these receptors (Nguyen et al., 2012; Tharun et al., 2015). In MCF7 cells, recruitment of SRC3 and RIP140 to the PGR gene is higher in cells expressing ERα, slightly lower in

ERα and ERβ expressing cells, and lowest in cells with ERβ only (Jiang et al., 2013).

Similar results were obtained for other ER target genes FOS, TFF1, and CASP7 (Chang et al., 2008). Recruitment of coregulators by estrogen receptors also differs for target genes that are up or down regulated. pS2, an ER target gene upregulated by E2 treatment in

MCF7 cells, shows increased recruitment of SRC1 and p/CIP coactivators and decreased recruitment of NCoR, NRIP1, and SMRT corepressors; the opposite results are observed for ER target genes that are downregulated (PSCA, SLC35A1) (Merrell et al., 2011). These changes are associated with histone deacetylation and RNA polymerase II docking.

Previous studies have demonstrated tissue specific expression of nuclear receptor coregulators and are hormonally regulated (Misiti et al., 1998; Shang and Brown, 2002;

Molenda et al., 2003). These differences have also been suggested to influence the tissue specific antagonist/agonist actions of tamoxifen (Smith et al., 1997; Shang and Brown,

2002). Knockdown of NCOA3 was shown in MCF7 cells to prevent estrogen-induced proliferation, but knockdown of other similar coactivators NCOA1 and NCOA2 had no such effect (Karmakar et al., 2009). Overexpression of CARM1 in MCF7 cells also results in inhibition of estrogen-induced proliferation and target gene regulation (Al-Dhaheri et al., 2011). Expression of NFIB and YBX1 can inhibit luciferase reporter activity and change target gene expression in ERα positive MCF7 breast cancer cells as well as promote estrogen independent growth (Campbell et al., 2018). Because expression of these coactivators and corepressors are known to have cell/tissue specific expression and can regulate estrogen receptor signaling, including modulation of proliferation, we hypothesize that individual expression of these coregulators could influence sensitivity to estrogen.

Estrogen Receptor Levels in Normal Breast Epithelium

ERα and ERβ are both expressed in the epithelial cells of the human breast (Figure

1.7). ERα is restricted to the luminal epithelial cells of the ducts, whereas ERβ is expressed in basal epithelial, luminal epithelial, stromal, lymphocytes, and endothelial cells of the

breast (Speirs 2002). Previous studies of ERβ expression in the normal human breast are limited by the availability of reliable ERβ antibodies (Andersson et al., 2017; Nelson et al.,

2017). Much work, however, has been done to quantify the ERα protein levels in the normal human breast. A summary of these studies is shown in Table 1.1. ERα protein is detectable in approximately 32-68% of human samples and is known to increase with age

(Battersby et al., 1992; Söderqvist et al., 1993; Gabrielson et al., 2016). The general consensus is that ERα positive cells make up no more than 5-10% of the normal human breast luminal epithelial population, in contrast to mouse epithelium where 15-40% of luminal epithelial cells may be ERα positive depending on strain (Montero Girard et al.,

2007; Raafat et al., 2012). The proportion of ERα positive luminal epithelial cell is also known to increase in premalignant breast lesions and in the majority of breast cancers

(Shoker et al., 1999a; Quong et al., 2002). Interestingly, it is thought that, although ERα positive cells respond to estrogen treatment by releasing growth factors that stimulate proliferation, only the ERα negative cell population responds by proliferating in the normal breast. This is referred to as the paracrine hypothesis for estrogen mediated proliferation

(Shoker et al., 1999b; Mallepell et al., 2006; Ciarloni et al., 2007). In breast cancer cells this is perturbed, as multiple ERα breast cancer cell lines increase proliferation when treated with estrogen.

Human Models for Estrogenic Responses in the Normal Breast

The original studies determining responses to estrogen exposure in the normal human breast were based on population studies using human breast tissue at different stages of the menstrual cycle or at different life stages. One large drawback to these studies is

that the levels of endogenous estrogen vary among women and responses to estrogen treatment may vary as well. To overcome this obstacle, 2D human breast epithelial cell models have been developed. Breast cancer cell lines, such as MCF7 and T47D, express high ERα levels and have been useful for studying the effects of estrogen receptor signaling in a breast cancer setting. They are not, however, an appropriate model for estrogen receptor signaling in the normal human breast due to many perturbations that occur during carcinogenesis and the heterogenous expression of ERα in normal human breast epithelium. Induced pluripotent stem cell lines have been used in the study of tissue development and function, but these cell lines are difficult to differentiate into breast epithelial cells (Cravero et al., 2015). Recently, improved protocols were developed to differentiate induced pluripotent stem cells into breast epithelial cells, but the widespread application of this technique is yet to be seen (Qu et al., 2017).

After optimization of culture conditions, it became possible to propagate normal human breast epithelial cells in 2D culture which now serve as an alternative to breast cancer cell lines. Early work with human mammary epithelial cells (HMECs) was hindered by early senescence. Methods to bypass this senescence were developed, including transformation with HPV E6 and E7, human telomerase (TERT), Myc, SV40 large T- antigen, zinc-finger protein ZNF217, and BMI-1. However, abnormal cell cycle progression in these immortalized cells is attributed to loss of p53, retinoblastoma protein, and p16INK4a, which perturb the underlying biology of these cells (Wang et al., 1998;

Dyson et al., 1989; Münger et al., 1989; Scheffner et al., 1990; Huschtscha et al., 2001;

Nonet et al., 2001; Dimri et al., 2002; Martinez-Zapien et al., 2016; Fischer et al., 2017).

These immortalized lines also can exhibit signs of transformation, including anchorage

independent growth, reduced apoptotic responses, and altered gene expression (Wang et al., 1998; Thibodeaux et al., 2009). Despite these changes, stably immortalized HMECs are still one of the better options for studying normal breast biology.

An alternative approach to stably immortalization called conditional reprogramming, or conditional immortalization, allows the expansion of HMECs through coculture with irradiated mouse fibroblasts and addition of a rho-associated protein kinase

(ROCK) inhibitor (Liu et al., 2012; Palechor-Ceron et al., 2013). With this method, removal of feeder cells and ROCK inhibitor results in removal of the conditionally reprogrammed phenotype. These reprogrammed cells are suggested to represent a stem cell like epithelial population (Suprynowicz et al., 2012, 2017), though it has been noted that HMEC cells in other culture methods can also lose expression of differentiation markers (Breindel et al., 2017). Comparison of these conditionally immortalized HMECs, or ciHMECs, isolated from breast tumors shows similar expression profiles to the original tissue sample (Mahajan et al., 2017). Some groups have successfully used this method to expand ERα positive luminal epithelial cells, but this population decreases in culture over passaging (Jin et al., 2017).

Although conditional immortalization and stable immortalization methods have enhanced the expansion of HMECS, one key issue that remains is the loss of estrogen receptor positive cells in 2D culture conditions. Multiple groups have reported the loss of

ERα expressing luminal epithelial cell populations during passaging (Fridriksdottir et al.,

2015; Jin et al., 2017; Lee et al., 2018), and it is known that luminal epithelial cells only make up a small population (10-25%) of epithelial cells isolated from human breast tissue

(Garbe et al., 2009). This population may be greater in breast tissue of older women (Lee

et al., 2015). As only 5-10% of luminal epithelial cells express ERα, it is not surprising that most epithelial cells isolated from breast tissue lack ERα expression. There are multiple hypotheses explaining why ERα positive luminal epithelial cells are difficult to grow in culture. It is thought that the luminal epithelial cell population is lost in 2D culture due to reduced adherence (Soule and McGrath, 1986; Keller et al., 2010), faster growth of basal epithelial cells (Keller et al., 2010; Fridriksdottir et al., 2015), triggering stress responses in luminal epithelial cells (Lee et al., 2018), changes to the microenvironment

(Miyano et al., 2017), or epigenetic reprogramming that results in the loss of the differentiated phenotype (Keller et al., 2010; Breindel et al., 2017). Calcium levels may also be important as one group noted the growth of free-floating cells in low calcium containing media, though these cells are thought to be less differentiated as they can form ductal structures when placed in collagen gels (Soule and McGrath, 1986). It has been noted that luminal epithelial cells grow better in serum containing media, as opposed to basal epithelial cells which grow in serum free media (Kao et al., 1995), though growth in serum containing media can lead to permanent senescence (Stampfer and Bartley, 1985).

Luminal epithelilal cells may be more sensitive to microenvironment changes and grow better in mixed cell type populations (Miyano et al., 2017). Even with optimal media conditions, ERα positive luminal epithelial cells are lost in early passages and the number of passages is ultimately limited by the original percentage (Lee et al., 2018).

More recent work has focused on the creation of 3D breast organoid cultures from primary HMECs of normal and cancerous breast tissue (Sokol et al., 2016; Sachs et al.,

2018; Djomehri et al., 2019; Florian et al., 2019; Goldhammer et al., 2019). Breast organoids contain both luminal and basal epithelial cells, retain the ERα expressing

population, and are responsive to hormone treatment. Our lab has found, however, that propagation of breasts organoids from normal human breast tissue is limited, as proliferation ceases after a few passages. We have had more success with human breast explant ex-vivo models, which can be maintained in culture for up to 4 weeks and retain all breast cell types (Zhuang et al., 2003; Eigeliene et al., 2006, 2008, 2012, 2016; Tanos et al., 2013; Dunphy et al., 2020). The ERα expressing population is maintained in this model as well as response to hormone treatment. Studies with these tissues, however, are limited by the amount of available breast tissue from donors. Ultimately, all these breast tissue and cell models remove some of the confounding factors of individual hormone levels when a single breast biopsy is taken, and each have unique advantages and disadvantages.

Hypothesis and Objectives

We hypothesize that a subset of women is uniquely sensitive to estrogen exposure, and thus may be at increased risk for breast cancer. This increased sensitivity may be due to differing levels of estrogen receptors, parity, or differing estrogen receptor coregulator expression. First, a human breast explant model is utilized to determine the level of variation in estrogen receptor expression among women, characterize transcriptional and functional responses to estrogen, and determine if these are correlated with estrogen receptor expression. Next, conditionally immortalized primary human mammary epithelial cells are explored as a model for capturing variation in estrogen sensitivity, as well as determining if xenoestrogen sensitivity also varies among individuals. Lastly, inducible

ERα human mammary epithelial cell lines are developed and utilized to probe estrogen

receptor induced activation of biological responses and differences in response between normal and breast cancer cells. If our hypothesis is correct, it has the potential to provide additional insight into the paradoxical roles of estrogen in breast cancer risk and suggests that this subset of individuals may be more susceptible to breast cancer through increased endogenous estrogen exposure. The models used in this work can be utilized in future studies to further identify factors mediating estrogen sensitivity.

Table 1.1: ERα expression in the human breast Literature review of ERα positivity in human breast epithelium.

Antibody First Author Year % ERα positive cells Dilution Used ER-ICA kit, Petersen 1987 7% Abbot Not Stated Laboratories Ricketts 1991 0-20%, but up to 65% H222 Not Stated ER-ICA kit, Battersby 1992 38-47% had detectable protein Abbot Not Stated Laboratories 32-68% had detectable protein, ER-ICA kit, Soderqvist 1993 5% in luteal phase, 18% in Abbot Not Stated follicular phase Laboratories Unspecified, ER-ICA kit, Khan 1994 57% of patients were positive Abbot Not Stated (defined as >50%) Laboratories Clarke 1997 10-15% 1D5, Dako UK 1:75 Khan 1998 7.7% H222 Not Stated Unclear Anderson 1998 12-13% Not Stated (possibly 1D5) 14% in Australian women, Lawson 1999 1D5, Dako UK Not Stated 9% in Japanese women 11% in < 46 years old, Shoker 1999 27% in 46-55 years old, 1D5, Dako UK Not Stated 34% in > 55 years old 6.8% in premenopausal, Shoker 1999 42% in postmenopausal (large 1D5, Dako UK 1:75 variation) 7% in Lob 1, 4% in Lob 2, 1% ER1D5, Russo 1999 1:400 in Lob 3 Amac Lab Khan 2002 5-20%, but up to 40% 6F11 1:100 67% of patients had ER, 10% 1D5, Dako Eigeliene 2006 positive after culturing explants 1:2000 Denmark (7 days) 21% in premenopausal, 1D5, Dako Eigeliene 2008 1:2000 17.3% in postmenopausal Denmark 1D5, Dako Eigeliene 2016 20% (in explant culture) 1:400 Denmark Detected protein in 61% of 1D5, Dako Gabrielson 2016 samples, mean 30.4% (range 4- 1:60 Sweden 56.8%), women 41-76 years old Premenopausal women: 0-50% F-10, Santa Chamberlin 2017 Postmenopausal women: 10- 1:200 Cruz (sc-8002) 45%

Figure 1.1: ERα and ERβ domains and homology Structural domains of Estrogen Receptor alpha (ERα) and Estrogen Receptor beta (ERβ). The DNA binding domains (C) are 97% homologous, indicating these receptors bind to the same regions of DNA. The ligand binding domain (E) is only 55% homologous, which allow for differences in ligand specificity between these two receptors. Adapted from Leitman et al. 2010, Curr Opin Pharmacol.

Figure 1.2: ERα and ERβ agonists 17β-estradiol is an estrogen receptor agonist with equal affinity for both ERα and ERβ. Propyl-pyrazole-triol (PPT) is a selective ERα agonist, while diarylpropionitrile (DPN) and ERB041 are both selective ERβ agonists. Xenoestrogens benzophenone-3 (BP3) and propylparaben (PP) are potential estrogen receptor agonists. Selectivity is indicated.

a b

Figure 1.3: Estrogen receptors mediate the actions of estrogen Canonical estrogen receptor signaling overview. (a) 17β-estradiol (E2) binds to ERα and/or ERβ in the cell nucleus, which triggers receptor dimerization. Homodimers or heterodimers then bind to estrogen response element (ERE) sequences in the DNA to trigger transcription of estrogen receptor target genes, such as AREG and PGR. (b) Estrogen receptor recruitment of transcriptional machinery to EREs depends on the presence of coactivators (such as FOXA1 and GATA3) or corepressors (example not shown).

Figure 1.4: Pathways for estrogen receptor signaling Overview of ligand-dependent and independent mechanisms of estrogen receptor signaling. Direct, ligand-dependent signaling involves ligand-bound estrogen receptor dimers binding to estrogen response elements in the DNA. Tethered estrogen receptor signaling involves ligand-bound receptor dimers binding to other transcription factors on the DNA, such as AP-1 or SP-1. Nongenomic estrogen receptor signaling is thought to involve membrane bound estrogen receptors or other membrane bound receptor which activates the estrogen receptors. Ligand-independent signaling involves other growth factor activation, which activates kinases and leads to phosphorylation of the estrogen receptors. Adapted from Heldring et al. 2007, Physiological Reviews.

Figure 1.5: Looping model of chromatin interaction in estrogen receptor signaling Proposed model of ERα signaling through chromatin interactions. Chromatin looping brings genes together for ERα induced transcriptional regulation. Genes near the anchored chromatin are more highly expressed, due to close proximity to transcriptional machinery. Genes far from the anchor site are expressed at lower levels. Adapted from Fullwood et al. 2009, Nature.

Figure 1.6: Coregulators involved in mammary gland development Overview of several coregulators of estrogen receptor signaling known to be required for each stage of mammary gland development, including pioneer factors FOXA1 and GATA3. Adapted from Manavathi et al. 2014, Frontiers in Cell and Developmental Biology.

Figure 1.7: Overview of human breast epithelium The breast epithelium consists of ductal structures formed by layers of breast epithelial cells. Myoepithelial (basal) cells make up the outer epithelial layer of the duct while the luminal epithelial cells are the inner layer, with a lumen in the center. Surrounding the ducts are stromal cells, including fibroblasts, as well as adipose tissue. Adapted from Visvader et al. 2009, Genes and Development.

CHAPTER 2

INTER-INDIVIDUAL VARIARION IN ESTROGEN-INDUCED RESPONSES IN

HUMAN BREAST EXPLANT CULTURES

Introduction/Rationale

Epidemiological studies have identified lifestyle, environmental, and reproductive risk factors for breast cancer. Early menarche (RR = 1.3), late menopause (RR = 1.2-2.0), or having the highest quartile of serum estrogen levels (RR = 1.8 – 5.0) are all reproductive risk factors linked to lifetime estrogen exposure (Brinton et al., 1983, 1988; Clemons and

Goss, 2001; Dall and Britt, 2017). Exogenous estrogens and use of hormone replacement therapy containing both estrogen and progesterone also increases breast cancer risk (RR =

1.2)(Rossouw et al., 2002). These results suggest that longer lifetime exposure to estrogen increases breast cancer risk. Estrogen has been demonstrated to contribute to breast tumor infiltration and promote the growth of breast cancer cells (Russo and Russo, 2006; Tian et al., 2018). In addition, treatment with selective estrogen receptor modulators to inhibit estrogen activity in the breast has been shown to reduce breast cancer incidence (Cuzick et al., 2013).

However, estrogen may also reduce breast cancer risk. A full-term pregnancy after age 35 increases breast cancer risk, but a full-term pregnancy before age 20 decreases risk by up to 50% (Albrektsen et al., 2005; Lambe et al., 1996; MacMahon et al., 1970;

Trichopoulos et al., 1983). This protective effect of early parity has been replicated in rodent models with estrogen and progesterone treatment (Dunphy et al., 2008; Guzman et al., 1999; Moon, 1969; Russo and Russo, 1980; Sivaraman, 1998; Thordarson et al., 1995).

High dose estrogen treatment can be an effective antitumor therapy in postmenopausal

women with breast cancer (Lonning et al., 2001). It is hypothesized that a decrease in endogenous estrogen after menopause or treatment with selective estrogen receptor modulators may sensitize tumors to estrogen (Yao et al., 2000). Accordingly, estrogen treatment can induce apoptosis in long-term estrogen deprived breast cancer cells (Song et al., 2001). Therefore, the effects of estrogen exposure appear to have paradoxical effects on breast cancer risk.

The effects of estrogen are mediated by the two estrogen receptors: estrogen receptor alpha (ERα) and estrogen receptor beta (ERβ). Estrogen binds as a ligand to the two estrogen receptors, which form homodimers or heterodimers, and then bind to DNA to initiate target gene transcription (Gruber et al., 2002; Heldring et al., 2007b; Yasar et al.,

2017). Target gene transcription is modulated by estrogen receptor dimer combination and the recruitment of transcriptional coregulators (Chang et al., 2008). ERα expression is limited to luminal epithelial cells in the breast, while ERβ is expressed in luminal epithelial cells, basal epithelial cells, and stromal cells (Speirs et al., 2002). Expression of ERα in normal breast epithelial cells is positively correlated with breast cancer risk and age

(Gulbahce et al., 2017; Shoker et al., 1999a, 1999b; Umekita et al., 2007). The expression of ERα is known to vary widely among women (Battersby et al., 1992; Söderqvist et al.,

1993), which suggests women vary in exposure to estrogen as well as estrogen receptor expression. Estrogen mediated responses are carried out by ERα and ERβ and having more receptor may indicate increased estrogen signaling.

It is estimated that 1 in 8 women will be diagnosed with breast cancer during their lifetime. This also implies, however, that 7 out of 8 women will not develop breast cancer.

Are some women more sensitive to estrogen exposure? It is known in rodent models that

certain mouse and rat strains are more sensitive to estrogen-induced responses and/or mammary tumors. Mice strains exhibit differences in estrogen-induced proliferation, apoptosis, and mammary tumor formation (Montero Girard et al., 2007; Aupperlee et al.,

2009; Jerry et al., 2018). Copenhagen and Brown Norway rat strains are resistant to estrogen-induced mammary tumors while the ACI strain is susceptible (Dennison et al.,

2015b; Shull et al., 2001, 2018).

We hypothesized that specific individuals may be more sensitive to estrogen exposure, and that this increased sensitivity may be due to increased ERα and ERβ levels, age, parity, or differing expression of estrogen receptor signaling coregulators. To examine this, we utilized an ex-vivo human breast explant model to compare estrogen receptor levels, transcriptional responses, and functional responses in individual women. Human breast explants retain luminal and basal epithelial cells, stromal cells, and adipose cells and the interactions between those cells. They have also been shown to be maintained in culture for up to 2-4 weeks and are responsive to hormone treatment (Eigeliene et al., 2016; Tanos et al., 2013; Eigeliene et al., 2012, 2008; Eigėlienė et al., 2006; Zhuang et al., 2003). The human breast explant model also allows for the clearing of endogenous hormones, so that responses to a single consistent dose of estrogen can be measured.

We observed significant changes in estrogen receptor expression with age in breast tissue from over 100 donors. In explant donors, levels of ESR1 and ESR2 varied by more than 80-fold. Estrogen-induced target gene responses (AREG, PGR, and TGFβ2), regulation of progesterone receptor protein levels, proliferation, and irradiation-induced apoptosis were not significantly correlated with estrogen receptor levels. Parity influenced sensitivity to estrogen treatment, as nulliparous donor explants had consistent up-

regulation of estrogen mediated target genes while parous had more variable expression.

Collectively, these data suggest that estrogen receptor expression does not correlate with sensitivity to estrogen exposure. We propose that differing expression of estrogen signaling coactivators and corepressors may contribute to individual estrogen sensitivity and estrogen-induced responses.

Materials and Methods

Sample Collection

Human breast tissue samples used for microarray analysis were collected from female donors, age 14-70 years, undergoing reduction mammoplasty surgery at Baystate

Medical Center in accordance with the Institutional Review Boards at Baystate Medical

Center and the University of Massachusetts, Amherst (Troester et al., 2009; Sun et al.,

2012; Rotunno et al., 2014). Donated tissue was snap frozen in liquid nitrogen and stored at -80oC at the Pioneer Valley Life Sciences Institute (PVLSI). Donor demographics and reproductive history were recorded after tissue collection.

Human breast tissue for explants was gathered through the Rays of Hope Center for Breast Cancer Research at the PVLSI according to IRB approval #286173 and #132204.

Tissue was collected from female donors (n=31) ages 18-62 years undergoing reduction mammoplasty surgery. Donor information including age and parity status were recorded

(Table 2.1).

Microarray Analysis

Microarrays were performed at the University of North Carolina at Chapel Hill as described previously (Troester et al., 2009; Sun et al., 2012; Rotunno et al., 2014). Briefly,

100mg of frozen tissues were homogenized for RNA isolation using RNeasy kits. RNA quality and concentration were determined using an Agilent 2100 Bioanalyzer and a ND-

1000 Nanodrop spectrophotometer. Two-color 4x44K Agilent whole genome arrays were performed. Data are publicly available through the Gene Expression Omnibus

(GSE:16113 and GSE:33526). When multiple probe sets for a gene were present, we used the probe with expression values for the most donor samples. Trends were analyzed using linear regression analysis in Graphpad Prism (GraphPad Prism version 7, La Jolla, CA).

Explant Culture

See Figure 2.3 for overview of model. Fresh breast tissue was cut into approximately 1x1x4mm sections using a Stadie-Riggs microtome. For each donor, tissue sections were frozen or fixed immediately for time zero (T=0) samples. The remaining sections were placed on top of surgical foam (Ethicon) in 60mm plastic dishes and cultured for a total of 7 days. The explants were kept in a humidified incubator with 5% CO2 at

37oC. Explants were cultured in basal media for 3 days to clear endogenous hormones and then cultured in basal or E2 media for an additional 4 days. Basal culture media consisted of phenol red free DMEM:F12 (Sigma-Aldrich), 10% charcoal stripped FBS (FB-04,

Omega Scientific), 10ng/mL human EGF (21-8356-U100, Tonobo Biosciences), and antibiotic-antimycotic (15240-062, Gibco). E2 treatment media included basal media with the addition of 10nM 17β-estradiol (E2758-250MG, Sigma-Aldrich). 4,4’,4”-(4-Propyl-

[1H]-pyrazole-1,3,4-triyl)trisphenol-treatment media (PPT; ERα specific agonist) or 7-

Ethenyl-2-(3-fluoro-4-hydroxyphenyl)-5-benzoxazolol (ERB041; ERβ specific agonist) included basal media with the addition of 200nM PPT or 200nM ERB041; respectively

(Tocris, Minneapolis, MN). Half of the tissue from each donor was irradiated with 5Gy 6

hours prior to tissue collection on day 7 of culture. Half of the explant sections intended for histological endpoints were fixed in 10% NBF overnight and transferred to 70% ethanol before processing and paraffin embedding. The other half of the explant samples were snap frozen and stored at -80oC for later RNA isolation. Table 2.2 summarizes the experimental analyses performed on each individual donor sample.

RT-qPCR

RNA isolation was performed using TRIzol according to manufacturer’s recommendation (15596018, ThermoFisher Scientific). cDNA was synthesized using 2 g of RNA with the Transcriptor First Strand cDNA synthesis kit (04-379-012-001, Roche). cDNA was diluted 1:10 before use in RT-qPCR. RT-qPCR was performed in a thermocycler (Mastercyler Epgradient S model 5345, Eppendorf). See Table 2.3 for primer sequences. Statistical analysis was performed using 2-way ANOVA in Graphpad

Prism (GraphPad Prism version 7, La Jolla, CA). Expression is relative to mean control or to an inter-run calibrator (IRC) consisting of pooled cDNA from a subset of T=0 (not cultured) samples.

Hematoxylin and Eosin Staining

Hematoxylin and eosin (H&E) staining were performed on paraffin-embedded 4- micron sections. Briefly, sections were deparaffinized in xylenes and rehydrated in graded ethanols. Sections were stained with Mayer’s modified hematoxylin (S216-32OZ, Poly

Scientific) and eosin phloxine alcohol working solution (S176-32OZ, Poly Scientific).

After dehydration in graded ethanols and clearing with xylenes, slides were cover-slipped.

Images were obtained under 200x magnification (BZ-X700 microscope, Keyence, Itasca,

IL).

Immunohistochemistry

Immunohistochemistry (IHC) was performed on paraffin-embedded 4-micron sections using a DakoCytomation autostainer (Dako, Carpinteria, CA) and the antibodies listed in Table 2.4. Sections were deparaffinized in xylenes, rehydrated in graded ethanols, and rinsed in phosphate buffered saline (PBS) before antigen retrieval. For ERβ detection, antigen retrieval was performed in heated 0.01M citrate buffer for 10 minutes. Using the

Acuity polymer detection system, primary antibody incubation was performed at room temperature for 30 minutes, followed by incubation with horse-radish peroxidase (HRP) polymer and chromogen detection. For ER, PR and PCNA detection the Envision HRP

Detection system was used (Dako). Antigen retrieval was performed by heating in 0.01M citrate buffer (pH 6) for 10 minutes for PR and 20 minutes for ER and PCNA. Primary antibody incubation was performed for 30 minutes. Primary antibody information and dilutions are listed in Table 2.4. Incubation with secondary antibody (K4001 or K4003,

Dako) was performed according to manufacturer’s instructions (Dako, Carpinteria, CA).

Immunoreactivity was visualized with chromogen diaminobenzidine (DAB) incubation for

10 minutes. Sections were counterstained with hematoxylin. Images were obtained under

200x magnification (model BZ-X700 microscope). Positive cells were quantified (ImageJ software, https://imagej.nih.gov/ij/) as the percentage of positive luminal cell nuclei in a total of 600-1200 cells).

Immunofluorescence

Freshly cut 4-micron paraffin-embedded sections were deparaffinized/rehydrated

(3X xylenes 5 min, 2X 100% ethanol for 5 min, 95% ethanol for 3 min, 70% ethanol for 3 min). Samples were rinsed with PBS. Antigen retrieval was performed by boiling slides in

0.1mM EDTA for 30 minutes. Samples were cooled to RT and treated with SSC 0.2X with gentle shaking at RT for 20 minutes. Samples were blocked in 5% BSA/PBS with 0.5%

Tween-20 for 1 hr. Primary antibody incubation was done with monoclonal S9.6 antibody

(1:100, ENH0001 Kerafast) or anti-H2AX (1:100, 9718S Cell Signaling ) overnight at

4oC. After primary incubation, samples were washed 3 times with PBS containing 0.5%

Tween-20 and then incubated with secondary antibody for 1 hr. Samples were washed 2-3 times with PBS containing 0.5% Tween-20 and then mounted with Vectashield mounting medium containing DAPI (H-1200, Vector Laboratories). Slides were imaged at 60X with

Nikon A1 Spectral Confocal microscope. Analysis of S9.6 or H2AX foci per nucleus were calculated using Nikon analysis software.

Terminal Deoxynucleotide Transferase dUTP Nick-End Labeling (TUNEL)

TUNEL was performed on 4-micron sections. Sections were deparaffinized in xylenes and rehydrated in graded ethanols. The Apoptag Plus Peroxidase in Situ Apoptosis

Detection kit was used according to manufacturer’s instructions to label apoptotic cells

(S7101, MilliporeSigma). Sections were rehydrated in ethanols and cleared in xylenes before adding cover slips. Images were taken at 200x (model BX40 microscope, Olympus,

Tokyo, Japan) using a MicroPublisher 3.3RTV camera (QImaging, Surrey, British

Columbia, Canada). Positive cells were quantified using ImageJ as the percentage of positive luminal epithelial cells.

Results

Variation in Estrogen Receptor Expression

Previous studies have demonstrated that ERα protein levels in normal human breast tissue increase with age. To determine if this trend could be observed also at the transcript level, fresh breast tissue was collected from 129 women undergoing reduction mammoplasty. We used cDNA synthesized from these samples in a microarray to examine

ESR1 mRNA levels. To account for variation in epithelial content between samples, we normalized ESR1 by KRT18 expression (Figure 2.1a). Linear regression analysis of this data revealed ESR1 levels significantly increase with age (p < 0.05), corresponding with observed protein expression data. Variation in ESR1 mRNA level was high, as indicated by a low R2 value for the line of best fit (R2 = 0.115). We next wanted to determine if

ESR2 expression was also correlated with age. Examination of ESR2 mRNA levels normalized to KRT18 (Figure 2.1b) in these same patients from our microarray data revealed a significant increase in expression with age (p < 0.001). There was also a weak, but positive correlation between ESR1 and ESR2 (Pearson Correlation Coefficient 0.378, p < 0.001) in this data set.

While ERα is expressed only in luminal epithelial cells of the breast, ERβ expression has been observed in luminal epithelial cells, myoepithelial cells, stromal cells, endothelial cells, and lymphocytes. To examine ERβ expression in luminal epithelial cells, we performed IHC using the PGG5/10 mouse anti-human ERβ1 monoclonal antibody in a subset of 45 samples (Figure 2.2). We observed ERβ positive cells in the stroma as others have found, as well as in basal and luminal epithelial cells of the ducts (Figure 2.2a).

Linear regression analysis of ERβ positive epithelial cells revealed a significant decrease

in ERβ protein expression with age (Figure 2.2b, p < 0.05). Both ESR2 mRNA and ERβ protein levels had a high level of variance with low R2 values of 0.068 and 0.106, just as observed with ESR1 mRNA levels (Figure 2.2 and Figure 2.1, respectively). This data demonstrates age-related changes in expression of both estrogen receptors and that the effect is highly variable among individuals.

Human Breast Explant Model

It has previously been shown that estrogen receptor expression fluctuates with the menstrual cycle. ER expression is highest at the follicular phase of the estrus cycle. Our prior microarray data on ESR1 and ESR2 expression was performed using tissue from both premenopausal and postmenopausal women, with the majority being premenopausal. As we do not know the stage of the estrus cycle for each donor, this could significantly impact the levels of ESR1 and ESR2 we observed (Figure 2.1). In order to observe responses to

E2 treatment ex vivo, we utilized a human breast explant model (Figure 2.3). This model allows us to study responses to estrogen in a human ex vivo system, where each donor can be compared to their own basal control. Donor information, including age and parity status, is shown in Table 2.1. The donor explant samples collected in our initial culture conditions will be referred to as set 1. A list of analyses performed on these samples is included in

Table 2.2.

One benefit of the human breast explant model is that stromal, epithelial, and adipose cells are all present in culture, so cell-cell interactions can occur between these cell types. Culture conditions are important for maintaining this tissue architecture. In order to confirm that our culture conditions were not leading to degradation of the breast tissue,

we compared tissue architecture from tissue that was fixed within 2 hours of the reduction mammoplasty (T=0) to explant tissue after 7 days in culture (Figure 2.4). In this first set of explant samples, we did not observe tissue degradation or gross apoptotic nuclei in the epithelial ducts. Normal breast architecture was maintained, with cuboidal luminal and basal epithelial cells, surrounding stromal cells, and adipocytes. Nulliparous individuals consisted of smaller lobular structures (Figure 2.4a-d) than parous explants, which had more complex ductal lobular structures (Figure 2.4e-f) as previously reported (Russo et al., 1992). This tissue architecture is retained by the explants after 7 days in culture.

Optimization for RT-qPCR in explants

Breast tissue composition is known to vary greatly between individuals, which affects the amount of epithelium collected in each tissue fragment and overall RNA yield.

Women with dense breast tissue have greater amounts of stromal and epithelial cells compared to women with less dense breast tissue, who have more adipose and less epithelium. ERα is known to be restricted to the luminal epithelial cells of the breast epithelium. Expression of ERα protein can be detected by immunohistochemistry (IHC), but quantification is complicated. Representative ERα IHC images from one explant donor illustrate the heterogenous distribution of ERα protein (Figure 2.5). One lobule from donor 234 shows relatively high and contiguous ERα positive cells (Figure 2.5a), but another lobule from the same donor contains almost no ERα positive cells (Figure 2.5b).

Because of the heterogenous distribution of ERα and the limited amount of donor tissue we receive for explant cultures, IHC quantification of ERα protein is unreliable in explant cultures.

For this reason, we sought to optimize our RT-qPCR procedure for our human breast explant model. To account for differing amounts of epithelium between donors and between tissue sections, we used keratin 18 as a normalizer for epithelial content, which is expressed in luminal epithelial cells of the breast epithelium. To test this method, we took multiple pieces of flash-frozen breast tissue from several donors which had not been cultured and used them in separate RNA isolations. We then used RT-qPCR to determine

KRT18, ESR1, and ESR1 normalized to KRT18 (Figure 2.6). While fragments from donors

177 and 185 were consistent, fragments from donor 178 demonstrated marked variability in both KRT18 and ESR1 expression. KRT18 expression in fragment 178b is 6-fold greater and fragment 178d is 2.7-fold greater than fragment 178a. This is also reflected in ESR1 expression, but when normalized to KRT18, ESR1 expression is consistent across all donor fragments.

Using this method, we then were able to compare the amount of ESR1 between multiple donors (Figure 2.7a). Donor 178 had the lowest ESR1 expression, while donor

177 had the highest expression. Donor 185 had 2-fold higher ESR1 expression than donor

178 (p < 0.05). Donor 177 has 3.8-fold higher ESR1 expression than donor 178 (p < 0.01) and 1.8-fold higher expression than donor 185 (p < 0.05). We next sought to examine whether ESR1 expression is lost in our explant cultures, as is the case in primary breast epithelial cell 2D culture. Comparing fresh tissue samples (T=0) to tissue which had been cultured for 5 days (T=5), we looked at ESR1 expression normalized to KRT18 for two donors, referred to as donor A and donor B (Figure 2.7b). Donor B had approximately 2- fold higher expression of ESR1, but both donor A and donor B ESR1 levels were similar

between T=0 and T=5. There was no difference in ESR1 expression in explant tissue which had been cultured in basal or 10nM E2 media.

Transcriptional Responses to Estrogen

In order to determine if estrogen receptor levels correspond with estrogen-induced transcriptional responses, we examined ESR1 and ESR2 expression in our donor explant samples. Although previous work in breast cancer cell lines has suggested that E2 treatment decreases estrogen receptor expression, we did not notice any difference in ESR1 or ESR2 levels between basal and E2 treated explants (Figure 2.8, p = 0.18). Four individual donors showed an increase of 1.6-3.1-fold for ESR1 and 1.2-5.6-fold for ESR2.

For 8 out of 12 donors, ESR1 and ESR2 expression was decreased. Parity status did not appear to influence ESR1 and ESR2 expression in this subset. In agreement with our previous microarray data, we observed variation in estrogen receptor mRNA levels among our explant donors. We observed an over 80-fold variation in ESR1 expression and over

100-fold variation in ESR2 expression.

It has previously been shown that the opening of chromatin during estrogen- induced transcription leads to the formation of R-loops, DNA: RNA hybrids, which if not resolved are sites of potential DNA damage (Stork et al., 2016). As a measure of global transcription, we examined R-loop formation after estrogen treatment in our explants using the S9.6 antibody in immunofluorescence (Figure 2.9). In basal control treated explants, we observed low levels of R-loop formation indicative of basal transcription levels (Figure

2.9a). After estrogen treatment, there was an increase in the overall number of detectable

R-loop foci per nucleus, suggesting an increased level of transcription (Figure 2.9b).

Quantification of R-loop foci per nucleus revealed no significant decrease in basal treated explants compared to T=0 samples, but there was a significant 4-fold increase after E2 treatment in the group of 5 donor samples (Figure 2.9c, p < 0.01). For each donor there was a similar fold increase in R-loop formation with estrogen treatment, apart from donor

237 which had significantly lower R-loop foci/nucleus than the other 4 donors (Figure

2.9d, p < 0.01). These results indicate an estrogen-induced transcriptional response in all

5 donors, with similar levels of induction in four out of five donor samples.

After determining that there was a global increase in transcription after estrogen treatment in a small subset of our explant samples, we sought to look at estrogen receptor target genes as a more specific endpoint of estrogen responsiveness. RT-qPCR of estrogen receptor target genes AREG, PGR, and TGFβ2 in a set of combined nulliparous and parous explant donors at first revealed no significant changes with estrogen treatment. However, after separating the nulliparous and parous explants, we observed significant increases in

AREG (p < 0.05) and TGFβ2 (p < 0.05) and an increase in PGR (p = 0.059) in nulliparous explants using 2-way ANOVA to account for individual variation (Figure 2.10a-c). In nulliparous explants, AREG expression increased from 0.39+/-0.23 to 1.14+/-0.34 (Figure

2.10a, p < 0.05). PGR increased from 0.30+/-0.15 to 1.53+/-0.37 (Figure 2.10b, p < 0.06).

TGFβ2 expression increased from 0.46+/-0.25 to 1.68+/-0.34 (Figure 2.10c, p < 0.05).

Response among parous women was more variable, with no overall significant increase in

AREG, PGR, or TGFβ2. Notably, in certain parous explants, expression of one or more target genes increased while other individuals had decreased expression or little change

(Figure 2.10a-c).

Pearson correlation analysis of ESR1 and ESR2 mRNA levels and fold-change in target gene (AREG, PGR, TGFβ2) expression after E2 treatment revealed no significant correlation (Figure 2.11 and Table 2.5). ESR1 levels were positively and significantly correlated with ESR2 (Table 2.5, p < 0.05). This data suggests that estrogen receptor levels do not correlate with individual sensitivity to estrogen-induced transcriptional responses.

Functional Responses to Estrogen

To determine if estrogen receptor levels correlate with functional responses to estrogen, we examined protein expression of PR, proliferation, and apoptosis in a larger set of explant donor samples. PR protein was detected in T=0, basal, and E2 treated explants (Figure 2.12a). Quantification of PR positive luminal epithelial cells revealed a mean of 11.52% +/-2.55 positive cells in 14 T=0 samples (Figure 2.12b). Basal treated explants had a mean of 7.9% +/-1.38 positive cells, which was not significantly different from the T=0. Estrogen treated explants had a mean of 3.44% +/-0.94 which was not significantly different from basal treated explants but was significantly lower than in T=0 samples (p < 0.05).

Although in other models estrogen induces a proliferative response, we did not detect significant differences in proliferation as measured by PCNA expression between

T=0 (24.64%+/-5.78), basal (17.21%+/-3.6) or E2-treated (26.74+/-3.47) explants (Figure

2.13a-b). There was a trending increase in proliferation in the E2-treated explants compared to the basal, but it was not significant (p = 0.067). We observed no significant changes in proliferation with age or parity status within this sample set using linear regression analysis (Figure 2.13c). Interestingly, some donors appeared insensitive to E2-

induced proliferation, despite having low or high initial proliferative levels. Other donors showed a marked increase in proliferation after E2 treatment.

Our last endpoint of estrogen mediated functional responses was apoptotic response detected by TUNEL assay (Figure 2.14a-b). In our basal and E2 treated explants, we examined both spontaneous apoptosis and irradiation induced apoptosis. Both basal and

E2 treated explants showed a significant increase in the number of apoptotic cells 6 hours after exposure to 5Gy of irradiation (Figure 2.14c, p < 0.05). Basal treated explants increased from 1.36 +/-0.34 to 4.53% +/-1.2 while E2 treated explants similarly increased from 1.14 +/-0.33 to 2.94% +/-0.55. This response was not enhanced by estrogen treatment, and there was no observable difference between nulliparous and parous explants.

To determine if there was an increase in DNA damage after irradiation due to estrogen treatment, we performed immunofluorescence for γH2AX in basal and E2 treated explants

(Figure 2.15a-b). While apoptotic cells showed high levels of γH2AX staining, we only quantified the γH2AX foci in viable nuclei. Basal treated explants had 0.56 γH2AX foci per nucleus at baseline levels but with irradiation increased to 1.06 foci per nucleus (Figure

2.15c, p < 0.05). Due to variable γH2AX foci levels among the E2 treated explants, there was no significant increase in γH2AX after E2 treatment. However, the E2 irradiated explants had significantly higher γH2AX foci per nucleus (2.3 foci/nucleus) than the basal unirradiated explants (p < 0.05).

Pearson correlation analysis of the functional responses to E2 treatment in this set of explant samples did not find any significant correlation with ESR1 or ESR2 levels. There was a significant positive correlation of spontaneous apoptosis and proliferation in the explants (Table 2.6). These results imply that functional E2 responses in human breast

explants are also not correlated with estrogen receptor mRNA levels, suggesting that ESR1 and ESR2 levels do not alone determine sensitivity to estrogen.

ERα and ERβ in human breast explants

As another model to study how both ERα and ERβ contribute to the regulation of estrogen-induced responses among individuals, we expanded the scope of our human breast culture model. In order to study ERα and ERβ induces responses separately, we treated with PPT, an ERα specific agonist, or ERB041, an ERβ specific agonist, in addition to estrogen and control media. We compared expression of ER target genes AREG, PGR, and TGFβ2 in two initial donors treated with these agonists (Figure 2.16). Donor 1 had higher baseline expression of all 3 genes compared to donor 2, though the relative levels varied by gene. E2 treatment only significantly increased PGR expression in donor 1, but

PPT treatment significantly increased AREG and PGR expression (p < 0.05). Interestingly,

ERB041 significantly decreased all 3 genes in donor 1 (p < 0.05). E2 treatment in donor 2 increased expression of AREG and TGFβ2, while PPT and ERB041 significantly increased all 3 genes (p < 0.05). These donors indicate that both ERα and ERβ are involved in the regulation of these ER target genes and that these transcriptional responses differ between individuals.

Discussion

Breast cancer is the most prevalent cancer type and second leading cause of cancer related death in women (Siegel et al., 2019). Reproductive risk factors for breast cancer, including early menarche, late menopause, high serum estrogen levels, and pregnancy after

age 35 are all related to lifetime estrogen exposure. Paradoxically, estrogen is also capable of decreasing breast cancer risk, including through early parity. Periods of increased estrogen exposure or prolonged lifetime exposure, depending on context, have differing consequences for breast cancer risk in women. We demonstrate that in addition to varying in their exposure to estrogen, women also vary in their sensitivity to estrogen treatment. It is worth noting that increased breast density is also a risk factor for breast cancer and is positively correlated with higher serum estrogen levels, and therefore may correlate with estrogen sensitivity (Brooks et al., 2018). Identification of women who are more sensitive to estrogen exposure may reveal a subpopulation at increased risk for breast cancer.

Mutations in ESR1, such as 908A/G (K303R) and 1608T/A (Y537N), are known to produce ERα proteins which are more sensitive to estrogen treatment or constitutively active and are associated with increased breast cancer risk (Abbasi et al., 2012; Conway et al., 2005, 2007; Herynk et al., 2007; Jeselsohn et al., 2014).

Our goal was to utilize the ex-vivo human breast explant model to examine variation in estrogen receptor expression and sensitivity to estrogen detected by estrogen- induced transcriptional and functional responses. We sought to determine if estrogen sensitivity is correlated with estrogen receptor levels using a single dose of 10nM E2 vs control media in each individual donor. Transcriptional responses to estrogen included the detection of R-loops and quantification of AREG, PGR, and TGFβ2 target genes.

Functional responses to estrogen were regulation of PR, proliferation, irradiation-induced apoptosis, and DNA damage. We then performed correlation analyses to determine if these responses were correlated with estrogen receptor levels.

Previous work quantifying estrogen receptor levels in luminal epithelial cells of normal human breast tissue have shown that ERα expression increases with age and is detectable in 32-68% of samples (Battersby et al., 1992; Söderqvist et al., 1993; Gabrielson et al., 2016). In our microarray data of fresh normal breast tissue from over 100 donors, we also observed a significant increase in ESR1 mRNA with age (Figure 2.1a). ESR2 mRNA expression was also significantly increased with age (Figure 2.1b), although ERβ expression decreased with age (Figure 2.2b). ERβ is expressed in luminal epithelial, basal epithelial, and stromal cells (Figure 2.2a) and the amount of stromal cells increases with age (Goyal et al., 2016). The increase in ESR2 mRNA with age is likely affected by the increased stromal cell population, even after normalizing to KRT18 for luminal epithelial cell levels. It is also possible that our ERβ antibody, PPG5/10, detected another nuclear protein as many ERβ antibodies are not specific for ERβ (Andersson et al., 2017; Nelson et al., 2017).

Recent publications have questioned the validity of multiple commonly used ERβ antibodies. Earlier reports with the PPG5/10 ERβ antibody confirmed effectiveness through western blots detecting a 60kDa protein, which correspond to the ERβ 1 isoform, and through in vitro work with ERβ overexpression in U2OS cells (Shaaban et al., 2008;

Wu et al., 2012). These studies also confirmed ERβ expression in both epithelial cells and stromal cells, as we observed in our work. More reports suggest that the PPG5/10 ERβ antibody is not specific for ERβ. Using rapid immunoprecipitation mass spectrometry of endogenous protein, a 2017 paper demonstrated that the PPG5/10 antibody does have specificity for ERβ but also binds non-specifically to a 77kDa band in western blots

(Nelson et al., 2017). Another 2017 paper states that the PPG5/10 antibody does not bind

ERβ but instead binds to other nuclear proteins (Andersson et al., 2017). Regardless of

ERβ antibody specificity, we observed a high degree of variability in both ESR2 and ERβ levels among women. This was also true for ESR1 expression, suggesting individual variation in estrogen receptor expression.

In our human breast explant samples, we sought to correlate estrogen receptor levels with estrogen-induced responses. We aimed to quantify ERα expression by IHC;

ERβ detection by IHC is unreliable with current available antibodies. We found that ERα detection was highly variable both within an individual sample and between donors

(Figure 2.5). One lobule within a patient sample exhibited high ERα expression while other lobules expressed almost no ERα (Figure 2.5a and b, respectively). This agrees with previous work showing that, in normal human breast tissue, expression of ERα may be contiguous, within a single lobule, or scattered (Battersby et al., 1992; Goyal et al., 2016).

Due to these issues with the quantification of ERα protein and the limited amount of donor tissues we received for explant cultures, we examined mRNA expression of ESR1 and

ESR2 using KRT18 normalization to optimize for the amount of epithelium in each sample

(Figure 2.6 and Figure 2.7). In 12 explant donor samples, we observed up to 80-fold variation in ESR1 and up to 100-fold variation in ESR2 expression, which was not significantly changed with 10nM E2 treatment (Figure 2.8).

Examination of individual explant donors revealed a trending decrease in ESR1 and

ESR2 expression in 8 donors and increase in 4 donors (Figure 2.8). This suggests that negative feedback inhibition of estrogen receptor may be prevalent in certain donors in our study. In human breast cancer cell lines, some studies have observed negative feedback of estrogen receptor while others are unchanged. T47D breast cancer cells show an increase

in ESR1 mRNA levels and no change in ERα protein with estrogen treatment (Berkenstam et al., 1989; Read et al., 1989; Pink and Jordan, 1996). MCF7 cells, in culture conditions with estrogen containing serum, decrease ESR1 mRNA and ERα protein (Eckert et al.,

1984; Saceda et al., 1988; Read et al., 1989; Pink and Jordan, 1996). However, MCF7 cells cultured in media with charcoal stripped serum, to remove estrogens, no longer decrease estrogen receptor expression with estrogen treatment (Read et al., 1989). Our results therefore fit in with these previous studies, as some individual donors demonstrate

E2-mediated repression of ESR1 while others show an increase. This regulation likely depends on previous estrogen exposure and genetic background.

Estrogen receptor signaling pathways activate transcription of numerous estrogen receptor target genes (Gruber et al., 2002; Carroll et al., 2005; Heldring et al., 2007b; Yasar et al., 2017). In breast cancer cells and normal mammary epithelial cells, estrogen treatment is known to cause the formation of transient DNA:RNA hybrids known as R- loops (Stork et al., 2016; Belotserkovskii et al., 2018; Vanoosthuyse, 2018; Majhi et al.,

2020). The formation of these R-loops is dependent on estrogen receptor induced transcription. R-loop formation has been observed at estrogen receptor target genes in

MCF7 cells (Stork et al., 2016). Our lab has shown the presence of R-loops in T47D cells and in normal mouse mammary epithelial cells (Majhi et al., 2020). Human mammary epithelial cells, such as MCF10A and 76N Tert, do not form R-loops with estrogen treatment (Stork et al., 2016; Majhi et al., 2020). However, addition of estrogen receptor into these cells results in measurable R-loop formation upon estrogen treatment (Majhi et al., 2020). As R-loops are dependent on estrogen receptor expression and estrogen-induced transcription, we used R-loop detection as a measurement of global transcription in our

human breast explants. We observed a significant increase in R-loop formation with estrogen treatment, which was consistent across donors (Figure 2.9). This demonstrates that explants response to estrogen treatment in culture by activating transcription, as in mouse mammary glands and breast cancer cell lines.

Next, we examined specific transcriptional responses in our explants by RT-qPCR analysis of ER target genes AREG, PGR, and TGFβ2. AREG and PGR are common ER target genes induced upon E2 treatment in mice and in human breast cancer cell lines

(Ciarloni et al., 2007; Peterson et al., 2015). TGFβ2 is also an ER target gene and is downregulated in MCF7 cells with E2 treatment (Putnik et al., 2012). In nulliparous explants, we observed significant increases in AREG and TGFβ2 and a trending increase in PGR (Figure 2.10). Parous explants had overall more variable responses in these 3 target genes, with some increasing, decreasing, or remaining unchanged (Figure 2.10).

These results were somewhat surprising, as we had expected to observe increases in AREG and PGR and a decrease in TGFβ2. This may be due to the limited number of ER positive cells in our heterogeneous explant samples and that many of these target genes were identified in breast cancer cell lines.

Human breast organoids are thought to contain 7-10% ER positive cells (Meng et al., 2019), which is comparable to the 5-20% in normal human breast tissue. In human breast organoids E2 treatment upregulates PGR expression, but not AREG (Meng et al.,

2019). In human breast explants, AREG protein expression was increased with 10nM E2 treatment, which agrees with our data at the mRNA level in nulliparous explants (Eigeliene et al., 2012). Expression of estrogen responsive genes, therefore, may be diluted by the limited number of ER positive cells in the overall cell population in human breast organoids

and our human breast explants. ER positive breast cancer cell lines have more homogenous expression of ERα, and accordingly have consistent target gene responses. Although we hypothesized that donors with greater estrogen receptor expression would be more sensitive to E2 treatment, there was no correlation between ESR1 and ESR2 expression and

E2-induced changes in AREG, PGR, and TGFβ2 (Figure 2.11 and Table 2.5). These estrogen receptor target genes are more diversely regulated in normal human breast explants than in ER positive breast cancer cell lines, suggesting individual estrogen mediated transcriptional responses.

ESR1 and ESR2 were also not correlated with function responses to E2 treatment.

PR protein expression is detectable by IHC in more than 80% of human breast samples

(Battersby et al., 1992; Söderqvist et al., 1993), but we did not observe a significant change in PR expression between our basal control and 10nM E2 treated explants (Figure 2.12).

E2 treated explants had significantly lower PR than T=0 fresh breast tissue. Previous work in human breast explants also demonstrated no change in control vs E2 treated samples

(Eigeliene et al., 2008). Additionally, PR expression studies in human breast tissue collected during the follicular and luteal phases of the menstrual cycle, where estrogen levels fluctuate, demonstrated no change in PR (Battersby et al., 1992; Söderqvist et al.,

1993). Responses in our human breast explants differed compared to human explants in nude mice, however, where E2-treatment induced increases in PR expression (Laidlaw et al., 1995). Of the 5 explant donors included in detection of both PGR and PR, one donor

(185) exhibited very little change in either mRNA or protein expression, one donor (238) had an increase in mRNA and no change in protein, and three donors (179, 231, 234) had an increase in mRNA and a decrease in protein with estrogen treatment.

These PGR results are unexpected, as previous reports using breast cancer cell lines have suggested progesterone receptor mRNA and protein levels are closely correlated

(Read et al., 1989; Cho et al., 1994; Fazzari et al., 2001). However, it has also been demonstrated that PR regulation is controlled by several factors including cyclic adenosine monophosphate, serum factors, and growth factors (Cho et al., 1994). One other study in

T47D cells also proposed, depending on the subline of breast cancer cells and basal PR protein levels, as their T47D line showed estrogen-induced increase in PGR but no change in PR protein (Vegeto et al., 1990). In most breast tumors examined, mRNA levels correlated with protein activity, but in a subset of tumors (8%) this was not the case.

Regardless, 3 of 5 of our explants illustrate an apparent dysregulation of PGR mRNA and

PR protein, which could be due to increased protein turnover. However, our small sample size makes it difficult to make any definitive conclusions.

E2-induced proliferation, detected by PCNA, was also not correlated with either

ESR1 or ESR2 mRNA levels (Figure 2.13). In human breast cancer studies, Ki67 is often used as a marker for proliferation. The range of Ki67 positive cells in the normal human breast, however, is very low (0.3-2.6%) (Shoker et al., 1999b). Another specific marker for S/M phase of the cell cycle, phospho-histone H3, is also very lowly expressed in normal breast (<0.05%) and did not report E2-induced proliferative responses (Tanos et al., 2013).

In our human breast explants, we found that Ki67 expression was very low (<2% of cells), and with the limiting number of epithelial cells in our samples decided to use PCNA instead. PCNA is expressed in the S and G2 phases of the cell cycle (Juríková et al., 2016;

Qiu et al., 2019), and has been used previously by our lab in ovariectomized mice to detect estrogen-induced proliferation (Becker et al., 2005). In mice, we observed 60% PCNA

positive cells with E2 treatment compared to 5% in control. Our human explants also showed much higher levels of PCNA than Ki67, with a range of 3-58% (Figure 2.13b).

We observed a trending increase in proliferation in E2 treated explants in most donors, though some with high basal levels were not responsive. Though it has been shown that proliferation in the human breast is reduced in parous individuals and with age, we did not find any relationship between PCNA positive cells and age or parity (Figure 2.13c).

Our lab has shown that parous BALB/c mice are more sensitive to irradiation induced apoptosis than nulliparous mice and that this effect is enhanced by estrogen and progesterone treatment (Becker et al., 2005; Dunphy et al., 2008). In our human breast explants, while we observed a significant increase in apoptosis with irradiation, we did not observe a significant difference in irradiation-induced or spontaneous apoptosis between basal control and estrogen treated explants (Figure 2.14). It is known that spontaneous apoptosis does not vary during the menstrual cycle, which supports the results observed here (Navarrete et al., 2005; Potten et al., 1988). However, estrogen treatment has been shown to repress spontaneous apoptosis in human breast explants (Eigėlienė et al., 2006).

Deficiency in DNA damage response pathways is correlated with increased breast cancer risk (Venkitaraman, 2001; Roy et al., 2011), so we also examined a marker of DNA damage

(γH2AX) in our explants to detect any accumulation of damage from irradiation or estrogen treatment. We observed a significant increase in DNA damage after irradiation in both control and estrogen treated samples, but there was no increase with estrogen treatment

(Figure 2.15). This suggests that, in our explants, there was no notable accumulation of

DNA damage after 4-day estrogen treatment.

We treated a preliminary set of explants with ER specific agonists, PPT and

ERB041, to characterize the role of ERα and ERβ in the variable estrogen-induced target gene activation. In our two parous explant donors, we observed variable responses in

AREG, PGR, and TGFβ2 (Figure 2.16). PPT, an ERα specific agonist, induced AREG and

PGR expression in both explants (p < 0.05). In donor 2, PPT treatment increased TGFβ2 expression (p < 0.05), but in donor 1 TGFβ2 was unchanged. Donor 1, who had higher basal expression of all 3 target genes, showed a significant decrease in all 3 genes with

ERB041, an ERβ specific agonist, treatment (p < 0.05). This agrees with reports that ERβ opposes the actions of ERα (80-82). Interestingly, in donor 2, who had lower basal expression of these genes, the reverse was observed; ERB041 significantly increased expression of all 3 genes (p < 0.05). Both donor 1 and donor 2 exhibited modest target gene induction with estrogen treatment, which could be attributed to the ratios of ERα and

ERβ in the breast epithelium of each individual (Chang et al., 2008; Covaleda et al., 2008).

Our work with the human breast explant model illustrates the inter-individual variation in estrogen receptor expression as well as estrogen-induced transcriptional and functional responses. In agreement with previous literature, we observed significant increases in ESR1 and ESR2 with age which had only been shown at the protein level.

While global transcription, reported by R-loop detection, increases similarly among individuals, specific estrogen receptor target gene responses vary and are influenced by parity. Similarly, functional responses to E2, including regulation of PR and induction of proliferation, varied by individual. Neither transcriptional nor functional responses were correlated with ESR1 and ESR2 expression, indicating that the level of estrogen receptors expressed do not dictate estrogen sensitivity. We propose that differences in estrogen

receptor signaling, mediated by coactivators and corepressors, may modulate estrogen sensitivity and have important implications for breast cancer risk.

Table 2.1: Donor information for first set of human breast explants Donor information for all nulliparous, parous, and unknown parity donors used in the human breast explant model (set 1). Collected by Dr. Karen Dunphy.

Nulliparous Parous Unknown

Donor Donor Age at Donor Age Age Age Number Number First Birth Number

237 18 186 26 20 239 35 237 18 189 27 25 176 37 240 18 225 35 27 231 26 234 35 23 184 27 181 39 19 249 28 178 43 39 187 29 243 45 21 180 47 185 49 17 179 61 241 50 18

Table 2.2: List of analyses performed on first set of explant donor tissues Summary donor information and experiments performed on RNA or tissue sections of explant donor tissues from set 1. Experiments included R-loop detection, qPCR analysis of ER and ER target genes, progesterone receptor (PR) IHC, PCNA IHC for proliferation, TUNEL assay for detecting irradiation induced apoptosis, and irradiation induced DNA damage by γH2AX detection.

R- Donor qPCR PR PCNA γH2AX Age Parity loop TUNEL: Number analysis: IHC: IHC: IF: IF: ss237 18 Nulliparous Yes Yes Yes Yes Yes Yes ss238 18 Nulliparous No Yes Yes Yes Yes No ss240 18 Nulliparous No No No No Yes No ss231 26 Nulliparous No Yes Yes Yes Yes No ss184 27 Nulliparous Yes No Yes Yes Yes Yes ss249 28 Nulliparous No No Yes Yes No No ss187 29 Nulliparous No No Yes Yes Yes No ss180 47 Nulliparous No Yes No No Yes No ss179 61 Nulliparous No Yes Yes Yes Yes No ss250 62 Nulliparous No No Yes No No No ss186 26 Parous No Yes No Yes Yes No ss189 27 Parous No No No Yes Yes No ss225 35 Parous No No No No Yes No ss234 35 Parous Yes Yes Yes Yes Yes Yes ss181 39 Parous No Yes Yes No Yes No ss178 43 Parous No No Yes Yes Yes No ss243 45 Parous Yes No No Yes No Yes ss185 49 Parous No Yes Yes Yes Yes No ss241 50 Parous Yes No No Yes Yes Yes ss244 56 Parous No No Yes No Yes No ss177 58 Parous No Yes No Yes Yes No Yes (ER ss239 35 Unknown No only) No No No No Yes (ER ss176 37 Unknown No only) No No No No ss379 21 Parous No Yes No No No No ss438 51 Parous No Yes No No No No

Table 2.3: RT-qPCR primer sequences Primer sequences used in RT-qPCR of human explant samples. Used on Eppendorf thermocycler.

Gene Sequence: 5’ to 3’ Fwd – CAC AGT CTG CTG AGG TTG GA Keratin 18 (KRT18) Rev – GAG CTG CTC CAT CTG TAG GG Fwd – TTA CTG ACC AAC CTG GCA GA Estrogen receptor 1 (ESR1) Rev – ATC ATG GAG GGT CAA ATC CA Fwd – GTT TGG GTG ATT GCC AAG A Estrogen receptor 2 (ESR2) Rev – TCC ATC CCC TTG TTA CTG G Fwd – CGG AGA ATG CAA ATA TAT AGA GCA C Amphiregulin (AREG) Rev – CAC CGA AAT ATT CTT GCT GAC A Fwd – TTT AAG AGG GCA ATG GAA GG Progesterone receptor (PGR) Rev – CGG ATT TTA TCA ACG ATG CAG Transforming growth factor Fwd – ATA GAC ATG CCG CCC TTC TT β2 (TGFβ2) Rev – CTC CAT TGC TGA GAC GTC AA

Table 2.4: Antibody information for IHC Antibodies used in IHC on human explant samples.

Name of Antibody, Dilution Target Manufacturer Species Catalog Number Used Estrogen Thermo Fisher Mouse, receptor alpha 1D5, MA5-13191 1:200 Scientific monoclonal (ERα) Estrogen Mouse, receptor beta PPG5/10, M7292 Dako 1:200 monoclonal (ERβ) Progesterone Progesterone receptor Rabbit, Cell Signaling 1:500 receptor (PR) A/B, D8Q2J monoclonal Proliferating cell PC10, Mouse, nuclear antigen Abcam 1:10,000 ab29 monoclonal (PCNA)

Table 2.5: E2-induced gene response does not correlate with estrogen receptor expression Results of Pearson correlation analysis comparing ESR1 or ESR2 mRNA expression with fold change in AREG, PGR, and TGFβ2 expression induced by estrogen treatment compared to basal controls. ESR1 and ESR2 were not positively or negatively correlated with E2-induced change in AREG, PGR, and TGFβ2 expression (p > 0.05) as indicated by correlation coefficients close to zero. ESR1 is positively correlated with ESR2 expression (p < 0.05), indicated by a correlation coefficient close to 1.

Correlation Coefficient ESR1 ESR2

AREG -0.138, p = 0.705 0.016, p = 0.965

PGR -0.126, p = 0.729 -0.036, p = 0.921

TGFβ2 -0.096, p = 0.792 0.005, p = 0.989

ESR2 0.702, p = 0.024 ---

Table 2.6: ER mRNA levels do not correlate with functional responses to E2 Results of Pearson correlation analysis comparing ESR1 or ESR2 mRNA expression with E2-induced increase in PGR protein expression, proliferation, spontaneous apoptosis, or irradiation induced apoptosis. ESR1 and ESR2 were not positively or negatively correlated with E2-induced functional endpoints (p > 0.05) as indicated by correlation coefficients close to zero. Spontaneous apoptosis was significantly positively correlated with proliferation.

Spontaneous Irradiation Correlation ESR1 ESR2 PGR PCNA apoptosis induced Coefficient (no IR) apoptosis 0.090, -0.368, -0.202, -0.744, -0.226, --- PGR p = 0.910 p = 0.632 p = 0.798 p = 0.256 p = 0.774 0.116, 0.222, -0.202, 0.817, 0.027, --- PCNA p = 0.784 p = 0.597 p = 0.798 p = 0.013 p = 0.954 Spontaneous 0.332, 0.404, -0.744, 0.817, 0.349, --- apoptosis p = 0.348 p = 0.248 p = 0.256 p = 0.013 p = 0.357 (no IR) Irradiation 0.126, 0.297, -0.226, 0.027, 0.349, --- induced p = 0.678 p = 0.438 p = 0.774 p = 0.954 p = 0.357 apoptosis

a b

Figure 2.1: ESR1 and ESR2 expression diversity among women. Agilent microarray data of human breast tissue collected from donors undergoing reduction mammoplasty surgery. Expression of (a) ESR1 is shown in 111 donor samples and (b) ESR2 from 102 donor samples, both normalized to KRT18 expression. Linear regression analysis showed a significant increase in both ESR1 (p < 0.001) and ESR2 (p < 0.01) levels with age. Dotted lines indicate 95% confidence intervals and solid lines indicate the best fit line. R2 values for best fit lines are indicated, and low values indicate high variability between donors. Microarray data provided courtesy of Dr. Melissa Troester.

a b

Figure 2.2: ERβ protein expression decreases with age. IHC using ERβ antibody (PPG5/10) was performed on paraffin-embedded sections from 45 donor samples. Representative image (a) with arrows indicating representative positive cells: yellow arrows indicate positive luminal epithelial cells, red indicate positive basal epithelial cells, and black indicate positive stromal cells. Scale bar is 50uM. (b) Quantification of ERβ positive ductal epithelial cells. Linear regression analysis shows a significant decrease in ERβ with age (p < 0.05). Dotted lines indicate 95% confidence intervals and solid lines indicate the best fit line. R2 values for best fit lines are indicated. Data provided courtesy of Dr. Karen Dunphy and Dr. Sallie Schneider.

Figure 2.3: Overview of human breast explant model Tissue obtained from donors undergoing reduction mammoplasty were sectioned into strips. Some strips were flash frozen or fixed in 10% neutral buffered formalin (NBF) for time 0 (T=0) samples. All other strips were placed on surgical foam in tissue culture dishes and cultured in basal control media for 3 days to clear endogenous hormones. After clearing, the explants were treated with either basal control or 10nM E2 media for an additional 4 days. Six hours prior to collection, half of the samples were irradiated. At collection, tissue strips were either flash frozen for later RNA isolation or fixed in 10% NBF for later IHC.

Figure 2.4: Explant tissue architecture at T=0 and T=7 Representative H&E images from four explant donors comparing tissue architecture at T=0 (a, c, e, g) and T=7 (b, d, f, h) for nulliparous (a-d) and parous (e-h). Donors are from set 1. No apoptotic nuclei were observed. Scale bar is 50uM.

a b

Figure 2.5: Heterogenous expression of ERα protein Representative IHC images from donor 234 showing one lobule with high (a) and one lobule with low (b) expression of ERα. Scale bar is 50uM.

Figure 2.6: Optimization of RT-qPCR with KRT18 3-4 separate 100ug pieces of flash-frozen tissue from three donors (177, 178, and 185) were treated as separate samples and used for RT-qPCR analysis of ESR1, KRT18 and ESR1 expression normalized to KRT18. Data provided courtesy of Dr. Karen Dunphy.

Donor A Donor B

Figure 2.7: Comparison of ESR1 expression between donors and during culture Gene expression analysis of (a) ESR1 normalized to KRT18 between three parous donors, demonstrating variation between individuals. Data provided courtesy of Dr. Karen Dunphy. (b) Normalized ESR1 expression of two explant donors using tissue collected at T=0 and tissue from T=5 treated with either basal or 10nM E2 media. ESR1 expression for each donor appears stable during culture.

Figure 2.8: ESR1 and ESR2 expression in nulliparous and parous explants RT-qPCR analysis of basal and 10nM E2 treated explant samples for ESR1 and ESR2 using KRT18 normalization. Expression shown relative to control mean. Circles represent nulliparous explants, squares parous explants, and triangles explants with unknown parity status. Parity and age are shown next to donor ID. Individual fold change in gene expression is indicated. No significant difference in ESR1 or ESR2 expression was observed between basal and E2 treated explants (p > 0.05).

Figure 2.9: E2 induces R-loop formation in human breast explants Representative images of S9.6 (R-loop) immunofluorescence in (a) basal and (b) E2 treated explants from donor 184. (c) Quantification of R-loop foci per nucleus in basal and luminal epithelial cells demonstrates increased R-loop formation with E2 relative to T=0 or basal treatment (n=5 for each treatment group). (d) R-loop foci per nucleus in basal and E2 treated explants from five donors. Error bars indicate SEM. Significance was determined using ANOVA (p < 0.01 **). Data collected in collaboration with Aman Sharma.

Figure 2.10: AREG, PGR, and TGFβ2 expression in nulliparous and parous explants RT-qPCR analysis of basal and 10nM E2 treated explant samples for AREG, PGR, and TGFβ2 normalized to KRT18. Expression is shown relative to basal mean. Circles represent nulliparous explants and squares represent parous explants. Donor age and parity status is shown in parentheses next to donor ID. Four-day E2 treatment significantly increased (a) AREG (p < 0.05), increased (b) PGR (p = 0.059), and significantly increased (c) TGFβ2 (p < 0.05) in nulliparous explants, but not in parous explants. Significance was calculated using 2-way ANOVA analysis. Data provided courtesy of Dr. Karen Dunphy.

Figure 2.11: E2-induced target gene responses do not correlate with ER mRNA expression Heatmaps representing (top) the fold change in PGR, AREG, and TGFβ2 expression in basal vs E2 treated explants and (bottom) the expression of ESR1 and ESR2 for each donor explant relative to the overall mean. All values are normalized to KRT18. Hierarchical clustering was used to group donors by target gene response. Correlation analysis was performed using Pearson correlation coefficients in Graphpad Prism. Heatmaps were created in R studio using heatmap2 package. Donor number, parity status, and age are indicated.

Figure 2.12: PR regulation by E2 in human breast explants Representative (a) progesterone receptor IHC and (b) quantification. E2 treatment significantly decreased PR expression compared to T=0 samples (p < 0.05), but there was no difference between T=0 and basal or basal and E2 treated explants. Circles represent nulliparous explants and squares parous explants. Parity and age are shown next to donor ID.

Figure 2.13: E2-induced proliferation in human breast explants Representative (a) proliferating cell nuclear antigen (PCNA) IHC and (b) quantification. E2 treatment did not significantly increase proliferation in nulliparous or parous explants (p > 0.05). Linear regression analysis (c) shows no significant changes in E2-mediated proliferation due to age. Each individual donor is indicated by a color (see key): square indicates E2-treated; circle indicates basal-treated. Black arrowhead identifies individuals that are insensitive to E2-mediated proliferation. White arrow identifies individuals that are highly sensitive to E2-mediated proliferation (compare to same color circle below square).

Figure 2.14: Irradiation induced apoptosis in human breast explants Representative terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) images from (a) basal and (b) E2 treated explants irradiated with 5Gy 6 hours prior to collection to induce apoptosis. Quantification (c) revealed a significant increase in apoptotic nuclei in irradiated explants compared to unirradiated (p < 0.05), but E2 treatment did not enhance the apoptotic response (p > 0.05). Circles represent nulliparous explants and squares parous explants. Parity and age are shown next to donor ID. Data provided courtesy of Dr. Karen Dunphy and Zida Li.

Figure 2.15: γH2AX foci formation in human breast explants Representative images of γH2AX foci from (a) basal IR and (b) E2 IR explants, with inset indicated. Scale bars are 50uM. Large arrows indicate apoptotic cells and small arrows show an example of γH2AX foci. Quantification (c) of γH2AX foci per nucleus in five donor explant samples showed an increase in γH2AX foci in basal treated explants (p < 0.05) but not in E2 treated explants (p > 0.05). Data collected in collaboration with Aman Sharma.

Figure 2.16: Effect of estrogen receptor specific agonists PPT and ERB041 on AREG, PGR, and TGFβ2 target gene response Two explant donors (379 and 438) from set 1 were treated with basal, E2, ERα specific agonist PPT, or ERβ specific agonist ERB041 and examined for expression of estrogen receptor target genes.

CHAPTER 3

COMPARISON OF ESTROGENIC RESPONSES IN PRIMARY HMECS

Introduction/Rationale

Many epidemiological studies have identified risk factors for breast cancer, such as early menarche (RR = 1.3), late menopause (RR 1.2-2.0), or having high serum estrogen levels (RR = 1.8-5.0), which are linked to increased levels of estrogen and/or increased duration of exposure (Brinton et al., 1983, 1988; Clemons and Goss, 2001; Dall and Britt,

2017). Conversely, estrogen exposure can also reduce breast cancer risk through early full- term pregnancy by up to 50% (MacMahon et al., 1970; Trichopoulos et al., 1983; Lambe et al., 1996; Albrektsen et al., 2005) and can be an effective antitumor therapy in postmenopausal women with breast cancer (Lonning et al., 2001). These results demonstrate that estrogen exposure, depending on context, can have conflicting effects on breast cancer risk. Exposure to xenoestrogens, a class of endocrine disrupting chemicals, may affect this delicate balance between the protective effects of estrogen and its contribution to breast cancer risk. Endocrine disrupting chemicals are able to disrupt normal hormone signaling by mimicking or antagonizing the effects of endogenous hormones and also by disrupting the synthesis and metabolism of endogenous hormones and their receptors (Fernandez and Russo, 2010; Sonnenschein and Soto, 1998b).

Xenoestrogens are present in food, drinking water, ambient air, and many are found in personal care products such as cosmetics and sunscreens (Soni et al., 2001; Mortensen et al., 2014). Xenoestrogens have estrogenic activity, meaning they act as ligands for estrogen receptors and thus can potentially regulate estrogen-responsive target genes in a manner similar to estrogen (Singleton and Khan, 2003).

Recent work has identified the potential for environmental chemical exposures to contribute to breast cancer risk. Two xenoestrogens, benzophenone-3 (BP3) and propylparaben (PP) have been detected in > 95% of the U.S. population (Ye et al., 2006;

Calafat et al., 2008, 2010). Not much is yet known about potential breast cancer risk from

BP3 or PP exposure. BP3 is a UV-filter used in personal care products, including sunscreens, while PP is an antimicrobial agent used in food packaging and personal care products. Studies with MCF7 and T47D breast cancer cell lines, which express ERα, have shown that 1uM concentrations of BP3 act as a weak agonist for ERα in reporter assays

(Schlumpf et al., 2001; Kerdivel et al., 2013; Schlotz et al., 2017). BP3 exposure during pregnancy and lactation in mouse models results in altered mammary gland ductal structures, and these alterations remain for weeks after the exposure has ended (LaPlante et al., 2018). While some studies have shown that BP3 does not increase proliferation of breast cancer cell lines (Schlotz et al., 2017; Majhi et al., 2020), others have indicated that long term high dose exposure may increase cancer cell motility (Alamer and Darbre, 2018).

PP is also an ERα agonist in reporter assays but was also able to increase estrogen responsive target gene expression and increase proliferation in MCF7 cells at 1uM levels

(Byford et al., 2002) and increased cell motility in short and long term exposures (Khanna et al., 2014). Our lab has confirmed that BP3 and PP act as agonists for ERα and that only

PP is able to increase cell proliferation (Majhi et al., 2020). BP3 and PP can also induce the formation of DNA:RNA hybrids (R-loops) in breast cancer cells, normal HMEC lines, mouse mammary epithelial cells in vivo, and human breast epithelial cells in explant culture which have the potential to create DNA double strand breaks if not resolved.

Rodent studies have identified strains which have increased susceptibility to estrogen-induced mammary tumor development (Shull et al., 1997; Spady et al., 1998;

Dennison et al., 2015; Shull et al., 2018) and have increased sensitivity to estrogen-induced responses (Montero Girard et al., 2007; Aupperlee et al., 2009). These strains do not have mutations in any well-known breast cancer risk genes such as BRCA1, BCRA2, or TP53

(Anglian Breast Cancer Study Group, 2000; Antoniou and Easton, 2006; Lalloo and Evans,

2012; Peto et al., 1999; Thompson and Easton, 2004). We have demonstrated in human ex-vivo breast culture that there is significant inter-individual variation in response to estrogen that is not correlated with estrogen receptor expression levels (Dunphy et al.,

2020). This data supports the hypothesis that a subset of women may be more sensitive to estrogen exposure, and potentially more sensitive to xenoestrogen exposure as well.

Genetic variation between individuals is known to affect responses to therapeutic drugs (Roden et al., 2011). Individual variation has also been demonstrated to influence cell metabolism and responses to cytokines in human mammary epithelial cell (HMEC) models (Schneider et al., 2017). Additionally, genetic variation affects irradiation sensitivity and apoptosis responses in human cells and mice (Smirnov et al., 2009, 2012;

Snijders et al., 2012; Yard et al., 2016). Gene by environment (gene-environment) interactions are also known to influence breast cancer risk. SNPs in regions associated with breast cancer risk increased risk further when combined with alcohol consumption, height, hormone replacement therapy, and high body mass (Gonzales et al., 2018; Rudolph et al., 2018). The increased risk from gene-environment interactions follows a multiplicative model of increased risk (Travis et al., 2010; Nickels et al., 2013; Barrdahl et al., 2014; Rudolph et al., 2015; Maas et al., 2016). In related research, SNPS in specific

genes involved in the metabolism of carcinogens predispose individuals who smoke or drink alcohol to colon cancer development (Garcia-Closas et al., 2005; Wu et al., 2012a), which emphasizes how gene-environment interactions can influence the potential risk of environmental compounds.

It is important, therefore, to create a normal human breast model in which exposure to xenoestrogens or other compounds can be characterized. Induced pluripotent stem cell

(iPS) lines are useful for differentiating into multiple cell types, but limited success with iPS lines has been observed in mammary epithelial cell development (Cravero et al., 2015;

Qu et al., 2017). 3D breast organoid models are reliable for breast cancer cell expansion and characterization, but we have found limited success with normal HMEC cells.

However, others have had success expanding basal and luminal mammary epithelial cells in organoid culture which retain luminal epithelial cells with expression of ERα (Sokol et al., 2016). Human explant culture models preserve all cell types in the mammary gland, maintain estrogen receptor levels, and are hormonally responsive but have a limited lifespan (2-4 weeks) which limits their use for characterization of xenoestrogen-induced responses (Dunphy et al., 2020; Eigeliene et al., 2008, 2012, 2016; Eigėlienė et al., 2006;

Tanos et al., 2013).

HMEC models for studying normal breast biology have been limited by the early senescence of HMECs in 2D culture. Transformation with adenovirus E6/E7, human telomerase (TERT), Myc, SV40 large T-antigen, zinc-finger protein ZNF217, and BMI-1 can be used to bypass this senescence, but ultimately these methods perturb the biology of the cells. Loss of retinoblastoma protein, p53, and p16INK14a in these methods results in abnormal bypass of cell cycle checkpoints (Wang et al., 1998; Dyson et al., 1989; Münger

et al., 1989; Scheffner et al., 1990; Huschtscha et al., 2001; Nonet et al., 2001; Dimri et al.,

2002; Martinez-Zapien et al., 2016; Fischer et al., 2017). Myc immortalized lines exhibit morphological changes, anchorage independent growth, reduced apoptotic responses, and altered gene expression (Wang et al., 1998; Thibodeaux et al., 2009).

In our work, we sought to use an alternative approach, which we refer to as conditional immortalization. The conditional immortalization process for mammary epithelial cells involves coculture of mammary epithelial cells with irradiated mouse fibroblasts and addition of a rho-associated protein kinase (ROCK) inhibitor (Liu et al.,

2012; Palechor-Ceron et al., 2013). Other research has characterized conditionally immortalized HMECs (ciHMECs) and found similar expression profiles to original tissue samples (Mahajan et al., 2017). Some studies found that conditionally reprogrammed cells are representative of an adult stem-like epithelial population and express specific stem cell markers (Suprynowicz et al., 2012, 2017). Using this method, one group was able to expand a population of ERα positive cells with luminal epithelial markers, but they noted that this population decreased dramatically with passaging (Jin et al., 2017). Expansion of

ERα positive epithelial cells may be restricted by transforming growth factor beta (TGFβ) signaling, as treatment with inhibitors for this pathway was shown to enhance the growth of ERα positive HMECs (Fridriksdottir et al., 2015).

The overall goal for this aim was to characterize inter-individual variation in estrogenic responses in primary cell lines, representative of individual patients, and to use this information to investigate possible mechanisms contributing to estrogen and xenoestrogen sensitivity. To do this, we utilized a ciHMEC panel. We performed dual luciferase assays after transiently transfecting ciHMEC lines with estrogen receptor and a

reporter construct to measure individual responses. While we did observe up to 3-fold variation in response to E2, BP3, and PP, we only observed subtle differences in responses between HMECs. Due to difficulties with the conditionally immortalized HMEC lines, further work was not continued.

Materials and Methods

Cell culture

All Rays of Hope Registry cell lines were obtained from Dr. Sallie Schneider at

PVLSI. Primary cells were passaged in F-media (250 mL DMEM (- pyruvate), 250 mL

Ham’s F12, 5% FBS, 250 ng/mL hydrocortisone (Sigma), 10 ng/mL Epidermal Growth

Factor murine submaxillary (Sigma), 8.6 ng/mL Cholera Toxin Vibrio (Sigma), 1 µg/mL human Insulin solution (Sigma), and 1X Antibiotic-Antimycotic (100X) (Lonza Basal,

Switzerland).

MCF10A, ME16C2, 76N Tert, HMECC, and H16N2 stably immortalized breast cell lines (references) were also grown in F-media. NIH 3T3 mouse fibroblasts (gift of

Bruce Jacobson) were cultured in DMEM with 10% FBS and 15 ug/mL gentamycin. All cells were incubated at 37°C with 5% CO2 and passaged every 2-4 days.

Conditional immortalization

NIH 3T3 cells were treated with 30-gray (Gy) of gamma radiation. Cells were then plated in F media and allowed to adhere for 1-2 hours. HMEC cells were plated on the feeder layer with 10uM ROCK inhibitor (Y27632, StemCell Technologies).

RT-qPCR

RNA was isolated from cells using TRIzol according to manufacturer’s instructions. RT-qPCR was performed using the Mx3005P real-time PCR system

(Agilent, Santa Clara, CA). Reactions were performed using the 1-Step Brilliant SYBRIII

Green QRT-PCR Master Mix Kit (Agilent) and 200 nM forward primer, 200 nM reverse primer, and 10 ng RNA. Expression was set relative to universal mRNA control consisting of 5 donors with normal tissue (BioChain, San Francisco CA). See Table 3.2 for primers used.

Mammosphere Extreme Limiting Dilution Assay

A preliminary mammosphere forming assay was conducted to identify cell lines that have primary mammosphere forming capability. Cells were cultured in ultra-low attachment 96-well plates (Corning, NY) in 100 µL of F-media at differing concentrations:

250, 500, 1000, and 2000 cells/well. Cell lines that developed primary mammospheres after 7 days were selected for extreme limiting dilution assay. For extreme limiting dilution assays, each cell line was cultured in 100 µL of F-media in 96-well ultra-low attachment plates for 7 days. Each line was cultured at 4 concentrations ranging from 25 to 2000 cells/well with 8 replicates per concentration. Concentrations varied between lines based on initial preliminary mammosphere forming assay. After 7 days, positive wells containing mammospheres were counted in each concentration for each cell line. Mammosphere forming cell frequency (±95% confidence interval upper and lower limit) was calculated using online software (bioinf.wehi.edu.au/software/elda16).

ERE-Luciferase Assays

HMECs were plated in 24 well tissue culture plates at densities ranging from 7 x

104 to 1 x 105 cells/well in F-media with 10 uM Y27632. After an overnight incubation, growth media was replaced with 0.3 mL/well Opti-MEM reduced serum media (11058-

021, Gibco). Plasmid DNA and Lipofectamine 2000 reagent (11668019, ThermoFisher) were diluted in Opti-MEM media before adding 200 µL/well. Cells were transfected with

0.125 µg of either ESR1, ESR2, or both ESR1 and ESR2 expression plasmids, 1.5 µg of 3X

ERE-TATA-Luciferase reporter plasmid, and 0.02 µg of pRL-CMV Renilla plasmid using lipofectamine 2000 (Invitrogen). As a receptor-negative control, 6 wells were transfected with 1.5 µg 3x ERE-TATA-Luc reporter plasmid and 0.02 µg pRL-CMV plasmid; half were treated with control media and the other half with 17β-estradiol (E2) media. For

Fugene 6 transfections, 0.01ug/well of pRL-CMV-Renilla, 0.6ug/well of 3X ERE-TATA-

Luciferase and 0.02ug/well ESR1 were as above. 1.8uL/well of Fugene 6 (E2691,

Promega) was used for transfection.

After 6 hours of transfection, transfection media was replaced with 0.5 mL of treatment media. Cells were incubated with treatment media for 24 hours. The Promega

Dual Luciferase Reporter Assay was used to perform luciferase assays (E1910, Promega).

Cells were lysed in 1x Passive Lysis Buffer and lysates stored at -20°C until reading.

Luciferase and Renilla activity in lysates were determined by using the Polar Star Optima plate reader (BMG Labtech). Luciferase values for each well were normalized to the amount of Renilla activity.

Cell Treatments

Control treatment media consisted of MEM media (51200-038, Gibco), 5% charcoal-dextran stripped serum (FB-04, Omega Scientific), and 2mM L-glutamine

(SH30034.01, Hyclone). 10 mM 17β-estradiol (E2758, Sigma-Aldrich) stock prepared in ethanol was diluted to prepare E2 treatment media. 10 mM 2-hydroxy-4- methoxybenzophenone (H36206, Sigma) and propyl 4-hydroxybenzoate (P53357, Sigma) stocks were prepared in DMSO (D8418, Sigma) and diluted to make BP3 and PP treatment media.

Immunofluorescence

Cells were plated on glass coverslips in 12 well plates at a density of 2x105 cells/well in normal growth media. After overnight incubation, the media was removed, cells were washed with 1x phosphate buffered saline (PBS), and then fixed in 4% paraformaldehyde for 15 minutes at room temperature. Cells were permeabilized with

0.5% TritonX-100 for 10 minutes at 4°C and washed 3 times with 1x PBS: Glycine

(130mM NaCl, 7mM Na2HPO4, 3.5mM NaH2PO4, 100mM Glycine) for 15 minutes each at room temperature. Blocking was performed in immunofluorescence buffer (130mM

NaCl, 7mM Na2HPO4, 3.5mM NaH2PO4, 7mM NaN3, 0.1% BSA, 0.2% TritonX-100,

0.005% tween-20) with 10% donkey and goat serum for 1-2 hours. A second blocking was performed using immunofluorescence buffer with 10% donkey and goat serum and

20ug/mL goat anti-mouse F(ab’)2 (ab6668, Abcam) for 30-40 minutes. Primary antibody incubation was performed with 1:500 human mitochondrial antibody (ab92824, Abcam) and 1:200 vimentin antibody (ab92547, Abcam) overnight at 4°C. Cells were washed with

1x PBS: Glycine 3 times for 15 minutes each, incubated with secondary antibody for 1

hour (1:500 Cell Signaling 4408S and Abcam ab150076), washed, and mounted using

Vectashield mounting media with DAPI (H-1500, Vector Laboratories). Images were obtained under 200x magnification (model BZ-X700 microscope).

Results

Characterization of HMECs

Using a previously published method for conditional immortalization of human breast epithelial cells, we expanded a subset of donor primary breast cells for characterization of estrogenic responses (Figure 3.1). Donor age, BMI, type of surgery, ethnicity, and parity status were recorded for each donor (Table 3.1). Briefly, primary human mammary epithelial cells (HMECs) were isolated from digestion of donor tissue samples and plated in 2D tissue culture. To conditionally immortalize these primary cell lines, we put them into co-culture with irradiated NIH 3T3 mouse fibroblasts and added a rock inhibitor (Y27632). This allowed for expansion of these primary lines for further analysis and are now referred to as conditionally immortalized HMECs (ciHMECs).

Our collaborator Stephanie Morin from the Schneider lab performed PCR characterization of these HMECs for basal epithelial markers keratin 14 (KRT14) and keratin 5 (KRT5), luminal epithelial marker keratin 18 (KRT18), and basal/mesenchymal cell marker vimentin (VIM). Upon analysis, this characterization revealed 3 subgroups: one group with low expression of all markers, one large group with low expression of

KRT18 and VIM but moderate to high levels of KRT14 and KRT5, and one with low expression of all keratins but high expression of VIM (Table 3.3 and Figure 3.2). Based on expression of these markers, we concluded that the cells isolated from this method were

largely basal epithelial cells, with variable expression of KRT14 and KRT5. This characterization was also performed on 4 stably immortalized HMEC lines obtained from other labs (MCF10A, 76N Tert, ME16C2, HME-CC), which all had low expression of

KRT18 and VIM while 3 of 4 also had low KRT14 and KRT5.

Further characterization of these ciHMEC lines demonstrated that some (5 of 6) lines retained the ability to form primary mammospheres in a limiting dilution mammosphere assay, indicating some cells in these populations may resemble basal stem/progenitor cells (Figure 3.3a-c). Five stably immortalized HMEC lines (MCF10A,

76N Tert, ME16C2, HME-CC, H16N2) were also tested, and 3 of 5 lines had some primary mammospheres (Figure 3.3a-b). Interestingly, the estimated number of mammosphere forming cells (MFC) in the ciHMECs (10.6 MFC/1000 cells) was significantly higher than in the stably immortalized HMEC lines (2.4 MFC/1000 cells, Figure 3.3c, p < 0.05). This suggests that ciHMEC culturing conditions may support the growth of basal epithelial cells with stem/progenitor characteristics.

Measuring Estrogenic Responses in HMECs

Because ERα positive luminal epithelial cells are lost when HMECs are plated in

2D tissue culture conditions, we sought to use transient transfection experiments with estrogen receptors and an ERE reporter to measure variation in estrogen-induced responses in our ciHMEC cell lines. To do this we first optimized our transient transfection protocol for HMECs, which we found were more sensitive to transfection media conditions than cancer cell lines. HMECs cultured in clearing media (10% charcoal stripped FBS,

10ug/mL insulin, 2mM L-Glutamine, MEM media) prior to transfections, to reduce

estrogenic background, were found to have higher numbers of dead floating cells and no successful uptake of plasmid DNA (data not shown), so this step was omitted for all future transfection experiments. Two alternative transfection protocols using either

Lipofectamine 2000 or Fugene 6 reagents were tested for optimal efficiency in HMEC cells

(Figure 3.4). Transfection efficiency was markedly improved with Lipofectamine 2000 reagent compared to Fugene 6. No luciferase readings were detected with Fugene 6 and significantly lower Renilla transfection control levels were observed (Figure 3.4a and b, respectively). When Lipofectamine transfected luciferase values were normalized to

Renilla, we were able to observe a significant increase in transactivation with 10nM E2,

30uM BP3, and 10uM PP treatment in ciHMEC line 910 (Figure 3.4c). The results of this experiment suggested that Lipofectamine transfection reagent was more suitable for transfection of ciHMEC cells, and thus was used for all subsequent experiments.

Further optimization was performed to determine the optimal amount of ESR1 plasmid DNA required to measure estrogenic responses in HMECs. In 76N Tert cells, transient transfections were performed using ESR1 plasmid DNA levels ranging from

0.03ug/well to 1.5ug/well while keeping ERE-luciferase reporter and renilla plasmid amounts consistent (Figure 3.5). At all levels of ESR1 used, a significant increase in transactivation was seen in 10nM E2 treated cells compared to control (Figure 3.5a). The highest amount of relative light units of luciferase was observed at 0.125ug/well ESR1 and decreased at higher plasmid levels. Renilla relative light units were also greatest at

0.125ug/well ESR1 and decreased with higher plasmid concentration, but renilla levels were consistent from 0.03ug/well to 0.125ug/well ESR1 (Figure 3.5b). When normalized to renilla transfection control, values for all control treated wells reflected similar amounts

of transactivation and the same was true for all 10nM E2 treated wells (Figure 3.5c). For future experiments, we decided to use 0.125ug/well ESR1; this amount of ESR1 appeared to be saturating.

BP3 and PP are Estrogen Receptor Agonists in HMECs

In order to determine if xenoestrogens benzophenone-3 (BP3) and propylparaben

(PP) are agonists for ESR1 and/or ESR2 in human breast epithelial cells, we first used 3 stably immortalized HMECs to generate dose responses to each compound. This was measured using ERE-luciferase assays, where the amount of transactivation was compared with ESR1 only, ESR2 only, and both ESR1 and ESR2. In all 3 lines, BP3 and PP were capable of transactivation with all combinations of receptors, and the level of transactivation increased with increasing doses (Figure 3.6). Activity observed with both xenoestrogens was similar for ESR1 only and both ESR1 and ESR2 transfections, suggesting that with these agonists, the activity of ESR1 is not enhanced or suppressed by

ESR2.

BP3 transactivation with ESR1 and both receptors in the 76N Tert line reached approximately 60% of the activity observed with 10nM E2 treatment and ESR1 only transfection (Figure 3.6a, gray dotted lines indicate 10nM E2 and ESR1 activity). In the other two lines, BP3 transactivation with ESR1 and both receptors reached the full activity seen with 10nM E2 (Figure 3.6b-c). With ESR2 only, 10nM E2 treatment induced lower overall levels of transactivation; this was also observed with BP3 and PP treatment. BP3 treatment and ESR2 transfection reached approximately 40% of transactivation levels induced with ESR2 and 10nM E2 treatment (green dotted lines) in the 76N Tert and HME-

CC lines (Figure 3.6a and c). This level was close to 100% in the ME16C2 line (Figure

3.6b). Our results show that BP3 can act as an agonist for both ESR1 and ESR2 but generates greater activity with ESR1.

PP induced transactivation with ESR1 alone and both receptors in the 76N Tert and

HME-CC lines which reached approximately 80% of the level induced by 10nM E2 treatment (Figure 3.6d and f). The ME16C2 line was more sensitive to PP treatment, as the amount of transactivation reached the same levels as 10nM E2 treatment (Figure 3.6e).

With ESR2 only, all 3 lines reached 100% transactivation levels compared to 10nM E2 and

ESR2 transfection (Figure 3.6d-f). Overall, this data suggests that BP3 and PP can act as agonists for both ESR1 and ESR2, but that most of the transactivation induced is due to

ESR1.

Limited Variation in Responses to E2, BP3, and PP in ciHMECs

After determining that BP3 and PP were acting primarily through ESR1 in our immortalized HMECs, we performed a luciferase reporter screen in our ciHMEC donor lines to identify individuals who are more sensitive to estrogen or xenoestrogen exposure.

Because little activity was seen with ESR2, we decided to transfect only ESR1 into these cell lines at saturating levels. We also chose single saturating doses of BP3 and PP; 30 uM

BP3 reflects the 95th percentile of exposure in pregnant women, while 10uM PP is 3-fold higher than the 95th percentile exposure level (Philippat et al., 2013). We performed luciferase assays in a total of 11 lines, including 4 immortalized (HME-CC, 76N Tert,

MCF10A, ME16C2) and 7 ciHMEC donor lines (Figure 3.7). All HMEC and ciHMEC lines responded to E2, BP3, and PP relative to control treated wells. This suggests BP3

and PP are estrogenic in ciHMEC lines. We observed a 2.5-fold variation in response to

E2 treatment, 2.3-fold variation to BP3, and 3.6-fold variation to PP treatment (Figure

3.7). Similar levels of activation were seen with most HMEC lines in response to BP3 and

PP treatment (p > 0.05), but the 876 and 698 primary HMEC lines were more sensitive to

BP3 than PP (P < 0.05). The 76N Tert, MCF10A, and 569 primary HMEC exhibited significantly more activity when treated with E2 compared to either xenoestrogen (P <

0.05), but the remaining 10 HMEC lines had similar activity levels with E2, BP3, and/or

PP (P > 0.05). The 811 line displayed a consistently lower response to E2, BP3, and PP compared to the other 12 HMEC lines.

Based on the results obtained from our human breast explant model, we had expected to see more inter-individual variation in responses in our ciHMEC donor cell lines. We hypothesized that the limited levels of variation could be due to using saturating levels of ESR1 or high doses of xenoestrogen treatment. It is possible that if we had performed dose response curves to estrogen and each xenoestrogen in every cell line, we would observe differing EC50 values for each patient, reflective of differing sensitivity.

Another possibility is that the coregulators involved in estrogen receptor signaling are present at very low levels in these cell lines, as it has been shown that specific coactivators can enhance luciferase responses in breast cancer cell lines (Wolf et al., 2007; Kong et al.,

2011). To investigate this possibility, we checked mRNA expression of FOXA1 and

GATA3 in our ciHMEC lines (Figure 3.8a-b). Expression of both coactivators was 10 to

12-fold lower than in MCF7 breast cancer cells, which may contribute to the limited responses in our luciferase assays. However, more work would be needed to make this

conclusion.

Feeder Cell Contamination in ciHMEC cultures

While our conditional immortalization method allowed us to expand multiple donor primary breast epithelial cell lines for phenotypic characterization, we encountered difficulties working with the mouse NIH 3T3 fibroblast cell line. Previous publications have described the use of irradiation or mitomycin c to halt proliferation in NIH 3T3 cells for use as feeder cell layers. We also used differential trypsinization to limit the number of feeder cells carried over during passaging. However, with these methods, we still observed NIH 3T3 feeder carry over in certain ciHMEC lines (Figure 3.9).

Immunofluorescence staining with a vimentin antibody, which marks fibroblasts and mesenchymal cells, and a human specific mitochondrial antibody in 4 higher passage ciHMEC lines revealed 3 of 4 lines carried some amount of NIH 3T3 contamination

(Figure 3.9d-f). Primary HMECs, which had never been co-cultured with NIH 3T3 cells, were used as a negative control for detection of the contaminating mouse cells (Figure

3.9a). Of the 3 contaminated lines, 753 had only a few observable NIH 3T3 cells, 698 contained approximately 50% human and 50% mouse cells, and 569 contained only a handful of human cells remaining (Figure 3.9d, e, and f, respectively). Compared to normal NIH 3T3 cells, the contaminating cells sometimes appeared to have altered morphology (Figure 3.9b vs f).

Discussion

Previous in vitro work using HMECs to study the normal biological responses of mammary epithelial cells was limited by the early senescence of HMECs in 2D culture and limited expansion capabilities. Transformation with TERT, adenovirus E6/E7, Myc, SV40 large T-antigen, ZNF217, and BMI-1 are methods which have been used to expand subsets of HMECs, but these methods ultimately disturb normal biological functions through loss of p53, retinoblastoma protein, p16INK4a, or oncogenic transformation (Wang et al., 1998;

Dyson et al., 1989; Münger et al., 1989; Scheffner et al., 1990; Huschtscha et al., 2001;

Nonet et al., 2001; Dimri et al., 2002; Martinez-Zapien et al., 2016; Fischer et al., 2017).

Our main concern with using these methods to characterize inter-individual variation in responses to estrogen or xenoestrogens was that these biological changes may influence the HMECs’ responses to stimuli. Therefore, we sought to utilize the conditional reprogramming method (Liu et al., 2012; Palechor-Ceron et al., 2013), or conditional immortalization, for expansion of our HMEC lines. This method is thought to gradually reprogram the epithelial cell population but does induce TERT expression, though the conditional reprogramming phenotype is believed to be reversible upon removal of the

ROCK inhibitor and feeder cells (Suprynowicz et al., 2012; Palechor-Ceron et al., 2013).

Using this method, we expanded a subset of HMEC lines from our panel of women donors from the Rays of Hope Registry. Our goal was to use this subset to characterize the variation in responses to estrogen and xenoestrogen treatment to determine if some women are uniquely sensitive to these exposures and then determine the mechanism(s) responsible.

A recent paper by Kumar et al. used conditional immortalization to expand low passage cultures of HMECs for permanent immortalization with TERT to generate

immortalized HMEC lines from a panel of normal breast tissue (Kumar et al., 2018).

Though this approach has the potential issue of subpopulation selection, they were able to preserve a small luminal cell population with ERα expression, luminal cell markers

(FOXA1 and GATA3), and mammosphere forming capability. Our method of HMEC isolation and culture promoted the growth of epithelial cells with primarily basal epithelial markers (Table 3.3 and Figure 3.2), but we did observe variability in the expression of

KRT5, KRT14, KRT18, and VIM markers between donor HMECs. These results were confirmed with flow cytometry for an epithelial cell marker EpCAM and Keratin 14 (data not shown). The low KRT18 expression in these HMEC lines suggests that any luminal epithelial cell populations were lost during culturing, which agrees with other published results (Fridriksdottir et al., 2015; Lee et al., 2018). This is likely due to reduced adherence of luminal epithelial cells (Soule and McGrath, 1986; Keller et al., 2010), slower growth rates compared to basal epithelial cell populations (Keller et al., 2010; Fridriksdottir et al.,

2015), increased stress responses in luminal cells (Lee et al., 2018), microenvironment changes (Miyano et al., 2017), or loss of differentiated phenotypes (Keller et al., 2010;

Breindel et al., 2017).

Consistent with previous reports that conditionally reprogrammed epithelial cells acquire an adult stem-like phenotype (Suprynowicz et al., 2012, 2017; Breindel et al.,

2017), we found that our ciHMECs had primary mammosphere forming cells in ultra-low attachment conditions (Figure 3.3) and the number of these cells was increased compared to spontaneous and TERT immortalized HMECs. Of the 6 ciHMEC lines tested, 5 were able to form primary mammospheres, compared to only 3 of 5 stably immortalized

HMECs. Interestingly, though previous reports have shown MCF10A are capable of

forming mammospheres (Arendt et al., 2014), we did not observe mammosphere formation. It is possible that the number of mammosphere forming cells in the MCF10A line were below our detectable limit, as others have noted their ability to form mammospheres is low (Qu et al., 2015).

Although our prior work has shown significant inter-individual variation in response to estrogen treatment (Dunphy et al., 2020) and others have shown effects on drug response, cell metabolism, cytokine response, and irradiation sensitivity (Roden et al.,

2011; Snijders et al., 2012; Yard et al., 2016; Schneider et al., 2017), we only detected low variation between the 7 ciHMEC and 4 stably immortalized HMEC lines in response to estrogen or xenoestrogen treatment in our luciferase reporter assays (Figure 3.7). There was up to 2.5-fold variation in response to E2 treatment, 2.3-fold variation to BP3, and 3.6- fold variation to PP treatment. It is possible that if we had performed dose response characterization on each HMEC line, we may be able to detect differences in EC50 values that could be more reflective of inter-individual variation. Another alternative is that we could not observe greater differences between lines due to using saturating levels of ESR1

(Figure 3.5) or low levels of estrogen receptor coactivators, including FOXA1 and GATA3

(Figure 3.8). Lastly, it is possible that we simply did not analyze a large enough number of HMECs to find the subpopulation with increased sensitivity to these compounds.

Our results are consistent with other previous work that BP3 and PP are estrogenic

(Schlumpf et al., 2001; Byford et al., 2002; Kerdivel et al., 2013; Schlotz et al., 2017; Majhi et al., 2020) and act as ligands for ERα and ERβ (Figure 3.6). While BP3 metabolites may have greater affinity for ERβ (Molina-Molina et al., 2008), we observed greater affinity for

ERα for both BP3 and PP. Though they have potential to act as ligands for both estrogen

receptors, it is unclear if this would translate into a biologically relevant response. BP3 is not known to induce breast cancer cell proliferation (Schlotz et al., 2017; Majhi et al., 2020) but can increase cell motility in long term exposure (Alamer and Darbre, 2018) and alter ductal formation in mammary gland development (LaPlante et al., 2018). PP can induce proliferation in breast cancer cells and increases cell motility (Byford et al., 2002; Khanna et al., 2014; Majhi et al., 2020). Both xenoestrogens form estrogen receptor dependent

DNA:RNA hybrids in response to estrogen treatment, but only BP3 treatment caused DNA damage in estrogen treated breast cancer cells (Majhi et al., 2020). Regardless of whether

BP3 and PP are able to induce all or some of the biological responses induced by estrogen treatment, their ability to act as estrogen receptor ligands in all HMEC lines tested highlights their potential risk during susceptible windows of development or to those that may be uniquely sensitive to estrogen and xenoestrogen exposure.

Our major concern when starting this work with ciHMEC lines was the possibility for contamination of mouse fibroblast feeder cells in the ciHMEC cultures. Though it is well documented that irradiation of mouse 3T3 feeder cells or treatment with mitomycin-

C is sufficient to arrest cell growth (Chugh et al., 2016, 2017), our work has highlighted the potential for a small population of 3T3 cells to undergo phenotypic changes and become resistant to irradiation induced senescence (Figure 3.9). While some ciHMEC lines tested were unaffected, others contained 50% or greater contamination. Differential trypsinization normally removes the 3T3 population from the ciHMEC culture, but this irradiation resistance subpopulation is also more resistant to trypsinization. Optimization of this method for future use could include fluorescent labeling of 3T3 cultures to identify contamination more easily, or infection with recombinant plasmid containing a

hygromycin phosphotransferase-thymidine kinase fusion gene to allow negative selection with ganciclovir.

Table 3.1: Donor information for primary human mammary epithelial cells Donor age, BMI, type of surgery, ethnicity, and parity status for all HMECs collected at PVLSI by Dr. Sallie Schneider and the limited donor information known about five previously established immortalized HMEC lines.

ID Number or Age BMI Surgery Ethnicity Parity Cell Line Reduction 563 19 26.8 Caucasian Nulliparous Mammoplasty Reduction 569 29 26.0 Caucasian Nulliparous Mammoplasty 698 48 48.9 Mastectomy Caucasian Nulliparous 753 42 24.7 Mastectomy Caucasian Parous 811 50 34.6 Mastectomy Caucasian Parous Prophylactic 833 42 36.5 Caucasian Parous Mastectomy 847 59 16.5 Mastectomy Latina Parous 876 44 24.8 Mastectomy Caucasian Parous MCF10A 36 Unknown Mastectomy Unknown Parous Reduction 76N Tert Unknown Unknown Unknown Unknown Mammoplasty ME16C2 53 Unknown Mastectomy Unknown Unknown HME-CC Unknown Unknown Unknown Unknown Unknown H16N2 36 Unknown Mastectomy Unknown Unknown

Table 3.2: RT-qPCR primers RT-qPCR primer pairs used on Mx3005P real-time PCR system.

Gene Forward Primer (5’ to 3’) Reverse Primer (5’ to 3’) Keratin 14 TTC TCC TCT GGA TCG CAG TC ATG ACC TTG GTG CGG ATT T Keratin 5 TGC AGA CTC AGT GGA GAA GG TCC AGA GGA AAC ACT GCT TG Keratin 18 CAC AGT CTG AGG TTG GA GAG CTG CTC CAT CTG TAG GG Vimentin ACG AAG AGG AAA TCC AGG AG CAG AGA GTC AGC AAA CTT GGA Gata3 GAG AGA GAG ACG GAG GGA GA GTC ACC TGG GTA GCG AAG AG Foxa1 GGA ACA GCT ACT ACG CAG AC ATG TTG CCG CTC GTA GTC AT

Table 3.3: Keratin expression profiles of HMEC lines RT-qPCR results for keratin 14 (KRT14), keratin 5 (KRT5), keratin 18 (KRT18), and vimentin (VIM). Expression is shown relative to human universal RNA from breast tissue, normalized to beta-actin (ACTB), and log2 transformed. Passage number is shown for ciHMEC lines. Data provided by Stephanie Morin.

Cell Line KRT14 KRT5 KRT18 VIM 563 P20 3.78 4.50 -3.32 -3.64 569 P8 4.07 4.37 -5.06 -4.32 698 P14 6.32 5.04 -2.94 -1.60 753 P12 1.06 2.18 -3.18 0.32 811 P4 4.81 4.28 -3.06 -1.43 833 P4 2.52 2.65 -3.06 -1.94 847 P1 -0.71 -1.51 -1.29 -1.15 876 P6 2.04 3.40 -3.32 -1.32 MCF10A 0.14 -2.18 -0.27 -6.64 76N Tert -0.10 -2.84 -1.51 -2.64 ME16C2 0.77 -1.15 -2.84 -0.22 HME-CC 2.93 2.99 -1.65 -5.46

Figure 3.1: Overview of Conditionally Immortalized HMEC model Primary human mammary epithelial cells (HMECs) were isolated from digestion of human donor breast tissue samples at PVLSI and put into 2D tissue culture conditions. For conditional immortalization, primary HMECs were cocultured with irradiated NIH 3T3 mouse feeder cells in the presence of a rock inhibitor (Y27632). This allowed expansion of conditionally immortalized HMECs (ciHMEC).

Figure 3.2: ciHMEC PCR Characterization Heatmap of the keratin expression profiles for HMEC lines shown in table 3.3 and additional HMEC lines (total of 61 lines analyzed, 51 unique HMECs and 4 immortalized HMEC lines). Heatmap was created from log2 transformed data using R studio with a scale of 5. Hierarchical clustering was performed. Yellow indicates high expression, black indicates average expression, and blue indicates low expression relative to universal human breast RNA.

Figure 3.3: ciHMEC Lines Can Form Primary Mammospheres Primary mammosphere forming assays were used to determine the number of mammosphere forming cells (MFC) in stably and conditionally immortalized HMEC lines. A preliminary mammosphere forming assay (a-b) was used to select HMEC lines for further analysis. Cells were cultured in ultra-low attachment 96-well plates for 7 days. (a) Representative images of mammosphere formation in three HMEC lines (563, HME-CC, and 876). (b) Examples of no mammosphere formation in three HMEC lines (76N TERT, MCF10A, 847). (c) Extreme limiting dilution assays (ELDA) were performed to determine the number of MFC in each HMEC line. Only lines that had mammosphere in preliminary mammosphere forming assay were tested. Error bars represent 95% confidence interval. Scale bars indicate 100uM. Data provided courtesy of Dr. Gat Rauner.

Figure 3.4: Comparison of Lipofectamine 2000 and Fugene 6 reagents ERE-luciferase results for ciHMEC 910 comparing the efficiency of lipofectamine 2000 and fugene 6 transfection reagents. Relative light units for (a) luciferase and (b) renilla transfection control showed higher readings for lipofectamine. No luciferase readings were detected for fugene transfected cells. (c) Ratio of luciferase to renilla, relative to lipofectamine control treated cells.

Figure 3.5: Increasing amounts of ESR1 DNA in ERE-luciferase assays ERE-luciferase assay performed with 76N Tert cells with increasing amounts of ESR1 plasmid DNA. Relative light units shown for (a) luciferase and (b) renilla transfection control. (c) Ratio of luciferase to renilla shows saturating levels of activity at 0.125ug of ESR1, which was used for further experiments. With higher amounts of DNA, decreases in both luciferase and renilla were observed.

a b

d e

Figure 3.6: Benzophenone-3 and propylparaben dose response curves in HMECs ERE-luciferase assay dose response curves for (a-c) benzophenone-3 (BP3) and (d-f) propylparaben (PP) in three stably immortalized HMEC lines: 76N Tert, ME16C2, and HME-CC. Activity is shown relative to transactivation with 10nM E2 and ESR1 (gray lines). Blue lines indicate ESR1 transfection, red for ESR2, and purple for both receptors. Green dotted lines indicate transactivation with 10nM E2 and ESR2.

Figure 3.7: Limited variation in response to E2, BP3, and PP in HMECs Four stably immortalized (HME-CC, 76N Tert, MCF10A, ME16C2) and eight conditionally immortalized HMEC lines were transfected with ESR1 for luciferase assays with control or single doses of E2, BP3, or PP. 10nM E2, 30uM BP3, and 10uM PP induced significant transactivation compared to control in all cell lines (p < 0.05). Parity status and age are shown for ciHMECs. Activity is shown relative to experimental control for each individual line. Luciferase readings were normalized to renilla transfection control.

Figure 3.8: FOXA1 and GATA3 expression in ciHMECs RT-qPCR analysis of FOXA1 and GATA3 expression in seven ciHMEC lines, with MCF7 cells as a comparison. Gene expression is not normalized.

Figure 3.9: NIH 3T3 contamination in ciHMEC lines Immunofluorescence staining for vimentin (red), a marker of fibroblasts and epithelial cells, and a species specific human mitochondrial marker (green), which only labels human cells. Differential trypsinization was performed before plating cells for staining. Representative images of (a) a primary HMEC line, which has never been cocultured with NIH 3T3 cells, shown in (b). Two ciHMEC lines tested had low or very low carryover of mouse feeder cells (c-d), one line had significant levels of contamination (e), and one line had almost no human cells remaining (f). Passage numbers indicated. Scale bars are 100uM.

100

CHAPTER 4

ESTROGEN-INDUCED RESPONSES IN INDUCIBLE ERΑ HMECS

Introduction/Rationale

Early studies examining the effects of estrogen in the human breast utilized breast biopsies from donors during different stages of the menstrual cycle. However, these studies were rarely able to follow individual donors across menstrual stages and comparisons had to be made using pooled population data. As the levels of endogenous estrogen (Dall and

Britt, 2017) and responses to estrogen (Dunphy et al., 2020) vary among women, a better model for studying estrogen signaling in the human breast was needed. Breast cancer cell lines, such as T47D and MCF7 cells, have been used extensively to study the effects of estrogen receptor signaling, but they are not reflective of the normal breast epithelium.

These cancer cells acquire mutations during carcinogenesis and express higher levels of

ERα than normal breast epithelium, which has heterogenous ERα expression (Battersby et al., 1992; Söderqvist et al., 1993; Gabrielson et al., 2016). Another key difference is that breast cancer cell lines use autocrine signaling pathways to induce proliferation after estrogen treatment (Tan et al., 2009). In normal breast epithelium, paracrine signaling is hypothesized to be responsible for estrogen-induced proliferation as proliferating cells do not express ERα (Clarke et al., 1997; Russo et al., 1999; Shoker et al., 1999b). ERα positive proliferating cells are only observed in hyperplasia and carcinoma samples (Shoker et al.,

1999b). Secretion of growth factors including AREG, FGF, and EGF are required for mediating ERα induced proliferation through paracrine signaling (Ciarloni et al., 2007;

Fillmore et al., 2010; Harrison et al., 2013). For these reasons, breast cancer cell lines are not an ideal model for studying estrogen-induced responses in the normal breast.

101

Primary human mammary epithelial cell (HMEC) models are a cell model which is reflective of the normal breast, but these were originally limited by senescence and loss of

ERα in 2D culture. Multiple methods to bypass HMEC senescence were developed including transformation with HPV E6 and E7, human telomerase (TERT), Myc, SV40 large T-antigen, zinc-finger protein ZNF217, and BMI-1. It is known, however, that loss of p53, retinoblastoma protein, and p16INK4a occurs as a result of these transformations

(Wang et al., 1998; Dyson et al., 1989; Münger et al., 1989; Scheffner et al., 1990;

Huschtscha et al., 2001; Nonet et al., 2001; Dimri et al., 2002; Martinez-Zapien et al., 2016;

Fischer et al., 2017). Anchorage independent growth, reduced apoptotic responses, and altered gene expression resulting from these methods also suggest transformation can occur

(Wang et al., 1998; Thibodeaux et al., 2009). However, these HMEC models are still one of the better options for studying normal breast biology.

As an alternative to stable immortalization, another model called conditional reprogramming was developed involving expansion of primary HMECs by treatment with a rho-associated protein kinase (ROCK) inhibitor and co-culture with irradiated mouse fibroblasts (Liu et al., 2012; Palechor-Ceron et al., 2013). Studies have suggested that conditionally immortalized HMECS (ciHMECs) created with this method represent a stem like epithelial population (Suprynowicz et al., 2012, 2017). Other culture methods for

HMECs have also noted to result in loss of differentiation markers (Breindel et al., 2017).

One group has reported successful use of the ciHMEC method to expand ERα positive luminal epithelial cells, but this population decreases during passaging (Jin et al., 2017).

Our experience with this method has proven that feeder cell contamination is a large potential issue. We also have found that stably immortalized cells appear to behave

102

similarly to ciHMECs in reporter assays of estrogen-induced transactivation. Others have used a combination of conditional reprogramming and Tert immortalization to obtain ERα positive HMECs (Kumar et al., 2018).

More recent models of the human breast have been developed which retain ERα positive luminal epithelial cells, but these have their own limitations. Recent studies have identified protocols for development of 3D breast organoids from primary HMECs and cancerous breast tissue (Sokol et al., 2016; Sachs et al., 2018; Djomehri et al., 2019; Florian et al., 2019; Goldhammer et al., 2019). Breast organoids contain luminal and basal epithelial cell populations, can organize into ductal structures, retain the ERα positive cell population, and are responsive to hormone treatment. We have found, however, that propagation of breast organoids from normal breast tissue is limited as they can only be expanded for a short number of passages. Another model, human breast explants, can also only be maintained for a limited amount of time (up to 4 weeks) but provide similar benefits to breast organoids. Breast explants from normal tissue retain all breast cell types, tissue morphology, ERα expressing luminal epithelial cells, and response to hormone treatment

(Zhuang et al., 2003; Eigeliene et al., 2006, 2008, 2012, 2016; Tanos et al., 2013; Dunphy et al., 2020). Ultimately, the use of breast explant models is limited by the amount of tissue obtained by donors and therefore only a finite number of experiments can be performed.

Due to current limitations of breast organoid and explant models, HMECs are still one of the better models for studying human breast biology. However, even after bypassing the senescence block to expand HMEC populations, the ERα expressing population of epithelial cells is lost over time (Fridriksdottir et al., 2015; Jin et al., 2017; Lee et al., 2018).

Expression of ERα is limited to luminal epithelial cells in the normal breast. Luminal

103

epithelial cells make up 10-25% of cells isolated from human breast tissue (Garbe et al.,

2009) but may be more prevalent in breast tissue of older women (Lee et al., 2015). Many have hypothesized that the ERα positive luminal cell population is lost in 2D culture due to reduced adherence (Soule and McGrath, 1986; Keller et al., 2010), faster growth of basal epithelial cells (Keller et al., 2010; Fridriksdottir et al., 2015), triggering stress responses in luminal epithelial cells (Lee et al., 2018), changes to the microenvironment (Miyano et al., 2017), or epigenetic reprogramming that results in the loss of the differentiated phenotype (Keller et al., 2010; Breindel et al., 2017). With optimal conditions, ERα positive luminal epithelial cells are still lost in early passages and the number of passages is ultimately limited by the original percentage present (Lee et al., 2018). This means that

HMEC models primarily represent basal epithelial cell populations, and in order to study the effects of estrogen in these cells we have to introduce exogenous ERα expression.

In order to study ERα induced responses in HMECs, many have tried to introduce expression of this receptor with mixed results. Previous studies have shown that overexpression of ESR1 in the ER negative breast cancer line MDA-MB-231 is not able to induce normal estrogen-induced proliferation (Garcia et al., 1992; Lazennec and

Katzenellenbogen, 1999) and instead causes activation of anti-proliferative pathways through aberrant cell cycle regulation (Moggs et al., 2005). Additionally, while exogenous expression of ESR1 in MDA-MB-231 cells can induce target gene TFF1 expression, PGR upregulation is not restored (Lazennec and Katzenellenbogen, 1999; Moggs et al., 2005).

In MCF10A cells, exogenous expression of ESR1 has been shown to lead to a modest increase in proliferation with E2 treatment when EGF is removed from media (Pilat et al.,

1996; Abukhdeir et al., 2006; Pugazhendhi and Darbre, 2010). Additionally, expression

104

of estrogen receptor targets genes TFF1 and PGR was increased by E2 treatment in

MCF10A cells (Abukhdeir et al., 2006). These studies illustrate that in breast cancer and normal HMEC cell lines, restored expression of ERα alone is not sufficient to restore estrogen responsiveness.

Exogenous expression of ERα may not be able to induce normal biological responses to estrogen for a number of reasons. For example, in ERα negative basal breast cancer cells, epigenetic silencing of estrogen receptor target genes, including PGR (Leu et al., 2004) is present, which may explain why exogenous expression does not impart normal

E2 responses. Abnormal responses induced in breast epithelial cells during overexpression of ERα may also be due to prolonged high expression of this receptor. MCF7 breast cancer cells demonstrate estrogen independent growth and expression of ER target genes with overexpression of exogenous ERα (Tolhurst et al., 2011). Overexpression of ERα in

MDA-MB-231 breast cancer cells only results in E2-induced proliferation and target gene response when used in combination with overexpression of other regulatory factors, including forkhead box A1 (FOXA1) and GATA binding protein 3 (GATA3) (Kong et al.,

2011). Expression of ESR1, FOXA1, and GATA3 in E2 treated MDA-MB-231 cells results in a gene expression profile more like E2 treated MCF7 cells, though these expression profiles are not identical. In HMECs, estrogen- induced proliferation can be induced when exogenous ESR1 is combined with exogenous expression of Polycomb-group gene BMI1

(Duss et al., 2007). Additionally, transduction with ESR1 and BMI1 in HMECs induces expression of target genes GREB1 and PRLR but does not affect PGR or TFF1 expression.

These results demonstrate that even expression of these factors in combination does not fully induce estrogen regulated gene response. A recent study in mouse mammary

105

epithelium demonstrated that forced expression of ERα is lost over time, suggesting either silencing of ERα or selection against ERα positive epithelial cells (Cornelissen et al.,

2019). Contrary to the results of Kong et al. in MDA-MB-231 breast cancer cells,

Cornelissen et al. found that exogenous expression of ERα combined with FOXA1 and

GATA3 was not sufficient to induce estrogen receptor target gene expression (Areg and

Ccnd1) or estrogen-induced proliferation in mouse mammary epithelial cells.

The goal of this study was to develop inducible ERα HMEC lines and utilize this model to study factors which may enhance or restrict estrogen signaling and sensitivity to estrogen. We used 4 stably immortalized HMEC lines: 76N Tert (Band and Sager, 1989;

Shamanin and Androphy, 2004), MCF10A (Soule et al., 1990), HME-CC (Troester et al.,

2004), and ME16C2 (Troester et al., 2004). The 76N Tert, HME-CC, and ME16C2 HMEC lines were immortalized by expression of telomerase (TERT). The MCF10A line is spontaneously immortalized, likely through the loss of p16INK4A, p14ARF, and p15INK4B tumor suppressors (Iavarone and Massague, 1997). All 4 HMEC lines are considered basal epithelial cell lines with expression of KRT5 and KRT14 (Prat et al.,

2013), although some groups have shown the 76N Tert and MCF10A lines also have some expression of luminal epithelial markers KRT8 and KRT18 (Zhao et al., 2010; Qu et al.,

2015).

In order to investigate estrogen-induced responses in our 4 HMEC lines and examine the role of estrogen receptor coregulators, we developed a doxycycline inducible

ESR1 expression construct using an inducible lentiviral backbone. We hypothesized that

ERα induced responses among individuals may be influenced by individual expression of estrogen receptor coregulators. Induced expression of ESR1 with 100ng/mL of

106

doxycycline was sufficient to increase E2-induced transactivation in luciferase reporter assays. Although 2 of our 4 inducible ERα HMEC showed even higher transactivation levels with increased doses of doxycycline, we limited our dose to 100ng/mL to avoid adverse metabolic effects of doxycycline (Ahler et al., 2013). This dose of doxycycline was sufficient to induce ERα protein levels comparable to MCF7 cells in all 4 lines. With induced expression of ERα, 3 of our 4 HMEC lines demonstrated modest proliferation with estrogen treatment. Expression of estrogen receptor target genes AREG and PGR was unregulated by estrogen treatment, however. Conditioned media from 2 iERα HMEC lines suggests that the estrogen-induced growth in these HMEC lines is due to cell intrinsic factors, supporting the idea that estrogen receptor coregulators may play an important role.

However, expression of FOXA1 and GATA3 did not correspond with E2-induced proliferation in our 4 HMEC lines. This work suggests FOXA1 and GATA3 are not sufficient to restore estrogen responses in HMECs and that there may be other coregulators or other cell intrinsic factors influencing estrogen sensitivity.

Materials and Methods

pIND-ESR1 Construct Design

An N-terminal FLAG tagged inducible ESR1 expression construct was generated using the pINDUCER14 vector (Meerbrey et al., 2011). First, FLAG sequence was amplified from pFLAG-CMV-2 (Andersson et al., 1989) using forward primer 5’-

ATACCGGTACCATGGACTACAAAGACGATGACGAC-3’ and reverse primer 5’-

TCGACCGGTACGCGTGCGATCGCTGAATTCGCGGCAAG-3’. This sequence was cleaned using the Monarch PCR and DNA Cleanup Kit (NEB #T1030) and ligated into

107

pINDUCER14 after digestion of both plasmids with AgeI, dephosphorylation with shrimp alkaline phosphatase (NEB #M0371S), and extraction from gel electrophoresis (DNAland

Scientific #GP1001). Sequencing was performed to confirm insertion.

ESR1 coding sequence was amplified from pIRES-hrGFPII-ESR1 containing the

ESR1 coding sequence (Open Biosystems #MHS6278-211691051) in the pIRES-hrGFPII vector (Stratagene #240157). ESR1 was amplified with forward primer 5’-

GCAGAAATGACCATGACCCTCCACACCAAAGC-3’ and reverse primer 5’-

TAAACGCGTTCAGACCGTGGCAGGGAAACCCT-3’. ESR1 was ligated into pINDUCER14-FLAG by digestion with EcoRI and MluI and cleaned as described above.

Two linker sequences were inserted between FLAG and ESR1 to keep the sequence in frame (Linker A: 5’-AATTGCGCGATCGCGG-3’ and Linker B: 5’-

AATTCCGCGATCGCGC-3’) . Sequencing of the final construct (pINDUCER14-FLAG-

ESR1, or pIND-ESR1) confirmed insert orientation, all inserts were in frame, and that the

ESR1 sequence was identical to the Homo Sapiens ESR1 gene (primers: forward 5’-

CGGTGGGAGGCCTATATAAG-3’, reverse 5’- ACTTATATACGGTTCTCCCC-3’). pIND-ESR1 expresses a constitutive GFP reporter and tetracycline inducible ERα with N- terminal FLAG tag. Cloning was performed by Puneet Singh in the AMB program.

Cell Culture

76N Tert, MCF10A, ME16C2, and HME-CC cells were grown in F-media: DMEM

-pyruvate (Gibco #11965-092), Ham’s F12 (Gibco #11765-054), 5% FBS, 250ng/mL hydrocortisone (Sigma #H4001), 10ng/mL human epidermal growth factor (Tonobo

Biosciences #21-8356-U100), 8.6ng/mL cholera toxin (Millipore Sigma #227035), 10 ug/mL human insulin (Sigma #I9278-5ML), and 1X antibiotic/antimycotic (Caisson Labs

108

ABL02-100ML). 239T cells were cultured in DMEM:F12 (Sigma #D8900) with 10%

FBS, 15ug/mL gentamycin (Gibco #15750-060), and 1X antibiotic/antimycotic. T47D cells were grown in DMEM:F12 supplemented with 10% FBS, 15ug/mL gentamycin,

2mM L-Glutamine, and 1X antibiotic/antimycotic and cleared in 10% CSS, 2mM L-

Glutamine, 10ug/mL human insulin, and MEM media for 48-72 hours prior to experiments.

All cells were incubated at 37˚C with 5% CO2 and passaged every 2-3 days.

Generation of inducible ERα HMEC lines

293T cells were lifted with 0.05% trypsin and plated in 60 mm tissue culture dishes at 2.5x106 cells/dish and left overnight. Transfection was performed next day using 3.5ug pIND-ESR1, 3ug psPAX2 (Addgene #12260, gag, pol, and rev packaging vector), and 2ug pMD2.G (Addgene #12259, vsv-g packaging vector) in antibiotic free media with

Lipofectamine 2000 (Invitrogen). After 24 hours, media was refreshed and replaced with

293T growth media. At 48 and 54 hours post initial transfection, viral media was collected and filtered using a 0.45-micron filter (Corning #431220) and added to HMEC lines in a

1:1 ratio with F-media twice, 6 hours apart. After 24 hours, viral media was replaced with

F-media. FACS was performed using pooled HMECs in 1% FBS/PBS by selecting for

GFP positive cells using FACSAria II (Becton-Dickinson). Uninfected parental HMECs were used as a negative control to set background fluorescence.

Cell Treatments

Control treatment media consisted of MEM media (51200-038, Gibco), 5% charcoal-dextran stripped serum (FB-04, Omega Scientific), and 2mM L-glutamine

(SH30034.01, Hyclone). 10 mM 17β-estradiol (E2758, Sigma-Aldrich) stock prepared in ethanol was diluted to prepare E2 treatment media, and 10mM ICI (Tocris #1047) in

109

ethanol was used to prepare ICI treatment media. Doxycycline was added to treatment media at 100ng/mL except in dose response experiments. Alamar blue proliferation assay control treatment media was the same as above with the addition of 10ug/mL human insulin.

ERE-Luciferase Assays

HMECs were plated in 24 well tissue culture plates at 1 x 105 cells/well in F-media.

After an overnight incubation, growth media was replaced with 0.3 mL/well Opti-MEM reduced serum media (11058-021, Gibco). Plasmid DNA and Lipofectamine 2000 reagent

(11668019, ThermoFisher) were diluted in Opti-MEM media before adding 200 µL/well.

Uninfected cells were transfected with 0.125 µg of pIND-ESR1, 1.5 µg of 3X ERE-TATA-

Luciferase reporter, and 0.02 µg pRL-CMV Renilla plasmid using Lipofectamine 2000

(Invitrogen). iERα HMECs were infected with 3X ERE-TATA-Luc and Renilla. After 6 hours of transfection, transfection media was replaced with 0.5 mL of control media. The next morning, treatment media containing 0-200ng/mL doxycycline was added. Promega

Dual Luciferase Reporter Assay was used to perform luciferase assays (E1910, Promega).

Cells were lysed in 1x Passive Lysis Buffer after 24-hour treatment and lysates stored at -

20°C until reading. Luciferase and Renilla activity in lysates were determined by using the Polar Star Optima plate reader (BMG Labtech). Luciferase values for each well were normalized to the amount of Renilla activity.

Western Blot for ERα

Cells were lysed in ice cold RIPA lysis buffer [50 mM Tris–HCl pH 8.0, 150 mM

NaCl, 1 mM EDTA, 1% Triton X-100, 1% Sodium deoxycholate, 0.1% SDS, 1% protease inhibitors (Sigma-Aldrich #P8340), 1% phosphatase inhibitor #2 (Sigma-Aldrich #P5726),

110

and 1% phosphatase inhibitor #3 (Sigma-Aldrich #P0044)]. Lysates were centrifuged at

13,000 rpm for 15 minutes at 4˚C to remove cellular debris. BCA protein assay was performed to quantify protein concentration (Thermo Scientific #23225). Equal amounts of protein (25ug or 28ug, as indicated) were separated using SDS-PAGE on 10% acrylamide gels under denaturing conditions and blotted onto PVDF membrane (Millipore

#IPVH00010). Blocking was performed using 5% nonfat dry milk in TBST (10mM Tris-

HCl pH 7.5, 150mM NaCl, 0.05% tween-20) for 1 hour. Blots were incubated with 1:100 anti-ERα (Abcam [Sp1] #ab16660, Lot #GR3202692-2) overnight at 4˚C. During optimization, other antibodies for ERα were also tested: MC-20 (1:500 Santa Cruz #sc-

542, Lot #H0415), F-10 (1:200 Santa Cruz #sc8002, Lot #D0708), and C1355 (1:1000

Millipore Sigma #06-935, Lot #3169865). See Table 4.1 for antibody information.

The next day, blots were washed 3 times with TBST and then incubated for 1 hour with HRP-conjugated secondary antibody (1:5000, GE Healthcare #NA934V). Bands were detected using enhanced chemiluminescence solution and imaged on G-box

(Syngene). The blots were washed with TBST and incubated with anti-β actin (1:5000,

Sigma #A1978, or Cell Signaling #4967 [for F-10 blot]) overnight at 4˚C. Washing, secondary antibody incubation (1:5000, GE Healthcare #NA931C), and detection were performed as described above. Expected molecular weights were 67 kDa for ERα and 42 kDa for β actin.

Alamar Blue Proliferation Assay

Cells were plated in 5% CSS, 2mM L-Glutamine, 10ug/mL insulin, and MEM media at 4000, 7000, or 10000 cells per well on five 96-well plates (one for each day), with

7 replicates per treatment and one empty row for a blank. The next day, treatment media

111

was added (control, 10nM ICI, 10nM E2) diluted in 5% CSS, 2mM L-Glutamine, 10ug/mL insulin, and MEM media. Each day, Alamar blue reagent was added (final concentration

10%) and plates were read 4, 8, and 24 hours after addition. BioTek Synergy 2 plate reader

(BioTek) readings were performed at 570nm and 600nm. Media was refreshed on day 3.

Plates were read for 5 days unless cells appeared confluent. Percent Alamar blue reduction was calculated using the Alamar blue protocol (DAL1100, Invitrogen).

Conditioned Growth Media

HMECs were plated at 1.6x106 cells/T75 in 5% CSS, 2mM L-Glutamine, 10ug/mL insulin, and MEM media. T47D cells were plated at 1.2x106 cells/T75 after clearing in

10% CSS, 2mM L-Glutamine, 10ug/mL insulin, and MEM media for 48-72 hours. The next day, treatment media was added (control, 10nM ICI, 10nM E2) diluted in 5% CSS,

2mM L-Glutamine, 10ug/mL insulin, and MEM media. 48 hours after treatment, media was collected and replaced with fresh media. Conditioned media was filtered with a 0.2- micron syringe filter, diluted 1:1 in fresh media, and used to treat cells in proliferation assays. After an additional 48 hours, this process was repeated, and conditioning cells were discarded.

RT-qPCR

RNA isolation was performed using TRIzol according to manufacturer’s recommendation (15596018, ThermoFisher Scientific). cDNA was synthesized using 1 ug of RNA with the Protoscript II First Strand cDNA synthesis kit (E6560L, New England

Biolabs). RT-qPCR was performed in a thermocycler (CFX96 Real-Time thermocycler,

BioRad). See Table 4.2 for primer sequences. Expression is relative to an inter-run calibrator (IRC) consisting of pooled cDNA from a subset of human breast tissue samples.

112

Results

pIND-ESR1 Construct is Functional in HMECs

To create HMEC lines with inducible ERα expression, a doxycycline inducible

ESR1 expression vector was created using the pINDUCER14 vector (Meerbrey et al.,

2011) and the Homo sapiens ESR1 coding sequence with an N-terminal FLAG tag, referred to as pIND-ESR1 (Figure 4.1). This construct contains an internal ribosome entry site

(IRES), which drives constitutive GFP expression (Figure 4.2). A human EF1-α promoter, from EEF1A1, drives expression of a reverse tetracycline-transactivator (rtTA), which will bind to doxycycline (Dox) when present and activate the tetracycline response element

(TRE2) to induce expression of FLAG-tagged ESR1. To determine if this construct was functional, we first performed transient transfection into the 76N Tert HMEC line and performed a luciferase reporter assay (Figure 4.3a-c). After 24-hour treatment with E2, there was a significant 3-fold increase in transactivation in the cells transfected with pIND-

ESR1 (p < 0.05), which was not observed in cells without the receptor (ERE reporter only controls). While there was a significant 1.8-fold increase in activity with E2 treatment in the cells without induced ESR1 expression (no Dox), there was a larger significant increase when 100ng/mL of doxycycline was added to induce ESR1 expression (p < 0.05). This suggested that the construct is indeed functional, but that there may be low levels of leaky expression in transient transfection experiments.

Generation of iERα HMEC lines

We next sought to generate 4 HMEC lines with inducible ESR1 expression (iERα) using the pIND-ESR1 vector. After lentiviral infection, each HMEC line was pooled and

113

FACS sorting was used to isolate populations which successfully incorporated pIND-ESR1 and therefore expressed GFP constitutively (Figure 4.4, Figure 4.6, Figure 4.8, Figure

4.10). Each HMEC line was also checked post FACS to determine the percentage of GFP expressing cells, and these lines were called 76N Tert-ESR1, MCF10A-ESR1, ME16C2-

ESR1, and HME-CC-ESR1 (Figure 4.5, Figure 4.7, Figure 4.9, Figure 4.11). The 76N

Tert-ESR1 were 90% GFP positive, MCF10A-ESR1 were 70% positive, ME16C2-ESR1 were 99% positive, and HME-CC-ESR1 were 90% positive. These 4 HMEC lines are our iERα HMECs.

Luciferase Reporter Assays in iERα HMECs

After generating the four iERα HMEC lines, we performed luciferase assays to test the function of the inducible ESR1 with increasing concentrations of doxycycline (Figure

4.12 through Figure 4.14). For all 4 iERα HMECs the amount of transactivation with

10nM E2 treatment was similar for each dose of doxycycline, except 0ng/mL. A significant dose dependent increase was observed at higher doses of doxycycline for the

MCF10A-ESR1 and HME-CC-ESR1 lines (Figure 4.13 and Figure 4.14, p < 0.05), but this was not present in the other 2 lines. In the MCF10A-ESR1 and ME16C2-ESR1 lines, there was significant activation upon treatment with E2 with 0ng/mL Dox, suggesting there may be leaky expression of ESR1 in these cell lines (Figure 4.13 and Figure 4.15, p <

0.05). This was not present in the 76N Tert-ESR1 and HME-CC-ESR1 lines (Figure 4.12 and Figure 4.14, p > 0.05). In the MCF10A-ESR1 line, the level of transactivation with leaky receptor expression and E2 treatment was not significantly different from 100ng/mL or 150ng/mL with E2 treatment after Renilla normalization (p > 0.05), but the raw

114

luciferase readings were significantly lower (Figure 4.13a and c, p < 0.05). Renilla readings were significantly lowered in the 0ng/mL doxycycline and E2 treated cells, which caused discrepancy between the normalized and raw data (Figure 4.13b, p < 0.05).

Similarly, the ME16C2-ESR1 line had significantly lower renilla values for cells treated with any dose of doxycycline and 10nM E2 (Figure 4.15b, p < 0.05). There was significant background activity in the ME16C2-ESR1 cells treated with control media and any dose of doxycycline. This background activity is not due to normalization, as the raw luciferase values were significantly higher than the no dox treated control cells (Figure 4.15a).

In order to compare activity between these 4 HMEC lines, we set transactivation relative to the 76N Tert-ESR1 cells. When activity for all HMECs was set relative to the

76N Tert-ESR1 line, transactivation was similar in the 76N Tert-ESR1 and HME-CC-ESR1 lines, slightly lower in the MCF10A-ESR1 line, and higher in the ME16C2-ESR1 line

(Figure 4.16 and Figure 4.17). However, because the MCF10A-ESR1 and ME16C2-ESR1 had decreases in renilla readings with doxycycline and E2 treatment, we decided to compare fold change for each cell line in raw luciferase values, renilla values, and normalized luciferase values. In the 76N Tert-ESR1 line, the normalized fold change with doxycycline and E2 treatment compared to control was 5 to 8-fold (Figure 4.18a) and was similar in the raw luciferase values (Figure 4.18b), which was expected because the renilla values remained unchanged. The MCF10A-ESR1 line had a normalized fold change of 3 to 5.6-fold (Figure 4.19a), which was like the fold change in raw luciferase values for the doxycycline treated cells (Figure 4.19b). The increase in transactivation observed in the uninduced cells treated with E2 was skewed by the drop in renilla values, as the fold change in raw luciferase values was 1.8 compared to 2.7 in the normalized values. This suggests

115

that there is a small increase in activity in the MCF10A-ESR1 cells with doxycycline treatment to induce ESR1, though the leaky expression is still present. The HME-CC-ESR1 cells had a 3 to 5-fold increase in activity with Dox and E2 treatment and, similarly to the

76N Tert-ESR1 cells, this did not differ when comparing the normalized fold change

(Figure 4.20a) and raw luciferase values (Figure 4.20b).

Lastly, in the ME16C2-ESR1 cells the normalized fold change in cells treated with doxycycline and E2 was 37 to 60-fold (Figure 4.21a) compared to 23 to 40-fold in the raw luciferase values (Figure 4.21b), indicative of the effects of lowered renilla values.

Interestingly, the uninduced ME16C2-ESR1 cells had the highest fold change in both normalized and unnormalized readings (11-fold for both). This is likely because in the doxycycline treated cells, the control treatment had significant background activity which was not observed in the uninduced cells (Figure 4.15). This background activity was reduced with 10nM ICI treatment, which suggests that there was a small amount of estrogenic activity in the control treated cells. This data suggests that the ME16C2-ESR1 cells are highly sensitive to E2 treatment in luciferase reporter assays, 76N Tert-ESR1 and

HME-CC-ESR1 are approximately equal in response, and the MCF10A-ESR1 cells may be less sensitive.

ERα Protein Expression in iERα HMECs

To check if ERα protein was expressed in the iERα HMEC lines, western blot analysis was performed. Antibody optimization was performed using a total of 4 anti-ERα antibodies: mouse monoclonal F-10, rabbit monoclonal MC-20, rabbit monoclonal Sp1, and rabbit polyclonal C1355. MCF7 cell lysates were used as a positive control and parental 76N Tert lysates for a negative control in blots to examine ERα expression in 76N

116

Tert-ESR1 lysates with 100ng/mL Dox or 200ng/mL Dox. The expected molecular weight for ERα was 67 kDa. MC-20 and C1355 antibodies detected bands in all samples, including the negative control (Figure 4.22b and d). The MC-20 antibody detected two bands between 50 kDa and 75 kDa, while the C1355 antibody detected 3 bands in the 76N

Tert lysates and 2 bands in the MCF7 lysate ranging from 37 kDa to 75 kDa. The F-10 antibody also detected 2 bands in the MCF7 lysate and 76N Tert-ESR1 lysates, one around

75 kDa and one between 37 kDa and 50 kDa (Figure 4.22a). Interestingly, the larger protein band was not detected in the 76N Tert parental lysate. The Sp1 antibody appeared to be the most specific for ERα, as only one band was detected, and this band was absent in the 76N Tert parental lysate (Figure 4.22c). Incubating these blots with anti-β-actin antibody confirmed protein was loaded evenly (Figure 4.23). For all future ERα westerns, the Sp1 antibody was used. It is likely that the other ERα antibodies were demonstrating nonspecific binding.

Next, each of our 4 iERα HMEC lines was examined for ERα protein expression.

In all 4 lines, ERα protein was detected upon doxycycline treatment and was not detected in the absence of doxycycline (Figure 4.24). Therefore, if any of the lines have leaky ESR1 expression (as indicated in luciferase assays), it does not lead to detectable protein levels.

Quantification of the ERα detected in these 4 lines, after normalization to β-actin, revealed that 10nM E2 treatment does not appear to downregulate ERα expression in these cell lines

(Figure 4.25, Figure 4.26, Figure 4.27, and Figure 4.28). 76N Tert-ESR1, MCF10A-

ESR1, and HME-CC-ESR1 appear to have ERα levels approximately equal to that in MCF7 cells with 100ng/mL doxycycline treatment. The ME16C2-ESR1 line may have more ERα than the MCF7 cells (Figure 4.24 and Figure 4.28), but this would need to be confirmed

117

with more replicates of MCF7 cell lysates. These results indicated that all 4 iERα HMECs express ERα proteins at levels which could carry out estrogen-induced responses.

Estrogen-Induced Proliferation and ERα Target Gene Responses

After confirmation that the 4 iERα HMECs produce ERα protein, we next sought to determine if this induced ERα could produce a biological response. In breast cancer cell lines and normal mammary gland epithelial cells, it is well documented that treatment with estrogen induces cell proliferation. Therefore, Alamar Blue proliferation assays were performed with each cell line. With induced ERα expression and 10nM E2 treatment for

4 days, the 76N Tert-ESR1, MCF10A-ESR1, and HME-CC-ESR1 HMEC lines showed modest but significant increases in proliferation compared to control or 10nM ICI treated cells by day 4 (Figure 4.29a-d and Figure 4.30a). The proliferation induced resulted in an approximately 20-25% increase in Alamar Blue reduction. The ME16C2-ESR1 cells, however, did not demonstrate any E2-induced proliferation (Figure 4.30b).

Next, E2 induction of ERα target genes AREG and PGR were examined after 24- hour treatment with E2 and 72-hour treatment with doxycycline. PGR was not detected in any of the iERα HMECs (data not shown). AREG was detected in all 4 lines, but its expression did not appear to be regulated by E2 treatment (Figure 4.31). Compared to untreated T47D breast cancer cells and normal breast tissue, AREG expression in the 76N

Tert-ESR1, HME-CC-ESR1, and ME16C2-ESR1 lines was constitutively high. MCF10A-

ESR1 expression of AREG was like that found in normal breast tissue and untreated T47D cells, but expression was not changed with 10nM E2 treatment. In 76N Tert-ESR1 cells without ERα induced, 10nM ICI treatment resulted in a small but significant increase in

118

AREG (p < 0.05). Treatment with 10nM E2 and doxycycline significantly reduced AREG expression (p < 0.05). In HME-CC-ESR1 cells, treatment with doxycycline and 10nM ICI increased AREG relative to control treated cells without ERα expression, though this was not significant when compared to the control doxycycline treated cells. Lastly, in the doxycycline treated ME16C2-ESR1 cells, treatment with 10nM ICI significantly increased

AREG while treatment with 10nM E2 significantly decreased expression (p < 0.05). It is clear from these results that the modest E2-induced proliferation observed in 3/4 iERα

HMECs does not correspond with AREG or PGR target gene regulation. In addition, the

76N Tert-ESR1 and ME16C2-ESR1 lines may exhibit negative regulation of AREG when

ERα is expressed after estrogen treatment.

E2-Induced Proliferation in iERα HMECs is Due to Cell Intrinsic Factors

We next performed conditioned growth media assays to determine if the modest

E2-induced proliferation observed in 3/4 iERα HMECs is due to cell intrinsic factors or secreted growth factor production. Previous studies in normal breast epithelium support a paracrine signaling model for estrogen-induced proliferation, where estrogen receptor positive luminal epithelial cells produce growth factors that stimulate the proliferation of nearby estrogen receptor negative cells. However, in breast cancer cells estrogen-induced proliferation can occur through autocrine signaling, as treatment with epidermal growth factor receptor (EGFR) inhibitors does not block proliferation.

In our iERα HMECs, we wanted to determine if autocrine or paracrine signaling was contributing to estrogen-induced proliferation. For this, we used 76N Tert-ESR1 and

MCF10A-ESR1 cells, as they both exhibited E2-induced growth. Conditioned growth

119

media assays were performed as outlined in Figure 4.32. Conditioned growth media was collected from these two iERα cell lines 2 days and 4 days after treatment with control,

10nM ICI, or 10nM E2 and treatment with 100ng/mL doxycycline. Media was also collected from T47D cells with the same treatments. The conditioned media filter sterilized and added to parental HMEC lines lacking ERα (76N Tert and MCF10A) on day 1 and day

3 of proliferation assays. 76N Tert parental cells exhibited no significant change in proliferation with conditioned media from doxycycline treated 76N Tert-ESR1 cells in our first experiment (Figure 4.33). In a follow up experiment with T47D conditioned media as an additional treatment, there was a small but significant increase in proliferation with

E2 treatment in the 76N Tert-ESR1 media treated cells (Figure 4.34), but analysis of the fold change between day 2 and day 5 Alamar Blue reduction revealed no significant increase from control treated cells. No significant change was observed with conditioned media from T47D cells (Figure 4.34). Comparison of the fold change in Alamar Blue reduction between normal growth media, T47D conditioned media, and 76N Tert-ESR1 conditioned media treated cells revealed a similar fold change between all treatment groups

(Figure 4.34d), suggesting that the 76N Tert parental cells are not responsive to the conditioned growth media from 76N Tert-ESR1 or T47D cells. In MCF10A parental cells, no significant change in proliferation was observed in treatment with MCF10A-ESR1 conditioned media (Figure 4.35b), and small but significant increases or decreases were observed in normal growth media or T47D conditioned media with E2 treatment (Figure

4.35a and Figure 4.35c). Comparison of fold change in Alamar Blue reduction among treatment groups revealed a small but significant increase in proliferation when treated with

T47D control or 10nM ICI conditioned media (Figure 4.35d). This data suggests that the

120

proliferation observed in our 3/4 iERα HMECs is due to intracellular factors and not secretion of factors into the media.

To confirm the results of our conditioned media experiments from 76N Tert and

MCF10A parental cell lines, we performed the same experiments with T47D breast cancer cell lines. Our lab and numerous others have shown that T47D cell proliferation is increased with estrogen treatment. We wanted to determine if conditioned media from

76N Tert-ESR1 or MCF10A-ESR1 cells could enhance the proliferative response of T47D cells, which would suggest there is a secreted factor present in the media. T47D cells treated with conditioned media from 76N Tert-ESR1 cells exhibited a significant increase in proliferation with 10nM E2 media, and this increase was enhanced with media from

10nM E2 treated 76N Tert-ESR1 cells (Figure 4.36a-b). However, this increased proliferation is likely due to the high amounts of AREG produced by the 76N Tert-ESR1 cells (Figure 4.31), as proliferation was also enhanced in T47D cells treated with conditioned media from control and 10nM ICI treated 76N Tert-ESR1 cells. Additionally,

T47D cells treated with conditioned media from 76N Tert-ESR1 cells without doxycycline treatment showed a significant increase, which suggests the increased proliferation is due to high basal AREG levels in the 76N Tert line (Figure 4.36c). T47D cells treated with conditioned media from MCF10A-ESR1 cells exhibited a significant increase in proliferation with 10nM E2 treatment and was slightly enhanced by treatment with conditioned media from MCF10A-ESR1 cells treated with 10nM E2 (Figure 4.37).

However, since the increase was also observed in cells treated with media from MCF10A-

ESR1 cells without doxycycline treatment, it is likely not the mechanism driving E2- induced proliferation in our MCF10A-ESR1 cells. These results indicate that any increased

121

proliferation observed in T47D cells treated with media from 76N Tert-ESR1 cells is likely due to constitutively high AREG expression in these cells, as the enhanced proliferation was not seen with MCF10A-ESR1 conditioned media, which are not producing high levels of AREG (Figure 4.31). This can be confirmed with further experiments using inhibitors of epidermal growth factor receptor to block proliferation due to AREG. Collectively, this data suggests that the E2-induced proliferation in 3/4 iERα HMECs is due to cell intrinsic factors, not secreted factors.

FOXA1 and GATA3 Expression is Not Sufficient for ERα Induced Responses

As recent literature has suggested that the ERα coactivators FOXA1 and GATA3 are essential for carrying out estrogen-induced biological responses, especially in previously ER negative cell lines with exogenous ERα expression, we checked mRNA expression of these two cell intrinsic factors in our parental HMEC lines. If this hypothesis is correct, we would expect the 3 HMEC lines which responded to E2 with proliferation

(76N Tert-ESR1, MCF10A-ESR1, HME-CC-ESR1) to have higher expression of these two factors or a lack of expression in the unresponsive ME16C2-ESR1 line. The 76N Tert cells had the highest expression of FOXA1, followed by the MCF10A line, HME-CC line, and the ME16C2 cells (Figure 4.38a). While the 76N Tert and MCF10A cells had significantly higher expression of FOXA1 than the ME16C2 cells, expression between the HME-CC and

ME16C2 were not significantly different. Additionally, the amount of FOXA1 expressed in the ME16C2 cells was similar to that seen in normal breast tissue. GATA3 expression was highest in the MCF10A cells, followed closely by the 76N Tert cells (Figure 4.38b).

The expression in these two cell lines was comparable to normal breast tissue. The HME-

122

CC and ME16C2 cell lines, in contrast, had significantly lower GATA3 expression.

However, there was also no significant difference between the HME-CC and ME16C2

GATA3 expression levels. This data suggests that, while FOXA1 and GATA3 may be necessary for ERα induced responses, expression of these factors is not sufficient to impart

E2-induced responses. A summary of responses from each iERα HMEC generated in this study is shown in Table 4.3.

Discussion

Although Tert and spontaneously immortalized HMECs lose expression of p16INK4A, p14ARF, and/or p15INK4B, our results suggest that they are a useful model for studying regulation of estrogen-induced responses in breast epithelial cells. The 4 iERα

HMEC lines developed here are capable of estrogen response element transactivation upon

E2 treatment, express levels of ERα similar to MCF7 cells, and 3 of 4 lines demonstrate modest E2-induced proliferation. Like many other primary breast epithelial cell models, these HMECs exhibit primarily basal epithelial keratin expression profiles (Prat et al.,

2013), although some have suggested that the 76N Tert and MCF10A lines have luminal

KRT8/KRT18 expressing subpopulations (Zhao et al., 2010; Qu et al., 2015). Even though these HMECs are primarily basal epithelial cells, our results illustrate that they can be used to study factors which enhance or restrict estrogen-induced responses.

Exogenous expression of ESR1 in both breast cancer cell lines (MDA-MB-231) and

HMECs has been shown to bind E2 as an agonist (Pilat et al., 1996; Lazennec and

Katzenellenbogen, 1999) and generate significant levels of transactivation with estrogen response element reporters (Pilat et al., 1996; Lazennec and Katzenellenbogen, 1999;

123

Pugazhendhi and Darbre, 2010; Kong et al., 2011), but this does not always equate to E2- induced biological responses. In MDA-MB-231 basal breast cancer cells, exogenous ESR1 expression has consistently been demonstrated to result in decreased proliferation (Jiang and Jordan, 1992; Zajchowski et al., 1993; Lazennec and Katzenellenbogen, 1999).

However, these cells can still control mRNA expression of some growth factors (TGFα and TGFβ2) associated with proliferation (Jeng et al., 1994) and increase expression of some ER target genes (TFF1) with E2 treatment (Lazennec and Katzenellenbogen, 1999).

In HMECs, response to E2 treatment after exogenous ESR1 expression has produced conflicting reports. Most studies report that MCF10A or other HMECs with exogenous ESR1 are capable of transactivation in reporter assays (Pilat et al., 1996; Wolf et al., 2007; Kong et al., 2011) but have no change in proliferation with E2 treatment

(Zajchowski et al., 1993; Pilat et al., 1996; Thomas et al., 1998). However, others report modest increases in MCF10A proliferation of 1.3-fold (Pugazhendhi and Darbre, 2010) or

1.5 to 2-fold increases in MCF10A clones expressing ESR1 in the absence of EGF

(Abukhdeir et al., 2006). All of our 4 iERα HMEC lines are responsive to E2 treatment in reporter assays (Figure 4.12, Figure 4.13, Figure 4.14, and Figure 4.15), but only 3 of 4

(76N Tert-ESR1, MCF10A-ESR1, and HME-CC-ESR1) demonstrated a 20-25% increase in proliferation with 10nM E2 treatment (Figure 4.29 and Figure 4.30), compared to ~60% in estrogen responsive T47D breast cancer cells (Figure 4.36a and Figure 4.37a). While our proliferation media does not contain added EGF, it is likely that our serum still contains residual amounts of EGF and treatment with EGFR inhibitors may increase our proliferative phenotype in these lines. In studies where no change in proliferation was observed, MCF10A cells demonstrated no change in expression of ER target genes TFF1

124

or PGR mRNA (Pilat et al., 1996) while our MCF10A-ESR1 and other HMEC lines expressed no detectable PGR (data not shown). Studies that showed proliferation with E2 treatment in MCF10A cells also observed induced expression of TFF1 and PGR

(Abukhdeir et al., 2006). We did not examine expression of TFF1 in our HMECs, but expression of AREG and PGR did not appear to be regulated by E2 treatment (Figure 4.31).

Recent papers have suggested that ERα mediated responses to E2 treatment can be induced through exogenous expression of other factors. Most primary HMECs do not express of ERα, but transduction with ESR1 and BMI1, a Polycomb-group gene which is capable of suppressing the p53 and Rb pathways and increases stem cell renewal, allows for growth of HMECs stably expressing ERα (Duss et al., 2007). These ERα/BMI1 expressing HMECs exhibit a 50% increase in proliferation with E2 treatment and increases in expression of ER target genes GREB1, PGR, and PRLR, while expression of ER target gene TFF1 remains unchanged. This suggests that many but not all E2-induced responses can be generated in HMECs with BMI1. Other work suggests that expression of ERα coregulators FOXA1 and GATA3 are necessary for generating normal E2-induced responses in ER negative breast cells. While expression of ESR1 alone is able to transactivate ERE luciferase reporters, co-expression of FOXA1 and GATA3 is required to induce proliferative responses (Kong et al., 2011). In MDA-MB-231 cells, a 2-fold increase in proliferation is observed with E2 treatment and these 3 factors; in BT-549 breast cancer cells, a 40% increase in proliferation is observed with ESR1, FOXA1, and GATA3 compared to a 30% decrease in proliferation with ESR1 alone and E2 treatment.

Comparison of E2-ER mediated transcriptional profiles of MDA-MB-231 cells with ESR1,

125

FOXA1, and GATA3 are closer to MCF7 profiles, but expression of key target genes is still missing (Kong et al., 2011).

Our conditioned media experiments support the hypothesis that other intrinsic cellular factors are involved in E2-induced proliferation in our 3 iERα HMECs (Figure

4.33 through Figure 4.37), but our data suggests that expression of FOXA1 and GATA3 are not sufficient to impart E2-induced proliferation in our ME16C2-ESR1 HMEC line

(Figure 4.38). Conditioned media from 2 iERα HMECs which demonstrate E2-induced proliferation is not able to increase proliferative responses in T47D cells (Figure 4.36 and

Figure 4.37) or increase proliferation in parental HMECs without ERα expression (Figure

4.33, Figure 4.34, and Figure 4.35). Expression of FOXA1 and GATA3 is significantly higher in the 76N Tert-ESR1 and MCF10A-ESR1 lines, but expression of these two factors is not significantly different between the ME16C2-ESR1 and HME-CC-ESR1 cells, which proliferate in response to E2 (Figure 4.29 and Figure 4.30).

It is interesting that Kong et al. suggested that lack of FOXA1 and GATA3 was responsible for the abnormal phenotypes in MDA-MB-231 breast cancer cells with exogenous ESR1 because others have demonstrated low basal expression of FOXA1 (Wolf et al., 2007) in these cells, but co-expression of ESR1 and GATA3 is not sufficient to generate E2-induced proliferative responses (Kong et al., 2011). This would suggest that increased expression of FOXA1 is needed, which our results contradict. Our results are supported by another recent paper that expressed exogenous ESR1 in mouse mammary epithelial cells, where co-expression of FOXA1 and GATA3 also was not sufficient to impart responsiveness to E2 treatment (Cornelissen et al., 2019). This group, however, attributed this result to the differences in number of FOXA1 and GATA3 binding sites

126

between mice and humans, suggesting mice rely less on these pioneer factors. Our results suggest that, in HMECs, FOXA1 and GATA3 expression is not sufficient to induce complete responsiveness to estrogen. These results in combination suggest the presence of other coregulators involved in regulation of E2-induced responses, which may influence sensitivity to estrogen in individuals as well.

127

Table 4.1: Antibody information for Western Blot Analysis Antibodies used in western blots on HMEC cell lysates.

Name of Antibody, Dilution Target Manufacturer Species Catalog Number Used Estrogen Rabbit, receptor alpha Sp1, #ab16660 Abcam 1:100 monoclonal (ERα) Estrogen Santa Cruz Rabbit, receptor alpha MC-20, #sc-542 1:500 (discontinued) monoclonal (ERα) Estrogen Mouse, receptor alpha F-10, #sc-8002 Santa Cruz 1:200 monoclonal (ERα) Estrogen Millipore Rabbit, receptor alpha C1355, #06-935 1:1000 Sigma polyclonal (ERα) Mouse, β-actin AC-15, #A1978 Sigma 1:5000 monoclonal Rabbit, β-actin #4967 Cell Signaling 1:5000 polyclonal

Table 4.2: RT-qPCR primer sequences Primer sequences used in RT-qPCR of HMECs. Used on BioRad CFX thermocycler.

Gene Sequence: 5’ to 3’ Fwd – CGG AGA ATG CAA ATA TAT AGA GCA C Amphiregulin (AREG) Rev – CAC CGA AAT ATT CTT GCT GAC A Fwd – TTT AAG AGG GCA ATG GAA GG Progesterone receptor (PGR) Rev – CGG ATT TTA TCA ACG ATG CAG Forkhead box protein A1 Fwd – GGA ACA GCT ACT ACG CAG AC (FOXA1) Rev – ATG TTG CCG CTC GTA GTC AT GATA binding protein 3 Fwd – CAC AAA ATG AAC GGA CAG AAC A (GATA3) Rev – GTT GTG GTG GTC TGA CAG TT

128

Table 4.3: Summary of HMEC responses Summary of data from chapter 4 illustrating the responses of each cell line across all assays performed. Clonality of each iERα HMEC line can be inferred from the initial percent GFP positive cell population during and after FACS sorting. Fold change in luciferase assays is relative to 10nM ICI treated cells. ERα protein expression is shown as a percentage of quantified levels in MCF7 cells (n = 2 for iERα HMECs, n = 1 for MCF7). AREG expression is shown relative to fresh breast tissue IRC. Significant E2-induced proliferation is also indicated.

ERα % GFP Fold Relative Initial Expression Cell positive change in Expression E2-induced % GFP (% of Line after luciferase of AREG proliferation positive MCF7 FACS assays and PGR level) AREG: 9- Similar 76N Tert 5% 90% 5-8 15 Yes (107%) PGR: none Similar AREG: 0.3 MCF10A 17% 70% 3-5.6 Yes (99%) PGR: none HME- Similar AREG: 5-8 19% 90% 3-5 Yes CC (130%) PGR: none AREG: 5- Increased ME16C2 72% 99% 37-60 23 No (720%) PGR: none

129

pINDUCER14-FLAG-ESR1 (pIND-ESR1)

Figure 4.1: Inducible ESR1 expression in pINDUCER14 backbone A FLAG tagged ESR1 coding sequence was cloned into the lentiviral pINDUCER14 backbone, to create pIND-ESR1. An internal ribosome entry site (IRES) drives constitutive GFP expression. The human EF1-α promoter (EEF1A1) drives expression of reverse tetracycline-transactivator (rtTA), which will bind to the tetracycline response element (TRE2) only in the presence of doxycycline to induce ESR1 transcription. ESR1 has a FLAG tag expressed on the N-terminal region. Cloning performed by Puneet Singh, former AMB master’s student.

130

Figure 4.2: GFP expression in 293T cells transfected with pIND-ESR1 293T cells were transfected with the pIND-ESR1 construct and confirmed constitutive GFP expression in all cells containing the plasmid. (a) Phase contrast image of 293T cells and (b) fluorescent image of cells with filter for GFP. Images were taken one day after transfection.

131

Figure 4.3: Transient transfection of 76N Tert cells with pIND-ESR1 Transfection of 76N Tert cells with pIND-ESR1. 76N Tert cells were transfected with pIND-ESR1, an ERE luciferase reporter, and a renilla transfection control. No receptor control cells (ERE-Ctrl and ERE-E2) were transfected with reporter but no pIND-ESR1. Cells were treated with control, 10nM ICI, or 10nM E2 for 24 hours. (a) Luciferase, (b) Renilla, and (c) normalized data shown. Results showed some leaky expression of ESR1 in the cells without doxycycline treatment, as a significant increase in transactivation with 10nM E2 was observed. However, transactivation with doxycycline treatment and 10nM E2 produced a significantly higher level of transactivation. Error bars indicate SEM. (* p < 0.05).

132

a b

Figure 4.4: FACS of 76N Tert cells after lentiviral infection with pIND-ESR1 After infection with pIND-ESR1, (a) 76N Tert cells (b) expressing GFP indicated successful incorporation of pIND-ESR1. This subpopulation was selected using FACS sorting (c) with the parameters shown. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Parental 76N Tert cells were used as a negative control to set background autofluorescence. Approximately 5% of 76N Tert cells were GFP positive. Images were taken with 10x objective.

133

Figure 4.5: Post FACS check of 76N Tert-ESR1 After FACS for GFP expressing 76N Tert cells, an aliquot was checked for purity. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Approximately 90% of the sorted cell population expresses GFP, which are considered the 76N Tert-ESR1 cell line.

134

a b

Figure 4.6: FACS of MCF10A cells after lentiviral infection with pIND-ESR1 After infection with pIND-ESR1, (a) MCF10A cells (b) expressing GFP indicated successful incorporation of pIND-ESR1. This subpopulation was selected using FACS sorting (c) with the parameters shown. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Parental MCF10A cells were used as a negative control to set background autofluorescence. Approximately 17% of MCF10A cells were GFP positive. Images were taken with 10x objective.

135

Figure 4.7: Post FACS check of MCF10A-ESR1 After FACS for GFP expressing MCF10A cells, an aliquot was checked for purity. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Approximately 70% of the sorted population expresses GFP, which are considered the MCF10A-ESR1 cell line.

136 a b

Figure 4.8: FACS of HME-CC cells after lentiviral infection with pIND-ESR1 After infection with pIND-ESR1, (a) HME-CC cells (b) expressing GFP indicated successful incorporation of pIND-ESR1. This subpopulation was selected using FACS sorting (c) with the parameters shown. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, and FSC-W and FSC-H confirmed selection of single cells. PE-Texas Red showed minimal autofluorescence. Parental HME-CC cells were used as a negative control to set background autofluorescence. Approximately 19% of HME-CC cells were GFP positive. Images were taken with 20x objective.

137

Figure 4.9: Post FACS check of HME-CC-ESR1 After FACS for GFP expressing HME-CC cells, an aliquot was checked for purity. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, and FSC-W and FSC-H confirmed selection of single cells. PE-Texas Red showed minimal autofluorescence. Approximately 90% of the sorted cell population expresses GFP, which are considered the HME-CC-ESR1 cell line.

138

a b

Figure 4.10: FACS of ME16C2 cells after lentiviral infection with pIND-ESR1 After infection with pIND-ESR1, (a) ME16C2 cells (b) expressing GFP indicated successful incorporation of pIND-ESR1. This subpopulation was selected using FACS sorting (c) with the parameters shown. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Parental ME16C2 cells were used as a negative control to set background autofluorescence. Approximately 72% of ME16C2 cells were GFP positive. Images were taken with 10x objective.

139

Figure 4.11: Post FACS check of ME16C2-ESR1 After FACS for GFP expressing ME16C2 cells, an aliquot was checked for purity. Side scatter area (SSC-A) and forward scatter area (FSC-A) were used to isolate viable cells, SSC width (W) and height (H) were used to remove doublets, FSC-W and FSC-H confirmed selection of single cells, and PE-Texas Red was used to detect autofluorescence. Approximately 99% of the sorted cell population expresses GFP, which are considered the ME16C2-ESR1 cell line.

140

Figure 4.12: Doxycycline dose response in 76N Tert-ESR1 Luciferase reporter assays in the 76N Tert-ESR1 line with 24-hour control, 10nM ICI, or 10nM E2 treatment and either 0ng/mL, 100ng/mL, 150ng/mL, or 200ng/mL doxycycline. (a) Luciferase, (b) Renilla, and (c) Luciferase normalized to Renilla. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

141

Figure 4.13: Doxycycline dose response in MCF10A-ESR1 Luciferase reporter assays in the MCF10A-ESR1 line with 24-hour control, 10nM ICI, or 10nM E2 treatment and either 0ng/mL, 100ng/mL, 150ng/mL, or 200ng/mL doxycycline. (a) Luciferase, (b) Renilla, and (c) Luciferase normalized to Renilla. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

142

Figure 4.14: Doxycycline dose response in HME-CC-ESR1 Luciferase reporter assays in the HME-CC-ESR1 line with 24-hour control, 10nM ICI, or 10nM E2 treatment and either 0ng/mL, 100ng/mL, 150ng/mL, or 200ng/mL doxycycline. (a) Luciferase, (b) Renilla, and (c) Luciferase normalized to Renilla. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

143

Figure 4.15: Doxycycline dose response in ME16C2-ESR1 Luciferase reporter assays in the ME16C2-ESR1 line with 24-hour control, 10nM ICI, or 10nM E2 treatment and either 0ng/mL, 100ng/mL, 150ng/mL, or 200ng/mL doxycycline. (a) Luciferase, (b) Renilla, and (c) Luciferase normalized to Renilla. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

144

Tert to76N Activity Relative

Figure 4.16: Luciferase Reporter Assays with 76N Tert-ESR1 and MCF10A-ESR1 ERE-luciferase reporter assays with 76N Tert-ESR1 and MCF10A-ESR1 cell lines with increasing concentrations of doxycycline. Cells were treated with control, 10nM ICI, or 10nM E2 for 24 hours. Data is normalized to renilla transfection control and set relative to 76N Tert-ESR1 no dox control. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

145

* * *

Tert to76N Activity Relative

Figure 4.17: Luciferase Reporter Assays with HME-CC-ESR1 and ME16C2-ESR1 ERE-luciferase reporter assays with HME-CC-ESR1 and ME16C2-ESR1 cell lines with increasing concentrations of doxycycline. Cells were treated with control, 10nm ICI, or 10nM E2 for 24 hours. Data is normalized to renilla transfection control and set relative to 76N Tert-ESR1 no dox control. Experiment was performed at the same time as Figure 4.12. Error bars indicate SEM. Significance is calculated with comparison to control treatment of same doxycycline dose unless otherwise indicated by lines. (* p < 0.05).

146

Figure 4.18: Fold change in ERE-Luc responses in 76N Tert-ESR1 Fold change in ERE-luciferase reporter data from 76N Tert-ESR1 with increasing concentrations of doxycycline relative to ICI treatment. Cells were treated with control, 10nm ICI, or 10nM E2 for 24 hours. Fold change is shown for (a) luciferase normalized to renilla and (b) both raw luciferase and ratio (normalized data). Error bars indicate SEM. (* p < 0.05).

147

Figure 4.19: Fold change in ERE-Luc responses in MCF10A-ESR1 Fold change in ERE-luciferase reporter data from MCF10A-ESR1 with increasing concentrations of doxycycline relative to ICI treatment. Cells were treated with control, 10nm ICI, or 10nM E2 for 24 hours. Fold change is shown for (a) luciferase normalized to renilla and (b) both raw luciferase and ratio (normalized data). Error bars indicate SEM. (* p < 0.05).

148

Figure 4.20: Fold change in ERE-Luc responses in HME-CC-ESR1 Fold change in ERE-luciferase reporter data from HME-CC-ESR1 with increasing concentrations of doxycycline relative to ICI treatment. Cells were treated with control, 10nm ICI, or 10nM E2 for 24 hours. Fold change is shown for (a) luciferase normalized to renilla and (b) both raw luciferase and ratio (normalized data). Error bars indicate SEM. (* p < 0.05).

149

Figure 4.21: Fold change in ERE-Luc responses in ME16C2-ESR1 Fold change in ERE-luciferase reporter data from ME16C2-ESR1 with increasing concentrations of doxycycline relative to ICI treatment. Cells were treated with control, 10nm ICI, or 10nM E2 for 24 hours. Fold change is shown for (a) luciferase normalized to renilla and (b) both raw luciferase and ratio (normalized data). Error bars indicate SEM. (* p < 0.05).

150

a F-10 b MC-20

1 2 3 4 4 3 2 1

Sp1 C1355 c d 1 2 3 4 4 3 2 1

Figure 4.22: ERα antibody optimization for Western blot analysis Summary of antibody optimization for Western blot analysis, using 35ug of protein per sample. Samples: 1- MCF7 (positive control), 2- 76N Tert parental cells (negative control), 3- 76N Tert-ESR1 with 100ng/mL doxycycline, 4- 76n Tert-ESR1 with 200ng/mL doxycycline. Results with (a) F-10 monoclonal antibody, (b) MC-20 monoclonal antibody, (c) Sp1 monoclonal antibody, and (d) C1355 polyclonal antibody for ERα.

151

a F-10 b MC-20

1 2 3 4 4 3 2 1

c Sp1 d C1355 1 2 3 4 4 3 2 1

Figure 4.23: β actin expression during ERα antibody optimization for Western blot Summary of antibody optimization for Western blot analysis, using 35ug of protein per sample. Samples: 1- MCF7 (positive control), 2- 76N Tert parental cells (negative control), 3- 76N Tert-ESR1 with 100ng/mL doxycycline, 4- 76n Tert-ESR1 with 200ng/mL doxycycline. β-actin protein expression in blots previously probed with (a) F-10 monoclonal antibody, (b) MC-20 monoclonal antibody, (c) Sp1 monoclonal antibody, and (d) C1355 polyclonal antibody for ERα.

152

a b

c d

Figure 4.24: ERα protein expression in iERα HMEC lines compared to MCF7 Western blot analysis showing detection of ERα protein using 1:100 dilution of ERα clone Sp1 antibody from Abcam. Amount of protein loaded was either 28ug (or 25 ug if indicated). β-actin is shown as a loading control. ERα protein was detected in (a) 76N Tert-ESR1, (b) MCF10A-ESR1, (c) HME-CC-ESR1, and (d) ME16C2-ESR1 cells in the presence of doxycycline. MCF7 cell lysates were used as a positive control, and parental HMECs were included as a negative control for ERα.

153

a ERα

b β-actin

c ERα normalized to β-actin

Figure 4.25: Quantification of ERα protein expression in 76N Tert-ESR1 Quantification of western blot data from for 76N Tert-ESR1 HMEC line for (a) ERα, (b) β-actin, and (c) ERα normalized to β-actin. 2 replicates were used for each condition, except for the MCF7 cell positive control. Error bars indicate SEM.

154

a ERα

b β-actin

c ERα normalized to β-actin

Figure 4.26: Quantification of ERα protein expression in MCF10A-ESR1 Quantification of western blot data from for MCF10A-ESR1 HMEC line for (a) ERα, (b) β-actin, and (c) ERα normalized to β-actin. 2 replicates were used for each condition, except for the MCF7 cell positive control. Error bars indicate SEM.

155

a ERα

b β-actin

c ERα normalized to β-actin

Figure 4.27: Quantification of ERα protein expression in HME-CC-ESR1 Quantification of western blot data from for HME-CC-ESR1 HMEC line for (a) ERα, (b) β-actin, and (c) ERα normalized to β-actin. 2 replicates were used for each condition, except for the MCF7 cell positive control. Error bars indicate SEM.

156

a ERα

b β-actin

c ERα normalized to β-actin

Figure 4.28: Quantification of ERα protein expression in ME16C2-ESR1 Quantification of western blot data from for ME16C2-ESR1 HMEC line for (a) ERα, (b) β-actin, and (c) ERα normalized to β-actin. 2 replicates were used for each condition, except for the MCF7 cell positive control. Error bars indicate SEM.

157

a b

c d

Figure 4.29: E2-induced proliferation in 76N Tert-ESR1 and MCF10A-ESR1 HMECs (a-b) 76N Tert-ESR1 and (c-d) MCF10A-ESR1 HMECs were treated with 100 ng/mL doxycycline and control, 10nM ICI, or 10nM E2 media for 4 days. The amount of proliferation was measured each day by Alamar blue reduction. Media was refreshed on day 3. Two experiments were performed in each cell line, shown in separate graphs. Error bars indicate SEM. Significance calculated using T-test (* p < 0.05).

158

Figure 4.30: E2-induced proliferation in HME-CC-ESR1 and ME16C2-ESR1 HMECs (a) HME-CC-ESR1 and (b) ME16C2-ESR1 HMECs were treated with 100ng/mL doxycycline and control, 10nM ICI, or 10nM E2 media for 4 days. The amount of proliferation was measured each day by Alamar blue reduction. Media was refreshed on day 3. Error bars indicate SEM. Significance calculated using T-test (* p < 0.05).

159

Figure 4.31: AREG expression in iERα HMEC lines After 2-day treatment with doxycycline to induce ESR1 expression, HMECs were treated with control, 10nM ICI, or 10nM E2 for 24 hours. Gene expression from RT-qPCR is shown relative to normal breast tissue. Untreated T47D cells are included for reference. Error bars indicate SEM. (* p < 0.05).

160

Figure 4.32: Overview of conditioned media assay Schematic overview of conditioned media assay. Briefly, iERα HMECs (76N Tert-ESR1 and MCF10A-ESR1) and T47D cells were treated with control, 10nM ICI, or 10nM E2 media for 48 hours. After 48 hours, the conditioned media was collected, filter sterilized using a 0.2-micron filter, diluted 1:1 with fresh media, and transferred to HMEC parental cells (76N Tert or MCF10A) and T47D cells. New media was added to iERα HMECs and T47D cells to condition for another 48 hours and again added to the parental HMECs and T47D cells. Alamar Blue growth assays were performed on parental HMECs and T47D cells, where conditioned growth media was added on day 1 and day 3 of treatment. Cells were also grown in normal growth media (no conditioned media) as a control.

161

Figure 4.33: Conditioned Growth Media Assay in 76N Tert Alamar Blue proliferation assay was performed on 76N Tert parental cells (without ERα) using (a) normal growth media or (b) conditioned growth media from doxycycline treated 76N Tert-ESR1 cells, diluted 1:1 with normal growth media. (c) Fold change in Alamar Blue reduction between day 5 and day 1 cells is shown for normal and conditioned media experiments. Error bars indicate SEM. (p < 0.05).

162

b * NS

Figure 4.34: Expansion of Conditioned Growth Media Assay in 76N Tert Alamar Blue proliferation assay was performed on 76N Tert parental cells (without ERα) using (a) normal growth media (b) conditioned growth media from doxycycline treated 76N Tert-ESR1 cells, or (c) conditioned growth media from T47D cells. (d) Fold change in Alamar Blue reduction between day 5 and day 2 cells is shown for normal and conditioned media (CM) experiments. Error bars indicate SEM. (* p < 0.05).

163

a * NS

c NS

Figure 4.35: Conditioned Growth Media Assay in MCF10A Alamar Blue proliferation assay was performed on MCF10A parental cells (without ERα) using (a) normal growth media (b) conditioned growth media from doxycycline treated MCF10A-ESR1 cells, or (c) conditioned growth media from T47D cells. (d) Fold change in Alamar Blue reduction between day 5 and day 2 cells is shown for normal and conditioned media (CM) experiments. Error bars indicate SEM. (* p < 0.05).

164

b *

Figure 4.36: Conditioned Growth Media Assay in T47D with 76N Tert-ESR1 media Alamar Blue proliferation assay was performed on T47D cells using (a) normal growth media or (b) conditioned growth media from doxycycline treated 76N Tert-ESR1 cells, diluted 1:1 with normal growth media. (c) Fold change in Alamar Blue reduction between day 5 and day 1 is shown for normal and conditioned media (CM, with and without Dox) experiments. Error bars indicate SEM. (* p < 0.05).

165

a *

Figure 4.37: Conditioned Growth Media Assay in T47D with MCF10A-ESR1 media Alamar Blue proliferation assay was performed on T47D cells using (a) normal growth media or (b) conditioned growth media from MCF10A-ESR1 cells, diluted 1:1 with normal growth media. (c) Fold change in Alamar Blue reduction between day 5 and day 1 is shown for normal and conditioned media (CM, with and without Dox) experiments. Error bars indicate SEM. (* p < 0.05).

166

Figure 4.38: FOXA1 and GATA3 expression in HMEC lines RT-qPCR expression of (a) FOXA1 and (b) GATA3 in untreated HMECs. Gene expression is shown relative to normal breast tissue. Error bars indicate SEM. (* p < 0.05).

167

CHAPTER 5

SUMMARY AND FUTURE

Chronic lifetime exposure to estrogen through early menarche, late menopause, hormone replacement therapy, and high serum estrogen levels all increase a woman’s risk for developing breast cancer. These reproductive risk factors suggest that longer lifetime exposure to estrogen increases breast cancer risk. Accordingly, estrogen is known to contribute to tumor growth/initiation and treatment with SERMs lowers incidence of breast cancer. However, estrogen exposure can also decrease breast cancer risk through early in life pregnancy, which decreases breast cancer risk by up to 50%. High dose estrogen treatment is also an effective antitumor therapy for postmenopausal women with breast cancer and is thought to sensitize tumors to estrogen-induced apoptosis. This data suggests that high levels of estrogen exposure, depending on context, have antagonistic effects on breast cancer risk.

Although all women are exposed to estrogen, only 1 in 8 women will develop breast cancer during their lifetime. Familial breast cancer risk accounts for 27-30% of overall breast cancer risk, but 60-70% of familial breast cancer risk is due to currently unknown inheritable factors. Rodent models demonstrate genetic susceptibility to estrogen-induced mammary tumor development and estrogen-induced responses in the mammary gland which are not linked to any known breast cancer risk genes, suggesting unique sensitivity to estrogen exposure. We hypothesized that a subset of women may be more sensitive to estrogen and that these women may be at increased risk for developing breast cancer.

Responses to estrogen exposure in the breast are mediated through estrogen receptors alpha and beta (ERα and ERβ). Estrogen acts as a ligand for these receptors to

168

activate transcription of target genes in coordination with estrogen receptor coregulators, which modulate the activity of ERα and ERβ. Estrogen receptor coregulators are expressed in a tissue specific manner. Through the use of human breast ex-vivo and primary cell models, we sought to characterize the level of variation in estrogen receptor expression and estrogen-induced responses to determine if some individuals are more responsive to estrogen and if this correlates with ER expression or is due to downstream regulation of the estrogen signaling pathway.

In microarray data from over 100 donor breast tissue samples, we observed significant variation in ESR1 and ESR2 expression, and expression of these receptors changes with age. Our data for ESR1 expression agrees with previously published data for

ERα expression, suggesting that ESR1 expression can be used as an endpoint. To examine if estrogen receptor expression correlates with estrogen-induced biological endpoints, we utilized a human breast explant model with over 20 donor tissue samples. The breast explant model allowed us to examine responses within a single donor to control or E2 treatment, which is advantageous to other studies of ERα expression which relied on data gathered from women at a single time point. We again observed significant variation in

ESR1 and ESR2 expression, with up to 80 or 100-fold variation between individuals. For estrogen-induced endpoints, we examined transcriptional and functional E2 responses in our explant samples. Global transcription levels indicated by R-loop formation were similar between donors, suggesting that E2 induces transcriptional responses regardless of individual genetic background. When we looked at specific transcriptional targets, however, we observed quite variable responses. Expression of ER target genes AREG,

PGR, and TGFβ2 were consistently increased in nulliparous donors, but in parous donor

169

expression of these genes with E2 treatment increased in some donors and decreased in others. Some individual donors also had high basal levels of target gene expression.

For functional endpoints of estrogen treatment, we examined PR, PCNA, and apoptotic responses among donors. PR expression was significantly decreased in E2 treated explants compared to fresh tissue timepoints, but there was no significance between control and E2 treated explants. This data agrees with other published results in human breast tissue but differ from what is known in breast cancer cell lines. E2-induced proliferation, assayed by PCNA expression, revealed a trending increase with E2 treatment but overall expression varied widely among individuals. In mice, our lab has shown that irradiation induced apoptotic responses are enhanced by E2 treatment. However, in our human breast explants we did not observe any significant increase in apoptotic cells compared to control treated tissues, though responses again varied by individual.

Correlation analysis with E2 mediated transcriptional and functional responses revealed no significant correlation of ESR1 or ESR2 with any endpoints. These results indicate that while there is significant variation in estrogen receptor expression among individuals and also variation in E2-induced responses, estrogen receptor expression alone is not dictating sensitivity to E2 treatment. In order to expand upon these results, we began working with human primary breast epithelial cell models. Conditionally reprogrammed primary human mammary epithelial cells (ciHMECs) and stably immortalized HMECs were used as individual donor samples to examine estrogen-induced responses as well as responses to environmental xenoestrogens BP3 and PP.

Our primary HMECs enriched primarily for basal epithelial cell populations, which corresponds with what others have noted about the limited growth of luminal epithelial

170

cells in 2D culture. Others have also noted that the conditional reprogramming culture method generates an epithelial population with some stem cell like properties.

Accordingly, in mammosphere forming assays, we found that most ciHMEC lines were able to form mammospheres in ultra-low attachment conditions, compared to half of permanently immortalized HMEC lines. One main drawback to HMEC models is that most HMECs do not express estrogen receptors, so for our HMEC studies we used transient transfection of the estrogen receptors along with reporter constructs as an endpoint. Dose response experiments with E2, BP3, and PP treatment revealed that BP3 and PP are agonists for both ESR1 and ESR2, but transactivation is significantly higher with ESR1.

Characterization revealed variation in ciHMEC responses to E2 (2.5-fold), BP3 (2.3-fold), and PP (3.6-fold) treatment, but the activity between donors cell lines was more consistent than what we had observed with our human breast explant model. This data reveals that xenoestrogens BP3 and PP act as ligands for both ERα and ERβ, but their main activity is through ERα, and that xenoestrogen exposure has the potential to activate estrogen receptors in the majority of individuals and therefore could influence the effects of estrogen signaling on breast cancer development.

Our last model for studying variation in estrogen-induced responses was our inducible ERα HMEC lines. Using a doxycycline inducible vector, we generated an inducible ESR1 expression construct and stably infected 4 HMEC lines. Characterization of these lines using luciferase reporter assays revealed significant transactivation with E2 treatment in all 4 HMECs, but the amount of transactivation and potential leaky construct expression varied between the 4 HMEC lines. We confirmed ERα protein expression by western blot, and expression levels of ERα were comparable to MCF7 breast cancer cells.

171

For biologically relevant responses, we examined proliferation and target gene expression in our 4 iERα HMECs. In 3 of 4 iERα HMECs, we observed modest but significant proliferation induced by E2 treatment resulting in a 20-25% increase in proliferation.

However, target gene expression was no longer regulated by estrogen, as PGR was not expressed in any lines and AREG expression was consistently high in 3 of 4 HMEC lines.

The data from our 4 iERα HMEC lines complimented our observations from our human breast explant model, where we observe some variation in biological responses with

E2 treatment. In our 4 iERα HMECs, we observed significant proliferative responses to

E2 in 3 of 4 lines, but clearly E2-induced responses are not consistent with what is known in breast cancer cell lines despite having similar levels of ERα. We hypothesized that coregulators of estrogen receptor signaling expressed in these 4 cell lines could be modulating E2-induced responses, which was supported by the work of other labs showing that co-expression of FOXA1 and GATA3 with ESR1 is required in ER negative breast cancer cell lines to restore some E2-induced responses. In order to test this hypothesis, we first determined if the proliferative responses observed in 3 of 4 iERα HMECs was due to cell intrinsic factors. Conditioned media assays in 2 iERα HMEC and one breast cancer cell line revealed that the iERα HMEC proliferation does not appear to be stimulated by secretion of growth factors, which suggests that intracellular factors are responsible for modulating these responses. Examination of FOXA1 and GATA3 expression in our HMEC lines, however, demonstrated that expression levels of these factors is not correlated with

E2-induced proliferative responses. Future work with these cell lines will involve somatic cell hybrid formation with breast cancer cell lines and our 4 iERα HMECs in order to

172

determine if there are intracellular factors in our HMECs that are restricting E2-induced responses.

Collectively, these results demonstrate significant variation in estrogen receptor expression and estrogen-induced responses among individuals. This data mirrors results in rodent models which have demonstrated genetic susceptibility to estrogen-induced responses and tumor formation, where individual genetic background appears to amplify or suppress the effects of estrogen exposure. Our human breast explant model data demonstrated that estrogen receptor expression level is not correlated with E2-induced responses, suggesting that other factors in the estrogen signaling pathway are mediating individual sensitivity to estrogen. As estrogen receptor coregulators are known to modulate estrogen receptor activity and are expressed in a tissue specific manner, we hypothesized that these coregulators may be responsible for determining individual sensitivity to estrogen. Our iERα HMECs are a useful model for studying estrogen-induced responses and illustrate that cell intrinsic factors appear to mediate E2-induced proliferative responses. However, unlike other studies which suggest FOXA1 and GATA3 are necessary and sufficient for E2-induced responses, we found that FOXA1 and GATA3 expression is not sufficient to restore E2-induced biological responses.

In order to determine whether specific coregulators of estrogen receptor signaling may alter E2-induced biological responses and expression of ER target genes, a single cell

RNA sequencing (scRNA-seq) approach can be used with the models outlined here. Many recent studies have utilized scRNA-seq to examine the differentiation of basal and luminal progenitors in the mammary gland. One group suggests that the differentiation of progenitor cells into mature basal and luminal epithelial cells occurs along a continuum

173

(Bach et al., 2017), while others seem to agree that epithelial cells form distinct clusters of basal and luminal epithelial cells which can be further classified by expression profiles (Pal et al., 2017; Nguyen et al., 2018; Sun et al., 2018; Chen et al., 2019). In human breast tissue scRNA-seq analysis, 2 luminal cell populations and 1 basal cell population are identified, with some inter-individual variability noted in certain sub clusters (Nguyen et al., 2018; Chen et al., 2019). The luminal epithelial subtypes are defined as either secretory, which do not express estrogen receptors, or hormone-responsive, which express

ESR1 and PGR (Bach et al., 2017; Nguyen et al., 2018). While ESR1 is most highly expressed in hormone-responsive luminal epithelial cells, some cells from basal and secretory luminal clusters also express low levels of ESR1 in the human breast (Chen et al., 2019). For our future work, we would be interested in isolating the hormone-responsive luminal epithelial cell cluster to examine expression of estrogen receptor coregulators.

As estrogen receptor coregulators control estrogen-induced signaling inside ERα positive cells, scRNA-seq can be used to isolate hormone-responsive luminal epithelial cells from digested human breast donor tissue samples using previously identified profiles in the human breast (Nguyen et al., 2018; Chen et al., 2019). This cluster of cells can be further examined to select those which express ESR1, and expression of estrogen receptor coregulators can be compared within this cell population. Once these cells are selected, expression of ER coregulators can be compared between human breast tissue donors and compared to E2-induced responsiveness determined in human breast explant or HMEC experiments. Alternatively, scRNA-seq can be used to analyze coregulator expression in our 4 iERα HMEC lines to compare ER coregulator expression in E2-proliferative cell lines (76N Tert-ESR1, MCF10A-ESR1, and HME-CC-ESR1) to our unresponsive cell line

174

(ME16C2-ESR1) and ERα positive breast cancer cell lines T47D or MCF7. We hypothesize that these experiments would lead to the identification of ER coregulators which correlate with increased or decreased E2-responsiveness. These experiments would determine if ER coregulators are rate-limiting for estrogen receptor signaling.

Overall, we demonstrate that some individuals are more sensitive to estrogen exposure, this sensitivity is most likely not due to estrogen receptor levels alone, and cell intrinsic factors such as estrogen receptor coregulators appear to modulate these responses.

Further work to expand upon the role of estrogen receptor coregulators on mediating estrogen sensitivity is still needed, but these results provide a potential mechanism for the antagonistic roles of estrogen in mediating breast cancer risk. This data also suggests that a subpopulation of women is more sensitive to estrogen and thus may be more at risk of breast cancer development through estrogen exposure.

175

REFERENCES

Abbasi, S., Rasouli, M., Nouri, M., and Kalbasi, S. (2012). Association of estrogen receptor-α A908G (K303R) mutation with breast cancer risk. Int. J. Clin. Exp. Med. 6, 39–49.

Abukhdeir, A.M., Blair, B.G., Brenner, K., Karakas, B., Konishi, H., Lim, J., Sahasranaman, V., Huang, Y., Keen, J., Davidson, N., et al. (2006). Physiologic estrogen receptor alpha signaling in non-tumorigenic human mammary epithelial cells. Breast Cancer Res. Treat. 99, 23–33.

Adank, M.A., Jonker, M.A., Kluijt, I., Mil, S.E. van, Oldenburg, R.A., Mooi, W.J., Hogervorst, F.B.L., Ouweland, A.M.W. van den, Gille, J.J.P., Schmidt, M.K., et al. (2011). CHEK2*1100delC homozygosity is associated with a high breast cancer risk in women. J. Med. Genet. 48, 860–863.

Ahlbory-Dieker, D.L., Stride, B.D., Leder, G., Schkoldow, J., Trölenberg, S., Seidel, H., Otto, C., Sommer, A., Parker, M.G., Schütz, G., et al. (2009). DNA Binding by Estrogen Receptor-α Is Essential for the Transcriptional Response to Estrogen in the Liver and the Uterus. Mol. Endocrinol. 23, 1544–1555.

Ahler, E., Sullivan, W.J., Cass, A., Braas, D., York, A.G., Bensinger, S.J., Graeber, T.G., and Christofk, H.R. (2013). Doxycycline Alters Metabolism and Proliferation of Human Cell Lines. PLoS ONE 8.

Ahmed, S., Thomas, G., Ghoussaini, M., Healey, C.S., Humphreys, M.K., Platte, R., Morrison, J., Maranian, M., Pooley, K.A., Luben, R., et al. (2009). Newly discovered breast cancer susceptibility loci on 3p24 and 17q23.2. Nat. Genet. 41, 585–590.

Alamer, M., and Darbre, P.D. (2018). Effects of exposure to six chemical ultraviolet filters commonly used in personal care products on motility of MCF-7 and MDA-MB- 231 human breast cancer cells in vitro. J. Appl. Toxicol. 38, 148–159.

Albrektsen, G., Heuch, I., Hansen, S., and Kvåle, G. (2005). Breast cancer risk by age at birth, time since birth and time intervals between births: exploring interaction effects. Br. J. Cancer 92, 167–175.

Al-Dhaheri, M., Wu, J., Skliris, G.P., Li, J., Higashimato, K., wang, Y., White, K.P., Lambert, P., Zhu, Y., Murphy, L., et al. (2011). CARM1 is an important determinant of ERα-dependent breast cancer cell differentiation and proliferation in breast cancer cells. Cancer Res. 71, 2118–2128.

Andersson, S., Davis, D.L., Dahlbäck, H., Jörnvall, H., and Russell, D.W. (1989). Cloning, structure, and expression of the mitochondrial cytochrome P-450 sterol 26- hydroxylase, a bile acid biosynthetic enzyme. J. Biol. Chem. 264, 8222–8229.

176

Andersson, S., Sundberg, M., Pristovsek, N., Ibrahim, A., Jonsson, P., Katona, B., Clausson, C.-M., Zieba, A., Ramström, M., Söderberg, O., et al. (2017). Insufficient antibody validation challenges oestrogen receptor beta research. Nat. Commun. 8.

Anglian Breast Cancer Study Group (2000). Prevalence and penetrance of BRCA1 and BRCA2 mutations in a population-based series of breast cancer cases. Br. J. Cancer 83, 1301–1308.

Antoniou, A.C., and Easton, D.F. (2006). Models of genetic susceptibility to breast cancer. Oncogene 25, 5898–5905.

Arendt, L.M., St. Laurent, J., Wronski, A., Caballero, S., Lyle, S.R., Naber, S.P., and Kuperwasser, C. (2014). Human Breast Progenitor Cell Numbers Are Regulated by WNT and TBX3. PLoS ONE 9.

Asselin-Labat, M.-L., Sutherland, K.D., Barker, H., Thomas, R., Shackleton, M., Forrest, N.C., Hartley, L., Robb, L., Grosveld, F.G., van der Wees, J., et al. (2007). Gata-3 is an essential regulator of mammary-gland morphogenesis and luminal-cell differentiation. Nat. Cell Biol. 9, 201–209.

Aupperlee, M.D., Drolet, A.A., Durairaj, S., Wang, W., Schwartz, R.C., and Haslam, S.Z. (2009). Strain-specific differences in the mechanisms of progesterone regulation of murine mammary gland development. Endocrinology 150, 1485–1494.

Bach, K., Pensa, S., Grzelak, M., Hadfield, J., Adams, D.J., Marioni, J.C., and Khaled, W.T. (2017). Differentiation dynamics of mammary epithelial cells revealed by single- cell RNA sequencing. Nat. Commun. 8.

Bado, I., Nikolos, F., Rajapaksa, G., Wu, W., Castaneda, J., Krishnamurthy, S., Webb, P., Gustafsson, J.-Å., and Thomas, C. (2017). Somatic loss of estrogen receptor beta and p53 synergize to induce breast tumorigenesis. Breast Cancer Res. BCR 19.

Bado, I., Pham, E., Soibam, B., Nikolos, F., Gustafsson, J.-Å., and Thomas, C. (2018). ERβ alters the chemosensitivity of luminal breast cancer cells by regulating p53 function. Oncotarget 9, 22509–22522.

Band, V., and Sager, R. (1989). Distinctive traits of normal and tumor-derived human mammary epithelial cells expressed in a medium that supports long-term growth of both cell types. Proc. Natl. Acad. Sci. U. S. A. 86, 1249–1253.

Barrdahl, M., Canzian, F., Joshi, A.D., Travis, R.C., Chang-Claude, J., Auer, P.L., Gapstur, S.M., Gaudet, M., Diver, W.R., Henderson, B.E., et al. (2014). Post-GWAS gene–environment interplay in breast cancer: results from the Breast and Prostate Cancer Cohort Consortium and a meta-analysis on 79 000 women. Hum. Mol. Genet. 23, 5260– 5270.

177

Battersby, S., Robertson, B.J., Anderson, T.J., King, R.J., and McPherson, K. (1992). Influence of menstrual cycle, parity and oral contraceptive use on steroid hormone receptors in normal breast. Br. J. Cancer 65, 601–607.

Becker, K.A., Lu, S., Dickinson, E.S., Dunphy, K.A., Mathews, L., Schneider, S.S., and Jerry, D.J. (2005). Estrogen and progesterone regulate radiation-induced p53 activity in mammary epithelium through TGF-β-dependent pathways. Oncogene 24, 6345–6353.

Belotserkovskii, B.P., Tornaletti, S., D’Souza, A.D., and Hanawalt, P.C. (2018). R-loop generation during transcription: Formation, processing and cellular outcomes. DNA Repair 71, 69–81.

Benecke, A., Chambon, P., and Gronemeyer, H. (2000). Synergy between estrogen receptor α activation functions AF1 and AF2 mediated by transcription intermediary factor TIF2. EMBO Rep. 1, 151–157.

Berkenstam, A., Glaumann, H., Martin, M., Gustafsson, J.-Å., and Norstedt, G. (1989). Hormonal Regulation of Estrogen Receptor Messenger Ribonucleic Acid in T47DCO and MCF-7 Breast Cancer Cells. Mol. Endocrinol. 3, 22–28.

Bernardo, G.M., Lozada, K.L., Miedler, J.D., Harburg, G., Hewitt, S.C., Mosley, J.D., Godwin, A.K., Korach, K.S., Visvader, J.E., Kaestner, K.H., et al. (2010). FOXA1 is an essential determinant of ERα expression and mammary ductal morphogenesis. Development 137, 2045–2054.

Bocchinfuso, W.P., and Korach, K.S. (1997). Mammary Gland Development and Tumorigenesis in Estrogen Receptor Knockout Mice. J. Mammary Gland Biol. Neoplasia 2, 323–334.

Breindel, J.L., Skibinski, A., Sedic, M., Wronski-Campos, A., Zhou, W., Keller, P.J., Mills, J., Bradner, J., Onder, T., and Kuperwasser, C. (2017). Epigenetic Reprogramming of Lineage-Committed Human Mammary Epithelial Cells Requires DNMT3A and Loss of DOT1L. Stem Cell Rep. 9, 943–955.

Brinton, L.A., Hoover, R., and Fraumeni, J.F. (1983). Reproductive factors in the aetiology of breast cancer. Br. J. Cancer 47, 757–762.

Brinton, L.A., Schairer, C., Hoover, R.N., and Fraumeni, J.F. (1988). Menstrual factors and risk of breast cancer. Cancer Invest. 6, 245–254.

Brooks, J.D., Sung, J.S., Pike, M.C., Orlow, I., Stanczyk, F.Z., Bernstein, J.L., and Morris, E.A. (2018). MRI background parenchymal enhancement, breast density and serum hormones in postmenopausal women. Int. J. Cancer 143, 823–830.

Byford, J.R., Shaw, L.E., Drew, M.G.B., Pope, G.S., Sauer, M.J., and Darbre, P.D. (2002). Oestrogenic activity of parabens in MCF7 human breast cancer cells. J. Steroid Biochem. Mol. Biol. 80, 49–60.

178

Calafat, A.M., Wong, L.-Y., Ye, X., Reidy, J.A., and Needham, L.L. (2008). Concentrations of the Sunscreen Agent Benzophenone-3 in Residents of the United States: National Health and Nutrition Examination Survey 2003–2004. Environ. Health Perspect. 116, 893–897.

Calafat, A.M., Ye, X., Wong, L.-Y., Bishop, A.M., and Needham, L.L. (2010). Urinary Concentrations of Four Parabens in the U.S. Population: NHANES 2005–2006. Environ. Health Perspect. 118, 679–685.

Campbell, T.M., Castro, M.A.A., Gonçalves de Oliveira, K., Ponder, B.A.J., and Meyer, K.B. (2018). ER alpha binding by transcription factors NFIB and YBX1 enables FGFR2 signalling to modulate estrogen responsiveness in breast cancer. Cancer Res. 78, 410– 421.

Carroll, J.S., Liu, X.S., Brodsky, A.S., Li, W., Meyer, C.A., Szary, A.J., Eeckhoute, J., Shao, W., Hestermann, E.V., Geistlinger, T.R., et al. (2005). Chromosome-Wide Mapping of Estrogen Receptor Binding Reveals Long-Range Regulation Requiring the Forkhead Protein FoxA1. Cell 122, 33–43.

Chang, E.C., Charn, T.H., Park, S.-H., Helferich, W.G., Komm, B., Katzenellenbogen, J.A., and Katzenellenbogen, B.S. (2008). Estrogen Receptors α and β as Determinants of Gene Expression: Influence of Ligand, Dose, and Chromatin Binding. Mol. Endocrinol. 22, 1032–1043.

Chen, W., Morabito, S.J., Kessenbrock, K., Enver, T., Meyer, K.B., and Teschendorff, A.E. (2019). Single-cell landscape in mammary epithelium reveals bipotent-like cells associated with breast cancer risk and outcome. Commun. Biol. 2.

Cho, H., Aronica, S.M., and Katzenellenbogen, B.S. (1994). Regulation of Progesterone Receptor Gene Expression in MCF-7 Breast Cancer Cells: A Comparison of the Effects of Cyclic Adenosine 3’,5’-Monophosphate, Estradiol, Insulin-Like Growth Factor-I, and Serum Factors. Endocrinology 134, 658–664.

Chugh, R.M., Chaturvedi, M., and Yerneni, L.K. (2016). Exposure cell number during feeder cell growth-arrest by Mitomycin C is a critical pharmacological aspect in stem cell culture system. J. Pharmacol. Toxicol. Methods 80, 68–74.

Chugh, R.M., Chaturvedi, M., and Yerneni, L.K. (2017). An optimization protocol for Swiss 3T3 feeder cell growth-arrest by Mitomycin C dose-to-volume derivation strategy. Cytotechnology 69, 391–404.

Ciarloni, L., Mallepell, S., and Brisken, C. (2007). Amphiregulin is an essential mediator of estrogen receptor α function in mammary gland development. Proc. Natl. Acad. Sci. U. S. A. 104, 5455–5460.

Clarke, R.B., Howell, A., Potten, C.S., and Anderson, E. (1997). Dissociation between Steroid Receptor Expression and Cell Proliferation in the Human Breast. Cancer Res. 57, 4987–4991.

179

Clemons, M., and Goss, P. (2001). Estrogen and the Risk of Breast Cancer. N. Engl. J. Med. 344, 276–285.

Coelingh Bennink, H.J., Verhoeven, C., Dutman, A.E., and Thijssen, J. (2017). The use of high-dose estrogens for the treatment of breast cancer. Maturitas 95, 11–23.

Colletti, J.A., Leland-Wavrin, K.M., Kurz, S.G., Hickman, M.P., Seiler, N.L., Samanas, N.B., Eckert, Q.A., Dennison, K.L., Ding, L., Schaffer, B.S., et al. (2014). Validation of Six Genetic Determinants of Susceptibility to Estrogen-Induced Mammary Cancer in the Rat and Assessment of Their Relevance to Breast Cancer Risk in Humans. G3 GenesGenomesGenetics 4, 1385–1394.

Conway, K., Parrish, E., Edmiston, S.N., Tolbert, D., Tse, C.-K., Geradts, J., Livasy, C.A., Singh, H., Newman, B., and Millikan, R.C. (2005). The estrogen receptor-α A908G (K303R) mutation occurs at a low frequency in invasive breast tumors: results from a population-based study. Breast Cancer Res. 7, R871.

Conway, K., Parrish, E., Edmiston, S.N., Tolbert, D., Tse, C.-K., Moorman, P., Newman, B., and Millikan, R.C. (2007). Risk factors for breast cancer characterized by the estrogen receptor alpha A908G (K303R) mutation. Breast Cancer Res. 9, R36.

Cornelissen, L.M., Henneman, L., Drenth, A.P., Schut, E., de Bruijn, R., Klarenbeek, S., Zwart, W., and Jonkers, J. (2019). Exogenous ERα Expression in the Mammary Epithelium Decreases Over Time and Does Not Contribute to p53-Deficient Mammary Tumor Formation in Mice. J. Mammary Gland Biol. Neoplasia 24, 305–321.

Cotrim, C.Z., Fabris, V., Doria, M.L., Lindberg, K., Gustafsson, J.-Å., Amado, F., Lanari, C., and Helguero, L.A. (2013). Estrogen receptor beta growth-inhibitory effects are repressed through activation of MAPK and PI3K signalling in mammary epithelial and breast cancer cells. Oncogene 32, 2390–2402.

Coutts, A.S., and Murphy, L.C. (1998). Elevated Mitogen-activated Protein Kinase Activity in Estrogen-nonresponsive Human Breast Cancer Cells. Cancer Res. 58, 4071– 4074.

Covaleda, A.M.S., van den Berg, H., Vervoort, J., van der Saag, P., Ström, A., Gustafsson, J.-Å., Rietjens, I., and Murk, A.J. (2008). Influence of Cellular ERα/ERβ Ratio on the ERα-Agonist Induced Proliferation of Human T47D Breast Cancer Cells. Toxicol. Sci. 105, 303–311.

Cravero, D., Martignani, E., Miretti, S., Accornero, P., Pauciullo, A., Sharma, R., Donadeu, F.X., and Baratta, M. (2015). Generation of Induced Pluripotent Stem Cells from Bovine Epithelial Cells and Partial Redirection Toward a Mammary Phenotype In Vitro. Cell. Reprogramming 17, 211–220.

180

Cuzick, J., Sestak, I., Bonanni, B., Costantino, J.P., Cummings, S., DeCensi, A., Dowsett, M., Forbes, J.F., Ford, L., LaCroix, A.Z., et al. (2013). Selective oestrogen receptor modulators in prevention of breast cancer: an updated meta-analysis of individual participant data. Lancet 381, 1827–1834.

Dall, G.V., and Britt, K.L. (2017). Estrogen Effects on the Mammary Gland in Early and Late Life and Breast Cancer Risk. Front Oncol 7, 110.

Dennison, K.L., Samanas, N.B., Harenda, Q.E., Hickman, M.P., Seiler, N.L., Ding, L., and Shull, J.D. (2015). Development and Characterization of a Novel Rat Model of Estrogen-Induced Mammary Cancer. Endocr. Relat. Cancer 22, 239–248.

Dimri, G.P., Martinez, J.-L., Jacobs, J.J.L., Keblusek, P., Itahana, K., Lohuizen, M. van, Campisi, J., Wazer, D.E., and Band, V. (2002). The Bmi-1 Oncogene Induces Telomerase Activity and Immortalizes Human Mammary Epithelial Cells. Cancer Res. 62, 4736–4745.

Djomehri, S.I., Burman, B., Gonzalez, M.E., Takayama, S., and Kleer, C.G. (2019). A reproducible scaffold-free 3D organoid model to study neoplastic progression in breast cancer. J. Cell Commun. Signal. 13, 129–143.

Dunphy, K.A., Blackburn, A.C., Yan, H., O’Connell, L.R., and Jerry, D.J. (2008). Estrogen and progesterone induce persistent increases in p53-dependent apoptosis and suppress mammary tumors in BALB/c-Trp53(+/- )mice. Breast Cancer Res. BCR 10, R43–R43.

Dunphy, K.A., Black, A.L., Roberts, A.L., Sharma, A., Li, Z., Suresh, S., Browne, E.P., Arcaro, K.F., Ser-Dolansky, J., Bigelow, C., et al. (2020). Inter-Individual Variation in Response to Estrogen in Human Breast Explants. J. Mammary Gland Biol. Neoplasia.

Duss, S., André, S., Nicoulaz, A.-L., Fiche, M., Bonnefoi, H., Brisken, C., and Iggo, R.D. (2007). An oestrogen-dependent model of breast cancer created by transformation of normal human mammary epithelial cells. Breast Cancer Res. 9.

Dyson, N., Howley, P.M., Munger, K., and Harlow, E. (1989). The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 243, 934–937.

Easton, D.F., Pooley, K.A., Dunning, A.M., Pharoah, P.D.P., Thompson, D., Ballinger, D.G., Struewing, J.P., Morrison, J., Field, H., Luben, R., et al. (2007). Genome-wide association study identifies novel breast cancer susceptibility loci. Nature 447, 1087– 1093.

Eckert, R.L., Mullick, A., Rorke, E.A., and Katzenellenbogen, B.S. (1984). Estrogen Receptor Synthesis and Turnover in MCF-7 Breast Cancer Cells Measured by a Density Shift Technique. Endocrinology 114, 629–637.

181

Eeckhoute, J., Keeton, E.K., Lupien, M., Krum, S.A., Carroll, J.S., and Brown, M. (2007). Positive Cross-Regulatory Loop Ties GATA-3 to Estrogen Receptor α Expression in Breast Cancer. Cancer Res. 67, 6477–6483.

Eigeliene, N., Harkonen, P., and Erkkola, R. (2006). Effects of estradiol and medroxyprogesterone acetate on morphology, proliferation and apoptosis of human breast tissue in organ cultures. BMC Cancer 6, 246.

Eigeliene, N., Härkönen, P., and Erkkola, R. (2008). Effects of estradiol and medroxyprogesterone acetate on expression of the cell cycle proteins cyclin D1, p21 and p27 in cultured human breast tissues. Cell Cycle 7, 71–80.

Eigeliene, N., Elo, T., Linhala, M., Hurme, S., Erkkola, R., and Härkönen, P. (2012). Androgens Inhibit the Stimulatory Action of 17β-Estradiol on Normal Human Breast Tissue in Explant Cultures. J. Clin. Endocrinol. Metab. 97, E1116–E1127.

Eigeliene, N., Kangas, L., Hellmer, C., Kauko, T., Erkkola, R., and Harkonen, P. (2016). Effects of ospemifene, a novel selective estrogen-receptor modulator, on human breast tissue ex vivo. Menopause J. North Am. Menopause Soc. 23, 719–730.

Eliassen, A.H., Missmer, S.A., Tworoger, S.S., Spiegelman, D., Barbieri, R.L., Dowsett, M., and Hankinson, S.E. (2006). Endogenous Steroid Hormone Concentrations and Risk of Breast Cancer Among Premenopausal Women. JNCI J. Natl. Cancer Inst. 98, 1406– 1415.

Fazzari, A., Catalano, M.G., Comba, A., Becchis, M., Raineri, M., Frairia, R., and Fortunati, N. (2001). The control of progesterone receptor expression in MCF-7 breast cancer cells: effects of estradiol and sex hormone-binding globulin (SHBG). Mol. Cell. Endocrinol. 172, 31–36.

Fernandez, S. V, and Russo, J. (2010). Estrogen and Xenoestrogens in Breast Cancer. Toxicol. Pathol. 38, 110–122.

Fillmore, C.M., Gupta, P.B., Rudnick, J.A., Caballero, S., Keller, P.J., Lander, E.S., and Kuperwasser, C. (2010). Estrogen expands breast cancer stem-like cells through paracrine FGF/Tbx3 signaling. Proc. Natl. Acad. Sci. U. S. A. 107, 21737–21742.

Fischer, M., Uxa, S., Stanko, C., Magin, T.M., and Engeland, K. (2017). Human papilloma virus E7 oncoprotein abrogates the p53-p21-DREAM pathway. Sci. Rep. 7, 1– 11.

Florian, S., Iwamoto, Y., Coughlin, M., Weissleder, R., and Mitchison, T.J. (2019). A human organoid system that self-organizes to recapitulate growth and differentiation of a benign mammary tumor. Proc. Natl. Acad. Sci. 116, 11444–11453.

Fridriksdottir, A.J., Kim, J., Villadsen, R., Klitgaard, M.C., Hopkinson, B.M., Petersen, O.W., and Rønnov-Jessen, L. (2015). Propagation of oestrogen receptor-positive and oestrogen-responsive normal human breast cells in culture. Nat. Commun. 6.

182

Fullwood, M.J., Liu, M.H., Pan, Y.F., Liu, J., Han, X., Mohamed, Y.B., Orlov, Y.L., Velkov, S., Ho, A., Mei, P.H., et al. (2009). An Oestrogen Receptor α-bound Human Chromatin Interactome. Nature 462, 58–64.

Gabrielson, M., Chiesa, F., Paulsson, J., Strell, C., Behmer, C., Ronnow, K., Czene, K., Ostman, A., and Hall, P. (2016). Amount of stroma is associated with mammographic density and stromal expression of oestrogen receptor in normal breast tissues. Breast Cancer Res Treat 158, 253–261.

Garbe, J.C., Bhattacharya, S., Merchant, B., Bassett, E., Swisshelm, K., Feiler, H.S., Wyrobek, A.J., and Stampfer, M.R. (2009). Molecular distinctions between the stasis and telomere attrition senescence barriers demonstrated by long-term culture of normal human mammary epithelial cells. Cancer Res. 69, 7557–7568.

Garcia, M., Derocq, D., Freiss, G., and Rochefort, H. (1992). Activation of estrogen receptor transfected into a receptor-negative breast cancer cell line decreases the metastatic and invasive potential of the cells. Proc. Natl. Acad. Sci. U. S. A. 89, 11538– 11542.

Garcia-Closas, M., Malats, N., Silverman, D., Dosemeci, M., Kogevinas, M., Hein, D.W., Tardón, A., Serra, C., Carrato, A., García-Closas, R., et al. (2005). NAT2 slow acetylation, GSTM1 null genotype, and risk of bladder cancer: results from the Spanish Bladder Cancer Study and meta-analyses. The Lancet 366, 649–659.

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

Goldhammer, N., Kim, J., Timmermans-Wielenga, V., and Petersen, O.W. (2019). Characterization of organoid cultured human breast cancer. Breast Cancer Res. 21, 141.

Gonzales, M.C., Grayson, J., Lie, A., Yu, C.H., and Shiao, S.-Y.P.K. (2018). Gene- environment interactions and predictors of breast cancer in family-based multi-ethnic groups. Oncotarget 9, 29019–29035.

Gould, M.N. (1986). Inheritance and Site of Expression of Genes Controlling Susceptibility to Mammary Cancer in an Inbred Rat Model. 46, 5.

Gould, K.A., Tochacek, M., Schaffer, B.S., Reindl, T.M., Murrin, C.R., Lachel, C.M., VanderWoude, E.A., Pennington, K.L., Flood, L.A., Bynote, K.K., et al. (2004). Genetic Determination of Susceptibility to Estrogen-Induced Mammary Cancer in the ACI Rat. Genetics 168, 2113–2125.

Goyal, R., Gupta, T., Gupta, R., Aggarwal, A., Sahni, D., and Singh, G. (2016). Histological and immunohistochemical study of estrogen and progesterone receptors in normal human breast tissue in adult age groups vulnerable to malignancy. Clin Anat 29, 729–737.

183

Gruber, C.J., Tschugguel, W., Schneeberger, C., and Huber, J.C. (2002). Production and actions of estrogens. N Engl J Med 346, 340–352.

Gruber, C.J., Gruber, D.M., Gruber, I.M.L., Wieser, F., and Huber, J.C. (2004). Anatomy of the estrogen response element. Trends Endocrinol. Metab. 15, 73–78.

Gulbahce, H.E., Blair, C.K., Sweeney, C., and Salama, M.E. (2017). Quantification of Estrogen Receptor Expression in Normal Breast Tissue in Postmenopausal Women With Breast Cancer and Association With Tumor Subtypes. Appl Immunohistochem Mol Morphol 25, 548–552.

Guzman, R.C., Yang, J., Rajkumar, L., Thordarson, G., Chen, X., and Nandi, S. (1999). Hormonal prevention of breast cancer: Mimicking the protective effect of pregnancy. Proc. Natl. Acad. Sci. U. S. A. 96, 2520–2525.

Haag, J.D., Shepel, L.A., Kolman, B.D., Monson, D.M., Benton, M.E., Watts, K.T., Waller, J.L., Lopez-Guajardo, C.C., Samuelson, D.J., and Gould, M.N. (2003). Congenic Rats Reveal Three Independent Copenhagen Alleles within the Mcs1 Quantitative Trait Locus That Confer Resistance to Mammary Cancer. 6.

Hall, J.M., and McDonnell, D.P. (1999). The Estrogen Receptor β-Isoform (ERβ) of the Human Estrogen Receptor Modulates ERα Transcriptional Activity and Is a Key Regulator of the Cellular Response to Estrogens and Antiestrogens. Endocrinology 140, 5566–5578.

Harris, H.A., Albert, L.M., Leathurby, Y., Malamas, M.S., Mewshaw, R.E., Miller, C.P., Kharode, Y.P., Marzolf, J., Komm, B.S., Winneker, R.C., et al. (2003). Evaluation of an Estrogen Receptor-β Agonist in Animal Models of Human Disease. Endocrinology 144, 4241–4249.

Harrison, H., Simões, B.M., Rogerson, L., Howell, S.J., Landberg, G., and Clarke, R.B. (2013). Oestrogen increases the activity of oestrogen receptor negative breast cancer stem cells through paracrine EGFR and Notch signalling. Breast Cancer Res. 15, R21.

Heery, D.M., Kalkhoven, E., Hoare, S., and Parker, M.G. (1997). A signature motif in transcriptional co-activators mediates binding to nuclear receptors. Nature 387, 733–736.

Heldring, N., Pike, A., Andersson, S., Matthews, J., Cheng, G., Hartman, J., Tujague, M., Strom, A., Treuter, E., Warner, M., et al. (2007). Estrogen receptors: how do they signal and what are their targets. Physiol Rev 87, 905–931.

Herynk, M.H., Parra, I., Cui, Y., Beyer, A., Wu, M.-F., Hilsenbeck, S.G., and Fuqua, S.A.W. (2007). Association Between the Estrogen Receptor α A908G Mutation and Outcomes in Invasive Breast Cancer. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 13, 3235–3243.

184

Homer-Bouthiette, C., Zhao, Y., Shunkwiler, L.B., Van Peel, B., Garrett-Mayer, E., Baird, R.C., Rissman, A.I., Guest, S.T., Ethier, S.P., John, M.C., et al. (2018). Deletion of the murine ortholog of the 8q24 gene desert has anti-cancer effects in transgenic mammary cancer models. BMC Cancer 18.

Hopp, T.A. (2004). Low Levels of Estrogen Receptor Protein Predict Resistance to Tamoxifen Therapy in Breast Cancer. Clin. Cancer Res. 10, 7490–7499.

Huschtscha, L.I., Neumann, A.A., Noble, J.R., and Reddel, R.R. (2001). Effects of Simian Virus 40 T-Antigens on Normal Human Mammary Epithelial Cells Reveal Evidence for Spontaneous Alterations in Addition to Loss of p16INK4a Expression. Exp. Cell Res. 265, 125–134.

Iavarone, A., and Massague, J. (1997). Repression of the CDK activator Cdc25A and cell-cycle arrest by cytokine TGF-(3 in cells lacking the CDK inhibitor p15. Nature 387, 6.

Jeng, M.-H., Jiang, S.-Y., and Jordan, V.C. (1994). Paradoxical regulation of estrogen- dependent growth factor gene expression in estrogen receptor (ER)-negative human breast cancer cells stably expressing ER. Cancer Lett. 82, 123–128.

Jerry, D.J., Shull, J.D., Hadsell, D.L., Rijnkels, M., Dunphy, K.A., Schneider, S.S., Vandenberg, L.N., Majhi, P.D., Byrne, C., and Trentham-Dietz, A. (2018). Genetic variation in sensitivity to estrogens and breast cancer risk. Mamm. Genome Off. J. Int. Mamm. Genome Soc. 29, 24–37.

Jeselsohn, R., Yelensky, R., Buchwalter, G., Frampton, G., Meric-Bernstam, F., Gonzalez-Angulo, A.M., Ferrer-Lozano, J., Perez-Fidalgo, J.A., Cristofanilli, M., Gómez, H., et al. (2014). Emergence of constitutively active estrogen receptor-α mutations in pretreated advanced estrogen receptor positive breast cancer. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 20, 1757–1767.

Jiang, S.Y., and Jordan, V.C. (1992). Growth regulation of estrogen receptor-negative breast cancer cells transfected with complementary DNAs for estrogen receptor. J. Natl. Cancer Inst. 84, 580–591.

Jiang, Y., Gong, P., Madak-Erdogan, Z., Martin, T., Jeyakumar, M., Carlson, K., Khan, I., Smillie, T.J., Chittiboyina, A.G., Rotte, S.C.K., et al. (2013). Mechanisms enforcing the estrogen receptor β selectivity of botanical estrogens. FASEB J. 27, 4406–4418.

Jin, L., Qu, Y., Gomez, L.J., Chung, S., Han, B., Gao, B., Yue, Y., Gong, Y., Liu, X., Amersi, F., et al. (2017). Characterization of primary human mammary epithelial cells isolated and propagated by conditional reprogrammed cell culture. Oncotarget 9, 11503– 11514.

Juríková, M., Danihel, Ľ., Polák, Š., and Varga, I. (2016). Ki67, PCNA, and MCM proteins: Markers of proliferation in the diagnosis of breast cancer. Acta Histochem. 118, 544–552.

185

Kao, C.-Y., Nomata, K., Oakley, C.S., Welsch, C.W., and Chang, C.-C. (1995). Two types of normal human breast epithelial cells derived from reduction mammoplasty: phenotypic characterization and response to SV40 transfection. Carcinogenesis 16, 531– 538.

Karmakar, S., Foster, E.A., and Smith, C.L. (2009). Unique Roles of p160 Coactivators for Regulation of Breast Cancer Cell Proliferation and Estrogen Receptor-α Transcriptional Activity. Endocrinology 150, 1588–1596.

Kato, S., Endoh, H., Masuhiro, Y., Kitamoto, T., Uchiyama, S., Sasaki, H., Masushige, S., Gotoh, Y., Nishida, E., Kawashima, H., et al. (1995). Activation of the Estrogen Receptor Through Phosphorylation by Mitogen-Activated Protein Kinase. Science 270, 1491–1494.

Keller, P.J., Lin, A.F., Arendt, L.M., Klebba, I., Jones, A.D., Rudnick, J.A., DiMeo, T.A., Gilmore, H., Jefferson, D.M., Graham, R.A., et al. (2010). Mapping the cellular and molecular heterogeneity of normal and malignant breast tissues and cultured cell lines. Breast Cancer Res. BCR 12, R87.

Kerdivel, G., Le Guevel, R., Habauzit, D., Brion, F., Ait-Aissa, S., and Pakdel, F. (2013). Estrogenic Potency of Benzophenone UV Filters in Breast Cancer Cells: Proliferative and Transcriptional Activity Substantiated by Docking Analysis. PLoS ONE 8, e60567.

Khanna, S., Dash, P.R., and Darbre, P.D. (2014). Exposure to parabens at the concentration of maximal proliferative response increases migratory and invasive activity of human breast cancer cells in vitro. J. Appl. Toxicol. 34, 1051–1059.

Kong, S.L., Li, G., Loh, S.L., Sung, W.-K., and Liu, E.T. (2011). Cellular reprogramming by the conjoint action of ERα, FOXA1, and GATA3 to a ligand-inducible growth state. Mol. Syst. Biol. 7, 526.

Kouros-Mehr, H., Slorach, E.M., Sternlicht, M.D., and Werb, Z. (2006). GATA-3 maintains the differentiation of the luminal cell fate in the mammary gland. Cell 127, 1041–1055.

Krege, J.H., Hodgin, J.B., Couse, J.F., Enmark, E., Warner, M., Mahler, J.F., Sar, M., Korach, K.S., Gustafsson, J.-A., and Smithies, O. (1998). Generation and reproductive phenotypes of mice lacking estrogen receptor. Proc. Natl. Acad. Sci. 95, 15677–15682.

Kuchenbaecker, K.B., Neuhausen, S.L., Robson, M., Barrowdale, D., McGuffog, L., Mulligan, A.M., Andrulis, I.L., Spurdle, A.B., Schmidt, M.K., Schmutzler, R.K., et al. (2014). Associations of common breast cancer susceptibility alleles with risk of breast cancer subtypes in BRCA1 and BRCA2 mutation carriers. Breast Cancer Res. BCR 16.

Kumar, R., and Thompson, E.B. (1999). The structure of the nuclear hormone receptors. Steroids 64, 310–319.

186

Kumar, B., Prasad, M., Bhat-Nakshatri, P., Anjanappa, M., Kalra, M., Marino, N., Storniolo, A.M., Rao, X., Liu, S., Wan, J., et al. (2018). Normal Breast-Derived Epithelial Cells with Luminal and Intrinsic Subtype-Enriched Gene Expression Document Interindividual Differences in Their Differentiation Cascade. Cancer Res. 78, 5107–5123.

Kumar, R., Zakharov, M.N., Khan, S.H., Miki, R., Jang, H., Toraldo, G., Singh, R., Bhasin, S., and Jasuja, R. (2011). The Dynamic Structure of the Estrogen Receptor. J. Amino Acids 2011.

Kurebayashi, J., Otsuki, T., Kunisue, H., Tanaka, K., Yamamoto, S., and Sonoo, H. (2000). Expression Levels of Estrogen Receptor-α, Estrogen Receptor-β, Coactivators, and Corepressors in Breast Cancer. Clin. Cancer Res. 6, 512 LP – 518.

Laidlaw, I.J., Clarke, R.B., Howell, A., Owen, A.W., Potten, C.S., and Anderson, E. (1995). The proliferation of normal human breast tissue implanted into athymic nude mice is stimulated by estrogen but not progesterone. Endocrinology 136, 164–171.

Lalloo, F., and Evans, D.G. (2012). Familial Breast Cancer. Clin. Genet. 82, 105–114.

Lambe, M., Hsieh, C.C., Chan, H.W., Ekbom, A., Trichopoulos, D., and Adami, H.O. (1996). Parity, age at first and last birth, and risk of breast cancer: a population-based study in Sweden. Breast Cancer Res. Treat. 38, 305–311.

Lan, H., Kendziorski, C.M., Haag, J.D., Shepel, L.A., Newton, M.A., and Gould, M.N. (2001). Genetic loci controlling breast cancer susceptibility in the Wistar-Kyoto rat. Genetics 157, 331–339.

LaPlante, C.D., Bansal, R., Dunphy, K.A., Jerry, D.J., and Vandenberg, L.N. (2018). Oxybenzone Alters Mammary Gland Morphology in Mice Exposed During Pregnancy and Lactation. J. Endocr. Soc. 2, 903–921.

Lazennec, G., and Katzenellenbogen, B.S. (1999). Expression of human estrogen receptor using an efficient adenoviral gene delivery system is able to restore hormone- dependent features to estrogen receptor-negative breast carcinoma cells. Mol. Cell. Endocrinol. 149, 93–105.

Lee, J.K., Garbe, J.C., Vrba, L., Miyano, M., Futscher, B.W., Stampfer, M.R., and LaBarge, M.A. (2015). Age and the means of bypassing stasis influence the intrinsic subtype of immortalized human mammary epithelial cells. Front. Cell Dev. Biol. 3.

Lee, J.K., Bloom, J., Zubeldia-Plazaola, A., Garbe, J.C., Stampfer, M.R., and LaBarge, M.A. (2018). Different culture media modulate growth, heterogeneity, and senescence in human mammary epithelial cell cultures. PLoS ONE 13.

Leitman, D.C., Paruthiyil, S., Vivar, O.I., Saunier, E.F., Herber, C.B., Cohen, I., Tagliaferri, M., and Speed, T.P. (2010). Regulation of specific target genes and biological responses by estrogen receptor subtype agonists. Curr. Opin. Pharmacol. 10, 629–636.

187

Leu, Y.-W., Yan, P.S., Fan, M., Jin, V.X., Liu, J.C., Curran, E.M., Welshons, W.V., Wei, S.H., Davuluri, R.V., Plass, C., et al. (2004). Loss of Estrogen Receptor Signaling Triggers Epigenetic Silencing of Downstream Targets in Breast Cancer. Cancer Res. 64, 8184–8192.

Levin, E.R. (2009). Plasma Membrane Estrogen Receptors. Trends Endocrinol. Metab. TEM 20, 477–482.

Li, S.A., Weroha, S.J., Tawfik, O., and Li, J.J. (2002). Prevention of solely estrogen- induced mammary tumors in female aci rats by tamoxifen: evidence for estrogen receptor mediation. J. Endocrinol. 175, 297–305.

Liu, X., Ory, V., Chapman, S., Yuan, H., Albanese, C., Kallakury, B., Timofeeva, O.A., Nealon, C., Dakic, A., Simic, V., et al. (2012). ROCK Inhibitor and Feeder Cells Induce the Conditional Reprogramming of Epithelial Cells. Am. J. Pathol. 180, 599–607.

Lonning, P.E., Taylor, P.D., Anker, G., Iddon, J., Wie, L., Jorgensen, L.M., Mella, O., and Howell, A. (2001). High-dose estrogen treatment in postmenopausal breast cancer patients heavily exposed to endocrine therapy. Breast Cancer Res Treat 67, 111–116.

Lubahn, D.B., Moyer, J.S., Golding, T.S., Couse, J.F., Korach, K.S., and Smithies, O. (1993). Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc. Natl. Acad. Sci. U. S. A. 90, 11162–11166.

Maas, P., Barrdahl, M., Joshi, A.D., Auer, P.L., Gaudet, M.M., Milne, R.L., Schumacher, F.R., Anderson, W.F., Check, D., Chattopadhyay, S., et al. (2016). Breast Cancer Risk From Modifiable and Nonmodifiable Risk Factors Among White Women in the United States. JAMA Oncol. 2, 1295–1302.

MacMahon, B., Cole, P., Lin, T.M., Lowe, C.R., Mirra, A.P., Ravnihar, B., Salber, E.J., Valaoras, V.G., and Yuasa, S. (1970). Age at first birth and breast cancer risk. Bull. World Health Organ. 43, 209–221.

Maggi, A. (2011). Liganded and unliganded activation of estrogen receptor and hormone replacement therapies. Biochim. Biophys. Acta BBA - Mol. Basis Dis. 1812, 1054–1060.

Mahajan, A.S., Sugita, B.M., Duttargi, A.N., Saenz, F., Krawczyk, E., McCutcheon, J.N., Fonseca, A.S., Kallakury, B., Pohlmann, P., Gusev, Y., et al. (2017). Genomic comparison of early-passage conditionally reprogrammed breast cancer cells to their corresponding primary tumors. PLoS ONE 12.

Majhi, P.D., Sharma Aman, Roberts Amy L., Daniele Elizabeth, Majewski Aliza R., Chuong Lynn M., Black Amye L., Vandenberg Laura N., Schneider Sallie S., Dunphy Karen A., et al. (2020). Effects of Benzophenone-3 and Propylparaben on Estrogen Receptor–Dependent R-Loops and DNA Damage in Breast Epithelial Cells and Mice. Environ. Health Perspect. 128, 017002.

188

Mallepell, S., Krust, A., Chambon, P., and Brisken, C. (2006). Paracrine signaling through the epithelial estrogen receptor α is required for proliferation and morphogenesis in the mammary gland. Proc. Natl. Acad. Sci. U. S. A. 103, 2196–2201.

Martinez-Zapien, D., Ruiz, F.X., Poirson, J., Mitschler, A., Ramirez, J., Forster, A., Cousido-Siah, A., Masson, M., Pol, S.V., Podjarny, A., et al. (2016). Structure of the E6/E6AP/p53 complex required for HPV-mediated degradation of p53. Nature 529, 541– 545.

Matthews, J., Wihlén, B., Tujague, M., Wan, J., Ström, A., and Gustafsson, J.-Å. (2006). Estrogen Receptor (ER) β Modulates ERα-Mediated Transcriptional Activation by Altering the Recruitment of c-Fos and c-Jun to Estrogen-Responsive Promoters. Mol. Endocrinol. 20, 534–543.

Meerbrey, K.L., Hu, G., Kessler, J.D., Roarty, K., Li, M.Z., Fang, J.E., Herschkowitz, J.I., Burrows, A.E., Ciccia, A., Sun, T., et al. (2011). The pINDUCER lentiviral toolkit for inducible RNA interference in vitro and in vivo. Proc. Natl. Acad. Sci. U. S. A. 108, 3665–3670.

Meng, P., Vaapil, M., Tagmount, A., Loguinov, A., Vulpe, C., and Yaswen, P. (2019). Propagation of functional estrogen receptor positive normal human breast cells in 3D cultures. Breast Cancer Res. Treat. 176, 131–140.

Merrell, K.W., Crofts, J.D., Smith, R.L., Sin, J.H., Kmetzsch, K.E., Merrell, A., Miguel, R.O., Candelaria, N.R., and Lin, C.-Y. (2011). Differential recruitment of nuclear receptor coregulators in ligand-dependent transcriptional repression by estrogen receptor- α. Oncogene 30, 1608–1614.

Meyers, M.J., Sun, J., Carlson, K.E., Marriner, G.A., Katzenellenbogen, B.S., and Katzenellenbogen, J.A. (2001). Estrogen Receptor-β Potency-Selective Ligands: Structure−Activity Relationship Studies of Diarylpropionitriles and Their Acetylene and Polar Analogues. J. Med. Chem. 44, 4230–4251.

Michailidou, K., Lindström, S., Dennis, J., Beesley, J., Hui, S., Kar, S., Lemaçon, A., Soucy, P., Glubb, D., Rostamianfar, A., et al. (2017). Association analysis identifies 65 new breast cancer risk loci. Nature 551, 92–94.

Misiti, S., Schomburg, L., Yen, P.M., and Chin, W.W. (1998). Expression and Hormonal Regulation of Coactivator and Corepressor Genes. 139, 8.

Miyano, M., Sayaman, R.W., Stoiber, M.H., Lin, C.-H., Stampfer, M.R., Brown, J.B., and LaBarge, M.A. (2017). Age-related gene expression in luminal epithelial cells is driven by a microenvironment made from myoepithelial cells. Aging 9, 2026–2051.

Moggs, J.G., Murphy, T.C., Lim, F.L., Moore, D.J., Stuckey, R., Antrobus, K., Kimber, I., and Orphanides, G. (2005). Anti-proliferative effect of estrogen in breast cancer cells that re-express ERα is mediated by aberrant regulation of cell cycle genes. J. Mol. Endocrinol. 34, 535–551.

189

Molenda, H.A., Kilts, C.P., Allen, R.L., and Tetel, M.J. (2003). Nuclear Receptor Coactivator Function in Reproductive Physiology and Behavior. Biol. Reprod. 69, 1449– 1457.

Molina-Molina, J.-M., Escande, A., Pillon, A., Gomez, E., Pakdel, F., Cavaillès, V., Olea, N., Aït-Aïssa, S., and Balaguer, P. (2008). Profiling of benzophenone derivatives using fish and human estrogen receptor-specific in vitro bioassays. Toxicol. Appl. Pharmacol. 232, 384–395.

Montero Girard, G., Vanzulli, S.I., Cerliani, J.P., Bottino, M.C., Bolado, J., Vela, J., Becu-Villalobos, D., Benavides, F., Gutkind, S., Patel, V., et al. (2007). Association of estrogen receptor-α and progesterone receptor A expression with hormonal mammary carcinogenesis: role of the host microenvironment. Breast Cancer Res. 9, R22.

Moon, R.C. (1969). Relationship between previous reproductive history and chemically induced mammary cancer in rats. Int. J. Cancer 4, 312–317.

Moore, J.T., McKee, D.D., Slentz-Kesler, K., Moore, L.B., Jones, S.A., Horne, E.L., Su, J.-L., Kliewer, S.A., Lehmann, J.M., and Willson, T.M. (1998). Cloning and Characterization of Human Estrogen Receptor β Isoforms. Biochem. Biophys. Res. Commun. 247, 75–78.

Mortensen, M.E., Calafat, A.M., Ye, X., Wong, L.-Y., Wright, D.J., Pirkle, J.L., Merrill, L.S., and Moye, J. (2014). Urinary Concentrations of Environmental Phenols in Pregnant Women in a Pilot Study of the National Children’s Study. Environ. Res. 129, 32–38.

Mucci, L., Hjelmborg, J., and Harris, J. (2016). Familial risk and heritability of cancer among twins in nordic countries. JAMA 315, 68–76.

Mukherjee, T.K., Reynolds, P.R., and Hoidal, J.R. (2005). Differential effect of estrogen receptor alpha and beta agonists on the receptor for advanced glycation end product expression in human microvascular endothelial cells. Biochim. Biophys. Acta BBA - Mol. Cell Res. 1745, 300–309.

Münger, K., Werness, B.A., Dyson, N., Phelps, W.C., Harlow, E., and Howley, P.M. (1989). Complex formation of human papillomavirus E7 proteins with the retinoblastoma tumor suppressor gene product. EMBO J. 8, 4099–4105.

Navarrete, M.A.H., Maier, C.M., Falzoni, R., Quadros, L.G. de A., Lima, G.R., Baracat, E.C., and Nazário, A.C. (2005). Assessment of the proliferative, apoptotic and cellular renovation indices of the human mammary epithelium during the follicular and luteal phases of the menstrual cycle. Breast Cancer Res. 7, R306–R313.

Nelson, A.W., Groen, A.J., Miller, J.L., Warren, A.Y., Holmes, K.A., Tarulli, G.A., Tilley, W.D., Katzenellenbogen, B.S., Hawse, J.R., Gnanapragasam, V.J., et al. (2017). Comprehensive assessment of estrogen receptor beta antibodies in cancer cell line models and tissue reveals critical limitations in reagent specificity. Mol. Cell. Endocrinol. 440, 138–150.

190

Nguyen, H.D., Phan, T.T.P., Carraz, M., and Brunsveld, L. (2012). Estrogen Receptor α/β–cofactor motif interactions; interplay of tyrosine 537/488 phosphorylation and LXXLL motifs. Mol. Biosyst. 8, 3134–3141.

Nguyen, Q.H., Pervolarakis, N., Blake, K., Ma, D., Davis, R.T., James, N., Phung, A.T., Willey, E., Kumar, R., Jabart, E., et al. (2018). Profiling human breast epithelial cells using single cell RNA sequencing identifies cell diversity. Nat. Commun. 9.

Nickels, S., Truong, T., Hein, R., Stevens, K., Buck, K., Behrens, S., Eilber, U., Schmidt, M., Häberle, L., Vrieling, A., et al. (2013). Evidence of Gene–Environment Interactions between Common Breast Cancer Susceptibility Loci and Established Environmental Risk Factors. PLoS Genet. 9.

Nilsson, S., Mäkelä, S., Treuter, E., Tujague, M., Thomsen, J., Andersson, G., Enmark, E., Pettersson, K., Warner, M., and Gustafsson, J.-Å. (2001). Mechanisms of Estrogen Action. Physiol. Rev. 81, 1535 LP – 1565.

Nonet, G.H., Stampfer, M.R., Chin, K., Gray, J.W., Collins, C.C., and Yaswen, P. (2001). The ZNF217 Gene Amplified in Breast Cancers Promotes Immortalization of Human Mammary Epithelial Cells. Cancer Res. 61, 1250–1254.

Novelli, F., Milella, M., Melucci, E., Di Benedetto, A., Sperduti, I., Perrone-Donnorso, R., Perracchio, L., Venturo, I., Nisticò, C., Fabi, A., et al. (2008). A divergent role for estrogen receptor-beta in node-positive and node-negative breast cancer classified according to molecular subtypes: an observational prospective study. Breast Cancer Res. 10, R74.

Ogawa, S., Inoue, S., Watanabe, T., Orimo, A., Hosoi, T., Ouchi, Y., and Muramatsu, M. (1998). Molecular cloning and characterization of human estrogen receptor betacx: a potential inhibitor ofestrogen action in human. Nucleic Acids Res. 26, 3505–3512.

Omoto, Y., Inoue, S., Ogawa, S., Toyama, T., Yamashita, H., Muramatsu, M., Kobayashi, S., and Iwase, H. (2001). Clinical value of the wild-type estrogen receptor β expression in breast cancer. Cancer Lett. 163, 207–212.

Pal, B., Chen, Y., Vaillant, F., Jamieson, P., Gordon, L., Rios, A.C., Wilcox, S., Fu, N., Liu, K.H., Jackling, F.C., et al. (2017). Construction of developmental lineage relationships in the mouse mammary gland by single-cell RNA profiling. Nat. Commun. 8.

Palechor-Ceron, N., Suprynowicz, F.A., Upadhyay, G., Dakic, A., Minas, T., Simic, V., Johnson, M., Albanese, C., Schlegel, R., and Liu, X. (2013). Radiation Induces Diffusible Feeder Cell Factor(s) That Cooperate with ROCK Inhibitor to Conditionally Reprogram and Immortalize Epithelial Cells. Am. J. Pathol. 183, 1862–1870.

Pedram, A., Razandi, M., Kim, J.K., O’Mahony, F., Lee, E.Y., Luderer, U., and Levin, E.R. (2009). Developmental Phenotype of a Membrane Only Estrogen Receptor α (MOER) Mouse. J. Biol. Chem. 284, 3488–3495.

191

Peterson, E.A., Jenkins, E.C., Lofgren, K.A., Chandiramani, N., Liu, H., Aranda, E., Barnett, M., and Kenny, P.A. (2015). Amphiregulin is a critical downstream effector of estrogen signaling in ERα-positive breast cancer. Cancer Res. 75, 4830–4838.

Peto, J., and Mack, T.M. (2000). High constant incidence in twins and other relatives of women with breast cancer. Nat. Genet. 26, 411–414.

Peto, J., Collins, N., Barfoot, R., Seal, S., Warren, W., Rahman, N., Easton, D.F., Evans, C., Deacon, J., and Stratton, M.R. (1999). Prevalence of BRCA1 and BRCA2 Gene Mutations in Patients With Early-Onset Breast Cancer. JNCI J. Natl. Cancer Inst. 91, 943–949.

Philippat, C., Wolff, M.S., Calafat, A.M., Ye, X., Bausell, R., Meadows, M., Stone, J., Slama, R., and Engel, S.M. (2013). Prenatal Exposure to Environmental Phenols: Concentrations in Amniotic Fluid and Variability in Urinary Concentrations during Pregnancy. Environ. Health Perspect. 121, 1225–1231.

Pilat, M.J., Christman, J.K., and Brooks, S.C. (1996). Characterization of the estrogen receptor transfected MCF10A breast cell line 139B6. Breast Cancer Res Treat 37, 253– 266.

Pink, J.J., and Jordan, V.C. (1996). Models of Estrogen Receptor Regulation by Estrogens and Antiestrogens in Breast Cancer Cell Lines. Cancer Res. 56, 2321–2330.

Potten, C.S., Watson, R.J., Williams, G.T., Tickle, S., Roberts, S.A., Harris, M., and Howell, A. (1988). The effect of age and menstrual cycle upon proliferative activity of the normal human breast. Br. J. Cancer 58, 163–170.

Prat, A., Karginova, O., Parker, J.S., Fan, C., He, X., Bixby, L., Harrell, J.C., Roman, E., Adamo, B., Troester, M., et al. (2013). Characterization of cell lines derived from breast cancers and normal mammary tissues for the study of the intrinsic molecular subtypes. Breast Cancer Res. Treat. 142, 237–255.

Pugazhendhi, D., and Darbre, P.D. (2010). Differential effects of overexpression of ERα and ERβ in MCF10A immortalised, non-transformed human breast epithelial cells. Horm. Mol. Biol. Clin. Investig. 1, 117–126.

Putnik, M., Zhao, C., Gustafsson, J.-Å., and Dahlman-Wright, K. (2012). Global identification of genes regulated by estrogen signaling and demethylation in MCF-7 breast cancer cells. Biochem. Biophys. Res. Commun. 426, 26–32.

Qiu, X., Wang, H., Wang, Z., Fu, Y., and Yin, J. (2019). Expression of PCNA, Ki-67 and COX-2 in breast cancer based on DCE-MRI image information. J. Infect. Public Health.

Qu, Y., Han, B., Yu, Y., Yao, W., Bose, S., Karlan, B.Y., Giuliano, A.E., and Cui, X. (2015). Evaluation of MCF10A as a Reliable Model for Normal Human Mammary Epithelial Cells. PLOS ONE 10, e0131285.

192

Qu, Y., Han, B., Gao, B., Bose, S., Gong, Y., Wawrowsky, K., Giuliano, A.E., Sareen, D., and Cui, X. (2017). Differentiation of Human Induced Pluripotent Stem Cells to Mammary-like Organoids. Stem Cell Rep. 8, 205–215.

Quong, J., Eppenberger-Castori, S., Moore, D., Scott, G.K., Birrer, M.J., Kueng, W., Eppenberger, U., and Benz, C.C. (2002). Age-Dependent Changes in Breast Cancer Hormone Receptors and Oxidant Stress Markers. Breast Cancer Res. Treat. 76, 221–236.

Raafat, A., Strizzi, L., Lashin, K., Ginsburg, E., McCurdy, D., Salomon, D., Smith, G.H., Medina, D., and Callahan, R. (2012). Effects of Age and Parity on Mammary Gland Lesions and Progenitor Cells in the FVB/N-RC Mice. PLOS ONE 7, e43624.

Rahman, N., Seal, S., Thompson, D., Kelly, P., Renwick, A., Elliott, A., Reid, S., Spanova, K., Barfoot, R., Chagtai, T., et al. (2007). PALB2, which encodes a BRCA2- interacting protein, is a breast cancer susceptibility gene. Nat. Genet. 39, 165–167.

Rajkumar, L., Kittrell, F.S., Guzman, R.C., Brown, P.H., Nandi, S., and Medina, D. (2007). Hormone-induced protection of mammary tumorigenesis in genetically engineered mouse models. Breast Cancer Res. 9, R12.

Read, L.D., Greene, G.L., and Katzenellenbogen, B.S. (1989). Regulation of Estrogen Receptor Messenger Ribonucleic Acid and Protein Levels in Human Breast Cancer Cell Lines by Sex Steroid Hormones, Their Antagonists, and Growth Factors. Mol. Endocrinol. 3, 295–304.

Renwick, A., Thompson, D., Seal, S., Kelly, P., Chagtai, T., Ahmed, M., North, B., Jayatilake, H., Barfoot, R., Spanova, K., et al. (2006). ATM mutations that cause ataxia- telangiectasia are breast cancer susceptibility alleles. Nat. Genet. 38, 873–875.

Roden, D.M., Wilke, R.A., Kroemer, H.K., and Stein, C.M. (2011). Pharmacogenomics: The genetics of variable drug responses. Circulation 123, 1661–1670.

Roger, P., Sahla, M.E., Makela, S., Gustafsson, J.Å., Baldet, P., and Rochefort, H. (2001). Decreased Expression of Estrogen Receptor ␤ Protein in Proliferative Preinvasive Mammary Tumors. 6.

Rosner, B., Colditz, G.A., and Willett, W.C. (1994). Reproductive Risk Factors in a Prospective Study of Breast Cancer: The Nurses’ Health Study. Am. J. Epidemiol. 139, 819–835.

Rossouw, J.E., Anderson, G.L., Prentice, R.L., LaCroix, A.Z., Kooperberg, C., Stefanick, M.L., Jackson, R.D., Beresford, S.A., Howard, B.V., Johnson, K.C., et al. (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. Jama 288, 321– 333.

193

Rotunno, M., Sun, X., Figueroa, J., Sherman, M.E., Garcia-Closas, M., Meltzer, P., Williams, T., Schneider, S.S., Jerry, D.J., Yang, X.R., et al. (2014). Parity-related molecular signatures and breast cancer subtypes by estrogen receptor status. Breast Cancer Res. BCR 16, R74.

Roy, R., Chun, J., and Powell, S.N. (2011). BRCA1 and BRCA2: different roles in a common pathway of genome protection. Nat. Rev. Cancer 12, 68–78.

Rudolph, A., Milne, R.L., Truong, T., Knight, J.A., Seibold, P., Flesch-Janys, D., Behrens, S., Eilber, U., Bolla, M.K., Wang, Q., et al. (2015). An investigation of gene- environment interactions between 47 newly identified breast cancer susceptibility loci and environmental risk factors. Int. J. Cancer J. Int. Cancer 136, E685–E696.

Rudolph, A., Song, M., Brook, M.N., Milne, R.L., Mavaddat, N., Michailidou, K., Bolla, M.K., Wang, Q., Dennis, J., Wilcox, A.N., et al. (2018). Joint associations of a polygenic risk score and environmental risk factors for breast cancer in the Breast Cancer Association Consortium. Int. J. Epidemiol. 47, 526–536.

Russo, J., and Russo, I.H. (1980). Susceptibility of the mammary gland to carcinogenesis. II. Pregnancy interruption as a risk factor in tumor incidence. Am. J. Pathol. 100, 497– 512.

Russo, J., and Russo, I.H. (2006). THE ROLE OF ESTROGEN IN THE INITATION OF BREST CANCER. J. Steroid Biochem. Mol. Biol. 102, 89–96.

Russo, J., Ao, X., Grill, C., and Russo, I.H. Pattern of distribution of cells positive for estrogen receptor and progesterone receptor in relation to proliferating cells in the mammary gland. Breast Cancer Res. Treatment 53, 217-227.

Saceda, M., Lippman, M.E., Chambon, P., Lindsey, R.L., Ponglikitmongkol, M., Puente, M., and Martin, M.B. (1988). Regulation of the Estrogen Receptor in MCF-7 Cells by Estradiol. Mol. Endocrinol. 2, 1157–1162.

Sachs, N., de Ligt, J., Kopper, O., Gogola, E., Bounova, G., Weeber, F., Balgobind, A.V., Wind, K., Gracanin, A., Begthel, H., et al. (2018). A Living Biobank of Breast Cancer Organoids Captures Disease Heterogeneity. Cell 172, 373-386.e10.

Saji, S., Omoto, Y., Shimizu, C., Warner, M., Hayashi, Y., Horiguchi, S., Watanabe, T., Hayashi, S., Gustafsson, J.-Å., and Toi, M. (2002). Expression of Estrogen Receptor (ER) βcx Protein in ERα-positive Breast Cancer: Specific Correlation with Progesterone Receptor. 6.

Samavat, H., and Kurzer, M.S. (2015). Estrogen Metabolism and Breast Cancer. Cancer Lett. 356, 231–243.

Saville, B., Wormke, M., Wang, F., Nguyen, T., Enmark, E., Kuiper, G., Gustafsson, J.- Å., and Safe, S. (2000). Ligand-, Cell-, and Estrogen Receptor Subtype (α/β)-dependent Activation at GC-rich (Sp1) Promoter Elements. J. Biol. Chem. 275, 5379–5387.

194

Schaffer, B.S., Lachel, C.M., Pennington, K.L., Murrin, C.R., Strecker, T.E., Tochacek, M., Gould, K.A., Meza, J.L., McComb, R.D., and Shull, J.D. (2006). Genetic Bases of Estrogen-Induced Tumorigenesis in the Rat: Mapping of Loci Controlling Susceptibility to Mammary Cancer in a Brown Norway × ACI Intercross. Cancer Res. 66, 7793–7800.

Schaffer, B.S., Leland-Wavrin, K.M., Kurz, S.G., Colletti, J.A., Seiler, N.L., Warren, C.L., and Shull, J.D. (2013). Mapping of Three Genetic Determinants of Susceptibility to Estrogen-Induced Mammary Cancer within the Emca8 Locus on Rat Chromosome 5. Cancer Prev. Res. Phila. Pa 6, 59–69.

Scheffner, M., Werness, B.A., Huibregtse, J.M., Levine, A.J., and Howley, P.M. (1990). The E6 oncoprotein encoded by human papillomavirus types 16 and 18 promotes the degradation of p53. Cell 63, 1129–1136.

Schlotz, N., Kim, G.-J., Jäger, S., Günther, S., and Lamy, E. (2017). In vitro observations and in silico predictions of xenoestrogen mixture effects in T47D-based receptor transactivation and proliferation assays. Toxicol. In Vitro 45, 146–157.

Schlumpf, M., Cotton, B., Conscience, M., Haller, V., Steinmann, B., and Lichtensteiger, W. (2001). In vitro and in vivo estrogenicity of UV screens. Environ. Health Perspect. 109, 239–244.

Schneider, S.S., Henchey, E.M., Sultana, N., Morin, S.M., Jerry, D.J., Makari‐Judson, G., Crisi, G.M., Arenas, R.B., Johnson, M., Mason, H.S., et al. (2017). Individual‐specific variation in the respiratory activities of HMECs and their bioenergetic response to IGF1 and TNFα. J. Cell. Physiol. 232, 2750–2765.

Schubert, E.L., Lee, M.K., Newman, B., and King, M.-C. (1999). Single nucleotide polymorphisms (SNPs) in the estrogen receptor gene and breast cancer susceptibility. J. Steroid Biochem. Mol. Biol. 71, 21–27.

Seal, S., Thompson, D., Renwick, A., Elliott, A., Kelly, P., Barfoot, R., Chagtai, T., Jayatilake, H., Ahmed, M., Spanova, K., et al. (2006). Truncating mutations in the Fanconi anemia J gene BRIP1 are low-penetrance breast cancer susceptibility alleles. Nat. Genet. 38, 1239–1241.

Shaaban, A., O’Neill, P., Davies, M., Sibson, R., West, C., Smith, P., and Foster, C. (2003). Declining Estrogen Receptor-β Expression Defines Malignant Progression of Human Breast Neoplasia. Am. J. Surg. Pathol. 27, 1502–1512.

Shaaban, A.M., Green, A.R., Karthik, S., Alizadeh, Y., Hughes, T.A., Harkins, L., Ellis, I.O., Robertson, J.F., Paish, E.C., Saunders, P.T.K., et al. (2008). Nuclear and Cytoplasmic Expression of ERβ1, ERβ2, and ERβ5 Identifies Distinct Prognostic Outcome for Breast Cancer Patients. Clin. Cancer Res. 14, 5228–5235.

Shamanin, V.A., and Androphy, E.J. (2004). Immortalization of Human Mammary Epithelial Cells Is Associated with Inactivation of the p14ARF-p53 Pathway. Mol. Cell. Biol. 24, 2144–2152.

195

Shang, Y., and Brown, M. (2002). Molecular Determinants for the Tissue Specificity of SERMs. Science 295, 2465–2468.

Shim, W.-S., Conaway, M., Masamura, S., Yue, W., Wang, J.-P., Kumar, R., and Santen, R.J. (2000). Estradiol Hypersensitivity and Mitogen-Activated Protein Kinase Expression in Long-Term Estrogen Deprived Human Breast Cancer Cells in Vivo. Endocrinology 141, 396–405.

Shoker, B.S., Jarvis, C., Sibson, D.R., Walker, C., and Sloane, J.P. (1999a). Oestrogen receptor expression in the normal and pre‐cancerous breast. J. Pathol. 188, 237–244.

Shoker, B.S., Jarvis, C., Clarke, R.B., Anderson, E., Hewlett, J., Davies, M.P.A., Sibson, D.R., and Sloane, J.P. (1999b). Estrogen Receptor-Positive Proliferating Cells in the Normal and Precancerous Breast. Am. J. Pathol. 155, 1811–1815.

Shull, J.D., Spady, T.J., Snyder, M.C., Johansson, S.L., and Pennington, K.L. (1997). Ovary-intact, but not ovariectomized female ACI rats treated with 17β-estradiol rapidly develop mammary carcinoma. 7.

Shull, J.D., Pennington, K.L., Reindl, T.M., Snyder, M.C., Strecker, T.E., Spady, T.J., Tochacek, M., and McComb, R.D. (2001). Susceptibility to Estrogen-Induced Mammary Cancer Segregates as an Incompletely Dominant Phenotype in Reciprocal Crosses between the ACI and Copenhagen Rat Strains. Endocrinology 142, 5124–5130.

Shull, J.D., Dennison, K.L., Chack, A.C., and Trentham-Dietz, A. (2018). Rat models of 17β-estradiol-induced mammary cancer reveal novel insights into breast cancer etiology and prevention. Physiol. Genomics 50, 215–234.

Siegel, R.L., Miller, K.D., and Jemal, A. (2019). Cancer statistics, 2019. CA. Cancer J. Clin. 69, 7–34.

Singh, B., Bhat, N.K., and Bhat, H.K. (2011). Partial Inhibition of Estrogen-Induced Mammary Carcinogenesis in Rats by Tamoxifen: Balance between Oxidant Stress and Estrogen Responsiveness. PLoS ONE 6.

Singleton, D.W., and Khan, S.A. (2003). Xenoestrogen Exposure and Mechanisms of Endocrine Disruption. Front. Biosci. 8, s110-118.

Sivaraman, L. (1998). Hormone-induced refractoriness to mammary carcinogenesis in Wistar- Furth rats. Carcinogenesis 19, 1573–1581.

Skliris, G.P., Munot, K., Bell, S.M., Carder, P.J., Lane, S., Horgan, K., Lansdown, M.R., Parkes, A.T., Hanby, A.M., Markham, A.F., et al. (2003). Reduced expression of oestrogen receptor β in invasive breast cancer and its re-expression using DNA methyl transferase inhibitors in a cell line model. J. Pathol. 201, 213–220.

Smirnov, D.A., Morley, M., Shin, E., Spielman, R.S., and Cheung, V.G. (2009). Genetic analysis of radiation-induced changes in human gene expression. Nature 459, 587–591.

196

Smirnov, D.A., Brady, L., Halasa, K., Morley, M., Solomon, S., and Cheung, V.G. (2012). Genetic variation in radiation-induced cell death. Genome Res. 22, 332–339.

Smith, C.L., Nawaz, Z., and O’Malley, B.W. (1997). Coactivator and Coprepressor Regulation of the Agonist/Antagonist Activity of the Mixed Antiestrogen, 4- Hydroxytamoxifen. 11, 10.

Snijders, A.M., Marchetti, F., Bhatnagar, S., Duru, N., Han, J., Hu, Z., Mao, J.-H., Gray, J.W., and Wyrobek, A.J. (2012). Genetic Differences in Transcript Responses to Low- Dose Ionizing Radiation Identify Tissue Functions Associated with Breast Cancer Susceptibility. PLOS ONE 7, e45394.

Söderqvist, G., von Schoultz, B., Tani, E., and Skoog, L. (1993). Estrogen and progesterone receptor content in breast epithelial cells from healthy women during the menstrual cycle. Am. J. Obstet. Gynecol. 168, 874–879.

Sokol, E.S., Miller, D.H., Breggia, A., Spencer, K.C., Arendt, L.M., and Gupta, P.B. (2016). Growth of human breast tissues from patient cells in 3D hydrogel scaffolds. Breast Cancer Res. 18.

Song, R.X.-D., Mor, G., Naftolin, F., McPherson, R.A., Song, J., Zhang, Z., Yue, W., Wang, J., and Santen, R.J. (2001). Effect of Long-Term Estrogen Deprivation on Apoptotic Responses of Breast Cancer Cells to 17β-Estradiol. JNCI J. Natl. Cancer Inst. 93, 1714–1723.

Soni, M.G., Burdock, G.A., Taylor, S.L., and Greenberg, N.A. (2001). Safety assessment of propyl paraben: a review of the published literature. Food Chem. Toxicol. 39, 513– 532.

Sonnenschein, C., and Soto, A.M. (1998). An Updated Review of Environmental Estrogen and Androgen Mimics and Antagonists. J. Steroid Biochem. Mol. Biol. 65, 143–150.

Soule, H.D., and McGrath, C.M. (1986). A simplified method for passage and long-term growth of human mammary epithelial cells. In Vitro Cell. Dev. Biol. 22, 6–12.

Soule, H.D., Maloney, T.M., Wolman, S.R., Peterson, W.D., Brenz, R., McGrath, C.M., Russo, J., Pauley, R.J., Jones, R.F., and Brooks, S.C. (1990). Isolation and Characterization of a Spontaneously Immortalized Human Breast Epithelial Cell Line, MCF-10. Cancer Res. 50, 6075–6086.

Southey, M.C., Park, D.J., Nguyen-Dumont, T., Campbell, I., Thompson, E., Trainer, A.H., Chenevix-Trench, G., Simard, J., Dumont, M., Soucy, P., et al. (2013). COMPLEXO: identifying the missing heritability of breast cancer via next generation collaboration. Breast Cancer Res. BCR 15, 402.

197

Spady, T.J., Harvell, D.M.E., Snyder, M.C., Pennington, K.L., McComb, R.D., and Shull, J.D. (1998). Estrogen-induced tumorigenesis in the Copenhagen rat: disparate susceptibilities to development of prolactin-producing pituitary tumors and mammary carcinomas. Cancer Lett. 124, 95–103.

Speirs, V., Skliris, G.P., Burdall, S.E., and Carder, P.J. (2002). Distinct expression patterns of ERα and ERβ in normal human mammary gland. J. Clin. Pathol. 55, 371–374.

Stampfer, M.R., and Bartley, J.C. (1985). Induction of transformation and continuous cell lines from normal human mammary epithelial cells after exposure to benzo[a]pyrene. Proc. Natl. Acad. Sci. U. S. A. 82, 2394–2398.

Stauffer, S.R., Coletta, C.J., Tedesco, R., Nishiguchi, G., Carlson, K., Sun, J., Katzenellenbogen, B.S., and Katzenellenbogen, J.A. (2000). Pyrazole Ligands: Structure−Affinity/Activity Relationships and Estrogen Receptor-α-Selective Agonists. J. Med. Chem. 43, 4934–4947.

Stork, C.T., Bocek, M., Crossley, M.P., Sollier, J., Sanz, L.A., Chédin, F., Swigut, T., and Cimprich, K.A. (2016). Co-transcriptional R-loops are the main cause of estrogen- induced DNA damage. ELife 5.

Sugiura, H., Toyama, T., Hara, Y., Zhang, Z., Kobayashi, S., Fujii, Y., Iwase, H., and Yamashita, H. (2007). Expression of Estrogen Receptor β Wild-type and its Variant ERβcx/β2 is Correlated with Better Prognosis in Breast Cancer. Jpn. J. Clin. Oncol. 37, 820–828.

Sun, H., Miao, Z., Zhang, X., Chan, U.I., Su, S.M., Guo, S., Wong, C.K.H., Xu, X., and Deng, C.-X. (2018). Single-cell RNA-Seq reveals cell heterogeneity and hierarchy within mouse mammary epithelia. J. Biol. Chem. 293, 8315–8329.

Sun, X., Casbas-Hernandez, P., Bigelow, C., Makowski, L., Joseph Jerry, D., Smith Schneider, S., and Troester, M.A. (2012). Normal breast tissue of obese women is enriched for macrophage markers and macrophage-associated gene expression. Breast Cancer Res Treat 131, 1003–1012.

Suprynowicz, F.A., Upadhyay, G., Krawczyk, E., Kramer, S.C., Hebert, J.D., Liu, X., Yuan, H., Cheluvaraju, C., Clapp, P.W., Boucher, R.C., et al. (2012). Conditionally reprogrammed cells represent a stem-like state of adult epithelial cells. Proc. Natl. Acad. Sci. U. S. A. 109, 20035–20040.

Suprynowicz, F.A., Kamonjoh, C.M., Krawczyk, E., Agarwal, S., Wellstein, A., Agboke, F.A., Choudhury, S., Liu, X., and Schlegel, R. (2017). Conditional cell reprogramming involves non-canonical β-catenin activation and mTOR-mediated inactivation of Akt. PLoS ONE 12.

Tan, H., Zhong, Y., and Pan, Z. (2009). Autocrine regulation of cell proliferation by estrogen receptor-alpha in estrogen receptor-alpha-positive breast cancer cell lines. BMC Cancer 9, 31.

198

Tanos, T., Sflomos, G., Echeverria, P.C., Ayyanan, A., Gutierrez, M., Delaloye, J.-F., Raffoul, W., Fiche, M., Dougall, W., Schneider, P., et al. (2013). Progesterone/RANKL Is a Major Regulatory Axis in the Human Breast. Sci. Transl. Med. 5, 182ra55-182ra55.

Tharun, I.M., Nieto, L., Haase, C., Scheepstra, M., Balk, M., Möcklinghoff, S., Adriaens, W., Dames, S.A., and Brunsveld, L. (2015). Subtype-Specific Modulation of Estrogen Receptor–Coactivator Interaction by Phosphorylation. ACS Chem. Biol. 10, 475–484.

Thibodeaux, C.A., Liu, X., Disbrow, G.L., Zhang, Y., Rone, J.D., Haddad, B.R., and Schlegel, R. (2009). Immortalization and transformation of human mammary epithelial cells by a tumor-derived Myc mutant. Breast Cancer Res. Treat. 116, 281–294.

Thomas, T.J., Faaland, C.A., Adhikarakunnathu, S., Watkins, L.F., and Thomas, T. (1998). Induction of p21 (CIP1/WAF1/SID1) by estradiol in a breast epithelial cell line transfected with the recombinant estrogen receptor gene: Breast Cancer Res. Treat. 47, 181–193.

Thompson, D., and Easton, D. (2004). The Genetic Epidemiology of Breast Cancer Genes. J. Mammary Gland Biol. Neoplasia 9, 221–236.

Thordarson, G., Jin, E., Guzman, R.C., Swanson, S.M., Nandi, S., and Talamantes, F. (1995). Refractoriness to mammary tumorigenesis in parous rats: is it caused by persistent changes in the hormonal environment or permanent biochemical alterations in the mammary epithelia? Carcinogenesis 16, 2847–2853.

Tian, J.-M., Ran, B., Zhang, C.-L., Yan, D.-M., and LI, X.-H. (2018). Estrogen and progesterone promote breast cancer cell proliferation by inducing cyclin G1 expression. Braz. J. Med. Biol. Res. 51.

Travis, R.C., and Key, T.J. (2003). Oestrogen exposure and breast cancer risk. Breast Cancer Res. BCR 5, 239–247.

Travis, R.C., Reeves, G.K., Green, J., Bull, D., Tipper, S.J., Baker, K., Beral, V., Peto, R., Bell, J., Zelenika, D., et al. (2010). Gene–environment interactions in 7610 women with breast cancer: prospective evidence from the Million Women Study. Lancet 375, 2143–2151.

Tremblay, A., Tremblay, G.B., Labrie, F., and Giguère, V. (1999). Ligand-Independent Recruitment of SRC-1 to Estrogen Receptor β through Phosphorylation of Activation Function AF-1. Mol. Cell 3, 513–519.

Trichopoulos, D., MacMahon, B., and Cole, P. (1972). Menopause and breast cancer risk. J. Natl. Cancer Inst. 48, 605–613.

Trichopoulos, D., Hsieh, C.-C., Macmahon, B., Lln, T.-M., Lowe, C.R., Mirra, A.P., Ravnihar, B., Salber, E.J., Valaoras, V.G., and Yuasa, S. (1983). Age at any birth and breast cancer risk. Int. J. Cancer 31, 701–704.

199

Troester, M.A., Hoadley, K.A., Sørlie, T., Herbert, B.-S., Børresen-Dale, A.-L., Lønning, P.E., Shay, J.W., Kaufmann, W.K., and Perou, C.M. (2004). Cell-Type-Specific Responses to Chemotherapeutics in Breast Cancer. Cancer Res. 64, 4218–4226.

Troester, M.A., Lee, M.H., Carter, M., Fan, C., Cowan, D.W., Perez, E.R., Pirone, J.R., Perou, C.M., Jerry, D.J., and Schneider, S.S. (2009). Activation of Host Wound Responses in Breast Cancer Microenvironment. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 15, 7020–7028.

Umekita, Y., Souda, M., Ohi, Y., Rai, Y., Sagara, Y., Sagara, Y., and Yoshida, H. (2007). Expression of estrogen receptor alpha and progesterone receptor in normal human breast epithelium. In Vivo 21, 535–539.

Vanoosthuyse, V. (2018). Strengths and Weaknesses of the Current Strategies to Map and Characterize R-Loops. Noncoding RNA 4.

Vegeto, E., Cocciolo, M.G., Raspagliesi, F., Piffanelli, A., Fontanelli, R., and Maggi, A. (1990). Regulation of Progesterone Receptor Gene Expression. Cancer Res. 50, 5291– 5295.

Venkitaraman, A.R. (2001). Functions of BRCA1 and BRCA2 in the biological response to DNA damage. J. Cell Sci. 114, 3591–3598.

Vrtačnik, P., Ostanek, B., Mencej-Bedrač, S., and Marc, J. (2014). The many faces of estrogen signaling. Biochem. Medica 24, 329–342.

Wang, J., Xie, L.Y., Allan, S., Beach, D., and Hannon, G.J. (1998). Myc activates telomerase. Genes Dev. 12, 1769–1774.

Webb, P., Nguyen, P., Shinsako, J., Anderson, C., Feng, W., Nguyen, M.P., Chen, D., Huang, S.-M., Subramanian, S., McKinerney, E., et al. (1998). Estrogen Receptor Activation Function 1 Works by Binding p160 Coactivator Proteins. Mol. Endocrinol. 12, 1605–1618.

Williams, C., Edvardsson, K., Lewandowski, S.A., Ström, A., and Gustafsson, J.-Å. (2008). A genome-wide study of the repressive effects of estrogen receptor beta on estrogen receptor alpha signaling in breast cancer cells. Oncogene 27, 1019–1032.

Witkowska, H.E., Carlquist, M., Engström, O., Carlsson, B., Bonn, T., Gustafsson, J.-Å., and Shackleton, C.H.L. (1997). Characterization of bacterially expressed rat estrogen receptor β ligand binding domain by mass spectrometry: Structural comparison with estrogen receptor α. Steroids 62, 621–631.

Wolf, I., Bose, S., Williamson, E.A., Miller, C.W., Karlan, B.Y., and Koeffler, H.P. (2007). FOXA1: Growth inhibitor and a favorable prognostic factor in human breast cancer. Int. J. Cancer 120, 1013–1022.

200

Wu, C., Kraft, P., Zhai, K., Chang, J., Wang, Z., Li, Y., Hu, Z., He, Z., Jia, W., Abnet, C.C., et al. (2012a). Genome-wide association analyses of esophageal squamous cell carcinoma in Chinese identify multiple susceptibility loci and gene-environment interactions. Nat. Genet. 44, 1090–1097.

Wu, X., Subramaniam, M., Negron, V., Cicek, M., Reynolds, C., Lingle, W.L., Goetz, M.P., Ingle, J.N., Spelsberg, T.C., and Hawse, J.R. (2012b). Development, Characterization, and Applications of a Novel Estrogen Receptor Beta Monoclonal Antibody. J. Cell. Biochem. 113, 711–723.

Yao, K., Lee, E.-S., Bentrem, D.J., England, G., Schafer, J.I.M., O’Regan, R.M., and Jordan, V.C. (2000). Antitumor Action of Physiological Estradiol on Tamoxifen- stimulated Breast Tumors Grown in Athymic Mice. Clin. Cancer Res. 6, 2028–2036.

Yard, B.D., Adams, D.J., Chie, E.K., Tamayo, P., Battaglia, J.S., Gopal, P., Rogacki, K., Pearson, B.E., Phillips, J., Raymond, D.P., et al. (2016). A genetic basis for the variation in the vulnerability of cancer to DNA damage. Nat. Commun. 7, 1–14.

Yasar, P., Ayaz, G., User, S.D., Gupur, G., and Muyan, M. (2017). Molecular mechanism of estrogen-estrogen receptor signaling. Reprod Med Biol 16, 4–20.

Ye, X., Bishop Amber M., Reidy John A., Needham Larry L., and Calafat Antonia M. (2006). Parabens as Urinary Biomarkers of Exposure in Humans. Environ. Health Perspect. 114, 1843–1846.

Yi, P., Driscoll, M.D., Huang, J., Bhagat, S., Hilf, R., Bambara, R.A., and Muyan, M. (2002). The Effects of Estrogen-Responsive Element- and Ligand-Induced Structural Changes on the Recruitment of Cofactors and Transcriptional Responses by ERα and ERβ. Mol. Endocrinol. 16, 674–693.

Zajchowski, D.A., Sager, R., and Webster, L. (1993). Estrogen Inhibits the Growth of Estrogen Receptor-negative, but not Estrogen Receptor-positive, Human Mammary Epithelial Cells Expressing a Recombinant Estrogen Receptor. 9.

Zaret, K.S., and Carroll, J.S. (2011). Pioneer transcription factors: establishing competence for gene expression. Genes Dev. 25, 2227–2241.

Zhao, X., Malhotra, G.K., Lele, S.M., Lele, M.S., West, W.W., Eudy, J.D., Band, H., and Band, V. (2010). Telomerase-immortalized human mammary stem/progenitor cells with ability to self-renew and differentiate. Proc. Natl. Acad. Sci. 107, 14146–14151.

Zhuang, Y.H., Saaristo, R., and Ylikomi, T. (2003). An in vitro long-term culture model for normal human mammary gland: expression and regulation of steroid receptors. Cell Tissue Res 311, 217–226.

201